Bulletin of the British Museum (Natural History). British Museum (Natural History) Converted as part of the ABLE project by Dauvit King London : BM(NH) Continued as: Bulletin of the Natural History Museum. Entomology series Vol.1 (1950) - vol.61 (1992) 2 4 This document has been converted to TEI XML as part of the ABLE project to make it more widely available to biodiversity researchers in a useful format. eng text No corrections have been made in the text. The original source has not been regularized or normalized. Quotation marks have not been processed. They are as in the original DjVu XML document. Hyphens, including end-of-line hyphens, have not been processed. They are as in the original DjVu XML document. The text has been segmented based purely on layout based on page breaks. No language level segmetation, such as sentences, tone-units or graphemic, has been applied. Additional mark up using taXMLit has been applied to the TEI XML based on analysis of the original source through the uBio and OpenCalais web services. (Add comment for fuzzy matching once this has been brought into the final workflow too.) n i SEUM Bulletin of the British Museum (Natural History) Entomology series Vol49 1984 British Museum (Natural History)London 1984 Dates of publication of the parts Nol 28 June 1984 No 2 26 July 1984 No 3 30 August 1984 No 4 20 December 1984 No 5 20 December 1984 ISSN 0524-6431 Printed in Great Britain by Henry Ling Ltd, at the Dorset Press, Dorchester, Dorset ContentsEntomology Volume 49 No 1 Afrotropical jumping plant lice of the family Triozidae (Homoptera:Psylloidea)By David Hollis 1 No 2 The taxonomy of the western European grasshoppers of the genusEuchorthippus, with special reference to their songs (Orthoptera:Acrididae)By D. R. Ragge & W. J. Reynolds 103 No 3 An historical review of the higher classification of the Noctuidae(Lepidoptera)By Ian J. Kitching 153 No 4 The Pimplinae, Xoridinae, Acaenitinae and Lycorininae (Hymenop-tera: Ichneumonidae) of AustraliaBy I. D. Gauld .235 No 5 The Palaearctic species of Ascogaster (Hymenoptera: Braconidae) By T. Huddleston 341 > n i Bulletin of the British Museum (Natural BRITISH Mi;(NATURAL HISTORY! 29JUN1984 Afro tropical jumping plant liceof the family Triozidae(Homoptera: Psylloidea) David Hollis Entomology seriesVol 49 No 1 28 June 1984 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique andever-growing collections of the Museum, both by the scientific staff of the Museum and byspecialists from elsewhere who make use of the Museum's resources. Many of the papers areworks of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself,available separately, and individually priced. Volumes contain about 300 pages and severalvolumes may appear within a calendar year. Subscriptions may be placed for one or more ofthe series on either an Annual or Per Volume basis. Prices vary according to the contents ofthe individual parts. Orders and enquiries should be sent to: t \S H ^^Publications Sales fs'** -"* """^^^w. v'J^'Ck _ ... , ,UCA i ;c/v s^XJ*7 vw British Musei\ n i.l j Museum (Natural History),Cromwell Road, London SW75BD,England. WorlcTList abbreviation: Bull. Br. Mus. nat. Hist. (Ent.) Trustees of the British Museum (Natural History), 1984 The Entomology series is produced under the general editorship of the Keeper of Entomology: Laurence A. Mound Assistant Editor: W. Gerald Tremewan ISBN 565 06003 1ISSN 0524-6431 British Museum (Natural History)Cromwell RoadLondon SW7 5BD Entomology seriesVol49Nolppl-102 Issued 28 June 1984 Afrotropical jumping plant lice of the familTriozidae (Homoptera: Psylloidea) BRITISH MU: 2 9 JU, \I934 David Hollis Department of Entomology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Contents Synopsis 1 Introduction 2 Checklist , host plants and distribution of Afrotropical Triozidae 4 Material, methods and terminology 7 Acknowledgements 8 Triozidae Low 8 Characters and their polarity 9 Identification 18 Key to species of Afrotropical Triozidae 18 Triozamia Vondracek 22 Afrotrioza gen. n 25 Trichochermes Kirkaldy 26 Pauropsylla Riibsaamen 27 The willcocks /-group 29 The trigemma- group 31 The sep//ma-group 32 Ungrouped species 33 Trioza Foerster 35 The erytreae-group 36 The litseae-group 43 The anomalicornis-group 44 The neoboutonia-group 46 The hargreavesi-group 50 The obsoleta-group 51 The cockerelli-group 55 The etiennei-group 56 The glabea-group 59 Ungrouped species 60 Doubtful species 69 References 69 Index to host plants 101 Index 101 Synopsis The family Triozidae is defined within the context of the superfamily Psylloidea. A list is given of the generaincluded in the family, on a world basis, with type-species and a summary of numbers of species, theirdistribution and host plant data. The family-group name Pauropsyllinae is placed within the Triozidae,necessitating the transfer of Paurocephala back to the Aphalaridae and Microceropsylla and Pelmato-brachia to the Calophyidae, and seven new combinations. The genus-group name Sympauropsylla issynonymised with Pauropsylla. A checklist of Afrotropical taxa is given, with host plant and distributiondata, and a key is provided for the identification of the 69 recognisable species. Two of the eight previouslydescribed species are treated as doubtful because their types are lost or untraced and the species cannot berecognised from their original descriptions; Trioza eastopi Orian is synonymised with T. litseae Bordage;one new genus and 63 new species are described. Bull. Br. Mus. nat. Hist. (Ent.) 49 (1): 1-102 Issued 28 June 1984 2 DAVID HOLLIS Introduction Throughout the world, species of the family Triozidae display a wide range of body form andhost plant preferences. Currently 48 genera are recognised but many of these are poorly definedand have no clear diagnostic features. Members of the family may be recognised easily, in theadult stage, by the unique venation and structure of the forewing (pp. 8-10, and Figs 40, 43).Fifth instar larvae of most species are also characteristic as the head and body margins normallyhave a complete fringe of specialised wax-producing setae called sectasetae (Figs 280, 289, 315).Table 1 gives, for each genus of Triozidae, a summary of the type-species, numbers of speciesincluded, distribution and host plant preferences. Table 1 Genera of Triozidae with type-species, numbers of species, distribution and host plant data. Numbers ofspecies recorded in parenthesis under one zoogeographical region also occur in another region. For the purposes of thistable species previously included under the generic names Megatrioza, Heterotrioza and Smirnovla are here includedunder Trioza. (Heterotrioza Dobreanu & Manolache, 1962: 258; type-species Trioza obliqua Thomson. MegatriozaCrawford, 1915: 264; type-species M. armata Crawford. Smirnovia Klimaszewski, 1968: 13; type-species Triozafemoralis Foerster.) GENUS NUMBERSOF SPECIES HOST PLANTFAMILIES 6 -> AFROTROPICAL JUMPING PLANT LICE 4 DAVID HOLLIS Like most Psylloidea, triozid species have very restricted trophic requirements, at least intheir larval stages. They are associated with dicotyledonous plants, and the host range oftriozids, as a whole, is greater than any other family of psyllids. It is worthy of note that thefamily does not colonise the Fabaceae (= Leguminosae), a plant family greatly exploited by theArytaininae/Euphalerinae complex of the Psyllidae. Two Trioza species in New Zealand, T.color ata (Ferris & Klyver) and T. dacrydii Tuthill, are known to develop on Dacrydium, agymnosperm (Podocarpaceae); and one species in Hawaii, Megatrioza palmicola Crawford,develops on Pritchardia spp. (Palmae). One Palaearctic species, Trioza nigricornis Foerster, isapparently polyphagous and has been recorded from Brassica (Cruciferae), Daucus (Apiaceae),Solarium (Solanaceae) and Beta (Chenopodiaceae), but recent work (Hodkinson, 1981) sug-gests a species complex may be involved. Several other triozid species are associated with plants of economic importance. Trioza alacrisFlor causes leaf-rolling in bay laurel (Laurus nobilis, Lauraceae) and has been introduced tovarious parts of the world, along with its host plant. In the New World Paratrioza cockerelli(Sulc) causes 'psyllid yellows' on potato (Solanum tuberosum, Solanaceae); T. anceps Tuthill, T.magnoliae Ashmead and T. perseae Tuthill feed on avocado pear (Persea americana, Lauraceae)in Central and South America. In the Oriental Region Trioza vitiensis Kirkaldy causes gallformation on the leaves of Malay apple (Eugenia malaccensis, Myrtaceae), and T. cinnamomi(Boselli) larvae form pit-galls on the leaves of Cinnamomum spp. (Lauraceae). In North Africaand the eastern Mediterranean basin Pauropsylla willcocksi Dcbski causes pit-gall formation onthe leaves of fig trees (Ficus sycomorus, Moraceae). T. litseae Bordage develops on Litseaglutinosa (Lauraceae) and feeds as an adult on Vanilla planifolia (Orchidaceae) on Reunion andMauritius, and T. erytreae (Del Guercio) is a serious pest of citrus (Rutaceae) in eastern andsouthern Africa where it is the principal vector of 'citrus greening disease'. Because of its pest status, T. erytreae has received considerable attention from biologists.Information on its life-cycle, bionomics, control, and further references may be found in Catling(19690; 19696; 1969c; 1970; 1971; 1972; 1973); Moran (1968a; 19686); Moran& Blowers (1967);Moran & Brown (1973); and Moran & Buchan (1975). This species is of considerable taxonomicinterest as it is part of a complex of species, all of which are difficult to define morphologically,but which have discrete host plant preferences (see pp. 36-42). Five other triozid species are recorded from the Afrotropical Region and the present paperadds a further 63 hitherto undescribed species to this list. A species and host plant checklist isgiven below. Checklist, host plants and distribution of Afrotropical Triozidae SPECIES HOST PLANTS DISTRIBUTION family species Ficus sp. Angola SPECIES AFROTROPICAL JUMPING PLANT LICEHOST PLANTS DISTRIBUTION T. cape/ier/sp. n. T. capensissp. n. T. carvalhoi sp. n. T. catling! sp. n. T. chiangae sp. n. 7 . dinaba sp. n. 7 . eafra sp. n. T. erytreae (Del Guercio) AraliaceaeSolanaceae AraliaceaeMenispermaceae Uapacaceae AraliaceaeRutaceae Seemannaralia gerrardii(Seeman.) Vig. Lycium salinicolaVerdoorn; L.?tetrandrum Thunb. Cussonia angolensis Hiern;C. paniculata Ecklon.; C.spicata Thunb. & Zeyher Cissampelos torulosa E.Mey. ex Harv. & Sond.;Cissampelos sp.;Stephania abyssinica(Dill. & Rich.) Walp Uapaca nitida Muell. Arg. Not known Cussonia spicata Thunb. Clausena anisata (Willd.)Oliv. ; Citrus spp. ; Fagara South Africa (Natal) South Africa (CapeProvince and OrangeFree State) Kenya, Angola, SouthAfrica (Transvaal andNatal), Swaziland Kenya, Tanzania, SouthAfrica (Transvaal) Angola Ghana, NigeriaKenya, Tanzania Tropical and South Africa,Sao Tome, St Helena, SPECIES DISTRIBUTION AFROTROPICAL JUMPING PLANT LICE HOST PLANTS DISTRIBUTION family species T. nestasimara sp. n.T. pitkinisp. n. T. schroederisp. n. 7 . seranistama sp. n. T. similis Heslop-Harrison / . tangae sp. n.7 . tavandula sp. n.T. tenuissp. n. T. theronisp. n.T. thibae sp. n. 7 . tiliacora sp. n.7. tundavalaesp. n. T. usambarica sp. n./ . xylopia sp. n. Triozamia lamborni (Newstead) 7 . usambarensissp. n. T. vondraceki sp. n. Sapotaceae Sapindaceae Icacinaceae MenispermaceaeMyrtaceae AnnonaceaeMoraceae Moraceae Moraceae Not known Chrysophyllum viridifoliumWood & Franks or C.pruiniforme Engl. Not knownNot knownNot known Not knownNot known Haplocoelum foliolosum(Hiern) Bullock Not known Apodytes dimidiata E.Mey. Tiliacora sp. Syzygium benguellense(Welw. ex Hiern) Not knownXylopia sp. Antiaris toxicaria Lesch.subsp. africana (Engl.)C. C. Bergvar. africana Antiaris toxicaria Lesch.subsp. africana (Engl.)C. C. Bergvar.usambarensis (Engl.) C.C. Berg Antiaris toxicaria Lesch.subsp. africana (Engl.)C. C. Bergvar.? TanzaniaKenya Tanzania, ZimbabweTanzania South Africa (CapeProvince) Tanzania Angola Angola South Africa (Natal, CapeProvince) Kenya TanzaniaAngola TanzaniaTanzania Senegal, Ivory Coast,Ghana, Nigeria, Zaire,Tanzania Tanzania Central African Republic,Uganda Material, methods and terminology The bulk of the material studied is deposited in the British Museum (Natural History) (BMNH),with additional material from the Museum fur Naturkunde der Humboldt Universitat, Berlin(MNHU); Musee Royal de 1'Afrique Centrale, Tervuren (MRAC); and the [National Collec-tion of Insects] Plant Protection Research Institute, Pretoria (NCI). Holotypes and paratypes ofthe species treated below are deposited in BMNH, MNHU, NCI, Museum National d'HistoireNaturelle, Paris (MNHN); the National Museums of Kenya (NMK); South African Museum,Cape Town (SAM); and the National Museum of Natural History, Washington (USNM). 8 DAVID HOLLIS All measurements are quoted in millimetres and were taken from slide-mounted material.Reference points, within which measurements were taken of the various structures, are given inHollis (1976), with additional information in Fig. 43. The length of the antennal flagellum is asummation of the lengths of individual flagellomeres. Unless otherwise stated all figures were drawn from slide-mounted material; scale linesrepresent 0-1 mm unless labelled otherwise. On the forewing figures the outer limits of patchesof spinules are indicated by pecked lines, as is the claval suture. Each paramere figure shows theinner surface of the right paramere. The figures of the larvae show the dorsal surface outline, leftside only; details are shown of the marginal setae, representing 0.1 mm, of the head, forewingbud and caudal plate. The detail of the anal pore area is shown from a ventral view, right sideonly. Structural terminology follows that of Vondracek (1957) and Hollis (1976). Wing venationterminology is explained in Figs 40, 43. The mi cell value is an expression of shape calculated bydividing the length of M 1+2 by the distance between the apices of M 1+2 and A/ 3+4 . Similarly thecu\ cell value is the distance between the apices of Cw la and Cw lb divided by the length of Cw lb .The Rs-Cui a line is an imaginary line connecting the apices of Rs and Cw la . Acknowledgements I thank Dr L. A. Mound, BMNH, and Dr M. R. Wilson, Commonwealth Institute ofEntomology, London, for offering constructive criticism of early drafts of the manuscript. Thefollowing colleagues provided valuable loans and gifts of material: A. L. Capener, formerly ofNCI; Jose Passes de Carvalho, presently with Departamento de Entomologia, Esta?ao Agrono-mica Nacional, Portugal; J. Etienne, Institut Senegalais Recherches Agricoles, Ziguinchor; DrU. Gollner-Scheiding, MNHU; A. van Harten, presently with Projecto Luta Integrada, Praia,Cape Verde Is; and Dr G. Schmitz, MRAC. TRIOZIDAE Low Triozinae Low, 1879: 605, 609; Maskell, 1890: 164; Froggatt, 1901: 272; Crawford, 1914: 64; 1919: 184; Pflugfelder, 1941: 78;Tuthill, 1943: 545; 1944: 144; 1952: 97; Zimmermann, 1948: 14; Heslop-Harrison, 1958: 562; 1961: 529; Ramirez-Gomez, 1960: 15; Mathur, 1975: 299. Type-genus: Trioza Foerster.Triozina Low; Low, 1886: 163.Triozaria Low; Puton, 1886: 93.Triozidae Low; Edwards, 1896: 251; Vondracek, 1957: 297; Dobreanu & Manolache, 1962: 253; Loginova, 1964: 473; Klimaszewski, 1973: 231; 1975: 201; Bekker-Migdisova, 1973: 113.Siphonaleyrodinae Takahashi, 1932: 48. Type-genus: Siphonaleyrodes Takahashi. [Synonymised by Mound & Halsey, 1978: 252.] DIFFERENTIATING DIAGNOSIS. Forewing (Fig. 40) without a costal break; R^ unbranched and pterostigmaabsent; M+Cu stem absent or very short so that R+M+Cu stem branches into its component veins atapproximately one point; R s not fused to M stem at any point. COMMENTS. This suite of characters separates the Triozidae from all other psyllid groups and Iregard it as autapomorphic for the family. Furthermore I^elieve the morphology of the triozidforewing has evolved to produce a wing movement which is fundamentally different to that ofother psyllids: there being no nodal line, from the costal break through to the break on the hindmargin at the apex of the claval suture, for the wing to flex. Many other groups of psyllids have some, but not all, of the triozid wing features. The costalbreak is lost in the Macrohomotomini, the Carsidarini and many members of the Arytainini; inEpipsylla the pterostigma is lost; and in many members of the Calophyidae M+Cu stem isreduced. In the Central and South American genus Synoza the triozid wing condition is achievedin an entirely different way. Rs is fused to M stem along its proximal two-thirds, the costal breakis absent but C+Sc is only very weakly sclerotised in the break area. The venation of this genuscan be derived easily from that of some South East Asian members of the Homotominae towhich Synoza is most closely related. AFROTROPICAL JUMPING PLANT LICE 9 As yet I have been unable to find synapomorphies linking the Triozidae with any one othergroup of psyllids to the exclusion of all others . Therefore I am regarding the rest of the Psylloideaas a paraphyletic group with respect to the Triozidae. The systematic scope of the Triozidae, as considered here, is similar to that of most recentauthors but with the addition of the genera Pauropsylla and Leptynoptera. This is a relativelyradical change as all previous authors, apart from White (1980), regarded Pauropsylla as amember of the Pauropsyllinae (or Pauropsyllini) within either the Carsidaridae or the Aphalar-idae (seep. 27). Some attempt has been made by previous authors to provide a subfamilial or tribalclassification for the Triozidae. Vondracek (1963) placed Triozamia in the Bactericerinae butretained the subfamily within the Ciriacreminidae (sensu Heslop-Harrison, 1958; nee Hollis,1976). Klimaszewski (1964: fig. 7) divided the group into two subfamilies, the Bactericerinaeand the Triozinae. Bekker-Migdisova (1973) retained these two subfamilies and divided theBactericerinae into two tribes, the monotypic Triozamini (Triozamia) and the Bactericerini(Bactericera, Rhinopsylla, Levidea and one unnamed genus). She divided the Triozinae into theHemischizocraniini (Hemischizocranium) , the Paracomecini (Paracomeca) and the Eutriozini(Eutrioza) . Presumably she regarded the remaining 40 or so genera as members of a fourth tribe ,the Triozini. Kwon (1983), in a review of the Korean psyllid fauna, erected the monobasic tribesEpitriozini (Epitrioza) and Trichochermini (Trichochermes) , apparently without reference tothe remaining world fauna. Most of the 47 previously described triozid genera are poorly defined and often based oninadequate or primitive characters. Species have been added to various genera withoutcomparison with relevant type-species, thus stretching generic limits beyond usefulness andlogic. The resulting network of genera is impossible to unravel without an overall examination oforiginal material. No attempt is made here to 'fit' species into existing genera other than Trioza,only one new genus is erected and the bulk of the Afrotropical species are placed in Trioza on atemporary basis. Where species-groups are recognisable these are defined, and if relationshipswith the triozid fauna outside Africa are apparent then these are noted. The characters used to differentiate taxa in this paper are discussed, together with their likelypolarity. When other faunas are examined in the light of these, and probably other charactersnot considered here, we may be able to produce a reasonably natural and predictive higherclassification for the Triozidae and offer useful biogeographical comment on the group. Characters and their polarity The integument has a general covering of setae of varying density which can be of use indistinguishing between closely related species. Often the density of pilosity on the vertex andthoracic dorsum is different to that of the genal cones and legs. Illustrations of dense, sparse,long and short pilosity are given in Figs 35, 38, 39. The head, in its primitive condition, is adpressed to the thorax and not declined from thelongitudinal axis of the body (Figs 32, 33). The occipital margin is sharply defined, the mediansuture of the vertex is well developed and the median ocellus is visible from above. Frontal lobeson the vertex tend to be better developed in the primitive head condition. Genal cones arenormally well developed in the group and cover the frons. Generally, in psyllids, it is thoughtthat the absence of genal cones is~a primitive condition but their absence may be a reversal inTriozidae. The more derived head condition shows it declined towards 90 from the longitudinalaxis of the body (Fig. 34), with a rounded occipital margin. The vertex becomes roundeddownwards and lacks frontal lobes. In some species, especially in the genus Pauropsylla, themedian suture of the vertex is lost. The antenna (Figs 9-29) shows many useful characters. In the primitive condition it has ascape, pedicel and a flagellum which has eight flagellomeres, with a single rhinarium presentsubapically on each of flagellomeres 2, 4, 6 and 7. Several species have developed supplementaryrhinaria, particularly on the first flagellomere (Triozamia lamborni, Trioza theroni and the T.hargreavesi-group). I suspect this feature is not uniquely derived. The rhinaria are normally 10 DAVID HOLLIS disc-shaped, with a convex centre, but may have complicated associated sensilla (Figs 22-28).Flagellomere 8 bears two terminal setae, which are subequal in length in the primitive condition.Throughout the group there is a tendency for one of the pair to become shortened and truncateapically, and this condition reaches its greatest development in Trioza anomalicornis where oneof the setae is reduced to a flat, circular sense organ (Fig. 17) . The overall length of the flagellumcompared to head width is of use in distinguishing between species, and the relative lengths ofindividual flagellomeres can also be useful. Reduction in the number of flagellomeres is rare butdoes occur in some Pauropsylla species, with 3 and 5, which do not normally bear rhinaria, beinglost. The mouthparts are relatively standard throughout the group. The clypeus most commonlyhas a pair of setae but in some species, notably those displaying primitive character states,several setae are present (Figs 5,7). The ultimate rostral segment always bears a pair of basalsetae and one or more pairs of setae along its length. It is rarely devoid of setae supplementary tothe basal pair (Figs 6,8). The thorax, in the primitive state, is relatively flat, the pronotum is clearly visible from aboveand rounded down anteriorly behind the occiput (Figs 30, 31). The propleural suture is diagonal,with the episternum somewhat larger than the epimeron, but the latter is in contact with theposterolateral corner of the pronotum. The mesopraescutum is wider than long, with a broadlyarcuate anterior margin which is gently rounded down to the pronotum. In the derived state thepronotum becomes strongly rounded down behind the occiput and is scarcely visible fromabove. The episternum becomes greatly enlarged and its posterior margin expanded laterallyand curved forward behind the eye; the epimeron being reduced and displaced ventrally. Themesopraescutum is elongated and produced forward, with a strongly arcuate anterior marginwhich is strongly rounded or angled down to the pronotum (Figs 36, 37). The/orewmg displays a number of distinct trends. The general shape varies from being ovatewith a rounded apex to elongate oval with an angular apex, but the obovate condition inPauropsylla is probably secondarily derived. The membrane may be heavily spinuled (Fig. 241)through to being devoid of spinules apart from the radular areas. The latter are normally presentat the margin of the wing in cells mi , ra 2 and cu\ , although more primitively they also occur in r 2(Fig. 65). They may be diffuse and broad or sharply defined and narrow triangular (Figs 249,178). The wing veins are usually sparsely clothed with short setae (Fig. 126) but may bear longsetae (Figs 183, 254); the latter condition is not always associated with a dense pilosity on theintegument. In the most primitive form of venation (Fig. 268) Rs is long, M branches proximallyto the Rs-Cui a line (see p. 16) with a consequently high m\ cell value, Cu is short with aconsequently low cu l cell value, and the claval suture reaches the wing margin close to the distalapex of Cw lb . In the most derived condition (Fig. 126) Rs is short, M branches distal to theRs-Cui a line with a consequently low mi cell value, Cu is long with a consequently high cui cellvalue and the claval suture reaches the hind margin distant from the apex of Cw lb and moreproximal to the base of the wing. These trends are not always associated and there are manycombinations. The hindwing, in the primitive state, is two-thirds or more as long as the forewing. The costalmargin is more densely setose, with the setae distal to the costal break not clearly divided intotwo groups (Fig. 44). In the derived state the hindwing is much shorter than the forewing and, inthe Trioza obsoleta group, it can be completely reduced and scale-like (Figs 179, 184, 189).Reduction in length is congruent with the displacement of the apex of the claval suture of theforewing towards the base of the wing. The costal margin is sparsely setose, with the setae distalto the costal break clearly divided into two groups (Fig. 42). The/ore and mid legs offer few characters of taxonomic significance, although the Pauropsyllawillcocksi-group has the extraordinary development of a ventroapical spur on the fore coxa (Fig.45). One or other of the tarsal segments may be elongated. The hind leg bears several useful characters. The coxa usually bears a well-developedmeracanthus, although this is weak in Triozamia lamborni and in the Trioza etiennei-group, andthere is often an anterior lobe developed (Figs 46, 47). The latter structure has, in the past, been AFROTROPICAL JUMPING PLANT LICE 11 used to group species but the feature has probably evolved several times. The three ventral senseorgans on the femur, which are probably stretch receptors, are usually in a medial position (Fig.49) but in Triozamia lamborni they are in the more primitive basal position (Fig. 48). The tibiahas a basal spine or tubercle, or a raised area bearing one or more small spines. The presence of alarge basal spine has been used in generic diagnoses, but I consider this to be the primitive state.Apical tibial spurs (Figs 52, 53) are normally present and are divided into an outer group of onespur and an inner group of up to four spurs. In Triozamia lamborni the outer spur is lost and theinner group consists of up to five spurs. There is a tendency towards reduction of the number ofinner spurs and there can be some variation within a species, e.g. some species which normallyhave three inner spurs occasionally have only two on one or other leg. Few species have thesespurs completely lacking but again I think this is a feature which has arisen more than once.Spurs at the apex of the basitarsus, a common feature in the rest of the Psylloidea, are, with theexception ofAfrotrioza, absent in the family. The abdomen, in psyllids generally, has setae on all visible abdominal tergites but thiscondition occurs in very few triozids (four species in the Afrotropical Region). The usualcondition in this family is for setae to be present only on the first, or the first two, visible tergites,i.e. tergites 2 and 3 in males and 3 and 4 in females. Trichochermes insleyi, some of the Triozaobsoleta-group and T. guiera, have setae on all tergites except the pregenital; T. mimusops hassetae only on tergites 6 to 8; T. chiangae has setae on tergites 3 to 5; and the abdominal tergites ofT. afrosersalisia are apparently devoid of setae. This character is not directly associated withgeneral body pilosity because T. medleri, probably the most setose Afrotropical species, hasonly the first visible abdominal tergite setose. The male genitalia provides the most useful character suite for defining species. As in mostpsyllids, the proctiger is normally unsegmented (unipartite), but a bipartite proctiger occurs inTriozamia species and is almost developed in the Trioza etiennei-group . I consider a bipartiteproctiger to be the derived condition (Fig. 59). The unipartite proctiger can be a simple, almostcylindrical tube (Fig. 242) but more commonly the basal part is swollen and often bears lateralexpansions which sometimes have modified setae on their inner surfaces (Figs 76, 97, 135). Theshape of the paramere and its chaetotaxy is usually definitive for a species. The aedeagus iscomposed of two articulated segments, as in most psyllids, but the apical segment is subdividedinto two parts in Triozamia lamborni and in the Trioza etiennei-group (Figs 64, 200). The apicalpart of the aedeagus is usually definitive for each species, and in some cases is quite complex(Figs 88, 91, 136). The female genital complex is usually conoid, with a subacute posterior apex (Fig. 127). Theanal pore ring usually consists of a double ring of wax-producing cells and is rarely modified; inTriozamia species it is very convoluted (Fig. 62), in the litseae-group and some species of theneoboutonia-group of Trioza it is reduced to a single ring, and in Trioza tundavalae it isincomplete anteriorly and forms multiple rows posteriorly (Fig. 274) . The base of the proctiger ismuch less sclerotised than in other groups of psyllids, and is sometimes membraneous aroundthe anterior arch of the anal pore ring. The apical part of the proctiger is usually narrowed to asubacute apex but may be modified in various ways (Figs 138, 253, 274). The subgenital plate istriangular in profile and usually has a narrowly arcuate or subacute posterior margin. Occa-sionally the posterior margin is truncate or emarginate (Figs 89, 138). The lower valves of theovipositor sometimes have teeth or ridges developed (Figs 128, 138, 273) which can be useful inspecies diagnosis. 12 DAVID HOLLIS Figs 1-8 Afrotropical Triozidae, head and mouthparts structure. 1, Afrotrioza bersama, head, antero-dorsal view. 2, 7, 8, Trioza erytreae; 2, head, anterior view; 7, clypeus, lateral view; 8, ultimate rostralsegment. 3, Trichochermes insleyi, head, lateral view. 4, Pauropsylla trichaeta, head, anterior view. 5, 6,Triozamia lamborni; 5, clypeus, lateral view; 6, ultimate rostral segment, (fl - frontal lobe; gc - genalcone.) Figs 1-4 drawn from dry mounted specimens. AFROTROPICAL JUMPING PLANT LICE 13 13 15 Figs 9-18 Afrotropical Triozidae, antennal structure. 9, Trioza erytreae, flagellum. 10, T. medleri,flagellum. ll,Pauropsyllatrigemma, flagellum. 12, P. breviantennata, flagellum. 13, Trioza afrobsoleta,apical flagellomere. 14, T. neoboutonia, apical flagellomere. 15, T. kakamegae, apical flagellomere. 16,T. dinaba, apical flagellomere. 17, T. anomalicornls , apical flagellomere. 18, T. tangae, subapical andapical flagellomeres. 14 DAVID HOLLIS 27 \ } e^/ 24 28 tlJ gf 29 Figs 19-29 Afrotropical Triozidae, antennal structure. 19, Triozamia lamborni, 1st and 2nd flagello-meres. 20, T. vondraceki, 1st and 2nd flagellomeres. 21, T. usambarensis , 1st and 2nd flagellomeres. 22,Trioza hargreavesi, 1st flagellomere. 23, T. mirificornis, 1st flagellomere. 24, T. theroni, apical third of1st flagellomere. 25, T. dinaba, 2nd flagellomere. 26, T. pitkini, 2nd flagellomere. 27, T. capensis, 2ndflagellomere. 28, T. ghanaensis, 4th flagellomere. 29, Pauropsylla ngongae, subapical and apicalflagellomeres. AFROTROPICAL JUMPING PLANT LICE 15 39 Figs 30-39 Afrotropical Triozidae, head and thorax structure. 30, 31, Triozamia vondraceki; 30, headand thorax, dorsal view; 31, same, lateral view. 32, 33, Afrotrioza bersama; 32, head and thorax, dorsalview; 33, same, lateral view. 34, 35, Trioza erytreae; 34, head and thorax, lateral view; 35, mesopraescu-tum, dorsal view. 36, 37, T. tenuis; 36, head and thorax, dorsal view; 37, same, lateral view. 38, T.gonjae, mesopraescutum, dorsal view. 39, T. medleri, mesopraescutum, dorsal view. Figs 30-34, 36, 37drawn from dry mounted specimens. 16 DAVID HOLLIS Rs-Cu. Rs-Cu 1a 44 42 Figs 40-44 Afrotropical Triozidae, wing structure. 40-42, Trioza erytreae; 40, forewing, showing veinand cell nomenclature; 41, hindwing; 42, costal setae of hindwing. 43, T. tennis, forewing (a/b = mi cellvalue, c/d = cui cell value). 44, Afrotrioza bersama, costal setae of hindwing. (ra - radular areas, cs -claval suture.) AFROTROPICAL JUMPING PLANT LICE 17 48 r\r \\ 51 50 Figs 45-53 Afrotropical Triozidae, leg structure. 45, Pauropsylla trichaeta, fore coxa, anteroventralview. 46, 50, Trioza etiennei; 46, hind coxa, lateral view; 50, hind tibia, anterodorsal view. 47, T. gonjae,hind coxa, lateral view. 48, Triozamia lamborni, hind femur, posteroventral view. 49, 52, Triozaerytreae; 49, hind femur, posteroventral view; 52, apex of hind tibia, posteroventral view. 51, T. karroo,hind tibia, anterodorsal view. 53, T. capensis, apex of hind tibia, posterodorsal view. 18 DAVID HOLLIS Identification The key presented below is intended to facilitate identification of Afrotropical triozid species; itdoes not reflect phylogenetic relationships. The concept of the 'genus' Trioza, used here, isadmittedly artificial, hence Trioza species key out both before and after the other includedgenera. Undoubtedly the user will have difficulty identifying species in the erytreae- and anomalicor-ms-groups of Trioza (couplets 57-65) because of their morphological homogeneity. As T.erytreae is of economic importance a summary of characters used to identify species of theerytreae-group is given in Table 4. Apart from anomaticomis, which has distinctive antennalcharacters, the other four species in this group can be separated on the basis of the shape of themale paramere and the apical segment of the aedeagus (see quadruplet 61, p. 22). No reliablecharacters were found which would discriminate between females of this group. T. bussei and T. similis are excluded from the key because of insufficient data (see p. 69). Key to species of Afrotropical Triozidae 1 Forewing with extensive brown-pigmented area either as a speckled pattern over most of wing, or as an infuscation along the medial vein and extending over posterior half of wing (Figs 69, 216,245,249) 2 Forewing hyaline, if brown-pigmented areas are present they are restricted to the very base ofwing (Fig. 188), or along course of R and R l veins (Figs 169, 194, 233, 272), or at points whereveins meet wing margin (Fig. 202) 5 2 Forewing with brown speckled pattern over most of surface (Fig. 69); long and narrow, more than 3-5 times longer than wide ; Rs strongly curved towards M } before reaching margin Trichochermes insleyi Capener (p. 26) Forewing with brown infuscation along M stem and across posterior half of wing; broader, atmost slightly more than 3 times longer than wide; Rs continuously divergent from M stem(Figs 216, 245, 249) 3 3 Small species, forewing not more than 1-8 mm long; antennal flagellum short, about as long as head width; genal cones absent Trioza pitkini sp. n. (p. 59) Larger species, forewing more than 2 mm long; antennal flagellum more than 1-5 times longer than head width ; genal cones well developed 4 4 Hind tibia with well-developed apical spurs; cu cell value 1-07-1-40; $ proctiger blunt, rounded apically in profile Trioza laingi sp. n. (p. 64) Hind tibia without apical spurs but with a crown of stout setae; cu cell value 1-48-1-79; $proctiger acute, extended apically in profile Trioza karroo sp. n. (p. 63) 5 Forewing with a radular area present in cell r 2 as well as in cells m\, ra 2 and cu 2 , but if absent then apex of claval suture in close proximity to apex of Cu\ b (Figs 56, 65); C+Sc greatly thickened along its whole length and more than twice as thick as R stem 6 Forewing with radular area absent from cell r 2 and apex of claval suture distant from apex ofCw lb ; C+Sc not or hardly thickened and at most slightly thicker than R stem 9 6 Metabasitarsus with a single apical spur; antenna short, flagellum 0-91-0-98 times as long as head width, 1st flagellomere without rhinaria; aedeagus 2-segmented Afrotrioza bersama sp. n. (p. 25) Metabasitarsus without apical spurs; antennal flagellum 1-5-2-0 times longer than head width,1st flagellomere bearing several rhinaria and often thickened (Figs 19-21); aedeagus3-segmented ( Triozamia spp.) 7 7 First flagellomere not at all thickened and bearing only 4-8 rhinaria basally; $ proctiger relatively short, 0-97-1-01 times as long as head width Triozamia usambarensis sp. n. (p. 24) First flagellomere thickened and almost completely covered with rhinaria; $ proctiger longer, 1-2-1-54 times longer than head width 8 8 First flagellomere 1-39-1-47 times longer than 2nd flagellomere; apical aedeagal segment 0-52-0-56 times as long as medial aedeagal segment Triozamia vondraceki sp. n . (p. 24) First flagellomere 0-63-0-76 times as long as 2nd flagellomere; apical aedeagal segment longerthan medial aedeagal segment Triozamia lamborni (Newstead) (p. 23) 9 Vertex dish-like, evenly and roundly concave, not at all raised or interrupted along median suture, margin of 'dish' complete, sharp, finely serrate Trioza afrosersalisia sp. n. (p. 54) AFROTROPICAL JUMPING PLANT LICE 19 Vertex convex, flat, or at most weakly concave on either side of median suture which clearlydivides vertex into two halves; if median suture absent then forewing, at most, 2-3 timeslonger than wide 10 10 M branching proximal to Rs-Cu^ line 11 M branching distal to Rs-Cui a line 26 1 1 Hindwing very reduced and scale-like 12 Hindwing not less than one-third the length of forewing 14 12 Angle at which RI branches from Rs greater than 90 (Fig. 178); genal cones short; terminal setae of apical flagellomere subequal Trioza afrobsoleta sp. n. (p. 52) Angle at which RI branches from Rs 90 or less; genal cones long; apical flagellomere with onelong and one short terminal seta 13 13 mi cell value 1-72-1-74, cu\ cell value 2-9-3-0; hindwing with broad, almost truncate apex (Fig. 189) Trioza boxi sp. n. (p. 53) m\ cell value 1-91-2-08, cu\ cell value 1-96-2-63; hindwing with narrow, subacute apex (Fig. 184) Trioza gonjae sp. n. (p. 53) 14 Cu stem not more than half (0-45) the length of Cw lb 15 Cu stem not less than half (0-6) the length of C lb 17 15 Densely pubescent species; genal cones long, well developed; forewing rounded apically (Fig. 254) Trioza medleri sp. n. (p. 65) Weakly pubescent species; genae rounded, cones not developed; forewing subacute apically (Figs 237, 265) 16 16 Forewing more than 3 times longer than wide; 1st flagellomere about twice as long as 2nd; RI short , about half as long as R stem ; inner apical spurs of both hind tibiae together numbering5 or 6; paramere as in Fig. 266; dorsal valve of ovipositor smooth apically Trioza tennis sp. n. (p. 67) Forewing about 2-5 times longer than wide; 1st flagellomere about 3 times as long as 2nd; RIabout as long as R stem; inner apical spurs of both hind tibiae together numbering 4;paramere as in Fig. 239; dorsal valve of ovipositor serrate apically Trioza ghanaensis sp. n. (p. 62) 17 Antennal flagellum not less than 1-5 times longer than head width 18 Antennal flagellum not more than 1-3 times longer than head width 19 18 Distal segment of aedeagus with a large basal expansion (Fig. 222); cell r 2 of forewing with spinules apically (Fig. 219) Trioza glabea sp. n. (p. 59) Distal segment of aedeagus with a small basal expansion (Fig. 224) ; cell r 2 of forewing devoid ofspinules Trioza usambarica sp. n. (p. 60) 19 First flagellomere with at least one rhinarium apically and usually two or three Trioza theroni sp. n. (p. 67)First flagellomere without rhinaria 20 20 Larger species, forewing 2-75-6-00 mm long; antennal flagellum more than twice as long as head width 21 Smaller species, forewing less than 2 mm long; antennal flagellum about as long as head width 22 21 Forewing about 6 mm long, less than 2-5 times longer than wide; course of R stem and R] brown-pigmented (Fig. 272) Trioza tundavalae sp. n. (p. 68) Forewing about 3 mm long and about 2-8 times longer than wide and unpigmented (Fig. 256) Trioza schroederi sp. n. (p. 65) 22 Pronotum with a median and two lateral raised tubercles (Fig. 203) 23 Pronotum without raised tubercles 24 23 Antennal flagellum 0-7 times as long as head width; mi cell value 1-62, cu\ cell value 2-9; brown-pigmented areas present where veins reach wing margin (Fig. 202) Trioza messaratina sp. n. (p. 57) Antennal flagellum 1-06-1-23 times longer than head width; mi cell value 1-25-1-45, cu\ cellvalue 1 -07-1-38; forewing hyaline (Fig. 198) Trioza etiennei sp. n. (p. 56) 24 Wing veins with short setae, m { cell without spinules; antennal flagellum 1 1 1-1 -23 times longer than head width; cf paramere long and slender, with a pair of teeth apically (Fig. 215) Trioza camerounensis sp. n. (p. 58) Wing veins with long setae; mi cell with spinules apically; antennal flagellum not more than0-91 times as long as head width; cf paramere 'thumb-shaped' with a transverse ridge inapical third (Figs 210, 213) 25 25 Rhinaria on flagellomeres 2, 4, 6 and 7 each with a short bifid sensillum; cf paramere with a 20 DAVID HOLLIS transverse ridge one-third below apex (Fig. 210) Trioza seranistama sp. n. (p. 58) Rhinarium of flagellomere 2 with a long bifid sensillum; inner transverse ridge of cf paramereapical (Fig. 213) Trioza nestasimara sp. n. (p. 58) 26 Forewing with course of R stem and R\ marked with brown pigment (Figs 169, 194, 233) 27 Course of R stem and RI unpigmented 29 27 mi cell value more than 1-8; forewing about twice as long as hindwing; abdomen with setae on tergites 6-8 Trioza mimusops sp. n. (p. 54) mi cell value less than 1-3; forewing at most 1-8 times longer than hindwing; abdomen withsetae on first two visible tergites at most 28 28 Antennal flagellum about twice as long as head width; ultimate rostral segment shorter than apical flagellomere Trioza fuscivena sp. n . (p. 61) Antennal flagellum about as long as head width; ultimate rostral segment about twice as long asapical flagellomere Trioza nachingweae sp. n. (p. 49) 29 Angle of R\ and Rs branch obtuse (Fig. 178) Trioza afrobsoleta sp. n. (p. 52) Angle of RI and Rs branch 90 or less 30 30 First flagellomere thickened and bearing about 40 rhinaria 31 First flagellomere not or hardly thickened, devoid of rhinaria 32 31 Larger species, forewing length of cf 4-4-5-0, of 9 5-0-5-4; m\ cell value 1-32-1-6; 3rd flagellomere without rhinaria; lower valves of ovipositor saw-like apically Trioza hargreavesi sp. n. (p. 50)Smaller species, forewing length 3-2-3-9 in cf and 4-1-4-6 in $; m\ cell value 1-13-1-29; 3rd flagellomere bearing several rhinaria; lower valves of ovipositor smooth Trioza mirificornis sp. n. (p. 51) 32 Forewing membrane with spinules in addition to radular areas 33 Forewing membrane devoid of spinules apart from radular areas 35 33 Forewing short and broad, 2-25-2-57 times longer than wide, with rounded apex (Fig. 241); antennal flagellum about as long as head width; apex of hind tibia without spurs Trioza guiera sp. n. (p. 62) Forewing more than 2-5 times longer than wide, with subacute apex; antennal flagellum morethan 1-5 times longer than head width ; apex of hind tibia with well-developed spurs 34 34 Apex of hind tibia with two inner apical spurs; forewing 2-71-3-43 times longer than wide; antennal flagellum 1-94-2-53 times longer than head width; ultimate rostral segment 1-0-1-4times longer than apical flagellomere Trioza litseae Bordage (p. 43) and T. xylopia sp. n. (p. 43) Apex of hind tibia with three inner apical spurs; forewing 2-60-2-89 times longer than wide;antennal flagellum 1-62-1-97 times longer than head width; ultimate rostral segment 1-3-2-0times longer than apical flagellomere Trioza eafra sp. n. (p. 42) 35 Hind tibia shorter than width of head, apex expanded, basal spine well developed, outer apical spur enlarged and displaced to a subapical position (Fig. 53); male genitalia as in Figs 226-228 Trioza capensis sp. n. (p. 55) Hind tibia longer than width of head, apex not expanded, basal tubercles small and poorlydeveloped, outer apical spur not enlarged or displaced; male genitalia another shape 36 36 Forewing narrow, not less than 2-5 times longer than wide, with angular or subangular apex, m\ cell value greater than 1-1 37 Forewing broadly rounded apically (Figs 75, 80, 83, 84, 86), not more than 2-3 times longer than wide , m \ cell value not more than 1 -05 (Pauropsylla species) 39 37 Second flagellomere with a double rhinarium (Fig. 25) 38 Second flagellomere with a single rhinarium 51 38 Cu stem long, more than 2-8 times longer than C lh , branching from R+M+Cu proximally to R branch (Fig. 165); cf paramere and aedeagus as in Figs 166-168 Trioza dinaba sp. n. (p. 49) Cu stem not more than 2-6 times longer than Cu\ b , branching from R+M+Cu at or slightlydistal to R branch; cf paramere and aedeagus as in Figs 162-164 .... Trioza bamendae sp. n. (p. 48) 39 Fore trochanter with a well-developed ventroapical spur (Fig. 45) 40 Fore trochanter without a ventroapical spur 42 40 Clypeus with a pair of setae; ultimate rostral segment with a pair of short setae; 9 genital segment rounded triangular in profile, subgenital plate without a ventral transverse groove Pauropsylla tatrichea sp. n. (p. 30)Clypeus with several short setae, ultimate rostral segment with a pair of long setae and a pair of AFROTROPICAL JUMPING PLANT LICE 21 short setae; $ genital segment angular in profile, subgenital plate with a ventral transversegroove 41 41 In final instar larva sectasetae present on dorsum, marginal sectasetae widely spaced and of uneven length (Fig. 280) Pauropsylla willcocksi Dcbski (p. 29) Final instar larva without sectasetae on dorsum , marginal sectasetae closely spaced and of evenlength (Fig. 282) Pauropsylla trichaeta Pettey (p. 30) 42 Antenna with 6 flagellomeres (Fig. 12) Pauropsylla breviantennata sp. n. (p. 31) Antenna with 8 flagellomeres 43 43 Basitarsus of each leg 1-5-2-0 times longer than corresponding apical tarsal segment; median suture of vertex absent Pauropsylla longipes sp. n. (p. 33) Basitarsus of each leg short, hardly longer than corresponding apical segment; median sutureof vertex present although sometimes incomplete posteriorly 44 44 cf proctiger with greatly expanded lateral lobes (Fig. 97) ; $ abdomen with sternite 6 much less sclerotised than other segments and membranous ventrally Pauropsylla eastopi sp. n. (p. 33) Cf proctiger flask-shaped, without expanded lateral lobes; $ abdomen with evenly sclerotisedsternites 45 45 Antennal flagellomeres 4 and 6 each bearing two apical rhinaria and each of these with a bifid sensillum (Fig. 11) 46 Antennal flagellomeres 4 and 6 each with a single apical rhinarium which does not bear a bifidsensillum 47 46 Forewing broader, about 1-8 times longer than wide, veins bearing long setae, at least in proximal half of wing (Fig. 83) Pauropsylla ngongae sp. n. (p. 31) Forewing narrower, about 2-25 times longer than wide, veins bearing short setae (Fig. 80) Pauropsylla trigemma sp. n. (p. 31) 47 Cf aedeagus with a complex apical segment (Figs 88, 91); $ subgenital plate with a ventroapical notch (Figs 89, 92) 48 Cf aedeagus with a simple apical segment (Figs 94, 103, 106); 9 subgenital plate with asmoothly rounded or straight posterior margin 49 48 Cf paramere broadening towards apex and with a transverse inner apical ridge (Fig. 87); posterior margin of $ subgenital plate weakly notched and bordered with short setae (Fig. 89) Pauropsylla septima sp. n. (p. 32) Cf paramere narrowing towards apex which is twisted and abcurved (Fig. 90) ; posterior marginof $ subgenital plate deeply notched and bordered with long setae (Fig. 92) Pauropsylla proxima sp. n. (p. 32) 49 Smaller species, forewing less than 1 -9 mm long in cf and 2- 1 in $ ; hind tibia less than 0-57 mm long; median suture of vertex incomplete and evanescing towards occiput; cf paramere as in Fig. 105 ; $ genital segment subrectangular in profile Pauropsylla senegalensis sp. n. (p. 34) Larger species, forewing more than 2-4 mm long; hind tibia more than 0-58 mm long; mediansuture of vertex complete to occiput; Cf paramere as in Figs 93, 102; $ genital segmentrounded triangular in profile 50 50 Larger species, hind tibia more than 0-75 mm long; cf paramere clavate, with an inner apical transverse ridge (Fig. 93); posterior margin of $ subgenital plate truncate Pauropsylla angolensis sp. n. (p. 33) Smaller species, hind tibia not more than 0-7 mm long; cf paramere conical with an apical pairof spines (Fig. 102); posterior margin of 9 subgenital plate rounded Pauropsylla mistura sp. n. (p. 34) 51 Cu stem relatively short , not more than 1-5 times longer than Cu lb 52 Cu stem longer, not less than 1-6 times longer than Cw lb 53 52 Hind tibia with one outer and three inner apical spurs; ultimate rostral segment narrow, almost twice as long as wide; d" paramere with a long abcurved anteroapical process (Fig. 231) Trioza ficicola sp. n. (p. 60) Hind tibia with one outer and two inner apical spurs; ultimate rostral segment broad, not morethan 1-5 times longer than wide; cf paramere with a simple rounded and notched apex(Fig. 157) Trioza hartenisp. n. (p. 47) 53 Apical flagellomere twice as long as flagellomere 7 Trioza tangae sp. n. (p. 66) Apical flagellomere at most slightly longer than flagellomere 7 54 54 Flagellomeres bearing long subapical setae, apical flagellomere with one apical seta and a flat discoid sense organ (Fig. 17) Trioza anomalicornis sp. n. (p. 44) 22 DAVID HOLLIS Flagellomeres with short subapical setae, apical flagellomere with one long and one shortterminal seta 55 55 Second flagellomere 1-76-2-35 times longer than 3rd; inner apical spurs of both hind tibiae together numbering at most 4 56 Second flagellomere about as long as 3rd; inner apical spurs of both hind tibiae togethernumbering 6 or rarely 5 57 56 mi cell value 1-10-1-23, cu\ cell value 1-6-1-9 Trioza neoboutonia sp. n. (p. 46) mi cell value 1-40-1-65, cw t cell value 2-00-2-82 Trioza chiangaesp. n. (p. 48) 57 All visible abdominal tergites with a transverse row of setae ; setae on wing veins twice as long as width ofveins Trioza tiliacora sp. n. (p. 41) Transverse row of setae present only on first two visible abdominal tergites; setae on wing veinsshorter than width of veins 58 58 Ratio of head width to length of ultimate rostral segment 4-6: 1 or more 59 Ratio of head width to length of ultimate rostral segment 4-5:1 or less 60 59 Smaller species, forewing length less than 2-6 in cf and 2-9 in $; cf paramere broadening towards apex which is truncate (Fig. 118) Trioza kilimanjarica sp. n. (p. 40) Larger species, forewing length more than 3-0 in cf and 3-3 in $ ; paramere broad medially butnarrowing to subacute apex Trioza ata sp. n. (p. 40) 60 Ventral valves of ovipositor with strong transverse ridges, giving the valves a saw-like appearance (Fig. 138) ; paramere as in Figs 141 , 144, 147, 150 61 Ventral valves of ovipositor smooth or with a few weak serrations apically; paramere as in Figs 110, 112, 114, 123 62 61 Paramere as in Fig. 141 Trioza kakamegae sp. n. (p. 45) Paramere as in Fig. 144 Trioza thibae sp. n. (p. 45) Paramere as in Fig. 147 Trioza tavandula sp. n. (p. 46) Paramere as in Fig. 150 Trioza luvandata sp. n. (p. 46) 62 cui cell value not more than 2-4 in cf and 2-45 in 9; forewing, at most, 1-58 times longer than hindwing 63 cu\ cell value not less than 2-55 ; forewing, at least 1-59 times longer than hindwing 64 63 cf paramere conical, in profile narrowing towards apex which is abcurved (Fig. 123); cf proctiger broader than long due to strong lateral expansions (Fig. 122) Trioza carvalhoi sp. n. (p. 41) Cf paramere ovoid, in profile broadening medially then narrowing towards apex (as inFig. 110); Cf proctiger narrower with less well-developed lateral lobes (as in Fig. 109) Trioza capeneri sp. n. (p. 42) 64 cf paramere and aedeagus as in Figs 114, 115 Trioza gregoryi sp. n. (p. 40) Cf paramere and aedeagus as in Figs 1 10-1 13 65 65 Ratio of head width to length of 1st flagellomere not more than 1-25:1 ... Trioza catling! sp. n. (p. 39)Ratio of head width to length of 1st flagellomere not less than 1-26:1 Trioza erytreae (Del Guercio) (p. 36) and T. menispermicola sp. n. (p. 40) TRIOZAMIA Vondracek Triozamia Vondracek, 1963: 266. Type-species: Rhinopsylla lamborni Newstead, by original designationand monotypy. DESCRIPTION . Head , from above , slightly narrower than mesoscutum , in profile at 90 to longitudinal axis ofbody; occipital margin angular, occiput concave; vertex with a median suture, without frontal lobes;median ocellus not visible from above, frons just visible in anterior view; genae slightly swollen ventrally;1st flagellomere bearing several rhinaria, flagellomeres 4, 6 and 7 bearing a single subapical rhinarium;clypeus prominent, densely setose. Thorax, in profile, weakly arched; pronotum clearly visible from aboveand strongly rounded down anteriorly behind occiput; propleural suture diagonal but both pleurites incontact with lateral margin of pronotum; forewing elongate oval with a subangular apex, C+Sc stronglythickened, M branching proximally to Rs-Cu^ line, claval suture reaching hind margin of wing at samepoint as apex of C lb , radular spines present in cells r 2 (sometimes weak or absent), m\, m 2 and cu\;hindwing well developed, costal margin densely setose both proximally and immediately distal to the costalbreak. Hind coxa with a weakly developed meracanthus and without anterior lobe; hind femur notthickened medially, ventral sense organs basally positioned with the most distal organ slightly separatedfrom the proximal pair; hind tibia with well-developed basal spine, and an inner apical group of AFROTROPICAL JUMPING PLANT LICE 23 well-developed spurs. Abdomen with setae on all tergites; cf proctiger bipartite; aedeagus 3-segmented;anal pore of $ proctiger with a double, convoluted ring of wax-producing cells. COMMENTS. When Vondracek described this genus he placed it in the Bactericerinae of theCiriacreminidae (sensu Heslop-Harrison, 1958; nee Hollis, 1976). Klimaszewski (1964) rightlyremoved the Bactericerinae to the Triozidae, recognising the true relationships of the group,and Bekker-Migdisova (1973) separated Triozamia from the other Bactericerinae on a suite ofboth derived (bipartite cf proctiger, 3-segmented aedeagus) and primitive (proximity of apex ofclaval suture to apex of Cw lb , radulae area in r^ characters. The genus has retained manyprimitive features in addition to those mentioned by Bekker-Migdisova, notably the flattenedthorax and unmodified prosternum, the basal position of the ventral sense organs on the hindfemur and the presence of setae on all abdominal tergites. Her statement of a 'distinctsclerotisation of the pterostigma' is probably a misinterpretation of the thickening of C+Sc, atrue pterostigma being absent. Other derived features of the genus include the large number ofrhinaria on the first flagellomere and the usual absence of a rhinarium on flagellomere 2, butthese offer no indication of the relationships of the genus to the rest of the Triozidae.Three included species are treated below. Triozamia lamborni (Newstead)(Figs 5, 6, 19, 48, 63, 64, 275, 276) Rhlnopsylla lamborni Newstead, 1914: 520; Eastop, 1961: 168. Holotype cf , NIGERIA: 'Lagos, 70m. E. nr Oni clearing' [not traced]. Triozamia lamborni (Newstead); Vondracek, 1963: 266, partim; Akanbi, 1981: 113.Triozamia lambourni [sic] (Newstead); Roberts, 1969: 78. DESCRIPTION. Adult. Having generic characters stated above. Integument of thoracic dorsum sparselyshort-haired but mesoscutellum and metascutum more densely haired. Vertex pentagonal, steeply slopingdownwards, slightly concave; ultimate rostral segment 4-5 times longer than apical flagellomere, denselyhaired; antennal flagellum 1-71-2-04 times longer than head width, 1st flagellomere 0-63-0-76 times as longas 2nd flagellomere, thickened and with a large number of rhinaria, 2nd flagellomere without a rhinarium,flagellomeres 4, 6 and 7 each with a single subapical rhinarium, apical flagellomere with one moderatelylong and pointed seta and one shorter and truncate seta terminally. Mesopraescutum, from above,rounded rhomboidal, wider than long and with a broadly arcuate anterior margin; forewing 2-66-2-81times longer than wide, apart from radular areas there is a patch of spinules in c+sc, veins moderatelydensely setose, R s long, M branching proximal to /? s Cu la line, m\ cell value 4-9-6-0, cu\ cell ratio 0-4-0-5;costal setae of hindwing, distal to costal break, divided into two groups; hind tibia with an inner apicalgroup of 5-6 strongly developed spurs and an outer apical group of about 10 thickened setae; hindbasitarsus elongate. Abdomen with a pair of lateral eversible sacs on segment 4; cf proctiger and genitaliaas in Figs 59, 63, 64, apical segment of aedeagus 1-12-1-23 times longer than medial segment; $ genitalsegment with proctiger 1-20-1-27 times longer than head width, ovipositor valves smooth. Measurements (7 cf, 5 <j>). Maximum width of head, cf 1-23-1.30, <j> 1-30-1-35; length of antennalflagellum, cf 2-22-2-51, $ 2-31-2-51; length of ultimate rostral segment, cf 0-41-0-46, $ 0-45-0-50; lengthof forewing, cf 4-05-4-33, $ 4-21-4-58; length of hind tibia, cf 0-90-1-00, $ 0-95-1-05. Fifth instar larva (Figs 275-277). Body clearly divided into head, thorax and abdomen, about 1-1 timeslonger than wide. Antenna with 8 flagellomeres. Small prothoracic sclerites present behind cephalo-prothorax, meso- and metathoracic sclerites separated. Forewing bud about 0-92 mm long, humeral lobesnot at all developed. Caudal plate about half as long as wide and does not include first 4 abdominalsegments; anus terminal or terminodorsal, anus and pore ring as in Fig. 277, caudal plate with acomplicated arrangement of wax pores on dorsal and ventral surfaces (Fig. 276). No sectasetae present butcaudal plate bears 8 small dorsomarginal lanceolate setae positioned as in Fig. 276. HOST PLANT. Larvae and adults collected from Antiaris toxicaria africana var. africana(Moraceae). The larvae are free-living and produce copious wax strands. DISTRIBUTION. Senegal, Ivory Coast, Ghana, Nigeria, Zaire and Tanzania. COMMENTS. T. lamborni may be distinguished from the other two known species in the genus bythe characters given in Table 2. 24 DAVID HOLLIS Table 2 Characters for the separation of the three African species of Triozamia. The material upon which Vondracek based his redescription of this species was a mixed series,the Ugandan specimens and his figs 18 and 26 being vondraceki. The presence of three clearly recognisable species of Triozamia in Africa raises someinteresting questions on the taxonomy of the host plant genus Antiaris. In a recent review ofAfrican Moraceae Berg (1977) recognises one African species, Antiaris toxicaria, which alsooccurs in Asia and Australia. Berg refers the African populations to the subspecies africana andfurther recognises three varieties, africana, welwitschii and usambarensis, all with partiallyoverlapping distributions (Berg, 1977: fig. 8). Of the three Triozamia species on Antiaris, lamborni is known from Senegal, Ivory Coast,Ghana, Nigeria, northern Zaire and Tanzania (Pare Mountains and Uluguru Mountains);vondraceki is described from Uganda and the Central African Republic; and usambarensis isdescribed from Tanzania (Usambara Mountains). It seems likely that the three psyllid speciesmaintain their genetic independence each on one of three 'varieties' of Antiaris toxicariaafricana. Triozamia usambarensis sp. n. (Figs 21, 56-58) DESCRIPTION. Adult. Very similar to T. lamborni but slightly larger. First flagellomere 1-10-1-26 timeslonger than 2nd flagellomere, not thickened and bearing only 4-12 rhinaria on basal half, a subapicalrhinarium rarely present on 2nd flagellomere. Apical segment of aedeagus 0-96-0-97 times as long asmedial aedeagal segment, paramere as in Fig. 57; $ proctiger short, 0-97-1-01 times as long as head width. Measurements (2 cf , 2 $). Maximum width of head, cf 1-41-1-43, $ 1-48-1-51; length of antennalflagellum, cf 2-4, $ antennae incomplete; length of ultimate rostral segment, cf 0-37-0-39, 9 0-40; lengthof forewing, Cf 5-51-5-57, $ 6-10-6-12; length of hind tibia, cf 1-11 $ 1-17-1-20. Larva. Unknown. HOST PLANT. Adults collected from Antiaris toxicaria africana var. usambarensis (Moraceae). Holotype cf, Tanzania: Lushoto, 23. vi. 1977, Antiaris usambarensis (Wadudu Msituni) (BMNH; slidemounted). Paratypes. Tanzania: 1 cf, 1 $, same data as holotype; 1 cf, 1 ?, Lushoto, 22. ix. 1977, 'ex scales onHirus' [probably an error of transliteration from a handwritten label]; 1 cf, 1 $, Lushoto Silv. Arb.22. ix. 1980, Antiaris usambarensis (Wadudu Msituni) (BMNH; slide and dry mounted). COMMENTS (see also under T. lamborni). This species is described from seven rather poorlypreserved specimens, all from approximately the same locality. However, they were collectedon three separate occasions over a period of three years and the population is clearly maintainingits morphological identity from the other two species of the genus. The type-locality is close tothat of Antiaris usambarensis (see Berg, 1977: 318). Triozamia vondraceki sp. n. (Figs 20, 30, 31, 59-62) [Triozamia lamborni Newstead; Vondracek, 1963: 268, partim (figs 18, 26). Misidentification.]DESCRIPTION. Adult. Very similar to T. lamborni and T. usambarensis but larger. First flagellomere AFROTROPICAL JUMPING PLANT LICE 25 1-39-1-47 times longer than 2nd flagellomere, swollen and almost completely covered with rhinaria, 2ndflagellomere without a subapical rhinarium. Apical aedeagal segment short, 0-52-0-56 times as long asmedial aedeagal segment; paramere as in Fig. 60; 9 proctiger long, 1-52-1-54 times longer than headwidth. Measurements (3 cT, 2 $). Maximum width of head, cf 1-53-1-63, 9 1-81-1-83; length of antennalflagellum, cf 2-64-2-77, $ 2-87-2-96; length of ultimate rostral segment, cf 0-47-0-50, $ 0-52; length offorewing, Cf 5-69-6-16, $ 6-96-7-26; length of hind tibia, cf 1-26-1-38, $ 1-45-1-57. Larva. Unknown. HOST PLANT. Adults collected from Antiaris toxicaria africana var. ?. Holotype cf , Uganda: Kampala, 8.x. 1915 (C. C. Gowdey) (BMNH; dry mounted). Paratypes. Uganda: 1 Cf, 1 9, same data as holotype; 5 cf, 6 9, 17. xi. 1915 (C. C. Gowdey); 2 cf,20.ii.1923 (H. Hargreaves); 2 cf, 27. ii. 1925, 'Kirundo' (Antiaris africana) (J. L. R. Hanwer). CentralAfrican Republic: 1 cf , Route Mbale, P.L., 12.ii.1969 (M. Boulard) (BMNH; slide and dry mounted). COMMENT. See under T. lamborni. AFROTRIOZA gen. n. Type-species: Afrotrioza bersama sp. n. DESCRIPTION. Head, from above, narrower than mesoscutum, in profile continuing axis of anterior part ofthorax; occipital margin sharp; median suture of vertex moderately developed, frontal lobes welldeveloped; median ocellus visible from above, frons completely covered by genal cones; latter large andwell developed, constricted basally and in parallel plane to vertex; antenna with a single subapicalrhinarium on each of flagellomeres 2, 4, 6 and 7; clypeus not prominent, with several short setae. Thorax, inprofile, strongly arched; pronotum clearly visible from above and scarcely rounded down behind occiput,prothoracic suture diagonal with both pleurites in contact with lateral margin of pronotum; forewingelongate oval, narrowing to subangular apex, radular areas present in cells r 2 , m\, m 2 and cu { , C+Scstrongly thickened, M branching proximal to Rs Ci a line, claval suture reaching hind margin a shortdistance proximal to apex of Cw lb ; hindwing well developed, costal margin densely setose proximal tocostal break, costal setae distal to costal break not clearly grouped; hind coxa with a well-developedmeracanthus, anterior lobe absent; ventral sense organs of hind femur medially placed; hind tibia with asmall basal spine, with one outer and four inner apical spurs; hind basitarsus with an outer apical spur.Abdomen with setae on all tergites; cf proctiger unipartite, aedeagus 2-segmented. COMMENTS. This monotypic genus shows a mixture of both primitive and derived character statesand is clearly defined from all other described triozids by the presence of an apical spur on thehind basitarsus. The shape of the genal cones is similar to that of Trichochermes but the thorax isstrongly arched. It shares such primitive character states with Trlozamia as the presence of aradular area in r and an unmodified prothorax. Afrotrioza bersama sp. n. (Figs 1, 32, 33, 44, 65-68, 278, 279) DESCRIPTION. Adult. Having generic characters stated above. Large species, head width 0-91-0-97 mm.Integument of thoracic dorsum almost devoid of setae but genal cones, legs and abdomen densely setose.Vertex pentagonal, with a deep concavity on either side of median suture; antennal flagellum short,0-91-0-98 times as long as head width, apical flagellomere with one moderately long pointed seta and onevery short truncate seta terminally; ultimate rostral segment about twice as long as apical flagellomere,sparsely haired. Mesopraescutum, from above, rounded rhomboidal and almost as long as wide, with anarcuate anterior margin; forewing 2-27-2-40 times longer than wide, membrane moderately denselycovered with spinules, veins densely short haired, Rs long, m\ cell value 2-73-3-07, cu\ cell value 0-73-0-83;hind tibia with one outer and four inner apical spurs; hind basitarsus short, cf genitalia and proctiger as inFigs 66-68; 9 genital segment conical, proctiger slightly attenuate apically, anal pore with a double ring ofwax-producing cells, ovipositor valves smooth. Measurements (6 cf , 3 9). Maximum width of head, cf 0-91-0-95, 9 0-91-0-97; length of antennalflagellum, cf 0-86-0-91, 9 0-87-0-89; length of ultimate rostral segment, cf 0-16-0-17, 9 0-16-0-18; lengthof forewing, cf 4-13-4-37, 9 4-48^-64; length of hind tibia, cf 0-74-0-80, 9 0-76-0-78. 26 DAVID HOLLIS Fifth instar larva (Figs 278, 279). Dorsal surface outline broadly oval, about 1-1 times longer than wide.Antenna with two flagellomeres. Forewing pad 2-2 mm long; narrow prothoracic sclerites present betweencephaloprothorax and mesothorax, meso- and metathoracic sclerites as in Fig. 278; humeral lobe stronglyextended forward and reaching beyond anterior margin of cephaloprothorax. Caudal plate about 0-6 timesas long as wide; anus ventral, a short distance from posterior margin of abdomen, anus and pore ring as inFig. 279. Pointed sectasetae forming a dense fringe along anterior margin of cephaloprothorax and caudalplate, wing buds with a sparse fringe of very short simple setae; sectasetae absent from dorsum. HOST PLANT. Larvae and adults collected from Bersama sp. (Melianthaceae); the larvae causesevere leaf-rolling and distortion. Holotype d", Tanzania: E. Usambara mts, Amani Res. Sta. 19-27. vi. 1974, c. 3,000' Bersama sp. (D.Hollis) (BMNH; dry mounted). Paratypes. 56 cf, 33 9> larvae, same data as holotype. (BMNH; slide and dry mounted, and stored in80% ethanol). TRICHOCHERMES Kirkaldy Trichopsylla Thomson, 1877: 823. Type-species: Trioza walked Forster, by monotypy. [Homonym of Trichopsylla Kolenati, 1863.]Trichochermes Kirkaldy, 1904: 280; Kuwayama, 1910: 54; Loginova, 1964: 473. [Replacement name for Trichopsylla Thomson.] DESCRIPTION. Head, from above, narrower than mesoscutum, in profile in same plane as longitudinal axisof thorax; occipital margin sharp, occiput concave; frontal lobes of vertex well developed, median suturewell defined; median ocellus visible from above, frons completely covered by genal cones; latter welldeveloped, in parallel plane to vertex, clavate and constricted basally; antenna with a single subapicalrhinarium on each of flagellomeres 2, 4, 6 and 7. Thorax, in profile, very weakly arched; pronotum clearlyvisible from above and not rounded down behind occiput, prothoracic suture diagonal, episternum greatlyenlarged, epimeron reduced and not in contact with lateral margin of pronotum; forewing narrow elongateoval, with subacute apex, membrane with extensive brown pattern, C+Sc not thickened, Rs sinuous orarched strongly toward M stem, M branching distal to Rs Cw la line, claval suture reaching hind margin along distance proximal to apex of Cu ib ; hindwing well developed, costal setae separated into groups distalto costal break; hind femur with ventral sense organs medially placed, hind tibia with apical spurs separatedinto two groups, hind basitarsus without spurs. Abdomen with setae on tergites 2-7 in cf , and 3-7 in $ ; cfproctiger unipartite, aedeagus 2-segmented; anal pore of $ with a double ring of wax-producing cells. COMMENTS. Trichochermes may be distinguished from other triozids by the shape of the genalcones, the flattened thorax, the patterned forewing and the sinuous or displaced Rs vein. Thegenus is similar in appearance to Leuronota but the latter has simple genal cones and a relativelystraight Rs. Four Palaearctic species, all apparently developing on different species ofRhamnus, and oneAfrotropical species are recognised. It is doubtful if either of the two Japanese species,described by Kuwayama (1910), are true Trichochermes and Crawford (1919: 185) regardedthem both as Trioza species. From Kuwayama's original description it seems that Trichochermesbicolor Kuwayama, 1910: 54 is identical with Petalolyma basalts (Walker, 1858: 275), fromIndia, but formal synonymy cannot be established until Kuwayama's type-material has beenexamined. Trichochermes insleyi Capener(Figs 3, 69-71) Trichochermes insleyi Capener, 1973: 59. Holotype $, SOUTH AFRICA: 'Rustenburg, Tvl, 15 Sept. 1971'(NCI) [not examined]. DESCRIPTION. Adult. Having generic characters given above. Medium size, head width 0-50. Integument ofhead, body and legs sparsely covered with long setae. Vertex pentagonal with a well-developed concavityon either side of median suture; genal cones well developed; antennal flagellum 1-22 times longer thanhead width; clypeus not prominent, bearing two setae. Anterior margin of pronotum angular medially;mesopraescutum about as long as wide, with arcuate anterior margin. Forewing (Fig. 69) 3-72 times longer AFROTROPICAL JUMPING PLANT LICE 27 than wide, with mottled brown pigmentation over whole membrane, veins densely long-haired, Rs archedstrongly toward M, mi cell value 1-31, cu\ cell value 0-45; hindwing with five setae on C+Sc proximal tocostal break, setae distal to costal break divided into two groups; hind coxa with a well-developedmeracanthus and incipient anterior lobe; hind tibia with a well-developed basal spine, with one outer andtwo inner apical spurs, cf proctiger narrow, lateral margins not expanded, genitalia as in Figs 70, 71; 9genital segment short, conical, ventral valves of ovipositor serrate at extreme apices. Measurements (19). Maximum width of head, 0-5; length of antennal flagellum, 0-61; length of ultimaterostral segment, 0-11; length of forewing, 2-97; length of hind tibia, 0-56. Fifth instar larva. See Capener, 1973: 60, figs 122, 123. HOST PLANT. Larvae and adults collected from Ziziphus mucronata (Rhamnaceae). DISTRIBUTION AND MATERIAL EXAMINED. Known only from the type-series of which 1 cf and 2 9 paratypes(BMNH) have been examined. PAUROPSYLLA Riibsaamen Pauropsylla Riibsaamen, 1899: 262; Kieffer, 1905: 167; Crawford, 1915: 258, partim; Crawford, 1919: 142,partim; Enderlein, 1921: 115, partim; Loginova, 1972a: 839; Mathur, 1975: 72 partim. Type-species:Pauropsylla udei Riibsaamen, by monotypy. Sympauropsylla Enderlein, 1921: 116. Type-species: Pauropsylla triozoptera Crawford, by originaldesignation. Syn. n. DESCRIPTION. Head, from above, as wide as mesoscutum, in profile inclined almost at 90 to longitudinalaxis of body; vertex pentagonal, with sharp occipital margin medially, rounded down anteriorly to genae,median suture weak or absent, dorsal depressions well developed, lateral ocelli placed on outer surfaces ofraised tubercles, median ocellus visible from above; frons visible from anterior view; genae rounded,sometimes weakly expanded laterally below eyes; latter prominent, rounded; antenna short, with 8 or,more rarely, 6 flagellomeres. Thorax strongly arched dorsally; pronotum just visible from above, stronglyrounded down below anterior margin of mesopraescutum and behind occiput, propleural suture diagonalwith epimeron reduced and not in contact with lateral margin of pronotum; mesopraescutum, from above,with broadly arcuate anterior margin, in profile anterior margin sharply rounded down to pronotum;forewing obovate with narrow proximal part, not more than 2-3 times longer than wide, a very short M+Custem often present, M branches distally to Rs-Cu ia line, w t cell value less than 1, Cu stem always morethan twice as long as C lb , wing membrane devoid of spinules except for radular areas; hindwing at leasttwo-thirds as long as forewing; legs slightly elongate, meracanthus present on hind coxa, hind femurnarrow with ventral sense organs medial, hind tibia elongate with a group of small weak spines basally andone outer and two inner spurs apically (2 + 2 in P. udei), hind basitarsus without apical spurs. Abdomenwith setae on tergites 2 and 3 in cf , and 3 and 4 in 9 ; cf proctiger unipartite, mostly simple, flask-shaped,basal part rarely expanded; anal pore of 9 normally with a double ring of wax-producing cells. COMMENTS. Pauropsylla may be distinguished from other triozid genera by the shape andvenation of the forewing (Figs 75, 80, 83, 84, 86). The most recent and comprehensive review of psyllid systematics (Bekker-Migdisova, 1973)places Pauropsylla in the tribe Pauropsyllini, together with the Microceropsyllini, in thesubfamily Pauropsyllinae and in the family Carsidaridae. This largely follows Klimaszewski(1964) rather than Heslop-Harrison (1958) who placed the genus within his diffuse group theCiriacreminae (see Hollis, 1976). In a recent revision of the Pauropsyllini, Loginova (19720) concludes that the group is closelyrelated to the Phacopteronini and contains the genera Pauropsylla, Sympauropsylla, Leptynop-tera, Microceropsylla, Pelmatobrachia and Paurocephala. Including Paurocephala in this groupis consistent with Crawford's (1915; 1919) interpretation, but other workers (Vondracek, 1957;Heslop-Harrison, 1958; Klimaszewski, 1964) disagreed with this and placed Paurocephala in theAphalaridae. After examining several species of Paurocephala, including the type-species, P.psylloptera Crawford, the structure of the adult head, thorax and wing venation, and the 5thinstar larva (Table 3) suggest this genus should be placed in the Aphalaridae close to the generaHaplaphalara and Moraniella. Of the five remaining genera in Loginova's (1972a) Pauropsyllini,Microceropsylla and Pelmatobrachia (Bekker-Migdisova's 1973 Microceropsyllini) are transfer-red to the Calophyinae; Sympauropsylla is synonymised with Pauropsylla which, together with 28 DAVID HOLLIS Leptynoptera, is transferred to the Triozidae. Following Klimaszewski (1964) the Phacoptero-nini is placed in the Aphalaridae. Mathur (1975), in his account of Indian psyllids, reviews the Pauropsyllinae (auctt.) andincludes four genera, Apsylla, Paurocephala, Pauropsylla and Phacopteron. Apsylla was placedin the Anomalopsyllini by Vondracek (1963) and into the Aphalaridae by Klimaszewski (1964).Of the species Mathur included in Paurocephala only minuta Crawford, psylloptera Crawford,phalaki Mathur and russellae Mathur should remain in that genus. P. menoni Mathur andtrimaculata Mathur are here transferred to Haplaphalara as H. menoni (Mathur, 1975) comb, n.,and H. trimaculata (Mathur, 1975) comb. n. Of the 14 species Mathur placed in Pauropsylla,depressa Crawford,/zc/co/fl Kieffer, globuli Kieffer , purpurescens Mathur and reticulata Mathurshould remain. P. brevicornis (Crawford) and nigra (Crawford) were placed in Microceropsyllaby Boselli (1930a); longispiculata Mathur, maculata Mathur and verrucosa Mathur are heretransferred to Microceropsylla as M. longispiculata (Mathur, 1975) comb. n. M. maculata(Mathur, 1975) comb. n. and M. verrucosa (Mathur, 1975) comb. n. P. spondiasae (Crawford)was placed in Pelmatobrachia by Enderlein (1921), and tuberculata (Crawford) was placed inPseudophacopteron, also by Enderlein (1921). P. beesoni Laing is here transferred to Trioza asT. beesoni (Laing, 1930) comb. n. P. stevensi Laing is here transferred to the genus Diceraeopsyl-la as D. stevensi (Laing, 1930) comb, n., and the genus is referred to the Aphalaridae. The description and figures of Pauropsylla shiwapuriensis Miyatake (1981: 53) show that thisspecies is wrongly assigned and, more likely, belongs in the Aphalaridae near the Haplaphalara/Diclidophlebia-group . Most Pauropsylla species are recorded from species of Ficus (Moraceae) although udei, thetype-species, was described from a member of the Rubiaceae. Uichanco (1921: 265) regards thishost record as a misidentification but Mathur (1975: 105) has described reticulata fromAnthocephalus indicus (Rubiaceae). I have examined specimens from the same sample as thetype-series of reticulata and suspect that this species is synonymous with udei. Further material,including larvae, from Anthocephalus indicus will be needed before this synonymy and hostplant record can be confirmed. Eleven described species are regarded here as congeneric in Pauropsylla, mainly from the Table 3 Comparison of morphological features and host plant preferences of Paurocephala species andPauropsylla species. Paurocephala Pauropsylla Adults In profile, head inserted at orbelow anterior margin ofpronotum. Pronotum strongly rounded downbehind occiput so that, in profile,it is below dorsal margin of head. From above, dorsum of thoraxbroad. From above, dorsum of thoraxnarrow. Forewing with costal break,pterostigma clearly defined,long M+Cu stem present, apexof claval suture adjacent toapex of Cw lb . Forewing without costal breakand pterostigma, M+Cu stem veryshort or absent, apex of clavalsuture distant from apex of Cw lb . Larvae Free-living; dorsal sclerites of thorax separate; sectasetaepresent on antenna. Gall-forming; dorsal thoracicsclerites fused to form a singleplate (except in proximo sp.n.)sectasetae absent from antenna. Host Malvales and Urticales (Moraceae, plants Ulmaceae, Tiliaceae, Malvaceae and Sterculiaceae). Ficus spp. (Moraceae), and possiblyRubiaceae. AFROTROPICAL JUMPING PLANT LICE 29 Oriental region, but two species of doubtful validity, nussex Carmin and biki Carmin, occur inthe Palaearctic Region. P. willcocksi and trichaeta occur in Africa and a further elevenAfrotropical species are described below. They may be identified using couplets 39-50 in the key(p. 20). The willcocksi-group Defined by the presence of a strong ventroapical spur on the fore coxa. Three species with anAfrotropical distribution belong in this group, willcocksi, trichaeta and tatrichea. Two poorlydescribed species from Palestine also belong here (see comment under willcocksi). Pauropsylla willcocksi Dcbski(Figs 72-74, 280, 281) Pauropsylla willcocksi Dcbski, 1918 [71923]: 14; Willcocks, 1922: 275; Samy, 1972: 458. Syntypes, cfs, $s, larvae, EGYPT [not traced].[Pauropsylla trichaeta Pettey; Awadallah & Swailem, 1971: 193;? Loginova, 1972o: figs 40, 41. Misiden- tifications.] DESCRIPTION. Adult. Having generic characters given above. Integument shiny, sparsely covered with veryshort setae. Median suture of vertex absent; antenna with eight flagellomeres, flagellum 1-34-1-61 timeslonger than head width, apical flagellomere with one long and one moderately long and truncate terminalseta; clypeus bearing several hairs, ultimate rostral segment bearing a long basal pair and a short subapicalpair of setae. Forewing 1-92-2-06 times longer than wide, veins very sparsely clothed with very short setae,mi cell value 0-79-0-99, cu\ cell value 1-15-1-69; costal margin of hindwing with 1-2 setae proximal tocostal break, setae distal to costal break clearly divided into two groups; forecoxa with a well-developedventroapical spur; tarsal segments subequal. cf proctiger flask-shaped with weak basal expansions, theinner surface of each expansion bearing up to 22 thickened peg-like setae (Fig. 72) ; paramere and aedeagusas in Figs 73, 77 (trichaeta); $ genital segment short, triangular in profile, subgenital plate with a ventraltransverse groove in apical third. Measurements (10 cf , 10 $). Maximum width of head, cf 0-44-0-51, $ 0-46-0-57; length of antennalflagellum, cf and $, 0-65-0-80; length of ultimate rostral segment, cf 0-07-0-10, $ 0-09-0-11; length offorewing, cf 1-75-2-15, $ 2-07-2-55; length of hind tibia, cf 0-51-0-67, $ 0-56-0-71. Fifth instar larva (Figs 280, 281). Dorsal surface outline broadly oval, about 1-3 times longer than wide.Antennal flagellum unsegmented. Cephalothorax entire. Forewing pad about 0-47 mm long, humeral lobeextending forward but not as far as anterior margin of eye. Caudal plate about 0-6 times as long as wide,anus ventral and distant from posterior margin of abdomen, anus and pore ring as in Fig. 281. Marginalfringe consisting of moderately dense truncate sectasetae of varying lengths; postocular truncate sectasetaepresent; dorsum sparsely covered with truncate sectasetae. HOST PLANTS. Adults and larvae collected from Ficus ?gnaphalocarpa and Ficus sp. in Senegal,and Ficus sycomorus in Egypt (Moraceae). The larvae live in pit-galls on the undersides of theleaves. DISTRIBUTION. Material has been examined from Cape Verde Is., Senegal, Egypt, Sudan andSaudi Arabia. COMMENTS. Pauropsylla willcocksi is very close to P. trichaeta. Adults of the two species appearto be almost indistinguishable, but males of trichaeta tend to have more peg-like setae on theinner surface of the lateral expansions of the proctiger (25-30 on each side as opposed to amaximum of 22 on each side in willcocksi). However, the larvae of the two species appear to bequite distinct (see Figs 280, 282). Carmin (1951: 1-3) described two species, P. nussex and P. bikii, from Palestine, but thedescriptions are not diagnostic from one another or from willcocksi and the type-material isapparently lost. In the BMNH collection are several adult specimens labelled 'Palestine, Drs. D.Scheinkin and J. Carmin, on Ficus sycamorus\ some bear the number S.108, others S.109. Ifthese specimens are not part of the original syntypic series of nussex and bikii they certainlyrepresent Carmin's two species. This material is indistinguishable from willcocksi. Recently 30 DAVID HOLLIS collected Pauropsylla specimens from Israel (Bet Dagan) on Ficus sycomorus, consisting ofadults and larvae, do not resolve the problem. The adults are not distinguishable from willcocksibut the larvae are distinct from both willcocksi and trichaeta. They have the narrower dorsalsurface outline and uneven marginal fringe of willcocksi, but lack sectasetae on the dorsalsurface, as in trichaeta. Thus we have a series of populations of Pauropsylla species from South Africa, throughoutAfrica and extending into the Middle East. The adults of these populations are morphologicallyvery similar but three distinct forms of larvae can be identified, and names are available for thesethree larval types. I have therefore decided to name those populations from Africa, south of theSahara, on Ficus sur (= F. capensis) and Ficus spp. as P. trichaeta; those populations fromSenegal, North Africa, Saudi Arabia and Cape Verde Is. on Ficus sycomorus and F. gnaphalo-carpa (H. C. D. de Wit in a personal communication, suggests these two species may besynonymous) as P. willcocksi; and the Palestinian populations on F. sycomorus, as either P.nussex or bikii or both. Pauropsylla trichaeta Pettey (Figs 4, 45, 75-78, 282) Pauropsylla trichaeta Pettey, 1924: 29; 1925: 137; Capener, 1970: 199. Holotype (? sex), SOUTH AFRICA:Tanzeen, Transvaal, on a native fig tree, associated with galls' (SAM) [not examined]. DESCRIPTION. Adult. Similar in most features to P. willcocksi. Differs in that 25-30 peg-like setae arepresent on inner surface of each lateral expansion of the cf proctiger (Fig. 76). Measurements (10 d", 10 <j>). Maximum width of head, cf 0-40-0-51, $ 0-44-0-56; length of antennalflagellum, cf 0-67-0-80, $ 0-63-0-80; length of ultimate rostral segment, cf 0-07-0-09, $ 0-08-0-11; lengthof forewing, cf 1-83-2-31, $ 2-12-2-77; length of hind tibia, cf 0-50-0-67, $ 0-53-0-69. Fifth instar larva (Fig. 282). Dorsal surface outline almost circular, about 1-2 times longer than wide.Antennal flagellum unsegmented. Cephalothorax entire. Forewing pad 0-64 mm long, humeral lobeextending forward but not reaching anterior margin of eye. Caudal plate about half as long as wide, anusventral and distant from posterior margin of abdomen, anus and pore ring as in willcocksi (Fig. 281).Truncate tubular sectasetae forming an even dense marginal fringe, postocular tubular sectaseta present,sectasetae absent from dorsum. HOST PLANTS. Adults and larvae collected from Ficus sur (= F. capensis) , Ficus spp. (Moraceae).The larvae from pit-galls on the undersides of the leaves. DISTRIBUTION. Material has been examined from South Africa, Mozambique, Zimbabwe,Angola, Tanzania, Kenya, Uganda, Sudan, Cameroun, Nigeria, Ghana, Ivory Coast, SierraLeone and Senegal. COMMENT. See under P. willcocksi. Pauropsylla tatrichea sp. n. (Fig. 79) DESCRIPTION. Adult. Very similar to P. willcocksi but larger. Glypeus with a pair of setae, ultimate rostralsegment with a pair of small setae subapically. Costal margin of hindwing with 3-4 setae proximal to costalbreak. $ genital segment (Fig. 79) rounded triangular in profile, subgenital plate without ventraltransverse groove, cf unknown. Measurements (3 $). Maximum width of head, $ 0-55-0-57; length of antennal flagellum, $ 0-79-0-88;length of ultimate rostral segment, $ 0-09; length of forewing, $ 2-77-3-08; length of hind tibia, $0-65-0-70. Larva and host plant unknown. Holotype $, Cameroun: Bamenda, 6.H.1957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: 1 $, same data as holotype. Ivory Coast: 1 $, 24. iv. 1969 (A. Pollet). (BMNH;slide mounted.) AFROTROPICAL JUMPING PLANT LICE 31 The fr/gemma-group Defined by having highly modified antennae which are reduced in length, some of theflagellomeres bearing double rhinaria with bifid sensilla; anal pore of 9 with an incomplete outerring of wax-producing cells. Three species are included: trigemma, ngongae and breviantennata. Pauropsylla trigemma sp. n. (Figs 11, 80-82) DESCRIPTION. Adult. Integument with a dense covering of short setae. Median suture of vertex present andcomplete; antennal flagellum (Fig. 11) very short, 0-84-0-88 times as long as head width, flagellomere 2with a single rhinarium bearing a very short bifid sensillum, flagellomeres 4 and 6 each with a pair ofrhinaria and each of these bears a long bifid sensillum, flagellomere 7 with a single rhinarium which bears abifid sensillum of moderate length, both terminal setae on apical flagellomere long; clypeus with a pair ofsetae, ultimate rostral segment without setae. Forewing (Fig. 80) 2-24-2-26 times longer than wide, veinsbearing short setae, m v cell value 0-94-1-04; fore coxa without a ventroapical spur; tarsal segmentssubequal. cf proctiger narrow tubular, without lateral expansions or thickened setae, paramere andaedeagus as in Figs 81, 82; 9 unknown. Measurements (2 cf). Maximum width of head, 0-48-0-55; length of antennal flagellum, 0-42-0-46;length of ultimate rostral segment, 0-06-0-08; length of forewing, 1-82-1-91; length of hind tibia,0-32-0-39. Larva and host plant unknown. Holotype cf , Angola: Bruco, 26.ii.-2.iii. 1972, at light (D. Hollis) (BMNH; slide mounted). Paratype. Tanzania: 1 cf , Kilimanjaro, Bismark Hut, 2,500-3,000 m, S. Mawenzi, at foot of highpasture, ii.1912 (Chr. Schroder] (MNHU; slide mounted). COMMENTS. P. trigemma and ngongae are regarded as sister-species. They may beseparated from one another by the shape of the forewing and the chaetotaxy of the forewingveins. Both species form the sister-group of breviantennata. Pauropsylla ngongae sp. n. (Figs 29, 83) DESCRIPTION. Adult. Very similar to trigemma. Integument covered with long setae. Antennal flagellum0-77 times as long as head width, flagellomere 7 (Fig. 29) with a supplementary long conical sensillumadjacent to rhinarium. Forewing (Fig. 83) 1-79 times longer than wide, veins bearing long setae in proximaltwo-thirds of wing, cf unknown. 9 genital segment short, triangular in profile; subgenital plate with atruncate posterior margin which bears long setae, inner fold connecting ventral ovipositor valve sclerotisedand extended as a triangular projection beyond posterior margin of subgenital plate. Measurements (1 9)- Maximum width of head, 0-48; length of antennal flagellum, 0-37; length ofultimate rostral segment, 0-05; length of forewing, 1-75; length of hind tibia, 0-46. Larva and host plant unknown. Holotype $, Kenya: Nairobi, June. Magadi-Langata Rd, c. 5,400', 26.vii.1974 (D. Hollis) (BMNH; drymounted).Paratypes. 2 9 , same data as holotype (BMNH; slide and dry mounted). COMMENT. See under P. trigemma. Pauropsylla breviantennata sp. n. (Figs 12, 84, 85) DESCRIPTION. Adult. Similar to P. trigemma. Integument covered with short setae. Median suture of vertexweak and incomplete towards occipital margin; antennal flagellum (Fig. 12) 6-segmented, 0-76-0-84, timesas long as head width, two rhinaria present apically on flagellomeres 2, 3 and 4, a single rhinarium presenton flagellomere 5, rhinaria of flagellomeres 2, 3 and 5 bear large bulbous sensillae, rhinaria of flagellomere4 bear bifid sensillae, rhinarium of flagellomere 5 with an associated conical sensillum, terminal setae ofapical flagellomere of equal length, very long, about 0-75 times as long as whole flagellum. cf proctiger, 32 DAVID HOLLIS paramere and aedeagus as in Fig. 85; 9 genital segment damaged in specimen available, posterior marginsof both proctiger and subgenital plate densely clothed with long setae. Measurements (1 d", 1 $). Maximum width of head, cf 0-37, 9 0-42; length of antennal flagellum , cf0-31, $0-32; length of ultimate rostral segment, $0-04; length of forewing, cf 1-13, 9 1-32; length of hindtibia, cf 0-36, $ 0-34. Larva and host plant unknown. Holotype, cf Ghana: Tafo, 29.V.1957 (V. F. Eastop) (BMNH; slide mounted).Paratype. Nigeria: 1 $, Mokwa, 6-12. ix. 1964, Moericke bowl (/. A' Brook) (BMNH; slide mounted). COMMENTS. P. breviantennata is regarded as the sister-species of both trigemma and ngongae. Itdiffers from both of these in that the antennal flagellum is even more reduced with the loss offlagellomeres 3 and 5, these being the non-rhinarium-bearing flagellomeres of the primitiveantenna. In all three species, congruent with the reduction in length of the antenna, is thedevelopment of complex rhinaria and lengthening of the terminal setae. The septima-group Defined by the complex form of the apical segment of the aedeagus and the emarginate posteriormargin of the female subgenital plate. Two species are included: P. septima and P. proxima. Pauropsylla septima sp. n. (Figs 86-89) DESCRIPTION. Adult. Integument sparsely covered with short setae. Median suture of vertex present;antennal flagellum 1-14-1-40 times longer than head width, rhinaria simple, apical flagellomere with onelong and one short and truncate terminal seta; clypeus with a pair of setae, ultimate rostral segment with apair of short setae. Forewing (Fig. 86) 2-00-2-25 times longer than wide, veins very sparsely clothed withshort setae; forecoxa without ventroapical spur; tarsal segments subequal. cf proctiger flask-shaped, withpatches of fine setae on inner surfaces of lateral expansions ; paramere and aedeagus as in Figs 87 , 88 , latterwith complex apical segment; 9 genital segment rounded triangular in profile, posterior margin ofsubgenital plate with a small emargination and bordered with short setae (Fig. 89). Measurements (8 cf , 7 $). Maximum width of head, cf 0-42-0-47, 9 0-43-0-49; length of antennalflagellum, cf 0-54-0-61, 9 0-55-0-61; length of ultimate rostral segment, cf and <J>, 0-06-0-07; length offorewing, cf 1-87-2-13, 9 2-16-2-49; length of hind tibia, cf 0-48-0-58, 9 0-52-0-58. Larva and host plant unknown. Holotype cf , Cameroun: Bamenda, 20-24.U957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: 3 cf, 2 9, same data as holotype; 17 cf, 11 9, 25.i.-6.ii.l957 (V. F. Eastop)(BMNH; slide mounted and stored in Berlese fluid). COMMENT. P. septima and proxima are regarded as sister species and may be distinguished fromone another by the form of the male paramere (Figs 87, 90) and the female subgenital plate(Figs 89, 92). Pauropsylla proxima sp. n. (Figs 90-92, 283, 284) DESCRIPTION. Adult. Very similar to septima. Differs in the shape of the cf paramere (Fig. 90), and thedeeply emarginate and long-haired posterior margin of the 9 subgenital plate (Fig. 92). Measurements (10 cf , 10 9)- Maximum width of head, cf 0-46-0-53, 9 0-50-0-53; length of antennalflagellum, cf 0-59-0-73, 9 0-63-0-72; length of ultimate rostral segment, cf 0-07, 9 0-07-0-08; length offorewing, cf 2-09-2-24, $ 2-32-2-49; length of hind tibia, cf 0-53-0-59, 9 0-54-0-61. Fourth instar larva (Figs 283, 284). No 5th instar larvae available for study. Dorsal surface outline ovalwith clearly defined head, thorax and abdomen, about 1-5 times longer than wide. Small prothoracicsclerites present behind cephaloprothorax, arrangement of sclerites of rest of thorax as in Fig. 283; humerallobe of forewing bud not developed. Caudal plate excludes first four abdominal segments; anus ventral,pore ring complex (Fig. 284). Sectasetae absent from margin of cephaloprothorax, pointed sectasetaepresent on margins of wing buds and abdomen, dorsal surface of abdomen with a dense covering oflanceolate setae. AFROTROPICAL JUMPING PLANT LICE 33 HOST PLANT. Adults and larvae collected from Ficus sp., and adults only collected from Ficusthonningii (Moraceae). Holotype cf , Cameroun: Bamenda, 25-31.1.1957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: 6 cf, 15 $, same data as holotype; numerous cf and $, 20-24.1.1957 and1-6.11.1957. Angola: 3 cf , 7 $, larvae, Chianga, 10.x. 1970, Ficus sp.; 4 cf , 3 $, 7.x. 1971, Ficus thonningii(A. van Harteri). Nigeria: 4 cf , Zaria, 28.11.1956, yellow tray (V. F. Eastop). (BMNH; slide and drymounted, and stored in Berlese fluid.) COMMENT. See under P. septima. Ungrouped speciesPauropsylla angolensissp. n. (Figs 93, 94) DESCRIPTION. Adult. Similar to senegalensis (p. 34) but larger. Median suture of vertex weak butcomplete; antennal flagellum 1-13-1-17 times longer than head width; clypeus with two pairs of setae.Forewing 1-99-2-09 times longer than wide, cf proctiger with very weakly expanded lateral lobes;paramere and aedeagus as in Figs 93, 94; 9 genital segment short, rounded triangular in profile; posteriormargin of subgenital plate truncate. Measurements (2 cf, 1 $) Maximum width of head, cf 0-60-0-64, 9 0-63; length of antennal flagellum,Cf 0.-72, $ 0-74; length of ultimate rostral segment, cf 0-07, 9 0-08; length of forewing, cf 2-80, 9 3-03;leng'th of hind tibia, cf 0-78-0-79, 9 0-88. Larva and host plant unknown. Holotype cf , Angola: 3 mis SW. Salazar, 15.iii.1972, at light (D Hollis) (BMNH; dry mounted).Paratypes. 4 cf , 1 $ , same data as holotype (BMNH; slide and dry mounted). Pauropsylla eastopisp. n. (Figs 95-97) DESCRIPTION. Adult. Integument shiny, sparsely covered with short setae. Median suture of vertex present;antennal flagellum 1-24-1-45 times longer than head width, rhinaria simple, apical flagellomere withsubequal terminal setae ; clypeus with several short setae ; ultimate rostral segment without setae . Forewing2-08-2-26 times longer than wide, veins very sparsely clothed with short setae, m l cell value 0-86-0-98;forecoxa without ventroapical spur; tarsal segments subequal. cf proctiger (Fig. 97) with greatly enlargedlateral expansions which do not bear peg-like setae, paramere and aedeagus as in Figs 95, 96; 9 genitalsegment very short triangular in profile, sternite 6 of abdomen very weakly sclerotised ventrally,membrane between tergite 8 and proctiger with lateral glandular areas. Measurements (7 cf , 7 9)- Maximum width of head, cf 0-47-0-50, 9 0-49-0-51; length of antennalflagellum, cf 0-65-0-71, 9 0-62-0-65; length of ultimate rostral segment, cf and 9 0-06-0-07; length offorewing, cf 2-25-2-40, 9 2-43-2-61; length of hind tibia, cf 0-55-0-60, 9 0-57-0-60. Larva and host plant unknown. Holotype cf , Cameroun: Bamenda, 1-5. ii. 1957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: numerous cf and 9> same data as holotype; numerous cf and 9, 20-31.1.1957and 6.H.1957 (V. F. Eastop} (BMNH; slide and dry mounted, and stored in Berlese fluid). COMMENTS. This species is easily distinguished from all other known Pauropsylla species by theform of the male proctiger and genitalia (Figs 95-97), and the lack of sclerotisation of the femaleabdominal sternite 6. I can find no other characters which will relate it more closely to otherspecies-groups within the genus. Pauropsylla longipessp. n. (Figs 98-101, 287, 288) DESCRIPTION. Adult. Integument granular, sparsely covered with short setae. Median suture of vertexabsent; antennal flagellum 1-41-1-78 times longer than head width, rhinaria simple, apical flagellomerewith one long and one moderately long truncate seta terminally; clypeus with a pair of setae, ultimate 34 DAVID HOLLIS rostral segment with 2-4 pairs of setae. Forewing 2-05-2-26 times longer than wide, veins very sparselyclothed with short setae, m\ cell value 0-77-1-05; costal margin of hindwing with 2-4 setae proximal tocostal break; forecoxa without ventroapical spur; basitarsal segments 1-5 times as long as apical segmentsof tarsi, cf proctiger with a small group of up to 12 peg-like setae on the inner surface of each lateralexpansion (Fig. 100), paramere and aedeagus as in Figs 98, 99; 9 genital segment (Fig. 101) elongatetriangular in profile, subgenital plate without ventral transverse groove. Measurements (9 cf , 7 9)- Maximum width of head, cf 0-43-0-50, 9 0-43-0-52; length of antennalflagellum, cf 0-68-0-94, 9 0-66-0-80; length of ultimate rostral segment, cf 0-08-0-10, 9 0-09-0-10; lengthofforewing, cf 1-94-2-24, 9 1-91-2-47; length of hind tibia, cf 0-63-0-78, 9 0-67-0-81. Fifth instar larva (Figs 287, 288). Dorsal surface outline elongate oval with a clear indentation on eitherside at base of abdomen, and a posteromedian prolongation, about twice as long as wide (includingprolongation). Antenna with three flagellomeres. Cephalothorax entire, posterior margin serrate; fore-wing pad 0-52 mm long, humeral lobes weakly produced forward but not reaching hind margin of eye.Caudal plate about 1-3 times longer than wide (including prolongation); anus ventral and distant fromposterior margin of abdomen, anus and pore ring as in Fig. 288. Small narrow tubular sectasetae form asparse fringe on head and wing buds; both large and small tubular sectasetae form an uneven fringe aroundabdomen; a simple postocular seta present; sectasetae on dorsum as in Fig. 287. HOST PLANT. Adults and larvae collected from Ficus sp. (Moraceae). Holotype cf , Tanzania: S. Pare mtns, hillside above Gonja, c. 3,000', 12-16.vi.1974, Ficus sp. (D. Hollis)(BMNH; dry mounted). Paratypes. Tanzania: 18 cf, 7 9> same data as holotype; 21 cf, 19 9> larvae, E. Usambara mtns,Amani-Muheza Rd, below Kiswani, c. 2,000', 21-24.vi.1974, Ficus sp. (D. Hollis). Cameroun: 5 cf , 6 9,Bamenda, 24.i.-6.ii.l957, yellow tray (V. F. Eastop). Nigeria: 1 9, W. State, Ile-Ife, 6.ix.l971 (J. T.Medler). Ghana: 2 cf, Tafo (B. N. Gerrard); 1 cf, 12 9, 15.V.1957; 15 cf, 15 9, vi.1957 (V. F. Eastop).(BMNH; slide and dry mounted, and stored in Berlese fluid.) COMMENTS. This species resembles those of the willcocksi-group in lacking a median suture onthe vertex and having peg-like setae on the inner surfaces of the lateral expansions of the maleproctiger, but differs in lacking a ventroapical spur on the forecoxa. It is readily distinguishedfrom other Afrotropical Pauropsylla species by the proportions of the tarsal segments. Pauropsylla mistura sp. n. (Figs 102-104) DESCRIPTION. Adult. Very similar to senegalensis (p. 34) but larger. Median suture of vertex present andcomplete; antennal flagellum 1-15-1-31 times longer than head width. Forewing 2-21-2-30 times longerthan wide, cf proctiger with very weak lateral expansions, paramere and aedeagus as in Figs 102, 103; 9genital segment short, rounded triangular in profile (Fig. 104), ventral margin of subgenital plate roundedin profile. Measurements (3 cf , 6 9)- Maximum width of head, cf 0-50-0-51, 9 0-51-0-53; length of antennalflagellum, cf 0-62-0-64, 9 0-61-0-67; length of ultimate rostral segment, cf 0-07, 9 0-07-0-08; length offorewing, cf 2-49-2-64, 9 2-58-2-78; length of hind tibia, cf 0-58-0-60, 9 0-59-0-66. Larva and host plant unknown. Holotype cf , Cameroun: Bamenda, 20-24.1.1957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: 1 9 > same data as holotype; 1 cf,3 9, 25. i. -5. ii. 1957. Nigeria: 1 cf, 1 9> Samaru,viii.1956; 1 9, Jos, 25-26. xi. 1956, yellow tray (V. F. Eastop). Ghana: 1 9, Tafo, v.1957 (V. F. Eastop).(BMNH; slide mounted.) COMMENTS. P. mistura is larger than senegalensis, has the male paramere and aedeagus of adifferent shape, and a rounded conical female genital segment. It may be separated fromangolensis on the form of the male genitalia. Pauropsylla senegalensis sp. n. (Figs 105- 108, 285, 286) DESCRIPTION. Adult. Integument granular, sparsely covered with very short setae. Median suture of vertexweak, evanescing before reaching occiput; antennal flagellum 1-06-1-27 times longer than head width, AFROTROPICAL JUMPING PLANT LICE 35 rhinaria simple, apical flagellomere with one long and one short and truncate seta; clypeus with a pair ofsetae, ultimate rostral segment without setae. Forewing 1-91-2-03 times longer than wide, veins verysparsely clothed with short setae, mi cell value 0-69-0-91. Forecoxa without ventroapical spur; tarsalsegments subequal. d" proctiger (Fig. 107) flask-shaped, with weak lateral expansions which bear simplesetae on their inner surfaces, paramere and aedeagus as in Figs 105, 106; $ genital segment (Fig. 108)short, subrectangular in profile, proctiger with a weak transverse groove posterior to anal pore, subgenitalplate with a narrowly arcuate posterior margin, ventral margin subangular in profile. Measurements (7 d", 4 $). Maximum width of head, cf 0-45-0-48, 0-48-0-50; length of antennalflagellum, cf 0-54-0-61, 9 0-51-0-58; length of ultimate rostral segment, cf and $ 0-06-0-07; length offorewing, cf 1-75-1-87, $ 1-93-2-09; length of hind tibia, cf 0-50-0-54, $ 0-54-0-56. Fifth instar larva (Figs 285, 286). Dorsal surface outline broadly oval, about 1-2 times longer than wide.Antenna with two flagellomeres. Cephalothorax entire; forewing pad 0-65 mm long, humeral lobeextending forward as far as anterior margin of eye. Caudal plate about 0-6 times as long as wide, anusventral and distant from posterior margin of abdomen, anus and pore ring as in Fig. 286. Truncate tubularsectasetae forming an even dense fringe; postocular seta absent, sectasetae absent from dorsum. HOST PLANT. Adults and larvae collected from Ficus sp. (Moraceae). Holotype cf , Senegal: Simbandi-Balante, 18.vi.1981, Ficus sp. (/. Etienne) (BMNH; dry mounted). Paratypes. Numerous adults and larvae, same data as holotype (BMNH; MNHN; slide and drymounted, and stored in 80% ethanol). COMMENTS. This species is very similar to mistura and angolensis, but it may be distinguishedfrom these by its smaller size, the incomplete median suture of the vertex and the form of themale and female genitalia (Figs 105-108). TRIOZA Foerster Trioza Foerster, 1848: 67. Type-species: Chermes urticae Linnaeus, by subsequent designation of Oshanin,1912: 128. Powellia Maskell, 1879: 223. Type-species: Powellia vitreoradiata Maskell, by monotypy. [Synonymised byMaskell, 1890: 167.] Phylloplecta Riley, 1884: 319. Type-species: Psylla tripunctata Fitch, by monotypy. [Synonymised byTuthill, 1943: 546.] Phyllopecta Ferris, 1926: 16. [Misspelling of Phylloplecta Riley; Tuthill, 1943: 546.] Spam'ozaEnderlein, 1926: 400. Type-species: Trioza galii Foerster, by original designation. [Synonymisedby Tuthill, 1943:546.] Colopelma Enderlein, 1926: 400. Type-species: Trioza thomasii Low, by original designation. [Synony-mised by Tuthill, 1943: 546.] Siphonaleyrodes Takahashi, 1932: 48. Type-species: Siphonaleyrodes formosanus Takahashi, by originaldesignation. [Synonymised by Mound & Halsey, 1978: 252.] Only established synonymy is stated above but the following references, given under thesynonymy of the family Triozidae (p. 8), contain regional revisions of Trioza: Vondracek,1957 (Czechoslovakia); Ramirez-Gomez, 1960 (Spain); Dobreanu & Manolache, 1962 (Ruma-nia); Loginova, 1964 (European U.S.S.R.); Klimaszewski, 1973 (Palaearctic check-list); 1975(Poland); Sulc, 1910; 1911; 1912; 1913 (revision of Palaearctic species); Crawford, 1919(Oriental, Australo-oriental and Pacific); Mathur, 1975 (India); Froggatt, 1901 (Australia);Tuthill, 1952 (New Zealand); Zimmermann, 1948 (Hawaii); Crawford, 1914 (New World);Tuthill, 1943 (N. America); 1944 (Mexico); 1959; 1964 (Peru). Tuthill's argument (1943: 547) that Phylloplecta Riley is a nomen nudum is not accepted hereas Riley's statement 'Phylloplecta tripunctata, which crumbles the tips of blackberry' makesclear he is referring to Psylla tripunctata and validates Phylloplecta Riley, 1884 under Article16(v) of the International Code of Zoological Nomenclature. As the following species are here placed in Trioza on a temporary basis, a detailed descriptionof the genus is of no systematic value. These species differ from Afrotropical species placed inother triozid genera as follows. Median suture of vertex present and normally complete; genal cones, when present, notconstricted basally. Propleural suture diagonal, episternum enlarged, epimeron reduced, 36 DAVID HOLLIS displaced ventrally and not in contact with lateral margin of pronotum. Forewing shape mostlyelongate elipsoid and narrowing to a subangular apex, if with rounded apex then more than 2-3times longer than wide (exception T. guiera, p. 62); radular areas present only in cells m\, m 2and cu\\ claval suture reaching hind margin of wing some distance from apex of Cw lb . Ventralsense organs of hind femur in median position; basal tarsal segment of hind leg without apicalspurs. Male proctiger unipartite. For the species treated below the grouped species are considered before those not placed ingroups. The erytreae-group is dealt with first as it contains the major pest species and, probably,is most closely related to the type-species. Other species-groups are treated in descending orderof apparent relationship to the erytreae-group. The ungrouped species are treated in alphabeti-cal order. The erjtreae-group A difficult group to define, the 10 species included having no single character which will delimitthem from other Trioza species, and therefore the grouping may be artificial. The description oferytreae serves to define the group. The species included are: erytreae, catlingi, menispermicola,gregoryi, ata, kilimanjarica, tiliacora, carvalhoi, capeneriandeafra. Known host plant families are Rutaceae, Menispermaceae, Araliaceae and possibly Sali-caceae. This is a rather wide range but there may be chemical similarities within these families(Thorne, 1976: 85; Dahlgren, 1977: 260). Trioza erytreae (Del Guercio) (Figs 2, 7-9, 34, 35, 40-42, 49, 52, 54, 55, 109-111, 126, 127, 289, 290) Citrus Psylla (Trioza); Lounsbury, 1897: 116. 'Psyllidengalle'; Citrus aurantium; Riibsaamen, 1899: 266. Aleurodes erytreae Del Guercio, 1918: 167. Syntypes, larvae, ETHIOPIA 'Eritrea' [not traced]. Trioza citri Laing; Waterston, 1922: 49, 55. [Nomen nudum.] Trioza merwei Pettey, 1923: 30. Syntypes SOUTH AFRICA 'Natal, Durban, on orange and Toddalia lanceolata" (SAM; USNM) [not examined]. [Synonymised by Pettey, 1933: 19.]Spanioza merwei (Pettey); Enderlein, 1926: 400. Spanioza erythreae (Del Guercio); Boselli, 1930ft: 228; Pettey, 1933: 19; Harris, 1936: 498.Citrus Psylla (Spanioza erytreae} Del Guercio; Van der Merwe, 1941: 5.Trioza erytreae (Del Guercio); Capener, 1970: 200. Only primary taxonomic references are quoted above. Further references on the biology andeconomic aspects of this species are given in the introduction (p. 4). DESCRIPTION. Adult. Integument sparsely covered with short setae. Head, in profile, almost at 90 tolongitudinal axis of body, from above almost as wide as mesoscutum; occipital margin rounded; vertexpentagonal with anterior margin deeply incised by median suture, rounded down to frons, lateral ocelli onouter sides of raised tubercles, a shallow concavity present on either side of median suture; median ocellusnot visible in dorsal view; frons completely covered by genae in anterior view; genal cones well developed,elongate conical with rounded apices; antennal flagellum 2-08-2-81 times longer than head width, headwidth to length of 1st flagellum in c? 1-26-1-70, in $ 1-30-1-82; a single rhinarium present subapically onflagellomeres 2, 4, 6 and 7, apical flagellomere with a long pointed seta and a short truncate seta apically;clypeus with a pair of setae, ultimate rostral segment with two pairs of setae. Thorax strongly arched;pronotum just visible from above, in profile strongly rounded down behind occiput. Mesopraescutumabout as wide as long, its anterior margin strongly arcuate in dorsal view, in profile strongly downcurved topronotum; forewing hyaline, elongate oval and narrowing to a rounded rectangular apex, 2-79-3-09 timeslonger than wide, radular areas elongate triangular, remainder of membrane devoid of spinules; veinsbearing short setae, R branch acutangular, M branch distal to RsCu^ line, Cu stem 2-75^-20 timeslonger than Cu lb , mi cell value 1-10-1-38, cu\ cell value 2-56-3-71; forewing 1-59-1-82 times longer thanhindwing, costal margin of hindwing with up to two setae proximal to costal break, setae distal to costalbreak clearly divided into two groups; hind coxa with a well-defined meracanthus and without anteriorlobe ; hind tibia with a moderately developed basal spine , with one outer and three (rarely two) inner apical AFROTROPICAL JUMPING PLANT LICE 37 spurs. Abdomen with setae on tergites 2 and 3 in cf , and 3 and 4 in $ ; cf proctiger (Fig. 109) with a laterallyexpanded basal part and a very short and narrow apical part; paramere as in Fig. 110; apical segment ofaedeagus simple (Fig. Ill); $ genital segment (Fig. 127) short, conical, subgenital plate with a ventralbulge, ventral valves of ovipositor weakly serrate apically. Measurements (19 cf , 21 $). Maximum width of head, cf 0-37-040, $ 0-38-0-46; length of antennalflagellum, Cf 0-85-1-10, $ 0-83-1 -10; length of ultimate rostral segment, cf 0-09-0-10, $ 0-09-0- 11; lengthof forewing, cf 2-40-2-96, $ 2-61-3-46; length of hind tibia, cf 0-50-0-62, $ 0-48-0-62. 54 0-30- 0-200-35 0-40head width (mm) A- erytreae - catlingi - menispermicola 55 no- 5 100- A A i A A A 0-6 length of forewing pad (mm) 0-9 1-0 Figs 54, 55 Scatter diagrams comparing characters of Trioza erytreae, T. catlingi and T. menispermicola.54, adults, length of 1st flagellomere against head width. 55, larvae, number of sectasetae on forewingpad against length of forewing pad. 38 DAVID HOLLIS AFROTROPICAL JUMPING PLANT LICE 39 Fifth instar larva (Figs 289, 290). Dorsal surface outline oval, about 1-5 times longer than wide. Antennawith 4-5 flagellomeres. Cephaloprothorax separate from rest of thorax which is entire. Forewing pad about0-8 mm long, humeral lobe extending forward beyond anterior margin of eye. Caudal plate about 0-65times as long as wide, anus ventral and distant from posterior margin of abdomen, anus and pore ring as inFig. 290. Truncate tubular sectasetae forming a dense, entire marginal fringe, postocular seta absent,sectasetae absent from dorsum. HOST PLANTS. Larvae, which form pit-galls on the lower surfaces of the leaves, and adults havebeen collected from Clausena anisata, Fagara capensis, Vepris undulata and Citrus spp. DISTRIBUTION. Widely distributed in the Afrotropical Region and specimens have been ex-amined from Cameroun, Zaire, Angola, Ethiopia, Ruanda, Uganda, Kenya, Tanzania, Mala-wi, Zimbabwe, South Africa, Sao Tome, St Helena, Reunion and Madagascar. COMMENTS. T. erytreae is the only member of this group known to develop on members of theRutaceae. Apart from citrus I have collected adults and larvae from Clausena anisata in EastAfrica but Van Der Merwe (1941) records the insect from Vepris undulata (as Toddalialanceolata), Fagara capensis and Clausena anisata (as C. inaequalis) in South Africa, stating thatthe last seemed to be the preferred host. Moran (1968) concludes that V. undulata and/or C.anisata were probably the indigenous host plants in South Africa. Members of the erytreae-group are morphologically homogeneous and male genitalia charac-ters are not always species diagnostic. These characters can be used to separate erytreae fromgregoryi (Figs 114, 115), kilimanjarica (Fig. 118), carvalhoi (Figs 122-124) and eafra (Fig. 125)but not the rest of the group. T. tiliacora is easily separated as it has setae on all abdominaltergites and a relatively broader forewing; capeneri has a lower cu cell value and a relativelylonger hindwing; in ata the clypeus has two pairs of setae and a relatively shorter ultimate rostralsegment. It is extremely difficult to separate catlingi and menispermicola from erytreae: incatlingi the first flagellomere is longer (head width to length of 1st flagellomere 1-06-1-23) andthe marginal sectasetae of the 5th instar larva are less dense (Fig. 54); in menispermicola theadults appear identical to erytreae but the marginal sectasetae are less dense (Fig. 55). Asummary of the characters used to separate members of the erytreae-group is given in Table 4. Trioza catlingi sp. n. (Figs 54, 55, 112, 113, 291, 292)Trioza sp.; Catling, 1969ft: 220; McDaniel & Moran, 1972: 299. DESCRIPTION. Very similar to T. erytreae. First flagellomere longer, head width to length of 1st flagellomereratio in cf 1-06-1-23 and 1-11-1-25 in 9- Marginal sectasetae of 5th instar larva less dense (Figs 55, 291). Measurements (10 cT, 8 <j>). Maximum width of head, cf 0-37-0-41, $ 0-40-0-43; length of antennalflagellum,cf 1-01-1-16, $ 1-05-1-22; length of ultimate rostral segment, cf 0-09-0-11, $ 0-10-0-11; lengthof forewing, cf 2-61-3-12, $ 2-74-3-24; length of hind tibia, cf 0-58-0-65, $ 0-57-0-68. HOST PLANTS. Larvae, which form pit galls, and adults have been collected from Stephaniaabyssinica and Cissampelos sp. (Menispermaceae). Holotype cf, Kenya: Kakamega For. sta., c. 5,000', 18-19. vii. 1974, Stephania abyssinica (D. Hollis)(BMNH; dry mounted). Paratypes. Kenya: 10 cf, 6 $, larvae, same data as holotype; 5 cf, 12 $, larvae, S. slope Mt Kenya,Thiba camp, c. 6,000', 9-10.vii.1974, Cissampelos sp.; 2 cf , 2 $, Tinderet For., 5 km along Lessos rd, c.8,000', 20. vii. 1974, Stephania abyssinica. Tanzania: 5 cf , larvae, Arusha NP, track to Ngurdoto, c. 5,000',7.vi.l974, Stephania abyssinica; 11 cf , 9 $, Ngurdoto Crater rim, c. 5,000', 8.vi.l974; 9 cf , 12 $, larvae,Stephania abyssinica; 6 cf , 1 $ , Meru FR, track to crater, 6,000-9,000' , 9.vi. 1974, Stephania abyssinica (D.Hollis). South Africa: 1 $, N. Tvl, Letaba, ix.1966, Cissampelos torulosa ( A. P. D. McCleah). (BMNH;NMK; slide and dry mounted and stored in 80% ethanol.) 40 DAVID HOLLIS Trioza menispermicola sp. n. (Figs 54, 55) DESCRIPTION. Very similar to erytreae and catlingi. Differs from catlingi in that ratio of head width to lengthof 1st flagellomere is 1-54 1-87 in cf and 1-52-1-80 in $. $ subgenital plate without ventral bulge. Differsfrom erytreae in the density of marginal sectasetae of the 5th instar larva (Fig. 55). Measurements (9 cf , 10 <j>). Maximum width of head, cf 0-37-0-43, $ 0-38-0-45; length of antennalflagellum, cf 0-83-0-89, $ 0-84-0-93; length of ultimate rostral segment, cf and $ 0-10-0-11; length offorewing, cf 2-27-2-64, $ 2-49-3-08; length of hind tibia, cf 0-47-0-53, $ 0-49-0-56. HOST PLANTS. Larvae, which form pit galls, and adults have been collected from Triclisiamacrophylla, T. patens and Cissampelos owariensis (Menispermaceae). Holotype cf , Ghana: 'Gold Coast; Bunsu, Jan. 1943, Triclisia macrophylla' (H. E. Box) (BMNH; drymounted). Paratypes. Ghana: 1 cf, 5 $ , larvae, same data as holotype. Nigeria: 100 cf, 100 $ , larvae, Ibadan, Moorplntn, 7.iv.l956, 'Pseudogalls on creeper' (R. Donald); 4 cf , 2 $, larvae, Ilesha, 30.xii.1943, Cissampelosowariensis (H. E. Box). No locality data: 3 cf, 2 $, larvae, 22.xii.1942, Triclisia patens (H. E. Box).(BMNH; slide and dry mounted and stored in 80% ethanol.) Trioza gregoryi sp. n. (Figs 114, 115) DESCRIPTION. Adult. Very similar to erytreae. Differs in shape of cf paramere (Fig. 114) and apical segmentof aedeagus (Fig. 115). $ subgenital plate smoothly rounded and without ventral bulge. Measurements (9 cf , 8 $). Maximum width of head, cf 0-35-0-38, $ 0.36-0-39; length of antennalflagellum, cf 0-72-0-86, $ 0-75-0-82; length of ultimate rostral segment, cf 0-09-0-11, $ 0-09-0-12; lengthof forewing, cf 2-30-2-74, cf 2-55-2-94; length of hind tibia, cf 0-44-0-57, $ 0-47-0-56.Larva and host plant unknown. Holotype cf, Nigeria: Umuahia, Umudike, 28.viii.1958, yellow tray (/. L. Gregory) (BMNH; slidemounted). Paratypes. Nigeria: 6 cf , 7 $, same data as holotype. Burundi: 1 Cf , 1 $, Gitega, 20.x. 1980 (Pointel).Tanzania: 2 cT, 1 $, E. Usambara mtns, Amani Res. sta., 19-27.vi.1974 (D. Hollis). (BMNH; MNHN;slide mounted.) Trioza ata sp. n. (Figs 116, 117, 128) DESCRIPTION. Adult. Very similar to erytreae but relatively larger (see measurements below). Clypeus withtwo pairs of setae; ultimate rostral segment relatively shorter, head width to ultimate rostral segmentlength ratio 4-6-5-22. cf proctiger less expanded laterally, paramere and apical segment of aedeagus as inFigs 116, 117. Ventral valves of ovipositor with more serrations along apical part (Fig. 128). Measurements (13 cf , 6 <J>). Maximum width of head, cf 0-43-0-49, $ 0-46-0-50; length of antennalflagellum, cf 1-04-1-20, $ 1-13; length of ultimate rostral segment, cf and $ 0-09-0-10; length of forewing,Cf 3-11-3-51, $ 3-41-3-61; length of hind tibia, cf 0-53-0-60, $ 0-55-0-57. Larva unknown. HOST PLANT. Adults collected from Salix sfl/sa/(Salicaceae). Holotype cf , Angola: 12 mis SW. Luimbale, c. 5,500', 20-2 l.iii. 1972, Salix safsaf (D . Hollis) (BMNH; drymounted). Paratypes. Angola: 19 cf, 9 $, same data as holotype; 7 cf, 3 9, Tundavala, 8-10 mis NW. Sa daBandeira, 27-29.iii.1972, Salix safsaf (D. Hollis). Tanzania: 2 cf , Kilimanjaro, Bismark Hut, 2,500-3,000m, S. Mawenzi, at foot of high pasture, ii. 1912 (Chr. Schroder). (BMNH; MNHU; slide and dry mounted.) Trioza kilimanjarica sp. n. (Fig. 118)DESCRIPTION. Adult. Similar to ata. Distinguished by its smaller size (see measurements below); head width AFROTROPICAL JUMPING PLANT LICE 41 to ultimate rostral segment length ratio 5-29-5-71. cf paramere as in Fig. 118. $ subgenital plate withoutventral bulge, ventral valves of ovipositor serrate only at extreme apex. Measurements (7 cf , 4 $). Maximum width of head, cf 0-37-0-38, $ 0-38-0-40; length of antennalflagellum, cf 0-73-0-80, $ 0-82; length of ultimate rostral segment, cf and $,0-07; length of forewing, cf2-37-2-52, $> 2-64-2-87; length of hind tibia, cf 0-40-0-45, $ 0-42-0-45. Larva and host plant unknown. Holotype cf , Tanzania: Kilimanjaro, Bismark Hut, 2,500-3,000 m, S. Mawenzi, at foot of high pasture,ii.1912 (Chr. Schroder} (MNHU; slide mounted).Paratypes. 6 cf , 4 $, same data as holotype (MNHU; BMNH; slide mounted). Trioza tiliacora sp. n. (Figs 119-121, 129, 293, 294) DESCRIPTION. Adult. Of similar general appearance to erytreae and catlingi. Integument slightly moredensely haired. Genal cones less elongate; antennal flagellum 2-05-2-41 times longer than head width.Forewing with an even yellowish infuscation in mature specimens, 2-46-2-61 times longer than wide; veinswith slightly longer setae; Cu stem 2-35-3-10 times longer than Cw lb ; m t cell value 1-08-1-26, cu\ cell value2-37-2-82. Abdomen with setae present on all visible tergites; cf genitalia as in Figs 119-121; $ proctigerslightly less elongate apically than in erytreae and subgenital plate with less pronounced ventral bulge. Measurements (4 cf , 7 $). Maximum width of head, cf 0-40-0-42, $ 0-41-0-44; length of antennalflagellum, cf 0-92-1-0, $ 0-88-1-01; length of ultimate rostral segment, cf 0-10-0-11, $ 0-11-0-12; lengthof forewing, Cf 2-46-2-81, $ 2-61-3-09; length of hind tibia, cf 0-56-0-62, $ 0-54-0-62. Fifth instar larva (Figs 293, 294). Dorsal surface outline broadly oval, slightly emarginate posteriorly,about 1-4 times longer than wide. Antenna with 5-6 flagellomeres. Cephaloprothorax separated from restof thorax which is entire. Forewing pad about 0-8 mm long, extending forward anterior to anterior marginof eye. Caudal plate about 0-56 times as long as wide, anus posteroventral (Fig. 294). Truncate tubularsectasetae forming a moderately dense marginal fringe but absent on posteromedial margin, postocularsetae absent, sectasetae absent from dorsum. HOST PLANT. Adults and larvae collected from Tiliacora sp. (Menispermaceae) in associationwith ants of the genus Technomyrmex . The larvae were causing severe leaf curl. Holotype cf , Tanzania: E. Usambara Mtns, Amani Res. sta., 19-27. vi. 1974, c. 3,000', Tiliacora sp., inassociation with Technomyrmex sp. (D. Hollis) (BMNH; dry mounted).Paratypes. 5 cf , 13 9 , larvae, same data as holotype (BMNH; slide and dry mounted). COMMENTS. Unlike other species in the erytreae-group, tiliacora has the chaetotaxy of theabdominal tergites in the primitive condition. Furthermore the anus of the 5th instar larva is in aposteroventral position which is normally regarded as a primitive state. However, in this case Ibelieve it to be secondarily derived and associated with ant-tending. Trioza carvalhoisp. n. (Figs 122-124, 130,295,296) DESCRIPTION. Adult. In general appearance very similar to erytreae. Antennal flagellum slightly shorter,1-95-2-27 times longer than head width. Cu stem 2-19-3-06 times longer than Cw lb , cu v cell value 1-6-2-26;hindwing relatively longer, ratio of length of forewing to length of hindwing 1-42-1-54. cf genitalia as inFigs 122-124; $ subgenital plate without ventral bulge. Measurements (13 cf , 19 $). Maximum width of head, cf 0-36-0-41, $ 0-40-0-43; length of antennalflagellum, cf 0-79-0-86, $ 0-81-0-96; length of ultimate rostral segment, cf and $,0-09-0-10; length offorewing, cf 2-40-2-55, $ 2-66-2-92; length of hind tibia, cf 0-48-0-52, $ 0-48-0-55. Fifth instar larva (Figs 295, 296). Very similar to erytreae. Antenna with 5-6 flagellomeres. Forewing padabout 0-66 mm long, humeral lobe extending forward to anterior margin of eye. Anus and pore ring as inFig. 296. HOST PLANTS. Adults and larvae collected from Cussonia spicata, C. angolensis and C.paniculata. The larvae form pit galls on the lower surfaces of the leaves and cause severe leafdistortion. 42 DAVID HOLLIS Holotype cf, Kenya: N. slope Mt Kenya, Naro Moru lodge area, c. 6,000', 11-14. vii. 1974, Cussoniaspicata (D. Hollis) (BMNH; dry mounted). Paratypes. Kenya: 15 d", 13 $, same data as holotype; 1 $, Thomson's Falls, path into gorge, c. 7,800',15-16.vii.1974; 1 9, Marmanet FR, 5-15 km N. Thomson's Falls, c. 8,000', 16.vii.1974. Angola: 1 cf , 12mis SW. Luimbale, c. 5,500', 20-2 l.iii. 1972; 11 cf, 23 $, larvae, Chianga, 2 1-24. iii. 1972, Cussoniaangolensis (D. Hollis) ; 7 cf , 7 $, larvae, 29.x. 1970, Cussonia angolensis; 1 cf, 4 $,i. 1971, vagrants (A. vanMarten); 3 cf, 5 9 8. i. 1973, Cussonia angolensis (L. Amorim). South Africa: 3 cf, 3 9, larvae, N. Tvl,Cussonia spicata (C. Malan); 2 cf, 3 $, Natal, Cathedral Peak, 18. i. 1964, Cussonia paniculata (A. L.Capener); 1 cf , 1 9, Swaziland, M.R.S., iv.1966, Cussonia sp. (R.C.H. Sweeney). (BMNH, NCI; NMK;slide and dry mounted, and stored in 80% ethanol.) Trioza capenerisp. n. (Fig. 297) DESCRIPTION. Adult. Very similar to erytreae and carvalhoi. Antennal flagellum 1-95-2-28 times longerthan head width. Cu stem 2-27-3-40 times longer than C lb ; m t cell value 1-26-1-55, cui cell value1-91-2-45. Forewing 1-46-1-58 times longer than hindwing. cf genitalia not distinguishable from erytreae. Measurements (6 cf , 6 $). Maximum width of head, cf 0-38-0-41, 9 0-41-0-43; length of antennalflagellum, cf 0-79-0-91, 9 0-84-0-89; length of ultimate rostral segment, cf 0-09-0-10, 9 0-10-0-11; lengthof forewing, cf 2-54-2-80, 9 2-93-3-12; length of hind tibia, cf 0-48-0-51, 9 0-50-0-53. Fifth instar larva. Very similar to carvalhoi, anal pore area as in Fig. 297. HOST PLANT. One sample of adults and larvae collected from Seemannaralia gerrardii (Ara-liaceae). The larvae were forming pit galls on the fruit. Holotype cf , South Africa: Natal, vii. 1972, from pit galls on fruits of Seemannaralia gerrardii (A. L.Capener) (NCI; slide mounted). Paratypes. 13 Cf , 18 $, larvae, same data as holotype (NCI; BMNH; slide mounted and stored in 80%ethanol). COMMENTS. The male genitalia of this species are very similar to erytreae but the cu\ cell valueand relatively longer hindwing are similar to carvalhoi. Its host plant is a member of theAraliaceae and I am regarding it as the sister species of carvalhoi. Trioza eafra sp. n. (Fig. 125) DESCRIPTION. Adult. Very similar to carvalhoi. Differs in having shorter antennae, antennal flagellum1-62-1-97 times longer than head width. Forewing membrane covered with spinules except for narrowareas bordering wing veins; cui cell value 1-61-2-45; forewing 1-37-1-56 times longer than hindwing. cfparamere as in Fig. 125. Measurements (11 cf, 10 9)- Maximum width of head, cf 0-35-0-40, 9 0-35-0-42; length of antennalflagellum, cf 0-59-0-71, 9 0-60-0-80; length of ultimate rostral segment, cf and $,0-08-0-09; length offorewing, cf 1-87-2-41, $ 1-99-2-70; length of hind tibia, cf 0-36-0-43, 9 0-36-0-47. Larva unknown. HOST PLANT. Adults collected from Cussonia spicata (Araliaceae). The leaves of the host treeswere heavily pitted but no larvae were found. Holotype cf , Kenya: L. Naivasha, W. shore rd, c. 6,200', 21-22. vi. 197 '4, Cussonia spicata (D. Hollis)(BMNH; dry mounted). Paratypes. Kenya: 9 cf, 16 $, same data as holotype; 5 cf , 8 $, Thomson's Falls, path into gorge, c.7,800', 15-16. vii. 1974, Cussonia sp.; 1 9, Mt Londiani For., 57 km W. Nakaru, c. 8,500', 20. vii. 1974 (D.Hollis); 2 cf , 2 $, Muguga, vi.1953, trapped. Tanzania: 5 cf , 6 9, Nachingwea, xi.1953-iii.1954, trapped(V. F. ostop);3cf,69,E.UsambaraMtns, Amani-Muhezard, below Kiswani,c. 2,000', 21-24.vi. 1974,Cussonia sp. (D. Hollis). (BMNH; NMK; slide and dry mounted.) COMMENTS. This species is the only member of the erytreae-group to have retained (or revertedto) the primitive condition of having spinules on the forewing membrane. It has the relativelylonger hindwing and low cu\ cell value of carvalhoi and capeneri and I regard it as thesister-species of this pair. AFROTROPICAL JUMPING PLANT LICE 43 The /itseae-group Very similar to the erytreae-group but body form and fore wings narrower, the hind tibia has twoinner apical spurs and the female anal pore has a single ring of wax-producing cells. Two veryclosely related species are included: litseae and xylopia. Host plants of both are in theAnnonales. Trioza xylopia sp. n. (Figs 131-134, 298, 299) DESCRIPTION. Adult. Integument sparsely covered with very short setae. Head, in profile, almost at 90 tolongitudinal axis of body, in dorsal view about as wide as mesoscutum; occipital margin rounded; vertexrounded pentagonal, strongly rounded downwards, with a deep median suture on either side of which is anirregular concavity, lateral ocelli on outer sides of raised humps; median ocellus not visible from above, inanterior view frons completely covered by genae; genal cones well developed, downward pointing, conicalwith rounded apices; antenna! flagellum 1 -97-2-53 times longer than head width, a single rhinarium presentsubapically on flagellomeres 2, 4, 6, and 7, apical flagellomere bearing one long pointed and one shorttruncate seta terminally; clypeus with a pair of short setae, ultimate rostral segment with a pair of setae.Thorax strongly arched; pronotum visible from above, strongly rounded down behind occiput; mesopraes-cutum, in profile, angled down to pronotum, in dorsal view anterior margin strongly arcuate; forewingelongate oval, narrowing to a subacutangular apex, 2-94-3-43 times longer than wide, membrane withspinules in all cells apart from c+sc and r l5 radular areas narrow, veins bearing very short setae, R branchacutangular, M branching distal to RsCui a line, Cu stem 1-78-2-17 times longer than Cw lb , m\ cell value1-09-1-38, cu\ cell value 1-73-2-19; forewing 1-54 times longer than hindwing, costal margin of hindwingwithout setae proximal to costal break, setae distal to costal break clearly divided into two groups. Hindcoxa with a well-developed meracanthus, without anterior lobe; hind tibia with 1-2 small basal spines andone outer and two inner apical spurs. Abdomen with setae on tergites 2-3 in cf and 3-4 in $ ; cf proctiger(Fig. 132) roundly expanded laterally, paramere and apical segment of aedeagus as in Figs 133, 134; $genital segment short, conical, ventral surface of ventral valve of ovipositor smooth. Measurements (7 cf, 6 $). Maximum width of head, cf 0-34-0-38, $ 0-36-0-41; length of antennalflagellum, cf 0-76-0-86, $ 0-75-0-84; length of ultimate rostral segment, cf 0-06-0-07, $ 0-07; length offorewing, cf 1-94-2-22, $ 2-34-2-59; length of hind tibia, cf 0-34-0-43, $ 0-41-0-44. Fifth instar larva (Figs 298, 299). Dorsal surface outline oval but with clear indentations at posteriormargin of eye and at base of abdomen, about 1-75 times longer than wide. Antenna with four flagello-meres. Cephaloprothorax separate from rest of thorax which is entire. Forewing pad about 0-6 mm long,humeral lobe weakly extended forward to just beyond posterior margin of eye. Caudal plate about 0-7times as long as wide; anus ventral, a short distance from posterior margin of caudal plate, anal pore area asin Fig. 299. Narrow tubular sectasetae forming an even and moderately dense marginal fringe, postoculartubular sectaseta present, sectasetae absent from dorsum. HOST PLANT. Larvae and adults swept from Xylopia sp. (Annonaceae) ; the larvae are apparentlyfree-living as no evidence of galls was observed. Holotype cf , Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27.vi.1974, Xylopia sp. (D.Hollis) (BMNH; dry mounted). Paratypes. Tanzania: 14 cf , 17 $ , same data as holotype. Kenya: 1 cf , W. slope Mt Kenya, Naro Morulodge area, c. 6,500', 11-14. vii. 1974, Cussonia spicata (D. Hollis). (BMNH; slide and dry mounted.) COMMENTS. T. xylopia is very closely related to litseae and the two species can be separated onlyby the form of the 5th instar larvae (Figs 298-301). Both are similar to eafra (p. 42), in theerytreae-group , but have only two inner apical spurs on the hind tibia. Trioza litseae Bordage(Figs 300, 301) Trioza litseae Bordage, 1898: 524; 1914: 409. Syntype[s] [? sex], REUNION: Litsea (Tetranthera) laurifolia [not traced].Trioza eastopi Orian, 1972: 4. Holotype cf , MAURITIUS: Litsea glutinosa (BMNH) [examined]. Syn. n. DESCRIPTION. Adult. Very similar to xylopia and, apart from its smaller size, not morphologically distinctfrom that species. 44 DAVID HOLLIS Measurements (5 cf , 8 9). Maximum width of head, cf 0-28-0-32, 9 0-30-0-34; length of antennalflagellum, cf 0-66-0-73, 9 0-61-0-70; length of ultimate rostral segment, cf 0-06-0-07, 9 0-07-0-08; lengthof forewing, cf 1-39-1-79, 9 1-78-2-06; length of hind tibia, cf 0-29-0-35, $ 0-31-0-36. Fifth instar larva (Figs 300, 301). Similar to xylopia but broader, about 1-5 times longer than wide indorsal view. Forewing pad about 0-5 mm long. Anus ventral and distant from posterior margin of caudalplate, anal pore area as in Fig. 301 . Marginal setae consisting of a moderately dense fringe of broad tubularsectasetae of varying lengths; postocular tubular sectaseta present; dorsum with a sparse covering ofsectasetae. HOST PLANT. Larvae and adults collected from Litsea glutinosa [= L. laurifolia]; the larvaedamage the flower buds and flowers of the host; adults are also known to damage Vanilla. MATERIAL EXAMINED. Reunion (Litsea glutinosa); Mauritius (L. glutinosa and 'citrus'). Bor-dage's type-series was not traced and may have never been preserved. Orian's holotype andsome of the paratypes, which are slide-mounted, are damaged in that the gum mountant hasbecome opaque. COMMENTS (see also p. 43). Orian (1972) regarded Trioza litseae Gd as a nomen dubium and Ihave not been able to trace this description. However, Bordage (1898) described the work of thisinsect so the name Trioza litseae Bordage, 1898, becomes available under Article 16 (a)(viii) ofthe Code. The anomalicornis-group Defined by the form of the ovipositor (Fig. 138), otherwise the five included species, anomali-cornis, kakamegae, thibae, tavandula and luvandata, are similar to those of the erytreae-group.Two species are known to develop on Apodytes dimidiata (Icacinaceae). Trioza anomalicornis sp. n. (Figs 17, 135-138) DESCRIPTION. Adult. Integument sparsely covered with long setae. In profile, head moderately depressedfrom longitudinal axis of body, in dorsal view almost as wide as mesoscutum; occipital margins rounded;vertex pentagonal, gently rounded downwards, median suture clearly defined, lateral concavities weak,lateral ocelli on outer sides of strongly raised humps, frontal lobes absent; median ocellus visible fromabove, frons completely covered by genae in anterior view; genal cones elongate conical, slender, withnarrowly rounded apices; antennal flagellum 1-84-2-0 times longer than head width, each flagellomerewith a pair of long setae apically, apart from apical flagellomere which bears one long pointed seta and onelarge campaniform sensillum apically (Fig. 17), a single rhinarium present subapically on flagellomeres 2,4, 6 and 7; clypeus with one pair of setae, ultimate rostral segment densely setose. Thorax strongly arched;pronotum just visible from above, in profile strongly downcurved behind occiput and well below plane ofvertex and mesopraescutum; mesopraescutum, in profile, strongly rounded down to pronotum, in dorsalview its anterior margin narrowly arcuate; forewing elongate oval and narrowing to rounded acutangularapex, 2-88-3-12 times longer than wide, radular areas narrow elongate, remainder of membrane devoid ofspinules; veins bearing very short setae, R branch acutangular,' M branching distal to RsCu^ line, Custem 2-81-3-94 times longer than Cwi b , m t cell value 1-08-1-17, cui cell value 2-14-3-00; forewing1-88-1-98 times longer than hindwing, costal margin of hindwing with 1-2 setae proximal to costal break,setae distal to costal break clearly divided into two groups. Hind coxa with a well-developed meracanthusand without anterior lobe; hind tibia relatively long, 1-49-1-69 times longer than head width, with a pair ofsmall spines basally and one outer and three inner apical spurs, the middle spur of the inner trio being lessdeveloped than the remaining two. Abdomen with setae on tergite 3 in cf , and 4 in $; cf proctiger with agreatly expanded basal part (Fig. 135), para'mere and apical segment of aedeagus as in Figs 136, 137; 9genital segment (Fig. 138) short, truncate, ventral valve of ovipositor with saw-like teeth on ventral andlateral surfaces, posterior margin of subgenital plate deeply incised ventrally. Measurements (1 cf , 3 $). Maximum width of head, cf 0-38, 9 0-35-0-40; length of antennal flagellum,Cf 0-70, 9 0-70-0-76; length of ultimate rostral segment, cf and 9 0-09; length of forewing, cf 2-72, 92-94-3-05; length of hind tibia, cf 0-59, 9 0-59-0-60. Larva and host plant unknown. AFROTROPICAL JUMPING PLANT LICE 45 Holotype cf, Nigeria: Ibadan, Moor plntn, 25-26. xi. 1957, yellow tray (V. F. Eastop) (BMNH; slidemounted). Paratypes. Nigeria: 1 9, vi.1956 (V. F. Eastop); 1 9, v.1959; 1 $, 2.vi.l961 (F. A. Squire). Cameroun:1 9, Bamenda, l.ii.1957, yellow tray (V. F. Eastop). (BMNH; slide and dry mounted.) COMMENT. T. anomalicornis and the following four species described below seem to form anatural group, having a similarly derived ovipositor and a slightly thickened antennal flagellum.In other respects they resemble the erytreae-group. T. anomalicornis is easily distinguished by itsantennal structure, the relatively short hindwing, the elongate hind tibia, the form of the cfproctiger, paramere and apical aedeagal segment, and the deeply emarginate $ subgenitalplate. The remaining four species in the group can only be separated from one another on cfgenital characters. Trioza kakamegaesp. n. (Figs 139-141, 302, 303) DESCRIPTION. Adult. Integument sparsely covered with moderately long setae. In profile, head depressedfrom longitudinal axis of body, in dorsal view as wide as mesoscutum; occipital margins rounded; vertexpentagonal, rounded downwards, without frontal lobes, median suture clearly defined, lateral ocelli onouter sides of raised humps; median ocellus visible from above, frons completely covered by genae inanterior view; genal cones elongate conical, broad, with broadly rounded apices; antennal flagellum2-05-2-47 times longer than head width, each flagellomere with a pair of short setae apically apart fromapical flagellomere which bears a long pointed seta and a short truncate seta, a single rhinarium presentsubapically on flagellomeres 2, 4, 6 and 7; clypeus with a pair of setae, ultimate rostral segment with a pairof setae. Thorax strongly arched; pronotum just visible from above, in profile strongly rounded downbehind occiput and well below plane of vertex and mesopraescutum; mesopraescutum with anteriormargin narrowly arcuate in dorsal view and strongly rounded down to pronotum in lateral view; forewingelongate oval, narrowing to rounded acutangular apex, 2-69-2-86 times longer than wide; Cu stem2-89-3-51 times longer than Cu ib , m l cell value 1-18-1-29, cu^ cell value 1-74-2-48; forewing 1-55-1-65times longer than hindwing, costal margin of hindwing with 0-1 seta proximal to costal break, setae distal tocostal break clearly divided into two groups; hind coxa with a well-developed meracanthus and withoutanterior lobe; hind tibia 1-16-1-36 times longer than head width, with a pair of small basal spines, with oneouter and three inner apical spurs , latter of equal development. Abdomen with setae on tergite 3 in cf and 4in 9 ; basal part of cf proctiger (Fig. 139) weakly expanded, apical segment of aedeagus and paramere as inFigs 140, 141; 9 genital segment short, truncate, posterior margin of subgenital plate truncate. Measurements (6 cf, 5 9). Maximum width of head, cf 0-34-0-42, 9 0-37-0-41; length of antennalflagellum, cf 0-83-0-99, $ 0-82-0-96; length of ultimate rostral segment, cf 0-09-0-10, 9 0-09-0- 11; lengthof forewing, Cf 2-46-2-95, $ 2-78-3-18; length of hind tibia, cf 0-46-0-53, 9 0-45-0-51. Fifth instar larva (Figs 302, 303). Dorsal surface outline broadly oval, about 1-4 times longer than wide.Antenna with four flagellomeres . Cephaloprothorax separate from rest of thorax which is entire . Forewingpad about 0-75 mm long; humeral lobe extending forward in front of anterior margin of eye. Caudal plateabout 0-65 times as long as wide ; anus ventral and distant from posterior margin of abdomen , anus and porering as in Fig. 303. Truncate tubular sectasetae forming an even dense marginal fringe; postocular setaabsent; sectasetae absent from dorsum. HOST PLANT. Larvae and adults swept from Apodytes dimidiata (Icacinaceae). Holotype cf, Kenya: Kakamega For. sta., c. 5,000', 18-19.vii.1974, Apodytes dimidiata (D. Hollis)(BMNH; dry mounted). Paratypes. Kenya: 6 cf, 5 $, larvae, same data as holotype; 1 $, 4.viii.l980 (Min. of Ag.); 3 cf, 6 $,Muguga, vi.1953; 1 cf , 1 9, vii.1954 (V. F. Eastop) (BMNH; slide and dry mounted.) Trioza thibae sp. n. (Figs 142-144) DESCRIPTION. Adult. Very similar to kakamegae. Differs in the shape of the cf proctiger, paramere andaedeagus (Figs 142-144).Measurements (3 cf , 3 9). Maximum width of head, cf 0-40-0-44, 9 0-41-0-44; length of antennal 46 DAVID HOLLIS flagellum, cf 0-94-1-05, 9 0-93-0-98; length of ultimate rostral segment, cf and $ 0-10; length of forewing,Cf 2-89-2-99, 9 3-11-3-28; length of hind tibia, cf 0-49-0-53, 9 0-50-0-52. Fifth instar larva. Very similar to kakamegae. Dorsal outline about 1-3 times longer than wide, forewingpad about 0-8 mm long, caudal plate about 0-6 times as long as wide. HOST PLANT. Larvae and adults swept from Apodytes dimidiata (Icacinaceae). Holotype d", Kenya, S. slope of Mt Kenya, Thiba camp, c. 6,000', 9-10. vii. 1974, Apodytes dimidiata (D.Hollis) (BMNH; slide mounted).Paratypes. 2 d", 4 $ , larvae, same data as holotype (BMNH; slide and dry mounted). Trio/.ii tavandula sp. n. (Figs 145-147) DESCRIPTION. Adult. Very similar to kakamegae and thibae. Integument bearing much shorter hairs.Antennal flagellum 1-90-2-32 times longer than head width. Thorax less strongly arched, pronotum moreclearly visible from above, anterior margin of mesopraescutum less narrowly arcuate. Forewing 2-83-3-16times longer than wide; hind tibia 1-07-1-16 times longer than head width, d" proctiger, paramere andaedeagus as in Figs 145-147; 9 subgenital plate with obtusangular posterior margin. Measurements (6 d", 3 9)- Maximum width of head, d" 0-38-0-40, $ 0-40-0-42; length of antennalflagellum, cf 0-76-0-88, $ 0-83-0-92; length of ultimate rostral segment, d" and $ 0-09-0-10; length offorewing, d" 2-53-2-59, $ 2-77-3-03; length of hind tibia, cf 0-43-0-44, 0-44-0-48. Larva and host plant unknown. Holotype d", Angola: Tundavala, 8-10 mis NW. Sa da Bandeira, 27-29. iii. 1972, general sweeping (D.Hollis) (BMNH; dry mounted).Paratypes. Angola: 4 C?, same data as holotype; 7 cf , 6 $, 5. iii. 1972 (BMNH; slide and dry mounted). Trio/a luvandata sp. n. (Figs 148-150) DESCRIPTION. Adult. Very similar to kakamegae and thibae. Differs in shape of cf paramere and aedeagus(Figs 148-150). Measurements (3 cf , 1 9)- Maximum width of head, d" 0-40-0-41, 9 0-40; length of antennal flagellum,d" (incomplete), $ 0-85; length of ultimate rostral segment, d" 0-08-0-10, 9 0-09; length of forewing, d"2-83-2-86, 9 2-75; length of hind tibia, d 1 0-44-0-48, 9 0-44. Larva and host plant unknown. Holotype d", Angola: Tundavala, 9 mis NW. Sa da Bandeira, 5.iii.l972, swept (D. Hollis) (BMNH; slidemounted).Paratypes. 2 d", 1 9 > same data as holotype (BMNH; slide mounted). The neoboutonia-group Defined by the relatively short 3rd flagellomere and the presence of only two inner apical spurson the hind tibia. Females of this group show a tendency towards a reduction of the outer ring ofwax-producing cells around the anal pore. Six species are included: neoboutonia, harteni,chiangae, bamendae, dinaba and nachingweae. Known host plants are in the Euphorbiaceae andUapacaceae. Trioza neoboutonia sp. n. (Figs 14, 151-155,304,305) DESCRIPTION. Adult. Integument sparsely covered with short setae. Head, in profile, moderately depressedfrom longitudinal axis of body, in dorsal view narrower than mesoscutum; occipital margin subangulardorsomedially, rounded dorsolaterally and laterally; vertex pentangular, with an irregular concavity oneither side of median suture, latter more strongly developed anteriorly, frontal lobes weakly developed;median ocellus just visible from above; frons completely covered by genae in anterior view; genal conesshort, conical, in lateral view their longitudinal axis slightly depressed from plane of vertex; clypeus with AFROTROPICAL JUMPING PLANT LICE 47 one pair of setae, ultimate rostral segment with two pairs of setae; antennal flagellum 1-95-2-12 timeslonger than head width, 3rd flagellomere very short, ratio of length of 2nd flagellomere to 3rd 1 -76-2-35, asingle rhinarium is present subapically on flagellomeres 2, 4, 6 and 7, apical flagellomere bearing one longpointed seta and one short truncate seta apically. Thorax moderately arched; pronotum clearly visible fromabove, its anterior margin strongly rounded down behind occiput; mesopraescutum about as long as wide,its anterior margin rounded; forewing elongate oval, narrowing to a rounded obtusangular apex, 2-7-2-91times longer than wide; radular areas narrow elongate triangular, remainder of membrane devoid ofspinules apart from a short narrow band at base of claval suture; veins bearing short setae, R branchacutangular, M branching distal to Rs-Cu^ line, Cu stem 1-63-1-84 times longer than C lb , m t cell value1-13-1-23, cui cell value 1-61-1-91; forewing 1-5-1-62 times longer than hindwing, costal margin ofhind wing with 2-4 straight setae proximal to costal break, setae immediately distal to costal break absent,2-3 curved setae present immediately proximal to retinaculum; hind coxa with a well-developed meracan-thus, without anterior lobe; hind tibia relatively long, 1-37-1-48 times longer than head width, with awell-developed basal spine, with one outer and two inner spurs apically. Abdomen with setae on tergite 3 inCf and tergite 4 in $ ; (3" proctiger, paramere and aedeagus as in Figs 152-154; 9 genital segment (Fig. 155)short, triangular in profile, ovipositor valves smooth, apex of proctiger upcurved, subgenital plate weaklyserrate apicoventrally, anus with a single ring of wax-producing cells. Measurements (5 cf, 7 $). Maximum width of head, cf 0-40-0-42, 9 0-42-0-44; length of antennalflagellum, cf 0-83-0-87, $ 0-82-0-88; length of ultimate rostral segment, cf and 9 0-09-0-10; length offorewing, cf 2-53-2-75, 2-75-2-96; length of hind tibia, cf 0-56-0-62, $ 0-59-0-63. Fifth instar larva (Figs 304, 305). Dorsal surface outline oval, about 1-5 times longer than wide. Antennawith 5-6 flagellomeres. Cephaloprothorax separate from rest of thorax which is entire. Forewing pad about0-65 mm long, humeral lobe extending forward to just behind anterior margin of eye. Caudal plate about0-65 times as long as wide, anus ventral and distant from posterior margin of abdomen, anal pore area as inFig. 305. Short, truncate, tubular sectasetae forming an even, dense marginal fringe; postocular setaabsent; sectasetae absent from dorsum. HOST PLANT. Larvae, which form pit galls on the lower surfaces of the leaves, and adults sweptfrom Neoboutonia sp. (Euphorbiaceae). Holotype cf , Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27. vi. 1974, Neoboutonia sp.(D. Hollis) (BMNH; dry mounted). Paratypes. Tanzania: 10 cf , 8 9, larvae, same data as holotype; 1 9, yellow tray (BMNH; slide and drymounted). COMMENTS. Within this species-group neoboutonia displays the most derived condition of headand thorax structure, wing venation and abdominal chaetotaxy. Of the three known larvae in thegroup, that of neoboutonia is the most highly evolved and adapted to living in a pit gall. Trioza hartenisp. n. (Figs 156-158, 306, 307) DESCRIPTION (only slide-mounted material available for study). Adult. Integument sparsely covered withshort setae. Head, from above, narrower than mesoscutum; frons not visible in anterior view; genae short,broadly rounded; antennal flagellum 2-07-2-72 times longer than head width, 3rd flagellomere short, ratioof length of 2nd flagellomere to 3rd 1-45-1-97, a single rhinarium present subapically on flagellomeres 2, 4,6, and 7, apical flagellomere bearing one long pointed seta and one short truncate seta apically; clypeuswith one pair of setae, ultimate rostral segment with one pair of setae. Pronotum strongly rounded downbehind occiput; mesopraescutum, from above, wider than long, its anterior margin broadly arcuate;forewing elongate oval, strongly narrowing to subangular apex, 2-62-2-84 times longer than wide, radularareas narrow elongate, remainder of membrane devoid of spinules; veins bearing short setae, R branchacutangular, M branching distal to Rs-Cu ia line, Cu stem 1-13-1-48 times longer than Cw lb , ra t cell value1-37-1-46, CM! cell value 1-48-1-91; forewing 1-75-1-86 times longer than hindwing, costal margin ofhindwing with up to four straight setae proximal to costal break, setae distal to costal break clearly dividedinto two groups; hind coxa with a well-developed meracanthus and without anterior lobe; hind tibiarelatively short, 1-19-1-25 times longer than head width, with a group of small spines basally, with 0-1outer and 1-2 (rarely 3) inner spurs apically. Abdomen with setae on tergites 3-8 in cf and 4-8 in 9; Cfproctiger with a moderately swollen basal part and a very short and narrow apical part, paramere as in Fig. 48 DAVID HOLLIS 157, apical part of aedeagus simple; 9 genital segment as in Fig. 158, ventral surface of ventral valve ofovipositor weakly serrate, outer ring of wax-producing cells around anus incomplete. Measurements (2 cf , 3 $). Maximum width of head, cf 0-57-0-59, 9 0-59-0-62; length of antennalflagellum, cf 1-48-1-55, 9 1-22-1-40; length of ultimate rostral segment, cf 0-10, 9 0-10-0-11; length offorewing, cf 3-99-4-07; 9 4-63^-73; length of hind tibia, cf 0-71-0-72, 9 0-70-0-75. Fifth instar larva (Figs 306, 307). Dorsal surface outline showing head, thorax and abdomen clearlydefined, about 1-6 times longer than wide. Antenna with seven flagellomeres. Cephaloprothorax,mesothorax and metathorax clearly defined. Forewing pad about 1-3 mm long, humeral lobe very weaklydeveloped and anterior margin of forewing pad reaching barely anterior to posterior margin of cephalo-prothorax. Basal two-thirds of abdomen not sclerotised, caudal plate about half as long as wide. Anusventral but close to posterior margin of abdomen, anal pore area as in Fig. 307. Marginal setae simple withsparsely scattered pointed sectasetae, anterior margin of Cephaloprothorax between eyes without secta-setae but with scattered lanceolate setae, four postocular pointed sectasetae present, sectasetae absentfrom dorsum. HOST PLANT. Larvae taken from large, enclosed galls on Uapaca nitida (Uapacaceae); adultsreared from these larvae (van Harten, pers. comm.). Holotype cf , Angola: Chinaga, 19.x. 1970, galls of Uapaca nitida (A. van Harten) (BMNH; slide mounted).Paratypes. 1 cf , 3 $, larvae, same data as holotype (BMNH; slide mounted). COMMENT. T. harteni displays primitive character states for the Cu stem and, therefore, a low cuicell value, and abdominal chaetotaxy. The larva is very non-triozid in general appearance butthis is almost certainly due to the fact that it develops in an enclosed gall. Trioza chiangae sp. n. (Figs 159-161, 308) DESCRIPTION (only slide-mounted material available for study). Adult. Very similar to harteni but smaller.Clypeus with a pair of short setae, ultimate rostral segment without setae. Forewing more elongate,2-87-3-18 times longer than wide, Cu stem 2-11-2-52 times longer than C lb , mi cell value 1-41-1-65, cu\cell value 2-0-2-82; hind tibia relatively short, 1-07-1-21 times longer than head width. Abdomen withsetae on tergites 3-5 in cf and 9 ; cf paramere as in Fig. 160; 9 genital segment (Fig. 161) with ovipositorvalves smooth apically, anal pore with a double ring of wax-producing cells. Measurements (6 cf , 6 $). Maximum width of head, cf 0-37-0-39, $ 0-40-0-44; length of antennalflagellum, cf 0-86-0-93, $ 0-85-0-92; length of ultimate rostral segment, cf and $ 0-06-0-07; length offorewing, cf 2-41-2-56, 2-93-3-08; length of hind tibia, cf 0-41-0-46, $ 0-45-0-48. Fifth instar larva (Fig. 308). Very similar to harteni. About 1-8 times longer than wide; forewing padabout 0-7 mm long; caudal plate not at all developed; anal pore area damaged in specimens available butsimilar in shape and structure to harteni. Long and short, narrow conical sectasetae with truncate apicesforming a sparse marginal fringe, including anterior margin of Cephaloprothorax; five postocular sec-tasetae present; sectasetae absent from dorsum. HOST PLANT. Larvae and adults collected from Uapaca nitida (Uapacaceae); the larvae arefree-living among unfolded leaves on the growing points (A. van Harten, pers. comm.). Holotype cf , Angola: Chianga, 12. ix. 1973, Uapaca nitida (A. van Harten) (BMNH; slide mounted).Paratypes. 5 cf , 6 9, larvae, same data as holotype (BMNH; slide mounted). COMMENT. T. chiangae has a relatively long Cu stem and, therefore, a relatively higher cui cellvalue than harteni, but the chaetotaxy of the abdominal tergites is intermediate between harteniand neoboutonia. The free-living larva is relatively unmodified for a triozid. Trioza bamendae sp. n. (Figs 162-164) DESCRIPTION. Adult. Integument very sparsely covered with short setae. Head, in profile, very slightlydepressed from longitudinal axis of body, from above at least as wide as mesoscutum; occipital margindeeply notched medially; vertex in same plane as thorax, median suture shallow, with a secondary furrowon either side originating from a point just above median ocellus and extending back diagonally to occipital AFROTROPICAL JUMPING PLANT LICE 49 margin, frontal lobes well developed, their anterior margins arcuate; median ocellus visible from above,frons not visible in anterior view; genal cones well developed, broadly conoid, in profile slightly depressedfrom plane of vertex; antennal flagellum 1-22-1-40 times longer than head width, ratio of length of 2ndflagellomere to 3rd 2-18-3-39, a double rhinarium present subapically on 2nd flagellomere and a singlerhinarium present subapically on flagellomeres 4,6, and 7, apical flagellomere with one long seta and a veryshort truncate seta apically; clypeus with a pair of setae, ultimate rostral segment with a pair of setae.Thorax flattened; pronotum mostly flat with only anterior margin sharply downcurved behind occiput;mesopraescutum, from above, longer than wide, with broadly rounded anterior margin; forewing elongateelipsoid, strongly narrowing distally to rounded acutangular apex, 3-12-3-50 times longer than wide,membrane devoid of spinules apart from narrow radular areas; veins almost devoid of setae, R branchacutangular, M branching distal to RsCui a line, Cu stem 1-42-2-58 times longer than Cw lb , mi cell value1-00-1-27, cu\ cell value 2-03-3-22; forewing 1-73-1-93 times longer than hindwing, costal margin ofhindwing with up to two setae proximal to costal break and with at most one seta distal to costal break; hindcoxa with a well-developed meracanthus and without anterior lobe; hind tibia with 1-2 small conoid spinesbasally, and one outer and two inner apical spurs. Abdomen with setae on tergite 3 in cf and 4 in $; cfproctiger, paramere and aedeagus as in Figs 162-164; $ genital segment short, conoid, anus with a singlering of wax-producing cells, ventral valve of ovipositor with three transverse ridges on ventrolateralsurface. Measurements (6 cf , 7 $). Maximum width of head, cf 0-30-0-34, $ 0-31-0-34; length of antennalflagellum, cf 0-39-0-44, $ 0-39-0-46; length of ultimate rostral segment, cf 0-05, $ 0-05-0-06; length offorewing, cf 1-84-2-08, $ 2-12-2-35; length of hind tibia, cf 0-45-0-50, $ 0-43-0-50. Larva and host plant unknown. Holotype cf , Cameroun: Bamenda, 20-24. i. 1957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: 3 cf , 7 $ , same data as holotype; 2 cf , 2 $ , 25-31.i.l957; 1 cf , 1 $ , l.ii.1957; 1 $ ,6.H.1957 (BMNH; slide and dry mounted). COMMENT. T. bamendae and the following species are regarded as sister-species as both share theunique character of the double rhinarium on the 2nd flagellomere (Fig. 25). Trio/.u dinaba sp. n. (Figs 16, 25, 165-168) DESCRIPTION. Adult. Very similar to bamendae. Secondary diagonal furrows on vertex much less welldefined, frontal lobes broad; antennal flagellum 1-22-1-59 times longer than head width, ratio of length of2nd flagellomere to 3rd 1-61-3-5, 2nd flagellomere as in Fig. 25. Forewing 2-93-3-38 times longer thanwide; Cu stem branching from R+M+Cu proximally to branch of R stem, 2-81-4-67 times longer thanCw lb , m\ cell value 1-06-1-29, cui cell value 2-31-4-0. cf proctiger, paramere and aedeagus as in Figs166-168; anus of 9 with an incomplete double ring of wax-producing cells. Measurements (10 cf , 10 $). Maximum width of head, cf 0-32-0-34, $ 0-32-0-36; length of antennalflagellum, cf 0-46-0-51, $ 0-43-0-50; length of ultimate rostral segment, cf and $ 0-06-0-07; length offorewing, cf 2-08-2-32, $ 2-30-2-70; length of hind tibia, cf 0-45-0-50, $ 0-44-0-51. Larva and host plant unknown. Holotype cf, Nigeria: Ibadan, Moor plntn, ll.viii.1956, yellow tray (V. F. Eastop) (BMNH; slidemounted). Paratypes. Nigeria: 9 cf, 11 $, same data as holotype; 1 cf, vi.1956; 1 cf, Umuahia, Umudike,28.viii.1958. Ghana: 2 cT, 2 $, Tafo, 15.V.1956, yellow tray (V. F. Eastop). Zaire ['Congo Beige']: 1 $,P.N.G., Miss, de Saeger, II/fd/5, 10.ix.1951; 1 cf II/fd/17, 9.vii.l952 (H. de Saeger). (BMNH; MRAC;slide and dry mounted.) Non-paratypic material. Uganda: 1 $ , Jinja, xii.1954-ii.1955, U.V. lamp (P. S. Corbetf) (BMNH); slidemounted. This specimen differs from the type-series in having a triple rhinarium present apically on the 2ndflagellomere. Trioza nachingweae sp. n. (Figs 169-172) DESCRIPTION (only slide-mounted material available for study). Adult. Very similar in general appearancetofuscivena (p. 61) but smaller. Antennal flagellum shorter, 1-03-1-40 times longer than head width, ratio 50 DAVID HOLLIS of length of 2nd flagellomere to 3rd 1-82-2-67; ultimate rostral segment relatively longer. Forewing(Fig. 169) 2-74-3- 12 times longer than wide, veins sparsely clothed with very short setae, Custem 1-59-2-65times longer than Cw lb , mi cell value 1-10-1-28, cu { cell value 1-59-2-50; forewing 1-54-1-58 times longerthan hindwing; hind tibia with one outer and two inner apical spurs, of which the second inner spur is weakand difficult to see. cf proctiger narrow, tubular, paramere and aedeagus as in Figs 170, 171; $ genitalsegment (Fig. 172) elongate triangular in profile, anus with a single ring of wax-producing cells, subgenitalplate with acutangular posterior margin. Measurements (6 cT, 3 $). Maximum width of head, cf 0-34-0-36, $ 0-37-0-40; length of antennalflagellum, cf 0-42-0-49, $ 0-38; length of ultimate rostral segment, cf 0-08-0-09, $ 0-09-0-10; length offorewing, cf 1-78-1-92, $ 2-18-2-30; length of hind tibia, cf 0-47-0-50, $ 0-51-0-53. Larva and host plant unknown. Holotype cf , Tanzania: Nachingwea, xi.1953-iii.1954, trapped (V. F. Eastop) (BMNH; slide mounted). Paratypes. Tanzania: 6 cT, 3 $, same data as holotype; 1 Cf, 1 $, x.l953-ii.!954. (BMNH; slidemounted.) COMMENTS. T. nachingeae is distinguished from other members of the neuboutonia-group by thepresence of brown pigmentation along the course of R and RI and at the bases of M and Cu onthe forewing. This feature has apparently arisen independently in many groups of triozids. The Aargreavesi-group Defined by the extraordinary development of supplementary antennal rhinaria which haveassociated bulbous or mushroom-shaped sensilla (Figs 22, 23). Two species, hargreavesi andmirificornis , are included. Larvae and host plants are not known but original collection datasuggest that the larvae of both species form pit galls on a forest climbing plant. Trioza hargreavesi sp. n. (Figs 22, 173-175) DESCRIPTION. Adult. Integument densely covered with short setae. Head, in profile, almost at 90 tolongitudinal axis of body, in dorsal view slightly narrower than mesoscutum; occipital margin sharpdorsomedially; vertex pentagonal, rounded downwards, with a deep irregular concavity on either side ofmedian suture which is deep and well defined, frontal lobes not developed; median ocellus visible fromabove, frons just visible between genae in anterior view; genal cones small, rounded; antennal flagellumapproximately 1-8 times longer than head width (no complete antennae in type-series), with a singlerhinarium subapically on flagellomeres 2, 4, 6 and 7, 1st flagellomere with 30-35 rhinaria along its length(Fig. 22), each bearing a small mushroom-shaped sensillum; clypeus with a pair of setae, ultimate rostralsegment with at least four pairs of setae. Thorax, in profile, strongly arched; pronotum, from above, hardlyvisible, in profile strongly rounded down behind occiput; mesopraescutum, from above, longer than wide,its anterior margin strongly arcuate and, in profile, strongly angled down to pronotum; forewing elongateoval, narrowing to rounded obtusangular apex, 2-36-2-45 times longer than wide, radular areas narrow andelongate, remainder of membrane devoid of spinules; veins bearing short setae, R branch acutangular, Mbranching distal to RsCui a line, Cu stem 1-24-1-72 times longer than Cw lb , m\ cell value 1-32-1 -6, cu^ cellvalue 1-78-1-98; forewing 1-66-1-77 times longer than hindwing, costal margin of hindwing with 0-2 setaeproximal to costal break, setae distal to costal break clearly divided into two groups; hind coxa with awell-developed meracanthus and without anterior lobe; hind tibia with 2-4 small spines basally and withone outer and two inner apical spurs. Abdomen with setae on tergite 3 or 3 and 4 in cf , and on 4 or 4 and 5 in$ ; Cf proctiger with weak lateral expansions, paramere and aedeagus as in Figs 174, 175; $ genital segmentshort, conical, ventral valves of ovipositor regularly saw-toothed on ventral and lateral surfaces in apicalhalf, subgenital plate short and with truncate posterior margin. Measurements (5 cf , 5 $). Maximum width of head, cf 0-65-0-74, $ 0-69-0-76; length of antennalflagellum, $ 1-25 (approx.); length of ultimate rostral segment, cf and $ 0-13-0- 16; length of forewing, cf4-45-5-02, $ 5-02-5-42; length of hind tibia, cf 0-72-0-81, $ 0-81-0-84. Larva and host plant unknown. (A label in the original tube containing the type-series from Ugandastated 'nymphs in pits in leaf of forest climber' but no larvae were present.) Holotype cf , Uganda: F. Kawanda, 14. ii. 1940, nymphs in pits in leaf of forest climber (H. Hargreaves)(BMNH; slide mounted). AFROTROPICAL JUMPING PLANT LICE 51 Paratypes. Uganda: 8 cf , 6 9, same data as holotype. Nigeria: 1 $, NE. State, Ngel Nyaki, 31.iii.1970,5,000'; 1 $, K[wara] State, 20 mis W. Lokaja, 21.iii.1970; 1 $, SE. State, Obudu CR, 21.iii.1971 (/. T.Medler). (BMNH; slide and dry mounted, and stored in 80% ethanol.) COMMENTS. T. hargreavesi and mirificornis are grouped together as sister-species because theyboth share the extraordinary development of multiple supplementary rhinaria on the antennalflagellum. At present I can find no further derived characters relating this group with otherspecies in the genus. Trioza mirificornis sp. n. (Figs 23, 176, 177) DESCRIPTION. Adult. Very similar to hargreavesi but smaller. Flagellomeres 1-7 bearing many rhinaria,each rhinarium bearing a bulbous sensillum (Fig. 23), 1st flagellomere with about 60 rhinaria, 2nd with18-25, 3rd with 7-15, 4th with 5-20, 5th with about 11, 6th with about seven, and 7th with one, apicalflagellomere with a long pointed seta and a short truncate seta apically; ultimate rostral segment with a pairof setae. Pronotum more clearly visible from above, anterior margin of mesopraescutum less stronglyarcuate and less pushed forward, rounded down to pronotum; forewing narrower, 2-51-2-7 times longerthan wide, Cu stem 1-70-2-32 times longer than Ci b , mi cell value 1-13-1-29, cu\ cell value 1-75-2-36;forewing 1-56-1-64 times longer than hindwing. cf paramere and aedeagus as in Figs 176, 177; $ subgenitalplate with acutangular posterior margin, ventral valves of ovipositor smooth. Measurements (7 cf , 4 $). Maximum width of head, cf 0-47-0-55, $ 0-51-0-60; length of antennalflagellum, cf 0-84 (only one complete specimen); length of ultimate rostral segment, cf and $ 0-08-0-10;length of forewing, cf 3-21-3-83, $ 4-17^-53; length of hind tibia, cf 0-54-0-65, $ 0-66-0-67. Larva and host plant unknown. (The Uganda type-material of this species was found in the BMNHbearing similar data relating to larvae as the previous species.) Holotype cf , Uganda: F. Kawanda, 24.x. 1939, nymphs in pits in leaf of forest climber (H. Hargreaves)(BMNH; slide mounted). Paratypes. Uganda: 16 cf , 25 9 , same data as holotype. Cameroun: 1 9 , Bamenda, 6.ii.l957, yellow tray(V. F. Eastop). (BMNH; slide mounted and stored in 80% ethanol.) The obsoleta-group A poorly defined group in which the species usually have greatly reduced hindwings, well-developed anterior lobes on the hind coxae and two inner apical spurs on the hind tibia. FiveAfrotropical species are included: afrobsoleta, gonjae, boxi, afrosersalisia and mimusops, thelatter two only tentatively. All five species have host plants in the Sapotinae (Ebenaceae andSapotaceae). Another Afrotropical member of this group is represented in BMNH by fivespecimens, which are too badly damaged for description, standing under a Heslop-Harrisonmanuscript name. A North American species, T. diospyri (Ashmead, 1881), is also known to develop onDiospyros. Although this species shows a general resemblance to those of the obsoleta-gToupand in particular to gonjae and boxi, its hindwings are more normally developed. It mayrepresent the sister-species of the Old World group. Trioza obsoleta (Buckton, 1900), described from India, and Megatrioza swezeyi Crawford,1927, described from Samoa, clearly belong to this group; the South East Asian species Triozaasiatica Crawford, 1915 and Megatrioza magnicauda Crawford, 1919 probably belong here, but Ihave no material for examination. Trioza diptera Crawford, 1919, from Singapore, also hascompletely reduced hindwings but, from its description, this species appears to be more closelyrelated to Leptynoptera sulfurea Crawford and both develop on Calophyllum inophyllum(Guttiferae). Earlier authors almost certainly would have placed all these species in Megatrioza but as Ihave not examined the type-species, M. armata Crawford, I am not sure of the status of thegenus. 52 DAVID HOLLIS Trioza afrobsoleta sp. n. (Figs 13, 178-182)Tsyllidengalle'; Diospyros mespiliformis; Riibsaamen, 1899: 271 (Eritrea [series not traced].) DESCRIPTION. Adult. Integument shiny, sparsely covered with short setae. Head, in profile, stronglydepressed from longitudinal axis of body, in dorsal view as wide as mesoscutum; occipital margin sharp;vertex pentagonal, rounded downwards, lateral concavities displaced posteriorly to occipital margin andgreatly reduced, median suture weak; median ocellus hardly visible in dorsal view, frons small but visiblebetween bases of genal cones in anterior view; genal cones in parallel plane to vertex, short, broadlyrounded apically; antennal flagellum 1-66-1-94 times longer than head width in cf and 1-44-1-77 timeslonger in $ , a single rhinarium present subapically on flagellomeres 2, 4, 6 and 7, apical flagellomere withtwo long setae apically ; clypeus with a pair of setae , ultimate rostral segment with a pair of setae . Thorax , inprofile, moderately arched; pronotum hardly visible from above, strongly downcurved behind occiput;mesopraescutum, in profile, strongly rounded down to pronotum, in dorsal view with arcuate anteriormargin; forewing elongate oval, narrowing to a rounded obtusangular apex, 2-46-2-87 times longer thanwide, radular areas very narrow and attenuate, remainder of membrane devoid of spinules, veins bearingshort setae, R branch obtusangular, M branching at or slightly distal to Rs C la line, Cu stem 2-0-2-7times longer than Cw lb , m t cell value 1-97-2-41, cu^ cell value 1 -42-2-05, claval suture reaching hind marginof wing very close to wing base; hindwing (Fig. 179) greatly reduced, scale-like, costal margin bearing oneseta proximal to costal break, setae distal to costal break clearly divided into two groups; hind coxa with awell-developed meracanthus and a moderately long anterior lobe; hind tibia with a group of 3-4 smallpointed tubercles basally, with one outer and two inner apical spurs. Abdomen with setae on tergites 3-7 inO" and 4-8 in $ ; cf proctiger, paramere and aedeagus as in Figs 180-182; $ genital segment short, conical,ovipositor valves smooth. Measurements (16 cf, 10 $). Maximum width of head, cf 0-42-0-56, $ 0-48-0-54; length of antennalflagellum, cf 0-74-0-93, $ 0-75-0-88; length of ultimate rostral segment, cf and 9 0-09-0-15; length offorewing, cf 2-75-3-79, $ 3-01-3-93; length of hind tibia, cf 0-59-0-78, $ 0-65-0-79. Fifth instar larva. Dorsal surface outline almost circular, about 1 1 times longer than wide. Antenna withsix flagellomeres. Cephaloprothorax incompletely separated from rest of thorax which is entire. Forewingpad about 0-9 mm long, humeral lobe extending well forward of anterior margin of eye, hindwing padgreatly reduced and without sectasetae on lateral margin. Caudal plate about twice as wide as long, anusventral and distant from posterior margin of abdomen. Truncate tubular sectasetae forming an even, densemarginal fringe, marginal sectasetae on cephaloprothorax longer than those on forewing pad andabdomen, postocular seta absent, sectasetae absent from dorsum. HOST PLANT. Larvae, which form pit galls on the lower leaf surface, and adults collected fromDiospyros mespiliformis (Ebenaceae). Holotype cf , Angola: Rocadas, R. Cunene, 19-22. ii. 1972, Diospyros mespiliformis (D. Hollis) (BMNH;dry mounted). Paratypes. Angola: 11 cf, 18 <j>, same data as holotype; 17 cf, 12 $, 3 mis N. Santa Clara,30.iii.-l.iv.1972; 1 $, Bruco, 26-29. ii. 1972, swept in riverine forest area (D. Hollis). Tanzania: 9 cf , 5 $,Nachingwea, x.l953-iii.!954, trapped (V. F. Eastop). Sudan: larvae, SW. side Jebel Marra, Wadi Golol,l.v.1981, Diospyros mespiliformis; 1 cf , swept, 28.iv.1981 (J. H. Martin). Nigeria: 1 cf , Bunga, 20.X.1956,Salix ledermanni (V. F. Eastop); 2 cf, Ibadan, 4.vi.l961, yellow tray; 1 cf, 2 $, 5-19.iv.1963; 1 cf,2.V.1963; 5 cf, 8 $, 3-9.iv.1964 (F. A. Squire); 1 $, B.P. State, Gindiri, 28.X.1968 (J. T. Medler); 1 cf,Zaria, Samaru, 24.xi.1970 (/. C. Deeming); 1 $, Lagos, Ikoyi, 7.iii.l975, at light (M. A. Comes). Ghana:1 $,Tafo,29.v.l957(V. F. Eastop). Ivory Coast: 2 cf, 1 ?, lO.iv. 1969 (A. Pollet). (BMNH; MNHN; slideand dry mounted and stored in 80% ethanol.) COMMENTS. This species is very closely related to T. obsoleta (Buckton, 1900) which is knownfrom India and Ceylon on Diospyros melanoxylon and D. tomentosa. The Indian species differsin having a longer and slightly sinuous Rs, and the paramere has two sclerotised teeth apicallyinstead of the transverse ridge as found in the Africa species. This latter point of difference,although slight, appears very consistent in all specimens examined. T. afrobsoleta showsconsiderable variation in the length of the ultimate rostral segment through its geographicalrange, that of specimens from Angola and Tanzania being relatively longer than in specimensfrom Sudan and West Africa. AFROTROPICAL JUMPING PLANT LICE 53 Trioza gonjae sp. n. (Figs 38, 47, 183-187, 309, 310) DESCRIPTION. Adult. Integument sparsely covered with long setae (Fig. 38). Longitudinal axis of head andbody in one plane; head, from above, almost as wide as mesoscutum; occipital margin sharp; vertexpentagonal, with an irregular concavity on either side of a well-defined median suture; median ocellusvisible from above, frons not visible between genae; genal cones well developed, elongate conical withrounded apices, densely setose, in profile their longitudinal axis is slightly inclined upwards from that ofvertex, in dorsal view cones slightly convergent apically; antennal flagellum 1-87-2-00 times longer thanhead width in cf and 1-68-1 -78 times longer in $ , a single rhinarium present subapically on flagellomeres 2,4, 6, and 7, apical flagellomere with a long pointed seta and a short truncate seta apically; clypeusprominent, with a pair of setae, ultimate rostral segment with two pairs of setae. Thorax, in profile, weaklyarched; pronotum narrow and inclined vertically down behind occiput, just visible in dorsal view;mesopraescutum, from above, longer than wide and with a subangular anterior margin, in profile anteriormargin angularly bent down to pronotum; forewing elongate oval, strongly narrowing to acutangular apex,2-98-3-29 times longer than wide, radular areas narrow, remainder of membrane devoid of spinules; veinsbearing a few long setae in proximal third of wing, R branch a right-angle, M branching proximal toRs-Cu^ line, Cu stem 2-36-3-21 times longer than Cw lb , raj cell value 1-91-2-08, ci cell value 1-96-2-63;hindwing greatly reduced and scale-like (Fig. 184); hind coxa (Fig. 47) with a well-developed meracanthusand a well-developed anterior lobe; hind tibia with a group of small pointed tubercles basally and one outerand two or three inner spurs apically. Abdomen with setae on tergites 3-7 in cf and 4 and 6-8 in $; cfproctiger, paramere and aedeagus as in Figs 185-187; $ genital segment conical, dorsal surface of dorsalvalve of ovipositor finely serrate apically. Measurements (5 cf, 4 $). Maximum width of head, cf 0-46-0-48, $ 0-47-0-51; length of antennalflagellum, cf 0-88-0-94, $ 0-81-0-91; length of ultimate rostral segment, cf 0-12-0-13, $ 0-12-0-15; lengthof forewing, cf 3-73-3-89, $ 4-01^-37; length of hind tibia, cf 0-70-0-77, $ 0-71-0-78. Fifth instar larva (Figs 309, 310). Dorsal surface outline elongate oval with clear indentations at anteriormargin of eye and at base of abdomen, about 1-7 times longer than wide. Antenna 3-segmented (flagellumnot divided). Cephaloprothorax separate from rest of thorax which is entire but does show partialseparation of prothoracic sclerite (Fig. 309). Forewing pad about 0-9 mm long, humeral lobe extendingalmost to anterior margin of eye. Caudal plate about 0-75 times as long as wide, anus ventral and close toposterior margin of abdomen, anal pore area as in Fig. 310. Truncate tubular sectasetae of varying lengthsforming a dense marginal fringe, a pointed postocular sectaseta or lanceolate seta present, dorsum bearinga sparse covering of tubular sectasetae. HOST PLANT. Larvae and adults swept from Diospyros squarrosus (Ebenaceae) ; no galling of thehost plant was observed and the larvae are apparently free-living. Holotype cf , Tanzania: S. Pare Mtns, hillside above Gonja, c. 3,000', 12-16. vi. 1974, Diospyros squarro-sus (D. Hollis} (BMNH; dry mounted). Paratypes. Tanzania: 11 cf, 10 $, larvae, same data as holotype; 1 $, ArushaNP,Ngurdoto Crater rim,c. 5,000', 8.vi.l974 (D. Hollis}. (BMNH; slide and dry mounted.) Trioza boxisp. n. (Figs 188, 189,311,312) DESCRIPTION. Adult. Similar to gonjae. Differs in that vertex has a longitudinal concavity on either side ofmedian suture; genal cones subconical with broadly rounded apices which are not convergent; ultimaterostral segment shorter and with only one pair of setae. Forewing 2-95-2-99 times longer than wide;proximal part of c+sc and cu 2 cells with brown pigmentation and membrane with a few scattered spinules inthis area; Cu stem 2-45-2-52 times longer than C lb , ra t cell value 1-72-1-74, cu { cell value 2-9-3-0;hindwing (Fig. 189) greatly reduced and scale-like, cf proctiger less extended laterally; dorsal surface ofdorsal valve of ovipositor smooth. Measurements (1 cf, 1 $). Maximum width of head, cf and $ 0-58; length of antennal flagellum, $ 1-03;length of ultimate rostral segment, cf and $ 0- 1; length of fore wing, cf 5-29, $ 5 -35; length of hind tibia, cf0-88,90-85. Fifth instar larva (Figs 311, 312). Very similar to gonjae. Dorsal surface outline less elongate, about 1-5times longer than wide. Forewing pad about 1-5 mm long, humeral lobe extending forward just in front ofanterior margin of eye. Caudal plate about 0-7 times as long as wide, anal pore area as in Fig. 312. Truncate 54 DAVID HOLLIS tubular sectasetae forming an even, dense marginal fringe, lanceolate postocular seta present, dorsalsurface devoid of sectasetae. HOST PLANT. Adults and larvae, collected on separate occasions, from Diospyros canaliculata[=xanthochlamys] (Ebenaceae). Holotype cf, Ghana: 'Gold Coast, Bunsu, 7.vii.l943, Diospyros xanthochlamys" (H. E. Box) (BMNH;slide mounted).Paratypes. Ghana: 2 9> same data as holotype; larvae, 15. i. 1943. (BMNH; slide and dry mounted.) Trioza afrosersalisia sp. n. (Figs 190-193, 313, 314) DESCRIPTION. Adult. Integument sparsely covered with long setae. Head, in profile, slightly depressedfrom longitudinal axis of body, from above slightly narrower than mesoscutum; occipital margin sharp;vertex oval, evenly and smoothly concave and surrounded by a sharp, finely serrate ridge, integumentwithin this depression shiny and devoid of setae, median suture just indicated; median ocellus just visiblefrom above, frons completely covered by genae; genal cones moderately developed, conical, extendingforward in parallel plane to vertex; antennal flagellum 1-75-2-00 times longer than head width, a singlerhinarium present subapically on flagellomeres 2, 4, 6, and 7, apical flagellomere with one short truncateseta and one very short truncate seta apically; clypeus with a pair of setae, ultimate rostral segment withtwo pairs of setae. Thorax, in profile, weakly arched; pronotum hardly visible from above, in profilestrongly rounded down behind occiput; mesopraescutum, in dorsal view, as long as wide and with anarrowly arcuate anterior margin, in profile strongly angled down to pronotum; forewing elongate oval,strongly narrowing to angular apex, 2-72-2-93 times longer than wide, radular areas very narrow,remainder of membrane devoid of spinules; veins bearing long setae in proximal half of wing, R branchacutangular, M branching at Rs C la line, Cu stem 3-29-4-38 times longer than Cw lb , m\ cell value1-77-2-07, cu\ cell value 2-47-3-38; forewing 2-14-2-32 times longer than hindwing, costal margin ofhindwing with 4-5 setae proximal to costal break, setae distal to costal break clearly divided into twogroups; hind coxa with a well-developed meracanthus and an incipient anterior lobe; hind tibia without orwith a very small basal spine, with one outer and two inner apical spurs. Abdominal tergites without setae;Cf proctiger, paramere and aedeagus as in Figs 191-193; 9 genital segment short, conical, ovipositor valvessmooth. Measurements (7 cf , 4 $). Maximum width of head, cf 0-48-0-52, $ 0-49-0-54; length of antennalflagellum, cf 0-93-1-04, $ 0-86-1-01; length of ultimate rostral segment, cf 0-08-0-10, $ 0-10; length offorewing, Cf 3-40-3-71, $ 3-61^4-13; length of hind tibia, cf 0-70-0-73, $ 0-66-0-73. Fifth instar larvae (Figs 313, 314). Dorsal surface outline oval, about 1-4 times longer than wide.Antenna with five flagellomeres. Cephaloprothorax separate from rest of thorax which is entire. Forewingpad about 1 -0 mm long , humeral lobe extending well forward of anterior margin of eye . Caudal plate about0-6 times as long as wide, anus ventral and distant from posterior margin of abdomen, anal pore area as inFig. 314. Truncate tubular sectasetae forming an even, dense marginal fringe, postocular seta absent,sectasetae absent from dorsum. HOST PLANT. Larvae and adults collected from Afrosersalisia sp. (Sapotaceae) ; galling of the hostplant was not noticed at the time the series, including the holotype, was collected. Holotype cf , Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27. vi. 1974, Afrosersalisia sp.(D. Hollis) (BMNH; dry mounted). Paratypes. Tanzania: 20 cf , 11 $, larvae, same data as holotype. South Africa: 1 $, C.P., Mossel Bay,ix.1921; 1 Cf , 3 $, vi.-vii.1930 (R. E. Turner). (BMNH; slide and dry mounted.) COMMENTS. T. afrosersalisia may be readily distinguished from other Afrotropical Trioza speciesby the peculiar form of the vertex. The species is tentatively placed in the obsoleta-groupbecause of its reduced hindwings and the presence of an incipient anterior lobe on the hind coxa.Furthermore the host plant is related to Diospyros. Trioza mimusops sp. n. (Figs 194-197, 315, 316)DESCRIPTION. Adult. Integument sparsely covered with long setae. Head, in profile, moderately depressed AFROTROPICAL JUMPING PLANT LICE 55 from longitudinal axis of body, from above narrower than mesoscutum; occipital margin sharp; vertexrounded rectangular, with a concavity on either side of median suture which is clearly defined, frontal lobesnot developed but a low transverse ridge is present; median ocellus clearly visible from above, fronscompletely covered by genae in anterior view; genal cones well developed, rounded conical, in profile inparallel plane to vertex; antennal flagellum 1-54-1-83 times longer than head width, a single rhinariumpresent subapically on flagellomeres 2, 4, 6 and 7, apical flagellomere with one short and one very short setaapically, both of which are truncate; clypeus with a pair of setae, ultimate rostral segment with two pairs ofsetae. Thorax, in profile, moderately arched; pronotum clearly visible from above, only its anterior margindowncurved behind occiput; mesopraescutum, from above, wider than long, its anterior margin arcuate, inprofile gently rounded down to pronotum; forewing elongate oval, narrowing to an acutangular apex,2-52-2-92 times longer than wide, radular areas narrow, remainder of membrane devoid of spinules,course of R and RI marked with brown pigment; veins bearing long setae in proximal third of wing, Rbranch acutangular, M branching distal to or, at most, at RsCu^ line, Cu stem 2-60-4-57 times longerthan Cui b , m\ cell value 1-82-2-20, c^ cell value 2-00-2-93; forewing 1-99-2-17 times longer thanhindwing, costal margin of hindwing with 2-6 setae proximal to costal break, no setae immediately distal tocostal break and two curved setae adjacent to retinaculum; hind coxa with a well-developed meracanthusand an incipient anterior lobe; base of hind tibia swollen dorsally and bearing several small, blunttubercles, with one outer and two inner spurs apically. Abdomen with setae on tergites 6 and 7 in cf and 6-8in $; cf proctiger, paramere and aedeagus as in Figs 195-197; $ genital segment short, conical, dorsalsurface of dorsal valve of ovipositor weakly serrate apically, ventral surface of ventral valve smooth. Measurements (6 cf, 6 $). Maximum width of head, cf 0-54-0-59, $ 0-57-0-61; length of antennalflagellum, cf 0-90-1-06, $ 0-88-1-00; length of ultimate rostral segment, cf and $ 0-09-0-10; length offorewing, cf 3-59-4-08, $ 3-93^-53; length of hind tibia, cf 0-61-0-79, $ 0-65-0-75. Fifth instar larva (Figs 315, 316). Dorsal surface outline very broadly oval, about 1-3 times longer thanwide. Antenna with five flagellomeres. Cephaloprothorax separate from rest of thorax which is entire.Forewing pad about 0-62 mm long, humeral lobe strongly extended forward to anterior margin ofcephaloprothorax. Caudal plate about 0-65 times as long as wide, anus ventral and distant from posteriormargin of abdomen, anal pore area as in Fig. 316. Narrow, tubular, truncate sectasetae forming amoderately dense fringe on anterior margin of cephaloprothorax and caudal plate and a sparse fringe onthe wing pads, postocular seta absent, sectasetae absent from dorsum. HOST PLANTS. Adults and larvae collected from Mimusops obovata and M. zeyheri, adults onlycollected from M. caffra (Sapotaceae). Holotype cf , South Africa: Transvaal, Hartebeestpoort Dam, Bet El Park, 17.viii.1974, Mimusops zeyheri(B. R. Pitkin) (BMNH; dry mounted). Paratypes. South Africa: 14 cf , 12 $ , same data as holotype; 7 cf , 6 $ , larvae, Natal, Durban BotanicalGardens, 23.viii.1974, Mimusops obovata (B. R. Pitkin); 1 $, Port St John, Pondoland, l-ll.vi.1923(R. E. Turner); 5 cf, 9 $, Transvaal, Rustenburg, 20.ii.1965, Mimusops zeyheri; 5 cf, 8 $, larvae,27-30.X.1966; 5 cf , 5 $, larvae, ll-15.ix.1971; 1 $, ll-15.ix.1971, Fagara capensis; 15 cf, 11 $, Pretoria,Hartebeestpoort, 20.V.1965, Mimusops zeyheri; 3 cf, 3 $, 6.viii.l965; 2 cf, 2 $, Natal, Umtentweni,29.ix.1965, Mimusops caffra; 1 $, Umkomaas, 29.ix.-3.x.l965, Bersama lucens (A. L. Capener).(BMNH; NCI; slide and dry mounted and stored in 80% ethanol.) COMMENT. This distinctive species is easily recognised by the shape, venation and pigmentationof the forewing (Fig. 194), and the unusual chaetotaxy of the abdominal tergites. It is tentativelyplaced in the obsoleta-group because the hindwing is only half the length of the forewing, thehind coxa has an incipient anterior lobe and the hind tibia has two inner apical spurs, andbecause of its host plant association. The cocfcere///-group For a discussion of this group see p. 56. Trioza capensis sp. n. (Figs 27, 53, 225-228) DESCRIPTION. Adult. Integument sparsely covered with short setae. Head, in profile, strongly depressedfrom longitudinal axis of body, from above narrower than mesoscutum; occipital margin sharp; vertex 56 DAVID HOLLIS trapezoidal, gently rounded down anteriorly, without frontal lobes or ridge, with a crescent-shapeddepression on either side of the shallow median suture; median ocellus visible from above, frons visible inanterior view; genae slightly swollen ventrally ; antennal flagellum 1 -36-1-49 times longer than head width,a single rhinarium present subapically on flagellomeres 2, 4, 6 and 7, that of 2 with a large curved sense cone(Fig. 27), apical flagellomere with a moderately long and a short truncate seta apically; clypeus with a pairof setae, ultimate rostral segment with two pairs of setae. Thorax moderately arched; pronotum clearlyvisible from above; mesopraescutum, from above, much wider than long, anterior margin broadly arcuate,in profile shallowly rounded down to pronotum; forewing elongate oval, strongly narrowing to roundedrectangular apex, 2-44-2-57 times longer than wide, radular areas narrow elongate, remainder ofmembrane devoid of spinules; veins bearing very short setae, R branch acutangular, M branching distal toRs Cui a line, Cu stem 3-54-4-71 times longer than Cw lb , m\ cell value 1-20-1-33, cu\ cell value 1-71-2-22;forewing 1-44-1-51 times longer than hindwing, costal margin of hindwing with 3-4 setae proximal to costalbreak, setae distal to costal break clearly divided into two groups; hind coxa with a well-developedmeracanthus and a very small anterior lobe; hind tibia with a well-developed basal spine, a verywell-developed preapical outer spur and two inner apical spurs (Fig. 53). Abdomen with setae on tergite 3in cf and 3 and 4 in 9; Cf proctiger, paramere and aedeagus as in Figs 226-228; $ genital segment veryshort, rounded conical, ovipositor valves smooth. Measurements (2 cf , 3 <j>). Maximum width of head, cf 0-53-0-54, $ 0-50-0-52; length of antennalflagellum, cf 0-79, $ 0-68-0-77; length of ultimate rostral segment, cf and $ 0-09-0-10; length of forewing,Cf 2-31-2-38, $ 2-54-2-72; length of hind tibia, cf 0-38, $ 0-38-0-40. Larva unknown. HOST PLANTS. Adults collected from Lycium salinicola and Lycium ? tetrandrum (Sola-naceae). Holotype cf, South Africa: Cape Province, Aliwal North, xii.1922 (R. E. Turner} (BMNH; slidemounted). Paratypes. South Africa: 1 $, same data as holotype; 1 $, 25 mis SSE. Merweville, 2. v. 1972, Lycium ?tetrandum (D. Hollis); 1 $, Nuwerus, 10.xi.1971 (/. G. Theron); 1 CM $, OFS, Philippolis, Vaalbank,19-30. xi. 1969, Lycium salinicolum (A. L. Capener). (BMNH; NCI; slide and dry mounted.) COMMENTS. T. capensis apparently belongs to the same species-group as the New WorldParatrioza cockerelli (Sulc, 1909) and the Palaearctic species of Paratrioza, which all develop onsolanaceous hosts. The paramere is very similar to lycii Loginova, 1970, described from Georgiaand Tadzhikistan, but capensis may be distinguished by the structure of the rhinarium on the 2ndflagellomere, the development and displacement of the outer apical spur on the hind tibia andthe form of the apical segment of the aedeagus. Furthermore capensis completely lacks genalcones and has a more primitive form of thorax, the Paratrioza species having the mesopraescu-tum more produced anteriorly. Previous authors would have placed this species in Paratrioza but I am not sure of the validityof the genus and prefer to place capensis under Trioza. The etfeiMiei'-group A distinctive group of small species defined by the absence of genal cones, bifid sensillaassociated with the antennal rhinaria, forewing with a broadly rounded apex, reduced meracan-thi and a 3-segmented (at least partially) aedeagus. Six species are included: etiennei, messarati-na, seranistama, nestasimara, camerounensis and pitkini. Host plants, both members of theSapotaceae, are known for two of the species, and the only known larva, that of etiennei, is verydistinctive (Fig. 317). Trioza etiennei sp. n. (Figs 46, 50, 198-201, 317, 318) DESCRIPTION. Adult. Integument sparsely covered with very short setae. Head, in profile, at 90 tolongitudinal axis of body, from above as wide as mesoscutum; occipital margin sharp; vertex pentagonal,rounded down to genae, with a median longitudinal ridge on either side of which is a parallel furrow, lateralmargins raised and with anterolateral tubercles, median suture evanescing just above median ocellus; AFROTROPICAL JUMPING PLANT LICE 57 latter not visible from above, frons visible in anterior view; genae smooth, rounded, without trace of cones;antennal flagellum short, 1-06-1-23 times longer than head width, a single subapical rhinarium present onflagellomeres 2, 4, 6 and 7, the proximal three each with a bifid sensillum, the distal one with a short conicalsensillum, apical flagellomere with two long setae apically; clypeus with a pair of setae, ultimate rostralsegment with two pairs of setae. Thorax, in profile, with mesopraescutum almost flat; pronotum stronglyrounded down behind occiput, with a sharp median tubercle, and a blunt tubercle at each dorsolateralmargin (cf. Fig. 203); mesopraescutum much wider than long, its anterior margin broadly arcuate;forewing elipsoid with broadly rounded apex, leading edge almost straight, 2-32-2-65 times longer thanwide, radular areas broad, claval suture bordered with spinules but remainder of membrane devoid ofspinules; veins bearing short setae, R branch acutangular, M branching proximal to RsCu\. A line, Cu stem1-60-2-33 times longer than Cu lb , m { cell value 1-25-1-46, cu v cell value 1-07-1-38; forewing 1-14-1-22times longer than hindwing, costal margin of hindwing with up to two setae proximal to costal break, setaedistal to costal break clearly divided into two groups; hind coxa (Fig. 46) with a reduced meracanthus andwithout anterior lobe; hind tibia without basal tubercles, with one outer and two inner apical spurs and witha double vertical row of up to five thickened setae (Fig. 50); hind basitarsus elongate. Abdomen with setaeon tergites 2-7 in cf and 3-8 in $; cf proctiger, paramere and aedeagus as in Figs 199-201; $ genitalsegment elongate conical, with apical half strongly elongate, ovipositor valves smooth. Measurements (10 cf , 5 $). Maximum width of head, cf 0-28-0-30, $ 0-31-0-33; length of antennalflagellum, cf 0-32-0-37, $ 0-33-0-38; length of ultimate rostral segment, cf 0-07-0-08, $ 0-08-0-09; lengthof forewing, Cf 0-99-1-14, 1-14-1-32; length of hind tibia, cf 0-29-0-36, $ 0-32-0-37. Fifth instar larva (Figs 317,318). Dorsal surface outline elongate oval, about 2-1 times longer than wide,head and thorax clearly differentiated from abdomen, latter with a median posterior prolongation.Antenna with three flagellomeres. Cephaloprothorax separate from rest of thorax which is entire.Forewing pad about 0-42 mm long, humeral lobe weakly extended forward but not reaching hind margin ofeye. Caudal plate about 1-35 times longer than wide, including posterior prolongation; anus ventral anddistant from posterior margin of abdomen, anal pore area as in Fig. 318. Truncate tubular sectasetaeinserted on elongate basal tubercles, and very small lanceolate setae forming a sparse marginal fringe, asingle truncate tubular postocular sectaseta mounted on a basal tubercle present, arrangement ofsectasetae on dorsum as in Fig. 317. HOST PLANT. Larvae and adults collected from Malacantha alnifolia (Sapotaceae). Holotype cf , Senegal: Djebelor, 26.xii.1979, Malacantha alnifolia (J. Etienne) (BMNH; slide mounted).Paratypes. Senegal: 50 cf, 50 $, larvae, same data as holotype; 10 cf, 10 9, larvae, Ziguinchor,20.vii.1981. Ghana: 2 cf, 1 9, Tafo, v.1957, yellow tray. Nigeria: 1 cf, 1 $, Ibadan, Moor Plntn,ll.vii.1956, yellow tray (V. F. Eastop); 1 cf , vi.1957; 11 cf, 6 $, iv.1960; 4 cf, 3 $, 19-27.iv.1961 (F. A.Squire); 1 $, Ikom, 11-12. ii. 1957 (V. F. Eastop). (BMNH; MNHN; slide mounted and stored in 80%ethanol.) COMMENT. T. etiennei and messaratina seem to be a sister pair, distinguished from the rest of thegroup by the presence of tubercles or projections on the pronotum. The two species may beseparated from one another by the characters given in key couplet 23. 'irio7.n messaratina sp. n. (Figs 202-206) DESCRIPTION. Adult. Very similar to etiennei. Median longitudinal ridge and parallel furrows of vertex lesswell developed and median suture weakly indicated just above median ocellus; ultimate rostral segmentwithout setae; antennal flagellum very short, 0-69 times as long as head width, all rhinaria with bifid sensillaalthough that of flagellomere 7 much shorter than the others. Pronotum as in Fig. 203; forewing (Fig. 202)2-95 times longer than wide; spinules present in posterior part of c+sc, completely covering cu 2 and analcell, along hind margin from cu\ to point at which Rs reaches wing margin; Cu stem 4-55 times longer thanCu lb , mi cell value 1-62, cu v cell value 2-89; hind tibia without vertical row of thickened setae, hindbasitarsus not elongate, cf proctiger, paramere and aedeagus as in Figs 204-206; $ unknown. Measurements (1 cf). Maximum width of head, 0-29; length of antennal flagellum, 0-20; length ofultimate rostral segment, 0-06; length of forewing, 0-96; length of hind tibia, 0-23. Larva and host plant unknown. Holotype cf, Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27.vi.1974, yellow tray (D.Hollis) (BMNH; slide mounted). 58 DAVID HOLLIS Trioza seranistama sp. n. (Figs 207-211) DESCRIPTION. Adult. Similar to etiennei. Head more rounded; vertex without longitudinal ridge andassociated furrows, median suture present and complete to occiput; antennal flagellum short, 0-86-0-91times as long as head width, with a single subapical rhinarium present on flagellomeres 2, 4, 6, and 7, eachwith a short bifid sensillum; ultimate rostral segment with one pair of setae. Pronotum without median andlateral tubercles; forewing (Fig. 207) 2-37-2-49 times longer than wide, radular areas weak and broad,spinules present in apical areas of cells r 2 , m t , m 2 , cu\ and occupying most of cu 2 ', veins bearing long setae,Cu stem 1-92-2-72 times longer than Ci b , m^ cell value 1-23-1-39, cui cell value 1-00-1-24; forewing1-19-1 -21 times longer than hindwing, costal margin of hindwing without setae proximal to costal break,setae distal to costal break clearly divided into two groups; hind tibia without vertical rows of thickenedsetae; hind basitarsus short. Abdomen with setae on tergites 2 and 3 in cf and 3 and 4 in $ ; cf proctiger,paramere and aedeagus as in Figs 208-210; apical half of 9 genital segment (Fig. 211) elongate. Measurements (3 cf , 1 $). Maximum width of head, d" 0-32-0-34, $ 0-36; length of antennal flagellum,Cf 0-28-0-31, $ 0-31; length of ultimate rostral segment, cf 0-09, 9 0-10; length of forewing, cf 1-29-1-38,$1-51; length of hind tibia, cf 0-31-0-33, 9 0-34. Larva and host plant unknown. Holotype cf , Tanzania: E. Usambara Mtns, Amani Res sta., c. 3,000', 19-27. vi. 1974, yellow tray (D.Hollis) (BMNH; slide mounted).Paratypes. 2 cf , 1 9 . same data as holotype (BMNH; slide mounted). Trioza nestasimara sp. n. (Figs 212, 213) DESCRIPTION. Adult. Very similar to seranistama. Differs in that rhinarium on flagellomere 4 bears a longbifid sensillum. Forewing 2-58 times longer than wide; Cu stem 1-76 times longer than Cu }b , mi cell value1-43, ciii cell value 0-92; forewing 1-15 times longer than hindwing. Paramere and aedeagus as in Figs 212,213, apical segment of aedeagus more clearly divided. Measurements (1 cf). Maximum width of head, 0-35; length of antennal flagellum, 0-31; length ofultimate rostral segment, 0-08; length of forewing 1-38; length of hind tibia, 0-31. Larva and host plant unknown. Holotype cf, Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27. vi. 1974, yellow tray (D.Hollis) (BMNH; slide mounted). COMMENT. The specimen upon which this species is based was collected in the same yellow tray,during the same period as the type-series of seranistama and, possibly, it is an aberrantspecimen of the latter. However, it would be most unusual for a psyllid species to have suchvariable antennal and cf genitalia structure, and for this reason two species are recognised here. Trioza camerounensis sp. n. (Figs 214, 215) DESCRIPTION. Adult. Very similar to seranistama. Antennal flagellum 1-11-1-23 times longer than headwidth, each rhinarium with a long bifid sensillum; ultimate rostral segment with two pairs of setae.Forewing 2-66-2-75 times longer than wide, spinules present in cells m l , m 2 , cu\ and cu 2 ; veins bearingshort setae, Cu stem 1-67-1-78 times longer than Cw lb , m\ cell value 1-28-1-35, cu\ cell value 1-04-1-29;forewing 1 -34-1 -37 times longer than hindwing, costal margin of hindwing with one seta proximal to costalbreak, cf paramere and aedeagus as in Figs 214, 215, apical segment of aedeagus subdivided. Measurements (1 cf, 2 9)- Maximum width of head, cf 0-31, 9 0-35; length of antennal flagellum, cf0-38, 9 0-39-0-42; length of ultimate rostral segment, cf and 9 0-09; length of forewing, 5 1-63, 91-86-1-89; length of hind tibia, cf 0-29, $ 0-31-0-34. Larva and host plant unknown. Holotype cf , Cameroun: Bamenda, 25-31. i. 1957, yellow tray (V. F. Eastop) (BMNH; slide mounted). Paratypes. Cameroun: 2 $, Bamenda, 21-24.i.l957, yellow tray (V. F. Eastop) (BMNH; slidemounted). AFROTROPICAL JUMPING PLANT LICE 59 Triozapitkinisp. n. (Figs 26, 216-218) DESCRIPTION. Adult. Integument sparsely covered with very short setae. Head, in profile, slightlydepressed from longitudinal axis of body; occipital margin sharp; vertex pentagonal with a weak concavityon either side of median suture, frontal lobes moderately developed; median ocellus not visible fromabove, frons hidden by genae which are slightly swollen but without cones; antennal flagellum short,1-03-1 -24 times longer than head width, a single rhinarium present subapically on flagellomeres 2, 4, 6, and7, that on 2 with a bifid sensillum (Fig. 26), apical flagellomere with two long setae apically; clypeus with apair of setae, ultimate rostral segment with two pairs of setae. Thorax weakly arched; anterior margin ofpronotum weakly downcurved behind occiput; mesopraescutum, from above, about as long as wide andwith a broadly arcuate anterior margin; forewing (Fig. 216) elongate oval with rounded apex, 2-97-3-10times longer than wide, radular areas broad, spinules present below R+M+Cu stem and M stem, this areaalso brown pigmented apart from hyaline areas on either sides of apices of M 3+4 , Cw la and Cw lb ; veins withshort setae, R branch acutangular, M branching distal to RsCu\ a line, Cu stem 3-09-3-63 times longerthan Cw lb , m\ cell value 1-11-1-22, cui cell value 2-22-2-74; forewing 1-26-1-31 times longer thanhindwing, costal margin of hindwing with one seta proximal to costal break, setae distal to costal breakclearly divided into two groups; hind coxa with a moderately developed meracanthus and without anteriorlobe; hind tibia without basal spine, with one outer and two inner apical spurs; hind basitarsus short.Abdomen with setae on tergites 2-7 in cf and 3-8 in $ ; cf paramere and aedeagus as in Figs 217, 218; $genital segment narrowed strongly in apical half and elongate, ovipositor valves smooth. Measurements (4 cf , 3 $). Maximum width of head, cf 0-28-0-29, $ 0-31; length of antennal flagellum,Cf 0-33-0-36, $ 0-32-0-33; length of ultimate rostral segment, cf 0-07-0-09, $ 0-08-0-09; length offorewing, cf 1-43-1-53, $ 1-65-1-71; length of hind tibia, cf 0-23-0-26, $ 0-25-0-26. Larva unknown. HOST PLANT. Adults collected from Chrysophyllum viridifolium (Ipruiniforme) (Sapotaceae). Holotype cf , Kenya: Nairobi Arboretum, c. 5,400', 25-26. vii. 1974, Chrysophyllum viridifolium (Ipruini-forme) (D. Hollis (BMNH; dry mounted). Paratypes. 32 cf , 15 $ , same data as holotype (BMNH; NMK; slide and dry mounted and stored in 80%ethanol). COMMENT. T. pitkini may be readily distinguished from other members of the etiennei-group bythe shape, pattern and venation of the forewing (Fig. 216). The g/aftea-group Differs from the etiennei-group in that the aedeagus is 2-segmented with the apical segmentthickened basally, and the forewing has a rounded acutangular apex. Two species, glabea andusambarica, are included but larvae and host plants are unknown. Trioza glabea sp. n. (Figs 219-222) DESCRIPTION. Adult. Integument sparsely covered with very short setae. Median suture of vertex presentand complete; genae rounded, without cones; antennal flagellum 1 -61-1-70 times longer than head width, asingle rhinarium present subapically on flagellomeres 2, 4, 6 and 7, those on 2 and 7 with a short conicalsensillum, those on 4 and 6 with a long bifid sensillum, apical flagellomere with two long subequal setaeapically; clypeus with a pair of setae, ultimate rostral segment with two pairs of setae. Forewing (Fig. 219)elongate elipsoid with a rounded acutangular apex, 2-88-2-92 times longer than wide, radular areas narrowtriangular, membrane with spinules at distal ends of cells r 2 and w 2 , almost completely filling m\ and cuiand following the course of claval suture; veins bearing moderately dense rows of short setae, R branchacutangular, M branching proximal to Rs-Cui a line, Cu stem 0-60-0-64 times as long as Cu lb , m\ cell value1-59-1-62, cuj cell value 0-80-0-85; forewing 1-44-1-48 times longer than hindwing, costal margin ofhindwing without setae proximal to costal break, setae distal to costal break clearly divided into twogroups; hind coxa with a well-developed meracanthus and an incipient anterior lobe; hind tibia with a smallbasal spine and one outer and two inner apical spurs. Abdomen with setae on tergites 2 and 3 in cf and 3and 4 in $; cf proctiger, paramere and aedeagus as in Figs 220-222, apical segment of aedeagus with a 60 DAVID HOLLIS heavily sclerotised and bilobed base; $ genital segment short, conical, dorsal surface of dorsal valve ofovipositor weakly serrate apically. Measurements (1 cf , 1 $). Maximum width of head, cf 0-40, $ 0-44; length of antennal flagellum, cf0-68, $ 0-71; length of ultimate rostral segment, cf 0-10, $ 0-11; length of forewing, cf 2-48, $ 2-82; lengthof hind tibia, Cf and $0-39. Larva and host plant unknown. Holotype cf , Angola: 7 mis W Gabela, 16-18.iii.1972, at light (D. Hollis} (BMNH; slide mounted). Paratypes. Angola: 1 $ , same data as holotype. Zaire ('Congo Beige'): 1 $ , P.N.G. , Miss. H. de Saeger,Pidigala, 23. iv. 1952 (H. de Saeger). (BMNH; MRAC; slide mounted.) COMMENTS. T. glabea and usambarica are obviously closely related and I regard them as a sisterpair. Both lack genal cones, have bifid sensilla on the antennal rhinaria, and the apical segmentof the aedeagus has a thickened base. The presence of bifid sensilla suggests a possiblesister-group relationship with the etiennei-group. A third species belonging to this group is represented by four specimens, from Zaire,deposited in MRAC, but this material is too badly damaged for description. Trioza usambarica sp. n. (Figs 223, 224) DESCRIPTION. Adult. Very similar to glabea. Integument more densely covered with slightly longer setae.Antennal flagellum 1-54 times longer than head width. Forewing with r 2 devoid of spinules; veins bearingslightly longer setae, Cu stem 0-66-0-73 times as long as C lb , mi cell value 1-57-1-58, cu\ cell value0-74-0-86. cf paramere and aedeagus as in Figs 223, 224, apical segment of aedeagus with a less expandedbase; dorsal surface of dorsal valve of ovipositor smooth. Measurements (1 cf , 1 $). Maximum width of head, cf 0-46, $ 0-51; length of antennal flagellum, cf0-71 ($ damaged); length of ultimate rostral segment, cf 0-13, $ 0-14; length of forewing, cf 2-64, $ 2-95;length of hind tibia, cf 0-41, $ 0-44. Larva and host plant unknown. Holotype cf , Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27. vi. 1974, yellow tray (D.Hollis) (BMNH; slide mounted).Paratype. Tanzania: 1 cf , similar data as holotype but swept (BMNH; slide mounted). Ungrouped species Twelve species are described in this section. Some may be grouped together in pairs on the basisof overall similarity due to common possession of characters in the primitive state. The speciesare ordered alphabetically and each may represent a distinct species-group. Trioza ficiola sp. n. (Figs 229-232, 319, 320) DESCRIPTION (only slide-mounted material available for study). Adult. Integument sparsely covered withmoderately long setae. Head with broadly conoid, well-developed genal cones; antennal flagellum1-27-1 -35 times longer than head width, a single rhinarium present subapically on flagellomeres 2, 4, 6 and7, apical flagellomere with one long and one short and truncate seta apically; clypeus with a pair of setae,ultimate rostral segment with several pairs of setae. Forewing elongate elipsoid, strongly narrowing distallyto a rounded obtusangular apex, 2-52-2-92 times longer than wide, radular areas narrow triangular,remainder of membrane devoid of spinules; veins bearing short setae, R branch acutangular, M branchdistal to Rs-Cu { . d line, Cu stem 1-21-1-53 times longer than Cw lb , m { cell value 1-53-1-64, cw t cell value1-67-2-20; forewing 1-73-1-78 times longer than hindwing, costal margin of hindwing with 0-2 setaeproximal to costal break, setae distal to costal break weakly divided into two groups; hind coxa with awell-developed meracanthus and without anterior lobe; hind tibia with a weak basal tubercle bearingseveral small conical spines, with one outer and three inner apical spurs. Abdomen with setae on tergites 2and 3 in cf , and 3 and 4 in 9 ; Cf proctiger, paramere and aedeagus as in Figs 230-232; $ genital segmentvery short, conoid, subgenital plate truncate apically, ventral surface of ventral valve of ovipositor saw-likeand dorsal surface serrate apically. AFROTROPICAL JUMPING PLANT LICE 61 Measurements (3 cf, 1 $). Maximum width of head, cf 0-62-0-66, $ 0-66; length of antennalflagellum,Cf 0-86-0-89, $ 0-84; length of ultimate rostral segment, cf and $ 0-11; length of forewing, cf 4-01-4-11, $4-40; length of hind tibia, cf 0-64-0-65, 0-64. Fifth instar larva (Figs 319, 320). Dorsal surface outline almost circular, about 1-1 times longer thanwide. Antenna about 0-3 mm long (segmentation not clear in specimens available). Cephaloprothoraxseparated from rest of thorax which is entire. Forewing pad 0-92 mm long, humeral lobe extended forwardbeyond anterior margin of eye. Caudal plate about 0-5 times as long as wide, anus ventral and distant fromposterior margin of abdomen, anal pore area as in Fig. 320. Elongate truncate tubular sectasetae forming adense, entire marginal fringe, post-ocular seta absent, sectasetae absent from dorsum. HOST PLANT. Adults and larvae collected from Ficus sp. (Moraceae). There is no information asto whether or not the larvae form pit galls on the host plant but this is highly likely judging fromtheir shape and the arrangement of sectasetae. Holotype cf, Mozambique: Musape River Valley, 7.vii.l968, Ficus sp. (C. J. Hodgeson) (NCI; slidemounted).Paratypes. 2 cf , 1 $ , larvae, same data as holotype (NCI; BMNH; slide mounted). COMMENTS. Apart from the male paramere and aedeagus this species has few diagnosticfeatures. It resembles species of the erytreae-group but the venation of the forewing has a highermi cell value and a lower cu\ cell value, and the structure of the female ovipositor is more derivedthan that of other members of the group. Trioza fuscivena sp. n. (Figs 233-236) DESCRIPTION. Adult. Integument sparsely covered with moderately long hairs. Head, in profile, slightlydepressed from longitudinal axis of body, from above at least as wide as mesoscutum; occipital marginobtuse; vertex flattened but deeply divided anteriorly by median suture, frontal lobes weakly developed;median ocellus not visible from above, frons just visible in anterior view; genal cones narrow, conical, welldeveloped, in profile depressed strongly from plane of vertex; antennal flagellum 1-59-1-98 times longerthan head width, a single rhinarium present subapically on flagellomeres 2, 4, 6 and 7, apical flagellomerewith two subequal terminal setae, the shorter with a truncate apex; clypeus with a pair of setae, ultimaterostral segment with two pairs of setae. Thorax weakly arched; pronotum, from above, with a rhomboiddorsal disc and narrowing strongly on each side; mesopraescutum, from above, longer than wide, anteriormargin strongly arcuate, in profile strongly rounded down to pronotum; forewing elongate elipsoid,narrowing to rounded acutangular apex, 2-65-2-92 times longer than wide, membrane devoid of spinulesapart from narrow triangular radular areas, course of R stem and R\ and bases of M stem and Cu stembrown pigmented; veins sparsely clothed with moderately long hairs in basal half of wing, R branchacutangular, M branch distal to /?s-Cw la line, Cu stem 1-51-1-77 times longer than Cw lb , mi cell value1-07-1-26, cu\ cell value 1-86-2-08; forewing 1-66-1-78 times longer than hindwing, costal margin ofhindwing with 4-7 setae proximal to costal break, setae distal to costal break sparse but clearly divided intotwo groups; hind coxa with a well-developed meracanthus and without anterior lobe; hind tibia with 3-4small but clearly developed spines, with one outer and two inner apical spurs. Abdomen with setae ontergite 3 in cf , and tergite 4 in $ ; cf proctiger very weakly expanded laterally, paramere and aedeagus as inFigs 234, 235; 9 genital segment (Fig. 236) short, conoid, proctiger strongly downcurved apically,subgenital plate with truncate posterior margin, lower valve of ovipositor with three dorsolateraltransverse ridges apically. Measurements (4 cf , 2 $). Maximum width of head, cf 0-50-0-54, $ 0-54-0-56; length of antennalflagellum, cf 0-94-1-05, $ 0-86-0-93; length of ultimate rostral segment, cf 0-07, $ 0-08; length offorewing, cf 3-25-3-46, 9 3-76-3-89; length of hind tibia, cf 0-66-0-70, $ 0-68-0-70. Larva and host plant unknown. Holotype cf , Cameroun: Bamenda, i-ii.1957, yellow trays (V. F. Eastop) (BMNH; slide mounted).Paratypes. 4 cf , 2 9 , same data as holotype (BMNH; slide and dry mounted). COMMENT. This species is superficially similar to nachingweae (p. 49), principally because of theforewing pigmentation. However, the third flagellomere is not greatly reduced and fuscivena isnot considered a member of the neoboutonia-group. Other differences from nachingweae may 62 DAVID HOLLIS be found in the short ultimate rostral segment, the longer setae on the fore wing veins and theform of the male and female genitalia (Figs 234-236). No close association with any other speciescan be found. Triozaghanaensissp. n. (Figs 28, 237-240) DESCRIPTION. Adult. Integument sparsely covered with very short setae. Head, from above, almost as wideas mesoscutum, in profile depressed at 90 from longitudinal axis of body; vertex pentagonal, flatteneddorsally, rounded down to frons, with sharp occipital margins and clearly defined median suture, lateralconcavities weak, frontal lobes absent; median ocellus not visible from above, frons just visible in anteriorview; genae rounded, without cones; antennal flagellum 1-44-1 -47 times longer than head width, a singlerhinarium present subapically on flagellomeres 2, 4, 6 and 7, those on 2 and 7 with an oval sense cone, thoseon 4 and 6 each with a long bifid sense cone, apical flagellomere with two long subequal terminal setae;clypeus with a pair of setae, ultimate rostral segment with two pairs of setae. Thorax strongly arched;pronotum strongly curved down behind occiput; mesopraescutum, from above, almost as long as wide,anterior margin broadly arcuate and strongly rounded down to pronotum; forewing elongate elipsoid,strongly narrowing to obtusangular apex, 2-54-2-55 times longer than wide, radular areas elongate andnarrow, rest of membrane devoid of spinules apart from small area at base of claval suture; veins bearingshort setae, R branch acutangular, M branch proximal to RsCu\ & line, Cu stem 0-36-0-42 times as long asCu lb , mi cell value 2-09-2-34, cui cell value 0-75-0-82; forewing 1-74-1-80 times longer than hindwing,costal margin of hindwing with 3-6 setae proximal to costal break, 10-12 setae immediately distal to costalbreak and 4-6 setae adjacent to retinaculum; hind coxa with a well-developed meracanthus and withoutanterior lobe; hind tibia with a pair of unequally developed basal tubercles and one outer and two innerapical spurs. Abdomen with setae on tergites 2 and 3 in cf and 3 and 4 in $ ; cf proctiger, paramere andaedeagus as in Figs 238-240; 9 genital segment conical, dorsal valves of ovipositor serrate apically. Measurements (2 cf , 2 <j>). Maximum width of head, cf 0-73-0-75, $ 0-75-0-78; length of antennalflagellum, cf damaged, $ 1-08-1-15; length of ultimate rostral segment, cf 0-22, $ 0-23; length offorewing, Cf 5-27-5-35; $ 5-64-5-81; length of hind tibia, cf 0-90, $ 0-91-0-92. Larva unknown. HOST PLANT. Adults collected 'in severely galled young terminal leaves of Malacantha sp.'(Sapotaceae). Holotype cf , Ghana: Tafo, xii.1942, on Malacantha sp. (H. E. Box) (BMNH; slide mounted). Paratypes. Ghana: 1 cf , 3 $ , same data as holotype. Zaire ('Congo Beige'): 4 9 , P.N. A., Nyasheke (vol.Nyamuragira), 1820 m, 14-26,vi.l935 (G. F. de Witte). (BMNH; MRAC; slide and dry mounted.) COMMENT. See under tenuis (p. 67). Trioza guiera sp. n. (Figs 241-244, 321, 322) DESCRIPTION. Adult. Integument covered with moderately long setae. Head, in profile, slightly depressedfrom longitudinal axis of body, from above slightly narrower than mesoscutum; occipital margin sharp;vertex with anterior margin incised by median suture, latter with an irregular concavity on either sidedorsofrontally; median ocellus visible from above, frons visible in anterior view; genal cones very short,rounded; antennal flagellum short, 0-9-1-21 times as long as head width, a single rhinarium presentsubapically on flagellomeres 2, 4, 6 and 7, apical flagellomere.with a long pointed seta and a short truncateseta terminally; clypeus with a pair of weak setae, ultimate rostral segment long, devoid of setae. Thoraxweakly arched; pronotum visible from above, its anterior margin rounded down behind occiput; meso-praescutum, from above, wider than long, with broadly arcuate anterior margin, in profile stronglydowncurved to pronotum; forewing elipsoid, with rounded apex, 2-25-2-57 times longer than wide,membrane densely covered with spinules, radular areas broadly triangular; veins bearing long setae at leastin proximal half of wing, C+Sc thickened at base, R branching from R+M+Cu slightly proximal to Cu, Rbranch acutangular, M branch distal to Rs-Cu la line, Cu stem 0-93-1-79 times as long as Cw lb , mi cellvalue 1-06-1-32, cui cell value 1-26-1-7; forewing 1-41-1-56 times longer than hindwing, costal margin ofhindwing with 3^4- setae proximal to costal break, no setae immediately distal to costal break and 2-3 setaeadjacent to retinaculum; hind coxa with a well-developed meracanthus and without anterior lobe; hind AFROTROPICAL JUMPING PLANT LICE 63 tibia with a group of 2-3 very small spines basally and without strongly developed apical spurs. Abdomenwith setae on tergites 2-7 in cf and 3-8 in 9 ; Cf proctiger, paramere and aedeagus as in Figs 242-244; $genital segment short, conical, apex of proctiger upcurved and hook-like, valves of ovipositor smooth. Measurements (10 cf , 10 $). Maximum width of head, cf 0-40-0-45, 9 0-43-0-47; length of antennalflagellum, cf 0-37-0-51, $ 0-41-0-50; length of ultimate rostral segment, cf 0-10-0-13, $ 0-11-0-13; lengthof forewing, cf 1-62-1-84, $ 1-86-2-14; length of hind tibia, cf 0-27-0-31, $ 0-30-0-34. Fifth instar larva (Figs 321, 322). Dorsal surface outline oval, about 1-4 times longer than wide, withindentations on either side at posterior margin of eye and at base of abdomen, posterior margin ofabdomen with a deep indentation medially. Antenna with 4-5 flagellomeres. Cephaloprothorax separatefrom rest of thorax which is entire. Forewing pad about 0-37 mm long, humeral lobe weakly extendingforward to posterior margin of eye. Caudal plate about 0-65 times as long as wide, anus apico-dorsal, analpore area as in Fig. 322. Pointed conical sectasetae of unequal length forming an uneven dense marginalfringe interrupted posteromedially; blunt conical sectasetae present on either side of dorsal mid line fromcephaloprothoracic suture to posterior margin of abdomen; a single conical sectaseta present on eachlateral margin of the abdominal tergite immediately anterior to caudal plate. HOST PLANT. Adults and larvae collected from Guiera senegalensis (Combretaceae). The larvaeare apparently free-living and one sample collected was associated with a species of the ant genusCrematogaster . Holotype cf, Senegal: Manpalago, 12.55N/16.00W, 13. i. 1981, Guiera senegalensis (J. Etienne} (BMNH;dry mounted). Paratypes. Senegal: 43 cf , 31 $, larvae, same data as holotype; 7 cf , 7 $, numerous larvae, Goudemp,23. vi. 1981, Guiera senegalensis (J. Etienne). Gambia: 2 cf, 2 $, larvae (V. F. Eastop). Chad: 5 $, larvae,Bebedjia8.40N/16.33E, 10.xi.l974(fl. M. Bink-Moenen). Sudan: 15 cf, 14 $, larvae, 50km W. ElObeid,19. iv. 1981, Guiera senegalensis, tended by Crematogaster sp.; 6 cf , 13 $, larvae, 140 km W. El Obeid,21.iv.1981; 9 cf, 8 ?, larvae S. Darfur, Ed Da'ein, 3.V.1981; 1 cf, 1 $, larvae, Ed Da'ein to BabanusaRoad, 4.V.1981; larvae, 40 km SW. El Muglad, 6.V.1981 (/. H. Martin}. (BMNH; MNHN; slide and drymounted and stored in 80% ethanol.) COMMENTS. The shape and venation of the forewing (Fig. 241) and the dense complete coveringof spinules on the membrane serve to distinguish this species from other Afrotropical triozids.Furthermore the anus of the 5th instar larva is in a dorsal position, possibly to facilitateant-tending. Trioza karroo sp. n. (Figs 51, 245-248) DESCRIPTION. Adult. Integument sparsely covered with short setae. Head, in profile, moderately depressedfrom longitudinal axis of body, from above narrower than mesoscutum; occipital margin obtuse but welldefined; vertex with a well-defined concavity on either side of median suture, frontal lobes weaklydeveloped; median ocellus visible from above, frons completely covered by genal cones, latter welldeveloped and with rounded apices, their longitudinal axis parallel with but not in same plane as vertex;antennal flagellum 1-89-2-16 times longer than head width, a single rhinarium present subapically onflagellomeres 2, 4, 6 and 7, apical flagellomere with one long and one short and truncate seta terminally;clypeus prominent and bearing a pair of setae, ultimate rostral segment without setae. Thorax weaklyarched; pronotum clearly visible from above, its anterior margin rounded down behind occiput; meso-praescutum, from above, slightly wider than long, anterior margin broadly arcuate, in profile angled downto pronotum; forewing (Fig. 245) elongate, narrow, rounded apically, 3-09-3-23 times longer than wide,membrane with a broad band of brown pigmentation extending across the longitudinal axis of the wing,following the course of R+M+Cu, M stem and M 1+2 and extending across posterior half of wing, evenlyspinuled and with diffuse radular areas; veins bearing short setae, R branch acutangular, M branch atRs-Cui a line, Cu stem 1-10-2-26 times longer than Cw lb , ra t cell value 1-58-1-87, cui cell value 1-47-1-79;forewing 1-23-1-32 times longer than hindwing, costal margin of hindwing bearing two setae proximal tocostal break, setae distal to costal break not clearly divided into two groups; hind coxa with a well-definedmeracanthus and without anterior lobe; hind tibia with a group of small tubercles on a raised bulge basally,without well-defined apical spurs but with an incomplete ring of thickened setae (Fig. 51). Abdomen withsetae on tergites 2 and 3 in cf and 3 and 4 in $ ; cf proctiger and aedeagus as in Figs 246, 247 (parameres 64 DAVID HOLLIS damaged in holotype); 9 genital segment (Fig. 248) short, conical, proctiger, in profile, with a prominentbulge posterior to anus and narrowing strongly to apex, ovipositor valves smooth. Measurements (1 d", 6 $). Maximum width of head, cf 043, 9 0-41-0-46; length of antennal flagellum,Cf 0-92, 9 0-85-0-98; length of ultimate rostral segment, cf 0-07, $ 0-08-0-09; length of forewing, cf 2-189 2-45-2-63; length of hind tibia, cf 0-36, $ 0-36-0-40. Larva and host plant unknown. Holotype cf , South Africa: C.P., Grabouw, 6.U971, swept (/. G. Theron) (NCI; slide mounted). Paratypes. South Africa: 4 9, same data as holotype; 4 $, Ceres, iv.1925 (R. E. Turner). (BMNH; NCI;slide and dry mounted.) COMMENTS (see also p. 64). T. karroo is easily recognised by the pigmentation of the forewing(Fig. 245) and the absence of hind tibial spurs. The male genitalia are not particularly distinctivebut the form of the female proctiger (Fig. 248) is very unusual. Trioza laingisp. n. (Figs 249-253) DESCRIPTION. Adult. Body sparsely covered with short setae, genal cones and legs bearing longer setae.Head, in profile, depressed almost at 90 to longitudinal axis of body, from above narrower thanmesoscutum; occipital margin obtuse but defined; vertex with a well-developed concavity on either side ofmedian suture which is deeper towards median ocellus, frontal lobes not developed; median ocellus notvisible from above, frons completely covered by genae; genal cones well developed, in profile depressedfrom axis of vertex; antennal flagellum 1-74 times longer than head width, a single rhinarium presentsubapically on flagellomeres 2, 4, 6 and 7, apical flagellomere with one long and one short and truncate setaapically; clypeus with a pair of setae, ultimate rostral segment with two pairs of setae. Thorax weaklyarched; pronotum clearly visible from above and rounded down behind occiput; mesopraescutum, fromabove slightly wider than long, anterior margin broadly arcuate, in profile gently rounded down topronotum; forewing (Fig. 249) elongate oval with rounded apex, 2-83-3-16 times longer than wide,membrane with an even brown pigmentation extending over posterior half of wing from the line ofR+M+Cu, M stem and M 1+2 , evenly covered with spinules and with diffuse radular areas; veins bearingshort setae, R branch acutangular, M branching at or proximal to Rs-Cu ia line, Cu stem 1-45-1-86 timeslonger than Cu lb , m l cell value 1-19-1-22, cu v cell value 1-07-1-40; forewing 1-23-1-25 times longer thanhindwing, costal margin of hindwing with 0-1 seta proximal to costal break, setae distal to costal breakclearly divided into two groups; hind coxa with a well-developed meracanthus and without anterior lobe;hind tibia with a weakly developed basal spine, with one strong outer and three inner apical spurs; basalsegment of hind tarsus elongate. Abdomen with setae on tergites 2 and 3 in cf and 3 and 4 in 9; Cfproctiger, paramere and aedeagus as in Figs 250-252; 9 genital segment (Fig. 253) very short, proctigerwith rounded apex, apicodorsal surface of ventral valves of ovipositor weakly serrate. Measurements (2 cf , 2 9). Maximum width of head, cf 0-55-0-57, $ 0-59-0-61; length of antennalflagellum, cf 0-96, $ 1-06; length of ultimate rostral segment, cf 0-10-0-11, 9 0-11-0-12; length offorewing, cf 2-91-3-14, 9 3-36-3-47; length of hind tibia, cf 0-45-0-48, 9 0-47-0-49. Larva unknown. HOST PLANT. Possibly Bartsia longiflora (Scrophulariaceae). Holotype cf , Kenya: Mt Elgon, 10,300', i.1974, on yellow flowers (S. Collins) (BMNH; slide mounted). Paratypes. Kenya: 1 9> same data as holotype. Ethiopia: 1 9 (without head), Simien, Lori, 11,500',beaten from low yellow-flowered bushes, Bartsia longiflora; 1 cf , 3 $, 'Abyssinia', Mt Zuquala, c. 9,000',22.x. 1926, beaten from trees near lake shore (H. Scott). Zaire ('Congo Beige'): 1 9 Ruanda, Lac N'Gando('pied vole. Karisimbi'), 2,400 m, 8. in. 1935 (G. F. de Witte 1216); 1 cf Terr. Rutshuru, 7.iv.l937 (Miss.Prophylactique). (BMNH; MR AC; slide and dry mounted.) COMMENTS. This species most closely resembles T. obscura Tuthill, 1952, and some undescribedspecies which develop on Hebe spp. (Scrophulariaceae) in New Zealand, but this resemblancemay be due to characters of the head, thorax and forewing being in the primitive state. Withinthe African fauna laingi is similar to karroo (p. 64) but the two species may be separated by thecharacters given in key couplet 4. The material from Mt Zuquala, Ethiopia, was deposited in the BMNH under the Laingmanuscript name ' Trioza semibrunneipennis' . AFROTROPICAL JUMPING PLANT LICE 65 Trioza medleri sp. n. (Figs 10, 39, 254, 7255) DESCRIPTION. Adult. Integument densely covered with long setae (Fig. 39). Head, in profile, depressed at90 to longitudinal axis of body, from above narrower than mesoscutum; occipital margin rounded; vertexdeeply divided into two bulbous halves by median suture, eyes very prominent and rounded; medianocellus not visible from above, frons completely covered by genae; genal cones long, conical, with subacuteapices; antennal flagellum (Fig. 10) 1-66 times longer than head width, a single rhinarium presentsubapically on flagellomeres 2, 4, 6 and 7, 1-6 with very long setae, apical flagellomere with one long andone extremely short and truncate seta apically; clypeus with a pair of setae, ultimate rostral segment withseveral pairs of setae. Thorax, in profile, strongly arched; pronotum, from above, with a rhomboid dorsaldisc and strongly narrowing laterally, in profile strongly rounded down behind occiput; mesopraescutum,from above, slightly wider than long, anterior margin broadly arcuate, in profile rounded down topronotum; forewing (Fig. 254) obovoid, broadly rounded distally, 2-29 times longer than wide, radularareas elongate, cell c+sc with a narrow diagonal band of spinules, remainder of membrane devoid ofspinules; veins densely clothed with long setae, R branch acutangular, M branching proximal to Rs Cw ialine, Cu stem 0-39 times as long as Cw lb , m t cell value 2-61, cui cell value 0-76; forewing 1-38 times longerthan hindwing , costal margin of hindwing with 0-1 seta proximal to costal break , setae distal to costal breakclearly divided into two groups; hind coxa with a well-developed meracanthus and without anterior lobe;hind tibia without basal spines and with one outer and two inner apical spurs, cf unknown; $ abdomenwith setae on tergite 3, genital segment short, conoid, proctiger strongly sclerotised apically, ovipositorvalves smooth. Measurements (1 $). Maximum width of head, 0-7; length of antennal flagellum, 1-16; length of ultimaterostral segment, 0-14; length of forewing, 3-85; length of hind tibia, 0-67. Larva and host plant unknown. Holotype $, Nigeria: NE. State, Ngel Nyaki, 31.iii.1970 (/. T. Medler} (BMNH; dry mounted). Paratypes. 2 $ , same data as holotype (BMNH; slide and dry mounted). Non-paratypic material. Zaire ('Congo Beige'): 1 cf , PNG, Miss. H. de Seager, II/fe/7, 4.vii. 1952, 3729;1 cf , PFNK 7/9, 28.vii.1952, 3842 (H. de Saeger] (MRAC). COMMENTS. This is a very distinctive species which may be easily recognised by the structure ofthe head, pronotum and forewing, the latter resembling those of the Hawaiian genera Hevahevaand Hemischizocranium. In overall appearance medleri is similar to the Indian species Petaloly-ma basalis (Walker, 1858) but the latter has a less cleft head which is more adpressed to thethorax, the pronotum is completely hidden dorsally, the mesonotum is less arched or producedforward, and the tibial spurs are much more strongly developed. Apart from the type-series I have examined 2 cf from Zaire, deposited in MRAC, which maybe conspecific but have a much shorter m\ cell (value 1-85). The genitalia are shown in Fig. 255. Trioza schroederi sp. n. (Figs 256-260) DESCRIPTION. Adult. Integument sparsely clothed with short setae. Head, in profile, depressed almost to90 from longitudinal axis of body, from above almost as wide as mesoscutum; occipital margin broadlyobtusangular; vertex rounded pentagonal, anteriorly incised by median suture and strongly rounded togenae, secondary diagonal grooves present from median ocellus backwards to lateral concavities, latterweak; median ocellus just visible from above, frons just visible in anterior view; genal cones elongateconoid, with rounded apices, in parallel plane to vertex; antennal flagellum 2-33-2-64 times longer thanhead width, a single rhinarium present subapically on flagellomeres 2, 4, 6 and 7, apical flagellomere withtwo long subequal setae apically; clypeus with a pair (rarely with a second very short pair) of setae, ultimaterostral segment with a pair of setae. Thorax weakly arched; pronotum strongly rounded down behindocciput; mesopraescutum, from above, about as long as wide, with arcuate anterior margin, in profilestrongly rounded down to pronotum; forewing (Fig. 256) elongate oval with rounded obtusangular apex,2-76-2-87 times longer than wide, membrane devoid of spinules apart from a small cloud around base ofclaval suture and triangular radular areas; veins bearing short setae, R branch acutangular, Rs long andsinuous, M branching proximal to RsCu ia line, Cu stem 1-32-1-71 times longer than Cw lb , m\ cell value1-47-1-65, cu } cell value 1-29-1-68; forewing 1-5 times longer than hindwing, costal margin of hindwing 66 DAVID HOLLIS with up to eight setae proximal to costal break, setae distal to costal break clearly divided into two groups;hind coxa with a well-developed meracanthus and without anterior lobe; hind tibia with 1-2 small tuberclesbasally and one outer and two inner apical spurs. Abdomen with setae on tergites 2 and 3 in cf and 3 and 4in 9 ; cf proctiger, paramere and aedeagus as in Figs 257-259; 9 genital segment (Fig. 260) short, conoid,proctiger with a weak transverse groove immediately posterior to anus, apex of ventral valve of ovipositorwith two ventrolateral serrations. Measurements (4 cf , 2 <j>). Maximum width of head, cf 0-49-0-56, $ 0-56-0-57; length of antennalflagellum, cf 1-19-1-48, $ 1-39; length of ultimate rostral segment, cf 0-13-0-16, $ 0-14-0-15; length offorewing, cf 2-85-3-68, $> 3-49; length of hind tibia, cf 0-47-0-62, $ 0-54-0-56. Larva and host plant unknown. One adult cf paratype was collected on Myrica conifer a (Myricaceae). Holotype cf , Tanzania: Kilimanjaro, Bismark Hut, 2,500-3,000 m, S. Mawenzi, at foot of high pasture,ii.1912 (Chr. Schroder} (MNHU; slide mounted). Paratypes. Tanzania: 6 cf , 3 $, same data as holotype. Zaire ('Congo Beige: Ruanda'): Lac N'Gando,pied Vole. Karisimbi, 2,400 m, 6.iii.l935 (G. F. de Witte). Zimbabwe ('S. Rhodesia'): 1 Cf, Harare('Salisbury'), iii.1957, Myrica conifera (N. L. H. Krauss). (BMNH; MNHU; MRAC; slide mounted andstored in 80% ethanol.) Non-paratypic material (damaged specimens). Tanzania: 1 $, 1 ?, 'D. O. Afrika, Kilimandscharo',3000-4000 m, i.1906; 2 $, Amani, xii.1905 (Chr. Schroder) (MNHU). Zaire ('Congo Beige'): P.N.A.,Secteur Tshiaberimu, riv. Kalivina affll., Talia Nord, 2,350 m, 29.iii.1954 (P. Vanschuytbroek & H.Synave) (MRAC). COMMENTS. This species shows no clear relationships to other Afrotropical species, havingseveral features of the wing in the primitive state. It is superficially similar to theroni (p. 67),from which it differs in having long, well-developed genal cones and lacking supplementaryrhinaria on the first flagellomere. Trioza tangaesp. n. (Figs 18, 261-264) DESCRIPTION. Adult. Integument covered with long setae. Head, in profile, depressed almost at 90 tolongitudinal axis of body; occipital margin obtusangular; vertex pentagonal, rounded down to genae,median suture weak but complete, lateral concavities well developed; median ocellus just visible fromabove, frons completely covered by genal cones; latter well developed, conoid, with rounded apices;antennal flagellum 1-76-1-83 times longer than head width, a single rhinarium present subapically onflagellomeres 2, 4, 6 and 7, apical flagellomere twice as long as 7th (Fig. 18) and bearing one long and oneshort and truncate seta apically; clypeus with a pair of setae, ultimate rostral segment with three pairs ofsetae. Thorax weakly arched; pronotum strongly rounded down behind occiput; mesopraescutum, fromabove, about as long as wide, with a moderately arcuate anterior margin, in profile gently rounded down topronotum; forewing (Fig. 261) elongate elipsoid, narrowing distally to rounded obtusangular apex,2-62-2-79 times longer than wide, membrane devoid of spinules apart from narrow radular areas and asmall patch at base of claval suture; veins bearing moderately long setae at least in proximal half of wing, Rbranch acutangular, M branching distal to Rs Cw la , Cu stem 2-26-2-81 times longer than Cw lb , m\ cellvalue 1-08-1-38, cu\ cell value 1-83-2-32; forewing 1-47-1-51 times longer than hindwing, costal margin ofhindwing with 0-1 seta proximal to costal break, setae distal to costal break clearly divided into two groups;hind coxa with a well-developed meracanthus and without anterior lobe; hind tibia with 3-4 small tuberclesbasally, with one outer and two inner apical spurs. Abdomen with setae on tergites 2 and 3 in cf and 3 and 4in 9; cf proctiger, paramere and aedeagus as in Figs 262-264; 9 genital segment short, conoid, ventralvalves of ovipositor arrowhead-like apically and finely serrate. Measurements (2 cT, 2 $). Maximum width of head, cf 0-41-0-47, $ 0-41-0-43; length of antennalflagellum, cf 0-75, $ 0-72-0-76; length of ultimate rostral segment, cf 0-08-0-10, 0-07-0-08; length offorewing, cf 2-30-2-61, $ 2-20-2-49; length of hind tibia, cf 0-48-0-52, $ 0-45-0-50. Larva and host plant unknown. Holotype cf, Tanzania: E. Usambara Mtns, Amani Res. sta., c. 3,000', 19-27. vi. 1974, yellow tray (D.Hollis} (BMNH; slide mounted). Paratypes. Tanzania: 2 $ , same data as holotype; 1 cf , light-trap (D. Hollis); 1 cf , Kilimanjaro, BismarkHut, 2,500-3,000 m, S. Mawenzi, at foot of high pasture, ii.1912 (Chr. Schroder). (BMNH; MNHU; slideand dry mounted.) AFROTROPICAL JUMPING PLANT LICE 67 COMMENT. This species is distinguished by its relatively long apical flagellomere and malegenitalia, the latter resembling those of the nigricornis-group (sensu Hodkinson, 1981) in thePalaearctic Region. Trioza tenuissp. n. (Figs 36, 37, 43, 265-267) DESCRIPTION. Adult. Integument very sparsely covered with very short setae. Head, in profile (Fig. 37), at90 to longitudinal axis of body, from above (Fig. 36) almost as wide as mesoscutum; occipital margin sharpdorsomedially and strongly indented by median suture of vertex; vertex also deeply cleft by median suture,lateral concavities weak; median ocellus just visible in dorsal view, frons visible between genae in anteriorview; genal cones very poorly developed, rounded; antennal flagellum 1-51-1-63 times longer than headwidth, a single rhinarium present subapically on flagellomeres 2,4,6 and 7, apical flagellomere with a longpointed seta and a short truncate seta apically; clypeus with a pair of setae, ultimate rostral segment longand densely setose. Thorax (Figs 36, 37) moderately arched; from above the pronotum is more clearlyvisible at its lateral margins, the mesopraescutum being strongly produced forward medially, in profilepronotum strongly rounded down behind occiput and mesopraescutum angled down to pronotum;forewing (Fig. 265) elongate oval, strongly narrowing to acutangular apex, 3-08-3-19 times longer thanwide, radular areas elongate triangular, remainder of membrane devoid of spinules; veins bearing shortsetae, R branch acutangular, M branching proximal to Rs Cw la line, Cu stem 0-28-0-35 times as long asCw lb , ra t cell value 2-33-2-55, cu\ cell value 1-01-1-08; forewing 1-71-1-75 times longer than hindwing,costal margin of hindwing with up to four setae proximal to costal break, setae distal to costal break clearlydivided into two groups; hind coxa with a well-developed meracanthus and an extremely well-developedanterior lobe; hind tibia with a well-developed basal spine and one outer and three (rarely two on one tibiaonly) inner spurs apically. Abdomen with setae on tergites 2 and 3 in cf and 3 and 4 in $ ; cf proctigerweakly expanded laterally, paramere and aedeagus as in Figs 266, 267; $ genital segment short, conical,subgenital plate with truncate apex, ovipositor valves smooth. Measurements (2 cf , 2 $). Maximum width of head, cf and $ 0-67-0-70; length of antennal flagellum,Cf 1-09, 9 1-04-1 -06; length of ultimate rostral segment, Cf 0-19-0-20, $ 0-20; length of forewing, cf 4-4, $4-66-4-74; length of hind tibia, cf 0-64-0-66, $ 0-65-0-66. Larva unknown. HOST PLANT. Few adults collected from Haplocoelum foliolosum (Sapindaceae). Holotype cf, Angola: Bruco, 28.ii.-2.iii. 1972, Haplocoelum foliolosum (D. Hollis) (BMNH; drymounted). Paratypes. Angola: 5 cf , 2 $, same data as holotype; 1 $, swept in riverine forest area, 26-29. ii. 1972;1 cf , Rocadas, 30.iii.1972 (D. Hollis). (BMNH; slide and dry mounted and stored in 80% ethanol.) COMMENTS. This species shares many primitive character states with ghanaensis (p. 62) but hasa more slender build, the forewing is narrower and the hind tibia has three inner apical spurs. Trioza theronisp. n. (Figs 24, 268-271) DESCRIPTION. Adult. Integument sparsely covered with short setae. Head, in profile, at almost 90 tolongitudinal axis of body, from above as wide as mesoscutum; occipital margin rounded; vertex pentago-nal, with a well-developed irregular concavity on either side of a shallow median suture, frontal lobesweakly developed; median ocellus just visible from above, frons just visible between genae in anteriorview; genae slightly swollen and genal cones poorly developed; antennal flagellum 2-60-3-23 times longerthan head width, 1st flagellomere (Fig. 24) with an apical group of 1-3 rhinaria and often with an extrarhinarium two-thirds along its length, a single rhinarium present subapically on flagellomeres 2, 4, 6 and 7,apical flagellomere with one long and one short and truncate seta apically; clypeus with several short setae,ultimate rostral segment with two pairs of setae. Thorax weakly arched; pronotum clearly visible fromabove but its anterior margin strongly rounded down behind occiput; mesopraescutum, from above,almost as long as wide, its anterior margin strongly arcuate and angled down to pronotum; forewing (Fig.268) elongate oval with narrowly rounded apex, 2-61-3-06 times longer than wide, radular areas narrowtriangular, remainder of membrane devoid of spinules except for cell cu 2 ; veins bearing short setae, Rbranch acutangular, M branching proximal to Rs-Cu la line, Cu stem 0-80-1-12 times as long as Cw lb , m l 68 DAVID HOLLIS cell value 1-79-2-30, cu\ cell value 0-78-0-97; forewing 1 -45-1 -54 times longer than hindwing, costal marginof hindwing with 1-2 setae proximal to costal break, setae distal to costal break clearly divided into twogroups; hind coxa with a well-developed meracanthus and without anterior lobe; hind tibia with 1-3 basalspines and one outer and two inner apical spurs. Abdomen with setae on tergite 3 in cf and 4 in $; cfproctiger, paramere and aedeagus as in Figs 269-271; $ genital segment short, conical, proctiger with ashallow transverse groove distal to anus, ventral valve of ovipositor with two small teeth on apicoventralsurface. Measurements (7 cf , 7 $). Maximum width of head, d" 0-42-0-59; $ 0-51-0-60; length of antennalflagellum, cf 1-20-1 -85,$ 1-41-1-71; length of ultimate rostral segment, cf 0-12-0-16, $ 0-14-0-17; lengthof forewing, cf 2-38-3-64, $ 3-00-3-96; length of hind tibia, cf 0-39-0-53, $ 0-45-0-55. Larva and host plant unknown. Holotype cf , South Africa: Cape Town, Milnerton, ii.1926 (R. E. Turner) (BMNH; dry mounted). Paratypes. South Africa: lcf,l $,samedataasholotype;6 $, 14-28.xii.l925(/?. E. Turner);l $,CapeProvince, Ceres, 13.xii.1969, swept; 2 cf, 3 $, 24.x. 1970; 3 cf, 3 $, Rawsonville, 7.xi.l970; 2 cf, 2 $>,l.i.1971; 1 Cf, 2 $, Stellenbosch, 3.xii.l970; 1 $, 4.U971; 1 cf , Knysna, 15. i. 1972 (J. G. Theron); 2 cf ,Natal, Kloof, viii.1926 (R. E. Turner). (BMNH; NCI; slide and dry mounted and stored in 80% ethanol.) COMMENTS. The arrangement of supplementary rhinaria on the first flagellomere in theroni isdistinctive, but the species shows no apparent relationships to other triozids which havedeveloped supplementary rhinaria. T. schroederi (p. 65} is superficially similar to theroni buthas well-developed genal cones and a primitive arrangement of rhinaria. Trioza tundavalaesp. n. (Figs 272-274, 323, 324) DESCRIPTION. Adult. Integument densely covered with long setae. Head, in profile, slightly depressed fromlongitudinal axis of body, from above narrower than mesoscutum; occipital margin obtusangular; vertexquadrate, its anterior margin rounded downwards and deeply incised by median suture, lateral concavitiesshallow but clearly defined; median ocellus visible from above, frons visible in anterior view; genal conesshort, broadly rounded, in profile depressed from plane of vertex; antennal flagellum 2-2 times longer thanhead width, a single rhinarium present subapically on flagellomeres 2, 4, 6 and 7, apical flagellomere with along pointed seta and a short truncate seta apically; clypeus with a pair of setae, ultimate rostral segmentwith four pairs of setae. Thorax strongly arched; pronotum clearly visible from above, in profile roundeddown to occiput; mesopraescutum, from above, slightly wider than long, anterior margin broadly arcuate;forewing (Fig. 272) ovoid with rounded apex, 2-33 times longer than wide, radular areas triangular,remainder of membrane devoid of spinules, course of/? and RI brown pigmented; veins bearing long setaein proximal third of wing, R branch acutangular, RI about 1-5 times longer than R stem, M branchingproximal to RsCui a line, Cu stem 0-88-1-11 times as long as Cw lb , m\ cell value 1-44-1-48, cu\ cell value1-29-1-34; forewing 1-72-1-76 times longer than hindwing, costal margin of hindwing with up to 11 setaeproximal to costal break, setae distal to costal break clearly divided into two groups; hind coxa with awell-developed meracanthus and without anterior lobe; hind tibia without basal spine and with one outerand 2-3 inner apical spurs, cf unknown; $ abdomen with setae on tergites 3 and 4; $ genital segment (Fig.273) short; proctiger (Fig. 274) short, rounded, emarginate posteriorly, anal pore ring of wax-producingcells incomplete anteriorly and forming multiple rows posteriorly; subgenital plate very short, withtruncate posterior margin; ventral valve of ovipositor triangularly expanded apically and this expansionwith 15-16 saw-like teeth on ventrolateral surface, dorsal surface with serrations. Measurements (2 9). Maximum width of head, 0-83; length of antennal flagellum, 1-84; length ofultimate rostral segment, 0-17; length of forewing, 6-13-6-25; length of hind tibia, 0-95-0-99. Fifth instar larva (Figs 323, 324). Dorsal surface outline very broadly oval with indentations on eitherside at base of abdomen and a weak indentation medially on the posterior margin, about 1-2 times longerthan wide. Antenna with six flagellomeres. Cephaloprothorax separate from rest of thorax which hasmesothorax and metathorax almost completely divided. Forewing pad about 2-1 mm long, humeral lobeextending forward to anterior margin of eye which is close to anterior margin of cephaloprothorax. Caudalplate 0-6 times as long as wide, anus ventral and close to posterior margin of abdomen, anal pore area as inFig. 324. Narrow, elongate, truncate, tubular sectasetae, with roundly swollen bases, forming an evendense marginal fringe; lanceolate postocular sectaseta present, sectasetae absent from dorsum. AFROTROPICAL JUMPING PLANT LICE 69 HOST PLANT. Larvae collected in midrib galls on upper surfaces of leaves of Syzygiumbenguellense (Myrtaceae); adults emerged from these galls. Holotype $, Angola: Tundavala, 8-10 mis NW. Sa da Bandeira, 27-29. iii. 1972, emerged from leaf gall ofSyzygium benguellense (D. Hollis) (BMNH; dry mounted).Paratypes. 3 $ , larvae, same data as holotype (BMNH; slide and dry mounted). COMMENT. This large, robust, pilose species has the distinctive feature that R { is considerablylonger than R stem, giving r 1 a very broad proximal area. It has a similar general appearance toTrioza palaquii (Laing, 1930), described from specimens reared from galls of Palaquium gutta(Sapotaceae) from Malaya, but this may be convergence as both species live in enclosed galls. Doubtful speciesTrioza bussei Zacher, 1915 Trioza bussei Zacher, 1915: 526; 1916: 419. Syntypes, larvae and adults, CAMEROON: Soppo, on Kickxia(not traced). The first published indication of this species was a brief description of the larva and a figure of theadult wing. Zacher (1916) gives a further description of the egg, all larval stages and the adultfemale. No diagnostic features are mentioned. The type-series was not traced and no reply wasreceived to enquiries at the Institut fur Pflanzenschutzforschung, Eberswalde. Trioza similis Heslop-Harrison, 1961 Trioza similis Heslop-Harrison, 1961: 530. Holotype $, SOUTH AFRICA: 'Cape Prov., Tzitzikama Forest,Stormsrivierpiek, 13. i. 1951, Loc. no. 137. In indigenous forest' (lost). The original description and fig. 12 (I) mentions an anteriorly projecting epiphysis on the ventralsurface of the subgenital plate, but this is not shown in fig. 12 (H). No other diagnostic featuresare mentioned. References Akanbi, M. 0. 1981. Preliminary notes on Triozamia lamborni (Newstead) (Hem. Psyllidae), a potentially dangerous pest oiAntiaris africana. Entomologist's Monthly Magazine 116: 113-115, 1 fig.Ashmead, W. H. 1881. On the Aphididae of Florida with descriptions of new species. (Paper No. 2). Canadian Entomologist 13: 220-225.Awadallah, K. T. & Swailem, S. M. 1971. On the bionomics of the Sycamore Fig psyllid Pauropsylla trichaeta Pettey. Bulletin de la Societe Entomologique d'Egypte 55: 193-199, 1 fig.Bekker-Migdisova, E. E. 1973. On the systematics of the Psyllomorpha and the position of the group within the Order Homoptera. Chteniya Pamyati Nakolaya Aleksandruriya Kholodovskogo 1971 (Leningrad): 90-177, 2 figs. [In Russian. English translation available from British Library Lending Division, no. RTS 8526.]Berg, C. C. 1977. Revisions of African Moraceae (excluding Dorstenia, Ficus, Musanga and Myrianthus). Bulletin du Jar din botanique national de Belgique 47: 267-407, 35 figs.Blanchard, E. 1852. V. Afidideos. In Gay, C. (Ed.), Historia Fisica y Politica de Chile, Zoologia 7: 306-320, pi. 3.Bordage, E. 1898. Notes d'entomologie agricole tropicale. II. Un ennemi du Vanillier. Revue Agricole de la Reunion!: 524-525.1914. Notes biologiques recueillies a 1'Ile de la Reunion. Bulletin Scientifique de la France et de la Belgique (7) 47: 407-410.Boselli, F. B. 19300. Studii sugli Psyllidi (Homoptera: Psyllidae o Chermidae), VI. Psyllidi di Formosa raccolti dal Dr. R. Takahashi. Bollettino del Laboratorio di Zoologia Generate et Agraria del R. Istituto Superiore Agrario di Portici 24: 175-210, 17 figs.19306. Studii sugli Psyllidi. IX. Descrizione di una Triozina galligena su agrumi in Eritrea. Bollettino del Laboratorio di Zoologia Generale et Agraria del R. Istituto Superiore Agrario di Portici 24: 228-232 , 2 figs.1931. Studii sugli Psyllidi (Homoptera: Psyllidae o Chermidae). X. Istituzione di un nuovo genere e 70 DAVID HOLLIS descrizione di Egeirotrioza ceardi (De Bergevin) euphratica n. var., Triozina galligena su Populuseuphratica in Mesopotamia. Bollettino del Laboratorio di Zoologia Generate e Agraria del R. IstitutoSuper lore Agrario di Portici 24: 267-278, 6 figs. Buckton, G. B. 1900. Description of a new species of Psylla destructive to forest trees. Indian MuseumNotes 5: 35-36. Caldwell, J. S. 1940. New genera and species of jumping plant-lice from the Hawaiian Islands withdescriptions of several immature stages (Homoptera: Psyllidae). Proceedings of the Hawaiian Entomo-logical Society 10: 389-397, 1 fig., pis 22, 23. - 1944. Psyllidae from tropical and semitropical America (Homoptera). Journal of the New YorkEntomological Society 52: 335-340, 1 pi. Capener, A. L. 1970. Southern African Psyllidae (Homoptera) - 1: A check list of species recorded fromSouth Africa, with notes on the Pettey collection. Journal of the Entomological Society of SouthernAfrica 33: 195-200. 1973. Southern African Psyllidae (Homoptera) - 3: A new genus and new species of South AfricanPsyllidae. Journal of the Entomological Society of Southern Africa 36: 37-61 , 123 figs. Carmin, J. 1951. Two new species of Pauropsylla (Psyllidae, Rhynchota). Bulletin of the IndependentBiological Laboratories Kefar-Malal9(2): 1-3, 10 figs. Catling, H. D. 1969a. The bionomics of the South African citrus psylla, Trioza erytreae (Del Guercio)(Homoptera: Psyllidae) 1. The influence of the flushing rhythm of citrus and factors which regulateflushing. Journal of the Entomological Society of Southern Africa 32: 191-208, 9 figs.19696. The bionomics of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera:Psyllidae) 2. The influence of parasites and notes on the species involved. Journal of the EntomologicalSociety of Southern Africa 32: 209-223, 4 figs. 1969c. The bionomics of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera:Psyllidae) 3. The influence of extremes of weather on survival. Journal of the Entomological Society ofSouthern Africa 32: 273-290, 8 figs. 1970. The bionomics of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera:Psyllidae) 4. The influence of predators. Journal of the Entomological Society of Southern Africa 33:341-348, 9 figs. - 1971. The bionomics of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera:Psyllidae) 5 . The influence of host plant quality. Journal of the Entomological Society of Southern Africa34: 381-391, 2 figs. 1972. The bionomics of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera: Psyllidae) 6. Final population studies and a discussion of population dynamics. Journal of the Entomolo-gical Society of Southern Africa 35: 235-251, 4 figs. 1973. Notes on the biology of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera: Psyllidae). Journal of the Entomological Society of Southern Africa 36: 299-306, 4 figs.Crawford, D. L. 1910. American Psyllidae I (Triozinae). Pomona College Journal of Entomology 2:228-237, figs 98-99. - 1911. American Psyllidae IV (A partial revision of subfamilies). Pomona College Journal ofEntomology 3: 480-503, figs 157-160. - 1914. A monograph of the jumping plant-lice or Psyllidae of the New World. Bulletin of the UnitedStates National Museum 85: ix+ 186pp. , 30 pis. - 1915. Ceylonese and Philippine Psyllidae (Homoptera). Philippine Journal of Science (D) 10 (4):257-267, 1 pi. 1918. The jumping plant-lice (Family Psyllidae) of the Hawaiian Islands. A study in insect evolution.Proceedings of the Hawaiian Entomological Society 3: 430-456, pi. 8. 1919. The jumping plant-lice of the palaeotropics and the south Pacific islands. Family Psyllidae, or Chermidae, Homoptera. Philippine J ournal of Science 15: 139-207, 3 pis. 1927. Hemiptera. Psyllidae (Chermidae). Insects of Samoa 2 (1): 29-33, 4 figs. Dahlgren, R. 1977. A commentary on a diagrammatic presentation of the angiosperms in relation to the distribution of character states. Plant Systematics and Evolution, Supplement 1: 253-283, 12 figs.Dc.bski, B. 1918 [?]. Liste des cecides signalees en Egypte jusqu'a ce jour. Memoires de la Societe Entomologique d' Egypte 1 (4): 3-38.Del Guercio, G. 1918. Note ed osservazioni di entomologia agraria. II cecidio delle foglie del limone ed il suo cecidozoo in Eritrea. Agricultura Colonials 1918: 167-169, 2 figs, 1 pi.Dobreanu, E. & Manolache, C. 1962. Homoptera Psylloidea. Fauna Republicii Populare Romine Insecta 8 (3): 1-376, 270 figs. AFROTROPICAL JUMPING PLANT LICE 71 Eastop, V.F. 1961. Some aspects of psyllid taxonomy. Annals and Magazine of Natural History (13) 4: 169.Edwards, J. 1896. The Hemiptera-Homoptera of the British Islands, Psyllina: 224261, pis 26-30. London.Emden, H. F. van (Ed.) 1972. Aphid technology with special reference to the study ofaphids in the field, xiv + 344pp. London.Enderlein, G. 1918. Psyllidologica V. Zoologische Jahrbiicher (Systematik) 41: 479-486, 7 figs, pi. 7. 1921. Psyllidologica VI. Zoologischer Anzeiger 52: 115-122, 2 figs. 1926. Psyllidologica VIII. Entomologische Mitteilungen 15: 397-401. Foerster, A. 1848. Uebersicht der Gattungen und Arten in der Familie der Psylloden. Verhandlungen des naturhistorischen Vereins der preussischen Rheinlande und Westfalens 5: 65-94.Ferris, G. F. 1926. Observations on the Chermidae (Hemiptera; Homoptera). Part III. Canadian Entomologist 58: 13-20, 5 figs.Froggatt, W. W. 1901. Australian Psyllidae. Part II. Proceedings of the Linnean Society of New South Wales 1901: 242-298, pis 14-16.Harris, W. V. 1936. Notes on two injurious psyllids and their control. East African Agricultural Journal 1: 498-500, 4 figs.Heslop-Harrison, G. 1958. Subfamily separation in the homopterous Psyllidae - III (a-c). Annals and Magazine of Natural History (13) 1: 561-579. - 1961. Hemiptera (Homoptera) Psyllidae. South African Animal Life 8: 487-532, 12 figs.Hodkinson, I. D. 1981. Status and taxonomy of the Trioza (Bactericera) nigricornis Forster complex (Hemiptera: Triozidae). Bulletin of Entomological Research 71: 671-679, 27 figs.Hollis, D. 1976. Jumping plant lice of the tribe Ciriacremini (Homoptera: Psylloidea) in the Ethiopian Region. Bulletin of the British Museum (Natural History) (Entomology) 34: 1-83, 191 figs.Kieffer, J. J. 1905. Etude sur de nouveaux insectes et phytoptides gallicoles du Bengale. II. Psyllides. Annales de la Societe Scientifique de Bruxelles 29: 159-182, figs. 5-14. 1908. Description de galles et d'insectes gallicoles d'Asie. Marcellia 7: 149-167, 4 figs, pis 3, 4. Kirkaldy, G. W. 1904. Bibliographical and nomenclatorial notes on the Hemiptera -No. 3. Entomologist 37: 279-283. Klimaszewski, S. M. 1962 Bemerkungen iiber einige neotropische Blattfloh-Arten (Homoptera, Psylli-dae). Bulletin de V Academic Polonaise des Sciences. Serie de Sciences Biologiques 10: 251-255, 9 figs.1964. Studien iiber die Systematik der Unterordnung Psylloidea. Annales Zoologici, Warszawa 22: 81-138. 1968. Stosunki pokrewienstwa srodkowoeurope jskich gatungkow z rodzaju Trioza Forst. (Homop-tera, Psyllodea) w swietle badari metodami taksonomii numerycznej . Annales Universitatis Mariae Curie -Sklodowska Lublin (C) 22: 1-20. 1973. The jumping plant lice or psyllids (Homoptera, Psyllodea) of the Palaearctic. An annotated check-list. Annales Zoologici, Warszawa 30: 155-286. 1975. Psylloidea koliszki (Insecta: Homoptera). Fauna Polski 3: 1-294, 512 figs. Kuwayama, S. 1910. Die psylliden Japans. II. Transactions of the Sapporo Natural History Society 3: 53-66 + [lj.pl. 2.Kwon, Y. J. 1983. Psylloidea of Korea (Homoptera: Sternorrhyncha). Insecta Koreana, Series 2: [i] + 1-181, 32 pis.Laing, F. 1923. On some Psyllidae (Hem. -Horn.) from the New World. Annals and Magazine of Natural History (9) 11: 696-705, 7 figs.1930. Some records of Indo-Malayan Psyllidae. Indian Forest Records (Entomological Series) 14 (8): 35-44, 5 figs.Loginova, M. M. 1964. Psyllinea. In Bei-Bienko, G. Ya. (Ed.), Keys to the insects of the European USSR.l. Opredeliteli po Faune SSR 84: 437-482, 220 figs. [In Russian, English translation, Israel program for scientific translations, Jerusalem 1967: 551-608.]1970. New species of psyllids (Homoptera, Psylloidea) from Middle Asia. Entomologischeskoe Obozrenie 49: 601-623, 19 figs. [In Russian, English translation, Entomological Review, Washington 49: 370-385(1971)].\972a. Revision of jumping plant lice of the tribe Pauropsyllini Crawf. (Homoptera, Psylloidea, Carsidaridae). Entomologischeskoe Obozrenie 51: 837-853, 49 figs. [In Russian, English translation, Entomological Review, Washington 51: 497-505 (1973)]. 19726. On the fauna of Psylloidea (Homoptera) from Morocco. Commentationes Biologicae 47: 1-37, 121 figs. Lounsbury, C. P. 1897. Psyllidae or jumping plant lice. Citrus Psylla. Report of the GovernmentEntomologist, Cape of Good Hope 1896: 115-118, 2 figs. 72 DAVID HOLLIS Low, F. 1879. Zur Systematik der Psylloden. Verhandlungen der kaiserlichkoniglichen zoologisch-botanischen Gesellschaft in Wien 28: 585-610, pi. 9. - 1886. Neue Beitrage zur Kenntniss der Psylliden. Verhandlungen der kaiserlich-koniglichen zoolo-gisch-botanischen Gesellschaft in Wien 36: 149-170, pi. 6. Maskell, W. M. 1879. On some Coccidae in New Zealand. Transactions and Proceedings of the New Zealand Institute (Zoology) 11: 187-228, pis 5-8. 1890. On some species of Psyllidae in New Zealand. Transactions and Proceedings of the New Zealand Institute 21: 157-170, pis 10-12. Mathur, R. N. 1975. Psyllidae of the Indian subcontinent, xii + 429 pp., 6 pis, 136 figs. New Delhi.\k Daniel. J. R. & Moran, V. C. 1972. The parasitoid complex of the Citrus Psylla Trioza erytreae (Del Guercio) [Homoptera: Psyllidae]. Entomophaga 17: 297-317, 3 figs.Miyatake, Y. 1972. Studies on the Philippine Psyllidae (Hemiptera: Homoptera) II. Bulletin of the Osaka Museum of Natural History 26: 11-34, 6 figs, 2 pis. - 1981. Studies on Psyllidae of Nepal I. Results of the survey in the Kathmandu Valley, 1979. Part 1(Hemiptera: Homoptera). Bulletin of the Osaka Museum of Natural History 34: 47-60, pi. 5, 7 figs. Moran, V. C. 1968a. The development of the citrus psylla, Trioza erytreae (Del Guercio) (Homoptera:Psyllidae), on Citrus limon and four indigenous host plants. Journal of the Entomological Society ofSouthern Africa 31: 391^02. - 1968ft. Preliminary observations on the choice of host plants by adults of the citrus psylla, Triozaerytreae (Del Guercio) (Homoptera: Psyllidae). Journal of the Entomological Society of Southern Africa31:403^410, 1 fig. Moran, V. C. & Blowers, J. R. 1967. On the biology of the South African citrus psylla, Trioza erytreae (Del Guercio) (Homoptera: Psyllidae). Journal of the Entomological Society of Southern Africa 30: 96-106, 8 figs.Moran, V. C. & Brown, R. P. 1973. The antennae, host plant chemoreception and probing activity of the citrus psylla Trioza erytreae (Del Guercio) (Homoptera: Psyllidae) . Journal of the Entomological Society of Southern Africa 36: 191-202, 7 figs. Moran, V. C. & Buchan, R. P. 1975. Oviposition by the citrus psylla, Trioza erytreae (Homop-tera: Psyllidae), in relation to leaf hardness. Entomologia Experimental et Applicata 18: 96-104, 3 figs.Mound, L. A. & Halsey, S. H. 1978. Whitefly of the world. A systematic catalogue of the Aleyrodidae (Homoptera) with host plant and natural enemy data, [vi] + 340 pp. Chichester.Newstead, R. 1914. VI. Homoptera (Psyllidae and Coccidae) collected in the Lagos District by W. A. Lamborn. Transactions of the Entomological Society of London 1913: 520-524, pi. 29.Orian, A. 1972. The Psylloidea of Mauritius, with a description of Trioza eastopi sp. nov. Fauna of Mauritius 1 (1): [i] + 1-6 + [ii], 10 pis.Oshanin, B. 1912. Katalog der paldarktischen Hemipteren (Heteroptera, Homoptera- Auchenorhyncha und Psylloidea), xvi + 187 pp. Berlin.Pettey, F. W. 1923. A new species of psyllid. South African Journal of Natural History 4: 30-33, 6 figs. - 1924. South African psyllids. Entomology Memoirs. Department of Agriculture, Union of South4/nco2:21-30,plsl-3. - 1925. New South African psyllids. South African Journal of Natural History 5: 125-142, pis 12, 13.1933. New species of South African psyllids, III. Entomology Memoirs. Department of Agriculture, Union of South Africa 8: 1-23, 2 pis.Pflugfelder, O. 1941. Psyllina. In Bronns, H. G., Klassen und Ordnungen des Tierreichs. Funfter Band: Arthropoda. 3. Abteilung: Insecta. VIII. 1-95, 75 figs. Leipzig.Puton, A. 1876. Notes pour servir a 1'etude des Hemipteres. 3 e Partie. Annales de la Societe Entomologique de France (5) 6: 275-290. - 1886. Catalogue des Hemipteres (Heteropteres, Cicadines et Psyllides) de la faune palearctiques.Revue d'Entomologie, Caen 5 (4-7): 1-100 [special pagination]. Ramirez-Gomez, C. 1960. Los Psilidos de Espana (Conclusion). Boletin de la Real Sociedad Espanola de Historia Natural (Biologica) 57: 1-87, 4 pis.Riley, C. V. 1884. The Psyllidae of the United States. Proceedings of the American Association for the Advancement of Science 32: 319. - 1885. Notes on North American Psyllidae. Proceedings of the Biological Society of Washington 2:67-79. Roberts, H. 1969, Forest insects of Nigeria with notes on their biology and distribution. Institute Paper,Commonwealth Forestry Institute, Oxford 44: 1-206. AFROTROPICAL JUMPING PLANT LICE 73 Riibsaamen, Ew. H. 1899. Mitteilungen uber neue und bekannte Gallen aus Europa, Asien, Afrika und Amerika. Entomologische Nachrichten, Berlin 25: 225-282, 2 pis, 18 figs.Samy, O. 1972. Psyllids of Egypt [Homoptera: Psyllidae]. Bulletin de la Societe Entomologique d'Egypte 56: 437-480, 88 figs.Scott, J. 18820. XVII. Description of a new genus and two new species of Psyllidae from South America. Transactions of the Entomological Society of London 1882: 443-448, pi. 18. - 18826. On certain genera and species of the group Psyllidae in the collection of the British Museum.Transactions of the Entomological Society of London 1882: 449-473, pis 18, 19. Sulc, K. 1909. Trioza cockerelli n. sp. , a novelty from North America, being also of economic importance.Casopis Ceske Spolecnosti Entomologicke 6: 102-108, 14 figs. - 1910. Monographia generis Trioza Foerster. Species Regionis Palaearcticae. Pars. I., No. 1-10.Sitzungsberichte der Koniglichen Bdhmischen Gesellschaft der Wissenschaften 1910 (17): 1-34, pis 1-10. - 1911. Monographia generis Trioza Foerster. Species Regionis Palaearcticae. Pars II., No. 11-20.Sitzungsberichte der Koniglichen Bdhmischen Gesellschaft der Wissenschaften 1911 (5): 1-34, pis 11-20. - 1912. Monographia generis Trioza Foerster. Species Regionis Palaearcticae. Pars III., No. 21-35.Sitzungsberichte der Koniglichen Bdhmischen Gesellschaft der Wissenschaften 1912 (16) : 1-63 , pis 21-35 . 1913. Monographia generis Trioza Foerster. Species Regionis Palaearcticae. Pars IV., No. 36-49. Sitzungsberichte der Koniglichen Bdhmischen Gesellschaft der Wissenschaften 1913 (1): 1-48, pis 36-68. Takahashi, R. 1932. Aleyrodidae of Formosa. Part 1. Report of the Department of Agriculture, Govern-ment Research Institute, Formosa, Japan 59: 1-52, 33 figs. Thomson, C. G. 1877. XXIX. Ofversigt af Skandinaviens Chermes-arter. Opuscula Entomologica (EdititC. G. Thomson), TrelleborgS: 820-841. Thome, R. F. 1976. A phylogenetic classification of the Angiospermae. Evolutionary Biology 9: 35-106, 4figs. Tuthill, L. D. 1938. Some new North American Psyllidae (Homoptera). Entomological News 49: 241-246,pi. 5. - 1939. Neotriozella and a new related genus (Homoptera: Psyllidae). Bulletin of the BrooklynEntomological Society 34: 51-54. - 1943. The psyllids of America north of Mexico (Psyllidae: Homoptera) (Subfamilies Psyllinae andTriozinae). Iowa State College Journal of Science 17: 443-660, 313 figs. - 1944. Contributions to the knowledge of the Psyllidae of Mexico. Journal of the Kansas Entomologi-cal Society 17: 143-159, figs 1-35. - 1945. Contributions to the knowledge of the Psyllidae of Mexico. Journal of the Kansas Entomologi-cal Society 18: 1-29, figs 36-93. - 1952. On the Psyllidae of New Zealand (Homoptera). Pacific Science 6: 83-125, 32 figs. 1956. A new genus of Psyllidae from Hawaii (Homoptera). Proceedings of the Hawaiian Entomolo- gical Society 16: 158-161, 2 figs. - 1959. Los Psyllidae del Peru Central (Insecta: Homoptera). Revista Peruana de EntomologicaAgricola 2 (2): 1-27, 28 figs. 1964. Conocimientos Adicionales sobre los Psyllidae (Homoptera) del Peru. Revista Peruana de Entomologia 7(1): 25-32, 11 figs.Tuthill, L. D. & Taylor, K. L. 1955. Australian genera of the family Psyllidae (Hemiptera: Homoptera). Australian Journal of Zoology 3: 227-251, 22 figs.Uichanco, L. B. 1921 . New records and species of Psyllidae from the Philippine Islands, with descriptions of some preadult stages and habits. Philippine Journal of Science 18: 259-288, 5 pis.Van Der Merwe, C. P. 1941 . The Citrus Psylla (Spanioza erytreae, del G.). Science Bulletin, Union of South Africa, Department of Agriculture and Forestry 233 (Entomology Series No. 8): 1-12, 1 pi.Vondracek, K. 1957. Mery-Psylloidea. Fauna CSR 9: 1-431, 265 figs. - 1963. Jumping plant-lice (Psylloidea - Homoptera) of Central Africa. Part 1 (Congo). Acta Entomologica Musei Nationalis Pragae 35: 263-290, 84 figs.Walker, F. 1858. List of specimens of homopterous insects in the collection of the British Museum. Supplement. [4] + 369 pp. London. Waterston, J. 1922. On the chalcid parasites of psyllids (Hemiptera, Homoptera). Bulletin of Entomologi-cal Research 13: 41-58, 7 figs.White, I. M. 1980. Nymphal taxonomy and systematics of the Psylloidea (Insecta: Homoptera). [xv], 340, 12, 8, 4, 20, 8, 1pp. Unpublished Ph.D. thesis, CNAA, Liverpool Polytechnic.Willcocks, F. C. 1922 [?]. A survey of the more important economic insects and mites of Egypt, viii + 482 + [i] pp. Cairo. 74 DAVID HOLLIS Zacher, F. 1915. Beitrag zur Kenntnis der westafrikanischen Pflanzenschadlinge. Tropenpflanzer, Berlin 18: 504-534, 35 figs. 1916. Neue und wenig bekannte Pflanzenschadlinge aus unseren Kolonien. Zeitschriftfiir angewandte Entomologie 3: 418-425, 15 figs.Zimmermann, E. C. 1948. Homoptera: Sternorhyncha. Superfamily Psylloidea. Insects of Hawaii 5- 12-38, 15 figs. Figs 56-64 Triozamia species. 56-58, T. usambarensis; 56, forewing; 57, paramere; 58, apical segmentsof aedeagus. 59-62, T. vondraceki; 59, cf proctiger, lateral view; 60, paramere; 61, apical segments ofaedeagus; 62, $ proctiger, dorsal view. 63, 64, T. lamborni, 63, paramere; 64, apical segments ofaedeagus. AFROTROPICAL JUMPING PLANT LICE 75 O5 O-5 68 71 Figs 65-71 Afrotropical Triozidae. 65-68, Afrotrioza bersama; 65, forewing; 66, d" proctiger, lateralview; 67, paramere; 68, apical segment of aedeagus. 69-71, Trichochermes insleyi; 69, forewing; 70,paramere; 71, apical segment of aedeagus. 76 DAVID HOLLIS 05 Figs 72-79 Pauropsylla willcocksi-group. 72-74, P. willcocksi; 72, O" proctiger, posterior view; 73, apicalsegment of aedeagus; 74, $ genital segment, lateral view. 75-78, P. trichaeta; 75, forewing; 76, cfproctiger, posterior view; 77, paramere; 78, apical segment of aedeagus. 79, P. tatrichea, genitalsegment, lateral view. AFROTROPICAL JUMPING PLANT LICE 77 82 88 86 89 Figs 80-92 Pauropsylla species. 80-82, P. trigemma; 80, forewing; 81, paramere; 82, apical segment ofaedeagus. 83, P. ngongae, forewing. 84, 85, P. breviantennata, 84, forewing; 85, C? genitalia, lateralview. 86-89, P. septima; 86, forewing; 87, paramere; 88, apical segment of aedeagus; 89, medianposterior margin of 9 subgenital plate, ventral view. 90-92, P. proxima; 90, paramere; 91, apicalsegment of aedeagus; 92, median posterior margin of $ subgenital plate, ventral view. 78 DAVID HOLLIS 102 105 96 103 106 107 100 104 108 Figs 93-108 Pauropsylla species. 93, 94, P. angolensis; 93, paramere; 94, apical segment of aedeagus.95-97, P. eastopi; 95, paramere; 96, apical segment of aedeagus; 97, c? proctiger, posterior view.98-101, P. longipes; 98, paramere; 99, apical segment of aedeagus; 100, C? proctiger, posterior view;101, 9 genital segment, lateral view. 102-104, P. mistura; 102, paramere; 103, apical segment ofaedeagus; 104, 9 genital segment, lateral view. 105-108, P. senegalensis , 105, paramere; 106, apicalsegment of aedeagus; 107, c? proctiger, posterior view; 108, 9 genital segment, lateral view. AFROTROPICAL JUMPING PLANT LICE 79 125 Figs 109-125 Trioza erytreae-group, cf genitalia. 109-111, T. erytreae; 109, proctiger, lateral view; 110,paramere; 111, apical segment of aedeagus. 112, 113, T. catlingi; 112, paramere; 113, apical segment ofaedeagus. 114, 115, T. gregoryi; 114, paramere; 115, apical segment of aedeagus. 116, 117, T. ata; 116,paramere; 117, apical segment of aedeagus. 118, T. kilimanjarica,parameTe. 119-121, T. tiliacora; 119,proctiger, lateral view; 120, paramere; 121, apical segment of aedeagus. 122-124, T. carvalhoi; 122,proctiger, lateral view; 123, paramere; 124, apical segment of aedeagus. 125, T. eafra, paramere. 80 DAVID HOLLIS 126 129 127 128 130 132 134 133 Figs 126-134 Trioza erytreae and litseae-groups. 126, 127, T. erytreae; 126, forewing; 127, $ genitalsegment, lateral view. 128, T. ata, apex of lower valve of ovipositor. 129, T. tiliacora, forewing. 130, T.carvalhoi, forewing. 131-134, T. xylopia; 131, forewing; 132, cf proctiger, lateral view; 133, paramere;134, apical segment of aedeagus. AFROTROPICAL JUMPING PLANT LICE 81 138 148 147 150 Figs 135-150 Trioza anomalicornis-group. 135-138, T. anomalicornis; 135, cf proctiger, lateral view;136, apical segment of aedeagus; 137, paramere; 138, 9 genital segment, lateral view. 139-141, T.kakamegae; 139, cf proctiger, lateral view; 140, apical segment of aedeagus; 141, paramere. 142-144, T.thibae; 142, cf proctiger, lateral view; 143, apical segment of aedeagus; 144, paramere. 145-147, T.tavandula; 145, cf proctiger, lateral view; 146, apical segment of aedeagus; 147, paramere. 148-150, T.luvandata; 148, cf proctiger, lateral view; 149, apical segment of aedeagus; 150, paramere. 82 DAVID HOLLIS Figs 151-161 Trioza neoboutonia-group. 151-155, T. neoboutonia; 151, forewing; 152, d" proctiger,lateral view; 153, apical segment of aedeagus; 154, paramere; 155, $ genital segment, lateral view.156-158, T. harteni; 156, forewing; 157, paramere; 158, $ genital segment, lateral view. 159-161, T.chiangae; 159, forewing; 160, paramere; 161, $ genital segment, lateral view. AFROTROPICAL JUMPING PLANT LICE 83 165 174 176 Figs 162-177 Trioza neoboutonia and hargrea ves /-groups. 162-164, T. bamendae; 162, d" proctiger,lateral view; 163, paramere; 164, apical segment of aedeagus. 165-168, T. dinaba; 165, forewing; 166, cfproctiger, lateral view; 167, paramere; 168, apical segment of aedeagus. 169-172, T. nachingweae; 169,forewing; 170, paramere; 171, apical segment of aedeagus; 172, 9 genital segment, lateral view.173-175, T. hargreavesi; 173, forewing; 174, paramere; 175, apical segment of aedeagus. 176, 177, T.mirificornis; 176, paramere; 177, apical segment of aedeagus. 84 DAVID HOLLIS Figs 17&-189 Trioza obsoleta-group. 178-182, T. afrobsoleta; 178, forewing; 179, hindwing; 180, C?proctiger, lateral view; 181, paramere; 182, apical segment of aedeagus. 183-187, T. gonjae; 183,forewing; 184, hindwing; 185, O" proctiger, lateral view; 186, paramere; 187, apical segment ofaedeagus. 188, 189, T. boxi; 188, forewing; 189, hindwing. AFROTROPIC^L JUMPING PLANT LICE 85 196 Figs 190-197 Trioza obsoleta-group. 190-193, T. afrosersalisia; 190, forewing; 191, cf proctiger, lateralview; 192, paramere; 193, apical segment of aedeagus. 194-197, T. mimusops; 194, forewing; 195, cfproctiger, lateral view; 196, paramere; 197, apical segment of aedeagus. DAVID HOLLIS 207 206 201 210 214 Figs 198-215 Trioza etiennei-group. 198-201, T. etiennei; 198, forewing; 199, c? proctiger, lateral view;200, apical segment of aedeagus; 201, paramere. 202-206, T. messaratina; 202, forewing; 203, pro-notum, anterodorsal view; 204, C? proctiger, lateral view; 205, apical segment of aedeagus; 206,paramere. 207-211. T. seranistama; 207, forewing; 208, cf proctiger, lateral view; 209, apical segment ofaedeagus; 210, paramere; 211, $ genital segment, lateral view. 212, 213, T. nestasimara; 212, apicalsegment of aedeagus; 213, paramere. 214, 215, T. earner ounensis; 214, apical segment of aedeagus; 215,paramere (slightly anterior view). AFROTROPICAL JUMPING PLANT LICE 87 221 220 223 224 Figs 216-224 Trioza etiennei and glabea-groups. 216-218, T. pitkini; 216, forewing; 217, paramere; 218,apical segment of aedeagus. 219-222, T. glabea; 219, forewing; 220, O" proctiger, lateral view; 221,paramere; 222, apical segment of aedeagus. 223, 224, T. usambarica; 223, paramere; 224, apical segmentof aedeagus. 88 DAVID HOLLIS 239 237 240 Figs 225-240 Trioza species. 225-228, T. capensis; 225, forewing; 226, cf proctiger, lateral view; 227,paramere; 228, apical segment of aedeagus. 229-232, T.ficicola; 229, forewing; 230, cf proctiger, lateralview; 231, paramere; 232, apical segment of aedeagus. 233-236, T. fuscivena; 233, forewing; 234,paramere; 235, apical segment of aedeagus; 236, $ genital segment, lateral view. 237-240, T.ghanaensis; 237, forewing; 238, cf proctiger, lateral view; 239, paramere; 240, apical segment ofaedeagus. AFROTROPICAL JUMPING PLANT LICE 89 253 Figs 241-253 Trioza species. 241-244, T. guiera; 241, forewing; 242, cf proctiger, lateral view; 243,paramere; 244, apical segment of aedeagus. 245-248, T. karroo; 245, forewing; 246, cf proctiger, lateralview; 247, apical segment of aedeagus; 248, $ genital segment, lateral view. 249-253, T. laingi; 249,forewing; 250, cf proctiger, lateral view; 251, paramere; 252, apical segment of aedeagus; 253, $ genitalsegment, lateral view. 90 DAVID HOLLIS Figs 254-264 Trioza species. 254, T. medleri, forewing. 255, Trioza sp. tmedleri, a" genitalia, lateral view(specimen from Zaire in MRAC). 256-260, T. schroederi; 256, forewing; 257, cf proctiger, lateral view;258, paramere; 259, apical segment of aedeagus; 260, $ genital segment, lateral view. 261-264, T.tangae; 261, forewing; 262, O" proctiger, lateral view; 263, paramere; 264, apical segment of aedeagus. AFROTROPICAL JUMPING PLANT LICE 91 272 274 Figs 265-274 Trioza species. 265-267, T. tennis; 265, forewing; 266, paramere; 267, apical segment ofaedeagus. 268-271, T. theroni; 268, forewing; 269, O" proctiger, lateral view; 270, paramere; 271, apicalsegment of aedeagus. 272-274, T. tundavalae; 272, forewing; 273, $ genital segment, lateral view; 274,9 proctiger, posterodorsal view. 92 DAVID HOLLIS 275 277 279 278 276 Figs 275-279 Afrotropical Triozidae, 5th instar larvae. 275-277, Triozamia lamborni; 275, dorsal view;276, dorsal (D) and ventral (V) views of caudal plate; 277, anal pore area. 278, 279, Afrotrioza bersama;278, dorsal view; 279, anal pore area. AFROTROPICAL JUMPING PLANT LICE 93 280 281 Figs 280-282 Pauropsylla willcocksi-group, 5th instar larvae. 280, 281, P. willcocksi; 280, dorsal view;281, anal pore area. 282, P. trichaeta, dorsal view. 94 DAVID HOLLIS 283 285 284 286 288 Figs 283-288 Pauropsylla species, 4th and 5th instar larvae. 283, 284, P. proximo; 283, 4th instar larva,dorsal view; 284, anal pore area of same. 285, 286, P. senegalensis; 285, 5th instar larva, dorsal view; 286,anal pore area of same. 287, 288, P. longipes; 287, 5th instar larva, dorsal view; 288, anal pore area ofsame. AFROTROPICAL JUMPING PLANT LICE 95 295 Figs 289-297 Trioza erytreae-group, 5th instar larvae. 289, 290, T. erytreae; 289, dorsal view; 290, analpore area. 291, 292, T. catlingi; 291, dorsal view; 292, anal pore area. 293, 294, T. tiliacora; 293, dorsalview; 294, anal pore area. 295, 296, T. carvalhoi; 295, dorsal view; 296, anal pore area. 297, T. capeneri,anal pore area. 96 DAVID HOLLIS Figs 298-303 Trioza species, 5th instar larvae. 298, 299, T. xylopia; 298, dorsal view; 299, anal pore area.300, 301, T. litseae; 300, dorsal view; 301, anal pore area. 302, 303, T. kakamegae; 302, dorsal view; 303,anal pore area. AFROTROPICAL JUMPING PLANT LICE 97 304 305 307 Figs 304-308 Trioza neoboutonia-group, 5th instar larvae. 304, 305, T. neoboutonia; 304, dorsal view;305, anal pore area. 306, 307, T. harteni; 306, dorsal view; 307, anal pore area. 308, T. chiangae, dorsalview. 98 DAVID HOLLIS 309 311 310 Figs 309-312 Trioza obsoleta-group, 5th instar larvae. 309, 310, T. gonjae; 309, dorsal view; 310, analpore area. 311, 312, T. boxi; 311, dorsal view; 312, anal pore area. AFROTROPICAL JUMPING PLANT LICE 99 315 314 Figs 313-316 Trioza obsoleta-group, 5th instar larvae. 313, 314, T. afrosersalisia; 313, dorsal view; 314,anal pore area. 315, 316, T. mimusops; 315, dorsal view; 316, anal pore area. 100 DAVID HOLLIS 322 323 324 Figs 317-324 Troza species, 5th instar larvae. 317, T. etiennei; 317, dorsal view; 318, anal pore area. 319,320. T. fidcola; 319, dorsal view; 320, anal pore area (from 4th instar larva). 321, 322, T. guiera; 321,dorsal view; 322, anal pore area from dorsal (D) and ventral (V) views. 323, 324, T. tundavalae; 323,dorsal view; 324, anal pore area. Afrosersalisia sp. 5, 54Anthocephalus indicus 28Antiaris toxicaria africana var. ? 7,25 var. africana 7, 23, 24 var. usambarensis 7, 24 var. welwitschii 24Apodytes dimidiata 6, 7, 44, 45, 46 Bartsia longiflora 6, 64Bersama sp. 4, 26Beta 4Brassica4 Calophyllum inophyllum 51Chrysophyllum pruiniforme 7, 59 viridifolium7, 59Cinnamomum spp. 4Cissampelos owariensis 6, 40 sp. 5, 39 torulosaS, 39Citrus spp. 4,5,39Clausena anisata 5, 39 inaequalis 39Cussonia angolensis 5, 41 paniculataS, 41 spicata5, 41, 42 Dacrydium 4Daucus 4DiospyrosSl, 54 AFROTROPICAL JUMPING PLANT LICE Index to host plants Invalid names are in italics. canaliculata5, 54melanoxylon 52mespiliformisS, 52squarrosus 6, 53tomentosa 52xanthochlamys 54 Eugenia malaccensis 4 Fagara capensis 5-6, 39Ficus capensis 30 gnaphalocarpa 5, 29, 30 spp. 4, 5, 6, 28, 29, 30, 33, 34, 35,61 sur5,30 sycomorus 4, 5, 29, 30 thonningii 4, 33 Guiera senegalensis 6, 63 Haplocoelum foliolosum 7, 67Hebe spp. 64 Kickxia sp. 5, 69 Laurus nobilis 4Litsea glutinosa 4, 6, 44 (Tetranthd) laurifolia 44Lycium salinicola 5, 56 tetrandrum 5, 56 Malacantha alnifolia 6, 57sp. 6, 62 101 Mimusops caffra 6, 55 obovata 6, 55 zeyheri 6, 55Myrica conifera 66 Neoboutonia sp. 6, 47 Palaquium gutta 69Persea americana 4Pritchardia spp. 4 Rhamnus 26 Salixsafsaf5,40Seemannaralia gerrardii 5, 42Solanum 4 tuberosum 4 Stephania abyssinica 5, 39Syzygium benguellense 7, 69 Tiliacora sp. 7, 41Toddalia lanceolata 39Triclisia macrophylla 6, 40patens 6, 40 Uapaca nitida 5, 6, 48 Vanilla planifolia 4, 44Vepris undulata 6, 39 Xylopia sp. 7, 43Ziziphus mucronata 5, 27 Index Invalid names are in italics; principal references are in bold. Aacanthocnema 2 acona 2 adaptata 3 afrobsoleta 5, 19,20,51,52 afrosersalisia 5, 11, 18, 51, 54 Afrotrioza 3, 4, 11,18,25 alacris 4 anceps 4 angolensis 4, 21 , 33, 34, 35 anomalicornis 5, 10, 18, 21, 44-45 Anomalocephala 2 Anomalopsyllini 28 Aphalaridae 9, 27, 28 Apsylla 28 armata 2,51 Arytainini 8 asiatica 51 ata 5, 22, 36, 38, 39, 40 baccarum 2Bactericera 2, 9Bactericerinae 9, 23Bactericerini 9bamendae 5, 20, 46,48-49basalis3,26,65beesoni 28 bersama 3, 4, 18, 25-26bessi 2 bicolor (Neotriozella) 2bicolor (Trichochermes) 26biki 29, 30bivittata 2 boxi 5, 19,51,53-54breviantennata 4, 21 , 31-32brevicornis 28bussei 5, 18, 69 Calinda 2 Calophyidae 8 Calophyinae 27 camerounensis 5, 19, 56, 58 capeneri 5, 22, 36,38,39,42 capensis 5, 20, 55-56 Carsidaridae 9, 27 Carsidarini 8 carvalhoi 5, 22, 36, 38, 39, 41-42 casuarina 2 catlingi5,22,36,37-39 ceardi 2 Cecidotrioza 2 Ceropsylla 2 Cerotrioza 2 chiangae 5, 11,22,46,48 cinnamomi 4 Ciriacreminae 27 Ciriacreminidae 9, 23 citri 36 cockerelli4, 55, 56 Colopelma 35 colorata 4 Crawforda 2 dacrydii 4 Dasymastix 2 depressa 28 Diceraeopsylla 28 Diclidophlebia 28 dinaba5,20,46,49 diospyriSl diptera 51 dunaliae 3 eafra5,20,36,38,39,42 eastopi (Pauropsylla) 4, 21, 33 eastopi (Trioza) 43 Egeirotrioza 2 elongagena 3 Engytatoneura 2 Epipsylla 8 Epitrioza 2, 9 Epitriozini 9 Eryngiofaga 2 erythreae 36 erytreae (Aleurodes) 36 erytreae (Spanioza) 36 102 DAVID HOLLIS erytreae (Trioza) 4, 5-6, 18, 22, 36-39, 44, 61 etiennei 6, 10, 1 1 , 19, 56-57, 59, 60eucalypti 3Eutrioza 2, 9Eutriozini 9 fasciatus 3 femoralis 2 ficicola (Pauropsylla) 28 ficicola (Trioza) 6, 21, 60-61 formiciformis 3 formosanus 35 fuscata 3 fuscivena 6, 20, 61-62 galii 35 ghanaensis 6, 19, 62, 67 gigantea 2 glabea 6,19, 59-60 globuli 28 gonjae 6, 19,51,53 gregoryi 6, 22, 36, 38, 39, 40 guiera 6, 11,20,36,62-63 Haplaphalara 27, 28hargreavesi 6, 9, 20, 50-51harteni6,21,46,47-48Hemischizocraniini 9Hemischizocranium 2, 9, 65Hemitrioza 2Heterotrioza 2Hevaheva 2, 65hiurai 3Homotominae 8 immaculata 3insleyi5,ll,18,26-27Izpania 2 johnsonii 3 kakamegae 6, 22, 44, 45karroo 6, 18, 63-64kilimanjarica 6, 22, 36, 38, 39, 40-41Kuwayama 2 laingi 6, 18, 64 lamborni 3, 7, 9, 10, 11, 18, 23-24 lambourni 23 Leptotrioza 2 Leptynoptera 3, 9, 27, 28, 51 Leuronota3, 26 Levidea 3, 9 lindbergi 2 lineata 3 litseae 4, 6, 11,20,43-44 longipes 4, 21, 33-34 longispiculata 28 luvandata 6, 22, 44, 46 lycii 56 machili 3 Macrohomotomini 8maculata (Leuronota) 3maculata (Microceropsylla) 28magnicaudaSlmagnoliae 4medicaginis 2medleri6, 11, 19,65Megatrioza 2, 4, 51menispermicola 6, 22, 36, 37, 38, 39,40 menoni 28merwei 36mesomela 2 messaratina 6, 19, 56, 57Metatrioza 3Microceropsylla 27, 28Microceropsyllini 27mimusops6, 11,20,51,54-55minuta 28 mirificornis 6, 20, 50, 51mistura4, 21, 34, 35mizuhonica 2Moraniella 27Myrmecephala 3 nachingweae 6, 20, 46, 49-50, 61 neoboutonia 6, 11, 22, 46-47, 48, 61 Neolithus 3 Neotrioza 3 Neotriozella 3 nestasimara 7, 20, 56, 58 ngongae 4, 21,31,32 nigra 28 nigricapatus 3 nigricapita 3 nigricornis (Stenopsylla) 3 nigricornis (Trioza) 4, 67 nussex 29, 30 obliqua 2 obscura 64 obsoleta 10, 11,51,52 ocellata 3 opima 2 Optomopsylla 3 Ozotrioza 3 palaquii 69palmicola 4Paracomeca3, 9Paracomecini 9Paratrioza3, 56Pariaconus 3Paurocephala 27, 28Pauropsylla 3, 4, 5, 9, 10, 20-21, 27-29 Pauropsyllinae 9, 27, 28Pauropsyllini 9, 27Paurotriozana 3Pelmatobrachia 27, 28perkinsi 2perrisii 2persea 4 Petalolyma3,26, 65Phacopteron 28Phacopteronini 27, 28phalaki 28Phyllopecta 35Phylloplecta 35pitkini7, 18,56,59Powellia 35prima 3 proxima 4, 21,28,32-33Pseudophacopteron 28Pseudotrioza 3psylloptera 27, 28pubescens 3purpurescens 28 reticulata 28Rhegmoza 3Rhinopsylla 3, 9 russellae 28 Schedoneolithus 3Schedotrioza 3schroederi 7, 19, 65-66, 68schwarzii 3semibrunneipennis 64senegalensis 5, 21, 33, 34-35septimaS, 21,32seranistama 7, 20, 56, 58shiwapuriensis 28sideroxyli 2similis7, 18,69Siphonaleyrodes 8, 35Siphonaleyrodinae 8Smirnovia 2sonchi 2Spanioza 35spondiasae 28Stenopsylla 3stevensi 28sulfurea3, 51Swezeyana 3swezeyi51Sympauropsylla 27Synoza 8 tangae7,21,66-67 tatrichea5,20, 29,30 tavandula 7, 22, 44, 46 tenuis 7, 19, 67 theroni 7, 9, 19, 66, 67-68 thibae 7, 22, 44, 45-46 thomasii 35 tiliacora 7, 22, 36, 38, 39, 41 tinctoria 3 trichaeta5,21,29, 30 Trichochermes 3, 5, 9, 18, 25, 26 Trichochermini 9 Trichopsylla 26 trigemma 5, 21, 31, 32 trimaculata 28 triopsyllina 2 Trioza 3, 5-7, 9, 18, 28, 35-36 Triozamia 3, 7, 9, 11, 18, 22-23, 25 Triozamini 9 Triozaria 8 Triozidae2, 8-17 Triozina 8 Triozinae 8, 9 Triozini 9 Triozoida 3 triozoptera 27 tripunctata 35 tuberculata 28 tundavalae 7, 11, 19, 68-69 udei 3, 27, 28 unica 2 urticae (Chermes) 35 urticae (Trioza) 3 usambarensis 7, 18, 24 usambarica 7, 19, 59, 60 verrucosa 28vitiensis 4vitreoradiata 35vondraceki 7, 18, 24-25 walked 3, 26 willcocksi 4, 5, 10, 21, 29-30, 34 xylopia 7, 20, 43 British Museum (Natural History) Milkweed butterflies: their cladistics and biology P. R. Ackery & R. I. Vane- Wright The Danainae, a subfamily of the Nymphalidae, contains only some 150 species, yet aspects oftheir biology have stimulated far more attention than can be justified by species numbersalone. In recent years, an expansive literature has grown, considering aspects of theircourtship and pre-courtship behaviour, migration, larval hostplant associations, mimicry andgenetics. The popularity of danaines among biologists can certainly be attributed to thiscombination, within one small group, of so many of the factors that make butterflies such aninteresting group to study. The obvious need to place this wealth of biological data within anacceptable systematic framework provided the impetus for this volume. Started eight years ago within the conventions of evolution by natural selection andHennig's phylogenetic systematics, the book is now largely about natural history (what theanimals have and do, where they live and how they develop) and natural groups - as revealedby a form of analysis approaching that practised by the new school of 'transformed cladistics'.The authors have prepared a handbook that will appeal to a wide range of biologists, frommuseum taxonomists to field ecologists. 424 pp (approx.), 12 pp colour, 73 h/vv plates, line and graphic illustrations, maps, extensive bibliography.ISBN 565 00893 5. Publication September 1984. Price 50, prepublication price 45. Titles to be published in Volume 49 Afrotropical jumping plant lice of the family Triozidae (Homoptera: Psylloidea). By David Hollis. The taxonomy of the western European grasshoppers of the genus Euchorthippus, with specialreference to their songs (Orthoptera: Acrididae). By D. R. Ragge & W. J. Reynolds An historical review of the higher classification of the Noctuidae (Lepidoptera). By Ian J. Kitching The Pimplinae, Xoridinae, Acaenitinae and Lycorininae (Hymenoptera: Ichneumonidae) ofAustralia. By I. D.Gauld The western Palaearctic species of Ascogaster (Hymenoptera: Braconidae) By T. Huddleston Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, SuffolkPrinted in Great Britain by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natura BRITISH MUSEUM(NATURAL HISTORY) 27JUL1984 History) The taxonomy of the westernEuropean grasshoppers of the genusEuchorthippus, with special referenceto their songs (Orthoptera: Acrididae) D. R. Ragge & W. J. Reynolds Entomology series Vol 49 No 2 26 July 1984 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique andever-growing collections of the Museum, both by the scientific staff of the Museum and byspecialists from elsewhere who make use of the Museum's resources. Many of the papers areworks of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself,available separately, and individually priced. Volumes contain about 300 pages and severalvolumes may appear within a calendar year. Subscriptions may be placed for one or more ofthe series on either an Annual or Per Volume basis. Prices vary according to the contents ofthe individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History),Cromwell Road, London SW75BD,England. - I World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.) Trustees of the British Museum (Natural History), 1984 The Entomology series is produced under the general editorship of the Keeper of Entomology: Laurence A. Mound Assistant Editor: W. Gerald Tremewan ISBN 565 06004 X ISSN 0524-6431 Entomology series Vol 49 No 2 pp 103-151British Museum (Natural History)Cromwell RoadLondon SW7 5BD Issued 26 July 1984 The taxonomy of the western European grasshoppersof the genus Euchorthippus, with special reference totheir songs (Orthoptera: Acrididae) D. R. Ragge & W. J. Reynolds Department of Entomology, British Museum (Natural History), Cromwell Road, LondonSW75BD Contents Synopsis 103 Introduction 103 Acknowledgements 104 Material 105 Methods 105 Morphology 105 Song 106 Presentation and evaluation of the data 110 Characters studied Ill Morphology Ill Colour pattern 117 Cytology 117 Song 117 Euchorthippus Tarbinskii 121 Key to the western European and North African species and subspecies of Euchorthippus 122 Descriptions of the species 125 General discussion 139 References 141 Index 151 Synopsis The taxonomy of the western European species and subspecies of the Gomphocerine genus Euchorthippusis re-examined on the basis of both morphological and song characters , making full use of such visual aids asbar-charts and scatter diagrams. It is shown that the past emphasis on the male subgenital plate as adiagnostic character has resulted in the frequent misidentification of the three main species, declivus,pulvinatus and chopardi. This has been particularly true in Spain, where pulvinatus and chopardi haveoften been misidentified as declivus and pulvinatus, respectively. It is also shown that the Jersey formelegantulus, previously synonymized with declivus, is actually a form of pulvinatus, and that a furtherpopulation of it occurs in southern Brittany. Analyses of the songs of these species, previously thought tobe of little taxonomic value, have provided strong support for these conclusions. The study embraces fourfurther taxa known only from the western Mediterranean islands or North Africa, and the song of theBalearic species angustulus is fully described and analysed for the first time. Keys are given for theidentification of all these species and subspecies, and their evolution and rather anomalous presentdistribution are briefly discussed. Introduction Euchorthippus belongs to the Gomphocerinae, a group that includes most of the commonEuropean grasshoppers of open grassland. They are well known for their characteristic - andtaxonomically useful - songs, and are frequently used in ecological and cytological studies. Thesubfamily includes about 150 European species, of which over 20 are pests in parts of their range,seven of them being regarded as major ones (Tsyplenkov, 1970). Bull. Br. Mus. nat. Hist. (Ent.) 49 (2): 103-151 Issued26 July 1984 104 D. R. RAGGE & W. J. REYNOLDS Euchorthippus is a widespread Palaearctic genus, including 14 species living in habitatsranging from moist grassland to steppe and barren hillside. Five species occur in western Europeand one of these, E. pulvinatus, becomes a pest when it occurs in sufficiently dense populations,causing damage to pastureland, hayfields, cereal grasses, alfalfa and other cultivated plants inwestern and, especially, eastern Europe (Bei-Bienko & Mishchenko, 1951; Tsyplenkov, 1970).This species sometimes reaches densities of over 10/m 2 in the Stipa-steppe areas of the Ukraineand northern Kazakhstan, where it is by far the most numerous grasshopper (Chetyrkina, 1954;Bei-Bienko, 1961). Euchorthippus is quite easy to recognize as a genus, but its species are much more difficult toidentify and have frequently been confused with one another. E. dedivus, E. pulvinatus and E.chopardi form a group of closely similar species that are very difficult to separate using the keysat present available. Among more recent studies of this group, those of Mafan (1957),Descamps (1968), Litvinova (1972) and Defaut (1982) have much improved our knowledge oftheir taxonomy, ecology and geographical variation. The songs of these species have beendescribed briefly by Descamps (1968), Luquet (1978) and Schmidt & Schach (1978), but theseauthors attached little or no taxonomic importance to the differences between them. The mainaim of our study has been to refine further the best methods for distinguishing morphologicallybetween these species and to demonstrate the taxonomic importance of their songs. Although we have concentrated our study on this group of three species, for each of which wehave been able to study the song, we have also included morphological information on all theother species occurring in western Europe and North Africa and, in the case of E. angustulus,have been able to give full information on the song for the first time. The geographical areacovered by our study consists of southern Europe and the Mediterranean Region from Italywestwards, including North Africa. Morphological recognition of the species of Euchorthippus depends mainly on differences insize and proportion, and we have chosen to present this information in the form of visual aidssuch as bar-charts and scatter diagrams so that the most important differences can be seen at aglance and a specimen can be identified quickly by entering its key measurements on theappropriate diagram. Similar visual aids have been used to illustrate the significance of thedifferences in song. E. pulvinatus pulvinatus, which does not occur in western Europe, is excluded from ourstudy, and where the specific name 'pulvinatus' is used in our text it refers only to the west-ern subspecies E. p. gallicus and E. p. elegantulus unless otherwise indicated. Acknowledgements We are much indebted to the following, who have kindly lent us type-specimens or othermaterial from their respective institutions: Drs A. Cejchan, G. Demoulin, M. Donskoff, M. J. Gonzalez, K. K. Giinther, L. Herrera, G.Kruseman, V. Llorente and E. Tremblay. We are also grateful to Dr A. Nadig for lending us the holotype of E. sardous from his privatecollection in Chur, Switzerland; to Dr B. Defaut, who lent us specimens of Euchorthippus fromBrittany and Teruel province in Spain; and to Mr D. W. Baldock, who gave us, among otherspecimens, a female Euchorthippus from Guerande in southern Brittany and thus alerted us tothe occurrence there of the interesting small form of E. pulvinatus (see p. 131). We thankMr J. F. Burton for very kindly making available to us a tape recording of the song of amale Euchorthippus from Jersey. Our thanks are due to Messrs W. G. Tremewan and B. C. Townsend, who collected live malesfor us from Spain and Jersey, respectively, and thus enabled us to make studio song recordingsfrom them; to Mr E. C. M. Haes, who collected specimens for us from France; to Mr W. B. Lee,who helped us to obtain data from specimens and song oscillograms; to Dr M. Hills of the BritishMuseum (Natural History) Biometrics and Computing Section, who advised us on statisticalmethods; and to Dr F. Willemse for information on the distribution of Euchorthippus in Greece. THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 105 Finally we should like to thank Mr P. H. Ward for his skilful preparation of the drawingsshown in Figs 25-30. Material Our study is based almost entirely on material in the British Museum (Natural History),amounting to about 1500 western European specimens of Euchorthippus . Many of thesespecimens were collected by us during recent visits to France and Spain. The song analyses were based on tape recordings made by us in the field in France and Spain,and in the BMNH Acoustic Laboratory from males brought back from Jersey, France, Spainand Majorca; we were also able to make use of a tape recording kindly made available to us byMr J. F. Burton. We have used the following abbreviations for depositories of type-specimens. BMNH British Museum (Natural History), London IE A Istituto di Entomologia Agraria, Portici IRSNB Institut Royal des Sciences Naturelles de Belgique, Brussels MNHN Museum National d'Histoire Naturelle, Paris MNHU Museum fur Naturkunde der Humboldt-Universitat, Berlin NM NarodniMuzeum, Prague Methods Morphology The morphological measurements were made using a Wild M5 microscope with a graduatedeyepiece. The lengths of the head and pronotum were measured along the mid-line. The lengthof the fore wing was measured from the tip to the point where the fused subcostal, radial andmedial veins of the flexed wing pass under the edge of the pronotum (Fig. 1 , f) . The length of thehind wing was measured (also when flexed) from the tip of the wing to the median posterior pointof the pronotum rather than to the base of the hind wing, since this is hidden under the fore wingwhen the wings are flexed (Fig. 1 , h) . The distance from the stigma to the tip of the fore wing wasmeasured from the centre of the stigma, which is more easily determined than its edge (Fig. 1, s).The distance by which the fore wing projected beyond the hind wing was determined byilluminating the flexed wings from below and measuring the distance between the tips of the foreand hind wings (Fig. 1, g). When the hind wing projected beyond the fore wing, themeasurement was regarded as negative. The length of the hind femur was measured from themost proximal to the most distal points. 1 2 3 Figs 1-3 Diagrams of (1) the flexed wings and (2, 3) male terminal abdominal segments in species ofEuchorthippus, showing the method of measuring the length of the fore wing (f), length of the hind wing(h), distance from the stigma to the tip of the fore wing (s), projection of the fore wing beyond the hindwing (g), length of the male subgenital plate (sp) and length of male abdominal tergites 9+10 (at). R =radius. (See also the explanation in the text.) 106 D. R. RAGGE & W. J. REYNOLDS When counting the stridulatory pegs it was sometimes necessary to relax the leg and to move itaway from the body so that the stridulatory file was visible. The pegs were viewed from above, inprofile, and were counted as they moved past a line on the microscope eyepiece scale. FollowingPerdeck (1957), bristles were included in the row and, in determining the ends of the file, a pegor bristle was excluded if its distance from the preceding peg was twice, or more than twice, thedistance between the two preceding ones. The length of the stridulatory file was measured fromone terminal peg or bristle (determined as above) to the other. The lengths of the male subgenital plate and abdominal tergites 9+10 were measuredfollowing the method used by Descamps (1968) (Fig. 2) except that, when the proximal dorsaledge of the subgenital plate was hidden, the length of only the visible part of the plate wasmeasured (Fig. 3). As we had at our disposal well over 500 specimens of each of the two common species dedivusand pulvinatus , we measured only a sample of them, following the principle that all the localitiesfrom which the specimens came should be represented by at least one measured specimen. Werestricted our measurements of albolineatus to 10 specimens of each sex, since this NorthAfrican species was outside the main scope of our study. Song All the field recordings of the songs used in this study were made while the insects were in fullsunshine using a Uher 4000, 4200 or 4200IC tape recorder and an AKG D202 microphone. Mostof the studio recordings were made in the BMNH Acoustic Laboratory using a Kudelski NagraIV tape recorder and Sennheiser MKH405 microphone. Recording 163 was made from a captiveinsect by Mr J. F. Burton using a Nagra III tape recorder. The only other recordings not made inthe field (179/6, 179/9, 180/1, 180/2) were recorded at the Station Biologique d'Arcachon,Gironde, France, using a Uher 4000 tape recorder and AKG D202 microphone. For all thestudio recordings a bench lamp was used to provide light and radiant heat. Further data are givenin Tables 1 and 2.The acoustic terms we have used are defined as follows. Calling song. The song produced by an isolated male. Courtship song. The special song produced by a male when close to a female. Syllable. The sound produced by one complete up and down movement of the hind legs (Fig. 4).Echeme. A first-order assemblage of syllables (Fig. 4). (Term introduced by Broughton, 1976.) All the song measurements were taken from oscillograms made with a Mingograf 34T ink-jetrecorder. Echeme duration was measured from the end of the first clearly visible syllable to theend of the last (Fig. 6). This measurement was divided by the number of syllables between thesetwo points to give the mean syllable duration for the echeme. 'Gaps' (momentary breaks in thesound - see Fig. 4) were counted if they occurred during the louder second part of the syllableand were at least 1-25 ms in duration (Fig. 6). Echeme repetition rate was determined only from periods of regular echeme repetition. Toput this on an objective basis we used only sequences of four or more echemes during which notwo successive intervals between echemes differed from one another by more than 10 per cent ofthe smaller of the two . The duration of each regular sequence of echemes was measured from theend of the first echeme to the end of the last; the number of echemes included (one fewer thanthe total number of echemes in the sequence) was then divided by this measurement to give theecheme repetition rate (Fig. 5). In the field singing usually stops when the sun is hidden by a cloud, and in the studio the lack ofradiant heat from a nearby bench lamp usually has a similar effect. If singing does continue insuch conditions, the echeme repetition rate is normally much lower and the duration of both theechemes and syllables much greater; the number of gaps per echeme does not changenoticeably. Because of this effect we have based our song data only on recordings made when the THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 107 Table 1 Data for the field recordings of male calling songs of Euchorthippus used in this study. All theserecordings were made from different males. Recordings 236/3, 236/4 and 236/6 were made at thetype-locality of E. chopardi. insect was in full sunshine or receiving radiant heat from a bench lamp, and we suggest that anyrecordings compared with our data should be made under similar conditions. Ambient airtemperatures are also important, but less so than radiant heat as the insects can regulate theirbody temperature to some extent by adjusting their orientation to the incident radiation. When combining the data from the various recordings of each species in order to find therange, overall mean and standard deviation, we treated the mean obtained from each male as anindividual observation, thus minimizing the effect of variation in the songs. When comparing asong with our data, it is thus desirable to use mean values taken from as much of the song aspossible. We excluded from our analysis any echemes that were obviously atypical, such as thoseproduced when the insect was just beginning to sing. We based our figures for echeme repetitionrate on all the regular sequences in each recording, but the number of echemes used for 108 D. R. RAGGE & W. J. REYNOLDS Table 2 Data for the studio recordings of male calling songs of Euchorthippus used in this study. All theserecordings were made from different males except for the following four pairs: (313/4, 316/2), (314/2,316/1), (179/6, 180/2), (179/9, 180/1). measuring the number of gaps per echeme, echeme duration and syllable duration was generallylimited to 50 per recording. When we had two recordings taken from the same male, the mean ofeach character was calculated separately for each recording and these means were thencombined to give a single grand mean for that male. THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUSECHEMES 109 / \ SYLLABLE GAPS Fig. 4 Oscillograms of the male calling song of Euchorthippus declivus, showing the terminology used inthis paper. DURATION OF SEQUENCE n e ECHEME REPETITION RATE = ECHEMES NUMBER OF ECHEMES (n e )DURATION OF SEQUENCE Fig. 5 Oscillogram of a sequence of echemes from the male calling song of Euchorthippus declivus,showing the method of determining the duration of the sequence and the echeme repetition rate. GAPS/ECHEME MEAN DURATION OF SYLLABLE = DURATION OF ECHEMENUMBER OF SYLLABLES (n s Fig. 6 Oscillogram of a typical echeme from the male calling song of Euchorthippus declivus, showing themethod of determining the duration of the echeme, the number and duration of the syllables and thenumber of gaps per echeme. 110 D. R. RAGGE & W. J. REYNOLDS Presentation and evaluation of the data We have chosen to present the morphological and acoustic data in the form of bar-charts (Figs20, 82-88) because they make it possible to compare at a glance our samples of the taxa and theirconstituent regional populations. The components of each bar are explained in Fig. 7 (but notethat in some bars the range does not extend beyond the standard deviation on one side of the MINIMUM MEAN STANDARD DEVIATION MAXIMUM I STANDARD ERROR OF MEAN Fig. 7 Diagram explaining the components of the bars used in the bar-charts (Figs 20, 82-88) . Note that insome of these bars the range does not extend beyond the standard deviation on one side of the mean andthat, when the number of observations is fewer than five, the bar is given as a simple line with nostatistical information beyond the range and mean. mean). The scale used for each bar-chart was chosen to make the fullest use of the spaceavailable, and the resulting differences in scale should be taken into account when comparingthe variability of different characters. All morphological measurements are given in millimetresand all temporal ones in milliseconds. For each sample, n represents the number of specimensmeasured or whose songs have been analysed, and on the bar-charts of acoustic data (Fig. 20) n erepresents the total number of echemes analysed. Where n<5 the bar is given as a simple linewith no statistical information beyond the range and mean. We found the bar-charts to demonstrate so clearly and quickly the differences between thetaxa, and which characters are most useful in separating them, that we decided it would bepointless to subject the same data to multivariate analysis. We have occasionally applied 't' teststo determine more objectively which of two characters (or combinations of characters) gave thebest separation between two taxa, but in every case these did no more than confirm what wasobvious from the bar-charts. The bar-charts also suggested combinations of characters that would give improved separa-tions and we have plotted a selection of these in the form of scatter diagrams (Figs 8-13, 21-24).Fig. 8 shows particularly clearly the contrast in the males between declivus and the remainingspecies in the projection of the fore wing beyond the hind wing. Fig. 9 gives a completeseparation of the males of declivus, pulvinatus and chopardi, especially if the French samples areconsidered alone. Figs 10 and 11 achieve a similar result for both sexes, but the French andSpanish samples have to be considered separately in order to obtain a complete separation in themales. Figs 21-24 show rather more clearly than the acoustic bar-charts (Fig. 20) the differencebetween the songs of declivus, pulvinatus and chopardi. The best morphological separation between males of declivus, pulvinatus and chopardi wasachieved by plotting three characters as percentages on a triangular graph (Fig. 12). Theprocedure for this is described by Mayr (1969) and is best explained by an example. For onespecimen the following figures were obtained: number of stridulatory pegs 148, length ofpronotum/length of head 1-09 and distance of stigma from tip of fore wing 2-68 mm. Since thefigures for the last two characters are small compared with the first, they were multiplied by 100and 25, respectively. The three figures used were thus 148, 109 and 67, giving the sum 324. Thethree characters represent 45-7 per cent, 33-6 per cent and 20-7 per cent of their sum and thesepercentages were used to enter a single point on the graph, the axes being arranged so that thesum of the three coordinates of each point was always 100. The graph thus shows proportionsrather than absolute sizes. Figure 12 gives perhaps the most convincing demonstration that declivus, pulvinatus andchopardi do not constitute a single, continuously variable species, though it does not of course THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 111 170160150 140 s g 130 _i 1 120 fc UL no OfLLJCO 3 100 9080 France ... f A rancpulvmatus 9 a " lcus \ T |b e ria France f rande f vus io Italy -0-5 0-5 1-0 1-5 PROJECTION OF FORE WING BEYOND HIND WING (mm) 2-0 Fig. 8 Scatter diagram showing a plot of the number of stridulatory pegs against the projection of the forewing beyond the hind wing in males of Euchorthippus . The negative values on the horizontal axis are forspecimens in which the hind wing projects beyond the fore wing. rule out polymorphism. Figure 13 shows that, using the same combination of characters,albolineatus and angustulus resemble chopardi, while sardous is nearer to pulvinatus . It should be noted that in some scatter diagrams the number of points entered is smaller thanthe number of specimens measured (as shown in the bar-charts), because in a few specimensboth (or, in the case of the triangular graphs, all four) of the characters were not available.Where two different population symbols would have coincided exactly, only one symbol isshown in the morphological scatter diagrams; the symbol chosen is the one considered to bemore useful or informative (e.g. near the edge of the cluster concerned). In the song scatterdiagrams (Figs 21-24), for which fewer points were available, coincident symbols have beenmoved apart slightly so that both can be seen. Characters studiedMorphology We have concentrated on morphological characters that are easy to use with pinned specimensand have therefore excluded those characters of the fore wing that are often hidden behind thehind legs (e.g. the width of the medial and cubital areas) or are difficult to measure withoutopening and flattening the wings (e.g. the width of the fore wing). We have also tried to avoidcharacters that tend to change as the body shrinks after death, and have therefore used thelength of the hind femur as a measure of size in preference to body length, which is often greatlyaltered by post mortem shrinkage (and in the females is also much influenced by gravidity). Wehave, however, included the lengths of the male subgenital plate and abdominal tergites 9+10because of the emphasis placed on these measurements by Descamps (1968), although we have 112 170160150 /> 140O 130 120 100 90 80 70 (A rancpu/v/nafus ga//icus| ,, . D. R. RAGGE & W. J. REYNOLDSFrance .. France - pu/vinafus e/egan; angustulus (o) a/bo/ineafus a/bo/ineafus () sardous (*) a/bo/ineatus sicu/us (o) ^. 123456 O^ DISTANCE FROM STIGMA TO TIP OF FORE WING (mm) Fig. 9 Scatter diagram showing a plot of the number of stridulatory pegs against the distance from thestigma to the tip of the fore wing in males of Euchorthippus . To facilitate comparison broken lines areused for the two Iberian samples. found that their usefulness is much reduced by shrinkage during drying. We suspect that thedistance by which the fore wing projects beyond the hind wing is also affected by post mortemshrinkage, but this does not prevent it from being a useful character. Head and pronotum The lengths of the head and pronotum are not very useful taxonomic characters other than asmeasures of size or when taken as a ratio of one to the other. This ratio is frequently useful forseparating chopardi from the other mainland species and was used by Descamps (1968) todistinguish chopardi from angustulus, albolineatus albolineatus and a. siculus. However, thepronotum/head ratios of Iberian chopardi show a large overlap with those of the other taxa (Figs82, 86). The prominence of the median and, especially, lateral carinae of the pronotum is often usefulin distinguishing between species, but this character is rather difficult to use until one is familiarwith the differences shown. We have attempted to show in diagrammatic cross-section (Figs31-34) the degree of prominence of these carinae in declivus, pulvinatus gallicus, chopardi andalbolineatus albolineatus. Figures 31-39 also show dorsal views of the pronotum of males of allthe taxa dealt with in this paper. In most species the lateral carinae are almost straight but insardous they are distinctly incurved in the prozona (Fig. 38). They are also somewhat incurved inangustulus, but less so in specimens from Majorca than in those from Formentera. 12 T10 THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS albolineatus albolineatus () 113 Francea ... ... I A Fran pulvinatus gamcus 4 . . Franio Italy ... ( Francedeclivus j angusfu/us (o) pulvinatus elegantulusi . I ^ Brittan sardous (*) 234DISTANCE FROM STIGMA TO TIP OF FORE WING (mm) Fig. 10 Scatter diagram showing a plot of the length of the hind femur against the distance from the stigmato the tip of the fore wing in males of Euchorthippus. To facilitate comparison broken lines are used forthe two Iberian samples. Fore wing The absolute and relative lengths of the fore wing are useful characters and have been much usedby previous authors (e.g. Defaut, 1982). The position of the tip of the fore wing in relation to thehind knee or, less reliably, the apex of the abdomen provides one of the easiest methods ofdistinguishing between typical declivus, pulvmatus gallicus and chopardi in the field (Figs25-30) . The length of the fore wing is more useful when taken as a ratio with the length of thehind femur; this is particularly true in the female, in which the range of our sample of p. gallicusdoes not overlap with those of chopardi or a. albolineatus (Fig. 87). The fore wings are short in declivus and become progressively longer in both sexes in thesequence declivus, p. gallicus, chopardi and a. albolineatus (Figs 83, 87). The distance from the stigma to the tip of the fore wing is, to some extent, proportional to thelength of the fore wing but provides an even better character, especially when used as a ratio withthe length of the hind femur (Figs 84, 85, 88). When these two characters are plotted against eachother in a scatter diagram, there is no overlap in our male samples of the populations of the threespecies occurring in France (declivus, pulvinatus and chopardi) or between the two occurringcommonly in Spain (pulvinatus and chopardi) (Fig. 10). In females there is no overlap at allbetween our samples of declivus, pulvinatus and chopardi (Fig. 11), though a. albolineatus isvery closely associated with chopardi in both sexes. E. p. elegantulus agrees with p. gallicusrather than with declivus when the measurements are plotted against the length of the hindfemur (Figs 10, 11) or, in males, against the number of stridulatory pegs (Fig. 9). 114 16 15 14 13 12 10 D. R. RAGGE & W. J. REYNOLDS i France i . . ... r A France ,.(m France pulvinafus ga/licus| . . x chopardcj ., . \ a/bolmeofus a/bo/mearus (*) . .. r Francedeclivus < .. , pulvinafus elegan a/bo/ineafus s/cu/us (o) angusfu/us sardous (*) 3456DISTANCE FROM STIGMA TO TIP OF FORE WING (mm) Fig. 1 1 Scatter diagram showing a plot of the length of the hind femur against the distance from the stigmato the tip of the fore wing in females of Euchorthippus . To facilitate comparison broken lines are used forthe two Iberian samples. Hind wing The length of the hind wing is often a useful character (Figs 83, 87), especially when taken as aratio with the length of the hind femur (Figs 84, 87) . In this ratio, for both sexes, p. elegantulus isagain closer top. gallicus than to declivus, which is distinct from all the other taxa. In the femalethis character separates chopardi from/7, gallicus. Projection of the fore wing beyond the hind wing The distinct gap between the tips of the flexed fore and hind wings provides the best character fordistinguishing declivus from the remaining taxa (Figs 8, 84, 88). This character has beenmentioned by several authors (Chopard, 1952; Descamps, 1968; Harz, 1975; Luquet, 1978;Defaut, 1982) but its significance in showing that elegantulus is distinct from declivus has notbeen previously realized (see p. 132). Hind femur The length of the hind femur (Figs 82, 86) is useful only as a measure of size, for which purposewe have used it in ratios with the length of the fore wing, the length of the hind wing and thedistance from the stigma to the tip of the fore wing (Figs 10, 11, 84, 85, 87, 88). Stridulatory file Counting the Stridulatory pegs takes at least a minute per specimen and ideally requires amicroscope with a movable stage. We have found the effort worthwhile, however, as thischaracter, in combination with stigma distance and pronotum/head ratio on a triangular graph, THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 115 France > > declivuso Italy J A France i >pu/v/nafus gallicusT Iberia J A Jersey 1 ^pu/vinafuj e/eganfu/usv Brittany J r France , > chopardiD Iberia J 65, 30 y LENGTH OF PRONOTUM/LENGTH OF HEAD (xlOO) Fig. 12 Triangular graph showing the relationship between males of Euchorthippus declivus, E. pulvina-tus and E. chopardi as determined by three diagnostic characters. (See the text for full explanation.) has provided one of the best means of separating males of declivus, pulvinatus and chopardi (Fig.12). The first two of these characters give almost as good a separation in a two-axis scatterdiagram (Fig. 9). The stridulatory pegs provide another character showing the close relationshipbetween p. elegantulus and/?, gallicus (Fig. 83). E. p. elegantulus has, on average, slightly fewerpegs than/?, gallicus, as would be expected from the difference in size, but in both of the two-axisscatter diagrams and in the triangular graph/?, elegantulus is clearly more closely associated with/?. gallicus than with declivus. We have found it necessary to examine the female stridulatory pegs only in order to find thebest means of separating a. siculus and pulvinatus . The pegs are smaller in females and are morelikely to be replaced by hairs, but it is still usually possible to count them by viewing the file inprofile. In cases where this is difficult a replica of the file can be made (see Pitkin, 1976). We did not find the length of the stridulatory file to be a useful character, the very short file ofsardous simply reflecting the generally small size of this species and in particular its short hindfemora (cf. Figs 82 and 83). 116 D. R. RAGGE & W. J. REYNOLDS a/bo/ineafus a/bo/ineafusO a/bo/ineafus sicu/us * sardous $ angusfulus 65 35 30. declivus a/bo/ineafus sicu/us angusfu/us a/bolineafus a/bo/ineafus . v t . . > \ . > LENGTH OF PRONOTUM/LENGTH OF HEAD (xlOO) Fig. 13 Triangular graph showing the relationship between males of Euchorthippus as determined bythree diagnostic characters. The detailed plots for E. declivus, E. pulvinatus and E. chopardi are shownin Fig. 12. (See the text for full explanation.) Male subgenital plate and abdominal tergites 9+10 The shape of the male subgenital plate has been used as a key character by all previous authors.Descamps (1968) tried to make this character more objective by comparing the length of thesubgenital plate with that of abdominal tergites 9+10, and concluded that in declivus andpulvinatus the dorsal part of the subgenital plate was longer than abdominal tergites 9+10 inprofile, while in chopardi, a. albolineatus, a. siculus, angustulus and sardous these tergites werelonger than, or subequal to, the subgenital plate. We have found that measuring the subgenital plate accurately in dried specimens is difficult,since post mortem shrinkage may cause the proximal end of the subgenital plate to be covered bythe supra-anal plate. Another problem is caused by the way in which variation in the degree ofcurvature of the abdomen affects the measurement of the length of tergites 9+ 10. Perhaps partlybecause of these difficulties, we have not found the ratio of the length of the subgenital plate tothat of abdominal tergites 9+10 to be particularly useful; for declivus and/?, gallicus the ratio isin fact less useful than the length of the subgenital plate alone (Fig. 85). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 117 We have found the shape of the male subgenital plate to be more useful than its dimensions. Itis always distinctly pointed in declivus, bluntly rounded in chopardi and intermediate inpulvinatus (Figs 25, 27, 29), but without some experience of the group it is easy to be misled bythis character. The frequent misidentification of Spanish p. gallicus as declivus has doubtlessbeen the result of using the male subgenital plate as the principal distinguishing character. Internal male genitalia We have examined the internal male genitalia of all the species included in this study and havefound no taxonomically useful character. Colour pattern Most specimens show some degree of dark longitudinal striping, especially on the head, and inchopardi these stripes are particularly conspicuous (Fig. 40). We have seen a few conspicuouslystriped specimens that seem to be pulvinatus on other characters, but well-developed stripes ofthe kind shown in Fig. 40 are characteristic of chopardi. Cytology In common with other Gomphocerinae of the Chorthippus-group, the species of Euchorthippusin which the chromosomes have been studied have complements of 17 in the male (XO) and 18 inthe female (XX) (Hewitt, 1979, Santos etal. , 1983). The chromosomes of pulvinatus have beenexamined by McClung (1932), Arana etal. (1980), Ferrer etal. (1981), Santos & Giraldez (1982)and Santos et al. (1983), while those of albolineatus have been described by Carlson (1936)(Carlson's specimens were identified by Hebard as pulvinatus, but as they came from Algeriathey were almost certainly albolineatus, the only species of Euchorthippus known to occur inNorth Africa). Santos et al. (1983) have also examined the chromosomes of Spanish chopardiand have shown that the C-banding patterns differ quite markedly from those of Spanishpulvinatus (these authors refer to chopardi as 'albolineatus' , but Dr Santos has kindly shown usthe specimens and we consider them to be chopardi). Although the number and grossmorphology of the chromosomes are likely to be uniform throughout the genus, it would clearlybe interesting to study the C-banding patterns in other species; careful comparison may alsoreveal differences in chiasma frequency, relative lengths of chromosomes or DNA content ofequivalent nuclei (John & Hewitt, 1966; White, 1972). Santos et al. (1983: 73) have succeeded in rearing artificial hybrids between pulvinatus andchopardi. This raises the possibility of natural hybrids occurring between these species, whichquite often occur together. We have seen a few specimens that are intermediate between the twoin some morphological characters, but in order to recognize natural hybrids with any confidenceit would be necessary to investigate them genetically and, in the case of males, to analyse thecalling song. Song There has been no comprehensive study of the acoustic behaviour of Euchorthippus, but severalauthors have published comments on the songs of one or more species. Chopard (1952) brieflyoutlined the calling song of pulvinatus and Faber (1953) described its complete acousticrepertoire, including calling, courtship and rivalry songs. Descamps (1968) published oscillo-grams of one echeme each of declivus, pulvinatus and chopardi, but found no noticeabledifference between them and gave a single description of the calling and courtship songs for allthree species. Luquet (1978) gave a diagram of one echeme for each of these species andsuggested that the echeme was longest in declivus, almost as long in chopardi and shortest inpulvinatus (our own analyses, however, show the echemes to be longer in chopardi than indeclivus - see Fig. 20). Luquet also found that pulvinatus had a more rapid 'cadence' (?echeme repetition rate) than the other two species, but he considered these differences to be 118 D. R. RAGGE & W. J. REYNOLDS insufficient to permit identification in the field. Schmidt & Schach (1978) gave oscillograms ofthe courtship song of pulvinatus and declivus, and described differences in the duration of thesyllables and echemes, but neither these authors nor any others have seriously suggested usingthe song as a taxonomic character. Fig. 17 14 declivus Fig. 19 16 chopardi 'It Hfr 17 declivus 18 pu/wnafus 19 chopardi 20 Hz 0-1 s Figs 14-19 Oscillograms of typical male calling songs of (14, 17) Euchorthippus declivus, (15, 18) E.pulvinatus and (16, 19) E. chopardi. Figs 17-19 are faster oscillograms of the single echemes indicated inFigs 14-16. The oscillograms were taken from recordings 233/6, 232/4 and 236/8, for which fullinformation is given in Table 1 (p. 107). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 119 Fig. 20 Data for four characters of the male calling songs of Euchorthippus declivus, E. pulvinatus and E.chopardi. For explanation of the bars see Fig. 7; n = number of males, n e = number of echemes fromwhich data were taken. Note that in compiling these charts the mean values obtained from each malewere treated as individual observations. (For further explanation see p. 110.) While our own observations are broadly in agreement with those of these authors, our moredetailed analyses have revealed good, taxonomically useful differences between the songs ofdeclivus, pulvinatus and chopardi. Figures 14-19 show oscillograms of typical sequences ofechemes from the male calling songs of these three species, followed by much faster oscillogramsof single echemes selected from these sequences. The full data from all our recordings are givenin Fig. 20 and scatter diagrams from four selected pairs of song characters are shown in Figs21-24. The most striking difference is between the large number of gaps per echeme in declivusand the much smaller number in pulvinatus. This difference has not previously been observedalthough it is clearly visible in the oscillograms given by Descamps (1968: Fig. 23 - but note thatthe oscillogram for declivus is shown backwards) . The number of gaps per echeme in chopardi isalso lower than in declivus, but the difference is less striking and our recordings of chopardi showa high degree of variability in this character. Both the syllables and echemes are of shorterduration in pulvinatus than in declivus, the syllables showing the clearer contrast (Fig. 20). In allthese characters the bar-charts and scatter diagrams show a close association between p. gallicusand p. elegantulus (Figs 20-24). 120 70 60~sr % 50 O 30 Q 20 10 D. R. RAGGE & W. J. REYNOLDS 1-6 . r France , .. frranc ch P ard( to Iberia pulvina decl/vus () 21 4 6 8 10 12 14 GAPS/ECHEME 1-4 1-2 1-0 0-8 0-6 0-4 chopardi . pu/vinafus ga/licus. pulvinaf us elegant u/us dec/ivus 22 4 6 8 10 12 14 GAPS/ECHEME 400 300 O20 100 23 chopardi -pu/vinafus ga/ficus pulvinatus e/eganfulus 6 8 10 12 GAPS/ECHEME 14 70 60 J.uj 50 CO s fc 40 30 20 10 24 chopardi declivus po/vinafus ga//icus pu/vinafus e/eganfufus 0-4 0-6 0-8 1-0 1-2 1-4ECHEMES/s 1-6 Figs 21-24 Scatter diagrams showing plots of four pairs of characters of the male calling songs ofEuchorthippus declivus, E. pulvinatus and E. chopardi. Note that each point represents the mean valuesof the two characters obtained from one male. The symbols are explained in Fig. 21. None of these differences in song can be measured without the help of recording and analysingequipment - indeed the gaps , which last for little more than a millisecond , are quite undetectableby the unaided human ear (though being easily resolved by the insects themselves - see, forexample, the review of Autrum, 1963). The only song character that can be used as an aid to fieldidentification is the echeme repetition rate, which can be quite easily measured using the secondhand (or digital count) of a wrist- watch. As shown in Fig. 20 this character does not show strikingcontrasts between the species but, when they are singing in similar warm, sunny conditions,chopardi has a slower rate than the other two, and pulvinatus usually has a faster rate thandeclivus. It should be emphasized that, for taxonomic use, measurements of all these characters shouldbe based on as many echemes as possible and should always be taken from songs produced byisolated males singing in warm sunshine (or in a warm laboratory with a source of radiant heat). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 121 The echeme repetition rate should be based only on regular sequences of echemes (see p. 106)and, for each male, it is the mean values of all these song characters that should be comparedwith the bar-charts and scatter diagrams (Figs 20-24). The number of syllables in each echeme (usually 6-7, occasionally 5 or 8) appears to have littleor no taxonomic value. We counted the syllables in the course of calculating the syllableduration, and it soon became clear that intraspecific variation would obscure any trend towardsan interspecific difference. EUCHORTHIPPUS Tarbinskii Euchorthippus Tarbinskii, 1925: 192; Uvarov, 1926: 341 [redescription]. Type-species: Oedipoda pulvi- nata Fischer de Waldheim, by original designation.Sinhippus Ramme, 1939: 132. Type-species: Sinhlppus alinl Ramme [= Euchorthippus unicolor (Ikonni- kov)], by original designation. [Synonymized by Bei-Bienko & Mishchenko, 1951: 543.] DIAGNOSIS, cf $ Fastigium of vertex rather rounded and relatively blunt; foveolae rather short andshallow, each with weak carinula leading inwards and backwards from its posterior end. Antennae notclubbed. Lateral carinae of pronotum straight or slightly incurved. Pleura coarsely rugose and punctate.Inner and outer tarsal claws of different lengths. Brachypterous to macropterous. Precostal area of forewings with small bulge on anterior margin; medial area not conspicuously widened, without intercalaryvein and with irregular cross-veins. Hind wings transparent and colourless. Subgenital plate bluntly toacutely conical. Ovipositor short, without lateral teeth. General coloration brown, straw-coloured or grey(sometimes with greenish tinge but never clearly green or reddish in western European species), withlongitudinal dark and light stripes on head and sometimes pronotum. Female usually with white lineascapularis. Male calling song. In the four species whose stridulation we have studied (declivus, pulvinatus, chopardiand angustulus) the calling songs have the same basic structure though differing in detail (Fig. 4). Theyconsist of a sequence of echemes repeated at a rate of 0-5-1-5 per second for an indefinite period,sometimes a minute or more. Oscillographic analysis shows that each echeme begins quietly, lasts about150-350 ms and is composed of 5-8 syllables, each of which has a quiet first part and a louder second part; inat least the later part of each echeme, the louder second part of each syllable contains momentary breaks inthe sound, referred to in this paper as 'gaps' (Fig. 4). DISCUSSION. Euchorthippus is an easily recognized genus to the experienced acridologist, but itsmorphological diagnostic characters are rather subtle and difficult to describe. In comparisonwith Chorthippus, with which it is most easily confused, Euchorthippus has a more elongatehead when viewed from the side, with the eyes produced anterodorsally into a more pronouncedand acute angle. The foveolae are rather weakly (sometimes very weakly) developed, and fromthe posterior end of each one arises a feeble carinula, usually extending inwards and backwardstowards the mid-line of the head, where there is often another feeble median carinula, especiallyon the fastigium; these carinulae are quite absent in Chorthippus. The difference in size betweenthe inner and outer tarsal claws has been used by some authors as a key character, but there issometimes a tendency towards this size difference in Chorthippus. The western Europeanspecies of Euchorthippus are always brown, straw-coloured or grey and, although sometimesshowing a greenish tinge, never have the clear green or reddish colours common in Chorthippusand other Gomphocerinae. The calling song of the male provides a particularly good character for recognizing Euchor-thippus in the field. The long sequences of rapidly repeated echemes are highly characteristicand quite unlike the songs of any other European Gomphocerinae we have heard. Some otherGomphocerinae (e.g. Chorthippus parallelus (Zetterstedt), C. montanus (Charpentier), C.dorsatus (Zetterstedt), C. dichrous (Eversmann), Chrysochraon dispar (Germar)) have callingsongs composed of long sequences of echemes, but in all these cases the repetition rate is muchlower, never higher than one every two seconds. The echeme-sequences of Euchorthippus are infact more strongly reminiscent of those produced by some Tettigoniidae, especially suchDecticinae as Metrioptera brachyptera (L.), Platycleis sabulosa Azam and P. albopunctata(Goeze), all of which are at least partially diurnal singers; P. sabulosa even has a similar numberof syllables per echeme (usually 6-7), but all these species have a higher echeme repetition rate 122 D. R. RAGGE & W. J. REYNOLDS when singing during the day, usually more than 2 echemes/s in warm sunshine. Althoughisolated males of Euchorthippus can maintain a fairly regular echeme repetition rate for quitelong periods, they never quite achieve the almost mechanical regularity and very long duration(often many minutes) of the echeme-sequences of these Decticine bush-crickets. There is no elaborate courtship song but in the presence of a female the male produces asomewhat modified song in which the echemes are quieter and the sequences longer than usual.Faber (1953) described a second stage in the courtship song of pulvinatus in which there is agradual increase in the intensity, repetition rate and duration of the echemes, and in the numberof syllables per echeme. Just before jumping on to the female, the male produces two or threequiet sounds, each consisting of one or two syllables, and then one loud and more extendedsound of uniform intensity. DISTRIBUTION. Southern Europe, most of the larger Mediterranean islands, Madeira, NorthAfrica and temperate Asia as far as China. INCLUDED SPECIES. E. albolineatus (Lucas), E. angustulus Ramme, E. arabicus Uvarov, E.chenbaensis Tu & Cheng, E. cheui Hsia, E. chopardi Descamps, E. declivus (Brisout), E.madeirae Uvarov, E. pulvinatus (Fischer de Waldheim), E. sardous Nadig, E. transcaucasicusTarbinskii, E. unicolor (Ikonnikov), E. weichowensis Chang, E. yungningensis Cheng. Key to the western European and North African species and subspecies of Euchorthippus The two mainland taxa pulvinatus gallicus and chopardi, both common in southern France and the IberianPeninsula, are particularly difficult to distinguish from each other, and for reliable identification ofspecimens from these regions (even including females thought to be declivus} we recommend taking therelevant measurements and plotting them on the scatter diagrams shown in Figs 10 and 11. For males theuse of the triangular graph shown in Fig. 12, although requiring more effort, will produce an even morereliable result. 1 Fore wings projecting beyond hind wings (when flexed) by at least 0-8 mm in cf , usually at least 0-5 mm in $. Male subgenital plate long and pointed, as in Fig. 25. Distance from stigmato tip of fore wing usually less than 0-21 times length of hind femur in cf , less than 0-25 in $ E. declivus (Brisout) (p. 125) Fore wings not projecting beyond hind wings (when flexed) or projecting by less than 0-7 mm inCf , usually less than 0-5 mm in 9 Male subgenital plate shorter or less pointed , as in Figs 27 or29. Distance from stigma to tip of fore wing usually more than 0-21 times length of hind femurin cf, more than 0-25 in $ 2 2 Pronotal lateral carinae distinctly incurved in prozona (Fig. 38) ; length of pronotum less than 2-1 mm in cf, less than 3 -Omm in $. (Sardinia only) E. sardous Nadig (p. 136) Pronotal lateral carinae straight or almost so in prozona; length of pronotum more than 2-1 mmin cf , more than 3 -0 mm in $ . (Not known from Sardinia) 3 3 Male subgenital plate relatively short and blunt, as in Fig. 29. Fore wings long, reaching at least base of genicular lobes of hind femora in both sexes. Distance from stigma to tip of fore wing usually more than 0-34 times length of hind femur in cf , more than 0-38 in $ 4 Male subgenital plate longer and slightly pointed, as in Fig. 27. Fore wings shorter, in female notreaching base of genicular lobes of hind femora. Distance from stigma to tip of fore wingusually less than 0-34 times length of hind femur in cf , less than 0-38 in $ 6 4 Known only from southern France and the Iberian Peninsula. Pronotal lateral carinae relatively low and broad, as in Fig. 33. Head and pronotum usually with conspicuous dark and light longitudinal stripes (Fig. 40) E. chopardi Descamps (p. 133) Known only from North Africa and the Balearic Islands. Pronotal lateral carinae relatively highand narrow, as in Fig. 34. Coloration more uniform 5 5 Known only from North Africa. Larger: length of hind femur usually more than 9-4 mm in cf , more than 12-6 mm in $ ; length of pronotum usually more than 2-7 mm in cf ; length of forewing usually less than 1 -22 times length of hind femur in $ E. albolineatus albolineatus (Lucas) (p. 135)- Known only from the Balearic Islands. Smaller: length of hind femur usually less than 9.4 mm in THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 123 Rl 10 mm Figs 25-30 Side view of typical specimens of each sex of (25, 26) Euchorthippus declivus, (27, 28) E.pulvinatus gallicus and (29, 30) E. chopardi. 124 D. R. RAGGE & W. J. REYNOLDS 5 mm Figs 31-39 Dorsal view of the pronotum of (31) Euchorthippus declivus, (32) E. pulvinatus gallicus, (33)E. chopardi, (34) E. albolineatus albolineatus , (35) E. pulvinatus elegantulus from Brittany, (36) E. p.elegantulus from Jersey, (37) E. albolineatus siculus, (38) E. sardous, (39) E. angustulus. Diagrammaticprofiles of cross-sections of the pronotal disc are also shown in Figs 31-34. Fig. 40 Dorsolateral view of a typical male of Euchorthippus chopardi, showing the characteristic colourpattern. Cf , less than 12-6 in $ ; length of pronotum usually less than 2-7 mm in cf ; length of fore wingusually more than 1 -22 times length of hind femur in 9 E. angustulus Ramme (p. 137) Known only from Sicily. Stridulatory file with fewer than 110 pegs in cf , fewer than 105 pegs orhairs in $ E. albolineatus siculus Ramme (p. 136) Not known from Sicily. Stridulatory file with more than 115 pegs in cf , more than 110 pegs orhairs in $ Larger: length of head usually more than 2-4 mm in cf , more than 3-2 mm in $ ; length of forewing usually more than 9-2 mm in cf , more than 11-3 mm in $. (Southern half of France,Iberian Peninsula) E. pulvinatus gallicus Mafan (p. 131) Smaller: length of head usually less than 2-4 mm in cf, less than 3-2 mm in 9; length of forewing usually less than 9-2 mm in cf, less than 11-3 mm in $. (Jersey and southern Brittany)E. pulvinatus elegantulus Zeuner (p. 131) THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 125 Descriptions of the species Euchorthippus declivus (Brisout) (Figs 25, 26, 31) Acridium declivum Brisout, [1849]: 420. LECTOTYPE $, FRANCE: near Paris, St-Germain (MNHN, Paris), here designated [examined].Stenobothrus pulvinatus var. gracilis Azam, 1901: 46. LECTOTYPE cf , FRANCE: near Clermont-Ferrand, Puy de Crouelle, viii.1896 (M. H. du Buyssoh) (MNHN, Paris), here designated [examined]. [Synony- mized by Chopard, 1952: 303.]Euchorthippus declivus (Brisout) Uvarov, 1926: 341.Euchorthippus declivus merldlonalis Jannone, 1937: 57. LECTOTYPE cf , ITALY: Puglia, Monte Angeli, 2.ix.l935 (TEA, Portici), here designated [examined]. [Synonymized by La Greca, 1959: 138.]Euchorthippus declivus stichai Mafan, 1954: 137. Holotype cf, CZECHOSLOVAKIA: Kamemin, viii.1952 (/. Mafan) (NM, Prague) [examined]. [Synonymized by Harz, 1975: 924.] DIAGNOSIS, cf $. Pronotal lateral carinae relatively low and broad; length of pronotum usually 0-98-1-41times length of head. Fore wings relatively short, not or hardly reaching apical quarter of hind femora in cf ,middle of hind femora in 9 , but usually projecting beyond hind wings by at least 0-8 mm in cf , at least 0-5mm in $ ; distance from stigma to tip of fore wing usually less than 0-21 times length of hind femur in cf , lessthan 0-25 in $ . Length of hind wing usually less than 0-87 times length of hind femur in cf , less than 0-78 in$. Male stridulatory file usually with 75-120 pegs. Male subgenital plate relatively long and pointed, as inFig. 25. (Full data for measurements, ratios and number of stridulatory pegs are given in the bar-charts onpp. 144-150.) Male calling song (see Figs 14, 17, 41-^45). Each echeme usually with 10-15 gaps. Other song charactersas shown in bar-charts (p. 119). Ml 41 180/3 # 42 232/2 43 233/6 44 241/1 45 484/4 20 Hz 0-1 s Figs 41-45 Oscillograms of single echemes from the calling songs of five French males of Euchorthippusdeclivus. The small numbers refer to the recordings from which the oscillograms were made and can beused to obtain the full data from Tables 1 and 2 (pp. 107, 108). 126 D. R. RAGGE & W. J. REYNOLDS DISCUSSION. This is the most easily recognizable of the mainland species of Euchorthippus inwestern Europe. As can be seen in Figs 8, 84 and 88, the projection of the fore wings beyond thehind wings (when both pairs of wings are flexed) enables the males, and usually the females, tobe identified at once. In all the other western Palaearctic species (except madeirae - see below)the tips of the two pairs of flexed wings are coincident or almost so in live or freshly killedspecimens, though a very small gap often develops in dried specimens. The length of the hindwing and the distance from the stigma to the tip of the fore wing, both taken as a ratio with thelength of the hind femur, also provide good characters in both sexes (see Figs 10, 11, 84, 85, 87,88). The male is further characterized by the relatively long and pointed subgenital plate (Fig.25). The endemic Madeiran species E. madeirae Uvarov resembles dedivus in having rather low,broad pronotal lateral carinae and short wings, with the fore wings extending well beyond thehind wings. However, the males have a much shorter and blunter subgenital plate than dedivus,and the pronotal lateral carinae are more clearly incurved in both sexes. There is a rare macropterous form of dedivus in which the fore wings reach a length of 11-13mm in the male and 13-15 mm in the female. This form has been recorded from thesouth-western Slovakian steppe in Czechoslovakia by Mafan (1957: 189) and from Gran Sasso,Abruzzi, Italy by Baccetti (1958: 426-428); we have seen further specimens from Rome andfrom near Banja Luka in Yugoslavia. In the specimens we have examined the fore wings projectbeyond the hind wings by much less than is typical of dedivus and the distance from the stigma tothe tip of the fore wing is usually larger, but they can be distinguished from pulvinatus by therelatively low pronotal lateral carinae and , in the male , the shape of the subgenital plate . Dr J . J .Presa has kindly informed us that he now believes that his records of this form from the Sierra deGuadarrama, Spain (Presa, 1978: 125) were based on misidentifiedpw/vmflfaj. The calling song of the male (see Figs 14, 41-45) cannot be recognized as dedivus with theunaided ear, but oscillographic analysis shows that it has a larger number of gaps per echemethan both pulvinatus and chopardi (see p. 119 for further discussion). In France dedivus occurs further north than pulvinatus and chopardi and seems, as one wouldexpect, to be more hygrophilous. It probably shows a similar ecological preference in northernSpain, but in Italy, where it is apparently the only species of Euchorthippus to occur on themainland, it seems to be able to tolerate drier conditions, tending to occupy the habitatsassociated with all three species in France and Spain. For discussion of the transfer of the subspecies elegantulus from dedivus to pulvinatus see p.131. We have examined seven female specimens, three from the MNHN, Paris and four from theIRSNB, Brussels, that have been regarded in these institutions as syntypes of Acridiumdedivum Brisout. We have also examined a number of further specimens of both sexes from theIRSNB that had previously been in the collection of de Selys Longchamps (who had earlieracquired Brisout's collection) and that could therefore also be regarded as possible syntypes ofthis species. Of all these specimens only one is labelled as being from one of the fourtype-localities named by Brisout (1849: 420); this is one of the three female specimens from theMNHN and is labelled 'St Germain'. We feel that this is the only specimen we have examinedthat can be regarded with confidence as a syntype and we have accordingly selected and labelledit aslectotype. We have examined the male and female syntypes of Stenobothrus pulvinatus var. gracilisAzam, and have selected and labelled the male as lectotype. We have also selected and labelled a male lectotype from the type-series of Euchorthippusdedivus meridionalis Jannone. This specimen, from Monte Angeli, is in much better conditionthan the male specimen labelled 'Tipo' by Jannone, which was from Altamura. MATERIAL EXAMINED Primary types (see synonymy) and 558 other specimens from the following localities. France. Marne (no further data). Paris. Yvelines: St-Germain. Essonne: Lardy. Seine-et-Marne:Fontainebleau; Fontainebleau Forest; Episy. Haute-Marne: Colombey-les-Deux-Eglises. Loire- THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 127 Atlantique: Varades. Yonne: 33 km E. of Sens, near Villeneuve-l'Archeveque. Cher: 23 km S. of Briare.Nievre: LaCharite. Puy-de-D6me: 12kmN. of Thiers; near Besse-en-Chandesse, Saurier; near Clermont-Ferrand, Puy de Crouelle; near Issoire, Clemensat. Dordogne: Bergerac; near Le Bugue, Campagne.Savoie: Flumet. Haute-Loire: St-Paulien; 5 kmN. of LePuy; 16kmS. of Cascades de la Baume. Lot: nearSouillac. Ardeche: nearPrivas. Lozere: 34 kmN. of Mende; near Mende, Col deMontmirat. Drome: nearValdrome. Landes: near Arcachon, Biscarrosse; near Morcenx. Vaucluse: Mont Ventoux; near Carpen-tras, Bedoin. Alpes-de-Haute-Provence: N. of Digne, La Rouine, 760-1070 m. Herault: 5 km E. ofMontpellier. Haute-Garonne: 15 km N. of Toulouse. Alpes-Maritimes: near St-Dalmas-de-Tende; Col deVence, 930 m; near Grasse, Greolieres; Mt Courmette; Thorenc; St-Martin-Vesubie; Auron, 1580 m. Var:c. 4 km W. of Cannes, Esveral; near Comps-sur-Artuby, Bargeme. Pyrenees-Orientales: Vernet-les-Bains, near Le Perthus. Spain. Santander: Playa de Cobreces. Leon: Riano. Huesca: Ordesa. Teruel: Montoro de Mezquita,Masada del Cerro, 1300 m; Villarluengo, 1270 m. Italy. Piemonte: Voltaggio. Liguria: Chiavari; Portofino, Vetta; near Sestri, Bracco Pass, 600 m. EmiliaRomagna: Vetto. Toscana: nearM. Amiata,4kmSW. of Bagni San Fillipo; 17kmE. of Siena; Siena, nearColleverde camp site; near Firenze, Futa Pass. Abruzzi: Gran Sasso d'ltalia, Fonte Cerreto, near StazFunivia, 340 m; G.S.I. , 2 km SW. of Valle Fredda; Civitella del Tronto. Lazio: Roma; near Casa del Corta;between Acquapendente and Lake Bolzena. Puglia: Monte Angeli; Altamura. Basilicata: Venosa, 420 m;Varco di Pietrastretta, near Potenza, 850 m. DISTRIBUTION (see Fig. 81). All France except for the north-western peninsulas and the extremenorth and north-east; northern and north-eastern Spain; southern Switzerland; most of Italy;Sicily and Sardinia. Central and eastern Europe (as far north as southern and eastern Austria,southern Czechoslovakia and south-western Ukraine); most of the Balkan Peninsula; AsiaMinor. E. declivus seems to be the only species of Euchorthippus occurring in mainland Italy,where it is widespread. There have been a number of published records of declivus from various parts of Spain; themore recent ones include those of Gangwere & Morales (1970: 58), Presa (1978: 125), Herrera(1979: 59), Gonzalez (1981: 61), Presa & Garcia (1982: 130) and Defaut (1982: 82, 85). In hisrecent catalogue Herrera (1982: 105) lists for declivus eight widely scattered Spanish provinces.Following an examination of specimens kindly lent to us by Drs L. Herrera and M. J. Gonzalez,we consider their records for the provinces of Navarra and Salamanca, respectively, to be basedin misidentifiedpw/vmarws, and Dr J. J. Presa has informed us that he now believes the same toapply to his records from the Sierra de Guadarrama and the record from Murcia province citedin his joint paper with Dr M. D. Garcia. Dr B. Defaut has kindly lent us the specimens on whichhis record from Teruel province was based and we agree with his identification. We have notbeen able to check the basis for the remaining records but, in view of the frequent confusionbetween these two species in Spain, are inclined to regard all Spanish records of declivus asneeding confirmation. Apart from those mentioned above from Teruel province, the onlySpanish specimens of declivus we have seen were from Santander, Leon and Huesca provincesand were kindly lent to us by Dr G. Kruseman of the Instituut voor Taxonomische Zoologie,Amsterdam, Dr V. Llorente of the Institute Espanol de Entomologia, Madrid, and Dr J. L.Santos of the Universidad Complutense, Madrid. The 97 specimens of Euchorthippus from theIberian Peninsula in the BMNH are all either pulvinatus or chopardi, and we have not seendeclivus during our own visits to Spain. Euchorthippus pulvinatus (Fischer de Waldheim) (Figs 27, 28, 32, 35, 36) Oedipoda pulvinata Fischer de Waldheim, 1846: 305. Type-material lost. Type-localities, U.S.S.R.: near Moscow; Kazan? ('Casan'); Karabag? ('Karabagh'); Caucasus.Euchorthippus pulvinatus (Fischer de Waldheim) Tarbmskii, 1926: 192. The nominate subspecies of pulvinatus does not occur in western Europe and has thus beenexcluded from this study. We have considered it best not to take it into account in the diagnosisgiven below, which is therefore based only on the two western subspecies gallicus and 128 D. R. RAGGE & W. J. REYNOLDS elegantulus. Most of the characters given in the diagnosis are probably equally true of p.pulvinatus, but this subspecies has longer wings, the fore wings usually reaching the tips of thehind femora in both sexes. The brief diagnoses given ioigallicus and elegantulus include only thecharacters by which these two subspecies differ from each other. DIAGNOSIS. cf $. Pronotal lateral carinae relatively high and narrow; length of pronotum usually 0-88-1-20times length of head in cf , 0-97-1-32 in $. Fore wings not usually reaching tips of hind femora in cf , notreaching bases of genicular lobes in $; distance from stigma to tip of fore wing usually 0-20-0-33 timeslength of hind femur in cf , 0-26-0-37 in $. Length of hind wing usually 0-91-1-16 times length of hindfemur in d", 0-79-1-00 in $. Stridulatory file usually with 115-165 pegs in d", 115-145 pegs or hairs in $.Male subgenital plate of moderate length and tending to be slightly pointed, as in Fig. 27. (Full data formeasurements, ratios and number of Stridulatory pegs are given in the bar-charts on pp. 144-150.) Male calling song (see Figs 15, 18, 46-64, 66-70). Echeme repetition rate usually 0-8-1-6/s. Syllablesusually lasting 20-35 ms. Other song characters as shown in bar-charts (p. 119). DISCUSSION. This species can be easily distinguished from declivus by the characters listed anddiscussed under that species. Distinguishing pulvinatus from chopardi where their rangesoverlap in southern France and the Iberian Peninsula is much more difficult, especially if areliable reference collection is not available. E. pulvinatus almost always lacks the stronglongitudinally striped colour pattern typical of chopardi (Fig. 40). The shorter fore wings (Figs27, 28, 83, 87) and prominent, narrow pronotal lateral carinae (Fig. 32) provide goodmorphological characters, and the distance from the stigma to the tip of the fore wing, taken as aratio with the length of the hind femur, is usually much smaller than in chopardi (Figs 85, 88). The best numerical method we have found for separating males of these two species is bycombining the number of Stridulatory pegs, the distance from the stigma to the tip of the forewing and the ratio of the length of the pronotum to the length of the head, and plotting the resulton a triangular graph (Fig. 12; see p. 110 for a full explanation). A simpler but rather lesseffective separation can be obtained by plotting two of these characters, Stridulatory pegs andstigma distance, on a two-axis graph (Fig. 9); this gives a particularly good separation for Frenchspecimens. Both sexes can be separated quite well by plotting stigma distance against length ofhind femur, as shown in Figs 10, 11; there is no overlap within our French and Iberian samples,but in the male there is a slight overlap between our samples of Iberian pulvinatus and Frenchchopardi. Because of the variable effect of shrinkage during drying, we have not found either the shapeor the dimensions of the male subgenital plate to be particularly useful in distinguishing betweendried specimens of pulvinatus and chopardi (see Fig. 85 for a comparison of the dimensions). The male calling song of pulvinatus can usually be distinguished from that of chopardi by thefaster echeme repetition rate, as shown in Figs 15, 16. Oscillographic analysis shows that theduration of the echemes, and especially of the syllables, is shorter than that of declivus andchopardi (Figs 18-20). In France the range of pulvinatus extends much further north than that of chopardi andDescamps (1968) has suggested that the former is less xerophilous than the latter (while beingmore so than declivus). There is, however, some indication from the material we have examinedthat this may not be so in Spain. Nearly all the specimens we have seen from the extreme south ofSpain have been pulvinatus and, although Pascual (1978) has shown that chopardi occurs in theSierra Nevada, he found it only at altitudes of 1800-2000 m, -whereas pulvinatus occurred at thelower altitudes of 1250-1450 m. DISTRIBUTION (see Fig. 81). In western Europe this species occurs in Jersey, France and theIberian Peninsula. The eastern subspecies, p. pulvinatus, occurs as far west as Czechoslovakia(Cejchan, 1981; 1982), eastern Austria and the Balkan Peninsula, and extends eastwardsthrough the Ukraine, southern Russia, Asia Minor, Kazakhstan and Central Asia to China. Therange of this species thus appears to be divided into two by quite a large gap, approximatelybetween longitudes 7E and 13E and including Switzerland, western Austria and the whole ofItaly; there have been some references in the literature to the occurrence of pulvinatus in Italy THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 129 * H-l ..... f'H 46 134/3 47 134/5 48 136/3 49 180/4 50 180/8 51 232/4 f 4' ill 52 232/5 53 234/8 54 235/2 20 Hz 0-1 s Figs 46-54 Oscillograms of single echemes from the calling songs of nine French males of Euchorthippuspulvinatus gallicus. The small numbers refer to the recordings from which the oscillograms were madeand can be used to obtain the full data from Tables 1 and 2 (pp. 107, 108). (e.g. La Greca, 1959: 139), but we have seen no Italian specimens of pulvinatus ourselves andare at present unconvinced by such references. If this gap is indeed a real one, it would clearlyaccount for the morphological divergence between/?, gallicus and the nominate subspecies (seealso p. 140). 130 D. R. RAGGE & W. J. REYNOLDS 55 272/3 56 272/9 57 274/3 58 309 59 313/1 4" "4 60 313/2 61 313/3 62 314/1 4 63 316/1 64 316/2 20 HzOls i Figs 55-64 Oscillograms of single echemes from the calling songs of ten Spanish males of Euchorthippuspulvinatus gallicus. The small numbers refer to the recordings from which the oscillograms were madeand can be used to obtain the full data from Tables 1 and 2 (pp. 107, 108). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 131 Euchorthippus pulvinatus gallicus Mar an (Figs 27, 28, 32) Euchorthippus pulvinatus gallicus Mafan, 1957: 187. Holotype cf, FRANCE: Draguignan, 1898 (N. Kheil)(NM, Prague) [examined]. DIAGNOSIS. cf $ . Length of head usually more than 2-4 mm in cf , more than 3-2 mm in $ . Length of forewing usually more than 9-2 mm in cf , more than 11-3 mm in $ . Length of hind femur usually more than 8-4mm in cf , more than 11-8 mm in $. (See also bar-charts on pp. 144-150.) DISCUSSION. For the best methods of distinguishing p. gallicus from the closely similar specieschopardi, see the species 'Discussion' for pulvinatus (p. 128). The relationship between thissubspecies and/?, elegantulus is discussed under that subspecies. MATERIAL EXAMINED Holotype (see above) and 573 other specimens from the following localities. France. Charente-Maritime: Saintes. Haute- Vienne: Limoges. Puy-de-D6me: near Clermont-Ferrand.Dordogne: near Le Bugue, Campagne; Bergerac. Lot: near Souillac. Lot-et-Garonne: near Agen.Ardeche: near Privas. Drome: near Nyons, 600 m; near Valdrome. Hautes-Alpes: near Laragne-Monteglin, Saleon. Aveyron: near Millau. Landes: Arcachon; near Morcenx. Gard: between Aries andNimes. Vaucluse: near Carpentras, Bedoin; Mont Ventoux, near Les Bruns; 3 km SE. of Cavaillon.Herault: several localities near Montpellier. Haute-Garonne: 15 km N. of Toulouse. Alpes-Maritimes:near Villefranche-sur-Mer; Cap d'Ail; Biot; Le Rouret; Sarree Valley, 600 m; near Nice, Cagnes.Bouches-du-Rhone: near Cassis; Petit Camargue, 22 km SE. of Aries, near Mas Thiberte; P.C.,Stes-Maries-de-la-Mer; 14 km E. of Salon. Var: Draguignan; Bagnoles; Montouroux; near Toulon, LaSeyne-sur-Mer; Hyeres; La St-Baume. Pyrenees-Orientales: near Banyuls-sur-Mer; near St-CyprienPlage; St-Cyprien; Vernet-les-Bains; near Le Perthus. Spain. Vizcaya: Sopelana. Burgos: 2 km S. of Lerma. Huesca: 1 km SE. of Seo de Urgel; near Ortedo, 7km SE. of Seo de Urgel; 2 km SE. of Ainsa; 5 km NW. of Solsona. Gerona: Port Bou; near Vilajuiga.Zaragoza: 12 km E. of Caspe. Salamanca: 5 km N. of Castraz, Rio Yaltes; Cervera; Cristo de La Laguna;Rodillo. Madrid: Sierra de Guadarrama; Madrid. Guadalajara: near Pastrana. Teruel: Albarracin.Cuenca: Belmonte; near Una. Valencia: Requena, 700 m. Jaen: Sierra de Cazorla, W. of Nava del Espino,1600-1700 m;S.C.,NavadeS. Pedro, 1400m;S.C., Puente de Las Harrerias. Murcia: Moratalla. Huelva:near Ayamonte. Malaga: near Torremolinos, RoyaMiel. Portugal. Guarda: Guarda. Leiria: Serra do Condieiros, near Alcanena, Ribatejo; Caldas da Rainha.Lisboa: Mafra; Ericeira; Lisboa, Parque Florestal de Monsanto. Evora: Evora. Setubal: R. Sado, nearMarateca. DISTRIBUTION (see Fig. 81). Central and southern France; Iberian Peninsula. The specimens onwhich Ebner (1931: 501) based his record of pulvinatus from Majorca have been lost and weconsider the occurrence of this species in the Balearic Islands to be doubtful. Euchorthippus pulvinatus elegantulus Zeuner(Figs 35, 36) Euchorthippus elegantulus Zeuner, 1940: 107. Holotype cf , JERSEY: Ouaine Bay, 4.ix.l938 (F. E. Zeuner) (BMNH, London) [examined]. Euchorthippus declivus elegantulus Zeuner; Mafan, 1957: 189; Descamps, 1968: 5.[Euchorthippus declivus (Brisout); Harz, 1975: 924. Erroneous synonymy.] DIAGNOSIS, c?$. Length of head usually less than 2-4 mm in cf , less than 3-2 mm in $. Length of fore wingusually less than 9-2 mm in cf, less than 11 -3 mm in $. Length of hind femur usually less than 8 -4 mm in cf,less than 11-8 mm in $ . (See also bar-charts on pp. 144-150.) DISCUSSION. This subspecies is essentially a small form of pulvinatus, the head showing the sizedifference most clearly (see Figs 82, 86). One of the more interesting conclusions to emerge from this study has been the fact thatelegantulus, which has hitherto always been associated with declivus, is much closer to pulvinatusin both morphological characters and song. When Zeuner (1940) originally described elegantu-lus (as a full species) he stated that it was related to declivus and compared the two in several 132 D. R. RAGGE & W. J. REYNOLDS characters while making no mention of pulvinatus. Chopard (1952) suggested that it 'n'est sansdoute qu'une sous-espece de dedivus' and, probably following this lead, Mafan (1957) andDescamps (1968) treated it formally as a subspecies of dedivus. Harz (1975) went even further,regarding it as no more than a synonym of dedivus. This association of elegantulus with dedivus probably stemmed from the fact that thesubgenital plate of the Jersey males is usually more pointed than is typical oi pulvinatus and issometimes nearer in shape to dedivus. However, in all other morphological characterselegantulus is closer \.o pulvinatus than to dedivus, and in some characters resembles pulvinatusvery closely; this is well shown by the bar-charts of the projection of fore wing beyond hind wing(Figs 84, 88), by those of ratios (Figs 81 , 83-87), in which the effect of size is largely eliminated,and by all the scatter diagrams, including the triangular graph (Figs 8-13). Zeuner (1940) evenincluded in his description of elegantulus the fact that the median and lateral carinae of thepronotum are 'very pronounced for the genus', another character shared with pulvinatus but notwith dedivus. The close relationship with pulvinatus is confirmed by the song of elegantulus,which agrees closely with that of pulvinatus gallicus in all respects, while showing a markedcontrast with dedivus in the number of gaps per echeme and the duration of both the echemesand the syllables (Figs 20-24, 65-70). I 65 dedivus 66 pulvinatus gallicus *-4 67 pulvinatus elegantulus (Brittany) 180/1 68 pulvinatus elegantulus (Brittany) 180/2 69 pulvinatus elegantulus (Jersey) 163 70 pulvinatus elegantulus (Jersey) 212 20 Hz 0-1 s Figs 65-70 Oscillograms of single echemes from the calling songs of four males of Euchorthippuspulvinatus elegantulus, with typical echemes from the male calling songs of E. dedivus and E. pulvinatusgallicus added for comparison. The small numbers refer to the recordings from which the oscillogramswere made and can be used to obtain the full data from Table 2 (p. 108). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 133 A further point of interest resulting from this study is that a form closely resembling the Jerseypopulation of elegantulus occurs in the southern part of Brittany, in the region extending fromQuiberon to Le Croisic. The Breton specimens are, on average, not quite as small as specimensfrom Jersey, and the male subgenital plate is usually blunter, more like typical pulvinatus.However, the bar-charts of absolute measurements (Figs 82-88) show that the Breton popula-tion is generally much closer to Jersey elegantulus than to French gallicus, and we think itreasonable, at least for the time being, to recognize these two populations as togetherconstituting the subspecies elegantulus. When the distribution and ecology of these forms havebeen more thoroughly studied, it may prove to be more sensible to regard the Jersey andBrittany populations as representing no more than island and coastal forms of gallicus, livingunder rather unfavourable conditions at the northern limit of the climatic tolerance of thissubspecies. A statistical comparison between our samples of gallicus and elegantulus gives tvalues of 9-5 for head length, 8-8 for fore wing length and 9-0 for hind femur length in the male,and corresponding values of 7-7, 7-3 and 6-7 in the female. The table given by Gery (1962) showsthat all these values indicate a non-overlap between these samples of between 75 per cent and 90per cent, which does not lend very strong support to their recognition as separate subspecies. MATERIAL EXAMINED Holotype (see p. 131) and 39 other specimens from the following localities. Jersey. Quennevais; Blanches Banques; St Ouen's Pond; Ouaine Bay; St Helier. France. Morbihan: Quiberon; Malanzac. Loire- Atlantique: 3 km SW. of Guerande; near Guerande, LeCroisic. DISTRIBUTION (see Fig. 81). Known only from Jersey and the extreme south of Brittany. Inaddition to the four Breton localities listed above, the record of pulvinatus from 'landes dePlescop, environs de Vannes' given by Sellier (1947) almost certainly refers to this subspecies. Euchorthippus chopardi Descamps(Figs 29, 30, 33, 40) Euchorthippus chopardi Descamps, 1968: 8. Holotype cf , FRANCE: Vaucluse, Cavaillon, route forestieredu Luberon, 150 m (M. Descamps) (MNHN, Paris) [examined]. DIAGNOSIS. O"$. Head and pronotum usually with conspicuous dark and light longitudinal stripes (Fig.40). Pronotal lateral carinae relatively low and broad, almost straight and parallel in metazona; length ofpronotum usually 0-85-1-25 times length of head. Fore wings relatively long, usually reaching at least tipsof hind femora in cf, bases of genicular lobes in $; distance from stigma to tip of fore wing usually0-33-0-46 times length of hind femur in cf , 0-40-0-54 in $. Length of hind wing usually 0-98-1-32 timeslength of hind femur in cf, 1-01-1 -29 in $. Male stridulatory file usually with 95-160 pegs. Male subgenitalplate relatively short and blunt, as in Fig. 29. (Full data for measurements, ratios and number ofstridulatory pegs are given in the bar-charts on pp. 144-150.) Male calling song (see Figs 16, 19, 71-77). Echeme repetition rate usually 0-5-0-7/s. Other songcharacters as shown in bar-charts (p. 119). DISCUSSION. For the easy separation of this species from declivus see the 'Discussion' under thatspecies. The much more difficult separation of chopardi from pulvinatus is discussed fully underthe latter species. The usual longitudinally striped colour pattern of chopardi (Fig. 40) provides astrong indication of its identity, and this can be confirmed for males by plotting their relevantmeasurements and peg-counts on the triangular graph shown in Fig. 12, and for either of thesexes by plotting their measurements on the scatter diagrams shown in Figs 9-11. For the bestmeans of distinguishing chopardi from albolineatus see the 'Discussion' under that species. MATERIAL EXAMINED Holotype (see above) and 122 other specimens from the following localities. France. Vaucluse: 3 km SE. of Cavaillon (type-locality); Mont Ventoux, 600-1200 m; near Carpentras, 3km S. of Malaucene. Herault: 5 km E. of Montpellier. Pyrenees-Orientales: Banyuls-sur-Mer; Vernet-les-Bains; near Vernet-les-Bains, Corneilla-de-Conflent. Spain. Orense: Leiro. Huesca: 8 km E. of Ainsa. Gerona: Port Bou. Zaragoza: 3 and 4-5 km E. of 134 D. R. RAGGE & W. J. REYNOLDS 72 236/4 73 236/6 * HHt 74 236/8 75 236/10 77 311/1 20 Hz 0-1 s Figs 71-77 Oscillograms of single echemes from the calling songs of five French (71-75) and two Spanish(76, 77) males of Euchorthippus chopardi. The small numbers refer to the recordings from which theoscillograms were made and can be used to obtain the full data from Tables 1 and 2 (pp. 107, 108). Notethat recordings 236/3, 236/4 and 236/6 were made at the type-locality. Caspe. Avila: Sierra de Gredos. Madrid: Sierra de Guadarrama. Cuenca: between Cuenca and Olivares;Belmonte. Valencia: near Gandia; Requena, 700 m. Alicante: near Alicante.Portugal. Leiria: Serra do Candieiros. DISTRIBUTION (see Fig. 81). A species occurring widely in the Iberian Peninsula and extending itsrange north-eastwards into the French departments bordering the Mediterranean Sea, andVaucluse. The only published record from the extreme south of Spain is from the Sierra Nevadaat altitudes of 1800-2000 m (Pascual, 1978). Euchorthippus albolineatus (Lucas) (Figs 34, 37)Oedipoda albo lineata Lucas, 1849: 38. In view of the major differences between the Sicilian taxon siculus and the North Africanpopulations of albolineatus (see especially Figs 8-11), we have strong doubts about the current THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 135 treatment of siculus as a subspecies of albolineatus (Descamps, 1968; Harz, 1975). We havetherefore not attempted to give a diagnosis that includes both these taxa, but have given each ofthem diagnoses as if they were distinct species. However, as this study is not primarily concernedwith either of these taxa, and we have no information on their songs, we prefer to retain thecurrently used nomenclature for them until their relationships can be more firmly established. Euchorthippus albolineatus albolineatus (Lucas)(Fig. 34) Oedipoda albo lineata Lucas, 1849: 38. LECTOTYPE cf , ALGERIA: Cercle de Lacalle, Boghar (H. Lucas} (MNHN, Paris), here designated [examined].Euchorthippus albolineatus (Lucas) Uvarov, 1926: 341.Euchorthippus albolineatus albolineatus (Lucas); Descamps, 1968: 5. DIAGNOSIS, cf $. Pronotal lateral carinae relatively high and narrow; length of pronotum usually 1-00-1-32times length of head. Fore wings relatively long, usually extending beyond tips of hind femora; distancefrom stigma to tip of fore wing usually 0-34-0-49 times length of hind femur in d" , 0-41-0-50 in 9 Length ofhind wing usually 1 -10-1-32 times length of hind femur in cf , 1 -04-1 13 in $ . Male stridulatory file usually3-28-4-70 mm long, with 100-140 pegs. Male subgenital plate relatively short and blunt (similar to Fig. 29).(Full data for measurements, ratios and number of stridulatory pegs are given in the bar-charts on pp.144-150.) DISCUSSION. E. albolineatus is at present the only species of the genus known from North Africaand so in practice identification presents no problems. We consider it doubtful that this speciesoccurs in the Iberian Peninsula (see 'Distribution' below), but if it did it could be quite easilyconfused with pulvinatus and chopardi. As shown by the triangular graph (Fig. 13) and two-axisscatter diagrams (Figs 9-11), a. albolineatus and chopardi can be distinguished from pulvinatusby the same combination of characters, but in all these diagrams a. albolineatus and chopardioverlap broadly. The best numerical characters for separating the males are the length of thestridulatory file and the ratio of pronotum to head (see Figs 82, 81), but this ratio is less effectivein separating the females. In qualitative characters both sexes of a. albolineatus usually lack thestrongly striped colour pattern typical of chopardi and have more prominent and sharply definedlateral carinae on the pronotum (cf. Figs 33, 34). E. a. albolineatus is clearly a close relative of both chopardi and pulvinatus , and is particularlyclose to chopardi in almost all its characters. If it does not occur in the Iberian Peninsula thequestion arises as to whether it might be better regarded as only subspecifically distinct fromchopardi. In the absence of any information on the song we can do no more than suggest this as apossibility to be borne in mind in future studies on the group. We have selected and labelled a male lectotype from the type-series of 10 males and 9 femalesof albolineatus. MATERIAL EXAMINED Lectotype (see synonymy) and 124 other specimens from the following localities. Morocco. Great Atlas Mts, Mouldirt, 1630 m; Oued Kroumane; Meknes district, Ifrane; 10 and 30 km S.of Fez; 12 km SE. of Azrou. Algeria. Near Algiers, Kouba; near Blida, Chrea, 1300 m; S. of Constantine, El Gehra; Djuradjura Mts,Bouira; Boghar; Boghari; Hauls Plateaux, Trolarel-Taza, 1100 m; Djelfa; Mascara. Libya. 30 km SW. of Tripoli, near Bianchi. DISTRIBUTION (see Fig. 81). North Africa, from western Morocco to Tripolitania. In view of theease with which this species can be confused with pulvinatus and chopardi, we consider that therecords of albolineatus from Portugal (Descamps, 1968: 6), Spain (Presa, 1978: 121; Gonzalez,1981: 61; Presa et /., 1983) and Majorca (Kruseman, in Harz, 1975: 929) need confirmation. 136 D. R. RAGGE & W. J. REYNOLDS Euchorthippus albolineatus siculus Ramme(Fig. 37) Euchorthippus pulvinatus siculus Ramme, 1927: 161. LECTOTYPE cf, SICILY: Fontanamurata, 17.vii.1924 (W. Ramme & W. Richter) (MNHU, Berlin), here designated [examined].Euchorthippus albolineatus siculus Ramme; Descamps, 1968: 5. DIAGNOSIS. cf 9 . Pronotal lateral carinae relatively high and narrow; length of pronotum usually 1-03-1-16times length of head in cf , 1-16-1-24 in $. Fore wings usually reaching bases of genicular lobes of hindfemora in cf , but usually failing to do so in $ ; distance from stigma to tip of fore wing usually 0-25-0-35times length of hind femur in cf , 0-28-0-34 in $ . Length of hind wing usually 0-97-1-09 times length of hindfemur in cf , 0-90-0-94 in $ . Stridulatory file usually with 85-110 pegs in cf , 60-90 pegs or hairs in $ . Malesubgenital plate of moderate length, sometimes tending to be slightly pointed (usually intermediatebetween the shapes shown in Figs 27 and 29). (Full data for measurements, ratios and number ofStridulatory pegs are given in the bar-charts on pp. 144-150.) DISCUSSION. The status of siculus is discussed above under albolineatus. The only other species of Euchorthippus known to occur in Sicily (according to Galvagni,1956: 352 and other authors) is declivus, which can be easily distinguished from siculus by thecharacters discussed on p. 126. To judge from the three male specimens we have examined, thissex of siculus can be distinguished from most other species of the genus by plotting the numberof Stridulatory pegs against the stigma distance, as shown in Fig. 9. The three female specimens we have examined can be separated from chopardi and a.albolineatus by the distance from the stigma to the tip of the fore wing, especially when taken as aratio to the length of the hind femur (Figs 1 1 , 88) . The only reliable character we have found forseparating females of siculus from this sex of pulvinatus is the number of vestigial pegs or hairs inthe Stridulatory file, as indicated in couplet 6 of the key (p. 124). We have selected and labelled a male lectotype from the type-series of siculus. MATERIAL EXAMINEDLectotype (see synonymy) , 2 cf , 3 9 paralectotypes from Fontanamurata and S . Maria de Gesu in Sicily . DISTRIBUTION. Known only from Sicily. Euchorthippus sardous Nadig(Fig. 38) Euchortippus [sic] sardous Nadig, in Nadig & Nadig, 1934: 18; La Greca, 1955: 4 [description of $ anddiscussion of relationships]. Holotype cf , SARDINIA: Gennargentu, 1900 m, 17.vii.1930 (A. Nadig) (Coll.Nadig, Chur, Switzerland) [examined]. DIAGNOSIS. cf $. Pronotal lateral carinae relatively high and narrow, distinctly incurved in prozona (Fig.38); length of pronotum less than 2-1 mm in cf, less than 3-0 mm in $, usually 0-85-1-05 times length ofhead. Length of hind femur usually 6-1-7-1 mm in cf , 8-2-10-4 mm in $ . Fore wings usually reaching basesof genicular lobes of hind femora in cf , but just failing to do so in $ ; distance from stigma to tip of fore wingusually 0-27-0-33 times length of hind femur in cf , 0-33-0-42 in $ . Length of hind wing usually 0-90-1-15times length of hind femur. Male Stridulatory file usually with 85-1 10 pegs. Male subgenital plate relativelyshort and blunt (similar to Fig. 29). (Full data for measurements, ratios and number of Stridulatory pegs aregiven in the bar-charts on pp. 144-150.) DISCUSSION. The small size of this endemic Sardinian species, and especially the short pronotumand hind femur (Figs 82, 86), enable both sexes to be easily distinguished from all other speciesof Euchorthippus. The only other species of the genus recorded from Sardinia is declivus, whichis equally easy to recognize (see p. 126). MATERIAL EXAMINEDHolotype (see above) and 30 other specimens from the type-locality. DISTRIBUTION. Known only from the type-locality. THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 137 Euchorthippus angustulus Ramme(Fig. 39) Euchorthippus angustulus Ramme, 1931: 191. Holotype cf , BALEARIC ISLANDS: Formentera, Can Marti,24-27.vii.1928 (M. Eisentraut) (MNHU, Berlin) [examined]. DIAGNOSIS, cf $ . Pronotal lateral carinae relatively high and narrow. Fore wings usually extending beyondtips of hind femora (usually 1 -25-1 -40 times length of hind femora) ; distance from stigma to tip of fore wingusually 0-43-0-56 times length of hind femur in cf , 0-50-0-61 in $ . Length of hind wing usually 1-22-1-37times length of hind femur in cf, 1-19-1-32 in $. Male stridulatory file usually with 103-133 pegs (mean of25 examined: 119-07-70). Male subgenital plate relatively short and blunt (similar to Fig. 29). Fig. 79 78 79 420 80 417 20 Hz 0-1 s Figs 78-80 Oscillograms of the calling songs of two males of Euchorthippus angustulus. Fig. 79 is a fasteroscillogram of the echeme indicated in Fig. 78. Fig. 80 is taken from the song of another male. The smallnumbers refer to the recordings from which the oscillograms were made and can be used to obtain the fulldata from Table 2 (p. 108). Male calling song (Figs 78-80). An analysis of songs recorded from two males from Palma Nova,Majorca (not available at the time the bar-charts and scatter diagrams were prepared) gave the followingmean values for each male. Echeme repetition rate: 0-84 and 0-85/s. Number of gaps per echeme: 5-6 and7-4. Duration of echeme: 161 and 164 ms. Duration of syllable: 31 and 32 ms. These figures are based on atotal of 179 echemes for the repetition rate and 125 echemes for the other song characters. Comparisonwith the bar-charts (Fig. 20) shows that the number of gaps per echeme agrees well with chopardi, theduration of both the echemes and the syllables is closer to pulvinatus , and the echeme repetition rate isintermediate between the two. This confirms Eisentraut's observations, quoted by Ramme (1931: 192),that the song of angustulus showed a slower echeme repetition rate than that of pulvinatus . MEASUREMENTS. The specimens of angustulus we have examined were not available at the timethe bar-charts were prepared and so their measurements are given separately below (number 138 D. R. RAGGE & W. J. REYNOLDS measured in parentheses; mean followed by standard deviation; all measurements in milli-metres). Of the 43 specimens measured, only 5 males and 7 females were available at the timethe morphological scatter diagrams were prepared. The measurements of four large females ofdoubtful identity from northern Majorca are excluded from this list; these specimens arediscussed on p. 139. Length of head Length of pronotum Pronotum/head Length of hind femur Length of stridulatory file Length of fore wing Length of hind wing Fore wing/hind femur Hind wing/hind femur Projection of fore wing beyond hind wing Distance from stigma to tip of fore wing Stigma distance/hind femur Length of subgenital plate Length of abdominal tergites 9+10 Subgenital plate/abdominal tergites 9+10 Males (26) 2-4-3-0, mean 2-770-20(25) 2-2-2-8, mean 2-5010-12(25)0-81-1-06, mean 0-91 0-07(25) 7-6-9-2, mean 8-420-42(25) 2-4-3-3, mean 2-880-22(25) 9-7-12-6, mean 11-3110-61(25)9-4-11-9, mean 10-840-55(25) 1-27-1-40, meanl-340-04(25) 1-22-1-37, meanl-290-04 (25) -0-32-0-24,meanO-010-12 (26) 3-7-4-6,mean4-160-25 (25) 0-43-0-56,mean 0-500-02 (23) 0-44-0-76,mean 0-600-09 (26) 0-44-0-68,meanO-530-06 (23)0-73-1-50,mean 1-14+0-20 Females (16)2-8-3-6, mean 3-21 0-24(16) 2-8-3-4, mean3-160-17(16)0-89-1-12, mean 0-990-07(16)9-7-11-6, mean 10-510-65 (16) 12-6-15-6, mean 13-8710-76(16) 12-4-14-3, mean 13-2810-61(16) 1-25-1-40, mean 1-3210-05(16) 1-21-1-32, mean 1-27 10-03(14) -0-36-0-52, mean 0-1010-22(15)5-6-6-3, mean 6-01 10-20(15)0-50-0-61, mean 0-5610-03 DISCUSSION. This species, originally described from Formentera and since found in Majorca andIbiza, is endemic to the Balearic Islands. With the possible exception of four large females fromnorthern Majorca (discussed below), all the specimens we have seen from these islands seem tobelong to angustulus, although pulvinatus and even albolineatus have been recorded fromMajorca (Ebner, 1931; Kruseman, in Harz, 1975: 929). Both sexes of angustulus can bedistinguished from pulvinatus by the relatively long fore and hind wings, and by the largedistance from the stigma to the tip of the fore wing, all these being taken as ratios to the length ofthe hind femur (Figs 10, 11); the small pronotum and hind femur enable both sexes to bedistinguished from albolineatus. From chopardi, which could also conceivably occur in theBalearic Islands, it can be distinguished by the prominent and narrow pronotal lateral carinae.The fact that angustulus males are closely associated with chopardi and well-separated frompulvinatus on the triangular graph shown in Fig. 13 suggests the possibility that angustulus is anisland derivative from Iberian stock of chopardi. Not surprisingly, the Formenteran and Ibizan specimens we have examined differ slightlyfrom the Majorcan ones. The Formenteran holotype and allotype show a tendency for thepronotal lateral carinae to be more incurved than is typical of Majorcan specimens, in whichthese carinae are usually straight in the prozona. The single Ibizan male we have examined (fromTalamanca) is unusually small, giving smaller values than any other male for most of the THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 139 measurements taken; the two Ibizan females examined (from Santa Eulalia del Rio) were ofmore typical size for angustulus. We have examined four female specimens from northern Majorca (one from Puerto dePollensa, on loan from the Instituut voor Taxonomische Zoologie, Amsterdam, and three fromAlcudia, on loan from the MNHU, Berlin) that are considerably larger than any other Balearicspecimens we have seen. The larger size is shown particularly by the following measurementstaken from these specimens: length of pronotum 3-6-3-9 mm, length of hind femur 12-1-12-6mm, length of fore wing 15-5-16-8 mm, length of hind wing 14-3-16-2 mm, distance from stigmato tip of fore wing 0-60-0-71 mm. The hind femora of the largest of these females are as long asthose of the smallest female we have examined of albolineatus , and there is a small overlap in thelength of the pronotum; however, the ratios of both the fore and hind wings to the hind femur inall four specimens give values that are well outside the range of these ratios in our sample ofalbolineatus . It is impossible to form definite conclusions about these four specimens withoutfurther material, and in particular associated males, but they could be no more than a large localvariant of angustulus. It may be significant that the Albufera marsh lies just south of Alcudia;this area of marshland, which is quite unique in the Balearic Islands, could conceivably havebeen the source of these specimens. MATERIAL EXAMINED Holotype (see p. 137) and 42 other specimens from the following localities. Balearic Islands. Formentera: Can Marti. Ibiza: Talamanca; Santa Eulalia del Rio. Majorca: Palma;Palma Nova; Seller; Puerto de Seller; Puerto de Pollensa; Capdella. DISTRIBUTION. Known only from the Balearic Islands. General discussion Groups of morphologically very similar species are quite common in the Gomphocerinae. Whenthe members of such a group intermingle in the same habitat, there are usually obviousdifferences between the calling songs of the males. This is not surprising since it is now generallybelieved - and has been proved experimentally in some cases - that such song differencesprovide the main reproductive isolating mechanism. The three main western European speciesof Euchorthippus - dedivus, pulvinatus and chopardi - are sufficiently similar morphologicallyfor misidentification to be frequent. In attempting to assign the 244 males we had at our disposalto one or other of these species on the basis of the shape and length of the subgenital plate (thecharacter on which most emphasis has been placed in the past), we soon found that they showeda continuous gradation from the longest and most pointed dedivus to the shortest and bluntestchopardi, and a large number of males defied identification on this character alone. (In terms oflength, the overlap in this character can be seen clearly in Fig. 85.) Since we knew it was notunusual for two of the species to be found together, and there was no obvious difference in themale calling songs, we began to wonder whether they were no more than forms of a singlespecies. The clear morphological separation achieved by the combination of characters used in Fig. 12made us more confident that three species were involved, and we then turned to the song in thehope of finding some evidence of an ethological barrier. Careful analysis gradually revealeddifferences that, although subtle and mostly undetectable by the unaided human ear, seemedsufficiently clear to provide a basis for mate recognition. It is perhaps significant that thedifference in the male calling song is smallest between dedivus and chopardi, which as far as weknow have never been found together; pulvinatus, which quite often occurs with either dedivusor chopardi, has a noticeably higher echeme repetition rate than either of them in addition to thefurther differences revealed by oscillographic analysis. The close resemblance in both morphology and song shown by these three species suggeststhat they evolved from a common ancestor quite recently, probably during the PleistocenePeriod. The Pleistocene glaciations, during which such an ancestor would no doubt have beenpushed southwards into the Iberian, Italian and Balkan Peninsulas, seem likely to have 140 D. R. RAGGE & W. J. REYNOLDS provided the geographical separations that would have tended to lead to morphological andethological divergence . As can be seen from Fig. 81 , there are some anomalies in the distributionof these species that we think can also be explained by the Pleistocene fluctuations in climate.These are the apparent absence of dedivus from most of the Iberian Peninsula, in spite of itsbeing widespread in Italy and the Balkan Peninsula; the opposite situation in pulvinatus , whichis common in the Iberian Peninsula but apparently absent from Italy and rare in the Balkans; andthe large apparent gap between the western and eastern populations of pulvinatus , i.e. betweenp. gallicus and p. pulvinatus. One possibility is that dedivus evolved from a population of ancestral stock isolated in theItalian Peninsula during a prolonged glacial period. In the course of subsequent warmer climaticphases such a population would have been able to spread northwards, eventually extending itsrange into much of western and eastern Europe, while leaving in the peninsula the existingItalian populations of dedivus (which would have had to become adapted to the amelioratingclimate). The poor penetration of dedivus into the Iberian Peninsula could well be explained bythe major barrier of the Pyrenees together with competition by other species of Euchorthippusalready well adapted to the very hot, dry summers of the Spanish Meseta. A population of Euchorthippus isolated in the Iberian Peninsula, perhaps during the sameglacial period, could have given rise to pulvinatus gallicus as a result of a rather smallerdivergence from the main body of the ancestral stock (now pulvinatus pulvinatus} occurring inthe Balkan Peninsula and further east. Competition with dedivus, perhaps by then well Fig. 81 Map showing the distribution of the western European and North African species of Euchorthip-pus. We regard as doubtful the past records of E. dedivus from central and southern Spain (see p. 127),E. pulvinatus from Italy and Majorca (see pp. 128, 131) and E. albolineatus from Portugal, Spain andMajorca (see p. 135), and have therefore not taken them into account in preparing this map. THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 141 established in northern Italy and Yugoslavia, could have prevented the western and easternpopulations of pulvinatus from coming together again. The evolution of albolineatus has no doubt been the result of isolation in North Africa, and wesuspect that chopardi is a comparatively recent northern derivative from albolineatus that hasmanaged to cross the Strait of Gibraltar (perhaps at a time when it was narrower than at present)and thence spread through Spain and past the eastern end of the Pyrenees into southern France,at the same time diverging somewhat from the parental North African stock. Isolation in theBalearic Islands, Sardinia and Sicily, respectively, is sufficient to account for the evolution ofangustulus, sardous and albolineatus siculus. Insular isolation would also account for themorphological divergence of the Jersey population of p. elegantulus, though the occurrence of asimilar population in southern Brittany suggests that elegantulus may be no more than a smallnorthern form of p. gallicus occurring in rather unfavourable conditions at the limit of its climatictolerance. Although we feel we have been able to throw some further useful light on the taxonomicproblems presented by Euchorthippus , many questions still remain to be answered. We shouldwelcome the opportunity to record and analyse the song of albolineatus, so that its relationshipwith the European species, particularly chopardi, could be better understood. An analysis of thesong of the eastern subspecies p. pulvinatus would help to establish whether this form is trulyconspecific with the western subspecies of pulvinatus whose songs are described and analysed inthis paper. Information on the songs of sardous and siculus would also be useful in elucidatingtheir affinities. An experimental test of the effectiveness as an isolating mechanism of the songdifferences we have detected would be most valuable, as would a further investigation into therelated question of hybridization; such studies should take account of the possibility ofpheromonal reinforcement of mate recognition, which could conceivably account for theunusually small interspecific song differences in this group. A comparative study of thechromosomes, and particularly the C-banding patterns, of all the western European speciesmight well throw some further light on their inter-relationships. Finally, a comprehensive studyof the eastern Palaearctic and Madeiran species of Euchorthippus, based on morphology,cytology and song, would lead nearer to a complete understanding of the taxonomy andevolution of this interesting genus. References Arana, P., Santos, J. L. & Giraldez, R. 1980. Chiasma interference and centromere co-orientation in a spontaneous translocation heterozygote of Euchorthippus pulvinatus gallicus (Acrididae; Orthoptera). Chromosoma 78: 327-340.Autrum, H. 1963. Anatomy and physiology of sound receptors in invertebrates [pp. 412-433]. In Busnel, R.-G. [Ed.], Acoustic behaviour of animals. xx+933pp. Amsterdam.Azam, J. 1901. Catalogue synonymique et systematique des Orthopteres de France (suite). Miscellanea Entomologica 9: 33-48.Baccetti, B. 1958. Notulae orthopterologicae. X. Indagini sugli Ortotteri del Gran Sasso d'ltalia per il Centre di Entomologia Alpina. Redia 43: 351^50.Bei-Bienko, G. Ya. 1961. Some features of changes in invertebrate fauna when virgin steppe is cultivated. Entomologicheskoe Obozrenie 40: 763-775. [In Russian. English translation: 1962, Entomological Review, Washington 40 (1961): 427-434.]Bei-Bienko, G. Ya & Mishchenko, L. L. 1951. Locusts and grasshoppers of the U.S.S.R. and adjacent countries. Part II. Opredeliteli po Faune SSR, Izdavaemye Zoologicheskim Muzeem Akademii Nauk, Leningrad 40: 381-667. [In Russian. English translation published by Israel Program for Scientific Translations, Jerusalem, 1964, iv+291+xxi pp.]Brisout de Barneville, L. 1849. Catalogue des Acridides qui se trouvent aux environs de Paris. Annales de la Societe Entomologique de France (2) 6 (1848): 411-425.Broughton, W. B. 1976. Proposal for a new term 'echeme' to replace 'chirp' in animal acoustics. Physiological Entomology 1: 103-106.Carlson, J. G. 1936. The intergeneric homology of an atypical euchromosome in several closely related Acridinae (order Orthoptera). Journal of Morphology 59: 123-161. 142 D. R. RAGGE & W. J. REYNOLDS Cejchan, A. 1981. On the orthopteroid insects (s.l.) of Czechoslovakia. I. Casopis Ndrodniho Muzea v Praze 149 (1980): 125-139. [In Czech with English summary.] 1982. On the ort'hopteroid insects (s.l.) of Czechoslovakia III. Casopia Ndrodniho Muzea v Praze 151: 1-13. [In Czech with English summary.]Chetyrkina, I. A. 1954. Acridoidea of steppes and deserts of the region of the river Ural. Trudy Zoologicheskogo Instituta, Akademiya Nauk SSSR, Leningrad 16: 229-284. [In Russian.]Chopard, L. 1952. Orthopteroides. Faune de France 56 (1951), 359 pp.Defaut, B. 1982. Quelques precisions sur la determination, la reparation geographique et 1'ecologie des especes franchises du genre Euchorthippus [Orth. Acrididae]. L' Entomologiste 38: 82-87.Descamps, M. 1968. Notes sur le genre Euchorthippus [Orth. Acrididae]. Sa repartition dans le Vaucluse et les departements adjacents. Annales de la Societe Entomologique de France (N.S.) 4: 5-25.Ebner, R. 1931. Einige Orthopteren von Mallorca. Boletin de la Sociedad Espanola de Historia Natural3l: 497-503.Faber, A. 1953. Laut- und Gebardensprache bei Insekten. Orthoptera (Geradfliigler). Teil I. 198 pp. Stuttgart.Ferrer, E., Lacadena, J. R. & Jodar, B. 1981. Analisis de la asociacion somatica de cromosomas homologos en Euchorthippus pulvinatus gallicus Maran (Acrididae, Orthopthera [sic]). Revista de la Real Academia de Ciencias Exactas, Fisicas y Naturales de Madrid 74: 639-651.Fischer de Waldheim, G. 1846. Orthopteres de la Russie. Nouveaux Memoires de la Societe Imperiale des Naturalistes de Moscou 8, iv+413 pp.Galvagni, A. 1956. Primo contributo alia conoscenza degli Ortotteroidei dei Colli Euganei (Veneto). Memorie del Museo Civico di Storia Naturale di Verona 5: 337-359.Gangwere, S. K. & Morales Agacino, E. 1970. The biogeography of Iberian Orthopteroids. Misceldnea Zoological (5): 1-67.Gery, J. 1962. Le probleme de la sous-espece et de sa definition statistique (a propos du coefficient de Mayr-Linsley-Usinger). Vie et Milieu 13: 521-541.Gonzalez Garcia, M. J. 1981. Contribucion al conocimiento de los Acridoidea (Orth.) de la Dehesa Salmantina. Boletin de la Asociacion Espanola de Entomologia 4 (1980): 55-64.Harz, K. 1975. Die Orthopteren Europas II. Series Entomologica 11, [viii+] 939 pp.Herrera Mesa, L. 1979. Contribucion al conocimiento de los Acridoideos (Orth. Acridoidea) de la provincia de Navarra. Boletin de la Asociacion Espanola de Entomologia 3: 45-64.1982. Catalogue of the Orthoptera of Spain. Series Entomologica 22, viii+162 pp.Hewitt, G. M. 1979. Orthoptera. Grasshoppers & crickets. Animal Cytogenetics 3 (Insecta 1), v+170pp.Jacobs, W. 1953. Verhaltensbiologische Studien an Feldheuschrecken. Zeitschrift fur Tier psychologic, Beiheftl,vii+228pp.Jannone, G. 1937. Contributi alia conoscenza dell'Ortotterofauna italica. 1. Nota preventiva su alcune specie e subspecie nuove opoco note della Puglia e di altre regione dell'Italia meridionale e insulare. Bolletino di Zoologia, Pubblicato dall'Unione Zoologica Italiana 8: 51-76.John, B. & Hewitt, G. M. 1966. Karyotype stability and DNA variability in the Acrididae. Chromosoma 20: 155-172.La Greca, M. 1955. Su alcuni Mantodei e Ortotteri italiani poco noti. Annuario dell' Istituto e Museo di Zoologia dell'Universitd di Napoli 6 (1954) (12): 1-11.- 1959. L'Ortotterofauna Pugliese ed il suo significato biogeografico. Memorie di Biogeografia Adriatica, Istituto di Studi Adriatici 4: 33-170. [This paper appears to have been issued in separate form well in advance of the relevant part of the journal; we have given the year cited on the separate (1959) in preference to that cited on the journal part (1962).]Litvinova, N. F. 1972. Patterns of geographic variability of morphological characters in the genus Euchorthippus (Orthoptera, Acrididae). Zoologicheskii ZhurnalSl: 821-827. [In Russian with English summary.]Lucas, H. 1849. Histoire naturelle des animaux articules. Cinquieme classe. Insectes. (Suite.) Deuxieme ordre. Les Orthopteres. Exploration Scientifique de I'Algerie pendant les annees 1840, 1841, 1842 (Sciences physiques, Zoologie) 3: 1-39.Luquet, G. C. 1978. La systematique des Acridiens Gomphocerinae du Mont Ventoux (Vaucluse) abordee par le biais du comportement acoustique [Orthoptera, Acrididae]. Annales de la Societe Entomologique de France (N.S.) 14: 415^50.Maran, J. 1954. Die Orthopterenfauna der staatlichen Naturschutzgebiete bei Sturovo in der Siidslowakei (CSR). Ochrana Pfirody 9: 132-139. [In Czech with Russian and German summaries.] THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 143 - 1957. Beitrag zur Kenntnis der europaischen Arten der Gattung Euchorthippus Tarb. (Orthoptera,Acrididae). Acta Entomologica Musei Nationalis Prague 31: 183-190. Mayr, E. 1969. Principles of systematic zoology, xi+428 pp. New York. McClung, C. E. 1932. Multiple chromosomes in the Orthoptera. Archivos de la Sociedad de Biologia deMontevideo 7: 1831-1848. Nadig, A. & Nadig, A. 1934. Beitrag zur Kenntnis der Orthopteren- und Hymenopterenfauna vonSardinien und Korsika. Jahresbericht der Naturforschenden Gesellschaft Graubundens 72: 3-39. Pascual, F. 1978. Estudio preliminar de los Ortopteros de Sierra Nevada, I: Introduction general einventario des especies. Boletin de la Asociacion Espanola de Entomologia I (1977): 163-175. Perdeck, A. C. 1957. The isolating value of specific song patterns in two sibling species of grasshoppers(Chorthippus brunneus Thunb. and C. biguttulus L.). [viii+] 75 pp. Leiden. Pitkin, L. M. 1976. A comparative study of the stridulatory files of the British Gomphocerinae (Orthop-tera: Acrididae). Journal of Natural History 10: 17-28. Presa, J. J. 1978. Los Acridoidea (Orthoptera) de la Sierra del Guadarrama. 277 pp. Madrid. Presa, J. J. & Garcia, M. D. 1982. Contribuci6n al conocimiento de los Acridoidea (Orth.) de la regionmurciana (II): Sierra seca de revolcadores. Boletin de la Asociacion Espanola de Entomologia 6:127-132. Presa, J. J., Montes, C. & Ramirez-Diaz, L. 1983. Tipificacion de poblaciones de saltamontes (Orth.Acrididae) en relation con la altitud, pisos y tipos de vegetation en la Sierra de Guadarrama (SistemaCentral, Espana). Boletin de la Asociacion Espanola de Entomolgia 6: 249-265. Ramme, W. 1927. Die Dermapteren und Orthopteren Siziliens und Kretas. Eos 3: 111-200. - 1931. Beitrage zur Kenntnis der palaearktischen Orthopterenfauna (Tettig. et Acrid.). [I.] Mit-teilungen aus dem Zoologischen Museum in Berlin 17: 165-200. - 1939. Beitrage zur Kenntnis der palaearktischen Orthopterenfauna (Tettig. u. Acrid.). III. Mit-teilungen aus dem Zoologischen Museum in Berlin 24: 41-150. Santos, J. L. & Giraldez, R. 1982. C-Heterochromatin polymorphism and variation in chiasma localization in Euchorthippus pulvinatus gallicus (Acrididae, Orthoptera). Chromosoma 85: 507-518.Santos, J. L., Arana, P. & Giraldez, R. 1983. Chromosome C-banding patterns in Spanish Acridoidea. Genetica 61: 65-74.Schmidt, G. H. & Schach, G. 1978. Biotopmassige Verteilung, Vergesellschaftung und Stridulation der Saltatorien in der Umgebung des Neusiedlersees. Zoologische Beitrage (N.F.) 24: 201-308.Sellier, R. 1947. Materiaux pour un catalogue des Orthopteres et Dermapteres de Bretagne. Premiere liste. Bulletin de la Societe Scientifique de Bretagne 21 (1946): 113-122.Tarbinskii, S. P. 1925. Materials concerning the Orthopterous fauna of the province of Altai. Russkoe Entomologicheskoe Obozrenie 19: 175-195. [In Russian with English summary.]Tsyplenkov, E. P. 1970. Harmful Acridoidea of the U.S.S.R. 272 pp. Leningrad. [In Russian. English translation published by Amerind Publishing Co., New Delhi, 1978, viii+208 pp.]Uvarov, B. P. 1926. New or less known Acrididae from Central Asia. Eos 2: 321-359.White, M. J. D. 1972. The value of cytology in taxonomic research on Orthoptera [pp. 27-33]. In Hemming, C. F. & Taylor, T. C. H. [Eds] , Proceedings of the International Study Conference on Current and Future Problems of Acridology, London, United Kingdom, 6-16 July 1970. xv+533pp. London.Zeuner, F. E. 1940. The Orthoptera Saltatoria of Jersey, Channel Islands. Proceedings of the Royal Entomological Society of London (B) 9: 105-110. 144 D. R. RAGGE & W. J. REYNOLDS SOURCE HEAD 2-5 3-0 3-5 2-0 PRONOTUM 2-5 3-0 3-5 declivus Italy pulvinatusgallicui Iberia pulvinatuselegantulus Jersey Brittany chopardi France Iberia a/bo/ineafus a/bo/ineafus NorthAfrica afbo/ineafus SICU/US Sicily sardous Sardinia 2-0 2-5 3-0 3-5 2-0 2-5 3-0 3-5 Fig. 82 Data for length of head, length of pronotum, length of hind femur, and length of pronotumdivided by length of head, in males of Euchorthippus . Measurements are given in millimetres. Forexplanation of the bars see Fig. 7 (p. 110). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 145 Fig. 83 Data for number of stridulatory pegs, length of stridulatory file, length of fore wing and length ofhind wing in males of Euchorthippus . Measurements are given in millimetres. For explanation of thebars see Fig. 7 (p. 110). 146 D. R. RAGGE & W. J. REYNOLDS Fig. 84 Data for length of fore wing divided by length of hind femur, length of hind wing divided by lengthof hind femur, projection of fore wing beyond hind wing, and distance from stigma to tip of fore wing, inmales of Euchorthippus . Measurements are given in millimetres. For explanation of the bars see Fig. 7(p. 110). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 147 Fig. 85 Data for distance from stigma to tip of fore wing divided by length of hind femur, length ofsubgenital plate divided by length of abdominal tergites 9+10, length of subgenital plate, and length ofabdominal tergites 9+10, in males of Euchorthippus . Measurements are given in millimetres. Forexplanation of the bars see Fig. 7 (p. 110). 148 D. R. RAGGE & W. J. REYNOLDS SOURCE HEAD 2-5 3-0 3-5 4-0 4-5 PRONOTUM 2-5 3-0 3-5 4-0 4-5 50 dec/ivus France Italy pu/vinafus gallicus France Iberia pu/vi'nafuse/eganfu/us Jersey Brittany chopardi France 36 Iberia a/bo/jnearusa/bo/ineatus NorthAfrica a/bo/ineafussiculus Sicily sardous Sardinia 2-5 3-0 3-5 40 4-5 25 3-0 3-5 40 45 5-0 Fig. 86 Data for length of head, length of pronotum, length of hind femur, and length of pronotumdivided by length of head, in females of Euchorthippus . Measurements are given in millimetres. Forexplanation of the bars see Fig. 7 (p. 110). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS 149 Fig. 87 Data for length of fore wing, length of hind wing, length of fore wing divided by length of hindfemur, and length of hind wing divided by length of hind femur, in females of Euchorthippus .Measurements are given in millimeters. For explanation of the bars see Fig. 7 (p. 110). 150 D. R. RAGGE & W. J. REYNOLDS Fig. 88 Data for projection of fore wing beyond hind wing, distance from stigma to tip of fore wing, anddistance from stigma to tip of fore wing divided by length of hind femur, in females of Euchorthippus .Measurements are given in millimetres. For explanation of bars see Fig. 7 (p. 110). THE TAXONOMY OF THE WESTERN EUROPEAN EUCHORTHIPPUS Index 151 Invalid names are in italics; principal page references in bold. albolineatus 111, 112, 113, 116, 117,122, 124, 133, 134, 138, 139, 141 albopunctata 121 alini 121 angustulus 111, 112, 116, 121, 122,124, 137, 141 arabicus 122 brachyptera 121 chenbaensis 122 cheui 122 chopardi 110, 111, 112, 113, 114,115,116,117,119,120,121,122,126, 127, 128, 133, 135, 136, 137,138, 139, 141 Chorthippus 121 Chrysochraon 121 declivus 110, 112, 113, 114, 115,116,117,118,119,120,121,122, 125, 128, 131, 132, 133, 136, 139, 140 dichrous 121dispar 121dorsatus 121 elegantulus 113, 114, 115, 119, 124,126, 128, 131, 141 gallicus 112, 113, 114, 115, 116, 117,119, 124, 127, 128, 129, 131, 132,133, 140, 141 gracilis 125, 126 madeirae 122, 126meridionalis 125, 126Metrioptera 121montanus 121 parallelus 121 Platycleis 121 pulvinatus 104, 110, 111, 113, 115,116, 117, 118, 119, 120, 121, 122,124, 126, 127, 135, 136, 137, 138,139, 140, 141 sabulosa 121 sardous 112, 115, 116, 122, 136, 141 siculus 112, 115, 116, 134, 135, 136, 141 Sinhippus 121stichai 125 transcaucasicus 122unicolor 121, 122weichowensis 122yungningensis 122 British Museum (Natural History) Milkweed butterflies: their cladistics and biology P. R. Ackery & R. I. Vane- Wright The Danainae, a subfamily of the Nymphalidae, contains only some 150 species, yet aspects oftheir biology have stimulated far more attention than can be justified by species numbersalone. In recent years, an expansive literature has grown, considering aspects of theircourtship and pre-courtship behaviour, migration, larval hostplant associations, mimicry andgenetics. The popularity of danaines among biologists can certainly be attributed to thiscombination, within one small group, of so many of the factors that make butterflies such aninteresting group to study. The obvious need to place this wealth of biological data within anacceptable systematic framework provided the impetus for this volume. Started eight years ago within the conventions of evolution by natural selection andHennig's phylogenetic systematics, the book is now largely about natural history (what theanimals have and do, where they live and how they develop) and natural groups - as revealedby a form of analysis approaching that practised by the new school of 'transformed cladistics'.The authors have prepared a handbook that will appeal to a wide range of biologists, frommuseum taxonomists to field ecologists. 424 pp (approx. ) , 12 pp colour, 73 b/w plates, line and graphic illustrations, maps, extensive bibliography.ISBN 565 00893 5. Publication September 1984. Price 50, prepublication price 45. Titles to be published in Volume 49 Afrotropical jumping plant lice of the family Triozidae (Homoptera: Psylloidea). By David Hollis. The taxonomy of the western European grasshoppers of the genus Euchorthippus, with specialreference to their songs (Orthoptera: Acrididae). By D. R. Ragge & W. J. Reynolds An historical review of the higher classification of the Noctuidae (Lepidoptera). By Ian J. Kitching The Pimplinae, Xoridinae, Acaenitinae and Lycorininae (Hymenoptera: Ichneumonidae) ofAustralia. By I. D. Gauld The western Palaearctic species of Ascogaster (Hymenoptera: Braconidae) By T. Huddleston Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, SuffolkPrinted in Great Britain by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natura BRITi :'. IN AU. 3 1 ; . i ^History) An historical review of the higherclassification of the Noctuidae(Lepidoptera) Ian]. Kitching Entomology series Vol 49 No 3 30 August 1984 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique andever-growing collections of the Museum, both by the scientific staff of the Museum and byspecialists from elsewhere who make use of the Museum's resources. Many of the papers areworks of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself,available separately, and individually priced. Volumes contain about 300 pages and severalvolumes may appear within a calendar year. Subscriptions may be placed for one or more ofthe series on either an Annual or Per Volume basis. Prices vary according to the contents ofthe individual parts. Orders and enquiries should be sent to: 30AUGI934 Publications Sales, British Museum (Natural History),Cromwell Road, London SW75BD,England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.) Trustees of the British Museum (Natural History), 1984 The Entomology series is produced under the general editorship of the Keeper of Entomology: Laurence A. Mound Assistant Editor: W. Gerald Tremewan ISBN 565 06005 8ISSN 0524-6431 British Museum (Natural History)Cromwell RoadLondon SW7 5BD Entomology seriesVol 49 No 3 pp 153-234 Issued 30 August 1984 An historical review of the higher classification of the Noctuidae (Lepidoptera) * Ian J. Kitching Department of Entomology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Contents Synopsis .................................................................................................... 153 Acknowledgements ..................................................................................... 153 Introduction .............................................................................................. 153 In the beginning .......................................................................................... 154 The American alternative .............................................................................. 165 Classifications in conflict ............................................................................... 168 Development of the Hampsonian monolith ....................................................... 180 The influence of immatures ............................................................................ 190 The main assault ......................................................................................... 195 The neglected arrangement ........................................................................... 210 Quo vadis, Noctual ..................................................................................... 214 References ................................................................................................ 227 Index ........................................................................................................ 233 Synopsis An historical review of the development of the higher classification of the Noctuidae is presented,emphasising the interrelationships between the various systems that have been put forward and thetaxonomic impact of the character complexes that have been employed. A provisional cladistic analysis ofthe higher noctuid taxa is performed and a tentative cladogram proposed. Many of the currently acceptedsubfamilies are rejected as potentially paraphyletic or polyphyletic assemblages. New, cladistic, definitionsof the families Arctiidae and Noctuidae are suggested, which would necessitate the establishment of theAganaidae and Herminiidae as separate families. Acknowledgements This study was carried out during the tenure of a Junior Research Fellowship awarded by the Trustees ofthe British Museum (Natural History), which is gratefully acknowledged. I thank my colleagues at theBritish Museum (Natural History) for their support and comments. Introduction It is exceptional to find any two authors who use the same combination of subfamilynames within the Noctuidae (Nye, 1975: 10). The classification of this family [Noctuidae] rests in a state of great confusion, and fewauthors appear to hold similar views regarding the suprageneric taxonomy. I haveconcluded after spending much time (perhaps I should say wasting time) on the problem,that it is impossible to present ... a correct suprageneric classification (Zimmerman,1958: 197-198). The subfamily divisions . . . are on the whole natural, but their defining charactershave far too many exceptions, and some genera may perhaps be completely misplaced(Forbes, 1954: 5). Much of the present grouping of species and genera is still debatable, and reasons forgroupings are often obscure (Birch, 1972a: 189). nj ov Bull. Br. Mus. nat. Hist. (Ent.) 49 (3): 153-234 Issued 30 August 1984 154 IAN J. KITCHING The moths of the family Noctuidae, sometimes known as 'owlets', constitute one of the threelargest families of Lepidoptera . Estimates of the number of known species vary considerably butthe most frequently quoted figure of 25,000 is still far in excess of such other groups as thebutterflies (15,000-20,000) and the land-living vertebrates (about 18,000). Not only are the noctuids a large family numerically, but they are also extremely importanteconomically. The larvae of many genera (e.g. Spodoptera, Heliothis, Euxoa, Earias andTrichoplusia) are well known as army- worms, cutworms, bollworms and stem-borers and causemany millions of pounds' worth of crop loss each year. Their control, being at present primarilychemical, is also expensive. A predictive classification can thus be seen to be an essentialprerequisite to the efficient use of the resources available to combat the pest species. However, constructing such a classification is no easy task. The sheer size of the group hasproved an effective deterrent (one in five lepidopteran genus-group names is noctuid; Nye,1975), and its worldwide although primarily tropical distribution has resulted in there being fewinstitutions where the work can be pursued on the required scale. Nevertheless, over the years,the genera of noctuids have been classified into tribes and subfamilies. Many of these taxonomicgroupings, particularly the earlier ones, were inadequately characterised and this, together withdifferences of opinion, extensive synonymies at all levels and a dependence upon superficialcharacters, has led to a far from satisfactory situation. This review is primarily historical and aims to trace and describe the development of thehigher (i.e. suprageneric) classification of the noctuids from its earliest beginnings to the presentday. The impact and use of the various authors' systems will be discussed, as will the charactersupon which they were based (where these were reported). Finally, a provisional cladisticanalysis is performed, using characters drawn from the literature. Each of the noctuid subfami-lies is considered in turn, and its relationships assessed and conclusions reached as to wherefuture research might be best directed. Several conventions will be used subsequently and these are listed below. 1. No authors' names are used (if required, these can be found in Nye, 1975, and Watson etal,1980). 2. The generic, tribal and subfamilial names used will be those employed in the works discussedat that point in the review. If subsequent changes have been made, the most recentlyaccepted name will be given in square brackets, immediately following the original, e.g.Gonoptera [Scoliopteryx]. Objective synonymies follow Nye (1975). Subjective synonymiesgenerally follow Franclemont & Todd (1983) for North American taxa, Kloet & Hincks(1972) for British taxa and the BMNH card index for all other groups. Some subjectivesynonymies (e.g. Heliothis/ Helicoverpa) may therefore not agree with the reader's opinions. 3. Where transfer of genera between subfamilies by other authors has been carried out, thename given in square brackets is the subfamily to which the genus was assigned by Nye(1975). Exceptions to this are: () the Agaristinae and Nolinae are treated here as subfamiliesof the Noctuidae; and (b) the Acronictinae (s.str.) is separated from the Amphipyrinae. In addition, the names Herminiinae, Hypenodinae and Rivulinae are not employed and itmust therefore be realised that many genera that correctly belong in such subfamilies are treatedas belonging to others (e.g. the Hypeninae s.l.). In the beginningEarly days Linnaeus (1758) divided the moths into two large genera, Sphinx and Phalaena. The latter wasfurther divided into seven 'subgenera': Bombyx, Noctua, Geometra, Tortrix, Pyralis, Tinea andAlucita. Forty species of true noctuids were contained in the section of Noctua described as'spirilingues, dorso cristate', together with the thyatirid, Phalaena Noctua [Achylia]flavicornis.However, due to Linnaeus' reliance upon superficial characters (e.g. degree of development of REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 155 the proboscis and the body vestiture), several noctuid species were placed in Bombyx (e.g. B.[Scoliopteryx] libatrix, B. [Colocasia] coryli and B. [Dicycla] oo). The family Noctuidae was probably first recognised as a distinct unit by Denis & Schiffermul-ler (1775). Their division 'Noctuae' was divided into 21 groups, based on characters drawn fromall developmental stages, including larval foodplants and habits. Each group was assigned aletter of the alphabet and accompanied by short diagnoses of the larvae and adults. Many of the groups were very heterogeneous and some contained species no longerconsidered to be Noctuidae. Most of the remainder combined representatives from severalcurrent subfamilies. The hypenines were included in the Pyralides as group A, 'PhalaenaePyralides Longipalpes'. Up to the nineteenth century, classifications were still largely based upon local faunas.However, Borkhausen (1792) considered all the European noctuid species and divided theminto two large groups, the Noctuae fasciatae and the Noctuae non fasciatae . These contained sixand 20 families respectively. Although characters were drawn from the morphology and habitsof several life-history stages, Borkhausen utilised only the wing patterns and resting positions ofthe adults and ignored all other structures. As a result, the classification was not particularlynatural. For example, one group consisted of those species with yellow hindwings, from suchgenera as Noctua, Anarta, Polyphaenis, Panemeria, Thalpophila and various quadrifines.Nevertheless, this arrangement, albeit imperfect, was an improvement. Hiibner (1805) divided the noctuids into three sections (the Bombycoides, Genuinae andSemigeometrae), while Haworth (1803-28) regarded the family as consisting of only threegenera: Noctua, Phytometra and Hemigeometra, subdivided into 22, three and two sectionsrespectively. This brief survey of the very early history of noctuid classification is by no means comprehen-sive. A more detailed review of this period (1720-1835) can be found in Guenee (1852-4:xlix-xc). Guenee's first classification In a paper of six parts, published between 1837 and 1841, Guenee presented the first detailedclassification of the noctuids, which were variously referred to as the family Nocturni, thedivision Noctuae, or, most frequently, the Noctuelides. As such, this arrangement will beconsidered in more detail than its predecessors. Unfortunately, the system 'evolved' during thepublication period, for reasons that were never stated, and this makes its description moredifficult. Initially, Guenee listed 18 tribes. In addition, he also recognised the Noctuo-Bombycidae (Table 1). This last tribe mainly contained species of the family Thyatiridae, butalso included three species of Cosmia [Amphipyrinae]. A fourth Cosmia species, C. trapezina,was excluded. Of the remaining 18 tribes, only the first six were dealt with in the first part of the paper(Guenee, 1837) and so these are the only tribes for which his original concepts are known. Although included in the 1837 list, Guenee discussed neither the Bombycoidi nor thePseudo-Bombycidi, but begin with the third tribe, the monobasic Bryophyagidi. This containedonly Bryophila [Cryphia], a genus that Guenee found difficult to place within the Noctuelides. The Nonagridi consisted of two genera, Nonagria and Gortyna, the larvae of which areroot- and stem-borers and, as such, show similar adaptations (smooth, supple integument;well-developed prothoracic and anal shields) which Guenee accepted as evidence of closerelationship. The Leucanidi comprised six genera, representing a mixture of noctuines, hadenines,acronictines and amphipyrines. They differed from the Nonagridi chiefly in that the larvae areleaf-feeders. The last tribe dealt with by Guenee in 1837 was the Noctuelidi, consisting of seven genera,which are mostly Noctuinae but also including a few hadenines. Guenee (1838a) continued his classification with the tribes Amphipyridi (four genera) and theMiselidi (three genera). The next tribe to be considered (Guenee, 18386) was the Hadenidi. Ten 156 IAN J. KITCHING Table 1 A comparison of the tribes employed by Guenee in his 'Essai' . The Roman numerals given inthe second column refer to the order in which the tribes were presented at the beginning of the work(Guenee, 1837), while the Arabic numbers given in the third column are those assigned to the tribesat the conclusion of the work (Guenee, 18416). TRIBE 1837 number 18416 number very diverse genera were included, from the Hadeninae, Cuculliinae, Amphipyrinae, Eute-liinae, Ophiderinae and Thyatiridae. Guenee recognised the heterogeneous nature of both thisand the next tribe (Orthosidi), accepting that they were almost completely artificial. Indeed, hefound difficulty in devising good characters to separate them and even went so far as to labelthem both Tribu IX'. The Orthosidi itself was described in 1839 and contained elements of allfour major trifine subfamilies. The second tribe to be considered in 1839, the Xylinidi, was composed primarily of cuculliinesbut also included several hadenines and a single heliothine (Chariclea [Periphanes] delphinii).This last species was regarded as a link between the Xylinidi and Orthosidi, and the Heliothidi,the tribe that contained the remaining heliothines, together with the hadenine genus, Anarta. The next tribe in Guenee's 1837 list was the Ctenoceridi. However, he must have reconsideredits validity, for its fails to appear in sequence (Guenee, 18410). Instead, the Heliothidi werefollowed by the Plusidi, a tribe corresponding to the current Plusiinae. In contrast, the Calpidi,absent from the 1837 list, was included and contained only a single species, Calpe [Calyptra]thalictri. This tribe was followed by the Ophiusidi, a mixture of catocalines and ophiderines (e.g.Lygephila and Minucia), and the Catocalidi, which consisted only of Catocala and theophiderine Catephia. The final part of the Essai (Guenee, 1841b) considered three tribes. For the first, Guenee usedthe name 'Phalaenoidi' rather than Noctuioidi (as used in Guenee, 1837). This was because hethought the sole included genus, Brephos [Archiearis] had much in common with the Pha-laenides [Geometridae] and formed a link between the two families. Guenee placed the genuswithin the Noctuelides because it was more similar to this family in all its life-history stages thanto either the Geometridae or the Pyralidae. The Acontidi and the Noctuo-Phalaenidi were similar in such characters as the form of thelarvae, antennae and the relative width of the wings. The latter tribe was seen as the natural linkbetween the Noctuelides and the Pyrales. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 157 The final classification of the noctuids proposed by Guenee (1841ft) (Table 1) contained anumber of changes in both the names and contents of the tribes compared to those employed inprevious parts of the paper. This was true even for that part with which the list appeared. Thecomposition of the Bombycoidi was now stated as Colocasia [Pantheinae], Acronicta and Moma[Acronictinae], while the Pseudo-Bombycidi had either been omitted or merged with anothertribe. The next few tribes had undergone extensive modification and rearrangement. The twogenera that formed the Nonagridi were separated, with Nonagria placed in the Leucanidi andGortyna included in a new tribe, the Apamidi. Other apamid genera included Glottula [Brithys],Xylophasia, Apamea (removed from the Leucanidi) and Mithymna [sic] [now containing only[Eriopygodes] imbecilla, not originally included in the genus]. . The Noctuelidi remained more or less intact, although it had gained Rusina from theLeucanidi. The Orthosidi gained Scoliopteryx from the Hadenidi, and Tethea [Zenobia] andCosmia from the Noctuo-Bombycidi, together with several other changes, while the Miselidiwere included within the Hadenidi. Otherwise this last tribe remained virtually unaltered, likemost of the others. The only significant alteration to the Ophiusidi and the Catocalidi was thetransfer of Catephia to the former tribe. Alternatives to Guenee Shortly afterwards, Duponchel (1844-6) constructed a classification of the Lepidoptera ofEurope. He recognised only three families: the Diurnes (butterflies), Crepusculares (mainlysphingids) and Nocturnes. The noctuids comprised 24 tribes within the last of these (Table 2),corresponding largely to those of Guenee (18416). There were, however, some notabledifferences. Firstly, Bryophila [Cryphia] was included within the Bombycoides. With regard to theNonagridi, Duponchel placed the genera Gortyna and Hydroeda [Hydraecia] in a separate Table 2 The tribes employed by Duponchel (1844-6) for the noctuid moths, within his lepidopterandivision, Nocturnes. Number Tribe XVI Bombycoides XVII Noctuo-Bombycides XVIII Orthosides XIX Gortynides XX Nonagrides XXI Leucanides XXII Caradrinides XXIII Apamides XXIV Hadenides XXV Noctuelides XXVI Amphipyrides XXVII Xylinides XXVIII Heliothides XXIX Calpides XXX Plusides XXXI Catocalides XXXII Ophiusides XXXIII Anthophilides XXXIV Agrophilides XXXV Anomalides XXXVI Phalenoides XXXVII Goniatides XXXVIII Acontides XXXIX Noctuo-Phalenides 158 IAN J. KITCHING tribe, the Gortynides, and Nonagria as the sole genus of the Nonagridi. The Apamides were thusequivalent to Guenee's 1841 Apamidi less Gortyna and Hydraecia, while the remaining generaof his 1841 Leucanidi were distributed between two tribes, the Leucanides and the Caradrinides. Duponchel's second innovation was the division of the Noctuo-Phalaenidi into five tribes. Thefirst, the Anthophilides, contained four acontiine genera while the Agrophilides included two.The latter tribe also included the amphipyrine Metaponia. A second amphipyrine genus(Metoptria [Synthymia]) was placed in the Goniatides, together with Euclidia, while two moreamphipyrines (Haemerosia and Erastria [Hapalotis]) , along with the acontiine Oratoscelis[Calymma] and the ophiderine Phytometra, constituted the Noctuo-Phalenides proper. Thefinal tribe, the Anomalides, included a single species, Timia [Axia] margarita, which is currentlyplaced in the Axiidae (Geometroidea). Thus, some of the heterogeneity of Guenee's Noctuo-Phalaenidi was removed by Duponchel, but not all. In the following year, Herrich-Schaffer (1845) proposed a series of groupings of the moths hecalled the Noctuidae. He excluded the species hitherto placed in the Noctuo-Bombycidi,including them in a separate family, the Cymatophoridae [Thyatiridae] . Nevertheless, theirposition at the head of the Noctuidae suggests that Herrich-Schaffer still considered them to beintermediate between this family and the preceding Bombyces. Two other families were associated with the Noctuidae. The first, the Nycteolidae, includedgenera presently placed in the Chloephorinae (Chloephora, Earias), Sarrothripinae (Nycteola),Ophiderinae (Rivuld) and Hypeninae (Schrankia [Hypenodes]) . The second family, the Brephi-dae, was equivalent to the Phalaenoidi. The Noctuidae were divided into 24 subfamilies (Table 3). Herrich-Schaffer followedDuponchel in placing Bryophila [Cryphia] in the Bombycoidae [mainly Acronictinae]. He alsoincluded Demas [Colocasia] coryli, which Duponchel put in the Liparides [Lymantriidae](although Guenee (18416) placed it in the Bombycoidi), and Diloba caeruleocephala, previously Table 3 Herrich-Schaffer's (1845) classification of the noctuid moths. Tribe 1: CymatophoridaeTribe 2: Noctuidae Subtribe 1: Bombycoidae 2: Orthosidae 3: Hadenidae 4: Xylinidae 5: Cucullidae 6: Gonopteridae 7: Cerastides 8: Amphipyridae 9: Noctuidae 10: Heliothidae 11: Maniidae 12: Eriopidae 13: Eurhipidae 14: Calpidae 15: Herminidae 16: Metoponidae 17: Plusidae 18: Erastridae 19: Ophiusidae 20: Acontidae 21: Goniatidae 22: Hypenidae 23: Leptosidae 24: AglenidaeTribe 3: NycteolidaeTribe 4: Brephidae REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 159 included in the Notodontides since Linnaeus (1758). Finally, Symira [Simyra] was removed fromits long-held relationship with the wainscots (Leucanid) (some of which Simyra superficiallyresembles) and placed nearAcronicta, its present position. The next two subfamilies of Herrich-Schaffer, the Orthosidi and the Hadenidi, containedmost of the remaining trifine noctuids less those placed in the Noctuidae, Xylinidae andCuculliidae. Mania [Mormo] maura was removed from the Amphipyridi and included withPlacodes [Eucarta] amethystina (from the Hadenidi) in the Maniidae. Mania [Naenia] typica wasplaced in the genus Neuria in the Hadenidae. The Noctuidae, Heliothidae, Acontidae, Amphipyridae (less Mania), Plusidae and Calpidaewere largely unchanged. Very small subfamilies were erected to accommodate Hoporina [Jodia]and Cerastis (Cerastides), Eriopus (Eriopides), Eurhipia [Eutelia] (Eurhipidae) and Gonoptera[Scoliopteryx] (Gonopteridae). New subfamilies of somewhat larger size were the Metoponi-dae, the Leptosidae and the Aglenidae. Erastria [Hapalotis] was removed from the Noctuo-Phalaenides of Duponchel and placed in a separate subfamily, the Erastridae. The Ophiusidae comprised Catephia and Ophiusa, but because this last genus also includedspecies currently assigned to such genera as Lygephila, the Ophiusidae was roughly equivalentto Guenee's (18416) Catocalidi and Ophiusidi combined. The most revolutionary aspect of Herrich-Schaffer's classification was his treatment of thedeltoids. These were split into two subfamilies, the Herminidae and the Hypenidae (this alsoincluding several acontiines). The decision to include the deltoid genera within the Noctuidaewas well ahead of its time and was only slowly to gain acceptance. Guenee's second classification The next major classification of the noctuids was that of Guenee (1852-4). The influence ofDuponchel and Herrich-Schaffer was evident, for the arrangement proposed (Table 4) incor-porated many of their groupings. This classification was to exert a significant influence on theideas of future workers and thus is discussed in detail. Two major divisions of noctuids were recognised, the Noctuelites and the Deltoides. Gueneeremoved this latter group from the Pyrales because he considered them to have more in commonwith the Noctuelites. However, despite the recognition that the two groups graded almost Table 4 Guenee's second classification (1852-4) of the Noctuidae. Two 'divisions' were recognised,of which the former was split into two 'phalanges'. Each phalange was further divided into sectionstermed 'tribus', which are not equivalent concepts to the present category of tribe. DIVISION: NOCTUELITES PHALANGE: TRIFIDAETRIBU: 1 Bombyciformes Family: 1 Noctuo-Bombycidae 2 Bryophilidae 3 BombycoidaeTRIBU: 2 Genuinae Family: 1 Leucanidae 2 Glottulidae 3 ApamidaeSubfamily: 1 Gortynides 2 Xylophasides 3 Episemides 4 Apamides propres 4 Caradrinidae 5 Noctuidae 6 Orthosidae 7 Cosmidae 8 Hadenidae 9 Xylinidae10 Heliothidae 160 IAN J. KITCHING Table 4 - cont. TRIBU: 3 Minores Family: 1 Haemerosidae 2 Acontidae 3 Erastridae 4 Anthophilidae 5 PhalaenoidaePHALANGE: QUADRIFIDAE TRIBU: 1 Sericiae Family: 1 Palindidae2 DyopsidaeTRIBU: 2 Variegatae Family: 1 Eriopidae 2 Eurhipidae 3 Placodidae 4 Plusidae 5 Calpidae 6 Hemiceridae 7 Hyblaeidae 8 GonopteridaeTRIBU: 3 Intrusae Family: 1 Amphipyridae 2 Toxocampidae 3 StilbidaeTRIBU:4Extensae Family: 1 Polydesmidae 2 Homopteridae 3 HypogrammidaeSubfamily: 1 Yridae 2 Hypogrammidae propreTRIBU: 5 Limbatae Family: 1 Catephidae 2 Bolinidae 3 Hypocalidae 4 Catocalidae 5 OphideridaeSubfamily: 1 Ophideridae propre 2 PhyllodidaeTRIBU: 6 Patulae Family: 1 Erebidae 2 Ommatophoridae 3 Hypopyridae 4 BendidaeSubfamily: 1 Hulodides 2 Bendides propresTRIBU: 7 SerpentiaeFamily: 1 Ophiusidae 2 Euclididae 3 Poaphilidae 4 RemigidaeTRIBU: 8 Pseudo-Deltoidae Family: 1 Focillidae 2 Amphigonidae 3 ThermesidaeDIVISION: DELTOIDES Family: 1 Platydidae 2 Hypenidae 3 Herminidae REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 161 imperceptibly into one another, Guenee did not regard this to be sufficient grounds for unitingthem into a single division. Thus a graded sequence was identified, passing from the Bombyci-formes (through which the Noctuelites were connected to the Bombyces) through to thePseudo-Delto'idae (by which the Noctuelites passed into the Deltoides and thence into thepyralids and geometrids). Noctuelites. Within this division, Guenee recognised two large groups, which he termed'phalanges', the Trifidae and the Quadrifidae. These were primarily separated on the now-infamous character of the degree of development and position of vein M 2 of the hindwing. In theTrifidae, this vein is generally much weaker than those following it, and originates at such aposition that crossveins mi~m 2 and m 2 -m 3 are approximately equal in length. This gives theveins forming the posterior part of the discal cell a trifid appearance. In the Quadrifidae, M 2 is astrong vein originating near the posterior angle of the cell, resulting in the quadrifid configura-tion. Trifidae. The Trifidae were divided into three tribes: the Bombyciformes, the Genuinae andthe Minores. Distinctions were based largely upon the superficial appearance of the adults andthe number of prolegs and hairiness of the larvae. It should be noted that Guenee's concept of a'tribe' (or tribu), to which the family was subordinate, is not equivalent to current usage, inwhich the tribe is subordinate to the family. Bombyciformes. The first tribe was split into three families. The Noctuo-Bombycidae[Thyatiridae] were envisaged as being intermediate between the Noctuelites and the bombycidfamily, Notodontides, while the Bombycoidae [Acronictinae, part] formed a link with theLiparidae [Lymantriidae]. Guenee considered substituting the name Acronyctidae for Bomby-coidae but felt that to do so would obscure their relationship with the Bombyces, even thoughthe term Acronyctidae would result in a more uniform nomenclature. The third family, theBryophilidae (the Bryophagidae of Guenee, 1837), did, however, have its name changed forexactly this reason. An advance upon the 1841 list, with regard to the Bombyciformes, was the unification of thepreviously widely scattered elements of the Thyatiridae (= Noctuo-Bombycidae): Thyatira,Leptina and Cymatophora [Achlya]. Guenee also followed Herrich-Schaffer (1845) by includingSimyra in the Bombycoidae. Genuinae. The Leucanidae, the first 'genuine' family, comprised those species collectivelyknown as 'wainscots'. The adults are pale-coloured, with a simple striate pattern, and the larvaeare either endophagous, boring in the stems and roots of grasses, sedges and other monocotyle-donous plants, or living concealed by day and feeding by night on Poaceae. The larval habitsformed the basis of two subdivisions. At present, the borers (Nonagrides) are placed in theAmphipyrinae, while the rest (Leucanides) are put in the Hadeninae. The larvae of the Glottulidae also feed internally, but in bulbous plants. The family containedonly four genera (e.g. Glottula [Brithys]} representing a heterogeneous assemblage of hade-nines, amphipyrines and ophiderines. The larvae of the next three families generally conceal themselves by day, emerging only afterdark to feed, although a few (Gortyna spp.) are root-borers. The first, the Apamidae, sharedcharacters of both the Noctuidae and the Hadenidae. Guenee recognised that the Apamidaewere heterogeneous and possibly subject to future modification and division. In order toindicate this, he separated the species into four subfamilies: the Gortynides, the Xylinides, theEpisemides and the 'Apamides propres'. In his 1841 classification, Guenee united the next family, the Caradrinidae, with theLeucanidae because the latter included the genus Simyra, which he considered to be intermedi-ate. However, in the present work, Guenee (1852-4) had removed Simyra to the Bombycoidaeand this allowed him to reinstate the Caradrinidae as a separate family. The fifth family, the Noctuidae, generally conformed to the current concept of the Noctuinae. 162 IAN J. KITCHING It consisted chiefly of three large genera: Agrotis (in the broadest sense), Triphaena [Noctua]and Noctua [Amathes, Graphiphora, Diarsia] . Members of the Noctuidae were characterised byholding their wings flat over the back at rest, one wing slightly overlapping the other, whichdistinguished them from the Leucanidae and Caradrinidae, in which the wings are heldroof- wise. Guenee found the Orthosidae very difficult to differentiate from both the Noctuidae and theHadenidae. Biologically, many of the Orthosidae share the feature of autumn or early springappearance as adults. The family consisted of a mixture of hadenines and cuculliines. The Cosmidae was erected to accommodate a group of genera (e.g. Dicycla, Cosmia) that, inhis Essai, Guenee had included in the Orthosidi and the Noctuo-Bombycidi, but that were nowconsidered to be distantly related. Like the Orthosidae, the Hadenidae proved very difficult to characterise. In larval features,the family approached the Apamidae and Xylinidae, whereas the adults resembled theOrthosidae. All the included genera are currently placed in either the Hadeninae or Cuculliinae,except for Phlogophora, which is an amphipyrine. In contrast to the preceding families of the Genuinae, the larvae of the remaining two feedexposed on plants (with rare exceptions). The Xylinidae contained Cucullia, Calocampa[Xylena], Xylina [Lithophane] and their relatives, with several smaller amphipyrine genera andone, Ny stalea, now placed in the Notodontidae. The Heliothidae was characterised additionallyby diurnally-active adults. The only alteration from the Heliothidi of the Essai (Guenee, 1839)was the inclusion of Chariclea [Periphanes], a genus previously placed in the Xylinidi. Minores. The last of the trifid tribes, the Minores, was divided into five families. All werecharacterised by the small size of the adult insects, which often resembled geometrids, pyralidsor tortricids, similarities that were not contradicted, in Guenee's opinion, by the early stages.The Minores comprised a very mixed group of families, subdivided primarily on the basis ofwhether the adult abdomen was slender or stout. Of the latter type, the Haemerosidae consisted of only two genera, Haemerosia (now placedin the Amphipyrinae) and Lepidomys (currently in the pyralid subfamily, Chrysauginae). Theother family with stout bodies, the Acontidae, was relatively larger and contained seven genera. The first of the slender-bodied families, the Erastridae, contained only four genera, mostlyacontiines but with at least one amphipyrine (Erastria [Elaphria] venustula). Most species of theAnthophilidae could be distinguished from the Erastridae by the lack of a forewing areole.Those possessing this structure could only be separated with difficulty. Eleven genera wereincluded within the Anthophilidae. The final family of the Minores was the Phalaenoidae [Archiearinae], which Gueneecontinued to retain in the Trifidae. Quadrifidae. The second 'phalange' of noctuids recognised by Guenee was the Quadrifidae.Although this group of tribes was 'notablement differente' from the Trifidae, the separation wasnot absolute. Guenee found that placing the quadrifid families into a linear sequence was muchmore difficult than for the Trifidae. Although he attempted to delimit the families as naturally aspossible, problems still arose, particularly in those groups that seemed to have closer affinitieswith certain trifid genera, for example, the Bolinidae and the Acontidae, and the Eurhipidae[Euteliinae] and certain hadenids. In contrast, several families were particularly well defined,notably the Catocalidae and the Plusidae. The Quadrifidae was divided into eight tribes on the basis of the wing venation and pattern,general facies and the labial palps. Sericiae. This tribe consisted of two small, tropical families, the Palindidae (Palindia[Eulepidotis] and Homodes) and the Dyopsidae (Dyops and Dyomyx). Both are currentlyplaced in the Ophiderinae. Variegatae. The second tribe was the largest in the Quadrifidae, comprising eight families. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 163 The first of these, the Eriopidae, was small, containing only five genera. On the basis of larvalcharacters, Guenee allied it closely with the next family, the Eurhipidae. However, he found itimpossible to synonymise the two on the grounds of differences between the adults. TheEurhipidae itself formed a well-defined family, corresponding to the present Euteliinae. Another family in which the constituent genera had, like those of the Eurhipidae, been placedin the Hadenidi was the Placodidae. Only two genera were included: Placodes [Eucarta;Amphipyrinae] and Diastema [ Acontiinae] . The largest family of the Variegatae was the Plusidae. In addition to the two genera currentlyassigned to the subfamily Plusiinae (Abrostola and Plusia s.l.), three other genera were alsoincluded, Basilodes, Thyria and Plusiodonta. Through this last genus, Guenee noted that thePlusidae graded into the Calpidae. This family had been extended to include not only Calpe[Calyptra] but also Oraesia, Gonodonta and Hapigia. The last of these was considered to have ahepialid or notodontid aspect, but Guenee placed it in the Calpidae because of the developmentof the proboscis and the palps. However, present classifications place Hapigia in the Notodonti-dae and so it is not surprising that Guenee found that this genus formed a convenient linkbetween the Calpidae and the Hemiceridae, another family containing current notodontids(Hemiceras and Canodia). Guenee commented on the resemblance between the larvae ofHemiceras and Cerura (Notodontidae) and concluded that despite the superficial similarity ofthe adult insects to noctuids, the genera of the Hemiceridae, and Hemiceras in particular, couldbelong elsewhere. The transfer of Hemiceras and Hapigia to the Notodontidae was effected byDruce (1887), while Canodia was moved by Schaus (1901). Of the three remaining hemiceridgenera, two (Arcyophora and Plusiodes [Westermannia]) are presently assigned to theChloephorinae while the third, Achantodes, (a genus Guenee described as having the overallappearance of a large species of Chilo) presently resides in the Pyralidae: Glaphyriinae. The seventh family of the Variegatae was a group of moths whose relationships are stilluncertain today. Until Guenee, the Hyblaeidae had been classified in such families as theCalpidae, Ophideridae or Gonopteridae. However, Guenee's general conclusion was that thegroup was of uncertain position within the Noctuelites. In present-day classifications, the twoconstituent genera are widely separated with Phy codes placed in the Glyphipterigidae. Thedevelopment of the current location of Hyblaea will be discussed in greater detail below. The last family in the Variegatae contained seven genera (e.g. Gonoptera [Scoliopteryx] ,characterised by angled wings. Intrusae. The third quadrifid tribe Guenee termed the Intrusae. It was divided into threefamilies on the basis of the general facies of the larvae and adults. The first of these, theAmphipyridae, contained four genera, the majority now being placed in the Trifinae (mainlyAmphipyrinae), although Barydia has been transferred to the Notodontidae (Nye, 1975). The Toxocampidae showed great similarity to the last family (in Guenee's opinion) and like it,was divided into two groups based upon the larvae. Five genera were included, all of which arecurrently assigned to the Ophiderinae. The final family, the Stilbidae, was composed of a singlespecies, Stilbia hybridata [anomala]. Consideration of all characters led Guenee to concludethat, like Brephos [Archiearis], Stilbia was totally isolated within the Noctuelites. At present,Stilbia is an amphipyrine. Extensae. Three families were recognised within this tribe on the basis of wing shape and sizeof the legs. The first family, the Polydesmidae, comprised three genera (Pantydia, Polydesma andDiatenes) which had marked affinities with certain genera of the Intrusae, from which they couldbe distinguished by the wing venation. The Polydesmidae were seen as a link between theAmphipyridae and the Homopteridae. This latter family was established by Boisduval (1840)for several genera, the larvae of which resemble those of the Catocalidae. The Hypogrammidae was considered to be somewhat heterogeneous and Guenee thoughtthat it might eventually have to be split. In order to draw attention to this, he established twosubfamilies, the Yridae and the 'Hypogrammidae propre'. 164 IAN J. KITCHING Limbatae. This tribe was another group Guenee considered might have to be split. It consistedof five families, of which the Catocalidae and the Ophiusidae were thought to perhaps warrant aseparate tribe. They were believed to be related, through several genera, to the Patulae, andthrough the Catephidae to the Extensae. The Catephidae and the Bolinidae were considered to have much in common. Seven generacomprised the former. Three (Stictoptera, Odontodes and Lophoptera) are currently assigned tothe Stictopterinae, while of the other four, three are ophiderines, and one (Cocytodes) is acatocaline. There were only four genera in the Bolinidae: Leucanitis, Panula, Bolina [Aleucani-tis] and Syneda [Drasteria], of which the first is presently a catocaline, the other threeophiderines. The genus Hypocala was the sole component of the Hypocalidae and superficially resembledthe hyblaeids. Guenee was uncertain as to its relationships, but on the basis of the form of thepalps suggested a link with the Bolinidae. Two genera, Parthenos [Euparthenos] and Catocala, were placed in the Catocalidae. Similar-ities in wing pattern suggested a relationship with the Hypocalidae and the Ophideridae. Thelatter family contained some species that showed similarities to the Calpidae and others thattended towards the Erebidae (Patulae) and the Ophiusidae (Serpentiae). Two subfamilies wererecognised, the 'Ophideridae propre' and the Phyllodidae. Patulae. All quadrifids that, despite their large size, presented an essentially phalaeniform (=geometriform) aspect, were placed in this tribe. Four families were recognised. The Erebidaewas a large family of many genera that was difficult to define using absolute structural characters,although Guenee felt that the overall facies was unmistakeable. The Ommatophoridae wassimilarly well defined in general appearance due to the large eyespot on the forewings of mostspecies. Guenee considered this family to be very natural, despite the aberrant wing venation ofgenera such as Cyligramma (the only quadrifid genus to lack an areole) and Argiva [Erebus,part] (in which the hindwing venation is much reduced in the males). Like the last family, the Hypopyridae was considered to form a link between the Erebidae andthe Ophiusidae (Serpentiae). Within the Bendidae, Guenee recognised two distinct subfamilies. The Hulodides comprisedspecies of the genera Homaea and Hulodes, which are South East Asian and resemble certainspecies of Hypopyra, while the 'Bendides propres' , inhabiting the Americas, included Itonia andBendis [Lesmone]. Serpentiae. The seventh tribe of the Quadrifidae principally contained species previouslyrecognised under the name Ophiusa, a genus that until then had been rather vaguely defined.The tribe was divided into four families, an arrangement Guenee considered provisional at best,because he knew so few of the larvae. By far the largest family in the tribe (and also in the Quadrifidae) was the Ophiusi-dae, comprising those species with elongate larval prolegs and large, impressive adults,with velvet-like, apically-pointed forewings. It included a mixture of catocalines and ophi-derines. The sole European representative of the Euclididae (Eudidia [Euclidia + Callistege] was onceplaced in the Noctuo-Phalaenidi. Duponchel (1844-6) had separated it into the Goniatides butGuenee rejected that name on the grounds that it was not based upon an included genus (but seehis own Bombycoidae). The six included genera were distinguished from other groups primarilyby a reduced number of larval prolegs (three pairs). The larvae of the Poaphilidae, in contrast, were characterised by four pairs of prolegs,although the adults were similar to those of the Euclididae. Seven genera were included, ofwhich four are currently catocalines and three ophiderines. Distinguished primarily by the form of the hindleg in the males, which is clothed with densehair arranged in a single compressed line on each surface, the Remigidae included genera thatare presently placed in both the Catocalinae and Ophiderinae. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 165 Pseudo-Deltoidae. The last tribe in the Noctuelites graded almost imperceptibly into thedivision Deltoi'des, although the two groups could apparently be distinguished using certaincharacters of the head. Three families were recognised. The Thermesidae were the largest and that which approached most closely in general faciesthe Delto'ides. Certain genera also showed some resemblance to genera of the Remigidae,Ophiusidae and Bolinidae. The other two families were characterised by features of the palps and body vestiture. TheFocillidae was recognisable by the form of the last palpal segment and by the somewhat angledwings. It contained four genera of ophiderines. The three genera of the second family, theAmphigonidae, bore a superficial resemblance to Gonoptera [Scoliopteryx] libatrix and are nowplaced in the Ophiderinae. Deltoi'des. The second great division of noctuid moths was termed the Delto'ides. Gueneeconsidered the distinction between the last listed Noctuelites (e.g. Palyna, now a hypenine) andthe first deltoids to be very fine but the two groups could be distinguished. As was discussedabove, such was not the opinion of Herrich-Schaffer (1845) but Guenee could not countenancesuch a decision. The Delto'ides were divided into three families that were not assigned to tribes. ThePlatydidae comprised only three genera: Trigonia [Claterna], Macrodes and Platydia [Yidalpta].As if to confirm Guenee's opinion regarding the fine distinction between the Noctuelites and theDelto'ides, these genera are currently classified in the Ophiderinae. He believed that thePlatydidae were the most noctuid-like of the deltoids, although on the basis of palp, antenna,body, leg and wing characters they were unequivocally deltoids. The Hypenidae were considered by Guenee to be intermediate between the other twofamilies and to epitomise the concept of the Delto'ides. Most of the Hypenidae are currently inthe Hypeninae but Rhodina, Madopa [Colobochyla] and Pterhemia are ophiderines. The Herminidae was the largest and most varied family of deltoids, and that which Gueneethought most closely approached the Pyrales. He also considered it to be the most interestingfamily because of the great diversity of structure found within it. Many genera have specialisedstructures on the antennae or eversible hair-pencils on the legs of the males, which have led onegroup to be termed the 'fan-foots'. Most of the herminid genera are presently classed ashypenines but two (Cyclopteryx, Rivula) are placed in the Ophiderinae. This then was the second arrangement of the noctuids proposed by Guenee. It was followedby Stainton (1857), as well as by Walker (1856-8) in his influential, if controversial, list ofspecimens in the British Museum. Not all workers, however, accepted Guenee's classification soreadily, and criticism was not long in coming. The American alternativeThe opening move Packard (1869) described many morphological characters of the family Noctuidae that heconsidered to be of use in classification , drawn mostly from the head, thorax and wings. It was byfar the most detailed comparative study conducted up to that time. Packard divided theNoctuidae into two subgroups, approximating Guenee's Trifidae and Quadrifidae. However, hecriticised Guenee's use of venational characters and condemned the divisions of the Trifidae inparticular, calling them 'trivial groups of genera'. Packard preferred characters such as thelength and narrowness of the clypeus and the form of the antennae to those from the venation orpalps. At least he felt this was true for the temperate noctuid fauna, but expressed no opinionregarding those from other areas. Consequently, Packard removed both the Sericiae and Variegatae (less then Gonopteridae)to the Trifidae. This combined group he termed the Noctuinae , while the remainder of Guenee'squadrifids became the Catocalinae. 166 IAN J. KITCHING Development of Crete's classifications (1874-1890) 1874. In his list of North American noctuids, Grote (1874) employed Packard's two subfami-lies. However, Grote also included the deltoids within the Noctuae and also associated two othergroups with the family: the Noctuo-Phalaenidae [Archiearinae] and the Bombyciae [Thyatiri-dae]. He also agreed with Packard's opinions concerning Guenee's tribal and family groups andemployed no subdivisions in the Noctuae. These criticisms had little or no effect on English authors (e.g. Butler, 1881; Meldola, 1881;Moore, 1881; Pryer, 1883-5), who continued to follow Walker (1856-8) and hence remainedfaithful to Guenee. 1882. A brief resume of the classification of the noctuids was provided by Grote (1882a). Hedisagreed with Lederer (1857), who had removed the Cymatophoriden [Thyatiridae] andBrephiden [Archiearinae] from the Noctuidae and established them as separate families, andchose to include the Cymatophorina [Thyatiridae] and Brephina [Archiearinae] within theNoctuae. This was followed by the New check list (Grote, 1882ft), in which the noctuids weresubdivided although no reasoning was given. Two major groups were recognised in the Noctuae: the Bombyciae [Thyatiridae] and theNoctuelitae (Table 5). Within the latter, two subgroups were recognised, the Nonfasciatae and the Fasciatae Table 5 The classification of the Noctuidae employed by Grote (18826) . NOCTUAE BombyciaeNoctuelitae (Non-Fasciatae) Dicopinae Bombycoidea Noctuinae Hadeninae Nonagriinae Pyrophilinae Taeniocampinae Orthosiinae Cuculliinae Nolaphaninae Anomiinae Litoprosopinae Euteliinae Ingurinae Calpinae Plusiinae Stiriinae Heliothinae Eustrotiinae Hyblaeinae Noctuo-Phalaenidi(Fasciatae) Catocalinae Ophiderinae Toxocampinae Erebiinae Brotiinae Pangraptinae Deltoides Hypeninae REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 167 (Packard's Noctuinae and Catocalinae), which were further divided into 21 and eight subfami-lies respectively. Many of the groups employed by Grote corresponded to Guenee's familiesbut, because the New check list was the first to subdivide the American noctuid fauna, Grotefound it necessary to erect a number of new subfamilies, which can be summarised as follows: Dicopinae: Eutolype, Dicopis [Psaphida, part], Copipanolis [Cuculliinae];Pyrophilinae: Pyrophila [Amphipyra], Caradrina [e.g. Athetis] [Amphipyrinae];Nolaphaninae: Adipsophanes [Catabena], Crambodes, Nolaphana [Balsa] [Amphipyrinae];Anomiinae: Anomis, Aletia [Anomis], Pteraetholix [Amyna], Chytoryza [Illatia] [the first two are ophiderines; the third an acontiine; and the last an amphipyrine] ; Litoprosopinae: Litoprosopus [Ophiderinae];Ingurinae: Ingura [Paectes] [Euteliinae]; Stiriinae: Stilbadium, Stiria, Acopa, Fala, Plagiomimicus [Amphipyrinae];Brotiinae: Brotis [Sphacelodes] vulneraria [Geometridae: Ennomiinae];Pangraptinae: Syllectra, Pangrapta, Phalaenostola [the first two are nowophiderines, the last a hypenine]. In addition, several genera were moved between subfamilies. For example, the Plusiinae nowincluded Anarta [Anarta, Sympistis, Hada], Telesilla [Eucarta] and Lepipolys. The deltoidgenera were divided into two groups. The first, the Deltoides, contained many hypenine generaand some ophiderines (e.g. Rivula) while the second, the Hypeninae, consisted of Hypena andits relatives. The deltoids were classified within the Noctuelitae but, unlike Herrich-Schaffer(1845), Grote kept the two subgroups together at the end of the list. In the same year, Smith (1882-3) published a synopsis in which he explicitly stated thecharacters used at each level of subdivision of the Noctuidae. He recognised three primarydivisions characterised by the eyes being naked, hairy or lashed. Finer divisions were based upontibial armature, clypeal modifications and the palps. However, Smith did not attach any formalnames to his groups. Indeed, he considered these divisions [to be] entirely artificial and thesequence of genera in the synopsis is not that to be followed in the systematic arrangement of thegroup, the object being only to enable the collector to place any unknown Noctuid into its propergenus with but little trouble. 1883. The next year, Grote (1883) considered in more detail the structural characters he hadused in his previous works. He reverted to Lederer's (1857) three families, in that theNoctuo-Phalaenidae were elevated to family status (as Brephidae). Also, the Bombyciae ofGrote (18826) was renamed the Thyatiridae (Cymatophoridae being inapplicable for reasonsgiven in Harvey, 1874). The Noctuidae themselves were divided as before but only the nonfasciate subfamilies werediscussed in detail. Most were unaltered from the New check list but several changes had beenmade. For consistency, the Bombycoidea were renamed the Apatelinae, while the Euteliinaeand Pyrophilinae became the Eurhipinae and Caradrinae respectively. The Bryophilinae(Bryophila [Cryphia] and its relatives) were split from the Apatelinae, while four genera(Trichotarache [Acontia, part], Tar ache [Acontia, part], Chamyris [Cerma] and Xanthodes[Bagisara]) were removed from the Eustrotiinae as the Acontiinae. Two new subfamilies were erected. The Arzaminae, comprising Sphida and Arzama [bothBellura] (previously placed in the Nonagriinae), and the Scolecocampinae, containing Scoleco-campa, Eucalyptera [Gabara, part], Cilia [Gabara, part], Amolita, Dory odes and Phiprosopus[Phyprosopus]. 1886. Meyrick (1886) was the first English author to adopt the characters advocated by Grote.However, Meyrick also employed Guenee's major divisions (based upon the development ofhindwing vein A/ 2 ), calling them the Noctuidae and Plusiadae. No further subordinate groupswere used. 168 IAN J. KITCHING 1890. A revised checklist of North American noctuids was published by Grote in 1890. Twosubfamilies were considered, the Thyatirinae and the Noctuinae (relegated from the rank offamily). The latter was divided into 25 tribes. Most of these corresponded to those of Grote'sprevious works but there were discrepancies. Arsilonche [Simyra] and Raphia were separated from the Apatelinae and, together withDemas [Colocasia] , comprised the Bombycoidi. Not only was this name potentially confusing (cfthe Bombycoidae of Guenee, 1852-4) but it was not based upon the name of an included genusand therefore ran counter to Grote's (1883) previous assertions regarding the correct formationof suprageneric names. In addition, the Noctuinae were renamed the Agrotini, and the Eurhipidae reverted to theEuteliini, with the Ingurinae being included within it. New tribes were formed for Lithophane,Calocampa [Xylena], Lithomia [Lithomoia] and Xylomiges [Egira] (Calocampini); andCleophana [Copicucullia] (Cleophanini). The Tarachinae lost Trichotarache [Acontia, part] tothe Heliothini, while interposed between the former tribe and the Eustrotiinae was the newmonobasic Cerathosiini (Cerathosia tricolor). This species had been described as an arctiid bySmith, near Utetheisa, but on the basis of wing venation and clypeal ornamentation, Grote feltcertain that it was actually a noctuid. Cerathosia currently resides in the Acontiinae. Grote (1890) gave only an outline of his classification of the rest of the family. More detailswere given in another paper (Grote, 1889-90). The Catocalinae were divided into two tribes.The Catocalini were characterised by their often brightly-coloured hind-wings and included theNew check list subfamilies Catocalinae, Ophiderinae and Toxocampinae. The second tribe, thePheocymini, comprised the Erebinae of the New check list. The wings of the included generawere generally concolorous and crossed by sinuous lines. Grote was unwilling to assign theremaining catocaline subfamilies of the New check list, the Brotiinae and the Pangraptinae,preferring to await a more thorough study of the Neotropical fauna. Two tribes were recognised in the Deltoidinae. The herminiines could be distinguished bytheir concolorous wings, marked with continuous lines. The Hypenini, in contrast, hadunicolorous hindwings while the forewings displayed the usual noctuid markings, albeitsubdued. The fourth subfamily of the Noctuidae, the Brephinae, now contained two genera, Brephos[Archiearis] and Leucobrephos . Grote recognised certain resemblances to some Geometridae(hairy abdomen, predominantly diurnal flight period, larvae with five pairs of prolegs [!] andsimilar habits) but still retained them in the Noctuidae. Thus, by 1890, there were two competing, if somewhat internally unstable and inconsistent,systems for the classification of the Noctuidae. The first, based upon the work of Guenee, waswidely accepted in Europe; the second, arising from the studies of Grote, was the primarysystem in use in North America. The stage was set for a comparative assessment. Classifications in conflictTutt's comparison After consideration of the available options, Tutt (1891-2) decided to adopt the system ofGuenee, not because he thought it was more correct but because it was based upon characters ofthe early stages as well as the adult insect. Grote's system relied entirely on adult features. Onlythe points of contention as seen by Tutt will be discussed further. Trifidae: Bombyciformes. Tutt considered as irrelevant the debate as to whether theCymatophoridae [Thyatiridae] were to be placed as the last tribe in the Bombyces (Staudinger &Wocke, 1871) or as the first in the Noctuae (Guenee, Grote). He did explicitly note, however,that the eggs were of the geometrid type, and that the removal of the Cymatophoridae from theNoctuae would render the latter group more homogeneous. Tutt concluded most authors to be in agreement over retaining the Bombycoidae (=Apatelini) and Bryophilidae in the Noctuae although Butler (1879) had dispersed the British REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 169 species of Acronycta [Acronicta] amongst the Arctiidae, Liparidae [Lymantriidae], Notodonti-dae and Noctuae (a position he later rescinded; Butler, 1893). Tutt also agreed with Staudinger& Wocke's (1871) inclusion ofMoma orion [alpinum] in the Bombycoidae but expressed doubtsregarding Demas [Colocasia] coryli and Diloba caeruleocephala, which he thought might belongelsewhere. He also rejected Grote's Bombycoidi on the grounds that Chapman (18930) haddemonstrated Arsilonche [Simyra] albovenosa [venosa} to be congeneric with Acronycta[Acronicta] rumicis and so Arsilonche henrici (the American species) could not therefore beplaced in a separate tribe to the genus Acronycta [Acronicta]. Trifidae: Genuinae. Tutt considered this group to be far more natural, although he hadreservations concerning the arrangements and contents of some of the families. He felt that, onthe basis of larval habits and mode of pupation, Staudinger & Wocke's (1871) placing of theGonopteridae in the Trifidae was correct. Tutt also believed that Mania [Mormo + Naenia] andAmphipyra might be better placed in the Trifidae, but that the Plusidae should remain in theQuadrifidae. This last point was the major difference between Guenee's Genuinae and Grote'sNoctuinae. Trifidae: Minores. Tutt acknowledged the anomalous nature of the Phalaenoidae (Brephi-dae) but retained it in the Trifidae. He considered Meyrick's (1892) transfer of this group to theGeometrina (as Monocteniadae) erroneous because the larvae have five pairs of prolegs (eventhough those on abdominal segments 3, 4 and 5 are reduced). Quadrifidae: Limbatae. Tutt disagreed with Grote's suggestion that the Ophiderinae be-longed in the Catocalini, considering that association to be based upon superficial characters. Healso believed that the Toxocampinae, allied with the Catocalinae by Grote (1890), were nearerthe Deltoides, a position about which 'there can be no doubt' when the larvae are examined. Quadrifidae: Intrusae. This was the subclass about which Tutt had most doubts, especiallywhen just the British representatives of the three families (Amphipyridae, Stilbidae andToxocampidae) were examined. He did not believe that Mania [Mormo] and Naenia belongedin the Hadenidae (where they had been placed by Staudinger & Wocke, 1871) although he wasreasonably satisfied with the position of Amphipyra in the Caradrinidae. Tutt also refused toaccept Aventia [Laspeyria] flexuosa and Boletobia [Parascotia] fuliginaria ('an admitted geo-meter' - now an ophiderine) in the Noctuae, and thought the Stilbidae were sufficiently isolatedto require special treatment, similar to the Brephides. Deltoides. Tutt agreed with those who placed this group as an integral part of the Noctuae, andconcurred with Grote's division of the subclass into the Herminiidae and Hypenidae, groupsthat Tutt considered to be very natural. Overall, Tutt's general conclusions supported Guenee's classification, but with the followingsuggestions. 1. Demas [Colocasia] and Diloba were not Noctuae. 2. The Bryophilidae had no close relationship with the Bombycoidae. 3. The Leucanidae was unnatural, Leucania belonged in the Noctuidae and the Nonagriae inthe Apamidae. 4. The Hadenidae and Apamidae were essentially identical. 5. The Xylinidae should be divided, because Xylina [Lithophane] and Calocampa ]Xylena]were not closely related. 6. The Amphipyridae should be divided, with Mania [Mormo] and Naenia being placed in theNoctuidae and Amphipyra near the Caradrinidae. 7. The Plusidae were less closely related to the Xylinidae than were the Heliothidae. 8. The Toxocampidae were closer to the Deltoides than to the Catocalidae. 9. The Deltoides were unequivocally noctuids. 170 IAN J. KITCHING Smith's comparison Smith (1891) held altogether different views. The Noctuina included three families: theThyatiridae, the Noctuidae and the Brephidae. No subdivision was used because Smithconsidered Grote's subfamilies to be of no use because of their 'unequal value and impossibilityof accurate definition'. Guenee's classification was not even considered. Smith's position hadnot altered two years later (Smith, 1893). Hampson's Fauna of British India The next major work on the higher classification of the Noctuidae introduced a new name butone which was to have a considerable impact in future years. The first arrangement of noctuidgenera proposed by Hampson (1893-5) pioneered an entirely novel approach to family levelinterrelationships. Hampson believed that the three families Agaristidae, Arctiidae and Noc-tuidae were very closely related and difficult to separate clearly in some instances. In particular,he thought that the 'primitive' forms of the last two families (the Nolinae and Sarrothripinaerespectively) graded into each other somewhat. The Nolinae were later considered to be worthyof a separate family, based upon larval characters (Packard, 1895). Also included within the Arctiidae, as the subfamily Nycteolinae, were nine genera currentlyreferable to the noctuid subfamily Chloephorinae. This group of genera had always beenenigmatic and had generally been placed in the Bombyces as a separate family (e.g. Smith,1891). Hampson (1893-5) noted that they approached the Acontiinae in the structure of thehindwing vein Sc + RI but on the basis of their 'tree-frequenting habits', retained them in theArctiidae. Table 6 The first classification of the Noctuidae proposed by Hampson (1893-5). NOCTUIDAE TrifinaeAcontiinae Palindidae Sarrothripinae Euteliinae Stictopterinae Gonopterinae Quadrifinae Focillinae Deltoidinae The Noctuidae were divided into 10 subfamilies (Table 6). Two (the Acontiinae and theTrifinae) were characterised by an obsolete hindwing vein A/ 2 , while in the other eight this veinwas well developed. Hampson thus used Guenee's major subdivisions rather than Packard's. The Agaristidae [Noctuidae: Agaristinae] were considered to be a development from theNoctuidae, the first occasion on which such a relationship had been postulated. Previously, theyhad been considered to be near the Arctiidae and in the Bombyces (e.g. Smith, 1891). Trifinae. This subfamily approximated Guenee's Genuinae and was defined primarily on thebasis of wing venation (as were many of Hampson's categories). It contained those generapresently referable to the Noctuinae, Heliothinae, Hadeninae, Cuculliinae and Amphipyrinae.The Trifinae also included Acronycta [Acronicta], but not the bryophilines (see below), and alsoToxocampa [Lygephila], a view almost diametrically opposed to that held by Tutt (1891-2;1902). Acontiinae. This subfamily was distinguished from the last by the presence of a slender vein M 2in the hindwing. The larvae generally have four pairs of prolegs but reduction of those on REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 171 abdominal segments 3 and 4 occurs in some species. Most of the 29 genera are still acontiines,but the subfamily also included two acronictines (Diphthera [Moma] and Bryophila [Cryphia\)\three amphipyrines (Megalodes, Cosmia and Pachylepis); two ophiderines (Rivula andTathodeltd) and a hypenine (Perciand) . Palindiinae. Ten genera were included in this , the first of the quadrifine groups , representing amixture of ophiderines and chloephorines, with one sarrothripine (Bryophilopsis) and oneamphipyrine (Callynd). Sarrothripinae. Seventeen genera constituted this subfamily. The majority are still sarrothri-pines, although Ariola is a chloephorine, Nolasena an ophiderine and Chlumetia a euteliine.Also included was the genus Hyblaea. None of the defining characters of this subfamily wasconstant but most had raised scales on the wings and a bar-shaped retinaculum in the male. Euteliinae. Only three genera were included in this subfamily (Ingura [Paectes], Anuga andEutelid) and all three presently reside here. Stictopterinae. This was another very small subfamily, of seven genera. Apart from venationalcharacters, the constituent genera also had fairly well-developed scale tufts on the forewing andfrequently a hyaline area at the base of the hindwing. Only four of the genera are stillstictopterines (Stictoptera, Odontodes, Sadarsa and Gyrtond). The others comprised achloephorine (Maceda), an ophiderine (Cymatophoropsis) and a sarrothripine (Risoba). Gonopterinae. Eighteen genera comprised this subfamily. Again, characterisation was vague,with most of the characters also occurring in some genera of other subfamilies. Seven of the 18genera listed are now ophiderines, the remainder being chloephorines. Quadrifinae. This subfamily was by far the largest in the Noctuidae, comprising 81 genera inall. Most are now placed in the Catocalinae or Ophiderinae but the subfamily also includedgenera assigned to the Plusiinae (Plusia, Abrostola), Amphipyrinae (Cetola), Pantheinae(Moma [Trichosea], Trisuloides] and Chloephorinae (Pseudocalpe [Arcyophora]). Focillinae. In common with many genera of the previous subfamily, the larvae of theFocillinae are semi-loopers. The legs of the adults are also modified, possessing short, stouttibiae and tarsi. Thirteen genera were included, all ophiderines. Deltoidinae. This group was relatively variable regarding wing venation and showed consider-able complexity in male secondary sexual characters. Hampson considered groups such as theTrifinae, Acontiinae, Focillinae and Quadrifinae to have arisen from this subfamily. The Deltoidinae contained 40 mainly ophiderine and hypenine genera but also two acon-tiines, Araeoptemm [Araeopteron] and Niaccaba. The Archiearinae are absent from India and were therefore not considered by Hampson. TheCymatophoridae [Thyatiridae] comprised a separate family, placed near the Notodontidae andthe Sesiidae [Aegeriidae]. Although not placed near the Noctuidae, Hampson stated that werehe to arrange the moth families in a linear sequence, then the Cymatophoridae would/o//ow theNoctuidae, rather than precede them as had been more or less standard practice in the past. Thiswas because he considered them to be a side-branch from the main stem leading from thetortricids, through the noctuids and notodontids, and thence to the Bombycoidea and Geo-metroidea (Fig. 1) . As can also be seen from Fig. 1 , Hampson dismissed a direct link between thenoctuids and either the pyralids or the geometrids. Further development of Grote's classification In the meantime, Grote had been reappraising his noctuid classification. On the basis of 172 Cymatophoridae- IAN J. KITCHING Sphingidae Bombycidae Saturniidae Geometroidea Notodontidae Noctuidae Trifidae 'aristidae LymantriidaeArctiidae Noctuidae Quadrifidae Hypeninae Nycteolinae [Chloephorinae] LimacodidaeEndromidaeasiocampidae Pyralidae Nolinae Tortricinae Tineidae Fig. 1 Phylogeny of the higher Lepidoptera, with particular reference to the Noctuidae, proposed byHampson (1893-5). Note that the Noctuidae are placed internally on the main stem leading from thetortricines to the geometroids and sphingids, while the arctiids form the terminus of a side-branch thatsplits off from the main stem below the noctuids. [unstated] larval characters, he elevated part of the Apatelini to family status (Grote, 18950) butcontinued to place them at the head of the Noctuidae. The Apatelidae, in addition to Acronictaand its relatives, also included Leptina [Baileya] (a sarrothripine), Raphia (an ophiderine), threepantheines (Demos [Colocasia], Panthea and Charadrd) and three cuculliines (Feralia,Arthrochlora [Feralia] and Momaphana [Momophana]). The Noctuidae became the Agrotidae on the grounds that the name Noctua was preoccupiedin the birds. This began a long argument as to the correct family-group name for the noctuids.The subfamilies Catocalinae and Deltoidinae were not discussed in this paper (Grote, 18950)but the remaining group (= Noctuinae) was divided into 24 tribes (Table 7), as follows. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 173 Table 7 The classification of the Agrotidae [Noctuidae] followed by Grote (1895a, b). The firstpublication dealt with the tribes up to and including the Hyblaeini, the second considered thesubfamily Catocalinae. THYATIRIDAEAPATELIDAEAGROTIDAEAgrotinae Jaspidiini Agrotini Psaphidini Hadenini Bellurini Nonagriini Heliophilini Scolecocampini Balsini Caradrinini Orthosiini Cleophanini Cuculliini Euteliini Anomiini Litoprosopini Plusiini Calpini Stiriini Heliothini Acontiini Cerathosiini Eustrotiini HyblaeiniCatocalinae Euclidiini Melipotini Eulepidotini Stictopterini Ascalaphini Catocalini Ophiderini Toxocampini Thysaniini Pheocymini Pangraptini HexeriniBoletobiini Jaspidiini. This tribe of seven genera (the Bryophilinae of Grote, 1883) had been excluded fromthe Apatelidae when this was raised to family rank. The name was based upon JaspidiaHiibner [1806], which Grote (1895a) had substituted for Bryophila Treitschke 1825. Agrotini. The large genus Agrotis and four smaller genera (Anytus [Sutyna], Richia, Eucoptoc-nemis, Agrotiphild) constituted this tribe, which thus approximated the Agrotini of Grote(1890). The following were transferred out: Adita (Hadenidi), Copablepharon (Helio-thini), Ufeus and Pteroscia [Anomogyna] (both Heliophilini). Psaphidini. This was the Dicopini of Grote (1890). Hadenini. Equivalent to the Hadenini of Grote (1890), the genera of this tribe had undergoneconsiderable rearrangement and renaming. 174 IAN J. KITCHING Bellurini. This was the Arzamini of Grote (1890). Nonagriini. This tribe now contained only the first five genera (Nonagria - Tapinostold) of thetribe of the same name in Grote (1890). Heliophilini. The remaining three nonagriines (Ommatostola, Heliophila [Mythimna] andZosteropodd) , together with Ufeus and Pteroscia [Anomogyna], constituted this new tribe. Scolecocampini. There had been no change from the Scolecocampini of Grote (1890).Balsini. This was the Nolaphanini of Grote (1890). Caradrinini. This tribe was essentially unchanged from the Caradrini of Grote (1890). However,Monodes [Elaphria] had been transferred to the Nonagriini (and sunk to Platysenta). Orthosiini. This tribe was equivalent to the combination of the Orthosiini and Calocampini ofGrote (1890). The remaining 13 tribes were unchanged from Grote (1890) although the Tarachini had beenrenamed the Acontiini and lost Trileuca [Schinia, part] to the Orthosiini. Smith on the deltoids Smith (1895), in his revision of the deltoids, recognised this group of moths to be devoid ofexclusive defining characters, although the general facies was relatively distinctive. As restrictedby him, Smith considered that the group might merit subfamily rank within the noctuids.However, he refrained from formally proposing such a grouping. Three (to Smith) very distincttribes were included: the Heliini (Epizeuxis only), the Herminiini and the Hypenini. Two generawere excluded, Pseudorgyia and Rivula, both of which are now considered as ophiderines.Smith also felt that the assignment of the latter to the Nycteolidae [Sarrothripinae] might nothave been the error most previous workers had considered it to be. Grote on the deltoids (and other noctuid groups) Grote (18956) considered that no stable classification could ever be achieved for the Noctuidaeuntil the North American and European faunas were studied together. In a paper written as aconclusion to that of 1883, many of his ideas concerning noctuid classification were expandedand reinterpreted. The Thyatiridae were omitted, support being drawn from Dyar's work on larvae, whichshowed the group to be structurally related to the Geometridae and Platypterigidae [Drepani-dae]. The noctuid moths, therefore, consisted of three families, the Apatelidae, Agrotidae andBrephidae. Agrotidae: Agrotinae. Harrisimemna was removed from the Jaspidiini and placed in theApatelidae, while Raphia was moved from the latter group to the Hadenini. Grote postulated arelationship between Raphia and Episema [Diloba] although he had not examined the Europeanspecies of those genera. The Calocampini was resurrected but only contained Lithomia [Lithomoia] and Calocampa[Xylena]. Lithophane remained in the Orthosiini. Pyrophila [Amphipyra] was removed to a monobasic tribe, the Pyrophilini. Grote agreed withSmith's placement of Pseudorgyia in the Scolecocampini, but having failed to find a satisfactoryposition for Rivula (he could not place it in either the Nycteolidae [Sarrothripinae] or thePseudoipsidae [Chloephorinae]), Grote was forced to erect a new tribe to receive it, theRivulini. Agrotidae: Catocalinae. Thirteen tribes were recognised in this subfamily (Table 7). Euclidiini: a mixture of eleven genera of catocalines and ophiderines (e.g. Parallelia, Drasteria,Phurys [Ptichodis], Poaphila [Argyrostrotis]) . REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 175 Melipotini: Grote included six genera in this tribe (e.g. Hypocala, Melipotes [Bolina]), all ofwhich are now ophiderines. He considered the distinction between this tribe and the last tobe difficult. Eulepidotini: Eulepidotis and Palindia [Eulepidotis] [Ophiderinae].Stictopterini: Stictoptera and Magusa (the latter now an amphipyrine). Ascalaphini: ten genera constituted this tribe, of which most are now ophiderines but Remigia isa catocaline, Argillophora an acontiine and Fagitana an amphipyrine. Catocalini: Allotria, Andrewsia [Catocala, part] and Catocala [Catocalinae].Ophiderini: Ophideres [Ophiderinae] and Euparthenos [Catocalinae].Toxocampini: Toxocampa [Lygephila] and Eutoreuma [Isogona] [Ophiderinae].Thysaniini: Erebus [Ascalapha] [Catocalinae], Thysania and Letts [Ophiderinae]. Pheocymini: this tribe was relatively large, containing 12 genera (e.g. Zale, Yrias [Metria]), ofwhich four are now placed in the Ophiderinae and the remainder in the Catocalinae. Pangraptini: Phalaenostola [Hypeninae], Zethes, Syllectra and Pangrapta [Ophiderinae].Hexerini: this tribe contained only Hexeris [Triprora], now classified in the family Thyrididae. Boletobiini: of the four included genera, two (Boletobia [Parascotia] and Aventia [Laspeyria]are ophiderines, Acherdoa is an amphipyrine and Dyaria [Coenodomus] is placed in thepyralid subfamily Epipaschiinae. Agrotidae: Hypeninae. Grote considered the subfamily name, Deltoidinae, which he hadproposed himself in 1890, to be objectionable because it was not based upon any included genus.He thus substituted Hypeninae. Two of Smith's three tribes were retained, the Heliini beingconsidered insufficiently distinct from the Herminiini to warrant separate status. The position of Nycteola The nycteolines had originally been placed among the tortricids but by the late nineteenthcentury were generally included in the Bombyces as a separate family, most commonly known asthe Nycteolidae (e.g. Smith, 1891). Frequently, a group of genera including Earias andPseudoips was also included, but more often these genera were placed in a distinct thoughrelated family, the Pseudoipsidae. Nycteola had also been referred to the Noctuidae but as a deltoid. Grote (1895b) disagreedwith this. Examination of the immature stages revealed similarities to Pseudoips [Bena]bicolorana [prasinana] but peculiarities in the wing shape and venation led Grote to believe thatNycteola did not belong in either the Pseudoipsidae [Noctuidae: Chloephorinae] (which Groterecognised as distinct) or the Noctuidae. A study of the larval chaetotaxy of Nycteola by Dyar, which was communicated by Grote,bore out the latter's opinions regarding the close relationship between this genus and thePseudoipsidae, and even suggested that the two groups might not be distinct. However, Groteretained the family rank of the Nycteolidae. Hampson (1893-5) considered the nycteolines to be a subfamily of the arctiids, closely relatedto the Nolinae and this postulated relationship was to recur later. The three classifications By 1900, there were three main arrangements of the noctuid genera in use, those of Grote,Hampson and Tutt. Grote's classification was the only one not based upon the families of Guenee, groups thatGrote still maintained had no value. The use of his classification was largely confined to NorthAmerica. 176 IAN J. KITCHING Hampson's classification was a less well-resolved version of Guenee's. It was followed withoutalteration by Leech (1900) while Staudinger & Rebel (1901) modified it slightly. They reintro-duced the Acronyctinae, with the usual constituent genera - Panthea, Trichosea, Moma, Demas[Colocasia], Acronycta [Acronicta], Simyra, etc. - but retained Bryophila [Cryphia] in theTrifinae. The latter subfamily was expanded to include the Acontiinae and Euteliinae ofHampson. With no European representative, the Palindiinae and Stictopterinae were notconsidered. The Sarrothripinae were united with the chloephorines (e.g. Earias) and placed in aseparate family, the Cymbidae, between the Nolidae and the Syntomidae [Arctiidae: Ctenuchi-nae]. Thus, where the sarrothripines were concerned, Staudinger & Rebel chose to acceptDyar's conclusions. The Gonopteridae and Deltoidinae were retained while the Focillinae wereincluded in the Quadrifinae. Whereas the other two systems had undergone only minor alterations, Tutt (1902) introducedconsiderable changes into his classification of the Noctuidae (Table 8). The Noctuides. Despite the opinions of Grote and Hampson, Tutt still clung to the earlierideas regarding the evolution of the noctuids. The Deltoides were considered to be the mostprimitive group from which at least two independent lines had evolved. One of these linesproceeded through the Minores, Genuinae and Bombycoidae to the Arctiidae and Lyman- Table 8 The classification of the Noctuidae followed by Tutt (1902) . It represents a modified versionof Guenee's second classification. Family: 1 BombycoidaeFamily: 2 Noctuidae Subfamily: 1 Noctuinae Tribe: 1 Leucaniidi 2 Agrotidi Subfamily: 2 CaradrinaeTribe: 1 Stilbidi 2 Caradrinidi Subfamily: 3 AmphipyrinaeSubfamily: 4 Apameinae Tribe: 1 Nonagriidi 2 Apameidi 3 Carpocapsidi 4 Bryophilidi 5 HadenidiSubfamily: 5 Orthosiinae Tribe: 1 Orthosiidi 2 CosmiidiSubfamily: 6 Calocampinae Tribe: 1 Xylocampidi 2 Calocampidi 3 Cuculliidi Subfamily: 7 HeliothinaeSubfamily: 8 Plusiinae Family: 3 Acontiidae Subfamily: 1 Acontiinae Subfamily: 2 Erastriinae Subfamily: 3 AnthophilinaeFamily: 4 Catocalinae Subfamily: 1 CatocalinaeFamily: 5 AventiidaeFamily: 6 PhalenidaeFamily: 7 EuclididaeFamily: 8 Poaphilidae REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 177 triidae, while the other passed through the Catocalinae to the geometrids. The Geometridaewere therefore highly evolved noctuids. Bombycoidae. Tutt considered that the larvae of this group resembled the Arctiidae andLiparidae [Lymantriidae] and that these families may therefore have evolved from bombycoidstock. Following Grote, rather than Hampson, Tutt accorded this group family status. Noctuidae. This family was essentially the equivalent of Guenee's Genuinae but with theaddition of the Bryophilini, Stilbidi, Amphipyrinae and Plusiinae. Eight subfamilies wererecognised. Noctuinae. Although Tutt divided the subfamily into three tribes (Leucaniidi, Agrotidi andNoctuidi), he treated the genera of the last two together as the Agrotidi. The Leucaniidicontained the hadenine wainscots and the Agrotidi coincided for the most part with the presentNoctuinae. Caradrininae. Two tribes of small inconspicuous moths formed this subfamily. The Stilbidiincluded Stilbia, Hydrilla [Hydrillula] and their relatives, while the Caradrinini included suchgenera as Laphygma [Spodoptera, part]. Amphipyrinae. This tribe contained only two species, Amphipyra pyramided and A. tragopo-gonis, characterised by their ability to secrete themselves into narrow crevices. Apaminae. Tutt considered this family to have been 'wonderfully muddled by Britishauthors'. He recognised the superficial similarities between the Nonagriidi and the Leucaniidi(Noctuinae) as convergent. Five tribes were included. The Nonagriidi consisted of those wainscots with stem-boring larvae. The Apameidi were dull-coloured and had larvae that were either root-borers or fed on lowplants, often on their roots (e.g. Gortyna, Xylophasia [Apamea] and Miana [Oligia]. Most arenow amphipyrines but there are some hadenines and noctuines.). The Carpocapsidi (Dianthoecia [Hadena]) were distinguished by the larvae feeding exclusive-ly on the seed-capsules of the Caryophyllaceae. Tutt removed Bryophila [Cryphia] from its usual place near Acronicta and placed it in aseparate tribe in the Apameinae, the Bryophilidi. He considered Bryophila to be a hadenoidgenus, related to Hecatera [Hadena]. The last tribe of this group, the Hadenidi, was one that had been merged with the Apameidi byother authors but which many British lepidopterists thought deserved family rank. Tuttconcurred with the former, being unable to find any significant distinction between the twogroups, but he nevertheless retained the Hadenidi as a separate entity. Included were amphipy-rines, cuculliines, and hadenines. Orthosiinae. The moths that comprised this subfamily were characterised by being on the wingeither very late or very early in the year. Two tribes were recognised. The Orthosidi was a mixed bag of genera, with representatives from the present Hadeninae(e.g. Taeniocampa [Orthosia]}, Cuculliinae (e.g. Xylina [Lithophane]) and Ophiderinae(Gonoptera [Scoliopteryx]}. The second tribe, the Cosmidi, was distinguished from the last by the larval habit of spinningleaves together. All are now placed in the Amphipyrinae and include such genera as Dicycla andCalymnia [Cosmia]. Calocampinae. Tutt followed Grote (1895b) by separating Xylina [Lithophane] from the othergenera previously placed in the 'Xylininae'. The remaining group was termed the Calocampinaeand consisted of three tribes: the Xylocampidi (Xylocampa) , the Calocampidi (Calocampa[Xylena] and Lithomia [Lithomoia]} and the Cuculliidi (Cucullia). 178 IAN J. KITCHING Heliothinae. The moths of this subfamily were characterised by the generally diurnally-active,colourful adults and flower-feeding larvae. All the British species currently assigned to theHeliothinae were included, together with the hadenine Anarta and the amphipyrine Heliaca[Panemeria]. Plusiinae. Tutt now considered this subfamily to be a highly-developed group of heliothines.The reduction in the number of larval prolegs (which also occurs in other groups of noctuids) wasexplained as the result of convergent evolution. Only the species currently placed in thissubfamily were included. Its position in the Trifidae agreed with Grote's system rather than withGuenee's. Acontiidae. This family was believed by Tutt to be closely related to the Deltoides, from whichthey might have arisen. Three subfamilies were recognised. The Acontiinae contained Agrophila [Emmelia] and Acontia \Tytd\. Erastria [Lithacodia,Elaphria] and Banksia [Eustrotia] comprised the Erastriinae while the Anthophilinae containedHydrelia [Eustrotia] and Thalpochares [Eublemma]. All except Tyta (Ophiderinae) andElaphria (Amphipyrinae) are currently acontiines. Geometriform noctuides. At this point in the family, Tutt's nomenclature and ranking brokedown, mainly because the groups concerned were poorly represented in the British fauna.Nevertheless, he considered the British species could still give some indications as to theevolutionary history and development of the Geometridae. The families discussed below,together with the Trifidae, were considered to have all arisen independently from thedeltoids. Catocalidae. Only one subfamily, the Catocalinae, was represented in Britain. This was thegroup Tutt considered to be closest to the geometrids, a relationship he thought was moreobvious in certain exotic groups than in the single British genus, Catocala. t Aventiidae. This family contained only Aventia [Laspeyria] flexula and was erected becauseTutt was unable to determine the nearest relatives of the species. The larvae have only threepairs of prolegs ('like some geometers') and a lateral row of fleshy filaments above the prolegs asin Catocala. The adult is deltoid in its general facies. Phalenidae. Tutt retained this name for the family in preference to Brephidae. He interpretedthe group as either being at a point early in the geometrid line of evolution or an isolated lineage,having undergone considerable parallel evolution with the Geometridae. Serpentides. This group consisted of two families. The Euclididae contained the single genusEudidia [Euclidia, Callistege], with larvae that Tutt considered to be 'remarkably geometri-form'. He did not comment on the second family, the Poaphilidae, with its single species,Prothymia [Phytometra] viridaria. The former family is now part of the Catocalinae, the latterpart of the Ophiderinae. Deltoides. Tutt interposed the Geometridae between this group and the Noctuidae. Heconsidered the Deltoides as a group, but not the individual species, to be ancestral to the othernoctuid families. Three deltoid families were recognised. Toxocampidae. This was a group of moths (Toxocampa [Lygephila] and Ophiodes [Minucia])which Tutt considered to be essentially deltoid, but that also had certain affinities with thenoctuids. Tutt was probably the only lepidopterist to associate these genera with the deltoids. Atpresent, the former genus is ophiderine, the latter catocaline. Hypenidae. Five genera (Madopa [Cclobochyla], Hypena, Bomolocha, Hypenodes and REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 179 Thylomiges [Tholomiges]) comprised this family. Of these, the last four are still hypenines,while the first is placed in the Ophiderinae. Herminiidae. This family was divided into two subfamilies: the monobasic Rivulinae (Rivula,now an ophiderine) and the Herminiinae, a group of three genera (Hermina [Zanclognatha,Paracolax], Zanclognatha and Pechipogon [Herminia]), commonly referred to as the 'fan-foots'. Miscellanea. The Nolidae were considered to be related to the lithosiine arctiids, whilst theNycteolidae [?Chloephorinae, ?Sarrothripinae] were a 'very highly developed group which hasevidently been evolved from the Plutellidae or the Tortrices'. Demas [Colocasia] coryli was placed in the Lymantriidae, while Diloba caemleocephala wasclassified in the notodontid subfamily, Dilobinae. Also placed in a subfamily of the Notodonti-dae (as the Asterocopinae) were the two species of Asterocopus [Brachionycha]. Althoughnoctuids according to virtually all previous authors, Tutt found it difficult to consider them astrue noctuids despite their being 'undoubtedly closely allied'. The Cymatophoridae [Thyatiridae] were considered to be equally closely allied to theNotodontidae and the Noctuidae (cf. Fig. 1). These then were the three main systems for classifying the noctuids in use at the beginning ofthe twentieth century. However, in the period since the publication of The Fauna of BritishIndia, Hampson had also reassessed the interrelationships of the noctuid genera. This culmin-ated in the publication of a system that was to become almost universally accepted and, in thelong term, put the higher systematics of the Noctuidae into virtual stasis. Demas [Colocasia} coryli- noctuid or lymantriid? Before proceeding with Hampson's second classification, I shall return to the early 1890s toconsider, in more detail, the controversy that surrounded the position of Demas [Colocasia]coryli. Chapman (18930) suggested that Demas should be returned to the family Liparidae [Lyman-triidae], basing his conclusions upon the eggs, larvae and pupae. He also associated Pantheacoenobita and Diphthera [Trichosea] ludifica with Demas, a group that was eventually to formpart of the subfamily Pantheinae (see below). Tutt (1895) criticised Smith (1893) for retaining Demas in its 'obsolete position . . . among theNoctuae', preferring to place the genus in the Lymantriidae. Dyar (1895ft) disagreed, drawing support from the studies of Poulton (1887). An investigationof the larvae of several Liparidae [Lymantriidae] had demonstrated the presence of dorsaleversible glands on abdominal segments 7 and 8 (more rarely on 8 only) in all species examinedexcept Demas. If these structures proved characteristic of the family, then Demas had to beexcluded. Dyar (1895a) had also found chaetotactic differences between the Noctuidae and theLymantriidae, and showed that Demas conformed to the noctuid type. In addition, he (1895ft)considered the adult to be more similar to the Noctuidae than to any other family. Tutt (1896) replied to this, basing his conclusions upon the work of Chapman (1893,ft), thenshortly afterwards, retired from the debate, having felt he had made his point. It thus befell Dyar(1896) to continue to press for the inclusion of Demas in the Noctuidae, by a reanalysis ofChapman's findings. The eggs of Demas are vertically ribbed as in other noctuids, rather than obscurely ribbed orsmooth as in the Lymantriidae. Chapman used characters such as the length and density of thesetae, the form of the tubercles, especially the anterior trapezoidal (Dl sensu Hinton, 1946), andthe colour. Dyar dismissed the first and last of these as being too likely to be subject toconvergence while he suggested Chapman had been mistaken regarding the tubercle. In fact,Demas shows the strong tubercle of the noctuids rather than the condition found in thelymantriids, in which it is reduced or lost. Dyar disputed Chapman's interpretation of homology between the dorsal glands of lymantriid 180 IAN J. KITCHING larvae and the mid-dorsal depressions found in Demas. To Dyar, the general fades of the larvaof Demas was, if anything, arctiid rather than lymantriid. Overall, he concluded that Chapman'sstudy supported the inclusion of Demas in the Noctuidae and not the Lymantriidae. Grote (1896) followed Dyar, retaining Demas (as well as several other genera now consideredto be pantheines) within the group he called the Apatelidae. However, no good adult characterwas given to differentiate the group from the Noctuidae and the larval characters upon which theApatelidae was based were insubstantial. Smith & Dyar (1898), using the conclusions reached previously by the latter author, restrictedthe family to include only Demas, Panthea, Charadra, Harrisimemna, Merolonche, Arsilonche[Simyra], and Acronycta [Acronicta]. Raphia and Feralia were rejected on larval characters,Momophana and Moma on adult characters. Of these four, the first is now an ophiderine, thenext two cuculliines and Moma an acronictine. Two series were recognised by Smith & Dyar. The first, the Pantheini, contained only Demas,Panthea and Charadra and was distinguished primarily by a quadrifine hindwing venation.Other characteristic features included hairy eyes, pectinate male antennae and a similar faciesand pattern. Raphia also possessed these features although the hairs on the eyes are microscopicand the male genitalia are distinct. Raphia could not be placed satisfactorily and was thereforeomitted from the Apatelidae. The second series, the Acronyctini, consisted of the residual apatelid genera. It wascharacterised by a trifine hindwing venation. As this is also found in several other subfamilies,only the distinct larvae allowed Smith & Dyar to maintain the Acronyctini as a separate entity.In contrast, the Pantheini was a very distinct group of moths as adults, resembling theAcronyctini only in the larval stage. Smith & Dyar concluded that there was no real basis for acontinued association between the two groups, as far as the adults were concerned. The twotribes were nevertheless collected into the noctuid subfamily Pantheinae. The position of Demas in the Noctuidae was consolidated by the works of Staudinger & Rebel(1901) and Hampson (1898-1913). Only Tutt (1902) continued to regard the genus as alymantriid. Development of the Hampsonian monolithThe precursor During the latter part of the 1890s, Hampson had been reappraising the classification of theNoctuidae. The new system, which appeared in part in Hampson (1900), was given in full inHampson (1902). A close relationship was still postulated between the Arctiadae [Arctiidae], Agaristidae[Noctuidae: Agaristinae] and the Noctuidae, and additionally, the Syntomiidae [Arctiidae:Ctenuchinae]. All these families agreed in having forewing vein M 2 approximating the lowerangle of the discal cell. This character was also found in the Pterothysanidae, Lymantriadae[Lymantriidae] and Hypsidae [Arctiidae: Aganainae] and defined what Hampson termed the'Noctuid group of families'. This grouping largely agrees with the present superfamily Noc-tuoidea (if one accepts the Notodontoidea as separate). The only exception to this is thePterothysanidae, which is now placed in the Geometroidea, although the sole species con-sidered by Hampson, Pterocerota virginea, may possibly be a eupterotid (R. Carcasson,manuscript note). Hampson (1902) did not detail the reasoning behind his new system. The characters used toseparate the subfamilies were given in a key (Table 9) and were based upon those of Lederer andGrote, although employed in a novel fashion. The Agrotinae [Noctuinae] also included some of the species currently referable to theHeliothinae (e.g. Melideptria [Heliothis]) . The other trifine subfamilies were the Mamestrinae[Hadeninae], Polianae [Cuculliinae] and Caradrininae [Acronictinae + Amphipyrinae]. The quadrifine subfamilies comprised the Eutelianae [Euteliinae], Stictopterinae, Sarrothri-pinae, Acontianae [Chloephorinae], Homopterinae [Catocalinae], Mominae [Pantheinae],Plusinae [Plusiinae], Noctuinae [Ophiderinae] (and the chloephorine, Earias), Erastianae REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 181 Table 9 Key to the subfamilies of the Noctuidae recognised by Hampson in his second classification of thefamily (modified from Hampson, 1902). 1 Maxillary palps absent 2 Maxillary palps present Hyblaeinae 2 M 2 hindwing obsolescent, crossveins m 1 -m 2 and m 2 -m 3 subequal 3 - M 2 hindwing well-developed 6 3 Mid- and hind-tibiae spined Agrotinae Mid- and hind-tibiae not spined 4 4 Eyes hairy Mamestrinae Eyes not hairy 5 5 Eyes with long, overhanging cilia Polianae Eyes not ciliated Caradrininae 6 Hindwing M 2 approximated to M 3 at base 7 - Hindwing M 2 parallel to M 3 Hypeninae 7 Female frenulum simple 8 - Female frenulum multiple 9 8 Lateral abdominal hair pencils present near anus Eutelianae Hair pencils absent, forewing with tufts of raised scales in cell Stictopterinae 9 Male retinaculum bar-shaped 10 Male retinaculum not bar-shaped 11 10 Forewing with tufts of raised scales in cell Sarrothripinae - Forewing without tufts of raised scales in cell Acontianae 11 Midtibiae spined Homopterinae Midtibiae not spined 12 12 Eyes hairy Mominae - Eyes not hairy 13 13 Eyes with long, overhanging cilia Plusianae Eyes not ciliated 14 14 Hindwing M 2 strong, from close to lower angle of cell Noctuinae Hindwing M 2 rather weak , from well above lower angle of cell Erastrianae [Acontiinae] (and several amphipyrines, e.g. Cnodifrontia and Callopistrid) and Hypeninae. Hampson retained the name Noctua but applied it to the ophiderine Noctua strix [Thysaniaagrippina]. This was the first species listed by Linnaeus (1758) in the 'genus' Noctua and so wasdesignated the type-species, following Hampson's personal rule (that is, the first listed specieswhen the genus was described; McDunnough, 1916). Thus, the group then currently known asthe Noctuinae was termed the Agrotinae. Hampson's Catalogue The basic system of subfamilies was unchanged in Hampson's major work, the monumentalCatalogue of the Lepidoptera Phalaenae in the British Museum (1898-1913). (Subsequentreferences to the Hampsonian system, unless qualified, refer to that proposed in this work.) Thescheme of the phylogeny of the Lepidoptera as a whole had undergone some rearrangementfrom that published in Fauna of British India. The relevant parts of this new phylogenetic treeare shown in Fig. 2. The major difference between Figs 1 and 2 is that, in the latter, the Noctuidae and relatedfamilies were considered as terminal on the main stem, rather than being intermediate stages ofdevelopment culminating in the Geometroidea. The relationship between the Noctuidae andGeometridae, as espoused by Tutt (1902), was thus rejected. The subgroups of the noctuids arediscussed in detail below. Volume 2 - Arctiadae: Nolinae [Noctuidae: Nolinae]. The nolines were retained by Hampsonas a subfamily of the Arctiadae [ Arctiidae] . They were considered to have arisen from very earlyarctiid stock, close to the noctuids Hypena and Sarrothripus [Nycteola] (with which they sharethe character of tufts of scales in the forewing cell). The similarity between the cocoons of the 182 IAN J. KITCHING Syntomidae Agaristidae Notodontidae Uraniadae Zygaenidae Lymantriidae, Pterothysanidae ^SphingidaeThyatiridae^Eupterotidae .GeometridaeBombycidae^Saturniidae Cossidae'LimacodidaeLLasiocampidae Rhopalocera r PyralidaeTortricidaeDrepanidae Tineidae Fig. 2 Phylogeny of the higher Lepidoptera, with particular reference to the Noctuoidea, proposed byHampson (1898-1913). There are several important differences from the previous phylogeny (Fig. 1).The noctuoid families are now considered to be terminal on the main stem and the arctiids are derivedfrom the noctuids. In addition, the notodontids are now ancestral to the noctuids (the reverserelationship was postulated in Fig. 1). (Redrawn from Hampson, 1898-1913, 1: 16.) nolines and certain sarrothripines and chloephorines was also noted. Hampson excluded thelatter two groups from the Arctiidae, considering them better placed in the Noctuidae becauseof the presence of ocelli and the free origin of hindwing vein Sc + R { . Volume 4 - Noctuidae: Agrotinae [Noctuinae]. This subfamily was characterised by a trifinehindwing venation and spined hind (and occasionally mid-) tibiae. Some species also had hairyor lashed (ciliated) eyes, characters of the Hadeninae and Cuculliinae respectively. Thishighlights a peculiar feature of the classificatory process as employed by Hampson. The major REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 183 (key) characters were treated as hierarchical. Thus, any trifine species with spined tibiae wasplaced in the Agrotinae irrespective of whether it also had the defining characteristics of theother subfamilies. Of the remaining species, any with hairy eyes were hadenines, though somealso had lashed eyes. The cuculliines were those of the remainder with lashed eyes, while theresidue , with none of these characters were , by default , acronyctines [ Acronictinae + Amphipy-rinae]. A similar but less hierarchical principle applied within the quadrifines. This method of allocating genera to subfamilies can thus be seen to be arbitrary in thoseinstances in which two or more characters conflict, due to the arbitrary order in which the keycharacters were ranked. Many of the species were misplaced (as considered by recent works)and it was from this realisation that the dissatisfaction with the system largely grew. The Agrotinae contained two groups of genera corresponding to the present Noctuinae andHeliothinae (less the pyrrhine amphipyrines). Volume 5 - Hadeninae. This group corresponded to the Mamestrinae of Hampson (1902) . Theincluded genera were hairy-eyed trifines with unspined tibiae, and a few (e.g. Trichopolia andStretchid) also had lashed eyes. Volume 6 - Cucullianae [Cuculliinae]. This trifine subfamily, previously referred to byHampson (1902) as the Polianae, was characterised by bare, lashed eyes and unspined tibiae. Inaddition, many of the species appear as adults in the autumn or early spring, sometimeshibernating. Volumes 7-9 - Acronyctinae [Acronictinae -I- Amphipyrinae]. This immense subfamilycontained almost half of the then described genera of trifine noctuids. It was characterised byabsences; no tibial spining and eyes devoid of hair and lashes. Hampson placed no specialemphasis on the diverse hairy larvae of the acronictas, which were incorporated within theamorphous mass. The Acronictinae were interpreted as the most primitive trifine subfamily from which theother three, together with the Agaristidae, had evolved as separate lineages. Volume 10 - Erastrianae [Acontiinae]. This subfamily was the first of the Hampson quadrifinegroups. Most genera had a strongly developed hindwing vein M 2 , approximated at the base toA/3, but some had an almost typical trifine venation (e.g. Eupseudomorpha, Cydosia, Graeperia[Heliodora], Tarachidia). The last three of these remain acontiines but the first is nowconsidered to be an agaristine. Overall, the subfamily was homogeneous in general facies butnot well defined structurally. In many ways, it was intermediate between the trifine and theremaining quadrifine subfamilies. The larvae of most of the genera had the first one or two pairsof prolegs reduced or missing. Volume 11 - Eutelianae [Euteliinae]. Characterised by a quadrifine wing venation, simplefemale frenulum and the presence of anal hair tufts in the males, this subfamily was essentiallysimilar to that in Hampson (1893-5). He now considered this subfamily to be a paralleldevelopment, with the Stictopterinae, from the Noctuinae [Ophiderinae]. Volume 11 - Stictopterinae. This subfamily had also undergone little change. Its constituentgenera also possessed the reduced female frenulum but had tufts of raised scales in the forewingcell and lacked the euteliine anal hair pencils. Volume 11 - Sarrothripinae. This group was characterised by a bar-shaped retinaculum in themale and by tufts of raised scales in the forewing cell. An exception was Eligma, a genus nowplaced in the Chloephorinae (see below and also Mell, 1943). The larvae of the Sarrothripinaeare sparsely covered in long hair and pupate in a boat-shaped cocoon. Volume 11 - Acontianae [Chloephorinae]. Many genera of this subfamily also have a 184 IAN J. KITCHING bar-shaped retinaculum in the male. However, in Earias, Tyana and several other genera, thishas been replaced by a tuft of hair. All are smooth-scaled except for Cerala [Kerala], which hasslight scale tufts in the forewing cell. (Despite this, it was not placed in the Sarrothripinae, andstill is not.) The larvae of the Acontiinae, like those of the previous subfamily, have five pairs of pro-legs and spin a boat-shaped cocoon on a twig, except for Acontia [Xanthodes], Leocyma andpossibly Armactica, in which the larvae are semi-loopers and pupation takes place under-Hampson considered these genera might be better placed in the Erastrianae, whichwould then take the name Acontianae, whence the original Acontianae would become theEariasinae. Volumes 12-13 - Catocalinae. The Catocalinae was the first of two large sections into whichHampson divided the Quadrifinae of 1893-5. They were characterised by bare, unlashed eyes,normal frenulum and retinaculum, and spines on the mid-tibiae. Those Plusianae [Plusiinae]with spined tibiae could be distinguished by the presence of lashed eyes. However, Hampson didobserve that many catocaline genera were closely allied to genera in the Noctuinae [Ophideri-nae] and he considered it probable that either the Catocalinae were polyphyletic, with mid-tibialspines being multiply convergent, or that the Noctuinae were polyphyletic, characterised bymultiple independent loss of spines. Either way, Hampson felt the two large subfamilies mighthave to be united as one large subfamily. Subdivision would then have to be on other grounds.Many of the larvae of the Catocalinae are semi-loopers, a feature Hampson considered to beconvergent in this group and others such as the Plusianae [Plusiinae], Noctuinae [Ophiderinae]and Hypeninae. Volume 13 - Mominae [Pantheinae]. This subfamily was equivalent to the tribe Pantheini ofSmith & Dyar (1898). Hampson considered hindwing venation to be a more important characterthan larval facies and so removed the group from its association with Acronicta and its allies to aplace among the quadrifine subfamilies. Hampson considered the Mominae to be very natural and well characterised with twoexceptions. Of these, Elydnodes is still a pantheine. However, Epicausis was transferred to theCuculliinae by Viette (1962). On the basis of male and female genitalia, he placed it betweenDaphoenura and Eudaphaenura, genera that Epicausis also resembles in colour pattern. Viettealso found that Hampson's interpretation of the hindwing venation was incorrect, for althoughM 2 is present, it is weaker than either MI or A/ 3 . However, Viette neglected the fact thatEpicausis has distinctly hairy eyes (which the other two do not; Kitching, pers. obs.) and thusresembles a third genus, the hadenine Adaphaenura, which also has a colour-pattern similar tothat of Epicausis. Viette characterised the Hadeninae and Cuculliinae as having hairy and lashedeyes respectively, yet still placed Epicausis in the latter. With the involvement of mimicry adistinct possibility, the position of Epicausis and its putative allies must still be consideredunsettled. Viette (1973) retained the genus in the Cuculliinae, as did Nye (1975). Volume 13 - Phytometrinae [Plusiinae]. Using his personal rules for type-species designationmentioned earlier, Hampson considered it necessary to rename the group of moths that hadbeen previously known generally as the Plusiinae. The subfamily consisted of those quadrifineswith lashed eyes and included the monobasic genus Diloba, in addition to Abrostola, Phy-tometra [Plusia s.l., part] and their relatives. Apart from Diloba, which Hampson considered tobe aberrant (for example, the larvae possessed five pairs of fully developed prolegs), thePhytometrinae were well circumscribed. Diloba itself had been considered to be notodontid, acronictine or even thyatirid. Chapman(1893a) found the genus very difficult to place. The egg is macroscopically similar to Acronicta,but the sculpturing is different. The larvae of the two groups have little in common, while thepupal cremaster has a slight resemblance to Bisulcia [Craniophora] (Acronictinae). Overall,Chapman considered Diloba to be more bombycid than noctuid and suggested that it mightrequire a family of its own. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 185 Hampson, however, considered Diloba to be noctuid. Thus, because of a quadrifine hindwingvenation and lashed eyes, it had to be placed in the Phytometrinae. Volumes [14-16] (unpublished) - Noctuinae [Ophiderinae]. The publication of the Cataloguewas suspended during the First World War for financial reasons. When Hampson retired in1920, he left the manuscript covering the subfamily Noctuinae [Ophiderinae] (Gahan, Prefaceto Hampson, 1926). It is from that manuscript that the following is taken. The Noctuinae were quadrifine noctuids with bare, unlashed eyes, which were distinguishedfrom the Catocalinae by the unspined tibiae, and from the Polypogoninae [Hypeninae] by thethird segment of the labial palp not being acuminate (most genera) and by hindwing vein M 2being approximated to the lower angle of the cell. The labial palp segment three was alwaysacuminate in the Polypogoninae while M 2 arises well above the lower angle of the cell and runsparallel to A/ 3 (except in the Mastigophorus-group). The larvae all move in semi-looper fashion, even if all the prolegs are present (except forRaphia, whose affinities are open to question). Hypeninae and Hyblaeinae. Hampson never dealt with these groups, the last two subfamiliesof the Noctuidae as he conceived it (see Table 9). The system proposed in Hampson's Catalogue, albeit unfinished, was to exert a profoundinfluence on the higher classification of the Noctuidae and is still generally accepted today. But itis possible that it too might have gone down in history as just another arrangement, were it notfor its acceptance and usage by the authors of Seitz' Die Gross-Schmetterlinge der Erde. This,more than anything else , was responsible for confirming the primacy of Hampson's classificationin noctuid systematics. Seitz' Die Gross-Schmetterlinge der Erde The publication of the parts of this huge work that dealt with the Noctuidae spanned the years1906 to 1944 and they were written by five authors - Jordan, Warren, Gaede, Draudt and Seitzhimself. Volume 3 - Fauna Palaearctica: Agaristidae (Jordan). Jordan (1906-14) followed Hampsonin allying the Agaristidae [Noctuidae: Agaristinae] with the Noctuidae, considering them asmerely 'day-flying Noctuidae which have preserved some generalised characters, being in otherrespects more spezialized [sic] than the Noctuids'. This interpretation was also espoused byDraudt (1919^4, Fauna Americana) and by Jordan & Gaede (1919-39, Fauna Africana).Originally the agaristines were placed at the head of the Bombyces, as a result of Linnaeus'(1758) opinion that antennal shape (i.e. clubbed) was of primary importance in the classificationof the Noctuidae (Seitz, 1909). Noctuidae (Warren). For convenience, Warren (1906-14) initially adhered to the fivesubfamilies used by Staudinger & Rebel (1901). The only change Warren introduced was to usethe terms Trifidinae' and 'Quadrifidinae' for 'Trifinae' and 'Quadrifinae' respectively, becausehe considered the syllable 'fid' to be an integral part of the word upon which the name was basedand not part of the family ending 'idae'. With the publication of the first noctuid volume of Hampson's Catalogue, Warren chose tofollow the new system as closely as possible but employing as few changes as possible also. As aresult, complete correspondence with Hampson's subfamilies was not achieved (Table 10). The Acronictinae, which also included the Pantheinae, was recognised as heterogeneous inwing venation but was accepted on the grounds that the included species had hairy larvae, whichfeed exposed, and that pupation takes place in a cocoon above ground. The Metachrostinae, equivalent to the Bryophilidi of Tutt (1902), included only Metachrostis[Cryphia]. For the remaining noctuids, Warren thought it generally advisable to followHampson's arrangement. 186 IAN J. KITCHING Table 10 Sequence of noctuid subfamilies adopted by Warren (1906-14) in Seitz' The Macrolepidop-teraofthe World. NOCTUIDAE Acronictinae Metachrostinae Euxoinae Hadeninae Cucullianae Amphipyrinae Melicleptrinae Heliothidinae Erastrianae Eutelianae Stictopterinae Sarrothripinae Acontianae Catocalinae Phytometrinae Noctuinae The Agrotinae was renamed the Euxoinae, for unstated reasons. The Hadeninae andCucullianae [Cuculliinae] were unaltered, while the Amphipyrinae represented the residue ofHampson's 'Acronyctinae' after removal of the Acronictinae (s.str.) and the Metachrostinae. The Melicleptrinae, roughly equivalent to the present Heliothinae, was composed primarilyof those trifine genera with curved fore-tibial claws. The next subfamily, the Heliothidinae, has no modern equivalent. The genera that comprisedit were characterised by small, reniform eyes (although Meyrick, 1912, regarded them as beingmore correctly termed ovate), a distinctly hairy vestiture and a primarily arctic-alpine distribu-tion. Six unnamed divisions were recognised. I. Anartomorpha. This genus combined the noctuine spined tibiae with the hadenine hairyeyes. Following Hampson's character hierarchy, it is currently placed in the Noctuinae(although it is treated as a hadenine, near Anarta, in the BMNH collection). II. Schoyenia, Orosagrotis [Euxoa subgenus], Grumia, Oxytypia. Characterised by noctuinefeatures, all these genera are currently placed in the Noctuinae (although Grumia is in theHeliothinae in the BMNH collection). III. Anarta, Panolis. This division was characterised by hairy eyes and unspined, unclawedtibiae. Both genera are presently placed in the Hadeninae. IV. Omia, Sympistis, Hypsophila, Cteipolia. These four lashed-eyed genera are currentlyplaced in the Cuculliinae. V. Heliothis [Schinia, part], Pyrocleptria [Schinia, part]. These two genera bear the spinedtibiae of the Noctuinae and the clawed fore-tibiae of the Melicleptrinae [Heliothinae]. Theyare currently placed in the latter subfamily. VI. Xanthothrix, Apaustis, Panemeria, Janthinea, Stenoecia, Micriantha, Stemmaphora, Omor-phina, Mesotrosta. These genera comprised those with unadorned eyes and tibiae, and thusqualified as amphipyrines. All are currently accepted as such except Omorphina, which wasclassified by Hampson in the Plusiinae (Kostrowicki, 1961 , considered it better placed in theAcontiinae). Warren recognised the heterogeneous nature of this new subfamily, in that its membersdisplayed all the features used to define the previous five groups. However, he considered theshared characters to be more important. The group was conceived to be the 'scattered remnantsof archaic types, the conditions of whose existence synchronised with the more rigorous climateand scantier light of an earlier cosmic period'. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 187 The Palaearctic genera of the Erastrianae [Acontiinae] were recognised to straddle thetrifine-quadrifine border, in that they possess a strongly-developed hindwing vein M 2 (incontradiction to Hampson's key, Table 9) but the larvae, as far as were known, had a reducednumber of prolegs. Warren treated them as intermediate between the other two large noctuidgroups. The remaining quadrifine subfamilies recognised by Warren coincided generally with those ofHampson. However, no distinction was made between the Noctuinae [Ophiderinae] and theHypeninae, the combined group being referred to under the former name. Diloba was includedwith reservations within the Cymatophoridae [Thyatiridae] (Warren & Seitz, 1906-12). Volume 7 - Fauna Americana: Noctuidae (Draudt). Draudt also largely adhered to Hamp-son's arrangement. However, the 'Acronyctinae' group [Acronictinae] was published before thevolume in which Hampson merged it with the Amphipyrinae and so the division was maintainedby Draudt. The Acronictinae was divided into two 'subordinate groups', the Mominae [Pan-theinae] and the Acronictinae, which Draudt recognised as probably not being closely relatedbut which were kept together in order to maintain the coherency of the entire work. Warren's heliothidine genera were distributed among the other subfamilies (e.g. Orosagrotis[Euxoa subgenus] to the Agrotinae [Noctuinae] and Anarta to the Hadeninae). As a result, thegroup Warren termed the Melicleptrinae, which now contained Heliothis [Schinia, part], wasrenamed the Heliothinae and must have added greatly to the confusion of those using the work. Unlike Warren, Draudt did separate the Noctuinae [Ophiderinae] and Hypeninae as distinctsubfamilies (1919-44: 17) although only the ophiderine part was ever published. Volume 11 -Fauna Indo-Australia: Agaristidae (Jordan). In this volume, Jordan (1912-14),in addition to his previous comments regarding this family, also added that it possibly ought to bemerged with Hampson's Acronyctinae. Indeed, the latter author (Volume 9) included severalagaristine genera in the Acronyctinae (e.g. Xerocerus to Zalissa inclusive). Jordan also thoughtit feasible that the Agaristidae were polyphyletic, with Old and New World forms representingdistinct lineages. Noctuidae (Warren, Gaede & Draudt). The trifine subfamilies were dealt with by Warren(1912-38), who retained the system he had used in the Palaearctic volume. The only alterationwas the renaming of the Metachrostinae as the Bryophilinae. The majority of the quadrifinesubfamilies, up to the Catocalinae, were described by Gaede (1937-38) and the remainder left toDraudt (1938). However, of the latter, only four lines of the introduction to the Phytometrinae[Plusiinae] were produced, the rest never being published. Both authors followed Hampson'sarrangement of subfamilies. Volume 15 - Fauna Africana: Noctuidae (Gaede). Several modifications were made to thearrangement employed in the previous volumes. Gaede (1913-39) followed Hampson inseparating the acronictines and the momines [pantheines] into distinct subfamilies, although thetwo were still placed contiguously at the head of the family. In addition, the Sarrothripinae weredivided into four informal groups: (1) Sarrothripus-group: mostly very small with broad wings; (2) Eligma-group: large and brightly coloured; (3) Gadirtha-group: allied to the last in having long, subuliform palps and a dorso-ventrallyflattened body, but being only half the size and cryptically coloured; (4) Blenina-group: with cylindrical bodies and very broad, cryptically coloured wings. Gaede also recognised the arbitrary nature of the division between the Catocalinae andNoctuinae [Ophiderinae] but considered that combining them, and perhaps also incorporatingthe Erastriinae [Acontiinae] and Hypeninae, would result in a very unwieldy group. Otherwise,Gaede saw no reason for separating such generic pairs as Hypocala-Catocala, Fodina-Grammodes and Arcte-Cocytodes . 188 IAN J. KITCHING This work, like many other volumes of Seitz, remained unfinished due to the destruction ofthe press and manuscript in the Second World War. Satisfaction and apparent stability The authors of 'Seitz' were not the only enthusiastic proponents of Hampson's subdivisions ofthe Noctuidae. Dyar (1904), in his review of the agrotine [noctuine] volume, consideredHampson's arrangement and use of characters 'highly commendable'. However, he was far fromcontent with the nomenclatural changes imposed upon his native American fauna. In particular,Dyar objected to Hampson's refusal to adopt the names used by Hiibner in the Tentamen. Meyrick (1912), however, concurred with Hampson, reasoning that Hiibner's names wereinvalid because they were published without accompanying descriptions. Meyrick did disagreewith Hampson regarding the potential use of sexual characters to define genera, characters thatwere totally eschewed by the latter author. As a result, Meyrick found it difficult to accept manyof Hampson's generic groupings. He did find some points of agreement with regard to the higherclassification of the Noctuidae, which he was able to incorporate into his own arrangement of theNew Zealand Caradrinina. Meyrick employed the term 'Caradrinina' to cover the group today known as the Noctuoidea(s.str.), maintaining his view that the generic name Noctua and its associated group-names wereinapplicable to this group of moths. In addition to the Arctiadae [Arctiidae] and the Hypsidae[Arctiidae: Aganainae], two other families were recognised. These were the Caradrinidae andthe Plusiadae, direct equivalents of the Noctuidae and Plusiadae of his earlier (1887) work. Inthe present study, Meyrick chose to divide the families into subfamilies, largely followingHampson's usage. He did employ different names, however (Table 11). Of these, the Plusiadesalso included the ophiderines. Meyrick (1928) later employed the same scheme in his BritishLepidoptera. Table 11 The classification of the superfamily Caradrinina proposed by Meyrick (1912), the contentsof which correspond to the current superfamily Noctuoidea, less the Lymantriidae and several smallgroups. CARADRININA ArctiadaeHypsidaeCaradrinidae Agrotides Poliades Melanchrides CaradrinidesPlusiadae Hypenides Catocalides Plusiades Forbes (1914) implicitly adopted the Hampsonian subfamilies although he did not employ thenames as such. He was also not impressed by the vagueness of the trifine-quadrifine border andreferred to those in which the condition was doubtful as 'intermediid'. Barnes & McDunnough (1917) were more faithful to Hampson's system. The only alterationwas the substitution of Erebinae for Noctuinae [Ophiderinae]. Turner (1920: 120) went much further in his praise of Hampson: 'It would be difficult toover-estimate the debt which we owe to Sir George Hampson's great work. By it the study of theNoctuidae as a whole has been for the first time placed on a scientific basis'. Despite these words,Turner found it difficult, as did many other authors, to define the noctuid subfamilies preciselywithout recourse to 'distinctions of relatively trivial importance' (e.g. hairy eyes, spined tibiae,etc.). Overall, however, Turner considered it advisable to adopt Hampson's arrangement,primarily because no better system was available. This attitude was to become entrenched. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 189 Turner formally relegated the agaristines to the rank of subfamily within the Noctuidae, andthus agreed with Mosher's (1916) findings on pupae (see below). He dismissed one of the usualdefining characters (clubbed antennae) and could thus include several genera placed byHampson in the Acronyctinae (as did Jordan, 1912-14). However, having done so, Turner wasthen quite unable to define the group. The remaining subfamilies dealt with (up to the Acontianae [Chloephorinae]) remainedunaltered, apart from the substitution of Hadeninae by Melanchrinae. Hampson's arrangement was also accepted by Lhomme (1923-35) although again, the namesof certain subfamilies had been changed (the Acronyctinae, Stictopterinae, Acontianae andMominae became the Zenobiinae, Odontodinae, Westermanniinae and Diphtherinae respec-tively). This was in line with Hampson's supplement to the Catalogue (Hampson, 1918). Similarly, minor nomenclatural changes apart, Hampson's subfamilies were adopted byBlackmore (1927). He agreed with Barnes & Benjamin (1923), in accepting that the Linnaeangeneric name Phalaena applied to the noctuids (rather than the geometrids, where it had morefrequently been used). Hence, the family became the Phalaenidae and the subfamilies Agro-tinae and Noctuinae [Ophiderinae] became the Phalaeninae and Erebinae respectively. (Thefamily name Noctuidae was not finally stabilised until the suppression of Phalaena Linnaeus,1758, by the International Commission for Zoological Nomenclature in 1957, Opinion 450.)Also, the Acronyctinae were referred to as the Apatelinae, presumably because Blackmoreaccepted Hiibner's Tentamen and thus considered Apatela [Apatele] to have priority overAcronycta [Acronicta]. Finally, as an example of a study following Hampson almost to the letter, that of Prout (1929)might be mentioned. The only alteration employed was that she followed Comstock (1925) andregarded the hyblaeines as a family distinct from the noctuids (see also below). Noctuoidea or Noctuidae? A rather interesting variant of the Hampsonian system of families and subfamilies of thenoctuoid group of moths was supported by some South African lepidopterists, notably Janse(1937-9). He considered the differences between the noctuoid families to be far too smallcompared to those found in other orders, and thus, following Handlirsch's (1929) suggestionsregarding ranking, reduced them all to subfamilial level in a more inclusive Noctuidae (Table12). Inevitably, there were conflicts with Hampson's arrangement. The sarrothripines (as Nyc-teolini) were removed from the Noctuidae on the basis of a venational character (despite the'shape of the forewings somewhat resembling those of the Acontinae' [Chloephorinae]) andplaced as a tribe in the Arctiinae [Arctiidae]. Also the hypenines were recognised as a tribe ofthe Noctuinae, equivalent in rank to the trifines (Trifini) and the quadrifines (Quadrifini).Consequently, most of Hampson's subfamilies were relegated to the status of subtribes. As to Hampson's subfamilial characters, Janse considered hairy eyes and tibial spines to befairly reliable, at least with regard to excluding genera from his Acronyctae, while lashed eyesand the condition of hindwing vein M 2 were much less dependable. He also disapproved of theuse Hampson made of thoracic and abdominal tufts and crests, preferring to delimit genera onthe basis of structural characters, particularly those drawn from the male genitalia. Only the Agaristinae and Noctuinae: Trifini: Acronyctae were covered by Janse (1937-9).The study was discontinued due to lack of material of non-South African type-species (Janse,1942). This arrangement of genera was also used by Van Son (1933), Janse (1939), and, in a modifiedform, by Kiriakoff (1963). The last of these works will be discussed in greater detail in a laterchapter. Hampson to the present day Despite considerable reservations regarding the naturalness or, in practical terms, even theusefulness, of Hampson's subfamilial classification, most workers have continued to use it, for 190 IAN J. KITCHING Table 12 The concept of the Noctuidae employed by Janse (1937-9), which is almost identical to thecurrent superfamily Noctuoidea. However, Janse did not consider the differences between theconstituent groups (his subfamilies) to be great enough to warrant family rank. NOCTUIDAE SyntominaeCallimorphinaeLiparidinaeArctiinae Lithosiini Nolini Arctiini NycteoliniAgaristinaeNoctuinae Trifini Quadrifini HadenaeAgrotidesCuculliaeAcronyctae Momae Euteliae Stictopterae Acontiae Catocalae Plusiae Noctuae Erastriae Hypenini example, Zimmerman (1958), Common (1968), Pinhey (1975) and Leraut (1980). Severalothers have made only small modifications: Viette (1962-7) separated the Melicleptrinae[Heliothinae] from the Noctuinae; Boursin (1964) additionally accepted the Apatelinae [Acro-nictinae] (moves that were subsequently followed by Kloet & Hincks, 1972 and Bretherton etal., 1979); while Forster & Wohlfart (1971) recognised the Apatelinae and the Bryophilinae butnot the Heliothinae. Thus it seems that, minor differences in opinion regarding the precise names of thesubfamilies apart, Hampson's subdivision of the family could be regarded as definitive anddominant. However, although apparently stable and imperturbable, the system was challenged,even during its publication. Hampson's classification was based largely upon superficial charac-ters of the adults and larvae. Detailed morphological and anatomical studies of all life-historystages were to cast serious doubt on the edifice. This doubt was eventually to lead to the proposalof a new system which, although still very much dependent upon that of Hampson, was to bemore securely based in structural morphology. The influence of immaturesEmergence Most classifications of Lepidoptera have been derived using primarily adult characters. How-ever, in several groups, of which the Noctuidae are one, the economic importance of the larvaeresulted in studies from which grew the realisation that not only could larval characters be usefulin identification but also in classification. Forbes (1910) was one of the first to investigate thoroughly larval morphology. Severalnoctuid subfamilies were considered in detail and within the Noctuinae (sensu the Trifinae) inparticular, Forbes discovered a highly uniform structure, despite much individual variation. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 191 Only the tribes Acronyctini [ Acronictinae] and Cuculliini [Cuculliinae] could be delimited usinglarval characters. Fracker (1915) also investigated larval morphology but with special emphasis upon thechaetotaxy. He agreed with Hampson that the Nycteolidae ought to be reduced to a subfamily ofthe Noctuidae (Sarrothripinae), having found no significant chaetotactic differences betweenmembers of the two families. The larvae of the Agaristidae [Noctuidae: Agaristinae] , apart fromtheir bright transverse striping (also found in some other noctuids, e.g. Acronicta alni), werealso indistinguishable from those of the Noctuidae and Fracker considered the family 'shouldmeet the same fate as the Nycteolidae'. Within the Noctuidae itself, Fracker found the same discouraging degree of uniformity. Thisprompted him to write: 'with the exception of half a dozen genera, noctuid larvae are so uniformthat one can often compare, part for part, every segment and appendage of larvae of two specieswithout rinding a difference either of kind or of degree' . Fracker could only divide the family intofour groups, three of which contained only members of the Acronictinae and Pantheinae, whilethe fourth comprised all the remaining subfamilies. Mosher (1916) found a greater range of variation in pupae and was able to characterisetentatively 10 subfamilies (largely Hampsonian). One, the Agaristinae, was included perhapsfor the first time as an explicit subfamily of the Noctuidae. As had been found for the larvae, nostructural differences of the pupae could reliably distinguish the agaristines from the noctuids atthe family level. The Acronyctinae were heterogeneous for, in addition to the acronictinesEulonche and A crony eta [both Acronicta} and the amphipyrine Achatodes, the division alsoincluded three ophiderines (Homopyralis [Metalectra], Plusiodonta and Anomis). However,Mosher did not consider that these genera formed a natural group. Of the remaining subfamilies, members of the Catocalinae were distinguished by the presenceof an alcohol-insoluble surface bloom, while the Sarrothripinae completely lacked a cremasterand spines (although Mosher did not consider this sufficient to warrant separation from theNoctuidae). Ripley (1923) performed an extensive study of the morphology of larval noctuids but did notinterpret the results taxonomically. The larvae of the Hypeninae were examined by Crumb (1934), who discovered two distincttypes. The first corresponded to Forbes' (1918) Herminiinae while the other was found not onlyin the remaining hypenines but also generally in the catocaline-erebine [ophiderine] complex. Inthis regard, the hypenoid type (as Crumb termed the latter form) was not particularly distinctfrom the majority of the Noctuidae. The herminioid type of larva was divided into twosubgroups and the hypenoid type into five. On the basis of the larva, it was suggested that Rivulamight merit a separate subfamily; it certainly was not hypenine. Dethier (1941), in his detailed study of lepidopteran larval, antennae, described those of themembers of 10 noctuid subfamilies (including the Rivulinae). Apart from noting the greatsimilarity between the Agaristidae and the Noctuidae, no further remarks were made regardingthe higher classification. The challenge from India A major contribution to the understanding of the structure of larvae from outside the Holarcticwas provided in a series of papers by Gardner on Indian Noctuidae. He subdivided the family,on the basis of previously defined larval characters (Gardner, 1941), into four groups, A-D, ofwhich the first three were split further (Gardner, 1946a). The groups were not allocated formalnames because of conflict with Hampson. Division A (considered mostly in Gardner, 19466) consisted of various trifine genera, theAgaristinae and some acontiines [chloephorines]. It comprised six sections. A I: (described in Gardner, 1946a): Acronictinae, Pantheinae and Cetola (Amphipyrinae).The larva of the last genus was not distinctly hairy but was included because it had sixsetae on the external surface of the prolegs rather than three. 192 IAN J. KITCHING A II: Amphipyrinae, Hadeninae; also Agrotis [Xestia] c-nigrum (Noctuinae) and Lyncestisamphix (Ophiderinae). A III: Noctuinae (Agrotis and Euxoa). A IV: (described in Gardner, 19460): Heliothinae (Heliothis, Pyrrhia andAdisura). A V: Acontiinae [Chloephorinae] (Aiteta, Carea, Maurilia, Pseudelydnd). A VI: Amphipyrinae (Callyna), Hadeninae (Tiracola), Chloephorinae (Churia), Agaristinae(Aegocera, Eusemia}. Division B (mostly dealt with in Gardner, 1947) comprised those larvae in which the prolegs ofabdominal segment 3 are abruptly reduced or absent. The distinction from the genera of divisionC, in which the prolegs are all equal or are only gradually reduced from segment 6 to 3, wasconsidered to be somewhat arbitrary. B I: (described in Gardner, 19460): Acontiinae [Chloephorinae] (Acontia [Xanthodes]) ,Amphipyrinae (Elydna, Chasmina, Androlymnid) , Erastriinae [Acontiinae] (Amyna)and Noctuinae [Ophiderinae] (Boculd). B II: (described in Gardner, 19460): Hypeninae (Hypena, Dichromia, Rhynchina, Bomo-lochd). B III: Plusiinae. B IV: Erastriinae [Acontiinae], Catocalinae (e.g. Parallelia, Achaea), Noctuinae [Ophideri-nae] (e.g. Othreis, Anomis) and Hypena [Sarobela] aurotincta [litterata] . Group B IV wasdivided into three subgroups: B IVa: lacking prolegs on abdominal segments 3 and 4;B IVb: lacking prolegs on 3 only;B IVc: prolegs present on 3 and 4, those of 3 distinctly reduced. Division C (Gardner, 19480), unlike A and B, was not subdivided. Instead, the genera weretreated according to their Hampsonian subfamily. Included were the subfamilies Euteliinae,Sarrothripinae, Stictopterinae, Herminiinae, some Catocalinae and Ophiderinae, together withWestermannia (Chloephorinae) and the amphipyrines Sesamia and lambia. The final division, D (Gardner, 1947), differed from the others in setal group VI (part of thesubventral group of Hinton, 1946) being bisetose on the meso- and metathorax rather thanunisetose. Gardner considered that division D was artificial (he was eventually to conclude thatthe entire system was artificial; Gardner, 19480), comprising as it did, Brithys (Hadeninae),Selepa (Sarrothripinae), and Earias and Eligma (Chloephorinae). These genera snared charac-ters with the Lithosiidae [Arctiidae: Lithosiinae], Hypsidae (Digama [now an arctiine]) and thegenera Argina and Utetheisa (Arctiidae: Arctiinae). Gardner (19480) considered that a natural classification of the noctuids should be based uponthe number of setae in group VII (part of the subventral group of Hinton, 1946) on the firstabdominal segment. This would divide the Noctuidae into two groups, representing roughlydivisions A plus B, and C plus D, but with many exceptions. The larval study was followed by a consideration of the pupae (Gardner, 1948ft). The pupae ofthe Hadeninae, Amphipyrinae, Agrotinae [Noctuinae] and Heliothinae were all found to bevery similar. Several genera (e.g. Brithys) that were distinct as larvae, proved equally distinct aspupae. The relationship between Cetola and the Acronictinae suggested by the larvae was notsupported by the pupae. The Euteliinae proved to be very well defined by pupal characters. Thegenera of Acontiinae [Chloephorinae], which had been widely dispersed on the basis of larvalcharacters, were all very close as pupae, except Acontia [Xanthodes]. They also shared featureswith the Sarrothripinae. No great distinction was found between the pupae of the Catocalinaeand the Noctuinae [Ophiderinae], thus giving Mosher's surface bloom character wider signi-ficance. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 193 Table 13 The division of the British Noctuidae proposed by Timlin (1955), based on larvae. NC = nocomparison; l 1 = uniordinal, 2 = biordinal; prolegs on abdominal segments 3 and 4 only partlydeveloped; 3 except for certain Cucullia species. Group Prolegnumber Secondarysetae Crotchets 1 SV group of setae abdominal segment 1 Contents 3 3 4 5 25 NC Most Acronyctinae [Acronictinae], Dasycamparubiginea (Cuculliinae) 2 Plusiinae (Plusia[Autographa] gamma, P. [Diachrysia] chrysitis) 3 Plusiinae (Polychrysiamoneta) 3 Ophiderinae (Parascotia fuliginaria) 3 Eustrotiinae [Acontiinae] 3 Hypeninae 3 Plusiinae (Episema [Diloba] caeruleocephala), Ophiderinae(part), Acronyctinae[Acronictinae] (Apatele[Acronicta] alni) 2 Agrotinae [Noctuinae], Hadeninae, Cuculliinae(except D. rubiginea)Amphipyrinae, Acronyctinae[Acronictinae] (Cryphia perla[domestica]), Sarrothripinae,Westermanniinae [Chloephorinae] Further conflict Timlin (1955) examined the larvae of 142 species of British noctuids, dividing the family intoeight equivalent groups (Table 13). This arrangement was then contrasted with those ofHampson, Meyrick (1928) and Tarns (unpublished). Tarns' classification differed from Hamp-son's only in separating the Acronictinae from the Amphipyrinae.Timlin found conflicts in the following areas: (a) the heterogeneity of the Plusiinae, especially regarding Episema [Diloba]; (b) the similarity between the Hylophilidae [Chloephorinae + Sarrothripinae] and the Noc-tuidae; (c) the reduction of the prolegs in the Eustrotiinae [Acontiinae]; (d) the lack of a distinction between the Catocalinae and the Ophiderinae; (e) the high degree of similarity between the trifine subfamilies (less the Acronictinae). In addition, Timlin considered the following genera might have been misplaced in classifica-tion based upon adult structures: 1, Dasycampa (Cuculliinae), which has secondary setae; 2, Parascotia (Ophiderinae), considered closer to the Hypeninae because of the reducednumber of prolegs; 3, Cucullia (Cuculliinae), which is heterogeneous with respect to crotchet ordination; 4, Scoliopteryx (Ophiderinae), in which seta XD1 is nearer the midline than Dl on theprothorax, contrary to Fracker's (1915) diagnostic character for the Noctuidae; 5, Episema [Diloba], which, unlike the other plusiines, has five well-developed pairs of prolegsand uniordinal crotchets; 194 IAN J. KITCHING 6, Apatele [Acronicta] alni and Cryphiaperla [domestica] (Acronictinae), the only acronictinesexamined lacking secondary setae. The next major work on noctuid larvae was that of Crumb (1956) . He attempted to classify thelarvae of as many North American genera and species as possible, in order to facilitate theidentification of crop pests. The arrangement of subfamilies adopted is given in Table 14. Crumb accepted the Agaristinae as a noctuid subfamily but did not discuss it further. He couldnot distinguish the Pantheinae and the Acronyctinae [Acronictinae] on structural grounds andso combined them under the latter name. The Heliothinae were found to be separable into twofairly good groups but Crumb did not apply the distinction. Table 14 Arrangement of the noctuid subfamilies followed by Crumb (1956). PHALAENIDAE Agaristinae Acronyctinae Acontiinae Cuculliinae Euteliinae Bagisarinae Hypeninae Phalaeninae Hadeninae Ufeiinae Amphipyrinae Plusiinae Lithacodiinae Sarrothripinae Catocalinae Herminiinae Within the remaining subfamilies, apart from the unorthodox order of listing, several otherinnovations were proposed. New subfamilies were erected for Bagisara (Bagisarinae) and Ufeus(Ufeiinae). They are currently considered to be tribes of the Acontiinae and Noctuinaerespectively (Franclemont & Todd, 1983). The Cuculliinae was restricted to the genera relatedclosely to Cucullia, the others being placed in the Amphipyrinae. The Acontiinae was restrictedto Acontia, Tarachidia, Heliocontia [Spragueia] and Pseudalypia (the Acontiini of Franclemont& Todd, 1983) while the remainder constituted the Lithacodiinae. The Amphipyrinae weredivided into eight informal groups (three of cuculliines, three of amphipyrines and twomixtures). One of the mixed groups (number 6, Pseudanarta, Proxenus, Anorthodes, Platyper-igea, Xanthia, Sunira andAnathix) was considered to be possibly natural. The Catocalinae (including the Ophiderinae) was divided into six informal groups, whichpartially correlated with the groupings of Forbes (1954) (see also below) as follows: 1, Erebinae: Synedini; 2, Catocalinae: group 3; Erebinae: second miscellaneous series (part) (Tathorhynchus); 3, Catocalinae: groups 1 and 2 (parts); Erebinae: Erebini and Panopodini (part); 4, Catocalinae: groups 1, 2 and 4 (parts); Erebinae: Panopodini (part); 5, Erebinae: first miscellaneous series; 6, Erebinae: Anomiini and Scoliopterygini. Beck (1960), in his study of the larvae of European noctuids, used a classification based uponHampson but with several differences. The Noctuinae, Hadeninae and Zenobiinae [Amphipyri-nae] were treated as tribes of the Noctuinae (following Borner's (1953) classification of theadults). The Cuculliinae and Heliothidinae [Heliothinae] were recognised as distinct trifinesubfamilies. The Rivulinae and Herminiinae were recognised as separate while the remainingophiderines were classified in the Catocalinae. Two tribes were recognised in the last subfamily, REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 195 the Catocalini (with two informal groups: Laspeyrial Catocalal Mormonial Minucia; and Eucli-dimera [Callistege]IEctypa [Euclidia\ILygephila) and the monobasic Scoliopteryginae (S. libat-rix). The ophiderine Aediafunesta was established in a separate subfamily, the Aediinae. Merzheevskaya (1967) generally accepted an arrangement that was very similar to that ofBeck (1960). However, the Agrotinae [Noctuinae] were divided into two tribes, the Triphaenini[Noctuini] and the Agrotini. One unusual move transferred Amphipyra to the Cuculliinae,resulting in the renaming of the Amphipyrinae as the Zenobiinae. Godfrey (1972), following Hampson's definition of the subfamily, examined the larvae ofmany of the North American Hadeninae and divided the subfamily into 21 informal groups. The most recent study of noctuid larvae (Arnold, 1982) was novel in that it looked not at themorphology of the larvae but at their haematocytes. Four basic types of cells were identified:plasmatocytes (of five varieties: standard, nematoform, lamellar, podoform and vesicular),granulocytes, spherulocytes and oenocytoids. The species of Agaristinae, Pantheinae and Acronictinae examined all had only the basic celltypes, in common with five arctiids, and this complement was interpreted as primitive. TheNoctuinae fell into three natural groups, which showed little agreement with Forbes' (1954)divisions. The hadenines could also be divided into three groups on the basis of plasmatocytetype. The few cuculliines examined agreed with Forbes' subdivision; the three Cuculliini hadonly the basic complement while Lithophane hemina (Xylenini, Franclemont & Todd, 1983)differed in having lamellar plasmatocytes. The Amphipyrinae proved to be heterogeneous whilethe two heliothidines [heliothines] shared the unusual feature of the absence of spherulocytestaining. Generally, classifications based upon noctuid larvae have proved to be at least partiallyincongruent with the Hampsonian system, while the degree of conformity with the newerarrangement exemplified by Franclemont & Todd (1983) remains to be seen. Larvae haveproved to be potentially very useful in elucidating the higher classification of nymphalidbutterflies (DeVries, Kitching & Vane-Wright, in prep.; Kitching, 1983, in press) whereprevious systems based upon adult characters have been shown to be incorrect by varyingamounts. It seems likely, therefore, that no satisfactory arrangement of the noctuid genera intotribes and subfamilies can be achieved without reference to the immature stages and much workstill remains to be carried out in this field. The main assaultThe gauntlet is thrown down Tympanal organs: the development Workers in noctuid systematics, from Guenee to Hampson, relied largely upon gross morpho-logical structures to divide the group-wing venation, tibial spining, etc. The prime reason for thelack of study of other features, such as the arrangement of sutures and sclerites, was the presenceof the characteristic scales of the Lepidoptera. Examination of surface structures wouldnecessitate the removal of this vestiture, which was an anathema because of its effect on theappearance of cabinet specimens. However, around 1910, attitudes began to change. The use of genitalia revolutionisedlepidopteran systematics at the generic and specific levels, but the system that was to have thegreatest impact on the higher levels was based on the tympanum. It had been known for some time that such an organ existed (it was first referred to by Swinton,1877) but it was Forbes (1916) who initially drew attention to the possibility of using it to helpdetermine taxonomic relationships. Forbes (1918) employed the form of the tympanic hood toargue for the resurrection of the Herminiinae as a subfamily separate from the Hypeninae.Tympanal structures were also employed to argue that the anomalous arctiid, Graphylesia, wasin fact a noctuid, related distantly to certain acontiine [chloephorine] genera (Forbes, 1924).The physiology and anatomy of the noctuid tympanum was further investigated by Eggers (1919;1925). 196 IAN J. KITCHING Application The wider systematic implications of the tympanum in the Noctuidae were left to Richards(1932). Drawing upon the studies of Shepard (1930), Richards made a highly detailed compari-son of the morphology and evolution of the structure and divided the family as follows. Herminiinae. Richards agreed with Forbes (1918) that this group was distinct from theremainder of the noctuids. The included genera all possessed a prespiracular tympanal hood (i.e.the first abdominal spiracle is posterior to and therefore outside the hood), as in the othernoctuoid families (except the Nolidae and Agaristidae, see below), rather than a postspiracularhood (in which the spiracle is concealed beneath the hood). The former condition wasconsidered to be primitive and led Richards to conclude that the Herminiinae were among theearliest offshoots from the noctuid lineage. Rivulinae. Although the members of this rather ill-defined subfamily possessed the derivedpost-spiracular hood, Richards considered them to represent the most primitive noctuidsubfamily. The remaining noctuid groups and the Herminiinae were interpreted as having beenderived from the Rivulinae. Rivula itself appeared to be the least specialised of all the Noctuidae examined by Richards,differing only slightly from the Hypsidae [Arctiidae: Aganainae] (although the latter has aprespiracular hood). Hypeninae. Richards found it immensely difficult to subdivide the large group of generacomprising the Hypeninae, Catocalinae and Erebinae [Ophiderinae]. He was able to distinguishthe genera near Hypena from the remainder of the Erebinae (except Plusiodonta and Scoliop-teryx), considering them as a separate subfamily, the Hypeninae. (By way of an aside, Richards,in a footnote (10), stated that 'the immediate Hypena group and Scoliopteryx are the onlyquadrifids with lashed eyes outside the Plusiinae'. If this is true, then one can only conjecture asto why Hampson did not place at least the latter in the Plusiinae.) The Hypeninae wereconnected to the Rivulinae via Pleonectyptera [Hemeroplanis] and to more typical Erebinae viaPangrapta-Herminodes . Three informal groups were recognised within the Hypeninae: 1, Pleonectyptera [Hemeroplanis], with an unspecialised tympanum; 2, the lashed-eyed hypenines, with a double pocket IV; 3, Hormoschista and its allies, with pocket IV yet further modified. The erebine-catocaline complex. This group comprised Hampson's Catocalinae and Noc-tuinae. Richards, like Gaede (1913-39), considered the distinction based upon tibial spining tobe completely artificial. However, unlike the latter author, Richards found he was able to dividethe combined group, albeit based upon a limited sample of genera. Richards was very aware ofthe limitation this placed upon his conclusions. The genera of the erebine-catocaline complex proved very diverse, falling into a number ofrelatively distinct groups which were connected by intermediate forms. Six such groups wererecognised. I. Pangrapta/Gabara/Herminodes group. This was the most primitive, from which the nextgroup arose. II. AnomislCalpel Plusiodonta group. This was considered to be rather artificial, connectingthe previous series to the 'higher' erebines. Alabama and Anomis were considered to beextreme developments, while Plusiodonta was tentatively placed here, although it hadmore in common with the Hypeninae. III. Melipotis/Syneda [Drasteria] group. Derived from the last mentioned, this was the mostwell-defined group of erebines, with a characteristic tympanal membrane and nodularsclerite. It includes Leucanitis and Syneda [Drasteria], genera currently assigned to the REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 197 Catocalinae and Ophiderinae respectively. These genera had tympana that Richardsconsidered to afford 'good evidence of the artificiality of any division on spines'. IV. EuparthenoslYrias [Metria]! Zale group. Characterised by an overhung or pouched pocketI, this group was considered to have given rise to the Stictopterinae and Plusiinae (seebelow). V. Catocalal Erebus [Ascalapha\IThysania group. Richards considered this group to be ratherartificially separated from the last. VI. DoryodeslEudidia group. This group, comprised of mainstream Catocalinae, was con-sidered to be an independent derivative of group II. Four genera were found to be very difficult to place: Scoliopteryx, Gonodonta, Noropsis[Diphthera] and Litoprosopus . The last of these was described as looking like 'a combination of aNotodontid and a Plusiid' and was highly anomalous. Stictopterinae. Richards considered this subfamily to be derived from the Erebinae of groupIV, and possibly to have given rise to the Plusiinae, 'for which they would serve as prototype'. Plusiinae. This subfamily proved to be the most homogenous and distinct of the quadrifines. Itwas characterised by a double hood and a swollen area of the epimeron ventral to the tympanalmembrane, which was the external manifestation of an enlarged pocket IV. (A similar bulge inother genera, notably the Herminiinae, is ventral to pocket IV.) Euteliinae. This subfamily proved enigmatic. Generally, it was considered to be a develop-ment from the Erebinae, and would thus be placed near the Stictopterinae. However, thetympana of the euteliines have much in common with those of the Erastriini. Richards wasundecided as to the position and affinities of the Euteliinae. Erastriinae [Acontiinae]. Richards derived this subfamily from the lineage linking theRivulinae and Hypeninae, and the Erebinae, via Eublemma [Eumicremma + Eublemma].From this group could be derived the acontiine [chloephorine]-sarrothripine series and thetrifine subfamilies. As mentioned above, the Euteliinae could be placed here also. Two tribes were recognised by Richards. The Erastriini, in which the alula is unmodified, andthe Tarachini, a very homogeneous group in which the alula is enlarged and strongly sclerotised.The tympanal membrane is concealed by the alula and the hood is correspondingly reduced orabsent. Acontiinae [Chloephorinae]. This subfamily combined the characters of both erastriine tribesand was considered to be derived from the last subfamily. Sarrothripinae. Richards agreed with Hampson by placing this group as a derivative of thelast, with which it shares many characters, especially internally. Pantheinae. The relationships of this tympanally homogeneous subfamily were obscure toRichards. The form of the tympanum could be derived from that of the trifines, erastriines oreven the erebines, and he was unable even tentatively to assign it a place on his phylogenetictree. The trifine subfamilies. These groups (Agrotinae [Noctuinae], Poliinae [Hadeninae], Cucul-liinae and Acronyctinae [Acronictinae + Amphipyrinae]) were not examined in detail butappeared to be highly similar. Richards considered them to be developments from the highererastriines with which they share many internal features. Agaristidae. This family was considered to be derived from the Acronyctinae. 198 IAN J. KITCHING Nolidae. Richards concluded that this group did not belong in the Arctiidae, but was in factvery close to the Erastriinae, from which it could possibly be derived. Richards synthesised his findings into a phylogenetic tree (Fig. 3). The arrangement, heconcluded, was not substantially different from that of Hampson apart from the position of thePlusiinae, the unknown position of the Pantheinae and Euteliinae, and the division of theCatocalinae + Ophiderinae. However, it should be noted that there are several other majorconflicts. 1. Hampson placed the Acontiinae [Chloephorinae] and Sarrothripinae between the Strictop-terinae and the Catocalinae whereas Richards considered them to be derivatives of theErastriinae [Acontiinae] near the trifines. 2. The Rivulinae, noctuines [ophiderines] according to Hampson, were given subfamilial statusand considered to be the most primitive noctuids by Richards. 3. The Herminiinae were separated from the Hypeninae. 4. The nolids were associated with the 'higher' noctuids rather than with the arctiids. Thus Richards' work provided evidence that cast serious doubt upon many of the Hampso-nian subfamilial relationships. Nevertheless, it was based, for the most part, on a single organsystem, the components of which are not readily observable. It was also based upon observa-tions on only a very small percentage of the species of Noctuidae, whereas Hampson'sarrangement was constructed after study of most of the then-known species. The influence thatRichards' work was to have, and its potential was great, remained to be seen. The cause is taken up Initial acceptance in the U.S.A. The main proponent of the taxonomic use of tympanal organs in noctuid classification continuedto be W. T. M. Forbes. However, even he was disinclined to change the accepted classification(i.e. Hampson's) to conform with the more recent evidence (cf . Forbes' work on the phylogenyof the butterfly subfamily Danainae, reviewed by Ackery & Vane-Wright, 1984). Thus, in hisstudy of the Lepidoptera of Barro Colorado Island (Forbes, 1939), he retained the nolids as aseparate family (placed between the Euchromiidae: Ctenuchinae [Arctiidae: Ctenuchinae:Euchromiini] and the Arctiidae: Lithosiinae), despite noting that they possessed a tympanumsimilar to that of the Erastriinae [Acontiinae] . Similarly, the structure of the agaristid tympanumwas mentioned as being 'of the Noctuid type' but the family status was retained and the groupwas left near the arctiids, between the Pericopidae [Arctiidae: Pericopinae] and the Lyman-triidae. McDunnough (1938) employed subfamily concepts (Table 15) based largely upon the work ofRichards (1932) and the results of comparative studies of male genitalia, which were then beingwidely used following the pioneering work of Pierce (1909). The Pantheinae were associatedwith the Acronictinae at the head of the noctuids, a return to the older concepts of relationshipsrather than anything novel (but see below). The heliothines were separated from the agrotines(or phalaenines as McDunnough preferred to call them) and placed relatively distantly. TheCatocalinae and Ophiderinae were considered to be one subfamily and the rivulines andherminiines were treated as distinct. The subfamily Hyblaeinae continued to be included in thenoctuids, as the last listed. These three changes were revolutionary insofar as they were includedin a major checklist for the first time and McDunnough's list was thus a step in the right direction . The Lepidoptera of New York and neighboring states The next stage in development was published by Forbes (1954). By including the results of themost recent studies on the Noctuidae, he was able to produce a classification that was the mostdetailed in its hierarchical structure since that of Tutt (1902) (Table 16). Based loosely uponHampson, the noctuids were split into 14 subfamilies, many of which were further divided intotribes. In addition, and contrary to normal taxonomic practice, Forbes established 'miscel- REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 199 Plusiinae Nolidae Agaristidae Group VI Erebinae Acronyctinae Acontiinae Tarachini Agrotinae Poliinae Cuculliinae Sarrothripinae Main body Herminiinae Basal herminiinegenera Hypena group other hypeninegenera Pleonectyptera Rivulinae Fig. 3 Phylogenetic tree of the Noctuidae based upon characters of the tympanal organ (redrawn fromRichards, 1932). Dashed lines indicate uncertain derivations, while the three arrows suggest threeequally plausible positions for the subfamily Euteliinae: from the stictopterine-plusiine branch; from thegroup IV Erebinae; or from the Erastriini. 200 IAN J. KITCHING Table 15 Classification of the Phalaenidae [Noctuidae] employed by McDunnough (1938). AGARISTIDAEPHALAENIDAE Pantheinae Acronictinae Phalaeninae Hadeninae Cuculliinae Amphipyrinae Heliothinae Acontiinae Euteliinae Sarrothripinae Plusiinae Catocalinae Hypeninae Rivulinae Herminiinae Hyblaeinae laneous series' and 'isolated genera', which were not assigned to formally named groups withinsubfamilies. Agrotinae. Forbes used the family-group name 'Noctuidae' on the basis of '150 years of almostunchallenged use here and nowhere else'. However, he felt that to employ 'Noctuinae' for asubfamily would only invite confusion because of the previously diverse applications and so theterm 'Agrotinae' was retained for the subfamily containing Agrotis, Euxoa and their allies. TheNoctuinae of Hampson was termed the Erebinae. Forbes divided the Agrotinae into two tribes. The first, the Heliothidini, was considered to beclosely allied to Agrotis, although some authors (e.g. McDunnough, 1938) preferred to separatethe two groups widely and derive the heliothines from the acronyctine tribe, the Pyrrhiini. TheAgrotini was divided into three sections based largely upon genitalic characters. The mainincluded genera in each were as follows: group I: Agrotis, Eucoptocnemis, Copablepheron;group 2: Peridroma, Pseudorthosia, Richia;group 3: Noctua [Xestia], Eurois, Actebia, Cerastis. The first and last were considered to be fairly homogeneous, despite previous minute subdivi-sion. Hadeninae. Defined primarily on the presence of hairy eyes, Forbes considered this group tobe fairly homogeneous although he had reservations regarding the amaryllidaceous-feedinggenus Xanthopastis . Three subgroups were recognised on the basis of superficial characters: 1. a Mamestra-type, with rough, mixed vestiture and heavy tufting; 2. a 'Taeniocampa'-type, with very smooth mixed vestiture, dull colouring and completepattern; and 3. a Leucania-type, with fine, smooth vestiture, striate pattern and a grass-feeding larva. However, the male genitalia did not agree with this division but instead suggested a largecentral homogeneous mass, from which individual genera and even species diverged. Also,many genera did not fit any of the above groups. Cuculliinae. Forbes considered this subfamily to be possibly polyphyletic, with two series oftribes being independently derived from the Acronyctinae [Amphipyrinae -I- Acronictinae].The first series (Cuculliini, Oncocnemidini and possibly the Psaphidini) were thought to berelated to Catabena, Oxycnemis and some Stiriini, while the second series (Lithophanini, REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 201 Table 16 The classification of the Noctuidae proposed by Forbes (1954). NOCTUIDAE Agrotinae Heliothidini Agrotini (divided into three informal groups)Hadeninae (divided into three informal groups)Cuculliinae Cuculliini Oncocnemidini Psaphidini Cleocerini Lithophanini AntitypiniAcronyctinae 'First series' Septidini Apameini (divided into two unnamed, intergrading subgroups) Arzamini Phlogophorini Dipterygiini Prodeniini'Second series' Apatelini Stiriini Eudryini Pyrrhiini (the Acronyctinae also included many isolated and unplaced genera)Acontiinae Eublemmini Erastriini Acontiini (plus several other unplaced genera)EuteliinaeSarrothripinaePantheinaePlusiinae Catocalinae (divided into four unnamed groups)Erebinae Erebini Panopodini Synedini Anomiini 'First miscellaneous series' Scoliopterygini 'Second miscellaneous series' Third miscellaneous series'HypenodinaeHypeninaeHerminiinae Cleocerini and Antitypini) were conceivably related to such genera as Andropolia, Rhizagrotisand some of the Septidini/ Apameini complex. The Cuculliini comprised Cucullia only, and appeared to be closely related to the next, theOncocnemidini, into which it intergraded in Europe. Containing six genera, the Psaphidinishowed similarities to the last two tribes, as well as to the Lithophanini and the Antitypini. TheCleocerini contained only Cleoceris [Brachylomia] and Litholomia, while the Lithophanini,characterised by the presence of a well-developed digitus, was considered to be the dominant 202 IAN J. KITCHING group of cuculliines, with 17 included genera. Although the Antitypini formed a well-definedtribe in North America, Forbes was of the opinion that such was not the case in Europe. Acronyctinae [Acronictinae 4- Amphipyrinae]. Despite its considerable size and degree ofhomogeneity, Forbes was able to subdivide this subfamily into two series of six and four tribeseach. In addition, there were many unplaced, isolated genera. First series. This group of tribes was polythetically defined by male genitalia with a normalcorona, free pleurite with muscle-plate, lobed penicillus and a well-developed digitus. Six tribeswere included: Septidini, Apameini (divided into two intergrading and unnamed groups),Arzamini, Phlogophorini, Dipterygiini and Prodeniini. In addition, eight isolated genera, whichhad some of the features of the Septidini, remained unplaced. Second series. This group of tribes was characterised by the more or less general reduction orabsence of a corona and digitus, an unlobed penicillus, no free ninth pleurite, a vinculumsometimes produced as a narrow strip, and larvae that are never borers or subterraneancutworms. Four tribes were recognised: Apatelini (equivalent to the most restrictive Acronicti-nae of previous authors), Stiriini (e.g. Stiria, Plagiomimicus) , Eudryini (Eudryas and Psycho-morphd) and Pyrrhiini (Pyrrhia and its allies). Three groups of isolated genera were alsorecognised: (A) e.g. Cosmia, Amphipyra, with high-feeding, usually green, larvae; (B) Proxenus and Anorthodes, with cryptically-coloured, brown larvae, feeding close to theground or as a cutworm; and (C) Galgula, with an unknown larva, about which Forbes would not conjecture. Acontiinae. With regard to this subfamily, Forbes followed Richards (1932), but recognisedthree tribes, not two. The extra tribe, the Eublemmini, was erected to include Eublemma[Eumicremma] only which Richards considered to be a link with the deltoids. In addition, theTarachini were renamed the Acontiini. Euteliinae. Containing three genera (Eutelia, Marathyssa and Paectes), this subfamily wasconsidered to be homogeneous and was unchanged from previous works. Sarrothripinae. Considered to intergrade with the Nolidae in the Old World, the Sarrothripi-nae were believed to be 'an offshoot of the ancestral Noctuid type' (cf. Richards, 1932). Pantheinae. This subfamily was considered to be generally well defined, although Raphiadiffered in its naked larva and in the hair on its eyes being microscopic. Plusiinae. Despite being a distinctive subfamily, Forbes considered the Plusiinae might wellbe combined with the primarily Old World group, the Stictopterinae. In addition to the usualgenera (Abrostola and Plusia s.l.), Forbes included Phyprosopus, because this aberrant genuswould key out to this subfamily and a better position could not be suggested. Catocalinae. Forbes recognised that the distinction between this subfamily and the Erebinae[Ophiderinae] broke down outside the Holarctic region but found it a useful concept within theNorth American fauna. On the basis of genitalic characters, the Catocalinae were divided intofour relatively distinct groups, which were not formally named. Erebinae. Forbes used Richards' analysis as a basis for his classification of this group. Heagreed with Richards that all the other quadrifine subfamilies, bar possibly the Herminiinae,could be derived from it. Because the North American genera represented diverse elements and REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 203 any classification based upon these alone would be meaningless, Forbes chose to remove onlyfive distinct groups as tribes and left the remainder in three 'miscellaneous series' correspondingto those groups to which they had been assigned by Richards. Erebini. This tribe was characterised by such features as very short discal cells and distinctivetarsal spining and included Erebus [Ascalapha], Thysania and Bendis [Lesmone]. Panopodini. From the North American fauna, only Panopoda was included in this tribe.However, the group was much larger in the tropics, where Forbes suggested it might needsubdivision. He also considered that the catocaline group 4 (Zale and Euparthenos) couldbe derived from a subgroup of the Panopodini. Synedini. This tribe coincided with Richards' erebine-catocaline group III and, in Forbesopinion, was very homogeneous, with the exception of Phoberia and Hypocala. Anomiini. This was another tribe Forbes considered to be well founded but for which he failed toelucidate the interrelationships within the subfamily. Consisting of Anomis and Alabama,the tribe corresponded to Richards' 'extreme development of group II'. Forbes also noted apossible affinity with the BagisaralElydnodes group of the Acronyctinae, and felt also thatthe genera Amyna and Xanthodes might belong in the Anomiini. 'First miscellaneous series'. This group represented those genera of Richards' group II thatlacked the enlarged alula of the Anomiini. Scoliopterygini. This erebine tribe (containing only Scoliopteryx libatrix) was characterised bystrongly lashed eyes and a mixture of erebine and hypenine features. 'Second miscellaneous series'. The genera comprising this series were interpreted as the mostprimitive erebines by Richards but Forbes considered them to grade into the first miscel-laneous series, via Hypsoropha. 'Third miscellaneous series' . Approximately equivalent to Richards' Rivulinae , this group was amixture of primitive erebines. Hypenodinae. This group was composed of the smallest of the noctuids and was characterisedby the absence of ocelli (also missing in the sarrothripine Comachard). Forbes was uncertainwhether the group deserved subfamilial or tribal rank. Hypeninae. After removal of the Hypenodinae and the Herminiinae from the deltoids, theresidue was placed in the Hypeninae, distinguished by long labial palps and lashed eyes. Herminiinae. Forbes included 13 genera in this subfamily, which was characterised by aprespiracular tympanal hood. The following were considered by Forbes (1960). Agaristidae. The great similarity between this family and the Noctuidae (they were 'hardlydistinct from the Eudryas-group of the Noctuidae'; but see below) did not influence Forbes andhe maintained the two as separate entities. Nolidae. This family was dealt with by Franclemont, who considered it to be derived from thesarrothripine noctuids and linked to them in the Old World by such genera as Barasa. Thus, anyresemblance to the Lithosiinae (Arctiidae) was convergent. Nevertheless, the Nolidae wereallocated their time-honoured position between the lithosiine arctiids and the Euchromiidae[Arctiidae: Ctenuchiinae: Euchromiini]. The system of Forbes was generally adopted by Inoue & Sugi (1958-61), although it differedin detail in several respects: 204 IAN J. KITCHING 1, the Heliothidinae [Heliothinae], Apatelinae and Cryphiinae [both Acronictinae] wereelevated to subfamily status; 2, Forbes' informal groupings of hadenines were recognised as distinct tribes: the Poliini,Orthosiini and Leucaniini; 3, no subdivisions were used in the Amphipyrinae, Acontiinae or Catocalini (which alsoincluded Forbes' Erebinae); 4, the Hypenodinae were not recognised. In addition, the Sarrothripinae and Chloephorinae were merged as a single subfamily, theNycteolinae. This usage was important insofar as it represented the first occasion on which the newAmerican classification had been applied to an Old World fauna. Tympanal organs in Europe While tympanal organs were having far-reaching consequences in noctuid taxonomy in NorthAmerica, they were being virtually ignored on the other side of the Atlantic. Only Kiriakoff paidthem any attention. In a series of 15 papers between 1948 and 1960, he studied many groups ofthe Noctuoidea, of which four are of direct relevance to the higher systematics of the Noctuidae. The first (Kiriakoff, 1953) concerned the Hyblaeidae and will be considered further later. The second (Kiriakoff, 1955) concluded that the Agaristidae were very close to the Noctuidaebut separable on the basis of several structures, such as the pleural bulla (a large swelling at thebase of the abdomen, visible externally). Richards (1932) was uncertain whether the tympanal hood of the nolids was pre- orpostspiracular, but Kiriakoff (1958) demonstrated the latter to be the case. He concurred withRichards and placed the nolids as part of the noctuid evolutionary line. The final groups to be studied (Kiriakoff, 1960) were the herminiines, hypenines andrivulines. He agreed with Richards' distinction of the Herminiinae on the basis of the position ofthe tympanal hood but could not find any specific characters to allow either the separation of theHypeninae from the Rivulinae or the differentiation of these two subfamilies from theerebine-catocaline complex. Characters such as the double pocket IV were considered to beunreliable because they 'occur in all the groups indiscriminantly'. As a result, and contrary toRichards, the Herminiinae rather than the Rivulinae were interpreted as the most primitivenoctuids. Kiriakoff (1963) eventually went even further. He proposed a classification of the Noctuoideabased upon the 'reduced rank' system of Janse (1937-9). Endrosa [Setina], a lithosiine with anunusual tympanum, was given family status in the Noctuoidea, equivalent in rank to a moreinclusive Noctuidae. This family was divided into two subfamilies, the Arctiinae and theNoctuinae. The former, characterised by a prespiracular hood and a noctuid-type tympanum,also contained, in addition to the infra-families Arctiidi and Lymantriidi, the Herminiini. Withinthe Noctuinae, three infra-families were recognised: the Nolidi, the Noctuidi and the Agaristidi. Kiriakoff considered his classification superior to older schemes 'because the tympanicstructures undoubtedly are the most significant single set of characters that can be found in theLepidoptera'. However, this confidence was not shared by other lepidopterists, who treated thesingle character complex-based classification with scepticism. It was never adopted as a seriousalternative. Recent developments in the U.S.A. The latest higher classification of North American Noctuidae was presented by Franclemont &Todd (1983) (Table 17). Contrary to both Richards and Forbes, they consider the Herminiinaeto be an advanced group of quadrifine noctuids, while the subfamilies closest to the ancestralstock of the Noctuidae are the Sarrothripinae and Acontiinae: Eustrotiini. Franclemont alsorecognises the very close similarity between the trifine subfamilies and would possibly advocatetheir amalgamation into a single subfamily, the Noctuinae. At most, perhaps two subfamilies REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 205 Table 17 The most recent classification of the Noctuidae, employed by Franclemont & Todd (1983).The 'Unassociated genera' of Amphipyrinae are those the authors did not assign to a particulartribe. NOCTUIDAE Herminiinae Rivulinae Hypenodinae Hypeninae Catocalinae Plusiinae Abrostolini PlusiiniEuteliinaeSarrothripinae Risobini SarrothripiniCollomenini NolinaeAcontiinae Cydosiini Eustrotiini Eublemmini Acontiini Bagisarini Pantheinae Acronictinae AcronictiniBryophilini Agaristinae Amphipyrinae Apameini Amphipyrini Stiriini Cuculliinae Hadeninae Noctuinae Heliothinae Nocloini'Unassociated genera' Xylenini Feraliini Psaphidini Oncocnemidini Cuculliini Hadenini Eriopygini Glottulini AgrotiniAnicliniNoctuiniUfeini [Heliothini]Grotellini could be recognised, the Noctuinae and the Acronictinae, the latter also including thepantheines. Deltoids and Catocalinae. The deltoid subfamilies (Herminiinae, Hypeninae and Hypenodi- 206 IAN J. KITCHING nae) are largely unaltered, but there are many discrepancies with Forbes' classificationregarding the remainder of the family. The Rivulinae represent the third miscellaneous series ofForbes' Erebinae, together with one genus from the second and three others. The remainingcatocalines and ophiderines are combined into one large subfamily, the Catocalinae, which isnot subdivided. Plusiinae. The division of this subfamily follows Eichlin & Cunningham's (1978) revisionalthough their Argyrogrammini and Autographini are merged to form one tribe, the Plusiini.Phyprosopus is transferred to the Catocalinae. Euteliinae. This small subfamily is unaltered. Sarrothripinae. Three tribes are recognised in this group. The Risobini comprises onlyBailey a; the Sarrothripini contains Characoma, Nycteola and Iscadia; while the Collomeniniincludes Motya and Collomena (two genera not considered by Forbes). Nolinae. Franclemont & Todd include this erstwhile arctiid group in the Noctuidae and thusfinally make one of the moves, which although mooted for decades, no-one had previously daredto perform. Acontiinae. In addition to the three tribes (Eublemmini, Eustrotiini (Erastriini in Forbes,1954) and Acontiini) previously employed, two more are also considered. The Cydosiiniincluded only Cydosia, briefly discussed by Forbes under Xanthoptera [Thioptera], a genus ofthe Erastriini, while the Bagisarini included only Bagisara. This genus had been variouslytreated as an amphipyrine and an acontiine (the suggestion that it may belong in the lattersubfamily was first made by Heinrich, 1926). Whereas Forbes treated Bagisara as an amphipy-rine (although as an 'isolated genus'), Franclemont & Todd consider it to be an acontiine. Pantheinae. This is the last quadrifine subfamily listed. The Pantheinae are removed fromtheir Hampsonian position (before the Plusiinae, a placing followed by Forbes) and put next tothe trifine subfamily Acronictinae. It would thus seem that the classification concerning therelative positions of these two groups has come full-circle. Acronictinae. This subfamily is interpreted in the strict sense and is divided into two tribes, theAcronictini (Acronicta, Simyra and their allies, Forbes' Apatelini) and the Bryophilini, contain-ing only Cryphia, a genus unconsidered by Forbes. Agaristinae. Despite Mosher (1916) and Turner (1920), only lip-service had been paid to thehigh degree of similarity between this group and the Amphipyrinae. The inclusion of theagaristines as a subfamily of the Noctuidae was, like the Nolinae, the first occasion on which thishad occurred in a major checklist. It is hoped that the false link with the Arctiidae and theunnecessary family rank have finally been laid to rest. Amphipyrinae. The largest trifine subfamily is divided into four tribes followed by a largegroup of 'unassociated genera'. The Apameini is approximately equivalent to the tribescomprising Forbes' 'first series' combined, but also includes some of the 'isolated genera'. Theclassification of the Stiriini follows Hogue (1963) and corresponds to Forbes' Stiriini except forthe removal of Stiriodes to the Amphipyrini. Seven genera constitute the Nocloini, none ofwhich were treated by Forbes. The Eudryini are transferred to the Agaristinae (which Forbesconsidered they resembled) while the Pyrrhiini are incorporated into the Heliothinae. Of theremaining 55 genera, 30 form the Amphipyrini and 25 are 'unassociated'. Cuculliinae. This subfamily is divided into five tribes. The Xylenini represents the combina-tion of Forbes' Cleocerini, Lithophanini and Antitypini. Feralia is removed from the Psaphidini REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 207 to a tribe of its own (the Feraliini). This former tribe, and the Oncocnemidini and the Cuculliini,are largely unchanged. Hadeninae. Unlike Forbes, Franclemont & Todd divide this group into three tribes: theHadenini, the Eriopygini and the Glottulini. However, these groups do not correspond toForbes' informal sections. The Eriopygini consists of a series of genera related to Orthodes,Tricholita and Ulolonche, while the Glottulini contains only Xanthopastis , a genus about whichForbes was unsure. Noctuinae. The heliothines are excluded from this subfamily and the remaining genera aresplit among four tribes. The first three, the Agrotini, the Aniclini and the Noctuini, correspondto the first, second and third sections respectively into which Forbes divided the subfamily. Thefourth tribe, the Ufeini, contains only the aberrant genus Ufeus. Heliothinae. This subfamily comprises two tribes, most genera being included in the presumedHeliothini (the name is omitted). Following Hardwick (1970), Forbes' acronyctine tribe, thePyrrhiini are also assigned here. The second heliothine tribe, the Grotellini, contains only Grotella, Hemigrotella and Neo-grotella, none of which was considered by Forbes. These genera were omitted fromthe Heliothidinae [Heliothinae] by Hardwick (1970), who suggested that they belongedin the Stiriinae [Stiriini] near Stiriodes (a genus Franclemont & Todd place in the Amphipy-rini). This then is the latest classification of the Noctuidae. It was produced from the results of manystudies and is highly commendable. However, it must only be considered a starting point and itremains to be seen how well it will stand up in the face of future studies of noctuids from otherparts of the world. What noctuid systematists outside North America cannot afford to do is tobury their heads in the Hampsonian sand and ignore this higher classification. Novel character complexes Scent brushes and hair pencils The presence of various brushes, hair pencils and coremata, presumed to be associated withdisseminating pheromones, have long been known in the Noctuidae (e.g. Pierce, 1909). Varley(1962) reviewed the structure and function of the brush organs of a number of noctuids and,unlike Pierce (1909), considered these organs to be potentially of great taxonomic value andurged further investigation. Such a study was eventually carried out by Birch (1972), who described the scent-brushes of anumber of trifine groups. In a second paper, Birch (1972ft) dealt with the relationship betweenchemistry and taxonomy. The brush-organs were found to be composed of five separate structures. Birch consideredthere to be a particular order in which these components were lost and from this evidenceconcluded that the brush-organs of the trifine noctuids had arisen only once and that if a specieslacked them, then it must be because of subsequent loss. The Acronyctinae [Acronictinae] werealso found to lack these organs but in their case, Birch believed that this was due to their havingsplit from the trifine lineage before the brush-organs first arose. On the basis of the most variedstructure, the Cuculliinae were considered to have diverged next, followed closely by theNoctuinae. Most genera of the latter subfamily then lost their brushes. The apex of thephylogenetic tree was composed of the hadenines and the amphipyrines, which were impossibleto subdivide. Hence, despite being extremely critical of the Hampsonian subfamilies, Birch,too, eventually had to use them in his final analysis. He therefore avoided the most controversialaspect of his work, that is, that the brush-organs could have suggested groupings of genera thatwere considerably at variance with those currently accepted. 208 IAN J. KITCHING Compound eyes In an impressive study of the lepidopteran compound eye, Yagi & Koyama (1963) examinedseveral noctuid species. Three groups of genera were recognised. Group 1 comprised catoca-lines and ophiderines only, while group 2 included five trifines, a plusiine and a hypenine. Thespecies in these two groups are all nocturnal. The third group comprised the diurnally-activegenus Hyblaea. On the basis of eye structure, Yagi & Koyama considered the hyblaeines (andalso the agaristines) ought to be separated as distinct families (see also below). This study, although superficial from the point of view of the Noctuidae, provided yet anothermeans by which the higher classification of the Noctuidae might in future be tested. A variation on the Sarrothripinae Most workers in noctuid systematics recognise the great similarity between the Sarrothripinaeand the Chloephorinae, in such features as the bar-shaped retinaculum (Hampson, volume 11),pupae (Gardner, 19486) and tympanal organs (Richards, 1932). On several occasions, they hadeven been placed together in a separate family, the Hylophilidae (e.g. Staudinger & Rebel,1901). Within the Hampsonian framework, the two groups were regarded as subfamilies of theNoctuidae, distinguished from each other on the basis of the presence or absence of tufts ofraised scales in the fore wing cell. However, Mell (1943) considered the genera comprising the Sarrothripinae and the Wester-manniinae [Chloephorinae] not to be sufficiently distinct to be worth regarding as separatesubfamilies. Nor did he consider that they warranted even tribal status. Instead, Mell divided themore-inclusive Sarrothripinae into eight tribes and one species group, thus: 1, Eligmini - Eligma, Selepa, Triorbis, Gadirtha, Lampothripa; 2, Sarrothripini - Sarrothripus [Nycteola], Bryophilopsis; 3, Risobini-Risoba; 4, Blenini- Blenina; 5, Hylophilini - Earias, Hylophilodes, Clethrophora, Hylophora [Pseudoips], Chloephora; 6, Cymatophoropsiini - Cymatophoropsis; 7, Ariolicini - Sinna, Gabala, Siglophora, Ariolica; 8, Species-group Tatothripa-Tympanistes; 9, Careini - Cam*, Nertobriga. Of these, 2-4 contain only sarrothripines; 5, 7 and 9 only chloephorines; 1 is mainlysarrothripine with one chloephorine (Eligma); 8 also contains representatives of both(Tympanistes is the chloephorine); while 6 contains a genus currently assigned to the Ophi-derinae. Unfortunately, no comparison is possible between the tribes of Mell and those ofFrancelement & Todd (1983) because there are no genera held in common except Sarrothripus[Nycteola]. This expanded concept of the Sarrothripinae was employed by Aubert & Boursin (1953),although the impoverished European fauna resulted in only two tribes being required, which,perhaps conveniently, corresponded to the previously accepted groups of the Sarrothripinae (asSarrothripini) and the Chloephorinae (as Benini, = Hylophilini sensu Mell). Probably becausethe genera are mostly tropical, Mell's divisions have been largely ignored, and their usefulness,if any, must await a world- wide reappraisal of the group. Mell (1943) also erected a new noctuid subfamily, the Camptolominae. This contained onlyCamptoloma, a genus that had until then been considered to be arctiid. Mell included it in theNoctuidae on the basis of fore wing venation, and considered it close to the hylophilinesarrothripines because of similarities in the genitalia and immature stages. Inoue & Sugi(1958-61) considered the genus to be worthy of a separate family, while Nye (1975) and Inoue etal. (1982) place it in the Arctiinae. The systematic position of this genus remains in doubt butInoue (pers. comm. to A. Watson) believes it to be more closely related to the Arctiidae than tothe Noctuidae. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 209 The removal of Hyblaea The hyblaeines are small moths that superficially resemble stocky tortricids or miniaturehypocalas (Noctuidae: Ophiderinae). On the basis of the latter resemblance, the group had longbeen considered to belong to the Noctuidae. Its position, as a subfamily of the Noctuidae, wasvirtually stabilised when it was accepted by Hampson. This was in spite of the well-developedmaxillary palps, a feature unique in the Noctuidae. The edict was first challenged by Comstock (1925) who considered the hyblaeines to be moreclosely related to the Thyrididae and placed the group accordingly after the pyraloids. The pupawas examined by Forbes (1933) and the supposed pyraloid affinities strengthened. A thorough morphological study of the adult teak moth, Hyblaea puera, was performed by DeKoning & Roepke (1949). They rejected the noctuid status of the species on the basis of: 1, absence of tympanum; 2, presence of three-segmented maxillary palps; 3, presence of an unusual secondary sexual organ on the male hind-tibia; 4, a tridentate uncus. This last character agreed with the Thyatiridae (Pierce, 1909) but De Koning & Roepke couldnot include Hyblaea in this family because of venational differences. They also disagreed withForbes (1933) contention that Hyblaea was pyraloid, although they were unable to suggest analternative position. Kiriakoff (1953) also noted a tympanum to be lacking in Hyblaea and therefore could notinclude the genus in either the Noctuoidea or in the Pyraloidea. Kiriakoff also reappraised theother structural characters of the Hyblaeidae. Although the family had much in common withthe Tortricoidea, he eventually decided it represented a new superfamily, the Hyblaeoidea,related to the Notodontoidea and the Noctuoidea ('dans la sous-cohorte des Noctuiformes') butwhich had diverged from the main stem before the development of the thoracic tympanal organin the latter groups. Later (Kiriakoff, 1963), he changed his mind, placing the Hyblaeidae in theTortricoidea, and concluding that the family had nothing to do with the Noctuiformes after all.Despite this, most authors continued to regard the hyblaeids as either a family near theNoctuidae (e.g. Inoue & Sugi, 1958-61; Yagi & Koyama, 1963) or as a subfamily thereof (e.g.Viette, 1961). Brock (1971) also excluded the hyblaeids from the Pyraloidea, including them in a newsuperfamily, the Aegerioidea, together with the Aegeriidae and the Dudgeoneidae. Pinhey(1975) noted the transfer but retained the hyblaeids at the head of the Pyraloidea. Although Common (1975) criticised some of Brock's conclusions, the Hyblaeidae were notmentioned. However, Nye (1975) considered the Dudgeoneidae and the Hyblaeidae aspyraloid. Until the early stages are better known, the exact position of the Hyblaeidae remainsdoubtful. Diloba: full circle - and back again? The early history of the aberrant species Diloba caeruleocephala was briefly discussed above.Several shifts in its systematic position had left it in the Plusiinae (Hampson, volume 13). Richards (1932) examined the tympanum and found it to be typical of the Acronyctinae[Acronictinae -I- Amphipyrinae] and particularly similar to that of Trachea. Beck (1960) alsoincluded Diloba in the Apatelinae [Acronictinae] although the larva was aberrant in no less thansix characters. However, most authors continued to consider it as either a pantheine (e.g.Aubert & Boursin, 1953; HesLop, 1960) or as a plusiine (South, 1961). Pierce & Beirne (1941)alone placed the species in the Lymantriidae, on the basis of the male and female genitalia. Kiriakoff (1970) disagreed with Richards' (1932) conclusions, considering the tympanal organof Diloba to be distinctly notodontid. However, Diloba differed from the Notodontidae inpossessing 'a paired abdominal bulla operculated by the counter-tympanum, not unlike thestructure found in the family Agaristidae' [Agaristinae]. Thus Kiriakoff found it necessary toplace Diloba in the monobasic family, Dilobidae. 210 IAN J. KITCHING Tarns (quoted in Varley, 1962, and pers. comm. to I. W. B. Nye, 1970) believed Diloba to berelated to the cuculliine Psaphida. Minet (1982) reappraised the previous work on thetympanum and concluded that Richards (1932) had been correct in his interpretation and thatKiriakoff (1970) had been wrong. In addition, Minet (1982) suggested that Diloba fittedperfectly into the Cuculliinae (lashed eyes, obsolescent M 2 on the hindwing, cf. Hampson) andthus supported Tarns' contention. Additional characters from the immature stages were used tosupport the inclusion of Diloba in the Noctuidae: eggs with radial ribs from the micropyle; larvawith a prothoracic gland produced 'vertically' (as opposed to 'transversely' as in the Notodonti-dae) and having only a single pair of MD setae in the last larval instar (although this lastcharacter is plesiomorphic and therefore 'n'est pas significatif). Minet did not apparentlycompare the genitalia of Diloba with those of cuculliines such as Psaphida. Overall, however, the current consensus places Diloba either in the Notodontidae (a positionfollowed by Kloet & Hincks, 1972; Bretherton et al., 1979; Nye, 1975) or as a separate family(followed by Watson etal., 1980; Leraut, 1980). It would thus seem that, like the Hyblaeidae, the systematic position of Diloba is open toquestion yet. However, unlike the former, the problem of the affinities of Diloba is still firmlywithin the field of noctuid systematics. The neglected arrangementPreamble By the 1920s, Hampson's system seemed to be the final word in noctuid higher classification.However, even before the dissatisfaction of workers such as Forbes set in, not all taxonomistswere happy to accept the Hampsonian subfamilies. In particular, the amateur English lepi-dopterists rejected them, preferring to follow Meyrick, or even Guenee. After all, a newarrangement would have meant them re-ordering their collections. Eventually a classificationarose that evolved into an arrangement of subfamilies which, although currently almostforgotten, bears much in common with the divisions of the Noctuidae employed recently byFranclemont & Todd (1983). The demise of the systems of Guenee, Staudinger & Rebel and Tutt The early part of the twentieth century saw the virtual disappearance of the noctuid classificationof Guenee (1852-4) and those derived directly from it. In his Moths of the British Isles, South (1908-9) followed the division of the Noctuidaeemployed by Staudinger & Rebel (1901). This survived until the revision by Edelsten & Fletcher(South, 1961), when the classification of Hampson was substituted. Following the death of Tutt in 1911, a supplement to his British Noctuae was produced byTurner (1926-48). Although he discussed the classifications of the noctuids used by Hampsonand in 'Seitz', in order to be consistent with Tutt's earlier work, Turner employed thearrangement used therein, namely that of Guenee. Thus Tutt's classification of 1902 fell intodisuse. It was eventually followed by that of Guenee; Turner's usage was its last. The English 'amateur' system Origins In 1907, Kirby published his Butterflies and Moths of Europe. In this, he acknowledged theclassification of Staudinger & Rebel (1901) but preferred to use one of his own. This divided theLepidoptera into 'macros' and 'micros'. The former was then further split into five 'superf ami-lies': Rhopalocera, Sphinges, Bombyces, Noctuae and Geometrae. Although Kirby removedthe Cymatophoridae [Thyatiridae] from the Noctuae to the Bombyces (this heterogeneousgroup also contained the Nycteolidae [Sarrothripinae] and Chloephoridae [Chloephorinae]),the Noctuae still included the Brephidae [Geometridae: Archiearinae]. The remaining noctuids REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 211 were distributed among 18 subfamilies (Table 18), which appeared to have something incommon with those of Herrich-Schaffer (1845) and part with those proposed by Grote in the1880s and 1890s (see Tables 5 and 7), but also included some apparent influence from Hampson. The Bombycoidae, a name which in various forms had previously been used to refer to severalgroups, here contained Diloba and the pantheines. These were followed by the Acronyctidae[Acronictinae] which, although it included Bryophila [Cryphia], omitted Arsilonche [Simyra]and Simyra. These last two genera were placed at the head of the subsequent group, theOrthosidae. This subfamily, which had previously been regarded as five (Leucaniidae, Caradri-nidae, Orthosiidae, Cerastidae, Amphipyridae; Herrich-Schaffer, 1845), consisted mainly ofhadenines, cuculliines (e.g. Agrochola) and amphipyrines, with minor representation from theAcronictinae (e.g. Simyra) and Ophiderinae (Scoliopteryx). The Agrotidae, comprising mainly Agrotis (s.l.) and Triphaena [Noctua], was broadlyequivalent to the Noctuinae, although it did include Brithys, a hadenine. The Hadeninaecontained the balance of the hadenines and amphipyrines. There followed several small subfamilies. The Xylinidae (e.g. Xylina [Lithophane], Calocam-pa [Xylena]), Cleophanidae (e.g. Calophasia, Cleophana [Calophasia, part]) and Cuculliidae(Cucullia only) contained most of the present Cuculliinae, although the first also includedXylomiges [Egira] (Hadeninae). The Euteliidae (cf. Eurhipiidae in the introduction to Kirby, 1907, and Table 18) and theHeliothidae corresponded to the present subfamilies of the same names (the latter includedPyrrhia and its relatives). The next subfamily, the Anartidae, conformed to the residue ofWarren's (1906-14) Heliothidinae after the removal of the true heliothines and it may be thatWarren took over the concept from Kirby and expanded it. The Anartidae included representa-tives of the Amphipyrinae (Euterpia [Enterpia], Heliaca [Panemeria]) , Noctuinae (Cyrebia),Hadeninae (Anarta) and Cuculliinae (Omia). The Plusiidae included Telesilla [Eucarta], as well as the usual genera, Abrostola and Plusia(s.l.), while the Calpidae contained only Calpe [Calyptra]. The Acontiidae, including Acontia[Tyta], Armada [Tarachepia] and Acontiola [Eustrotia and Ozarba], was a mixture ofophiderines and acontiines. Most of the remaining catocalines and ophiderines were dividedbetween the Catocalidae (e.g. Catocala, Zethes, Aedia) and the Ophiusidae (e.g. Apopestes,Ophiusa [Lygephila]). These two groups also contradicted those that Kirby, in his introduction(Table 18), said he was going to use (Ophiusidae and Toxocampidae). Table 18 The subdivisions of the Noctuae [Noctuidae] employed by Kirby (1907). NOCTUAE Bombycoidae Acronyctidae Orthosidae Agrotidae Hadenidae Xylinidae Cleophanidae Cuculliidae Eurhipiidae Heliothidae Anartidae Plusiidae Calpidae Acontiidae Ophiusidae Taxocampidae Noctuophalaenidae Deltoidae Brephidae 212 IAN J. KITCHING The Noctuophalaenidae included the rest of the acontiines and the ophiderine, Rivula. TheNoctuae finished with the Deltoidae (including the ophiderines Laspeyria and Parascotia) andthe Brephidae. Development The arrangement of Kirby (1907) formed the basis of the system adopted by Heslop (1945) in hischecklist of British Lepidoptera. It should be noted that Heslop, like many other earlier authorsof checklists and catalogues, employed no new research, relying entirely upon the studies ofothers. His arrangement of the various groups was probably governed as much by personal tasteas it was by explicit characters. Whatever his reasons, Heslop, again like many of hispredecessors and several workers subsequently, never published his argumentation, thusrendering it virtually immune from objective criticism. The noctuids were classed in the superfamily Agrotides and, following Meyrick (1928),divided into two families, the Caradrinidae and the Plusiidae. Ten subfamilies were recognised in the Caradrinidae. Eight corresponded to those of Kirby:Mominae (Kirby 's Bombycoidae), Acronictinae (Acronyctidae), Hadeninae, Caradrininae(Orthosiidae), Xyleninae (Xylinidae and also including the Cleophanidae), Cuculliinae, Anarti-nae and Heliothinae. The Agrotidae was split into two groups, the Agrotinae and theTriphaeninae. The Euteliinae have no British representatives. The Plusiidae contained five subfamilies. The Catocalidae and Ophiusidae were combinedunder the Catocalinae, while the Acontiidae and Noctuophalaenidae merged to become theEustrotiinae [Acontiinae]. The Deltoidae became the Hypeninae while the Plusiinae remainedunaltered. Scoliopteryx was removed from the Orthosiidae and placed in a separate quadrifinesubfamily, the Gonopterinae. The Brephidae, although listed after the Hypeninae, had beenmoved across to head the Geometridae as a subfamily of the Monoctenidae. In common withKirby, Heslop placed the Westermanniinae [Chloephorinae] and Sarrothripinae, as subfamiliesof the Hylophilidae, in the Bombyces. Several changes were evident in the classification employed by Heslop (1960) in his revisedchecklist. Some subfamilies had been renamed as a result of considerable reassessment of theapplication of generic names that had been performed in the interim. The superfamily was now known as the Noctuoidea, and the Caradrinidae as the Noctuidae.Similarly, the Triphaeninae had become the Noctuinae. The Hadeninae and Caradrininae hadundergone extensive reassortment and had been divided into no less than seven subfamilies: 1, Hadeninae - those currently recognised hadenines, less the next two groups; 2, Orthosiinae - Orthosia and Panolis; 3, Leucaniinae - the hadenine wainscots; 4, Nonagriinae - the amphipyrine wainscots and Stilbia; 5, Caradrininae - Caradrina and its close relatives, Meristis and Laphygma [Spodoptera, part]; 6, Apameinae -Apamea, Luperina and their allies, including also Prodenia [Spodoptera, part] ; 7, Amphipyrinae - the remaining amphipyrines, e.g. Hydraecia, Cosmia, Mormo. Two series, 1-3 and 4-7, can be recognised as equivalent to the present Hadeninae andAmphipyrinae respectively. The Acronictinae became the Apatelinae and finally, the cuculliine section of Kirby'sCaradrinidae (e.g. Antitype, Agrochola, Cirrhia) was removed and established as the Dasy-poliinae. The second family of noctuoids was the Hylophilidae, finally recognised by Heslop ascorrectly belonging with the noctuids, and the third was the Plusiidae. The subfamilies of thelatter group now largely agreed with those of Hampson, except that the Gonopterinae wasretained. Consequently, the Pantheinae was moved into the Plusiidae from the Caradrinidae,and Parascotia and Laspeyria were transferred from the Hypeninae to the Ophiderinae. Apart from combining the Xyleninae and the Dasypoliinae under the former name, thearrangement of Heslop (1960) was faithfully followed by Chalmers-Hunt (1962-8). REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 213 Was Heslop on the right track? The most fascinating aspect to Heslop's (1960) classification of the Noctuidae is the degree towhich it parallels that recently employed by Franclemont & Todd (1983). The two are comparedin Table 19, demonstrating the considerable concordance, especially in the trifine subfamilies.There are, however, several discrepancies, which ought to be discussed further. The British deltoid fauna is impoverished relative to that in North America (13 species,excluding Laspeyria, as opposed to 82) and so it is not perhaps surprising that Heslop did notconsider it necessary to divide the Hypeninae, particularly as the study of the tympanal organhad not been influential in British noctuid systematics. Table 19 A comparison of the higher classifications of the Noctuidae proposed by Heslop (1960) andFranclemont & Todd (1983). Heslop Franclemont & Todd PLUSIIDAE Hypeninae Ophiderinae CatocalinaePlusiinaeEustrotiinaePantheinae [Nolidae (Bombycoidea)]HYLOPHILIDAE Sarrothripinae WestermanniinaeNOCTUIDAE Apatelinae CuculliinaeXylenini j Dasypoliinae IAmphipyrinae JCaradrininae IApameinae j Nonagriinae ILeucaniinae \Orthosiinae I HadeninaeAnartinae / HeliothinaeNoctuinaeAgrotinae NOCTUIDAE Herminiinae Hypenodinae Hypeninae Rivulinae Catocalinae (part) Catocalinae (part) Plusiinae Acontiinae Pantheinae Euteliinae Nolinae Sarrothripinae AcronictinaeAgaristinaeCuculliinae: Cuculliini Cuculliinae: XyleniniAmphipyrinae: AmphipyriniAmphipyrinae: Apameini Hadeninae: Hadenini Heliothinae: HeliothiniNoctuinae Heslop also retained Hampson's division into Catocalinae and Ophiderinae, including in thelatter, the Rivulinae, another subfamily recognised mainly on the basis of tympanal characters.Heslop's reasoning probably coincided with that of Forbes (1954) on the same subject.However, Heslop did separate Scoliopteryx into the Gonopterinae. Franclemont & Toddincluded this genus in the Catocalinae. The Plusiinae, Acontiinae, Sarrothripinae and Acronic-tinae were recognised by both systems in equivalent terms (allowing for Heslop's inclusion ofTelesilla [Eucarta] in the Plusiinae and his separation of the sarrothripines, together with theWestermanniinae [Chloephorinae], into the family Hylophilidae). Within the trifines, the amphipyrine genera were divided into four subfamilies by Heslop. Of 214 IAN J. KITCHING these, the Apameinae and the Nonagriinae, are approximately equal to Franclemont & Todd'sApameini, although this tribe also includes Hydraecia, a genus Heslop placed in the Amphipyri-nae. The second two of Heslop's subfamilies, the Amphipyrinae and the Caradrininae, roughlyequate to Franclemont & Todd's Amphipyrini, with a few exceptions (e.g. Prodenia [Spodop-tera] litura, which is in Heslop's Apameinae). The cuculliine genera were placed by Heslop into three subfamilies. The Cuculliinae isequivalent to Franclemont & Todd's Cuculliini, while the Dasypoliinae and Xyleninae com-bined approximate their Xylenini. Interestingly, iiJodia and Eupsilia are incorporated from theXyleninae, the Dasypoliinae equates to Forbes' (1954) Antitypini. The residual xylenines arethen equivalent to Forbes' Lithophanini. Heslop divided the hadenines into four subfamilies, all of which are included in Franclemont& Todd's Hadeninae: Hadenini. The two concepts of the subfamily Heliothinae agree althougha more accurate equivalence would be between Heslop's Heliothinae and Franclemont &Todd's Heliothinae: Heliothini. There are no North American representatives of those genera (Mesogona, Euschesis [Noctua,part], Noctua and Lampra [Noctua, part]) that Heslop placed in the Noctuinae, and thus hisAgrotinae equates with the Noctuinae of Franclemont & Todd. However, on nomenclaturalgrounds, the Noctuinae (or at least Noctua) would belong in the Noctuini of Franclemont &Todd, which also includes several agrotines sensu Heslop. The two systems are therefore indisagreement. Forbes (1954) had already noticed this discrepancy between the European andAmerican systems and considered the reason to be that 'European tradition has treated thewhole group [Noctuinae] as a single genus, except a few species (not in fact closely related toeach other) with yellow hindwings. There is no agreement [with his classification] as to thissubdivision'. However, overall and considering that the two systems are based upon more or less exclusivefaunas, the classifications of Heslop (1960) and Franclemont & Todd (1983) are remarkablysimilar. It seems likely that a reassessment of the European fauna along the lines of that of NorthAmerica would result in relatively few changes in subfamily/tribal placings with respect toHeslop's groupings. The inescapable conclusion is that the English amateur lepidopterists, inrefusing to accept Hampson's system, and choosing to follow Kirby and Heslop instead, weremore correct than perhaps they imagined. Quo vadis, Noctua!Introduction That the higher classification of the Noctuidae is in disarray and in need of extensive re-evaluation is probably the only point on which all workers in the field agree. The Hampsoniansystem still maintains a considerable influence, despite many efforts to loosen its stranglehold,and in this respect very little advance has been made since 1920. However, current changes inattitudes probably mean that the days of the Hampsonian system per se are numbered, althoughthere remains a vast amount of work to be done before an adequate replacement can beproposed. To this end, I will first consider the individual subfamilies, illuminating those areas where Iconsider future research could be most usefully directed. Secondly, I present an outline cladisticanalysis of the higher classification of the Noctuidae. This should not be considered as a 'newsystem', but merely as the first step towards the production of one. I do not expect everyone (oreven anyone) to agree with all of my interpretations and conclusions, but I hope the result will beto generate renewed interest in noctuid higher systematics. The individual subfamilies Arctiidae: Aganainae. This group of moths has been consistently placed in the Arctiidae orelse treated as a separate family closely related to the arctiids. Its importance regarding thehigher systematics of the Noctuidae lies in its possible future transfer to that family. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 215 A relationship between the Noctuidae and the Aganainae was first tacitly suggested byGardner (1941) on the basis of larval characters. The two current aganaine species that heexamined, Hypsa [Asota\ alciphron and H. [Psephea]ficus , agree with the Noctuidae in having asingle subventral seta on the meso- and metathorax (but see below). This contrasts with thebisetose condition found in the Arctiidae (the third 'aganaine' examined by Gardner, Digamahearseyana, which has a bisetose SV group, is currently placed in the Arctiinae). However,Gardner refrained from uniting the Noctuidae and Hypsidae [Aganainae], mainly because thelatter lacked the ventral prothoracic gland found in the larvae of the former. The discovery of the tymbal organ in the Arctiidae (Forbes & Franclemont, 1957; Blest et al. ,1963) provided a potential apomorphy by which the family could be characterised. However, itsoccurrence is not universal. The structure is conspicuously absent in the aganaines and a fewwasp-mimicking ctenuchines (A. Watson, pers. comm.). It can be argued that the loss of thetymbal organ (and also the tympanal organ in a few species) in the latter group is concurrent withthe general reduction and high degree of modification of the thoracic sclerites that has occurredin order to produce the hymenopteran facies. Tymbal loss cannot be so convincingly argued forin the aganaines, especially as it is lacking in all genera and not just a few as in the Ctenuchinae.Thus, if the Arctiidae were to be redefined, employing the presence of the tymbal organ as asynapomorphy, then the Aganainae would have to be excluded. Franclemont & Todd (1983)considered the aganaines to be 'probably an aberrant group of Noctuidae'. However, theaganaines have a pre-spiracular tympanal hood (Richards, 1932), and for this reason, I wouldchoose to exclude them also from the Noctuidae (see Herminiinae below). The net result ofthese character interpretations would be the reinstatement of the Aganaidae as a separatefamily. Herminiinae. Recognised as a distinct subfamily of noctuids by several pre-Hampsonianauthors (e.g. Herrich-Schaffer, 1845; Grote, 1890; Smith, 1895), the Herminiinae were consi-dered by Hampson as part of the Hypeninae. Forbes (1918) resurrected the subfamily on thegrounds of the possession of a pre-spiracular tympanal hood but the exact position of theHerminiinae within the Noctuidae remained unsettled. The primitive position of the hood was used by Kiriakoff (1963) to argue for the exclusion ofthe Herminiinae from the Noctuidae and their placement within a more-inclusive 'Arctiinae'.This conclusion, however, is disputed. Richards (1932) did not consider that the Herminiinaewere the most primitive noctuids, on the grounds that the 'basal group' of genera (Paraherminia[Paracolax] and Dercetis [Redectis]} , which were determined as the most primitive herminiineson other characters, had the spiracle 'slightly under (ventro-anterior to) the greatly reducedhood'. Thus, Richards implicitly treated the pre-spiracular hood in the herminiines as acharacter reversal, a position supported recently, without further elaboration, by Franclemont& Todd (1983). I remain unconvinced and maintain that the herminiine pre-spiracular hooddoes represent the plesiomorphic state . If this is accepted , then , in the absence of other evidenceto the contrary, the subfamily must be considered to form the sister-group of the remainingnoctuids, and represents an analogous situation to that between the Arctiidae and Aganainae. Ifthe Noctuidae were to be defined by the possession of a post-spiracular hood, the herminiineswould have to be excluded. Characterisation of the resultant 'Herminiidae' would be possible onthe basis of the swollen metepimeron ventral to pocket IV and perhaps the modifications of theforelegs and antennae in the males. The results of a cladistic analysis under the alternative interpretation of polarity are lesssatisfactory. Following the strict cladistic approach to loss characters (Patterson, 1982),reversion of the tympanal hood to a pre-spiracular position could not be used to justify themonophyly of the Herminiinae and would also cast doubt on the use of the post-spiracular hoodto characterise the family Noctuidae. If this latter character state was then rejected, we shouldbe left in the extremely unsatisfactory position of having a family of 25,000 species completelyuncharacterised. Cladistics aside, however, there is little doubt that a pre-spiracular hood is a good diagnosticcharacter for recognising the herminiines (assuming that they can be differentiated from the 216 IAN J. KITCHING aganaines). Their separation from the other 'deltoid' groups has so far only been performed forthe North American fauna. Rivulinae. First proposed by Richards (1932), and then also including the genera currentlyassigned to the Hypenodinae (q.v.), the Rivulinae was interpreted as being the most primitivenoctuid subfamily. However, it was also the most difficult for Richards to characterise and thusits status is highly questionable. Forbes (1954) rejected the Rivulinae, referring most of itsgenera to the Erebinae, as his third miscellaneous series. Likewise, Kiriakoff (1960) consideredthe Rivulinae to be insufficiently differentiated and placed both it and the Hypeninae in theCatocalinae-Erebinae complex. Recently, only Franclemont & Todd (1983) have employed thegroup. At present, there appear to be no good characters on which to base the Rivulinae. Within thesubfamily, Richards included those 'primitive' noctuids that could not be placed in either theHerminiinae or Hypeninae. The rivuline genera were thus characterised by a post-spiracularhood, unlashed eyes and short labial palps, characters that can all be interpreted as inapplicableat the level of universality relevant to the Rivulinae. Thus, the monophyly of the subfamily isunsubstantiated and it remains to be seen whether future studies will reveal any additionalcharacters to suggest the Rivulinae is not a non-group. Hypenodinae. Like the preceding two subfamilies, the Hypenodinae had been recognisedonly in the North American fauna. Characterisation of the group is weak. Richards (1932)included the hypenodine genera in the Rivulinae, which, given the vague nature of that group, isunsatisfactory. Forbes (1954) defined the hypenodines primarily on the lack of ocelli but, asmentioned above, such a feature is inadmissable in a strict cladistic framework. Other unifyingcharacters may exist and were hinted at by Forbes (the subfamily is 'rather homogeneousin other structures'; 1954: 381) but their exact nature is unknown. Overall, while it re-mains possible that the Hypenodinae is a real entity, at present it cannot be stated withcertainty. Further work on other faunas is necessary before a definitive conclusion can bereached. Hypeninae. The Hypeninae was characterised by Hampson primarily by the hindwing vein M 2arising from well above the lower angle of the cell and running parallel to M 3 . Under such adefinition, the subfamily also included the hypenodines and the herminiines. Richards (1932) found the Hypeninae (s.str.) to be reasonably well characterised tympanally,and that, in addition, they possessed lashed eyes. The latter state is also found in the Plusiinae,Cuculliinae and the ophiderine Scoliopteryx. The tympanum and larvae of the last genus alsoresembled those of the Hypeninae (Richards, 1932: 14). Forbes (1954) added the character stateof 'long and obliquely porrect [palps], normally twice as long as [the] head, with a rather long,porrect third segment'. Although it is probable that the Hypeninae, as restricted by Richardsand Forbes, and employed recently by Franclemont & Todd (1983), does represent amonophyletic unit (sensu Farris, 1974), more work remains to be done. Further studies of theother deltoid subfamilies are also required in order to clarify the interrelationships of these muchunderworked and neglected yet phylogenetically important noctuids. Catocalinae. The general consensus of opinion, faunal advantages notwithstanding, is thatthe division of this very large group of moths into the Catocalinae and the Ophiderinae, on thebasis of mid-tibial spining, is entirely artificial and should be abandoned. However, the result ofsuch action is a subfamily containing in excess of 10,000 named species, and probably many moreawaiting discovery and description. Subdivision of this large group is therefore necessary. Thereare homogeneous groups of genera contained within the Catocalinae (s.l.); based around, forexample, the genera Catocala, Erebus, Parallelia, Anomis and Drasteria, which can be definedby various structural features. However, the genera concerned represent only a very smallproportion of the subfamily and the work required to completely order the Catocalinae (s.l.) isimmense. Doubtless, further knowledge of the immature stages will prove invaluable - for REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 217 example, the Anomiini have malvaceous-feeding larvae; Forbes (1954: 367) - but so few arecurrently known that even their potential is largely unknown. What is required is a piecemeal dissection of the group; the removal of the homogeneousgeneric groupings, perhaps as provisional tribes; followed by studies of the numerous isolatedgenera; leading finally to a coherent system. The task is so vast as to almost deny the possibilityof success, especially when it is remembered that the Catocalinae may be polyphyletic withrespect to the Hypeninae, Rivulinae, Hypenodinae and Acontiinae. Nevertheless, the attemptmust be made, for the current state of knowledge, Richards' and Forbes' work on the NorthAmerican fauna aside, is negligible. Plusiinae. Of this subfamily, Richards (1932) stated that 'this is the most homogeneous anddistinct of all the quadrifid groups'. Certainly, the plusiines have been recognised from the veryearliest days of noctuid systematics as a natural group, but even so, their separation from theother subfamilies has been difficult. To Hampson, the plusiines were the lashed-eyed, non-deltoid, quadrifine noctuids. However, as previously noted, exceptions to this rule exist. BothScoliopteryx libatrix and Phyprosopus callitrichoides have lashed eyes, while the eyes of theplusiine genus Pseudeva 'do not appear to be lashed' (McDunnough, 1944: 213). Mosher (1916) found that the pupae of the Plusiinae showed a number of differences fromthose of other noctuids, most notably in the position of the labrum and the ventral extension ofthe wings and proboscis beyond the posterior margin of abdominal segment 4. The Plusiinaewere further characterised by Richards (1932), who discovered that its members possessed adouble tympanic hood and a swollen metepimeron formed by a greatly enlarged pocket IV. Recently, Eichlin & Cunningham (1978) proposed three tribes within the Plusiinae. Of these,the Abrostolini is the best candidate for monophyly, based on the form of the clavus in the malegenitalia. The Argyrogrammini is demonstrably paraphyletic, while there are doubts as to themonophyly of the Plusiini (Eichlin & Cunningham's 'Autographini') because it is primarilybased on a character loss (absence of prolegs on abdominal segments 3 and 4). However, this isperhaps excusable given that the authors did not propose a cladistic classification. All threetribes may yet prove to be monophyletic but the study needs to be extended to the tropicalfaunas before this can be confirmed. As to the position of the subfamily within the Noctuidae, the Plusiinae lack the proposedapomorphies of the Catocalinae (the fused pleural sclerite, J. D. Lafontaine, pers. comm.; andthe bloom on the pupa, Mosher, 1916) and so it is likely that their affinities will prove not to bewith that subfamily. It is possible that the plusiines will be demonstrated to be related to certaintrifine groups, possibly parts of the Cuculliinae, Amphipyrinae or Heliothinae (see also below),but this is presently speculative. Stictopterinae. Hampson characterised this subfamily primarily by the simplified femalefrenulum (the frequently used description 'single' is misleading, as the frenulum in femalestrictopterines often consists of more than one bristle, which are closely appressed and verydifficult to discern; A. H. Hayes, pers. comm.) and the presence of tufts of raised scales in theforewing cell. Richards (1932) considered the stictopterines to be very close to the plusiines buthe only examined two species of the genus Stictoptera (S. melanistis, from the Old World, and S.clara, a Neotropical species that should be referred to a separate genus; J. D. Holloway, pers.comm.). Both were found to possess a tympanal hood with a ventral second lobe. In S.melanistis, this lobe was large enough to give the impression of a double tympanal hood, similarto that found in the Plusiinae. The lack of stictopterines in North America has resulted in very little structural informationbeing collected for the group (5. clara probably never reaches north of Mexico and Cuba,despite a paratype of S. phryganealis [a synonym of 5. clara] in the BMNH bearing the locality'West Coast of America'. This designation of Walker's often meant the west coast of CentralAmerica, or even the Galapagos Islands; Hayes, 1975: 165-7). This situation will be partiallyalleviated in a forthcoming revision of the Bornean stictopterines by J. D. Holloway. Neverthe-less, further study is required to ascertain the interrelationships of the genera and the position of 218 IAN J. KITCHING the subfamily within the Noctuidae. It is likely that the Stictopterinae will prove to be closelyrelated to the Plusiinae, but whether it represents the sister-group of the latter remains to bedemonstrated conclusively. Euteliinae. Unlike many other noctuid subfamilies, several good apomorphies are known forthe Euteliinae. However, none of these occurs in the morphology of either the tympanal organsor the larvae. Richards (1932) found that the tympana of the euteliines could be derived fromeither the catocaline (s.l.) or the acontiine types and he was unable to place them on hisphylogenetic tree (Fig. 3). Similarly, Gardner (1948a) discovered no unifying characters in thelarvae. In contrast, he (Gardner, 19486) found a combination of characters in the pupae thatsharply defined the Euteliinae, of which the complete lack of a cremaster was the mostcharacteristic (although the cremaster is also absent in the Sarrothripinae and Chloephorinae;Mosher, 1916; Gardner, 19486). Forbes (1954) noted that the larvae were almost completelyrestricted in their foodplants to members of the Anacardiaceae (although some feed onCombretaceae and Hamamelidaceae; Mell, 1943) and that the adults typically 'rest with the forewings crumpled and partly rolled about the hind wings, and standing out obliquely to thestrongly upcurved abdomen'. The subfamily is probably monophyletic on the basis of the above characters and other, as yetundescribed, structural features (J. D. Holloway, pers. comm.). The problem of where theEuteliinae fit within the Noctuidae remains, however. The reduced female frenulum couldrepresent a synapomorphy linking the euteliines and stictopterines, but as a loss character, itdoes not form very strong evidence. Similarly, the lack of a cremaster, a feature also shared withthe Sarrothripinae, Chloephorinae and possibly the Nolinae is also weak. Otherwise, theeuteliines are very distinct and their interrelationships with other noctuid groups are far fromclear. Chloephorinae. With the transfer of Bagisara to the Acontiinae (Heinrich, 1926) or Amphipy-rinae (Forbes, 1954), and Ipimorpha to the Amphipyrinae (Forbes, 1954), the Chloephorinaeceased to be represented in the North American fauna. Consequently, little is known of themorphology of the group. Richards (1932) found the chloephorines combined the characters ofboth the acontiine tribes (Erastriini [Eustrotiini] and Tarachini [Acontiini] and believed thegroup to be derived from the higher Acontiinae. Mell (1943) considered the Chloephorinae tobe indistinct from the Sarrothripinae and treated the two groups as a single subfamily. A study ofthe larvae led Gardner (1946; 1948) to disperse the genera of chloephorines widely among hisnoctuid groups but subsequent examination of the pupae and cocoons (Gardner, 19486) causedhim to revise his decision. He finally placed all the chloephorines in a single group near thesarrothripines, although he considered Acontia [Xanthodes] to perhaps belong elsewhere. There appears to be little doubt now that the Chloephorinae is very closely related to theSarrothripinae. They share such probable apomorphies as a bar-shaped retinaculum in themales and the characteristic boat-shaped cocoon. Whether the Chloephorinae deserves sub-familial rank (based at present on the lack of the tufts of raised scales in the fore-wing cell foundin the Sarrothripinae - not only an absence but also a highly homoplasious character in thatsimilar tufts occur in the nolines and stictopterines) or merely tribal rank within the Sarrothripi-nae, or whether, as Mell (1943) proposed, the two-way division itself is artificial and should bereplaced, is as yet debatable. Sarrothripinae. This subfamily was separated from the last by Hampson on the basis of thepresence of raised scales in the forewing cell. Richards (1932) found that the Sarrothripinae andChloephorinae were tympanally very close and, like Hampson, considered the former to bederived from the latter. Gardner (19486) could find only minor differences between the twosubfamilies while Forbes (1954) suggested that the Sarrothripinae intergraded in the Old Worldwith the Nolinae. Apart from Mell's (1943) subdivision, the Sarrothripinae had always been treated as a singlehomogeneous entity. However, Franclemont & Todd (1983) divided the subfamily into three REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 219 tribes: the Risobini, the Sarrothripini and the Collomenini. Forbes (1954) considered Baileya(the sole member of the Risobini) to be totally isolated within the North American fauna andmost closely allied to the Old World Risoba. The larvae of both genera were said to resemblethose of the Euteliinae and might prove to form some sort of link between the two groups. Thejustification of the Collomenini is not clear. The two constituent genera (Collomena and Motyd)were not dealt with by either Mell or Forbes, and Franclemont & Todd (1983) do not give anyreasons for employing the tribe. Nolinae. The Nolinae also has affinities with the Sarrothripinae. Forbes (1960) derived theformer from the latter, while Richards (1932) (Fig. 3) preferred an independent developmentfrom the Acontiinae. However, the boat-shaped cocoon of the nolines, together with tuftedsetae (similar to those found in certain sarrothripines; Forbes, 1960: 52) point towards arelationship with the Sarrothripinae, and hence the Chloephorinae. There appears to be apotentially monophyletic group of three subfamilies, which may also possibly be related to theeuteliines, through the Risobini. However, such a relationship is as yet largely unverified. Acontiinae. This subfamily was Hampson's least well-defined group as adults, falling as it didacross the trifine-quadrifine boundary. However, he did note that 'the larvae of such as areknown [have] the anterior [pair of] prolegs aborted, which is the essential distinction betweenthe two subfamilies' [Acontiinae and Amphipyrinae + Acronictinae]. This character was oflimited use, however, because so few acontiine larvae were known at the time, and reduction ofthe anterior prolegs is widespread in other noctuid subfamilies. Also, there are severalacontiines (e.g. Neoerastria [Homophoberia]) in which all the prolegs are present (Forbes, 1954:270). Richards (1932: 23) considered the acontiines 'to have been derived from some point betweenthe Rivulinae and Hypeninae with Eublemma [Eumicremma] as the connecting link' (but seeFig. 3), and which then gave rise to the sarrothripine-chloephorine series, the Nolinae, thetrifine subfamilies and the Agaristinae. The Acontiinae itself was divided into three sections:Eublemma [Eumicremma], the Erastriini [Eustrotiini] and the Tarachini [Acontiini]. Of the twotribes, probably only the second is monophyletic on the basis of the enlarged, chitinised alula.The Eustrotiini was characterised by the lack of such an alula and is therefore undoubtedly atleast paraphyletic. Forbes (1954) characterised the Eublemmini (Eublemma [Eumicremma]) using probableplesiomorphies ('pocket IV double as in the deltoids, hood and alula normal, no corona orpenicillus' [on the valve of the male genitalia]) and did not improve upon the definition of theErastriini [Eustrotiini]. Franclemont & Todd (1983) largely accepted Richards' and Forbes'classification but employed two additional monobasic tribes, the Cydosiini and the Bagisarini,erected to contain two presumably, somewhat aberrant genera. The higher classification of the Acontiinae leaves much to be desired. Only the Acontiini andthe monobasic tribes are likely to be monophyletic. The Eustrotiini appears to represent aheterogeneous assemblage of genera that do not fit any of the other tribes. However, the crucialposition of the subfamily in understanding the higher systematics of the Noctuidae as a wholemeans that a thorough (cladistic) analysis should be accorded high priority. Pantheinae. As adults, apart from dubious differences in hindwing venation, the members ofthe Pantheinae share the same Hampsonian defining characters as the Hadeninae: unspinedtibiae and hairy eyes. The larvae, however, are quite different, being clothed, except for Raphia,in dense tufts of secondary setae on the body and head. In the latter respect, the Pantheinaediffer from the Acronictinae. Richards (1932) found them tympanally very homogeneous buthighly isolated from the rest of the noctuids; so much so that he was unable to even tentativelyassign them a place on his phylogenetic tree. Forbes (1954) also noted the well-defined nature ofthe Pantheinae, although he had reservations regarding Raphia. This genus corresponded to theother pantheines in its tympanum, wing venation and general facies, but differed in themicroscopic hair on the eyes and the larva lacking secondary setae. Nye (1975), following 220 IAN J. KITCHING Hampson, placed Raphia in the Ophiderinae. Raphia may belong in the Panth.einae, but it couldequally belong elsewhere and only a thorough understanding of the rest of the family will finallyresolve the question. Franclemont & Todd (1983) follow Forbes (1954) but also suggest the possibility of mergingthe Pantheinae and Acronictinae. If Richards (1932) were correct regarding the tympanalstructure of the two groups, then such a move would be premature, which is not to rule it out.The two subfamilies were long considered closely related (e.g. Smith & Dyar, 1898) and thepresence of secondary setae in the larvae may be a synapomorphy. However, Mosher (1916)noted that the pupa of Charadra deridens (the only species examined) was more arctiid thannoctuid, in the shape of the body, the presence of setae arranged around the scars of the larvalverrucae, the absence of an epicranial suture and in the labial palps and prothoracic femorabeing visible. However, the cremaster is more noctuid than arctiid, in being long and providedwith hooked setae. Additionally, the pantheines have never been recorded as possessing tymbalorgans. The sum total of these characters, together with Mosher's highly restricted sample,suggest that, until further information has been gathered, the pantheines are best left in theNoctuidae. Their position with regard to the other subfamilies, however, remains unknown. Acronictinae. The Acronictinae is the first of the trifine subfamilies, all of which are highlyuniform structurally. This has resulted in much confusion in their classification and has led to anover-reliance upon superficial characters (hairy/lashed eyes, spined tibiae). However, someorder can be discerned within the trifines and one group of genera that has long been recognisedas a distinctive subgroup is the Acronictinae. Originally, the acronictines were distinguished by the presence of secondary setae on the bodyof the larvae. The subsequent inclusion of the bryophilines confused matters, because the larvaeof these genera possess only long, primary setae. Hampson placed no emphasis on larvalvestiture and included the acronictines, together with the amphipyrines and certain heliothinesin a more inclusive 'Acronyctinae', where they have generally remained. Gardner (1946a) distinguished the acronictines as his group AI. He also included thepantheine Diphthera [Trichosea] champa; the amphipyrine Cetola dentata and the ophiderineThiacidas postica (this genus appears superficially to resemble Raphia and may thus be betterplaced in the Pantheinae - assuming Raphia belongs in that subfamily). Examination of thepupae led Gardner (1948ft) to exclude Cetola from the Acronictinae and he placed it instead inthe Amphipyrinae. Mosher (1916) found the pupae to be of little use in distinguishing theAcronictinae; so much so, that she also included an amphipyrine (Achatodes) and threecatocalines (s.l.) (Homopyralis [Metalectra] , Anomis and Plusiodonta) . However, she was wellaware of the unnatural nature of this grouping. Forbes (1954) followed Hampson and treated the acronictines as a tribe (the Apatelini) withinhis second series of 'Acronyctinae', but the trend was reversed by Franclemont & Todd (1983).They reinstated the Acronictinae as a separate subfamily, containing two tribes, the Acronictiniand the Bryophilini. As mentioned above, Franclemont suggested that the trifine noctuids mightbe better treated as comprising only two subfamilies: the Acronictinae (including the Pan-theinae) and the Noctuinae (comprising the remainder). Unfortunately, he did not elaborate onhis reasoning. The Acronictinae, as restricted by Franclemont & Todd, probably represents a monophyleticunit but its limits are presently poorly defined. Whether it indeed is related to the Pantheinae, oris a convergent offshoot from somewhere within the Amphipyrinae, remains to be discovered. Agaristinae. For a long time, these brightly-coloured, largely diurnal and aposematic mothswere accorded family rank near the Arctiidae. However, Mosher (1916) could not discover anydifferences in the pupae to distinguish the agaristines from the Noctuidae and thus placed themin the latter as a subfamily. Richards (1932) found the agaristine tympanal organ to be similar tothose of the Acronyctinae [Acronictinae + Amphipyrinae], although the counter-tympanalcavities of the agaristines were much enlarged. Nevertheless, he retained the family status of thegroup and considered the agaristines to be derived from the Acronyctinae (Fig. 3). REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 221 Gardner (1946ft) examined two agaristine genera (Aegocera and Eusemia [Episteme]) andplaced them in his group AVI, together with Callyna (Amphipyrinae), Tiracola (Hadeninae)and Churia (Chloephorinae). In his subfamily summaries, Gardner (19480) listed the Agaristi-nae between the Noctuinae and the Cuculliinae, suggesting that he considered the group to bemerely aberrant trifine noctuids. Forbes (1960) retained the family status whereas Franclemont& Todd (1983) treated them as a noctuid subfamily. It is probable that the Agaristinae ismonophyletic, although the involvement of mimicry might complicate superficial resemblances.The relationships of the subfamily to other trifine groups is currently poorly understood. Amphipyrinae. To Hampson, this subfamily (including the Acronictinae and the Pyrrhia-group heliothines) was characterised entirely by absences: hindwing vein M 2 reduced, the tibiaeunspined and the eyes bare and unlashed. As such, the group can be seen to be a prime candidatewithin the Noctuidae for paraphyly or polyphyly. Unfortunately, like the Catocalinae (s.l.), thelarge size of the group (it contained nearly 2,400 species in Hampson's Catalogue, a number thathas considerably increased since; Forbes, 1954) has resulted in a long period of classificatorystasis. Richards (1932) found a high degree of uniformity in the tympanum and could not suggestany interrelationships beyond that they appeared to be derived from the higher Acontiinae, agroup that itself is probably not monophyletic. Gardner (1946a, ft; 1947; 1948a) found a great deal of diversity within the larvae and allocatedamphipyrines to his groups AI, All, AVI, BI and C. In contrast, the pupae of this subfamilywere, for the most part, indistinguishable from those of the Hadeninae, Noctuinae andHeliothinae that he examined. The first attempt to subdivide the Amphipyrinae, in which the characters used were stated,was provided by Forbes (1954). On the basis of the genitalia and larvae, he erected two series,each divided into several tribes. Franclemont & Todd (1983) reduced the number of tribes tofour. The Apameini corresponded to Forbes' first series less the Prodeniini. This latter tribe,together with many of Forbes' isolated genera, formed the Amphipyrini, while the Stiriini waslargely unchanged. Franclemont & Todd's fourth tribe, the Nocloini, the basis for which isunclear, contained seven genera not dealt with by Forbes. Of these four tribes, the Stiriini is almost certainly monphyletic (Hogue, 1963) and theApameini (and Nocloini?) may also prove to be. The Amphipyrini appears to be a somewhatmore restricted 'dustbin' than was the Amphipyrinae and is therefore likely to be at leastparaphyletic. It may even be polyphyletic given that the other trifine subfamilies are thought tohave arisen from within its limits, as presently defined. Cuculliinae. The Cuculliinae is also a prime candidate for having a polyphyletic origin(Forbes, 1954). It was circumscribed by Hampson on the basis of bare, lashed eyes and unspinedtibiae, not the most convincing of characters, especially as, venation apart, it applies equally wellto the majority of the Plusiinae. Lack of material prevented both Mosher (1916) and Gardner(19480) from reaching more than highly tentative conclusions regarding both the internal andexternal relationships of the Cuculliinae based on larvae. Forbes (1954) divided the subfamily into two series, each comprising three tribes, which hethought to have been independently derived from the 'Acronyctinae'. The Cuculliini, Oncocne-midini and Psaphidini were considered to be related to certain Stiriini (q.v.) and the generaOxycnemis and Catabena (the latter an 'isolated genus' of acronyctines, the former notmentioned elsewhere). This latter relationship was recently strengthened by the inclusion ofCatabena and Oxycnemis in the Oncocnemidini (Franclemont & Todd, 1983). The remainingthree tribes, the Lithophanini, Cleocerini and Antitypini, apparently showed affinities with suchgenera as Andropolia, Rhizagrotis and some elements of the Apamea-Septis complex (all ofwhich are currently placed in the Apameini; Franclemont & Todd, 1983). The latter authorsretained the Cuculliini, Oncocnemidini and Psaphidini but considered the other three asconstituting a single tribe, the Xylenini. In addition, Feralia was placed in the monobasicFeraliini. Given the highly uncertain nature of the trifine subfamilies in general, to find the Cuculliinae 222 IAN J. KITCHING polyphyletic would be no great surprise. Of the constituent tribes, the Psaphidini (fore-tibialmodification) and the Feraliini (by virtue of it containing a single, assumed monophyletic genus)are probably monophyletic taxa. The Cuculliini and Oncocnemidini taken together may alsoprove monophyletic but the latter may be paraphyletic with respect to the former ('they tend tointergrade in Europe'; Forbes, 1954: 122). The Xylenini, as currently conceived, is probablyparaphyletic, although the Lithophanini (sensu Forbes, 1954) may prove a monophyletic unit,based upon the form of the digitus in the male genitalia and the biological characteristic of anautumn adult emergence followed by hibernation. These tentative conclusions are unaffected bythe nature of the Cuculliinae as a whole. Hadeninae. The hadenine genera were originally separated into a number of subfamilies byearly authors (e.g. the Orthosiinae, Leucaniinae, Hadeninae; see also Heslop, 1960). However,they were all brought together into a single trifine subfamily by Hampson, defined by hairy eyesand unspined tibiae. As with the other trifine groups, tympanal organs and immature stagesprovided little information. Mosher (1916) characterised the Hadeninae by pupae possessing'stout straight setae or spines at the caudal end of the body'. However, she also included theheliothines Chloridea [Heliothis], Pyrrhia and Rhodophora [Schinia], the noctuine Lycophotiaand the amphipyrines Eriopus [Callopistria] and Laphygma and Prodenia [both Spodoptera]. Forbes (1954) recognised three general facies within what he considered 'a homogeneousgroup', but also noted that a large proportion of the genera did not fit into any of them.Consequently, he did not formally subdivide the hadenines, unlike Franclemont & Todd (1983),who recognised three tribes. The Glottulini comprised only Xanthopastis , an aberrant genusForbes considered might not be closely related to the other hadenines. The Eriopygini containedOrthodes, Tricholita and their relatives, while the Hadenini consisted of the residue. TheGlottulini are probably monophyletic given, for example, their amaryllidaceous-f ceding larvae(this tribe undoubtedly also includes the Old World genus Brithys). From the availableinformation, the basis of the Eriopygini is unknown and consequently its status cannot becommented upon. Such is therefore also the case for the Hadenini but it is almost certainly atleast paraphyletic, even assuming the Eriopygini to be adequately characterised. Within theHadeninae, there are 'centres of monophyly', around such genera as Xanthopastis, Mythimna,Hadena and Orthosia, but their limits and interrelationships have yet to be established. Noctuinae. This was the fourth and last trifine subfamily recognised by Hampson, on the basisof spined tibiae. Both Mosher (1916) and Gardner (19466; 1948a) found no characters by whichthey could distinguish the larvae and pupae of the noctuines from those of most other trifinegenera. The heliothine section of the Agrotinae (sensu Hampson) had long been recognised byearly authors as a discrete group but only slowly re-emerged as a distinct entity, which variedfrom being classed as a tribe of the Noctuinae (Forbes, 1954) to a separate subfamily that wasisolated from the Noctuinae (e.g. McDunnough, 1938). Discussion of the 'heliothine Noctuinae'is deferred to the next section. The remaining genera of Noctuinae were split by Heslop (1960) into two groups, followingEuropean tradition, one (Noctuini) characterised by yellow hindwings, the other (Agrotini) byhindwings of a different colour (usually brown). Forbes (1954) disagreed with this system,preferring instead to recognise three informal groups based around Agrotis, Peridroma andNoctua, which were formalised by Franclemont & Todd (1983) as the Agrotini, Aniclini andNoctuini respectively. In addition, they recognised a fourth tribe, the Ufeini, to accommodatethe aberrant genus Ufeus (The genus is not really an Agrotid, but fits no better elsewhere';Forbes, 1954: 74). The three large tribes are relatively well defined by larval and genitalic characters but it is notclear how well these would stand up to a critical cladistic analysis. Indeed, the monophyly of theNoctuinae itself has yet to be adequately demonstrated. Heliothinae. The Heliothinae was formed largely from the union of two subgroups of trifinenoctuids, classified widely apart by Hampson, based around the genera Heliothis and Pyrrhia. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 223 The former, because of their spined tibiae, were placed in the Noctuinae, while the latter, whichlacked all Hampson's 'definitive characters', were relegated to 'acronyctine' obscurity. However, the larvae of the Heliothinae are relatively distinctive. Their habit of feeding moreor less exclusively on the flowers and fruits of low-growing plants had long been recognised, butthis information was subsequently supplemented by structural features (Crumb, 1926; 1956;Gardner, 1946a,6). Of prime importance among these were the biordinal crotchets, a featurethat the group shares with most Plusiinae and some Cuculliinae. The North American Heliothinae were revised by Hardwick (1970). He excluded Grotella,Neogrotella and Hemigrotella from the subfamily, because of their slender build, distinctlyquadrifine hindwing venation and the presence of multiple cornuti on the vesica, features thatHardwick considered linked the three genera with the Stiriini. However, Franclemont & Todd(1983) retained them in the Heliothinae as the Grotellini, possibly as an interim measure. The relationship between the Stiriini, the Heliothinae and, possibly, the Plusiinae warrantsfurther consideration, due to its potential profound effects upon the higher classification of theNoctuidae generally. Hogue (1963) considered the Stiriini were derived from generalisednoctuine stock via forms similar to either certain Heliothinae or Oncocnemidini, with the lattermore likely (both groups possess an angled vesica bearing numerous cornuti). Furthermore,Hogue treated any resemblance to the Plusiinae to be entirely convergent. Hardwick (1970) discussed the relationship between the Stiriini and the Heliothinae inconsiderable detail and concluded that there were 'a sufficient number of features in common tosuggest some immediate common ancestry'. No mention was made of the Plusiinae withreference to a relationship with the stiriines and heliothines. That such a possibility can beentertained rests on somewhat equivocal evidence. Both the Heliothinae and the Plusiinae havelarvae with biordinal crotchets, which could be interpreted as synapomorphic. However, thelarvae of the tentative sister-group of the plusiines, the Stictopterinae, have uniordinal crotchets(Gardner, 19480), while the larvae of most stiriines are unknown and the crotchets of such thatare have never been examined. Biordinal crotchets may, therefore, be homoplasious in the twogroups. Perhaps the only overall conclusion that can be reached, like so many concerning thehigher classification of the Noctuidae, is that it all depends on the results of investigations yet tobe done. The higher classification of the Noctuidae - fact or fiction? No adequate higher classification of the Noctuidae can yet be proposed that will serve as areplacement for the Hampsonian system. However, it is possible to construct a tentativecladogram (Fig. 4) to serve as a suitable starting-point. Only those apomorphies that can bereasonably positively identified are included, although those on several branches are highlysuspect. The relative paucity of synapomorphies results in a number of extensive polychotomiesand several of the groups being uncharacterised (see below). The individual branches, 1-34, willbe discussed and justified seriatim, and their relative merits assessed. 1. The Arctiidae generally can be characterised by three apomorphies, none of which arepresent in all genera. Two are perhaps quite reliable: the presence of the tymbal organ(although this is absent/lost in some of the ctenuchines and a few other groups) and thepresence of two subventral (SV) setae in the larvae on the meso- and metathorax. Only oneSV seta is present in this position in the aganaids, herminiids and noctuids (Gardner, 1941),and examination of five notodontids (Stauropusfagi, Notodonta dromedarius, Eligmodontaziczac, Peridea anceps and Pheosia tremula; Kitching, unpublished) showed these alsopossess only a single seta. The latter evidence was used in an outgroup comparison toestablish the bisetose condition as apomorphic. The third arctiid apomorphy, a swollenhindwing vein Sc + R^ is more uncertain, especially as 'swollen' has never been preciselydefined. 2. No apomorphies have been discovered in the literature for the aganaid genera. As such,they are of uncertain position and it is possible for some to be more closely related to the 224 IAN J. KITCHING ArctiidaeAganaidaeHerminiidaeRivulinaeHypenodinaeHypeninaeCatocalinaeAnomis/ AlabamaCatocala/Othreis Fig. 4 Cladogram illustrating the relationships between the various noctuid subgroups. For detailsregarding the characters defining each of the numbered branches, 1-34, see pp. 223-226. Sevenbranches are undefined by apomorphies; those subtending the Rivulinae, Catocalinae, Acontiinae,Amphipyrinae, Acronictinae, Cuculliinae and Hadeninae. For discussion of the undefined branch 2, seepp. 214-215,223-225. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 225 arctiids and some to the noctuids. It may even be that some represent the sister-group of theherminiids. 3. The exclusion of the 'fan-foots' and their relatives from the Noctuidae (branch 4) isprimarily on the basis of the plesiomorphic pre-spiracular hood. In addition, they may becharacterised by two apomorphies: modified fore-tibiae in the males of most genera(although the exact nature of the modification is deliberately left unspecified) and a swollenmetepimeron ventral to pocket IV. 4. The Noctuidae is restricted to those groups possessing a post-spiracular hood, although thisis absent in the Pantheinae (q.v.) and is greatly reduced in many other genera scatteredamong the other subfamilies. These can be interpreted as secondary losses but requirefurther analysis. Two uncharacterised paraphyletic/polyphletic assemblages are recognisedwithin the heptachotomy that terminates branch 4, the Rivulinae and the Acontiinae. 5 . The Hypenodinae are characterised by the lack of ocelli , an unsatisfactory situation but thebest that can be done at present. 6. A similar situation pertains to the Hypeninae. Its putative apomorphies of long, 'deltoid'palps and lashed eyes are not particularly convincing. 7. The Catocalinae (s.l.) are better characterised (fused pleural sclerite in the male genitaliaand pupa with a whitish bloom) but the majority of genera and species of this vast subfamilyhave yet to be examined, especially with regard to the pupal character. 8 & 9. These branches represent two of the many genus-groups that can be recognised in theCatocalinae (s.l.). Anomis and Alabama (the Anomiini of Forbes, 1954) are defined by anenlarged but unsclerotised alula and malvaceous-feeding larvae. Catocala and Othreis bothpossess a chitinous projection from the inner margin of the tympanal frame (the 'Biigel';Eggers, 1919). 10. Of the acontiine groups, only the Acontiini is well-characterised: tympanal hood reduced/lost, alula enlarged and sclerotised, male with paired anal hair-masses; cf. Euteliinae(Forbes, 1954: 271). 11. This branch includes four subfamilies, tentatively grouped on the absence of a cremaster.However, this has not been confirmed for the Nolinae. The Euteliinae also bear ageneral resemblance in venation and the larvae of certain sarrothripine genera (Forbes,1954: 288). 12. The Euteliinae are perhaps the most well-defined subfamily of noctuids: frenulum reducedin the female; larvae feeding mainly on Anacardiaceae; pair of anal hair pencils present inthe male; and the attitude of the adults at rest. There can be little doubt they represent amonophyletic taxon. 13. The Nolinae, Sarrothripinae and Chloephorinae are united by the common possession of aboat-shaped cocoon with a vertical exit slit. 14-17. Unfortunately, the relationships within the above group of subfamilies are unclear. TheChloephorinae and Sarrothripinae agree in the possession of a bar-shaped retinaculum inthe male but the latter also agrees with the Nolinae in the presence of tufts of raised scales inthe forewing cell. As the second character also occurs elsewhere in the Noctuidae (e.g. theStictopterinae and the Plusiinae: Abrostolini), the first character is preferentially taken torepresent a synapomorphy. However, it is probable that the group needs to be split up in adifferent manner, possibly along the lines suggested by Mell (1943). In addition, the nolineslack ocelli in the adult and have larvae with tufted setae. 18. The 'higher noctuids' can generally be characterised by the presence of a clavus in the malegenitalia. However, this structure is absent in many genera and uncertainties regardinggenitalic homologies decrease confidence in this character. Secondly, the larval silk-pore of 226 IAN J. KITCHING the included subfamilies is not concealed (Crumb, 1956) and may represent a goodapomorphy, although the condition in the Stictopterinae is unknown. 19-21. The Stictopterinae and Plusiinae share a double tympanal hood and tufts of raised scalesin the forewing cell (present in the Plusiinae only in the Abrostolini). The stictopterines canbe distinguished by the reduced female frenulum and possibly by the caudal extremity of thepupa being produced as 'two divergent attenuated spines borne together on a thickermedian stem' (Gardner, 1948ft: 88). The plusiines are relatively well defined, with fourapomorphies: lashed eyes, a metepimeral bulge formed by an enlarged pocket IV, biordinalcrotchets in the larvae and pupae in which the wings and proboscis project beyond theposterior margin of abdominal segment 4 ventrally. 22. The subfamilies subtended by this branch, the Trifinae', are held together on ratherdubious grounds. Of the obsolescent hindwing vein M 2 , nothing more needs to be said. Thetympanal organs of all forms (except the Pantheinae, see branch 24) are extremely uniformand similar to those of the 'higher Erastriinae' [Acontiinae] (Richards, 1932: 29). The exactnature of this homogeneity and whether it includes any structures that can be regarded assynapomorphies was not elucidated by Richards. 22-24. The Acronictinae and Pantheinae are grouped on the basis of the presence of secondarysetae on the larval trunk. However, the absence of such setae in the Acronictinae:Bryophilini weakens this argument, and the two subfamilies may not be closely related atall. The pantheines are further distinguished by their hairy eyes, the presence of secondarysetae on the larval head, the reduction/absence of the tympanal hood and the highlymodified tympanal morphology (Richards, 1932: 28). 25-30. Three branches (25, 26 and 29) subtend three of the currently recognised trifinesubfamilies, the Noctuinae, Cuculliinae and Hadeninae respectively. This was done, notbecause there are good synapomorphies for the included genera (the characters employedare the classic ones of spined tibiae, lashed eyes and hairy eyes), but because to omit themwould create a vast, uncharacterised group of genera (cf . the Catocalinae). In addition, it isprobable that some of the groups will eventually be able to be defined by good apomor-phies. Two tribes are split out of the Cuculliinae: the Lithophanini (well-developed digitusin the male genitalia, adult emergence generally autumnal, followed by hiberation) and thePsaphidini (fore-tibia with a terminal claw on the inner side, usually with an oblique, flatplate continuous with it; Forbes, 1954: 127); and one tribe from the Hadeninae: theGlottulini (larvae brightly-coloured, black, transversely spotted with yellow/white; feedingon the bulbs of the Amaryllidaceae). 31-33. The amphipyrine tribe, the Stiriini, and the Heliothinae are associated by the clawedfore-tibiae (single in the former, multiple in the latter) and the larval preference for feedingon flowers and young fruits. The Stiriini are characterised by an angled vesica with multiplecornuti and a heavily chitinised frons in the adult, with a raised ring and various projections.The larvae of the Heliothinae have biordinal crotchets. 34. The final noctuid subfamily to be considered is the Agaristinae. This is a highly apomorphicgroup: counter-tympanum several times the size of the tympanal membrane; hood veryreduced/absent; adults active by day, generally brightly coloured, usually with clubbedantennae; larvae also brightly coloured, feeding largely on Vitaceae and Onagraceae. I do not claim that the apomorphies employed in Fig. 4 are all that could be used to reconstructthe higher classification of the Noctuidae. However, they do represent all those extracted fromthe literature in which I have more than minimal confidence regarding the polarity, although itmust be remembered that few are known to be present in all members of the group that they arebeing used to characterise. Many more characters are known at the subfamily level (see, forexample, those cited by Forbes, 1954) but their usefulness has yet to be assessed. In addition, REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 227 there are certainly a considerable number of features known that are as yet undocumented,which must therefore necessarily fall outside the scope of this analysis. To return to the title of this section, it would perhaps be more appropriate to refer to theHampsonian classification as 'classifiction'. The system proposed by Franclemont & Todd(1983), which ranks as the main contender as an alternative, has much to commend it. However,it is of limited application until extended to encompass the world fauna and, unfortunately, untilsuch time as the defining characters/apomorphies of the included groups are reported, suchdevelopment is impossible. That there is a higher classification for this group of moths is not inquestion, but until careful character analyses, performed within a cladistic methodologicalframework, are published, the present state of confusion will reign. I hope that noctuidsystematists will rise to meet the challenge. References Ackery, P. R. & Vane- Wright, R. 1. 1984. Milkweed butterflies: their cladistics and biology. London.Arnold, J. W. 1982. Larval hemocytes in Noctuidae (Insecta: Lepidoptera). International Journal of Insect Morphology & Embryology 11: 173-188.Aubert, J.-F. & Boursin, C. 1953. Les phalenides du Jura (Revision de la nomenclature et introduction a la faune des macrolepidopteres du Jura). Bulletin Mensuel de la Societe Linneene de Lyon 22: 115-126.Barnes, W. & Benjamin, F. H. 1923. Nomenclature notes and new species. Contributions to the Natural History of the Lepidoptera of North America 5: 53-61.Barnes, W. & McDunnough, J. H. 1917. Checklist of the Lepidoptera of Boreal America, viii + 392 pp. Decatur, Illinois. Beck, H. 1960. Die Larvalsystemdtik der Eulen (Noctuidae). 406 pp. Berlin.Birch, M. C. 1972a. Male abdominal brush organs in British noctuid moths and their value as a taxonomic character. Part I. Entomologist 105: 185-205.1972ft. Male abdominal brush organs in British noctuid moths and their value as a taxonomic character. Part II. Entomologist 105: 233-244.Blackmore, E. H. 1927. Check-list of the Macrolepidoptera of British Columbia (Butterflies and moths). 47 pp. Victoria, British Columbia.Blest, A. D., Collett, T. S. & Pye, J. D. 1963. The generation of ultrasonic signals by a New World arctiid moth. Proceedings of the Royal Society of London (B) 158: 196-207. Boisduval, J. A. 1840. Genera et index methodicus europaeorum Lepidoptorum. vii + 238 pp. Paris.Borkhausen, M. B. 1792. Naturgeschichte der europdischen Schmetterlinge 4, xvi + 809 pp. Frankfurt.Borner, C. 1953. Lepidoptera, pp. 382-421. In Brohmer, P., Fauna von Deutschland. Edn 7, ix + 591 pp. Heidelburg.Boursin, C. 1964. Les Noctuidae Trifinae de France et de Belgique. Bulletin Mensuel de la Societe Linneene de Lyon 33: 204-240.Bretherton, R. F., Goater, B. & Lorimer, R. I. 1979. Noctuidae, pp. 120-278. In Heath, J. & Emmet, A. M., The moths and butterflies of Great Britain and Ireland. 9, Sphingidae-Noctuidae (Part I). 288pp. London.Brock, J. P. 1971. A contribution towards an understanding of the morphology and phylogeny of the ditrysian Lepidoptera. Journal of Natural History 5: 29-102.Butler, A. G. 1879. On the natural affinities of the Lepidoptera hitherto referred to the genus Acronycta of authors. Transactions of the Entomological Society of London 12: 313-317.1881. Descriptions of new genera and species of heterocerous Lepidoptera from Japan. Transactions of the Entomological Society of London 14: 1-23, 171-200, 401-426, 579-600.1893. Notes on the genus Acronycta of authors and its position in the classification of heterocerous Lepidoptera. Annals and Magazine of Natural History (6) 11: 396-402.Chalmers-Hunt, J. M. 1962-1968. The butterflies and moths of Kent. 2, Heterocera (Sphingidae - Plusiidae). 380 pp. Arbroath. [Published in parts in the Entomologist's Record and Journal of Variation 74-80.]Chapman, T. A. 1893a. The genus Acronycta and its allies, iv + 116 pp. London. 1893ft. The genus Acronycta and its allies. Entomologist's Record and Journal of Variation 4: 97. Common, I. F. B. 1968. Lepidoptera (Moths and butterflies), pp. 765-866. In Mackerras, I. M., Insects of Australia, xiii + 1029 pp. Canberra. 1975. Evolution and classification of the Lepidoptera. Annual Review of Entomology 20: 183-203. Comstock, J. H. 1925. An introduction to entomology. Second edition, xix + 1044 pp. Ithaca, New York. 228 IAN J. KITCHING Crumb, S. E. 1926. The Nearctic budworms of the lepidopterous genus Heliothis. Proceedings of theUnited States National Museum 68: 1-8. 1934. A classification of some noctuid larvae of the subfamily Hypeninae. Entomologica Americana14: 133-196. 1956. The larvae of the Phalaenidae. Technical Bulletin. United States Department of Agriculture no. 1135: 1-365.De Koning, H. S. & Roepke, W. 1949. Remarks on the morphology of the teak moth, Hyblaeapuera Cr. (Lep. Hyblaeidae). Treubia2Q: 25-30.Denis, J. N. C. M. & Schiffermiiller, I. 1775. Ankiindung eines systematischen Werkes von den Schmetter- lingen der Wienergegend. 323 pp. Wien.Dethier, V. G. 1941. The antennae of lepidopterous larvae. Bulletin of the Museum of Comparative Zoology 87: 455-507.DeVries, P. J., Kitching, I. J. & Vane-Wright, R. I. In prep. The systematic position of Antirrhea and Caerois with comments on the classification of the Nymphalidae.Draudt, M. 1919-1936. Agaristidae and Noctuidae. In Seitz, A., The Macrolepidoptera of the World. 7: 3-411. (Pages 412-508 consulted in the German edition, published 1919-1944.) Stuttgart. 1938. Phytometrinae [part]. In Seitz, A., The Macrolepidoptera of the World 11: 496. Stuttgart. Druce, H. 1887. Lepidoptera-Heterocera. In Biologia Centrali- Americana. 1: 249. London. Duponchel, P. A. J. 1844-1846. Catalogue methodique des Lepidopteres d'Europe. xxx + 523 pp. Paris.Dyar, H. G. 1895a. Additional notes on the classification of lepidopterous larvae. Transactions of the New York Academy of Sciences 14: 49-63. 1895ft. Larva of Demos propinquilinea; its systematic position. Journal of the New York Entomologi-cal Society 3: 130-131. 1896. On the position of genus Demas. Canadian Entomologist 2%: 103-104. 1904. Review of 'Catalogue of the Lepidoptera Phalaenae in the British Museum. Vol. IV.' G. F. Hampson. Canadian Entomologist 36: 27-28.Eggers, F. 1919. Das thoracale bitympanale Organ einer Gruppe der Lepidoptera Heterocera. Zoolo-gische Jahrbiicher 41: 273-376. -1925. Versuche iiber das Gehor der Noctuiden. Zeitschrift fur vergleichende Physiologic 2: 297-314.Eichlin, T. D. & Cunningham, H. B. 1978. The Plusiinae (Lepidoptera: Noctuidae) of America north of Mexico, emphasizing genitalic and larval morphology. Technical Bulletin. United States Department of Agriculture Agricultural Research Service no . 1567: 1-222. Farris, J. S. 1974. Formal definitions of paraphyly and polyphyly. Systematic Zoology 23: 548-554.Forbes, W. T. M. 1910. A structural study of some caterpillars. Annals of the Entomological Society of America 3: 94-143. 1914. A table of the genera of Noctuidae of northeastern North America. Journal of the New YorkEntomological Society 22: 1-33. 1916. On the tympanum of certain Lepidoptera. Psyche 23: 183-192. 1918. Guenee's Herminidae revived. Journal of the New York Entomological Society 26: 224-225.- 1924. The family position of Graphelysia. Psyche 31: 146-147. 1933. The pupa of Hyblaea (Lepidoptera, Hyblaeidae). Annals of the Entomological Society ofAmerica 26: 490. - 1939. The Lepidoptera of Barro Colorado Island, Panama. Bulletin of the Museum of ComparativeZoology 85: 95-322. 1954. Lepidoptera of New York and neighboring states. Part 3. Noctuidae. Memoirs. Cornell University Agricultural Experiment Station no. 329: 1960. Lepidoptera of New York and neighboring states. Part 4. Agaristidae through Nymphalidae including butterflies. Memoirs. Cornell University Agricultural Experiment Station no. 371: 1-188.Forbes, W. T. M. & Franclemont, J. G. 1957. The striated band (Lepidoptera, chiefly Arctiidae). Lepidopterists' News 11: 147-150.Forster, W. & Wohlfahrt, T. A. 1971. Die Schmetterlinge Mitteleuropas. 4, Eulen (Noctuidae). vii + 329 pp. Stuttgart. Fracker, S. B. 1915. The classification of lepidopterous larvae. Illinois Biological Monographs 2: 1-170.Franclemont, J. G. & Todd, E. L. 1983. Noctuidae, pp. 120-159. In Hodges, R. W.,etal. Check list of the Lepidoptera of America north of Mexico. London.Gaede, M. 1913-1939. Noctuidae. In Seitz, A., The Macrolepidoptera of the World. 15: 30-286. Stuttgart. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 229 1937-1938. Eutelianae [Noctuidae] - Catocalinae. In Seitz, A., The Macrolepidoptera of the World 11: 352-496. Stuttgart.Gardner, J. C. M. 1941. Immature stages of Indian Lepidoptera (2) [Noctuidae, Hypsidae]. Indian Forest Records (N.S.) 6: 253-296.1946a. On the larvae of the Noctuidae (Lepidoptera) - I. Transactions of the Royal Entomological Society of London 96: 61-72.1946ft. On the larvae of the Noctuidae (Lepidoptera) - II. Transactions of the Royal Entomological Society of London 97: 237-252.1947. On the larvae of the Noctuidae (Lepidoptera) - III. Transactions of the Royal Entomological Society of London 98: 59-90.1948o. On larvae of the Noctuidae (Lepidoptera) - IV. Transactions of the Royal Entomological Society of London 99: 291-318. 1948ft. Notes on the pupae of the Noctuidae. Proceedings of the Royal Entomological Society of London (B) 17: 84-92.Godfrey, G. L. 1972. A review and reclassification of larvae of the subfamily Hadeninae (Lepidoptera, Noctuidae) of America north of Mexico. Technical Bulletin United States Department of Agriculture Agricultural Research Service no. 1450: 1-265.Grote, A. R. 1874. List of the Noctuidae of North America. Bulletin of the Buffalo Society of Natural Science 2: 1-77.1882a. An illustrated essay on the Noctuidae of North America. With 'A colony of butterflies'. 85 pp. London.1882ft. New check list of North American moths. 73 pp. New York. 1883. Introduction to a study of the North American Noctuidae. Proceedings of the American Philosophical Society 21: 134-176. 1889-1890. The Noctuidae of Europe and North America compared. Canadian Entomologist 21: 121-126, 154-157, 188-193, 226-230; 22: 26-30, 69-72, 105-109, 145-149. 1890. North American Lepidoptera, revised check list of the North American Noctuidae, part I, Thyatirinae, Noctuinae. 52 pp. Bremen. 1895o. List of North American Eupterotidae, Ptilodontae, Thyatiridae, Apatelidae and Agrotidae. Abhandlungen des naturwissenchaftlichen Vereins zu Bremen 14: 44-128. 1895ft. The hypenoid moths and allied groups. Proceedings of the American Philosophical Society 34: 416-436. 1896. Die Apateliden. Mitteilungen aus dem Roemer- Museum, Hildesheim 3: 1-18. Guenee, A. 1837. Essai pour servir a la classification des Noctuelides (suite). Annales de la Societe Entomologique de France 6: 311-367.1838o. Materiaux pour servir a la classification des Noctuelides (suite) (1). Annales de la Societe Entomologique de France 7: 107-125.1838ft. Materiaux pour servir a la classification des Noctuelides. Annales de la Societe Entomologique de France 7: 201-230.1839. Essai sur la classification des Noctuelides (suite). Annales de la Societe Entomologique de France 8: 473-522. 1841fl. Essai sur la classification des Noctuelides (suite) (1). Annales de la Societe Entomologique de France 10: 53-83. 1841ft. Essai sur la classification des Noctuelides (suite et fin). Annales de la Societe Entomologique de France 10: 217-250. 1852-1854. In Boisduval, J. A. & Guenee, A., Histoire naturelle des Insectes - Species general des Lepidopteres. 5 Noctuelites. I. xcvi + 407 pp. 6 Noctuelites. II. 444 pp. 7 Noctuelites. III. 442 pp. 8Deltoides et Pyralites. 448 pp. Paris. Hampson, G. F. 1893-1895. The fauna of British India, including Ceylon and Burma. 1, viii + 527 pp. 2, xxii+ 609 pp. 3, xxviii + 546 pp. London 1898-1913. Catalogue of the Lepidoptera Phalaenae in the collection of the British Museum. 1 Syntomidae. xii + 599 pp. 2 Arctiadae (Nolinae, Lithosianae). xx + 589 pp. 3 Arctiadae (Arctianae) andAgaris.idae. xix + 690 pp. 4 Agrotinae. xx + 689 pp. 5 Hadeninae. xvi + 634 pp. 6 Cucullianae. xiv + 532pp. 7 Acronyctinae. xv + 709 pp. 8 Acronyctinae continued, xiv + 583 pp. 9 Acronyctinae continued, xv+ 552 pp. 10 Erastrianae. xix + 829 pp. 11 Eutelianae, Stictopterinae, Sarrothripinae, Acontianae. xvii+ 689 pp. 12 Catocalinae. xiii + 626 pp. 13 Catocalinae continued, Mominae, Phytometrinae. xvi -I- 609pp. London. 1900. Lepidoptera Phalaenae, pp. 63-74. In Andrews, C. W., A monograph of Christmas Island (Indian Ocean), xiii + 337 pp. London. 230 IAN J. KITCHING - 1902. The moths of South Africa (Part II). Annals of the South African Museum 2: 255-446. 1918. Descriptions of new genera and species of Amatidae, Lithosidae and Noctuidae. Novitates Zoologicae 25: 93-217. 1926. Descriptions of new genera and species of Lepidoptera Phalaenae of the subfamily Noctuinae (Noctuidae) in the British Museum (Natural History). 641 pp. London.Handlirsch, A. 1929. Gegen die ubermassige Zerplitterung der systematischen Gruppen. Zoologischer Anzeiger 84: 85-90.Hardwick, D. F. 1970. A revision of the North American Heliothidinae (Lepidoptera: Noctuidae). Memoirs of the Entomological Society of Canada 73: 1-59.Harvey, L. F. 1874. Rectification of Treitschke's use of Hiibner's generic term 'Cymatophora'. Bulletin of the Buffalo Society of Natural Science 1: 276-278.Haworth, A. H. 1803-1829. Lepidoptera Britannica. Part I: xxxvi + 136 pp. Part II: pp. 137-376. Part III: pp. 377-511. Part IV: pp. 512-609. London.Hayes, A. H. 1975. The larger moths of the Galapagos Islands (Geometroidea: Sphingoidea & Noc- tuoidea). Proceedings of the California Academy of (Natural) Science 15: 145-208.Heinrich, C. 1926. In Bottimer, L. J., Notes on some Lepidoptera from eastern Texas. Journal of Agricultural Research 33: 797-819 [799].Herrich-Schaffer, G. A. W. 1845. Systematische Bearbeitung der Schmetterlinge von Europa, zugleich als Text, Revision und Supplement zu Jakob Hiibner's Sammlung europdischer Schmetterlinge. 2. Die Schwarmer, Spinner und Eulen (Hepialides-Cossides-Zygaenides-Sesiides-Sphingides-Bombycides- Noctuides-Nycteolides). 450pp. Regensburg.Heslop, I. R. P. 1945. Checklist of the British Lepidoptera with the English names of each of the 2299 species. 35 pp. London. 1960. Revised indexed check-list of the British Lepidoptera. Part I. Entomologist's Gazette 11:55-66.Hinton, H. E. 1946. On the homology and nomenclature of the setae of lepidopterous larvae, with some notes on the phylogeny of the Lepidoptera. Transactions of the Royal Entomological Society of London 97: 1-37.Hogue, C. L. 1963. A definition and classification of the tribe Stiriini (Lepidoptera: Noctuidae). Contributions in Science. Los Angeles County Museum 64: 1-129.Hiibner, J. 1805. Sammlung europaischer Schmetterlinge. pp. 155-194. Angsburg.Inoue, H. & Sugi, S. 1958-1961. Checklist of the Lepidoptera of Japan. 5 Noctuidae and 6 Hyblaeidae. pp. 431-683. Tokyo.Inoue, H., Sugi, S., Kuroko, H., Moriuti, S., & Kawabe, A. 1982. Moths of Japan. 2. Plates and synonymic catalogue. 532 pp. Tokyo.Janse, A. J. T. 1937-1939. The moths of South Africa. 3. Cymatophoridae, Callidulidae and Noctuidae (partim). xv + 435 pp. Durban.1939. On the South African Cuculliae (Noctuidae). Journal of the Entomological Society of South Africa 1: 78-130. 1942. The moths of South Africa. 4, Part 1. Jugatae. xxv -I- 78 pp. Durban. Jordan, K. 1906-1914. Agaristidae. In Seitz, A., The Macrolepidoptera of the World 3: 5-8. Stuttgart. - 1912-1914. Agaristidae. In Seitz, A., The Macrolepidoptera of the World 11: 1-30. Stuttgart.Jordan, K. & Gaede, M. 1913-1939. Agaristidae. In Seitz, A. , The Macrolepidoptera of the World 15: 1-29. Stuttgart. Kirby, W. F. 1907. The butterflies and moths of Europe. Ixxii + 432 pp. London.Kiriakoff, S. G. 1953. Recherches sur les organes tympaniques des lepidopteres en rapport avec la classification. X. Hyblaeidae. Bulletin et Annales de la Societe Roy ale Entomologique de Belgique 89: 258-263.1955. Onderzoekingen over de Gehoororganen bij Vlinders met Betrekking tot de Classificatie. XI. Agaristidae. Verhandelingen Konincklijke vlaamse academic voor wetenschappen, letter en en schone kunsten van Belgie'47: 1-70.1958. Recherches sur les organes tympaniques des lepidopteres en rapport avec la classification. XIV. Nolidae. BiologischJaarboek26: 69-75. 1960. Recherches sur les organes tympaniques des lepidopteres en rapport avec la classification. XV. Hypeninae et Herminiinae. Biologische Jaarboek 28: 140-153.1963. The tympanic structures of the Lepidoptera and the taxonomy of the order. Journal of the Lepidopterists' Society 17: 1-6. 1970. Die Stellung von Diloba caeruleocephala (L.) im Lepidopteren-System. Nachrichtenblatt der Bayerischen Entomologen 19: 101-104. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 231 Kitching, I. J. 1983. An analysis of danaid classifications based on differing character sets and methods. Unpublished Ph.D. thesis, University of London. 602 pp.In press. Early stages and the classification of the milkweed butterflies (Lepidoptera: Danainae). Zoological Journal of the Linnean Society.Kloet, G. S. & Hincks, W. D. 1972. A checklist of British Insects. Second edition (revised). Part 2. Lepidoptera. viii + 153 pp. London.Kostrowicki, A. S. 1961. Studies of the Palaearctic species of the subfamily Plusiinae (Lepidoptera, Phalaenidae). Acta Zoologica Cracoviensia 6: 367-472.Lederer, J. 1857. Die Noctuinen Europas. xv + 251 pp. Wien.Leech, J. H. 1900. Lepidoptera Heterocera from northern China, Japan and Corea. Parts III & IV. Transactions of the Entomological Society of London 33: 9-161, 511-663.Leraut, P. 1980. Liste systematique et synonymique des Lepidopteres de France, Belgique et Corse. 334 pp. Paris. Lhomme, L. 1923-1935. Catalogue des Lepidopteres de France et de Belgique. 1. 800 pp. Douelle, Lot.Linnaeus, C. 1758. Systema Naturae Edn 10 Regnum Animale. 1: 824 pp. Holmiae.McDunnough, J. H. 1916. On the types of certain noctuid genera occurring in North America (Lepidop-tera). Entomological News 27: 393-400.1938. Checklist of the Lepidoptera of Canada and the United States of America. Part 1. Macrolepi- doptera. Memoirs. Southern California Academy of Sciences 1: 1-272. 1944. Revision of the North American genera and species of the phalaenid subfamily Plusiinae (Lepidoptera). Memoirs. Southern California Academy of Sciences 2: 175-232.Meldola, R. 1881 . The developmental characters of the larvae of the Noctuae as determining the position of that group. Transactions of the Epping Forest and County of Essex Naturalists' Field Club. 1881: 1-10.Mell, R. 1943. Beitrage zur Fauna sinica. xxiv. Ueber Phlogophorinae, Odontodinae, Sarrothripinae, 'Westermannianae' und Camptolominae (Noctuidae, Lepid.) von Kuangtung. ZoologischeJahrbiicher 76: 171-226.Merzheevskaya, O. I. 1967. The larvae of Noctuidae, their biology and morphology (with keys). [In Russian.] 452 pp. Minsk.Meyrick, E. 1886. Monograph of New Zealand Noctuina. Transactions of the New Zealand Institute 19: 3^0. 1892. On the classification of the Geometrina of the European fauna. Transactions of the Entomo-logical Society of London 1892: 53-140.1912. A revision of the classification of the New Zealand Caradrinina. Transactions of the New Zealand Institute 44: 88-107. 1928. A revised handbook of British Lepidoptera. vi + 914 pp. London. Minet, J. 1982. Elements sur la systematique des Notodontidae et nouvelles donnees concernant leur etude faunistique a Madagascar. Bulletin de la Societe Entomologique de France 87: 354-370.Moore, F. 1881. On the genera and species of the lepidopterous subfamily Ophiderinae inhabiting the Indian region. Transactions of the Zoological Society of London 11: 63-76.Mosher, E. A classification of the Lepidoptera based on characters of the pupa. Bulletin of the Illinois State Laboratory of Natural History 12: 17-159.Nye, I. W. B. 1975. The generic names of moths of the World. 1, Noctuoidea (part): Noctuidae, Agaristidae, and Nolidae. 568 pp. London.Opinion 450. 1957. Suppression under the plenary powers of the generic name 'Phalaena' Linnaeus, 1758, and validation as of subgeneric status (a) as from 1758, of the terms 'Bombyx', 'Noctua', 'Geometra', Tortrix', 'Pyralis', Tinea' and 'Alucita', as used by Linnaeus for groups of species of the genus 'Phalaena' and (b) as of 1767 of the term 'Attacus' similarly published by Linnaeus and matters incidental thereto (Class Insecta, Order Lepidoptera). Opinions and Declarations rendered by the International Commission on Zoological Nomenclature 15: 253-328.Packard, A. S. 1869. The characters of the lepidopterous family Noctuidae. Proceedings of the Portland Society of Natural History 1: 153-156. 1895. On a new classification of the Lepidoptera. American Naturalist 29: 636-647, 788-803. Patterson, C. 1982. Morphological characters and homology. In Joysey, K. A. & Friday, A. E. Problems in phylogenetic reconstruction. Systematics Association special volume 21: 21-74. London.Pierce, F. N. 1909. Thegenitalia of the group Noctuidae of the Lepidoptera of the British Islands, xii + 88 pp. Liverpool.Pierce, F. N. & Beirne, B. P. 1941. The genitalia of the British Rhopalocera and the larger moths, xiv + pp. 15-66. Oundle.Pinhey, E. C. G. 1975. Moths of Southern Africa. 273 pp. Cape Town. 232 IAN J. KITCHING Poulton, E. B. 1887. Notes in 1886 upon lepidopterous larvae, etc. Transactions of the Entomological Society of London 20: 281-321.Prout, A. E. 1929. A catalogue of the Lepidoptera of Hainan. - Noctuidae. Bulletin of the Hill Museum 3: 1-12.Pryer, H. 1883-1885. A catalogue of the Lepidoptera of Japan. Transactions of the Asiatic Society of Japan 11: 216-242; 12: 35-103; 13: 22-67, 228-235.Richards, A. G. Jr 1932. Comparative skeletal morphology of the noctuid tympanum. Entomologica Americana 13: 1-43.Ripley, L. B. 1923. The external morphology and postembryology of noctuid larvae. Illinois Biological Monographs 8: 1-102.Schaus, W. 1901. A revision of the American Notodontidae. Transactions of the Entomological Society of London 34: 257-344.Seitz, A. 1909. III. Section: Noctuiformes, noctuiform moths. In Seitz, A., The Macrolepidoptera of the World*. 3-4. Stuttgart.Shepard, H. H. 1930. The pleural and sternal sclerites of the lepidopterous thorax. Annals of the Entomological Society of America 23: 237-254.Smith, J. B. 1882-1883. A synopsis of the North American genera of the Noctuidae. Bulletin of the Brooklyn Entomological Society 4: 47-52; 5: 3-6, 11-14, 19-22, 27-30, 33-36, 43-46, 53-56.1891. List of the Lepidoptera of boreal America, v + pp. 6-124. Philadelphia. 1893. Catalogue of the lepidopterous superfamily Noctuidae found in boreal America. Bulletin of the United States National Museum 44: 1-424. 1895. Contribution toward a monograph of the insects of the lepidopterous family Noctuidae of boreal North America. - A revision of the deltoid moths. Bulletin of the United States National Museum 48: 1-129.Smith, J. B. & Dyar, H. G. 1898. Contributions toward a monograph of the lepidopterous family Noctuidae of boreal North America; a revision of Acronycta (Ochsenheimer) and of certain allied genera. Proceedings of the United States National Museum 21: 1-194.South, R. 1908-1909. The moths of the British Isles. 1, vi + 355 pp. 2, vi + 376 pp. London. - 1961. The moths of the British Isles. 1. Edited and revised by H. M. Edelsten & D. S. Fletcher. 427 pp. London. Stainton, H. T. 1857. A manual of British butterflies and moths. 1, xii + 338 pp. London.Staudinger, O. & Rebel, H. 1901. Catalog der Lepidopteren des palaearctischen Faunengebietes. Theil I: Famil. Papilionidae-Hepialidae. xxx + 411 pp. Berlin.Staudinger, O & Wocke, M. F. 1871. Catalog der Lepidopteren des europaeischen Faunengebietes. xxxviii + 426pp. Dresden. Swinton, A. H. 1877. On an organ of hearing in insects with special reference to Lepidoptera. Entomolo-gist's Monthly Magazine 14: 121-126.Timlin, J. S. 1955. The classification of the larval stages of some British Agrotidae. Unpublished Ph.D. thesis, University of London. 298 pp.Turner, A. J. 1920. A revision of the Australian Noctuidae. Transactions of the Royal Society of South Australia 44: 120-189.Turner, H. J. 1926-1948. Supplement to Tutt's British Noctuae and their varieties. 1, 364 pp. 2, 284 pp. 3, 295 pp. 4, 94 pp. London.Tutt, J. W. 1891-1892. The British Noctuae and their varieties. 1, xvi + 164 pp. 2, xviii + 180 pp. 3, xxiv + 140 pp. 4, xxviii + 144 pp. London.1895. Review: Smith, J. B., Catalogue of the lepidopterous superfamily Noctuidae found in boreal America. Entomologist's Record and Journal of Variation 6: 69-72. 1896. On the structural affinities of the genus Demas. Canadian Entomologist 28: 81-82.1902. British Moths, xii + 368 pp. London. Van Son, G. 1933. A revision of South African moths of the tribe Nolini. Annals of the Transvaal Museum 15: 181-232.Varley, G. C. 1962. A plea for a new look at Lepidoptera with special reference to the scent distributing organs of male moths. Transactions of the Society for British Entomology 15: 29-40.Viette, P. 1962. Les Noctuidae Hyblaeinae de Madagascar (Lep.). Bulletin Mensuelde la Societe Linneenede Lyon 30: 191-194. - 1962-1967. Noctuelles trifides de Madagascar, ecologie, biogeographie, morphologic et taxonomie(Lep.). Annales de la Societe Entomologique de France 131: 1-825. -1973. Nouvelles noctuelles de Madagascar (Lepidoptera). Nouvelle Revue d' Entomologie 3: 185-191. REVIEW OF THE HIGHER CLASSIFICATION OF THE NOCTUIDAE 233 Walker, F. 1856-1858. List of specimens of lepidopterous insects in the collection of the British Museum. 9: 1-252; 10: 253-491; 11: 493-764. London. Warren, W. 1906-1914. Noctuidae. In Seitz, A., The Macrolepidoptera of the World 3: 9-508. Stuttgart.1912-1938. Noctuidae: Acronictinae - Amphipyrinae. In Seitz, A., The Macrolepidoptera of the World 11: 31-352. Stuttgart.Warren, W. & Seitz, A. 1906-1912. Cymatophoridae. In Seitz, A., The Macrolepidoptera of the World 2: 321-332. Stuttgart.Watson, A., Fletcher, D. S. & Nye, I. W. B. 1980. In Nye, I. W. B., The generic names of moths of the World. 2, xiv + 228 pp. London.Yagi, N. & Koyama, N. 1963. The compound eye of Lepidoptera: approach from organic evolution. 319 pp. Tokyo.Zimmerman, E. C. 1958. Insects of Hawaii. 7, Macrolepidoptera. ix + 542 pp. Honolulu. Index Only the more important references to higher taxon names are included and those taxa mentioned inpassing are omitted. The indexed authors are those whose works are discussed at length. Abrostolini217Acontianae 183Acontiidae 178Acontiinae 170, 197, 202, 206, 219, 225 Acontiini219,225Acronictinae 206, 220, 226Acronictini 220Acronyctinae 183, 197, 202Acronyctinae: first series 202Acronyctinae: second series 202Aganainae214,223Agaristidae 170, 185, 187, 197, 203Agaristinae 206, 220, 226Agrotidae 172, 174Agrotinae 174, 182, 197, 200Agrotini 222Amphipyrinae (sensu Franclemont & Todd)206,221 Amphipyrinae (sensu Tutt) 177Amphipyrini221Aniclini 222Anomiini217,225Antitypini 221Apameini 221Apaminae 177Apatelidae 172, 174Arctiadae 181Arctiidael70,214,223Argyrogrammini 217Autographini217Aventiidae 178 Bagisarini219Bombyciae 166Bombyciformes 161, 168Bombycoidae 177Brephidae 158, 167, 170, 174Bryophilini220,226 Calocampinae 177CaradrininalSSCaradrininae 177 Catocalidae 178 Catocalinae (sensu Franclemont & Todd)206,216,225Catocalinae (sensu Hampson) 184, 202 Catocalinae (sensu Packard) 165, 174Chloephorinae 197, 218, 225Cleocerini 221Collomenini219Colocasia 158, 168, 169, 172, 179Cucullianae 183 Cuculliinae 197, 200, 206, 221, 226Cuculliini221,222Cydosiini219Cymatophoridae 158, 167, 171, 179 Deltoides/Deltoides 165, 169, 178 Deltoidinae 171 Diloba 169, 179, 184, 187, 193, 209 Epicausis 184 Erastrianae 183 Erastriinae 197 Erebinae 202 Erebine-catocaline complex 196 Eriopygini 222 Eublemmini219 Eustrotiim'219 Eutelianae 183 Euteliinae 171 , 197, 202, 206, 218, 225Extensae 163 Fasciatae 166Feraliini 22 1,222Focillinae 171FORBES, 1954 198FRANCLEMONT &TODD, 1983204,213 GARDNER, 1946-1948 191Genuinae 161, 169Geometriform noctuides 178 Glottulini222,226Gonopterinae 171Grotellini223GUENEE, 1852-1854 159 Hadeninae 183, 200, 207, 222, 226 Hadenini 222 HAMPSON, 1898-1913 181 Heliothidinae (sensu Warren) 186 Heliothinae 178, 207, 222, 226 Herminiidae 179 Herminiinae 196, 203, 205, 215 Hyblaea208,209 Hyblaeidae 209 Hyblaeinae 185, 189 Hypenidae 178 Hypeninae 175, 185, 196,203,205, 216, 225Hypenodinae 203, 205, 216, 225 Immature stages 190Intrusae 163, 169 Limbatae 164, 169Lithophanim'221,226 Minores 162, 169Mominae 184 Nocloini221 Noctua 154, 172, 188 Noctuae 155, 166 Noctuae fasciatae 155 Noctuae nonfasciatae 155 Noctuelides 155 Noctuelitae 166 Noctuelites 161 Noctuidae (sensu Hampson) 170, 182,185,187,225Noctuidae (sensu Herrich-Schaffer) 158,170 Noctuidae (sensu Janse) 189, 204Noctuidae (sensu Meyrick) 167 234 Noctuidae (sensu Tutt) 177 Noctuides 176 Noctuina 170 Noctuinae (sensu Franclemont & Todd)207,222,226Noctuinae (sensu Hampson) 185Noctuinae (sensu Packard) 165Noctuinae (sensu Tutt) 177Noctuini 222Noctuoideal89,204Nocturnes 157Nolidael79,198,203Nolinael81,206,219,225Nonfasciatae 166Nycteola 175Nycteolidae 158, 179Nycteolinae 170 Oncocnemidini 221 , 222, 223Orthosiinae 177 Palindiinae 171 IAN J. KITCHING Pantheinae 180, 197, 202, 206, 219, 225,226Patulae 164Phalaena 154, 189Phalaenidae 189Phalenidae 178Phytometrinae 184Plusiadae 167Plusiinae 178, 197, 202, 206, 217, 223, 226 Plusiini217Poliinae 197Polypogoninae 185Psaphidini221,222,226Pseudo-Deltoidae 165 Quadrifidael61,162, 169Quadrifinae 171 RICHARDS, 1932 196 Risobini 219 Rivulinae 196, 206, 216, 225 Sarrothripinae 171, 183, 197, 202, 206,208,218,225Sarrothripini219SEITZ, Die Gross-Schmetterlinge der Erde 185Sericiae 162Serpentiae 164Serpentides 178 Stictopterinae 171 , 183, 197, 217, 226Stiriini221,223,226 Thyatiridae 167, 170Toxocampidae 178Trifidael61,168,169Trifinae 170Tympanal organs 195 Ufeini222Variegatae 162Xylenini221,222 British Museum (Natural History) Milkweed butterflies: their cladistics and biology P. R. Ackery & R. I. Vane- Wright The Danainae, a subfamily of the Nymphalidae, contains only some 150 species, yet aspects oftheir biology have stimulated far more attention than can be justified by species numbersalone. In recent years, an expansive literature has grown, considering aspects of theircourtship and pre-courtship behaviour, migration, larval hostplant associations, mimicry andgenetics. The popularity of danaines among biologists can certainly be attributed to thiscombination, within one small group, of so many of the factors that make butterflies such aninteresting group to study. The obvious need to place this wealth of biological data within anacceptable systematic framework provided the impetus for this volume. Started eight years ago within the conventions of evolution by natural selection andHennig's phylogenetic systematics, the book is now largely about natural history (what theanimals have and do, where they live and how they develop) and natural groups - as revealedby a form of analysis approaching that practised by the new school of 'transformed cladistics'.The authors have prepared a handbook that will appeal to a wide range of biologists, frommuseum taxonomists to field ecologists. 424 pp (approx. ), 12 pp colour, 73 b/w plates, line and graphic illustrations, maps, extensive bibliography.ISBN 565 00893 5. Publication September 1984. Price 50, prepublication price 45. Titles to be published in Volume 49 Afrotropical jumping plant lice of the family Triozidae (Homoptera: Psylloidea). By David Hollis The taxonomy of the western European grasshoppers of the genus Euchorthippus, with specialreference to their songs (Orthoptera: Acrididae). By D. R. Ragge & W. J. Reynolds An historical review of the higher classification of the Noctuidae (Lepidoptera). By Ian Kitching The Pimplinae, Xoridinae, Acaenitinae and Lycorininae (Hymenoptera: Ichneumonidae) ofAustralia. By I. D.Gauld The Palaearctic species of Ascogaster (Hymenoptera: Braconidae) By T. Huddleston Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, SuffolkPrinted in Great Britain by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) The Pimplinae, Xoridinae,Acaenitinae and Lycorininae(Hymenoptera: Ichneumonidae)of Australia I. D. Gauld Entomology series Vol49 No 4 20 December 1984 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique andever-growing collections of the Museum, both by the scientific staff of the Museum and byspecialists from elsewhere who make use of the Museum's resources. Many of the papers areworks of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself,available separately, and individually priced. Volumes contain about 300 pages and severalvolumes may appear within a calendar year. Subscriptions may be placed for one or more ofthe series on either an Annual or Per Volume basis. Prices vary according to the contents ofthe individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History),Cromwell Road, London SW75BD,England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.) Trustees of the British Museum (Natural History), 1984 The Entomology series is produced under the general editorship of the Keeper of Entomology: Laurence A. Mound Assistant Editor: W. Gerald Tremewan ISBN 565 06006 6ISSN 0524-6431 British Museum (Natural History)Cromwell RoadLondon SW75BD Entomology seriesVol49No4pp235-339 Issued 20 December 1984 The Pimplinae, Xoridinae, Acaenitinae andLycorininae (Hymenoptera: Ichneumonidae) ofAustralia I. D. Gauld Department of Entomology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Contents Synopsis 235 Introduction 235 Material examined 236 Checklist of Australian Pimplinae , Xoridinae , Acaenitinae and Lycorininae 237 Subfamily Pimplinae 238 Zoogeography 239 Biology 240 Key to genera of Pimplinae occurring in Australia 242 Tribe Ephialtini 244 Tribe Polysphinctini 254 Tribe Pimplini 268 Tribe Delomeristini 310 Tribe Rhyssini 316 Subfamily Xoridinae 319 Zoogeography 320 Biology 320 , Subfamily Acaenitinae 322 Zoogeography 322 Biology 322 Subfamily Lycorininae 327 Acknowledgements 331 References 331 Index to hosts 338 Index to Ichneumonidae 338 Synopsis The Australian species of the ichneumonid subfamilies Pimplinae, Xoridinae, Acaenitinae and Lycorini-nae are revised and keys provided to the 20 genera and 76 species occurring on the continent. A total of 35species are newly described but four are not formally named as their status needs further investigation. Theremaining 41 species are redescribed and their diagnostic features emphasized. Five new specific synonymsare proposed. For each species notes are given about geographical distribution, habitat preference andhost range. Introductory sections for each subfamily include notes on zoogeography and a brief discussionof group biology. A checklist of Australian species, and indexes to hosts and parasitoids, complete thework. Introduction Amongst the most important of the natural enemies of insect pests are the parasitic Hymenop-tera, a very large group of animals whose larvae develop at the expense of other insects (Askew,1971). Under normal circumstances the populations of many injurious insects are severelylimited by the attacks of Parasitica and in several countries, including Australia, the ravages ofaccidentally imported pests have been curtailed by the introduction of one or more hymenopter-ans (Muldrew, 1967; Taylor, 1978). Other introduced pests have been severely attacked by Bull. Br. Mus. not. Hist. (Ent.) 49 (4): 235-339 Issued 20 December 1984 236 I. D. GAULD native Australian parasitoids (Tryon, 1900). During the past 20 years there has been an upsurgein interest in using Hymenoptera and other organisms for purposes of pest control (Wilson,1960; Huffaker & Messenger, 1976) as an alternative to costly, ineffective and environmentallydestructive chemical methods (Bosch, 1978). However, for biological control programmes to besuccessful an intimate knowledge is necessary of the life history and interactions of the pest andits parasites. A sound taxonomic basis is vital for the development of such knowledge (Hardy,1982), for such work permits the accurate identification of an organism and hence providesconstancy and universality in the usage of names, a prerequisite for the national and internation-al communication of information. The present work is a taxonomic study of four rather distantly related subfamilies ofAustralian Ichneumonidae, one of the largest groups of Parasitica. The primary aim of the paperis to provide practical keys for the identification of these insects and to collate informationconcerning their distribution, habitat preference and host range. These subfamilies, the Pimplinae, Acaenitinae, Lycorininae and Xoridinae, contain anumber of species common in agricultural, horticultural and forest ecosystems, includingseveral that parasitize notorious lepidopterous and symphytan pests. No comprehensiveaccount of the Australian species of these subfamilies has ever been published, although one ortwo of the pimpline genera have been revised recently for the whole of the Indo-Australianregion (Gupta, 1962;'Townes & Chiu, 1970; Gupta & Tikar, 1978). Several new Australianspecies of these genera are described herein, and simpler local keys provided for identification. The inclusion of the four taxa Pimplinae, Acaenitinae, Lycorininae and Xoridinae together ina single work is largely a matter of convenience. The first subfamily is by far the largest andconstitutes the major part of the study; the remaining three taxa are very small and share certainsuperficial similarities, making it sensible to deal with them at the same time. Although notclosely related, these four subfamilies, together with the Banchinae and part of the Labeninae,were included by older authors (e.g. Morley, 19130; Szepligeti, 1914) in one of the five classicalichneumonid subfamilies, the Pimplinae s.l. This heterogeneous taxon was characterized bypossession of a fairly broad first tergite with subcentral or antecentral spiracles (sessile gaster)and, in females, by possession of a long ovipositor (though in fact the ovipositor is very short insome species ofXanthopimpla). All Australian species will run to 'Pimplinae' in Riek's (1970)key. The subfamilies may easily be separated by reference to Gauld's (1984) recent monographon Australian ichneumonids. The Lycorininae, Xoridinae and Acaenitinae are extremelydistinctive taxa and easily recognized by the possession of unique features (the co-adapted'catch' between the metathorax and propodeum of lycorinines; the transcarinate scuto-scutellargroove and geniculate female flagellum of xoridines; the elongate female subgenital plate andaccessory tooth on the tarsal claw of acaenitines). Care is required to separate some Pimplinaefrom Labeninae but species of the latter group have a larger clypeus and higher position ofinsertion of the gaster than pimplines (Townes, 1969; Gauld, 1984). None of these subfamiliesshould be confused with the other 'pimpline' taxon, the Banchinae, as all banchines have adorsal subapical notch in the ovipositor; such a notch is not present in any Pimplinae,Acaenitinae, Lycorininae, Xoridinae or Labeninae. The terminology in this work follows that of Gauld (1984). Material examined The study is based on examination of almost all specimens available in collections of Australianichneumonids. Special attention was paid to collections in agricultural institutions that containlarge numbers of reared specimens. Examination of these collections has been supplemented byextensive collecting, particularly in Tasmania and the south-east. Although the resulting sampleis thought to be fairly representative of the fauna of the more humid eastern part of thecontinent, relatively little material has been seen from the north and west. Smaller, crypticspecies, such as many polysphinctines, are probably very unrepresented.The following abbreviations have been used for museums containing Australian material. AM Australian Museum, Sydney, New South Wales, Australia ICHNEUMONIDAE OF AUSTRALIA 237 ANIC Australian National Insect Collection, Canberra, Australian Capital Territory, Australia BMNH British Museum (Natural History), London, U.K. BPBM Bernice P. Bishop Museum, Honolulu, Hawaii, U.S.A. DAR Department of Agriculture, Rydalmere, New South Wales, Australia DAT Department of Agriculture, Hobart, Tasmania, Australia DFT Department of Forestry, Hobart, Tasmania, Australia DPIQ Department of Primary Industries, Indooroopilly, Queensland, Australia LSL Linnean Society, London, U.K. MNHN Museum National d'Histoire Naturelle, Paris, France MNHU Museum fur Naturkunde der Humboldt-Universitat, Berlin, D.D.R. NM Naturhistorisches Museum, Vienna, Austria NMV National Museum of Victoria, Melbourne, Victoria, Australia PANS Philadelphia Academy of Natural Sciences, Philadelphia, Pennsylvania, U.S.A. QM Queensland Museum, Fortitude Valley, Queensland, Australia QUM Queensland University Museum, Brisbane, Queensland, Australia TC Townes Collection, Ann Arbor, Michigan, U.S.A. TM Termeszettudomanyi Museum, Budapest, Hungary UL Universite Laval, Quebec, Canada UM University Museum, Oxford, U.K. USNM United States National Museum, Washington D.C. , U.S.A. UZM Zoologisk Museum, Copenhagen, Denmark WAM Western Australian Museum, Perth, Western Australia, Australia ZMA Zoologisch Museum, Amsterdam, The Netherlands Checklist of Australian Pimplinae, Xoridinae, Acaenitinae and Lycorininae PIMPLINAE EPHIALTINIGenus A CROPIMPLA Townes xantha sp. n. Genus CAMPTOTYPUS Kriechbaumerbicolor Kriechbaumer atropos (Morley) syn. n.lachesis (Morley)scllatus Kriechbaumer flaviceps (Cameron) syn. n.clotho (Morley) syn. n.Genus PARVIPIMPLA Gauld petita Gauld Genus SERICOPIMPLA Kriechbaumeraustralis Townes, Townes & Gupta annulipes (Cameron)crenator(F.) pilosella (Cameron)consimilis (Morley)lu ten sp. n. Genus ZAGLYPTUSFoersterglabrinotum (Girault)hollowayisp. n. POLYSPHINCTINIGenus ACRODACTYLA Haliday cursor sp. n. in/cans sp. n. quadrisculpta (Gravenhorst) zekhem sp. n.Genus DREISBACHIA Townes lutcii sp. n.Genus ERIOSTETHUS Morley carinatus Baltazar maximussp. n.minimus sp. n.perkinsi (Baltazar)pulcherrimus MorleyGenus ZATYPOTA Foersterbingilisp. n.celersp. n.dandiensis sp. n.kauros sp. n.phraxos sp. n.rennefersp. n.stcllatu sp. n.velata sp. n. PIMPLINIGenus ALOPHOPIMPLA Momoi kluia sp. n. Genus ECHTHROMORPHA Holmgrenagrestoria (Swederus) melioratorius (F.) interrupta (Brulle) insidiator (Smith) platymischa (Vachal) striata Krieger conopleura Krieger immaculate Krieger notulatoria var. immaculate Morley diversor Morleyintricatoria (F .) excavata (Guillou) 12-guttata (Ashmead)nigricornis (Smith) maxima Krieger fastigata Krieger 238 I. D. GAULD Genus LISSOPIMPLA Kriechbaumer atra Girault excelsa (Costa) semipunctata (Kirby)10-notata Kriechbaumer8-guttata Kriechbaumerrufipes (Tryon)priocnemidea Vachal obesa sp. n. scutata Krieger species 1Genus XANTHOPIMPLA Saussure anton sp. n. unkhu sp. n. arealis Krieger gracilis Kriegerbeauforti Cameronpapuana Cameron australis Kriegersimitts Krieger haralfsn n Wnod^Townes & Chiu ecaudate Krieger ida Krieeer Subgenus NOMOSPHECIA Gupta melanosoma MorleySubgenus PAREMA Guptapenetrans (Smith) cephalotes Krieger fumata Krieger papuana CameronSubgenus THERONIA Holmgrenfraucai sp. n.maculosa Krieger viridicans Morleysteindachneri Krieger dubia Krieger teiae (Cameron) antherae (Cameron) fumipennis Morley syn. n. claripennis Morley ffavolineata Cameron hyaloptiaKnegeTxanthostiema GiraultChessman frafercu7ijs Townes & Chiuhiatus Townes & Chiu Genus EPIRHYSSA Cresson feiroi Mocsary comb. rev.Genus ^EGARHYSSA Ashmead nortom (Bresson quebecemis (Provancher) Genus RHYSSA Gvenhotpersuasona (L.) XORIDINAE Ggnus joiUDISSLatieille Subgenus CYANOXORIDES Cameron ,. . /c , ,. 4 .,ausfra*ens,s (Szephgeti) crude/I* (Turner) . ~ . ,7^,^,,^ A rACAENITINAE , . /c ...\caudata (Smith) T^ Vcrassa Krieeer ochracea peterseni Townes & Chiu syn. n.pubidorsis Townes & Chiuquadridens Townes & Chiurhopaloceros Krieger xanthopimploides (Girault)srriata Townes & Chiu5ummervi//ei (Girault)fenninato (Brulle) DELOMERISTINIGenus THERONIA Holmgren amaryllyxsp. n. . / 7 . "L, >.apicipciwis (Turner) ' "' taiyxsp. n. Genus canfeerrae sp. n. splendid u In sp. n.turner] sp. n. species 1 Subfamily PIMPLINAE The Pimplinae is a moderately large subfamily, but as certain genera contain one or two large,strikingly patterned and common species, it is often the numerically best represented subfamilyin general collections. Such collections will be found to be dominated by only three or fourspecies, typically Pimplini, that are common in suburban situations, and also, less frequently, byrhyssines, conspicuous on account of their exceptional size and long ovipositors. Such a situationapplies throughout most of the world. In Australia, collections are dominated by species of thegenera Lissopimpla and Echthromorpha, whilst in Palaearctic collections the related and ICHNEUMONIDAE OF AUSTRALIA 239 biologically similar genera Pimpla and Itoplectis predominate. The numerous other smaller ormore cryptic pimplines are less often collected and, although many appear to be quitewidespread, few are usually present in most museums. Malaise traps generally yield smallnumbers, but good series can often only be taken by more specialist collecting, such as carefulexamination of tree bark inhabited by spiders, or 'stalking' the brightly coloured arborealXanthopimpla species in forest understorey vegetation. Many species are only ever encounteredin one particular habitat and these 'rare' taxa may prove to be very common when one knowsjust where to look. This ecological specialization coupled with the great range of biologicaldiversity exhibited by the subfamily makes the Pimplinae one of the most attractive of allichneumonid groups for biological investigation. The Pimplinae is currently subdivided into seven tribes, Delomeristini, Ephialtini, Rhyssini,Pimplini, Neoxoridini, Polysphinctini and Diacritini, but authors have not reached agreementabout their exact limits (compare Finlayson, 1967 and Townes, 1969) . The Ephialtini is probablythe most primitive group, and is the paraphyletic 'grade-group' from within which all the othergroups have probably arisen. The Delomeristini appears to be a polyphyletic assemblage and, asit is not yet possible to unambiguously assign its members to other taxa, for the present it is leftintact as a tribe. The remaining five tribes seem to be natural holophyletic (in the sense ofHennig, 1966) groups, although the definition of one, the Polysphinctini, can be altereddepending upon whether a classification is formulated on larval or adult characters (Townes,1969; Gupta & Tikar, 1978). Throughout the world approximately 76 genera are recognized, butdisagreement over some generic limits means this number is liable to fluctuate for some time.Only 17 of these genera are known to occur in Australia (Gauld, 1984), and several of these (e.g.Alophopimpla, Acropimpla and Camptotypus] are restricted to tropical Queensland; others,such as Xanthopimpla and Zatypota, appear to be more diverse in the north. In the following account 62 Australian named species are formally described, of which 23 arenew. One morphological segregate of Lissopimpla is keyed separately and described but itstaxonomic status remains uncertain. Zoogeography For the Indo-Australian region as a whole the general picture of pimpline distribution is one ofrapid decrease, both in number of species and genera, from the Asian continental shelfeastwards. This is particularly the case for the Ephialtini which is represented by 20 genera inSouth East Asia, six in New Guinea and five in Australia, and for the Rhyssini which has sevengenera in Asia, five in New Guinea but only one present in Australia. Acropimpla is representedby 27 species in South East Asia, two in New Guinea and one in Australia whilst Gupta (1962)demonstrated a progressive west-east reduction in numbers of species and species-groups ofTheronia. In some groups New Guinea has apparently acted as a centre for secondary radiation.This is particularly noticeable in the case of Xanthopimpla for the splendens-group (Townes &Chiu, 1970) and is possibly also true for Camptotypus. In both these cases the genera seem tohave had an Asian origin. There is a marked relationship between the pimpline fauna of Australia and New Guinea. Ofa total of 42 species (excluding polysphinctines) found in Australia, 17 (i.e. 40%) also occur inNew Guinea. Many of the species apparently endemic to Australia have their closest or veryclose relatives to the north. This is particularly obvious for many species of Xanthopimpla,Zaglyptus, Acropimpla and Alophopimpla. In none of these genera is there any indication ofsubstantial radiation in Australia, though in a few cases (e.g. Zaglyptus) sister-species pairs dooccur there. The most obvious interpretation of this data is to suggest that the Australian fauna has beenderived from that of the Old World tropics, presumably within the last five million or so years, asbefore that time the continent was remote from South East Asia (Audley-Charles et al. , 1981).Any hypothesis that advocates a southern origin for these pimplines needs to account for theirrapid spread into South East Asia and their extreme radiation, despite possible competitivepressure from numerous Old World pimplines (as many genera are not present in Australia). 240 I. D. GAULD Furthermore there are no obvious affinities between the Australian pimpline fauna and that ofthe southern neotropics. The only endemic Australian genus, Parvipimpla, seems to be relatedto Camptotypus and most pimpline and ephialtine Australian genera are not present in SouthAmerica. The few that are (e.g. Xanthopimpla, Zaglyptus) are tropicopolitan taxa and theAustralian species belong to quite different species-groups from New World taxa. The same istrue for Theronia; Neotropical species belong to an exclusively Neotropical subgenus (Townes& Townes, 1966) and are not at all closely related to Australian species. The polysphinctines have been excluded from the above discussion for the reason thatvirtually nothing is known of them in any region other than north temperate areas. TheAustralian Dreisbachia is probably very closely related to a New Guinea species (Momoi, 1966)and Eriostethus has numerous northern congeners. This latter genus was thought by Gauld(1984) to have possible close phylogenetic affinity with some Neotropical species, but subse-quent study suggests that the apparent similarities may be the result of evolutionary convergence(seep. 259). Biology Biologically the Pimplinae is both one of the most diverse and one of the more primitivesubfamilies of Ichneumonidae. Pimplines utilize a wide range of holometabolous insect larvaeand pupae as hosts, though a few even attack spiders' egg sacs. One specialized group, thePolysphinctini, is unique amongst ichneumonids in parasitizing immature spiders. In generalpimplines attack relatively mature hosts and usually curtail further host-development byenvenomation during parasitization (Chrystal & Skinner, 1932; Morgan & Stewart, 1966;Spradbery, 1968; Price, 1973). The pimpline's venom, produced by accessory glands associatedwith the reproductive system (Togashi, 1963), is injected by the ovipositor prior to oviposition.It may kill the host or induce permanent paralysis (Cushman, 1926). Smithers (1956) observedthat larvae of a psychid stung by Sericopimpla (but not parasitized) remained immotile butusually 'fresh' for over two months. The majority of pimplines are ectoparasites of concealedendopterygotes and such immobilization is presumably necessary to prevent the host damagingthe pimpline egg or young larva. Smithers (1956) observed that Sericopimpla eggs artificiallyplaced on unparalysed psychid larvae were destroyed prior to eclosion. Envenomation by theendoparasitic Pimplini may be necessary to reduce the host's haemocyte reaction which isusually more pronounced in more mature larvae (Puttier, 1961; Bosch, 1964). However, Carton(1973) observed that careful placement of the egg was necessary for Pimpla to minimizeencapsulation, whilst Fiihrer & Kilincer (1972) noted that the newly hatched larvae of someother species migrate rapidly to the host's head and destroy the cephalic ganglion by histolysis.Unlike other pimplines, polysphinctines generally do not curtail host development. Species ofPolysphincta andAcrodactyla have a powerful paralysing venom which they use to immobilize aspider. After stinging and attaching an egg the pimpline departs and the spider recovers and livesactively (Nielsen, 1923; Cushman, 1926). Parasitoids, such as most Pimplinae, that adopt the strategy of attacking mature hosts,generally lay large, well-developed eggs (Price, 1975) and undergo quite rapid larval develop-ment. The size of pimpline eggs can readily be appreciated from the tables given by Iwata (1958).A species of Sericopimpla has 'sausage-shaped' eggs 2-4 x 0-4 mm whilst a specialist endoparasi-toid of similar body size (e.g. Meniscus) has eggs between 0.33 x 0.08 mm and 0.71 x 0. 16 mm,that is around one-twentieth of the volume. Even very small pimplines, such as Zatypota, haveeggs 0.74 x 0.22 mm in size. Many very large species, such as Megarhyssa, have extremelyelongate eggs (up to 14 mm long) which deform to facilitate passage down the long ovipositor(Spradbery, 1970). To successfully form such large eggs many female pimplines need protein-rich food. In addition to feeding at flowers, like the males, these females frequently consumehaemolymph exuding from the punctures they make in pupae used as hosts (Graham, 1947;Leius, 1960). Some pimplines seek out and kill prey to obtain such fluid nourishment (Cole,1967) and, as the prey is often the larvae or pupae of the host species, the ichneumonid is often amore important mortality factor than estimation of percentage of parasitism may suggest. ICHNEUMONIDAE OF AUSTRALIA 241 As pimpline eggs are large, adult females have relatively few mature oocytes in the oviduct atany one time. Iwata (1960) calculated the average number to be about six with a maximum ofgenerally less than 20, except in a few species such as Iseropus which are gregarious parasites.Banchines, which are similar-sized ichneumonids, but specialist endoparasites, have on average53 mature oocytes, with a maximum of 176. The eggs of many pimplines are deposited on or nearthe paralysed larva (Smithers, 1956; Morgan & Stewart, 1966) though those of Pimplini areplaced within the host's haemocoel, often in the thorax (Carton, 1973). One striking feature ofsome pimplines is their apparent inefficiency in oviposition. Smithers (1956) observed thatSericopimpla may attempt to sting a psychid larva over 100 times, and is often engaged for anhour or more attacking a single individual. Morgan & Stewart (1966) observed that Rhyssa madebetween five and 12 insertions of the ovipositor to make a single successful attack. Pimpline eggs hatch quite rapidly, often one or two days after oviposition, and early larvaldevelopment proceeds very rapidly with generally only one or two days between ecdyses(Rosenberg, 1934; Smithers, 1956; Rojas-Rousse & Benoit, 1977). The final larval instar is oflongest duration, frequently exceeding the length of the sum of all others. This contrastsmarkedly with some specialist endoparasitoids where the first larval instar is extremelyprotracted (Tothill, 1922). The exact number of instars pimpline larvae pass through is difficultto ascertain (Rojas-Rousse & Benoit, 1977) but may often be five or six; individual intraspecificvariation may occur (Smithers, 1956). The cocoon of pimplines is generally only rudimentarythough some polysphinctines spin thick and very characteristic ones (Figs 98, 99). Diapause,when it occurs, is usually in the prepupal stage though at least one European species aestivates asan adult (Cole, 1967). As most pimplines incapacitate their hosts they need little physiological or cytologicalcompatibility with the latter as do many specialist endoparasites (Salt, 1968). Consequentlypimplines often utilize a very wide range of hosts. This is particularly true for some Pimpliniwhich search disturbed habitats for pupae. For example Echthromorpha intricatoria has beenreared from pupae of Anthelidae, Lymantriidae, Noctuidae and Nymphalidae. Many species ofEphialtini and some Theroniini are specialized, not immunologically, but in their adaptationsfor reaching particular hosts. In such groups the apex of the ovipositor is often characteristic of aspecies (Townes & Townes, 1960). Although these morphological adaptations are associatedwith a particular niche (e.g. decaying wood), within each niche quite different species may beused as hosts. For example, one Nearctic species of Dolichomitus is restricted to timber-borersin certain trees, but will parasitize larvae of Buprestidae, Cerambycidae and Sesiidae if available(Townes & Townes, 1960). Pimplinae not only often utilize a wide range of hosts, but manyattack a variety of different-sized hosts. Consequently size variation in adults of some speciesmay be very large. For example, females of Rhyssa persuasoria in Europe have a fore winglength range of 9 to 25 mm. Females of some species are facultatively arrhenotokous, layingfertilized (i.e. female) eggs in large hosts and unfertilized (male) eggs in smaller ones. Thus thefemales are, on average, larger than the males. The evolutionary biology of the Pimplinae is a fascinating subject, and although incompletelyknown both from phylogenetic and biological standpoints, more information exists for the groupthan probably any other ichneumonid subfamily. The ancestral ichneumonoids are believed tohave been parasites of wood-borers (Konigsmann, 1978) such as cerambycid or siricid larvae.Such habits are retained by a number of structurally rather primitive ephialtines. The Rhyssiniand Poemeniini are biologically similar though the adults are somewhat more specializedmorphologically than ephialtines. They are difficult to place phylogenetically (each beingcharacterized by several autapormorphies) but may represent early branches from the ephial-tine stock. Several evolutionary lineages of ephialtines have given rise to single species thatparasitize aculeates nesting in old borings in timber (e.g. Jussila & Kapyla, 1975). Such a switchfrom timber-boring larvae to larvae (usually aculeate) inhabiting timber borings is a commonfeature in ichneumonids and occurs amongst labenines (Gauld, 1983) and gabuniine mesoste-nines (Gupta & Gupta, 1983). A large number of species related to ephialtine parasites oftimber-borers have become associated with endopterygotes that pupate on cracks in trees,under bark etc. (Rosenberg, 1934; Perkins, 1942). Further evolutionary change has followed 242 I. D. GAULD one of three major biological pathways - progressive association with cocoons; association withborers in non-ligneous or partially lignified tissue; or specialization on more or less exposedhosts. Whilst in evolutionary terms the latter may have been derived from either of the formerthese different biological adaptations represent three alternative areas of specialization forpimplines. One apparent phyletic lineage of Ephialtini (the Tromatobia-subgroup) has special-ized in being cocoon parasites. Oviposition is apparently stimulated by the presence of silk onwhich the ichneumonid alights prior to probing for a host. Several species in this group aregregarious (Iwata, 1961). Amongst the hosts used are psychid larvae, various lepidopterouscocoons and also spider egg-sacs. The polysphinctines are probably derived from this lineage.Presumably the Tromatobia-subgroup arose from ephialtines that probed for cocoons throughbark or other plant tissue. Other ephialtines attack borers in non-lignified or partially lignified tissue such as Pseudopim-pla, which attacks cephids (Bruzzese, 1982) mining Rubus stems, and Alophosternum, whichparasitizes leaf miners (Townes, 1969). Many Scambus species attack hosts of this type. OneEuropean species attacks immature larvae of a bud-mining tortricid (Winter, 1979) whilstanother species probably attacks gall-forming Diptera in reed beds. Some species are gregarious(Iwata, 1961). The species that attack fairly exposed hosts include the majority of Pimplini. They areendoparasitic (a common feature of parasitoids of exposed hosts) but exhibit few of thespecialized immunological mechanisms of more advanced endoparasitic ichneumonids. Thehost is little more than 'a piece of meat' and is either virtually killed by envenomation prior tooviposition or has its cephalic region destroyed at the earliest opportunity by the newly hatchedparasitoid larva (Fiihrer & Kilincer, 1972). One genus of Pimplini, Itoplectis, includes a numberof species that are facultative hyperparasites. A similar adaptation occurs in Australia in somespecies of Theronia, a genus whose species exhibit a considerable range of biological adapta-tions. It is possible that all Theronia species are parasites of hymenopterous larvae, either ofParasitica as hyperparasites through Lepidoptera, or in aculeate nests (Gupta, 1962; Short,1978). A rather similar host range is found in the family Trigonalidae (Clausen, 1940). It is quitepossible that all of the Delomeristini are parasitic only on Hymenoptera and the initiallyapparently heterogeneous facies presented by the genera of this group is an indirect result ofspecialization on different hymenopterous hosts, such as sawflies in the case of Delomerista(Furniss & Dowden, 1941), rhyssines in the case of Pseudorhyssa (Spradbery, 1969), andstem-nesting aculeates in the case ofPerithous (Danks, 1971). A great deal of additional work needs to be undertaken to develop this subject. Particularlyvaluable would be studies of the comparative anatomy of the reproductive systems, analyses ofvenoms and studies of larval morphology. Biological observations on the poorly knownAustralian fauna would also be invaluable, especially if they involved such virtually unstudiedgenera as Alophopimpla, Acropimpla, Camptotypus, Parvipimpla and Epirhyssa. One of theprincipal aims of this work is to stimulate such investigations. Key to genera of Pimplinae occurring in Australia The tribal groupings in the Pimplinae are based to a large extent on structures of the final instar larvae(Finlayson, 1967; Short, 1978). The characters used to place adults in tribes are subtle, subject to variationand difficult for an inexperienced person to appreciate. For ease of identification the key given below isdirect to genus. 1 Mesoscutum without sharp, transverse wrinkles or rugae; last visible tergite of $ gaster not cornute 2 Mesoscutum with sharp transverse wrinkles or rugae; last visible tergite of $ gaster ending in ashort horn 15 2 Epicnemial carina entirely absent; posterior margins of tergites 3-5 incised. 3r-m lacking in fore wing and hind wing with first abscissa of Cu\ shorter than cu-a PARVIPIMPLA(p. 249) Epicnemial carina present, at least ventrally; posterior margins of tergites 3-5 not conspicu-ously incised 3 ICHNEUMONIDAE OF AUSTRALIA 243 3 Hind wing with cu-a more than 1-8 times as long as first abscissa of Cui (Fig. 7) (or sometimes with distal abscissa of Cu^ and M basally united (Fig. 6)); distal abscissa of Cui present; foretarsal claws of 9 simple 4 - Hind wing with cu-a about equal to or shorter than first abscissa of Cu\ (Fig. 8), or with distal abscissa of Cu\ absent; fore tarsal claws of 9 with a basal lobe 8 4 Hind femur with a ventral tooth ; face with vertical impression either side of raised mid-line LISSOPIMPLA(p. 273)Hind femur without a ventral tooth ; face without vertical impressions 5 5 Mandibles not twisted; labrum concealed when mandibles closed; clypeus transverse and entire 6 Mandibles strongly twisted; labrum exposed when mandibles closed; clypeus elongate, or iftransverse then divided by a transverse ridge 7 6 Propodeal spiracle almost circular; propodeum dorsally without carinae; occipital carina dorsally absent ALOPHOPIMPLA(p. 269) - Propodeal spiracle elliptical; propodeum dorsally with lateral and lateromedian longitudinal carinae discernible ; occipital carina complete THERONIA (p. 310) 7 Malar space equal to or longer than basal manidbular width; marginal cell or fore wing with distal infumate spot; fore wing with cu-a distal to Rs&M ECHTHROMORPHA (p. 269) Malar space shorter than basal mandlbular width; marginal cell of fore wing with cu-a more orless opposite base of Rs&M XANTHOPIMPLA(p.216) 8 Fore wing with 3r-m present , enclosing a subtriangular or rhombic areolet 9 - Fore wing with 3r-m absent 11 9 Pterostigma bright yellow, contrasting with infumate wings and dark veins; pleural carina absent (Fig. 9) ; occipital carina obsolescent or absent mediodorsally. . . . CAMPTOTYPUS(p. 245) - Pterostigma blackish, concolorous with other veins; pleural carina present (Fig. 10); occipital carina present mediodorsally 10 10 Eye with a strong indentation opposite base of antenna (Fig. 4); hind ocellus separated from eye by about its own diameter; head behind eyes very strongly narrowed SERICOPIMPLA(p. 249) - Eye with weak indentation opposite base of antenna (Fig. 5); hind ocellus separated from eye by more than its own diameter; head behind eyes evenly rounded ACROPIMPLA (p. 244) 11 Propodeum with lateral subapical tubercles (Fig. 11); ovipositor shaft in profile parallel-sided, the apex abruptly tapered with proximal tooth bearing an elongate, free tip; segment 5 of hind tarsus quite slender, about 2-5 times as long as broad ZAGLYPTUS(p. 252) - Propodeum without lateral subapical tubercles; ovipositor shaft in profile evenly tapered from centre to apex, apical teeth indistinct, the most proximal never with an elongate free tip;segment 5 of hind tarsus stout, 2-0 times or less as long as broad 12 12 Mesoscutum with a small crest near anterior end of notaulus; propodeum in profile rather long, often coarsely sculptured (Fig. 85) ACRODACTYLA(p. 254) - Mesoscutum without an anterior crest; propodeum in profile rather evenly rounded, generally smooth and polished (Figs 88-91) 13 13 Mesoscutum with central lobe evenly and closely pubescent; surface of eye bearing scattered long conspicuous hairs; ovipositor slightly up-curved DREISBACHIA (p. 258) - Mesoscutum without hair centrally; surface of eye with, at most, inconspicuous short hair; ovipositor straight or up-curved near apex 14 14 Hind tibia with a longitudinal hairless furrow on inner surface; ovipositor projecting beyond apex of gaster by more than 0-9 times length of hind tibia ERIOSTETHUS(p. 258) - Hind tibia without a hairless furrow internally; ovipositor projecting beyond apex of gaster by less than 0-5 times length of hind tibia ZATYPOTA(p. 263) 15 Pronotum mediodorsally with a depression centrally, but without a deep transverse furrow separating an anterior lip-like portion; first sternite not fused to tergite; $ with sternites 2-4 with a median pair of tubercles RHYSSA (p. 318) - Pronotum mediodorsally with deep transverse groove separating off a recurved anterior lip; first sternite fused to tergite ; $ with sternites 2-4 bearing tubercles near anterior edge 16 16 Fore wing without 3>r-m; tergite 1 without dorsolateral carinae anteriorly; tergite 2 with thyridia continguous with anterior margin EPIRHYSSA (p. 316) Fore wing with 3r-m present, enclosing rhombic areolet; tergite 1 with dorsolateral carinae onanterior 0- 1 ; tergite 2 with thyridia separated from anterior margin MEGARHYSSA (p. 317) 244 I. D. GAULD Tribe EPHIALTINI(= Pimplini sensu Townes) The Ephialtini is undoubtedly a paraphyletic assemblage as it is the group from within which thePolysphinctini has arisen. It also is possibly paraphyletic with respect to the Pimplini. As aparaphyletic 'group' the Ephialtini can only be characterized by its lack of the apomorphicfeatures of other tribes, but like that notorious paraphyletic taxon, the Reptilia, the Ephialtini isboth practically recognizable and a functionally useful category (vide Charig, 1981). Ephialtines are mostly moderately large to large insects and usually possess an ovipositor thatis as long as or longer than the gaster; this organ is slender, parallel-sided and tapered onlyapically, unlike that of the Polysphinctini which is elongately tapered. The fore wing may or maynot lack 3r-m (this vein is usually present in the Pimplini) and the first abscissa of Cu\ in the hindwing is more than 0-6 times the length of cu-a (it is less than 0-5 in Pimplini). Ephialtines do nothave specialized pulvilli as do polysphinctines, nor the reduced larval hypostoma characteristicof pimplines. Currently 32 genera are recognized; these are divisible into four generic groups, theEphialtes-group, the Alophosternum-group, the Pseudopimpla-group and the Camptotypus-group. The Ephialtes-group, characterized by a more or less complete occipital carina and basallobes present on all female claws, comprises two ill-defined subgroups, the Ephialtes-subgroupwith a medianly dipped occipital carina, flat clypeus, centrally membranous female subgenitalplate and less oblique ovipositor teeth, and the Tromatobia-subgroup with a less obviouslydipped occipital carina, more convex clypeus, fairly evenly sclerotized female subgenital plateand a very oblique proximal tooth on the ovipositor. The former subgroup is primarily Holarcticand Neotropical and contains 15 genera, including Scambus, Ephialtes and Dolichomitus. It isnot represented in Australia. The Tromatobia-subgroup is cosmopolitan and contains sevengenera, three of which, Acropimpla, Sericopimpla and Zaglyptus, occur in Australia. Two othergenera, Tromatobia and Clistopyga, occur in New Guinea. The other three generic groups differfrom the Ephialtes-group in having a broadly incomplete occipital carina. The monobasicPseudopimpla-group is Palaearctic whilst the Alophosternum-group, which contains twogenera, is primarily eastern Palaearctic and Nearctic. Only the Camptoty pus -group, charac-terized by the incomplete occipital carina, glabrous mesosternum and long straight ovipositor, isrepresented in Australia. This group is tropicopolitan. It comprises seven genera, three of whichare Old World/Australian and the remaining four are Neotropical. Of the Palaeotropicalgenera, Hemipimpla is restricted to Africa, Parvipimpla to Australia, whilst Camptotypus isIndo- Australian. Biologically there is some difference between the five groupings listed above, though little isknown of the Camptoty pus-group. The Alophosternum-group is associated with leaf-mininglarvae (Cushman, 1933), whilst Pseudopimpla is a parasite of cephid stem-miners (Bruzzese,1982). The Ephialtes-subgroup is parasitic on a variety of holometabolous insect larvae andpupae concealed in plant tissue (e.g. stem-borers, fruit-miners); oviposition is thus through theplant tissue. The Tromatobia-subgroup is parasitic in 'cocoons' of various kinds, includingpsychid larval cases, lepidopterous cocoons and spider egg-sacs. Not only is oviposition throughsilk but, unlike the previous subgroup, the adult ichneumonid is usually in contact with the silk.Many members of this group are gregarious. Gauld (1984) recognized five ephialtine genera as occurring in Australia; a species ofLiotryphon, purported to have been introduced from Europe was in fact not liberated inAustralia (Froggatt, 1909). The majority of Australian ephialtines occur in tropical Queenslandwhere they are encountered fairly frequently. Some species of Sericopimpla and Zaglyptus aremore widely distributed, particularly in the wet sclerophyll forests of the eastern part of thecontinent. ACROPIMPLA Townes Selenaspis Roman, 1910: 191. Type-species: Hemipimpla alboscutellaris Szepligeti, by original designa-tion. [Homonym of SelenaspisBletker, 1858.] ICHNEUMONIDAE OF AUSTRALIA 245 Acropimpla Townes, 1960ft: 159. Type-species: Charitopimpla leucostoma Cameron, by original designa-tion. Medium-sized species, fore wing length 7-8 mm; clypeus with a median apical notch; eye weakly indentedopposite antennal socket (Fig. 5); malar space 0-3 times as long as basal mandibular width; occipital carinacomplete. Epomia short; epicnemial carina strong; propodeum without carinae, spiracle circular. Femalewith claws basally lobate; fore wing with 3r-m present, enclosing a broad rhombic areolet; hind wing withfirst abscissa of Cu\ longer than cu-a. Lateromedian carina of tergite 1 not reaching to end of segment;tergites 2-4 with lateromedian swellings; ovipositor straight, projecting beyond apex of gaster by 1-5 timeslength of hind tibia. Wings infumate with pterostigma black. REMARKS. A moderately large genus most species of which occur in the Oriental region. A feware Holarctic and some Ethiopian. Gupta & Tikar (1978) recognized 26 Oriental species butnone from south-east of Mindanao. I have seen two apparent species from east of Weber's line, one from New Guinea and asecond, closely related one from Australia. Acropimpla species are known to parasitize a variety of Microlepidoptera, especially Pyrali-dae (Gupta & Tikar, 1978). Oviposition may be through silk (Townes, 1969). Acropimpla xantha sp. n. (Fig- 5) Face very weakly convex, sparsely punctate; malar space 0-3 times basal mandibular width; occipital carinavery weakly dipped mediodorsally. Mesoscutum with notauli fairly strong, reaching posteriorly to nearlevel of centre of tegulae; scutellum smooth, convex; mesopleuron highly polished, virtually impunctatewith subtegular ridge normally convex; metapleuron smooth and polished, submetapleural carina presentbut only slightly raised. Propodeum abruptly declivous, smooth and polished; lateromedian longitudinalcarinae absent, pleural carina present. Fore wing with cu-a opposite base of Rs&M; hind wing with firstabscissa of Cu\ longer than cu-a, distal abscissa of Cui present but weak. Gaster with tergite 1 withlateromedian carinae extending about 0-8 times length; lateral carina present, virtually complete except forpetiolar spiracle; tergites 2+ with weak tubercles, with coarse shallow punctures on tergite 2, the punctureson succeeding tergites progressively more obscure; subgenital plate centrally membranous; ovipositorprojecting beyond apex of gaster by 1-5 times length of hind tibia, its most proximal tooth oblique,subtending an angle of 30 to shaft. Female yellow, scape dorsally, pedicel and flagellum entirely, frons above antennal insertion, vertex andmost of genae above lower margin of eye, shining black; ovipositor sheath black. Wings strongly infumate,pterostigma dark brown. Male unknown. REMARKS. A distinctive species easily recognized from all other Indo- Australian species by itscolour. It belongs to group A of Gupta & Tikar (1978), a group of species characterized by thesmooth convex propodeum.A. xantha is only known from a single specimen collected in tropical Queensland. HOST RECORDS. None. MATERIAL EXAMINEDHolotype 9, Australia: Queensland, Atherton, ii.1975 (Howden) (TC). CAMPTOTYPUS Kriechbaumer Camptotypus Kriechbaumer, 1889: 311. Type-species: Camptotypus sellatus Kriechbaumer, by subsequent designation, Viereck, 1914: 27.Erythropimpla Ashmead, 1900a: 57. Type-species: Erythropimpla abbottii Ashmead (= Pimpla olynthia Cameron), by original designation.Trichiothecus Cameron, 1903: 136. Type-species: Trichiothecus ruficeps Cameron (= Ichneumon rugosus DeGeer), by monotypy. Medium-sized to moderately large species, fore wing length 9-14 mm; clypeus with a median apical notch;eye weakly indented opposite antennal socket; malar space 0-5-1-0 times as long as basal mandibularwidth; occipital carina obsolescent or absent dorsally, present laterally and ventrally. Epomia weak;epicnemial carina present, not strong; propodeum virtually without carinae, spiracle subcircular (Fig. 9). 246 I. D. GAULD Female with claws basally lobate; fore wing with 3r-m present, enclosing a broad, rhombic areolet; hindwing with first abscissa of Cu t longer than cu-a . Lateromedian carinae of tergite 1 reaching end of segment ;tergites 2, 3 and to a lesser extent 4 with diagonal furrows delimiting raised central areas; tergites 3-5sometimes with a small incision laterally so posterolateral corner is produced as a blunt tooth; ovipositorstraight, projecting beyond apex of gaster by 2-2-3-2 times length of hind tibia. Wings strongly infumatewith pterostigma yellow. REMARKS. Camptotypus is a moderately large genus containing 24 recognized species restrictedto the Indo-Australian region. It is very closely related to the Afrotropical genus HemipimplaSaussure. Although now treated as separate genera (Gupta & Tikar, 1978; Gauld, 1984), someauthors (e.g. Townes, 1969) consider Hemipimpla a subgenus of Camptotypus. Gauld (1984) listed five nominate species as occurring in Australia but remarked that twonames, sellatus Kriechbaumer and bicolor Kriechbaumer, had remained unassociated as thewhereabouts of the type-material was not known (Gupta & Tikar, 1978). Study of the speciesnow known to occur in Australia, and examination of Kriechbaumer 's original descriptions, hasenabled these two species to be identified with reasonable certainty and this has necessitatedsynonymizing two of the other names. Camptotypus species are conspicuous ichneumonids that can be observed flying in clearings inrain forest. Their colour pattern and mottled wings are reminiscent of some braconines and Ihave collected both flying in the same clearing. Few host records exist for the genus. One Asianspecies have been reared from Hyblaea sp. (Lepidoptera: Hyblaeidae) (Sonan, 1930), whilstone African species of Hemipimpla parasitizes an Epicampoptera (Lepidoptera: Drepanidae)(LePelley, 1954); a second species has been reared from Belonogaster nests (Hymenoptera:Vespidae) (Keeping & Crewe, 1983). Key to Australian species of Camptotypus 1 Hind tibia strongly flattened and broadened so it is less than 6-0 times as long as apically broad (Fig. 55); tergite 4 of gaster closely punctate, the punctures virtually contiguous (Fig. 63); $ with ovipositor 2-2-2-6 times as long as hind tibia bicolor Kriechbaumer (p. 246) Hind tibia subcylindrical, not obviously flattened, at least 7-0 times as long as apically broad(Fig. 54); tergite 4 or gaster more sparsely punctate, the punctures separated by at least 0-5times their own diameters (Fig. 64); $ with ovipositor 2-8-3-2 times as long as hind tibia 2 2 Gaster more or less entirely black, or with tergite 1 brownish; head dorsally whitish yellow; hind legs black sellatus Kriechbaumer (p. 248) Gaster with anterior and posterior tergites brownish, the 9 w ith tergites 3 and 4, theCf with tergites 4-6 blackish or infuscate; head dorsally black; hind legs brown, only tarsusblackish lachesis (Morley) (p. 247) Camptotypus bicolor Kriechbaumer(Figs 55, 63) Camptotypus bicolor Kriechbaumer, 1889: 311. Holotype $ , AUSTRALIA (depository unknown).Hemipimpla atropos Morley, 1914: 94. Lectotype 9, AUSTRALIA (BMNH), designated by Townes et al., 1961: 21 [examined]. Syn. n. Camptotypus (Camptotypus) atropos (Morley) Townes etal., 1961: 21.Camptotypus atropos atropos (Morley) Gupta & Tikar, 1978: 213. Fore wing length 12-14 mm ; upper tooth of mandible slightly broader than the lower but approximately thesame length. Mesoscutum with notauli strongly impressed anteriorly, reaching to level of centre of tegulae;submetapleural carina fairly strongly broadened and thickened anteriorly; propodeum abruptly declivous,posteriorly slightly convex. Hind tibia less than 6-0 times as long as maximally broad, flattened, with spursslightly shorter than apical breadth (Fig. 55). Tergite 1 of gaster in profile dorsally centrally angulate, thesurface of the petiole and postpetiole meeting at about 130, in dorsal aspect with carinae extending toposterior margin, the area between them smooth, with scattered punctures posteriorly; tergites 3 and 4centrally coarsely and very closely punctate (Fig. 63); posterolateral corners of tergites 3-5 slightlytruncated, not developed into teeth; ovipositor projecting beyond apex of gaster by 2-2-2-6 times length ofhind tibia.Female and male head, alitrunk, tergites 1, 2 and often 3 of gaster yellowish brown; flagellum brown, ICHNEUMONIDAE OF AUSTRALIA 247 centrally infuscate, distally pale; legs orange, hind tarsus sometimes infuscate; remainder of gasterblackish. Wings strongly infumate, clear at extreme proximal end and below pterostigma; pterostigmayellow. REMARKS. C. bicolor has previously remained as an unrecognized species as the location of thetype is unknown (Gupta & Tikar, 1978). However, Kriechbaumer's original description leaveslittle doubt as to its identity. The phrase 'Fulvis . . . abdomine basi excepta . . . nigris' clearlyidentifies it as this species and although not all specimens have the hind tarsus black some do. C.lachesis, which to some extent resembles this species, is unlikely to be conspecific with C. bicoloras in lachesis the head is black dorsally. Kriechbaumer does not state head colour exceptby the general statement 'fulvis'. Furthermore, the gaster of lachesis is fulvous anteriorly andposteriorly. Gupta & Tikar (1978) recognized three subspecies on minor colour differences, but for thepresent these are ignored as one may expect colour variation in any sexually reproducing speciesthat has a fairly wide and discontinuous geographical distribution. The rather shortened, flattened hind tibia and tarsus make C. bicolor one of the mostdistinctive species of Camptotypus . This species is uncommon in Australia; it also occurs in the Moluccas, New Guinea and NewBritain (Gupta & Tikar, 1978). HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (lectotype ofatropos), Queensland, Mackay (Turner) (BMNH). Australia: 1 $, Northern Territory, Daly River Mission, x.1974 (Hutchinson) (ANIC); 2 $, Queens-land, Redlynch, xi.1938 (Sternitzky) (BMNH); 3 $ (paralectotypes) , same data as lectotype (BMNH). Camptotypus lachesis (Morley)(Fig. 54) Hemipimpla lachesis Morley, 1914: 94. Lectotype 9> AUSTRALIA (BMNH), designated by Townes et al., 1961: 23 [examined]. Camptotypus (Camptotypus) lachesis (Morley) Townes etal. , 1961: 23.Camptotypus lachesis (Morley); Gupta & Tikar, 1978: 209. Fore wing length 9-13 mm; upper tooth of mandible slightly broader than the lower, of more or less thesame length. Mesoscutum with notauli strongly impressed anteriorly, reaching to level of centre of tegulae;submetapleural carina anteriorly parallel-sided, posteriorly evenly tapered; propodeum abruptly decli-vous, posteriorly biconcave. Hind tibia cylindrical, about 7-0 times as long as apically broad, with spursslightly longer than apical breadth (Fig. 54). Tergite 1 of gaster in profile dorsally from fairly abruptlyrounded to angulate, in dorsal aspect with carinae complete to posterior margin of segment, the areabetween them smooth, with scattered punctures posteriorly; tergites 3 and 4 centrally coarsely punctate,the punctures separated by about 0-5 or more times their own diameter; posterolateral corners of tergites3-5 slightly truncated, not developed into teeth; ovipositor projecting beyond apex of gaster by 3-0-3-2times length of hind tibia. Female head dorsally black, ventrally yellow; antenna black, distal apices yellowish brown; alitrunk,tergites 1-3 and 6+ of gaster and legs brownish orange; hind tarsi and tergites 3 and 4 of gaster black. Wingsinfumate distal to about cu-a except for an area below pterostigma; pterostigma yellow. Male similar tofemale but with tergites 4-6 infuscate, remainder of gaster fulvous. REMARKS. This species is quite easily recognized on account of its colour pattern, especially thedorsally black head and centrally black gaster. The holotype is labelled 'Victoria' but I doubtthat this is correct as all other specimens of Camptotypus are tropical. However, I havepreviously had no reason to doubt the veracity of French's locality data. C. lachesis is otherwise known only from Northern Territory and north Queensland. MATERIAL EXAMINED Australia: 1 $ (lectotype), 'Victoria' (French) (BMNH); 1 cf (paralectotype), Queensland, Mackay(Turner) (BMNH). Australia: 14 $ , 5 O" , Northern Territory, Queensland (AM; ANIC; BMNH; TC) (Map 4). 248 I. D. GAULD Camptotypus sellatus Kriechbaumer(Figs 9, 64) Camptotypus sellatus Kriechbaumer, 1889: 311. Holotype $, AUSTRALIA (depository unknown).Erythropimpla flaviceps Cameron, 1911a: 206. Lectotype 9 > AUSTRALIA (ZMA), designated by Townes et al., 1961: 22. [Junior secondary homonym of Camptotypus flaviceps (Smith).] Syn. n.Hemipimpla clotho Morley, 1914: 93. Lectotype $, AUSTRALIA (BMNH), designated by Townes et a/., 1961: 22 [examined]. Syn. n. Camptotypus (Camptotypus) clotho (Morley) Townes etal. , 1961: 22.Camptotypus (Camptotypus) flaviceps (Cameron) Townes et al. , 1961: 22.Camptotypus (Camptotypus) sellatus Kriechbaumer; Townes et al. , 1961 : 24.Camptotypus flaviceps (Cameron); Gupta & Tikar, 1978: 204. Fore wing length 13-14 mm; upper tooth of mandible slightly broader and longer than the lower.Mesoscutum with notauli weakly impressed anteriorly; submetapleural carina quite narrow, slightlybroadened anteriorly; propodeum abruptly declivous, posteriorly rather flat. Hind tibia subcylindrical,about 7-0 times as long as maximally broad, with spurs longer than apical breadth. Tergite 1 of gaster inprofile dorsally convex, without a central angulation, in dorsal aspect with carinae extending more or less tohind margin, the area between them smooth, at most with superficial punctures posteriorly; tergites 3 and 4centrally with large punctures separated by 0-5-1-0 times their own diameters (Fig. 64); posterolateralcorners of tergites 3-5 slightly incised, produced into blunt teeth; ovipositor projecting beyond apex ofgaster by 2-8-3-2 times length of hind tibia. Female and male head whitish yellow; flagellum blackish, often with scape reddish brown; alitrunk andanterior two pairs of legs orange brown; gaster and hind legs black. Wings strongly and relatively uniformlyinfumate; pterostigma yellow. VARIATION. In some specimens tergite 1 of the gaster is reddish brown; the mid tarsus may beinfuscate. A small hyaline area may be present below the pterostigma. One New Guineaspecimen has the propodeum strongly infuscate. REMARKS. C. sellatus has previously remained as an unrecognized species because the location ofthe type is unknown (Gupta & Tikar, 1978; Gauld, 1984). Only three species of Camptotypusseem to occur in Australia, and as the type-locality of sellatus is given as Australia (Kriech-baumer, 1889) presumably the name must refer to one of them. Kriechbaumer's originaldescription states 'pedibus posticis et abdomine nigris' and this is the only Australian species thathas the hind legs and gaster black. This species has previously been known under the name C.flaviceps (Cameron) (Gupta & Tikar, 1978), despite the fact thai flaviceps (Cameron) is a juniorsecondary homonym of flaviceps (Smith). The fact that flaviceps (Smith) is a rejected name as itis a junior primary homonym of Pimpla flaviceps Brulle does not affect its availability forpurposes of secondary homonomy so therefore Gupta & Tikar (1978) should have used clotho(Morley) as the name for this species. Structurally C. sellatus is very similar to C. basalis Cameron, a New Guinea species. Probablybasalis is only a colour variation of sellatus. In Australia C. sellatus is easily recognized on account of its colour pattern. In profile tergite 1of the gaster is dorsally more evenly rounded than either other Australian species. C. sellatus occurs in New Guinea and tropical Queensland. HOST RECORDS. In New Guinea this species was found 'attacking larva on leaf of Cassia grandis'(BMNH). MATERIAL EXAMINED Australia: 1 $ (lectotype of clotho), Queensland, Cairns, Hi. 1902 (BMNH); 2 $, same data (BMNH);6 $, 6 C?, Mackay (Turner) (BMNH) (all paralectotypes). Australia: 13 $, 3 d", Northern Territory, Queensland (AM; ANIC; BMNH; TC) (Map 3). Papua NewGuinea: 3 $ (BMNH). ICHNEUMONIDAE OF AUSTRALIA 249 PARVIPIMPLA GauldParvipimpla Gauld, 1984: 67. Type-species: Parvipimpla minuta Gauld, by original designation. Lower face subquadrate; clypeus flat, apically bilobed with a median apical notch; apical margin of clypeusthin ; malar space shorter than basal mandibular width ; mandibles of moderate length , weakly tapered withupper tooth slightly the longer. Occipital carina absent. Antennae rather short, those of male withouttyloids. Alitrunk highly polished, very sparsely punctate and with only scattered hairs; epomia present but short;mesoscutum in profile abruptly rounded, notauli strongly impressed, reaching back to level of centre oftegulae; epicnemial carina absent; propodeum without carinae; submetapleural carina absent. All tarsal claws of female with conspicuous basal lobe, those of male simple; distal tarsal segmentsslightly broadened. Fore wing with 3r-m absent; cu-a more or less opposite base ofRs&M. Hind wing with distal abscissa ofCui present; first abscissa of Cui about 0-6 times length of cu-a. Gaster smooth and highly polished; tergite 1 relatively long, without obvious lateromedian longitudinalcarinae; tergite 2 with oblique grooves cutting off anterior corners and a pair of weaker diagonalimpressions posteriorly cutting off hind corners so that the centre of the tergite is a convex rhombus;tergites 3 and 4 with grooves on posterior 6 ; hind margin of tergites 3-5 incised , the membranous incisionnarrow medially, broadest close to lateral margin but not extending laterally right to corner. Malesubgenital plate transverse, simple; female with ovipositor about as long as gaster, cylindrical, with a nodusand with about seven strong oblique teeth on lower valve. REMARKS. Parvipimpla belongs to the Camptotypus genus-group (Townes, 1969), a grouprepresented in the Indo-Australian region by only this genus and Camptotypus. The absence of3r-m, lack of epicnemial carina and long cu-a in the hind wing serve to distinguish Parvipimplafrom Camptotypus.A single endemic Australian species is known. Nothing is known of the biology of this insect. Parvipimpla petita GauldParvipimpla petita Gauld, 1984: 68. Holotype $, AUSTRALIA (DPIQ) [examined]. Fore wing length 6-7 mm. Head highly polished with few scattered punctures; ocelli forming a nearlyequilateral triangle; flagellum of female with 20 segments, male with 19. Alitrunk highly polished, almostwithout punctures and with long pale sparse hairs. Legs with conspicuous long hairs. Gaster smooth andpolished with a few minute hair-bearing punctures posteriorly and laterally on the tergites. Female head , alitrunk and anterior two pairs of legs reddish brown ; pedicel , flagellum , tergites of gaster ,hind tibia and tarsus and ovipositor sheath black; remainder of hind leg blackish red; membranous incisionof tergites 3+ white. Wings infumate, pterostigma blackish. Male similar except face white, all coxaeyellowish brown, hind femur and tibia yellowish with indistinct infumation. REMARKS. The striking colour pattern is like that of many other similar-sized AustralianParasitica (e.g. Eriostethus, one species of Philogalleria, some Braconinae etc.). It is character-istic of Australia. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), New South Wales, Urbenville, Tooloom scrub, 22-23. iii. 1975 (Cantrell)(DPIQ); 1 cf (paratype), Queensland, Mt Nebo, ii. (TC). SERICOPIMPLA Kriechbaumer Sericopimpla Kriechbaumer, 1895: 135. Type-species: Pimpla sericata Kriechbaumer, by monotypy.Charitopimpla Cameron, 1902: 48. Type-species: Charitopimplaflavobalteata Cameron, by monotypy.Philopsyche Cameron, 1905: 137. Type-species: Philopsyche albobalteata Cameron (= Pimpla sagraeSnellen van Vollenhoven), by monotypy. Moderately large species, fore wing length 11-15 mm; clypeus with median apical notch; eye large, stronglyindented opposite antennal socket (Fig. 4); malar space very short; occipital carina complete. Epomiastrong; epicnemial carina complete; propodeum without distinct carinae except laterally (Fig. 10);propodeal spiracle circular. Female with claws basally lobate; fore wing with 3r-m present, enclosing 250 I. D. GAULD triangular areolet; first abscissa of Cu\ in hind wing, longer than cu-a (Fig. 8). Lateromedian carinae oftergite 1 not reaching to posterior margin; tergites 2-4 of gaster with lateromedian swellings; ovipositorslightly decurved, projecting beyond apex of gaster by 1 -8-2-2 times length of hind tibia. REMARKS. Sericopimpla is a moderately large Palaeotropical genus containing 13 Indo-Austra-lian species and one widespread Afrotropical species. One of the Oriental species extends as farnorth as Korea and Japan and in Australia species occur in the temperate south. The species arevery similar to each other, differing primarily in colour and superficial sculpture, and a numbermay prove to be extreme variants of other, more widespread species. Sericopimpla belongs to the Acropimpla-group of genera (Townes, 1969), species of whichusually oviposit through silk. As far as is known the hosts of Sericopimpla are psychids. Only thebiology of one species, the Afrotropical 5. sericata Kriechbaumer (= abdominalis Morley), hasbeen studied in detail (Skaife, 1921; Smithers, 1956). This species was observed to repeatedlyinsert its ovipositor into the psychid 'bag' in an attempt to sting the larva. If stung, the larvabecame paralysed and remained alive but immotile. An egg would then be laid either on the hostlarva or else enmeshed in the innermost strands of the 'bag'. After eclosion the parasitoid larvacommenced feeding, usually on the abdomen of the host. It underwent four or five rapid larvalinstars and then a more protracted final instar. The mature larva is furnished with clusters ofhooks on the dorsum of the abdominal segments which enable it to move around within the bag.Although the Afrotropical species is apparently solitary (and Smithers (1956) observed destruc-tion of supernumerary Sericopimpla larvae by the most developed parasitoid larva), at least oneOriental species, 5. albicincta (Morley), is gregarious as there is a psychid case (lEumetd) in theBMNH from which 10 male and five female specimens emerged. Three species occur in Australia. One, 5. lutea, is particularly distinctive, both in colour andmorphology, whilst the other two are structurally similar and seem to be very closely related toseveral Oriental species. These two species are widely distributed in Australia and also occur inNew Guinea; the third, 5. lutea, is only known from tropical Queensland. Key to Australian species of Sericopimpla 1 Alitrunk and gaster predominantly yellow; mesopleuron swollen below subalar prominence; scutellum strongly convex in profile (Fig. 61) lutea sp. n. (p. 252) Alitrunk and gaster black with tegulae and posterior margins of gastral tergites yellow-marked;mesopleuron not swollen below subalar prominences ; scutellum almost flat (Fig. 62) 2 2 Hind tibia with black band near proximal 0-2; mesoscutum with sparse punctation, the punctures widely interspaced (Fig. 57) australis Townes, Townes & Gupta (p. 250) Hind tibia with proximal 0-6 white, without a proximal black band; mesoscutum denselypunctate, the punctures interspaced by only about their own diameters (Fig. 56) crenator(F.)(p. 251) Sericopimpla australis Townes, Townes & Gupta(Fig. 57) Philopsyche annulipes Cameron, 1912: 186. Holotype $, AUSTRALIA (BMNH) [examined]. [Secondary homonym of Charltopimpla annulipes Cameron, 1905.] Sericopimpla australis Townes et al. , 1961: 17. [Replacement name for annulipes Cameron, 1912.]Sericopimpla annulipes (Cameron) Gupta & Tikar, 1978: 176. Fore wing length 1 1-14 mm. Mandible evenly tapered with lower tooth very slightly shorter than the upper;lower face elongate; about 0-8 times as broad as long; face sparsely punctate; genae moderately stronglynarrowed behind eyes. Mesoscutum centrally sparsely and shallowly punctate, the punctures separated bymore than their own diameters (Fig. 57); scutellum coarsely punctate posteriorly. Mesopleuron quite finelypunctate, not swollen below subalar prominences. Propodeum centrally smooth, laterally and posteriorlyclosely punctate, with lateromedian carinae vestigial. Tergite 1 of gaster with scattered punctures, tergites2-4 fairly closely punctate. Ovipositor projecting beyond apex of gaster by 2-0-2-2 times length of hindtibia. Black; scape ventrally in part, palpus, hind corner of pronotum, tegula, anterior two pairs of legs, hindtrochanter, trochantellus, much of tarsus and hind margin of gastral tergites, yellow; hind tibia whitish ICHNEUMONIDAE OF AUSTRALIA 251 yellow, distal 0-3 and band near proximal 0-2 black; distal hind tarsal segments blackish. Wings hyaline,pterostigma black. REMARKS. This species is structurally very similar to 5. crenator and, although the differencesgiven in the key work well for material to hand, I have some doubt as to whether they are reallydistinct. They appear to be synchronous and sympatric and attack, at least some of the time, thesame host species. 5. aus trails is somewhat less common than 5. crenator but it is similarly widely distributed,though I have seen no material from Northern Territory, South Australia or Tasmania.Although Gupta & Tikar (1978) mentioned that this species occurs in New Guinea they did notvalidate their record by reference to any material or collection. There is in the BMNH a recentlycollected female of this species from Popondetta, Papua New Guinea, confirming that S.aus trails occurs on that island. HOST RECORDS. Clanla Ignobllls (Walker) (Chadwick & Nikitin, 1976); Hyalarcta huebnerl(Westwood) (Chadwick & Nikitin, 1976), both Lepidoptera: Psychidae. In both cases thematerial on which these records were based has been examined in DAR and the identity of theichneumonids confirmed. MATERIAL EXAMINED Australia: 1 $ (holotype of annulipes) , New South Wales, Mittagong (Froggatt) (BMNH). Australia: 19 $ , 14 cf , Australian Capital Territory, New South Wales, Queensland, Victoria, WesternAustralia (AM; ANIC; BMNH; DAR) (Map 1). Papua New Guinea: 1 $ (BMNH). Sericopimpla crenator (F.)(Figs 1,56, 62) Pimpla crenator F., 1804: 114. Holotype $, AUSTRALIA (Kiel). Ichneumon crenator (F.) Thunberg, 1822: 277. Philopsyche pilosella Cameron, 19116: 337. Holotype d", AUSTRALIA (BMNH) [examined]. [Synonymized byTownese?a/.,1961: 18.]Exeristes consimilis Morley, 1914: 27. Holotype d", AUSTRALIA (BMNH) [examined]. [Synonymized by Betrem, 1932: 23.] Charitopimpla pilosella (Cameron) Betrem, 1932: 23.Sericopimpla crenator (F.)Townesef a/., 1961: 17. Fore wing length 12-15 mm. Mandible evenly tapered, with lower tooth shorter than the upper; lower faceelongate, about 0-8 times as broad as long; face sparsely but distinctly punctate; genae strongly narrowedbehind eyes. Mesoscutum centrally closely punctate, the punctures separated by only about their owndiameters or less, laterally more sparsely punctate, but still distinctly so (Fig. 56); scutellum coarselypunctate. Mesopleuron anteroventrally finely, evenly punctate, not swollen below subalar prominence.Propodeum centrally smooth, laterally and posteriorly closely punctate, with lateromedian carinaevestigial. Tergite 1 of gaster with quite coarse punctures, tergites 2-4 quite coarsely and closely punctate.Ovipositor projecting beyond apex of gaster by 2-0-2-2 times length of hind tibia. Black; scape ventrally, palpus, hind corner of pronotum, tegula, anterior two pairs of legs, hindtrochanter, trochantellus and tarsus and posterior parts of gastral tergites, whitish yellow; hind tibiawhitish with distal 0-25 black; distal hind tarsal segments often infuscate. Wings hyaline, pterostigmablack. REMARKS. This species is very similar in general appearance to S. australis but is distinguishableby the colour of the hind tibia and coarse sculpture of the mesoscutum. 5. crenator is probablythe commonest Australian Sericopimpla. I have seen material from all states except SouthAustralia. It is also widely distributed throughout New Guinea and the New Hebrides (Gupta &Tikar, 1978). HOST RECORDS. Clania ignobilis (Walker) (ANIC); C. tenuis Rosenstock (ANIC); Hyalarctahuebneri (Westwood) (Heather, 1976); H. nigrescens (Doubleday) (Chadwick & Nikitin, 1976);Oiketicus elongatus Saunders (Chadwick & Nikitin, 1976) (all Lepidoptera: Psychidae). 252 I. D. GAULD MATERIAL EXAMINED Australia: 1 9 (holotype of pilosella), New South Wales, Mittagong (BMNH); 1 cf (holotype ofconsimilis) , Queensland, Mackay (BMNH). Australia: 85 $, 49 cf, from all states except South Australia (AM; ANIC; BMNH; NMV; QM; TC)(Map 2). Sericopimpla lutea sp. n. (Figs 4, 10,58,61) Fore wing length 11-13 mm. Mandible strongly narrowed with lower tooth shorter than the upper; malarspace virtually obliterated; lower face subquadrate, about 0-9 times as broad as long, 1-2 times medianheight from clypeal suture to level of antennal bases; face virtually impunctate. Mesoscutum centrallysparsely punctate, lateral lobes with few isolated punctures, highly polished (Fig. 58); scutellum impunc-tate, polished, in profile very convex (Fig. 61). Mesopleuron with coarse punctures, pyramidally swollenbelow subalar prominences. Propodeum smooth, polished, with well-developed lateromedian longitudinalcarinae. Tergite 1 of gaster with few punctures, tergites 2+ coarsely and quite closely punctate. Ovipositorprojecting beyond apex of gaster by 1-8 times length of hind tibia. Bright yellow; flagellum, scape internally, vertex of head, mesoscutal marks and ovipositor sheath black;hind tarsi infuscate. Wings weakly infumate with apices of fore wings strongly infumate; pterostigma black. REMARKS. A distinctive species on account of its colour pattern. The swollen mesopleuron,convex scutellum, broad face and weakly punctate mesoscutum are other characteristic featuresof this species. It is only known to occur in northern Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, 31 km NW. by N. of Cooktown (1518'S; 14501'E), 250 m,20.V.1977 (Common & Edwards) (ANIC). Paratypes. 1 $, Australia: Queensland, Mackay, 1909 (BMNH); 1 9, Cape York, iv.1973 (Monteith)(ANIC); 1 9, Mackay, 8.vi.l931 (Burns) (NMV); 1 $, Mackay (BMNH). ZAGL YPTUS Foerster Zaglyptus Foerster, 1869: 166. Type-species: Polysphincta varipes Gravenhorst, by subsequent designa-tion, Woldestedt, 1877: 17. Medium-sized species, fore wing length 4-8 mm; clypeal margin concave; occipital carina complete.Epomia weak; epicnemial carina complete; propodeum without carinae but with lateral subapicaltubercles (Fig. 11); propodeal spiracle subcircular. Female with claws basally lobate, male with fore femurventrally with a small tubercle near distal end, the area distal to this concave; fore wing with 3r-m absent;hind wing with distal abscissa of Cu\ present or absent, if present then with first abscissa of Cui about equalin length to cu-a or very slightly shorter, or if absent with Cu\ and cu-a confluent, smoothly arcuate.Lateromedian carinae of tergite 1 not reaching to posterior margin; tergite 2 of gaster with oblique groovesdelineating a central, raised, rhombic area; tergites 3-4 with transverse tubercles; ovipositor straight,projecting beyond apex of gaster by about 1-2 times length of hind tibia, with most proximal tooth of lowervalve with an elongate free tip. REMARKS. Zaglyptus is a moderately small cosmopolitan genus with nine described species in theIndo-Australian region. Two species occur in Australia, Z. glabrinotum and Z. hollowayi. Theyare closely related and belong to the nigrolineatus-group as defined by Gupta (1961). The twoAustralian species seem to be very closely related to the New Guinea species Z. grandis Gupta.Zaglyptus species are believed to attack spiders in retreats; the spider is stung to death and theichneumonid larvae are believed to consume both the spider's eggs and body (Nielsen, 1935;Townes, 1969). European species have been reared from Clubionidae and Araneidae (Aubert,1969). Key to Australian species of Zaglyptus 1 Hind wing with distal abscissa of Cu\ present; Cu\ and cu-a angled at junction (Fig. 59); punctures on tergites 3-4 of gaster coarse, close and deep glabrinotum (Girault) (p. 253) ICHNEUMONIDAE OF AUSTRALIA 253 Hind wing with distal abscissa of Cui absent; Cui and cu-a together forming a slightly arcu-ate reclivous vein (Fig. 60); punctures of tergites 3-4 moderately coarse, scattered andshallow hollowayi sp. n. (p. 253) Zaglyptus glabrinotum (Girault)(Figs 11, 59) Polysphincta glabrinotum Girault, 1925: 541. Holotype $, AUSTRALIA (QM) [examined].Zaglyptus? glabrinotum (Girault) Townes et al., 1961: 20.Zaglyptus glabrinotum (Girault) Townes, 1971a: 470. Fore wing length 5-8 mm. Mandible with upper tooth stout, tapered to a sharp point, about 2-0 timeslength of the lower; clypeus convex in profile; ocelli moderately small, the posterior ones separated fromeye by about their own diameter. Alitrunk highly polished, virtually impunctate; notauli stronglyimpressed, submetapleural carina reduced to an anterior tooth. Fore wing with cu-a opposite base ofRs&M; hind wing with distal abscissa of Cu v present (Fig. 59). Caster with tergites 3-4 with coarse, closedeep punctures centrally and on lateral tubercles; ovipositor projecting beyond apex of gaster by 1-2 timeslength of hind tibia. Yellow; dorsal surface slightly brownish yellow; flagellum dorsally infuscate, distal apex pale; extremeposterior end of propodeum with a pair of black areas near insertion of gaster; distal apex of hind tibia andhind tarsi 1-3 infuscate; ovipositor sheath black. Wings hyaline; pterostigma black, proximal and distalcorner pale. REMARKS. A distinctive species easily recognized by the hind wing venation. It appears to be themore common of the two Australian species. It is known from Queensland, Northern Territoryand northern New South Wales. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype of glabrinotum), Queensland, Yeronga, vi.1924 (QM). Australia: 1 <j>, New South Wales, Blue Mt, 150 m, i. (TC); 1 cf, Grosevale nr Richmond, iii.1972(McAlpine & Holloway) (AM); 1 $, Northern Territory, 14kmNW. Cape Crawford (1634'S; 13541'E),xi.1975 (Cardale) (ANIC); 1 $, 1 cT, Queensland, Biggenden, Bluff Rg., viii. 1971 (Fraucd) (ANIC); 1 $,Brisbane, i-vi. (TC); 3 , Bundaberg, 1., vi., viii. 1971-1973 (Fraucd) (ANIC); 1 $, Fernvale, xi-xii. (TC);1 $, 1 cf, Mt Glorious, v. (TC); 1 $, Mt Nebo, iii. (TC); 3 $, 1 cf, Mt Tambourine, ix-x. 1976-78(Galloway} (BMNH); 1 $, Stanthorpe ii-iii. (TC). Zaglyptus hollowayisp. n. (Fig. 60) Fore wing 4-6 mm ; mandible with upper tooth stout , tapered to a sharp point , about 1 8 times as long as thelower; clypeus moderately strongly convex in profile; ocelli small, the posterior ones separated from eye bymore than their own diameter. Alitrunk very highly polished, impunctate; notauli strongly impressed;submetapleural carina reduced to an anterior tooth. Fore wing with cu-a more or less opposite Rs&M; hindwing with distal abscissa of Cu\ absent, Cui&cu-a forming a smooth, very slightly arcuate line (Fig. 60).Gaster with tergites 3-4 with moderately coarse, scattered, shallow punctures; ovipositor projectingbeyond apex of gaster by 1-1-1-2 times length of hind tibia. Similar in colour to Z. glabrinotum but with interocellar area of holotype yellowish and hind tarsi onlyslightly infuscate at distal apices of segments 1-3. The female paratype has the interocellar area black. REMARKS. This species is named in honour of Mr Geoff Holloway as a gesture of thanks fororganizing the trip on which the holotype was collected. Z. hollowayi is very similar to Z. glabrinotum except for the characters mentioned in the keyand in having less extensively infuscate hind legs. It is known from forests in Queensland andNew South Wales. HOST RECORDS. None. MATERIAL EXAMINEDHolotype $, Australia: New South Wales, Mt Royal Nat. Pk., ii.1983 (Gauld) (AM). 254 I. D. GAULD Paratypes. Australia: 1 9, Queensland, Mt Tambourine, xi.1977 (Galloway) (BMNH); 1 cf , Shipton'sFlat, x.1980 (Cardale) (ANIC). Tribe POLYSPHINCTINI The Polysphinctini is a holophyletic group of genera characterized by the possession of enlargedpulvilli, inflated tarsal segments and an evenly tapered, very sharply pointed ovipositor. Mostspecies lack vein 3r-m in the fore wing. Larval polysphinctines are external parasites of spidersand they are furnished with 'holdfast organs' used to maintain the larva on the host (Nielsen,1923; Short, 1978). These holdfast organs are single or paired dorsal tubercles covered in curvedspines. The cocoon of many polysphinctines is unusual in having a small hole in the ventral endvia which the meconium is voided. The supposedly primitive polysphinctines (e.g. Schizopygd)attack spiders in retreats whilst many genera attack free-living araneids. A prospective host isstung and immobilized before an egg is attached to the body. Subsequently, the spider recoversfrom envenomation and continues its normal life but with a polysphinctine egg and eventuallylarva developing externally. As the larva grows it is visible as a whitish 'muffler' usually near theanterior end of the opisthosoma. The spider host is not killed until the parasite reaches its finallarval instar (Townes, 1969). The Polysphinctini is a specialized offshoot of the Ephialtini and many steps in the biologicaland morphological evolution of the group are extant (Townes, 1969). This means that it isdifficult to decide where to delineate the tribe. Larval holdfasts are found in several genera ofEphialtini such as Sericopimpla, though this is traditionally regarded as an ephialtine genus(Baltazar, 1961; Townes, 1969; Gupta & Tikar, 1978). More controversial is the placement ofthe genera Tromatobia, Zaglyptus and Clistopyga. Some authors (Finlayson, 1967; Gupta &Tikar, 1978) place these genera in the Polysphinctini; others (Townes, 1969; Aubert, 1969) treatthem as ephialtines. Like the Polysphinctini these genera often lack an areolet and areassociated with spiders, though apparently as carnivores devouring eggs in egg sacs. Unlike thepolysphinctines none has a specialized tarsus and pulvilli. Tromatobia and Zaglyptus havetypical ephialtine ovipositors with a parallel-sided shaft abruptly tapering to a point distally . Theovipositor of Clistopyga is more like that of a polysphinctine but is finely sculptured, not smooth.Although the delineation of the Polysphinctini is somewhat arbitrary it is suggested that the termbe used for only the genera known (or believed) to attack spiders (rather than their eggs), thathave modified pulvilli and distal tarsal segments and that possess tapered, polished ovipositors.This definition excludes Tromatobia, Zaglyptus and Clistopyga. World-wide the Polysphinctini contains 13 genera, four of which, Acrodactyla, Dreisbachia,Eriostethus and Zatypota, occur in Australia. Eriostethus is a moderately large genus confined tothe Indo-Australian region. The other taxa are probably cosmopolitan but they are ratherinfrequently collected and consequently little is known about their distribution other thanisolated records from widely separated areas. In Australia polysphinctines are predominantly tropical and moist forest insects. Except forsome species of Eriostethus, which can be very common in Queensland, most are ratheruncommon and seldom collected though Malaise traps at suitable sites generally take severalspecies in a year. Nielsen (1923) remarks on a high incidence of parasitism being observedamongst spiders but I believe this was an exceptional observation. Collecting evidence, basedmostly on Palaearctic experience, suggests some species may be locally common for a shortperiod, then not be seen in numbers for several years. The large number of Dreisbachia taken onBlack Mountain, Canberra in November-December, 1978 and the subsequent virtual dis-appearance of the species (two specimens collected since then despite considerable collectingactivity) suggests such population fluctuations may occur in Australia. ACRODACTYLA Haliday Acrodactyla Haliday, 1839: 117 [as a subgenus ofPimpla F.]. Type-species: Pimpla (Acrodactyla) degener Haliday, by subsequent designation, Westwood, 1840: 57.Acrodactyla Haliday; Westwood, 1840: 57 [raised to genus]. ICHNEUMONIDAE OF AUSTRALIA 255 Colpomeria Homgren, 1859: 126. Type-species: Colpomeria laevigata Holmgren (= Ichneumon quadris- culptus Gravenhorst), by monotypy.Symphylus Foerster, 1871: 105. Type-species: Symphylus hadrodactylus Foerster (= Pimpla (Acrodactyla) degener Haliday), by original designation. [Homonym of Symphylus Dallas, 1851.]Polemophthorus Schulz, 1911: 22. [Replacement name for Symphylus Foerster.] Small to medium-sized insects, fore wing length 3-6 mm; palp formula 5, 4 or 5, 3; clypeus convex, apicallyslightly flattened; eye surface bare; occipital carina complete. Epomia strong; mesoscutum polished,glabrous to sparsely pubescent, with a crest near anterior end of notaulus; epicnemial carina present;propodeum in profile long, dorsally with some traces of carinae, often coarsely sculptured. Tarsal claws offemale basally lobate; hind tibia of female internally uniformly hirsute; fore wing with 3r-m absent; hindwing with distal abscissa of Cui present or absent, first abscissa of Cu\ longer than cu-a. Tergites 2-4polished, sparsely punctate with weakly defined central rhombic areas; ovipositor straight or slightlyup-curved, projecting beyond apex of gaster by 0-3-0-6 times length of hind tibia. REMARKS. A moderate-sized genus with most species in the Holarctic and Oriental regions.Townes (1969) recognized three species-groups - the madida-group, the degener-group and thequadrisculpta-group. Several authors (e.g. Aubert, 1969; Carlson, 1979) treat the latter group asa separate genus, Colpomeria, but intermediates do occur (Townes, 1969). Furthermore, thequadrisculpta-group seems to grade into Pterinopus Townes, a Madagascan genus and the statusof this as a distinct group needs critical re-evaluation in the light of recent collecting in India andSouth East Asia. Four species of Acrodactyla occur in Australia. Two, A. quadrisculpta and A. micans, belongto the quadrisculpta-group; A. cursor apparently belongs to the degener-group whilst A. zekhemappears to be intermediate. No hosts are recognized for Australian Acrodactyla species but in Europe species have beenreared from Araneidae, Linyphiidae and Tetragnathidae (Aubert, 1969). The latter family alsoserves as host for a Nearctic species (Howell & Pienkowski, 1972). Key to Australian species of Acrodactyla 1 Ocelli large, the posterior one separated from the eye by about 0-3 times its minimum diameter (Fig. 81); face narrow, its minimum width 0-5-0-6 times distance from apex of clypeus toinsertion of antenna zekhem sp. n.(p. 257) - Ocelli not exceptionally large, the posterior one separated from the eye by at least its minimum diameter (Fig. 82); face moderately wide, its minimum width 0-7-0-8 times distance from apexof clypeus to insertion of antenna 2 Metapleuron fairly smooth, at most with traces of wrinkling; fore and mid femora slender, not inflated and specialized ventrally; hind wing with distal abscissa of Cui absent, not evenrepresented by an angulation at junction of Cu\ and cu-a (Fig. 79) cursor sp. n. (p. 255) - Metapleuron coarsely reticulo-rugose (Fig. 85); fore and mid femora inflated, distally con- stricted ventrally and with a blunt median ventral tooth (Fig. 80); hind wing with distalabscissa of Ci present, at least discernible as a short stub at junction of Cu\ and cu-a (Fig. 78) 3 3 Alitrunk entirely black; tergite 1 of gaster rather stout, 1-4-1-5 times as long as posteriorly broad; hind leg with third tarsal segment 1 -7-1 -8 times as long as broad quadrisculpta (Gravenhorst) (p. 256) Alitrunk predominantly orange , only propodeum , metapleuron and mesosternum black ; tergite1 of gaster slender, 1-7-1-8 times as long as posteriorly broad; hind leg with third tarsalsegment about 2-0 times as long as broad micans sp. n. (p. 256) Acrodactyla cursor sp. n. (Fig. 79) Fore wing length 3 mm; face moderately wide, about 0-8 times as broad as long; clypeus apically truncate;malar space about 0-8 times basal mandibular width; ocelli small, the lateral one separated from eye byabout 2-0 times its minimum diameter. Mesopleuron smooth and shining; metapleuron more or lesssmooth, with isolated punctures and traces of wrinkling posteriorly; propodeum fairly smooth withlateromedian longitudinal carinae obsolescent. Fore and mid legs with femora cylindrical, slender,unspecialized; hind leg with trochantellus bearing a small ridged protuberance ventrally; third hind tarsal 256 I. D. GAULD segment 2-0-2-3 times as long as broad. Hind wing with distal abscissa of Cu\ absent, not even discernibleas an angulation at junction of Cu\ and cu-a (Fig. 79). Gaster with sternite 1 smooth; tergite 1 about 1-8times as long as posteriorly broad, with weak longitudinal carina reaching to near centre, tergites 2-4 withweak convex areas; ovipositor straight, projecting beyond the apex of the gaster by 0-4-0-5 times length ofhind tibia. Female head, alitrunk and tergite 1 of gaster black, tergites 2+ blackish brown; fore and mid legsbrownish orange, coxae, distal ends of trochanters and mouthparts whitish; hind legs brownish, trochanterand trochantellus slightly paler. Male unknown. REMARKS. This small species is easily recognized on account of its dark colour and slenderfemora. It belongs to the degener-group of Townes (1969) though it differs from any describedspecies in having only sparse hair on tergite 4. It is only known to occur in Tasmania. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Tasmania, Roseberry, iv. (TC). Paratypes. Australia: 1 $, Tasmania, Hellyer Gorge, i-ii. (TC); 1 $, Waldheim, 800m, i-ii.(BMNH). Acrodactyla micans sp. n. (Figs 80, 82, 85) Fore wing length 3-6 mm; face moderately wide, 0-7-0-8 times as broad as long; clypeus apically truncate;malar space about 1-0 times basal mandibular width; ocelli moderately small, the lateral one separatedfrom eye by about 1-0 times its minimum diameter (Fig. 82). Mesopleuron smooth and polished,metapleuron coarsely reticulo-rugose (Fig. 85); propodeum rugulose, with lateromedian carinae long andstrong, usually with a small area superomedia separated from an elongate area petiolaris. Fore and mid legswith femora inflated, abruptly tapered distally and ventrally, bearing a median tooth-like process (Fig. 80);hind leg with trochantellus ventrally unspecialized; third hind tarsal segment about 2-0 times as long asbroad. Hind wing with distal abscissa of Cu\ present, rather faintly pigmented but reaching almost tomargin of wing. Gaster with sternite 1 rugose; tergite 1 1-7-1-8 times as long as posteriorly broad, withquite well-defined lateromedian longitudinal carinae extending for 0-8 of its length; tergites 2-4 withobvious convex areas and conspicuous shallow transverse furrows; ovipositor very slightly up-curved,extending beyond apex of gaster by 0-5-0-6 times length of hind tibia. Female and male head, mesosternum, propodeum, metapleuron and first segment of gaster black,remainder of alitrunk orange, tergites 2+ of gaster brownish; mouthparts, coxae and trochanters white,remainder of fore and most of mid legs orange or brownish, mid tarsus infuscate; hind femur, tibia andtarsus dark brown or blackish. VARIATION. Some females have an indication of a median pale band on the hind tibia. REMARKS. A. micans is very similar to A. quadrisculpta except that it is more slender, thepropodeal sculpture is slightly coarser, the terminal four segments of the gaster are more evenlyhirsute and the tarsi are more elongate. The most striking difference is in colour pattern.This species seems to be restricted to the south-east of Australia. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: New South Wales, Kiandra, Alpine Creek, iii.1962 (Riek) (ANIC). Paratypes. Australia: 1 <j>, Australian Capital Territory, Blundell's, iv.1930 (Graham) (ANIC); 1 $,Blunders, iv.l930(7omw>) (ANIC); 2 $,Blundell's,xi. 1930 (Gra/mm)( ANIC); 1 $, New South Wales,Kiandra, Alpine Creek, iii.1962 (Riek) (BMNH); 1 $, Kiandra, ii.1963 (Riek) (ANIC); 1 cf, Tasmania,Mole Ck., ii.1923 (Edwards) (BMNH). Acrodactyla quadrisculpta (Gravenhorst)(Fig. 78) Ichneumon quadrisculptus Gravenhorst, 1820: 378. Holotype cf , ITALY (lost).Colpomeria quadrisculpta (Gravenhorst) Schmiedeknecht, 1907: 1178. ICHNEUMONIDAE OF AUSTRALIA 257 Polysphincta quadrisculpta (Gravenhorst) Meyer, 1934: 96.Acrodactyla quadrisculpta (Gravenhorst) Townes etal., 1965: 35. Fore wing length 3-4 mm; face moderately wide, 0-80-0-85 times as broad as long; clypeus apicallytruncate; malar space about 1-0 times basal mandibular width; ocelli small, the lateral one separated fromeye by 1-2-1-4 times its own minimum diameter. Mesopleuron finely punctate, polished, metapleuroncoarsely reticulo-rugose; propodeum rugulose, with lateromedian longitudinal carinae strong, usually witha small rectangular area superomedia separated from an elongate area petiolaris. Fore and mid legs withfemora inflated, abruptly tapered distally and ventrally, bearing a median tooth-like process; hind leg withtrochantellus ventrally unspecialized, third hind tarsal segment 1-7-1-8 times as long as broad. Hind wingwith distal abscissa of Cu\ present but faint (Fig. 78). Gaster with sternite 1 rugulose; tergite 1 1-4-1-5 timesas long as posteriorly broad, with strong lateromedian carina extending about 0-8 of its length; tergites 2-4with moderately weak convexities; ovipositor very slightly up-curved, extending beyond apex of gaster by0-3-0-5 times length of hind tibia. Female and male black; legs pale yellowish, hind tibia centrally white, proximally and distally infuscate;mouthparts yellowish. REMARKS. This species was described from Europe and is known to occur quite widelythroughout the western Palaearctic region (Aubert, 1969). Presumably it was inadvertentlyintroduced into Australia. The Australian specimens differ from 'typical' European specimensin having the scape dark, not ventrally pale marked. HOST RECORDS. In Europe this species attacks tetragnathid spiders (Nielsen, 1937; Capener,1938). MATERIAL EXAMINED Australia: 1 $, 4 cT, Queensland, Kuranda, vi-vii.1913 (Turner) (BMNH). Europe: 5 $, 5 cf , variouslocalities (BMNH). Acrodactyla zekhem sp. n. (Fig. 81) Fore wing length 3 mm; face narrow, 0-5-0-6 times as broad as long; clypeus narrowly truncate, almostevenly convex; malar space short, less than 0-4 times basal mandibular width; ocelli large, the lateral oneseparated from eye by about 0-3 times its own minimum diameter (Fig. 81). Mesopleuron smooth andpolished; metapleuron coriaceous, with few rugae posteriorly; propodeum almost smooth with laterallongitudinal carinae strong, defining an elongate area, the area superomedia not clearly delineated. Foreand mid legs with femur very slightly inflated, otherwise unspecialized; hind leg with tibia slightly flattenedventrally; third hind tarsal segment about 3-0 times as long as broad. Hind wing with distal abscissa of Cu\exceptionally weak, discernible as a slight angling and stub at junction of Cu\ and cu-a. Gaster with sternite1 rugose; tergite 1 1-8-1-9 times as long as broad, with strong lateromedia longitudinal carinae reaching 0-8of its length; tergites 2-4 with weak convexities; ovipositor straight, its apex up-curved, ovipositorprojecting beyond apex of gaster by 0-4-0-5 times length of hind tibia. Female yellowish brown, head black, gaster infuscate; flagellum brown, scape yellowish, mouthpartswhitish; legs concolorous with alitrunk and coxae slightly paler. Male similar to female but with base ofantenna and coxae white. REMARKS. This small species from tropical Queensland is immediately recognizable by itsenlarged ocelli and short malar space. The enlarged ocelli and eyes suggest a nocturnal habit.The holotype was collected at light. HOST RECORDS. None. MATERIAL EXAMINED Holotype 9j , Australia: Queensland, Moses Ck, 45 km N. by E. Mt Finnigan (1547'S; 14517'E), x. 1980(Car dale) (ANIC). Paratypes. Australia: 1 $, Queensland, same data as holotype (ANIC); 3 C?, Gap Ck, 5 km ESE. MtFinnigan v.1981 (Naumann) (ANIC); 1 d" , Mt Webb N.P., 50 km N. Cooktown, vii.1976 (Monteith &Monteith) (ANIC); 1 cf , 4 km SSE. Cape Tribulation, nr Daintree, xi.1981 (Colless) (ANIC). 258 I. D. GAULD DREISBACHIA Townes Law/e/oTosquinet, 1903: 381. Type-species: Law/e/a w/raTosquinet, by monotypy. [Homonym of Laufeia Simon, 1889.]Dreisbachia Townes, 1962: 38. [Replacement name for Laufeia Tosquinet.] Small insects, fore wing length 4-5 mm; palp formula 5,3; clypeus of female flat, in same plane as face, itsapical margin truncate, that of male more clearly separated from face; surface of eye bearing long hairs;occipital carina complete. Epomia present; epicnemial carina present but with upper end rather far fromanterior margin of pleuron; mesoscutum subpolished, with close pubescence; propodeum in profile evenlyrounded, dorsally without distinct carinae, polished. Tarsal claws of female with basal lobes; hind tibia offemale internally uniformly hirsute; fore wing with 3r-m absent; hind wing with distal abscissa of Cu\absent; tergites 2-4 polished, with weak central rhombic areas; ovipositor weakly but evenly up-curved,projecting beyond apex of gaster by 0-6 times length of hind tibia. REMARKS. A small genus with few species scattered widely throughout the world. The majorityhave 3r-m present in the fore wing though Momoi (1966) described a New Guinea specieslacking this vein, and the Australian species is similar. The possession of hair on the eyes and uniformly on the mesoscutum, the rather flattenedmesoscutum, the inflated fore femora and the broad, evenly narrowed genae, together with theflat face and clypeus of the female suggest that Dreisbachia may be related to Afrosphincta andthe more specialized Schizopyga. Afrosphincta is only known from a single male and if a femalewere available it may prove to be congeneric with Dreisbachia - thus necessitating the placementof the name Dreisbachia as a junior synonym. A single species occurs in eastern Australia. Its hosts are not known, but the ichneumonid'sexternal morphology suggests that it attacks spiders in retreats. European species are known toattack spiders of the families Drassodidae (Aubert, 1969) and Clubionidae (BMNH collec-tions). Dreisbachia lutea sp. n. Fore wing length 4-5 mm; ocelli small, the lateral one separated from eye by 1-6 times its own maximumdiameter; malar space about 0-3 times basal mandibular width; head in dorsalview with genae moderatelybroad, evenly narrowed behind eyes; occipital carina complete. Mesoscutum with notauli moderatelystrongly impressed; mesopleuron smooth, with few fine scattered punctures; metapleuron highly polished,almost entirely smooth; propodeum with transverse carinae absent, metapleural carina complete, spiracleadjacent to this carina. Mid legs with tarsal segments 2-4 very short, together not as long as basitarsus;tergite 1 of gaster centrally somewhat striate, tergites 2+ quite coarsely punctate centrally; gaster quiteslender; ovipositor weakly up-curved, projecting beyond apex of gaster by 0-6 times length of hind tibia. Female and male head black, face, clypeus, mouthparts and ventral surface of antenna proximallyyellow; remainder of antenna brownish; alitrunk and gaster yellowish brown, laterally brighter yellow;tergites 7+ black; legs yellowish, hind femur distally black, hind tarsi infuscate. Wings hyaline. VARIATION. In a few individuals a trace of 3r-m is discernible in the fore wing. It is usually notcomplete and even if it is, it is only present in one wing. REMARKS. A distinctive little species on account of its colour pattern. It is only known fromMalaise trap catches on Black Mt in Canberra. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Australian Capital Territory, Canberra, xi.1978 (Tidemanri) (ANIC). Paratypes. Australia: 12 $, 2 cf, Australian Capital Territory, Canberra, xi-xii.1978 (Tidemanri)(ANIC; BMNH); 1 $, Canberra, i.1979 (Short) (ANIC); 1 cf , Canberra, ix.1981 (Gauld) (BMNH). ERIOSTETHUS Morley Eriostethus Morley, 1914: 34. Type-species: Eriostethus pulcherrimus Morley, by monotypy.Millironia Baltazar, 1964: 394. Type-species: Millironia trifasciata Baltazar, by original designation. Medium-sized insects, fore wing length 4-13 mm; palp formula 4, 3; clypeus in profile convex, its margin ICHNEUMONIDAE OF AUSTRALIA 259 slightly rounded; eye surface bare, occipital carina present and strong, or weak or absent. Epomia absent;epicnemial carina present; mesoscutum polished, glabrous; propodeum in profile evenly rounded, dorsallysmooth, without carinae except rarely in some males. Tarsal claws of female basally lobate; hind tibia offemale with an internal longitudinal glabrous groove; hind wing with distal abscissa of Cu\ absent. Tergites2-4 smooth to with rounded lateromedian prominences; ovipositor straight or slightly up-curved nearapex, projecting beyond apex of gaster by 0-9-1-2 times length of hind tibia. REMARKS. A moderate-sized genus centred in New Guinea, with a number of species inAustralia, South East Asia and Samoa. The systematic position of this genus is somewhat perplexing. Gauld (1984) suggested it mightbe related to some Neotropical genera, including Hymenoepimecis and Acrotaphus which itresembles in general appearance. All are elongate insects and have small rather globose headson long prothoraxes which lack epomia; the ovipositors are very long (for polysphinctines) inboth cases. However, close examination suggests these similarities may be the result ofevolutionary convergence; both genera have specializations not found in the other. The mostsignificant is the possession of the tibial groove in Eriostethus; this structure is presumably usedto manipulate the ovipositor. Such a feature is not present in the Neotropical genera. Eriostethuslacks the distal abscissa of Cui in the hind wing. This feature is particularly significant as someEriostethus are quite large insects and the loss of this vein is, in polysphinctines, closelycorrelated with small size. This loss of a wing vein coupled with the reduction in palp number(another feature associated with small size) suggest that Eriostethus may have been derived froma group of small polysphinctines such as Zatypota. The males of species of these two genera arevery similar and differ from other polysphinctines in having a 4, 3 palp formula. Hyme-noepimecis and Acrotaphus on the other hand show no reduction in palp formula and have veinCi present in the hind wing. Apart from obvious specializations these two genera have anumber of features in common with Polysphincta. Thus both Eriostethus and Hymenoepimecis/Acrotaphus are undoubtedly highly specialized tropical forest polysphinctines but it is suggestedthey are derived from different ancestors spreading south. Five species of Eriostethus occur in Australia, though one, E. maximus, exhibits a consider-able range of variation and may eventually be found to comprise more than one species. There are no host records from Australia but recently I have seen a specimen from NewGuinea reared from Acanthosoma or a similar genus of jewel spiders (Araneidae) (Fig. 97). Key to Australian species of Eriostethus 1 Gaster with lateral part of sternites 2-3 weakly sclerotized , not pigmented ; fore wing with Rs&M conspicuously bowed or angled so that the concave side is internal; cu-a usually proximal tobase of Rs&M, oblique; basal cell glabrous except for a band of hair along anterior margin (Fig. 83) maximus sp. n. (p. 260) Gaster with lateral part of sternites 2-3 with black-pigmented sclerotized patches ; fore wing withRs&M more or less straight; cu-a about opposite base of Rs&M, subvertical; basal cell hirsuteor with a glabrous central area, always with hairs in posterodistal corner (Fig. 84) 2 Occipital carina present as a distinct though often narrow flange 3 - Occipital carina absent or sometimes vestigial but not raised 4 3 Ocelli very large, the lateral one separated from eye by less than 0-2 of its minimum diameter; interocellar area yellowish orange; propodeum uniformly orange perkinsi (Baltazar) (p. 262) - Ocelli of moderate size, the lateral one separated from eye by 0-4-0-5 of its minimum diameter; interocellar area blackish; propodeum blue-black carinatus Baltazar (p. 260) 4 Ocelli small, the $ with lateral one separated from eye by at least its own minimum diameter, that of cf by 0-8-0-9 times its own diameter; propodeum and mid coxa orange minimus sp. n. (p. 261) Ocelli of moderate size, the lateral one separated from eye by 0-4-0-8 times its own minimumdiameter; propodeum and mid coxa bluish black pulcherrimus Morley (p. 262) 260 I. D. GAULD Eriostethus carinatus Baltazar (Fig. 90)Eriostethus carinatus Baltazar, 1964: 392. Holotype $, AUSTRALIA (BMNH) [examined]. Fore wing length 6-9 mm; ocelli of moderate size, the lateral one separated from eye by 0-4-0-5 times itsminimum diameter; malar space 0-9-1-0 times basal mandibular width; genae evenly narrowed behindeyes; occipital carina complete but not strongly raised. Pronotum mediodorsally rather short. Fore wingwith cu-a very slightly oblique , opposite base of Rs&M; Rs&M virtually straight , almost vertical ; basal cellwith a glabrous central area, peripherally hirsute; hind wing with Cu\+cu-a almost vertical so anterodistalcorner of subbasal cell is 80-85. Gaster quite stout, tergite 2 of female about 0-7 times as long as posteriorlybroad; lateral part of sternites 2-3 sclerotized, with conspicuous pigmented spots; ovipositor projectingbeyond apex of gaster by 1-0 times length of hind tibia. Female head, prothorax, mesoscutum and anterior part of mesopleuron reddish brown; antenna,interocellar area and remainder of alitrunk black; gaster black; tergite 1 margined laterally and posteriorlyand tergites 2+ posteriorly margined with white; fore leg reddish brown, hind two pairs black; face, forecoxa and tarsus infuscate; wings strongly infumate, especially near base. Male similar to female but withtergite 2 very extensively white. Cocoon: silvery grey, fibrous, 16 mm long, 2 mm maximum diameter, tapering evenly from centre;posterior end terminating in small hole, 1 mm diameter and anterior to this are seven small fin-like ribs;anterior end elongately tapered to a fibrous 'tufted' top, which is almost completely cut off duringemergence (Fig. 98). REMARKS. This species resembles E. pulcherrimus in colour and has often been mistaken for it incollections. E. carinatus is recognizable on account of the presence of a sharp occipital carinaand by possession by a black interocellar area. The stub of a vein on 2m-cu which Baltazar(1964) mentions as a distinguishing feature of this species is a teratological deformity of theholotype. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), Victoria (French) (BMNH). Australia: 1 $, New South Wales, Mt Kosciusko, xii.1924 'on snow on summit' (Dunstari) (BMNH);3 $, 2 cf, Tasmania, Hobart (Cole) (ANIC); 1 cf, Western Australia, Yallingup, ix-x.1913 (Turner)(BMNH). Eriostethus maximussp. n. (Figs 2, 83) Fore wing length 4-13 mm; ocelli very large, the lateral ones contiguous or almost contiguous with eye;malar space 0-6-0-7 times basal mandibular width; genae very strongly narrowed behind eyes; occipitalcarina complete, very strongly raised laterally and dorsally. Pronotum mediodorsally very long. Forewing with cu-a oblique, proximal to base of Rs&M; Rs&M conspicuously bowed or angled centrally soconcave side is proximal; basal cell glabrous except for a band of hair anteriorly (fig. 83); hind wing withCi -I- cu-a slightly oblique so anterodistal corner of subbasal cell is 75-80. Gaster relatively slender,tergite 2 of female 1-0-1-3 times as long as posteriorly broad; lateral part of sternites 2-3 weaklysclerotized, not pigmented; ovipositor projecting beyond apex of gaster by 1-0-1-2 times length of hindtibia. Female orange-brown with antenna, except at extreme base, blackish, tergites 3-1- blackish, all tergitesmargined posteriorly with white; hind legs brownish, tibia and tarsus infuscate. Wings slightly infumate.Male similar to female though often slightly paler. VARIATION. Although the majority of specimens closely conform in colour to the description,occasional small specimens (particularly females) have been found with tergite 2 and eventergite 1 blackish and the hind trochanter and femur also nigrescent. In some extreme cases eventhe hind coxa is infuscate. These specimens have tergite 2 quadrate and have the ovipositorprojecting beyond the apex of the gaster by about the length of the hind tibia, putting them onthe extremities of the ranges of both these characters. However, they do not differ in thesefeatures from small-sized but 'normally coloured' individuals, suggesting that they are merely ICHNEUMONIDAE OF AUSTRALIA 261 extreme variants. I have chosen to exclude these somewhat atypical specimens from theparatype material. REMARKS. E. maximus is easily distinguished from other Australian Eriostethus on account ofthe unusual bend in Rs&M and the posteriorly glabrous basal cell. It is apparently closely relatedto two tropical species, E. nodata (Baltazar) and E. krombeini (Baltazar), which occur in thePhilippines and New Guinea respectively. E. maximus is larger and differs in colour from thesespecies. This species seems to be restricted in Australia to tropical Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, 1 km N. of Rounded Hill (1517'S; 14513'E), v.1981 (Naumann)(ANIC). Paratypes. Australia: 1 cf , Queensland, Annan R. , 3 km W. by S. Black Mt (1541'S; 14512'E), ix.1980(Cardale) (ANIC); 2 $, 1 Cf, Annan R., 3 km W. by S. Black Mt (1541'S; 14512'E), iv.1981 (Naumann)(ANIC); 1 $, Capt. Billy Ck (1140'S; 14250'E), vii.1975 (Monteith) (ANIC); 1 $, 1 cf , Dividing Rg.,15 km W. Capt. Billy Ck, vii.1975 (Monteith) (ANIC); 2 $, 1 cf , 14 km W. by N. Hope Vale Mission(1516'S; 14459'E), x.1980 (Cardale) (ANIC); 1 cf, 7 km N. Hope Vale Mission (1514'S; 14507'E),x.1980 (Cardale) (ANIC); 4 $, 14 km W. by N. Hope Vale Mission (1516'S; 14459'E), v.1981(Naumann) (ANIC); 1 cf , 1-5 km SE. Kuranda, v.1980 (Naumann & Cardale) (ANIC); 4 $, Leo Creek,Mcllwraith Rg, 30 km NE. Coen, vi-vii.1976 (Monteith) (ANIC); 1 $ , Lockerbie Area, iv.1973 (Monteith)(ANIC); 2 Cf , 3-5 km SW. by S. Mt Baird (1510'S; 14507'E), v.1981 (Naumann) (ANIC); 1 $ , 4 cf , 1 kmSE. Mt Cook (1530'S; 14516'E), x.1980 (Cardale) (ANIC); 2 , 1 cf, Mt Cook Nat. Pk (1529'S,14516'E), v.1981 (Naumann) (ANIC); 1 $>, 3 km NE. Mt Webb (1503'S; 14509'E), v.1981 (Naumann)(ANIC); 1 $, 1 Cf , Peach Ck. Xing, 25 km NNE. Coen, vii.1976 (Monteith) (ANIC); 2 $, 2 cf , 1 km N.Rounded Hill, nr Hope Mission (1517'S; 14513'E), x.1980 (Cardale) (BMNH); 5 $ , 4 cf , 5 km W. by N.Rounded Hill (1517'S; 14510'E), x. 1980 (Cardale) (ANIC) ; 1 $ , 1 km N. Rounded Hill, nr Hope Mission(1517'S; 14513'E), v.1981 (Naumann) (ANIC); 1 cf , Sarina, xii. (TC); 1 $, 1 cf , Shipton's Flat (1547'S;14559'E), v.1981 (Naumann) (ANIC); 1 $ , Split Rock, 14 km S. Laura, vi.1975 (Monteith) (ANIC); 1 $,Upper Lankelly, Ck, vi.1971 (Monteith) (ANIC); 1 cf, Wenlock R. at Moreton, vi.1975 (Monteith)(ANIC); 1 cf , Northern Territory, Casuarina Beach, Darwin, x.1972 (Colless) (ANIC). Non-paratype material. Australia: 7 $ , 4 cf , Queensland (ANIC). Eriostethus minimus sp. n. Fore wing length 4-6 mm; ocelli small, the lateral one separated from eye by 1-0-1 -3 times its minimumdiameter; malar space 1-2-1-3 times basal mandibular width; genae long, quite evenly narrowed behindeye; occipital carina absent dorsally and laterally except at ventro-lateral extremity. Pronotum mediodor-sally moderately long. Fore wing with cu-a almost vertical, subopposite base of Rs&M; Rs&M slightlycurved, somewhat oblique; basal cell uniformly hirsute; hind wing with Cui+cu-a moderately oblique soanterodistal corner of subbasal cell is 75-80. Gaster moderately stout with tergite 2 of female 0-8-0-9 timesas long as posteriorly broad; lateral part of sternites 2-3 sclerotized, with pigmented spots; ovipositorprojecting beyond apex of gaster by 1-0 times length of hind tibia. Female head, alitrunk and anterior two pairs of legs orange; antenna and hind leg black; gaster black,tergites margined laterally and posteriorly with white; wings weakly infumate. Male similar to female butwith ocelli slightly larger, fore and mid coxa whitish and tergite 1 and 2 extensively pale. REMARKS. This species is close to E. pulcherrimus . The males of both species are very similar, themost obvious difference being in colour, but the females are readily distinguishable on accountof their small ocelli. E. minimus is only known from Queensland and New South Wales. MATERIAL EXAMINED Holotype $, Australia: Queensland, Bundaberg, vii.1971 (Franco) (ANIC). Paratypes. Australia: 1 cf , New South Wales, 1920 (Froggatt) (ANIC); 1 $, Springwood, i. (TC); 1 cf ,Taralga, i. (TC). Queensland. 1 $, Brisbane, ix.1962 (Lisle) (UQM); 1 $, Bundaberg, vii.1971 (Frauca)(BMNH); 4 cf, Mackay, on citrus, v.1959 (Campbell) (ANIC); 1 , Maryborough, on guava, iv.1959(Campbell) (ANIC); 1 cf, Mt Glorious, xii. 1976 (Boucek) (BMNH); 2 $>, Mt Tambourine, xii-i. (TC);1 $,Stanthorpe,i. (TC). 262 I. D. GAULD Eriostethus perkinsi (Baltazar) Millironia perkinsi Baltazar , 1964: 400. Holotype $, AUSTRALIA (BMNH) [examined].Eriostethus perkinsi (Baltazar) Gauld, 1984: 71. Fore wing length 6-10 mm; ocelli very large, the lateral one virtually contiguous with the eye; malar space0-7-0-8 times basal mandibular width; gena strongly narrowed behind eye; occipital carina complete, quitestrongly raised dorsally . Pronotum mediodorsally of moderate length. Fore wing with cu-a almost vertical,virtually opposite base of Rs&M; Rs&M slightly oblique, straight; basal cell fairly uniformly hirsute; hindwing with Cui+cu-a moderately oblique so anterodistal corner of subbasal cell is 60-65. Gastermoderately slender with tergite 2 of female 0-9-1-1 times as long as posteriorly broad; lateral part ofsternites 2-3 sclerotized, with pigmented spots; ovipositor projecting beyond apex of gaster by about 1-0times length of hind tibia. Female head and alitrunk orange, gaster black, tergites margined posteriorly with white, anterior twoalso margined laterally with white; antenna black; anterior two pairs of legs orange, hind legs black. Wingshyaline. Male similar to female but usually with face and fore and mid coxae pale yellowish. REMARKS. The very large ocelli distinguish E. perkinsi from all other species of Eriostethus with astraight Rs&M. This species is quite clearly intermediate between the 'genera' Eriostethus andMillironia as defined by Baltazar (1964). E. perkinsi is quite widely distributed in Australia but israther uncommon in collections. It is known from Australian Capital Territory, Queensland andWestern Australia. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), Queensland, Kuranda, vi-vii.1913 (Turner) (BMNH). Australia: 5 $, Australian Capital Territory, Canberra, i-ii. 1959-60 (Riek) (ANIC). Queensland; 1 $,Bluff Rg, nr Biggenden, viii.1971 (Franco) (ANIC); 2 $>, Brisbane, xii.1972 (Sedlacek) (TC); 1 $,Einasleigh R, 22 km S. by W. Lyndhurst, xi.1981 (Colless) (ANIC); 2 cf, Mackay, v.1959 (Campbell)(ANIC); 1 $, Mt Cook Nat. Pk, v.1981 (Naumann) (ANIC); 1 $, Shipton's Flat, x.1980 (Cardale)(ANIC); 1 $, Yeppoon, i.1975 (Howden) (TC); 1 $, Western Australia, Drysdale R., viii.1975 (Cardale}(ANIC). Eriostethus pulcherrimus Morley (Fig. 84)Eriostethus pulcherrimus Morley, 1914: 35. Holotype $, 'AUSTRALASIA' (BMNH) [examined]. Fore wing length 6-12 mm; ocelli of moderate size, the lateral one separated from the eye by 0-4-0-8 timesits own minimum diameter; malar space 0-9-1-1 times as long as basal mandibular width; genae long, fairlyevenly narrowed behind eye; occipital carina absent dorsally and laterally though junction of gena andocciput is usually discernible as a low ridge. Pronotum mediodorsally moderately long. Fore wing with cu-aalmost vertical, subopposite to base of Rs&M; Rs&M slightly curved, somewhat oblique; basal cell with amedian glabrous patch (Fig. 84); hind wing with Cui+cu-a quite strongly oblique so anterodistal corner ofsub-basal cell is 60-65. Gaster moderately stout with tergite 2 of female 0-8-1-0 times as long as posteriorlybroad; lateral part of sternites 2-3 sclerotized, with pigmented spots; ovipositor projecting beyond apex ofgaster by 0-9-1-1 times length of hind tibia. Female head, alitrunk except propodeum and fore leg orange; antenna, propodeum and hind leg black;mid leg infuscate, coxa blackish; gaster black, tergites 1 and 2 extensively and 3+ marginally white. Malesimilar to female. VARIATION. In specimens from Queensland and New South Wales tergites 1 and 2 are almostentirely white and the latter is clearly transverse; in specimens from Tasmania and southernVictoria tergite 2 is quadrate and both tergites 1 and 2 are black with a pale hind margin. REMARKS. E. pulcherrimus is very similar to E. minimus, but may be distinguished not only bythe characters given in the key but also by having Cu^+cu-a more oblique in the hind wing andby having a glabrous centre to the basal cell of the fore wing. Morley (1914) expressed doubts that the holotype came from Australia. I see no reason todoubt an Australian origin as the holotype agrees almost exactly in structure and colour pattern ICHNEUMONIDAE OF AUSTRALIA 263 with recent material collected in southern Queensland. This species is widely distributedthroughout eastern Australia, from tropical Queensland south to Tasmania. HOST RECORDS. None. MATERIAL EXAMINED 'Australasia': 1 $ (holotype) (Darnel) - (BMNH). Australia: 63 $ , 19 d", Australian Capital Territory, New South Wales, Queensland, Tasmania, Victoria(ANIC; BMNH; QUM; TC). ZAlTPOTAFoerster Zatypota Foerster, 1869: 166. Type-species: Ichneumon percontatorius Miiller, by subsequent designation,Viereck, 1914: 156. Polysphinctopsis Habermehl, 1917: 167. Type-species: Polysphincta examina Schmiedeknecht (= Glyptaalbicoxa Walker), by monotypy. Lycorinopsis Haupt, 1954: 110. Type-species: Lycorinopsis rhombifer Haupt (= Ichneumon percontator-ius Miiller), by original designation. Small insects, fore wing length 4-5 mm; palp formula 4, 3; clypeus moderately convex in profile, its apicalmargin slightly rounded; ocelli normal; eye surface almost bare; occipital carina usually present, veryrarely incomplete or even absent. Epomia present; epicnemial carina strong; mesopleuron subpolished,virtually glabrous; propodeum in profile evenly rounded, smooth, with lateromedian longitudinal carinae.Tarsal claws of female basally lobate; hind tibia of female without a glabrous groove internally; fore wingwith 3r-m absent; hind wing with distal abscissa of Cw t absent or faint, if present then first abscissa of Ci ismuch longer than cu-a. Tergites 2-4 with oblique grooves delimiting an almost rhombic central area;ovipositor straight, projecting beyond apex of gaster by less than 0-5 times length of hind tibia. REMARKS. Zatypota is a moderately large genus widely distributed throughout the world. It maywell be the largest polysphinctine taxon as most species are small, inconspicuous insects whichfavour damp habitats. They are seldom collected in any number, but Malaise traps in virtuallyany locality will yield a few specimens. I have seen eight undescribed species from Australia but I have to hand three males thatcannot be placed. It is likely that they represent additional species. The hosts of Australian species are not known but in western Europe specimens are notuncommonly reared as ectoparasites of immature Theridiidae. Key to Australian Species of Zatypota II Propodeum with posterior transverse carina entirely absent, lateromedian carina absent(Fig. 95); occipital carina absent or incomplete mediodorsally 2- Propodeum with posterior transverse carina present, usually complete, and with, at the veryleast, traces of lateromedian carinae extending forward from it, often with lateromediancarinae reaching anterior end of propodeum (Figs 94, 96); occipital carina complete,mediodorsally strongly raised 3 2 Hind wing with distal abscissa of Cu v absent (Fig. 86); head in dorsal aspect with genae long, evenly narrowed; hind tibia black rennefer sp. n.(p. 267) - Hind wing with distal abscissa of Cu\ present (Fig. 87); head in dorsal aspect with genae evenly rounded; hind tibia white, proximally and distally black velate sp. n. (p. 267) 3 Juxtacoxal carina present on metapleuron , usually complete (Fig .89) 4 - Juxtacoxal carina absent (Figs 90, 91) 5 4 Vertex of head extensively black, frontal orbits bright yellow; ovipositor sheath projecting beyond apex of gaster by 0-45 times length of hind tibia; female with mid leg with third tarsalsegment about 1-5 times length of fourth tarsal segment stellate sp. n.(p. 268) - Vertex of head and upper orbits entirely orange; ovipositor sheath projecting beyond apex of gaster by 0-35 times length of hind tibia; female with mid leg with third tarsal segment 1-8-2-2times length of fourth tarsal segment bingili sp. n. (p. 264) 5 Metapleuron virtually smooth and highly polished (Figs 88, 91); head in dorsal view predomi- nantly or at least partly orange , sometimes with interocellar area black 6 - Metapleuron rugose or rugulose , weakly polished (Fig. 90) ; head in dorsal view entirely black or black with pale orbital stripes 264 I. D. GAULD 6 Propodeal spiracle adjacent to metapleural carina (Fig. 88); latero-median longitudinal carinae complete from posterior transverse carina forward to propodeal margin (Fig. 94); tergites 2-4 of gaster black with large white triangular or rhombic central area dandiensis sp. n. (p. 265) Propodeal spiracle separated from metapleural carina by more than its own diameter (Fig. 91);lateromedian longitudinal carinae present on posterior transverse carina, anteriorly absent(Fig. 96); tergite 2 of gaster black, 3-4 black with hind margin white 2 phraxos sp. n. (p. 266) 7 Malar space very long, 1-5-1-8 times basal mandibular width (Fig. 92); gaster with ter- gites 2-4 unicolorous or with posterior margin narrowly pale-marked; head dorsally entirely black kauros sp. n. (p. 266) Malar space moderately long, 1-1 times basal mandibular width (Fig. 93); gaster with tergites2-4 dark brown with conspicuous cream-coloured central areas; head dorsally with frontalorbits yellow, the yellow mark curving backwards behind the ocelli celer sp. n. (p. 264) Zatypota bingili sp. n. (Fig. 89) Fore wing length 3-4 mm; ocelli small, the lateral one separated from the eye by 1-6-2-0 times its ownminimum diameter; malar space 0-6-0-8 times basal mandibular width; head in dorsal aspect with genaemoderately broad, fairly evenly rounded behind the eye; occipital carina complete, sharply raised;flagellum with 26-27 segments. Mesoscutum smooth, with notauli very strongly impressed to anteriormargin; mesopleuron smooth with few fine punctures ventrally; metapleuron smooth; juxtacoxal carinamore or less complete. Propodeum with posterior transverse carina complete, lateromedian carinaeextending from posterior transverse carina to anterior margin, anteriorly slightly convergent, enclosing asomewhat wedge-shaped area; propodeal spiracle separated from metapleural carina by almost its owndiameter, but connected to it by a small carina-like ridge (Fig. 89). Mid leg of female with third tarsalsegment 1-8-2-2 times as long as the fourth segment. Fore wing with 2r-m about 0-2 times as long asabscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cui absent entirely. Gaster withlateral part of sternites 3-6 very weakly sclerotized , unpigmented ; tergite 2 with strongly impressed , almostsmooth grooves defining rhombic central area. Ovipositor projecting beyond apex of gaster by 0-35 timeslength of hind tibia. Female head and alitrunk orange, antenna blackish, clypeus pale yellowish near margin; gaster withtergite 1 brownish orange, tergites 2+ black, 6 and 7 with narrow white margins; fore and mid legs orange,hind leg orange with tibia and tarsus black, femur slightly infuscate; wing hyaline, pterostigma blackishbrown. Male unknown. REMARKS. Z. bingili resembles Z. rennefer and Z. phraxos in general colour pattern but differsfrom both in possessing a well-developed juxtacoxal carina and lacking strongly sclerotizedsternal plates. The only other Australian Zatypota with a juxtacoxal carina is Z. stellata whichdiffers from bingili in many features, including colour pattern, relative lengths of tarsal segmentsand length of ovipositor.Z. bingili is only known from tropical Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, Cedar Ck, via Bingil Bay, v.1980 (Naumann & Cardale) (ANIC).Paratype. Australia: 1 $ , Queensland, Moses Ck, 4 km N. by E. Mt Finnigan, x. 1980 (Cardale) (ANIC). Zatypota celer sp. n. (Fig. 93) Fore wing length 4 mm; ocelli moderately small, the lateral one separated from the eye by 1-2 times its ownminimum diameter; malar space 1-1 times basal mandibular width (Fig. 93); head in dorsal aspect withgenae rather short, evenly narrowed behind eyes; occipital carina complete, distinctly raised; flagellumwith 21 segments. Mesoscutum smooth with notauli very strongly impressed to anterior margin; meso-pleuron highly polished; metapleuron weakly polished, rugose; juxtacoxal carina absent. Propodeum withposterior transverse carina complete, lateromedian longitudinal carinae present, delimiting an elongaterectilinear area; propodeal spiracle contiguous with metapleural carina. Mid leg with third tarsal segment1-8 times as long as the fourth segment. Fore wing with 2r-m obliterated by fusion of Rs and M in left wing, ICHNEUMONIDAE OF AUSTRALIA 265 in right wing with 2r-m very short, about 0-1 times as long as abscissa of M between 2r-m and 2m-cu; hindwing with distal abscissa of Cui absent entirely. Gaster with lateral part of sternites 3-6 weakly sclerotized,pigmented; tergite 2 with deep trans-striate grooves delimiting a rhombic central area. Ovipositorprojecting beyond apex of gaster by about 0.5 times length of hind tibia. Female head black, mouthparts, clypeus, facial and frontal orbits and lower part of gena yellow; antennabrownish, proximally yellow; alitrunk and gaster blackish, mesoscutal margin, pronotal margin, scutellum,mesepimeron, central areas of tergites 2-5 and most of legs whitish yellow; hind tibia proximally anddistally infuscate. Male similar to female but with central area of tergite 2 blackish. REMARKS. Z. celer is immediately recognizable on account of its colour pattern. Only it and Z.dandiensis have whitish central marks on the gastral tergites. Z. celer differs from dandiensis inthe sculpture of the metapleuron and in having striate grooves on the gastral tergites.Z. celer is only known from Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, Brisbane, ix-x.1972 (Sedlacek) (TC).Paratype. Australia: 1 cf , Queensland, Shipton's Flat (1547'S; 14514'E), x.1980 (Cardale) (ANIC). Zatypota dandiensis sp. n. (Figs 88, 94) Fore wing length 3-5 mm; ocelli small, the lateral one separated from eye by 1-6-1-8 times its ownminimum diameter; malar space 1-0-1-3 times basal mandibular width; head in dorsal aspect with genaequite broad and evenly tapered behind the eye; occipital carina complete, sharply raised; flagellum with20-23 segments. Mesoscutum smooth with notauli rather shallow; mesopleuron smooth and highlypolished, with sparse fine pubescence ventrally; metapleuron similar but with very fine punctures at basesof hairs; juxtacoxal carina absent (Fig. 88). Propodeum with posterior transverse carina complete;lateromedian carinae complete from posterior transverse carina to anterior margin, enclosing an elongaterectangular area (Fig. 94); propodeal spiracle contiguous with metapleural carina. Mid leg of female withthird tarsal segment 1-8-2-3 times length of fourth segment. Fore wing with 2r-m 0-1-0-3 times as long asabscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cu\ entirely absent. Gaster withlateral part of sternites 3-6 sclerotized and pigmented; tergite 2 with strongly impressed, smooth groovesdelineating rhombic central area. Ovipositor projecting beyond apex of gaster by 0-4-0-5 times length ofhind tibia. Female head orange, interocellar area, marks on genal orbits and antenna blackish; clypeus at leastmarginally yellow; alitrunk anteriorly orange, posteriorly with much of mesopleuron and entire meta-pleuron and propodeum black; gaster black, tergites 2-4 centrally white, tergite 5 usually white marked;fore leg orange, mid and hind legs black, with coxae and trochanters white and usually with small, indistinctwhitish marks on outer side of tibiae; wings infumate; pterostigma black. Male similar to female butalitrunk orange except for propodeum and metathorax which are black; mid femur orange-brown, hindfemur black with distal apex white, hind tibia black, centrally white. VARIATION. Two small females in the Townes collection and the female from Canberra have acolour pattern like that of the males. The wings of these specimens are less densely infumatethan other females. Cocoon: found under bark; 10 mm long, 2-5 mm in maximum diameter, cylindrical, tapered toboth ends, one of which is open as a small hole 1 mm diameter; outer surface irregularly fluted,whitish with loops of thick golden silk protruding randomly (Fig. 99). REMARKS. Z. dandiensis is immediately recognizable on account of its striking and handsomecolour pattern. The only other Australian species with a similarly coloured gaster is Z. celer, butthis species has a different coloured head and alitrunk and a rugose metapleuron.Z. dandiensis is widely distributed throughout the southern half of Australia. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Australian Capital Territory; Canberra, iv.1962 (Rlek) (ANIC) (with cocoon).Paratypes. Australia: 1 d", Australian Capital Territory. Canberra, xii.1958 (Riek) (ANIC); 1 $, 266 I. D. GAULD Canberra, ii.1978 (Tidemann) (ANIC); 1 cf, New South Wales, Wambool Common, 18 km ESE.Bathurst, iv.1980 (Cardale) (ANIC); 1 $ , Queensland, Stanthorpe, xii. (TC); 1 $ Tasmania, Mt Field NP,250 m, i-ii.1983 (Gauld) (BMNH); 1 $, Port Arthur, i-ii. (TC); 1 $, Western Australia, Yallingup,xii. 1913 (Turner) (BMNH). Zatypotakaurossp. n. (Figs 90, 92) Fore wing length 2-5-3-0 mm; ocelli small, the lateral one separated from eye by 1-2-1-3 times its ownminimum diameter; malar space 1-5-1-8 times as long as basal mandibular width (Fig. 92); head in dorsalview with genae moderately long, strongly narrowed behind eye; occipital carina complete, present as astrong raised flange; flagellum with 14-16 segments. Mesoscutum smooth with notauli strongly impressedto anterior margin; mesopleuron smooth and polished; metapleuron finely rugose; juxtacoxal carinaabsent. Propodeum with posterior transverse carina present; lateral longitudinal carinae present fromtransverse carina forward to anterior margin of propodeum, enclosing an elongate rectilinear area;propodeal spiracle contiguous with metapleural carina (Fig. 90). Mid leg with third tarsal segment 1-6-1-8times as long as fourth segment. Fore wing with 2r-m obliterated by the fusion of Rs and M, or if presentthen thicker than long; hind wing with distal abscissa of Cu\ entirely absent. Gaster with lateral part ofsternites 3-6 sclerotized and pigmented; tergite 2 with quite strongly impressed, smooth grooves defining arhombic central area. Ovipositor projecting beyond apex of gaster by about 0-4 times length of hind tibia.Female head, black; antenna infuscate, base of antenna, clypeal margin and mouthparts yellow;alitrunk, gaster and legs brownish, mesoscutal margin, subalar prominence and coxae yellowish. Wingshyaline; pterostigma dark brown. Male similar to female but with propleuron, mesopleuron, metathorax,propodeum and gaster blackish, tergites 2+ paler margined; coxae whitish, the hind one infuscateproximally; hind tarsus infuscate. VARIATION. A single female collected with the holotype differs in having much of the meso-pleuron, metathorax and propodeum blackish; the anterior end of tergite 1 is infuscate as are thelast two tergites. It is on the basis of this variation I have associated the male. I doubt that theextreme difference in colour pattern is normal sexual dichromatism but probably variation, asthe male was collected further south. REMARKS. A distinctive species on account of its more or less totally black head, long malarspace and short antennae. It is known from southern Queensland and Australian CapitalTerritory. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, Stanthorpe, 1000m, i. (TC). Paratypes. Australia: 3 $, Queensland, Stanthorpe, 1000 m, xii., iii. (TC): 1 cf, Australian CapitalTerritory, Canberra, Black Mt, x.1981 (Gauld) (BMNH). Zatypotaphraxossp. n. (Figs 91, 96) Fore wing length, 4-5-5-5 mm; ocelli small, the lateral one separated from eye by 1-3-1-4 times its ownminimum diameter; malar space 0-6-0-7 times basal mandibular width; head in dorsal aspect with genaebroad, evenly tapered behind eye; occipital carina complete, sharply raised; flagellum with 25-26segments. Mesoscutum polished, notauli quite strongly impressed, reaching anterior margin; mesopleuronhighly polished, smooth; metapleuron similar; juxtacoxal carina absent. Propodeum with posteriortransverse carina complete, lateromedian longitudinal carinae vestigial, present only at junction withposterior transverse carina (Fig. 96); propodeal spiracle separated from metapleural carina by more thanits own diameter, the two joined by a raised carina-like ridge (Fig. 91). Mid leg with third tarsal segment1 -6-1-9 times as long as fourth segment. Fore wing with 2r-m 0-1-0-2 times as long as abscissa of M between2r-m and 2m-cu; hind wing with distal abscissa of Cu\ entirely absent. Gaster with lateral part of sternites3-6 sclerotized and pigmented; tergite 2 with quite strongly impressed, smooth grooves defining a rhombiccentral area. Ovipositor projecting beyond apex of gaster by 0-4 times length of hind tibia. Female head, alitrunk and anterior part of tergite 1 orange; antenna blackish; posterior part of tergite 1and tergites 2+ black, tergite 2 with posterolateral corner white, tergites 3+ with posterior margin ICHNEUMONIDAE OF AUSTRALIA 267 narrowly white; anterior two pairs of legs orange; hind leg with coxa orange; distally infuscate, remainderof leg blackish. Wings weakly infumate; pterostigma pale translucent brown. Male unknown. REMARKS. Z. phraxos is distinctive in having the posterior transverse carina of the propodeumcomplete but the lateromedian longitudinal carinae vestigial. In colour it resembles Z. renneferand to some extent Z. bingili.It is only known from tropical Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, 1-5 km SE. Kuranda, v.1980 (Naumann & Cardale) (ANIC). Paratypes. Australia: 1 $, Queensland, Kuranda, vi-vii.1913 (Turner) (BMNH); 1 $, Mt Cotton,iv.l966(C/tort)(QUM). Zatypota rennefer sp. n. (Figs 86, 95) Fore wing length 6 mm; ocelli small, the lateral one separated from eye by 1-5 times its own minimumdiameter; malar space 0-8 times basal mandibular width; head in dorsal aspect with genae moderatelybroad, evenly tapered behind eye to margin of occiput where there is an abrupt corner, but the occipitalcarina is not present; flagellum with 30 segments. Mesoscutum smooth, with notauli quite stronglyimpressed to anterior margin; mesopleuron smooth and polished; metapleuron smooth; juxtacoxal carinaabsent. Propodeum with posterior transverse and lateromedian carinae entirely absent (Fig. 95); pro-podeal spiracle separated from metapleural carina by its own diameter, the two joined by a raisedcarina-like ridge. Mid leg with third tarsal segment 1-4 times as long as the fourth segment. Fore wing with2r-m 0-2 times as long as abscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cientirely absent (Fig. 86). Gaster with lateral part of sternites 3-6 sclerotized and pigmented; tergite 2 withmoderately impressed smooth grooves defining a rhombic central area. Ovipositor projecting beyond apexof gaster by 0-3 times length of hind tibia. Female head, alitrunk and anterior two pairs of legs orange; antenna except scape ventrally, gaster andhind legs except coxa, mainly black; tergites 3+ posteriorly white margined, tergite 2 with white spots inposterolateral corners; tergite 1 laterally pallid; hind coxa orange; distally infuscate. Wings weaklyinfumate; pterostigma blackish. Male unknown. REMARKS. Z. rennefer resembles Z. phraxos in shape and colour pattern, but differs in lacking anoccipital carina and having reduced propodeal carination. The only other Australian specieswith these carinae incomplete is Z. velata which is immediately recognizable by the presence of adistal abscissa for Cu\ in the hind wing.Z. rennefer is only known from southern Queensland. HOST RECORDS. None. MATERIAL EXAMINEDHolotype $, Australia: Queensland, Stanthorpe, 1000 m, iii. (TC). Zatypota stellata sp. n. Fore wing length 3-5 mm; ocelli small, the lateral one separated from eye by 1-6 times its own minimumdiameter; malar space 0-9 times basal mandibular width; head in dorsal aspect with genae moderatelywide, quite strongly narrowed behind eyes; occipital carina complete, sharply raised; flagellum with 26segments. Mesoscutum smooth with notauli strongly impressed, not quite reaching anterior margin;mesopleuron smooth and polished; metapleuron with upper part smooth, lower part with slight rugosity;juxtacoxal carina complete. Propodeum with posterior transverse carina complete, lateromedian longitu-dinal carinae present, extending from posterior transverse carina forward to anterior margin of pro-podeum, anteriorly convergent and defining an elongate wedge-shaped area, unusual in having lateralcarina almost complete; propodeal spiracle separated from metapleural carina by 0-5 times its owndiameter. Mid leg with third tarsal segment about 1-5 times as long as fourth segment. Fore wing with 2r-mabout 0-1 times as long as abscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cu\absent entirely. Gaster with lateral part of sternites 3-6 sclerotized, weakly pigmented; tergite 2 withmoderately strongly impressed grooves defining a rhombic central area, the grooves tending to be weakly 268 I. D. GAULD trans-striate. Ovipositor projecting beyond the apex of gaster by 0-35 times length of hind tibia. Female head black, facial and frontal orbits, clypeal margin, area below antennal insertion and lower 0-2of gena, yellow; antenna brownish; scape ventrally paler. Alitrunk and gaster reddish brown; anteriormargin of mesoscutum, scutellum, margin of pronotum, subalar prominence, mesepimeron and tegulabright yellow; ovipositor sheath darker brownish. Legs reddish brown with coxae and trochanteralsegments yellow; hind tibia centrally pale, proximally and distally infuscate. Wings almost hyaline;pterostigma dark brown. Male unknown. REMARKS. Z. stellata is the only Australian species of Zatypota, apart from Z. bingili, which has ajuxtacoxal carina. It is most easily separated from the latter on account of its colour pattern.Z. stellata is only known from north Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, 5 km W. by N. Rounded Hill, nr Hope Vale Mission (1517'S;14510'E), x.1980 (Cardale) (ANIC). Zatypota velata sp. n. (Fig. 87) Fore wing length 3-6 mm; ocelli moderately large, the lateral one separated from the eye by about 1-0 timesits own minimum diameter; malar space 0-5-0-7 times basal mandibular width; head in dorsal aspect withgenae moderately short, abruptly narrowed behind eyes; occipital carina interrupted mediodorsally;flagellum with 20-22 segments. Mesoscutum smooth with notauli weakly impressed; mesopleuron smoothand polished; metapleuron smooth; juxtacoxal carina absent. Propodeum with posterior transverse andlateromedian longitudinal carinae absent; propodeal spiracle close to but not quite contiguous withmetapleural carina. Mid leg with third tarsal segment 1-8-1-9 times as long as fourth segment. Fore wingwith 2r-m 0-2-0-3 times as long as abscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa ofCui present, reaching almost to margin of wing (Fig. 87). Gaster with lateral part of sternites 3-6sclerotized and pigmented; tergite 2 with deep, smooth grooves defining a central rhombic aiea. Ovipositorprojecting beyond apex of gaster by 0-5 times length of hind tibia. Female and male head, antenna, propleuron, propodeum and most of gaster black, remainder ofalitrunk orange-brown; tergites 3+ of gaster with hind margin white marked, tergite 2 with white spotslaterally; fore leg brownish to blackish with indistinct pale patches on femur; tibia and often with tarsipallid; mid and hind legs black; tibia broadly white centrally. Wings infumate; pterostigma blackish. REMARKS. A very distinctive species on account of the complete distal abscissa of Cui in the hindwing. This vein is completely absent in other Australian species.Z. velata is only known from Canberra. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Australian Capital Territory, Canberra, x-xi.1946, ex bark-living spider (Riek)(ANIC). Paratypes. Australia: 2 9, 16 O", Australian Capital Territory, Canberra, x-xi.1946, ex bark-livingspider (Riek) (ANIC; BMNH); 1 $, Canberra, iii.1960 (Riek) (ANIC); 1 $, Canberra, iv.1961 (Riek)(ANIC). Tribe PIMPLINI(= Ephialtini sensu Townes) The Pimplini constitutes a holophyletic group of genera, whose component taxa are most easilyrecognized by their final instar larvae; these have a strongly developed epistomal arch and atotally reduced hypostoma. The adults are small to large insects with the ovipositor neverconspicuously longer than the gaster, an almost straight mesopleural suture and the first abscissaof Ci in the hind wing very short. World-wide the Pimplini contains nine genera, four of which, Alophopimpla, Echthromor-pha, Lissopimpla and Xanthopimpla, occur in Australia. With the exception of Alophopimpla ICHNEUMONIDAE OF AUSTRALIA 269 these are large tropicopolitan taxa. The other pimpline genera are mainly sub-boreal and northtemperate insects though a few species occur on higher ground in the tropics. In Australia, the Pimplini includes some of the commonest and most conspicuous of allichneumonids. Several are quite large, brightly patterned insects, frequently observed insuburban habitats. The majority of the rare species are apparently restricted to humid forestsand consequently are most commonly encountered along the eastern coastal ranges of thecontinent. ALOPHOPIMPLA MomoiAlophopimpla Momoi, 1966: 160. Type-species: Alophopimpla polia Momoi, by original designation. Medium-sized species, fore wing length 6 mm; clypeus with margin slightly concave; mandible not twisted,moderately narrowed; occipital carina dorsally absent. Epicnemial carina present; mesopleuron withouttrans-striate grooves; mesopleural suture not centrally angled; propodeum without carinae dorsally,spiracle subcircular. Tarsal claws of female without basal lobes, with spatulate bristles; hind femur simple.Fore wing with 3r-m present enclosing a rhombic areolet; hind wing with first abscissa of Cui about 0-3times length of cu-a. Tergites of gaster smooth and polished; ovipositor projecting beyond apex of gasterby 1-0-1-4 times length of hind tibia. REMARKS. A small genus with two species, one in Australia and the other, the type-species, inNew Guinea (Momoi, 1966). Nothing is known of the biology or habits of these insects. Alophopimpla kluia sp. n. Mandible with upper tooth slightly the longer; posterior ocellus separated from eye by about 0-5 times itsmaximum diameter, interocellar distance 0-6 times maximum ocellar distance. Antenna slightly clavate;scape apically obliquely truncate 60; flagellum with 23 segments, the penultimate distal ones slightlylonger than broad. Epomia virtually absent; epicnemial carina present. Scutellum carinate laterally only atextreme anterior end. Submetapleural carina weak, not raised anteriorly; propodeum with concavitypresent above and before spiracles. Fore tibia distally inflated; fore tarsal segments 3 and 4 bearing longstout bristles. Fore wing with cu-a opposite base of Rs&M; areolet moderately small. Gaster with tergite 1of female barely longer than posteriorly broad, the sternite extending 0-2 of its length. Ovipositor slightlydecurved, projecting beyond apex of gaster by 1-4 times length of hind tibia; lower valvulae simple, notenclosing the upper. Head and antenna black; alitrunk reddish brown, posteriorly infuscate; gaster black, profusely white-marked on most of tergites 1 and 2, 3 posteriorly and laterally, 4+ posteriorly, and on all sternites. Anteriortwo pairs of legs white, femora and tibiae black-striped, tarsi reddish; hind leg black with large irregularwhite marks on coxa, trochanter and femur. Wings strongly and uniformly infumate. REMARKS. A distinctive pimpline in Australia on account of its colour pattern. It resembles theNew Guinea species A. polia except that the epomia is much shorter and weaker, the epicnemialcarina is present, the areolet is smaller and the ovipositor longer. HOST RECORDS. None. MATERIAL EXAMINEDHolotype $, Australia: Queensland, Mt Tambourine, xi.1977 (Galloway} (ANIC). ECHTHROMORPHA Holmgren Echthromorpha Holmgren, 1868: 406. Type-species: Echthromorpha maculipennis Holmgren (=Ichneumon agrestorius Swederus), by subsequent designation, Ashmead, 1900a: 57. Syene Snellen van Vollenhoven, 1878: Ixxvi. Type-species: Cryptus notulatorius F. (= Ichneumonagrestorius Swederus), by subsequent designation, Townes, 1960a: 43. Stagmopimpla Saussure, 1892: 16. Type-species: Stagmopimpla hyalina Saussure (= Ichneumon agrestor-ius Swederus), by subsequent designation, Viereck, 1914: 136. Rhynchopimpla Kriechbaumer, 1894a: 51 . Type-species: Pimpla interrupta Brulle (= Ichneumon agrestor-ius Swederus), by monotypy. Polyamma Kriechbaumer, 18946: 304. Type-species: Pimpla continua Brulle sensu Kriechbaumer (=Ichneumon agrestorius Swederus), by monotypy. 270 I. D. GAULD Chrysopimpla Cameron, 1899: 185. Type-species: Chrysopimpla ornatipes Cameron (= Ichneumonagrestorius Swederus), by subsequent designation, Viereck, 1914: 32. Allotheronia Ashmead, 1900a: 55. Type-species: Allotheronia 12-guttata Ashmead (= Cryptus intricatoriusF.), by original designation. Glyptogastra Ashmead, 1900a: 57. Type-species: Glyptogastra hawaiiensis Ashmead, by original designa-tion. Polyhamma Dalla Torre, 1901: 455. [Unjustified emendation.] Medium to large-sized species, fore wing length 8-22 mm; clypeal margin slightly concave; mandiblesstrongly narrowed, twisted; occipital carina complete. Epicnemial carina complete; mesopleuron withouttrans-striate grooves; mesopleural suture very slightly angled centrally; propodeum without distinctcarinae dorsally, sometimes with hypophyses, spiracle elliptical. Tarsal claws of female large, without astrong basal lobe, but with a spatulate bristle; hind femur simple. Fore wing with 3r-m present, enclosing asubpetiolate, rhombic areolet; marginal cell with infumate spot distally; hind wing with first abscissa of Cu\very short or obsolete (Fig. 6). Tergites of gaster somewhat polished, with or without punctures; ovipositorprojecting beyond apex of gaster by 0-9-1-4 times length of hind tibia, usually slightly decurved. REMARKS. Echthromorpha is a moderately large genus centred in the Papuan subregion but witha few species in Africa. One species, E. agrestoria is extremely widespread. It appears to migrate(Common, 1954) and has spread across the south Pacific, colonizing many remote islands(Perkins, 1952; Mason, 1974). Three species occur in Australia. One, E. intricatoria, is particularly common in the coolerparts of the south-east and may be observed in large numbers searching low vegetation and treeroots in wooded areas. The related but largely allopatric species E. agrestoria is more common inthe north. The third species, E. nigricornis, is predominantly a tropical insect. None of theAustralian species is endemic. Echthromorpha species are known to parasitize a variety of lepidopterous pupae andprepupae including those of Noctuidae, Hesperiidae, Agaristidae, Anthelidae, Bombycidae,Lycaenidae, Lymantriidae, Nymphalidae, Psychidae, Xyloryctidae, Papilionidae, Tortricidaeand Saturniidae (Gauld, 1984). Key to Australian species of Echthromorpha 1 Malar space 0-8-1-0 times as long as basal mandibular width (Fig. 14); infumate spot at distal apex of fore wing very weak ; alitrunk and gaster uniformly orange nigricornis (Smith) (p . 272) Malar space 1-3-1-5 times as long as basal mandibular width (Fig. 15); infumate spot at distalapex of fore wing distinct; alitrunk and usually gaster, at least, marked with black 2 2 Propodeal apophyses distinct; metapleuron uniformly punctate (Fig. 12); alitrunk and gaster black, with ivory markings intricatoria (F.) (p. 271) Propodeal apophyses absent; metapleuron with a large, smooth area anteriorly (Fig. 13);alitrunk and gaster yellowish with black markings agrestoria (Swederus) (p. 270) Echthromorpha agrestoria (Swederus)(Fig. 13) Ichneumon agrestorius Swederus, 1787: 279. Holotype 9, TAHITI (BMNH) [examined]. Ichneumon melioratorius Fabricius, 1793: 147. Holotype $ , TAHITI (BMNH) [isotypic with /. agrestorius] [examined] .Pimpla interrupta Brulle, 1846: 91. Holotype $, NEW CALEDONIA (MNHN) [examined]. [Synonymized by Townes <?/fl/., 1961:38].Pimpla insidiator Smith, 1863: 9. Lectotype $, MISOOL (UM), designated by Townes et al., 1961: 37 [examined]. [Synonymized by Townes, 1958: 42.]Notiopimpla platymischa Vachal, 1907: 120. Lectotype $, NEW CALEDONIA (MNHN), designated by Townes etal., 1961: 38 [examined]. [Synonymized by Townes, 1958: 43.]Echthromorpha striata Krieger, 1909: 306. Lectotype $ , TAHITI (NM), designated by Townes et al., 1961: 34. [Synonymized by Morley, 1913a: 44.]Echthromorpha conopleura Krieger, 1909: 321. Lectotype $ , MARIANA Is. (not Brazil as stated) (MNHU), designated by Townes etal., 1961: 35. [Synonymized by Townes, 1958: 42.] ICHNEUMONIDAE OF AUSTRALIA 271 Echthromorpha immaculata Krieger, 1909: 331. Lectotype cf , FIJI (NM), designated by Townes et al, 1961: 37. [Synonymized by Townes, 1958: 42.]Echthromorpha agrestorius (Swederus) Morley, 1909: 135.Echthromorpha agrestoria (Swederus); Morley, 1913a: 44.Echthromorpha notulatoria var. immaculata Morley, 1913a: 46. Lectotype cf , 'INDIA' (probably Mariana Is.) (BMNH), designated by Townes et al., 1961: 46. [Synonymized by Townes, 1958: 42.] [Junior primary homonym of Echthromorpha immaculata Krieger.]Echthromorpha diversor Morley, 1913a: 47. Holotype cf , 'SOLOMON Is.' or 'NEW HEBRIDES' (BMNH) [examined]. [Synonymized by Perkins, 1952: 536.] Medium to large-sized species, fore wing length 8-18 mm. Malar space 1-3-1-5 times as long as basalmandibular width; face centrally very sparsely punctate. Mesoscutum moderately closely punctate,scutellum strongly convex; post-scutellum 1-5-2-0 times as long as broad. Mesopleuron ventrally closelypunctate, dorsally almost impunctate; epicnemial carina reaching almost to subalar prominence; meta-pleuron anteriorly smooth, posteriorly punctate (Fig. 13). Propodeum dorsally punctate anteriorly,posteriorly smooth, without apophyses. Tergites 2-4 of female gaster very closely and coarsely punctatecentrally, those of male more sparsely punctate. Ovipositor projecting beyond apex of gaster by 0-9-1-0times length of hind tibia, its apex depressed with weak dorsal ridges. Yellowish orange, black-marked on frons, interocellar area, occiput, mesoscutum in three stripes,pronotum, epicnemium and tergites 2-6 except posteriorly. Flagellum infuscate, legs, propodeum andposterior apex of gaster orange. Wings hyaline with distinct infumate spots present on distal apices of forewings. VARIATION. Considerable variation in the extent of the black colour exists; many smaller speci-mens and some large ones have only the frons, interocellar area and mesoscutal vittae black. REMARKS. Townes et al. (1961) recognized a number of subspecies largely on colour pattern. Theones included above in synonymy fall within the range of variation I have observed in Australianspecies. Almost certainly many of the other 'subspecies' are also mere colour varieties andprobably the best course of action is to treat the whole agrestoria complex as a single specieswithout recognizing subspecies. As recognized above, E. agrestoria is widespread throughout the south Pacific and Melanesia,extending west to about the Moluccas. In Australia it is primarily restricted to Queensland withisolated records from northern New South Wales. It is apparently common on Lord Howe Island(Map 5). HOST RECORDS. Hesperiidae: Parnara amalia (Semper) (DPIQ). Noctuidae: Anomisflava (F.)(DPIQ); A. lyona (Swinhoe) (DPIQ). Tortricidae: Cryptophlebia ombrodelta (Lower) (Iron-side, 1974). MATERIAL EXAMINED Tahiti: 1 $ (holotype of agrestorius and melioratorius) (BMNH). New Caledonia: 1 $ (holotype ofinterrupta) (MNHN). Indonesia: 1 $ (lectotype of insidiator) , Misool (UM). 'India' (probably MarianaIs.): 2 cf (lectotype of notularia var. immaculata} (BMNH). 'Solomon Is.' or 'New Hebrides': 1 cf(paralectotype of notularia var. immaculata) (Liddell) (BMNH). Australia: 34 $, 27 cf , New South Wales, Queensland (ANIC, BMNH, QM, QUM). Echthromorpha intricatoria (F.)(Figs 6, 12, 15) Cryptus intricatorius Fabricius, 1804: 77. Holotype 9, 'NovA CAMBRIA' (UZM).Ichneumon intricatorius (Fabricius) Thunberg, 1822: 278.Pimpla excavata Guillou, 1841: 322. Holotype $, AUSTRALIA (lost).Pimpla intricatoria (Fabricius) Erichson, 1842: 254.Echthromorpha intricatoria (Fabricius) Krieger, 1899: 59. Allotheronia 12-guttata Ashmead, 1900a: 55. Holotype $, NEW ZEALAND (USNM) [examined]. [Synony-mized by Parrott, 1952: 163.] Medium to large-sized species, fore wing length 9-18 mm. Malar space 1-3-1-5 times as long as basalmandibular width (Fig. 15); face centrally coarsely punctate. Mesoscutum coarsely punctate, scutellum 272 I. D. GAULD abruptly declivous posteriorly; postscutellum 1-2-1-3 times as broad as long. Mesopleuron fairly evenlyclosely and coarsely punctate; epicnemial carina reaching about 0-7-0-8 of way up mesopleuron;metapleuron uniformly closely punctate (Fig. 12). Propodeum dorsally closely reticulopunctate, gradingposteriorly to trans-striate, apophyses distinct. Tergites 2-4 of gaster virtually impunctate centrally,smooth and polished. Ovipositor projecting beyond apex of gaster by 1-0-1-2 times length of hind tibia, itsapex depressed and with upper valve bearing strong ridges. Black; face, genae, antenna, fore leg, mid leg excluding coxa and hind leg except for coxa andtrochanteral segments, orange; subalar prominences, axillae, upper and lower posterior corners ofmesopleuron, postscutellum, a spot on metapleuron, propodeal apophyses and paired spots on tergites 1-6ivory or pale yellow. Wings almost hyaline, with a strongly infumate spot at apex of fore wing. REMARKS. This is one of the commonest southern insects in Australia. It can often be observed inlarge numbers in damp, shaded habitats such as woodland understorey. It is found in drierlocalities also, though less frequently. E. intricatoria is also quite common in New Zealand(Parrott, 1952). It has been recorded from New Guinea (Morley, 1915) and India (Morley,19136) though almost certainly these records are based on misidentifications. In Australia this species is one of the more commonly reared ichneumonids and is known froma considerable range of lepidopterous hosts, suggesting it has little host specificity. It is ofpossible economic importance as frequently it can be found parasitizing pests such as cutworms(Froggatt, 1910). HOST RECORDS. Agaristidae: Phalaenoides glycinae Lewis (ANIC). Anthelidae: Anthela denticu-lata (Newman) (Cameron, 1912); A. xantharcha (Meyrick) (QM). Bombycidae: Bombyx mori(L.) (Chadwick & Nikitin, 1976). Hesperiidae: Hesperilla chrysotricha (Meyrick & Lower)(Parrott, 1957); H. donnysa Hewitson (Parrott, 1957). Lycaenidae: Ogyris olane Hewitson(Parrott, 1957). Lymantriidae: Euproctis edwardsi (Newman); Olene mendosa Hiibner (Chad-wick & Nikitin, 1976); Teia anartoides Walker (ANIC; DAT). Noctuidae: Persectania ewingii(Westwood) (Martyn etal, 1977); Spodoptera exempta (Walker) (Chadwick & Nikitin, 1976).Nymphalidae: Vanessa itea (F.) (ANIC). Psychidae: Hyalarcta huebneri (Westwood) (Chad-wick & Nikitin, 1976); Lomera caespitosa (Oke) (Chadwick & Nikitin, 1976). Xyloryctidae:Neodrepta luteotactella (Walker) (DPIQ). In New Zealand it has been recorded from a variety of other hosts including Epiphyaspostvittana (Gourlay, 1926; Dumbleton, 1932). MATERIAL EXAMINED New Zealand: 1 $ (holotypeof 12-guttata) (USNM). Australia: 156 $, 112 d", Australian Capital Territory, New South Wales, Queensland (SE.), SouthAustralia, Tasmania (including Bass Strait Islands), Victoria, Western Australia (Map 6). Echthromorpha nigricornis (Smith)(Fig. 14) Pimpla nigricornis Smith, 1865: 64. Holotype cf , PAPUA NEW GUINEA (UM) [examined].Echthromorpha maxima Krieger, 1909: 334. Holotype $, MOLUCCAS (NM). [Synonymized by Perkins, 1952: 536.]Echthromorpha fastigata Krieger, 1909: 336. Holotype 9> MOLUCCAS (NM). [Synonymized by Townes et al, 1961:45.]Echthromorpha nigricornis (Smith) Perkins, 1952: 536. Very large species, fore wing length 20-22 mm. Malar space 0-8-1-0 times as long as basal mandibularwidth (Fig. 14); face centrally coarsely punctate. Mesoscutum coarsely and closely punctate, scutellum inprofile almost pyramidal; postscutellum tranverse, about 2-5 times as broad as long. Mesopleuronpolished, with large close punctures posteroventrally, the punctures sparser dorsally; epicnemial carinareaching about 0-6 of way up mesopleuron; metapleuron anteriorly entirely smooth, posteroventrallycoarsely and closely punctate. Propodeum dorsally coarsely punctate in anterior 0-4, posteriorly smooth,apophyses absent. Tergite 2 of gaster with few punctures centrally, tergites 3 and 4 more closely punctatebut still with punctures separated by usually more than their own diameter. Ovipositor projecting beyondapex of gaster by 1-4 times length of hind tibia, its apex depressed and with upper valve bearing fineridges. ICHNEUMONIDAE OF AUSTRALIA 273 Orange; flagellum and interocellar area infuscate, legs golden yellow. Wings very weakly infumate, withweak trace of darker infumate spot in distal end of marginal cell. REMARKS. This large species is easily distinguished by the shorter malar space and weak spot onthe fore wing. It is widely distributed from the Moluccas to the Solomon Is. but in Australia it isrestricted to Queensland and northern New South Wales. HOST RECORDS. Papilionidae: Ornithoptera priamus euphorion (Gray) (Morley, 19130). Satur-niidae: Antheraea saccopoea Turner (AM). MATERIAL EXAMINED Papua New Guinea: 1 cf (holotype of nigricornis) (UM). Australia: 1 $, New South Wales, Eltham, ii.1929 (Chadwick) (DAR); 2 $, 2 cf, Queens-land, Middle Claudie River, ix-x.1974 (Daniels) (AM); 1 $, Townsville, xii.1901 (Dodd)(ANIC); 2 Cf, same data (BMNH); 1 $, Mt Webb Nat. Pk, 1504'S, 14507'E, iv.1981(Naumann) (ANIC). LISSOPIMPLA Kriechbaumer Lissopimpla Kriechbaumer, 1889: 309. Type-species: Lissopimpla 8-guttata Kriechbaumer (= Pimpla excelsa Costa), by subsequent designation, Ashmead, 1900a: 55.Xenopimpla Cameron, 1898: 28. Type-species: Rhyssa semipunctata Kirby (= Pimpla excelsa Costa), by monotypy.Trichrus Tosquinet, 1903: 373. Type-species: Trichrus stupenda Tosquinet (= Pimpla basalis Snellen van Vollenhoven), by monotypy.Notiopimpla Vachal, 1907: 118. Type-species: Notiopimpla priocnemidea Vachal (= Pimpla excelsa Costa), by subsequent designation, Viereck, 1914: 101. Medium to large-sized species, fore wing length 7-18 mm; clypeus divided into basal and apical parts by atransverse suture, its margin convex; mandibles strongly narrowed, twisted; occipital carina complete.Epicnemial carina present; mesopleuron with impressed trans-striate grooves; mesopleural suture angledslightly centrally; propodeum with traces of carinae, usually with apophyses and a central low crest,spiracle elliptical. Tarsal claws of female large, without basal lobes, with spatulate bristle; hind femur witha ventral tooth. Fore wing with 3r-m present, enclosing a large rhombic areolet; hind wing with firstabscissa of Cu\ very short or obliterated. Tergites of gaster polished, almost smooth; ovipositor projectingbeyond apex of gaster by 0-75-1-70 times length of hind tibia, slightly decurved. REMARKS. A small genus occurring in the Indo-Australian region with most species in Australiaand nearby south Pacific Islands. They are morphologically very alike and differ most con-spicuously in colour pattern. The genus is widely distributed throughout Australia, and one species, L. excelsa, is verycommon. It can frequently be observed in gardens probing with its ovipositor in grass tussocks.The hosts of Lissopimpla species are the pupae and prepupae of various Lepidoptera, especiallyNoctuidae. Key to Australian species of Lissopimpla 1 Median transverse groove of mesopleuron impressed to mesopleural furrow, trans-striate for its entire length (Fig. 17); gaster tricoloured, tergites 1-4 black with lateral yellow spots, tergites5 reddish; $ with ovipositor projecting beyond apex of gaster by 1-4-1-7 times length of hindtibia ;cf with median yellow triangle on frons below ocellus excelsa (Costa) (p. 274) - Median transverse groove of mesopleuron not extending to mesopleural furrow (Fig. 16), or if reaching the furrow then posteriorly weak, not trans-striate; gaster bicoloured or rarelyunicolorous; $ with ovipositor projecting beyond apex of gaster by 1-0 or less times length ofhind tibia; cf with frons centrally black 2 All coxae black; face of $ entirely black, that of cf pale yellow with a central black mark; fore wing with cu-a proximal to base of Rs&M by about 0-5 times its length afra Girault(p. 274) - All coxae reddish brown, rarely the hind ones infuscate basally; face of $ marked with red or pale yellow, that of cf entirely pale yellow; fore wing with cu-a proximal to base of Rs&M by0-1-0-2 times its own length 3 274 I. D. GAULD 3 Subalar prominence concolorous with mesopleuron; gaster unicolorous brownish red, with- out white marks; ovipositor projecting beyond apex of gaster by 1-0 times length of hind tibia species 1 (p . 276) Subalar prominence conspicuously white-marked; gaster with more or less distinct pale markslaterally on tergites 1-4 at least ; ovipositor less than 1 times length of hind tibia 4 4 Gaster of $ rather stout, tergite 3 about 2-1 times as broad as long; ovipositor projecting beyond apex of gaster by less than 0-8 times length of hind tibia; posterior transverse carina of propodeum indistinct obcsa sp. n. (p. 275) Gaster of $ slender, tergite 3 1-7-1-8 times as broad as long; ovipositor projecting beyond apexof gaster by 0-85-0-95 times length of hind tibia; posterior transverse carina of propodeumsharp, usually complete between apophyses scutate Krieger (p. 276) Lissopimpla atra Girault Lissopimpla atra Girault, 1924a: 1; Townes, 1971a: 464. Holotype $, AUSTRALIA (QM) [examined].Lissopimpla atra Girault, 19246: 1. [Reprint of original description.] Medium-sized species, fore wing length 8-10 mm. Flagellum with a distinct white mark centrally.Mesopleuron with transverse median groove strongly impressed to near centre; metapleuron posterodor-sally evenly convex. Propodeum dorsally coarsely transversely wrinkled, laterally coriaceous anterior tospiracle; lateral longitudinal and posterior transverse carinae virtually complete, central area raised,elongately rectangular. Forewing with cu-a proximal to base of Rs&M by 0-4-0-5 times its own length.Gaster of female with tergite 31-6 times as broad posteriorly as long. Ovipositor projecting beyond apex ofgaster by 1-0 times length of hind tibia. Female black, only femora reddish. Wings quite strongly infumate. Male similar to female but withlower face except centrally, orbits, prominences on alitrunk and sometimes by corner of tergites 1-3 ofgaster, white-marked. REMARKS. A rather uncommon species most easily recognized by the entirely black coxae.HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), Queensland, Brisbane, iv.1914 (Hacker) (QM).Australia: 1 $, 7 cf , Queensland, Bundaberg, various dates (Fraucd) (ANIC). Lissopimpla excelsa (Costa)(Figs 3, 17) Pimpla excelsa Costa, 1864: 69. Holotype $, AUSTRALIA (depository unknown). Rhyssa semipunctata Kirby, 1883: 202. Holotype $, NEW ZEALAND (BMNH) [examined]. [Synonymized by Schulz, 1912:353.]Lissopimpla 10-notata Kriechbaumer, 1889: 310. Syntypes 9, c? AUSTRALIA (Radoszkowsky Coll., ? in Jagielloriski University, Cracow, Poland). [Synonymized by Schulz, 1912: 353.]Lissopimpla 8-guttata Kriechbaumer, 1889: 310. Holotype $, AUSTRALIA (Radoszkowsky Coll., ? in Jagiellonski University, Cracow, Poland). [Synonymized by Schulz, 1912: 353.]Xenopimpla semipunctata (Kirby) Cameron, 1898: 26.Theronia rufipes Tryon, 1900: 140. Syntypes 9> Cf , AUSTRALIA (depository unknown). [Synonymized by Morley, 1914: 46.] Lissopimpla semipunctata (Kirby) Dalla Torre, 1901: 406.Notiopimpla priocnemidea Vachal, 1907: 119. Lectotype $, AUSTRALIA (MNHN), designated by Townes etal, 1961: 48 [examined]. [Synonymized by Townes etal, 1961: 48.]Lissopimpla excelsa (Costa) Schulz, 1912: 353. Medium-sized to very large species, fore wing length of female 9-18 mm, of male 7-12 mm. Flagellumwithout a distinct white band. Mesopleuron with transverse median groove strongly impressed, reaching tomesopleural furrow; metapleuron with convexity at posterodorsal end (Fig. 17). Propodeum dorsallyfinely trans-striate, laterally finely striate anterior to spiracle, lateral longitudinal and posterior transversecarinae obsolescent, central raised area cornute, somewhat flattened anteriorly to form a narrow invertedisosceles triangle. Fore wing with cu-a proximal to Rs&M by 0-2-0-3 times its own length. Gaster of femalewith tergite 3 1-6-1-7 times as broad posteriorly as long. Ovipositor long, projecting beyond apex of gasterby 1-4-1-7 times length of hind tibia. ICHNEUMONIDAE OF AUSTRALIA 275 Female brownish orange, yellow-marked on orbits, notaular crests, laterally before scutellum and oftenon margin of postscutellum and metapleuron. Gaster black, tergites 1-4 with paired pale yellow spotsposteriorly, tergites 5+ orange. Ovipositor sheath black. Flagellum and hind tibia and tarsus infuscate.Wings proximally very strongly infumate, bluish black, distally less strongly coloured. Male similar tofemale but more extensively yellow-marked on subalar prominence, mesepimeron, scutellum and onanterior lateral margin of tergite 1. Frons black with triangular yellow mark below ocellus. Wings lessstrongly infumate. VARIATION. The most obvious variation is in size though morphologically this is a rather uniformspecies. The smaller males generally have tergites 1-4 paler than large specimens. Some aspectsof variation are discussed by Parrott (1952). REMARKS. This is one of the commonest and most conspicuous of Australian ichneumonids. It isoften found in gardens and females can sometimes be observed probing with their ovipositors ingrass tussocks. L. excelsa is commonly reared as a parasite of a variety of Lepidoptera,particularly noctuids, and accounts of its biology are given by Tryon (1900) and Smith &Caldwell (1947). Males of L. excelsa are known to pollinate the orchid Cryptostylis leptochila byattempting to mate with the flowers (Coleman, 1928) and occasional specimens may be takenwith the orchid pollinia attached to the tip of the gaster. L. excelsa is, on account of its colour pattern, the most distinctive species in the genus. It iswidely distributed throughout Australia though it seems to be relatively uncommon in thewestern part of Tasmania. The species also occurs in New Zealand, Fiji and the Kermadec Is.(Morley, 19130; Parrott, 1952). HOST RECORDS. Anthelidae: Anthela denticulata (Newman) (Cameron, 1912). Noctuidae:Achaea Janata (L.) (Morley, 19130); Mythimna convecta (Walker) (Chadwick & Nikitin, 1976);M. separata (Walker) (Tryon, 1900); Spodoptera exempta (Walker) (ANIC); 5. mauritia(Boisduval) (Tryon, 1900); Tiracola plagiata Walker (Temperley, 1930). Pyralidae: Cactoblastissp. (Chadwick & Nikitin, 1976).A number of other host records from New Zealand are given by Parrott (1952). MATERIAL EXAMINED New Zealand: 1 $ (holotype of semipunctata (BMNH). Australia: 1 $ (lectotype of priocnemidea)Melbourne (MNHN). Australia: 197 $ , 164 cf , from all states (Map 7) (AM, ANIC, BMNH, DAH, DAR, NMV, QM). Lissopimpla obesa sp. n. Medium-sized species, fore wing length 9 mm. Flagellum with white ring from segments 7-14. Mandiblesvery strongly tapered, lower tooth minute; malar space 1-5 times length of basal mandibular width.Mesopleuron with transverse median groove rather broad and shallow, reaching only to centre; meta-pleuron postero-dorsally entirely convex. Propodeum dorsally coarsely trans-striate, laterally with areaanterior to spiracle coriaceous, lateral longitudinal carinae complete, posterior transverse carina weak;central area slightly raised, rectangular. Fore wing with cu-a proximal to Rs&M by about 0-1 times its ownlength. Gaster rather stout, tergite 3 about 2-1 times as broad posteriorly as long. Ovipositor projectingbeyond apex of gaster by 0-75 times length of hind tibia. Orange, marked with yellowish white on orbits, margin of pronotum, subalar prominence, mesoscutumanteriorly, centrally and before scutellum, spots on mesopleuron, postscutellum, propodeal apophysesand laterally on tergites 1-5. Flagellum infuscate except for band. Wings hyaline. REMARKS. This species is distinctive on account of its relatively stout gaster, short ovipositor andindistinct propodeal transverse carina. It appears to be closely related to L. scutata. HOST RECORDS. None. MATERIAL EXAMINED Holotype $ , Australia: Queensland, Murchie's Scrub, Watalgan Forest, 9 km off Rosedale Rd, xii.1973(Franco) (ANIC). 276 I. D. GAULD Lissopimpla scutata Krieger(Fig. 16) Lissopimpla scutata Krieger, 1899: 50. Lectotype 9, AUSTRALIA (MNHU), designated by Townes et al.,1961: 49 [examined]. Medium-sized to moderately large species, fore wing length of female 7-11 mm, of male 7-9 mm.Flagellum with a distinct white mark centrally. Mesopleuron with transverse median groove impressed tocentre; metapleuron posterodorsally evenly convex (Fig. 16). Propodeum dorsally transversely coarselystriate, laterally coriaceous or somewhat striate anterior to spiracle, lateral longitudinal and posteriortransverse carina sharp and virtually complete, central part raised into a tubercle, not obviously extendedanteriorly. Fore wing with cu-a proximal to base of Rs&M by 0-2 times its own length. Gaster of femalewith tergite 3 1-7-1-8 times as broad posteriorly as long. Ovipositor projecting beyond apex of gaster by0-85-0-95 times length of hind tibia. Female black with profuse pale marks on orbits, notaular crests, mesoscutum centrally and laterallybefore scutellum, subalar prominences and mesopleuron between grooves, scutellum, postscutellum,propodeal apophyses and hind corners of tergites 1-5; legs reddish, hind tibia and tarsus infuscate. Wingweakly infumate. Male similar to female but with lower face entirely pale. VARIATION. There are a number of small specimens, in ANIC and BMNH, which have the forewing length 7-8 mm, and the gaster and the lower part of the alitrunk predominantlyreddish-coloured. These individuals may represent a distinct species but no characters otherthan colour could be found to separate them from smaller specimens of typical scutata. REMARKS. L. scutata is another relatively common species though its range is mostly morenorthern than that of L. excelsa. It is easily recognized by its characteristic colour pattern whichresembles that of Echthromorpha intricatoria, although the two species are allopatric. HOST RECORDS. None. Material examined Australia: 1 9 (lectotype), Queensland, Cooktown (MNHU).Australia: 27 $, 19 cf , New South Wales, Queensland (Map 8) (AM, ANIC, BMNH, QM, QUM). Lissopimpla species 1 Medium-sized species, fore wing length 9 mm. Flagellum with white central band. Mesopleuron withmedian transverse furrow, shallow, reaching only to centre; metapleuron posterodorsally evenly convex.Propodeum very like that of L. scutata. Fore wing with cu-a proximal to base of Rs&M by 0-2 times its ownlength. Gaster with tergite 31-9 times as broad posteriorly as long. Ovipositor projecting beyond apex ofgaster by 1 -0 times length of hind tibia. Reddish brown, only pale-marked on notaular crest and mesepirneron. Ovipositor sheath black. Wingsinfumate. REMARKS. This species is very similar to L. scutata except that it is strikingly different in colour. Itis also similar to the New Caledonian species, L. pacifica, from which it principally differs in size. HOST RECORDS. None. MATERIAL EXAMINEDAustralia: 1 $, Queensland, Mt Tambourine, x.1978 (Galloway) (BMNH). XANTHOPIMPLA Saussure Xanthopimpla Saussure, 1892: 13. Type-species: Xanthopimpla hova Saussure, by subsequent designa-tion, Ashmead, 1900a: 56. Chloropimpla Saussure, 1892: 13. Type-species: Chloropimpla dorsigera Saussure, by monotypy. Notopimpla Krieger, 1899: 106. Type-species: Plmpla terminate Brulle, by monotypy. Neopimploides Viereck, 1912: 151. Type-species: Neopimploides syleptae Viereck (= Ichneumon punc-tatus F.), by original designation. Austrapophua Girault, 1926: 135. Type-species: Austrapophua xanthopimploides Girault (= Xanthopim-pla rhopaloceros Krieger), by subsequent designation, Walkley, 1963: 116. ICHNEUMONIDAE OF AUSTRALIA 277 Small to moderately large-sized species, fore wing length 4-13 mm; clypeus transversely divided, marginstraight; mandibles strongly narrowed and twisted about 90; occipital carina complete. Epicnemial carinapresent; mesopleural suture generally not angled centrally; propodeum usually with strong carinae,spiracles oval to elliptical. Tarsal claws of female without basal lobes, usually large with spatulate bristles.Fore wing with 3r-m present or absent; hind wing with first abscissa of Cw t about 0-3 or less times as long ascu-a (Fig. 7). Tergite 1 moderately slender; tergites 2-5 polished, usually rather sparsely punctate;ovipositor from barely projecting beyond apex of gaster to projecting by 1-5 times lengh of hindtibia. REMARKS. Xanthopimpla is a very large tropicopolitan genus, most species of which occur in theIndo-Papuan area. Townes & Chiu (1970) revised the genus and erected 19 species-groups.These groups are difficult to recognize unless one has access to an extensive collection, but theyhave held up well for the considerable additional material collected recently. Most Australianspecies belong to the rhopaloceros-group or the splendens-group, two species complexes whichhave undergone considerable radiation in New Guinea. Xanthopimpla species are conspicuous bright yellow ichneumonids. Most are profuselyblack-speckled and early authors (e.g., Krieger, 1899) placed considerable emphasis on thepatterns of black spots on the gastral tergites. However, not only are these patterns generallysexually dimorphic, but there is individual variation within a sex. Frequently spots may bemissing in smaller individuals, though some species can be characterized by apparently neverhaving spots on certain tergites. The majority of species of Xanthopimpla are restricted toforests at various altitudes, from coastal mangrove swamp to montane moss-forest. A fewspecies, such as X. flavolineata, are associated with disturbed areas and are not uncommon inagricultural ecosystems where they parasitize a variety of lepidopterous pests. Gauld (1984) recorded 15 named and two undescribed species from Australia. A furtherundescribed species has been seen and these three taxa are described below. The majority ofspecies are found in the tropical parts of Australia. Key to Australian species of Xanthopimpla 1 Fore wing with 3r-m absent (Fig. 42) 2 Fore wing with 3r-m present, enclosing a rhombic areolet (Figs 39-41) 6 2 Area superomedia complete laterally and posteriorly (Fig. 33); pleural carina of propodeum extending only as far forward as level of spiracle (Fig. 21) ; scutellum in profile pyramidal A/rsute(Girault)(p. 304) - Area superomedia not delineated (Figs 28, 30, 35); pleural carina of propodeum complete, extending to front edge of propodeum (Fig. 22); scutellum in profile weakly to stronglyconvexly rounded 3 3 Anterior transverse carina of propodeum present laterally (Fig. 28); mid coxa with a blunt prominence on anterolateral side rhopaloceros Krieger (p. 307) - Anterior transverse carina of propodeum entirely absent (Figs 30, 35); mid coxa without a blunt prominence 4 4 Propodeum without a trace of tubercle above and just behind spiracle (Fig. 30); posterior transverse carina of mesosternum centrally broadened and with a deep median V-shapednotch; punctures of tergite 3 centrally very fine summervillei (Girault) (p. 309) - Propodeum with a tubercle above and just behind spiracle (Fig. 35) ; posterior transverse carina of mesosternum broadened lateral to centre, medially with wide U-shaped notch; punctures of tergite 3 centrally coarse 5 5 Submetapleural carina distinct; propodeal tubercle very strong (Fig. 22); ovipositor long quadridens Townes & Chiu (p. 306) - Submetapleural carina absent (Fig. 23) ; propodeal tubercle weak ; ovipositor very short binodus Townes & Chiu (p. 301) 6 Lateral longitudinal carinae of scutellum present only at extreme anterior end, the scutellum flat; propodeum with lateromedian longitudinal and anterior transverse carinae virtuallyabsent (Fig. 36); distal flagellar segment truncate with a distinct elliptical flat area on apex(Fig. 47) terminate (Brulle) (p. 309) - Lateral longitudinal carinae of scutellum distinct for more than 0-5 times scutellar length, the scutellum convex to pyramidal; propodeum with part of lateromedian or anterior transverse 278 I. D. GAULD carinae or both discernible (Figs 37, 38); distal flagellar segment rounded or somewhattruncate, but if truncate sharply ended with a flat area 7 7 Fore wing with 2m-cu joining M very close to outer corner of areolet (Fig. 40) ; lower anterior corner of pronotum sharply angled about 95 and $ with ovipositor projecting 0-9 times length of hind tibia hiatus Townes & Chiu (p. 304) Fore wing with 2m-cu joining M almost equidistant between 2r-m and 3r-ra (Fig. 39) or closerto 2r-m; lower anterior margin of pronotum rounded off, or if rarely rather sharply angledthen with ovipositor less than 6 times as long as hind tibia , ovipositor otherwise various 8 8 Notauli very strongly impressed, reaching to or behind the level of the hind edge of tegulae; mesopleuron with a strongly impressed sternaulus (Fig. 18) 9 Notauli weak to strongly impressed but short, at most reaching to level of centre of tegulae;mesopleuron with sternaulus weak or indistinct (Figs 19, 20) 13 9 Lower anterior part of metapleuron diagonally striate (Fig. 18); scutellum with lateral carinae reaching about 0-8 times scutellar length 10 Lower anterior part of metapleuron without diagonal striae; scutellum with lateral carinaereaching to apex 11 10 Clypeus with a median swelling near upper margin which is most obvious in the $ ; propodeum with anterior transverse carina present laterally (Fig. 26); tergite 1 of gaster usually with a pair of black marks; hind tibia with 5-9 preapical bristles striate Townes & Chiu (p. 308) Clypeus medially almost flat; propodeum with anterior transverse carina vestigial or absentlaterally (Fig. 31); tergite 1 of gaster entirely yellow; hind tibia with 3-4 preapical bristles a man sp. n. (p. 279) 11 Hind tibia with 5-7 preapical bristles (Fig. 44); area superomedia not defined laterally, confluent with areae lateralis (Fig. 34) harak sp. n. (p. 300) Hind tibia with 0-3 preapical bristles (Fig. 43); area superomedia laterally defined bylateromedian longitudinal carina , separated from areae lateralis (Figs 25 , 38) 12 12 Wings strongly and uniformly infumate; ovipositor sheath barely protruding beyond apex of gaster ; mesopleuron ventrally closely and coarsely punctate ecaudate Krieger (p . 302) Wings hyaline with apices slightly infumate; ovipositor sheath protruding beyond apex ofgaster by 0-5 times length of hind tibia; mesopleuron ventrally very finely and sparselypunctate puhidorsis Townes & Chiu (p . 306) 13 Propodeum with area superomedia large, elongate (though sometimes absent laterally), more than 1-1 times as long as wide and with section of lateromedian longitudinal carina betweenanterior and posterior transcarinae (or if absent the distance) more than 0-6 times as long aspart of lateromedian longitudinal carina anterior to anterior transcarina (Fig. 29) ; tergite 1 ofgaster slender, at least 1-3 times as long as posteriorly broad; Rs in fore wing strongly sinuate (Fig. 39) 14 Propodeum with area superomedia various, either small and subquadrate or larger andtransverse, usually with section of lateromedian longitudinal carina between anterior andposterior transcarinae very short (Figs 24, 32, 37); tergite 1 of gaster subquadrate, less than1 -2 times as long as posteriorly broad ; Rs in fore wing weakly sinuate (Fig. 41) 15 14 Gaster more or less entirely yellow; tergite 1 with carina from spiracle to anterior margin of segment (Fig. 49) ; notaulus weak, not reaching to centre of tegulae . . . ffavolineata Cameron (p. 302)Gaster with black spots at least on tergites 3 , 4 and 5 , sometimes on all tergites (Fig. 50) ; tergite1 without carina to spiracle (Fig. 48); notauli strongly impressed, reaching to level of centreof tegulae area/is Krieger (p. 299) 15 Scutellum in profile pyramidal (Fig. 19); propodeum with a swelling anterior to spiracle; mesopleuron strongly swollen centrally; area superomedia very transverse, about 2-7 times as broad as long, with carinae very strong (Fig. 24) fraterculus Townes & Chiu (p. 303) Scutellum in profile convex (Fig. 20); propodeum without a conspicuous swelling anterior tospiracle; mesopleuron weakly swollen centrally; area superomedia various, if transversethen usually less than 2 5 times as broad as long and often with some carinae weak 16 16 Posterior transverse carina of mesosternum broadened into rounded lobes, with a wide central notch (Fig. 45); ovipositor very long, projecting beyond apex of gaster by more than 1-0 times length of hind tibia (Fig. 51) ; proximal 0- 1 of hind tibia black aust rails Krieger (p. 300) Posterior transverse carina of mesosternum centrally produced into a pair of high posteriorlydirected keels with a deep V-shaped central notch (Fig. 46); ovipositor shorter, projectingbeyond apex of gaster by no more than 0-8 times length of hind tibia; hind tibia entirelyyellow 17 ICHNEUMONIDAE OF AUSTRALIA 279 17 Area superomedia small, slightly longer than broad (Fig. 32); tergite 3 of gaster highly polished, with sparse punctures ankhu sp. n. (p. 279) Area superomedia moderately large, transverse (Fig. 37); tergite 3 of gaster with moderatelydense punctures ochracea (Smith) (p. 305) Xanthopimpla amon sp. n. (Figs 18, 31) Clypeus very weakly convex, medially flat; face shallowly and rather sparsely punctate. Flagellum with 36segments, the distal one slightly flattened, apically rounded. Lower anterior corner of pronotum ratherabruptly rounded. Mesoscutum with notauli very strongly impressed, reaching far behind level of hindmargin of tegulae; central lobe of mesoscutum with fine hairs laterally, in the middle glabrous. Scutellumvery convex, carinate laterally to 0-7-0-8 of its length. Mesopleuron quite strongly swollen centrally with apronounced constriction just below the swelling, anteroventrally sparsely punctate; sternaulus deep,extending almost entire length of mesopleuron (Fig. 18); metapleuron with diagonal striae present in loweranterior corner, otherwise smooth; submetapleural carina strong. Posterior transverse carina of mesoster-num centrally broadened into two rounded lobes, with a median V-shaped cleft. Propodeum moderatelylong, posterior transverse carina complete, strong; lateromedian carinae present anteriorly; anteriortransverse carina either present but very weak, or absent; tubercle vestigial (Fig. 31). Mid coxaunspecialized; hind tibia with 3-4 preapical bristles, largest bristle on hind tarsal claws spatulate. Fore winglength 5-6 mm; 3r-m present enclosing a rather broad petiolate rhombic areolet; 2m-cu joining areoletslightly proximal to centre; Rs arcuate; cu-a opposite base ofRs&M. Gaster with tergite 1 about 1-2 timesas long as posteriorly broad, without a carina extending from near spiracle to anterior end of segment.Tergite 3 centrally almost impunctate. Ovipositor projecting beyond apex of gaster by 0-2-0-3 times lengthof hind tibia, its apex almost cylindrical. Female yellow, black-marked on interocellar area, mesoscutum in three stripes, propodeum in ananterior transverse band, tergites 3-5 in paired spots, tergites 7 and 8 with transverse band. Anterior end ofhind tibia infuscate; tergite 2 with weak lateral spots. Apex of wing infumate. Male similar to female. VARIATION. The paratype female has the mesoscutal stripes confluent to form a large black area,tergite 6 has paired black spots in the holotype but only weak marks in the paratypes. REMARKS. This small species belongs to the splendens-group of Townes & Chiu (1970). It is quitesimilar to X. striata in having a striate metapleuron, incompletely carinate scutellum, strongnotauli and a deep sternaulus. It differs from X. striata in being smaller and more delicate, in nothaving a centrally swollen clypeus, having a vestigial anterior propodeal transverse carina,having fewer tibial bristles, being more sparsely punctate, generally, having a more convexmesopleuron and in the position of black gastral marks. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Queensland, Mt Tambourine, x. (TC). Paratypes. Australia: 1 C?, Queensland, Brisbane, i-vi.1971 (Sedlacek) (TC); 1 $, 'North Queensland'(BMNH). Xanthopimpla ankhu sp. n. (Figs 20, 32, 46) Clypeus very weakly convex; face strongly and coarsely punctate. Flagellum with 35 segments, the distalone slightly flattened, apically rounded. Lower anterior corner of pronotum fairly abruptly rounded.Mesoscutum with notauli moderately shallow, reaching about the level of fore edge of tegulae; central partof mesoscutum sparsely hirsute. Scutellum moderately convex, carinate laterally to apex. Mesopleuronweakly swollen centrally, anteroventrally quite coarsely punctate, sternaulus vestigial (Fig. 20); meta-pleuron smooth, submetapleural carinae present. Posterior transverse carina of mesosternum veryabruptly raised near centre, produced into two acute lobes separated by a narrow V-shaped cleft (Fig. 46).Propodeum moderately long, carinae strong with a small elongate area superomedia discernible that isabout 0-9 times as broad as long; tubercle weak (Fig. 32). Mid coxa unspecialized; hind tibia with 6-8preapical bristles, largest bristle on hind tarsal claw spatulate. Fore wing length 6-7 mm; 2>r-m present, 280 I. D. GAULD Fig. 1 Sericopimpla crenator, $ . ICHNEUMONIDAE OF AUSTRALIA 281 Fig. 2 Eriostethus maximus, $ . Fig. 3 Lissopimpla excelsa, 9 . 282 I. D. GAULD 15 Figs 4-15 4, 5, head, anterior view, of (4) Sericopimpla lutea; (5) Acropimpla xantha. 6-8, hind wing of(6) Echthromorpha intricatoria; (7) Xanthopimpla hiatus; (8) Sericopimpla lutea. 9-13, propodeum,lateral view, of (9) Camptotypeus sellatus; (10) Sericopimpla lutea; (11) Zaglyptus glabrinotum; (12)Echthromorpha intricatoria; (13) E. agrestoria. 14, 15, head, lateral view, of (14) E. nigricornis; (15) E.intricatoria. 20 Figs 16-21 16-20, alitrunk, lateral view, of (16) Lissopimpla scutata; (17) L. excelsa; (18) Xanthopimplaamon; (19) X. fraterculus; (20) X. ankhu. 21, propodeum, lateral view, X. hirsuta. 284 I. D. GAULD Figs 22-38 22, 23, propodeum, lateral view, of (22) Xanthopimpla quadridens; (23) X. binodus. 24-38,propodeum, dorsal view, of (24) X. fraterculus; (25) X. pubidorsis; (26) X. striata; (27) X. hiatus; (28) X.rhopaloceros; (29) X. arealis; (30) X. sutnmervillei; (31) X. amon; (32) X. ankhu; (33) X. hirsuta; (34) X.barak; (35) X. quadridens; (36) X. terminalis; (37) X. ochracea; (38) A', caudata. ICHNEUMONIDAE OF AUSTRALIA 285 Figs 39-49 39-42, fore wing of (39) Xanthopimpla flavolineata; (40) X. hiatus; (41) X. australis; (42) X.quadridens. 43, 44, hind tibia of (43) X. ecaudata; (44) X. barak. 45, 46, posterior transverse carina ofmesosternum of (45) X. australis; (46) X. ankhu. 47, tip of antenna, X. terminate. 48, 49, segment 1 ofgaster, lateral view, of (48) X. arealis; (49) X. flavolineata. I. D. GAULD Figs 50-64 50-53, gaster, dorsal view, of (50) Xanthopimpla arealis, $; (51) X. australis, $; (52) X.rhopaloceros , $; (53) X. summervillei, $. 54, 55, hind leg (54) Camptotypus lachesis; (55) C. bicolor.56-58, mesoscutum, dorsal view, of (56) Sericopimpla crenator; (57) 5. australis; (58) S. lutea. 59, 60,hind wing of (59) Zaglyptus glabrinotum ; (60) Z. hollowayi. 61 , 62, scutellar profile of (61) Sericopimplalutea; (62) S. crenator. 63, 64, tergite of gaster of (63) Camptotypus bicolor; (64) C. sellatus. ICHNEUMONIDAE OF AUSTRALIA 287 75 Figs 65-77 65, 66, propodeum and tergite 1, lateral view, of (65) Theronia fraucai; (66) T. steindachneri.67, propodeum, lateral view, T. maculosa. 68, 69, scutellum and propodeum, dorsal view, of (68) T.maculosa; (69) T. steindachneri. 70, 71, tip of flagellum, $ of (70) T. steindachneri; (71) T. fraucai. 72,73, clypeus and mandibles of (72) T. melanosoma; (73) T. maculosa. 74, 75, tip of ovipositor of (74) T.penetrans; (75) T. maculosa. 76, 77, head, dorsal view, of (76) T. fraucai; (77) T. penetrans. 288 I. D. GAULD 89 Figs 78-93 78, 79, hind wing of (78) Acrodactyla quadrisculpta; (79) A. cursor. 80, fore leg, A. micans. 81 ,82, head, dorsal view, of (81) A. zekhem; (82) A. micans. 83, 84, fore wing of (83) Eriostethus maximus;(84) E. pulcherrimus. 85, propodeum, lateral view, Acrodactyla micans. 86, 87, hind wing of (86)Zatypota rennefer; (87) Z. velata. 88-91, propodeum, lateral view, of (88) Z. dandiensis; (89) Z. bingili;(90) Z. kauros; (91) Z. phraxos. 92, 93, head, anterior view, of (92) Z. kauros; (93) Z. ce/er. ICHNEUMONIDAE OF AUSTRALIA 289 106 Figs 94-110 94-96, propodeum and anterior gastral tergites, dorsal view, of (94) Zatypota dandiensis;(95) Z. rennefer; (96) Z. phraxos. 97, larval remains of Eriostethus on spider. 98, 99, cocoons of (98)Eriostethus carinatus; (99) Zatypota dandiensis, 100, 101, segment 1 of gaster, lateral view, of (100)Yezoceryx coelyx; (101) Y. apicipennis. 102, 103, ovipositor tip of (102) Y. apicipennis; (103) Y. tantalyx.104, hind trochantellus, Y. tantalyx. 105, 106, submetapleural carina of (105) Y. amaryllyx; (106) Y.coelyx. 107, 108, scutellum and propodeum, dorsal view, of (107) Lycorina turneri; (108) L. splendidula.109, 110, $ claws of (109) L. turneri; (110) L. canberrae. 290 I. D. GAULD Map 1 Distribution of Sericopimpla australis. Map 2 Distribution of Sericopimpla crenator. ICHNEUMONIDAE OF AUSTRALIA 291 Map 3 Distribution of Camptotypus sellatus. Map 4 Distribution of Camptotypus lachesis. 292 I. D. GAULD Map 5 Distribution of Echthromorpha agrestoria. Map 6 Distribution of Echthromorpha intricatoria. ICHNEUMONIDAE OF AUSTRALIA 293 Map 7 Distribution of Lissopimpla excelsa. Map 8 Distribution of Lissopimpla scutata. 294 I. D. GAULD Map 9 Distribution of Xanthopimpla arealis. Map 10 Distribution of Xanthopimpla australis. ICHNEUMONIDAE OF AUSTRALIA 295 Map 11 Distribution of Xanthopimpla ecaudata. Map 12 Distribution of Xanthopimpla ochracea. 296 I. D. GAULD Map 13 Distribution of Xanthopimpla flavolineata. Map 14 Distribution of Xanthopimpla rhopaloceros. ICHNEUMONIDAE OF AUSTRALIA 297 Map 15 Distribution of Xanthopimpla summervillei. Map 16 Distribution of Xanthopimpla terminalis. 298 I. D. GAULD Map 17 Distribution of Theronia steindachneri. Map 18 Distribution of Theronia maculosa. ICHNEUMONIDAE OF AUSTRALIA 299 enclosing a moderately small areolet; 2m-cu joining areolet near centre; Rs slightly sinuate; cu-a oppositebase of Rs&M. Gaster with tergite 1 about as long as posteriorly wide, without a carina joining spiracle toanterior margin of tergite. Tergite 3 with large shallow, rather sparse punctures. Ovipositor projectingbeyond apex of gaster by 0-6-0-7 times length of hind tibia, slightly decurved. Female yellow. Scape, most of flagellum, interocellar area and ovipositor sheath, black; mesoscutum,propodeum and tergites 3-5, 7 and 8 with brownish spots. Wings hyaline. Male like female except thattergites 7 and 8 are black-marked. REMARKS. X. ankhu belongs to the ochracea-group. It can be distinguished from ochracea (theonly other Australian species in this group) by the shape of the area superomedia. The gaster isalso far less strongly sculptured. HOST RECORDS. None. MATERIAL EXAMINED Holotype $ , Australia: Queensland, Bundaberg Ck, in mangroves, x.1977 (Franca) (ANIC).Paratypes. Australia: 1 $, 2 d", same data as holotype; 1 9, same locality, iii.1976 (Frauca) (ANIC). Xanthopimpla arealis Krieger(Figs 29, 48, 50) Xanthopimpla gracilis Krieger, 1899: 74. Holotype d", PAPUA NEW GUINEA (MNHU). [Synonymized by Townes & Chiu, 1970: 70.]Xanthopimpla arealis Krieger, 1899: 93; Townes & Chiu, 1970: 70. Holotype $, AUSTRALIA (MNHU) [examined].Xanthopimpla beauforti Cameron, 1907: 45. Lectotype cf, IRIAN JAVA (BMNH), designated by Townes et al, 1961: 51 [examined]. [Synonymized by Townes & Chiu, 1970: 70.]Xanthopimpla papuana Cameron, 1907: 46. Holotype $, IRIAN JAVA (ZMA). [Synonymized by Townes et al, 1961: 51.] Clypeus flat; face finely punctate. Flagellum with 37-39 segments, the distal one slightly flattened,truncate. Lower anterior corner of pronotum obtusely rounded. Mesoscutum with notauli very stronglyimpressed but reaching only to about level of centre of tegulae; central part of mesoscutum sparselyhirsute. Scutellum moderately convex, carinate to posterior margin. Mesopleuron weakly swollencentrally, anteroventrally quite finely punctate, sternaulus impressed; metapleuron smooth, submetapleu-ral carina complete. Posterior transverse carina of mesosternum strongly raised, with a deep V-shapednotch centrally. Propodeum long with posterior transverse carina complete, anterior one absent centrally,laterally strong and angled at junction with anterior part of lateromedian carinae; lateromedian longitudin-al carinae vestigial between transverse carinae but area superomedia discernible, 1-1-1-2 times as long asbroad; tubercle weak (Fig. 29). Mid coxa unspecialized; hind tibia with 4-7 preapical bristles, largest bristleon hind tarsal claw spatulate. Fore wing length 8-10 mm; 3r-m present, enclosing a broad rhombic areolet;2m-cu joining areolet in centre; Rs strongly sinuous; cu-a opposite base of Rs&M. Gaster with tergite 11-3-1-5 times as long as posteriorly broad, without a carina extending from near spiracle to anterior end(Fig. 48). Tergite 3 closely and coarsely punctate. Ovipositor projecting beyond apex of gaster by 0-4-0-5times length of hind tibia. Female yellow with black marks on interocellar area, spots on lobes of mesoscutum and anterolateralareas of propodeum and tergites 3-5 of gaster (Fig. 50). Wings hyaline. Male similar but generally withblack marks on all tergites of gaster except rarely the first. VARIATION. There is a considerable range of variation in the spots on the gaster of the female. Apair of spots are present on at least tergites 3-5, sometimes 2-6. Tergite 7 generally has a blackbar whilst tergite 1 may occasionally have one so that at its most extreme all but the last twotergites will be black-marked too. REMARKS. This Australo-Papuan species was placed by Townes & Chiu (1970) in the splendens-group. Unlike other Australian representatives of this complex it has rather short notauli thatreach only to about the level of the centre of the tegulae and has less pronounced scutellarcarinae and a weaker sternaulus. X. arealis resembles X. flavolineata in being slender, having alarge elongate area superomedia and a sinuous vein Rs. Whether or not X. arealis should beplaced in the citrina-group in preference to the splendens-group is likely to be debated for sometime. 300 I. D. GAULD HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype of arealis), New South Wales (MNHU). Irian Jaya: 1 C? (lectotype ofbeauforti), Etna Bay (BMNH). Australia: 34 $, 17 cf , Queensland, New South Wales (Map 9). Papua New Guinea: 2 $, 1 cT. (ANIC;BMNH.) The single record for this species from Victoria (Townes & Chiu, 1970) is based on a correctlyidentified specimen but possibly it is erroneously labelled as recent collecting has not yielded specimensfrom south of about 35. Xanthopimpla australis Krieger (Figs 41, 45) Xanthopimpla australis Krieger, 1899: 102; Townes & Chiu, 1970: 215. Lectotype $, AUSTRALIA (MNHU), designated by Townes etal., 1961: 51 [examined].Xanthopimpla similis Krieger, 1899: 103. Holotype 9, AUSTRALIA (MNHU). [Synonymized by Krieger, 1914: 133.] Clypeus flat; face weakly but closely punctate. Flagellum with 37-39 segments, the apical one roundeddistally. Lower anterior corner of pronotum rounded. Mesoscutum with notauli quite strongly impressedbut short, not reaching to level of anterior edge of tegulae; central part of mesoscutum with fine sparsehairs. Scutellum moderately convex, carinate laterally to hind margin. Mesopleuron barely swollencentrally, antero-ventrally finely and sparsely punctate; metapleuron smooth, submetapleural carinastrong. Posterior transverse carina of mesosternum broadened centrally into rounded lobes with a widemedian notch (Fig. 45). Propodeum rather short, with posterior transverse carina entire and anteriortransverse carina complete laterally; area superomedia laterally defined, short and strongly transverse,about 0-4 times as long as broad; tubercle vestigial. Mid coxa unspeciali/ed; hind tibia with 6-8 subapicalbristles, largest bristle on hind tarsal claw not explanate. Fore wing length 5-12 mm; 3r-m present,enclosing a shortly petiolate, broadly rhombic areolet; 2m-cu joining slightly distal to centre; Rs sinuous(Fig. 41); cu-a opposite base otRs&M. Gaster with tergite 11-1 times as long as posteriorly broad, withouta carina extending from spiracle to anterior margin; tergite 3 with strong, moderately dense punctation.Ovipositor projecting beyond apex of gaster by 1-2-1-3 times length of hind tibia, its apex slightlydecurved, slightly flattened. Female yellow, interocellar area, transverse band on mesoscutum, proximal 0-1 of hind tibia, centralspot on tergite 1 and lateral spots on tergites 3 and 7, black. Wings hyaline. Male similar to female but withtergites 1 and 3-7 black-marked. VARIATION. Some larger females have tergite 5 black-marked and rarely, very small specimenslack the black mark on tergite 1. REMARKS. This species belongs to the punctata-group. In Australia X. australis is easilyrecognized by the long ovipositor and rather characteristic colour pattern (Fig. 51). Townes &Chiu (1970) recognize three subspecies of which only the nominate one occurs in Australia. Theother two are Melanesian and though I do not consider these warrant distinction, I have avoidedlisting them as synonyms pending further study. HOST RECORDS. Gelechiidae: Pectinophora scutigera (Holdaway) (DPIQ). MATERIAL EXAMINED Australia: 1 $ (lectotype of australis) , Queensland, Cooktown (MNHU). Australia: 44 <j>, 28 cf, Northern Territory, Queensland, Western Australia (Map 10) (AM; ANIC;BMNH;TC). Xanthopimpla baraksp. n. (Figs 34, 44) Clypeus flat; face finely, sparsely punctate. Flagellum with 35-36 segments, the apical one roundedterminally but flattened laterally. Lower anterior corner of pronotum rounded. Mesoscutum with notaulivery strongly and deeply impressed, reaching behind level of hind margin of tegulae; mesoscutum centrallypunctate, with dark sparse pubescence. Scutellum strongly convex, with lateral carinae reaching to hindmargin. Mesopleuron moderately swollen centrally, anteroventrally finely but quite closely punctate; ICHNEUMONIDAE OF AUSTRALIA 301 sternaulus distinct; metapleuron smooth, submetapleural carina distinct. Posterior transverse carina ofmesosternum centrally broadened into rounded lobes, with an indistinct median notch. Propodeum rathershort, with anterior and posterior transverse carinae present, the lateromedian ones absent so areasuperomedia is confluent with areae lateralis; tubercle weak (Fig. 34). Mid coxa unspecialized; hind tibiawith 5-7 preapical bristles (Fig. 44), largest bristle on hind tarsal claw spatulate. Fore wing length 6-7 mm;3r-m present, enclosing a shortly petiolate transverse areolet; 2m-cu joining areolet distal to centre; Rssinuous; cu-a opposite base of Rs&M. Gaster with tergite 1 about 1.1 times as long as broad posteriorly;spiracle not joined to anterior margin by carina; tergite 3 with close coarse punctures. Ovipositorprojecting beyond apex of gaster by 6.0 times length of hind tibia, its apex slightly depressed and truncatedso in dorsal view it is blunt. Female yellow, only interocellar area and ovipositor sheath black; antenna infuscate. Wings hyaline,distally infumate. Male similar to female. REMARKS. X. barak belongs to the splendens-group. It can easily be separated from the otherAustralian species of this group by the characters given in the key but especially by the morespinose hind tibia. HOST RECORDS. None. MATERIAL EXAMINED Holotype $ , Australia: Queensland, 14 km W. by N. of Hope Vale Mission (15' 16S, 144' 59E), v.1981(Naumann) (ANIC). Paratype. 1 cf , same data as holotype. Xanthopimpla binodus Townes & Chiu (Fig. 23)Xanthopimpla binodus binodus Townes & Chiu, 1970: 290. Holotype cf , NEW CALEDONIA (BPBM). Clypeus with a faint median dorsal swelling, otherwise flat; face strongly and closely punctate. Flagellumwith about 35 segments, with distal one rounded apically. Lower anterior margin of pronotum fairly evenlyrounded. Mesoscutum with notauli deep, reaching to about level of centres of tegulae, mesosternumcentrally glabrous. Scutellum strongly convex, carinate laterally about 0.6 of its length. Mesopleuroncentrally weakly swollen, anteroventrally finely punctate; sternaulus vestigial; metapleuron smooth;submetapleural carina absent (Fig. 23). Posterior transverse carina of mesosternum weakly raised with abroad median U-shaped notch. Propodeum moderately long, only posterior transverse carina distinguish-able dorsally and this is obsolescent centrally; tubercle weak but clearly raised, punctate. Mid coxaunspecialized; hind tibia with 6-8 preapical bristles, largest bristle on hind tarsal claw not spatulate. Forewing length 5-8 mm; 3r-m absent; Rs fairly evenly arcuate; cu-a opposite base of Rs&M. Gaster withtergite 1 1-1-1-2 times as long as posteriorly broad, without a carina joining spiracle to anterior margin oftergite. Tergite 3 with close coarse punctures. Ovipositor projecting beyond apex of gaster by about 0-3times length of hind tibia, its apex subcylindrical. Female yellow, only interocellar area, mesoscutum centrally, scutoscutellar groove, paired spots ontergite 1 and proximal end of hind tibia, blackish; tergite 4 with pale brownish spots. Wings slightlyinfumate distally. Male similar to female but with black marks on at least tergites 3-6, often all tergites ofgaster. REMARKS. This species belongs to the rhopaloceros-group. It is the only Australian species with3r-m absent that has a coarsely punctate third tergite and lacks the submetapleural carina. InAustralia this species exhibits striking sexual dichromatism, the gaster of the female havingblack or brownish markings only on tergites 1 and 4 (or also on tergites 3 and 7 in NewCaledonian specimens), and on at least tergite 3-6 in the male. Townes & Chiu's (1970)recognition of two subspecies based on colour differences of the gaster is thus highly question-able as they overlooked sexual colour difference and have seen only one sex of the 'subspecies'from the Loyalty Is. HOST RECORDS. None. MATERIAL EXAMINED New Caledonia: 1 $, 1 Cf (paratypes), Puebo Coast, ix-x.1949 (Cheesman) (BMNH). Australia: 1 $, Queensland, Biggenden, Bluff Range, vii.1971 (Franco) (ANIC); 1 cf, CunninghamPass, iii (TC); 1 cT, Toowoomba, iii (TC). 302 I. D. GAULD Xanthopimpla ecaudata Krieger(Figs 38, 43) Xanthopimpla h ispida Krieger, 1899: 70. Holotype $, 'W. MALAYSIA' (MNHU). [Synonymized by Townes & Chiu, 1970: 93.]Xanthopimpla ecaudata Krieger, 1899: 71; Townes & Chiu, 1970: 93. Lectotype $, MOLUCCAS (MNHU), designated by Townes etal, 1961: 54 [examined].Xanthopimpla minor Krieger, 1899: 73. Holotype cf , PAPUA NEW GUINEA (MNHU). [Synonymized by Townes & Chiu, 1970: 93.] Clypeus rather flat; face quite coarsely punctate. Flagellum with 44-45 segments, the distal one flattened,terminally rounded. Lower anterior corner of pronotum obtusely rounded. Mesoscutum with notauli verystrongly impressed, reaching almost to hind margin, central part of mesoscutum with sparse darkpubescence. Scutellum very strongly convex, almost pyramidal, laterally carinate to hind margin.Mesopleuron weakly swollen centrally, anteroventrally coarsely and closely punctate, sternaulus mod-erately impressed; metapleuron smooth, punctate posteriorly, submetapleural carina complete. Posteriortransverse carina of mesosternum strongly raised, with a very narrow, V-shaped median notch. Pro-podeum moderately long, with posterior transverse carina complete, anterior one strong, absent centrally;lateromedian longitudinal carinae usually not present or weak between transverse carinae, rarely distinctin larger females; area superomedia elongate; tubercle weak (Fig. 38). Mid coxa unspecialized; hind tibiawith 0-3 preapical bristles (Fig. 43), largest bristle on hind tarsal claw spatulate. Fore wing length 6-7 mm;3r-m present, enclosing a rhombic areolet; 2m-cu joining areolet slightly distal to centre; Rs sinuate; cu-aopposite base of Rs&M. Gaster with tergite 11-1 times as long as broad posteriorly, with a strong carinaextending from near spiracle to anterior end of segment. Tergite 3 closely punctate. Ovipositor barelyprojecting beyond apex of gaster. Female yellow, only interocellar area and antenna black in Australian specimens. Wings stronglyinfumate. VARIATION. The few Australian specimens examined are quite alike in having dark wings, butelsewhere this species does not have such strong alar infumation. REMARKS. A widespread Melanesian-tropical Australian species occurring as far west as theMoluccas and east to the Solomons. X. ecaudata belongs to the splendens-group of Townes &Chiu (1970). It is a group easily recognized by the very long, deep notauli. In Australia X.ecaudata is easily recognized by its very short ovipositor. HOST RECORDS. None. MATERIAL EXAMINED Indonesia: 1 $ (lectotype of ecaudata), Moluccas, Kai (MNHU). Australia: 12 $ , 4 cf , Queensland. Indonesia: 9 $ , 7 cf , Moluccas; Irian Jaya. Papua New Guinea: 2 $ ,4 cf . Solomon Is: 4 $ (Map 11). Xanthopimpla ttavolineata Cameron(Figs 39, 49) Xanthopimplaflavolineata Cameron, 1907: 48; Townes & Chiu, 1970: 114. Holotype $ , IRIAN JAVA (ZMA) [examined.]Xanthopimpla emaculata Szepligeti, 1908: 256. Holotype $, JAVA (TM). [Synonymized by Townes & Chiu, 1970: 115.]Xanthopimpla immaculata Morley, 19136: 115. Holotype 9, INDIA (depository unknown, ?New Delhi) [paratype 9 examined.] [Synonymized by Cushman, 1925: 46.]Xanthopimpla hyaloptila Krieger, 1914: 35. Holotype cf , AUSTRALIA (MNHU). [Synonymized by Townes & Chiu, 1970:115.]Xanthopimpla xanthostigma Girault, 1925: 38. Holotype $, AUSTRALIA (QM) [examined]. [Synonymized by Townes & Chiu, 1970: 115.] Xanthopimpla xara Cheesman, 1936: 179. Holotype $, NEW HEBRIDES (BMNH) [examined]. [Synony-mized by Townes & Chiu, 1970: 115.]Metopius sesamiae Rao, 1953: 184. Holotype cf , INDIA (Dehra Dun). [Synonymized by Townes & Chiu, 1970: 115.] ICHNEUMONIDAE OF AUSTRALIA 303 Clypeus slightly convex near upper margin; face fairly evenly punctate. Flagellum with 36-40 segments, thedistal one flattened and terminally rounded. Lower anterior corner of pronotum rounded off about 90.Mesoscutum with notauli weak, barely reaching to level of anterior margin of tegulae; central part ofmesoscutum virtually glabrous. Scutellum moderately convex; carinate to posterior margin. Mesopleuronweakly convex centrally, anteroventrally finely and closely punctate, sternaulus indistinct; metapleuronsmooth, submetapleural carina complete. Posterior transverse carinae of mesosternum moderatelystrongly raised, with a median notch. Propodeum long, posterior transverse carina complete, anterior onepresent laterally, lateromedian carinae complete, defining a large elongate area superomedia that is1-1-1-2 times as long as broad; tubercle vestigial. Mid coxa unspecialized; hind tibia with 5-6 preapicalbristles, largest bristle on hind tarsal claw spatulate. Fore wing length 4-8 mm; 3r-m present, enclosing arhombic petiolate areolet; 2m-cu joining areolet near centre; Rs strongly sinuous (Fig. 39); cu-a oppositebase of Rs&M. Gaster with tergite 1 1-3-1-5 times as long as broad posteriorly, with a carina extendingfrom near spiracle to anterior end (Fig. 49). Tergite 3 moderately sparsely punctate. Ovipositor projectingbeyond apex of gaster by about 0-5 times length of hind tibia.Female yellow, interocellar area and ovipositor sheath black. Wings hyaline. Male similar to female. REMARKS. X. flavolineata is a widespread Indo-Australian species that frequents open areas ofcoarse grassland and cereal crops. It is placed by Townes & Chiu in the citrina-gToup, a mainlyAfrotropical complex. In Australia X. flavolineata may be recognized by its slender appearance,large elongate area superomedia and sinuous Rs. It is rather similar to X. arealis from which itmay be distinguished by characters given in the key. HOST RECORDS. In Asia this species is a common parasite of pyralids and noctuids that bore inrice, maize and sugar-cane (Townes & Chiu, 1970), though it has not been reared in Australia. MATERIAL EXAMINED Indonesia: 1 $ (holotype of flavolineata), Irian Jaya, Merauke (ZMA). India: 1 $ (paratype ofimmaculata), Bengal, Chapra (Mackenzie) (BMNH). Australia: 1 $ (holotype of xanthostigma) , Queens-land, Amamoor Forest (QM). New Hebrides: 1 $ (holotype of xara), Malekula (Cheesman) (BMNH). Australia: 33 $, 14 cf, New South Wales; Northern Territory; Queensland (Map 13) (AM; ANIC;BMNH;TC). Xanthopimpla fraterculus Townes & Chiu (Figs 19, 24)Xanthopimpla fraterculus Townes & Chiu, 1970: 33. Holotype cf , AUSTRALIA (TC) [examined]. Clypeus slightly convex near upper end; face strongly punctate. Flagellum with 37-41 segments, the distalone slightly flattened, apically rounded. Lower anterior corner of pronotum evenly rounded. Mesoscutumwith notauli deep but short, barely reaching to level of anterior margin of tegulae; central part ofmesoscutum sparsely hirsute. Scutellum pyramidal, carinate laterally to hind margin. Mesopleuronstrongly swollen just above centre, anteroventrally coarsely punctate; sternaulus vestigial (Fig. 19);metapleuron smooth, submetapleural carina strong. Posterior transverse carina of mesosternum stronglyraised, centrally with a broad V-shaped notch. Propodeum rather short with carinae strong, defining ashort transverse area superomedia that is about 2-7 times as broad as long; propodeum unusual in having astrong swelling in front of spiracle; tubercle vestigial (Fig. 24). Mid coxa unspecialized; hind tibia withoutpreapical bristles; largest bristle on hind tarsal claws spatulate. Fore wing length 9-11 mm; 3r-ra enclosinga fairly large rhombic areolet, 2m-cu joining it about in centre; Rs sinuate; cu-a opposite Rs&M. Gasterwith tergite 11-1-1 -2 times as long as broad posteriorly, usually with a carina extending forwards from nearanterior margin of spiracle. Tergite 3 sparsely but coarsely punctate. Ovipositor projecting beyond apex ofgaster by about 0-7-0-9 times length of hind tibia, noticeably decurved. Female yellow, upper part of scape, interocellar area, vertex, mesoscutum in 3 stripes and a transverseposterior bar and ovipositor sheath, black. Wings slightly infumate, apices more darkly so. Male similar tofemale though with at least tergite 3 bearing indistinct blackish spots. VARIATION. This is one of the most variable species in the extent of black marking. Some femaleshave paired black spots on tergites 2-4 and spots on the hind femora whilst a few have spots ontergites 2-6 and black bars on tergites 7 & 8 and paired spots on the hind femora. One specimeneven has indistinct spots on tergite 1. REMARKS. X. fraterculus belongs to the regina-group (Townes & Chiu, 1970) and is one of the 304 I. D. GAULD easiest Australian species to recognize on account of its pyramidal scutellum and the swelling onthe propodeum in front of the spiracle. The only other species with a relatively long ovipositorand strongly transverse area superomedia with which it could be confused is X. australis, whichhas a quite different colour pattern with a single transverse black stripe in the centre of themesoscutum rather than the four marks characteristic of X. fraterculus. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 C? (holotype), Western Australia, Fitzroy R. crossing, v.1969 (Alexander & Otte) (TC); 1 $(paratype), Northern Territory, Brock's Creek, iv.1931 (Campbell) (ANIC); 1 C? (paratype), Queens-land, Brisbane, xi.1918 (Hacker) (QM). Australia: 13 $, 3 C?, Northern Territory, Queensland, Western Australia. Xanthopimpla hiatus Townes & Chiu (Figs 27, 40)Xanthopimpla hiatus Townes & Chiu, 1970: 256. Holotype $ , AUSTRALIA (QM) [examined]. Clypeus fairly flat; face strongly punctate. Flagellum with 38-39 flagellar segments, the distal one slightlyflattened and slightly truncated. Lower anterior corner of pronotum sharply angled about 90. Mesoscutumwith notauli deeply impressed, reaching to level of hind margins of tegulae; central part of mesoscutumfairly evenly hirsute. Scutellum quite strongly convex, carinate laterally for its entire length. Mesopleuronweakly swollen centrally, anteroventrally finely and closely punctate, sternaulus vestigial; metapleuronsmooth, submetapleural carina complete. Posterior transverse carina of mesosternum strongly raised,centrally with a deep V-shaped notch. Propodeum of moderate length, anterior transverse carinae presentlaterally, curved forward lateromedially to join anterior part of lateromedian longitudinal carinae,centrally absent; posterior transverse carina complete, median portion of lateromedian longitudinal carinavestigial, indicating a transverse area superomedia that is about 1-6 times as broad as long; tubercle presentat lateral end of anterior transverse carina above and slightly behind propodeal spiracle (Fig. 27). Mid coxaunspecialized; hind tibia with 2-4 preapical bristles; largest bristle of hind tarsal claws spatulate. Fore winglength 10-12 mm; 3r-m present, enclosing a small obliquely trapezoidal petiolate areolet; 2m-cu joiningareolet near outer corner (Fig. 40); Rs moderately arcuate; cu-a virtually opposite Rs&M. Gaster withtergite 1 1-1-1-2 times as long as posteriorly broad, without a distinct carina joining spiracle to anteriormargin of segment. Tergite 3 with close coarse punctures centrally. Ovipositor projecting beyond apex ofgaster by 0-9-1-0 times length of hind tibia, slightly decurved with its dorsal apex flattened weakly. Female yellow, black-marked on interocellar area, mesoscutum in 3 almost contiguous spots, posteriorpart of tegula and ovipositor sheath. Wings hyaline with infumate distal apex. Male unknown. REMARKS. X. hiatus belongs to the elegans-group as defined by Townes & Chiu (1970). It is theonly Australian representative of this group which is easily recognizable by the sharply angledpronotal corner. X. hiatus is also distinctive in having 2m-cu joining the areolet very close to theouter corner. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), New South Wales, Tooloom, i.1926 (Hacker) (QM); 1 $ (paratype),Queensland, Cairns (BMNH); 1 $ (paratype), Mackay, v.1909 (Turner) (BMNH); 1 $ (paratype), MtGlorious, ii.1961 (Gressitt & Gressitt) (TC). Australia: 6 $ , New South Wales, Queensland. Xanthopimpla hirsuta (Girault)(Figs 21, 33) Austrapophua hirsuta Girault, 1930: 2. Holotype $, AUSTRALIA (QM) [examined].Xanthopimpla hirsuta (Girault) Townes etal, 1961: 56; Townes & Chiu, 1970: 272. Clypeus flat, more strongly transverse than is usual for the genus; face closely punctate. Flagellumincomplete. Lower anterior corner of pronotum about 95, slightly rounded. Mesoscutum with notaulistrongly impressed, reaching about to level of centre of tegulae; central part of mesoscutum sparsely ICHNEUMONIDAE OF AUSTRALIA 305 hirsute. Scutellum pyramidal with carina extending almost to posterior margin. Mesopleuron moderatelyweakly swollen centrally, anteroventrally closely and quite coarsely punctate, sternaulus absent; meta-pleuron smooth, unusual in not being defined anterodorsally as pleural carina extends only to level ofspiracle; submetapleural carina strong (Fig. 21). Posterior transverse carina of mesosternum weak, onlyslightly broadened centrally. Propodeum moderately long, anterior transverse carina complete laterally,posterior transverse carina complete, strong, lateromedian carinae defining an area superomedia which ismuch narrower posteriorly than centrally, the entire area being as long as centrally broad; tubercle weak(Fig. 33). Mid coxa unspecialized; hind tibia with about 4 preapical bristles; largest bristle on hind tarsalclaw slender, not spatulate. Fore wing length 11 mm; 3r-m absent; Rs simply arcuate; cu-a slightlyproximal to base of Rs&M. Gaster with tergite 11-2 times as long as posteriorly broad, with carinaextending from near spiracle to anterior end. Tergite 3 with coarse contiguous punctures centrally.Ovipositor projecting beyond apex of gaster by 0-1 times length of hind tibia, the ovipositor apex slightlydecurved, the lower valve with a distinct median ventral keel, not enclosed by upper valve. Female yellow, black-marked on interocellar area, back of head, mesoscutum in three stripes and onscuto-scutellar groove, paired spots on tergites 1, 3-5 and transverse black band on tergite 7. Wingshyaline, distally infumate. Male unknown. REMARKS. A very distinctive species on account of the complete area superomedia andincomplete areolet. Townes & Chiu (1970) placed X. hirsuta in the incompleta-group, but theovipositor is not very like that of the other species as the lower valve is not enclosed by the uppervalve. I believe it should be placed in the rhopaloceros-group. HOST RECORDS. None. MATERIAL EXAMINEDAustralia: 1 $ (holotype), Queensland, Herbert River, Macknade? v.1912 (QM). Xanthopimpla ochracea (Smith)(Fig. 37) Pimpla ochracea Smith, 1859: 172. Holotype $, MOLUCCAS (UM) [examined]. Pimpla caudata Smith, 1863: 10. Holotype , MISOOL (UM) [examined]. [Synonymized by Townes & Chiu, 1970: 126.] Xanthopimpla (?) caudata (Smith) Krieger, 1899: 65.Xanthopimpla (?) ochracea (Smith) Krieger, 1899: 65.Xanthopimpla crassa Krieger, 1899: 97. Lectotype $, PAPUA NEW GUINEA (MNHU), designated by Townes etal, 1961: 53. [Synonymized by Townes & Chiu, 1970: 126.]Xanthopimpla ochracea ochracea (Smith); Townes & Chiu, 1970: 126.Xanthopimpla ochracea peterseni Townes & Chiu, 1970: 125. Holotype $, NEW IRELAND (UZM). Syn. n. Clypeus weakly convex; face quite coarsely punctate. Flagellum with 35-38 segments, the distal oneflattened, rounded off. Lower anterior corner of pronotum fairly abruptly rounded. Mesoscutum withnotauli short, deep, reaching nearly to level of -centre of tegulae; central part of mesoscutum sparselyhirsute. Scutellum moderately strongly convex, carinate laterally to apex. Mesopleuron weakly swollencentrally, anteroventrally quite coarsely punctate, sternaulus weak, discernible; metapleuron smooth orwith some striation, submetapleural carina distinct. Posterior transverse carina of mesosternum veryabruptly raised near centre, produced into two acute lobes separated by a narrow V-shaped cleft, theselobes usually strengthened by a thicker outgrowth from the sternum. Propodeum short, anterior andposterior transverse carinae strong; lateromedian carinae present between transverse carinae, enclosing atransverse area superomedia that is 1-8-1-9 times as broad as long; tubercle vestigial (Fig. 37). Mid coxaunspecialized; hind tibia with 6-8 preapical bristles; largest bristle on hind tarsal claw spatulate. Fore winglength 6-9 mm; 3r-m present, enclosing a moderately small, slightly oblique areolet; Im-cu joining areoletnear centre; Rs very slightly sinuate; cu-a opposite base of Rs&M. Gaster with tergite 1 about as long asposteriorly broad, slightly more slender in males; petiolar spiracle without a carina extending from near itto anterior margin of segment. Tergite 3 moderately densely punctate. Ovipositor projecting beyond apexof gaster by 0-8 times length of hind tibia. Female yellow, black-marked at least on interocellar area and ovipositor sheath. Wings weaklyinfumate. Male similar to female but with a pair of lateral spots on propodeum and spots on tergites 3-7black. VARIATION. Many females are as extensively black marked as the males. Some individuals of 306 I. D. GAULD both sexes have black marks on the mesoscutum and on tergite 1 a pair of faint brownish spots.Occasionally males have only the interocellar area black. REMARKS. This species belongs to the ochracea-gioup of Townes & Chiu (1970) and isdistinguished by the strongly raised posterior transverse carina of the mesosternum. Townes &Chiu (1970) recognized seven subspecies. Several of these would seem to warrant specific statusand in the present work I have restricted ochracea to the ochracea ochracea and ochraceapeterseni of Townes & Chiu. HOST RECORDS. None. MATERIAL EXAMINED Indonesia: 1 $ (holotype of ochracea), Moluccas, Aru (UM); 1 $ (holotype ofcaudata), Misool (UM).Australia: 2 $ (paratypes of ochracea peterseni) , Queensland, N. Stradbroke Is. (QM). Australia: 15 $, 10 cf , Northern Territory, Queensland (Map 12). Indonesia: 3 $, 2 cf , Irian Jaya,Misool. Papua New Guinea: 2 $ , 3 cf . (BMNH; TC) Xanthopimpla pubidorsis Townes & Chiu (Fig. 25)Xanthopimpla pubidorsis Townes & Chiu, 1970: 86. Holotype 9, IRIAN JAVA (RHN) [examined]. Clypeus flat; face finely punctate. Flagellum with 36-37 segments, the distal segment only slightlyflattened, terminally rounded abruptly. Lower anterior margin of pronotum obtusely rounded. Mesoscu-tum with notauli deeply impressed, reaching to level of hind margin of tegulae; median lobe of mesoscutumwith fine pale pubescence. Scutellum convex, carinate laterally to entire length. Mesopleuron weaklyswollen centrally, anteroventrally finely and very sparsely punctate; sternaulus distinct, posteriorly fairlyshallow; metapleuron smooth, submetapleural carina distinct. Posterior transverse carina of mesosternumvery broad, with a small median notch. Propodeum moderately long, anterior and posterior transversecarinae complete, lateromedian carinae delimiting a large area superomedia that is slightly longer thanbroad; tubercle moderately weak (Fig. 25). Mid coxa unspecialized; hind tibia with 0-1 preapical bristles;largest bristle on hind tarsal claw spatulate. Fore wing length 7-8 mm; 3r-ra present enclosing a ratherbroad rhombic areolet; 2m-cu joining very slightly distal to centre; Rs slightly arcuate; cu-a opposite baseof Rs&M. Gaster with tergite 1 elongate, 1-2 times as long as broad posteriorly, without carina extendingfrom near spiracle to anterior end. Tergite 3 with very superficial punctures. Ovipositor projecting beyondapex of gaster by about 0-5 times length of hind tibia. Female yellow, only scape partially, interocellar area and ovipositor sheath blackish. Wings hyaline,apices slightly infumate. Male similar to female. REMARKS. X. pubidorsis belongs to the splendens-group and may be recognized by thecombination of strong notauli, deep sternaulus, complete scutellar carinae, smooth meta-pleuron and finely punctate mesopleuron. Unlike X. barak this species always has the areasuperomedia delineated laterally. HOST RECORDS. None. MATERIAL EXAMINED Irian Jaya: 1 $ (holotype), Sigi Camp, 1350m, ii.1939 (Toxopeus) (RNH). Australia: Queensland, 1 cf , Capsize Ck, 64 km N. of Archer Xing, Cape York, vi.1975 (Monteith)(ANIC); 1 $, 1 cf, Claudie R., v.1966 (McAlpine) (AM); 1 $, Claudie R., 2 km W. of Mt Lomond,xii.1971 (McAlpine & Holloway) (AM); 1 cf, Gap Ck, 1550'S, 14520'E, v.1981 (Naumann) (ANIC);1 $, Peach Ck. Xing, 25 km NNE. Coen, vii.1976 (Monteith & Monteith) (ANIC). Xanthopimpla quadridens Townes & Chiu (Figs 22, 35, 42)Xanthopimpla quadridens Townes & Chiu, 1970: 278. Holotype cf , AUSTRALIA (NMV) [examined]. Clypeus flat; face strongly punctate. Flagellum with 36-38 segments, the distal segment evenly roundedapically. Lower anterior corner of pronotum very obtusely angled. Mesoscutum with notauli deeplyimpressed, reaching almost to level of hind edges of tegulae; central part of mesoscutum with few sparse ICHNEUMONIDAE OF AUSTRALIA 307 hairs. Scutellum weakly convex, carinate laterally for its entire length. Mesopleuron weakly swollencentrally, anteroventrally quite closely and coarsely punctate; sternaulus vestigial; metapleuron obsoletelypunctate, submetapleural carina distinct, discernible as a wide flat ridge (Fig. 22). Posterior transversecarina of mesosternum weakly raised, centrally with a broad U-shaped notch. Propodeum short; anteriortransverse and lateromedian longitudinal carinae absent, posterior transverse carina distinct, centrallyinterrupted, laterally raised into flanges; lateral tubercle present above and slightly behind propodealspiracle, the tubercle with conspicuous fine setiferous punctures (Fig. 35). Mid coxa unspecialized; hindtibia with 5-8 preapical bristles; largest bristle of hind tarsal claws slender. Fore wing length 11-12 mm;3r-m absent (Fig. 42); Rs almost straight; cu-a opposite base ofRs&M. Gaster with tergite 1 1-4-1-5 timesas long as broad posteriorly, without distinct carina joining spiracle to anterior margin of segment. Tergite3 with coarse close punctures centrally. Ovipositor projecting beyond apex of gaster by 1-0-14 timeslength of hind tibia, slightly decurved, its tip cylindrical. Female yellow, black-marked on interocellar area, frons and occiput continuously, posterior part oftegula, mesopleuron in a stripe below subalar prominence, three adjacent spots on mesoscutum, scutellargroove, anterior of propodeum except centrally, hind coxa with a spot, distal apices of femur and tibia,proximal apex of tibia, hind tarsal segment 5, gaster with central transverse bar on tergite 1, bars on tergite1, bars on tergites 3, 4, 6-8; tergites 5 and 9 with lateral black spots. Wings hyaline. Male similar tofemale. VARIATION. Some specimens may have minute black spots laterally on tergite 2 and the spots oftergite 5 centrally confluent to form a dumb-bell-like mark. A few males are less extensivelyblack and only have lateral black spots on tergites 3-6. REMARKS. This species is the largest one without an areolet. It is easily recognized by theconspicuous propodeal tubercles, long ovipostor and complete submetapleural carina. X.quadridens belongs to the rhopaloceros-group as defined by Townes & Chiu (1970). HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 cf (holotype), Victoria, Warburton, i.1955 (Neboiss) (NMV). Australia: 1 $, New South Wales, Mt Keira, xii.1981 (Holloway) (AM); 5 $, 1 cf, Queensland, MtGlorious, x-xi. (TC); 1 cf, Mt Tambourine, x. (TC); 1 $, Victoria, Burnley (Prescotf) (NMV); 1 <j>,Kallista, iii. 1952 (Burns) (NMV). Xanthopimpla rhopaloceros Krieger(Figs 28, 52) Xanthopimpla rhopaloceros Krieger, 1914: 23; Townes & Chiu, 1970: 285. Lectotype cf, AUSTRALIA (MNHU), designated by Townes et al., 1961: 66 [examined].Austrapophua xanthopimploides Girault, 1926: 135. Holotype $, AUSTRALIA (QM) [examined]. [Synony- mized by Townes & Chiu, 1970: 285.]Xanthopimpla xanthopimploides (Girault) Townes et al., 1961: 72. Clypeus flat; face strongly closely punctate. Flagellum with 33-34 segments, the terminal segment evenlyrounded apically. Lower anterior corner of pronotum evenly rounded. Mesoscutum with notauli deep,extending only to level of centre of tegulae; central lobe of mesoscutum sparsely hirsute. Scutellum weaklyconvex, carinate almost to hind margin. Mesopleuron centrally very weakly swollen, anteroventrallysparsely punctate; sternaulus vestigial; metapleuron smooth, submetapleural carina absent. Posteriortransverse carina of mesosternum moderately broadened, with a wide V-shaped central notch. Propodeummoderately long; anterior and lateromedian carinae absent; posterior transverse carina complete; tuberclequite strong (Fig. 28). Mid coxa with a blunt prominence on anterolateral side; hind tibia with 9-15preapical bristles; largest bristle on hind tarsal claw not spatulate. Fore wing length 4-11 mm; 3r-m absent;Rs almost straight; cu-a opposite base of Rs&M. Gaster with tergite 11-2 times as long as posteriorlybroad, without a carina joining spiracle to anterior margin of tergite. Tergite 3 with coarse close punctures.Ovipositor projecting beyond apex of gaster by about 0-4 times length of hind tibia, its apex subcylindrical.Female yellow, black marks on interocellar area, occiput, mesoscutum in three stripes, scutoscutellargroove, on anterior transverse band on propodeum, a central band on tergites 1, 3 and 7, paired spots ontergites 2, 4 and 6, hind tibia proximally and distally and hind basitarsus proximally. Wings hyaline. Malesimilar to female but with mesoscutal stripes usually fused to form a single black area. 308 I. D. GAULD VARIATION. Some specimens have faint black dots on tergite 5 whilst others have no black markon tergite 2. Not infrequently tergites 6 and 7 are almost entirely black. REMARKS. This species belongs to the rhopaloceros-group of Townes & Chiu (1970). X.rhopaloceros is easily recognized by the characteristic shape of the mid coxa, the lack of anareolet, and the colour pattern. It differs from the related X. summervillei most obviously inhaving a strong black band on tergite 1 (Fig. 52) and in having the occiput black. X. summervilleinever has a black occiput and at the most only a vague brownish central mark on tergite 1. HOST RECORDS. Nolidae: Uraba lugens Walker (DPIQ). Tortricidae: Epiphyas postvittana(Walker) (Dumbleton, 1940); Merophyas divulsana (Walker) (DPIQ). MATERIAL EXAMINED Australia: 1 0" (lectotype of rhopaloceros), South Australia, Adelaide (MNHU); 1 9 (holotype ofxanthopimploides), Queensland, Riverview (QM). Australia: 74 9> 31 cf , Australian Capital Territory, New South Wales, Queensland, South Australia,Tasmania, Victoria, Western Australia (Map 14) (AM; ANIC; BMNH; DAH; QM; TC). Xanthopimpla striata Townes & Chiu (Fig. 26)Xanthopimpla striata Townes & Chiu, 1970: 91. Holotype 9, AUSTRALIA (BPBM) [examined]. Clypeus swollen centrally near upper margin; face coarsely but sparsely punctate. Flagellum with 41segments. Lower anterior corner of pronotum fairly evenly rounded but with lower corner of epomiaright-angled and projecting, appearing superficially like the corner of pronotum. Mesoscutum with notaulivery strongly impressed almost to hind margin; central part of mesoscutum with sparse dark pubescence.Scutellum very strongly convex, carinate laterally about 0-8 of its length. Mesopleuron weakly swollencentrally, anteroventrally coarsely and closely punctate; sternaulus deeply impressed; metapleuronanteroventrally striate, posteriorly punctate, submetapleural carina complete. Posterior transverse carinaof mesosternum moderately broad, with a deep median V-shaped notch. Propodeum of moderate length,with posterior and anterior transverse carinae strong, the latter absent centrally; lateromedian longitudinalcarinae indistinct but area superomedia discernible transverse; lateral tubercle indistinct (Fig. 26). Midcoxa unspecialized; hind tibia with 6-9 preapical bristles, fewer in male; largest bristle on hind tarsal clawspatulate. Fore wing length 7-8 mm; 3r-m present, enclosing a moderately large rhombic areolet; 2m-cujoining areolet at centre; Rs arcuate; cu-a virtually opposite base ofRs&M. Gaster with tergite 11-3 timesas long as broad posteriorly, with a strong carina extending from near spiracle to anterior end of segment.Tergite 3 with fine, close, rather elongate punctures. Ovipositor projecting beyond apex of gaster by0-3-0-4 times length of hind tibia, its apex subcylindrical. Female yellow, black-marked on interocellar area, vertex, mesoscutum centrally and indistinctlylaterally, and gaster with spots laterally on tergites 1-3 and 7 centrally. Apex of wings infumate. Malesimilar to female but with only interocellar area, mesoscutum centrally, propodeum centrally, paired spotsof tergites 2-6 and remaining tergites transversely banded black. VARIATION. Some females have paired black spots present on tergites 4 and 5 and have someblack on the propodeum anteromedially, like the male. REMARKS. X. striata belongs to the splendens-group as defined by Townes & Chiu (1970).Superficially this species may be mistaken for members of the elegans-group on account of theapparently sharply angulate anterior corner of the pronotum. However, this angulation is thelower corner of the epomia and not the pronotal margin which underlies this projection and isevenly rounded. In species of the elegans-group it is the actual pronotal margin that is sharplyangled. Unlike X. hiatus, the only Australian member of the elegans-group, X. striata has astriate metapleuron, regularly rhombic areolet with 2m-cu joining in centre, a rather shortovipositor and strongly impressed sternaulus. HOST RECORDS. None. MATERIAL EXAMINEDAustralia: 1 9 (holotype), Queensland, Cairns (Kusche) (BPBM). ICHNEUMONIDAE OF AUSTRALIA 309 Australia: 1 cf , Queensland, Ellis Beach, N. of Cairns, iv.1967 (Colless) (ANIC). Papua New Guinea:2 $, Dumpu, 250 m, i-ii.1979 (Sedlacek) (TC). Xanthopimpla summervillei (Girault)(Figs 30, 53) Austrapophua summervillei Girault, 1926: 136. Holotype $, AUSTRALIA (QM) [examined]. Xanthopimpla summervillei (Girault) Townes etal., 1961: 70. Xanthopimpla summervillei summervillei (Girault) Townes & Chiu, 1970: 295. Clypeus flat; face quite strongly, closely punctate. Flagellum with 32-34 segments, the terminal segmentevenly rounded apically. Lower anterior corner of pronotum evenly rounded. Mesoscutum with notaulideep, extending beyond level of centre of tegulae; central lobe of mesoscutum very sparsely hirsute.Scutellum moderately strongly convex, carinate for 0-8 of its length. Mesopleuron centrally very weaklyswollen, anteroventrally smooth; sternaulus vestigial; metapleuron smooth, submetapleural carina absent.Posterior transverse carina of mesosternum quite strongly broadened, with a deep median V-shapednotch. Propodeum moderately long, anterior transverse carina present laterally, often reaching laterome-dian carinae which are only present anteriorly; posterior transverse carina complete; tubercle entirelyabsent, its possible former position indicated only by long hairs (Fig. 30). Mid coxa unspecialized; hindtibia with 6-10 preapical bristles; largest bristle on hind tarsal claw not spatulate. Fore wing length 4-8 mm;3r-m absent ;Rs slightly sinuous \cu-a opposite base of Rs&M. Gaster with tergite 1 1-1-1 -3 times as long asposteriorly broad, without a carina joining spiracle to anterior margin of tergite. Tergite 3 with very finesparse setiferous punctures. Ovipositor projecting beyond apex of gaster by 0-7-0-8 times length of hindtibia, its apex subcylindrical. Female yellow, black-marked on interocellar area, mesoscutum in a broad transverse band, anteriorpart of propodeum, bands on tergites 3, 4 and 7 and hind tibia proximally. Male similar to female but withblack spots on tergites 4-6 and proximal black mark on hind tibia less distinct. Apex of fore wing infumate. VARIATION. In some females the black band on tergite 3 is present as two barely confluent spots.A few have a faint brownish mark on tergite 1 centrally. REMARKS. X. summervillei belongs to the X. rhopaloceros-group of Townes & Chiu (1970). Liketwo of the other Australian species of this group (X. binodus and X. rhopaloceros) it lacks thesubmetapleural carina. It can be distinguished from these other species most easily by itscharacteristic colour pattern (Fig. 53).Townes & Chiu (1970) recognize four subspecies. The nominate one is restricted to Australia. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), New South Wales, Tweed Heads (QM).Australia: 17 $, 11 cf , New South Wales, Queensland (Map 15) (ANIC; BMNH: TC). Xanthopimpla terminalis (Brulle) (Figs 36, 47) Pimpla terminalis Brulle, 1846: 96. Holotype $, AUSTRALIA (MNHN) [examined]. Notopimpla terminalis (Brulle) Krieger, 1899: 107. Xanthopimpla terminalis (Brulle) Townes etal., 1961: 70; Townes & Chiu, 1970: 300. Clypeus flat; face rather finely punctate. Flagellum with 37-39 segments, the most distal segmentsomewhat flattened, distally truncate so antenna has a small elliptical flat area at the end (Fig. 47). Loweranterior corner of pronotum produced into a sharp tooth. Mesoscutum with notauli deep but short, barelyreaching to level of anterior margin of tegulae; central part of mesoscutum virtually glabrous. Scutellumrather flat, carinate laterally only at extreme base. Mesopleuron very weakly swollen centrally, anteroven-trally very finely and sparsely punctate, sternaulus absent; metapleuron smooth, submetapleural carinapresent. Posterior transverse carina of mesosternum slightly raised, with a minute central notch. Pro-podeum moderately long, posterior transverse carina complete, anterior and lateromedian carinae absent;tubercle strong (Fig. 36). Mid coxa unspecialized; hind tibia with three preapical bristles arranged in astraight line; largest bristle on hind tarsal claws spatulate. Fore wing length 12-13 mm; 3r-m present,enclosing a small areolet, 2m-cu joining areolet close to proximal side, Rs sinuate; cu-a opposite base of 310 I. D. GAULD Rs&M. Gaster with tergite 1 longer than posteriorly broad, without a carina joining spiracle to anteriormargin of tergite. Tergite 3 with large, shallow, close punctures. Ovipositor projecting beyond apex ofgaster by 1 -0-1 1 times length of hind tibia, exceptional in having a small swelling on upper valve near apex.Female yellow, scape, interocellar area, head behind vertex, three mesoscutal marks and ovipositorsheath only, black. Wings slightly infumate, apices more distinctly so. Male similar to female, butmesoscutal marks more extensive and tergite 1 with a pair of black spots, tergite 3 with traces of spots. REMARKS. X. terminalis belongs to the terminalis-group as defined by Townes & Chiu (1970). InAustralia it is one of the most distinctive species, easily recognized by the flat acarinatescutellum, apically specialized antennae (in both sexes) and nodulate ovipositor. It is quiteclosely related to the Papuan species X. interceptor (Smith). X. terminalis has recently beencollected flying amongst the understorey vegetation in wet forest. It seems to prefer very shadedhabitats and often flies quite high, 2-3 m from the ground. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), no further data (MNHN).Australia: 9 $ , 2 cf , New South Wales, Queensland (Map 16) (AM; ANIC; BMNH; TC). Tribe DELOMERISTINI (= Theroniini sensu Townes) This tribe, a rather heterogeneous assemblage of taxa, includes six genera, Atractogaster,Delomerista, Hybomischos, Pseudorhyssa, Perithous and Theronia, placed together on accountof the absence of a basal tooth on the fore tarsal claw of the female, the unusually elongate malesubgenital plate and the often delineated, rather long, area superomedia. The final instar larvaeall possess a large internal tooth on the mandible and have a well developed hypostoma (Short,1978). Pseudorhyssa, Delomerista and Atractogaster are north temperate genera whilst Perithousand Hybomischos also occur on some higher mountains in the Oriental region (Gupta, 19820,b). Theronia, a predominantly tropicopolitan genus, is the only delomeristine genus representedin Australia. THERONIA Holmgren Medium to moderately large-sized species, fore wing length 7-14 mm; clypeus with margin slightlyconcave; mandible not twisted, weakly to moderately narrowed; occipital carina complete. Epicnemialcarina present; mesopleural suture centrally angled; propodeum with strong carinae, spiracles elongatelyoval. Tarsal claws of female without basal lobes, usually large with spatulate bristle. Fore wing with 3r-mpresent, enclosing a rhombic areolet; hind wing with first abscissa of Ci about 0-3 times length of cu-a.Tergite 1 rather slender; tergites 2-5 smooth and highly polished; ovipositor projecting beyond apex ofgaster by 1-2-1-9 times length of hind tibia. REMARKS. Theronia is a large, mainly tropicopolitan genus with a few species occurring intemperate areas. Traditionally it has been divided into a number of subgenera (Gupta, 1962;Townes, 1969) although these are more reasonably comparable with the species-groups used inrelated genera (e.g. Xanthopimpld) . Gauld (1984) recognized three subgenera as occurring inAustralia, T. (Theronia), T. (Parema) and T. (Nomosphecia}. Theronia species are mostly yellowish or green insects which seem to be restricted to foresthabitats. In flight they are conspicuous amongst the vegetation but some species spendprotracted periods host-searching on foot, particularly amongst the leaf litter and along earthenbanks. When caught, Theronia species sink their large claws into their captor (or his net). Theseclaws are difficult to disengage and usually break near their bases at the level of an apparentlyfluid filled cavity. Townes (1940) postulated these may function as poison fangs to deterpredators but as yet this has not been proven. Species of the subgenus Nomosphecia are known to be parasites of vespids (Gupta, 1962) butthe majority of Theronia species are associated with Lepidoptera. Although recorded as primary ICHNEUMONIDAE OF AUSTRALIA 311 parasites (Gupta, 1962) there is increasing evidence to suggest many are hyperparasitic,particularly on other pimplines (Townes, 1940; Short, 1978). Whether hyperparasitism isobligatory or facultative is not known but in the field T. (T.) maculosa is found in the samemicrohabitat as Echthromorpha intricatoria, at the same time, but in lesser numbers. Five species of Theronia occur in Australia. One, T. (P.) penetrans is also widespread inSulawesi, the Moluccas and New Guinea whilst the other four are endemic. The tropical speciesT. (N.) melanosoma is very closely related to two Melanesian species, T. (N.) elegans Gupta andT. (N.) diligens (Smith). As the subgenera occurring in Australia have recently been redefined (Gauld, 1984) the keygiven below is direct to species though the subgenera are shown in brackets. Key to Australian species of Theronia 1 Lower tooth of mandible about twice as long as the upper (Fig. 72); ovipositor pro- jecting beyond apex of gaster by 1-8-1-9 times length of hind tibia; hind tibia blackish(Nomosphecia) melanosoma Morley(p. 311) - Lower tooth of mandible almost equal in length to upper (Fig. 73) ; ovipositor projecting beyond apex of gaster by 1 2-1 5 times length of hind tibia ; hind tibia yellowish or brownish 2 2 Lateral longitudinal carina of $ extending more than 0-6 of length of scutellum; ovipositor with lower valve partially enclosing the upper (Fig. 74); posterior ocelli separated by aboutminimum ocellar diameter, so ocellar triangle is a narrow-based isosceles triangle (Fig. 77)(Paremo) penetrans (Smith) (p. 312) - Lateral longitudinal carina of $ extending at most 0-5 of length of scutellum; ovipositor with lower valve simple, not enclosing the upper (Fig. 75); posterior ocelli separated by at leastmaximum ocellar diameter, the ocellar triangle thus being almost equilateral (Fig. 76)(Theronia) 3 3 Propodeum with anterior parts of lateromedian longitudinal carinae subparallel, the area superomedia more or less closed posteriorly (Fig. 68); submetapleural carina fairly evenlybroadened anteriorly (Fig. 67); tergites of gaster with greenish tinge (most pronounced infresh material) and with pairs of black spots on at least tergites 2-4 maculosa Krieger(p. 314) - Propodeum with anterior parts of lateromedian longitudinal carinae posteriorly divergent, the area superomedia not closed posteriorly (Fig. 69); submetapleural carina abruptly broadenedanteriorly into an acute flange (Figs 65 , 66) ; tergites of gaster yellow to orange-brown , withoutpaired spots 4 4 Sternite 1 of $ reaching to or slightly beyond centre or tergite 1 (Fig. 65); hind femur with close, short pubescence ventrally ; distal fiagellar segments slightly longer than broad, theflagellum slightly tapered towards end (Fig. 71); mesoscutum with three black longitudinalstripes fraucai sp. n. (p. 313) - Sternite 1 of $ reaching 0-2-0-3 of length of tergite 1 (Fig. 66); hind femur with long sparse pubescence ventrally; distal flagellar segments slightly broader than long, the flagellum parallel-sided towards end (Fig. 70) ; mesoscutum without black marks steindachneri Krieger (p. 315) THERONIA subgenus NOMOSPHECIA Gupta Theronia (Nomosphecia) Gupta, 1962: 68. Type-species: Theronia zebroides Krieger, by original designa-tion. Mandible moderately tapered, with lower tooth almost twice the length of the upper (Fig. 72). Scutellumwith lateral longitudinal carinae present only at extreme anterior end. Lower valve of ovipositor partiallyenclosing the upper. Theronia (Nomosphecia) melanosoma Morley(Fig. 72) Theronia melanosoma Morley, 1914: 47. Lectotype $ , AUSTRALIA (BMNH), designated by Townes et al, 1961: 74 [examined].Theronia (Nomosphecia) melanosoma Morley; Gupta, 1962: 76. 312 I. D. GAULD Fore wing length 9-14 mm. Genae strongly constricted behind the eyes; posterior ocellus separated fromeye by 1-1 times maximum ocellar diameter, interocellar distances 0-7 times maximum ocellar diameter,the ocellar triangle higher than broad. Occipital carina complete. Antenna setaceous; scape apicallytruncate 45; flagellum with 37-39 segments, the penultimate distal segments transverse. Scutellumcarinate laterally only at extreme anterior end. Propodeum with lateral longitudinal carinae complete,lateromedian carinae divergent posteriorly and confluent with posterior transverse carina which is absentcentrally; propodeal spiracle elliptical; submetapleural carina strongly broadened anteriorly. Fore tibiasimple; fore tarsal segments 3 and 4 without stout bristles; hind femur with short close hairs ventrally. Forewing with cu-a slightly distal to base ofRs&M; areolet large and slightly transverse. Gaster of female withtergite 1 1-4-1-5 times as long as posteriorly broad, its sternite reaching 0-4 of its length. Ovipositorprojecting beyond apex of gaster by 1-8-1-9 times length of hind tibia. Head, alitrunk and most of fore legreddish brown; tarsi, antenna, most of hind leg and gaster black; tergites 2-7 margined posteriorly withyellow. Wings strongly infumate. VARIATION. Some specimens have less extensive black areas of coloration and have the hindfemora, trochanters, coxae and the scapes reddish brown. REMARKS. A very distinctive species on account of its mandible. T. (N.) melanosoma appears tobe restricted to the tropical north-east of Australia. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), 2 $, 3 cf (paratypes), Queensland, Kuranda near Cairns, ii.1902 (Turner)(BMNH). Australia: 1 $, Queensland, Cape York Pen., Iron Rg., v.1974 (Walford-Huggins) (AM); 1 O", CapeYork Pen., Lockerbie Area, iv.1973 (Monteith) (ANIC); 3 $, 1 cf , Claudie, 8 km W. Mt Lamond, i.1972(Me Alpine & Holloway) (AM); 1 cf , Dividing Rg., 1140'S, 14245'E, vii.1972 (Monteith) (ANIC); 1 $,Gordonvale, x.1924 (BMNH); 1 cf , 7 km N. of Hope Vale Mission, 1514'S, 14507'E, x.1980 (Cardale)(ANIC); 5 $, 2 C? Middle Claudie R., ix.1974 (Daniels & Moulds) (AM). THERONIA subgenus PAREMA Gupta Theronia (Parema) Gupta, 1962: 54. Type-species: Theronia nigrobalteata Cameron, by original designa-tion. Mandibles weakly tapered with teeth about equal. Scutellum with lateral longitudinal carina extending0-6-0-8 of its length. Lower valve of ovipositor partially enclosing the upper (Fig. 74). Theronia (Parema) penetrans (Smith)(Fig. 74) Pimpla penetrans Smith, 1859: 173. Lectotype $ , MOLUCCAS (UM), designated by Townes etal, 1961: 75 [examined] . Xanthopimpla(>.) penetrans (Smith) Krieger, 1899: 65.Theronia cephalotes Krieger, 1899: 116. Lectotype $, SULAWESI (MNHU), designated by Townes et al, 1961: 75. [Synonymized by Townes etal., 1961: 75.]Theronia fumata Krieger, 1899: 118. Lectotype $, PAPUA NEW GUINEA (MNHU), designated by Townes et al., 1961: 75. [Synonymized by Townes et al., 1961: 75.]Theronia papuana Cameron, 1911a: 209. LECTOTYPE <j>, IRIAN JAVA (BMNH), here designated [examined]. [Synonymized by Townes et al., 1961: 75.]Theronia penetrans (Smith) Morley, 1914: 45.Theronia penetrans cephalotes Krieger; Townes etal., 1961: 75.Theronia penetrans penetrans (Smith); Townes etal., 1961: 75. Fore wing length 6-11 mm. Genae strongly constricted behind eyes; posterior ocellus separated from eyeby 1 -2-1-3 times maximum ocellar diameter, interocellar distance 0-8 times maximum ocellar diameter, theocellar triangle higher than broad (Fig. 77). Occipital carina complete. Antenna slightly setaceous; scapeapically truncate 50; flagellum with 38-40 segments, the penultimate distal segments slightly broader thanlong. Scutellum carinate laterally 0-6-0-7 of its length. Propodeum with lateromedian carinae presentanteriorly, parallel, lateral longitudinal carinae complete and posterior transverse carina weak, centrallyabsent; propodeal spiracle elliptical; submetapleural carina strongly and abruptly widened anteriorly into ICHNEUMONIDAE OF AUSTRALIA 313 an acute lobe. Fore tibia unspecialized; fore tarsal segments 3 and 4 without stout bristles; hind femur withshort close hair ventrally. Fore wing with cu-a distal to base of Rs&M; areolet of moderate size, oblique.Gaster of female with tergite 1 1-5-1-7 times as long as broad, its sternite reaching almost to centre.Ovipositor projecting beyond apex of gaster by 1 -3 times length of hind tibia.Orange-brown, the flagellum and hind tarsus blackish. Wings strongly infumate. VARIATION. The Australian examples are slightly smaller than the Papuan specimens butmorphologically otherwise virtually identical, although there is a little variation in the shape ofthe area superomedia. Over its range from Sulawesi to Australia there is considerable variationin propodeal sculpture, depth of facial punctures, infumation of wings and extent of darkermarkings on the body. Gupta (1962) recognized two subspecies though remarking thatintergradation occurred at least on many characters. REMARKS. This species can be recognized easily by the form of the ovipositor and closeness of theposterior ocelli. It is only known to occur in northern Queensland though it is widespread fromSulawesi to New Guinea. HOST RECORDS. None. MATERIAL EXAMINED Indonesia: 1 9 (lectotype ofpenetrans), Moluccas, Aru (UM); 1 $, (lectotype ofpapuana), Irian Jaya,Alkmaar(BMNH). Australia: 1 $, 1 C?, Queensland, Moses Ck, 4 km N. by E. Mt Finnigan (1547'S, 14517'E), x.1980(Cardale) (ANIC); 1 cf , Shipton's Flat (1547'S, 14514'E), x.1980 (Cardale) (ANIC); 1 cf , same locality,v.1981 (Naumann) (ANIC). THERONIA subgenus THERONIA Holmgren Theronia Holmgren, 1859: 123. Type-species: Pimpla flavicans F. (= Ichneumon atalante Poda), by monotypy.Pseudacoenites Kriechbaumer, 1892: 219. Type-species: Pseudacoenltes moravicus Kriechbaumer (= Pimpla laevigata Tachek), by monotypy. Poecilopimpla Cameron, 1903: 141. Type-species: Poecilopimpla lucida Cameron, by monotypy.Orientotheronia Morley, 1913&: 146. Type-species: Orientotheronia rufescens Morley (= Pimpla zebra Snellen van Vollenhoven), by original designation.Theronia (Theronia) Holmgren; Gupta, 1962: 9. Mandible weakly to moderately narrowed, subequally bidentate (Fig. 73). Scutellum with lateral carinaeextending 0-2-0-5 of its length. Lower valve of ovipositor not enclosing the upper (Fig. 75). Theronia (Theronia) fraucaisp. n. (Figs 71, 76) Fore wing length 6-7 mm. Mandible moderately narrowed, the teeth almost equal. Genae evenly roundedbehind eyes; posterior ocellus separated from eye by 1-2-1-3 times its maximum diameter, interocellardistance 1-0 times maximum ocellar diameter, the ocellar triangle equilateral (Fig. 76). Occipital carinacomplete. Antenna setaceous; scape truncate apically 50; flagellum with 33-34 segments, the penultimatedistal segments slightly longer than broad (Fig. 71). Scutellum with lateral carinae extending 0-3-0-5 of itslength. Propodeum with lateral longitudinal carinae complete, lateromedian carinae present anteriorly,divergent posteriorly and contiguous with ends of posterior transverse carina which is absent centrally;propodeal spiracle elliptical; submetapleural carina strongly broadened anteriorly. Fore tibia unspecial-ized; fore tarsal segments 3 and 4 without bristles; hind femur with short hairs ventrally. Fore wing withcu-a distal to base of Rs&M; areolet large, almost rhombic. Gaster of female with tergite 1 1-6-1-7 times aslong as posteriorly broad, dorsally flattened and with its sternite reaching to or slightly beyond centre(Fig. 65). Ovipositor projecting beyond apex of gaster by 1-3-1-4 times length of hind tibia, the lower valvesimple. Yellowish, with three black stripes on mesoscutum; yellow colour slightly infuscate close to ocelli and onscape. Flagellum dark brown; femora dorsally and tergites extensively reddish yellow. Ovipositor sheathblack. Wings almost completely hyaline. REMARKS. A small insect not obviously closely related to other Australian species. The rather 314 I. D. GAULD slender tergite 1 with the elongate sternite immediately distinguishes fraucai from steindachneriand although it resembles maculosa more in this feature the propodeal carination is quitedifferent. T. (T.) fraucai is known only from Queensland. HOST RECORDS. None. MATERIAL EXAMINED Holotype 9, Australia: Queensland, coast range near Biggenden, viii.1974 (Fraucd) (ANIC). Paratypes. Australia: 1 c?, same locality as holotype, x.1976 (Fraucd) (ANIC); 1 9, same locality,viii. 1977 (Frauca} (ANIC); 1 c? , Brisbane-Indooroopilly Road, xi. 1976 (5oucdt)(BMNH); 1 $,PineCk,S. of Bundaberg, ix.1976 (Frauca) (BMNH). I hcronin (Theronia) maculosa Krieger(Figs 67, 68, 73) Theronia maculosa Krieger, 1906: 239. Lectotype $, AUSTRALIA (MNHU), designated by Townes et al., 1961:74.Theronia viridicans Morley, 1914: 46. Holotype $, AUSTRALIA (BMNH) [examined]. [Synonymized by Townes etal., 1961:74.]Theronia (Theronia) maculosa Krieger; Gupta, 1962: 46. Fore wing length 8-12 mm. Mandible weakly narrowed, the teeth almost equal. Genae evenly constrictedbehind the eyes; posterior ocellus separated from eye by 1-0 times its maximum diameter, interocellardistance 0-8 times maximum ocellar diameter, the ocellar triangle almost equilateral. Occipital carinacomplete. Antenna setaceous; scape apically truncate 55; flagellum with 40-42 segments, the penultimatedistal segments subquadrate. Scutellum carinate 0-2-0-4 of its length. Propodeum with lateral longitudinalcarinae complete, lateromedian carinae anteriorly strong, almost parallel and with rectilinear areasuperomedia almost closed posteriorly (Fig. 68); propodeal spiracle elliptical; submetapleural carinaevenly broadened anteriorly (Fig. 67). Fore tibia unspecialized; fore tarsal segments 3 and 4 without strongbristles; hind femur with rather short hairs ventrally. Fore wing with cu-a opposite or slightly distal to baseofRs&M; areolet rather narrow, oblique. Gaster of female with tergite 1 1-6-1-8 times as long as broad, itssternite reaching almost to centre. Ovipositor projecting beyond apex of gaster by 1-4-1-5 times length ofhind tibia. Yellowish brown, with flagellum, marks on base and apex of hind femur and hind tarsus blackish; gasterfrequently with black spots on tergites 1-5. Wings moderately infumate. VARIATION. Whilst alive most specimens have a greenish gaster and frequently also green hindlegs but this colour fades rapidly after death. Some variation in the distribution of black spotsoccurs but all specimens seem to have them on tergites 2-4. REMARKS. A common and rather widespread endemic Australian species (Map 18) found mostlyin cooler damp woodlands. It is readily distinguished from other Theronia species by the blackspots on some gastral tergites. It is the only Australian species with a pronounced greenish tinge,and freshly emerged females are often brilliant emerald-green. HOST RECORDS. Anthelidae: Anthela acuta (Walker). Saturniidae: Antheraea astrophela Walker(Chadwick & Nikitin, 1976). Short (1978), on the basis of dissection of a cocoon oiAntheraea sp.from which Theronia had emerged, suggested Theronia is hyperparasitic, using Pimplini ashosts. MATERIAL EXAMINED Australia: 1 $ (holotype of viridicans) , Tasmania, 1877 (Atkinson) (BMNH). Australia: 97 9? 34 cf, Australian Capital Territory, New South Wales, south-eastern Queensland,South Australia, Tasmania, Victoria (Map 18) (AM; ANIC; BMNH; DAT; NMV; TC). ICHNEUMONIDAE OF AUSTRALIA 315 Theronia (Theronia) steindachneri Krieger(Figs 69, 70) Theronia steindachneri Krieger, 1906: 320. Lectotype 9, AUSTRALIA (NM), designated by Townes et at., 1961:76.Theronia dubia Krieger, 1906: 320. Lectotype 9 , AUSTRALIA (MNHU) , designated by Townes et al. , 1961 : 76. [Synonymized by Townes etal, 1961: 76.]Neotheronia teiae Cameron, 1912: 183. Holotype cf , AUSTRALIA (BMNH) [examined]. [Synonymized by Townes etal., 1961:76.]Neotheronia antherae Cameron, 1912: 184. Holotype $, AUSTRALIA (BMNH) [examined]. [Synonymized by Townes etal., 1961:76.] Theronia steinbachneri Krieger; Morley, 1914: 39. [Misspelling.][Theronia viridicam Morley; Morley, 1914: 47. Misidentification, in part.]Theronia fumipennis Morley, 1914: 47. Holotype $, AUSTRALIA (BMNH) [examined]. Syn. n.Theronia daripennis Morley, 1914: 48. Holotype d" , AUSTRALIA (BMNH) [examined]. [Junior secondary homonym of Theronia daripennis Cameron, 1911a.] [Synonymized by Townes et al., 1961: 76.]Habropimpla anthereae (Cameron) Girault, 1925: 39. [Unjustified emendation.]Theronia (Theronia) steindachneri Krieger; Gupta, 1962: 34.Theronia (Theronia) fumipennis Morley; Gupta, 1962: 35. Fore wing length 7-12 mm. Mandible weakly narrowed, the teeth almost equal. Genae rounded behindeyes, in larger specimens somewhat inflated; posterior ocellus separated from eye by 1-1-1-2 times itsmaximum diameter, interocellar distance 0-7-0-8 times maximum ocellar diameter, the ocellar triangleslightly higher than broad. Occipital carina complete. Antenna stout, not distally tapered; scape apicallytruncate 45-50; flagellum with 37-39 segments, the penultimate distal segments slightly broader than long(Fig. 70). Scutellum carinate 0-2-0-5 of its length, generally most extensively carinate in the male.Propodeum with lateral longitudinal carinae complete; lateromedian carinae present anteriorly, rather farapart and divergent posteriorly, confluent with lateral remnants of posterior transverse carina (Fig. 69);propodeal spiracle elliptical; submetapleural carina very strongly broadened anteriorly, often bearingcoarse striae. Fore tibia unspecialized; fore tarsal segments 3 and 4 without stout bristles; hind femur withlong rather sparse hairs ventrally. Fore wing with cu-a slightly distal to base of Rs&M; areolet quite large,oblique. Gaster of female with tergite 1 1-4-1-5 times as long as broad, its sternite extending 0-2-0-3 of itslength (Fig. 66). Ovipositor projecting beyond apex of gaster by 1-2-1-3 times length of hind tibia, thelower valve simple.Orange, with antenna and hind tarsus mainly blackish. Wings weakly to moderately infumate. VARIATION. Specimens from northern Queensland have more strongly infumate wings and oftenslightly narrower genae. REMARKS. Gupta (1962) treated fumipennis as a separate species on account of the slightlynarrower genae, paler wings and less strongly ridged submetapleural carina. The latter characteris so variable as to be useless, whilst many specimens of a variety of species have slightly moreinfumate wings in the more equatorial region of their ranges. This leaves only the head shapeand here intermediates exist so I have chosen to regard fumipennis as a synonym. T. steindachneri appears to be a more northerly species than T. maculosa and is mostfrequently collected in Queensland and northern New South Wales. The more southern records(e.g. Australian Capital Territory) all refer to specimens collected in late summer. HOST RECORDS. Lymantriidae: Teia anartoides Walker (BMNH). Noctuidae: Pericyma cruegeri(Butler) (DPIQ). Psychidae: Hyalarcta huebneri (Westwood) (QM). Saturniidae: Antheraeaastrophela Walker (Cameron, 1912). MATERIAL EXAMINED Australia: 1 cf (holotype of teiae). New South Wales, Tenterfield, reared ex Teia anartoides (Froggatt)(BMNH); 1 $ (holotype of antherae). New South Wales, Richmond, reared ex Antheraea astrophela(Froggatt) (BMNH); 1 $ (holotype of fumipennis), Queensland, Townesville (Dodd) (BMNH); 2 $>(paratypes), Mackay (Turner) (BMNH); 1 d* (holotype of daripennis) , New South Wales, Sydney, 1856(Camming) (BMNH); 1 cf (paratype), Queensland, Mackay (Turner) (BMNH); 1 $ (paratype), Victoria(French) (BMNH). 316 I. D. GAULD Australia: 53 $, 17 cf , Australian Capital Territory, New South Wales, Queensland (Map 17) (AM;ANIC; BMNH; DAR; DPIQ; QM; QUM). Tribe RHYSSINI The Rhyssini is an apparent holophyletic group of pimplines characterized by the possession of acornute process on the last gastral tergite of the female and by the presence of transverse rugaeon the mesoscutum. The group includes some of the largest of ichneumonids and females areparticularly distinctive on account of their very long ovipositor. Rhyssines are all believed to beparasites of wood-boring holometabolous insect larvae, and recorded hosts include Siricidae,Xiphydriidae, Syntexidae and Cerambycidae. The hosts of the numerous tropical species arevirtually unknown. The tribe includes eight genera which are clearly divisible into two groups. These are themonobasic Rhyssa-group, characterized by possession of numerous primitive features such ashaving sternite 1 free, possessing a relatively unspecialized pronotum and not having a ridge onthe median trochantellus, and the Epirhyssa-group containing the genera Lytarmes, Epirhyssa,Cyrtorhyssa, Rhyssella, Megarhyssa, Myllenyxis and Triancyra. This latter group is charac-terized by several apomorphic features including having sternite 1 fused to the tergite, having aflange along the fore margin of the pronotum and possessing a ridge on the median trochantel-lus. The least specialized genus, Rhyssa, is Holarctic whilst many of the most highly specializedgenera of the Epirhyssa-group are tropical. Three genera are known to occur in Australia.Rhyssa and Megarhyssa were introduced to Victoria and Tasmania in an attempt to control Sirexinfestations in Pinus plantations; Epirhyssa is apparently a recently colonist of the north ofQueensland. EPIRHYSSA Cresson Epirhyssa Cresson, 1865: 39. Type-species: Epirhyssa speciosa Cresson, by subsequent designation, Viereck, 1914: 52. Rhyssonota Kriechbaumer, 1890: 489. Type-species: Rhyssonota tristis Kriechbaumer, by monotypy.Hierax Tosquinet, 1903: 255. Type-species: Hierax raptor Tosquinet, by monotypy. [Homonym of Hierax Vigors, 1826.]Sychnostigma Baltazar, 1961: 75. [Replacement name for Hierax Tosquinet.] Large species, fore wing length (9)15-17 mm; clypeus with a trace of a median apical tooth, laterallytuberculate; occipital carina complete, mediodorsally dipped. Pronotum mediodorsally with a deeptransverse groove separating off recurved anterior lip; epicnemial carina present; propodeum withoutdistinct carinae, spiracle elliptical. Female with claws simple; fore wing with 3r-m absent; hind wing withfirst abscissa of Cw t short, less than 0-2 times length of cu-a. Sternites 2-4 of female bearing tubercles nearanterior edge; tergite 2 with thyridia contiguous with anterior margin; tergites 2-4 smooth; ovipositorstraight, projecting beyond apex of gaster by about 4-0 times the length of hind tibia. REMARKS. Epirhyssa is a large tropicopolitan genus, the majority of species of which seem toinhabit lowland tropical rain forest. Kamath & Gupta (1972) recognized 39 Indo-Australianspecies but a number of unplaced specimens are to hand. These may represent additional speciesor the described species may encompass a greater range of morphological variation than Kamath& Gupta discerned. Recently collected material favours the latter suggestion. Two species occur in New Guinea and it is with slight hesitation that I consider the soleAustralian specimen conspecific with one of these. Epirhyssa biroi Mocsary comb. rev. Epirhyssa biroi Mocsary, 1905: 17. Holotype $ , PAPUA NEW GUINEA (TM) [examined]. Ahyborhyssa biroi (Mocsary) Enderlein, 1919: 152. Sychnostigma biroi (Mocsary) Townes et al., 1961: 87; Kamath & Gupta, 1972: 169. Face transversely striate, tending to vertically striate centrally adjacent to clypeus; fron transverselysemicircularly wrinkled, vertex sparsely punctate. Flagellum with 37 segments. Mesopleuron with subtegu-lar ridge convex, without a pronounced posterior concavity; epicnemial carina present, weak, its upper end ICHNEUMONIDAE OF AUSTRALIA 317 about on level of mesopleural pit, remote from anterior margin of pleuron; mesopleuron sparsely punctateon upper part, ventrally becoming more coarsely and closely obliquely punctate; epicnemium closelypunctate, the area behind fore coxae striate; metapleuron sparsely punctate. Propodeum dorsally smooth,evenly convex, at most with minute punctures. Tergites of gaster smooth and polished, the posteriortergites with fine punctures. Yellow with mandibles and antenna black; vertex black, the area of this colour extending to base ofantenna, to orbits and down onto the occiput above the foramen. Alitrunk black-marked on hind margin ofpronotum, fore margin of mesopleuron, mesoscutum anteriorly and centrally and metanotum narrowly;fore femur black-striped; hind tarsus brown, the distal segment black; gaster yellowish with anterior andoften posterior margins of tergites infuscate; ovipositor sheath blackish brown, distally yellowish brown.Wings infumate, apices very strongly so; pterostigma black. REMARKS. Only a single Australian specimen is at hand. It appears to be conspecific with theNew Guinea material studied but the combination of all specimens encompasses a greater rangeof variation than previously recognized in the genus. However, a considerable range of variationis known to occur in one of the few well-studied rhyssines, Rhyssa persuasoria (Spradbery &Ratkowsky, 1974), so similar extensive variation may be expected in related taxa. TheAustralian specimen differs from the Papuan examples in having more finely punctate meso- andmetapleurae and a weaker epicnemial carina. Amongst the Papuan specimens there is consider-able variation in the degree of infuscation of the wings. Some exceptionally small specimensfrom Wau have the wings hyaline except distally. Some specimens have well-developed blackmarks at the anterior and posterior margins of the gastral tergites whilst in one specimen thegaster is more or less entirely yellow. The hind tarsi are usually infuscate though almostimperceptibly so in some individuals. E. biroi is known in Australia only from tropical Queensland. It is widely distributedthroughout New Guinea and is not uncommon in lowland rain forest. It is here recorded for thefirst time from New Britain. HOST RECORDS. None. MATERIAL EXAMINED Papua New Guinea: 1 $ (holotype), Simbang on Huon Gulf (Biro) (TM). Australia: 1 $, Queensland, Shipton's Flat (1547'S; 14514'E), x.1980 (Cardale) (ANIC). New Britain:1 $, Cape Hoskins, vi.1970 (Stibick) (TC). Papua New Guinea: 1 $, Popondetta, x.1968 (Hassan)(BMNH); 1 $, Wau, 1200 m, vii-viii.1970 (Robinson) (TC); 1 ?, 1 cT, Wau, 1000 m, x.1979 (Gauld)(BMNH). MEGARHYSSA Ashmead Thalessa Holmgren, 1859: 122. Type-species: Ichneumon clavatus F. (= Ichneumon gigas Laxmann), by subsequent designation, Ashmead, 1900a: 53. [Homonym of Thalessa Adams, 1843.]Megarhyssa Ashmead, 1900ft: 368. [Replacement name for Thalessa Holmgren.]Megalorhyssa Schulz, 1906: 115. [Unjustified emendation.] Very large species, fore wing length 20-30 mm (Australian specimens); clypeus with a weak median apicaltooth, laterally tuberculate; occipital carina usually obsolescent centrally. Pronotum mediodorsally with adeep transverse groove cutting off anterior recurved lip; epicnemial carina present; propodeum withoutdistinct carinae, spiracle oval. Female with claws simple; fore wing with 3r-m present, enclosing large,triangular areolet; hind wing with first abscissa of Cu^ very short, less than 0-2 times length of cu-a.Sternites 2-4 of female bearing tubercles near anterior edge; tergite 2 with thyridia separated from anteriormargin; tergites 2-4 smooth; ovipositor straight, projecting beyond apex of gaster by about 6-0 times lengthof hind tibia. REMARKS. A moderate-sized Holarctic and Oriental genus with a single North Americanspecies, M. nortoni, introduced into Tasmania and Victoria to control Sirex infestations in Pinusradiata plantations (Taylor, 1976; Neumann & Minko, 1981). A second species, M. emargina-toria (Thunberg), was also introduced into Tasmania in 1964, but the females displayed littleinterest in the Sirex infested timber (Taylor, 1967) and the species has not become established.M. nortoni has become an important factor in controlling the population of Sirex in thesouth-east of Australia (Taylor, 1978). 318 I. D. GAULD Megarhyssa nortoni (Cresson) Rhyssa nortoni Cresson, 1864: 317. Holotype $, U.S.A. (PANS). Thalessa nortoni (Cresson) Cresson, 1870: 169. Thalessa Quebecensis Provancher, 1873: 447. Lectotype $, CANADA (UL), designated by Barren, 1975: 543. [Synonymized by Morley, 1913o: 16.] Megarhyssa nortonii (Cresson) Dalla Torre, 1901: 481. [Unjustified emendation.]Megarhyssa quebecensis (Provancher) Dalla Torre, 1901: 481.Megarhyssa nortoni (Cresson); Townes & Townes, 1960: 417. Face slightly coriaceous; frons transversely wrinkled; vertex punctate finely, centrally tending to striatebehind ocellar triangle. Mesopleuron ventrally coarsely punctate; metapleuron smooth, almost impunc-tate. Scutellum tranversely striate; propodeum dorsally smooth. Head black, facial and frontal orbits and gena yellow; alitrunk brown, notauli, margins of most sclerites,most of mesopleuron and axilla black; pronotum centrally, metapleuron in part, scutellum, postscutellumand hind corners of propodeum yellow; coxae black, remainder of legs yellowish or brownish; gasterbrown, tergites black-margined and with tergites 3-7 with a pair of yellow spots. Wings weakly, uniformlyinfuscate; pterostigma brown. REMARKS. Carlson (1979) suggests this North American native may be conspecific with thePalaearctic species M. gigas (Laxmann) but confirmation of this requires more detailed study.M. nortoni is believed to be the only established species of Megarhyssa in Australia. It differsfrom the European species, M. emarginatoria Thunberg, in having a slightly shorter ovipositor(that of emarginatoria is about 8 times the length of the hind tibia) and in colour. The genal orbitsof emarginatoria are black and the coxae orange. A considerable number of observations have been made on the biology of this species in NorthAmerica (e.g., Essig, 1926; Madden, 1968) whilst its mating behaviour in New Zealand isdocumented by Nuttall (1973). HOST RECORDS. Sirex species. In North America it is also known to parasitize other Siricinae suchas species of Xeris and Urocerus (Carlson, 1979). MATERIAL EXAMINED Australia: 5 $, 2 cT, Tasmania, Hobart, ii.1967 (Taylor) (BMNH). U.S.A., Canada: 10 $, 10 cf , variouslocalities including material compared with the holotype by H. K. Townes (BMNH; TC). RHYSSA Gravenhorst Rhyssa Gravenhorst, 1829: 260. Type-species: Ichneumon persuasorius L., by subsequent designation, Westwood, 1840: 59. Cryptocentrum Kirby, 1837: 260. Type-species: Cryptocentrum lineolatum Kirby, by monotypy.Pararhyssa Walsh, 1873: 109. Type-species: Ichneumon persuasorius L., by subsequent designation, Viereck, 1914: 111. Medium-sized to very large species, fore wing length 7-25 mm; clypeus with a median apical tooth, withoutlateral tubercles; occipital carina mediodorsally incomplete. Pronotum with a mediodorsal depression,without a deep transverse furrow separating anterior lip-like portion; epicnemial carina present; pro-podeum without distinct carinae, often transversely striate, spiracle oval. Female with claws simple; forewing with 3r-m present enclosing a triangular areolet; hind wing with first abscissa of Cu\ short, less than0-3 times length of cu-a. First sternite not fused with tergite; sternites 2-4 of female bearing tubercles nearcentre; tergite 2 with thyridia near to anterior margin; tergites 2-4 smooth; ovipositor straight, projectingbeyond apex of gaster by more than 4-0 times length of hind tibia. REMARKS. Rhyssa is a relatively small genus, species of which are common parasites of siricids inthe north temperate region. A species, R. persuasoria, was first introduced into the southernhemisphere in 1928-31 in an attempt to control Sirex noctilio in New Zealand (Miller & Clark,1935). The discovery of S. noctilio in a Pinus radiata plantation near Hobart, Tasmania in March1952 (Gilbert & Miller, 1952) prompted the Tasmanian Department of Agriculture to approachthe New Zealand authorities (DSIR and the Forest Research Institute) for shipments of R.persuasoria. In 1957 this rhyssine was liberated in Tasmania and by 1959 it seemed to havebecome established (Taylor, 1967). In 1961 Sirex was discovered in Victoria (Irvine, 1962) and ICHNEUMONIDAE OF AUSTRALIA 319 the following year the National Sirex Fund was established to promote an extensive programmeof research on Sirex and its parasites. In subsequent years many parasites were introduced intoAustralia, including eight species of Rhyssa. These were - R. alaskensis Ashmead (49 9 fromsouth western U.S.A.); R. amoena Gravenhorst (12 $ from Europe); R. crevieri (Provancher)(31 9 from eastern Canada); R. hoferi Rohwer (34 9 from Arizona and New Mexico); R.howdenorum Townes (95 9 from the south-eastern U.S.A.); R. jozana Matsumura (26 9 fromJapan); R. lineolata (Kirby) (30 9 from New Zealand and 30 9 from Canada and the U.S.A.);R. persuasoria (L.) (1622 9 from Europe, Turkey, Morocco, North America, Japan and India)(Taylor, 1976). Of these, only R. persuasoria seems to have become firmly established(Neumann & Minko, 1981). Rhyssa persuasoria (L. ) Ichneumon persuosorius L., 1758: 562. Lectotype cf, 'EUROPE' (LSL), designated by Fitton, 1978: 371 [examined]. Pimpla persuasoria (L.) F., 1804: 112.Rhyssa persuasoria (L.) Gravenhorst, 1829: 267; Townes & Townes, 1960: 406. Face somewhat coriaceous; frons smooth; vertex punctate, more closely and finely so laterally. Meso- andmetapleurae quite closely punctate; scutellum transversely striate; propodeum transversely weakly striate,with a median longitudinal depression. Male with sternites 2-4 undivided. Female blackish; orbits, marks on pronotum, subalar prominence, tegula, hind corner of mesopleuron,scutellum, postscutellum, hind part of metapleuron, propodeum posteriorly, hind margin of tergite 1 andtwo pairs of spots on hind margins of tergites 2+ , whitish; legs orange, hind leg distal to femur progressivelyinfuscate. Wings more or less hyaline; pterostigma blackish. Male much less extensively pale-marked thanfemale, black; face except clypeus, frontal and genal orbits partially, pronotum dorsally and double pairedspots on tergites 2+ , white. Legs and wings similar in colour to those of female. REMARKS. This introduced species is established in Pinus plantations in Tasmania and Victoria,but there is some danger that others of the various Rhyssa species introduced and presumed notto have become established may be overlooked or misidentified as R. persuasoria. This hashappened in New Zealand where R. lineolata (Kirby), an introduced Nearctic species, hasrecently been rediscovered and appears to have become established (Zondag & Nuttall, 1961).It is therefore worth emphasizing the following features. R. persuasoria is mainly black and has paired spots or sometimes bars on the hind margins oftergites 2-4, the temple is at least partly pale-marked, the antennae are unicolorous blackish andthe male has sternites 2-4 of the gaster entire. All other species introduced at various times toAustralia differ in one or more features. R. hoferi is basically brownish and, like R. howdenor-um, has continuous pale bands on the hind margins of tergites 2-4 of the gaster. R. alaskensis, R.crevieri, R. lineolata, R. amoena and R. jozana have sternites 2-4 of the male gaster longitudinal-ly divided into three and often have pale-banded antennae; R. alaskensis has the temple entirelyblack. A detailed account of the biology ofR. persuasoria is given by Morgan & Stewart (1966) whilstSpradbery (1970) details the host-finding behaviour. HOST RECORDS. In Australia, Sirex noctilio F. (Siricidae). In the north temperate region R.persuasoria is recorded as a parasite of a variety of other Siricinae. In the U.S.A. it is alsorecorded from Syntexidae (Carlson, 1979) whilst there are a number of early European recordsof this species attacking cerambycid larvae (Aubert, 1969). MATERIAL EXAMINED Australia: 10 $, 8 cf , Tasmania, (ANIC; BMNH). Europe: 187 $, 174 cf , various localities.'Europe': 1 9 (lectotype) (LSL). Subfamily XORIDINAE The Xoridinae is one of the smaller subfamilies containing, world-wide, four genera. Three ofthese are North Temperate, the fourth, Xorides, is quite large and cosmopolitan. It isrepresented in Australia by three species. 320 I. D. GAULD Zoogeography Of the four genera in the Xoridinae only Xorides occurs outside the North Temperate region.Species of this genus are not uncommon in temperate and tropical forests. Xorides is currentlydivided into 10 subgenera (Townes, 1969), four of which occur in the Oriental region. Only onesubgenus, X. (Cyanoxorides), extends to Australia and I have only seen species of this groupfrom Queensland and northern New South Wales. Biology Xoridines are parasites of wood-boring Coleoptera and Symphyta. Most usually larvae serve ashosts but pupae and even adults within their cocoon may be parasitized. The families Ceramby-cidae and Siricidae are the commonest hosts though a number of other coleopterous families arealso attacked. There are a number of studies on the biology of xoridines in Europe and NorthAmerica and that of Chrystal & Skinner (1931) serves as a typical example. These authorsobserved that the female xoridine oviposited through the bark into the host gallery. A single eggwas deposited on or near the cerambycid larva. Although it was never observed they consider ithighly likely that the host larva was stung and paralysed prior to oviposition. Larvae that hadbeen oviposited on were virtually immotile but alive, and even if the xoridine egg failed to hatchthe cerambycid larva remained unmoving for weeks before eventually dying. The elongate,fusiform egg hatched about eight days after oviposition and the first instar larva immediatelybegan attacking the host. It fed by burying its mandibles in the host's integument and sucking thebody fluids. Growth was rapid and after six or seven days the larva moulted. The second instarlarva resembled the first and lasted a further four or five days. At about this time the host larvawas found to be dead. The third larval instar was stouter with a less well-developed head. Thiscontinued to feed on the host for six to seven days before moulting. The fourth and final instarlarva consumed the entire host except for the head capsule and chitinous exoskeleton in five orsix days. It then remained motionless for several days before spinning a capacious cocoon. Thelarva was observed to overwinter in this cocoon and pupated in the spring of the following year.The duration of the pupal stage was short, with the adult emerging after nine or ten days. The head of the final instar xoridine larva is quite strongly sclerotized and the antennae arepapilliform, which is probably the plesiomorphic condition for ichneumonids. The epistomalarch is incomplete but the other sclerites are well formed. The mandibles are large withnumerous small accessory teeth. Short (1978) considered xoridine larvae to be structurallyprimitive. XORIDES subgenus CYANOXORIDES CameronCyanoxorides Cameron, 1903: 141. Type-species: Cyanoxorides brookei Cameron, by monotypy. Spiloxorides Cameron, 1903: 143. Type-species: Spiloxorides ruficeps Cameron, by monotypy.Xorides (Cyanoxorides) Cameron; Townes & Townes, 1960: 491. Flagellum of female with subapical angulation, the segment proximal to this angulation bearing peg-likeseta; flagellum of male filiform with conspicuous long fine pubescence; pronotum specialized with dorsalcarinae and strong epomia; epicnemial carina present; scuto-scutellar groove with a median carina;posterior transverse carina of mesosternum interrupted before mid coxae. Fore wing with cu-a proximal tobase of Rs&M. Gaster with tergite 2 with anterior corners impressed, with traces of diagonal furrows and acentral depression which may be raised to form a median longitudinal ridge posteriorly. REMARKS. The flagellum, pronotum, scuto-scutellar groove and gastral sculpture are distinctivefeatures enabling these insects to be easily recognized. Key to Australian species of Xorides (Cyanoxorides) 1 Alitrunk and gaster entirely orange-brown, only distal margin of tergite 7 pale-banded;flagellum unicolorous black; propodeum with carinae sharply raised and straight, clearlydelineating regular areae species 1 (p. 322) ICHNEUMONIDAE OF AUSTRALIA 321 Alitrunk and gaster partly to almost completely black; flagellum with a conspicuous white band in distal part ; propodeum with irregular and often indistinctly delineated carinae 2 2 Alitrunk black; gaster predominantly black, tergite 1 anteriorly reddish, tergites 7+ posteriorly white-banded ; propodeum with lateral areae reticulo-rugose australiensis (Szepligeti) (p. 321) Alitrunk black with large pale marks laterally; gaster black with tergite 1 anteriorly and ter-gites 1+ posteriorly pale-banded; propodeum with lateral areae punctate with scatteredrugae crudelis (Turner)(p. 321) Xorides (Cyanoxorides) australiensis (Szepligeti) Cyanoxorides australiensis Szepligeti, 1914: 420. Holotype $, AUSTRALIA (TM) [examined].Xorides (Cyanoxorides) australiensis (Szepligeti) Townes etal, 1961: 107. Fore wing length 11-13 mm. A shining species with head and much of alitrunk very finely punctate.Propodeum with carinae rather weak and irregular with numerous subsidiary carinulae radiating fromthem; lateral areae reticulo-rugose. Gaster almost matt, tergite 1 centrally puncto-coriaceous, posteriorlybecoming striato-coriaceous. Female black, face except centrally, upper orbits, genae in part and flagellar bands, white; tergite 1anteriorly reddish, tergites 7+ posteriorly white-banded. Legs golden yellow, distal 0-5 of hind femur,distal 0-4 of hind femur and hind tarsal segment 5 black, remaining hind tarsal segments white. Male similarto female except with smaller pale markings on head. REMARKS. X. australiensis is easily recognized by its colour pattern, but it is chromatically verysimilar to a mesostenine, Gotra gilberti (Turner). Both species can be found resting on treetrunks in rain forest and they are not easily distinguished in the field. However, X. australiensisand G. gilberti are quite unalike in detailed morphology (Gauld, 1984). The Gotra species can bedistinguished by its wedge-shaped areolet and simple female flagellum. Xorides lacks an areoletand the female has a subapical angulation in the flagellum. X. australiensis is apparently the mostcommon Australian xoridine and is widely distributed in Queensland and south into northernNew South Wales. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype of australiensis), 'New South Wales' (TM). Australia: 1 $, Queensland, Brisbane, xi.1976 (Boucek) (BMNH); 1 $, Brisbane, Long Pocket, 1977(Galloway} (BMNH); 1 cT, Bundaberg, Baldwin Swamp, ii.1973 (Frauca) (ANIC); 2 $, Mt Glorious,xii.1976 (Boucek) (BMNH); 1 $, Mt Tambourine, i.1936 (Hacker) (BMNH); 1 $, Mt Tambourine,x-xi.1978 (Galloway) (BMNH); 3 ?, Watalgan Rg., ix-xii.1972, v.1974 (Frauca) (ANIC). Xorides (Cyanoxorides) crudelis (Turner) Xylonomus crudelis Turner, 1919: 553. Holotype $, AUSTRALIA (BMNH) [examined].Xorides (Cyanoxorides) crudelis (Turner) Townes & Townes, 1960: 494. Fore wing length 12-14 mm. A weakly polished species with fine punctures laterally on gaster, dorsallyalmost impunctate. Propodeum with carinae rather weak and irregular with numerous subsidiary radiatingcarinulae; lateral areae punctate with scattered rugae. Gaster almost matt, tergite 1 centrally punctate,posteriorly becoming puncto-reticulate. Female black with flagellar band, face, upper orbits, genae in part, marks on vertex, much of pronotum,mesopleuron and propodeum laterally, tergite 1 anteriorly and posteriorly and tergites 2+ posteriorly,white-marked. Legs orange, all coxae white-marked; hind trochanteral segments, distal 0-2 of hind femur,distal 0-4 of hind tibia and hind tarsal segment 5 blackish, remainder of hind tarsus white. Male unknown. REMARKS. X. crudelis can be easily separated from australiensis by the striped black and whitegaster. It is a rather uncommon insect and possibly restricted to northern Queensland. HOST RECORDS. None. MATERIAL EXAMINED Australia: 1 $ (holotype), Queensland, Kuranda (Turner) (BMNH). Australia: 1 $ , Queensland, 6 km S. Daintree, in rain forest, x.1966 (Taylor) (ANIC); 1 $ , Gordonvale,1919 (BMNH). 322 I. D. GAULD Xorides (Cyanoxorides) species 1 Fore wing length 4 mm. A shining species with alitrunk dorsally rather coarsely punctate, laterally finelybut strongly punctate. Propodeum with carinae very strong, sharply raised, without carinulae, anddelineating regular areae; lateral areae smooth with few scattered rugae. Gaster weakly polished, tergite 1centrally puncto- reticulate, posteriorly tending to obsolescent striations. Male with head black, face, upper orbits and genae white-marked; flagellum unicolorous blackish;alitrunk and gaster orange-brown, only distal margin of tergite 7 whitish. Legs brownish orange, all tibiaedistally infuscate; hind femur with distal 0-1 blackish, hind basitarsus proximally and tarsal segment 5black, remainder of tarsus white. REMARKS. The single specimen known is a male. Considering the chromatic differences it isunlikely to be the unknown male ofcrudelis although structurally the species are fairly similar. Ihave avoided formally naming the species until a female is collected when proper comparisonmay be made with other Australian species. HOST RECORDS. None. MATERIAL EXAMINEDAustralia: 1 d", Queensland, Shipton's Flat (1547'S; 14514'E), x.1980 (Cardale) (ANIC). Subfamily ACAENITINAE The Acaenitinae is a moderately large subfamily, but often one of the most poorly representedin general collections. The majority of species are moderately large insects and the females havea conspicuous long ovipositor. They are usually brightly coloured and are amongst the moststriking of all ichneumonids. Acaenitines are parasites of wood-boring insect larvae, particularlythose of Coleoptera (Townes & Townes, 1962), and most species appear to be restricted toforest habitats. The subfamily is divided into two tribes, the Coleocentrini and the Acaenitini. The latter isrepresented in Australia by a single genus, Yezoceryx. Zoogeography The Coleocentrini comprises seven small genera most of which are restricted to the Palaearcticregion. Two, Coleocentrus and Mesoclistus, are represented in North America. The former alsooccurs in the mountains on the Palaearctic/Oriental interface (Townes, 1971b). The Acaenitinicomprises 18 genera and includes most species of the subfamily. All genera occur in the OldWorld although three, Arotes, Spilopteron and Yezoceryx, are also represented in NorthAmerica. The former extends as far south as about 15N but otherwise acaenitines are absentfrom the Neotropical region. The greatest diversity of species appears to be in everwet SouthEast Asia and one genus in particular, Yezoceryx, is very large. A recent series of shortcollecting trips to Brunei yielded over 20 species of Yezoceryx, most of which were representedby single specimens. Yezoceryx seems to be very closely related to the Afrotropical genusPhorotrophus; the former is probably paraphyletic with respect to the latter. Several species ofYezoceryx occur in New Guinea and in the present work six are believed to occur in Australia.Almost certainly the group has spread rather recently to Australia from Asia. Biology Virtually nothing is known about the biology of acaenitines except that they are parasites ofwood-boring holometabolous insect larvae. The most usual hosts appear to be Coleoptera, butthere are records from Sesiidae and Siricidae (Aubert, 1969). It is not known for certain whetheracaenitines are endo- or ecto-parasites. Townes (1969: 34) includes them in his 'section IF(internal parasites) together with such known endoparasitoids as Diplazontinae, Anomaloni-nae, Metopiinae, Collyriinae and Ichneumoninae. However, Aubert (1969) states that severalEuropean species are 'sans doubte ectoparasite larvaire'. Neither author actually says on whatevidence his statement is based and both Townes, in later works (e.g. 19716), and other authorswho have examined larval remains (e.g. Short, 1978) have conspicuously avoided stating ICHNEUMONIDAE OF AUSTRALIA 323 whether they are ecto- or endoparasites. Observations appear to have been made only on finalinstar larvae found with host remains (Baumann, 1927, 1933; Finlayson, 1970).Several features suggest acaenitines may be endoparasitic. 1. The ovipositor tip is rather unspecialized and shows little difference between closely relatedspecies. Those of many ectoparasites differ strikingly. 2. The ovaries usually contain relatively large numbers of rather small mature oocytes. Iwata(1960) gave an average value of 46 for Japanese Acaenitini and stated 'in comparison with itsbody size the mature egg of this group is generally small in size . . .'. Endoparasites generallyhave larger numbers of smaller eggs than similar-sized ectoparasites. 3. The mandibles of the final instar larvae are rather small and simple. Those of manyectoparasites are large and often denticulate. The general form of the cephalic capsule of thefinal instar larva is quite different from that of most ectoparasites (Short, 1978). This evidence is of course circumstantial and the controversy will not be settled until firmbiological observations are made. YEZOCER YX UchidaYezoceryx Uchida, 1928: 25. Type-species: Yezoceryx scutellaris Uchida, by original designation. Small to moderately large insects, fore wing length 6-15 mm; occipital carina complete across midline.Propodeum with some carinae but generally with area superomedia and area petiolaris confluent. Tarsalclaws with a small to large accessory tooth near apical tooth. Fore wing lacking 3r-m. First sternite with aweak to strong ventral swelling, the swelling without conspicuous hairs. Female subgenital plate large, inprofile triangular and extended posteriorly; ovipositor apex without a notch. REMARKS. Yezoceryx is a large genus with numerous species in the Indo-Australian region,several in the eastern Palaearctic and one in North America. Gauld (1984) suggested sevenspecies occur in Australia but re-examination of initial sorting and some progress in associatingsexes has reduced the number to five or possibly six, the putative sixth being only known frommales. Three of the species, Y. amaryllyx, Y. coelyx and species A, are closely related and form anendemic Australian species-complex. Y. apicipennis belongs to a monobasic species-groupwhilst the remaining two species, Y. tantalyx and Y. dinyx, appear to be closely related to someundescribed Indo-Papuan species. Key to Australian species of Yezoceryx 1 Females; ovipositor projecting far beyond apex of gaster 2 - Males 6 2 Posteroventral margin of hind trochantellus raised into a slightly protruding crest (Fig. 104); lower valve of ovipositor with two weak teeth, widely separated from each other and distantfrom the distal teeth (Fig. 103); clypeus with a strong transverse ridge above median tooth;tergites 1-7 yellow and blackish-banded tantalyx sp. n. (p. 326) - Posteroventral margin of hind trochantellus simple ; lower valve of ovipositor with teeth only at distal apex, or teeth not discernible; clypeus without a transverse ridge, or if ridge present it isabsent above median tooth or the median tooth projects from a ridge; tergites 1-7 usuallyunicolorous , if yellow and black banding is present then it is not on tergite 1 3 3 Ovipositor very long, projecting beyond apex of gaster by 2-6 times length of hind tibia; ovipositor sheath dark-coloured with a white subapical band; distal 0-3 of flagellum paleyellow, remainder brown; metapleuron closely and coarsely punctate dinyx sp. n. (p. 326) - Ovipositor projecting beyond apex of gaster by 2-2 or less times length of hind tibia; ovipositor sheath blackish with extreme distal apex pale; flagellum blackish, brown-marked on proximal2 or 3 and the extreme distal segments only; metapleuron sparsely punctate or closely andfinely punctate 4 4 Ovipositor projecting beyond apex of gaster by approximately 1-0 times length of hind tibia, the apex of the upper valve bearing weak teeth (Fig. 102); gaster black with white spots on hindmargins of tergites; mesopleuron with sparse scattered punctures; sternite 1 with a weak bluntprotuberance (Fig. 101) apicipennis (Turner) (p. 324) 324 I. D. GAULD - Ovipositor projecting beyond apex of gaster by approximately 2-0 times length of hind tibia, theapex simple; gaster almost entirely ochre-yellow; mesopleuron moderately closely punctate;sternite 1 with a strong pointed protuberance (Fig. 100) 5 5 Metapleuron finely and closely punctate, the punctures tending to be oval and form lines, thus creating impression of striations; mesoscutum yellow with a small black mark in middle ofcentral lobe; submetapleural carina strongly expanded anteriorly, but centrally without a small tooth (Fig. 105) amaryllyx sp. n.(p. 324) Metapleuron with scattered fine punctures; mesoscutum yellow with a large black mark on bothlateral and central lobes; submetapleural carina expanded anteriorly but with a small toothcentrally (Fig. 106) coelyx sp. n. (p. 325) 6 Tergite 1 of gaster black ; flagellum orange ; metapleuron with irregular wrinkles posteriorly apicipennis (Turner) (p. 324) Tergite 1 of gaster yellow or predominantly yellow; flagellum blackish; metapleuron posteriorlypunctate 7 7 Gaster yellow with tergites 2-7 black-marked near anterior end species A (p . 327) Gaster entirely yellow 8 8 Face laterally with punctures close, tending to form striae radiating to centre; mesoscutum with three black marks coelyx sp. n. (p. 325) Face laterally with punctures discrete, without any trace of striae; mesoscutum with a smallcentral black spot amaryllyx sp. n.(p. 324) Yezoceryx amaryllyx sp. n. (Fig. 105) Fore wing length 11-15 mm. Orbits moderately strongly, divergent ventrally; face with coarse closepunctures, without striae ventrolaterally; clypeus quite blunt, without a transverse ridge above the medianapical tooth; labrum slightly bilobate; mandibles quite stout, striate, the lower tooth about 1-3 times as longas the upper; malar space 0-70-0-85 times as long as basal mandibular width, genal sulcus moderatelystrongly impressed. Distal end of flagellum almost parallel-sided, apically rounded. Epicnemium andlower part of mesopleuron with close fine punctures that tend to be arranged in rows suggesting striae;metapleuron similar but with striae pronounced; submetapleural carina narrow, strongly expandedanteriorly (Fig. 105). Propodeum with anterior and posterior transverse carinae developed as lateralcrests, the area between them concentrically striate. Hind trochantellus unspecialized. Fore wing with2r-m generally more than 2-0 times length of M between 2r-m and 2m-cu, often with a slight angulationnear top; hind wing with distal abscissa of Cu\ slightly curved. Sternite 1 in midline reaching about 0-6 ofway to spiracles, with a strong, flattened very acutely pointed protuberance. Ovipositor projecting beyondapex of gaster by 2-0-2-1 times length of hind tibia, its apex with fine inconspicuous teeth. Female and male yellow-ochre, flagellum black with first two segments orange-marked, mesoscutumwith a small black mark centrally; proximal end of hind femur, distal apex of hind tibia and hind tarsusblackish; ovipositor sheath blackish, proximally and distally paler. Wings slightly infumate, fore wing withdistal apex strongly infumate. REMARKS. This species is very similar to Y. coelyx both in colour and structure. They are the onlytwo Australian acaenitines that are predominantly yellow. Y. amaryllyx can be separated fromcoelyx not only by the differences given in the key but also in being slightly larger and 'morerobust'. Y. amaryllyx is only known from Queensland. MATERIAL EXAMINED Holotype Cf , Australia: Queensland, Coolangatta, xii.1912 (QM). Paratypes. Australia: 1 $, Queensland, Brisbane, xi.1972 (Sedlacek) (TC); 1 $, 2 cf, Moggil, xi-xii.(TC); 1 $, Mt Cootha, xi-xii. (TC); 1 cf , Mt Glorious, i. (TC); 1 C?, Mt Nebo, ii. (TC). Yezoceryx apicipennis (Turner)(Figs 101, 102) Chorischizus apicipennis Turner, 1919: 36. Holotype $ , AUSTRALIA (BMNH) [examined].Yezoceryx apicipennis (Turner) Townes et al., 1961: 325. ICHNEUMONIDAE OF AUSTRALIA 325 Fore wing length 6-5-9-5 mm. Orbits barely divergent ventrally; face with coarse sparse punctures, with aslight trace of striation centrally; clypeus with traces of ridge only at lateral extremities, the area above thetooth flat; labrum slightly bilobed apically; mandibles stout, striate, with lower tooth about 1-5 times lengthof the upper; malar space 0-75 times as long as basal mandibular width; genal sulcus strongly impressed.Distal apex of flagellum slightly clavate. Epicnemium and lower half of mesopleuron with fine very sparsepunctures separated by more than twice their own diameters; metapleuron smooth, with very sparsescattered punctures, posteriorly somewhat wrinkled; submetapleural carina narrow, quite stronglybroadened anteriorly. Propodeum with anterior transverse carina complete, the posterior transversecarina centrally incomplete, the area between these carinae with a few transverse striae. Hind trochantellusunspecialized. Fore wing with 2r-m 1-2 times length of M between 2r-m and 2m-cu; hind wing with distalabscissa of Cu t curved down near distal end. Sternite 1 very short, in the mid-line reaching only 0-4 of wayto level of spiracle, with a rather weak, fairly blunt protuberance (Fig. 101). Ovipositor projecting beyondapex of gaster by 0-9-1-0 times length of hind tibia, its apex bearing about 7 weak teeth on lower valve andabout 5 on the upper (Fig. 102). Female head and alitrunk orange-brown, interocellar area partially, mesoscutum centrally, scutoscutel-lar groove , hind end of scutellum , anterior part of propodeum and posterior part of metapleuron blackish ;gaster black, tergites 1 and 2 with white spots near hind margin, tergites 3+ with a pair of such spots,tergites 5 and 6 additionally with whitish membranous area present; sternites black, margined with white,the anterior ones almost entirely pale; ovipositor sheath black, extreme distal end slightly paler. Legsorange, hind tarsus infuscate. Wings hyaline, the fore wing with an infumate patch only at distal apex. Malesimilar to female but with protuberance on sternite 1 more pronounced. Dorsal surface of head andalitrunk more extensively dark-marked, propodeum black; pronotum and mesopleuron brownish-marked. REMARKS. This distinctive species is easily recognizable by its short ovipositor and characteristiccolour pattern. The wings differ from those of other Australian acaenitines in having shorter,stouter and more sparsely distributed microtrichia, and the ovipositor is unusual in having teethon the upper valve. The male herein associated with the female differs strikingly in the colour pattern of the headand alitrunk though that of the gaster is similar. The association of sexes is based on locality,colour pattern of gaster, similarity in punctation of meso- and metapleurae and resemblance invenation. y. apicipennis is only known from the south-west of Australia; it is the only acaenitine knownto occur outside Queensland. MATERIAL EXAMINED Australia: 1 $ (holotype), Western Australia, Yallingup, ix.1913 (Turner) (BMNH). Australia: 1 C?, Western Australia, Yallingup, xii.1913-i.1914 (Turner) (BMNH); 1 $,70-75 km ENE.Norseman, xi.1978 (Houston) (WAM). Yezoceryx coelyx sp. n. (Fig. 106) Fore wing length 7-10 mm. Orbits moderately strongly divergent ventrally; face with coarse closepunctures, these punctures forming rows on lower part of face, giving appearance of radiating striaeventrolaterally; clypeus quite blunt, without a transverse ridge above the median apical tooth; labrum veryslightly bilobed or truncate; mandibles moderately slender with lower tooth barely longer than the upper;malar space 0-8-0-9 times as long as basal mandibular width, genal sulcus weakly impressed. Distal end offlagellum not clavate. Epicnemium and lower part of mesopleuron closely punctate, metapleuron withrather superficial moderately sparse fine punctures; submetapleural carina narrow, broadened anteriorlyand with a small tubercle just behind centre (Fig. 106). Propodeum with both anterior and posteriortransverse carinae present, incomplete, and not strongly rounded into crests, the area between themtransversely striate. Hind trochantellus unspecialized. Fore wing with 2r-m 1-2-1-5 times as long asabscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cu\ curved. Sternite 1 in midlinereaching about 0-7 of way to spiracle of tergite 1, with a strong, flattened, acutely pointed protuberance(Fig. 100). Ovipositor projecting beyond apex of gaster by 2 -0-2-1 times length of hind tibia, its apex withfine inconspicuous teeth. Female and male yellow, antenna black, scape and first flagellar segments yellowish-marked laterally,interocellar area, mesoscutum in three stripes, hind end of scutellum, proximal end of hind femur, distal 326 I. D. GAULD end of hind tibia and all hind tarsus, black. Wings almost hyaline, the apex of the fore wing infu-mate. REMARKS. This species is similar to Y. amaryllyx from which it can be separated by the charactersgiven in the key. It is only known from Queensland. MATERIAL EXAMINEDHolotype $, Australia: Queensland, O'Reilly's Guest House, on edge of rain forest, via Canungra, Paratypes. Australia: 5 $, Queensland, Mt Glorious, xii.1976 (Boucek) (BMNH); 1 $, Mt Glorious,vi.1977 (Miller) (BMNH); 1 $, 1 cf , Mt Glorious, x-xi. (TC); 3 cT, Mt Tambourine, xi. (TC); 1 $, MtTambourine, xi.1977 (Galloway} (BMNH); 1 $, Mt Tambourine, x-ix.1978 (Galloway) (BMNH). Yezoceryx dinyx sp. n. Fore wing length 6 mm. Orbits strongly divergent ventrally; face with coarse shallow punctures, withoutstriae; clypeus with transverse ridge present laterally, centrally absent so part above median tooth is flat;labrum concealed; mandibles weakly tapered, slightly striate, with lower tooth about 1-5 times as long asthe upper ; malar space 1 times as long as basal mandibular width , genal sulcus quite strong . Distal apex offlagellum slightly clavate. Epicnemium and lower part of mesopleuron with very coarse puncturesseparated by about their own diameter; metapleuron with very coarse scattered punctures; submetapleuralcarina moderately strong. Propodeum with both anterior and posterior transverse carinae present, widelyseparated, the area between them rugose. Hind trochantellus unspecialized. Fore wing with 2r-m equal tolength of M betweeen 2r-m and 2m-cu; hind wing with distal abscissa of Cu\ curved down near distal end.Sternite 1 reaching virtually to level of spiracle of tergite 1 , swollen centrally, the swelling with a sharp edgeanteriorly. Ovipositor projecting beyond apex of gaster by 2-6 times length of hind tibia, its apex smooth,without apparent teeth on either upper or lower valves. Female head blackish, face and postorbital marks yellowish; malar space brownish; antenna blackishbrown, scape ventrally yellowish, distal 9 or so flagellar segments bright yellow; alitrunk and gasterblackish brown, hind corners of meso- and metapleurae brownish, scutellum and posterior end ofpropodeum orange-brown; posterior margins of gastral segments narrowly yellow. Legs orange, hindtrochanter, trochantellus, tibia and tarsus infuscate. Ovipositor sheath blackish with a white subapicalband. Wings hyaline, the fore wing with distal apex and region adjacent to stigma infumate. Maleunknown. REMARKS. Y. dinyx is a very distinctive little species on account of the long ovipositor, thewhite-banded ovipositor sheath and the rather extensive area of coarse punctures, especially onthe lateral parts of the alitrunk. The first segment of the gaster is more slender than that of otherAustralian acaenitines and the sternite correspondingly more elongate; the subgenital plate issomewhat shorter than that of other Yezoceryx species. MATERIAL EXAMINEDHolotype $, Australia: Queensland; Mt Spec, 800 m, iii.1964 (Common & Upton) (ANIC). Yezoceryx tantalyxsp. n. (Figs 103, 104) Fore wing length 7-12 mm. Orbits strongly divergent ventrally; face coarsely punctate, ventrolaterallyarranged in rows giving the effect of striae; clypeus with a strong transverse ridge above the median tooth;labrum almost evenly rounded apically; mandibles moderately strongly tapered, striate, with lower toothabout 1-3 times length of the upper; malar space 0-9-1-0 times as long as basal mandibular width, with astrong genal sulcus. Distal apex of flagellum slightly clavate. Epicnemium and mesopleuron with very closefine punctures separated by about their own diameter; metapleuron fairly uniformly sparsely punctate;submetapleural carina narrow. Propodeum with anterior transverse carina more or less complete, theposterior transverse carina discontinuous but discernible. Hind trochantellus with posteroventral marginraised into a slightly protruding crest (Fig. 104). Fore wing with 2r-m about equal to or shorter thanabscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cu t exceptionally sinuous.Sternite 1 short, reaching about 0-5 of way to spiracle, somewhat swollen, the swelling convexly notangularly rounded. Ovipositor projecting beyond apex of gaster by 2-7-3-1 times length of hind tibia, theovipositor apex with lower valve bearing two weak teeth, widely separated from each other and distantfrom the distal teeth, upper valve simple (Fig. 103). ICHNEUMONIDAE OF AUSTRALIA 327 Female head yellow, dorsally black; antenna black, scape ventrally yellow, distal apex of flagellumpallid. Mesoscutum, much of mesopleuron, anterior part of propodeum and metapleuron near insertion ofhind coxa brownish, black; pronotum, scutellum, subalar prominence, mesopleuron centrally andventrally, remainder of propodeum and metapleuron, yellow. Gaster blackish brown with posterior 0.3 oftergite 1, posterior 0-4 of tergites 2-4 and lateral extremities of other tergites yellow; sternites andsubgenital plate predominantly yellowish. Legs yellow-brown, anterior two pairs of coxae yellow, hindcoxa, tibia and tarsus infuscate; ovipositor sheath blackish. Wings hyaline, fore wing with apex and areaadjacent to pterostigma infumate. Male unknown. VARIATION. Two specimens are large (fore wing length ca 12 mm) whilst three from the Townescollection are notably smaller (fore wing length ca 1 mm). The smaller specimens are slightlymore extensively yellow-marked, especially on the mesoscutum and mesopleuron. However, instructure they closely resemble the larger specimens. REMARKS. Y. tantalyx is probably the most distinctive Australian Yezoceryx on account of themodified hind trochantellus, but it also has a number of other unusual features including thepositioning of the ovipositor teeth, the sinuous I A in the hind wing and the clypeal ridge.This species is only known from Mt Tambourine, Queensland. MATERIAL EXAMINED Holotype $, Australia: Queensland, Mt Tambourine (Davidson) (QM). Paratypes. Australia: Queensland, 1 $, Mt Tambourine, i.1936 (Hacker) (BMNH); 3 $, Mt Tam-bourine, xii. (TC). Yezoceryx species A Fore wing length 10-13 mm. Orbits rather weakly divergent ventrally; face coarsely punctate, tendingtowards being striate latero ventrally; clypeus with margin blunt, without a ridge above median apicaltooth; labrum weakly bilobed; mandibles moderately stout, with lower tooth about 1-3 times length of theupper; malar space 0-7 times as long as basal mandibular width, the genal sulcus weak. Distal end offlagellum not clavate. Epicnemium and lower part of mesopleuron closely punctate, metapleuron moresparsely so; submetapleural carina moderately wide, broadened anteriorly and with a small thickeningcentrally. Propodeum with anterior and posterior transverse carinae represented as crests, not continuous,the area between them transversely striate. Hind trochantellus unspecialized. Fore wing with 2r-m morethan 2-0 times length of abscissa of M between 2r-m and 2m-cu; hind wing with distal abscissa of Cuicurved. Sternite 1 in midline reaching 0-7 of way to spiracle, with a strong flattened, acutely pointedprotuberance. Female yellow, with most of flagellum, scape dorsally, vertex of head, three mesoscutal stripes,scuto-scutellar groove, anterior edges of tergites 2-7 (either as spots or bars) black. Legs yellow, hindfemur proximally black, hind tarsus black, hind tibia infuscate. Male unknown. REMARKS. This group of specimens is structurally rather similar to Y. coelyx, although it is bothlarger and has a slightly different colour pattern. It may be merely an extreme 'form' of coelyxbut for the present it was thought better to treat it as being tentatively distinct until females canbe collected. Males of most species of Yezoceryx are structurally very similar and difficult toseparate. MATERIAL EXAMINEDAustralia: 3 d", Queensland, Stanthorpe, 700 m, i. (TC). Subfamily LYCORININAE This small subfamily is taxonomically rather isolated. Townes (1970) placed it close to theBanchinae, but the characters supporting this association are plesiomorphic features such as thepossession of an unspecialized petiole with spiracles placed antecentrally. The egg is mostcurious, being leech-like and sinuous (Iwata, 1958), quite unlike that of any other ichneumonid.The structure of the cephalic capsule of the final instar larva (Finlayson, 1976; Short, 1978) isunlike that of the Banchinae as the mandibular teeth are denticulate, like those of many speciesof Tryphoninae and Phygadeuontinae. Unlike these ectoparasitoids the Lycorininae lack adistinct labral sclerite. 328 I. D. GAULD Some species have been reared from various Microlepidoptera, most usually pyralids inleaf-rolls (Finlayson, 1976), but no details of the biology of the group are known, not evenwhether the larva is endo- or ectoparasitic. Were such observations to be made they would beinvaluable jn helping to place this enigmatic subfamily. Townes (1970) recognized three closely related genera, but Gauld (1984) treated all species asbelonging to a single genus, Lycorina. LYCORINA Holmgren Lycorina Holmgren, 1859: 126. Type-species: Lycorina triangulifera Holmgren, by monotypy.Toxophoroides Cresson, 1873: 406. Type-species: Lycorinal apicalis Cresson, by original designation.Chlorolycorina Cushman, 1920: 9. Type-species: Glypta scitula Cresson, by original designation.Gonioglyphus Seyrig, 1932: 22. Type-species: Lycorina (Gonioglyphus) fid Seyrig, by monotypy. Small to medium-sized insects, fore wing length 3-7 mm; clypeus separated from face by groove, withmargin slightly concave medially; mandible bidentate; malar space with an impressed groove running fromeye to mandible; frons simple. Occipital carina complete or narrowly interrupted mediodorsally; genalcarina joining hypostomal carina above base of mandible. Flagellum cylindrical, unspecialized. Notaulivestigial; sternaulus absent; posterior transverse carina of mesosternum absent. Propodeum with dorso-lateral corner projecting anteriorly and engaging a small hook on the metanotum; propodeum with orwithout carinae. Fore tibia without a tooth on apical margin; tarsal claws strongly pectinate. Fore wing with3r-m absent; 2r-m longer than abscissa of M between 2r-m and 2m-cu\ 2m-cu with one bulla; pterostigmabroadly triangular; marginal cell moderately long. Hind wing with first abscissa of Rs longer than r-m;distal abscissa of Cu\ present or absent. Gaster with tergite 1 broad, dorsally convex with spiracles beforethe centre, sternite not reaching to spiracles; tergites 2-4 with strongly impressed, striate grooves definingtriangular central areas. Female with subgenital plate large, triangular, centrally membranous; ovipositorprojecting beyond apex of gaster by 1-15-1 -70 times length of hind tibia, its apex with a strong nodus. Lycorina is one of the more distinctive ichneumonid taxa on account of the grooves on tergites2-4 and their unique propodeal/metathoracic 'catch'. It is a moderate-sized genus with speciesoccurring in almost all zoogeographical regions. Three or perhaps four species occur inAustralia. These represent two species-groups, the turneri-group with a carinate propodeumand long tarsal pectinae, and the canberrae-group with a smooth propodeum and short tarsalpectinae. The latter group resemble the Neotropical species (formerly considered a separategenus, Toxophoroides) in the lack of propodeal carinae, but it is not possible to decide whetherthis results from evolutionary convergence or is indicative of phylogenetic affinity. Theturneri-group is tropical, and resembles some Indo-Papuan species. Key to species of Lycorina occurring in Australia 1 Propodeum with distinct carinae, at least with posterior transverse one more or less complete (Figs 107, 108); hind tarsal claw with long pectinae (Fig. 109); 9 with anterior coxae pale 2 Propodeum without discernible carinae, rarely with a trace of lateromedian ones anteriorly;hind tarsal claw with short sparse pectinae (Fig. 110);$ with all coxae black 3 2 Scutellum slightly longer than broad anteriorly, centrally sparsely punctate (Fig. 108) ; mesoscu- tum black with two yellow longitudinal stripes; upper end of epomia raised into strong crest; wings hyaline splendidula sp. n. (p. 329) - Scutellum transverse, slightly broader anteriorly than long, centrally smooth, impunctate(Fig. 107); mesoscutum unicolorous orange-brown; upper end of epomia not raised into astrong crest; wings infumate turner! sp. n. (p- 330) 3 Alitrunk entirely blue-black ; head orange , interocellar area black species 1 (p . 330) Alitrunk predominantly reddish brown ; head dorsally extensively blackish canberrae sp . n . (p . 328) Lycorina canberrae sp. n. (Fig. 110) Fore wing length 5-7 mm. Face moderately strongly convex with a weak median tubercle just belowantennal insertion, genal carina joining hypostomal carina well away from base of mandible; occipitalcarina mediodorsally obsolescent. Pronotum with upper end of epomia raised into a small crest;mesoscutum moderately long, in dorsal view 1-2-1-3 times as long as broad, without discernible notauli, ICHNEUMONIDAE OF AUSTRALIA 329 smooth and shining, virtually impunctate except for minute setiferous punctures anteriorly and anter-olaterally; scutellum distinctly broader anteriorly than long, with isolated minute punctures, lateral carinaewell developed, reaching about 0-7 of length of scutellum; mesopleuron smooth and polished with isolatedfine punctures, posteromedially moderately impressed below speculum; metapleuron polished, with finesparse punctures, juxtacoxal carina usually complete. Propodeum convex, without carinae, or at very mostwith only vestiges of lateromedian longitudinal carinae present anteriorly. Hind tarsal claws with short,sparse pectinae (Fig. 110). Gaster with triangular areas defined on tergites 2-4, those on 1 and 5 notcomplete, the grooves defining these areas all strongly trans-striate; laterotergites 2 and 3 broad, foldedunder, with strongly pigmented sclerotized central areas; ovipositor projecting beyond apex of gaster by1-60-1-65 times length of hind tibia; posterior margin of tergite 8 mediodorsally produced to form a bluntprocess that projects almost as far as cercus. Female with head blackish, face, mouthparts and marks above eyes yellow, flagellum distally brown;alitrunk orange-brown, margin of pronotum, propleuron and ventral part of metapleuron blackish; gasterblack with posterior and often lateral margins of tergites whitish, tergite 1 usually extensively white withcentral black spot. Legs black, fore and mid femora and tibiae striped longitudinally with pale yellow, tarsimore brownish; ovipositor sheath black, distally pale. Wings infumate, pterostigma brown. Male similar tofemale but with propodeum extensively black, genae ventrally and fore and mid coxae and trochanterswhite-marked. REMARKS. This species is easily distinguished from L. turneri and L. spendidula by the lack ofpropodeal carinae and indistinct pectinae on the tarsal claws. It is morphologically more or lessidentical to 'species 1'. This is the most widely distributed Australian species, occurring in Australian CapitalTerritory, New South Wales, Queensland, Victoria and Western Australia. HOST RECORDS. None. MATERIAL EXAMINED Holotype $, Australia: Australian Capital Territory, Canberra, Black Mt, ix.1981 (Gauld) (ANIC). Paratypes. Australia: 1 $ , New South Wales, Mt Keira, on edge of rain forest, ii.1983 (Gauld} (BMNH);1 <J>, Queensland, Mapleton, x.1967 (BMNH); 1 $, Victoria, no further locality data, 1912 (French}(BMNH); 1 cT, Western Australia, 50 km S. Coolgardie, x.1958 (Riek) (ANIC). Lycorina splendidula sp. n. (Fig. 108) Fore wing length 3-5 mm. Face moderately convex, without a distinct median tubercle; genal carinameeting hypostomal carina just before base of mandible; occipital carina complete. Pronotum with upperend of epomia raised into a strong tooth; mesoscutum rather long, in dorsal view 1-3-1-4 times as long asbroad, with weak notaular indentations and some punctures, in larger specimens the punctures runningtogether to form transverse striae; scutellum slightly longer than broad anteriorly, punctate, with stronglateral carinae; mesopleuron with isolated punctures, posteromedially strongly impressed below specu-lum; metapleuron almost impunctate with juxtacoxal carina almost complete. Propodeum with small butdistinctly defined area superomedia; posterior transverse carina complete (Fig. 108). Hind tarsal clawswith long, strong pectinae. Gaster with triangular area defined on tergites 1-5, the grooves transverselystriate; laterotergites 2 and 3 narrow, folded under, weakly sclerotized and inconspicuous; ovipositorprojecting beyond apex of gaster by 1-15-1-25 times length of hind tibia; posterior margin of tergite 8dorsally truncate forming a more or less straight line between cerci. Female yellow; dorsal surface of antenna, frons centrally, vertex, occiput, mesocutum in threelongitudinal vittae, ventral part of mesopleuron and extreme anterior corner of propodeum, black;pronotum centrally, tergites 6+ of gaster, subgenital plate, ovipositor sheath except distal apex, brownish.Wings hyaline; pterostigma brown. Male unknown. REMARKS. This distinctive species is easily recognized by its extensively yellow colour pattern. Itis more slender than other Australian lycorinines and has a more strongly sculptured alitrunkand a more rigid gaster with slightly coarse sculpture. It appears to be closely related to L. turneriwhich it resembles in the structure of the claws and carination of the propodeum. L. splendidulahas a smaller, more completely denned area superomedia than L. turneri and differs also inhaving the genal carina joining the hypostomal carina slightly away from the base of themandible. 330 I. D. GAULD This species occurs in tropical Queensland. MATERIAL EXAMINED Holotype $, Australia: Queensland, Cooktown, Mt Cook N.P. (1529'S; 14516'E), x.1980 (Cardale)(ANIC). Paratype. Australia: 1 $, Queensland, Mackay, 1909 (Turner) (BMNH). Lycorina turner! sp. n. (Figs 107, 109) Fore wing length 4-6 mm. Face moderately convex, without a distinct median tubercle; genal carinameeting hypostomal carina at base of mandible; occipital carina complete. Pronotum with upper end ofepomia slightly raised to form a weak flange; mesoscutum moderately long, in dorsal view 1-1-1-2 times aslong as broad, without discernible notauli, smooth and shining, impunctate; scutellum distinctly broaderanteriorly than long, with isolated minute punctures, lateral carinae moderately well developed, reachingto near centre; mesopleuron smooth and polished with isolated fine inconspicuous punctures, post-eromedially only slightly impressed below speculum; metapleuron smooth with isolated fine punctures,with juxtacoxal carina almost complete. Propodeum with a weakly indicated transverse area superomedia;posterior transverse carina usually complete but other carina may be absent, weakly indicated or in largerspecimens distinct (Fig. 107). Hind tarsal claws with long strong pectinae (Fig. 109). Gaster with triangulararea defined on tergites 2-4, that of tergites 1 and 5 indistinct, the grooves defining these areas not clearlytrans-striate; laterotergites 2 and 3 moderately wide, folded under, weakly sclerotized, inconspicuous;ovipositor projecting beyond apex of gaster by 1-55-1-70 times length of hind tibia; posterior margin oftergite 8 dorsally truncate, forming a more or less straight line between cerci. Female orange-brown; head black with mouthparts, face except orbits and marks on vertex pale yellow.Gaster with tergites 1 and 2 laterally yellow, 1-5 centrally black, 3+ black also laterally and often alsoposteriorly except for extreme posterolateral margin which is yellow-marked; ovipositor sheath darkbrown, its distal apex paler. Hind leg blackish, coxa dorsally brownish orange. Wings strongly infumate,pterostigma dark brown. Male similar to female but with fore and mid coxae pallid, almost white. REMARKS. The propodeal carination of this species is rather variable but at least the posteriortransverse one is distinct, enabling it to be separated from L. canberme. It differs from L.splendidula most obviously in colour pattern but the two species also differ in many subtlemorphological details as can be seen by contrasting the descriptions. This species is named in honour of the late R. E. Turner who collected a short series inKurandain 1913. It occurs in tropical Queensland. MATERIAL EXAMINED Holotype $, Australia: Queensland, Gap Ck., 5 km ESE. Mt Finnigan (1550'S; 14520E), v.1981(Naumann) (ANIC). Paratypes. Australia: 1 cf , Queensland, same data as holotype (ANIC); 1 9, Mossman, i. (TC); 2 $,1 cf , Kuranda, 350 m, v-vi.1913 (Turner) (BMNH). Lycorina species 1 Morphologically this is virtually identical to L. canberme though the pleural punctures in theTasmanian specimens are sparser. It differs strikingly in colour pattern. Female head orange, interocellar area and proximal part of antenna black, flagellum distally brownish;alitrunk, legs and gaster black, fore and mid femora and tibiae white-striped, segments of gaster broadlyinfumate, pterostigma dark brown. Male similar to female but with fore and mid coxae and hind tibiawhite-marked. REMARKS. The characteristic colour pattern occurs in a variety of small temperate AustralianHymenoptera. However, as this group of specimens are so similar to L. canberrae I am scepticalthat they warrant specific recognition. They do not have a clearly separate range and I have notformally described them pending further study. MATERIAL EXAMINED Australia: 1 , New South Wales, Araluen Hill, 100 m, x.1981 (Gauld) (BMNH); 1 $, 1 O", Tasmania,Coles Bay, ii-iii.1981 (TC). ICHNEUMONIDAE OF AUSTRALIA 331 Acknowledgements I thank the Australian Government who, through the agency of ABRS, provided me with agrant to study in Australia. I am particularly grateful to the following Australian entomologistsfor their help and encouragement in a variety of ways: Dr Ian Naumann and Miss Jo Cardale inCanberra; Dr Ian Galloway, Dr Ted Dahms, Dr Geoff Monteith and Ms Margaret Schneider inBrisbane; Mr Geoff Holloway in Sydney; Mr Dennis Farrugia, Dr Artur Neboiss and Dr KenWalker in Melbourne; Ms Margaret Williams in Hobart and Dr Terry Houston in Perth. Thewhole insect figures were drawn by Jonathan Carter and Andrew Atkins. Dr Ian Common andMr Ted Edwards gave invaluable help concerning Australian Lepidoptera. I am especiallygrateful to Dr Henry Townes for his comments at various stages in the work and to my colleaguesat the BMNH, particularly Dr Mike Fitton, Mr Mick Day, Dr Andy Austin and Dr GadenRobinson for their helpful discussion of numerous points. Finally I would like to thank DarrenGauld for drawing the maps and Ms Pam Mitchell for collating data, typing the manuscript andchecking references. References Ashmead, W. H. 1900a. Classification of the ichneumon-flies, or the superfamily Ichneumonoidea. Proceedings of the United States National Museum 23: 1-220. 1900&. Some changes in generic names in Hymenoptera. Canadian Entomologist 32: 368. Askew, R. R. 1971. Parasitic Insects, xvii+316 pp. London. Aubert, J. F. 1969. Les ichneumonides ouest palearctiques etleurs holes 1. 299 pp. Paris. Audley-Charles, M. G., Hurley, A. M. & Smith, A. G. 1981. Continental movements in the Mesozoic and Cenozoic. In Whitmore, T. C. (ed.), Wallace's line and plate tectonics. 91 pp. Oxford.Baltazar, C. R. 1961. The Philippine Pimplini, Poemeniini, Rhyssini and Xoridini (Hymenoptera, Ichneumonidae, Pimplinae). Monographs of the National Institute of Science and Technology, Manila 7: 1-130. 1964. Eriostethus Morley, and a new polysphinctine genus. Pacific Insects 6: 388-401. Barron, J. R. 1975. Provancher's collection of insects, particularly those of Hymenoptera, and a study of the types of his species of Ichneumonidae. Naturaliste canadien 102: 387-591 .Baumann, C. 1927. Beobachtungen iiber die Metamorphose der Schlupfwespen Coleocentrus excitator Poda und Ephialtes manifestator L. Zeitschrift fur Morphologie und Okologie der Tiere 9: 313-332.- 1933. Zur Kenntnis der Metamorphose von Ephialtes manifestator, Coleocentrus excitator und Echthrus reluctator. Zoologischer Anzeiger 102: 143-155.Betrem, J. G. 1932. Eine neue aus Manga-friichten geziichtete Pimpline (Fam. Ichneumonidae). Treubia 14: 21-28.Bosch, R. van den 1964. Encapsulation of the eggs of Bathyplectes curculionis (Thomson) (Hymenoptera; Ichneumonidae) in larvae of Hypera brunneipennis (Boheman) and Hypera postica (Gyllenhal) (Coleoptera; Curculionidae). Journal of Insect Pathology 6: 343-367. 1978. The pesticide conspiracy. 226 pp. New York. Brulle, M. A. 1846. In Lepeletier de Saint-Fargeau, A. L. M., Histoire naturelle des insectes 4. Hymenop- teres. viii+680 pp. Paris. Bruzzese, E. 1982. Observations on the biology of Pseudopimpla pygidiator Seyrig (Hym., Ichneumoni-dae), a parasite of the blackberry stem boring sawfly Hartigia albomaculatus (Stein) (Hym. , Cephidae). Entomologist's Monthly Magazine 118: 249-252.Cameron, P. 1898. Notes on a collection of Hymenoptera from Greymouth, New Zealand, with descriptions of new species. Memoirs and Proceedings of the Manchester Literary and Philosophical Society 42(1): 1-53. - 1899. Hymenoptera Orientalia: or contributions to a knowledge of the Hymenoptera of the Oriental zoological region, part 8. The Hymenoptera of the Khasia Hills, first paper. Memoirs and Proceedings of the Manchester Literary and Philosophical Society 43(1): 1-220.1902. On the Hymenoptera collected by Mr Robert Shelford at Sarawak, and on the Hymenoptera of the Sarawak Museum. Journal of the Straits Branch of the Royal Asiatic Society 37: 29-140. 1903. Description of new genera and species of Hymenoptera taken by Mr Robert Shelford at Sarawak, Borneo. Journal of the Straits Branch of the Royal Asiatic Society 39: 89-181. 332 I. D. GAULD 1905. On some new genera and species of Hymenoptera collected by the Revd. J. A. O'Neill, S.J.chiefly at Dunbrody, Cape Colony. Records of the Albany Museum 1: 245-265. - 1907. Hymenoptera of the Dutch expedition to New Guinea in 1904 and 1905. Part 3. ParasiticHymenoptera. Tijdschrift voor Entomologie 50: 27-57. 1911fl. Hymenoptera (except Anthophila and Formicidae). Nova Guinea 9: 185-248. - 19116. On a collection of parasitic Hymenoptera (chiefly bred) made by W. W. Froggatt, in NewSouth Wales, with descriptions of new genera and species. Part 1. Proceedings of the Linnean Society ofNew South Wales 36: 333-346. 1912. On a collection of parasitic Hymenoptera (chiefly bred) made by W. W. Froggatt, in New South Wales, with descriptions of new genera and species. Part 3. Proceedings of the Linnean Society of New South Wales 37: 172-216.Capener, A. L. 1938. Notes on a spider parasite, believed to be Colpomeria quadrisculpta Gr. Journal of the Society for British Entomology 1: 209-210.Carlson, R. W. 1979. Family Ichneumonidae. In Krombein, K. V., Hurd, P. D., Smith, D. R. & Burks, B. D., Catalog of the Hymenoptera in America north of Mexico 1: 1-1198. Washington.Carton, Y. 1973. Biologic de Pimpla instigator F. 1793 [Ichneumonidae, Pimplinae]. Entomophaga 18: 25-39.Chadwick, C. E. & Nikitin, M. 1. 1976. Records of parasitism in the families Ichneumonidae, Braconidae and Aulacidae. Journal of the Entomological Society of Australia (N.S. W.) 9: 28-38.Charig, A. 1981. Cladistics: a different point of view. Biologist 28: 19-20.Cheesman, L. E. 1936. Hymenoptera of the New Hebrides and Banks Islands. Transactions of the Royal Entomological Society of London 85: 169-195.Chrystal, R. N. & Skinner, E. R. 1931. Studies in the biology of Xylonomus brachylabris Kr. and X. irrigator F. , parasites of the larch longhorn beetle, Tetropium gabrieli Weise. Forestry 5: 21-33. 1932. Studies on the biology of the woodwasp Xiphydria prolongata Geoffr. and its parasite Thalessa curvipes Grav. Scottish Forestry Journal 46: 36-51.Clausen, C. P. 1940. Entomophagous Insects. 688 pp. New York.Cole, L. R. 1967. A study of the life-cycles and hosts of some Ichneumonidae attacking pupae of the green oak leaf roller moth, Totrix viridana (L.) (Lepidoptera: Tortricidae). Transactions of the Royal Entomological Society of London 119: 267-281.Coleman, E. 1928. Pollination of an Australian orchid by the male ichneumonid Lissopimpla semipunctata Kirby. Transactions of the Entomological Society of London 76: 533-539.Common, I. F. B. 1954. A study of the ecology of the adult Bogong moth Agrotis infusa (Boisd.) (Lep., Noctuidae), with special reference to its behaviour during migration and aestivation. Australian Journal of Zoology 2: 223-263.Costa, A. 1864. Descrizione di taluni insetti stranieri all'Europa. Annali Museo Zoologica della Universite Napoli 2 (1862): 66-155.Cresson, E. T. 1864. Descriptions of North American Hymenoptera in the collection of the Entomological Society of Philadelphia. Proceedings of the Entomological Society of Philadalphia 3: 257-321.- 1865. On the Hymenoptera of Cuba. Proceedings of the Entomological Society of Philadelphia 4: 1-200. 1870. Descriptions of new species belonging to the subfamily Pimplariae found in America north of Mexico. Transactions of the American Entomological Society 3: 143-172. 1873. Descriptions of Mexican Ichneumonidae. Proceedings of the Academy of Natural Sciences of Philadelphia 25: 374-413. Cushman, R. A. 1920. The North American ichneumon-flies of the tribes Lycorini, Polysphinctini andTheroniini. Proceedings of the United States National Museum 58: 7^48. 1926. Some types of parasitism among the Ichneumonidae. Proceedings of the Entomological Societyof Washington 28: 25-51. 1933. Descriptions of new ichneumon-flies with taxonomic notes. Proceedings of the United States National Museum 82: 1-16.Dalla Torre, C. G. 1901. Catalogus Hymenopterorum 3. vii+1141 pp. Leipzig.Danks, H. V. 1971. Biology of some stem-nesting aculeate Hymenoptera. Transactions of the Royal Entomological Society of London 122: 323-399.Dumbleton, L. J. 1932. The apple leaf-roller (Tortrix postvittana Walk.). New Zealand Journal of Science and Technology 14: 83-92.- 1940. Tortrix postvittana Walk, and its parasites in Australia. New Zealand Journal of Science and Technology (A)21: 322a-336a.Enderlein, G. 1919. Beitrage zur Kenntnis aussereuropaischer Ichneumoniden 4. Einige neue Pimpliden. ICHNEUMONIDAE OF AUSTRALIA 333 Sitzungsberichte der Gesellschaft Naturforschender Freunde zu Berlin 1919: 146-153.Erichson, W. F. 1842. Beitrage zur Inseckten-fauna von Vandiemensland. Archiv fur Naturgeschichte 8: 83-287. Essig, E. 0. 1926. Insects of western North America, ix+1035 pp. New York.Fabricius, J. C. 1793. Entomologica Systematica. viii+519 pp. Hafniae. 1804. Systema Piezatorum . . . adjectissynonymis, locis, observationibus. 439 pp. Brunsvigae. Finlayson, T. 1967. A classification of the subfamily Pimplinae (Hymenoptera: Ichneumonidae) based on final-instar larval characteristics. Canadian Entomologist 99: 1-8.1970. The final-instar larval characteristics of Coleocentrus rufus (Hymenoptera: Ichneumonidae). Canadian Entomologist 102: 905-907. 1976. Cephalic structures and spiracles of final-instar larvae of the genus Toxophoroides (Hymenop- tera: Ichneumonidae: Lycorinae). Canadian Entomologist 108: 981-984.Fitton, M. G. 1978. The species of "Ichneumon" (Hymenoptera) described by Linnaeus. Biological Journal of the Linnean Society 10: 361-383.Foerster, A. 1869. Synopsis der Familien und Gattungen der Ichneumonen. Verhandlungen des Naturhis- torischen Vereins der Preussischen Rheinlande und Westfalens 25: 135-221.1871. Ubersich der Gattungen und Arten der Plectiscoiden. Verhandlungen des Naturhistorischen Vereins der Preussischen Rheinlande und Westfalens 28: 71-123.Froggatt, W. W. 1909. Report on parasitic and injurious insects. Report of the Department of Agriculture of New South Wales 1907-1908: 64-65. 1910. Friendly insects. Agricultural Gazette of New South Wales 21: 334-350. Fiihrer, E. & Kilincer, N. 1972. Die motorische Aktivitat der endoparasitischen Larven von Pimpla turionellae L. und Pimpla flavicoxis Ths. [Hym., Ichneum.] in der Wirtspuppe. Entomophaga 17: 149-163.Furniss, R. L. & Dowden, P. B. 1941. Western hemlock sawfly, Neodiprion tsugae Middleton, and its parasites in Oregon. Journal of Economic Entomology 34: 46-52.Gauld, I. D. 1983. The classification, evolution and distribution of the Labeninae, an ancient southern group of Ichneumonidae (Hymenoptera). Systematic Entomology 8: 167-178. 1984. An introduction to the Ichneumonidae of Australia. 413 pp. London. Gilbert, J. M. & Miller, L. W. 1952. An outbreak of Sirex Noctilio in Tasmania. Australian Forestry 16: 63-69.Girault, A. A. 1924a. Lese majeste, new insecta and robbery. 1 p. Gympie. 19246. New insecta from Queensland. 1 p. Gympie. 1925. Records of Australian ichneumon-flies (Hymenoptera). Insecutor Inscitiae Menstruus 13: 35^0. 1926. A miscellany of new species of the lower Hymenoptera from Australia, with notes. Insecutor Inscitiae Menstruus 14: 133-137. 1930. New pests from Australia VIII. 5 pp. Brisbane. Gourlay, E. S. 1926. Echthromorpha intricatoria F. in New Zealand. Entomologist's Monthly Magazine 62: 169-171.Graham, A. R. 1947. Feeding of Pimpla examinator Ratz. on host pupae exposed for parasitism. Annual Report of the Entomological Society of Ontario 1947: 44-45.Gravenhorst, J. L. C. 1820. Monographia ichneumonum Pedemontanae regionis. Memoir es des Academic Sciences Torino 24: 275-388. 1829. Ichneumonologia Europaea 3. 1097 pp. Vratislaviae. Guillou, M. 1841. [Untitled, in Societes Savantes.] Revue Zoologique par la Societe Cuvierienne 1841: 321-325.Gupta, S. & Gupta, V. 1983. The tribe Gabuniini (Hymenoptera: Ichneumonidae). Oriental Insects Monographs 10: 1-313. Gupta, V. K. 1961. A revision of the oriental species of the genus Zaglyptus (Hymenoptera: Ichneumoni-dae). Indian Journal of Entomology 22: 244-257.1962. Taxonomy, zoogeography and evolution of Indo-Australian Theronia (Hymenoptera: Ichneumonidae). Pacific Insects Monographs 1: 1-142. 1982o. A review of the genus Perithous, with descriptions of new taxa (Hymenoptera: Ichneumoni- dae). Contributions of the American Entomological Institute 19(4): 1-20.19826. A study of the genus Hybomischos (Hymenoptera: Ichneumonidae). Contributions of the American Entomological Institute 19(5): 1-5.Gupta, V. K. & Tikar, D. T. 1978. Ichneumonologia Orientalis, part 1. The tribe Pimplini (Hymenoptera: Ichneumonidae: Pimplinae). Oriental Insects Monographs 1: 1-312. 334 I. D. GAULD Habermehl, H. 1917. Beitrage zur Kenntnis der palaarktischen Ichneumonidenfauna. Zeitschrift fur Wissenschaftlichte Insecktenbiologie 13: 161-168.Haliday, A. 1839. Descriptions of new British insects, indicated in Mr Curtis's guide. Annals of Natural History 2 (1838): 112-121.Hardy, D. E. 1982. The role of taxonomy and systematics in integrated pest management programmes. New Bulletin of the Entomological Society of Queensland 10: 19-24.Haupt, H. 1954. Fensterfange bemerkenswerter Ichneumonen. Deutsche Entomologische Zeitschrift (N.F.) 1:99-116.Heather, N. W. 1976. An outbreak of the leaf bagworm Hyalarcta huebneri (Westw.) (Lepidoptera: Psychidae) in forest plantations ofPinus radiata in Queensland. Queensland Journal of Agricultural and Animal Sciences 33: 145-154. Hennig, W. 1966. Phylogenetic Systematics. xiii+263 pp. Urbana, Illinois.Holmgren, A. E. 1859. Conspectus generum Pimplariarum Sueciae. Ofversigt af Kongliga Vetenskaps- Akademiens Forhandlingar 16: 121-132. 1868. Hymenoptera. Kongliga Svenska Fregatten Eugenics Res omkring jorden Zoologie 6: 391-442.Howell, J. O. & Pienkowski, R. L. 1972. Notes on the biology of a spider parasite Colpomeria kincaidii Ashmead [Hymenoptera, Ichneumonidae]. Entomophaga 17: 5-7.Huffaker, C. B. & Messenger, P. S. (eds) 1976. Theory and practice of biological control. 788 pp. New York.Ironside, D. A. 1974. Biology oiMacadamia nut-borer (Cryptophlebia ombrodelta (Lower)). Queensland Journal of Agricultural and Animal Sciences 31: 201-212. Irvine, C. J. 1962. Forest and timber insects in Victoria, 1962. Victoria's Resources 4: 40-43.Iwata, K. 1958. Ovarian eggs of 233 species of the Japanese Ichneumonidae. Acta Hymenopterologica Fukuoka 1: 63-74.- 1960. The comparative anatomy of the ovary in Hymenoptera. Part V. Ichneumonidae. Acta Hymenopterologica Fukuoka 1: 115-169. 1961. Biological observations on Ichneumonidae. Acta Hymenopterologica Fukuoka I: 315-325. Jussila, R. & Kapyla, M. 1975. Observations on Townesia tenuiventris (Hlmgr.) (Hym., Ichneumonidae) and its hosts Chelostoma maxillosum (L.) (Hym., Megachilidae) and Trypoxylon figulus (L.) (Hym., Sphecidae). Annales Entomologici Fennici 41: 81-86. Kamath, M. K. & Gupta, V. K. 1972. Ichneumonologia Orientalis, Part 2. The tribe Rhyssini (Hymenop-tera: Ichneumonidae: Pimplinae). Oriental Insects Monographs 2: 1-300.Keeping, M. G. & Crewe, R. M. 1983. Parasitoids, commensals and colony size in nests of Belongaster (Hymenoptera: Vespidae). Journal of the Entomological Society of Southern Africa 46: 309-323.Kirby, W. F. 1837. In Richardson, J., Swainson, W. & Kirby, W. F., Fauna Boreali- Americana 4. xxxiv+325 pp. Norwich. 1883. Notes on new or little-known species of Hymenoptera chiefly from New Zealand. Transactions of the Entomological Society of London 1883: 199-203.Konigsmann, E. 1978. Das phylogenetische System der Hymenoptera 3. Terebrantes (Unterordnung Apocrita). Deutsche Entomologische Zeitschrift (N.F.) 25: 1-55.Kriechbaumer, J. 1889. Nova genera et species Pimplidarum. Entomologische Nachrichten 15: 307-312.1890. Ichneumonen-Studien. Neue Ichneumoniden des Wiener Museums. Annalen des Naturhistoris- chen H of museums Wien 5: 479-491. 1892. Xylonomiden und Pimpliden-Studien. Die Xylonomiden-Gattung Perosis Frst. und deren mirbekannte Arten. Pseudacoenitesn.g., Pimplidarum. Entomologische Nachrichten 18: 211-220. 1894a. Hymenoptera ichneumonidea a medico nautico Dr Joh. Brauns in itinere ad oras Africae occidentalis lecta. Berliner Entomologische Zeitschrift 39: 43-55. 1894/J. Hymenoptera Ichneumonidae. Berliner Entomologische Zeitschrift 39: 297-318. 1895. Hymenoptera nova exotica Ichneumonidea. Sitzungsberichte der Naturforschenden Gesell- schaftzu Leipzig 1893/4: 124-136.Krieger, R. 1899. Uber einige mit Pimpla verwandte Ichneumonidengattungen. Sitzungberichte derNaturforschenden Gesellschaft zu Leipzig 1897/8: 47-124. 1906. Ueber die Ichneumonidengattung Theronia Holmgren. Zeitschrift fur Systematische Hymenop-terologie und Dipterologie 6: 231-240. 1909. Ueber die Ichneumonidengattung Echthromorpha Holmgren. Mitteilungen aus dem Zoologis-chen Museum in Berlin 4: 295-344. 1914. Ueber die Ichneumonidengattung Xanthopimpla Saussure. Archiv fur Naturgeschichte 80(6): 1-148. ICHNEUMONIDAE OF AUSTRALIA 335 Leius, K. 1960. The attractiveness of different foods and flowers to the adults of some hymenopterous parasites. Canadian Entomologist 92: 369-376. LePelley, R. H. 1954. Agricultural Insects of East Africa, x+307 pp. Nairobi.Linnaeus, C. 1758. Systema Naturae . . . Edition 10. ii+1384 pp. Holmiae.Madden, J. L. 1968. Behavioural responses of parasites to the symbiotic fungus associated with Sirex noctilio F. Nature (London) 218: 189-190.Martyn, E. J., Hudson, N. M., Hardy, R. J., Terauds, A., Rapley, P. E. L., Williams, M. A., Ireson, J. E. & Miller, L. A. 1977. Insect pest occurrences in Tasmania 1975-76. Insect Pest Survey 9: 1-27.Mason, W. R. M. 1974. An endemic subspecies of Echthromorpha agrestoria on Easter Island (Hymenop- tera: Ichneumonidae). Canadian Entomologist 106: 935-936.Meyer, N. F. 1934. Schulupfwespen die in Russland in den letzten Jahren aus Schadlingen gezogen sind. Zeitschrift fur angewandte Entomologie 20: 611-618.Miller, D. & Clark, A. F. 1935. Sirex noctilio (Hym.) and its parasites in New Zealand. Bulletin of Entomological Research 26: 149-155.Mocsary, A. 1905. Rhyssae sociarumque species in collectione Musei Nationalis Hungarici. Annales Musei Nationalis Hungarici 3: 1-20.Momoi, S. 1966. Some new Ichneumonidae (Hymenoptera) from New Guinea and adjacent areas. Pacific Insects 8: 152-164.Morgan, D. & Stewart, N. C. 1966. The effect of Rhyssa persuasoria (L.) (Ichneumonidae) on a population of Sirex noctilio F. (Siricidae). Transactions of the Royal Society of New Zealand 8: 31-38.Morley, C. 1909. On the Ichneumonidae of the Banksian collection in the British Museum. Entomologist 42: 131-137.1913a. A revision of the Ichneumonidae ... 2. Tribes Rhyssides, Echthromorphides, Anomalides and Paniscides. ix+140pp. London. 19l3b. The fauna of British India, including Ceylon and Burma. Hymenoptera 3. Ichneumonidae 1; Ichneumones Deltoidei. 532 pp. London. 1914. A revision of the Ichneumonidae . . . 3. Tribes Pimplides and Bassides. ix+149 pp. London. 1915. In Turner, R. E. , Meade-Waldo, G. & Morley, C. , Report on the Hymenoptera collected by the British Ornithologist's Union Expedition and the Wollaston Expedition in Dutch New Guinea. 112 pp. London.Muldrew, J. A. 1967. Biology and initial dispersal of Olesicampe (Holocremnus) sp. nr. nematorum (Hymenoptera: Ichneumonidae), a parasite of the larch sawfly, recently established in Manitoba. Canadian Entomologist 99: 312-321.Neumann, F. G. & Minko, G. 1981. The Sirex woodwasp in Australian radiata pine plantations. Australian Forestry 44: 46-63.Nielsen, E. 1923. Contributions to the life history of the pimpline spider parasites (Polysphincta, Zaglyptus, Tromatobid) (Hym., Ichneum.) Entomologiske Meddelelser 14: 137-205.- 1935. A third supplementary note upon the life histories of the polysphinctas (Hym., Ichneum.). Entomologiske Meddelelser 19: 193-215. 1937. A fourth supplementary note upon the life histories of the polysphinctas (Hym., Ichneum.). Entomologiske Meddelelser 20: 25-28.Nuttall, M. J. 1973. Pre-emergence fertilization of Megarhyssa nortoni (Hymenoptera: Ichneumonidae). New Zealand Entomologist 5: 112-117. Parrott, A. W. 1952. New Zealand Ichneumonidae 2. The tribe Echthromorphini (Pimplinae). Transac-tions of the Royal Society of New Zealand 80: 155-170.1957. Notes on the host relation of some Australian Ichneumonidae, with a description of a new species. Memoirs of the National Museum of Victoria 21: 79-82.Perkins, J. F. 1942. On a new species of Ephialtes parasitizing the codling moth. Bulletin of Entomological Research 33: 171-174.1952. Echthromorpha from the Marquesas and Society Islands (Hymenoptera: Ichneumonidae). Proceedings of the Hawaiian Entomological Society 14: 533-536. Price, P. W. 1973. Reproductive strategies in parasitoid wasps. American Naturalist 107: 684-693.1975. In Price, P. W. (ed.), The evolutionary strategies of parasitic insects and mites. 244 pp. New York.Provancher, L. 1873. Les ichneumonides de Quebec avec description de plusiers especes nouvelles. Naturaliste canadien 5: 435^52.Puttier, B. 1961. Biology of Hyposoter exiguae (Hymenoptera: Ichneumonidae), a parasite of lepidopter- ous larvae. Annals of the Entomological Society of America 54: 25-30.Rao, S. N. 1953. Notes on the Indian Ichneumonidae. Indian Forest Records (Entomology) 8: 159-225. 336 I. D. GAULD Riek, E. F. 1970. Hymenoptera. In: Insects of Australia xiii+1029 pp. Melbourne. Rojas-Rousse, D. & Benoit, M. 1977. Morphology and biometry of larval instars of Pimpla instigator (F.) (Hymenoptera: Ichneumonidae). Bulletin of Entomological Research 67: 129-141.Roman, A. 1910. Notizen Schlupfwespensammlung des schwedischen Reichmuseums. Entomologisk Tidskrift 31: 109-196.Rosenberg, H. T. 1934. The biology and distribution in France of the larval parasites of Cydia pomonella L. Bulletin of Entomological Research 25: 201-256.Salt, G. 1968. The resistance of insect parasitoids to the defence reactions of their hosts. Biological Reviews of the Cambridge Philosophical Society 43: 200-232.Saussure, H. de 1892. Hymenopteres. In Grandidier, A., Histoire physique naturelle et politique de Madagascar 2. xxi+590 pp. Paris. Schmiedeknecht, 0. 1907. Opuscula Ichneumonologica 15: 1121-1200.Schulz, W. A. 1906. Spolia Hymenopterologica. 356 pp. Paderborn. 1911. Zweihundert alte Hymenopteren. Zoologischen Annalen (Wurzburg) 4: 1-220.1912. Einige Bemerkungen iiber Schlupfwespen. Zeitschrift fur Wissenschaftliche Insektenbiologie 8: 353-354.Seyrig, A. 1932. Les ichneumonides de Madagascar 1. Ichneumonidae pimplinae. Memoires de L'Academie Malgache 11: 1-183.Short, J. T. R. 1978. Larvae of Ichneumonidae. Memoirs of the American Entomological Institute 25: 1-508.Skaife, S. H. 1921. Some factors in the natural control of the wattle bagworm. South African Journal of Science 17: 291-301.Smith, F. 1859. Catalogue of hymenopterous insects collected by A. R. Wallace at the islands of Aru and Key. Journal and Proceedings of the Linnean Society of London (Zoology) 4: 132-178.- 1863. Catalogue of the hymenopterous insects collected by A. R. Wallace in the islands of Mysol, Ceram, Waigiou, Bouru and Timor. Journal and Proceedings of the Linnean Society of London (Zoology) 6: 6-48. 1865. Description of new species of hymenopterous insects from the islands of Sumatra, Sula, Gilolo, Salwatty and New Guinea collected by A. R. Wallace. Journal and Proceedings of the Linnean Society of London (Zoology) 8: 61-94.Smith, J. H. & Caldwell, N. E. H. 1947. Army worm and other noctuid outbreaks during 1946-47. Queensland Agricultural Journal 65 : 396-401 .Smithers, C. N. 1956. On Philopsyche abdominalis Morley (Hym., Ichneumonidae), a parasite of Acanthopsyche junodi Heylaerts (Lep., Psychidae). Journal of the Entomological Society of Southern Africa 19: 225-249.Snellen van Vollenhoven, S. C. 1878. Especes nouvelles ou peu connues d'Hymenopteres terebrants. Tijdschrift voor Entomologie 21: 153-177.Sonan, J. 1930. A few host known Ichneumonidae found in Formosa (Hym.) (2). Transactions of the Natural History Society of Formosa 26: 103-107.Spradbery, J. P. 1968. A technique for artificially culturing ichneumonid parasites of woodwasps (Hymenoptera: Siricidae). Entomologia Experimental etApplicata 11: 257-260.1969. The biology oiPseudorhyssasternata Merrill (Hymenoptera: Ichneumonidae), a cleptoparasite of siricid woodwasps. Bulletin of Entomological Research 59: 291-297. 1970. The immature stages of European ichneumonid parasites of siricine woodwasps. Proceedings of the Royal Entomological Society of London 45(A): 14-28.Spradbery, J. P. & Ratkowsky, D. A. 1974. An analysis of geographic variation in the parasitoid Rhyssa persuasoria (L.) (Hymenoptera: Ichneumonidae). Bulletin of Entomological Research 64: 653-668.Swederus, N. S. 1787. Fortsattning af Beskrifningen pa 50 nya species af insecter. Kungliga Svenska Vetenskapsakademiens Handlingar 8: 276-290.Szepligeti, G. V. 1908. E. Jacobons'sche Hymenopteren aus Semarang. Notes from the Leyden Museum 29: 209-260.- 1914. Ichneumoniden aus der Sammlung des ungarischen National-Museums. Annales Musei Nationalis Hungarici 12: 414434.Taylor, K. L. 1967. The introduction, culture, liberation and recovery of parasites of Sirex noctilio in Tasmania, 1962-67. CSIRO Division of Entomology Technical Paper 8: 1-19. 1976. The introduction and establishment of insect parasitoids to control Sirex noctilio in Australia. Entomophaga 21: 429-440. 1978. Evaluation of the insect parasitoids of Sirex noctilio (Hymenoptera: Siricidae) in Tasmania. Oecologiatt: 1-10. ICHNEUMONIDAE OF AUSTRALIA 337 Temperley, M. 1930. Life history notes on the banana fruit-eating caterpillar (Tiracola plagiata Walk.). Queensland Agricultural Journal 33: 251-261.Thunberg, C. P. 1822. Ichneumonidea, Insecta Hymenoptera illustrata. Memoires de I' Academic Im- periale des Sciences de Saint Petersbourg 8: 249-281.Togashi, 1. 1963. A comparative morphology of the poison glands in the adults of ichneumon-flies (Hym., Ichneumonidae). Kontyu3l: 297-304.Tosquinet, J. 1903. Ichneumonides nouveaux. Memoires de la Societe Royale Entomologie de Belgique 10: Tothill, J. D. 1922. The natural control of the fall webworm (Hyphantria cunea Drury) with an account ofits several parasites. Bulletin of the Department of Agriculture of Canada, Entomology Branch 19: 1-107. Townes, H. 1940. A revision of the Pimplini of eastern North America (Hymenoptera: Ichneumonidae).Annals of the Entomological Society of America 33: 283-323. - 1958. Hymenoptera: Ichneumonidae, Stephanidae and Evaniidae. Insects of Micronesia 19(2):35-87. - 1960a. The application of the name Syene (Hymenoptera: Ichneumonidae). Proceedings of theEntomological Society of Washington 62: 43. - 19606. In Townes, H. & Townes, M., Ichneumon-flies of America north of Mexico: 2, subfamiliesEphialtinae, Xoridinae, Acaenitinae. Bulletin of the United States National Museum 216(2): 1-676. - 1962. A new generic name in the polysphinctine ichneumonids. Proceedings of the EntomologicalSociety of Washington. 64: 38. - 1969. Genera of Ichneumonidae 1. Memoirs of the American Entomological Institute 11: 1-300. - 1970. Genera of Ichneumonidae 3. Memoirs of the American Entomological Institute 13: 1-307.1971a. A review of the Ichneumonidae described by Girault. Proceedings of the Entomological Society of Washington 72: 458-470. 19716. Genera of Ichneumonidae 4. Memoirs of the American Entomological Institute 17: 1-372. Townes, H. & Chiu, S.-C. 1970. The Indo-Australian species of Xanthopimpla (Ichneumonidae). Memoirs of the American Entomological Institute 14: 1-372.Townes, H., Momoi, S. & Townes, M. 1965. A catalogue and reclassification of eastern Palaearctic Ichneumonidae. Memoirs of the American Entomological Institute 5: 1-661.Townes, H. & Townes, M. 1960. Ichneumon-flies of America north of Mexico: 2, subfamilies Ephialtinae, Xoridinae, Acaenitinae. Bulletin of the United States National Museum 216(2): 1-676.1966. A catalogue and reclassification of Neotropic Ichneumonidae. Memoirs of the American Entomological Institute 8: 1-367.Townes, H., Townes, M. & Gupta, V. K. 1961. A catalogue and reclassification of Indo-Australian Ichneumonidae. Memoirs of the American Entomological Institute 1: 1-522.Tryon, H. 1900. Caterpillar plague. Queensland Agricultural Journal 6: 135-147.Turner, R. E. 1919. Notes on the Ichneumonidae in the British Museum 1. Annals and Magazine of Natural History 9(3): 550-558.Uchida, T. 1928. Dritter Beitrage zur Ichneumoniden-fauna Japans. Journal of the Faculty of Agriculture of Hokkaido Imperial University 25: 1-115. Vachal, J. 1907. Hymenopteres de la Nouvelle-Caledonie. Revue d' Entomologie, Caen 26: 113-123.Viereck, H. L. 1912. Descriptions of five new genera and twenty six new species of ichneumon-flies. Proceedings of the United States National Museum 42: 139-153.1914. Type species of the genera of ichneumon-flies. Bulletin of the United States National Museum 31: 1-186.Walkley, L. M. 1963. The type-species of Austrapophua Girault. Proceedings of the Entomological Society of Washington 65: 116.Walsh, B. D. 1873. Descriptions of north American Hymenoptera. Transactions of the Academy of Sciences of Saint Louis 3: 65-166.Westwood, J. O. 1840. An introduction to modern classification of insects, 2 and synopsis of the genera of British insects. 587+158 pp. London.Wilson, F. 1960. A review of the biological control of insects and weeds in Australia and Australian New Guinea. Technical Communication of the Commonwealth Institute of Biological Control 1: 1-102.Winter, T. G. 1979. On the biology of Rhyacionia duplana logaea Durrant (Lepidoptera: Tortricidae). Entomologist's Gazette 30: 257-265.Woldestedt, F. W. 1877. Ueber eine Sammlung schlesischer Ichneumoniden. Bulletin de V Academic des Sciences de Saint Petersbourg 22: 1102.Zondag, R. & Nuttall, M. J. 1961. Rhyssa lineolata (Kirby) (Hymenoptera: Ichneumonidae: Pimplinae). A species new to New Zealand. New Zealand Entomologist 2(6): 40-44. 338 I. D. GAULD Index to hosts Achaea Janata 275Agaristidae 270Anomis flava 271Anomis lyona 271Anthela acuta 314Anthela denticulata 272, 275Anthela xantharcha 272Anthelidae241,270Antheraea astrophela 314, 315Antheraea saccopoea 273Antheraea sp. 314Araneidae 252, 259 Belonogaster 246Bombycidae 270Bombyx mori 272Buprestidae 241 Cactoblastis sp. 275Cephidae 242Cerambycidae 241, 316Clania ignobilis 251Clubionidae 252, 258Cryptophlebia ombrodelta 271 Drassodidae 258 Epicampoptera 246Epiphyas postvittana 272, 302Euproctis edwardsl 272 Hesperiidae 270Hesperilla chrysotricha 272Hesperilla donnysa 272Hyalarcta huebneri 251, 272, 315 'Hyalarcta nlgrescens 251Hyblaea sp. 246 Lomera caespitosa 272Lycaenidae 270Lymantriidae 241, 270 Merophyas divulsana 308Mythimna convecta 275Mythimna separata 275 Neodrepta luteotactella 272Noctuidae241,270, 303Nymphalidae241,270 Ogyris olane 212 Oiketicus elongatus 251 Olene mendosa 212 Ornithoptera priamus euphorion 273 Papilionidae 270Parnara amalia 271Pectinophora scutigera 300Pericyma cruegeri 315 Persectania ewingii 212Phalaenoides glycinae 212Psychidae 244, 250, 270Pyralidae 245, 303, 328 Saturniidae 270Sesiidae241,321Sirex noctilio 318, 319Sirex sp. 317, 318Siricidae 316, 321Spodoptera exempta 212, 275Spodoptera mauriiia 275Syntexidae 316, 319 Teia anartoides 272, 315Tetragnathidae 257Theridiidae 263Tiracola plagiata 275Tortricidae 242, 270 Uraba lugens 308Uroceros sp. 318 Vanessa itea 212 Xeris sp. 318Xiphyriidae 316Xyloryctidae 270 Index to Ichneumonidae Invalid names are in italics; principal references are in bold. Acaenitinae 236, 238, 322Acrodactyla 237, 240, 243, 254Acropimpla 237, 239, 242, 243, 244,250 agrestoria 237, 270, 271, 282, 292Alophopimpla 237, 239, 242, 243, 268, 269 amaryllyx 238, 289, 323, 324, 326amon 238, 278, 279, 283, 284ankhu 238, 278, 279, 283, 284, 285, 299 annulipes 237, 250antherae 238, 315 apicipennis 238, 289, 323, 324, 325arealis 238, 278, 284, 285, 286, 294, 299, 303 atra 238, 273, 274atropos 237, 246australiensis 238, 321australis Krieger 238, 279, 285, 286, 294, 300, 304 aqstralis Townes, Townes & Gupta237, 250, 290 barak 238, 278, 284, 285, 300, 301, 306 beauforti 238, 299bicolor 237, 246, 247, 286bingili 237, 263, 264, 267, 268, 288binodus 238, 277, 284, 301, 309biroi 238, 316, 317 Camptotypus 237, 239, 240, 242,243, 244, 245, 246, 247, 248,249 canberrae 238, 289, 328, 330 carinatus 237, 259, 260, 289 caudata 238, 305 celer 237, 264, 265, 288- cephalotes 238, 312 daripennis 238, 315 dotho 237, 248 coelyx 238, 289, 323, 324, 325 conopleura 237, 270 consimilis 237, 251 crassa 238, 305 crenator 237, 250, 251, 280, 286, 290 crudelis 238, 321cursor 237, 255, 288Cyanoxorides 238, 320 dandiensis 237, 264, 265, 288, 289decem-notata 238, 274Delomeristini 238, 239, 242, 310dinyx 238, 323, 326diversor 237, 270 Dreisbachia 237, 240, 243, 254, 258dubia 238, 315duodecem-guttata 237, 271 ecaudata 238, 278, 284, 285, 295, 302Echthromorpha 237, 238, 241, 244, 268, 269, 270emaculata 238, 302Ephialtini 237, 239, 241, 242, 243, 244,254E'pirhyssa 238, 242, 243, 316 ICHNEUMONIDAE OF AUSTRALIA 339 Eriostethus 237, 240, 243, 254, 258, 259, 289 excavata 237, 271excelsa 238, 273, 274, 275, 281, 283, 293 fastigata 237, 272flaviceps 237, 248flavolineata 238, 277, 278, 285, 296, 299, 302, 303fraterculus 238, 278, 283, 284, 303, 304 fraucai 238, 287, 31 1,313, 314fumata 238, 312fumipennis 238, 315 glabrinotum 237, 252, 253, 282, 286gracilis 238, 299 hiatus 238, 278, 282, 284, 285, 304, 308 hirsuta 238, 277, 283, 284, 304, 305hispida 238, 302hollowayi 237, 252, 253, 286 .hyaloptila 238, 302 immaculata Krieger 237, 271immaculata Morley, Echthromorpha 237, 271immaculata Morley, Xanthopimpla 238, 302insidiator 237, 270intermpta 237, 270intricatoria 237, 270, 271, 272, 276, 282, 292, 311 kauros 237, 264, 266, 288kluia 237, 269 lachesis 237, 246, 247, 286, 291Lissopimpla 238, 239, 243, 268, 273Lissopimpla species 1 238, 274, 276lutea Dreisbachia 237, 258lutea Sericopimpla 237, 250, 252, 282, 286 Lycorina 238, 328Lycorina species 1 238, 328, 330Lycorininae 236, 238, 327 maxima 237, 272 maximus 237, 259, 260, 261, 281, 288Megarhyssa 238, 240, 243, 316, 317, 318 melanosoma 238, 287, 311, 312melioratorius 237, 270micans 237, 255, 256, 288minimus 237, 259, 261, 262minor 238, 302 nigricornis 237, 270, 272, 282Nomosphecia 238, 310, 311nortoni 238, 317, 318 obesa 238, 274, 275 ochracea 238, 279, 284, 295, 299, 305 ochracea peter seni 238, 305, 306octo-guttata 238, 274 papuana Cameron, Parema 238, 312papuana Cameron, Xanthopimpla 238, 299 Parema 238, 310, 312Parvipimpla 237, 240, 242, 244, 249penetrans 238, 287, 311, 312perkinsi 237, 259, 262persuasoria 238, 317, 318, 319petita 237, 249phraxos 237, 264, 266, 267, 288, 289 pilosella 237, 251Pimplinae 236, 237, 238, 240, 241Pimplini 237, 238, 239, 240, 241, 242, 244, 268, 314platymischa 237, 270Polysphinctini 237, 239, 240, 244, 254 priocnemidea 238, 274pubidorsis 238, 278, 284, 306pulcherrimus 237, 259, 260, 261, 262, 288 quadridens 238, 277, 284, 285, 306, 307 quadrisculpta 237, 255, 256, 288quebecensis 238, 318 rhopaloceros 238, 277, 284, 286,296,301,305,307,308,309Rhyssa 238, 241, 243, 316, 318Rhyssini238, 239, 241,316rufipes 238, 274 scutata 238, 274, 275, 276, 283, 293sellatus 237, 245, 246, 248, 282, 286, 291 semipunctata 238, 274Sericopimpla 237, 240, 241, 243, 244, 249, 250, 254sesamiae 238, 302simUis 238, 290 splendidula 238, 289, 328, 329, 330steindachneri 238, 287, 298, 311, 314, 315 stellata 237, 263, 264, 267, 268striata 237, 270striata 238, 278, 279, 284, 308summervillei 238, 277, 284, 286, 297, 308, 309 tantalyx 238, 289, 323, 326, 327 teiae 238, 315 terminalis 238, 277, 284, 285, 297, 309, 310 Theronia 238, 239, 240, 242, 243, 310, 313 turneri 238, 289, 328, 329, 330 velata 237, 263, 267, 268, 288viridicans 238, 314 xantha 237, 245, 282xanthostigma 238, 302Xanthopimpla 236, 238, 239, 240, 243, 268, 276, 277, 310xanthopimploides 238, 307xara 238, 302Xorides 238, 319, 320Xorides species 1 238, 320, 322Xoridinae 236, 238, 319, 320 Yezoceryx 238, 322, 323Yezoceryx species A 238, 324, 327 maculosa 238, 287, 298, 311, 314,315 rennefer 237, 263, 264, 267, 288,289 Zaglyptus 237, 239, 240, 243, 244, 252, 254Zatypota 237, 240, 243, 254, 259, 263zekhem 237, 255, 257, 288 British Museum (Natural History) Milkweed butterflies: their cladistics and biology P. R. Ackery & R. I. Vane- Wright The Danainae, a subfamily of the Nymphalidae, contains only some 150 species, yet aspects oftheir biology have stimulated far more attention than can be justified by species numbersalone. In recent years, an expansive literature has grown, considering aspects of theircourtship and pre-courtship behaviour, migration, larval hostplant associations, mimicry andgenetics. The popularity of danaines among biologists can certainly be attributed to thiscombination, within one small group, of so many of the factors that make butterflies such aninteresting group to study. The obvious need to place this wealth of biological data within anacceptable systematic framework provided the impetus for this volume. Started eight years ago within the conventions of evolution by natural selection andHennig's phylogenetic systematics, the book is now largely about natural history (what theanimals have and do, where they live and how they develop) and natural groups - as revealedby a form of analysis approaching that practised by the new school of 'transformed cladistics'.The authors have prepared a handbook that will appeal to a wide range of biologists, frommuseum taxonomists to field ecologists. 424 pp (approx.), 12 pp colour, 73 b/w plates, line and graphic illustrations, maps, extensive bibliography.ISBN 565 00893 5. Publication September 1984. Price 50, prepublication price 45. Titles to be published in Volume 49 Afrotropical jumping plant lice of the family Triozidae (Homoptera: Psylloidea). By David Hollis. The taxonomy of the western European grasshopper of the genus Euchorthippus, with specialreference to their songs (Orthopetera: Acrididae). By D. Ragge & W. J. Reynolds An historical review of the higher classification of the Noctuidae. By Ian Kitching The Pimplinae, Xoridinae, Acaenitinae and Lycorininae (Hymenoptera: Ichneumonidae). By I. D. Gauld The Palaearctic species of Ascogaster (Hymenoptera: Braconidae) By T. Huddleston Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, SuffolkPrinted in Great Britain by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) UM The Palaearctic species of Ascogaster(Hymenoptera: Braconidae) T. Huddleston Entomology series Vol 49 No 5 20 December 1984 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique andever-growing collections of the Museum, both by the scientific staff of the Museum and byspecialists from elsewhere who make use of the Museum's resources. Many of the papers areworks of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself,available separately, and individually priced. Volumes contain about 300 pages and severalvolumes may appear within a calendar year. Subscriptions may be placed for one or more ofthe series on either an Annual or Per Volume basis. Prices vary according to the contents ofthe individual parts. Orders and enquiries should be sent to: ' T. C* ^ , . -" L^" 5 " 5 ^ Publications Sales > ' J ?^W British Museum (Natural History) ,Cromwell Road, London SW75BD, 20DECI984 I - U England. _ World List abbreviation: fiw//. 5r. Mus. nat. Hist. (Ent.) Trustees of the British Museum (Natural History), 1984 The Entomology series is produced under the general editorship of the Keeper of Entomology: Laurence A. Mound Assistant Editor: W. Gerald Tremewan ISBN 565 06007 4ISSN 0524-6431 British Museum (Natural History)Cromwell RoadLondon SW7 5BD Entomology seriesVol49No5pp341-392 Issued 20 December 1984 The Palaearctic species of Ascogaster(Hymenoptera: Braconidae) T. Huddleston Department of Entomology, British Museum (Natural History), Cromwell Road, LondonSW75BD Contents Synopsis 341 Introduction 341 Biology 342 List of host records 342 Taxonomic history 343 Synonymic list of species 344 Taxonomic characters 345 Depositories 346 Characteristics of subfamily Cheloninae 347 Ascogaster Wesmael 348 The species-groups 348 Key to species 348 The semenovi-group 352 The caucasica-group 353 The abdominator-group 357 The annularis-group 360 The bidentula-group 364 The quadridentata-group 371 Species inquirendae 380 Excluded species 381 Acknowledgements 381 References 382 Index... 392 Synopsis The cosmopolitan genus Ascogaster, which is parasitic on Microlepidoptera, is revised for the Palaearcticzoogeographical region. A key to species is given. Thirty species are recognized; four of them are describedas new. Twenty- two synonyms are newly established; one species (nigrator) is transferred to Ascogasterfrom Chelonus and one (maculata) from Ascogaster to Phanerotoma. Introduction It is imperative that biological studies are founded on a basis of sound taxonomy; the results ofsuch studies are otherwise vitiated. The taxonomy of the Braconidae has been composedpiecemeal over the last 200 years or so and relatively few taxa have received comprehensiveup-to-date attention. Many taxa therefore need to be completely revised, at least within a singlezoogeographical region. This revision is concerned with the Palaearctic species of Ascogaster, achelonine braconid genus. Ascogaster is a cosmopolitan genus parasitic upon Microlepidoptera, principally Tortricidae.A. quadridentata has frequently been recorded as a parasite of the codling moth (Cydiapomonella) and other pests of fruit trees (see Evenhuis & Vlug, 1983). This species and probablyseveral others merit consideration for use in biological control programmes. Bull. Br. Mus. nat. Hist. (Ent.) 49 (5): 341-392 Issued 20 December 1984 342 T. HUDDLESTON Biology Very little work has been done on the biology of Ascogaster and, as far as is known, the speciesare solitary endoparasites of Microlepidoptera, principally Tortricidae. They lay their eggs intothe egg of the host and, as van Achterberg (1976: 47) has pointed out, this is clearly anadaptation to the use of hosts whose larvae live in concealment, thereby deriving someprotection against the attacks of parasites and predators. In his studies of Ascogaster quad-ridentata, Rosenberg (1934) found that the development of the parasite is slow and its larva is inits first instar when the host larva has completed its development. If the host larva then goes intohibernation the parasite remains within as a first-stage larva, otherwise it passes rapidly throughtwo further instars and emerges from the host to pupate. Rosenberg's work and that of Cox(1932), Boyce (1936) and Allen (1962) are fairly full accounts of the biology of A. quadridentataand they also give much information on the larval morphology of the parasite. The larvalcharacteristics of chelonines are discussed in more general terms by Short (1952) and Capek(1970). Host-searching and egg-laying behaviour has recently been studied in reticulata parasiticon the smaller tea tortrix (Adoxophyes sp.), an important pest of tea in Japan (Kainoh &Tamaki, 1982; Kainoh, Hiyori & Tamaki, 1982). There are many host records in the literature on Ascogaster, collated by Shenefelt (1973).Much of this information, however, is of little value because of doubt about the accuracy ofidentification of the parasite species involved; undoubtedly host species are also sometimesmisidentified. Host data cited here are therefore only from the specimens examined for thisrevision except for the hosts from which the type-specimens of reticulata were reared (as given byWatanabe, 1967: 43). These records are listed below. List of host records HOSTS LEPIDOPTERA TINEIDAE Infurcitinea argentimaculella (Stainton)YPONOMEUTIDAE Yponomeuta padella (L.)COLEOPHORIDAE Coleophora hornigi TollOECOPHORIDAE Tubuliferola subochreella (Doubleday)GELECHIIDAE Recurvaria leucatella (Clerck)MOMPHIDAE Sorhagenia lophyrella (Douglas)TORTRICIDAE Cydia pomonella (L.) Cydia funebrana (Treitschke) Cydia pallifrontana (Lienig & Zeller) Spilonota ocellana (Denis & Schiffermiiller) Epiblema uddmanniana (L.) Epiblema roborana (Denis & Schiffermiiller) Epinotia cruciana (L.) Endothenia gentianaeana (Hiibner) Endothenia quadrimaculana (Haworth) Pandemis corylana (Fabricius) Pandemis cerasana (Hiibner) Pandemis heparana (Denis & Schiffermiiller) Pandemis sp. Archips oporana (L.) Archips issikii Kodama Adoxophyes orana (Fischer von Roslerstamm) SPECIES CF Ascogaster grahami rufidens, quadridentata armata klugii annularis grahami quadridentata quadridentata quadridentata quadridentata quadridentata bidentula bidentula dispar canifrons rufidens rufidens rufidens, annularis bidentula reticulata reticulata reticulata THE PALAEARCTIC SPECIES OF ASCOGASTER 343 Archips pulchra (Butler) reticulata Croesia bergmanniana (L.) rufidensGEOMETRIC AE Eupithecia venosata (Fabricius) bidentula Taxonomic history Ascogaster was erected by Wesmael (1835: 226) for 10 species which he described as new.Wesmael also correctly recognized that several of Nees von Esenbeck's species of 'Chelonus'came within the definition of the new genus but he also included Chelonus dentata [=Phanerotoma dentata]. This obviously anomalous placement was based upon the examination ofonly a single specimen and was corrected by Wesmael's (1838: 165) erection of the genusPhanerotoma for dentata. Wesmael possessed outstanding taxonomic skill and his discrimina-tion of species was always sound even though he worked with few specimens of each species.This was the general practice in the nineteenth century and the results are generally consideredto be more or less inadequate when judged by modern standards. Latreille (1809) was the first to describe (as Sigalphus) a species which can be unequivocallyassigned to Ascogaster. The description would no doubt be adequate to differentiate the speciesif one had only Latreille's material of the genus available for study. Unfortunately I have beenunable to locate Latreille's specimen and the description alone is insufficient for the certainidentification of the species. Its interpretation must therefore rest upon inference; this questionis discussed in the text upon rufipes (Latreille). Nees von Esenbeck (1816) described (as Sigalphus) four species which were subsequentlytransferred by him (1834) to Chelonus and by Wesmael to Ascogaster. Three of Nees vonEsenbeck's species are certainly identifiable from his descriptions but the other, similis, is not. Curtis (1837) described four species of Ascogaster. Fortunately his collection is still availablefor study and his inadequate descriptions can therefore be interpreted with confidence. One ofhis species is here considered valid. Herrich-Schaffer (1838) described (as Chelonus) 10 species of Ascogaster; his originalmaterial appears to be lost and his descriptions are generally difficult to interpret, butReinhard's paper (see below) is of considerable help in this matter. Two of Herrich-Schaffer'sspecies are here considered valid. In his revision of Ascogaster, Reinhard (1867) described three new species, one of which Iconsider to be valid. Despite Thomson's (1892: 1715) dismissal of Reinhard's revision as 'unfitfor use', it includes much useful information. Reinhard's most significant contribution lies in hisredescriptions and placements of species more or less inadequately described by Herrich-Schaffer and Ratzeburg. Reinhard's descriptions are much superior to those of the originalauthors and were often based on syntypic material which no longer exists. Without Reinhard'spaper the placement of many of those species would be difficult, perhaps impossible. Thomson (1874; 1892) revised the Swedish species of Ascogaster, describing nine species asnew. However, Thomson appears not to have fully understood Wesmael's species and all hisnames are here placed in synonymy, except for lapponicus which could not be identified. There have been only two previous attempts to monograph the Palaearctic species ofAscogaster. Marshall (in Andre, 1888) and Fahringer (1934) included the genus in their reviewsof the Palaearctic braconid fauna. The admirably ambitious scope of these works severelylimited their usefulness; neither author was able to give sufficiently rigorous attention to thelarge number of species involved. Their interpretation of the species was in general based on thedescriptions rather than on a critical re-examination of original material. Marshall at leastrealised that some descriptions were inadequate and this left him unable, for instance, to includeThomson's species of Ascogaster in his key. Both Marshall and Fahringer exhibited a too facileacceptance of the significance of any character proposed as a taxonomic discriminant by originalauthors and made little attempt to search for worthwhile characters. Moreover, Marshall andprobably Fahringer worked with small numbers of specimens and this made it impossible forthem to assess intraspecific variation. Thus the contribution of these major works to theunderstanding of the species of Ascogaster has been insignificant. 344 T. HUDDLESTON Telenga (1941) revised the Ascogaster species of the U.S.S.R.; his taxonomy was muchsuperior to that of Marshall and Fahringer and his work is still useful. Hellen (1953) reviewed theFinnish Ascogaster fauna and established two new synonymies. Graham (1955), in an unpublished doctoral thesis on the European species of Ascogaster (andChelonus), was the first worker to approach the subject systematically. His arrangement of thespecies was based on the examination of a large amount of material, including the criticalreappraisal of the available type-specimens. Graham's thesis was considerably superior toprevious work on the genus and only Tobias (1976) has produced anything of comparable qualityon Ascogaster (mainly of the eastern Palaearctic region). However, because Graham's work wasnever published, his nomenclatural decisions were never validated. I carried out my revision ofthe Palaearctic species of Ascogaster without first examining his manuscript so as to avoid beinginfluenced when solving taxonomic difficulties. Many of the conclusions at which I have thusindependently arrived agree with those reached by Graham. I have, however, ascribed to somespecies wider limits of variation than did Graham and I have therefore placed in synonymy somenames which he regarded as representing valid species, albeit sometimes tentatively. Isolated descriptions by various authors are scattered throughout the literature (Dahlbom,1833; Lucas, 1849; Ratzeburg, 1844; 1848; 1852; Ruthe, 1855; Wollaston, 1858; Marshall, 1897;Szepligeti, 1896; 1908; Ashmead, 1906). Many of these authors appear to have had but littleunderstanding of the species. Synonymic list of species semenovi-group excavate Telenga kasachstanicus Tobias syn. n.semenovi Telenga kyushuensis Yoneda syn. n. caiicas/ca-group bicarinata (Herrich-Schaffer) mlokossewitchi Kokujevcaucasi'ca Kokujev rostrata Szepligetiexcisa (Herrich-Schaffer) longiventris Tobias syn. n.kasparyani Tobiasabdominator-group abdominator (Dahlbom) instabilis Wesmael fulviventris Curtis femoralis (Herrich-Schaffer) rufiventris (Herrich-Schaffer) pallida Ruthedentifer Tobias punctulator Kirchnernachitshevanica Abdinbekova annu/aris-group annularis (Nees von Esenbeck) exigua sp. n. gonocephala Wesmael grahamisp. n. klugii (Nees von Esenbeck)ruficeps Wesmaelneesii Reinhard syn. n. bidentula-group albitarsus Reinhardleptopus Thomson arisanica Sonanbidentula Wesmael multiarticulatus Ratzeburg gibbiscuta Thomson syn. n. fuscipennis Thomson syn. n. atamiensis Ashmead syn. n.consobrina Curtislongicornis sp. n.per/tins/ sp. n.rufidens Wesmael rufipes (Herrich-Schaffer) laevigator (Ratzeburg)varipes Wesmael cavifrons Thomson syn. n. sternalis Thomson jaroslawensis Kokujev syn. n. qfuadridenfa fa-grouparmata Wesmael pulchellus (Curtis) esenbeckii Curtis syn. n. luteicornis (Herrich-Schaffer)brevicornis Wesmael monilicornis (Herrich-Schaffer)canifrons Wesmael graniger Thomson syn. n.dispar Fahringer spinifer Tobias syn. n. koslovi Tobias syn. n.quadridentata Wesmael impressus (Herrich-Schaffer) nigricornis Thomson syn. n. cynipum Thomson syn. n. egregius Kokujev syn. n. nigrator (Szepligeti) syn. n. epinotiae Watanabe syn. n. THE PALAEARCTIC SPECIES OF ASCOGASTER 345 reticulate Watanabe contracta (Ratzeburg) rutipes Latreille dentiventris Telenga elegans (Nees von Esenbeck) erythrothorax Marshall fasciatus (Dahlbom) kabystanica Tobias ratzeburgii Marshall syn. n. lapponica Thomson soror Telenga syn. n. pallidicornis Curtis scabricula (Dahlbom) quadridens (Herrich-Schaffer) limitatus Wesmael syn. n. mbripes (Lucas) clypealis Thomson syn. n. similis (Nees von Esenbeck) SPECIES INQUIRENDAE tersa Reinhardatriceps (Ratzeburg) Taxonomic characters Sexual dimorphism. A few species ofAscogaster show little sexual dimorphism; most, however,exhibit some generally rather subtle differences between the sexes. Female antenna oftenshorter and broader than that of male, generally with areas of felt-like pubescence on undersideof medial flagellar segments. Male antenna generally more copiously covered with hairs andmore richly supplied with tyloids than in female. Carapace of female often shorter, broader andmore rounded than that of male, with postero ventral rim often closer to apex and therefore lessrounded postero ventrally. In several species in which the female has large genitalia thepostero ventral rim of the carapace is modified in various ways in order, presumably, to facilitatethe extrusion of the ovipositor (e.g. flanged in excavata, semenovi; emarginate in annularis;produced in kasparyani); these characteristics are not present in the males of these species. Inmany species facial hair of male longer and much denser than in female. Ocelli. Size of ocelli varies little and therefore of limited use in discrimination of species;however, relative positions of ocelli are characteristic at species-group level. Two character-states of the ocellar triangle are defined as follows: where a straight line drawn between theanterior borders of posterior ocelli also touches anterior ocellus, the ocelli are referred to asbeing 'on line'; where a line between posterior ocelli passes behind anterior ocellus withouttouching it, the ocelli are referred to as being 'not on line'. Variations in size and protuberance ofthe eyes are sometimes useful characteristics for differentiation of species; in none, however, dothe eyes exhibit a significant degree of convergence. Clypeus. Shape, sculpture and apical armature of clypeus generally of great significance. Apicalborder of clypeus often bears one or more tubercles which, in some species, are tooth-like.There is little variation within a species in the clypeal tubercles which are therefore ofconsiderable value as a taxonomic character. The dentate clypeal armature is extremely rare inChelonus although a characteristic feature of many species ofAscogaster. It appears to be largelyconfined to the Palaearctic species of the latter genus and is commoner in those of the easternpart of the region. Facial hair. Hair on face generally pointing downwards except for a patch between bases ofantennae; in some species, however, the hair points upwards, at least on upper part efface. Thischaracteristic varies little within a species and is of considerable use for differentiation ofspecies-groups. It is more easily seen in males because their facial hair is generally denser thanthat of females, but it always requires oblique lighting to be properly examined. Caster. Shape of carapace sometimes varies within a species but nonetheless furnishes usefulcharacters for species differentiation if the possibility of variation is constantly borne in mind.Specimens with a deformed carapace are of not infrequent occurrence. Hypopygium generallysmall, not projecting beyond apex of carapace, at least in dead specimens. In some species (e.g.excisa, semenovi) hypopygium large, projecting beyond apex of carapace. Ovipositor in thesespecies generally long and often exserted. 346 T. HUDDLESTON Genitalia. In chelonines, the genitalia are often retracted beneath the carapace and are thusdifficult to see, particularly in dried specimens. It is difficult to determine the sex of suchspecimens, especially in those species which exhibit little secondary sexual dimorphism. If a viewcan be obtained into the carapace, however, the female can be distinguished as such because thehypopygium has a convex apical border whereas the corresponding male sternite is stronglyemarginate. Also, the hypopygium often has a medial longitudinal fold revealing the course ofthe ovipositor beneath. When the latter projects beyond the hypopygium it is usually possible torecognise it because the ovipositor sheaths are nearly always close together and apicallyrounded, whereas the corresponding structures in the male genitalia are generally widelyseparated, strongly tapered and curved. Colour. In Ascogaster, as in most groups of Braconidae, undue importance has been given tocolour-characteristics by many authors. In my opinion differences in colour are not sufficient todifferentiate species of Ascogaster. Shu-Sheng & Carver (1982) have provided experimentalconfirmation of the lability of colour in one species of Aphidiinae and have briefly reviewedprevious studies of this phenomenon in Parasitica. In some species (e.g. varipes, canifrons) thecarapace of some females is black, in others distinctly banded with yellow at the base. There areno intermediate stages between the two colour-patterns nor any associated morphologicaldiscriminants. In abdominator, on the other hand, all intermediates between extreme colour-variants occur. Presence or absence of strongly contrasted light-coloured bands on the legsappears to vary but little within a species, though the colour itself is subject to variation; this istherefore a useful characteristic for differentiation of species. Two species in the Palaearcticfauna have predominantly light-coloured heads (klugii, gonocephala); although there is con-siderable variation in head-colour within these species it is a useful ancillary character for theiridentification. Wing venation. In much previous work on Ascogaster significance has been ascribed to relativelyminor differences in venation. Wharton (1980), writing about Alysiinae, has shown theprobability of error inherent in the use of small differences in venation as taxonomic discrimi-nants, and his conclusion is probably equally valid when applied to the chelonines. Yoneda(1978) has analysed and illustrated variation in forewing venation in semenovi (= kyushuensis). Sculpture. In chelonines, sculpture is more strongly developed than that of other groups ofBraconidae. At first sight the taxonomic significance of the sculpture of Cheloninae is obscuredby its richness. Differences in sculpture in Ascogaster, however, are of considerable significance,particularly at species-group level. Thus sculptural characteristics of face, vertex, mesonotumand mesopleuron are of first importance. Sculpture of the carapace is sometimes useful in thediscrimination of species but the differences are generally rather too subtle to be usefultaxonomic characters. Presence or absence of rugosity on the hind coxa is sometimes a usefulspecific character. The morphological terms used are defined in Richards (1977). Depositories AS Academy of Sciences, Leningrad. BMNH British Museum (Natural History), London. CC Capek collection, Prague. EH Haeselbarth collection, Munich. ELKU Entomological Laboratory, Kyushu University, Fukuoka. HNHM Hungarian Natural History Museum, Budapest. IP Institut fur Pflanzenschutzforschung, Eberswalde. IPO Instituut voor Plantenziektenkundig Onderzoek, Wageningen. IRSNB Institut Royal des Sciences Naturelles de Belgique, Brussels. LC Lukas collection, Trencin. MC Maeto collection, Ibaraki. MNHN Museum National d'Histoire Naturelle, Paris. THE PALAEARCTIC SPECIES OF ASCOGASTER 347 NCM Norwich Castle Museum, Norwich. NM V National Museum of Victoria , Melbourne . NM Naturhistorisches Museum, Vienna. NR Naturhistoriska Riksmuseet, Stockholm. RNH Rijksmuseum van Natuurlijke Historic, Leiden. RSM Royal Scottish Museum, Edinburgh. SC Stelfox collection in USNM, Washington. TC Townes collection, American Entomological Institute, Ann Arbor. UEI Entomological Institute, Hokkaido University, Sapporo. USNM [U.S. National Museum] National Museum of Natural History, Washington D.C. ZC Zaykov collection, Plovdiv. ZI Zoological Institute, The University, Lund. ZMB Zoologisches Museum, Humboldt Universitat, Berlin. ZMC Zoologisk Museum, Copenhagen. Characteristics of subfamily Cheloninae More than 10,000 species of Braconidae have been described and it is probable that there are atleast 40,000 species in nature; it is therefore one of the largest families in the Hymenoptera. TheCheloninae, with at least 800 described species, is one of the larger subfamilies of the 22 intowhich the Braconidae is at present divided (see van Achterberg, 1976). The Cheloninae are distinguished from all other subfamilies of Braconidae by the possessionof a complete posterior mesopleural carina. This structure is generally absent in other Braconi-dae and, when present, is broadly interrupted in front of the mid coxae. All Cheloninae have aheavily sculptured gastral carapace formed by the fusion of tergites 1-3, the remaining parts ofthe gaster being normally concealed beneath the carapace. This feature occurs in a number ofgenera in other subfamilies, most of them not closely related to the Cheloninae; these taxatherefore bear a superficial resemblance to the Cheloninae, undoubtedly as a result ofconvergence. This matter has been dealt with in great detail in an excellent review byDudarenko (1974). In the Palaearctic region the genera most likely thus to be confused withCheloninae are Symphya (Alysiinae), easily distinguished by its exodont mandibles, Triaspisand Schizoprymnus (Helconinae sensu van Achterberg, 1976), which have only two submargin-al cells in the fore wing whereas chelonines have three. In other zoogeographical regions thereare genera possessing a gastral carapace also in, for example, Braconinae (Physaria), Rogadinae(Spinaria), Hormiinae (Cedrid), Microgasterinae (Fornicia, Buluka), Opiinae (Coleopius),Meteoridiinae (Urosigalphus). The gastral carapace is nonetheless a valuable characteristic forthe differentiation of Cheloninae if the possible exceptions are borne in mind. Chelonines alsoalways have three submarginal cells in the forewing and the ovipositor is generally not exsertedor, if it is, it is most often short and always has a slender needle-like apex. In the majority ofspecies the ovipositor is thick at the base, abruptly narrowed shortly before the apex, short andrarely exserted. In a few species the ovipositor is slender throughout its length and requires butlittle reduction in breadth to attain a fine point; this type of ovipositor is often exserted. In somespecies of Chelonus the ovipositor is thick at the base, evenly tapered to the apex and oftenstrongly upcurved. Two groups of genera of the Cheloninae are represented in the Palaearctic region; thePhanerotoma-group, in which the carapace is divided by crenulate sutures into three tergites andthe body colour is predominantly testaceous, and the Chelonus-group, in which the carapace isentire with no trace of sutures (except in occasional deformed specimens) and the body colour ispredominantly black. The latter group is represented in the Palaearctic region by two genera,Ascogaster and Chelonus, which can be distinguished as follows. 1 First submarginal cell always divided from first discal cell by vein Rs+M (Fig. 2); eyes neverconspicuously hairy ASCOGASTER Wesmael - First submarginal cell and first discal cell confluent, vein Rs+M never present (Fig. 1); eyes alwaysconspicuously hairy CHELONUS Panzer 348 T. HUDDLESTON There are several ancillary characters which are useful in placing some species in the two genera.The clypeus often has apical tubercles in Ascogaster but rarely in Chelonus. The face is generallystrongly rugose in Chelonus, sometimes punctate in Ascogaster. The hair on the face generallypoints downwards in Ascogaster. upwards in Chelonus. The males of some species of Chelonushave a supragenital aperture at the apex of the carapace; this characteristic is never present inAscogaster. ASCOGASTER Wesmael Ascogaster Wesmael, 1835: 226. Type-species: Ascogaster instabilis Wesmael (= abdominator (Dahl- bom)), by subsequent designation (Foerster, 1862: 244).Cascogaster Baker, 1926: 482. Type-species: Cascogaster fullawayi Baker, by original designation. [Synonymized by Watanabe, 1937: 75.] The species-groups The 30 Palaearctic species of Ascogaster are here divided into six informal groups based on thesculpture and vestiture of the face and the armature of the clypeus. I have made use of thespecies-group concept because it has the advantage of being flexible - its limits can be extendedor restricted without nomenclatural disturbance. It therefore seems to me to be ideally suited foruse in the Braconidae in which the fragmentary knowledge of the species in most generaprecludes their sensible division into formal subgeneric units. The semenovi- and caucasica-groups have in common a rather slender habitus with long legsand long slender gasters and concomitantly elongate female genitalia. The female carapace isoften modified posteroventrally either by the development of bizarre flanges or by the strongdepression of this part of the carapace. Two species in the caucasica-group do not have suchmodification of the carapace, but in these species the hypopygium is grossly elongated incaucasica and the ovipositor is elongate in bicarinata, the least modified species of its group. Themales in the two groups have conspicuously more elongate, slender gasters than their femalesand the male antennae are also considerably more elongate. These species-groups thereforeexhibit much stronger sexual dimorphism than is usual. The habitus of the abdominator- andannularis-groups has certain features in common with the preceding groups. Thus species of theabdominator-group are rather slender with elongate legs, though the female carapace is notmodified and the ovipositor and hypopygium are short. The degree of sexual dimorphism istherefore less in the abdominator-group. All the species of the latter group share the upwardlydirected facial vestiture of the semenovi- and caucasica-groups. The annularis-group alsoconsists of species which are rather elongate and slender, but the female carapace is generallybut little modified and the hypopygium and ovipositor only moderately elongate. The degree ofsexual dimorphism is even less marked in the annularis-group than in the abdominator-groupand only three of the five species of the former group have upwardly directed facial vestiture.The species of the semenovi-, caucasica-, abdominator- and annularis-groups are evidently moreclosely related to each other than to the remaining species of the genus. The bidentula- andquadridentata-groups of species have in common a generally short, squat habitus with rathershort, thick legs, a short, often deep carapace and a short hypopygium and ovipositor. The facialvestiture in these groups is always downwardly directed. Key to species 1 Ovipositor sheaths broad, flat (Figs 4, 6); ventral opening of carapace strongly narrowed posteriorly and with a strong flange posterolaterally but this not projecting beyond apex . of carapace; clypeus with dentate flanges laterally (Fig. 5) or a broad, spatulate medial projection (Fig. 3) 2 Ovipositor sheaths not conspicuously broad and flat; if ventral opening of carapacestrongly narrowed posteriorly and with a posteroventral flange (excisd) then ovipositorsheaths conspicuously slender and flange projecting beyond apex of carapace (Figs 23,24); clypeus with no dentate flanges laterally or broad spatulate projection medially(except longicornis, cf . Figs 41 , 42) 3 THE PALAEARCTIC SPECIES OF ASCOGASTER 349 2(1) Clypeus with conspicuous dentate flanges laterally (Fig. 5); carapace more pointed posteriorly (Fig. 6); ovipositor sheaths broader, knife-like (Fig. 6) semenovi Telenga (p. 352)Clypeus with no dentate flanges laterally but with a broad, spatulate medial projection(Fig. 3); carapace more rounded posteriorly (Fig. 4); ovipositor sheaths less broad, more evenly rounded at apex (Fig. 4) excavate Telenga(p. 352) 3(1) Apical border of clypeus with one or two large dentate tubercles medially (Figs 7-9, 13-15,17), laterally often strongly reflexed inwards; face always distinctly punctate, neverrugose, often reticulate-punctate, sometimes finely reticulate-rugulose; hairs on upper part of face directed upwards 4 Apical border of clypeus with very small medial tubercles (Figs 37, 51 , 54) or with none, orif with a single large, dentate tubercle medially than face strongly irregularly rugose(except perkinsi, Fig. 43); hairs on upper part efface directed downwards (except for asmall area between bases of antennae), or if occasionally directed upwards, face strongly , transversely rugose (exigua, gonocephala, klugii) 10 4(3) Clypeus with a single apical tooth (Figs 7-9) ; hypopygium short, rarely projecting beyondapex of carapace and then by little (Figs 10-12); hind coxa generally strongly rugose, at least dorsally 5 Clypeus with two median apical teeth (Figs 13-15, 17): hypopygium generally long,projecting beyond apex of carapace (Figs 20, 22-24) (but generally not in bicarinata, Fig.21); hind coxa generally conspicuously smooth with minute punctures except dorsally reticulate-punctate 7 5(4) Apical border of clypeus produced into apical tooth (Fig. 7) strong, dentate tuberclebetween antennae; face shining, more sparsely punctate; carapace shorter more round-ed in lateral view (Fig. 12) dentifer Tobias (p. 359) Apical border of clypeus truncate (Figs 8 , 9) ; if tubercle present between antennae then it isweak; face matt, reticulate-punctate or rugulose; carapace longer, less rounded in lateral view (Figs 10,11) 6 6(5) Notaulices obsolescent , generally almost indistinguishable from punctures on mesonotum ;antenna longer, 33-36 segmented, dilated medially; ovipositor sheaths thicker (Fig. 10) ahdominator (Dahlbom)(p. 357)Notaulices weak but distinct; antenna shorter, 27-30 segmented, not dilated medially; ovipositor sheaths slender (Fig. 11) nachitshevanica Abdinbekova (p. 359) 7(4) Carapace of $ strongly narrowed distally (Fig. 19), strongly flattened dorsoventrally (Fig.20); finely and regularly reticulate rugose; $ hind leg orange-testaceous kasparyani Tobias (p. 356) Carapace of $ not strongly narrowed or dorsoventrally flattened distally, normallyrounded (Figs 21-24); more coarsely reticulate rugose, at least anteriorly, the longitu-dinal element predominant ; $ hind leg black , only tibia lighter in colour 8 8(7) Carapace of $ with a bifurcate flange posteroventrally (Figs 23, 24); propodeum broadlyimpressed medially; hypopygium long, narrow in lateral view (Fig. 23); ovipositorabruptly upcurved distally; labiomaxillary complex not exserted. excisa Herrich-Schaffer(p. 355) Carapace of $ with no bifurcate flange posteroventrally; propodeum not impressedmedially; hypopygium shorter, broader (Figs 21, 22); ovipositor evenly and slightlyupcurved over its whole length, not abruptly upcurved distally; labiomaxillary complex generally exserted , at least slightly (Figs 15-18) 9 9(8) Labiomaxillary complex strongly exserted , labial palps inserted at a point distinctly beyondmandibles (Figs 15, 16); hypopygium generally extruded (Fig. 22), always at leastslightly longer than hind basitarsus, often considerably so; genae in face view contracted(Fig. 15), face and clypeus not protuberant, rather straight in profile (Fig. 16) caucasica Kokujev(p. 354) Labiomaxillary complex weakly exserted, labial palps inserted at a point distinctly behindmandibles (Figs 17, 18), hypopygium generally not extruded, always distinctly shorterthan hind basitarsus (Fig. 21); genae in face view more rounded (Fig. 17); face and clypeus protuberant, rounded in profile (Fig. 18) bicarinata Herrich-Schaffer (p. 354) 10(3) Antenna short, generally less than 30 segments in $ (except sometimes in gonocephala andin exigua); apical border of clypeus rounded, neither produced nor with teeth; carapacerather flat, sometimes with a posteroventral notch (Figs 33, 34); ovipositor slender, long;sheaths slender; radius emitted from distal third of pterostigma (Fig. 25) 11 350 T. HUDDLESTON Antenna long, generally more than 30 segments in $ (except brevicornis); apical border ofclypeus either produced or with one, two or three teeth (rounded in varipes, reticulata,armata); carapace rounded, generally deep, without posteroventral notch (Figs 60, 61,69, 70, 74). Ovipositor short, thick, abruptly narrowed shortly before apex; sheathsshort , wide ; radius generally emitted from about middle of pterostigma (Fig .2) 15 11(10) Face completely irregularly rugose, transverse-striate element often predominating; hairson upper part of face directed upwards; head generally pale in colour, sometimescompletely so except for ocellar triangle , sometimes partly so , rarely completely black ... 12Face sometimes punctate, shining, sometimes finely regularly reticulate-rugose; hairs onupper part of face directed downwards (except sometimes between bases of antennae);head always completely black save only occasionally the clypeus 14 12(11) Temple straight, strongly contracted, produced posterolaterally into strongly backwardlydirected flanges (Fig. 26); vertex rather depressed; strong dentate tubercle between bases of antennae gonocephala Wesmael (p. 362) Temple rounded, not strongly contracted, not produced posterolaterally (Figs 27, 28);vertex convex, not depressed ; no tubercle between bases of antennae 13 13(12) Carapace in 9 not conspicuously long and narrow, at most about twice as long as broad,reaching maximum breadth in distal half and roundly tapered to apex (Fig. 35) klugii (Nees von Esenbeck)(p. 363) Carapace in $ elongate, narrow, about 2-5 times as long as broad, broadest in proximalthird and tapering to apex (Fig. 36) exigua sp. n.(p. 361) 14(1 1) Carapace of $ emarginate posteroventrally (Fig. 34) , often tapered and slightly truncate indorsal view; temple at least equal to eye in dorsal view (Fig. 29); carapace always with two yellow patches at base annularis (Nees von Esenbeck)(p. 360) Carapace of $ at most slightly emarginate posteroventrally, rounded in dorsal and lateralview (Figs 31, 32); temple shorter than eye in dorsal view (Fig. 30), strongly contracted;carapace sometimes without yellow patches at base grahami sp . n . (p . 362) 15(10) Face generally completely smooth , punctate , sculpture of clypeus not contrasting stronglywith that of face; face sometimes finely and regularly reticulate-rugose (varipes);mesonotum generally predominantly punctate; clypeus rounded apically or produced(Fig. 45) or emarginate (Fig. 42), sometimes with two (Figs 51, 54) or three (Fig. 37) medial teeth, occasionally with one (Fig. 43) 16 Face completely, strongly irregularly rugose, never finely reticulate, sculpture of clypeusoften smooth, punctate in contrast to that efface; mesonotum generally coarsely rugoseso that notaulices indistinct; clypeus generally produced medially with a single medialtubercle (except in armata, reticulata) , never with more (Figs 62 , 72 , 79) 23 16(15) Mandible at base with a deep, semicircular depression (Fig. 38); apical border of clypeustransversely impressed, without medial teeth or excision; hind coxa always strongly, transversely striate, at least in part varipes Wesmael (p. 370) Mandible at base with a vertical, parallel-sided depression; if apical border of clypeustransversely impressed then medially excised or with teeth or tubercles (Figs 37, 51);hind coxa largely smooth punctate 17 17(16) Apical border of clypeus medially with three small but distinct dentate tubercles (Fig. 37) ;mandible broad, not strongly twisted, ventral border reflexed forward into a distinct flange (Fig. 40); carapace broad, short (Fig. 39) rufidens Wesmael (p. 369) Apical border of clypeus medially with a small excision flanked by two small tubercles (Figs51, 54), or produced into a blunt point (Fig. 45), or emarginate (Fig. 42); mandiblerelatively slender, twisted, without a distinct flange on ventral border but often with acarina from between the teeth to the ventral border at base; carapace generally lessbroad (Figs 44, 53), sometimes longer (Figs 46, 48, 52) 18 18(17) Antenna very long, 47-50 segmented; apical border of clypeus broadly shallowly emargin-ate medially (Fig. 42); mesopleuron dorsally completely coarsely reticulate-rugose, precoxal suture not distinct longicornis sp. n.(p. 368) Antenna shorter, at most 41-segmented; apical border of clypeus medially either excisedwith two small tubercles or produced (Figs 45, 51, 54); mesopleuron dorsally at least inpart smooth, punctate , precoxal suture distinct 19 19(18) Carapace elongate, clavate, always broadest in posterior third (Figs 46, 48); apical border of clypeus produced medially , generally without medial excision or tubercles (Fig. 45) ... 20Carapace short, broadest at about mid-point (Figs 44, 53), if elongate clavate (Fig. 52) then THE PALAEARCTIC SPECIES OF ASCOG ASTER 351 clypeus not produced medially and with a distinct medial excision flanked by twotubercles (Fig. 54) 21 20(19) Carapace finely rugose and with a downwardly directed anterior flange (Fig. 47), antennashorter, 33-34 segmented; hind leg yellow, except coxa at base black, apex of femur and sometimes of tarsus infuscate consobrina Curtis (p. 367) Carapace more coarsely rugose and with anteroventral flange not directed downwards;antenna longer, 37-39 segmented; hind leg dark with base of tibia and of tarsus ivory orpale yellow albitarsus Reinhard(p. 364) 21(19) Clypeus with a single medial apical tooth (Fig. 43); clypeus not distinctly divided from face; propodeum and carapace finely regularly reticulate-rugose perkinsi sp. n. (p. 368) Clypeus with two medial apical teeth, generally with a small excision between them (Figs51, 54); if excision obsolete than medial clypeal border a narrow, spatulate projection,never a medial tooth; clypeus distinctly divided from face; propodeum and carapacemore coarsely reticulate-rugose 22 22(21) Head massive, distinctly rounded behind eyes (Fig. 49), broader than mesonotum; carapace long (Fig. 52) ; propodeum long dorsally arisanica Sonan (p. 365) Head less massive, contracted behind eyes (Fig. 50), about equal in breadth to mesono-tum; carapace short (Fig. 53); propodeum short dorsally bidentula Wesmael(p. 366) 23(15) Antenna very short, 21-23 segmented in female; apical segments of flagellum distinctly separated (Fig . 67) with copious long erect bristles brevicornis Wesmael (p . 373) Antenna longer, generally more than 30 segmented in female; flagellar segments notdistinctly separated and hairs shorter, adpressed 24 24(23) Head behind eyes strongly contracted , temples rather straight (Fig. 59) ; genae in face view contracted, long (Fig. 58); $ antenna long, 39-40 segmented canifrons Wesmael (p. 374) Head behind eyes not strongly contracted, often at least slightly expanded, always rounded(Figs 56, 63, 66, 68, 78) genae in face view rounded (Figs 62, 76, 79); $ antenna shorter,30-36 segmented 25 25(24) Interantennal carina strongly raised into an erect triangular flange between scapes (Fig.57); apical border of clypeus rounded with at most a weak trace of a medial tubercle;propodeum with a weak medial transverse carina which is never raised into strong medial tubercles armata Wesmael(p. 372) Interantennal carina present but never strongly raised into a triangular flange betweenscapes; apical border of clypeus produced medially, generally with a strong medialtubercle (except reticulata, scabricula); propodeum with strongly raised medial tubercles (except rufipes) 26 26(25) Temple generally equal in length to eye in dorsal view (Figs 63 , 66) ; carapace short , deep in lateral view, posteroventral rim distinctly in front of apex (Figs 60, 61) 27 Temple generally conspicuously longer than eye in dorsal view, rounded (Figs 68, 78); carapace longer, posteroventral rim less conspicuously in front of apex (Figs 69, 70, 74) .. 2827(26) Apex of clypeus not impressed, produced, with a distinct medial tubercle (Fig. 62); hindtibia pale at base but never with a distinct pale band medially; hind coxa striate dorsally, generally black quadridentata Wesmael (p. 376) Apex of clypeus transversely impressed, its edge more or less straight with no medial tubercle (Fig. 65); hind tibia with a medial pale-coloured band; hind coxa smooth, punctate dorsally , generally yellow , sometimes infuscate at base reticulata Watanabe (p . 377) 28(26) Carapace rather narrow, almost parallel-sided in dorsal view (Fig. 77), slender in lateral view (Fig. 69); carapace of $ pale yellow in proximal third; legs pale, all coxae yellow, at least in part rufipes (Latreille) (p. 378) Carapace more rounded or tapered in dorsal view (Figs 73, 75), stouter in lateral view (Figs 70, 74); carapace of 9 always black; legs darker, all coxae black 29 29(28) Clypeus not distinctly divided from face, apex with a conspicuous dentate tuberclemedially (Fig. 72); mandible produced ventrally at base with no flange beneath dispar Fahringer(p. 375) Clypeus divided from face by a deep groove, apex forming a blunt point medially but withno dentate tubercle (Fig. 76); mandible not produced ventrally at base, with a flangebeneath (Fig. 71) scabricula (Dahlbom)(p. 379) 352 T. HUDDLESTON The semenovi-group Face punctate, generally reticulate-punctate, hairs on upper part efface pointing upwards. Clypeus withlateral dentate flanges or a broad medial projection, never with dentate tubercles. The two species of this group are known only from the eastern part of the Palaearctic region, andare uncommon. They both have unusually stout female genitalia and the posteroventral part ofthe female carapace is strongly excised and laterally flanged, presumably to facilitate the use ofthese enlarged organs. Ascogaster excavata Telenga(Figs 3, 4) Ascogaster excavatus Telenga, 1941: 311. Lectotype <J>, U.S.S.R.: Voronezh, Shipovo, 17.vi.1898 (Silan- t'ev) (AS) [examined]. [Lectotype selected by Tobias.]Ascogaster kasachstanicus Tobias, 1964: 183. Holotype cf , U.S.S.R.: Kazachstan, Tselinogr. oblast, 17 km from mouth of river Shapdara, 21. vi. 1957 {Tobias) (AS) [examined]. Syn. n. The data of the lectotype agree with those cited for one of Telenga's (1941: 312) syntypes ofexcavata. Tobias selected and labelled the lectotype but I have been unable to discover where orwhether the designation was published. The holotype of kasachstanica, a male, of course lacks the extremely modified carapace whichmakes excavata so distinctive. Tobias was evidently misled by this extreme dimorphism and sodescribed kasachstanica as distinct. The latter species, however, has most of the characteristicswhich distinguish excavata and I have no doubt that they are conspecific. In the original materialof excavata, Telenga had specimens of this species from Kazachstan. 9- Antennae missing. Head contracted behind eyes but not strongly so. Temple about equal in length toeye in dorsal view, not strongly rounded. Ocelli moderately large, OO = 3-0 OD; ocellar triangle acute.Frons behind antennae not strongly excavate, with a strong medial carina from face almost to anteriorocellus. Eyes not strongly protuberant. Face about twice as broad as high, not strongly protuberant,reticulate-punctate. Hair on upper part of face directed upwards and forwards, sparse. Clypeus about twiceas broad as high, not strongly protuberant, reticulate-punctate; apical border produced medially into abroad, blunt projection without tubercles. Mandibles long, slender, strongly twisted. Pronotum projects infront of mesonotum; lateral surface smooth anteroventrally with minute punctures, strongly reticulate-rugose dorsally. Notaulices foveolate, rest of mesonotum reticulate-punctate. Precoxal suture distinct,foveolate; rest of mesopleuron densely punctate, dorsally reticulate-punctate except anterodorsallyrugose. Propodeum reticulate-rugose. Carapace long, oval, rather pointed distally in dorsal view;posteroventrally rather truncate, the ventral rim only narrowly separated posteriorly and expanded into astrong flange which is deeply notched medially; smooth, shining punctate in distal half, dorsally andanteriorly reticulate-punctate with some weak rugosity. Hypopygium long, narrow, projecting beyondventral rim of carapace. Ovipositor thick at base, abruptly narrowed shortly before apex; sheaths broad,flat. Hind coxa smooth, punctate. Colour black; gaster and legs brown except tibiae yellow. Cf. Same as $ except carapace narrower and with posteroventral flange less strongly raised, notemarginate medially. HOSTS. Unknown. REMARKS. This species is closely related to semenovi from which it can be distinguished by thecharacteristics cited in the key to species. A. excavata is conspicuously smooth - no otherPalaearctic species of Ascogaster has as smooth and punctate a carapace. Ascogaster semenovi Telenga(Figs 5, 6) Ascogaster semenovi Telenga, 1941: 310. Holotype 9, MONGOLIA: Alashan, Dyn-juan-in, 23. vi. 1908 (Koslov) (AS) [examined].Ascogaster kyushuensis Yoneda, 1978: 291. Holotype $, JAPAN: Kyushu, Fukuoka Pref., Fukuoka city, Wakozaki, 22.V.1975 (Yoneda) (ELKU) [not examined]. Syn. n. THE PALAEARCTIC SPECIES OF ASCOGASTER 353 Telenga cited data for only one specimen of semenovi. I have examined the specimen bearingprecisely these data and, agreeing with Telenga's description of semenovi, no doubt thisspecimen is the holotype. It has lost its head but is otherwise well preserved. Yoneda'sdescription and excellent figures of kyushuensis are quite sufficient for this species to beidentified and the name placed as a synonym of semenovi. $. Antenna short, 22-segmented. Flagellum dilated medially, strongly tapered to apex, distally serratebeneath; medial segments at least as broad as long, some broader than long, rest of segments distinctlylonger than broad but only three basal ones as much as twice as long as broad. Head broad, rounded behindeyes. Temple about equal to eye in dorsal view. Ocelli small, OO = 4OD; ocellar triangle acute. Fronsbehind antennae moderately depressed, strongly transversely striate-rugose. Eyes moderately protuber-ant. Genae contracted below. Face slightly protuberant, about twice as broad as high, reticulate-punctate;densely hairy, the hairs on the upper part of face pointing upwards. Clypeus moderately protuberant,densely punctate though slightly less densely than face; apical border rounded medially, without teeth orincision, produced laterally into broad dentate flanges which point forwards and downwards in front of restof clypeus. Mandible long, twisted, the teeth blunt, broad, not distinctly separated. Pronotum projectinglittle in front of mesonotum; laterally rugose-foveolate. Notaulices deep, foveolate; rest of mesonotumpunctate except where notaulices coalesce in a reticulate-rugose area. Precoxal suture deep foveolate; restof mesopleuron punctate but with a deep foveolate groove anterodorsally, the anterior end of this grooveand middle of precoxal suture joined by a further foveolate groove. Propodeum strongly reticulate-rugose,with a medial transverse carine dividing it into a short dorsal surface and a long posterior surface; the dorsalsurface with a medial areola formed by two short medial longitudinal carinae. Carapace long, slender,sharply pointed in dorsal and lateral views, finely reticulate-rugose except posterolaterally polished,punctate; ventral rim strongly converging posteriorly, with a broad flange which is strongly incisedmedially. Hypopygium very long, projecting beyond ventral rim of carapace. Ovipositor long, broad,abruptly narrowed shortly before apex. Hind coxa smooth, punctate. Colour black; sternites and legs brown except fore tibia yellow, ovipositor sheaths translucent, palebrown. Cf . Same as $ except antenna longer, 27-28 segmented; flagellum less strongly dilated, not distallyserrate, all segments longer than broad with copious tyloidae and a conspicuous band of light-colouredhairs at junction of each segment; occipital carina joins genal carina a little way behind mandibles andexpanded into a prominent flange beneath base of mandible; ventral rim of carapace not expandedposteriorly into a flange, generally less strongly pointed. MATERIAL EXAMINED 9 Cf, 3 $. Japan: 7 cf , Fukuoka city, Hakozaki, 22.V.1978 (Maeto) (BMNH; MC); 2 cF, Fukuoka city,Ohori, 22.V.1976 (Maeto) (MC). Mongolia: holotype $ of semenovi; 2 <j>, 'de Cha Tcheou a Kan Tcheou'Nan Chan, north slope, 1000-2000 m, vi.1908 (Vaillanf) (BMNH; MNHN). HOSTS. Unknown. REMARKS. This remarkable species is easily distinguished by the dentate flanges on its clypeusand by the characteristics of the ovipositor sheaths. The caucasica-group Face punctate, generally reticulate-punctate, hairs on upper part of face pointing upwards. Clypeusmedially excised, with two prominent dentate tubercles. The four species of this group are commoner in the eastern part of the Palaearctic region thoughthe range of excisa extends as far west as France and that of bicarinata to Spain. All species in thisgroup have exceptionally long, slender female genitalia and a large hypopygium. The clypealteeth are least well developed in kasparyani which also has the most highly modified femalecarapace of any in the group. 354 T. HUDDLESTON Ascogaster bicarinata (Herrich-Schaffer) (Figs 17, 18, 21) Chelonus bicarinatus Herrich-Schaffer, 1838: 154. Syntypes. GERMANY (lost). Ascogaster mlokossewitchiKokujev, 1895: 78. Holotype $,U.S.S.R.: 'Lagodechi' (Mlokossewitsch) (AS)[examined]. [Synonymised by Sz6pligeti, 1908: 409.] Reinhard (1867: 368) redescribed bicarinata after having examined Herrich-Schaffer's originalmaterial. This detailed redescription is precise and I have based my interpretation of the speciesupon it. I do not believe that Reinhard could have overlooked the conspicuously exsertedlabiomaxillary complex of rostrata (= caucasica) or that he would have failed to mention it had itbeen present in the syntype of bicarinata. I therefore reject Fahringer's (1934: 517) placement ofrostrata in synonymy with bicarinata. 9- Antenna long, 25-26 segmented. Flagellum generally evenly tapered from base to apex, occasionallyslightly dilated medially; all segments considerably longer than broad except terminal three or four whichare quadrate. Head broad, distinctly broader than mesonotum, roundly contracted behind eyes. Templeabout equal to eye in dorsal view. Occiput deeply concave. Ocelli small, OO = 5-0 OD; ocellar triangleobtuse but ocelli not on line. Frons behind ocelli slightly depressed. Eyes protuberant. Malar space long;genae in face view moderately convex. Face moderately protuberant, slightly less than twice as broad ashigh, reticulate-punctate, often with some weak rugosity medially; hair on upper part of face pointsupwards. Clypeus protuberant, slightly less densely punctate than face; apical border folded inwards withtwo strong dentate tubercles medially. Labiomaxillary complex projecting more or less beneath the head;exserted part never as long as malar space; labial palps inserted behind mandibles. Mandibles long,slender, strongly twisted. Prothorax projecting in front of mesonotum; pronotum dorsolaterally rugose,rest punctate. Notaulices foveolate, coalescing posteriorly in a reticulate-rugose area; rest of mesonotumpunctate. Precoxal suture shallow, foveolate; rest of mesopleuron densely punctate except anterodorsallyrugose. Propodeum not distinctly divided into dorsal and posterior surfaces; strongly rugose, often thelongitudinal rugae predominating; with two weak posterolateral tubercles. Carapace oval, ratherflattened; strongly reticulate-rugose, the longitudinal element predominating except distally where thesculpture is weaker and more regular reticulate-rugose. Hypopygium short, not projecting beyondposteroventral rim of carapace; distinctly shorter than basal segment of tarsus. Ovipositor long, thin,gradually tapered, upcurved. Hind coxa smooth, punctate, sometimes with weak rugae dorsally. Colour black. Carapace sometimes orange-testaceous at base; apex of femur, tibia and tarsus of foreleg,tibia of midleg and medial band of tibia of hindleg testaceous. Cf . Same as 9 except antenna longer, 30-32 segmented; all flagellar segments distinctly longer thanbroad, terminal segments not quadrate; carapace more slender in dorsal view, flatter in lateral view; legsdarker, generally completely black except for femur at apex and tibia testaceous, occasionally mid and hindtibiae somewhat lighter but never testaceous. MATERIAL EXAMINED32 cf , 35 9. Bulgaria, Greece, Spain, Turkey, U.S.S.R., Yugoslavia. HOSTS. No reared material examined. REMARKS. This species is closely similar to caucasica with which it has generally been confused.The labiomaxillary complex is often more or less exserted in bicarinata but never as strongly as incaucasica', the characteristics of the hypopygium, ovipositor and head that are cited in the keyserve amply to differentiate the two species. Both bicarinata and caucasica are generally darkerthan most other species of Ascogaster but the 9 carapace of bicarinata is sometimes tinged withorange-testaceous, that of caucasica never is. Ascogaster caucasica Kokujev(Figs 15, 16, 22) Ascogaster caucasicus Kokujev, 1895: 82. Holotype 9, U.S.S.R.: 'Lagodechi', v.1881 (AS) [examined].Ascogaster rostratus Szepligeti, 1896: 178. Lectotype cf, HUNGARY: Budapest, Gellerth, 19. v. 1895(Szepligeti) (HNHM), designated by Papp in Shenefelt, 1973: 817 [examined]. Kokujev stated that he had a single female of caucasica; this specimen, bearing the data cited inthe description, is now in AS, Leningrad. Fahringer (1934: 517) synonymised rostrata with THE PALAEARCTIC SPECIES OF ASCOGASTER 355 bicarinata Herrich-Schaffer and the two species have often been confused in collections, but thecharacteristics distinguishing them are given above in the key to species. Telenga (1941: 316)synonymised caucasica with bicarinata, evidently following Fahringer's misinterpretation of thelatter species. $. Antenna moderately long, 25-26 segmented. Flagellum not conspicuously dilated medially, slightlytapered apically; basal segment distinctly more than three times as long as broad, each following segmentslightly shorter until preapical segment which is about as long as broad. Head contracted behind eyes.Occiput deeply concave. Temple about equal to length of eye in dorsal view. Ocelli minute, OO = 5 OD;widely separated. Eyes small, round, protuberant. Malar space long, about 1-5 times basal breadth ofmandible. Genae in face view contracted, not strongly rounded. Face almost twice as broad as high, ratherflat, not strongly protuberant, densely reticulate-punctate, often with a distinct rugose element. Hair onupper part of face directed upwards. Clypeus not protuberant, weakly and evenly convex, apical borderreflexed backwards at sides, medially produced into two prominent dentate tubercles; polished anddensely punctate, sometimes reticulate. Mandible long, strongly twisted, ventrally expanded at base.Labiomaxillary complex elongate, projecting conspicuously beneath head; exserted part distinctly longerthan malar space; labial palps inserted at least one third of distance along exserted part, distinctly beyondmandibles. Prothorax projecting in front of mesonotum with a strongly impressed foveolate groovemedially. Notaulices shallow, foveolate, coalescing posteriorly in a coarsely rugose area; rest of mesono-tum polished, reticulate-punctate. Precoxal suture irregularly rugose; rest of mesopleuron reticulate-punctate. Propodeum rounded, not abruptly divided or posteromedially impressed; coarsely irregularlyrugose with no distinct carinae but with small lateral tubercles. Carapace long, slightly acuminate, notrounded posteroventrally, reticulate-rugose, the longitudinal element predominating anteriorly. Hypopy-gium exceptionally long, sometimes projecting beyond the posteroventral rim of carapace. Ovipositor verylong, slender, gradually tapering to apex; generally covered by hypopygium for most of its length butoccasionally protruding to a distance at least equal to the length of the hind tibia. Colour black; wings infumate, legs yellow-marked. Cf . Same as $ except that antenna longer, 32-33 segmented; carapace narrower, flatter, slightly roundedposteroventrally . MATERIAL EXAMINED 21 Cf, 15 $>. Bulgaria: 2 cf, 1 $, Rhodopi, Sh. poljana, v.1976 (Zaykov) (ZC); 1 $>, Rhodopi, D.Lukovo, v.1977 (Zaykov) (ZC). Cyprus: 1 $, Yermosoyia, 3.iv.l978 (Teunissen) (RNH); 1 $, EpiscopiForest, 20-23.iv.1950 (BMNH); 1 cf, Yerolakko, 25.iv.1948 (Longfield) (BMNH); 3 cf, Limassol,l-S.v.1934 (Mavromoustakis) (BMNH); 1 cf, Kellaki, 2000 ft, 10.iv.1952 (BMNH); 1 $, Nicosia,20-23.iv.1934 (Mavromoustakis) (BMNH); 2 cf, Yerasa, 24.iii.1947 (Mavromoustakis) (BMNH).Czechoslovakia: 1 cf , 1 $, Sturovo, 25.V.1959 (Strejcek) (CC). Greece: 3 cf , 1 $>, Prov. Attiki, Marathon,21.iv.1978 (Papp) (HNHM); 3 cf , 1 $, Athens, Mt Pendeli, 25.iv.1980 (Teunissen) (RNH); 1 cf , Delphi,22-23.iv.1980 (Teunissen) (RNH); 1 cf, Lesbos, Milies, 21.V.1980 (Teunissen) (RNH); 1 $, Kerkyra,16-30. v. 1971 (Aartsen & Wolschrjn) (RNH). Hungary: lectotype cf of rostratus. Iran: 1 Cf, 3 $, Susa(Escalera) (BMNH). Spain: 1 cf, Ventas, St Barbara, 29.iii.1980 (Teunissen) (RNH). Turkey: 1 $,Adana, 7.v. (Vachat) (MNHN). U.S.S.R.: holotype $ of caucasicus. Yugoslavia: 1 $, Lake Ochrid,16.vi. 1958 (Coe) (BMNH). HOSTS. No reared material examined. REMARKS. The strongly exserted labiomaxillary complex of this species immediately disting-uishes it from any other Palaearctic species of Ascogaster. The strongly dentate clypeus, theupwardly pointing facial vestiture and the elongate, slender habitus of caucasica show its closerelationship with bicarinata, excisa and kasparyani. Ascogaster excisa (Herrich-Schaffer)(Figs 14, 23, 24) Chelonus excisus Herrich-Schaffer, 1834: 154. Syntypes, GERMANY (lost). Ascogaster longiventris Tobias, 1964: 184. Holotype cf, U.S.S.R.: Kazachstan, Kokshetay, 15.vi.1957(Tobias) (AS) [examined]. Syn. n. The identity of this species is clear from Herrich-Schaffer's description and from his excellentfigure. The remarkable posteroventral modification of the female carapace makes excisa one ofthe least liable to misinterpretation of all species of Ascogaster. 356 T. HUDDLESTON $ . Antenna long, 28-31 segmented. Flagellum not dilated medially, strongly tapered apically ; basal five orsix segments at least twice as long as broad, rest distinctly longer than broad except apical seven or eightwhich are almost quadrate. Head not strongly contracted behind eyes. Temple slightly shorter than eye indorsal view. Occiput deeply concave. Ocelli minute, OO = 4 OD; not on line. Eyes protuberant. Malarspace only slightly greater than basal breadth of mandible. Face moderately protuberant, about twice asbroad as high, densely and regularly reticulate-punctate; hairs on upper part of face directed upwards.Clypeus moderately protuberant, reticulate-punctate; apical margin incised medially with two bluntdentate tubercles flanking the incision, laterally weakly reflexed backwards. Mandible large, only slightlytwisted, punctate. Pronotum not projecting strongly in front of mesonotum; laterally reticulate-punctateexcept for a medial longitudinal foveolate groove. Notaulices deep, foveolate, coalescing posteriorly in arugose-foveolate area; rest of mesonotum reticulate-punctate. Precoxal suture broad, reticulate-foveolate; rest of mesopleuron reticulate-punctate, except dorsally rugose. Propodeum posteromediallyimpressed; anteriorly with two short longitudinal carinae which terminate in two blunt tubercles one oneach side of medial depressed area; reticulate-rugose. Carapace about twice as long as broad, ventral rimproduced posteriorly into a broad, bifurcate flange. Hypopygium long and rather narrow in side-view,generally extruded and projecting behind the carapace. Ovipositor very long, slender; abruptly upcurvedshortly before apex. Hind coxa generally smooth, polished with only weak puncturation, occasionallyreticulate-punctate anterodorsally, sometimes also with obsolescent transverse rugae dorsally. Colour black; legs brownish-testaceous except all coxae and trochanters black, hind femur and all tarsiinfuscate. Cf . Same as $ except antenna longer; flagellum not tapered apically, all segments distinctly longer thanbroad; carapace narrower in dorsal view, posteroventral rim distinctly in front of apex; legs darker, midand hind femur black, fore femur dark proximally. MATERIAL EXAMINED 23 cf, 19 $. Bulgaria: 22 cf, 2 $>, Rhodopi, Bojno, 6.vii.l976 (Zaykov) (ZC); 5 $, same data except24.vii.1975. France: 3 $, Drome, Col de Macuegne, i.viii.1979 (Graham) (BMNH); 2 $, Lozere, R. Lot,Mende, 3-8.vii.1924 (Moseley) (BMNH); 1 $, Chateau Arnoux, 10.viii.1972 (Boudek) (BMNH); 4 $,Dept. Vaucluse, Castellet, 20.vii.1974 (Gijswijt) (RNH); 1 $, Var, Montauroux, l.vii.1960 (Van derVecht) (RNH); 1 $, Var, Ste Baume, 19.vii.1951 (Granger) (MNHN). U.S.S.R.: holotype cf oflongiventris . HOSTS. No reared material examined. REMARKS. The rounded carapace with the posteroventral rim distinctly in front of the apex is auseful supplementary character for distinguishing the male of excisa from those of rostrata andbicarinata in which the carapace of the male is not strongly rounded posteriorly and theposteroventral rim is only slightly in front of the apex. In describing longiventris, Tobiasevidently had not appreciated the considerable sexual dimorphism of this species; nonethelesshe correctly pointed out its relationship with bicarinata. Ascogaster kasparyani Tobias(Figs 13, 19, 20) Ascogaster kasparyani Tobias, 1976: 233. Holotype cf, U.S.S.R.: Gruzia, Bogdanovka, 28.vi.1967(Kasparyan) (AS) [examined] . $. Antenna long, slender, more than 28-segmented. Flagellum slender, not dilated medially, first segmentabout three times as long as broad, next 10 or so segments at least twice as long as broad, rest at least slightlylonger than broad. Head not strongly contracted behind eyes. Temple slightly shorter than eye in dorsalview. Occiput deeply concave. Ocelli small, OO = 3-0 OD; not on line. Eyes small, protuberant. Malarspace long, about twice basal breadth of mandible. Face about twice as broad as high, moderatelyprotuberant, coarsely reticulate-punctate; hairs on upper part of face directed upwards. Clypeus protuber-ant, narrower than face; apical margin incised medially, weakly reflexed laterally. Mandible moderatelytwisted. Pronotum projecting strongly in front of mesonotum; laterally reticulate-rugose. Notaulicesshallow, indistinct reticulate-rugose, coalescing posteriorly in a strongly reticulate-rugose area; rest ofmesonotum reticulate-punctate. Precoxal suture indistinct, most of mesopleuron laterally irregularlyreticulate-rugose, ventrally reticulate-punctate. Propodeum rounded, completely regularly reticulate-rugose with two weak medial longitudinal carinae at base. Carapace long, slender, about 2-5 times as longas broad, gradually diminished in breadth from about proximal third; narrow and rounded distally in dorsal THE PALAEARCTIC SPECIES OF ASCOGASTER 357 view; in lateral view rather flat, the proximal tenth strongly dorsoventrally flattened; completely finelyregularly reticulate-rugose. Hypopygium at least slightly exserted. Ovipositor long, slender, stronglyupcurved. Hind coxa polished with minute punctures. Colour black, antenna at base and legs orange-testaceous, carapace more or less light-marked. Cf. Same as female except carapace rounded, not dorsoventrally flattened at apex; colour darker,carapace black, legs black except fore tibia yellow, mid and hind tibiae brown. MATERIAL EXAMINED 1 Cf, 2 $. Greece: 1 $, Rhodes, Kremasti Hills, 24-30.vi.1958 (Mavromoustakis) (BMNH); 1 $,Empona, 18-19.vi.1968 (Mavromoustakis) (BMNH). U.S.S.R.: holotype cf . HOSTS. Unknown. REMARKS. The long, exserted hypopygium, the long, slender, upcurved ovipositor and theupwardly directed hair on the upper part of the face show the relationship of kasparyani,caucasica, excisa and bicarinata. The characteristics of the carapace distinguish kasparyani fromall other Palaearctic species of the genus. The abdominator-group Face punctate, often reticulate-punctate, hairs on upper part of face pointing upwards. Clypeus with asingle prominent dentate tubercle. Two of the three species of this group are widely distributed in the more northerly parts of thePalaearctic region, the third, nachitshevanica, is known only from the U.S.S.R. and Mongolia.All species in the group have a short ovipositor and a short hypopygium which are generallyconcealed beneath the carapace. A. dentifer is rather different in head and body-shape from theother two species which form a close pair differentiated on small differences in the antennae, thenotaulices and the genitalia. Ascogaster abdominator (Dahlbom)(Figs 9, 10) Chelonus abdominator Dahlbom, 1833: 165. Holotype %, SWEDEN: 'Esperod', Fallen colln (ZI) [ex-amined]. [Synonymised with instabilis by Thomson, 1874: 583.] Ascogaster instabilis Wesmael, 1835: 227. Syntypes, BELGIUM: Wesmael colln (IRSNB) [examined]. Syn.n. Ascogaster fulviventris Curtis, 1837: folio 672. Holotype cf , GREAT BRITAIN: England, Curtis colln (NMV)[examined]. [Synonymised with instabilis by Fischer, 1965: 11.] Chelonus femoralis Herrich-Schaffer, 1838: 154. Syntypes, GERMANY (lost). [Synonymised with instabilisby Reinhard, 1867: 364.] Chelonus rufiventris Herrich-Schaffer, 1838: 154. Syntypes, GERMANY (lost). [Synonymised with instabilisby Reinhard, 1867: 364.] Ascogaster pallida Ruthe, 1855: 293. LECTOTYPE $, GERMANY: 'Freienwalde', 16.vii.1854, Ruthe colln(BMNH), here designated [examined]. [Synonymised with instabilis by Reinhard, 1867: 364.] There is a single $ in Fallen's collection, labelled 'C. abdominator' in Dahlbom's handwriting.There is no indication that Dahlbom had more than one specimen before him when describingthis species and I therefore accept this specimen as holotype of abdominator. Dahlbomobviously mistook the sex of his specimen, citing it as male in his description - the holotype bearsa label upon which this error is repeated. Thomson (1874: 583) Synonymised abdominator withinstabilis despite the fact that the latter species is clearly a junior synonym of abdominator.Thomson's placement has been followed by all subsequent workers, but there is no justificationfor continuing this incorrect usage. Wesmael (1835: 229) stated that he possessed six males andtwo females of instabilis. At present there are six males and five females labelled 'Ascogasterinstabilis mini' in Wesmael's collection; all of these specimens are conspecific and agree withWesmael's description of instabilis. Obviously, however, there are more female specimenslabelled as 'instabilis' than Wesmael claimed to possess, thus they cannot all belong to thesyntype series. A lectotype has not been selected because the species is so distinctive. There is 358 T. HUDDLESTON one male in Curtis' collection named fulviventris . It agrees in every particular with what Curtiswrote about fulviventris and I regard this specimen as the holotype of that species. The specimenwas labelled as 'type' by Nixon in 1948; it is well within the limits of variation of abdominator. Reinhard placed femoralis Herrich-Schaffer and rufiventris Herrich-Schaffer in synonymywith instabilis without indicating that he had seen the original material of femoralis. (Herrich-Schaffer himself stated in his description of rufiventris, that his original specimen of this specieshad been destroyed.) There is nothing in Herrich-Schaffer's descriptions of femoralis andrufiventris which precludes Reinhard's placement of them. There are two female specimens from Ruthe's collection in the BMNH which agree with hisdescription of pallida. One of them bears a label in Ruthe's handwriting 'Asc. pallidus m.Freienwalde 16.7.54'; these are precisely the data cited by Ruthe in his description of pallidaexcept that he gave the date of capture as '18.7.54'. This discrepancy is probably due to an errorof transcription, all too easily made when copying figures from a barely legible handwrittenlabel. These specimens are undoubtedly syntypic material and the female bearing Ruthe'shandwritten label is here designated lectotype of pallida. $. Antenna long, 32-36 segmented. Flagellum dilated medially, strongly tapered to apex; all segments atleast as long as broad. Head large, roundly contracted behind eyes. Temple about equal to eye in dorsalview. Occiput strongly concave. Ocelli small, OO = 3-0-3-5 OD; not on line. Eyes small, protuberant.Malar space about half height of eye. Face protuberant, about twice as broad as high, reticulate-punctate,sometimes rugose medially and occasionally completely so; pubescence short, fine and directed upwardsand outwards on upper part of face. Clypeus protuberant, reticulate-punctate, ventral border reflexedbeneath so that in front view clypeus appears truncate; medially with a large dentate tubercle. Mandiblestrongly twisted with numerous long inwardly-directed hairs at base. Pronotum projecting strongly in frontof mesonotum; dorsally with a strong transverse fold which becomes laterally a foveolate groove.Notaulices largely obsolete, visible at the anterior border of the mesonotum as a small reticulate-rugosearea, otherwise indistinguishable from the puncturation of the mesonotum. Mesonotum smooth reticulate-punctate. Precoxal suture broad, shallow, foveolate-rugose; mesopleuron laterally reticulate-rugose,ventrally punctate. Prepectal carina strongly raised into a flange behind each fore coxa, deeply notchedmedially. Propodeum rather flat, reticulate-rugose, carinae inconspicuous; generally with small medialand lateral tubercles. Carapace rounded posteriorly, only slightly rounded posteroventrally. Hypopygiumshort, truncate, never projecting strongly beyond end of carapace. Ovipositor short, straight, abruptlytapered shortly before apex. Colour black; antenna at base, forelegs and proximal part of carapace yellow or sometimes the latterivory, mid and hind legs at least partly light-marked; carapace sometimes completely yellow. Occasionallyclypeus, face and pronotum orange-yellow, at least in part. Cf . Same as $ except antenna longer, flagellar segments longer, thinner though the number of segmentsis not greater; pubescence on face longer and white in colour so that it appears much denser than that of $ ;carapace longer, narrower, sometimes almost parallel-sided anteriorly; colour darker, antenna generallyblack, carapace often completely dark. MATERIAL EXAMINED 267 cf , 127 9. Belgium, Bulgaria, Czechoslovakia, France, Germany, Great Britain, Ireland, Nether-lands, Sweden, U.S.S.R., Yugoslavia. HOSTS. No reared material examined REMARKS. This species is common and widespread in the region. The characteristics of theclypeus and the face and the virtual absence of notaulices easily differentiate abdominator fromall other Palaearctic species. Indeed, the latter characteristic alone is almost diagnostic though insome species in the caucasica-gTOup the notaulices are rather weak and in those species of thequadridentata-group in which the mesonotum is strongly rugose the notaulices are thereforeindistinct. This species is extremely labile in colour, varying from almost completely black toalmost completely yellow, though the majority of specimens are largely black-bodied exceptthat the carapace is more or less pale-marked. THE PALAEARCTIC SPECIES OF ASCOGASTER 359 Ascogaster dentifer Tobias(Figs 7, 12) Ascogaster dentifer Tobias, 1976: 236. Holotype cf, U.S.S.R.: Armenia, Kefansky region, Tsav. wood, 28.vi.1971 (Kulitskij) (AS) [examined].Chelonus punctulator Kirchner, 1867: 119. [Nomen nudum; attributed to Foerster.] $. Antenna long, 36-37 segmented, pedicel often protuberant posterolaterally. Flagellum weakly dilatedmedially, proximal segments distinctly longer than broad, 2-0-3-0 times as long as broad at base offlagellum, rest of segments about as long as broad, those at apex about half breadth of medial ones. Headnot strongly contracted behind eyes, rounded. Temple about 1-5 times length of eye in dorsal view. Occiputstrongly concave. Ocelli small, OO = 3-5-4-0 OD, not on line but in an obtusely angled triangle. Fronsdeeply excavate behind antennae, the depressed area smooth, polished. Strong dentate lamella betweenbases of antennae. Face protuberant, about 1-5 times as broad as high, shining, punctate; hair on upperpart of face directed upwards. Clypeus about twice as broad as high, convex, shining, punctate; apicalborder slightly reflexed inwards laterally, medially produced into a strong dentate tubercle. Mandiblestrongly twisted. Pronotum projecting distinctly in front of mesonotum; rugose. Propleurae reticulate-punctate. Notaulices distinct, foveolate, coalescing posteriorly in a reticulate-rugose area; rest of mesono-tum closely, even reticulately, punctate. Precoxal suture shallow, foveolate-rugose; rest of mesopleuronpunctate except anterodorsally rugose. Propodeum rather depressed, finely and regularly reticulate-rugose with two medial longitudinal carinae and one medial transverse carina, the latter raised into bluntcrests postero-laterally. Carapace oval, rather broadly rounded posteriorly, postero ventral rim consider-ably in front of apex. Ovipositor short, straight, stout, abruptly tapered shortly before apex. Sheathsclavate with brush of long hairs at apex. Hypopygium short. Hind coxa transversely rugose. Colour black; base of antenna testaceous, carapace with two yellow patches anterolaterally; pronotumsometimes more or less yellow-marked; fore trochanter, femur at apex, tibia, mid and hind trochanters andtibiae at base yellow or testaceous; all tarsi infuscate. Cf . Same as $ except antenna longer, all segments at least as long as broad, most distinctly longer thanbroad; carapace narrower in dorsal view with no yellow patches anteroventrally; legs often darker, lessextensively pale-marked. MATERIAL EXAMINED 77 cf , 1 1 9- Bulgaria, Czechoslovakia, France, Germany, Great Britain, Hungary, Italy, Switzerland,U.S.S.R., Yugoslavia. HOSTS. Unknown. REMARKS. This species is superficially similar toperkinsi; both have smooth, punctate faces and awell-developed medial tubercle on the clypeus and are similar in stature and colour. The facialhair of dentifer, however, always points upwards on the upper part of the face while that ofperkinsi does not, but the sparseness of the facial vestiture in dentifer may result in thischaracteristic being overlooked. However, in dentifer there is a well-developed dentate flangebetween the antennae which is not present in perkinsi, and the latter species has a conspicuouslongitudinal flange on the basal segment of the hind tarsus which is lacking in dentifer. I have examined four specimens (1 BMNH; 3 MNHN) of dentifer which are labelled'Chelonus punctulator Foerster' in Foerster's handwriting. No description was published byFoerster but the name Chelonus punctulator was listed by Kirchner (1867: 119). Ascogaster nachitshevanica Abdinbekova(Figs 8, 11) Ascogaster nachitschevanicus Abdinbekova, 1969: 69. Holotype 9> U.S.S.R.: Azerbaidzhan, Ordubad,20. v. 1967 (AdamidzK) (AS) [examined]. 9- Antenna short, 27-30 segmented. Flagellum not dilated medially, strongly narrowed at apex, allsegments distinctly longer than broad except terminal segments which are moniliform. Head broad,transverse, roundly contracted behind eyes. Temple 1-0-1-5 times length of eye in dorsal view. Occiputstrongly concave. Ocelli small, OO = 4-0 OD, almost on line. Eyes small, protuberant. Face slightlyprotuberant, about twice as broad as high, reticulate-punctate, sometimes weakly rugulose medially; hairon upper part of face directed upwards. Clypeus transverse, apically impressed and truncate with a strong 360 T. HUDDLESTON dentate tubercle medially; reticulate-punctate. Mandible long, slender, twisted. Pronotum projecting infront of mesonotum; rugose laterally. Notaulices shallow but distinct; rest of mesonotum reticulate-punctate. Precoxal suture broad, reticulate-foveolate; mesopleuron anterodorsally rugose, rest denselypunctate. Propodeum completely coarsely reticulate-rugose with a medial transverse carina and twomedial longitudinal carinae dorsally but these sometimes indistinct; no distinct propodeal tubercles.Carapace strongly narrowed posteriorly, generally also more or less dorsoventrally flattened in lateralview; posteroventral rim at apex of carapace; coarsely reticulate-rugose at base, more finely rugosetowards apex. Hypopygium at most slightly exserted. Ovipositor sheaths conspicuously slender. Hind coxasmooth, punctate except dorsally weakly rugose. Colour black; antenna at base, tegula, sometimes carapace anterolaterally and legs orange-testaceousexcept hind coxa dark, at least at base; sometimes mid and hind femora dark medially, hind tibia dark atapex; all tarsi infuscate. Cf . Unknown. MATERIAL EXAMINED 4 $. Mongolia: 3 $, Chovsgol aimak, 8 km N. of Somon Burenchaan, 1450 m, 20.vi.1968 (Kaszab)(HNHM). U.S.S.R.: holotype $. HOSTS. Unknown. REMARKS. This species is closely similar to abdominator . The characteristics cited in the key tospecies, together with differences in the shape of the carapace, are sufficient to distinguish thetwo species, but whether these differences would hold good in a longer series of nachitshevanicais open to question. The a/mu/aris-group Face punctate, hairs on upper part pointing downwards or rugose, hairs on upper part pointing upwards.Clypeus without medial excision or tubercles, apical border more or less convex. The five species in this group are all more or less small in size, rather flat and have long, slenderovipositors; the emission of the radius from conspicuously beyond the middle of the stigma is avenational feature characteristic of the group. Four of the species of this group are widelydistributed, at least in north-western Europe, although gonocephala is uncommon. The fifthmember of the group, exigua, is known only from the holotype from Mongolia. The latterspecies is the most highly modified of the group, having an exceptionally elongate, slendercarapace; it is also the only one with long antennae. Ascogaster annularis (Nees von Esenbeck) (Figs 25, 29, 33, 34)Sigalphus annularis Nees von Esenbeck, 1816: 264. Syntypes, GERMANY (lost). Nees von Esenbeck's description of annularis is clear and detailed enough for the species to beidentified with confidence as the one described below. $ . Antenna short, 21-24 segmented. Flagellum sometimes slightly dilated medially, moderately tapered toapex, all segments distinctly longer than broad except terminal three or four about as long as broad, oftenmore or less moniliform. Head broad, slightly broader than mesonotum, roundly contracted behind eyes.Temple about as long as eye in dorsal view or slightly longer. Occiput moderately concave. Ocellimoderately large, OO = 2-5-3-0 OD; not on line. Frons behind antennae shallowly excavate, polished. Aweak medial carina extends from a tubercle on upper part of the face, between antennae. Eyes large,protuberant. Malar space equal to about half height of eye. Face weakly convex, about twice as broad ashigh, shining, punctate, with sparse, downwardly pointing hairs. Clypeus weakly convex, shining,punctate; apical border slightly reflexed, evenly rounded or occasionally more or less straight medially,never with any trace of tooth or excision. Mandible small, strongly twisted. Pronotum projecting distinctlyin front of mesonotum; rugose except ventrolaterally smooth, punctate. Notaulices shallow, foveolate,coalescing posteriorly in a finely rugose area; rest of mesonotum smooth punctate. Precoxal suturereticulate- foveolate; rest of mesopleuron smooth punctate except anterodorsally rugose. Propodeumfinely irregularly rugose; divided by a transverse medial carina into dorsal and posterior surfaces, the dorsal THE PALAEARCTIC SPECIES OF ASCOGASTER 361 rather flat, with a distinct central areola. Carapace distinctly longer than thorax and propodeum; elongateoval and often truncate posteriorly in dorsal view, rather shallow in lateral view; posteroventrallyemarginate. Hypopygium long, often projecting beyond carapace. Ovipositor long, sheaths slender,clavate. Hind coxa finely rugose; hind tibia somewhat swollen. Radius emitted from stigma conspicuouslybeyond middle. Colour black; antenna brown, lighter at base; mandibles, foreleg testaceous, midleg testaceous excepttibia and tarsus infuscate, tibia with a light medial band; hindleg dark except trochanter and femur yellowor testaceous, at least in part, and tibia with an ivory or pale yellow band medially; carapace always withtwo yellow patches at base. Cf . Same as $ except antenna longer, 22-25 segmented; flagellum not dilated medially, all segmentsabout twice as long as broad; carapace not truncate posteriorly or emarginate posteroventrally. MATERIAL EXAMINED 65 cf , 74 9 Bulgaria, Czechoslovakia, France, Germany, Great Britain, Greece, Hungary, Netherlands,Sweden. HOSTS. Pandemis heparana (Denis & Schiffermuller) (Lepidoptera: Tortricidae), Recurvarialeucatella (Clerck) (Lepidoptera: Gelechiidae). REMARKS. This is one of the smallest Palaearctic species of Ascogaster, averaging about 2-5 mmin length. The small size of annularis together with its characteristic colour pattern of yellow atthe base of the gaster and pale-banded hind tibia make it readily recognisable. The only specieswhich shares these characteristics, though it often lacks the yellow on the carapace, is grahamiwhich is very closely related to annularis and can be differentiated by the characters cited in thekey. Ascogaster exigua sp. n. (Figs 28, 36) $ . Antenna long, 35 segmented. Flagellum weakly dilated medially, strongly tapered to apex, all segmentslonger than broad, apical ones only slightly longer than broad. Head broader than mesonotum, roundedbehind eyes. Temple slightly shorter than length of eye in dorsal view. Ocelli large, OO = 2-5 OD; not online. Face protuberant, about twice as broad as high, conspicuously transversely striate-rugose; hair onupper part pointing upwards. Clypeus narrow, convex, apical part weakly transversely depressed; smooth,punctate; apical border rounded, with no trace of medial tubercles. Mandibles small, strongly twisted.Pronotum projecting but little in front of mesonotum; laterally densely punctate, ventrally weakly rugose.Notaulices foveolate, coalescing posteriorly in a reticulate-foveolate area; rest of mesonotum shining,punctured, rather flat dorsally. Precoxal suture shallow, foveolate; rest of mesopleuron smooth, punctate.Propodeum flat dorsally, depressed, with medial transverse carina, two weak medial longitudinal carinaeand lateral longitudinal carinae; completely finely reticulate-rugose. Carapace elongate, about three timesas long as broad, broadest at about a third of its length from base; reticulate-rugose proximally, thelongitudinal element predominant, distally smooth shining with minute punctures; posteroventral rimreaching apex. Hypopygium not exserted. Hind coxa largely smooth, punctate. Colour black: antenna brown except testaceous at base; all legs testaceous except hind tibia, all coxaeand all tarsi infuscate. Cf . Unknown. MATERIAL EXAMINED Holotype $ , Mongolia: Zavchan aimak, Choit chunch, 26 km ENE. from See Tehnen nuur, 13.vii.1968(Koszab) (HNHM). HOSTS. Unknown. REMARKS. The shape of the head, the sculpture and vestiture of the face and the point where theradius joins the pterostigma show the relationship of this species to klugii. The shape of thecarapace and the strongly depressed body distinguish exigua from all other Palaearctic species ofthe genus. 362 T. HUDDLESTON Ascogaster gonocephala Wesmael(Fig. 26) Ascogaster gonocephalus Wesmael, 1835: 240. Holotype cf , BELGIUM: Brussels, Wesmael colln (IRSNB)[examined]. Wesmael stated that he had a single male of this species. Two males in Wesmael's collection bearlabels in Wesmael's handwriting 'Ascogaster gonocephalus mihi cf '; both are conspecific andagree precisely with Wesmael's description except that on the specimen which has completeantennae the segments number 31, not 30 as stated as Wesmael. This latter specimen, however,bears a small label '12 Jn.' and as Wesmael's original specimen was collected in June, I acceptthis specimen as the holotype of gonocephala. $ . Antenna long, 30 segmented. Flagellar segments at base at least twice as long as broad; apical segmentsvery small, much shorter and narrower than preceding ones. Head broad, strongly contracted behind eyes.Temple straight, about half length of eye in dorsal view; produced backwards on each side into a strongflange; vertex behind ocelli depressed, occipital carina obsolete. Ocelli large, OO = 2-0-2-5 OD. Eyeslarge, protuberant. Face broad, 2-5-3-0 times as broad as high, not strongly protuberant, rugose, thetransverse striae predominant; short medial vertical carina produced into a strong dentate tuberclebetween bases of antennae; hair on face pointing upwards. Clypeus not strongly protuberant, distinctlydivided from face; narrow, about half as broad as face; rugose-punctate; apical border reflexed forwards,rounded, without tubercles. Mandibles strongly twisted. Notaulices deep, foveolate, coalescing posterior-ly in a broad reticulate-rugose area. Precoxal suture broad, reticulate-foveolate. Propodeum rugose,dorsally reticulate with strong medial and lateral tubercles. Carapace rather long, slightly taperedposteriorly, not distinctly rounded posteroventrally. Ovipositor thin, slightly upcurved, evenly tapered.Ovipositor sheaths slender. Colour black; scape ventrally, face (except sometimes medially), genae ventrally, tegula and legsreddish-yellow. Coxae, hind femur distally, hind tibia distally and tarsus black or at least infuscate; hindtibia at base whitish yellow. Cf . Same as $ except that antenna longer, narrower, the apical segments not conspicuously shorter ornarrower than preceding ones; hindleg more extensively infuscate. MATERIAL EXAMINED 3 Cf , 3 $. Belgium: 2 cf (including holotype), Brussels, colln Wesmael (IRSNB). France: 1 cf , Var,Montauroux, l.viii.1970 (Van der Vecht) (RNH). Great Britain: 1 $>, Hampshire, Totton, 14.viii.1952(Vardy) (BMNH). Yugoslavia: 1 $, Srbija, Drazevac, nr Belgrade, 27-28.vi.1981 (Day & Fitton)(BMNH). ? Hungary: 1 $, no data (HNHM). HOSTS. No reared material examined. REMARKS. Morphologically gonocephala is closely similar to klugii; it differs from all otherspecies of Ascogaster by the curious posterior prolongation of the genae. Ascogaster grahami sp. n. (Figs 30-32) $. Antenna short, 20-24 segmented. Flagellum only slightly dilated, moderately contracted to apex; allsegments distinctly longer than broad, apical two or three only slightly longer than broad, almost quadrate.Head slightly broader than mesonotum, contracted behind eyes, strongly contracted ventrally in face view.Temple shorter than eye in dorsal view. Ocelli large, OO = 2-5-3-0 OD; not on line. Eyes large,protuberant. Face about twice as broad as high, moderately protuberant, shining, more or less denselypunctate, sometimes with weak rugae laterally; hair pointing downwards except between bases ofantennae ..Clypeus not protuberant, more or less densely punctate; apical border weakly rounded withoutteeth or excision. Mandibles slender, strongly twisted. Pronotum projecting but little in front ofmesonotum; reticulate-punctate and rugose laterally. Notaulices foveolate, indistinct anteriorly, coalesc-ing posteriorly in a broad, finely reticulate-rugose area; rest of mesonotum densely punctate, shining.Precoxal suture foveolate-rugose; rest of mesopleuron punctate except anterodorsally rugose. Propodeumdivided by a medial transverse carina with lateral longitudinal and often medial longitudinal carinae; finely,regularly reticulate-rugose. Carapace oval, rounded posteriorly, at most slightly emarginate postero- THE PALAEARCTIC SPECIES OF ASCOGASTER 363 ventrally. Hypopygium long but generally not exserted. Ovipositor long, about half as long as carapace,evenly upcurved, slender; ovipositor sheaths slender. Hind coxa rugose laterally and ventrally. Colour black except antenna at base testaceous, palps pale yellow or ivory; tegula and legs yellow ortestaceous except coxae and tarsi infuscate, at least in part; femur of mid leg sometimes infuscate dorsally;femur of hindleg infuscate in dorsal and ventral longitudinal bands; tibia of midleg sometimes and ofhindleg always infuscate but with a medial pale band. Carapace generally black but occasionally suffusedwith yellow at base, rarely with distinct pale areas. Cf . Same as $ except antenna longer, 22-26 segmented; all flagellar segments conspicuously longer thanbroad; carapace more elongate, less rounded; legs often lighter in colour. MATERIAL EXAMINED 58 Cf , 112 $ . Holotype $ , Great Britain: England, Hampshire, New Forest, Beaulieu Road, 25.vi.1959(Clark) (BMNH). Paratypes. Czechoslovakia: 1 $, Praha, Kosire, 27.vii.1961 (Strejcek) (CC); 1 $, Kamenica, ll.vii.1961(Strejtek) (CC); 1 $, Banska Stiavnica, 24.vii.1958 (Capek). France: 1 $, Brout Vernet (du Buysson)(MNHN); 3 $, Vien, 1877 (Giraucf) (MNHN). Germany (West): 1 $, Geierlambach, Heidelbeere(Haeselbarth) (HC); 1 $, Ober-Bayern, Wessling, 12.viii.1975 (Haeselbarth) (HC); 2 $, Ruthe colln(BMNH). Great Britain: 4 $, same data as holotype (BMNH); 1 $, Hampshire, Awbridge, vii.1981(Vardy) (BMNH); 1 cf, Droxford (ex Sorhagenia lophyrella) 20.V.1977 (Langmaid) (RSM); 6 cf, 4 $,Kent, Faversham (ex Infurdtinea argentimaculelld) 9.vii.l977 (Bradford) (RSM); 18 cf , 26 $, London,Bedford Park, vii.1929 (Waterston) (BMNH); 23 cf , 3 $, same data except 'on leaves of Tilia' 7.vi.l925(BMNH); 3 cf, same data but no date (BMNH); 3 cf, 9 $, London, Earls Court, 26.vi.1937 (Nixon)(BMNH); 1 $, Cambridgeshire, Monks Wood NNR, 17-29.vii.1978 (Fitton & Noyes) (BMNH); 2 $,Northamptonshire, Spratton, vii.1975 (Gauld) (BMNH). Italy: 4 cf , 3 $ , Ancona, ex Ficus, 1978 (Cola &Freude) (HC). Netherlands: 45 $, Wijster (Dr.), opposite Biol. Station, various dates, vii-viii (vanAchterberg) (RNH). Sweden: 1 $, Skane, Hoor district, 22.vi.1938 (Perkins) (BMNH); 1 $, Skane,Loderup, 22.vii.1938 (Perkins) (BMNH). HOSTS. Infurdtinea argentimaculella (Stainton) (Lepidoptera: Tineidae) [larvae feeding in mosson wall]; Sorhagenia lophyrella (Douglas) (Lepidoptera: Momphidae). REMARKS. This species is closely similar to annularis with which it has been generally confused incollections. The two species are not easy to distinguish and the only differences between themare summarised in the key. Waterston (1926: 174) misidentified this species as annularis. Ascogaster klugii (Nees von Esenbeck)(Figs 27, 35) Sigalphus klugii Nees von Esenbeck, 1816: 263. Syntypes, GERMANY (lost). Ascogaster ruficeps Wesmael, 1835: 242. Holotype $, BELGIUM: Brussels, Wesmael collection (IRSNB) [examined]. [Synonymised by Hellen, 1953: 86.]Ascogaster neesii Reinhard, 1867: 368. Syntype cf , GERMANY (ZMB) [examined]. Syn. n. Nees von Esenbeck's description of klugii applies equally well to both ruficeps and gonocephalasave only that he makes no mention of the gross modifications of the head characteristic of thelatter species. Nees von Esenbeck would certainly have mentioned so striking a feature had itbeen present in the species he was describing, and I therefore accept Hellen's synonymy ofruficeps with klugii. Wesmael stated that he had a single female of ruficeps. One specimen inWesmael's collection bears his label 'Ascogaster ruficeps mihi 9' and agrees precisely with hisdescription of ruficeps. I accept it as the holotype, which, although in rather poor condition, isnonetheless clearly identifiable. Two specimens (cf $) in Reinhard's collection, named as neesii, are conspecific and agreewith Reinhard's description. The locality data of the 9 do not agree with those cited byReinhard; the specimen is therefore not syntypic. The specimens assignable to neesii are allsmaller than typical klugii, their eyes are relatively larger, temples shorter and more stronglycontracted and the thorax flatter and more depressed. I consider these differences to besize-related and neesii therefore to be within the limits of variation of klugii. 364 T. HUDDLESTON $. Antenna 27-29 segmented. Flagellum sometimes dilated, the medial segments broader than long,sometimes not dilated, the medial segments then longer than broad; strongly tapered, apical six or sosegments much shorter and narrower than rest of flagellum, generally almost moniliform. Head broad,distinctly broader than mesonotum, more or less roundly contracted behind eyes. Temple slightly shorterthan eye in dorsal view. Occiput strongly concave. Ocelli small, OO = 3-0-3-5 OD; not on line. Frons notstrongly excavate. Face broad, at least twice as broad as high, slightly protuberant, medially distinctlyrugose, the transverse element predominating; hair on face fine, inconspicuous, pointing upwards onupper half of face. Clypeus only weakly convex, narrow, about half as broad as face, densely rugose-punctate with a row of foveae just behind thickened apical border. Mandibles small, strongly twisted.Pronotum projecting in front of mesonotum, transverse striate dorsally, laterally largely reticulate-punctate with fine rugosity ventrally. Notaulices anteriorly indistinct foveolate, merging with the reticu-late-punctate sculpture of mesonotum; posteriorly coalescing in a broad rugose, foveolate area; mesono-tum reticulate-punctate. Precoxal suture narrow, rugose foveolate, rest of mesopleuron punctate. Prop-odeum rather depressed; strongly, irregularly reticulate-rugose, generally only a medial transverse carinadistinct though sometimes weak medial longitudinal carinae visible, at least in part, only weak tuberclespresent laterally. Carapace elongate, oval, coarsely reticulate-rugose anteriorly, more finely reticulate-rugose posteriorly; rather flat in lateral view; posteroventral rim slightly in front of apex. Ovipositorslender; ovipositor sheaths long, slender. Hypopygium not long, produced medially to an apical point.Hind coxa smooth, punctate. Colour black; antenna pale at base, generally yellow or testaceous, occasionally basal flagellar segmentsivory; head varies from completely pale testaceous or yellow (except stemmaticum black) to completelyblack, except inner orbits pale-marked; pronotum in part and carapace anterolaterally yellow or tes-taceous. Legs completely pale yellow or testaceous except sometimes (in darker specimens) apex of hindtibia infuscate and occasionally, hind leg completely brown. Cf. Same as 9 except flagellum not dilated medially, not strongly tapered apically, the six apicalsegments not conspicuously shorter and narrower than preceding ones; carapace narrower. MATERIAL EXAMINED 93 O*, 124 9. Austria, Belgium, Bulgaria, Czechoslovakia, Finland, France, Germany, Great Britain,Sweden. HOSTS. Tubuliferola subochreella (Doubleday) (Lepidoptera: Oecophoridae). REMARKS. This species and gonocephala are closely similar in structure; the only differences arefound in the head and are cited in the key. Extensive pale coloration of the head is found only inklugii and gonocephala in Palaearctic Ascogaster, thus both species are fairly easy to recognise.The colour of the head is very variable, however, and it is not infrequently largely black. The bidentula-group Face punctate, hairs on upper part pointing downwards. Clypeus with one, two or three small, dentatetubercles or medially produced or emarginate. Four of the eight species of this group are common and widespread; albitarsus is uncommon,arisanica is known from China and Japan andperkinsi and longicornis only from Japan. This isthe most disparate group of species; varipes, perkinsi and longicornis could each be regarded asspecies sola. Indeed, varipes could also be regarded as transitional with the quadridentata-group. Ascogaster albitarsus Reinhard(Fig. 48) Ascogaster albitarsus Reinhard, 1867: 364. LECTOTYPE cf , POLAND: Gdansk (ZMB), here designated[examined]. Ascogaster leptopus Thomson, 1874: 584. LECTOTYPE $, SWEDEN: Thomson colln (ZI), here desig-nated [examined]. [Synonymised by Hellen, 1953: 86.] The single male specimen named albitarsus in Reinhard's collection agrees precisely withReinhard's description and is therefore accepted as a syntype and here designated lectotype. Four specimens (3 O* , 1 9) are in Thomson's collection above the name leptopus; they areconspecific and agree precisely with Thomson's description of that species. I here designate as THE PALAEARCTIC SPECIES OF ASCOG ASTER 365 lectotype the female which Graham restricted as type - his decision was not validated bypublication. I have no doubt that Hellen was correct to synonymise leptopus with albitarsus. $. Antenna long, 39 segmented. Flagellum tapered, medial segments about as long as broad, segments atapex much narrower, slightly longer than broad. Head not strongly contracted behind eyes. Templeslightly longer than eye in dorsal view. Occiput concave. Ocelli almost on line, OO = 2-5 OD. Fronsweakly excavate; strongly striate-rugose. Face about 1-5 times as broad as high; weakly convex;reticulate-punctate with a weak median tubercle dorsally. Clypeus weakly protuberant; apical borderproduced medially and slightly reflexed forwards, occasionally with a weak indication of two tubercles;smooth, shining, densely punctate but not reticulate. Mandible long, moderately twisted. Pronotum notprojecting strongly in front of mesonotum; finely, densely rugose laterally. Notaulices foveolate, coalesc-ing posteriorly in a large reticulate-rugose area; rest of mesonotum densely reticulate-punctate. Precoxalsuture broad, strongly reticulate-rugose anteriorly; narrow and rather weak posteriorly. Mesopleuronanterodorsally strongly rugose; rest of mesopleuron densely punctate. Propodeum short, broad, dividedby a medial transverse carina into dorsal and posterior areas, dorsal area coarsely reticulate-rugose;posterolaterally with stout, blunt teeth, medially with two rather weak teeth. Carapace long, clavate,reaching its maximum breadth in posterior third; coarsely reticulate-rugose; posteroventral rim somewhatbefore apex. Hypopygium and ovipositor short. Hind coxa densely punctate. Colour black; antenna testaceous in basal half; legs dark except apex of femur, tibia and tarsus of forelegyellow, base of tibia and of tarsus of mid-leg, base of tibia and basal segment of tarsus of hind leg paleyellow or ivory. Cf. Same as $ except antenna longer, all flagellar segments distinctly longer than broad; fore andmidlegs more or less light in colour, sometimes completely pale yellow. MATERIAL EXAMINED 13 cf , 3 $. Ireland: 1 cf , S. Tipperary, Burnt Woods, l.x.1947 (Daltry) (SC). Poland: lectotype cf ofalbitarsus. Sweden: 3 cf , 1 $ (lectotype of leptopus), Thomson colln (ZI); 5 cf, 1 $, Narke, Barsta,ll.ix.1931 (Roman) (NR); 1 cf, Uppland, 23.viii. (Roman) (NR); 1 $, Varmso (NR); 2 cf, dataindecipherable (NR). HOSTS. Unknown. REMARKS. I have examined 3 cf , 1 $ from Japan which differ from albitarsus in that the head ismore rounded behind the eyes, the mesopleuron is less strongly and less extensively rugose, thecarapace is smooth, punctate posteriorly and the light-coloured basal bands on hind tibia andtarsus are less distinct. These differences, too slight to warrant the description of a new species,are nevertheless enough to exclude these specimens from albitarsus, at least at present. Thelimits of variation given for albitarsus have, however, been based on the examination of so littlematerial that they will probably require revision when more specimens are available for study. Ascogaster arisanica Sonan(Figs 49, 52, 54) Ascogaster arisanicus Sonan, 1932: 79. Holotype cf , TAIWAN: Arisan, 2.V.1917 (Shiraki & Sonan) [notexamined] . Reinhard synonymised bidentula with rufipes but I have retained them as separate species.Watanabe (1937: 77) synonymised arisanica with rufipes, following Reinhard's interpretation ofthe latter species. I have based my interpretation of arisanica on specimens named as rufipes byWatanabe; they are not, however, conspecific with bidentula (= rufipes sensu Reinhard) or withrufipes as I have interpreted it. $. Antenna long, 36-39 segmented. Flagellum dilated medially, moderately contracted at apex; allsegments longer than broad but medial ones only slightly so. Head distinctly broader than mesonotum;often slightly expanded behind eyes; strongly rounded. Temple at least 1-5 times length of eye in dorsalview. Occiput strongly concave. Ocelli moderately large, OO = 3-0 OD; almost on line. Face about 1-5times as broad as high, evenly convex, densely punctate, sometimes reticulate-punctate, sometimes withweak rugosity medially. Clypeus evenly convex, about twice as broad as high, not distinctly divided fromface; densely punctate, the punctures larger than those on face; apical border produced, with a distinctmedial excision flanked by two tubercles. Mandible stout, weakly twisted, strongly striate-punctate at 366 T. HUDDLESTON base. Pronotum projecting strongly in front of mesonotum, with a strongly impressed Y-shaped groovedorsally; rugose laterally. Notaulices broad, shallow, rugose, coalescing posteriorly in a reticulate-rugosearea. Mesonotum narrow; densely punctate, generally reticulate-punctate medially. Precoxal suturebroad, shallow, reticulate-foveolate; rest of mesopleuron smooth, punctate except dorsally rugose.Propodeum rather long dorsally, divided medially by a weak, transverse carina which is raised into strongmedial and lateral flanges; dorsally coarsely reticulate-rugose. Carapace elongate, distinctly longer thanthorax and propodeum; generally broadest in distal half; posteroventral rim conspicuously in front of apex.Hypopygium and ovipositor short. Hind coxa smooth, punctate. Colour black; antenna at base testaceous, palps pale yellow; legs testaceous except mid tibia and tarsusgenerally infuscate apically, hind coxa infuscate at base, hind femur, tibia and tarsus infuscate at apex, hindtibia generally ivory in basal half, basal segment of hind tarsus ivory or pale yellow except sometimes atapex of segment. Hind coxa and femur sometimes almost completely black; in these darker specimens thelegs are brown rather than testaceous and the mid coxa, femur and tibia are extensively black-marked. Cf. Same as $ except antenna shorter, 33-37 segmented; flagellum not dilated medially; head notexpanded behind eyes; carapace narrower, less rounded, flatter. MATERIAL EXAMINED29 cf, 40$. Japan. HOSTS. No reared material examined. REMARKS. This species is structurally very close to bidentula, but it can be distinguished by the more massive head and the more elongate carapace. Ascogaster bidentula Wesmael(Figs 50, 51, 53) Ascogaster bidentulus Wesmael, 1835: 230. Holotype $, BELGIUM: 'Charleroy', Wesmael colln (IRSNB)[examined]. Chelonus multiarticulatus Ratzeburg, 1852: 25. Syntypes, GERMANY (lost). [Synonymised by Reinhard,1867: 365.] Ascogaster gibbiscuta Thomson, 1874: 586. LECTOTYPE $, SWEDEN: Skane, 'F (?) 14.viii., Thomsoncolln (ZI), here designated [examined]. Syn. n. Ascogaster fuscipennis Thomson, 1892: 1718. LECTOTYPE $>, SWEDEN: 'Hbg.' [= Helsingborg] Thom-son colln (ZI), here designated [examined]. Syn. n. Ascogaster atamiensis Ashmead, 1906: 191. Holotype $. JAPAN: 'Atami' (Koebele) (USNM) [examined].Syn. n. Wesmael (1835: 231) stated that he had only one female specimen when describing bidentulus.At present, two specimens in Wesmael's collection bear Wesmael's label ' 'Ascogaster bidentulus'mihi $ '. One of these has obviously been misplaced , as it is a female of rufidens ; the other agreesperfectly with Wesmael's description of bidentulus and is here accepted as the holotype. Reinhard (1867: 365) regards bidentula and rufipes as conspecific, but this is rejected here (seerufipes). Reinhard, however, had correctly identified bidentula and placed multiarticulatus insynonymy with it after having examined Ratzeburg's original material. This synonymy isaccepted here. Thomson (1874: 586) did not state what material he possessed when describing gibbiscuta. Inhis collection there are six specimens so named (2 cT, 4 9). The four females are conspecific andI designate as lectotype that which bears Thomson's labels 'gibbiscuta' and 'Skan' and a squarered label 'type' added by Graham. I can find no structural difference between gibbiscuta andbidentula; although gibbiscuta is more extensively pale-marked than typical bidentula I considerthis to be infraspecific. The specimen in Thomson's collection which agrees with his description of fuscipennis andwhich was labelled as lectotype by Graham is here designated lectotype. The only significantstructural difference between fuscipennis and bidentula is the form of the clypeus. In fuscipennisthe apical border is produced medially but without the paired tubercles or the excision ofbidentula. However, I have seen specimens in which the structure of the clypeus is intermediate-the tubercles are present but not divided by an excision. I therefore consider this characteristicto be variable in bidentula and so regard it and fuscipennis as conspecific. THE PALAEARCTIC SPECIES OF ASCOGASTER 367 $. Antenna long, 33-36 segmented. Flagellum slightly dilated medially, contracted in apical third, medialsegments about as long as broad, apical ones distinctly longer than broad. Head contracted behind eyes;more or less strongly rounded. Temple slightly longer than eye in dorsal view. Frons strongly rugose exceptsometimes medially smooth, punctate (particularly in small specimens). Occiput concave. Ocelli mod-erately large, OO = 3-0 OD, ocellar triangle obtuse, ocelli almost on line. Face about 1-5 times as broad ashigh, moderately convex with a small medial tubercle on upper half from which runs a weak carina whichextends back to the lower part of frons; reticulate-punctate, pubescence very long, silvery. Tentorial pitsminute with small but distinct tubercles dorsally. Clypeus weakly divided from face; not stronglyprotuberant, smooth, densely punctate; ventral border somewhat retracted but produced medially intotwo small but distinct dentate tubercles. Mandible moderately twisted. Pronotum projects in front ofmesonotum; weakly rugose. Notaulices distinct foveolate, rest of mesonotum densely punctate exceptposteriorly reticulate-rugose where notaulices coalesce. Precoxal suture coarsely foveolate-rugose, mostof mesopleuron dorsally strongly rugose (so that precoxal suture is often rather indistinct), ventrallysmooth, punctate. Propodeum strongly rugose, divided by a strong medial transverse carina which is raisedmedially and laterally into strong dentate flanges. Carapace short, oval, broadest at mid point; coarsely,irregularly reticulate-rugose; postero ventral rim distinctly in front of apex. Hypopygium short. Ovipositorshort, straight, thick at base, abruptly narrowed shortly before apex. Ovipositor sheaths short, stronglyclavate at apex. Hind coxa smooth, punctate, sometimes reticulate-punctate dorsally but never stronglytransversely rugose. Colour black; proximal half of antenna testaceous, rest brown or brownish-testaceous; legs yellowexcept all coxae dark, at least in part, fore and mid femur slightly infuscate, hind femur dark brown exceptat base, tibiae at least slightly infuscate at apex, tarsi infuscate except hind basitarsus pale yellow or ivory atbase. MATERIAL EXAMINED 162 cf, 92 9. Belgium, Bulgaria, Finland, France, Germany, Great Britain, Ireland, Japan, Nether-lands, Sweden. HOSTS. Epiblema roborana (Denis & Schiffermiiller), Epinotia cmciana (Linnaeus), Pandemissp. (Lepidoptera: Tortricidae), Eupithecia venosata (Fabricius) (Lepidoptera: Geometridae). REMARKS. I have examined 10 specimens (4 cf , 6 9) in which the female antenna is longer (39-41segmented), the facial sculpture slightly coarser and the legs more extensively yellow-markedthan in typical specimens of bidentula. They are also slightly larger and the differences may thusbe related to size. Ascogaster consobrina Curtis (Figs 45-47)Ascogaster consobrina Curtis, 1837: folio 672. Holotype d", GREAT BRITAIN: England (NMV) [examined]. Curtis did not state how many specimens he had before him when describing consobrina, but 1Cf , 2 $ are in his collection above this name. The two females are not accepted as syntypesbecause they differ from Curtis' brief but precise description in several features. The male agreesexactly with Curtis' description and I accept this specimen as the holotype of consobrina as thereis no indication that Curtis had more than one specimen of the species. The two females areconspecific and belong to rufidens; one was labelled 'type' by Nixon in 1948. $ . Antenna long, 33-34 segmented. Flagellum slightly dilated medially, not strongly narrowed at apex, allsegments longer than broad, generally distinctly so but sometimes medial segments only slightly longerthan broad. Head not strongly contracted behind eyes. Temple distinctly longer than eye in dorsal view.Occiput deeply concave. Ocelli moderately large, 00 = 3-0 OD; almost on line. Eyes moderatelyprotuberant. Malar space slightly less than twice basal breadth of mandible. Face about 1-5 times as broadas high, strongly projecting dorsally, rather straight in profile, reticulate-punctate, often with a few verticalrugae medially and with a strong medial carina running from upper part efface between antennal sockets.Clypeus raised medially; apical border not retracted, produced into a broad, blunt point with noimpression or tubercle; slightly less densely punctate than face. Mandible twisted distally. Pronotum notprojecting in front of mesonotum; finely, densely rugose laterally. Notaulices distinct, foveolate anteriorly,coalescing from about mid-point of mesonotum in a broad, reticulate-rugose area; rest of mesonotumpunctate. Precoxal suture distinct, foveolate. Mesopleuron laterally finely rugose-punctate except general- 368 T. HUDDLESTON ly for a polished, largely impunctate area just dorsal to precoxal suture, ventrally always sparsely punctate,shining. Propodeum finely, irregularly rugose, transverse carina strongly raised into medial and lateraltubercles. Carapace rather elongate, clavate, widened in posterior third; anterolateral flange dentate,projecting distinctly beneath the rim of carapace; posteroventral rim only slightly in front of apex; finelyreticulate-rugose. Hypopygium short. Ovipositor sheaths clavate. Hind coxa punctate, often weaklyrugose dorsally. Colour black; antenna at base and mandible brown; legs yellow except tarsi infuscate, hind coxa black atbase, generally hind femur and tibia infuscate at apex. Cf . Same as $ except antenna 31-33 segmented, flagellum not dilated medially. MATERIAL EXAMINED 63 cf , 28 $. Belgium, Czechoslovakia, France, Germany, Great Britain, Ireland, Japan, Netherlands,Sweden. HOSTS. No reared material examined. Ascogasterlongicornissp. n. (Figs 41, 42) $ . Antenna long, 47-50 segmented. Flagellum weakly dilated medially. Strongly narrowed at apex, medialsegments about as long as broad, rest distinctly longer than broad. Head broad, rounded behind eyes, notstrongly contracted. Temple slightly longer than eye in dorsal view. Occiput deeply concave. Ocelli in line,moderately large, OO = 3-5 OD. Eyes large, not protuberant. Malar space short, about one-third height ofeye. Face about 1 -3 times as broad as high, evenly convex, reticulate-punctate. A weak carina extends froma tubercle on the upper part efface to the frons; expanded into a blunt tubercle between antennae. Clypeusalmost as broad as face; weakly convex, reticulate-punctate; apical border rounded except in medialquarter a weakly emarginate projection. Mandible large, slightly twisted. Pronotum projecting in front ofmesonotum; dorsolaterally reticulate-foveolate. Notaulices foveolate; rest of mesonotum densely punc-tate except posteromedially reticulate-rugose. Precoxal suture not distinct because upper part of meso-pleuron completely coarsely rugose, generally reticulate, ventral part smooth, punctate. Propodeumcompletely coarsely and evenly reticulate-rugose with prominent medial and lateral tubercles, the medialpair pyriform, the lateral ones blunt, dentate. Carapace oval in dorsal view, clavate in lateral view,posteroventral rim distinctly in front of apex; finely reticulate-rugose. Ovipositor short, straight, abruptlynarrowed shortly before apex; ovipositor sheaths clavate. Hind coxa smooth, punctate; hind tibiaincrassate; hind tarsus slightly laterally compressed. Colour black: palps ivory, mandibles and antenna brown except antenna at base testaceous; carapace atbase yellow; foreleg yellow, tarsus infuscate apically; midleg pale except femur sometimes, tibia and tarsusat apex infuscate; hindleg dark except coxa (infuscate only dorsally) and trochanter yellow. Mid and hindtibia with ivory band near base. Cf . Same as $ except that antenna longer but sometimes with slightly fewer segments; not dilatedmedially; all flagellar segments distinctly longer than broad; legs generally lighter in colour. In two of themales examined apical border of clypeus rufous. MATERIAL EXAMINED 6 cf , 7 $. Holotype $, Japan: Mt Tachibana, Fukuoka city, 22.ix.1979 (Maeto) (ELKU). Paratypes. Japan: 1 cf , same data as holotype; 1 $, same data except 14.vii.1979 (MC); 5 cf, 3 $,Hiroshima Pref., Shobara, 23.viii.1976 (Maeto] (2 cf , 2 $ , BMNH; rest MC); 2 $ , Fukuoka city, MinamiPark, 19.V.1977 (Maeto) (BMNH; MC). HOSTS. Unknown. REMARKS. This species is perhaps most closely related to bidentula. A. longicornis is easilydistinguished from all other Palaearctic species by its conspicuously long antennae and by thecharacteristics of its clypeus. It is a large (up to 7-0 mm) stout species though the males tend to berather smaller on average. Ascogasterperkinsisp. n. (Figs 43, 44) 9- Antenna long, 36-38 segmented. Flagellum weakly dilated medially, contracted apically, medialsegments about as broad as long, rest distinctly longer than broad. Head contracted behind eyes. Temple THE PALAEARCTIC SPECIES OF ASCOGASTER 369 slightly longer than eye in dorsal view. Occiput deeply concave. Ocelli moderately large, OO = 3-0-3-5OD, ocellar triangle obtuse but ocelli not on line. Frons with two depressed, polished impunctate areasbehind antennae. Eyes moderately protuberant. Face about 1-5 times as wide as high, protuberant, evenlyconvex, reticulate-punctate, shining, with a medial tubercle from which a weak carina runs to anteriorocellus. Clypeus weakly divided from face, narrow, about half as broad as face; protuberant; shining,densely punctate, sometimes reticulate-punctate; apical border strongly impressed laterally, medial arearaised and produced forwards into a distinct tooth. Mandible small, moderately twisted. Pronotumprojecting in front of mesonotum, laterally reticulate-punctate dorsally with a deep pit medially.Notaulices weak, foveolate; rest of mesonotum reticulate-punctate. Precoxal suture foveolate; meso-pleuron above precoxal suture reticulate-foveolate anteriorly, smooth, punctate posteriorly; mesopleuronventrally densely punctate. Propodeum completely finely reticulate-rugose, divided medially by a trans-verse carina with medial and lateral tubercles, the medial pair of tubercles broad, stout and not stronglyraised particularly in small specimens, the lateral pair more prominent, pyriform. Carapace short, oval indorsal view, clavate in lateral view; posteroventral rim distinctly in front of apex. Hind coxa smooth,punctate. Hind tibia incrassate. Hind tarsus slightly laterally compressed, basal segment with a conspi-cuous flange ventrally which has a row of short bristles along its edge and a row of longer bristles at its baseon each side. Colour black; antenna at base, mandible and tegula testaceous; legs yellow except mid tibia dark at apex,hind femur and tibia dark at apex, mid and hind tarsi infuscate; carapace yellow at base. Cf Same as $ except antenna shorter, 32-34 segmented; flagellum not dilated medially; carapace lessextensively pale-marked. MATERIAL EXAMINED 27 cf , 9 $. Holotype $, Japan: Mt Tachibana, Fukuoka City, 23.vi.1979 (Maetd) (ELKU). Paratypes. Japan: 11 cf , 2 $ same data as holotype (BMNH, MC); 3 cf , same data except 18.vi.1978(MC); 2 cf , 4 $ , Nagano Pref. , Shimashima-dani, 1000-1300 m, 28.vii.1980 (Takemoto) (BMNH, MC); 5Cf , Oiya Pref. , Kuju, Bogazuru (Maetd) (MC); 2 cf , 1 $ , Kamikochi, 22-30.vii. 1954 (Townes) (TC); 4 cf ,Mt Norikura, 2000 m, 30.vii.1954 (Townes) (TC). HOSTS. Unknown. REMARKS. This species is morphologically very similar to dentifer (q.v.). A. perkinsi is not closelyrelated to any species in the bidentula-group and in the structure of its clypeus it shows affinitieswith the quadridentata-group. I dedicate this species to the memory of the late Dr J. F. Perkins, agifted and generous former colleague. Ascogaster rufidens Wesmael(Figs 37, 39, 40) Ascogaster rufidens Wesmael, 1835: 231. Holotype $, BELGIUM: Charleroy, colln Wesmael (IRSNB) [examined].Chelonus rufipes Herrich-Schaffer, 1838: 154. Syn types, GERMANY (lost). [Junior secondary homonym of rufipes (Latreille, 1809); synonymised by Reinhard, 1867: 366.]Chelonus (Ascogaster) laevigator Ratzeburg, 1852: 25. Syntypes, GERMANY (lost). [Synonymised by Reinhard, 1867: 365.] Wesmael (1835: 233) stated that he had a single specimen of rufidens. One specimen inWesmael's collection bears his label 'Ascogaster rufidens mihi cf ' and agrees precisely with thedescription of that species except that the antennae are now more mutilated than originallystated. It is here accepted as holotype. Wesmael believed this specimen to be a male and solabelled it, but I consider it to be a female though the retraction of the genitalia beneath thecarapace makes it difficult to discern the sex, particularly so as rufidens exhibits no sexualdimorphism. Reinhard stated that he had examined Ratzeburg's original specimen of laevigator and hecited characters for rufidens which show that he had correctly identified that species. The abovesynonymy is therefore accepted. Reinhard (1867: 366) placed rufipes Herrich-Schaffer in synonymy with rufidens althoughapparently without having seen the original material. Herrich-Schaffer's description and figureare entirely consistent with Reinhard's placement. 370 T. HUDDLESTON 9- Antenna long, 33-35 segmented. Flagellum not distinctly dilated medially, all segments at least as longas broad, generally longer than broad. Head contracted behind eyes but not strongly so. Temple aboutequal to eye in dorsal view. Occiput strongly concave. Ocelli small OO = 3 OD, ocellar triangle obtuse,ocelli almost on line. Eyes protuberant but not large. Malar space about half height of eye. Face notstrongly protuberant, densely punctate, sometimes reticulate-punctate, densely hairy; a weak tuberclemedially below antennal bases, generally with a weak carina extending back between the antennae,sometimes almost reaching anterior ocellus. Clypeus not protuberant, sculpture same as face; apicalborder convex, with medially three small dentate tubercles. Mandibles stout, only slightly twisted, ventralborder reflexed, produced into a stout flange extending from mandibular articulation to base of ventraltooth. Pronotum transversely rugose dorsally with a distinct medial pit and a small raised triangular area offine punctures immediately in front of mesonotum; lateral surface of pronotum punctate dorsally, rugoseventrally with a medial foveolate groove. Propleuron reticulate-punctate. Notaulices deep, foveolate.Mesoscutum broad, reticulate-punctate. Precoxal suture broad, anteriorly rugose-foveolate, posteriorlynarrow foveolate; mesopleuron rugose dorsally, rest smooth punctate. Propodeum completely stronglyreticulate-rugose, divided into dorsal and posterior surfaces by a medial transverse carina which isproduced posterolaterally into two stout blunt crests. Carapace broad, rugose, sometimes with one or twofaintly impressed lines in positions suggesting the suturiform articulation (such a condition was noted byWesmael in his description of rufidens and is indeed visible in the holotype). Hypopygium short.Ovipositor sheaths short, widened apically where there is a brush of long hairs. Hind coxa smooth,punctate. Colour black; antenna at base, mandibles, palps yellow or testaceous. Legs yellow except coxae black, atleast in part, all tarsi infuscate, hind femur and tibia infuscate. Cf. Same as 9- MATERIAL EXAMINED 81 cf , 4 1 9 Belgium, Czechoslovakia, France, Germany, Great Britain, Hungary, Ireland, Netherlands,Sweden. HOSTS. Croesia bergmanniana (L.), Pandemis cerasana (Hiibner), Pandemis corylana (Fabri-cius), Pandemis heparana (Denis & Schiffermuller) (Lepidoptera: Tortricidae). Yponomeutapadella (L.) (Lepidoptera: Yponomeutidae). REMARKS. The unique clypeal armature of rufidens has ensured that it is one of the leastmisinterpreted species oi Ascogaster. The characteristics of the mandibles, which appear to havebeen overlooked by previous workers, are also invaluable taxonomic discriminants. Evenhuis &Vlug (1983) have pointed out that the principal hosts of rufidens belong to the Tortricidae,Tortricinae which lay their eggs in batches. In contrast to this quadridentata attacks Tortricidae,Olethreutinae the females of which lay their eggs separately on the host plants. Ascogaster varipes Wesmael(Fig. 38) Ascogaster varipes Wesmael, 1835: 234. LECTOTYPE 9, BELGIUM: Brussels, colln Wesmael (IRSNB),here designated [examined]. Ascogaster cavifrons Thomson, 1874: 585. LECTOTYPE 9, SWEDEN: Skane, Torekov, vii.1860 (Thom-son) (ZI), here designated [examined]. Syn. n. Ascogaster sternalis Thomson, 1874: 587. LECTOTYPE 9, SWEDEN: Smaland (Thomson) (ZI), heredesignated [examined]. [Synonymised by Telenga, 1941: 322.] Ascogaster jaroslawensis Kokujev, 1895: 86. Holotype 9. U.S.S.R.: 'Jaroslaw' (AS) [examined]. Syn. n. Wesmael (1835: 235) stated that he possessed two males and six females of varipes. Two malesand four females are present above this name in Wesmael's collection, each bearing Wesmael'slabel 'Ascogaster varipes mihi'. They are conspecific and agree precisely with Wesmael'sdescription of varipes. Wesmael assigned two of his series of six females to a 'var. 2'; thesespecimens are among the surviving syntypes and there is also a male labelled 'var. 1'. I designateas lectotype of varipes the better-preserved of the two females from the main syntypic series. Thomson did not state how many specimens he had before him when describing cavifrons; atpresent six are in his collection above this name. One of these is a male of abdominator andcannot be a syntype of cavifrons because it does not agree with the original description; THE PALAEARCTIC SPECIES OF ASCOGASTER 371 Thomson's remarks upon abdominator show him to have been thoroughly familiar with thatspecies. Of the five remaining specimens (2 cf , 3 $) four are certainly conspecific and agree withThomson's description of cavifrons, the fifth (cf) is in poor condition and is impossible toidentify. I designate as lectotype the female bearing Thomson's labels Tkov., 7 '60' '$'cavifrons m.' and a square red label Type' added by Graham. A. cavifrons comes well within therange of variation of varipes, agreeing closely with Wesmael's 'var. 2'. Of the four specimens in Thomson's collection above the name sternalis, I designate aslectotype the best preserved specimen that was labelled as 'type' by Graham. The strongly raisedacetabular carina used by Thomson to differentiate sternalis seems to me to be only ofinfraspecific value; apart from this characteristic, sternalis agrees with varipes in every particularand Telenga's synonymy is therefore accepted. The holotype 9 ofjaroslawensis has lost its headbut I have no doubt that it is conspecific with varipes. $. Antenna long, 34-35 segmented. Flagellum dilated medially, strongly tapered to apex, most distalsegments about as long as broad. Head rounded behind eyes, often expanded, particularly in largerspecimens. Temple about 1-0-1-5 times length of eye in dorsal view. Occiput strongly concave. Ocelli inobtuse triangle but not on line, OO = 3-0-3-5 OD. Frons behind antennae excavate, the depressed areasmooth with a weak medial carina which runs from between the antennae to just in front of the fore ocellus.Eyes not strongly protuberant. Genae generally convex in face view. Face protuberant, always distinctlyand evenly convex, about twice as broad as high, generally reticulate-punctate but sometimes finely andregularly reticulate-rugose. Clypeus transversely convex dorsally, the ventral half transversely depressed,its border slightly reflexed without tooth or impression medially; reticulate-punctate. Mandible large,strongly twisted with a deep semicircular depression at base. Pronotum projecting slightly in front ofmesonotum; rugose-punctate. Notaulices distinct, foveolate, rest of mesonotum punctate except post-eriorly reticulate-rugose and ventrally punctate; precoxal suture not easily distinguished from rugosity onmesopleuron. Propodeum completely reticulate-rugose, divided by a medial carina which is expanded intoprominent dentate flanges laterally and two broad low protuberances medially between which dorsalsurface of propodeum weakly depressed, sometimes with a short, medial longitudinal carina. Carapaceoval, generally rather deep distally, posteroventral rim distinctly in front of apex; reticulate-rugose at base,rugose-punctate apically. Hypopygium short. Ovipositor short, straight, abruptly narrowed shortly beforeapex; ovipositor sheaths short, clavate. Hind coxa strongly transversely rugose. Colour black; carapace sometimes pale yellow at base, legs generally predominantly black, only femoraat apex and tibiae yellow; hind tibia infuscate at apex, occasionally legs predominantly yellow with onlyapex of hind femur and apex of mid and hind tibiae and all tarsi dark. Cf . Same as $ except antenna not dilated medially; carapace only rarely pale-marked and then weaklyand not in a distinct basal patch; legs often darker. MATERIAL EXAMINED 165 cf , 127 $. Austria, Belgium, Bulgaria, Czechoslovakia, France, Germany, Great Britain, Hungary,Netherlands, Sweden, U.S.S.R., Yugoslavia. HOSTS. No reared material examined. REMARKS. This species is readily recognisable by the semicircular depression at the base of themandible and by the characteristics of its face and clypeus. The mandibular depression is bestseen if the head is examined in anterolateral view under oblique illumination; it is often invisiblein full anterior view. A. varipes is a conspicuously short, stout species with a short, deep, stronglyrounded carapace; reticulata and quadridentata share this characteristic, however, and may thusbe confused with varipes, particularly those specimens of the latter species which have ratherrugose faces. Neither reticulata nor quadridentata, however, has a semicircular depression at thebase of the mandible, quadridentata always has a medial tubercle on the clypeus, and reticulatahas a smooth, punctate hind coxa. The quadridentete-group Face strongly, irregularly rugose, hair on upper part pointing downwards. Clypeus generally with medialdentate tubercle, occasionally with none but never with more than one. The eight species of this group are all heavily sculptured, having the head and thorax more or lesscompletely rugose. The dentate tubercle on the clypeus is absent in armata and reticulata and is 372 T. HUDDLESTON often weak in scabricula; these species are readily distinguished from all other Palaearcticspecies by the characteristics of the head and carapace, as cited in the key. The species of thisgroup are generally widely distributed throughout the Palaearctic region - indeed quadridentatais known to be Holarctic - but armata, canifrons and scabricula are uncommon and reticulata isas yet known only from Japan and Czechoslovakia. Ascogaster armata Wesmael(Figs 56, 57) Ascogaster armatus Wesmael, 1835: 233. LECTOTYPE $, BELGIUM: Brussels, colln Wesmael (IRSNB), here designated [examined] . Chelonus pulchellus Curtis, 1829: 105 [Nomen nudum. ]Ascogaster esenbeckii Curtis, 1837: folio 672. LECTOTYPE $, GREAT BRITAIN: England, Glanvilles Wootton, vii, grass in meadows (NMV), here designated [examined]. [Replacement name for pulchellus Curtis, 1829.] Syn. n.Chelonus luteicornis Herrich-Schaffer, 1838: 154. Syntypes, GERMANY (lost). [Synonymised by Reinhard 1867: 263.] Wesmael (1835: 234) stated that he possessed three females and one male of armata. There arethree females and two males in Wesmael's collection each bearing his label 'Ascogaster armatusmini'; they are conspecific and agree precisely with the description of armata. I designate thebest preserved female as lectotype and the other two females as paralectotypes. No doubt one ofthe two males is the male syntype referred to by Wesmael but it cannot be differentiated. I have examined two specimens from Curtis's collection, one labelled 'pulchellus' and theother Type of Ascogaster esenbeckii Curtis det G. E. J. Nixon 1948'. These specimens areconspecific and agree precisely with Curtis' description; that labelled 'Type' by Nixon bears nodata whereas the other specimen has a label with the data published by Curtis. I have thereforedesignated the latter specimen as lectotype of esenbeckii, and it comes well within the limits ofvariation of armata. Marshall (1885: 142) synonymised esenbeckii with instabilis, giving noreason for his action but probably basing it on Curtis' description. I have no doubt that thisplacement is incorrect. It is quite clear from the descriptions of luteicornis and annularis that Herrich-Schaffer'sfigures of these two species were transposed. Thus figure 154.7 depicts luteicornis and 154.8annularis. Reinhard (1867: 263) synonymised luteicornis and armata based on an examination ofspecimens sent to him as luteicornis by Herrich-Schaffer. It is clear from Reinhard's descriptionthat he had correctly identified armata and I therefore accept his synonymy. $. Antenna long, 36-39 segmented. Flagellum dilated medially, medial segments distinctly broader thanlong, apical five or six segments nearly moniliform; densely covered in short, thick adpressed bristles. Headsubcubic, not strongly contracted behind eyes. Temple 1-5-2-0 times length of eye in dorsal view. Occiputstrongly concave. Ocelli small, OO = 4 OD; ocellar triangle obtuse, ocelli almost on line. Frons behindantennae deeply excavate, generally smooth, shining and delimited laterally by a weak carina. A medialcarina extends from upper part of face between antennae to anterior ocellus; grossly expanded betweenantennae into an erect triangular lamina. Eyes small. Malar space about half height of eye. Faceprotuberant, about twice as broad as high, strongly rugose, generally striate-rugose but sometimesreticulate medially with a weak tubercle medially at which point arises the interantennal carina. Clypeusprotuberant though not strongly so, deeply divided from face; apical border convex, not protrudingmedially and without distinct tooth or tubercle; largely reticulate-punctate. Mandible stout, not twisted,densely punctate with long hairs. Pronotum projecting slightly in front of mesonotum; with a medialtransverse groove dorsally, reticulate-rugose laterally; propleuron reticulate-rugose. Notaulices presentbut indistinct because mesonotum strongly rugose, posteriorly reticulate; medial lobe of mesoscutum oftenwith a reticulate-punctate area anteriorly. Precoxal suture indistinct; mesopleuron completely reticulate-rugose except sometimes reticulate-punctate ventrally, but mesolcus rugose or reticulate. Propodeumstrongly reticulate-rugose, divided into dorsal and posterior surfaces by a weak transverse carina which isproduced posterolaterally into stout blunt teeth. Carapace long, rather narrow; finely reticulate-rugose,posteriorly rugulose-punctate. Ovipositor short, straight, abruptly tapered just before apex; ovipositorsheaths with apical brush of long hairs. Hind coxa transversely striate on dorsal and outer surface; hindtibia densely reticulate-punctate. Wings rather short, infumate. THE PALAEARCTIC SPECIES OF ASCOGASTER 373 Colour black, basal segments of antenna light in colour; fore tibia completely, mid and hind tibia at leastpartly and gaster proximally yellow. Cf . Same as $ except that flagellum not expanded medially, medial segments about as long as broad,gaster almost always completely black. MATERIAL EXAMINED 13 cf, 13 $. Belgium: 2 cf, 3 $ (including lectotype), Wesmael colln (IRSNB). France: 2 cf, St M.Vesubie, 27-29. vii. 1950 (Granger) (MNHN); 1 $, Chavith; 30.vii.1882 (de Gaulle) (MNHN). GreatBritain: 4 cf, Stephen's colln (BMNH) 2 $, Curtis colln (lectotype & paralectotype of Ascogasteresenbeckii) (NMV); 1 $ , Marshall coin (BMNH); 1 $ , Hampshire, Portodown nr Portchester, 16.vii.1971(Else) (BMNH); 1 cf, Sussex, Hailsham marshes, 27.vii.1949 (Ford) (BMNH); 1 $, Surrey, Coulsdon,Happy Valley, 13.viii.1978 (Noyes) (BMNH); 1 $, Kent, Chattenden, 12.vii.1949 (Ford) (BMNH); 1 cf ,Buckinghamshire, Buttlers Hangings nr West Wycombe, 24.vii.1978 (Shaw) (RSM); 1 ^Leicestershire(Matthews) (BMNH); 1 cf , Cardiganshire, Tresaith coll. vi.1979 [ex Coleophora paripennella] (Simpson)(RSM). Hungary: 1 $, Nagyvisnyo, 1-8. viii. 1956 (Mihalyi) (HNHM); 1 cf, Nemetbanyo Jager-volgy,17.vii.1973 (Papp) (HNHM). Italy: 1 cf , Campi, Riva s Garda, 550 m, 7.vii.l966 (Haeselbarth) (HC).Netherlands: 1 $ , Bemelerberg Exc. St Petersberg, 20.vii.1950 (RNH). HOSTS. Coleophora hornigiToll [= paripennella auctt.] (Lepidoptera: Coleophoridae). REMARKS. The massive head of this species together with its strongly excavate frons distinguish itfrom all other species of Ascogaster. The strong, dentate flange between the scapes is also adistinctive characteristic shared only by dentifer which is otherwise quite a different insect. A.armata has no medial tooth on the clypeus, a characteristic found otherwise only in reticulata andscabricula in the quadridentata-group. Ascogaster brevicornis Wesmael(Fig. 67) Ascogaster brevicornis Wesmael, 1835: 239. Syntypes, BELGIUM: Wesmael colln (IRSNB) [examined].Chelonus monilicornis Herrich-Schaffer, 1838: 154. Syntypes, GERMANY (lost). [Synonymised byReinhard, 1867: 367.] Wesmael (1835: 240) stated that he had three males and one female of this species. Two malesand two females in Wesmael's collection are labelled 'Ascogaster brevicornis mini' in Wesmael'shandwriting; they are conspecific and agree with the original description. Clearly, however,there is a discrepancy between the sex of his specimens as given by Wesmael and that of thespecimens which now stand above brevicornis in his collection. Thus there is doubt about thesyntype status of the females and, as the characteristics which differentiate brevicornis are bestexhibited by that sex, I have not designated a lectotype. Reinhard placed monilicornis and brevicornis together in synonymy with similis. It is clearfrom Reinhard's description that he had correctly identified brevicornis and Herrich-Schaffer'sdescription and figure of monilicornis support Reinhard's placement of these species. I do not,however, believe that similis and brevicornis are conspecific because the thorax and scutellum ofbrevicornis are always coarsely rugose whereas those of similis were described as finely andclosely punctulate. $. Antenna short, 22-23 segmented. Flagellum slightly dilated medially, all segments considerablybroader than long, preapical segments slightly broader than long, apical segment longer than broad,pointed, all segments in distal half of flagellum distinctly separated. Head roundly contracted behind eyes.Temple about equal in length to eye in dorsal view. Vertex strongly rugose, in part reticulate. Ocelli small,OO = 4 OD; ocellar triangle obtuse, the ocelli almost on line. Eyes fairly large but not protuberant. Facenot strongly protuberant, about twice as broad as high, strongly rugose, generally reticulate-rugosemedially with a small tubercle from which a weak carina extends back between antennae and ends atanterior ocellus. Clypeus not protuberant polished, punctate; apical border projecting slightly forwards,produced into a blunt tooth medially. Mandible not twisted, punctate at base. Pronotum dorsally with adeeply impressed foveolate transverse groove; laterally strongly rugose-reticulate. Propleuron reticulate-rugose. Notaulices deep, foveolate. Mesoscutum strongly rugose. Mesopleuron almost completely reticu-late-rugose, precoxal suture therefore indistinct. Propodeum completely reticulate-rugose, divided by a 374 T. HUDDLESTON weak, transverse carina produced into medial and postero-lateral pairs of stout dentate crests. Carapaceviewed dorsally widest at about mid-point, tapering evenly to apex; laterally angulate posteroventrally, theventral rim reaching almost to apex; completely strongly reticulate-rugose. Ovipositor short. Ovipositorsheaths wide, polished, with a brush of long hairs apically. Hind coxa rugose on outer surface. Colour black: antenna at base, fore tibia completely, mid and hind tibia at base yellow. Cf . Same as 9 except that antennal segments longer, less distinctly separated; ventral rim of carapace notreaching apex. MATERIAL EXAMINED 15 Cf , 29 9 Belgium, Czechoslovakia, France, Germany, Great Britain, Hungary, Ireland, Netherlands,Switzerland. HOSTS. No reared material examined. REMARKS. This species closely resembles quadridentata; the female of brevicornis is easilydistinguished from quadridentata by its short, thick apically moniliform antenna. Further, thenumber of antennal segments (cf, 9) is different in the two species; 21-23 segments inbrevicornis and more than 30 in quadridentata. The carapace also differs; in brevicornis theventral rim reaches almost to the apex, in quadridentata it is always distinctly before the apex.The shape of the carapace viewed dorsally is also generally rather different, that of brevicornisbeing much more distinctly tapered and less rounded than that of quadridentata. The shape ofthe carapace in the latter species, however, is rather variable. Ascogaster canifrons Wesmael(Figs 58, 59) Ascogaster canifrons Wesmael, 1835: 236. LECTOTYPE 9, BELGIUM: Liege (Robert) (IRSNB), heredesignated [examined]. Ascogaster graniger Thomson, 1892: 1791. LECTOTYPE 9, SWEDEN: Skane, Palsjo (ZI), here desig-nated [examined]. Syn. n. Wesmael (1835: 237) stated that he possessed one female and one male of canifrons', two suchspecimens are in his collection above the name canifrons, both labelled 'Ascogaster canifronsmini'. They are conspecific and agree precisely with Wesmael's description; I consider themsyntypes and designate the female as lectotype. There are five specimens (2 cf , 3 9) in Thomson's collection above the name graniger. Thethree females and one of the males are conspecific and agree with the original description; thesecond male mounted on the same pin below the first belongs to the bidentula-gioup. Thelectotype is from Palsjo (type-locality) and was selected and labelled as 'type' by Graham; it iswell within the limits of variation of canifrons. One of the female specimens bears the labels 'L-d'(Lappland) and 'lapponicus' in Thomson's handwriting. At first sight, therefore, this specimenappears to be the holotype of lapponica but it disagrees with Thomson's description of thatspecies in several respects. It has been pointed out by Huddleston (1980: 3) and Fitton (1982: 4)that species labels in Thomson's handwriting were not placed on the specimens by him and thatthe presence of such a label on a pin is not reliable evidence of the identity of the specimen. 9- Antenna long, 39-40 segmented. Flagellum not dilated medially, strongly contracted distally, proximaland distal segments distinctly longer than broad, medial segments broader than long. Head large, stronglycontracted behind eyes. Temple not rounded, slightly longer than eye in dorsal view. Occiput deeplyconcave. Ocelli small, OO = 3-0-3-5 OD, almost on line. Frons depressed behind antennae, stronglytransversely rugose. Eyes not strongly protuberant. Face about 1-5 times as broad as high, moderatelyprotuberant, completely rugose with a strong medial tubercle from which a weak carina extends backbetween the antennae to the anterior ocellus. Clypeus narrow, about half breadth of face; moderatelyprotuberant, shallowly reticulate-punctate, apical border weakly convex with a distinct sharp tuberclemedially. Mandible stout, not strongly twisted. Pronotum projecting slightly in front of mesonotum;rugose. Mesonotum completely densely rugose; notaulices not distinct. Mesopleuron completely coarselyreticulate-rugose; precoxal suture not distinct. Propodeum completely coarsely reticulate-rugose, notdistinctly carinate but for a medial transverse carina which is expanded medially into two small blunt teethand laterally into two larger ones. Carapace long, oval, densely and irregularly reticulate-rugose, THE PALAEARCTIC SPECIES OF ASCOGASTER 375 posteroventral rim slightly in front of apex. Hypopygium short. Ovipositor short, straight, abruptlynarrowed shortly before apex; ovipositor sheaths short, broad, clavate with a copious tuft of hairs distally.Hind coxa distinctly rugose. Colour black; mouthparts, antenna at base, foreleg (except coxa and tarsus infuscate) and sometimesanterior third of carapace yellow; midleg yellow except coxa, femur at base, tibia at apex and tarsusinfuscate; hind leg dark except trochanter, trochantellus, base and apex of femur, base of tibia andsometimes tarsus in part yellow. Cf . Same as $ except antenna longer, medial flagellar segments as long as broad; carapace completelyblack, its posteroventral rim distinctly in front of apex. MATERIAL EXAMINED 4 cf, 11 $. Belgium: lectotype cf of canifrons. Czechoslovakia: 1 $, Trencin, Kostolna, ll.vii.1976(Lukas) (LC); 1 cf, Bohemia, K. Studenec, 18.vii.1955 (Sedivy) (CC). Great Britain: 1 $, Cam-bridgeshire, Abbots Ripton, Monks Wood NNR, 17-28. viii. 1978 (Fitton & Noyes) (BMNH). Ireland: 2 $> ,Co. KD., Royal Canal, 3.viii.l949 & 15.vii.1950 (Stelfox) (SC); 1 cf , same data except 30.vi.1950; 1 $ , Co.KD, Landenstown, 20.viii.1941 (Stelfox) (SC); 1 $, Co. KD., Landenstown, 20.viii.1941 (Stelfox) (SC); 1$ Co. WX., Killurin, 22.vii.1937 (Stelfox) (SC). Mongolia: 1 $, Central aimak, 11 km ESE. from Somon,Bajanzogt 1600-1700 m, 26.viii.1968 (Kaszab) (HNHM). Sweden: lectotype $ of graniger; 1 cf, 2 $(paralectotypes ofgraniger), Thomson coll. HOSTS. Endothenia quadrimaculana (Haworth) (Lepidoptera: Tortricidae). REMARKS. Papp (1967; 1971) has previously recorded this species from Mongolia. The longantenna and contracted temples and genae distinguish canifrons from all other species in thequadridentata-group. The strong transverse rugae on the frons are a striking feature of thespecies. Ascogaster dispar Fahringer(Figs 70, 71, 73) Ascogaster dispar Fahringer, 1934: 524. LECTOTYPE $, AUSTRIA: '13.V.12' (NM), here designated [examined]. Ascogaster spinifer Tobias, 1964: 185. Holotype cf , U.S.S.R. : 9.vi. 1958 (Tobias) (AS) [examined]. Syn. n.Ascogaster koslo vi Tobias, 1972: 601. Holotype $>, MONGOLIA: 3-4. vii. 1968 (Koslov) (AS) [examined]. Syn. n. There are four specimens (1 cf , 3 9) above the name dispar in the NM collection bearing labelswith dates from 13. v. 1912 to 25. vi. 1912 and the number 7-144; one also has a label bearing thehost data cited by Fahringer and another a label 'Ascogaster dipsaris m.' in Fahringer'shandwriting. All these specimens are conspecific and agree precisely with Fahringer's descrip-tion of dispar and I have no doubt that they are syntypes of that species. Fahringer obviouslychanged the specific epithet before publication. The holotype of spinifer and that of koslovi are rather smaller than typical specimens of disparbut they are within the limits of variation of that species. $. Antenna 33-35 segmented. Flagellum slightly dilated medially, strongly tapered to apex; medialsegments about as long as broad, following segments slightly longer than broad. Head broad, often slightlybroader than mesonotum; slightly expanded then strongly roundly contracted behind eyes. Temple alwaysdistinctly longer than eye in dorsal view. Occiput strongly concave. Frons moderately depressed,reticulate-rugose. Ocelli almost on line, OO = 3-5-4-0 OD. Face about twice as broad as high, weaklyconvex, completely finely rugose, medially slightly raised and more closely rugose. Clypeus weaklyconvex, smooth, punctate; apical border retracted except medially produced and with a dentate tubercle.Mandibles moderately twisted, stout, produced ventrally at base so that generally they project beneath thehead capsule; strongly rugose-punctate with no flange beneath at the base. Pronotum projects distinctly infront of mesonotum; striate-rugose except laterally often a small central smooth, punctate area. Notaulicesindistinct. Most of mesonotum strongly reticulate-rugose, often with small densely punctate areas laterallyand anteriorly but sometimes transversely rugose anteriorly. Precoxal suture indistinct; mesopleuronlaterally strongly reticulate-rugose, ventrally with smooth, punctate areas. Propodeum completely strong-ly reticulate-rugose; divided by a medial transverse carina which is strongly raised into a medial pair and alateral pair of dentate flanges. Carapace elongate oval, generally strongly narrowed posteriorly in dorsal 376 T. HUDDLESTON view; slightly depressed and rounded in lateral view; posteroventral rim slightly in front of apex.Ovipositor short, straight. Ovipositor sheaths clavate. Hind coxa finely transversely striate. Colour black; apex of femur, and tibia of foreleg, apex of femur of midleg testaceous. Cf . Same as 9 except antenna longer, 33-36 segmented; flagellum not dilated medially, all segmentsdistinctly longer than broad; propodeum rounded, the flanges generally much weaker; posteroventral rimof carapace distinctly more in front of apex. MATERIAL EXAMINED 25 cf, 21 9- Austria, Bulgaria, Czechoslovakia, France, Great Britain, Greece, Hungary, Iran,Mongolia, Switzerland, Turkey, U.S.S.R., Yugoslavia. HOSTS. Endothenia gentianaeana (Hiibner) (Lepidoptera: Tortricidae). REMARKS. The shape of the head is rather variable in this species; whereas in most specimens it israther massive, expanded behind the eyes and strongly rounded, in smaller specimens it is notexpanded behind the eyes, less strongly rounded, even weakly contracted. Thus some of thesmaller specimens of dispar rather resemble quadridentata; in dispar, however, the mandible isalways ventrally expanded at the base and never has a flange beneath, and in quadridentata themandible is not expanded at the base and always has a small flange ventrally. Further, in disparthe clypeus is not distinctly differentiated from the face except by a change in sculpture; inquadridentata the clypeus is distinguishable from the face by a fold. The medial apical tooth ofthe clypeus is always long and pointed in dispar, less so in quadridentata. Ascogaster quadridentata Wesmael(Figs 60, 62, 63, 64) Ascogaster quadridentata Wesmael, 1835: 237. Lectotype $, BELGIUM: Brussels, colln Wesmael (IRSNB), designated by Shaw, (1984) [examined].Chelonus impressus Herrich-Schaffer, 1838: 153. Syntypes, GERMANY (lost). [Synonymised by Reinhard, 1867: 367.]Ascogaster nigricornis Thomson, 1892: 1719. LECTOTYPE $, SWEDEN: Thomson colln (ZI), here designated [examined]. Syn. n. Ascogaster cynipum Thomson, 1892: 1720. Holotype cf , SWEDEN: Thomson colln (ZI) [examined]. Syn.n.Ascogaster egregius Kokujev, 1895: 83. Holotype cf , U.S.S.R. (AS) [examined]. Syn. n.Chelonus nigrator Szepligeti, 1896: 303. Holotype $, YUGOSLAVIA: Buccari, 16.viii.1889 (Biro) (HNHM) [examined]. Syn. n.Ascogaster epinotiae Watanabe, 1937: 76. Holotype $, JAPAN: Hokkaido, Sapporo, 14.vii.1927 (Uchida) (UEI) [examined]. Syn. n. Wesmael (1835: 239) described quadridentata from five males and two females. Seven specimensin his collection bear the label 'Ascogaster quadridentata mini', agree with the originaldescription and are conspecific. Reinhard examined Herrich-Schaffer's original material of Chelonus impressus and hisplacement of this species in synonymy with quadridentata is therefore accepted. Three specimens (1 cT, 2 9) are in Thomson's collection above the name nigricornis', they areprobably conspecific but the male and one of the females are in poor condition and are notreadily identifiable. I designate as lectotype the better-preserved female, labelled 'type' byGraham. It agrees precisely with Thomson's description and is within the limits of variation ofquadridentata. The holotype of Chelonus nigrator Szepligeti bears a label with the data cited in thedescription and is within the limits of variation of quadridentata. The holotype of epinotiaeWatanabe is smaller than average for quadridentata but is within the limits of variation of thatspecies. One male in the collection of the AS, Leningrad agrees precisely with what Kokujev wroteabout egregius, is labelled 'Asc. egregius m.' in Kokujev's handwriting and is here accepted asthe holotype. It comes within the limits of variation of quadridentata. $. Antenna long, 29-33 segmented. Flagellum slightly dilated medially, weakly tapered to apex, medialflagellar segments at most slightly longer than broad, generally slightly broader than long. Head rounded THE PALAEARCTIC SPECIES OF ASCOGASTER 377 behind eyes. Temple contracted (more strongly so in smaller specimens); about equal in length to eye indorsal view. Occiput moderately concave. Frons slightly excavate behind eyes, generally strongly rugose.Vertex rugose, sometimes reticulate-rugose. Ocelli on line, OO = 3-0-3-5 OD. Eyes fairly large, notprotuberant. Malar space about half height of eye. Face protuberant, about twice as broad as high; finelyirregularly rugose, occasionally almost reticulate, with a small medial tubercle on the upper part of face, afine carina extending from this to the anterior ocellus. Clypeus about half breadth of face, not stronglyprotuberant; polished, punctate except sometimes rugose laterally, apical border produced medially to apoint and often with a small but distinct dentate tubercle. Mandibles moderately twisted, always with asmall dentate flange formed at the junction of hypostomal and genal carinae, projecting beneath the baseof the mandible. Pronotum projecting slightly in front of mesonotum; strongly rugose-reticulate laterally.Notaulices rather indistinct, obscured by the strong reticulate-rugose sculpture of mesonotum though deepand easily seen under oblique illumination. Precoxal suture likewise obscured; mesopleuron completelycoarsely reticulate-rugose except random small areas smooth, punctate. Propodeum coarsely regularlyreticulate-rugose with a medial transverse carina raised into a medial pair and a lateral pair of prominentdentate flanges. Carapace oval, generally laterally compressed posteriorly so that it appears sharplypointed in dorsal view, sometimes the point is produced into a tubercle, particularly in specimens of smallerthan average size, in larger specimens the apex is often rounded, not pointed in dorsal view; posteroventralrim distinctly in front of apex. In some specimens, particularly those of smaller than average size, thecarapace narrower in dorsal view and flatter. Hind coxa always strongly transversely striate-rugose. Colour black, antenna at base brown or testaceous, mandibles brown, base of femur and tibia of foreleg,base of tibia of midleg and base of tibia of hind leg testaceous. Sometimes legs more extensively testaceous,in the lightest specimens only hind coxa black and mid and hind legs only lightly infuscate in the areas whichare normally dark. Occasional specimens have a small pale mark at base of carapace on its lateral borderbut it is always diffuse. Cf . Same as $ except antenna longer, flagellum not dilated medially, all flagellar segments longer thanbroad. MATERIAL EXAMINED 130 Cf, 121 $. Belgium, Bulgaria, Czechoslovakia, France, Germany, Great Britain, Greece, Hungary,Italy, Japan, Netherlands, Sweden, Turkey, U.S.S.R., Yugoslavia. HOSTS. Cydia pomonella (L.), Cydia funebrana (Treitschke), Cydia pallifrontana (Lienig &Zeller), Epiblema uddmanniana (L.), Spilonota ocellana (Denis & Schiffermiiller) (Lepidop-tera: Tortricidae). Yponomeuta padella (L.) (Lepidoptera: Yponomeutidae). REMARKS. It has been common practice for any specimen of Ascogaster with prominent flangeson the propodeum to be identified as quadridentata; this characteristic, however, is possessed bymany species in the quadridentata- and bidentula- groups so it is not a reliable means ofdiscriminating quadridentata. A. quadridentata is a common and widespread Holarctic species,rather variable in colour, size and certain other characteristics, probably because of its manyhost species. The smaller specimens tend to be darker in colour, more slender-bodied, with aprominent apical tubercle on the carapace, more coarsely sculptured and with shorter temples;Watanabe (1937) described the small form of quadridentata as epinotiae. I am convinced thatthese differences are infraspecific, resulting from the small size. The characteristics cited aboveoccur sporadically in otherwise typical quadridentata and in small specimens, one or more of thedifferentiating characteristics is often of the normal quadridentata form. There is therefore nocombination of characteristics by which the small form can be maintained as a distinct species.Larger specimens are stouter in build, the carapace generally more rounded apically and thehead relatively more massive; they bear a close resemblance to scabricula. A. quadridentata alsohas a superficial resemblance to brevicornis (q.v.). Ascogaster reticulata Watanabe(Figs 61, 65, 66) Ascogaster reticulatus Watanabe, 1967: 41. Holotype cf [cited as $], JAPAN: Hokkaido, Asahigawa,15.vi.1966 (Kamijo) (UEI), [examined]. Watanabe (1967: 42) confused the sexes of this species, describing the male as female and viceversa. The carapace of reticulata is very deep and the ventral opening short and it is therefore 378 T. HUDDLESTON difficult to examine the genitalia. Watanabe reasonably concluded that the specimens with thelonger antennae were male, as is usual in Braconidae. In reticulata, however, as in several otherspecies of Ascogaster, the converse is true. $. Antenna long, 35-38 segmented. Flagellum weakly dilated medially, tapered to apex, two or three ofmedial segments about as long as broad, rest distinctly longer than broad. Head strongly contracted behindeyes. Temple about as long as eye in dorsal view or slightly shorter. Occiput moderately concave. Ocellilarge, OO = 2-5 OD; on line. Eyes moderately large. Malar space about half height of eye. Face about 1-5times as broad as high, protuberant, coarsely irregularly rugose; a strong carina running from upper part offace between antennae to fore ocellus; ventral border of antennal scrobes produced, forming a flangewhich projects forwards and upwards from face. Clypeus not strongly protuberant, narrower than face;transversely depressed in apical half, strongly punctate, matt, apical border almost flat without any trace oftooth or tubercle. Mandible fairly stout, slightly twisted. Pronotum projecting slightly in front ofmesonotum, coarsely irregularly rugose laterally. Notaulices indistinct, foveolate. Rest of mesonotumreticulate-rugose posteriorly and laterally; transversely rugose and foveolate anteromedially. Precoxalsuture indistinct among strong reticulate-rugose sculpture of mesopleuron, ventrally mesopleuron general-ly punctate. Propodeum completely coarsely reticulate-rugose, divided by a medial transverse carinawhich is raised medially and laterally into prominent dentate flanges. Carapace short, deep; postero ventralrim conspicuously in front of apex so that ventral opening is only slightly more than two-thirds total lengthof carapace. Hind coxa smooth punctate; basitarsus of hindleg with a ventral longitudinal keel bearing arow of short, stout bristles on its edge and with a row of long, erect bristles on either side. Colour black; antenna at base, mandible, palps testaceous; foreleg testaceous except tarsus infuscate,midleg testaceous except tibia at apex and tarsus infuscate, hindleg testaceous except base of coxa, apex offemur, tibia and tarsus more or less heavily infuscate but tibia with an ivory or pale yellow medial band. Cf . Same as $ except antenna shorter, 32-34 segmented, flagellum barely dilated medially and weaklytapered to apex. MATERIAL EXAMINED 7 cf , 6 . Czechoslovakia: 4 cf, 4 $ , Banska Stiavnica (Capek) (CC; BMNH); 1 cT, Klak, 1959 (Capek)(CC); 1 $, Kajlovka, 1962 (Capek) (CC). Japan: holotype cf; 1 Of, 1 $ (paratypes), same data asholotype. HOSTS. Adoxophyes orana (Fischer von Roslerstamm), Archips issikii Kodama, Archipsoporana (L.), Archips pulchra (Butler) (Lepidoptera: Tortricidae). REMARKS. The exceptionally short, deep carapace of this species is most distinctive and is onlylikely to be confused with that of varipes from which reticulata is easily distinguished by itsstrongly, irregularly rugose face and smooth, punctate hind coxae. I have examined one specimen from Korea (Yeson (Paik) (RNH)), reared from an appleleafroller (probably Adoxophes orana), that is probably conspecific with reticulata although ithas a coarsely reticulate-rugose face, the carapace yellow at the base and the legs lighter incolour. Ascogaster rufipes (Latreille)(Figs 69, 77-79) Sigalphus rufipes Latreille, 1809: 14. Syntypes, FRANCE (lost). Sigalphus elegans Nees von Esenbeck, 1816: 264. Type-material, GERMANY (lost). [Synonymised by Thomson, 1892: 1716.]Chelonus fasciatus Dahlbom, 1833: 163. LECTOTYPE $, SWEDEN (ZI), here designated [examined]. [Synonymised by Thomson, 1874: 583.]Ascogaster ratzeburgii Marshall, 1885: 146. Holotype cf, GREAT BRITAIN: Norfolk, Brundall, 3.vii.l881 (Bridgman) (NCM) [examined]. Syn. n.Ascogaster soror Telenga, 1941: 324. Lectotype cf, U.S.S.R.: 'Sinelnikovo', vi.1930 (AS) [examined]. [Lectotype selected by Tobias.] Syn. n. I have not been able to locate Latreille's specimens of rufipes and the description of the species isinsufficient for it to be identified with certainty. Several interpretations of the name have beenused and this has led to some confusion. Thomson's (1892: 1716) remarks upon rufipes clearedaway some of the confusion, showing that Nees von Esenbeck's intrepretation of rufipes was THE PALAEARCTIC SPECIES OF ASCOGASTER 379 incorrect. Thomson put forward his own interpretation of rufipes and so tactitly rejected that ofReinhard (1867: 365) who regarded rufipes and bidentula as conspecific. It is clear from hisdescription that he had identified Wesmael's species correctly, but I cannot accept that bidentulafits Latreille's description of rufipes and so reject Reinhard's interpretation and follow Thom-son's. Nees von Esenbeck's description of elegans fits perfectly well with Thomson's interpreta-tion of rufipes. Four specimens are named fasciatus in the Fallen collection. Two are conspecific and onebears a label 'C. fasciatus Dbm var a' in Dahlbom's handwriting; I designate this specimen aslectotype. The third specimen bearing a small label 'var b' in Dahlbom's handwriting belongs tovaripes and the fourth bearing the label 'var c' in Dahlbom's handwriting is a male abdominator. TTie holotype of ratzeburgii comes well within the limits of variation of rufipes. The lectotypeof soror has three labels, one bearing the data published by Telenga, a second in Telenga'shandwriting, 'Ascogaster soror sp.n. cf' and a third bearing Tobias' lectotype designation(though I have not been able to find where, or whether this was published); soror comes wellwithin the limits of variation of rufipes. 9 . Antenna 34-37 segmented. Flagellum generally slightly broader medially than at base, then tapering toapex, most segments longer than broad, a few medial segments about as long as broad. Head slightlyexpanded behind eyes, then roundly contracted. Temple about 1-5 times length of eye in dorsal view.Occiput strongly concave. Frons behind antennae excavate. Ocelli small, OO = 3-5 OD; ocellar triangleobtuse, ocelli almost on line. Weak carina extends from upper part of face to fore ocellus. Eyes notprotuberant. Genae in face view rounded. Face protuberant, about twice as broad as high, reticulate-rugose. Clypeus weakly protuberant; smooth, punctate; apical border produced medially into a largedentate tubercle. Mandibles short, stout, moderately twisted. Pronotum projecting distinctly in front ofmesonotum, laterally completely rugose except for a polished medial area. Mesonotum almost completelystrongly reticulate-rugose except sometimes a small punctate area laterally. Notaulices indistinguishable.Precoxal suture indistinguishable, almost all mesopleuron strongly reticulate-foveolate. Propodeumcompletely regularly reticulate- rugose, divided by a medial transverse carina which is raised into strongflanges laterally and weak ones medially; dorsal surface rather long. Carapace rather long and narrow,coarsely reticulate-rugose, the longitudinal element predominant, expecially anteriorly; posteroventralrim not reaching apex of carapace. Hypopygium short. Ovipositor short, straight, abruptly narrowedshortly before apex. Ovipositor sheaths thick. Hind coxa polished, punctate, with fine transverse rugaedorsally. Colour black, apical border of clypeus often rufous, mandibles at base testaceous; carapace at baseyellow for slightly more than a third of its length; legs yellow or testaceous except all tarsi generallyinfuscate, at least in part, hind coxa at base and mid and hind tibiae often also infuscate. C?. Same as 9 except antenna 32-35 segmented; carapace longer, narrower, always completely dark;legs generally darker. MATERIAL EXAMINED 91 cf, 39 9- Czechoslovakia, France, Great Britain, Hungary, Ireland, Netherlands, Spain, Sweden,U.S.S.R., Yugoslavia. HOSTS. No reared material examined. REMARKS. Smaller and more delicate than most other species in the quadridentata-group; onlybrevicornis and the smaller specimens of quadridentata are of comparable size. A. rufipes alsoexhibits more sexual dimorphism than others in the quadridentata-group; it was Marshall'sfailure to appreciate this which led him to describe the male of rufipes as a distinct species eventhough he collected it with the female (for which he used the name elegans). Ascogaster scabricula (Dahlbom)(Figs 68, 71, 74-76) Chelonus scabriculus Dahlbom, 1833: 166. LECTOTYPE $ , SWEDEN: Fallen colln (ZI), here designated [examined].Ascogaster limitatus Wesmael, 1838: 163. LECTOTYPE $, BELGIUM: Wesmael colln (IRSNB), here designated [examined]. Syn. n. 380 T. HUDDLESTON Ascogaster clypealis Thomson, 1892: 1719. LECTOTYPE $, SWEDEN: Oland, Thomson colln (ZI), heredesignated [examined]. Syn. n. Four specimens in Fallen's collection are named scabriculus; I designate as lectotype thespecimen labelled 'C. scabriculus Dbm. var. a' in Dahlbom's handwriting. It agrees withDahlbom's description of 'scabriculus var. a' and is labelled 'cf' although I believe it to be 9-The other three specimens belong each to a different species; the only one which bears a label inDahlbom's handwriting 'var. c' belongs to rufidens. Fahringer (1934: 544) synonymised scabri-cula with similis which, from his description he obviously interpreted as being conspecific withbrevicornis', I reject this placement of scabricula. Wesmael (1838: 164) stated that he hadexamined two specimens of limitatus. In Wesmael's collection there is one specimen labelled'Ascogaster limitatus mini'; it agrees precisely with Wesmael's description and I accept it as asyntype. Wesmael gave the number of antennal segments as 33; in the material I have examinedthe number ranges from 30-32. Three females (one without a head) are above the name clypealis in Thomson's collection;they are conspecific and agree with the original description. I designate as lectotype thebest-preserved female, which Graham labelled as type despite the presence on the pin of a label'4-dentatus' in Thomson's handwriting; the significance of such labels is discussed undercanifrons. $. Antenna 30-32 segmented. Flagellum dilated medially, tapered to end, most segments in distal halfquadrate, in proximal half longer than broad (only three or four at base as much as twice as long as broad).Head rounded, more or less expanded behind eyes. Temple about 1-5 times length of eye in dorsal view.Occiput strongly concave. Ocelli small, OO = 3-5 OD; almost on line. Frons behind antennae slightlyexcavate, strongly rugose. Eyes small, not protuberant. Face protuberant, about twice as broad as high,completely regularly reticulate-rugose, divided from clypeus by a deep furrow. Anterior tentorial pitsdeep. Clypeus protuberant, polished, densely punctate, but with some rugae laterally; apical borderproduced medially with no distinct tubercle. Mandible moderately twisted, densely punctate at base.Genal carina joining hypostomal carina at base of mandible, forming a small flange which projects beneathbase of mandible. Pronotum projecting but little in front of mesonotum, completely coarsely rugose.Mesonotum broad, almost completely rugose, the notaulices more or less indistinguishable. Mesopleuroncompletely coarsely reticulate-rugose, precoxal suture indistinguishable. Propodeum divided into dorsaland posterior surfaces by a weak medial transverse carina which is raised into two low broad humpsmedially and into two strong flanges laterally; completely, regularly reticulate-rugose with no distinctcarinae. Carapace rounded, about 1-5 times as long as broad, deep; posteroventral rim not significantly infront of apex. Hypopygium short. Ovipositor short, straight, thick at base, abruptly narrowed shortlybefore apex. Hind coxa punctate except dorsally transversely striate. Colour black, fore tibia pale, at least in part. Cf. Same as $ except posteroventral rim of carapace distinctly before apex; head more stronglycontracted. MATERIAL EXAMINED 5 Cf , 12 $. Belgium: 1 $ (lectotype of limitatus). France: 3 cf , Lent, Ain (Audras) (MNHN); 1 cf ,Megere, vi.1948 (Granger) (MNHN); 1 cf , 'env. Paris', 20.vi.1889 (Chretien) (MNHN); 1 $, Chartrettes,21.vi.1942 (MNHN); 1 $, Drome, Chap-en- Vercors, vii.1938 (MNHN); 1 $, St. M-Vesubie, 27.vii.1950(Granger) (MNHN); 1 $, Le Vesinet, 1867 (Sichel) (MNHN). Sweden: 3 $, Solna, Bergshamra, viii.1976(Quinlan & Huddleston) (BMNH); 1 $ (lectotype of scabricula) (ZI); 1 $ (lectotype of clypealis) (ZI); 2 $(paralectotypes of clypealis), Thomson colln (ZI). HOSTS. No reared material examined. Species inquirendae Ascogaster atriceps (Ratzeburg) Chelonus atriceps Ratzeburg, 1844: 33. ?Syntype, GERMANY (IP) [examined].The purported syntype of C. atriceps is in too poor condition to be identified. THE PALAEARCTIC SPECIES OF ASCOGASTER 381 Ascogaster contractus (Ratzeburg)Chelonus contractus Ratzeburg, 1848: 24. Syntypes, GERMANY (lost). Ascogaster dentiventris TelengaAscogaster dentiventrisTelenga, 1941: 311. Syntypes, U.S.S.R. (lost). Ascogaster erythrothorax MarshallChelonus erythrothorax Marshall, 1898: 171. Syntypes, FRANCE (lost). Ascogaster kabystanicus Tobias Ascogaster kabystanicus Tobias, 1976: 235. Holotype cf, U.S.S.R.: Azerbaidzhan, 19. v. 1972 (Kasparyari)(AS) [examined] . This species belongs to the caucasica-group in which the males are extremely difficult todifferentiate. Ascogaster lapponicus ThomsonAscogaster lapponicus Thomson, 1874: 588. Syntypes, SWEDEN (lost). Ascogaster pallidicornis CurtisAscogaster pallidicornis Curtis, 1837: folio 672. Syntypes, GREAT BRITAIN (lost). One specimen named pallidicornis in Curtis's collection disagrees in most particulars with theoriginal description and cannot therefore be a syntype. Ascogaster quadridens (Herrich-Schaffer)Chelonus quadridens Herrich-Schaffer, 1838: 154. Syntypes, GERMANY (lost). This species has been placed in synonymy with quadridentata but the latter never has pale marksat the base of the carapace as described for quadridens. Ascogaster rubripes (Lucas)Chelonus rubripes Lucas, 1849: 339. Syntypes, ALGERIA (lost). Ascogaster similis (Nees von Esenbeck)Chelonus similis Nees von Esenbeck, 1816: 262. Syntypes, GERMANY (lost). Ascogaster tersus ReinhardAscogaster tersus Reinhard, 1865: 366. Syntypes, GERMANY (lost). One specimen named tersus in Reinhard's collection disagrees with the original description andcannot therefore be a syntype. Excluded species Phanerotoma maculata (Wollaston ) comb. n. Ascogaster maculata Wollaston, 1858: 24. Holotype $, MADEIRA (BMNH) [examined]. Acknowledgements I thank the following for their kindness in lending me the type-specimens and much of the other materialupon which this revision is based: Dr C. van Achterberg (Leiden); Dr A. A. Allen (Reigate); Mr S.Belokobylskij (Leningrad); Dr M. Capek (Prague); Mr R. Danielsson (Lund); Dr P. Dessart (Brussels);Dr M. Fischer (Vienna); Dr M. W. R. de V. Graham (Oxford); Dr E. Haeselbarth (Munich); Dr K.-J.Hedqvist (Stockholm); Dr A. Irwin (Norwich); Dr J. Lukas (Trencin); Dr K. Maeto (Kyushu); Dr P. M. 382 T. HUDDLESTON Marsh (Washington); Prof. Morge (Berlin); Dr J. Papp (Budapest); Dr B. Petersen (Copenhagen);Dr M. R. Shaw (Edinburgh); Dr S. R. Shaw (Washington); Dr B. Sigwalt (Paris); Dr S. Takaga (Sapporo);Dr V. I. Tobias (Leningrad); Dr H. K. Townes (Ann Arbor); Dr H. J. Vlug (Wageningen); Dr A. Zaykov(Plovdiv). I also thank Mr U. Gardenfors (Lund), Dr G. E. J. Nixon and Mr C. R. Vardy (BMNH) for helpwith translations, and my wife, Linda, for typing the manuscript. References Abdinbekova, A. A. 1969. Braconids (Hymenoptera, Braconidae) of the lesser Caucasus in Azerbaidzan. Isvestya Academii Nauk Azerbaidzenskoi SSR 1969 (5): 69-76.Achterberg, C. van 1976. A preliminary key to the subfamilies of the Braconidae (Hymenoptera). Tijdschrift voor Entomologie 119: 33-78.Allen, H. W. 1962. Parasites of the oriental fruit moth in the eastern United States. Technical Bulletin of the United States Department of Agriculture no. 1265: 136 pp.Ashmead, W. H. 1906. Descriptions of new Hymenoptera from Japan. Proceedings of the United States National Museum 30: 169-201.Baker, C. F. 1926. Braconidae-Cheloninae of the Philippines, Malaya, and Australia. The Philippine Journal of Science 31: 451^*89.Boyce, H. R. 1936. Laboratory breeding of Ascogaster carpocapsae Viereck with notes on biology and larval morphology. Canadian Entomologist 68: 241-246.Capek, M. 1970. A new classification of the Braconidae (Hymenoptera) based on the cephalic structures of the final instar larva and biological evidence. Canadian Entomologist 102: 846-875.Cox, J. A. 1932 Ascogaster carpocapsae Viereck , an important parasite of the codling moth and the oriental fruit moth. Technical Bulletin of the New York State Agricultural Experimental Station no. 188: 1-26.Curtis, J. 1829. A guide to an arrangement of British insects vi pp. + 256 columns. London. 1837. British Entomology 14: folio 672. London. Dahlbom, A. G. 1833. Forsok till beskrifning ofver Hymenopterslagtet Chelonus, med dertill horande Skandinaviske arter. Kongliga Swenska Vetenskaps-Academiens Handlingar 53: 147-167.Dudarenko, G. P. 1974. Formation of the abdominal carapace in braconids (Hymenoptera, Braconidae) and some aspects of the classification of the family. Entomological Review, Washington. 53: 80-90 [translated from the Russian in Entomologicheskoye Obozneniye].Evenhuis, H. H. & Vlug, H. J. 1983. The hymenopterous parasites of leaf-feeding appletortricids (Lepidoptera, Tortricidae) in the Netherlands. Tijdschrift voor Entomologie 128: 109-135.Fahringer, J. 1934. Opuscula braconologica 3 Palaearktische Region 2 (5-8): 321-594.Fischer, M. 1965. Die Braconiden des Steiermarkischen Landesmuseums Joanneum in Graz (Hymenop-tera, Braconidae). Abteilungfiir Zoologie und Botanik am Landesmuseum Joanneum 21: 3-29.Fitton, M. G. 1982. A catalogue and reclassification of the Ichneumonidae (Hymenoptera) described by C. G. Thomson. Bulletin of the British Museum (Natural History) (Entomology) 45: 1-119.Foerster, A. 1862. Synopsis der Familien und Gattungen der Braconen. Verhandlungen des natur- historisches Vereines der preussischen Rheinlande und Westfalens 19: 225-288. Graham, M. W. R. de V. 1955. A taxonomic revision of the Chelonine Braconidae (Insecta, Hymenop-tera). Unpublished thesis, Oxford. Hellen, W. 1953. Die Ascogaster-arten Finnlands (Hym., Brae.). Notulae Entomologicae 33: 84-88.Herrich-Schaffer, G. A. W. 1838 In Panzer, G. W. F., Deutschlands Insecten Heft 153-54. Regensburg.Huddleston, T. 1980. A revision of the western Palaearctic species of the genus Meteorus (Hymenoptera: Braconidae). Bulletin of the British Museum (Natural History) (Entomology) 41: 1-58.Kainoh, Y., Hiyori, T. & Tamaki, Y. 1982. Kairomone of the egg-larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Applied Entomology and Zoology 17: 102-110.Kainoh, Y. & Tamaki, Y. 1982. Searching behaviour and oviposition of the egg-larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Applied Entomology and Zoology 17: 194-206. Kirchner, L. 1867. Catalogus Hymenopterorum Europae 285 pp. Vindobonae.Kokujev, N. 1895. Fragments Braconologiques. Horae Societatis Entomologicae Rossicae 29: 77-95.Latreille, P. A. 1809. Genera Crustaceorum etlnsectorum 4: 399 pp. Paris.Lucas, H. 1849. Exploration Scientifique de I'Algerie, Zoologie 3: 527 pp. Paris.Marshall, T. A. 1885. Monograph of British Braconidae, Part 1. The Transactions of the Entomological Society of London 1885: 1-280. 1888. In Andre, E. , Species des Hymenopteres d' Europe et d'Algerie 4: 603 pp. 1897-1900. In Andre, E. , Species des Hymenopteres d' Europe et d'Algerie 5 bis: 373 pp. THE PALAEARCTIC SPECIES OF ASCOGASTER 383 Nees von Esenbeck, C. G. 1816. Ichneumonides adsciti, in genera et familias divisi. Magazin. Gesellschaft Naturforschender Freunde zu Berlin 7 (1813): 243-277.1834. Hymenopterorum Ichneumonibus affinium monographiae, genera Europaea et species illus- trantes 1: 320 pp. Stuttgart & Tubingen.Papp, J. 1967. Ergebnisse der zoologischen Forschungen von Dr. Z. Kaszab in der Mongolei Braconidae (Hymenoptera) 1. Acta Zoologica Academiae Scientiarum Hungaricae 13: 191-226.1971. Results of the Zoological Explorations of Dr. Z. Kaszab in Mongolia. Hymenoptera: Braconidae II. Acta Zoologica Academiae Scientiarum Hungaricae 17: 51-90.Ratzeburg, J. T. C. 1844. Die Ichneumonen der Forstinsekten inforstlicher und entomologischer Beziehung 1:224 pp. Berlin.1848. Die Ichneumonen der Forstinsekten in forstlicher und entomologischer Beziehung 2: 238 pp. Berlin. 1852. Die Ichneumonen der Forstinsekten in forstlicher und entomologischer Beziehung 3: 272 pp. Berlin.Reinhard, H. 1867. Beitrage zur Kenntniss einiger Braconiden-Gattungen. Berliner entomologische Zeitschrift 11: 251-274.Richards, O. W. 1977. Hymenoptera. Introduction and key to families 2nd ed. Handbooks for the Identification of British Insects 6(1): 100 pp.Rosenberg, H. T. 1934. The biology and distribution in France of the larval parasites of Cydia pomonella L. Bulletin of Entomological Research 25: 201-256. Ruthe, J. F. 1855. Beitrage zur Kenntniss der Braconiden. Stettiner entomologischer Zeitung 16: 291-294.Shaw, S. R. 1984. A taxonomic study of Nearctic Ascogaster and a description of a new genus Lepto- drepana (Hymenoptera: Braconidae). Entomography 2: 1-54.Shenefelt, R. D. 1973. Catalogue Hymenopterorum (Nov. ed.) part 10 Braconidae 6: 813-936. S'Gravenhage.Short, J. R. T. 1952. The morphology of the head of larval Hymenoptera with special reference to the head of Ichneumonoidea including a classification of the final instar larvae of the Braconidae. Transactions of the Royal Entomological Society of London 103: 27-84.Shu-Sheng, L. & Carver, M. 1982. The effect of temperature on the adult integumental coloration of Aphidius smithi. Entomologia Experimental etApplicata 32: 54-60.Sonan, J. 1932. Notes on some Braconidae and Ichneumonidae from Formosa with descriptions of 18 new species. Transactions of the Natural History Society of Formosa 22: 66-86.Szepligeti, G. 1896. Adatok a Magyar fauna Braconidainak ismerethehez. Termeszetrajzl Fiizetek 19: 165-186.1908. Braconiden aus der Sammlung des ungarischen national Museums II. Annales Historico- Naturales Musei Nationalis Hungarici 6: 397-427.Telenga, N. A. 1941. Insects Hymenoptera, Family Braconidae, Subfamily Braconinae (continued) and Sigalphinae. Fauna SSSR 5 (3): 466 pp.Thomson, C. G. 1874. Ofversigt af Sveriges Sigalpher. Opuscula entomologica 6: 553-588. 1892. Bidrag till Braconidernas kannedom. Opuscula entomologica 16: 1659-1751. Tobias, V. I. 1964. New species and genus of Braconids (Hymenoptera, Braconidae) from Kazachstan. Trudi Zoologischeskovo Instituta, Leningrad 34: 177-234.1972. New species of braconids (Hymenoptera, Braconidae) from Mongolia. Nasekome Mongolii 9: 585-612. 1976. Braconidae of the Caucasus. Opredelitel Po Faune SSSR 110: 1-286. Watanabe, C. 1937. A contribution to the knowledge of the Braconid fauna of the Empire of Japan. Journal of the Faculty of Agriculture Hokkaido Imperial University 42 (1): 1-188.Waterston, J. 1926. Ascogaster annularis Nees (Hym.) in the London district and a note on its probable host. Entomologist 59: 174.Wesmael, C. 1835. Monographic des Braconides de Belgique. Nouveaux Memoires de I' Academic Royale des sciences et belles-lettres de Bruxelles 9: 1-252.1838. Monographic des Braconides de Belgique. Nouveaux Memoires de I' Academic Royal des sciences et belles-lettres de Bruxelles 11: 1-166.Wharton, R. A. 1980. Review of the Nearctic Alysiini (Hymenoptera, Braconidae) with discussions of generic relationships within the tribe. University of California publications in entomology 88: 1-112.Wollaston, T. V. 1858. Brief diagnostic characters of undescribed Madeiran insects. Annals and Magazine of Natural History (3) 1: 18-28.Yoneda, Y. 1978. A new species of the genus Ascogaster Wesmael (Hymenoptera, Braconidae) from Japan. Kontyu 46: 291-296. 384 T. HUDDLESTON 8 Figs 1-9 1 , Chelonus sp. , forewing, anterodorsal. 2, Ascogaster rufidens Wesmael, forewing, anterodor-sal. 3, 4, A. excavataTelenga; (3) head, frontal; (4) carapace, distal. 5, 6, A. semenoviTelenga; (5) head,frontal; (6) carapace, distal. 7, A. dentifer Tobias, head, frontal. 8, A nachitshevanica Abdinbekova,head, frontal. 9, A. abdominator (Dahlbom), head, frontal. THE PALAEARCTIC SPECIES OF ASCOGASTER 385 10 Figs 10-18 10, Ascogaster abdominator (Dahlbom), carapace, lateral. 11, A. nachitshevanica Abdin-bekova, carapace, lateral. 12, A. dentifer Tobias, carapace, lateral. 13, A. kasparyani Tobias, head,frontal. 14, A. excisa (Herrich-Schaffer), clypeus, anteroventral, 15, 16, A. caucasica Kokujev; (15)head, frontal; (16) head, lateral. 17, 18, A. bicarinata (Herrich-Schaffer); (17) head, frontal; (18) head,lateral. 386 T. HUDDLESTON Figs 19^-28 19, 20, Ascogaster kasparyani Tobias; (19) carapace, dorsal; (20) carapace, lateral. 21, A.bicarinata (Herrich-Schaffer), carapace, distal. 22, A. caucasica Kokujev, carapace, distal. 23, 24, A.excisa (Herrich-Schaffer); (23) carapace, lateral; (24) carapace, posterolateral. 25, A. annularis (Nees),forewing, anterodistal. 26, A. gonocephala Wesmael, head, dorsal. 27, A. klugii, (Nees), head, dorsal.28, A. exigua sp. n., head, dorsal. THE PALAEARCTIC SPECIES OF ASCOGASTER 387 Figs 2-40 29, Ascogaster annularis (Nees), head, dorsal, 30-32, A. grahami sp. n.; (30) head, dorsal;(31) carapace, dorsal; (32) carapace, lateral. 33, 34. A. annularis (Nees); (33) carapace, lateral; (34)carapace, posterolateral. 35, A. klugii (Nees), carapace, dorsal. 36, A. exigua sp. n., carapace, dorsal.37, A. rufidens Wesmael, clypeus, antero ventral. 38, A. varipes Wesmael, head, lateral. 39, 40, A.rufidens Wesmael; (39) carapace, dorsal; (40) head, lateral. T. HUDDLESTON Figs 41-51 41 , 42, Ascogaster longicornis sp. n. ; (41) head, dorsal; (42) clypeus, anteroventral. 43, 44, A.perkinsi sp. n. ; (43) head, frontal; (44) carapace, dorsal. 45-47, A. consobrina Curtis; (45) head, frontal;(46) carapace, dorsal; (47) carapace, lateral. 48, A. albitarsus Reinhard, carapace, dorsal. 49, A.arisanica Sonan, head, dorsal. 50, 51, A. bidentula Wesmael; (50) head, dorsal; (51) clypeus, anteroven-tral. THE PALAEARCTIC SPECIES OF ASCOGASTER 389 52 61 Figs 52-61 52, Ascogaster arisanica Sonan, carapace, dorsal. 53, A. bidentula Wesmael, carapace, dorsal,54, A. arisanica Sonan, clypeus, anteroventral. 56, 57, A. armata Wesmael; (56) head, dorsal; (57) head,anterolateral. 58, 59, A. canifrons Wesmael; (58) head, frontal; (59) head, dorsal. 60, A. quadridentataWesmael, carapace, lateral. 61, A. reticulata Watanabe, carapace, lateral. 390 T. HUDDLESTON 72 71 Figs 62-72 62-64, Ascogaster quadridentata Wesmael; (62) head, frontal; (63) head, dorsal; (64)carapace, dorsal. 65, 66, A. reticulata Watanabe; (65) clypeus, anteroventral; (66) head, dorsal. 67, A.brevicornis Wesmael, antenna, distal. 68, A. scabricula (Dahlbom), head, dorsal. 69, A. rufipes(Latreille), carapace, lateral. 70, A. dispar Fahringer, carapace, lateral. 71, A. scabricula (Dahlbom),mandibles, anteroventral. 72, A dispar Fahringer, mandibles, anteroventral. THE PALAEARCTIC SPECIES OF ASCOGASTER 391 79 75 Figs 73-79 73, Ascogaster dispar Fahringer, carapace, dorsal. 74-76, A. scabricula (Dahlbom); (74)carapace, lateral; (75) carapace, dorsal; (76) head, frontal. 77-79, A. rufipes (Latreille); (77) carapace,dorsal; (78) head, dorsal; (79) head, frontal. 392 T. HUDDLESTON Index abdominator 357albitarsus 364an n ularis 360arisanica 365armata 372atamiensis 366atriceps 380 bicarinata 354bidentula 366brevicornis 373 canifrons 374caucasica 354cavifrons 370clypealis 380consobrina 367contracta 381cynipum 376 dentifer 359dentiventris 381dispar 375 egregius 376elegans 378epinotiae 376erythrothorax 381esenbeckii 372excavata 352excisa 355exigua 361 Synonyms are in italics. femoralis 357fulviventris 357fuscipennis 366 gibbiscuta 366gonocephala 362grahami 362graniger 374 impressus 376instabilis 357 jaroslawensis 370 kabystanica 381kasachstanicus 352kasparyani 356klugii 363koslovi 375kyushuensis 352 laevigator 369lapponica 381leptopus 364limitatus 379longicornis 368longiventris 355luteicornis 372 maculata 381mlokossewitchi 354monilicornis 373multiarticulatus 366 nee?// 363nigrator 376nigricornis 376 pallidicornis 381perkinsi 368pulchellus 372punctuator 359 quadridens 381quadridentata 376 ratzeburgii 378 reticulata 377 rostrata 354 rubripes 381 ruficeps 363 rufidens 369 ru/zpes (Herrich-Schaffer) 369 rufipes (Latreille) 378 rufiventris 357 scabricula 379semenovi 352similis 381soror 378spinifer 375sternalis 370 terra 381 fasciatus 378 nachitshevanica 359 varipes 370 British Museum (Natural History) Milkweed butterflies: their cladistics and biology P. R. Ackery & R. I. Vane- Wright The Danainae, a subfamily of the Nymphalidae, contains only some 150 species, yet aspects oftheir biology have stimulated far more attention than can be justified by species numbersalone. In recent years, an expansive literature has grown, considering aspects of theircourtship and pre-courtship behaviour, migration, larval hostplant associations, mimicry andgenetics. The popularity of danaines among biologists can certainly be attributed to thiscombination, within one small group, of so many of the factors that make butterflies such aninteresting group to study. The obvious need to place this wealth of biological data within anacceptable systematic framework provided the impetus for this volume. Started eight years ago within the conventions of evolution by natural selection andHennig's phylogenetic systematics, the book is now largely about natural history (what theanimals have and do, where they live and how they develop) and natural groups - as revealedby a form of analysis approaching that practised by the new school of 'transformed cladistics'.The authors have prepared a handbook that will appeal to a wide range of biologists, frommuseum taxonomists to field ecologists. 424 pp (approx.), 12 pp colour, 73 b/w plates, line and graphic illustrations, maps, extensive bibliography.ISBN 565 00893 5. Publication September 1984. Price 50. Titles to be published in Volume 49 Afrotropical jumping plant lice of the family Triozidae (Homoptera: Psylloidea). By David Hollis. The taxonomy of the western European grasshopper of the genus Euchorthippus, with specialreference to their songs (Orthopetera: Acrididae). By D. Ragge & W. J. Reynolds An historical review of the higher classification of the Noctuidae. By Ian Kitching The Pimplinae, Xoridinae, Acaenitinae and Lycorininae (Hymenoptera: Ichneumonidae). By I. D.Gauld The Palaearctic species of Ascogaster (Hymenoptera: Braconidae) By T. Huddleston Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, SuffolkPrinted in Great Britain by Henry Ling Ltd, Dorchester