Bulletin of the British Museum (Natural History). British Museum (Natural History) Converted as part of the ABLE project by Dauvit King London : BM(NH) Set of bird parts only held at TOS 112. General Library missing Vol 55-56, 1989-90 Vol.1 (1950)- 4 5 6 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.) Bulletin of the ,, British Museum (Natural History) LIBRARY , Zoology series Vol 44 1983 British Museum (Natural History)London 1983 Dates of publication of the parts No 1 27 January 1983 No 2 24 February 1983 No 3 31 March 1983 No 4 28 April 1983 No 5 26 May 1983 No 6 30 June 1983 ISSN 0007-1 498 Printed in Great Britain by Henry Ling Ltd, at the Dorset Press, Dorchester, Dorset ContentsZoology Volume 44 Page No 1 Observations on the systematics of the genus Difflugia in Britain(Rhizopoda, Protozoa).By Colin G. Ogden 1 No 2 Miscellanea Cirolana cranchi Leach, 1818 (Crustacea: Isopoda: Cirolanidae) redescribed, with notes on its distribution. By N. L. Bruce & Joan Ellis 75 Valettieta, a new genus of deep-sea amphipod (Gammaridae: Lysia-nassidae) with descriptions of two new species from the NorthAtlantic Ocean.By Roger J. Lincoln & Michael H. Thurston . . . .85 Three new genera of misophrioid copepods from the near-bottom plankton community in the North Atlantic Ocean. By G. A. Boxshall 103 Larval development of British prawns and shrimps (Crustacea:Decapoda: Natantia) 4. Palaemon (Palaemon) serratus (Pennant,1777) and functional morphology of swimming.By A. A. Fincham . T ...... 125 The larval development of the Angular Crab, Goneplax rhomboides (Linnaeus) (Decapoda: Brachyura). By R. W. Ingle & P. F. Clark 163 The larval and first crab stages of three Inachus species (Crustacea: Decapoda: Majidae); a morphological and statistical analysis. By Paul F. Clark 179 No 3 A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin R. Curds & Irene C. H. Wu 191 No 4 The Opthalmotilapia assemblage of cichlid fishes reconsidered. By Peter Humphry Greenwood . . . . . 249 No 5 Osteology, genitalia and relationships of the Acanthodactylus (Rep-tilia: Lacertidae).By E. N. Arnold 29 1 No 6 Morphological studies on some Difflugiidae from Yugoslavia (Rhi-zopoda, Protozoa).By Colin G. Ogden & Andjelija Zivkovic . 341 Bulletin of the British Museum (Natural History) Observations on the systematics of thegenus Difflugia in Britain (Rhizopoda,Protozoa) Colin G. Ogden Zoology series Vol 44 No 1 27 January 1983 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, andan 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 SW7 5BD,England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Zool.) Trustees of the British Museum (Natural History), 1983 The Zoology Series is edited in the Museum's Department of ZoologyKeeper of Zoology : Dr J. G. ShealsEditor of Bulletin : Dr C. R. Curds Assistant Editor: Mr C. G. Ogden ISSN 0007- 1 498 Zoology series Vol 44 No 1 pp 1-73British Museum (Natural History)Cromwell RoadLondon SW7 5BD Issued 27 January 1983 Observations on the systematics of the genusDifflugia in Britain (Rhizopoda, Protozoa). F GENERAL 28 JANW83 ^4L Colin G. Ogden Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Contents Synopsis Introduction Materials and methods Systematic descriptions Pyriform and elongate species . . . . Pointed species or those with protruberances . Ovoid or spherical species Compressed species Discussion References Index . 11 222 274559666973 Synopsis Detailed descriptions of the shell structure in thirty-eight species of Difflugia are given, andfurther information on one species which has been redescribed in an earlier report (Ogden &Fairman, 1979) is included. Two new species, Difflugia hiraethogii and D. stoutii, are described,and other systematic changes include the following new combinations: D. cylindrus (Thomas,1953), D. lacustris (Penard, 1899), D. microclaviformis (Kourov, 1925), D. microstoma (Thomas,1954), D. parva (Thomas, 1954), D. tennis (Penard, 1890), D. tricornis (Jung, 1936), D. venusta(Penard, 1902) and D. distenda nom, nov., D. gassowskii nom, nov., D. paulii nom. nov., D. rotundanom. nov. Seventeen of these redescriptions are new records for the British Isles. The structure of theshell is discussed and the patterning of the organic cement which binds the particles together is shown tobe a useful taxonomic character. Introduction The taxonomy of specimens belonging to the genus Difflugia is based mainly ondifferences in size and shape of the agglutinated shells constructed by these animals.Comparison of cytoplasmic features are either difficult or impossible because mostof it is encased by the shell which is often opaque. Differences in the shell featureshave resulted in about three hundred named species, varieties and forms beingcurrently attributed to the genus. This proliferation of species is due to a combination of thelack of good diagnostic features and inadequate descriptions. The problem is clearly shownin the only comprehensive survey of the genus, based on African specimens, by Gauthier-Lievre and Thomas (1958). In this work the authors had difficulties with several groups ofindividuals which shared common features, and as a result, about half of the 129 speciesdescribed were designated as varieties or forms. Some earlier studies (Ogden, 1979, 1980; Ogden & Fairman, 1979) were con-cerned with the variation of specimens having a pyriform shell, and the surface Bull. Br. Mus. nat. Hist. (Zool.)44 (1): 1-73 Issued 2 7 January 1983 2 C. G. OGDEN ultrastructure as revealed by the scanning electron microscope. The results showedthat there was usually a limited variation in size, shape and composition of the shellwithin a species, whilst in some instances the patterning of the organic cement whichbinds the particles together was a reliable specific character. The present accountuses the experience gained from these previous studies to extend the examination ofshell structure in Difflugia, and to establish specific features for ten of the varieties ofDifflugia oblonga listed by Gauthier-Lievre and Thomas (1958). Thirty nine speciesare described, of which some are new records for the British Isles. Some 600specimens have been examined and over 3000 micrographs representing differentaspects of the shells are retained in the Protozoa Section, Department of Zoology aspart of the study collection. Materials and methods Samples have been collected from several localities in England and Wales during the lastfour years. The material gathered varied from mosses, water plants including the substrateassociated with the roots, and clumps of algae. Type of habitat ranged from areas of bog,banks of streams and small ponds in the New Forest, Lake District and North Wales to thedykes and rivers of Norfolk and Suffolk. Information relating to locality, date and type ofsample is given with the description of each species as several have been found in more thanone habitat. Specimens of Difflugia were selected by searching through small isolates of material in apetri dish. Specimens were extracted using a glass micropipette, washed in several transfersthrough distilled water, and then individual shells were manipulated with a single-hairbrushonto a small drop of Araldite on a previously cleaned cover slip. When about twentyspecimens were positioned on each cover slip it was glued with Araldite onto a standardaluminium stub. In a few instances the very small delicate specimens collapsed before orduring manipulation, this problem was overcome by transferring the washed specimens in asmall drop of water onto cleaned cover slips and allowing it to dry. Prepared stubs werecoated evenly with gold or gold/palladium, using a conventional sputter coating device, andexamined in a Cambridge Stereoscan SI 80 operating at lOkV. The results were recorded onIlford HP5 film. Systematic descriptions In a previous review (Gauthier-Lievre & Thomas, 1958) of the genus the species weredivided into ten groups, namely; lobed, collared, compressed, urceolate, globose, ovoid-globose, elongate, acute angled, horned and pyriform, but no particular significance wasattached to these groupings. However, it does show the diversity of shell shape that has beenincluded in the genus. Similarly in this report the species have been grouped together, thepyriform and elongate species are described first, followed by those which are pointed orhave aboral protruberances, then the ovoid or spherical and finally the two compressedspecies. Pyriform and elongate speciesDifflugia bryophila (Penard, 1 902) Jung, 1 942 Difflugia pyriformis var. bryophila Penard, 1 902 Difflugia oblonga var. bryophila (Penard, 1902) Gauthier-Lievre & Thomas, 1958 DESCRIPTION. The shell is brown, pyriform, with the sides usually tapering evenly to theaperture (Figs, la & d), although the occasional specimen may be slightly mis-aligned or DIFFLUGIA IN BRITAIN Fig. 1 Difflugia bryophila: a, lateral view of shell with large particles obscuring the basic outlinex780; b, apertural view x760; c, portion of shell surface showing the distribution of organiccement (arrowed) x 5100; d, lateral view of specimen with typical basic outline x440; e, detailof organic cement x 24 000. 4 C. G. OGDEN Table 1 Range of measurements (in um) of pyriform specimens. *measurements quoted from an earlier report (Ogden & Fairman, 1979) have a large particle obscuring the even-tapering (Fig. la). It is composed mainly of amixture of small to medium pieces of quartz and the occasional diatom frustule or siliceousflagellate cyst. In common with most rough, thick shells, organic cement is seen infrequently(Fig. Ic), it appears as an open network, with a mesh about 350-450 nm in diameter. Thewalls of the mesh are not pronounced but blend with the matrix, the mesh openings are oftencovered by inner strands of cement which appear to form a smaller mesh (Fig. le). Theaperture is circular and surrounded by small particles (Fig. Ib), in several of the specimensexamined small flagellate cysts had also been incorporated to give an irregular margin. MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED Specimens were collected from samples of Sphagnum moss gathered atMately Bog, Lyndhurst, New Forest, Hampshire in March, 1980 and at Mynnd Hiraethog,Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Belgium(Chardez & Gaspar, 1976; Couteaux, 1969), British Isles (Cash el al, 1919), Chile (Jung,1942), Congo (Chardez, 1964), Czechoslovakia (Rosa, 1957), France (Thomas, 1954),Gambia (Decloitre, 1947), Germany (Jung, 1936), Roumania (Godeanu et al., 1973), Russia(Kourov, 1925), Spain (Gracia, \912a), Switzerland (Penard, 1902), Tunisia (Gauthier-Lievre & Thomas, 1958), West Africa (Decloitre, 1948). REMARKS. This description is in good agreement with Penard (1902) who stated that the sideswere rather straight and that it was formed of stones which were usually large and angular. Amongst the sample from Matley Bog were specimens of Pontigulasia which wereindistinguishable in size and shape from those of D. bryophila. Observations of the formerspecimens by optical microscopy to determine the presence of an inner diaphragm, thedistinguishing generic character for Pontigulasia, are often difficult. However, detailedexamination by scanning electron microscopy showed that, unless the apertural opening isblocked, it was easy to identify the inner diaphragm and that additionally there are differ-ences in their organic cement patterns (a review of the genus Pontigulasia is in preparation). DIFFLUGIA IN BRITAIN 5 The random selection of these specimens from the sample gave a ratio of 3 : 1 in favour ofD. bryophila. Difflugia cylindrus (Thomas, 1953) comb. nov.Difflugia oblonga var. cylindrus Thomas, 1953 DESCRIPTION. The shell is usually opaque, cylindrical, tapering evenly from the aboralregion to the aperture (Fig. 2a). It is composed mainly of medium to large pieces of quartzwith the occasional diatom frustule on the rough surface, but the latter are seldomincorporated into the thick structure. Small areas of organic cement in the form of a networkare seen infrequently (Fig. 2d). Each mesh of the network is between 300-500 nm indiameter, has thick walls and is covered by a thin perforated layer of cement, the pores ofwhich are evenly distributed and are about 30 nm in diameter (Figs. 2e & f)- Theaperture is irregular in both outline and composition, being roughly circular and usuallysurrounded by small particles but often incorporating medium particles that produce ajagged margin (Fig. 2b & c). MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from a sample of aquatic plants taken at thebanks of the River Brett, near Hadleigh, Suffolk in August, 1979. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Austria(Laminger, 19736, 1975, 1976), Belgium (Chardez, 1980; Chardez & Caspar, 1976), Congo(Chardez, 1964), France (Thomas, 1953, 1954; Thomas & Mabille, 1956), Germany(Voeltz-Hohn, 1971), Ivory Coast (Gauthier-Lievre & Thomas, 1958), Morocco(Gauthier-Lievre & Thomas, 1958), Poland (Moraczewski, 1965). REMARKS. This species was initially described by Thomas (1953) as a new variety of D.oblonga, he also listed what he considered to be previous descriptions of this variety underthe general descriptions of D. oblonga given by earlier workers. These are not repeated here,but suggest that this variety may be found throughout Europe. The shape, structure and sizeof the shell, together with the distinct patterning of the organic cement matrix as describedabove, are considered to be good specific characters sufficient to warrant the raising of thisvariety to species rank. Difflugia gassowskii nom. nov. Difflugia pyriformis longicollis Gassowsky, 1936 Difflugia longcollis (Gassowsky, 1936) Ogden & Hedley, 1980 DESCRIPTION. The shell is pyriform, with a distinct short neck about one-third of the bodylength, and a rounded aboral region (Fig. 3a). It is rough and composed of small to mediumangular pieces of quartz, small areas of organic cement are sometimes seen between particles(Fig. 3b). The cement is in the form of a network, the mesh of which is about 400-550 nm indiameter and the walls 125-200 nm thick, each mesh enclosure has a covering with smallperforations about 50 nm in diameter (Fig. 3c). The aperture is circular. MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from samples of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire in March, 1980 and at ClocaenogForest, Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Belgium (Chardez, 1980; Couteaux, 1969), British Isles(Ogden & Hedley, 1980), Germany (Voeltz-Hohn, 1971), Netherlands (Hoogenraad &Groot, 1940a), Nigeria (Gauthier-Lievre & Thomas, 1958), Poland (Moraczewski, 1961,1965), Roumania (Godeanu et al, 1973), Russia (Gassowsky, 1936), Spain (Gracia, 1972a;Margalef, 1955). C. G. OGDEN Fig. 2 Difflugia cylindrus: a, lateral view x450; b, side view of aperture to illustrate the irregularmargin x770; c, apertural view x410; d, shell surface with areas of organic cement(arrowed) x4200; e, typical arrangement of organic cement network x 8700; d, detail of organiccement x 27 000. DIFFLUGIA IN BRITAIN Fig. 3 Difflugia gassowskii: a, lateral view x810; b, shell surface with small areas of organiccement x 7600; c, detail of organic cement network x 24 000. REMARKS. This species has been redescribed recently (Ogden & Hedley, 1980), but isincluded here on two counts, the added detail of the organic cement and the change of name.The name D. longicollis was used initially by Ehrenberg (1854) to describe specimens whichnow are not considered to belong to the genus Difflugia, nevertheless, under the Rules ofZoological Nomenclature the name is preoccupied. I am grateful to Dr. E. G. Merinfeld,Dalhousie University, Halifax, Nova Scotia, for drawing my attention to this point. ENTYMOLOGY. This species is named after Dr G. N. Gassowsky who first described it fromthe Kareliens Lakes. Difflugia glansPenard, 1902 DESCRIPTION. The shell is dark, elongate ovoid, tapering towards the aperture and evenlyrounded aborally (Fig. 4a). It is composed mainly of small to medium pieces of quartz, C. G. OGDEN a Fig. 4 Dijjlugia glans: a, lateral view x 1400; b, detail of small unidentified cyst x3000; c,apertural view x 1000; d, portion of shell surface showing the close packing of particles x 3500. packed closely together with only a minimum amount of organic cement visible (Fig. 4d). Asmall cyst, of unknown origin, is seen adhering to the surface of the illustrated specimen(Fig. 4b), otherwise the outline is usually well denned. The aperture is circular andsurrounded by both small and medium particles (Fig. 4c). MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from a sample of water plants taken at the DIFFLUGIA IN BRITAIN 9 banks of the River Brett, near Hadleigh, Suffolk in August, 1979 and a gathering ofSphagnum moss from Mynnd Hiraethog, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Austria^ (Laminger, 19736), Congo (Chardez, 1964),Czechoslovakia (Rosa & Lhotsky, 1971; StSpanek, 1952), Germany (Jung, 1936; Schonborn,1962), Italy (Grandori & Grandori, 1934; Rampi, 1950), Russia (Kourov, 1925). REMARKS. This species has not been widely reported, which may be because it falls withinthe range of three more common species namely, D. penardi, D. manicata and D. rubescens.It differs from these three species in being a distinct ovoid shape, from D. rubescens in beingdark and not transparent although Penard's description of a thin, fragile shell for D. glansdoes not seem compatible with a dark structure which usually suggests that it is robust andstrong and from D. penardi and D. manicata in aperture size and shape. Difftugia lacustris (Penard, 1 899) comb. nov. Difflugia pyriformis var. lacustris Penard, 1899Difflugia oblonga var. lacustris Cash & Hopkinson, 1 909 DESCRIPTION. The shell is transparent or hyaline, elongate, cylindrical or slightly pyriform(Figs. 5a & b). It is composed of small to medium pieces of quartz, diatom frustules and smallsiliceous flagellate cysts blended together to form a thin structure intermediate betweensmooth and rough. Only small areas of organic cement occur at the junction of the shellcomponents (Fig. 5d). The cement is in the form of thick-walled rings, between 700-800 nmin diameter, perforated with either three or four holes, 120-160 nm in diameter, which givesthese units a similar shape to a button (Fig. 5e). The cement may occasionally be seen eitheras rings with a slight indentation or as a network of joined rings. When organised as anetwork the walls of individual rings may be fused together but the typical button-likeform are usually seen at the edges. The aperture is usually circular and surrounded by smallparticles so that the margin is smooth (Fig. 5c). MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from three samples in the same locality,aquatic plants at the edge of a pond in Burley, and two gatherings of Sphagnum fromopposite banks of a small stream at Holmsley Lodge, Burley, New Forest, Hampshire inMarch, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina(Dioni, 1970; Lena & Zaidenwerg, 1975), Austria (Laminger, 19736 1974, 1975), Belgium(Chardez, 1980; Chardez & Gaspar, 1976; Oye, 1953), British Isles (Cash & Hopkinson,1909), Congo (Chardez, 1964; Gauthier-Lievre & Thomas, 1958), Czechoslovakia(Stepanek, 1967), France (Thomas, 1954; Thomas & Mabille, 1956), Germany (Jung, 1936;Schonborn, 1962a & b), Guatemala (Laminger, 1973a), Italy (Rampi, 1950), Ivory Coast(Gauthier-Lievre & Thomas, 1958), Mexico (Laminger, 1973a), Morocco (Gauthier-Lievre& Thomas, 1958), Poland (Moraczewski, 1961, 1965), Roumania (Godeanu et al., 1973),Russia (Kourov, 1925), Switzerland (Penard, 1902), United States of America (Lamingeretal., 1979;Wailes, 1912). REMARKS. The specimens described here agree well with the original description (Penard,1899) shell long, cylindrical, larger in the rear, rarely with a small constriction of theneck and those given later by Penard (1902) and Cash & Hopkinson (1909). Nevertheless,it would appear that these earlier descriptions were based on groups of similar specimens,hence Penard's reference to a constriction of the neck and the diverse illustrations providedby Cash & Hopkinson. As a result of the latter diagrams, two specimens were tentativelyidentified as D. lacustris in a previous publication (Ogden, 1980). Additional specimens to 10 C. G. OGDEN Fig. 5 Difflugia lacustris: a, lateral view x 450; b, alternative view of same specimen (a.) toillustrate the uniform shape x280; c, apertural view x500; d, shell surface showing thedistribution of organic cement x 5800; e, detail of organic cement network x 24 000. DIFFLUGIA IN BRITAIN 1 1 the two mentioned above have been examined and are clearly seen to differ from D. lacustris,they are now referred to D. linearis (see below). The present account shows that D. lacustris has a well denned shape and a distinctivepatterning of organic cement which are considered sufficient to designate this a distinctspecies. Difflugia lanceolata Penard, 1 890 DESCRIPTION. The shell is yellow or hyaline, lanceolate, tapering from the widest diametersituated about two- thirds of the body-length from the aperture, to give a clean outline that isrounded aborally and evenly angled towards the aperture (Fig. 6a). It is composed of smallto medium flattish pieces of quartz and some flat diatom frustules so arranged that the shell isthin and smooth, the surface frequently appearing as though it had been polished, a featurethat often permits easy identification. An angular piece of quartz may occasionally protrudefrom the surface but these are uncommon and limited to one or two in any one shell. As theshell components are usually arranged so that they are in close contact with each other thereare no large areas of organic cement, nevertheless, a network of small rings of organic cementmay be seen between these particles (Figs. 6d, e, f)- The rings are 240-300 nm in diameterand have a distinct wall about 1 50-200 nm in thickness with a smooth membrane over themesh. When several rings fuse to form a sheet the thick wall is still usuallv apparent. Theaperture is circular and well defined because the rim has a thin covering of organic cement(Figs. 6b & c). Variation appears to be limited to cigar-shaped specimens which have almost parallelsides, one such specimen is illustrated by Ogden & Hedley (1980). MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED The majority of specimens came from a sample of Sphagnum mosscollected close to a small stream at Holmsley Lodge, Burley, New Forest, Hampshire, a fewcame from a similar sample on the opposite bank where they were equally abundant, bothsamples were taken in March, 1 980. GEOGRAPHICAL DISTRIBUTION. Argentina (Vucetich, \913a, /?), Austria (Laminger, 1972c),Belgium (Chardez, 1961; Oye, 1953), Brazil (Green, 1975), British Isles (Cash & Hopkinson,1909; Ogden & Hedley, 1980), Canada (Decloitre, 1965), China (Decloitre, 1965), Congo(Chardez, 1964), Czechoslovakia (Opravilova, 1974), France (Thomas, 1954; Thomas &Mabille, 1956), Germany (Schonborn, 1975), Hungary (Gal, 1969), Java (Bartos, 1963a),Morocco (Gauthier-Lievre & Thomas, 1958), Poland (Golemansky, 1970; Moraczewski,1961, 1965), Roumania (Godeanu et al, 1973), Sudan (Gauthier-Lievre & Thomas, 1958),Switzerland (Penard, 1902), Venezuela (Grospietsch, 1975), West Africa (Decloitre, 1965),United States of America (Decloitre, 1965). REMARKS. The organic rim surrounding the aperture is shared with one other pyriformspecies of Difflugia, namely D. rubescens where the cement is in the form of tooth-likeprojections (see PI. 66 in Ogden & Hedley, 1980). The only other species of which we areaware that has an organic rim is D. oviformis, but this was transferred to a new genus Netzeliaby Ogden, 1979. D. lanceolata is characterized by its uniform size and outline. Difflugia linearis (Penard, 1 890) Gauthier-Lievre & Thomas, 1958 Difflugia oblonga var. linearis Penard, 1890Difflugia lacustris in Ogden, 1980 DESCRIPTION. The shell is transparent, flask-shaped or elongate pyriform, having a long thinneck with parallel sides and a slightly swollen, rounded aboral region (Fig. 7 a). The surface issometimes slightly uneven because of projecting particles, but generally it has a regular 12 C. G. OGDEN Fig. 6 Difflugia lanceolata: a, lateral view x 930; b, apertural view x 760; c, side view of apertureto illustrate the thin covering of organic cement around the rim x 1500; d, shell surface withsmall isolates of organic cement x4400; c, small rings of organic cement between particlesx 7600; f, detail of organic cement x 24 000. DIFFLUGIA IN BRITAIN 13 Fig. 7 Difflugia linearis: a, lateral view x950; b, apertural view x!200; c, portion of shellsurface showing areas of organic cement (arrowed) x 14 000; d, detail of organic cementX 24 000. outline. It is composed of a mixture of flattened pieces of quartz, small whole, flat diatomfrustules, fragments of flattish frustules, small siliceous shell plates and round flagellate cysts.Small areas of organic cement, in the form of a network with an open mesh, are occasionallyseen (Fig. 7c), The open mesh has a diameter of about 300 nm and walls 100 nm thick (Fig.7d). In appearance, the openings of the mesh suggest that it may have been covered at sometime and has subsequently been broken, but only the examination of further specimens willestablish its normal condition. The aperture is circular and usually surrounded by smallparticles (Fig. 7b). MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from Sphagnum moss gathered at Holmsley 14 C. G. OGDEN Lodge, Burley, New Forest, Hampshire on two occasions, May, 1978 and March, 1980, andat Myndd Hiraethog, Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Austria (Laminger, 19736, 1975), Belgium (Chardez, 19616),British Isles (Ogden, 1980), Bulgaria (Golemansky, 1967), Congo (Chardez, 1964), France(Thomas, 1954), Germany (Penard, 1890), Ivory Coast (Gauthier-Lievre & Thomas, 1958),Nepal (Laminger, 19726), Roumania (Godeanu et at, 1973). REMARKS. The initial description of this species (Penard, 1890) was brief, and relies mainlyon the diagrams. The scarcity of subsequent reports may be due to this inadequatedescription or the difficulty in finding this species, which by being thin, long and transparent Fig. 8 Difflugia manicata: a, lateral view x 1300; b, apertural view x 1 100; c, and d, detail oforganic cement with the ill-defined inner structure (arrowed) x 30 000. DIFFLUGIA IN BRITAIN 15 Fig. 9 Difflugia minutissima: a, lateral view to show the arrangement of flattish particles x 6700;b, detail of aperture x 10 000; c, latero-apertural view x4900. makes it unusually elusive. Two specimens described earlier (Ogden, 1980) and tentativelyidentified as Difflugia lacustris, because of their similarity to the description given by Cash &Hopkinson (1909) and especially to one figure (PI. XIX Fig. 1), are now redescribed as D.linearis. Additional specimens, plus the benefit of being able to compare these withspecimens of D. lacustris (see p. 9), allows the former identification to be rectified and showthat D. linearis is a distinct species. Difflugia linearis can be differentiated from other pyriform species by its distinctiveflask-like shape, thin, transparent shell and small aperture. 16 C. G. OGDEN Difflugia manicata Penard, 1 902 DESCRIPTION. The shell is yellow or brown, pyriform, tapering evenly and gradually from arounded aboral extremity towards the aperture (Fig. 8a). The surface is rough andcomposed mainly of small to medium pieces of quartz, although the occasional specimenmay have large particles added. Small areas of organic cement are seen infrequently, due tothe close packing of the shell material, but when present they appear either as strands or aspores in a matrix (Fig. 8c). These pores are about 300 nm in diameter and have an ill-definedinner structure which appears to have smaller pores about 1 30 nm in diameter (Figs. 8c & d).The aperture is circular and surrounded by a distinct pattern of small particles(Fig. 8b). MEASUREMENTS (in um). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from samples taken at three sites,Sphagnum moss gatherings at Holmsley Lodge, Burley, New Forest, Hampshire in March,1980 and Mynnd Hiraethog, Clwyd, North Wales in August, 1980, and from aquatic plantstaken at the banks of the River Brett, near Hadleigh, Suffolk in August, 1979. GEOGRAPHICAL DISTRIBUTION. Austria (Laminger, 1971, 1972c), Belgium (Chardez, 196 \b\British Isles (Cash el al, 1919; Ogden & Medley, 1980), Congo (Stepanek, 1963), Italy(Grandori & Grandori, 1934), Poland (Moraczewski, 1965; Pateff, 1926, 1927), Switzerland(Penard, 1902). REMARKS. Although this species was thought to be uncommon by Cash, Wailes &Hopkinson (1919), it is suggested that this has been one of the overlooked species and that itwill probably be found as one of the most commonly distributed. The wide range of habitatalready reported, lake, river and Sphagnum moss would appear to support this opinion. Difflugia minutissima Penard, 1904Sexangularia minutissima (Penard, 1 904) Deflandre, 1931 DESCRIPTION. The shell is transparent, elongate or ovoid, rounded aborally and taperingslightly towards the aperture (Figs. 9a & c). It is composed of thin flattish pieces of quartz,and the occasional diatom frustule, which are arranged so that they overlap, but the result isstill a fragile structure. Only small strands of organic cement have been seen at somejunctions. The aperature is basically circular, any variation is usually due to the irregulararrangement of particles around the opening (Fig. 9b). MEASUREMENTS (in urn). See Tables 1 and 2. MATERIALS EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Mynnd Hiraethog, Clwyd, North Wales in August 1980. GEOGRAPHICAL DISTRIBUTION Switzerland (Penard, 1904) REMARKS. This species appears to be known only from the initial description (Penard, 1 904),although Deflandre (1931) using Penard's description suggested that it should be transferredto the genus Sexangularia Awerintzew, 1906. The main features of this latter genus are'chitinous shell, with or without the addition of extraneous particles; polygonal in transversesection, most often hexagonal'. At present it is considered that these features are insufficientto differentiate this genus from Difflugia, and therefore refer the included species toDifflugia-D. minutissima Penard, 1904; D. parvula (Awerintzew, 1906) and D. polyderaDeflandre, 1931. The specimens described here were found adhering to extraneous particles when preparedby air-drying and because of this it is not possible to examine them in transverse section, DIFFLUGI A IN BRITAIN 17 nevertheless, they do not appear to be polygonal. The value of this feature in relation to afragile shell is questionable, especially as Penard's observations would have been carried outon specimens under a cover slip, which would allow a degree of compression. The specimensshare similar dimensions to those given by Penard for D. minutissima and are so designated. Difflugia parva (Thomas, 1954) comb. nov.Dijjlugia oblonga var. parva Thomas, 1954 DESCRIPTION. The shell is pyriform, tapering evenly from the swollen and rounded aboralthird, to the aperture for the remaining two-thirds (Fig. lOa). It is composed mainly of amixture of small to medium pieces of quartz, often with the addition of two or three largepieces. Organic cement is seen between particles usually as a series of single units (Fig. 1 Oc),which may overlap but are seldom fused to form a network. Each unit is a ring about650-750 nm in external diameter, 300-380 nm internal diameter, with walls about 150 nmthick. A small mesh covers the inner portion of each ring (Fig. lOd). The aperture is circularand surrounded mainly by small particles (Fig. lOb). MEASUREMENTS (in urn). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from Sphagnum moss gathered at MatleyBog, Lyndhurst, New Forest, Hampshire in March, 1980; Mynnd Hiraethog, Denbigh,Clwyd, North Wales in August, 1980 and aquatic plants from a pond at Hurley, New Forest,Hampshire in March, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina (Lena& Zaidenwerg, 1975), Austria (Laminger, 19736), Belgium (Chardez, 1980; Chardez &Caspar, 1976; Couteaux, 1969), Bulgaria (Golemansky, 1967), France (Thomas, 1954;Thomas & Mabille, 1956), Germany (Schonborn, 1965; Voeltz-Hohn, 1971), Ivory Coast(Gauthier-Lievre & Thomas, 1958), Poland (Golemansky, 1970; Moraczewski, 1965),Venezuela (Grospietsch, 1975). REMARKS. There have been several reports of D. oblonga var. parva since Thomas (1954)described his new variety, although size was the only diagnostic feature used and illustrated(PI. Ill, Fig. 1) by him to differentiate this variety from D. oblonga. The figure given alsoshows the difference in shell composition, described below as a specific feature. This variety is considered as a distinct species from Z>. oblonga because of its clean outline,relatively smooth surface and detailed cement pattern. Difflugia paulii nom. nov.Difflugia oblonga var. elongata Oye, 1953 DESCRIPTION. The shell is transparent, slim and elongate, tapering evenly from just anteriorof the mid-body region towards the aperture, the even-tapering is more apparent in Fig. libwhereas the upper side of Fig. 1 la has a misleading hump, the posterior region is slightlyswollen, curving sharply and smoothly at the extremity (Fig. 1 la). It is composed of flattishpieces of quartz to give a smooth appearance, with small areas of organic cement oftenapparent as part of the matrix (Fig. lie). The cement is in the form of small perforated conesabout 600 nm in diameter, the perforations being about 100-1 50 nm in diameter (Fig. 1 Id).The aperture is circular and surrounded by small pieces of quartz (Fig. 1 Ib). MEASUREMENTS (in nm). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Mynnd Hiraethog, Denbigh, Clwyd, North Wales in August, 1980. 18 C. G. OGDEN Fig. 10 Difflugia parva: a, lateral view x 710; b, apertural view x 790; c, shell surface showingareas of organic cement x 9800; d, detail of organic cement x 30 000. DIFFLUGIA IN BRITAIN J V * ,. f '^ tRS- .y .-.!'*^i-J*'..3L'^L .-, Fig. 11 Difflugia paulii: a, lateral view x800; b, apertural view x790; c, portion of shellsurface x 4600; d, detail of organic cement x 1 5 000. GEOGRAPHICAL DISTRIBUTION. Austria (Laminger, 19736), Belgium (Oye, 1953), Congo(Gauthier-Lievre & Thomas, 1958), Spain (Gracia, 19720). REMARKS. This species has been described on two occasions, from Belgium by Oye (1953)and from Africa by Gauthier-Lievre & Thomas (1958). Oye (1953) stated that thesespecimens were slender in comparison with D. lacustris, and considered that this feature andthe limited grains of quartz in the shell was sufficient to warrant a new variety. Gauthier-Lievre and Thomas (1958) agreed with this earlier description and designation. Of the more elongated pyriform species, D. paulii in body length appears to occupy aposition mid-way between D. linearis and D. lacustris, the present work shows that it isdistinct from these two species in outline, elemental composition and patterning of the 20 C. G. OGDEN organic cement, in addition, it can be differentiated from D. linearis by the size of theaperture (Table 1). ETYMOLOGY. This species is named after Dr Paul van Oye whose original description wasbased on specimens from a pond in Belgium. DifflugiapetricolaCash, 1909 This species has been redescribed recently by Ogden & Fairman (1979). It is reported hereonly to demonstrate the regularity of dimensions between specimens from differentlocalities, see Tables 1 and 2, because these are often used as taxonomic features. MATERIAL EXAMINED. Specimens were collected from Sphagnum moss gathered at two sites,Holmsley Lodge, Burley, New Forest, Hampshire in March, 1980 and Mynnd Hiraethog,Denbigh, Clwyd, North Wales in August, 1980. Dijflugia pristis Penard, 1902 DESCRIPTION. The shell is brown or opaque, ovoid, tapering from the mid-body positiontowards the aperture and gracefully curved aborally (Fig. 12a). It is thin, smooth andcomposed of flattish pieces of quartz (Fig. 1 2a-c) or infrequently including flat pieces ofdiatom frustule. The particles are packed closely together and only small connections oforganic cement are visible (Figs. 12d & e). The aperture is circular with a regular margin(Fig. 12b). MEASUREMENTS (in urn). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from samples of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire in May, 1978 and at Myndd Hiraethog,Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. British Isles (Cash & Hopkinson, 1909), Czechoslovakia(Stepanek, 1967), Hungary (Varga, 1963), Italy (Grandori & Grandori, 1934), Java (BartoS,1963a), Poland (Moraczewski, 1965), Roumania (Godeanu et al, 1973), Switzerland(Penard, 1902). REMARKS. Cash and Hopkinson (1909) considered their specimens to be in good agreementwith Penard's (1902) description, except for the absence of refractive particles that made theshell appear opaque or black. This they attributed to the difference in habitat of theirsamples, the former being found amongst floating vegetation in clear water and Penard'sfrom a muddy lake bottom. Some differences in size have been reported since Penard whogave 45-65 um as the range of length, Cash and Hopkinson gave 60-65 um, whilst morerecently Varga (1963) found specimens measuring 44-48 um in length. The specimens reported here are smaller than any previously described, although in allother respects agree with Penard's description. The main features that distinguish D. pristisfrom D. pulex are the darker colour and the more regular shape. A single larger specimen, 56 um long, 31um wide, aperture 10 um in diameter, with thetypical pyriform shape (Fig. 13a & b) was found in the sample from Wales. This is reportedhere as being questionably D. pristis (?), being atypical in having a definite neck, thediscovery of more specimens may result in a more accurate identification. This specimen issimilar to those identified by Cash & Hopkinson (1909) as (?) D. pulex which they describedas being 'oval, tapering suddenly to a short neck'. DIFFLUGIA IN BRITAIN 21 J Fig. 12 Difflugia pristis: a, lateral view x2400; b, apertural view x2400; c, alternative view ofspecimen shown in a., to illustrate regular ovoid shape x 1700; d and e, portions of shell surfacewith small connections of organic cement (arrowed) x 9300 and x,14J)00. 22 C. G. OGDEN a Fig. 13 Difflugia pristis (?): a, lateral view to show distinct neck x 860; b, apertural view x 1 600. Difflugia pulex Penard, 1902Difflugia minuta minor Godeanu, 1972 DESCRIPTION. The shell is transparent, elongate or ovoid (Figs. 14a & d). It is composedmainly of a mixture of small thin pieces of flat quartz and pieces of diatom frustule, oftenwith whole frustules or round flagellate cysts adhering to the surface (Figs. 14a-e). Thearrangement of these particles is such that only small strands of organic cement are visible.The aperture is usually circular (Fig. 14c) but may vary due to the arrangement of thesurrounding particles (Fig. 14b). MEASUREMENTS (in urn). See Tables 1 & 2. MATERIAL EXAMINED Specimens were collected from samples of Sphagnum moss gathered atSubberthwaite, Broughton in Furaess, Cumbria in June 1979 and at Mynnd Hiraethog,Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION Argentina (Lena & Zaidenwerg, 1975), Australia (Playfair,1918), Austria (Laminger, J9736), Belgium (Chardez, 19616), British Isles (Cash &Hopkinson, 1909), Congo (Stdpanek, 1963), Czechoslovakia (Stepanek, 1967), Germany(Schonborn, \962a & b), Italy (Grandori & Grandori, 1934; Rampi, 1950), Java (BartoS,1963a), Netherlands (Hoogenraad & Groot, 1940), Roumania (Godeanu et al, 1972), Spain(Gracia, 19726), United States of America (Laminger et al, 1979), West Africa (Decloitre,1948). REMARKS. Penard's original description stated that the shell was-' pyriform, with or withoutnarrowing of the mouth, chitinoid, slightly yellow, covered with small scales or particles ofamorphous silica, plates, sufficiently transparent to examine the contents' and thatspecimens rarely exceeded 30 urn in length. Cash & Hopkinson (1909) described specimensbetween 65-70 urn in length that they tentatively identified as D. pulex, but on the basis ofboth Penard's and the present description it is now suggested they should be reassigned. The description of D. minuta minor given by Godeanu (1972), shell colourless circular incross section but with an irregular outline because of added quartz particles and a circularaperture, is so similar to that of D. pulex that it must be considered a synonym. The range ofmeasurements given are also similar to those quoted here (see Table 1 ). Although Penard gave 30 urn as the maximum length for this species, in the describedsample two specimens were just in excess of 40 urn and two 30 urn, but in all other respectswere similar to the smaller specimens, and all are considered to represent D. pulex. DIFFLUGIA IN BRITAIN Fig. 14 Difflugia pulex: a, lateral view x 1800; b, apertural view of specimen with irregularapertural opening x 1300; c, apertural view of specimen (a.) with circular aperture x 1500; d,lateral view of specimen mainly made of flattish particles x 1600; e, shell surface showingmixture of flat components x 5 100. 24 C. G. OGDEN Difflugia tenuis (Penard, 1 890) comb. nov. Difflugia pyriformis var. tenuis Penard, 1890Difflugia oblonga var. tenuis Wailes & Penard, 1911 DESCRIPTION. The shell is usually transparent, cylindrical or slightly pyriform, composed ofa mixture of mainly small to medium pieces of angular quartz, but with an occasional largeparticle added (Figs. 15a & b). Sometimes additional particles give a pointed outline to theaboral region, instead of the usual rounded contours. Organic cement is seen frequently assingle units between shell components (Fig. 15d), and occasionally in small patches wherethese units are in a network with a mesh diameter of 350-400 nm and walls 180-220 nmthick (Fig. 1 5e). The mesh surface is characterised by a thin, usually central, inner ring about1 50-200 nm in diameter (Fig. 15e). The aperture is roughly circular, often having anirregular outline because of the mixture of particles surrounding it (Fig. 1 5b). One specimen with projections similar to the rigid 'filament' structures described byPenard (1890), occurred in the present sample (Fig. 15c). These projections usually arisefrom a common base, on the surface of the quartz particles, and vary in size and number (Fig.16a). The projections are about 250 nm in diameter and may be up to 7 um in length, asmany as fifteen have been seen sharing a common base (Fig. 1 6b). On the present evidencethey seem to be rosettes of bacterial rods rather than the parasitic organisms suggested byPenard (1890). MEASUREMENTS (in urn). See Tables 1 and 2. MATERIAL EXAMINED. Specimens were collected from Sphagnum moss gathered at HolmsleyLodge, Burley, New Forest, Hampshire in March, 1980 and Mynnd Hiraethog, Denbigh,Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION Argentina (Dioni, 1970), British Isles (Cash et al, 1919),Germany (Penard, 1 890), Poland (Golemansky, 1970), Venezuela (Grospietsch, 1975). REMARKS. Although the specimens described here are a little longer than those reported byPenard (1890) they are otherwise in good agreement with his description. This species hasbeen identified previously on only a few occasions, but this may be due to its being mistakenfor the more common species in this size range, for example D. penardi and D. rubescens.There remains the question of specific biological requirements, food, temperature, pH etc.,which may be the reason why although the site at Holmsley has been sampled regularly forfive or six years, March 1980 was the first time that this species has been seen there and thenit appeared in significant numbers. This species is distinguished by its shape, size of apertureand the pattern of the organic cement. Difflugia viscidula Penard, 1 902 DESCRIPTION. This species has been redescribed recently by Ogden & Hedley (1980) but isreported again with some additional information. The shell is opaque, pyriform or elongate ovoid, aborally it is usually rounded (Fig. 17a)but may occasionally be pointed. It is composed of a mixture of different sizes of angularquartz, with organic cement seen infrequently as part of the surface matrix. The cementappears as single units squashed together so that they are adjacent or overlap (Fig. 1 7c). Theaperture is circular and usually surrounded by small particles which give it a characteristicwell-defined outline (Fig. 17b). The illustrated aperture has a cyst plug, which althoughbroken at one edge has in general a smooth surface, suggesting that it is mainly organic butreinforced by particles of quartz. About 10% of the examined specimens had an aperturalcyst plug. MEASUREMENTS (in um). See Tables 1 and 2. DIFFLUGIA IN BRITAIN 25 Fig. 15 Difflugia tennis: a, lateral view x 1000, b, apertural view x780; c, specimen with'filament-like' structures projecting from the surface x 780; d, shell surface showing small areasof organic cement x 8 100; e, detail of organic cement network x 24 000. 26 C. G. OGDEN Fig. 16 Difflugia tennis: a, portion of shell surface to show distribution of 'filament-like'structures x2500; b, rosette of 'filaments', tentatively identified as bacterial rods x 7000. MATERIAL EXAMINED Specimens were collected from a sample of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire in March, 1980. GEOGRAPHICAL DISTRIBUTION. Argentina (Boltovskoy & Lena, 1974), Austria (Laminger,1971), British Isles (Ogden & Hedley, 1980), Germany (Schonborn, \962a, 1965, 1975), Java(BartoS, 1963a),Roumania(Godeanu^0/., 1973), Switzerland (Penard, 1902). REMARKS. The present material extends the measurements we gave previously (Ogden &Hedley, 1980) to encompass those given by Penard (1902). There remains the problem ofpriority of names for these specimens because Penard (1902) described two distinct species,namely D. lemani Blanc, 1892 and D. viscidula sp. nov., which he differentiated mainly onsize; the former being 50-85 um long (three individuals were 100-200 um) and the latter180-260 um long. When he later found (Penard, 1905) that he had used incorrect measure-ments for D. lemani he suggested that D. viscidula should be regarded as a synonym.However, this proposal left his description of specimens under the name D. leman /-Penard,1902 p. 249, without a proper designated name. They were divided into two series, the firstslim, between 75-85 um long and the second wider, but smaller about 50 urn long. It wouldappear that no subsequent report has rectified this situation, and it would seem that now istoo late to change this because of the proliferation of species that have since been describedand fall within the size range of the earlier description. Furthermore, it appears that Blanc's(1892) original description of D. lemanii might contravene Article 8 of the InternationalCode of Zoological Nomenclature, because when first issued it would seem that it was notavailable by purchase or free distribution. There are no records of this publication in eitherthis Museum's Libraries nor in the British Library, although I have recently obtained aphotocopy from the Universitaire Lausanne. In Blanc's initial description of D. lemanii theshell dimensions varied as follows: body length between 180-310 um, breadth 130-150 um.and the aperture diameter was given as 1 lOum, it had a cylindrical body with almost parallelsides and the composition was such that it was friable and delicate. These measurements arein good agreement with Penard's 180-260 um body length for D. viscidula, and those givenhere in Table 1 except for the diameter of the aperture, this latter feature is given as beingequal to about half the breadth of the shell by Penard (1905). The diameter of the apertureand the pyriform rather than cylindrical shape of the body are considered sufficient todifferentiate the present specimens and those of Penard (1902, 1905) from D. lemanii Blanc,1892. This therefore leaves Penard's designated name of D. viscidula as valid and thespecimens described here are so named. DIFFLUGIA IN BRITAIN 27 Fig. 17 Difflugia viscidula: a, lateral view x410; b, detail of aperture which is blocked by a cystplug x 1000; c, detail of organic cement x 13 000. Pointed species or those with protruberances Difflugia amphoralis Cash & Hopkinson, 1 909 DESCRIPTION. The shell is transparent, squat pyriform with the aboral extremity taperingevenly to a point (Figs. 19a & b). It is composed mainly of medium pieces of quartz withsmall pieces mixed and so arranged to produce an intermediate smooth surface. Organiccement is seen in small patches between particles (Fig. 19d) and appears as rings fused to linearis I \gassowskii/ \ bryophila I V petricola Fig. 18 Diagrams of pyriform and elongate species to illustrate the basic, outline based on measurements given in Table 2. Table 2 Average dimensions of pyriform specimens listed inTable 1 and illustrated in Fig. 18. DIFFLUGIA IN BRITAIN 29 Fig. 19 Difflugia amphoralis: a, lateral view x850; b, lateral view to show even aboraltapering x 500; c, apertural view x 760; d, shell surface showing areas of organic cementx 4500; e, detail of organic cement network x 24 000. 30 C. G. OGDEN form a network, each mesh of which is about 380-450 nm internal diameter and the raisedwalls are 130 nm thick, a smooth membrane covers each enclosure (Fig. 19c). The circularaperture is surrounded by a small rim of mainly small particles to give a poorly definedborder (Figs. 19a,b&c). MEASUREMENTS (in urn). A single specimen: body length 109, breadth 62, diameter ofaperture 28. MATERIAL EXAMINED. The specimen was found in Sphagnum moss gathered at HolmsleyLodge, Burley, New Forest, Hampshire in March, 1980. GEOGRAPHICAL DISTRIBUTION. British Isles (Cash & Hopkinson, 1909), Tashkent(Pashintowa, 1929). REMARKS. The structure of the shell in the present specimen differs from the original (Cash &Hopkinson, 1909) by being composed mainly of quartz particles, rather than 'amorphous(?siliceous) scales'. However, it should be noted that the specimens described by Leidy (1879)and quoted as synonyms of this species by Cash & Hopkinson (1909) are also composedmainly of 'quartz sand'. Difflugia bicruris Gauthier-Lievre & Thomas, 1958 DESCRIPTION. The shell is elongate ovoid, the sides being almost parallel with a slighttapering towards the aperture and the aboral extremity, the latter is rounded and has twosmall, equally spaced, protruberances or horns (Figs. 20a & c). It is composed of medium tolarge pieces of angular quartz, with some small particles being used in shaping the horns (Fig.20d). The surface is rough but patches of organic cement are seen to form part of the shellmatrix (Fig. 20e). Organic cement is arranged in the form of a regular network whose meshhas a diameter of about 300-350 nm and the distance between each enclosure is about300 nm (Fig. 20f). The aperture is circular and surrounded by an even arrangement of smallparticles (Fig. 20b). MEASUREMENTS (in um). Two specimens: body length 202-207, breadth 95-1 1 5, diameter ofaperture 4 1-5 8. MATERIAL EXAMINED. Specimens were collected from aquatic plants taken at the banks of theRiver Brett, near Hadleigh, Suffolk, in August, 1979. GEOGRAPHICAL DISTRIBUTION Ivory Coast (Gauthier-Lievre & Thomas, 1958), Poland(Golemansky, 1970). REMARKS. Slight differences exist between the present specimens and those described byGauthier-Lievre & Thomas (1958). They have slightly larger general body measurements,although the proportions are directly comparable, and the horns are reduced in length,20 urn here compared with 30-33 um in the African specimens. Nevertheless, thedescriptions are in good agreement showing that D. bicruris is distinctly ovoid with twoaboral spines or horns. Difflugia distenda nom. nov.Difflugia acuminata var. inflata Penard, 1899 DESCRIPTION. The shell is transparent, pyriform with the aboral extremity acutely curvedtowards a small central tubular horn (Fig. 2 la), although the extent of the angle may be lessacute in a few specimens. It has an intermediate smooth surface and thickness, beingcomposed mainly of small to medium pieces of quartz, with occasional diatom frustulesadded. Areas of organic cement are sometimes seen in the shell matrix as a network (Fig.21c), with a mesh 350-400 nm in diameter and walls 150-200 nm thick (Fig. 21d). The DIFFLUGIA IN BRITAIN Fig. 20 Difflugia bicruris: a, lateral view illustrating the two aboral horns x 420; b, aperturalview x400; c, alternative lateral view with aboral horns not easily seen x430; d, detail of aboralhorn x2500; e, portion of shell surface showing distribution of organic cement x2700; f, detailof organic cement network x 25 000. C. G. OGDEN Fig. 21 Difflugia distenda: a, lateral view to show small aboral horn x430; b, aperturalview x350; c, shell surface showing large areas of organic cement x2400; d, detail of organiccement network x 1 1 000. aperture is circular and usually surrounded by an even arrangement of small particles (Fig.21b). MEASUREMENTS (in urn). Based on ten specimens: body length 217-270, breadth 109-135,diameter of aperture, 58-64; B/L 0-53 0-04, d/L 0-26 0-02. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire in March, 1980. DIFFLUGIA IN BRITAIN 33 GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Austria(Laminger, 1971, 19736), Belgium (Chardez, 196 la), British Isles (Cash & Hopkinson,1909), Congo (Chardez, 1964; Gauthier-Lievre & Thomas, 1958), France (Deflandre, 19626;Thomas & Mabille, 1956), Gabon (Gauthier-Lievre & Thomas, 1958), Ivory Coast(Gauthier-Lievre & Thomas, 1958), Mexico (Laminger, 19736), Morocco (Gauthier-Lievre& Thomas, 1958), Poland (Golemansky, 1970; Moraczewski, 1965), Switzerland (Penard,1902), Tashkent (Pashintowa, 1967). REMARKS. The variations of D. acuminata, and its twelve varieties has recently beenillustrated by Chardez (1961). More recently it has been shown (Ogden, 1979) that D.acuminata Ehrenberg, 1838 has a distinctive elongate shape and an apparently uniqueorganic cement pattern. Specimens described as var. inflata differ from this species in shape,size and organic cement pattern, the former is clearly demonstrated when the measurementsare expressed as ratios, the comparable ratios for D. acuminata are B/L 0-36 0-03 and d/L0-1 5 0-03 (Ogden, 1979). The specimens described here agree well with Penard's (1899, 1902) original descriptions,and the differences from D. acuminata are considered sufficient to warrant specificdesignation. Our normal practise is to raise the variety name to specific rank, but on thisoccasion it is impracticable as the terminology inflata has been used several times inconnection with specimens of Difflugia, for example D. curvicaulis var. inflata Decloitre,1951. ETYMOLOGY The specific name has been selected to reflect the inflated condition of the shell(L. distenda = swell out). Difflugia labiosa Wailes, 1919Difflugia amphora Leidy 1879 of Penard, 1901; 1902 & 1905 DESCRIPTION. The shell is opaque or dark brown, ovid or elongate ovoid, tapering from themid-body region sharply to the rounded or pointed aboral extremity and more graduallytowards the aperture (Fig. 22a). It is relatively thick and composed of quartz pieces soarranged that small particles fill the interstices between the larger pieces which form amainly smooth surface (Fig. 22d). Organic cement is seen infrequently but is in the form of anetwork, each mesh being about 250-350 nm in diameter and is usually covered by a smoothmembrane (Fig. 22e). The aperture is roughly circular but is sinuous with as many as six orseven undulations or lobes (Fig. 22b). It is slightly recessed into the main body, as though itwas surrounded by a groove, the margin or lips, are thin and bordered by tiny particles ofquartz. (Fig. 22b & c). MEASUREMENTS (in um). Based on five specimens: body length 150-21 1, breadth 112-158,diameter of aperture 50-63. MATERIAL EXAMINED. Specimens were collected from aquatic plants taken at the banks of theRiver Brett, near Hadleigh, Suffolk in August, 1979. GEOGRAPHICAL DISTRIBUTION. Austria (Laminger, 1971, 1975), Belgium (Chardez, 1980),British Isles (Cash et al, 1919; Ogden & Hedley, 1980), Czechoslovakia (Ertl, 1965;Stepanek, 1952, 1967), France (Thomas, 1954; Thomas & Mabille, 1956), Germany(Grospietsch, 1958; Schonborn, 19620, & 6), Netherlands (Hoogenraad & Groot, 1940),Poland (Moraczewski, 1961, 1965), Russia (Kourov, 1925), South Africa (Oye, 1931),Switzerland (Penard, 1902), Venezuela (Deflandre 1926a; Grospietsch, 1975). REMARKS. There has been some confusion over the correct name for this species, most recentauthors using D. amphora Leidy, 1879, basing their identifications on the description givenby Penard (1902). However, this name is preoccupied as it was used by Ehrenberg (1854,1872) to describe a specimen which is no longer considered to be a species of Difflugia. It 34 C. G. OGDEN ? ** v fc :4-^"""X-. Fig. 22 Difflugia labiosa: a, latero-apertural view to show arrangement of particles x 580; b,apertural view illustrating the undulations or lobes x470; c, lateral view of aperture to showslight groove and small particles on margin x990; d, portion of shell surface showing the closepacking of particles x 1 300; e, detail of organic cement x 24 000. DIFFLUGIA IN BRITAIN 35 follows that the later reports by Leidy (1874, 1879) default for the same reason. The latterreport being more confusing by quoting D. amphora as a synonym of D. urceolata, whilstgiving a figure of D. ureceolata var. amphora. Wailes (in Cash et ai, 1919) noted these earlierreports and proposed the new name D. labiosa, giving Penard's (1901 , 1902, 1905) reports assynonyms. Earlier, Cash & Hopkinson (1909) had suggested that part of Leidy 's (1879)description of/), urceolata var. amphora was a synonym of their new species D. amphoralis.The specimens described here are in good agreement with Penard (1902) who gave a range ofbody lengths 150-270 urn, but usually about 200-210 urn, and Wailes who suggested that itwas a rare species from his single 265 um long specimen. This species is distinct in its ovoid-conical shape plus the lobed aperture with distinctmargin and recessed base. Difftugia mamillaris Penard, 1 893 DESCRIPTION. The shell is colourless or hyaline, ovoid elongate, swollen or arched in themid-region but tapering at both extremities, to give a rounded protruberance aborally andgradually near the aperture to give a slightly pronounced neck (Fig. 23a). Irregularities ingeneral shape are not uncommon, for example one specimen although tapered did not havean aboral protruberance, whilst another (Fig. 23c) tapered markedly from the mid-bodyregion. It is composed mainly of small to medium pieces of quartz so arranged that the largerparticles tend to be in the mid-body region whilst the extremities have the smaller particles,overall it usually produces an intermediate thickness of a single layer and a relatively smoothoutline. Organic cement occurs in small patches as a network, which has walls about80-1 30 um thick between each mesh but sometimes 230 um thick at junctions. Each mesh isabout 250nm in diameter and is covered by a smooth membrane which is distinct in havingthree or four small white spots on the surface of each enclosure (Figs. 23d & e). The apertureis circular, composed of small particles, and roughly finished so that the margin appearsuneven or serrated (Fig. 23b). MEASUREMENTS (in um). Based on twenty-three specimens: body length 93-111, breadth54-72, diameter of aperture 23-3 1 . MATERIAL EXAMINED. Specimens were collected from a sample of aquatic plants taken at thebanks of the River Brett, near Hadleigh, Suffolk in August, 1979. GEOGRAPHICAL DISTRIBUTION. Austria (Laminger, 1975), Congo (Chardez, 1964),Czechoslovakia (St6panek, 1967), Germany (Grospietsch, 1957), Switzerland, (Penard,1901), Venezuela (Grospietsch, 1975). REMARKS. This species was initially described by Penard (1893) and redescribed in moredetail by the same author (Penard, 1 902). The present specimens are in good agreement withthe latter description, including measurements which gave the body length as being generallybetween 90-110 um with a few large specimens up to 130 um long. More recently,Grospietch (1957) has shown a similar regularity in size of specimens from Lake Maggiore,giving length 90-1 13 um and breadth 48-63 urn. Penard (1902) suggested that this specieswas rare in Swiss Lakes, and this appears to be the case in other localities judging by itsreported incidence. Nevertheless, if the two recent reports (Grospietch, 1957 and thepresent) are used as indicators, it would appear that when present this species is usuallyabundant. D. mamillaris is distinct in outline, even though the thin structure may be subject todistortion, and the unusual organic cement pattern. Difftugia microdaviformis (Kourov, 1925) comb. nov.Difflugia oblonga var. microdaviformis Kourov, 1925DESCRIPTION. The shell is brown, pyriform with a distinct aboral protruberance (Fig. 24a). It 36 C. G. OGDEN w^S, * a Fig. 23 Difflugia mamillaris: a, lateral view x 1 100; b, apertural view x 790; c, lateral view ofirregular shaped shell x 440; d and e, detail of organic cement network, note the small whitespots in each enclosure which is a regular feature x 24 000. DIFFLUGIA IN BRITAIN Fig. 24 Difflugia microclaviformis: a, lateral view x 230, b, apertural view x 530; c, shell surfaceillustrating the arrangement of particles and organic cement x2600; d, detail of organic cementnetwork, note the regular distribution of small pores in each enclosure x 33 000. is composed of small to medium pieces of quartz, some flattish diatom frustules, and anetwork of organic cement is often seen as part of the shell matrix (Fig. 24c). The result is asmooth surface and a well defined outline. The mesh of the organic cement has a diameter of300 nm with walls 150nm thick, and a smaller network, with pores about 30 nm indiameter, covers each mesh enclosure (Fig. 24d). The aperture is circular and surrounded bysmall particles (Fig. 24b). 38 C. G. OGDEN MEASUREMENTS (in um). Based on two specimens: body length 202-206, breadth 88-89,diameter of aperture 27-28. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire, in March, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina(Vucetich, 1978), Austria (Laminger, 19736), Congo (Gauthier-Lievre & Thomas, 1958),France (Thomas, 1954), Ivory Coast (Gauthier-Lievre & Thomas, 1958), Mexico (Laminger,1973a), Roumania (Godeanu et al, 1973), Russia (Kourov, 1925), Spain (Gracia, 1972a). REMARKS. This species when initially described (Kourov, 1925) had a body length notgreater than 185 urn, whereas Gauthier-Lievre & Thomas, 1958 gave dimensions similar tothose given here. It is redescribed to show its marked similarity to D. claviformis, lack ofmaterial prevents a complete comparison and the identification must be treated as tentative.D. microclaviformis appears to differ from D. claviformis (see Ogden, 1979) in size andpossibly the organic cement pattern. Difflugia molesta Penard, 1902 DESCRIPTION. The shell is brown, ovoid or ovoid-elongate, sometimes with a small aperturalcollar and arched aborally (Fig. 25a). It is composed of a mixture of quartz particles anddiatom frustules, the former usually being predominant. Organic cement is seen as a networkbetween particles, either as part of the shell matrix or occasionally at junctions. The ringsthat form the network are about 650-750 nm internal diameter with the dividing wall being100-1 50 nm thick, a second thin, inner wall lies close to the main wall and in the illustratedspecimen the covering membrane is either holed or figured (Fig. 25c). The aperture isroughly circular, irregular in outline, and mainly surrounded by small particles (Fig. 25b). MEASUREMENTS (in um). Based on four specimens: body length 106-114, breadth 61-87,diameter of aperture 28-43. MATERIAL EXAMINED. Specimens were collected from Sphagnum moss gathered at HolmsleyLodge, Burley, New Forest, Hampshire in March, 1980 and Mynnd Hiraethog, Denbigh,Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Roumania (Godeanu et al, 1973), Russia (Kourov, 1925),Switzerland (Penard, 1902). REMARKS. These specimens are tentatively identified here as D. molesta, the query arisingdue to their similarity with D. amphoralis. Of the four specimens examined, the smallest(Fig. 25a) bears the closest resemblance being slim, with a small aperture and differing in theabsence of a pointed aboral protruberance. The three broader specimens have a wideraperture, a narrow neck and are arched aborally. In addition, there are differences in theorganic cement patterns between the two species, but altogether there is insufficientinformation to satisfactorily resolve the problem. Difflugia tricornis (Jung, 1936) comb. nov.Difflugia elegans forma tricornis Jung, 1936 DESCRIPTION. The shell is pyriform, with a slight broadening near the aperture, and ischaracterised by three, evenly spaced, aboral spines (Fig. 26a). The surface is rough andmainly composed of a mixture of medium and large pieces of angular quartz, the spines havemedium particles at their wide bases but small particles are used progressively as they taperto a point. Small areas of organic cement are seen between the particles (Fig. 26c), usually inthe form of a regular network (Fig. 26d), having a mesh between 350^00 nm internal DIFFLUGIA IN BRITAIN 39 Fig. 25 Difflugia molesta: a, lateral view x910; b, apertural view x740; c, detail of organiccement to show arrangement of rings and figured centre x 24 000. diameter and dividing walls about 1 50 nm thick although at some junctions there are largerareas. The aperture is circular and surrounded by an irregular assortment of particles (Fig.26b). MEASUREMENTS (in um). One specimen: body length 1 16, breadth 82, diameter of aperture40. 40 C. G. OGDEN Fig. 26 Dijjlugia tricornis: a, lateral view to show the three equally spaced aboral spines x 690; b,apertural view x 530; c, portion of shell surface showing small areas of organic cement x2300;d, detail of organic cement network x 14 000. MATERIAL EXAMINED. The specimen was collected from aquatic plants taken at the banks ofthe River Brett, near Hadleigh, Suffolk in August, 1979. GEOGRAPHICAL DISTRIBUTION. Germany (Jung, 1936), Sudan (Gauthier-Lievre & Thomas,1958). REMARKS. In the earlier descriptions (Jung, 1936; Gauthier-Lievre & Thomas, 1958) thespecimens resembled D. elegans except for the three aboral spines which were stated to berandomly placed, although Jung (1936) cited Penard's figures referring to much smalleranimals. Differences in the shape and structure of D. elegans examined recently by the DIFFLUGI A IN BRITAIN 41 author (Ogden, 1979; Ogden & Hedley, 1980) have been confined to the size and structure ofthe single aboral spine or horn, whilst the body length was usually of a standard size110-160 jim. The present specimen has three, equally spaced, aboral spines and the body breadth ismarkedly wider, features that are considered sufficiently different from D. elegans to warranta specific designation. The earlier reports are placed, with reservations, in synonymy. Difflugia ventricosa Deflandre, 1926 DESCRIPTION. The shell is colourless, elongate, with a slight swelling in the aboral half of thebody which then tapers to a sharp point (Fig. 27a). It is composed of a mixture of quartz,diatom frustules and flagellate cysts to give a thin, irregular surface. Organic cement isfrequently seen between particles in the form of a network (Fig. 27c), made of rings about600-680 nm in diameter with walls 1 80-250 nm thick (Fig. 27d). In some instances the ringsare fused and appear to have lost or merged their walls (Fig. 27e). The aperture is circularand usually surrounded by small particles (Fig. 27b). MEASUREMENTS (in urn). Two specimens: body length 177-199, breadth 64-66, diameter ofaperture 30-3 1 . MATERIAL EXAMINED. The specimens were collected from a sample of Sphagnum mossgathered at Holmsley Lodge, Burley, New Forest, Hampshire in March, 1 980. GEOGRAPHICAL DISTRIBUTION. Belgium (Chardez, 1973), Congo (Gauthier-Lievre &Thomas, 1958), France (Thomas & Mabille, 1956), Ivory Coast (Gauthier-Lievre &Thomas, 1958), Venezuela (Deflandre, 1926a). REMARKS. A redescription of D. ventricosa has recently been given by Chardez (1973) whocompared it with other species having a pointed aboral extremity. It is interesting to notethat he made no comparison with D. venusta, although from the description given here(below) there would seem to be some similarities. The tabulated measurements given byChardez (1973) from earlier descriptions, are in good agreement with the exception of thosegiven by Thomas & Mabille (1956). This species is distinct in having a thin, elongate outline which is sharply pointed aborally. Difflugia venusta (Penard, 1902) comb. nov. Difflugia pyriformis var. venusta Penard, 1902 Difflugia oblonga var. venusta (Penard, 1902) Cash & Hopkinson, 1909 DESCRIPTION. The shell is pale yellow or hyaline, cylindrical, gradually swelling from theaperture for about two-thirds of the body length to the broadest diameter and then taperingsharply in the last third to the bluntly pointed apex (Fig. 28a & b). It is composed mainly ofsmall to medium pieces of quartz and diatom frustules arranged to give a relatively regular,intermediate smooth, outline apart from the occasional addition of a larger angular piece ofquartz or diatom frustule. Small areas of organic cement are sometimes visible as a thickwalled network with a covered mesh (Fig. 28d), but more often as thick walled rings about450-600 nm in diameter and walls 150-220 nm (Fig. 28c). The aperture is usually circularand surrounded by small particles that give it an irregular margin (Fig. 28c). MEASUREMENTS (in um). Based on three specimens: body length 174-188, breadth 68-76, diameter of aperture 30-32. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gathered at Holmsley Lodge, Burley, New Forest, Hampshire in March, 1980. GEOGRAPHICAL DISTRIBUTION. Argentina (Dioni, 1970), Belgium (Chardez & Caspar, 1976), 42 C. G. OGDEN . a Fig. 27 Difflugia ventricosa: a, lateral view to illustrate tapering of aboral spine x 520; b,apertural view x 760; c, shell surface showing small areas of organic cement x4700; d, detail oforganic cement network of rings x 24 000; e, organic cement network with fused rings x 28 000. DIFFLUGIA IN BRITAIN 43 Fig. 28 Difflugia venusta: a and b, lateral views to show sharp tapering of aboral region x570and x340; c, apertural view x590; d, shell surface showing distribution of organiccement x 9200; e, detail of organic cement network x 24 000. 44 C. G. OGDEN mamillaris / \ amphoralis Fig. 29 Diagrams of pointed species or those with protruberances to illustrate the basic outline, based on measurements given in Table 3. DIFFLUGIA IN BRITAIN 45 Table 3 Average dimensions of pointed species or thosewith protruberances used to give basic outlines illustratedin Fig. 29 France (Thomas, 1954), Haute Volta, W. Africa (Gauthier-Lievre & Thomas, 1958),Switzerland (Penard, 1902). REMARKS. The present specimens agree well with the descriptions given by Penard (1902),Cash & Hopkinson (1909) and Gauthier-Lievre & Thomas (1958) who considered it avariety of D. pyriformis/oblonga, although the shells described by Cash & Hopkinson (1909)were slimmer than those reported here and by Gauthier-Lievre & Thomas (1958). This species is distinct in the graceful outline with bluntly pointed aboral extremity andshell structure. Ovoid or spherical species Difflugia ampullula Playfair, 1918 DESCRIPTION. The shell is hyaline, ovoid and circular in cross section (Fig. 30a). It has amedium thickness and is composed mainly of small to medium pieces of quartz, arranged togive a clean outline with a smooth surface. Small areas of organic cement in the form of anetwork, are often seen as part of the shell structure (Fig. 30d). The mesh of the network issmall about 300-350 nm in diameter with thin walls 50-100 nm thick, and an even smallerdistinctive network covering each mesh enclosure (Figs. 30e & f)- The aperture is circular,surrounded by a slightly raised collar of small particles, and the edge of the collar is oftenirregular (Figs. 30b & c). MEASUREMENTS (in urn). Range of 39 specimens: body length 54-95, breadth 35-72,diameter of aperture 16-29. MATERIAL EXAMINED. Specimens were collected from aquatic plants taken at the banks of theRiver Brett, near Hadleigh, Suffolk in August, 1979. GEOGRAPHICAL DISTRIBUTION. Australia (Playfair, 1918). REMARKS. The specimens described here differ slightly from the original description(Playfair, 1918) in the absence of a 'minute, pointed apiculate process' on the aboralextremity. However, this process was reported as being present sometimes, and as there isotherwise good agreement between the two reports, the specimens are designated as D.ampullula. 46 C. G. OGDEN Fig. 30 Difflugia ampullula: a, lateral view x 1000; b, latero-apertural view of aperture toillustrate the small collar x 1200; c, apertural view x 770; d, portion of shell surface to show thearrangement of particles and organic cement x5000; e, detail of shell surface x7700; f, smallarea of organic cement network, note that each enclosure has an inner network x 25 000. DIFFLUGIA IN BRITAIN 47 Dijflugia angulostoma Gauthier-Lievre & Thomas, 1958 DESCRIPTION. The shell is transparent, spherical and composed mainly of diatom frustules(Fig. 3 la & b). The particles are packed close together with many overlapping, to give arough surface. Organic cement is seen infrequently as small strands between particles (Fig.3 Ic). The aperture is usually circular (Fig. 3 la), but it may have irregularities depending onthe arrangement of surrounding diatom frustules. Two groups of specimens with identical shell features but differing dimensions wereexamined, the 'a' specimens are from all four listed localities and 'b' specimens are from Three Shires Stone only. * MEASUREMENTS (in urn) body diameter of length breadth aperture B/L d/L 8 specimens 'a' 40-56 40-48 18-23 0-94 0-06 0-42 0-06 8 specimens 'b' 60-82 50-73 28-51 0-870-09 0-560-05 MATERIAL EXAMINED. Specimens were collected from samples of Sphagnum moss gatheredat four localities: Cranes Moor in May, 1977, Holmsley in May, 1978 both in the NewForest, Hampshire; Three Shires Stone, Wrynose Pass and Lanthwaite, both in Cumbria,June, 1979. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina(Vucetich, \913a & b), Congo (Gauthier-Lievre & Thomas, 1958). REMARKS. This species was described by Gauthier-Lievre & Thomas (1958), from specimensfound in Algeria, who considered that it differed from D. minuta by the large size of theaperture, quoted as being about one-third of the breadth in diameter, and the coveringdiatoms. Both sets of the present specimens, 'a' and 'b', agree well with this description,having a large aperture about half the breadth diameter and are composed of diatoms. The'b' specimens share almost similar dimensions to D. angulostoma, the latter having a bodylength 60-95 um and aperture 30-45 um, whereas the 'a' specimens are generally smaller.The difference in size between specimens 'a' and 'b' is more apparent when the ratios d/L arecompared, such differences are usually significant. However, lack of similar data from theoriginal description of D. angulostoma does not allow a comparison to be made. In theabsence of this information and the otherwise similarity of the two groups of specimens, theyare both designated as D. angulostoma. Dijflugia decloitrei Godeanu, 1972 Difflugia levanderi Playfair, 1918 (in part)Difflugia acuminata Levander, 1 894 (in part) DESCRIPTION. The shell is transparent, ovoid, tapering evenly from the mid-body positiontowards the aperture and aboral extremity, there is the suggestion of a collar near theaperture due to the tapering ending prior to the apertural opening (Fig. 32a). In somespecimens there is an apparent lateral compression, but it is usually slight and probablyrelated to the fragility of the structure. It has a well defined outline, and the arrangement offlattish pieces of quartz give it a smooth surface. A network of organic cement is seen at mostjunctions of these particles (Fig. 32c). The mesh is about 280-350 nm in diameter and thewalls 350 nm thick (Fig. 32d). The aperture is circular with often a rugged outline due to theplacement of the flattish particles (Fig. 32b). MEASUREMENTS (in |im). Based on ten specimens: body length 77-95, breadth 39-55,diameter of aperture 20-27. C. G. OGDEN Fig. 31 DiJJlugia angulostoma: a, apertural view x 1600; b, lateral view, note the coveringdiatom frustules x 1000; c, shell surface with strands of organic cement x 8700. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Myndd Hiraethog, Denbigh, Clywdd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Australia (Playfair, 1918), Germany (Levander, 1894),Roumania (Godeanu, 1972). REMARKS. Levander (1894) described four different forms of D. acuminata, one ofwhich-type 'b'-was considered by Playfair (1918) to represent a distinct species., D.ievanderi. Both of these authors suggested that there were two sizes of these specimens,'110x60 and 70x40um', the larger being rough and 'stony' whilst the smaller werechitinous with small, scattered granules. Recent descriptions of D. Ievanderi, for example DIFFLUGIA IN BRITAIN 49 Fig. 32 Difflugia decloitrei: a, lateral view x 1 300; b, apertural view x 900; c, portion of shell toshow arrangement of flat particles to give a smooth surface x2400; d, detail of organic cementnetwork x 22 000. that given by Gauthier-Lievre & Thomas (1958) stated a range of body length of 85-140 ^msuggesting that these are the larger specimens of the earlier authors. The specimens describedby Godeanu (1972) as D. decloitrei appear to be similar to the group of smaller specimens,having a similar structure and large aperture. Those described here are in good agreementwith this latter description and share similar measurements; body length 62-86 breadth40-56 and diameter of aperture 20-23 (Godeanu, 1972). In the absence of larger shells forcomparison the present specimens are referred to D. decloitrei. 50 C. G. OGDEN Difflugia gramen Penard, 1 902 This species has recently been redescribed (Ogden, 1980), but is included here because thenumbers examined allow a comparison of dimensions between specimens from differenthabitats and localities. Both samples were collected in August, 1979, from sites which areabout fifty miles apart. Specimens 'A' were selected from a sample of algae and water plantsin stationary water (see Ogden, 1980), specimens 'B' from aquatic plants at the banks of theRiver Brett, near Hadleigh Suffolk, which in summer is a slow moving, small watercourse. MEASUREMENTS (in urn). 'A' thirty-five specimens; 'B' forty-four specimens. body diameter of length breadth aperture B/L d/L d/B 'A' 89-117 70-112 23-39 0-960-07 0-340-04 0-36 + 0-03 'B' 61-97 42-75 18-33 0-770-07 0-320-04 0-43 0-04 avg. 4 A' 98-8 94-3 33-6 avg. 'B' 78-7 60-5 26-2 REMARKS. In shell construction the 'A' specimens are larger, spherical and more regular,only one aperture not appearing typically trilobed. Whilst, 'B' specimens are ovoid and nine(about 20%) had four lobes or were irregular in outline. It is interesting to note that thecommon feature between these specimens is the ratio of the aperture to the body length. Difflugia masaruzzi Oye, 1958 DESCRIPTION. The shell is transparent, ovoid and composed of a mixture of flattish siliceousparticles including some diatom frustules, to give a fragile structure with an irregular surfaceand outline (Fig. 33a). Organic cement is seen at some junctions (Fig. 33c), but there is adegree of overlapping with most particles. It appears as a network having a mesh of about450-600 nm in diameter with walls 200 nm thick, each enclosure having a smaller networkwith a mesh about 90 nm in diameter (Fig. 33d). The aperture is circular, wide, usually withan irregular margin (Fig. 33b). MEASUREMENTS, (in um). Two specimens: body length 66-67, breadth 39^43, diameter ofaperture 25. MATERIAL EXAMINED. Specimens were collected from a sample of aquatic plants taken fromthe edge of a pond at Burley, New Forest, Hampshire in March, 1980. GEOGRAPHICAL DISTRIBUTION. Congo (Oye, 1958), 'Costa Rica (Laminger, 1973a), Mexico(Laminger, 1973a). REMARKS. In the initial description of D. mazaruzii it was stated by Oye (1958) to be similarto D. rubescens and D. lucida, although it only shares a transparent shell with these twospecies. The present specimens agree well with his description of a shell having some largedistinct particles attached, and an aperture devoid of a regular margin more or less wavybecause parts of the shell extend to the edge. Two specimens formed the basis for the earlierreport and were somewhat larger, 72 and 78 |im in body length, 44 and 55 um in breadth, 28and 30 nm diameter of aperture. Although this species is similar in size to D. glans Penard,1902 (see p. 7), it is distinct in having a transparent, fragile shell, wide aperture and apatterned organic cement. Difflugia mica Frenzel, 1892DESCRIPTION. The shell is brown, spherical or ovoid with a shallow apertural collar (Fig. DIFFLUGIA IN BRITAIN 51 Fig. 33 Difjlugia masaruzii: a, lateral view x 1400; b, apertural view x 1 100; c, shell surfaceillustrating the distribution of organic cement x4200; d, detail of organic cement networkx 26 000. 34a). It is composed of small flattish pieces of quartz (Fig. 34d), packed tightly together toform a strong structure with a smooth surface and positive outline. Only small strands oforganic cement are visible between the particles (Fig. 34e). The aperture is circular andusually well defined by the collar, which has a thin, even layer of organic cement around it 52 C. G. OGDEN Fig. 34 Difjlugia mica: a, latero-apertural view showing the shallow collar x 1600; b, aperturalview, note that the aperture is blocked by a cyst plugx 1 100; c, portion of apertural collar toillustrate the organic cement covering x3900; d, shell surface with close packing ofparticles x 3800; e, detail of organic cement x 24 000. DIFFLUGIA IN BRITAIN 53 (Figs. 34b & c). The illustrated specimen has a broken cyst membrane, made mainly oforganic cement, just inside the apertural opening. MEASUREMENTS, (in um). Based on seven specimens: body length 44-58, breadth 36^9,diameter of aperture 12-1 8; B/L 0-81 0-10, d/L 0-31 0-05. MATERIAL EXAMINED. Specimens were collected from a sample of aquatic plants taken at thebanks of the River Brett, near Hadleigh, Suffolk, in August, 1979. GEOGRAPHICAL DISTRIBUTION. Argentina (Frenzel, 1892), Germany (Schonborn, 1962a & b,1965), Poland (Moraczewski, 1961, 1965); Roumania (Godeanu et al., 1973), Switzerland(Penard, 1902). REMARKS. This species was initially described as Difflugia sp. by Frenzel (1892), the specificname being added as a footnote (p. 135). In redescribing the species Penard (1902) used thename D. mica? Frenzel, the query has been dropped by subsequent authors and the nameconsidered to be valid. Difflugia microstoma (Thomas, 1954) comb. nov.Difflugia globularis var. microstoma Thomas, 1954 DESCRIPTION. The shell is ovoid or subspherical, composed mainly of a mixture of small tomedium pieces of flattish quartz and diatom frustules. The particles are arranged to give arelatively smooth outline (Fig. 35a), with the diatom frustules being in general additions tothe main structure (Fig. 35d). The close packing of materials is such that only small strandsof organic cement are seen (Fig. 35c). The aperture is circular and usually surrounded by aborder of small particles (Figs. 35b & e). MEASUREMENTS (in um). Based on fifteen specimens: body length 76-105, breadth 63-83,diameter of aperture 1 8-29; B/L 0-79 + 0-07, d/L 0-26 0-03. MATERIAL EXAMINED. Specimens were collected from Sphagnum moss gathered at HolmsleyLodge, Burley, New Forest, Hampshire in July, 1978; March, 1980 and at Myndd Hiraethog,Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), France (Thomas,1954). REMARKS. Thomas (1954) considered when describing the variety D. globularis var.microstoma that it was similar to specimens of/), globulosa illustrated by Penard (1902, p.258 Fig. 6), although he later (Gauthier-Lievre & Thomas, 1958) proposed both assynonyms of D. minuta Rampi, 1950. The examples of D. minuta described in this reportshow that D. microstoma is distinct in having a larger ovoid shell with a small aperture(compare ratios B/L and d/L, below), the latter feature also differentiates it from D. globulosaDujardin, 1837. Difflugia minuta Rampi, 1950 DESCRIPTION. The shell is ovoid or spherical, composed mainly of small pieces of flattishquartz and the occasional fragment or diatom frustule (Fig. 36a). The particles are packed soclosely, to give a robust structure, that organic cement is visible only as small strands (Fig.36c). The aperture is small and often surrounded by a narrow lip of organic cement (Fig.36d), the lip is not apparent in side view but makes the apertural opening distinct whenviewed en face (Figs. 36b & d). MEASUREMENTS, (in um). Based on six specimens: body length 44-53, breadth 34-48,diameter of aperture 9-12; B/L 0-98 + 0-08, d/L 0-25 0-04. 54 C. G. OGDEN Fig. 35 Difflugia microstoma: a, lateral view of shell with smooth surface x980; b, aperturalview x770; c, portion of shell surface with strands of organic cement x 13 000; d,lateral view of shell with added diatom frustules x 770; e, apertural view x 580. DIFFLUGIA IN BRITAIN 55 Fig. 36 DiJJlugia minuta: a, lateral view x 1400; b, apertural view x970; c, portion of shellsurface showing close packing of particles x7700; d, detail of aperture to show narrow lip oforganic cement x3700. MATERIAL EXAMINED. Specimens were collected from samples of Sphagnum moss gatheredat three locations, Cranes Moor, in May, 1977; Holmsley, in July, 1978, both in the NewForest, Hampshire; Myndd Hiraethog, North Wales in August, 1980; and aquatic plantstaken at the bank of a pond near Burley, New Forest in March, 1980. GEOGRAPHICAL DISTRIBUTION. Brazil (Green, 1975), Costa Rica (Laminger, 1973a),Germany (Schonborn, 1965), Italy (Rampi, 1950), Roumania (Godeanu / a/., 1973). REMARKS. The initial report (Rampi, 1950) of this species is brief, consisting of onefigure and a few lines of description. These note that it has a globular shell made mainly ofquartz particles and concludes that it differs from D. globulosa by its small size, length53 um, breadth 48 urn. Unfortunately no dimensions for the aperture are given. Thespecimens referred to this species by Gauthier-Lievre & Thomas (1958) are all much largerthan the measurements given by Rampi (1950), and are here considered to represent D.microstoma (see p. 53). D. minuta is considered a distinct species in having a circular shell composed mainly ofquartz, with a small aperture surrounded by a narrow lip or rim of organic cement. 56 C. G. OGDEN Fig. 37 Difflugia rotunda: a, apertural view x270; b, lateral view x240; c, apertural view ofspecimen made mainly of quartz particles, note the regular outline of the aperture x290; d,portion of shell surface of 'diatom' specimen x 3300; e, shell surface of specimen made mainly ofquartz x 2900. Difflugia rotunda nom. nov.Difflugia globularis var. sphaerica Chardez, 1956 DESCRIPTION. The shell is brownish, spherical or hemispherical, with the outline frequentlydistorted by the addition of large diatom frustules (Figs. 37a & b). The basic structure is made DIFFLUGIA IN BRITAIN 57 mainly of quartz (Fig. 37c), but diatom frustules or fragments of frustules, are often mixedwith this in different proportions (Fig. 37b). This material is usually packed tightly togetherso that only small strands of cement are seen (Figs. 37d & e). The aperture is circular,sometimes slightly irregular, but usually surrounded by a shallow rim of small particles (Figs.37a&c). MEASUREMENTS (in um). Based on sixteen specimens: body length 133-204, breadth138-193, diameter of aperture 79-1 13; B/L 0-98 + 0-12, d/LO-550-07. MATERIAL EXAMINED. Specimens were collected from samples of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire in May, 1978; March, 1979; 1980 andMyndd Hiraethog, Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Argentina (Vucetich, \913a & b), Belgium (Chardez, 1956). REMARKS. In the original description of this variety, D. globularis var. sphaerica, Chardez(1956) noted that it differed in both size and diameter of aperture, the latter feature beingabout half the breadth, from his concept of/), globularis. According to Cash & Hopkinson(1909) the name globularis was used in error by Wallich (1864) for D. globulosa Dujardin,1837. Nevertheless, these distinguishing features are used here to differentiate thesespecimens from other spherical species. Again a new name is proposed because the termsphaerica has been widely used for varieties in the terminology of this genus. ETYMOLOGY. The specific name has been chosen to reflect the shape of the shell (L.rotunda round circular or orbicular). Difflugia stoutii sp. nov. DESCRIPTION. The shell is ovoid or ovoid elongate, composed of mainly small, flattishparticles of siliceous material, including quartz, diatom frustules and shell plates fromsmaller testate amoebae (Figs. 38a & d). It is extremely fragile, several specimens havingcollapsed in preparation, and hence the apparent lateral flattening of the specimen shown inFig. 38c. Organic cement is seen only as small threads due to the regular overlapping of theshell components (Fig. 38e). The aperture is roughly circular, small and appears to berecessed, but this latter feature may be due to structural fragility (Figs. 38b & d). MEASUREMENTS (in um). Based on four specimens: body length 47-59, breadth 33-36,diameter of aperture 9-12. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Myndd Hiraethog, Denbigh, Clwyd, North Wales in August, 1979. REMARKS. The present specimens are similar to three species recently described fromGermany, namely D. stechtinensis Schonborn, 1962, D. sudiformis Schonborn, 1966 and D.szczepanskii Schonborn, 1965. They differ from the two former species in generaldimensions, D. stechtinensis is almost spherical, with an aperturual diameter equal to halfthe body breadth, whilst D. sudiformis is an elongate, very slender species. D. szczepanskii isa slightly larger species but differs mainly in having an aperture size two-thirds of the bodywidth. All three species are described as having a hyaline shell covered with a meagrescattering of particles. D. stoutii is distinct in having a fragile, elongate ovoid shell composed of flattish particlesand a small aperture. ETYMOLOGY. This species is named after the late Dr John Stout in recognition of hiscontributions to recent advances in protozoology. Difflugia urceolata Carter, 1864DESCRIPTION. The shell is opaque, ovoid or rotund, often having one or more irregular blunt 58 C. G. OGDEN Fig. 38 Difflugia stoutii: a, lateral view x 1 700; b, apertural view x 1 700; c, lateral view ot"specimen slightly compressed anteriorly x 1 100; d, latero-apertural view of ovoid specimen withslightly recessed aperture x 1 300; e, shell surface, note the overlapping of particles x 5800. DIFFLUGIA IN BRITAIN 59 aboral protruberances, and a pronounced apical rim or collar (Fig. 39a). The rim has arecurved appearance the edge of which is usually well denned (Figs. 39b & c), with anabundance of organic cement apparent as part of the rim matrix (Fig. 39e). The body iscomposed of small to medium particles of quartz, blended together so that the smallerparticles and organic cement fill the gaps between the larger particles and give a relativelysmooth surface. Diatom frustules or parts of them are occasionally included in the structure.Organic cement in the form of a network is seen as part of the matrix (Fig. 39d), the mesh hasa diameter of about 240-290 nm with walls 100-180 nm thick (Fig. 39f). The aperture isusually circular (Fig. 39a). Variation in this species is not uncommon. Although usually limited to the presence orabsence of aboral protruberances, an occasional deformed shell may be seen. The specimenillustrated here (Figs. 40a & b) has a depressed apical rim, malformed body and definedaboral protruberances are absent. MEASUREMENTS (in jim). Based on twenty-one specimens: body length 204-398, breadth193-^26, diameter of aperture 87-198; B/LO-920-10,d/L 0-44 + 0-06. MATERIAL EXAMINED. Specimens were collected from samples of Sphagnum moss gatheredat Holmsley Lodge, Burley, New Forest, Hampshire on several occasions, May, 1977; 1978and March, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina(Boltovskoy & Lena, 1971, 1974; Dioni, 1970; Vucetich, 19730 & b), Australia (Playfair,1918), Austria (Laminger, 1972c, 19736), Belgium (Chardez, 1960, 19616; Chardez &Gaspar, 1976), Brazil (Green, 1975), British Isles (Cash & Hopkinson, 1909; Ogden &Hedley, 1980), Chile (Decloitre, 1954), China (Decloitre, 1954), Congo (Chardez, 1964;Stepanek, 1963), Czechoslovakia (Stepanek, 1952), France (Deflandre 1962b; Thomas,1954), Germany (Jung, 1936), Hungary (Bereczky, 1973), Italy (Grandori & Grandori,1934), Java (Bartos, 1963a), Netherlands (Hoogenraad & Groot, 1940), Poland(Moraczewski, 1965), Russia (Kourov, 1925), Spain (Margalef, 1955), Sudan (Gauthier-Lievre & Thomas, 1 958), Switzerland (Penard, 1 902), United States of America (Laminger etal, 1979), Venezuela (Deflandre, 19260;Grospietsch, 1975). REMARKS. This is one of the most widely reported specimens ofDifflugia, probably due to itssize and distinctive shape. However, variation in shell construction has led to the descriptionof several varieties. Thomas (1954) used the presence of aboral protruberances todifferentiate the variety olla Leidy, 1879; whilst Gauthier-Lievre & Thomas (1958) list fourwhich differed in rim construction, namely lageniformis (Wallich), lageniformis formaminor forma nov., minor Deflandre and sphaerica Playfair; and more recently descriptionsof specimens which differed in shape and material have produced two more-var. chayuensisWang Jiagi, 1977 and forma subureceola Chardez & Gaspar, 1976. These reports of natural variation can have little value until they are thoroughlyinvestigated, and the present specimens are therefore referred to D. urceolata. Compressed species Difftugia himethogii sp. nov. DESCRIPTION. The shell is light yellow or transparent, thin pyriform with a distinct neck orcollar which often has parallel sides (Figs. 4 la & b). The neck region is made of angularquartz and usually has a rough appearance (Fig. 41b), whilst the remainder of the body iscomposed of small to medium pieces of flattened quartz and has a smooth appearance.Organic cement is frequently seen in small areas as part of the shell matrix (Fig. 4 Id). It is inthe form of a network, made of fused rings each having an internal diameter of about250-320 nm and walls 200-260 nm thick (Fig. 41e). The aperture is circular and surroundedby assorted particles of quartz to give it an irregular outline (Fig. 41c). 60 C. G. OGDEN Fig. 39 Dijjlugia urceolata: a, apertural view x 180; b, lateral view, note the apertural collar andsmall aboral protruberances x 260; c, part of apertural collar, note the well-defined edge of smallparticles x 790; d, portion of shell surface to illustrate the distribution of organic cement x 5600;e, detail of apertural collar shown in c. x 3700; f, detail or organic cement network x 1 5 000. DIFFLUGIA IN BRITAIN 61 Fig. 40 Difflugia urceolata specimen with a deformed shell; a, lateral view x 1 70; b, latero-apertural view x 1 60. Some of the examined specimens had cyst plugs in their apertural openings. These plugsvaried from being either an uneven mixture of angular quartz (Fig. 42a) or flattish pieces(Fig. 42b), in both instances the sealing cement was similar to that binding the shell walls(Figs.42c&d). MEASUREMENTS (in jim). Based on twenty-six specimens: body length 137-171, breadth87-1 37, depth 57-84 diameter of aperture 35-52; B/L 0-67 0-06, d/L 0-26 0-02. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum moss gatheredat Myndd Hiraethog, Denbigh, Clwyd, North Wales in August, 1980. REMARKS. This species is similar to two other compressed species namely, D. compresscf!and D. lingula Penard, 1911. Complications regarding the species D. compressa Carter, 1 864should have been resolved by Cash & Hopkinson (1909) who suggested that from Carter'sfigures he was 'beyond question' referring to a species of Pontigulasia. Nevertheless, thename has been used subsequently to refer to compressed specimens of Difflugia, either as D.compressa or D. oblonga/pyriformis var compressa. Whether or not there are some genuinespecimens of Difflugia amongst these descriptions is difficult to know, but the namecompressa is preoccupied by Carter's description and is no longer valid, and most refer tomuch longer, broader specimens than those described here. The present specimens aredistinct from D. lingula Penard, 1911 and D. lingula var regularis Gauthier-Lievre &Thomas, 1958 because these have a more rounded shape which tapers sharply from themid-body region to the aperture, and D. lingula also has an aboral horn. D. hiraethogii can be recognised by its lateral compression, distinct circular collar andaperture. ETYMOLOGY. This species is named after the area of North Wales in which it was found. Difflugia lucida Penard, 1890 DESCRIPTION. The shell is transparent, ovoid, gracefully curved aborally but tapering moregradually towards the aperture to give a well defined outline (Fig. 43a), and laterallycompressed (Fig. 43c). It is thin, smooth and composed mainly of flattish pieces of quartzwith an occasional siliceous shell plate or diatom frustule added, these particles are usuallyarranged so that they meet but do not overlap. Small areas of organic cement, in the form of anetwork, are seen as part of the shell matrix (Fig. 43d). The network is often an arrangement 62 C. G. OGDEN Fig. 41 Dijjlugia hiraethogii: a, lateral view x 730; b, lateral view to illustrate the distinct circularneck and compressed body x430; c, apertural view x540; d, portion of shell surface showingsmall areas of organic cement x 3500; e, detail of organic cement x 1 3 000. DIFFLUGIA IN BRITAIN Fig. 42 Difflugia hiraethogii: a, detail of aperture with cyst plug composed mainly of angularquartz, organic cement at edges x 1 500; b, specimen with cyst plug composed mainly of organiccement x 1100; c, portion of cyst plug shown in b., note that the particles appear to be wellembedded in organic cement x 3500; d, detail of organic cement of cyst plug x 1 7 000. of rings whose internal diameter is about 38(M80 nm with walls 95-125 nm thick (Fig. 43e).The aperture is elliptical and surrounded by irregularly arranged particles which give arough outline to the immediate apertural region (Figs. 43a & b). Several presumably encysted specimens were present in the sample, and easilydistinguished optically by the dark -appearance around the aperture. On detailedexamination this dark area was seen to be a concentration of flat particles projecting from theapertural openings (Figs. 44a & b). MEASUREMENTS (in urn). Based on thirty-six specimens: body length 67-91, breadth 40-55,depth 23-37, diameter of aperture 23-29, depth of aperture 13-19. MATERIAL EXAMINED. Specimens were collected from a sample of Sphagnum gathered atMyndd Hiraethog, Denbigh, Clwyd, North Wales in August, 1980. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina(Vucetich, 1972), Austria (Laminger, 1972a, 19736, 1974, 1975), Belgium (Chardez, 19616;Couteaux, 1969), British Isles (Cash & Hopkinson, 1909), Bulgaria (Golemansky, 1967),Canary Isles (Gracia, 1965a & b), China (Bartos 19636), Congo (Chardez, 1964; Stepanek,1963), Costa Rica (Laminger, 1973a), Czechoslovakia (Rosa, 1957; Stepanek, 1952, 1967), 64 C. G. OGDEN Fig. 43 Difflugia lucida: a, lateral view to illustrate basic outline x 1 300; b, apertural viewx 1 100; c, view showing lateral compression and smooth surface x 840; d, part of shell surfacewith small areas of organic cement x 5900; e, detail of organic cement network x 23 000. DIFFLUGIA IN BRITAIN 65 Fig. 44 Difflugia lucida specimen with cyst plug, note the irregular arrangement of particles inthe apertural opening: a, lateral view x620; b, apertural view x 1 100. angulostomai \ angulostoma minuta mica \ stout/ i / \ masaruziil \ decloitrei / \ ampullula Fig. 45 Diagrams of ovoid, spherical and compressed species to illustrate the basic outline, basedon measurements given in Table 4. Note that D. rotunda and D. urceolata are drawn to thereduced scale. 66 C. G. OGDEN Table 4 Average dimensions of ovoid or spherical species andcompressed species used to give basic outlines illustrated in Fig. 45 lucidahiraethogii length 76150 breadth depth 46104 3067 diameter ofaperture 17x2541 France (Thomas, 1954), Germany (Schonborn, 19620 & b\ Guatemala (Laminger, 19730),Hungary (Varga, 1963), Italy (Grandori & Grandori, 1943; Rampi, 1950), Java (Bartos,19630; Hoogenraad & Groot, 19406), Mexico (Laminger, 19730), Morocco (Decloitre,1961), Nepal (Laminger, 19726), Netherlands (Hoogenraad & Groot, 19400), Poland(Golemansky, 1970; Pateff, 1926), South Shetland Isles (Smith, 1972), Spain (Gracia, 1964),Switzerland (Penard, 1902). REMARKS. Some differences in dimensions are worth noting from earlier descriptions; Penard(1 890) gave a range of body length 50-70 um, but later stated that specimens ranged between50-60 and rarely greater than 65 um (Penard, 1902), Cash & Hopkinson quoted 60-80 um,whilst Gauthier-Lievre & Thomas (1958) suggested that there might be three groups (a)44-50 (b) 55-70 (c) 83-90. In the present group of specimens only seven are outside of therange 70-80 um and they are remarkable for their similarity.This species is distinct in having a well defined shape and by being evenly compressed. Discussion One of the main difficulties encountered in trying to identify specimens ofDifflugia is due tothe irregular shape of the shell. It is a problem shared with other agglutinate species ofprotozoa, such as the foraminifera. In general most species have a regular basic outline,which may be altered by either natural variation or obscured by the addition of extraneousmaterial. Both natural and additional variation are probably related to the composition ofthe shell, fragile shells being more likely to be influenced by disturbances in the environmentduring or after construction, whilst robust shells may be so encrusted by particles that anyresemblance of a specific shape is lost. Fragile shells are usually made of small particles arranged in a single layer and often haveorganic cement as a major component of the shell matrix. In some species a smooth surfacecomposed of flattish particles is constructed, for example D. mamillaris, where variation in DIFFLUGIA IN BRITAIN 67 general shape is frequently seen especially in the aboral region (see Fig. 23). Nevertheless, ina species with a similar surface but different shape, D. lanceolata which is rounded ratherthan pointed in the aboral region, there is a relatively constant shape. In the present report athird of the specimens of D. mamillaris differ from the basic outline given in Fig. 29, whilstall of the specimens of D. lanceolata agree with the outline in Fig. 1 8. Although robust shellsare usually made of angular quartz which do not lend themselves to being arranged in aregular manner, if enough specimens of a species is present in a sample it is possible toillustrate a basic outline. The problems of subsequent recognition of such species from thebasic outline is complicated when the diagnostic feature is obscured by the arrangement ofparticles. For instance the diagnostic feature may be the presence of a neck, but if this ishidden it may be identified incorrectly. Amongst species with this type of shell, examples ofspecimens incorporating a single large particle with similar dimensions to the whole shellhave been observed and a not infrequent sight is to see two similar shells united. In theselatter instances the shell is usually of similar size and composition, but these are notnecessarily species of Difflugia but can be other agglutinate forms like Pontigulasia (pers.observation). To assist in resolving the question of what represents the basic outline in the speciesdescribed here, three sets of line drawings are provided (Figs. 18, 29, 45) which are based onthe average dimensions of the specimens examined. In the previous studies on pyriform species of Difflugia it has been suggested (Ogden,1979) that measurements are useful in distinguishing species, with the body length anddiameter of aperture perhaps being the more stable dimensions. However, it was emphasisedthat these features alone are not usually sufficient to warrant specific diagnoses. The problemof using dimensions as a diagnostic character is that they may be valid for a proportion ofspecies in a genus, but do not hold for all especially in the present instance with a genussupposedly comprised of over three hundred species. Possibly this is best illustrated by thevariability in size exhibited within a species of testate amoebae, the smallest often beingreported as half the size of the largest, which does not pose problems of identification whenthe body length is under 80 um, but for those of larger dimensions the difference between200 jim and 400 urn can often be interpreted as representing two separate species. Theextreme example is as we have noted previously (Ogden & Fairman, 1979) the range ofmeasurements quoted for the body length of D. oblonga, 60-580 um, which is so variablethat it could embrace most of the genus or almost all of the described testate amoebae.Nevertheless, there are examples of consistent dimensions within a species, for instancethose of D. lanceolata and D. mamillaris described here are in good agreement withpreviously published results (Penard, 1902; Grospietsch, 1957). The regularity of bodylength in D. lanceolata is such that over 87% of the specimens fall within a range of 10% ofthe average value given in Table 2, whilst in D. mamillaris 96% fall within the same range. Ovoid or spherical specimens of Difflugia present the same problem. In certain casesgroups of similar species may only be distinguished by dimensions, for example D. anchlora,D. gramen and D. lobostoma (see Ogden, 1980) which may represent a phylogenetic series.Whilst is other cases, like the D. globulosa/globularis species complex, size variation is sogreat that it is difficult not to include any ovoid or spherical specimen between 50-1 50 um inthis complex. Part of the difficulty is illustrated by the two groups of specimens describedhere as D. angulostoma (p. 47), where the shells are identical in construction and essentiallythey share the same shape, but there are differences in dimensions especially the diameter ofthe aperture. It is possible to separate some of these small spherical species using the latterfeature in addition to other differences, as shown in the descriptions of D. minuta, D. micaand D. microstoma (see p. 53, p. 50 & p. 53). Structural differences in basic outline asmentioned earlier are mainly related to shell components and deformities, the formerconcerns the choice of materials and will be dealt with later, but the latter using D. urceolataas an example may be due to its large size. Perhaps it is easier to understand if one considersthat in all probability the shell components are not cemented together until the final shape 68 C. G. OGDEN has been moulded by cytoplasmic movements. As this process takes about sixty minutes in asmall siliceous species (Ogden, 1981), it will probably take considerably longer in a largeranimal, during which time in a natural environment there is a continual motionand hence a possibility of disruption. The result of such a disturbance may produce the shellillustrated in Fig. 40, sufficiently different from normal but not enough for the animal toabhort and discard the shell prior to the final stiffening of the cement. Studies on clonal cultures of both siliceous and proteinaceous species (Ogden, 1981 &pers. observations) show that variation in dimensions are small, differences outside the normusually being attributed to abnormal development and even here it is usually below 5%.Such abnormal development is thought to be associated with cultural differences and not afrequent natural occurrence. One feature of siliceous species behaviour which may explainsome changes in dimensions, is the occasional production of a shell having a doublecomplement of shell plates (Hedley & Ogden, 1973). However, this results in an increase involume of an ovoid structure, which means that the enlargement in body length is probablyno greater than a third. At present there are only two reports (Jennings, 1916, 1937) on the development ofDifflugila corona in the laboratory. Both have shown that there can be some variation inshell construction and until further observations are available on other species of Difflugia,the question of shell size and composition as diagnostic features will remain a subject ofspeculation. Mention has already been made in the literature of differences in shell construction andthe three categories which are readily identified, robust, intermediate and fragile (Ogden,1980). But the choice of materials, other than a comment (Ogden, 1980) on the influence ofpH, has not been discussed. That a system of choice is available to the animal is clearlydemonstrated by the composition of certain shells. For example, D. minutissima, D.lanceolata, D. mamillaris and D. decloitrei all use flattish pieces of quartz, in some instancessmall flat particles of diatom frustules may be substituted, but the components used appearto be restricted in size and thickness. The function of the cytoplasm to identify and selectthese particles may appear to be extreme. Nevertheless, it can be measured against the abilityof siliceous testate amoebae to hold each shell plate during shell construction, place it inposition so that there is an even amount of overlap between plates and in some specimensmanipulate spines into definite positions (Ogden, 1981). This selectivity is not restricted toflat particles, but probably includes the choice of diatoms or angular particles, as well asmixtures of all types in the composition of Difflugia shells. An additional factor thatinfluences the choice of particles is undoubtedly the structure and extent of organic cementin the shell matrix. It has already been suggested (Ekert & McGee-Russell, 1974) that the organic cementwhich binds the shell particles together in Difflugia lobostoma imparts both strength andflexibility to the structure. This may seem obvious from the different type of shellsconstructed by these animals, but it has an importance related to the material used. Forinstance, when the cement becomes part of the surface matrix, usually in species with flatparticles, it is found at each facet as part of the shell wall. The strength of the shell is thendirectly related to the tenacity of the cement at these junctions. Furthermore, in species where there is some overlapping of particles the cement isinterwoven with the material and can be likened to the structure of a brick wall, in which thestrength is dramatically increased by the combination of bricks and mortar beyond thestrength of the individual materials assessed on their own. The importance of the organiccement in shell structure is easily demonstrated by treating a robust individual with either achelating agent or concentrated sulphuric acid, in each case within a short time it is reducedto a small residue of particles. Strength is not directly due to the composition of the acidmucopolysaccharide material that forms the basic organic cement, but to the properties ofthis material. It has already been shown that inorganic elements incorporated with this typeof material in the proteinaceous shells of testate amoebae (Hedley et al, 1976; pers. observ.) DIFFLUGIA IN BRITAIN 69 and areanceous foraminifera (Hedley, 1963) are thought to strengthen the shell. Somespecimens of Difflugia have been examined by X-ray microanalytical techniques (pers.observ.), and found to have a significant amount of ferrous iron associated with areas oforganic cement. This probably accounts for the reports of yellow or brown specimens, thedegree of colouration being proportional to the amount of inorganic elements bound to theorganic cement. There is no doubt that the density of this colouration in proteinaceousspecimens can be used as a measure of the degree of reinforcement that the inorganicelements impart to the structure, newly formed shells being light coloured and fragile, whilstolder shells are dark and strong. This measure can probably be applied to agglutinate formsas well. The differences in the network structure of the organic cement are harder to understand,especially as the examination is limited to surface detail. A need for porosity in some of theindividual organic cement units is puzzling. If they are definite pores to the interior of theshell they might function as pressure valves for aqueous interchange, because often when theanimal is moving or feeding the apertural opening is completely blocked by cytoplasmicextrusions. Such a scheme could ensure that the internal volume unoccupied by cytoplasm isnot isolated and allowed to stagnate. Alternatively it may be associated with the hardeningprocess by inorganic elements. This process seems to be directly related to the environmentand the chemical composition of the cement, the activities of the animal apparently havingno effect on this association. The strengthening process has some degree of justificationbecause the pores often seen between individual proteinaceous units is newly-formed, lightcoloured, shells ofArcella, are not seen in older darker specimens. Examination of the wallsof such specimens show that they are thick and stronger, the implication being that theinorganic elements have strengthened the shell not the deposition of further organic material(pers. observ.). The diversity in the construction of the organic cement units is considered to be a goodtaxonomic feature, although they are beyond the limit of optical microscopy, the appearanceof some being particularly unique for example the button-type of D. lacustris (see Fig. 5ep. 9). The sharing of the same type of unit between different species may suggest somephylogenetic relationship, possibly linked to the type of shell construction whether smooth,rough, fragile or strong. However, at present only a quarter of the described species havebeen examined and it is too early to make proposals on such relationships. That is apart fromthe apparent sharing of the same organic cement pattern between most ovoid species. It ishoped that further studies in progress on this genus will help to unravel the complicationsattributed to describing so many different shapes and forms, and allow a comprehensivedivision based on shell structure. ReferencesAwerinzew, S. 1906. Die Strucktur und die chemische Zusammensetzung der Gehause bei den SiiBwasser-rhizopoden. Arch. Protistenk. 8: 95-1 11.Bartos, E. 1963a. Rhizopoden einiger moosproben aus Java. Acta Univ. Carol. 1 19-190 19636. Die Rhizopoden einiger Moosproben aus China. V$stn. csl. Spol. zool. 27: 85-96. Bereczky, M. Cs. 1973. Beitrage zur Kenntnis der im Eprofundal des Balaton lebenden testaceen. Annies Univ. Scient. bpest. Rolando Eotvos (Sect. Biol.) 15: 1 17-127.Blanc, H. 1892. Les Difflugies de la faune profonde du Lac Leman. Recuiel inaugural de 1'Umversite de Lausanne 8. Lausanne.Carter, H. J. 1864. On Freshwater Rhizopoda of England and India; with illustrations. Ann. Mag. nat. Hist. (3)13: 18-39.Cash, J. & Hopkinson, J. 1909. The British Freshwater Rhizopoda and Hehozoa. Vol. II Rhizopoda, part 2. 166 pp. The Ray Society, London.Cash, J., Wailes, G. H. & Hopkinson, J. 1919. The British Freshwater Rhizopoda and Hehzoa. Vol. IV. Supplement to the Rhizopoda. 1 30pp. The Ray Society, London. 70 C. G. OGDEN Chardez, D. 1956. Thecamoebiens de la region Vervietoise. Revue verviet. Hist. nat. 13 (3-4): 23-32. - 1 960. Etudes sur deux Difjlugia. Hydrobiologia 16: 1 1 8-125. - 196 la. Sur Difjlugia acuminata Ehrenberg (Rhizopoda, Testacea). Bull. Inst. agron. Stns Rech. 19616. Catalogue des Thecamoebiens de Belgique. Protozoa, Rhizopoda, Testacea. Bull. Inst.agron. Stns. Rech. Gembloux 29: 269-300. 1964. Thecamoebiens (Rhizopoda, Testacea). Expl. hydrobiol. L. Bangweolo-Luapula 10 (2);1-77. 1973. Sur Dijjlugia ventricosa Deflandre, 1926 (Rhizopoda, Thecamoebien). Revue verviet. Hist.nat. 30 (7-9): 57-60.-. 1980. Thecamoebiens de 1'Etang de Thommen (Protozoa, Rhizopoda, testacea). Naturalistes belg. 61: 88-91.Chardez, D. & Caspar, Ch. 1976. Thecamoebiens aquatiques du domaine des Epioux (Ardenne, Belgique). Biol. Jaarb. 44: 86-100.Couteaux, M. M. 1969. Thecamoebiens muscicoles de Gaume et de Moyenne-Belgique. Revue Ecol. Biol. SV>/.60:413-428.Decloitre, L. 1947. Rhizopodes de la Station de Tanaf. Feuille Nat. 2: 91-94. 1948. Materiaux pour une faune rhizopodique d'A.O.F. Bull. Inst.fr. Afr. noire 10: 235-285. 1961. Materiaux pour une faune thecamoebien du Maroc: 2 e note. Thecamoebiens des sols aeriens des palmiers de Marrakech. Bull. Soc. Sci. nat. Maroc 41: 121-1 36. 1965. Contribution a la faune du Congo (Brazzaville) Mission A. Descarpentardes et A. Villiers. Ill Rhizopodes, Thecamoebiens. Bull. Inst.fr. Afr. noire 27 A: 165-184.Deflandre, G. 1926a. Notes sur quelques Rhizopodes et Heliozoaires du Venezuela. Bull. Soc. zool. Fr. 51: 5 15-530.19266. Materiaux pour la faune rhizopodique de France. Rhizopodes du Confolentais. Feuille Nat. 33: 169-170. 193 1 . Thecamoebiens nouveaux ou peu connus 1 . Annls. Protist. 3: 81-95. Dioni, W. 1 970. Taxocenos de tecamebianos en cuencas islenas del Parana Medio 1 . Acta zool. lilloana 27:201-240.Eckert, B. S. & McGee-Russell, S. M. 1974. Shell structure in Difjlugia lobostoma observed by scanning and transmission electron microscopy. Tissue Cell 6: 2 1 5-22 1 .Ehrenberg, C. G. 1938. Die Infusionsthierchen als vollkommene Organismen. Bin Blick in das tiefere organische Leben der Natur. 547 S. Folio, Leipzig. 1854. Mikrogeologie. Fol. Leipzig. 1872. Uebersichte der seit 1847 fortgestzten Untersuchungen iiber das in der Atmosphare unsichtbar getragene reiche organische Leben. Phys. Math. Abh. K. Acad. Wiss Berlin (1 87 1 ): 1-1 50.Ertl, M. 1965. Zur Kenntnis der Testaceenfauna der slowakischen Reisfelder. Hydrobiologia 26: 13-20.Frenzel, J. 1892. Untersuchungen iiber die mikroskopische Fauna Argentiniens. I. Die Protozoen. I & II Die Rhizopoden und Helioamoeben. Biblthca. zool. Stuttgart 12, 162pp Xpl.Gal, D. 1969. Zoolplanktonuntersuchungen im Ostlichen-Hauptkanal. Acta biol. SzegedlS: 93-100.Gassowsky, G. N. 1936. Quelques rhizopodes nouveau des lacs du groupe de Kontchesero (En Karelie). Trudy p res nov. biol. Sta. S-petreb. Obshch Estest. 8 (2): 101-121.Gauthier-Lievre, L. & Thomas R. 1958. Les genres Dijjlugia, Pentagonia, Maghrebia et Hoogenraadia (Rhizopodes, testaces) en Afrique. Arch. Protistenk. 103: 241-370.Godeanu, S. 1972. Especes nouvelles de thecamoebiens (Protozoa, Rhizopoda, Arcellinida). Revue roum. Biol. (Ser. Zool.) 17: 227-236.Godeanu, S., Godeanu, M. & lonescu, V. 1973. Varia^a anualS a planctonului din incintS indiguita Obretin (Delta Dunarii). Peuce3: 21 1-260.Golemansky, V. 1967. Etudes de la faune de Rhizopodes (Sarcodina, Rhizopoda) des mousses epiphytes et terricoles en Bulgarie. Izv. zool. Inst. Sof 24: 103-1 19.1970. A list of Testacea (Protozoa, Rhizopoda) from the Duszatyriskie Lakes in Poland. Fragm. faun. 16: 21-25.Gracia, M. P. 1964. Tecamebas muscfcoles del valle de Ribas (Gerona). Publnes Inst. Biol. apl. Barcelona 37: 67-71. 1965a. Tecamebas muscicoles de Tenerife. Publnes Inst. Biol. apl. Barcelona 39: 123-127. 19656. Tecamebas muscfcoles de Gran Canaria. Publnes Inst. Biol. apl. Barcelona 38: 93-96. 1972a. Tecamebas esfagnicolas de la Peninsula Iberica. Publnes Inst. Biol. apl. Barcelona 52: 5^2. DIFFLUGIA IN BRITAIN 71 19726. Tecamebas musicolasde la Peninsula Iberica. Publnes Inst. Biol. apl. Barcelona 53: 5-16.Grandori, R. & Grandori, E. 1934. Prime richerche sui Protozoi del terreni e delle acque dell'alta Montagna Alpina. Boll. Lab. Zool. agr. Bachic. R. 1st. sup. agr. Milano 5: 1-340.Green, J. 1975. Freshwater ecology in the Mato Grosso, Central Brazil, IV: Associations of testate Rhizopoda. J. nat. Hist. 9: 545-560.Grospietsch, T. 1957. Beitrag zur Rhizopodenfauna des Lago Maggiore. Arch. f. Hydrobiol. 53: 323-331.1958. Beitrage zur Rhizopodenfauna Deutschlands I. Die Thekamoben der Rhon. Hydrobiologia 10: 305-322. 1975. Beitrag zur Kenntnis der Testaceen-Fauna des Lago Valencia (Venezuela). Verh. Internal Verein. Limnol. 19: 2778-2784.Hedley, R. H. 1963. Cement and iron in the arenaceous foraminifera. Micropaleontology9: 433-441.Hedley, R. H., Ogden, C. G. & Mordan, N. J. 1976. Manganese in the shell of Centropyxis (Rhizopodea: Protozoa). Cell Tiss. Res. 171: 543-549.Hoogenraad, H. R. & Groot, A. A. de. 19400. Fauna van Nederland Vol. 9. Zoetwaterrhizopoden en Heliozoen. 302pp Leiden.19406. Moosbewohnende thekamoebae rhizopoden von Java und Sumatra. Treubia 17: 209-259.Jennings, H. S. 1916. Heredity, variation and the results of selection in the uniparental reproduction of Difflugia corona. Genetics 1: 407-534.1937. Formation, inheritance and variation of the teeth in Difflugia corona. A study of the morphogenetic activities of rhizopod protoplasm. /. exp. Zool. 77: 287-336.Jung, W. 1936. Thekamoben urspriinglicher, lebender deutscher Hochmore. Abh. Landesmus. Prov. 1942. Siidchilenische Thekamoben (Aus dem siidchilensichen Kiistengebiet, Beitrag 10). Arch. Protistenk. 95: 253-356. Kourov, O. 1 925. Faune rhizopodique des bassins de Kossino. Trudy kosin. biol. Sta. 2: 43-68.Laminger, H. 1971. Sedimentbewohnende Schalenamoben (Rhizopoda, Testacea) der Finstertaler Seen (Tirol). Arch Hydriobiol. 69: 106-140.. 19720. Terrestrische Testaceen (Protozoa, Rhizopoda) in der Umgebung von Obergurgl (Osterrich, Tirol). Pedobiologia 12: 16-22.19726. Notes on some terrestrial Testacea (Protozoa, Rhizopoda) from Nepal, Himalaya (Lhotose Shar). Arch. Protistenk. 114: 486^88.1972c. Die profundale Testaceeenfauna (Protozoa, Rhizopoda) alterer und jiingerer Bodensee-Sedimente. Arch. Hydrobiol. 70: 108-129.19730. Die Testaceen (Protozoa, Rhizopoda) einiger Hochgebirgsgewasser von Mexiko, Costa Rica und Guatemala. Int. Revue ges. Hydrobiol. 58: 273-305.19736. Zur Kenntnis der Testaceenfauna in den jungsten Sedimenten des Bodenseeprofundals und-litorals. Schweiz. Z Hydrol. 35: 239-246.1974. Ein Beitrag zur Kenntnis der Protozoenfauna der Donau. 1. Die Testaceen (Protozoa, Rhizopoda) im Abschnitt Obernzell-Linz (Oberosterreich). Arch. Hydrobiol. Suppl. 44 3: 330-337.1975. Die Sukzession der Testaceen- Assoziationen (Protozoa, Rhizopoda) im rezenten und subfossilen Sphagnum des Obersees bei Lunz (Neiderosterreich). Hydrobiologia 46: 465^87.Laminger, H., Zisette, R. Phillips, S. & Breidigam, F. 1979. Contribution to the knowledge of the protozoan fauna of Montane (USA): 1. Testate amoebae (Rhizopods) in the surrounding of Flathead Lake valley. Hydrobiologia 65: 257-27 1 . Leidy, J. 1 874. Notes on some New Fresh-water Rhizopods. Proc. Acad. nat. Sci. Philad. 77-79.1879. Freshwater Rhizopods of North America in Vol. 12. United States Geological Survey of the Territories. 324pp. Washington.Levander, K. M. 1894. Materialien zur Kenntniss der Wasserfauna in der Umgebung von Helsmgtors, mit besonderer berucksichtigung der Meeresfauna. I. Protozoa. Acta Soc. Fauna Flora Fenn. 12 No. Margalef, R.I 955. Contribution al estudio de la fauna de las aguas dulces del norveste de Espana. Publnes Inst. Biol. apl. Barcelona 21:137-171.Moraczewski, J. 1961. Testacea du littoral peu profond du lac Kisajno (Region des lacs de Mazune). Polskie Archwm. Hydrobiol. 9: 175-194.1965. Taxocenses des Testacea de quelques petit bassins de terrains mondables de la Narew. Acta Protozool.3: 189-213. 72 C. G. OGDEN Ogden, C. G. 19790. Comparitive morphology of some pyriform species of Difflugia (Rhizopoda). Arch. Protistenk. 122: 143-153.1979ft. Siliceous structures secreted by members of the subclass Lobosia (Rhizopodea, Protozoa). Bull. Br. Mm. nat. Hist. (Zool.)36: 203-207.1980a. Shell structure in some pyriform species of Dijjlugia (Rhizopodea). Arch. Protistenk. 123: 455-470.1980ft. Notes on some Difflugiidae from Norfolk (Rhizopodea, Protozoa). Bull. Br. Mus. nat. Hist. (Zool) 39: 125-1 38. 1981. Observations on clonal cultures of Euglyphidae (Rhizopoda, Protozoa). Bull. Br. Mus. nat. Hist. (Zool.) 41: 137-151.Ogden, C. G. & Fairman, S. 1979. Further observations on pyriform species of Dijjlugia (Rhizopoda). Arch Protistenk. 122: 372-381.Ogden, C.G. & Hedley, R. H. 1980. An Atlas of Freshwater Testate Amoebae. British Museum (Nat. Hist.), London & Oxford University Press, Oxford 222pp.Opravilova, V. 1974. Testacea (Protozoa: Rhizopoda) of the river Bobrava in Moravia. Vestnik csl. Spol. zool. 38: 127-147.Oye, P. van 193 1 . Rhizopoda from South Africa. Rev. Zool. hot. afr. 21: 54-73. 1953. Faune Rhizopodique de Petang de Beernem. Biol. Jaarb. 20: 1 54-205. 1958, Etude sur les rhizopodes des marais du Sud-ouest d'Uvira (Congo-Beige). Hydrobiologia 10: 85-137.Pashitnowa, Z. A. 1929. Materialen zur Erforschung der Mikrofauna der Reisfelder und die Biologic der Anopheles-larvae auf den Reisfeldern. Acta Univ. Asiae mediae Ser. Villa 10: 1-42.Pateff, P. 1926. Siisswasser-Rhizopoden aus der Hohle Salzlocher (Schlesien). Mitt, Hohlen-u Karstforsch. 2: 46-49.1927. Die von Romer und Schaudinn im Reliktensee Mogilnoje gesammelten Siisswasserrhizqpoden. Zool. Anz. 70: 36-38. Penard, E. 1 890. Etudes sur les Rhizopodes d'eau douce. Mem. Soc. Phys. Hist. nat. Geneve 31 : 1-230.1893. Pelomyxa palustris et quelques autres organismes inferieures. Archs. Sci. phys. nat. 29: 165-182. 1 899. Les Rhizopodes de Faune profunde dans le lac Leman. Revue suisse Zool. 1: 1-142. 1901 . Notes complementaires sur les Rhizopodes du Leman. Revue suisse Zool. 9: 225-241 . 1 902. Faune Rhizopodique du Bassin du Leman. Geneva 700pp. 1904. Quelques nouveaux Rhizopodes d'eau douce. Arch. Protistenk 3: 391-422. 1 905. Les Sarcodines des Grands Lacs. Libraire de 1'Institut, Geneve 1 33pp. 1911. Rhizopode d-eaux douce. Brit. Antarct. Exped. 1907-1909. Reports on Scient. Investig. vol.1. Playfair, G. J. 1918. Rhizopods of Sidney and Lismore. Proc. Linn. Soc. N.S. W. 42: 633-675.Rampi, L. 1950. Su alcuni Laghetti Alpini del Massiccio dell'Asbisso (Alpi Marittime). Boll. Pesca Pisci. Idrobiol. 26: 207-224.Rosa, K. 1957. Bodenmikrofauna undmikrofauna im Fichtenbestande am PradSd (Altoater). Prirodov. Sb. ostrav. Kraje 18: 7-75.Rosa, K. & Lhotsky, O. 1971. Edaphische Algen und Protozoen im Isergebirge Tschechoslowakei. 0/A:0s22:21-29.Schonborn, W. 1962. Die Okologie der Testaceen im oligotrophen See, dargestellt am Biespiel des GroBen Stechlinsees. Limnologica 1: 1 1 1-182. 1965. The sediment-inhabiting Testacea from some Masurian Lakes. Acta. Protzool. 3: 297-309. 1966. Untersuchungen iiber die Testaceen Schwedisch-Lapplands. Ein Beitragzur Systematik und Okologie der beschalten Rhizopoden. Limnologica 4: 5 1 7-559.Smith, H. G. 1972. The terrestrial protozoa of Elephant Island, South Shetland Islands. Bull. Br. Antarct. Surv. 31: 55-62.Stepanek, M. 1952. Testacea of the pond of Hradek at Kunratice (Prague). Sb. nar. mus. PrazeSE (3): 1-55. 1963. Die Rhizopoden aus Katanga (Kongo-Afrika). Annls Mus. r. Afr. cent. 117: 9-91 . 1967. Testacea des Benthos der Talspere Vranov am Thayafluss. Hydrobiologia 29: 1-66. Thomas, R. 1953. Sur deux formes critiques du genre Dijjlugia Leclerc. Bull. Soc. zool. Fr. 78:132-136. 1954. Thecamoebiens de la region Bordelaise. Bull. Soc. Hist. nat. TolouseS9: 245-264. Thomas, R. & Mabille, J. 1956. Rhizopodes thecamoebiens observes dans le Department de 1'Aisne. Cah. Nat. 12: 26-32. DIFFLUGIA IN BRITAIN 73 Varga, L. 1 963. Weitere Untersuchungen iiber die aquatile Microfauna der Baradla-Hohle bei Aggtelek (Ungarn). (Biospeologia Hungarica, XVIII). Acta Zool. hung. 9: 439-458.Voeltz-Hohn, J. 1971. Ein Beitrag zur Testaceenfauna (Protozoa, Rhizopoda) des Hochmoores im Naturpark "Hoher Vogelsberg". Philippa 1: 80-84. Vucetich, M. C. 1970. Algunos tecamebianos de la provincia de Formosa. Neotropica 16: 42-48.1972. Tres nuevas tecamebas muscicoles para la Argentina (Rhizopoda Testacea). Neotropica 18: 126-128.1973a. Tecamebas muscicoles de la selva marginal de punta lara (Provincia de Buenos Aires, Argentina). Physis (B) 32: 6 1-66.1973&. Estudio de tecamebianos Argentinos, en especial los del Dominio Pampasico. Revta. Mus. La Plata (Zool.) 11: 287-322.1978. Neuvos aportes al conocimiento de los tecamebianos del Dominio Subtropical. Neotropica 24: 79-90.Wailes, G. H. 1912. Freshwater Rhizopoda and Helizoa from the States of New York, New Jersey and Georgia, U.S.A.; with supplemental note on Seychelles species. J. Linn. Soc. 32: 121-161 .Wallich, G. C. 1864. On the extent, and some of the principal causes, of structural variation among the Difflugian Rhizopods. Ann. Mag. nat. Hist. (3) 13: 215-245.Wang Jiaji 1977. Protozoa from some districts of Tibetan Plateau. Acta zool. sin. 23: 131-160. Manuscript accepted for publication 4 June 1982 Index of species and synonyms Difflugia acuminata inflata Penard, 1 899 30 amp hora Leidy, 1879 33amphoralis Cash & Hopkinson, 1909 27 ampullula Playfair, 1918 45angulostoma Gauthier-Lievre & Thomas, 1958 47bicruris Gauthier-Lievre & Thomas, 1958 30 bryophila (Penard, 1902) 2 cylindrus (Thomas, 1953) 5 decloitrei Godeanu, 1972 47 distenda nom. nov. 30 elegans tricornis Jung, 1 936 38 gassowskii nom. nov. 5glans Penard, 1902 globularis microstoma Thomas, 1 954 53 globularis sphaerica Chardez, 1956 56 gramen Penard, 1902 50 hiraethogii sp. nov. 59 labiosa Wailes, 1919 33 lacustris (Penard, 1899) 9 lanceolata Penard, 1 890 1 1 lemanii Blanc, 1892 26 levanderi Playfair, 1918 47 linearis (Penard, 1890) 11 longicollis(Gasso\vsky, 1936) 5 lucida Penard, 1 890 61 mamillaris Penard, 1893 35 manicata Penard, 1902 16 masaruzziOye, 1958 50 mica Frenzel, 1 892 50 microclaviformis(Kourov, 1925) 35 microstoma (Thomas, 1954) 53 minuta Rampi, 1950 53 Difflugia minuta minor Godeanu, 1972 22 minutissima Penard, 1904 16 molesta Penard, 1902 38 oblonga bryophila Penard, 1 902 2 oblonga cylindrus Thomas, 1953 5 oblonga elongata Oye, 1953 17oblonga lacustris Cash & Hopkinson, 1909 9 oblonga linearis Penard, 1 890 1 1oblonga microclaviformis Kourov, 1925 35 oblonga parva Thomas, 1 954 1 7oblonga tenuis Wailes & Penard, 1911 24oblonga venusta Cash & Hopkinson, 1909 41 parva (Thomas, 1954) 17 paulii nom. nov. 1 7 petricola Cash, 1909 20 pristis Penard, 1902 20pulex Penard, 1902 pyriformis bryophila Penard, 1 902 2 pyriformis lacustris Penard, 1 899 9pyriformis longicollis Gassowsky, 1936 pyriformis tenuis Penard, 1 890 24 pyriformis venusta Penard, 1902 41 rotunda nom. nov. 56 stoutii sp. nov. 57tenuis (Penard, 1890)tricornis (Jung, 1936)urceolata Carter, 1 864 ventricosa Deflandre, 1 926 4 1 venusta (Penard, 1902) 41 viscidula Penard, 1902 24 Sexangularia minutissima (Penard, 1 904) 1 6 British Museum (Natural History) An Atlas of Freshwater Testate Amoebae C. G. Ogden & R. H. Medley 1980, Hardcovers, 222pp, 17,50 (18.00 by post). Co-published by British Museum(Natural History) and Oxford University Press. This book illustrates, using scanning electron micrographs, most of the commonspecies of testate amoebae that are found in freshwater habitats. Information onthe biology, ecology, geographical distribution and a classification are followed bydescriptions of ninety-five species. Each of these is illustrated by several views ofthe shell. The text is designed not only to enable biologists to identify species of testateamoebae, but to serve as an introduction to students interested in the taxonomyand biology of these freshwater protozoa. It will be of special interest toprotozoologists, ecologists, limnologists, water treatment specialists andmicropalaeontologists interested in recent sediments. British Museum (Natural History)Publication Sales,Cromwell Road,London SW7 5BD. Titles to be published in Volume 44 Observations on the systematics of the genus Difflugia inBritain (Rhizopoda, Protozoa). By Colin C. Ogden Miscellanea A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin Curds & Irene C. H. Wu Osteology, genitalia and relationships of the Acanthodactylus(Reptilia: Lacertidae). By E. N. Arnold The Opthalmotilapia assemblage of cichlid fishes reconsidered. By Peter Humphrey Greenwood Morphological studies on some Difflugiidae from Yugoslavia(Rhizopoda, Protozoa). By Colin G. Ogden & Andjelija Zivkovic Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) Miscellanea Zoology series Vol 44 No 2 24 February 1983 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, andan 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 SW7 5BD,England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Zool.) Trustees of the British Museum (Natural History), 1983 The Zoology Series is edited in the Museum's Department of ZoologyKeeper of Zoology : Dr J. G. ShealsEditor of Bulletin : Dr C. R. CurdsAssistant Editor : Mr C. G. Ogden ISSN 0007-1498 Zoology series Vol44 No 2 pp 75-1 90 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 24 February 1983 Miscellanea Contents Cirolana cranchi Leach, 1818 (Crustacea: Isopoda: Cirolanidae) redescribed, withnotes on its distribution. By N. L. Bruce & Joan Ellis Valettieta, a new genus of deep-sea amphipod (Gammaridea: Lysianassidae) withdescriptions of two new species from the North Atlantic Ocean. By Roger J.Lincoln & Michael H. Thurston Three new genera of misophrioid copepods from the near-bottom planktoncommunity in the North Atlantic Ocean. By G. A. Boxshall Larval development of British prawns and shrimps (Crustacea: Decapoda: Natantia)4. Palaemon (Palaemon) serratus (Pennant, 1 777) and functional morphology ofswimming. By A. A. Fincham The larval development of the Angular Crab, Goneplax rhomboides (Linnaeus)(Decapoda: Brachyura). By R. W. Ingle & P. F. Clark The larval and first crab stages of three Inachus species (Crustacea: Decapoda:Majidae); a morphological and statistical analysis. By Paul F. Clark . Page 75 85 103 125 163 179 Cirolana cranchi Leach, 1818 (Crustacea: Isopoda:Cirolanidae) redescribed, with notes on itsdistribution Niel L. Bruce Department of Zoology, University of Queensland, St Lucia, Queensland, 4067, Australia Joan Ellis British Museum (Natural History), Cromwell Road, London SW7 5BD Cirolana cranchi is considered to be the type of the genus Cirolana Leach (by monotypy,Bruce, 1981) and as Cirolana is the type genus for the family Cirolanidae (Harger, 1880;Hansen, 1890), the species is of some significance. It comes therefore as some surprise to findthat C. cranchi has not been effectively described since the work of Hansen (1890). Thespecies has frequently been considered to be indistinguishable from Cirolana parva Hansen1890 (Stebbing, 1917; Nordenstam, 1946; Monod, 1976), and though these authors havediscussed at some length the similarities of the two species, none resorted to redescription. Hale (1925) described a variety of Cirolana cranchi from South Australia. Examination ofthat species suggested that not only was Hale's variety quite distinct from C. cranchi, but thatseveral closely similar species exist in the seas around Australia. It was therefore necessary toexamine the holotype and specimens of Cirolana cranchi from England in order to be certainthat none of the Australian species is C. cranchi. To prevent any further confusion of Cirolana cranchi with C, parva and similar relatedspecies, a new description is given here with full figures, from specimens in the BritishMuseum (Natural History) collections. The holotype, initially stored in the BritishMuseum's dry collections (Ellis, 1981) is in extremely poor condition, lacking mostappendages, setae and spines, and is in two pieces. The holotype, taken from Cornwall wasclosely compared to Norman's (1904) specimen to assure conspecificity, and the descriptionand drawings are taken from the latter. Cirolana cranchi Leach(Figs 1-3) Cirolana cranchii Leach, 1818 : 347; Gosse, 1855 : 134, Fig. 230; Hesse, 1866 : 257; Bate & Westwood,1867:296, Fig. 5; Delages, 1881 : 156; Chevreux 1884:519; Koehler, 1886:25, 61; Bonnier,1887: 134;Heape, 1888: 1 76; Robertson, 1888 : 76; Hansen, 1890 : 341, PI. 3, figs 3-3i; 1905:350,PI. 33. fig 3a; Stebbing, 1893 : 343; 1906 : 275; Norman, 1904 : 438; 1907 : 362; Norman & Scott,1906:40, PI. 4; Monod, 1923: 14; 1930: 137, 145, Figs 2, 5B; 1976: 151; Larwood, 1940:33;Barrett & Yonge, 1958 : 99, Fig. 59; Crothers, 1966 : 58; Naylor, 1972 : 28, Fig. 9A-C; Ryland &Nelson-Smith, 1975 : 252; Kussakin, 1979: 191, Figs 71, 72; Bruce. 1981 : 949. Nelocira swainsonii Leach, 1818 : 347; Desmarest, 1825 : 302, PI. 48, fig. 2. Eurydice swainsonii: Milne-Edwards, 1840 : 236. Conilera grampoides Gourrett, 1 89 1 : 1 1 , PI. 1 , fig. 7, PI. 3, figs 4-1 1 . Cirolana cranchi: Marine Biological Association, 1931 : 183; 1957 : 195; Ellis, 1981 : 123. Cirolana borealis: Clarke, 1971 : 103 (Non Natatolana &0ra2//s(Lilljeborg)). Part Cirolana cranchii, Nordenstam, 1946:3, Figs 1-5. [More than one species is involved inNordenstam's description.] Non Cirolana cranchii, Barnard, 1920 : 346; 1940 : 392, 49, 499, Fig. 66; Kensley, 1978 : 65, Fig. 27B,C.[= Cirolana vicina Barnard, 1914]. TYPE. The holotype is held by the British Museum (Natural History). Bull. Br. Mus. nat. Hist. (Zool.) 44(2): 75-84 Issued 24 February 1 983 76 N. L. BRUCE &J. ELLIS Fig. 1 Cirolana cranchi: (a)-(e), rf 17-0 mm, Polperro; (f) 9 13-3 mm, Polperro; remainder, d14-0 mm. Plymouth, (a) lateral view; (b) cephalon, dorsal view; (c) pleon and pleotelson, dorsalview; (d) pleon, lateral view; (e) clypeal region; (0 pleon and pleotelson; (g) pereopod 1 ; (h)pereopod 7; (i) antennal peduncle; (j) pereopod 1, propodus; (k) pereopod 2; (1) pereopod 2,dactylus; (m) antennule. Scale line represents 4-5 mm. CIROLANA CRANCHI 77 Fig. 2 Cirolana cranchi: all figs d 14-0 mm except (iHO 9 13-3 mm. (a) pleopod 1; (b) appendixmasculina, apex; (c) pleopod 2; (d) pleopod 3; (e) pleopod 2, medial margin of peduncle; (0pleopod 4; (g) pleopod 5; (h) uropod, ventral view; (i) uropodal exopod, ventral view; (j) uropod,dorsal view; (k) uropodal endopod, apex; (1 ) uropodal exopod, apex; (m) sternite 7; (n) penes. 78 N. L. BRUCE &J. ELLIS Fig. 3 Cirolana cranchi: all figs from cf 14-0 mm. (a) maxilliped; (b) maxillule; (c) right mandible;(d) antenna, flagellar articles 9-11; (e) antennule, flagellar articles 4-7; (f) left mandible, incisor;(g) mandibular palp; (h) maxilla. TYPE LOCALITY. Leach (1818) gives 'Grande Bretagne' as the source of his specimens. Ellis(1 98 1 ) records the locality as Falmouth, Cornwall. MATERIAL. 2cf (17-0, 13-3 mm), Polperro, Cornwall. Coll. A. M. Norman. BM(NH) Reg1911. 11. 8:7840-49. 2rf(14-0, 9-5 mm) 9 (12-6 mm), Plymouth, Devon. Coll. A. M.Norman. BM(NH) Reg 191 1. 11.8: 7828-30. <S (17-2 mm), 2 9 (13-0, 13-5 mm), Torquay,Devon. Coll. A. M. Norman. BM(NH) Reg 191 1. 1 1. 8: 7831-33. These specimens form partof those reported on by Norman (1904). DESCRIPTION OF MALE. Body about 2-75-3 times longer than wide. Cephalon without rostralprocess, interocular carina on anterior margin, dorsal interocular furrow extending from thedorsal medial margin of each eye; posterior margin of cephalon with groove on each sideindicating presence of maxillipedal somite. Pereonite 1 longest, with 2 horizontal furrows oneach side; pereonites 2-7 approximately subequal in length; coxal plates each with distinctcarina, posterior margins of coxae 5-7 straight, projecting beyond posterior margin ofsegment. Pleonite 1 entirely concealed by pereonite 7; pleonite 3 with posterolateral marginsmoderately produced, those of pleonite 4 rounded; dorsal surfaces of pleonites with posteriormargins minutely crenulate, pleonite 5 with additional small tubercles. Pleotelson slightlymore than 0-66 as long as greatest width, lateral margins sinuate, converging to narrowly CIROLANA CRANCHI 79 rounded apex; posterior margins densely setose, with about 21 spines; dorsal surface flatexcept for two oblique anterolateral ridges. Antennule peduncle Inarticulate, articles 1 and 2 appearing fused, although sutureevident; peduncular article 3 equal in length to combined lengths of articles 1 and 2;flagellum extends to posterior margin of eye. Antenna with peduncular articles 1 and 2 short,peduncular article 3 about half as long as 4, which in turn is half as long as 5; flagellumcomposed of about 38 articles, extends to pereonite 4. Frontal lamina pentagonal, apex not overlapped by rostral process, lateral margins slightlyconcave, diverging slightly, anterior margins straight; about 0-5 times as long as greatestwidth. Clypeus about 5-75 times wider than long. Mandibles with asymetrical incisors, thatof right mandible with 3 distinctly formed subequal cusps, that of left mandible withposterior cusp prominent, central cusp broad and shallow; molar process with about 25teeth, inferior distal margin setose, lacina mobilis with about 7 spines; mandibular palp withterminal article curved ventrolaterally, lateral margins with numerous stiff setae. Maxillulewith about 10 stout spines on gnathal surface of exopod, 3 robust plumose setae on proximalhalf of medial margin. Maxilliped with continuous marginal setae on palp articles 3-5,marginal setae on distal margins only of palp article 2; endite with 3 terminal and 3 lateralplumose setae, and with 2 coupling hooks. Pereopod 1 with slender spines at posterior distal angle of basis; ischium with 2 setae onposterior margin and 3 setae at anterior distal angle; merus with about 6 setae at anteriordistal angle, posterior margin with 3 acute spines and 5 tubercular submarginal spines;carpus with a single spine on posterior margin, set within a conspicuous indentation;propodus with 2 acute spines on palm, each spine set distally to tooth like projection, thirdrobust spine opposes dactylus; margin of propodus minutely denticulate between spines.Pereopods 2 and 3 similar, pereopod 2 with 3 acute spines at anterior distal angle of ischium,2 blunt spines at posterior distal angle, and third spine on the distal lateral margin; meruswith 5 spines at anterior distal angle, posterior margin bisinuate, with 8 stout spines; carpuswith a single stout spine and single seta on posterior distal angle, spine present on lateraldistal margin; propodus with 3 spines on palm, fourth spine opposing the dactylus; dactyluswith weakly developed but distinct secondary unguis, as in all pereopods. Pereopods 5-7similar, pereopod 4 intermediate in form between anterior (1-3) and posterior (5-7)pereopods. Pereopod 7 with about 2 setae and anterior distal angles of propodus, otherwisewithout setae; distal angles of ischium, merus and carpus each with a group of spines;posterior margin of ischium and produs with further 3 groups of 1-4 spines; posteriormargins of merus and carpus with further group of spines; propodus has spine opposingdactylus. Penes set together on posterior of sternite 7, separated from each other by about 0-05 thewidth of the sternite; penes are not robust, but rather lamellar flaps of cuticle which originateposteriorly, and project anteriorly, lying against sternite. Pleopods 3-5 with exopods with partial suture. Peduncles of pleopods 1-5 becomingprogressively shorter towards posterior, peduncle of pleopod 1 twice as wide as long,peduncle of pleopod 4 3-5 times as wide as long; lateral distal angles each with a single spine,medial margin 3-5 coupling hooks on pleopods 1-4, pleopod 5 without coupling hooks.Pleopod 1 with rami subequal in length, endopod with margins parallel, exopod with spineat proximal lateral angle. Pleopod 2 with endopod fractionally longer than exopod; appendixmasculina arises basally, extends beyond endopod by 0-1 of its length, narrows smoothly toan acute apex. Uropods extend distinctly beyond apex of pleotelson. Exopod slightly shorter thanendopod, lateral margin smoothly convex, with continuous marginal setae and about 9spines, distal half of lateral margin with dense mass of setae extending on to dorsal surface;medial margin with distinct angle half way along its length, distal half with 6 spines anddense mass of marginal setae; apex not bifid. Endopod with lateral margin angled at about0.33 of the way along its length, proximal 0-66 densely setose, setae extending onto dorsalsurface; medial margin convex with dense marginal setae and about 8 spines; apex not bifid. 80 N. L. BRUCE &J. ELLIS Fig. 4 Map showing the distribution of C. cranchi. Type locality of Nelocira swainsonii Leach, 18 18 given as Sicily. FEMALE. The only differences from the male are in the shape of the pleotelson which is wider,in the shape of the uropods which are not angled, and the lack of the dense setae on thepleotelson and uropods. DEVELOPMENT. Young males are similar to females, and the characteristic shape and setationof the uropods of large males is acquired gradually. One female has oostegites, and measured13-3 mm. CIROLANACRANCHI 81 VARIATION. From the specimens examined, it would appear that the minute crenulations onthe pleonites are not visible in large males. The presence and distribution of tubercles onthe dorsal surface of the pleonites is erratic. SIZE. Hansen (1905) records the largest specimen as a female of 18.0 mm. Museum materialhad adult males from 9-0-1 9-1 mm, females ranged from 9-6-1 9-2 mm. COLOUR. In alcohol, all a pale tan. Barrett and Yonge (1958) describe the colour as 'very palegrey, minutely dotted on first three segments; rear edge of first seven segments marked bytransverse line'. REMARKS. The shape and setation of the pleotelson and uropods are unique, andimmediately separate Cirolana cranchi from all other species of Cirolana. Other charactersuseful in separation include the shape of the frontal lamina, the shape of the posterolateralmargins of the pleonites, the length and shape of appendix masculina, the shape of theendopod of pleopod 1 , and the form of the penes. Cirolana cranchi can be separated from C. parva by the lack of a rostral point, by nothaving the frontal lamina overlapped by a rostral projection, and the very different shape,setation and spination of the pleotelson and uropods. DISTRIBUTION. Reliably recorded only from the eastern North Atlantic and Mediterranean.These records are summarized here together with new records from the collections of theBritish Museum. Fig. 4 illustrates the present distribution of Cirolana cranchi. Firth of Clyde: Cumbrae (Bate & Westwood, 1867); Fairland Point, Cumbrae (Robertson, 1888). Galway Bay, Eire: Spiddal (Ryland and Nelson-Smith, 1975); near Galway (Naylor, 1972); North Sound (Clark, 197 1 , as Cirolana borealis). Nymphe Bank: Ballycotton, County Cork, Eire (BM(NH) Collections). St George's Channel: Dale, Pembroke (Crothers, 1966). English Channel: Falmouth (BM(NH), holotype); Gwyllyn Vase, Falmouth (Stebbing, 1906); Polperro, Cornwall; Torquay (Norman, 1904); Plymouth, outside breakwater (Bate & Westwood, 1867); Batten, Mewstone Ledge, Tinside and Stoke Point, Plymouth (Marine Biological Association, 1957); Knapp Buoy, Plymouth (Heape, 1888); Torbay (BM(NH) collections); Anstis Cove, near Torquay (Stebbing, 1893); Jersey (Koehler, 1886); off St Sampson's Harbour, Guernsey (Norman, 1907); Roscoff(Delages, 1881). Bay of Biscay: Minou, Brittany (Hesse, 1866); Concarneau (Bonnier, 1887); Grands-Carneaux and east of Belle-Isle, Croisic (Chevreux, 1884); Le Croisic; Belle Isle; He d'Yeu; Guethery, near Biarritz (Hansen, 1905); Capbreton (Norman, 1904). Mediterranean: Toulon; Cannes; Porto Vecchio, Corsica; Gabes, Tunis (Hansen, 1905); Rade d'Hyeres (BM(NH) collections); Brusq, Var (Gourret, 1891, as Conilera grampoides); Villefranche (Hansen, 1890); Monaco (Monod, 1923); Sicily (Leach, 1818, as Nelocira swainsonii)', Alexandria (Larwood, 1940). Discussion Leach (1818) recorded Cirolana cranchi, the second species of what was to becomethe family Cirolanidae. The species then received little attention up to the revision ofthe family by Hansen (1890). In that publication and a later one Hansen (1905) reviewed allprevious records. Records published since Hansen's two publications have basically con-tributed little towards an increased knowledge of the species. Monod (1930) figured thepleopods, and Kussakin (1979) gave new figures for the antenna, pereopods 1 and 7 and themale second pleopod, his other figures being taken from Hansen (1890). From this it can beseen that although the figures given by Hansen (1890) are of a high standard, no moderndescription has been given. 82 N. L. BRUCE &J. ELLIS As a consequence of the lack of detailed description Cirolana cranchi has becomeconfused with Cirolana parva. The initiator of this confusion was Stebbing (1917)who considered C. cranchi and C. parva as likely to be synonymous. This opinionwas later followed by Nordenstam (1946) who went to some length to demonstratethat the two species were one. Unfortunately, he chose to do this by illustratingpereopod characteristics, the one character likely to lead to the conclusion to whichhe came. Pereopods in the genus Cirolana (sens. str. Bruce, 1981) vary very littlebetween species. Monod (1976) entered the argument in describing a Cirolana sp.from Togo, West Africa. In this paper Monod reviewed the arguments of previousauthors, and discussed Hansen's (1890) diagnoses. Monod came to the conclusionthat the West African species could be assigned neither to cranchi nor to parva andnor could they be said not to belong to those species. In effect, that parva and cranchiare not separable. When reviewing this debate, the most surprising aspect is that the problem existedat all. Hansen's (1890) figures clearly separate the two species. The differences infrontal lamina shape, pleotelson and uropods, presence and absence of rostral process are allclearly shown. Hansen also states that in parva the uropod apices are bifid, and in cranchientire. It also seems remarkable that no author encountering this problem has sought toredescribe the species involved as a solution. Comparison of the figures, description andremarks given here to those of Bruce & Bowman (1982) show that Cirolana cranchi andCirolana parva are two readily separable species. As there has been some confusion over the species included under the synonymiesof Cirolana cranchi, the most important of these are now listed and discussed. Cirolana swainsonii: Miers, 1881. Examination of Miers' specimens shows that they represent a species related to, but distinct from cranchi that has yet to be described. Cirolana vicina Barnard, 1914. This species is very similar to C. cranchi and should be redescribed before final judgement on its status is passed. From Barnard's (1914) description it differs in having 'sub-bifid' uropod apices, and lacks the dense mass of setae on the uropods. Cirolana cranchii var. australiense Hale, 1925. This species is in no way a race or variety of C. cranchi. It differs in the shape of the frontal lamina, pleotelson and posterolateral margin ofpleonites2-4, and has a distinct rostral process. Cirolana cranchii: Nordenstam, 1946. Nordenstam's material came from Europe, South Africa and the Pacific, and consisted of C. cranchi., C. vicina and a species of unknown identity from the Pacific. Cirolana australiense Naylor, 1961. This species is closest to the variety described by Hale (1925), but may be a distinct species. Naylor's specimens were from the Chatham Islands, New Zealand. Cirolana sp.: Monod, 1976. Monod figures two 'Cirolana sp', one from Togo, and one from the Congo. One of these is Cirolana chaloti Bouvier, 1901 (Bruce, in press), and the other species may well be new. Monod's (1931) record of C. cranchi may be of one of these species. In conclusion, it should be emphasized that in identifying or describing speciesrelated to C. parva and C. cranchi particular care and attention should be given todetails of the frontal lamina, the relative position and shape of the penes, the shapeof the posterolateral margins of pleonites 2-4, the first and second pleopods of themale, and most importantly details of the pleotelson and uropods. Acknowledgements We are grateful to Dr R. J. Lincoln for his comments on the manuscript and for the loan ofspecimens to one of us (N. L. B.). This study was supported by a CommonwealthPostgraduate Study Award to N. L. Bruce. CIROLANA CRANCHI 83 References Barnard, K. H. 1914. Contributions to the crustacean fauna of South Africa. 3. Additions to the marineIsopoda, with notes on some previously incompletely known species. Ann. S. Afr. Mus. 10 :325a-358a, 359-442. 1920. Contributions to the crustacean fauna of South Africa. 6. Further additions to the list of marine Isopoda. Ann. S. Afr. Mus. 17 : 319-348. 1940. Contributions to the crustacean fauna of South Africa. 12. Further additions to the Tanaidacea, Isopoda and Amphipoda, together with keys for the identification of hitherto recorded marine and freshwater species. Ann. S. Afr. Mus. 32 : 38 1-5 15. Barrett, J. & Yonge, C. M. 1958. Pocket guide to the Sea Shore. London (Collins), 272 pp.Bate, C. S. & Westwood, J. O. 1867. A history of British sessile-eyed Crustacea 2 (16-19): 209-400. London (John van Voorst.)Bonnier, J. 1887. Catalogue des crustaces malacostraces recueillis dans la Baie de Concarneau. Bull. scient. Dep. AW (2)10: 1-190.Bruce, N. L. 1981. Cirolanidae (Crustacea: Isopoda) of Australia: Diagnoses of Cirolana Leach, Metacirolana Nierstrasz, Neocirolana Hale, Anopsilana Paulian & Debouteville, and three new genera-Natatolana, Politolanaand Cartetolana. Aust. J. mar. Freshwat. Res. 32: 945-966.in press. Cirolana chaloti Bouvier (Isopoda, Cirolanidae), a long overlooked species from West Africa. Bull. Mus. natn. Hist. nat. Paris. & Bowman, T. E. 1982. The status of Cirolana parva Hansen, 1890 (Crustacea: Isopoda: Cirolanidae) with notes on its distribution. Proc. biol. Soc. Wash. 95 : 325-333.Chevreux, E. 1884. Crustaces amphipodes et isopodes des environs du Croisic. C. r. Ass. fr. Avanc. Sci. 12:517-520.Clark, F. L. 1971. Cirolana borealis Liljb. (Isopoda) on living fish from Galway Bay. Ir. Nat. J. 17: 103.Crothers, J. H. (ed.) 1966. Dale Fort marine fauna (2nd edition). London (Field Studies Council), 169pp.Delages, Y. 1881. Catalogue des crustaces edriophthalmes et podophthalmes qui habitent les plages de RoscofT. Archs Zool. exp. gen. 9 : 1 52-1 58.Desmarest, A.-G. 1 825. Considerations generates sur la classe des Crustaces, et description des especes de ces animaux, qui vivent dans la mer sur les cotes, ou dans les eaux douces de la France. Paris (F. G. Levrault), 446 pp.Ellis, J. 1981. Some type specimens of Isopoda (Flabellifera) in the British Museum (Natural History), and the isopods in the Linnaean Collection. Bull. Br. Mus. nat. Hist. (Zool)40 : 121-128.Gosse, P. H. 1855. A manual of marine zoology for the British Isles Part 1 . London (John van Voorst), 203 pp.Gourret, P. 1891. Les lemodipodes et les isopodes du Golfe de Marseille. Annls Mus. Hist. nat. Marseille. (Zool.) 4 Travaux scientifiques. Memoire No. 1, 44 pp.Hale, H. M. 1925. Review of Australian isopods of the cymothoid group. Part 1. Trans. R. Soc. S. Aust. 49: 128-185.Hansen, H. J. 1890. Cirolanidae et familiae nonnullae propinquae Musei Hauniensis. K. dansk. Vidensk. Selsk. Skr. 5 : 239-426.1905. Revision of the European marine forms of the Cirolaninae, a subfamily of Crustacea Isopoda. J. Linn. Soc. (Zool.) 29 : 337-373. Harger, O. 1880. Notes on New England Isopoda. Proc. U.S. natn. Mus. 2 : 297-462.Heape, W. 1888. Preliminary report upon the fauna and flora of Plymouth Sound. J. mar. biol. Ass. U.K.I: 153-193.Hesse, M. 1 866. Observations biologiques sur quelques crustaces des cotes de Bretagne. Annls Sci. nat. (Zool.) (5) 5: 24 1-264.Kensley, B. 1978. Guide to the marine isopods of southern Africa. Cape Town (South African Museum), 173 pp.Koehler, R. 1 886. Contribution a 1'etude de la faune littorale des lies Anglo-Normandes. Annls Sci. nat. (Zool.) (6) 20, Art. 4. 62pp.Kussakin, O. G. 1979. Marine and brackish water isopod Crustacea. Suborder Flabellifera. Leningrad (Academy of Science, U.S.S.R.), 470 pp. (In Russian.)Larwood, H. J. 1940. The fishery grounds near Alexandria. XXI. Tanaidacea and Isopoda. Notes Mem. Fouad I Inst. Hydrobiol. Fish. 35, 72 pp.Leach, W. E. 1818. Cymothoadees In: Cuvier, F. (ed.) Dictionnaire des sciences naturelles. Paris & Strasbourg (Leurault), Vol. 12: 338-354. 84 N. L. BRUCE &J. ELLIS Marine Biological Association. 1931. Plymouth Marine Fauna 2nd edition. Plymouth: 371 pp. 1957. Plymouth Marine Fauna 3rd edition. Plymouth: 457 pp. Miers, E. J. 1881. On a collection of Crustacea made by Baron Hermann-Maltzan at Goree Island, Senegambia. Ann. Mag. nat. Hist. (5)8 : 364-377. Milne-Edwards, H. 1840. Histoire Naturelledes Crustaces. Paris (Roret), Vol. 3, 638 pp.Monod, T. 1923. Notes carcinologique (Parasites et commensaux). Bull. Inst. oceanogr. Monaco No. 427,23pp. 1930 Contribution a Petude des 'Cirolanidae'. Annls Sci. nat. (Zool.) (10) 13 : 129-183. 193 1. Sur quelques crustaces aquatiques d'Afrique (Cameroun et Congo). Revue Zool. Bot. afr. 21 1976. Remarques sur quelques cirolanides (Crustaces, Isopodes). Bull. Mus. natn. Hist. nat. Paris (Zool.) No. 251 : 133-161.Naylor, E. 1961. Some Isopoda from the Chatham Is., including two species of Cirolana new to New Zealand. Trans. R. Soc. N.Z. (Zool.) 1 : 7-17. 1972. British Marine Isopods. Synopses Br. Fauna (N.S.) No. 3, 86 pp. Nordenstam, A. 1946. Marine Isopoda from Prof. Sixten Bock's Pacific Expedition 1917-1918. Ark. Zool. 37 A (7): 1-31.Norman, A. M. 1904. British Isopoda of the families Aegidae, Idoteidae, and Arcturidae. Ann. Mag. nat. Hist. (7) 14 : 430-448. 1907. Notes on the Crustacea of the Channel Islands. Ann. Mag. nat. Hist. (7) 20 : 356-371. & Scott, T. 1906. The Crustacea of Devon and Cornwall. London (William Wesley & Son), 232 pp.Robertson, D. 1888. A contribution towards a catalogue of the Amphipoda and Isopoda of the Firth of Clyde. Proc. nat. Hist. Soc. Glasgow (N.S.) 2 : 9-99.Ryland, J. S. & Nelson-Smith, A. 1975. Littoral and benthic investigations on the west coast of Ireland. IV. (Section A: Faunistic and ecological studies.) Some shores in Counties Clare and Galway. Proc. R. Ir. Acad. 75 Sect. B: 245-266.Stebbing, T. R. R. 1893. A history of Crustacea. London (Keegan Paul, Trench, Trubner & Co. Ltd.), 466 pp.1906. Crustacea. In: Page, W. (ed.) The Victoria history of the Counties of England. A history of the County of Cornwall. 1 : 255-289.1917. South African Crustacea (Part IX of S.A. Crustacea, for the marine investigations in South Africa). Ann. S. Afr. Mus. 17 : 23-46. Manuscript accepted for publication 14 June 1982 Valettietta, a new genus of deep-sea amphipod(Gammaridea: Lysianassidae) with descriptions oftwo new species from the North Atlantic Ocean Roger J. Lincoln Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Michael H. Thurston Institute of Oceanographic Sciences, Brook Road, Wormley, Godalming, Surrey GU8 5UB Introduction Within the gammaridean family Lysianassidae the combination of a strongly toothedmandibular incisor and unspecialized gnathopods is shared by only 4 genera, ValettiaStebbing, 1888, Alicella Chevreux, 1899, Valettiopsis Holmes, 1908 and ValettiellaGriffiths, 1977. Each is monotypic with the exception of Valettiopsis which comprises 4species, dentata Holmes, 1908, macrodactyla Chevreux, 1909, anacantha Birstein &Vinogradov, 1963 and multidentata Barnard, 1961. Barnard (1969) included one othergenus, Onesimoides Stebbing, 1888, in his key to lysianassids having a toothed incisor, butthere is some doubt as to the true nature of the mandibular margin. Stebbing (1888), in hisdescription of O. carinatus says '. . . cutting edge . . . seemingly of the usual form . . .'. In O.cavimanus, the mandible has '. . . bord tranchant presque simple, renforce a chacun de sesangles par des bourrelets de chitine. . . .' (Pirlot, 1933), while in O. chelatus the cutting edgeis simple (Pirlot, 1933). Those genera with a strongly dentate incisor share a broadly similar facies and can beregarded as forming a natural group, with the omission of Valettia coheres which has aquite different morphology, especially in the structure of the mouthparts. The choice ofnames for these genera has turned out to be rather unfortunate since the genus name Valettiawas used as the stem term for the later taxa Valettiopsis and Valettiella with which it does nothave close affinity. Valettiopsis and its allies live at moderate to great ocean depths and have as a conse-quence been infrequently recorded. In fact, all six species were first described from uniquetypes, and only dentata and macrodactyla have since been redescribed from additionalmaterial (Barnard, 1967; Chevreux, 1935). Recent deep-water collections from the NorthAtlantic made during cruises of RRS Discovery and RRS Challenger have produced 5mature individuals of this rare lysianassid group, one belonging to Valettiopsismacrodactyla, the other 4 representing two species new to science. The combination ofcharacters shared by the two new species puts them close to Valettiopsis, but withsufficient disparity in the configuration of the coxal plates, pereopodal bases, andmandibular palp armature to justify the erection of a new genus for which we propose thename Valettietta gen. nov. One existing species of Valettiopsis, namely V. anacantha fromthe Philippine Trench in the Pacific, is transferred to the new genus. SystematicsFamily LYSIANASSIDAE Genus VALETTIOPSIS Holmes, 1908 DIAGNOSIS. Body robust, compressed, pleosome well developed; urosome segment 1 with Bull. Br. Mus. not. Hist. (Zool.)44(2): 85-101 Issued 24 February 1983 86 R. J. LINCOLN & M. H. THURSTON large, acute, mid-dorsal tooth, segment 3 broad and dorsally flattened with lateral marginsraised. Antenna 1 and 2 elongate, slender, subequal length, peduncle articles 2-3 of antenna1 compressed, flagellum article 1 conjoint, accessory flagellum well developed, multi-articulate. Upper lip weakly notched; lower lip without inner lobes, mandibular lobeselongate. Mandible having robust incisor, strong spine row interspersed with plumose setae,and large triturative molar; palp attached level with molar, article 2 elongate with onlydistomarginal setae. Maxilla 1 inner plate densely setose along entire inner margin, palprobust, 2-articulate. Maxilla 2 inner and outer plates subequal length, inner plate with densemediodistal and facial setae. Maxilliped basic, outer plate with short inner marginal spinesgrading distally to robust elongate spines. Coxal plate 1 much shorter than 2 and partly con-cealed; plate 4 with only shallow posterior emargination. Coxal plate 5 anterior lobe deeperthan posterior lobe. Epimeral plate 2 distal angle with tooth. Gnathopods 1 and 2 subchelate;gnathopod 1 palm transverse. Pereopod 7 basis expanded, lacking posterodistal lobe.Uropods biramous, lanceolate, spinose; uropod 3 outer ramus 2-articulate. Telson deeplycleft, each lobe with several large apical spines. Branchial lobes bearing small accessory lobeat the base. TYPE SPECIES. Valettiopsis dentata Holmes, 1908 (original designation). REMARKS. Three of the Valettiopsis species names (dentata, anacantha, multidentata) havebeen corrected to give the epithets feminine terminations. Holmes (1908) derived the nameValettiopsis '. . .from Valettia, a genus of Amphipods, and 6vj/is, appearance'. Greek nounsending in ops, genitive opsis, are feminine and adjectival specific names must agree in genderwith their genus name (International Code of Zoological Nomenclature, Articles 1 1 (g) (i)(D,34(b)). Valettiopsis macrodactyla ChevreuxValettiopsis macrodactyla Chevreux, 1909 p. 1 , figs 1-2; 1935 p. 8, pi. 2, fig. 1 . MATERIAL EXAMINED. 1 rf Bay of Biscay abyssal plain, about 47 15'-28'N 8 9'^46'W; 4300metres; collected by Dr A. G. Macdonald during RRS Challenger cruise, October, 1978,using baited trap. BM(NH) reg. no. 1979 : 8 : 1. DESCRIPTION. Figs la-h; 2a-h; 3a-k. Length 17 mm. Body robust, compressed, pleosomesegments strongly developed; urosome segment 1 with prominent dorsal tooth, upper marginof tooth weakly sinous, apex acute. Epimeral plates 2-3 (Fig. 2h) with posterodistal tooth.Head (Fig. Ib) with triangular lateral lobe apically produced and with sinuous lower margin;postantennal sinus very shallow; eyes absent. Antenna 1 elongate, peduncle article 1 slender,longer than 2-3 combined; flagellum 30-articulate; accessory flagellum 9-articulate, reachingslightly beyond end of basal conjoint article of flagellum; conjoint article of flagellum equalto length of peduncle, densely setose on inner surface; remaining flagellar articles sparselysetose. Antenna 2 longer than 1 , peduncle article 5 slightly longer and more slender than 4,flagellum 40-articulate, proximal flagellar articles with erect setules on inner margin. Upperlip (Fig. 2a) rounded with minute apical notch. Lower lip (Fig. 2b) outer lobes elongate,robustly spinulose on inner distal margin; inner lobes absent, mandibular lobes welldeveloped. Right mandible (Fig. 2c, d), incisor robustly 7-dentate, lacinia also stronglytoothed, comprising two plates with 7 and 4 teeth; spine row with 1 1 large spinesinterspersed with long plumose setae, distal spines dentate; molar strongly triturative; palprobust, article 3 oval with inner distal margin robustly setose, article 2 elongate, inner distalmargin with long setae, inner proximal margin naked. Maxilla 1 (Fig. 2e) inner plate withentire inner margin densely setose; palp large, article 2 distal margin with stout short spinesand row of submarginal setae. Maxilla 2 (Fig. 20 inner and outer plates subequal, denselysetose, inner plate also with row of facial setae. Maxilliped (Fig. 2g) inner plate with 3 shortapical spines; outer plate inner margin bearing row of spines that are short and stoutproximally grading to elongate and plumose distally; palp elongate, article 3 with row ofstrong facial setae, article 4 inner margin bearing 2 short spines. VALETTIETTA GEN. NOV. 87 Fig. 1 Valettiopsis macrodactyla Chevreux. Male, a, habitus; b, head and antennae; c, gnathopod1; d, gnathopod 1, palmar region; e, coxal plate 1, anterodistal margin; f, gnathopod 2; g,gnathopod 2, palmar region; h, coxal plate 2, posterodistal margin. Bar scales: a, b, 2-0 mm; c, f,1 -0 mm; d, e, g, h, 0-2 mm. 88 R. J. LINCOLN & M. H. THURSTON Fig. 2 Valettiopsis macrodactyla Chevreux. Male, a, upper lip; b, lower lip; c, right mandible; d,lacinia mobilis and spine row, right mandible; e, maxilla 1 ; f, maxilla 2; g, maxilliped; h, pleon.Bar scales: a-g, 0-5 mm; h, 2-0 mm. VALETTIETTA GEN. NOV. 89 Gnathopod 1 (Fig. 1 c, d) subchelate, coxal plate short, triangular, apically rounded, antero-distal margin with small tooth (Fig. le), distal margin setulose; basis with long setae onanterior and posterior margins; ischium elongate; merus short with mat of short setules onposterior margin; carpus little longer than propodus, setose on posterior margin; propodusrectangular, strongly setose, palm transverse (Fig. Id), delimited by group of stout spines,palmar margin minutely toothed; dactylus short, slightly overlapping palm, inner marginwith small tooth. Gnathopod 2 (Fig. If, g) subchelate; coxal plate rectangular, posterodistalmargin with 2 small teeth (Fig. Ih); basis with many long setae on anterior and posteriormargins; ischium elongate; merus small; carpus equal to length of propodus, posteriormargin densely setose; propodus robust, densely setose, palm oblique (Fig. Ig) convex,delimited by group of large dentate spines, palmar margin smooth; dactylus stout with smalltooth on inner margin. Pereopod 3 (Fig. 3a), coxal plate rectangular, distal margin weaklysinuous, posterodistal angle with 2 small teeth (Fig. 3b); basis curved, merus robust; carpusshorter than propodus. Pereopod 4 (Fig. 3c) similar to 3, except coxal plate with broadshallow posterior emargination, distal margin without tooth. Pereopods 5-7 (Figs 3d, e, f)robust, spinose; basis with tapering posterior lobe, distal angle weakly produced on 5, notproduced on 6-7, bearing 1-2 slender submarginal spines, posterior margin weakly serrate,distal margin of basal lobe on pereopod 7 bevelled. Uropods biramous, spinose; uropod 1(Fig. 3g) rami subequal, inner margins of rami minutely serrate; uropod 2 (Fig. 3h) outerramus little shorter than inner; uropod 3 (Fig. 3i) distal article of outer ramus about one-thirdlength of proximal article, inner margin of inner ramus setose. Telson (Fig. 3j) elongatetriangular, cleft to three-quarters length, inner apical margin rounded, outer apical angle(Fig. 3k) with 4 graduated spines. DISTRIBUTION. Known only from the North Atlantic in the region of the Azores (Chevreux,1 935) at 1 692-1 919m, and from the present Biscay record at 4300 m. VALETTIETTA gen. nov. DIAGNOSIS. Body robust, compressed; pleosome well developed; urosome segment 1 withweak dorsal process, segment 3 broad and flattened dorsally with lateral margins raised.Antenna 1 and 2 elongate, slender, about equal length; peduncle articles 2-3 of antenna 1compressed, flagellum article 1 conjoint, accessory flagellum well developed, multi-articulate. Upper lip weakly notched. Lower lip without inner lobes, mandibular lobesprominent. Mandible with robustly dentate incisor, spine row strong, interspersed withplumose setae, molar large and triturative, palp attached level with molar, article 2 elongatewith proximal and distal margin setose. Maxilla 1 inner plate densely setose along entireinner margin, palp robust, 2-articulate. Maxilla 2 inner and outer plates subequal length,inner plate with dense mediodistal and facial setae. Maxilliped basic; outer plate with shortstout inner marginal spines grading distally to robust elongate spines. Coxal plates 1-4forming continuous series; plate 4 with deep posterior emargination. Coxal plate 5 anteriorlobe not deeper than posterior lobe. Gnathopod 1 subchelate; palm oblique; gnathopod 2subchelate or simple. Pereopods 5-7 basis expanded with prolonged rounded posterodistallobe. Uropods biramous, lanceolate, spinose. Telson triangular, deeply cleft. Branchial lobeswith small accessory lobe close to base. TYPE SPECIES. Valettietta lobata sp. nov. ETYMOLOGY. The affinity of the new genus to Valettiopsis is recognized by adding thediminutive ending -etta to the common stem. Gender feminine. Valettietta lobata sp. nov. MATERIAL EXAMINED. Holotype d, Bay of Biscay abyssal plain, about 4715'-28'N89'^6'W; 4300 metres; collected by Dr A. G. Macdonald during RRS Challenger cruise1980, using baited trap. BM(NH) reg. no. 1982 : 204. 90 R. J. LINCOLN & M. H. THURSTON Fig. 3 Valettiopsis macrodactyla Chevreux. Male, a, pereopod 3; b, coxal plate 3, posterodistalmargin; c, pereopod 4; d, pereopod 5; e, pereopod 6; f, pereopod 7; g, uropod 1 ; h, uropod 2; i,uropod 3; j, telson; k, apex of telson lobe. Bar scales: a-f, 1 -0 mm; g-j, 0-5 mm. VALETTIETTA GEN. NOV. 91 Fig. 4 Valettietta lobata sp. nov. Holotype. a, habitus; b, head and antennae; c, gnathopod 1 ; d,gnathopod 1, palmar region; e, gnathopod 2; f, gnathopod 2, palmar region; g, coxal plate 2,posterodistal margin. Bar scales: a, 5-0 mm; b, 2-0 mm; c, e, 1 -0 mm; d, f, g, 0-2 mm. 92 R. J. LINCOLN & M. H. THURSTON Fig. 5 Valetlietta lobata sp. nov. Holotype, a, left mandible; b, left mandible, incisor, lacinia andspine row; c, right mandible, incisor, lacinia and spine row; d, upper lip; e, lower lip; f, maxilla 1 ;g, maxilla 2; h, maxilliped; i, pleon. Bar scales: a, d-h, 0-5 mm; b, c, 0-2 mm; i, 2-0 mm. VALETTIETTA GEN. NOV. 93 Paratype d 1 , Discovery station 9541 # 19, north west of Cape Verde Islands, RMT 1 +8combination net, 18 April 1977, 4040-3970 m (fished 0-20 metres off bottom); 2019-T N2151-3' W-20184' N2140-5' W. BM(NH) reg. no. 1982 : 205. ETYMOLOGY. The epithet refers to the expansive posterior lobe of the pereopod basis. DESCRIPTION. Figs 4a-g; 5a-i; 6a-m. Holotype. Length 21-5 mm. Body robust, compressed,pleosome segments strongly developed; urosome segment 1 with small rounded medianknob-like process bearing pair of minute apical spinules. Epimeral plate 2 (Fig. 5i) distalangle quadrate, plate 3 distal angle acute. Head (Fig. 4b) large, lateral lobe triangular,postantennal sinus very shallow; eyes absent. Antenna 1 elongate, peduncle article 1 slender,posterodistal angle with small tooth, articles 2-3 short, flagellum 34-articulate; accessoryflagellum 10-articulate, reaching little beyond end of basal conjoint article of flagellum;conjoint article 1 equal to length of peduncle, densely setose on inner surface, remainingflagellar articles sparsely setose. Antenna 2 little longer than 1, peduncle articles 4-5subequal length, flagellum 48-articulate, proximal articles with erect setules on posteriormargin. Upper lip (Fig. 5d) asymmetrically rounded with small apical notch, distal surfaceminutely setulose. Lower lip (Fig. 5e) outer lobes elongate, robustly spinulose on inner distalmargin, inner lobes absent, mandibular lobes elongate. Mandible (Fig. 5a, b, c), left incisorrobustly 8-dentate and closely applied to 6-dentate lacinia; right incisor 7-dentate, wellspaced from double bladed lacinia bearing numerous small teeth; spine row with 12-13 largespines interspersed with long plumose setae, distal spines dentate; molar strongly triturative;palp robust, article 3 oval with inner margin spinose, article 2 elongate with regular row oflong inner distal setae and irregular groups of shorter proximal setae. Maxilla 1 (Fig. 50 innerplate with entire inner margin densely setose, outer plate with 2 rows (6 and 5) of pectinatespines; palp large, article 2 distal margin with stout spines and row of submarginal setae.Maxilla 2 (Fig. 5g) inner and outer plates subequal, distally setose, inner plate also with rowof facial setae. Maxilliped (Fig. 5h) inner plate with 3 short apical spines, outer plate innermargin with row of short stout spines becoming gradually more elongated and plumosedistally; palp elongate, article 3 with row of strong facial setae, article 4 inner margin with 2small teeth. Gnathopod 1 (Fig. 4c, d) subchelate; coxal plate rounded, distal margin withsmall tooth, inner distal surface with about 8 groups of setae; basis short, anterior andposterior mid-margins with long setae; ischium elongate, setose; merus short, margin withmat of short setules; carpus shorter than propodus and robustly setose; propodus sub-rectangular, robust, densely setose, palm weakly oblique (Fig. 4d), smooth, delimited byfan-like group of 6 spines (3 on inner face, 3 on outer face); dactylus short, just reaching toend of palm. Gnathopod 2 (Fig. 4e, subchelate, longer and more slender than 1 ; coxal platesubrectangular, posterodistal margin with 2 small teeth (Fig. 4g), inner distal surface withabout 6 groups of 2-5 setae; basis slender, anterior and posterior margins setose; ischiumelongate; merus small; carpus equal to length of propodus, densely setose; propodus sub-rectangular, setose, palm weakly oblique (Fig. 4f) convex, delimited by fan-like group of 6spines; dactylus short, not reaching end of palm; gnathopod 2 with characteristic twist inappendage between basis and merus that reverses the orientation of distal articles. Pereopod3 (Fig. 6a), coxal plate slender, rectangular, posterodistal angle with 2 small teeth (Fig. 6b);basis curved; merus elongate slender and much longer than carpus; dactylus small, straight.Pereopod 4 (Fig. 6c) similar to 3, except coxal plate with broad moderately deep posterioremargination, posterodistal margin with small notch (Fig. 6d). Pereopods 5-7 (Fig. 6e, f, g),robust, spinulose; basis broadly expanded with large posterodistal lobe that becomes moreangular from 5 to 7, posterior margin finely serrate; coxal plate of pereopod 6 with posteriorlobe distinctly angular. Uropods biramous; uropod 1 (Fig. 6h) rami subequal, robustlyspinose; uropod 2 (Fig. 6i) inner ramus just shorter than outer, spinose, adjacent marginsminutely serrate; uropod 3 (Fig. 6j) distal article of outer ramus about one-third length ofproximal article, apex with pair of minute inset setules (Fig. 6k). Telson elongate triangular(Fig. 61), cleft to three-quarters length, inner apical margin acute, outer angle with 2 spinesset in groove (Fig. 6m); dorsal surface of telson flattened with lateral margins downturned. 94 R. J. LINCOLN & M. H. THURSTON Fig. 6 Valettietta lobata sp. nov. Holotype. a, pereopod 3; b, coxal plate 3, posterodistal margin;c, pereopod 4; d, coxal plate 4, posterodistal margin; e, pereopod 5; f, pereopod 6; g, pereopod 7;h, uropod 1; i, uropod 2; j, uropod 3; k, apex inner ramus of uropod 3; 1, telson; m, apex oftelson lobe. Bar scales: a, c, e-j, 1,1-0 mm; b, d, 0- 1 mm. VALETTIETTA GEN. NOV. 95 Fig. 7 Valettietta gracilis sp. nov. Holotype. a, habitus; b, gnathopod 1 ; c, gnathopod 1 palmarmargin; d, gnathopod 2; e, gnathopod 2 propodus, distal margin; f, urosome, telson. Bar scales: a,2-0 mm; b, d, f, 0-5 mm; c, e, 0- 1 mm. 96 R. J. LINCOLN & M. H. THURSTON REMARKS. The paratype male, body length 18 mm, has essentially similar morphology to theholotype; antenna 1 flagellum 25-articulate with especially obvious tooth on pedunclearticle 1; antenna 2 flagellum 36-articulate bearing calceoli. The structure of the calceolus(Fig. 10) was examined by SEM and was found to approximate to the basic lysianassid formdescribed by Lincoln & Hurley (1981), but with certain unique features that set it apart fromall other lysianassid calceoli described to date. These include the 4 broad crescentic platesthat form the distal element and the lack of a cuticular pit at the point of origin of the stalkfrom the flagellar article. Valettietta gracilis sp. nov. MATERIAL EXAMINED. Holotype rf, Bay of Biscay abyssal plain, about 4715'-28'N89'^6'W; 4300 metres; collected by Dr A. G. Macdonald during RRS Challenger cruise1980, using baited trap. BM(NH) reg. no. 1982 : 206. Paratype cf, Discovery Station 9541 * 19 north west of Cape Verde Islands, RMT 1 +8combination net, 18 April 1977, 4040-3970 m (fished 0-20 m off bottom); 2019-T N2151-3' W-2018-4'N2140-5' W. BM(NH) reg. no. 1982 : 207. ETYMOLOGY. The epithet alludes to the slender condition of the gnathopod 2 propodus. DESCRIPTION. Figs 7a-f; 8a-h; 9a-h. Holotype. Length 1 1 mm. Body compressed, pleosomesegments well developed; urosome segment 1 with rounded median process. Epimeral plates1-2 posterodistal angle obtuse (Fig. 8h), plate 3 acutely produced. Head large (Fig. 8a),lateral lobes triangular, apically rounded; postantennal sinus very shallow; eyes absent.Antenna 1 elongate, flagellum 3 1 -articulate, sparsely setose; accessory flagellum 7-articulate,reaching beyond end of basal conjoint article of flagellum; conjoint article equal to length ofpeduncle article 1 , densely setose on inner surface. Antenna 2 little shorter than 1 , pedunclearticles 4 and 5 subequal length, flagellum 30-articulate, proximal flagellar articles with erectsetules on posterior margin. Upper lip asymmetrically rounded with small apical notch,distal surface minutely setulose. Lower lip (Fig. 8b) outer lobes elongate, inner distal marginrobustly spinulose, inner lobes absent, mandibular lobes elongate. Mandible (Fig. 8c, d), leftincisor strongly 8-dentate and closely applied to 7-dentate lacinia; right incisor 8-dentate,well spaced from 6-dentate lacinia; spine row with 13 large spines interspersed with longplumose setae, distal spines dentate, molar strongly triturative; palp robust, article 1 small,article 2 extremely elongate with regular row of inner distal setae and irregular groups ofproximal setae, article 3 oval with robust marginal setae. Maxilla 1 (Fig. 8e, inner platesetose along entire inner margin, outer plate bearing two rows of pectinate spines; palp large,article 2 distal margin with stout spines and row of long submarginal setae. Maxilla 2 innerand outer plates subequal, distally setose, inner plate also with row of facial setae. Maxillipedinner plate with 3 short apical spines; outer plate inner margin with row of short stout spines(Fig. 8g) becoming gradually elongate and plumose distally; palp elongate, setose.Gnathopod 1 (Fig. 7b, c) subchelate; coxal plate rectangular, anterior margin angular, distalmargin setose; ischium long and setose; merus small, posterior margin with mat of shortsetules; carpus much shorter than propodus, posterior margin densely setose; propoduselongate, tapering distally, anterior and posterior margins with long setae; palm oblique (Fig.7c) convex, dentate, delimited by group of short spines; dactylus overlapping palm, innermargin toothed. Gnathopod 2 simple (Fig. 7d, e), coxal plate rectangular, distal marginsetose, smooth; basis curved, anterior and posterior margins setose; ischium extremelyelongate; merus small; carpus slender, sparsely setose; propodus slender, tapering distally,margin with groups of long setae that curve inwards to form a setal basket, palm absent butpropodal margin with solitary spine close to the closing point of the dactylus. Gnathopod 2with characteristic twist in appendage between basis and merus that reverses the orientationof the distal articles. Pereopod 3 (Fig. 9a) coxal plate rectangular, distal margin setose; basiscurved; merus longer than carpus; propodus and carpus subequal length; all articles withlong marginal spinules. Pereopod 4 (Fig. 9b) similar to 3, except coxal plate very broad and VALETTIETTA GEN. NOV. 97 Fig. 8 Valettietta gracilis sp. nov. Holotype. a, head and antennae; b, lower lip; c, left mandible;d, right mandible; e, maxilla 1 ; f, maxilla 1 palp; g, maxilliped outer plate; h, pleon. Bar scales: a,h, 1-0 mm;b-g, 0-2 mm. 98 R. J. LINCOLN & M. H. THURSTON Fig. 9 Valettietta gracilis sp. nov. Holotype. a, pereopod 3; b, pereopod 4; c, pereopod 5; d,pereopod 6; e, pereopod 7; f, uropod 1; g, uropod 2; h, uropod 3. Bar scales: a-e, 1 -0 mm; f-h, 0-5 mm. Fig. 10 Valcttietta lobata sp. nov. Scanning electron micrographs of antennal calceoli; a, b, dsectional photographs of same calceolus, bar scale: 5-0 (im; c, entire calceolus, bar scale 10 |im. 1 00 R. J. LINCOLN & M. H. THURSTON deeply excavate posteriorly, distal margin straight and setose; merus and propodussubequal and longer than carpus. Pereopods 5-7 (Fig. 9c, d, e) robust, spinulose; basisbroadly expanded with large posterodistal lobe that becomes more angular from 5 to 7,posterior margin minutely serrate; coxal plate of pereopod 6 with subangular posterior lobe.Uropods biramous; uropod 1 (Fig. 90 and uropod 2 (Fig. 9g) inner ramus shorter than outer,spinose, adjacent margins minutely serrate; uropod 3 (Fig. 9h) distal article of outer ramustwo-thirds length of proximal article, apex of proximal article with triangular tooth, innermargin of inner ramus setose, apex of both rami with inset small setule. Telson (Fig. 70triangular, cleft beyond three-quarters length, apex acute with small spinule. DISCUSSION. The new genus Valettietta shares a general appearance and many specialfeatures with Valettiopsis, but is characterized by the following combination of characters:fully developed coxal plate 1; obtuse epimeral plate 2; produced posterodistal lobes onpereopods 5-7 bases; groups of proximal setae on article 2 of mandibular palp. The speciesanacantha described by Birstein & Vinogradov from a deep-sea station in the Pacific south ofthe Philippines is transferred to Valettietta; it is very close to gracilis but can be distinguishedby the rounded shape of coxal plate 4 and the presence of a short palm on the propodus ofgnathopod2. Key to species of Valettiopsis and Valettietta gen. nov. 1 Coxal plate 1 reduced; urosome with strong acute tooth . . . (VALETTIOPSIS) 2Coxal plate 1 not reduced; urosome lacking strong acute tooth (VALETTIETTA gen. nov.) 4 2. Pereon segments 5-7 and pleosome segments 1-3 dorsally dentate .... multidentataPereon segments 5-7 and pleosome segments 1-3 not dorsally dentate 3 3. Gnathopod 2 propodus elongate, tapering dentata Gnathopod 2 propodus stout, ovo-rectangular macrodactyla 4. Gnathopod 2 simple, or with very small palm, propodus slender, tapering .... 5Gnathopod 2 subchelate, propodus not slender, ovo-rectangular .... lobata sp. nov. 5. Gnathopod 2 palm oblique, coxal plate 4 distal margin convex anacantha Gnathopod 2 lacking palm, coxal plate 4 distal margin straight .... gracilis sp. nov. Valettietta gracilis and V. anacantha can be regarded as vicarious species having disjunctdistributions, one from the Atlantic Ocean and the other from the Pacific Ocean. Thedifferences separating them, although considered valid at species level, are of a minor nature,and are much less marked than those separating either species from Valettietta lobata.Within Valettiopsis, the Pacific species dentata and the Atlantic macrodactyla form a similarspecies pair. Other Atlantic/Pacific species pairs are known; Paracallisoma albertiChevreux, 1903 and P. coecum (Holmes, 1908), and Crybelocephalus birsteini Thurston,1976 and C. obensis Birstein & Vinogradov, 1964. The separation of the Atlantic and Pacificelements of these species pairs may have occurred in the geologically recent past. All aremeso- to abyssopelagic, and have been found in areas that were contiguous prior to theemergence of the Isthmus of Panama about 3-5 x 10 6 yearsB.P. (Keigwin, 1978). References Barnard, J. L. 1961. Gammaridean Amphipoda from depths of 400 to 6000 meters. Galathea Rep. 5 : 23-128.1967. Bathyal and abyssal gammaridean Amphipoda of Cedros Trench, Baja California. Bull. U.S. natn. Mus. 260 : 1-205. 1969. The families and genera of marine gammaridean Amphipoda. Bull. U.S. natn. Mus. 271 : 1-535.Birstein, J. A. & Vinogradov, M. E. 1963. The deep-sea pelagic amphipods of the Philippine Trench. Trudy Inst. Okeanol. 71 : 8 1-93 (In Russian).1964. Pelagic gammarids of the northern part of the Indian Ocean. Trudv Inst. Okeanol. 65: 152-195. VALETTIETTA GEN. NOV. 101 Chevreux, E. 1899. Sur deux especes geantes d'Amphipodes provenant des campagnes du yacht Princesse Alice. Bull. Soc. zool. Fr. 24: 1 52-1 58.1903. Note preliminaire sur les Amphipodes de la famille Lysianassidae receuillis par la Princesse Alice dans les eaux profondes de TAtlantique et de la Mediteranee. Bull. Soc. zool. Fr. 28: 8 1-97.1909. Diagnoses d'Amphipodes nouveaux provenant des campagnes de la Princesse Alice dans 1'Atlantique nord. Bull. Inst. oceanogr. Monaco 150 : 1-7. 1935. Amphipodes provenant des campagnes du Prince Albert l er de Monaco. Result. Camp. sclent. Prince Albert I 90 : 1-2 14.Griffiths, C. 1977. The South African Museum's Meiring Naude cruises. Part 6 Amphipoda. Ann. S. Afr. Mus. 14(4): 105-123.Holmes, S. J. 1908. The Amphipoda collected by the U.S. Bureau of Fisheries Steamer 'Albatross' off the west coast of North America, in 1903 and 1904, with descriptions of a new family and several new genera and species. Proc. U.S. natn. Mus. 35 : 489-543.Keigwin, L. D., Jr. 1978. Pliocene closing of the Isthmus of Panama, based on biostratigraphic evidence from nearby Pacific Ocean and Caribbean Sea cores. Geology, Ashtead6(\0) : 630-634.Lincoln, R. J. & Hurley, D. E. 1981. The calceolus, a sensory structure of gammaridean amphipods (Amphipoda: Gammaridea). Bull. Br. Mus. nat. Hist. (Zool.) 40 (4) : 103-1 16.Pirlot, J. M. 1933. Les amphipodes de 1'expedition du Siboga. Deuxieme partie. Les amphipodes gammarides. II. Les amphipodes de la mer profonde. I. Siboga Exped. 33c (1 20): 1 1 5-1 67.Stebbing, T. R. R. 1888. Report on the Amphipoda collected by H.M.S. Challenger during the years 1873-1876. Rep. sclent. Results. Voy. Challenger (Zoology) 29 : 1-1737.Thurston, M. H. 1976. New pelagic amphipods (Crustacea: Amphipoda) collected on the Sond cruise. J. mar. biol. Ass. U.K. 56: 143-159. Manuscript accepted for publication 10 June 1982 Three new genera of misophrioid copepods from thenear-bottom plankton community in the NorthAtlantic Ocean G. A. Boxshall Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Introduction Only three species of the copepod order Misophrioida have been described, Misophriapallida Boeck 1864, Benthomisophria palliata Sars 1909 and B. cornuta Hulsemann & Grice1964. Despite the small number of species the misophrioids have attracted interest becauseof the combination of characters that they exhibit, drawn from both the gymnoplean andpodoplean lineages within the Copepoda. Attention has recently been drawn to the uniquecharacters that misophrioids display; the possession of a carapace-like posterior extension ofthe cephalosome, the lack of a nauplius eye in a free living copepod, the distensibility of thegut and the retention of the antennary glands as the functional excretory organs of the adult(Boxshall, 1982). These characters, together with the abbreviated lecithotrophic naupliusphase of the life cycle, can be interpreted collectively as evidence of a bathypelagic origin ofthe Misophrioida. This interpretation is supported by the discovery of several newmisophrioid taxa from the deep North Atlantic Ocean, which was reported at the FirstInternational Conference on Copepoda held at Amsterdam in August 1981 (Boxshall, inpress). A total of 37 misophrioids was taken in a single haul fished near the bottom in 3000 m ofwater to the south west of the Azores. Fourteen of these were B. cornuta, 8 were B. palliataand 15 represented previously undescribed taxa. Three new genera and species are heredescribed on the basis of 13 of these specimens, the remaining 2 specimens being too badlydamaged for description. The new genera are of great phylogenetic significance as theyexhibit some very plesiomorphic characters which provide fresh insights on the nature of theappendages of the common ancestor of the Copepoda as a whole. The new records of B.cornuta and B. palliata further extend their known geographical ranges. Family MISOPHRIIDAE Genus ARCHIMISOPHRIA nov.DIAGNOSIS. As for type species.TYPE SPECIES. Archimisophria discoveryi gen. et sp. nov. Archimisophria discoveryi gen. et sp. nov. Adult female (Fig. 1A) body length 1-1 to 14 mm. Prosome large, apparently 4-segmentedbut with first free thoracic somite entirely concealed beneath a carapace-like extension fromthe posterior margin of the maxilliped-bearing somite. Nauplius eye absent. Prominentanteriorly-directed rostrum visible from dorsal aspect, not fused to labrum (Fig.1 B). Cone organs not observed but large mass of glandular tissue present on sides ofcephalosome beneath usual location of cone organs. Urosome 6-segmented (Fig. 1C). Surface Bull. Br. Mus. nat. Hist. (Zool.) 44(2): 103-124 Issued 24 February 1983 104 G. A. BOXSHALL Fig. 1 Archimisophria discovery! gen. et sp. nov. Holotype 9. A, dorsal view; B, rostrum andlabrum, ventral; C, urosome, ventral; D, antennule, dorsal. Scales 100 urn unless otherwisestated. MISOPHRIOID COPEPODS 105 of prosome and urosome somites 1 to 5 ornamented with a reticulum of epicuticularlamellae. Urosome somite 6 without reticulate markings. Caudal rami longer than wide;armed with 2 long distal margin setae, 2 medium-length distal angle setae, a dorsal seta nearthe inner margin and a distally located lateral seta. Antennule (Fig. ID) 27-segmented, articulating proximally with an expanded area ofventral cephalic surface (Fig. IB). Armature elements as follows: 1-2, II-2, III-2+1aesthetasc, IV-2, V-2, VI-2, VII-2 + 1 aesthetasc, VIII-2, IX-2, X-2, XI-2+1 aesthetasc,XII-2, XIII-2, XIV-2, XV-2, XVI-2+1 aesthetasc, XVII-2, XVIII-2, XIX-2, XX-2,XXI-2, XXII-1, XXIII-1, XXIV-2, XXV-2, XXVI-2 + 1 aesthetasc, XXVII-5+1aesthetasc. First segment also with patch of minute spinules. Labrum (Fig. IB) small, posteriorly directed, not fused with rostrum, with a posterior rowof marginal denticles. Antenna (Fig. 2A), basis lacking inner distal seta; endopod 3-segmented, exopod 8-segmented. Endopod segment 1 apparently unarmed; segment 2 with 4 unequal unilaterallyplumose setae at inner distal angle; segment 3 with 6 long subequal unilaterally plumosesetae along distal margin and with several transverse rows of spinules. Exopod segment 1with a short naked seta at inner distal angle; segments 2 and 3 unarmed; segment 4 with along plumose seta at inner distal angle; segment 5 with 2 long plumose setae on inner margin;segments 6 and 7 small, unarmed; segment 8 with 3 long unilaterally plumose setae on distalmargin and areas of spinules subapically. Mandible (Fig. 2B) with well developed gnathobase bearing distally 2 multicusped blades, 5 strong spines and an extensive fringe of pinnules. Mandibular palp comprising basis,2 -segmented endopod and 4-segmented exopod. Basis armed with a naked seta at inner distalangle. Endopod segment 1 with a short unilaterally plumose seta at inner distal angle;segment 2 with 8 unequal plumose setae along distal margin. Exopod segment 1 unarmed;segments 2 and 3 each with 1 long seta at inner distal angle; segment 4 with 3 similarunilaterally plumose setae and a short naked seta. Maxillule (Fig. 2C), gnathobase with 14 distal elements; endite 1 with 1 spiniform and 3setiform armature elements, endite 2 with 3 spiniform elements. Outer lobe rudimentary,represented by 6 plumose setae on outer surface of segment. Maxillulary palp biramous with2 -segmented endopod and 1 -segmented exopod. Endopod segment 1 with 3 unequalarmature elements at inner distal angle; segment 2 with a long and a short seta proximally oninner surface and an apical armature of 3 long unilaterally plumose setae, 1 long and 2 shortnaked setae. Exopod with a proximal fringe of pinnules and 6 plumose setae on inner marginand with 3 long unilaterally plumose setae and a naked seta on distal margin. Maxilla (Fig. 3A) 6-segmented; segment 1 with 6 plumose setae on proximal endite and 3on distal endite; segment 2 with 3 similar setae on both proximal and distal endites; segment3 produced medially into a curved claw armed with 3 naked setae near its base; segments 4 to 6 with a total of 10 setae. Maxilliped (Fig. 2D) 8-segmented, with a 3-segmented, robust proximal portion and aslender 5-segmented distal portion. Segment 1 with 1 seta on inner surface; segment 2 with 2medial setae and a row of pinnules along the outer margin; segment 3 with proximal enditebearing 1 strong spine and 3 setae, distal endite with 1 naked seta and a long plumose seta, 2other setae on inner margin, a plumose seta at inner distal angle and a long row of pinnulesalong outer margin; segments 4 to 6 with 1, 2 and 1 medial setae respectively, each armedwith short spinules; segment 7 with an inner margin spinulate seta and an outer plumose setaon which the pinnules decrease markedly in length towards the apex; segment 8 with 3similar plumose setae and a naked seta. Legs 1 to 4 incomplete in holotype 9 and paratype 9, assumed to be similar to thosedescribed below for a paratype rf. Leg 5 (Fig. 1C) uniramous, 3-segmented and positioned midventrally with inner marginsalmost touching at base. Segment 1 with 1 naked seta at outer distal angle; segment 2 with ashort naked seta in same position; segment 3 elongate with 2 unequal distal margin setae, thelonger armed with spinules bilaterally. G. A. BOXSHALL Fig. 2 A. discoveryi. A, antenna, anterior; B, mandible, anterior; C, maxillule, posterior; D,maxilliped, posterior. Scales 100 |im unless otherwise stated. MISOPHRIOID COPEPODS 107 Fig. 3 A. discoveryi. A, maxilla, anterior; B, Paratype cf, dorsal view; C, urosome, ventral; D,antennule, dorsal. Scales 100 fim unless otherwise stated. 108 G. A. BOXSHALL Leg 6 (Fig. 1C) reduced to a semicircular flap closing off the opening of the genital antrum;bearing an outer plumose seta and a short inner spine. Adult male (Fig. 3B) body length 1-1 to 1-3 mm (based on 3 specimens). Prosome andurosome (Fig. 3C) as in adult female. Appendages as in female except for antennules and legs5 and 6. Antennules (Fig. 3D) 25-segmented, unigeniculate with the articulation between segmentsXIX and XX. Armature elements as follows: 1-2, II-2, III-2, IV-2, V-2, VI-2, VII-2, VIII-2,IX-2, X-2, XI-2 + 1 aesthetasc, XII-2, XIII-2, XIV-2, XV-4, XVI-2+1 aesthetasc,XVII-2, XVIII-2, XIX-2, XX-0(?), XXI-1, XXII-2, XXIII-2, XXIV-2+1 aesthetasc,XXV-3 + 1 aesthetasc. Segment XIII with a spinous process at posterolateral angle. Legs 1-4 (Figs 4A-D) biramous with 3-segmented rami; armature formula as follows: coxa basis endopod exopod legl 0-1 1-1 0-1;0-1;1,2,3, I-1;I-1;III,I,3 leg 2 0-1 1-0 0-1;0-2;1,2,3 I-1;I-1;1 11,1,4 leg 3 0-1 1-0 0-1;0-2;1,2,3 I-1;I-1;III,I,4 Ieg4 0-1 1-0 0-l;0-2;missing I-1;I-1;1 11,1,4 Pinnule rows present on inner and outer margins of endopod segments and inner margins ofexopod segments; spinules present along margins of exopod segments. All outer marginexopodal spines armed with bilateral strips of fine membrane, apical spine with membraneexternally and pinnules internally. All setae plumose except for distalmost 2 inner marginsetae on exopod of leg 4. These setae with blunt tips and rows of short stout pinnules,possibly representing a male dimorphic character. Leg 5 (Fig. 4E) uniramous, 4-segmented and with bases of legs almost touching at ventralmidline as in 9. Segment 1 bearing 1 plumose seta at outer distal angle; segment 2 with nakedseta in same position; segment 3 with plumose seta at inner distal angle; segment 4 with shortinner margin plumose seta and 2 unequal plumose setae on distal margin. Leg 6 (Fig. 3C) represented by a flattened plate bearing a long outer plumose seta and ashort inner spine. MATERIAL EXAMINED. Holotype 9, 3 paratype cfd 1 , 19,2 Copepodid IV and 3 Copepodid IIIstages all from Discovery Stn 10379*37 (345T N 3255' W) in the North Atlantic to thesouthwest of the Azores. Collected in RMT1+8M net system fished 23 to 56 m off thebottom in a water depth of about 3000 m. BM(NH) Registration Nos Holotype 9 1982.128,paratype 9 1982.129, rfrf 1982.130-132,Co. IV 1982. 133-134 and Co. Ill 1982.135-137. REMARKS. The new genus differs from all known misophrioids, including those describedherein, in the possession of an anteriorly directed rostrum. In other genera the rostrum iseither ventrally directed (Misophria and Misophriopsis gen. nov.) or postero ventral lydirected and fused to the labrum (Benthomisophria and Misophriella gen. nov.). Anotherremarkable feature of this genus is the large number of segments in the antennules of bothsexes. The twenty-seven segments found in the female is the largest number recorded for anycopepod, including the calanoids in which 25 is the largest number known. Thephylogenetic significance of the multi-segmented antennules is discussed below. The developmental stages of A. discovery! will not be described as only the third (Co. Ill)and fourth (Co. IV) copepodid stages have been found. As in other misophrioids (Boxshall &Roe, 1980) the copepodid stages can be determined by the number of urosome somites, theCo. Ill having 3 and the Co. IV having 4. It is interesting to note that the segmentation of theantennule is complete (27 segments) at the Co. IV stage whereas in Benthomisophria palliatathe complete complement of 1 8 segments is not achieved until the last moult into the adult.The third copepodid of A. discoveryi has a 24-segmented antennule. The presence of a 3-segmented leg 5 in female and 4-segmented leg 5 in male A. discoveryi MISOPHRIOID COPEPODS 109 Fig. 4 A. discoveryi. A, leg 1 , anterior; B, leg 2, anterior; C, leg 3, anterior; D, leg 4, anterior; E,leg 5, anteroventral. Scales 100 urn unless otherwise stated. 110 G. A. BOXSHALL is the first documented example of sexual dimorphism in the fifth leg in misophrioids. InMisophria and Benthomisophria sexual dimorphism is restricted to the antennules and sixthlegs. The presence of modified setae on the inner margin of the third exopod segment of leg 4in male A. discoveryi may also represent sexual dimorphism but this cannot be confirmeduntil more material is obtained, as the only 2 females in the present material had incompletefourth legs. Genus MISOPHRIELLA nov. DIAGNOSIS. As for type species. TYPE SPECIES. Misophriella tetraspina gen. et sp. nov. Misophriella tetraspina gen. et sp. nov. Adult female (Fig. 5 A) body length 0-87 mm (Holotype 9). Prosome apparently 4-segmentedbut, as in all misophrioids, with the first free thoracic somite entirely concealed beneath acarapace-like extension of the posterior margin of the maxilliped-bearing somite. Naupliuseye absent. Rostrum posteroventrally directed and fused to anterior surface of labrum, notvisible from dorsal aspect. Cone organs present in lateral areas on cephalosome. Urosome6-segmented (Figs 6A-B), somites 2 to 5 each with a hyaline frill around posterior border.Anal somite with paired lobes dorsally, either side of anus and row of minute spinulesaround posterior margin. Pairs of pores present on both dorsal and ventral surfaces of analsomite. Caudal rami about as long as wide, armed with 2 long distal margin setae, 2 mediumlength distal angle setae, a dorsal seta near the distal margin, a seta near the middle of thelateral margin and a proximal setule on this margin. Antennule (Fig. 5B) 19-segmented, articulating basally with a raised area of ventralcephalic surface. Armature elements as follows: 1-1, II-9, III-2, IV-2, V-6, VI-2, VII-2,VIII-2, IX-2+ 1 aesthetasc, X-2, XI-2+ 1 aesthetasc, XII-2, XIII-2, XIV-2 + 1 aesthetasc,XV-1, XVI-1, X VII-2, X VIII-2 + 1 aesthetasc, XIX-6+ 1 aesthetasc. Segment I with 1 andsegment II with 3 strong curved setae, each strongly sclerotized with an expanded base and arow of spinules along its convex margin. Labrum large, posteriorly directed and fused with rostrum. Antenna (Fig. 5C); basis with inner distal seta; 3-segmented endopod and 6-segmentedexopod. Endopod segment 1 with a short plumose seta at inner distal angle; segment 2 with 2short naked setae near middle of inner margin and a long and a short seta at inner distalangle; segment 3 with 6 long unilaterally plumose setae along its distal margin. Exopodsegment 1 unarmed; segments 2 to 5 each with a single long, unilaterally plumose seta on itsinner margin; segment 6 with 2 similar setae and a short plumose seta on its distal margin. Mandible (Fig. 7A) with well developed gnathobase bearing 4 multicusped blades andsome pinnules distally. Mandibular palp comprising basis, 2-segmented endopod and4-segmented exopod. Basis apparently lacking inner distal angle seta. Endopod segment 1with 1 naked seta at inner distal angle; segment 2 with 1 short naked seta and 4 long plumosesetae apically. Exopod segment 1 unarmed; segments 2, 3 and 4 with 1 , 2 and 3 long plumosesetae respectively. Maxillule (Fig. 7B) with armature of gnathobase reduced, comprising only 7 curvedspinous elements, 1 hirsute seta and 2 slender naked setae. Endites 1 and 2 with 5 and 4slender setae respectively; all setae sparsely armed with short spinules bilaterally. Outer lobeapparently absent. Maxillulary palp biramous with 1 -segmented exopod and 3-segmentedendopod. Endopod segment 1 fused to basis, with 3 unequal plumose setae at inner distalangle; segment 2 with 2 inner margin plumose setae; segment 3 small, bearing 4 unequalsetae apically. Exopod with 3 long plumose setae distally and a short plumose seta and a rowof pinnules along inner margin. Maxilla (Fig. 7C) 6-segmented; segment 1 with 2 hemispherical endites, proximal enditewith 1 naked and 4 plumose setae, distal endite with 2 spinulate setae; segment 2 with a MISOPHRIOID COPEPODS 111 Fig. 5 Misophriella tetraspina gen. et sp. nov. Holotype 9. A, dorsal view; B, antennule, dorsal;C, antenna, anterior. Scales 50 Jim unless otherwise stated. 112 G. A. BOXSHALL Fig. 6 M. tetraspina. A, urosome, dorsal; B, urosome, ventral. Scale 100 |im. single elongate endite bearing 3 unequal naked setae at its apex; segment 3 produced into along medial claw with a fringe of minute pinnules along its concave margin and 3 setae nearits base; segments 4 to 6 each with 1 long, robust claw-like seta armed with a fringe ofpinnules, segment 6 also bearing 2 slender apical setae. Maxilliped (Fig. 7D) 7-segmented; segments 1 and 2 long and robust. Segment 1 armedwith 1 proximal seta, 3 midmargin setae and 2 distal setae all on inner margin; segment 2with 3 slender setae at middle of inner margin; segments 3 to 5 with 1, 2 and 1 slender innermargin setae respectively; segment 6 with an articulated seta distally and segment 7 with 2similar articulated setae, plus 2 short naked setae. Legs 1-4 (Figs 8 A-D) biramous, presumably with 3-segmented rami; armature formula asfollows: leglleg 2leg 3leg 4 coxa 0-1 0-1 0-1 0-1 basis1-11-01-01-0 endopod0-1 ;0-2; 1,2,30-1 ;missing0-l;0-2;missing0-l;0-2;missing exopod I-l;missingI-l;missing Pinnule rows present on inner and outer margins of endopod segments and on inner marginsof exopod segments. Leg 1 with accessory digitiform processes on outer margin of exopodsegments 2 and 3 between bases of spines. Exopod spines armed with bilateral strips ofserrate membrane.Leg 5 (Fig. 8E) uniramous, 4-segmented. Segment 1 broader than long, unarmed; segment MISOPHRIOID COPEPODS 113 2 broader than both segments 1 and 3, armed with a naked seta on outer margin; segment 3with spinous process at outer distal angle; segment 4 with 1 plumose seta either side ofcentral spine in distal margin, also with an inner margin plumose seta on left leg but not onright. Leg 6 (Fig. 6B) forming a curved plate on ventral surface of genital somite; armatureincomplete. MATERIAL EXAMINED. Holotype 9 collected at Discovery Stn 10379*37 (3457 N 3255'W)in the North Atlantic southwest of the Azores. Caught in RMT1 + 8M net system fished 23to 56 m off the bottom in a water depth of about 3000 m. BM(NH) Registration No.1982-138. REMARKS. The new genus differs from all known misophrioids, including those describedherein, in the form of the maxillule (which has a 3 -segmented endopod, no outer lobe and anelongate exopod bearing only 4 setae), in the presence of only a single endite on the secondsegment of the maxilla and in the number of segments in the antennule. This genus exhibits ageneral reduction in the numbers of armature elements on most of the mouthparts, particu-larly on the maxillulary palp and the maxilla. In addition to these quantitative differencesthere are also qualitative differences, such as the spiniform nature of 4 setae on the proximalsegments of the antennule and the presence of articulated setae on the maxilliped. The armature elements of the fifth legs provide some indication of the homology of thesegments. The second segment carries an outer seta at its distal angle and it is also muchwider than the other segments. It probably represents the basis. Segment 1 thereforerepresents the coxa, and segments 3 and 4 the 2-segmented exopod. The holotype exhibitsbilateral asymmetry in the armature of leg 5 but it is assumed that this is an aberrant con-dition and is not indicative of a true asymmetry as displayed by the fifth legs of manycalanoids. Genus MISOPHRIOPSIS nov.DIAGNOSIS. As for type species.TYPE SPECIES. Misophriopsis dichotoma gen. et sp. nov. Misophriopsis dichotoma gen. et. sp. nov. Adult female (Fig. 9 A) body length 0-9 mm (Holotype 9). Prosome apparently 4-segmentedbut with first free thoracic somite entirely concealed beneath a carapace-like extension fromthe posterior margin of the maxilliped-bearing somite. Nauplius eye absent. Rostrum small,ventrally directed with its apex adjacent to, but not fused to, the labrum (Fig. 12). Coneorgans present in lateral areas on either side of cephalosome. Urosome (Fig. 9B)6-segmented. Caudal rami wider than long, armed with 2 long distal margin setae, a mediumlength seta at both inner and outer distal angles, another on the dorsal surface near bases ofdistal setae, and a short lateral seta. Antennule (Fig. 9C) 18-segmented. Armature elements as follows: 1-1, II-l 1, III-2, IV-6,V-2, VI-2, VII-2, VIII-2 + 1 aesthetasc, IX-2, X-2+ 1 aesthetasc, XI-2, XII-2, XIII-2+ 1aesthetasc, XIV-1, XV-1, XVI-2, XVII-2 + 1 aesthetasc, XVIII-6+1 aesthetasc. Spinulespresent on posterior surface of segment II. Labrum (Fig. 12) large, posteriorly directed but not fused with rostrum; armed with 2 largemedially directed spinous processes on its posterior margin. Antenna (Fig. 9D) basis lacking inner distal seta; endopod 3 -segmented, exopod6-segmented. Endopod segment 1 with 2 inner distal setae; segment 2 with 3 setae spacedalong inner margin; segment 3 with 5 long distal margin setae. Exopod segment 1 unarmed;segment 2 with 2 inner margin setae; segments 3 to 5 each with a single seta at inner distalangle; segment 6 with 3 plumose setae. Mandible (Fig. 9E) with well developed gnathobase bearing distally 4 multicusped blades, 114 G. A. BOXSHALL Fig. 7 M. tetraspina. A, mandible, anterior; B, maxillule posterior; C, maxilla, anterior; D, maxilliped, posterior. Scale 100 |im. MISOPHRIOID COPEPODS 115 Fig. 8 M. tetraspina. A, leg 1 , anterior; B, leg 2, anterior; C, leg 3, anterior; D, leg 4, anterior; E,leg 5, antero ventral. Scales 100 ^im unless otherwise stated. 116 G. A. BOXSHALL Fig. 9 Misophriopsis dichotoma gen. et sp. nov. Holotype 9 A, dorsal view; B, urosome, ventral;C, antennule, dorsal; D, antenna, anterior; E, mandible, posterior; F, detail of mandibulargnathobase. Scales 100 (im unless otherwise stated. MISOPHRIOID COPEPODS 117 several strong spines and a small subapical patch of pinnules. Mandibular palp comprisingbasis, 2-segmented endopod and an indistinctly 5-segmented exopod. Basis armed with aplumose seta midway along inner margin. Endopod segment 1 with plumose seta at innerdistal angle; segment 2 elongate with 4 unequal setae on distal margin. Exopod segments 1and 2 incompletely separated; segment 1 unarmed; segments 2 and 3 each with a plumoseseta at inner distal angle; segment 4 probably with inner seta, missing from dissected appen-dage but its presence indicated by a scar on the surface of the segment; segment 5 with 1 innerand 2 distal margin setae. Maxillule (Fig. 10 A), gnathobase with 7 distal margin spines, 2 hirsute setae and 3 nakedsetae subapically on the posterior surface, and 2 plumose setae on a spinulate swelling on theanterior surface. Endite 1 short and slightly furrowed on its posterior surface, armed with 6apical plumose setae. Endite 2 long, with 3 apical plumose setae. Outer lobe rudimentary,represented by 8 plumose setae on outer surface of segment. Maxillulary palp biramous with2-segmented endopod and 1 -segmented exopod; segment 1 of endopod fused to basis, withjunction marked by 2 subapical setae. Endopod segment 1 with 4 plumose setae at innerdistal angle; segment 2 with 3 naked setae arising proximal to the midpoint of the innermargin, 3 similar setae arising subapically on same margin, and 5 setae on distal margin.Exopod with 9 plumose inner and distal margin setae of varying lengths and with fringes oflong pinnules proximally. Maxilla (Fig. 1OB) 6-segmented; segment 1 with 5 plumose setae on proximal endite and 3on distal endite; segment 2 with 3 plumose setae on both proximal and distal endites;segment 3 produced medially into a curved claw armed with 2 naked setae near its base;segments 4 to 6 with a total of 7 naked setae. Maxilliped (Fig. IOC) 7-segmented, although proximal segment showing some signs ofsubdivision at midlength. Segment 1 bearing 4 plumose setae and a short naked seta alonginner margin, and some long pinnules proximally on outer margin; segment 2 with 3 innermargin plumose setae; segments 3 to 6 each with 2 long, unilaterally plumose setae at innerdistal angle; segment 7 with 3 distal setae. Legs 1-4 (Figs 1 1 A-D) biramous, with 3-segmented rami. Armature formula as follows: coxa basis endopod exopod legl 0-1 I-I 0-1;0-2;1,2,3 I-1;I-1;III,I,4 Ieg2 0-1 1-0 0-1;0-2;1,2,3 I-1;I-1;III,I,5 leg 3 0-1 1-0 0-l;0-2;missing I-1;I-1;III,I,5 leg 4 0-1 1-0 0-l;0-2;missing I-1;I-1;III,I,5 Outer margins of all exopod segments with strips of serrated membrane. Rows of pinnulespresent on inner margins of all exopod segments and inner and outer margins of endopodsegments. Outer margin spines of leg 1 armed bilaterally with fine strips of smoothmembrane. Apical spines with short pinnules along inner margins, and strip of smoothmembrane on outer margin in leg 1. Outer margin element on basis spinous on leg 1,setiform on legs 2 to 4. Leg 5 (Fig. 1 1 E) biramous, comprising unsegmented protopod, 2-segmented exopod and1 -segmented endopod. Basal seta present at outer distal angle of protopod. Exopod segment1 unarmed, segment 2 with 3 distal margin elements, a long plumose outer seta, a medianspine and an inner naked seta. Endopod with single plumose seta apically. Leg 6 (Fig. 1 1 F) with transverse intercoxal sclerite joining members of leg pair reduced to aslender bar. Leg comprising an outer process with a long apical seta, a median spine and aninner spinous process. MATERIAL EXAMINED. Holotype 9 collected at Discovery Stn 10379*37 (3457 N 3255' W)in the North Atlantic to the southwest of the Azores. Caught in RMT1 +8M net systemfished 23 to 56 m off the bottom in a water depth of about 3000 m. BM(NH) Registration No.1982-139. 118 G. A. BOXSHALL Fig. 10 M. dichotoma. A, maxillule, posterior; B, maxilla, anterior; C, maxilliped, anterior. Scales 100|im. MISOPHRIOID COPEPODS 119 Fig. 11 M. dichotoma. A, leg 1 , anterior; B, leg 2, anterior; C, leg 3, anterior; D, leg 4, anterior; E,leg 5, anteroventral; F, leg 6, ventral. Scales 1 00 ^im unless otherwise stated. 120 rostrum G. A. BOXSHALL antennule antenna labrum paragnath intermaxillaryswelling mandibularpalp Fig. 12 M. dichotoma. Ventral view of mouthparts of left side, with antennule, left paragnath andmandibular gnathobase removed. Scale 100 urn. REMARKS. The new genus differs from all known misophrioids, and from all knownpodopleans, in the possession of a biramous fifth leg. It also differs from other misophrioidgenera in the presence of a pair of large spinous processes on the posterior margin of thelabrum. The arrangement of the mouthparts (Fig. 12) is very similar to that ofBenthomisophria palliata. The antennae and mandibular palps are both reflexed so that theirsetae will sweep over the areas of cone organs located laterally on the cephalosome as inBenthomisophria. Key to genera and species of the Misophrioida 1. Large anteriorly directed rostrum visible in dorsal view; antennule with 25 (rf) or 27 (9) segments Archimisophria discovery! Rostrum ventrally or postero ventral ly directed, not visible in dorsal view; antennule withless than 25 segments 2 2. Leg 5 biramous, with 1 -segmented endopod and 2-segmented exopod .... Misophriopsis dichotoma Leg 5 uniramous, with 1 to 4 segments 3 3. Leg 5 4-segmented, antennule with 19 segments (9) . . . MisophriellatetraspinaLeg 5 with less than 4 segments, antennule with less than 1 9 segments (9) .... 4 4. Leg 5 3-segmented; antennule with 1 3 segments (d) or 1 6 segments (9) Misophria pallidaLeg 5 2-segmented, comprising a triangular proximal segment and short distal segment; antennule with 18 (9) or 16 (d 1 ) segments Benthomisophria palliata Leg 5 1 -segmented; antennule with 16 segments (9 & rf) B.cornuta MISOPHRIOIDCOPEPODS 121 Discussion These three new genera exhibit between them an unusual array of plesiomorphic characters,many of which are present in a state approaching that attributed to the hypothetical ancestorof the Copepoda as a whole (see discussion in Boxshall et al., in press). The antennae, forexample, are biramous with an 8-segmented exopod and a 3-segmented endopod inArchimisophria, and the mandibles have a well developed gnathobase plus a biramous palpwith a 5 -segmented exopod and 2 -segmented endopod in Misophriopsis. The basic structureof both these limbs is the same as that proposed for the ancestral copepod. The detailedstructure of the maxillule was not considered in the discussion reported by Boxshall et al. (inpress) but in my opinion the misophrioid pattern of large gnathobase, 2 other endites, 1setose outer lobe, a 1 -segmented exopod and a 3-segmented endopod, with the first segmentfused to the basis, may well be similar to that possessed by the ancestral copepod. Themisophrioid maxilla comprises 6 segments, the first 2 each bear a pair of setose endites, thethird a claw-like endite, and the fourth to sixth variable number of inner and distal setae.This is close to what may be considered to be the ancestral copepod pattern. The8-segmented maxilliped of Archimisophria consisting of a 3-segmented protopod and5-segmented endopod is also very similar to the 9-segmented basic copepod maxillipedfavoured by Gurney (193 1) in his analysis of copepod appendages. All these misophrioid features closely approximate to those exhibited by theplesiomorphic calanoids. It is the common possession of these calanoid-likegnathostomatous mouthparts and the possession of a heart that indicates that theMisophrioida diverged from the podoplean lineage soon after its separation from thegymnoplean lineage. The discovery of a 27-segmented antennule in Archimisophria and of abiramous fifth leg in Misophriopsis clearly demonstrates that the Misophrioida has divergedless from the common ancestral stock of the Copepoda than any other podoplean group. The possession of a biramous fifth leg is of great phylogenetic significance. The differencebetween the normal biramous fifth swimming leg of gymnopleans and the reduceduniramous fifth leg of podopleans led Giesbrecht (1899) to suggest the possibility that theyare not homologous. Gurney (1931) rejected this and suggested that the typical uniramousleg of podopleans represents the exopod of an originally biramous limb. This interpretationhas been widely adopted and the presence, in Misophriopsis, of a biramous fifth leg in whichthe endopod is reduced to a single segment bearing a single seta provides furtherconfirmation. The fifth leg undergoes considerable reduction within the Misophrioida. InMisophria the endopod is represented by a single median seta on the distal margin of theunsegmented protopod, although the exopod is similar to that found in Archimisophria. InBenthomisophria cornuta the fifth leg is reduced to a single segment. The 27-segmented antennule of female Archimisophria is of interest because of the con-siderable controversy that exists (see Boxshall et al, in press) concerning the nature of thislimb in the ancestral copepod. Giesbrecht (1892 & 1899) analysed the segmentation andarmature of the antennules of many calanoid and other copepods in an attempt to reduce theantennule of all copepods to a common type. Giesbrecht's basic copepod antennule was25-segmented and by studying the arrangement of the armature elements he was able todetermine which segments had fused in those forms with fewer segments. This basic limbclosely resembles that ofCalanusfinmarchicusGunnerus, 1770 both in number of segmentsand in setation. The typical armature present on each antennulary segment is 2 setae and 1aesthetasc, at least in the female, although one or more of these elements is often lost, mostcommonly the aesthetasc. Even the arrangement of these 3 elements, which Giesbrechtcalled a 'trithek', follows a constant pattern. One seta, the proximal seta, is positioned aboutmidway along the anterior margin of the segment whereas the other seta, the distal seta, andthe aesthetasc are positioned close together at the distal angle of the anterior margin. Thetypical trithek may have been different for the male, because a proximal and a distal setaplus 2 distal aesthetascs are commonly found, as for example, in Eucalanus attenuatusDana, 1849. Some of the more distal segments have modified tritheks. In female Calanus 122 G. A. BOXSHALL MISOPHRIOID COPEPODS 123 finmarchicus segments 20 and 21 have no proximal seta, segments 22, 23 and 24 have noproximal seta but have instead a distal seta on the posterior margin, and the terminalsegment 25 has an increased armature of up to 6 setae and an aesthetasc (see Table 1). AsGurney (193 1) noted, many calanoids possess 3 complete tritheks on the second segment anda single proximal seta plus a distal trithek on the first segment. He interpreted this asevidence that the second segment of calanoid antennules is derived from 3 fused segmentsand that the first segment may be derived from 2 fused segments. On the basis of thisinterpretation he postulated that the ancestral copepod antennule comprised 27 or possibly28 segments. The discovery of Archimisophria with its 27-segmented antennules provides aremarkable corroboration of Gurney 's hypothesis. Comparison between the antennules of female Calanus finmarchicus and Archimisophriadiscoveryi is made in Table 1. The precise correspondence in the position of typical tritheks,denoted by T or t, and of modified tritheks, denoted by D or d and P or p, in these two taxa isremarkable as they are drawn from the 2 main copepod lineages, Gymnoplea and Podoplea.It is possible from the comparison in Table 1 to identify a common pattern from which bothmay be derived by reduction. I suggest that this pattern may well represent that found in thecommon ancestor of the Copepoda. Functional interpretation of the unique characters exhibited by misophrioids (Boxshall,1982 & in press) suggests that the ancestral misophrioid stock became adapted to abathypelagic existence and to gorging as a feeding strategy. The group appears to haveradiated in the deep-sea near-bottom environment and it is probable that many newmisophrioid taxa will be discovered as the near-bottom community is subject to moreintense study. Despite their obvious specializations the misophrioids also retain manycharacters of the presumed ancestral copepod stock and it is clear that they diverged from theancestral podoplean stock soon after it had attained its characteristic division into prosomeand urosome. Acknowledgements I would like to thank P. M. David and Dr Howard Roe of the Institute of OceanographicSciences for permission to work on this material and for arranging for it to be donated to theBM(NH) collections. I am also grateful to Dr Roger Lincoln for reading and commenting onthe manuscript. References Boeck, A. 1864. Oversigt over de ved Norgs Kyster iagttagne Copepoder henhevende tie Calanidernes, Cyclopidernes og Harpacticidernes Familiar. Fork. VidenskSelsk. Krist. 1864 : 226-28 1 .Boxshall, G. A. 1982. On the anatomy of the misophrioid copepods, with special reference to Benthomisophria palliata Sars. Phil. Trans. R. Soc. Lond. B. 297 : 125-181.in press. The functional morphology of Benthomisophria palliata Sars, with a consideration of the evolution of the Misophrioida. Crustaceana, Suppl. in press.Boxshall, G. A., F. D. Ferrari & H. Tiemann. in press. The ancestral copepod: towards a consensus of opinion at the First International Conference on Copepoda. Crustaceana, Suppl. in press.Boxshall, G. A. & H. S. J. Roe. 1980. The life history and ecology of the aberrant bathypelagic genus Benthomisophria Sars, 1909 (Copepoda: Misophrioida). Bull. Br. Mus. nat. Hist. (Zool.)38 : 9-41.Dana, J. D. 1849. Conspectus Crustaceorum quae in orbis terrarum circumnavigatione, Carolo Wilkes e classe Reipublicae Foederatae duce, lexit et descripsit Jacobus D. Dana. Proc. Am. Acad. ArtsSci.2: 8-61.Giesbrecht, W. 1892. Systematik und Faunistik des pelagischen Copepoden des Golfes von Neapel und der angrenzenden Meeresabschnitte. Fauna Flora Golfo Napoli 19: 1-83 1 .1899. Die Asterocheriden des Golfes von Neapel und der angrenzenden Meeresabschnitte. Fauna Flora Golfo Napoli 25 : 1-2 1 7.Gunnerus, J. E. 1770. Nogle smaa rare mestendelen nye norske S0dyr beskrevene. Skr. Kiobenhavnske Selsk. Laerd. og Videnskab. Elsk. 10 : 175. 124 G. A. BOXSHALL (in nicy, R. 1931. British Fresh-water Copepoda Vol. 1 . Ray Society, Lond. Hulsemann, K. & Grice, G. D. 1964. A new bathypelagic species of Benthomisophria (Copepoda: Misophriidae) from the North Atlantic. Zoo/. Anz. 173 : 259-264.Sars, G. O. 1909. Note preliminaire sur trois formes remarquables de copepodes provenant des Campagnes de S.A.S. Le Prince Albert de Monaco. Bull. Inst. Oceanogr. Monaco. 147 : 1-8. Manuscript accepted for publication 9 June 1 982 Larval development of British prawns and shrimps(Crustacea: Decapoda: Natantia) 4. Palaemon(Palaemon) serratus (Pennant, 1777) andfunctional morphology of swimming A. A. Fincham Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Introduction Palaemonid shrimps are widely distributed in the N.E. Atlantic being found in fresh water,in estuaries, intertidally and down to 40 m offshore, but their larvae are rare in the plankton.All decapod larvae have a natural tendency to sink (Foxon, 1934) and in order to maintainposition at a particular depth (Savage, 1926) and perform daily vertical migrations (Russell,1925, 1927; Hardy & Bainbridge, 1954), active upward swimming is necessary also. Foxonmeasured rates of movement in various decapods including pandalid carideans and severalauthors have noted the effect of light and gravity on the orientation and movement ofdecapod larvae (Sollaud, 1921;Gurney, 1942; Forward &Cron in, 1978). The aim of this paper is to review the larval development of Palaemon (Palaemon)serratus (Pennant, 1777) and report on morphological adaptations, and a mechanism usingmany-jointed plumose setae fringing the thoracic exopods, for larval swimming. Materials and MethodsRearing Ovigerous Palaemon (Palaemon) serratus were trawled from 12 m in April 1979 fromPlymouth Sound, Devon (Grid reference: SX 475512). Similar rearing techniques to thosereported previously (Fincham 1977, 1978, 1979) were used with the following modifications: 1 . The controlled temperature room was at 1 4 C. 2. Antibiotics were used for the first three stages only (Fincham, 1979). Larval material has been deposited in the Crustacea collection of the BM(NH), registrationnumber 1982 : 186. Telson morphology The telson was removed from larvae by a cut at the narrow junction with the abdomen,rinsed in distilled water, freeze dried, mounted end-on with Araldite on stubs, coated withgold and examined with a scanning electron microscope. Palaemon (Palaemon) serratus (Pennant, 1 777) Astacus serratus Pennant, 1777 Melicerta triliana Risso, 1816 Palaemon trilianus Risso, 1826 Palaemon treillianus H. Milne Edwards, 1837 Leander latrei/lianus Czerniavsky, 1 884 Bull. Br. Mus. not. Hist. (Zool.)44(2): 125-161 Issued 24 February 1983 125 126 A. A. FINCHAM LARVAL DEVELOPMENT OF P. (P.) SERRA TVS 1 27 CN! O O rf. Tt VO ON -T^^ OO^M "-> + + -^ + + +C* + + + + + + I + I+ + + ++ 1^+^^ A A A /A T^- A A >/-> o _k co so r-> rvi ON , m rn vor^l -II TJ- + + ^+ + + + + + + + + +1 + I +-~ + + + I ^ + J. ' NO Tt -o i^^t x . r (N rs S o^ I + 0+ ^7 ~ + 5^ 22 ps) -)- , , ~ ON 7 -ON- ^ + + -0 + + + + + + + + + +I + i ed i i i xj + + 2 2 JH rN i +1 +01 \+ +O+ r- "8.2 a-oo oo ac x UJ PU ^ 1 t- (/) - ^*4 'V" 3 ^5 + c o 'a . 45 ^^ a <u ~ D. D. 00C ^ O C 'a E 5jtr o C/5 OO xC -J -J 128 A. A. FINCHAM Leander serratus Sharp, 1 893Leander treillianus Adensamer, 1 898Leander serratus var. treillianus De Man, 1915Palaemon (Palaemon) serratus Holthuis, 1950 SYNOPSIS OF LARVAL DATA FROM PUBLISHED WORK. Thompson, 1 836 zoeae I, III & V, p. 22 1 ,Figs 1-5, laboratory reared and plankton. Couch, 1845 zoea I, p. 20, no Figs, laboratoryreared. Mayer, 1877 zoea I, telson, p. 250, Fig. 49. Keeble and Gamble, 1904 zoea I,chromatophores, p. 316. Sollaud, 1912 zoeae I-IX, p. 664, no Figs, laboratory reared.Williamson, 1915 zoea II (not I), Figs 120-125, zoea I Figs 126-128, last zoea Figs 129-132,p. 396, plankton. Sollaud, 1923 zoeae I-IX & post larva I, p. 530, plates 16-18 (Figs 1^only), laboratory reared and plankton. Reeve, 1969 zoeae I-V (XI), p. 77, no Figs,laboratory reared. Sutton et at., 1969 zoea I, counting larvae, p. 433, Fig. 1, laboratoryreared. Fincham & Williamson, 1978 key to larval stages. In the following short descriptions of the key characters of the larval stages, setal countshave been omitted usually, but they are recorded in Table 1 . Description of larval stages Key characters are printed in italic type and are useful for separating stages in British species.ZOEA 1 (Fig. 1)3-3 mm (3-2-3-5 mm) Head (Figs la, b): eyes sessile. Carapace (Figs la, b): without spines, rostrum straight or downcurved at tip, tapering distally, ventral margin with minute retrorse teeth distally, equal to, or greater than, length of peduncle of antenna 1 but not reaching to end of antenna I (excluding terminal aesthetascs and setae). Antenna 1 (Fig. Ic): peduncle bearing single flagellar segment with three aesthetascs distally, usually two narrow and one wide, occasionally 1 narrow and 2 wide. Antenna 2 (Fig. Id): exopodite as a broad lamina divided into 5 short segments distally, with 9 + 2 plumose setae on inner and distal margins. Endopodite of one segment (0-67 length of exopodite), with terminal plumose seta and short spine. Mandibles (Fig. le): asymmetrical. Maxillipeds 1-3 (Figs Ih-j): with natatory exopodites. Pereiopods 1,2 (Figs Ik, 1): rudimentary, biramous. Pereiopods 3-5: absent. Abdomen (Figs 1 a, b): somite 5 with posterior margin rounded, not produced into spines, somite six continuous with telson. No trace of pleopods. Telson (Fig. Ip): fans out distally, posterior margin bears 7 + 7 plumose spines, with minute spines between four innermost spines. ZOEA 2 (Fig. 2) 3-7 mm (3-5-3-9 mm) Head (Figs 2a, b): eyes stalked. Carapace (Figs 2a, b): one dorso-medial and a pair of supra-orbital spines all bent forward with small retrorse teeth, rostrum without teeth, downturnedat end to form small hook. Antenna 1 (Fig. 2c): two peduncle segments, stylocerite forming on proximal external margin of first segment; single flagellar segment with four terminal aesthetascs, two wide and two narrow. Antenna 2 (Fig. 2d): exopodite with 4 or 5 short segments distally. Pereiopods 1 , 2 (Figs 2k, 1): developed with natatory exopodite. Pereiopods 3, 4 (Figs 2m, n): rudimentary, biramous. Pereiopod 5 (Fig. 2o): rudimentary, uniramous. Abdomen (Figs 2a, b): somite 5 with posterior margin produced into a pair of conspicuous spines, somite 6 continuous with telson. Telson (Fig. 2p): developing uropods visible beneath exoskeleton alongside telson proper; in central group of small spines, one pair longer than others. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 129 Fig. 1 Zoea 1: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles; (f)maxilla 1; (g) maxilla 2; (h) maxilliped 1; (i) maxilliped 2; (j) maxilliped 3; (k) pereiopod 1; (1)pereiopod 2; (m) telson. Bar scales: a, b = 0-5 mm; c, d, h-m = 0-2 mm; g, f= 0-1 mm; e = 0-05mm. 130 A. A. FINCHAM Fig. 2 Zoea 2: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles; (0maxilla 1; (g) maxilla 2; (h) maxilliped 1; (i) maxilliped 2; (j) maxilliped 3; (k) pereiopod 1; (1)pereiopod 2; (m) pereiopod 3; (n) pereiopod 4; (o) pereiopod 5; (p) telson. Bar scales: a, b = 0-5mm;c, d, h-p = 0-2 mm;g, f=0-l mm;e = 0-05 mm. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 131 Fig. 3 Zoea 3: (a) dorsal view; (b) lateral view; (c) mandibles; (d) maxilla 1; (e) maxilla 2. Barscales: a,b = 0-5 mm; c = 0-05 mm; d, e = 0-l mm. ZOEA 3 (Figs 3,4) 4-0 mm (3-7-4-3 mm) Carapace (Fig. 3a, b): two dorso- medial spines and a small fronto-lateral spine at edge of carapace beneath the eyes, former with retrorse teeth ventrally. Antenna 1 (Fig. 4a): conspicuous spine medially, stylocerite more pronounced; distal segment of peduncle bearing first segment of internal flagellum, single segment of external flagellum bearing 3 wide aesthetascs distally. Antenna 2 (Fig. 4b, c): exopodite with distal part divided into 3 short segments; endopodite of 3 segments. 132 A. A. FINCHAM Fig. 4 Zoea 3: (a) antenna 1 ; (b) antenna 2; (c) distal part of exopodite of antenna 2; (d) maxilliped1; (e) maxilliped 2; (0 maxilliped 3; (g) pereiopod 1; (h) pereiopod 2; (i) pereiopod 3; (j)pereiopod 4 ;(k) pereiopod 5;(l)telson. Bar scales: a, b, d-I=0-2 mm; c = 0-05 mm. LARVAL DEVELOPMENT OF P. (P.) SERRA TVS 1 33 Abdomen (Figs 3a, b, 41): somite 6 divided from telson by suture. Uropod endopodite rudimentary with no marginal setae, exopodite with marginal setae. Telson (Fig. 41): narrower but still broader distally, outer pair of spines on posterior margin reduced. ZOEA 4 (Figs 5, 6) 4-5 mm (4-1-5-0 mm) Carapace (Fig. 5a, b; Fig. 6a, b): three dorso- medial spines with retrorse teeth ventrally; pair of small fronto-lateral spines at edge of carapace beneath the eyes; rostrum still downturned at end to form a small hook. Antenna 1 (Fig. 5c): single segment of external flagellum bearing 3 wide and 1 narrow aesthetascs distally. Antenna 2 (Fig. 5d): distal part of exopodite no longer divided into segments. Pereiopod 3 (Fig. 6h): developed with natatory exopodite. Pereiopod 4 (Fig. 6i): rudimentary, biramous. Pereiopod 5 (Fig. 6j): developed, uniramous. Abdomen (Figs 5a, b): endopodite and exopodite of uropod both with marginal plumose setae. Telson (Fig 6k): narrower but still broader distally; posterior margin weakly concave with 4 + 4 large spines, the 3 outer spines on the latero-distal margin reduced (outer pair sometimes absent). ZOEA 5 (Figs 7-9) 5-5 mm (5-2-5-8 mm) Carapace (Figs 7a, b): rostrum still downturned to form a small hook. Antenna 1 (Fig. 8a): rudiment of 'circular statocyst visible on first segment of peduncle. Antenna 2 (Fig. 8b): endopodite of 3 or 4 segments. Pereiopods 1 , 2 (Figs 9a, b): endopodite with internal distal margin of propodus produced slightly forward (will become fixed finger of chela). Pereiopod 4 (Fig. 9d): developed, exopod with rudimentary fringing setae. Telson (Fig. 9f): a little broader distally than proximally; spine formula as in Zoea 4 but with further reduction or even loss of small outer spines. Abdomen (Fig. 7b): somites 1-5 with rudimentary pleopods. ZOEA 6 (Figs 10-12) 5-7 mm (5-2-7-4 mm) Carapace (Figs lOa, b): rostrum weakly hooked at tip; short plumose seta in angle of anteriordor so- medial spine. Antenna 1 (Fig. 11 a): single external flagellum with four aesthetascs distally, additionalgroup of 2 or 3 narrow aesthetascs on internal margin. Antenna 2 (Fig. lib): increase in number of segments of endopodite flagellum,approximately equal to scaphocerite in length, small spine on distal margin of pedunclesegment. Maxilla 2 (Fig. 1 le): occasional increase in number of setae on basis 1 .Maxilliped 1 (Fig. 1 If): one plumose seta on proximo-lateral margin of exopod.Pereiopods 1, 2 (Figs 12a, b): endopodite with internal distal margin of propodus producedforeward to almost half length ofdactylus (excluding terminal setae).Pereiopod 4 (Fig 12d): exopodite occasionally with fringing plumose setae reduced.Abdomen (Fig. lOb): pleopods on somites 1-5 rudimentary, biramous. ZOEA 7 (Figs 13-15)6-1 mm (5-5-6-7 mm) Carapace (Figs 1 3a, b): rostrum straight or weakly hooked at tip; 2 or 3 short plumose setae in angle of anterior dorso- medial spine. Antenna 1 (Fig. 14a): two or occasionally three groups of aesthetascs on internal margin of external flagellum. Maxilla 2 (Fig. 14e): up to 6 setae on basis 2. Maxilliped 1 (Fig. 14f): 2-5 plumose setae on proximo-lateral margin of exopodite. Pereiopods 1, 2 (Figs 15 a, b): endopodite with internal distal margin produced forward to half length ofdactylus (excluding terminal setae). 134 A. A. FINCHAM Fig. 5 Zoea 4: (a) dorsal view; (b) lateral view; (c) antenna 1; (d) antenna 2; (e) mandibles; (0maxilla l;(g) maxilla 2. Bar scales: a, b = 0-5 mm;c, d = 0-2 mm;e = 0-05 mm; f, g = 0-l mm. LARVAL DEVELOPMENT OF P. (PJSERRATUS 135 Fig. 6 Zoea 4: (a) tip of rostrum; (b) fronto-lateral corner of carapace; (c) maxilliped 1; (d)maxilliped 2; (e) maxilliped 3; (f) pereiopod 1; (g) pereiopod 2; (h) pereiopod 3; (i) pereiopod 4;(j)pereiopod 5;(k)telson. Bar scales: a, b = 0-05 mm;c-k = 0-2 mm. 136 A. A. FINCHAM Fig. 7 Zoea 5: (a) dorsal view; (b) lateral view. Bar scales: 0-5 mm. LARVAL DEVELOPMENT OF P. (P.) SERRA TVS 137 Fig. 8 Zoea 5: (a) antenna 1; (b) antenna 2; (c) mandibles; (d) maxilla 1; (e) maxilla 2; (f)maxilliped 1 ; (g) maxilliped 2; (h) maxilliped 3. Bar scales: a, b, f-h = 0-2 mm; c = 0-05 mm; d,e = 0-l mm. 138 A. A. FINCHAM Fig. 9 Zoea 5: (a) pereiopod 1 ; (b) pereiopod 2; (c) pereiopod 3; (d) pereiopod 4; (e) pereiopod 5; (f) telson. Bar scale: 0-2 mm. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 139 Fig. 10 Zoea 6: (a) dorsal view; (b) lateral view. Bar scales: 0-5 mm. 140 A. A. FINCHAM Fig. 11 Zoea 6: (a) antenna 1; (b) antenna 2; (c) mandibles; (d) maxilla 1; (e) maxilla 2; (0maxilliped 1; (g) maxilliped 2; (h) maxilliped 3. Bar scales: a, b, f-h = 0-2 mm; c=0-05 mm; d,e = 0-l mm. Pereiopod 4 (Figs 1 5d, e): exopodite occasionally with fringing plumose setae much reduced.Abdomen (Figs 13b, 15g-k): pleopods 1-5 still rudimentary, biramous with traces ofterminal setae on exopodites.Telson (Figs 1 3a, 1 5 1 ): posterior margin straight or slightly convex. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 141 Fig. 12 Zoea 6: (a) pereiopod 1 ; (b) pereiopod 2; (c) pereiopod 3; (d) pereiopod 4; (e) pereiopod 5; (Otelson. Bar scale: 0-2 mm. ZOEA 8 (Figs 16-1 9) 6-6 mm (5-8-7-1 mm) Maxilliped 1 (Fig. 1 8a): up to 6 plumose setae on proximo- lateral margin ofexopodite. Pereiopod 1 , 2 (Figs 1 8d, e): endopodite with internal distal margin produced forward to over half length ofdactylus (excluding terminal setae). Telson (Fig. 1 9i): posterior margin convex. 142 A. A. FINCHAM Fig. 13 Zoea 7: (a) dorsal view; (b) lateral view. Bar scales: 0-5 mm. ZOEA 9 (Figs 20-23) 7-3 mm (6-9-7-9 mm) Carapace (Fig. 20b): 3 short plumose setae in angle of anterior dorso- medial spine, 1 in angle of second dor so- medial spine. Antenna 1 (Fig. 2 la): internal flagellum of 1 or 2 segments, external flagellum of 2 segments; 3 or 4 groups of aesthetascs on internal margin of external flagellum; statocyst fully developed. Antenna 2 (Fig. 21b): increase in number of segments of endopodite flagellum, now longer than scathocerite. Maxilliped 1 (Fig. 21g): 14 to 20 setae on internal margin of basis, 5 to 9 plumose setae on proximo- lateral margin ofexopodite. LARVAL DEVELOPMENT OF P. (P.) SERRA TVS 143 Fig. 14 Zoea 7: (a) antenna 1; (b) antenna 2; (c) mandibles; (d) maxilla 1; (e) maxilla 2; (0maxilliped 1; (g) maxilliped 2; (h) maxilliped 3. Bar scales: a, b, f-h = 0-2 mm; c=0-05 mm; d,e = 0-l mm. 144 A. A. FINCHAM Fig. 15 Zoea 7: (a) pereiopod 1 ; (b) pereiopod 2; (c) pereiopod 3; (d) pereiopod 4; (e) variant ofpereiopod 4 exopod with reduced setae; (0 pereiopod 5; (g) pleopod 1 ; (h) pleopod 2; (i) pleopod3; (j) pleopod 4; (k) pleopod 5; (1) telson. Bar scale = 0-2 mm. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 145 Fig. 16 Zoea 8: (a) dorsal view; (b) lateral view. Bar scales: 0-5 mm. 146 A. A. FINCHAM Fig. 17 Zoea 8: (a) mandibles; (b) maxilla 1; (c) maxilla 2; (d) antenna 1; (e) antenna1 - enlargement of proximal array of sensory hairs showing thread-like connections remaining incast exoskeleton; (0 antenna 2. Bar scales: a, e = 0-05 mm; b, c = 0- 1 mm; d, f = 0-2 mm. LARVAL DEVELOPMENT OF P. (P.) SERRA TUS 147 Fig. 18 Zoea 8: (a) maxilliped 1 ; (b) maxilliped 2; (c) maxilliped 3; (d) pereiopod 1 ; (e) pereiopod 2. Bar scale: 0-2 mm. 148 A. A. FINCHAM Fig. 19 Zoea 8: (a) pereiopod 3; (b) pereiopod 4; (c) pereiopod 5; (d) pleopod 1 ; (e) pleopod 2; (0pleopod 3; (g) pleopod 4; (h) pleopod 5; (i) telson. Bar scale: 0-2 mm. LARVAL DEVELOPMENT OF P. (P.) SERRA TVS 149 Fig. 20 Zoea 9: (a) dorsal view; (b) lateral view. Bar scales: 0-5 mm. Pereiopods 1, 2 (Figs 22c, d): endopodite with immovable finger of propodus produced forward to almost length ofdactylus (excluding terminal setae) Abdomen (Figs 20b, 23c-g): pleopods with rudimentary setae, a few fully plumose; endopodite of pleopods 2 to 5 with rudiment of appendix interna (stylamblys). Telson (Fig. 23h): further narrowing distally, posterior margin with 4 + 4 large spines (no small spines between) and with 3 reduced spines on latero-distal margin. 150 A. A. FINCHAM Fig. 21 Zoea 9: (a) antenna 1 ; (b) antenna 2; (c) antero-lateral teeth of carapace; (d) mandibles; (e)maxilla l;(f) maxilla 2; (g) maxilliped 1. Bar scales: a, b, g = 0-2 mm; c, d = 0-05 mm; e, f=0-lmm. LARVAL DEVELOPMENT OF P. (P.) SERRATUS 151 Fig. 22 Zoea 9: (a) maxilliped 2; (b) maxilliped 3; (c) pereiopod 1; (d) pereiopod 2; (e) pereiopod 3. Barscale = 0-2 mm. 152 A. A. FINCHAM Fig. 23 Zoea 9: (a) pereiopod 4; (b) pereiopod 5; (c) pleopod 1; (d) pleopod 2; (e) pleopod 3; (0pleopod 4; (g) pleopod 5; (h) telson. Bar scale: 0-2 mm. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 153 Fig. 24 Post larva 1: (a) rostrum; (b) rostrum with characters intermediate between larval andpost larval phases; (c) antenna 1; (d) antenna 2; (e) mandible; (0 maxilla 1; (g) maxilla 2. Barscales: a, b = 0-5 mm;c, d = 0-2 mm; e = 0-05 mm; f, g = 0-l mm. 154 A. A. FINCHAM Fig. 25 Post larva 1 : (a) maxilliped 1 ; (b) maxilliped 2; (c) maxilliped 3; (d) pereiopod 1 chela; (e)pereiopod 2 chela; (0 pereiopod 3; (g) pereiopod 4; (h) pereiopod 5. Bar scale: 0-2 mm. POST LARVA 1 (Figs 24-26) 7-9 mm (7-0-8-6 mm) Most specimens had metamorphosed to post larvae or intermediate stages at thismoult. Meristic characters of the more advanced individuals are described here and includedin Table 1. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 155 Fig. 26 Post larva: (a) pleopod 1; (b) pleopod 2; (c) pleopod 3; (d) pleopod 4; (e) pleopod 5; (0somite 5 with postero-lateral spines (plumose setae of pleopod 5 not shown); (g) uropodexopodite (plumose setae not shown); (h) telson. Bar scale: 0-2 mm. Fig. 27 Telson of palaemonid Zoea 1 . (a) Dorsal view of posterior margin. Bases of large plumosespines with row of smaller spines between, (b) Single large spine ornamented with spinules. (c)Articulation at base of large spines with circlet of small spines, (d) Ventral view of posteriormargin showing rows of small spines arranged with gaps. LARVAL DEVELOPMENT OF P. (P.)SERRATUS 157 Fig. 28 Single plumose seta from natatory exopodite of pereiopod. Shaft is a differentiallythickened cylinder hinged at intervals and bears two rows of thin lateral setules. (a) Setastationary, (b) Beginning of recovery stroke, (c) End of recovery stroke. Bar scale: 0-05 mm. Carapace (Fig 24a): rostrum with 6 or 7 dorsal and 2 ventral teeth, rostral tip straight. Supra- orbital spines missing. Intermediate condition of rostrum shown in Fig. 24b. Antenna 1 (Fig. 24c): internal Jlagellum of 5 or more segments and external Jlagellum of 6 or more segments. Antenna 2 (Fig. 24d): endopodite multisegmented, at least twice as long as scaphocerite. Mandible (Fig. 24e); divided into pars incisiva and pars molaris, lacinia mobilis no longer present and palp (three jointed in adult) not yet appeared. Maxilla 2 (Fig. 240: reduction of setae on coxal endite and endopodite, increase in setae on basal endites. Maxilliped 2 (Fig. 25b): endopodite with dactylus, propodus and merus flattened. Maxilliped 3 (Fig. 25c): endopodite dactylus shortened. Pereiopods 1 , 2 (Figs 25d, e): immovable finger of propodus same length as dactylus. Pereiopods 3, 4 (Figs 25f, g): dactylus shortened. Pleopod 1 (Fig. 26a): ratio of endopodite to exopodite 1:4; endopodite bearing terminal plumose setae, exopodite fringed with long plumose setae. Pleopods 2-5 (Figs 26b-e): endopodite over half length of exopodite, both with long, marginal plumose setae, endopodite with appendix interna bearing well developed intero- distal coupling hooks. Abdomen (Fig. 260: fifth abdominal somite with posterior margin still produced into pair of lateral spines. 158 A. A. FINCHAM Telson (Fig. 26g): intermediate condition (extreme narrowing with posterior margintapering to point not yet developed). Discussion Most of the specimens reared by Sollaud (1912) at the Marine Laboratory in Roscoff reachedmetamorphosis after 8 larval stages. But he found also '. . . quelques individus, en effet,peuvent presenter un stade IX supplementaire avant de se transformer. . .' He concluded,however, that the normal larval development for Palaemon serratus was eight stages(Sollaud, 1923). In the present work the differences between stages 8 and 9 were slight andconfirm, therefore, the conclusion reached by Sollaud. The insertion of extra moults whichprobably represents an adaptation to sub-optimal conditions by prolonging larval existenceis a recurrent feature of the development of Palaemoninae (Fincham, 1977, 1979). Sandiferand Smith (1979) indicated that in addition to being affected by environmental factors thetendency of individual palaemonid larvae to pass through a given number of larval stagesmay be inherited. They suggested also that variation in development may enhance thegeneral advantages of an extended planktonic larval phase, a common feature of 'r'strategists. These include greater potential for dispersion and the ability to colonize newhabitats quickly if favourable conditions occur. A variable planktonic existence mayproduce a wider spread of individuals of a given brood and average the risks of survival. Atthe population level Sandifer and Smith emphasize that early metamorphosis will enhancethe possibility of gene flow between populations. Many environmental factors affect development including photoperiodicity. Wickins(1972) reported work on larval Palaemon serratus in which growth was improved andmetamorphosis reached sooner in those reared in continuous light. Eight hours dark and 16hours light produced improved growth compared with 8 hours light and 16 hours dark;continuous darkness produced slowest growth and development. Dalley (1979) working withPalaemon elegans concluded that greatly increased mortality during larval development innon-circadian light regimes was due to desynchronization of the circadian rhythms ofmetabolic processes. One feature commonly found in exuviae of the various larval stages are thin strandsextending from the bases of sensory hairs (see Figs 17d, e). In a paper on the ultrastructure ofthe antennal sensilla of the shrimp Acetes Ball & Cowan (1977) describe dense strands ofunknown composition crossing the base of their type 1 seta, which they consideruninnervated. Tracts leading from the bases of their other four types of seta contain axons.The exact nature of the strands in the exuviae of Palaemon serratus is not clear and will beexamined further. Swimming: adaptations and a mechanism When palaemonid larvae first hatch they usually swim upside down and telson first. At restthe larvae tend to sink and there are several adaptations which slow the rate of descent. Ateither end of the body plumose setae or spines increase drag. Fringing plumose setaeeffectively double the area of the broad exopodite (scaphocerite) of antenna 2 at the anteriorend of the body (Figs la, 2a, 3a, 5a, 7a, lOa, 13a, 16a, 20a). These antennal setae arethickened at the base (Fig. 4c) and their rigidity increases drag when the exopodites arespread. At the posterior end the telson bears stout spines which are ornamented with rows of smallspines and are also plumose (Figs 27a-d). The function of the ornamentation is uncertain butmight further slow the rate of sinking. Body size increases as larval development progressesand the importance of even the large spines on the telson for increasing drag, diminisheswhen the sixth abdominal segment develops its appendages. The exopodites of these uropods LARVAL DEVELOPMENT OF P. (P.) SERRA TVS \ 59 appear at zoea 3 (Fig. 41) and endopodites at zoea 4 (Fig. 6k) and are broad and flat andfringed with plumose setae. The spreading of these uropods slows the rate of sinking duringperiods when the larva is not swimming. The telson, however, retains its important function as a stabilizing hydrofoil throughoutlarval life and is supplied with powerful muscles that enable it to assist in orientation control.Its gradual transformation from a triangular shape at zoea 1 with a 7 + 7 spine formula (Figs1m, 27a-d) to the narrow shape at the last larval stage with spine formula 4 + 4 and threepairs of lateral spines (Fig. 23h) accompanies the gradual development of more thoracicnatatory exopodites. In the present rearing programme food is supplied to the larvae in the form of Anemianauplii. In the wild it is likely that copepods and other small planktonic organisms form themain source of food (Sollaud used copepods in his rearing work at Roscoff). The pursuit ofprey- Anemia or copepods - necessitates accurate orientation and direction control. Thepropulsive locomotory force is provided by the natatory exopodites developed sequentially.Zoea 1 has only three natatory exopodites - those of the maxillipeds-but biramouspereiopods are developed later and are also integrated into a regular beating pattern of thelimbs. Pereiopod 5 is in fact developed by stage 4 before pereiopod 4 in Palaemon serratusbut has no exopodite and is therefore not involved in larval swimming. The effective area ofthe propulsive exopodites is extended by the fringing plumose setae. Preliminary analysis ofthe swimming action from cine film shows that there is a power stroke and a recovery stroke.The morphological adaptations of these plumose setae and their role in swimming behaviourwas determined following the routine examination by light microscopy of hundreds ofmoults during the course of the rearing programme. The shaft of the plumose setae fringing the natatory exopodites is a differentially thickenedcylinder with flexible cuticular hinges at intervals along its length (Fig. 28a). The hinges onlypermit bending towards the unthickened side of the shaft. During the propulsive powerstroke the setae remain straight with the two rows of thin, lateral setules set at an obtuse angleto the shaft, thus providing maximum surface area and purchase in the water. On therecovery stroke the flexible exopodite bends and the marginal plumose setae fold back alongthe many hinge lines with their thin side branches streaming out behind. This offers the leastpossible resistance by the exopodite to the water and repositions the limb ready to begin thenext power stroke. The rhythmic beating of the six pairs of setose thoracic exopodites in the three larvalstages, and swimmerets or pleopods in the post larvae of lobsters, has been analysed fromcine film by Neil et al. (1976), Macmillan et al. (1976) and Laverack et al. (1976). Nocomparable study has been made for a caridean with regular sequential addition of limbs (inlobsters all limbs are present on hatching). The rarity of palaemonid larvae in the planktonremains an enigma especially as they are apparently so well adapted for swimming. Detailedanalyses of cine film of swimming in larval palaemonids, together with plankton samplingusing a static bottom net and experimental work on the periodicity of larval swimming, all ofwhich are in progress, should shed some light on the problem. References Adensamer, T. 1898. Decapoden gesammelt auf S.M. Schiff Pola in den Jahren 1890-1894. Berichte der Commission fur Erforschung des ostlichen Mittelmeeres. XXII Zoologische Ergebnisse. XI. Denkschr. Akad. Wiss. Wien6S : 597-628.Ball, E. E. & Cowan, A. N. 1977. Ultrastructure of the antennal sensilla of Acetes (Crustacea, Decapoda, Natantia, Sergestidae). Phil Trans. R. Soc. Lond. B 277 : 429-456.Couch, R. Q. 1845. On the metamorphosis of the crustaceans, including the Decapoda, Entomostraca and Pycnogonidae. Rep. R. Cornwall polytech. Soc. 12 : 17-46.Czerniavsky, W. 1884. Crustacea Decapoda Pontica littoralia, Materialia ad Zoographiam Ponticam comparatum. II. Trans. Soc. Univ. Kharkow, 13 (suppl.) : 1-268. 160 A. A. FINCHAM Dalley, R. 1979. Effects of non-circadian light cycles on the survival and development of Palaemon elegans Rathke reared in the laboratory. In: Proc. 13th Eur. mar. Biol. Sym. 13 : 157-163 (Eds E. Naylor& R. Hartnoll. Oxford: Pergamon Press). Edwards, H. Milne 1837. Histoire naturelledes Crustaces. II. Paris. 53 1 pp.Fincham, A. A. 1977. Larval development of British prawns and shrimps (Crustacea: Decapoda: Natantia). 1. Laboratory methods and a review of Palaemon (Palaeander) elegans Rathke, 1837. Bull. Br. Mus. nat. Hist. (Zool.)31 (1) : 1-28.1978. Larval development of British prawns and shrimps (Crustacea: Decapoda: Natantia). 2. Palaemonetes(Palaemonetes) varians (Leach, 18 14) and morphological variation. Bull. Br. Mus. nat. ///s/.(Zool.)35(2): 163-182. 1979. Larval development of British prawns and shrimps (Crustacea: Decapoda: Natantia). 3. Palaemon (Palaemon) longirostris H. Milne Edwards, 1837 and the effect of antibiotic on morphogenesis. Bull. Br. Mus. nat. Hist. (Zool.) 37 (1 ) : 1 7^46.Fincham, A. A. & Williamson, D. I. 1978. Crustacea, Decapoda: Larvae. VI Caridea, Families Palaemonidae and Processidae. Ficht. Ident. Zooplancton, 159/160: 8 pp.Forward, R. B. & Cronin, T. W. 1978. Crustacean larval phototaxis: possible functional significance. In: Proc. 12th Eur. mar. Biol. Sym. 12:253-261 (Eds D. S. McLusky & A. J. Berry. Oxford: Pergamon Press).Foxon, G. E. H. 1934. Notes on the swimming methods and habits of certain crustacean larvae. J. mar. biol. Ass. U.K. 19 : 829-849. Gurney, R. 1942. Larvae of decapod Crustacea. 306 pp. Ray Society, London.Hardy, A. C. & Bainbridge, R. 1954. Experimental observations on the vertical migrations of plankton animals. J. mar. biol. Ass. U.K. 33 : 409-448.Holthuis, L. B. 1950. The Palaemonidae collected by the Siboga and Snellius expeditions with remarks on other species. 1 . Subfamily Palaemoninae. Siboga Exped. 39a : 1-268.Keeble, F. & Gamble, F. W. 1904. The colour physiology of the higher Crustacea. Phil. Trans. R. Soc. Lond. 8196:295-388.Laverack, M. S., Macmillan, D. L. & Neil, D. M. 1976. A comparison of beating parameters in larval and post-larval locomotion systems of the lobster Homarus gammarus (L.). Phil. Trans. R. Soc. Lond. 8274:87-99.Macmillan, D. L., Neil, D. M. & Laverack, M. S. 1976. A quantitative analysis of exopodite beating in the larvae of the lobster Homarus gammarus (L.). Phil. Trans. R. Soc. Lond. B 274 : 69-85.Man, J. G. De 1915. On some European species of the genus Leander Desm., also a contribution to the fauna of Dutch waters. Tijdschr. ned. dierk. Vereen. Ser. 2, XIV : 115-1 79.Mayer, P. 1877. Zur Entwicklungsgeschichte der Dekapoden. Jena. Z. Med. Naturw. 11 : 188-269.Neil, D. M., Macmillan, D. L. & Laverack, M. S. 1976. The structure and function of thoracic exopodites in the larvae of the lobster Homarus gammarus (L.). Phil. Trans. R. Soc. Lond. B 274 : 53-68. Pennant, T. 1777. British Zoology. 4 : 136 pp.Reeve, M. R. 1969. Growth, metamorphosis and energy conversion in the larvae of the prawn, Palaemon serratus. J. mar. biol. Ass. U.K. 49 : 77-96.Risso, A. 1816. Histoire naturelle des Crustaces des environs de Nice. Paris a la libraire Greque-Latine-Allemande. 175pp. A. 1 826. Histoire naturelle des principales productions de {'Europe meridionale. 5 : 403 pp. Russell, F. S. 1925. The vertical distribution of marine macroplankton. An observation on diurnal change. /. mar. biol. Ass. U.K. 13 : 769-809.1927. The vertical distribution of marine macroplankton. V. The distribution of animals caught in the ring-trawl in the daytime in the Plymouth area. J. mar. biol. Ass. U.K. 14 : 557-608.Sandifer, P. A. & Smith, T. I. J. 1979. Possible significance of variation in the larval development of palaemonid shrimp. J. exp. mar. Biol. Ecol. 39 : 55-64. Savage, R. E. 1926. The plankton of a herring ground. Fish. Invest. Lond. 9 : 1-35.Sharp, B. 1893. Catalogue of the crustaceans in the museum of the Academy of Natural Sciences of Philadelphia. Proc. Acad. nat. Sci. Philad. 104-127.Sollaud, E. 1912. Les metamorphoses du 'Bouquet', Leander serratus Pennant. C. R. Acad. sci. Paris 154 : 664-666.1921. Le comportment des larves de Palaemonetes varians microgenitor Boas. Changement de signe du phototropisme apres la metamorphose. Bull. Ass. fr. Avanc. Sci. Congres de Rouen 1921:671-673.1923. Le developpement larvaire des Palaemoninae. Bull. biol. Fr. Belg. 57 : 509-603. LARVAL DEVELOPMENT OF P. (P.) SERRATUS 161 Sui ton, A. H., Main, G. & Ronald, A. 1969. An instrument for counting the larvae of the prawn Palaemon serratus and the brine shrimp Anemia salina. Lab. Pract. 18 : 433^36.Thompson, W. V. 1836. Memoir on the metamorphosis in the Macrourae or long-tailed Crustacea, exemplified in the prawn (Palaemon serratus). Edinb. New phil. J. 21 : 221-223.Wickins, J. F. 1972. Developments in the laboratory culture of the common prawn Palaemon serratus Pennant. Fishery Invest. Lond. Series 2 27 (4) : 1-23.Williamson, H. C. 1915. Nordisches Plankton. VI. Crustacea Decapoda. Larven. Nord. Plankt. 6:315-588. Manuscript accepted for publication 1 1 June 1982. The larval development of the Angular Crab,Goneplax rhomboides (Linnaeus) (Decapoda:Brachyura) R. W. Ingle & Paul F. Clark Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Introduction The Angular Crab, Goneplax rhomboides (Linnaeus) occurs in the Eastern Atlantic Oceanand Mediterranean Sea. It has been reported from the northern Irish Sea to as far south as theCape region of South Africa (see Barnard, 1950 : 285; Ingle, 1980 : 109); the species occursfrom the lower shore to depths of about 700 m. Capart (1951 : 1 69) discussed regional variation of this species with respect to the degree ofdevelopment of the posterior pair of carapace anterio-lateral teeth and differences in shapesof the male first pleopod. He suggested two 'varieties' may exist. Specimens attributed to G.rhomboides (L.) have a very rudimentary pair of posterior teeth and are reported only fromthe Mediterranean Sea, coast of Mauritania and the Canary Islands whereas materialassigned to G. angulata (Pennant) has been recorded from various Atlantic Ocean localities(see Manning & Holthuis, 1981 : 164). The larval stages of G. rhomboides have been described previously (see larval andpost-larval references below), but these accounts are generally inadequate for use in detailedcomparative studies of larval morphology. The recent rearing of G. rhomboides to third crabstage has provided sufficient material for redescribing the complete larval development ofthis species and an account is given here of the four zoeal and megalop stage. Materials and Methods After several unsuccessful trawling attempts, SCUBA diving was used to collect ovigerousmaterial. On the advice of Alan Howard (MAFF) members of the BM(NH) Diving Unitsearched the sandy-mud substrate to a depth of 18m off Shoalstone Point (SX937568),Brixham, Devon. An ovigerous crab was collected on 10 July 1981 and transported to therearing laboratory of the Crustacea Section, BM(NH). The larvae were reared using methodsdescribed by Rice & Ingle (1975) and Ingle & Clark (1977), except that sea water wasuntreated. Drawings and measurements were made with the aid of a camera lucida.Measurements are as follows: total lengths of zoeae (T.T.) measured from tip of dorsal to tipof rostral spine and carapace length (C.L.) measured from between eyes to posterio-lateralmargin of carapace (for zoeae) and from rostral tip for megalop. All material was fixed inSteedman's preservative (Steedman, 1976 : 148) and later transferred to 70% ethanolalcohol. The female and larval stages are deposited in the collections of the BM(NH),accession numbers 1981 : 540 & 1982 : 55 respectively. Descriptions Goneplax rhomboides (Linnaeus, 1 758) non Gonoplax rhomboides:-Cano, 1891, Tav. XI, figs ID (or ?E), IX d , IX e (megalops); Brachynotussexdentatus:-Cano, 1891, Tav. XI, fig. IF (crab stage); non Gonoplax rhomboides:-Wi\\iamson, 1915, Bull. Br. Mm. not. Hist. (Zool.) 44(2): 1 63-1 77 Issued 24 February 1983 164 R. W. INGLE & P. F. CLARK Fig. 1 Goneplax rhomboides (L.): a-d lst-4th zoeae; e megalopal carapace from left lateralaspect f; megalop from dorsal aspect; g moult of megalopal abdomen slightly flattened to showsetation; scale, each division = 0- 1 mm. LARVAL DEVELOPMENT OF ANGULAR CRAB Goneplax rhomboides (L.): a-d antennule of 1 st-4th zoea respectively and e of megalop;f-i antenna of lst-4th zoea and j of megalop respectively; scale = 0- 1 mm. fig. 398 (after Cano); Gonoplax angulata:- Caroli, 1927: 161 (lst-4th zoeae, megal., describ.);Gonoplax rhomboides:- Lebour, 1928 : 534, figs 4 (6-9), 5 (22-24), PI. II, fig. 6, PI. XI, fig. 10, PI. XII,figs 1-4 (lst-4th zoeae, megal., lst-4th crab); Gonoplax angulata.-Bourdi lion-Casanova, 1960 : 180,figs 57a-c (1st zoea, megal.); Goneplax rhomboides:- Rice & Williamson, 1977 : 55, fig. 29 (3rd zoea). FIRST ZOEA Dimensions: 1.1. 1-5 mm,C.L. 0-5 mm. Carapace (Fig. la): Dorsal, rostral and lateral spines present; a pair of posterio-dorsal 166 R. W. INGLE & P. F. CLARK Fig. 3 Goneplax rhomboides (L.): a-c abdomen and telson of lst-3rd zoea respectively fromdorsal aspect and d-f same from lateral aspect; g, h left half of mandible of 1st and 4th zoearespectively (drawn from scanning EM photographs); scale, each division = 0-1 mm except LARVAL DEVELOPMENT OF ANGULAR CRAB 167 Fig. 4 Goneplax rhomboides (L.): a abdomen and telson of 4th zoea from dorsal aspect; b ofanother specimen from lateral aspect; c, d 1st maxilliped of 1st and 2nd zoea respectively; scale,each division = 0-1 mm. setules; dorso-median elevation present; posterior margin of carapace minutely serrate and with 3-4 setules. Eyes: Partly fused to carapace. Antennule (Fig. 2a): Exopod unsegmented, with 2 terminal aesthetascs and one seta. Antenna (Fig. 2f): Exopod with very minute spinules distally and with 2 median spinules and 2 setules; spinous process distally spinulate, slightly longer than exopod. 168 R. W. INGLE & P. F. CLARK Fig. 5 Goneplax rhomboides (L.): a, b 1st maxilliped endopods of 3rd and 4th zoea respectively;c, d 2nd maxillipeds of 1st and 2nd zoea and e, f 2nd maxilliped endopods of 3rd and 4th zoearespectively; g maxillule of 1 st zoea; scale = 0- 1 mm. LARVAL DEVELOPMENT OF ANGULAR CRAB 169 Mandible (Fig. 3g): Incisor and molar processes developed. Maxillule (Fig. 5g): Endopod 2 -segmented, proximal segment with one seta, distal with 2 sub-terminal and 4 terminal setae; basal endite with one seta and 4 spines on distal margin; distal and inner margins of coxal endite with a total of 6 setae. Maxilla (Fig. 7a): Scaphognathite with 4 long plumose setae and one distal stout posterior process; endopod bilobed, with 5 + 3 setae; basal endite unequally bilobed, with 4 + 5 setae; coxal endite bilobed with 4 + 4 setae. First maxilliped (Fig. 4c): Exopod incipiently 2-segmented, with 4 terminal plumose setae; endopod 5-segmented, with 3,2, 1,2,4+1 setae; margin of basis with 2, 2, 3, 3 setae. Second maxilliped (Fig. 5c): Exopod incipiently 2-segmented, with 4 terminal plumose setae; endopod 3-segmented, with 1, 1,4+1 setae; margin of basis with 4 setae. Third maxilliped'. not developed. Pereiopods: not developed. Abdomen (Figs 3a, d): 5-segmented + telson, segments 2-4 each with a pair of lateral processes decreasing in size on each respective segment; posterio-lateral margins of segments with minute denticles as shown in inset to Fig. 3d; margin of segment 2 produced and rounded, those of 3-5 with acute processes; each posterio-dorsal margin of segments 3-5 with minute denticles and of 2-5 with a pair of small setules. Telson broad, one long dorsal and one lateral spine on each fork; posterior margin concave, with 3 spines on each outer half, outermost pair longest; middle portion of telson forks invested with minute spinules. SECOND ZOEA Dimensions: T.T. 1-9-2-0 mm, C.L. 0-6-0-7 mm. Carapace (Fig. Ib): Now with 2 pairs of anterio-dorsal setules, 4-6 setules on posterior margin and a prominent dorso-median elevation; eyes free. Antennule (Fig. 2b): Exopod now with 3 terminal aesthetascs. Antenna (Fig. 2g): Exopod setules longer than in previous stage, an incipient endopod bud present. Mandible: Unchanged. Maxillule (Fig. 6a): Endopod now conspicuously stepped distally; basal endite with a prominent plumose seta on outer margin, distal and inner margins with a total of 3 setae and 5 spines; distal and inner margins of coxal endite with a total of 7 setae. Maxilla (Fig. 7b): Scaphognathite now with 12 marginal setae. First maxilliped (Fig. 4d): Exopod now with 6 terminal plumose setae. Second maxilliped (Fig. 5d): Exopod now with 7 terminal plumose setae. Third maxilliped: represented as a small bud. Pereiopods: represented as small buds. Abdomen (Figs 3b, e): Dorsal surface of 1st segment with one seta, posterio-lateral margin now slightly produced, lateral processes on segment 2 and posterio-lateral processes on 3-5 longer than in previous stage. THIRD ZOEA Dimensions: T.T. 2-8-2-9 mm, C.L. 1-1-1-2 mm. Carapace (Fig. Ic): Now with 5 pairs of anterio-dorsal setules and 7-10 setules on posterior margin. Antennule (Fig. 2c): Exopod now with 3 setules and 3 setae. Antenna (Fig. 2h): Endopod bud well developed. Mandible: Incisor sub-divided. Maxillule (Fig. 6b): Basal endite now with 3 setae on distal margin and with a total of 9 setae on distal and inner margins of coxal endite. Maxilla (Fig. 7c): Scaphognathite now with 20 setae, basal endite with 5 + 5 and coxal with 4 + 5 setae. First maxilliped (Fig. 5a): Exopod now with 8 terminal plumose setae; distal segment of endopod now with 5 + 1 setae. 170 R. W. INGLE & P. F. CLARK Fig. 6 Goneplax rhomboides (L.): a-c maxillule of 2nd-4th zoea respectively; scale = 0- 1 mm. Second maxilliped (Fig. 5e): Exopod now with 8 terminal plumose setae; distal segment of endopod with 5 + 1 setae. Third maxilliped: represented as a conspicuous biramous bud. Pereiopods: rudimentary but conspicuous, first pair incipiently chelate. Abdomen (Figs 3c, 0: Now 6-segmented + telson; a minute lateral process on segment 5 in some specimens; posterio-lateral processes on segments 3-5 longer than in previous stage; LARVAL DEVELOPMENT OF ANGULAR CRAB 171 Fig. 7 Goneplax rhomboides (L.): a, b maxilla of 1 st and 2nd zoea; c, d endopod, basal and coxalendites of maxilla of 3rd and 4th zoea respectively; scale = 0- 1 mm. dorsal surface of 1st segment now with 3 setae; rudimentary paired pleopods on segments 2-5. FOURTH ZOEA Dimensions: T.T. 3-5-3-6 mm, C.L. 1-4-1-5 mm. Carapace (Fig. Id): Now with 8 or more pairs of anterio-dorsal setules, 2 pairs at base of 172 R. W. INGLE & P. F.CLARK rostral spine and sometimes a small setule on each eye; 12-1 5 setules on posterior margin of carapace. Antennule (Fig. 2d): Exopod now with 4 terminal aesthetascs and short setae; endopod represented as an incipient bud. Antenna (Fig. 2i): Exopod now with conspicuous distal spinules; endopod bud more than half length of exopod. Mandible (Fig. 3h): Incisor and molar processes sub-divided as shown. Maxillule (Fig. 6c): Distal and inner margins of basal endite now with a total of 7 setae and 7 spines; margins of coxal endite with a total of 10 setae. Maxilla (Fig. 7d): Scaphognathite now with 31 marginal setae; margins of basal endite with 6 + 6 and coxal with 4-1-6 setae respectively. First maxilliped (Fig. 5b): Exopod now with 9 terminal plumose setae; distal segments of endopod proportionally slightly longer than in previous stage. Second maxilliped (Fig. 50: Exopod now with 10-1 1 terminal plumose setae; segments of endopod proportionally longer than in previous stage. Third maxilliped'. more conspicuous than in previous stage. Pereiopods: more developed than in previous stage. Abdomen (Figs 4a, b): Segment 6 now with minute denticles on posterio-dorsal margin and with a pair of pleopods; segments 1, 2, 3 with 5, 4, and 3 dorsal setae respectively; pleopods biramous. Medio-posterior margin of telson with 3 setae, dorsal surface with a pair of median setae. MEGALOP Dimensions: C.L. 1-8-1-9 mm. Carapace (Figs le-f): Rostrum small, slightly deflected ventrally; mesogastric region with a prominent longitudinal carina; each half of protogastric region with a prominent curved spine; cardiac and intestinal regions with carinae and broad tubercles arranged as shown in Fig. If; margin of carapace with numerous small setules. Antennule (Fig. 2e): Peduncle 3-segmented, with 2 setae on each segment; exopod 4-segmented with 0, 5, 4, 3 aesthetascs and 0, 0, 2, 2 setae respectively; endopod unsegmented, with one sub-terminal and 5 terminal setae. Antenna (Fig. 2j): Peduncle 3-segmented, with 1,1,0 setae and flagellum 7-segmented with 1 , 0, 2, 5, 0, 4, 4 setae respectively. Mandible (Fig. 9e): Molar process now reduced, palp 3-segmented, with 0, 1, 8 setae respectively. Maxillule (Fig. 8a): Endopod now reduced and unsegmented, with 2 terminal setae; margins of basal endite with a total of 1 3 setae and 8 spines; margins of coxal endite with a total of 1 5 setae/spines. Maxilla (Fig. 8b): Scaphognathite with 51 marginal setae and 4 setae on dorsal surface, posterior margin sub-truncate; endopod reduced to a sub-acute lobe with setae on outer margin; basal endite with 8 + 7 marginal setae and with additional setae on dorsal and ventral surfaces as shown; coxal endite with 6 + 9-10 setae. First maxilliped (Fig. 8c): Coxal segment with 6-7 setae, basis with 26-28 setae; endopod represented as a broad sub-acute lobe invested with 3-4 setae; exopod 2-segmented, with 3 and 5 setae respectively; epipod well developed, with 6 long setae. Second maxilliped (Fig. 8d): Coxal segment hardly differentiated from basis, with 5-6 setae, ischium to dactylus differentiated, with 0, 3, 1, 4, 5 setae respectively in addition to 4 spines on dactylar margin; exopod 2-segmented, with 2 and 4 setae respectively; 2 setae at basis-exopod junction; epipod short, with 3 distal setae. Third maxilliped (\g. 9a): Coxa not differentiated from basis, with 4-5 setae as shown; outer margin of ischium with 1-2 broad, acute spines and with 19-20 setae; merus to dactylus well differentiated and with 1 1-12, 5, 7, 6 setae respectively; exopod 2-segmented, with 1 and 5 setae respectively; epipod long, with numerous short setae in proximal half (circa 18) and 1 5 long medio- to distally placed setae. LARVAL DEVELOPMENT OF ANGULAR CRAB 173 Fig. 8 Goneplax rhomboides (L.): megalop-a maxillule; b maxilla; c 1st maxilliped; d 2nd maxilliped; scale = 0-1 mm. Pereiopods (Figs 9b-d, lOa, b, i): Cheliped stout, invested with numerous setae as shown inFig. 9b; one large and one small ischial spine present; inner margin of propodal extensionwith 2-3 processes, inner margin of dactylus without processes. Pereiopods stout, setose asshown in Figs 9c, d & lOa, b; coxal-ischial segments of pereiopods 2-4 each with a welldeveloped spine; dactylus of 5th pereiopod with 3 long setae on inner distal margin. 174 R. W. INGLE & P. F. CLARK Fig. 9 Goneplax rhomboides (L.): megalop - a 3rd maxilliped; b left cheliped; c 2nd pereiopod; d3rd pereiopod; e mandible; each division of scale = 0-1 mm. Cephalothorax (Fig. lOi): Second to 4th sternites each with a prominent curved spine and aseta, first segment of sternum also with a small spine and numerous setae.Abdomen (Figs If, g & lOh): 6-segmented + telson; posterio-lateral margin of first segmentsub-acute, of 2nd truncate, of 3rd-5th acutely produced and of 6th sub-truncate. Surfaces ofsegments invested with numerous setae distributed as shown in Fig Ig. Well developed LARVAL DEVELOPMENT OF ANGULAR CRAB 175 Fig. 10 Goneplax rhomboides (L.): megalop-a-b 4th and 5th pereiopods and c-f lst-4thpleopods respectively; g telson and left uropod from dorsal aspect; h abdomen from right lateralaspect; i sternites and coxal-ischial segments of pereiopods from left side, ventral aspect; scale,each division = 0- mm. biramous pleopods on segments 2-5, exopods with 17, 17, 16, 15 long plumose setae onlst^4th pairs (Figs lOc-f) respectively; inner distal margin of endopod of each with 3coupling hooks. Uropods (Fig. lOg) well developed, distal segment with 7-8 long plumosesetae. Telson much broader than long, dorsal surface with a pair of lateral and 2 pairs ofmedian setules, ventral surface with 3-4 small setules. 176 R. W. INGLE & P. F. CLARK Remarks The present laboratory reared material of G. rhomboides differs in a number of features fromthe accounts given by Bourdil Ion-Casanova (1960) and Rice & Williamson (1977). Thesedifferences are tabulated below. Bourdillon-Casanova ZOEA! Denticles on proximal part of posterio-lateralmargin of abdominal segments very pronounced. MEGALOP Disto-lateral margins of rostrum very acute.Protogastric spines of carapace stout and straight.A pair of widely spaced tubercles on meta-branchial-intestinal regions.Groups of setae on posterior region of carapace Exopod of uropod with 8-9 setae Rice & Williamson ZOEA IIIAntennal exopod with a single mid-point seta. Scaphognathite of maxilla with 1 8-19 setae.Endopod of 2nd maxilliped with 1,1,5 setae. ZOEAE 1-IV Exopod terminal segment of 2nd maxilliped with4, 7,9,11 setae in respective stages. Present material ZOEA I Denticles on proximal part of posterio-lateralmargin of abdominal segments very minute. MEGALOP Disto-lateral margins of rostrum not acute. Protogastric spines of carapace thin and curved. A pair of tubercles placed near to median line on cardiac region. Without groups of setae on posterior region of carapace. Exopod of uropod with 7-8 setae. ZOEA III Antennal exopod with more than one seta/spinuleat mid-point. Scaphognathite of maxilla with 20 setae.Endopod of 2nd maxilliped with 1,1,6 setae. ZOEAE I-IV Exopod terminal segment of 2nd maxilliped with4, 7, 8, 10-1 1 setae in respective stages. With the exception of Geryon tridens (Kroyer), the zoeae of Goneplax rhomboides canbe distinguished from those of other known brachyrhynchs occurring in the N.E. Atlantic seaarea (see Ingle, 1980) by the following combined features: (1) A pair of small but prominentdorso-lateral processes on the 4th segment of the abdomen and sometimes a minute pair onthe 5th segment in the 3rd and 4th stages. (2) The antennal exopod with spinules and setaesub-terminally placed. Features separating zoeae of G. rhomboides from those of Geryontridens were tabulated by Ingle (1979 : 229). The following amendments must now be madeto this table with respect to G. rhomboides: (1) The antennal exopod-the spinous process islonger than the exopod in all stages. (2) Maxilla of ZHI-scaphognathite with a maximum of20 setae on margin. (3) 1st maxilliped-endopod setae of ZIII, G. tridens 2, 2, 1, 2, 5+ 1 andG. rhomboides 3, 2,1,2,5+1 respectively. Acknowledgements We express our thanks to the following persons. Alan Howard, Fisheries Laboratory, MAFF,Burnham-on-Crouch, Essex, for detailed locality information that enabled us to collectmaterial. Drs David George and Howard Platt without whose support the diving programmecould not have been realized; John Tapp and Brian Maddock, ICI Marine Laboratory,Brixham, Devon, who provided facilities for keeping live crabs and helped us in many otherways. We also thank Don Claugher for providing scanning photographs that confirmed somemorphological features. LARVAL DEVELOPMENT OF ANGULAR CRAB 1 77 References Barnard, K. H. 1950. Descriptive catalogue of South African Decapod Crustacea (Crabs and Shrimps). Ann. S. Afr. Mus. 38 : 1-837.Bourdillon-Casanova, L. 1960. Le meroplancton du Golfe de Marseille: Les larves de crustaces decapodes. Reel. Trav. Stn mar. Endoume30 : 1-286.Capart, A. 1951. Crustaces Decapodes, Brachyures. Result, sclent. Exped. oceanogr. belg. Eaux cot. afr. Atlant. Sud. Ill (I) : 1 1-205.Caroli, E. 1927. Sviluppo larvale della Gonoplax angulata (Pennant). Boll. Soc. Nat. Napoli 38: 161-166.Ingle, R. W. 1979. The larval and post-larval development of the brachyuran crab Geryon tridens Kroyer (Family Geryonidae) reared in the laboratory. Bull. Br. Mus. nat. Hist. (Zool.) 36 : 2 1 7-232.1980. British Crabs, vi + 222 pp. British Museum (Natural History) & Oxford University Press, London & Oxford. & Clark, P. F. 1977. A laboratory module for rearing crab larvae. Crustaceana 32 : 220-222. Lebour, M. V. 1928. The larval stages of the Plymouth Brachyura. Proc. zool. Soc. Lond. 2 : 473-560.Manning, R. B. & Holthuis, L. B. 1981. West African Brachyuran Crabs (Crustacea: Decapoda). Smithson. Contr. Zool. 306 : i-xii, 1-379.Rice, A. L. & Ingle, R. W. 1975. The larval development of Carcinus maenas (L.) and C. mediterraneus Czerniavsky (Crustacea, Brachyura, Portunidae) reared in the laboratory. Bull. Br. Mus. nat. Hist. (Zool.) 28 : 101-1 19.& Williamson, D. I. 1977. Planktonic stages of Crustacea Malacostraca from Atlantic Seamounts. Meteor ForschErgebn. D 26 : 28-64.Steedman, H. F. (Ed.) 1976. Zooplankton fixation and preservation. In: Monographs on oceanographic methodology. 350 pp. Paris. Manuscript accepted for publication 1 1 June 1982 The larval and first crab stages of three Inachusspecies (Crustacea: Decapoda: Majidae); amorphological and statistical analysis Paul F. Clark Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Introduction Several authors have acknowledged difficulties in distinguishing between congenericbrachyuran crab- larvae (see Lebour, 1928 : 546; Hartnoll, 1961 : 181; Christiansen, 1969;Rice & Ingle, \915a & 19756; Ingle, 1982). These observations were based on a limitedamount of material that did not permit statistical analysis of larval characters. Rearing largenumbers of crab larvae (see Ingle & Clark, 1977) provided sufficient material for thisstatistical study. A multivariate technique was used to examine the larvae and first crab stageof three species of spider crabs belonging to the genus Inachus. Material & Methods Berried females of Inachus dorsettensis and /. phalangium were collected by trawl fromlocalities off Port Erin, Isle of Man, and Plymouth and /. leptochirus was trawled fromModiolus beds four miles south of Spanish Head, Isle of Man. The adult females, togetherwith the larvae used in this study are deposited in the BM(NH). Larvae were reared usingmethods described by Ingle & Clark (1977), then fixed and preserved in 80% alcohol. Twentyspecimens of each stage were dissected and mounted as permanent slide preparations inlignin pink/poly vinyl lactophenol. For the multivariate study 43, 142 & 178 characters wereused for zoeal, megalop and first crab stages respectively. The majority of these charactersare meristic, e.g. setal counts on appendages, but a few were present or absent scores.Accurate measurements of spines and carapace dimensions, as well as carapace setal countsof megalops and crab stages proved to be impracticable and were not used. Setal counts werescored for one side of the body although on occasions it was necessary to combine parts ofboth left and right appendages to form a complete score. The data was subjected to Principalco-ordinate analysis according to Gower (1966). This method summarizes similaritiesbetween OTUs as a 2-dimensional plot. The computation was carried out using a varian V72 computer. Each larval stage was analysed separately to avoid major differences betweenstages swamping any specific variation. ResultsStatistical Analysis First and second principal co-ordinates were plotted for zoea II, megalops and first crab stageof each species. The zoea I stage data was not computed because only two characters wereconsidered significant on inspection. The OTUs in zoea II (Fig. 1) can be separated into twogroups, /. leptochirus and /. dorsettensis/I. phalangium whereas in both megalops (Fig. 2)and first crab stage (Fig. 3) they are clearly separated into three groups which correspond tothe three species. Bull. Br. Mas. nat. Hist. (Zool)44(2): 179-190 Issued24 February 1983 179 180 +7 +6 +4+3 e -2 o +1 o 'I H06 r. "2 -3-4-5-6-7 - 6,19,2 P. F. CLARK ^6,19,0"346,19,0 6,19,0 .33,375^20,06,18,0 22,23,24,27,43,50,58 7190 < 2 7,16,0 36,46 7,18,0 \ 35,51,52,53,55,56,57 f -|Q Q X V 45 ' 4ai59 '. 60 i 1 7,16,0\7,19,0 --7,17,0 .47,49,54 7,18,0 .87,19,2.12,13 7,19,2 3 7,19,2 7,20,2 .\ N7;i8,2 S.10 "1.30*7,18,0 1,6,15,7,18,2 4,7,16,18^18,2 -7,18,2 i i i -5 -4 -3 -2 -1 +1 +2 +3 +4 +5 SECOND PRINCIPAL CO-ORDINATE +6 Fig. 1 A plot of the 1st & 2nd Principal co-ordinates of zoea II. Scores of variate 1 (range 5, 6, 7),variate 1 1 (range 16, 17, 18, 19, 20) and variate 36 (range 0, 2) are plotted against their OTUs.OTUs 1-20 /. leptochirus, 21-40 /. dorsettensis, 41-60 /. phalangium. The dotted line dividesthe OTUs into two groups, group 1 =1. leptochirus and group 2 1. dorsettensis/I. phalangium.Variate 36 is the only diagnostic character separating the two groups. Note that the dotted linehas no statistical significance. The zoea I stages can be divided into two groups, /. leptochirus and /. dorsettensis/I.phalangium, using the basal article of the second maxilliped and the posterio-dorsal marginof the first abdominal somite. /. leptochirus has one seta on the basis (Fig. 4b) and two on thefirst abdominal somite (Fig. 4c) whereas /. dorsettensis and /. phalangium have no setae ateither site (Fig. 4a, d). The first zoeal stages of/, dorsettensis and /. phalangium cannot beseparated on setal characters. STATISTICAL ANALYSIS OF CRAB LARVAE 181 WH H Mfe +3 +2 O k +1 -1 -2 - -3 L- 1 14 5 17,20 0-1512 *f 13 16 35 31 32" 29 51 53^ 27 28 43 59* 41 4847* 56 44 24 3736.3_3.._30 37 55 34 23*38 21*22 50 52 46* 54 I I I -.4 -3 -2 -1 +1 +2SECOND PRINCIPAL CO-ORDINATE + 3 +4 Fig. 2 A plot of the 1 st & 2nd Principal co-ordinates of megalops using 68 variates. OTUs 1-20 /.leptochirus, 21-40 /. dorsettensis, 41-60 /. phalangium. These OTUs fall into 3 distinct groupswhich correspond to the three British Inachus species. The second zoeal stages can also be split into the same two groups (Fig. 1). The threecharacters showing variation were the number of terminal aesthetascs on the antennule, thenumbers of setae on the margin of the maxillary endopod and on the posterio-dorsal marginof the first abdominal somite. The scores of these three characters are shown respectively onFig. 1 in bold type, adjacent to their respective OTUs. Separation of the two groups shown bythe dotted line is determined only by one character, the number of posterio-dorsal marginalsetae on the first abdominal somite. /. leptochirus has 2 setae (Fig. 4c) whereas /. dorsettensisand /. phalangium have none (Fig. 4d). The megalops can be divided into three groups which correspond to the three species usingonly 68 out of the original 142 characters (Fig. 2). Group separation remains constant how-ever, if only nine characters are used. By using combinations of the means of these nine characters it seems possible to separatethe megalops of the three species. Nevertheless, their overlapping distribution (Table 2) makes separation very difficult inpractice as no single character separates all three species. For example, OTU 21 (I.dorsettensis) is grouped with OTUs 52 and 53 (I. phalangium) on its overall similarity asthey have identical scores for all nine characters. Overlapping ranges of variation in numbersof setae makes it impossible to differentiate all three species with absolute confidence at the 182 P. F. CLARK WH +2 p ? +1 o PuH -1 n fe "2 -3 -4 12* 19 4 .. ir 3 133 14 53* 57 "44 58* 28.3822* 37 oc , * 26 29 21 . *23.. -* " n 31 25 -4 -3 -2 -1 + 1 +4 SECOND PRINCIPAL CO-ORDINATE Fig. 3 A plot of the 1 st & 2nd Principal co-ordinates of the first crab stages using 1 38 variates.OTUs 1-20 /. leptochirus, 21-40 /. dorsettensis, 41-60 /. phalangium. These OTUs fall into 3distinct groups which correspond to the three British Inachus species. megalop stage. Only the number of setae on the proximal exopod segment of the antennule(Figs. 4e, f) displays no intraspecific variability and serves as a diagnostic characterseparating /. leptochirus from /. dorsettensis/I. phalangium. The number of spines on themerus of the first peraeopd is also a relatively good character separating /. leptochirus fromthe other two species, but this is a particularly difficult character to observe because of thethickness of the merus. Differentiation of the first crab stage OTUs (Fig. 3) was achieved using 1 38 characters fromthe original 178, but clustering of the OTUs corresponding to the three Inachus species doesnot alter when the number of characters is reduced to 21 (Fig. 5). These 21 characters arelisted in Table 3. As with the megalops, separation of the three species is marked when the means of thevariates (Table 4) are used, but again the distributions show considerable overlap. In STATISTICAL ANALYSIS OF CRAB LARVAE 183 Fig. 4 2nd maxilliped, zoea I (a) /. dorsettensis & I. phalangium (b) /. leptochirus; abdomen ofzoea I & II (c) /. leptochirus (d) /. dorsettensis & I. phalangium; antennule of megalop (e) /.leptochirus (f) /. dorsettensis & /. phalangium. 184 P.F.CLARK Table 1 A list of characters that may separate the megalops of the three Inachm species Variate no. Characters 47 444862678195109 number of setae on proximal exopod segment of antennulenumber of setae on first segment of antennanumber of setae on merus of 3rd maxillipednumber of setae on epipodite of 3rd maxillipednumber of spines on merus of 1st peraeopodnumber of setae on propodus of 2nd peraeopodnumber of setae on propodus of 3rd peraeopodnumber of setae on propodus of 4th peraeopodnumber of setae on propodus of 5th peraeopod Table 2 Studying the means of each character from Table 1 , the megalops in theory are separableusing combinations of characters. However, if the distribution of each character is tabulated onlyvariate 4 is a good diagnostic character, but this only separates 7. leptochirus from /. dorsettensis/I.phalangium (see Figs. 4e, 0- (L = 7. leptochirus, D = I. dorsettensis & P = 7. phalangium) distribution of variate 4 1 2 L 20 D 20 P 20 variate 4 separates L from D & P distribution of variate 7 1 L 20 D 2 18 P 15 5 variate 7 separates P from D & L distribution of variate 44 4 5 L 1 19 D 18 2 P 16 4 variate 44 separates L from D & P distribution of variate 481 2 LOO 7 13 D 1 9 8 2 P 5 10 5 variate 48 separates L from D & P distribution of variate 62 2 3 L 00 D 20 P 415 variate 62 separates L from D & P 20 1 mean 21 1 mean1 1 mean544 mean3 22 mean433 distribution of variate 67 16 17 18 mean L 3 17 17 D 19 1 16 P 19 1 17variate 67 separates D from L & P distribution of variate 81 15 16 17 mean L 2 18 17 D 1 19 16 P 3 17 17Variate 8 1 separates D from L & P distribution of variate 95 15 16 17 mean L 6 14 17 D 4 16 16 P 1 10 9 16variate 95 separates L from D & P distribution of variate 109 14 15 16 mean L 1 2 17 16 D 20 15 P 1 18 1 15variate 1 09 separates L from D & P Note that the means have been rounded up to the nearest whole number. STATISTICAL ANALYSIS OF CRAB LARVAE 185 Table 3 List of characters that may be used to separate the 1 st crab stages Variate no. Characters 6 aesthetascs on 2nd exopod segment of antennule 9 setae on endopod of antennule 1 3 number of setae on 3rd segment of antenna 14 number of spines on 1st segment of antenna 2 1 number of setae on distal segment of mandibular palp 45 number of setae on basis of 2nd maxilliped 57 number of setae on coxa of 3rd maxilliped 94 number of setae on basis of 2nd peraeopod 99 number of spines on merus of 2nd peraeopod 1 00 number of spines on ischium of 2nd peraeopod 1 14 number of setae on ischium of 3rd peraeopod 1 1 5 number of setae on basis of 3rd peraeopod 1 20 number of spines on merus of 3rd peraeopod 1 2 1 number of spines on ischium of 3rd peraeopod 1 35 number of setae on ischium of 4th peraeopod 1 36 number of setae on basis of 4th peraeopod141 number of spines on merus of 4th peraeopod146 number of hooks on propodus of 4th peraeopod 1 56 number of setae on ischium of 5th peraeopod 1 57 number of setae on basis of 5th peraeopod 1 67 number of hooks on propodus of 5th peraeopod practice, only a combination of characters can be used to distinguish between the species atfirst crab stage. Morphology The general morphology of/, dorsettensis was described and illustrated by Ingle (1977).Most of the setal counts fall within the variation recorded during the present study. Thisindicates little or no temporal variation. Some of the discrepancies may be due to thedifficulty of classifying and objectively defining elements, as between a seta and a spine,when the structures grade one into the other (Gurney, 1931 : 38). However, some disparitybetween Ingle's study and the present work could not be accounted for. Discussion Williamson (1965 : 390) listed the presence of a seta on the outer margin of the maxillulebasal endite as one often characters for separating brachyuran larvae from anomuran larvae.In previous descriptions of zoea II in majids this character is shown as present, for example,Ingle (1977) records this seta as present in zoae II and megalops of Inachus reared fromPlymouth material this was confirmed by re-examining Ingle's material and by rearingfresh material from the Plymouth area. In this study the seta was absent from zoea II of/.dorsettensis reared from the Isle of Man and from all three megalops. Lebour* (1928) suggested that larvae of the three Inachus species could be separated onsize, chromatophore patterns and length of dorsal spines, but none of these claims could beverified. Only the megalops and first crab stage of/, leptochirus in the present study proved tobe larger than those of the other two species. Larval inachinids have apomorphic zoealcharacters which were listed by Rice (1980 : 307), to which can now be added the absence of *Lebour's material is no longer extant; Ingle, pers. comm. 186 P. F. CLARK STATISTICAL ANALYSIS OF CRAB LARVAE 187 -) Q cu QQ- Qcu Q cu J - o ON oo O 3 -c> o o cu"08 Q.C j ^ O ct (N ON O O^ $ 2 a --oo -oo oo O 05 a 3 & 7$ O C O tN ON O r^J "3 eS cu pO . a 2 -s. rS n ^^o-^ 2So i _ Qou Qeu Qo- Qcu 188 P. F. CLARK Q05OI OO 05 (X, HCO05 +7 +6+5+4+3+2+1 -1-2-3-4-5-6-7 L 25 33 24 31. 23 38* 29 28 36 57 .58.48 44 ' 51 53 4743 54 13, I I 5 14*9*4 A 16 ,10,11,18,20 -6 -5 -4 -3 -2 -1 +1 +2 +3 SECOND PRINCIPAL CO-ORDINATE +4 + 5 Fig. 5 A plot of the 1st and 2nd Principal co-ordinates of first crab stage using 2 1 variates. OTUs1-20 7. leptochirus, 21-40 /. dorsettensis, 41-60 /. phalangium. Although the number of variatesis reduced from 138 to 21, the OTUs still fall into 3 distinct groups which correspond to the 3British Inachus species. The grouping of the OTUs is similar to that in Fig. 3 and thereforeillustrates that the 21 characters selected contribute to the separation of the 3 groups. Note thatthe groups /. leptochirus and /. dorsettensis have changed positions when compared with Fig 3;this has no significance in the analysis. STATISTICAL ANALYSIS OF CRAB LARVAE Genus INACHUS 189 Genus MACROPODIA group 1 I. leptochirusI. thoracicus group 2 I. dorsettensisI. phalangium M. tenuirostris M. rostrata 2 setae presenton firstabdominal somite First abdominal somitewithout setae. Zoea I without aseta on basis of2nd maxilliped Distal endopodsegment of 2ndmaxilliped with3 terminal setae. Distal endopod segment of 2nd maxillipedwith h setae. 1 seta present on basis of 2nd maxilliped in zoea Iand absent in zoea II. Zoea II with mandible palp and antennule endopod absent. Fig. 6 Cladogram of known Inachus & Macropodia larval descriptions. a mandibular palp and endopod bud on the antennule of zoea II. Similarly, when comparedwith other majid larva characters the loss of paired dorsal setae on the first abdominal somiteand the absence of setae on the basis of the 2nd maxilliped in /. dorsettensis and /.phalangium, can be considered as derived traits. The present study failed to reveal characterswhich separate the larvae of/, dorsettensis from /. phalangium, but demonstrated that thelarvae of/, leptochirus can be easily recognized. Adult males of/, leptochirus share one important feature with two other species of Inachus(i.e. /. thoracicus & I. aquiarii) in having a sternal callosity, a character that is absent inmales of/, dorsettensis, I. phalangium and /. communissimus. Such a separation of Inachusspecies into two groups, those with and those without a sternal callosity, is supported by thepresent study. Heegaard (1963) studied the zoeae of/, thoracicus and clearly figures two setaeon the first abdominal somite (p. 475, Fig. 83), but not a seta on the basis of the 2ndmaxilliped (p. 475, Fig. 82). Unfortunately Heegaard's material is no longer extant.Re-examination of /. thoracicus zoeae may well show that they are inseparable from /.leptochirus, adding support to the suggestion that there are two natural groups in the genusInachus. Present larval evidence supports the view that Inachus and Macropodia are the mostderived of all majids since they show the greatest reduction in numbers of setae; consideredby Rice (1980) to be the derived condition. A suggested phylogeny of well documentedlarvae from the genera Inachus and Macropodia is shown in Fig. 6. Setal studies of other brachyuran genera have shown that the larvae of species accepted as 190 P. F. CLARK closely related are not usually separable on quantitative characters. Therefore meristic setalincongruities within genera, as shown here for Inachus, may be the only morphologicalevidence of phylogenetic non-homogeniety. This project formed part of an M.Sc. degree in Modern Taxonomy. All relevant data andlarval figures were deposited in the Crustacea Section, BM(NH) and the Library of thePolytechnic of Central London. Acknowledgements I wish to thank Drs Roger Lincoln, Ray Ingle and Geoff Boxshall, Crustacea SectionBM(NH), for their help and critical comments on earlier drafts of this paper. For help withcollection of ovigerous crabs, I thank Mr Tony Mattacola, Plymouth Marine Laboratory, thecrews of the research vessels, Sarsia, Sepia & Squilla (Plymouth Marine Laboratory) andCuma & Silver Spray (Port Erin, Isle of Man) and Drs D. I. Williamson & Richard Hartnoll(Port Erin Marine Station, University of Liverpool). Miss Joan Ellis, Crustacea Section,kindly sorted through large quantities of Modiolus, and Dr M. Hills & Miss Kay Shaw,Biometrics Section, gave invaluable statistical advice. References Christiansen, M. E. 1969. Marine invertebrates of Scandinavia. No. 2. Crustacea, Decapoda, Brachyura. Universitetsforlaget, Oslo pp. 1-143. Oslo.Clark, P. F. 1980. British Spider Crabs of the genus Inachus; a morphological study of larval development. M.Sc. Modern Taxonomy (C.N.A.A.) Thesis. Polytechnic of Central London/City of London Polytechnic.Gower, J. C. 1966. Some distance properties of latent root and vector methods used in multivariate analysis. Biometrika 53 (3-4) : 325-358. Gurney, R. 1931. British Fresh- Water Copepoda. Vol. 1, 239 pp, Ray Society, London.Hartnoll, R. G. 1961. A re-examination of the Spider Crab, Eurynome Leach from British waters. Crustaceana2(3) : 171-182.Heegaard, P. 1963. Decapod larvae from the Gulf of Napoli, hatched in captivity. Vidensk. Meddr dansk. nalurh. Foren. 125 : 449-493.Ingle, R. W. 1977. The larval and post-larval development of the Scorpion Spider Crab, Inachus dorsettensis (Pennant) (Family: Majidae) reared in the laboratory. Bull. Br. Mus. nat. Hist. (Zool.) 30 (9) : 329-348.1982. Larval and post larval development of the slender- legged Spider Crab, Macropodia rostrata (Linnaeus) (Oxyrhyncha: Majidae: Inachinae), reared in the laboratory. Bull. Br. Mus. nat. Hist. (Zool.) 42 (3) : 207-225.Ingle, R. W. & Clark, P. F. 1977. A laboratory Module for rearing crab larvae. Crustaceana 32 (2): 220-222. Lebour, M. V. 1928. The larval stages of the Plymouth Brachyura. Proc. zool. Soc. Lond. 2 : 473-560.Rice, A. L. 1980. Crab zoeal morphology and its bearing on the classification of the Brachyura. Trans. zool. Soc. Lond. 35 : 271-424.Rice, A. L. & Ingle, R. W. 1975a. The larval development of Carcinus maenas (L.) and C. mediterraneus Czerniavsky (Crustacea, Brachyura, Portunidae) reared in the laboratory. Bull. Br. Mus. nat. Hist. (Zool.) 28 (3) : 101-1 19.19756. A comparative study of larval morphology of the British Portunid crabs Macropipus puber(L.) and M. holstatus (Fabricius), with a discussion of generic and sub-familial larval characters within the Portunidae. Bull. Br. Mus. nat. Hist. (Zool.) 28 (4) : 1 2 1-1 5 1 .Williamson, D. I. 1965. Some larval stages o"f three Australian crabs belonging to the families Homolidae and Raninidae, and observations on the affinities of these families (Crustacea: Decapoda). Aust. J. mar. Freshw. Res. 16 : 369-398. Manuscript accepted for publication 18 June 1982. British Museum (Natural History)British Marine Amphipoda: Gammaridea by R. J. Lincoln 658pp 2,300 figures 4to hard boundISBN 565 00818 50.00 Amphipods are both numerous and diverse in numbers of genera and species inBritish coastal waters, but in the absence of any form of modern systematic synopsisor key this group of crustaceans has acquired the reputation of being notoriouslydifficult to identify. This monograph, which is the first comprehensive and illustratedtext on British gammaridean amphipods to be published in more than a century,should go a long way towards solving the problem. The systematic section of the book contains descriptions and figures of all 271 speciesof marine and brackish water amphipods, in 123 genera and 36 families, recordedfrom British coasts and the adjacent continental shelf to a depth of 200 metres. Keysare provided at all levels, as well as relevant synonymies and diagnoses of genera andfamilies. The text is illustrated with about 2,300 separate figures which have beendrawn by the author from Museum and other material, in many cases with referenceto type specimens. The work has been carefully edited to bring corresponding descriptions, keys andfigures into close proximity within the text. The systematic section is supported bychapters dealing with morphology, systematics, geographical distribution, biologyand ecology, the latter being presented in the form of an annotated subject index ofresearch literature. Finally, there is an extensive bibliography of about 1,200references that includes most of the British marine amphipod literature published todate. Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. Titles to be published in Volume 44 Observations on the systematics of the genus Difflugia inBritain (Rhizopoda, Protozoa).By Colin C. Ogden Miscellanea A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin Curds & Irene C. H. Wu Osteology, genitalia and relationships of the Acanthodactylus(Reptilia : Lacertidae). By E. N. Arnold The Opthalmotilapia assemblage of cichlid fishes reconsidered. By Peter Humphrey Greenwood Morphological studies on some Difflugiidae from Yugoslavia(Rhizopoda, Protozoa). By Colin G. Ogden & Andjelija Zivkovic Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) A review of the Euplotidae(Hypotrichida, Ciliophora) Zoology series Vol 44 No 3 31 March 1983 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, andan 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 SW7 5BD,England. World List abbreviation: Bull Br. Mus. not. Hist. (Zool.) Trustees of the British Museum (Natural History), 1982 The Zoology Series is edited in the Museum's Department of ZoologyKeeper of Zoology : Dr J. G. ShealsEditor of Bulletin : Dr C. R. CurdsAssistant Editor : Mr C. G. Ogden ISSN 0007-1498 Zoology series Vol 44 No 3 pp 191-247British Museum (Natural History)Cromwell RoadLondon SW7 5BD Issued 31 March 1983 A review of the Euplotidae (Hypotrichida,Ciliophora) Colin R. Curds & Irene C. H. Wu Zoology Department, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Contents IntroductionKey to Genera .Genus DiscocephalusGenus DiophrysGenus Uronychia .Genus Certesia .Genus GastrocirrhusGenus Euplotaspis .Genus Euplotidium .Genus Paraeuplotes .Genus Swedmarkia .Genus GruberellaGenus Cyatharoides.References .Index . 191192193197 215227228232234239240242243244247 Introduction Guides to the species of two genera of Euplotid ciliates, Euplotes Ehrenberg in Hemprichand Ehrenberg, 1831 and Aspidisca Ehrenberg, 1830, have been published in recent years(Curds, 1975; Wu & Curds, 1979). The other genera in the family Euplotidae Ehrenberg,1838 contain fewer species but there are still identification problems in some. The presentwork is intended to aid the specific identification of the remaining nine genera which Borror(1972) grouped together into the Euplotidae, another more recently described genus and oneother addition. Although four of these genera only contain single species it was thoughtworthwhile to include them for completeness. The genera included here are in chronologicalorder of authority. Discocephalus Ehrenberg in Hemprich and Ehrenberg, 1828; DiophrysDujardin, 1841; Uronychia Stein, 1859; Certesia Fabre-Domergue, 1885; GastrocirrhusLepsi, 1928; Euplotaspis Chatton and Seguela, 1936; Euplotidium Noland, 1937;Paraeuplotes Wichterman, 1942; Swedmarkia Dragesco, 1954; Gruberella Corliss, 1960 andCyatharoides Tuffrau, 1975. Members of the Euplotidae are hypotrichs with the characteristically prominent adoralzone of membranelles (AZM) arranged at the anterior left of the ventral surface bordering awide peristome. In certain genera the AZM may continue over the apical end of the bodyonto the dorsal surface. Unfortunately there is no set of characters which will include all ofBorror's (1972) genera into the family. This could well indicate that it is not a natural family.The cirri on the ventral surface are arranged in distinct groups but they are not the onlyhypotrichs to have this feature. For example, the Oxytrichidae Ehrenberg, 1838 also havegrouped cirri, and Borror (1972) distinguished them from the Euplotidae by their possessionof only a few left marginal and no right marginal cirri, even so there are several exceptions to Bull. Br. Mus. nat. Hist. (Zool.)44(3): 191-247 Issued 31 March 1983 191 192 C. R. CURDS & I.C. H. WU this rule. Furthermore, while right caudal cirri are present in Diophrys, Discocephalus,Euplotes and Uronychia, they are absent in others. However, transverse cirri are present inall but one Euplotid genus. With this state of affairs, perhaps it is not surprising that there issome disagreement in the literature as to what constitutes the family Euplotidae. Borror(1972) included Aspidisca in the family but Stein (1859a), Biitschli (1889), Kahl (1932),Faure-Fremiet (1961) and Corliss (1961, 1977, 1979) all maintain it in a separate family theAspidiscidae Ehrenberg, 1838. Corliss (1979) included four genera in the familyAspidiscidae; Aspidisca, Euplotaspis, Onychaspis Stein, 1859 and Paraeuplotes (as anincertae sedis). The genus Onychaspis differs from Aspidisca only by the former's possessionof a larger number of transverse cirri and most authors consider it to be a synonym ofAspidisca (Borror, 1972; Wu and Curds, 1979). Indeed, Corliss (1979) indicated thispossibility in his classification. One of the features of the Aspidiscidae is the great reductionof oral membranelles, however this description would not fit the AZM of Euplotaspis orParaeuplotes. Corliss (1979) also placed three of the genera being considered here(Cirrhogaster Ozaki and Yagui, 1941; Euplotidium and Gastrocirrhus) into the familyGastrocirrhidae Faure-Fremiet, 1961 . Here the single species genus Cirrhogaster is regardedas a member of the genus Gastrocirrhus as it only differs slightly in cirral number. Borror(1972) included Gastrocirrhus in the Euplotidae but noted that its position in that family wasprovisional. It is clear from this brief synopsis that the position of these genera within three or a singlefamily group is still in a state of flux and to a large extent a matter of conjecture. It is notthe purpose of the present paper to attempt to assign the genera to any old, current or newclassification scheme; more data, particularly concerning their comparative morphogenesis,are required before anything useful can be suggested. The aim here is to aid the marineecologist and protozoologist to identify the species of those genera which conveniently fallwithin the single family Euplotidae. Key to Genera 1 AZM conspicuous, in single anterior part 2 AZM inconspicuous, in 2 parts ASPIDISCA (see Wu& Curds, 1979) 2 Without frontoventral cirri, caudals never extremely large . . . GRUBERELLA(p.242)With frontoventral cirri (when frontoventrals reduced and inconspicuous then caudals usually very large) 3 With caudal cirri 4 Without caudal cirri 11 4 With marginal cirri, caudals sometimes very large 5 Without marginals, caudals present but never very large 5 Caudals very large and prominent 6 Caudals weak EUPLOTES (see Curds, 1975) 6 5- 10 conspicuous frontoventral cirri present, moves forwards . . DIOPHRYS (p. 197)3 inconspicuous cirri present, moves backwards URONYCHIA(p.2l5) 1 Elongate with discoid 'head' region DISCOCEPHALUS (p. 193) Ovoid to elongate, when elongate never with discoid 'head' region 8 8 Ovoid, dorsoventrally flattened 9 Elongate, rounded in section, often cup-shaped with funnel-shaped peristome opening apically and ventrally 10 9 Cirri mainly in 2 rows, found on coral PARAEUPLOTES (p. 239) Cirri in well-defined groups, found in sea-squirts EUPLOTASPIS (p. 232) 10 With 5-6 transverse cirri EUPLOTIDIUM (p. 234) Without transverse cirri GASTROCIRRHUS (p. 228) 11 Without marginal cirri, rounded in section EUPLOTIDIUM (p. 234) With marginal cirri, dorsoventrally flattened 12 1 2 Marginal cirri on both right and left sides, confluent posteriorly . SWEDMARKIA (p. 240)Marginal cirri on either right or left sides 13 13 Marginal cirri on left, not planktonic CERTESIA(p.221) Marginal cirri on right, planktonic CYATHAROIDES(p.243) REVIEW OF EUPLOTIDAE 193 Genus DISCOCEPHALUS Ehrenberg, 1 828 Introduction The genus Discocephalus was erected by Ehrenberg in Hemprich and Ehrenberg (1828) andalthough his description and illustration of the type species D. rotatorius were crude, it wasclear that the organism had a distinctive discoid 'head' formed by a constriction at theanterior end of the elongate oval body. Butschli (1889) expressed some doubt concerning theobservations of Ehrenberg (1831) but it was not until Kahl (1932), that Discocephalus wasadequately described. The latter description concerned an organism which Kahl (1932)considered to be D. rotatorius Ehrenberg and of which he thought Polycoccon octangularisSauerbrey, 1928 to be a junior synonym. Later, however, Dragesco (1960) isolated anorganism from Roscoff which although identical to that described by Kahl (1932) wasdifferent in several respects from D. rotatorius Ehrenberg. Dragesco (1960) therefore namedthe species found at RoscoffZ). ehrenbergi and designated D. rotatorius Kahl to be its juniorsynonym. However, Dragesco (1960) agreed that Polycoccon octangularis Sauerbrey was asynonym of D. rotatorius Ehrenberg. In addition to the two species mentioned above twoothers, D. grandis Dragesco, 1954 and D. minimus Dragesco, 1968, have been described. There is a well developed AZM in all four of the above species which borders the smallventral peristome on the left of the discoid 'head'. In one species, D. ehrenbergi, there are 5-7large membranelle-like structures lying along the right border of the peristome whichDragesco (1968) called 'pre-membranelles'. This character enables the diagnosis andseparation of D. ehrenbergi Dragesco from D. rotatorius Ehrenberg which lacks these'membranelles'. The cirral patterns on the ventral surface of Discocephalus differ greatly from those ofEuplotes and as their morphogenesis is yet to be published fully it is difficult to interpretthem adequately. Furthermore as the cirri vary considerably from species to species, it ispossible that Discocephalus is really a polyphyletic group. The present confusion in differentterms used for the same cirri by different authors can be illustrated by reference to 'marginal'cirri. In D. rotatorius and D. ehrenbergi there are two 'marginal' cirri on the left body edgejust below the peristome. However, in D. grandis, Dragesco (1954) refers to the presence ofthree rows of 'marginal' cirri which is more characteristic of members of the Oxytrichidaerather than the Euplotidae. Furthermore, in D. minimus there is one row of cirri on the rightbody edge called ventral cirri and one row on the left edge called 'marginal' cirri. Untilfurther morphogenetic information becomes available the present authors prefer to follow inpart the system used by Hartwig and Parker (1977) which distinguishes left, central and rightventral cirri rather than to attempt to distinguish marginal from ventral rows simply on thebasis of their position on the ventral surface. However, it is preferable to call the mostposterior cirri, left and right caudal cirri, rather than left and right posterior 'marginals' asused in Hartwig and Parker (1977). The dorsal surface of Discocephalus has received lessattention than that of either Euplotes (Curds, 1975) or Aspidisca (Wu & Curds, 1979) butboth Dragesco (1965) and Kahl (1932) have illustrated D. ehrenbergi with six kinetics on thedorsal surface. Dorsal argyrome patterns as found in Euplotes and Aspidisca (Curds, 1975:Wu & Curds, 1979) have yet to be described. This means that the four species must currentlybe distinguished on the basis of their size, numbers and distribution of cirri, the presence orabsence of membrane-like structures on the right of the peristome and the nature of themacronuclear apparatus. Corliss (1979) described this as a curious genus which he includedonly 'tentatively' in the Euplotidae and indeed, the unique body shape, the cirral arrange-ment and large number of macronuclear parts may well be considered sufficient for placing itin a separate family. Diagnosis of Discocephalus Oval marine hypotrichs with an anterior constriction of the body which forms a discoid 194 C. R. CURDS & I. C. H. WU 'head' containing the peristome and AZM on the left. Ventrally there are 4-8 frontal, 5-1 1transverse (including satellites) and variable numbers and arrangements of ventral cirri.Posterio-dorsally there are 3-4 right caudal and 9-30 left caudal cirri. The size variesbetween 50-200 um long and there are numerous macronuclei. Key to the species of Discocephalus \ With 2 left ventral cirri With 7 to many left ventral cirri 2 With 5 transverse cirri but without 'pre-membranelles'.With 8-9 transverse cirri and several 'pre-membranelles' 3 Large (200 urn), row of many (about 40) left ventral cirriSmall (50 urn), row of few (7-8) left ventral cirri . 23 D. rotatorius D. ehrenbergi D. grandis D. minimus Species descriptions Discocephalus rotatorius Ehrenberg, 1828 in Hemprich and EhrenbergPolycoccon octangulus Sauerbrey, 1928 DESCRIPTION (Fig. 1). Medium sized (70-100 um long) marine species whose elongate, ovalbody is typically discocephalic. There is a well developed AZM bordering the left of theperistome area. The 7-8 frontal cirri are restricted to the anterior 'head' region. The ventralcirri are in two groups: there are 2-4 on the right body edge and 2 on the left immediatelybehind the peristome. There are 5 prominent transverse cirri. The caudal cirri are also in twogroups and arise dorso-laterally: on the right are 3-4 large prominent caudals and on the left 20yu m Fig. 1 Discocephalus rotatorius: (a, b) after Sauerbrey, 1928 (called Polycoccon octangulus); (c)unpublished figure after Faure-Fremiet (in Dragesco, 1960). REVIEW OF EUPLOTIDAE 195 is a row of many (12-20) smaller caudal cirri. The macronucleus is divided into manyrandomly distributed spherical parts. NOTES. The description is based on that of Sauerbrey (1928) and on the illustration given inDragesco (1960) but attributed to unpublished diagrams of Faure-Fremiet. Discocephalus grandis Dragesco, 1954 DESCRIPTION (Fig. 2). Large (200 um long) marine species with an uncharacteristically largenumber (120-150) of ventral cirri arranged in three longitudinal rows: one lies on the leftbody edge and two on the right. There are 4 frontal cirri which are all located alongthe right margin of the discoid 'head'. Immediately above the first of the long transverse cirri,on the left, are two small 'satellite' transversal cirri. There are two small right caudal andabout 13 left caudal cirri. The macronucleus is divided into many (10-20) spherical partseach being 3-6 um in diameter. NOTE. This description is based on those of Dragesco (1954, 1960) who noted that thespecies is very similar to Holosticha discocephalus Kahl. Discocephalus ehrenbergi Dragesco, 1960Discocephalus rotatorius Kahl, 1932 misidentification Fig. 2 Discocep ''alus grandis, after Dragesco, 1 960. 196 C. R. CURDS & I. C. H. WU DESCRIPTION (Fig. 3). This is a medium sized (90-120 um long) marine species. The presenceof 7-9 transverse cirri (with 1 or 2 'satellites', Fig. 3f-h) and 5-7 'pre-membranelles' on theright of the peristome serve to distinguish it from the type species D. rotatorius Ehrenberg.The AZM consists of 18-22 membranelles arranged along the left of the peristome and thereis an undulating membrane on the right. The 4-8 frontal cirri are of variable size and thereare 2 left ventral, 2 mid-ventral and 4 right ventral cirri. The caudal cirri are in two groups:3-4 right caudal and 14-20 caudal cirri on the left. There are 6 dorsal kineties with only thethird from the right extending into the 'head' region (Dragesco, 1965). There is a large 30K Fig. 3 Discocephalus ehrenbergi: (a, b) ventral and dorsal surfaces, after Dragesco, 1965; (c, d)ventral and, dorsal surfaces, after Kahl, 1932 (called D. rotatorius); (e, ventral surface andvariations in transverse cirral arrangement, after Dragesco, 1960. REVIEW OF EUPLOTIDAE 197 number of spherical parts to the macronuclear apparatus which are scattered throughout thebody but apparently not in the 'head' region. NOTES. This species was first described by Kahl (1932) as D. rotatorius Ehrenberg but wasrenamed and redescribed by Dragesco (1960, 1965). Later Dragesco (1968) expressed doubtsabout his conclusions and suggested that the species described as D. ehrenbergi in 1960 (Fig.3e-h) might be a different species from that described in 1965 (Fig. 3a-b), although there wasinsufficient information to be certain. The description above is based on that of Dragesco(1965) but variations noted in other descriptions are included. Discocephalus minimus Dragesco, 1968 DESCRIPTION. Discocephalus minimus (Fig. 4) is characterised by its small size (50-60 urnlong) and possession of left ventral cirri. The species has 7 frontal, 4-6 right ventral, 2 mid-ventral and 7 left ventral cirri. There are 4 right caudal and 9-10 left caudal cirri. Themacronucleus is divided into relatively few (3-1 5) spherical parts. lOyu m Fig. 4 Discocephalus minimus, after Dragesco, 1 968. NOTES. The single description of this species is based on the observation of 5 individuals(Dragesco, 1968) and there are no data on the dorsal aspect of this organism. Genus DIOPHRYS Dujardin, 1841Introduction It was Dujardin (1841) who first briefly described an organism called Diophrys marina anddefined the genus Diophrys as having a discoid shape with groups of long 'cilia' at the twoextremities of the body which was without a mouth. Later Stein (18590) stated that both D.marina and the earlier Stylonychia appendiculata Ehrenberg, 1838 were synonyms of 198 C. R. CURDS & I. C. H. WU Styloplotes appendiculatus Stein, 1859 (see Stein 1859a) which he characterised by thepresence of three large right caudal cirri. Biitschli (1889) recognised Diophrys as a distinctgenus and his list of synonyms included certain species of the genera, Stylonychia Ehrenberg,Ploesconia Dujardin, Euplotes Claparede & Lachmann, Schizopus Claparede & Lachmann,Styloplotes Stein, Styloplotes Quennestedt, Styloplotes Fresenius, Styloplotes Rees,Styloplotes Fabre-Domergue, and Styloplotes Andrusova. Although Biitschli (1889) statedthat there were two species in the genus, he only illustrated Diophrys (Styloplotes) grandisRees, 1 88 1 and failed to mention D. appendiculata. Over the next few years the combinationD. appendiculata was used occasionally, for example, by Wallengren (1901) and by Calkins(1902) who established the combination Diophrys (Styloplotes) appendiculatus Stein.However, it was Kahl (1932) who first traced the taxonomic history of the species correctly.Borror (1972) recently listed eleven species of the genus with their synonyms and hedesignated Diophrys scutum (Dujardin, 1841) Kahl, 1932 the type species and listed D.marina as a synonym of D. appendiculata (Ehrenberg) Kahl. In view of the evidence givenabove, the present authors are of the opinion that Diophrys (Stylonychia) appendiculata(Ehrenberg, 1838) Kahl, 1932 is the correct name and authority for the type species of thegenus Diophrys. Kahl (1932) defined the genus Diophrys as those members of the family Euplotidae withconspicuous sturdy transverse cirri and a single group of three large caudal cirri. In additionto D. appendiculata, Kahl (1932) described, and gave keys, to three other species, Diophrys(Ploesconia) scutum (Dujardin, 1841) Kahl, 1932; Diophrys hystrix Buddenbrock 1920 andDiophrys irmgard Mansfeld, 1923. Of the eleven species listed by Borror (1972) the presentauthors accept ten nominal species. D. tetramacronucleata Kattar, 1970 and D.multinucleata Hartwig, 1973 are two later additions to the genus, characterised by theirpossession of four and over twenty macron uclei respectively. The authors accept the latter asa distinct species but suspect that the former is a synonym of D. appendiculata. Two of the species, D. appendiculata and D. scutum, have been described by many authorsover the past century but most of the other species are relatively recent additions. In mostcases there are rather few data on the extent of intraspecific variation that might be found inpotentially useful diagnostic structures. Even so, it was thought worthwhile to present asummary of the data that are available and to discuss the possible diagnostic importance ofthe various morphological features. (a) Shape. The typical body shape of Diophrys is an ovoid in which there is often a posteriorright lateral indentation where the right caudal cirri are located. The anterior of D. hystrix isparticularly truncate and the body of D. kahli is elongate. The dorsal surface of Diophrys isgenerally smooth, but in D. irmgard it is described with a 'trapeziform elevation'(Mansfeld, 1923). These variations are apparently distinctive although the keys here do notrely upon them for identification of these species. (b) Size. Most Diophrys species are between 50 and 120 um long. D. scutum at 1 50-200 (j.mlong is the largest and D. hystrix at 30-40 um long is the smallest. Size variation withinspecies appears to be small (Hartwig, 1973) so the size differences between D. hystrix, D.scutum and D. appendiculata (50-100 um long) are likely to be of diagnostic value. (c) Adoral zone of membranelles (AZM). Diophrys has a wide peristome which extends athird to two-thirds down the body length. On the right of the peristome there is a large, wideundulating membrane. On the left, there is a well-developed AZM which continuesanteriorly over onto the dorsal surface: in some species it can return to the ventral surfacedown the right side of the body. Borror (1965a) stated that statistical analysis showed that thelength of the right portion of the AZM in D. scutum was significantly longer from that of D.peloetes. He noted that this 'terminal portion' of the AZM in D. scutum was 0-4 the bodylength but only 0-3 the body length in D, peloetes. The present authors would notrecommend species recognition on such a small difference as this without resort to statisticalmorphometric analysis but it is a useful feature for distinguishing between certain species. REVIEW OF EUPLOTIDAE 199 For example, in D. appendiculata the AZM hardly extends onto the right of the body whilein D. scutum it extends almost to the central region. Rees (1883) and Kahl (1932) alsoconsider this right extension of the AZM to be an important diagnostic feature by which theyidentified Styloplotes quennerstedti and D. scutum respectively. (d) Cirri. The type species, D. appendiculata, has 7-8 fronto ventral, 5 transverse, 1-3 leftmarginal and 3 right caudal cirri. The frontoventral cirri are arranged in two distinct groups,with 5 in the anterior right frontal group and 2-3 in the ventral group, often much smallerthan the frontal cirri, lying in close proximity to the first transverse cirrus on the right. Thisdistribution pattern is seen in all species but D. hystrix, D. irmgard and D. kahli have 9-10frontoventrals in groups of 7 frontals and 2-3 ventrals which the authors consider to be ofdiagnostic importance. Some authors are in agreement with this others are not. For example,Agamaliev (1967) established a new species D. scutoides which differs from D. scutum onlyin having five instead of seven frontoventral cirri. However, Borror (1963) identified anorganism as D. irmgard even though it possessed only five frontoventrals rather than the ninein Mansfeld's (1923) original description. Borror (1963) neither established the organism as anew species nor did he suggest the cirral difference to be due to intraspecific variation. Themorphogenesis of cirri in D. appendiculata was described by Wallengren (1901). Thefrontoventral-transverse cirri arise from six streaks of kinetosomes as in Euplotes, with theI/I , 11/2, II/3, HI/2 and IV/2 cirri forming the frontals: V/2 and VI/2 the ventrals and II-VI/1the transversals (Fig. 5). The transverse cirri of Diophrys are noticeably larger than those of Euplotes. There arealmost invariably five transverse cirri but there are only four in D. irmgard and D.multinucleata. The presence of large, sickle-shaped, dorso-laterally attached right caudalcirri is a characteristic feature of the genus Diophrys. During morphogenesis, these cirri arisefrom basal bodies at the posterior ends of the ciliary rows on the right of the dorsal surface(Borror, 1972). Most of the species described have three of these cirri but D. quadricaudatusAgamaliev, 1967 has four and D. kahli Dragesco, 1963 has only one. Intraspecific variationin the number of these cirri has not yet been reported and so it is thought that this might be auseful diagnostic character. In most species, the left marginal cirri are located just posterior Fig. 5 Arrangement and numbering of cirral streaks in Diophrys appendiculata, after Wallengren, 1901. 200 C. R. CURDS & I.C. H. WU to the peristome, but in D. irmgard, D. kahli and D. multimicronucleata, they are in the'caudal position' as are those in Euplotes and Uronychia. Hartwig (1974) believed that thepositioning of these cirri in these three species was sufficient to constitute a separate genus.While the present authors agree that the caudal positioning of the left marginal cirri appearsto be a significant character apparently associated with the presence of only four transversecirri, they hesitate to erect a new genus on this alone. Mansfeld (1923) and Kahl (1932)distinguished D. irmgard from other species, not by the position of the left marginal cirri butby their number. Although most species have two left marginal cirri and D. irmgard hasthree, variation between 1-3 has been observed in D. appendiculata. (e) Nuclear features. The most commonly found nuclear arrangement in Diophrys is twomacronuclei and 2-6 micronuclei. Exceptions to this include D. quadricaudatus, D.tetramacronudeata and D. multinucleata which have three, four and over 20 macronucleirespectively, with the diagnosis of the latter two species resting heavily on this feature. Themacronuclei may be rod-shaped, ovoid or moniliform. Fig. 6 Nuclear arrangement in Diophrys: (a) Diophrys magnus, after Raikov & Kovaleva, 1968;(b-d) Diophrys scutoides, after Agamaliev, 1967; (c) Diophrys quadricaudatus, after Agamaliev,1967; (e, Diophrys hystrix, after Buddenbrock, 1920. Raikov and Kovaleva (1968) separated D. magnus from D. scutum principally on the basisof its macronuclei being moniliform, however they are similar to those of D. scutum andother species which are often shown to be slightly nodular (Fig. 6a-c). Here D. magnus andD. scutum are therefore regarded to be synonymous. Summers (1935) described the reorganisation and division of the macronuclei of D.scutum, misidentified as D. appendiculata. He noted (Fig. 7) that in the normal resting stage,the species possesses two macronuclei without a visible strand between them and a variablenumber of micronuclei. He observed that fragmentation of the reorganised parts of themacronuclei was not uncommon but 'fragments of the macronuclei have never been foundfree in the cytoplasm after the several parts fuse to form the rod-like mass'. One cannot becertain if the tripartite macronuclear structure of D. quadricaudatus (Fig. 6d) is the result offragmentation or if it is the true resting stage, but it is likely that the sausage-shaped nucleuswhich Buddenbrock (1920) described in certain specimens of D. hystrix to be a divisionalstage in the normally ovoid macronuclei (Fig. 6e). REVIEW OF EUPLOTIDAE 201 d Fig. 7 Division of the macronucleus of Diophrys scutum, after Summers, 1935 (called D.appendiculatd): (a) nucleus at rest; (b) first appearance of a reorganisation band at the outer poleof the posterior macronucleus; (c) reorganisation bands about to disappear at inner poles of themacronuclei; (d) fusion of the macronuclei; (e) after fusion; (0 macronucleus beginning to divide;(g) macronuclei completely divided just before daughter cells separate. Kisselbach (1936) also illustrated various stages in the nuclear development of D.appendiculata (Fig. 8a-e) one of which shows a quadripartite stage similar to that illustratedby Kattar (1970) in D. tetramacronucleata (Fig. 8f, g). In view of this D. tetramacronucleatashould be strongly suspected as being a synonym of the type species. However, since oneillustration by Kattar (1970) and another by Hartwig (1974) show the four ovoidmacronuclei to be completely distinct (Fig. 8g) the species has been provisionally includedhere awaiting further data. As mentioned above, there are commonly 2-6 micronuclei inDiophrys. Borror (19650) accorded little significance to this feature and referred to Ito (1963)who stated that macronuclei may vary considerably in number within a species. (/) Dorsal silver-line system. Borror (19650) pointed out that few workers had mentioned thedorsal ciliature in Diophrys species descriptions and observed that Kahl (1932) was the firstto note the presence of five rows of short cilia in D. scutum. Using the Chatton-Lwoff(1930)technique, Borror (19650) was able to show that '. . . the kinetosomes in the dorsal rows ofDiophrys behave during cell division in a manner similar to the behavior of the dorsalkinetosomes in Euplotes, hence the proter and opisthe usually have the same number ofrows. Within a population, with practically no exceptions, all members of the genus presenthave the same number of rows of cilia dorsally, and this is apparently not related with bodysize.' 202 C. R. CURDS & I.C. H. WU Several authors have used the number of dorsal kinetics as a diagnostic character (Borror,19650, b: Agamaliev, 1967: Raikov & Kovaleva, 1968). However, the dorsal silver-linesystems of other species including D. hystrix, D. irmgard, D. kahli, D. tetramacronudeataand D. multinucleata still remain to be described. For this reason the key here onlydistinguishes between D. oligothrix and D. peloetes on this character. All dorsal argyromesthat have been described so far consist of a meshwork pattern and are therefore of little valuefor specific identification. Fig. 8 20yur Nuclear arrangement in Diophrys: (a-e) Diophrys appendiculata, after Kisselbach, 1936;(f, g) Diophrys tetramacronudeata, after Kattar, 1970. Diagnosis of Diophrys Marine hypotrichs 30-200 um in length. Ovoid body usually with prominent rightposterio-lateral concavity from which arise three large sickle-shaped right caudal cirri. Thereare 5-10 fronto ventral, 4-6 transverse and usually 2-3 left marginal cirri. There are oftentwo elongate, sometimes nodular, macronuclei but in some species there may be four or overtwenty macronuclei. There is a variable number of micronuclei. The dorsal silver-linesystem consists of 4-8 dorsolateral kinetics and a mesh-like argyrome. Key to the species of Diophrys 1 With 5 frontoventral cirri With more than 5 frontoventral cirri 2 With 7-8 frontoventral cirri With 9- 10 frontoventral cirri 3 With 4 right caudal cirri With 3 right caudal cirri 4 With 1-3 left marginal cirri Without left marginal cirri 5 With 2 macronuclei With 4 or more macronuclei 6 With 5-6 dorsolateral kinetics With less than 5 or greater than 6 dorsolateral kinetics . 7 AZM extends almost to centre of right border, 1 50-200 um long.AZM hardly extends at all down right border, 50- 1 00 um long . 8 With 4 dorsolateral kinetics With 8 dorsolateral kinetics I), scutoides . . 2 . . 3 . . 10 D. quadricirratus . . 4 . . 5 D. salina . . 6 9 1 D. scutum D. appendiculata D. oligothrix D. peloetes REVIEW OF EUPLOTIDAE 203 9 With 4 macronuclei D. tetramacronucleata With 20 or more macronuclei D. multinucleata 10 With 2 left marginal cirri posterior to the peristome, 30-40 urn long .... D. hystrixLeft marginal cirri in the 'caudal' position, 80-100 um long. ....... 11 1 1 With 1 right caudal and 2 left marginal/caudal cirri D. kahli With 3 right caudal and 3 left marginal/caudal cirri D. irmgard Species descriptions Diophrys appendiculata (Ehrenberg, 1838)Kahl, 1932 Stylonychia appendiculata Ehrenberg, 1838 Diophrys marina Dujardin, 1 84 1 Schizopus norwegicus Claparede & Lachmann, 1858 Styloplotes appendiculatus Stein, 1 859 Styloplotes fresenii Rees, 1883 Styloplotes appendiculatus var. pontica Andrusova, 1 886 Planiplotes wagneri Andrusova, 1 886 Diophrys appendiculatus (Stein, 18 59) Calkins, 1902 DESCRIPTION (Figs 9, 10). This, the type species of the genus, is 50-100 (im long. The bodyshape is typically ovoid with the characteristic lateral concavity at the posterior where thethree sickle-shaped right caudal cirri arise. There are 7-8 frontoventral cirri - 5 anterior and2-3 close to the transversals, 5 transverse and 1-3 left marginal cirri. The AZM extendsdown half to two-thirds of the length of the body on the left but hardly at all on the right. Thetwo macronuclei are usually elongate to ovoid and may be smooth or nodular. There are 2-4micronuclei and 5-6 dorsolateral kinetics each carrying 6-10 cilia. NOTES. Some of the nomenclatural history of this species has already been outlined in theintroduction to the genus. Kahl (1932) transferred Stylonychia appendiculata Ehrenberg,1838 to the genus Diophrys and redescribed the species. However, he made no reference tothe fact that Stein (1859#) had already erroneously redefined the species as Styloplotesappendiculatus which combination was subsequently used by many workers (Fresenius,1865: Quennerstedt, 1867: Kent, 1881: Rees, 1883: Fabre-Domergue, 1885). Stein (18590)suggested that Ehrenberg (1838) had overlooked the frontoventral cirri and suggested thatSchizopus norwegicus Claparede & Lachmann, 1858 was probably a synonym, even thoughClaparede & Lachmann (1858) had distinguished their species from Stylonychiaappendiculata Ehrenberg by the absence of marginal cirri. It seems likely that the two leftmarginal cirri were mistaken identified to be satellite transverse cirri. Calkins (1902) latertransferred Styloplotes appendiculatus Stein to the genus Diophrys. Rees (1883) briefly 'described a species of Styloplotes which he stated was the same asStyloplotes appendiculatus Stein as described by Fresenius (1865) which he (Rees, 1883)proceeded to call Styloplotes fresenii. Earlier Rees (1881) had described the new speciesStyloplotes grandis but later (Rees, 1883) concluded that it was identical to Styloplotesnorwegicus Quennerstedt, 1867 which he then erroneously called Styloplotes quennerstedti.In the present authors, opinion, Quennerstedt (1867) was mistaken in making Styloplotesnorwegicus Quennerstedt a synonym of Schizopus norwegicus Claparede & Lachmann. It ishere considered that the former species is a synonym of D. scutum because of the extent towhich the AZM is developed on the right side of the body. Rees (1883) also used thischaracter to distinguish Styloplotes quennerstedti, a synonym of D. scutum, from Styloplotesfresnii, a synonym of D. appendiculata. Diophrys scutum (Dujardin, 1841) Kahl, 1932 Ploesconia scutum Dujardin, 1841 (in part) 204 C. R. CURDS & I.C. H. WU 15;urn * * " - J --1.9 jfa. - * :- : fto i>*- % feiS-VS*!'><V^ V^-^'V- 1 ^-ni^.s Fig. 9 Diophrys appendiculata: (a) after Ehrenberg, 1838 (called Stylonychia appendiculatd);(b, c) after Claparede & Lachmann, 1 858 (called Schizopus norwegicus); (d, e) after Stein, 1 859a(called Styloplotes appendiculatus); (0 after Fresenius, 1865 (called Styloplotes appendiculatus);(g) after Calkins, 1902 (called D. appendiculatm; (h) after Kahl, 1932; (i,j) after Pierantoni,1909; (k) after Andrusova, 1886 (called Styloplotes appendiculatm var. pontica; (1) afterAndrusova, 1886 (called Planiplotes wagneri). 205 Fig. 10 Diophrys appendiculata, after Borror, 1963: (a) ventral surface; (b) dorsal surface; (c) nuclei. 20 um 20 urn Fig. 11 Diophrys scutum: (a) after Dujardin, 1841 (called Ploesconia scutum); (b) after Butschli,1 889 (called Diophrys grandis); (c) after Dragesco, 1963; (d-f) ventral surface, dorsal surface, andnuclei, after Borror. 1965a. 206 C. R. CURDS & I. C. H. WU Styloplotes norwegicus Quennerstedt, 1 867Styloplotes grand is Rees, 1881Styloplotes quennerstedti Rees, 1 883Diophrys grandis Butschli, 1889Diophrys magnus Raikov & Kovaleva, 1968Diophrys kasymovi Agamaliev, 1 97 1 DESCRIPTION (Figs 11, 12). This is the largest species (150-200 urn long) of the genus so fardescribed. It may be distinguished from the type species, D. appendiculata, by its size and bythe AZM which extends to the central body region on the right side. The body shape is ovoidwith an indentation in the posterior right. The dorsal surface is smooth and arched. Thereare 3 large sickle-shaped right caudal cirri, 7-8 fronto ventral, 5 transverse and 2 left marginalcirri. The two elongate macronuclei may be curved, nodular or moniliform and areaccompanied by up to six micronuclei. The dorsal silver-line system consists of 5-6dorsolateral kinetics interspersed with mesh-like argyromes. f Fig. 12. Diophrys scutum: (a, b) ventral and dorsal surfaces with section showing argyromes, afterAgamaliev, 1968; (c) after Raikov & Kovaleva, 1968 (called D. magnus); (d, ventral argyrome,nuclei and dorsal argyrome after Agamaliev, 1 97 1 (called D. kasymovi). REVIEW OF EUPLOTIDAE 207 NOTES. Claparede & Lachmann (1858) and Stein (\S59a) believed that the three illustrationsof Ploesconia scutum by Dujardin (1841) were of two different species. Claparede &Lachmann (1858) thought that one represented a Euplotes species whereas Stein (1859a)considered it to be Styloplotes appendiculatus Stein. The present authors doubt both of theseopinions but agree with Kahl (1932) who identified the species as Diophrys scutum andconsidered the AZM originating in the middle of the right border to be an importantdiagnostic feature. It is mainly by this character that D. scutum can be distinguished from D.appendiculata. Raikov and Kovaleva (1968) distinguished D. magnus from D. scutum by thedifference in the shapes of their macronuclei which is not a significant feature. The recentaddition D. kasymovi Agamaliev, 1971 has been included here in spite of it being rathersmall for this species. Fig. 13 Diophrys hystrix, after Buddenbrock, 1 920: (a) ventral surface; (b, c) nuclear features. Diophrys hystrix Buddenbrock, 1920 DESCRIPTION (Fig. 1 3). Diophrys hystrix is a small (30^0 urn long) species. The body outlineis generally oval but it is truncated anteriorly and there is the usual concavity on theposterior right accommodating three large sickle-shaped right caudal cirri. The dorsalsurface is strongly arched. The ten frontoventral cirri are arranged in two distinct groupswith 7 'frontals' at the anterior on the right of the peristome and 3 'ventrals' near thetransverse cirri. One of the 'ventrals' is adjacent and similar in size to the transversals, but itpoints in the opposite direction and is used in the creeping movements of the animal. Thereare 4 long transverse cirri and 2 small left marginal cirri located just behind the peristome.There are two ovoid macronuclei. NOTES. Kahl (1932) gave an almost identical description and illustration of this species as inBuddenbrock's (1920) original. Kattar (1970) identified a small (35-40 urn) species with twoovoid macronuclei as D. hystrix but this is a dubious identification since the author gave aninadequate description and the illustration was of a different shape and cirral number to thatof Buddenbrock (1920). Diophrys irmgard Mansfeld, 1 923 DESCRIPTION (Fig. 14). This is a medium sized (75-135 urn long) marine species. The bodyshape is characteristic, being broadly oval in outline but tending to be rectangular. The wideperistome is approximately half the body length and there is a prominent undulating 208 C. R. CURDS & I. C. H. WU m Fig. 14 Diophrys irmgard: (a, b) after Mansfeld, 1923; (c, d) after Kahl, 1932; (e) after Dragesco, 1 963 ;(f) after Borror, 1963. membrane on the right. The AZM bordering its left edge curves around the anterior end butonly just extends to the right side of the body. There are 3 right caudal, 9 frontoventral, 4-5transverse and 3 left marginal cirri. The two ovoid macronuclei are 8 urn in diameter andeach is associated with a micronucleus. NOTES. This species may be distinguished from the type species, D. appendiculata, by thepresence of 9 instead of 7 frontoventral cirri and because the left marginal cirri are in the'caudal' position. The shape of its body is consistently observed to be widely oval,rectangular and lacks the posterior lateral concavity on the right which is usually charac- REVIEW OF EUPLOTIDAE 209 teristic of the genus. Additionally, the right caudal cirri are attached dorsally like those ofUronychia. Horror (1963) described an organism which he called D. irmgard but as it hadonly 5 frontoventral cirri perhaps future studies will show this to be a separate species. Diophrys salina Ruinen, 1938 DESCRIPTION (Fig. 1 5). This is a small (3CMO um long) species. The outline shape of thebody is characteristically oval without any obvious lateral concavities but there is a posteriorindentation where the 3 large caudal cirri arise. The dorsal surface is strongly convex, but theventral surface is flattened. The peristome region extends to the centre of the body, with anAZM on the left and a well developed undulating membrane on the right. The frontoventralcirri are arranged in two groups with 4 right anterior 'frontals' and 4 'ventrals' situatedimmediately behind, and to the right of, the peristome. There are no left marginal cirri. Fivetransverse cirri lie between the caudal and 'ventral' cirri. The original description did notinclude any details concerning the nuclear apparatus. Feeds on diatoms, algae and bacteria. Fig. 15 Diophrys salina: (a, b) ventral surface and lateral view, after Ruinen, 1 938. Diophrys kahli Dragesco, 1 963 DESCRIPTION (Fig. 16). This species is about 80 um long, the body is elongate with a shortposterior narrowed tail region. The peristome is small, bordered on the right by a welldeveloped undulating membrane and on the left by a small AZM. The frontoventral cirriconsist of a group of 7 long 'frontals' and 2 small 'ventrals'. The 5 transverse cirri are long,the 2 left marginal cirri are in the 'caudal' position and there is only one right caudal cirrus.The two oval macronuclei are connected by a nuclear membrane and there are 4-6micronuclei. A few cilia have been observed along the left border of the animal. NOTES. Dragesco (1963) identified an organism that he considered to be identical to oneincompletely studied by Kahl (1932) which he called D. kahli. Untypically, the body of thisspecies is rather elongate and its left marginal cirri emerge caudally instead of immediatelyposterior to the peristome. Hartwig (1974) grouped this species with D. irmgard and D. 210 C. R. CURDS & I.C. H. WU Fig. 16 Diophrys kahli: (a, b) ventral surface and dorsal plaques, after Dragesco, 1963; (c) dorsalsurface, after Kahl, 1932 (called an 'incompletely studied Diophrys species'). multinucleata because of this latter feature. In contrast to other species in the genus, D. kahlihas only one right caudal cirrus, its peristome is rather small and the AZM is under-developed. Dragesco (1963) described the presence of small oval plaques on the dorsalsurface (Fig. 14b) which he suspected to be 'protrichocysts'. Diophrys oligothrix Borror, 1 965 DESCRIPTION (Fig. 17). Diophrys oligothrix is a medium sized (79-106 urn long), ovoidspecies. The concavity at the posterior right edge, so typical of the genus, is here incon-spicuous and the dorsal surface smooth. The AZM bordering the left edge of the largeperistome continues dorsally along the anterior region to the right side of the body. Theciliature and nuclear features are also typical of the genus, there are 7 fronto ventral, 5transverse, 2 left marginal and 3 right caudal cirri. The two irregular, elongate macronucleiare usually accompanied by four micronuclei. The silver-line system consists of 3 dorsalkinetics, one lateral kinety on the right and a fine mesh-like argyrome. The dorsal kinetiesbear, from left to right, 13-20, 16-24, 12-18 cilia and the ventral kinety 9-13 cilia. NOTE. This species is distinguished from D. scutum on the basis of its different number ofkineties. Diophrys peloetes Borror, 1965 DESCRIPTION (Fig. 18). D. peloetes is a medium sized (95-135 urn long) species typical of thegenus. It has a wide peristome which is about two-thirds of the body length, bordered by awell developed AZM on the left and an undulating membrane on the right. The AZM travelsalong the anterior dorsal edge onto the right side of the body where it extends about 3/10 ofthe body length. There are 3 prominent right caudal cirri, 5 long transverse, 7 frontoventraland 2 left marginal cirri. The two macronuclei are elongate. There are 8 dorsolateral REVIEW OF EUPLOTIDAE 211 Fig. 17 Diophrys oligothrix, after Borror, 19656: (a) ventral surface; (b) dorsal surface; (c) frontal section showing nuclei. kinetics, each bearing 6-17 cilia. The number of kinetics distinguishes this species from D.scutum which has only five. NOTES. Borror (1965a) established D. peloetes after a morphometric comparison with D.scutum. He stated that the new species was 'extremely similar to D. scutum except for havingeight dorsal rows of cilia instead of five, having fewer cilia per row, and having a significantlyshorter AZM'. He also pointed out that he had identified it incorrectly as D. scutum in anearlier study (Borror, 1963). Diophrys quadricaudatus Agamaliev, 1967 DESCRIPTION (Fig. 19). This is a medium sized (100-1 10 urn long) species, characterised bythe presence of 4 right caudal cirri arising from a particularly prominent concavity on theposterior right of an otherwise oval body. The wide peristome is about 48 um long and on itsright there is a conspicuous undulating membrane. The AZM is composed of 50-60membranelles which extend forward over to the right side of the body. There are 7fron to ventral, 5 transverse, and 3 left marginal cirri. The dorsal silver-line system consists of5 or 6 dorsolateral kinetics and a finely-meshed dorsal argyrome. There are two anterior andone posterior macronuclei. 212 C. R. CURDS & I. C. H. WU Fig. 18 Diophrys peloetes: (a, b) ventral and dorsal surfaces, after Borror, 1963 (called D.scutum); (c-e) ventral and dorsal silver-line systems, and nuclei, after Borror, 1965a. NOTES. Agamaliev (1967) considered the presence of four right caudal cirri and threemacronuclei to be the diagnostic features of this species. The uneven number of macron ucleiis rather strange and could perhaps represent an interdivisional state. Diophrys scutoides Agamaliev, 1967 DESCRIPTION (Fig. 20). D. scutoides is 1 10-120 urn long. The body is oval with a prominentconcavity on the posterior right edge. The AZM of 75-80 membranelles borders the left edgeof a wide peristome and continues along the anterior edge to extend a third of the body length REVIEW OF EUPLOTIDAE 213 20yum Fig. 19 Diophrys quadricaudatus, after Agamaliev, 1967: (a) ventral surface of living cell (b-d)stained ventral surface, dorsal surface and nuclei. down the right side. There are 3 right caudal, 5 transverse and 2 left marginal cirri but only 5frontoventral cirri. The two elongate macronuclei are nodular and are accompanied by threemicronuclei. The silver-line system consists of 5 dorsolateral kinetics with mesh-like dorsaland ventral argyromes. 214 C. R. CURDS & I. C. H. WU Fig. 20 Diophrys scutoides, after Agamaliev, 1967: (a) ventral surface; (b) dorsal surface; (c, d) ventral and dorsal argyromes. NOTES. Agamaliev (1967) described this species from the Caspian Sea. Apart from beingslightly smaller in size, it differs from D. scutum in having five instead of seven frontoventralcirri. Diophrys tetramacronucleata Kattar, 1970 DESCRIPTION (Fig. 21). This is a small (55-65 urn long) species. The body is oval except forthe concavity at the posterior right edge. The AZM consists of about 36 membranelles whichextend about halfway down the body on the left but hardly at all on the right. There are 7frontoventral, 5 transverse, 2 left marginal and 3 right caudal cirri. There are 5 dorsolateralkinetics and 4 ovoid macronuclei. NOTES. The original description of this species was brief and only the presence of fourmacronuclei distinguished it from the type species. Although it is suspected that Kattar(1970) described an interdivisional specimen, it is provisionally included here as a nominalspecies. Kattar (1970) stated that protargol impregnation showed the presence of threemedian and two dorsolateral kinetics he provided no diagram of these structures. The REVIEW OF EUPLOTIDAE 215 Fig. 21 Diophrys tetramacronucleata, after Kattar, 1 970: (a) ventral surface; (b) nuclei. redescription and photographs of this species by Hartwig (1974) are also incompletealthough the presence of four macronuclei was clearly demonstrated. Diophrys multinucleata Hartwig, 1973 DESCRIPTION (Fig. 22). D. multinucleata is a medium sized (76-11 Sum long) highlythigmotactic species. The body is oval to rectangular in outline with the dorsal surface beingarched and the ventral surface flattened. There is a small but distinct projection on theanterior right corner of the body. The peristome is about two-thirds of the body length. TheAZM consists of about 30 membranelles lying along the left of the peristome and a few largermembranelles along the anterior edge of the body. The frontoventral cirri are arranged intwo distinct groups: the 5-6 large 'frontals' are in the anterior and the two very small'ventrals' lie close to the transverse cirri. There are 4 long transverse, 3 right caudal and 3 leftmarginal/caudal cirri. There are at least 20 macronuclei present. NOTES. This species is characterised by the presence of over twenty macronuclei which isunique in the genus, otherwise it is similar to D. irmgard in cirral arrangement, particularlyin the caudal positioning of the left marginal cirri which Hartwig (1973, 1974) considered tobe a feature of diagnostic importance. Genus URONYCHIA Stein, 1859 Introduction Stein (1859#, 18596) established and described the genus Uronychia and transferredTrichoda transfuga Muller, 1786 to the genus as the type species. Between 1901 and 1928,five new species were described and Kahl (1932) included a key to them in his classicalcompendium. Since then only one addition, Uronychia bivalvorum Fenchel, 1965, has been 216 C. R. CURDS & I.C. H. WU Fig. 22 Diophrys multinucleata, after Hartwig, 1973: (a) dorsal surface; (b) ventral surface showing nuclei. made which he found in the mantle cavity of certain lamellibranch molluscs. Uronychia ischaracterised by its enormous peristome region occupying half of the ventral surface withconspicuous peristomial membranes, the three posterior groups of large cirri, the apparentabsence of frontoventral cirri and by its rapid backward motion (Bullington, 1940). How-ever, there are few features which clearly distinguish the described species from one another.Many of the features used in the past are variable so that here only four nominal species arerecognised. (a) Shape. The typical Uronychia body is oval. Dorsal ribs, ridges or striations are oftenmentioned in descriptions but there are few data on the variability of these structures.Ventrally, there are two large posterior cavities which accommodate the transverse and leftmarginal cirri. In general, the right caudal cirri of Uronychia are attached more dorsally thanthose in Diophrys, so the oval body outline is without the dorsolateral concavity at theposterior right which is characteristic of the latter genus. However, Buddenbrock (1920)described U. heinrothi with an 'Ausschnitt' (notch) on the posterior right edge which canmake the posterior border S-shaped. Similarly, Taylor (1928) described an 'uncinus' on U.uncinata which he stated apparently owes its origin in part to an attenuation of the remnantwhich holds proter to opisthe during the final stages of division. It seems likely that the 'Ausschnitt' and 'uncinus' are the same structure and asBuddenbrock (1920) noted that it was variable in extent and absent from some individualsthen it cannot be used as a reliable diagnostic character. Indeed if it is formed during divisionthen it could only be present in the proter of a dividing cell. (b) Size. The distinctive sizes of U. magna and U. setigera are here considered to be REVIEW OF EUPLOTIDAE 217 diagnostically significant. The former at 450 um long is the largest species so far described inthe family Euplotidae, and the latter at 40-50 jim long is the smallest Uronychia species. Thesizes of the remaining species fall within the range of 70-250 um long. The size of the typespecies U. transfuga has been noted to vary between 50-1 50 um long (Kahl, 1932) butBullington (1940) separated U. heinrothi from U. transfuga by the larger size (129-264 umlong) of the species he observed. However, Buddenbrock (1920) described U. heinrothi ashaving the range 70-250 urn in length. It is apparent that the size variation of these twospecies is great and their ranges overlap. Since there are no other significant distinguishingfeatures, these two species are here regarded as being synonymous. (c) Adoral Zone of Membranelles. The ventral surface of Uronychia is dominated by anenormous peristome with conspicuous membranelles and membranes. Unlike the rest of theEuplotidae, the AZM of Uronychia consists of large membranelles along the anterior of thebody and 4-5 smaller paroral membranelles at the posterior left of the peristome and thereare undulating membranes along both sides of the peristome. Fenchel (1965) was of theopinion that it was the fusion of the adoral membranelles which forms the membrane alongthe left peristomial border. These peristomial membranes and membranelles may beobserved to be closed over the peristome region or spread out and wing-like. The paroralmembranes are often found in a pocket-like invagination of the peristomial wall. Buddenbrock (1920) described the presence of two large peristomial membranes in U.heinrothi but these appear to be split into five parts in one of his diagrams. Bullington (1940)observed four peristomial membranes on his specimens of U. heinrothi plus another largemembrane close to the posterior right edge of the body. This marginal membrane has beendescribed only on this occasion so it has not been used here as a diagnostic character.Although the peristomial membranes of U. heinrothi appear to be larger and more numerousthan has been observed in U. transfuga, it is not considered to be a reliable character. On theother hand, Calkins (1902) described 'flagella-like' cirri in the peristome of U. setigera whichhave subsequently been noted by other authors (Buddenbrock, 1920; Young, 1922; Kahl,1932; Kattar, 1970). This feature is apparently distinctive and consistent and is thereforeuseful in the identification of U. setigera. Fig. 23 Arrangement and numbering of cirral streaks in Uronychia transfuga, after Wallengren, 1901. 218 C. R. CURDS & I.C. H. WU (d) Cirri. One of the most distinctive generic features of Uronychia is the apparent absence offrontoventral cirri. Wallengren (1901) studied cirral morphogenesis in U. transfuga and wasfirst to number the cirral streaks (Fig. 23). He found that there was a small residuum of cirralelements I/I, II/2 and 111/2 grouped closely together at the anterior right near the adoralmembranelles which he described as membranous cirri. Young (1922) noted these in threeUronychia species and described them as cirri-like membranelles, but it was Buddenbrock(1920) who illustrated three similar small cirrus-like structures at the anterior left as well ason the right. Wallengren (1901) showed that the cirri V/2 and VI/2 which form the 'ventrals'in Euplotes and Diophrys lie among the transversals in U. transfuga and they appear like'satellites' in this group of cirri. These cirri have not yet been observed by other authors butBuddenbrock (1920), Young (1922) and Taylor (1928) described the presence of one slendercirrus at the right of the transversals in U. setigera, U. heinrothi and U. uncinata. All species of Uronychia have 4 or 5 transverse cirri. Young (1922) considered thepossession of four transversals in U. binucleata to be a specific character even though herecorded that the number of transversals in U. transfuga varied between 4 and 5. To the leftof the ventral cavity from which the transversals emerge, there is another cavity whichaccommodates the two enormous left marginal cirri with a variable number of small cirri orcilia. The right caudal cirri emerge dorsally on the posterior right and are invariably sickle-shaped. All of the described species have three right caudals except U. magna which wasshown as having two by Pierantoni (1909), although he also observed only two right caudalsin U. transfuga which others have shown to possess three. Bullington (1940) discussed theattachment and movement of the right caudal cirri in Uronychia (Fig. 24). In addition to these three groups of large cirri, two small right marginal cirri have beenobserved in most Uronychia species. Fenchel (1965) distinguished U. bivalvorum from U.transfuga partly by the absence of these two small cirri but it should be noted that both U.transfuga and U. setigera have also been described without right marginals. Fig. 24 Arrangement of the right caudal cirri of Uronychia transfuga, after Bullington, 1940(called U. heinrothi): (a) showing two large cirri attached in the mid-dorsal region of the posteriorbody end with a cirrus on the right; (b) showing two defimbriated cirri attached to either side ofthe mid-line and one cirrus attached in a pocket near the right edge and slightly below the othertwo cirri. REVIEW OF EUPLOTIDAE 219 Fig. 25 Stages in cell division of Uronychia transfuga, after Calkins, 1911: (a) resting stage; (b)condensation of the macronucleus; (c, d) later stages in division; (e) just before cell separation;(f-h) daughter cells immediately, 15 minutes and 1 hour after separation. Kahl (1932) noted the presence of a long slender right caudal cirrus in U. binucleata whichhe considered to be of diagnostic value, similarly Young (1922) described such a cirrus in U.setigera although no others have reported its presence. It is evident that cirral numbers andarrangements in this genus are highly variable and therefore of little diagnostic importance. (e} Nuclear features. Kahl (1932) relied heavily on the nuclear features of Uronychia in hiskey to the species. For example, he stated that U. transfuga had, without exception, twosausage-shaped macronuclei with a micronucleus and that U. magna and U. heinrothi hadtheir nuclei split into many fragments. However, U. transfuga had already been seen with itsmacronucleus in many fragments. Calkins (1911) described the nuclear reorganisation of U.transfuga and showed that the macronucleus just after cell division was in two parts but thatit was a fragmented C-shape when at the resting stage (Fig. 25). The macronucleus of U.transfuga has been described as having a fragmented C-shape, like a string of beads orpartially fused, and in two parts (Fig. 26) (Buddenbrock, 1920; Bullington, 1940). Borror(1972) has also noted that the macronuclear shape and number were variable and for thisreason considered U. heinrothi to be synonymous with the type species. Calkins (1902) described U. setigera with a single ovoid macronucleus but Young (1922)observed two irregular macronuclei in his specimens of the same species. Fenchel (1965)distinguished U. bivalvorum from the type species partly on the basis of its nucleus which hedescribed as 'an irregular structure divided into two or more parts of unequal size': thiswould lie within the range of variability already noted above for U. transfuga. Kattar(1970)distinguished between U. transfuga and U. setigera by their possession of 9-15 and 2 nodularmacronuclei respectively, although he also concluded that the six species recognised by Kahl(1932) including U. setigera could be varieties of U. transfuga. Although the number andshape of the macronuclei in Uronychia appear to be highly variable this is not the case with 220 C. R. CURDS & I. C. H. WU ^0 Fig. 26 Nuclear variation in Uronychia transfuga, after Buddenbrock, 1920 (called V. heinrothi). the micronucleus. Most species possess a single micronucleus but U. binucleata is charac-terised by the presence of two micronuclei. Only one other species, U. magna, has beenrecorded with two micronuclei. (/) Dorsal silver-line system. Kattar (1970) noted the presence of four kinetics on the dorsalsurface of U. transfuga and five on U. setigera, while Reiff (1968) illustrated five dorsolateralkinetics on the former species. The kinetics which have only recently been described appearto correspond in number and position with the dorsal ridges and striations that aresometimes described by earlier authors. However, far more data are required before thediagnostic importance of the silver-line systems of Uronychia can be assessed. Diagnosis of Uronychia Marine hypotrichs of variable size, mostly within the range 50-250 urn long, but up to450 um long. Body oval and smooth in outline, dorsal surface smooth or with ridges.Peristome large with conspicuous membranes. The AZM is limited to the anterior borderand the paroral region. There are 4-5 transverse and 2 left marginal cirri which emerge fromprominent ventral cavities. The 3 large right caudal cirri are attached to the dorsal surfaceand there are sometimes 2 right marginal cirri present. The frontoventral cirri are reduced toa field of 3 near the origin of the AZM. The macronuclei are variable in number and shapeand there are 1 or 2 micronuclei. Characteristically moves backwards rapidly. Key to the species of Uronychia 50-250 um long Smaller than 50 um long or larger than 250 um long With a single micronucleus With two micronuclei Peristomial cirri present Peristomial cirri absent 40-50 um long About 450 um long About 450 urn long, macronucleus moniliform and C-shaped60-80 um long, macronucleus in 3-5 pieces .... 2435 U. setigera U. transfuga U. setigera U. magna U. magna U. binucleata Species descriptionsUronychia transfuga (Muller, 1 786) Stein, 1 859 Trichoda transfuga Muller, 1 786Ploesconia scutum Dujardin, 1841 (in part)Campylopus paradoxus Claparede & Lachmann, 1 858Uronychia heinrothi Buddenbrock, 1920Uronychia uncinata Taylor, 1928Uronychia bivalvorum Fenchel, 1965 REVIEW OF EUPLOTIDAE 221 f -M- P Fig. 27 Uronychia transfuga: (a, b) ventral and dorsal surfaces, after Stein, 1859a; (c, d) ventraland dorsal surfaces, afier Claparede & Lachmann, 1858 (called Campylopus paradoxus); (e)dorsal surface, after Dujardin, 1841 (called Ploesconia scutum); (f, g) ventral and dorsal surfacesshowing nuclei, after Pierantoni, 1909; (h) ventral surface and nuclei, after Calkins, 1911; (i)dorsal surfaces and nuclei, after Young, 1922. 222 C. R. CURDS & I. C. H. WU b FMZ vFM rc IMC Fig. 28 Uronychia transfuga: (a-c) ventral and dorsal surfaces showing nuclei, after Kahl, 1 932 (ccalled U. uncinata); (d-f) ventral and dorsal surfaces and nuclei, after Fenchel, 1965 (called V.bivalvorum); (g, h) ventral surface and dorsal surfaces showing nuclei, after Reiff, 1968. REVIEW OF EUPLOTIDAE 223 Fig. 29 Uronychia transfuga: (a, b) ventral surface and nuclei, after Kattar, 1970; (c) ventralsurface, after Fenchel, 1965 (called U. bivalvorum)', (d-g) ventral surface, dorsal surface showingnucleus, ventral view and lateral view of peristomial membranes, after Buddenbrock, 1920(called U. heinrothi); (h) ventral surface, after Bullington, 1940 (called U. heinrothi); (i) dorsalsurface, after Taylor, 1928 (called U. uncinata). 224 C. R. CURDS & I.C. H. WU DESCRIPTION (Figs 27, 28 & 29). This is a cosmopolitan species. The oval shaped body isvariable in size (50-260 urn long), its dorsal surface is arched and may be smooth or with 3-4ridges. At the posterior right there are the 3 characteristically sickle-shaped right caudal cirri.The large peristome and two posterior cavities occupy most of the ventral surface.Prominent adoral membranelles are situated along the anterior edge of the body and theseemerge dorsally. At the posterior left of the peristome, there are 4-5 paroral membranelleslying in a pocket-like invagination. Large undulating membranes border two sides of theperistome. There are 4-5 large transverse cirri implanted in the larger posterior cavity on theright, sometimes with 1 or 2 satellite cirri. In the posterior left cavity there are 2 largemarginal cirri sometimes accompanied by a few, usually 2, small cirri or cilia. Themacronucleus varies from being a moniliform (5-13 segments) C-shape at resting stage, tobeing 2 irregular masses immediately after cell division. The silver-line system consists of3-5 dorsolateral kinetics. It has been observed in the mantle cavities of the lamellibranchmolluscs Thyasiraflexuosa and T. sarsi. 10JJ Fig. 30 Uronychia setigera: (a) after Calkins, 1902; (b-d) after Young, 1922; (e) after Kattar, 1970. REVIEW OF EUPLOTIDAE 225 NOTES. Stein (1859<2, b) established the genus Uronychia and redescribed U. (Trichoda)transfuga Muller as the type species. U. transfuga has since been identified by many workersfrom various localities (see Hartwig 1973, 1974). As discussed earlier U. heinrothi, U.uncinata and U. bivalvorum were established on characters that are now known to bevariable and unreliable. These three species are here regarded as synonyms of the typespecies until more convincing data become available. Uronychia setigem Calkins, 1 902 DESCRIPTION (Fig. 30). This is the smallest (40-50 urn long) species yet described. The dorsalsurface is arched and is sculptured longitudinally by 3-4 ridges. There are 2-3 flagella-likecirri in the large peristome and there are about 4 paroral membranelles. Wide undulatingmembranes lie on each side of the peristome. There are 3 curved right caudal cirri attacheddorsally and a long thin satellite cirrus has been observed. In the two ventral cavities thereare 4-5 transverse cirri and 2 large left marginal cirri. The macronucleus is usually in 2 band-like parts with a micronucleus between them but a single spherical macronucleus has alsobeen observed in this species. There are 5 dorsolateral kinetics. NOTES. This species is characterised by the presence of flagella-like cirri extending from theposterior left margin into the peristome (Calkins, 1902). This author further distinguished itfrom the type species by its single ovoid macronucleus; however later authors (Young, 1922;Kattar, 1970) reported two macronuclei. Kattar (1970) impregnated the cell with protargolbut failed to clearly illustrate the silver-line system. cl ..ctrr. cl SOyum Fig. 31 Uronychia magna, after Pier^.ntoni, 1 909; (a) ventral surface; (b) dorsal surface. 226 C. R. CURDS & I. C. H. WU Uronychia magna Pierantoni, 1909 DESCRIPTION (Fig. 3 1 ). This is the largest species yet recorded (450 um long). The body shapeis approximately oval in outline but is ornamented at either end. The anterior dorsal edge isserrated and there are two short spines on the posterior left dorsal border. The dorsal surfaceis longitudinally striated. The peristome is extensive, occupying most of the ventral surface.The AZM is located anteriorly in a posterior pocket in the peristome. There are two largeundulating membranes, one on each side of the peristome. There are 6 transverse, 2 leftmarginal and 3 right caudal cirri. The macronucleus is in many pieces forming a string ofbeads which follow the curve of the left side of the body so that it is C-shaped. Anteriorlythere are 2 micronuclei. NOTE. This species has apparently been described on a single occasion. Uronychia binucleata Young, 1 922DESCRIPTION (Fig. 32). Uronychia binucleata is 60-80 urn long. The dorsal surface is arched 10yum Fig. 32 Uronvchia binucleata: (a-c) ventral surface, dorsal surface and nuclear features, after Young, !922;(d)afterKahl, 1 932. REVIEW OF EUPLOTIDAE 227 and decorated with many small pits. The peristome is bordered on either side by a largeundulating membrane and at the anterior left there are 3 delicate sickle-shapedmembranelles. About 4 paroral membranelles lie in the 'buccal' pocket. There is a longslender cirrus accompanying the 3 curved right caudal cirri, 4 transverse, 2 left marginal and2 right marginal cirri. The macronucleus is in 3-5 fragments and 2 micronuclei. NOTES. Young (1922) established this species which he showed to be almost identical to U.setigera in his comparative study. He distinguished his species by the absence of peristomialcirri, the presence of 4 or less transverse cirri, and the presence of a slender right caudal cirrusin addition to the two micronuclei. Although Young (1922) described a satellite cirrus in U.setigera which he considered to be significant, curiously he failed to mention it in U.binucleata even though he illustrated it. Here the presence of two micronuclei is used tocharacterise the species: should future studies show that the number of micronuclei to bevariable, then it would become synonymous with U. transfuga. Genus CERTESIA Fabre-Domergue, 1885 Introduction Fabre-Domergue (1885) described an organism that was very similar to Euplotes except thatit had a row of left marginal cirri, no caudal cirri and the macronucleus was in four parts. Inhis view this was sufficient to create the new genus Certesia although later both Biitschli(1 889) and Sauerbrey (1928) considered it to be a subgenus of Euplotes. Since the descriptionof the type species, Certesia quadrinucleata Fabre-Domergue, 1885, only one other species,Certesia ovata Vacelet, 1960, has been described. Vacelet (1960) distinguished it from thetype species on account of its smaller size, its more oval shape and its two curved lefttransverse cirri. Here, these differences are not considered to be sufficient to treat C. ovata asa separate species. Diagnosis of Certesia Oval marine hypotrichs with an anterior nose-like projection on the right anterior bodyedge. There is a conspicuous AZM which extends a third to halfway down the left side of thebody. There are 1 1-13 frontoventral, 5 large transverse, and a variable number (6-1 1) of leftmarginal cirri. There are no caudal cirri. Macronucleus in several parts. Single species genus. 20/jm Fig. 33 Certesia quadrinucleata: (a) after Fabre-Domergue, 1885; (b) after Sauerbrey, 1928; (c)after Kahl, 1932;(d, e) after Vacelet, 1960 (called C. ovata). 228 C. R. CURDS & I. C. H. WU Species description Certesia quadrinucleata Fabre-Domergue, 1885Certesia ovata Vacelet, 1 960 DESCRIPTION (Fig. 33). Certesia quadrinucleata is a medium sized (75-80 fim long), oval,rigid and colourless marine hypotrich. The dorsal surface is arched and smooth: the ventralsurface flat. The right body border is convex and the left may be slightly concave. There is adistinct nose-like projection on the right of the extreme anterior body edge. The peristome isa third to a half of the body length. The AZM consists of about 5 large membranelles alongthe anterior border and 1 5-20 smaller ones along the left edge of the peristome. There are1 1-13 frontoventral cirri, 5 large transverse cirri, a row of 6-1 1 left marginal cirri but nocaudal cirri. Four ovoid macronuclei are located in pairs, one pair on the right anterior andone pair on the left below the AZM. Genus GASTROCIRRHUS Lepsi, 1928 Introduction Lepsi (1928) created the genus Gastrocirrhus and stated that the type species Gastrocirrhusintermedium Lepsi, 1928 possessed characters that were intermediate between the oligotrichsand the hypotrichs. The organism had ventral cirri arranged in groups like those ofhypotrichs, but its large anterior funnel-shaped peristome and spiral AZM were more likethose of oligotrich genera such as Stentor. Kahl (1932) thought that the organism wasprobably a piece of a fragmented Oxytricha, but soon after Bullington (1940) establishedanother species Gastrocirrhus stentoreus Bullington, 1940. This has been followed by severalother species descriptions. The six cirri along the right side of G. intermedius are here interpreted to be frontoventralsand the remaining ten cirri are called caudals. Bullington (1940) recorded the presence offour marginals, four ventral and two oral/anterior cirri on G. stentoreus. These areconsidered to be frontoventral cirri so that this species may be diagnosed by the presence ofonly five caudal cirri. In Gastrocirrhus adhaerens Faure-Fremiet, 1954, there are 16 cirriarranged in two rows on the right of the peristome, and 12 in a semi-circle at the posterior.Here, these are interpreted to be frontoventral and caudal cirri respectively. Gastrocirrhustrichocystus Ito, 1958 has 18 frontoventral and 13 caudal cirri arranged similarly to those inG. adhaerens but the species is characterised by the presence of zones of trichocysts(mucocysts?) on the dorsal and ventral surfaces. Faure-Fremiet (1961) created the family Gastrocirrhidae and included the single speciesgenus Cirrhogaster Ozaki & Yagui, 1942 in the family. The type species, Cirrhogastermonilifer Ozaki & Yagui, 1942 is similar to G. adhaerens, but the former species has 10rather than 16 frontoventral cirri. Dragesco (1965) suggested that these two species weresynonymous but here, although C. monilifer is considered to be a species of Gastrocirrhus itis not considered to be a synonym of G. adhaerens. Diagnosis of Gastrocirrhus Marine hypotrichs that may be dorso vent rally flattened or cup-shaped. There is a largeanterior funnel-shaped peristome which opens both anteriorly and ventrally. A welldeveloped AZM borders the C-shaped anterior body edge and winds anti-clockwise down theleft of the peristome. There are 5-18 frontoventral cirri, usually arranged in two obliquerows on the right of the peristome: 5-13 caudal cirri along the posterior pole of the bodywhich in some species curve forwards ventrally on the left to give the appearance oftransverse cirri. Macronucleus may be oval or moniliform in 11-15 pieces. REVIEW OF EUPLOTIDAE 229 Key to the species of Gastrocirrhus \ With 10 fronto ventral cirri 4 With 6, or more than 10 frontoventral cirri 2 2 With 6 frontoventral cirri G. intermedius With more than 10 frontoventral cirri 3 3 With 1 6 frontoventral cirri and 1 2 thigmotactic caudal cirri, without 'trichocyst' zones G. adhaerensWith 1 8 frontoventral cirri, 1 3 caudal cirri and with dorsal and ventral zones of 'trichocysts' G. trichocystus 4 With 5 caudal cirri G. stentoreus With 12-13 caudal cirri G.monilifer 10/jm Fig. 34 Gastrocirrhus intermedius, after Lepsj, 1 928.Species descriptions Gastrocirrhus intermedius Lepsi, 1928 DESCRIPTION (Fig. 34). This species is a dorso ventral ly flattened cup-shape, approximately70 jim long. The well developed AZM borders the anterior funnel-shaped peristome. Thereare 16 cirri in three distinct groups. Near the right body margin there are 6 frontoventralcirri. The caudals are situated along the posterior pole, comprised of a group of 3 sickle-shaped cirri, rather like the right caudal cirri in Diophrys, and 7 others to their left which aresmaller and straighten The macron ucleus is oval. NOTES. There is only a single description of this, the type species of the genus. Theobservations of Lepsi (1928) appear to have been made exclusively on living specimens. Thearrangement of the cirri and nuclear features do not appear to be as typical as those of thespecies described later. Gastrocirrhus stentoreus Bullington, 1940 DESCRIPTION (Fig. 35). Gastrocirrhus stentoreus is a medium sized (100 urn long, 70-80 umwide) perfectly cup-shaped species. The large funnel-shaped peristome has a ventral,elongate opening about two-thirds of the body length and is bordered by a prominent AZM.There are 1 1 frontoventral cirri on the right of the peristome in groups of two, four and fivecirri. At the posterior right margin, there are 5 sickle-shaped caudal cirri. 230 C. R. CURDS & I. C. H. WU^ SOiim Fig. 35 Gastrocirrhus stentoreus, after Bullington, 1 940. Fig. 36 Gastrocirrhus adhaerens: (a) ventral surface; (b) attached to substratum showing nuclei. Both after Faure-Fremiet, 1954. Gastrocirrhus adhaerens Faure-Fremiet, 1954 DESCRIPTION (Fig. 36). This is a medium sized (100 urn long), typically cup-shaped species.The peristome diameter is slightly less than the body length. The AZM, which winds aroundthe anterior of the cell and down the left side of the peristome, consists of about 150membranelles. The species is characterised by its 12 very long thigmotactic caudal cirrialong the posterior border of the cell which can be retracted when not attached to asubstratum. There are 16 frontoventral cirri arranged in two rows of 8 cirri on the right of the REVIEW OF EUPLOTIDAE 231 peristome. The moniliform macronucleus is composed of about 12 pieces: 3-5 micronucleihave been observed. Gastrocirrhus monilifer n. comb.Cirrhogaster monilifer Ozaki & Yagui, 1942 DESCRIPTION (Fig. 37). Gastrocirrhus monilifer is 95-105 urn long, 75-90 urn wide and cup-shaped. The ventral opening into the large funnel-shaped peristome is about a half the bodylength. The 10 frontoventral cirri, arranged in two rows, are restricted to the small area onthe right of the peristome. There are 12 long caudal cirri. The moniliform macronucleusconsists of 1 1-1 5 pieces and there are 8 micronuclei scattered along its length. NOTES. Dragesco (1965) suggested that this species should be a synonym of G. adhaerens.Here it is considered to be a separate species until further information concerning thevariability in numbers of frontoventral cirri become available. Fig. 37 Gastrocirrhus monilifer, after Ozaki & Yagui, 1 942 (called Cirrhogaster monilifer). Gastrocirrhus trichcystus Ito, 1958 DESCRIPTION (Fig. 38). Gastrocirrhus trichocystus is 90-103 um long, 82-90 urn wide andcup-shaped. The ventral surface becomes flattened when starved. The large peristome isbordered by numerous long adoral membranelles. There are 18 frontoventral and 13 caudalcirri separated by longitudinal ridges on the ventral surface. The sickle-shaped macronucleusis moniliform with 10-12 pieces. There are usually 8 micronuclei. The species ischaracterised by bands of 'trichocysts' (mucocysts?) along the anterior and right margin ofthe dorsal surface, also they are found along the left margin and at the anterior right on theventral surface. 232 C. R. CURDS & I. C. H. WU II oo O Q/ Fig. 38 Gastrocirrhus trichocystus, after Ito, 1958: (a) ventral surface; (b, c) ventral and dorsal surfaces showing trichocyst distribution. Genus EUPLOTASPIS Chatton & Seguela, 1936 Introduction Chatton & Seguela (1936) described a Euplotes-like hypotrich found in the branchial cavityof the sea squirt, dona intestinalis. They noted that the organisms were never observed onor around their host and they did not survive without their host for more than 36 hours in amedium which had been used successfully for the culture of many marine Euplotes. This species was found to have many characters similar to those of Euplotes but Chatton &Seguela (1936) noted that its body shape and very thick short cirri were more like those ofAspidisca. Furthermore, the AZM was entirely ventral so that it did not border the anteriordorsal surface as in Euplotes, and four of the frontoventral cirri were split into two partswhich is a distinctive feature. Corliss (1961) placed this genus along with Paraeuplotes in thefamily Paraeuplotidae but later (Corliss, 1977) he transferred it to the family Aspidiscidae. REVIEW OF EUPLOTIDAE 233 Diagnosis of Euplotaspis Marine oval hypotrichs living in sea squirts. There is a prominent AZM restricted to theventral surface. There are 9 frontoventral, 5 transverse and 34 caudal cirri. The fourfrontoventral cirri towards the right body border are split longitudinally into two parts.Macronucleus C-shaped. The dorsal argyrome is like that of Euplotes vannus (see Curds,1975). Species description Euplotaspis cionaecola Chatton & Seguela, 1936 DESCRIPTION (Fig. 39). Euplotaspis cionaecola is a 60-70 um long marine hypotrich found in rv-'x ; <N > \ ::..x A u ,/ /--T ,-(' 1A _^ ^ ri 4---"\- J ! ' * A ,.-\ f- v 4 <v ..*.-.-4 I-_>- ) v - V""' i Fig. 39 Euplotaspis cionaecola, after Chatton & Seguela, 1936: (a) ventral surface of living cell;(b, c) ventral and dorsal surfaces of silver-impregnated specimens. 234 C. R. CURDS & I. C. H. WU the branchial cavity ofCiona intestinalis. The body outline is perfectly oval, convex dorsallyand flattened or slightly concave ventrally. The AZM of about 50 membranelles, is aprominent band parallel to, and at a short distance from, the left body margin. There are 10short, wide frontoventral cirri: the four on the right being split longitudinally. The 5transverse cirri are separated by marked ridges on the ventral surface. The 3-4 caudal cirriare small and indistinct in vivo but can be seen in silver impregnated specimens. The dorsalsilver-line system consists of 7-8 dorsal kinetics, each carrying about 1 5 cilia, with simplecross-links between the kinetics. The macronucleus is C-shaped and a single micronucleushas been observed at its posterior left. Genus EUPLOTIDIUM Noland, 1937 Introduction Noland (1937) described a hypotrich found in sponges from the Gulf of Mexico and statedthat 'the organism differs from Euplotes, its nearest relative, in the absence of caudal cirri,and in the more cylindrical shape of the body'. Consequently, he erected the genusEuplotidium Noland, 1937 and called the organism Euplotidium agitatum Noland, 1937because of its erratic movements. Ito (1958) also described a cylindrical Euplotes-\ikeorganism, but in this species there was one left caudal cirrus and more frontoventral andtransverse cirri than in E. agitatum, this he called Euplotidium itoi Ito, 1958. A furtherspecies with different numbers of frontoventral and transverse cirri, Euplotidium arenariumMagagnini & Nobili, 1964, was later described and Borror (1972) transferred Euplotespsammophilus Vacelet, 1961 to the genus. More recently Hartwig (1980) added anotherspecies to the genus. Diagnosis of Euplotidium Small to large (65-200 um long) marine hypotrichs. Outline shape slightly elongated oval.Rounded in cross-section, never dorso ventrally flattened. The peristome is a wide funnel-shape with a prominent AZM that borders both the anterior semi-circular body edge and theleft margin of the peristome. There are 7-12 frontoventral, 5-6 transverse and, when present,a reduced number (1-2) of caudal cirri. Key to the species of Euplotidium 1 With less than 10 frontoventral cirri 2 With 10 or more frontoventral cirri 4 2 With 9 frontoventral cirri but without caudal cirri E. agitatum With 7 frontoventral and 1 or 2 caudal cirri 3 3 With 1 caudal cirrus, macronucleus in many parts E. helgae With 2 caudal cirri, macronucleus elongate E. psammophilus 4 With 10 frontoventral and more than 1 caudal cirri E. arenarium With 12 frontoventrals and 1 caudal cirrus E. itoi Species descriptions Euplotidium agitatum Noland, 1937 DESCRIPTION (Fig. 40). This, the type species, is 65-95 um long and its cylindrical bodyshape is most apparent when viewed from the anterior. There is a funnel-shaped peristomewhich is lined around its anterior rim and left side by a prominent AZM. There areapproximately 40 membranelles on the anterior part and about the same number in the REVIEW OF EUPLOTIDAE 235 Fig. 40 Euplotidium agitatum, after Noland, 1937: (a) ventral surface; (b) dorsal surface; (c) view from anterior showing cylindrical body. ventral part of the AZM. There are 9 frontoventral cirri, aligned in 2 rows, near the rightbody edge and 5 conspicuous transverse cirri but caudal cirri are not present. NOTE. Originally isolated from water squeezed from sponges in the Gulf of Mexico. Euplotidium itoi Ito, 1958 DESCRIPTION (Fig. 41). This is a medium sized (89-95 um long), cylindrical hypotrich with aslightly flattened ventral surface. The AZM borders the anterior and left edges of theperistome forming a sigmoid shape. Nine of the 12 frontoventral cirri are in 2 oblique rowssituated near the anterior right of the peristome while the other 3 are scattered along the rightbody edge. There are 6 large transverse cirri and a single small left caudal cirrus. Themacronucleus is in 2 ribbon-like parts with 4 micronuclei. NOTES. The position of the reorganisation band in Ito's (1958) diagram indicates that it couldhave been at some stage of division. Originally found in seaweed in the Inland Sea of Japan. Euplotidium arenarium Magagnini & Nobili, 1964DESCRIPTION (Fig. 42). This is a medium sized (71-120 um long) oval species with a wide 236 C. R. CURDS & I. C. H. WU Fig. 41 Euplotidium itoi, after Ito, 1958. triangular peristome. The AZM consists of about 75 membranelles and it extends around theanterior semi-circular rim of the peristome down to the left side. There are 10 fron to ventral s,5 transversals and a single left caudal cirrus. The transverse cirri are not conspicuous as inthe other species of the genus and could be overlooked. The macronucleus is moniliform, in5-10 pieces, and is curved towards the right. The silver-line system consists of 2 marginalkinetics confluent posteriorly and there are dorsal and ventral mesh-like argyromes. NOTE. Originally isolated from sand in the Gulf of Naples. Euplotidium psammophilus (Vacelet, 1961)Borror, 1972Euplotes psammophilus Vacelet, 1 96 1 DESCRIPTION (Fig. 43). This is a large (125 urn long) species in which the wide peristomeextends about two-thirds down the length of the body. There are 7 frontoventral cirriarranged in 2 groups on the right of the peristome. Three are closely packed together at theanterior and 4 are in a row behind them. The 5 transverse cirri are long and there are 2caudal cirri. Vacelet (1961) also described the presence of a row of short cilia along theposterior right body edge. The elongated curved macronucleus is rod-like with an adjacentmicronucleus. NOTES. This was originally described as a species of Euplotes', Borror (1972) transferred it toEuplotidium on account of the shape of the body and peristome and because of the reducednumber of caudal cirri. Euplotidium helgae Hart wig, 1980DESCRIPTION (Fig. 44). This is the largest (up to 200 urn long) of the species. The peristome Fig. 42 Euplotidium arenarium, after Magagnini & Nobili, 1964; (a) ventral surface of living cell;(b, c) ventral and dorsal surfaces of silver-impregnated specimens; (d) stages in nuclearreorganisation. 238 C. R. CURDS & I. C. H. WU 20iim Fig. 43 Euplotidium psammophilus, after Vacelet, 1961 (called Euplotes psammophilus): (a) ventral surface; (b) nuclear apparatus. 50yu m Fig. 44 Euplotidium helgae, after Hartwig, 1 980. extends about two-thirds down the body length. The original author (Hartwig, 1980) was nottoo sure about the cirral pattern and the interpretation used here is based on the assumptionthat there are 5 transverse and 7 frontoventral cirri rather than 4 transverse and 8frontoventral cirri. There is a single caudal cirrus on the left and the fronto ventral s arearranged in a single group on the right of the peristome. The macronucleus consists of 1 1oval parts arranged in the shape of the letter C. There are several micronuclei. NOTE. Originally found in sand in Bermuda. REVIEW OF EUPLOTIDAE Genus PARAEUPLOTES Wichlerman, 1942 239 Introduction Wichterman (1942) described a hypotrichous ciliate which he found in abundance on thecoral Eunicea crassa in the Tortugas. The species resembled Euplotes in that it had a welldeveloped AZM and a C-shaped macronucleus, but its rather peculiar ciliature bore noresemblance to that of Euplotes. He called the genus Paraeuplotes Wichterman, 1942 andplaced it in a new family, the Paraeuplotidae Wichterman, 1942. Here, following Horror(1972) and Corliss (1977) the genus is provisionally included in the Euplotidae. Diagnosis of Paraeuplotes Discoid marine hypotrich with a well-developed AZM which originates on the dorsal surfaceand curves down three-quarters of the body length on the ventral surface. Undulatingmembrane absent. There are no marginal cirri but there is a small group of short caudal cirri.There is an extensive arc of transverse cirri parallel with the right body edge. Anteriorly,there is an arc of cilia parallel with the apical body edge and an isolated pair of cirri in themidventral position. The macronucleus is C-shaped. Contains numerous zooxanthellae. 20>um Fig. 45 Paraeuplotes tortugensis, after Wichterman, 1942: (a) ventral surface; (b) dorsal surface; (c) nucleus. 240 C. R. CURDS & I. C. H. WU Species description Paraeuplotes tortugensis Wichterman, 1942 DESCRIPTION (Fig. 45). Small to medium discoid species measuring on average 80 um indiameter. The well-developed AZM forms a collar-like structure on the anterior dorsalsurface which extends ventrally down about three-quarters of the body length. There is noundulating membrane present. Anteriorly there is an arc of cilia parallel with the body edgeand a pair of isolated cirri lying in a midventral position. There is an extensive arc oftransverse cirri parallel with the right body edge and a group of 5-6 short caudal cirri slightlyto the right of the posterior body pole. The macronucleus is C-shaped. The cell is packedwith numerous yellow-brown zooxanthellae. NOTE. Originally isolated from the coral Eunice crassa in the Tortugas. Genus SWEDMARKIA Dragesco, 1954 Introduction Dragesco (1954, 1960, 1965) studied and described this genus over a period of several years.He (Dragesco, 1960, 1965) likened the genus to Euplotidium and Gastrocirrhus andconsidered it to be a possible evolutionary link between the Holostichidae Faure-Fremiet,1961 and the Euplotidae. Both Faure-Fremiet (1961) and Corliss (1977) placed Swedmarkiain the family Gastrocirrhidae but it is here provisionally included in the Euplotidae. Diagnosis of Swedmarkia Medium sized (100-1 10 jim long) marine hypotrichs with conspicuous AZM arrangedaround a wide triangular peristome. There are two particularly long membranelles at the leftof the apex of the peristome. A wide undulating membrane lies on the right of the peristome.There are many (54-58) cirri present with the following distribution. Numerousfrontoventrals include a row along the right peristome edge, an apical group and an irregularmidventral row. There are 5 large transverse and rows of right and left marginal cirri that areconfluent posteriorly. The macronucleus is divided into numerous (about 100) portions andthere are 5-9 micronuclei. Species description Swedmarkia arenicola Dragesco, 1954 DESCRIPTION (Fig. 46). This is a medium sized (100-1 10 urn long) marine hypotrich, similarin shape to Euplotes. The body is oval to triangular in shape and there is a large triangularperistome which extends down three-quarters of the body length where it occupies about halfof the ventral surface. The AZM consists of 54-68 membranelles of which two, at theextreme left apex of the peristome, are particularly long. There is a wide undulatingmembrane on the right peristome border. The total of 54-58 cirri are arranged as follows, arow of fron to ventral (peristomial) cirri is spaced evenly along the right peristomial edge. Agroup of 6 frontoventral cirri are situated around the extreme apex of the cell and there is anirregular row of 5 frontoventrals lying in the midventral position. Additionally, a short rowof left marginal cirri is present with an irregularly spaced row of right marginals, of which theanterior half arise from the dorsal surface. The marginals are continuous along the posteriorborder. The macronucleus is divided into numerous (about 100) spherical pieces and thereare 5-9 micronuclei. The three dorsal kinetics bearing double cilia are illustrated in Fig. 46b. REVIEW OF EUPLOTIDAE 241 Fig. 46 Swedmarkia arenicola, after Dragesco, 1965: (a) ventral surface; (b) dorsal surface; (c, d)ventral surface showing variations in form, number and arrangement of certain cirri. 242 C. R. CURDS & I. C. H. WU Genus GRUBERELLA (Gruber, 1884) Corliss, 1960 Stylocoma Gruber, 1 884Introduction Gruber (1884) described the marine hypotrich Stylocoma oviformis Gruber, 1884 whichKahl (1932) redescribed and placed in the family Euplotidae. Soon after, another species,Stylocoma adriatica Kiesselbach, 1936, which lacked transverse cirri, was described. Corliss(1960) pointed out that the generic name was preoccupied by Stylocoma Lioy, 1864 adipteran insect, and proposed that it should be replaced by Gruberella Corliss, 1960. Borror(1972) considered the genus to be of questionable status but Corliss (1977) included it as anincertae sedis in the Sporadotrichina. Here, the genus is provisionally included in theEuplotidae on account of its funnel-like peristome, which is reminiscent of Gastrocirrhus,and the arrangement of cirri which, although reduced, are more like those in the Euplotidaethan in any other family of hypotrichs. Diagnosis of Gruberella Ovoid, marine hypotrichs, rounded in cross-section. There is a centrally placed funnel-shaped peristome bordered by a conspicuous AZM anteriorly and down the left side. Cirrireduced. Caudal cirri present, transverse cirri may or may not be present. Frontoventral andmarginal cirri absent. Macronucleus in two parts. Key to the species of Gruberella 1 With 6 caudal and 7 transverse cirri .With 6 caudal but without transverse cirri G. oviformisG. adriatica Species descriptions Gruberella oviformis (Gruber, 1 884) Corliss, 1960Stylocoma oviformis Gruber, 1 884 DESCRIPTION (Fig. 47). Ovoid marine hypotrich with wide, centrally placed, funnel-shapedperistome which has a slightly raised edge forming an anterior collar-like region. The Fig. 47 Gruberella oviformis: (a) after Gruber, 1 884; (b) after Kisselbach, 1936. REVIEW OF EUPLOTIDAE 243 prominent AZM consisting of many large membranelles originates on the ventral surface onthe right of the peristome and continues around the apex down the left side. There are only 2groups of cirri, 7 transverse and 6 long caudal cirri. No frontoventral or marginal cirri arepresent. Shape of macronucleus unrecorded. Gmberella adriatica (Kisselbach, 1936) Corliss, 1960Stylocoma adriatica Kiesselbach, 1936 DESCRIPTION (Fig. 48). Medium sized (80 um long) triangular shaped marine hypotrich withfunnel-shaped, centrally positioned peristome. Prominent AZM borders the anterior and leftperistomial edges. Cirri reduced to a single group of 6 long caudals. There are no transverse,frontoventral or marginal cirri present. Macronucleus divided into 2 ovoid pieces with amicronucleus between the pair. Fig. 48 Gruberella adriatica, after Kisselbach, 1 936.Genus CYATHAROIDESTuffrau, 1975 Introduction Tuffrau (1975) described a species from Antarctica which closely resembled Euplotes incertain respects but not in others. It is characterised by the presence of a row of 12 large rightmarginal cirri in addition to the usual frontoventral and transverse cirri. Furthermore, thereis an extensive paroral membrane composed of a single kinety of long cilia which is unlikethe undulating membrane of Euplotes and there is also endoral ciliature which is not foundin Euplotes. Diagnosis of Cyatharoides Irregularly oval marine hypotrich with a very large peristomial funnel occupying much ofthe ventral surface onto which it opens. The peristome is surrounded by a collar-likeswelling of the anterior left body edge, with a characteristic 'niche' or invagination of theright anterior peristome region. The AZM is composed of many membranelles lining the leftperistomial border. On the right, at the bottom of the peristomial funnel there is a paroralmembrane composed of a single kinety of long cilia and an arc of endoral cilia. There are 10frontoventral, 5 transverse and a row of 1 2 right marginal cirri along the right body edge.Dorsally there are 12-26 kinetics of cilia and the macronucleus is C-shaped. 244 C. R. CURDS & I. C. H. WU Species description Cyatharoides balechi Tuffrau, 1 975 DESCRIPTION (Fig. 49). Irregularly oval, large (140-200 urn long) marine, planktonichypotrich with a very large peristomial funnel opening over much of the ventral cell surface.Anteriorly the peristome is surrounded by a collar-like swollen ridge of the anterior left bodyedge, with a characteristic invagination of the peristome on the right anterior edge. TheAZM is prominent and composed of many large membranelles. There are paroral andendoral membranes on the posterior right peristomial edge. With 10 frontoventral, 5transverse and 12 right marginal cirri. Dorsal surface with 12-26 kinetics of many cilia.Macronucleus large, open C-shaped. NOTE. Originally isolated from plankton sample taken from Antarctic Ocean near base ofArgentina by Prof. Balech. Mr Fig. 49 Cyatharoides balechi, ventral surface, after Tuffrau, 1975. References Agamaliev, F. G. 1967. Faune des cilies mesopsammiques de la cote ouest de la Mer Caspienne. Cah. Biol. mar. 8 : 359^02.1 968. Materials on morphology of some psammophilic ciliates of the Caspian Sea. Acta Protozool. 6 (20) : 225-244. 1971 . Complements to the fauna of psammophilic ciliates of the western coast of the Caspian Sea. Acta Protozool. 8 (30) : 379-404.Andrusova, J. 1886. Uber die Infusoria der Bucht von Kertsch. Trudy imp. S.-Peterb. Obshch. Estest.17 : 263-259. REVIEW OF EUPLOTIDAE 245 Borror, A. C. 1963. Morphology and ecology of the benthic ciliated protozoa of Alligator Harbor, Florida. Arch. Protistenk. 106 : 465-534.1965a. Morphological comparison of Diophrys scutum (Dujardin, 1841) and Diophrys peloetes n.sp. (Hypotrichida, Ciliophora). J. Protozool. 12 : 60-65. \965b. New and little-known tidal marsh ciliates. Trans. Am. microsc. Soc. 84 : 550-565. 1972. Revision of the order Hypotrichida (Ciliophora, Protozoa). J. Protozool. 19 : 1-23. Buddenbrock, W. von 1920. Beobachtungen iiber einige neue oder wenig bekannte marine Infusorien. Arch. Protistenk. 41 : 341-364. Bullington, W. E. 1940. Some ciliates from Tortugas. Pap. Tortugas Lab. 32 : 179-222.Biitschli, O. 1887-1889. Protozoa. Abt III. Infusoria und System der Radiolaria. In Bronn, H. G. (Ed.), Klassen und Ordnungdes Thiers-Reichs Vol. 1 , pp. 1098-2035. Leipzig, C. F. Winter.Calkins, G. N. 1902. Marine protozoa from Woods Hole. Bull. U.S. Fish Comm. 21 (yr 1901): 41 3-468. 1911. Regeneration and cell division in Uronychia. J. exp. Zool. 10 : 95-1 16. Chatton, E. & Lwoff, A. 1930. Impregnation par diffusion argentique de 1'infraciliature des cilies marins et d'eau douce apres fixation cytologique et sans dessication. C. r. Seanc. Soc. Biol. 104 : 834-836.Chatton, E. & Seguela, J. 1936. Un hypotriche de la branchie de dona intestinalis L., intermediaire entre les Euplotidae et les Aspidiscidae: Euplotaspis cionaecola n.g., n.sp. Bull. Soc. zool. Fr. 61 : 332-340.Claparede, E. & Lachmann, J. 1858 (1857). Etudes sur les infusoires et les rhizopodes. Mem. Inst. natn. genev. 5 : 1-260.Corliss, J. O. 1960. The problem of homonyms among generic names of ciliated protozoa, with proposal of several new names. J. Protozool. 1 (3) : 269-278.1961. The ciliated protozoa: characterization, classification and guide to the literature. 310 pp. Pergamon Press, New York.1977. Annotated assignment of families and genera to the orders and classes currently comprising the Corlissian scheme of higher classification for the phylum Ciliophora. Trans. Am. microsc. Soc. 96(1): 104-140. 1979. The ciliated protozoa: characterization, classification and guide to the literature. 455 pp. Pergamon Press, Oxford.Curds, C. R. 1975. A guide to the species of the genus Euplotes (Hypotrichida, Ciliatea). Bull. Br. Mus. nat. Hist. (Zool.) 28 (1): 1-61.Dragesco, J. 1954. Diagnoses preliminaries de quelques cilies nouveaux des sables. Bull. Soc. zool. Fr. 79 : 62-70. 1960. Cilies mesopsammiques littoraux. Systematique, morphologie, ecologie. Trav. Stn biol. RoscoffU : 1-356. 1963. Complements a la connaissance des cilies mesopsammiques de RoscofF. II. Heterotriches III. Hypotriches. Cah. Biol. mar. 4 : 251-275. 1965. Complements a la connaissance de Swedmarkia arenicola et Discocephalus ehrenbergi cilies hypotriches. Anneebiol. 4 : 187-204. 1968. Metopus jankowskii n.sp., Sonderia sinuata Kahl et Discocephalus minimus n.sp. cilies nouveaux ou mal connus. Annls Fac. Sci. Univ. fed. Cameroun 1 : 77-Dujardin, F. 1841. Histoire Naturelle des Zoophytes Infusoires. 678 pp. Paris.Ehrenberg, C. G. 1830 (1832). Beitrage zur Kenntriis der Organisation der Infusorien und ihrer geographischen Verbreitung, besonders in Siberien. Abh. Akad. Wiss. DDR year 1832 : 1-88.1831 (1832). Uber die Entwicklung und Lebendauer der fnfusionsthiere; nebst ferneren Beitragen zu einer Vergleichung ihrer organischen Systeme. Abh. Akad. Wiss. DDR year 1832 : 1-1 54.1838. Die Infusionsthierchen als Vollkommene Organismen.6\2 pp. Leipzig. Fabre-Domergue, P. L. 1885. Note sur les infusoires cilies de la baie de Concarneau. /. Anal. Physioi, Paris 21 -.554-56%. Faure-Fremiet, E. 1954. Gastrocirrhus adhaerens n.sp. Anais Acad. bras. Cienc. 26 : 163-168.1961. Remarques sur la morphologie comparee et la sytematique des ciliata Hypotrichida. C.r. hebd. Seanc. Acad. Sci., Paris 252 : 35 1 5-3519. Fenchel, T. 1965. Ciliates from Scandinavian molluscs. Ophelia 2:71-1 74.Fresenius, G. 1865. Die Infusorien des Seewasseraquariums. Zool. Gart., Frankf. 6 (3): 8 1-89. 6(4): 121-129.Gruber, A. 1884. Die Protozoen des Hafens von Genua. Nova Acta Acad. Caesar. Leop. Carol. 46 : 473-539. 246 C. R. CURDS & I.C. H. WU Hartwig, E. 1973. Die Ciliaten des Gezeiten-Sandstrandes der Nordseeinsel Sylt. 1. Systematik. Mikrofauna Meeresbodens 18 : 1-69.1974. Verzeichnis der im Bereich der deutschen Meereskiiste angetroffenen interstitiellen ciliaten. Mitt. hamb. zool. Mus. Inst. 71 : 7-21. The interstitial ciliates of Bermuda with notes on their geographical distribution and habitat. Cah. Biol.mar.2l : 409-441.Hartwig, E. & Parker, J. G. 1977. On the systematics and ecology of interstitial ciliates of sandy beaches in North Yorkshire. J. mar. biol. Ass. U.K. 57 : 735-760.Hemprich, F. G. & Ehrenberg, C. G. 1828 (1831). Animalia Evertebrata exclusis Insectis. In Symbolae Physicae seu Iconis et Descriptiones Animalium Evertebratorum Sepositis Insectis. 126 pp. Berolini ex officina Academica. (date of plates 1 828).1831. Animalia Evertebrata exclusis Insectis. In Symbolae Physicae seu Iconis et Descriptiones Animalium Evertebratorum Sepositis Insectis. 126 pp. Berolini ex officina Academica. (date of text 1831).Ito, S. 1958. Two new species of marine ciliates Euplotidium itoi sp.nov., and Gastrocirrhus trichocystus sp.nov. Zool. Mag. Tokyo 61 : 184-187.1963. Cytological observations of nuclear behavior in the conjugation of Diophrys scutum. Zool. Mag. Tokyo 72: 230-234.Kahl, A. 1932. Urtiere oder Protozoa. I. Wimpertiere oder Ciliata (Infusoria), eine Bearbeitung der freilebenden und ectocommensalen Infusorien der Erde, unter Anschluss der marinen Tintinnidae. 3. Spirotricha. In Dahl, F. (Ed.), Die Tierwelt Deutschlands, pt. 25 : 399-650. G. Fischer, Jena.Kattar, M. R. 1970. Estudo dos protozoarios ciliados psamofilos do litoral Brasileiro. Bolm Fac. Filos. Cienc. Univ. S. Paulo Zool. Biol. mar. NS21 : 123-206. Kent, W. S. 1880-1882. A Manual of the Infusoria. Vols 1-3: 913 pp. David Bogue, London.Kisselbach, A. 1936. Zur Ciliatenfauna der Nordlichen Adria. Thalassia2(5): 1-53.Lepsi, J. 1928. Un nouveau protozoaire marin; Gastrocirrhus intermedium. Annls Protist. 1 : 195-197.Lioy, P. 1864. I ditteri distribuiti secundo un nuovo metodo di classificazione naturale. Atti 1st. veneto Sci. Ser. 3.9(2) : 989-1 126.Magagnini, G. & Nobili, R. 1964. Su Euplotes woodruffi Gaw e su Euplotidium arenarium n.sp. (Ciliatea, Hypotrichida). Monitore zool. ital. 12 : 1 78-202. Mansfeld, K. 1923. 16 neue oder wenig bekannte marine Infusorien. Arch. Protistenk. 46 : 97-140.Miiller, O. F. 1 786. Animalcula Infusoria Fluviatilia et Marine. 367 pp. Havniae et Lipsiae.Noland, L. E. 1937. Observations on marine ciliates of the gulf coast of Florida. Trans. Am. microsc. Soc.56: 160-171.Ozaki, Y. & Yagui, R. 1942. A new marine ciliate Cirrhogaster monilifer n.g. n.sp. Annotns zool. jap. 21:79-81. Pierantoni, N. 1909. Su alcuni Euplotidae del Golfo di Napoli. Boll. Soc. Nat. Napoli 1909 : 53-64.Quennerstedt, A. 1867. Bidragtill a Sveriges Infusorie-fauna. Ada Univ. lund. 4 : 1-47.Raikov, I. B. & Kovaleva, V. G. 1968. Complements to the fauna of psammobiotic ciliates of the Japan Sea (Posjet Gulf). Ada Protozool. 6 (27) : 309-333.Rees, J. van 1881. Zur Kenntnis der Bewimperung der Hypotrichen Ifusorien, nach Beobachtungen an Styloplotes grandis n.sp. und Euplotes longipes Claparede & Lachmann. Ellerman & Harms, Amsterdam. 1883-1884. Protozoaires de 1'escautde Test. Tijdschr. ned. dierk. Vereen. Suppl. 1 : 592-673. Reiff, I. 1968. Die genetische Determination mutlipler Paarungstypen bei dem Ciliaten Uronychia transfuga (Hypotricha, Euplotidae). Arch. Protistenk. 110 : 372-397. Ruinen, J. 1938. Notizen uber Ciliaten aus konzentrierten Salzgewassern. Zool. Meded. 20 : 243-256.Sauerbrey, E. 1928. Beobachtungen uber einige neue oder wenig bekannte marine Ciliaten. Arch. Protistenk. 62 : 355-407.Stein, F. 18590. Der Organismus der Infusionsthiere nach eigenen Forsuchungen in Systematischer Reihenfolge Bearbeitet I. 206 pp. Leipzig. 185%. Charakteristik neuer Infusoriengattungen. Lotos 9 : 2-5. 9 : 57-60. Summers, F. M. 1935. The division and reorganization of the macronuclei of Aspidisca lynceus Miiller, Diophrys appendiculata Stein, and Stylonychia pustulata Ehrbg. Arch. Protistenk. 85: 173-208.Taylor, C. V. 1928. Protoplasmic reorganization of Uronychia uncinata sp.nov. during binary fission and regeneration. Physiol. Zool. 1(1): 1-25.Tuffrau, M. 1975 (1974). Un nouvel Euplotidae: Cyatharoides balechi, n.g., n.sp., cilie Hypotriche de 1'Anarctique. Prostistologica 10 (3) : 31 1-317. REVIEW OF EUPLOTIDAE 247 Vacelet, E. 1960. Note preliminaire sur la faune infusorienne des 'Sables a Amphioxus' de la biae de Marseille. Reel Trav. Stn mar. Endoume33 (9) : 53-57.1961. La faune infusorienne des 'Sables a Amphioxus' des environs de Marseille. Bull. Inst. oceanogr. Monaco 3 (\202) : 1-12.Wallengren, J. 1901 (1900). Zur Kenntnis der vergleichenden Morphologic der hypotrichen Infusorien. Bih. K. svenska VetenskAkad. Handi 26 (4) : 1-31.Wichterman, R. 1942. A new ciliate from coral of Tortugas and its symbiotic Zooanthellae. Pap. Tortugas Lab. 33 : 107-111.Wu, I. C. H. & Curds, C. R. 1979. A guide to the species ofAspidisca. Bull. Br. Mus. nat. Hist. (Zool.) 36(1): 1-34.Young, D. B. 1922. A contribution to the morphology and physiology of the genus Uronvchia. J. exp. Zool. 36 : 353-395. Manuscript received for publication 4 August 1982 Index to Species (Names given in roman refer to synonyms) Campylopus paradoxus 220Certesia ovata 228 quadrinucleata 228Cirrhogaster monilifer 23 1Cyatharoides balechi 244 Discocephalus ehrenbergi 195 grandis 195 minimus 197 rotator ius 194 rotatorius 195Diophrys appendiculata 203 appendiculatus 203 grandis 206 hystrix 207 irmgard2Ql kahli 209 kasymovi 206 magnus 206 marina 203 multinucleata2\5 oligothrix 2 1 peloetes 2 1 quadricaudatus 2 1 1 salina 209 scutoides 2 1 2 scutum 203 tetramacronucleata 2 1 4 Euplotaspis cionaecola 233Euplotes psammophilus 236Euplotidium agitatum 234 arenarium 235 Helgae236 /to/ 23 5 psammophilus 236 Gastrocirrhus adhaerens 230 intermedius 229 monilifer 231 stentoreus 229 trichocystus23\Gruberella adriatica 243 oviformis 242 Paraeuplotes tortugensis 240Planiplotes wagneri 203Ploesconia scutum 203, 222Polycoccon octangulus 194Schizopus norwegicus 203Stylocoma adratica 243 oviformis 242 Stylonychia appendiculata 203Styloplotes appendiculatus 203 appendiculatus var pontica 203 fresenii 203 grandis 206 norwegicus 206 quennerstedti 206 Swedmarkia arenicola 240Trichoda transfuga 220 Uronvchia binucleata 226bivalvorum 220heinrothi 220magna 226setigera 225transfuga 220uncinata220 British Museum (Natural History) An Atlas of Freshwater Testate Amoebae C. G. Ogden & R. H. Hedley 1980, Hardcovers, 222pp, 17.50 (18.00 by post). Co-published by British Museum(Natural History) and Oxford University Press. This book illustrates, using scanning electron micrographs, most of the commonspecies of testate amoebae that are found in freshwater habitats. Information onthe biology, ecology, geographical distribution and a classification are followed bydescriptions of ninety-five species. Each of these is illustrated by several views ofthe shell. The text is designed not only to enable biologists to identify species of testateamoebae, but to serve as an introduction to students interested in the taxonomyand biology of these freshwater protozoa. It will be of special interest toprotozoologists, ecologists, limnologists, water treatment specialists andmicropalaeontologists interested in recent sediments. British Museum (Natural History)Publication Sales,Cromwell Road,London SW7 5BD. Titles to be published in Volume 44 Observations on the systematics of the genus Difflugia inBritain (Rhizopoda, Protozoa). By Colin C. Ogden Miscellanea A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin Curds & Irene C. H. Wu Osteology, genitalia and relationships of the Acanthodactylus(Reptilia: Lacertidae). By E. N. Arnold The Opthalmotilapia assemblage of cichlid fishes reconsidered. By Peter Humphrey Greenwood Morphological studies on some Difflugiidae from Yugoslavia(Rhizopoda, Protozoa). By Colin G. Ogden & Andjelija Zivkovic Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) The Ophthalmotilapia assemblage oftichlid fishes reconsidered Peter Humphry Greenwood Zoology series Vol 44 No 4 28 April 1983 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, andan 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 SW7 5BD,England. World List abbreviation: Bull. Br. Mus. not. Hist. (Zool.) <C) Trustees of the British Museum (Natural History), 1983 The Zoology Series is edited in the Museum's Department of ZoologyKeeper of Zoology : Dr J. G. ShealsEditor of Bulletin : Dr C. R. CurdsAssistant Editor : Mr C. G. Ogden ISSN 0007-1498 Zoology series Vol 44 No 4 pp 249-290British Museum (Natural History)Cromwell RoadLondon SW7 5BD Issued 28 April 1983 The Ophthalmotilapia assemblage of cichlid fishesreconsidered Peter Humphry Greenwood^ Department of Zoology, British Museum (Natural History), Cromwell RoaSW7 5BD Contents Synopsis 249 Introduction 249 Material examined 250 A review of group characters for the Ophthalmotilapia assemblage . . . 254 The Ophthalmotilapia assemblage reconsidered 262 Relationships within the Ophthalmotilapia assemblage .... 263The nature of the apophysis for the upper pharyngeal bones in the Ophthalmotilapia assemblage 274 A review of other schemes of relationship suggested for members of the Ophthalmotilapia assemblage 275 Summary of the taxonomic conclusions and a discussion of the sister-group problem in these and other lake cichlids 278 The status of Cyathopharynx Regan, 1 920 and Cardiopharynx Poll, 1 942 . 282 Cyathopharynx Regan, 1920 284 Description 284 Acknowledgements 289 References 290 Synopsis The Ophthalmotilapia assemblage of cichlid fishes from Lake Tanganyika, first defined by Liem(1981), originally comprised the genera Asprotilapia, Ectodus, Lestradea, Cunningtonia andOphthalmotilapia (with which was synonymized Ophthalmochromis). The characters on which theassemblage was based are reviewed and revised in the light of information derived from large-scaleoutgroup comparisons. As a result of this review, five additional genera can be included in the assemblage (viz.Aulonocranus, Grammatotria, Callochromis, Xenotilapia and Cyathopharynx (with which Cardio-pharynx is synonymized). Two lineages within the Ophthalmotilapia assemblage are defined. Previous schemes of supposed relationships for these taxa are discussed, as is the problem ofidentifying the sister-group for the assemblage. No sister-group can be identified amongst the endemictaxa of Lake Tanganyika, but the possibility of an endemic taxon from Lake Malawi being the sister-group is explored. Problems arising from possible homoplasy and thus the misidentification of sister-groups areillustrated by examples involving cichlid species from Lakes Tanganyika and Malawi, and from theselakes and Lake Fwa (Zaire drainage basin). Introduction During the last thirty years there has been a marked increase in our knowledge of taxonomicand ecological diversity amongst African cichlid fishes. Unfortunately there has been lessprogress made in our understanding of phyletic relationships between the various elementsof those faunas, or even amongst members of the so-called species flocks of the major Africanlakes (see Greenwood, 1980). Bull. Br. Mm. nat. Hist. (Zool.) 44 (4): 249-290 Issued 28 April 1 983 250 P- H. GREENWOOD Two recent papers are welcome and important contributions to the field of phyleticstudies since both are concerned with interrelationships in a single species flock, that of LakeTanganyika (Liem & Stewart, 1976;Liem, 1981). Liem's (1981) paper is directed at establishing the monophyletic origin of five endemicgenera from that lake, viz. Asprotilapia, Ectodus, Lestradea, Cunningtonia and Ophthalmo-tilapia. Although Liem was able to argue a case for recognizing the monophyly ofthese taxa (the Ophthalmotilapia assemblage), and the intragroup relationships of itsconstituent genera, he was unable to recognize a sister-group for the whole assemblage. Hedid, however, suggest that among the endemic Tanganyika genera, Aulonocranus,Xenotilapia, Callochromis and Cardiopharynx share some of the derived featurescharacterizing the Ophthalmotilapia assemblage (Liem, 1981 : 206; 208). My interest in these species, and the Ophthalmotilapia lineage, stems from my currentresearch into the levels of relationship existing between the endemic cichlid genera of LakesVictoria, Malawi and Tanganyika (Greenwood, 1979; 1980). Using information acquired inthis search, it seemed that Liem's concept of the Ophthalmotilapia assemblage could beextended to include several other Tanganyika taxa, including those suspected of suchrelationship by Liem. It was also apparent that the group characters for the assemblageshould be reviewed in the light of more extensive outgroup comparisons than were employedoriginally. Finally, my interest was aroused by what seemed to be the unusually clear light that somemembers of the Ophthalmotilapia assemblage could throw on the longstanding and oftenintractable problem of homoplasy and its effects on hypotheses of relationship amongstAfrican cichlid fishes. Like Liem, I have been unable to identify a sister-group for the Ophthalmotilapiaassemblage from amongst the Tanganyika cichlids. A very tentative suggestion can be made,however, for a possible sister-group relationship between the assemblage and certainmembers of the Lake Malawi flock. Material examined Dissections were made of the dorsal gill-arch musculature in one, or usually 2, specimens ofeach genus now included in the Ophthalmotilapia assemblage (see p. 278). The specimen ofAsprotilapia leptura (BMNH 1906.9.6: 157) was that used by Liem (1981) but the jawmuscles of the left side were freshly dissected. The nature of the gut and its coiling pattern were checked in several specimens of eachOphthalmotilapia assemblage species (except Asprotilapia leptura where only the type andthe specimen noted above could be used). Jaw and dorsal gill-arch muscles were dissected in unregistered specimens of Astatotilapiaelegans and in A. burtoni. All available dry skeletal material in the BM(NH) collections was examined, in particularthat prepared for the revisions of the Lake Victoria, Edward-George, Kivu, and Turkanahaplochromine species (see Greenwood, 1980). Additional material, prepared for thispaper, and alizarin transparencies not previously listed, are given below. The taxa arefirst grouped geographically, and then alphabetically, within the categories: Dry skeleton(DS) and Alizarin transparency (AT). Lake Tanganyika DS: A sprotilap ia leptu ra 671 Aulonocranus dewindti 1960.9.30 : 4629-641 Callochromis macrops 1906.9.8: 178 Cardiopharynx schoutedeni 1950.4.1. : 1854-81; 1960.9.30 : 1647-56 Cyathopharynxfurcifer 1900.9.8 : 25 1 ; 148; 1950.4. 1 : 1 7 14^7 THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 251 Cunningtonia longiventralisEctodus descampsiEretmodus cyanostictusGrammatotria lemaireiLestradea perspicax s tappers iLimnochromis auritusLimnochromis leptosomaLobochilotes labiatusNeotilapia tanganicaeOphthalmotilapia boopsOphthalmotilapia ventralisPerissodus microlepisPerissodus paradoxusPetrochromisfamulaPetrochromisfasciolatusPetrochromis polyodonSimochromis babaultiSimochromis curvifronsSimochromis dardenneiSimochromis diagrammaSimochromis loockiTrematocara marginatumTrematocara unimaculatumTropheus mooreiXenotilapia boulengeri AT: Astatotilapia burtoniAulonocranus dewindtiCallochromis macrops macropsCallochromis macrops melanostigmaCardiopharynx schoutedeniCtenochromis horeiCunningtonia longiventralisCyathopharynx furciferEctodus descampsiEretmodus cyanostictusGrammatotria lemaireiHemibates stenosomaLamprologus brevisLamprologus cunningtoniLamprologus elongatusLamprologus fasciatusLamprologus tetracanthusLamprologus werneriLestradea perspicaxLimnochromis abeeleiLimnochromis auritusLimnochromis dhanisiLimnochromis otostigmaLimnochromis pfefferiLimnochromis permaxillarisOphthalmotilapia boopsOphthalmotilapia ventralisPerissodus hecquiPerissodus microlepisPerissodus paradoxusPetrochromisfasciolatusPetrochromis polyodon 3066-67 1900.12.13:351960.9.30: 1720-1724 1906.9.8 : 144Uncatalogued1906.9.8. : 2671950.4.1 : 7642-521960.9.30: 1368-71; 1372-75 1 898.9.9 : 66; 1 950.4. 1 : 7608-6 1 1 ; uncataloguedUncatalogued 1955.4.12 : 47-66; uncatalogued 1906.9.8 : 244; uncatalogued (4 specimens) 1906.9.8: 217; uncatalogued 1950.4.1 : 7674-7701; 7702-7728 1960.9.30:4835^850 1906.9.6: 119 1950.4.1 : 907-67; uncatalogued 1961.11.22:290-293 1960.9.30:2415-33 1950.4.1 : 4695-775; 4642-656 1960.9.20:2821-2823 1960.9.30:2845-2859 1960.9.30: 1574-1615 1960.9.30:2526-28 1960.9.30: 1896-1907 1960.9.30: 1792-1802 1961.11.22: 113-119 1950.4.1 : 5171-5182 1960.9.30 : 3317-30; 3276-79; 3208-21 1961.11.22:976-989 1960.9.30:7181-7225 1950.4.1 : 6701-6707 1960.9.30:6851-6860 1950.4.1 :7075 1960.9.30:7309-7336 1976.5.21 : 46-65 1960.9.30: 1468-84 1961.11.22:56-58 1960.9.30: 1981-85 1960.9.30:2001-04 1960.9.30: 1989-97 1960.9.30: 1923-26 1961.11.22:41-46 1960.9.30: 1716-18; 1720-24 1960.9.30: 1689-94 1960.9.30:6364-69 1960.9.30:6386-91 1960.9.30:6468-6482 1960.9.30: 1392-93; 1394-96; 1390 1960.9.30: 1359-63 252 Simochromis curvifronsSimochromis dardenneiSimochromis diagrammaSimochromis loockiTelmatochromis temporalisTrematocara capartiTrematocara kufferathiTrematocara marginatumTrematocara nigrifronsTrematocara stigmaticumTrematocara unimaculatumTropheus mooreiXenotilapia boulengeriXenotilapia melanogenysXenotilapia ochrogenysXenotilapia simaXenotilapia tenuidentata Lake MalawiDS: Astatotilapia callipteraAulonocara nyassaeAulonocara rostrataChilotilapia rhoadesiiCorematodus shiranusCorematodus taeniatusCyathochromis obliquidensCynotilapia afraDocimodusjohnstoniGenyochromis mentoGephyrochromis lawsi'Haplochromis ' ahli'Haplochromis ' annectens'Haplochromis ' argyrosoma'Haplochromis ' atritaeniatus'Haplochromis ' auromarginatus'Haplochromis ' breviceps'Haplochromis ' caeruleus'Haplochromis ' chrysonotus'Haplochromis ' dimidiatus'Haplochromis ' ericotaenia'Haplochromis ' euchilus'Haplochromis ' eucinostomus'Haplochromis ' fuscotaeniatus'Haplochromis' guentheri'Haplochromis ' intermedium'Haplochromis ' johnstoni'Haplochromis ' kirkii'Haplochromis ' kiwinge'Haplochromis ' labifer'Haplochromis ' labridens'Haplochromis ' lateristriga'Haplochromis ' lepturus'Haplochromis ' longimanus'Haplochromis ' macrostoma'Haplochromis ' marginatus'Haplochromis ' moorii'Haplochromis ' nototaenia'Haplochromis ' ornatus P. H. GREENWOOD 1961.11.22:2-8 1960.9.30: 1061 1960.9.30: 1170-1174; 1188-91 1950.4.1 : 7702-28 1960.9.30:6531-6538 1961.11.22:703-721 1961.11.22:883-910 1960.9.30:4881-90 1961.1 1.22 : 693-703; 1960.9.30 : 4990-5019 1960.9.30:5143-171 1961.11.22:519-528 1961.11.22: 13 1961.11.22:225 1950.4.1 : 4035-40; 1960.9.30 : 3449-455 1960.9.30:3820-855 1961.11.22:208-211 1960.9.30:7998-8014 1893.11.15:4 1935.6.14: 2259-63; 681. 5A 681. 5A 681.2; 1935.6.14:2103-211 68 1.4 A 681.4B;681.4C 1935.6.14 : 282-295; 681. 12A : 681. 12B 1893.1.17: 8 (syntype) 681.3 1965.10.26:24-29 1965.11.2: 14-22(paratypes) 1935.6.14: 1469-71 1935.6.14:847-52 1935.6.14: 1657-61 1935.6.14: 1426-9 1935.6.14: 1476-78 1935.6.14:870-72 1935.6.14: 1267-69 1935.6.14: 1823-32 1935.6.14: 1154-71 1935.6.14:2405-2411 1972.9.13:70 1962.10.18: 1-10 1935.6.14:494-95 1921.9.6: 154-62 1972.9.13:91-94 1935.6.14:523-32 1935.6.14:953-62 1935.6.14: 1031^0 1972.9.13:77-81 1935.6.14:991-1000; 1001-05 1935.6.14: 1209-18 1935.6.14: 1340-56 1972.12.20:35-78 1935.6.14:605-7 1935.6.14:769-77 1935.6.14: 1692-1700 1935.6.14: 1378-84 1972.12.18:31-33 THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 253 'Haplochromis ' orthognathus'Haplochromis ' placodon'Haplochromis ' pleurotaenia'Haplochromis ' polystigma'Haplochromis ' quadrimaculatus'Haplochromis ' prostoma'Haplochromis ' rhoadesii'Haplochromis ' rostratus'Haplochromis ' semipalatus'Haplochromis ' spilonotus'Haplochromis ' spilorhynchus'Haplochromis ' subocularis'Haplochromis ' tetrastigmaHemitilapia oxyrhynchusLabeotropheus fuelleborniLabidochromis vellicansLabidochromis zebroidesLethrinops auritusLethrinops lethrinusLethrinops longimanusLethrinops parvidensLethrinops praeorbitalisMelanochromis melanopterusMelanochromis vermivorousPetrotilapia tridentigerPseudotropheus fuscusPseudotropheus livingstoniPseudotropheus lucernaPseudotropheus macrophthalmusPseudotropheus tropheopsPseudotropheus williamsiPseudotropheus zebraTrematocranus microstomaAT: Astatotilapia callipteraAulonocara nyassaeLethrinops lethrinus ElsewhereDS: Astatotilapia bloyetiAstatotilapia nubilaChromidotilapia batesiiChromidotilapia kingsleyaeOreochromis niloticusTilapia rendalliTilapia zilliAT: Astatotilapia bloyetiAstatotilapia macropsoidesOreochromis niloticus Radiographs used in this study were:Lake Tanganyika Asprotilapia lepturaAulonocranus dewindtiCallochromis macrops macrops 1973.3.26: 189; 1969.3.11 : 19 1935.6.14: 1736^15 1935.6.14:911-16 1935.6.14:433-37 1935.6.14: 1960-69 1962.10.18 : 59-71; 1979.1 1.26 : 5-9 681T 1935.6.14:549-54 1956.6.12:9-10 1969.3.11 : 11-14 1935.6.14: 1260-^ 1935.6.14: 1180-89 1935.6.14: 1556-66; 1567-77 1906.9.7 : 39 1972.8.11 : 3-5; 681. 11 1965.10.26: 14-21 1981.1.9: 102-1 06 (paratype) 1930.1.31 : 84-86 1930.1.31 : 150-4; 1906.9.7:35 1969.3.11 : 20-23 1935.6.14:2070-73 696J 1935.6.14 : 303 (syntype); 1971.9.13 : 32-36 1935.6.14: 307-16 (syntype) 1981.2.2: 212-20; 681.9 1965.10.25: 131^1 1935.6.14: 128-30 1935.6.14: 165-9 (syntype) Uncatalogued 1965.10.25: 1 15-24; 681. 8C 1965. 10.25: 20-26; 68 1.8B 1935.6.14: 135-40; 681.8A 1935.6.14:2232-6 1966.7.26: 17-421935.6.14:2259-631930.1.31 : 109-118 1961.12.1 : 341-353 1911.3.3: 148 1912.6.29:4 1912.4.1 :526; 1908.5.30: 186 662 D; Uncatalogued 1906.9.7:32 1907.12.2:3767 UncataloguedUncatalogued1907.12.2:3533-534 1906.9.6: 156-157 (FW 677)1950.4.1 : 4843-93 (FW678)1950.4.1 : 3258-3 12 (FW683) 254 Callochromis macrops melanostigmaCallochromis pleurospilusCardiopharynx schoutedeniCunningtonia longiventralisCyathopharynxfurciferEctodus descampsiGrammatotria lemaireiLes traded perspicaxOphthalmotilapia hoopsOphthalmotilapia ventralisXenotilapia boulengeriXenotilapia melanogenysXenotilapia ochrogenysXenotilapia ornatipinnisXenotilapia sima Lake Fwa Callopharynx microdonCyclopharynxjwaeNeopharynx schwetzi P. H. GREENWOOD1960.9.30 : 2882-96 (FW683) 1940.4.1 1960.9.30: 1574-1615 (FW674) 1950.4.1950.4. 3458-662; 1920.5.25 : 152-53 (FW682) 1 282-86 (FW676) 1605-37; 1906.9.8 : 249-251 (FW673) 1 96 1.1 1.22: 76-87 (FW677)1950.4.1 3703-3727 (FW684)1950.4.1 153 1^16 (FW676)1960.9.30: 1 720-24 (FW675)1950.4.1 1291-1465 (FW675)196 1.1 1.22: 290-93 (FW679)1960.9.30: 3419-48 (FW681)1950.4.1 : 4047-1 36 (FW679)1960.9.30 : 3685-728 (FW681)1961.11.22: 1 90-202 (FW680) MRAC71300(FW671)MRAC71301(FW671)MRAC 71290, 76201 (FW671); 71291-99 (FW672) A review of group characters for the Ophthalmotilapia assemblage As originally defined by Liem (1981), the Ophthalmotilapia assemblage 1 comprised thefollowing genera: Asprotilapia Blgr (1901), Ectodus Blgr (1898), Lestradea Poll (1943),Cunningtonia Blgr (1906) and Ophthalmotilapia Pellegrin (1904). Ophthalmochromis Poll(1956) was shown by Liem (1981 : 210) to be a synonym of Ophthalmotilapia, a decisionwith which I fully agree. Liem's concept of the OA, and his grounds for considering it a monophyletic group, werebased on eight uniquely congruent apomorphic characters present in all the constituent taxa(Liem, 1981 : 207-208). With one possible exception, however, Liem believed that none ofthese characters is an autapomorphy for the group (Liem's character 2, the outline shape ofthe palatine bone, is the exception). The apomorphic status of the eight group-characters was justified by Liem (198 1 : 205) onthe grounds of their being derived relative to the character state found in various generalizedtaxa (such as Astatotilapia burtoni and A. elegans) and in other taxa from Lake Tanganyika. I have been able to extend Liem's outgroup comparisons to include the cichlid genera ofLakes Malawi, Victoria and Edward, taxa from various river systems and their associatedsmall lakes, and additional genera from Lake Tanganyika itself. Wherever possible, severalspecies of a genus were examined. The review of the eight OA group-characters which follows takes this extra material intoaccount. It should be noted, however, that the new members of the 0A (see p. 262) are, forthe purposes of the review, not treated as elements of the assemblage. Thus, unless otherwiseindicated, all references to the OA in this section of the paper are to the assemblage asoriginally defined by Liem. Since the sequence in which group apomorphies are treated here differs somewhat fromthat used by Liem (1981 : 207), the number he gave to a character is given, in squarebrackets, after the number used in this review. The same convention is followed throughoutthe paper. (1) [1]. The entopterygoid is widely separated from the palatine (see Liem,1981: 205-206, 207; fig. 4; and Figs 1 A-F below). 1 For the sake of brevity, Liem's shorthand term for the assemblage, the OA, will be used in this paper as well. THE OPHTHALMOTILAP1A ASSEMBLAGE OF CICHLID FISHES 255 Scales '3 mm Fig. 1 Suspensoria (left) of: A, Callochromis macrops melanostigma; B, Xenotilapiatenuidentata; C, Cyathopharynx furcifer; D, Cyathopharynx schoutedeni (see p. 282); E.Grammatotria lemairei; F, Aulonocranus dewindti. As far as I can determine, this character, except for its occurrence in the Malawian genusLethrinops (and those species from Tanganyika which I propose to include in the 6>A) isunique to the Ophthalmotilapia assemblage. The gap is created, in part, by a reduction in the depth of the entopterygoid, and in part bya reduction of its extension above the quadrate. As a. result of these proportional andpositional changes, the anterior entopterygoidal margin lies below the level of the palatine'sposterior margin, and below the upper part of the ectopterygoid as well (with both of whichbones it would otherwise articulate). The space between the three bones (that is, the 'gap') isactually filled by dense, translucent connective tissue. 256 P. H. GREENWOOD It could be argued (and has been argued by Barel, pers. comm.} that the palatopterygoidgap is a correlate of the large and often ovoid eye characterizing all members of the OA. Thisis a complex problem to resolve since the influence of eye size and shape on cranialarchitecture is the result of several interacting factors, and there are few data available on thedynamics of syncranial ontogeny. In particular we have no information on the influence ofultimate eye size on the ontogeny of the whole suspensorium, of which the palatine andentopterygoid bones are but a part (see below p. 257). Taking the maximum diameter of the eye (expressed as a proportion of head length) as ameasure of eye size it is difficult to find a simple correlation between that ratio and the natureof the palatine-entopterygoid articulation. For example, in Hemibates, Trematocara,Reganochromis, certain ' 'Limnochromis'' species and in Perissodus (all from LakeTanganyika) the eye is as large as that in members of the OA, and the eye in Trematocara isas markedly ovoid as it is in any member of the OA. Yet, in none of these species is there apalatopterygoid gap. (Recently, Poll (1981) has revised the genus Limnochromis which henow divides into four genera. Since several of the older references in the literature are just tothe genus Limnochromis, I have adopted the convention of referring to the taxon as'Limnochromis', except when reference is made to species retained by Poll (1981) in thatgenus.) Conversely, in many species of the Malawian genus Lethrinops (Fig. 15) there is apalatopterygoid gap (albeit a less extensive one than in some though not all OA species), butthe eye is smaller than in members of the OA. Since, amongst the taxa with enlarged eyes only the OA species have developed apalatopterygoid gap, the gap would seem to be a shared derived character for the assemblage. That argument would hold even if future research shows the 'gap' to be one of severalcorrelated features associated with the evolution of a large eye. The presence of a gap inLethrinops (which does not have enlarged eyes), on the other hand, fails to support anyargument suggesting that eye size and a palatopterygoid gap are necessarily correlated. (2) [3]. The slender hyomandibula has a long symplectic process and no, or a veryreduced, hyomandibular flange. On the basis of extensive outgroup comparisons I cannot agree with Liem on the relativelength of the symplectic process. In fact, most OA species have a process which is no longerthan that in Astatotilapia burtoni, A. elegans (or in other Astatotilapia species), or that inother Tanganyika genera (Fig. 2). Unfortunately this character is difficult to quantify and isone affected by the level to which the anterior margin of the hyomandibular flange isproduced ventrally. To this extent the apparent length of the process is an optical illusion. Itis also affected by the stoutness of the flange which, in this region of the bone, is easilydamaged; if partially broken its absence may add to the apparent length of the symplecticprocess. As compared with the hyomandibular flange in many taxa (both within and outside LakeTanganyika) that in all OA species is reduced. But, I would question that it is ever absent,even in those species with the greatest reduction in flange area. A reduction in flange area comparable with that found in OA species does occur in severalother species as well, and these are mostly taxa with large eyes. Thus, amongst the LakeTanganyika endemics a reduced OA-like flange is found in Hemibates, Trematocara,Haplotaxodon, Grammatotria, Callochromis, Xenotilapia, Cardiopharynx, Cyathopharynxand Aulonocara. In contrast, Reganochromis calliurus, whose eye is as large as that in someOA species, has a moderately well-developed flange which is larger than that in any OAspecies. The hyomandibular flange is not reduced in Lethrinops (Lake Malawi) nor in any of thespecies examined whose modal eye size (i.e. eye diameter as a proportion of head length) isless than that of any OA species. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 257 i mm Fig. 2 Hyomandibula (right), in lateral view, of: A, Ectodus descampsi; B, Astatotilapia macropsoides. There is thus some suggestion that a reduction in hyomandibular flange area may becorrelated with a large eye size. This correlation could result, ontogenetically, from thedeveloping eye preempting some of the space which otherwise would be available for thelevator arcus palatini muscle whose origin is principally from the flange. It is noteworthythat in those species with a reduced flange, the levator arcus palatini has, relatively speaking,a reduced volume, and that its origin has shifted largely to the metapterygoid (Liem,1981 : 19 5-6; personal observations). Incidentally (pace Liem, 1981 : 207), in three of the four Oreochromis niloticus specimensexamined, hyomandibular flange area is proportionately equivalent to that in the OAspecies; in the fourth specimen it is noticeably larger. The two smallest O. niloticus (ca. 42and 52 mm standard length) have the narrowest flanges of the four specimens examined,again suggesting that flange area may be related to eye size since the smallest fishes haveproportionately the largest eyes. (3) [5] The vertical depth of the metapterygoid is shallow. It is difficult to assess the value of this character as an indicator of phyletic relationships.The metapterygoid is noticeably shallow in all the OA species recognized by Liem, and inthose which I would add to the assemblage. However, an equally shallow or even shallowermetapterygoid is known from several Tanganyika taxa (for example, in species ofLamprologus, Reganochromis, Trematocara, 'Limnochromis', and also in Perissodushecqui, P. microlepis and P. eccentricus). In some of these species the eye is large (equivalentin size to those of the OA species), in others it is smaller. From this information it isimpossible to decide whether a shallow metapterygoid is a correlate of eye enlargement, ahomoplastic feature, or one which might be indicative of phyletic relationship at a higherlevel than that under consideration. (4) [2] The posterior and dorsal margins of the palatine form a 90 angle. Elsewhere,Liem (1981: 206) expands this statement and notes that The 90 posterodorsal anglesurrounding a posterodorsal expansion of the palatine is not found in any otherTanganyika cichlids and deviates from the condition in generalized cichlids (e.g.Astatotilapia burtoni, Liem and Osse, 1975 and A. elegans, Barel, et al, 1976).' 258 P. H. GREENWOOD Liem (1981 : 208) places particular importance on this character, considering it as possiblythe only autapomorphic feature of the assemblage. My observations on the OA species indicate that the posterodorsal angle is not alwaysrectangular and that it shows some intra- and interspecific variability. Also, I would contestthe statement that the OA type of palatine shape is not found in other Tanganyika species. Itdoes occur, for example, in Limnochromis abeelei, and is closely approached in other speciesas well. Furthermore it occurs in species outside the lake (e.g. Astatotilapia macropsoides[Lakes Edward and George] and in some Lethrinops species [Lake Malawi]). I would agree, however, that the overall type of palatine morphology in OA species isrelatively uncommon amongst African cichlids, combining as it does a straight or virtuallystraight posterior margin meeting the dorsal margin at, or almost at, an angle of 90, and witha posteriorly expanded body of the bone. In combination these features give to the elongatevertical part of the bone a distinctive and near rectangular outline. Like character (3) [5], the palatine shape is difficult to evaluate as an indicator, or potentialindicator, of close phylogenetic relationship. However, since a similarly shaped bone israrely present amongst taxa other than members of the OA (and in those genera where it doesoccur it is not manifest by all member species), and since non-OA taxa with this type ofpalatine are not closely related to the OA, it might well indicate a shared common ancestryfor the Ophthalmotilapia assemblage. (5) [4]. The anterior margin of the pterosphenoid is notched. This character is so widely distributed amongst African cichlids (including the mostgeneralized taxa) that it cannot be treated as an apomorphy at this level of phyletic analysis. That the notch, or rather the tongue which delimits one aspect of the notch, has not beencommented upon before, or been shown in figures of cichlid neurocrania, may well be due toits fragility and hence loss during preparation of the skull. (It is of course absent in somespecies and is not invariably present in all members of a genus.) Also, the process can berather small and is then virtually invisible unless the skull is carefully cleaned of connectivetissue. Pace Liem (1981 : 207), the ligament connecting the sclerea with the pterosphenoid or itsnotch is present in generalized cichlids. In fact, it is present in all the cichlids I havedissected, and also in several other teleostean groups as well (including non-percoids). (6) Liem's two myological characters, viz. [7] the transversus dorsalis is reduced, and [8]the obliquus posterior is enlarged, may be taken together. At the outset of any discussion it must be made clear that neither character is easilyassessed, partly because of insufficient comparative data from outgroups, and partly becausethey are not readily quantified and are thus particularly subjective. In the discussion which follows, Asprotilapia is excluded from any generalizations aboutthe transversus dorsalis in the OA; Asprotilapia does show unequivocally clear-cutreduction of the transversus dorsalis anterior and the posterior head of the muscle is notdeveloped at all. From my observations on dorsal gill-arch muscles in OA taxa, and in other cichlids fromTanganyika and elsewhere (including generalized species such as Astatotilapia elegans, and'derived' taxa such as Bathybates, Hemibates and Trematocara (see also Stiassny, 1981), Iwould not consider the condition of the transversus dorsalis or the obliquus posterior in OAspecies to be trenchantly distinct. For example, although Liem (1981 : 207) considers thetransversus dorsalis complex in OA species to be reduced relative to that in A. elegans, andthe obliquus posterior to be hypertrophied, I could see no obvious differences when makingthe same comparisons. That there are differences in the extent to which these and other dorsal gill-arch musclesare developed amongst African cichlids cannot be denied (see for example Liem, 1973; Liem& Osse, 1975; Stiassny, 1981). But, the differences are rarely trenchant and in some cases THE OPHTHALMOT1LAPIA ASSEMBLAGE OF CICHLID FISHES 259 (including manifestly trenchant ones) are the result of environmentally induced individualvariations (see Liem & Osse, 1975 : 442, fig. 11; for environmental effects, see Greenwood1965). Thus, until considerably more comparative data are available, and until some means ofquantifying apparent differences is employed, the use of relative muscle size would seem tobe of very limited value, and could even be misleading. For those reasons I would not useeither of the muscle characters in attempting to unravel the phylogeny of the OA (except, asnoted earlier, with regard to Asprotilapia; see below, p. 263). (7) [6] The operculum has a distinct auricular process. Although Liem (1981 : 207) notes that a well-developed auricular process occurs on theoperculum in several taxa, it is only in Perissodus hecqui, 'Limnochromis' dhanisi and thesix genera now included in the OA (see p. 262) that I would consider the process equivalentto that found in the OA species. In all members of the assemblage, the opercular process,although interspecifically variable in form and size, is always a noticeable feature of thebone. That fact, coupled with the infrequent occurrence of a distinct process amongst othergenera (and its mosaic interspecific distribution in those taxa), would seem to enhance itsvalue as an indicator of monophyletic origin for the OA. In brief, of the eight supposedly apomorphic characters cited by Liem as suggesting amonophyletic origin for the OA, only three would seem to fulfil the necessary requirementsfor such features, and then mainly because of their unique congruence in the species con-cerned. The characters, as numbered above, are: the palatopterygoid gap (1), the morphologyof the palatine bone (4), and the presence of an auricular process on the posterodorsalmargin of the operculum (7). The remaining characters (i.e. 2, 3, 5 & 6) are either of no value, are possible homoplasies,or cannot be fully evaluated within the scope of our present knowledge. There are, however, two other characters noted by Liem, and used by him at a differentlevel of universality (the intragroup level), which I consider to be group apomorphies forthe whole assemblage. These are: (8) [15] The morphology of the lachrymal (1st infraorbital) bone. The lachrymal in members of the OA has a very distinctive appearance (see Liem, 1981,fig. 5; and Figs 3A-H), one which, apparently, is not replicated in any other African cichlid.It differs from the generalized condition (as seen, for example, in Astatotilapia macropsoides,Fig. 3J) in its overall protraction and relative shallowness. Because the anterior portion isnoticeably deeper than the posterior part, the bone has almost the appearance of a short butdeep handle extending from its expanded and near rhomboidal anterior region (Figs 3A-H);although the anterior region is much deeper than the posterior part, it is by no means asexpansive as it is in the majority of African cichlids. An elongate lachrymal occurs in a few other taxa (e.g. certain 'Limnochromis' species; Fig.31) but here the bone is uniformly protracted and so lacks the contrast between its shallowerposterior third to half and the deeper anterior part, which is so characteristic of the OA typelachrymal. Also, in these non-OA taxa with an elongate, or relatively elongate lachrymal,the dorsal margin of the bone is but slightly concave, whereas in the OA species it ismarkedly so. The anterior margin of the lachrymal is straight or very slightly concave, its anteroventralangle produced into a slight but distinctive peak (which is lacking in most species with anelongate lachrymal but is present in some, e.g, 'Limnochromis' permaxillaris and 'L'.pfefferi). When the bone is in situ its anterior margin slopes upwards at an angle of 50-60 to thehorizontal (the 'modal' slope in African cichlids would seem to be one close to thehorizontal, but there are several and apparently unrelated species where the slope is between45 and 60). 260 P. H. GREENWOOD Fig. 3 Lachrymal (left), in lateral view, of: A, Cyathopharynx furcifer; B, Callochromis macropsmelanostigma; C, Xenotilapia tenuicaudata; D, Xenotilapia boulengeri; E, Grammatotrialemairei; F, Cyathopharynx schoutedeni (see p. 282); G, Asprotilapia leptura; H, Aulonocranusdewindti: I, Limnochromis auritus; J, Astatotilapia macropsoides. Liem's figure of the OA species Cunningtonia longiventralis (and his comments on thebone, Liem, 1981 : 206, fig. 5C) represents an unusual condition in that species since theanterior lachrymal margin is shown as markedly concave and aligned almost horizontally. Inall 12 specimens I examined, the orientation of the bone and the shape of its anterior margin(including the anteroventral peak) are typically of the OA type (see Fig. 4). There are, with very few individual exceptions, six lateral line canal pores in thelachrymal of OA species (but see p. 263 below). The modal number of these pores in allAfrican cichlids is five; the occasional departure from that number seems to be attributableto individual variability. Thus, both in its outline shape (including the anteroventral peak)and in having six lateral line pores, the lachrymal appears to be a unique apomorphic featurefor the OA. Liem (1981 : 208) used the presence of an anteroventral peak on the lachrymal as asynapomorphy differentiating Ectodus, Lestradea, Ophthalmotilapia and Cunningtoniafrom Asprotilapia which, by implication, lacked this process. In the three specimens ofAsprotilapia leptura (the sole species) I examined, a typical OA peak is present and theoutline shape of the bone also conforms with that of other OA species (see Fig. 3G). THE OPHTHALMOT1LAPIA ASSEMBLAGE OF CICHLID FISHES 261 omm Fig. 4. Cunningtonia longiventralis, left lateral view of lachrymal in situ. (9) [9]. The presence on the anguloarticular bone of a well-developed, anteroposteriorlyaligned fossa for the insertion of the adductor mandibulae muscle A 2 . This well-developed fossa is noted by Liem (1981 : 195) in his anatomical description ofEctodus descampsi, and is illustrated in the six species depicted in his figure 4. Liem did not,however, use the feature as a major apomorphy for the OA as a whole. Indeed, he used thepresence of an extensive fossa in Asprotilapia as an autapomorphic character for that genus(Liem, 1981 : 208). In all <9A species the fossa is clearly defined, and is limited anteriorly by a prominent,near-vertically aligned ridge on the anguloarticular. The principal surface for muscleinsertion is situated on the lateral aspect of the bone, and is thus unlike the generalizedcondition where it lies on and across the bone's posterior margin. The insertion face, andhence the area of the fossa, varies in size amongst members of the OA. It is largest inAsprotilapia, smallest in Cunningtonia, with the other species occupying various inter-mediate positions in the range. Parenthetically, it may be noted that amongst the species Iwould now include in the 6>A, the fossa is small in Cyathopharynx, of intermediate size inXenotilapia and Grammatotria, and largest in Callochromis. Amongst the outgroup taxa examined, an 0A type fossa is found only in Lethrinops (LakeMalawi), Trematocara (Lake Tanganyika), Chromidotilapia batesi and C. kingsleyae (WestAfrica) and, in a poorly developed state, in the following Lake Malawi 'Haplochromis'species 'H. ' prostoma, 'H. ' johnstoni, 'H. ' breviceps and 'H. ' tetrastigma. (For the use of thename 'Haplochromis' see Greenwood [ 1 979 : 3 1 7]). In all other outgroup taxa examined the adductor mandibulae A 2 inserts along the some-what medially expanded posterior margin of the anguloarticular. There is no obvious evidence to suggest a recent common ancestry shared by the O\ andthe Chromidotilapia species, nor one between any of the Lake Malawi 'Haplochromis' andeither the OA or the Chromidotilapia species. 262 P. H. GREENWOOD In the current state of our knowledge one can be less certain about making a similar state-ment with respect to the OA and either Trematocara (also from Lake Tanganyika) or theMalawian genus Lethrinops. Thus a suspended judgement only can be given on whether thepresence of an OA-type fossa in these three taxa is a homoplasy, or whether it is asynapomorphy for a group of higher universality than the one under discussion (see alsop. 257). Because the fossa is present in all OA species (including the new additions, see below) andbecause it is congruent with other apomorphic features shared by them, it can, I would con-sider, be taken as further evidence for the monophyly of the lineage. To summarise this review of group characters in the Ophthalmotilapia assemblage (asdefined by Liem, 1981), the congruent apomorphic features are:(i) The presence of a palatopterygoid gap (see p. 254)(ii) The morphology of the palatine bone (see p. 257)(iii) The presence of an auricular process on the opercular bone (see p. 259)(iv) The shape of the lachrymal bone (1st infraorbital); see p. 259)(v) The presence of a well-defined, laterally placed fossa for the insertion of the A 2adductor mandibulae muscle on the anguloarticular bone (see p. 261). The Ophthalmotilapia assemblage reconsidered On the grounds of all their constituent species sharing the five features listed above, andbecause none possesses any feature which might suggest other relationships, six furthergenera can be included in the Ophthalmotilapia assemblage (see Figs 1,3, 10 and 1 1). The new additions are: (i) Xenotilapia Blgr, 1899 (type species X. sima Blgr) (ii) Callochromis Regan, 1920 (type species Pelmatochromis macrops Blgr) (iii) Grammatotria Blgr, 1899 (type species G. lemairii Blgr) (iv) Cyathopharynx Regan, 1920 (type species Tilapia grandoculis Blgr; but see p. 284).(v) Cardiopharynx Poll, 1942 (type species C. schoutedeni Poll) (vi) Aulonocranus Regan, 1920 (type species Paratilapia dewindti Blgr) Although not apparently relevant to the question of their phyletic relationships, it may benoted that these six genera also share with members of Liem's original OA all the otherfeatures discussed in the previous section (i.e. characters 2, 3, 5 & 6). For taxonomic and biological details of these genera and their contained species, referenceshould be made to Poll (1946 & 1956). Since I can find no grounds for maintainingCyathopharynx and Cardiopharynx as separate genera (see p. 282), all further references tothese taxa will be made under the name of the senior synonym, Cyathopharynx. Liem (198 1 : 208) has already suggested that Xenotilapia, Callochromis and Aulonocranusmight be the sister lineage of the Ophthalmotilapia assemblage, but he felt that moreinformation was needed before their 'precise relationships' to the others could bedetermined. His reason for making this suggestion was that all three genera share with theOA a palatopterygoid gap and various derived features of the hyomandibula, ie characters (1)and (2) above. That the three taxa also shared the other six and supposedly apomorphicfeatures originally used to define the OA, was not noted by him. Although I would be chary of claiming that the 'precise relationships' of Xenotilapia,Callochromis, Aulonocranus, Grammatotria and Cyathopharynx have been determined, Iwould submit that an a priori case can be established for including them within theOphthalmotilapia assemblage itself, and not just as a sister-group to that lineage. Before going on to consider intralineage relationships within the expanded OA, somecomments must be made about certain features in Aulonocranus and Xenotilapia. In Aulonocranus the lachrymal has the characteristic shape of that bone in other OAspecies (Fig. 3H), but it lacks the anteroventral peak; the anteroventral angle is rounded and THE OPHTHALMOT1LAPIA ASSEMBLAGE OF CICHLID FISHES 263 so resembles the condition found in most cichlid taxa. It is possible that this atypical anteriorprofile might be attributed to the greatly inflated laterosensory canals in the lachrymal ofAulonocranus. In most Xenotilapia species too, the lachrymal is atypical for the OA because in thesespecies there are, modally, five and not six openings to the laterosensory canal system (Figs3C & D). There is, however, considerable inter- and intraspecific variation in pore number;some species have six pores, others only four, and some individuals have a different numberof pores on each side of the head. Clearly, pore number is an unstable characteristic inXenotilapia. The genus also differs from all other OA taxa, and all other African cichlids Ihave examined, in having the posterior opening to the lachrymal laterosensory systempositioned below and not opposite the anterior canal opening in the second infraorbitalbone. As in Aulonocranus, the outline shape of the lachrymal in Xenotilapia is a typical OA one(Figs 3C & D); not surprisingly, considering the number of species (11), there is rather morevariation on that basic shape in Xenotilapia than in other members of the assemblage. Relationships within the Ophthalmotilapia assemblage In its original form, the assemblage was divided into two major sublineages, one comprisingonly Asprotilapia leptura, the other containing the four remaining genera, Ectodus,Lestradea, Ophthalmotilapia and Cunningtonia (see Liem, 198 1 : 208 & fig. 9). As defining features for the larger sublineage, Liem employed two supposedlysynapomorphic characters: (i) the morphology of the lachrymal, and (ii), the dominance ofthe A, division of the adductor mandibulae muscle complex (Liem's characters 15 & 16respectively). The Asprotilapia lineage was recognized both by the absence of those features, and, moreimportantly, by its having six presumed autapomorphic characters (see Liem, 198 1 : 208). As argued above (character (8)[15], page 259), the features of the lachrymal must now beconsidered an apomorphic character for the whole OA (including the new additions andAsprotilapia itself)- The muscle character, according to Liem, has two components. First, that the A, divisionof the adductor mandibulae has become the dominant component of the complex, its crosssectional area surpassing '. . . that of the other parts', and second, that '. . . its origin hasexpanded ventrally at the expense of the adductor mandibulae part A 2 ' (Liem, 1981: 208). It is difficult to test the first claim adequately, and my attempts to do so failed to confirmLiem's claims, especially if, as his statement implies, the A, division is dominant to boththe A 2 and A 3 divisions combined. However, the area of origin of A, in the taxon concernedis very clearly greater than that of A 2 when measured by its extent along the vertical arm ofthe preoperculum (see fig. 6 in Liem, 1981). When, however, the additional OA taxa are taken into account, the second feature shows acontinuous range of variation from a state where the origins of both A, and A 2 occupy anapproximately equal depth on the vertical preopercular limb, to one where the origin of A 2is virtually excluded from that limb and thereby is almost confined to the horizontal part ofthe bone. Furthermore, even within a single genus (as in Xenotilapia and Callochromis}some species have A, and A 2 with almost equal depths of vertical origin (the plesiomorphcondition in cichlids), others have the depth of A, greater than A 2 , and yet others have A 2with a much greater vertical depth of origin than A, (Fig. 5B). Thus there would no longerseem to be any grounds for maintaining the unity of the Ectodus- Cunningtonia sublineageon the basis of its myological characters. The peculiar arrangement of the adductor mandibulae muscles in Asprotilapia still standsas a well-defined autapomorphy for the genus. Here, division A, is markedly reduced, bothin the depth of its origin on the preoperculum, and in its overall bulk (Fig. 5A), a conditionnot found elsewhere in the OA. Other features, however, suggest that Asprotilapia is closelyrelated to some of the newly incorporated members of the assemblage (p. 265). 264 P. H. GREENWOOD B Fig. 5 Superficial adductor mandibulae muscles (left side) of: A, Asprotilapia leptura; B,Xenotilapia sima. AM, & AM 2 : adductor mandibulae divisions 1 and 2; TAM, : tendon ofinsertion for AM,. Although a primary dichotomy within the OA cannot be made on the charactersemployed by Liem, such a dichotomy can be established on differences in the length of thegut and the manner in which it is arranged within the visceral cavity (see Figs 6 & 7). Liem (1981 : 209; character [19]) used intestinal length to define a group of taxa(Ophthalmotilapia, Cunningtonia and Lestraded) within the original 0A, but he did notcomment on the spatial arrangement of the alimentary tract in those species. Within the expanded OA, the member taxa can be grouped into those with an intestinallength less than 3 times the standard length of the body, modally 2-3-2-5 times SL, and thosewith an intestinal length 3-6 times the standard length. Species in the first group have the relatively short gut coiled into a few loops whosearrangement is in an essentially anteroposterior direction. In contrast, species with a long guthave the intestine much coiled and the coils are arranged in an essentially transversedirection (see Figs 6 & 7). At first sight this transverse arrangement gives an erroneousimpression of the intestine actually being coiled around the stomach (Fig. 7). A long and complexly coiled gut is a derived feature, occurring in several cichlid lineages.But, since these lineages are not closely related it can be treated as a synapomorphy at thelevel of universality involved here. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 265 mm Fig. 6 Alimentary tract (left lateral view) of an Asprotilapia subassemblage taxon (Grammatotria lemairei). mm Fig. 7 Alimentary tract (left lateral view) of an Ophthalmotilapia subassemblage taxon (Ophthalmotilapia boops). Taxa of the OA belonging to the group with a long and transversely coiled intestine are:Lestradea, Ophthalmotilapia, Cunningtonia and Cyathopharynx - hereafter referred to asthe Ophthalmotilapia subassemblage; their interrelationships will be considered later (see p.271). The group with a short and longitudinally coiled intestine comprises the generaAsprotilapia, Callochromis, Xenotilapia, Grammatotria, Ectodus and Aulonocara-hereafter referred to as the Asprotilapia subassemblage. Four genera in the Asprotilapia subassemblage, viz. Asprotilapia, Callochromis,Xenotilapia and Grammatotria, share an apomorphic feature which suggests their sharedcommon ancestry. This character is the presence of a fully developed pharyngeal hangingpad, with its associated modifications to the superficial anatomy of the gill-rakers; see Figs 8&9. The pad is an hypertrophied and well-circumscribed, forwardly directed, and turgid fold ofthe buccopharyngeal tissues. It lies immediately anterior to the upper pharyngeal bones andextends forward and downward, as a visor-like projection, for a short distance in front of thefirst gill-arch. Posterolaterally the pad is fused with the thickened tissue covering theepibranchial gill-rakers of the first gill-arch, but over most of its width the visor-like part is 266 P. H. GREENWOOD Fig. 8 Pharyngeal hanging pad in Xenotilapia boulengeri. Left side, seen from a slightly dorsolateral viewpoint. mm Fig. 9 Ventral portions of the gill-arches, and the lower pharyngeal bone of Xenotilapiaboulengeri, viewed from above to show the nature of the gill-rakers. separated from the buccal roof by a distinct transverse groove. This groove is open anteriorlyand laterally. In the buccal midline the left and right halves of the visor-like portion areconfluent with the buccopharyngeal roof. When the mouth is closed, and the ventralgill-arch skeleton is adducted, the visor occludes the pharynx, leaving only a narrow channelcoincident with the medial area of confluence between the two halves of the visor and thebuccopharyngeal epithelium. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 267 The pharyngeal surface of the visor is thrown into a large number of broad-based butterminally acute papillae. When the gill-arches are adducted the papillose area is broughtinto close contact with the upper surface of the cerato- and hypobranchial regions of thegill-arches. These, in turn, are covered by a greatly thickened and soft epithelium. The innerand outer gill-rakers on the ceratobranchial of each arch are joined transversely by anvil-shaped folds of thickened tissue so that, superficially, there appears to be but a single series ofrakers, with each raker extending across the breadth of the arch (Fig. 9). A pharyngeal hanging pad is known from two other, and very dissimilar, African genera,Chromidotilapia and Tylochromis, and in the South American Geophagus (see Trewavas,1 974 : 389-392). Since there are several features strongly indicating that none of these taxa isclosely interrelated, and that none is closely related to the Ophthalmotilapia assemblage, therepeated occurrence of a hanging pad can only be interpreted as the result of convergentevolution. Any other explanation would be most unparsimonious. It would be equallyunparsimonious to assume that its occurrence in four of the OA taxa was the result ofindependent evolution in each genus. This is particularly so since three of the four generashare another apomorphy, a high number of caudal vertebrae in Xenotilapia, Asprotilapiaand Grammatotria, and a second derived character, the shape of the dentary, is shared byXenotilapia, Callochromis and Grammatotria, of which taxa Callochromis alone does notshare the vertebral apomorphy. Thus, on the basis of their all possessing a pharyngeal hanging pad, Xenotilapia,Callochromis, Grammatotria and Asprotilapia 1 are taken to form a natural group within the<9A. The group can be further subdivided on the distribution within its members of certainother derived features. Xenotilapia, Callochromis and Grammatotria all have a peculiarly shaped dentary (Fig.10). When viewed laterally, the dorsal margin of the bone is seen to dip downwardsimmediately behind the last tooth in the outer row. It continues posteriorly at this lowerlevel until it curves upwards to form the anterior margin of the coronoid process. Thealveolar surface is confined to that part of the bone preceding the step, behind which it isedentulous. The dentary of Asprotilapia will be discussed below. This type of dentary, as far as I am aware, is not found in any other cichlid from LakeTanganyika, or for that matter from Lake Victoria either. It is, however, closely approachedby the dentary in the Malawian genus Lethrinops and in some 'Haplochromis' species fromthat lake. The significance of this similarity, and the occurrence in Lethrinops of apalatopterygoid gap, is discussed on page 279. Within the group comprising Xenotilapia, Callochromis and Grammatotria it isimpossible to determine which two genera are the more closely related since no clear-cutlinking synapomorphies can be recognized. It is accepted that the diagnostic 'generic'characters for each genus are autapomorphies for that taxon. At present the trio can only betreated as an unresolved trichotomy, but with the suggestion that further research may showXenotilapia and Grammatotria to be sister taxa. Asprotilapia, the fourth member of the group, is a most distinctive taxon, in which Liem(1981 : 208) identified six autapomorphies. These must now be reviewed in the context ofthe expanded Ophthalmotilapia assemblage. As in earlier discussions, Liem's apomorphy number is given in square brackets. (i) [9]. The elongate, slender mandible has an expanded adductor fossa for the A 2division of the adductor mandibulae muscle. As noted earlier (p. 261) the adductor fossa is expanded in several members of the OA,particularly in species of the group to which Asprotilapia belongs. However, even amongstthose species the fossa is most expansive in Asprotilapia. 'Microbranchiospines are present in all four of these Tanganyika genera; in this respect they resemble Tylochromisand differ from Chromidotilapia (see discussion in Trewavas, 1 973 : 1 7 & 1 974 : 388). 268 P. H. GREENWOOD ADDF Fig. 10 Dentary and anguloarticular in two species of the Asprotilapia subassemblage : A,Grammatotria lemairei (lateral view); B & C, Xenotilapia boulengeri (lateral and ventral viewsrespectively). ADDF : adductor fossa. NF : nerve foramen. 1mm Fig. 11 Left dentary and anguloarticular of Asprotilapia leptura in : A, lateral view; B, occlusalview. In B the teeth have been restored (based on a spirit specimen), but in A only the tooth scarsare shown. THE OPHTHALMOTILAP1A ASSEMBLAGE OF CICHLID FISHES 269 The mandible is not, in my view, especially elongate, particularly when it is comparedwith that element in other members of the Asprotilapia subassemblage (cf. Figs 10 & 1 1). Itsgross morphology differs from that in Xenotilapia, Callochromis and Grammatotria since itlacks a 'stepped' alveolar margin, and because each ramus of the jaw has a more abrupt andstronger medial curvature towards the symphysis. Also, in Asprotilapia, immediately beforethe dentary curves inwards, the alveolar surface is produced laterally so as to form anoticeable, shelf-like overhang of the underlying ramus (Fig. 1 1). Overall, the dentary in Asprotilapia bears a fairly close resemblance to that bone inLabeotropheus of Lake Malawi, a resemblance enhanced by the tricuspid, slender-necked,procumbent and movably implanted teeth present in both genera. Asprotilapia has, how-ever, a shallower dentary, and the anguloarticular is taller and more expansive than inLabeotropheus. Parenthetically it may be noted that Asprotilapia, like Labeotropheus, hasan enlarged and similarly shaped cartilaginous meniscus underlying the premaxillaryascending process, and a similar fleshy medial projection overlying the broad palato-premaxillary ligaments. Observations made from radiographs of the two species, and frommanipulation of preserved specimens, indicate that the protrusile mechanisms in the twospecies are very similar. There are, however, no reasons to doubt that these similaritiesshould be treated as homoplasies. (ii) [10]. The posterior head of the transversus dorsalis anterior muscle is absent. This seems to be a clear-cut autapomorphy, but should be checked in more examples thanthe single specimen available to Liem and myself. (iii) [11]. Lateral ethmoids greatly enlarged. This again is an apparently good autapomorphy.(iv) [12]. Interorbital width greatly reduced. The interorbital width is reduced in some Xenotilapia species, but in none is it as narrowas in Asprotilapia. (v) [13]. The reduced articular process of the premaxilla is in a more forward position. There is considerable and continuous variation in the relative position of this process inXenotilapia species, and indeed within the taxa of the entire OA. It would not, therefore,seem to be a character of particular value. (vi) [14]. The greatly enlarged cranial condyle and the premaxillary process constitutethe bulk of the maxilla. The maxilla of Asprotilapia is less outstanding when compared with that bone in othermembers of the subassemblage, especially Xenotilapia (Fig. 12). In X. boulengeri, forexample, the process is larger than in Asprotilapia. The cranial condyle, however, is largestin Asprotilapia and, as compared with all other <9A species, the whole bone is relativelyforeshortened. In the context of the expanded <9A, I would consider that at least three of theautapomorphies originally proposed (i.e. nos. i, ii and iii above) retain their validity (if, thatis, the apomorphic features of the lower jaw are interpreted as has been done here). Although not listed as an autapomorphy by Liem (198 1), the peculiar condition of the A,division of the adductor mandibulae muscle in Asprotilapia would seem deserving of thatstatus. In Asprotilapia, as compared with all other OA taxa, A, is a very short and narrowmuscle with an extremely long tendon of insertion (Liem, 1981; fig. 6; also Fig. 5), and anarea of origin much smaller in all respects than that of the A 2 division. No other species inthe OA has this arrangement of the adductor mandibulae muscles. 270 P. H. GREENWOOD mm 1mm Fig. 12 Right maxilla, in lateral view, of: A, Xenotilapia boulengeri; B, Asprotilapia leptura (thedamaged ventral margin of the premaxillary saddle is indicated by a broken line). When commenting on the unusual A, muscle in Asprotilapia, Liem (1981:203)makes particular reference to a tendon stemming from the muscle's principal tendon ofinsertion, and which joins the tendinous part of adductor division A^. This emphasis mightgive an impression that the vertically directed interconnecting tendon is a unique (orunusual) feature of Asprotilapia. That is not so because the tendon is present in all cichlidswhose jaw musculature has been examined (see for example Liem & Osse, 1975: fig. 6;Stiassny, 1981 : 80, and figs 8 & 1 1 [tA.b]; also personal observations on species other thanthose studied by these authors). Asprotilapia is, however, unusual in having a discrete,ligament-like condensation in the connective tissue between the premaxilla and dentary,inserting on the lateral (and not the medial) face of the premaxilla. Interestingly, despite the resemblances in lower jaw morphology between Asprotilapia andLabeotropheus (see above p. 269), the pattern of adductor muscles in the two genera is quitedifferent. Labeotropheus has the typical generalized cichlid arrangement. The various autapomorphies of Asprotilapia serve to indicate the taxon's isolation withinits group. For the moment it can only be treated as the sister taxon to the other three genera(i.e. Callochromis, Xenotilapia and Grammatotria) combined. The two remaining taxa in the Asprotilapia subassemblage (see p. 265), Ectodus andAulonocranus, are 'interrelated' only at the level of their sharing with their supposed sister-group, Asprotilapia, Callochromis, Xenotilapia and Grammatotria, the plesiomorphicfeature of a short intestine. With one possible exception, neither Ectodus nor Aulonocranushas any uniquely shared derived features suggestive of their recent common ancestry. The exceptional feature concerns the marked elongation of the first, and to a lesser extentthe second pelvic fin rays in adult males. In Aulonocranus the first ray extends to about themiddle of the anal fin, and in Ectodus to a point beyond its spinous part, but never as far asthe middle of the fin. Other members of the Asprotilapia subassemblage have variouslymodified relative proportions of different pelvic fin rays (see Poll, 1956) but none has the firstray elongated to the degree found in Ectodus and Aulonocranus. The significance of thisfeature is, however, somewhat questionable (see p. 272). THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 271 In Liem's scheme (1981 : 208-9, fig. 9), Ectodus was allied with Lestradea, Ophthalmo-tilapia and Cunningtonia (now considered to be a separate sublineage, see below)and not with Asprotilapia. Liem's grounds for this alliance were based on Ectodus sharingtwo apomorphic characters with Lestradea, Ophthalmotilapia and Cunningtonia, namely adistinct antroventral process on the lachrymal bone, and a dominant Aj division in theadductor mandibulae muscle complex. As was argued above (pp. 262-263) the former character is one shared by all members ofthe OA (except Aulonocranus) and the latter is a variable feature of little value as anindicator of relationships within the OA. Aulonocranus was not included in the original OA, although Liem (1981 : 206 & 208) didsuggest that it might be related to that assemblage. The genus is readily identified by one outstanding apomorphy, the hypertrophy of itscephalic laterosensory canal system, and as a probable correlate, the enlarged saccular bullain the basioccipital and prootic bones. Ectodus has one autapomorphy, viz. the vertical and horizontal limbs of the pre-operculum are of equal or almost equal length (see Liem, 1981 : 209). Liem also ranked theenlarged saccular bulla as an autapomorphy, but the bulla is enlarged to an equal extent inLestradea and Ophthalmotilapia amongst members of the original OA, and in Aulonocranusand Cyathopharynx among the new additions to the assemblage. Indeed, apart fromCallochromis, Xenotilapia and Grammatotria the saccular bulla is noticeably enlarged in allOA species, the degree of enlargement showing a continuous increase from the condition inAsprotilapia through to that in Aulonocranus, with Lestradea, Ectodus and Ophthalmo-tilapia all close to Aulonocranus. Turning now to the second lineage of the primary dichotomy, the Ophthalmotilapiasubassemblage (p. 265), one again finds difficulty in establishing intragroup relationships,but little difficulty in recognizing autapomorphies for the constituent genera, viz. Lestradea,Ophthalmotilapia, Cunningtonia and Cyathopharynx. Since Lestradea lacks the various derived features shown by its congeners in thesubassemblage it would appear to be the plesiomorph sister taxon to the other three generacombined. Liem (1981 : 209) singled out two autapomorphies for Lestradea (i) the edentulousanterior process of the lower pharyngeal bone is only half as long as the toothed part (hisapomorphy 20), and (ii), the body of the maxilla is stout, and has a prominent postmaxillaryprocess (his apomorphy 2 1 ). I cannot agree with Liem's statement about the relative proportions of the anterior processof the lower pharyngeal bone* The Lestradea specimens I have examined all have thisprocess as long as, or almost as long as the toothed part, a condition approaching that inOphthalmotilapia and Cunningtonia. This condition must be considered derived relative tothat in Astatotilapia and many other African cichlids. In these various taxa the edentulousanterior process of the bone is indeed half or less than half as long as the toothed part. The body of the maxilla in Lestradea is stout and the posterior process is prominent, butin both features the bone does not differ from the maxilla in Ophthalmotilapia where theprocess is a little less prominent. A relatively short maxilla with a prominent posteriorprocess is, it would seem, a common feature in the whole Ophthalmotilapia assemblage. Cyathopharynx is readily distinguished by its autapomorphous lower pharyngeal bone(Fig. 23), with its deeply concave occlusal surface, heart-shaped dentigerous area, and greatlyinflated body below that surface. There are, however, no synapomorphies (except groupones) shared by Cyathopharynx and any other member of the subassemblage. Liem (1981:209) recognized three synapomorphies which suggested to him thatOphthalmotilapia and Cunningtonia were members of a monophyletic unit. The presumedsynapomorphies were (i) jaw teeth with long stalks, and movably implanted, (ii) the firstpelvic ray greatly elongate (Liem's apomorphies 22 & 23 respectively), and (iii) the posteriormargin of the vertical preopercular limb is straight and forms a 90 angle with the horizontallimb. 272 P. H. GREENWOOD The teeth in all members of the OA (sensu lato) are movably implanted, and in all speciesthe teeth could be described as having 'long stalks', albeit with specifically distinct butvarying degrees of slenderness. For example, in both these features the outer jaw teeth inOphthalmotilapia are very like those in Lestradea, but those in Cunningtonia have a muchmore slender and elongate neck than do the teeth in any other OA species. The first and second pelvic rays are elongate (much more so in males than in females) inboth Ophthalmotilapia and Cunningtonia, with, in the former, the first ray somewhat longerthan the second. But, when other species in the expanded OA are taken into account onefinds that the condition of the two rays in Cyathopharynx furcifer is like that inCunningtonia longiventralis, and that Aulonocranus also has elongate rays but which extendonly to the middle of the anal fin base and not to the posterior margin, or slightly beyond, asthey do in Cunningtonia and Cyathopharynx furcifer. Cyathopharynx schoutedeni haselongate first and second pelvic rays too, but in this species it is the second ray which is thelonger and, unlike the elongate first ray in C. furcifer, it extends posteriorly only a shortdistance beyond the spinous part of the anal fin. Thus it is difficult to treat marked elongation of the first, or first and second pelvic rays as asynapomorphy for Ophthalmotilapia and Cunningtonia. That it is a character showingcontinuous variation and one that is incongruent with other apomorphies would alsopreclude its use as an indicator of recent shared common ancestry for the genera exhibitingit, namely Aulonocranus, Ophthalmotilapia, Cunningtonia and Cardiopharynx (see below). Some comment on the spatulate, bifid tips to the first ray in males of Ophthalmotilapiaspecies would be appropriate here. The feature appears to be a unique apomorphy for Ophthalmotilapia, indeed it is one ofthe reasons given by Liem (1981 :210) for synonymizing Ophthalmochromis with thatgenus. Spatulate tips have not been mentioned in the formal descriptions of any otherspecies, nor have I seen such modifications in any of the taxa, other than Ophthalmotilapia,which I have examined. However, Brichard (1978 : 187) describes the pelvic fin inCyathopharynx furcifer as having '. . . a long filament tipped with a yellow double spatula'.On page 190 of the same book he comments on 'The ventral filaments, reaching the end ofthe anal fin are each tipped with a double yellow-orange spatula'. Finally, Brichard(1978 : 148 & 149) provides two illustrations of a live individual (or individuals) showingwhat certainly looks like a spatulate tip to the protracted pelvic fin rays. Brichard is an experienced underwater-naturalist and one is hesitant to suggest he hasmisidentified his material. On the other hand, in his description (Brichard, 1978 : 326) ofOphthalmotilapia ventralis (as Ophthalmochromis ventralis) he remarks that The males . . .have very long pelvic filaments also forked like O. nasutus, but the tips are without spatulae(although they are pale yellow like those of O. nasutusY; bold type added. Clearly there issome terminological confusion here since none of the male O. ventralis specimens I haveexamined, or those which have been described elsewhere, lacks well-defined spatulae. The tips of the elongate pelvic rays in Cyathopharynx are brightly coloured, and seem toserve the same ethological function as the spatulae in Ophthalmotilapia. Perhaps theconfusion stems from that colour similarity rather than a structural one? Until spatulateCyanopharynx are described formally, the uncertainty will remain. In itself, very marked elongation of the first and second pelvic rays is a derived feature (seebelow), but in the context of the OA it is not clear at what level of universality it can berecognized as a synapomorphy. For example, if used to unite Aulonocranus, Cyathopharynx,Ophthalmotilapia and Cunningtonia, pelvic fin length would be incongruent with the groupsynapomorphy of an elongate and complexly folded intestine shared by all exceptAulonocranus, which has a short and simply folded intestine. If, on the other hand, becauseof its short gut Aulonocranus was taken to be the plesiomorph sister taxon of the others, thenwhere would Lestradea which has a long gut but short pelvic fins be placed? The problem is further compounded by Ectodus which has a short and simply folded gutbut elongate first and second pelvic rays produced to a degree almost comparabale with thosein Aulonocranus. THE OPHTHALMOT1LAPIA ASSEMBLAGE OF CICHLID FISHES 273 In no case are there any known synapomorphies which are uniquely congruent withprotracted pelvic fin length, and which would thereby establish a strong case for arguing thatthe gut character is a homoplasy. Finally, it must be recalled that some relative elongation ofthe first and second pelvic rays is of relatively common occurrence in African cichlids (seeGreenwood, 1981), and that the degree of elongation must be treated as a continuousvariable, albeit one rarely reaching the extremes found in Cunningtonia and Ophthalmo-tilapia. Liem's third synapomorphy (involving the shape of the preoperculum; see page 271) isalso difficult to substantiate, particularly in the context of the expanded OA. In allconstituent species of the O\ sensu law the vertical and horizontal limbs of the preopercularbone meet at, or very nearly at, a right angle, and in all except Ophthalmotilapia andCunningtonia the posterior margin of the vertical limb curves inwards for a short distancenear its dorsal extremity. Because Ophthalmotilapia and Cunningtonia have the bonyflange behind the vertically aligned laterosensory canal tube narrowing imperceptibly, ratherthan abruptly, the entire posterior margin of the preoperculum does give the impression ofbeing straight in these two species. The difference between the two types of posterior marginis, however, very slight and is almost obliterated by the condition in Cyathopharynx. Herethe dorsal extremity of the margin is slightly indented, but less noticeably so than in mostother OA taxa. Thus, of the three apomorphies under review, only the nature of the preopercular margin,a not particularly trenchant character, would seem to be a synapomorphy linkingOphthalmotilapia and Cunningtonia. On the morphological evidence alone, it is thus impossible to hypothesize preciseintragroup relationships for the Ophthalmotilapia subassemblage, except to note that on asimple summation of derived features Lestradea would seem to be the most plesiomorphtaxon. At the generic level the other taxa are readily identified by their particularautapomorphies (see below). It is possible that these currently obscure and therefore uncertain intralineage relation-ships will be clarified when more data are available on the breeding habits of its constituentspecies. Brichard (1978 : 108), for example, groups Cyathopharynx with Ophthalmotilapiaas polygamous spawners in which there is no contact between the sexes at the nest site. Theimport of Brichard's statement is not really clear, and the phyletic importance of mostethological characters still awaits evaluation. Nevertheless it is suggestive that Brichard didsingle out these taxa as forming a distinctive reproductive class. Returning now to the autapomorphic features of the genera constituting the sub-assemblage, one finds that Ophthalmotilapia (sensu Liem, 1981 : 210) is distinguished bythe bifid spatulae in which each elongate first pelvic ray terminates, and the subdivision ofthe retractor dorsalis muscle of the upper gill-arches into two distinct heads (see Liem,1981 :201,fig.8D). Cunningtonia has as its principal autapomorphies the nature of its oral dentition, the stoutforeshortened dentary (Fig. 1 3) and the stout premaxilla. 5mm Fig. 13 Left dentary and anguloarticular of Cunningtonia longiventralis in lateral view. 274 P. H. GREENWOOD Most of the jaw teeth are very slender and tall, with strongly recurved, tricuspid and broadcrowns. The teeth are arranged in wide bands over the entire alveolar surface of thepremaxilla and on the transverse part of the dentary, but on the lateral alveolar surface theteeth are much stouter and are unicuspid. The palatine of Cunningtonia is also unique in having its facet for articulation with thelateral ethmoid expanded medially into a shelf-like projection. When viewed from the side,especially in specimens prepared as alizarin transparencies, the shelf has a spine-likeappearance (see Liem, 198 1 : 2 10; apomorphy 26). Liem considered that the symplectic in Cunningtonia was '. . . very elongate' (hisapomorphy 28), but I find that it is of virtually equal relative length in all members of theOA. In Cyathopharynx the principal autapomorphy is the peculiar, heart-shaped and deeplyconcave dentigerous surface of the lower pharyngeal bone and, of course, the correlatedchanges in the shape of the upper pharyngeal elements (see p. 288). The nature of the apophysis for the upper pharyngeal bones in the Ophthalmotilapiaassemblage Like Liem (1981), I have not taken into account the nature of the pharyngeal apophysiswhen analysing intragroup relationships amongst members of the OA (see Greenwood,1978, for an evaluation of this character in determining phyletic relationships; also Fryer &lies, 1972:504^5^.). Based on material examined personally, the distribution of apophyseal types (defined as inGreenwood, 1978) within the two major subassemblages of the 0A is as follows: Asprotilapia subassemblage Ophthalmotilapia subassemblage Hap. 1 Troph. Tilapia Hap. Troph. Tilapia Callochromis Xenotilapia Cyatho- Lestradea (a few spp.) pharynx Xenotilapia Aulono- (most spp.) cranus Ophthalmotilapia Grammatotria* Ectodus Cunningtonia l ffap. = Haplochromis type; Troph. = Tropheus type; Tilapia = Tilapia type. *In all specimens the apophysis is of the Haplochromis type on one side, and the Tropheus type on the other. It is interesting that no true Tilapia type apophysis occurs amongst members of theAsprotilapia subassemblage, but that three of the four taxa in the Ophthalmotilapiasubassemblage do have an apophysis of that type. In contrast, a Haplochromis typeapophysis occurs in three members of the Asprotilapia subassemblage but not in any taxa ofthe Ophthalmotilapia subassemblage. It must be borne in mind, however, that the Tropheus apophyseal type is structurallyintermediate between the Haplochromis and Tilapia types (see Greenwood, 1978) and thatin the Asprotilapia subassemblage certain taxa have more than one type of apophysis.Indeed, the three specimens of Grammatotria lemairei examined have a Haplochromis typeapophysis on one side of the skull and a Tropheus type on the other. It is also interesting to note that if the Tropheus category (of Greenwood, 1978), whichRegan did not recognize, is ignored and the taxa in that category are returned to the one inwhich they were placed by Regan (1920), then the entire Asprotilapia subassemblage is of theHaplochromis type, and the Ophthalmotilapia subassemblage becomes exclusively of theTilapia type. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 275 Such a pattern might well be taken to argue against my earlier criticisms of the pharyngealapophysis as an indicator of phyletic relationships at a high level of universality(Greenwood, 1978) and that, on the contrary, African cichlids can be divided into 'Tilapicfand 'Haplochromis" 1 supralineages as suggested by Regan (1920). If the latter argument is accepted, then the synapomorphies delimiting the Ophthalmo-tilapia assemblage as a whole must be considered as homoplasies developed independentlyin the two subassemblages recognised here. That assumption would produce a scheme ofrelationships less parsimonious than the one proposed above and earlier by Liem(1981). Also, the existence of intermediate conditions (i.e. the Tropheus type) in thestructure of the apophysis, and of other cases where a classification based on apophysealstructure is incongruent with different and apparently synapomorphic characters (seeGreenwood, 1978; Liem & Stewart, 1976) would seem to support the rejection of apophysealstructure as an indicator of phyletic relationships at the level proposed by Regan. A review of other schemes of relationship suggested for members of theOphthalmotilapia assemblage Regan (1920 : 52) did not present detailed arguments for his views on the interrelationshipsof the Lake Tanganyika genera, which were strongly influenced by his assumption of therebeing a fundamental dichotomy of African cichlids into those with a 'TilapicC typeapophysis, and those with a 'Haplochromis'' type (see above). As a result of this basicdifference in approach, it is difficult to make direct comparisons between Regan's ideas andthose put forward in this paper. However, some comments can be made on certain of Regan'ssuggested relationships involving OA members and taxa outside that assemblage. For example, Ophthalmotilapia was grouped with Cyathopharynx, Cunningtonia,Asprotilapia, Petrochromis and the Malawian genus Petrotilapia (then considered a speciesof Petrochromis); furthermore, Regan suggested that Ophthalmotilapia was closely related toLimnotilapia (now synonymised with Simochromis, see Greenwood, 1979) and that it had'. . . given rise to Cyathopharynx'. Petrochromis and Petrotilapia were included in this grouping because, in their dentalmorphology and pattern, they are strikingly similar to Cunningtonia (see p. 280 below). No reasons were given for including Asprotilapia, a taxon quite unlike the others in itsgross morphology and in its dentition; presumably the reason lay in Regan's (1920 : 42)belief that the 'Skeleton (is) essentially similar to that of Ophthalmotilapia ventralis . . .'. Theskeletal features noted by Regan (1920 : 41 & 42) were, it should be emphasised, not thoseused in this paper; most can be treated as plesiomorph characters when used at the level ofanalysis involved here. Presumably it was the same suite of skeletal characters which led Regan to suggest a closerelationship between Limnotilapia (i.e. Simochromis) and Ophthalmotilapia, a relationshipwhich I cannot accept since ''Limnotilapia'' apparently shares no derived features with anymembers of the OA.. Ectodus (as a putative ancestral morphotype) was grouped by Regan (1920: 53) withCallochromis, Xenotilapia and Grammatotria,\hQ taxa being given that order of increasingmorphological derivation. Again no detailed reasons are given for this grouping, save thatall its taxa have a 'Haplochromis'' type pharyngeal apophysis and small conical teeth. Ineffect, however, it approximates closely to the arrangement proposed in this paper. Regan (1920 : 53) also included, albeit implicitly rather than explicitly, Aulonocranus andTrematocara with those genera listed in the previous paragraph. The association ofAulonocranus with Trematocara was, presumably, based on both genera having hyper-trophied cephalic laterosensory canal systems (Regan, 1920:47); no other relationshipwith Aulonocranus was suggested, save that it is 'Intermediate between Haplochromis andTrematocara.'' The linking of Aulonocranus and Trematocara with the Ectodus-Grammatotria group (see above) was apparently based on the common possession of smallconical teeth and a 'Haplochromis' type of pharyngeal apophysis. 276 P. H. GREENWOOD That Regan did not consider there to be any relationship between his Asprotilapia-Ophthalmotilapia and his Ectodus-Grammatotria ( + Aulonocranus) groups is doubtless dueto his basic assumption that the endemic genera of Lake Tanganyika were derived from'. . .two ancestral types, one nearly related to Limnotilapia and the other to Haplochromis'(Regan, 1920:53). After Regan's initial analysis of the Tanganyika cichlids, no further attempt to interrelatethe endemic genera of the lake was made for more than fifty years. In 1972 Fryer & lies paidconsiderable attention to this problem, in particular to the assumption that there was a basicdiphyletic origin of the flock. However, despite their professed uncertainty about the value ofthe pharyngeal apophysis as an indicator of phyletic relationships, Fryer & lies (1972 : 506,fig. 337) virtually followed Regan's (1920) scheme. They were, of course, able to includethree genera described since that time, namely Lestradea, Cardiopharynx and Ophthalmo-chromis. Cardiopharynx is now considered a synonym of Cyathopharynx (see p. 282)and Ophthalmochromis was synonymised with Ophthalmotilapia by Liem (1981: 210-21 1). As would be expected, Fryer & lies considered Ophthalmochromis and Cardiopharynx tobe the sister taxa of Ophthalmotilapia and Cardiopharynx respectively. Interestingly, theyassociated Lestradea with Asprotilapia, Cunningtonia and the Ophthalmotilapia-Ophthalmochromis pair, but gave no reasons for doing so. Their tentative alliance ofCyathopharynx -{-Cardiopharynx with Lobochilotes and Limnotilapia is not explainedeither, and only partly follows Regan who implied some relationship between Limnotilapiaand Lobochilotes, but also included Gephyrochromis, Simochromis and Tropheus in thesame group -again without a detailed explanation (Regan 1920:52). I can find nosynapomorphic characters to support the idea of a close relationship between Lobochilotesand any member of the Ophthalmotilapia assemblage. Like Regan, Fryer & lies (1972 : fig. 337) treat Aulonocranus and Trematocara as closerelatives, but give the two genera an origin separate from that of the taxa currently groupedin the 0A. Leptochromis (now renamed Reganochromis, see Whitley, 1928), a genus not mentionedin Regan's analysis, is included by Fryer & lies as a member of their Ectodus, Callochromis,Xenotilapia lineage, but again no reasons are given. This suggested relationship is discussedon p. 278 below. In effect, the main difference between the schemes proposed by Regan (1920) and Fryer &lies (1972) lies in the latter authors not portraying any taxa in an ancestor-descendentrelationship, as was implied, or stated explicitly, in Regan's treatment. Also, Fryer & liesindicate a more distant relationship than did Regan between Limnotilapia (i.e.Simochromis) and other members of the latter author's Ophthalmotilapia group (see above,p. 275). In their final analysis Fryer & lies are less definite in their suggested relationships than wasRegan (see figure 337 in Fryer & lies, 1972 : 507); their phylogram was to be '. . . regarded asextremely tentative'. Liem's (1981) wide ranging review of the 0A is, in its treatment of anatomical andmorphological detail, far more thorough than either of the other two reviews. It was also thefirst to employ a basically cladistic (sensu Hennig, 1966, phylogenetic sensu Wiley, 1981)methodology. Liem brought together certain taxa from Regan's two major groups (see above p. 275),namely Ectodus with some elements of Regan's Ophthalmotilapia- Asprotilapia group, butexcluded other taxa from his Ectodus-Grammatotria assemblage (see p. 275 above). The present analysis (also cladistically based) finally brings together, in a single lineage, allbut one pair of taxa from Regan's two groups, the exceptions being Petrochromis and theMalawian genus Petrotilapia (see above, p. 275). It also includes Aulonocranus (but notTrematocara) from a third group which Regan implied had some relationship with hisCallochromis-Grammatotria lineage (Regan, 1920 : 53). Effectively it hypothesizes that thetwo major Regan groups are sister lineages within a larger taxon which, following Liem(1981), can be named, informally, the Ophthalmotilapia assemblage (see Fig. 1 4). Asprotilapia SA ..Ophthalmotilapia SA. (U 1-5 Fig. 14 Cladogram for the Ophthalmotilapia assemblage based on the 10 apomorphic charactersdiscussed in this paper. Page numbers, given in brackets after each numbered character, refer tothose pages on which the character is discussed in detail.An interrogation mark precedes character 9 since its apomorphic status is doubtful (see p. 272). ( 1 ) Palatopterygoid gap (p. 254) (2) Auricular process on the operculum(P. 259) (3) Morphological features of the palatinebone (p. 257) (4) Outline shape of the lachrymal (1st infra-orbital bone), and the presence of sixlaterosensory canal pores (p. 259) (5) Adductor fossa on the lateral aspect of theanguloarticularbone (p. 261) (6) Intestine long and transversely coiledp. 264) (7) Presence of a pharyngeal hanging pad andassociated modifications to the gill-rakermorphology (p. 265) (8) Dentary with a distinct 'step' (p. 267) (9) First branched pelvic fin ray produced(P. 272) (10) Dorsal part of the flange behind the verticalpart of the preopercular laterosensorycanal not narrowing abruptly (p. 273) 278 P. H. GREENWOOD Summary of the taxonomic conclusions and a discussion of the sister-group problem in these and other lake cichlids The Ophthalmotilapia assemblage, originally comprising the genera Ectodus, Lestradea,Asprotilapia, Cunningtonia and Ophthalmotilapia (Liem, 1981) can now be expanded toinclude Xenotilapia, Callochromis, Grammatotria, Aulonocranus and Cyathopharynx (withwhich is synonymized Cardiopharynx). Taxonomically, the assemblage is recognized as a monophyletic lineage on the grounds ofits member species sharing a unique congruence of five apomorphic characters (pages254-262; Fig. 14). Two major sublineages can be recognized within the assemblage. One, characterized by having a short and simple coiled intestine, comprises the generaEctodus, Aulonocranus, Asprotilapia, Xenochromis, Callochromis and Grammatotria. It canbe further subdivided on the basis of various synapomorphies shared by four of its members(see pages 263-274 and Fig. 14). The other major sublineage contains Lestradea, Ophthalmotilapia, Cunningtonia andCyathopharynx, species in which the intestine is long and complexly coiled (see pages264-265). Attempts to subdivide this lineage have not proved satisfactory, but it would seemthat Lestradea is the plesiomorph sister-group of the other three genera (pages 271-274 andFig. 14); possibly amongst these three genera Ophthalmotilapia and Cunningtonia are sistertaxa, but the entire group is, for the moment, probably best treated as an unresolvedpolychotomy. No sister-group, or even a single taxon, has so far been satisfactorily identified amongst thecichlids of Lake Tanganyika. Regan's (1920 : 52) suggestion of close relationship betweenLimnotilapia (i.e. Simochromis, see Greenwood, 1979) and Ophthalmotilapia (and henceCyathopharynx, Asprotilapia and Cunningtonia as well) cannot be corroborated on the basisof shared synapomorphies. When comparisons are made with other Tanganyika taxa the results are equallyunproductive except that two OA group synapomorphies occur, singly, in some species of' Limnochromis* and Trematocara. For example, in Trematocara marginata, but in no other species, there is a poorlydifferentiated OA-like projection from the anteroventral angle of the lachrymal (see p. 259).A similar projection occurs in 'Limnochromis'' permaxillaris and 'L.' pfefferi, both of whichare now placed in Poll's genus Gnathochromis. Both the ' 'Limnochromis* species andTrematocara marginata have only 5 pores in the lachrymal, but other 'Limnochromis'species which lack the projection, V otostigma, L. auritus and L. abeeli, have 5 or 6 pores. Neither Trematocara nor ' ' Limnochromis'' has the characteristically shaped lachrymal ofthe OA species, and the distribution pattern of the two OA group-features indicates anindependent (i.e. homoplastic) origin in the two genera. Regan's (1920:53) suggested relationship between Aulonocranus and Trematocaraapparently stems from the hypertrophied laterosensory canals, and pores, present in bothtaxa. Apart from that feature, and the weak OA-type lachrymal peak in one species ofTrematocara, there are no derived features uniquely shared by the two genera. Admittedlythere is a trend amongst the OA species for there to be some degree of hypertrophy in thelaterosensory canal system of the lachrymal. But, in the absence of other synapomorphies itwould be unrealistic to use a trend character as a basis for suggesting group relationships,especially when that trend occurs in several other lineages as well. In this particular instance,too, the lowest degree of canal enlargement is found in those OA species with the leastnumber of derived features (i.e. Ectodus and Lestradea), and which are therefore taken to bethe plesiomorph members of their respective subassemblages. The tentative phylogenetic schemes proposed by Fryer & lies (1972 : 507, fig. 337) are notalways arranged so as to suggest sister-group relationships with taxa outside the OA. Theydo, however, show Leptochromis (i.e. Reganochromis) as a sister taxon of Ectodus,Callochromis, Xenotilapia and Grammatotria, and Aulonocranus is paired with THE OPHTHALMOT1LAPIA ASSEMBLAGE OF CICHLID FISHES 279 Trematocara. Once again, there are no shared derived features to substantiate such relation-ships, and none which might indicate that either Reganochromis or Trematocara is thesister-group, or part of the sister-group, to the Ophthalmotilapia assemblage. The only taxon which consistently shows more than a single OA group synapomorphyamongst a number of its species is the Malawian genus Lethrinops. I have examined alizarin preparations and dry skeletons of five Lethrinops species, thetype species Lethrinops lethrinus, and L. praeorbitalis, L. parvidens, L. auritus and L.longimanus. In all there is a distinct palatopterygoid gap and a well-defined, although notextensive OA type adductor fossa on the anguloarticular bone (Figs 1 5 & 16). The palatopterygoid gap is relatively smaller than in most OA species, but in none of theLethrinops species examined is there any contact between the palatine and theentopterygoid. Unlike members of the OA, the Lethrinops species have a much deeperentopterygoid, and one that either rests along the upper margin of the quadrate or slightlyoverlaps that bone medially. In OA species, most of the entopterygoid lies medial to thequadrate, and is thus largely obscured by it in lateral view. As a consequence of this spatialrelationship the dorsal margin of the entopterygoid in Lethrinops lies at a level nearer thepalatine head than it does in OA species. * mm Fig. 15 Left suspensorium of Lethrinops lethrinus. The shape of the palatine (Fig. 15) in Lethrinops differs somewhat from that in the OAtaxa(see p. 257 above). Its posterior margin is slightly concave, and the angle between thismargin and the head of the bone is less nearly rectangular; the posterodorsal margincontributing to the angle is also less acute in Lethrinops; indeed, in some specimens andspecies it is almost rounded. But, as in the OA species the body of the bone is expandedposteriorly so that the bone's proportions are nearer those of the OA type than thatcommonly found amongst African cichlids. The occurrence of this particular palatine shape in association with a palatopterygoid gapraises the question of whether or not the two characters are correlated. That a similarlyshaped palatine does occur in at least two species without a palatopterygoid gap (viz.Limnochromis abeeli [Lake Tanganyika] and Astatotilapia macropsoides [Lakes Edwardand George]), would seem to argue against correlation, but the possibility requires furthertesting. The adductor fossa in Lethrinops (Fig. 16) is well defined but, as compared with the fossain members of the Asprotilapia subassemblage amongst the OA (p. 261) it is less extensive. Itis, however, comparable with the fossa in members of the Ophthalmotilapia subassemblage(p. 265). 280 P. H. GREENWOOD ommFig. 16 Left dentary and anguloarticular ofLethrinops lethrinus, lateral view. An adductor fossa of this type is not restricted to members of the OA, but also occurs inseveral seemingly unrelated taxa, including some 'Haplochromis* species from Malawi (seep. 261). Thus, in itself, the fossa cannot be considered a unique apomorphy; its value as agroup synapomorphy stems solely from its congruence with other apomorphic characters. There is a third derived character found in the five Lethrinops species which is also presentin one subgroup of the OA, namely a stepped dorsal margin to the dentary, with the teethconfined to the higher level of the step (Fig. 16). Amongst the OA taxa this feature is foundonly in Xenotilapia, Callochromis and Grammatotria (all members of the Asprotilapiasubassemblage; see p. 267). Its restricted distribution within the OA considerably reduces itspotential significance as a character indicative of a possible relationship between Lethrinopsand the OA. That it apparently occurs only in these OA species and in Lethrinops, and thatboth groups have a palatopterygoid gap is, nevertheless, intriguing and requires furtherinvestigation. Attempts to evaluate the two apomorphic features shared by Lethrinops and the OA in itsentirety (i.e. the palatopterygoid gap and the adductor fossa) are hampered by lack ofcomparative data from the Malawi cichlids as a whole. A relatively superficial survey shows,however, that the adductor fossa is present only in some haplochromine species (see p. 26 1 ).Whether or not these haplochromines are closely related to Lethrinops has not beenadequately tested, but preliminary investigations do not suggest that this is the case. The situation regarding the palatopterygoid gap is different. I have examined specimens ofall the described genera of Malawi fishes, but by no means all their included species. In noneis there a palatopterygoid gap. Thus, the character appears restricted to Lethrinops and theOA (see also p. 255). In summary, it seems that no group of Lake Tanganyika cichlids consistently shows one ormore of the group synapomorphies for the OA. On the other hand, in Lake Malawi at leastsome species of Lethrinops share two derived features with all members of the OA, andonly one of these features (the adductor fossa) occurs in other Malawi taxa. The Lethrinopsspecies also have a third derived feature, the shape of the dentary, which is present in onelineage amongst the OA. As it stands, this indication of a possible sister-group relationship between the OA andLethrinops is no more than suggestive. It could well be refuted as an example of convergencewhen more is known about the interrelationships of the cichlids from Lake Malawi and thoseof Lake Tanganyika. There are several cases of close similarity existing between certain features in endemic taxaof each lake, but few between species in those lakes and endemic taxa from Lake Victoria. As examples one may mention that the dentition in Cunningtonia is virtually identicalwith that in Petrotilapia and there are marked similarities in the jaw anatomy of the twogenera (Petrochromis from Lake Tanganyika can also be included in this example); or, themorphology of the dentary in Asprotilapia which is like that of Labeotropheus a Malawiangenus also sharing several features with Tropheus from Lake Tanganyika. Finally, one cancite the many similarities in syncranial architecture existing between Simochromis ofTanganyika and the Pseudotropheus species complex of Lake Malawi. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 281 These are by no means the only cases that can be, or have been, cited of supposedconvergence or parallelism between the cichlid faunas of the two lakes (see Fryer & lies,1 972). Adequate explanations for these similarities are, however, far more difficult to find. An example of similarity involving much greater geographical separation than thatbetween Tanganyika and Malawi, and one which incorporates a greater number of speciesas well as a mosaic distribution of similar characters amongst the species, involves twomembers of the OA, another, unrelated, genus from Lake Tanganyika, a Malawian taxonand Neopharynx schwetzi, a monotypic genus from the lower Fwa (Kasai drainage in south-western Zaire; for a full description of N. schwetzi, see Poll, 1948). >mm Fig. 17 Lower pharyngeal bone of Neopharynx schwetzi, occlusal view (from a specimen in paratypical series, MRAC 7 1 29 1-7 1 299). Neopharynx schwetzi has a lower pharyngeal bone morphologically and dentally almostidentical with that of Cyathopharynx (see Figs 17 & 24). Its oral dentition, in contrast, isvirtually identical with that in Petrochromis (Tanganyika) and Petrotilapia (Malawi), and isquite unlike that in Cyathopharynx. As far as the morphology of the teeth is concerned, butnot their distribution on the dentary, Neopharynx also closely resembles Cunningtonia ofLake Tanganyika. It differs from all three taxa in the shape of its premaxilla, but themorphology of the premaxillary teeth is, as might be expected, very similar in all four genera. The relationships of Neopharynx have yet to be established; probably they lie with twoother Fwa endemics, Cyclopharynx and Callopharynx, genera having an extremedevelopment of the Neopharynx-Cyathopharynx type of lower pharyngeal bone (for detailssee Poll, 1948). Neopharynx has neither an OA type lachrymal, a palatopterygoid gap, nor an OA-type ofadductor fossa, and the shape of its palatine bone is close to the generalized form. Thus thesimilarities between Neopharynx, Cyathopharynx and Cunningtonia are undoubtedlyhomoplastic, as most probably are the similarities shared with Petrochromis andPetrotilapia. The Neopharynx example underlines the problems involved in attempting to work outinterrelationships amongst cichlid fishes, as do the repeated appearances of certain derivedfeatures shown by members of the Ophthalmotilapia assemblage in species which appear tobe but distantly related to the OA. Surmises about the interrelationships of African lake cichlids have, I believe, been undulyinfluenced, perhaps even inhibited, by three major factors. Firstly, the idea that there are two 282 P. H. GREENWOOD basic lineages, a ' Tilapia" line and a 'Haplochromis' one (see discussion in Greenwood, 1 978;also p. 274 above). Secondly, that the major lakes are, faunistically, closed basins and havebeen so almost since their inception, with the result that the cichlids of a lake are presumedto have evolved from one or a few ancestral species originally trapped there (see discussion inFryer & lies, 1972; and Greenwood, 1974). In other words, assumed histories for the lakeshave been given too great a weight in deciding whether a feature was the result ofconvergence, parallelism, or the consequence of common ancestry. The third, and overriding, factor is a paucity of specific and critical studies on thephylogeny of the fishes. Overall resemblances, or the use of characters without adequateoutgroup comparisons, are often major weaknesses influencing decisions on relationships. The effect of these three factors has resulted in a tendency to restrict the search for sister-groups to a single lake and to the appropriate 'Tilapia' or 'Haplochromis' lineage.Admittedly, at lower levels of universality sister species are generally to be found within thesame lake or proto-lake system (Greenwood, 1980); the problems arise when attempting toestablish relationships at somewhat higher taxonomic levels. The Ophthalmotilapiaassemblage, and the search for its sister-group are good examples of problems encountered atthese two levels. The existence of endemic species flocks, the superficially close similarity of species indifferent flocks, and indeed of many different taxa outside the lakes, all help to complicatethe issue. Under such circumstances the possibility and probability of homoplasy aretheoretically enhanced, as in practice are the problems associated with their resolution. What is needed to resolve these problems are tests of the assumption that a so-calledspecies flock is really of monophyletic origin. If a lake's cichlid fauna was derived from a fewspecies which are not true sister species, and if the true sister taxa were the ancestors ofanother flock, then the situation suggested by the apparent relationship of Lethrinops(Malawi) and the 0A (Tanganyika) could well be a real one. The idea first put forward byRegan (1922) that the Malawi 'flock' carries indicators of its monophyly must be seriouslyquestioned (Greenwood, in press). No indicators of monophyly have been suggested forthe Tanganyika 'flock' (even at the levels of the supposed 'Tilapia' and "Haplochromis" typebasic stocks). The possibility of a close relationship between the faunas of the two lakesdeserves very careful examination. The status of Cyathopharynx Regan, 1920 and Cardiopharynx Poll, 1942 In his original description of the monotypic genus Cardiopharynx, Poll (1942 : 346) notedthe great similarity between its peculiarly shaped lower pharyngeal bone and the lowerpharyngeal of Cyathopharynx. He differentiated the two genera because Cardiopharynx has,as compared with Cyathopharynx: (i) Larger scales (36-38 c/48-64 in a longitudinal series (see Poll, 1956:1 27). (ii) Jaw teeth in two rows, the teeth, in both jaws, of equal size (c/3-5 rows; teeth in the outer row larger than those of the inner rows),(iii) Dentigerous surface of the lower pharyngeal bone cardiform (cf rounded in Cyathopharynx)(iv) Supraoccipital extending forward to a level above the anterior margin of the orbit (c/to the mid-orbital region only)(v) Parietal crests ending above the centre of the orbit (c/extending to a point above the posterior part of the orbit)(vi) More vertebrae (36, i.e. 17+19) c/32-34 (i.e. 1 6 or 1 7 + 1 6 or 1 7) Poll gives no reasons why these characters should be used to separate the taxa at a genericlevel, and neither does he indicate why the great similarity in pharyngeal morphology, andits uniqueness, should be outweighed as an indicator of close phyletic relationship by thediagnostic characters he enumerates. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 283 None of these latter features can be considered uniquely apomorphic for Cardiopharynx,and now that more material is available several are found to be less trenchant than was firstthought to be the case, as the following comments show. The teeth in both jaws of all Cardiopharynx specimens I examined are invariably arrangedin two rows, but in Cyathopharynx there is a greater variation than was intimated by Poll(1942 & 1956). In the majority of specimens examined, the inner premaxillary row is, inplaces, irregularly arranged so as to give the appearance of a double row wherever theirregularities occur. Occasionally there are specimens in which the inner tooth row is clearlyand regularly double, thus giving a total count of three tooth rows; also occasionally it isdistinctly single, giving a total of two rows. In none of the specimens is there a total of morethan three premaxillary rows. The inner row of teeth in the dentary is generally single; in afew fishes, however, it is somewhat irregular and so comes to resemble the modal conditionin the premaxilla. Contrary to Poll (1946), I can find no marked difference between the taxa in the relativesize of inner and outer teeth. In both genera the outer teeth, in both jaws, are clearly tallerand stouter than those of the inner row or rows, and not of equal size in Cardiopharynx asclaimed by Poll. The most that can be said is that in Cardiopharynx the size differencebetween outer and inner row teeth is a little less marked than in Cyathopharynx. In a later redescription of both taxa, Poll (1956) commented on the outer teeth ofCardiopharynx being more or less tricuspid in young fishes, but he gave no size-range overwhich tricuspid teeth are found. In specimens 60-105 mm SL I have examined, there are noouter tricuspids, whereas in specimens of Cyathopharynx of a comparable size range and upto 1 12 mm SL, distinctly, and also weakly, tricuspid teeth do occur. In larger Cyathopharynxspecimens the teeth are exclusively and clearly unicuspid. The difference in the shape of the dentigerous surface of the lower pharyngeal bone (iiiabove, p. 282) is due entirely to a marked median depression in the posterior face of the boneand of the toothed area in Cardiopharynx. In Cyathopharynx this margin of the bone isslightly and more broadly indented, and there is no indentation of the posterior tooth row(Fig. 23). In all other respects the lower pharyngeal bone in both taxa is identical. As noted above (p.27 1 ) it represents a uniquely derived condition amongst the cichlids of Lake Tanganyika. The shape of the pharyngeal teeth, tall and slender, with spatulate crowns that are notbroader than the neck, and their cardiform pattern on the alveolar surface of the bone, is thesame in both genera (Fig. 24). Likewise, the morphology of the principal upper pharyngealbones is identical, as is the morphology and pattern of their teeth. Unlike the lower teeth,those on the upper pharyngeal bones do have a slight, shoulder-like cusp at the base of thespatulate crown surface. The anterior point reached by the supraoccipital and parietal crests (iv & v above) isvariable intragenerically. In the skulls I have examined there are specimens from each genusshowing conditions intermediate between those originally used as diagnostic features for thetwo genera. Finally, there are the supposedly intergeneric differences in vertebral number (vi above). Ihave examined 15 specimens of Cardiopharynx schoutedeni (13 from radiographs, 2 asalizarin preparations), and 12 of Cyathopharynx furcifer (10 from radiographs and 2 dryskeletons), and obtained the following counts: Cardiopharynx: Total number (excluding the fused Ui-PUi centra) 33 (f5) and 34(flO), comprising 1 5 (fl), 16 (f7) or 17 (f7) abdominal and 16 (fl), 17(f9) or 18 (f5) caudalelements. Cyathopharynx: Total number (excluding the fused Ui-PUi centra) 32 (f5) 33 (f6) or 34(fl ), comprising 1 6 (f9) or 1 7 (D) abdominal and 1 6 (f7) or 1 7 (f5) caudal elements. The differences in range are not very marked and there is a complete overlap in othercounts, but with a slight difference in the modal number for total [34 c/33] and abdominalcounts [1 7 cf\ 6] for Cardiopharynx and Cyathopharynx respectively. Thus, of Poll's (1942) original diagnostic features, only the difference in scale size remains. 284 P. H. GREENWOOD There are, however, other differences which were noted in the original diagnosis.Cyathopharynx has proportionately longer pelvic fins, in males the tip of this fin reaches thelast anal fin ray or even to as far as the caudal fin fork; in Cardiopharynx it reaches only toabout the middle of the anal fin. Also, in Cyathopharynx the first pelvic ray is clearly thelongest whereas in Cardiopharynx either the first and second rays are equally protracted orthe second ray may be the longest. Other differences involve neurocranial shape (Figs 18 & 19). Cardiopharynx has ashallower skull than does Cyathopharynx, the prootic portion of its otic bulla is moreinflated, and the interorbital region is much narrower. Undoubtedly the two taxa are distinguishable. The problem is to decide at whattaxonomic level their separation should be recognized. To recognize two genera on the basis of the differences discussed above is to obscure thefact that, amongst the Lake Tanganyika cichlids, Cyathopharynx and Cardiopharynx share aunique apomorphy (the form of the lower pharyngeal bone) which would indicate a commonancestry not shared with any other taxon. That relationship is, I believe, best indicated bytreating the two species as members of a single genus, Cyathopharynx Regan, 1920. Cyathopharynx Regan, 1 920 Cyathopharynx Regan, 1920. Ann. Mag. nat. Hist. (9), 5: 42-43.Cardiopharynx Poll, 1942. Revue Zool. Bot. afr., 36: 346-347. TYPE SPECIES. Tilapia grandoculis Boulenger, 1899. Trans, zool. Soc. Lond. 15: 94, pi. XIX,fig. 6. Poll (1946 : 283^) has synonymized this species with C.furcifer (Blgr), 1 898. DIAGNOSIS. A member of the Ophthalmotilapia assemblage, distinguished from othermembers of that group by having a cardiform dentigerous surface to the lower pharyngealbone, the body of which is inflated and nearly cardiform in outline. Cyathopharynx is distinguished from other species with a cardiform alveolar surface to thelower pharyngeal bone by, among other features, its <9A group characters (see p. 262) and bythe marked elongation of the first, or first and second pelvic fin rays in adult male fishes;these protracted rays extend to at least the middle of the anal fin, and in one species, some-times as far as the fork of the caudal fin. Description Neurocranium (Figs 1 8 & 1 9). In its general outline, the skull differs little from that of thegeneralized haplochromine type in which the preorbital profile is slightly decurved. Theorbit, however, is relatively larger than in a generalized skull, and the otico-occipital regionis shorter, a correlate, probably, of the enlarged orbit since the ethmoid region retains thesame proportions as in a generalized skull. The ventral apophysis for the upper pharyngeal bones is of a weak 'Tropheus 1 type, inwhich the basioccipital barely contributes to the articular surface (see Greenwood, 1978; alsop. 274 above). Suspensorium (Figs 1 C & D). As in other members of the Ophthalmotilapia assemblage,there is a distinct palatopterygoid gap (see p. 255), and the entopterygoid is shallow, withonly one-third to one-quarter of its depth visible above the quadrate margin. Thehyomadibula has a narrow flange anterior and dorsal to the symplectic process. The shapeand proportions of the symplectic are typically those of an OA species (see p. 256). Infraorbital series (Figs 3 A & F). The lachrymal bone (1st infraorbital) has the typicaloutline shape, and the anteroventral process, of an OA taxon (see p. 259). There are sixpores opening from the laterosensory canal system, the tubular part of which is somewhat THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 285 5mm Fig. 18 Neurocranium of Cyathopharynx furcifer in : A, left lateral view; B, dorsal view. inflated. The pores, however, do not show a corresponding enlargement (i.e. they do notdeviate noticeably from the generalized condition). The other infraorbital bones are littlemore than tubular ossifications around the sensory canal, but do have low dorsal and ventralkeels. Myology. Division I of the adductor mandibulae complex has an extensive origin alongthe vertical limb of the preoperculum, but division II has its preopercular origin mainly fromthe horizontal limb of that bone; only a small area extends onto the vertical limb. The dorsal gill-arch muscles compare closely, in most respects, with those of other 0Aspecies (see Liem, 1981 : 196-7 & 205, & fig. 8; also p. 258 above). The retractor dorsalismuscles are especially well developed with, in some individuals of both species, indicationsof a subdivison into dorsal and ventral components. Dentition (Fig. 20). Some aspects of the oral dentition have been commented upon already(p. 283 above). The outer row teeth in both jaws are tall and slender. In C. furcifer the crownis slightly broader than the neck and shaft of the tooth, whereas in C. schoutedeni the crownis no wider than the shaft. The crowns are slightly recurved in both species. Premaxillary outer row teeth are aligned vertically to the alveolar surface, but in thedentary the teeth situated anteriorly and anterolaterally are procumbent. The posteriordentary teeth are vertical, those of C. furcifer continuing for some distance up the coronoidprocess. Inner row teeth in both jaws, and both species, are noticeably smaller than those of theouter row, and are implanted so as to lie almost horizontally. Mouth. The lips are thin, and the gape is horizontal. 286 P. H. GREENWOOD 5mm Fig. 19 Neurocranium of Cyathopharynx schoutedeni in : A, left lateral view; B, dorsal view. mm Fig. 20 Outer row jaw teeth (drawn in situ) from : A & B, Cyathopharynx schoutedeni, specimen103 mm SL (premaxillae and dentary respectively); C & D, C. furcifer, specimen 1 13 mm SL(premaxilla and dentary, respectively). Dentary teeth viewed from below. Jaws. The dentary in both species is a moderately slender bone, shallower and less robust inC. schoutedeni than in C. furcifer (Fig. 2 1 ). Anteriorly and anterolaterally the dorsal aspect ofthe bone is expanded into a broad surface which extends outwards to overhang, as a narrowshelf, the body of the bone. The tooth rows occupy only the outermost part of the surface,with the result that there is a wide expanse of bone lying medial to them. The laterosensorycanal system and its openings in C. furcifer are more cavernous than those in C. schoutedeni.The anguloarticular has a well-defined but short fossa for the adductor mandibulaemuscle, with the ridge delimiting its anterior margin particularly deep and prominent.The premaxilla (Fig. 22) has no especially outstanding features. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 287 Fig. 21 Dentary and anguloarticular, in left lateral and occlusal views respectively, of : A & B,Cyathopharynxfurcifen C & D, C. schoutedeni. 5mm Fig. 22 A & B, premaxilla ofCyathopharynxfurcifer, anterior and lateral view; C, maxilla (left) seen from a slightly ventrolateral viewpoint. 288 P. H. GREENWOOD As compared with the generalized type of maxilla, that in Cyathopharynx is foreshortenedand has a well-developed, long-based posterior process (see also p. 269). The pharyngeal bones. The shape and other peculiar features of the lower pharyngeal bone(Fig. 23) are described on page 283. The upper pharyngeal bones differ less markedly from the usual condition seen in Africancichlids. The outline of the alveolar surface of the major element (pharyngobranchial 3) isnoticeably ovoid, but otherwise differs little from the generalised condition. Their principaldifference lies in the relatively greater alveolar surface area, and its more elongateproportions. Other differences are found in the less prominent facets for articulation with the3rd and 4th epibranchials, and in the lower summit facet (nomenclature following Barel etal, 1976: 214, fig. 26). The lower pharyngeal teeth (Fig. 24) are slender, near cylindrical in cross-section and areclosely packed. Those at the periphery of the dentigerous area are weakly curved, the others i mm Fig. 23 Lower pharyngeal bone, in occlusal view of: A, Cyathopharynx furcifer, B,C. schoutedeni, and, in ventral view, C, of C. furcifer. B I \ 0.25mm Fig. 24 Cyathopharynx furcifer, pharyngeal teeth (lateral and semiocclusal views) from : A, the posterior; and B, anterior dental fields. THE OPHTHALMOTILAPIA ASSEMBLAGE OF CICHLID FISHES 289 erect. There is a noticeable and rapid increase in the height of the teeth forming the posterior5 or 6 rows, with the teeth of the posterior row tallest and stoutest. The elongate crown ofeach tooth is flat, and slopes gently upwards and backwards. There is no indication of a lowshoulder-like projection or cusp at the base of the crown. The shape and dense arrangement of the upper pharyngeal teeth are very similar to thoseof the lower bone, the teeth differing only in having a small shoulder or cusp at the base of theposteriorly directed crown. Vertebrae. Regan (1920 : 43), in his original description of Cyathopharynx, noted that thethird vertebra lacks an inferior apophysis, from which the retractor dorsalis musclesoriginate. In four of the five dry skeletons and alizarin preparations I examined, a lowapophysis is present on the fourth centrum in one fish and on the fifth centrum in threeothers, but none is present in the fifth specimen.Vertebral counts for the two species are given on page 283. Squamation. Scales on the body are weakly ctenoid except for the cycloid scales on the chestand belly. Those covering the thoracic region are small, and are fairly abruptly demarcatedfrom the larger scales on the ventral flanks and the belly. The two Cyathopharynx species differ, disjunctly, in the size of their body scales, withC. schoutedeni having 36-38 scales in a longitudinal series, and C. furcifer 48-64. Thespecies also differ, slightly, in the posterior extension of the upper lateral line pore scales,those in C. furcifer usually extending almost to the caudal fin base, whereas in C. schoutedenithe pore scales terminate at a level 3 or 4 scale rows anterior to the caudal base. There is,however, some interspecific overlap in this feature. Fins. Little can be added to the description (p. 272) of the protracted first or first and secondpelvic fin rays in adult males. In females and juvenile males these rays are also noticeablylonger than the others, but usually do not extend beyond the level of the anus; exceptionallythey may reach the spinous part of that fin. The caudal fin is forked; adult males have the two upper- and lowermost principalbranched rays produced into fine filaments. Rows of small, barely overlapping scales arepresent on the fin membrane (except between the three middle rays), and extend to the levelof the fork. Gut. The intestine is long (ca 3 to 4 times SL) and complexly coiled in a predominantlytransverse direction (see p. 264). Contained species Cyathopharynx furcifer (Blgr) 1898Cyathopharynx schoutedeni (Poll) 1 942 Both are lacustrine species endemic to Lake Tanganyika; for detailed descriptions, figuresand biological data see Poll (1956: 1 30-1 37 & 1 25-1 30 for the species respectively). Acknowledgements Gordon Howes has once again devoted much time and patience to producing the figures, andfor that I am especially grateful. I also owe him a great deal of gratitude for his help with allthe numerous, and monotonous, tasks he has undertaken in helping with the production ofthis paper. I am greatly indebted to Dr Thys van den Audenaerde of the Koninklijk Museum voorMidden-Afrika, Tervuren, who so graciously allowed me to borrow the types and othermaterial of the river Fwa cichlids. Finally, it is a great pleasure to thank my colleague Professor Karel Liem of the MCZwhose original paper on the <9A has proved both a stimulus and a focal point for my studieson the interrelationships of the cichlid species flocks from the African lakes. 290 P. H. GREENWOOD References Barel, C. D. N., Witte, F. & van Oijen, M. J. P. 1976. The shape of the skeletal elements in the head of a generalized Haplochromis species: H. elegans Trewavas 1933 (Pisces, Cichlidae). Neth. J. Zool. 26 (2): 163-265.Boulenger, G. A. 1898. Report on the collection of fishes made by Mr J. E. S. Moore in Lake Tanganyika during his expedition, 1895-96. Trans, zool. Soc. Lond. 15 (1) : 1-30.Brichard, P. 1978. Fishes of Lake Tanganyika. T.F.H. Publications Inc., New Jersey & London.Fryer, G. & lies, T. D, 1972. The cichlid fishes of the Great Lakes of Africa. Their biology and evolution. Oliver & Boyd. Edinburgh.Greenwood, P. H. 1965. Environmental effects on the pharyngeal mill of a cichlid fish, Astatoreochromis alluaudi and their taxonomic implications. Proc. Linn. Soc. Lond. 176: 1-10. 1974. Cichlid fishes of Lake Victoria, East Africa: the biology and evolution of a species flock. Bull. Br. Mus. nat. Hist. (Zool.) Suppl. 6: 1-134. 1978. A review of the pharyngeal apophysis and its significance in the classification of African cichlid fishes. Bull. Br. Mus. nat. Hist. (Zool.) 33: 297-323. 1979. Towards a phyletic classification of the 'genus' Haplochromis (Pisces, Cichlidae) and related taxa. Part I. Bull. Br. Mus. nat. Hist. (Zool.) 35 : 265-322. 1980. Towards a phyletic classification of the 'genus' Haplochromis (Pisces, Cichlidae) and related taxa. Part II: the species from Lakes Victoria, Nabugabo, Edward, George and Kivu. Bull. Br.Mus. nat. Hist. (Zool.) 39 : 1-101. 1981. The haplochromine fishes of the East African Lakes. Kraus-Thomson Organization GmbH. Munich & London.Hennig, W. 1966. Phylogenetic systematics. University of Illinois Press, Urbana.Liem, K. F. 1973. Evolutionary strategies and morphological innovations: cichlid pharyngeal jaws. Syst. Zool. 22 (4): 42 5-441. 1981. A phyletic study of the Lake Tanganyika cichlid genera Asprotilapia, Ectodus, Lestradea, Cunningtonia, Ophthalmochromis, and Ophthalmotilapia. Bull. Mus. comp. Zool. Harv.149(3): 191-214. & Osse, J. W. M. 1975. Biological versatility, evolution and food resource exploitation in African cichlid fishes. Am. Zool. 15 (2) : 427-454. & Stewart, D. J. 1976. Evolution of the scale-eating cichlid fishes of Lake Tanganyika: a generic revision with a description of a new species. Bull. Mus. comp. Zool. Harv. 147 (7) : 3 19-350.Poll, M. 1942. Cichlidae nouveaux du Lac Tanganyika appartenant aux collections du Musee duCongo. Revue Zool. Bot. afr. 36 (4) : 343-360. 1946. Revision de la faune ichthyologique du lac Tanganyika. Annls. Mus. Congo beige C. Zool.Serl,4(l): 141-364. 1948. Descriptions de Cichlidae nouveaux recueillis par le Dr. J. Schwetz dans la riviere Fwa (Congo beige). Revue Zool. Bot. afr. 41 : 91-104. 1956. Poissons Cichlidae. Result, sclent. Explor. hydrobiol. lac Tanganyika (1946-1947), 3 fasc.Sb: 1-619. 1981. Contribution a la faune ichthyologique du lac Tanganyika. Revision du genre Limnochromis Regan 1920. Description de trois genres nouveaux et d'une espece nouvelle: Cyprichromis brieni. Annals Soc. r. zool. Belg. Ill: 163-179.Regan, C. T. 1920. The classification of the fishes of the family Cichlidae-I. The Tanganyika genera. Ann. Mag. nat. Hist. (9) 5 : 33-53. 1922. The cichlid fishes of Lake Nyassa. Proc. zool. Soc. Lond. 1921 : 675-727. Stiassny, M. L. J. 1981. Phylogenetic versus convergent relationship between piscivorous cichlid fishes from Lakes Malawi and Tanganyika. Bull. Br. Mus. nat. Hist. (Zool.) 40 : 67-101.Trewavas, E. 1973. On the cichlid fishes of the genus Pelmatochromis with a proposal of a new genus for P. congicus; on the relationship between Pelmatochromis and Tilapia and the recognition of Sarotherodon as a distinct genus. Bull. Br. Mus. nat. Hist. (Zool.) 25 : 1-26.~^==-4574. The freshwater fishes of rivers Mungo and Meme and Lakes Kotto, Mboandong and Soden, West Cameroon. Bull. Br. Mus. nat. Hist. (Zool.) 26 : 299^1 9.Whitley, G. P. 1*928. Studies in ichthyology. No. 3. Rec. Aust. Mus. 17 : 101-143.Wiley, E. O. \9jH?Phylogenetics. John Wiley & Sons. New York. Manuscript accepted for publication 9 September 1982 East African Cichlid Fishes The haplochromine fishes of the east African lakesP. H. Greenwood, British Museum (Natural History) May 1981, 840 pp, 3 plates, 350 figures Hardback Price 25.00 This volume brings together, for the first time, Dr. Greenwood's various papers (somenow out of print) on the taxonomy and biology of the species of haplochromine fishesfrom Lake Victoria, east Africa. Also reprinted are his papers on the haplochrominesfrom Lakes Nabugabo, George and Turkana (Rudolph), two papers dealing withthe classification of the genus Haplochromis, and a recent essay on the explosiveevolution of cichlid fishes in Africa. An index to the 200 species dealt with in this book (and their current generic place-ment), and a general introduction to the evolutionary and taxonomic problems posedby these biologically complex and fascinating fishes, are included. Published jointly by British Museum (Natural History)(exclusive U.K. rights) and Kraus-Thomson Organization Ltd.(all other rights). ISBN 3 601 00438 6 Publications SalesBritish Museum (Natural History) Cromwell Road London SW7 5BD England Titles to be published in Volume 44 Observations on the systematics of the genus Diffiugia inBritain (Rhizopoda, Protozoa). By Colin G. Ogden Miscellanea A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin Curds & Irene C. H. Wu Osteology, genitalia and relationships of the Acanthodactylus(Reptilia: Lacertidae). By E. N. Arnold The Opthalmotilapia assemblage of cichlid fishes reconsidered. By Peter Humphry Greenwood Morphological studies on some Difflugiidae from Yugoslavia(Rhizopoda, Protozoa). By Colin G. Ogden & Andjelija Zivkovic Printed by Henry Ling Ltd, Dorchester Osteology, genitalia and the relationshipsof Acanthodactylus (Reptilia: Lacertidae) E. N. Arnold Zoology series Vol 44 No 5 26 May 1983 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, andan 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 SW7 5BD,England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Zool.) Trustees of the British Museum (Natural History), 1983 The Zoology Series is edited in the Museum's Department of ZoologyKeeper of Zoology : Dr J. G. ShealsEditor of Bulletin : Dr C. R. CurdsAssistant Editor : Mr C. G. Ogden ISSN 0007-1498 Zoology series Vol 44 No 5 pp 291-339British Museum (Natural History)Cromwell RoadLondon SW7 5BD Issued 26 May 1983 Osteology, genitalia and the relationships ofAcanthodactylus (Reptilia: Lacertidae) 2 6 h AY 198 E.N.Arnold t Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Contents Synopsis 291 Introduction 292 The reality of the genus Acanthodactylus 293 Systematic position of Eremias guineensis 296 Characters varying within Acanthodactylus 297 Osteological characters 297 Hemipenial characters 300 Polarity of hemipenial features 303 Origin of hemipenial differences 304 Description of the hemipenes of different species 305 External features 307 Species boundaries and species groups 311 A. micropholis 311 The A. cantons group, A. gongrorhynchatus and A. haasi. . . . 311 A. schreiberi and A. boskianus 315 The A. grandis complex 316 The A. tristrami group 318 The /I. erythrurus group 318 The A. pardalis group 319 The A. scutellatus group 322 The A. opheodurus group 329 Inter-relationship of the species of Acanthodactylus 329 Acknowledgements 336 References 336 Addendum 338 Synopsis Acanthodactylus is reassessed, supplementing the external features previously used with new data fromthe skeleton and hemipenis. The genus appears to be closely related to Eremias and Mesalina ratherthan to Latastia as was previously thought, and 'Eremias' guineensis is confirmed as anAcanthodactylus. It is suggested that genitalia have an enhanced propensity to 'store' evidence of shared evolutionaryexperience in the form of common characters among descendants, when compared with other organsystems. Such characters deserve relatively high weight in assessing relationships. However, not allgenital characters can be interpreted in this way for some seem likely to have evolved as physicalisolating mechanisms between similar species and, in such cases, closely related forms may haveradically different genitalia. Differences of this kind, together with some osteological features, haveproved important in establishing the species status of several forms usually regarded as subspecies orvarieties. Thus the four subspecies of A. cantoris are now accorded full species status as A. cantoris, A.blanfordii, A. schmidti and A. arabicus and a similar upgrading may well be appropriate for two taxausually subsumed in A. tristrami: A. (/.) tristrami and A. (t.) orientalis. Within the A. pardalis complex,A. pardalis, A. maculatus and A. spinicauda are regarded as separate species and the form listed by Bull. Br. Mm. nat. Hist. (Zool.) 44(5): 29 1-339 Issued 26 May 1983 292 E. N. ARNOLD Boulenger (1921) as var. bedriagai is treated as a subspecies of A. pardalis; a similar un-named WestMoroccan population is also assigned to this species. Recently available material makes it probablethat A. grandis and A. fraseri are closely related and perhaps allocatable to the same highly variablespecies. Within the A. scutellatus group the following taxa recognized by Bons and Girot (1962) areassigned to A. scutellatus itself: A. s. scutellatus, A. s. audouini, A. s. hardyi, A. i. inornatus and A.dumerilii. A. longipes is retained as a separate species and the same status is given to A. aureus whichBons and Girot regarded as a subspecies of A. inornatus. Geographical variation requires considerablefurther study in a number of taxa including the A. grandis complex, A. boskianus and the A. scutellatusgroup but a number of currently recognized subspecies are invalid such as A. tristrami iracensisSchmidt, 1939 ( = A. (/.) orientalis), A. pardalis latastii ( = A. maculatus) and probably several of theforms in the A. scutellatus group. An attempt has been made to estimate a phylogeny for the species of Acanthodactylus, althoughconsiderable character conflict exists. Introduction The lacertid lizard genus Acanthodactylus Wiegmann, 1834 contains about 26 species. Itoccurs from Spain and Portugal across the Sahara desert and its periphery to the Red Sea,over most of Arabia and as far north as Cyprus and the Syrian-Turkish border; it alsoextends through Iraq, south and east Iran, south Afghanistan, Pakistan and northwest India.All its members are quite small, maximum body sizes of populations varying from about52 mm to around 105 mm from snout to vent. They are diurnal and essentially ground-dwelling lizards usually found in relatively flat, often sandy situations and occurring in awide variety of dry habitats which usually have at least some vegetation and range from openwoodland to the borders of quite severe desert. Substrates occupied vary from quite hardloess and clay to aeolian sand. All species appear to be mainly active hunters and feed largelyon small invertebrates including ants. Body temperatures of normally active animals areusually in the range 36 to 41 C (personal observations; Duvdevani & Borut, 1974a). Likemost open-country lizards, Acanthodactylus species are probably very prone to birdpredation and are also taken by other vertebrates such as snakes and monitor lizards(Varanus). Predator avoidance seems to depend largely on crypsis (colour match withsubstrate is often very good), fleeing, tail autotomy and the use of burrows. All species areoviparous and usual clutch size varies from two to about seven eggs, females in manypopulations appearing to produce more than one clutch annually. The majority of formsseem to mature within a year of hatching although some, like A. erythrurus in Spain andPortugal, may take two seasons. The most detailed synopsis of the genus to date is that of Boulenger (192 1) which althoughinevitably to some extent out dated, remains extremely useful, providing detailed descrip-tions of many forms. However, the amount of material available for study has increasedmassively over the past fifty years. In particular, specimens from many areas previouslyunsampled, such as much of the Sahara desert and Arabia, have been obtained. Furthermore,both Boulenger and subsequent workers have depended almost entirely on externalcharacters. As there are reasons for believing that these on their own can be misleading whenjudging lacertid relationships (see for example Arnold, 1973) it seems worthwhile to increasethe range of characters considered and, in the present paper, osteology and the structure ofthe hemipenis are taken into account. What follows is not a formal revision but an overview of Acanthodactylus is given payingespecial attention to the objective reality of the genus, species boundaries and, so far as theycan be judged, inter-relationships of its members. Although the results presented here differconsiderably from previously held opinions, the genus still requires a more detailed appraisalbased on the large but scattered collections available in continental Europe and NorthAmerica. RELATIONSHIPS OF ACANTHODACTYLUS 293 In this paper, the methods of phylogeny estimation employed are those discussed byArnold (1981 a). They depend largely but not entirely on Hennig's (1950, 1 966) precept thatjoint possession of a derived character state is prima Jade evidence of relationship. The wordrelationship is used in its genealogical sense: two species are more closely related to eachother than to a third if they share a common ancestor not shared by that form. The followingterms coined by Hennig are used, together with the adjectives derived from them:apomorphy a derived character state; synapomorphy a derived character state shared bytwo or more species and possibly indicating their relationship to each other; plesiomorphya primitive character state; symplesiomorphy a primitive character state snared by two ormore species, it does not indicate their relationship. Monophyletic and holophyletic are usedin the sense of Ashlock (1974). Abbreviations used BM(NH) British Museum (Natural History), London; CAS California Academy ofSciences, San Francisco; CM Carnegie Museum, Pittsburgh; EBD Estacion Biologica deDonana, Seville); INHM Iraq Natural History Museum, Baghdad; JUM JordanUniversity Museum; MCZ Museum of Comparative Zoology, Harvard; NMWNaturhistorisches Museum, Vienna; RSM Royal Scottish Museum; USNM UnitedStates National Museum, Washington. The reality of the genus Acanthodactylus On the basis of external features, Boulenger (19 18a) described the genus Acanthodactylus as'un des plus naturels et des plus nettement delimites de la famille des Lacertides'. Exam-ination of skeletal and hemipenial characters provides additional support for Boulenger'sview and Acanthodactylus is confirmed as a clearly denned assemblage that shows nointegration into other genera. Most of its features are found in the lizard usually known asEremias guineensis but it is certain that this form should really be included in Acantho-dactylus (see p. 296). Features present in all or most species are listed below. 1 . Frontal bones completely fused in adults and at least largely so in juveniles. 2. Often a fontanelle in the anterior wall of the orbit between the frontal and prefrontal bones. 3 . A backwardly directed spur of jugal bone absent. 4. Parietal fontanelle present, 5. Parietal bone not projecting backwards over the supraoccipital. 6. Postorbital and postfrontal bones usually separate (fused only in A. cantoris). 7. Postorbital bone not filling supratemporal foramen. 8. Parietal and squamosal bones not usually in contact. 9. Pterygoid teeth present or absent (often considerable intraspecific variation). 1 0. Epipterygoid not usually in direct contact with pro-otic bone. 1 1 . Fourteen scleral ossicles present in each eye. 12. Scleral ossicle number 14 (following numbering system of Gugg, 1939) lacks a radially directed peripheral section (see Fig. 1). 13. Number of presacral vertebrae ranges from 23 to 27, in most cases 23-26. 14. In the post-thoracic series of free dorsal ribs, the longer anterior ribs are fewer than the shorter posterior ones. 1 5. Ribs on last presacral vertebra very reduced or absent. 294 E. N. ARNOLD Fig. 1 Scleral ossicles of Acanthodactylus; arrow indicates scleral ossicle 14 which lacks a radially directed peripheral section. 16. Sternum with a more or less heart-shaped fontanelle, which is sometimes divided into left and right sections. 1 7. Sternal: xiphisternal rib formula 3 : 2 or less commonly 3:1. 18. Clavicle expanded medially with a large fenestra in this region, so that the bone forms an uninterrupted loop. 19. Interclavicle cruciform, the lateral arms usually directed obliquely forwards. 20. Only the principal coracoid foramen present. 21. Caudal vertebrae of the C-type (Arnold, 1973: 305); that is, the most anterior autotomic vertebrae have two pairs of transverse processes, the members ofthe posterior one being longer and directed obliquely backwards. 22. Hemipenis and armature usually asymmetrical with medial side reduced (see p. 300). 23. Hemipenis with an armature consisting of a flat intramuscular plate, one or two clavulae and short connectors. 24. In species where the hemipenis and armature is not very reduced there may be more than four connectors. 25. Nostril closable by a valve hinging on the posterior side of the opening. 26. Nostril nearly always situated between a postnasal, an internasal and the first upper labial (not in A. guineensis, although the situation in this species is derivablefrom that found in other Acanthodactylus, see Fig. 2). 27. First upper labial broad above but sides converge downwards (not in A. guineensis', see comments about previous character). 28. Occipital scale reduced or, more usually absent. 29. A distinct collar of imbricate scales beneath neck. 30. Dorsals small and granular or larger, imbricate and keeled with rounded posterior borders. 3 1 . Maximum number of ventral scales in a row across the belly varies from 8 to 1 8. 32. Toes with three longitudinal rows of scales around them and fingers with three or four rows. 33. Subdigital lamellae keeled. 34. Lateral rows of scales on the toes and sometimes fingers forming pectinations. 35. Femoral pores present. 36. Tail more or less cylindrical, longer than body. RELATIONSHIPS OF ACANTHODACTYLUS 295 Fig. 2 Right nasal regions of Eremias and Acanthodactylus'. (a) 'Eremias' ( = Acanthodactylus)guineensis, adult: (b) Eremias nitida, adult: (c) 'Eremias'' ( = Acanthodactylus) guineensis,juvenile: (d) Acanthodactylus erythmrus lineomaculatus, juvenile: (e) A. e. lineomaculatus,atypical juvenile, BM 1966.430. Arrows indicate first upper labial scales. Of these features, the following are probably derived, rather than primitive, within theLacertidae as a whole: 1,2,3,5,7,8, 10, 12, 15, 16, 21, 22,24, 25,27, 28, 31, 32, 33 and 34.Only number 27 appears to be unique to Acanthodactylus but 22, hemipenial asymmetrywith the medial side of the organ reduced, is found elsewhere only in Philochortus Matschie,1893 which, on other grounds, does not seem to be closely related. Monophyly is alsosuggested by (i) the great similarity between the species of Acanthodactylus so that they areinterconnected by high levels of resemblance; (ii) possession of a unique combination ofderived features even if they themselves are not unique to the genus; (iii) a coherent geo-graphical range. The relationship of Acanthodactylus to other lacertids is difficult to judge outside the 296 E. N. ARNOLD context of a comprehensive revision of the Lacertidae. Boulenger (1921) suggests that thegenus is related to Latastia Bedriaga, 1884 but the two share only a couple of the derivedfeatures listed above, namely 1 and 33. Most derived features are shared with EremiasWiegmann, 1834 (used in the restricted sense of Shcherbak, 1974, for the Palaearctic species)and Mesalina Gray, 1838 (used for the north African and southwest Asian species originallyreferred to Eremias see Arnold, 19806). Eremias shares 1, 2, 3, 7, 8, 14, 15, 16, 28, 32, 33and often 25 and 34 while Mesalina shares 1, 2?, 3, 5, 7, 8, 10, 15, 16, 21, 33 and sometimes28. More characters need to be examined before it can be decided which of these is likely tobe the sister taxon of Acanthodactylus or whether this is Eremias plus Mesalina. Systematic position of Eremias guineensis Eremias guineensis Boulenger, 1887a was described on the basis of a single hatchling,supposedly from 'Brass, Mouths of Niger' (Nigeria). In fact, as Schmidt (1919) has alreadysuggested, this locality is almost certainly erroneous, or the result of accidental trans-portation, for all the specimens collected since have come from places far to the north ofBrass in the Doka and Sudan woodland areas (vegetation classification of Rosevear, 1965) ofNigeria and Ghana (BM(NH) specimens), Niger (Pappenfuss, 1969) and Cameroon. Materialfrom the latter country was described as a separate species, Eremias benuensis Monard, 1949but Pappenfuss correctly synonymized this form with E. guineensis. The differences notedby Monard (p. 740) result largely from comparing adult Cameroon animals with Boulenger'sdescription of a single juvenile, as is clearly apparent now that a number of adults areavailable from Nigeria and other more western localities. The ostensible differences in headand frontal scale proportions and relative head and leg lengths all result from allometricchanges during growth, and the supposedly distinctive conditions of the supraocular scales,supraciliary granules and subnasal scale in E. benuensis can all be matched in E. guineensisfrom Nigeria. Finally the supposed difference in pattern (six white stripes on dorsum in E.benuensis and allegedly five in E. guineensis results from an error in Boulenger's description. Boulenger placed E. guineensis in Eremias sens. lat. because it has the combination of finedorsal scaling, keeled subdigital lamellae and nostril separated from the first upper labialscale that this author regarded as diagnostic of the genus. But he felt that this speciesoccupied an isolated position within the assemblage and assigned it to a new subgenus,Taenieremias Boulenger, 19186. This separation of E. guineensis from the rest of Eremiassens. lat. appears valid, for it differs in a number of features from the three main groupsincluded in Boulenger's concept of the genus. It can be distinguished from Eremias proper (aPalaearctic assemblage) by having its postorbital and postfrontal bones unfused, by absenceof a radially directed peripheral section on scleral ossicle number 14, possession of C-typecaudal vertebrae, a reflectable collar and an asymmetrical hemipenis; from members ofMesalina by absence of a radially directed peripheral section of scleral ossicle number 14,absence of an occipital scale, possession of three rows of scales around the digits, which arepectinate, and an asymmetrical hemipenis; and from the Ethiopian species usually referredto Eremias and the probably related Meroles by C-type caudal vertebrae and an asym-metrical hemipenis which differs in detail from any found in that group. It also differs fromthe great majority of them in its heart-shaped sternal fontanelle and in absence of aperipheral section of scleral ossicle 14 and of an occipital scale. On the other hand, E. guineensis agrees with all the features of Acanthodactylus listed onpp. 293-294 with the exception of the arrangement of scales around the nostril (features 26 &27). Yet even here the situation in E. guineensis is not very like that found in speciesassigned to Eremias sens. lat. and in size, shape and pattern of contact with each other, thescales bear a close overall resemblance to those in Acanthodactylus, the only obviousdifference being the presence of an extra suture in E. guineensis running across the areaoccupied by the first upper labial scale in Acanthodactylus to produce a smaller, nominalfirst upper labial and a subnasal scale that is radically different in form from that found in RELATIONSHIPS OF ACANTHODACTYLUS 297 members of Eremias sens. lat. (Fig. 2). Evidence that the E. guineensis condition is easilyderived from that found in Acanthodactylus is provided by a juvenile A. erythruruslineomaculatus (BM(NH) 1966.430) which, although normal in other respects, possesses anextra suture that virtually duplicates the condition found in E. guineensis (Fig. 2e). In addition to possessing the overwhelming majority of the features found in all or mostAcanthodactylus, E. guineensis bears a more detailed resemblance to A. boueti of northDahomy and Ghana and to A. erythrurus and A. savignyi of northwestern Africa (p. 318).Given this degree of similarity, it seems best to transfer E. guineensis to Acanthodactylus, assuggested elsewhere (Arnold, 1980/?). Characters varying within Acanthodactylus The following features may vary between species of Acanthodactylus and are thereforepotentially useful in defining species and in attempting to work out their inter-relationships.Species names mentioned in the course of description are used in the senses employedsubsequently in this paper. A summary of many of the features is given in Tables 8 and 9. Osteological characters Shape ofthepremaxillary region (Fig. 3). In many Acanthodactylus the outer edge of the premaxilla (viewed from above or below) issmoothly continuous with those of the maxillae, but in some species this bone is abruptlynarrowed and there may be a slight constriction where its lateral surfaces meet the maxillae. Fig. 3 Premaxillary regions of Acanthodactylus skulls from beneath: (left) A. erythrurus outeredge of maxilla and premaxilla smoothly continuous; (centre) A. schmidti premaxilla abruptlynarrowed; (right) A. scutellatus premaxilla abruptly narrowed and premaxillary teeth reducedto five, pm-premaxilla m-maxilla. This condition is strongly developed in A. blanfordii, A. schmidti, A. arabicus, A.gongrorhynchatus, A. haasi, A. masirae and especially the members of the A. scutellatusgroup: it is at least indicated in many A. boskianus, A. cantoris and A. opheodurus. A smoothlateral border to the anterior part of the skull is the usual condition in lacertids and thenarrowing found in some Acanthodactylus is almost certainly apomorphic. It is bestdeveloped in species habitually living on soft sand. These forms usually have very acutelypointed snouts and at least some probe the sand when hunting, perhaps in response to tactileor aural cues originating from hidden invertebrates. Narrowing of the premaxilla may berelated to producing the acuminate snout-tip necessary for this activity. Number of premaxillary teeth. The usual number of premaxillary teeth in most Acanthodactylus species is seven withoccasional individual variation to six or eight. The three specimens of A. boueti checked aresingular in apparently having nine premaxillary teeth while many members of the A. 298 E. N. ARNOLD scutellatus group have only five; this is true of the great majority of A. scutellatus and A.longipes examined but of only a few A. aureus (see p. 328). As the numberof premaxillary teeth in most lacertids is about seven to nine, five seems likely to be anapomorphic condition within Acanthodactylus. Its development may well be connected withthe narrowness of the premaxilla in these forms. Postorbital and postfrontal bones. These elements are separate in nearly all Acanthodactylus but they are fused in A. cantoris, even juvenile ones. Number ofpresacral vertebrae. There are from 23 to 27 vertebrae between the skull and the sacrum in Acanthodactylus. Theinterspecific variation encountered during this study is set out in Table 1. From this it willbe seen that, in most species, males usually have more presacral vertebrae than females, theaverage difference in vertebral number between them varying from about 0-66 to 1-23. Themain exceptions to this are A. pardalis bedriagai and members of the A. scutellatus groupwhere average sexual differences are very small, ranging from 0-07 to 0-28. Although mostpopulations of A. schmidti show sexual differentiation in count, this is very reduced inpopulations from the United Arab Emirates, eastern Arabia. In the majority of species, thereare 24 presacral vertebrae in most males and 25 in most females, but counts are higher in A.(t.) tristrami, A. (t.) orientalis, A. robustus, A. erythrurus, A. savignyi, A. boueti and A.guineensis: here males typically have about 25 presacral vertebrae against usually 26 or even27 in females. A. pardalis bedriagai generally has 26 vertebrae in both sexes while in maleand female A. schmidti from the United Arab Emirates, A. scutellatus and A. aureus averagecounts approach 24; the average is even lower in A. longipes and in some parts of its rangemay be around 23. Nearly all other lacertids show marked sexual variation in vertebral number, so itsreduction is likely to be apomorphic. Counts of 23 in males and 23 or 24 in females are at thelower limit for lacertids as a whole and may well be derived conditions. Other counts are wellwithin the normal range for the family but the higher ones, 25 in males, 26 or 27 in females,may possibly also be apomorphic as they commonly occur only in a minority of forms whichon other grounds may be closely related. In general, species with high presacral vertebral counts are found in relatively mesichabitats while those with low vertebral numbers occur in drier, more open places withsparser vegetation. It may be that the former habitat type requires more body flexion inpassing through plant cover and more vertebrae facilitate this. Certainly in lacertids as awhole there is a distinct if imprecise correlation between vertebral number and habitatstructure. The selective pressures that cause loss of sexual differentiation in vertebral number areobscure. Possibly the difference in average count found in most lacertids reflect sex-correlated difference in microhabitat, males might, for instance, spend more time in opensituations than females. If this were so, uniformity of vertebral number might result fromlack of such microhabitat differences, something that seems quite probable in the relativelyuniform, open environments occupied by A. schmidti and the A. scutellatus group. Anotherpossibility is that the extra vertebrae of most female lacertids increase the length of the bodycavity for carriage of eggs and that loss of sexual differentiation is a correlate of small clutchmass. However, although egg number in the A. scutellatus group at least is quite small (n = 2or 3) there is no clear evidence that a clutch occupies a smaller proportion of the body cavitythan in most other species of Acanthodactylus. Sternal ribs. In the great majority of Acanthodactylus species the usual sternal rib formula is 3 : 2, that isthree pairs of ribs attach directly to the sternum and two to the xiphisternum. However, innearly all of these, the fifth sternal rib is interrupted on at least one side in a minority ofindividuals, albeit often only briefly. In contrast, interruption is the commoner condition in RELATIONSHIPS OF ACANTHODACTYLUS 299 300 E. N. ARNOLD intra muscular portion of armature medial connectors upper surface oflateral clavula TS of lateral clavula lateral connectors upper surface ofmedial clavula TS of medial clavula medial clavula lateral lobe reflexed lateralsulcal lip medial lobe reflexed medial sulcal lip medial lip of sulcus sulcus lateral lipof sulcus Fig. 4 Left armature and uneverted hemipenis of Acanthodactylus micropholis, diagramatic viewfrom below, the base of the hemipenis is at the bottom of the illustration. The m. retractor penismagnus and ventral wall of the hemipenis have been removed, the flattened hemipenial lobesspread out and the armature disconnected: normally the points marked A coincide, with thearmature lying largely above and behind the retracted hemipenis. members of the A. pardalis and A. scutellatus groups, sometimes overwhelmingly so (Table1 .) Furthermore, the fifth sternal rib is frequently reduced to quite a small vestige. Hemipenial characters Like many other lacertids, Acanthodactylus has a hemipenis with an armature, that is adiscrete and complex supporting structure of dense connective tissue situated mainly in andaround the penis retractor magnus muscle. Its basic anatomy and that of the lacertid RELATIONSHIPS OF ACANTHODACTYLUS d e f Fig. 5 Uneverted left hemipenes of Acanthodactylus spp. showing variation in form andsymmetry. In all except (b), the lobes are flattened and normally complexly folded, but they havebeen spread out for illustration, (a) lobes subequal (A. micropholis); (b) lobes subequal buttubular and unflattened (A. cantoris); (c) medial lobe somewhat reduced (A. blanfordii. A.pardalis generally similar); (d) medial lobe very small (some A. (t.) orientalist (e) medial lobeminute, lateral lobe large (e.g. A. schmidti); (f) medial lobe absent, lateral lobe extremely large(e.g. A. scutellatus). hemipenis in general are described elsewhere (Arnold, 1973; and in press) but Fig. 4 showsthe salient features in Acanthodactylus. A number of variations between species aredescribed below, and are illustrated in Figs 5 and 6. Size. Hemipenes are small in A. cantoris, A. boueti, A. maculatus and A. spinicauda. Asymmetry in the lobes of the hemipenis. In many species, the two lobes are subequal in size but in others the medial lobe is reducedwhile the lateral one increases in size and becomes more fan-like in the uneverted organ. Theamount of medial reduction varies and in some cases the medial lobe is absent or minute. Lobes subequal: A. micropholis, A. cantoris, A. boskianus, A. schreiberi, A. grandiscomplex, A. (t.) tristrami, A. robustus, A. erythrurus, A. savignyi, A. boueti. Medial lobe somewhat reduced: A. blanfordii. Medial lobe more strongly reduced: A. pardalis, some A. (t.) orientalis. Medial lobe very small: some A. (t.) orientalis. 302 E. N. ARNOLD Fig. 6 Left armatures of Acanthodactylus species. Main views from beneath; arrangement ofdistal views of clavulae and of their transverse sections as in Fig. 4: (a) A. cantoris: (b) A.blanfordii', (c) A. schmidti; (d) A. arabicus', (e) A. gongrorhynchattus; (f) A. (t.) tristrami. A number of species have generally similar armatures to A. (t.) tristrami but differ mainly inthe form of their lateral clavula, consequently in (g)-(j) only this is shown, ventral and dorsalviews and a cross section being given: (g) A. robustus', (h) A. grandis; (i) A. boskianus; (j) A. e.erythrurus, A. pardalis; (k) A. (t.) orientalis; (1) A. guineensis\ (m) A. maculatus; (n) A. scutellatus',(o)A. opheodurus. Medial lobe minute or absent: A. schmidti, A. arabicus, A. gongrorhynchatus, A. haasi, A.guineensis, A. maculatus, A. spinicauda, A. aureus, A. longipes, A. scutellatus, and the A.opheodurus group. Lobe shape. In most Acanthodactylus, the hemipenial lobes are flattened and fan-shaped, but complexly folded in the uneverted organ. However, in A. cantoris they are narrow and tubular. Form of the proximal lip of the medial branch of the sulcus. This is usually well developed and flap-like in species where the medial side of the hemi- RELATIONSHIPS OF ACANTHODACTYLUS 303 penis is relatively large, although it is usually reduced in forms where this is not so. A.boskianus, A. schreiberi and A. grandis are exceptional in having a well-developed medianlobe but a weak proximal lip to the medial branch of the sulcus. Reduction of the medial side of the armature. In most species, the medial side of the armature is reduced in size, simpler in form than thelateral side and indeed may be entirely absent. The extent to which this reduction occursvaries and shows some correlation with reduction of the medial lobe of the hemipenis,although how precise this is varies. Sides subequal: A. micropholis, some A. cantoris. Medial side somewhat reduced: some A. cantoris, A. boskianus, A. schreiberi, A. grandiscomplex, A. (t.) tristrami, A. robustus, A. erythrurus, A. savignyi, A. boueti, A. pardalis. Medial side more strongly reduced: A. blanfordii, some A. schmidti, A. (t.) orientalis. Medial side reduced to a thread or absent: most A. schmidti, A. arabicus, A. gongrorhyn-chatus, A. haasi, A. guineensis, A. maculatus, A. spinicauda, A. aureus, A. scutellatus, A.longipes, A. opheodurus, A.felicis, A. masirae. In these forms the medial branch of the sulcus is short or absent. Shape of the medial clavula. In species where a medial clavula is present there are two main patterns. Clavula fairly broad, flat and blunt-tipped with at least the inner edge turned upwards: A.micropholis, A. cantoris, A. blanfordii and some A. schmidti. Clavula narrow and pointed with a ^ -shaped cross section: A. boskianus, A. schreiberi, A.grandis complex, A. (t.) tristrami, A. robustus, A. erythrurus, A. savignyi, A. boueti, A.pardalis. In A. (tristrami) orientalis the median clavula is like this or very small and flattened. Shape of lateral clavula. Substantial differences are found in the form of this structure. 1 . Simple, flattened, often with one or both edges turned dorsally: A. micropholis, A.cantoris, some A schmidti, A. robustus, A. opheodurus, A.felicis, A. masirae. 2. Not flattened, blunt, edges folded dorsally, a central lobe on the upper surface with aproximally directed pocket: A. blanfordii, many A. schmidti. 3. Very narrow, hollowed above: A. arabicus, A. gongrorhynchatus (A. haasi has a similarbut shorter lateral clavula). 4. Complexly structured with multiple lobes below: A. boskianus, A. schreiberi (mostmedial lobe often single), A. grandis complex (most medial lobe often divided). 5. Complexly structured although not conspicuously lobed below, sometimes veryslightly bifurcate: A. (t.) tristrami. 6. Complexly structured and not conspicuously lobed below but clearly bifurcate at tip: A.(t.) orientalis, A. erythrurus, A. savignyi, A. boueti, A. pardalis. 1. Lateral clavula simple but often bifurcate and folded to give a D-shaped cross section:A. maculatus, A. spinicauda. 8. Lateral clavula flattened and not bifurcate but folded to give a ID-shaped cross section:A. aureus, A. scutellatus, A. longipes. Connectors. In some species the connectors are all relatively thin but in others the most medial one on thelateral side is usually thickened. This is found in A. boskianus, A. schreiberi, the A. grandiscomplex, A. (t.) tristrami, A. (t.) orientalis. A. robustus, A. erythrurus, A. savignyi, A. boueti,A. guineensis, A. pardalis, A. maculatus and A. spinicauda. Polarity of hemipenial features The only available criterion for the polarity of hemipenial features found in Acanthodactylusis outgroup comparison using other lacertids as the outgroup. In these, the hemipenis is 304 E. N. ARNOLD usually fairly large and symmetrical with well developed sulcal lips on both sides. Wherepresent the armature is also typically symmetrical with simple, often flattened and ratherblunt clavulae. On this basis, small size, asymmetry of the lobes, sulcal lips and armature andthe development of peculiar and often complex clavulae and thickened connectors are alllikely to be derived features. Origin of hemipenial differences It is common in many animal groups for genital structure to vary between taxa and for atleast some of its features to be considered good indicators of relationship, especially as theyoften correlate with characters from other organ systems. The origin of such differences andpossible reasons why they should often reflect relationship are discussed at length elsewhere(Arnold, 1973; in press). It is suggested that genitals differ from other organ systems in theirpotential to retain changes that develop in them as a result of pleiotropic effects or transientselective forces. With most organ systems, any pleiotropically induced change is likely toresult in a loss of functional efficiency. Provided this lowering or efficiency does notoutweigh the original selective advantage of the genetic change concerned, the pleiotropicalteration will initially become fixed. But there will then be normalizing selection modifyingthe genotype so that the pleiotropic change will tend to be suppressed and the organ willreturn to its original state and level of efficiency. Similarly, if the organ changes in responseto direct but transient selective pressure, it will tend to revert to its original condition oncethis pressure is relaxed. The situation with genitalia is different since their efficiency cannot be considered inisolation but only in relation to their co-ordinated functioning with the genitals of theopposite sex. Efficiency does not depend on a particular conformation or size but on goodmatch between the sexes. Consequently, if there were, say, a pleiotropic change affecting themale organs which reduces copulatory efficiency, there would not only be normalizingselection acting on the male genotype to bring the male organ back towards its originalcondition but also, simultaneously, selection acting on the female genotype to producegenital changes adapting to the pleiotropic alterations that have already taken place in themale system. This means that the two sets of genitals might return to their previous mutualefficiency without reverting to their original morphological state. The same sort of effectwould be expected if transient selective forces acted directly on the genitalia themselves.Genitals are thus likely to 'store' changes, which in other organ systems would probably beeradicated. So they may be on the whole more likely to incorporate evidence of sharedevolutionary experience than many other organs. An important factor in the production of hemipenial differences in lizards, includingAcanthodactylus, may be the development of physical isolating mechanisms when pre-viously allopatric species come into contact. Such isolating mechanisms could reduce theloss of reproductive effort which is likely if two species interbreed. That such loss can occuris evidenced by the occasional discovery of hybrids between lacertid species (cases in Lacertaand its relatives are summarized by Arnold, 1973, and in Acanthodactylus by Mertens,1968). Physical isolating mechanisms would be at a premium in situations where similarspecies with no experience of each other come into contact since, in this situation,premating isolating mechanisms may not be well developed. Interspecific mating would bereduced if considerable mismatch evolved in the shape and dimensions of the genitalia of theforms concerned. The often radically different hemipenes of otherwise similar species ofAcanthodactylus may be examples of this, especially as they correspond in shape and size tothe cloacal bursa of their own females into which the male organ is inserted. Such differencesoccur between forms which, on other grounds, are likely to be closely related, for instanceAcanthodactylus pardalis and A. maculatus and, because they reduce the possibility ofsuccessful copulation between the forms, are prima facie evidence of separate species status.Some of the derived features found in the hemipenes of Acanthodactylus can be interpretedas contributing to such isolating mechanisms; they include not only small size but also RELATIONSHIPS OF ACANTHODACTYLUS 305 asymmetry of the hemipenis and armature since loss of its medial side reduces the diameterof the organ. Although they are likely to be 'stored' beyond their period of function (presumablypremating isolating mechanisms are likely to develop in the long run), changes that areprobably associated with physical isolating mechanisms are not likely to be very reliableindicators of relationship. This is partly because, by their nature, they tend to producemarked differences between very similar and often closely related species. Also, means ofproducing incompatible genitalia are rather restricted in lizards and it is very likely thatsmall size and asymmetry may have been produced several times. This is supported by theirpatchy occurrence in Acanthodactylus, turning up in a number of groups which on othergrounds do not seem to be closely related. Furthermore these features have developed inother lacertid genera. Other hemipenial features may not be subject to such strongconvergent selective forces and, as such are more likely to be good indicators of relationshipfor the reasons given earlier in this discussion. Description of the hemipenes of different species Figures in parentheses refer to the number of individual organs examined. A. micropholis (2). Lobes more or less symmetrical or medial rather reduced, sulcusbifurcate, the proximal borders of its branches formed by backwardly directed flaps;armature fairly symmetrical although one or other clavula larger than the other, clavulaerather expanded and obtusely pointed, the medial one with its inner edge turned upwards,lateral one with both edges turned upwards; often three lateral connectors and two medialones, all narrow. A. cantoris (7). Size reduced; medial lobe usually rather smaller than lateral, lobes notcomplexly folded, the walls often corrugated and lack regular plicae, stem of hemipenis long,sulcus bifurcate, the proximal borders of its branches formed by a soft flap in at least somecases although this is not always clear; armature symmetrical or medial side somewhatreduced, slender, the clavulae expanded and laminar, medial with inner edge turnedupwards, lateral with both edges turned upwards, at least one connector on each side. A. blanfordii (10). Lateral lobe large and medial one reduced, sulcus bifurcate, proximalborder of lateral but not medial side formed by a backwardly directed flap; armature clearlyasymmetrical, the median side reduced with a flattened clavula of which the inner edge isturned upwards, lateral clavula large and obtusely pointed, not flattened, edges turnedupwards and a central lobe present on the upper side below which there is a proximallydirected pocket (the lobe is least developed in southern populations), connectors as in A.micropholis. A. schmidti (10). Lateral lobe large and fan-shaped, medial lobe absent or minute, sulcusbifurcate but medial branch very short; medial side of armature absent or reduced to a threadin most cases; lateral clavula large, either rather pointed, slightly bifurcate or rather blunt,edges turned upwards and clavula may be rolled to some extent, a central lobe present inmany cases on the upper side, below which is often a proximally directed pocket as in A.blanfordii but is less well developed. In one case (BM(NH) 1931.7.16.48) there is a flattenedmedial clavula. A. arabicus (6). Essentially like A. schmidti but intramuscular part of armature muchnarrower and lateral clavula very narrow, simple and pointed with a concavity above or themedial edge rolled upwards. A. gongrorhynchatus ( 1 ). Similar to A. arabicus but medial edge of clavula rolled onto dorsalsurface. 306 E- N. ARNOLD A. haasi (1). Like A. gongrorhynchatus but clavula shorter, broader and more dorso-ventrally flattened. A. (tristrami) tristrami (3). Lobes more or less symmetrical, sulcus bifurcate, the proximalborders of its branches formed by distinct flaps; armature asymmetrical, the medial sidereduced with a pointed clavula that has a ^ -shaped cross section; lateral clavula expandedand, when spread out, essentially flat with a very weakly bifurcate tip formed of overlappinglaminae, lower surface fairly smooth but with a strap-shaped area with free edges towardsouter side, upper surface with a medially directed pocket and edges rolled upwards; threelateral and two medial connectors, the more medial of the lateral ones thickened. A. (tristrami) orientalis (4). Different from A. (t.) tristrami: lateral lobe very large and medialstrongly reduced (as in Fig. 5c or d), sulcus bifurcate, proximal borders of branches formedby distinct flaps, the median one rather small; armature very asymmetrical, medial clavulaquite small, either pointed with a <= -shaped cross-section or blunt and flattened; lateralclavula rather similar to that of A. (t.) tristrami and has free strap-shaped area on lowersurface but is smaller with a more bifurcate tip and no clear pockets. A. robustus (2). Generally like A. (t.) tristrami but lateral clavula much simpler with edgesturned on to dorsal surface. A. boskianus (10), A. schreiberi (5). Generally like A. (t.) tristrami but proximal border ofmedial branch of sulcus is a relatively weak, soft fold rather than a flap; lateral clavula hassimilar general form but is more swollen and lobed and there is an area with free edgestowards the medial side of the ventral surface; upper surface has medial edge rolled upwardsand greatly swollen. In A. schreiberi and A. boskianus the median area of the ventral surfacemay be divided by a longitudinal cleft. A. erythrurus erythrurus (3), A. e. lineomaculatus (3), A. savignyi savignyi (2), A. s. bland(2). Similar to A. (t.) tristrami but lateral clavula usually clearly bifurcate at tip and the moremedian cleft in A. (t.) tristrami, that separates a strap shaped area with free edges in this form,is usually not clearly marked. A. e. belli (6). As A. e. erythrurus but lateral clavula is smaller, flatter and usually simpler. A. boueti (1). Size reduced; lobes subequal; armature asymmetrical, the medial side reducedwith a pointed clavula that has a <c -shaped cross section; lateral clavula expanded andflattened, bifurcate at tip, with lateral edge turned upwards. A. guineensis (1). Lateral lobe very large, medial lobe absent, sulcus lacks medial branch,proximal border of its lateral branch formed by a backwardly directed flap; medial side ofarmature absent, lateral clavula superficially narrow but in fact laminar and tightly rolled,lower surface smooth but with a cleft near each margin, dorsum with a medially directedpocket; two connectors, the more medial one broad. A. pardalis (20). Lateral lobe large and medial small but plicate, sulcus bifurcate, theproximal border of its branches formed by backwardly directed flaps; armature very like thatof A. erythrurus and A. savignyi. A. maculatus (20), A. spinicauda (2). Small; lateral lobe large, medial lobe absent; medialbranch of sulcus very short, proximal border of lateral branch only of sulcus formed by back-wardly directed flap; medial side of armature absent or reduced to a thread, lateral sidenarrow and clavula small but sometimes bifurcate; compared to condition in A. pardalis,lateral clavula has been rotated and folded to give a ^-shaped cross section so that thebifurcation, if present, is in the vertical plane; two or three connectors, the most medial oneusually thickened. A. aureus (3). Lateral lobe large and fan-shaped, medial lobe absent or very small andconfluent with lateral one, medial branch of sulcus absent or very short; medial side of RELATIONSHIPS OF ACANTHODACTYLUS 307 armature absent or reduced to a thread, lateral side broad; lateral clavula flattened butrotated sideways, as in A. maculatus, and folded to give a D -shaped cross section, with twoconnectors. A. scutellatus (5), A. longipes (3). Generally similar to A. aureus but folded clavula usuallybroader and sometimes apparently only one connector present. A. opheodurus (10), A. felicis (4), A. masirae (2). Lateral lobe extremely large, medial onevery reduced, sulcus with relatively short medial branch, proximal border of lateral branchusually visible as a rather weak flap or fold; medial side of armature reduced to a thread orabsent, lateral clavula usually flattened with the margins rolled over onto dorsal surface;lateral connectors usually two or sometimes three, slender. External features These have nearly all been used by Boulenger (1918a, 1921) and subsequent authors and,until now have formed the basis of Acanthodactylus classification. Scales around nostrils. A. guineensis has a peculiar perinasal scale arrangement that seems to be derived from thetypical Acanthodactylus condition, it also occurs as a rare variant in A. erythrurus (seep. 296). Frontonasal scale. This is usually intact but is frequently divided in two by a longitudinal suture in some populations of A. erythrurus and A. savignyi. The condition occurs as a rarity in a few other species. Azygos scales between prefrontals. One or more azygos scales between the prefrontals occurs commonly in A. boueti, A.savignyi and in some populations of A. erythrurus. This arrangement is also found as anuncommon variant in A. guineensis, A. schreiberi and A. boskianus and even more rarely insome other forms. Supraocular scales. There is considerable variation in the extent to which the supraocular scales roofing the orbitare fragmented in Acanthodactylus. 1. A. schreiberi has four large, more or less intact scales;this condition is approached by members of the A. grandis complex. 2. In the A. cantoris, A.scutellatus and A. opheodurus groups and in A. gongrorhynchatus, A. haasi and many A.boskianus there is usually an area of granules wedged between the third and fourth supra-oculars and, in some cases, between the first and second too; the fourth supraocular may alsobe broken up, at least to some extent. A. boskianus from north Egypt and A. haasi mayhave the first supraocular divided into two or three sections. 3. Members of the A.pardalis group have the first supraocular sometimes divided into two or even three and thefourth is very fragmented. 4. In the A. tristrami and A. erythrurus groups the first supraocularis divided into at least three sections and often into many granules and this usually applies tothe fourth too. 5. A. boueti has the first, second and third supraoculars all fragmented. Thereis however a significant amount of variation within taxa. It is probable that the plesiomorphic condition is four more or less undivided supraocularscales, as this is by far the commonest state in the Lacertidae as a whole, but fragmentationseems to have occurred independently in some forms assigned to Eremias and in Ichnotropis. Subocular scale. A number of species have a subocular scale that extends ventrally to the edge of the mouth,separating the, in these cases, usually four anterior upper labial scales from the remainder.This condition is found in A. (t.) tristrami, A. (/.) orientalis, A. masirae, A. boueti and A.guineensis, in many A. micropholis, A. erythrurus and A. savignyi, in a few A. boskianusfrom the Iraq region and in occasional examples of some other species such as members of 308 E. N. ARNOLD the A. pardalis group and A. opheodurus. Elsewhere the subocular is shallower and the upperlabials form a continuous series beneath it. The former state is likely to be plesiomorphic asit is very widespread in the Lacertidae as a whole, but the latter occurs in Eremias proper, insome species usually assigned to Eremias from the Ethiopian region and in Meroles Gray,1838, Aporosaura Boulenger, 18876 and Ichnotropis Peters, 1854. The supposed apomor-phic condition may apparently arise by the subocular becoming increasingly narrowedventrally or by splitting to produce a small ventral scale that forms an extra upper labial. In general, the subocular is separated from the lip in forms from drier, more open habitatswhich tend to have particularly large eyes and the character may be at least partly afunctional correlate of this size increase. In species where the subocular scale usually bordersthe mouth, the lateral surface below the anterior eye, formed from the jugal and maxillarybones, tends to be fairly vertical, so that the scale can cover it but still remain roughly in thesame plane as the upper lip of which its lower section forms part. But, with increase in eyesize, the jugal arch bulges outwards and its outer face and that of the section of the maxillabelow it is directed more obliquely downwards. In this situation, a single scale covering thissurface and forming part of the lip would be strongly bent and possibly not able to providethe necessary flexibility for labial movement. In forms living in dry areas, the need for suchmovement is increased, for the upper lip projects further ventrally and is more bulky, so itand the scales covering its outer surface are displaced more during closure of the mouth,presumably producing a more efficient seal against water loss and incursion of sand particlesthan is present in more mesic species. Because of these changes, the replacement of theoriginal subocular by two functional parts would probably be an advantage. Number of upper labial scales anterior to the centre of the eye. In the majority of forms there are typically four upper labials anterior to the centre of the eyebut in some, such as the members of the A. cantoris group, A. scutellatus and A. longipes, fiveis the usual number. As four is found in forms which on other grounds seem primitive, it maybe the plesiomorphic condition for Acanthodactylus. When present, the extra upper labialsometimes seems to develop by being split off the subocular but this may not always be itsorigin. In many species exceptions to the usual number occur. Parietal scales. A. boueti is singular in having the parietal scales reduced and the areas normally occupied by their lateral and posterior parts replaced by small scales. Ear opening. This is usually quite large but in A. gongrorhynchatus and A. haasi it is reduced in size, beingoccluded by skin from above and from the front. Similar reduction is found in some otherlacertids that live on loose sand such as Aporosaura and some species of Meroles. Dorsal body scales. There is great variation in size and shape of the dorsal scales. In many species they are fineand more or less granular and either smooth or carinate, but in others size increasesposteriorly and the scales become flat, more keeled and strongly imbricate. In such cases, thelateral scales tend to be smaller than the more medial ones. Transverse counts at mid-bodymay be as high as 100 in the most fine-scaled species, A. longipes, or as low as 18 in some A.boskianus asper. There is often substantial variation within and, more obviously, betweenpopulations of a species; thus counts vary from about 35 to 80 in A. scutellatus and from 1 8to 52 in A. boskianus. This suggests that dorsal scalation is very labile so that the assessmentof a primitive condition for the genus is impossible. All that can be said is that the extremesencountered in Acanthodactylus are near the extremes for the Lacertidae as a whole and thatthe primitive condition is perhaps more likely to be somewhere in between. In some other sections of the Lacertidae, relatively fine granular scaling is usuallyassociated with open environments while, large, keeled, strongly imbricate scales aretypically found in species that live, or at least take refuge in, dense often spiny vegetation. RELATIONSHIPS OF ACANTHODACTYLUS 309 The latter scale arrangement provides much better mechanical protection against damage tothe flexible skin between the scales, something that is potentially likely to occur whenrunning in densely vegetated habitats. The same selective pressure may well act onAcanthodactylus; the most fine-scaled forms, like many populations of the A. scutellatusgroup, tend to live in the most open habitats, whereas forms like A. boskianus do frequentlyseem to take refuge in spiny vegetation (see p. 3 1 5). Dorslateral tracts of enlarged scales. These are found on the posterior body of A. schmidti and are at least indicated in some A. blanfordii. Ventral scale number. The number of ventral scales in the longest transverse row across the belly varies from about8 to 18, 10 being the commonest figure. As the majority of lacertids have low numbers, theseare likely to be primitive in Acanthodactylus. In general low numbers are most frequent informs occupying more mesic habitats and higher ones in species living in relatively aridareas, but this correlation is imprecise. Intraspecific variation occurs, especially in formswith higher counts. Tessellation of ventral scales. Forms with high transverse counts of ventrals often have the scales tessellated, that is, theyare staggered so that they do not form regular longitudinal rows. In many cases, thisstaggering is confined to the sides of the belly, but it may affect the median areas too. It is bestand most extensively developed in most populations of the A. scutellatus group and in the A.grandis complex; it is found to a lesser extent in other members of these assemblages and isat least indicated in the A. cantoris group and A. gongrorhynchatus. There is significantvariation in the extent of staggering within species where it is present. Tessellation is foundin some other lacertid groups such as Eremias proper, Meroles and Aporosaura. Inter gradation of dorsal and ventral scales. In most Acanthodactylus, dorsal and ventral scales are easily distinguished from each other,but in A. gongrorhynchatus they intergrade. This trend is also apparent in A. haasi and to alesser extent in some A. arabicus. Number of rows of scales running along fingers. In many species there are three rows of scales running along fingers one to four, the samenumber that runs along each toe. One row is dorsal, one ventral and the third posterior. Onfinger five, and sometimes at the base of the others, there may be four rows, the additionalone being anterior. In contrast, a number of forms have four regular and continuous rows ofscales on all fingers; these include the A. cantoris and A. scutellatus groups, the A. grandiscomplex, A. gongrorhynchatus and A. haasi. In fact, these two conditions are not as clear cutas they appear, for a number of individuals of species that usually have three scale rowspossess additional scales forming an irregular fourth row on at least some fingers; this occursin A. micropholis, A. robustus and A. opheodurus. A. masirae habitually has four rows butthese are often irregular, and A. cantoris in which four regular rows are usually presentoccasionally has the anterior one strongly reduced. It seems likely that three scale rows running along the fingers is the primitive condition inAcanthodactylus for it comes closest to the two rows typical of the majority of Lacertidae andother scincomorph lizards. Three or four scale rows on the fingers also occur in Eremiasproper, Meroles and Aporosaura. The presence of a regular fourth row is largely associatedwith sandy habitats. It may well allow better development of a pectination on the anteriorface of the digit that matches the posterior one. This results in a wider digit that is moreefficient for digging in light, unconsolidated substrates such as fine sand whereAcanthodactylus often scrabbles for food as well as excavating burrows. 310 E.N.ARNOLD Pectination of the digits. The scales along the posterior surfaces of the digits are pointed and project to form a comb-like fringe or pectination. In some instances, this is inconspicuous but in others the scalesbecome very long and flattened in the horizontal plane and in extreme cases, such as A.longipes, the fringe on the toes may be wider than the digits themselves. The scales formingthe row running along the upper surfaces of the digits may project slightly forwards toproduce a weak anterior fringe as well, but in species where there is an anterior scale row onthe fingers, this may form a fringe approaching the posterior one in development. There is a continuum between the weakest fringes and the most pronounced ones. Degreeof development correlates very closely with what is known about the ecology of the speciesconcerned, being least in forms living on relatively hard ground and greatest in forms fromsoft sand. The relationship is so strong in taxa for which some ecological information isavailable that it seems reasonable to use degree of pectination to predict substrate type incases where this is unknown. The importance of fringes on the fingers for digging in loosesand has already been mentioned; fringes also enable lizards to travel across soft sandsurfaces efficiently by increasing the area covered so that unit pressure is low and effort is notwasted pushing the sand backwards as the lizard moves forwards. Keeling on upper caudal scales. Most Acanthodactylus, like the majority of lacertids, have keels on the dorsal scales of the tailbase but these are absent in A. (/.) tristrami, A. (t.) orientalis and A. robustus. The restricteddistribution of this feature in the Lacertidae suggests it is an apomorphy, although it hasdeveloped independently elsewhere, for instance in Mesalina rubropunctata. Lateral processes on tail base. In A. spinicauda the lateral scales at the base of the tail are greatly enlarged, especially in males, with their keels produced to form tubercles. This feature is unique in the Lacertidae. Tail length. In most Acanthodactylus the tail is more than 1-5 times the length of the head and body together but in A. (t.) tristrami, A. (t.) orientalis and A. robustus it is less. Pattern. In the great majority of Acanthodactylus species, juveniles have a striking pattern of narrow,dark and light longitudinal stripes. The only certain exceptions are A. schmidti, A. longipesand some A. scutellatus in which the newly hatched young are uniform or dappled. Thiscondition is likely to be derived as striped patterns are very widespread among juvenilelacertids. When present, the number of dark stripes varies considerably and there may be asmany as six pairs, although not all these run the whole length of the body. In the no-menclature of Lantz (1928), developed for Palaearctic Eremias, these are 1. the spinaloriginating near the mid-line and usually fusing to produce a single stripe on the body, 2. theoccipital arising from the hind edge of the parietal scale, 3. the parietal arising from thelateral edge of that scale, 4. the temporal beginning behind the eye and passing over theupper edge of the ear (equivalent to Lantz's upper and lower temporals), 5. the maxillarypassing through the ear and 6. the costal passing below it. It is usually acknowledged that a high number of stripes is plesiomorphic in the Lacertidae(Eimer, 1881; Boulenger; 1921; Lantz, 1928) and this condition is certainly commoner informs which on other grounds seem primitive. Reduction in number seems to take placemost usually by loss or fusion of bands at the mid-line. As we have seen, a few Acantho-dactylus have a pair of spinal bands on the neck that fuse on the body; reduction from thisstate appears to follow a clear sequence: 1. the spinals fuse throughout their length; 2. theresultant composite stripe shortens while the occipital stripes fuse posterior to it; 3. thespinal stripe is reduced to a vestige or disappears and fusion of the occipitals extends for-wards; 4. the occipitals fuse completely to produce a simple mid-line stripe. This means ofcourse that mid-line stripes in Acanthodactylus are not always homologous as they can be RELATIONSHIPS OF ACANTHODACTYLUS 3 1 1 composed either of fused spinal bands or of occipitals. There may also be loss of the outer-most bands, especially the costals on the body. Presence of a pair of spinal stripes on the neck is largely confined to some A. micropholis,A. erythrurus and A. pardalis. The partial development of a mid-dorsal stripe made up of theoccipitals occurs in some individuals of the following species: A. boskianus, A. schreiberi, A.pardalis, A. scutellatus, A. opheodurus, A. felicis and A. masirae, but complete fusion seemsto be confined to some A. boskianus and A. scutellatus and most A. opheodurus. A. micropholis retains a simple striped pattern throughout life. In other forms this is oftenmodified, but predominantly striped individuals, or ones with longitudinal rows of spots, arecommon in most species. However not all the stripes present in juveniles may be discerniblein these adults. A. pardalis and A. maculatus are singular in that some individuals of each have spots ofbrownish red pigment on their backs that do not fade or disappear in alcohol. In A. (t.) tristrami, A. (t.) orientalis and A. robustus some individuals have two rows of largeocellar markings along the back. Species boundaries and species groups The introduction of osteological and especially hemipenial characters provides a partial testof the homogeneity of the species presently recognized within Acanthodactylus. For even ifthe external features on which these nominal taxa are based show little obvious variation,discontinuity in characters from the new sources may indicate that more than one form isinvolved. The species are mostly discussed below in groups that consist of species with a highlevel of overall similarity and which, in most cases, probably form holophyletic assemblages,although there are exceptions. A. micropholis A. micropholis Blanford, 1874a, occurs in southeast Iran and southwest Pakistan, in thelatter country occupying sandy places along water courses in hilly areas (Minton, 1966). Ithas a high proportion of primitive features including the following: premaxilla with aboutseven teeth and not abruptly narrowed, 24 presacral vertebrae in males and 25 in females,sternal ribs usually intact; hemipenis and armature fairly symmetrical, clavulae flattened andsimple, connectors slender; subocular often reaching mouth, frequently four upper labialsanterior to centre of eye, eight or ten ventrals in longest row across belly, ventrals arranged instraight longitudinal rows, usually three longitudinal rows of scales on fingers, pectination ondigits rather weak, upper caudal scales keeled, young with numerous dorsal stripes which areretained by adults. On the other hand, the first and fourth supraoculars are broken up and, insome individuals, the subocular may be separated from the mouth, five upper labials aresometimes present anterior to the centre of the eye and there may be indications of a fourthscale row on the fingers. The A. cantons group, A. gongrorhynchatus and A. haasi A. cantons Giinther, 1864ft: NW. India, Pakistan, E. Afghanistan. A. blanfordii Boulenger, 191 8a: S. Afghanistan, SW. Pakistan, SE, Iran, Oman (Muscat area). A. schmidti Haas, 1957: Arabia except extreme west; SW. Iran (Anderson, 1963, 1974). A. arabicus Boulenger, 1918a: SW. Arabia. A. gongrorhynchatus Leviton & Anderson, 1967: E. and SE. Arabia. A. haasi Leviton & Anderson, 1967: E. Arabia (Dhahran). The first four forms, which constitute the A. cantoris group, are all associated with loosesand habitats (A. cantoris Minton, 1966; Mertens, 1969, A. blanfordii Blanford, 18746;Anderson, 1963; Clark et ai, 1969; M. D. Gallagher, pers. comm. A. schmidti Anderson,1963; personal observations, A. arabicus Anderson, 1895). Soft ground types are also 312 E.N.ARNOLD suggested by the extensive pectination of the digits. Members of the A. cantoris group sharethe following features: premaxilla with about seven teeth and usually abruptly narrowed(least in A. cantoris), usually 24 presacral vertebrae in males, 25 in females (less sexualdifferentiation in some populations of A. schmidti), fifth sternal rib often intact; hemipenisand armature usually showing marked asymmetry (not in most A. cantoris), clavulae notvery complex, connectors relatively slender; first supraocular intact or not much divided andthe fourth large or rather fragmented with a group of granules wedged between it and thethird, subocular nearly always separated from the mouth, usually five upper labials anteriorto the centre of the eye, posterior dorsals coarse and carinate, 12-18 ventrals in longest rowacross belly, ventrals arranged in fairly straight longitudinal rows, although sometimesstaggered at sides, nearly always four longitudinal rows of scales on fingers (anterior rowsometimes reduced in A. cantoris), pectination on digits strong, upper caudal scales keeled,young usually with numerous dorsal stripes (not in A. schmidti) that are typically lost inadults.In the past it has been usual to treat these four forms as subspecies of A. cantoris. But, Table 2 Variation in the A. cantoris group A. cantoris A. blanfordii A. schmidti A. arabicus Hemipenissize medial lobe smallunreduced medial side of armature virtually unreduced lateral clavula flat andsimple Postorbital and postfrontal bones fused Sexual variation in number of presacral vertebrae yes Dorsal scales across mid-body 26-36Dorsolateral tracts of enlarged scales on no hind backNumber of ventral scales in longest row across belly 1 2( 1 4)Juveniles with dorsal stripes yes Adults striped or with rows of dark spots sometimes Tail with transverse bands no Maximum size withinpopulations(snout-vent, mm) 77 large somewhatreduced somewhatreduced lobed andpocketed separateyes 30-46 indicated insome cases (11)12-14(16)yes sometimes(all animalsfrom Muscat) no 60-75 large minute or absent usually absent lobed andsometimespocketed separate not in populationsfrom United ArabEmirates 33-51 yes 13-18 no no in juveniles andmany adults 67-105 large minute orabsent absent very narrow separate yes 27-35no (12)14-16yes in west ofrange no55-63 RELATIONSHIPS OF ACANTHODACTYLUS 3 1 3 although they have a strong overall resemblance to each other and, so far as is known, areallopatric, each has a number of distinctive characters that are often more pronounced thanthose separating accepted species of Acanthodactylus (see Table 2). Because of this, it seemsmore consistent to give them full species status. A. cantoris is widespread in northwest India and in Pakistan is found in the southeast andin the Indus drainage whence it reaches east Afghanistan (Kabul River area, CAS 96200-01 ,120358-60); it also extends westwards along the Pakistan coast at least to Ormara (BM(NH)1904.12.7.2; Minton, 1966, records A. blanfordii from this locality but his specimens, RSM1 964.58.2/6, are in fact A. cantoris). There is some geographical variation, for instance in thePunjab the anterior row of scales on the fingers is incomplete in some animals. This is true ofa female from Cambellpur (BM(NH) 1933.5.16.8) which is also peculiar in having a ratherblunt snout and elongate, strongly overlapping dorsal scales. A. blanfordii reaches its easternmost limit in the Helmand drainage of south Afghanistanand the neighbouring border area of Pakistan (BM(NH) 86.9.21.77-80; Mertens, 1969; it isalso probable that the specimens listed by Clark et al, 1969, and Minton, 1966 are A.blanfordii). A. blanfordii and A. cantoris may be separated by intervening high ground in thisarea but it is possible that they meet on the coast for A. blanfordii extends eastwards as far asPasni (Mertens, 1969), only about 130 km from the nearest known A. cantoris locality. Thedifferences in dorsal and gular scale counts on which these two forms were originallyseparated do not always hold but they can be distinguished by the fusion of postorbital andpostfrontal bones in A. cantoris and by hemipenial features. A. blanfordii extends intoeastern Iran at least as far as the Bandar Abbas region (Anderson, 1963) and an isolatedpopulation has recently been found by M. D. Gallagher near Muscat, southeast Arabia(BM(NH) 1973.723-26, 1976. 1462-64, 1977.68-69); here all the adults examined retain atleast traces of dorsal striping. A. schmidti is widespread in Arabia where it shows substantial regional variation in bodysize (Arnold, 1981/7). It also extends into southwest Iran (Anderson, 1974) but withoutapparently contacting A blanfordii. A. arabicus is restricted to the littoral area of southwest Arabia and seems to be separatedfrom A. schmidti by hilly, not very sandy country occupied by such species as A. boskianus,A. felicis and A. opheodurus. Although A. arabicus and A. schmidti both have the medialside of the hemipenis and armature very reduced, they differ in their other apparentapomorphies (A. arabicus very narrow clavula; A. schmidti often very large size,dorsolateral tracts of enlarged scales on hind back, lack of stripes in juvenile and adultpatterns) and in dorsal scale number. In fact A. arabicus is more similar to A. blanfordii,especially Muscat specimens, differing mainly in its more slender build, greater hemipenialasymmetry, narrower clavula and higher number of ventral scales across the belly (usually1416 as against 11-13(14) at Muscat). A. arabicus from Aden and its vicinity arecomparatively large and usually retain stripes or spots when adult but animals from futhereast (as far as Gischin ( = Qishn), NMW 11814/1-16) are smaller and often become uniformin colouring with maturity. Relationships within the A. cantoris group are not easy to discern with great certainty,since most of the apomorphies available for judging affinities are rather weak (see p. 324), butA. blanfordii, A. schmidti and A. arabicus all differ from A. cantoris in their more narrowedpremaxillae, more asymmetrical hemipenes and usually higher number of ventrals, whilelacking its fused postorbital and postfrontal bones and its hemipenial peculiarities (p. 305),so they may well form a distinct unit with A. blanfordii as the least modified species. Withinit, A. schmidti and A. arabicus share almost total loss of the medial side of the hemipenis andarmature, but there are no other apparent synapomorphies joining them, apart from increasein ventral scale count. As extreme hemipenial asymmetry has developed a number of timeswithin Acanthodactylus, its value as an indicator of relationship is quite low. It is possiblethat A. schmidti is the sister species of A. blanfordii for the latter occasionally showsindications of the dorsolateral tracts of enlarged scales on the hind-back that are typical of 314 E.N.ARNOLD the former and they often share a peculiar lateral clavula shape in which there is a proxi-mally directed pocket. Moreover, A. schmidti occasionally has a medial clavula exactly likethat of A. blanfordii. Certainly A. arabicus and A. schmidti are more like A. blanfordii thanlike each other and may well have been independently derived from a A. blanfordii-\ikestock. A. gongrorhynchatus is sympatric with A. schmidti in east and southeast Arabia and sharesmany features of the A. cantoris group but it differs in its very slender adult build, convexpileus, short but abruptly acuminate snout, weakly keeled supratemporal scales that arefragmented posteriorly, four upper labials anterior to the centre of the eye, an ear hole that isreduced in size and fine dorsal scaling that sometimes grades into the ventral scales. Recentobservations by Mr W. Ross (personal communication) show that A. gongrorhynchatus liveson aeolian sand. Its morphology depressed fingers, strong digital pectination, very pointedsnout and partly occluded ear opening is appropriate for such a substrate and suggests thatit is more strictly confined to this habitat than A. schmidti which occupies the same generalarea. A. gongrorhynchatus may well be paedomorphic in some features which are typical ofjuvenile Acanthodactylus, for instance the rounded pileus, short snout and lack of strongkeeling on the supratemporal scales. In fact it has substantial resemblance to immaturespecimens of A. arabicus and may well be closely related to this species, especially as theirhemipenes are very similar including the very narrow lateral clavula. The present situation in Arabia may have arisen by a second stock of the A. cantoris groupinvading to exist alongside one already present. It is possible that much of the peninsula wasoriginally occupied by A. arabicus-\ike populations and that the precursor of A. schmidtientered the area from the east to become sympatric with them, except in the extreme south-west littoral region which is isolated by geographical barriers. Here A. arabicus survives butelsewhere coexistence of two species in soft-sand habitats favoured by the A. cantoris groupwould be expected to produce considerable character displacement, the A. arabicus-likestock giving rise to A. gongrorhynchatus and the invading one to A. schmidti, the two mostdivergent species in the assemblage. An analagous situation exists among the nocturnalground geckoes of the genus Stenodactylus that occupy soft sand habitats (Arnold, 1980a).As in Acanthodactylus there is a single species in the southwestern littoral region of Arabia(S. pulcher) which has a close relative widespread in the peninsula (S. arabicus), where it issympatric with another more distantly related species (S. doriae). As with A. gongrorhyn-chatus, S. arabicus is more strongly adapted to sandy environments and smaller than itssouthwestern relative whereas S. doriae, like A. schmidti is relatively large. A. haasi is known only from the male holotype obtained at Dahran, eastern Saudi Arabia.In spite of being collected in 1946, no further examples have appeared, even though a greatdeal of Acanthodactylus material has been taken in the general area. A. haasi has much incommon with A. gongrorhynchatus and shares a number of synapomorphies with it thatoccur nowhere else in the genus; these include very slender habitus, similar lateral clavulastructure and reduced ear opening. It differs in its less acuminate snout and less narrowedpremaxilla, shorter lateral clavula, five upper labials anterior to the centre of the eye, largeand weakly keeled dorsal scales, less pectination on the fingers and numerous keels on thesubdigital lamellae. In most of these features and many others A. haasi resembles members ofthe A. cantoris group from which A. gongrorhynchatus is probably derived. This togetherwith the synapomorphies shared with the latter may indicate that A. haasi and A.gongrorhynchatus are sister species. Another possibility, suggested partly by the fact that no further specimens have beencollected, is that A. haasi is a hybrid between A. gongrorhynchatus and one of the otherAcanthodactylus species in eastern Arabia. Assuming that a hybrid would probably beintermediate in many features, the various candidates can be considered in turn. A. schmidtiis unlikely because its premaxilla is constricted, it has a high number of ventral scales in thelongest row across the belly (13-18 compared with 10-12 in A. gongrorhynchatus and 12 inA. haasi), pectination on the fingers is extensive, multiple carination on the digits is absent RELATIONSHIPS OF ACANTHODACTYLUS 3 1 5 (present in A. haasi) and, in east Arabia, it is a far larger animal than A. gongrorhynchatus, sointerspecific copulation is unlikely; size difference would probably also exclude A. boskianusas well. A. scutellatus has a narrow premaxilla with only five teeth, dorsal scaling is fine andagain there is no multiple carination on the digits. A. opheodurus has this feature in at leastsome cases and possesses all the other characteristics to be expected if A. haasi is a hybridbetween it and A. gongrorhynchatus. But the status of A. haasi will probably only be decidedby further collection in the area of the type locality. A. schreiberi and A. boskianus A. schreiberi Boulenger, 1878; Cyprus, Lebanon, N. Israel. A. boskianus (Daudin, 1802): Much of north Africa as far south as N. Nigeria and Eritrea, Arabia butnot the R'ub al Khali, S. Israel, Jordan, Iraq, Syria and adjoining Turkey. These species are rarely found on very soft sand surfaces but occur on more stable sand andother harder substrates. A. boskianus is often associated with quite dense, sometimes spinyvegetation, although it also forages in more open areas (Doumergue, 1901; Flower, 1933;field labels attached to BM specimens; personal observations in southeast Arabia). Bothspecies share a relatively high number of primitive features including the following:premaxilla with about seven teeth and not abruptly narrowed, usually 24 presacral vertebraein males and 25 in females; fifth sternal rib often intact, hemipenis more or less symmetrical,medial side of clavula not strongly reduced; supraoculars more or less intact (first sometimesdivided in A. boskianus), usually four upper labials anterior to centre of eye, typically 10ventral scales in longest row across belly (12 in many Arabian A. boskianus), ventralsarranged in straight longitudinal rows, three longitudinal rows of scales on fingers,pectination on digits not exceptionally strong, upper caudal scales keeled, young withnumerous dorsal stripes that are sometimes retained in modified form by adults. Apparentlyapomorphic characters include reduction of the proximal lip of the medial branch of thehemipenial sulcus, some reduction of medial side of armature, medial clavula pointed with a^ -shaped cross section, lateral clavula complexly lobed, most medial of the lateralconnectors broad; subocular does not usually reach mouth (exceptions in A. boskianus) anddorsal scales are often large. The close resemblance of A. schreiberi and A. boskianustogether with the close match of their hemipenes in apomorphic features makes it veryprobable that they are sister species. A. schreiberi has two allopatric populations: A. s. schreiberi on Cyprus and A. schreiberisyriacus Boettger, 1879 in Lebanon and north Israel, the latter having coarser dorsal scaleswith sharper keels. A. boskianus usually differs most obviously from A. schreiberi inpossessing larger dorsal scales. However there is slight overlap in the transverse scale countsbetween the hind limbs, but not in Israel where the two species appear to contact each otherwithout introgression (Duvdevani and Borut, 19746). A. boskianus is the most widelydistributed species of its genus but the range is not continuous in northern Africa where thislizard is mainy restricted to the periphery of the Saharan desert and to oases and otherfairly mesic areas within it. As the species is relatively uniform in this region, its spreadto these isolated areas may have been quite recent, presumably during one of the episodes inthe Quaternary when conditions in north Africa were less extreme. A. boskianus has been divided into three subspecies: A. b. boskianus in the Nile delta areaand parts of Sinai; A. b. euphraticus Boulenger, 1919 described from Ramadieh ( = Ramdi,Iraq) and A. b. asper (Audouin, 1829) which covers almost the whole of the species' range.This simple, tripartite division is not satisfactory, for some of the supposedly distinctivefeatures of A. b. euphraticus are not consistent and there is some differentiation within thepopulations assigned to A. b. asper. An adequate intraspecific treatment of A. boskianus isbeyond the scope of this paper but variation of some features within the species will be 316 E.N.ARNOLD briefly described. Over most of north Africa, the number of dorsal scales in a transverse rowat mid-body varies from about 26 to 41, although maxima are less in many areas, the firstsupraocular scale is usually intact and the subocular does not reach the mouth in mostindividuals. In the Nile delta and north Sinai, populations assigned to A. b. boskianus havehigh dorsal scale counts (34-52) and the supraocular scale is divided, but these populationsintergrade with those typical of the rest of north Africa and with similar ones in Israel andwest Jordan. Animals with high dorsal counts (38-48) also occur in northeast Jordan, northand central Iraq, east Syria and adjoining Turkey. In some cases, such as the types of A. b.euphraticus, they also have the first supraocular divided and the subocular often meeting themouth, but these conditions are absent in many individuals from neighbouring populations.Arabian animals are rather similar to those widespread in north Africa but are characterizedby low transverse dorsal scale counts at mid-body (usually 18-27 but up to 36 in the south-western mountains) and a large body size (up to 93 mm from snout to vent against maxima ofabout 82 elsewhere; specimens from the western periphery of the peninsula tend to besmaller). Arabian A. boskianus contrast strongly with those found to the north. In Jordan ananimal from the southeast of the country (El Inab, JUM R505) is of the Arabian typewhereas those from elsewhere are less coarsely scaled (31-37 scales across mid-back) andrelatively small (only up to about 65 mm from snout to vent). Animals of this type occur atWadi Rum (JUM R69), only about 130 km from El Inab. Whether the two forms intergradeand, if so, whether the intergradation is abrupt is not known. Geographical variation in A. boskianus may well reflect differences in niche across itsrange. As stated, this species is often associated with dense vegetation and large dorsal scalesmay well be protective where shrubs are rigid and spiny; the fine-scaled populations are inrelatively mesic areas (Nile Delta, Tigris-Euphrates river system) where vegetation would beexpected to be less damaging than in more arid regions. Division of the first supraocular scaleand a subocular that reaches the mouth are also characters typical of more mesic situations(see p. 307). The distinctive features of lowland Arabian A. boskianus may be related to thefact that they coexist here with A. opheodurus, a small form originally confused with A.boskianus that occupies similar substrates and occurs in strict sympatry with it, at least insome areas (Arnold, 19806). The presence of A. opheodurus might have produced displace-ment or restriction of the niche available to A. boskianus with consequent morphologicalchange (Arnold, 19816). Thus, increase in body size may allow adult A. boskianus to takelarger prey, reducing competition for small food items; very large scale size could indicatethat this species spends a higher proportion of time in vegetation than elsewhere. It isperhaps significant that in extreme southwest Arabia, where no A. opheodurus are known, A.boskianus is smaller with higher dorsal scale counts than elsewhere in the peninsula. In spite of the considerable variation encountered in A. boskianus, as presently under-stood, there is as yet no firm evidence that it consists of more than one species. However, A.schreiberi may well have originated as an isolate of A. boskianus. The A. grand is complex A. grandis Boulenger, 1909 complex (including A. fraseri Boulenger, 1918c): Syria, E. Lebanon,Jordan, NW. Arabia, Iraq, SW. Iran (Khuzistan and Pars provinces Anderson, 1974). Nothing certain is known about the ecology of these lizards but the restricted pectination onthe digits suggests that they are not usually found on very soft sand surfaces; however there issome variation in this feature between populations which may indicate that they occupy avariety of ground types. The members of the A. grandis complex share many features with A.schreiberi and A. boskianus, especially the former, the most significant differences beingthat, in the A. grandis complex, the ventrals are tessellated and the number in the longestrow across the belly is higher (14-18), there are four longitudinal series of scales runningalong the fingers and the hemipenis, although generally very similar often differs slightly inthe pattern of lobing on the upper surface of the lateral clavula (p. 306); also, five upperlabial scales are more frequently present anterior to the centre of the eye. RELATIONSHIPS OF ACANTHODACTYLUSTable 3 A. grandis complex: variation between samples 317 A. grandis was originally described from the Damascus area of Syria while the only othernominal species assignable to the complex, A. fraseri, is based on material from Zobeya,Shariba ( = Shu'aiba, northeast of Basra) far to the east in southeastern Iraq. Materialcollected subsequently in the intervening areas shows a variety of conditions intermediatebetween the named forms but the pattern of variation appears to be irregular and sometimesanimals from adjacent localities show considerable differences in such features as adult size,strength of keeling on the dorsal scales, and pattern (summarized in Table 3). In general,members of the eastern populations tend to be smaller than western animals with fewer,more strongly keeled dorsal scales and single points on each subdigital lamella. It is possible that the A. grandis complex is best regarded as a single species but availablesamples are too small and scattered to be certain about this. The irregular variation ofpopulations intermediate between typical A. grandis and A. fraseri may reflect the geographyof Mesopotamia, for here the comparatively arid country favoured by Acanthodactylus isdivided up by the Tigris and Euphrates rivers and their tributaries which flood seasonally, sopopulations may be substantially discontinuous. Other Acanthodactylus species, especiallyA. boskianus, also show considerable variation in this area. 318 E.N.ARNOLD The A. tristrami group A. (/.) tristrami (Giinther, 1864): Lebanon, SW. Syria, NW. Jordan. A. (/.) orientalis Angel, 1936: E. Syria, W. and central Iraq. A. robustus Werner, 1929: S. Syria, SW. Iraq, Jordan and N. Arabia. The weak digital pectination in these forms suggests that they are found on fairly hardsubstrates. A. (t.) tristrami seems to be confined to steppe-type habitats on the edge of theAnti-Lebanon range. A. (t.) orientalis may also be restricted to relatively mesic habitats sincemost records are from localities on or near the Tigris-Euphrates river system (Angel, 1936;Schmidt, 1939; Haas and Werner, 1969), although Angel also records it from Palmyra. A.robustus, on the other hand seems to occupy more desertic regions; Riney (1953) gives briefecological notes that refer to this species, although they are attributed to A. (t.) orientalis. Typical A. tristrami has many features in common with A. schreiberi (p. 3 1 5) but thereare a number of differences: there are more presacral vertebrae, with about 25 in males and 26or 27 in females, the proximal lip of the medial branch of the hemipenial sulcus is notreduced, the lateral clavula is complex but not lobed and is very slightly bifurcate at its tip,the first and fourth supraoculars are fragmented, the subocular reaches the mouth, the dorsaland upper caudal scales lack keels and the tail is short. Of these features, fragmentation of thesupraoculars, unkeeled upper caudals, short tail, clavula shape, and perhaps vertebralnumber may be regarded as apomorphies. Nearly all occur in other populations of the A.tristrami group and unkeeled upper caudals and the short tail are confined to it. Two other subspecies assigned to A. tristrami have been named, A. t. orientalis fromeastern Syria and A. t. iracensis, Schmidt, 1939, described from Haditha on the Euphrates inwest Iraq. The latter is said to have a lower mid-body dorsal scale count (45-46 against48-56) and a weaker dorsal pattern than A. t. orientalis but other material from the same area(Haas & Werner, 1969) and from Najaf, further down the Euphrates (INHM, with 51 dorsalscales at mid-body and a bold pattern) suggest that this distinction is illusory. A. t. iracensis istherefore referred to the synonymy of orientalis. As understood here, this form isdistinguished from typical tristrami by its smaller size (up to 63 mm from snout to ventinstead of up to 92 mm) and fewer dorsal scales (43-56 against 54-65). The hemipenes arealso distinctive, orientalis having the medial lobe and medial side of the armature veryreduced. These differences suggest that tristrami and orientalis might be better regarded asfull species but more information is needed before this can be confirmed. A. robustus shares most of the distinctive features of A. (t.) tristrami but the lateral clavulais relatively simple, the snout is shorter and more pointed, the subocular frequentlyseparated from the mouth, there are often 12 instead of 10 ventrals in the longest row acrossthe belly and there may be a rudimentary fourth longitudinal row of scales on the fingers. A. ery thrums group A. erythrurus (Schinz, 1833): Spain, Portugal, Morocco, N. Algeria. A. savignyi(Audou\n, 1829): N. Algeria, N. Tunisia. A. boueti Chabanaud, 1 9 1 7: N. Dahomey, N. Ghana. A. guineensis (Boulenger, 1887a): Ghana, Nigeria, Niger and Cameroon. These species are typical of relatively mesic areas and usually have weak digital pectination.They are similar to A. (t.) tristrami in many features but all lack the short tail and unkeeledupper caudal scales of this form and also frequently have an azygos shield between theprefrontal scales. In A. erythrurus and A. savignyi the lateral clavula of the hemipenis is deeply bifurcate atits tip, the frontonasal scale is quite often divided and the subocular may be separated fromthe lip. A. erythrurus has three widespread forms: A. e. erythrurus in Spain and Portugal A. e.lineomaculatus Dumeril & Bibron, 1839 in lowland western Morocco and A. e. bellii Gray, RELATIONSHIPS OF ACANTHODACTYLUS 3 1 9 1845 in the adjoining higher country and in north Algeria. Each of these is well denned onexternal features and A. e. bellii is further distinguished by its usually simple lateral clavula.Two other subspecies are sometimes recognized: A. e. mauritanicus, Doumergue, 1901 ofextreme northern Algeria resembles A. e. bellii in most features including its lateral clavulaand differs only in its more obviously keeled dorsals; A. e. atlanticus Boulenger, 1918# has arestricted distribution in north Morocco between the ranges of A. e. bellii and A. e.lineomaculatus and may be an intermediate between them. A. savignyi bland Doumergue,1 90 1 of north Tunisia is very similar to A. e. lineomaculatus but A. s. savignyi which coexistswith A. erythrurus in north Algeria is more distinctive. It has a clearly pointed snout and abetter developed pectination on the digits than is usual in its close relatives which may berelated to its occupying more sandy habitats (Doumergue, 1901). It is possible that thedifferences between the two subspecies of A. savignyi are a result of character displacement(Arnold, 198 16). A. boueti and A. guineensis are generally like the above but the frontonasal is undividedand the subocular scale is never separated from the lip. Each species has a number ofdistinctive features. In A. boueti there are nine premaxillary teeth, the hemipenis is verysmall, the second supraocular scale is broken up and the parietal scales are partlyfragmented. A. guineensis has the medial side of the hemipenis and armature absent and apeculiar arrangement of nasal scales (p. 296), a feature that occurs rarely also in A.erythrurus. The A. pardalis group A. p. pardalis (Lichtenstein, 1823): Cyrenaica, north Egypt, Israel. A. p. bedriagai Lataste, 1881: High plateaux of north Algeria (provinces of Oran, Algiers and Constantine); closely related populations in west Morrocco and west Tunisia.A. maculatus (Gray, 1838): northeast Morocco, north Algeria, Tunisia, Tripolitania.A. spinicauda Doumergue, 1901: Arba Tahtani and El Abiodh Sidi Sheikh, northwest Algeria. Members of the A. pardalis group are largely confined to relatively hard compact substratessuch as clayey-sandy soils, clay, loess and salt flats (Blanc, 1980; Gauthier, 1967; Mosauer,1 934; data on labels attached to BM(NH) specimens). They all agree in the following characters: premaxilla with about seven teeth, notnarrowed and separated from the maxillae by a constriction, presacral vertebrae usually24-26 in males and 25-26 in females, fifth sternal rib often interrupted; hemipenis andarmature asymmetrical, often markedly so, lateral clavula often bifurcate at tip, lateralconnectors often thickened; first supraocular intact or not very much divided and the fourthfragmented, subocular separated from mouth, four or five upper labials anterior to centre ofeye, dorsals small and often more or less smooth, 12 or more ventrals in longest row acrossbelly, ventrals arranged in fairly straight longitudinal rows, three longitudinal rows of scaleson fingers, pectination on digits relatively weak although somewhat variable, young withnumerous dorsal stripes. Boulenger (1918a, 1921) treated the A. pardalis group as a single species within which herecognized five varieties: pardalis (his forma typica), bedriagai, maculatus, latastiiBoulenger, 1918a and spinicauda. Pasteur & Bons (I960) regarded spinicauda as a distinctspecies but felt that the other varieties were no more than individual variations and placedthem in the synonymy of the monotypic species A. pardalis. In fact, the introduction ofosteological and hemipenial data helps to confirm the validity of the majority of Boulenger'sdivisions and it appears that at least five entities within the A. pardalis group can berecognized (see Table 4 & Fig. 7), A. maculatus, as understood here, includes Boulenger'svar. latastii. This author restricted maculatus to often relatively small-bodied populations inwhich the posterior dorsal scales are frequently keeled and some individuals have lightstripes or rows of light spots. These are distributed along the coastal areas of Tripolitania and 320 E. N. ARNOLD Table 4 Variation within the Acanthodactylus pardalis group A. p. A. p. pardalis bedriagai Un-namedwest Moroccanpopulation A. A. maculatus spinicauda Hemipenis lacks medial lobe and medial section of armatureUsual number of presacral vertebrae: malesfemalesApproximate number of animals in which 5th sternal rib is interruptedNumber of ventral scales in longest transverse rowLongitudinal rows of ventral scalesKeeling on dorsal scales often quite strongEnlarged, tubercular scales on sides of tail base 24/2525(26) 2626 20/24 7/14 2626 4/9 2425 2426 12(14) 14(12,16) 12 27-34 31-36 30-34 48/62 7/10 12(14) 12 27-33 27-33 'excluding population from Tamesmida, see below. Tunisia and in northeast Algeria. The name latastii was applied to inland populations lyingmainly to the south in which body size is sometimes greater, dorsal scales are usually more orless smooth, the dorsal pattern tends to lack light stripes or spots and the snout is morepointed. However, there is considerable variation and many intermediates occur, so it seemsbest to regard these populations as a single unit. The geographical differences within thismay reflect change in vegetation and substrate conditions; the pattern of maculatus-typeanimals is more likely to be cryptic where vegetation is relatively dense whilst that typical oflatastii is probably cryptic in more open areas. The relationships between the forms described in Table 4 are not entirely clear, but it iscertain that they cannot all be assigned to a single species. Although no areas of certainsympatry are known, bedriagai, the form of the high plateaux of north Algeria, approachesthe more southern maculatus near Biskra (bedriagai Aures Mountains north of Biskra,BM(NH) 91.5.4.70-72; maculatus near Biskra, BM(NH) 1907.4.6.10-25, BM(NH)1920.1.20.791, BM(NH) 1969.2121-23). The two localities are probably within 100 km ofeach other but the forms retain their differences in hemipenial structure, usual number ofpresacral vertebrae and of ventral scales in the longest row across the belly, and in pattern. Itseems likely therefore that bedriagai and maculatus are distinct species, especially as thehemipenial differences between them could act as an isolating mechanism. The un-named west Moroccan population (north of Agadir, BM(NH) 1970.246^7; 20 kmnorth of Tiznit, BM 1970.248; 30km southwest of Goulmime, BM(NH) 1970.249-50) isgenerally similar to bedriagai and is probably conspecific with it but differs in the followingfeatures: snout often more pointed, posterior dorsal scales more pointed and keeled, ventralsin longest row across belly often 12 (not 14), upper labials anterior to centre of eye often four(frequently five in bedriagai), range of dorsal patterns somewhat different. It is possible thatthis population also contacts A. maculatus although evidence is lacking. Another A. bedriagai-\ike population occurs at Tamesmida (33.05N 8.23E) in westernTunisia (BM(NH) 1920.1.20.3018, discussed by Boulenger, 1921, p. 67). The two malespecimens available resemble this form very closely in osteological features and externalsbut the hemipenis is single-lobed and the median side of the armature is lacking. Close by at RELATIONSHIPS OF ACANTHODACTYLUS 321 11 zS * o J=0c/>w O. Si c S <L> 03 J H <U O g?t: JH c 43 O. && P <u c -r t/5 ^ Q -Tl 322 E. N. ARNOLD Furryanah (34.57N 8.35E), A. maculatus occurs (BM(NH) 1920.1. 20.30 18e) but, as theTamesmida specimens are not like this form in other respects, a hybrid origin for them seemsunlikely. Possibly they represent an isolate of the A. bedriagai stock that has undergonehemipenial modification in response to some previous threat of introgression, as seems tohave happened several times in Acanthodactylus (p. 304). Tamesmida is on the edge of theeastern extension of the high ground on which A. bedriagai is found. The form spinicauda is known only from Arba Tahtani ( = Arbaouat, 33.05N 00.35E) andnearby El Abior Sidi Sheikh ( = E1 Abiodh Sidi Cheikh, 32.53N 00.34E), oases at thesouthern foot of the Saharan Atlas in northwest Algeria. In general form, hemipenialstructure and pattern, it resembles A. maculatus populations in the region (Mecheria andLaghouat) but differs in having the lateral scales at the base of the tail greatly enlarged,especially in males, with their keels produced to form tubercles, this condition occurring inno other lacertid. The hind legs are also rather longer than in nearby A. maculatuspopulations, although this is partly due to the fact that relative hind-leg length decreaseswith size and spinicauda is quite small; three females radiographed all have 26 presacralvertebrae instead of the 25 usual in A. maculatus. The two forms do not appear to besympatric, so a direct test of the species status of spinicauda is not possible. It is certainlyvery like neighbouring A. maculatus but the three distinctive features, especially the uniquetail structure, suggest it may be best to regard it as a full species, at least for the present. In thesouth of its range, maculatus appears to have a disjunct distribution, occurring mainlyaround oases such as Ouargla, Ghardia and Leghouat. Presumably its range was once morecontinuous but has become fragmented since the climate of the area has become moreextreme. It is probable that A. spinicauda originated from such an oasis isolate. Typical A. pardalis show some variation, for instance animals from the Jebel el Akhdarregion of Cyrenaica appear to be smaller than samples to the south and east. Also, althoughEgyptian animals usually have 24 presacral vertebrae in males and 25 in females, inCyrenaica 25 is common in males and females sometimes have 26. While there is evidencethat bedriagai and maculatus are specifically distinct, it is less clear how they are related totypical A. pardalis. This is partly because neither is known to contact this form, there being agap in the known range of the A. pardalis group as a whole south of the Gulf of Sirte, Libya,although whether this is real or a collecting artifact is uncertain. Typical A. pardalisresembles bedriagai in hemipenial structure but differs in usual number of presacralvertebrae and ventral scales in the longest row across the belly (although the bedriagai-\ikepopulation in west Morocco matches pardalis in this last characteristic). On the other hand,A. pardalis resembles maculatus in these features and the nearest populations of each formare quite similar, although coastal maculatus west of the Gulf of Sirte are smaller and havemore strongly keeled dorsal scales than the closest pardalis. However, these two forms differradically in hemipenial structure which, coupled with size-difference in this area might wellpreclude interbreeding. This being so, it seems reasonable to regard maculatus as specificallydistinct from pardalis, although it cannot be ruled out that connecting populations may yetbe found in north Libya with intermediate hemipenes. Typical bedriagai and the similarwest Moroccan population are tentatively regarded as conspecific with pardalis as they arehemipenially similar and the differences between them are mainly in features showingsubstantial variation within each population. On hemipenial evidence, the Tamesmidapopulation may be specifically distinct from A. bedriagai but it seems safest to leave its statusundecided until more is known about it. The 1. scutellatus group A. scutellatus (Audouin, 1829): North Africa south to Senegal, Mauretania, Mali, Niger, N. Chad and N. Sudan; S. Israel, N. Arabia, S. and central Iraq.A. longipes Boulenger, 1918a: North Africa including parts of S. Morocco, Algeria, Libya, Mauretania, Mali, Niger and Chad. RELATIONSHIPS OF ACANTHODACTYLUS 323 A. aureus Gunther, 1903; Coasatal areas of NW. Africa from S. Morocco to Mauretania; also someneighbouring inland areas. The A. scutellatus group is largely associated with soft-sand habitats (Anderson, 1898;Doumergue, 1901; Flower, 1933; Mosauer, 1934; Gauthier, 1967; Blanc, 1980; field labelson BM(NH) material); this is also suggested by the extensive pectination usually present onthe toes. However occasional populations are found on harder substrates (Scortecci, 1946).In north Africa the group occupies many desert areas and, unlike A. boskianus in this region,is not largely confined to relatively mesic enclaves. A. scutellatus and its relatives have fewfeatures plesiomorphic within Acanthodactylus but many apomorphic ones including thefollowing: premaxilla often with only five teeth and abruptly narrowed, usual number ofpresacral vertebrae 23 or 24 with very little sexual variation, fifth sternal rib interrupted in ahigh proportion of individuals; medial lobe of hemipenis, medial branch of sulcus andmedial side of armature absent or greatly reduced, clavula with a ID -shaped cross section;subocular scale separated from mouth, five or six upper labial scales anterior to centre of eye(four in A. aureus), (10)12-18 ventral scales in longest row across belly; ventrals tessellated,four longitudinal rows of scales on fingers, pectination on digits usually strong; young oftenuniform above. The populations that constitute the A. scutellatus group show great variation in somefeatures: maximum adult size ranges from about 50 mm from snout to vent to nearly 80 mm,dorsal scales may be coarse and keeled or fine and smooth, transverse counts at mid-bodyranging from 36 to 100, number and degree of tessellation of ventral scales varies as doesamount of pectination on digits and dorsal pattern ranges from stripes or rows of spots toreticulation or uniformity. From analogy with other species in the genus, keeled dorsalscales, low scale counts, reduced ventral tessellation, limited pectination and striped patternsare all likely to be associated with habitats that have harder substrates and more extensivevegetation, whereas contrasting conditions may be typical of soft, open sand. Although the A. scutellatus group is easily defined, recognition of taxa within it is difficult.Boulenger, (19180, 1921) treated it as a single species with six varieties. Bons & Girot (1962)pointed out that, as some of these were sympatric, they could not be regarded as meresubspecies and suggested the following arrangement for Boulenger's taxa; they also includeda form named by Haas (1957) and another described by themselves. A. scutellatus scutellatus (Audouin, 1829): Egypt, Israel, Sudan. A. scutellatus hardyi Haas, 1957: Arabia, S. Iraq. A. scutellatus audouini Boulenger, 1918#: S. Egypt, Libya, S. Tunisia, S. Algeria. A. longipes longipes Boulenger, 1918#: S. Libya, Algerian Sahara. A. longipes panousei Bons & Girot, 1962: SE. Morocco. A. inornatus inornatus (Gray, 1838): N. Libya, Tunisia, N. Algeria, S. Morocco. A. inornatus aureus Gunther, 1903: Atlantic coast from S. Morocco to Port Etienne. A. dumerilii (Milne-Edwardes, 1829): Senegal, Mauretania, Niger. But this system too presents problems for there is no real evidence that A. scutellatus, A.inornatus and A. dumerilii act as good species, nor is it clear that A. i. inornatus and A. i.aureus are conspecific. Difficulties in the interpretation of the A. scutellatus group arisepartly because intrapopulational variation is great and there are large areas in its huge rangewhere material is unavailable. Problems also stem from the geographical structure of thegroup. The soft-sand habitats favoured by these lizards are by no means continuous andconsist of a series of isolates and semi-isolates which have almost certainly had a complicatedhistory of contraction and expansion, and contact and isolation (see, for instance Sarnthein,1 978). The lizard populations associated with these different, partly discrete habitat units arelikely to have been subjected to differing selective regimes and even neighbouringpopulations may show differences. However, although the A. scutellatus group exhibits great 324 E. N. ARNOLD variation, it is mainly in a few characters that are likely to be ecologically correlated and itseems possible that similar morphotypes may have developed independently in some cases. Sometimes, populations may have evolved differences while isolated that enable them toact as good species if contact is restored. Such areas of sympatry or parapatry would providethe best evidence of speciation within the A. scutellatus group but, in the stringent environ-ments inhabited by these lizards, the niche space they usually occupy may often be able tosupport only one species, so that areas of extensive sympatry may be uncommon and regionsof contact will frequently be very restricted. Given the poor sampling available, places wheretwo species occur together will not be very easy to find. Nevertheless they do exist and thereare others where quite different populations approach each other very closely. On the basisof such localities, the A. scutellatus group is best divided into three species A. scutellatus,A. longpipes and A. aureus. A. scutellatus As understood here, this widespread species consists of populations in which there aretypically less than 70 dorsal scales in a transverse row at mid-body and not usually more than1 4 ventral scales in the longest row across the belly (exceptions to both frequent in Egypt andneighbouring areas), five or six upper labial scales anterior to the centre of the eye,premaxilla typically with five teeth, most usually 24 presacral vertebrae and dorsalcoloration highly variable. Variation is summarized in Table 5. The name A. scutellatus wasoriginally applied to the often large, frequently reticulated, fine scaled populations ofnorthern Egypt. The var. audouini of Boulenger is said to differ in usually having spots orvermiculations and in its coarser, keeled dorsal scales but individuals assignable to these twoforms occur in the region of Wadi Haifa, Sudan, alongside intermediates and animals fromKufra cannot be convincingly assigned to one form or the other. Other examples attributedby Boulenger to var. audouini occur on the coast of Tripolitania but extensive material nowin the British Museum (Natural History) shows that these intergrade with other samplesassignable to the small var. inornatus. The snout-length difference between these formsmentioned by Boulenger is very difficult to demonstrate and is complicated by allometricchanges. It seems very improbable, on present evidence, that two species are represented bythese samples. Animals from Mauretania and Senegal were assigned by Boulenger to var. dumerilii,regarded as a full species by Bons and Girot. This form is characterized by usually small size,coarse dorsal scales, low ventral number and sometimes striped dorsal patterns. But, ifsamples are compared, there is a gradual change from south to north the numbers of dorsalsand ventrals increasing and striping disappearing (Table 5, localities N-J). The mostnorthern animals are extremely similar to the next available samples in northern Algeria andthere are certainly no differences that would suggest they are not conspecific. In summary, the forms scutellatus, audouini, inornatus and dumerilii are not discrete and,on present evidence, cannot be separated at the species level. The pattern of geographicalvariation in A. scutellatus seems too complex to assign subspecies names in any consistentway, at least at present. A. longipes This species can be distinguished from sympatric or parapatric A. scutellatus by its highermid-dorsal transverse scale count (usually over 70 and always over 65), typically greaternumber of ventral scales in the longest row across the belly (often more than 14) andfrequently lower presacral vertebral count (usually 23); the dorsum is typically very pale,either plain or with a reticulation on the flanks. These differences, together with its usuallygreater pectination on the toes, suggest that it occupies softer, more open sand than A.scutellatus, where they occur together. A number of contact or approach areas are known.In northern Algeria, there may be considerable geographical overlap but in Mauretania onlya narrow abutment seems to be present with A. scutellatus in the west and A. longipes in the RELATIONSHIPS OF ACANTHODACTYLUS 325 RELATIONSHIPS OF ACANTHODACTYLUS 327 S II a a Q. ^I "^ II j=.s ffl ftu la c^|1 II c So ? o a I rj 00 J3 la ^ J3 .2 00 328 E. N. ARNOLD east. In south Libya sympatry may be quite extensive, as indicated by Scortecci, 1946 (hisgroup A is apparently A. scutellatus and group B is A. longipes). This author's observationsgive some support to the ecological separation between the two species suggested above. Localities for A. longipes are very scattered and, given the disjunct nature of the sort ofhabitat that it occupies, this species may consist of several quite isolated populations. A. aureus The populations found along the Atlantic coast of northwest Africa are all rather similar andin many respects resemble neighbouring A. scutellatus (referred by Bons and Girot to A. i.inornatus). They differ most obviously from this species in having only four upper labialscales anterior to the centre of the eye, seven teeth in the premaxilla and the snout is oftenmore acuminate. Such animals, typical of A. aureus, may occur some way from the coast(e.g. at Uedi Taamia, 26.0 IN 13.12W; EBD 2450) but at present they are not known tocontact A. scutellatus. However, at two localities in Mauretania, the latter occurs alongsideanother form that may well be conspecific with A. aureus. If so this provides some evidencethat the latter is a good species. The animals concerned are from Fort Gouraud (cfcf,BM(NH) 1982.292-93) and Atar (9, BM(NH) 1982.294) and are compared with sympatricA. scutellatus in Table 7. They are very similar to typical A. aureus in most features and itseems likely that they are referrable to this species. The inter-relationships of the three species that make up the A. scutellatus group areunclear. A priori, A. aureus might be considered the most primitive, often having sevenpremaxillary teeth and four upper labials anterior to the centre of the eye. But, if theMauretanian specimens with five premaxillary teeth are assignable to A. aureus, thisapomorphic feature cannot be used to unite A. scutellatus and A. longipes as sister species.Also, the presence of four, instead of five or six upper labials, may be secondary, for otherspecies with acuminate snouts (A. gongrorhynchatus, A. masirae] have lower labial scalecounts than expected from their phylogenetic relationships judged by other features. Table 7 A. aureus: comparison of samples with neighbouring A. scutellatus Fort Gouraud, Atar Coastal area scutellatus aureus aureus Upper labial scales anterior to centre of eye 5 or 6 4 4 Dorsal scales across mid-body 45-56 38-46 41-60 Lower temporal scales large, often large, usually small, keeled smooth smooth Enlarged gular scales bordering 4th and 5th chin shields frequent none rare Usual number of rows of granules beneath centre of collar 2 or 3 or 1 Usual number of premaxillary teeth 5 57 Pattern of males Fine spots or coarse spots coarse spots or reticulation vermiculation RELATIONSHIPS OF ACANTHODACTYLUS 329 The A. opheodurus group A. opheodurus Arnold, 19806: Arabia (but not apparently the southern desert or northern Oman), S. Israel, Jordan, parts of Iraq. A.felicis Arnold, 19806: S. Arabia (Republic of South Yemen and Dhofar, S. Oman)A. masirae Arnold, 19806: S. Oman. This assemblage consists of three small species, all wholly or largely confined to Arabia. Allhave only recently been described for, although A. opheodurus and A. felicis have beenpresent in museum collections for some time, they have been confused with A. boskianus.These lizards appear to be associated with relatively hard substrates and have restrictedpectination on the digits. The group possesses a high proportion of plesiomorphic features: premaxilla with aboutseven teeth and not abruptly narrowed (except in A. masirae), usually 24 presacral vertebraein males and 25 in females, fifth sternal rib usually intact; lateral clavula simple, hemipenialconnectors unthickened; first supraocular more or less intact, typically four upper labialscales anterior to the centre of the eye, eight or ten ventral scales in longest row across belly,ventrals arranged in straight longitudinal rows, three longitudinal rows of scales aroundfingers (a fourth row in A. masirae), pectination on digits not strong, upper caudal scaleskeeled, young striped dorsally, the stripes sometimes persisting in adults. However thehemipenis is highly modified in the A. opheodurus group, the medial lobe and medial side ofthe armature being greatly reduced, the fourth supraocular scale is at least partly broken upand the subocular scale is often separated from the mouth (not in A. masirae) and thenumber of stripes in the juvenile pattern shows some reduction. Dorsal scaling is relativelycoarse, there being 25-42 scales in a transverse row across the mid-body. The three species are grouped together largely on the grounds of overall similarity. Inter-relationship of the species of Acanthodactylus The approach used for estimating the phylogeny of the species of Acanthodactylus is brieflydiscussed on p. 293. Probably derived character states shared by two or more species(provisional synapomorphies) are listed in Table 8 and their distribution shown in Table 9.Polarity, that is which state of a character is primitive and which state or states derived, hasbeen decided largely by outgroup comparison, using the rest of the Lacertidae as theoutgroup. Some assessment of the relative reliability of characters as indicators of relation-ship (character weighting) was attempted employing the indicators mentioned by Arnold(198 la). No features scored very well on this basis and most scored quite badly but amongthe better ones were the following (numbers refer to Tables 8 & 9, and Fig. 9): premaxillaryteeth reduced to five (1); hemipenial features not thought to be connected directly with thedevelopment of physical isolating mechanisms against interbreeding discussed on p. 304 (9,1 1-1 7); reduction in size of the ear opening (24); no keeling on proximal dorsal caudal scales(30). These features are consequently given some precedence in situations where theevidence of different derived character states conflicts. Because many features are poly-morphic, with both the primitive and a derived state occurring together in the same species,Le Quesne's (1969) method for determining compatible characters could not be used. As wellas joint possession of derived character states, strong overall resemblance is also taken asprima facie evidence of close relationship. Using these indicators, a provisional partialphylogeny of Acanthodactylus was constructed and is shown in Fig. 9. The relationships of the A. cantoris group, A. gongrorhynchatus and A. haasi are discussedon p. 3 1 1 . By and large, the hypothesis of their relationships shown in Fig. 9 involves apattern of shared apomorphies that does not include much homoplasy. As stated, the mainconflict involves the position of A. schmidti: is it more closely related to A. arabicus or to A.blanfordiil The former relationship is supported by total loss of the medial side of thehemipenis and armature (8d, lOc) and some increase in the number of ventral scales (26b), 330 E. N. ARNOLD Table 8 Probable derived character states shared by two or more species of Acanthodactylus. Wherefeatures appear to constitute a transformation series, they are listed in their supposed order of originand denoted by a consecutive letter thus: a, b, c, d. 1 . Premaxilla narrowed: a. somewhat; b. distinctly. 2. Premaxillary teeth reduced to five. 3. Presacral vertebrae 23 or 24 in females. 4. Presacral vertebrae 26 or 27 in females. 5. On average, females have less than one more presacral vertebra than males: a. 0-66-0-84 morevertebrae; b. 0-07-0-33 more vertebrae. 6. Fifth sternal rib interrupted in over 50% of individuals. 7. Hemipenis small. 8. Medial lobe of hemipenis reduced: a. somewhat reduced; b. more strongly reduced; c. very small;d. minute or absent. 9. Proximal lip of medial branch of hemipenial sulcus reduced to a fold in symmetrical hemipenes. 10. Medial side of armature reduced: a. somewhat reduced; b. more strongly reduced; c. reduced to athread or absent. 1 1 . Medial clavula narrow and pointed with a ^ -shaped cross section. 12. Lateral clavula with a proximally directed pocket. 13. Lateral clavula very narrow. 14. Lateral clavula complexly structured with multiple lobes below. 1 5. Lateral clavula complexly structured and divided at tip. 1 6. Lateral clavula folded with D-shaped cross section. 1 7. Most medial connector on lateral side of hemipenis thickened. 1 8. Nostril separated from first upper labial scale by a subnasal. 19. Frontonasal scale longitudinally divided. 20. One or two azygos scales present between the prefrontal scales. 2 1 . Supraocular scales not all intact: a. an area of granules wedged between the third and fourthsupraoculars and the latter sometimes divided; b. first supraocular divided into two or three, fourthsupraocular very fragmented; c. first and fourth supraoculars very fragmented; d. first, second andfourth supraoculars very fragmented. 22. Subocular scale usually separated from mouth. 23. Upper labial scales anterior to centre of eye more than four. 24. Ear opening reduced in size. 25. Dorsolateral tracts of enlarged scales on posterior body. 26. Maximum number of ventral scales in a transverse row: a. 12; b. 14 or more. 27. Ventral scales tessellated: a. at sides only; b. generally. 28. Ventral scales grade into dorsals: a. to a small extent; b. more generally. 29. Four longitudinal rows of scales on the fingers: a. anterior row irregular; b. anterior row regularand continuous. 30. No keeling on proximal dorsal caudal scales. 3 1 . Tail less than 1-5 times snout-vent distance. 32. Occipital stripes fuse to form a mid-dorsal band (often not visible in adults): a. fusion does notextend to occiput; b. fusion extends to occiput. 33. Young without pattern of light and dark stripes. 34. Reddish-brown spots present in pattern that do not fade in alcohol. 35. Two rows of large ocellar markings along back. while affinity to A. blanfordii is suggested by similarity in lateral clavula structure (12) andthe presence of dorsolateral tracts of enlarged scales (25). The former characters appearweaker indicators of relationship, especially as they have developed several times in theLacertidae, while characters 12 and 25 are known nowhere else. A. schmidti is consequentlyregarded as the sister species of A. blanfordii. Whether A. haasi is the sister species of A.gongrorhynchatus, which is suggested by its slender habitus, reduced ear opening (24) andsome features of the lateral clavula (p. 306), or a hybrid between this species and anothersuch as A. opheodurus, is discussed elsewhere (p. 314). 331 332 E. N. ARNOLD RELATIONSHIPS OF ACANTHODACTYLUS 333 V V ""X / T-1b,2.3.5. 19,22\ / 27b. 29, 33 \ / 16 -30.31.35 \ / . 15, 21 c Fig. 9 Tentative hypothesis of the relationships of the species of Acanthodactylus. Figures refer toderived characters listed in Tables 8 and 9. Lines above species names join forms that have avery close overall resemblance, vindicates alternative position for the A. scutellatus group (A.aureus, A. scutellatus and A. longipes). A number of Acanthodactylus species share a distinctive hemipenial structure whichincorporates two derived features not found elsewhere: the medial clavula is narrow andpointed with a ^ -shaped cross section (1 1) and the most medial connector on the lateral sideis often thickened (17); in addition there is moderate asymmetry of the armature (lOa). Thispattern occurs in A. boskianus, A. schreiberi, the A. grandis complex, A. (t.) tristrami, A. (t.)orientalis, A. robustus, A. erythrurus, A. savignyi, A. boueti and A. pardalis. It seemsprobable that these forms constitute a monophyletic group in Ashlock's (1974) sense withinAcanthodactylus. Among them A. boskianus, A. schreiberi and the A. grandis complex allhave a reduced proximal lip to the medial branch of the hemipenial sulcus (9) and a lateralclavula that is complexly lobed (14). Within this trio, A. boskianus and A. schreiberi maywell be sister species, for some populations are extremely similar. A. grandis shows conflictof evidence as to its relationships. It has some similarity in derived features to members of theA. cantoris group, viz. high number of ventral scales (26) which are tessellated (27b) and fourlongitudinal rows of scales along the fingers (29b). However, weighting criteria (especiallyfrequent occurrence in groups not closely related to Acanthodactylus} suggest these featuresare relatively weak indicators of relationship, whereas it has been argued (p. 329) that most ofthe conflicting hemipenial features (9, 11, 14, 17) that ally the A. grandis complex to A.boskianus and A. schreiberi are likely to be of greater reliability. Moreover, any detailedplacement of the A. grandis complex in the A. cantoris group would involve additionalhomoplasies. Members of the A. tristrami group share features not found elsewhere in the genus, viz. nokeeling on the proximal, dorsal caudal scales (30), a short tail (31) and an often distinctivepattern (35). The very similar A. (t.) tristrami and A. (t.) orientalis could well be sister species.A. robustus has some features that suggest relationships elsewhere: subocular scale separatedfrom the mouth (22), increased number of ventrals (26a) and an irregular fourth scale row onthe fingers (29a), but none of these derived states is very strongly developed and all seemlikely to be relatively weak indicators of relationship compared with those allying A.robustus with A. (t.) tristrami and A. (t.) orientalis. The A. tristrami and A. erythrurus groups share some distinctive features, in particular ahigh number of presacral vertebrae (4) and fragmented first and fourth supraocular scales 334 E. N. ARNOLD (2 1 c). Furthermore, some species of each have a lateral clavula that is bifurcate at its tip ( 1 5),so it seems reasonable to regard the two assemblages as closely related. The members of theA. erythrurus group are all generally similar and at least some individuals of all species haveone or more azygos shields between the prefrontal scales (20) although this condition alsooccurs occasionally in other forms. It is suggested that A. erythrurus and A. savignyi are sisterspecies on the evidence of their great similarity and frequent division of the frontonasal scale(19). They are also the only members of the group in which the subocular scale may beseparated from the lip (22). A. boueti and A. guineensis both have distinctive features of theirown (A. boueti nine premaxillary teeth, fragmented parietal scales, small hemipenis (7),supraocular scales very fragmented (2 Id); A. guineensis very asymmetrical hemipenis andarmature (8d, lOc), lateral clavula rolled) but their precise relationships to each other and toA. erythrurus and A. savignyi are not clear. All A. guineensis examined (n = 20) have apeculiar arrangement of nasal scales (18) which occurs as a rarity in A. erythrurus (p. 296).However, this feature cannot be regarded as strong evidence of the close relationship of thesespecies within the group, for its rarity in A. erythrurus raises the possibility that its apparentabsence in A. savignyi and A. boueti may be due to inadequate sampling or to pseudoreversal(Arnold, 198 la). The A. pardalis group seems likely to be a monophyletic assemblage on the basis of thestrong overall similarity of its members. They also have more hemipenial asymmetry (8b or8c) than the species placed in the A. grandis complex and the A. boskianus, A. tristrami andA. erythrurus groups, share a high incidence of interruption of the fifth sternal rib (6) and thenumber of ventral scales is also regularly high (26). The main reason for allying the A.pardalis assemblage with these groups is that the species with the least modified hemipenis,A. pardalis, has an armature just like the others with derived features lOa, 11 and 17. Theabsence of these features in the other two species in the A. pardalis group, A. maculatus andA. spinicauda, may be secondary, the result of the development of physical isolatingmechanisms (p. 304). The A. pardalis group shares with the A. tristrami and A. erythrurusgroups frequent bifurcation of the tip of the lateral clavula (15), some fragmentation of thesupraocular scales (2 Ib) and some tendency for high presacral vertebral counts (4). However,the last feature is not universal and the supraoculars are less broken up than in these forms.Because of this, the A. pardalis group is tentatively placed as a sister assemblage to them.Within the A. pardalis group, A. maculatus and A. spinicauda may be closely related havingstrongly asymmetrical hemipenes and armatures (8d, lOc) with the lateral clavula foldedsideways (16). Holophyly of the A. scutellatus group is supported by the close similarity of its membersand their possession of a unique synapomorphy; premaxillary teeth reduced to five (2). Itsmembers have several other derived features that occur elsewhere and these suggestconflicting hypotheses as to the closest relatives of the group. The main candidates are A.schmidti, particularly the populations in the United Arab Emirates that have little sexualvariation in the number of presacral vertebrae, and all or part of the A. pardalis group.Derived features that these share with the A. scutellatus assemblage are set out in Table 10;the A. grandis complex shares a much smaller number, namely 26, 27 and 29. Most derivedfeatures, eleven, are shared with A. schmidti populations from the United Arab Emirates andonly seven with the A. pardalis group of which no more than six occur in any one species.Thus on simple count of shared characters A. schmidti would be considered clearly the morelikely sister taxon, but the likely quality of the features as indicators of relationship shouldalso be taken into account. In general, the features shared with A. schmidti seem rather weakas they score badly on the weighting criteria listed by Arnold (198 la). For instance, numbers3, 5, 6, 8d, lOc, 23, 26, 27, 29 and 33 occur in a number of other lacertid stocks; there areadditional functional reasons for thinking 8d and lOc may have evolved more than once(p. 304); if the argument on p. 333 that the A. grandis complex is more closely related to theA. boskianus group than to the A. cantoris assemblage is accepted, then characters 23, 26, 27and 29 must have evolved twice within Acanthodactylus and a third independent origin RELATIONSHIPS OF ACANTHODACTYLUS 335 Table 10 Comparison of derived features shared by the A. scutellatus group with A. schmidti and theA. pardalis group. A. schmidti A. pardalis group Main United Arab A. scutellatus A. maculatus range Emirates group A. pardalis A. spinicauda Ib Premaxilla narrow + + + 3 Presacral vertebrae 23 or 24 in females + + 5b Little variation in number of presacral vertebrae + + + 6 Fifth sternal rib often interrupted + + + 8 Medial lobe of hemipenis very reduced + + + + 1 Oc Medial side of armature very reduced + + + + 1 6 Lateral clavula folded to produce a z> -shaped cross section + + 23 High number of upper labials + + + + + 26 Increased number ofventrals + + + + + 27 Ventrals tessellated, at least at sides + + + 29 Four scale rows on fingers + + + 33 Young without stripes + + +,Strong pectination on toes + + + would not seem unlikely; a number of features seem to be functionally related to theproblems of living in the open, soft-sand habitats occupied by A. schmidti and the A.scutellatus group, this is true of Ib, 3, 29, 33 and strong pectination on the digits; the coerciveselective forces likely to produce convergence of these features are discussed in the relevantcharacter descriptions. Finally, two features, 3 and 5, are confined in A. schmidti to a verysmall part of the geographical range, the United Arab Emirates. The restricted distributionof these characters suggest they may well have developed independently within the species. Ifso, the number of features to be considered is reduced to nine. On the other hand therelationship of the A. scutellatus assemblage to the A. pardalis group is supported by anapparently unique hemipenial feature (16) and the high incidence of interrupted sternal ribs,a condition not found elsewhere in the genus. Assessing such conflicting evidence isinevitably difficult but, on balance, the author is inclined to believe that the A. scutellatusgroup is most closely related to the A. pardalis assemblage. As stated, the members of the A. opheodurus group are placed together largely on thegrounds of overall similarity which is especially marked between A. opheodurus and A.felicis. Shared derived features are 8d, lOc and in some individuals of each species 32a; noneof these are unique to the group. Evidence for the inter-relationship of the species isconflicting: A. felicis and A. opheodurus share 5a and 22 while the latter species and A.masirae share 1 and 29, although 29 occurs only in a reduced form in A. opheodurus andonly in a minority of individuals. The lack of strong derived characters makes it difficult to relate the A. opheodurus groupto other Acanthodactylus, but absence of a thickened hemipenial connector (1 7) suggests thatit is not directly related to the A. grandis-A. guineensis sequence (see Fig. 9). Nor is it clearhow this sequence relates to A. micropholis and the A. cantoris assemblage. 336 E. N. ARNOLD Acknowledgements The author is very grateful to the various collectors who, in the past few years have donatedoften critical series of Acanthodactylus to the British Museum (Natural History). Theyinclude D. Vesey Fitzgerald, M. D. Gallagher, J. and P. Gasperetti, D. J. Greathead,K. M. Guichard, M. C. Jennings, M. R. K. Lambert, J. P. Mandeville, S. Moult, G. V.Popov, T. D. Rogers, T. D. Stoner, W. Thesiger and J. O. Wade. The following curators were kind enough to lend material in their care: E. R. Brygoo,(Museum Nationale d'Histoire Naturelle, Paris), J. Castroviejo and P. W. Hopkins (EstacionBiologica de Donana, Seville), M. S. Hoogmoed (Rijksmuseum van Natuurlijke Historic,Leiden), A. E. Leviton (California Academy of Sciences, San Francisco), C. J. McCoy(Carnegie Museum, Pittsburgh), H. Marx (Field Museum of Natural History, Chicago),F. Tiedemann (Naturhistorisches Museum, Vienna), E. E. Williams and P. Alberch(Museum of Comparative Zoology, Harvard), H. Marx (Field Museum, Chicago). Some of the radiographs used in this study were produced by B. T. Clarke. References Anderson, J. 1895. On a collection of reptiles and batrachians made by Colonel Yerbury at Aden andits neighbourhood. Proc. zool. Soc. Land.: 63 5-663. 1898. Zoology of Egypt: Reptilia and Batrachia. London, lxiv+ 371 pp. Anderson, S. C. 1963. Amphibians and reptiles from Iran. Proc. Calif. Acad. Sci. 31 : 417-498. 1974. Preliminary key to the turtles, lizards and amphisbaenians of Iran. Fieldiana Zool. 65 : 27-44. Angel, F. 1936. Reptiles et batraciens de Syrie et de Mesopotamie recoltes par M. P. Pallery. Bull. Inst. EgyptelS: 107-116.Arnold, E. N. 1973. Relationships of the Palaearctic lizards assigned to the genera Lacerta, Algyroides and Psammodromus (Reptilia: Lacertidae). Bull. Br. Mus. nat. Hist. (Zool.), 25 : 289-366. 1980<2. Reptiles of Saudi Arabia: A review of the lizard genus Stenodactylus (Reptilia: Gekkonidae). Fauna of Saudi Arabia, vol. 2 : 368-404. Basle: Pro Entomologica and Ciba Geigy. 19806. The reptiles and amphibians of Dhofar, Southern Arabia. J. Oman Stud. Spec. Rep. No. 2 : 273-332. 198 la. Estimating phylogenies at low taxonomic levels. Zeit. zool. Syst. Evolut.-Forsch. 19 : 1-35. 19816. Competition, evolutionary change and montane distributions. In Greenwood, P. H. & Forey, P. L. (Eds) The Evolving Biosphere. London, 2 1 7-228.in press. Structure and function of the lacertid hemipenis. Ashlock, P. D. 1974. The uses of cladistics. Ann. Rev. Ecol. Syst. 5 : 8 1-99.Audouin, J. V. 1829. Description de I'Egypte, Reptiles. Supplement. Paris.Bedriaga, J. V. 1884. Die neue Lacertiden-Gattung Latastia und ihre Arten (L. doriai n. sp., var. Martensi m., Samharica Blanf. und Boscai n. sp.). Annali Mus. civ. Stor. nat. Giacomo Doria 20 : 307-324.Blanc, C. P. 1980. Studies of the Acanthodactylus of Tunisia IV: geographic distribution and habitats. J.Herpet.14: 39 1-398.Blanford, W. T. 1874a. Descriptions of new lizards from Persia and Baluchistan. Ann. Mag. nat. Hist. (4)13:453-5.1 8746. The Zoology and Geology. In Goldsmid, F. J. Eastern Persia, an account of the journeys of the Persian Boundary Commission, 1870-72. London. Boettger, O. 1879. Reptilien und amphibien aus Syrien. Ber. senckenb, naturf.Ges. 1878-79 : 57-84.Bons, J. & Girot, B. 1962. Revision de 1'espece Acanthodactylus scutellatus (Lacertide-Saurien). Bull. Soc. Sci. nat. Maroc. 42 : 3 1 1-334.Boulenger, G. A. 1 878. Sur les especes $ Acanthodactylus des bords de la Mediterranee. Bull. Soc. zool. Fr.3: 179-201.1887a. Descriptions of new reptiles and batracians in the British Museum (Natural History) Part 3. Ann. Mag. nat. Hist. (5) 20 : 50-53.18876. Catalogue of the lizards in the British Museum (Natural History), vol. 3. London, xii + 575 pp.1909. Description of a new lizard of the genus Acanthodactylus from Syria. Ann. Mag. nat. Hist. (8)4: 188-189. RELATIONSHIPS OF ACANTHODACTYLUS 337 1 9 1 8a. Sur les lezards du genre Acanthodactylus Wiegm. Bull. Soc. zool. Fr. 43 : 143-1 55. 19186. A synopsis of the lizards of the genus Eremias. J. Zool. Res. 3 : 1-12. 1918c. Description of a new lizard of the genus Acanthodactylus from Mesopotamia. J. Bombaynot. Hist. Soc. 25 : 373-374, 1919. On a new variety of Acanthodactylus boskianus, Daud., from the Euphrates. Ann. Mag.nat.Hist. (9)3:549-550. 1921. Monograph of the Lacertidae, vol. 2. London, viii + 451. Chabanaud, M. P. 1917. Enumeration des reptiles non encore etudies de 1'Afrique occidentale, appartenant aux collections du Museum, avec la description des especes nouvelles. Bull. Mus. Hist. nat. Paris 191 7: 83-105.Clark, R. J., Clark, E. D., Anderson, S. C. & Leviton, A. E. 1969. Report on a collection of amphibians and reptiles from Afghanistan. Proc. Calif. Acad. Sci. 36 : 279-3 16.Daudin, F. M. 1802. Histoire naturelle . . . . des Reptiles, vol. 3. Paris.Doumergue, F. 1901. Essai sur la faune erpetologique de 1'Oranie. Soc. Geogr. Arch. Oran. 19: 197-260,501-32,Dumeril, A. M. C. & Bibron, G. 1839. Erpetologie generale ou histoire naturelle complete des reptiles, vol. 5, Paris 854 pp.Duvdevani, I. & Borut, A. 1974a. Mean body temperature and heat absorption in four species of Acanthodactylus lizards (Lacertidae). Herpetologica 30 : 1 76-18 1 .19746. Oxygen consumption and evaporative water loss in four species of Acanthodactylus (Lacertidae). Copeia 1974: 155-164.Elmer, T. 1881. Untersuchungen iiber das Variiren der Mauereidechse, ein Beitrag zur Theorie von der Entwicklung aus constitutionellen Ursachen, sowie zum Darwinismus. Arch. Naturgesch. 47 : 239-517.Flower, S. S. 1933. Notes on the recent reptiles and amphibians of Egypt, with a list of the species recorded from that kingdom. Proc. zool. Soc. Lond. 1933 : 735-85 1 .Gauthier, R. 1967. Ecologie et ethologie des reptiles du Sahara Nord-occidental (region de Beni- Abbes). Annls Mus. roy. Afr. centr. 155 : 1-83.Gray, J. E. 1838. Catalogue of the slender-tongued saurians, with descriptions of many new genera and species. Ann. Mag. nat. Hist. 1 : 274-283.1845. Catalogue of the specimens of lizards in the collection of the British Museum. London, xxviii + 289. Gugg, W. 1939. Der Skleralring der plagiotremen Reptilien. Zool. Jb. Abt. Anal. 65 : 339-416.Giinther, A. C. L. G. 1864a. Report on a collection of reptiles and fishes from Palestine. Proc. zool. Soc. Lond. 1864:488^93. 1 8646. The reptiles of British India. London, xxvii + 452. 1903. Reptiles from Rio de Oro, Western Sahara. Novit. zool. 10 : 298-299. Haas, G. 1957. Some amphibians and reptiles from Arabia. Proc. Calif. Acad. Sci. 39: 47-86.and Werner, Y. L. 1969. Lizards and snakes from southwestern Asia, collected by Henry Field. Bull. Mus. comp. Zool. Harv. 138 : 327-405.Hennig, W. 1950. Grundzuge einer Theorie der phylogenetischen Systematik. Berlin. 1966. Phylogenetic systematics. Urbana. Lantz, L. A. 1928. Les Eremias de TAsie occidentale. Bull. Mus. Georgie, 4 and 5. Lataste, F. 1881. Diagnosis des reptiles nouveaux d'Algerie. Naturaliste 1881 : 357-59. Leviton, A. E. & Anderson, S. C. 1967. Survey of the reptiles of the Sheikhdom of Abu Dhabi, Arabian Peninsula, part 2. Systematic account of the collection of reptiles made in the sheikhdom of Abu Dhabi by John Gasperetti. Proc. Calif. Acad. Sci. 35 : 157-192.Lichtenstein, M. H. C. 1823. Verzeichniss der Doubletten des Zoologischen Museums der ... Universitdt. Berlin.Matschie, P. 1893. Uber einige von Herrn Oscar Neumann bei Aden gesammelte u. beobachtete Saugthiere, Reptilien und Amphibien. Sbr. Ges. naturf. Freunde Berl. 1893 : 27-31.Mertens, R. 1968. Uber Reptilienbastarde, IV. Senckenberg. biol. 49 : 1-12. 1969. Die Amphibien und Reptilien West-Pakistans. Stuttg. Beitr, Naturk. 197 : 1-96. Milne-Edwardes, M. H. 1829. Recherches zoologiques pour servir a 1'histoire des lezards extraites d'une monographic de ce genre. Ann. Sci. nat. 16 : 50-89.Minton, S. A. 1966. A contribution to the herpetology of West Pakistan. Bull. Am. Mus. nat. Hist. 134 : 27-184.Monard, A. 1949. Vertebres nouveaux du Cameroun. Revue Suisse Zool. 56 : 73 1-745. 338 E. N. ARNOLD Mosauer, W. 1934. The reptiles and amphibians of Tunisia. Publs Univ. Calif. Los Ang. biol. Sci. 1 : 49-64.Pappenfuss, T. J. 1969. Preliminary analaysis of the reptiles of arid central west Africa. Wasmann J. Biol. 27 : 249-336.Pasteur, G. & Bons, J. 1960. Catalogue des reptiles actuels du Maroc. Trav. Inst. sclent, cherif. Serie Zoo/. 21: 1-132.Peters, W. 1854. Diagnosen neuer Batrachier welche zusammen mit der friiher (24. Juli und 17. August) gegedenen Ubersicht der Schlangen und Eidechsen mitgetheilt werden. Monatsb. Akad. Wiss. Berlin 1854 : 614-628.Quesne, W. J. Le, 1969. A method of selection of characters in numerical taxonomy. Syst. Zoo/. 18 : 201-205. Riney, T. 1953. Notes on the Syrian lizard Acanthodactylus tristrami orientalis. Copeia 1953 : 66-67.Rosevear, D. R. 1965. Bats of West Africa. London.Sarnthein, M. 1978. Sand deserts during glacial maximum and climatic optimum. Nature Lond. 272 : 43-46. Schinz, H. R. 1833. Naturgeschichte und Abbildungen der Reptilien. Leipzig, iv + 240.Schmidt, K. P. 1919. Contributions to the herpetology of the Belgian Congo based on the collection of the American Congo Expedition 1 909- 1915. Bull. A m. Mus. nat. Hist. 39 : 3 8 5-624. 1939. Reptiles and amphibians from southwestern Asia. Fieldiana Zoo/. 24: 49-89. Scortecci, G. 1946. Tentative di analisi biologica condotto sulla specie Acanthodactylus scutellatus Audouin. Riv. Biol. colon. 7 : 5-15. Shcherbak, N. N. 1974. Yashchurki Palearktiki. Kiev, 295 pp. Werner, F. 1929. Beitrage zur Kenntnis der Fauna von Syrien und Persien. Zoo/. Anz. 81 : 238-245.Wiegmann, A. F. A. 1834. Herpetologia Mexicana. Berlin, vi + 64 pp. Manuscript accepted for publication 4 August 1982 Addendum Since this paper was submitted for publication, I have had the opportunity to see the recentarticle on Acanthodactylus by Dr Alfredo Salvador. (1982). In the main, we are in agreementover species boundaries within the genus and a number of the divergencies that exist can beattributed to using different criteria for deciding if allopatric populations merit species statusor not, always a rather subjective decision. These and other more important disagreementsare briefly considered below. 1. The A. yemenicus group of Salvador is the same as the A. opheodurus group of Arnold(19806 and this paper) but with the addition of a new taxon, A. yemenicus. This form hasmuch in common with A. felicis. Unlike the other members of the group, its specific statuscannot be confirmed by sympatry, as its range is completely separate from the others. Thereare considerable differences between the two populations from which A. yemenicus has beendescribed, Ta'izz and the Sheikh Osman-Aden area, even although these are only separatedby some 1 30 km. 2. A. erythrurus group. Salvador treats bland as a full species rather than a subspecies of A.savignyi. Given the differences between this form and typical savignyi, this course is notunreasonable. 3. A. pardalis group. A. pardalis, as understood here, is divided by Salvador into three fullspecies: A. p. pardalis becomes monotypic as A. pardalis, A. p. bedriagai becomes A.bedriagai and the Moroccan population related to bedriagai becomes A. busacki n. sp.These forms overlap in many morphological features. 4. A. tristrami sens. lat. Salvador regards A. tristrami, in its wide sense, as consisting of twosubspecies: A. t. tristrami, with which A. t. orientalis is synonymized, and A. t. iracensis. Thiscontrasts with the arrangement adopted here where orientalis is regarded as separable fromtristrami and iracensis is synonymized with orientalis. A careful examination of the types ofall three forms and of most other available material convinces me that the latter course ismore appropriate. A. t. iracensis cannot be separated from orientalis on the basis of the RELATIONSHIPS OF ACANTHODACTYLUS 339 features mentioned by Schmidt (1939), namely lower transverse dorsal count, less vividcolour pattern and more sharply pectinate toes; nor does it differ significantly in any of theother features investigated. On the other hand, typical tristrami differs from orientalis,including iracensis, not only in its larger size, more numerous dorsal scales and hemipenialstructure, but usually also in its deeper snout with convex upper profile, absence of adenticulation on the anterior edge of the ear and virtual lack of pectination on the toes. Inaddition, the first supraocular is more fragmented (tristrami-n 10, average number offragments on each side 5.4, range 3-10; orientalis-n= 17, average number of fragments oneach side 2.8, range 1-6). 5. A. scutellatus group. Salvador divides A. scutellatus, as understood here, into A.scutellatus in the east and A. dumerilii in the west. The two supposed species are said toapproach each other in Mali and Algeria but not to integade. I have re-examined theBM(NH) material from Mali that Salvador assigns to A. scutellatus and A. dumerilii and canfind no differences in the features that are said to separate them, namely transverse ventralcount, snout shape and number of rows of supraciliary granules. Where A. scutellatus and Adumerilii approach each other around the Algerian-Tunisian border, I again find noconsistent differences in snout shape as mentioned by Salvador, and although there areaverage differences in degree of fragmentation of the fourth supraocular scale, there issubstantial overlap. I consequently regard the separation of A. dumerilii from A. scutellatusas unproven. Salvador divides his A. dumerilii into two subspecies: A. d. dumerilii in Senegal, southwestMauretania and Mali and A. d. exiguus in Algeria and Morocco. The former is said to have12 instead of 14 longitudinal rows of ventral scales, two instead of one row of supraciliarygranules and longer legs. In material I have examined, I find that two rows of supraciliarygranules are often absent in the area where A. d. dumerilii is reported and leg length isvariable. Furthermore, there is a very large intergrade area between the two subspecies whereother characters vary clinally. It scarcely seems worth while naming such poorly definedentities, especially when A. scutellatus varies so substantially elsewhere in its large NorthAfrican range and no subspecies names are assigned. 6. The diagnosis of Acanthodactylus given on p. 8 does not fully differentiate the genusfrom Meroles or Eremias. Salvador, A. 1982. A revision of the lizards of the genus Acanthodactylus (Sauria: Lacertidae). Bonn,zool. Monogr. 16: 1-167. British Museum (Natural History)1881-1981 Centenary PublicationsChance, change & challenge Two multi-author volumes from one of the foremost scientific institutions in the world.General Editor: P. H. Greenwood The Evolving Earth Editor: L. R. M. Cocks The Evolving Biosphere Editor: P. L. Forey In the first volume, The Evolving Earth, twenty scientists have been asked to reviewthe present state of knowledge in their particular field, ranging from the origin ofthe Earth, through ocean sediments and soils to continental drift and palaeogeography. In the companion volume, The Evolving Biosphere, museum scientists have chosenan evolutionary concept speciation, coevolution, biogeography etc. and relatedthis to the group of animals or plants in which they are specialising. Thus beetlesand birds exemplify sympatric and allopatric speciation, butterflies mimicry andcertain fishes explosive evolution. In both volumes the text is supplemented by over one hundred specially-commissionedpieces of two-colour artwork. These two books will be invaluable to all sixth-form and undergraduate biology andgeology students. The Evolving Earth: 276x219 mm, 280pp, 138 line illustrations, 42 halftonesThe Evolving Biosphere: 276x2 19 mm, approx. 320pp, 133 line illustrationsPublished: May 1981 Co-published by the British Museum (Natural History), London and CambridgeUniversity Press, Cambridge. Titles to be published in Volume 44 Observations on the systematics of the genus Difflugia inBritain (Rhizopoda, Protozoa).By Colin G. Ogden Miscellanea A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin Curds & Irene C. H. Wu The Opthalmotilapia assemblage of cichlid fishes reconsidered* By Peter Humphry Greenwood Osteology, genitalia and relationships of the Acanthodactylus(Reptilia: Lacertidae). By E. N. Arnold Morphological studies on some Difflugiidae from Yugoslavia(Rhizopoda, Protozoa). By Colin G. Ogden & Andjelija Zivkovic Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) Morphological studies on some Difflugiidae from Yugoslavia (Rhizopoda, Protozoa) Colin G. Ogden & Andjelija Zivkovic Zoology series Vol 44 No 6 30 June 1983 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in fourscientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, andan 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 SW7 5BD,England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Zool.) Trustees of the British Museum (Natural History), 1983 The Zoology Series is edited in the Museum's Department of ZoologyKeeper of Zoology : Dr J. G. ShealsEditor of Bulletin : Dr C. R. CurdsAssistant Editor : Mr C. G. Ogden ISSN 0007-1498 Zoology series Vol44No6pp341-375British Museum (Natural History)Cromwell RoadLondon SW7 5BD Issued 30 June 1983 Morphological studies on some Difflugiidae fromYugoslavia (Rhizopoda, Protozoa) Colin G. Ogden Department of Zoology, British Museum (Natural History), Cromwell Road, LondonSW7 5BD Andjelija Zivkovic Institute for Biological Research, University of Belgrade, 29, Novembra 142, 11000Belgrade, Yugoslavia Contents Introduction 341 Materials and methods 341 Systematic descriptions 342 Difflugia 342 Cucurbitella 369 Pontigulasia 369 Summary 373 References 374 Introduction Recent studies (Ogden, 1979, 19800, b, 1983; Ogden & Fairman, 1979) on the shellmorphology of specimens belonging to the family Difflugiidae, have shown that detailedexamination of these structures aid identification at the species level. It is now possible usingthe scanning electron microscope to demonstrate differences in shell structure and toexamine more easily the type and arrangement of materials used in these complex con-structions. Earlier workers were handicapped in similar examinations by the limitedresolution of reflective optical microscopy. This made comparisons of shell structuredifficult due to the different densities of the shell ranging from opaque to transparent, andsometimes the shape made a complete survey impracticable, ovoid or circular shells beingparticularly awkward. The present work is based on specimens collected in Serbia,Yugoslavia, from an area of peat bogs located on the high plateau at 1200 m, now submergedby the artificial lake 'Vlasina', which was formed by damming the river Vlasina and floodinga depression. A limnological study of this lake by Milovanovic & Zivkovic (1956) givesinformation relating to the chemical and biological conditions during the initial formation ofthis feature. There are several reasons for our interest in the Difflugiidae of Yugoslavia, no previousrecords of testate amoebae are available for this region, there appears to be a similarity of thisfauna with that reported from Africa by Gauthier-Lievre & Thomas (1958) and specimens ofthe so-called 'cosmopolitan' species are available for comparison with those alreadydescribed from the British Isles. Furthermore, the presence of several compressed species ofDifflugia is unusual, as these forms appear to be rare. Materials and methods Samples were collected in September, 1947 from a pond 'Godzina bistrica' located at the Bull. Br. A/MS. nat. Hist. (Zool.) 44(6) : 34 1-375 Issued 30 June 1 983 342 C. G. OGDEN & A. 2lVKOVIC margin of a peat bog, and preserved in formalin. The samples were divided into two in 1980,one half was deposited in the Zoology Department, British Museum (Natural History) andthe other was retained in Yugoslavia. This report is based mainly on the BM(NH) specimensbut reference is also made to those in the other sample. Selected shells from the sample werewashed in several changes of distilled water, then individuals were prepared for scanningelectron microscopy using the technique described by Ogden (1979). The prepared stubswere examined using a Cambridge Stereoscan SI 80 operating at 10 kV and the resultsrecorded on Ilford HP5 film. Systematic descriptions The species ofDifflugia are listed in alphabetical order, except for the new species which aredescribed last and D. gramen which is described with D. schurmanni for comparativepurposes. The single species of the genus Cucurbitella and two species of Pontigulasia aredescribed after the species of Dijflugia. The measurements for total body length includesaboral processes, unless otherwise stated, breadth and diameter of aperture are taken as thewidest point, the latter is an internal measurement. Genus DIFFLUGIA Leclerc, 1815 Difftugia acuminata Ehrenberg, 1838 A single specimen, 199 urn long, 85 urn in breadth, with an aperture diameter of 37 urn, wasexamined and seen to be identical to those already described (Ogden, 1979). Significantly thedistinctive organic cement pattern of this species, a network in which each mesh is furtherdivided by a smaller network, was present. Difflugia acutissima Deflandre, 193 1 DESCRIPTION. The shell is transparent, pyriform with the sides tapering evenly from aboutthe mid-body region and terminating usually with a sharp point (Fig. la). It is composedmainly of flattish pieces of quartz to give a smooth surface, with organic cement sometimesseen as part of the shell matrix (Fig. Ic). The cement is usually in the form of a network about400 nm in diameter with walls 200 nm thick, and each mesh is covered by a smoothmembrane (Fig. Id). The aperture is circular and surrounded by small to medium particles(Fig. Ib). One specimen had an oblique aperture and a misaligned aboral protuberance (Fig.le), the latter being only partially sealed (Fig. 10, but was considered to agree with thegeneral description of this species in all other respects. MEASUREMENTS (in um). Four specimens: body length 178-217, breadth, 80-107, diameterof aperture 47-51. GEOGRAPHICAL DISTRIBUTION. Argentina (Dioni, 1970), Chad (Gauthier-Lievre & Thomas,1958), Congo (Chardez, 1964), Gold Coast, Morocco (Gauthier-Lievre & Thomas, 1958),United States of America (Leidy, 1879), Venezuela (Deflandre, 1931). REMARKS. This species was initially described by Deflandre (1931) who considered that itwas distinct from D. acuminata and its varieties by the sharpness of the aboral extremity. Itdiffers from D. ventricosa, also described by Deflandre (1926) from Venezuela, by having amuch wider body and aperture. The general body dimensions are similar to those given forD. distenda by Ogden (1983), which was a variety of D. acuminata but this species is curvedaborally to a small tubular horn. The problems of differentiating between species with aboralhorns or spines are emphasized in two publications by Chardez (1961 & 1973), which showseveral figures supposedly pertaining to the same species. It is clear that concise specific DIFFLUGIDAE FROM YUGOSLAVIA 343 Fig. 1 Difflugia acutissima : a, lateral view x600; b, apertural view x360; c, detail of organiccement pattern x 14000; d, portion of shell surface to show organic cement between particlesX 6600; e, specimen with oblique aperture and non-central aboral spine x 420; f, detail of partialclosure at extremity of aboral spine x 1900. 344 C. G. OGDEN & A. 2IVKOVIC identification of specimens must await morphogenetic evidence of natural variation of thisfeature within a species. Difflugia bicornis Penard, 1890 DESCRIPTION. The shell is transparent, ovoid or spherical with usually two, but occasionallyone aboral spines (Figs 2a & b). Of the three specimens examined only one had a single longaboral spine. The surface is rough, thin and composed of a mixture of small to mediumparticles of quartz, with the occasional addition of either a portion of diatom frustule or asiliceous flagellate cyst. Each aboral spine is roughly pointed and composed of smallparticles. A network of organic cement is sometimes seen between particles (Fig. 2d). It is inthe form of a sheet with each mesh being separated by small walls, about 80 nm thick. Themesh is irregular, about 300-400 nm in diameter, and each enclosure often has a small innercircular wall with arms connecting it to the mesh wall (Fig. 2e). The aperture is usuallycircular and surrounded by a mixture of particles (Fig. 2c). MEASUREMENTS (nm). Three specimens: body length 75-85, breadth 39-53, diameter ofaperture 19-23. GEOGRAPHICAL DISTRIBUTION. Germany (Penard, 1890; Jung, 1936), Guatemala, Mexico(Laminger, 1973), Switzerland (Penard, 1902). REMARKS. This species was initially described as a distinct species by Penard (1890) but waslater considered (Penard, 1902) to be a small form of D. elegans, and he illustrated the vari-ability of the shell in this form with several figures, including specimens with either one ortwo aboral spines. The specimens reported here have a fragile shell and are in good agree-ment with the original description of D. bicornis, the body lengths without the spines orhorns being about 60 um which is near to Penard's 50-60 urn. The contrast between thesespecimens and those of D. elegans studied recently (Ogden, 1979), which had robust shellsand varied in body length between 1 1 3-1 58 um, is marked. In the absence of information onother small forms identified as D. elegans, we have decided to use the earlier description ofD. bicornis and consider the present specimens as a distinct species. Difflugia bryophila (Penard, 1902) A single specimen, 124 um in body length, 53 um broad with an aperture diameter of 1 7 um.It was identical to those described by Ogden (1 983). Difflugia capreolata Penard, 1 902 DESCRIPTION. The shell is opaque, thick, pyriform with a restriction of the neck at a positionabout one-third of the total body length, before it swells into the main body (Fig. 3a). It iscomposed of small to medium pieces of angular quartz, with small areas of organic cement aspart of the matrix (Fig. 3c). This cement is in the form of a smooth sheet with irregularperforations (Fig. 3d), the reason that there is no apparent pattern to these perforations maybe due to these areas being in thin strips rather than a more open arrangement. Eachperforation has a mean diameter of 200 nm. The aperture is circular and surrounded by aregular distribution of medium particles (Fig. 3b). MEASUREMENTS (in um). One specimen: body length 225, breadth 128, diameter of aperture58. GEOGRAPHICAL DISTRIBUTION. Argentina (Dioni, 1970; Lena & Zaidenwerg, 1975),Germany (Schonborn, 1965), Russia (Kourov, 1925), Tunisia (Gauthier-Lievre & Thomas,1958), Sudan (Gauthier-Lievre & Thomas, 1958), Switzerland (Penard, 1902). REMARKS. This species although it is large and has a distinctive outline, does not appear to becommon and is rarely found in large numbers. DIFFLUGIIDAE FROM YUGOSLAVIA 345 Fig. 2 Difjlugia bicornis: a, lateral view of specimen with two aboral spines x 1400; b, lateralview of specimen with single aboral spine x 760; c, apertural view x 820; d, portion of shellsurface to illustrate the organic cement (arrowed) x7500; e, detail of organic cement x 35000. 346 C. G. OGDEN & A. 2IVKOVIC Fig. 3 Difflugia capreolata: a, lateral view x400; b, apertural view x340; c, shell surfaceshowing small areas of organic cement x 3000; d, detail of organic cement pattern x 1 3000. Difflugia corona Wallich, 1864 DESCRIPTION. The shell is brown, spherical or ovoid with distinct cone-like spines projectingfrom the aboral half of the body (Fig. 4a-d). The main body is composed of a mixture ofsmall to medium particles of quartz arranged to give a relatively smooth shell. The spines arerandomly arranged, varying between two and eight in number, made of small particles andare usually finely pointed. Organic cement in the form of a network is sometimes seen whereparticles meet. The aperture is circular and surrounded by a distinct denticular collar. There DIFFLUGIIDAE FROM YUGOSLAVIA 347 , , ^~'-^~r ^%'j^^& Fig. 4 Difflugia corona: a, lateral view of specimen with six spines x430; b, apertural view of a,note the regular arrangement of twelve teeth x 330; c, lateral view of another specimen x 240; d,apertural view of c, to show sixteen thickened teeth x 240. are usually between ten to twelve tooth-like projections, although one specimen in thepresent sample had sixteen (Fig. 4d). They are arranged evenly, usually being finelypointed, but as seen in Fig. 4d compression of the teeth makes the projections thicker and theouter curvature of the collar more rounded. MEASUREMENTS (in um). Fourteen specimens: body length 126-190, breadth 126-177,diameter of aperture 53-86. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina(Boltovskoy & Lena, 1974; Lena & Ziadenwerg, 1975; Vucetich, 1970), Austria (Laminger,1975), Belgium (Chardez, 1961, 1980), Brazil (Green, 1975), British Isles (Ogden & Hedley,1980; Wallich, 1864), Chad (Gauthier-Lievre & Thomas, 1958), Congo (Chardez, 1964;Gauthier-Lievre & Thomas, 1958), Czechoslovakia (Ertl, 1965), Germany (Penard, 1890),Hungary (Bereczky, 1973), India (Wallich, 1864), Java (BartoS, 1963), Netherlands(Hoogenraad & Groot, 1940), Roumania (Godeanu et al, 1973), Russia (Kourov, 1925), 348 C. G. OGDEN & A. 2IVKOVIC Fig. 5 Difflugia difficilis: a, lateral view x 950; apertural view x 760; c, detail of aperture to showshort collar and organic cement (arrowed) between particles x 2 100; d, detail of organic cementx 18000. Senegal, Sudan and West Africa (Gauthier-Lievre & Thomas, 1958), South Africa (Oye,1931), Switzerland (Penard, 1902), United States of America (Leidy, 1879). REMARKS. This species is truly 'cosmopolitan' having been reported from most continents,but these sightings are probably due to the ease in observing the prominent features of thislarge, distinctive shell in any sample. Although Jennings (1916, 1937) showed that variationof both teeth and spines occurred under cultural conditions, subsequent authors havecontinued to designate forms and varieties based on these structures. It could be argued thatJenning's observations were made on rough cultures and that clonal cultures would behavedifferently, or that the medium used was limiting in some way. Nevertheless, some of thereported features could easily be considered to represent natural variation. DIFFLUGIIDAE FROM YUGOSLAVIA 349 Difflugia decloitrei Godeanu, 1 972 A single specimen similar to those recently described by Ogden (1983) was found. It ispossibly an encysted form as the aperture was blocked with several flat pieces of quartzbound by organic cement. MEASUREMENTS (in um). One specimen: body length 79, breadth 52, diameter of aperture 22. Difflugia difficilis Thomas, 1954 DESCRIPTION. The shell is transparent, ovoid, thin with a small apertural collar (Fig. 5a). It iscomposed of small to medium pieces of angular quartz arranged to make a relatively smoothsurface, with the occasional projection of some particles. A network of organic cement isfrequently seen as part of the shell matrix (Fig. 5c). The mesh of this network is open, eachopening being about 350 nm wide with dividing walls 200 nm thick (Fig. 5d). A short collarmade mainly of regularly arranged smallish particles, surrounds the circular aperture (Figs5b & c). MEASUREMENT (in um). One specimen: body length 8 1 , breadth 56, diameter of aperture 18. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina (Lena& Zaidenwerg, 1975), Congo (Chardez, 1964), France (Thomas, 1954), Ivory Coast(Gauthier-Lievre & Thomas, 1958), Roumania (Godeanu et al, 1975), Sudan (Gauthier-Lievre & Thomas, 1958). REMARKS. This specimen has similar dimensions to those given for D. difficilis and D.kabylica by Gauthier-Lievre & Thomas (1958). It differs from D. kabylica in having asmaller aperture which does not have a flattened irregular outline. It is most similar to D.difficilis, although it does not have 'une excoissance peu proeminente' as initially describedby Thomas (1954). However this species was later reported (Gauthier-Lievre & Thomas,1958) to vary a great deal in body outline and often the aboral protuberance is absent. Thefeature that is characteristic of this species is the small collar and in the absence of furtherspecimens the present specimen is considered to be D. difficilis. Difflugia gramen Seep. 357. Difflugia labiosa Wailes, 1919 Two specimens of this species identical to those recently redescribed by Ogden (1983) werefound. MEASUREMENTS (in um). Two specimens: body length 176 & 177, breadth 111, diameter ofaperture 45 & 52. Difflugia lata Jung, 1942 Difflugia oblonga forma lata Jung, 1942 DESCRIPTION. The shell is opaque, pyriform having a rough surface composed of a mixture ofmainly small to medium pieces of quartz, with an occasional large particle added (Fig. 6a).One specimen has a curved or malformed aboral extremity (Fig. 6c), but in all other respectsis similar to the original description. Organic cement in the form of a network is seen as partof the shell matrix (Fig. 6d). The mesh of the network is open and has a mean diameter of350 nm with walls 200 nm thick, although the walls often fuse to give larger areas of cement(Fig. 6e). The aperture is circular and surrounded by both small and medium particlesarranged to give a somewhat irregular outline (Fig. 6b). 350 C. G. OGDEN & A. ZlVKOVIC Fig. 6 Dijjlugia lata: a, lateral view x 590; b, apertural view x 420; c, specimen with malformedaboral extremity x 400; d, shell surface with organic cement as part of matrix x 3600; e, detail oforganic cement x 13500. MEASUREMENTS (in urn). Two specimens: body length 137 & 149, breadth 90 & 103,diameter of aperture 42 & 46. GEOGRAPHICAL DISTRIBUTION. Chile (Jung, 1942). REMARKS. This species is known apparently only from the initial description (Jung, 1942) asa new form of D. oblonga. Although measurements are not given in the original text, from DIFFLUGIIDAE FROM YUGOSLAVIA 351 Fig. 7 ^^^^^^^^^^^^^^^^^^l ''I Difflugia levanderi: a, lateral view x730; b, apertural view x530; c, portion of shellsurface with organic cement (arrowed) x 7000. the figure these are estimated to be body length 157 um, breadth 81 urn and diameter ofaperture 43 urn, which agree well with the present specimens. Note that our Fig. 6ccompares well with that given by Jung (1942). This species is considered to be distinct fromD. oblonga (see Ogden, 1979) in its stout pyriform body, wide aperture and organic cementpattern. Difflugia levanderi Playfair, 1918 DESCRIPTION. The shell is ovoid or almost spherical, composed of flattish and angular piecesof quartz to give a regular outline (Fig. 7a). The particles appear to overlap and produce arobust structure with a network of organic cement, which is seen infrequently, binding theparticles (Fig. 7c). Details of the surface are restricted because all the examined specimenshave a slight covering of small debris, this latter material is clearly no part of the shellstructure. The aperture is circular, well defined and surrounded by small particles (Fig. 7b). MEASUREMENTS (in um). Five specimens: body length 95-104, breadth 76-92, diameter ofaperture 32-40. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Australia(Playfair, 1918), Chile (Jung, 1942), Morocco and Tunisia (Gauthier-Lievre & Thomas,1958). REMARKS. The descriptions of both Levander (1894) and Playfair (1918) refer to specimenshaving a similar shape but differing in size and composition. The larger being robust with acoarse appearance, whilst the smaller had a chitinous shell with a scattering of particles.Unable to differentiate them Playfair (1918) described them as a new species D. levanderi. C. G. OGDEN & A. 2IVKOVIC Fig. 8 Difflugia lismorensis: a, latero-apertural view x 640; b, apertural view to illustrate thethickened teeth and collar x470; c, part of shell surface with overlay of extraneous materialX 3400. Fortunately, he created a precedence by initially describing the larger specimens, which arenow considered to represent D. levanderi. Specimens similar to D. levanderi except for beingsmaller and having a thin shell whose particles did not overlap, were described by Godeanu(1972) as a new species D. decloitrei, and these are considered to be identical to the smalleranimals described by Levander (1894), Playfair (19 18) and redescribed by Ogden (1983). Difflugia lismorensis Playfair, 1918 Difflugia lismorensis var. quinquelobata Gauthier-Lievre & Thomas, 1958Difaugia lismorensis var. elongata Gauthier-Lievre & Thomas, 1 958 DIFFLUGIIDAE FROM YUGOSLAVIA 353 DESCRIPTION. The shell is either elongate or ovoid, thick, composed of small to mediumpieces of quartz arranged to give a rough surface (Fig. 8a). All the specimens examined had aproportion of the surface covered with an overlay of small extraneous material (Fig. 8c),nevertheless, the portions of actual surface visible had only strands of organic cement as partof the shell matrix. The aperture has five lobes which are equally spaced, strong, blunt,tooth-like extensions (Fig. 8b). Often the strengthening around these lobes, usually withsmall particles, gives it a collar-like appearance (Fig. 8b). MEASUREMENTS (in um). Three specimens: body length 132-144, breadth 96-104, diameterof aperture 3 6-4 5. GEOGRAPHICAL DISTRIBUTION. Argentina (Vucetich, 1970), Australia (Playfair, 1918), Brazil(Green, 1973), Chad (Gauthier-Lievre & Thomas, 1958), Congo (Chardez, 1964; Gauthier-Lievre & Thomas, 1958), Gold Coast, Morocco and Sudan (Gauthier-Lievre & Thomas,1958). REMARKS. Playfair (1918) described D. lismorensis as having an ovoid or sub-globular shellwith a six-lobed aperture, and added two varieties trilobulata and crucifera which had threeand four lobes respectively. Since that time two more varieties with five lobes, from Africa,have been described by Gauthier-Lievre & Thomas (1958). They suggested the namequinquelobata for the ovoid variety and elongata for the specimens with an elongate body,the former has subsequently been reported again in Africa by Chardez (1964). Vucetich(1970) after examining about 50 specimens with seven lobes from Argentina, concluded thatthese were otherwise identical with D. lismorensis and she did not consider that differencesin the number of lobes warranted specific designation. Until examples of this species fromAustralia, Africa and South America are compared, we agree with Vucetich (1970) that herspecimens, plus those of Gauthier-Lievre & Thomas are best described as D. lismorensis. Difflugia lithophila (Penard, 1902), Gauthier-Lievre & Thomas, 1958 Difflugia hydrostatica var. lithophila Penard, 1902 A single specimen identical to that described by Ogden & Hedley (1980) was examined. It isslightly smaller than previous records being 93 um in body length, 67 urn breadth and havingan apertural diameter of 32 um; however, it is proportionally similar. African specimensappear to have a larger range of body length, 100-170 um (Gauthier-Lievre & Thomas,1958), to those from Europe 99-1 40 um (Penard, 1902; Thomas, 1954). Difflugia lucida Penard, 1 890 A single specimen was examined, it measured 61 um in body length, 39 um broad, 19 umdepth with an aperture diameter of 22 um. Although slightly smaller than those examined inan earlier study (Ogden, 1983) it was otherwise identical. Difflugia manicata Penard, 1 902 Two specimens 76 & 78 um long, 39 & 43 um in breadth, with aperture diameters of 14 &17, were examined and found to be identical to those recently redescribed by Ogden (1983). Difflugia mica Frenzel, 1892 DESCRIPTION. The shell is brownish, spherical or ovoid sometimes with a shallow aperturalcollar (Fig. 9a). It is composed of flattish pieces of quartz arranged to give a relatively smoothsurface, although one specimen has most of the surface obscured by extraneous material.Organic cement is seen infrequently between particles but is more evident around theaperture. The cement around the aperture is apparent as a thin layer on the surface of someouter particles (Fig. 9c), but is a concentration of strands, some apparently fused, on the 354 C. G. OGDEN & A. 2IVKOVIC Fig. 9 Difjlugia mica: a, lateral view showing the arrangement of flattish particles x 1400; b,apertural view x 1 100; c, organic cement at outer limit of apertural concentration x 17000; d,concentration of organic cement on apertural lip, note the fusion of some strands x 20000; e,general appearance of organic cement network x 29000. DIFFLUGIIDAE FROM YUGOSLAVIA 355 f Fig. 10 Difflugia nodosa: a, apertural view x270; b, view to illustrate lateral compressionx 160; c, lateral view showing the lateral wings and aboral protuberance x970; d, portion ofshell surface close to aperture, note that the organic cement is torn (arrowed) x4600; e, & f,detail of organic cement pattern x 14000 & x 2 1000. apertural lip (Fig. 9d). The general structure is of a network with an irregular mesh about 300to 500 nm in diameter, with dividing walls about lOOnm thick (Fig. 9e). The aperture iscircular, well denned, usually with an organic margin and sometimes with a collar (Fig. 9b). MEASUREMENTS (in um). Two specimens: body length 55 & 60, breadth 46 & 55, diameter ofaperture 19 & 18. 356 C. G. OGDEN & A. 2lVKOVIC REMARKS. These specimens are in good agreement with the description given by Ogden(1983). They are described again here to include information of the organic cement pattern. Difflugia microclaviformis (Kourov, 1925) The two specimens examined here fall within the range of body lengths given by Kourov(1925) of 139-1 60 um, their respective measurements are: body length 141 & 163 urn,breadth 66 & 75 um and diameter of aperture 19 & 21 um. Furthermore, they are in goodagreement with the recent redescription of slightly larger specimens given by Ogden (1983). Difflugia nodosa (Leidy, 1879) comb. nov.Difflugia pyrif or mis var. nodosa Leidy, 1879 and in Penard, 1902Dijflugia oblonga var. nodosa Leidy, 1 879 in Gauthier-Lievre & Thomas, 1958 DESCRIPTION. The shell is thick, opaque, pyriform with the main body compressed to givetwo large lateral wings and an aboral extension or protuberance (Figs lOb & c). It iscomposed of a mixture of various sized pieces of quartz, which unfortunately in the presentspecimen is overlaid with debris. Nevertheless, some organic cement was observed close tothe aperture (Fig. lOd); it is in the form of a network with the open mesh having a meandiameter of 200 nm and walls 200 nm thick (Figs lOe & f)- The aperture is circular andsurrounded by a regular arrangement of small particles (Fig. 1 Oa). MEASUREMENTS (in um). One specimen: body length 367, breadth 241, depth 140, diameterof aperture 53. GEOGRAPHICAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Belgium(Chardez, 1980; Chardez & Gaspar, 1976), Congo (Gauthier-Lievre & Thomas, 1958),Germany (Laminger, 1973; Schonborn, 1962), Ivory Coast (Gauthier-Lievre & Thomas,1958), Netherlands (Hoogenraad & Groot, 1940), Roumania (Godeanu et al, 1973), Russia(Kourov, 1925), Sudan (Gauthier-Lievre & Thomas, 1958), United States of America (Leidy,1879). REMARKS. This species has consistently been described as a variety of/), oblonga, initiallyLeidy (1879) remarked that it was a 'striking variety' distinguished by its unusually large size,compressed body and three conical eminences which varied a great deal in their develop-ment. Although Leidy (1879) reported it as being abundant on one particular occasion, sothat one drop of ooze contained several dozen individuals, since that time no other reportshave been so fruitful with specimens. Nevertheless, we consider that this variety is besttreated as a distinct species which differs from the typical D. oblonga in the featuresdescribed above and the organic cement pattern. Difflugia oblonga Ehrenberg, 1838 Two specimens with the typically rough shell recently redescribed by Ogden & Fairman(1979) were examined; they measured: body length 232 & 263 um, breadth 1 12 & 109 umand diameter of aperture 46 & 41 um. Difflugia parva (Thomas, 1954) The specimens of this species agree well with the redescription given by Ogden (1983) havinga relatively smooth shell and areas of organic cement as part of the shell matrix; body length143-203 urn, breadth 78-94 urn and diameter of aperture 24-32 um. Difflugia pristis Penard, 1902 Four specimens, body length 37-62 um, breadth 27-38 um and diameter of aperture13-16 um, having shells composed mainly of quartz particles but with some added diatomfrustules, were examined. DIFFLUGIIDAE FROM YUGOSLAVIA 357 Difflugia pulex Penard, 1902 Two small ovoid specimens, body length 30 & 32 um, breadth 22 & 24 um with diameter ofaperture 10 & 1 1 um were examined. Difflugia rubescens Penard, 1 89 1 Two specimens, body length 5 & 62 um, breadth 34 & 40 um and diameter of aperture1 5 um were examined. Difflugia gramen Penard, 1902 DESCRIPTION. The shell is transparent or light brown, spherical tapering towards theaperture (Fig. 1 la). It is composed of a mixture of small to medium pieces of quartz, boundby an organic cement network (Fig. lie). A ring of small pores surrounds the aperture (seeOgden, 1 980b) which is trilobed and bordered by an irregular raised rim (Fig. 1 1 b). MEASUREMENTS (in um). See Table 1 . REMARKS. A brief description of this species is included so that a direct comparison can bemade with specimens of Difflugia schurmanni Oye, 1 932 described below. Table 1 Range of measurements (in um) of four ovoid species of Difflugia. n = number of specimens; E = Norfolk, England (see Ogden, 1 9806); Y = Yugoslavia (present work) Difflugia schurmanni Oye, 1 932 DESCRIPTION. The shell is transparent, elongate ovoid, tapering equally to the aperture andaboral extremity (Fig. 1 Id). It is composed mainly of flattish pieces of quartz arranged to givea smooth and rather fragile appearance. The particles are bound by organic cement, similarto that described for D. gramen, but because the particles fit closely together cement is notseen as frequently as in that species. The aperture is trilobed with the dividing projectionsless pronounced in apertural view (Fig. lie) compared with those of D. gramen, and thisgives it a more open outline especially as the shell has a smaller breadth (see Table 1). Inaddition, the surrounding ridge is often not as well defined although in lateral view the lobesare more prominent (Fig. 1 Id). A ring of small pores surrounds the aperture, similar to thosedescribed for D. gramen (see Ogden, 19806). MEASUREMENTS (in um). See Table 1. GEOGRAPHICAL DISTRIBUTION. Brazil (Green, 1975), Congo (Gauthier-Lievre & Thomas,1958), Guatemala (Laminger, 1973a), Senegal (Gauthier-Lievre & Thomas, 1958), SouthAfrica (Oye, 1932), Sudan (Gauthier-Lievre & Thomas, 1958). 358 C. G. OGDEN & A. ilVKOVIC K- r : f ;<;"*'.ry-^'^t *'-^/; ;r 4 'H Fig. 11 Difflugia gramen: a, lateral view x 1000; b, apertural view, note the small apertureopening but well denned surrounding ridge x610; c, detail of organic cement pattern x7500Difflugia schurmanni; d, lateral view, note the different arrangement of particles and dennedapertural lobes x 1500; e, apertural view to illustrate wide opening and less pronounced ridgexlOOO. DIFFLUGIIDAE FROM YUGOSLAVIA 359 REMARKS. Reports of this species appear to be confined to tropical countries, but this may bedue to it being reported elsewhere as D. gramen. The measurements of specimens for thethree similar species, D. achlora, D. gramen and D. lobostoma, redescribed recently (Ogden,1980&) from England are listed in Table 1 for comparison. It would appear that theinformation given does not help to resolve specific identification, which still depends on themeasurement of overall body length and differences in shell structure. The significance ofthese latter features must await morphological studies on clonal cultures. D. schurmanni is considered to be distinct from D. gramen in being smaller, moreelongate, having a smooth surface composed of flattish particles and a more open aperture. Difflugia smilion Gauthier-Lievre & Thomas, 1958 DESCRIPTION. Unfortunately the only specimen was broken during preparation. The shell istransparent, elongate with a distinct aboral conical protuberance. It is composed mainly ofmedium to large, angular or flat pieces of quartz to give a rough surface. Organic cement isseen at the junctions of these particles as enclosed spheres, or spheres with small openings, oras a thick rimmed circle with four equal openings (Fig. 12b & c). The spheres are about600 nm in diameter, the walls of the circle are about 1 50 nm thick and the internal openingsvary between 1 50-220 nm. The aperture is circular and surrounded mainly by medium sizedparticles (Fig. 12a). Fig. 12 Difflugia smilion: a, apertural view to show arrangement of surrounding particles; b & c,detail of organic cement pattern x 2 1 000 & x 1 3000. Table 2 Range of measurements (in um) of D. tuberculata and D. wailesi n = number of specimens; (i) = typical specimens with indented aperture; (ii) = specimens with circular aperture;E = Norfolk, England (see Ogden, 1 9806); Y = Yugoslavia (present work) 360 C. G. OGDEN & A. ZlVKOVIC GEOGRAPHCAL DISTRIBUTION. Algeria (Gauthier-Lievre & Thomas, 1958), Argentina (Dioni,1970), Belgium (Chardez, 1980; Chardez & Caspar, 1976), Brazil (Green, 1975), Congo(Chardez, 1964), France (Thomas, 1953, 1954), Ivory Coast (Gauthier-Lievre & Thomas,1958), Poland (Moraczewski, 1965), Roumania(Godeanu^a/., 1973). REMARKS. This specimen is similar in all respects to the description given by Thomas (1953),the body length being about the same, 226 urn, although here it is an estimated figure becauseof the state of the specimen. No value is given for the breadth, but it was seen to be onlyslightly larger than the aperture which has a diameter of 4 1 u,m. Dijflugia tuberculata (Wallich, 1864) Specimens with the typical protuberances were examined and the measurements areincluded in Table 2. Dijflugia wailesi Ogden, 1980Difflugia tuberculata var. minor Wailes, 1919 DESCRIPTION. Two distinct forms of this species are present in the examined material, bothare figured and described. In the typical form the shell is transparent, ovoid but tapering slightly near the aperture(Fig. 13a). It is composed mainly of flattish pieces of quartz and siliceous elements arrangedto give a smooth surface (Fig. 1 3a). The aperture is polygonal with usually five but occasion-ally six indentations and bordered by a small lip (Fig. 13b). A ring of about ten small pores isoften seen just posterior to the apertural rim. In the other form the shell is ovoid but more markedly tapered towards the aperture (Fig.13c). The aperture is circular with no indentations (Fig. 13d), but with a similar small ridgeas in the typical form. There are nine or more small tooth-like projections inside theapertural rim on a level with the main body of the shell (Fig. 13e), these tooth-like structuresare similar to those sometimes seen at the apex of each apertural projection in the polygonalaperture specimens. MEASUREMENTS (in um). See Table 2. REMARKS. The similarity of dimensions between the present specimens and those reportedrecently from Norfolk, England, highlight the difficulty of differentiating the two species D.tuberculata and D. wailesi. A further division of D. wailesi is not desirable unless supportedby strong morphological evidence, therefore, although the specimens are reported as twogroups no particular significance is attached to these differences. Difflugia balcanica sp. nov. DESCRIPTION. The shell is transparent, ovoid or roughly circular with an aboral cone or spineand laterally compressed, the shape resembles either an arrow-head or a leaf (Figs 14a & c).It is composed of a mixture of small to large pieces of mainly flattish quartz, arranged to givea relatively smooth but irregular surface, unfortunately the illustrated specimen has someoverlying debris scattered on the anterior third of the body. An open network of organiccement is seen between the particles. The aperture is roughly circular and surrounded bysmall or medium pieces of quartz (Fig. 14b). MEASUREMENTS (in um). Two specimens: body length 1 1 1-1 14, breadth 79-82, depth 50,diameter of aperture 3 1 . REMARKS. There does not appear to be any prior descriptions of any similar compressedspecies of Difflugia, and D. balcanica is considered to be distinct in having a leaf-like shape,with an angular surface and a pointed aboral extremity. DIFFLUGIIDAE FROM YUGOSLAVIA 361 Fig. 13 Difflugia wailesi: a, lateral view of typical specimen x 760; b, apertural view of a, notethe six indentations x470; c, lateral view of elongate specimen x 1 100; d, apertural view of; c, toshow circular aperture and surrounding collar x 720; e, detail of teeth on inner rim of circularaperture x5100. 362 C. G. OGDEN & A. 2lVKOVIC Fig. 14 Difflugia balcanica sp. nov.: a, view to illustrate general outline with aboral pro-tuberance x570; b, apertural view x620; c, lateral view to show compression, note thetapering at the apertural and aboral extremities x 460. Difflugia bistrica sp. nov. DESCRIPTION. The shell is ovoid or roughly circular, thin and laterally compressed (Fig. 15a& b). It is composed of medium flattened pieces of quartz with smaller particles between togive a smooth surface (Fig. 15d). Small areas of organic cement occur infrequently, and areusually in the form of a network (Fig. 1 5e). The aperture is a regular oval and surrounded bysmall particles (Fig. 1 5c). MEASUREMENTS (in um). One specimen: body length 104, breadth 84, depth 54, diameter ofaperture 38. REMARKS. Although this specimen has an encrustation of small debris, which is not part ofthe shell matrix (see Fig. 15d), it is still possible to describe the surface as smooth. Thisspecies is similar to D. balcanica in dimensions and degree of compression, but differs signifi-cantly in having a rounded, smooth shell which is curved gently at the aboral extremity.Notwithstanding that this description is based on a single specimen, it is still considered torepresent a distinct species and is so designated. Difflugia dragana sp. nov. DESCRIPTION. The shell is opaque, elongate ovoid, thick, tapering gradually from themid-body position to a gracefully curved aboral extremity and anteriorly to the suggestion ofa small apertural collar, it is laterally compressed (Figs 16a & b). Medium pieces of quartzappear to make up most of the surface with smaller particles added. Only small strands oforganic cement have been seen (Fig. 16d) between these particles. The aperture is circularand surrounded by mainly small particles of quartz (Fig. 16c). MEASUREMENTS (in um). One specimen: body length 195, breadth 1 19, depth 96, diameter ofaperture 48. DIFFLUGIDAE FROM YUGOSLAVIA 363 Fig. 15 Difflugia bistrica sp. nov.: a, lateral view to show regular, circular shape x 760; b, view toillustrate lateral compression, note the gentle aboral curvature x 720; c, apertural view showingoval-shaped aperture x 660; d, portion of shell surface to show flat particles overlaid withextraneous debris x 2900; e, organic cement (arrowed) partially obscured by debris x 13000. 364 C. G. OGDEN & A. 2lVKOVIC Fig. 16 Difflugia dragana sp. nov.: a, lateral view to illustrate the general outline x490; b, viewof lateral compression to show gradual curving at both extremities x 320; c, apertural viewx 400; d, part of shell surface with strands of organic cement (arrowed) x 9400. REMARKS. This compressed specimen is similar to D. lingula Penard, 1911, D. avellanaPenard, 1890 and D. hiraethogii Ogden, 1983. It differs from D. lingula and D. avellanawhich have pyriform shells, because of the narrow, elongate body which is almost cigar-shaped in lateral view (Fig. 16b), and wide aperture. Although it is most similar to D.hiraethogii, it lacks the distinct neck of this species and is again more elongate and not evenlycompressed. D. dragana is considered to be distinct in having an elongate, compressed shellwith a wide aperture. DIFFLUGIDAE FROM YUGOSLAVIA 365 Fig. 17 Difflugia serbica sp. nov.: a, lateral view of specimen encrusted with extraneous debrisx680; b & c, additional specimens to illustrate the variations of tapering x310 & 240; d,apertural view x420; e, view to show lateral compression x400; f, detail of organic cementnetwork x 21000. 366 C. G. OGDEN & A. 2IVKOVIC ETYMOLOLGY. This species is named after Dragana Dorothea Zivkovic, who has been mosthelpful in correcting the language shortcomings of both authors. Difflugia serbica sp. nov. DESCRIPTION. The shell is opaque, shaped like a spear-head, tapering evenly from the bodyto the pointed aboral extremity and gradually, anteriorly towards the short apertural collar,it is laterally compressed (Figs 1 7a-e). The body is composed mainly of medium to largeflattish pieces of quartz with a mixture of small to medium pieces at the aperture and aboralextremity. Although the surface is usually smooth, two of the specimens are covered with alayer of small particular debris which appears to have been added after the shell was con-structed (Fig. 1 7e). The organic cement is not frequently seen but is in the form of a network,the mesh being about 400 nm in diameter with walls 50 nm thick, each mesh being covered(Fig. 1 70- The aperture is circular and usually surrounded by small particles (Fig. 1 7d). Fig. 18 Difflugia serbica sp. nov. Diagrams of four specimens to show the variation in outline. MEASUREMENTS (in um). Five specimens: body length 169-189, breadth 110-113, depth68-79, diameter of aperture 31-41. REMARKS. The dimensions of five specimens from the Yugoslavian sample were somewhatlarger: body length 180-280, breadth 1 10-160, depth 75-1 15 (only two measured), diameterof aperture 40-70. Nevertheless, they demonstrate a similar variability in general shape andstructure (Fig. 18). This species is similar to D. soudanensis Gauthier-Lievre & Thomas(1958) and D. kempyi Stepanek, 1953 in general dimensions, but both of these species have adistinct apertural collar, are curved aborally and have a median pronounced spine or horn.Furthermore, of these two species only D. soudanensis is compressed. Difflugia serbica isconsidered to be distinct in having an irregular spear-shaped shell which is laterallycompressed, with the suggestion of a small apertural collar and usually pointed aborally. Difflugia serrata sp. nov.DESCRIPTION. The shell is transparent, ovoid, composed of small to medium flattish particles DIFFLUGIDAE FROM YUGOSLAVIA 367 a Fig. 19 Dijjlugia serrata sp. nov.: a, lateral view, note the serrated apertural margin x 1300; b,apertural view x 930; c, portion of shell surface with organic cement (arrowed) x 9400. of quartz arranged to give a smooth surface (Fig. 1 9a). Organic cement is seen as smallstrands or as a smooth sheet, with regular perforations about 50 nm in diameter (Fig. 1 9c).The aperture is circular with a slight irregular or serrated margin (Figs 19a & b). MEASUREMENTS (in um). One specimen: body length 66, breadth 56, diameter of aperture 28.REMARKS. This species is distinct from other ovoid species of similar dimensions, like D. 368 C. G. OGDEN & A. 2lVKOVIC Fig. 20 Dijflugia styla sp. nov.: a, lateral view to illustrate the aboral spine and smooth surface X 700; b, apertural view x 960. Fig. 21 Difflugia elegans. Illustrations of four specimens, the typical constriction of the neck can be seen in a. minuta Rampi, 1950 and D. pristis Penard, 1902 which have recently been redescribed byOgden (1983), in having a thin smooth shell and a wide aperture. It is most similar to D.pristis, but the more ovoid or rounded shell, with a large and serrated aperture serve todistinguish D. serrata. Difflugia styla sp. nov. DESCRIPTION. The shell is transparent, ovoid with a distinct aboral spine (Fig. 20a). It iscomposed of small to medium flattish pieces of quartz, and a network of organic cement isoften seen around the aperture as part of the shell matrix. Elsewhere on the surface theorganic cement is seen as strands between particles. The aperture is circular and surroundedby smallish particles (Fig. 20b). DIFFLUGIDAE FROM YUGOSLAVIA 369 MEASUREMENTS (in um). One specimen: body length 97, breadth 54, diameter of aperture 29. REMARKS. This species is similar to D. elegans or D. bicornis (see p. 000) in having a singleaboral spine or horn. However, both of these latter species are described as having a rough,irregular surface made of angular quartz particles and diatom frustules. In addition, D.elegans has a slight constriction of the neck which gives the aperture a flared appearance(Ogden, 1979), and it was found in the Yugoslavian portion of the sample to show thesefeatures (Fig. 21). Although D. styla shares similar dimensions to D. serrata (see p. 000) thepresence of an aboral spine separates these species. D. styla is distinct in being elongate ovoidwith an aboral spine, and having a smooth shell composed of flat particles. Genus CUCURBITELLA Penard, 1902 Cucurbitella vlasinensis sp. nov. DESCRIPTION. The shell is brown or opaque, subcircular or ovoid, with a distinct aperturalcollar (Fig. 22a). It is composed of a mixture of small to medium pieces of quartz to give arough surface, but arranged so that the outline is more or less regular. The particles arepacked close together (Fig. 220, with only small areas of organic cement visible. The cementis in the form of a network whose mesh is covered by a smooth membrane (Fig. 22e). Thecollar is trilobed and composed of small pieces of quartz arranged randomly (Fig. 22b). Adouble thickness of particles strengthens the three tooth-like projections where they form adividing barrier with the inner apertural opening; these 'teeth' are usually well denned andcomposed of small particles. Each lobe has a small recess or cavity so that the internalopening is smaller than the external collar, the floor of these recesses appears as a con-tinuation of the shell matrix (Fig. 22d). The apertural opening is trilobed in sequence withthe collar and lined with flattish pieces of quartz with smaller pieces filling the junctions (Fig.22c). MEASUREMENTS (in um). Forty-one specimens: body length 81-1 13, breadth 69-97, diameterof collar 36-5 1 , depth of collar 9-16, diameter of aperture 1 9-35. REMARKS. In a review of the genus Cucurbitella by Gauthier-Lievre & Thomas (1960) thenumber of lobes surrounding the aperture was used to differentiate species into a number ofvarieties and forms. More recently, Ogden (19806) considered that the three and four lobedspecimens of C. mespiliformis were otherwise identical in all but that feature, and shouldtherefore be designated as a single species until adequate morphological information wasavailable on the variability of lobe formation. Nevertheless, observations on the many speci-mens of the present sample show that there is little or no variation in the number and shapeof the lobes. C. vlasinensis is similar to C. modesta Gauthier-Lievre & Thomas, 1960 and C.modesta forma trilobata Gauthier-Lievre & Thomas, 1960. It differs from C. modesta, whichhas four lobes, in that feature and general measurements, and from C. modesta formatrilobata in having well-defined teeth, a more extensive inner lining to each lobe and in over-all larger general dimensions. An interesting ecological point to note is that C. modesta forma trilobata was reportedonly from Morocco which has similar climatic conditions to the region of Yugoslaviawhere the present samples were collected. ETYMOLOGY. This species is named after the location where the samples were collected, LakeVlasina, and which, after a visit by both authors in 1 982, has become especially significant. Genus PONTIGULAS1A Rhumbler, 1895 Pontigulasia bryophila Penard, 1902 DESCRIPTION. The shell is pyriform, tapering from the aperture to the slightly swollenposterior third of the body (Fig. 23a). It has a rough surface composed of a mixture of quartzparticles with an occasional diatom frustule added. Organic cement is seen at the junctions C. G. OGDEN & A. 2IVKOVIC Fig. 22 Cucurbitella vlasinensis sp. nov.: a, lateral view x 830; b, apertural view x 540; c, view toillustrate the arrangement of particles around the collar x 1200; d, detail of aperture to showtooth-like projections and smaller internal opening x 1200; e, detail of organic cement(arrowed) x 1 3000; f, portion of shell surface showing the close packing of particles x 1 500. DIFFLUGIDAE FROM YUGOSLAVIA 371 Fig. 23 Pontigulasia bryophila: a, lateral view x 470; b, view inside external aperture to show theinternal aperture opening (arrowed), micrograph obtained by 'expanded contrast' facilityx4100; c, apertural view x760; d, detail of organic cement to illustrate the smaller irregularnetwork of each mesh enclosure x 29000. between particles as a network. This network has a mesh about 600 nm in diameterseparated by walls 200 nm thick. Each enclosure has a smaller irregular network over itssurface (Fig. 23d). The external aperture is small, surrounded by small or medium particlesoften arranged to give a serrated or tooth-like margin (Fig. 23c). The single, internal apertureis roughly circular (Fig. 23b) and positioned about a quarter of the body length from theexternal aperture. MEASUREMENTS (in urn). See Table 2. REMARKS. This species is similar in general appearance to Difflugia bryophila but may bedistinguished from it by the presence of an internal aperture and a different organic cementpattern. The present specimens are similar in most respects to specimens of P. bryophilafound in England (C.G.O. pers. observ.) except for a difference in organic cement patterns.This feature alone is not considered sufficient for specific recognition. 372 C. G. OGDEN & A. ZlVKOVIC Pontigulasia montana sp. nov. DESCRIPTION. The shell is pyriform, bluntly rounded aborally for one-third of its length andusually tapering gradually from the widest point towards the aperture (Fig. 24a), althoughsometimes there is a distinct constriction at the position of the internal aperture (Fig. 24c). Itis composed of a mixture of mainly medium to large pieces of quartz arranged to give a roughsurface. Organic cement is often seen as part of the shell matrix (Fig. 24d), in the form of an Fig. 24 Pontigulasia montana sp. nov.: a, lateral view of specimen without neck constrictionx 730; b, apertural view with the internal aperture arrowed x 730; c, lateral view of specimenwith distinct neck constriction x430; d, portion of shell surface showing organic cement as partof the shell matrix x4400; e, detail of organic cement network x 14000. DIFFLUGIDAE FROM YUGOSLAVIA 373 open network with walls 200 nm thick and a mesh diameter of 400 nm (Fig. 24e). Theexternal aperture is roughly circular and surrounded by small or medium particles randomlypositioned to give it an irregular margin (Fig. 24b). The smaller, single, internal aperture isalso circular and surrounded by small particles overlaid with organic cement. MEASUREMENTS (in um). See Table 3.Table 3 Range of measurements (in um) of two species ofPontigulasia REMARKS. The specimens described here are similar to Pontigulasia bryophila Penard, 1902,but the latter species has a more streamlined, elongate shell, with smaller internal andexternal openings, plus a distinct organic cement pattern. Comparison of the measurementsgiven in Table 3 emphasizes the difference in body and apertural size between these twospecies, furthermore, P. bryophila appears to be the only previously described species ofPontigulasia with a single internal aperture. ETYMOLOGY. This species is named after the geographical topography surrounding LakeVlasina, which is mountainous (L. montana). Summary The information provided by this study extends our knowledge of the family Difflugiidae, inparticular the finding of compressed species of Difflugia. The scarcity of these forms isperhaps highlighted by the review of African species in which Gauthier-Lievre & Thomas(1958) list twelve compressed species, of these nine were new, out of a total of one hundredand thirty-three. Thirty-three species of Difflugia are listed here of which thirteen are re-described, including two former varieties of D. oblonga namely D. lata and D. nodosa, sixnew species are described D. balcanica, D. bistrica, D. dragana, D. serbica, D. serrata andD. styla the first four having compressed shells. In addition, two other new species aredescribed Cucurbitella vlasinensis sp. nov. and Pontigulasia montana sp. nov. It has been mentioned before (Ogden, 19806) that the size and shape of the aperture inpyriform species of Difflugia is a relatively uniform character, but in ovoid specimensappears more variable as shown here in the description of D. lismorensis. Morfoloska studija nekih Difflugiidae iz Jugoslavije (Rhizopoda, Protozoa) REZIME Podatke date u ovom radu prosiruju nasa znanja o familiji Difflugiidae, narocito u pogleduotkrivanja pljosnatih vrsta Difflugia. Retkost ovih vrsta mozda se najbolje ogleda u radu oafrickim vrstama, Gauthier-Lievre i Thomas (1958), u kome je dat spisak od ukupno 133vrsta; medju njima je dvanaest pljosnatih vrsta, od kojih je devet novih. Ovaj rad sadrzispisak od tridesettri vrste Difflugia, od kojih su trinaest ponovo opisane, ukljucujuci i dyaranija varijeteta: D. oblonga, odnosno D. lata i D. nodosa, a opisano je i sest novihvrsta D. balcanica, D. bistrica, D. dragana, D. serbica, D. serrata i D. styla, od kojih prve 374 C. G. OGDEN & A. 2IVKOVIC cetiri poseduju pljosnate ljusture. Opisane su i dve nove vrste Cucurbitella vlasinensis sp.nov. i Pontigulasia montana sp. nov. U ranijim publikacijama ukazano je (Ogden, 19806) da su velicina i oblik pseudostoma upiriformnim vrstama Difflugia relativno konstantna, medjutim, pojavljuje se veca vari-jabilnost u ovoidnim uzorcima, kao sto je slucaj u D. lismorensis. Prouceni materijal sakupljen je 9 septembra 1947. godine iz jedne bare sa prostranesfagnumske tresave na Vlasini (1200 m) koja je izgradnjom akumulacionogjezera izcezla. References Bartos, E. 1963. Rhizopoden einiger moosproben aus Java. Ada Univ. Carol, 1 19-190. Bereczky, M. C. 1973. Beitrage zur Kenntinis der im Eprofundal des Balaton lebenden testaceen. Annales Univ. Sclent, bpest Rolando Eotvos (Sect. Biol.) 15:1 17-127. Boltovskoy, E. & Lena, H. 1974. Tecamebas del Rio de la Plata. Armada Argentina, H660 32 pp.Cash, J., Wailes, G. H. & Hopkinson, J. 1919. The British Freshwater Rhizopoda and Heliozoa. Vol. IV. Supplement to the Rhizopoda. The Ray Society, London. 1 30 pp.Chardez, D. 1961. Sur Difflugia acuminata Ehrenberg (Rhizopoda, Testacea). Bull. Inst. agron. Stns. Rech. Gembloux29 : 301-308. 1964. Thecamoebiens (Rhizopoda, Testacea). Expl. hydrobiol. Bangweolo-Luapula 10 (2): 1-77. 1973. Sur Difflugia ventricosa Deflandre, 1926 (Rhizopode, Thecamoebien). Revue Verviet. Hist. nat. 30 (7-9) : 57-60. 1980. Thecamoebiens de 1'Etang de Thommen (Protozoa, Rhizopoda, testacea). Naturalistes belg.61 : 88-91.Chardez, D. & Caspar, C. 1976. Thecamoebiens aquatiques du domaine des Epioux (Ardenne, Belgique). Biol. Jaarb. 44 : 86-100.Deflandre, D. 1926. Notes sur quelques Rhizopodes et Heliozoaires du Venezuela. Bull. Soc. zool Fr. 51: 51 5-530. 1931. Thecamoebiens nouveaux ou peu connus. 1 . Annals. Protist. 3 : 81-95. Dioni, W. 1970. Taxocenos de tecamebianos en cuencas islenas del Parana Medio 1. Acta zool. lilloana 21: 201-240.Ertl, M. 1965. Zur Kenntnis der Testaceenfauna der slowakischen Reisfelder. Hydrobiologia 26 : 13-20.Frenzel, J. 189 . Untersuchungen iiber die mikroskopische Fauna Argentiniens. I. Die Protozoen. I & II Rhizopoden und Helioamoeba. Bibliotheca Zoologica, Stuttgart 12 : 162 pp., X pi.Gauthier-Lievre, L. & Thomas, R. 1958. Les genres Difflugia, Pentagonia, Maghrebia et Hoogenraadia (Rhizopodes, testaces) en Afrique. Arch. Prostistenk. 103 : 241-370. & 1 960. Le genre Cucurbitella Penard. Arch. Protistenk. 104 : 569-602. Godeanu, S. 1972. Especes nouvelles de thecamoeniens (Protozoa, Rhizopodea, Arcellinida). Revue roum. Biol. (Ser. Zool.) 17 : 227-236.Godeanu, S., Godeanu, M. & lonescu, V. 1973. Variata anualS a planctonului din incinta indiguita Obretin (Delta DunSrii). Peuce 3:21 1-260.Green, J. 1975. Freshwater ecology in the Mato Grosso, Central Brazil, IV. Associations of testate Rhizopoda,/. nat. Hist. 9 : 545-560.Hoogenraad, H. R. & Groot, A. A. de 1940. Fauna van Nederland Vol. 9. Zoetwaterrhizopoden en Heliozoen. 302 pp. Leiden.Jennings, H. S. 1916. Heredity, variation and the results of selection in the uniparental reproduction of Difflugia corona. Genetics 1 : 407-534. 1937. Formation, inheritance and variation of the teeth in Difflugia corona. A study of the morphogenic activities of rhizopod protoplasm. J. exp. Zool. 77 : 287-336. Jung, W. 1936. Thekamoben urspriinglicher, lebender deutscher Hochmoore. Abh. Landesmus.Prov. Westf. 1 (4): 3-80. 1942. Siidchilenische Thekamoben (Aus dem siidchilenischen Kiistengebiet, Beitrag 10), Arch. Protistenk. 95 : 253-356. Kourov, O. 1925. Faune rhizopodique des bassins de Kossino. Trudy kosin. biol. Sta. 2 : 43-68.Laminger, H. 1973. Die Testaceen (Protozoa, Rhizopoda) einiger Hochgebirgsgewasser vonMexiko, Costa Rica und Guatemala. Int. Revue ges. Hydrobiol. 58 : 273-305. 1975. Die Sukzession der Testaceen-Assoziationen (Protozoa, Rhizopoda) im rezenten und subfossilen Sphagnum des Obersees bei Lunz (Niederosterreich). Hydrobiologia 46 : 465-487. DIFFLUGIDAE FROM YUGOSLAVIA 375 Leidy, J. 1879. Freshwater Rhizopods of North America in Vol. 12. United States Geological Survey of the Territories. 324 pp. Washington.Lena, H. & Zaidenwerg, S. J. 1975. Tecamebas del delta del Parana (Argentina). Revta esp. Micropaleont. 1 : 519-537.Levander, K. M. 1894. Materialien zur Kenntniss der Wasserfauna in der Umgebung von Helsingfors, mil besonderer beriicksichtigung der Meeresfauna. I. Protozoa, Acta Soc. Fauna Flora Few?. 12 No. 2: 1-H5.Milovanovic, D. & Zivkovic, A. 1956. Limnolska ispitivanja baraznog jezera na Vlasini. Zborn. Rad. Inst. Ekol. Biogeogr. 1 (5) : 1^7.Ogden, C. G. 1979. Comparative morphology of some pyriform species of Difjlugia (Rhizopoda). Arch. Protistenk. 122 : 143-153. 1980a. Shell structure in some pyriform species of Difflugia (Rhizopoda). Arch. Protistenk. 123:455^70. 19806. Notes on some Difflugiidae from Norfolk (Rhizopoda, Protozoa). Bull. Brit. Mus. nat. Hist. (Zool.) 39: 125-138. 1983. Observations on the systematics of the genus Difflugia in Britain (Rhizopoda, Protozoa). Bull. Brit. Mus. Nat Hist (Zool.) 44 : 1-73.Ogden, C. G. & Fairman, S. 1979. Further observations of pyriform species of Difflugia (Rhizopoda). Arch. Protistenk. 122 : 372-381.Ogden, C. G. & Hedley, R. H. 1980. An Atlas of Freshwater Testate Amoebae. British Museum (Natural History), London & Oxford University Press, Oxford. 222 pp. Oye, P. van 193 1 . Rhizopoda from South Africa. Rev. Zool. hot. afr. 21 : 54-73. 1932. Neue Rhizopoden aus Africa. Zool. Anz. 99 : 323-328. Penard, E. 1890. Etudes sur les Rhizopodes d'eau douce. Mem. Soc. Phys. Hist. Nat. Geneve 31 : 1-230. 1891. Rocky Mountain Rhizopods. Am. Nat. 25 : 1070-1083. 1902. Faune Rhizopodique du Bassin du Leman. Geneve. 700 pp. 191 1. On some Rhizopods from Sierra Leone. J. Quekett. microsc. Club Ser. II. 11 : 299-306. Playfair, G. J. 1 9 1 8. Rhizopods of Sydney and Lismore. P roc. Linn. Soc. N.S. W. 42 : 633-675.Rampi, L. 1950. Su alcuni Laghetti Alpini del Massiccio dell'Abisso (Alpi Marittime). Boll. Pesca Piscic. Idrobiol. 26 : 207-224. Schonborn, W. 1965. The sediment-inhabiting Testacea from some Masurian Lakes. Acta_ Protozool. 3 : 297-309.Stepanek, M. 1953. Rhizopoda jako biologicke indikatory znecisteni vod. Priord. Shorn. Ostravsk. Kraje 14: 470-505.Thomas, R. 1953. Sur deux formes critiques du genre Difflugia Leclerc. Bull. Soc. zool. Fr. 78: 132-136. 1954. Thecamoebiens de la region Bordelaise. Bull. Soc. Hist. nat. Toulouse 89 : 245-264. Vucetich, M. C. 1970. Algunos tecamebianos de la provincia de Formosa. Neotropica 16 : 42-48.Wallich, G. C. 1864. On the extent, and some of the principal causes, of structural variation among the Difflugian Rhizopods. Ann. Mag. nat. Hist. (3) 13 : 215-245. Manuscript accepted for publication 23 July 1982 British Museum (Natural History) An Atlas of Freshwater Testate Amoebae C. G. Ogden & R. H. Hedley 1980, Hardcovers, 222pp, 17.50 (18.00 by post). Co-published by British Museum(Natural History) and Oxford University Press. This book illustrates, using scanning electron micrographs, most of the commonspecies of testate amoebae that are found in freshwater habitats. Information onthe biology, ecology, geographical distribution and a classification are followed bydescriptions of ninety-five species. Each of these is illustrated by several views ofthe shell. The text is designed not only to enable biologists to identify species of testateamoebae, but to serve as an introduction to students interested in the taxonomyand biology of these freshwater protozoa. It will be of special interest toprotozoologists, ecplogists, limnologists, water treatment specialists andmicropalaeontologists interested in recent sediments. British Museum (Natural History)Publication Sales,Cromwell Road,London SW7 5BD. Titles to be published in Volume 44 Observations on the systematics of the genus Difflugia inBritain (Rhizopoda, Protozoa).By Colin G. Ogden Miscellanea A review of the Euplotidae (Hypotrichida, Ciliophora). By Colin Curds & Irene C. H. Wu The Opthalmotilapia assemblage of cichlid fishes reconsidered. By Peter Humphry Greenwood Osteology, genitalia and relationships of the Acanthodactylus(Reptilia: Lacertidae). By E. N. Arnold Morphological studies on some Difflugiidae from Yugoslavia(Rhizopoda, Protozoa). By Colin G. Ogden & Andjelija Zivkovic Printed by Henry Ling Ltd, Dorchester BOUND 2 JUL 1988