" Bulletin of the British Museum (Natural History) Zoology series Vol 34 1978-1979 British Museum (Natural History) London 1979 Dates of publication of the parts Nol No 2 No 3 No 4 No 5 No 6 29 June 1978 30 November 1978 21 December 1978 22 February 1979 29 March 1979 26 April 1979 ISSN 0007-1498 Printed in Great Britain by Henry Ling Ltd, at the Dorset Press, Dorchester, Dorset Contents Zoology Volume 34 Page No 1 The anatomy and relationships of the cyprinid fish Luciobrama macrocephalus (Lacepede) G. J. Howes 1 No 2 A new species ofPteralopex Thomas, 1888 (Chiroptera: Pteropodidae) from the Fiji Islands J. E. Hill & W. N. Beckon 65 No 3 A revision of the spider genus Portia (Araneae: Salticidae) F. R. Wanless 83 No 4 Anatomical specimens of birds in the collections of the British Museum (Natural History) J. S. Blandamer & P. J. K. Burton 125 No 5 Amphibians and reptiles from northern Trengganu, Malaysia, with descriptions of two new geckos : Cnemaspis and Cyrtodactylus J. C. M. Dring 181 No 6 A revision of the species of Sertulariidae (Coelenterata : Hydroida) recorded from Britain and nearby seas P. F. S. Cornelius 243 Bulletin of the British Museum (Natural History) The anatomy and relationships of the cyprinid fish Luciobrama macrocephalus (Lacepede) G. J. Howes Zoology series Vol34Nol 29 June 1978 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology and Zoology, and an Historical series. Parts are published at irregular intervals as they become ready. Volumes will contain about four hundred pages, and will not necessarily be completed within one calendar year. Subscription orders and enquiries about back issues 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), 1978 ISSN 0007-1498 Vol 34 No 1 pp 1-64 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 29 June 1978 The anatomy and relationships of the cyprinid fis Luciobrama macrocephalus (Lacepede) G. J. Howes Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD Contents Synopsis. ............. 1 Introduction ............. 1 List of study material 2 Abbreviations used in text figures ......... 5 Luciobrama macrocephalus .......... 7 Osteology 7 Cranial myology 21 Comparative analysis ........... 25 Osteological characters 26 Myological characters 52 Considerations of functional morphology of Luciobrama ..... 58 Summary ............. 58 Relationships of Luciobrama .......... 59 Relationships of Elopichthys .......... 59 The aspinine group of cyprinids ......... 59 Discussion ............. 61 Acknowledgements ............ 62 References ............. 62 Synopsis The osteology and cranial myology of the long-headed cyprinid fish, Luciobrama macrocephalus, is described and aspects of its cranial functional morphology are considered. On the basis of shared derived characters the closest relatives of Luciobrama are identified as Aspiolucius, Pseudaspius, Aspius and Elopichthys which together form a monophyletic assemblage termed the aspinine group. It is suggested that this group is related to an as yet unidentified monophyletic aggregate of the 'Leuciscinae'. The osteological characters of many other cyprinid genera are compared and commented upon in the light of their useful- ness as indicators of relationship and in terms of function. The cranial myology of Elopichthys bambusa and Barilius bo la is described and figured. Introduction Although the Cyprinoidei display a great degree of diversity in their morphology and trophic specializations, there appear to be fewer piscivorous taxa than amongst their presumed sister group of ostariophysans, the Characoidei. This is undoubtedly due, in part at least, to the fact that jaw teeth are lacking in cyprinoid fishes. On the other hand, a characteristic of cyprinoids is the protusile mechanism of the upper jaw and the correlated development of a suction method of feeding (see Alexander, 1964, 1966 & 1967). Because of this particular type of jaw mechanism, the evolution of pike-like piscivores, a habitus (of piscivore) so prevalent amongst the characoids and other teleost groups, has been severely limited. To my knowledge, only one such fish has evolved amongst the cyprinids, namely, Elopichthys (which is discussed in this paper). The reasons for the absence of this type of predatory facies, and those which account for the other types manifest in present day old-world cyprinid lineages, are presented on page 61. Bull. Br. Mus. nat. Hist. (Zool.) 34 (1): 1-64 Issued 29 June 1978 2 G. J. HOWES In a previous paper (Howes, 1976) I described, in part, the cranial anatomy of a piscivorous cyprinid, Macrochirichthys macrochirus. From this initial study my attention was drawn to other piscivorous cyprinids and in particular to Luciobrama macrocephalus. Because Luciobrama is so obviously specialized, it seemed that it might offer clues to the recognition of primitive and advanced character states in other groups of cyprinids. In order to detect these and to find the closest relatives of Luciobrama on the basis of shared specializations, comparisons have been made with a number of other cyprinid genera (see p. 25). As pointed out later (p. 61) the classification of cyprinid fishes is in an unsatisfactory state and it would be futile and pretentious at this stage to offer any phylogenetic scheme for subfamilial classification. What has become clear during this study is that few of the presently recognized cyprinid subfamilies are monophyletic groups. List of study material Species BMNH register number Type of preparation and standard lengths (mm) of alcohol preserved specimens Locality Abramis brama 1864.4.2:12 Skeleton Holland Abramis brama Unregistered Skeleton Unknown Alburnus alburnus Unregistered Alizarins; 37, 40, 58 Thames Aspidoparia morar 1872.4.17:81 92 Bengal Aspiolucius esocinus (Types) 1897.7.5:31-32 244, 107 Amu-Daria Aspiopsis merzbacheri (Type) 1914.3.2:1 166 Central Asia Aspius aspius Unregistered Two skulls Danube Aspius aspius 1860.3.7:34 225 Danube Aspius aspius 1976.4.1:1 110 Mures R. Aspius vorax 1920.3.3:127-146 235 Basra Aspius vorax 1874.4.28:30 Skeleton Tigris R. Barbus altianalis Unregistered Skeleton Lake George Barbus altianalis Unregistered 142 Lake George Barbus altus 1882.2.11:314 145 Thailand Barbus argenteus 1907.6.29:217 Skeleton Angola Barbus aspilus 1909.4.29:14 Skeleton JaR. Barbus barbus Unregistered Skeleton England Barbus barbus 1908.12.28:123 Skeleton England Barbus callensis 1869.1.29:4 Skeleton Algeria Barbus callipterus 1975.1.17:201-210 53 Togo Barbus camptacanthus Unregistered Skeleton W. Africa Barbus canis 1864.8.23:24 Skeleton Jordan Barbus capensis Unregistered Skeleton BurgR. Barbus fritschii 1904.11.18:59 Skeleton Morocco Barbus gonionotus 1974.10.10:823-824 144 E. Java Barbus guirali 1902.11.12:119 Skeleton Kribi R. Barbus harterti 1902.7.28:35 Skeleton Morocco Barbus harterti 1903.10.29:16 Skeleton Morocco Barbus holotaenia 1902.11.12:122 Skeleton Kribi R. Barbus intermedius 1902.12.13:291 Skeleton Errer R. Barbus kersteni 1961.12.1:156 Skeleton Aswa R. Barbus longiceps Unregistered Skeleton Galilee Barbus mariae 1936.12.22:35-39 370 Athi R. Barbus nasus 1902.1.4:22 Skeleton Morocco Barbus natalensis 1862.8.28:8 Skeleton Natal Barbus orphoides 1974.10.10:865-872 90 Java Barbus oxyrhynchus 1893.12.2:31 Skeleton Thikathike R. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) Species BMNH register number Type of preparation and standard lengths (mm) of alcohol preserved specimens Locality Barbus paludinosus 1902.1.4:72 Skeleton Morocco Barbus perince 1907.12.2:3745 Skeleton Nile Barbus progeny s 1903.7.28:155 Skeleton Cameroon Barbus radiatus 1961.12.10239 Skeleton Aswa R. Barbus reinii 1903.10.29:10 Skeleton Morocco Barbus rocadasi 1911.6.1:26 Skeleton Angola Barbus sarana 1889.9.26:99-103 94 Deoli Barbus setivemensis 1869.1.29:21 Skeleton Unknown Barbus somereni 1971.1.5:96-99 Alizarins; 76, 60 Ruimi R. Barbus tor 1889.2.1:523 Skeleton Assam Barbus trimaculatus 1907.4.9:98 Skeleton Transvaal Barilius bendelisis 1889.10.29:37 Skeleton Dehra Dun Barilius bola 1889.2.1:1205 Skeleton Assam Barilius bola 1889.9.26:118-127 235 Deoli Barilius bola 1867.5.12:20-22 135 Norar R. Barilius gatensis 1889.2.1:1135-1139 107 Nilgherriss Barilius loati 1907.12.2:1503-1512 132 Gondokoro Barilius loati 1907.12.2.3748 Skeleton Nile Barilius microcephalus 1906.9.7:4 Skeleton L. Malawi Barilius microcephalus 1974.1.15:25-26 210 L. Malawi Barilius moori 1974.3.19:1-5 102-137 L. Kivu Barilius niloticus Unregistered Alizarins; 23-5-40 L. Rudolf Barilius niloticus 1907.12.2:3764-3767 Skeletons Nile Barilius ornatus 1893.6.30:61-70 75-93 Shan States Barilius ubangensis 1902.11.12:142-148 88 Kribi R. Barilius ubangensis 1903.7.28:165 Skeleton Kribi R. Barynotus luteus 1874.4.28:23 Skeleton Tigris R. Catla catla 1908.12.28:122 Skeleton Hooghly R. Catostomus teres 1866.12.13:5 Skeleton Montreal Chela laubuca 1889.2.1:1356-1359 60 Madras Chelaethiops sp. Unregistered Alizarins; 20-5-30-0 L. Rudolf Chondrostoma nasus 1976.4.1:4-6 148 Romania Chondrostoma polylepis 1885.1.29:30 Skeleton Coruna Ctenopharyngodon idellus 1888.5.15:25 Skeleton Tchang Culler brevicauda 1891.1.31:28 Skeleton China Cyprinus carpio Unregistered 122 Lincolnshire Cyprinus carpio Unregistered Skeleton Unknown Cyprinus carpio Unregistered Skeleton Amoy Cyprinus carpio Unregistered Skull Unknown Danio aequipinnatus 1894.5.21:56-60 72 Byingi Elopichthys bambusa 1889.6.24:51 Skeleton Kiu Kiang Elopichthys bambusa 1936.10.19:20 Alizarin; 116 Tatung Elopichthys bambusa 1936.10.19:13-19 185, 250 Tatung Elopichthys bambusa Unregistered 220 China Erythroculter ilishaeformis 1936.10.19:31-34 176 Hankow Erythroculter mongolicus 1889.6.24:57 Skeleton Kiu Kiang Esomus danricus 1863.12.8:108-114 85 Madras Garra blandfordi 1902.12.13:420 Skeleton Jerrer R. Garra lamta Unregistered Skeleton Unknown Hemiculter leucisculus 1938.12.1:26 54 Chusan Hemiculterella polylepis 1921.7.26:21-23 132 Yunnan Lake Hypophthalmichthys molitrix 1895.5.31:22 Skeleton China Labeo bata 1889.2.1:206 Skeleton Assam G. J. HOWES Species BMNH register number Type of preparation and standard lengths (mm) of alcohol preserved specimens Locality Labeo coubie 1907.12.2:3744 Skeleton Kosheh, Nubia Labeo coubie 1911.3.31:21-29 Alizarin; 52 Aboina R. Labeo chrysophekadion 1898.11.8:115 Alizarin; 63 Menam R. Labeo cylindricus 1902.5.26:23 Skeleton Tana Labeo diplostomus 1889.2.1:163 Skeleton Hardwar Labeo macrostoma 1904.5.2:158-160 140 Angola Labeo niloticus 1907.12.2:975-980 122 L. Menzaleh Labeo rohita 1889.2.1:138-139 143 Calcutta Labeo rohita 1858.8.15:50 Alizarin, 83 India Labeo stoliczke 1891.11.30:286 Skeleton Sittang R. Leuciscus cephalus 1867.4.2:15 Skeleton Holland LeuciscUs idus 1867.4.2:6 Skeleton Holland Luciobrama macrocephalus 1889.6.24:48 Skeleton Kiu Kiang Luciobrama macrocephalus 1928.4.24:15 273 Nanking Luciobrama macrocephalus 1896.6.24:46 420 Kiu Kiang Luciobrama macrocephalus 1888.5.15:31-32 458 Tchang Luciosoma bleekeri 1898.11.8:114 Skeleton Menam R. Macrochirichthys macrochus 1898.11.8:121 Skeleton Menam R. Macrochirichthys macrochus 1898.4.2:243 212 Menam R. Megalobrama bramula 1936.10.19:21 111 Hankow Megalobrama macrops Unregistered Skeleton Formosa Myxocyprinus chinensis 1889.6.24:10 Skeleton Kiu Kiang Notropis hudsonius 1925.2.3:121-125 68 Mississippi R. Ochetobius elongatus 1936.10.19:35-38 147 Tatung, China Ochetobius elongatus 1889.6.8:56 330 Kiu Kiang Opsariichthys uncirostris 1901.3.6:9 Skeleton Ningpo Opsariichthys uncirostris 1923.3.5:6-12 152 Seoul Oreinus sinuatus 1889.2.1:64-72 177-205 Simla Oreoleuciscus pewslowi 1975.1.17:259-265 118-178 Mongolia Oreoleuciscus potanini 1891.10.7:26-27 174 Mongolia Oxygaster anomalura 1881.3.21:3 195 Sarawak Parabramis pekinensis 1936.10.19:22-23 125 Hankow Parabramis pekinensis 1889.6.8:46-53 235 Kiu Kiang Parapelecus argenteus Unregistered 166 China Pelecus cultratus 1879.11.14:36 Skeleton Syr Darya Pelecus cultratus 1966.2.22:1-2 174, 175 Romania Phoxinus lagowskii 1974.8.6:21-30 81 Onon R. Mongolia Phoxinus phoxinus 1967.12.18:1-13 66 Kysuka R. Pseudaspius leptocephalus 1925.8.6:28 137 Amur R. Pseudolaubuca sinensis 1889.6.24:61 Skeleton Kiu Kiang Pseudolaubuca sinensis 1889.6.24:59-60 195 Kiu Kiang Pseudoxygaster gora 1934.10.7:54 137 Allahabad Rasbora argyrotaenia 1974.10.10:1801-1805 50-79 Bali Rutilus friesii Unregistered Skeleton L. Derkus Rutilus rutilus Unregistered Skeleton England Rutilus rutilus Unregistered Alizarins; 66, 76, 77 England Salmostoma bacaila 1889.9.26:145-154 90, 107 Rajputana Salmostoma sardinella 1891.11.30:374-383 85,98 Sittang R. Saurogobio dumerili 1889.6.24:21 Skeleton Kiu Kiang Scardinius erythropthalmus 1867.4.2:7 Skeleton Holland Schizothorax esocinus 1870.11.30:40 Skeleton Kashmir Schizothorax esocinus 1870.11.30:39 260 Kashmir Schizothorax grahami 1969.4.15:118 184 Kuan Hsien THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) Species BMNH register number Type of preparation and standard lengths (mm) of alcohol preserved specimens Locality Semiplotus macclellandi 1889.2.1:869 Skeleton Assam Squaliobarbits curriculus 1889.6.8:34-38 138 Kiu Kiang Squaliobarbus curriculus 1888.5.15:29 Skeleton Tchang Varicorhinus beso 1968.7.24:18-19 Alizarins; 80, 114 L. Tsana Varicorhinus steindachneri 1910.11.28:158 Skeleton Lucalla R. Varicorhinus tanganicae 1906.9.6:11 Skeleton L. Tanganyika Zacco platypus 1865.5.2:30 Skeleton China Zacco platypus Unregistered 80 Locality unknown Zacco spilurus 1939.3.23:14-16 74 Kowloon Zacco spilurus (Types) 1956.2.25:1-5 34-5^6 Hong Kong Zacco temmincki 1905.6.7:61-65 150 Japan In addition to being dissected, all the alcohol preserved specimens were radiographed. Abbreviations used in text figures Skeletal elements AA Anguloarticular AHY Anterohyal APTE Autopterotic BB 1-4 Basibranchials BO Basioccipital BSR Branchiostegal rays C 1-5 Ceratobranchials CIM Cranial intermuscular bones CL Cleithrum CLA Claustrum COR Coracoid D Dentary DHY Dorsohyal DPT Dermopterotic DSP Dermosphenotic E Ethmoid ECT Ectopterygoid ENT Entopterygoid EP Epural EPI \-^ Epibranchials EPO Epioccipital ES Extrascapular EX Exoccipital F Frontal H Hyomandibula HB 1-4 Hypobranchials HF Hyomandibular fossa HY 1-6 Hypurals 1C Intercalar IF Infrapharyngobranchial INC Intercalarium IO Infraorbitals IOP Interoperculum KE Kinethmoid LE Lateral ethmoid 6 G. J. HOWES LF Lateral foramen LP1 Lateral process of the 1st vertebra LP2 Lateral process of the 2nd vertebra MAX Maxilla MC Mesoco racoid MET Metapterygoid METP Metapterygoid process N Nasal NC Neural complex of Weberian apparatus NP2 Neural plate of 2nd vertebra NP3 Neural plate of 3rd vertebra OP Operculum OS Orbitosphenoid OSS Os suspensorium PA Parietal PAL Palatine PE Preethmoid PC Postcleithrum PH Parhypural PHY Posterohyal PMX Premaxilla PO Preoperculum PP Pharyngeal process PR4 Lateral process of 4th vertebra (modified pleural rib) PRO Prootic PS Parasphenoid PTF Posttemporal fossa PTS Pterosphenoid PIT Posttemporal PU1 + Ul Fused preural and ural centra Q Quadrate RA Retroarticular SCA Scaphium SCP Scapula SE Supraethmoid SN Supraneural SO Supraoccipital SOR Supraorbital SP Autosphenotic STF Subtemporal fossa SUB Suboperculum SY Symplectic TR Tripus UN Uroneural V Vomer VHY Ventrohyal Muscles and soft tissues Al, A2, A3 Divisions of the adductor mandibulae muscle aap Adductor arc us palatini ah Adductor hyomandibulae ct Connective tissue do Dilatator operculi hh Hyohyoidei im Intermandibularis km lig Kinethmoid-maxillary ligament lap 1, 2 Divisions of the levator arcus palatini Hg Ligament lo Levator operculi THE ANATOMY OF LUCfOBRAMA MACROCEPHALUS (LACEPEDE) J Isa Ligamentous sheet connecting supraneural to supraoccipital Isb Ligamentous sheet connecting neural complex to supraoccipital obv 1-3 Obliqiti vent rales pee Pharyngoclavicularis externus pci Pharyngoclavicularis internus ph Protractor hyoidei rv 1-3 Recti ventrales sb Swimbladder sth Sternohyoideus tf Tendinous fascia of adductor mandibulae A3 tlap Ventral tendon of levator arcus palatini tv Transversus Luciobrama macrocephalus (Lacepede) (Fig. 1) Synodus macrocephalus Lacepede, 1803, Hist. Nat. Poiss. 5 : 322, pi. IX, fig. 1 (described from a Chinese drawing). Luciobrama typus Bleeker, 1870, Versl. Meded. K. Akad. wet. Amst. (2) 4 : 252 (Yangtse-Kiang). Luciobrama macrocephalus'. Bleeker, 1873, Ned. Tijdschr. Dierk 4 : 89 (re-description). Luciobrama is a monotypic genus (see p. 60) of east Asian and Chinese piscivorous cyprinids (see Nichols, 1925 & 1943 for a general account of this habit in the cyprinids). Stomach contents that I have examined have revealed the remains of small (ca 80-100 mm SL) cyprinid fishes. Specimens of Luciobrama macrocephalus grow to large size; Kimura (1934) records total lengths of over 700 mm and weights of 1000 g. The external morphology of this fish has been described adequately by Bleeker (1873), Rendahl (1928), Kimura (1934) and Lin (1935). Kimura (op. cit.) also cited all references to the species up to that date. Since then the following accounts have been published. Chu (1935) gave an account of scale morphology and described the pharyngeal bones and teeth; Nichols (1943) and Wu (1964) both gave descriptions and noted the distribution of the species. It is unnecessary to repeat the detailed descriptions of external characters given by these authors, save to note the absence of barbels, the ellipsoid shape of the orbit and the small scales (up to 155 in the lateral line). Nuptial tubercles have not been detected in any specimens examined, but have been reported as occurring in this species (see review by Wiley & Collette, 1970). Osteology Circumorbital series (Fig. 2) The first infraorbital (lachrymal) is an almost square plate bearing 12 pores of the lateralis canal along its ventral border. The second infraorbital is very narrow and borders the entire ventral margin of the orbit. It is joined to the third just below the posterior border of the eye. The third infraorbital is lamellate, the lateralis canal bearing 6-7 pores. It is an elongate bone extending in an almost horizontal plane to a point well beyond the posterior margin of the eye before joining the fourth infraorbital. The fourth infraorbital is reduced to the canal tube. It diverges from the third at an angle of about 45° across the postorbital part of the head. The fifth is minute and is sometimes fragmented. It consists of only the canal tube. (See page 26 for further discussion of these bones.) The supraorbital (SOR, Figs 3 & 5) is large being bordered anteriorly by the lateral ethmoid and latero-posteriorly by the frontals. Ethmo-vomerine region The kinethmoid (KE, Fig. 7) is a short columnar bone, the dorsal surface bears a wedge-shaped groove and the ventral surface is rounded. It is connected by two ventral ligaments to the heads of the vomer and by laterally extending ligaments to the maxillary ascending processes. G. J. HOWES Fig. 1 Outline drawings of: A. Luciobrama macrocephalus, B. Aspiolucius esocinus, C. Pseudaspius leptocephalus, D. AspiUs vorax, E. Elopichthys bambusa. Scale = 5 cm. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 9 The supraethmoid (SE, Figs 3, 4 & 5) is narrow with thin wing-like extensions posteriorly. Laterally the bone is bordered by the nasals and posteriorly by the frontals. The anterior border is rounded with a median notch. The ethmoid (E, Figs 3 & 4) underlies the supraethmoid and overlies the vomer. It is bifurcated anteriorly, the two short arms provide the medial surfaces of the fossae for the preethmoids. Antero-dorsally there is a small foramen separating the bone from the supraethmoid. Posteriorly there is a wide synchondrosis with the lateral ethmoid. The lateral ethmoids (including fused prefrontals; LE, Figs 3, 4 & 5) extend to protrude well beyond the lateral margins of the cranium. Basally each bone is triangular, supporting anteriorly a thick, curved ascending wall which connects the supraorbital. Posteriorly it extends as a thin wall meeting the orbitosphenoid in a synchondrosis. Dorsally, each bone is overlain by its correspond- ing frontal; ventrally, contact is made with the parasphenoid. Medially, the lateral ethmoids are in contact. Fig. 2 Luciobrama macrocephalus, infraorbital series. The vomer (V, Figs 4 & 5) is a thin, lamellate bone extending posteriorly to a point well beyond the centre of the orbit. It is overlain by the parasphenoid and the ethmoid. Anteriorly it flares into the shape of a double club, the arms of which provide the lower surfaces of the preethmoid fossae. The preethmoids (PE, Figs 3 & 4) are irregular ovate bones articulating with the ethmoid and the vomer. They are covered by the cartilage upon which rest the autopalatines. The nasals (N, Figs 3 & 4) border the supraethmoid anteriorly and the frontals posteriorly. They are in the form of long tubes bearing 9-10 pores. Laterally each nasal is attached by skin to the first infraorbital. The frontals (F, Figs 3, 4, 5 & 6) are extremely long and narrow, sutured for their entire length. Anteriorly their lateral borders slope downward to contact the supraorbitals. The canal tube runs along the lateral edge of each bone from the anterior edge of the pterotic in an almost straight line to the nasal. It bears 22 pores. Orbital region The orbitosphenoids (OS, Figs 4, 5 & 6) are greatly depressed bones 'sandwiched' between the frontals and the parasphenoid. Contact with the parasphenoid is along a third of that bone's orbital length. Anteriorly, the orbitosphenoids join the lateral ethmoids by a wide synchondrosis. Dorso-posteriorly, each bone extends as a long arm which diverges slightly from its fellow. These arms contact similar anterior extensions of the pterosphenoid. Posteriorly, the lower part of each orbitosphenoid is directed medially and meets its partner in the midline to form a narrow, wedge-shaped septum. (This is not the 'orbitosphenoid septum' referred to later in this 10 G. J. HOWES LJJ M to THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) \\ paper as occurring in other cyprinids (p. 32), in those cases the septum extends from the ventral surfaces of the bones.) The pterosphenoids (PTS, Figs 4, 5 & 6) are extensive, forming almost half the length of the orbito-otic region. Each bone is produced into an anterior process along the ventral surface of the frontal which contacts the corresponding posterior extension from the orbitosphenoid. The posterior dorsal margin is sutured to the autosphenotic. Latero-posteriorly the bone extends as a wide branch which forms the anterior wall of the leading hyomandibular fossa. Below this the pterosphenoid border is concave, this concavity forming part of the border of the trigemino- facialis foramen; the lateral surface at this point is deeply grooved to allow for the nerve path ways. Ventrally, there is an extensive connection with the wide ascending wing of the parasphenoid. The pterosphenoids are narrowly separated from each other leaving only a small optic foramen. The parasphenoid (PS, Figs 4, 5 & 6) is wide anteriorly, the ventral surface below the lateral ethmoids is flat but becomes concave below the orbitosphenoids; beyond this point the bone narrows and deepens, the sides becoming thin walls which rise gradually until, as wide ascending processes, they contact the ventral margins of the pterosphenoids. The posterior border of the ascending process is separated from the pterosphenoid and prootic by an extensive hypophysial foramen. The dorso-posterior part of the parasphenoid is first overlapped by, and then runs abutted with, the prootic, followed by the anterior part of the exoccipital. The posterior border of the parasphenoid is sutured against the basioccipital. The ventral surface of the parasphenoid is rounded below the ascending processes, flaring slightly and becoming flattened below the prootic with a slight medial groove developing posteriorly. This groove leads into the aortic foramen of the basioccipital. The posterior part of the parasphenoid forms the floor and walls of the myodome. Otic region The prootics (PRO, Figs 4, 5 & 6) are large, forming long dorsal connections with the auto- sphenotics. Anteriorly, the border of each prootic with the parasphenoid is interrupted by the large trigemino-facialis foramen. Postero-ventrally, part of the prootic extends to overlap the parasphenoid. Anteriorly, each bone is in the form of a wedge-shaped arm which inserts partly between the pterosphenoid and the parasphenoid ; dorso-posteriorly, it contacts the pterotic. The area between the sphenotic and pterotic is bevelled and forms part of the hyomandibular facet. Postero-medially, the prootic forms the lower medial wall of the subtemporal fossa. Its posterior border is sutured to the epioccipital. A posterior myodome is present (ascertained by radiographs) and appears similar to that described for Aspius by Oliva and Skofepa (1968). The pterotics (APTE, Figs 3, 4, 5, 6 & 32) dorsally border the parietals and the frontals extending forward as narrow triangles overlying the autosphenotics. The outer margin of each bone bears the canal which contains 13 pores. Posteriorly, the pterotic is recessed as a facet for the hyomandibula (a continuation of that feature on the autosphenotic and prootic). Ventro- laterally, the surface is arched, the posterior foot of the arch joining the epioccipital and forming the outer roof of the subtemporal fossa (STF, Fig. 5). The lateral border continues posteriorly as a thick spine. Medially, the pterotic meets the epioccipital to form the high vaulted subtemporal fossa. The pterotic also contributes substantially to the formation of the posttemporal fossa (PTF, Figs 4 & 32) where it provides the lateral wall, part of the roof and the floor. The basioccipitals (BO, Figs. 4, 5 & 32) are sutured to the parasphenoid anteriorly and to the epioccipitals dorsally. Medially they extend forward between the prootics and form part of the roof of the posterior myodome. There is no obvious bulla acoustica lagenaris. From the ventral surface of each bone there extends posteriorly two processes which fuse distally to form the pharyngeal process (PP, Figs 3, 4 & 5). The 'masticatory plate' is virtually absent. The supraoccipital (SO, Figs 3, 4 & 32) is bordered anteriorly by the parietals and laterally by the epioccipitals, the whole area forming an almost flat platform. Medially, the bone rises as a low ridge which extends posteriorly as a thin plate-like process. This supraoccipital process barely rises above the highest level of the skull roof. 12 G. J. HOWES THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 13 The intercalar is absent. The parietals (PA, Figs 3 & 4) are elongate and join the frontals anteriorly. Laterally they are bordered, for part of their length, by the pterotics. Posteriorly, the parietals are deeply recessed, the roof of this recess providing insertion for the epaxial musculature, and the lateral border origin for the cranial intermuscular bones. The parietals gently slope posteriorly to join the supraoccipital. The presence of the transverse occipital sensory canal is indicated by four pores on the surface of each bone. The posttemporals (PTT, Fig 3). Each is a paddle-shaped bone, the stem of which extends dorso-anteriorly at an angle of 45° to rest along the epioccipital. The lamellate 'paddle' of the bone is joined to the supracleithrum. Ventrally, between the posttemporal and the pterotic border, there lies a long narrow extrascapular (ES, Fig 3). The autosphenotics (SP, Figs 3, 4, 5 & 6) are long depressed bones overlain by the pterotics. The anterior part of each sphenotic projects laterally from beneath the cranial border as a sloping shelf. Posteriorly, the dorsal surface of the bone presents two undulations which form a lateral cavity roofed by the pterotic. From this cavity arises the dilatator operculi muscle. The undulations of the bone also provide, ventro-laterally, the roofs of two deep facets for the hyomandibular condyles. The posterior facet is continued onto the pterotic and ventrally onto the prootic. No dermosphenotic can be identified. (See page 28 for comments on this bone.) The exoccipitals (EX, Figs 3, 4, 5 & 32) are seen laterally as almost square bones. Dorso- medially, each bone is in contact with the epioccipital and forms the inner surface of the sub- temporal fossa. Posteriorly, the arm containing the semi-circular canal is compressed and is directed laterally to contact the descending arch of the pterotic. Medially, the exoccipital is curved around to form the border of the lateral occipital foramen. Its dorsal border is sutured to the epioccipital and supraoccipital. Ventrally, it is bordered by the epioccipital and the basi- occipital. The glossopharyngeal foramen is situated in the centre of the bone's lateral face. The epioccipitals (= epiotic; see Patterson, 1975) (EPO, Figs 3, 4 & 32). The lateral face of each bone forms the medial wall of the posttemporal fossa. Dorso-medially, it is suturally united with the supraoccipital and together the bones form a platform posterior to the parietal. Ventrally, the epioccipital contacts the exoccipital and prootic. Dorso-laterally, it joins the pterotic and forms the roof and the upper part of the lateral wall of the subtemporal fossa. The upper jaw (Fig. 7) Each premaxilla (PMX) is in the form of a slender rod with a large anterior ascending process which is ligamentously connected to the kinethmoid. Each premaxilla is curved gently mesad to meet its counterpart, to which it is attached by a thick band of ligamentous tissue. Each maxilla (MAX) is a large triangular bone. Anteriorly, it is produced into a large knob-like process which fits into a concavity on the premaxilla. Ventrally there is a process which runs mesad below the premaxilla to contact its fellow from the opposite side. The dorsal border of the maxilla is expanded into a large plate-like process (termed here the palatine process). Between this and the anterior ascending process runs the ligament joining the maxilla and the palatine. The posterior margin of the bone is concave, ventrally it contacts the premaxilla. The lower jaw (Fig. 8) Each dentary (D) is a long canoe-shaped bone curving gently mesad to join its partner. The coronoid process is deep with a narrow convex border. Ventrally the mandibular lateral line canal bears 20 pores. The anguloarticular (AA) is a large bone sloping gently dorsad to meet the coronoid process of the dentary. The articular surface for the quadrate is almost rectangular. The anguloarticular bears 5 pores of the mandibular canal. The retroarticular (RA) is a very thin wedge of bone lying ventrally. Hyopalatine arch (Fig. 9) The hyomandibula (H) is broad and flat dorsally, the border shaped into two broad articular 14 G. J. HOWES MAX B Fig. 7 Luciobrama macrocephalus, upper jaw. A. Lateral view. B. Ventral view. 5mm D Fig. 8 Luciobrama macrocephalus, lower jaw, lateral view. heads. The anterior of these heads fits into the facet formed by the sphenotic, posterior part of the pterosphenoid and the dorsal margin of the prootic. The posterior head fits into the facet formed by the sphenotic, the prootic and the pterotic. The lower limb, or shaft, is long and broad, directed at an angle of 45° to the vertical. The anterior border of the hyomandibula is concave, the posterior is produced into a small condyle which articulates with the operculum. The lateral face bears a weak depression. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 15 The quadrate (Q) is a low elongate bone which is produced ventro-posteriorly into a triangular extension covered partially by the symplectic and preoperculum. The dorsal margin is horizontal and extends anteriorly to above the anguloarticular. Just posterior to the articulation with the anguloarticular is a large foramen. There appears to be no nerve or vessel of any kind passing through this aperture but only fibres of the connective tissue which line the floor of the branchial cavity. There is a wide synchondrosis between the posterior border of the quadrate and metapterygoid. The symplectic (SY) is a lanceolate bone and lies between the metapterygoid and the quadrate. The lateral surface bears a ridge from which originate some of the fibres of the adductor mandibulae muscle complex (see p. 21). The autopalatine (PAL) is a thick rod-shaped element, forked anteriorly. The lateral fork provides the insertion for the maxillary ligament, the mesial process contacts the cartilage overlying the preethmoid. H PAL 10mm Fig. 9 Luciobrama macrocephalus, hyopalatine arch, lateral view. No dermopalatine is present. The ectopterygoid (ECT) is a thin gutter-like bone applied to the anterior border of the quadrate. Dorsally it contacts the entopterygoid. It is separated from the palatine by a wide area of cartilage. The entopterygoid (ENT) is an extensive sheet of bone, the dorsal border of which is horizontal. Laterally it is overlapped by the quadrate and metapterygoid. The metapterygoid (MET) overlaps the entopterygoid and ventrally partially overlaps the symplectic. The dorsal border is horizontal. The bone slopes mesad to the parasphenoid and a ridge is produced along the lateral face of the bone at the base of the slope. The opercular series (Fig. 10) The preoperculum (PO) is a large, broadly crescentic bone overlapping the anterior edge of the operculum and most of the inter operculum. There are 10 pores of the opercular-mandibular lateral line canal visible along its lower margin with 3 or 4 along the posterior border. The inter operculum (IOP) is shaped like a broad knife blade. Its posterior border is rounded and overlaps the anterior margin of the suboperculum. The suboperculum (SOP) is a narrow curved sheet of bone, its dorsal edge, apart from the posterior point, covered by the operculum. 16 G. J. HOWES The operculum (OP) is extensive. The dorsal margin is long and concave and is produced anteriorly into a long finger-like process to which is attached the dilatator operculi muscle. Mesially a thin strut, pierced ventro-anteriorly by three large foramina, extends caudal from the articular facet. Hyoid arch (Fig. 1 1 A) The interhyal (IH) is a short flat element with a concave posterior border. The head bears a dorsally extended projection from which runs a ligament which passes between the symplectic and the shaft of the hyomandibula to insert on the ventro-posterior tip of the metapterygoid. This ligament is overlain by the connective tissue extending between the symplectic and the hyo- mandibula. PO SOP 10mm Fig. 10 Luciobrama macrocephalus, opercular series, lateral view. Another ligament runs from the anterior edge of the interhyal to join the medial face of the preoperculum. The posterohyal (PHY) is a thick triangular bone bearing one branchiostegal ray. The anterohyal (AHY) is a thin-waisted element articulating anteriorly with the dorso- and ventrohyals. It bears two branchiostegal rays. The dorsohyal (DHY) is small, and is in contact with the first basibranchial (basihyal); the ventrohyal (VHY) is thick, its lower surface bearing a fossa for the ligament connecting the urohyal. Its medial surface is joined to that of its opposite member by a ligament. The urohyal (Fig. 11B) is extremely elongate, the ventral surface is flat, but bears a slight ridge posteriorly. A shallow vertical plate runs along the midline of the dorsal surface. Anteriorly the bone narrows into a neck from which arise two heads ; these are connected to the ventrohyal by thick ligaments. The branchiostegal rays (BSR 1, 2 & 3) are long lamellate bones, the first of which is thin and articulates with the ventral medial surface of the anterohyal ; the second is expanded proximally into an anteriorly directed process. Articulation is on the lateral face of the anterohyal. The third is also expanded proximally and articulates with the lateral face of the posterohyal. The branchial arches (Fig. 12) There are two ossified infrapharyngobranchials (IF 2, 3), recognized as infrapharyngobranchials 2 and 3 (see Nelson, 1969). Cartilaginous elements are present between the first epibranchial and THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 17 the second infrapharyngobranchial and between the fourth epibranchial and third infra pharyngo- branchial. These elements most probably represent the 1st and 4th infrapharyngobranchials. The epibranchials (EPI 1-4) number four. The first is wide with a blade-like posterior border overlapping the second. The third epibranchial bears a dorsal process overlapping the fourth. The certaobranchials number the usual five (C 1-5). The first bears 6 long finger-like gill rakers ; the second 9; the third 10 plus 6 along the medial surface; the fourth 7 plus 7. The fifth is the lower pharyngeal bone which bears a single row of 4-5 curved teeth. The bone is very thin and anteriorly elongate, curving gently mesad to ligamentously join its fellow. The pharyngeal bone and teeth have been described and figure by Chu (1935). IH DHY '—3 AHY VHY BSR1 B 10mm Fig. 11 Luciobrama macrocephalus. A. Hyoid arch, lateral view. B. Urohyal seen in (top) lateral and (bottom) dorsal views. There are four basibranchials (BB 1-4). The first (basihyal) is a long thin rod in contact with the hypohyals. The second is flat and in contact with the 1st hypobranchials (HB1); the third is long, narrow-waisted and in contact with the 2nd hypobranchials; the fourth is thin and is curved ventrad. Weberian apparatus and swimbladder (Fig. 13). The 1st vertebra is a thin disc bearing lateral processes (LP1) from which a ligament extends to contact the medial face of the cleithra. The second and third centra are fused dorsally, but ventrally they are clearly separated. The 2nd vertebra bears thick lateral processes which are inclined upward (LP2). The 3rd vertebra contains a lateral fossa for the articulation of the tripus. Extending dorsad from the second and third centra is the third neural plate (NP3). The dorsal border is triangular. On the anterior margin lies the second neural plate (NP2) which extends forward to contact the supraoccipital. Lying below the anterolateral border of the second neural plate is the claustrum (CLA) which is a cartilaginous cup-shaped structure. A ligament runs from each claustrum to insert upon the basioccipital. 18 G. J. HOWES Along the posterior margin of the third neural plate there rests part of the large neural complex (NC); see below. The 4th vertebra bears stout lateral processes which become ventrally directed (pleural ribs, PR4). The medial surface of each extends inwards as an os suspensorium (OSS), a thin plate curving anteriorly so that its tip underlies the posterior edge of the 3rd vertebra. There is a short neural spine on the 4th vertebra, the dorso-anterior surface of the spine supports the posterior half of the neural complex. BB1 IF3 EPI1 10mm A B — 5 Fig. 12 Luciobrama macrocephalus, branchial arches. A. Upper arch, left side, dorsal view. B. Lower arch, right side and basibranchials, dorsal view. The scaphium (SCA) is rounded and capped dorsally by the claustrum. Ventrally it articulates with a groove along the top of the first centrum and from a depression on its posterior face a liga- ment runs to connect this bone with the intercalarium. The intercalarium (INC) articulates in a fossa situated below the third neural plate which overlaps the second vertebra. A ligament (a continuation of that extending from the scaphium) attaches ventrally and continues to insert upon the leading edge of the tripus. The tripus (TR) is a thin triangular plate articulating with the third vertebra. Anteriorly it contacts the lateral process of the second vertebra (LP2); posteriorly its tip connects with the medial face of the process emanating from the 4th vertebra (PR4). THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 19 The neural complex (NC) which lies across the 3rd and 4th vertebrae is an irregularly shaped bone. The dorsal surface is grooved and posteriorly a supraneural plate (possibly two fused supraneurals) slides into the groove. This plate is connected to the supraoccipital by a ligamentous sheet (Isa) which runs across the anterior part of the grooved upper margin of the neural complex. The anterior border of the neural complex is concave and a separate ligamentous sheet (Isb) connects this to the supraoccipital. The swimbladder (sb) is an elongate cigar-shaped structure extending posteriorly to above the first anal fin ray. It is bipartite. The pneumatic duct is very long and runs along the dorsal surface of the gut from its exit at the anterior of the alimentary canal to its entry into the posterior division of the swimbladder. 10mm Fig. 13 Luciobrama macrocephalus, Weberian apparatus, lateral view. Pectoral girdle (Fig. 14A) The cleithrum (CL). The horizontal limb of the cleithrum is narrow and bifurcated anteriorly. The tip of the limb lies on a perpendicular with the posterior margin of the prootic. The ascending limb has a slightly curved hind margin ; it is aligned almost vertically. The postcleithrum (PC) is a short spine-like process. The supracleithrum (SCL) is a small blunted crescentic element attached to the upper third of the cleithral limb. It attaches to the inner face of the posttemporal. 20 G. J. HOWES The coracoid (COR) is a narrow, flat bone posteriorly joined to the cleithrum along a flat lateral extension; anteriorly, there is a narrow area of attachment along the leading edge of the cleithrum. The coracoids diverge from each other and meet only along the anterior margin. The mesocoracoid (MC) is a wide bridge of bone extending between the cleithrum and the coracoid. The scapula (SCP) lies against the medial face of the cleithrum below the mesocoracoid. It joins the posterior border of the coracoid and provides an articular surface for the four plate-like proximal radials. SCL SCP COR B 10mm Fig. 14 Luciobrama macrocephalus. A. Pectoral girdle, lateral view. B. Pelvic girdle, left side, dorsal view. Pelvic girdle (Fig. 14B) The pelvic bones are deeply forked, the thin dagger-like arms narrowly diverging. The ischiac process is wide and is joined to its fellow along the apposed medial face. Vertebral column There is a total of 55 vertebrae (including the four forming the Weberian apparatus), comprising THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 21 30 precaudal (without a haemal spine), 20 caudal and the fused preural and first ural vertebrae. All the centra are of almost the same length. There are 12 supraneurals lying between the neural spines of the 5th to 18th vertebrae. The first two supraneurals are expanded and contact the neural complex of the Weberian apparatus (see p. 18). The supraneurals become progressively thinner posteriorly; on a radiograph they are barely visible. The cranial intermuscular bones (CIM, Figs 3 & 4) extend from the medial margin of the parietals and epioccipitals as thin rods which become lamellate and expand into 7 or 8 bones to lie above the Weberian apparatus. The first of the epipleurals is associated with the 15th rib. They are exceedingly thin needle-like bones numbering 11-12. There are numerous intermuscular bones above the anal fin. The first pterygiophore of the dorsal fin is expanded anteriorly. There are 9 dorsal and 1 1 anal pterygiophores. Two radials connect the pterygiophore with the fin ray (see p. 52). Caudal fin skeleton (Fig. 15) There are six hypurals (HY 1-6) of which the first is greatly expanded. The fused preural and ural centra (PU1 + Ul) bear a knife-like neural spine. There is one large epural and a pair of small uroneurals (UN) above hypural 6. The parhypural (PH) bears only a slight hypurapophysis. The principal fin ray formula is 19 + 91. Cranial myology Jaw and suspensorial muscles (Figs 16-19) The postorbital region of Luciobrama is covered by thin skin, when this is removed there is exposed a large adductor mandibulae muscle. Two major divisions of this muscle can be distinguished, namely Al and A2. Adductor mandibulae Al extends from the quadrate, symplectic and preoperculum. The antero- ventral fibres run dorsad at an angle of 30°, those forming the dorsal border of the muscle run almost horizontally. Below the orbit the muscle is greatly thickened but becomes abruptly compressed prior to its insertion. The fibres insert upon a thick tendinous band which runs along the ventral border of the maxillary and is attached to that bone by connective tissue. The maxilla has been described elsewhere (p. 13). A cartilaginous mass (car) fills the area bordered anteriorly by the concave dorsal edge of the maxilla, dorsally by the lateral ethmoid and medially by the dentary. The large A2 extends from the lateral face of the preoperculum, the hyomandibula and the metapterygoid. The fibres running from the preoperculum and hyomandibula are orientated horizontally and form the lateral face of the muscle ; those running from the metapterygoid are directed laterally at an angle of 40° to join the body of horizontal fibres. Anteriorly A2 is divided, each division inserting upon its own tendon. The tendon of the lateral division inserts upon the rim of the coronoid process of the dentary, that of the inner division on to the rim of the angulo- articular, just posterior to the outer tendon. Adductor mandibulae Aw is absent. The medial face of the lower jaw is covered by a thick connective tissue which forms a cushion along the dorsal edge of the jaw (the lower lip), and at the articulation of the jaw is continuous with that tissue and skin covering the upper jaw. The levator arcus palatini (lap) is an exceptionally well-developed muscle and to my knowledge is the most extensive described for any teleost although that of Arapaima gigas approaches this size (see Kershaw, 1976). It originates from the ventral surface of the frontal, the pterotic and the sphenotic to insert upon the length of the entopterygoid, metapterygoid and on a sheet of thick connective tissue connecting the metapterygoid with the hyomandibula (ct, Fig. 17). The ventral surface of the levator is bevelled to accommodate adductor mandibulae A2. The anterior border of the muscle forms the posterior border of the orbit. When the outer layer of the muscle is removed (lap 1), two inner sections are revealed. The first (lap 2, Fig. 17) lies posteriorly and runs from a dorsal aponeurosis from which the dilatator 22 G. J. HOWES PUH-U1 HY1 10mm Fig. 15 Luciobrama macrocephalus, caudal fin skeleton, lateral view. operculi also stems. The direction of its fibres is the same as that of the outer layer. Insertion is upon the lateral face of the hyomandibula. The second element (ah, Fig. 17) lies against the pterosphenoid and parasphenoid. The fibres are orientated in the opposite direction to those of the other layers. Anteriorly, the muscle is bordered by a sheet of connective tissue which covers the lateral face of the pterosphenoid; posteriorly, it originates from the deep subtemporal fossa and inserts upon the medial face of the hyomandibula. Ventral insertion of all the sections is along the medial dorsal edge of the ento- and metapterygoid. The inner element is well differentiated from the rest of the levator arcus palatini and I interpret it as being the adductor hyomandibulae (which is also found in Aspius and some other genera; see p. 53). THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 23 0.0 "o 1.3 *— c 1-1 .^ QJ ^ a -5 LU -g 3 -O |S I < s Is II u a 1JS IS a -a ~ — -o o •2 -° 24 G. J. HOWES The adductor arcus palatini is absent, but it would appear that the anterior extension of the adductor hyomandibulae is in fact fulfilling the function of the adductor arcus palatini. The dilatator operculi (do) originates anteriorly from the aponeurosis which also gives rise to the inner section of the levator arcus palatini, and posteriorly from the lateral border of the pterotic. Some fibres also stem from the sphenotic process. Insertion of the fibres is into a long tendon which forms the ventral border of the muscle and which joins the anterior process of the operculum. A thick band of tissue connects the lateral face of the opercular process with the pterotic. The levator operculi (lo, Fig. 16) is a flat sheet of muscle running from the pterotic to the medial face of the operculum. The fibres run almost perpendicularly. The adductor operculi is a thin conical muscle originating from the deep subtemporal fossa to insert upon the medial leading edge of the operculum anterior to the insertion of the levator. Hyoid muscles (Fig. 18) The intermandibularis (im) is very thin and ellipsoidal in cross-section. It is covered dorsally by skin and connective tissue, ventrally by the protractor hyoideus. ' im Ph IH Fig. 18 Luciobrama macrocephalus, hyoid musculature. The hyal bones are those of the left side shown in lateral view. The first branchiostegal ray and ventral borders of other hyal bones are indicated by dashed lines. The dentary is of the right side and is shown in medial view. The protractor hyoidei (ph) arise from the medial and lateral surfaces of the anterohyal and posterohyal. Those parts stemming from the internal surfaces of the bones overlap the dorsal edge of the anterohyal to pass laterally into the muscle body. The lateral origin is as far back as the interhyal. Small bundles of fibres also originate from the second and third branchiostegal rays. The two halves of the muscle join together and run as an elongate cone between the dentaries. Insertion is posterior to, and below, the intermandibularis. The hyohyoidei (hh) are weakly developed. They lie as sheets of fibres between the branchiostegal rays. From the first branchiostegal ray the fibres run into tendinous bands which meet along a raphe below the dorso- and ventrohyals. Insertion is from the third branchiostegal ray onto the suboperculum. It is not possible to distinguish abductores and adductores sections of this muscle and it would appear as Winterbottom (1974) noted in Cyprinus that the function of the adductores is taken over, in this case to a great extent, by the protractor hyoidei. The sternohyoideus (sth, Fig. 19) originates from the forked leading edge of the cleithrum, the dorsal arm of the fork contributing a separate bundle of fibres which is directed ventrally into the main mass of horizontally arranged fibres. The lateral border of the muscle is marked by tendinous bands. Insertion is along the ventral and lateral faces of the urohyal. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (tACEPEDE) 25 Branchial arch muscles (Fig. 19) I have not made a thorough investigation of the branchial arch muscles due to lack of material for deep dissection. As far as I can see, the arrangement of this musculature is essentially that described for Opsariichthys by Takahasi (1925). The obliqui ventrales (obvl-3) are present on the first three ceratobranchials, they are very elongate well-developed muscles. The fourth ceratobranchial bears a transversus muscle which meets its fellow along a median raphe into which inserts the pharyngoclavicularis interni. obv3 obv2 rv3 obvl pci sth pee Fig. 19 Luciobrama macrocephalus, ventral branchial muscles of the right side shown in ventro-lateral view. Connecting the second and third hypobranchials is a thin almost tendinous muscle, the recti ventrales (rvl); another larger element runs from the third to the fourth hypobranchial (rv2), a thin tendon then connecting the fourth hypobranchial to the recti ventrales of the fifth (pharyngeal) ceratobranchial. These muscles are the arcualis-hyoideus of Takahasi (1925). Winterbottom (1974) refers to these as the recti ventrales and remarks that there are four to five in the cyprinids. Comparative analysis In order to determine the interrelationships of Luciobrama, it has been necessary to examine a wide range of cyprinid genera and to review a series of anatomical features to decide if they are derived or primitive characters. The species that have been examined are listed on pages 2-5. These were chosen to represent those groups currently recognized as subfamilies (see p. 61). In referring to large genera such as Barilius, Barbus and Labeo it should be made clear that in the context of this paper such reference 26 G. J. HOWES is only to those species examined and does not imply that any particular feature occurs in all congeners. Although many genera have been examined, not all are cited in the following analysis. An initial study suggested those that could possibly be related to Luciobrama, those that displayed parallel features and those exhibiting marked differences. Examples of genera in all three categories have been used in this analysis. IO5 B Fig. 20 Infraorbital bones of A. Aspiolucius esocinus, B. Pseudaspiiis leptocephalus, C. Aspius vorax. Osteological characters Circumorbital series In Luciobrama all the infraorbitals, apart from the lachrymal, are reduced to a slender ossification around the sensory canal tube. This is not an unusual condition in cyprinids and is found in many genera (e.g. Alburnus, Labeo, Barbus, Notropis, Chrosomus). However, in all species of three genera, namely Aspiolucius, Pseudaspius and Aspius, the posterior infraorbitals are of similar morphology and arrangement to those in Luciobrama (cf. Fig. 2, Figs 20A, B & C). In these three genera the first infraorbital is large, and the canal bears 9-10 pores in Aspiolucius, 11 in Pseudaspius and 9-10 in Aspius. The second infraorbital is short in Aspiolucius, but in Pseudaspius and Aspius it approaches the proportions of that bone in Luciobrama. The third and fourth infraorbitals are expanded, the fourth being shield-shaped. The canal carried by the fourth THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 27 infraorbital is diverted across the postorbital region as in Luciobrama. In all these species the fifth infraorbital is minute and reduced to an ossification around the canal tube. In Elopichthys (Fig. 21 A), although the reduction of ossification is similar to that found in the above-cited genera, the fourth infraorbital is orientated vertically, and the fifth curved dorso- posteriorly to join the pterotic canal. This arrangement is found in a number of genera (e.g. Leuciscus, Alburnus, Culler, Pelecus, Paralaubuca, Oxygaster and Ochetobius; Fig. 2 ID). However, in Paralaubuca and Oxygaster the fourth infraorbital is expanded. A different situation is found in Opsariichthys, Zacco, Barilius and some other genera (see below, p. 29). Here all elements in the infraorbital series are expanded. In Barilius bola the second, B Fig. 21 Infraorbital bones of A. Elopichthys bambusa, B. Erythroculter mongolicus, C. Schizothorax esocinus, D. Pelecus cult rat us. 28 G. J. HOWES third and fourth bones cover the entire cheek. In Opsariichthys the fourth and fifth infraorbitals are expanded posteriorly. Expansion of the second and third bones occurs in some Rasbora (Ramaswami, 1955b). The fifth infraorbital is reduced to an ossification around the canal tube in Opsariichthys and is remote from the supraorbital, but in Barilius it is large and connected to the supraorbital. A similar arrangement is found in Salmostoma, Luciosoma and Squaliobarbus (see below, p. 29). The supraorbital is variously developed in cyprinids (see Ramaswami, 1955b : 208). In Aspiolucius, Pseudaspius and Aspius (Fig. 20) it is, as in Luciobrama, relatively narrow, the frontal widening posteriorly to it and preventing its contacting the infraorbital series. In narrow headed cyprinids, such as Oxygaster, Pseudolaubuca and Macrochirichthys, the bone is narrow and extends for almost the length of the lateral margin of the frontal but fails to make contact with the fifth infraorbital. In most Barbus species the supraorbital is small and well separated from the infraorbital series by the frontal, but in Barbus tor the bone is very long and meets the fifth infraorbital. Gosline (1974 : 3) also noted the variability of contact in certain south-east Asian species of Barbus. Fig. 22 Infraorbital bones of Oreoleuciscus pewslowi. Comments on circumorbitals Roberts (1973) states that in cyprinids there is an antorbital and six infraorbitals. I count only six including the first, here identified as the lachrymal (following Harrington, 1955). Gosline (1961) considered the first infraorbital in cyprinids as possibly a compound structure comprising the lachrymal and antorbital. It is not clear if Roberts considered the first infraorbital to represent two fused elements. Harrington (1955) identified the sixth infraorbital in Notropis bifrenatus as the dermosphenotic, noting that it was reduced to a bony tube and was often lacking in that species. Gosline (1975) discussed the dermosphenotic of cyprinids and thought the degree of develop- ment of this bone and its proximity to other circumorbitals could be an aid in assessing the primitive or specialized condition of a particular taxon. However, he considered that in Aspius there was a certain degree of difficulty in the identification of this bone. Gosline (1975 : 2; pi. 2, fig. 3) described a membranous tube connecting the fifth infraorbital with the cranium. In fact his fifth infraorbital is the fourth in the series and the 'membranous tube' is the fifth. In a series of specimens of Aspius vorax I have found quite some variation in the development of this bone. In some specimens there is a single completely ossified canal connecting the fourth infraorbital with the pterotic canal, in others the bony tube is fragmented into two components, the upper THE ANATOMY OF LUCIOBKAMA MACROCEPHALUS (LACEPEDE) 29 one of which probably represents the true dermosphenotic (sixth infraorbital) and overlies the autosphenotic. In a specimen of 1 10 mm SL on one side of the head there is no sign of an ossified element between the fourth infraorbital and the pterotic, although the canal is present as an epidermal tube. Yet, on the other side of the head the fifth infraorbital is well ossified. In Oreoleuciscus pewslowi (Fig. 22) it is interesting to note that there is marked fragmentation of the posterior infraorbital. In one specimen the fourth infraorbital is fragmented into three separate lamellate elements and the sixth infraorbital (the dermosphenotic) is also present (see Jollie, 1975, on the fragmentation of these bones.). In Luciobrama connection between the fifth infraorbital and the dermopterotic is made through an epidermal canal. This connection occurs far in front of the autosphenotic because of the forward extension of the overlying dermopterotic. Thus, there is no infraorbital bone connected with any part of the autosphenotic and which could be interpreted as a dermosphenotic. In Esomus, Ramaswami (1955b) pointed out that the sphenotic occurred as a roofing bone. The bone he was referring to is, in fact, the dermosphenotic and in preparations to hand I have been able to separate this canal bearing bone from the underlying autosphenotic. Greenwood et al. (1966) refer to this feature in Esomus as being specialized, but it is probably a primitive condition for a cyprinid. Gosline (1975) noted that in Salmostoma the dermosphenotic was large and contacted the supraorbital (which it also does in Barilius, Cyprinus, Squaliobarbus, Luciosoma and some other genera). He was of the opinion that contact between the dermosphenotic and supraorbital was a primitive character (Gosline, 1975 : 6) because such contact is found in the 'generalized' characoid Brycon. It is so that in Brycon and other characoids the sixth infraorbital (dermosphenotic) is well developed and makes contact with the supraorbital (Weitzman, 1962; Roberts, 1969). However, this has little bearing on the situation in cyprinids. Indeed, if Opsariichthys is to be considered the 'primitive' cyprinid this argument fails because no such contact is found between the infra- orbitals and supraorbital. In Salmostoma and other cyprinid genera in which such contact occurs, it is between the fifth infraorbital and supraorbital. No cyprinid I have examined shows any evidence of the interposition of a sixth infraorbital. This would suggest that either the dermo- sphenotic has been lost altogether in these genera or else it has become incorporated with the fifth infraorbital. Incorporation into another dermal head bone is suggested by an observation on Chelaethiops. A specimen of Chelaethiops sp. (29-5 mm SL) was found to possess a well-developed dermosphenotic, but in four other specimens of the same series (alizarin preparations) it was absent. However, in a fifth specimen (24-5 mm SL) a fragment of the dermosphenotic was visible, apposed to the posterior edge of the frontal (Figs 23A & B). The dermosphenotic is also well developed in Esomus danricus (see above), and the fifth infra- orbital, although small, maintains contact with the supraorbital; that area lying postero-dorsally to the infraorbital (i.e. above the dilatator operculi muscle) is covered by a 'normal' body scale, one bearing concentric radii. This scale appears to be in no way associated with any cranial bone. The area covered by the fifth infraorbital is that which houses the dilatator operculi and levator arcus palatini muscles, and it seems likely that reduction of dermal bones in that region would be a necessary preadaptation or a response to the reorientation and expansion of the underlying musculature. It may be noted here that in some Barilius species where the fifth infraorbital is large, the dilatator operculi is covered by the adductor mandibulae muscle (see p. 55). Tretiakov (1946) placed much emphasis in classifying the cyprinids on the development of the infraorbital series and suggested that those cyprinid genera with the broadest posterior bones (presumably in contact with the supraorbital) were the most primitive. He included Cyprinus in that category. Gosline (1974) considered the cephalic canals of cyprinids and divided the old world genera into two groups on the basis of 'presence or absence of a gap between the supraorbital and infraorbital canals'. He stated (loc. cit. : 1 1) that all south-east Asian and African cyprinids have the supraorbital and infraorbital canals connected. The connection between the infraorbital and supraorbital canals is dependent on the form of the last infraorbital (or dermosphenotic) already discussed above. There is always a connection 30 G. J. HOWES between the two canal systems, be it through an ossified or an unossified tube. The 'gap' observed by Gosline is presumably the unossified condition. Such an unossified connection is found in some African Barbus (e.g. B. somereni), a group which Gosline included amongst those genera with a connection between the canal systems; whereas Rutilus and Pelecus, included in the group with a break between the canal systems, have a complete connection with the pterotic canal. Ethmo-vomerine region The kinethmoid. I have been unable to determine the condition of this bone in Aspiolucius and Pseudaspius owing to lack of material for dissection. In Aspius it is short and blunt as in Luciobrama. DSP IO5 DPTE APTE 1mm Fig. 23 Upper circumorbital bones of Chelaethiops sp. A. A specimen of 29-5 mm SL, B. A specimen of 24-5 mm SL. In Elopichthys the dorsal surface of the kinethmoid is greatly expanded into a flat diamond- shaped plate separating the premaxillae posteriorly (Fig. 35, KE). It is attached by ligaments to the premaxillae and maxillae and rotates against the ethmoid when the jaw is opened. In Opsariichthys, Zacco and some species of Barilius the kinethmoid is an elongate gutter- shaped bone the dorsal edges of which are slightly flared. However, in Barilius bola the dorsal surface is expanded as a solid plate similarly to that of Elopichthys although not as extensively (see p. 43). In other species of Barilius the kinethmoid is a rod-shaped structure which is notched dorsally. This is the 'usual' condition of the kinethmoid in other cyprinid genera and is probably the primitive one. Exceptionally, in the genus Barbus, the species B. tor and B. mariae possess very long kinethmoids, the posterior borders of which are convex, enabling the bone to rock in the anterior ethmoid groove. Current research on the genus Macrochirichthys has indicated an unusual feature of the kinethmoid (which is a triangular bone) whereby it articulates directly with the premaxillae. THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) 31 I could only examine the features of the ethmo-vomerine region in Aspiolucius and Pseudaspius from radiographs. Both genera resemble Luciobrama in possessing a similar depression, elongation and contact of the lateral ethmoids with the parasphenoid. Ramaswami (1955b) noted that in many genera including Barbus, Opsariichthys and Aspius the preethmoid articulated with the ethmoid only. However, in all the cyprinid genera I have examined the preethmoid is supported in a lateral fossa formed by the ethmoid and the vomer (the condition noted in Labeo by Starks, 1926 : 174). Ramaswami (1955b) also states that the preethmoid is completely lacking in, amongst other genera, Esomus, Leuciscus and Notropis. I can report that it is present in both the former genera and Harrington (1955) notes that it is present in Notropis but is supported only by the vomer. The size of the preethmoid varies considerably, being a very large laterally protruding structure in Opsariichthys to minute, partially ossified pads in Chela and Esomus. The supraethmoid is variously developed in cyprinids. One condition appears for it to be broad and short with the anterior border notched medially; this is the type found in Opsariichthys, and some Barilius species. Although basically similar, the supraethmoid in the cultrines is narrow and the medial notch much deeper. In Barbus and Labeo the bone can become extensively developed (e.g. Labeo cylindricus) and the anterior border produced medially. In Barbus tor and Barbus mariae there is, however, a very deep medial groove which accepts the expanded kinethmoid; see above page 30. The usual condition is for the supraethmoid to interdigitate posteriorly with the frontals, the posterior border is mostly straight or somewhat irregular but without the long lateral forks found in Luciobrama and Aspius. However, in Macrochirichthys, Chela and some other genera the supraethmoid is overlain by the frontals. Work currently in progress suggests that this is a derived feature associated with the oblique orientation of the jaws and that it is indicative of close relationship of those genera in which it occurs. Variability within the ethmoid appears to be mainly one of depth; being very deep in some genera such as Pelecus and shallow in others such as Chela. In all the cyprinids I have examined the ethmoid makes some contribution to the preethmoid fossa. The vomer is usually short and wide as it is in Opsariichthys but in this genus and in some Barilius it is greatly thickened anteriorly. An extreme of this condition is found in Elopichthys where the ventral surface is swollen and posteriorly folds over to contact the parasphenoid (Fig. 24). The vomer in the majority of cyprinids is thin and the ventral surface is either flat or bears a shallow groove. The nasals in Aspiolucius are long decurved bones containing 8 pores (Fig. 25) ; in Pseudaspius they are shorter, bearing 5 pores, and in Aspius, long with 6 pores (Fig. 26). Elopichthys resembles Luciobrama in possessing long, narrow nasals bearing 9 pores. In the majority of cyprinid genera studied the nasals are found to be short bones with 2-4 pores. In some cultrines they may be long as in Erythroculter mongolicus where the nasal bears a lateral flange and has 6 pores. The frontals in Pseudaspius, but even more in Aspiolucius, are narrow, elongate and anteriorly are slightly separated from one another (Fig. 25). The lateral border posterior to the orbit, like that in Luciobrama, is markedly concave. No other cyprinid genus I have encountered has such elongate frontals as are present in Aspiolucius and Luciobrama. In most genera examined the frontals are relatively short and broad but in some Barilius species are narrow and elongate with a concave border above the orbit. The dorsal surface of the frontals is mostly flat or convex but in Macro- chirichthys, Pseudoxygaster, Pelecus and some species of Oxygaster the frontals are medially depressed to allow for the cranial extension of the epaxial musculature (see Howes, 1976), and in Nematabramis there are transverse lamellate ridges across the frontal surfaces. Orbital region Each orbitosphenoid in Aspiolucius and Pseudaspius appears similar to that in Luciobrama; for some distance it is joined to the parasphenoid but lacks the dorsal posterior extensions seen in Luciobrama. (These observations were made entirely from radiographs.) In Aspius the orbitosphenoids are short, deep and widely divergent (Fig. 27B). Medially, they fuse to form an interorbital septum which extends ventrally to join the parasphenoid. A similar development is found in Elopichthys but here the septum is reduced (Fig. 27 A). 32 G. J. HOWES The orbitosphenoid septum is variable in its development within the Cyprinidae. Its purpose is to provide wide separation between the cranial roof and the parasphenoid. Such separation appears to have little to do with the size of the orbit but more with the angle at which the cranium is aligned to the vertebral column and the size of the buccal cavity. In piscivorous cyprinids the orbitosphenoids make direct contact with the parasphenoid without the intervention of a septum. This is also the case in those genera such as Labeo and Garra which are characterized by their depressed crania. In both cases this close union has resulted in increased rigidity of the cranium (in Labeo cylindricus the orbitosphenoids extend lateral wings which join similar processes from the parasphenoid), and increased area of the buccal cavity. When the orbitosphenoids are reduced in depth they often exhibit a cancellous surface and bear lateral ridges (e.g. Labeo, Schizothorax, Barilius). PS PAL Fig. 24 Elopichthys bambusa, ethmo-vomerine region, ventral view. The pterosphenoid in Aspiolucius and Pseudaspius is an extensive bone and is depressed as in Luciobrama, but in Aspius it is less extended antero-posteriorly, and that part of the bone forming the wall of the hyomandibular facet less well developed (Fig. 27B). The pterosphenoids of Aspius diverge widely and their borders are close to the lateral margins of the frontals. In Elopichthys (Fig 27A, 28 & 29), the pterosphenoids present a condition not encountered in any other cyprinid. The bones are extensively developed and diverge to reach the lateral margins of the frontals. Unlike the genera mentioned above, the sphenotic is not continuous with the frontal and these two bones are separated by the intervention of the pterosphenoid, whose surface at this point is depressed to form a basin. The pterosphenoid basin provides the site of origin for the adductor mandibulae A3 muscle (see p. 53). As in Luciobrama there is an extensive connection with the parasphenoid. The pterosphenoids of other cyprinid genera examined are generally small, almost hexagonal in outline and make contact with both the prootic and parasphenoid. However, in Zacco, Cyprinus, Catla, Rutilus and some species of Barilius, the pterosphenoid does not contact any THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) 33 10 mm N SOR F SP DPTE Fig. 25 AspioluciUs esocinus, part of the dorsal surface of the cranium. (Holotype.) SP IT SOR Fig. 26 Aspius vorax, neurocranium, dorsal view. part of the parasphenoid but instead is sutured entirely to the prootic. (See later note on prootic, p. 35.) The parasphenoid in Aspiolucius and Pseudaspius is thick and horizontally aligned, contacting the medial sections of the lateral ethmoids and the orbitosphenoids as in Luciobrama. In Opsariichthys the parasphenoid is thin, flared anteriorly, and bears a deep groove on its ventral surface. The lateral ascending wing of the parasphenoid which contacts the prootic and ptero- sphenoid is wide. Below the prootic the parasphenoid broadens into a triangular platform which 34 G. J. HOWES is extended laterally by the contribution of the flattened ventral surfaces of the prootics. A similar contribution from the prootics to the parasphenoid platform is found in most Barilius species (Fig. 30). The lateral ascending wings of the parasphenoid in Zacco, some Barilius species, and Leuciscus are narrow, and as mentioned above, make contact only with the prootic and not the pterosphenoid. Although Ramaswami (1955b) stated that the parasphenoid did not show any variation, in fact it does. In some genera the anterior part is very wide (e.g. some Labeo species) and there are present in others well-developed medial dorsal and ventral ridges, Again, in Labeo the ascending Fig. 27 Ventral views of the orbital regions of A. Elopichthys bambusa, B. Aspius vorax. PTS APTE PTS PRO Fig. 28 Elopichthys bambusa. Lateral view of the orbital-otic region of the neurocranium. (Composite from three specimens.) THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) 35 wing may be greatly extended laterally and there is sometimes developed a medial strut of the ascending wing which contacts the pterosphenoid. Catla and Hypophthalmichthys have the posterior part of the parasphenoid bent upwards, and the midline below the ascending wings bears a strong ventral process. Otic region The prootic in Aspiolucius, Pseudaspius, Aspius and Elopichthys is of a similar elongate shape to that in Luciobrama and the anterior foramen of the trigemino-facialis chamber is likewise situated on the anterior border of the bone which contacts the pterosphenoid. Also, as in Luciobrama the lateral commissure is wide. The large hypophysial foramen (Ramaswami, 1955b) between the parasphenoid and prootic in Luciobrama is absent in the above-mentioned genera. In most cyprinid genera examined, the anterior foramen of the trigemino-facialis chamber interrupts the anterior border of the prootic, which is bounded by the pterosphenoid. However, in some species of Barilius the prootic bears the anterior opening of the trigemino-facialis chamber entirely in its lateral face, remote from the anterior border and in those species the lateral commissure is a narrow structure. There is much variation in the width of the lateral commissure throughout the cyprinids. Besides Luciobrama and the genera mentioned above, it is a wide structure in Schizothorax, Erythroculter and some Barbus species (all those taxa in fact, which have an elongation of the postorbital cranium). On the other hand, it is reduced to a narrow strut in Opsartichthys, Zacco, Leuciscus and Labeo. The floor of the prootic forms the roof of the posterior myodome in all the genera examined. Again, the extent and depth of the myodome show great variation. As far as can be ascertained from radiographs, the myodome in Aspiolucius and Pseudaspius resembles that of Luciobrama which in turn bears similarity to that described in Aspius by Oliva and Skofepa (1968). The topographic relationship between the prootic, parasphenoid and pterosphenoid in the cyprinids appears to have some significance in establishing phylogenetic relationships between various taxa. A particular study is being made of these bones in connection with current work on the genera Opsariichthys, Zacco and Barilius. The autosphenotic is extensive in Aspiolucius and extends laterally from below the border of the cranium as a long shelf on which the dilatator operculi muscle originates (Fig. 25). In Aspius (Fig. 26) the sphenotic is not roofed by any part of the frontal or pterotic. Together with part of the frontal it extends laterally and forms the fossa for the dilatator operculi muscle. The sphenotic in Elopichthys is bordered anteriorly by the basin-like pterosphenoid (see above, p. 32), and it extends laterally as a wide platform, the posterior ventral surface of which forms the anterior hyomandibula facet (Figs 27A, 28 & 29). In Opsariichthys, Zacco and some Barilius the bone is overlapped along its medial margin by the frontal and forms a deep dilatator fossa. The anterior lateral process of the sphenotic in these genera is short and lamellate, in contrast to that of other species of Barilius (e.g. bola, loati) where the lateral process is long and thick. In Barilius microcephalus the posterior dorsal border of the sphenotic is separated from the overlying pterotic to form a lateral foramen (LF, Fig. 31). Part of the adductor mandibulae A2 muscle originates from the ventral surface of the pterotic and passes through this foramen. Some genera display a condition in which the laterally directed process of the sphenotic is separated from the overlying frontal, contact between the two bones being along their lateral margins. Thus, a foramen is formed which provides a passage for the dilatator operculi muscle which originates on the ventral surface of the frontal. This feature is found in Esomus, Cyprinus and Catla. It also occurs in some Barbus species and appears to be present in all European and north African species examined (Barbus barbus, B. callensis, B. nasus and B. reinii), a middle eastern species (B. canis), some eastern and southern African species (B. altianalis, B. intermedius, B. oxyrhynchus, B. rocadasi, B. progenys, B. natalensis and B. capensis) and in some Asian species (B. altus and B. tor). However, it is absent in all the other species of Barbus examined (see list of species on p. 2) where the dilatator fossa is of the 'usual' type (see p. 56). In all the African and Asian Labeo species examined the sphenotic process is separated from the frontal in the same way as it is in Barbus and the other genera cited above. However, the sphenotic 36 G. J. HOWES PIT PTS SP ES LE N SOR SP APTE 10mm Fig. 29 Elopichthys bambusa, neurocranium, dorsal view. The position of infraorbitals 4 and 5 are indicated. BO Fig. 30 Barilius bola. Ventral view of the parasphenoid and prootic. The parasphenoid platform is unshaded. THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) 37 process, which in Labeo is usually reduced to a thin strut of bone, is also perforated by a foramen. These two openings allow for passage of a divided dilatator operculi muscle (see p. 57). A sub- sidiary foramen is also present in Catla and here too the dilatator muscle is divided through both apertures. In Squaliobarbus the lateral process of the sphenotic is a wide platform covered for half its width by the frontal. The ventral surface of the bone provides a fossa for the articulation of the anterior condyle of the hyomandibula. Anteriorly, the sphenotic is deeply recessed, leaving only the thinnest wall between the orbital cavity and the dilatator fossa, perhaps an incipient condition for the development of this feature. so APTE 5mm STF Fig. 31 Barilius microcephalus. Dorso-posterior section of neurocranium, lateral view. Further discussion of the dilatator fossa is reserved until page 56. The pterotics in Aspiolucius and Elopichthys (Figs 25 & 28) are wide bones contributing to the cranial surface and bordering the parietals and frontals as in Luciobrama. Also, as in that species, they overlie and extend anteriorly beyond the sphenotic shelf. In Aspius there is only a narrow region of contact between the pterotics and the frontals. The fossa for the posterior condyle of the hyomandibular bone is almost entirely confined to the pterotics. The cranial surface of the pterotics is narrow in Opsariichthys and most Barilius species. Medially the bones form a flat roof to the subtemporal fossa. Ventro-posteriorly their connection with the exoccipitals, through which passes the semi-circular canal, is greatly compressed. The posterior spine of each pterotic is short and is directed ventrally at an acute angle. The posterior hyomandibular fossa lies mainly in the pterotic, but the anterior third extends onto the sphenotic. In the long-jawed species of Barilius (e.g. B. bola, B. loati) this fossa is confined almost entirely to the pterotic. 38 G. J. HOWES The posttemporal fossae in Luciobrama are well developed (see p. 1 1 & Fig. 32A). As far as I can ascertain from radiographs such is also the case in Aspiolucius and Pseudaspius. Certainly in Aspius (Fig. 32B), Elopichthys (Fig. 33A), Megalobrama, Culler, Erythroculter and Schizothorax deep posttemporal fossae are developed. In other genera such as Opsariichthys (Fig. 33B) and Pelecus the fossae are present but are shallow. They are absent in Leuciscus, most Barbus species, Labeo, Garra, Paralaubuca and Macrochirichthys. In some Barbus species (B. tor, B. longiceps) 'pseudo-posttemporaP fossae are developed. That is to say, instead of the lateral wall of the fossa being formed from the pterotic, it is provided by the enlarged posttemporal, the pterotic contributing only slightly to the anterior part of the fossa (Fig. 33C). EPO RTF APTE PTF Fig. 32 Posterior views of the neurocrania of: A. Luciobrama macrocephalus, B. Aspius vorax. A rather different situation is found in some Barilius species (B. bola, B. microcephalus, B. loati, B. niloticus} where there are no posttemporal fossae as such but a posterior (posttemporal) opening which invades the deep subtemporal fossa. Here, the pterotic contacts the epioccipital along the cranial surface and posterior border, forming the roof of the subtemporal fossa (Fig. 31). Ramaswami (1955b : 222) notes that 'In some form or other all genera possess a posttemporal fossa in Cyprinidae', while Roberts (1973) stated that posttemporal fossae are entirely closed in Cyprinidae. Neither of these statements is correct. Weitzman (1962) considered the presence of posttemporal fossae in cyprinids to be a specialized feature. Certainly the fossae take on a specialized form in Luciobrama and in long-headed repre- sentatives of other genera, but its presence in such relatively 'primitive' cyprinids as Opsariichthys, Zacco and Barilius, and the fact that this feature is much more widespread amongst the cyprinids than had previously been supposed, would indicate that it is a plesiomorph character. The exoccipitals show little variation in the genera examined. In all of them the lateral occipital foramen of each bone is extensive, its border being defined by a narrow strip of bone. The lateral limb of the exoccipital is directed at an angle of 45° in Luciobrama (Fig. 32A) and in other long- headed cyprinids (Figs 32B & 33A). In Labeo the bone is markedly modified. The lateral limb is directed horizontally (as it is in Barbus, Fig. 33C) and the lateral occipital foramen is reduced, being margined by a thickening of the bone. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 39 The epioccipital in Aspius has the posterior face of this bone produced into a thick, caudally directed process. Such a feature is also noted in Erythroculter and Schizothorax but appears to be absent in Aspiolucius and Pseudaspius. Except in Bar bus tor, where similar processes are present, such well-developed epioccipital features have not been found in the other genera examined. It is a feature no doubt associated with the elongation of the skull and the need to produce an extended surface for the attachment of epaxial muscle fibres. The dorsal surface of the epioccipital in Elopichthys covers a large area which, together with the parietals and supraoccipital forms an extensive posterior cranial platform (see p. 13). The basioccipital could not be examined in either Aspiolucius or Pseudaspius and the shape of the masticatory plate of the pharyngeal process cannot be ascertained. The plate is weakly developed in Aspius and Elopichthys as it is in Luciobrama, and the pharyngeal process itself is short and laterally compressed distally. 1C APTE 1C Fig. 33 B Posterior views of the neurocrania of: A. Elopichthys bambusa, B. Opsariichthys Uncirostris, C. Barbus oxyrhynchus. Opsariichthys, Zacco and some Barilius have a pharyngeal process which is laterally compressed and steeply angled, the masticatory plate being moderately developed. The masticatory plate is particularly well-developed in many Barbus and Labeo species and in Hypophthalmichthys (see Ramaswami, 1955b), which in some cases, completely hides the aortic foramen. It appears that in those taxa which possess slender pharyngeal bones and teeth the masticatory plate is weakly developed (e.g. Pelecus, Macrochirichthys and long-jawed Barilius species). The bulla acoustica is not well developed in any of the genera studied. The supraoccipital is basically similar in all genera examined. The crest of the bone is variously developed; for example in Labeo coubie there are two diverging wings arising from the medial lamellae; in Pelecus the bone forms a high point to the cranium but the medial crest is virtually absent and in Chela the entire crest is truncated. Direct contact of the supraoccipital with the neural complex appears to occur only in Labeo (Reid, pers. comm.), connection normally being effected by ligamentous sheets (see p. 19). The intercalar was said by Ramaswami (1955b : 216 -as opisthotic) to be 'normally absent'. However, in Aspius (1C, Fig. 32B) and Erythroculter the intercalar is extremely well developed and covers the area between the pterotic and epioccipital both dorsally and ventrally. Each intercalar in Schizothorax is reduced to a discoid bone lying ventral to the junction of the pterotic and epioccipital. The bone is also present in Opsariichthys, Barilius, Squaliobarbus, Culler, Leuciscus and some Barbus species. I have not found it in Labeo, Pelecus or Macrochirichthys. The parietals are especially elongate in Luciobrama, Aspiolucius and Pseudaspius, a condition also encountered in Hypophthalmichthys. The 'normal' condition in the Cyprinidae is for the 40 G. J. HOWES parietals to be short and wide, and the most extreme form of this condition is to be found in some Labeo species. As mentioned earlier (p. 39) the parietals in Aspius, Elopichthys and Erythroculter contribute to the formation of a posterior cranial platform (Figs 26 & 29). The posttemporals in Aspius, Pseudaspius, Aspiolucius and Elopichthys are like those in Lucio- brama (Figs 3, 26 & 29), namely, lamellate ventrally with a long anterodorsal extension, bordered ventrally by the lateral extrascapula. B Fig. 34 10mm Upper jaws, in lateral view of: A. Aspius vorax, B. Opsariichthys uncirostris, C. Barilius bola. The extrascapula is variously developed in the cyprinids (noted by Ramaswami, 1955b : 218 -as supratemporal). I have found it as an ellipsoid bone in Aspius, thin and elongate in Elopichthys and varying from a plate-like element lying over the posterior part of the pterotic (as in Squaliobarbus) to a lamellate bone running between the posttemporal and pterotic (as in Paralaubucd). The jaws The upper jaw in Aspiolucius and Pseudaspius appears, from radiographs, to be similar to that of Luciobrama (Fig. 7). In Aspius the premaxilla is thin with a large plate-like ascending anterior process. The premaxillae of Elopichthys are massive beak-like structures sutured along their midlines, except for a small anterior foramen. The ventral border of the bones is very sharp edged. The THE ANATOMY OF LVCIOBRAMA MACROCEPHALUS (LACEPEDE) 41 maxillae possess very long anterior medial processes, but the palatine processes are much reduced. The ventral border of the maxilla lies medial to the premaxilla and posteriorly it curves to project ventrally beyond the end of the premaxilla (Fig. 35). The premaxillae in Opsariichthys are very slender bones. At the symphysis the ventral border is rounded. The dorsal border of the maxilla displays a marked concavity anteriorly which is the point of insertion of the maxillary-palatine ligament. The anterior medial process of each maxilla contacts its fellow from the opposite side along a narrow face. The maxilla is separated quite widely for part of its length from the premaxilla. The bones contact each other at their anterior and posterior margins. 10 mm Fig. 35 Elopichthys bambusa, upper jaw. Lateral, dorsal and ventral views. In Barilius the anterior tip of each maxilla is extended forward and curved medially so as to almost contact its counterpart. The medial process of each bone, which underlies the anterior ascending process of the premaxilla, meet each other along the midline and maintain contact along a wide surface. These medial processes (maxillary rostral process of Ramaswami, 1955b) do not appear to contact each other in other cyprinid genera examined but are connected across the midline by a ligament. The upper jaw of Barilius bola is greatly elongate when compared with that of Opsariichthys (Figs 34B & C). The premaxilla is extremely long and thin, and dorso-laterally is overlapped for almost its entire length by the maxilla. The anterior ascending process of the premaxilla is beak- like and resembles that of Elopichthys (cf. Fig. 35) but is not so extensively developed. 42 G. J. HOWES Another cyprinid which is characterized by its long jaws is Macrochirichthys (see Howes, 1976). Here the upper jaw has a complex symphysial joint, mesial extensions being developed on the maxillae which serve to separate the two halves of the jaw to allow for the accommodation of the symphysial knob of the dentary when the jaws are closed. In most Barbus and Labeo species examined the maxilla possesses a wide palatine process. The mesial processes are short and are connected with each other by a long ligament. The ventral border of the maxilla overlaps the premaxilla for only part of its length. D Fig. 36 Lower jaws of: A. Pseudaspius leptocephalus, B. Oreoleuciscus pewslowi, C. Aspius vorax, D. Elopichthys bambusa, E. Opsariichthys unicirostris, F. Barilius bola, G. Schizothorax esocinus, H. Erythroculter mongolicus. The lower jaw in Aspiolucius and Pseudaspius is of the same narrow, canoe-shape as in Lucio- brama. There are 14 pores in the mandibular canal on the dentary, plus 2 on the anguloarticular. In Aspius (Fig. 36C) the dentary is proportionately thicker with an expanded symphysial process. The posterior border of the coronoid process is almost perpendicular, with the angulo- articular extending backwards from it at right angles. The dentary-anguloarticular junction in Aspius most closely resembles that of Erythroculter and Opsariichthys (Figs 36E & H). In Elopichthys (Fig. 36D) the dentary is very thin, the tapered and hammer-shaped symphysial process fiting into the corresponding curvature of the upper jaw when the mouth is closed. Posteriorly the coronoid process rises steeply to meet the anguloarticular. Like Luciobrama, the dorsal border of this process is continuous and not broken as it is in the other genera. The dentary of Opsariichthys (Fig. 36E) bears a strong symphysial process followed by a notch and then a curved thin wall which rises gradually to form the ascending process. A similar notch can be found in Paralaubuca and Macrochirichthys (see Howes, 1976 : 244). THE ANATOMY OF LVCIOBRAMA MACROCEPHALUS (LACEPEDE) 43 In Barilius bola (Fig. 36F) the dentary is exceedingly thin and gently concave just before it enlarges to form the symphysial knob. There is a low ascending process. The other Barilius species examined exhibit a lower jaw similar to that of Opsariichthys. The number of mandibular pores ranges from 6 to 12 in both genera. The dentary of Leuciscus, Barbus and Labeo is invariably short and narrow, tapering anteriorly in Barbus and Leuciscus but expanded in Labeo. In these genera the jaws curve sharply medially, whereas in all those genera mentioned previously the curve is a shallow one (see below). The dentaries of Luciobrama, Aspiolucius and Pseudaspius have weakly developed symphysial processes; posteriorly the curve of the dorsal margin of the coronoid process is continued by the anguloarticular instead of there being an abrupt discontinuity as noted in Aspius (Fig. 36C). This form of articulation of the dentary with the anguloarticular is encountered in many 'primitive' teleosts (Flops, Esox, Hoplias and most of the characoids). In the case of Luciobrama, etc. it is likely that this type of lower jaw has been derived from the condition found in Aspius and Opsariichthys, the larger adductor mandibulae muscles requiring a greater area of bone for insertion. It is of interest to note that the posterior part of the lower jaw of Oreoleuciscus (Fig. 36B) greatly resembles that of Luciobrama, although the anterior part of the jaw is narrowed and curved like that of a 'typical' leuciscine. Matthes (1963) and Liem (1970) pointed out that the presence of a high coronoid process on the lower jaw increased the power and speed at which the jaws could be closed. Only in Elopichthys, Schizothorax, Erythroculter , Pelecus, Leuciscus and some species of Barbus and Labeo can the coronoid process be considered as relatively high. In such exclusively piscivorous cyprinids as Luciobrama, Barilius bola and Macrochirichthys the process is low. In this respect it is probably the distance and angle of the coronoid process from the point of articulation with the quadrate which is more important. It will be noted by referring to Fig. 36 that in the long-jawed species in which the jaw is aligned at an angle (represented here by Barilius bold), the coronoid process is placed well posteriorly and is relatively low, whereas in Elopichthys, where the jaw is aligned horizontally, the coronoid is high, although still placed well back. In some cyprinids the coronoid process is placed close to the anterior tip of the jaw. This apparent forward movement is due in part to the greater mesial curvature of the dentary and its subsequent foreshortening, seen particularly in Labeo and Barbus species where it is correlated with the broadening of the head. In these fishes the higher the coronoid region the greater the area available for the adductor mandibulae muscles which operate at a greater angle than those in the long-jawed species, where the muscle fibres lie partially horizontally. Dr Keith Banister has pointed out to me that in some species of Barbus this forward movement of the coronoid process is more real than apparent, and that it seems correlated with a definite shortening of the ethmoid region. Ramaswami (1955b : 223) remarks that the upper jaw bones of cyprinids show a uniform morphology. Generally speaking this is true. Most cyprinids rely on protrusion of the upper jaw and employ suction feeding (Alexander, 1964, 1966, 1967 & 1969), and this requires a particular association between the maxilla and the premaxilla. The extent of movement of the upper jaw in Luciobrama appears to be rather restricted. The ascending processes of premaxillae are not particularly long, and manipulation of preserved material shows that the upper jaw is not very protractile. From measurements of the bones and manipulations of the jaws of Barilius it would seem that the shorter jawed species possess more mobile upper jaws than do the longer jawed species such as B. bola where, like Macrochirichthys, there appears to be very restricted movement of the premaxilla. It must be emphasized that these observations have been made on preserved material only and so must be open to doubt. In Elopichthys the upper jaw is so modified that the only kind of movement possible is rotation against the ethmoid (see p. 41), and jaw action here must be centred upon fast snapping action (an almost parallel jaw action can be seen in the African characoid, Phagoborus}. Greenwood (1974) noted that in piscivorous cichlids the long lower jaw is accompanied by a highly protrusile upper jaw (this is particularly noticeable in the South American cichlid genus Petenid). But in the long-jawed piscivorous cyprinids the reverse seems to be the case. Here the 44 G. J. HOWES elongation of the jaws has been by posterior extension and this has necessitated a correlated movement of the hyopalatine arch so that the shaft of the hyomandibula is perpendicular and the jaw articulation is moved backward and downward to below the orbit. There has been no great extension of the anterior ascending process of the upper jaw nor of the ethmoid-vomerine region of the cranium. Only in Elopichthys has any marked modification been achieved in the forward extension of the upper jaw. However, I consider this fish to be a 'long-headed' rather than a 'long-jawed' cyprinid and in these species the hyopalatine arch is differently orientated (see further discussion on p. 61). ECT SY 10mm Fig. 37 Hypopalatine arches shown in lateral view of: A. Aspius vorax, B. Elopichthys bambusa, C. Erythroculter mongolicus. Hyopalatine arch (Figs 37 & 38) I am unable to determine the precise nature of the elements constituting the hyopalatine arch in Aspiolucius and Pseudaspius. From radiographs it is possible to follow the outline of the bones, and these closely resemble the description given below for Aspius. In Aspius (Fig. 37A) the ventral limb of the hyomandibula is short, the posterior border of the bone is convex and the lateral face bears a shallow depression. By contrast, the hyomandibula of Elopichthys (Fig. 37B) has a well-developed flange in this position. The hyomandibula in Opsariichthys and Barilius (Figs 38A & B) is vertically aligned; the ventral limb is thick and long. Here too the lateral face bears a wide flange along which the dilatator operculi muscle runs. A lateral flange is also present on the hyomandibula of Barbus tor but is absent in other species of Barbus I have examined. In Macrochirichthys the hyomandibula has a lateral process on which is inserted the levator arcus palatini muscle (see Howes, 1976). Similar lateral processes serving for muscle attachment are present in Pseudolaubuca. THE ANATOMY OF LUCfOBRAMA MACROCEPHALUS (LACEPEDE) 45 In Hypothalmichthys the lateral surface of the hyomandibula is strongly curved outward so that the suspensorium lies beyond the cranial border. The quadrate of Aspius (Fig. 37A) is a short high bone separated by a wide area of cartilage from the metapterygoid. The anterior border is perpendicular, the dorsal border curved. A small foramen is present posterior to the articulatory condyle, a feature shared with Luciobrama (see p. 15). I have found such a foramen in four other genera: Elopichthys and Erythroculter, in which it is large (in the skeleton of Elopichthys examined, the foramen is present only in the right quadrate) ; Pseudoxygaster and Macrochirichthys, in which it is minute. METP B Fig. 38 Hyopalatine arches shown in lateral view of: A. Opsariichthys uncirostris, B. Barilius bold. The quadrate of Opsariichthys is a thin, curved bone, the dorsal border being concave and forming the margin of a large fenestra. This feature is shared with Zacco, where the fenestra is not so well developed, with Macrochirichthys, in which it is about equally developed, with Pseudoxygaster and with Salmostoma. Further discussion of the metapterygoid-quadrate fenestra appears on page 46. The symplectic is an extensive bone in most of the genera examined. Medially it overlaps the posterior extension of the quadrate. In Opsariichthys the symplectic forms the postero-ventral border of the metapterygoid-quadrate fenestra. In Barilius it is a small element lying medial to the quadrate (Fig. 38B). Ramaswami (1955b : 224) comments on the symplectic in other cyprinid genera. 46 G. J. HOWES A dermopalatine has not been found in any of the genera examined. The autopalatine varies little in the genera examined. Differences involve the degree of develop- ment of the articular head with the ethmoid; it may be very wide and flat in this region as in Hypophthalmichthys or rod-shaped as in Luciobrama. The medial face of the palatine is almost always concave. There is also variation in its articulation with the entopterygoid. Usually this is by way of a right-angled indentation in the entopterygoid bnt sometimes as in the case of Barilius bola (Fig. 38 B) and Elopichthys (Fig. 37B) the edge of the palatine may be sloped and partly overlapped laterally by the entopterygoid. In both Opsariichthys and Barilius the palatine is laterally compressed and in the former genus bears a ventral keel. The ectopterygoid in most cyprinids is a thin sheet which partly underlies the medial face of the quadrate and the ventral margin of the entopterygoid. The 'generalized' condition of this bone appears to be wide with a slightly concave to convex anterior mrgin. In the long-jawed taxa such as Elopichthys, Macrochirichthys and Barilius bola the ectopterygoid is very narrow with a markedly concave anterior border, Figs 37B & 38B. Only in Luciobrama (and possibly in Aspiolucius) does the ectopterygoid appear to be rod-shaped. The entopterygoid is generally short and deep, its dorsal border rounded. In Luciobrama, Aspiolucius, Pseudaspius and Elopichthys the bone is narrow and long with a straight dorsal margin. In broad-headed species of the genera Barbus and Labeo the entopterygoid is markedly medially directed and often bears a thick ridge along the fold for the insertion of the adductor arcus palatini muscle. Ramaswami (1955a) notes the presence in Labeo macrostoma and Cyprinus carpio of a mesial facet on the entopterygoid which articulates with the lateral ethmoid. Other Labeo species I have examined have this facet variously developed and it seems to be well developed in those species with a long ethmoid but is virtually absent in those with a short ethmoid. In other cyprinid genera this part of the entopterygoid is usually a wedge-shaped platform and varies considerably in size. The metapterygoid of almost all genera examined is deep with the lateral ridge developed to varying degrees depending on the angle at which the bone is directed mesially toward the parasphenoid. The dorsal border is invariably concave (the one exception to this rule being Luciobrama). In Opsariichthys (Fig. 38A) the metapterygoid is narrow and axe-shaped, its ventral border forming the dorsal margin of the metapterygoid-quadrate fenestra. Posteriorly it is sutured to the hyomandibula. There is a small ascending process on the dorso-posterior edge of the bone (METP). Barilius bola has a similar metapterygoid process but in this case it is developed as a long anteriorly directed spine which serves both as the point of origin for the adductor mandibulae A3 and for the insertion of the levator arcus palatini muscles. Posteriorly, the metapterygoid has a long perpendicular border in contact with the hyomandibula (Fig. 38B). A metapterygoid spine appears to be present in all the species of Barilius I have examined but in none it is developed to the same degree as in B. bola. There is no metapterygoid spine in Zacco. The significance of the metapterygoid-quadrate fenestra has been discussed by Regan (191 1)1 and remarked on again by Ramaswami (1955b), Weitzman (1962), Gosline (1973, 1974 & 1975) and Howes (1976). Most authors, including Regan (1911), have assumed it to be a generalized or primitive character. Gosline (1973), however, expressed doubts about this, and was inclined to attach more significance to the state of the entire pterygoid complex. In those genera with a long postorbital cranium (Luciobrama, Aspiolucius, Pseudaspius) the hyomandibula has a short ventral limb which is directed anteriorly, and there is also an increase in the length of the pterygoid bones but not in their depth. In the majority of cyprinids the hyomandibula is a deep bone with its shaft aligned almost perpendicular to the skull. The pterygoid bones are also short and deep and in those genera with wide and depressed crania the metapterygoid is often acutely directed mesad toward the parasphenoid (e.g. Labeo) thus providing a large space to accommodate the thick adductor mandibulae muscles. 1 Regan (191 1) stated that a fenestra is present in Chela but he did not designate the species. I have been unable to detect it in any species now assigned to the genus Chela (BanaYescu, 1968) and I suspect that Regan was in fact referring to Macrochirichthys, which he knew as Chela macrochir. It would seem that the skeleton of Macrochirichthys macrochirus in the BMNH collection which was available to Regan was the only skeleton bearing the generic name Chela at the time he would have examined it. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 47 If a comparison is made with the situation in characoids where the metapterygoid-quadrate fenestra is widespread the following points emerge. In the majority of characoids there does not seem to be the same degree of dorsal extension of the pterygoid bones as occurs in cyprinids. The ectopterygoid and overlying entopterygoid are long (a correlate of the generally more elongate snouts and jaws of characoids). However, in broad-headed characoids such as Erythrinus, Lebiasina, Citharinus, and in some leporinids and prochilodontids the pterygoid bones are deep and directed acutely mesad as in the broad-headed cyprinids. In these cases the metapterygoid-quadrate fenestra is either absent or reduced. Furthermore, it is noted that the adductor mandibulae and levator arcus palatini muscles in the characoids do not utilize the surfaces of the pterygoid bones to the same degree as in the cyprinids and in those characoids where the adductor mandibulae muscles are particularly extensive (Hepsetus, Ctenolucius, Acestrorhynchus, Salminus) there has been a forward extension of the hyomandibula along the dorsal margin of the metapterygoid thereby providing the additional surface area of attachment (see Roberts, 1969). The fenestra between the quadrate and metapterygoid is certainly a functional device which would seem to serve either to relieve stresses by directing forces generated in the lower jaw around the perimeter of the pterygoid bones and into the cranium or perhaps, more importantly, it acts as a type of hinge which enables the pterygoid bones to undergo lateral rotatory movements. Again, it is noted that those characoids in which the fenestra is reduced are those in which there is little or no development of the symphysial articulation of the lower jaw - as in most of the cyprinids. There are also differences in the articulation of the palatine with the ento- and ectopterygoids in the broad-headed and narrow-headed cyprinids which indicate different degrees of lateral rotation. The exact functional significance of this fenestra is not obvious and may only be realized when all the vectors have been analysed. Whatever its use as a stress-relieving or force-directing device, one possible advantage of this feature seems that in Macrochirichthys, Pseudoxygaster and Salmostoma its presence has allowed the jaw articulation to move further forward than in other cyprinids. The jaws of Macrochirichthys are proportionately as long as those of Barilius bola and it might be supposed that retention of such a fenestra would also have been an advantage to this bariliine. However, in Barilius there is a well-developed adductor mandibulae A3 muscle extending from the dorsal process of the metapterygoid (see p. 55), whereas in Macrochirichthys and Pseudoxygaster A3 originates from the hyomandibula (see Howes, 1976 : 242). The develop- ment of this muscle and bone in Barilius as a functional unit could account for a ventral expansion of the hyomandibula so as to occlude any opening that may have been present in the ancestral form. Some confirmation of this comes from my current studies on Salmostoma in which there is a reduction in the size of the metapterygoid-quadrate fenestra in those species with the longer jaws. My current researches on Macrochirichthys indicate that neither it, Salmostoma nor Pseudoxy- gaster are at all closely related to Opsariichthys or Zacco and so the metapterygoid-quadrate fenestra cannot therefore be regarded as a synapomorph feature. I conclude that it represents a primitive cypriniform character (widespread occurrence in the Characoidei) and that in Opsariichthys and those genera in which it occurs it should be considered as plesiomorph. Possibly the potentiality for its development was inherited in several lineages and was realised (as in Macrochirichthys) under the necessary combination of selective pressures. Opercular series In Aspiolucius, Pseudaspius and Aspius the morphology of the opercular elements closely resembles that mLuciobrama; the pre-, inter- and subopercula are extensive bones. The operculum has a long concave dorsal border and a well-developed anterior extension for the insertion of the dilatator operculi muscle. The medial strut contains two ventral foramina. The preoperculum has 9-10 pores along the ventral border. In Elopichthys the vertical limb of the preoperculum is longer and the horizontal limb shorter, than in the genera mentioned above. The ventral border bears 6 pores. The medial strut of the operculum is feeble and contains a single large foramen. 48 G. J. HOWES In Opsariichthys and the majority of Barilius the lower limb of the preoperculum does not extend so far forward and the ascending process of the limb is almost vertical. Three pores of the mandi- bular lateral line canal are present on the ventral border. The operculum is deep, with a short dorsal border; it lacks the prominent anterior extension for the dilatator operculi muscle. The interoperculum in Barilius bola has been greatly reduced in length, a change in proportion correlated with the backward shift of the jaw articulation (see above, p. 43). Most Bar bus and Labeo species have a wide, vertically orientated preoperculum which has a short ventral limb. The operculum generally is deep, with a short dorsal border, but Barbus tor, B. longiceps and B. barbus all have a long dorsal border. A large operculum with a long dorsal border appears characteristic of some predatory cyprinids, e.g. Pelecus, Macrochirichthys and Erythroculter, where it is associated with a shallow, acutely angled preoperculum. This is the situation encountered in Luciobrama, Aspiolucius, Pseudaspius and Aspius in which are present noticeably elongate pre-, inter- and subopercula. There is an opposite situation in the long-jawed Barilius species where an operculum with a short dorsal border is associated with deep and narrow opercular elements. Hyoid and branchial arches HYOID ARCH. There is little variation in the morphology of the elements comprising the hyoid arch in the genera studied. All the branchiostegal rays, apart from the first, are expanded proximally. The urohyal is variously developed, sometimes being channelled ventrally as in Aspius and Opsariichthys, or flat, as in Barilius and Barbus. The medial plate can be short and high as in Schizothorax or reduced to a slight ridge as in Elopichthys. The basihyal is greatly elongated in Luciobrama (p. 17), a condition also encountered in Aspiolucius and Elopichthys. THE BRANCHIAL ARCHES. Intcrgencric comparison reinforces Ramaswami's (1955b) opinion that there is a general uniformity in the branchial elements of cyprinids. I have not found either first or fourth infrapharyngobranchials represented as an ossified element. There is some variation in the development of the dorsal processes on the epibranchials. The pharyngeal teeth of Aspiolucius are arranged in two rows with 3-5 teeth, those of Pseudaspius in two rows (2-4), of Aspius in two rows (3-5) and of Elopichthys in three rows (5-3-2 or 5-4-2). The whole problem of what constitutes a primitive pharyngeal dentition in the cyprinids has been discussed at length by numerous authors (see Nelson, 1969 : 513). In the case of Luciobrama it seems evident that a single row of pharyngeal teeth is a derived condition. Weberian apparatus and swimbladder In Aspius, Elopichthys and, as far as I can tell from radiographs, in Aspiolucius and Pseudaspius, the os suspensorium is curved anteriorly, the lateral processes of the fourth centrum (PR4) are short and thin, and those of the second centrum (LP2) are curved slightly dorsad. In all these genera, except Aspius, the second neural plate contacts the supraoccipital, and the fused neural complex is in contact with a supraneural plate. In Luciobrama, Aspius and Elopichthys (Figs 39A & B) the second and third centra are not completely fused. Dorsally a distinct separation can be detected; ventrally this is not so clearly defined since the suture line is always visible and there is never the complete fusion seen in other cyprinids where the division between the centra has become completely obliterated. In Opsariichthys and Barilius (Fig. 39D) the lateral processes of the fourth centrum (PR4) are weakly developed and directed posteriorly. The lateral processes of the second centrum are greatly expanded laterally and are not curved. The tripus in these genera is very elongate. The second and third centra are separated in Opsariichthys but they are fused in Barilius. In Schizothorax, Barbus and Labeo, and in the majority of genera examined, the os suspen- sorium is curved anteriorly and the lateral processes of the fourth centrum are well developed ; those of Barbus are often expanded distally. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 49 In Macrochirichthys and Pseudoxygaster the os suspensorium is very short and vertically directed; the tripus is elongate and the second and third centra are separated and in Pelecus (Fig. 39C) the os suspensorium is almost horizontally directed. Although Pelecus has been included in the Cultrinae along with Macrochirichthys and Pseudoxygaster (see Banarescu, 1967) there are many differences in the structure of the anterior part of the vertebral column. The first and second centra of Macrochirichthys and Pseudoxygaster are greatly modified, whereas in Pelecus these centra are of a generalized nature. (Work is in progress on the description and analysis of these elements in the Cultrinae.) LP2 PR4 5mm OSS B D Fig. 39 Weberian apparatus shown in ventral view of: A. Aspius vorax, B. Elopichthys bambusa, C. Pelecus cultratus, D. Barilius microcephalus. Sorescu (1972) studied the Weberian apparatus of representatives of the subfamilies Danioninae and Cultrinae. Her ideas of primitive and specialized characters exhibited by these skeletal elements are ill-defined and her conclusions concerning the relationships of the Cultrinae and Danioninae - neither of which are monophyletic assemblages - are, in my opinion, invalid. As far as I am aware no comparative study has been made of the swimbladder throughout the Cyprinidae. Tchang (1933) frequently notes the condition of the swimbladder in various genera. 50 G. J. HOWES He mentions that the organ in Opsariichthys is enclosed in a bony capsule. I can find no evidence to justify this statement. The swimbladder in most of the genera studied is a simple bipartite structure. In some genera at present included in the Cultrinae there is a further posterior division. Pectoral girdle In Aspiolucius, Pseudaspius and Aspius the pectoral girdle is similar to that of Luciobrama. In all these genera the horizontal limb of the cleithrum is directed forward to a point below the anterior border of the prootic, and (except in Aspius) it is bifurcated anteriorly. The coracoids are joined only anteriorly. The mesocoracoid in Aspius is a thin bridge of bone. The postcleithrum in Aspiolucius and Aspius is short, like that of Luciobrama but in Pseudaspius it is elongate. The supracleithrum is elongate in Aspius, while that of Aspiolucius and Pseudaspius is of a length inter- mediate between that of Aspius and Luciobrama. In Elopichthys the anterior margin of the ventral limb of the cleithrum is very wide, the ascending limb short. The mesocoracoid is wide and the coracoids are joined medially for half of their lengths. The postcleithrum is very long and the supracleithrum as elongate as that of Aspius, In both Opsariichthys and Barilius the vertical limb of the cleithrum is narrow. The coracoids are joined anteriorly, the mesocoracoids are long and narrow, the supracleithra elongate and the postcleithra very long and spine-like. A similar kind of arrangement, but with the upright and horizontal arms of the cleithrum broadened, is found in Barbus and Labeo. Throughout the cyprinids there appears to be considerable variation in the degree of develop- ment of the cleithral-coracoid fenestra which may, as in the case of Chela, be entirely lacking. The development of this feature appears to be associated with the variation of the pectoral fin muscles and ventral extension of the coracoids. In Salmostoma, Oxygaster, Pseudolaubuca, Paralaubuca, Pelecus, Pseudoxygaster, Chela and Macrochirichthys there is a great expansion and complete, or almost complete, medial contact of the coracoids (see description of pectoral girdle of Pelecus by Rauther, 1950). Regan (1911 : 28-29) states that 'Barilius, Danio etc. agree with Opsariichthys in the form of the cleithra, rounded or pointed anteriorly, and these are connected by genera such as Aspius with Leuciscus and its allies, in which the cleithra are more expanded and truncated anteriorly . . . the Barbus group differs in that the cleithra are distinctly emarginate anteriorly'. Broadly speaking, the cleithra of all these genera are similar, but in Barbus, Leuciscus and Alburnus the curved dorso-lateral margin of the cleithral limb is raised into a blade-like ridge. Furthermore, it is not the case, as stated by Regan (see above), that all Barbus exhibit anteriorly emarginate cleithra. In Barbus tor, for example, the cleithra are forked anteriorly as in Luciobrama, providing two places of origin for the sternohyoideus muscle (see p. 19). One variable element of the pectoral girdle is the postcleithrum. It is usually found as a long medially curved bone which reaches its most extensive development in Macrochirichthys. In some genera (e.g. Salmostoma) it is reduced to a short spike and in Barilius appears as a small scale-like bone, whilst in Pseudoxygaster it appears to be entirely lacking (although it may possibly be identified as a modified external scale above the pectoral fin). Sorescu (1968) uses the morphology of the pectoral girdle as a principal character in differentiating cyprinid subfamilies. She has placed reliance on similarity of shape of the elements to indicate affinity. This has led, for example, to placing Barilius zambezensis in the Cultrinae without taking account of other Barilius species. Sorescu's failure to utilize other and probably more significant cranial characters, and her apparent failure to recognize parallelism, renders her conclusions doubtful. Pelvic girdle There is little variation in the pelvic girdle. In some genera the pelvic bone is only shallowly forked, but as far as I am aware there are no cyprinids with an unforked pelvic bone, the usual condition in characoids. THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 51 Appendicular skeleton VERTEBRAL COLUMN. The total number of vertebrae in Aspiolucius, Pseudaspius, Aspius and Elopichthys ranges from 51 to 54 (cf. 55 in Luciobrama). Of the other genera examined, only Pelecus, Hemiculterella and Ochetobius possess more than 50 vertebrae. It is to be noted that in Ochetobius the increased number is in the posterior abdominal region (i.e. between the dorsal and anal fin bases). No other member of the Cyprinidae has such a distance between the posterior ray of the dorsal and the origin of the anal fin. In all remaining genera examined the total number of vertebrae is from 41 to 49. The more elongate cyprinid species show an increase in vertebral number (see Lindsey, 1975). The proportions of the centra in all cyprinid genera I have examined are virtually identical. Only in Macrochirichthys and Pseudoxygaster is there any appreciable elongation of the precaudal centra, but in none have the caudal centra been lengthened. Table 1 Vertebral counts of some Cyprinid genera. These are arranged in groups of what I believe to be related genera. W = Weberian vertebrae. * = Second and third centra separate or partially separated. ? = Condition unknown. A = Abdominal. C = caudal vertebrae. T = total number. SN = Supraneurals Genus W SN Luciobrama 4* 30 21 55 12 Aspiolucius 4? 27 23 54 13 Pseudaspius 4? 26 21 51 11 Aspius 4* 25-26 24 51-52 13-14 Elopichthys 4* 25-27 22-23 52-54 15-16 Barilius spp 4 16 23-25 43^5 11 Barilius bola 4 22 21 48 13 Opsariichthys 4* 19 21 44 6-7 Oxygaster 4* 16 23 43 13 Pseudolaubuca 4* 17 26 47 15 Pseudoxygaster 4* 19 21 44 12 Macrochirichthys 4* 23 22 49 16 Hemiculterella 4* 17 23 44 10 Pelecus 4 24 24 52 22-23 Erythroculter 4 19 23 46 7 Parabramis 4 19 21-22 45^16 5-6 Paralaubuca 4 16 22 42 12 Schizothorax 4 22-25 19-20 45-49 11 Oreinus 4 22-24 20-21 47-48 12-13 Leuciscus 4 21-22 20-22 46-48 10-12 Abramis 4 20 22 46 10 Ochetobius 4 34 22-23 60-61 14 The total number of vertebrae in Barbus and Labeo never exceeds 48 (Banister and Reid, pers. comms). The supraneurals vary considerably in their development and in their number, ranging from 5 to 6 in Parabramis to 23 in Pelecus. In Luciobrama, Aspiolucius, Pseudaspius, Aspius and Elopichthys they are thin rod-like structures, numbering 11-16, whereas those of Schizothorax, Oreinus, Barbus and Labeo are plate-like, numbering 10-13. In Semiplotus the supraneurals reach their most marked degree of development, expanding between the neural spines to form a rigid pre-dorsal septum along the vertebral column. In Macrochirichthys the anterior supraneurals are horizontally aligned and are in contact with the enlarged neural spines (Howes, 1976 : 244). A similar arrangement is present in Pseudoxy- gaster and Pseudolaubuca. 52 G. J. HOWES MEDIAN FIN SKELETON. In all cyprinid genera examined the first proximal dorsal pterygiophore is expanded. In those with well-ossified anterior dorsal rays the corresponding proximal pterygio- phores bear lateral struts. Roberts (1973) comments on the number of radials (pterygiophores) supporting the dorsal and anal fin rays in cyprinids. He notes that there are three in Opsariichthys but that according to Bridge (1896) there are usually two in cyprinids. However, Bridge (op. cit.) did in fact identify three elements in all the cyprinids he examined, namely species of Barbus, Cyprinus, Abramis and Tinea. In all the species I have examined there are three elements, distal, medial and proximal pterygiophores. PMX Fig. 40 Elopichthys bambusa, cranial musculature, superficial elements. CAUDAL FIN SKELETON. In all the genera examined there is a marked consistency in the construction of the caudal fin skeleton. The only variation appears to be in the size of the first hypural and the number of epurals (either one or two). Roberts (1973) gives 10 + 9 as the count for the principal caudal rays, but I count 9 + 9 in all genera examined. Shukla & Verma (1972) have described the caudal fin skeleton of Barilius bola. They have, however, misidentified many of the elements and have used published descriptions of other cyprinids in their comparative treatment. Thus their conclusions concerning the 'primitive' nature of this species are very dubious. Myological characters Cranial muscles Several workers have described the cranial muscles of various cyprinids (e.g. Takahasi, 1925' several genera; Ping et al., 1958, Cyprinus; Saxena, 1960, Garra; Matthes, 1963, Labeo, Barbus, Barilius and other genera; Munshi & Singh, 1967, Labeo and Cirrhina; Meinel et al. 1970, Barbus nasus, Ctenopharyngodon and Squalius). THE ANATOMY OF LVCIOBRAMA MACROCEPHALUS (LACEPEDE) 53 From a superficial dissection of the type specimen of Aspiolucius esocinus it would appear that the cranial muscles of this species are like those in Luciobrama. The dilatator operculi and levator arcus palatini are developed in the same way and the adductor mandibulae muscles are similarly arranged. I was unable to examine the arrangement of the deeper muscles. Pseudaspius and Aspius both exhibit the same basic arrangement as Luciobrama. A2b(cut) A 2 (cut) A2b(cut) Fig. 41 Elopichthys bambusa, cranial musculature, deeper layers. Adductor mandibulae A2 is cut through to expose A3, which in turn is cut to show the tendinous fascia. The cranial muscles of Aspius aspius have been described by Suslowska (1971). I have compared the muscles of Aspius vorax with those of A. aspius and find them to be of almost identical morphology. Suslowska (1971) does not recognize the conventional divisions of the adductor mandibulae. Although she points out that the external portion (Al) inserts on the maxilla, she does not state that it is bifurcated at this insertion. The levator arcus palatini is extensive ; a tendinous sheet extends through the lateral surface of the muscle to join the preoperculum and this provides a site of origin for part of the adductor mandibulae (A2). The levator is divided by A3. An adductor arcus palatini is present. Suslowska (1971) noted the presence of an adductor hyomandibulae. The form of this and the other muscles she describes for A. aspius are as those in A. vorax. The musculature of Elopichthys (Figs 40-44) resembles that of Aspius in that the adductor mandibulae is separable into its various parts only anteriorly. The adductor mandibulae Al inserts via two tendons onto the maxilla (Fig. 40). A2 inserts onto the rim of the anguloarticular. Aw is poorly developed. The morphology of the inner element, A3, differs quite considerably from that described in any other cyprinid. This muscle originates partly from the frontal, partly 54 G. J. HOWES from the pterosphenoid basin (see p. 32 & Fig. 42) and partly from the dorsal aspect of the sphenotic. Laterally it stems from the face of the sphenotic and pterotic and from a tendinous fascia covering the hyomandibula (tf, Fig. 41). PTS H Fig. 42 Elopichthys bambusa, cranial musculature. A. Deep dissection to show levator arcus palatini; the adductor mandibulae muscles having been removed. B. Inner section of levator arcus palatini. The circle indicates that part of the tendon of the inner muscle which is visible before the outer layer is removed. The levator arcus palatini in Elopichthys runs from the lower border of the frontal, the faces of the pterosphenoid and the sphenotic. The muscle is a complex one; the anterior bundle of fibres, which runs from the frontal, joins a medial pinnate sheet of fibres which originates from the face of the pterosphenoid and sphenotic to insert on the hyomandibula (Fig. 42A). The ventral border of this element is a stout tendon. When this outer layer is removed there is revealed a pinnate medial section with its fibres running in the opposite direction (Fig. 42B). When this medial layer THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 55 is removed there is yet another, triangular sheet bordered anteriorly by a wide tendon (tlap, Fig. 43). This tendon stretches from the ventral surface of the frontal to insert upon the edge of the metapterygoid. The muscle fibres arise in part from the frontal but mostly from the sphenotic. Insertion is along the medial face of the hyomandibula. I would identify this medial element as an adductor hyomandibulae. The adductor arcus palatini is a stout muscle which stems from the base of the prootic to insert ventrally upon the metapterygoid. The dilatator operculi runs across the upper border of the levator arcus palatini from the pterotic and sphenotic to the anterior process of the operculum. Medially the muscle originates in part from the face of the hyomandibula (a condition also found in Pseudaspius). Suslowska (1971) states that the dilatator inserts upon the hyomandibula in Aspius. aap MET Fig. 43 Elopichthys bambusa, cranial musculature. The outer layers of the levator arcus palatini have been removed to expose the inner layer and the adductor arcus palatini. In Opsariichthys, adductor mandibulae Al is a single element inserting on the maxilla; A2 is a large muscle inserting on the anguloarticular and separated medially from A3 by the levator arcus palatini. Aw is present. The dilatator operculi originates from a discrete lateral fossa formed by the sphenotic. It is not covered by any part of the adductor mandibulae since this complex muscle originates below its ventral border. In Barilius the adductor mandibulae Al is a narrow, divided element (Matthes, 1963, recognized divisions Ala and Alb) the anterior part of which is bordered ventrally by a strong ligamentum primordium. It inserts along the lateral face of the maxilla. The larger element inserts via a thick tendon medial to the outer element. (In Barilius bola the medial section of Al joins the maxilla along its distal border; there is no tendon of insertion, see Fig. 44.) The A2 section of the adductor in Barilius bola and other long-jawed Barilius species is extensive, and originates from the preoperculum, pterotic and sphenotic; it gives rise to a reduced Aw section. A3 originates from the dorsal process of the metapteryoid (see p. 47). This process also provides the insertion area for the levator arcus palatini which originates from the ventral margin of the frontal and from the sphenotic. The dilatator operculi runs from the posterior part of the sphenotic process and is covered by A2. The adductor arcus palatini is well developed. In other, shorter-jawed Barilius species (e.g. ubangensis, bendelisis, ornatus) adductor mandibulae A2 is not 56 G. J. HOWES as extensive, does not cover the dilatator operculi and does not originate from the sphenotic; the metapterygoid process is reduced and the adductor arcus palatini is small. In other words, the cranial muscle arrangement greatly resembles that of Opsariichthys (see Takahasi, 1925). Indeed, this basic type of morphology has been found in all the other cyprinid genera examined. There are, of course, modifications, as for example in Oxygaster and Pseudolaubuca where the anterior part of the adductor mandibulae Al is narrowed and tubular; the dorsal part of the levator arcus palatini in Pseudolaubuca extends to insert upon the operculum together with the dilatator operculi; the dilatator operculi is divided in Labeo, and the levator arcus palatini is complexly divided, as described in this paper for Elopichthys and in Macrochirichthys (Howes, 1976). METR do A2&A3 5mm Fig. 44 Barilius bola, cranial musculature. Adductor mandibulae A2 and A3 are reflected to expose the deeper muscles. A ligamentum primordium is not well developed in the syprinids studied. Only in Barilius have I have found it to be clearly differentiated and this not to the same extent as it is in most characoids. It may be a primitive character of cyprinids. Earlier (p. 35), differences in the morphology of the dilatator operculi fossa were pointed out; four main types of fossa can be distinguished : 1. That involving mostly the sphenotic, with part of the frontal and pterotic, and lying laterally on the cranium, sometimes roofed by part of the frontal. This type of fossa is usually small and found in many cyprinid genera (e.g. Opsariichthys, Zacco, Rasbora, Danio, some Barilius and Bar bus species and most cultrine species). 2. That involving a broad sphenotic process and a large area of the frontal, and which extends onto the cranial roof. This type of fossa is variously developed and can be extensive as in Hypophthalmichthys, where it occupies a large area of the frontal. It is characteristic of most genera currently assigned to the Leuciscinae. In Characoids a similar fossa is found in the Cynodontini (Howes, 1976). THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) 57 3. That in which the sphenotic is developed below the frontal as a shelf. This type of fossa occurs in Luciobrama and its relatives. In characoids a similar type of fossa is found in Salminus (Roberts, 1969) and in Brycon alburnus (pers. obs.). 4. That in which the frontal and underlying sphenotic have become partly separated to form a foramen. This type of fossa is found in some Barbus species in Cyprinus, in Labeo and what are believed to be related genera (Reid, unpublished information). In characoids a similar foramen is found in some erythrinids, Acestrorhynchinae and Ctenoluciidae (Roberts, 1969; pers. obs.). Luciobrama Aspiolucius Pseudaspius Aspius Elopichthys ?LEUCISCINAE Fig. 45 Cladogram to illustrate the relationships of the aspinine genera. There is also another condition, whereby the dilatator fossa is virtually absent. This occurs in the long-jawed Barilius species, e.g. bola, loati, and in Macrochirichthys. In these taxa the dilatator operculi muscle has either been so reduced (Barilius) or runs almost perpendicularly that the need of an inclined or horizontal shelf has been eliminated. Roberts (1973) places little value on the condition of the dilatator fossa, believing it to be 'extremely labile'. Admittedly, we must be aware of parallelism in the formation of this as in any other character, but I believe the particular characteristics of this feature are indicators of relationships. Gosline (1975) also draws attention to the varying conditions of the dilatator fossa. Hyoid and branchial muscles During this present study no extensive comparison of these groups of muscles has been made. It is noted that the development of the hyoid muscles in Luciobrama and Aspius is relatively 'weak' when compared with that of the Barbus and Labeo groups of genera, where the hyohyoidei are well developed (see Matthes, 1963). 58 G. J. HOWES Various arrangements of the hyoid muscles are found in the cyprinids (i.e. intermandibularis absent; protractor hyoidei divided anteriorly), but the basic plan is little modified from that described in Opsariichthys by Takahasi (1925). Concerning the branchial arch muscles, the only comments I am able to make at present concern the retractor pharyngeus of the upper branchial arches. Winterbottom (1974) points out two conditions of this muscle, one as observed in Opsariichthys, where it is undivided, and the other as in Cyprinus, where there are two divisions. In Luciobrama, Aspius and Elopichthys there appears to be a single element stemming from the basioccipital process and inserting along the medial edge of ceratobranchial 5. Considerations of functional morphology in Luciobrama The head of Luciobrama is enormously elongate, but as described in this paper, this elongation is entirely postorbital in extent. The preorbital part of the head is 'normal' and in fact greatly resembles that of Aspius, both in shape and proportions. Suslowska & Urbanowicz (1957) have commented on the feeding mechanism of Aspius, and their comments may be of help in elucidating those of Luciobrama. These authors compared Aspius with Cyprinus and Eso x and considered the morphology of the cranium and jaw in Aspius to be in accordance with development 'from an omnivore into a carnivore'. (See remarks on p. 61.) Suslowska (1971) again compared Aspius with Cyprinus and Esox. She comments that the form of the levator arcus palatini in Aspius closely resembles that of Cyprinus. However, there is a difference in the orientation of the fibres, those of Aspius being more horizontally directed (as in Luciobrama, see p. 21). Suslowska (op. cit.) also demonstrated a close resemblance between the arrangement of the adductor hyomandibulae in Aspius and Esox. Woskoboinikoff (1932), Yeremeyeva (1950), Alexander (1964), Suslowska (1971) and Howes (1976) all draw attention to the form of the dilatator operculi and note that this muscle is developed to the greatest degree in those fishes with elongate heads in which the operculum has become greatly extended. The need for strong dilatation of the opercula is two-fold, providing an increased flow of water for respiration and enabling the prey to be moved into the buccal cavity. In Luciobrama, the large operculum and long dilatator operculi may not necessarily be indicative of their principal role in the feeding action. The movement of the operculum at its articulation with the hyomandibula seems to be restricted and it seems likely that the muscle's action is concerned primarily with strong breathing movements (see Alexander, 1969). The enormous elongation of the postorbital skull region in Luciobrama and the correlated development of both the pterygoid series and the levator arcus palatini suggest, as a consequence, a powerful abduction of the hyopalatine arch, which would provide a suction corridor and thus reinforce the feeding action. Summary The following characters in Luciobrama are considered to be specialized (apomorph): 1. Reduced and elongate fourth infraorbital, diverted across the postorbital face (pp. 7 & 26). 2. Elongate postorbital region (including lengthened and narrowed parietals) (pp. 1 1 & 35). 3. Long tubular nasals, bearing 9-10 pores (pp. 9 & 31). 4. Orbitosphenoid and pterosphenoid bearing posterior and anterior extensions respectively (pp. 9 & 31). 5. Autosphenotic underlying the frontal as a shelf (pp. 13 & 35). 6. Posttemporal fossa extending well forward (pp. 13 & 38). 7. Extensive postparietal platform (pp. 13 & 40). 8. Specialized form of the lower jaw and large number of pores in the mandibular lateral line canal (pp. 13 & 42). 9. Short, inclined hyomandibula (pp. 13 & 44). 10. Elongate pterygoids (p. 15 & 45). THE ANATOMY OF LUCIOBRAMA MACROCEPHALUS (LACEPEDE) 59 11. Extensive opercular series (pp. 15 & 47). 12. Second neural plate contacting the cranium (pp. 17 & 48). 13. Total number of vertebrae 50 or more (pp. 20 & 51). 14. Enlarged and complex levator arcus palatini muscles (pp. 21 & 53). 15. Scales minute and numerous (p. 7). Of these characters only 4 is confined to Luciobrama. Characters 2 and 10 are shared only with Aspiolucius. Characters 8 and 9 are shared with Aspiolucius and Pseudaspius. All the remaining characters are shared with Aspiolucius, Pseudaspius and Aspius. Relationships of Luciobrama Bleeker (1870) suggested that Luciobrama was near to Aspius. Since that perceptive statement was made no other worker has speculated upon the relationships of this genus. Berg (1964) thought that Pseudaspius was close to Leuciscus but implied that Aspiolucius and Aspius were related. Luciobrama has been placed in the Leuciscinae by Rendahl (1928), Chu (1935) and Lin (1935), and in the Cyprininae by Tchang (1933). Aspiolucius, Pseudaspius and Aspius are placed in the Leuciscinae by Chu (1935) and Nikolsky (1954). From the summary of synapomorph characters presented above (p. 58) it would appear that Luciobrama is, as Bleeker (1870) supposed, closely related to Aspius, but with even closer ties to Aspiolucius and Pseudaspius. I regard Luciobrama and Aspiolucius as a sister group and Pseudaspius as the closest relative of those two genera. In turn, all three are the sister group to Aspius which, by virtue of the morphology of the jaw and relatively unmodified cranium, I consider to be the least specialized representative of this group of genera. The relationships of the aspinine group of genera are discussed below, page 61. Relationships of Elopichthys Elopichthys poses a difficult problem concerning relationships and I have included it here because I believe it belongs to the aspinine assemblage. Gosline (1974 : 12) stated '. . . Elopichthys (with Ochetobius) and Hypophthalmichthys (with Aristichthys) seem to represent highly specialized cyprinid groups without close relatives'. From this statement I assume that Gosline is considering Ochetobius to be related to Elopichthys. I have examined specimens of Ochetobius elongatus, but apart from an enlargement of the levator arcus palatini muscle and the numerous vertebrae (see p. 51) I am unable to find any specializations that would suggest close affinity with Elopichthys. The upper jaw of Ochetobius appears to be highly protractile, the premaxillae possess long anterior ascending processes, there is no expansion of the pterosphenoid and the lower jaw is short and deep. All these features represent marked differences between Elopichthys and Ochetobius. I believe Ochetobius to be a specialized leuciscine. I agree with Gosline (1974) that Elopichthys is a highly specialized taxon and has diverged sufficiently from its ancestral lineage to almost 'stand alone'. Nevertheless, Elopichthys shares many derived characters with the aspinine genera (see list on p. 58). Particularly important are those characters shared only with Aspius: the lateral expansion of the pterosphenoid to the cranial borders, the close resemblance of the orbitosphenoids and the elongate supracleithrum. I consider Elopichthys to be most closely related to Aspius and thus, together, these form the sister group to Luciobrama, Aspiolucius and Pseudaspius. The aspinine group of cyprinids For the present I do not intend to assign any formal taxonomic status to the assemblage of genera considered here as the aspinine group. The reasons for this are discussed below (p. 61). 60 G- J- HOWES No single character of those enumerated below will distinguish the aspinines from other groups of cyprinids but the following combination of characters will identify this assemblage. Character Character state Barbels absent ? Plesiomorph Scales small; 65-155 in lateral line Apomorph Vertebrae 51-55 Apomorph Cranium elongate; in some cases the postorbital cranium is three times the length of the preorbital part Apomorph Sphenotic exposed as a shelf below the frontal margin Apomorph Orbitosphenoid making extensive contact with parasphenoid Apomorph Infraorbitals 2, 3, 4 and 5 narrow, the fourth elongate Apomorph Operculum antero-posteriorly extended with a long dorsal border Apomorph Posttemporal fossa present, extending well forward Presence: Plesiomorph Condition: Apomorph Pterosphenoid sometimes extended to the lateral margin of the frontal Apomorph Pterosphenoid makes extensive contact with the parasphenoid Apomorph 12 or more pores in the mandibular lateral line canal Apomorph Nasals elongate with 9-10 pores Apomorph Second neural plate contacts the cranium Apomorph Levator arcus palatini muscle extensive and complex Apomorph Dilatator operculi muscle extended Apomorph The genera and species comprising the aspinine group are: Aspius Agassiz, 1832 Aspius aspius (Linn.) 1758 Distribution: Europe Two subspecies are recognized by Berg (1964), A. aspius aspius (Europe) and A. aspius taeniatus (Caspian and Aral seas). Aspius vorax Heckel 1843 Distribution: Tigris R. Aspiolucius Berg 1907 Aspiolucius esocinus (Kessler), 1 874 Distribution: Amu-Darya Berg (1964) and Nikolsky (1954) mention a second species, A. harmandi (Sauvage) from Tonkin. This is an error. The species originally described as Gymnognathus harmandi by Sauvage (1884) is a synonym of Elopichthysbambusa (see synonymy in Lin, 1935 and Wu, 1964). Tchang (1933) placed Aspiopsis merzbacheri Zugmayer, 1921 in the genus Aspiolucius. He gave no reason for this action. I have examined the type and can find no characters which would justify inclusion in this genus. I agree with Berg (1964 : 541) in treating Aspiopsis as a synonym of Leuciscus (sensu latu). Pseudaspius Dybowski, 1869 Pseudaspius leptocephalus (Pallas), 1776 Distribution: Amur basin Luciobrama Bleeker, 1870 Luciobrama macrocephalus (Lacepede), 1803 Distribution: Southern China Luciobrama longiceps Pellegrin, 1907 Distribution: Hanoi Rendahl (1928) considered L. longiceps as possibly a subspecies or other populational variant. THE ANATOMY OF LUC1OBRAMA MACROCEPHALUS (LACEPEDE) 61 Since the species is known from only a single specimen more material from a wide range of localities will have to be available before its supposed subspecific status can be evaluated. Elopichthys Bleeker, 1859 Elopichthys bambusa (Richardson), 1844 Distribution: China Discussion Luciobrama is a highly specialized member of the aspinine group of cyprinids. Although this group can be seen as a monophyletic assemblage it is difficult to relate it to other groups; the crux of this difficulty is the fact that various classifications of the Cyprinidae have been made by attaching significance to superficial resemblances and by utilizing only single characters or a series of too few characters. In some cases the significance of these features has defied interpreta- tion. For example, Saxena & Khanna (1965) in their work on the osteology of Catla state 'It is impossible to indicate any specific features of the osteocranium as representative of primitive or evolved conditions.' Because cyprinids tend to present a uniformity in those characters previously analysed, several authors have concluded that the family cannot be divided readily into subfamilies or other well-defined groupings (Sagemahl, 1891; Regan, 1911; Ramaswami, 1955b; Hensel, 1970; Gosline, 1973). Hensel (1970) has presented a history of the classification of the Cyprinidae. It seems that certain genera were assigned to a subfamily on a purely arbitrary basis (see p. 59 concerning Luciobrama). In view of this unsatisfactory state of affairs it is not possible to place the aspinine genera in any relevant framework of related groups. Although I suspect that the aspinines can be related to the 'Leuciscines', until the Leuciscinae can be identified on the basis of shared specializations it will not be possible to say exactly how they are related. From the anatomical evidence presented in this paper it is apparent that the piscivorous facie s characterizing the aspinines is a derived condition - derived no doubt from an ancestral form possessing a reasonably protractile jaw and narrow cranium, i.e. a 'leuciscine-type' fish. One representative of this related group may be Oreoleuciscus which shares affinities with the aspinines in the morphology of the lower jaw, infraorbitals and dilatator fossa; see page 43. Some authors have considered a piscivorous or carnivorous habit to be a primitive character for the Cyprinidae (see Hubbs & Black, 1947 and Matthes, 1963). Roberts (1969) also noted this possibility in the characoids, referring specifically to Hepsetus. I take the contrary view to these authors and maintain that it is the omnivores of the respective cyprinid lineages which represent the 'primitive' type' Evidence for this point of view is presented in this paper where it is shown that the piscivorous facies are the result of derived anatomical features, or 'specializations'. (See also the remarks of Suslowska & Urbanowicz, 1957; quoted here on page 58. Several lineages of old-world cyprinids have evolved as piscivores. One, the aspinines, is considered in this paper. The predatory morphology of this group has been achieved not by any marked changes in jaw structure (apart for Elopichthys which is considered below), but by an increase in length of the postorbital part of the cranium and modification of the hyopalatine arch so as to improve or modify the suction feeding method. Another, exemplified by Barilius, has evolved by the more 'usual' method of elongating the jaws but because of the particular type of protractile mechanism of the cyprinid upper jaw (involving the kinethmoid and not the median ethmoid), the jaw elements have lengthened posteriorly. This has meant a correlated deepening of the hyopalatine series coupled with a slightly increased length to the postorbital part of the cranium. Macrochirichthys represents another type in which there has also been a lengthening of the jaws but coupled with a forward movement of their articulation and modification to the anterior part of the vertebral column which has allowed an upward movement of the head (pers. obs.). In this case there has been no lengthening of the postorbital cranium. Some other genera such as Erythroculter and Schizothorax display attributes of both the aspinine type of morphology (increased postcranial length) and bariliine type (increased jaw length and deepening of the pterygoid series). 62 G. J. HOWES Only one cyprinid, the aspinine Elopichthys, has developed what may be termed the pike-like fades so characteristic of other teleost piscivores. This has been achieved by 'sacrificing' the evolutionary potential of the protrusile upper jaw. The 'early stages' of this particular evolu- tionary pathway may, however, be detected in Barilius bola. Acknowledgements I am greatly indebted to Dr P. H. Greenwood for his encouragement, advice and criticism of many earlier drafts of the manuscript. I want to express my gratitude to Dr K. E. Banister and Mr G. McGregor Reid for so many helpful discussions concerning the phylogeny of cyprinids and for providing me with much unpublished information concerning their respective studies on the genera Barbus and Labeo. Mr M. van Oyen of Leiden University provided much help by discussing functional anatomical problems as did Dr Richard Vari in discussing characoids. I am most grateful to them both. To Mr J. Chambers go my thanks for preparing so many alizarin specimens and to Mrs Margaret Clarke for assisting with radiographs and numerous other jobs. Finally, I wish to thank Dr P. J. P. Whitehead for his helpful advice and criticism of the final draft of the manuscript. References Alexander, R. McN. 1964. Adaptation in the skull and cranial muscles of South American characinoid fish. J. Linn. Soc. (Zoo/.) 45 : 169-190. 1966. The functions and mechanisms of the protrusible upper jaws of two species of cyprinid fish. /. Zoo/. Land. 149 : 288-296. 1967. Functional design in fishes. London. 160pp. 1969. Mechanics of the feeding action of a cyprinid fish. /. Zoo/. Lond. 159 : 1-15. Banarescu, P. 1967. Studies on the systematics of Cultrinae (Pisces, Cyprinidae) with description of a new genus. Revue roum. Biol. (Zoo/.) 12 (5) : 297-302. 1968. Remarks on the genus Chela Hamilton-Buchanan (Pisces, Cyprinidae) with description of a new subgenus. Annali Mus. civ. Stor. nat. Giacomo Doria 77 : 53-64. Berg, L. S. 1964. Freshwater fishes of the USSR and adjacent countries. 2. (English translation of the fourth ed. of 1949.) 496 pp. Bleeker, P. 1870. Nieuw vischsoorten van China. Versl. Meded. K. Akad. wet. Amst. 4 : 251-256. 1873. Notice sur le Synodus macrocephalus Lac. (Luciobrama typus Blkr). Ned. Tijdschr. Dierk 4 : 89. Bridge, T. W. 1896. The mesial fins of ganoids and teleosts. /. Linn. Soc. (Zoo/.) 25 : 530-602. Chu, Y. T. 1935. Comparative studies on the scales and on the pharyngeals and their teeth in Chinese cyprinids, with particular reference to taxonomy and evolution. Biol. Bull. St. John's Univ., Shanghai 2 : 1-225. Gosline, W. A. 1961. Some osteological features of modern lower teleostean fishes. Smithson. Misc. Coll. 142 (3) : 1-142. 1973. Considerations regarding the phylogeny of cypriniform fishes, with special reference to structures associated with feeding. Copeia, 1973 (4) : 761-776. 1974. Certain lateral-line canals of the head in cyprinid fishes, with particular reference to the derivation of North American forms. Jap. J. Ichth. 21 (1) : 9-15. 1975. The cyprinid dermosphenotic and the subfamily Rasborinae. Occ. Pap. Mus. Zoo/. Univ. Mich. 673 : 1-13. Greenwood, P. H. 1974. The cichlid fishes of Lake Victoria, East Africa: the biology and evolution of a species flock. Bull. Brit. Mus. nat. Hist. (Zool.), Suppl. 6 : 1-134. Greenwood, P. H., Rosen, D. E., Weitzman, S. H. & Myers, G. S. 1966. Phyletic studies of teleostean fishes, with a provisional classification of living forms. Bull. Am. Mus. nat. Hist. 131 : 339-456. Harrington, R. W. 1955. The osteocranium of the American cyprinid fish, Notropis bifrenatus with an annotated synonymy of teleost skull bones. Copeia, 1955 (4) : 267-290. Hensel, K. 1970. Review of the classification and the opinions on the evolution of Cyprinoidei (Eventognathi) with an annotated list of genera and subgenera described since 1921 . Annotnes Zool. Bot. Bratislava, 57 : 1-45. THE ANATOMY OF LUCIOBRAMA MACROCEPHALVS (LACEPEDE) 53 Howes, G. J. 1976. The cranial musculature and taxonomy of characoid fishes of the tribes Cynodontini and Characini. Bull. Brit. Mus. nat. Hist. (Zool.) 29 (4) : 203-248. Hubbs, C. L. & Black, J. D. 1947. Revision of Ceratichthys, a genus of American cyprinid fishes. Misc. Publs. Mus. Zool. Univ. Mich. 66 : 1-56. Jollie, M. 1975. Development of the head skeleton and pectoral girdle in Esox. J. Morph. 147 (1) : 61-88. Kershaw, D. R. 1976. A structural and functional interpretation of the cranial anatomy in relation to the feeding of osteoglossoid fishes and a consideration of their phylogeny. Trans, zool. Soc. Lond. 33 : 173- 252. Kimura, S. 1934. Description of the fishes collected from the Yangtse-Kiang, China by the late Dr K. Kishinouye and his party in 1927-1929. /. Shanghai Sci. Inst. 1 (3) : 11-247. Liem, K. F. 1970. The comparative anatomy of the Nandidae (Pisces : Teleostei). Fieldiana Zool. 56 : 1- 166. Lin, S. Y. 1935. Contribution to the study of cyprinidae of Kwangtung and adjacent provinces. Lingnan Sci. J. 14 (4) : 651-663. Lindsey, C. C. 1975. Pleomerism, the widespread tendency among related fish species for vertebral number to be correlated with maximum body length. /. Fish Res. Bd. Can. 32 (12) : 2453-2469. Matthes, H. 1963. A comparative study of the feeding mechanism of some African Cyprinidae (Pisces, Cypriniformes). Bijdr. Dierk. 33 : 3-35. Meinel, W. Von, Holl, A. & Schulte, E. 1970. Vergleichend-anatomische und funktionsanalytische Unter- suchungen an der splanchnischen Muskulatur einiger Cyprinidae (Teleostei). Zool. Beit. 16 : 93-128. Munshi, J. S., Datta. & Singh, B. R. 1967. The cranial muscles and the natural mechanism of opening and closing of mouth in two Indian major carps. Zool. Anz. 178 (1-2) : 49-60. Nelson, G. J. 1969. Gill arches and the phylogeny of fishes, with notes on the classification of vertebrates. Bull. Am. Mus. nat. Hist. 141 art. 4 : 479-552. Nichols, J. T. 1925. Chinese freshwater fishes. Nat. Hist. N. Y. 25 (4) : 346-352. 1943. Freshwater fishes of China. Nat. Hist, of Central Asia. 9. American Museum of Natural History, New York. 322 pp. Nikolsky, G. V. 1954. Special Ichthyology. 2nd ed. Moscow. 458 pp. Oliva, O. & Skorepa, V. 1968. The myodome in carp (Cyprinus carpio Linnaeus) and Repfen (Aspius aspius Linnaeus). Vest. csl. Zool. Spol. 32 (1) : 39-45. Patterson, C. 1975. The braincase of pholidophorid and leptolepid fishes, with a review of the actino- pterygian braincase. Phil. Trans. Roy. Soc. B 269 (899) : 275-579. Ping, C., Pao, S. & Yang, H. Y. 1958. The skeletal musculature of the carp (Cyprinus carpio). Acta zool. sin. 10 : 289-315. Ramaswami, L. S. 1955a. Skeleton of cyprinoid fishes in relation to phylogenetic studies: 6. The skull and Weberian apparatus in the subfamily Gobioninae (Cyprinidae). Acta Zool. 36 (2) : 127-158. 1955b. Skeleton of cyprinoid fishes in relation to phylogenetic studies: 7. The skull and Weberian apparatus of the Cyprininae (Cyprinidae). Acta Zool. 36 (3) : 199-242. Rauther, M. 1950. Ober Pelecus cultratus (L.) in Vergleich mit anderen 'Beilfischen'. Zool. Anz. 145 : 794- 804. Regan, C. T. 1911. The classification of the teleostean fishes of the order Ostariophysi. 1. Cyprinoidei. Ann. Mag. nat. Hist. (8) 8 : 13-32. Rendahl, H. 1928. Beitrage zur kenntnis der Chinesischen siiswasserfische. 1. Systematischer Teil. Ark. Zool. 20A (1) : 1-194. Roberts, T. R. 1969. Osteology and relationships of characoid fishes, particularly the genera Hepsetus, Salminus, Hoplias, Ctenolucius and Acestrorhynchus. Proc. Calif. Acad. Sci. (4) 36 : 391-500. 1973. Interrelationships of ostariophysans. In: Interrelationships of fishes, Greenwood, P. H., Miles, R. S. & Patterson, C. (Eds) : 373-395. London and New York. Sagemahl, M. 1891. Beitrage zur vergleichenden Anatomic der Fische, IV. Das Cranium der Cyprinoiden. Morph. Jahrb. 17 : 489-495. Sauvage, H. E. 1884. Contribution a la faune ichthyologique du Tonkin. Bull. Soc. zool. Fr. 9 : 209-215. Saxena, S. C. 1960. The cranial musculature of a hill-stream cyprinid fish Garra mullya (Sykes). Proc. Nat. Inst. Sci. India B (4) 26 : 176-188. Saxena, D. B. & Khanna, S. 1965. Osteology of the freshwater teleost Catla catla (Ham.). Part I. Osteo- cranium. Vest. csl. Zool. Spol. 29 : 127-145. Shukla, G. R. & Verma, S. R. 1972. Appendicular skeleton of Barilius bola (Ham.). (With the remark of phylogenetic consideration.) Anat. Anz. 130 : 560-570. Sorescu, C. 1968. Vergleichende Untersuchungen iiber den Schultergiirtel der Cyprinidae. Senck. biol. 49 (5) : 387-397. 64 G. J. HOWES 1972. Comparative studies on the Weberian apparatus in the subfamily Danioninae and Cultrinae (Cyprinidae - Pisces). Revue roum. Biol. (Zool.). 17 (6) : 391-398. Starks, E. C. 1926. Bones of the ethmoid region of the fish skull. Stanf. Univ. Public. Biol. Sci. 4 (3) : 139- 338. Suslowska, W. 1971 . A comparative study of the muscles of the head of Aspius aspius (Linnaeus, 1758) and Esox lucius Linnaeus, 1758. Acta Zool. Cracov. 16 (16) : 695-713. & Urbanowicz, K. 1957. Badania prownaweze nad wybranymi elementami szkieletu szezupala - Esox lucius L., bolenia - Aspius aspius (L.) i karpia - Cyprinus carpio L. Zesz. nauk. Univ. Lodz (2) 3 : 71-93. Takahasi, N. 1925. On the homology of the cranial muscles of the cypriniform fishes. /. Morph. 40 : 1-109. Tchang, T. L. 1933. The study of Chinese cyprinoid fishes, Part 1. Zool. Sinica 2 (1) : 1-247. Tretiakov, D. K. 1946. Systematic groups of Cyprinidae. Zool. Zh. 25 (2) : 149-156. Weitzman, S. H. 1962. The osteology of Brycon meeki, a generalized characin fish, with an osteological definition of the family. Stanford ichthyol Bull. 8 (1) : 3-77. Wiley, M. L. & Collette, B. B. 1970. Breeding tubercles and contact organs in fishes: their occurrence, structure and significance. Bull. Am. Mus. nat. Hist. 143 art. 3 : 143-216. Winterbottom, R. 1974. A descriptive synonymy of the striated muscles of the teleostei. Proc. Acad. nat. Sci. Philad. 125 (12) : 225-317. Woskoboinikoff, M. M. 1932. Der Apparat der Kiemenatmung bei den Fischen. Zool. Jb. Anat. Jena 55:315^88. Wu, Hsien-Wen. 1964. Cyprinid fishes of China. 1. Shanghai. 228 pp. Yeremeyeva, E. F. 1950. Zavistimost miezdu stroieniem cherepa i stroiemiem rotovogo i golotochnogo apparatov kostistich ryb. Trudy Inst. Morf. Zhivot. 3 : 34-41. British Museum (Natural History) Monographs & Handbooks The Museum publishes some 10-12 new titles each year on subjects including zoology, botany, palaeontology and mineralogy. Besides being important reference works, many, particularly among the handbooks, are useful for courses and students' background reading. Lists are available free on request to : Publications Sales British Museum (Natural History) Cromwell Road London SW7 5BD Subscriptions placed by educational institutions earn a discount of 10% off our published price. Titles to be published in Volume 34 The anatomy and relationships of the cyprinid fish Luciobrama macrocephalus (Lacepede). By G. J. Howes. A new species of Pteralopex Thomas, 1 888 (Chiroptera : Pteropodidae) from the Fiji Islands. By J. E. Hill & W. N. Beckon. A revision of the species of Sertulariidae (Coelenterata : Hydroida) recorded from Britain and nearby seas. By P. F. S. Cornelius. Anatomical specimens of birds in the collection of the British Museum (Natural History). By J. S. Blandamer & P. J. K. Burton. Amphibians and reptiles from northern Trengganu, Malaysia, with descriptions of two new geckos : Cnemaspis and Crytodactylus. By J. M. C. Dring. A revision of the spider genus Portia (Araneae : Salticidae). By F. R. Wanless. Type set by John Wright & Sons Ltd, Bristol and Printed by Henry Ling Ltd, Dorchester ulletin of the British Museum (Natural History) A new species of Pteralopex 888 (Chiroptera: he Fiji Islands Hill & W, Zoology series Vol 34 No 2 30 November 1978 itish Museum (natural History), instituted in 1949, • .< rie Botany, Entomology, Geology and Zoology, and an Historical series, ure published at irregular intervals as they become ready. Volumes will contain about nundred pages, and will not necessarily be completed within one calendar year. Subscription orders and enquiries about back issues should be sent to : Publications Sales, ish Museum (Natural History), Cromwell Road, London SW7 5BD, England. Worl (>reviation: Bull Br. Mus. nat. Hist, (Zool.) of the British Museum (Natural History), 1978 ISSN 0007-1498 British Museum (Natural History) Cromwell Road London SW7 5BD Zoology series Vol 34 No 2 pp 65-82 Issued 30 November 1978 A new species of Pteralopex Thomas, 1888 (Chiroptera : Pteropodidae) from the Fiji Islands J. E. Hill Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD W. N. Beckon Peace Corps, Suva, Fiji Contents Synopsis ............. 65 Introduction ............ 65 Systematic descriptions ........... 66 Genus Pteralopex Thomas, 1888 66 Pteralopex acrodonta sp. nov. ......... 68 Relationships ............ 75 Dental homologies and dental evolution ........ 76 Acknowledgements ........... 81 References ............. 81 Synopsis The unusual megachiropteran genus Pteralopex is briefly reviewed and considered to include three species, two named many years ago from the Solomon Islands, and a third, here described as new, from the Fiji Islands, whence the genus is recorded for the first time. Cuspidation of the molariform teeth in Pteralopex is considered in detail, with especial reference to the new species, and compared with the similarly cuspi- date condition of the molariform teeth of Harpy ionycter is. Presumed homologies with the cusps of the dilambdodont teeth of the Microchiroptera are examined, with a discussion of the relevance of molariform cuspidation in the Megachiroptera to theories of their dental evolution. Introduction The known megachiropteran fauna of the Fiji Islands has been limited hitherto to two species of the widespread genus Pteropus, one the Pacific fruit bat P. samoensis, the other the Polynesian fruit bat P. tonganus, and to the long-tailed fruit bat Notopteris macdonaldi. None is endemic: P. samoensis is represented on the Fiji Islands by an endemic subspecies, P. s. nawaiensis, the other, nominate subspecies occurring in the Samoa Islands (Wodzicki & Felten, 1975), while P. tonganus is more widely distributed, its subspecies occurring variously from Dampier (Karkar) Island, off the northeastern coast of New Guinea and from Rennell Island, in the Solomon Islands, eastward to the Tonga and Samoa Islands and to Niue Island. This species may occur or have occurred even further to the east, in the Cook Islands, whence bats, apparently pteropodids, have been reported (Smith, 1902, Krzanowski, 1977 : 271) from Raratonga Island (21° 14' S, 159° 46' W) and where there is evidence of bats (Gill, 1876, Krzanowski, 1977 : 271) on Mangaia Island (21° 55' S, 157° 55' W). The subspecies of P. tonganus on the Fiji Islands, P. t. tonganus, occurs also in the Tonga and Samoa groups to the east, but to the west is replaced by P. t. geddiei on the Loyalty Islands, New Caledonia and the New Hebrides. The representatives of Pteropus on the Fiji Islands thus display a closer affinity to those of the more easterly islands than to their congeners on the islands to the west of the Fiji group. Notopteris macdonaldi has a rather different distributional pattern : one subspecies, N. m. macdonaldi, occurs on the Fiji Islands and in the New Hebrides (specimens reported from Ponape, Caroline Islands by Jentink (1887 : 268, 1888 : 158) are referred to this subspecies by Andersen (1912 : 798) who also (p. Ixxiv) queries Bull. Br. Mus. nat. Hist. (Zool.) 34 (2) : 65-82 Issued 30 November 1978 65 66 J. E. HILL & W. N. BECKON the veracity of the record) with a second subspecies, N. m. neocaledonica, in New Caledonia. The microchiropteran fauna of the Fiji Islands is similarly sparse with records only of Embal- lonura semicaudata and Tadarida jobensis. As among the Megachiroptera there is no endemic species : E. s. semicaudata occurs also in the Tonga and Samoa groups and in the New Hebrides, with a second subspecies in the Palau Islands, while the distribution of T. jobensis extends from the Fiji Islands to New Guinea and Australia. Specimens of the latter species from the Fiji Islands seem likely to represent T. j. bregullae, described originally from the New Hebrides (Felten, 1964a : 12). The islands to the east of the Fiji group have fewer species, with Pteropus tonganus and Embal- lonura semicaudata in the Tonga Islands, occurring together with Pteropus samoensis in the Samoa Islands. To the west of Fiji, however, the islands and island groups have a more diverse bat fauna. Apart from Notopteris macdonaldi on New Caledonia and the New Hebrides, the Megachiroptera are represented by several species of Pteropus, all except P. tonganus usually restricted to particular islands or groups of islands. Pteropus ornatus occurs on the Loyalty Islands and New Caledonia, on the latter island with P. vetulus (Felten, 1964c); P. anetianus, its several subspecies (Felten, 19646, Felten & Kock, 1972) and P.fundatus occur on the New Hebrides, while P. tuberculatus, P. vanikorensis and P. nitendiensis are found on various of the Santa Cruz Islands. By contrast, the Microchiroptera of these islands and island groups consist chiefly of species with a much wider general distribution, although one, Miniopterus robustior, is known only from the Loyalty Islands. Most are Australasian : Minopterus australis and M. medius extend to the Loyalty Islands and New Caledonia, Hipposideros galeritus, Aselliscus tricuspidatus, Myotis adversus and Miniopterus tristis to the New Hebrides. Emballonura semicaudata, widely distributed among the islands of the Pacific to the east, occurs also on the New Hebrides and Chalinolobus gouldi, otherwise an Aus- tralian species, is represented on New Caledonia (Koopman, 1971 : 4). The bat species so far recorded from the Fiji Islands are quite clearly consistent with the position of the group in the long island chain that stretches eastward from New Guinea into the Pacific Ocean, the Fiji Islands lying near the eastern limit of bat distribution along the chain and somewhat widely separated from their nearest neighbours to the west. Although four of the five bat species hitherto known from the Fiji Islands occur also on the islands to the west, the absence from the Fijian fauna of the majority of the widespread microchiropteran species that extend eastward to the New Hebrides or even to New Caledonia and the Loyalty Islands suggests that for many the relatively wide oceanic strait between these islands and the Fiji group is an effective barrier to further dispersal. Of bat species found both east and west of this obstacle, only one megachiropteran and one microchiropteran occur further west than New Caledonia and the New Hebrides. Furthermore, differentiation is unusual among Fijian bats, until now there being no endemic species reported from this island group, and but one endemic subspecies. In these circumstances it is of particular interest to report the existence in the Fiji Islands of a hitherto undescribed and very distinct species of the aberrant genus Pteralopex, itself previously known only from the Solomon Islands. Systematic descriptions Genus PTERALOPEX Thomas, 1888 Pteralopex Thomas, 1888 : 155, 1889 : 473, pi. 20, fig. 3, pi. 21, figs 4-7. Pteralopex atrata Thomas, 1888. The genus Pteralopex is characterized externally by the insertion of the wings at or near the mid-line of the back along the spinal line; cranially by an unusually well-developed sagittal crest, long postorbital processes which reach or nearly reach the zygomata, rather upwardly directed orbits and short, broad, nearly parallel-sided rostrum, but especially dentally by massive upper canines which have a prominent posterior supplementary cusp and by the exceptional size of the outer lower incisors (i2_2)-* The molariform teeth of Pteralopex, moreover, are variously cuspi- date, on occasion displaying an extreme of the tendency towards lateral cuspidation seen in various * The dental notation adopted in this paper is that of Miller (1907). NEW SPECIES OF PTERALOPEX 67 ways and to varying degrees in some species of the related genus Pteropus, in Hypsignathus, Nyctimene, Paranyctimene, Dobsonia or in Harpyionycteris, although surface cusps or ridges such as occur in some of Cynopterus, in Ptenochirus, Latidens, Dyacopterus, Thoopterus, Dobsonia or in Harpyionycteris are lacking. 21 32 The dental formula of Pteralopex is i -, c -, pm -, m - = 34, as in Pteropus; the upper incisors M A •? .3 (i2~3) have very broad posterior ledges and the upper canines are short, very thick antero-pos- teriorly, with a large and prominent posterior secondary cusp extending halfway along the length of the tooth and a wide internal cingulum bearing small internal and postero-internal basal cusps, the anterior cusp sometimes indistinct. The first upper premolar (pm2) is rudimentary, its crown only slightly differentiated; the second (pm3) and third (pm4) upper premolars and the first upper molar (m1) have prominent, shelf-like raised anterior and posterior basal ledges, the labial lateral elevation in these teeth raised into a cuspidate structure, the lingual elevation usually similarly so but on occasion more ridge-like, while the second upper molar (m2) lacks the anterior basal ledge and the cuspidate appearance is usually less evident. The inner lower incisor (ix) is very small, subterete, its edge slightly lobed, the outer lower incisor (i2) much enlarged to twelve or fifteen times the bulk of i1? its posterior ledge very long antero-posteriorly, the longitudinal diameter of the tooth greater than the transverse diameter of its crown, the cutting edge tricuspid or incipiently so. The lower canine is relatively short and stout, its cingulum with generally a small raised postero-external tubercle; the first lower premolar (pm2) subequal in crown area to i2, with similar broad inner ledge and tricuspid cutting edge, the central cusp the largest; the second lower premolar (pm3) has a large main cusp, sometimes with a subsidiary anterior cusp, a short posterior basal ledge forming labially a small posterior basal cusp separated from the main cusp by a notch. The third lower premolar (pm4) and the first (nij) and second (m2) lower molars are short and broad, with strong posterior basal ledges more developed lingually than labially and thus oblique, and their lateral elevations are variously cuspidate to differing degrees, the lingual elevation on occasion more ridge-like; third lower molar (m3) usually subcircular, with a concave crushing surface, generally with a shallow notch in its labial margin, but sometimes may be more definitely cuspidate, its crown pattern more nearly similar to that of m2. A comprehensive review of Pteralopex as then understood was provided by Andersen (1912 : 432), who had earlier (1909a : 213) studied its affinities in considerable detail. The genus has hitherto included two named forms, P. atrata Thomas, 1888 (the type species) and P. anceps Andersen, 19096, the former described from Guadalcanar Island, in the eastern Solomon Islands and later reported (Sanborn, 1931 : 21) from the nearby island of Ysabel, the latter apparently known only from Bougainville Island and from Choiseul Island (Phillips, 1968 : 792), in the more westerly part of the Solomons group. Andersen (1912 : 437) considered the two to be distinct species but Laurie & Hill (1954 : 40) and Phillips (1968 : 790) considered anceps a subspecies of P. atrata. There is much, however, to commend the original arrangement. The ears of the larger anceps have a trace of a blunt tip and although for the most part its pelage is blackish or seal brown, the hairs on the posterior part of the ventral surface are tipped with drab brown. The fur is long and rather woolly, extending to the dorsal surface of the tibia and of the metatarsals, which dorsally are densely haired. In contrast, the ears of atrata are more rounded and its pelage uniformly blackish or dark seal brown, rather short and not especially woolly, the fur not extending dorsally to the distal end of the tibia (the last fourth is naked) or to the metatarsals. Although many dental features such as the enlargement of i2~3 and of i2, and the presence of a secondary canine cusp are shared in equal measure by anceps and atrata, in others there is con- siderable divergence. In particular, the anterior basal ledges of pm4 and m1, and to a lesser extent of pm3 are less developed in anceps than in atrata, and the lingual margins of pm4, mt and m2 are more ridge-like and less cusp-like. According to Andersen (1912 : 437, 438, fig. 22) the anterior basal ledge of pm4 in anceps does not extend internally on to the inner surface of the lingual cusp as it does in atrata, but the lingual faces of pm4~4 have been damaged in the subadult holotype and only available specimen of anceps, and the observation cannot now be confirmed from this example: according to Phillips (1968:792) the anterior basal ledge of pm4 in adult anceps 68 J. E. HILL & W. N. BECKON extends to the labial surface. Certainly the anterior basal ledges of pm3 and m1 in anceps extend much less obviously on to the inner face of the lingual cusp than in the corresponding teeth of atrata. In anceps the lingual cusp of pm4 is long and slightly ridge-like but in atrata the correspond- ing cusp is much more nearly conical: this contrast is more pronounced in m1 which in anceps has a long, ridge-like lingual elevation, scarecely separated from the raised anterior rim of the tooth, but in atrata has a prominent, sharply defined antero-internal cusp, or in m2 where the similarly long, rather low, ridge-like lingual elevation of anceps differs from the short, slightly raised corresponding elevation of atrata. Furthermore, the labial elevations of pm4, n^ and m2 in anceps are much less definitely divided into two cusps than in atrata, the dividing fissure in pm4 and mt scarcely reaching the labial face of the teeth and barely perceptible in m2. In general terms the molariform teeth of anceps are nearer in appearance to those of many species ofPteropus than are those of atrata : the external and dental differences between anceps and atrata suggest that the two forms must be considered specifically distinct. Key to species of Pteralopex 1 Smaller (length of forearm 11 6-1 20 mm); pelage brown overall; labial margin (excluding raised posterior ledge) of pm4 and m1 divided into three cusps, the anteriormost very small ; pm3 lacking any lingual cusp; crowns of m! and m2 closely similar . . P. acrodonta sp. nov. Larger (length of forearm 139-171 mm); pelage predominantly black or blackish seal brown, especially dorsally; labial margin (excluding raised posterior ledge) of pm4 and m1 forming a single large cusp; pm3 with prominent lingual cusp; crowns of ir^ and m2 dissimilar . 2 2 Smaller (length of forearm 139-144 mm); ventral pelage uniformly black or blackish seal brown; fur short, not especially woolly, extending to entire dorsal surface of tibia and to dorsal surface of metatarsals ; pm4 with subconical lingual cusp; lingual elevations of ml and m2 short, in ir^ forming a sharply defined subconical antero-internal cusp . . P. atrata Larger (length of forearm 160-171 mm); ventral pelage black or blackish seal brown anteriorly, brownish posteriorly; fur long, woolly, not extending dorsally to distal end of tibia or to dorsal surface of metatarsals; pm4, mx and m2 with lengthened or long ridge-like lingual elevations ............. P- anceps Pteralopex acrodonta sp. nov. HOLOTYPE. BM(NH) 77.3097. Adult £. Ridge about 300 m NE of the Des Voeux Peak Radio Telephone Antenna Tower, Taveuni Island, Fiji Islands, 16° 50£' S, 179° 58' W, c. 3840ft (1170 m). Collected 3 May 1977 by W. N. Beckon, died 6-7 May 1977. Caught in mist net on ridge summit : bulldozed land with secondary scrubby growth, adjacent to primary forest. Original number 104. Skin and skull. OTHER MATERIAL. No. 101. Adult ?. Des Voeux Peak, Taveuni Island, c. 3900 ft (1190 m). Col- lected 9 November 1976 by W. N. Beckon, died 12 November 1976. Caught in mist net on ridge summit. Skin and skull. To be deposited in the Fiji Museum. DIAGNOSIS. Smaller externally (Table 1) than either Pteralopex anceps or P. atrata, differing sharply from these in overall brown rather than black or blackish seal brown pelage; skull smaller (Table 1) than in either of the related species, but with relatively wider interorbital region and rela- tively more massive zygomata which have a more pronounced upward deflection. Teeth, excepting m2 and m3, smaller (Table 2) than in P. anceps or P. atrata, pm4, m1"2 and m^g more cuspidate, (Fig. 1) the labial margin (excluding raised posterior ledge) of pm4 and m1 divided into three cusps rather than elevated into a single cusp as in these species; m2 little reduced, similar in size to m2 of P. anceps or P. atrata; pm3 lacking the prominent internal cusp of pm3 in P. anceps or P. atrata, its internal vertical ridge merging smoothly into the tip of the tooth; ml and m2 closely resembling each other, not markedly dissimilar in size and cuspidation as in P. anceps and P. atrata; crown of m3 less basin-like than in the related species, its cusp pattern similar to that of m^ and m2. DESCRIPTION. Ears small, short, almost concealed by surrounding fur; upper margin of ear semicircular as in Pteralopex atrata, with no indication of any tip; outer surface of conch nearly naked, a few sparse long hairs on its inner surface, clustered a little more thickly near the anterior NEW SPECIES OF PTERALOPEX 69 6 « (HM)PM36 2 IW68 (HN)WS (HM)PMaP 2 ''I l 101 sj ifij '- (HM)WaP - OS .0 ^ *7« co co 00 ^ — oo •o «o o"l «0 O< — 1 ^H O4 04 CO IO 04 04 OS CO t — OS ooior-'^co «of~-^^Tj-Os so 04 ~ s 0) m Os 04 rt CO OS 00 04 1 TJ|-o4so|oob |O°°ob|oo0 O O4 O4 •— i ^H O4'rt O4"* "^ ob 04 0 0 o Os CO oo s r- 04 OS CO OO SO CO OS S^^s;00-^^^^^ K (N O oo 04 OS o ^ 2 so CO rn OS 04 rt CO OO SO CO OS -(NSO-OOO-O^OS^^ ^ 04 OS CO r}- O4 O O 0 O ( — rn SO O SO CO OO ^O CO OO O4o4t~~SOOO^H— HOSa\vorr> ^° ob rt SO CO 00 Os 04 OS OO OO T}" OS ^" oo — Os so SO — oo O4 OO «O O4 OO O4o4t~~'OOSOOOOooi^Tt' IO fi. 3; so en ~ ^ ^04^-,^ ^,04- (N1^ «^ (N (^ oo so OS •0 <0 TJ- oo SO o Tt r- ~ ^t r- Oosvocncn«ocno\soofr> ^ 04 ^ os r~- SO 1— 1 SO •o r^i ^\ oo f^J */^ "^4 <^ Tf CO OO OS ^^ *^ SO O4 CO >O T^- c^ 1—1 10 >0 CO 04 ^^ ^ —.CO "SJ fN^ i-p^ f-H 1~H T"H ^NJ ^^ ^" 04 s £ G 0 •0 VO o 04 04 (^ uo ^ r4 «0 «0 IO SO O4 CO ^^ *^" *^ ^H OS OS *O ^* 04 so OO t — O4 r- •o CO OS F*- O4 O4 >— iOT;l"COOOOS^lOSO^H^ SO ^j- ^ ^ <0 >0 m 04 rt ^H ^H CO 0404^^ -H»-H 04"T Tt 04 O 1_ 0 o <*-< 0 cn u i- '^ X— S ^ O „ ' in o ^^ S" c S" "^ S "^ Length of forearm Total length of skull Condylobasal length £ 60 _0> 13 5 13 Pu 'o 0 c "So foramina Length front of orbit t nasals Lachrymal width Interorbital width Postorbital width Orbital diameter Zygomatic width c*3 c^u§) S 3 4) ^uC^OtH^^» O E irs^-a^^lll e-£ e?i|-Sl||| 3^-^ •0 ^ t) t) t) •*' ^ f "S NG c* S c* "^ "*&" ^2 iH rH iH 5 C G ^" 1 1 LJ 1 O ^" rt O O O Qi CH ji ^^ O ^J fji — J fi I 1 Holotype. 2 Subadult. 70 J. E. HILL & W. N. BECKON T3 •s •j sijiAUtBSnog (HN)Wa 6 j, \Sdd3llV 'J (HN)wa (HM)WS 6 (HM)W9 si ifH "I ! 101 'OM 6 si ?r!j *'i i z.6oe-/,z, (HM)wa ^ •rf 3 T3 ^'^^ & 1 1 NEW SPECIES OF PTERALOPEX 71 margin of the ear; no antitragai lobe; wings inserted on the back at or near the spinal line, and posteriorly at the base of the first and second toes; second phalange of fourth digit subequal in length to first phalange, not conspicuously longer; uropatagium or interfemoral membrane a very narrow flange; calcar short. Pelage brown overall, dorsally with the hairs mid-brown at the base and for much of their length, over the head, mantle and rump tipped with paler shining brown I I 5mm Fig. 1 Pteralopex acrodonta. Holotype c? BM(NH) 77.3097. a. Labial aspect of left toothrows. b. Ventral aspect of left upper toothrow. c. Dorsal aspect of left lower toothrow. d. Lingual aspect of right molariform teeth. to give a slightly bronzed appearance, paler hair tips on rump a little less evident than those of mantle; hairs in mid-dorsal region immediately behind mantle bright brown for their entire length, smooth, closely adpressed, forming a longitudinal band about 18-20 mm in width over the inner- most part of the wing membranes at their insertion along the mid-line of the body; ventral pelage a drab brown, rather paler than the dorsal pelage, on the lower part of the neck and on the flanks with paler light brown tips. The colour of the new species contrasts sharply with the black coloration of P. atrata or with the black dorsal surface of the head and body in P. anceps; only in 72 J. E. HILL & W. N. BECKON this latter species the blackish brown colour of the underparts is relieved by drab tipping to the otherwise dark seal brown hairs over the lower part of the chest and over the belly. Fur long and woolly as in Pteralopex anceps', upper surface of forearm thickly clothed with moderately long, adpressed brownish hairs for the proximal two thirds of its length, the hair covering a little more extensive and denser near and at the elbow; tibia densely covered dorsally with thick long hair to the ankle, with a very thin scattering of moderately long, brownish hairs on the upper surface of the phalanges. Dorsal surface of wing and uropatagium or interfemoral membrane largely naked but a narrow band of quite dense, closely adpressed, long bright brown hairs extending along the junction of the mesopatagium with the proximal two thirds of the fore- arm, around the elbow and across the endopatagium to the rear of the mantle, the median longi- tudinal mid-dorsal band of long, smoothly adpressed bright brown hairs above the insertion of the endopatagium on the body extending over the proximal part of the membrane; a thin cluster- ing of moderate, blackish brown hairs near the hind margin of the endopatagium at and near its junction with the foot; ventrally a denser band of moderately long brown hairs along the proximal two thirds of the junction of the mesopatagium with the forearm; ventral surface of forearm and tibia with no more than a few very sparse hairs. Pteralopex anceps has a rather dense cover of long black hairs on the proximal third of the dorsal surface of the forearm and the dorsal surface of the tibia is densely clothed with quite long blackish chestnut brown fur, extending to the dorsal surface of the foot over the metatarsals and, more sparsely, to the dorsal surface of the phalanges; there is a band of moderate hairs on the mesopatagium both dorsally and ventrally adjacent to the lower part of the forearm, blackish above, brownish below, but the dorsal band does not extend across the surface of the endopata- gium to the rear of the mantle. In P. atrata the proximal third of the dorsal surface of the forearm and the dorsal surface of the tibia have only a very sparse cover of moderate black hairs which extend and scatter on to the dorsal surface of the foot over the metatarsals and phalanges; there is no definite band of fur on the dorsal surface of the mesopatagium or of the endopatagium, but the ventral surface of the mesopatagium has a band of brownish hairs along the proximal two thirds of its junction with the forearm. Skull similar to that of Pteralopex anceps or of P. atrata, with broad, nearly parallel-sided rostrum, but smaller; sagittal crest well developed but not especially prominent; interorbital region relatively wide; orbits less markedly upwardly directed than in its congeners; zygomata massive, deep, their upper margin forming a strongly pronounced upwardly sweeping curve; nasals terminating anteriorly at a line vertically above the rear of i3 as in P. anceps rather than above the centre of this tooth as in P. atrata. Inner upper incisor (i2) with narrow posterior shelf, less sharply demarcated from the vertical cusp than in Pteralopex anceps or P. atrata, the junction smoothly curved rather than angular; i3 with posterior shelf more prominently developed than in i2 and more sharply demarcated from the vertical cusp, the internal cingulum of the tooth slightly raised as in P. anceps and lacking the postero-internal cingulum cusp of i3 in P. atrata. Upper canine very similar to that of P. anceps or P. atrata, substantial, massively based, its antero-external face shallowly grooved, the principal cusp with sharp anterior and internal ridges and with a large, heavy posterior supporting cusp extending along about two thirds of the length of the tooth, a very small, poorly defined posterior cingulum cusp at its base; a small internal cingulum cusp at and just behind the base of the internal ridge, larger than the corresponding indistinct internal cingulum cusp of P. anceps but rather less developed than the corresponding cusp in P. atrata, in the unworn dentition followed by two small internal cusplets; postero-internal cingulum cusp small and undeveloped, in contrast to the sub- stantial postero-internal cingulum cusp of the related species, but in these there is little or no trace of any posterior cingulum cusp. Anterior upper premolar (pm2) small, terete, its circular crown very slightly larger than its shaft; pm3 with larger labial cusp and smaller lingual cusp as in Pteralopex anceps or P. atrata, its anterior basal ledge narrower than in those species but nevertheless also extending to the inner face of the tooth, a small basal supplementary postero-external cingulum cusp at rear of main labial cusp; pm4 strongly cuspidate, labially with a higher central cusp supported by a small anterior subsidiary cusp and a rather more prominent posterior subsidiary cusp, separated from the NEW SPECIES OF PTERALOPEX 73 posterior basal ledge by a small, undeveloped basal cusp, lingually with a lower but rather more massive cusp supported posteriorly by a small postero-internal basal cusp; anterior and posterior basal ledges well developed, raised, the anterior ledge extending to the inner face of the tooth as in P. atrata, the posterior ledge oblique as in P. anceps. First upper molar (m1) very similar to pm4 with its labial elevation divided into a larger central cusp with smaller anterior and posterior supporting cusps, separated from the raised posterior ledge by a very small basal cusp integral with the ledge and with large lingual cusp, an incipient subsidiary postero-internal cusp at the base of its posterior face; m2 relatively large, the tooth labially with a large anterior cusp supported posteriorly by a lower subsidiary cusp, the lingual elevation divided into two cusps, the anteriormost much the longer and higher, completely separated anteriorly from the labial elevation by a moderate fissure, the posterior basal ledge well developed, raised and cusp-like. In profile, the labial faces of pm4 and m1 present a more or less tricuspid appearance, the anteriormost cusp small and sometimes rudimentary, the second cusp much the largest, supported posteriorly by a well-developed subsidiary cusp, the basal cusp at the rear of this structure barely evident in profile, overshadowed by the large, massive elevated pos- terior basal ledge. In the same way, m2 is bicuspid in labial profile, with a large anterior cusp and smaller posterior subsidiary cusp: as in pm4 and m1, the strongly developed, elevated posterior basal ledge appears in profile to be a further posterior cusp. There is a sharp contrast in labial profile between pm4, m1 and m2 in Pteralopex acrodonta and the corresponding teeth in P. anceps and P. atrata, which present but a single large cusp, with posteriorly the prominent elevated posterior basal ledge : in P. atrata the extreme elevation of this ledge produces the appearance of a posterior cusp, especially in m1 and m2. Inner lower incisor (i2) as in Pteralopex anceps or P. atrata, very small, about one twelfth to one fifteenth the bulk of i3, its edge slightly widened, faintly and irregularly lobed, its crown triangular in cross-section; i3 much as in P. anceps or P. atrata, large, with high, rounded, chisel-like cutting edge, shallowly divided into a small inner and larger central cusp, the latter flanked at its base by a small external basal cusp, and with strong posterior shelf to give the crown a triangular outline; lower canine short and relatively massive, with narrow postero-internal shelf, not dif- fering appreciably from the lower canine in the related species. First lower premolar (pm2) similar in cross-section to i3, as it is in Pteralopex anceps or P. atrata, but a little smaller than that tooth as in P. anceps, its edge with larger central cusp, a rudimentary anterior cusp and a rather more developed posterior cusp ; pm3 faintly bicuspid in labial profile, its large labial cusp with a small anterior subsidiary cusp, the tooth lacking any trace of an internal lingual cusp, its internal ridge integral to the tip, not forming an incipient internal cusp as in P. anceps or a well-developed internal cusp as in P. atrata, the labial cusp in these species single, with no secondary or subsidiary anterior cusp; posterior basal ledge narrow as in the related species, terminating labially in a small postero-external basal cusp separated from the principal labial cusp by a distinct notch ; pm4 similar to pm4 of P. anceps or P. atrata, labially with two well developed cusps, lingually with a single large cusp, the anterior part of the crown divided completely by moderate fissures, the labial cusps divided totally as in P. atrata, not partially as in P. anceps, the posterior basal ledge broad and oblique as in these species, elevated labially into a postero-external cusp-like structure separated by a distinct notch from the posterior of the larger labial cusps, but undeveloped lingually, as in P. anceps and P. atrata. First lower molar (m^) quite different from mx in either Pteralopex anceps or P. atrata, its crown anteriorly elevated as in these species but divided anteriorly by deep longitudinal and transverse fissures into four cusps, two labial and two lingual, the anteriormost labial cusp in the unworn dentition with a slight curving of its postero-external edge which may indicate a further slight degree of cuspidation. The anterior part of the crown thus displays a condition contrasting sharply with P. anceps in which the elevated part of the crown of ml is only partially divided by shallow fissures and is rather basin-like, the lateral ridges more or less integral with the raised anterior rim, which is divided by a shallow antero-internal groove; the labial ridge has a shallow fissure just extending to its outer face and the lingual ridge is rather long, with a similarly shallow fissure in its internal face. In P. atrata the fissures are deeper than in P. anceps, creating labially two well-developed cusps and lingually a single large cusp. As in P. anceps and P. atrata, the 74 J. E. HILL & W. N. BECKON posterior basal ledge of ml in P. acrodonta is well developed, wide and oblique, elevated labially into a postero-external basal cusp but low lingually, the cusp rather less developed than in the related species. Second lower molar (m2) closely resembling ir^ but differing in the unworn dentition in having the labial elevation divided into three cusps, the large anterior labial cusp having a small sub- sidiary cusp shallowly divided from its postero-external face, the whole deeply separated from a more posteriorly placed large labial cusp, the two anterior cusps eroding with wear to a single, large flattened structure. The anterior part of the crown differs widely from the corresponding part of the crown of m2 in Pteralopex anceps from which cusps are effectively lacking, the labial ridge only faintly divided and then incompletely, the lingual ridge long and undivided, both ridges integral with the raised anterior margin to form an elevated rim round much of the anterior part of the tooth. The anterior part of the crown of m2 in P. atrata is similar to that of P. anceps, but the labial ridge is very shallowly divided by a slight fissure and the lingual ridge is relatively shorter and slightly cusp-like, faintly divided from the raised anterior margin. The posterior basal ledge of m2 in P. acrodonta resembles that of m1 and is similarly well developed, wide and oblique, elevated labially into a low basal cusp. In P. anceps the posterior basal ledge of m2 is much developed postero-externally into a large cusp-like structure not clearly separated from the labial ridge but is similarly low lingually, while in P. atrata the ledge is elevated postero-exter- nally into a distinct large cusp, separated from the main labial ridge by a deep notch. In P. anceps, therefore, the crown of m2 is basin-like, its central depression opening postero-internally but otherwise surrounded by elevated ridges : in P. atrata shallow fissures appear in the anterior and labial of these ridges, the lingual ridge is shorter and more cusp-like and the labial ridge is quite clearly separated from the small cusp that forms the labial termination of the posterior basal ledge. In P. acrodonta the basin-like pattern is absent, the tooth with a group of elevated cusps anteriorly and a broad, low posterior basal ledge that is only slightly elevated labially. Third lower molar (m3) relatively large, similar in size to m3 in Pteralopex anceps or P. atrata, the anterior cusp pattern in the unworn dentition similar to that of m2, with three labial and two lingual cusps, the anterior labial cusp large, with a small subsidiary cusp shallowly divided from its postero-external face, deeply separated from a more posteriorly situated large posterior labial cusp, and much of the crown of the tooth traversed by relatively deep longitudinal and transverse fissures; posterior basal ledge narrow and short, elevated, shallowly divided into two small unequal basal cusps, the outer smaller, the inner rather larger. The cusp pattern is greatly eroded in the worn dentition to produce two flattened cusps anteriorly, one labial, one lingual, separated by a shallow longitudinal fissure, both separated by a deeper transverse fissure from the low, cusp- like remnants of the posterior labial and lingual cusps and of the raised posterior basal ledge. The crown of m3 in P. acrodonta contrasts sharply with that of m3 in P. anceps in which no cusps are evident and which has instead very short lateral ridges, contiguous with the raised anterior margin, the posterior margin greatly developed and elevated, separated from the labial ridge by a shallow notch, the central depressed area of the tooth opening postero-internally as in m2 of that species. There are similar contrasts with the crown of m3 in P. atrata, which has very similar short lateral ridges, the lingual ridge slightly cusp-like, separated from the raised anterior margin by a faint indentation, the posterior margin elevated labially into a rounded cusp, separated from the labial ridge by a shallow notch, and the central area of the tooth opening postero-internally as in P. anceps. The anterior parts of pm4 and ml in Pteralopex acrodonta are strongly bicuspid in labial profile, with m2 and m3 more or less tricuspid in the unworn dentition when the large anterior cusp and its associated subsidiary cusp are clearly defined, but bicuspid when wear has occurred. The small external basal cusp terminating the posterior basal ledge in pm4, ml and m2 suggests an additional, low posterior cusp behind the main elevation. The lower molariform teeth of P. acrodonta differ sharply in profile from those of P. anceps in which only a slight indication of a division of the labial ridge into two cusps can be seen in pm4 and m^ is imperceptible in m2 and absent from m3. In labial profile they more nearly resemble the lower molariform teeth in P. atrata where the labial ridge in pm4 and ml is clearly bicuspid, that of m2 faintly so but m3 lacks any division of the labial ridge. The lingual profile of pm4 is similar in the three species, but P. acrodonta differs NEW SPECIES OF PTERALOPEX 75 markedly from P. anceps and P. atrata in the lingual profile of ir^ and m2. The lingual elevation of these teeth in P. acrodonta, although long, is deeply divided into two cusps but in P. anceps presents an uninterrupted ridge-like lingual elevation which is shorter and raised into a single large cusp in P. atrata. The lingual profile of m3 differs similarly. In the unworn dentition of P. acrodonta its lingual elevation is bicuspid although in the worn dentition the posterior cusp tends to erode into the posterior basal ledge : in P. anceps there is a low lingual ridge which in P. atrata is anteriorly higher and a little more cusp-like. ETYMOLOGY. The specific name is derived from aKpos, pointed, and oocbv = oSovs, tooth, in allusion to the many pointed summits of the molariform teeth in the new species. BIOLOGY. Little is known of the biology of this species : the male was not reproductively active when captured, and the female not pregnant. REMARKS. This interesting new species was discovered by the junior author and his wife, who obtained a single example on Taveuni in the latter part of 1976. They visited the island again in 1977 when two more specimens were caught, one of which escaped. It is possible that the species may be the 'white' fruit bat described by Mr Vasu Shankaran, an Indian resident of Taveuni, and known as the 'beka lulu' by the local population. This 'white' bat was reported in the lower forest of the Nasinu area, about 13| km SSW of Waiyevo, but while camped there no specimens were netted although many bats were seen circling in the forest during the early evening. Mr Shankaran remarked that this bat usually roosted in pairs in the fern clumps growing some 6-10 m from the ground on the trunks of the larger trees in the open, tall forest, leaving the clump when disturbed but flying only a short distance before landing again, unlike the other Fijian fruit bats which are more colonial and which desert their roost when an intruder approaches. This observation is supported by Mr Robin Mercer, a planter and naturalist of Savusavu in Vanua Levu, who said that the Fijians of that island use the term 'beka lulu' for a light coloured fruit bat that circles the roost instead of flying away when disturbed : he had thought such bats to be old, hoary individuals of the known species. A different pattern of erratic, manoeuvring flight, suggesting the hunting of insects, was noted in large bats over Des Voeux Peak on Taveuni. The name 'beka lulu' or 'mbeka lulu' appears in a list of Fijian names (Macdonald, 1857 : 267) collected during an expedition up the Rewa River and its tributaries in 1856 and also in the New Fijian Dictionary of Capell (1973) where it is defined as a 'species of bat'. The term apparently refers to its reputed owl-like colour and large eyes : in life the eyes of Pteralopex acrodonta, although not unusually large, are bright orange and very conspicuous. From the pattern of bird distribution in the Fiji Islands, too, it seems possible that in due course the species will be found in Yanua Levu, across the narrow Somosomo Strait from Taveuni Island. Relationships Andersen (1909« : 218) discussed the diagnostic characters and affinities of Pteralopex in detail and concluded that it was closely related to the pselaphon group of Pteropus (including P. insular is, P. phaeocephalus, P. pselaphon, P. pilosus, P. tuber culatus and P. leucopterus), some members of this group displaying to a greater or lesser extent many of those features that appear in more exaggerated form in Pteralopex anceps or in P. atrata. The newly described species adds support to this opinion. The distribution of the fur, the shape of the skull, its long postorbital processes that do not reach the zygomata, the lack of postorbital zygomatic processes, its short, broad rostrum, heavy premaxillae and its high, broad coronoid rising at about a right angle from the horizontal ramus, with broad, steeply sloping gonys are all features of Pteralopex that are fore- shadowed in the pselaphon group of Pteropus. The dentition of Pteralopex acrodonta, although more extreme and further removed from that of Pteropus than the dental structure either of Pteralopex anceps or of P. atrata, has nevertheless a number of interesting features that extend the parallels drawn by Andersen between the dental architecture of Pteralopex and that of the members of the pselaphon group of Pteropus, especially of P. pselaphon, P. pilosus, P. tuberculatus and P. leucopterus. 76 J. E. HILL & W. N. BECKON The posterior shelf of i2 is narrower in Pteralopex acrodonta than in P. anceps or P. atrata and is slightly less sharply demarcated from the vertical cusp, with a smooth, more rounded transitional area rather than a right-angled junction such as occurs in these species, and i3 is relatively smaller than in P. anceps or P. atrata, about one and one half times greater in bulk than i2 rather than two times or more its bulk as in the other species of Pteralopex. In these features j2-s of Pteralopex acrodonta resemble those of the members of the Pteropus pselaphon group and, indeed, approach the condition found in P. leucopterus. The upper canine in Pteralopex acrodonta resembles the upper canine of P. anceps or P. atrata in the structure of the principal cusp and its major subsidiary cusp, but the postero-internal basal cingulum cusp is smaller and lower than in these species. In the Pteropus pselaphon group internal cingulum cusps at the base of the canine are when present usually low and irregular, except in P. pilosus which has a large postero-internal basal cusp and a smaller internal basal cusp. The lower incisors and lower canines of the new species are characteristically those of Pteralopex. Certain features of the post-canine dentition of Pteralopex acrodonta also find a precedent among the members of the pselaphon group of Pteropus. The second lower premolar (pm3) in lacking any well-defined internal cusp in Pteralopex acrodonta differs sharply from pm3 in P. anceps and P. atrata : in the Pteropus pselaphon group, P. pselaphon, P. pilosus and P. tuber culatus have an internal shoulder on pm3 that clearly represents such a cusp but in P. leucopterus the internal ridge of pm3 merges smoothly into the summit of the tooth, as in the new species. The third lower premolar (pm4) and m^ are also of especial interest in Pteralopex acrodonta. In this species, as in Pteralopex anceps and P. atrata, these teeth are short and broad with oblique, labially more developed posterior ledge, but while pm4 in P. acrondonta is otherwise very similar to pm4 of P. anceps and more especially of P. atrata, the lingual elevation of ml and m2 (and also of m3) is divided into two cusps in contrast to the undivided lingual ridge of mx and m2 in these species. As Andersen (1909 : 221) pointed out, in the Pteropus pselaphon group it is the inner or lingual ridge of pm4 and nij that is divided in P. pselaphon, while in P. leucopterus the inner ridge of nij and m2 is faintly divided, with a lesser or scarcely perceptible division in the outer or labial ridge of the same teeth. This author commented that in Pteralopex, as it was then understood, there was a further development of a tendency already apparent in the Pteropus pselaphon group but that division had shifted, so to say, from the inner or lingual elevation to the outer or labial elevation (there is in fact a shallow internal fissure in the internal face of the inner ridge of nij in Pteralopex anceps that does not extend to the edge of the tooth), but in P. acrodonta the cuspid condition appears very strongly in both the lingual and the labial elevations of m.l and m2. The crowns of these teeth in P. acrodonta present a multicuspid appearance that finds a weak parallel in Pteropus leucopterus, albeit very much less strongly emphasized. Furthermore, in Pteralopex acrodonta the much increased degree of cuspidation extends to m2 and m3, which are less reduced than in P. anceps or P. atrata, with m2 at least three quarters rather than about one half or less the size of m1 and larger than the corresponding tooth in the related species, and m3 one half rather than one third the size of m2, about the same size as m3 of P. anceps or P. atrata. These features suggest a tendency to obtain a maximum of occlusal area in the teeth, despite the smaller size of the newly described bat when compared with its congeners. The geographical distribution of the dental characters among the three species of Pteralopex is also of interest. The least cuspidate molariform dentition, most like that of Pteropus, is to be found in the westernmost species, P. anceps. The most cuspidate dentition, least like that of Pteropus, is found in the easternmost species, P. acrodonta. A condition intermediate between these extremes occurs in the geographically central species, P. atrata. Dental homologies and dental evolution The larger of the pteropodid post-canine teeth have generally a rather characteristic appearance, consisting basically of a rectangular or more or less square crown, with a large labial and a small lingual elevation : these elevations are higher and more developed on the anterior teeth and less so in those that lie towards the rear of the toothrows, particularly in the last lower molar where they may be little more than longitudinal ridges along each side of the tooth. The elevations are NEW SPECIES OF PTERALOPEX 77 higher in the anterior part of each tooth, with an oblique crushing surface, and lower in the posterior part, the crushing surface more nearly flat. The median division between the two eleva- tions in the first of the larger teeth in each jaw is either obscured by a high, ridge-like commissure, or, often, they merge to form a single large cusp. Miller (1907 : 41) suggested that it may be assumed (from analogy with the fruit-eating phyllo- stomatids) that in the upper molars the labial of these elevations is the paracone, the lingual elevation the protocone, while in the lower molars the labial elevation is the protoconid, its lingual counterpart the metaconid. More recently, Slaughter (1970 : 77) considered further that in the upper molars the metacone has been incorporated into the labial ridge and the hypocone into the lingual ridge, while the ridges of the lower molars similarly incorporate the hypoconid labially and the entoconid lingually. Additional small subsidiary cusps are present in several of the Pteropodidae, reaching an extreme in Harpyionycteris and Pteralopex but also present to some degree in Hypsignathus, Dobsonia, Cynopterus, Ptenochirus, Dyacopterus, Thoopterus, Nyctimene and Paranyctimene, while some division of the lateral ridges is apparent in a few species of Pteropus. The view adopted by Slaughter (p. 56, figs II, 1H, p. 77) is that the small subsidiary cusp on the posterior slope of the main labial cusp of the upper molariform teeth of Dobsonia, Nyctimene and sometimes Cynopterus is a rudimentary metacone and that likewise the small subsidiary cusp on the posterior slope of the principal labial cusp of the corresponding lower teeth is a rudimentary protostylid. On this basis Slaughter suggests or infers (p. 56, fig. 1H, p. 78) that in Harpyionycteris the last upper premolar (pm4) consists labially of the paracone and meta- cone, lingually the protocone, pm4 consisting of a tall protoconid flanked postero-labially by a well-developed protostylid, and lingually by a prominent metastylid, there being no metaconid. According to this author, the upper molars in Harpyionycteris retain the paracone and protocone, the metacone lying behind the paracone, and on occasion supporting a posterior metastyle. Lingually, the posteriormost cusp is the hypocone. The lower molars have anteriorly the proto- conid and metaconid, the labial protoconid followed by a protostylid, the lingual metaconid by a metastylid : the third labial cusp is a small hypoconid, the third lingual cusp the entoconid, the stylids being as well developed as the other cusps. Support for such close homology between the multiple cusps of the molariform teeth in certain of the Megachiroptera and the cusps of the Microchiropteran dentition is lacking or contradictory. Convincing palaeontological evidence has yet to be found: such cusps in the Megachiroptera occur sporadically, antero-internally, postero-externally, laterally, or on the crown of the tooth, sometimes as a short, raised ridge. In one form or another, or in combination, they are to be found in varying degrees in several of the megachiropteran genera. Even within the species their occurrence is irregular and variable: Peterson & Fenton (1970 : 5) have pointed out that in eight examples of Harpyionycteris whiteheadi no two specimens can be said to be even close to identical in the cusp pattern of m1 and m2. In this series the cusps vary in size, position and number, with the addition of accessory cusps to the basic pattern, or with the division of primary cusps into two elements, even between the corresponding right and left teeth of individuals, to the extent that these authors considered the variability of cusp patterns in Harpyionycteris to be unique among bats and certainly ranking high among mammals. It is difficult to avoid the conclusion that such multiple cusps cannot be properly homologized with the cusps of the Microchiroptera. Thomas (1889 : 473), although admitting in an expanded description of Pteralopex that it might be thought at first sight that the genus was a highly specialized offshoot of Pteropus, considered that its cuspidate teeth most probably represented a survival from the cuspidate den- tition that the ancestors of the Pteropodidae might be presumed to have possessed, largely on account of the 'tuberculo-sectorial' appearance of the third premolar and first molar. Later, Thomas (1896 : 243, 1898 : 384) also suggested that the cuspidate canines of Harpyionycteris might also owe their origin to a presumably cuspidate-toothed condition among the ancestors of the Pteropodidae. Miller (1907 : 41), although suggesting homology between the anterior of the molar cusps of the Pteropodidae and the protocone, paracone, protoconid and metaconid of the microchiropteran tooth, considered that additional cusps and ridges were not homologous and noted that the tendency to produce supernumerary cusps reached its extreme in Pteralopex and Harpyionycteris. A similar view was adopted by Andersen (19090 : 222; 1912 : xxix, 435) 78 J. E. HILL & W. N. BECKON who considered Pteralopex to be a very specialized offshoot of Pteropus, more particularly of the Pteropus pselaphon group, the dental peculiarities of Pteralopex deriving in his opinion directly from tendencies already latent in the members of that group. Winge (1923 : 263; 1941 : 305) was also unable to recognize the tooth structure of the insectivorous bats in the multicuspidate cheek- teeth of Pteralopex and, indeed, found nothing else in its dentition to indicate primitive conditions, repeating the opinions of Miller and Andersen that Pteralopex is a highly modified pteropodid. Similarly, Tate (1951 : 4), in considering the dentition of Harpyionycteris in relation to the mega- chiropteran assemblage as a whole, suspected that its multicuspid molars, and the entire dentition, must be regarded as secondary rather than as a surviving example that represented a formerly widespread condition in the Megachiroptera. However, Phillips (1968 : 790) thought Thomas probably correct in considering Pteralopex to be an isolated relic. Slaughter (1970 : 51) has examined possible evolutionary trends in the dentition of the Chirop- tera. In discussing the Megachiroptera, this author (p. 77) reviewed the features reported for the fragmentary dentition of reputedly the earliest known megachiropteran, Archaeopteropus transiens Meschinelli, 1903, from the Oligocene of Italy, and concluded that only Harpyionycteris among living genera had a cuspidate molariform dentition that in any way approached the denti- tion ascribed to Archaeopteropus. However, little is known of the true nature of the ancestral dentition of the Megachiroptera, and Smith (1976 : 53) has remarked that there is apparently no dentition remotely similar to the 'primitive' dilambdodont condition among either the living or the fossil megachiropterans. The argument for megachiropteran dental evolution that Slaughter based on a supposed similarity between the dentitions of Archaeopteropus and Harpyionycteris was thought by Smith to be weak. In developing his theme, Slaughter (1970 : 78) considered that the dentition of Pteralopex indicated that emphasis shifted very early from a cuspid state to one of U-shaped lophs, minimal dental erosion exposing a rodent-like pattern of dentine. Thus, in Pteralopex, the crown of a partially worn upper molar will present (Slaughter, p. 65, fig. 3H) a U-shaped fossette, opening labially, the lower molars presenting two similar lophs or fossettes, the anterior opening to the rear, the posterior opening forward. He considered that the origin of the dental patterns of Pteralopex could be easily understood by comparison with Harpyionycteris and, indeed, would derive the dentition of Pteralopex from that of a Harpyionycteris-like form. This presumed loph pattern is not entirely supported by the specimens examined in the course of the present study. In Pteralopex anceps m1 and m2 have clearly a labially opening fossette, but occasionally in P. atrata the high, anterior cusp-like part of each tooth is divided lingually from the elevated pos- terior basal ledge (hypocone and metastylar cusp of Slaughter) as deeply or almost as deeply as it is labially so that wear will expose an anterior and a posterior loph, rather than a U-shaped rim, or the fossette so formed may be rimmed lingually by a low, narrow unworn ridge rather than the broad ridge figured by Slaughter. The lingual ridge of pm4 in P. atrata may also rarely be similarly deeply divided from the posterior basal ledge. The upper molariform dentition of P. acrodonta differs quite sharply from the concept of labially opening fossettes : the lingual division between the anterior part of the tooth and the posterior basal ledge in pm4, m1 and m2 is deeper than the labial division so that the fossette opens lingually rather than labially, but with a depression in the postero-external part of its labial rim. In the mandible, the anterior cusps of m^ m2 and m3 in Pteralopex anceps, P. atrata and P. acrodonta provide the necessary basis for the posteriorly directed, U-shaped anterior loph postulated by Slaughter but it is more difficult to establish a foundation for a similar but anteriorly directed posterior loph in the specimens examined. There is little trace of such a loph in m^ of any of these species: in this tooth the posterior basal ridge consists of little more than a low labial postero-external cusp, with no lingual elevation, the postero-internal part of the tooth flattened, the internal fossette thus opening postero-lingually. The posterior part of m2 in P. atrata and P. acrodonta is similar to the corresponding area of IT^, except that in P. atrata the posterior basal ridge and its postero-external cusp is larger and more massive : in both species the crown is low and flat postero-internally, as in m1} the posterior basal ridge extending rather more than half- way across the rear of the tooth, the internal fossette opening postero-lingually. In P. anceps the posterior basal ridge of m2 has a very large postero-external cusp forming an internal part of the NEW SPECIES OF PTERALOPEX 79 labial ridge, and extends across the rear of the tooth almost to its internal corner. Thus with wear the crown of the tooth will become more basin-like, its central fossette opening postero-lingually on to a small, flattened area. Posteriorly, m3 in P. atrata is broadly similar to m2 but the posterior basal ridge is a little heavier and extends a little further across the tooth, the internal fossette opening postero-lingually. In P. anceps the posterior basal ridge of m3 is heavy, more or less integral with the labial ridge and extending to the inner corner of the tooth, much as in m2, the fossette opening lingually, while in P. acrodonta the posterior basal ridge of m3 although a little elevated is narrow and short and can scarcely enclose any internal fossette. The suggestion that the dentition ofPteralopex might derive from a Harpyionycteris-like form also deserves close examination, especially since the extremely cuspidate molariform teeth of Pteralopex acrodonta have a number of similarities with the corresponding teeth in Harpyionyc- teris. They do, however, differ from the molariform teeth of Harpyionycteris in several features. The last upper premolar (pm4) in P. acrodonta closely resembles its counterpart in Harpyionyc- teris, its major labial cusp with a small posterior subsidiary cusp as in that genus, but has addi- tionally a smaller anterior cusp, not evident in Harpyionycteris, while the lingual cusp has small anterior and posterior basal cusps which are barely if at all evident in that genus. The anterior basal ledge of the tooth is a little wider and flatter in P. acrodonta, and the posterior basal ledge larger and heavier, not clearly divided into two cusps as in Harpyionycteris. The upper molars (m1"2) of P. acrodonta, although basically with the same cusp pattern as those of Harpyionycteris, have more elevated crowns, with the cusps less clearly divided : the main labial cusp of m1 has an additional rudimentary subsidiary cusp on its anterior face and in both teeth the anterior basal ledge is more developed, the posterior basal ledge more elevated, heavier and more massive, not divided into two cusps. In Harpyionycteris the posterior basal ledge in m1 is divided into two cusps but in m2 it forms an indistinctly divided postero-internal cusp : accessory cusps are sometimes present in these teeth (Peterson & Fenton, 1970 : 7, fig. 2) but usually lingually. As in P. acrodonta, m2 is relatively unreduced. The lower molariform dentitions of Pteralopex acrodonta and Harpyionycteris have similar resemblances and differences. The principal cusps of pm4 in both form an anterior 'trigonid', but in Harpyionycteris the tooth has additionally an anterior basal cusp and there is a small sub- sidiary cusp on the posterior face of the principal lingual cusp. The first lower molar (m^) of P. acrodonta has anteriorly a group of four well-divided cusps arranged laterally in two pairs, the antero-labial of these with perhaps a faint trace of further cuspidation. Posteriorly, the tooth has a small low external basal cusp and its internal part is low and flattened. This configuration resembles the crown of ml in Harpyionycteris except that in this genus the anterior part of the tooth has three rather than two lingual cusps. The second lower molar (m2) in P. acrodonta is similar to m^ with the anterior part of the crown clearly divided, but in the unworn dentition the large antero-labial cusp has a smaller subsidiary cusp divided from its postero-external face, so that effectively there are three labial cusps : lingually, m2 in P. acrodonta has two cusps, as in mj. Posteriorly, there is a low external basal cusp flanking the low, flattened internal part of the tooth. In contrast, the anterior part of m2 in Harpyionycteris has two labial and three lingual cusps : posteriorly, the tooth resembles m2 of P. acrodonta. The unworn cuspidation of the anterior part of m3 in P. acrodonta is similar to that of the anterior part of m2, with a group of three labial and two lingual cusps, the first two labially consisting of a larger cusp with a smaller subsidiary cusp divided from its postero-external face. Posteriorly, however, the tooth is strongly elevated and slightly cuspidate, its internal part not low and flattened. In Harpyionycteris the anterior part of m3 has two labial and three lingual cusps : posteriorly, there is a low external basal cusp but the internal part of the tooth is low and flattened as it is in m2. However, m3 in Harpyionycteris is relatively unreduced, in this respect resembling m3 in P. acrodonta. It is clear, therefore, that Pteralopex acrodonta resembles Harpyionycteris quite closely in the major details of its molariform teeth, differing chiefly in a slightly greater degree of development of the basal ledges of pm4, m1 and m2, in the greater elevation of the crowns of the latter two teeth, and in certain details of cuspidation. The degree and pattern of cuspidation, especially where accessory cusps are concerned, may prove variable to some extent: only the holotype of H. whiteheadi whiteheadi has been available for comparison but variability in the cusps of m1 and m2 80 J. E. HILL & W. N. BECKON in H. w. negrosensis has been clearly demonstrated by Peterson & Fenton (1970 : 5, 7, fig. 2). In other respects the dentition of Harpyionycteris differs widely from that of P. acrodonta, as in its reduced number of incisors with the specialized structure and proclivity of the upper pair, this latter a character shared with the upper canines, and the effective obsolescence of lower incisors with their apparent substitution by strongly tricuspidate, rather incisiform, slightly procumbent lower canines, supplemented by well-developed anterior lower premolars (pm2_2). These extensive differences militate against the view that the dentition of Pteralopex can be derived from a Harpyionycteris-\ike form: Andersen (1909a : 220) considered that the structure of pm3, pm4 and m1 in Pteralopex could be derived very easily from that of the corresponding teeth of any species of Pteropus and had most probably originated from teeth in which the posterior basal ledge was already more than usually developed, as for instance in the members of the pselaphon group of Pteropus or in Pteropus samoensis; in the P. pselaphon group the anterior cingulum of pm3 and pm4 is also raised. Furthermore, this author (pp. 217, 220) presented a detailed case for the progressive division of the lateral ridges of certain of the lower molariform teeth in Pteropus. In the majority of species the ridges are simple: in P. samoensis and P. pilosus a faint depression in the inner or lingual ridge of pm4 suggested an initial division to Andersen, leading to P. pselaphon in which the inner ridges of pm4 and m1 are very distinctly divided and to P. leucopterus in which there is a depression in the inner ridges of mx and m2, with usually a shallower depression in the outer or labial ridges of these teeth, although it may sometimes be absent or scarcely detectable. This same process can be discerned more vividly in the three species of Pteralopex, from a least emphatic, Pteropus-like condition in Pteralopex anceps, approaching the pselaphon group of Pteropus as exemplified by P. pilosus, P. pselaphon and P. leucopterus, through an inter- mediate stage in Pteralopex atrata, to the strongly multicuspidate condition of P. acrodonta. The hypothesis that the multicuspidate molariform teeth of Pteralopex are an extreme in the Pteropodine dentition is attractive : an opposite extreme is to be found in the low, broadened and rounded cusps and ridges of the molariform teeth of Styloctenium or of Aproteles Menzies, 1977. Although Andersen (19090 : 220) suggested that the multicuspidate condition could be derived from the molariform dentition of Pteropus, it is perhaps more plausible to suggest that the smoother or laterally ridged crown represents a derived condition, the cuspidate crown a less modified state. The links between the dentition of Pteralopex and the pselaphon group of Pteropus, rather than indicating Pteralopex to be a specialized offshoot of Pteropus as Anderson (1909a : 222) thought at one time, may well indicate that Pteralopex and the Pteropus pselaphon group derive from a close common ancestor. Andersen (1912 : 1) himself later adopted a similar view, con- sidering it scarcely open to doubt that Pteralopex had developed from a bat closely related to the living species of the Pteropus pselaphon group, or, in other words, that the genus (as it was then understood) was a peculiarly modified representative of that group in the Solomon Islands. The adaptive significance of the multicuspidate dentition in Harpyionycteris was discussed briefly by Peterson & Fenton (1970 : 7) who speculated that, like most of its allies, Harpyionyc- teris is a fruit eater, perhaps adapted for a particular type of tough-textured fruit, the multicus- pidate teeth being valuable in extracting the juice. Similar considerations may apply to Pteralopex (Sanborn, (1931 : 21) reported P. atrata feeding on green coconuts) and especially to P. acrodonta which closely resembles Harpyionycteris in the extent of cuspidation and in the lack of reduction of the last molars, both adaptations that apparently maintain a maximum of occlusal area despite relatively small overall size. Harpyionycteris is rather isolated within the Pteropodidae, having been given subfamilial status by Miller (1907 : 77) and by Andersen (1912 : 799), who, however, remarked (p. 803) that but for the fact that the plan of his Catalogue of 1912 had been pre- determined before detailed work had been undertaken the genus ought to have been classed in the Pteropodinae, immediately after Dobsonia. Tate (1951 : 4) doubted the association with Dobsonia (this author misinterpreted Andersen's action in according subfamilial rank to Harpyioncteris as a lack of conviction in his association of the genus with Dobsonia) and suggested a connection with Nyctimene. Koopman & Cockrum (1967 : 1 16) also accorded subfamilial rank to Harpyionyc- teris but more recently Koopman & Jones (1970 : 23) preferred tribal status for the genus, within the Pteropodinae. On the other hand, Schultz (1970) noted its isolated systematic position in the Megachiroptera and considered that its multicuspidate molars indicated that it did not live NEW SPECIES OF PTERALOPEX 8 1 exclusively on fruit. He concluded that certain morphological features of the digestive tract (notably the structure of the intestinal mucosa) pointed to an early separation from other ptero- podids, and consequently postulated familial rank as the Harpyionycteridae. Hitherto among the Megachiroptera, extreme cuspidation of the molariform teeth has been unique and diagnostic to Harpyionycteris and for any higher category based upon it. In contrast, the multicuspidate condition in Pteralopex acrodonta can be linked through P. atrata and P. anceps to the condition more general among the other members of the Pteropodinae. Acknowledgements Our thanks are due to Mrs W. N. Beckon for much essential assistance and support, particularly with the collection of specimens; to Mr Peter Thomson, District Officer, Taveuni Island, and his wife Marijcke, who offered hospitality and gave invaluable assistance with transport, accommo- dation and food during successive visits to the island, and to Mr Vasu Shankaran of Taveuni and Mr Robin Mercer of Vanua Levu for providing local information. We are grateful also to Dr J. D. Smith, of the Department of Biological Sciences, California State University, Fullerton, California, whose views in correspondence with the senior author have been most valuable during the preparation of this paper. References Andersen, K. 1909a. On the characters and affinities of ' Desmalopex" and Pteralopex. Ann. Mag. nat. Hist. (8), 3 : 213-222. 19096. Two new bats from the Solomon Islands. Ann. Mag. nat. Hist. (8), 3 : 266-270. 1912. Catalogue of the Chiroptera in the Collection of the British Museum, 2nd ed. I. Megachiroptera. London. Capell, A. 1973. A New Fijian Dictionary. 4th ed. Suva. Felten, H. I964a. Zur Taxionomie indo-australischer Fledermause der Gattung Tadarida (Mammalia, Chiroptera). Senckenberg biol. 45 : 1-13, 7 figs. 19646. Flughunde der Gattung Pteropus von den Neuen Hebriden (Mammalia, Chiroptera). Senckenberg. biol. 45 : 87-92, 6 figs. 1964c. Flughunde der Gattung Pteropus von Neukaledonien und den Loyalty-Inseln (Mammalia, Chiroptera). Senckenberg. biol. 45 : 671-683, 6 figs. & Kock, D. 1972. Weitere Flughunde der Gattung Pteropus von den Neuen Hebriden, sowie den Banks- und Torres-Inseln, Pazifischer Ozean. Senckenberg. biol. 53 : 179-188. 3 figs, 1 tab. Gill, W. W. 1876. The Mammalia of the Pacific. The Leisure Hour, 308-309. [N.V.]. Jentink, F. A. 1887. In H. Schlegel & F. A. Jentink, Museum d'Histoire Naturelle des Pays-Bas. Revue methodique et critique des Collections deposees dans cet Etablissement. Tome 10. Catalogue Osteologique des mammiferes. Leiden. 1888. In H. Schlegel & F. A. Jentink, Museum d'Histoire Naturelle des Pays-Bas. Revue methodique et critique des Collections deposees dans cet Etablissement. Tome 12. Catalogue Systematique des mam- miferes (Rongeurs, Insectivores, Cheiropteres, Edentes et Marsupiaux). Leiden. Koopman, K. F. 1971. Taxonomic notes on Chalinolobus and Glauconycteris (Chiroptera : Vespertili- onidae). Am. Mus. Novit. No. 2451 : 1-10, 1 fig. & Cockrum, E. L. 1967. In S. Anderson & J. K. Jones, Jr (Eds), Recent Mammals of the World. A Synopsis of Families. 6. Bats, pp. 109-150, 11 maps. New York. & Jones, J. K., Jr 1970. Classification of bats. In B. H. Slaughter & D. W. Walton (Eds), About Bats. A Chiropteran Biology Symposium, pp. 22-50, 1 1 figs, 2 tabs. Dallas. Krzanowski, A. 1977. The easternmost occurrence of bats in Polynesia. Acta theriol. 22 : 271-272. Laurie, E. M. O. & Hill, J. E. 1954. List of Land Mammals of New Guinea, Celebes and adjacent Islands J 758-1 95 2. London. Macdonald, J. D. 1857. Proceedings of the Expedition for the exploration of the Rewa River and its tributaries, in Na Viti Levu, Fiji Is. // R. geogr. Soc. 27 : 232-268, map. Menzies, J. I. 1977. Fossil and subfossil fruit bats from the mountains of New Guinea. Aust. J. Zool. 25 : 329-336, 3 figs, 3 tabs. Meschinelli, L. 1903. Un nuovo chirottero fossile (Archaeopteropus transiens Mesch.) dellelignitideMonte- viale. Atti R. Inst. veneto Sci. 62 (2) : 1329-1344, 1 pi. 82 J. E. HILL & W. N. BECKON Miller, G. S. 1907. The families and genera of bats. Bull. U.S. natn. Mus. 57 : i-xvii, 1-282, 49 figs, 14 pis. Peterson, R. L. & Fenton, M. B. 1970. Variation in the bats of the genus Harpyionycteris, with the descrip- tion of a new race. Contr. Life Sci. Div. R. Ont. Mus. No. 17 : 1-15, 6 figs, 3 tabs. Phillips, C. J. 1968. Systematics of the Megachiropteran bats of the Solomon Islands. Univ. Kans. Publs Mus. not. Hist. 16: 777-837, 17 figs., 6 tabs. Sanborn, C. C. 1931. Bats from Polynesia, Melanesia, and Malaysia. Publs Field Mus. not. Hist. Zool. Ser. 18 : 7-29. Schultz, W. 1970. Einige Bemerkungen zum Bau des Verdauungstraktes und der systematischen Stellung des Spitzzahnflughundes, Harpyionycteris whiteheadi Thomas, 1896 (Megachiroptera). Z. Saugetierk. 35 : 81-89, 7 figs. Slaughter, B. H. 1970. Evolutionary trends of Chiropteran dentitions. In B. H. Slaughter & D. W. Walton (Eds), About Bats. A Chiropteran Biology Symposium, pp. 51-83, 5 figs. Dallas. Smith, J. D. 1976. Chiropteran evolution. In R. J. Baker, J. K. Jones, Jr & D. C. Carter (Eds), Biology of bats of the New World family Phyllostomatidae. Part I. Spec. Publs Mus. Texas Tech Univ. No. 10 : 49-69, 2 figs. Smith, S. P. 1902. Niue Island and its people. /. Polynes. Soc. 11 : 81-106. [N.V.] Tate, G. H. H. 1951. Harpyionycteris, a genus of rare fruit bats. Am. Mus. Novit. No. 1522 : 1-9, 4 figs, 2 tabs. Thomas, O. 1888. Diagnoses of six new mammals from the Solomon Islands. Ann. Mag. not. Hist. (6) 1 : 155-158. 1889. The mammals of the Solomon Islands, based on the collections made by Mr C. M. Woodford during his second expedition to the archipelago. Proc. zool. Soc. Land. (1888) : 470-484, 3 pis. 1896. On mammals from Celebes, Borneo, and the Philippines recently received at the British Museum. Ann. Mag. nat. Hist. (6) 18 : 241-250. 1898. On the mammals obtained by Mr John Whitehead during his recent expedition to the Philip- pines. Trans, zool. Soc. Land. 14 : 377-412, 7 pis. Winge, H. 1923. Pattedyr-Slaegter. I. Monotremata, Marsupialia, Insect ivora, Chiroptera, Edentata. Kobenhavn. 1941. The Interrelationships of the Mammalian Genera (Ed. A. S. Jensen, R. Sparck & H. V01s0e, translated E. Deichmann & G. M. Allen). /. Monotremata, Marsupialia, Insectivora, Chiroptera, Edentata. K0benhavn. Wodzicki, K. & Felten, H. 1975. The peka, or fruit bat (Pteropus tonganus tonganus) (Mammalia, Chirop- tera), of Niue Island, South Pacific. Pacif. Sci. 29 : 131-138, 2 figs, 4 tabs. British Monographs & The Museun- r including zool< Besides beii .~ the handbook- reading. Lists are available Publica British Crc Standing ore of 10% ofT our publisher 1 lie ai macro nKBHH Pteroi ByJ.~: wt Bv ished in Volume 34 ^^^^H ) } n lationships of the cyprinid fish Luciobrama ! (L 10 p< de). B> G. f. Howes, [ Pteralopex ['homas, 1888 (Chiroptera : v; ' ron b : ' iji L lands. 5 -•/ s| Beckon. ^acies of Sertularii^ae (Coelenterata : Hydroida) tnd •: arby seas. By P. F. S. Cornelius. i .: • • imens oi >ircls in ibe collection of the ij History). P. J, K. Burton. otiles from northern Trengganu, Malaysia, two new geckos: Cnemaspis and Crytodactylus. j ;.» -s>j J ./ (. Portia (Araneae : Salticidae). Type set by John Wright & Sons Ltd. Bristol and Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) A revision of the spider genus Portia (Araneae: Salticidae) F. R. Wanless Zoology series Vol 34 No 3 21 December 1978 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany Entomology, Geology and Zoology, and an Historical series. Parts are published at irregular intervals as they become ready. Volumes will contain about three hundred pages, and will not necessarily be completed within one calendar year. Subscription orders and enquiries about back issues should be sent to : Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviations: Bull. Br. Mus. nat. Hist. (Zool.) Trustees of the British Museum (Natural History), 1978 ISSN 0007-1498 Zoology series Vol 34 No 3 pp 83-124 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 21 December 1978 A revision of the spider genus Portia (Araneae : Salticidae) F. R. Wanless r*— Department of Zoology, British Museum (Natural History), Cromwell Road, London SW7 5BD Contents Synopsis ............. 83 Introduction ............ 83 The genus Portia 84 Definition ............ 84 Biology. ............ 85 Affinities 85 Species list ............ 86 Key to species ............ 86 The schultzi-group ........... 87 Species Sola 109 The kenti-group . . . . . . . . . . . . Ill Species Sola . . . . . . . . . . . . 116 Species Incertae sedis . . . . . . . . . . . 116 Acknowledgements 117 References . . . . . . . . . . . . . 117 Synopsis The spider genus Portia Karsch is revised. All the 16 known species (of which 5 are new) are described and figured. Biological and distributional data are given and a key to the species is provided. Two species groups based on the structure of the male genitalia are proposed. The type-material (including 14 holo- types) of 22 nominate species was examined and 7 lectotypes and 1 neotype are newly designated. One genus is revived, 2 generic and 4 specific names are newly synonymized and 6 new combinations are proposed. Introduction Portia is a distinctive genus of long-legged ornate spiders which have attracted the attention of naturalists on account of their conspicuous leg fringes and abdominal hair tufts. The genus is represented in both the Oriental and Ethiopian regions and at present includes 16 species. Portia, as recognized in this paper, has previously been divided between two genera Portia and Linus which Simon (1901) placed in separate suprageneric groups, the Boetheae comprising Portia and Boethus, and the monogeneric Lineae. In the same work Simon synonymized Brettus Thorell, 1895 with Portia, but recent studies on the type species of Brettus, B. cingulatus Thorell, have shown that this genus is valid, and furthermore that several species at present in Portia will have to be transferred back into Brettus. Simon's understanding of Portia seems to have been based in part on Portia semifibriata (Simon), from India and not on the type species P. schultzii Karsch from South Africa. P. semifimbriata agrees well with Simon's concept of Boetheae but it is not congeneric with P. schultzii and will have to be referred back into Brettus, i.e. its original combination. P. schultzii, in spite of uncertainties in respect of adult females (see below, p. 88), is clearly congeneric with Portiafimbriata (Doleschall), the type species of Linus. In fact a Madagas- can record for P. fimbriata (Simon, 1901) probably refers to P. schultzii or even P. africana (Simon) both of which are now known to occur in Madagascar. Bull. Br. Mus. nat. Hist. (Zool.) 34 (3) : 83-124 Issued 21 December 1978 83 84 REVISION OF PORTIA The Ethiopian species have been revised by Roewer (1965) who recognized both Portia and Linus as valid, distinguishing them by the curvature of the first eye row and carapace shape. Unfortunately he based his descriptions and figures almost entirely on the literature and did not examine many of the type specimens. His new genus Neccocalus, proposed for Cocalus africanus Thorell, is a synonym of Portia as C. africanus is conspecific with the widespread West African species P. africana. Genus PORTIA Karsch Sinis Thorell, 1878, June: 269. Type species Salticus fimbriatus Doleschall, by original designation [Junior homonym of Sinis Heer, 1862 : 31]. Petrunkevitch, 1928 : 246. Bonnet, 1958 : 4061. Portia Karsch, 1878 [December] : 774. Type species Portia schultzii Karsch, by original designation and monotypy. Peckham & Peckham 1885 : 267. Simon, 1901 : 400-403. Petrunkevitch, 1928 : 182. Bonnet, 1958 : 3766. Roewer, 1965 : 10. Linus Peckham & Peckham, 1885 : 289 [Replacement name for Sinis Thorell]. Simon, 1901 : 400, 408, 410. Petrunkevitch, 1928 : 181. Sherriffs, 1931 : 538. Bonnet, 1957 : 2482. Roewer, 1965 : 14. Syn. n. Boethoportia Hogg, 1915 : 501. Type species Boethoportia ocellata Hogg, by monotypy. Petrunkevitch, 1928: 181. Strand, 1929: 15. Roewer, 1954:933. Bonnet, 1955:892 [Synonymized by Proszynski, 1971 : 385]. Neccocalus Roewer, 1965 : 20. Type species Cocalus africanus Thorell, by original designation and monotypy. Syn. n. Simon (1901), without giving his reasons, synonymized Brettus Thorell with Portia Karsch. However, preliminary studies on the holotype of the type species Brettus cingulatus Thorell have shown that the genus is valid and it is here removed from synonymy gen. rev. The genus will be revised in a paper in preparation. DEFINITION. Medium to large spiders ranging from about 4-5 to 9-5 mm in length. Sexes alike in general body form, but colour markings sometimes showing slight sexual dimorphism. Usually ornate with tufts of hair and leg fringes; colour patterns composed on setae (easily rubbed). Carapace: high, usually with marked slope from posterior lateral eyes to anterior row and to posterior thoracic margin; fovea elongate, just behind posterior lateral eyes; sculpturing not marked. Eyes: anteriors subcontiguous, apices procurved to recurved; posterior median eyes relatively large, about midway between anterior lateral and posterior lateral eyes or nearer to anterior laterals; posterior row usually narrower than anterior row; quadrangle length between 38 and 55 per cent of carapace length. Clypeus: high, concave. Chelicerae: medium to large, more or less vertical; promargin with three teeth, retromargin with three to six. Maxillae: elongate, usually divergent. Labium: subtriangular, about as long as broad. Sternum: scutiform to elongate scutiform. Pedicel: short. Abdomen: usually ovoid to elongate ovoid, rarely elongate; scuta lacking; spinnerets subequal in length, anteriors and posteriors robust, medians slender, usually hirsute; colulus apparently lacking, its position indicated by scanty hair tuft in front of spinnerets; trachea (Fig. ID) unbranched, arising from transverse slit just in front of spinnerets and apparently limited to the abdomen. Legs: long and slender, usually with conspicuous fringes; spines numerous, generally robust; claw tufts present, scopula lacking, but female metatarsi I and tarsi I with minute ventral setae (PI. 5c, d). Female palp: usually hirsute with terminal claw. Male palp: femoral apophysis usually lacking; tibia with prolateral and ventral apophyses; cymbium usually modified proximally and often with a prolateral flange (Fig. IB, F); embolus moderately long and slender; conductor sometimes present; tegulum with peripheral seminal reservoir, a deeply curved furrow and a tripartite membraneous apophysis (adjacent to the embolic base) that sometimes extends laterally to form a small tegular apophysis (Fig. 10A, D); median apophysis lacking. Epigyne: relatively simple openings usually indistinct; seminal ducts generally short, wide and very dark, opening into large, dark, rounded spermathecae. Remarks. The conductor (c) is very closely associated with the embolic shaft and would appear to have a supporting or protective function. In P. labiata (Thorell) and P. crassipalpis (Peckham & Peckham) the conductor is well developed (Figs 9B, 10A), but in P.fimbriata (Doleschall) it is greatly reduced, its presence being indicated by a shallow groove across the embolic shaft (PI. 3a, F. R. WANLESS 85 b). This groove is not joined to another groove which extends along the inside of the embolus and terminates near the tip (PL 3e, f). In P. schultzii and several other species the conductor is apparently absent. The tripartite membraneous apophysis (m) is generally indistinct (PI. 3b, c) and its prolateral extension is not always evident. As might be expected preliminary observations with the scanning electron microscope have revealed additional characters which will undoubtedly be of phylogenetic significance. For example, the species examined in this study all have the same type of setae (PI. 4, 5) which appear to show inter- and intraspecific differences. Unfortunately a detailed analysis of microsculpture cannot be presented as several species are known only from one or two specimens that are in poor condition and cannot be used for SEM studies. BIOLOGY. Gravely (1921) reports an Indian species of 'Linus' stalking and pouncing on a pholcid, Smeringopus sp., in its web. Another species (Bristowe, 1941) was seen devouring an Indian web- building pisaurid, Euprothenops ellioti (O. P.-C). Bristowe also records 'Linus fimbriatus' feeding on Smeringopus elongatus (Vinson), on an Araneus and on a Theridion. In each case the 'Linus' met with success in capturing its prey and remained in the victim's web to eat the owner before retiring. P.fimbriata has also been recorded from the web ofNephila malabarensis (Sherriffs, 1931). An African Portia has been found in association with another spider's web by John and Frances Murphy, who collected two immature specimens of P. schultzii Karsch from an extensive diplurid web in Kenya. Important observations (previously unpublished) were made by Frances Murphy who successfully reared the above juveniles through four or five moults. The female died after its penultimate or final moult (see p. 88), but the male reached adulthood. In captivity both spiders made several webs which were apparently used for trapping and locating prey. The male's web was more flimsy than that of the female, but neither web appeared to be essential as both spiders readily caught wingless fruit flies when the cages were cleaned and the webs destroyed. Moulting was not observed, but several exuvia were found hanging upside down below the sheet- web and it seems reasonably certain that P. schultzii moults in the open. The adult male did not spin a web, but when placed on that of the female (PI. la, b) found no difficulty in running about and adopted poses which had not been previously noticed and were taken to be part of a courtship routine. Another species, believed to be P. durbanii Peckham & Peckham collected by the author and Mr A. Russell-Smith in Durban, was also reared through several moults by Frances Murphy, but there was no web spinning activity. The specimen, apparently a subadult male, died just before its final moult. Bristowe (pers. comm.) quite naturally assumed that Portia invaded spider webs for the purpose of feeding, but the above observations on captive specimens suggests that some Portia species may build and live in their own webs and use them for prey capture, either in isolation or in association with the webs of other spiders. Careful field observations are needed to resolve this aspect of behaviour as web building has not to my knowledge been reported for salticid spiders. AFFINITIES. The affinities of Portia cannot be fully reviewed at present as numerous related genera have yet to be revised. The structure of the genitalia suggests that Portia is closely related to Brettus and some species of Cocalus (e.g. C. lancearius Thorell). Unfortunately the type species of Cocalus, C. concolor Koch, known only from a single male specimen, has no palps and I am unable to present a diagnosis at the moment. Preliminary observations suggest that Portia, Brettus, Cocalus and other genera with large posterior median eyes may be related to lyssomanid spiders. For example, P. adonis (Simon), P. albolimbata (Simon) and P. semifimbriata (Simon) do not belong in Portia but represent another genus Brettus (type species Brettus cingulatus Thorell), which may form a link between Asemonia and Portia. If one considers Simon's concept of Portia (based in part on P. semifimbriata) and its affinities with Boethus (sensu Simon) then Simon's comment (Simon, 1901) that: 'Les Boethus (Nealces E. Sim.) me paraissent faire le passage des Lyssomanes aux Cocalodes et aux Linus' become significant. The affinities of Cocalodes are uncertain but the genus is probably close to Brettus. The taxonomic status of lyssomanids has yet to be resolved. The presence of four eye rows has until recently been considered diagnostic of lyssomanids but it has already been shown (Galiano, 86 REVISION OF PORTIA 1976 and Wanless, 1978a) that several 'typical' salticid genera have similar eye formulae. The most aberrant example of a salticid with four eye rows is Athamas whitmeei O. P.-Cambridge, which has the anterior lateral eyes behind and almost exactly above the anterior medians. Even Simon (1901) remarked on the similarity with Lyssomanes, but as far as I am aware he has never suggested that the genera were closely related, presumably because the genitalia of A. whitmeei are of the type frequently found in the Salticidae. If Portia and related genera were transferred to the Lyssomanidae, then the large posterior median eyes might be considered as generally diagnostic of Lyssomanidae. However Pandisus, a small madagascan genus closely related to Asemonia, is exceptional in having small posterior median eyes. Platnick (1971) has suggested that courtship behaviour in Lyssomanes bradyspilus Crane, as described by Crane (1949), indicates that lyssomanids should have family status, but Galiano (1976) holds the view that on anatomical grounds the group merits no more than subfamily rank. The decision is made difficult as our knowledge of the group is very limited. However, the web spinning behaviour of P. schultzii, the flimsy brooding webs of Lyssomanes jemineus Peckham & Wheeler (Eberhard, 1974) and Asemonia sp. n. (Murphy coll, vials 1549, 3661; F. Murphy un- published observations) are in contrast to known salticid behaviour. Furthermore, the branched tracheal systems of several 'typical' Salticidae (Lamy, 1902; Hill, 1977 and Wanless, 1978a) are more complex than the unbranched systems found in Asemonia (Wanless unpublished observa- tion), Portia (Fig. ID) and Lyssomanes (Lamy, 1902 and Forster, pers. comm.). Forster (1970) has already argued that the general complexity of the tracheal system (i.e. branched or unbranched) is of more evolutionary significance than the presence or absence of tracheal intrusions into the cephalothorax. There is thus evidence to support Platnick's view, or at least suggest that lysso- manids should be accorded a higher taxonomic rank than other salticid subfamilies as they are understood at the present time. List of species in the genus Portia Karsch, 1878 Portia africana (Simon, 1886) P, albimana (Simon, 1900) P. alboguttata (Lawrence, 1938) P. assamensis sp. n. P. cazomboensis sp. n. P. crassipalpis (Peckham & Peckham, 1907) P. durbanii Peckham & Peckham, 1903 P. falcifera sp. n. P. fimbriata (Doleschali, 1859) P. kenti Lessert, 1925 P. labiata (Thorell, 1887) P. madagascarensis sp. n. P, oreophila sp. n. P. russata Simon, 1900 P. schultzii Karsch, 1878 P. solitaria Lessert, 1927 Key to species of Portia Males 1 Tibial apophysis jointed (Figs 14C, F; 16B, C) (Africa, Madagascar) - Tibial apophysis not joined (Ethiopian and Oriental regions) ...... 4 2 Tibial apophysis very robust (Fig. 16B-D) (Madagascar) . . madagascarensis sp. n. (p. 114) - Tibial apophysis slender (Fig. 14C, F) 3 Embolus short and slender (Fig. 14A) (South Africa) . . . kenti Lessert (p. Ill) - Embolus long and robust (Fig. 14B) (Uganda) .... falcifera sp. n. (p. 1 1 1 ) 4 Femora of palp with a distal blunt apophysis (Fig. 13B) (South Africa) durbanii Peckham & Peckham (p. 109) F. R. WANLESS 87 - Femoral apophysis lacking 5 5 Embolus relatively short (Fig. 12B) (Sri Lanka) .... albimana (Simon) (p. 107) - Embolus relatively long ............ 6 6 Tegular apophysis present (Fig. 10A, D, C, F) 7 - Tegular apophysis absent ............ 8 7 Tibial apophysis slender (Fig. 10B) (Sri Lanka, India, Burma, Thailand, Malaysia, Sumatra) labiata (Thorell) (p. 103) - Tibial apophysis robust (Fig. 10F) (Assam, Nepal) .... assamensis sp. n. (p. 105) 8 Conductor well developed (Fig. 9B) (Malaysia, Borneo) crassipalpis (Peckham & Peckham) (pi 00) - Conductor poorly developed or lacking .......... 9 9 Cymbial flange as in Fig. IB, F; clypeus with a white spot below each AM (Zaire, Kenya, Tanzania, South Africa, Madagascar) ...... schultzii Karsch (p. 88) - Cymbial flange otherwise; white spots lacking . . . . . . . . 10 10 Cymbial flange as in Fig. 4D; clypeus with curved white 'moustache' (West and Central Africa) ufricana (Simon) (p. 93) - Cymbial flange as in Fig. 7G ; 'moustache' lacking . . . fimbriata (Doleschall) (p. 99) Females 1 Epigyne as in Fig. 17B; body with white longitudinal bands (Fig. 17D) (Madagascar) oreophyla sp. n. (p. 116) - Epigyne and body otherwise ........... 2 2 Epigyne as in Fig. 6B, spermathecae large, ducts apparently lacking; (Madagascar) r us sat a Simon (p. 97) - Epigyne otherwise 3 3 Anterior eye row strongly recurved in frontal view 4 - Anterior eye row procurved in frontal view ......... 6 4 Epigyne with central curtain-like membrane (Fig. 2C) (Angola, South Africa) cazomboensis sp. n. (p. 90) - Epigyne without central membrane .......... 5 5 Epigyne with delicate septum (Fig. 1 3D) (South Africa) . durbanii Peckham & Peckham (p. 109) - Septum lacking; epigyne as in Fig. 3D-F (Guinea, Ivory Coast, Zaire) solitaria Lessert (p. 91) 6 Epigyne with sclerotized septum as in Fig. 5A, B, C (Africa) 7 - Epigyne otherwise ............. 8 7 Septum central, ducts not extended posteriorly (Fig. 5A, B, F) . . africana (Simon) (p. 93) - Septum distal, ducts extended posteriorly (5C, D, H) . . alboguttata (Lawrence) (p. 96) 8 Epigyne small and pale; spermathecae indistinct (Fig. 1G; PI. 3b) (Africa) [? subadult ?] schultzii Karsch (p. 88) - Epigyne otherwise (Oriental region) 9 9 Epigyne as in Fig. 8A-F; clypeus not thickly white haired . fimriata (Doleschall) (p. 99) - Epigyne otherwise; clypeus thickly white haired . . . . . . . . 10 10 Epigynal orifice undivided (Fig. 11 B) labiata (Thorell) (p. 103) - Epigynal orifice divided (Fig. HE) assamensis sp. n. (p. 105) The schultzii-group The schultzii-group occurs in both the Ethiopian and Oriental regions and is comprised of ten species. It is characterized by the presence of a fixed male palpal tibial apophysis that lacks the membraneous joint found in the kenti-group. The latter group is known only from males and it is not known if the epigynes of schultzii-group females show any diagnostic features relative to the kenti-group. Two species, P. cazomboensis sp. n. and P. solitaria Lessert, resemble the kenti-group in having the first eye row recurved in frontal view. However, I have not placed them in the kenti-group as the males of both species are unknown and the resulting definition based on the female genitalia could be misleading. Furthermore, P. russata (Simon) is somewhat intermediate as the first eye row is only slightly recurved and not procurved as in other species of this group. 88 REVISION OF PORTIA Portia schultzii Karsch (Fig. 1A-G; Pis 1, 2, 4a, b) Portia schultzii Karsch, 1878 : 774, ?. Holotype ?, South Africa, Port Natal (MNHU, Berlin) [examined]. P. schultzi: Simon, 1901 : 402, 403 [Unjustified emendation]. Petrunkevitch, 1928 : 182. Roewer, 1954 : 934. Bonnet, 1958 : 3767. Roewer, 1965 : 12. Proszyriski, 1971 : 461. Brettus martini Simon, 1900: 31, ?, South Africa, Natal, Zululand (? MNHN, Paris) [Not examined, presumed lost; synonymized by Simon, 1901 : 402, 403]. Roewer, 1954 : 934. Bonnet, 1958 : 3767. Linus lesserti Lawrence, 1937:254, fig. 22, #. Holotype oYa giganteus 3A*,3A 6S,7K,St /zera's 20A Fulmarus Pelagodroma glacialis 5A.4A* 9S,K,10St marina 3A*,9A 3S,3St Thalassoica Fregetta antarctica 6A 2S,2K,St tropica 2S £>ap//o/z Hydrobates capense 2A 4S,K,St pelagicus 5A 6S Pagodroma Oceanodroma nivea 3A 3S tethys A 3S Pterodroma Cflj/ro A 4S,2K,St macroptera 10S leucorhoa A 15S,K /ftSttw/ 4S,3K,Sf hornbyi S Cfl/ZOH> S* melania 2A inexpectata 12S,K solandri S brevirostris S PELECANOIDIDAE neglecta S One genus, 4 species magentae 2K One species unrepresented arminjoniana A S Pelecanoides wo///5 4A S magellani 2A phaeopygia 2S,K georgicus 7A externa S urinatrix 3A*,17A 6S,2K,2St coo/:/ K Halobaena caerulea 4S PELECANIFORMES Pachyptila PHAETHONTIDAE vittata 3 A S,12K,4St 1 genus, 3 species salvini 5S Phaethon desolata 13A 12S,10K aethereus 8A 6S belcheri 9A,2A* 7S,3K rubricauda A* ST /mwr A 8S,St /e/j/wrw^ A*,8A 5S,2St,K 130 J. S. BLANDAMER AND P. J. K. BURTON PELECANIDAE 1 genus, 8 species Spirit Skels Pelecanus onocrotalus a 5S roseus S rufescens K philippensis S,2St,K crispus S conspicillatus 2S,K erythrorhynchos S,2K occidentalis 3A 4S,2K SULIDAE 2 genera, 9 species 2 species unrepresented Morus bassanus Sula nebouxii variegata abbotti dactylatra sula leucogaster 5A*,4A A 2A*,5A 10S,6St,K 2K S K 3S,St 8S.K St 2Ak,2A*,7A S,2K PHALACROCORACIDAE 3 genera, 33 species 14 species unrepresented Phalacrocorax auritus olivaceus sulcirostris carbo nigrogularis aristotelis urile magellanicus bougainvillii varius carunculatus verrucosus atriceps albiventer Halietor melanoleucos africanus niger pygmeus Nannopterum harrisi 3A 3A,a 3A*,3A Ak,2A* Ak 2A* A,Ak 3A 4A* 3S S,K,St 3S 12S,/K,4St 4S,4K 16S,4K,5St S 3S* S,St,4K 2S S,2K S,2K S 2a,4A 2S,St,4K ANHINGIDAE 1 genus, 4 species 1 species unrepresented Anhinga rufa 3A* St melanogaster 2S,K anhinga 7 A*, 7 A S FREGATIDAE 1 genus, 5 species 1 species, unrepresented Spirit Skels Fregata aquila A*,2Ak,9A 3S,K magnificens 3A S minor A 2K ariel S CICONHFORMES ARDEIDAE 16 genera, 64 species 30 species unrepresented BOTAURINAE Botaurus stellaris 5A 4S,K,St poiciloptilus 2A Ixobrychus exilis K minutus 3A 3S,K sinensis A flavicollis 2A ARDEINAE Tribe TIGRIORNITHINI Tigrisoma lineatum Ak,3A S Tribe BYCTICORACINI Gorsachius melanolophus Nycticorax nycticorax caledonicus 3A pileatus violaceus 6A Tribe COCHLEARIINI Cochlearius cochlearius A*,A Tribe ARDEINI Ardeola 2K 8S.4K K 2St 3S,St S,2K,St ralloides A S grayii 3S rufiventris A ibis 3A*,A 2S,K,2St Butorides virescens s,st striatus 10A 2S Hydranassa caerulea 2A 3S tricolor 2A*,4A S Egretta sacra 3A 2S gularis 2A* garzetta 2A* 2S intermedia 2A alba A 3S.3K Ardea purpurea 6S,K ANATOMICAL SPECIMENS OF BIRDS 131 ARDEINI (cont.) Ardea (cont.) Spirit Skels novaehollandiae 3A 2K,3S pacifica 2A cinerea 11A*,4A 13S cocoi 2S melanocephala A S goliath 4S imperialis st Balaeniceps rex Scopus umbrella BALAENICIPITIDAE 1 genus, 1 species SCOPIDAE 1 genus, 1 species A*,3A 3S,3K 3S,K CICONIIDAE 6 genera, 17 species 2 species unrepresented Tribe MYCTERHNI Mycleria ibis 3K leucocephala 2A 3S,St Anaslomus oscilans 2A 2S,3K lamelligerus 2S Tribe CICONHNI Ciconia nigra 2S,K abdimii 5A*,2A S episcopus 3S,K maguari 2S ciconia A*,A 4S,K Tribe LEPTOPTILINI Ephippiorhynchus asialicus A 4S senegalensis S,3K /a6/>« mycleria 4S Leploplilos javanicus A 2S,K,St ^w6/M5 4S,3K crumeniferus 3S,4K THRESKIORNITHIDAE 20 genera, 33 species 8 genera & 16 species unrepresented THRESKIORNITHINAE Threskiornis aelhiopica 2A 5S,4K melanocephala 5A 3S,K molucca 5A Carphibis spinicollis Pseudibis papillosa A* Spirit Skels Geronlicus calvus S Hagedashia hagedash A S Harpiprion caerulescens 2Ak K Therislicus caudalus Ak,A 2S melanopis A Eudocimus albus 2S ruber A*,5A 2S,St Plegadis falcinellus 2A 5S,S Lopholibis crislala A S PLATALEINAE Plalalea leucorodia A*,A 4S,4K,St regia 3A Ajaia ajaja 2S,St,K PHOENICOPTERIDAE 3 genera, 6 species 1 species unrepresented Phoenicopterus ruber 2A,A* 7S,St,2K chilensis A 2S,K Phoeniconaias minor A,Ak S,2K Phoenicoparrus andinus A jamesi 4Ak,4A S ANSERIFORMES ANHIMIDAE 2 genera, 3 species 1 genera & 1 species unrepresented Chauna torquata 3A 3S,St chavaria A*,Ak,2A S,3St ANATIDAE 43 genera, 146 species 24 species unrepresented ANSERANATINAE Anseranas semipalmata 4S ANSERINAE Tribe DENDROCYGNINI Dendrocygna eyloni A bicolor 2A* 2S arcuata S javanica 2K viduala A*,A 2S,K arborea S aulumnalis 2A S 132 j. S. BLANDAM ANSERINAE(cont.) Tribe ANSERINI Spirit Skels Cygnus (Cygnus) olor 3A*,2A 9S,St,4K atratus At 2S melanocoryphus Ak 4S (Olor) cygnus A 4S,7St,2K columbianus Ak 5S,St Coscoroba coscoroba A S Anser cygnoides A 2S.5K fabalis 5S,K,St albifrons 2A 68^,2*: anser Ak 4S,9K,St indicus S caerulescens 3A* K rossi S canagicus A brachyrhynchus 2A hyperboreus K Branta (Nesocheri) sandvicensis 2S (Branta) canadensis A 3S leucopsis A 7S,2St bernicla a,5A 6S,2K,St ruficollis 3A 2S Cereopsis novaehollandiae 5A* 5S,St Tribe STICTONETTINI Stictonetta naevosa A 2S ANATINAE Tribe TADORNINI Cyanochen cyanopterus A* 2S.K Chloephaga melanoptera S picta At,A 4S.32K hybrida A S poliocephala Ak,A 2S rubidiceps A S Neochen jubatus A S Alopochen aegyptiacus A*,A 2S,2K,St Tadorna (Casarcd) ferruginea A* 7S,St ca/w A*,2A S tadornoides 5S (Tadorna) tadorna 2A*,2A 4S radjah 9A S Spirit Tribe TACHYERINI Tachyeres patachonicus 6 A* brachypterus Tribe CAIRININI Plectropterus gambensis Cairina moschata scutulata Sarkidiornis melanotos A,2A Pteronetta nartlaubi Nettapus pulchellus 7A coromandelianus A auritus 2A Aix sponsa 3A*,2A galericulata a,A*,3A Chenonetta jubata 2A Amazonetta brasiliensis 2A*,A Tribe ANATINI Hymenolaimus malacorhynchus A Merganetta armata 2A Anas (Anas) penelope 3Ak,A americana sibilatrix 3A falcata A strepera A formosa spinicauda wyvilliana superciliosa A crecca A*, 4 A flavirostris A*,A capensis A gibberifrons A castanea aucklandica 2A platyrhynchos 4A*,5A rubripes melleri poecilorhyncha luzonica A specularis 2A specular ioides 2A flcwta 6A*,At,A georgica 2A bahamensis erythrorhyncha At Skels 2S S 5S S,2K S 3S K S,2K,St S 4S.K 3S 3S 2S S 6S.3K S 2S,2K 5S,K S S,St S 2S 8S,2K,St 3K S 2S 15S,18K,3St 3S S S,K S S,St K 2S S s,st 3S ANATOMICAL SPECIMENS OF BIRDS 133 ANATINI (cont.) Spirit Skels Anas (cont.) Spirit Skels Oxyura versicolor 2A S,K (Oxyura) querquedula A*,2A S,2K jamaicensis 4A discors 4A vittata 2S cyanoptera a,2A S Biziura clypeata A*,A 8S,K lobata 3S Malacorhynchus Thalassornis tnembranaceus 3Ak,A 2S leuconotos 2 A*, 2 A Marmaronetta angustirostris 2S,2K FALCONIFORMES CATHARTIDAE Tribe AYTHYINI 5 genera, 7 species Rhodonessa 2 species unrepresented caryophyllacea 3A s,st Cathartes Netta aura 2A 6S,K,St rufina A* st,s Coragyps erythrophthalma 6A* atratus A 4S peposaca A Sarcoramphus Aythya papa 2A 7S,2K,St valisineria S Gymnogyps ferina A 5S californianus 2S australis S Vultur nyroca 2S gryphus A,A* 4S.2K innotata A*,A novaeseelandiae S PANDIONIDAE fuligula 2A 7S 1 genus, 1 species marila At 6S,K Pandion haliaetus 2A,A* 4S,3St,K Tribe MERGINI Somateria ACCIPITRIDAE mollissima 2A*,7A 14S,K,2St 64 genera, 217 species spectabilis st 14 genera unrepresented Polysticta 106 species unrepresented stelleri Aviceda Histrionicus subcristata A histrionicus A 2S leuphotes St Clangula Leptodon hyemalis 8A*,5A 10S,K cayanensis K Melanitta Pernis nigra 4A 5S,St apivorus 2A 4S.2K perspicillata A Elanoides fusca 3A 4S,K forficatus A 2S Bucephala Machaerhamphus albeola A*,2A alcinus A S islandica 2S Elanus clangula 3A 9S leucurus A*, A S Mergus caeruleus 2A 7S,K,St (Lophodytes) notatus A cucullatus A Chelictinia (Mergellus) riocourii A albellus A 6S Rostrhamus (Mergus) sociabilis 4A senator 2A*,A 6S,St Harpagus merganser 3A 6S bidentatus K australis E,A 3S Ictinia plumbea 4S Tribe OXYURINI Milvus Heteronetta migrans A* 4S,3St,K atricapilla S milvus A*,A 3S,2K 134 J. S. BLANDAMER AND P. J. K. BURTON ACCIPITRIDAE Spirit (cont.) Skels Erythrotriorchis Spirit Skels Haliastur radiatus S sphenurus A 5S Accipiter imlus A 4S,St gentilis A.E.A* 6S,St,3K Haliaeetus henstii 2St leucogaster S,K gularis A vocifer A 3S,2K virgatus A K leucoryphus st cirrhocephalus 2S leucocephalus 4S,2K nisus E.8A.5A* 21S,3St,3K albicilla 6S,2St,K striatus S Ichthyophaga tachiro A 2S,St ichthyaetus St trivirgatus A 2S,2St Gypohierax fasciatus S angolensis A 8S,K novaehollandiae S Neophron soloensis S percnopterus 6S badius 4A 2S,St Gypaetus francesii s.st barbatus HS,4St,K cooperii S Necrosyrtes bicolor A St monachus A,Ak 2S,K Butastur Gyps liventer S bengalensis A 2S,2St teesa A st africanus K indicus S rueppellii 2S Kaupifalco himalayensis K monogrammicus 3A S,K fulvus A* 3S,2K,St Leucopternis Torgos polionota S tracheliotus Sarcogyps calvus 2A Aegypius monachus Trigonoceps occipitalis Circaetus 3S,K S 2S,2K,St 2S,K,St Buteogallus anthracinus urubitinga Heterospizias meridionalis Busarellus nigricollis A A A S 3S S gallicus E 4S Geranoaetus cinereus S melanoleucus 3S cinerascens S Parabuteo Terathopius unicinctus 4S ecaudatus A 4S Buteo Spilornis nitidus 2A 4S cheela 6S magnirostris 4A 3S Eutriorchis lineatus 2S astur A platypterus A Polyboroides swainsonii 3A typus A 2S albicaudatus A radiatus S,3St jamaicensis 2S.K Circus buteo 4A 6S.K assimilis S,St lagopus 4A*,A S aeruginosus At,4A* 5S,K,St rufinus 7S,K ranivorus S regalis S cyaneus A*,3A 4S,St rufofuscus A 5S,St cinereus A St,K Morphnus pygargus 2S,2St guianensis S Melierax Harpia met abates At S harpyja 3S canorus 2S Pithecophaga gabar 2A jefferyi A 2S ANATOMICAL SPECIMENS OF BIRDS 135 ACCIPITRIDAE (cont.) Spirit Skels Ictinaetus malayensis S Aquila pomarina S,2St clanga 2S,St rapax 4S.2K heliaca s,st wahlbergi S chrysaetos A 8S,St,3K audax A 3S,K,St verreauxi S Hieraaetus fasciatus A* 5S,2St pennatus E s,st morphnoides A Lophoaetus occipitalis 3S Spizaetus cirrhatus 2S nipalensis K ornatus S Stephanoaetus coronatus s,st Polemaetus bellicosus 3S,K Spirit Sagittarius serpentarius SAGITTARIIDAE 1 genus, 1 species At,A 10S,K FALCONIDAE 10 genera, 61 species 1 genus & 26 species unrepresented Daptrius ater S,K americanus Ak,3A Phalcoboenus megalopterus A S australis 2S,K Polyborus plancus 4A,A* 3S Milvago chimango 4A* 2S,K chimachima 3A Herpetotheres cachinnans Ak,3A S,K Micrastur ruficollis 2A Polihierax semitorquatus 7A S insignis A Microhierax caerulescens 9A S fringillarius A S melanoleucus A Falco naumanni 4A sparverius 13A tinnunculus 12A,13A* newtoni cenchroides 4A vespertinus A chicquera 2A columbarius 3A*,A berigora 4A novaezeelandiae A subbuteo 3A longipennis A eleonorae femoralis 2A biarmicus A mexicanus jugger A cherrug rusticolus A deiroleucus A peregrinus 8A Skels S,St,K 6S,St,K 21S,3St,2K S S 4S 2S 14S 2S s,st 6S,5K S St 3S S S,St 4S 5S,K S 14S,5St,K GALLIFORMES MEGAPODIIDAE 7 genera, 12 species 3 genera & 1 species unrepresented Megapodius freycinet A*,8A 2S pritchardii 5A*,A,t4A Leipoa ocellata S Alectura lathami Macrocephalon maleo 3S,K 2S CRACIDAE 8 genera, 44 species 1 genus & 18 species unrepresented Ortalis vetula S cinereiceps A garrula A ruficauda 2S canicollis 3K motmot 2S Penelope argyrotis A montagnii S marail 6A S superciliaris Ak 4S jacquacu A* S purpurascens 2S pileata 2S Aburria pipile Ak 136 Penelopina nigra Oreophasis derbianus Nothocrax urumutum Crax (Mitu) tomentosa mitu (Pauxi) pauxi (Crax) rubra alberti alector globulosa fasciolata J. S. BLANDAMER AND P. J. K. BURTON CRACIDAE (cont.) Spirit A" Ak 3A 2At PHASIANIDAE MELEAGRIDINAE 2 genera, 2 species gallopavo A Agriocharis ocellata TETRAONINAE 6 genera, 16 species 6 species unrepresented Dendragapus canadensis Lagopus lagopus mutus Tetrao tetrix A* 2A,5A 2A urogallus Bonasa bonasia umbellus 2A urophasianus Tympanuchus phasianellus cupido Hybrids: Tetrao tetrix x Phasianus colchicus At Lagopus lagopus x Tetrao tetrix 2 A* ODONTOPHORINAE 10 genera, 33 species 5 genera & 24 species unrepresented Lophortyx californica 3A* 2S,K > Spirit Skels Skels Philortyx fasciatus A St Colinus virginianus 4S 2S,2St leucopogon A cristatus S 2S Odontophorus gujanensis 3A guttatus A 2S,K Cyrtonyx 4S.2K montexumae A ocellatus S S PHASIANINAE 40 genera, 155 species 4S 6 genera & 69 species unrepresented 4S Tribe PERDICINI 3S Lerwa 3S 4S lerwa s,st HO Ammoperdix griseogularis A heyi A 3S Tetraogallus caspius S tibetanus A 6S,7K altaicus A S,K himalayensis 4S S Alec tor is graeca 2A*,6A 5S,St barbara 2S rufa 2A 3S,St,K melanocephala 2A S,2St S Francolinus francolinus A 5S,St 13S,13K pictus 2S,St HS,6K,St pintadeanus A S swainsonii A 8S,2St,8K leucoscepus jacksoni A S 4S,4SM3K ahantensis 2A* squamatus St 5S bicalcaratus K 2S clappertoni S natalensis S S capensis S sephaena A 2St levaillantii S 2S albogularis At,A pondicerianus A 2S,Sf gularis S Margaroperdix madagarensis Melanoperdix nigra Corturnix coturnix coromandelica 16A*,2A,A 6S,St,6K 3A A 6A 4S.SM5K 2A* ANATOMICAL SPECIMENS OF BIRDS 137 PERDICINI Spirit Skels Corturnix (cont.) Spirit Skels Phasianus delegorguei A colchicus 24A*,10At, 7S,16K,St pectoralis A 2S 10A Synoicus versicolor S ypsilophorus 4A*,4A 2S.K Chrysolophus Excalfactoria pictus A,A* 4S,2K adansonii St amherstiae A* S,K chinensis 7A s Polyplectron Perdicula chalcurum 3A S,K asiatica 3S germaini A Arborophila bicalcaratum 2A,3A* St torqueola S malacense A rufogularis A*,2At,2A emphanum 3A javanica S Rheinartia Tropicoperdlx ocellata A 4S,St charltonii A K Argusianus Rollulus argus 3A 3S rouloul 2A*,3A 6S,St Pavo Ptilopachus cristatus A,4A* 4S,St,K petrosus 5A muticus At 5S,K Bambusicola Afropavo fytchii 2A* congensis 2A S thoracica A,A*,2A S Hybrid: Galloperdix Gallus gallus x lunulata S Phasianus colchicus A bicalcarata 2S NUMIDINAE Tribe PHASIANINI 5 genera, 7 species Ithaginis 2 species unrepresented cruentus s,st Phasidus Tragopan niger St satyra A,At 6S,K Numida temmincki S meleagris 9A* 3At,2A 7S,K Pucrasia Guttera macrolopha A 2S edouardi 3S Lophophorus pucherani 3At impeyanus A 5S,St Acryllium Callus vulturinum A gallus lafayettei 33A*,At 34S,16K,5S S OPISTHOCOMIDAE sonnerati Lophura leucomelana 2A* 2A 2S 3S,St 1 genus, 1 species Opisthocomus hoazin 18A*,8A 4S,St nycthemera A* 3S,2K GRUIFORMES imperialis A,3A* edwardsi 3A*,3A MESITORNITHIDAE swinhoei 5A* S 2 genera, 3 species erythrophthalmus At 2S Mesitornis ignita A 3S variegata S diardi A* S unicolor At,3A S,K bulweri A* Manias Crossoptilon benschi 4A 2S mantchuricum 2A 2S,St TURNICIDAE auritum A 2S 2 genera, 14 species Catreus 7 species unrepresented wallichi 2S,St Turnix Syrmaticus sylvatica A*,6A S mikado 2A S tanki 3A* S,St soemmerringi 3S suscitator 8 A S reevesi A*,10t,2A 2S 138 J. S. BLANDAMER AND P. J. K. BURTON TURNICIDAE (cont.) Turnix (cont.) Spirit Skels nigricollis A*,A s.st varia 2A.5A* 3S velox 2A*,A 2S Ortyxelos meiffrenii A*,5A 2S PEDIONOMIDAE 1 genus, 1 species Pedionomus torquatus GRUIDAE 4 genera, 1 5 species 5 species unrepresented GRUINAE Grus grus A* S,2K japonensis A S,K americana S antigone 5A*,2A 5S rubicunda A* 2S,K leucogeranus S,K Bugeranus carunculatus A 2S Anthropoides virgo A* 7S paradisea A* 3S BALEARICINAE Balearica pavonina 2At,2A 7S,K Aramus guarauna Psophia crepitans viridis ARAMIDAE 1 genus, 1 species 2A S PSOPHHDAE 1 genus, 3 species species unrepresented A a,2A 3S RALLIDAE 53 genera, 141 species 18 genera & 79 species unrepresented RALLINAE Rallus (Rallus) longirostris S elegans S limicola A aquaticus 9A*,27A 5S,3St,4K madagascariensis S mirificus* A striatus A,6A* (Hypotaenidia) philippensis 7 A*, 2 A 2S,K torquatus A S Spirit Skels Atlantisia rogersi A,a S Ortygonax sanguinolentus A nigricans 2A Tricholimnas sylvestris 2S Pardirallus maculatus A Dryolimnas cuvieri 2A 3S,St Rallina fasciata 8A St eurizonoides A Ar amides cajanea 4A s,st ypecaha A*,4A 3S,K,St calopterus S Nesoclopeus poeciloptera E ?,A Gallirallus australis 2A*,3A 4S Himantornis haematopus A Canirallus oculeus 3A Mentocrex kioloides 2A S,2St Crecopsis egregia A K Crex crex 7A,8A* S,K,St Limnocorax flavirostra A*,4A 2S,K Porzana pava A S pusllla A 2S porzana A 4S Carolina A,3A* fusca A.3A* 2St tabuensis A Porzanula palmeri E,A*,2A S Pennula millsi E,4A*,3A sandwichensis E,4A*,3A Nesophylax ater 2A Laterallus viridis 3A leucopyrrhus S Coturnicops noveboracensis K Sarothrura pulchra A*,2A insularis A Poliolimnas cinereus A ANATOMICAL SPECIMENS OF BIRDS 139 RALLINAE (cont.) Porphyriops melanops Tribonyx vent rails mortierii Amaurornis phoenicurus Gallicrex cinerea Gallinua chlloropus angulata Porphyriornis nesiotis comeri Pareudiastes pacificus Porphyrula alleni martinica Porphyrio porphyrio madagascariensis A poliocephalus A Notornis mantelli FULICINAE Fulica atra 5A*,3A 6S,2K americana 7A* leucoptera A,A HELIORNITHTOAE 3 genera, 3 species Podica senegalensis At,4A S Heliopais personata 4A* Heliornis fulica 11A S OTIDIDAE Spirit Skels 11 genera, 24 species 10 species unrepresented S Spirit Skels Tetrax 3S tetrax 3S,K 2A 3S,S* Otis tarda A*,2t,3A 3S,K A*,3A 2S Neotis nuba K Ak,4A Choriotis arabs K 30A*,At,9A 7S,2St,5K kori K A australis A 2S,K Chlamydotis S* undulata 3S,K A*,At,3A S,4St Lophotis ruficrista A S Afrotis E ?,A atra S Eupodotis 4A 3S vigorsii S 8A S senegalensis At,A S Lissotis A,3A* melanogaster K A Houbaropsis A 7S bengalensis S Sypheotides 2S indica S* Rhynochetos jubatus Eurypyga helias Cariama cristata Chunga burmeisteri RHYNOCHETIDAE 1 genus, 1 species Ak,A*,2A 5S EURYPYGIDAE 1 genus, 1 species 2A 2S CARJAMIDAE 2 genera, 2 species Ak,A 6S,St A 2S CHARADRIIFORMES Suborder CHARADRII JACANIDAE 6 genera, 8 species 1 genus & 1 species unrepresented Actophilornis africana A*,8A S albinucha 2S Irediparra gallinacea 8A Hydrophasianus chirurgus 4A st Metopidius indicus 6A 2S,2St Jacana spinosa 4A*,7A 3S jacana 2A*,3A ROSTRATULIDAE 2 genera, 2 species 1 genus & 1 species unrepresented Rostratula benghalensis 4A*,12A 3S,3K,3St Dramas ardeola DROMADIDAE 1 genus, 1 species 4A*,5A*, S,2K 2At,2A 140 J. S. BLANDAMER AND P. J. K. BURTON HAEMATOPODIDAE 1 genus, 7 species CHARADRIIDAE 8 genera, 64 species 5 species unrepresented 2 genera & 1 5 species unrepresented Spirit Skels Spirit Skels Haematopus Vanellus ostralegus A*,61A 18S,15K,2St vanellus 14A*,13A 6S leucopodus 2A K crassirostris 2A armatus 4A IBIDORHYNCHIDAE spinosus A 1 genus, 1 species tectus 3A Ibidorhyncha malabaricus A K struthersii 3A*,A 2S,2K,St albiceps Ak lugubris A RECURVIROSTRIDAE coronatus 3A,2A* 3 genera, 13 species senegallus 2A*,A 5 species unrepresented leucurus A Himantopus cayanus 2A himantopus 10A 6S,3K chilensis 13A*,6A mexicanus 3A indicus Ak,2A leucocephalus 2A tricolor 5A 3S,K Cladorhynchus miles 4A leucocephala 2S,K Recurvirostra Pluvialis avosetta 5A 3S,2K apricaria 35A 7S,4St,8K . sominica 8A 6K americana A novaehollandiae 2A K,S squatarola 20A 5S,3St,4K andina K obscura 2K Charadrius BURHINIDAE hiaticula 4A*,17A 5S,9St,3K 2 genera, 9 species semipalmatus 2A 1 genus & 4 species unrepresented placidus K Burhinus dubius 7A*,2A 4K oedicnemus 7A*,2A 7S,St,3K wilsonia 4A 3S,St,K senegalensis A,A* vociferus 4A 3K vermiculatus A melodus 2K capensis S pecuarius 2A*,10A K magnirostris 3A 2S,5K,2St sanctaehelenae 2S,K tricollaris 8A S,K GLAREOLIDAE alexandrinus 9A 5K.3S 5 genera, 8 species marginatus 2At 4 species unrepresented peronii 3A 2K CURSORIINAE collaris K Pluvianus bicinctus A K aegyptius 2A*,8A 2S,K falklandicus 2A*,3A K Rhinoptilus mongolus A 4S,2K africanus A leschenaultii 7A,4Ak 4K cinctus A asiaticus 2A 2K chalcopterus A* ,3 A S veredus K Cursorius modestus 4A K cursor A,A* S,2K melanops 2A*,23A S coromandelicus A S,2K cinctus 4A K temminckii 4A K rubricollis K GLAREOLINAE Anarhynchus S/tf/fo Isabella 5A frontalis A S Glareola Peltohyas pratincola 2A,A* 2S,K australis A*,6A maldivarus 2S,K Eudromias nuchalis A*,5A morinellus 2A* St,K cinerea A* S ruficollis 4A 2S,K ANATOMICAL SPECIMENS OF BIRDS 141 SCOLOPACIDAE 9 genera, 31 species 7 species unrepresented TRINGINAE Tribe NUMENIINI Limosa limosa haemastica lapponica fedoa Numenius minutus borealis phaeopus arquata americanus Bartramia longicauda Tribe TR1NGIN1 Tringa (Totanus) erythropus totanus stagnatilis nebularia melanoleuca flavipes (Tringa) ochropus solitaria glareola Catoptrophorus semipalmatus Xenus cinereus Act it is hypoleucos macularia Heteroscelus brevipes incanus Spirit 8A E ? A*,6A 13A*,33A 3A*,5A 11 A*, A 31A 3A A*,4A 3A 4A 5A 3A 3A*,4A 4A 3A 3A*,16A 9A 4A 4A Tribe PROSOBONIINI Prosobonia cancellata E,2A ARENARIINAE 1 genus, 2 species 1 species unrepresented Arenaria interpres 20A PHALAROPODINAE 1 genus, 3 species Phalaropus tricolor A lobatus 4A fulicarius 5A Skels 8S 12S,6K,2St K K K 8S,5K,St 10S,3St,5K S,2K S,2St 7S,5K,St S,4K 2K S 2S,3K,6St S St 2S,K 3K St,2K S,K 2S,K St,K 3S,9St,2K 3S,K S,K SCOLOPACINAE 1 genus, 6 species 3 species unrepresented Scolopax Spirit Skels rusticola 8A*,15A 4S,4St,14K saturata 2A minor A GALLINAGONINAE 4 genera, 20 species 8 species unrepresented Coenocorypha aucklandica Ak S Gallinago stenura 3A K megala A macrodactyla S media S,St,K gallinago 7A*,39A 7S,2St,6K paraguaiae A nobilis 2A undulata K Lymnocryptes minimus 20A 5S,K,St Limnodromus griseus 7A 3K semipalmatus K CALIDRIDINAE 7 genera, 24 species 1 species unrepresented Aphriza virgata 2A S Calidris canutus A*,33A 8S,2K,St tenuirostris A K alba 11A 4S,K pus ilia 12A S,2K mauri 2A S ruficollis 7A 3S minuta 7 A, 5 A* 3S,K temminckii 2A St subminuta 5A minutilla 2A K fuscicollis 5A bairdii A melanotos 2A S acuminata 8A S maritima 4A*,10A 4S,K alpina 2A*,71A 12S,9K ferruginea 2A S Eurynorhynchus pygmeus 2A S Limicola falcinellus 2A K Micropalama himantopus 8A K Tryngites subruficollis 7A K Philomachus pugnax 1 2 A* , 1 0 A 11S.6K 142 J. S. BLANDAMER AND P. J. K. BURTON THINOCORIDAE Spirit Skels 2 genera, 4 species Creagrus 1 species unrepresented furcatus 2S Spirit Skels Xema Attagis sabini At 2S gayi K Thinocorus STERNINAE orbignyianus 4A 2S,K rumicivorus 10A 2S,K,St Chlidonias bybrida leucoptera A 2S CHIONIDIDAE nigra 6A 2S 1 genera, 2 species Hydroprogne r>* Chionis caspia St alba Ak,9A 4S,St,3K minor 9A*,7A S,K Sterna hirundo 9A*,4A 5S,2St paradisaea 2A* 12S.2K Suborder LARI vittata 19A*,3A STERCORARIIDAE dougallii 3A,A* 2 genera, 5 species striata 3S 1 species unrepresented repressa 7S Caiharacta sumatrana 2S skua 18A*,2A 6S.3K anaethetus 8A*,5A Stercorarius fuscata A,A* 3S pomarlnm A albifrons 2S parasiticus 2A*,3A 6S,K,2St Thalasseus longicaudus A*,A bergii 5A* 5S,3K maximus 2A LARIDAE bengalensis S 1 7 genera, 90 species sandvicensis A,A* 6S,2K 3 genera & 39 species unrepresented Larosterna LARINAE inca A 2S Gabianus Procelsterna pacificus S cerulea S scoresbii 2A S Anous Pagophila stolidus 15A*,4A 5S,St alba 4A*,2A tenuirostris S Larus minutus 2A modestus 7A* Gygis hemprichii 2A* S alba A*,11A 6S delawarensis 2A canus 5S,St,5K argentatus A,A* 12S,2St,4K fuscus 3A*,4A 3S dominicanus At S marinus At,A,A* 6S,7K hyperboreus 2A* 5S,2K,St glaucoides 2S ichthyaetus 3S atricilla S brunnicephalus K cirrocephalus A* novaehollandiae A 4S,K melanocephalus 2S maculipennis A S ridibundus 4A*,7A 11S.4K genei A 4S,K Philadelphia S minutus A S 10A*,5A 7S,St,6K Rynchops niger albicollis RYNCHOPIDAE 1 genus, 3 species 1 species unrepresented 4S At,2A tridactyla Suborder ALCAE ALCIDAE 13 genera, 23 species 3 genera & 7 species unrepresented Alle alle Alca torda Uria lomvia 2S aalge 4A*,2At,5A 17S,St,9K Cepphus grylle A HS.St carbo A At,15A 8S,3St,K 7A*,5A 17S.SM2K ANATOMICAL SPECIMENS OF BIRDS 143 ALCIDAE (cont.) Streptopelia Spirit Skels Spirit Skels turtur 5A 5S,St Synthliboramphus orientalis A antiquus A bitorquata 28 Ptychoramphus decaocto A*,5A 78 aleuticus A roseogrisea S Cyclorrhynchus decipiens A,A* psittacula 8,8* semitorquata 2A*,2A 2S Aethia capicola 4A cristatella A 8 vinacea S pusilla A chinensis 3A 4S,St pygmaea S senegalensis 3A S,K Fratercula Aplopelia arctica 3A*,5At,5A 118,8* larvata St corniculata A 8 Macropygia Lunda unchall S cirrhata 2A S,K amboinensis A* phasianella A S COLUMBIFORMES ruficeps 28 PTEROCLIDIDAE Turacoena 2 genera, 16 species manadensis S 9 species unrepresented Turtur Syrrhaptes chalcospilos 4A paradoxus A 48 abyssinicus A Pterocles afer 10A S alchata 2A* 48 tympanistria 5A 48 exustus 2A* St,6K brehmeri A senegallus 2A* 28* Oena orientalis 38 capensis 8A S lichtensteinii 2A 28 Chalcophaps quadricinctus 6A 8 indica 2A*,14A 9S,K stephani 3A RAPHIDAE Henicophaps 2 genera, 3 species albifrons A 1 genus & 1 species unrepresented Phaps /?OP/ZW5 chalcoptera 4A S cucullatus S elegans 2S solitarius 8* historionica S Ocyphaps COLUMBIDAE lophotes 4A 3S 42 genera, 303 species Petrophassa 6 genera & 180 species unrepresented plumifera 7A S Columba ferruginea A livia A 54S,4St,5K scripta A* guinea 3A 38 smithii 7A oenas A*,2A 2S,St,4K rufipennis A palumbus 6A*,3A 18S,3St,3K albipennis 4A S /rocaz 2A 8,8* Geopelia unicincta A cuneata 12A 68 arquatrix S striata A*,8A 48 hodgsonii 8* humeralis 7A K,S leucocephala A S Leucosarcia squamosa S melanoleuca S picazuro S Ectopistes maculosa S migratorius E S fasciata A Zenaida nigrirostris A auriculata 6A*,A S Atesoe/ia.s aurita S,2K mayeri 2A* galapagoensis A 144 COLUMBIDAE (cont.) Spirit Columbina passerina A*,7A minuta 2A talpacoti 5A picui 3A Claravis pretiosa 2A Metriopelia melanoptera Scardafella squammata 2A,A* Leptotila verreauxi 4A rufaxilla A wellsi jamaicensis cassini A Geotrygon versicolor A chrysia A violacea A montana Starnoenas cayanocephala Caloenas nicobarica A Gallicolumba luzonica 2A criniger rubescens A beccarii Otidiphaps nobilis A*,3A Goura cristata scheepmakeri A victoria Didunculus strigirostris 6A Phapitreron leucotis Treron fulvicollis olax 3A vernans 3A bicincta A pompadora curvirostra 5A phoenicoptera waalia A australis 2A calva 3A apicauda 2A oxyura sphenura Ptilinopus cinctus A J. S. BLANDAMER AND P. J. K. BURTON Skels 2S S 2S 2S 3S S S S S 3S S S 2S,3St,2K S,2St s,st st S S,St,K 3S.2K S S St 2S st S St Ptilinopus (cont.) Spirit Skels jambu 4A magnificus 2A ornatus A superbus A porphyraceus 2A*,3A rarotongensis A roseicapilla 2S regina A S greyii 2A dupetithouarsii A 3S coronulatus A* S puchellus S rivoli S melanospila 3A,A* S Alectroenas madagascariensis A pulcherrima A S,St Ducula badia 2S forsteni S aenea A 5S galeata S whartoni 2S lacernulata 2S bicolor At 3S,St spilorrhoa 6A S Lopholaimus antarcticus S Hemiphaga novaeseelandiae 2A PSITTACIFORMES LORIIDAE 81 genera, 344 species 18 genera & 181 species unrepresented Chalcopsitta atra A sintillata A S Eos squamata S bornea S Trichoglossus ornatus 2S haematodus 9A 3S,K flavoviridis A chlorolepidotus 3A euteles S versicolor 2A S Lorius lory S domicellus S garrulus 2S Phigys solitarius A Vini australis A peruviana S ultramarina 2S ANATOMICAL SPECIMENS OF BIRDS 145 LORIIDAE (cont.) Spirit Skels Glossopsitta concinna 4A 3S pusilla 3A porphyrocephala 3A Charmosyna palmarum A placentis 4A 2S papou 2A Neopsittacus musschenbroekii A CACATUIDAE CACATUINAE Probosciger aterrimus 2S,3K Calyptorhynchus funereus 1 3 A 10S,4K magnificus 3A 2S Callocephalon fimbriatum 2S,St Eolophus roseicapillus 3 A 3S Cacatua leadbeateri 2S,K sulphurea 2S.K galerita 3S,K moluccensis 2S,St alba 2S sanguinea 3 A 4S tenuirostris S ducorps NYMPHICINAE Nymphicus hollandicus 7A 4S,St PSITTACIDAE NESTORINAE Nestor notabilis 2A 2S meridionalis 2A S MICROPSITTINAE Micropsitta keiensis S finschii 6A PSITTACINAE Pacific Taxa Opopsitta diophthalma 3A S Psittinus cyanurus 5A S Geoffroyus geoffroyi A Prioniturus luconensis A Tanygnathus megalorynchos A lucionensis S Eclectus Spirit roratus 4A Psittrichas fulgidus Ak,2A Prosopeia tabuensis A Alisterus scapularis 4A chloropterus 4A Aprosmictus erythropterus 2A*,7A jonquillaceus Polytelis swainsonii anthopeplus A alexandrae A,2A* Purpureicephalus spurius 3A Barnardius barnardi zonarius 5A Platycercus caledonicus elegans eximius 7A adscitus 7A venustus A*,3A icterotis 4A Psephotus haematonotus 2A varius 3A,A* Cyanoramphus unicolor A novaezelandiae A auriceps Eunymphicus cornutus A Neophema bourkii 2A chrysostoma 2A,At elegans 8A pulchella 3A splendida A,A* Lathamus discolor 6A Melopsittacus undulatus 20A Pezoporus wallicus Afro-Asian Taxa Coracopsis vasa A nigra Psittacus erithacus 2A Poicephalus robustus 3A cryptoxanthus A senegalus A Skels 4S St 3S 3S S S 4S S 6S,K 5S K,2S 2S,St,K S,K 2S 3S 2S K S St 3S K 2S S 10S,2St,2K K S 2S 9S,2K S s,st 146 J. S. BLANDAMER AND P. J. K. BURTON PSITTACINAE (cent.) Poicephalus (cont.) Spirit Skels rufiventris S meyeri 2A 3S flavifrons A Agapornis cana 2A 2S pullaria 2A 5S,St,K taranta 5A 2S roseicollis 4A 3S personata 2A*,16A 6S,2St lilianae 3A nigrigenis A Loriculus vernalis A St beryllinus 3K galgulus 2A S aurantiifrons St Psittacula eupatria A* S krameri 3A 22S,2St,S* himalayana A cyanocephala 3A 12S,4St columboides S calthorpae 2A S alexandri 4A 23S,7St longicauda 2A S echo A New World Taxa Anodorhynchus hyacinthinus A 5S Cyanopsitta spixii S Ara ararauna S militaris 2S,K macao 4S,K chloroptera S,2St auricollis A manilata A St maracana S nobilis A 3S Aratinga acuticaudata A 3S guarouba 2A.A* holochlora S mitrata S erythrogenys A 2S leucophthalmus 3A 2S pertinax 3A cactorum A Nandayus nenday S Conuropsis carolinensis E 2S Cyanoliseus patagonus At S Pyrrhura cruentata frontalis leucotis picta Enicognathus ferrugineus leptorhynchus Myiopsitta monachus Bolborhynchus aymara Forpus cyanopygius passerinus coelestis Brotogeris versicolurus pyrrhopterus sanctithomae Pionites melanocephala Pionus menstruus maximiliani seniloides fuscus Amazona leucocephala ventralis albifrons xantholora viridigenalis dufresniana festiva aestiva ochrocephala amazonica farinosa guildingii Deroptyus accipitrinus Triclaria malachitacea STRIGOPINAE Strigops habroptilus Spirit 3A A A 2A 7A A 2A A A A 3S A A 2A Skels S S S 2S S S 2S S S 3S S S 2S 2S 3S S S 4S,K 3S 4S 3S S,ST 4S,St Corythaeola cristata Crinifer africanus CUCULIFORMES MUSOPHAGIDAE 5 genera, 19 species 9 species unrepresented 2A 2S,St 2S,2St ANATOMICAL SPECIMENS OF BIRDS 147 MUSOPHAGIDAE (cont.) Spirit Skels Corythaixoides concolor A*,4A personata A leucogaster A Musophaga violacea 4A 2S Tauraco corythaix 2A schuetii A macrorhynchus 2A S leucotis 2A S porphyreolophus A persa A* 4S livings tonii 2 A St CUCULIDAE 38 genera, 130 species 13 genera & 80 species unrepresented CUCULINAE Clamator glandarius A,At 2S coromandus 2A* st jacobinus 2A cafer K Cuculus varius A clamosus A canorus 3A*,35A 8S,5St,2K poliocephalus A pallidus 7A 2S Cacomantis merulinus 3A 2S,2St variolosus A* pyrrhophanus 5A Misocalius osculans S Chrysococcyx cupreus 4A klaas 3A 2S caprius 4A 2S Chalcites xanthorhynchus St basalts 4A 4S lucidus A S Surniculus lugubris 2A Eudynamys scolopacea 7A 2S,K Vrodynamis taitensis 2A Scythrops novaehollandiae A St PHAENICOPHAEINAE Coccyzus erythropthalmus 2S americanus A 2S minor 2A melacoryphus 3A Spirit Skels Piaya (Piayd) cayana 4A 4S,K (Coccycua) minuta A Saurothera vetula K Ceuthmochares aereus 3A Rhopodytes diardi S tristis A S Taccocua leschenaultii Rhinortha chlorophaea 2A S Zanclostomus javanicus 4S Rhamphococcyx calyorhynchus 3S curvirostris 2A 4S CROTOPHAGINAE Crotophaga major 3At,3A S ani 12A S,2K sulcirostris A Guira guira 2A*,5A 3S NEOMORPHINAE Geococcyx californiana A 4S,St Carpococcyx radiceus A renauldi A S COUINAE Coua cristata 3A caerulea A S CENTROPODINAE Centropus phasianius 5A 4S sinensis 5A*,A 5S toulou 2A 3S bengalensis 3A S monachus A,A* K senegalensis A S,K superciliosus A,4A* S STRIGIFORMES TYTONIDAE 2 genera, 12 species 8 species unrepresented TYTONINAE Tyto alba 4A*,11A 17S,3K,St novaehollandiae S,K capensis S 148 J. S. BLANDAMER AND P. J. K. BURTON TYTONIDAE (cont.) Spirit Skels PHODILINAE Phodilus badius A K Speotyto cunicularia STRIGIDAE 27 genera, 134 species 1 1 genera & 83 species unrepresented BUBONINAE Otus rufescens A 2S spilocephalus 3A 6S scops 2A*,9A 6S,2St senegalensis 5A rutilus 3S insularis S bakkamoena 4A 4S,St asio S nudipes S leucotis A S virginianus 4A*,3A 6S bubo 4A*,6A 10S,St,2K capensis 2S africanus A*,4A 5S,K nipalensis 2St sumatrana A lacteus A S coromandus s,st blakistoni S flavipes St ketupu 5S,St (Pulsatrix) perspicillata 3A 3S Nyctea scandiaca 2A,A* S,K ulula 2A 5S Glaucidium brodiei S passerinum A 2S jardinii A brasilianum 5A s,st perlatum 2A radiatum St cuculoides 2A S,2St connivens A novaeseelandiae 6A 5S scutulata 5A S,2St philippensis S Sceloglaux albifacies 2At brama A*, ISA 3 A 15S,3K,St Spirit Skels 3A*,12A 8S woodfordii 5A STRIGINAE seloputo leptogrammica aluco Ak,A*,4A S St 9S,7K 2S uralensis A 2S nebulosa 2A 2St Rhinoptynx clamator Asio otus madagascariensis flammeus Aegolius funereus Steatornis caripensis 2A*,3A 5A*,9A 5A 9S.4K s,st 10S,K,3St 2S CAPRIMULGIFORMES STEATORNITHIDAE 1 genus, 1 species 12A*,5A 2S PODARGIDAE 2 genera, 13 species 8 species unrepresented strigoides 7A 3S,3K papuensis 2A S Batrachostomus auritus 5A harterti A stellatus 3A Nyctibius griseus Aegotheles cristatus NYCTmilDAE 1 genus, 5 species 4 species unrepresented 2A K AEGOTHELIDAE 1 genus, 8 species 7 species unrepresented 3A CAPRIMULGIDAE 19 genera, 76 species 10 genera & 54 species unrepresented CHORDEILINAE Chordeiles acutipennis 2A S minor 3A S Podager nacunda 3A ANATOMICAL SPECIMENS OF BIRDS 149 CAPRIMULGIDAE (cont.) Spirit Skels CAPRIMULGINAE Eurostopodus guttatus 2A S mystacalis s temminckii S Nyctidrornus albicollis 3A S,K Phalaenoptilus nuttallii A S Caprimulgus vociferus S ruficollis A indicus A,A* st europaeus 2A*,7A 3S,2K,St madagascariensis A S macrurus 2A St pectoralis A natalensis A affinis S enarratus 3S,St Scotornis climacurus 3A,A* 2K Macrodipteryx longipemis 5A*,6A 2S Semeiophorus vexillarius A,4A* APODIFORMES APODIDAE 18 genera, 82 species 1 1 genera & 65 species unrepresented CYPSELOIDINAE Streptoprocne (Streptoprocne) zonaris 3A APODINAE Collocalia (Aerodramus) francica vanikorensis fuciphaga maxima (Collocalia) esculenta Raphidura leucopygialis Hirundapus caudacuta gigantea Chaetura (Chaeturd) pelagica Cypsiurus parvus A*,8A 4A 21A 5A 13A A A A 2A 4A*,7A S St 3S Apus (Tachymarptis) melba aequatorialis apus Spirit Skels 10A pacificus A coffer 2A*,A affinis 2A* S HEMIPROCNIDAE 1 genus, 4 species 2 species unrepresented St S 12S,2K,25*, 4S S Hemiprocne longipennis comata 2A 5A 3S TROCHILIDAE 116 genera, 338 species 75 genera & 274 species unrepresented Clauds hirusuta Threnetes ruckeri Phaethornis superciliosus hispidus longuemareus bourcieri Campylopterus rufus Eupetomena macroura Melanotrochilus fuscus Colibri delphinae thalassinus serrirostris Anthracothorax prevostii Eulampis jugularis Sericotes holosericeus Chrysolampis mosquitus Orthorhyncus cristatus Abeillia abeillei Lophornis ornata magnifica Discosura longicauda Chlorestes notatus 4A 7A A A A 4A 2S 2S 2A 2A 2A 9A A 30A 5A S 3S K 150 J. S. BLANDAMER AND P. J. K. BURTON TROCHILIDAE (cont.) Spirit Spirit Skels Chlorostilbon aureoventris 23A K ricordii A S Thalurania furcata 16A S watertonii A glaucopis A Panterpe insignis S Damophila julie S Hylocharis (Hylocharis) sapphirina A cyanus A chrysura A Trochilus polytmus s,i Amazilia (Polyeratd) versicolor 43A fimbriata S lactea A (Saucerottia) beryllina 2A (Amazilia) tzacatl st Urochroa bougueri A Patagona gigas A Lafresnaya lafresnayi 8S Coeligena coeligena A torquata A 11! lutetiae 8S iris 3S Ensifera ensifera 5S Sephanoides sephaniodes A Boissonneaua ma thews ii 2A 2S Heliangelus strophianus S exortis 8S viola S Eriocnemis vestitus 8S luciani S Ocreatus underwoodii 2A Lesbia victoriae S nuna 2A Metallura tyrianthina ' 5S Aglaiocercus kingi Heliomaster furcifer A Archilochus colubris 5A Calliphlox amethystina 13A Calypte anna Aces tr lira mulsant A Selasphorus rufus 2A Skels S Colius (Colius) striatus castanotus colius (Urocolius) indicus macrourus COLIIFORMES COLIIDAE 1 genus, 6 species 1 species unrepresented 4A*,14A A 4A 2A 2A 9S,K S 3S TROGONIFORMES TROGONIDAE 8 genera, 37 species 4 genera & 1 8 species unrepresented Pharomachrus mocinno A K fulgidus S auriceps S,K Trogon (Curucujus) massena A melanurus A Trogon) viridis A citreolus A S (Trogonurus) collaris 1 8 A rufus A j«m/cwa curucui violaceus Apaloderma narina A Harpactes reinwardtii kasumba 4A diardii 2A duvaucelii 4A erythrocephalus At,5A Mwdf A S,K S 2S S 4S 6S S,2Sf ANATOMICAL SPECIMENS OF BIRDS 151 CORACIIFORMES ALCEDINIDAE 14 genera, 90 species 1 genus & 43 species unrepresented Spirit Skels CERYLINAE CERYLE (Megaceryle) maxima A S torquata 5A,A* st alcyon 3A 2S,St (Ceryle) rudis 2At,5A St,K Chloroceryle amazona 4A S americana 13A S inda 3A S aenea A ALCEDININAE Alcedo (Alcedo) atthis 5A 4S,2S meninting 48A S coerulescens 19A ( Corythornis) cristata 12A 2S,St leucogaster 4A hpidina picta 21A S madagascariensis 2A Ceyx lepidus 3S azureus 5A S erithacus 17A S rufidorsum 4A 2S DACELONINAE Pelargopsis capensis 5A S,2S Lacedo pulchella A Dacelo novaeguineae 6A,A* 8S leachii 3A 2S gaudichaud 2A S Clytoceyx rex A Melidora macrorrhina A Cittura cyanotis S Halcyon coromanda A badia A smyrnensis 6A S2t,K pileata A cyanoventris 4A leucocephala 4A*,2A senegalensis 2A*,6A S malimbica 3A Halcyon (cont.) Spirit Skels albiventris 2A* chelicuti 7A macleayii 6A K leucopygia A farquhari 3A pyrrhopygia 3A 4S sancta 29A 4S,K Moris 34A 2S saurophaga A recurvirostris 3A*,3A concreta 3A 5S Tanysiptera galatea 2A Todus todus TODIDAE 1 genus, 5 species 4 species unrepresented 10A S MOMOT1DAE 6 genera, 9 species 2 genera & 4 species unrepresented Aspatha gularis 2K Electron platyrhynchum 3A Baryphthengus ruficapillus S.2K martii A Momotus momota 6A 3S,K MEROPIDAE 3 genera, 24 species 1 genus & 9 species unrepresented Nyctyornis amicta 3A*,7A S,K athertoni S* Merops gularis 2A muelleri A bulocki A pusillus 1 8 A variegatus 7A hirundinaceus A albicollis 9A S viridis ISA* superciliosus A*,5A 2S ornatus 5A*,5A 4S apiaster 2A,3A S,K malimbicus A nubicus S Coracias garrulus abyssinica caudata benghalensis CORAC1IDAE 2 genera, 1 1 species 4 species unrepresented A* 3A 2A,2A* A 2S,K 2St 152 J. S. BLANDAMER AND P. J. K. BURTON Eurystomus glaucurus gularis orientalis CORACIIDAE (cont.) Spirit Skels 5A 2A,A* A*,9A 2S 6S,2St BRACHYPTERACIIDAE 3 genera, 5 species Brachypteracias leptosomus A squamigera A Atelornis pittoides 2A 2S cross ley i S* Uratelornis chimaera A Leptosomus discolor Upupa epops LEPTOSOMATIDAE 1 genus, 1 species 2A UPUPIDAE 1 genus, 1 species S,St Anthracoceros Spirit Skels malayanus S,K malabaricus 3A* 6S,St coronatus 2S Bycanistes bucinator A 2S.K cylindricus A subcylindricus A* Ceratogymna atrata 2S,K elata A Buceros rhinoceros 4A* 2S.2K bicornis A 2S,5K hydrocorax A 2S Rhinoplax vigil St,K Bucorvus abyssinicus 2A 4S,2K leadbeateri A,At 2K Bycanistes brevis 7A,At S.K A,11A,5A* 2S,2K,2St PHOENICULIDAE 2 genera, 8 species 4 species unrepresented Phoeniculus pur pure us 2 A 3S bollei A aterrimus A PICIFORMES GALBULIDAE 5 genera, 17 species 1 genus & 10 species unrepresented Galbalcyrhynchus leucotis A S Jacamaralcyon tridactyla 3A S Galbula Rhinopomastus galbula 3A S cyanomelas A S ruficauda 2A*,8A BUCEROTIDAE leucogastra 2S 12 genera, 44 species dcu 8A 1 genus & 14 species unrepresented Jacamerops Tockus aurea 4A birostris K fasciatus A K alboterminatus 4A* BUCCONIDAE nasutus A*,7A 2S 10 genera, 34 species hemprichii K 2 genera & 20 species unrepresented griseus 2A K Notharchus camurus 3A macrorhynchos 3 A 2S erythrorhynchus 3A S,St tectus A deckeni A* Nystalus Berenicornis radiatus 3A comatus 2A chacuru A S Anorrhinus maculatus A galeritus 3A* S Malacoptila Penelopides striata S panini 2S fusca S Acer os panamensis 3A nipalensis K Micromonacha corrugatus S lanceolata S undulatus 2A* S.S'.K Nonnula plicatus 3S,3K ruficapilla 3A ANATOMICAL SPECIMENS OF BIRDS 153 BUCCONIDAE (cont.) Monasa atra nigrifrons morphoeus Chelidoptera tenebrosa Spirit A A 8A 9A Skels CAPITONIDAE 13 genera, 80 species 1 genus & 32 species unrepresented Capita niger A S Eubucco richardsoni S bourcierii A Semnornis 3S 2S,K K 2St,2K 2S 5S 2S S.2K S S 3S 2S,2St 2S,St 5S 3S frantzii A ramphastinus 3A Psilopogon pyrolophus 3A Megalaima virens A lagrandieri 2A zeylanica At,8A viridis A faiostricta 3A corvina chrysopogon 2A rafflesii A mystacophanos 3A javensis flavifrons A franklinii A Megalaima oorti A asiatica 3A australis 3A rubricapilla 2A haemacephala 5A henrici Calorhamphus fuliginosus 2A Gymnobucco peli 6A bonapartei 2A Smilorhis leucotis 2A*,2A Pogoniulus duchaillui 3A scolopaceus 3A*,2A chrysoconus A bilineatus 9A subsulphureus 2A Tricholaema lacrymosum 2A diadematum A melanocephalus 2A hirsutum A 2S Lybius vieilloti torquatus guifsobalito leucocephalus melanopterus bidentatus dubius Trachyphonus purpuratus vaillantii darnaudii margaritatus Spirit 3A 3A A 2A A 3A 2A A,At 2A 2A Skels 4 genera, 14 species 2 genera & 9 species unrepresented Prodotiscus insignis a regulus At Indicator maculatus A indicator 3A K minor A S archipelagicus A Melichneutes robustus A RAMPHASTIDAE 6 genera, 33 species 15 species unrepresented A ulacorhynchus prasinus A S,St Pteroglossus viridis 4 A 3S inscriptus 3 A flavirostris A aracari A S castanotis Ak,2A torquatus 2A S Selenidera maculirostris 2A S Sellnidera langsdorffii S reinwardtii K,S spectabilis A Baillonius bailloni A S Andigena laminirostris S Ramphastos discolorus vitellinus sulfuratus toco tucanus ambiguus 3A 7A A A A A 2S 3S,2St 4S S 154 J. PICIDAE S. BLANDAMER AND P. J. K. BURTON Spirit Skels 27 genera, 204 species Colaptes 3 genera & 128 species unrepresented auratus S JYNGINAE Spirit Skels campesfis 3A 4S,St Jynx Celeus torquilla 4A 4S,4K loricatus A PICUMNINAE brachyurus A St Picumnus elegans 2A squamulatus S Dryocopus olivaceus A lineatus 2S innominatus A javensis 2A Sasia martius A 3S ochracea 2A Campephilus abnormis 9A K melanoleucos 2Ak,2A Nesoctites rubricollis 2S micromegas A magellanicus 5A PICINAE Picas Melanerpes miniaceus A S candidus A puniceus 3A 3S lewis 2A S chlorolophus A 2St formicivorus cruentatus 4A A S mentalis flavinucha 2A S A 2St flavifrons pucherani 3A 2S S vittatus xanthopygaeus 2S 5A rubricapillus 2A 2S canus S,K aurifrons 2A viridis HA 9S.5K Sphyrapicus Dinopium varius 3A rafftesii 2A Campethera nubica A javanense benghalense 6A 3A S bennettii 2A Chrysocolaptes punctuligera nivosa 3A 4A lucidus Gecinulus A 2S Dendropicos fuscescens namaquus 3A A grantia Blythipicus rubiginosus S 2A S xantholophus Picoides A pyrrhotis Rein wardtipicus 2At,A maculatus A validus 2A obsoletus 2A Meiglyptes canicapillus 3A tristis 5A S minor macei 2A 2S,3K,St 3S tukki Hemicircus 4A,Af 3S auriceps A concretus A dorae 2A Mulleripicus darjellensis S fulvus 2S major 2A*,9A 8S,4K pulverulentus 2A nuttallii S EURYLAIMIDAE pubescens A 2S 8 genera, 14 species villosus tridactylus 2A 3A S S,St,2K 2 species unrepresented EURYLAIMINAE arcticus A Smithornis Veniliornis capensis 2A*,2A nigriceps S rufolateralis S passerinus st sharpei 2S affinis A Pseudocalyptomena Piculus graueri At flavigula A Corydon aurulentus A S sumatranus 2S ANATOMICAL SPECIMENS OF BIRDS 155 EURYLAIMINAE (cont.) Cymbirhynchus Spirit Skels macrorhynchos 8A*,15A 2S Eurylaimus javanicus S,K ochromalus 6A Serilophus lunatus 2S* Psarisomus dalhousiae 3A 2St Calyptomena viridis 3A 8S whiteheadi A S 4 genera Dendrocincla fuliginosa homochroa Deconychura longicauda Sittawmus griseicapitlus Glvphorhynchus spirurus Xiphocolaptes albicollis Dendrocolap tes picumnus platyrostris Xiphorhynchus picus guttatus flavigaster Lepidocolaptes angusiirostris fl$7m.s fuscus soulevetii Camp vlorhamphus trochilirostris DENDROCOLAPTIDAE 13 genera, 52 species a & 36 species unrepresented 2A A 8 A K A 7A A 3A 4A S S 2S FURNARIIDAE 34 genera, 218 species 14 genera & 178 species unrepresented FURNARIINAE Geositta cunicularia 6A isabellina A Upucerthia ruficauda 4A dumetaria 3A Cbtdodes fuscus A patagonicus 2A nigrofumosus S Furnarius leucopus SYNALLAXINAE Aphrastura spinicauda Leptasthenura platensis aegithaloides Synallaxis (Schoeniophylax) phryganophila (Synallaxis) Spirit 3A 2A A Skels ruficapilla 2A frontalis 2A albescens A spixi A gujanensis cinnamomea 6A Certhiaxis cinnamomea 2A Thripophaga pyrrholeuca Phacellodomus rufifrons 2A striaticollis 2A ruber A Spartonoica maluroides 4A Phleocryptes melanops 2A Anumbius annumbi 2A*,5A PHILYDORINAE Lochmias nematura A Pseudoseisura lophotes 2A Pseudocolaptes lawrencii A boissonneautii Philydor (Philydor) rufosuperciliatus A lichtensteini 2A rufus 2A Sclerurus albigularis A* caudacutus 3A guatemalensis 4A Xenops minutus A rutilans 2A Pygarrhichas albogularis A 2S 2S 2A 2A,2A* K Batara cinerea FORMICARUDAE 51 genera, 228 species 27 genera & 192 species unrepresented 156 J. S. BLANDAMER AND P. J. K. BURTON FORMICARIIDAE (cont.) Mackenziaena Spirit Skels severa S Taraba major 5A Sakesphorus canadensis 2S Thamnophilus doliatus 2A 2S nigriceps 4A punctatus 9A 4S caerulescens A S ruficapillus 4A Dysithamnus mentalis 2A Thamnomanes caesius A Myrmotherula surinamensis A fulviventris 4A axillaris 3A Herpsilochmus longirostris A Formicivora grisea 3A Drymophila ferruginea 4A squamata A Cercomacra tyrannina A Pyriglena leuconota 2A atra 2A Myrmoborus leucophrys 3A Gymnocichla nudiceps 7A S Myrmeciza longipes 2A exsul 7A S ferruginea 2A albifrons 2A Gymnopithys rufigula A leucaspis 7A 2S Hylophylax naevioides 5A S poecilonota 2A Phaenostictus mcleannani 2A Formicarius analis 2A S campanisona S Grallaria (Grallarid) varia A S Hylopezus perspicillatus A CONOPOPHAGIDAE 2 genera, 1 1 species 8 species unrepresented Spirit Skels Conopophaga lineata melanops Corythopis torquata 2A A RHINOCRYPTIDAE 12 genera, 30 species 9 genera & 28 species unrepresented Pteroptochos megapodius A 3S Scelorchilus albicollis St COTINGIDAE 27 genera, 79 species 1 6 genera & 58 species unrepresented Amp el ion rubrocristata 2A Pipreola riefferii A chlorolepidota A Lipaugus subalaris A S vociferans 2A K Pachyramphus viridis A rufus 2A cinnamomeus 3A S polychopterus 2A minor S cayana 2A semifasciata A S,K inquisitor S cayana K Gymnoderus foetidus A purpurata 2A S Perissocephalus tricolor 4A S rocnias 3A,At nudicollis A 2S Rupicola rupicola 2A S peruviana A* S PIPRIDAE 19 genera, 57 species 1 1 genera & 39 species unrepresented Pipra aureola A erythrocephala 10A 3S rubrocapilla 2A ANATOMICAL SPECIMENS OF BIRDS 157 P1PRIDAE (cont.) Pipra (cont.) Spirit Skels mentalis A 2S pipra 5A coronata 5A S serena 2A Antilophia galeata A Chiroxiphia lanceolata A pareola 17A 3S caudata A Masius chrysopterus 2A Ilicura militaris 2A Corapipo gutturalis 2A leucorrhoa 2A Manacus manacus 4A vitellinus 6A 2S Chloropipo holochlora A TYRANNIDAE 112 genera, 374 species 64 genera & 302 species unrepresented FLUVICOLINAE Xolmis coronata A,A* irupero A Muscisaxicola macloviana 2S maculirostris A Lessonia rufa 9A Myiotheretes striaticollis 3A Sayornis phoebe 2A nigricans 2S soya A Colonia colonus 2A Knipolegus nigerrimus A Hymenops perspicillatus 7A Fluvicola pica A S nengeta A Arundinicola leucocephala 12A S Pyrocephalus rubinus 15A S Satrapa icterophrys 3A Machetornis rixosus 4A Spirit Sk< Sirystes sibilator 2A Muscivora forficata S tyrannus 4A 2S Tyrannus tyrannus 6A S melanocholicus 4A,2A* 2S dominicensis 8A st caudifasciatus 2A Megarhynchus pitangua A S Myiodynastes maculatus 4A S Myiozetetes cayanensis 3A 2S similis Pitangus sulphuratus A*,10A S MYLARCHINAE Myiarchus ferox 2A tyrannulus 2A stolidus 2A magnirostris 9A Attila spadiceus 4A S cinnamomeus A Laniocera hypopyrrha S Contopus virens A cinereus 4A caribaceus A Empidonax virescens 2A hammondii A euleri A Terenotriccus erythrurus A Myiobius villosus A barbatus 2A S Myiophobus fasciatus A*,3A Onychorhynchus coronatus A K PLATYRINCHINAE Platyrinchus mystaceus 5A Tolmomyias sulphurescens 7A Rhynchocyclus olivaceus 3A 2S EUSCARTHMINAE Todirostrum poliocephalum 3A 158 J. S. BLANDAMER AND P. J. K. BURTON EUSCARTHMINAE (cont.) Todirostrum (cont.) Spirit cinereum A plumbeiceps 2A sylvia A Oncostoma olivaceum A EUSCARTHMINAE Colopteryx galeatus A Myiornis auricularia A Hemitriccus diops A Pogonotriccus eximius A Phylloscartes ventralis A SERPOPHAGINAE Tachuris rubrigastra A Anairetes parulus Serpophaga subcristata Mecocerculus stictopterus ELAENINAE Elaenia martinica spectabilis albiceps pallatangae Phyllomyias fasciatus Leptopogon amaurocephalus 5A 6A OXYRUNCIDAE 1 genus, 1 species Skels 2S 6A A A 2A 2A olivaceus Pipromorpha oleaginea 3S Oxyruncus cristatus Phytotoma rutila rara Pitta soror oatesi PHYTOTOMIDAE 1 genus, 3 species 1 species unrepresented 2S PITTIDAE 1 genus, 24 species 12 species unrepresented A 4A Pitta (cont.) Spirit Sk caerulea S erythrogaster A granatina 7A cyanea A guajana 2A S baudii A S sordida 24A brachyura 12A S angolensis 2A versicolor A K XENICIDAE 2 genera, 4 species 1 species unrepresented Acanthisitta Moris Xenicus longipes gilviventris 6A 7A 4A Philepitta castanea schlegeli Neodrepanis coruscans PHILEPITTIDAE 2 genera, 4 species 1 species unrepresented 4A S 2A MENURIDAE 1 genus, 2 species 1 species unrepresented Menura novaehollandiae S,St,2K None 1 genus ATRICHORNITHIDAE 1 genus, 2 species & 2 species unrepresented ALAUDIDAE 1 5 genera, 77 species 3 genera & 53 species unrepresented Mirafra javanica 3 A 8S hova S rufocinnamomea 2A poecilosterna 2A nigricans A Eremopterix leucotis A signata 3A nigriceps 2A Ammo manes cincturus A Alaemon alaudipes A Ramphocoris clotbey A ANATOMICAL SPECIMENS OF BIRDS 159 ALAUDIDAE (cont.) Melanocorypha calandra maxima leucoptera yeltoniensis Calandrella cinerea rufescens Chersophilus duponti Galerida cristata theklae Lullula arborea Alauda arvensis gulgula Eremophila alpestris Spirit 2A* A 4A 13A 3A 29A 2A Skels 2S St 4S S S 2S 3S,K 7S,St,7K 2S 3A HIRUNDINIDAE 20 genera, 80 species 5 genera & 50 species unrepresented PSEUDOCHELIDONINAE Pseudochelidon eurystomina At K HIRUNDININAE Tachycineta bicolor 7A leucorrhoa 2A Progne tapera 3A chalybea 2A modesta 4A Notiochelidon cyanoleuca A Atticora fasciata A Stelgidopteryx ruficollis 3A Cheramoeca leucosternum S Pseudhirundo griseopyga 2A Riparia paludicola S,St riparia 2A 4S,K cincta A S Ptyonoprogne obsoleta A fuligula 2 A* Hirundo rustica 62A 16S,6K,St,S* tahitica A*,12A 2S,St Cecropis cucullata 2A semirufa 3A Cecropis (cont.) senegalensis daurica Petrochelidon rufigula nigricans pyrrhonota ariel Delichon urbica Psalidoprocne {Psalidoprocne) albiceps holomelaena obscura Spirit 3A 3A A 3A 5A 9A A 3A A Skels St 7S S 2S 12S,2K,S* 6S.2K 2St 2S,2St,4K HS,5St,4K MOTACILLIDAE 5 genera, 5 species 1 genus & 28 species unrepresented Dendronanthus indicus Motacilla flava citreola cinerea alba aguimp capensis flaviventris Macronyx croceus ameliae Anthus (Group A) novaeseelandiae campestris similis leucophrys pratensis trivialis hodgsoni roseatus cervinus spinoletta (Group B) berthelotii sokokensis (Group C) furcatus lutescens correndera antarticus 2A 17A 5A 5A A A 3A 2A 16A 3A 4A 2A 22A A A A 5A A A 3A 6A A 4S 2K St 4S,St,4K S,2K 2St S 2S,2K CAMPEPHAGIDAE 9 genera, 70 species 3 genera & 42 species unrepresented Coracina novaehollandiae 9A lineata 2A leucopygia papuensis A 9A 9S 2S 160 J. S. BLANDAMER AND P. J. K. BURTON CAMPEPHAGIDAE (cont.) Coracina (cont.) Spirit Skels robusta A caesia 3K pectoralis 2A cinerea 2A S azurea A S tenuirostris 4A S morio A melaena A melaschistos A 3St fimbriata S ^nfgra 5A K sueurii 27A 5S leucomela 11A S maculosa 8A Campephaga phoenicea 7A S Pericrocotus divaricatus A cinnamomeus 2A S brevirostris S,2St flammeus 35A 2St solans 6S Hemipus picatus 8A hirundinaceus 4A Tephrodornis gularis 3A S,2St pondicerianus A PYCNONOTIDAE 15 genera, 123 species 4 genera & 60 species unrepresented Spizixos semitorques A zeylanicus A melanoleucos 2A melanicterus 3A S,2St squamatus A S cyaniventris A iocosus 3A 8S xanthorrhous A leucogenys 3A S cafer A 2S aurigaster 11A S xanthopygos 2A St barbatus 24A 2S finlaysoni 2A flavescens 2A goiavier 3A S luteolus A plumosus 12A blanfordi A brunneus 2A erythropthalmos 3A 2S virens 20A S gracilis 3A Pycnotus (cont.) Spirit Skels curvirostris 2A importunus A latirostris ISA S gracilirostris 4A simplex A S Baeopogon indicator A Ixonotus guttatus 3A Chlorocichla simplex 9A flavicollis 5A flaviventris 4A Phyllastrephus terrestris 5A strepitans S flavostriatus A debilis 2A S albigularis 2A fishceri 13A 2S icterinus 4A xavieri 3A cinereiceps A Bleda syndactyla 5A eximia 5A canicapilla A Nicator chloris 9A S vireo A Criniger finschii S barbatus A calurus 8A pallidus 2A ochraceus 3A 12S bres ISA 5S phaeocephalus 19A 12S Setornis criniger 3A Hypsipetes viridescens A charlottae 3A criniger 11A 9S,2St mcclellandii 2A 10S flavala 2A 2S,St amaurotis A madagascariensis 5A 2S,St Tylas eduardi 2A S Aegithina tiphia viridissima lafresnayei IRENIDAE 3 genera, 14 species 5 species unrepresented ISA A 3A ANATOMICAL SPECIMENS OF BIRDS 161 IRENIDAE (cont.) Chloropsis sonnerati cyanopogon cochinchinensis aurifrons hardwickei Irena puella Spirit 6A A 5A A 2A 17A Skels 3S 2S S,2S 3S 5S LANIIDAE 12 genera, 74 species 40 species unrepresented PRIONOPINAE Eurocephalus ruppelli A PRIONOPINAE Prionops plumata 3A S caniceps A S retzii A scopifrons A MALACONOTINAE Lanioturdus torquatus 6A Nilaus afer 8A gambensis 3A cubla 4A sabini 2A Tchagra senegala 6A australis A Laniarius ferrugineus 2A barbarus 4A funebris A leucorhynchus Telophorus bocagei A sulfureopectus multicolor 2A Malaconotus cruentus A blanchoti LAMINAE Corvinella corvina A melanoleuca A Lanius tigrinus 6A cristatus 4A collurio 7A collurioides A schach 6A minor A ludovicianus A 3S 2K St S S 7S,4K 4St,K S Lanius (cont.) Spirit Skels excubitor 2A*,11A 4S,K excubitoroides A collaris 10A*,9A senator A*,10A 2S,K PITYRIASINAE Pityriasis gymnocephala 7A 2S VANGIDAE 9 genera, 13 species 2 species unrepresented Calicalicus madagascariensis 5A 2A 4A curvirostris Xenopirostris xenopirostris Falculea palliata Leptopterus viridis chabert madagascarinus Oriolia bernieri Euryceros prevostii Hypositta corallirostris 3A 3A 4A A 2A 2S S 2St S st BOMBYCILLIDAE 5 genera, 8 species 1 genus & 4 species unrepresented Spirit Skels PTILOGONATINAE Ptilogonys caudatus A Phainopepla nitens 4A Bombycilla garrulus 7A*,3A 5S cedrorum 2 A*, 4 A K HYPOCOLIINAE Hypocolius ampelinus At,3A Dulus dominicus Cinclus cinclus pallasii DULIDAE 1 genus, 1 species 2A CINCLIDAE 1 genus, 5 species 3 species unrepresented 3A*,7A 3S St 162 J. S. BLANDAMER AND P. J. K. BURTON TROGLODYTIDAE Erythropygia Spirit Skels 14 genera, 59 species leucophrys 5A 10 genera & 49 species unrepresented galactotes 3A st Campylorhynchus Spirit Skels quadrivirgata A zonatus S Drymodes Thryothorus brunneopygia A st,s fasciatoventris 2A Pogonocichla coraya 2A stellata 6A nigricapillus 4A S Erithacus ludovicianus 3A erythrothorax 4A leucotis 4A rubecula 9A*,10A 8S,2St longirostris A sibilans A Troglodytes luscinia 2A S troglodytes 5A*,16A 5S,3K megarhynchos 5A 2S,K aedon A*,3A S calliope A S Henicorhina svecicus 7A 2S,K leucosticta 5A pectoralis 3A pectardens A MIMIDAE brunneus A 13 genera, 31 species 5 genera & 19 species unrepresented Dumetella cyane cyanurus chrysaeus 19A At,2A 2A 4S carolinensis 4A 2S Cossypha Melanoptila glabrirostris A Melanotis heuglini cyanocampter caffra 2A A A caerulescens A ansorgei 2St Minius niveicapilla 4A polyglottos 4A 2S thenca S albicapilla Cichladusa 2A saturninus 3A 2S arquata 2A A 10tht> Nesomimus Si 1C I llC trifasciatus 6A diademata Copsychus 14A rufum 2A 2S saularis 13 A, A* 4S,2S curvirostre A albospecularis 2A redivivum A malabaricus 21A 5S,K stricklandii A atricapillus 6A pyrropygus A*,A,At 3S Ma/-j?a/m>s Phoenicurus fuscatus A ochruros 7A,A* 3S phoenicurus 4A*,34A 9S,K frontalis A PRUNELLIDAE moussieri A 1 genus, 12 species 9 species unrepresented erythrogaster Rhyacronis 9A Prunella collaris 2S fuliginosus Cinclidium A montanella S leucurum 2A modularis A*,3A 6S,2K,2St frontale A Grandala MUSCICAPIDAE coelicolor A S,2St 48 genera, 308 species Sialia 12 genera & 183 species unrepresented stalls 2A 3S,2St TURDINAE mexicana A S £rac/;>y7/-4>f ir y$:z •~ ^ - S 8 P i_-« ftt! w> 2- O '43 i § 6 H^ 2.S d> ^ C (— ( 11 8J 4> § 12 CO CM r- KHz 184 J. C. M. DRING Megophrys aceras Megalophrys montana var. aceras Boulenger, 1903. MATERIAL. BM. 1974. 4276-4288 (4 •-> •5 G T3 C O W) o X) 60 C •.s I o I & ri § VI -* — r AMPHIBIANS AND REPTILES FROM MALAYSIA 1 99 Rana chalconota raniceps Polypedates raniceps Peters, 1871. MATERIAL. BM. 1974. 4551-4626 (41 &J, 19 ??, 15 juveniles, larvae). HABITAT. This riparian species was found in all areas near water at the Sungei Kelebang, including the flooded areas of logging tracks, and small streams in logged forest. Most were low on vege- tation but some were up to 1| m above ground. Males call from these positions. The call is a soft 'pink pink' like dripping water. The species was also found at the Sungei Petuang at 250 m. Rana hosei Rana hosii Boulenger, 18916. MATERIAL. BM. 1974. 4627-4640 (8 &?, 6 ?$). HABITAT. Six specimens were found at the Sungei Kelebang, on the river banks and up to 20 m from the river on vegetation. The other specimens were found along the Sungei Petuang (250 m), on the rocky river banks and up to 5 m from the water and 2 m above ground on shrubs. The <&£ all have nuptial asperities and the females are gravid or have convoluted oviducts. Rana kuhli Rana kuhlii Dumeril and Bibron, 1841. MATERIAL. BM. 1974. 4641-4654 (2 <$& 4 ??, 8 immature and larvae). HABITAT. Rana kuhli was found only on the east ridge of G. Lawit (790 m) where it occupied small, rocky, mountain streams. Most specimens were found in shallow gravel-bottomed pools along the streams. Some were caught in a side pool of clear water with a light layer of silt and decomposing leaves. Two larvae were caught in a similar pool. REMARKS. The 2 males have nuptial pads on the first finger covered with minute, white spicules. They have small unpigmented testes and are 61-4 and 93-0 mm in length. Only the larger male has significantly enlarged mandibular processes, which are 3-4 mm in height. Four gravid females are 66-0-8 1-3 mm in length. The mature ova are one-third pigmented and about 2-5 mm in diameter. The larvae have 1 : 1-1 / 1-1 : 11 labial teeth, all the toes broadly webbed to the tips, the tibiae strongly tuberculate and the tail blotched and barred with black. They differ from Bornean larvae, however, in having acutely pointed tails twice as long as the body and in having 3 rows of papillae on the lower lip. Rana laticeps Rana laticeps Boulenger, 1882. MATERIAL. BM. 1974. 4655-4702 (7 &J, 5 gravid ??, immature specimens). HABITAT. As previously mentioned, this species and Amolops larutensis were the only frogs collected at all camps 43-1280 m. These records seem to extend both the upper and lower altitude limits slightly. In Borneo specimens come from 100 to 920 m. In the Malay Peninsula previous records are from 900 to 1220 m. At the Kelebang R. laticeps could only be found in one stream and only four specimens, all caught on the same night, were found, despite frequent collecting both at night and in the day. This was the same rocky hillside stream in which Amolops larutensis and Ansonia sp. were found. The frogs were in shallow pools or between rocks at the stream edge. Thirty-nine of these frogs came from two streams on the east ridge of G. Lawit (790 m). The species was found in areas of shallow gravel-bottomed pools and in shallow reaches with many dead leaves and some other detritus in the upper parts of the streams. During the night many 200 J. C. M. DRING were found in the water, from which the males seem to call. They were also found on the banks, mostly near the water but up to 2 m away in a few cases. During daylight the frogs hide among dead leaves in the water, in crevices under boulders and among the leaf litter along the banks. Juveniles were found in daylight among the leaf mould in seepages along the banks. Only 2 specimens came from the summit ridge (1280 m). Both were in the small, semi-stagnant stream under closed canopy forest, along which most of the collecting was conducted. One was resting on matted roots in the edge of a pool between rocks, the other was in a shallow pool on a smooth rock base. A few calls were also heard along this stream. Thus R. laticeps occupies a similar habitat throughout its altitudinal range. Apparently it also occupies a very similar habitat in Borneo (Inger, 1966). REMARKS. The adult males are 34-5-46-9 mm (mean 43-8 mm, N=7) in snout-vent length. Gravid females are 36-7-45-0 mm long (mean 40-8 mm, N = 5) and contain ova 2-8-3-0 mm in diameter, with only about one-fifth of the surface pigmented. The ova are therefore larger and less pigmented than those of kuhli. The call is a rising gurgle (Fig. 7). Rana limnocharis limnocharis Rana limnocharis Boie in Weigmann, 1835. MATERIAL. BM. 1974. 4703-4724 (15 -