riba Ths Pema Ltt an > me OLD een Pe en ee ae —, gee Rte SPS Nise ern n f qetse: Putn ie , ee ey — PEF Penh Si i ares Ketel eh ictal oo trie Mala at a ee WET Cray te He ie ee PEL Dalhvigc ieee a ‘ 4 J wr DD = f x Og OP : 4 ae ae - wer om + a Pe PE FHT ar PG epee ee Ane ee ee 5 - . . ’ . er ee nk ak aaa - mary ee ee ed ee a oo : os Sia = A . - neon ODP eS eee oF J Vee ee eS . ys ote Gp hate sala _ mo > nee ete oe fo ae hh : . s ©. S (Ses Rye “Pinorsied {uh &-B hae o“ Abeta) ete Oe jhe hee ehh onan» 81 SP 1, be ANNALS OF THE SOUTH AFRICAN MUSEUM VOLUME XLII ¥ Ves, at + i | spy yty © Ae " - p ne ' S Pad me cs eae ae a i cs i" = fa = . py ty - i 7 A 4 % ‘ S és a f 7 ' Pedal see ae ae 4 ‘ « ' y ul 7A 7 : tele we. aor a a ee he a et | pte ‘ me YS Lae 3 ~J ce YO ANNALS OF THE Sew ii AtWi@AN MUSEUM VOLUME XLIII TRUSTEES OF THE SOUTH AFRICAN MUSEUM LOO may ® PRINTED IN THE UNION OF SOUTH AFRICA BY THE RUSTICA PRESS, PTY., LTD., WYNBERG, CAPE LIST OF CONTRIBUTORS K. H. BARNARD Additions to the Fauna-list of South African Crustacea and Pycnogonida (published December 1955) ; ss ae P. L. G. BENorrT Ichneumonidae Nouveaux ou Intéressants de l’Afrique du Sud (published December 1956) .. L. D. BoonstRA The Skull of Tapinocephalus and its near relatives (published December 1956) Pareiasauriér-studies. Deel XII. °’n Lewensrekonstruksie van Bradysaurus seeleyi (published December 1956) J. P. Harpinc The South African Cladoceran Euryalona colletti (Sars) and other South African species (published May 1957) . ; ? N. A. H. MiLtarp The Hydrozoa of False Bay, South Africa (published January 1957) _D. P. Murray The Species of Lepidochrysops and Cupidopsis of Southern Africa. The Genitalia (Lepidoptera-Lycaenidae) (published December 1956) ae ae os J. E. Wess On the Lancelets of South and East Africa (published May 1957) Page 245 7h) 109 249 NEW GENERIC NAMES PROPOSED IN THIS VOLUME _— es Cypsiphimedia (Acanthonotozomatidae) Ilyograpsus (Grapsidae) Miophrixus (Bopyridae) A pxcetes ... Actumnus Afrephialtes .. Afrochiltonia . . Aglaophenia .. Alpheus Ampelisca.. Amphisbetia . Antennella Antennopsis Antilibinia Artopoles Asphragis Athelges Bopyrella Brachiella Brachiostoma.. Bradysaurus Byblis .. Caligus Camorta Campanularia Caprellina Charitopimpla Chlorodiella .. Chondracanthus Chorinaeus Cirolana Cleantis Clistopyga Clytia .. ; Conilorpheus .. Corallana Coryne Cryptochirus . Cryptocnemus Cryptodromia Cupidopsis Cymo.. Cymodoce LIST OF GENERA PAGE Cyphocarcinus Cypsiphimedia Delaulax Dictyocladium Dies Ebalia .. Elasmopus Euchrysops Eudendrium .. Eumedonus Euonyx Euryalona Eurystheus Exetastes Glypta Halecium Haliophasma . . Halopteris Hapalocarcinus Hebella Hermilius Holcopimpla .. Hyale . Hydractinia Hydrocorella .. Hydrodendron Hymenosoma. . Ianiropsis Ilyograpsus Iphimedia Isocladus Itoplectis Keratocephalus Kirchenpaueria Lanocira Leander Lemboides Lembos Lepidochrysops Leptochelia Leuckartiara .. Liljeborgia Lovenella Lytocarpus Macrophthalmus Medesicaste Melita. . Menaethiops .. Miophrixus Mormosaurus. . Myriothela Nemertesia Nikoides Nototropis Obelia. . Oralien Orthocentrus . . Palaemon Pallenoides Paracaprella .. Paracilicaea Paracleistostoma Paracollyria Paragattya Paranthura Paratymolus .. Pelosuchus Periclimenes .. Petrolisthes Pilumnopeus .. Pimpla Pinnixa Plumularia LIST OF GENERA PAGE Pontophilus 61 Porcellana 47 Processa 95 Propallene Q4. Pseudodromia III Pycnogonum .. 49 Pycnotheca 182 a Queubus 241 Reticularia 21 103 Salacia oe Sarmatium We . Scandia 79 Sertularella 156 Sertularia 186 Sjostedtiella Sphaeramene . . Sphaeroma 334 Sphaerozius oo Sphyrion go Squilla se Symplectoscyphus 198 Synidotea 103 Synthecium 124 Talorchestia .. A7 Tapinocephalus 105 ‘Taurops : 99 Thecocarpus .. 70 Theronia 23 Tryphosa 126 Tubularia 230 Typhlocarcinodes 51 8 whee 164 Unciolella 47 Urothoe AI 30 Xenopthalmodes 131 19 230 Zuzara PAGE 48 40 43 105 3i/) 106 234 107 203 207 28 202 207 221 124 66 61 38) 104 49 219 74 203 93 152 164 240 127 80 179 35 g6 83 ot 65 ANNALS a SOUTH AFRICAN MUSEUM VOLUME XLII i) PART I, containing:— : 1. Additions to the Fauna-list of South African Crustacea and Pycnogonida. By — K. H. Barnarp. (With fifty-three text-figures.) ZTE ON y JUN13 1953 di i & i Re : oe) é eat ISSUED DECEMBER 1955 PRICE 20s. od. S | PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY THE RUSTICA PRESS (PTY.) LIMITED, COURT ROAD, WYNBERG, CAPE ier ANNALS OF THE POUTH AFRICAN MUSEUM VOLUME XLII 1. Additions to the Fauna-list of South African Crustacea and Pycnogonda. By K. H. BARNARD. (With fifty-three text-figures.) This paper forms an appendix to previous papers on South African Crustacea and Pycnogonida. The additions are mainly the results of intensive collecting by the University of Gape Town Ecological Survey (U.C.T.) carried out by the Zoology Department under the direction of Professor J. H. Day; and by the Zoology Department of the University of the Witwatersrand (U.W.). The former has extended the survey of the estuaries of the Union of South Africa to that of Inhambane, Portuguese East Africa; and the latter has operated from the Marine Biological Station on Inhaca Island, Delagoa Bay. As might be expected several interesting new records have been collected, but it is quite certain that many more will be added in the future. The marine fauna of Portuguese East Africa has not yet been thoroughly investigated; nor in fact has that of the Union of South Africa. I wish to express my thanks to Professor Day and his team of enthusiasts (U.C.T.), and to Mrs. M. Kalk (U.W.) for submitting these collections for identification; and for presenting to the South African Museum duplicates of the species not represented in its collections. A full list of the species collected is not given, only notable additions to the localities which extend the already known distribution of the respective species, new records, and new species. The employment of new methods of collecting usually produces abundant, and often surprising, results. The University of Cape Town has recently adopted ‘Operation Frogman’ for investigating the shallow-water fauna of False Bay. Mr. J. F. GC. Morgans, a member of Professor Day’s staff, has pioneered this method in South Africa, and has demonstrated its advantages over the old-fashioned ‘naturalist’s dredge’. Many of the following records and discoveries are due to his enthusiasm for submarine collecting. In some cases species, hitherto considered rare, have been found to be plentiful when their particular habitat has been discovered. For example, in the course of an investigation into the species of Patella living on the Cape coast (Koch, H. J., 1949, Ann. Natal Mus., xi, 3, p. 491), the Isopod Dyna- VOL. XLIII. PART I. JUN 4 1956 EEE EEE EE’? ned CQO OO OOOO OOOO OEE 2 ANNALS OF THE SOUTH AFRICAN MUSEUM menella australis Rich. and the Amphipod Calliopiella michaelsent Schell. were found in abundance under the shells of these molluscs. The following papers containing references to the South African Crustacean fauna have appeared in recent years. Barnard, K. H. 1951. Ann. Mag. Nat. Hist. (12), iv, pp. 698-709, figs. (Isopoda and Amphipoda). . 1952. Trans. Roy. Soc. S. Afr., xxxill, pp. 279-82, figs. (Grandidierella species). : 1954. Ann. Mus. Congo. Tervuren, 4to, zool. 1, pp. 120-131, figs. (Decapoda, Isopoda, Amphipoda). . 19546. Mem. Inst. Sct. Madagascar. A. ix. pp. 95-104, figs. (Decapoda). . 1955a. Ann. Natal Mus., xiii, 2, p. 247 (record of a Cirripede). . 1955). Ann. Natal Mus., xiii, 2, p. 249 (a new freshwater Isopod). Capart, A. 1951. Exp. océanogr. Belge, III, 1, pp. 1-205, figs. (Brachyura). Forest, J. 1951. Bull. Soc. zool. Fr., \xxvi, p. 83 (Calcinus). . 1954. Ann. S. Afr. Mus., xli, 4, pp. 159-213, figs. (Paguristes). Holthuis, L. B. 1950. Stboga Exp. monogr., xxxix, a, 9, pp. 1-126, figs. (Palaemoninae). ——. 1951. Altlantide Rep. No. 2, pp. 7-187, figs. (Caridea). . 1952. Stboga Exp. monogr., Xxxix, a, 10, pp. 1-254, figs. (Pontoninae). . 1952. Exp. océanogr. Belge, III, 2, pp. 1-88, figs. (Macrura). Reid, D. M. 1951. Ailaniide Rep. No. 2, pp. 189-291, figs. (Amphipoda). Ruffo, 8. 1947. Mem. Mus. Civ. Verona. I. pp. 113-30. figs. (Amphipoda). . 1953. Rev. Kool. Bot. Afr. xlvii. pp. 120-36. figs. (Amphipoda). The following new genera and new species are described in the present paper. DECAPODA. Menaethiops natalensis and delagoae, Paracleistostoma fossula, Ilyograpsus (n.g. Grapsidae), I. rhizophorae, Xenophthalmodes brachyphallus, Ebalia pondoensis (tuberculata Brnrd. 1950 non Miers 1881) and agglomus, Porcellana delagoae, Acetes natalensis, Alpheus bullatus. ISOPODA. Paranthura latipes, Cirolana imposita, pilula, and capitella, Conil- orpheus blandus, Corallana furcilla, Isocladus otion and mimetes, Sphaeramene muicro- tylotos, Cymodoce lis and alts, Paracilicaea teretron and clavus, Artopoles capensis, Taniropsis bisbidens, Athelges caudalis, Miophrixus (n.g. Bopyridae), M. latreutzdis. AMPHIPODA. Evuonyx conicurus, Tryphosa normalis and africana, Urothoe serrulidactylus, pinnata and pinnata var. femoralis, Cypsiphimedia (n.g. Acanthonoto- zomatidae), C. gibba, Melita machaera, Afrochiltonia (n.g. Talitridae) for Chilionza capensis Brnrd. 1916, Lembos teleporus, Unciollela foveolata. COPEPODA. Chondracanthus colligens, Brachiella supplicans. PYCNOGONIDA. Propallene similis, Pallenowdes proboscideum. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 3 REcorpDs OF ADDITIONAL SPECIES AND NEw LOCALITIES DECAPODA Macropodia rostrata (Linn.). Zwartkops estuary, Port Elizabeth; Bushmans River mouth (Alexandria-Bathurst Division) (U.C.T.). Barnard 1954. Achaeus lacertosus Stmpsn. Inhambane, Portuguese East Africa (U.C.T.). Inachus angolensis Capart 1951. 0°-16° 30’ S. Inachus dorsettensis (Penn.). False Bay, 40 fms. (U.C.T.). Menaethius monoceros (Latr.). Durban Bay (U.C.T.). Naxioides hirta M. Edw. Inhambane (U.C.T.). Platylambrus quemvis Stebb. Inhambane (U.C.T.). Eumedonus granulosus MacGilch. Delagoa Bay (U.W.). Barnard 19540. Hymenosoma orbiculare Desm. Delagoa Bay (U.W.); Inhambane (U.C., JD) Dotilla fenestrata Hilg. Keiskama River mouth, and Breede River mouth (U.C.T.). Barnard 1954. Cleistostoma edwards McLeay. West coast: Saldanha Bay. South coast: mouths of Sundays River, Bushmans River, Keiskama River, Bashee River (U.C.T.). Barnard 1954. Cleistostoma algoense Brnrd. 1954. West coast: Saldanha Bay. South coast: mouths of Breede River, Great Brak River, Knysna, Zwartkops River, Bushmans River (U.C. ae Me Tylodiplax blephariskios (Stebb.). Durban Bay and St. Lucia Bay (U.C.T.). Barnard 1954. Inhambane (U.C.T.). See infra fig. 7f. Cardisoma carnifex (Herbst). Richards Bay (U.C.T.). Barnard 1954. Pachygrapsus plicatus (M. Edw.). Delagoa Bay (U.W.). Varuna litterata (Fabr.). Port St. Johns (U.C.T.). Barnard 1954. Varuna tomentosa Pfeff. Port Edward, Umtamvuna River, 300 yards from mouth (U.C.T.). Barnard 1954. Pseudograpsus erythraeus Kossm. Delagoa Bay (U.W.). Sesarma longipes Krauss. Inhambane (U.G.T.). Sesarma meinertt de Man. Beira (Chace, 1953); Delagoa Bay (U.W.). Sesarma catenata Ortm. Mouths of Bushmans River, Sundays River, Zwart- kops River, and Breede River. The last is the most westerly locality hitherto recorded. Barnard, 1954. Sesarma elongata M. Edw. Inhambane (CLC T,): Portumnus meleayt Brnrd. Angola, 9° S. (Capart, 1951). Barnard, 1954. Ovalipes punctulatus de Haan. Walfish Bay (Capart, 1951). Tristan d’Acunha (a.60) x 77 mm. 5) At Mus:)? Portunus tuberculatus Roux. Barnard 1950 as Elliptodactylus rugosus Doflein. Atlantic, southwards to 26° S. (Capart, 1951). Barnard, 1954. - Monomia gladiator (Fabr.). Inhambane (U.C.T.). Cycloachelous granulatus (M. Edw.). Durban Bay (U.C.T.). Barnard, 1954. Thalamita integra Dana. Durban Bay and Inhambane (U.C.T.). Thalamita investigators Alcock. Durban Bay (U.C.T.). Barnard, 1954. Panopeus africanus M. Edw. Durban Bay (U.C.T.). Barnard, 1954. Kozymodes xanthoides (Krauss). Delagoa Bay (U.W.). Chlorodopsis pilumnoides (White). Delagoa Bay (U.W.). Phymodius ungulatus (M. Edw.). Delagoa Bay (U.W.). Phymodius monticulosus (Dana). Durban Bay (U.C.T.). Xantho hydrophilus = exaratus (M. Edw.). (?) Halimede delagoae Brnrd. 1954. Portuguese East Africa (Lourengo Marques Museum). : ANNALS OF THE SOUTH AFRICAN MUSEUM Lybia plumosa Brnrd. Durban Bay (U.C.T.); Delagoa Bay (U.W.). Barnard, 1954. See infra fig. 13d. Euriippellia annulipes (M. Edw.). Kosi Bay and St. Lucia Bay (U.C.T.). Barnard, 1954. Delagoa Bay (U.W.). Pseudozius caystrus (Ad. & White). St. Lucia Bay (U.C.T.). Pilumnoides perlatus (Poeppig). Walfish Bay (Capart, 1951). Barnard, 1954. Epixanthus frontalis (M. Edw.). Kosi Bay and Inhambane (U.C.T.). Pilumnus hirsutus Stmpsn. var. (Barnard, 1950, p. 265). Saldanha Bay (U.G.T.). Barnard, 1954. Parapilumnus pisifer (McLeay). Delagoa Bay (U.W.). Goneplax angulata (Pennant). See Barnard, 1954. Eucrate sulcatifrons (Stmpsn). Delagoa Bay (U.W.). Typhlocarcinus sp. 92 only. Inhambane (U.C.T.). Thaumastoplax spiralis Brnrd. Estuaries of Great Brak River (Mossel Bay) and Zwartkops River (Algoa Bay), in Echiuroid burrows; Durban Bay (U.C.T.). Barnard, 1954. Nautilocorystes ocellata (Gray). Walfish Bay (Capart, 1951). Barnard, 1954. Conchoecetes artificiosus (Fabr.). Inhambane (U.C.T.). Dromidia dissothrix Brnrd. False Bay, 11-12 metres (U.C.T.). Barnard, 1954. Calappa japonica Ortmann. Kentani coast (S. Afr. Mus.). Barnard, 1954. Calappa gallus (Herbst). Inhambane (U.C.T.). Ebalia tuberculata Miers. See Capart, 1951 and Barnard, 1954. Capart con- siders that the specimens assigned by me in 1950 to Miers’s species represent a different, perhaps new, species. Since then Dr. Monod has sent me West African specimens undoubtedly referable to tuberculata, which prove that barnardi is a synonym of Miers’s species. For the former, therefore, I propose the name pondoensis n. sp. (Pondoland coast). Leucosia marmorea Bell. Inhambane (U.C.T.). Dorippe lanata (Linn.). Delagoa Bay (U.W.). Emerita austroafricana Schmitt. Occurs as far south as the Kentani coast. Barnard, 1954. I have recently seen specimens from the west coast of Madagascar (Madagascar Institute). Paguristes spp. See Forest, 1954. Aniculus aniculus (Fabr.). Delagoa Bay (U.W.). Troglopagurus jousseaumi Bouvier. Delagoa Bay (U.W.). Clibanarius clibanarius (Herbst). Inhambane (U.C.T.). Clibanarius padavensis de Man. Durban Bay and Inhambane (U.C.T.). Clibanarius longitarsus (de Haan). Richards Bay and Inhambane (U.C.T.); Delagoa Bay (U.W.). Diogenes custos (Fabr.). Delagoa Bay (U.W.); Inhambane (U.C.T.). Diogenes avarus Heller. Delagoa Bay (U.W.); Inhambane (U.G.T.). Diogenes costatus Hend. Inhambane (U.C.T.). Diogenes senex Heller. Delagoa Bay (U.W.); Inhambane (U.C.T.). Spiropagurus spirifer (de Haan). Inhambane (U.C.T.). Pachycheles natalensis Krauss. Delagoa Bay (U.W.). Polyonyx sp. In Chaetopterus tubes, Durban Bay (U.C.T.). Callianassa rotundicauda Stebb. Saldanha Bay, and Durban Bay (U.C.T.). Upogebia africana (Ortm.). West coast : Saldanha Bay (U.C.T.); South coast: mouth of Keiskama River and Sundays River (U.C.T.). Barnard, 1954. Upogebia africana var. (Barnard, 1950, p. 520). Inhambane (U.C.T.). ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 5 Upogebia assist Brnrd. False Bay, 10 fms. (U.C.T.). Upogebia savignyt Strahl. False Bay, 8 metres (U.C.T.); Delagoa Bay (U.W.). Scyllarides elisabethae (Ortm.). Off Inhambane (Lourengo Marques Museum). Penaeopsis hilarulus de Man. Delagoa Bay (U.W.); Inhambane (U.C.T.). Acetes erythraeus Nobili. Durban Bay (U.C.T.). Leucifer penicillifer Hansen. Durban Bay (U.C.T.). Caridina africana Kingsley. Inhambane (U.C.T.). Latreutes mucronatus (Stmpsn.). Barnard, 1950, p. 706: for ‘135 mm.’ read 13°5 mm. Durban Bay (U.C.T.). Latreutes pygmaeus Nobili. Inhambane (U.C.T.). Angasia armata (Paulson). Inhambane (U.C.T.). Eippolysmata vittata Stmpsn. Inhambane (U.C.T.); Delagoa ee CULW2)): Ogyrides saldanhae Brnrd. Walfish Bay, up to 32 mm. in length (Holthuis, 1952). Table Bay (R. W. Rand, Guano Islands Administration, 1955). Betaeus jucunda Brnrd. Mouths of Bashee River, Bushmans River, Zwartkops River, and Breede River (U.C.T.). Barnard, 1954. Athanas cf. minikoensis Gout. Delagoa Bay (U.W.). Palaemon in Barnard, 1950 = Macrobrachum; and Leander ibid. = Palaemon, subgen. Paleander and Palaemon. ‘Palaemon’ sundaicus and delagoae = M. equidens (Dana). ‘Palaemon’ dolichodactylus = M. scabriculum (Heller). See Holthuis, 1950. ‘Leander’ Palaemon concinnus Dana. Umkomaas and Durban Bay (U.C.T.). Barnard, 1954. Delagoa Bay (U.W.). Anchistus inermis (Miers) = custos (Forskal.). Holthuis, 1952. ‘Palaemonetes’ natalensis Stebb. is considered to belong to the genus Peri- clymenaeus by Holthuis, 1952. Paranchistus ornatus Holthuis 1952. Mozambique. Harpilius a subgen. of Periclymenes. H. depressus and beaupresii placed in gen. Harpiliopsis. Holthuis, 1952. Focaste n.g. for Coralliocaris lucina. Holthuis, 1952. TANAIDACEA Leptochelia savignyi (Kroyer). Inhambane (U.C.T.). ISOPODA Cyathura estuaria Brnrd. St. Lucia Bay (U.G.T.). Panathura serricauda Brnrd. Saldanha Bay (U.C.T.). Exanthura macrura Brnrd. Lamberts Bay (U.C.T.). Apanthura dubia Brnrd. East London (U.QC.T.). Mesanthura catenula (Stmpsn.). East London (U.C.T.). Haliophasma pseudocarinata Brnrd. False Bay and Algoa Bay (U.C.T..). Eurydice longicornis (Studer). Olifants River mouth and Saldanha (U.C.T.); Zwartkops estuary, Algoa Bay (Macnae, Rhodes Univ.). Pontogeloides latipes Brnrd. Saldanha Bay, Durban Bay, Kosi Bay, and Inhambane (U.C.T.). Cirolana undulata Brnrd. Various localities from Port Nolloth to East London (Weal): Cirolana bovina Brnrd. East London (U.C.T.). Cirolana saldanhae Brnrd. 1951. Saldanha Bay (U.C.T.). Cirolana cingulata Brnrd. Off Umkomaas (Natal), 40 fms. (S. Afr. Mus.) ; Arniston and Algoa Bay (U.C.T.). See also infra p. 53. Aega monilis Brnrd. Saldanha Bay area (Fisheries Survey). Nerocila trichiura (Miers). Durban, on Exocoetus (S. Afr. Mus.). 6 ANNALS OF THE SOUTH AFRICAN MUSEUM Livoneca raynaudi M. Edw. Durban, on a wrasse (S. Afr. Mus.). TIrona melanosticta Sch. & M. Durban, on Tylosurus (S. Afr. Mus.); Delagoa Bay, on Tylosurus (U.W.). Sphaeroma serratum (Fabr.). Durban Bay (U.C.T.). Barnard, 1951. Delagoa Bay (U.W.). Sphaeroma walkeri Stebb. Durban and Inhambane (U.C.T.). Exosphaeroma planum Brnrd. Various localities from Port Nolloth to Jeffreys Bay (U.C.T.). Exosphaeroma porrectum Brnrd. Lamberts Bay, Saldanha Bay, and Port Elizabeth (U.C.T.). Exosphaeroma pallidum Brnrd. Saldanha Bay (U.C.T.). Exosphaeroma estuarium Brnrd. 1951. Umkomaas, Natal (U.C.T.). Pseudosphaeroma barnardi Monod. Hermanus, Wilderness (George Division) and Knysna (U.C.T.). Barnard, 1951. Sphaeromene polytylotos Brnrd. Liideritzbucht (S. Afr. Mus.); Port Nolloth UC): Dies monodi Brnrd. 1951. Bashee River, Port St. Johns, Umkomaas, and Kosi Bay (U.C.T.). See also infra p. 72. Dynamenella australis Rich. Paternoster Bay and Danger Point (U.C.T.). Occurs under the shells of Patella. Cymodocella sublevis Brnrd. Saldanha Bay, Plettenberg Bay, and East London (UEC): Cymodocella pustulata Brnrd. Saldanha Bay, Still Bay, and East London (U-CoP.). Cymodocella magna Brnrd. 1954. Groen River (south of Port Nolloth) (U.C.T.). Cymodocella eutylos Brnrd. 1954. Mossel Bay, Jeffreys Bay, and Kleinmond (Bathurst Division) (U.C.T.). Idotea ziczac Brnrd. 1951. Saldanha Bay, and False Bay, 8 fms. (U.C.T.). Paridotea reticulata Brnrd. Port Nolloth and Lamberts Bay (U.G.T.). Paridotea rubra Brnrd. Port Nolloth, Lamberts Bay, and East London (U.C.T.). Paridotea fucicola Brnrd. Port Nolloth, Still Bay, and Cape Infanta (U.C.T.). Idarcturus platysoma Brnrd. Paternoster Bay (U.C.T.). Arcturella corniger (Stebb.). One 9, Morumbane, Inhambane (U.C.T.). Antias uncinatus Vanh. Saldanha Bay (U.C.T.). Originally described from Simonstown. Stenetrium crassimanus Brnrd. Plettenberg Bay and Quolora (U.C.T.). Stenetrium diazt Brnrd. Plettenberg Bay (U.C.T.). Janira capensis Brnrd. Lamberts Bay, Saldanha Bay, False Bay, and Knysna (UG). Janira exstans Brnrd. Saldanha Bay, and East London (U.C.T.). Taniropsis palpalis Brnrd. Various localities from False Bay to East London (WEEE Iais pubescens (Dana). Durban, on Sphaeroma walkeri; and Inhambane, on Sphaeromids (U.C.T.). Protojanira perbrinckt Brnrd. 19556. Natal and Zululand, freshwater. Peraeo- pod 1 of 9 resembles that of 3. Alloniscus pigmentatus B-L. Inhambane (U.C.T.). AMPHIPODA Microlysias xenokeras Stebb. False Bay (U.C.T.). Ampelisca diadema (Costa). Saldanha Bay (U.C.T.).. Leucothoe spinicarpa (Abildg.). Portuguese East Africa (S. Afr. Mus.). ADDITIONS TO FAUNA-LIST OF S.A. GRUSTACEA AND PYCNOGONIDA 7 Leucothoe richiardit Less. Saldanha Bay and False Bay (U.C.T.). Stenothoe gallensis Wikr. Knysna (U.QC.T.). Stenothoe assimilis Chevr. = valida Dana. Temnophlias capensis Brnrd. Buffels River mouth (south of Port Nolloth) to Still Bay (U.C.T.). Barnard, 1954. Temnophiias hystrix Brnrd. 1954. Hondeklip Bay (west coast), Gape Hangklip (south coast) (U.C.T.). Palinnotus natalensis Brnrd. Port Shepstone and Richmond (Natal) (U.C.T.). Colomastix pusilla Grube. Saldanha Bay (U.C.T.). Perioculodes longimanus (B. & W.). Saldanha Bay (U.C.T.). Synchelidium haplocheles (Grube). Inhambane (U.QC.T.). Recorded by Walker 1904 from Ceylon as brevicarpa B. & W. Calliopiella michaelsent Schell. Saldanha Bay and False Bay (U.C.T.). Occurs under the shells of Patella harmonizing in colour (mauve) with the animal’s foot. Nototropis guttatus (Costa). Olifants River mouth (U.C.T.). Recorded from the Mediterranean, Canaries, and Senegal. Paramoera capensis (Dana). Various localities from Port Nolloth to Natal (W-E.T:).: Eriopisa chilkensis (Chilton). Knysna, and St. Lucia Bay (U.C.T.). Barnard, IQ5I. Megaluropus agilis Hoek. Saldanha Bay (U.C.T.). Melita zeylanica Stebb. Various localities from Olifants River mouth (west coast) to Inhambane (U.C.T.). Melita fresnel: (Aud.). Inhambane (U.C.T.). Maera subcarinata (Hasw.). Algoa Bay (U.C.T.). Dexamine spiniveniris (Costa). Saldanha Bay (U.C.T.). Paradexamine pacifica (Thomson). False Bay (U.C.T.). See Barnard, 1930, ‘Terra Nova’ Rep., viii, 4, p. 389, figs. Polycheria atolli Wikr. Lamberts Bay, and Algoa Bay (U.Q.T.). Talorchestia ancheidos Brnrd. Inhambane (U.C.T.). Talorchestia inaequalipes Brnrd. 1951. Saldanha Bay (U.C.T.). Parorchestia notabilis Brnrd. 1935. Inhambane (U.C.T.). Originally des- -cribed from Cochin State, India. Hyale maroubrae Stebb. Knysna (U.C.T.). Hyale saldanha Chilton. Port Elizabeth (U.C.T.). FAyale grandicornis (Kroyer). Various localities from Lamberts Bay to Natal (W:@ Et). Sce also: znfraip: 933) figs 46: Grandidierella bonniert Stebb. Durban, and Inhambane (U.C.T.). Grandidierella chelata Brnrd. 1951. Port St. Johns and Bashee River mouth (CECeL:).. Grandidierella lutosa Brnrd. 1952. Hermanus (U.C.T.); Kleinmond (Bat- hurst Division) (Rhodes Univ. Zool. Dept.). Lembos jassopsis Brnrd. 1951. Saldanha Bay (U.C.T.). Lembos podoceroides Wikr. Inhambane (U.C.T.). Recorded from Ceylon, Maldives, Red Sea, Paumotu Archipelago. Eurystheus semidentatus Brnrd. Algoa Bay (U.C.T.). Cheiriphotis megacheles (Giles). Saldanha Bay, False Bay and Algoa Bay (ola i Chevalia aviculae Wikr. Inhambane (U.C.T.). Ampithoe ramondi Aud. Inhambane (U.C.T.). Cymadusa australis (Brnrd.). Paternoster Bay and False Bay (U.C.T.). Delagoa Bay (U.W.). 8 ANNALS OF THE SOUTH AFRICAN MUSEUM Erichthonius brasiliensis (Dana). Delagoa Bay (U.W.). Cerapus abditus Templ. False Bay, Algoa Bay, and Inhambane (U.C.T.). Siphonoecetes orientalis Wikr. Inhambane (U.C.T.). Delagoa Bay (U.W.). Siphonoecetes dellavalle: Stebb. St. Helena Bay and False Bay (U.C.T.). Pseudoprotella phasma (Montagu). 29° 16’ S., 14° 48’ E. s.s. ‘Africana’ (per WECaie): Metaprotella haswelliana (Mayer) var. taprobanica Mayer. Inhambane (U.C.T.). STOMATOPODA Conchoderma huntert (Owen). Natal, on jellyfish. Barnard, 19554. Poecilasma (Temnaspis) amygdalum Auriv. Delagoa Bay, on Palinurid (U.W.). COPEPODA PARASITICA Caligus pelamydis Kroyer. False Bay, on snoek, Thyrsites atun; Table Bay, on Trigla capensis (U.Q.T.). Caligus lunatus Wilson, 1928. False Bay, on Seriola lalandei (U.C.T.). Pre- viously recorded from Red Sea, on Seriola aurovittata (Wilson). Caligus. coryphaenae St. & L. Table Bay, on a shark (U.C.T.). Caligus rapax M. Edw. Table Bay, on Merluccius capensis and Trachurus trachurus (U.C.T.). Caligus mauritanicus Brian 1924. False Bay, on Pomatomus saltator (U.C.T.). Caligus (?) bonito Wilson 1905. False Bay, on Sarda sarda. § only; record should be checked on 2 material (U.C.T.). Caligus zei Norm. & Scott 1906. False Bay, on Thyrsites atun. Agrees with the description, but a check on further material is desirable (U.C.T.). Caligus brevicaudatus A. Scott 1901. Table Bay, on Trigla capensis (U.C.T.). Caligus labracis 'T. Scott 1902. ‘Table Bay, on Clinus superciliosus (U.C.T.). Pennella exocoeti (Holten). Length 40 mm. On Exocoetus. (Fish. Research Ship ‘Africana II’, 1951.) Charopinus ramosus Kroyer. ‘Table Bay, on a skate (U.C.T.). Clavellisa scombri (Kurz). ‘Table Bay, on gills of Scomber colias (japonicus) (WU). ©. ike): | Parabrachiella insidiosa (Heller). ‘Table Bay, on Merluccius capensis (U.C.T.). P. australis Wilson (see: Barnard, Ann. S. Afr. Mus., xli, p. 298, 1955) should probably be made a synonym. DECAPODA Fam. INACHIDAE Gen. Paratymolus Miers 1879. Miers, Proc. Zool. Soc. Lond., p. 45. 1880. Haswell, Ann. Mag. Nat. Hist (5), v. p. 302. 1882. id., Cat. Austral. Crust., p. 142. 1884. Miers, Zool. H.M.S. ‘Alert’, Crust., p. 261. 1893. Ortmann, ool. Jahrb. Syst., vil, p. 34. 1895. Alcock, 7. Aszat. Soc. Bengal, \xiv, p. 173. 1929. Balss, Denkschr. Ak. Wiss. Wien., cii, p. 3 (placed in group Camposcioidea). Carapace ovoid or subpentagonal, convex, anterior third declivous. Rostrum prominent but short, apically bifid, dorsally grooved. Eye-stalks fairly long, ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 9 straight, movable but not retractile, cornea subglobular, scarcely oblique; orbits incomplete ventrally. Ant. 1 folding longitudinally, sockets separated by aseptum. Ant. 2 peduncle slender, basal joint short. Mxp. 3 suboperculiform, flagellum well developed. Chelipeds (2) short, wrist with strong upstanding spine. Legs not elongate, dactyls slender, nearly straight. Genital openings of @ sternal. Abdomen ¢ 5-segmented, 2 7-segmented. Differs from Achaeus in the short basal joint of ant. 2, less oblique and more globular cornea, and short legs. Paratymolus pubescens Miers Bigs.) iy 2) to79. | Miers, loc. cit., p..45, pl. 2, fig. 6. 1880. Haswell, loc. cit., p. 303, pl. 16, figs. 1, 2 (bztuberculatus). 1882. id., loc. cit., p. 142 (bztuberculatus). 1893. Ortmann, loc. cit., p. 35, pl. 3, fig. 2 (ventral view rostrum and mxp. 3). 1910. Rathbun, Vid. Selskr. Skr. Copenhagen (7), v. p. 317. All exposed surfaces of carapace and appendages covered with a close velvety ‘pubescence’ (see Remarks). Carapace pentagonal; posterior two-thirds nearly flat, anterior third declivous. Rostrum short, broadly triangular, dorsally grooved, apically bifid, lobes rounded. A supraorbital tooth. Antero- lateral margin sharply bidentate, hinder tooth the larger, forming the angle between antero- and postero-lateral margins; postero-lateral margin fading out posteriorly on the dorsal surface, with 2 inconspicuous tubercles; lateral margin (as seen in dorsal view) with a conical tooth on the branchial region. Dorsal surface with 2 conical projections in front of the gastric region, and a median rounded inconspicuous hump on the cardiac-intestinal regions. Lower margin of carapace with feebly plumose setae in front, followed by closely set straight, tubular setae. Basal joint of ant. 2 short, and joint longer, 3rd longer than and, clavate setae on lower margin of 2nd joint, and on lower and upper margins of grd joint, those on lower margin of 3rd joint especially long; flagellum about equal to last 2 peduncular joints, with a few long setae. Eye-stalks short, cornea subglobular, a short clavate seta on anterior apex of stalk at base of cornea. Mxp. 3 external surface of 3rd and 4th joints and exopod with ‘pubescence’. Cheliped with ‘pubescence’, anterior margin of arm with 4 clavate setae, the strong acute process of the wrist with large clavate setae on posterior margin, smaller ones on anterior margin, hand and fingers also with clavate setae, cutting edges of finger and thumb with several small denticles, stronger on thumb than on finger, and numerous short simple and longer clavate setae. All legs of all specimens detached; the longer ones being presumably the anterior ones. Dactylus longer than preceding joint, with 3 or 4 retrorse denticles, the middle one the largest; a series (5-6) of long clavate setae on both anterior and posterior surfaces. IO ANNALS OF THE SOUTH AFRICAN MUSEUM 225° 59-80 9.959 Sere. its Ht lditey ir Paratymolus pubescens Miers. a. dorsal view of carapace. b. lateral view, with portion of lower margin further enlarged, and one seta further enlarged. c. surface view and section of one of the scale-like setae forming the pubescence on carapace. d. antenna 2, with seta further enlarged. e. maxilliped 3, with two kinds of setae further enlarged. f. exopod of maxilliped 3. g. right cheliped. h, clavate seta from cheliped further enlarged. 7. dactylus of leg. ADDITIONS TO FAUNASLIST OF S.A. CRUSTACEA AND PYCNOGONIDA II Pleopod 1 §: the basal joints of the two appendages are fused to form a transverse plate, to which the separate distal joints are articulated. The latter very stout, especially the posterior flange of the seminal channel proximally; apex curved outwards, thickly setose. Pleopod 2: the two appendages are separate and normal in shape. : Length ovig. 2 6:5, breadth 5-5 mm. As preserved, pinky-brown, eyes maroon. Fic. 2. Paratymolus pubescens Miers. Sternal view of g abdomen with 1st (fused) and 2nd pleopods (apical setae omitted on left pleopod 1). Locality. Inhambane, Portuguese East Africa. 1 ¢, 2 ovig. 99, 2 immature, in Zostera beds (U.C.T. Jan. and July 1954). Distribution. Japanese Seas, Siam (pubescens); Queensland (bituberculatus). Remarks. ‘The ‘pubescence’ would be better described as a ‘lepidosis’, the setae being flattened and broadly ovate, and more or less imbricate. These specimens correspond so closely with the descriptions and figures of Miers’s pubescens and Haswell’s bituberculatus, and Alcock’s hastatus, that there seems no justification for a n. sp. They have broader shoulders than in Miers’s and Alcock’s figures, thus corresponding better with Haswell’s figure. ‘The upper distal apex of the hand is more prominent than in Miers’s figure, but 12 ANNALS OF THE SOUTH AFRICAN MUSEUM less prominent than in Haswell’s, approximately as in Alcock’s figure. The 4 small teeth on posterior border of the arm, mentioned by Haswell, are not present; whereas the 4 conspicuous clavate setae on the anterior margin apparently were not present in Haswell’s specimen, but are shown in Alcock’s figure. Alcock’s hastatus apparently is more tuberculate than either Miers’s or Haswell’s or the present specimens, but I think it should be regarded as synonymous. The size of the plate formed by the fusion of the coxopodites (Cochran 1935) or ‘bridge’ joining the protopodites (Flipse, 1930) of the 1st pleopods is remarkable. Fam. ACANTHONYCHIDAE Gen. Antilibinia McLeay 1950. Barnard, loc. cit., p. 36, fig. 7c, d. Through the kindness of Dr. H. B. S. Cooke (Geology Dept. Witwatersrand University) the Museum has received a fine ovigerous 2 of A. smithii McLeay from the coast of Pondoland, measuring 62 X 57 mm. This specimen confirms Krauss’s statement and figures that there are seven abdominal segments in both sexes. Gen. Menaethiops Alcock 1950. Barnard, loc. cit., p. 38. Two dozen specimens of this genus have been examined. They appear to belong to three species. One is fascicularis, but the other two cannot easily be assigned to any species recorded from the western Indian Ocean, viz. brevicornis M. Edw. 1868, acutifrons M. Edw. 1868, bicornis Alcock 1895, and Lenz 1905, nodulosa Nobili 1905, corniculata Klunz. 1906, and contiguicorms Klunz. 1906. The descriptions of these species make no mention of some characters which appear to be of specific value. ‘Iwo n. spp. are therefore instituted, with local names. The validity of these three species may appear doubtful because of the overlap in distribution. ‘Twenty-four is a very small number of specimens; gradations may later be collected; but on the available evidence they cannot be referred to only one species. The sternal sculpture in the § may not prove to distinguish all the species of this genus, but it is very distinctive in two of the present species. Menaethiops fascicularis (Krauss) Fig. 3a-c. 1929. Balss, Denkschr. Ak. Wiss. Wien., cii, p. 9, fig. 3. 1950. Barnard, loc. cit., p. 39, fig. 8. This species is well characterized by the absence of a projecting anterior corner on the supraorbital margin, which passes in a sinuous curve into the ADDITIONS TO FAUNA-LIST OF 8.A. CRUSTACEA AND PYCNOGONIDA 13 rostral prong. Suborbital margin denticulate. Outer apex of basal joint of ant. 2 strongly produced. The distal margin of the basal joint, adjoining the insertion of the following joint, is nodulose or crenulate, and these crenulations continue on a short longitudinal ridge in the middle of the basal joint. The rim of the antennulary socket is also crenulate. These crenulations are particu- larly distinct in the Delagoa Bay 6 mm. 9. The postocular tooth (hind corner of the supraorbital margin) is nearly linear or narrowly triangular, serrulate (at least on hind margin). A slight swelling on ocular peduncle, but no definite spine. The sternite between the chelipeds is smooth, polished, and concave in front of a transverse series of inconspicuous granules, behind which the surface is covered with impressed punctae. The other sternites are also punctate, without any granules. The above characters are constant in 3 gg (6-7 mm.) and (except the sternite) in two non-ovigerous 99 (4:5 and 6 mm.). Additional locality. Delagoa Bay. 1 non-ovig. 2 (6 mm.). (U.W.) Menaethiops natalensis n. sp. Fig. 3d—f. In general similar to fascicularis, but rostral prongs a little longer. Supra- orbital margin with well-marked acute anterior projection; postocular tooth serrulate as in fascicularis, but larger and more triangular, though not always as broadly triangular as here figured. Suborbital margin very feebly (if at all) denticulate, usually nearly straight, but sometimes convex in the middle. Outer apex of basal joint of ant. 2 not strongly produced (but somewhat variable). Rim of the antennulary socket feebly crenulate, but no ridge on the basal joint of ant. 2. A well-marked spine on ocular peduncle. Sternite between chelipeds polished in front and punctate behind with a scarcely perceptible ridge between the two areas. Other sternites punctate, without granules. Pleopod 1 ¢ as in fascicularis. Length 10 mm. Localities. Umpangazi (Impengazi), Natal. 1 g (U.C.T); Delagoa Bay. 2 $5 (7 mm.) 5 ovig. 92 (7-10 mm.) g immature (4-7 mm.) (U.W.) Remarks. In one specimen the right anterior corner of the supraorbital margin is rather strong and curved outwards, approximating to delagoae. In two other specimens the anterior corner on one side (right or left) is bevelled off as in fascicularis. M. Edwards’s figure of brevicornis shows a transverse line of granules not only on the sternite between the chelipeds, but on the three following sternites. Menaethiops delagoae n. sp. Fig. 39-1. In general similar to fascicularis, but rostral prongs longer, distally divergent. Supraorbital margin sinuous, the acute anterior corner larger than the acute 14 ANNALS OF THE SOUTH AFRICAN MUSEUM “i % ey oN Fic. 3. Menaethiops fascicularis (Krauss). a. dorsal view of rostral prongs and antero-lateral portion of carapace. b. ventral view of same. c. sternite between chelipeds, and apex of abdomen. Menaethiops natalensis n. sp. d,e, f. as ina, b, ¢c respectively. Menaethiops delagoaen. sp. g, h, i. as in a, b, ¢ respectively. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA ES posterior corner (postocular tooth), both non-serrulate. Suborbital margin straight, denticulate posteriorly. Outer apex of basal joint of ant. 2 acutely, but not strongly, produced. Rim of antennulary socket not crenulate, and no ridge on basal joint of ant. 2. No spine on ocular peduncle. Sternite between chelipeds with 2 groups of 6—7 granules in front of the abdominal groove, polished and concave in front of a transverse series of granules, punctate behind. Other sternites punctate, without granules. The suture between Ist and 2nd sternites opposite the middle of the 6th abdominal segment, not as in the other species opposite the suture between 6th and 7th segments. Pleopod 1 ¢ as in fascicularis. Length 10-5 mm. Locality. Delagoa Bay. 1 3 (U.W.) Remarks. M. Edwards does not figure the sternites of acutzfrons. Fam. MAMAIDAE Gen. Cyphocarcinus M. Edw. 1868. M. Edwards, Nouv. Arch. Mus. Paris, 1V, p. 73 (9). 1875. Paulson, Red Sea Crust, p. 3 (3) (Ixton). 1889. Cano, Bol. Soc. Napoli (1), 3, pp. 86, 180 (Podohuenia). 1923. Gravier, Bull. Mus. N. H. Paris, 1923, 3, p. 214 (g) (Stenocarabus). 1929. Balss, Denkschr. Ak. Wiss. Wien., cli, p. 22. 1945. Stephensen, Dan. Sci. Invest. Iran, 4, p. 108, fig. 21 A, B (plp. 1, 2. 3). — Cyphocarcinus capreolus (Paulson) Fig. 4. 1875. Paulson, loc. cit., p. 3, pl. 2, figs. ra—f (g) (Lxion c.). 1923. Gravier, loc. cit., p. 214, figs. 1-8 (¢) (Stenocarabus suspensus). Carapace of both sexes resembling the figures given by Paulson and Gravier, i.e. without conspicuous knobs, only a rostral knob followed by 2 small granules in the median line; some tiny granules along the costate lateral margin, dis- cernible more by their white colour than by their relief; a rather conspicuous white granule just in front of the faint cervical groove. A transverse section through the hinder third of the carapace would show an almost evenly curved semicircular profile in the g, but in the 2 a somewhat squarish profile owing to the presence of an obscure blunt longitudinal ridge, on which are 2-3 little granules; below this the lateral wall of the branchial region is nearly vertical. All these granules are better developed in 9 than in g; in addition 9 has 4 granules en carré in the middle of the carapace and a medio-dorsal series (3-5) towards the hind end. Orbit circular, with a slit on anterior, posterior, and inferior margins; upper margin anteriorly with a small setiferous tubercle. Surface closely pitted (as in Paulson’s and Gravier’s figures), and covered with short fine pile; longer setae on supraorbital ridge and on rostral knob and prongs. 16 ANNALS OF THE SOUTH AFRICAN MUSEUM f | e Fic. 4. Cyphocarcinus capreolus (Paulson). a. side view of g carrying a piece of Cymodocea. b. dorsal view of carapace (the tips of the upper branches of the rostral prongs have been drawn too far apart). c. view of the left orbital region. d. ventral view of left antennulary and antennal region: ant. 1 removed from socket; lettering corresponding with that in fig. c. e. chela of J. f. chela of 9. g. pleopod 1 g, with apex further enlarged. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 1 Abdomen in both sexes of 7 segments; in 3g 1st short, 2nd and 3rd much wider, remainder gradually decreasing, but 6th slightly widened across distal margin, 7th longer than its basal width; in 2 (nonovigerous) widest across suture between 3rd and 4th segments, 7th shorter than its basal width; 2nd segment in both sexes with a small blunt median knob. Chelipeds subequal, larger and more robust in 3 than 9, arm with 4 granules on upper margin, and 5-6 on lower margin (not shown in Paulson’s figure). In ¢ finger with strong tooth near base, followed by a semicircular gap, and a distal cutting-edge with 5 rounded denticles, thumb with 8-9 denticles, one larger than the others, distal to which the cutting-edge resembles that of the finger. In 9 finger with no tooth (or only a very small one) at base, cutting- edge with 9 rounded denticles, thumb with 9 similar denticles. Paulson’s figure of the cheliped of his 3 almost corresponds with that of the present 99, from which one may assume that the toothed and gaping finger seen in the present $¢ is developed only when the animal is fully grown. The curve of the cutting-edge of the thumb in Paulson’s figure is not exactly like that of the present 99, but seems to indicate that at a future, perhaps the next, moult it would assume that of the present gg. Gravier’s specimen was in the same stage as Paulson’s. Dactylus of all legs with 7-8 strong spiniform teeth, increasing in size distally, in addition to the unguis. Pleopod 1 ¢ as in Stephensen’s figure; the process on the basal transverse joint not so strong as in Gravier’s figure 8. Length ¢ @ 11°5 from rostral knob, 13-14 mm. including the rostral prongs. As preserved, madder-brown, pinkish or pale salmon, abdomen of 2 speckled, chelipeds and legs paler, finger and thumb of chelipeds white, eyes maroon. Locality. Inhaca Island, Delagoa Bay. 4 3d, 9 ovigerous and 1 non- ovigerous 99, 2 juv. (U.W.) Distribution. Red Sea (capreolus); Madagascar (suspensus). Remarks. Laurie (1915. 7. Linn. Soc. Lond., xxxi, p. 433), Nobili (1906 and 1907), and Balss (1929) regard capreolus as a synonym of minutus M. Edw. 1868. Balss accepted two species: the knobbly minutus and the smooth suspensus, but nevertheless made the smooth capreolus a synonym of the former. I am inclined to concur with Dr. Gordon (in litt. July 1954) that in all probability there is only one somewhat variable species whose distribution extends from the Red Sea, Persian Gulf, Andaman Is., to the Seychelles, Madagascar, and Delagoa Bay (Stephensen misquoted M. Edwards and included Zanzibar). For the present, however, I prefer to record these speci- mens, smooth (except for the granules) in both sexes, as capreolus. Laurie and Gravier both refer to M. Edwards’s fig. 8 which shows in the 9 the rostral prongs united for a distance about equal to the length of the anten- nulary sockets, and then divergent, with a V-shaped sinus; very different from Paulson’s fig. 15 and the present specimens. Rathbun (i1g11. Trans. Linn. Soc. Lond., zool., xiv, p. 255), however, suggests that M. Edwards over- 18 ANNALS OF THE SOUTH AFRICAN MUSEUM looked a narrow slit separating the bases of the prongs. But that does not obviate the difference in the shape of the sinuses. Nobili and Rathbun record the 3 as carrying a long cylindrical sponge ‘projecting’ from its rostral prongs; but it is not stated (I have not seen Nobili) whether the sponge was growing on the crab or had been picked up by the crab. One of the present ¢ specimens has a length (22 mm.) of the stem of the marine flowering plant Cymodocea ciliata firmly wedged between the prongs. The stem is oval in section and will only fit one way between the prongs. Another ¢ (total length 11 mm.) carries edgwise a piece of the leaf of Cymo- docea, approximately 11 X 8mm. A Q (total length 13 mm.) carries, edgewise, a piece of leaf 10 » FGA, SAA? \RRY Atos SSN Zan. SN Kz Fic. 16. Typhlocarcinodes piroculatus (Rathbun). Antennae 1 and 2. Abdomen of g. Pleopod 1 4, posterior view of right appendage, with lateral view of apex (to same scale). Length ¢ 8-5 mm. (by calipers, not over the curve), 9 7 mm.; breadth g I1mm., 29mm. As preserved, dirty white, the fringes on the legs brown with a reddish tinge. Locality. Delagoa Bay. 1 3, 1 non-ovig. 9. (U.W.) Distribution. Amirante Is., Seychelles, 34 fms.; Philippine Is., 36 metres. Remarks. In spite of a few minor differences from the descriptions of Rathbun and Tesch (e.g. the flagellum of antenna 2 is longer, especially in the ¢) these specimens appear to belong to this species. Both are larger than the other known specimens. Tesch did not describe the pleopods of the g. Nor have they been pre- viously described in any species of this genus. Stephensen (1945, Dan. Sez. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 37 Invest. Iran, pt. 4, p. 227) notes that, as far as recorded, pleopod 2 in the Rhizopinae is short except in Notonyx. The present species forms another exception. Fam. DROMIIDAE Pseudodromia integrifrons Hend. Figs 17: 1888. Henderson, ‘Challenger’ Rep., xxvii, p. 16 footnote (sine descr.). 1893. id., Trans. Linn. Soc. Lond. (2), v, p. 406, pl. 38, figs. 7-9. 1950. Barnard, loc. cit., p. 316 (in key only). Pubescence on surface of carapace chiefly in the shallow grooves marking the regions; the lateral and anterior margin of carapace, however, with thick fringe of plumose setae, especially noticeable on the rostrum. The 6th joint of 5th leg has 2 strong spines on outer surface of apex (not 3 as in Henderson’s description). Length of carapace 3 7°5 mm., ovig. 9 65 mm. 36 ova, about I mm. in diameter. Locality. Inhambane. §¢ and ovig. ee (UCT) _ Distribution. Ceylon; Obock. Cryptodromia tomentosa (Heller) 1861. Heller, SB. Ak. Wiss. Wien., sive P-.214- Fic 17. 1875. Paulson, Red Sea Crust., p. 83. Pseudodromia integrifrons Hend. Left 5th : . leg, outer surface (on left) and anterior 1878. Hilgendorf, MB. Ak. Wiss. (appeny (icra wiew- (onbnieHol Berlin, p. 813, pl. 2, figs. 3-5. 1880. Kossmann, Reise. Roth. Meeres., pt. 2, p. 68. mooe:) de Man, 7. Linn. Soc. Lond., xxii, p. 212. ?1903. Borradaile, Fauna Geogr. Mald. Lacc. Arch., Il, p. 577, pl. 33, fig. 3 (hirsuta). 1942. Ward, Mauritius Inst. Bull., II, 2, p. 70. A non-ovigerous 9, 14 X 14 mm., is more like Hilgendorf’s figure of penta- gonalis in general outline, but the large antero-lateral tooth is truly marginal, and the median rostral tooth is only just visible in dorsal view as in his figure of tomentosa. ‘The tooth behind the branchial groove is slightly less prominent than in the figure of the former species, and the supraorbital margin is less concave than in that of the latter species. Dorsal surface of carapace com- pletely covered with tomentum. Localities. Ibo (Hilgendorf); Inhambane. 1 9. (U.C.T.) Distribution. Red Sea, Chagos Archipelago. 38 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. C. pentagonalts Hilg. (loc. cit, p. 814, pl. 2, figs. 1, 2) should, I think, be united with tomentosa. According to Hilgendorf the antero-lateral tooth is truly marginal in tomentosa, but inframarginal in pentagonalis. Both species were found together at Ibo; pentagonalis has been recorded also from Mauritius and India (Henderson, 1893). Alcock (1899) and Henderson also thought the two species were not distinct. C. hirsuta Borrad. is also very similar. Regarded as synonymous with canaliculata Stimpson 1858 by de Man (1887, Arch. Naturg., lili) and by Ortmann (1892). Rathbun (1911) records both canaliculata and pentagonalis from the Seychelles. = sh ry ally s 2? a si >» 9 eer o ’ Se AR ST Wiens Se ie ey SA6 a a 2 > Fic. 18. Ebalia agglomus n. sp. a, 6. dorsal and lateral views. c. mxp. 3. d. abdomen ¢. e. pleopod 1 @. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 39 Fam. LEUCOSIIDAE Ebalia agglomus n. sp. Bice re: g—Carapace as broad as long, but postero-lateral margins slightly con- verging behind; branchial regions separated from the median gastro-cardiac- intestinal regions by a well-marked furrow (not as deep as in glomus), 2 large medio-dorsal cardiac tubercles, 2 smaller intestinal ones, smaller tubercles and granules on the cardiac and branchial regions, and minute granules on intestinal region; gastric region and front quite smooth (with sparse impressed punctae); front with shallow median groove; supraorbital and antero-lateral margins beaded; pterygostomial ridge prominent, beaded, separated by the shallow hepatic-branchial gap from the beaded postero-lateral margin; hind margin with slightly enlarged median and lateral granules. Cheliped about 14 times length of carapace; arm shorter than carapace, granulate on upper, lower, and outer surfaces especially basally; wrist, palm and fingers minutely granulate (appearing smooth even under x8 lens). Abdomen with 3rd—5th segments fused, lateral margins of 6th and 7th segments crenulate. Pleopod 1 straight, distally enlarged, with pointed apex. Length and breadth 7-5 mm. Locality. Inhambane. 1 ¢ (U.C.T.). Delagoa Bay. 1 3 (U.W.). Remarks. At first glance resembling a less highly sculptured form of glomus or diadumena, but distinguished by the 1st pleopod . Gen. Cryptocnemus Stimpson 1858. Stimpson, Proc. Ac. Nat. Sci. Philad., x, p. 161. 1918. Ihle, Siboga Exp. monogr., xxxix, b2, p. 285 (with key to species) and p. 317 (references). i Cryptocnemus holdswortht Miers Fig. 19: 1877. Miers, Trans. Linn. Soc. Lond. (2), zool., I, p. 241, pl. 38, figs, 30-2. 1906. Laurie, Herdman’s Ceylon Pearl Oyster Rep., Suppl. Rep. V, p. 356. 1910. Lenz, Voelizkow Reise Ostafr., Il, p. 544. 1918. Ihle, loc. cit., p. 286 (in key) and p. 317 (in list of species). Laurie remarks that the shape of the carapace is somewhat variable; and that Miers’s 9 specimen and his two 99 were the only three specimens recorded to that date. Lenz had one 9, whose abdomen agreed with Miers’s figure. In the present specimen the abdomen is for the most part parallel-sided, and all 7 segments are distinct, but the 1st and 2nd are immovably fused; the hind margin of the Ist segment (probably, but may be the front margin of the 2nd) with a transverse ridge. A strong ridge on the pterygostomial region, anteriorly forming the border of the inhalent channel. 40 ANNALS OF THE SOUTH AFRICAN MUSEUM IMCs HO. Cryptocnemus holdsworthi Miers. Carapace and abdomen of immature (? 9). Length 6:25 mm., breadth 10 mm. As preserved, white. Locality. Inhambane. 1 immat. ? 9, dredged in 4-1-5 metres, bottom sand, with patches of Cymodocea weed. (U.C.T. July 1954). Distribution. Ceylon. Fam. PORCELLANIDAE Porcellana delagoae n. sp. Hic 20. Median frontal tooth rather narrow, length subequal to basal width; margins of frontal teeth thickly fringed with short plumose setae. Carapace with scattered plumose setae, chiefly anteriorly and on the lateral branchial regions; the 2 semicircular postfrontal lobes rather prominent, with rather deep depression in front of them. Lateral margin with 3-4 spines above insertion of Fic. 20. Porcellana delagoae n. sp. a. carapace. b. ventral view of basal joint of left ant. 1. c. ventral view of peduncle of left ant. 2. d. right cheliped, the vertical flange on 4th joint seen foreshortened. e. external view of 3rd and 4th joints of cheliped. £ dactylus of leg. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 41 peduncle of ant. 2 in front of lateral indent, 5 behind it, increasing in size posteriorly. First peduncular joint of ant. 1 thick, dentate, the ventral apical transverse margin with a couple of small denticles and some long setae, dorsal apical margin with 1 spiniform tooth and some small denticles, outer apical angle with 3-4 strong teeth. Basal peduncular joint of ant. 2 with margin entire except at apex which is serrulate; following two joints each with apical tooth on front margin. Inner margin of grd joint of cheliped rather prominent, with tuft of plumose setae; 4th joint with strong spine on lower inner margin, upper inner margin produced as a rounded, denticulate flange projecting vertically, perpendicular to the plane of upper surface of 5th joint, the median keel spinulose; both margins of hand, and outer margins of finger and thumb spinulose, and with plumose setae, the latter especially numerous on outer margin of hand; median keel of hand spinulose, 2—3 denticles on base of finger; cutting-edges of finger and thumb furry on inner (lower) side. (Only the right cheliped present, which seems to be the smaller.) Legs with moderately numerous plumose setae on 2nd—4th joints; dactyls with 4-5 spines, the distal- most one the largest, followed by a tuft of setae. Length and breadth 4 mm. Pinkish-orange, somewhat faintly mottled, 2 red dots on the cardiac region. Locality. Delagoa Bay. 1 ovig. 9. (U.W.) Remarks. At first glance deceptively like dehaanii, but distinguished by the spinulose and setose cheliped. The upstanding flange on the 4th joint of cheliped, if normal, is curious. Among a large number of both sexes of dehaaniz I have found no tendency of this flange to become vertical and perpendicular to the plane of the upper surface of the 5th joint. Petrolisthes virgatus Paulson ISI, Qi. 1875. Paulson, Red Sea Crust., p. 87, pl. xi, fig. 4 (Paulson gives a reference to to a figure ‘pl. 1, fig. 9’ in another paper by himself). 1894. Ortmann in Semon, Austral. Reise, v, p. 28, pl. 2, fig. 5 (trvirgatus). 1905. Lenz, Abh. Senckenb. Ges., xxvii, p. 375 (trivirgatus). Integument not strongly calcified. Carapace, chelipeds and legs completely covered with short spinate spine-setae. No supraorbital or epibranchial spines. First peduncular joint of ant. 1 distally tridentate, inner and outer teeth acute, the broad median lobe with serrulate margin and 2-3 larger denticles. Second (ist free) peduncular joint of ant. 2 squarely lobed on front margin. Inner apex of 4th joint of cheliped sharply dentate, 5th joint with 3 sharp teeth on inner (front) margin; outer margin of hand proximally with sharp spiniform denticles, hidden by the thick marginal fringe of rather long spinate setae, which is continued along outer margin of thumb. Fourth joints of gnd-4th legs without spines on front margin, lower (hind) margin of 4th joint of 2nd leg with acute apical tooth, 4th joint of 4th leg 14 times as long as broad; both margins of legs strongly fringed with spinate setae. 42 ANNALS OF THE SOUTH AFRICAN MUSEUM Length 8-5 mm., breadth 8 mm. (As preserved in formalin) dark crimson- maroon dorsally, more crimson ventrally, sternum pale but speckled; carapace with 3 broad longitudinal pale, dark-bordered stripes, a pair of dark rings between the gastric-cardiac regions ending posteriorly in a pair of small shallow glabrous pits; abdomen with transverse crimson and pale bands; legs with pale stripes; chelipeds with pale stripe on hinder margin of 4th and 5th joints, hand with faint pale stripe along inner (upper) margin continued along finger, a more conspicuous pale stripe along middle of upper surface, con- tinued along cutting-edge of finger. Ortmann gave the. ground-colour as dark violet, paler below, with pattern as in the present specimens, but without mentioning the dark rings which are shown in Paulson’ figure. Locality. Delagoa Bay. 2 specimens. (U.W.) Distribution. Red Sea, Dar-es-Salaam, Zanzibar. hod oF 3 4 3 AG a * = Y i Ry q ae % % a AD % o f SSS edad SPORE és y PPE INE ATS ogee i 3 t ’ ras wget LE a v Wcengiaaloe’e's & \ RB yn. ise oe Bre. 32 Petrolisthes virgatus Paulson. a. carapace, denuded, the spinate setae shown only around the margin. b. ventral view of basal joint of left ant. 1. c. ventral view of peduncle of left ant. 2, with spinate seta further enlarged. d. spine-seta from front margin of hand of cheliped. Remarks. There can be little doubt that Ortmann’s species is the same as Paulson’s. From the figure (x 2) Ortmann’s specimens measured 10 X 7:5 mm. but the length has almost certainly been exaggerated in relation to the width. Paulson’s much better figure ( X 3) gives the size as (approx.) 6°17 X 5°7 mm. Fam. SERGESTIDAE Gen. Acetes M. Edw. 1950. Barnard, loc. cit., p. 822 Key to the South African species 1. Telson extending beyond middle of inner ramus of uropod, apically pointed. . erythraeus 2. Telson not extending beyond middle of inner ramus of uropod, apically truncate. MP ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 43 Acetes natalensis n. sp. Very close to znsularis Kemp (1917 Rec. Ind. Mus., xiii, p. 54, figs.) (= serru- latus (Kréyer)) from the East Indies, differing only in the 3 petasma. Outer flagellum of ant. 1 ¢ as in Kemp’s fig. 2¢ of insularis; 2 spinules proximal to a retrorse process on the side of the flagellum remote from the 2 large clasping spines. Petasma: proximal and distal extensions of the inner lobe more like those of indicus (Kemp, loc. cit., fig. 4g), but between the two portions an angular lobe or tooth on inner margin, distal portion with a pointed process beyond which extends the apical cylindrical portion, the latter with several granules or ‘pits’ (Kemp). Telson not reaching beyond middle of inner ramus of uropod, apex truncate but slightly convex, with a denticle at each corner. Length ¢ ca. 20 mm., 9 22-3 mm. Locality: Durban Bay, 4 63; 5 9° (U.G.T., 1951); also a single specimen from the same locality submitted in 1955 by Mr. W. D. Oliff of the Natal River Survey. Remarks. Examination of the latter specimen showed that I had not heeded Kemp’s warning (loc. cit., p. 44): ‘. .. two species are often found together. In such cases . . . identification is a tedious process, for each individual must be separately and carefully examined.’ U.C.T. collected many juveniles of the large erythraeus and I assumed that all the small examples were the same species. The character of the telson will at once distinguish the two species, reinforced in the case of gg by the flagellum of antenna 1 and the petasma. Fam. PROCESSIDAE Key to the South African genera 1. First peraeopod without exopod. oe eA LenOcessa: 2. First peraeopod with exopod. MMS hen NeKOtdeS, Gen. ProcessA Leach 1950. Barnard, loc. cit., p. 715 1951. Holthuis, Atlantide Rep., 2, p. 37 Key to the South African species 1. Rostrum narrow, slender. a. 2nd legs unequal. i. Body and 3rd—5th legs slender. 5th abdominal segment with hind corner rounded. . austroafricana ii. Body and 3rd—5th legs robust. 5th abdominal segment With ae corner dentate. ey mere cdules b. and legs equal (or subequal). i cae Margiimand 2. Rostrum broad at base, triangular in dorsal view. Nie Wea sie ya sonica 44. ANNALS OF THE SOUTH AFRICAN MUSEUM Processa aequimana (Paulson) 1875. Paulson, Red Sea Crust., p. 97, pl. 14, figs. 6, 6a (Nika ae). 1906. Nobili, Ann. Sci. Nat. zool. Paris (9), iv, (no specimens, translation of Paulson’s description). 1922. de Man, Siboga Exp. monogr., xxxix, a4, p. 44, pl. 4, figs. 19-19. 1937. Gurney, Proc. Zool. Soc. Lond., cvii, B, p. 87, pl. 1, figs. 1-10; pp. 91, 92 (in key); and pp. 92, 95, 97; pl. 2, figs. 33-5, pl. 3, fig. 36 (develop- ment). , Distinguished by the 2nd legs being equal, or nearly so, in length. Length 17 mm. Locality. Inhambane. 1 ovig. 9, in Zostera bed. (U.C.T. Jan. 1954). Distribution. Red Sea; East Indies. Remarks. Gurney records ovigerous 99 in late February. At Ghardaqa, Red Sea, the shrimps inhabit ‘the shallow reef-flat, which is largely laid bare at the lowest tides’. They appear to burrow in the sand during the day, and were only taken at night (loc. cit., p. 86). Processa japonica (de Haan) 1849. De Haan, Fauna Japonica, p. 184, pl. 46, fig. 6. 1920. De Man, Siboga Exp. monogr., xxxix, a 3, p. 208, pl. 18, figs. 53-534. 1937. Gurney, Proc. Zool. Soc. Lond., B, cvii, p. 88, pl. 1, figs. 16-19. Three ovig. 99, 30-36 mm., have the broad triangular rostrum and minute telsonic spines which Gurney notes as distinctive of this species. Recorded as ‘yellow (bright golden) shrimps burrowing in sand’ (Mrs. M. Kalk, U.W.). Locality: Delagoa Bay (U.W. 1955). Distribution. Japan; East Indies; east coast of Africa (Gurney does not state the exact locality of the specimen he examined from the John Murray Expedition). Gen. Nikowdes Paulson 1875. Paulson, Red Sea Crust., p. 98. Nikowdes danae Paulson 1875. Paulson, loc. cit., p. 98, pl. 14, figs. 5-5d. 1906. Nobili, Ann. Sci. zool. Paris (9), iv, p. 79, pl. 5, figs. 1-1f- 1937. Gurney, loc. cit., p. 89, pl. 1, figs. 20-5, pl. 2, figs. 26-9. One specimen, ca 27 mm. (if extended) is provisionally assigned to this species. The basal process of ant. 1 is short and broad, apex obliquely truncate, the internal corner projecting farther forwards than the external corner; thus differing from Paulson’s fig. 5d. Gurney does not figure this appendage, and Nobili’s paper is not available to me; de Man (1920, loc. cit., p. 193, pl. 16, figs. 50-50/) also gives no figure of it for his stbogae, and without a figure his description is not very enlightening. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 45 Locality: Delagoa Bay (U.W.). Distribution. Red Sea; Djibouti. Gurney (loc. cit., pp. 89, 91) considers sibogae from the East Indies synonymous with danae. Fam. ALPHEIDAE Alpheus bullatus n. sp. Pismo) 1950. Barnard, loc. cit, p. 742 (banded species from St. Lucia Bay). Rostrum extending to end of 1st peduncular joint of ant. 1, not strongly keeled, similar to that of crassimanus. Supraorbital spines present. Basal process of ant. 1 reaching to (or a trifle beyond) apex of 1st peduncular joint. Basal joint of ant. 2 with ventral (external) spine; blade of antennal scale reaching half way along spine, which does not quite reach end of last peduncular joint of ant. 2. First leg: margins of 4th joint entire, upper apex blunt; large chela ¢ (right) about twice as long as broad, outer half of upper surface with rounded granules, more or less in transverse rows, inner half with smaller granules and punctae; lower sur- face with scattered large punctae; finger knob-like, apex truncate; small chela J, a short denticle at upper apex of 5th joint, a short blunt tooth on upper surface of palm at base of finger, outer half of upper surface with rounded granules, inner half with punctae, lower surface with a few large punctae, finger subequal to palm. Both chelae in 9 similar to, and almost as robust as in g; the large chela on the left Be. on. side. Alpheus bullatus n. sp. Dorsal view of Third and fourth legs, 4th joint 3 times Hots, hewge Cael) al Gy Koen eis i i upper surfaces. as long as broad, lower margin with spaced groups of 2—3 setules, no tooth at lower apex; 6th joint of 3rd leg with 7, of 4th leg with 5 stout spines. Fifth leg more slender than 3rd and 4th. Dactyls of 3grd—5th legs biunguiculate. Telson not strongly narrowed distally; with median groove between well- marked ridges on which are inserted 2 pairs of strong spines. Length 3 22 mm., 2 27 mm. As preserved, two broad transverse bands on carapace and one on each abdominal segment brick-red; no medio-dorsal abdominal pale stripe; spine on outer ramus of uropod dark brown or blackish. Localities. Delagoa Bay. 1 g, 1 9. (U.W.) St. Lucia Bay (S. Afr. Mus.). 46 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. Although the chelae are missing, all three specimens from St. Lucia Bay agree with the Delagoa Bay specimens in all other features, including the colour pattern. These specimens do not seem to agree with any described species. The large chela, although broader, seems to resemble in general that of lutini Cout. and of phrygianus Cout.; but both these species have a tapering telson and no supraorbital spines. ! The large chela is indeed very similar to that of malleator Dana, as figured by Coutiére (1899, fig. 262), and in other characters also there is a close resemblance to the specimen described by Balss (1914, ate. D. Zentr. Afr. Exp., I, p. 98, figs. 1-5) as ‘tuberculosus’ (non Osorio) and later (1916, Beitr. Meeresf. Westafr., II, p. 22) identified as malleator Dana var. edentatus Zimmer 1913. A. malleator is recorded from the Eastern Pacific, West Indies, Brazil, and West Africa. The finger of the large chela of the present specimens, however, is more knob-like than hammer-shaped: the proximal end only slightly bulbous and the apex abruptly truncate. The strongly bicarinate telson of the Delagoa Bay specimens seems to be an unusual feature in the genus, but it is scarcely noticeable in the smaller specimens from St. Lucia Bay. Fam. PALAEMONIDAE Subfam. Palaemoninae Key to South African genera and subgenera (after Holthuis) I. Branchiostegal spine present. A. Branchiostegal groove absent. Lower margin of rostrum with double row of setae, which arise some distance above the bases of the teeth. 6th joint of Sth leg without transverse rows of setae distally. —Two median spines on telson very strong. Pleopod 1 § with well developed appendix interna. Mandibular palp 2-jointed. . . . . Leander L. tenuicornis (Say), circumtropical. B. Branchiostegal groove present. Lower margin of rostrum with (usually) a single row of setae (if two rows, the setae arise near the margin). 6th joint of 5th leg with transverse rows of setae distally. ‘Two median spines on telson slender. Pleopod 1 3 appendix interna rudimentary or absent. . . Palaemon 1. Mandibular palp 2-jointed. . . . . subgen. Palaeander elegans Rathke, maculatus (Thallw.), squilla (Linn.) (adspersus Rathke), west coast of southern Africa. 2. Mandibular palp 3-jointed. . . . . subgen. Palaemon a. Pleopod 1 g appendix interna rudimentary. . +s» Concinnus b. Pleopod 1 g appendix interna absent. i. Rostrum very slender, curving upwards. Free part of antennulary flagellum subequal to fused part. Peraeopod 2 slender, 5th joint twice as long as chela. 2 ss Gebales ii. Rostrum curving upwards. Free part of antennulary flagellum 3% times the fused part. Peraeopod 2 5th joint shorter than chela. . pacificus iii. Rostrum straight. Freshwater. os + a OapeRsEs II. Branchiostegal spine absent. Hepatic spine present. . « « « Macrobrachium ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 4.7 Leander tenuicornis (Say) more. Say, Proc. Ac. Sc. Philad., I, p. 249. 1875. Paulson, Red Sea Crust., p. 116, pl. 17, figs. 3-5 (torensis). 1925. Kemp, Rec. Ind. Mus., xxvii, p. 302, fig. 11 (append. intern., plp. 1 ). 1950. Holthuis, Szboga Exp. monogr., xxxix, a 9, p. 26, figs. 1, 2. 1952. id., Allan Hancock Publ. Occ. Pap., No. 12, p. 155, pls. 41, 42 (synonymy and references). Two specimens with rostral teeth *¢° and 7+. Locality. Delagoa Bay. (U.W.) Distribution. Red Sea, Réunion. Indo-Pacific and Atlantic. Common in floating Sargassum weed and among seaweeds in shallow water (Holthuis). Palaemon (Palaemon) debilis Dana 1852. Dana, Proc. Ac. Nat. Sct. Philad., vi, p. 26. 1950. Holthuis, Szboga Exp. monogr., xxxix, a. 9, p. 66, fig. 13 (references). Holthuis records this species from mangrove swamps at Durban. Also recorded from Chagos, Maldives, and Indo-Pacific. Subfam. Pontoniinae 1952. Holthuis, Szboga Exp. monogr., xxxix, a 10, pp. 1-254 (key to genera and subgenera). Holthuis regards Ancylocaris Schenkel 1902 as a synonym of Harpilius Dana 1852. The two species recorded by me (1950) as species of Harpilius should be included in Harpiliopsis Borradaile 1915. Gen. Periclimenes Costa Subgen. Periclimenes Periclimenes rex Kemp 1922. Kemp, Rec. Ind. Mus., xxiv, p. 158, text-fig. 25, and pl. 5, fig. 5. An ovig. 9 24 mm. in length agrees with Kemp’s description except in two features, probably varietal or sexual. The upper edge of the rostrum has a series of 28 teeth, excluding the apical point, and the two teeth immediately preceding the latter are smaller than the others; the posterior 3 teeth are postorbital, but there is no isolated tooth on the carapace. The outer margin of the external lobe of the basal joint of ant. 1 ends in an acute point (as in Kemp’s fig. 255), but the distal margin projects beyond this point and ends in one (left) or two (right) sharp points (somewhat similar to the right ant. 1 in fig. 39¢ of Paranchistus ornatus Holthuis 1952). Only one (detached) leg of the second pair is present. Kemp records the gorgeous coloration of this species; as preserved, the present specimen is uniform deep orange, with a hint of a pale dorsal patch on 48 ANNALS OF THE SOUTH AFRICAN MUSEUM the carapace, and the chelae of the first pair of legs and of the one remaining leg of the second pair still retain a purple tinge (as described by Kemp). Locality: Delagoa Bay. (U.W.) Distribution. Andaman Islands. Subgen. Harpilius Key to the South African species I. Supraorbital spine present. 4th joint of 2nd leg with spine-tooth on lower apex. a. Spine of antennal scale extending well beyond apex of scale. oe joint of and leg shorter than palm (9). : . grandis b. Spine of antennal scale not, or scarcely, extending beyond apex of scale. 5th joint of 2nd leg subequal to (9), longer than ($) palm. . demani 2. Supraorbital spine absent. 4th joint of and leg without apical spine- tooth. 5th joint of 2nd leg shorter than palm. - « +. 6 breutcarpales Pertclimenes (Harpilius) grandis (Stimpson) 1922. Kemp, loc. cit., p. 210, text-figs. 58, 59, and pl. 7, fig. 10. 1952. Holthuis, loc. cit., pp. 11, 79 (references). Locality: Delagoa Bay, 1 3, 1 ovig. 9. (U.W.) Distribution. Japan, East Indies, India, Ceylon, Red Sea, Zanzibar. Periclimenes (Harpilius) demani Kemp 1915. Kemp, Mem. Ind. Mus., v, p. 279, pl. 13, fig. 10, and text-figs. 27 a-1. 1922. id., Rec. Ind. Mus., xxiv, pp. 171 (in key), 2109, fig. 64 (antennal scale). 1952. Holthuis, loc. cit., pp. 11, 83. Using Kemp’s 1922 key, these specimens are easily identified as this species; and they agree with his 1915 detailed description. The rostral formula varies: 8-9 teeth above, 2-3 below. Length 20 mm. Localities. Durban Bay. 34, ovig. 92 July 1950. 1 ovig. 9, Jan. 1951; and Inhambane. 34, ovig. 99, and immature, Jan. 1954. (U.C.T.) Distribution. Chilka Lake; Madras; Mergui Archipelago. Inhabits both sea and brackish water. Fam. CRANGONIDAE Pontophilus pilosus Kemp 1916. Kemp, Rec. Ind. Mus., xii, p. 367, text-fig. 4, and pl. 8, fig. 4. Of 6 specimens the smallest (9g mm.) agrees with Kemp’s description, but the others (12-13 mm.) have 4 medio-dorsal spines on the carapace instead of 3. The basal process of ant. 1 and other characters agree with the description. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 49 Mottled and banded with brown as described by Kemp. Locality: Delagoa Bay, 6 specimens, incl. ovig. 99, from among Cymodocea plants. (U.W.) Distribution. Southern India. STOMATOPODA Fam. SQUILLIDAE Squilla harpax de Haan 1849. de Haan, Fauna Japon. Crust., vii, p. 222, pl. 51, fig. 1. 1913. Kemp, Mem. Ind. Mus., iv, p. 88, pl. 7, fig. 77 (raphidea, part) 1950. Barnard, Ann. S. Afr. Mus., xxxvill, p. 851, fig. 1c, g (raphidea, non Fabr.). 1952. ‘Tiswari and Biswas, Rec. Ind. Mus., xlix, p. 358, fig. 3b, d, f- The last-mentioned authors, Holthuis concurring, have separated harpax from raphidea, and given the differential characters. S. harpax appears to be a smaller species, not exceeding 200 mm. in length. The cornea is even wider proportionately to the length of the whole eye than in raphidea; the lateral margin of the carapace is concave from the antero- lateral spines to the angular projection in hinder third (sinuous in raphidea). The 5th thoracic segment is laterally obtuse (not armed with a spine); the submedian keels on the thoracic segments and first five abdominal segments are obsolete. The intermediate keels on the thoracic segments do not end in spines. The keels supporting the marginal teeth on the telson are not massive or inflated. The 5th thoracic segment is the crucial character. The single specimen from Durban in the South African Museum, measuring 195 mm. in length, shows the above diagnostic characters. Distribution. Japan, Hong Kong, Singapore, Andaman Islands, Bay of Bengal, Bombay. TANAIDACEA Fam. ‘TANAIDAE Leptochelia mirabilis Stebb. ice 23% These specimens are identified with mirabilis solely on geographical reasons, Ceylon being the nearest locality from which these ‘dolichochelous’ Tanaids have been recorded. The other species to be considered are: minuta Dana 1853 (loc. ?), rapax Harger 1879 (east coast of N. America and Bermuda) and foresti (Stebb.) 1896 (West Indies). Stebbing (1900, Willey’s Zool. Res., pt. 5, p- 616) suggests that Paratanais erythraea Kossm. 1880 (Red Sea) may be synonymous with minuta. These species have been distinguished by minute differences in the relative lengths of the joints, and the number of flagellar joints in 1st and 2nd antennae. 50 ANNALS OF THE SOUTH AFRICAN MUSEUM A one-jointed outer ramus of the uropod is said to distinguish rapax (see: Richardson, 1905, Bull. U.S. Nat. Mus., no. 54, p. 23 in key), but Stebbing (1905, loc. cit., p. 6) states that even the same species may have a one- or two- jointed outer ramus. Stebbing’s single specimen of mirabilis was a § with the 1st gnathopods not quite symmetrical, and both of them damaged; he was therefore unable to describe the apex of the thumb. In the present specimens the 1st gnathopods are symmetrical in all the gg, and the thumb is similar to those of rapax and forestt (see: Richardson, loc. cit., figs. 31 and 25). Length, g 3 mm., Ist gnathopod 5-6 mm.; 2 2-5 mm. As preserved, white, eyes black. Locality. Inhambane, in Zostera beds. 34, ovig. 99. (UEC) Distribution (mirabilis). Ceylon. ISOPODA Fam. ANTHURIDAE Gen. Haliophasma Hasw. 1940. Barnard, Ann. S. Afr. Mus., xxxii, pp. 382, 498 (key to species). The following key seems preferable to that given in the above work on p. 408. | 1. Telson clearly distinct from pleon segment 6. a. Telson narrow-ovate, strongly tricarinate. .. . . éricarinata b. Telson broadly-ovate. i. Telson widest proximally, with basal lateral flanges, more or less distinctly tricarinate. ... . foveolata Fie: ate ii. Telson widest in middle, not tricarinate. . . . . coronicauda Leptochelia mirabilis ae : 5 Stebb. Finger and iii. Telson obovate, apically upturned. .. . « hermani thumb of 1st peraeo- 2. Telson fused with pleon segment 6. . . . . pseudocarinata pod g. Halwophasma foveolata Brurd. Fig. 24a-c. 1940. Barnard, loc. cit., p. 384, fig. 2. My suggestion that this species might be the female of éricarinata is put entirely out of court by the present fine male specimen. Foveolae and pits present on dorsal surface of head and peraeon segments 1-3 (only a few on 3), demarcated lateral areas on segments 2—5 (one or two on segments 6 and 7), and strongly marked on sternites of segments 2—7. Antenna 1, flagellum 12-jointed (the apical one minute), with dense fascicles of setae. Antenna 2, flagellum 7-jointed, with fascicles of shorter and stiffer setae (not ‘furry’ as in antenna 1). ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 51 Peraeopod 1, 5th joint with lower apex blunt, 6th joint cylindrical but curved, lower margin concave from base to a triangular tooth near finger hinge. B Pleopod 2 J stylet inserted slightly beyond middle of inner ramus, extending slightly beyond apex of ramus, slightly expanded distally, apex acute. Telson as in the original figure: basal half with marginal flanges, distinctly narrowed in distal half; lateral ridges present but no trace of the central ridge; ventral surface as a whole concave, but with a slight median convexity. Length 26 mm. wekocalty. False Bay. 1 g. (U.C.T.) a G f af FIG. 24. Haliophasma foveolata Brnrd. a. peraeopod 1 g. b. transverse section of telson. c. pleopod 2 g stylet. Paranthura latipes n. sp. d. peraeopod 5. e. telson and uropod. f. pleopod 2 ¢ stylet. Gen. Paranthura B. & W. 1925. Barnard, 7. Linn. Soc. Lond., xxxvi, p. 152. Paranthura latipes n. sp. Fig. 24d-f. Telson ovate-lanceolate, lateral margin slightly sinuous, greatest width a little before midway, margin and dorsal surface with long, minutely plumose setae. Uropod, peduncle with a transverse series of long setae at distal inner corner, ramus longer than basal width, outer ramus broadly ovate, margins with long, minutely plumose setae. Antenna 1, flagellum 6-jointed, setose. Peraeopod 1 as in punctata (1914, Barnard, Ann. S. Afr. Mus., x, pl. 29 C) but basal tooth on palm not prominent. Peraeopods 4-7 with and and grd 2 ANNALS OF THE SOUTH AFRICAN MUSEUM joints broadly oval. Pleopod 2 ¢ stylet a little longer than inner ramus, slightly constricted at two-thirds the length, the distal third slightly lanceolate, apex acute. Length, ¢ 7-5 mm. As preserved, yellowish-grey; head, a medio-dorsal line and the hind margins of peraeon segments, pleon segments laterally, telson, peduncle and outer ramus of uropods darker grey, eyes black, antennae and peraeopods white. Locality. Maxixe (Machiche), Inhambane Bay. Low tide, sand banks and concrete piles. (U.C.T.) Remarks. Distinguished by the broadly oval 2nd and grd joints of peraeopods 4-7, and the densely setose telson. In my references to punctata (loc. cit., p. 348a) the inclusion of Hilgendorf’s opinion (1878, MB. Berl. Ak. Wiss., p. 847) that punctata was a Paranthura, might be construed as a record of Stimpson’s species from Ibo (Mozambique). Hilgendorf merely recorded an ‘Anthura sp.’ without any description. Fam. EuRYDICIDAE (CIROLANIDAE) Cirolana imposita n. sp. Fig. 25¢, d. g—Head, peraeon and pleon smooth, but, when viewed semidried under a x 16 lens, sparsely sprinkled with minute granules. Frontal lamina pentagonal, separate (but barely) from rostral point. Telson with 2-3 granules at base on either side of a broad median ridge which extends to apex; this ridge with a median groove, the edges of which are approximate basally but divergent in the distal two-thirds, and enclose a subsidiary median ridge; lateral margins distally and apex with plumose setae (spines may have been present but are now broken off). Pleopod 1, outer ramus twice as broad as inner ramus. Uropod rami extending slightly beyond apex of telson; outer ramus with 2 groups of a spine and setae near the outer margin in distal two-thirds, inner margin and apex with plumose setae; inner ramus with spines (6 or 7) and plumose setae. Penial processes on 7th sternite as in sulcata (Barnard, 1939, Ann. S. Afr. Mus., XXxll, p. 392, fig. 6f). Stylet on pleopod 2 extending slightly beyond apex of ramus, slender, slightly curved outwards distally, apex acute. Length 9-10 mm. Localities. 1 § from stomach of cormorant shot off Danger Point, 1954. (R. W. Rand, Biologist to the Guano Islands Administration); 1 g, 1 9 off Cove Rock (East London), 80 fms. (s.s. Pzeter Faure coll. in S. Afr. Mus.). Remarks. In regard to the sculpturing of the telson, this species is allied to sulcata Hansen 1890, but invites closer comparison with cingulata Brnrd., 1920. The telson of sulcata has a median groove between two ridges, open from close to the base of the telson, without any indication of a median ridge (“imposita’) (fig. 25a). ee oe ie an ~~ ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 53 The granules on the pleon segments are usually distinct, but are usually obsolete on the ridges bounding the telsonic furrow. I have seen only one specimen as strongly granulate as in Hansen’s figure, which shows in addition a median line of granules in the furrow. C. sulcaticauda Stebb. 1904 has the tuberculate ridges nearly parallel, and also a median row of tubercles in the furrow. In other features sulcaticauda and sulcata are quite distinct. FIG. 25. Cirolana sulcata Hansen. a. telson Cirolana palifrons Brnrd. 6. malformed telson of the Type. Cirolana imposita n. sp. c. telson and right uropod. d. frontal lamina. Cirolana cingulata Brnrd. e. telson and right uropod (marginal plumose setae omitted). My original description of the telson of cingulata (‘a median ridge with two punctate-striate grooves’) is correct; but for comparison with the present species would be better worded: a median ridge with a groove which is slit-like in basal third, but widens in distal two-thirds, where it is nearly completely filled with a median somewhat granular ridge (fig. 25¢). Another species with a similarly sculptured telson is palifrons Brnrd. 1920. Although the sculpture had been disrupted by injury in the only known 54 ANNALS OF THE SOUTH AFRICAN MUSEUM specimen, the normal sculpture seems to have been very similar to that of cingulata (fig. 255). ‘The two species are easily distinguished by other characters. — Cirolana pilula n. sp. Fig. 26a-c. Body smooth, very convex. Frontal lamina and peraeopods as in hirtipes. Eyes round. Impressed line on hind margin of eyes continuous across median area of head. Telson broader than long, length about ? basal width, broadly rounded, distal margin minutely crenulate, with a plumose short seta in each notch. Uropods not reaching telsonic apex, inner ramus broad, distal margin with plumose setae and 6—7 spines, outer ramus shorter than inner, ovate, apex with plumose setae and 5-6 spines. Vasa deferentia opening by pores flush with sternal surface (as in hirtzpes). Stylet on pleopod 2 ¢ slender, straight but slightly curved distally, apex serrulate, with a minute digitiform process. Length ca. 11 mm., breadth 4-5 mm. As preserved, pale with pinkish or greyish speckles. Locality. False Bay (94° 17 6" S. 18° 399 2" E. and 34° 12 6” Sate sa@e, ene 25-34 fathoms, shelly sand. 3 specimens, including ¢ but no ovig. 9. (U.C.T.) Remarks. ‘Vhis species appears to be able to ‘conglobate’ almost as com- pletely as the terrestrial Tylos, and in fact bears a strong resemblance to this littoral Isopod. It belongs to the hirtzpes group, but is at once distinguished by the short, rounded telson. Cirolana capiiella n. sp. Fig. 26d-g. A fringe of setae on hind margins of the peraeon segments laterally and on side-plates, smaller specimens not so setose as the largest (in fact almost glabrous); pleurae of pleon segments sparsely setose; pleon and telson with very fine pile. Head with an upstanding median process between the eyes, apically slightly bifid, and near the inner margin of eye a very low rounded tubercle; median process not developed in juveniles 7 mm. in length, but indicated in specimens Q-11 mm. in length, in which latter also the tubercles can just be seen if specimen examined semi-dry. Frontal lamina obovate, anterior margin rounded, freely projecting, separated from small rostral point by the contiguous bases of Ist antennae. Hind margins of peraeon segments each with a series of small sharp denticles, 5 or 6 on either side of a median one, feeble and obscure on Ist and 2nd seg- ments, but becoming more prominent on hinder segments. Pleon segments 3 and 4 each with 3 or 4 denticles on either side of a median one; in the largest specimen the denticles on segment 4 stronger than on segment ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 55 3, and the submedian pair stronger than the others; segment 5 with a median and 2 (or 3) lateral denticles. ‘Telson with an obscure median groove bounded by very feeble ridges; this sculpture can only be seen when specimen is semi- dry; a strong acute tooth laterally over the articulation of uropod; apex narrowly rounded, with plumose setae and about 6 spines. Uropod, inner ramus broad, outer ramus ovate, of about equal length, margins with plumose setae and spines. Vasa deferentia opening by pores flush with surface. Fic. 26. Cirolana pilula n. sp. a. pleon segments 4 and 5, telson and uropod. Bb. frontal lamina. c. pleopod 2 § with apex of stylet further enlarged. Cirolana capitella nn. sp. d. lateral and dorsal views of head. e. frontal lamina. f. peraeon segment 7, pleon, telson and uropods. g. lateral view of pleon segments 3~—5, basal corner of telson, and peduncle of uropod. Pleopod 1, inner ramus % as broad as outer ramus. Pleopod 2 ¢ stylet as long as ramus, slender, apically acute. Largest specimen 12 X 4°5 mm. Preserved, uniform pale yellowish, eyes black. 56 ANNALS OF THE SOUTH AFRICAN MUSEUM Locality. Inhambane, and in the Lagoa Poelela at Inharrime. 1 adult 3, 3 immature gd, 7 juv. (U.G.T. 1954) Remarks. Closely resembling willey: Stebb. (1904, Spol. eylan., II, p. 11, pl. 3) but pleon segment 3 not overlapping segment 4, and frontal lamina not pentagonal with anterior point meeting rostral point (see: Barnard, 1935, Rec. Ind. Mus., xxxvii, pp. 309, 312, fig. 185). In this last respect the present species is comparable with fluviatilis (see: Barnard, loc. cit., p. 310, fig. 19). Gen. Conilorpheus Stebb. The coinciding of three characters, viz. transverse impressed lines on the peraeon segments, a particular kind of sculpture on the telson, and a narrow inner ramus on pleopod 1, in Conilorpheus scutifrons and Cirolana cingulata led to a check-up on all the South African species of Cirolana. A narrow inner ramus on pleopod 1 was recorded in the original descriptions of cingulata and palifrons, and is found in imposita n. sp. (supra), but the trans- ference of these three species to Conilorpheus is precluded by the frontal lamina being separate from the rostral point, and the bases of 1st antennae contiguous. In C. scutifrons the bases of 1st antennae are definitely separated by the junction of the rostral point and the strongly produced frontal lamina. A narrow inner ramus on pleopod 1 is also found in C. sulcata. But in this species—with the same basic type of telsonic sculpture in its simplest form— the contiguity of the bases of 1st antennae is an ambiguous character: in some specimens the bases are sufficiently bulbous to meet over the lower-lying septum joining the rostral point and the frontal lamina, in others the rostral point and the frontal lamina meet, or almost meet, on the surface. One cannot go farther than say: there is a group of four species of Czrolana with closely comparable telsonic sculpture: sulcata, palifrons, imposita and cingulata, the last of which shows a truly remarkable convergence with Conilorpheus scutifrons. In the course of examining the Crrolana species one or two features were noted suggestive of a possible division of the rather numerous species into groups. It would, however, be premature to attempt this. Conilorpheus scutifrons Stebb. Fig. 27a-—d. 1908. Stebbing, Ann. S. Afr. Mus., vi, p. 46, pl. 5 (Crustacea pl. 31). 1920. Barnard, ibid., xvii, pp. 351, 352. An unfortunate error occurred in the latter description; the ¢ and 2 symbols were transposed. The first paragraph (p. 352) applies to the 9, the second to the g. As is often the case, the 3 is the more highly sculptured of the two sexes, but not to a very great extent. The sculpturing typical of the species is the same in both sexes. There are traces sometimes of feeble tubercles on peraeon segment 2, but usually one of the 3 or 4 transverse impressed lines is merely undulate or ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 57 crenulate; these crenulations are more marked on segments 3 and 4, and on segment 5 become definite denticles or tubercles. There are usually accessory denticles between the major tubercles on segments 5-7. On the pleon the median pair of tubercles on segments 3-5 are more promi- nent than any of the other tubercles, of which there appear to be constantly 2 on either side, and some minor denticles. For reasons given by Stebbing the full development of the telsonic sculpture was not present in his (type) specimen, and as no mention was made in my 6 f e FIG. 27. Conilorpheus scutifrons Stebb. a. peraeon segment 7, pleon, telson and uropods (marginal spines and setae on telson and uropods omitted). b. peraeopod 1. c. ¢ penial processes on peraeon segment 7. d. pleopod 2 ¢. Conilorpheus blandus n. sp. e. pleon, telson and uropods. /f. peraeopod 1. 58 ANNALS OF THE SOUTH AFRICAN MUSEUM note, a description and figure is given here. As will be seen from the figure the sculpturing is distinctly reminiscent of that of Cirolana cingulata (fig. 25e), which species also has transverse impressed lines on the peraeon segments. The median ridge is overlapped at the base by two adnate lobules, there is a slight ridge (sometimes obscure) in the middle of the central groove, each margin of the latter being ornamented with 4 little tubercles, often apically bifid. There are several other little denticles and points on the telson, as also on the inner ramus of uropod. Minor variations in arrangement occur among the smaller denticles. Peraeopod 1 as described by Stebbing. The 4th joint has 4 (sometimes 5) bluntly rounded teeth as shown in his figure (not ‘spines’ as in his description) ; 5th and 6th joints each with 2 spine-setae on inner apex; inner margin of 7th joint crenulate, ending in a ‘squared’ or rounded lobe at base of unguis. Penial processes on 7th peraeon sternite separated by about twice the basal width of one of them; apices curving inwards. Stylet on pleopod 2 in fully adult § arising at basal quarter of inner ramus, not midway as in the ‘subadult’; uniform in width, distally curved outwards, apex rounded. One @ has large brood lamellae on peraeopods 2-5. Only one embryo, 2-5 mm. in length, remains in the pouch. Length ¢ 9 mm., 2 10 mm. As preserved, the Algoa Bay specimen has greyish irregular mottling on the peraeon; the other specimens are yellowish, eyes black. Additional localities. East London, littoral. 1 3, 1 9 (Prof. Stephenson, 1937) ; False Bay and Algoa Bay. (U.C.T.) Conilorpheus blandus n. sp. Fig. 27¢, f. ©—Head and body covered with very fine and short pile, which becomes thicker on pleon and telson; when semi-dried the head and peraeon are seen to be obscurely rugulose. Postero-inferior corners of side-plates 1-4 with a few plumose setae; whole inferior margins of 5—7 thickly set with plumose setae, as are also the pleural margins of the (exposed) pleon segments. Frontal lamina as in scutifrons, meeting the rostral point and separating the bases of ist antennae, tridentate, with an accessory denticle on each side proximally. Flagellum of 1st antenna 5~-7-jointed, of and antenna 9—10-jointed. Maxilla 1 inner lobe with 3 strongly setose spines. Peraeopod 1, 4th joint with undulate inner margin, feebly crenulate proxi- mally, 5th joint with 4 blunt teeth increasing in size distally, 6th with 2 large blunt curved teeth, 7th with a strong curved tooth at base of unguis. Pleopod 1, inner ramus only half the width of outer ramus. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 59 Uropod, inner ramus obovate, outer distal corner extending slightly beyond telson apex, outer ramus ovate, not reaching level of telson apex, both rami with marginal spines and thick fringe of plumose setae. Telson apex more narrowly rounded than in scutifrons, with about 6 spines, and thick fringe of plumose setae. 8-8-5 mm. X 4mm. As preserved, yellowish, eyes black. Locality. Algoa Bay. 2 subadult 99. (U.C.T.) Remarks. Both specimens are subadult, preparing for ecdysis. The contrast between this smooth velvety species and the sculptured scutifrons, both with a tridentate frontal lamina, is very striking. Fam. COoRALLANIDAE Gen. Corallana Dana 1879. Schiddte and Meinert, Nat. Tidsskr. (3), xii, p. 286. 1904. Stebbing, Fauna Geogr. Mald. Lacc. Archip., ii, p. 703. 1914. Barnard, Ann. S. Afr. Mus., x, p. 358a. 1931. Nierstrasz, Siboga Exp. monogr., Xxxil, c., p. 163. Corallana furcilla n. sp. Fig. 28c-e. 2 with embryos—Body with scattered setules, chiefly on side-plates and hind margins of peraeon segments. Surface obscurely rugulose or foveolate, pleon and telson with irregularly scattered pits. Peraeon segments 1-4 without sculpture; segment 5 with 4-5 obscure flattened tubercles laterally; segment 6 with 3 tubercles, the submedian pair obscure; segment 7 with 5-6 tubercles, the lateral ones more prominent than the central ones. Pleon segment 2 smooth; segments 3 and 4 each with a median pair of tubercles and a single lateral one, upper margin of pleuron of segment 4 crenulate; segment 5 with a median pair of longitudinal tubercles. Telson with a very obscure transverse series of granules at base, outer basal corner upturned, carinate; surface slightly concave between the lateral margins but with a very slight central convexity, apex with a pair of short slightly upturned points, distal margin with plumose setae, a pair of spines in the sinus, and one spine on outer margin of each prong. Basal joints of 1st antennae contiguous, not anteriorly dilated, flagellum 13-jointed. Antenna 2 flagellum 23-jointed. Epistome transverse, vertical and not visible until upper lip is depressed. Peraeopods 1-3, inner apex of 3rd joint with 3 blunt tubercles, 5th with 4 blunt tubercles on inner margin; these tubercles do not lie in the plane of the limb, but curve backwards towards the posterior surface. Uropod, inner ramus extending almost to telson apex, obovate, apically truncate with 2 spines near inner corner, outer and apical margins with 60 ANNALS OF THE SOUTH AFRICAN MUSEUM plumose setae; outer ramus slightly longer than inner, narrow, ovate, articu- lating vertically, upper margin thick, both margins with plumose setae. 13 X 5 mm. As preserved, slaty-grey, somewhat mottled, eyes black. Embryo 3:3 mm. Telson with broadly rounded, subtruncate apex. Juvenile 8-5 mm. Peraeon and pleon without sculpture. Telson as in embryo, but apical margin with plumose setae and 6 spines; distal margin of Fic. 28. Lanocira latifrons Stebb. a. head. 6. telson and uropod. Corallana furcilla n. sp. c. peraeon segments 5-7, pleon, telson and uropods of adult 9. d. telson and uropods of juvenile from brood pouch, 3-3 mm. e. telson and uropods of juvenile 8-5 mm. ge a ee on et ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 61 inner ramus of uropod with plumose setae and 6 spines; outer ramus with 3 spines on outer and 1 on inner margin distally, apex with 1 stout spine and plumose setae. Locality. Inhambane, from holes in roots of mangroves. 1 2 with embryos, ago. (U.C.T.) Remarks. Similar to nodosa 8. & M. and /ursuta S. & M. as regards the tubercular sculpture of the hinder peraeon segments. The juvenile, presumably, is conspecific with the adult. The recurved tubercles or stout spines on the anterior peraeopods appear to be well adapted to excavating holes or burrows. Lanocira latifrons Stebb. Fig. 28a, 5. Tons otebbineg, 7. Linn. Soc. Lond. xxxi, yp). 217, pl. 21. g—Agreeing with Stebbing’s description except in two features: one obvious, the other rather obscure. The telson has not a ‘rather narrowly rounded apex’. The margins are convex, without any concavity, and the apex is broadly rounded. Moreover there are no marginal spines among the plumose setae, though marginal spines are present on both rami of uropods. The second feature is one which may not have been developed in Stebbing’s smaller ¢ (7.75 mm.), or may have been overlooked. There are two pairs of tufts of setae on the head, one near inner hind margin of eye, the other near inner front margin of eye. Unless examined semi-dry these little tufts might easily be overlooked. Bases of 1st antennae contiguous, separating the pentagonal frontal lamina from the rostral point. Length 9:5 mm. As preserved, yellowish, eyes black. Locality: Inhaca Island, Delagoa Bay. 1 3. (U.W.) Distribution. Sudanese Red Sea. Remarks. In a large number of specimens the growth of the rostral process in the g could be followed. The suggestion seems not unreasonable that latufrons may be only the fully adult ¢ of the earlier species gardinert Stebb. 1904, and zeylanica Stebb. 1905. The latter was also found in the Sudanese Red Sea; and gardineri and zeylanica at Ceylon. One point deserves attention: Stebbing’s figure (pl. 51A) of gardinert seems to indicate that the frontal lamina separates the bases of the Ist antennae. Fam. SPHAEROMIDAE Sphaeroma annandale: Stebb. 1940. Barnard, Ann. S. Afr. Mus., xxxii, p. 405. A ¢ and an ovigerous 2 from Inhambane (U.C.T.) have the transverse ridges on the peraeon segments continuous from side to side, not broken up into 62 ANNALS OF THE SOUTH AFRICAN MUSEUM tubercles; there are 2 circular tubercles on pleon segment 4, very indistinct in g; in @ the left-hand one only of one of the normal 2 pairs of submedian tubercles on telson is present, the other pair and the median tubercle being obsolete; in ¢ there is a mere suspicion of one submedian pair of tubercles on telson. Gen. Isocladus Miers 1914. Barnard, Ann. S. Afr. Mus., x, p. 384. The two species here described are the first representatives of this genus to be recorded from South African waters. Fic. 209. Isocladus otion n. sp. a. 5th—7th peraeon segments and pleotelson with uropod ¢. 6. lateral view, 6. ¢. 5th—7th peraeon segments, pleotelson with uropod 9. d. lateral view 9. e. ventral view of apex of telson. f. epistome. The sharp denticle on either side of the median process in g is a distinctive character. The ridge formed by the bulbous junctions of the peraeon segments with their side-plates does not occur in tristensis (Leach) or integer (Heller), in both of which the transverse profile forms a perfectly even curve. It occurs, however, to a lesser degree in the Australian Zuzara venosa (Stebb.) of which I have seen specimens, and in the Cape <. furcifer Brnrd.; and appears to be | ‘ , 7! es 2 hl — | = ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 63 quite well marked in <. (Isocladus) excavata Baker judging by his figure (1910, Trans. Roy. Soc. S. Austr., xxxiv, pl. 24, fig. 4). ‘ Isocladus otion n. sp. Bis: 20: Body glabrous, inconspicuously rugulose (best seen when removed from liquid and semi-dry). Epistome without crest, almost in contact with rostral process. Hind margins of peraeon segments rather prominent. Antero-lateral corner of peraeon segments bulbous at junctions with epimera, the latter also bulbous at the junctions so as to form a rather prominent lateral ridge on each side of the peraeon. On 7th segment the epimeron stands up conspicuously, more so in g than in 9, like a little ear (hence the specific name). Seventh segment in § with a median process, its apex obscurely trifid, and flanked on each side (midway between median line and epimeron) by a sharp denticle; in 9 hind margin only slightly more convex than in preceding seg- ments, with a mere suggestion of a pair of granules corresponding with the sharp denticles in the g. Pleon smooth in g, but with a submedian pair of inconspicuous knobs. Telson in ¢ smooth, with bluntly pointed apex; in 2 more strongly convex proximally, with 2 submedian parallel ridges and the rugulosity becoming almost granular; a shallow ventral groove in both sexes. Uropods not extending to apex of telson in either sex, ovate, broader in J than in 9. Penial processes on 7th peraeon segment close together but bases not conti- guous. Stylet on pleopod 2 straight, longer than endopod, apex acute. Length, 6-5 mm. As preserved, chalky white. Locality. False Bay, 4-10 fathoms. 2 §¢, 1 ovig. 9. Sept. 1946; 2 immat. eee rebr. 1947. (U.C.T.) Remarks. ‘The apices of the uropods falling short of the telsonic apex may perhaps indicate that the only two males are not quite fully grown, although in other respects (separation of stylet on pleopod 2) they seem to be mature. Isocladus mimetes n. sp. Fig. 30. Body glabrous. Head and peraeon segment 1 granulate. Epistome as in olion, without crest, almost in contact with rostral process. Posterior half of “segments 2—6 rather conspicuously ridged in side view, each segment with 2 submedian small granules, segments 2-4 also with a less conspicuous median granule; additional intervening minute granules can be seen when examined semi-dry. Segment 7 in ¢ inconspicuously granulose, with median process, its apex forked, flanked on either side by a prominent conical denticle; in 9 similar to segment 6, with the 4 well-marked granules, and additional minute ones. 64 ANNALS OF THE SOUTH AFRICAN MUSEUM Junctions of peraeon segments with their side-plates bulbous, forming a rather prominent lateral ridge. Side-plates vertical, scarcely visible in dorsal view; side-plate on 7th segment not so prominent as in otzon. Pleon in $ granulose, but without definite granules, in 9 distinctly granulate, and with a submedian pair of short ridges or longitudinally oval tubercles. Telson in g obscurely granulose, apex slightly raised, with a rather wide open notch; seen from behind the ventral groove is rather deep; in @ distinctly granulate with 2 submedian parallel ridges, apex bluntly pointed, dorsally raised, scarcely carinate but with a few granules; the profile of telson but not that of pleon closely resembles that of Zuzara furcifer (fig. 31¢), in hind view the ventral groove not so deep as in 3. 7B FIG. 30. Isocladus mimetes n. sp. a. 5th—7th peraeon segments and pleotelson with uropod ¢. 6. lateral view. c¢, d. posterior and ventral views of apex of telson. Uropods in ¢ extending beyond telsonic apex, inner ramus rather narrow, outer margin in distal } nearly straight, outer ramus subtriangular, basally rather broad, distal half of outer margin nearly straight and sloping to the narrowly pointed apex; in 9 not extending to telsonic apex, ovate. Penial processes on 7th peraeon segment close together but bases not contiguous. Stylet on pleopod 2 ¢ straight, longer than endopod, apex acute. Length, 5 mm. (the g appears larger than 2 on account of the projecting uropods). As preserved, amber-coloured, with minute pigment specks, uropods paler. Locality. Off Roman Rock, False Bay, 12-14 metres. 3 gg. 19. (U.C.T. 1954.) ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 65 Remarks. The forked process on peraeon segment 7 in the ¢ gives a remark- ably deceptive likeness to Zuzara furcifer. ‘The absence of the median process on telsonic apex, and the shape of the uropods, however, at once distinguish the present species. The @ is distinguished from that of otion by the four longitudinal series of granules on the peraeon, and by the tubercles on the pleon being more longi- tudinally oval than circular. The telson is more rugged and in this respect resembles that of <. furcifer (infra). These three species were found at slightly different sites, but in each case both sexes were found in association; otherwise the correlation of the sexes would have been difficult if not impossible. Kuzara furcifer Brnrd. Bigs Sir 1920. Barnard, Ann. S. Afr. Mus., xvii, p. 361, pl. 15, figs. 26, 27 (3). Adult g. On 6th peraeon there is a feeble tubercle about midway between the median line and the side-plate. This was overlooked in the original description and figure. Fic. 31. Kuzara furcifer Brnrd. a. dorsal view of 6th and 7th peraeon segments and pleotelson with uropods of immature ¢ 4:5 mm. 6. the same of 2 5°5 mm. . lateral view 9. Immature 3, 4°5-5 mm. in length: 7th peraeon segment with a submedian tubercle on either side of the short, quadrangular median process, which does not extend beyond the hind margin of pleon segment 4; the second tubercle clearly seen in the adult, is very faint. Telson relatively a little shorter than in adult, apex acute, subapical margin slightly concave; small tubercles or granules as in adult. Rami of uropods not reaching telsonic apex. ®, 5°5-6 mm.: more strongly granulose than 4; no trace of tubercles on 6th peraeon segment; 7th peraeon segment with an obscure tubercle about midway between median line and side-plate; pleon segment 4 with a small median 66 ANNALS OF THE SOUTH AFRICAN MUSEUM granule, flanked by a larger tubercle. Telson shield-shaped, apex acute, margins evenly convex; a submedian pair of elongate tubercles followed by a pair of round tubercles, apex upturned and bluntly keeled, with 2 granules. Rami of uropods not reaching telsonic apex. Additional localities. Keurbooms River estuary (K.H.B. 1931. 1 39); Klein- mond (Caledon Division) (K.H.B. 1 ¢); False Bay, 7 fathoms (U.C.T. g, Pe; Juv.). Gen. Sphaeramene Brnrd. 1914. Barnard, Ann. S. Afr. Mus., x, p. 405. Sphaeramene microtylotos n. sp. Fig. 32a-e. In general very similar to polytylotos Brnrd. 1914, but the larger tubercles, especially the submedian ones on the pleo-telson, are conical instead of button- like, and the smaller ones are rounded granules (cf. fig. 32f). Telson with 3 pairs of submedian conical tubercles, increasing in size and diverging posteriorly; apex bluntly pointed, with a rounded tubercle dorsally, and a groove ventrally. Epistome pentagonal, the bluntly pointed apex meeting the rostral point. Antenna 1, Ist joint L-shaped, with prominent conical boss on anterior margin. Uropods extending to telsonic apex, inner ramus oblong, margins sub- parallel, with a median series of 6—7 granules. Distal margin truncate, slightly emarginate; outer ramus ovate, apex turned outwards, outer distal margin feebly serrulate. Length 6-7 mm. As preserved, pinkish. Localities. False Bay, from stomach of cormorant. 3 specimens (R. W. Rand, Biologist to Guano Islands Administration, 1954); False Bay, 14-17 metres. 1 ovig. 9. (U.C.T. 1954.) Remarks. The great likeness to polytylotos obviously suggests inclusion in the same genus, although the characters of the J, when discovered, may necessitate the removal of the new species to another genus. The 3 specimens are fairly well preserved in spite of their having been taken from the stomach of a cormorant, in all probability having been first swallowed by a fish. The dorsal integument of two of the specimens has not suffered from the gastric juices, and the pleopods are intact. All three specimens appear to be immature, because neither brood lamellae, penes, nor ¢ stylets on pleopod 2 could be found. The tubercles and granules are not quite symmetrically arranged, and differ slightly in the 3 specimens. The ovig. 2 was received later, and confirms the description based on the three specimens. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 67 Cymodoce zanzibarensis Stebb. 1910. Stebbing, Trans. Linn. Soc. Lond., zool., xiv, p. 105, pl. 9, fig. D. g—Agreeing with Stebbing’s description. It is a little tempting to make this species a synonym of the earlier bicarinata Stebb., described from Minikoi, Ceylon and Zanzibar; but the fact that Stebbing himself (1910, loc. cit., pp. 105, 106) has recorded both species from Zanzibar precludes this. More- over zanzibarensis is a larger species, less setose, especially anteriorly, and the stylet on pleopod 2 ¢ is straight and relatively shorter than the long, inward curving stylet in bicarinata (see: Stebbing, 1904, Fauna Geogr. Mald. Lacc. Arch., ip: 712, pl. 52, fig. B). BiG 3o" Sphaeramene microtylotos n. sp. a. head with peduncles of tst antennae. 6. 7th peraeon segment, pleotelson with uropods. c. posterior view of apex of telson. d. dorsal view of first two joints of peduncle of Ist antenna. e. epistome with bases of Ist antennae. Sphaeramene polytylotos Brnrd. f. head for comparison with a. 11-12 X 6 mm. As preserved, pinkish-brown, buff, or greyish, the smooth telsonic boss rose-pink or pale orange. In Stebbing’s specimens the rose- coloured boss on the telson had retained its colour for seven or eight years. Q—At one station 5 ovig. 99 were taken in the same haul as 5 gg, and may be presumed to belong to this species. 9-10 mm. Telson with 2 small sub- median tubercles or low rounded bosses, each surmounted by a granule; the median lobe of the trilobate apex bears no dorsal tubercle or boss, but is slightly gibbous, and more densely setose than the surrounding area. Rami of -uropods of equal length, extending to level of telsonic apex. 68 ANNALS OF THE SOUTH AFRICAN MUSEUM On another occasion 1 Q was taken with ¢¢ of both C. zanzibarensis and Paracilicaea clavus; which raises a little doubt as to whose wives these females really are. Locality. Inhambane. 9 gg. (U.G.T.) Distribution. Zanzibar. Cymodoce natalensis Brnrd. 1920. Barnard, Ann. S. Afr. Mus., xvii, p. 366, pl. 16, figs. 1, 2 (japonica Rich. var. n.). Additional locality. Algoa Bay. 2 gg. (U.C.T.) These two specimens correspond exactly with the Natal specimens, which were formerly considered as a variety of japonica. I now think this form should be given specific status. The 3 pairs of tubercles (1 on 4th pleon segment and 2 on telson) are all very nearly of the same size; in fact the anterior pair on the telson are usually larger than the posterior pair; and the closely fitting opposed surfaces of the former pair and the pair on the 4th pleon segment is distinctive; neither Richardson’s nor Thienemann’s figures bear the least resemblance to this arrangement. Cymodoce cryptodoma Brnrd. Fig. 33d. 1920. Barnard, loc. cit., p. 368, pl. 16, figs. 6, 7. Additional locality. Algoa Bay. 1 4, 1 ovig. 9. (U.C.T.) The present ¢ has the 4th pleon segment more prominent than in the type 3; it might almost be said to have a submedian pair of low bosses. A lateral view is given here, as the shape of the telsonic projections is not fully indicated in a dorsal view. On the outer side of each ridge is a shallow submarginal hollow. Cymodoce lis n. sp. Fiss3oan0: g—Head, peraeon and pleo-telson perfectly smooth and glabrous, obscurely and sparsely punctate (seen when examined semi-dry). Telson apically broadly rounded. Antenna 1, basal peduncular joint entire (not dentate). Epistome broadly /\-shaped. Penial processes on 7th peraeon segment close together, elongate, slender and tapering to fine points. Pleopod 2 ¢ stylet tapering to a fine point, extending slightly beyond apex of ramus. Uropod, inner ramus obovate, apically truncate, inner distal angle extending to level of telsonic apex, outer ramus much longer, narrow ovate, apex narrowly rounded. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 69 11 X 6mm. As preserved, creamy whitish, eyes black. Locality. Algoa Bay. 1 g. (U.C.T.) Remarks. A species with the telson of an Exosphaeroma (cf. truncatitelson) combined with the broad, arrowhead-like epistome and the long slender penial processes of a Cymodoce. ir a Fic. 33. Cymodoce lis n. sp. a. peraeon segment 7, pleotelson and uropods. 6b. epistome. Cymodoce alisn. sp. c. pleotelson and uropods. Cymodoce cryptodoma Brnrd. d. lateral view of peraeon segment 7, pleotelson and uropod. Cymodoce alts n. sp. Big 39C. g—Setulose, becoming densely setose on pleo-telson and uropods. Peraeon segments 2—7 each with a transverse series of granules on the thickened hind margin, becoming more conspicuous posteriorly, almost tubercles on segments 6 and 7. Pleon, some small granules on pleura of segment 3, segment 4 with a submedian pair of large conical tubercles on hind margin overlapping base of 70 ANNALS OF THE SOUTH AFRICAN MUSEUM telson, splayed outwards, their inner margins more or less crenulate, flanked externally by 2 small granules; on anterior portion of segment a series of 4-6 granules. Telson basally with a transverse (curved) row of 4 tubercles, centrally with 2 widely separated apically bifid tubercles; laterally a (longitudinal) series of 3 tubercles, the hindmost one apically bifid; apex trilobate, median lobe with a submedian pair of small setiferous tubercles and a larger conical median one, apex truncate, with a small medio-dorsal conical tubercle; lateral lobes apically bifid; minute granules basally, centrally and apically. A strong fringe of plumose setae on pleura of segment 4 and on the submedian tubercles. The bifid tubercles on telson setiferous, and the apical lobes with stiff golden-brown setae. Penial processes on 7th peraeon segment close together, slender. Pleopod 2 3 stylet 14 times as long as ramus, moderately stout, tapering to an acute apex. Uropods extending beyond telsonic apex, inner ramus oblong, apex obliquely truncate, outer apical corner with an acute point, outer ramus longer, narrow ovate, apex with acute point; peduncle and rami with stiffish brown setae on upper surfaces, margins with long plumose setae. First peduncular joint of antenna 1 not denticulate. 9-9°5 X 4-4:25 mm. As preserved, dull yellowish-grey, eyes black. Locality. Algoa Bay. 3 gg. (U.C.T.) Paracilicaea mossambicus Brnrd. 1914. Barnard, Ann. S. Afr. Mus., x, p. 397, pl. 34, fig. D. g—Agreeing with the type, but: peraeon segments 2—7 each with 2 trans- verse rows of granules on hind margin, feeble on segment 2, becoming stronger posteriorly, but not conspicuous. Each boss on telson with 2 denticles more or less concealed among setae. ©—Whole body setulose, especially pleon segment 4 and telson. Telson with 2 low bosses, each surmounted by an inconspicuous glabrous point; no depression between the bosses at base of telson; apex trilobate, the lobes broad and separated by short clefts. Rami of uropods not extending beyond telsonic apex, subequal, ovate, apices blunt. Length ¢ 9 mm., 2 8 mm. As preserved, ¢ amber-coloured, the telsonic bosses and the subcircular basal depression between them castaneous; 9 pale greyish-cream; eyes black. Locality. Inhaca Island, Delagoa Bay. 1 4, 2 ovig. 99. (U.W.) Paracilicaea teretron n. sp. Fig. 34a, 0. Setulose. Head, peraeon and pleon without tubercles or granules; in a few specimens there is a hint of a submedian pair of granules on hind margin of 4th pleon segment. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 71 Telson in g with a submedian pair of small tubercles in centre ; In some specimens these granules are the end points on very feeble and obscure ridges; apex trilobate, median lobe not bulbous dorsally; in 9 evenly convex, without any tubercles, apex not so deeply trifid as in @. Uropod, inner ramus extending to level of telsonic apex in 4, slightly beyond in 9, oblong, obliquely truncate, outer apex acute; outer ramus in ¢ twice as long as inner ramus, terete, outer margin gently convex, tapering to an acute apex; in @ a little longer than inner ramus, ovate, apex acute. FIG: 34. Paracilicaea teretron n. sp. a, 6. pleon, telson and uropods, ¢ and 9. Paracilicaea clavus n. sp. c. pleon, telson and uropods 3. Penial processes on 7th peraeon segment close together, elongate, slender, tapering to acute points. Pleopod 2 ¢ stylet one third longer than ramus, scarcely tapering, apex blunt. Length g 7-5 mm., 2 6 mm., breadth 3 mm. As preserved, slaty-grey or buff coloured. Locality. Inhambane. gg, 99. (U.C.T.) Remarks. Very close to mossambicus Brnrd. (1914, loc. cit., supra) but differing by the simpler sculpturing on the telson, and the narrower, terete outer rami of the uropods. : There can be little doubt that the present ¢g and 99 are conspecific, as they were found on several occasions, and always together. The 2 of hansent Stebb. 1910 (Zanzibar) is unknown. “Cymodoce’ hamaia Baker (1908, Trans. Roy. Soc. S. Austr., XXxli, p. 141, pl. 4, figs. 1-11) appears to belong to Paracilicaea, but Baker has described the 9 as being very similar to the ¢ both as regards the telson and the outer rami of uropods. A small g, 6 mm., with penial processes and pleopod 2 stylet already developed, appears to belong to this species, but the telsonic apex is not so deeply incised, and the inner margin of outer ramus of uropod is gently convex, consequently the ramus is not so narrow and terete. Delagoa Bay (U.W.). 72 ANNALS OF THE SOUTH AFRICAN MUSEUM Paracilicaea clavus n. sp. LS BVI g—Smooth; when semi-dried the whole surface is seen to be shallowly and rather coarsely pitted; glabrous except for a few setae on the margins of the side-plates, pleura, telson and uropods. Pleon segment 4 with a submedian pair of low conical tubercles; the angle of the pleuron where it articulates with the telson is also often rather prominent. Telson with a submedian pair of large bosses; apex trilobate, the median lobe not dorsally gibbous. Uropod, inner ramus not extending to apex of telson, somewhat curved, outer distal corner sharp, distal and inner margins forming a continuous curve; outer ramus elongate, outer and inner margins straight, terete, with apical point, rather thickly setose on inner margin. Penial processes on 7th peraeon segment close together, elongate, slender, tapering to acute points. Pleopod 2 ¢ stylet one third longer than ramus, proximally rather stout, distally tapering to a subacute apex, inner margin proximally and the apex minutely setulose (cf. hanseni Stebb., loc. cit., pl. 9, fier Cine ae Length (largest) 11 mm., breadth 6 mm. As preserved, cream or greyish, mottled, or slaty-grey. Locality. Inhambane. 10 gg. (U.C.T.) Remarks. Similar to mossambicus in having submedian tubercles on 4th pleon segment and submedian bosses on telson, but smooth (to the naked eye) and almost glabrous. The ten specimens were taken on three different occasions, but no females referable to this species were caught. Dies monod: Brnrd. 1951. Barnard, Ann. Mag. Nat. Hist. (12), iv, p. 701, fig. 3. Additional locality. Inhambane estuary. (U.C.T.) Remarks. Dr. Th. Monod has suggested to me that Cassidina pulchra Chilton (1924, Mem. Ind. Mus., v, p. 888, fig. 8, and pl. 60, fig. 4) might belong to the genus Dies. This may well be so, but it cannot be decided without further details, e.g. if pulchra has a single penial process, it should be transferred to Dies, if double it can remain in Cassidina (see: M. Edwards, Hist. Nat. Crust., pl. 32, fig. 11, which shows two processes on the 7th sternite). Gen. Ariopoles Brnrd. 1920. Barnard, Ann. S. Afr. Mus., xvii, p. 376. Artopoles capensis n. sp. Fig. 35. g$—Body depressed, almost circular; surface granulose, margin setose. Rostral point curved downwards, abutting against the projecting apex of epistome which is visible in dorsal view. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 73 Antenna 1, 1st and 2nd joints each with a triangular lobate expansion on front margin. Peraeopod 1 dissimilar from the other peraeopods, much stouter, 4th joint produced at lower apex, lower margin crenulate, upper apex with a biserrate spine, 5th joint underriding 6th and partly enclosed in 4th joint, with stout biserrate spine on lower apex, 6th joint with a similar spine on lower apex. Fic. 35. Artopoles capensis n. sp. a. dorsal view. 5), c, d. frontal, ventral and lateral views of epistome; in b sockets of 1st antennae shaded. e. inner view of peduncle of left 1st antenna. /. anterior view of left peraeopod 1 3, with biserrate spine further enlarged. g. pleopod 2 3, with apex of stylet further enlarged. Penial processes on 7th peraeon sternite contiguous. Stylet on pleopod 2 longer than inner ramus, rather stout, margins minutely spinulose, apex acute. Pleopod 3 outer ramus obscurely 2-jointed. Pleopods 4 and 5 both rami without marginal setae, inner ramus with 3 or 4 transverse undulations (scarcely folds or pleats); pleopod 5 outer ramus 2-jointed, squamiferous processes distinct. Telson with small apical notch, no ventral groove. Uropod (peduncle plus inner ramus) broad, almost semicircular, outer ramus minute. Length 6 mm., breadth 5 mm. Locality. False Bay (south of Kogel Bay), 14-17 metres, rock. 1 g. (U.C.T.) 74 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. ‘This 3 specimen is obviously a species of Artopoles, a genus which — was based on a single 2 specimen; no other specimens have been recorded, and no other species has been added to the genus. The generic diagnosis needs some slight emendations to accommodate sexual dimorphism, viz.: telson with small apical notch in 3; peraeopod 1 in § with enlarged 4th joint. The small differences in the pleopods may or may not be important; for the present they can only be recorded. The appearance of the inner rami of pleopods 4 and 5 at first glance seem to indicate affinity to the Hemibranchiate group. The question arises whether this specimen is the ¢ of A. natalis. The ¢ is very much broader than the 2 (6:5 and 4 : 3:25 resp.); if the reverse were the case, one might accept the difference as being merely sexual. The notch at the telsonic apex and also the different shape of the uropods might well be characteristic of the male. But the difference in shape of the lobate expansions on Ist and 2nd joints of the 1st antennae is too great to be considered as sexual. This ¢ specimen is therefore described as a new species. Fam. IDOTEIDAE Synidotea variegata Cllge. 1917. Collinge, Rec. Ind. Mus., xii, p. 2, pl. tr. 1936. Barnard, ibid., xxxvili, p. 185, fig. 16a. 1940. | 1d!, Ann. S: Ayr Mus.) xxxu, 1420. Additional localities. Durban, Empangazi, Richards Bay, St. Lucia Bay; Inhambane. (U.C.T.) Remarks. In the key, loc. cit., 1940, p. 507, in the first couplet the words ‘Peraeopods . . . peduncle’ should be omitted in both sets of characters. An additional character for separating hirtipes and variegata is the width relative to the length: width 24-23 in length in the former, at least 3 in the latter. Synidotea setifer Brnrd. 1914. Barnard. Ann. S. Afr. Mus., x, p: 205, plesnen The fully developed stylet on pleopod 2 of a §, 21 mm. in length, is rather remarkable, but the specimen was received too late for figuring in this paper. It extends beyond the apex of the inner ramus of the pleopod; the inner edge (i.e. remote from the ramus and facing its fellow on the opposite side) is formed by a rather strongly chitinized rib, ending in a short point; from about midway a thin flange extends on one side to the apex, widening distally; this flange is armed with numerous close-set parallel lines of spinules, appearing under a low magnification as if obliquely striated. There is a single median penial process, as in S. hzrtzpes and variegata (Barnard, 1920, Ann. S. Afr. Mus., xvii, p. 380, footnote; and 1936, Rec. Indian Mus., XXXVIll, p. 185). ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 75 Largest specimen 23 mm., an ovigerous 9 16 mm. Colour of some freshly preserved specimens: cream, with a row of tiny black dots along the lateral margins and on upper surfaces of antennae and legs. Additional localities: 33° 53’ S., 25° 51’ E. 26 fms. 6 specimens, incl. 1 ovig. ©, on red Fan-coral (Gorgonia). s.s. Pieter Faure; 34° S., 20° 49’ E. gt metres. meeerijcana (per U-C.T., 1948); 34° 49 S., 20° 21’ E. 47 fms. 1 specimen 23 mm., 1 ¢ 21 mm. (trawler, 1947, per U.C.T.). Cleantis natalensis Brnrd. 1925. Barnard, Ann. S. Afr. Mus., xx, p. 394. 1Q30.) 1G..)hvec, Indian Mus) xxxvi, p. 186, fig. 17. This is the first recorded g. A pair of penial processes. Stylet on pleopod 2 half as long again as inner ramus, rather stout, triquetral in section at base, distal third curving gently outwards, apex subacute. Length 8 mm. Locality: Richards Bay, Natal. 1 g. (U.C.T.) Distribution: Described from Durban Bay, and later recorded from India. Fam. JAERIDAE Taniropsis bisbidens n. sp. Fig. 36. Head, peraeon and pleon sparsely setose dorsally, profusely laterally, setae on lateral margins of posterior peraeon segments and pleon especially long. Head produced in a prominent triangular rostral process, apex bluntly pointed. Side-plates on segments 1 and 5-7 unilobate, on segments 2-4 bilobate. Antenna 1 flagellum 15-16-jointed. Mouth-parts normal. Maxilliped with external distal angle of 2nd and 3rd joints acutely produced; 4th and 5th joints broad. Peraeopod 1 ¢ more robust than the following peraeopods, distal half of lower margin of 4th joint and whole lower margin of 5th and 6th joints with dense fringe of long, very fine (non-plumose) setae. In the larger ¢ the right peraeopod is thus enlarged and furry, but the left is similar to peraeopod 2; in a smaller 3, 5 mm. in length, the reverse is the case. In Q all peraeopods similar (except the hinder ones increase in length, as in @). Pleopod 1 ¢ peduncles fused, rami well developed, adnate but not fused, apices of peduncle and rami each with a tiny scarious projection. Pleopod 2 6 normal. Pleopods 1-2 @ forming a circular operculum with shallowly emarginate distal margin. Pleopod 3 ¢ operculiform, inner ramus with truncate apex, 1 plumose seta on inner angle, 2 on outer; outer ramus enormously enlarged, ovate, incom- pletely 2-jointed, the suture indicated only on inner and outer margin. In 9 also operculiform, but outer ramus not enlarged to such an extent as in g. 76 ANNALS OF THE SOUTH AFRICAN MUSEUM Fic. 36. Taniropsis bisbidens n. sp. a. whole animal, flagella of 2nd antennae omitted. 6. maxilliped. c. 4th—7th joints of peraeopod 1 g. d. pleopod 1 g, with apices of peduncle and ramus further enlarged. e. anterior view of left pleopod 3 3. f/. pleopod 3 9. g. pleopod 4 (¢ and 9). ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA Wil Pleopod 4 g and @ inner ramus with 7-8 plumose setae on the oblique distal margin. Pleopod 5 normal, uniramous. Uropods well developed, rami in both sexes well developed, narrowly ovate, strongly setose, especially the distal half of the inner ramus. Length, 2 gg 5 and 8 mm., 2 ovig, 92 7 and 9 mm. Pale cream becoming white laterally where the integument is thin, probably semitransparent when alive; a black medio-dorsal stripe which bifurcates in front on the head and behind on the pleon; in front the black colour is continued on to the peduncles of 1st antennae, and peduncles of 2nd antennae except the last (6th) joint; behind it is continued on to the peduncles and proximal halves of the uropods. Eyes black. Locahty. False Bay (Oatland Point, south of Simonstown), 4-64 metres, rock. maa zovig. 92. (U.C.T. 1953.) Remarks. Remarkable at first sight on account of its striking colour pattern. It would be interesting to discover the particular habitat and habits of this animal. The coloration may be concealing, adapted possibly to that of some Hydroid or Alcyonarian. The colour appears to be fast in alcohol; the speci- mens were caught in June 1953 and submitted to me in November 1954, and the colour is still (1955) intense black. Morphologically the shape of the head is noticeable, but there are no unusual features in the appendages except the 3rd pleopods. These are oper- culiform in both sexes; in the 9 the outer ramus is moderately enlarged, some- what similar to that of Jaera marina (Sars, 1897, Crust. Norw. I, pt. 5-6, pl. 43, labelled ‘plp. 1’), but in the g the outer ramus is enormous. Fam. BoPpyRIDAE Gen. Athelges Gerst. 1861. Hesse, Ann. Scr. Nat. (4), zool., xv, p. 91 (Athelgue ? vernac.). 1862. Gerstaecker, Arch. Naturg., xxviii, Abt. 2, p. 558. 1898. Sars, Crust. Norw., II, p. 209. As the genus Parathelges Bonnier 1900 does not seem to be differentiated by very tangible and distinctive characters from Athelges, the present species may be assigned to the older genus. Athelges caudalis n. sp. Fig. 37a. 9-First pair of oostegites projecting beyond head like a double ‘cocked hat’, the outer angles curving backwards (cf. Parathelges whiteleggei N. & B., 1931, Vidensk. Medd. Dansk. nat. Foren., xci, pl. 1, fig. 4). Pleon narrower than peraeon; 5th segment and ‘pleo-telson’ freely projecting (not embraced by pleura of 4th segment as in P. weberi N. & B. 1923), with a 78 ANNALS OF THE SOUTH AFRICAN MUSEUM pair of circular uropods. The two pleopodal lamellae on each pleuron arise from a common stalk; both lamellae of pleopods 1 and 2 are ovate; the anterior one of pleopods 3 and 4 is subcylindrical and the other ovate. g—Attached on outside of bases of last peraeopods on left side of 9. Similar to that of A. pagun. The pleon segments fused, longer than basal width, not so elongate as in A. paguri (cf. Sars, 1898, pls. 88, 89), but more elongate than in c Fies/37. a. Athelges caudalis n. sp. 6, c. Bopyrella hodgarti Chopra 9 and @. d. Miophrixus latreutidis n.g., n. sp. P. whiteleggei N. & B. (1931, loc. cit., fig. 92) and P. webert (cf. Nierstrasz & Brandis, 1929, ool. Anz., Ixxxv, p. 302, fig. 11). Eyes distinct. Length, 9 ca. 9 mm., ¢ 1:3 mm. Locality. Inhambane. 1 ovig. 2 with g, on the Pagurid Duogenes senex. (U.C.T.) Gen. Bopyrella Bonnier 1900. Bonnier, Trav. Sta. zool. Wimereux, vill, p. 347. 1923. Nierstrasz & Brandis, Siboga Exp. monogr., xxxii, b, p. 95. 1923. Chopra, Rec. Ind. Mus., xxv, pp. 467, 540. 1929. Nierstrasz & Brandis, Vid. Medd. Dansk. Naturh. Foren., Ixxxvii, p. 29. 1930. Chopra, Rec. Ind. Mus., xxxii, p. 132. 1931. Nierstrasz & Brandis, Vid. Medd. D. Naturh. Foren, xci, p. 150. 1936. Shiino, Mem. Coll. Sc. Kyoto Umv., B. xi, 3, p. 157. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 79 Bopyrella hodgartt Chopra Hig 27D. c 1923. Chopra, loc. cit., p. 473, pl. 14, figs. 7-12. 923 Pp P: 473> P gs. 7 Locality. Inhambane. One ovig. 9, with dg, in left branchial cavity of Alpheus crassimanus. (U.C.T.) Distribution. Vizagapatam, Madras, on Alpheus crassimanus. One ovig. 9, without 3, on Hippolyte, may also belong to this genus. Length 1:25mm. Pleon segments fused, outline entire but slightly undulate on the one side. Miophrixus n. g. 9—Body asymmetrical. All 7 peraeopods present on the shorter side, but only peraeopods 1 and 2 on the longer side. Pleon 3-segmented, segments 1 and 2 distinct, followed by a larger segment representing the fused 3rd-5th segments and telson. A pair of bilobed processes on each side of pleon segments 1 and 2; the dorsal processes (side-plates or exopods of pleopods) more or less pedunculate, the ventral processes (uniramous pleopods or endopods of pleopods) smaller, sessile. Uropods distinct. g—Pleon segments fused. Eyes distinct. , Remarks. ‘The presence of only two free pleon segments, with pleopods, seem to differentiate this form from the several closely related genera or sub- genera of the original genus Hemuiarthrus (olim Phryxus). If Hemiarthrus Giard & Bonnier 1887 is adopted for the preoccupied name Phryxus Rathke, as Chopra (1923, Rec. Ind. Mus., xxv, p. 428) has done, Phrixus proposed by Caroli (1929, and adopted by Nierstrasz & Brandis 1931) falls into synonymy. Miophrixus latreutidis n. sp. Bissso7d. Length, 2 3-3:-5 mm. Attached between the pleopods and facing telson of host. g 0:75 mm. Locality. Inhambane. 3 ovig. 99, 1 3, on Latreutes pygmaeus. (U.C.T.) Fam. DETONIDAE (SCYPHACIDAE) Camorta nicobarica Brnrd. 1936. Barnard, Rec. Ind. Mus., xxxviii, p. 190, fig. 19. Two g¢ and 2 ovig. 99, 5-6 mm. in length, appear to belong to this species. The short telson has the minute apical indent (loc. cit., fig. 19b). The mouth- parts also agree, except that the spines on the apex of outer lobe of maxilla 1 are not so blunt, and the lobe on outer distal corner of 2nd joint of the -maxilliped is more broadly rounded. This last feature should be checked in 80 ANNALS OF THE SOUTH AFRICAN MUSEUM the type specimens; the acute lobe shown in my fig. 19 f may be an error in observation or draftsmanship. Locality. Inhambane, among mangroves. (U.C.T.) Distribution. Nicobar Islands. Remarks. The genus Olibrinus Budde-lund (1912, Trans. Linn. Soc. Lond. (2), zool., xv, p. 389), placed by its author in a separate subfamily of the Oniscoidea, appears to have some similarity with the above genus. Both have a multi- articulate flagellum in antenna 2, but the apex of inner lobe of maxilla 1 has 2 plumose setae in Olibrinus, and only a few setules in Camorita. The three species of Olibrinus have been recorded from the Malay Peninsula, Djibouti, and the Chagos Archipelago. AMPHIPODA Fam. LysIANASSIDAE Euonyx conicurus n. sp. Fig. 38. Closely resembling the North Atlantic : chelatus Norman (see: Sars, Crust. Norw., pl. 40, fig. 1), but differing as follows: Pleon segment 3 dorsally ridged, but not sharply keeled; postero-interior angle quadrate with a short sharp point; segment 4 dorsally depressed at base, then raised into an acute forwardly directed process (cf. Tryphosa onconotus Stebb. 1908). Antenna 1, basal joint prominently lobed in distal half. Gnathopod 2, 6th Fic. 38. joint about two-thirds length of 5th joint, © Ewonyx conicurus n. sp. Pleon seg- upper distal corner gibbous, palm short, ™cn's ae a oid ae rounded, finger weak. Length 6-5 mm. As preserved, white. Locality. Fish Hoek Bay (False Bay), 15 metres. (U.C.T.) Tryphosa normalis n. sp. Fig. 39. No eyes. Lateral angles of head moderately acute. Postero-lateral angles of pleon segment 3 quadrangular, the angle rounded. Pleon segment 4 with convex dorsal hump, but not sharply carinate. Epistome moderately pro- jecting, rounded above, front margin straight. Gnathopod 1, palm of 6th joint very oblique and almost as long as hind margin, angle between the two about (oa | ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 81 160°, defined by two slender spines. Telson with 2 pairs of dorsal spines, and a pair of small spines on apex of each lobe. 8-9 mm. As preserved, white. Locality. Langebaan, Saldanha Bay; False Bay, 26 metres. (U.C.T.) Fic. 39. Tryphosa normalis n. sp. a. lateral view of head. b. epistome and upper lip. c. pleon segments 3 and 4. d. 1st gnathopod. e. telson, with apex of one lobe further enlarged. Remarks. A very ordinary and featureless species. The best differential character seems to be the very oblique palm of gnathopod 1, not unlike that of onconotus Stebb., but the whole joint is more slender and parallel-sided than in the latter species. The epistome is similar in these two species. The dorsal profile of the 4th pleon segment resembles that of the eyeless bispinosa Schell. from the Falkland Islands. Tryphosa africana n. sp. No eyes. Lateral angles of head moderately acute. Postero-lateral angle of pleon segment 3 as in trigonica (Stebbing, Challenger Rep., pl. 9g) but rather more produced (cf. Tryphosites longipes, Sars, Crust. Norw., pl. 29, fig. 1); pleon segment 4 with rounded carina. Antenna 2 ¢ 4th joint broadly oval, width more than half the length. Epistome rounded above, front margin straight. Gnathopod 1 6th joint slightly ovoid, palm very oblique, subequal to hind margin. Gnatho- pod 2, 5th joint ovoid, width slightly more than half length of anterior margin. Telson tapering, with 2 pairs of dorsal spines, and a spine and seta on apex of each lobe. Length 7-8 mm. Colour as preserved white. Locality. 32° 5’ S., 17° 52’ E. (off Lamberts Bay), 123 metres. $3, ovig. 99. 5 Febr. 1953. (Trawler, per U.C.T.) Remarks. Rather like trgonica from Kerguelen Island as regards pleon segment 3, and like normalis as regards gnathopod 1. 82 ANNALS OF THE SOUTH AFRICAN MUSEUM Fam. AMPELISCIDAE Ampelisca excavata Brnrd. Fig. 404. 1925. Barnard, Ann. S. Afr. Mus., xx, p. 336, pl. 34, figs. 5-7. This species grows to a length of 10 to 13 mm. The 2nd joint of peraeopod 5 in the original specimen appears to have been abnormal, and a figure is here given of the normal form. Additional localities. Gordon’s Bay (False Bay), 36 metres; Algoa Bay, 14 metres. (U.C.T.) Byblis amsuropus Stebb. Fig. 405. 1908. Stebbing, Ann. S. Afr. Mus., vi, p. 72, pl. 10 (Crust., pl. 36). A co-type from Stebbing, and three other examples from approximately the same locality (off East London) have been examined. These confirm Stebbing’s description. The present specimen also agrees with Stebbing’s description except as regards the 5th peraeopod and the length of the 3rd uropods. It is smaller than the above-mentioned specimens, but about the same size as Stebbing’s type. The 3rd uropod reaches beyond the apices of the other two pairs of uropods by only half, instead of the whole, length of its rami. Fic. 40. a. Ampelisca excavata Brnrd. 5th peraeopod. b. Byblis anisuropus Stebb. 5th peraeopod, with 6th and 7th joints further enlarged. c. Nototropis granulosa Wlkr. 2nd joint of 5th peraeopod. In general, peraeopod 5 agrees with Stebbing’s description, but the 2nd joint is broader, the 5th joint is relatively larger and the front and hind apices, especially the hind apex, embrace the base of the 6th joint, which is shorter and stouter, 7th joint much smaller. Peraeopods 3 and 4 of both sides missing. Locality. False Bay. 1 specimen, 7 mm. in length. (U.C.T., Sept. 1953.) Remarks. When more specimens are available, a separate specific name may be justified for this form. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 83 Fam. HaAvusTorRIIDAE Gen. Urothoe Dana 1891. Stebbing, Trans. Zool. Soc. Lond., xiii, p. 1. @ The tuft of plumose setae on the pleura of pleon segment 2 is on the outer Wsurface in the specimens I have examined (cf. Ruffo, loc. cit., infra, fig. I, 3, 4, 7), not on the inside as Stebbing (loc. cit., p. 4) says. In some cases these tufts Bare composed of numerous very long setae, and are very noticeable in a dorsal B view of the animal. Key to the South African species 1. Peraeopod 3, 4th and 5th joints longer than broad. seeclezan’s | Peraeopod 3, 4th and 5th joints broader than long. Sie @2. Peraeopod 3, 5th joint 14 times as broad as long. Dactyl slender, ae denticulate. . pulchella Peraeopod 3, 5th joint twice (or nearly so) as broad as long. Dactyl aul | form. Ly or Re 3. Peraeopod 3, dactyl spinose. ... . grimaldi | Peraeopod 3, dactyl serrulate or crenulate. bid Bae G! 4. Peraeopod 5, 2nd joint subcircular, as broad as long. ... . serrulidactylus Peraeopod 5, 2nd joint oval, about 1} times as long as broad. i pnnatasand var. femoralis Urothoe elegans Bate 1925. Chevreux & Fage, Faune de France. Amphip., p. 101, fig. 95. 1949.* Ruffo, Ann. Mus. Civ. St. Nat. Genova, \xiii, p. 80, fig. I, 1-4. 1951. Barnard, Ann. Mag. Nat. Hist. (12), iv, p. 704. Length 2-5-3°5 mm. m 8 ©Localities. 21 miles off Cape St. Blaize, 44 fathoms. 2 99 (mutilated), (S. B Afr. Mus., s.s. ‘Pieter Faure’ coll.); False Bay (near Seal Island), larva traps at surface. gd and 99 (s.s. Africana); Still Bay, 38 fms.; Langebaan, Saldanha B Bay. (U.C.T.) & Remarks. The minute denticulations on the dactyl of peraeopod 3 are often B very obscure and difficult to observe. Urothoe pulchella (Costa) Fig: 41a: 1916. Barnard, Ann. S. Afr. Mus., xv, p. 143. 1925. Chevreux & Fage, loc. cit., p. 99, fig. 92. 1949. Ruffo, loc. cit., p. 82, fig. I, 5-7. Localities. 33 miles off Cape St. Blaize, 60 fathoms. 1 ¢ (Barnard); 12 miles off Cape St. Blaize, 42 fathoms. 1 @ (S. Afr. Mus., s.s. ‘Pieter Faure’ coll.) ; False Bay, 40 metres. 1 g, 1 9; Great Brak R. mouth; Knysna; Sundays R. estuary; Port St. Johns. (U.C.T.) *Date of publication of reprints 11th September, 1947. See Vol. Ixiii, pp. 303, 304. 84 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. Reid (1951, Atlantide Rep. II, p. 221, fig. 21) has described U. leone from the coast of Monrovia, a species so close to pulchella as almost to amount to synonymy. The quadrate, slightly produced and acute postero-inferior corner of and joint of peraeopod 3 is unusual and may be abnormal (only 1 @ was obtained). The oblique series of spine-setae near the apices of 4th and 5th joints of peraeopods 1 and 2 are not shown in the figure (in any case they are not specific characters), nor are any stout spines shown on the 4th joint of peraeopod 3. Ayyyiy oe My byp spt iN Min Fic. 41. Outer view of 3rd peraeopod of: a. Urothoe pulchella (Costa); b. U. grimaldii Chevr. c. U. serrulidactylus n. sp. Urothoe grimaldit Chevr. Fig. 410. 1895. Chevreux, Mem. Soc. zool. Fr., viii, p. 428, figs. 1-4 (3). 1925. Chevreux & Fage, loc. cit., p. 99, fig. 93 (¢ 2); and var. inermis, p. 100. 1935. Chevreux, Res. Sct. Camp. Monaco, fasc. xc, p. 69, pl. 16, figs. 10, 16-18 (g) (2 stated to be unknown). 1951. Reid, Atlantide Rep., II, p. 220 (localities only). Peraeopods 1 and 2, 4th joint is slightly oval, the front and hind margins being slightly convex, 14 times as long as 5th joint (in pulchella the 4th joint is twice as long as 5th, and parallel-sided). The usual oblique rows of long slender setae near apices of 4th and 5th joints, on outer surface. ADDITIONS TO FAUNA-LIST OF S.A. CRUSTACEA AND PYCNOGONIDA 85 Peraeopod 3, 4th joint with 5 spines immediately above articulation of 5th joint, and in its hinder half a fan-like series of 6 slender spine-setae, plumose setae on hind margin but not on distal margin (inner surface); marginal plumose setae on hinder half of distal margin of 5th and 6th joints; dactyl cultriform, with (6) 7-8 slender spines in a single row on front margin in both SEXES. Length 5 mm. As preserved (U.C.T. specimens), white, eyes black. Localities. 21 and 33 miles off Cape St. Blaize, 44 and 60 fathoms. 1 3,5 99 (S. Afr. Mus., s.s. “Pieter Faure’ coll.); False Bay, 26 metres; 32° 12’ S., 16° 38’ E., 215 fms.; from sponge; Langebaan, Saldanha Bay. (U.C.T.) Distribution. North and west coast of France; Moroccan coast (Mediter- ranean) ; off Monrovia, Ivory Coast and French Guinea. Remarks. These specimens seem referable to Chevreux’s species without much doubt. Chevreux & Fage describe the dactyl of peraeopod 3 as having 12 spines in a double row in Q, in contrast with the single row of 6 in g. In the present specimens both sexes have only a single row. The variety inermis (French coast and coast of Sahara) has no spines on the dactyl. Urothoe serrulidactylus n. sp. Big. Aue. 1951. Barnard, Ann. Mag. Nat. Hist. (12), iv, p. 704 (ruber, non Giles). Antenna 1, flagellum 6—7-jointed, accessory flagellum 3-—4-jointed. Peraeopods 1 and 2, front and hind margins of 4th joint slightly convex, 5th joint as broad as long. Peraeopod 3, 4th joint with a series of 10 spines in front half and one of 5 spines in hind half, plumose setae on hind margin and a tuft of 4-5 in middle of distal margin on inner surface (not shown in figure); marginal plumose setae on hind half of distal margins of 5th and 6th joints; dactyl cultriform, front margin serrulate. Peraeopod 4, 2nd joint about 1% longer than wide, hind margin entire, with a few setules, and a submarginal row of long plumose setae on inner surface. Peraeopod 5, 2nd joint nearly as wide as long, hind margin convex, entire, with a few setules. Pleon segment 2, postero-inferior corner quadrate, outer surface with tuft of plumose setae. Lower margins of segments 1 and 3 rounded. Telson with 3 spinules on apex of each lobe. Length 4-5 mm. As preserved, yellowish, eyes black. Locality. Kosi Bay, Zululand. 10 99, some ovig. (U.C.T.) Remarks. ‘These specimens were erroneously recorded as ruber. The character of the dactyl of peraeopod 3 is against such identification. In the present specimens it is serrulate, whereas Giles’s figure shows two spines on the front margin; probably there are, normally, more than two spines. I still consider that spinidigitus Wlkr. 1904 will prove to be a synonym of ruber. 86 ANNALS OF THE SOUTH AFRICAN MUSEUM The somewhat oval shape of the 4th joint, and the broad 5th joint of peraeopods 1 and 2 seem to be differential characters in the present specimens. Urothoe pinnata n. sp. Fig. 42a, b. Antenna 1, flagellum 6—8-jointed, accessory flagellum 5-6-jointed. Peraeopods 1 and 2, front and hind margins of 4th joint slightly convex, 5th joint longer than broad. HY \ tC x S 03 Zoology gua se ns os a at XL. 1952- Botany Part 1, 1s. Part 2, 175. 6d. Part 3 XLI. 1952-1955 Zoology is Be aS oe XLIT.. 1953- Palaeontology Part 1, 12s. 6d. Part 2 ‘ Copies may be obtained from— The LIBRARIAN, Soutu ArricAN Museum, Cape Town, GOVERNMENT PRINTER, PRETORIA. NN OO 0oOW NO = wm Ne m1 G9 O29 69 DR OO OD OD ) & WN oO moat OR ON HOO ANNALS | s “AFRICAN MUSEUM H, containing : — 2. < The Species of Lepidochrysops, Euchrysops and Cupidopsis of Souther Africa. The Genitalia (Lepidoptera - Lycaenidae). By DeEsmMonD PP. MurRAY, F.R.E.S. e as Nouveaux ou Interessanis de VlAfrique du Sud. _ Par P. L. G. Benorr, Musée Royal du Congo Belge, Tervuren. ONS ISSUED DECEMBER 1956 PRICE 9520 PRINTED oA THE ANNALS ©).E 7 TELE: BOUTH AFRICAN MUSEUM VOLUME XLII 2. The Species of Lepidochrysops, Euchrysops and Cupidopsis of Southern Africa. The Genitalia (Lepidoptera-Lycaemdae). By Drsmonp P. MurRAy, F.R.E.S. (With Plates I-III) INTRODUCTION With the recent publication of South African Butterflies by D. A. Swanepoel (1953) one is now able to deal with the twenty-eight species of this genus in a more comprehensive way than was previously possible. Swanepoel does not give any information about the genitalia, although the very detailed accounts he gives of each species of the Lycaenidae (which com- prise a third of the volume of 320 pages) are very exhaustive, showing long experience in the field and careful note-keeping. The present author dealt with the same matter twenty years ago (Murray, 1935) from a different view- _ point, but since then a number of new species have been added to the list. Bethune-Baker (1922) is the only author who has previously studied the genitalia of the Lepidochrysops (Catochrysops—nom. preocc.) (see also the same author’s paper on the value of this study, 1914), giving us photographs of all the then known African species. The three magnificent coloured plates (by Knight) are the finest drawings that have ever been made of Lycaenids. Stempffer has already dealt with most of the tropical species. Drawings of the genitalia are much more satisfactory for comparison of these organs. The present ones have all been made direct from mounts with the aid of the camera lucida, to assure accuracy; a photo often conceals the very part one wishes to examine. The present author has dissected the species of all the other genera of the South African Lycaenids; the published findings will be found in two papers (Murray, 1944, 1948). The present genus and the species of Phasis alone remain to complete the study. The disorganization caused by the late war was the ma‘n cause of the delay, but perhaps this now proves to be an advantage since the new species and those hitherto very rare but lately | rediscovered, can be included. It has been recently shown that it is of great 109 VOL. XLIM. PART 2. IIo ANNALS OF THE SOUTH AFRICAN MUSEUM importance in this study to examine first the type species, for all species under any given genus should naturally correspond with the genotype before being placed therein (Murray, 1954). Hf the genitalia had been examined first instead of last when a new species was discovered, we should have been saved a terrible lot of confusion in names, to say nothing of the consequent contro- versies to which such procedure gives rise. Unfortunately it is found that the European, Indian, North American (and no doubt other) species are very mixed up, simply because in the past wing-colouring was practically the only standard by which a species was judged and placed accordingly. With the Lycaenidae this is a much more uncertain criterion than with most other families. It is to be hoped that we have now passed this stage and make use of more modern and better methods for differentiation and classification. Many students still do not seem to appreciate how essential is the present study. Nomenclature is, of course, another great stumbling-block; yet here again if the genotype were always taken as the standard to which all species within the genus should conform, we should be saved the multiplication of small genera, often quite unnecessary and certainly adding to confusion rather than elucidation. In this paper we give short descriptions of the genitalia, after setting down the genotypes; under each species the place and date of capture, from the author’s own collection except where otherwise stated, and a few necessary notes accompanied by clear line-drawings of the male genitalia of all the species for comparison. The same order is used as given by Swanepoel, and Peters’ Check List of Ethiopian Butterflies (1952) can be consulted. At the end a list of food-plants is added, as this important information has been gathered together over a number of years by various collectors and is now practically complete. Detailed description of the genitalia of each species is not necessary as they nearly all resemble each other very closely. It is by the shape of the valves (harpes), particularly in this genus, that the different species are distinguished; these are therefore shown clearly on the plates. These are the two club-shaped claspers which are nearly always sym- metrical but occasionally show asymmetry as in L. asteris and caffrariae. Sometimes the aedoeagus (penis) varies in shape, but not sufficiently to mark out a species as distinct. It is also often forgotten that before and after coitus this organ is apt to show differences, according as it is extended or contracted. The falces are two prominent hooks always present on and being part of the gnathos, attached to the uncus, which here are bilobed hairy pads. In this genus they do not vary much in form or length. The anellus is the central cone-like tube which bears on its ventral surface a shield or place called the juxta, often holding (in some genera) hairs or scales (i.e. the cristae). The juxta probably acts as a guide to the penis, through which this organ passes (Pierce, 1914; Beirne, 1942). In the species of Lepidochrysops the anellus is sometimes large and of varied shape. — GENITALIA (LEPIDOPTERA—LYCAENIDAE) III Gen. Lepidochrysops, Hed. 1923. Hedicke, Disch. ent. Keit., p. 226. *1. L. parsimon parsimon (Fabr.). Genotype. No tail. Exp. 3 37-39 mm. 9 38-46 mm. Potchefstroom, Transvaal, Dec., 1933. For a number of years this species was confused with plebeja (which has a tail), until Bethune-Baker (loc. cit.) showed clearly that it was a distinct species. From the drawing it will be seen that the apex of the valve is quite different in each case. 2. L. plebgja (Butler). With tail. Exp. ¢ 48mm. 2 48-50 mm. Drawing made from Bethune-Baker’s photo-plate. This species has six spots on the underside of hindwing (as also parsimon), whereas patricia and glauca have only five. 3. L. patricia (Trimen). With tail. Exp. 3 38-41 mm. 2 41-45 mm. Pretoria, Transvaal, Febr., 1932. Natal, Oct., 1942. This species is generally more common than the two former ones, but often flies with them. The apex of the valve is rounded, almost club-shaped; the aedoeagus is sometimes bulbous at the point. 4. L. aethopia (Bethune-Baker). With tail. Exp. g and 3 50 mm. Drawing made from Bethune-Baker’s photo-plate as no specimen was available. It is a larger species than the above. 5. L. glauca (Trimen). With tail. Exp. 40-45 mm. 9 40-47 mm. Pretoria, Transvaal, Febr., 1932. Distinguished externally by its pale blue colour and five black spots on the underside of hindwings. This species has often been confused with patricia which it closely resembles, and although the apex o’ the valve is also rounded, it is of a different form to patricia. 6. L. jeffery: (Swierstra). No tail. Exp. 5 and 2 29 mm. Barberton, Transvaal, Oct., 1945 (Swanepoel). This is the measurement given in the original description, but specimens I received are at least 10 mm. larger, 1.e. 38-40 mm. similar to patricia. * The numbers of the species correspond with the numbers of the figures on the plates. I12 ANNALS OF THE SOUTH AFRICAN MUSEUM It was first discovered by Mr. Jeffery near Ulundi, Barberton, October, 1906, and at Noordkioof, Transvaal, November, 1908, the type-specimen being in the Transvaal Museum, Pretoria. It was rediscovered by Mr. Swanepoel in the same district October to November, 1945-7. His very interesting account of his long search for this species, which had not been collected for about thirty- seven years, should be read. 7. L. hypopolia (Trimen). No tail. Exp. 3 38 mm. 2 40 mm. Potchefstroom, Transvaal, 1920 (Ayres). This rare and dull species, only found in the Transvaal, was first taken by Mr. Morant in Natal, September, 1870, and again by Mr. Ayres in 1903 and 1920 near Potchefstroom, and refound there after ten years’ search by Mr. Swanepoel, September, 1950-1. The specimen I dissected was one of those taken by Ayres in 1920, now in the Feltham Coll., Witwatersrand University, Johannesburg. It is possible that two species have been confused here, as the figures given by Swanepoel do not look like ¢ and Q of the same species, and he also queries whether Ayres’ species is identical with his own later captures. The point cannot be decided until comparison is made of the genitalia. 8. L. procera (Trimen). No tail. Exp. 929 mm. 9 30 mm. Natal, Dec., 1941. Estcourt and Weenen, Natal, Oct., 1946. This elusive species, although first named by Trimen as early as 1893, did not become well known until it was found in Natal by Messrs. Pennington and Swanepoel in October, 1943, and again in October and November, 1943-6. The underside markings of the hindwing are almost identical with letsea, which is slightly larger. This led me to confuse the two species in the description given in the Monograph of the S. African Lycaenidae (Murray, 1935). The shape of the valves, however, is quite distinct: in procera very long and tapering; in letsea much shorter and not so pointed. Anellus produced forward with an irregular front edge. g. L. vansont Swan. No tail. Exp. 615mm. Swartklip, Vivo, Transvaal, March, 1948 (Swanepoel). This small species was first found by Mr. Swanepoel on the flats between the Zoutpansberg and the Blaauwberg Mts., Northern Transvaal, in March, 1947; he kindly sent me two males for examination. The valves are very long and slender, the apex (in the two specimens received) ends in a distinct hook. 10. L. letsea (Trimen). No tail. Exp. 3 32-33mm. 9 32-38 mm. Potchefstroom, Transvaal, Dec., 1933. The Transvaal examples show a much lighter colouring on the upper side than the Cape specimens, which in both sexes is a glossy-brown, not greyish as GENITALIA (LEPIDOPTERA—LYCAENIDAE) 11133 those taken by myself near Potchefstroom on the border of the O.F.S. (speci- men now at Pretoria). Mr. Swanepoel says there is very little difference in the colouring of the g and 9, but the Transvaal examples are distinctly grey, not brown as those from the Cape. I suspect two different species have been confused here or overlooked. 11. JL. wrvingt Swan. No tail. Exp. ¢ 20 mm. @ 22 mm. Barberton, Transvaal, Nov., 1946 (Swanepoel). Smaller and darker than the last species. First taken by Mr. Swanepoel on the Nelshoogte Range near Barberton, November, 1946. It flies together with lacrimosa and tantalus. The valve is similar in shape to that of letsea, but shorter; the anellus produced into two long pointed arms. ee denote ((lirinien): No tail. Exp. 3 31-33 mm. @ 31-36 mm. Natal, 12 Nov., 1942. Estcourt, Natal, Dec., 1942. The valve, which is very long, has a sickle-shaped apex, sharply dentate, according to Bethune-Baker, but my specimens do not show this feature. 13. L. lerothodi (Trimen). No tail. Exp. 3 32-33 mm. 9 33-36 mm. Mokhotlong, Basutoland, Jan., 1944 (Pennington). This very rare and dark species (g and 2 both dark brown) was first found by Major Crawshay in the Maluti Mts., Basutoland, in January, 1902, and remained unknown until rediscovered by Mr. Pennington in January, 1940, in the same area and by Mr. Swanepoel in a kloof of the same range in January, 1949, flying with methymna and ortygia. The former collector very kindly sent me some specimens from which the slide and drawing were made. It has not been found outside Basutoland. Valve long with apex turned into a sharp hook; anellus produced into two curved prongs. 14. L. loewenstein Swan. No tail. Exp. g 18 mm. Basutoland, Jan., 1949 (Swanepoel). First taken by Mr. Swanepoel, January, 1949, at the same place as the former species. Although resembling lerothodi very closely, it is much smaller and has a lunule in the hindwing, which the former species lacks, and five prominent black spots on the underside of the hindwing. The valves are much larger than in the former species, the apex being slightly curved but not into a curved prong. It4 ANNALS OF THE SOUTH AFRICAN MUSEUM 15. L. tantalus (Trimen). No tail. Exp. $ 33-37 mm. 941 mm. Letaba District, Transvaal, Oct., 1943 (Swanepoel). Estcourt, Natal, Oct., 1946. After being lost sight of for a great number of years this species was found again in 1943-6, both in the Transvaal and in Natal. The valves are very long and narrow, tapering at apex into a slight hook with a fine point; anellus large, the front edge being greatly curved. 16. L. pephredo (Trimen). No tail. Exp. 3 32-36 mm. 236mm. Estcourt, Natal (Tring Mus.). First discovered by Mr. C. W. Morrison at Estcourt in October and December, 1888, and lost sight of until October, 1943-6, when it was collected again in the same area. This species closely resembles grahami. The valves in both species are very similar, with a distinct difference in the shape of the apex, the former being broad, the latter narrow in comparison ; anellus large, excavated above in front and curved laterally. 17. L. puncticilia (Trimen). No tail. Exp. d 25-32 mm. 2 32-35 mm. Withhock [sic] Springs, Transvaal, Jan., 1938. A very dark brown species, both g and ¢, on both sides of the wings. The valves of this species, pephredo, and methymna resemble each other very closely; anellus front edge sub-conical. 18. L. methymna (‘Trimen) No tail. Exp. S 20-31 mm. 2 32-99 mm. Cape, Och wnosr This is also a very dark species. Valve long, apex narrow, tapering into a slightly curved fine point; anellus front edge well curved. A much larger variety of this species was discovered by Mr. C. G. Dickson at the Cape, November, 1936: Exp. 42-44 mm., with very dark brown, almost black, uppersides. It is as large as triment, which it resembles closely in the underside markings and with which it has previously been confused. The apex of the valve shows a more distinctly hooked shape than that of the smaller methymna; the apex of the valve is given for comparison in the figure alongside fig. 18. The author thinks it should be recognized as a distinct species on account of these differences and proposes the name major sp. nov. Cape Penin- sula, November, 1936 (Dickson). cf. Trans. Roy. Entom. Soc. Lond., \xxxvii, pt. 10, pl. 2, figs. 48-50, 1938, where methymna and puncticilia are shown together with the following variety (bacchus). ‘There are at least three or four forms of methymna; until these have been carefully compared and more material is available, it is not possible to say whether they are distinct species or not. GENITALIA (LEPIDOPTERA—LYCAENIDAE) 115 19. L. bacchus Riley No tail. Exp. 9 13 mm. 9 15 mm. Tygerberg Mt., Cape, Sept., 1936 (Dickson). This very small example, described as a new species in 1938, was found on Tygerberg Mt. near Cape Town in September, 1936, by Mr. Dickson, who very kindly sent me a specimen. Although smaller than puncticilia and much smaller than methymna, it resembles both these species very closely. My mounts show very little difference, if any, in the shape of the valves from methymna, though slightly larger and broader in the latter. “The aedoeagus alone seems to distinguish them.’—(Riley). But a difference in form of this organ alone does not warrant specific separation. As already shown in the introduction to this paper, this organ sometimes varies in form before and after coitus. For these reasons, in the author’s opinion bacchus is a small variety of methymna, probably confined locally to dry, arid hillsides and therefore a local variety but not a distinct species. 20. L. ortygia (Trimen). No tail. Exp. 3 34-39 mm. 2 39-42 mm. Muizenberg, Cape, Febr., 1930. Kalk Bay Mts., Cape, Dec., 1942 (Dickson). The unusually long-arm valves, with the sudden excised sickle-shaped apex, distinguishes at once this species from all the others. Some examples show a slightly different shaped apex. on eo wasiens(Godart): With tail. Exp. 3 40-45 mm. 92 45 mm. Cape, Nov., 1942. Until Bethune-Baker (loc. cit.) pointed out the mistake in identification, this species was confused with the following. Valves sometimes asymmetrical, very near in shape to trimeni, but distinct; anellus with front edge well arched. 22. L. trimem (Bethune-Baker). With tail. Exp. gd and 2 38-45 mm. Cape, Jan., 1936. Bethune-Baker named this insect after Trimen, but there are two species involved, not one only; the underside also shows a difference from asteris. Valve apex broader than the previous species; size varies somewhat in both forms. 23. L. caffrariae (Trimen). With tail. Exp. 3 38-41 mm. 2 38-45 mm. Natal, Dec., 1941. Cape, Dec., 1935; One valve in this species is nearly always asymmetrical, a distinction, as well as being quite different from the previous two species in form; the left valve shown in the drawing is the normal form. 116 ANNALS OF THE SOUTH AFRICAN MUSEUM 24. L. swanepoeli Pennington No tail. Exp. 3 20 mm. Barberton, Transvaal, April, 1948. This small species was first taken by Mr. Swanepoel near Sheba Mines, Barberton, in October, 1945, flying with jeffery. The valves are very narrow and long, the most delicately formed of all the South African species. 25. L. grahami (Trimen). No tail. Exp. 5 33-36 mm. 2 38 mm. Cala, Cape, Dec., 1950 (Swanepoel). It was first taken by Mr. F. Graham in the north-east Cape Province in 1891. As already noted, the valves are almost the same as in pephredo, the apex in the present species being only slightly narrower. 26. L. lacrmosa (Bethune-Baker). No tail. Exp. 3 30-38 mm. 9 35-40 mm. Greytown, Natal, 1920 (Pennington). Harding, Natal, Oct., 1922. This and the two following species show a form of valve quite distinct from all the former examples; they approximate more to those of Euchrysops, to which genus the author thinks they should be moved, if we are to adhere to the rule that species should conform to the genotype. In the present case the valve is much broader and armed with a hood of fine spines near the apex, distin- guishing it from miobe; the aedoeagus is large with a fine hooked point; the anellus forms two curved arms on each side of the penis. 272) HE niobe! | Mirinien))|: No tail. Exp. g 26-33 mm. 92 26-38 mm. Natal, Jan., 1941 (Clark). Natal, Dec 1942: This species is slightly smaller than the previous one. The genitalia similar in most respects, except that the valves are not armed with spines; aedoeagus extended to a long hooked point with two fine spines or cornuti (Pierce) ; the anellus is the same as in lacrimosa. 28. L. artadne, Butler. No tail. Exp. g and 9 39-42 mm. Karkloof, Natal, April, 1947. This very rare species was first taken near Karkloof Falls, Natal, in 1897; it was not until forty-five years afterwards that it was found again in 1941 near Belgowan, Natal, by Mr. Clark. It has never been found outside Natal. The photo Bethune-Baker gives of the genitalia clearly shows how inadequate photographs of these organs often are in comparison with a drawing, as in his picture the form of the valve cannot be seen at all. Compared with niobe this species shows a distinct shape of valve; the aedoeagus is also different, it is nearer in general form to lacrimosa. ‘Through GENITALIA (LEPIDOPTERA—LYCAENIDAE) 1E7 the kindness of Mr. Swanepoel, who took it at Karkloof in April, 1947, and sent me a male for dissection, the genitalia can now be compared. Gen. Euchrysops Butler 1900. Butler, Entomologist, xxxiil, p. 1. 29. FE. cnejus (Fabr.) Genotype. With tail. Exp. g and 2 27-32 mm. Cawnpore, India, Aug., 1935. The valves are strongly made, long and straight, excavated at the apex forming a head like the head of a spanner; aedoeagus long, point divided; anellus developed into a large cone-shaped organ. 30. EE. barkert (Trimen). With tail. Exp. 3 30-33 mm. 2 29-36 mm. Natal, Tring Mus. No date. This is a species that has rarely been found, though where it occurs, it seems to be locally common, both in Natal and in the Transvaal. The valves are very long and narrow; anellus a hood-shaped cone; aedoeagus long with a bulbous apex. This species is definitely wrongly placed here; with dolorosa it should be transferred to Lepidochrysops. (See under No. 33.) 31. E. osiris (Hopffer). With tail. Exp. 3 34-38 mm. 2 34-39 mm. Klerksdorp, Transvaal, May, 1932. Natal, Dec., 1941. Valves short and broad; apex a rounded dentate head; aedoeagus short, sometimes holding spines or cornuti; anellus front edge rounded. 32. E. malathana (Boisd.) (syn. asopus Hopffer). No tail. Exp. J 25-35 mm. 2 25-31 mm. Naboomspruit, N. Transvaal, Nov., 1932. Potchefstroom, Dec., 1936. Valve fairly short and broad, the apex book-shaped; aedoeagus bulbous at apex with two fine spines; anellus cone-like, well developed. _ 33. £. dolorosa (‘Trimen). No tail. Exp. J 25-29 mm. 9 29 mm. Magaliesberg, Transvaal, Oct., 1927. Valves very long and narrow, tapering into a fine hooked point; aedoeagus divided at apex; anellus cone-shaped. This species, with barkert (No. 30) is definitely wrongly placed here; the genital form is identical with that of all the other species of Lepidochrysops, to which by affinity both species naturally belong. 118 ANNALS OF THE SOUTH AFRICAN MUSEUM Gen. Cupidopsis Karsch. 1895. Karsch, Ent. Nachr., xxi, p. 298. 34. C. cissus (Godart) Genotype. No’ tail) Exp. 3) 30-33 mm. 2 30-38) mm" Irene, Pretoria Wcemmmaens Witpoortje, Krugersdorp, ‘Transvaal, Jan., 1930. In this genus the valves are of quite different formation from those of Lepidochrysops, which feature naturally separates the two genera. The form of the valves in the two species resemble each other closely. 35. C. jobates (Hopffer). With tail. Exp. J 25-30 mm. 9 26-30 mm. Pretoria, Transvaal, Jan., 1931. Klerksdorp, Transvaal, May, 1931. Natal, Jan., 1941. Two new species from Southern Rhodesia have recently been described and figured, viz. L. ruthica Penn. and barnest Penn. (1953, jf. Entom. Soc. S. Afr., xvi, pt. 2.) SUMMARY The paper has dealt with all the present known species of the Lepidochrysops found in South Africa, accompanied by drawings of the male genitalia for comparison and distinction. The great importance of examining these organs and comparing them with the genotype before a given species is placed in any particular genus is stressed. One species only seems doubtful, i.e. L. bacchus, for the reasons stated. The author considers from the reasons given that L. lacrimosa and L. niobe should be placed in the genus Euchrysops. E. barkert and E. dolorosa are shown to be definitely wrongly placed and should be transferred to Lepidochrysops. One new species is named under No. 18. The author would like to state that his collection of the Lycaentdae of Europe (Pethers Coll.), South Africa, India, and of North America has been recently donated to the South London Entomological Society’s collections now housed at Burlington House, Piccadilly. FOOD-PLANTS Species Plant Family I. parsimon (1) Lantana salvifolia Jacq. (Swanepoel, Verbenaceae 1953) (2) Salvia sp. Transvaal (Murray, 1935) Labiatae 2. plebeja (1) Lantana salvifolia Jacq. (Swanepoel) Verbenaceae (2) Salvia sp. (cultivated) Flowers. Labiatae Transvaal (Murray) 2 (1) mM oO BP OO GENITALIA (LEPIDOPTERA—LYCAENIDAE) 119 Species Plant Family . patricia Lantana salvifolia Jacq. (Irving) Verbenaceae . aethiopia = =F . glauca The same as for patricia? . jefferyi Becium obovatum N.E. Br. (Swanepoel, Labiatae 1953) . hypopolia Becium sp. (Swanepoel, 1943) 5 . procera Becium sp. Flowers. Transvaal (Swane- ee poel, 1953) . vansoni Lantana salvifolia Jacq. (Swanepoel, Verbenaceae 1950) yoletsea Lippia scaberrrma. Sond. ‘Transvaal. Verbenaceae (Murray, 1934) . irvingi Becium sp. Flowers. ‘Transvaal (Swane- Labiatae poel, 1940) . ignota » (Swanepoel, 1953) - . lerothodi ete 38 1948) 3 . loewensteini speek E 1948) e . tantalus aN at 9 1943) >»? . pephredo Becium sp. Flowers. Transvaal (Swane- . poel, 1953) . puncticilia Dischima ciliatum Choisy (Dickson, Selaginaceae Cape, 1943) In the Transvaal food-plant must be different as Dischima_ does not occur. . methymna Selago sp. Flowers. Cape (Dickson, Selaginaceae Cape, 1953) : . bacchus S. fructicosa Linn. (Dickson, Cape, 1944) 33 . ortygia (1) Selago fructiculosa Rolfe. Cape if (Dickson, Cape, 1945) (2) Becium burchellianum N.E. Br. Cape Labiatae (Clark) . asteris Selago serrata Berg. Cape Selaginaceae . trimeni Selago serrata (and two other sp.) (Lunt) x . caffrariae Becium burchellianum N.E. Br. Labiatae Salvia sp. Gape (Dickson, 1953) a Selago sp. (Dickson) Selaginaceae 120 ANNALS OF THE SOUTH AFRICAN MUSEUM Species Plant Family 24. swanepoeli Becium sp. (Swanepoel, 1953) Labiatiae 25. grahami Saluia sp. Cape. Flowers (Swanepoel, si 1953) 26. lacrimosa Indigofera sp. ‘Transvaal (Swanepoel, Leguminosae 1953) 27. niobe Indigofera sp. Natal (Swanepoel, 1953) uh 28. ariadne I. astraglina. Natal (Swanepoel, 1953) ry 29. cnejus ome rsa 30. barker i ue 31. Osiris (1) Rhynchosia D.C. Flowers. Transvaal Leguminosae (Murray) (2) Vigna monophylla Taub. Flowers. Kenya (Jackson, 1937) 32. malathana Vigna monophylla ‘Taub. Flowers. Kenya 3 (Jackson, 1937) 33. dolorosa Becium sp. Transvaal (Swanepoel, Labiatae 1953) 34. CiSSUS (1) Eriosema cordaium KE. Mag. Flowers Leguminosae Transvaal (Murray, 1935) (2) E. cynoides Benth. Flowers (Murray, 1935) 35. jobates is ss 23 REFERENCES Beirne, B. P. 1942. ‘Morphology of male genitalia in Lepidoptera.’ Ent. Rec., liv, pp. 17-22, 37-39, 1 pl. 1942. ‘Notes on morphology and taxonomy of genitalia of British Rhopalocera.’ Entomo- logist, Ixxv, pp. 211-16, 17, figs.; and 1943. ibid., lxxvi, pp. 50-4. Bethune-Baker, G. T. 1914. “The taxonomic value of genital armature in Lepidoptera.’ Trans. Ent. Soc. Lond., pp. 314-37, pls. 55-65. 1922. ‘Monograph of the genus Catochrysops.’ Trans. Ent. Soc. Lond., pp. 275-366, pls. 12-32 (3 coloured). Murray, D. P. 1935. South African Butterflies. A monograph of the family Lycaenidae. With 12 colour and 6 line plates. London. A new issue was published later (without date and without the author’s permission) in which the order of the plates was mixed up. 1944. “The genus Cupido: genitalia.’ 7. Entom. Soc. S. Afr., vil, pp. 82-95, 2 pls. 1948. ‘The genitalia of some S. African Lycaenids.’ 7. Entom. Soc. S. Afr., x, pp. 182-92, 2 pis., 2 figs. 1954. ‘The genera of European Lycaenids.’ Entomologist, Ixxxvii, pp. 4-11, 2 pls. Peters, W. 1952. Provisional check-list of butterflies of the Ethiopian Region. Lycaenidae. 201 pp. Pierce, F. N. 1914. The genitalia of the Geometridae. xxix, 88 pp., 48 pls. (Introduction for the terms used.) GENITALIA (LEPIDOPTERA—LYCAENIDAE) 121 Stempffer, H. 1936. ‘Contribution a létude des Lycaenidae éthiopiens.’ Bull. Entom. Soc. Fr., xli, pp. 283, 284. — 1938. ‘Mission Scientifique de ’Omo. Lepidoptera. Lycaenidae.’ (2). Mem. Mus. Hist. Nat., Paris. viii, pp. 179-212, 32 figs. Stevenson, R. H. R. 1939. ‘A new variety of Lepidochrysops glauca with notes on genus Lepidochrysops.’ Proc. Rhodes. Sc. Assoc., Xxxvil, pp. 41, 42. Swanepoel, D. A. 1953. Butterflies of South Africa. 320 pp., 17 colour plates. Cape Town. falty ates Ag BY i ihe Rie. >; Ann. S. Afr. Mus., Vol. XLIII Plate I 1. Lepidochrysops parsimon parsimon (Fabr.). 7. L. hypopolia (Trimen). Genotype. Africa. 8. L. procera (‘Trimen). 2. L. plebeja (Butler). g. L. vansoni Swan. 3. L. patricia (Trimen). 10. L. letsea (Trimen). 4. L. aethiopica (Bethune-Baker). 11. L. wvingt Swan. 5. L. glauca (Trimen). 12. L. ignota (Trimen). 6. L. jeffereyi (Swierstra). reps Las & Ann. S. Afr. Mus., Vol. XLITI Plate 11 Fig. Fig. 13. L. lerothodi (Trimen). 19. L. bacchus Riley. 14. L. loewensteinti Swan. 20. L. ortygia (Trimen) 15. L. tantalus (Trimen). a1. L. asteris (Godart). 16. L. pephredo (Trimen). 22. L. trimeni (Bethune-Baker). 17. L. puncticilia (Trimen). 23. L. caffrariae (Trimen). 18. L. methymna (Trimen). 24. L. swanepoeli Penn. Inset: major sp. nov. Apex of valve. Ann. S. Afr. Mus., Vol. XLIII Plate II1 fg A tomes, Sv FSS 25. L. grahami (Trimen). 26. L. lacrimosa (Bethune-Baker). 27. L. niobe (Trimen). 33. EE. dolorosa (Trimen). 28. L. ariadne Butler. 34. Cupidopsis cissus (Godart). Genotype. Africa. 29. Euchrysops cnejus (Fabr.). Genotype. India. 35. C. jobates (Hopffer). 30. E. barker: (Trimen). 31. E. osiris (Hopffer). 32. E. malathana (Boisd.) (syn. asopus Hopffer). 3. Ichneumonidae Nouveaux ou Interessants de V Afrique du Sud. Par P. L. G. BEnorr, Musée Royal du Congo Belge, ‘Tervuren. Les insectes qui font Vobjet du présent mémoire m’ont été confiés pour étude par le South African Museum a Cape ‘Town. J’adresse mes remerciements les plus chaleureux a Monsieur le Dr. A. J. Hesse qui a bien voulu mettrea ma disposition cette collection d’Ichneumonides d’une qualité exceptionnelle. Le travail a pu étre mené a bien grace a Paide de Monsieur A. J. Hesse par le prét de types de Cameron et de Morley, de Monsieur R. Malaise du Natur- historiska Riksmuseum de Stockholm par le prét de types de Szepligeti, et de Madame Delfa Guiglia du Museo Civico di Storia Naturale de Génes par le prét de types de Tosquinet. A tous ma profonde gratitude. Les types de toutes les espéces décrites dans les lignes qui suivent se trouvent dans les collections du South African Museum a Cape Town. Subf. METOPIINAE Chorinaeus capensis sp. n. Téte jaune a bande frontale et occiput noirs. Thorax et abdomen noirs. Pattes jaunes, les femurs partiellement testacés. Antennes brunes, noircies au dessus. Ailes hyalines. Flagellum des antennes composé de 21 articles, tous plus longs que larges. Face et clypéus bombés dans tous les sens, luisants, couverts de ponctuation peu profonde mais serrée. Front et vertex lisses et luisants; les ocelles limités a Pextérieur par un sulon. Tempes réguli¢rement arrondies. Mésonotum finement ponctué, la ponctuation densea la base, plus clairsemée vers apex; la seconde moitié du mésonotum est aplatie et luisante. Scutellum entierement rebordé, lisse et luisant sauf quelques points a la base. Méso- pleures luisants, a fine ponctuation dispers¢ée, la partie postérieure présente une succession de fines rides longitudinales; aucune caréne entre les mésopleures et le mésosternum. Métapleures lisses et luisants; il existe de fortes rides longitudinales sur la partie inférieure. Le propodeon présente une aire médiane rectangulaire, deux fois plus longue que large, couverte de sculpture finement ruguleuse. Les aires latérales couvertes de ponctuation dense et profonde. Tous les tergites 4a ponctuation forte et serrée. Le tergite 1 est 1,25 fois aussi long que large a l’apex, il porte deux fortes carénes médianes paralléles et deux latérales. Tergite 2 peu plus large que long et muni de trois carénes entiéres, dont les deux latérales sont légérement sinueuses. Le tergite 3 porte trois carénes qui s’estompent vers le milieu. Tibias II avec deux longs éperons. Les fémurs III sont 2, 5 fois aussi longs que leur plus grand diamétre. Les tibias III sont 3,2 fois aussi longs que leur diameétre apical. 123 124 ANNALS OF THE SOUTH AFRICAN MUSEUM Ongles des tarses I et II lobés, ongles des tarses III simples. Longueur: 5 mm. Cap: Mossel bay 12, X, 1941 (R. Turner), 3, holotype au South African Museum. Par les tibias II munies d’un seul long éperon cette espéce s’apparente a _ C. latitarsis Benoit et C. quadricarinatus Benoit; elle s’en distingue par l’absence — d’une séparation caréniforme entre les mésopleures et le mésosternum. Subf. ORTHOCENTRINAE Orthocentrus limpidus Seyrig Cape: Mossel bay 20, X, 1941 (R. Turner), 3. Cette espéce semble décidément la plus répandue du genre dans la faune éthiopienne. Jusqu’a présent, elle n’était connue que du Centre africain. Subf. BANCHINAE Exetastes peronatus Cameron Cap: Augusfontein (Calvinia) IX, 1947 (Mus. Exped.) 3 & . Dikbome, Merweville, Koup, X, 1952 (Mus. Exped.) ¢. Brandkop, Nieuwoudtville, IX, 1941 (Mus. Staff), 3. Quatre exemplaires identiques au Type. Le male se distingue de la femelle par les mandibules tachetées de jaune. L’espéce était connue par deux exemplaires (2 et ¢) ayant servi a la des- cription et originaires du Namaqualand: O’okiep et Klipfontein. Glypta hispida Benoit Uganda: Mulange novembre, 1922 (R. Dummer), 3, allotype. Le male différe de la femelle par la ponctuation faciale disparue au milieu, ce qui laisse le tegument lisse et luisant a cet endroit. Les petites épines du scutellum sont atténuces. J'ai décrit cette espéce sur un exemplaire originaire de Basoko, localité située en pleine forét équatoriale. Il est étonnant de la retrouver en Uganda, ou les conditions écologiques sont absolument différentes. Sjostedtiella pulchella Szepligeti Cap: Mossel bay, 3, XII, 1941 (R. Turner) Goedehoop (Heidelberg District), X, 1951. S76stedtiella negripectus Brullé Glypta Brullé, Hist. Nat. Ins., Hym., 1846, p. 108, 2 3. Morley, Ann. S. Afr. Mus., XV, 1916, p. 388. ICHNEUMONIDAE DE L’ AFRIQUE DE SUD 125 Glypta nigro-ornata Cameron, Ann. S. Afr. Mus., V, 1906, p. 117, gd (syn. nov.). Sjostedtiella nugripectus Morley, ibid., XV, 1916, p. 388. Sjéstedtiella unirufa Morley, ibid., XXIII, 1926, p. 460, ¢ (syn. nov.). Le type de Brullé se trouve au Museum National d’Histoire Naturelle de Paris et son bon état de conservation a permis d’établir la présente synonymie. Les types de Cameron et de Morley se trouvent au South African Museum, Cette espéce est connue de la Province du Cap: Stellenbosch, Lammerfontein in the Moordenaars Karoo, Het Kruis (Piquetberg), et Wit River (Wellington). Zululand: Eshowe (Bell-Marley). Sjostedtiella vittenotum Morley Jai eu VPheureuse surprise de pouvoir étudier de cette espéce encore nele exemplaires capturés par Lightfoot en méme temps que les types, sur uf Matroosberg (alt. 3,500 ft.) dans la Province du Cap. Comme les deux types, il s'agit de femelles uniquement; leur coloration ainsi que les caractéres plastiques sont stables. Asphragis densepunctatus sp. n. A. flavidorbitalis Morley, Ann. S. Afr. Mus., XVII, 1917, p. 217. Téte jaune aux marques noires suivantes: une strie médiane faciale incom- pléte, une bande frontale médiane, celle-ci s’étend sur le vertex et envahit Pocciput. Antennes noires. Mésonotum jaune a trois bandes longitudinales noires. Scutellum jaune au milieu, noir sur les cotés. Propodeon noir. Pro- pleures jaunes a bande noire longitudinale. Méso- et métapleures jaunes sauf la partie supérieure noire. Les sternums noirs en grande partie. Abdomen d’un rouge brique sauf le tergite I noir en grande partie et une tache noire sur les deux ou trois tergites suivants. Tous les coxas jaunes, les autres parties des pattes testacées sauf les tarses III rembrunis et les trochanters parfois noirs. Ailes hyalines. Clypéus lisse et luisant. Face densément marquée de points profonds, sa partie médiane nettement bombée et saillante. Front et vertex également couverts de ponctuation profonde, mais celle-ci n’est pas si serrée que sur la face. LTempes rétrécies en ligne droite. Flagellum des antennes composé de 36 articles. Mésonotum de peu plus long que large, densément marqué d’une forte ponctuation, l’espace entre les points est lisse et luisant, sans microsculpture. Scutellum avec une ponctuation dense, ainsi que les méso- et meétapleures. Propodeon couvert de ponctuation profonde et extrémement serrée, sans aucun espace entre les points; cette sculpture est uniforme. La caréne trans- versale normale, réguli¢rement courbée. Les carénes métapleurales entieres. Le tergite I est 1,84 fois aussi long que large a l’apex, il porte a sa base deux courtes carénes le long de la fossette basale, i! est couvert d’une ponctuation 126 ANNALS OF THE SOUTH AFRICAN MUSEUM forte; sur les cotés, cette ponctuation est entremélée de quelques rides; l’espace entre les points est sans aucune microsculpture. Le tergite 2 est 4 peine plus long que large a apex, densément marqué de gros points, mais ici l’espace entre les points est couvert d’une microsculpture alutacée. Les tergites 3 et 4 portent également une ponctuation trés serrée mais un peu plus fine que sur le tergite 2. Tariére aussi longue que le corps. Ongles des tarses avec une dent a la base. Longueur: 8-8,5 mm. Tariére: 8 mm. Province du Cap: Gt. Winterhoek, Tulbagh, 3,600 ft., XI, 1916, 9, holotype au South African Musuem, et IV, 1916, 9, paratype au Musée Royal du Congo Belge. Du point de vue plastique, cette espéce se rapproche de A. rufipes Cam., elle s’en distingue immeédiatement par la coloration de la téte et celle du mésonotum ainsi que par la ponctuation faciale, les dimensions du tergite I et la sculpture différente de ce dernier. Asphragis recalcitrans sp. n. Espéce trés proche de A. marliert Benoit; elle ne s’en distingue que par les caractéres suivants: Mésonotum d’un roux testacé avec deux bandes jaunes. Le propodeon est dépourvu d’un sulcus médian. Chez A. marliert Benoit, ce sulcus part de la base et s’estompe peu au dela du milieu. Le tergite 2 est entiérement aciculé et ne présente aucune trace de ponctuations dans les angles basaux. La tariére est nettement plus longue que chez l’espéce citée; elle atteint ici la longueur du corps. La taille du corps est plus petite. Les autres caractéres plastiques sont rigoureusement les mémes que chez A. marliert Benoit. Longueur: 5 mm. Tariére: 5 mm. Cap: Mossel bay, 23, I, 1940 (R. Turner), 9, holotype au South African Museum. Subf. PIMPLINAE Paracollyria calosoma sp. n. Téte et thorax d’un rouge brique. Tergites 1 et 7 d’un rouge brique, les tergites intermédiaires noirs. Pattes I et II rouges. Coxas, trochanters et fémurs III rouges, les autres parties des pattes III noires. Le tiers basal des ailes antérieures est d’un jaune doré, les deux tiers apicaux sont violacés; cette derniére partie renferme une petite tache jaune a la pointe basale du ptéro- stigma et deux vagues taches hyalines au dela du milieu. Antennes noires, sauf le dernier article du funicule roux. Antennes composées de 29 articles, le dernier article tronqué au sommet. Comme chez les autres espéces du genre, le labre dépasse légérement le clypéus; ce dernier aplati, luisant avec quelques gros points, son bord inférieur IGHNEUMONIDAE DE L’ AFRIQUE DE SUD 127 tronqué. Espace malaire peu plus court que la base des mandibules. Face grossiérement ponctuée, elle n’est pas plus haute que large, munie d’une forte bosse en dessous des antennes. Front et vertex lisses et luisants. Le lobe médian du mésonotum fortement saillant et enti¢érement lisse, les lobes latéraux a ponctuation trés dispersée. Les notaules sont crénelés sur la partie de leur trajet qui longe le lobe médian. Scutellum lisse et luisant. Propodeon luisant a quelques points trés dispersés, muni d’une faible caréne postérieure, semicirculaire; son plan médian accuse un sulcus longitudinal. Caréne métapleurale absente; la séparation propodeon—meétapleures est indiquée par une faible rangée de points. Méso- et métapleures luisants, éparsement marques de gros points. Le tergite I est deux fois aussi long que large, fortement convexe, lisse et luisant. Les tergites suivants sont tous transversaux, lisses et luisants. ‘Tariere aussi longue que |’abdomen. Ongles des tarses I et II munis d’une dent supplémentaire, les ongles des tarses III simples. Fémurs III densément ponctués. Longueur: 16 mm. ‘Tariére: 8 mm. Province du Cap: Cookhouse, III, 1954, 9, holotype au South African Museum. Cette espéce prend place parmi le groupe d’espéces dont la base de Vaile est jaune doré et l’extrémité violacée et qui compte encore P. flavipennis Cam., P. pulchripennis Cam. et P. dapsilis Tosq. Outre la coloration qui lui est spécifique, cette espéce se distingue encore par la face pas plus haute que large, la caréne postérieure du propodeon ainsi que par la présence d’un sulcus longitudinal sur le ce dernier. Theronia hildebrandti Krieger Province du Cap: Mossel bay, 15, XII, 1941 (R. Turner). Cape Town, IX, 1918 (K. H. Barnard). Robertson, 21, IV, 1927 (J. GC. Le Roux). Cette espéce a été décrite de Madagascar d’aprés un exemplaire portant une étiquette de fausse provenance et c’est seulement depuis Seyrig (Mem. Ac. Malg., XIX, 1934) que nous conaissons les caractéres qui la différencient de T. lurida 'Tosq., espéce trés répandue et commune partout. II est plus que probable que les Theronia lurida 'Yosq., signalés par Morley de l’ Afrique du Sud (Ann. S. Afr. Mus., XV, 1916, p. 387) renferment en partie des spécimens de cette espece. Jusqu’a présent elle n’avait jamais été signalée de |’Union Sud- Africaine. Theronia abyssinica Jacobson Péninsule du Cap: Claremont, 23, IV, 1943 (D. E. Willis). Newlands, I/II, 1931 (K. H. L. Key). I] m’a été donné de réétudier les insectes cités par Morley (Ann. S. Afr. Mus., XXIII, 1926, p. 458) comme JT. melanocera. Holmgren et obtenus a Kabete 128 ANNALS OF THE SOUTH AFRICAN MUSEUM (Afrique Orientale Anglaise) comme parasites probables de Plusza orichalcea. Ces exemplaires appartiennent sans aucun doutea la présente espéce. Dans ce cas également, les citations de Morley relatives a 7. melanocera Holmgren doivent se rapporter en partie a T. abyssinica Jac. La distinction entre ces deux espéces est assez subtile, comme di’ailleurs le cas pour tous les Theronia actuellement connus; sauf en ce qui concerne la seul forme mélanique: T. ruwenzorica Benoit. Ce groupe est en pleine évolution. Chez TJ. abyssinica Jac., les tempes sont rétrécies en ligne droite derriére les yeux et les ailes sont légérement jaunies. Chez 7. melanocera Holmgren, les tempes sont nettement arrondies, rétrécies en ligne courbe et leur niveau atteint presque celui des yeux; les ailes sont légerement enfumées. La forme du clypéus est la méme chez les deux espéces. Theronia interrupta Szepligeti Zululand: Mfongosi (W. E. Jones). Cette espéce n’etait connue que par le seul holotype originaire du Ruwenzori- Meru. J’en ai réecemment trouvé un troisieme exemplaire dans les collections du Musée Royal du Congo Belge, il est originaire du Congo Belge. Kivu: Rwankwi, 15, II, 1952 (J. V. Leroy). Ces trois stations démontrent la trés vaste aire de distribution géographique de cette espéce. Afrephialtes violaceus Kriechbaumer Province du Cap: Graafwater, X, 1947. Nardouw, Clanwilliam, IX, 1941. Espéce typiquement sud-africaine; décrite par Kriechbaumer de la ‘Caffrerie’ elle fut rédecrite par Gameron comme Perithous violaceus d’apres un exemplaire originaire de Stellenbosch (Cap). Charitopimpla sericata Kriechbaumer Province du Cap: Burghersdorp, XI, 1939 (Mus. Staff). Natal: Clan Syndicate wattle plantations, 25, VI, 1914 (C. B. Hardenberg). Zululand: Mfongosi (WV. E. Jones). Espece inféodée au parasitisme des Psychides. Sans étre commune, elle est répandue dans toute l’Afrique Noire. Philopsyche abdominalis Morley (Ann. S. Afr. Mus., XV, 1916, p. 388) est probablement synonyme de cette espeéce. Les spécimens ayant servi a la description furent obtenus a Kentani (Cap) comme parasites du Psychide: Chalzoides junodi. Itoplectis melanospilus Cameron Province du Cap: Mossel bay, 5, XII, 1941 (R. Turner). Cette espéce occupe toute la partie méridionale de la faune éthiopienne; la limite de son aire de dispersion doit se situer au Ruanda et au Congo Belge (Kivu) d’ot j’ai pu étudier quelques exemplaires. ICHNEUMONIDAE DE L’ AFRIQUE DU SUD 129 Ttoplectis tosquineti D.T. Transvaal: Louis Trichardt, Zoutpansberg, 4,500 ft., II, 1928 (R. F. Lawrence). Espéce trés rare, connue jusqu’a présent par le holotype, originaire du Cap de la Bonne-Espérance et par un exemplaire récolté récemment au Ruanda (Kagogo). Les deux exemplaires du Transvaal sont morphologiquement identiques au type, mais se distinguent par la présence d’une bande noire sur les cétés du pronotum. Leur coloration est identique a celle de JI. saudus Tosquinet. Un matériel plus abondant est nécessaire pour pouvoir conclure sil s’agit ici d’une race distincte ou simplement une question de variabilité individuelle. Itoplectis rugosus sp. n. Téte noire. La moitié basale des antennes est noire, la moitié apicale rousse. Thorax et abdomen d’un rouge foncé. Coxas du méme rouge foncé; les autres parties des pattes noiratres, les femurs et tibias I et II blancs en majeure partie. Tibias III munis d’une bande blanche a la base. © Funicule des antennes composé de 28-32 articles. Face couverte d’une ponctuation forte et dense, en certains endroits cette sculpture est confluente et forme des rides. Front finement strié en travers, sans points. Occiput faible- ment découpé. Mésonotum enticérement couvert de forte ponctuation serrée. Scutellum renflé, gibbeux; sa ponctuation est atténuc¢e. Mésopleures entiére- ment couverts de ponctuation dense et forte sauf sur le spéculum postérieur qui reste lisse et luisant. Métapleures ponctués comme les mésopleures; sur leur partie inférieure au dessus des coxas III cette ponctuation conflue et forme des rides longitudinales. Le propodeon présente une aire horizontale nettement sépar€e de l’aire verticale; la premiere porte une sculpture rugueuseridée transversalement; la seconde se présente sous la forme d’un miroir semi- circulaire lisse et luisant, et limité sur les cotés par une carene lamelliforme. La base du tergite 1 est lisse et luisante, la partie postérieure grossi¢rement ponctuée, les tergites suivants a ponctuation forte et uniforme. Coxas III mats, enti¢rement couverts de points serrés. Longueur: 6,5-11 mm. ‘Tariere: 1,5-3 mm. @: identique a la femelle. Le 3 paratype de Mossel bay posséde les sternums noirs. Longueur: 8,8-9 mm. Natal: Stella Bush (Durban), II, 1914 (Marley), 9, holotype. Cap: Mossel bay, 13, I, 1940 (R. Turner), 4, allotype, ibidem 17, I, 1940, 9, paratype. Holotype et allotype au South African Museum. Un paratype au Musée Royal du Congo Belge. Holcopimpla vitellina ‘Tosquinet Pimpla, Tosquinet, Mem. Soc. Ent. Belg., 5, 1896, p. 325, 6. Holcopimpla nigricornis Cameron, Ann. S. ‘Afr. Mus., V, 1906, p. 113, @ (syn. nov.). 130 ANNALS OF THE SOUTH AFRICAN MUSEUM Pimpla concolor Szepligeti, Sjést. Kilim.-Meru Exp., Il, 8, 1910, p. 85, ° (syn. nov.). Cap: Wit River Valley, Bain’s Kloof (Wellington), X, 19409. Zuluiand: Mfongosi, V, 1916 (W. E. Jones). Transvaal: Louis Trichardt, Zoutpansberg, 4,500 ft., II, 1923 (R. F. Lawrence). Mozambique: Chai Chai, XI, 1923 (R. F. Lawrence). Cette espéce se caractérise par la présence d’un large sillon médian sur le propodeon et par la ponctuation dense de la base du propodeon. Ce sillon médian est toutefois atténué chez les exemplaires de taille réduite et il subsiste dans ces cas seulement un léger défoncement médian. Chez le male le sillon du propodeon n’existe pas; il est remplacé par une zone médiane lisse, luisante et imperceptiblement défoncée. Chez le male le propodeon en entier, est couvert de ponctuation dense. La taille de cette espéce varie de 7 a 12 mm. Delaulax rufus Cameron Cameron, Ann. S. Afr. Mus., V, 1906, p. 112, 3. Q—Téte, thorax et abdomen d’un rouge vineux sauf la zone médiane du front ainsi que le stemmaticum noirs. Deux petites taches noires a apex du propodeon et deux a lapex du tergite 2. Ailes noircies a reflets violacés. Pattes rouges, les coxas III noirs. Funicule des antennes composé de 30 a 32 articles, a l’extrémité les articles sont plus courts mais ils restent linéaires. Face luisante avec quelques fins points trés dispersés. Le clypéus séparé de la face par un fin sillon. Tempes élargies et arrondies. Occiput a profonde excavation. Mésonotum entiérement lisse et luisant, il est 1,25 fois aussi long que large (rapport 5/4). Scutellum lisse et luisant, normalement convexe. Propodeon a courbure réguliere, sa base est grossitrement ponctuée. Meéso- et métapleures lisses et luisants; la caréne métapleurale entiére. Le tergite 1 est exactement aussi long que large a l’apex, sa cavité basale lisse et luisante, mais le reste de sa superficie est densément marquee de gros points. Le tergite 2 transversal, il est 1,72 fois aussi large a l’apex que long, grossiére- ment ponctué saufa la base et a l’apex ou subsistent des zones lisses. Le tergite 2 est 1,91 fois aussi large a l’apex que long, la ponctuation profonde mais les points nettement séparés; a la base et a l’apex subsistent de grandes zones lisses. Le tergite 4 est deux fois aussi large que long, couvert de la méme ponctuation. Tariére beaucoup plus longue que l’abdomen. Longueur: 10 mm. ‘Tariére 7,5 mm. Province du Cap: Ceres District, Cold Bokkeveld, 15/30, X, 1934 (M. Vers- feld), 9, allotype au South African Museum; Laingsburg District, Koup Siding, X, 1952 (Mus. Exped.), 9. Note—Certains exemplaires possédent le mésosternum noir ainsi que deux bandes noires sur le mésonotum, ces bandes sont confluentes devant. Le seul exemplaire connu a ce était originaire de Cape Town. IGHNEUMONIDAE DE L’ AFRIQUE DU SUD I31I Aanthephialtes hyalina sp. n. Téte noire sauf clypéus et mandibules rouges. Thorax et abdomen rouges aux marques noires suivantes: une petite tache triangulaire a l’apex du mésono- tum, méso- et métasternum, une tache aux mésopleures sous l’insertion des ailes postérieures, la partie inférieure des métapleures, deux petites taches a apex du propodeon et deux a l’extrémité du tergite 2. Pattes rouges, les tarses III noirs. Ailes hyalines. Funicule des antennes composé de 29 4 31 articles, les derniers sont graduelle- ment raccourcis vers l’apex. Espace malaire alutacé, plus court que la moitié de la largeur basale des mandibules. Face densément ponctuée, l’espace entre les points est lisse et luisant. Front luisant, muni de deux petites fossettes centrales, juste au dessus de l’insertion des antennes. Tempes élargies et bombées, elles atteignent les deux tiers de la Jargeur des yeux. Occiput réguli¢rement excave. Mésonotum lisse et luisant, il est 1,2 fois aussi long que large, le lobe médian tronqué. Scutellum lisse et luisant, normalement convexe. Propodeon normalement bombé, lisse et luisant sauf quelques gros points espacés dans les angles antérieurs. Epicnémies enti¢res. Mésopleures a ponctuation serrée sur leur partie antérieure, elle est plus espacé sur la seconde moitié des méso- pleures. Meétapleures bombés, ponctués sur leur partie supérieure. Tergite I, typique du genre. ‘Tergite 2 de peu plus court que large a l’apex (rapport 6/7), luisant avec quelques gros points au milieu. Tergite 3 lisse et luisant sauf une é€troite zone basale ponctuée, il est 1,3 fois aussi large que long. La méme sculpture existe sur les tergites suivants. Tariére aussi longue que le corps. Longueur: 15mm. Tariére: 15 mm. Province du Cap: Upper sources of Olifants River, Ceres, XII, 1949 (Museum Expedition), 2 99, holotype au South African Museum, paratype au Musée Royal du Congo Belge. Cette espéce se distingue de toutes les espéces actuellement connues par les ailes hyalines et abdomen lisse et luisant avec une étroite zone ponctuée a la base des tergites ainsi que par la taille réduite. Delaulax natalensis Cameron Zululand: St. Lucia Lake, 3, XI, 1920 (Marley). Montclair (Durban), 20, VI, 1945 (Marley). Une revision du genre Delaulax Cameron est en préparation. Hemipimpla bifasciata Morley est probablement synonyme de cette espéce. Pimpla bicolor Brullé Brullé, Hist. Nat. Ins. Hym., 4, 1846, p. 98, 8. Lissonota trochanterata Holmgren, Eugenies Resa, 1868, p. 407, ¢ (syn. nov.). Pimpla rufiventris Szepligeti, Sjost. Kilim.-Meru Exp., U1, 8, 1910, p. 82. 132 ANNALS OF THE SOUTH AFRICAN MUSEUM Cap: Mossel bay, 12, I, 1940 (R. Turner). Katberg, III, 1954. Tradouw Pass, Swellendam District, XI, 1925. Zululand: Mfongosi, IV, 1917 et IV, V, 1934 (W. E. Jones). Espéce rarement récolté et repandue du Cap jusqu’au Congo Belge (Kivu). Je remercie Monsieur R. Malaise du Naturhistoriska Riksmuseum de Stock- holm qui a bien voulu me soumettre le type de Holmgren. Pimpla conchyliata ‘Tosquinet Tosquinet, Mem. Soc. Ent. Belg., 5, 1896, p. 288, 9. Pimpla spiloaspis Cameron, Rec. Albany Museum, I, 1904, p. 146, 9 g (syn. nov.). Pimpla brunnewentris Cameron, ibid., I, 1905, p. 238 (syn. nov.). Pimpla properata subsp. kivuensis Benoit, Rev. Zool. Bot. Afr., 47, 1953, P- 155 (syn. nov.). Transvaal: Louis Trichardt, I/II, 1928 (R. F. Lawrence). Potchefstroom sy Nyittes) Natal: Durban, VII, 1916 (Marley). Cap: Robertson, 21, IV, 1927 (J. CG. le Roux). Rhodésie du Sud: Insiza, I, 1919 (G. French). Je remercie Madame Delfa Guiglia du Museo Civico di Storia Naturale de Génes qui a bien voulu m’envoyer le type de Tosquinet en étude, ainsi que Monsieur A. J. Hesse du South African Museum qui me permit étude des types de Cameron. Pimpla conchyliata, subsp. properata Tosquinet Pimpla properata 'Tosquinet, Mem. Soc. Ent. Belg., 5, 1896, p. 203. Pimpla pleuralts Buysson, Ann. Soc. Ent. France, 1897, p. 352. Pimpla madecassa Morley, Rev. Ichn. Brit. Mus., 3, 1914, p. 72. Pimpla properata subsp. aequatorialis Seyrig, Mém. Mus. H. N. Paris, 4, 1936, p. 13 (syn. nov.). Cette forme se distingue de la précédente par les femurs III entiérement — noirs. Il est 4 supposer quil s’agisse ici d’une race biologique qui cohabite parfois avec la véritable P. conchyliata Tosq. Nous ne disposons pas encore d’un matériel assez abondant pour trancher la question avec certitude. Rhodésie du Sud: Salisbury, 19, [X, 1908. Transvaal: Potchefstroom, I, 1930 (J. Joubert). Cap: Jackals Water, Namaqualand (Lightfoot). Outiep, Garies, Namaqualand, IX, 1953 (J. du Toit). Cookhouse, III, 1954 (Mus. Staff). Newlands, I, II, 1931 (K. H. L. Key). Stellenbosch, 30, III, 1927 (J. J. Strijdom). Cape Peninsula, 1941 (Mus. Staff); roadside, Chapman’s Peak, II, 1934. (Wood and Barnard), 2 (2) ICHNEUMONIDAE DE L’AFRIQUE DU SUD 133 Pimpla vumbana Benoit subsp. meridionalis nov. Cette race géographique se distingue de la série typique originaire des Mont Vumba (Rhodésie du Sud) par le thorax noir, sans reflets métalliques et par les fémurs III noirs. Les fémurs I et II sont également plus foncés que chez les exemplaires ayant servia la description. Province du Cap: Katberg, III, 1954, 9, holotype au South African Museum. Natal: Bulwer, 1914 (W. Haygarth), 9, paratype au Musée Royal du Congo Belge. Gen. Clistopyga Gravenhorst Ce genre est représenté par quelques espéces paléarctiques et quelques espéces néarctiques. Une de ces espéces européennes C’. zncitaior F’. fut signalé par Seyrig du Kenya et plus tard par moi-méme du Congo Belge. Aucune espéce typiquement éthiopienne ne fut décrite jusqu’a présent. Le genre fut classé par les auteurs anciens parmi les tribu des Pimplini puis par les auteurs récents, tels Seyrig, et Townes, parmi les Polysphinctim. En réalité ses caractéres morphologiques ne permettent son introduction dans aucune de ces deux tribus sans devoir élargir le sens de ces derniéres. I] se distingue des Pimplini par ja tariére sortant directement de l’extrémité de Pabdomen et par le dernier sternite entier. Il est exclu des Polysphinctini a cause de sa tariére a pointe non acérée et le dernier sternite trés grand pas éloigné de l’extrémité de abdomen comme chez les Polysphinctini mais trés erand, dépassant méme l’extrémité de ’abdomen et formant une large gout- tiére sous la base de la tariére. Il est intéressant de noter que les Polysphinctini sont tous ectoparasites d’Araignées, tandis que les Clistopyga sont endoparasites de chenilles de Lépidoptéres. J’ai obtenu en Belgique Clistopyga incitator F. comme endoparasite larvaire d’un microlépidoptere indéterminé. Je propose en conséquence de considérer ce genre comme type d’une tribu nouvelle: CLIsTOPYGINI répondant aux carateéres essentiels suivants: Clypéus aplati pas comprimé latéralement. Notaules profonds, longs, et convergents. Scutellum normalement bombé. Propodeon sans aréolation, parfois muni d’une faible caréne transversale. Carénes métapleurales entiéres. Tergites a ponctuation profonde et tubercules atténués. Tariére courbée vers le haut, sa pointe pas acérée et pas muni d’‘ecailles’ comme chez les Pimplinz; son extrémité apicale lisse. Dernier tergite tres grand dépassant l’extrémité de Pabdomen et formant une goutti¢re sous la base de la tariére. Ailes san aréole. La nervure récurrente est percée de deux ‘fenétres’. Ongles des tarses fortement lobés a la base. Yeux non échancrés. Clistopyga africana sp. n. Entiérement roux aux marques jaunes suivantes: Clypéus, moitié inférieure de la face, deux stries orbitales, deux paires de taches allongées et jumelées sur le mésonotum, les angles antérieurs du scutellum et son extrémité apicale, 1394 ANNALS OF THE SOUTH AFRICAN MUSEUM partie médiane du métanotum, deux taches sur le propodeon au dessus des coxas III, langle infratégulaire des mésopleures ainsi qu'une large bande longitudinale sur ces derniers. Métasternum noir. Les propleures sont tri- colores: jaunes au dessus et en dessous, le milieu noir et roux: le tergite 2 présente deux taches noires a son extrémité. Pattes jaunes en majeure partie aux marques brunes suivantes: une tache a la face inferieure des coxas III, une strie sur tous les femurs, les tibias III en majeure partie et les tarses III. Ailes hyalines. Téte entiérement lisse et luisante. Une étroite zone faciale est légérement bombée et se détache des parties latérales de la face. Tempes rétrécies immédiatement derriére les yeux suivant une ligne trés faiblement courbée. Mésonotum trés finement marqué de points espacés, l’espace entre les points est lisse et luisant. Notaules profonds, séparant un lobe médian saillant; ils se rejoignent aux 2/3 de la longueur du mésonotum. Le mésonotum est 1,23 fois aussi long que large et hautement bombé. Sillon scutellaire lisse et luisant. Scutellum lisse et luisant, normalement bombé. Propodeon sculpté de fines stries transversales serrées, vers l’insertion de Pabdomen cette sculpture s’efface progressivement; sur les cdtés, les stries sont subponctuées. Propodeon, sans trace de carene postérieure. Mésopleures lisse au dessus, 4 ponctuation trés espacée sur la moitié inférieure. Meétapleures lisses et luisants. Tergite 1 aussi long que large a l’apex, ses bords réguli¢érement divergents; couverts de sculpture ponctucée fine et subruguleuse; la fossette basale lisse. Tergite 2 aussi long que large a lapex, grossi¢rement ponctué, les points se touchent. Tergite 3 de peu plus court que large a l’apex, a ponctuation dense mais un peu atténuée par rapport au tergite précédent, les tubercules nets. Les tergites suivants progressivement plus étroits, a ponctuation plus atténuée. Taille proportionnelle des 6 premiers tergites. tergite I 2 3 4 5 69) Mariere Jongueur 26 32 31 29 27 23 68 largeur 26 DD) OS 32 29 26 Tariére courbée vers le haut et plus longue que les deux premiers tergites. Nervulus interstitiel. Nervellus de l’aile postérieure brisée en dessous de son milieu. gd Identique a la femelle sauf la conformation de l’abdomen qui est plus étroit. Funicule des antennes composé de 24 articles. Taille proportionnelle des 6 premiers tergites: tergite I 2 3 4 5 6 longueur 25 28 25 25 25 27 largeur 7 20 20 20 20 20 ICHNEUMONIDAE DE L’ AFRIQUE DU SUD 135 Longueur 2: 7,5 mm. Tariére: 1,8 mm. o: 7mm. Transvaal: Louis Trichardt, Zoutpansberg, 4,500 ft., II, 1928 (R.F. Lawrence), Q, holotype au South African Museum. Province du Cap: Mossel bay, 22, I, 1940 (R. Turner), ¢ allotype au Musée Royal du Congo Belge. 2(2A) The ANNALS OF THE SOU TH AFRICAN M USEUM are ‘taeued in parts at at inegular { intervals as material becomes available. Some volumes and parts are out of print, and bce are only sold as ee of a set, or ee 3 respectively. The prices of parts published prior to 1940 have been increased. Out of print: Vols. I, II (Paris 1-3, 5, 7, 8), III (Part 1), V (Parts 1, 2, 5, 8, 9), VI (Part 1, - Index), VII (Parts 1, 2, 3, 5), VIII, LX (Parts 1, 2), * (Part 2), XI (Pans 2. 7 vege: — q (Part 1), XXIV (Part 2), XXXI (Parts I-3). Vol. II. 1900-1902 Zoology and Geology (excl. Parts 1-3, 5, 7, 8) III. 1903-1905 Zoology .. os Bion alee ..(excl. Part. Le IV. 1903-1908 Palaeontology os V. 1906-1910 Geology, Palaeontology, Zoology, Anthropology (excl. Parts 1, 2, 5, 8, 9) VI. 1908-1910 Zoology e se: Part a aes VII. 1908-1913 Palaeontology . (Parts 4 and 6 only) IX. 1911-1918 Botany Ge. Parts 1 and 11) X. sm911-1914 Zoology . (excl. Part 2) © XI. 1911-1918 Zoology (a Pats a as Index) XII. 1913-1924 Palaeontology and Geology si rad aie XIII. 1913-1923 ae ae and Zoology _—. .... ae on ae XIV. 1915-1924 Zoology : a ss Sige a e os XV. 1914-1916 Zoology .. ol ee ob Ot eae XVI. 1917-1933 Botany a2 os ae Aga Shes): ue he ae XVII. 1917-1920 Zoology ; ee XVIII 1921 Zoology XIX. 1924-1925 Zoology XX. 1924-1926 Zoology .. ea etones ae a a ey XXI. 1925-19027 Zoology os uf a .. (excl. Part 1) XXII. 1925-1928 Paldcontatagy | : oy See XXIII. 1925-1926 Zoology .. aie ds?) Ree XXIV. 1929-1938 Anthropology and Fetnoloey ea ..(excl. Part 2) XXV. 1927-1928 Zoology ad aro eee XXVI 1928 Zoology a XXVIT 1929 Anthropology oe XXVIII. 1929-1932 Palaeontology XXIX. 1929-1931 Zoology XXX. 1931-1935 Zoology ; INDEX of papers, authors, and subjects, Ppabliched; in Vols. - XXX.. XXXII. 1934-1950 Palaeontology : ‘ gs 4 ‘nly XXXII. 1935-1940 Zoology .. we os ie Aes a XXXIII. 1939 Zoology / A XXXIV. 1938 Zoology ae ok a *. XXXV. 1956 Zoology .. ahs ue As a XXXVI. 1942-1948 Zoology XXXVII. 1947-1952 Archacolaay XXXVIII. 1950 Zoology XXXIX. 1952 Zoology ex wie ue XL. 1952-1956 Botany .. (excl, Index) XLI. 1952-1955 Zoology 4 ee ae XLII." 1953-1956 Palaeontology : 55 .s ae Vs XLII 1955- Zoology .. Ae ie HAW gf Te (Part 1) Copies may be obtained from— The LIBRARIAN, Soutu Arrican Museum, Care Town. HOR ENDO HN ANNHWOOWNKNNHHHNHNHNOBWWRWOR DAN 5 - el © ~ wn Le) io) ms é an 2 10am oo _ OMowr~rH ORODRMOGCODOOOIOMDS © 09 ooo OM By LiEUWE | ie oe XIL n i eee van Brady- ee “Deur LiuwE Dirk Boonstra., (Met plaat IV.) © Ae PRICE bei: ea ee cent A, TEES OF ° THE SOUTH . AFRICAN MUSEUM 5 es cn (Pry.) LIMITED, : COURT ROAD, _WYNBERG, (CAPE i | Na ERRATUM In the paper ‘Fossil Carnivora from Hopefield’, vol. XLII, pp. 335-347 on p. 340 the teeth whose dimensions are given in Table 3 are P;, P, and M, mand not P,, P; and M,: on p. 341 the caption to Table 4 should read ‘Dimen- sions of teeth of fossil Lycaon compared with those of a sample of 11 extant b) 4. The Skull of Tapinocephalus and its near Relatives. By Liztuwe Dirk Boonstra. (With Plate [V and seventeen text-figures.) INTRODUCTION All the known cranial material of the South African 'Tapinocephalidae is to be considered in this communication. I have examined the specimens in the British and American Museums of Natural History and those housed in the South African Museum. Adequate material exists of the genera Tapinocephalus, Mormosaurus, Keratocephalus and Phocosaurus, whereas of Taurops there is only a snout and of Pelosuchus only a single weathered dentary. HIsTORICAL In 1876 Owen!? made a snout the type of Tapinocephalus atherstoner and, although a part of the occiput and a part of the supraorbital region were associated, they were not described or figured. Seeley in 188814 founded Phocosaurus megischion on a pelvis and this specimen with other associated bones were included by Lydekker™ in the genus Tapino- cephalus until I* revived Seeley’s name Phocosaurus. Broom’s genus Pelosuchus,®? founded on a weathered dentary, includes among other bones a fairly good femur and it was on the structure of these postcranial bones that I* recently included this genus in the Tapinocephalidae. The first good Tapinocephalid skull was described by Broom® in 1909, in as Watson terms it ‘a . . . somewhat inaccurate account’ and referred it to Owen’s Tapinocephalus atherstonei and subsequently (1914) it was made the type of the genus Mormosaurus by Watson.'® In the meantime (1912) Broom’ had created the genus Taurops on a very inadequate weathered snout, and in 1913 Haughton! had published some poor photographs of a very good skull of Tapinocephalus in the South African Museum (No. 2344). : } In 1928 Broom® described an incomplete skull of Tapinocephalus, which I refigured in 1936.1 In 1931 von Huene™ founded the genus Keratocephalus on some pieces of a skull associated with some postcranial bones, and in 1951 I? figured an imperfect skull in the South African Museum which I referred to the same genus. In 1936 I? arranged the then known genera of the South African Tapino- cephalia into a number of groups in an attempt to facilitate our study of the various described forms. In this admittedly arrangement of convenience I grouped Taurops and Keratocephalus with Tapinocephalus, and included Mormo- 137 VOL. XLIII, PART 3. 138 ANNALS OF THE SOUTH AFRICAN MUSEUM saurus in another group containing Taurocephalus and Struthiocephalus. Recently, in a study of the girdles and limbs, I* have included in the former group Pelosuchus and Phocosaurus and am here also transferring to it Watson’s Mormosaurus. In the Tapinocephalidae, thus including the genera Tapinocephalus, Phoco- saurus, Mormosaurus, Taurops, Keratocephalus and Pelosuchus, the structure of the skull, considered as a group, can be described as follows: THE GENERAL SHAPE AND FORM OF THE SKULL The Tapinocephalid skull is large and massive (length 516-600 mm., width 456-543 mm., height 315-384 mm.). The snout is relatively weak, short, low to fairly low and fairly wide to wide (preorbital length 273-330 mm., height 144-186 mm., width 222-330 mm.). The orbit, which is relatively small, lies wholly in the posterior half of the skull. ‘The bones of the preorbital half of the skull are relatively unthickened, whereas the posterior half has the bones, particularly of the roof, greatly pachyostosed. In the latter the surface is rough, whereas in the former it is smooth. In this group the thickening also affects the ‘cheek’, whereas in Struthiocephalus and its allies this remains relatively unthickened and the surface is smooth. The pachyostosis of the dorsa] cranial bones is accentuated in centres in Keratocephalus (vide Struthiocephalus), whereas in the other genera these centres coalesce, so that in the former the distinct naso-frontal boss and a parietal mound are present, whereas in the other genera these are incorporated in the general pachyostosis. This pachyostosis extends furthest anteriorly in Tapinocephalus and extends less anteriorly in Phocosaurus, Mormosaurus and Keratocephalus in this order. ‘There is a distinct and abrupt step from the facial to the cranial surface. The postorbital bar is greatly thickened and massive, but its postfrontal part is not boss-like, but has its surface flowing evenly into the general thickening of the upper cranial surface. The orbits are of moderate size and look forwards and outwards, posteriorly bounded by the massive outwardly thickened postorbital bar, dorsally over- hung by the thick pre- and postfrontals; anteriorly the border is much less thickened, except in Tapinocephalus where the dorso-anterior part is also thickened. The nostrils are fairly large and oval, situated on the dorsal surface and directed mainly dorsally and but little laterally; they lie well back from the anterior tip of the snout and are separated from each other by a strong inter- narial bar. The temporal fossa is slitlike in all but Keratocephalus, where the antero- posterior diameter is not so reduced; dorsally the bay, characteristic of most species of Struthiocephalus, is obliterated in all except Keratocephalus, where remains of a bay are still apparent. The interparietal width is fairly to very large (120-240 mm.) with only a slight pinching-in to form a bay in Keratocephalus. THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 139 Due to the forward position of the quadrate the lower jaw is short and the gape of the jaws fairly small. In dorsal view the occipital edge is nearly straight except in Tapinocephalus where it is concave. THe BONES OF THE DORSAL AND LATERAL SURFACES The premaxillaries (P.M.) form the upper median surface of the snout, stretching from the anterior border posteriorly nearly to or beyond the plane of the antorbital border. The posterior extent is least in Tapinocephalus and Keratocephalus and most in Phocosaurus and Mormosaurus. The premaxillaries form a strong internarial bar and then taper in posterior direction — becoming very narrow in Phocosaurus and Mormosaurus—to lie in a grove in the nasals. The abrupt step from snout to cranial roof causes the posterior part of the premaxillaries to be sharply turned up in Tapinocephalus, but less sharply in Keratocephalus, Mormosaurus and Phocosaurus in this order. The nasals (N.) are a pair of long, narrow bones tapering somewhat anteriorly; anteriorly separated by the premaxillaries and posteriorly forming a groove to receive the posterior ends of the paired premaxillaries; posterior to the limits of the premaxillaries the nasals meet in the median line and are here greatly to very greatly thickened to form the antero-median part of the pachyostosis of the cranial roof and the greater part of the distinct naso- frontal boss of Keratocephalus; at the line of thickening the surface of the nasal is bent more or less sharply upwards, thus facing more or less anteriorly; here each nasal carries a depression presumably to house some facial gland. In all but Keratocephalus the nasal thickening is confluent with that of the prefrontal, but in Keratocephalus these two thickenings are separated by a shallow groove. The septomaxillaries (S.M.) are small bones forming the lower border and floor of the nostrils and extend posteriorly as narrow tapering splints between the nasal and maxilla. The maxillaries (M.) are the largest bones of the snout, being long but low; posteriorly a dorsal prong meets the prefrontal, and a ventral prong extends further posteriorly with its upper edge meeting the ventro-anterior edge of the jugal. In the indentation between these two prongs lie the anterior ends of the jugal and the lacrimal. The ventral edge of the maxilla is fairly straight and this straight edge curves very slightly upwards anteriorly, where it is continued by the lower edge of the premaxilla. The lacrimal (L.) is a small, low and relatively short bone. It is shortest in LTapinocephalus, somewhat larger in Mormosaurus and Phocosaurus and unknown in Keratocephalus. It forms only a small part of the median part of the anterior orbital border, which is relatively unthickened. It is overhung by the pre- frontal—greatly so in Tapinocephalus, but much less in the other genera, where the anterior part of the prefrontal is not greatly swollen. It has no contact with the nasal, because of the intercalation of the dorsal prong of the maxilla. 140 ANNALS OF THE SOUTH AFRICAN MUSEUM The jugal (J.) is a strong bone; like the bones of the snout it is not greatly — thickened and its surface is smooth, being without pits and rugae. It forms the antero-ventral, relatively unthickened, border of the orbit. It extends far ventrally, decreasing in width, with its anterior edge curving down sharply — from the contact with the maxilla. Its ventral edge meets the quadratojugal in a firm suture in all the genera except Mormosaurus, where there is a deep incisure reminiscent of that shown in Séruthiocephalus. Its anterior prong is short in Tapinocephalus, somewhat longer in Mormosaurus and long and high in Phocosaurus. Its posterior edge presents a shallow curve, where it meets the anterior wedge-like prolongation of the squamosal; this indentation is deepest in Mormosaurus and shallowest in Tapinocephalus. The prefrontal (Pr. F.) is very greatly thickened along its lateral border to form the very thick rugose overhanging antero-dorsal border of the orbit. In Tapinocephalus it is swollen right up to its antero-ventral edge and thus strongly overhangs the lacrimal but in the other genera its anterior quarter is relatively unthickened. In all the Tapinocephalid genera it meets the maxilla, and the lacrimal is thus excluded from contact with the nasal. In dorsal view its width is seen to vary, this also in specimens of the same genus. Posteriorly it meets the postfrontal so that the frontal is excluded from the orbital border, but a longer or shorter tongue of the frontal represents the original participation of the frontal in the orbital border before the overgrowing swelling of the pre- frontal and postfrontal pushed it out. Medially the swelling of the prefrontal is confluent with that of the nasal and frontal in all except in Keratocephalus. Here the prefrontal is mediaily thinner and a shallow groove-like hollow clearly demarcates it from the naso-frontal boss, as is also the case in the genus Struthiocephalus. An anterior tongue of the frontal is intercalated between the nasal and the prefrontal for a considerable distance in Keratocephalus but less in the other genera and least in Tapinocephalus. The frontal (F.) is the largest bone of the dorsal] skull roof. Its median part is roughly rectanguiar and from this body three tongues diverge; the posterior tongue is the strongest and extends as a fairly thin to fairly wide wedge, lying between the postfrontal and parietal to enter the border of the temporal fossa in its antero-dorsal part; the anterior tongue, already mentioned as a fairly narrow tapering wedge separating the posterior part of the nasal from the prefrontal, is much weaker than the posterior tongue and in Keratocephalus lies in the deepest part of the groove, Jateral to the frontonasal boss; the lateral tongue is weak and only separates the pre- from the postfrontal for a short distance medially. Anteriorly the frontals are greatly thickened at their junction with the nasals, where, in Keratocephalus they form the posterior fifth of the naso-frontal boss, and in the other genera coalesce in the general pachyo- — stosis. Behind the naso-fronta! swelling the frontals remain thick and swollen in YLapinocephalus and Phocosaurus, but in Mormosaurus and particularly in Keratocephalus are reduced in thickness to form a saddle between the elevated naso-frontal boss and a mound on the parietal. Laterally the frontals are THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES I41 elevated near their contact with the postfrontals, forming a distinct ridge in Mormosaurus, which is reminiscent of a similar condition in Struthiocephalus akraalensis. Posteriorly the frontals meet the parietals in a shallowly to deeply curved suture just anterior to the pineal foramen. The postfrontal (Po. F.) forms the dorso-posterior corner of the orbital border, which is here greatly thickened and rugose, and forming the upper part of the postorbital bar, extends to the margin of the post-temporal fossa. (Its posterior extent is inconstant and it is sometimes excluded from the margin of the temporal fossa by the overgrowing of the postorbital. Even in the same skull this inconstancy is shown, where it reaches the temporal fossa on one side but not on the other.) The postfrontal, although greatly thickened and massive, does not form a distinct boss (as in Anteosaurus) because the frontal above and the postorbital below are equally thickened with the swellings coalescing. The great thickening of the postfrontal coalescing with that of the prefrontal overgrows the lateral tongue of the frontal and thus excludes that bone from the orbital border. ‘The postfrontal forms a greater part of the dorso-posterior thick orbital border in Phocosaurus, a small part in Tapinocephalus, a small or great part in Keratocephalus and a moderate part in Mormosaurus. The parietals (P.) as a pair form a much smaller part of the cranial root than do the two frontals. A moderate to large pineal foramen lies near the frontal suture. ‘The antero-posterior length varies considerably even in the same genus; in Mormosaurus and Phocosaurus it is very short, moderate in Tapinocephalus and moderate to fairly long in Keratocephalus. In three of the known Keratocephalus specimens the pineal foramen is surrounded by a low moundlike boss, whereas in a fourth this area is flat to depressed. In one specimen of Tapinocephalus this area is depressed, whereas in another there is a mound; in Mormosaurus there is a low mound and in Phocosaurus this area is flat. There is a considerable variation in the intertemporal width of the parietals; the intertemporal width is great in Tapinocephalus (240 mm.), fairly great in Mormosaurus (201 mm.), moderate in Phocosaurus (177 mm.) and moderate to fairly small in Keratocephalus (120-158 mm.). In Keratocephalus there is still some indication of the pinching in to form a bay to the temporal fossa, as in some species of Struthiocephalus; and in one specimen of Tapino- cephalus there is a narrow niche in the upper part of the temporal fossa. The parietal forms a sharp edge to the upper border of the temporal fossa; from here the parietal presents a lateral more or less vertical face within the temporal fossa and here meets the posterior flange of the postorbital and the dorsal flange of the squamosal. Laterally the parietal enters the posttemporal arch, extending as a thin wedge between the squamosal and the tabular in most of the skulls examined, but in one specimen of Keratocephalus the tabular enters the border of the temporal fossa and the parietal does thus not enter between the squamosal and the tabular, at least on the surface. ‘he degree to which this parietal tongue, between the tabular and squamosal, is visible on the surface of the posttemporal arch is dependent on the amount of overgrowth 142 ANNALS OF THE SOUTH AFRICAN MUSEUM of the tabular. Posteriorly the parietals are buttressed by the tabulars and interparietal. The postorbital (P.O.) is a massive bone forming the lower part of the broad and massive postorbital bar; anteriorly it forms the greatly thickened postero- ventral part of the orbital border and posteriorly most of the thickened anterior border of the temporal fossa. Ventrally it abuts against the squamosal in a long curved suture, and anteriorly meets the jugal along a short suture. Dorsally it meets the postfrontal along a curved suture running across the postorbital bar, but posteriorly the postorbital sometimes ascends along the border of the temporal fossa to meet the frontal and thus excludes the post- frontal from the border of the temporal fossa. (This inconstancy is even seen in the same skull, where the postorbital excludes the postfrontal from the margin of the temporal fossa on the one side but not on the other.) Within the temporal fossa the postorbital sends a short flange upwards and backwards to meet the descending face of the parietal and at a lower level it meets the anteriorly directed flange of the squamosal, which lies within the temporal fossa of which it forms most of the posterior lining. The squamosal (Sq.) is the main constitutent bone of the ‘cheek’. It is a strong element, greatly to very greatly thickened. Its outer surface is swollen in Tapinocephalus, flat and smooth in Keratocephalus, with a strong ridge in Mormosaurus, and with a ridge, knobs and hollows in Phocosaurus. Anteriorly it meets the jugal in a long curved squamous suture—shallowly indented in Tapinocephalus and Keratocephalus, but deeply in Phocosaurus and very deeply in Mormosaurus. In its antero-ventral corner it receives the quadratojugal, which appears to be wedged into it, with a tongue of the squasmosal overlapping the postero-lateral surface of the quadratojugal. Dorsally it is overlapped by the strong postorbital along a long curved suture and further posteriorly it forms the thick lower and most of the thick posterior border of the temporal fossa and much of the posttemporal arch. From the edge of the temporal fossa it sweeps inwards to form most of the facing of the anterior surface of the post- temporal arch and here its dorsal edge overlaps the lower part of the lateral face of the parietal and at a lower level meets the postorbital. On the surface of the posttemporal arch the squamosal meets the dorso-lateral tongue-like process of the parietal, which is usually intercalated between the squamosal and the tabular for a longer or shorter distance; but in one specimen of Kerato- cephalus a thickening of the tabular prevents the squamosal from meeting the parietal on the surface of the ‘posttemporal arch. Lower down the posterior face of the squamosal is covered by the tabular. Postero-ventrally the squamo- sal forms the thickly rounded postero-ventral edge of the skull. This rounded border forms the outer wall of the ‘auditory’ groove; medially of this groove the squamosal forms a strong and prominent curved ridge which forms the inner wall of the ‘auditory’ groove. The tabular (T.) in dorsal view presents only its upper and lateral edge, which forms the postero-lateral corner of the skull, lining the posterior face of THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 143 the parietal and the upper part of the upsweeping squamosal. In lateral view the tabular is seen to present its edge, which forms most of the posterior edge of the skull. The interparietal (I1.P.) in dorsal view just shows its upper edge which lines the median part of the posterior parietal edge. Its width from side to side is small in Tapinocephalus and Mormosaurus, fairly great in Phocosaurus and variable in Keratocephalus. The quadratojugal (Q.J.) in lateral view presents a small to moderate surface. It tapers postero-dorsally and this end is firmly wedged into the antero-dorsal surface of the squamosal. It has quite a strong contact with the jugal, except in one specimen of Mormosaurus, where there is a notch at the junction of jugal with quadratojugal. The quadratojugal is firmly clasped by the squamosal especially along its posterior edge where the squamosal has a long ventral process. The quadrate (Q.) in lateral view has only the outer cotylar face exposed antero-ventrally of the limits of the quadratojugal. THE OcciIPut The occiput is adequately preserved only in TYapinocephalus, fairly well in Mormosaurus and poorly in Phocosaurus and Keratocephalus. The occiput of the Tapinocephalidae forms a large surface, very much broader than high. It is shallowly concave from side to side. Along its median line the occiput is nearly vertical, with its dorsal edge, however, lying some- what further posteriorly than its ventral edge. Along the median line the plane of the occiput makes an angle of slightly more than go° with the plane in which the alveolar borders of the maxillaries lie. In the median line there is a robust ridge on the interparietal which runs to the upper rounded edge of the occiput. On both sides of the median ridge there lies a deep depression — very deep in one specimen (S.A.M. 11294) of Keratocephalus and not very deep in Phocosaurus (S.A.M. 11997). The upper edge of the occiput is thick and rounded and overhangs the lower part of the occiput. The condyle is directed postero-ventrally so that the skull would hang downwards. The condyle forms a stout rounded knob that has its dorsal surface excavated by a shallow groove leading into the foramen magnum. The dorso-lateral corners of the condyle are formed by the exoccipitals. The foramen magnum is large and nearly circular. The posttemporal fossae are small slits, bounded dorsally by the supraoccipital and ventrally by the paroccipital. The lateral outer border of the skull is formed by the squamosal and medial to this edge lies a shallow ‘auditory’ groove that has its inner wall formed by a strong curved ridge formed wholly by the squamosal with no participation by the paroccipital and tabular as in Struthiocephalus. Ventrally the condyles of the quadrates lie far anteriorly of the plane of the occiput. In occipital view the basioccipital part of the condyle forms the median part of the ventral edge, which is laterally continued by the quadrate ramus of the pterygoid, the 144 ANNALS OF THE SOUTH AFRICAN MUSEUM quadrate condyles and the quadratojugal. Looking at the occiput at right angles to the plane in which the alveolar borders of the maxillaries lie, much of the surfaces of the parietal, postorbitals, postfrontals and frontal is seen. The basioccipital (B.O.) in occipital view shows a T-shaped surface with the stem, lying in the median line, excavated by a shallow groove leading into the foramen magnum, and the cross-member forming the ventral edge of the condyle carrying a large notochordal pit. The dorso-lateral corners of the condylar surface are formed by the exoccipitals. Each exoccipital (E.O.) forms a small triangular part of the occipital face lying lateral to the lower half of the foramen magnum and then stretches posteriorly to overlap on to the dorso-lateral corner of the condyle. The supraoccipital (S.O.) is a low, broad bone forming the upper half of the border of the foramen magnum. Laterally it tapers to a point extending just past the posttemporal fossa, and forms its dorsal edge. Ventrally its edge is overlapped by the exoccipital, dorsally it meets the interparietal and, laterally, the tabular with only small contacts with the paroccipital on both sides of the posttemporal fossa. The median occipital ridge is low or absent on the supra- occipital but is a prominent feature on the interparietal. The interparietal (dermosupraoccipital) (I.P.) is the main median bone of the occiput. It is a fairly large bone—nearly square in outline—and carries a strong and prominent ridge along the median line of the occiput. Lateral to this ridge its surface is shallowly to very deeply excavated. Dorsally it forms the strong rounded edge of the middle part of the upper occipital edge, meeting the parietal on the dorsal surface of the skull. The tabular (T.) is the bone with the largest surface of all the bones of the occiput. Dorsally it forms the major part of the strong rounded overhanging edge of the occiput and meets the parietal on the dorsal surface of the skull and on the dorsal face of the posttemporal arch. Dorso-laterally it meets the upsweeping process of the squamosal along the posterior edge of the post- temporal arch. It stretches far laterally but does not contribute to the formation of the ridge lying medially of the ‘auditory’ groove. Ventrally it overlaps the dorsal paroccipital edge lateral to the posttemporal fossa. The paroccipital (P.Oc.) in occipital view is seen to be a strong bar medially abutting against the basioccipital and overlapped by the exoccipital and stretching far laterally to meet the squamosal mesial of the ‘auditory ridge’ and overlapping the dorsal part of the posterior face of the quadrate, which it buttresses very firmly. Its proximo-ventral corner is seen to form the posterior part of the rim of the fenestra ovalis. In occipital view it obscures much of the stapes—especially that bone’s distal end. The quadrate (Q.) in occipital view presents a roughly rectangular face with, ventrally, a pair of strong rounded knobs separated by a groove forming the articulatory condyles for the articular. Laterally the quadrate is over- lapped by the squamosal and flanked by the quadratojugal. Dorsally the posterior face of the quadrate is overlapped by the distal end of the par- 3(@) THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 145 occipital firmly applied to it. Medially the short anterior process of the quadrate is firmly overlapped by the distal end of the quadrate ramus of the pterygoid and dorsally to this the postero-distal process of the stapes is securely applied to the posterior face of the quadrate. Latero-ventrally to the postero- distal corner of the stapes the quadrate carries a low rounded tubercle. An oval quadratic foramen cuts a notch into the lateral edge of the quadrate. The quadratojugal (Q.J.) in posterior view presents a small triangular face;: its inner edge is notched by the oval quadratic foramen; dorsally the squamosal overlaps the posterior face of the quadratojugal. The stapes (St.) can only be partly seen in occipital view, as the paroccipital obscures its upper and dorso-distal corner. The stapedial foramen can just be seen. The postero-distal process, directed towards the tubercle of the quadrate, is seen to be applied to the quadrate and firmly wedged in between the par- occipital and the quadrate and its distal end securely overlapped by the end of the quadratic ramus of the pterygoid. Proximally the footplate of the stapes fits into the ventrally situated fenestra ovalis, which has the posterior part of its rim formed by a downgrown process of the paroccipital and its ventral edge formed by the basioccipital. The stapes is seen to lie diagonally, with its proximal end appreciably higher than its distal end. Finally, the quadrate ramus of the pterygoid (Pt.) is seen to be firmly applied to the postero-medial face of the anterior process of the quadrate and overlapping the ventro-distal end of the stapes. THE VENTRAL SURFACE OF THE SKULL The ventral surface of the skull in the Tapinocephalidae is well exposed in only one specimen of Tapinocephalus (S.A.M. 2344), fairly well in one specimen (B.M. R.3594) of Mormosaurus, poorly in another specimen (S.A.M. 9082), whereas little is known of this aspect in Phocosaurus and Keratocephalus. The bones of the palate and the basis crani lie in practically the same plane, with the fairly weak lateral pterygoidal rami extending ventral to this plane, and the bones of the suspensorium lying still further ventrally. Striking is the very anterior position of the articulatory condyles of the quadrates, which lie far anterior to the plane in which the basioccipital condyle lies, nearly half-way up the ventral surface of the skull. The subtemporal fossae are small, short but fairly wide, the choanae are large, short but wide and the inter-pterygoidal vacuity is a fairly narrow slit just extending anteriorly to between the posterior end of the prevomers. A very small suborbital opening lies between the palatine and transversum. In the projection on to the plane of the maxillary alveolar borders the surface of the occiput is seen as the occiput makes an angle of more than go° with this plane. The basioccipital (B.O.) is, in ventral view, seen to form the greater part of the occipital condyle. The condyle is a strong rounded knob with its postero- ventral corners formed by the exoccipitals. The basioccipital has a fairly large 146 ANNALS OF THE SOUTH AFRICAN MUSEUM and deep notochordal pit facing postero-ventrally so that from the atlas the skull would hang downwards. Anteriorly to its condylar part the basioccipital has a short but apparently wide face, which is slightly tilted down in anterior direction and there carries two low tubera lying just posterior to the basioccipital-basisphenoidal suture. The lateral edge of these tubera forms the ventral rim of the fenestra ovalis, which thus lies far ventrally. Between the lateral corner of the tuber and the jugular foramen the basioccipital abuts against the antero-medial corner of the paroccipital. Further posteriorly the basioccipital meets the flange of the exoccipital, which overlaps the proximal part of the paroccipital face. The basisphenoid (B.S.) posteriorly meets the basioccipital in a digitate transverse suture at an angle so that the ventral surfaces of these two bones subtend a very obtuse angle. On its postero-lateral processes the basisphenoid carries low tubera and their postero-lateral edge forms the antero-ventral rim of the fenestra ovalis. ‘The anterior extension of the basisphenoid, wedged in between the quadratic rami of the pterygoids, is short. Along the median line the basisphenoid carries a low keel lateral to which, near the anterior edge of the bone, lie the internal carotid foramina. Lateral to each tuber lies the foramen for the carotid. The paroccipital (P.Oc.) is a strong and massive bone which forms a strong and firm connecting link between the cranial base (more particularly the basioccipital, with the exoccipital overlapping on to its ventral and ventro-— posterior face) and the tabular, lying dorsally, and the squamosal, laterally, and its firm support of the quadrate is most important. Its antero-lateral corner also affords a firm support to the postero-distal end of the stapes, and its antero-medial corner forms the postero-ventral part of the rim of the fenestra ovalts. Concomitant with the forward shift of the quadrate a rotation of the par- occipital on its long axis has taken place with the result that the paroccipital presents a much greater face in ventral view than it presents in occipital view. The two pterygoids (Pt.) together form a relatively small part of the centre of the ventral surface of the skull. As each pterygoid has no anterior process, only a weak lateral flange, a short quadrate ramus and in the median line extending only a short distance posteriorly, it is a not very prominent bone. Posterior to the interpterygoid slit the pterygoid meets its fellow in the median line to form a low keel, which is continuous with that formed by the basi- sphenoids. Lateral to this median keel each pterygoid is deeply excavated to form a wide and deep diagonally situated groove, which is then laterally bounded by the quadrate ramus, which is a deep sheet of bone lying obliquely in the skull and this sheet lies nearly at right angles to the plane of the palate. Although the quadrate ramus of the pterygoid is comparatively short, it has a large and firm overlap on the quadrate, being firmly applied to the mesial face of the anterior quadratic process and extending well posterior of the plane in which the condyle of the quadrate lies. The upper edge of the quadrate THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 147 ramus lies ventrally to, but with a fairly firm contact with the anterior part of the distal end of the stapes, which, passing above it, is firmly applied to the mesial face of the anterior quadratic process. The quadratic ramus is connected to the lateral ramus by a strong web of bone. The lateral ramus of the pterygoid is weak and does not descend far ventrally nor extend much laterally, which is in strong contrast to the condition in the carnivorous Titanosuchia and Anteosauria. It is also relatively weaker than in the Moschopids, but about the same as in the Struthiocephalids. Laterally the flange is supported by the robust transversum which, because of the small lateral extent of the lateral ramus, presents a large ventral face. ‘This is also seen in Struthiocephalus. Anteriorly the pterygoids do not appear to extend anterior to the plane in which the anterior edge of the lateral rami lie. The transversum (Tr.) is a robust bone forming a strong link between the lateral ramus of the pterygoid and the side wall of the skull, being firmly buttressed against the inner face of the jugal and, to a lesser degree, against the maxilla. Anteriorly it meets the palatine in a straight suture, with only a small suborbital foramen separating them. The palatine (Pal.) stretches antero-laterally from the anterior edge of the transverse pterygoid ramus as a thick sheet of bone to form much of the lateral border of the choana, and has its lateral edge applied to the inner maxillary surface, where it flanks the alveolar border. In the median line a tongue of the prevomers apparently prevents the two palatines from meeting each other. On its postero-median corner the palatine carries a mound-like ridge which does not, however, appear to be dentigerous. The prevomers (vomers) (P.V.) are strong but relatively short bones, which together form a massive interchoanal bar. Their anterior bevelled ends under- lie the inner face of the premaxillaries. Posteriorly they widen to meet the palatines postero-laterally, but in the median line apparently send a tongue posteriorly which is intercalated between the medial edges of the palatines. Anteriorly, along the median line, the interchoanal bone is grooved, whereas posteriorly a low keel is developed. The premaxillary (P.M.) has a very massive alveolar face but in all the known specimens little of the teeth is known and where infilled alveolar sockets can be seen they appear to be irregular with some indication of replacement lingually. The maxilla (M.) has its alveolar border anteriorly massive and wide but it then tapers rapidly in posterior direction with a sharp edge continuous with the sharp ventral edge of the jugal. Little is known of the teeth but it would appear that they were even more degenerate than those on the premaxilla. The jugal (J.) in ventral view presents a sharp ventral edge curving sharply downwards to meet the quadratojugal, and an internal face forming, together with the transversum, the outer border of the subtemporal fossa. Anteriorly it meets the massive transversum in a sigmoid suture. 148 ANNALS OF THE SOUTH AFRICAN MUSEUM The quadrate (Q.) in ventral view presents its articulatory surface as a very prominent feature. The condyle is bipartite with two fairly robust rounded cotyli separated by a shallow groove. The edges of the cotyli overhang both the anterior and posterior faces of the dorsal part of the quadrate. Running obliquely inwards from the inner cotylus is the anterior quadratic process and it is against its mesial face that the quadrate ramus of the pterygoid is very firmly applied. Lateral to the outer cotylus the quadrate edge has a rounded step up before meeting the quadratojugal, which is applied to its outer surface, with a fairly small foramen quadrat: lying half-way along the suture. Since the quadrate lies in a forwardly inclined plane much of its posterior face is seen in ventral view. On this surface there is a small low tubercle above the inner cotylus, to which the postero-lateral sharp process of the stapes is all but applied. In ventral view it is very clearly shown how firmly the more dorsal part of the quadrate is overlapped by the down-sweeping squamosal outwardly and inwardly by the strongly down-sweeping antero-ventral part of the strong paroccipital. Above the upper edge of the quadratic ramus of the pterygoid the distal end of the stapes abuts very firmly against the mesial face of the quadrate. Another part of the distal end of the stapes passes above the antero-lateral corner of the paroccipital to abut against the quadrate. The quadratojugal (Q.J.) in ventral view is seen to be a small element wedged in between the outer face of the quadrate and the lower overlapping edge of the down-sweeping squamosa]. A small foramen quadrati notches the mesial edge of the quadratojugal. The squamosal (Sq.) is seen to form the angle of the ‘cheek’ and sweeping downwards it overlaps the quadrate and quadratojugal. The strong and prominent ‘auditory’ ridge is well shown with the ‘auditory’ groove lying lateral to the ridge. The stapes (St.) lies obliquely; from the ventrally situated fenestra ovalis it is inclined both forwards and downwards. In the ventral view the stapes is seen to present two distinct faces. The ventral face is elongate with a central waist, a knoblike proximal end firmly fixed in the fenestra ovalis and an expanded distal end, which has a long pointed postero-lateral process terminating just above the tubercle on the posterior face of the quadrate and wedged in between the quadrate ramus of the pterygoid and the paroccipital. ‘The posterior face, which is perforated by the round stapedial foramen, is triangular in outline, with its base applied to the quadrate, after passing outwards above the down- wardly grown antero-ventral corner of the paroccipital. By this rather complex system of wedging it is clear that the distal end of the stapes is all but wholly immovable. The tabulars (T.), interparietal (I.P.) and the supraoccipital (S.O.) are visible in ventral view, because the occiput-face is inclined somewhat anteriorly from above downwards. These bones lie well posterior to the basioccipital condyle with the first two forming the posterior edge of the skull, which is slightly concave in Tapinocephalus and Mormosaurus, convex in Phocosaurus and THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 149 Keratocephalus (except in one specimen), whereas in the Struthiocephalids this edge is nearly straight or convex, in Moschopids convex and deeply concave in the Titanosuchia and very deeply concave in the Anteosauria. THE BRAINCASE IN LATERAL VIEW Hitherto the braincase of a specimen of Tapinocephalus atherstoner has been figured! and now a specimen of Keratocephalus, in which both sides of the brain- case have been exposed, is to be described. The braincase in Keratocephalus is short and low with the nearly horizontal basis crani formed by the basioccipital and basisphenoid. The posterior part of the side wall is formed by inwardly directed plates of the squamosal, supraoccipital and paroccipital. The antero-ventral corner of the paroccipital meets the basioccipital to form the posterior part of the rim of the fenestra ovalis. ‘The anterior half of the rim of the fenestra ovalis is formed by the basisphenoid. Anterior to the plates of the squamosal, supraoccipital and paroccipital lie the prootic and epipterygoid, but the sutures here cannot be determined with certainty, but appear to lie where indicated in broken line in the figure (fig. 15). The epipterygoid (Ep.) appears to be incorporated into the side wall with the obliteration of the cavum epiptericum. Its base rests on the pterygoid with possibly a small contact with the basisphenoid. The posterior edge of the epipterygoid has not been determined with any degree of certainty, but probably lies where indicated by broken line in the figure, in which case it is firmly applied to the antero-internally directed plates of the paroccipital and supraoccipital. Dorsally the dumbbell-shaped vacuity for the trigeminal nerve (V.) causes the epipterygoid to bifurcate; the posterior processus oticus meets the prootic, whereas the anterior processus ascendens meets the under surface of the parietal _ with its upper edge and the sphenethmoid with the upper part of its anterior edge. The lower half of the anterior edge of the epipterygoid forms the posterior edge of the fenestra leading into the fossa hypophyseos. The prootic (Pot.), as I have tentatively determined its limits, is a small bone wedged into the dorso-posterior corner of the side wall of the braincase. Dorsally it meets the parietal just mesial to this bone’s contact with the post- orbital. Posteriorly its edge is applied to the plate of the squamosal. Its ventral edge meets the supraoccipital and the upper edge of the processus oticus of the epipterygoid. Its anterior edge is applied to the posterior edge of the processus ascendens of the epipterygoid. The prootic forms the upper edge of the trigeminal foramen. Of the parasphenoid (Ps.) only the dorsal branch is present, the anterior horizontal branch is not preserved. It stands with its thickened base on the upper surface of the basisphenoid and also, anteriorly, on the pterygoids. Immediately above its base the bone narrows and then thickens in dorsal 150 ANNALS OF THE SOUTH AFRICAN MUSEUM direction so that its two lateral faces diverge to where the sphenethmoids rest on its outer dorsal edges. Its dorsal edge forms the anterior border of the fenestra of the fossa hypophyseos. In lateral view its dorsal edge meets the sphenethmoid in a long curved suture. The sphenethmoid (Se.) forms a large part of the lateral wall of the brain- case. In their postero-dorsal part the two sphenethmoids diverge to form a fairly narrow space to house the fore-end of the brain. Antero-dorsally the two thin sphenethmoidal plates are applied to each other to form a thin median septum, which apparently extends far anteriorly. Lateral to this median septum the sphenethmoid thus forms a recess, with the tractus olfactorius emerging in its posterior part. Dorsally of the recess in the sphenethmoid the under surface of the prefrontal is also deeply excavated and in this hollow the bulbus olfactorius was apparently housed. OPENINGS IN THE SIDE WALL OF THE BRAINCASE The fenestra (Hyp.) in the side wall for the fossa hypophyseos is fairly large, narrowly oval and is bounded by the parasphenoid, sphenethmoid, epiptery- goid and ? pterygoid. The fenestra for the trigeminus and abducens nerves (V.) is in outline dumbbell-shaped and it lies in the upper part of the epipterygoid and causes this bone to bifurcate into a processus oticus and processus ascendens. The truncus olfactorius (1) emerges from the posterior corner of the recess in the sphenethmoid. The foramen for the n. trochlearis (IV) pierces the sphenethmoid near its upper edge, just below its contact with the under surface of the frontal. The foramen for the n. opticus (II) emerges from the upper part of a shallow depression lying in the sphenethmoid above the large fenestra hypophyseos. A venous foramen (Ven.) pierces the prootic near its dorso-anterior corner just under this bone’s contact with the parietal. A foramen (Ca.) for the exit of the accessory branch of the carotid lies at the base of the parasphenoid. The fenestra ovalts (F.O.) lies very low down and its rim is formed by the basioccipital, basisphenoid and paroccipital. THe Lower JAw In all the known Tapinocephalid genera the lower jaw is only adequately known in one specimen (S.A.M. 9082) of the genus Mormosaurus. Owing to deformation due to a simple shear the disposition of the elements of the lower jaw has been disturbed in that by telescoping the dentary has been forced backwards over the posterior bones. Because of the great forward shift of the quadrate the lower jaw is very short, viz. at most only 60% of the maximum skull length. It is massive with a strong symphysis and the articulatory facets on the articular lie far ventrally —not much above the plane of the lower border of the jaw. THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES Syl The dentary (D.) is the largest bone of the jaw with a strong, fairly square mentum and it forms by far the greater part of the strong symphysis. Its dorso-posterior end forms a very low coronoid process. On its alveolar border the dentition is very poorly preserved, with only three roots of teeth present anteriorly, but what appear to be emply sockets indicate that in life a fairly regular row of teeth must have been developed, with the anterior ones much larger than the posterior ones. The splenial (S.) is a small bone with a weak symphysis with its fellow. At the symphysis there is a deep groove between dentary and splenial. The angular (An.) presents a large external face. There is a deep angular notch and the reflected lamina (lamina reflexa) is a robust sheet of bone, which in its posterior part lies well away trom the inner sheet of the angular to form a roomy recessus angularis. ‘The reflected lamina forms the free postero-ventral edge of the jaw immediately anterior to the articular. The surangular (S.A.) forms the curved dorso-posterior border of the jaw and this edge is rounded and robust. Dorsally the surangular abuts against the dentary and ventrally overlaps the articular. Internally the surangular presents a large face between the postero-ventral edge of the dentary, the articular and the postero-dorsal edge of the prearticular. This is the area of insertion of the m. temporalis. The prearticular (Pa.) is a long sheet of bone anteriorly intercalated between the dentary and splenial, and posteriorly overlapping and strongly supporting the articular. The anterior half of its upper edge meets the coronoid and then posteriorly it has a free edge lingual to the insertion of the m. temporalis. Its ventral edge overlaps the inner face of the angular. The articular (Ar.) is a robust bone wedged in and securely clasped exter- nally by the surangular from above and the angular anteriorly, and internally by the prearticular. The articulatory facets on its dorso-posterior surface are situated low down. The outer transversely oval concave facet is separated by a low ridge from the inner nearly circular concave facet. A strong but short retroarticular process is directed postero-ventrally and to it was inserted a strong m. depressor mandibulae. The coronoid (C.) is a flat sheet of bone applied to the inner face of the dentary and meeting the upper edge of the prearticular; it does not extend far posteriorly and the dorso-posterior corner of the lower jaw is formed by the dentary. ‘TAXONOMIC ‘TAPINOCEPHALIDAE Skull Characters of the Family Skull large (length 488-585? mm.; breadth 400-540 mm.); relatively short and broad to fairly long and fairly narrow (length varies from 104-127?% of the breadth); relatively low to fairly low (height varies from 46-69% of the length); snout short to fairly short (snout length varies from 51-67% of the Sy ANNALS OF THE SOUTH AFRICAN MUSEUM total skull length); the snout is low to fairly high and fairly broad to broad (height of snout from 44-76% of the width of the snout). The orbit lies in the posterior half of the skull. The transition from the facial to the cranial surface sloping or very abrupt. The dorsal cranial bones are strongly pachyostosed with the centres of thickening coalesced in all but one genus (Keratocephalus); this thickening occupies from 48-63% of the total median length of the skull. Postorbital bar wide to very wide and massive; the postfrontal does not form a prominent boss but its surface curves evenly on to the generally thickened dorsal surface; temporal fossa small with small to fairly small. antero-posterior diameter; intertemporal region moderately to very wide (120-240 mm.) ; parietals enter upper border of temporal fossa; frontal excluded from the orbital border; lacrimal not reaching nasal. Quadrate ramus of pterygoid is short. Dentition in most cases feeble and undifferentiated. ‘TAPINOCEPHALUS Skull Characters of the Genus 1. The preorbital length is 51-52% of the total median length; the snout is thus short (preorbital length 273 mm.) and it is broad (330 mm.) and low (144 mm.), with the height 44% of the width; the dorsal facial surface of the snout is short and does not extend to the plane of the anterior orbital border; the transition from the facial to the cranial surface is very abrupt. 2. ‘The dorsal cranial bones are very strongly pachyostosed, with the centres of thickening coalesced and the transition on to the face is by an abrupt step along a U-shaped line; the thickening extends very far forward—well anterior to the plane of the anterior orbital border; along the median line the pachy- ostotic cranial roof is 339 mm. in length and this is 58-63% of the total median length of the skull; this great anterior extension is due to the greatly thickened naso-frontal boss, which does not, however, stand out dorsally above the general cranial surface. On the sharply up-bent face of the posterior end of each nasa] there is a fairly deep hollow. The prefrontal is greatly thickened and this is confluent with the pachyostosis of the nasal and frontal and it overhangs the orbit forming the greatly thickened antero-dorsal part of the orbital border and limits direct anterior vision. The area around the pineal foramen is raised, but its general surface is lower than that of the surrounding bones. 3. The postorbital bar is very wide and very massive. 4. The posttemporal arch is very thick and the temporal fossa is small with its antero-posterior diameter greatly reduced. ) 5. The dorsal parietal surface is very broad (210-240 mm.) with in one specimen a narrow constricted bay to the temporal fossa; and the interorbital width is 88—100% of the intertemporal width. 3 (2) a THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 153 6. Antero-posterior length of the parietal is fairly long (105-114 mm.); frontal fairly large. 7. The intersquamosal width is very great (495-540 mm.), and is g9-100% of the median length and the skull is thus nearly or quite as broad as long. 8. The quadrate is situated well forward; the mandibular length is 57% of the median length of the skull; the quadrate ramus of the pterygoid, although it extends to well posterior of the condyles of the quadrate and underlies the Fic. 1. Yopinocephalus atherstonet. Lateral view of skull. x %. S.A.M. 2344. Slight dorso-ventral compression not corrected. Note.— All the figures in this paper are not perspective drawings but projections obtained by pantograph. Those labeiled ‘lateral view’ are orthoprojections on to the median plane; ‘dorsal’ and ‘ventral views’ are orthoprojections on to the plane in which the alveolar borders of the makxillaries lie; the ‘occipital views’ are orthoprojections at right angles to the plane in which the maxillary alveolar borders lie. distal end of the stapes, is short; the lateral pterygoid rami form a weak trans- verse bar of small width. 9g. The dentition is poorly preserved but probably feeble. Genotype. Tapinocephalus atherstonei, Owen 1876. Specific diagnosis as for the genus. Holotype. B.M. R.1705. Snout (there are also other parts of the skull associated). Near Janwillemsfontein, Prince Albert. Low Tapinocephalus zone. Coll. Atherstone. Referred Specimens A.M.N.H. 5626. Posterior half of a weathered skull. Ganskraal, Prince Albert. High Tapinocephalus zone. Coll. van Wyk. 154 ANNALS OF THE SOUTH AFRICAN MUSEUM Fic. 2. Tapinocephalus atherstonei. Dorsal view of Jal x 4. S.A.Ms 2344- Fic. 3. Tapinocephalus atherstonei. Occipital view of skull. x }. S.A.M. 2344. Slight dorso-ventral compression not corrected. THE SKULL OF TAPINOCEPHALUS AND ITS NEAR RELATIVES 155 S.A.M. 2344 (figs. 1-4). A good skull without lower jaw (with parts of the postcranial skeleton associated). Uitkyk, Beaufort West. Low T apinocephalus zone. Coll. le Roux. S.A.M. 11998 (fig. 5). Weathered piece of skull-roof. Sandriver, Beaufort West. Low? Tapinocephalus zone. Coll. Jack. Fic. 4. Tapinocephalus atherstonet. Ventral view of skull. x %. S.A.M. 2344. PHOCOSAURUS Skull Characters of the Genus 1. The preorbital length is 54% of the total median length as reconstructed; the snout as reconstructed is thus fairly short (preorbital length 282? mm.) and it is fairly narrow (222 mm.) and fairly high (168 mm.) with the height 76% of the width; the facial surface of the snout is not very short and extends to the plane of the anterior orbital border; the transition from the facial to the cranial surface is not abrupt but slopes up evenly. 2. The dorsal cranial bones are strongly pachyostosed, with the centres of thickening coalesced and the transition on to the face is by a slope along an 156 ANNALS OF THE SOUTH AFRICAN MUSEUM indistinct indented line; the cranial thickening does not extend very far forward —just reaching the plane of the anterior orbital border; along the median line the pachyostotic cranial roof is 267 mm. in length and this is 51% of the total median length; this small anterior extension of the thickening is due to the fact that the naso-frontal swelling is situated posterior to the plane of the anterior e) 140 bh, 11 A, 142 B. FAL 42 U, 66 E, 78 L, ros K, eee 160 U, 176 W, 274 P, 288 F. Description. Very common in intertidal region and shallow water, forming a short feathery growth reaching a maximum height of 2-5 cm. Many minor variations occur. In some colonies all the stem internodes bear one hydro- cladium each, in others the lower stem internodes may bear as many as 4 hydrocladia. Very occasionally in young colonies hydrocladia may arise separately from the hydrorhiza. On the hydrocladia athecate internodes are occasionally present between 2 thecate ones; and sometimes an extra I or 2 athecate internodes may be present between the basal athecate internode and the first thecate one. These may be devoid of nematothecae, or may bear 1 or 2 each. Gonophores very variable, younger ones normally smooth, older ones ribbed, with or without spines. Borne on stem or on hydrorhiza. Present in February, April to July, and September to October. Remarks. Kirchenpaueria unilateralis (Ritchie 1907a) is considered to be a synonym for A. pinnata. ‘This conclusion was reached after a careful study of type material kindly loaned by the Royal Scottish Museum, Edinburgh. Its distinguishing features are said to be: (i) The stem internodes never bear more than one hydrocladium each. But this is the normal condition in many stems of A. pinnata. Further, careful study of the type material of A. unilateralis shows that at least two of the stem internodes bear two hydrocladia each. (ii) ‘The two rows of hydrocladia are borne on one side of the stem, and so lie in two planes. This unilateral arrangement is obviously only a growth- form of A. pinnata, and in False Bay is often found on stems growing in the same colony as the normal flattened type. (ii) The hydrocladia occasionally branch. Branching hydrocladia have never again been observed in all the wealth of material examined, neither in the normal flat stems, nor in the unilateral type. If A. unilateralis were 234 ANNALS OF THE SOUTH AFRICAN MUSEUM: retained as a separate species, this would be its only distinguishing feature. I am inclined to regard it, however, as an abnormal condition and possibly due to injury. In Ritchie’s material the internodes of the stem and hydrocladia are rather shorter, and the hydrothecae more crowded than in the False Bay material (hydrotheca length/internode length = 0-40-0-59 mm.), but not sufficiently so for the maintenance of a separate variety. Broch gives the proportions for var. typica as § to 4. All the False Bay material falls well within the range of var. typica. Pycnotheca mirabilis (Allman) 1883 Diplocheilus mirabilis Allman 1883, p. 49; Pl. VIII, figs. 4-7. Kirchenpaueria mirabilis Bale 1894. Pycnotheca mirabilis ‘Totton 1930, p. 216, fig. 55. Records. FAL 78 U, 108 J. Description. A fair number of unbranched stems, reaching a maximum height of 3 cm. Basal 2-3 mm. of stem devoid of hydrocladia and containing I—4 transverse nodes. Remainder with 2 hydrocladia per internode in lower part, and one in distal part. Gonophores present in January, as described by Bale, borne on hydrorhiza and on basal part of stem replacing hydrocladia. Measurements (mm.) False Bay Richard’s Bay Hydrocladium, internode length oe 2 O°94=0r42 O°32-0- a5 Hydrotheca, height .. Se bs .. 0°25-0°30 0:20—-0'21 diameter at margin ih 5.) 0722-0727 0-18-0'19 adcauline side, length of free oe 22) 0-06_O712 0:06-0:08 Gonotheca, length ae : a .. 2:12-9-G6 maximum diameter oe a .. 1°02—1-:06 Remarks. ‘The material from False Bay agrees well both in size and structure of the hydrotheca with the typical form of the species as figured by Totton. In the material previously reported from Richard’s Bay, Natal (Millard and Harrison 1954) as Airchenpaueria mirabilis, the hydrothecae are smaller, and the opening into the medial nematotheca is in an oblique position rather than vertical. This material must thus be referred to var. warren. ‘The measurements of the two forms are given for comparison. Nemertesia cymodocea (Busk) 1851 Antennularia decussata Kirchenpauer 1876, p. 52; Pl. II, fig. 24; Pl. III, fig. 24; Pl. Vil; fig. 24. Antennularia hartlaubi Ritchie 1907a, p. 542; Pl. III, fig. 4. Records. PF 337 A. FB 112, 115 A, 129 A, 191 A, 132,.134 Ay 130"geaO B. FAL 95 R, 209 F, 211 F, 217 K, 222 E, 268 F, 274 N, 279 G, 290 A. THE HYDROZOA OF FALSE BAY, SOUTH AFRICA 235 Description. Massive colonies abundant in shallow water and reaching a maximum height of 49 cm. The decussate arrangement, with the hydrocladia borne in alternate pairs, is found in the lower parts of the branches, and the hexastichous arrangement near the ends, i.e. just the opposite of the arrange- ment described by Marktanner-Turneretscher. In some colonies only the decussate arrangement is present, and in others only the hexastichous type. Occasionally the small branches arising near the base of the colony have only one hydrocladium to an internode, and these are arranged alternately, the two rows being in one plane. One or two athecate internodes may be present at the base of each hydrocladium. The hydrocladium-bearing apophysis bears one pair of bithalamic nematophores, and a ‘mamelon’ more or less in the centre of the upper surface. | Gonophores include the normal elongated type, and young ones which are shorter and truncated at the ends. Present in February, September and November. é; Nemertesia ramosa Lamouroux 1816 Antennularia ramosa Hincks 1868, p. 282; Pl. LXII. Nemertesia ramosa Broch 1933, p. 38. Records. PF 351 A. FAL 207 G. Description. ‘Two specimens of unbranched fragments with a maximum length of 6-7 cm., some of them fascicled in lower region. Hydrocladia in whorls of 3 in upper parts, but becoming irregular lower down. Hydrocladial apophyses stouter than usually figured, but narrowing to distal ends where the diameter is the same as that of the hydrocladia. Athecate internodes not limited to distal ends of hydrocladia, but irregularly scattered in any region. One specimen possesses male gonophores and the other female, and these are slightly different in appearance. The young male gonophore is elongated, with a rounded end, and is quite symmetrical. Older ones have open truncated ends which are sometimes at right angles to the stem and sometimes placed slightly oblique, facing towards the stem. Female gonophores always have the opening on the side facing the stem and practically parallel to it. They contain several large eggs or embryos. Both sexes present in September. Antennopsis scotiae Ritchie 1907a Antennopsis scotiae Ritchie 1907 a, p. 543; Pl. III, fig. 3. 1909, p. go, fig. 8. Records. PF 15675 A. FAL 64 M, 207 H, 211 G, 217 L, 222 G, 230 N, 291 A. Description. Colonies reaching a maximum height of 13:7 cm. Gonophores present in March and September. Aglaophenia pluma (Linnaeus) 1758 Aglaophenia pluma Hincks 1868, p. 286; Pl. LXIII, fig. 1. Broch 1933, p. 44. General Discussion. A number of species of Aglaophena have been recorded from South Africa, all closely related to A. pluma, namely A. dichotoma (M. Sars), 236 ANNALS OF THE SOUTH AFRICAN MUSEUM A. parvula Bale, A. conferta Kirchenpauer, A. heterodonta Jaderholm and A. chalerocarpa Allman. Of these A. parvula, A. conferta and A. heterodonta are accepted as synonyms by Ritchie 1909, Bedot 1921 and Stechow 1925. A. chalerocarpa is regarded as a synonym for A. pluma by Bedot 1921, Stechow 1925 and Vervoort 1946. This leaves three South African species: (i) A. pluma var. typica. (ii) A. dichotoma. (iii) A. parvula. A. dichotoma is distinguished from A. pluma by its dichotomous branching, and A. parvula by the double nature of the second paired tooth of the hydro- theca. However, there is some confusion about this distinction, for Ritchie 1909 includes both dichotomous and unbranched forms under A. heterodonia, saying that the minute characters are identical. Leloup 1937a describes a dichotomously branched form with the double hydrothecal tooth as A. parvula but suggests that the latter may be a variety of A. dichotoma. Bedot 1921 suggests that A. parvula is a variety of A. pluma. Examination of a large series of specimens in the University collection suggests that the three forms should be regarded as varieties of the same species, for neither of the distinguishing features is clear-cut and there are many inter- mediate types. The ‘parvula’ form can, and does, branch dichotomously, although its stems generally form a short, dense turf in shallow water and intertidal regions. The ‘dichotoma’ form, with its long-drawn-out, dichoto- mously branched stems is more typical of deeper water and occurs abundantly in dredgings round the coast. But all intermediate stages occur. The double point of the second hydrothecal tooth is most pronounced in the ‘parvula’ form (fig. 15), but by no means constant, the second point may be only a hump on the side of the tooth, or the tooth may be single-pointed and triangular. All variations may occur in the same colony, but in general the second point is better developed in the lower parts of the hydrocladia. A similar variation occurs in the ‘dichotoma’ form, where the second hydrothecal tooth is usually triangular, but may have an accessory point on one side. Only seldom does the “pluma’ type occur, where there is no sign of a double-pointed tooth. The corbulae are sexually dimorphic and are identical in all three forms. Nor could any difference, other than minor variations in length and breadth, be found in the internodes of the stem or hydrocladia. The length of the median nematotheca is variable in each form, but in general is longer in the ‘parvula’ form than in the ‘pluma’ and “dichotoma’ forms. For convenience a brief description of the composite species as now envisaged follows. Description. Stem unfascicled, branched or unbranched, distinctly segmented. The basal part of the stem, which is of variable length and often very short, is THE HYDROZOA OF FALSE BAY, SOUTH AFRICA 237 devoid of hydrocladia and nematothecae, and divided into internodes of irregular length by transverse nodes. ‘This region is terminated by 2 oblique hinge-joints (occasionally 3) with one extra large nematotheca between them. The rest of the stem is divided by slightly oblique nodes into regular internodes each bearing a hydrocladium. The first internode of this region is longer than the others and bears 2 hydrocladia. The stem may branch in two ways, and both types may be present in the same colony: (i) The hydrocladia-bearing part may divide dichotomously, the two branches being of equal diameter and arising from apophyses on a broad stem internode, which also bears a hydrocladium. The internodes of the branches all bear hydrocladia. (ii) Subsidiary branches may be given off from the stem; these are always short and usually of smaller diameter. They do not arise dichotomously, but replace hydrocladia. On these branches the first few internodes are without hydrocladia, then follow 2 oblique hinge-joints and then hydro- cladia-bearing internodes, exactly as on the base of the stem. Hydrocladia alternately arranged, with the two rows directed towards one side of stem, divided into internodes by transverse or slightly oblique nodes. Each internode with 2 internodal septa and 1 hydrotheca. Hydrotheca with 9 marginal teeth, including 1 median tooth on abcauline side directed inwards, and 4 pairs of lateral teeth, of which the second may be double. Median nematotheca narrow at base, but widening distally, adnate for most of length and then free, reaching to hydrothecal margin or just below. Openings 2 (1 into hydrotheca, 1 extending along free adcauline edge on to distal end) or 3 (1 into hydrotheca, 1 on adcauline surface at base of free part, 1 on distal ' end). Supracalycine nematothecae swollen, reaching to hydrothecal margin or just below. Cauline nematothecae 4, of which one is on hydrocladial apophysis. Corbula replacing hydrocladium, with one hydrotheca between it and stem. Female corbula closed, except for first rib, with 4-9 ribs on each side, each with about 10 nematothecae along one edge. First rib free at base, bearing about 6 nematothecae on each edge, and a free lateral branch on one side running parallel to the axis of the corbula and bearing about 6 nemato- thecae on each edge. The side of the free branch corresponds to the side of the corbula, i.e. when the corbula replaces a right-hand hydrocladium the free branch is on its right. Very occasionally female corbulae may lack the free lateral branch, though present in others of the same colony. Male corbula partly or completely open, with 5-9 ribs on each side. Ribs usually fused in basal part (except for first rib which is free), and bearing nematothecae along one edge, free in distal part and bearing nematothecae along both edges. Old empty corbulae often with ribs free for entire length. 238 ANNALS OF THE SOUTH AFRICAN MUSEUM Normally no lateral branch on first rib, but very occasionally a small stunted one present. Male and female corbulae on separate stems. var. typica Bedot 1919 Fig. 15 A ? Aglaophenia chalerocarpa Allman 1886, p. 150; P. XXI, figs. 1-4. Aglaophenia pluma forma typica Broch 1933, p. 44, fig. 18a. Aglaophema pluma Vervoort 1946a, p. 335, fig. 8. Records. FB 133 A. FAL 13 CG, 20 R, 26 H. Description. Stems short, usually about 5 cm., generally unbranched. Second paired tooth of hydrotheca triangular or truncated, but not double. Median nematotheca not reaching to hydrothecal margin. Stems dark brown, hydro- cladia lighter. Ripe corbulae in March. LESSEE SE NE eA IN Se CO OLS YO aiLarIs Fic. 15. Aglaophenia pluma (Linnaeus). A. var. typica Bedot, FB 133 A. B and C. var. dichotoma (M. Sars), FB 116. In B the second paired tooth is single and triangular, in C it has a subsidiary point. D and E. var. paroula Bale (not from False Bay material). Second paired tooth bifid. F. var. parvula Bale, FAL 62 L. Subsidiary point of second paired tooth almost obsolete. THE HYDROZOA OF FALSE BAY, SOUTH AFRICA 239 Measurements (mm.) Stem internode, length A Hf at of .. 0°38-0°41 Hydrocladial internode, length .. at oe o/-))) 10:20-0-38 Hydrotheca, length bie a ee .. 0'27-0°34 length above median Henlatornees ae Me .. O°09-0°12 width at margin ANG ay 8 .. Ot17-0'21 Remarks. Allman’s specimens of A. chalerocarpa thank probably be included here. var. dichotoma (M. Sars) 1857 Fig. 15 B and C Aglaophenia pluma var. B Hincks 1868, p. 287. Aglaophenia dichotoma Kirchenpauer 1872, p. 30; Pl. I, fig. 7; Pl. II, fig. 7; Pl. III, fig. 7. Ritchie 1907a, p. 539; Pl. III, fig. 2. Vervoort 1946a, Pp. 338. Aglaophenia tubiformis Marktanner-Turneretscher 1890, p. 269; Pl. VII, figs. 4, iy Ue Aglaophenia heterodonta Ritchie 1909 (pp.), p. 96. Aglaophenia dichotoma f. typica Broch 1933, p. 54, fig. 18b. ecanram Ol 2960). HB 110,111, 116, 117, 120 A; 121 A, 127, 128) 146 A, IV OVAL, 200 Ef, 217 J. Description. Stems elongated and branching profusely in a dichotomous manner, though diameter remains the same; often reaching 20-30 cm. in length. Second paired tooth of hydrotheca usually single-pointed and tri- angular, but often with a subsidiary point on one side. Median nematotheca not reaching hydrothecal margin. Colour somewhat lighter than var. typica, with hydrocladia cream. Ripe corbulae in February, September and November. Measurements (mm.) Stem internode, length a me ue ae ty) 0:37-0:50 Hydrocladial internode, length .. ste us -. 0°29-0°37 Hydrotheca, length . oe be ny ». 0°29—-0°35 length above i enit nemiatotiees a he Ae! 1 0;00- On 0 width at margin om ay a i .. 0°18—0°25 Corbula, male, length A A ie: He 22) 1 O0—3°53 wane lay 9 e as Ay vo sit Ne .. 0°84-1-25 Corbula, female, length .. ay 1M ob ai 2248 B02 WAGIEI iy 2). na a at 6 aie -., 1:02-2:28 var. parvula Bale 1882 Fig. 15, D-F ?Aglaophenia conferta Kirchenpauer 1872, p. 32; Pl. I, fig. 4; Pl. II, fig. 4; a JOD, sites 21s 240 ANNALS OF THE SOUTH AFRICAN MUSEUM Aglaophena parvula Bale 1882, p. 23; Pl. XIV, fig. 3. 1884, p. 165; Pl XIV, fig. 3; Pl. XVII, fig. 10. Stechow 1925, p. 516. Vervoort 1946a, p. 339, fig. gb. Aglaophenia heterodonta Jaderholm 1903, p. 296; Pl. XIII, figs. 10-12; Pl. XIV, fig: 1. Ritchie 1909 (pp:),) p. 96. Records. F 49, 253 (reported by Eyre 1939 as A. dichotoma). CP 18, 356 D, 392 P, 410 X, 415 X. FB 119 A, 130. FAL 62 L, 66 G, 78 OQ, 125 V, 160 7, 176 X, 262 A, 279 G. Description. Stems short and clustered, usually unbranched, but sometimes branching by either of the methods described above, seldom more than one branch to a stem. Length usually 3-5 cm., maximum length recorded 11-6 cm. Second paired tooth of hydrotheca usually double, though sometimes one point is much smaller than the other, and occasionally the tooth is single- pointed and triangular. Median nematotheca usually reaching hydrothecal margin. Colour darker than var. dichotoma and more like var. typica. Ripe corbulae in February, and April to September. Measurements (mm.) Stem internode, length ai oh De ak -. 0°26-0°55 Hydrocladial internode, length .. eis ae ». 0°25-0°29 Hydrotheca, length . se a ae ». (0°25—079i length above eaclian nematoriees ae on .. 0°06-0-10 width at margin e a apt hs - > | OF On2e Corbula, male, length oe ok se ae «187-2 width .. At aM Ay: a a .. 0°96-1°07 Corbula, female, length .. Be dhe Hes .. 1:69-1:98 width .. ne bi as ae a .. I°00—1°05 Thecocarpus giardi Billard 1907 Thecocarpus Giardi Billard 1907, p. 381, fig. 21; Pl. XXV, fig. 9; Pl. XXVI, figs. 11-16. Thecocarpus giardi Vervoort 1946a, p. 335. Records. PF 306 A. Description. Colony reaching 23:8 cm. in height. Basal part of stem thick and woody with a maximum diameter of 15 mm., and branching irregularly. At about 12-15 cm. from the base each branch assumes the spirally twisted form described by Billard and gives off alternate pinnae which bear the hydrocladia. Hydrocladial internodes with two well-developed internodal septa, the lower one continued as an intrathecal septum across the hydrotheca about a third from its base, the upper one opposite the base of the lateral nematotheca. The lateral hydrothecal teeth are seldom bifurcated as in many of Billard’s specimens, with the exception of the pair nearest the hydrocladium which is THE HYDROZOA OF FALSE BAY, SOUTH AFRICA 241 invariably double, so that one could more accurately describe the hydrotheca as having 4 pairs of lateral teeth, and 1 median tooth of variable length. Corbulae as described by Billard, present in October. Lytocarpus filamentosus (Lamarck) 1816 Lytocarpus filamentosus Billard 1907, p. 371; figs. 15-17. FHlalicornaria segmentata Warren 1908, p. 328; Pl. XLVIII, figs. 33-36. Records. GP 224. PF 18293 A. FAL 13 B, 42 V, 52 W, 58 W, 66 A, 78 P. Description. Colonies reaching a maximum height of 9.9 cm. The gono- cladium varies in length, possessing only 3 or 4 segments to the nematocladium in some specimens and as many as 10 in others. Gonophores present in October and December. Measurements (mm.) Hydrocladial internode, length .. bid ae .. 0°22-0°26 Hydrotheca, length, to tip of median tooth oh -. 0°25-0:28 width at margin bse as i : .. O*1I-0'16 Median nematotheca, length abcauline .. i .. 0°24-0°30 length adcauline, free part .. ay a, .. 0°08-0°14 REFERENCES Allman, G. J., 1876. ‘Diagnoses of new genera and species of Hydroida.’ 7. Linn. Soc. Lond. (Zool.), vol. XII, pp. 251-284. Allman, G. J., 1883. ‘Report on the Hydroida dredged by H.M.S. Challenger during the years 1873-76. Part I. Plumularidae.’ Rep. Sci. Res. H.M.S. Challenger, Zool., vol. VII, pp. 1-54. Allman, G. J., 1886. ‘Description of Australian, Cape and other Hydroida, mostly new, from the collection of Miss H. Gatty.’ 7. Linn. Soc. Lond., Zool., vol. XIX, pp. 132-161. Allman, G. J., 1888. ‘Report on the Hydroida dredged by H.M.S. Challenger during the years 1873-76. Part II. The Tubularinae, Corymorphinae, Campanularinae, Sertularinae and Thalamophora.’ Rep. Sci. Res. H.M.S. Challenger, Zool., vol XXIII, pp. 1-87. Bale, W. M., 1882. ‘On the Hydroida of South-Eastern Australia, with descriptions of supposed new species, and notes on the genus Aglaophenia.’ 7. Micr. Soc. Victoria, vol. II, pp. 15-48. Bale, W. M., 1884. ‘Catalogue of the Australian Hydroid Zoophytes.’ Austr. Mus. Sydney, pp. 1-198. Bale, W. M., 1888. “On some new and rare Hydroida in the Australian Museum Collection.’ Proc. Linn. Soc. N.S. Wales, ser. 2, vol. III, pp. 745-799. Bale, W. M., 1894. ‘Further notes on Australian Hydroids, with descriptions of some new species.” Proc. Roy. Soc. Victoria (n.s.), vol. VI, pp. 93-117. Bale, W. M., 1913. ‘Further notes on Australian Hydroids. II.’ Proc. Roy. Soc. Victoria (n.s.), vol. XXVI, pp. 114-147. Bale, W. M., 1914. ‘Report on the Hydroida collected in the Great Australian Bight and other Localities. Part I.’ Biol. Res. ‘Endeavour’, 1909-1914, vol. II, pp. 3-62. Bale, W. M., 1915. “Report on the Hydroida collected in the Great Australian Bight and other Localities. Part III.’ Biol. Res. ‘Endeavour’, 1909-1914, vol. III, pp. 241-336. Bale, W. M., 1924. ‘Report on some Hydroids from the New Zealand coast, with notes on New Zealand Hydroida generally, supplementing Farquhar’s list.’ Trans. Proc. N.&. Instit. vol. LV, pp. 225-268. Bedot, M., 1921. ‘Notes systématiques sur les Plumularides.’ Rev. Suisse Zool., vol. 28, pp. 311-356. Billard, A., 1906b. ‘Hydroides.’ Expéd. Sci. ‘Travailleur’ et du ‘Talisman’, vol. VIII, pp. 153-241. Billard, A., 1907. ‘Hydroides de Madagascar et du Sud-Est de I’Afrique.’ Arch. Zool. exp. gén., ser. 4, vol. VII, pp. 335-396. 24.2 ANNALS OF THE SOUTH AFRICAN MUSEUM Billard, A., 1925. “Les Hydroides de l’expédition du Siboga. II. Synthecidae et Sertularidae.’ Siboga-Expeditie, monogr. VII b, pp. 1-116. Broch, H., 1914. “Hydrozoa benthonica.’ Michaelsen’s Beitrége zur Kenntnis der Meeresfauna Westafrikas, vol. I, pp. 19-50. Broch, H., 1918. ‘Hydroida II.’ Danish Ingolf Exped., vol. V, pp. 1-205. Broch, H., 1933. “Zur Kenntnis der adriatischen Hydroidenfauna von Split.’ Norske Videns.- Akad. Oslo, I. Mat.-Naturv. Klasse, no. 4, pp- 1-115. Eyre, J., 1939. “The South African Intertidal Zone and its relation to Ocean Currents. VII. An Area in False Bay.’ Ann. Natal Mus., vol. IX, pp. 283-306. Fraser, C. McL., 1944. ‘Hydroids of the Atlantic Coast of North America.’ Toronto. Hartlaub, C., 1905. “Die Hydroiden der magalhaensischen Region und chilenischen Kiiste.’ Fauna Chilensis, vol. III, pp. 497-714. Hincks, T., 1868. ‘A History of the British Hydroid Zoophytes.’ London. Jaderholm, E., 1903. ‘Aussereuropaische Hydroiden im schwedischen Reichsmuseum.’ Ark. Xool., vol. I, pp. 259-312. Jaderholm, E., 1904. ‘Hydroiden aus den Kiisten von Chile.’ Ark. Zool., vol. II, pp. 1-7. Kirchenpauer, G. H., 1872. ‘Ueber die Hydroidenfamilie Plumularidae, einzelne Grupp derselben und ihre Fruchtbehalter, I. Aglaophenia.’ AbA. Naturw., Hamburg, vol. V, pp. 1-52. Kirchenpauer, G. H., 1876. ‘Ueber die Hydroidenfamilie Plumularidae, einzelne Gruppen derselben und ihre Fruchtbehalter, II. Plumularia und Nemertesia.’ Abh. Naturw. Hamburg, vol. VI, pp. 1-59. Leloup, E., 1935. “Hydraires Calyptoblastiques des Indes Occidentales.? Mém. Mus. Roy. d Hist. Nat. Belgique, ser. 2, fasc. 2, pp. 1-73. Leloup, E., 1937a. ‘Résultats Scientifiques des Croisiéres du Navire-Ecole Belge ‘‘Mercator’’. Hydroidea, Siphonophora, Ceriantharia: Hydropolypes.’ Mém. Mus. Roy. d’Hist. Nat. Belgique, ser. 2, fasc. 9, pp. 91-121. Manton, S. M., 1940. ‘On two new species of the Hydroid Myriothela.’ Sci. Rep. Brit. Graham Land Exped. 1934-1937, vol. I, pp. 255-294. Marktanner-Turneretscher, G., 1890. ‘Die Hydroiden des k.k. naturhistorischen Hofmuseums.’ Ann. k.k. naturh. Hofmus., vol. V, pp. 195-286. Millard, N. A. H. and Harrison, A. D., 1954. “The Ecology of South African Estuaries. Part V: Richard’s Bay.’ Trans. Roy. Soc. S. Afr., vol. XXXIV, pp. 157-179. Millard, N. A. H., 1955. “New species of Hydrozoa from South Africa.’ Ann. S. Afr. Mus., vol. XLI, pp. 215-222. Nutting, C. C., 1900. ‘American Hydroids. Part I. The Plumularidae.’ Spec. Bull. Smithsonian Inst., Washington, No. IV, pp. 1-285. Rees, W. J., 1938. “Observations on British and Norwegian Hydroids and their Medusae.’ J. Mar. Biol. Ass. U.K., vol. XXIII, pp. 1-42. Rees, W. J., 1946. “A Revision of the Hydroid Genus Perigonimus M. Sars, 1846.’ Bull. Brit. Mus. (Nat. Hist.), Zool., vol. III, pp. 337-362. Ritchie, J., 1907. ‘On Collections of the Gape Verde Islands Marine Fauna, made by Cyril Crossland, M.A. (Cantab.) B.Sc. (Lond.), F.Z.S., of St. Andrews University, July to September, 1904.’ Proc. Zool. Soc. Lond., pp. 488-514. Ritchie, J., 1907a. “The Hydroids of the Scottish National Antarctic Expedition.’ Trans. Roy. Soc. Edinburgh, vol. XLV, pp. 519-545. Ritchie, J., 1909. ‘Supplementary Report on the Hydroids of the Scottish National Antarctic Expedition.’ Trans. Roy. Soc. Edinburgh, vol. XLVII, pp. 65-101. Ritchie, J., 1911. “‘Hydrozoa (Hydroid Zoophytes and Stylasterina) of the “Thetis’ Expedition.’ Mem. Austr. Mus. Sydney, vol. 1V, pp. 807-869. Stechow, E., 1913a. “Hydroidpolypen der japanischen Ostkiiste. Teil II.’ Abh. Bayer Akad. Wiss., Math.-phys. Kl., supple. vol. III, pp. 1-162. Stechow, E., 1919a. ‘Zur Kenntnis der Hydroidenfauna des Mittelmeeres, Amerikas und anderer Gebiete.’ ool. Jahrb. Jena, vol. XLII, pp. 1-172. Stechow, E., 1923b. “Ueber Hydroiden der Deutschen Tiefsee-Expedition, nebst Bemerkungen uber einige andre Formen.’ ool. Anz. Leipzig., vol. LVI, pp. 97-119. Stechow, E., 1923c. ‘Zur Kenntnis der Hydroidenfauna des Mittelmeeres, Amerikas und anderer Gebiete. II Teil.’ Zool. Jahrb. Jena, vol. XLVII, pp. 29-270. — THE HYDROZOA OF FALSE BAY, SOUTH AFRICA 243 Stechow, E., 1925. ‘Hydroiden der Deutschen Tiefsee-Expedition.’ Wissenschaftliche Ergebnisse der Deutschen Tiefsee-Expedition auf dem Dampfer ‘Valdivia’ 1898-1899, vol. XVII, pp. 383-546. Stechow, E., 1932. ‘Neue Hydroiden aus dem Mittelmeer und dem Pazifischen Ozean, nebst Bemerkungen iiber einige wenig bekannte Formen.’ ool. Anz. Leipzig., vol. C, pp. 81-92. Totton, A. K., 1930. ‘Coelenterata. Part V.i—Hydroida.’ Brit. Antarct. (‘Terra Nova’) Expedition 1910. Nat. Hist. Rep., Xool., vol. V, pp. 131-252. Trebilcock, R. E., 1928. ‘Notes on New Zealand Hydroida.’ Proc. Roy. Soc. Victoria (n.s.), vol. XLI, pp. 1-31. Vanhoffen, E., 1910. ‘Die Hydroiden der Deutschen Stidpolar-Expedition, 1901-1903.’ Deutschen Stidpolar-Exped. 1901-1903, vol. XI, Zool., pp. 269-340. Vervoort, W., 1946. ‘Fauna van Nederland. Vol. XIV. Hydrozoa (C 1), A. Hydropolypen.’ Leiden. Vervoort, W., 1946a. ‘Exotic Hydroids in the Collections of the Rijksmuseum van Natuurlijke Historie and the Zoological Museum at Amsterdam.’ ool. Meded., vol. XX VI, pp. 287-351. Vervoort, W., 1946b. “‘Hydroiden uit de Waddenzee.’ Arch. Néerl. Zool., vol. VII, pp. 334-352. Warren, E., 1906a. ‘On Tubularia solitaria sp. n., a Hydroid from the Natal Coast.’ Ann. Natal Mus., vol. I, pp. 83-96. Warren, E., 1908. “On a Collection of Hydroids, mostly from the Natal Coast.’ Ann. Natal Mus., vol. I, pp. 269-355. Warren, E., 1909. ‘On Lafoea dispolians sp. n., a Hydroid parasitic on Sertularia bidens, Bale.’ Ann. Natal Mus., vol. II, pp. 105-112. tortie tie a HO hey as ee PVG RNIN oO 1 TED MOO es Biv by : tance Li Tala Selita! ADR AB 4 F i i = é , i gb ' ; Mae Pee se a coeltigat mise i i Ay a ‘* “ i af ; i ey ta’) is Go ne eae ya aie tee, Nb The ANNALS OF THE SOUTH AFRICAN MUSEUM are ree n parts. intervals as material becomes available. Ke + tenn Some volumes and parts are out of print, and sikeis are only pals as parts ofa set, or 0 respectively. The prices of parts published prior to 1940 have been increased. Out of print: Vols. I, II (Parts 1-3, 5, 7, 8), Ill (Part 1), V Index), VII (Parts 1, 2, 3, 5), VIII, 1X (Parts 1, 2), x (Part 2), XI er 2, e ee XxX (Part 1) >» XXIV (Part 2), XXXI (Parts 1-3). Vol. “3 ie haa, aa II. 1900-1902 Zoology and Gaul, (excl. Parts Hi rere 7,8) x } III. 1903-1905 Zoology - «(exh Part. ae ae ee IV. 19031908 Palaeontology i oan V. 1906-1910 Geology, Palaeontology, Zoology, Anthropology (excl. raat Parts 1, 2,.5, 8, 9) g cae VI. 1908-1910 Zoology ieee Pace ic Tadeo ere VII. 1908-1913 Palaeontology . (Parts 4 and 6 only) | ; IX. 1911-1918 Botany ent Parts1 and2) 2 X. &Ig1ti1-1914 Zoology ae ; . (exer Part2) . XI. 1911-1918 Zoology tk lexal Parts 2, 7s indeele wee XII. 1913-1924 Palaeontology and ‘Geolbay ve eae 2 a nen XIII. 1913-1923 Seeley and areas LA ee XIV. 1915-1924 Zoology j 3 XV. 1914-1916 Zoology .. Sheen Laced se ae Je wa igs XVI. 1917-1933 Botany .. 4 te Re i XVIL 1919-1920, Zodlagy = 3.4.3. ae ae oo) Dr XV. 1921 Zoology .. man « os haa wf hd ac eae XIX. 1924-1925 Zoology .. PR ak teh ef au Rear XX. 1924-1926 Zoology .. ol ee XXI. 1925-1927 Zoology . (excl. Bart 1) he XXII. 1925-1928 Palachutolney ee te ee XXIII. 1925-1926 Zoology .. wie Aig SN arena XXIV. 1929-1938 Anthropology and Ethnology 0°25 2 52-6 9°64 53:0 10°12 O-000n | |= -0-25, 3 4°9 0°89 5°5 0-90 ETS Fit hi 25 4 8-4. I-14 8°3 1-06 0-156 >0°25 5 0°387 0-021 0°36 0-029 2°638 0-02 6 49:0 1°03 48-9 IO 0°234 >0°25 7 18°7 0°79 18-3 O67 1.) O-5R5 >0:25 8 12°25 0-69 Iovl 0:88 6-666 <0°001 9 30-0 oy (IC: 1°34 4°331 <0°001 A statistical analysis showing to what extent the differences between the Peruvian and the Galapagos samples of B. elongatum may be held to characterize the populations from which the samples were taken. It is seen from Table I that the differences between the samples of B. elongatum from Peru and from the Galapagos Islands are not significant (assuming ‘P’ =0-05 to be the limit of significance) except in characters 5, 8 and g. In character 5, the relative length of the postatrioporal region, the difference, although significant, is not great, but in character 8, the number of myotomes in the tail, the difference is highly significant. In accordance with the greater number of myotomes in the tail in the Peruvian lancelet there is also a significant difference in the total number of myotomes (character g) in the two samples. Ignoring, therefore, the rather slight difference in the relative lengths of the postatrioporal region, it is clear that the only important feature distinguishing the lancelets of Lobos de Tierra and those of the Galapagos is the number of myotomes in the tail. The difference in this character is equivalent to the population differences found in B. lanceolatum from different regions (Webb, 1956b) and is insufficient to justify the recognition of the Galapagos lancelets as a distinct species. The comparison shows that the sample of B. elongatum from Peru can be considered representative of the species and, as it contains a number of mature specimens, it is more suitable for comparison with the Cape lancelet than the Galapagos form. A statistical analysis of the Peruvian sample of B. elongatum and the sample of B. capense is given in Table II. 258 ANNALS OF THE SOUTH AFRICAN MUSEUM TABLE II B. capense B. elongatum CHARACTER (Peruvian sample) 7 te i NUMBER MEAN S.D. MEAN 5.) I 4.20°9 14°67 381-0 18°50 5766 <0-001 2 71-2 6-11 64:0 10°35 1°904 0:05 3 4°55 0-64. 3°6 1°03 2-616 0:02 4. Q°4 1°36 10:0 I-41 1-061 0°25 5 0°405 0-014 0°387 0-021 2°204. 0-02 6 46-0 1°26 49:0 1°03 6-524 <0-001 Gi 1Q°2 0.75 18-7 0°79 1-587 Or! 8 10°1 0°32 12°25 0-69 9282 <0-‘001 9 753 1°45 80:0 1°57 7.586 <0-001 A statistical analysis showing to what extent the differences between B. capense and the _ Peruvian sample of B. elongatum may be held to characterize the populations from which the samples were taken. In Table II the Means and Standard Deviations for each of the nine diagnostic characters of B. capense and the Peruvian sample of B. elongatum are given and from them ‘/’ and hence ‘P’ have been calculated to show how far the differences between the samples are significant. The chief characters in which the Cape lancelet differs from B. elongatum are the number of dorsal fin chambers (character 1), the number of myotomes before the atriopore (character 6) and the total number of myotomes (character 9) for each of which the value of ‘?’ is high and that for ‘P’ correspondingly low. Nevertheless in all of these characters the range of variation is such that there is some degree of overlap between the Cape sample and either the Peruvian or the Galapagos samples. Thus, on these characters alone, individual specimens of the Cape lancelet cannot necessarily be separated from specimens from one or other of the populations of B. elongatum. In the number of preanal fin chambers the Cape lancelet, with a rather higher number, is different from the Peruvian lancelet, but the difference is barely significant. In character 3, the proportions of the dorsal fin chambers, the Cape lancelet is not significantly different from the small specimens of the Peruvian lancelet although it is different from the mature forms. The height of the dorsal fin relative to the depth of the body (character 4) is a very variable character in all forms and differences between them are not significant. A difference in the relative length of the postatrioporal region exists in the two forms, but again it is ON THE LANCELETS OF SOUTH AND EAST AFRICA 259 barely significant. In character 7, the number of myotomes between the atriopore and the anus, there is a close similarity between the Cape lancelet and the Peruvian and Galapagos lancelets. In character 8, the number of myotomes in the tail, the Cape lancelet is significantly different from the Peruvian lancelet, but not from the Galapagos form. Although, as would be expected, there is a greater difference between the Cape and the Peruvian than between the Peruvian and the Galapagos lance- lets, on numerical characteristics alone the differences between any of the samples are not such that individual specimens from each are readily distin- guishable. As the numerical characters do not provide a clear basis for separation, it is necessary to take into consideration other features not con- veniently assessed on a numerical basis. From a comparison of figures 1 and 2, it is seen that the form of the rostrum and the oral region, the dorsal fin and also the caudal and preanal fins is quite different in the two lancelets. In general appearance B. elongatum has resemblances with B. belchert whereas the Cape lancelet resembles the West African forms. Moreover, the changes in the proportions of the dorsal fin chambers, the relative height of the dorsal fin and the number of preanal fin chambers which occur in B. elongatum with advancing age are features not found in the Cape lancelet where the characters of large and small specimens do not differ significantly. ‘The writer attaches considerable importance to these characters and, in view of the fact that the two forms can be separated by them, is of the opinion that the Cape lancelet should be regarded as a distinct form in which the similarity of numerical characteristics with B. elongatum is fortuitous. The comparison of B. capense with the Indo-Pacific species B. belchert, B. bazarutense and B. arabiae can be dismissed very briefly. In each of these three species the number of dorsal fin chambers is much fewer and the height of the dorsal fin relatively less than in B. capense. In consequence, the dorsal fin chambers are short and broad in these species, but long and narrow in B. capense. In B. bazarutense and B. belcheri, but not in B. arabiae, the postatrioporal region is relatively longer than in B. capense. An outstanding difference, how- ever, lies in the number of myotomes before the atriopore which is much higher in B. capense than in the other three species. The difference in the number of anterior myotomes is also largely responsible for the clear difference in the total myotome number (see Webb, 1955 and 1956). Moreover the form of the rostrum, the oral region and the caudal and preanal fins in B. capense gives this lancelet a general appearance by which it can be distinguished immediately from the three Indo-Pacific species. It has already been pointed out that close relationships exist between B. belcheri, B. bazarutense and B. arabiae (see Webb, 1956) and it is evident that B. capense does not belong to this group. 260 ANNALS OF THE SOUTH AFRICAN MUSEUM Tasle III B. capense B. nigeriense CHARACTER —___,——______—_____ es x NUMBER MEAN S.D. MEAN S.D. I 4.20°9 14°67 34.43 15°30 II‘1Q5 <0-00I 2 71-2 6-11 53°9 3°07 7°548 <0:001 3 4°55 0-64 471 0-74 1421 0°25 4. Q°4 1°36 9°3 106 0-178 >0°25 5 0°405 0-014 0-417 0:030 1-194 0°25 6 46-0 1-26 41°7 1-70 6-298 <0-001 7 1Q°2 0°75 15°4 0°52 12°714 <0°001 8 1O°1 0°32 10°9 0°57 3°816 0-001 9 75:3 TAG | O79 1°45 II‘I1O <0-00I A statistical analysis showing to what extent the differences between samples of B. capense and B. nigeriense may be held to characterize the populations from which the samples were taken. The only other species with which B. capense can be compared with profit are the West African species B. africae, B. nigeriense, B. takoradiu and B. leonense which form a closely related group from the Guinea Coast (see Webb 1955, 1956a and 1956b). It is worth noting that, in describing B. africae, Hubbs (1927) compared this species with B. californiense, B. capense, B. lanceolatum and B. tatiersalli thus evidently recognizing a degree of relationship between B. capense and B. africae. Unfortunately specimens of B. africae from which a statistical analysis of the species can be prepared are not available. However B. nigeriense is a species very close to B. africae and for which there is an abun- dance of material. B. nigeriense occurs in two forms, one from the sea and the other from brackish water. An analysis of the marine form has already been published (Webb, 1956a) and these figures are here compared with those for B. capense in Table III. A detailed comparison of B. capense with B. takoradi and B. leonense is considered unnecessary in view of the close relationship existing between all the Guinea Coast lancelets. In Table III, for each of the characters numbered 1-9 in the diagnoses for B. capense and B. nigeriense, the Mean and the Standard Deviation is given and from these ‘t? and hence ‘P’ have been calculated as a measure of the extent of the differences between the samples. High values for ‘t? and cor- respondingly low values for ‘P’ show that B. capense and B. nigeriense are significantly different in all characters except the height of the dorsal fin, the proportions of the dorsal fin chambers and the relative length of the postatri- ON THE LANCELETS OF SOUTH AND EAST AFRICA 261 oporal region (characters 4, 3 and 5 respectively). In some characters (the number of preanal chambers, the number of myotomes before the anus and the total myotome number) the difference between the two samples is con- siderable. In the number of myotomes in the tail (character 8), however, it is not so great and in the number of dorsal fin chambers (character 1) there is a tendency among the Guinea Coast lancelets for this to be high, 400 chambers occasionally being exceeded in B. leonense. With regard to characters other than those treated numerically, it has already been mentioned that there is a striking similarity between the form of the rostrum, the oral region and the caudal fin in B. capense and that in B. nigertense and B. leonense. In general appearance B. capense can be mistaken for B. nigeriense except for a difference in number and length of the preanal fin chambers. It seems, therefore, that these two lancelets possess a considerable number of identical characters which strongly suggests that they have diverged from a common stock. Thus, whereas there is no doubt that B. capense is distinct from the Guinea Coast lancelets, it is also evident that it is an extreme member of that series and is more closely related to B. nigeriense or perhaps B. africae than to any other known species. Lancelets have not yet been recorded from the coasts of French Equatorial Africa, Angola, or South West Africa where forms intermediate between B. capense and the Tropical West African species might be expected to occur. The proposed relationship between B. capense and the West African lancelets is in accordance with their geographical location and supports the contention that B. capense has no close affinities with B. elongatum in spite of the remarkable agreement between their numerical characteristics. The view, therefore, is taken that B. capense is a valid species not to be confused with the lancelets of the western seaboard of the Americas. Branchiostoma belchert (Gray) Amphioxus belcheri Gray (1847) Branchiostoma belcheri Gray (1851) (for a full list of references see Franz, 1922, and Hubbs, 1922). Material examined. Nine specimens (MOR. 43. K. 20.1.54) dredged at Linga-linga, Morrumbene Estuary, Portuguese East Africa, common on a sandy bottom. One specimen (P.E.A. 148. —.11.55) dredged off Mozambique Island, Portuguese East Africa. These specimens are deposited in the Depart- ment of Zoology, University of Gape ‘Town. Distribution. Coasts of Borneo, China and Japan, Singapore, Philippine Islands, Torres Strait, Madras, Ceylon and East Africa. Remarks. The nine specimens from Linga-linga are a uniform series, but the specimen from Mozambique Island shows minor differences from these. It has therefore been considered desirable to give a separate diagnosis for each sample. 262 ANNALS OF THE SOUTH AFRICAN MUSEUM Diagnosis of the Linga-linga lancelet. In the following diagnosis counts and — measurements have been made on the 9 specimens in this sample and the ~ results have been analysed statistically. 1. Dorsal fin chambers number 258-306: Mean 287-0: Standard Deviation 15:1: S.D. = 5:26 per cent of the Mean. 2. Preanal fin chambers number 70-88: Mean 80-6: Standard Deviation 5°61: S.D. = 6-06 per cent of the Mean. 3. Tallest of dorsal fin chambers 3:0—3-6 times as high as broad: Mean 3:17: Standard Deviation 0-24: $.D. = 7-48 per cent of the Mean. 4. Height of the dorsal fin contained 7-10 times in the depth of the body in the mid-atrial region: Mean 8-8: Standard Deviation 0-97: $.D. = 11-05 per cent of the Mean. 5. Postatrioporal region 0.43-0°47 the length of the preatrioporal region: Mean 0-450: Standard Deviation 0-012: $.D. = 2-72 per cent of the Mean. 6. Myotomes from anterior end to atriopore 36-39: Mean 37:1: Standard Deviation 0:79: S.D. = 2-13 per cent of the Mean. 7. Myotomes from atriopore to anus 17-19: Mean 17-7: Standard Deviation O-71. 9.0. 4-0) per centiol the Mican. 8. Myotomes posterior to anus 9-11: Mean 9-9: Standard Deviation o-60: S-D)\==6-07, per cent of the Mean. g. Total myotomes 63-66: Mean 64:7: Standard Deviation 1:12: $.D. = 1°73 per cent of the Mean. 10. Maximum length in sample examined 48 mm.; minimum length 42 mm. Diagnosis of the Mozambique Island lancelet. Dorsal fin chambers number 276. Preanal fin chambers number 68. Tallest of dorsal fin chambers 4:0 times as high as broad. Height of the dorsal fin contained 8 times in the depth of the body in the mid-atrial region. Postatrioporal region 0-46 the length of the preatrioporal region. Myotomes from anterior end to atriopore 34. Myotomes from atriopore to anus 17. Myotomes posterior to anus II. Total myotomes 62. Length of specimen 62 mm. mem Oo N SO OI DY =! The following characters also assist in the recognition of this species and are illustrated in fig. 3 which has been prepared from camera lucida drawings of the Linga-linga lancelet. The rostrum is of moderate size and is separated from the dorsal fin by a post-rostral notch. The comparatively low dorsal fin is of approximately the same height throughout the length of the body. The caudal ON THE LANCELETS OF SOUTH AND EAST AFRICA 263 fin is rather small and lacks the subterminal indentations in outline present in B. capense. ‘The anus is situated in advance of the centre of the lower lobe of the caudal fin. The preanal fin is long but comparatively narrow and the preanal chambers extend almost the full width of the fin. Seen from the ventral aspect the preanal fin chambers are broad and, in some specimens, there is incipient doubling of the fin either at the anterior end only or through- out its entire length. The presence of a double preanal fin has been mentioned by Gilchrist (1923) as a character of B. bazarutense. The anterior end of the body is slender and the oral hood comparatively shallow dorso-ventrally. In all the specimens examined well-formed gonads were present. The body is more or less circular in cross-section in contrast with the marked lateral flattening of most species other than B. belchert. 0:25 5 0°45 0-012 0°50 0:024. 5808 <0-00I 6 SPI 0°79 36°5 0:82 1°844. Or! 7 77] O-71 17:0 0-76 2°330 >0:02 8 9:9 0-60 10:0 0°54. 0:448 >0°25 9 64:7 on 63°56 1-00 2°746 0-01 A statistical analysis showing to what extent the differences between samples of B. belcheri from Linga-linga in Portuguese East Africa and Amoy in China may be held to characterize the populations from which the samples were taken. The single specimen from Mozambique Island has been shown to differ from those from Linga-linga more particularly in the rather low number of myotomes before the atriopore. Whereas it is impossible to draw any con- clusions from such differences in a single specimen, the occurrence of this specimen does suggest that there might be other populations of B. belchert on the East African coast where isolation may have given rise to slight differences which are maintained. It is also possible that Gilchrist’s species B. bazarutense should more properly be referred to as a variant of B. belcheri. In his description of B. bazarutense Gilchrist was more concerned with separating his species from B. capense and B. elongatum than any other species, and he overlooked the similarity with B. belcherr, the existence on the East African coast of this species not being suspected at that time. Gilchrist placed some weight on the peculiar form of the preanal fin in B. bazarutense, but the approach to a similar con- dition in the Linga-linga lancelet shows that this character is, in fact, a further ON THE LANCELETS OF SOUTH AND EAST AFRICA 265 link with B. belcheri. Nevertheless B. bazarutense can be recognized from a number of slight differences which involve most of the characters, but whether these differences are sufficient to warrant specific recognition is another matter which, as the types of this species appear to have been lost, can only be settled when a new sample has been obtained. Branchiostoma banceolatum (Pallas) Limax lanceolatus Pallas (1774) Branchiostoma lubricus Costa (1834) Amphioxus lanceolatus Yarrell (1836) Branchiostoma lanceolatum Gray (1851) Material examined. Two specimens (MOR. 43.K. 20.1.54) dredged at Linga-linga, Morrumbene Estuary, Portuguese East Africa, from sandy bottom. These specimens are deposited in the Department of Zoology, University of Cape Town. Distribution. Norwegian coast, North Sea, Kattegat, Heligoland, English Channel, Irish Sea, Mediterranean Sea, Suez Canal, Arabian Sea, Madagascar, Portuguese East Africa. Remarks. These two specimens of B. lanceolatum were included in the same sample as the nine specimens of B. belcherr from Linga-linga, Portuguese East Africa. Apparently the two species have been taken from the same locality at the same time. Diagnosis of the Portuguese East African sample of Branchiostoma lanceolatum 1. Dorsal fin chambers number 270 and 280. 2. Preanal fin chambers number 32 and 34. 3. Tallest of dorsal fin chambers 2-5 times as high as broad. 4. Height of dorsal fin contained 11 and 12 times in the depth of the body in the mid-atrial region. 5. Postatrioporal region 0-40 and 0-42 times the length of the preatrioporal region. 6. Myotomes from anterior end to atriopore 34 and 35. 7. Myotomes from atriopore to anus 13. 8. Myotomes posterior to anus Io. g. Total myotomes 57 and 58. o. Length of the specimens examined 31 mm. The two specimens examined were not in a very good condition, but, as far as could be determined, in general appearance and in the form of the rostrum and the caudal fin they resembled B. lanceolatum from the central Mediterranean region (see Webb, 1956b). 266 ANNALS OF THE SOUTH AFRICAN MUSEUM This is the third record of B. lanceolatum from a region east of the Suez Canal where this species is said to occur in abundance (Gruvel, 1933). Prenant (1928) first recorded lancelets which he held to be B. lanceolatum from the north-west coast of Madagascar. In 1933 the “John Murray’ Expedition collected a single specimen of the species in a haul containing a large number of B. arabiae (Webb, 1956). The present record from Portuguese East Africa is similar in that the two specimens were taken together with another species, but this time B. belchert. ‘This sporadic occurrence of B. lanceolatum outside European waters suggests that this species is spreading from the Suez Canal to the Arabian Sea and Indian Ocean. TABLE V Range of variation in Indian Ocean and Arabian Sea samples of B. lanceolatum Total range of CHARACTER -—_,—-__—__——__ variation in Euro- NUMBER Portuguese | Madagascar Arabia pean samples of East Africa B. lanceolatum I 270-280 220-280 264 200-270 2 32-34 35 47 29-48 3 2°5 — 1:6 I-4—2°7 4 II-12 — 12 8—16 5 0°40—0°42 — 0°52 0°38-0°51 6 34-35 35 oH 33-38 7 13 15 The 12-16 8 IO 12 14 10-14 9 57-58 62 64 58-65 A comparison of the recorded range in variation of the numerical diagnostic characters in the Indian Ocean and Arabian Sea samples of B. lanceolatum with the total range in variation in the European samples of that species. In Table V the diagnoses for the Portuguese East African, Madagascan and Arabian samples of B. lanceolatum are compared with that of the European samples of the species. It is seen that the total range in variation in each character in the European samples covers the range for the Indian Ocean and Arabian Sea forms with very few exceptions. These exceptions are a rather high number of dorsal fin chambers in the Portuguese East African and the Madagascan :samples (character 1), the relatively long postatrioporal region in the Arabian specimen (character 5) and a low total number of myotomes in one of the Portuguese East African specimens (character 9). In each case, 2B ON THE LANCELETS OF SOUTH AND EAST AFRICA 267 however, these figures fall outside the European range by so little that it is doubtful how far they can be considered significant. ‘Two facts emerge from this comparison which may be of value. First the diagnoses for the Indian Ocean and Arabian Sea samples do not conform to any of the populations of the European lancelet for which detailed figures are available (Webb, 1956b). They may be similar, however, to the Suez Canal population a diagnosis for which has not been published. Second, the numerical range shown by the characters of the three samples from East of the Suez Canal, when taken together, exceeds in certain respects that of any of the European populations determined from larger samples. The probability is, therefore, that the Portuguese East African, Madagascan and Arabian samples are not all taken from the same population. The first two of these samples probably belong to one population, taking into account both the similarity of the diagnoses and their geographical location, but the Arabian specimen almost certainly belongs to another. On the assumption that the populations of B. lanceolatum in the Indian Ocean and Arabian Sea have been derived from the eastern Mediter- ranean population at a time subsequent to the opening of the Suez Canal in 1869, it can be presumed that any differences between populations in the Indian Ocean and in the Arabian Sea have arisen during a period of about sixty years. It is known that the English Channel population of B. lanceolatum has remained stable for more than seventy years (Webb, 1956b), so that any — changes that may have occurred in the eastern Mediterranean lancelets since their migration through the Red Sea may be due either to the change in environmental conditions accompanied by isolation o1 to some degree of hybridization with the endemic lancelets. In the Arabian Sea B. lanceolatum is sympatric with B. arabiae and in Portuguese East Africa with B. belcheri. Hybridization with B. arabiae might account for rather higher numbers of dorsal and preanal fin chambers than are commonly found in lancelets at Naples, but not, perhaps, for the relatively long postatrioporal region. Similarly, hybridization with B. belchert could account for a high dorsal fin chamber number and a rather tall dorsal fin. The evidence, however, is not at all clear and further samples of B. lanceolatum from Portuguese East Africa, the Arabian Sea and, in particular, the Suez Canal are required before an opinion can be given. CONCLUSIONS An examination of the types of Branchiostoma capense and new collections from South Africa of this species has been made. It is concluded that B. capense is a valid species and is not synonymous with B. californiense or B. elongatum as has been proposed by Tattersall (1903) and Franz (1922, 1922a and 1930). The remarkable similarity between many of the characters of B. capense and B. elongatum is thought to be due to convergence and not to close affinity of 268 ANNALS OF THE SOUTH AFRICAN MUSEUM these geographically widely separated forms. Comparisons have been made between B. capense and the neighbouring species B. belcheri, B. bazarutense and B. arabiae from the east coast of the African continent and B. africae, B. nigeriense, B. takoradu and B. leonense from West Africa. B. capense has been shown to be related to the Guinea Coast lancelets of West Africa and to'be an extreme member of that series. ‘The Cape lancelet extends in range from the south-west tip of the Cape Province (Simons Bay and False Bay) eastward to Port Elizabeth (Algoa Bay). On the east coast of Africa a different species, B. belcheri, is established and has been recorded from the Morrumbene Estuary and Mozambique Island in Portuguese East Africa and also by Franz (1922) from a locality near Zanzibar. The population of B. belcheri at the Morrumbene Estuary is very similar to that at Amoy on the China coast and the two forms can only be distinguished by minor differences. B. bazarutense, described by Gilchrist (1923) from the Bazaruto Islands off Portuguese East Africa, is a close relative of B. belcheri differing from the Morrumbene form, but not, perhaps, sufficiently to warrant specific distinction. Another species, B. arabiae, which is related to B. belcheri, has been described by Webb (1956) from the Arabian coast. Sympatric with B. arabiae and B. belcheri from the Morrumbene Estuary are populations of B. lanceolatum which have evidently been derived from a Suez Canal population of that species. A further record of B. lanceolatum has also been made by Prenant (1928) from the north-west coast of Madagascar. The lancelets of South and East Africa, therefore, appear to have been derived from several different sources and consequently do not form a closely related group as, for instance, seems to be the case with the lancelets of the west Atlantic or, to a lesser extent, those of the east Atlantic. An explanation of the occurrence of such a variety of different lancelets on these coasts is probably to be found in the pattern of currents in the Indian Ocean. B. capense, being related to the lancelets of the west side of the contingnt, is probably endemic. Its distribution seems to be limited to a region of the Cape coast just beyond the reach of the south-west warm Agulhas Current which is formed from components from the north-south Mozambique Current, running between Madagascar and Portuguese East Africa, and the east-west Equatorial Current from Australia to the south of Madagascar. The direction of flow of the Agulhas Current may well prove a barrier to the spread of the planktonic larvae of B. capense northward beyond Port Elizabeth. The reason for the abrupt termination of the range of this species to the west at the Cape (Simons Bay and False Bay) might be due to the presence of the cold water of the Benguela Current which originates south of this point and passes up the west coast of the continent. The Benguela Current, however, provides a connection between the Cape and the Guinea Coast of tropical West Africa which could account for a relationship between B. capense and the Guinea Coast lancelets as there must be some intermixing of waters between the Benguela and the ON THE LANCELETS OF SOUTH AND EAST AFRICA 269 Guinea Coast Current at the level of the equator. From the direction of flow of the Benguela Current it would seem more likely that the Guinea Coast lancelets are derived from the Cape lancelet than that the reverse is true. B. belcheri, which is an Asiatic form common in the China Seas, East Indies and North Australia, appears to have been carried to East Africa by the broad east-west Equatorial Current which strikes the east coast of Madagascar and turns both to the north and to the south, passing around the island. The direction of flow of this current could account for the occurrence of B. belchert at Zanzibar and in Portuguese East Africa, and, perhaps, elsewhere on the East African coast north of Port Elizabeth. This species should also occur in Madagascar. The close similarity between the East African and the Amoy forms of B. belchert suggests that a migration from east to west in the Equatorial Current is made with some frequency and that the East African populations are not fully isolated from the Asiatic. B. arabiae is connected with the East African population of B. belchert by the seasonal north-east Monsoon Drift, but has come to occupy a somewhat isolated position on the Arabian coast at the entrance to the Gulf of Oman. B. bazarutense similarly has only been recorded once from the Bazaruto Islands. Both of these forms have evidently been derived from B. belchert and presumably could only have arisen through occupation of an isolated area. There must, therefore, be regions along the East African coast where such isolation is possible. The mode of distribution of B. lanceolatum in the Arabian Sea and Indian Ocean can be similarly explained. Once the species reached the Gulf of Aden from Suez by means of the reciprocal currents of the Red Sea, one current runs along the coast of Somaliland to meet the seasonal south-west Monsoon Drift passing the East African coast to Madagascar and the Mozambique Current, while another runs north-east along the Arabian coast. Thus the south-west Monsoon Drift could account for the presence of B. lanceolatum on the north-west coast of Madagascar and in Portuguese East Africa, and the Arabian Current for the presence of that species in the locality of B. arabiae. The distribution of lancelets on the East African coast, therefore, is in accor- dance with the ocean currents, and may give an indication of the geographical range of other marine organisms similarly distributed by means of planktonic larvae in the Indian Ocean. REFERENCES Andrews, E. A. 1893. ‘An undescribed acraniate, Asymmetron lucayanum.’ Stud. biol. Lab. Johns Hopkins Univ., V, pt. 4, 213-247. Beebe, W. and Tee-Van, J. 1941. ‘Fishes from the Tropical Eastern Pacific (from Cedros Island, Lower California, South to the Galapagos Islands and Northern Peru). Part I, Lancelets and Hag-fishes.’ Zoologica, New York, 26, 89-91. Cooper, J. G. 1868. In Cronise, Natural wealth of California, San Francisco, 498. 270 ANNALS OF THE SOUTH AFRICAN MUSEUM Costa, O. G. 1834. Annuario zoologico: Cenni zoologici ossia descrizione delle specie nuove.’ 49, Napoli. Drach, P. 1948. In Grassé, Traité de Zoologie, 11, Paris. Franz, V. 1922. ‘Systematische Revision der Akranier. (Fauna et Anatomia ceylanica Nr. 10.”) Jena &. Naturw. 58, 369-452. 1922a. ‘Kurzer Bericht tiber systematische Acranierstudien.’ Zool. Anz. 54, 241-249. 1930. Acranier (Lanzett-Fische). Tabul. biol. 6, 573-581. Gilchrist, J. D. F. 1902. “South African Fishes.’ Marine Investig. S. Africa, 11, 7, 101-113. 1923. ‘A new species of Amphioxus from the Bazaruto Islands (Branchiostoma bazarutense).’ Ann. Mag. nat. Hist. (9), 12, 62-65. Gray, J. E. 1847. ‘Description of a new species of Amphioxus from Borneo.’ Proc. zool. Soc. Lond. 15, 35-36. 1851. Cat. Brit. Mus. Fish, 7, 149-150. Gruvel, A. 1933. “‘Abondance du Branchiostoma lanceolatum (Pallas) dans le canal de Suez.’ C.R. Acad. Sci. Paris, 197, 92-93. Hubbs, C. L. 1922. ‘A list of the lancelets of the world with diagnoses of five new species of Branchiostoma.’ Occ. Pap. Mus. Zool. Univ. Mich., No. 105, 1-16. — In Monod, 1927. ‘Contribution a la faune du Cameroun. Pisces I.’ Faune Colon. Frang. Paris, 1, 644-645. Pallas, P. S. 1774. Spicilegia zoologica, 10, 19. Prashad, B. 1934. ‘Ona collection of Indian Cephalochordates, with notes on the species from the Indian waters in the Indian Museum, Calcutta.’ Rec. Indian Mus. 36, 329-334. Prenant, M. 1928. ‘Sur un Amphioxus de la céte Malgache.’ Bull. Soc. zool. Fr., 53, 202-204. Snodgrass, R. E. and Heller, E. 1905. ‘Shore fishes of the Revillagigedo, Clipperton, Cocos and Galapagos Islands.’ Proc. Wash. Acad. Sci., 6, 333-427. Sundevall, C. J. 1852. ‘Ny art af Amphioxus.’ Ofvers. VetenskAkad. Férh. Stockh., 9, 147-148. 1853. ‘Ny art af Branchiostoma.’ Ofvers. VetenskAkad. Forh. Stockh., 10, 11-13. Tattersall, W. M. 1903. ‘Notes on the classification and geographical distribution of the Cephalochorda.’ Trans. Liverpool biol. Soc., 17, 269-302. 1903a. ‘Report on the Cephalochorda collected by Professor Herdman at Ceylon in 1902.’ Rep. Ceylon Pearl Oyster Fisheries, Part I, Suppl. Rep., 6, 209-226. Webb, J. E. 1955. ‘On the lancelets of West Africa.’ Proc. zool. Soc. Lond., 125, 421-443. 1956. ‘Cephalochordata.’ Sci. Rep. ‘Fohn Murray’ Exped., X (3), 121-128. 1956a. ‘Cephalochordata of the coast of Tropical West Africa.’ Aflantide Rep., IV, 167-182. — 1956b. ‘On the populations of Branchiostoma lanceolatum and their relations with the West African lancelets.’ Proc. zool. Soc. Lond., 127, 125-140. Yarrell, W. 1836. History of British Fishes, 468, London. Pee ae te # ~ aF Sts + Liens 7 rhe: ANNALS OF THE SOU TH AFRICAN M USEUM. are issued in parts at jirreg intervals as material becomes available. — . Some volumes and parts are out of print, and others are cen wilde as parts of a set, or val respectively. The prices of parts published prior to I 1940 have been increased. — . Out of print: Vols. I, II (Parts 1-3, 5, 7, 8), III (Part 1), V (Parts 1, 2, 5,8, a VI (ean 1 Index), VII (Parts 1, 2, 3, 5), VIII, EX (Parts 1, 2), X (Part 2); XI (Parts Phen Indes), rasa (Part 1), XXIV (Part 2), XXXI1 (Parts I-3). Vol. . a 1900-1902 III. 1903-1905 IV. 1903-1908 _V.. 1906-1910 VI. 1908-1910 VII. 1908-1913 IX. 1911-1918 X. IQII-I9I4 XI. 1911-1918 XII. 1913-1924 MIM. r91g—1923 XIV. 1915-1924 XV. 1914-1916 XVI. 1917-1933 XVII. 1917-1920 XVIII 1921 XIX. 1924-10925 XX. 1924-1926 XXI. 1925-1927 XXII. 1925-1928 XXIII. 1925-1926 XXIV. 1929-1938 XXV. 1927-1928 XXVI 1928 XXVII 1929 XXVIII. 1929-1932 XXIX. 1929-1931 XXX. 1931-1935 INDEX XXXII. 1934-1950 XXXII. 1935-1940 XXXIII. 1939 XXXIV. 1938 XXXV. 1956. XXXVI. 1942-1948 _ XXXVII. 1947-1952 XXXVITI. 1950 XXXIX. 1952 XL. 1952-1956 XLI. 1952-1955 XLIT. 1953-1956 XLITI. 1955- . Zoology and Geology \ (eel Parts 1-3, | 4, aye Zoology .. oe AG .. (excl. Part. 1) Palaeontology A Geology, Palcontaioses Zoology, Anthropology (excl. Parts 1, 2, 5, 8, 9) Zoology en ‘Part I, nde Palaeontology . (Parts 4 and 6 only) — Botany Scee Meee ee. Parts 1 and 2) © Zoology : . (excl. Part 2) Zoology (exall Pars Beas Index) Palaeontology and Geolens ; Bera and oa Zoology a Zoology Botany Zoology Zoology Zoology .. gi se Zoology .. AOA Zoology . . (excl. ‘Part 1) Pale cucesloey gar ee ON ae oes Sheba th aur Zoology : Anthropology and Ethnology Zoology .. ae we Zoology : Anthropology Palaeontology Zoology Zoology - of papers, authors, and ents published in an L XXK Palaeontology Zoology Zoology Zoology Zoology Zoology Anteecion,. Zoology .. Rito aes Be Cae A. Zoology Botany Zoology a fe ; Palaeontology he DEN Nr sth ea Zoology and Palaeontology . (excl. Index) Part 1 £1 os, od, Parts 2 & 3, each £0 55. od. Part i ‘ Copies may be obtained from— The LIBRARIAN, Soury Arrican Museum, Cape Town. .. (excl. Part 2) : Ze 4 only) ; | 2 me an Bi OHNO RENAN NWOOW KN ~ mw Bmw m 12 G9 HD GO GO ob 69 Oop wb PT ae gee i) Ss OMOwW IH WO Oo ne * a aes ° oo oo on om oOo °° xo) oon {<<} ow oo = peers Cwm i Wie oe ay ih aA 4 22 ytd $f ; p.foip iba | byiis ; : at oe SMITHSONIAN INSTITUTION LIBRARIES (INUIT Ht | ae ; i Hin 3 9088 01206 5777