27 3 9 FS Haws hha | of ys abe 4 4 y 4 Pi te reat BO a Sgt oe OMarwtan lage pag (i ANAM Tai LOA i % it One ae ate io. ae pi ‘ eA st ‘ 4 dh ee Ue Po La} * 4 z Speke et s 4 * Has Fie ack Lp cee cows Capeed if ro ac Me a 64 oils os ig ob a es = Lee Sais Ses: 2 oA dt A Be (eek iat vaatueegonnn aa fi Waly tae Mi ante =e oso stats Rae j Vane ot 6 Ln Lt ae 7h et Haar Pa ek ea 4G i) Cae ee 3 Wot i oe [rey ato kat bi ) eA) et Noy ts thay ean CL ee oe cea eA a ai Puen renee Pari Wee? Hi VEL eaten ARCS ‘ 14 aha 1 4 " \ ot aye iy ve 4 ‘ 4 AM) WP bk Pads ot ag d L* He n ‘ qa) Loe Md ‘ Poe Wien iri at ty St 2G. sin yy ut Hot TOO BD oy Ca aes CATO TO Sy de ae ay aA AAS Ke Mette a tel he Hrd i thy f ‘W 5k Oe i Ge hb 4 a 7 Ae the Sa ka PCH daltekellbay v ‘ sd Bafta SEM ORM ITAA RENTS ieee netlist Dey hy heed At hats tk ree Met ae ey iy ea Ne oly dattet ve 4 i cs via wea yaa AS AN pH ded , Sune anit Pin Oy pete Hut . rd Now Mg 4 ic) ol Ons Hate Nh , 1G ‘ rutin ner ren ent Af h aus aS" SMaTiona WZ 2g 4f Ain ANNALS OF THE SOU AFRICAN MUSEUM VOLUME XLI S : ~ ibang 2 ites SS) : ) o> ¢, o Ran ate US nat * a 3 £ = MSO a. my i, oN \\ ‘ J oy WAy, ~\ ‘sf f coh nen ont 4 i iI5RPAR { A“ wy get Fase PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY THE RUSTICA PRESS (PTY.) LIMITED, COURT ROAD, WYNBERG, CAPE 1952 = 1955 TRUSTEES OF THE SOUTH AFRICAN MUSEUM Pror. R. S. ApAmson, F.R.S.S.Afr. Pror. C. G. S. DE VILLIERS Mr. CounciLLor H. E. GEARING Pror. D. L. ScHOLTz Mr. C. J. SIBBETT, J.P. SCIENTIFIC STAFF OF THE SOUTH AFRICAN MUSEUM KEPPEL HArcourT BARNARD, M.A., D.Sc., F.L.S., Director. ALBERT JOHN Hesse, B.Sc., Ph.D., Assistant in Charge of the Entomological Department. LizuweE Dirk BoonstrA, D.Sc., Assistant in Charge of the Palaeontological Department. Miss E. MARGARET SHAW, B.A., Assistant in Charge of the Ethnological and Numismatic Departments. Miss G. Joyce Lewis, Ph.D., Assistant in Charge of the Botanical Department. A. J. H. Goopwin, M.A., Honorary Keeper of the Archaeological Collections. ‘JUL 2 8 1955 LIST OF CONTRIBUTORS H. ANDREAE On South African Coccinellidae The genus Anobium K. H. BARNARD South African Pycnogonida ‘ South African Parasitic Copepoda S. BREUNING Neue Lamiinae aus dem S.A. Museum.. G. J. BROEKHUYSEN The breeding and growth of Hymenosoma orbiculare (Crustacea, Brachyura) J. Forest Les Paguristes des cétes occidentales et méridionales d’Afrique A. D. Harrison Four new species of Gregarines from mountain cockroaches of the Cape Peninsula J. W. JAckson A revision of some South African Brachiopoda, with descriptions of new species .. Z. KaszAB Studien tiber siidafrikanischen Meloiden D. LEsTon Notes on Ethiopian Pentatomidae. A new genus and species from the Cape Province Notes on Ethiopian Pentatomoidea. Some specimens from Southern Africa in the South African Museum N. A. H. MILiarp New species of Hydrozoa from South Africa K. M. F. Scotr Some new Caddis flies from the Western Cape Province Bo ‘'TJEDER Two new species of Bortomyia from South Africa E. UHMANN Finf Dactylispa-Typen Péringuey’s W. WITTMER Beitrag zur Kenntnis der Malacodermata Afrikas PAGE 215 367 381 407 359 INDEX OF NEW GENERIC NAMES INTRODUCED IN THIS VOLUME Carvalhocoris (Insecta, Hemiptera, Pentatomidae) Leston Falsotragiscus (Insecta, Coleoptera, Cerambycidae, Lamiinae) Breuning Mimogmodera (Insecta, Coleoptera, Gerambycidae, Lamiinae) Breuning Scapogoephanes (Insecta, Coleoptera, Cerambycidae, Lamiinae) Breuning Tambusoides (Insecta, Coleoptera, Cerambycidae, Lamiinae) Breuning. . DALE OF ISSUE, OF THE, PARTS Part 1, October, 1952. Part 2, October, 1953. Part 3, April, 1954. Part 4, June, 1954. Part 5, February, 1955. Part 6, April, 1955. 46 345 352 S7/ 348 LIST OF PLATES Plate I. II. South African Brachiopoda. III. IV. Paguristes. Acanthochondria Achelia Achtheinus Achtheres Aeptus Aethemenes Agonoscelis Agulhasia Ainigma Alebion Ammothea Amphimachus Anaphia Anauxesis Ancistrotos Anobium ne Anoplodactylus Antestia Antestiopsis Anthosoma Argulus Atelocera Attalus Austroraptus .. Blepisanis Boerias Bohmia Bolbocoris Boriomyia .. Botryllophilus Brachiella Caligus Callipallene Cancerilla Carbula Cardiodectes Carvalhocoris. . Cecrops Ceroplesis Chaetonymphon Charopinus Chilophoxus A. INDEX OF GENERA PAGE Chondracanthus Chonopeltis Clavella Clavellopsis Cloniocerus Coccinella Colotes Colossendeis .. Coponia Coridius Crania Cryptacrus Dactylispa Dalsira Dalterus Dasytes Decapotoma .. Delegorguella. . Deroplax Dinematura Dinemoura Dinemurus Dinometopus Diploxys Discoarachne Dolichopsis Dolops Doropygus Dorycoris Durmia Dysgamus Echthrogaleus Elvisura Endeis Enterocola Entobius Epilachna Eubrachiella Eunidia Exochomus Falsotragiscus. . Farnya 445 Gelastaspis Gregarina Gryphus Gunenotophorus Gynenica Halosoma Halydicoris Hannonia Hatschekia Hedybiinus Hedybius Heteropallene Hydractinia Hyllisia Hymenosoma Hyperaspis Idactus Iselma Kraussina Kyphomia Lamproglena Legnotus i Lepeophtheirus Lerida 3 Lernaea Lernanthropus Lerneopoda Lingula Lytta .. Martinella Medesicaste Megathiris Megerlina Menida 5 Metapallene .. Mimogmodera Miihlfeldtia Necrobia Nemesis Neococalus Nesippus Nogagus Nogaus ee Notodelphys .. Notostenus Nymphon Nymphopsis Obereopsis Oeax .. INDEX OF GENERA 126, 129 Olenecamptus Orthagoriscicola Paguristes Pallene Pallenoides Pallenopsis Pandarus Pantellaria Parabrachiella Parapellene Pausias. . Pennella Perissopus Petroplax Phanodemus Philhedonus Philorthragoriscus Phoxichilidium Phoxichilus Phytoecia Piezodorus Pipetta Procidella Prosopocera Prosymnus Pseudochariesthes Pseudopallene Pycnogonum .. Queubus Rhopalorhynchus Sabelliphilus Scapogoephanes Sitaris ‘ Sphaeronella . Sphyrion Stathmodera .. Stenoria Stenozygum .. Strabax Tambusoides Tanystylum Terebratella .. Terebratulina Tetraulax Thriocera Trebius Veterna Zanclopus Zonitomorpha Zygophylax 107, PAGE 170 105 4, 21 108 age - oy Sf rae ae ANNALS _ OF THE SOUTH AFRICAN MUSEUM ~ VOLUME XLI PART I, iiiae: — 7 1. A Revision fs some South ie Brash toda with Dunne: of New - Species—By J. WILFRED es D.Sc., F.S.A., F.G.S. (With Plates I-III.) > | figs = Be A. if ee al % sll 4 & 32S art _ ISSUED OCTOBER 1952 - PRICE 4s. 6d. : PRINTED FOR THE _ TRUSTEES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN BY THE RUSTICA PRESS LIMITED, COURT ROAD, WYNBERG, CAPE PXo NaN Ayo OF THE SOUTH AFRICAN MUSEUM \ _ VOLUME XLI 1. A revision of some South African Brachiopoda; with Descriptions of New Species.— By J. WiLrrep Jackson, D.Sc., F.S.A., F.G.S. (Late of Manchester Museum). (With Plates I-III.) 1. Introduction More than twenty years ago I received through the courtesy of the Director of the South African Museum a small collection of South African Brachiopoda for report and description. Various unavoidable circumstances prevented me from giving the necessary time and attention to the collection, and, in conse- quence, the report on the same has been considerably delayed.* The collection, though a small one, is of great interest as it contains several hitherto undescribed species and thus adds materially to our knowledge of the brachiopod fauna of the South African Province. ‘The specimens came from several stations extending from the neighbourhood of Saldanha Bay on the west coast to Delagoa Bay on the east. Those from the Saldanha Bay area include a new species of Megathiris and one of Crana. Hitherto Megathiris has not been reported from Cape waters. From False Bay there is also a strongly- -costate Kraussina which differs sufficiently from the well-known Avaussina rubra (Pallas) as to merit specific recognition. In addition there is a small species of Gryphus from the south coast, off Cape St. Blaize and Cape St. Francis, which appears to be undescribed. ‘The remaining species comprise a new Terebratu- lina (formerly referred to Terebratulina septentrionalis Couth.), and Terebratulina abyssicola Adams and Reeve, previously recorded from the Cape. To these are to be added Lingula (two species) from Durban and Delagoa Bay, and Agulhasia davidson from two new stations, one being off Durban. The description of the above series provides an opportunity of discussing and revising a number of other South African species preserved in the British Museum (Natural History) and in other collections. I have to thank the authorities of the above institution for the privilege of examining many of their * Additional material was sent to the author in 1949, and his revised report was received in 1950. Post-war difficulties have further delayed the publication of this report. Ed. VOL. XEI. PART I. pec 8 1952 2 ANNALS OF THE SOUTH AFRICAN MUSEUM specimens; also Messrs. H. McClelland and J. R. le Brockton Tomlin, M.A., for the loan of material. My own private collection has also been fully utilized. | The most important results achieved through the close study of this material is the placing of the Kraussina-series on a better basis, and the removal of certain forms erroneously ascribed to that genus. For many years this series has been in a state of confusion and a stumbling block to students of the Brachiopoda through the erroneous identifications that have been made. Some particulars with regard to punctation are added in the hope that these may prove useful in future researches on this group. 2. Previous Research The Brachiopod fauna of South Africa is not a very large one, but it includes some interesting species. Very little research has been conducted upon these in recent years, and nearly all the records date from the latter part of the eighteenth century or the first two-thirds of the nineteenth century. The first notice of Brachiopoda in Cape waters appears to be that of Pallas, who, in 1766,!* described Anomia rubra (now Kraussina rubra). Chemnitz, in 1785,” described the same species as Anomia striata promontorit bonae sper, and another species, also from the Cape, as Cognata Anomia craniolaris basi perforata. Owing to Chemnitz not being a binomial writer, his species are not recognized. G. B. Sowerby, in 1847,? redescribed Chemnitz’s second species as Terebratula cognata, which is now recognized as AKraussina cognata (Sow.) In 1819, Lamarck? described a species from Mauritius as Terebratula pisum. This species has given rise to much confusion and has been misidentified on several occasions. It is mentioned here as being an undoubted member of the fauna of the Natal coast. ‘The form attributed to this species by Davidson® is not that of Lamarck: it is referred to on a subsequent page. G. B. Sowerby, in 1847,3 described a Terebratula algoensis from Algoa Bay, coll. J. S. Bowerbank. The specimen, a bleached ventral valve, is in the British Museum. It has been referred to Kraussina rubra by Reeve, while Krauss regarded it as the same as his Terebratula natalensis. A discussion of this species is given later. Dr. F. Krauss, in 1848,° gave three species for South African waters, viz. Terebratula capensis (Gmel.), Terebratula cognata (Chem.) Sow., and Terebratula natalensis Krauss (Natal). ‘The 7. capensis (Gmel.) is Kraussina rubra (Pallas) ; T. cognata (Chem.) Sow. is Kraussina cognata (Sow.); T. natalensis Krauss has been referred to Kraussina and considered to be the same species as Terebratula pisum Lamarck. ‘The question as to its generic and specific position is discussed on a later page. In 1850,7 A. Adams and Lovell Reeve described as new two species of brachiopods as follows:— Terebratula capensis, dredged off the Cape of Good Hope at a depth of 120 fathoms; and Terebratula abyssicola, Cape of Good Hope, * Superior numbers refer to the Bibliography at end. REVISION OF SOME SCUTH AFRICAN BRACHIOPODA 3 120 fathoms. The 7. capensis Ad. and Rve. is not that of Gmelin (= Kraussina rubra): it was changed by Davidson in 18528? to Kraussia deshayesit (later Krauss- sina) and regarded by him as nearly related to Terebratula pisum Lam. Like the latter, it is not a Kraussina, but goes with pisum Lam. and pisum Dav. The Terebratula abyssicola Ad. and Rve. is now Terebratulina abyssicola (Ad. and Rve.). In 1870,° the late Dr. Dall described a Terebratella, which he later, 1871,1° called Terebratella rubiginosa, from Simons Bay, Cape of Good Hope. The single specimen is in the United States National Museum. Little can be said concerning this species until more examples are forthcoming. I have been unable to examine the original figure, but the description of the loop suggests that it is of a peculiar type. A rather remarkable species was described and figured by W. King in 187114 under the name of Agulhasia davidson, the examples having been obtained on the Agulhas Bank, South Africa, from a depth of 45 to 60 fathoms. It is a very small species with a curious and long beak to the ventral valve. ‘The genus is said to have Cretaceous representatives in Europe. During the ‘Challenger’ Expedition three species of brachiopods were met with at the Cape. These were described by Davidson in 1880° as follows: Terebratula vitrea var. minor Philippi (p. 29, pl. II, figs. 5-6); Terebratulina caput- serpentis var. septentrionalis Couth. (p. 33, pl. I, figs. 6-9); and Avaussina pisum (Val, apud Lam.) (p. 54, pl. IV, figs. 7-8). All were dredged at Station 142, off the Cape of Good Hope, lat. 35°4'S., long. 18°37’E., December 1873. 150 fathoms. Terebratula vitrea var. minor is now known as Grpphus affinis (Calcara) and is common in the Mediterranean. The specimens dredged at the Cape are not that species. They are referred on a later page to a new form of Gryphus. Terebratulina septentrionalis is a distinct species and not a variety of T. caput-serpentis (now retusa L.). It occurs at many stations in the North Atlantic, off Northern Europe and off the eastern states of America. The Cape specimens are distinct from that form and are dealt with later in this memoir. The examples referred by Davidson to Kraussina pisum are specifi- cally distinct from Lamarck’s species. These two species are also generically distinct from Kraussina. In 1892,” Fischer and Oehlert gave a list of nine species as inhabiting Cape waters and remarked on the great dissimilarity between this fauna and that of the Magellanic Province. They called attention to the recorded presence of two species of Kraussina in the Australian region, viz. K. lamarckiana and Kf. atkinsom. These two species, however, do not belong to the genus Araussina: the first is the genotype of Megerlina and the second may also belong to that genus. In the Journal of Conchology for 1901 E. A. Smith recorded Kraussina atkinson. (YT. Woods)—a Tasmanian species—for Algoa Bay, Cape Colony (Brit. Mus. presented by J. H. Ponsonby). I have examined the specimens in question—four in number (B.M. 1900.6.13.5-8)—and find them to be un- doubted juveniles of the Terebratula prsum Lam., which is not a true Araussina. 4 ANNALS OF THE SOUTH AFRICAN MUSEUM In my opinion, the Tasmanian species also does not belong to the genus Kraussina, as it possesses a different type of cardinalia and brachidium. In 1908,14 Blochmann recorded the discovery by the ‘Valdivia’ Expedition of a small species of Liothyrina (=Gryphus) in company with Kraussina rubra and Terebratulina abyssicola (not septentrionalis) on the Agulhas Bank. He figured the brachidium of the Liothyrina (pl. 39, fig. 31) and regarded it as distinct from Liothyrina (now Gryphus) affins. It is possibly the same species as that referred to Gryphus sp. nov. in this memoir. Dr. J. Allan Thomson, in 1918,)° listed ten species as occurring at the Cape, and called attention to the fact that no additions had been made to the brachiopgd fauna of South Africa during the previous thirty years. He remarked ‘especially on the absence of Crania, and discussed the relation of the fauna with that of Australia and other places. Dr. W. H. Dall, in 1920,!* recorded Pantellaria (olim Miihlfeldtia) echinata (F. & O.) as from the Cape of Good Hope, 224 fathoms (Jeffreys coll.), one specimen in the United States National Museum. The type locality of this species is “Off Cape Bojador, Sudan [szc.] coast of West Africa, 407 fathoms’. In a paper in 1921,1” the present author referred to the presence in the British Museum (Natural History) of an immature example of Muihlfeldtia truncata labelled ‘South Africa. J. H. Ponsonby coll., 1900. 6.13.4.’ ‘The occurrence at the Cape of this and the previously mentioned species—both Northern forms, one Mediterranean and east Atlantic, the other West African—is very remark- able, but one feels somewhat doubtful as to the provenance of the specimens. In 1932, Lt.-Col. W. H. Turton*™ published an account of the marine shells of Port Alfred. These were largely dead and drifted specimens and included a few brachiopods, one of which resembled a Cranza, PLATE I Figs. 1-4. Crania roseoradiata sp. nov. Syntypes, A. 5665. Dredged in 45 fathoms 7 miles SSW. of Constable Hill (Saldanha Bay). Four dorsal valves. South African Museum. Figs. 5-7. Terebratulina abyssicola (Adams and Reeve). A. 5617. Dredged in 85 fathoms, 33°6’S., 28°11’E., off SE. coast of the Cape. South African Museum. Fig. 7, exterior view; figs. 5-6, interior views. Fig. 8. Terebratulina meridionalis sp. nov. Holotype. Station 142, ‘Challenger’ Expedition ere miles due south of Cape Point), 150 fathoms. British Museum (Nat. Hist.). 0. 78.6.15.3. Fig. 2: Teta septentrionalis Couthouy. Casco Bay, Maine. British Museum (Nat. ist.). Figs. 10-13. Gryphus capensis sp. nov. Holotype, A. 5627. Dredged in 75 fathoms, 29 miles SW. of Cape St. Francis. Paratype, A. 5626. Dredged in 125 fathoms, 73 miles S. by W. of Cape St. Blaize. South African Museum. Figs. 10-11: Holotype. Fig. 12: Paratype. Fig. 13: interior of Holotype. Plate t. Ann. S. Afr. Mus., Vol. XLI. South African Brachiopoda. REVISION OF SOME SOUTH APRICAN BRACHIOPODA 5 3. Description of Spectes ORDER ATREMATA BEECHER, 1891 Superfamily LINGULACEA Waagen, 1885. Family LINGULIDAE Gray, 1840. Genus Lingula Bruguiere. 1797. Ency. Meths. Vers. I, p. 151, II, pl. 250. Lingula sp. aff. murphiana (King MS.) Reeve. Lingula murphiana (King MS.) L. Reeve, ‘Mon. of Lingula’, Conch. Icon., November 1859, pl. I, fig. 3. Davidson, ‘Mon. Rec. Brach.’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. pt. III, 1888, p. 215, pl. X XIX, fig. 11, and pl. XXX, figs. 1-3. In 1920, when studying the brachiopods in the British Museum (Natural History), I noted a spirit specimen in the zoological department labelled: ‘Lingula anatina Lam. Durban. J. H. Ponsonby. 99.4.14.1.’ Many years ago, Mr. R. Cairns presented me with a large example of Lingula with a label: “Durban, r900. J. P. Cregoe.’ On making inquiries about this collector, Dr. K. H. Barnard tells me that Cregoe contributed quite a lot to the Invertebrate collection of the South African Museum about that period, chiefly from Natal and Lourenco Marques; but the Museum files contain no biographical details. The Lingula in question agrees closely in shape and size (viz. L. 58-5: W. 26:1 mm.) with the L. murphiana figured by Reeve (op. cit., pl. 1, fig. 3) and has the same coppery-red colour with green border anteriorly. It also agrees with two other large examples in my collection from New South Wales (precise locality unknown). Reeve’s description of his species is as follows: ‘Murphy’s Lingula. Shell oblong-ovate, anteriorly square, posteriorly attenuated, umboes small, sharp, yellowish-green, stained with coppery-red; valve closed throughout. King MS. Mus. Cuming. Hab. Moreton Bay, Australia; Strange.’ The specimen figured is, L. 57-8: W. 25:7 mm. He further remarks: “Whether this should be regarded as an Australian form of L. anatina or as a distinct species, it is certain that the differences are obvious and constant. ‘The late Captain Phillip King, who was an excellent conchologist . . . was struck by the peculiarities of this Lingula, and sent speci- mens home to Mr. Cuming with the above name. More have been collected by Mr. Strange in nearly the same locality, and all are distinguished from L. anatina collected abundantly by Mr. Cuming in the Bay of Manila, by a more square outline, and by a peculiar coppery-red tone of colour.’ Davidson (op cit.) figures two large examples from Moreton Bay in his collection at the British Museum, and gives an amplified description and the colour as coppery-red, with bands of different shades of green and brown. He 6 ANNALS OF THE SOUTH AFRICAN MUSEUM gives the size as: L. 2 in. 6 lines2 W..c,1n. 1, dine (=L. 63:5; W227-7isasane The size of the specimen in his plate XXX, figs. 1-3, 1s somewhat less, being: L. 53°5: W. 24 mm. Lingula murphiana, though first described from Queensland, appears to have a wide distribution. According to W. H. Dall’s List bof 1920, 16 the United States National Museum possesses the following: N.E. Australia, coll. Stearns, 2 specimens; Australia, coll. Cuming, 2 specimens; Fiji Islands, U.S. Ex. Exp., 2 specimens; Vita Leva, near Rawa River, U.S. Ex. Exp., 2 specimens; and off Shimbaya Gulf, Japan, E. S. Morse, 1 specimen. One wonders how much reliance has been placed on colour in these determi- nations. It seems to be a variable character in some forms. I have in my possession four very large specimens of the same shape and colouring as the Queensland form collected at Karachi by Lt.-Col. H. C. Winckworth in September 1932 and given to me by his brother Mr. R. Winck- worth. The sizes of these are: 64 X 28-1; 62-4 x 29; 60°8 X 29; and 59°3 X 27°7 mm. If these had been found at the Queensland locality, they would have been referred to L. murphiana Rve. From the same collector I have a smaller Lingula found at Trimenuaai Ceylon, October 1929, also rather dark in colour with a green border. It measures: L. 43:3; W. 18-8 mm. _ It might very well be a younger stage of the above examples. Dall!® cites Lingula translucida Dall—a ruddy-brown species—from aes coll. Captain Shopland, 2 specimens and Fulton, 2 specimens, but does not figure the species. He also cites the same form from Java, coll. Ward, 2 speci- mens and Palmer, 1 specimen, and from Tataan Island, Philippines, 12 fms., coll. Bureau of Fisheries, 1 specimen. It is uncertain which is the type station. Dall gives Java, but cites the catalogue number for Karachi. With regard to the Lingula anatina of authors, this Eastern Seas species (Type station, Amboyna, Moluccas) was referred to the Patella unguis of Linne by Dall in 1920,%¢ but the shell to which Linné refers in Rumphius’ D’ Amboin- ische Rariteitkamer, 1741, t. 40, fig. L, is a Parmophorus or Scutus and not a brachiopod. Solander, in his Portland Catalogue, 1786, No. 1718, p. 77, mentions Mytilus lingua or green Duck’s-bill Limpet, $., Humphrey's Conchology, pl. 2, fig. 2, from Amboyna. This last work was published in 1770 and the figured shell is certainly a Lingula. Though the name anatina has been in use for a long time, the species should be known as Lingula lingua (Sol.). Lingula sp. aff. exusta Reeve Lingula exusta Reeve, ‘Mon. of Lingula’, Conch. Icon., November 1859, pl. II, fig. 9. Davidson, ‘Mon. Rec. Brach.’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. pt. III, 1888, p. 217, pl. XXVIII, figs. 20-214. Dr. K. H. Barnard has recently submitted to me two specimens of a Lingula which he collected in 1912 at Delagoa Bay, Indian Ocean, S.E. Africa. The specimens are rather small and measure: 20 x 8 and 17:2 X 64mm. Both REVISION OF SOME SOUTH AFRICAN BRACHIOPODA yi are light brown in colour and resemble the Lingula exusta of Reeve, though about half the size of the type. Reeve describes the species from a specimen in the Mus. Cuming as follows: “The Swarthy Lingula. Shell oblong-ovate, rather thin, reddish-yellow, deeply stained with brown towards the margin, shining, umboes rather sharply beaked. Hab. Moreton Bay, Australia; Strange.’ He further states: ‘If L. murphiana be an Australian form of L. anatina, this might be regarded as the representative in the same locality of L. hians. Both species exhibit a peculiar coppery-redness, heightened in this to a dark, shining, swarthy tone of colour.’ Davidson (op. cit.) gives a fuller description mentioning its shining, darkish, coppery yellow-brown colour. He reproduces Reeve’s figure in fig. 20; fig. 21 being from a specimen from the same locality in his collection at the British Museum. He gives the length as 1 in. 7 lines and breadth 8 lines (10-6 < 17-1 mim.). Dall, in his 1920 List (#*p. 265) gives Brammo Bay, Dunk Id., N. Queens- land, coll. E. J. Banfield, 1 specimen, in the United States National Museum. ° This is much farther north than the original locality. I possess some half-dozen specimens of a Lingula from Singapore with similar colouring and shape to the Delagoa Bay examples, one or two being only slightly larger. There is need for further research in connection with the various forms of Lingula, as there appears to be much confusion with regard to species. The form known as L. anatina (now referred to lingua), though normally of a light green colour (as in Reeve, 1841, pl. II, fig. 10), is at times suffused with a bronzy colour, especially near the margins, or may be of a dark brown colour all over the surface, as in specimens in my collection from Amboyna, Moluccas (‘Siboga’), and Zamboanga, Philippines (coll. F. G. Pearcey, 1875). OrDER NEOTREMATA BEEcHER, 1801. Superfamily cRANIACEA Waagen, 1885. Family CRANIIDAE King, 1846. Genus Crania Retzius. 1781. Schrift. Berlin Ges. nat. Fr. 2 (4), p. 72. Crania roseoradiata sp. nov. (Pl. I, figs. 1-4.) Description: Shell (upper, or dorsal, valve) small, depressed, somewhat ovate in outline, with a short, almost straight posterior margin, apex low, subcentral, directed posteriorly; surface eroded, strongly wrinkled by growth-lines, and radiated by rose-coloured streaks or flames on a whitish background. Interior very finely granulose, margin thin and narrow beyond the heavy submarginal encircling ridge. Muscular impressions of posterior adductors well-marked, 8 ANNALS OF THE SOUTH AFRICAN MUSEUM oval, obliquely disposed on low tubercles within the posterior angles of the shell; imprints of protractors in small pits on external side and near the anterior extremities of posterior adductors impressions; imprint of median odd muscle very slight and situated on floor of valve between the posterior adductors; imprints of anterior adductors very distinct in ovate pits with strong posterior ridges, closer together than posterior pair and situated just behind the centre of the valve; small pits for insertion of retractors of arms immediately adjacent and external to the last; imprints of protractors of arms very slight on floor of valve in front of anterior adductors. Anterior to the muscle impressions are two shallow, ovate, cavities (one on either of the median line) showing indis- © tinctly the grooves of the pallial sinuses. Dimensions: Syntypes, Length, 9 mm., breadth, to mm. Mee yay soqueous 6 g mm. Catalogue No. A5665, in the South African Museum. Type-locality: 7 miles SSW. of Constable Hill, Saldanha Bay, dredged in 45 fms., with Megathiris sp. nov. Remarks: This is the first authentic Crania to be reported from South Africa. In 1931, Lt.-Col. W. H. Turton submitted for my opinion a very imperfect dead valve, 12 X 11 mm., from Port Alfred which I tentatively referred to Crania. In 1932, Turton figured the specimen as Crania sp. (pl. LXX, fig. 1843). Port Alfred lies nearly 500 miles to the east of Saldanha Bay. The nearest record of the genus appears to be a species from West Africa assigned by Reeve?’ to Crania rostrata Hoeninghaus. Four examples of this form are in the Cuming Collection at the British Museum (Natural History). Two are attached to fragments of dark-grey limestone and are the originals of the two lower figures of Reeve (8pl. I, fig. 3). Davidson!® places Crania rostrata among the synonyms of C. turbinata (Poli), but says ‘the so-termed C. rostrata figured by Reeve in his Conch. Icon. does not agree with Hoening- haus’s figures of his species’. He misquotes Reeve by giving ‘South Africa’ instead of ‘West Africa’ (vide Cuming) as the habitat. One of the British Museum specimens shows interior details and the ventral valve agrees closely with the figure of the interior given by Hoeninghaus.2® The position and number of the pallial sinuses are the same and there is the same arrangement PLATE II Figs. 1, 2, 7, 8, 14, 15. Megathiris detruncata (Gmelin). Messina, Mediterranean (J. W. J. coll.). Figs. 1-2: interior views of ventral and dorsal valves. Figs. 7-8: exterior views of same. Fig. 14: interior view of dorsal valve (fig. 2). Fig. 15: interior view of ventral valve (fig. 1). Figs. 3-6, 9-13, 16,17. Megathiris capensis sp. nov. Holotype, A. 5616a. Paratype, A. 5616b. Dredged in 45 fathoms, 7 miles SSW. of Constable Hill (Saldanha Bay). South African Museum. Figs 3, 4: interior views of ventral and dorsal valves of Holotype. Figs. 5, 6: interior views of ventral and dorsal valves of Paratype. Figs. 9, 10: exterior views of ventral and dorsal valves of Holotype. Figs. 11, 12: exterior views of ventral and dorsal valves of Paratype. Fig. 13: exterior view of dorsal valve of Holotype. Fig. 16: interior view of dorsal valve of Holotype. Fig. 17: lophophore in spirit specimen. Ann. S. Afr. Mus., Vol. XLI. Plate 1. South African Brachiopoda. REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 9 with regard to the muscular impressions and prominent rostrellum. The dorsal valve was not figured by Hoeninghaus so no comparison can be made here. Until a much-desired revision is made of the various forms attributed to either Crania anomala (Miill.) or Crania turbinata (Poli) little advance in knowledge can be attained. Hoeninghaus’s type of Crania rostrata appears to have come from the Mediterranean. The absence of the ventral valve in the new Cape species renders it impossible for a complete comparison to be made with the West African C. ‘rostrata’; but a comparison of the dorsal valve shows that the two forms are sufficiently distinct to merit a new name being given to the Cape species. It is of some interest to compare C. roseoradiata with the C. suessi Reeve.1§ The latter species was founded on five specimens in the Cuming Collection reported to have been collected by Mr. Strange at Sydney. It is described as being faintly tinged with orange-red. ‘The five specimens are in the British Museum and are now rather bleached: they possess somewhat yellowish rays. They bear a label in German? handwriting (?Pfeiffer’s) ‘Sidney, M. Strange’ (note misspelling). In general form they somewhat resemble the West African examples described and figured by Reeve as C. rostrata Hoen., and the coloured rays of the upper valve suggest a relationship with the C. roseoradiata of this memoir. Cyania suessi Reeve has also been recorded from Mast Head Reef, Queensland, 17-20 fathoms.”4 This record is based upon a few worn valves and there is some hesitation as to species. The occurrence of Crania at the Cape is an interesting addition to the South African fauna, but until a further study has been made it is difficult to ascertain whether it is of northern or southern facies. OrvEeR TELOTREMATA BeEEcHER, 1891. Superfamily TEREBRATULACEA Waagen, 1883. Family TEREBRATULIDAE Gray, 1840. Subfamily CANCELLOTHYRINAE Thomson, 1926. Genus Agulhasia King. 1871. Ann. Mag. Nat. Hist. (4), vil, p. 109. Agulhasia davidsom King Agulhasia davidson King, loc. cit., 1871, p. 111, pl. XI, figs. 1-7. Terebratulina (Agulhasia) davidson’ King. Davidson, ‘Monograph of Recent Brachiopoda’, Trans. Linn. Soc., ser. 2, vol. 1V, Zool. pt. I, 1886, pp. 36-7, pl. 7, figs. 1-5. Agulhasia davidson King. Thomson, Brachiopod Morphology and Genera (Recent and Tertiary), 1927, pp. 182-3, fig. 52 (after Davidson). Remarks: The original specimens of this interesting species were obtained from 45 to 60 fathoms, on the Agulhas Bank, South Africa, and were described by Professor W. King in 1871 (op. cit., supra). I have been unable to examine the original examples, but the excellent description by Davidson (pp. 36-7) 10 ANNALS OF THE SOUTH AFRICAN MUSEUM shows that the species is a remarkable one and sufficiently distinct from the young of some species of Terebratulina, as was at one time thought. This extremely small species is characterized by the beak being produced into a tubular rostrum and by the possession of a long pedicle-collar (see Davidson,” pl. 7, figs. 1a and 2). : Through the kindness of Dr. K. H. Barnard I have been able to examine three specimens of this species dredged by the Cape Government trawler s.s. Pieter Faure from two new stations, one off the south-eastern Province, and the other off Natal. The details are as follows: Lat. 34°S., long. 25°44’E. dredged (no depth given). 2 specimens. A. 5633. South African Museum. Size: L. 4:2, W. 2°5 and L. 4:0, W. 2-3 mm. Cape Natal (i.e. Durban), W. x N. 6 miles, 54 fathoms, 1 specimen. A. 5634. South African Museum. Size: L. 4:2, W. 2:7 mm. These specimens are smaller than the size given by Davidson, viz. 3 lines x 2 lines (= 6-3 X 4:6 mm.). The specimens are interesting, not only in providing evidence of an extended eastern range of the species, but also on account of the opportunity they offer of studying the peculiar foraminal characters. As pointed out by Thomson (8p. 74 and footnote), ‘the long beak is apicate, with a very long delthyrium which is partly filled by lateral trigonal deltidial plates, leaving an elongate pedicle-opening. ‘The pedicle-collar extends nearly to the cardinal margin, and has the effect of restricting the pedicle mainly to an opening on the cardinal margin.’ Davidson’s figures ('*pl. 7, figs. 1a and 2, repeated by Thomson, fig. 52) show this quite clearly. King (op. cit.) confused the deltidial plates with the area, and the pedicle- collar with the deltidium. Beecher (?°p. 390) referred to the deltidial plates as being obsolete; this being probably due to the imperfect descriptions by King and Davidson. Unfortunately details are lacking concerning the character of the animal, though the cardinalia and brachidium have been figured. The range of Agulhasia has been given as Cretaceous to Recent, but no Cretaceous forms have been cited. Terebrirostra or Lyra from the Cretaceous possesses a somewhat similar tubular beak, but with the foramen at the apex: it belongs to an entirely different Family. Genus Terebratulina d’Orbigny. 1847. C.R. Ac. Sc. Paris, XXV, p. 268. Terebratulina abyssicola (Adams and Reeve) (Pl. I, figs. 5-7) Terebratula abyssicola Adams and Reeve, Voy. Samarang, Moll., 1850, p. 72, pl. 21, fig. 5: Reeve, “Terebratula’, Conch. Icon., 1860, pl. 4, fig. 14. Terebratula radiata Reeve, “Terebratula’, Conch. Icon., 1860, pl. 3, figs. 7a-b. REVISION OF SOME SOUTH AFRICAN BRACHIOPODA II Terebratulina abyssicola, Davidson, ‘Mon. Rec. Brach.’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. pt. I, 1886, pp. 37-8, pl. 5, fig. 54. Terebratulina radiata, ibid., 1886, pp. 34-5, pl. 6, figs. 9-14. Terebratulina africana Turton, The Marine Shells of Port Alfred, South Africa, 1932, p. 260, pl. LXX, fig. 1842. Non Terebratulina radiata Baily, Quart. Journ. Geol. Soc., vol. XIV, 1858, p. 136, pl. VIII, figs. 3a-d (= Jurassic species from Crimea). Non Terebratulina radiata Moore, The Geologist, vol. III, 1860, p. 444, pl. XIII, figs. 11-14. (= Great Oolite species from Hampton Cliff, England). Habitat: South-east coast of South Africa, lat. 33°6’S., long. 28°11’E., 85 fathoms (= opposite East London), 3 large examples A. 5617, South African Museum: also 4 miles off Cape Morgan, 17 fathoms, 8 small examples, A. 5618, South African Museum. Remarks: ‘There has been much confusion in the past with regard to this species, its type locality, and its relation to the T. radiata Reeve. The latter is merely the adult form of 7. abyssicola Ad. and Rve., which antedates 7. radiata by ten years. The type-locality is ‘Gape of Good Hope (dredged at the depth of 120 fathoms); Belcher’, as given by Reeve. The reference to Corea in Reeve’s ‘Corrigenda’ is evidently a mistake. In the Zoological Department (Cuming Collection) of the British Museum are four specimens (three juveniles and one large example attached to some extraneous object) of T. abyssicola, with a legend on the bottom of the box ‘Hab. Cape of Good Hope, 120 fms. M.C. ?Type’. The attached specimen looks like that figured in the ‘Samarang’ Report (pl 21, fic. 5) and in Conch. Icon. (pl. 4, fig. 14), but is smaller - and more globular. It shows the beginning of a sulcus in the dorsal valve, and the striae are like those on the umbonal part of a South African Museum specimen (A. 5617). The British Museum also possesses specimens of T. abyssicola from Port Alfred, Cape Colony (Coll. Lt.-Col. Turton, 1903, 12.19. 463-5). Davidson’s figure of T. abyssicola (pl. 5, fig. 54) is a poor copy of that of Adams and Reeve. As far as one can judge from a casual microscopic examination, the punctae of the “Type-specimen’ of 7. abyssicola are the same as in Reeve’s type of T. radiata, being small and slit-like and of the same density. ‘The type-specimen of 7. radiata is preserved with two other examples in the Zoological Department at the British Museum (Cuming Coll.), while the originals of Davidson’s figures of the same species are in the Geological Department (Davidson Coll.) of the same Institution. The original of jDavidson’s T. radiata (pl. 6, fig. 11) has small, ovate punctae (slit-like) with the same density as Reeve’s type. The exact density of the punctae of the British Museum specimens has not been counted, but examples of T. abyssicola from other sources have been examined. Five somewhat dwarfed specimens from East London yielded the following results: in one the density was 352-368 per sq. mm., with dimen- sions of 30-40 X 15 »; in three others the density ranged from 448 to 512 per 12 ANNALS OF THE SOUTH AFRICAN MUSEUM sq. mm., the dimensions being 20-30 X 15 pu; while the fifth example was very densely punctate with 576 pores per sq. mm., with dimensions of 15-20 X 15 p. Four others from East London yielded 384 per sq. mm., and two from Port Elizabeth, 384-432 per sq. mm. The three large examples, A. 5617 (pl. I, figs. 5-7), in the South African Museum are larger than any figured by Davidson (pl. 6, figs. 9-14), their dimensions being: Length Breadth Thickness No. 1 24°0 1Q°2 15;0 mm. No. 2 27°5 19°3 LO-5) as No. 3 28-4 29°0 16:0 E The type of folding is sulciplicate (as Davidson, pl. 6, fig. 10), and all are rayed with black flames or streaks. ‘The beaks are mesothyrid and the deltidial plates disjunct. The number of pores per sq. mm. in these three examples ranges from 308-400, the dimensions being 30-40 X 15 p, on the outer side, and 10 X I0 pn on the inner. ) The eight small examples, A. 5618, in the South African Museum range in length from 4:5 to 13-4 mm., the beaks are submesothyrid to mesothyrid and the deltidial plates disjunct. They possess a uniplicate type of folding, the larger ones showing the beginning of the sulciplicate stage. ‘Some are marked with black streaks, and all agree closely with shallow water forms in my collection from East London and Port Elizabeth. Some of these appear to be stunted or dwarf examples and show irregularities of growth: the shell also is much thickened. This species under the name of 7. radiata was cited as a new brachiopod for Western Australia by W. B. Alexander, in 1914.26 The solitary specimen obtained had united deltidial plates, and this fact, together with its habitat, seems to suggest the Terebratulina (olim Terebratula) cancellata (Koch), since transferred to a new genus Cancellothyris by J. A. Thomson,?’ who also changed Puiate III Figs. 1, 2. Kraussina rubra (Pallas). Agulhas Bank, South Africa, 22 fathoms (J. W. J. coll.). Fig. 1: exterior view. Fig. 2: interior view of dorsal valve. Figs. 3-5. Kraussina crassicostata sp. nov. Holotype, A. 5615. Dredged in 23 fathoms, False Bay, Cape. South African Museum. Fig. 3: exterior view. Figs. 4, 5: interior views of dorsal and ventral valves. Figs. 6 and 9. Megerlina striata sp. nov. Holotype, Station 142, ‘Challenger’ Expedition (nearly 40 miles due south of Cape Point), 150 fathoms. British Museum (Nat. Hist.). No. 78.6.15.27. Fig. 6: exterior view. Fig. 9: punctation. Figs. 7 and 10. Megerlina pisum (Lam.). South Africa, J. H. Ponsonby coll. British Museum (Nat. Hist.). No. 99.4.14.3765. Fig. 7: exterior view. Fig 10: punctation in specimen in J. W. J. coll. (ex. Agnes Crane, from Cape, 150 fathoms). Fig. 8. Megerlina capensis (Adams and Reeve). Cape. Lombe Taylor coll. British Museum (Nat. Hist.). No. 74.12.11.386. Ann. S. Afr. Mus., Vol. XLI. Plate 1. - a fe» e",a' vo e¢? © + git,2n%s.s Fen”) § | PF Cette d. ahh eo % 4 , § t; - - as P a a & South African Brachiopoda. REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 13 the specific name to australis on account of cancellata being preoccupied by Tere- bratula cancellata Eichwald, 1829. But, unless the 7. cancellata Eichwald turns out to belong to Cancellothyris also, there appears to be no reason for the change of the specific name. Turton, in 1932 (pl. LXX, fig. 1842), figured as Terebratulina africana sp. nov. a small example from Port Alfred, near which place TJ. abyssicola is common. His figure shows a shell without apparent folding, the ventral valve having a submesothyrid foramen, disjoint deltidial plates and a pedicle-collar: the dorsal valve has rather long crura and an almost completed loop. The shell is thin and white with radiating striae and fine growth-lines, giving it a cancellated appearance, and does not seem to show evanescence of the striae, as in the next species. The absence of information regarding punctation renders it difficult to decide whether Turton’s shell is to be referred to T. abyssicola, which occurs in the near neighbourhood. Turton’s T. radiata (*4p. 260) is a synonym of the last species. Terebratulina meridionalis sp. nov. (Pi le fie. 3) Terebratulina caput-serpentis var. septentrionalis Davidson (non Couthouy), Voy. Challenger, Zool. vol. I, 1880, pp. 33-6, pl. I, figs. 6-9. Terebratulina septentrionalis Davidson (pars), ‘Mon. Rec. Brach.’, Trans. Linn. bee ser: 2. vol. LV, Zool. pt. I, 1886, pp. 26-32. Description: Shell of medium size, ovate, with curved sides and slightly truncated front, widest at the middle. Valves regularly convex, ventral slightly deeper than the dorsal and possessing a slight, narrow, distinct sulcus anteriorly. Folding uniplicate. Surface of valves finely capillate, striae strong and simple in umbonal region, increasing by bifurcation, and becoming - evanescent later; finely punctate. Beak short, suberect, foramen large, almost circular, submesothyrid, attrite; deltidial plates disjunct; pedicle-collar short. Hinge teeth without dental plates or swollen bases. Cardinalia in dorsal valve as is usual in Terebratulina, with socket-ridges fused to the crural bases and forming prominent ridges which project a little behind the dorsal umbo; small transverse cardinal process; no median hinge-plates. Crura long, descending branches of loop short, transverse band arched ventrally, crural processes united and transforming the loop into a ring. Muscular impressions feeble. Dimensions: Holotype, length 22 mm., breadth 18 mm., thickness 9:5 mm. Holotype and Locality: Specimen in the British Museum (Nat. Hist.), dredged by the ‘Challenger’ Expedition at Station 142, Cape of Good Hope (= nearly 40 miles due south of Cape Point) in 150 fathoms. Remarks: In his ‘Challenger’ Report, Davidson (p. 33, pl. I, figs. 6-9) recorded the Terebratulina septentrionalis Couthouy (as a variety of T. caput- serpentis = retusa) as abundant at Station 142, lat. 35°4’S., long. 18°37’E., off the Cape of Good Hope, in 150 fathoms, together with ‘Terebratula vitrea var. 14 ANNALS OF THE SOUTH AFRICAN MUSEUM minor’ and ‘Kraussina pisum’. In the same work he recorded the species from lat. 46°40’S., long. 37°50’E. (= Marion Island) in 150 fathoms, associated with ‘Platydia anomioides’ and *Waldheimia kerguelensis’ . In his later work (p. 28), Davidson appears to cast some doubt upon the authenticity of the Cape locality, but at the same time gives the Marion Island latitude and longitude in error for the Cape. By the courtesy of the British Museum authorities I have been able to examine the ‘Challenger’ examples and to compare them with typical T. septentrionalis Couthouy from Casco Bay, Maine, and other places. Some of these are preserved in the Zoological Department of that Institution, and I was surprised to find that one box, registered 78.6.15.28, contained two distinct species, one with evanescent striae and the other (a large, solitary example) agreeing with the true T. septentrionalis from Halifax, 83 fathoms (‘Challenger’ specimens, 78.6.15.24). I came to the conclusion that the true T. sepien- trionalis had probably been used for comparison and accidentally left in the same box with the others. At the same Institution I also had the privilege of examining some specimens preserved in spirit and coming from Station 142. These I found agreed exactly with the examples mentioned above which showed evanescent striae; and like them differed sufficiently from T. septentrionalis to merit the erection of the present new species. Compared with a spirit specimen of T. septentrionalis from Casco Bay, Maine, of almost the same size, viz. length 22 mm., breadth 17 mm., thickness 11 mm. (pl. I, fig. 9), the new species shows important differences in striation as well as in punctation. The striation of 7. meridionalis is evanescent (pl. I, fig. 8) and contrasts strongly with that of T. sepientrionalis (pl. I, fig. 9). The number of striae is about the same in both species. On the outer surface of the ventral valve, about the middle, 7. meridionalis has small ovate pores, 20-30 X 15 pw in diameter, and 256-288 per sq. mm. In T. septentrionalis from Casco Bay, the pores are small and rounded, 15-20 X 15-20 p, and 320-384 per sq. mm. at the middle of the ventral valve. Anteriorly both species are very similar in the possession of a slightly truncated front and low dorsal uniplication. The foramen in each species is also submesothyrid and there are rudimentary deltidia. In addition to the British Museum material I have had the advantage of comparing others in my own collection as well as examples in the South African Museum. These are all from essentially the same neighbourhood as the ‘Challenger’ specimens and confirm the conclusion reached above. Specimens from ‘Off Cape Point, S. Africa, 190 fathoms’ (in my own collection, ex South African Museum) agree exactly with the ‘Challenger’ example from Station 142 (p. I, fig. 8). They are smaller in size, but low uniplication is visible. They have small, ovate pores 256-320 per sq. mm., and 20-30 X 15 p in Size. Three small specimens in the South African Museum, A. 5623, dredged in 85 fathoms, off Cape Point, 10 miles S. 16° W., belong to the same species, and REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 15 a very young example in the same Museum, A. 5632, dredged at the same depth and place as the above, is probably the same. The T. ‘septentrionalis’ and the two other species of brachiopods recorded by Davidson from Marion Island require further investigation in the light of present knowledge. With regard to ‘Waldheimia kerguelensis’, the figures given by Davidson (pl. III, figs. 3-5) suggest a somewhat different species from that under the same name from Kerguelen Island (pl. III, fig. 1). Terebratulina sp. indet. Included in the South African Museum material are several juvenile examples of a Terebratulina which must remain unidentified pending further material. The specimens are all from the western side of the Cape—the opposite side from which T. abyssicola was dredged. ‘Three lots are from the neighbourhood of Table Bay, and two lots are from off Saldanha Bay. Some notes on these may be useful in connection with further dredgings in these waters. A. 5620. ‘Table Bay, 22 fathoms (South African Museum). There are three juvenile examples from this locality which possess a rather coarser and more distinct ornament than T. meridionalis. Near the umbo there are 13 or more strong, rounded striae which soon break up into fascicules of finer striae by bifurcation and trifurcation. ‘There are also a few inter- polated fine striae. All extend to the anterior margin with occasionally a second bifurcation at a late stage. The beak is short, suberect; foramen large and almost circular, submesothyrid, attrite. The deltidia in two examples are just touching at the points, but in the largest example they are disjunct (perhaps broken or worn by the umbo of the dorsal valve). In the smallest example the folding is uniplicate and in the two others slightly sulciplicate. The species presents some resemblance to 7. abyssicola, but differs in punctation. It has long oval pores on the outer surface ranging from 224 to 288 per sq. mm., the dimensions being 30 (-45) X 15-20 pw. On the inside of the valve they are round and measure 15 X 15 wp. A. 5621. 28 miles S. 76° W. of Lion’s Head, Cape Peninsula, 140 fathoms. Four juvenile examples (South African Museum). The punctation in these ranges from 272 to 288 per sq. mm., the diameters being 20 X 15-20 p, on the outer side. A. 5619. 34 miles S. 63° W. of Lion’s Head, Cape Peninsula, 154 fathoms. Four juvenile examples (South African Museum). The punctation ranges from 272 to 304 per sq. mm., with diameters of 15-20 X 10-15 p, on outer side. These two lots seem to agree with A. 5620, but are too young for critical comparison. ‘The punctae are near those of 7. meridionalis and less than in T. abyssicola. A. 5628. Off Saldanha Bay, 20 fathoms. Five juvenile examples (South African Museum). 16 ANNALS OF THE SOUTH AFRICAN MUSEUM The punctation ranges from 320 to 352 per sq. mm., with diameters of ?30 x 20 w (outer) and 10 X 10» (inner). One specimen has spicules in the genital sinuses of the ventral valve near those of T. valdiviae BI. A. 5622. 5 miles W. by S. of Constable Head (Saldanha Bay), 47 fathoms. One juvenile example (South African Museum). The punctation ranges from 308 to 320 per sq. mm., with diameters of 15 x 15 » (outer). ‘The striation is somewhat of the evanescent type and not that of T. abyssicola. The above two lots are puzzling, but are too immature for a definite con- clusion to be reached. Subfamily GRYPHINAE Sahni, 1929 (nom. nov., to replace Yerebratulinae Dall, 1870, which conflicts with the genus Terebratulina). Genus Gryphus Megerle von Mihlfeldt. 1811. Ges. nat. Fr. Berlin, Mag. V., p. 64. Gryphus capensis sp. nov. (Pl. I, figs. 10-13.) Description: Shell small, longitudinally oval, with somewhat straight sides and very slightly truncated front, widest at the middle and tapering posteriorly. Valves regularly convex, ventral deeper than dorsal; anterior commissure rectimarginate. Beak short, rounded and with no apparent ridges, incurved dorsalwards and truncated by a small, circular, marginate foramen, separated from the hinge-line by a symphytium: the foramen is also slightly labiate and has a short, but distinct, pedicle-collar. Interior of ventral valve smooth and with fairly strong muscle-impressions in the umbonal region; teeth small, situated at the basal angles of the symphy- tium: dorsal valve with cardinalia consisting of thin, concave hinge-plates uniting the socket-ridges and crural bases; small accessory outer socket-ridges; no median hinge-plates; cardinal process small and transverse; crural processes some distance down the almost parallel descending branches of the loop; transverse band narrow and arched ventrally; thin, thread-like septum separa- ting the distinct muscle impressions in the umbonal region. Dimensions: A. 5626 (Paratype), length 9-9 mm., breadth 6-5 mm. A. 5627 (Holotype), length 13:5 mm., breadth 9-7 mm. Type-specimens: As above; both in the South African Museum. Type-localities: Holotype: 29 miles SW. of Cape St. Francis, 75 fathoms. Paratype: 73 miles S. by W. of Cape St. Blaize, 125 fathoms. Remarks: In his Annotated List of the Recent Brachiopoda, the late Dr. W. H. Dall, in 1920,!* showed that the familiar generic name Liothyrina Oehlert must give way to the earlier name of Gryphus of Megerle von Mihl- feldt, 1811, the type species of which is the Anomia vitrea Born, 1778, a common Mediterranean species. two REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 17 The general characters of the new Cape species seem to place it in the Gryphus group rather than in that of Liothyrella Thomson. As pointed out in a previous paper,?® the Gryphus (olim Liothyrina) group comprises fairly large shells with a somewhat truncated front, and, in some cases, broad dorsal uniplication, together with a loop with long, almost parallel, descending branches, with crural processes a little distance down, and a broad transverse band, as in Gryphus vitreus, sphenoideus, cubensis, bartletti, and stearnsi: the Liothy- rella group, on the other hand, comprises broadly dorsally uniplicate oval. shells, usually of small size, with a rounded front, and possessing a loop of a short triangular form, with crural processes close in, and a very narrow trans- verse band, as in Liothyrella uva, notorcadensis, antarctica, etc. ‘The features in these two groups are correlated with certain types of spiculation in the animal (see Blochmann,!4” 1°), Though Gryphus capensis is smaller than the other species assigned to the genus, its loop bears considerable resemblance to that of G. stearnst or G. sphen- oideus, as figured by Blochmann (1*°pl. 39, figs. 29 and 23a); the transverse band is, however, much narrower. A small Gryphus-like species was dredged by the ‘Challenger’ Expedition in 1873, off the Cape of Good Hope, at Station 142, lat. 35°4’S., long 18°37’E., at a depth of 150 fathoms. Davidson (*pp. 29-30, pl. II, figs. 5-6) referred the species to Terebratula vitrea var. minor Philippi (now Liothyrella affinis Calcara). The specimens are in the British Museum: they are not so elongate as G. capensis, being slightly broader. A similar species was dredged by the ‘Valdivia’ Expedition in 1898, from the Agulhas Bank, and was described, but not named, by Blochmann in 1906 ('#4p. 699) and in 1908 ('*°pp. 605, 613, and 630, pl. 39, fig. 31). He compared it with the ‘Challenger’ specimen and concluded that both were specifically distinct from L. affinis in the character of the loop and in other features. In 1906 he referred it to the Gryphus (= Liothyrina) series and ranged it with the larger species G. vitreus, etc.—a group without certain spicules penetrating the bases of the cirri. In his distributional map of 1908 (‘*°pl. 40), however, it is placed with the series containing Liothyrella affinis, antartica, etc. Unfortunately the ‘Valdivia’ example has not been fully figured, the loop only having been illustrated by Blochmann (!*°pl. 39, fig. 31). Without a further examination and a comparison of the ‘Challenger’ and ‘Valdivia’ examples with the present species it is not possible to be sure of their identity. The habitat of the two former is some distance to the west of that of G. capensis, but does not rule out the probability of the species being identical. In connexion with the relationship of the South African brachiopod fauna, mention might be made here of other occurrences of species of either Gryphus or Liothyrella in localities not far removed from the South African region. Of some interest is the record of Terebratula cernica Crosse,2* which was obtained from the stomach of a fish taken at a depth of 80 fathoms off the Island of Mauritius. It is a larger species than G. capensis but nothing is known of its 18 ANNALS OF THE SOUTH AFRICAN MUSEUM loop or interior features. It was refigured by Davidson in 1886 ('’p. 16, pl. I, fig. 19) as Liothyris cernica, and was later referred tentatively to the Gryphus vitreus series by Blochmann ('*°p. 625, pl. XL). According to the latter, Studer also mentions a fragment of a similar Terebratulid obtained at Mauri- tius by the ‘Gazelle’ Expedition, 1874-6. Further, a Liothyrina (= Gryphus) sp. indet. was recorded by Dall (**p. 439, pl. 26, figs. 1-2) from south of Saya de Malha Banks, NNE. of Mauritius. In outward form it approached G. bartlettz (West Indies) ; its loop showed affinity with G. sphenoideus (West Mediter- ranean and East Atlantic, off Portugal, etc.); and the spiculae were closely allied to G. vitreus (Mediterranean and East Atlantic) (!’p. 46). Two examples of a further interesting species of Gryphus were dredged by the ‘Challenger’ Expedition in 1876, off Ascension Island, in 420 fathoms, and were regarded by Davidson (*p. 28, pl. II, figs. 10-11) as agreeing in every respect with the G. cubensis (Pourtales) from 100 to 300 fathoms off the Florida Reefs. Blochmann, in 1908 (!*°pp. 622-3, pl. 38, fig. 20) refigured one of the ‘Challenger’ specimens, and remarked upon its somewhat asymmetric outline. More examples are required in order to establish the exact status of the Ascension Island form. Family TEREBRATELLIDAE King, 1850. Subfamily MEGATHYRINAE Dall, 1870. Genus Megathiris d’Orbigny. 1847. C.R. Ac. Sc., Paris, xxv, p. 269. Megathiris capensis sp. nov. (Pl. II, figs. 3-6, 9-13, 16, 17.) Description: Shell small, transversely ovate, with a long straight hinge-line, almost or quite equal to the full width of the shell; rounded anteriorly; strongly convex, the ventral valve being considerably deeper than the dorsal; surface smooth except for rather strong growth-lines; test thick and visibly punctate; cardinal area in each valve, that on the dorsal valve being small and visible only at the cardinal extremities. Ventral valve very deep, with truncated rostrum, large submesothyrid foramen, rudimentary deltidial plates separated by linear grooves from the adjacent area which is high and triangular; interior furnished with two short, lateral septa near the anterior border, and one long, thin, median septum extending backwards from just within the anterior border to the umbo where it is high-standing and supports a distinct pedicle-collar; hinge-teeth fairly large, widely separated, and without dental plates; internal surface covered with white, rounded, granulations like shagreen. Dorsal valve shallow with slightly protuberant apex, furnished with three prominent septa thickened and denticulated along their crests; the two lateral septa, which commence within the anterior border, possess bulbous triangular apices with points directed towards the lateral borders of the shell, they extend backwards as low ridges to about the middle of the valve: the median septum, which has a bulbous rounded apex, commences at the same distance from the anterior REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 19 border and extends backwards as a low ridge to the hinge platform: cardinalia consisting of two prominent socket ridges, widely separated but united across the umbonal cavity by two broad, excavate, hinge-plates resting on the posterior prolongation of the median septum; no true cardinal process, the diductor muscles being attached to a vertically striated apical portion of the interior of the valve; crura very short, crural processes large, convergent; loop as in M. detruncata: lophophore ptycholophous. The shell shows no sign of folding, the anterior commissure being recti- marginate. | The lophophore is ptycholophous, as in Megathiris detruncata (Gmelin). (See plate II, fig. 17.) Dimensions: Holotype, A. 5616a, 7°71 X 5 mm. Paratype, A. 5616b, 7°5 x 5 mm. Type-specimens: As above, both in South African Museum. Type-locality: Dredged in 45 fathoms, with Crania roseoradiata sp. nov., 7 miles SSW. of Constable Hill (Saldanha Bay). Remarks: ‘The genus has not hitherto been recorded from South Africa, and its occurrence there is of extreme interest since it appears to be peculiarly northern as regards both recent and fossil occurrences. A Tertiary fossil form with some external resemblance to Megathiris has been recorded from Chile by Dr. A. Philippi (“p. 218, pl. 49, fig. 11). The species is described as Terebra- tula depressa and comes from Lebu, and the author suggests, from its extremely small size, that it is possibly not fully adult. ‘It resembles’, he says, ‘the T. detruncata of European Seas very much, which belongs to the genus Megathiris d’Orbigny.’ D. P. Oehlert, in 1888 (*p. 814), referred T. depressa Phil., to Megathiris. Von Ihering, in 1903 (*°p. 338), also placed it in the same genus along with Argiope barretiana Davidson which occurs in the West Indies and Florida (43- 461 fathoms) and off Rio de Janeiro (7o fathoms). The latter is now regarded as Argyrotheca barretiana. Von Thering (*p. 341) regarded the presence of a species of Megathiris in the Tertiary of Chile as confirming the Atlantic character of the Tertiary fauna of that country, the presence of this and other forms being explained easily by the free communication of seas in Central America, certainly in Eocene and probably in Miocene times. Until other, and more adult, examples of the ‘Terebratula’ depressa Philippi are forthcoming, there remains some doubt as to its generic relationship. It may be pointed out that Philippi’s name is pre-occupied by the Terebratula depressa Faujas 1799, a Cretaceous (Senonian) species from France and Holland.*” This was referred to Megathiris by d’Orbigny in 1847 (*p. 149) and to Cistella by de Morgan in 1883 (*p. 10 of reprint). Cistella is now Arg yrotheca Dall. The Cape species is very distinct from the genotype Megathiris detruncata (Gmelin) (pl. II, figs. 1, 2, 7, 8, 14, 15), which occurs in the Mediterranean and East Atlantic from Guernsey and the Scilly Is. to Madeira (16-100 fathoms), 20 ANNALS OF THE SOUTH AFRICAN MUSEUM and is recorded as a fossil from the Eocene, Miocene and Pliocene of south and eastern Europe.°? Megathiris capensis is larger and more transverse, in addition to being entirely smooth externally and more rugose internally. Jeffreys (*'p. 124, pl. 5, figs. ga-c) records a form of M. detruncata from Guernsey which has weaker costation: it resembles a horse’s hoof in shape, being longitudinally oval, instead of transversely oblong (as in Mediterranean examples); the costae are much fainter and do not extend to the anterior margin. He also refers to specimens in the M’Andrew collection from Madeira, which, though smaller, have the same form and sculpture. These, he considers, may therefore belong to a distinct and undescribed species. Sacco, in 1902 (“pp. 30-1), figured and briefly described several varieties of M. detruncata from the Italian Tertiary rocks. ‘These include a var. semi- laevis (>*pl. 6, figs. 17-19) in which the costation does not reach the margin which is smooth: it occurs frequently in the Middle Miocene and Lower Pliocene, and suggests Jeffreys’s form previously mentioned and an Italian Eocene species, referred to M. detruncata by Davidson (*pl. XXI, figs. 6-8). Another of Sacco’s varieties is var. perlaevis (>pl. 6, fig. 20), in which the valves are smooth or nearly so: it occurs somewhat rarely in the Middle Miocene and Lower Pliocene. Sacco’s figure of this variety is an external view of the dorsal valve: it strongly suggests the Cape species, but in the absence of an interior view of this fossil a comparison of the cardinalia, etc., with those of the present species cannot be made. Subfamily KRAUSSININAE Dall, 1870. The subfamily position of the genus Krausszna and some allied genera has been the subject of much difference of opinion in the past. Dall in 1870 (?p. 138) created the subfamily Kraussininae for Kraussina rubra and four other species including lamarckiana which later (in 1884) became the genotype of Megerlina. The same subfamily was adopted by Davidson in 1887 ('p. 118), and by Schuchert in March 1893 (*p. 160). But Beecher in 1893 (*’p. 391) placed Kraussina and Megerlina in his new subfamily Magellaniunae along with Magellama, Terebratella, and others, and at the same time created a new subfamily Dallininae to embrace Dallina, Macandrevia, Terebratalia, Laqueus, Miihlfeldiia, etc. At that time the genus Muiihlfeldtia included truncata and what is now known as Frenulina sanguinolenta. ‘The latter species, under the name M. sanguinea, was used by Beecher in illustration of one of the stages in the ontogeny and morphology of the Dallininae, the stage being called the ‘Muhlfeldtiform’ stage. In studies made in 1916 (°p. 24) I found that Miihlfeldiia truncata differed fundamentally from Frenulina sanguinolenta and I transferred the former genus to the Magellaniinae and changed Beecher’s stage name in the Dallininae from ‘Miuhlfeldtiform’ to ‘Frenuliniform’. This is now generally accepted. My reasons for the transfer were based upon the absence of dental plates in Miahlfeldtia truncata, as well as in the resemblance of the early loops stage to a similar stage of Terebratella dorsata, and the appearance of the REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 2 secondary loop before the appearance of the primary lamellae. It was pointed out for the first time in that paper that in the Dallininae the descending branches of the loop (i.e. the primary loop) were united to the septum at an earlier stage than in the Magellaniinae. In the same paper I also emphasized a hitherto unrecognized feature in the Dallininae of the universal presence (except in adults of the genus Dallina) of dental plates in the ventral valve. Though possessing close affinities with the Magellaninae, I was not fully satisfied in 1916 that Muhlfeldtia rightly belonged to that subfamily or to a new one altogether. The matter was taken up later in the same year by J. A. Thom- son (*!p. 498) and again in 1918 (p. 7). In these valuable papers he foreshadowed a separate subfamily to include Araussina, Megerlina and Miihl- feldtia, and in his later Manual (78p. 219) placed these three genera together with Pantellaria and Aldingia in the emended but almost forgotten subfamily Miihlfeldtiinae Oehlert 1887 (p. 1314). It is unfortunate that Oehlert in proposing this subfamily was not more precise in specifying the genera he proposed to assign to it, but one can assume, from the arrangement of the genera on later pages (pp. 1322-3), that it comprised Muihlfeldtia (M. truncata) [with Section Megerlina (M. lamarckiana) and Ismenia (I. pectunculus)], Kraussina (K. rubra), and Platidia (P. anomiozdes). The close relationship of Mihlfeldtiia, Megerlina and Kraussina as regards loop- development and spiculation was clearly demonstrated by Deslongchamps in 1884 (778p. 122). Davidson in 1887 ('’p. 118), placed Kraussina and Megerlina in the sub- family Kraussininae, and Dall in 1920 (!*pp. 374-6), followed the same pro- cedure, at the same time placing Muihlfeldtia and Pantellaria in the subfamily Mihlfeldtiunae. Though Kraussina, Megerlina and Miihlfeldtia are so closely related in loop characters, there are certain other minor points in which they differ. Kraussina does not possess a pedicle-collar in large adult shells: in these there is only a narrow thickened band fused to the floor of the umbonal cavity (as in Copto- thyris, Terebratalia, and other higher long-looped forms). In very young shells about 4 mm. and 6 mm. in size the band is slightly free anteriorly, which suggests that in still earlier stages there may be a free pedicle-collar. Megerlina and Miihlfeldtia possess a free or true pedicle-collar up to the adult stage (see 2°p. 25). Pending further detailed study of the animals I feel disposed at present to place Kraussina, Megerlina and Miuihlfeldtia in Dall’s subfamily Kraussininae which antedates Mihfeldtunae Oehlert by several years. Genus Kraussina Davidson. 1859. SB. Ak. Wien. Math.-naturw., pl. xxxvui, p. 189, 210 (pro Kraussia preocc.). From a careful study of numerous specimens and of the literature upon the subject, the only species which I consider as coming within the genus Kraussina 22 ANNALS OF THE SOUTH AFRICAN MUSEUM proper are A. rubra (Pallas), K. cognata (Sowerby), K. gardineri Dall, and K. crassicostata sp. nov., described in a later page. The area of distribution of this genus is of peculiar interest. So far as is known at present it ranges from the Cape of Good Hope to the Indian Ocean south of the Saya de Malha Banks. The species restricted to the Cape itself are three in number, viz. A. rubra, cognata and crassicostata. ‘The Indian Ocean form is K. gardinerr. tis seen from the above that the distribution of the genus is somewhat discontinuous, there being no record of its occurrence between Port Alfred and the Saya de Malha Banks, except a reference by Davidson, in 1887 (1%p. 120), to some small specimens erroneously described by Gray in 1872 under the mistaken name of Terebratula truncata (= Miihlfeldtia truncata). They were found attached to Ascidia, and to the stems of large algae, off the coast of Natal. ‘These specimens are referred to Kraussina rubra by Davidson. The occurrence of K. rubra here appears to me to be extremely doubtful in view of the direction taken by the Mozambique Current, viz. towards the Cape. Gray may have been correct in his attribution of the species to Miihl- feldtia truncata. In 1921 (*p. 49) I referred to the presence in the British Museum (Natural History) of an immature example of this typical Lusitanian species, labelled ‘S. Africa. J. H. Ponsonby coll. 1900. 6.13.4’. Unfortunately full details as to habitat and depth have not always been recorded, hence it is impossible at present to give the bathymetric range of Kraussina, but, from the available data and from the thick-shelled character of the Cape species, the genus appears to be one of fairly shallow water. It is also to be regretted that in most of the literature dealing with K. rubra one finds no more precise locality than ‘Cape of Good Hope’. More data on these points are badly needed. The genotype of Araussina is the Anomia rubra Pallas 1766. ‘The original generic title proposed by Davidson in 1852 (®°p. 369) for the reception of this species was Araussia, but this name having been used by Dana for Crustacea earlier in the same year, Kraussina was suggested by Suess in 1859 (*%p. 210) in collaboration with Davidson, who confirmed it in 1861 (*4p. 39). Davidson, in 1887 (}*p. 118), included seven species in the genus, five in Kraussina proper (viz. rubra, cognata, deshayesi, pisum and atkinsom), and two in the subgenus Megerlina (viz. lamarckiana and davidson). Only two of these are strictly referable to Kraussina, viz. rubra and cognata. ‘The generic position of the remaining five forms will be dealt with in the sequel. Kraussina rubra (Pallas) (Pl UT: figs. m7) Anomia rubra Pallas, Misc. Zool., 1766, p. 182, pl. XIV, figs. 2-11. Terebratula capensis Kuster (not Adams and Reeve), ‘Mart. and Chem.’, Conch. Cab., 1848, p. 32, pl. 3, figs. 15-17. Terebratula capensis Krauss (not Adams and Reeve), Siidafr. Moll., 1848, p. 32, pl. Il, fig. 10. REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 23 Terebratula (Kraussia) rubra Reeve, ‘Mon. Terebratula’, Conch. Icon., 1861, pl. IX, fig. 37. Kraussina rubra Davidson, ‘Mon. Rec. Brach.’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. pt. II, 1887, pp. 119-21, pl. XX, figs. 19-23. Remarks: This well-known species has been met with at several localities around the Cape, but little is known as to its range in depth. It has been seen by the writer from Port Elizabeth, Algoa Bay and Port Alfred, and from the Agulhas Bank, 22 fathoms. A figure of a specimen from the last locality, agreeing closely with Pallas’s original, is given here (pl. III, fig. 1) for com- parison with the new species described in a later page. Kraussina rubra has been described and figured by several authorities and appears to present a certain amount of variation in outward form. ‘The Anomia rubra, as figured by Pallas, is a transversely oval and costate shell, without apparent fold or sinus; but many specimens of equal size which I have examined show a strong ventral fold and dorsal sulcus: these features are well marked from the umbonal region. In these the shells are ventrally uniplicate or sulcate. Some forms tend to become more elongate than others. ‘There is also some slight variation in the intensity of the costae. It is not my intention to burden this paper with a full revised description of Kraussina rubra, but some important details are necessary in order to emphasize certain differences between this genus and that of Megerlina dealt with in a later page. In the interior of the dorsal valve of K. rubra the cardinalia consist of two divergent socket-ridges bounding the dental-sockets. Lying between these ridges is an umbonal callosity with two eye-like depressions on its surface. There are no excavate hinge-plates, even in shells as small as 4 mm. in size. At the apex a small cardinal process is present. A grooved median septum rises from the umbonal callosity and extends some distance down the interior of the valve. From its anterior extremity arise two short, deviating lamellae exten- ding towards the ventral valve: these fork at their ends, giving rise to anterior and posterior processes. ‘These processes are usually quite short, but in some specimens the posterior processes are in the form of long narrow ribbons extending backwards and slightly inwards towards each other: they possess hook-like extremities. The loop is essentially adult in shells 6 mm. long. The dorsal valve also possesses a slight, but distinct, cardinal area, a feature of unusual occurrence, found also in the genera Miihlfeldtia and Megerlina (see 1”). There are a few spines within the margin. In the ventral valve the features of importance are the absence of dental- plates and of a free pedicle-collar. As previously stated, in adult shells, there is a thickened band in place of the latter, but in very small specimens this band may be free anteriorly. The dense character of the shells of this species renders it somewhat difficult to make a study of the punctation. I have, however, succeeded in one or two cases with interesting results. Adult shells have large oval pores on the 24 ANNALS OF THE SOUTH AFRICAN MUSEUM exterior, these pores being denser and larger in the grooves than in the costae. In specimens from Port Elizabeth I have counted as many as 256 pores per sq. mm. in the grooves, as against 168 pores in the costae. In the former they measure 60 X 30-35 » , and in the latter, 45-50 x 15-20 wu. On the inner surface the dimensions are 15 X 15 p for all pores. The young specimens, however, exhibit some interesting differences both as regards punctation and interior details. ‘These are perhaps worth recording. In a specimen 6 mm. long, on which fine costation is visible, the punctae appear to be evenly distri- buted with no differentiation in the grooves or on the costae: they average 244 per sq. mm. at the middle of the ventral valve. Internally this valve has a narrow band within the apex which is slightly free anteriorly. The dorsal valve is like the adult, without excavate hinge-plates; and has a loop consisting of two very broad divergent lamellae, slightly bifurcated at their extremities. Widely-spaced spinules occur just within the margin of the valve. In a smaller specimen 4 mm. long, the punctae are moderately round with no differentiation, being 188 per sq. mm. _ Internally it is much the same as the other, but the extremities of the divergent lamellae are without bifurcations. The spiculation of the mantle of Kraussina rubra was studied by Deslong- champs in 1864 (*p. 25, pl. II, figs. 10-12), and again in 1884 (3*°pp. 121 et seq. and 160, pl. XIX, fig. 7). The spicules are of a very special form and very small. ‘Two systems exist, one in the mantle, the other in the pallial sinuses. Spicules also exist in the walls of the visceral chamber and in the arms. Kraussina cognata (Sowerby). Terebratula cognata Sowerby, Thes. Conch., vol. 1, 1847, p. 346, pl. 68, figs. 12-14. Terebratula cognata Kiister, in ‘Mart. and Chemn.’, Conch. Cab., Bd. VII, I, 1848, p. 46, pl. 4, figs. 5-6 (as figs. 3-4 in text). Terebratula (Kraussia) cognata Reeve, ‘Monogr. Tereb.’, Conch. Icon., 1861, pl. IX, figs. 38a, b. Kraussina cognata Davidson, ‘Mon. Rec. Brach.’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. pt. II, 1887, pp. 121-2, pl. XX, figs. 24-6 (var.? 27-30). Remarks: According to figures and descriptions this species is somewhat subtrapezoidal or elongated, with a nearly straight hinge-line almost as long as the width of the shell. The colour is pale yellow and the surface covered with radiating costae. Chemnitz, in 1785 (#p. 78, pl. 76, figs. 688a, b) described and figured a shell from the Cape as ‘Cognata Anomiae craniolaris basi perforata’, but from his poor description and illustrations and from the fact of his not being a binomial writer, the Terebratula cognata of Sowerby is regarded as the type. Davidson gives copies of the figures of Sowerby and Reeve which show the type of folding as sulcate. Unlike Kraussina rubra, this species increases in length but retains the mega- thyrid type of cardinal margin. REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 25 This species requires more careful study and more material. I have only been able to examine one or two indifferent specimens which appear to agree with the description of the species. In these the cardinalia, etc., agree essen- tially with K. rubra, and the interior of the dorsal valve is very spinous just within the margin. Spicules are present in the ventral and dorsal mantles. Owing to the almost smooth condition of the surface of the valves, there is no differentiation in the punctation. ‘The pores about the middle of the ventral valve are small, even, and widely-spaced, and number 160-176 per sq. mm. They are 15 X. 15 p» in size, but a few reach 20 X 20 wp. Habitat: South Africa, near the Cape of Good Hope. Note. Dr. K. H. Barnard has recently submitted a large example of a Kraussina containing the dried animal dredged in 10-14 fathoms at Saldanha Bay by the Preter Faure. Its registered number in the South African Museum is A. 5607. It is said to have been identified as Araussina cognata by G. B. Sowerby and may be this species. The dimensions are: L. 32:2, W. 29:5, D. 18-3 mm. The earlier growth-lines suggest a shell wider than long, as in Kraussina rubra. ‘The shell is hardly sulcate and of a pale yellow colour, and possesses moderately strong radiating costae. The occurrence of the genus at Saldanha Bay is of extreme interest: it probably reached the South Atlantic coast by the’aid of the Benguela Current. Two other interesting brachiopods are recorded in this memoir from this neighbourhood, viz. Crania roseoradiata sp. nov. and Megathiris capensis sp. nov. Kraussina gardinert Dall Kraussina gardinert Dall, “Brach. of the Sea Lark Exped.’, Trans. Linn. Soc. Lond., ser. 2, Zool. vol. 13, 1910, p. 440, pl. 26, figs. 3-6. Description: Shell rude and solid, greyish white, similar in shape to Muhl- Jfeldtia truncata when young with shell 12 x 13 mm. in size. During growth, the outline elongates proportionally more that it widens, and an adult measures 24°5 mm. in length by 23 mm. in width at the broadest part. Outline then roughly rhombic. Dorsal valve slightly sulcate; ventral valve with corre- sponding convexity. Latter valve slightly more convex than dorsal: total thickness of adult about halfits length. Beak short and wide; peduncle short; foramen wide and incomplete, margins more or less eroded or defective; wide flattened area of irregular shape on each side of foramen. Surface of valves with coarse, rounded costae, strongest mesially, with sub- equal roundly excavated interspaces: costae mostly continuous from beaks, and increasing rather by bifurcation than by intercalation: about forty costae in adult, the laterals being finer and closer-set than the others. Interior of valves strongly calcified; spinules within margin. Habitat: Station C1, Indian Ocean, south of the Saya de Malha Banks, in 123-153 fathoms (dredged by J. Stanley Gardiner, after whom it is named). 26 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks: Judging from Dall’s figures, the cardinalia and brachidium of this species are essentially of the Araussina rubra type, but there appears to be more excessive calcification in the umbonal cavity. The occurrence of Araussina in the Indian Ocean is of great interest. Asso- ciated with the species was a curious intraplicate Rhynchonellid (Hemithyris sladem Dall) which is tentatively referred to the Early Tertiary New Zealand genus Aetheia by Thomson (8p. 157); and a Liothyrina sp. indet., having some affinities with Mediterranean forms (see also Jackson,’ p. 46). Kraussina crassicostata sp. nov. (Pl. III, figs. 3-5.) Description: Shell solid, subcircular, about as broad as long, with a broad, straight hinge-line somewhat shorter than the greatest width of the shell (= submegathyrid). Valves convex, the ventral much more so than the dorsal; sulcate (i.e. with a shallow sulcus in the dorsal valve opposed to a fold in the ventral valve); margins waved by alternate multicostation. Cardinal area in both valves. Surface covered with 10-12 coarse, angular, radial ridges, strongest mesially and becoming obsolete near the cardinal angles, separated by angular interspaces: ridges continuous from beak, broadening anteriorly, and increasing by bifurcation: they are crossed by fairly strong growth lines. Colour of surface reddish; punctate. Beak well-truncated, foramen very large (due to wear), and incomplete anteriorly,* slight deltidial plates, narrow sessile pedicle collar (i.e. thickened band) within the umbonal cavity. Interior of ventral valve fairly smooth, with distinct muscle impressions posteriorly; bluntly ridged round the margin (= reverse of external ornament) ; row of small spines just within the margin; teeth small, without dental plates, situated at anterior angles of inner margins of the two widely-separated portions of the transversely-grooved area. Interior of dorsal valve smooth, with well- defined muscle impressions, bluntly ridged margin, and row of spinules, as in ventral valve: cardinalia and brachidium essentially of the Kraussina rubra type: no hinge-plates proper, their place being occupied by two eye-shaped pits (for insertion of the pedicle-muscles) in a thick callus in the umbonal cavity between the distant and thick socket-ridges. Cardinal process small and prominent, united laterally to the posterior bases of the socket-ridges by narrow, vertically striated, muscular impressions, within the posterior margin of the umbonal cavity; rudimentary outer socket-ridges bounding the widely separated portions of the area of this valve. Mesial septum distinct, extending from the umbonal callus (where it appears to be grooved on its upper surface) to about the centre of the valve, where it supports two, short and narrow, divergent and ventrally directed lamellae.+ Dimensions: A. 5615 (Holotype), length 16:2 mm., breadth 16-2 mm., thickness 11:3 mm. * Owing to wear it is not possible to define the type of foramen. + In the holotype the extremities are broken, but in another example the extremities are flattened and forked (as in K. rubra). REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 237) Type-specimens: Holotype and two metatypes, No. A. 5615, in the South African Museum, Cape Town. Type-locality: False Bay, 23 fathoms. Remarks: This species differs from the genotype, Kraussina rubra, in posses- sing much coarser and less numerous costae which tend to become obsolete near the cardinal angles; the ventral valve is also deeper. Avaussina cognata (Sow.) is a longer species with a flattened dorsal valve, and of a pale yellow colour: Kraussina gardineri Dall is greyish in colour and possesses a more calcified cardinalia, etc. The external surface of the ventral valve has large ovate punctae in the grooves with diameters of 45 < 30 », and about 256 per sq. mm., while the costae have smaller ovate punctae of less density: internally the punctae are very small and round, about 10 X I0 wp. Spicules are present in fragments of the ventral mantle adhering to the interior of the shell, especially in the sinuses. In some specimens in the Manchester Museum of apparently the same species and dredged with Allopora nobilis Kent near the Cape of Good Hope, in 26 fathoms, the punctae are similarly diversified, but are somewhat larger and rather less in relative density than in A. 5615. Genus Megerlina Deslongchamps. 1859. Ktud. crit. Brachiop., p. 159. 1884. Bull. Soc. Linn. Normand., viii, pp. 210, 242. Having dealt with the species appertaining to the genus Kvaussina proper, attention may be directed to other forms which in the past have been errone- ously referred to that genus. Three related South African species are included in this group which may be termed the ‘Terebratula pisum series’. Davidson, in his description of the ‘Challenger’ Brachiopoda (°p. 54), committed a most remarkable error in attributing to the Terebratula pisum Lamarck an altogether different species. I have examined the Cape specimens figured and described by Davidson as Kraussina pisum (Lam.), in the ‘Challenger’ Report (°pl. IV, figs. 7, 7a, 7b and 8) and later in his “Monograph of Recent Brachiopoda’ ('%p. 123, pl. XXI, figs. 1-4). The figure 7a of pl. IV of the former report is reversed in printing from the stone (as are some others in the same work): that in the latter report (pl. XXI, fig. 2) is not reversed. The latter, and adjacent figures, however, are not so correct in detail as those in the earlier work, as they do not show the strong ventral plication of the species in question. Davidson’s specimens are preserved in the British Museum: figs. 7, ya and 7b, are the same shell drawn in different views and repeated in ‘Recent Brachiopoda’, pl. X XI, fig. 2; this shell is in the Zoological Department of the above Institution and is registered No. 78.6.15.27. Figures 1, 1a, 1b, 3 and ga of pl. X XI are| drawn from other specimens now preserved in the Geological Department of the same Institution (Reg. No. B. 12405). Figure 8 (interior 28 ANNALS OF THE SOUTH AFRICAN MUSEUM view) of pl. IV (‘Challenger’ Report) is repeated as fig. 4 of pl. X XI, in the Monograph. ! The above species is certainly not the Terebratula pisum of Lamarck, 18109, as, according to the original description (*pp. 245-6), this is a small, smooth, subglobose red shell, resembling a cherry stone, and 9 mm. wide. The habitat given is ‘Isle de France’ (= Mauritius). Davidson’s ‘Kraussina pisum’ bears costae, somewhat like Megerlina lamarckiana (Dav.), but finer, and presents no resemblance to a cherry stone. It is also of a uniform light yellow colour, with no red markings at all. This species does not possess the Kraussina-type of cardinalia and brachial support, but shows a greater resemblance to Megerlina in these features. What I regard as the true Terebratula pisum Lamk. possesses essentially the same type of cardinalia and brachial support and must be placed in the same genus. A new specific name, therefore, is required for Davidson’s shell. One other species, viz. T. capensis Ad. and Rve. = deshayest Dav., also comes into the same genus, viz. Megerlina. Megerlina striata sp. nov. (Pl. 111; figs. 6; 9.) Kraussina pisum (non Lamk.) Davidson, Voy. Challenger, Zool. vol. I, 1880, p- 54, pl. IV, figs. 7-8. ‘Mon. Rec. Brach.’, Trans Linn. Soc., ser. 2, vol. IV, Zool. pt. II, 1887, p. 123, pl. XXI, figs. 1-4. Description: Shell small, transversely oval (except in young); colour yellowish-white. Dorsal valve slightly convex, with a distinct central longi- tudinal depression extending from a prominent umbonal swelling to the anterior margin and increasing in width rather rapidly. Hinge line nearly straight, about two-thirds the width of the valve. Ventral valve deeper than dorsal, longitudinally keeled. Folding sulcate. Beak slightly incurved, with a rather large incomplete foramen, bounded laterally by two small deltidial plates which curve upwards towards the dorsal umbo but fail to meet; beak- ridges very distinct, leaving a small flattened cardinal area between them and the deltidial plates. Foramen submesothyrid. Surface of valves covered with numerous distinct radii which increase by bifurcation and interpolation: cardinal angles smooth (i.e. radii absent on these portions) ; concentric growth- lines present at variable intervals. Interior of ventral valve with traces of radii like the exterior but reversed; spinous within margin. Pedicle-collar sessile, closely appressed to apex of umbonal cavity but slightly free in front especially near bases of teeth. The latter are comparatively small and not supported by dental plates. Interior of dorsal valve covered with rows of pustules radiating from apex; spinous within margin. Rudimentary area present in this valve. The cardinalia (as seen in immature shell) consist of two rather stout diver- gent socket-ridges, excavate below, and supported on inner sides anteriorly by two spurs which extend to the median septum and enclose two eye-shaped muscular pits. Cardinal process feeble. Median septum extending from near REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 29 apex to centre of valve where it gives off two broad, thin diverging lamellae, the upper extremities of which are slightly forked, the posterior forks curving inwards to some extent. Midway down each lamella, on the exterior or dorsal face, an accessory process, or ledge, is present, analogous to that in Megerlina lamarckiana, but less developed. Shell with large subcircular punctae. Dimensions : Length Breadth Thickness Holotype 1. (pl. III, fig. 6) I4°1 16-2 5°9 mm. Paratype 2. 13°5 14°9 Boles. Ny 3. (interior details) 9°3 8:5 Ce NaeweiNo. 1 — pl. IV, fig. 7 of ‘Challenger’ Report, and pl. XXI1, fig. 2 of Recent Brachiopoda. Habitat: Station 142, lat. 35°4’S., long. 18°37’E., off Gape of Good Hope, 150 fathoms (‘Challenger’ Expedition). Type in the British Museum (Nat. Hist.) Registered No. 78.6.15.27. Remarks: ‘The above description is based upon three examples in the Zoological Department of the British Museum, the largest of which was figured by Davidson, as stated above. Owing to the risk of damage in opening the two large examples, the details of the cardinalia and brachidium are taken from the smallest specimen (N. 3), but these have been checked to some extent by No. 2 which was partly opened. ‘There is reason to believe that No. 1 would exhibit a slightly more adult condition of the loop, etc. Two other examples of this species were also figured by Davidson (**pl. 21, figs. 1 and 3) and are now in the Geological Department of the above Institu- tion. These have also been examined externally and found to agree with those above. The shells of this species have much larger punctae than those of M. pisum and M. capensis. The pores are subcircular in outline and range from 208 to 240 per sq. mm. about the middle of the ventral valve. The size of the pores in No. 1 is 45-50 X 45-50 » (outer surface) (pl. III, fig. 9). They appear to be evenly distributed over the radii and grooves. The South African Museum possesses the following specimens which seem to belong here: A. 6410. 1 specimen (broken), dredged in 40 fathoms, St. Sebastian Bay, Agulhas Bank (K. H. Barnard, 1922). A. 5663. 5 specimens dredged off East London in 32 fathoms, by the Preter Faure. A. 7690. 2 specimens (locality and depth unknown). Preter Faure coll. A. 5605. 1 specimen (broken) dredged off Cape St. Blaize (depth unknown). Pieter Faure coll. The shells are mainly transversely oval, sulcate, and of a light yellowish colour, except one shell of A. 7690 which is white, and possess rather strong radii. In A. 5605 the radii are evanescent. 30 ANNALS OF THE SOUTH AFRICAN MUSEUM The four larger shells of A. 5663 measure: 10 X 10°7, 9°9 X 10, 10 X 9°83 and 8-5 x 8-7 mm. Megerlina pisum (Lamarck) (PIM fos) 7, 10.) Terebratula pisum Lamarck, Animaux sans Vert., vol. VI, 1819, p. 245. (Text written by Valenciennes owing to Lamarck’s blindness.) Terebratula natalensis Krauss in Kiister, Conch. Cab., von Mart. u. Chem., Bd. VII, I, 1844 and 1848, p. 36, pl. 2b, figs. 4-7. (Plate published, 1844, text, 1848.) Terebratula pisum Lamk. Sowerby, Thes. Conch., I, 1846, p. 345, pl. 69, figs. 37-9- Terebratula algoensis Sowerby, Thes. Conch., 1, 1847, p. 362, pl. 71, figs. g1-2. Terebratula natalensis Krauss, Siidafrikanischen Mollusken, 1848, p. 33, pl. 2, figs. Ila-c. ? Terebratula (Kraussia) pisum Lamk. Reeve, Conch. Icon., vol. XIII, 1861, pl. 9, fig. 36a (non 36b). 7 Kraussina atkinsont (non ‘T. Woods), E. A. Smith, Journ. Conch., vol. X, 1901, p. 116. Description: Lamarck’s description (*pp. 245-6) is as follows: ‘Térébratule pois. Terebratula pisum. T. testa minuta, subglobosa, laevi, subantiquata, rubella margine integro antice valde sinuato. Habite a ’Ue-de-France. Par. M. Mathieu. Mus. no. Petite coquille semblable a un noyau de cerise, ne le surpassant pas en grosseur. Elle a 9 millimétres de largeur. The following is an amplified description, based upon South African examples. Shell small, transversely oval, with a subterebratulid type of cardinal mar- gin; valves about equal in depth; dorsal valve sulcate, ventral broadly carinate; type of folding sulcate. Surface nearly smooth, with faint traces of costation; test finely punctate. Colour milk-white or slightly suffused with red. Beak suberect; foramen large, submesothyrid, incomplete; deltidial plates discrete, small, trigonal. Pedicle-collar sessile, slightly free in front. Hinge-teeth without dental plates. In the dorsal valve the cardinalia consist of two promi- nent, divergent, socket-ridges curling over the dental sockets posteriorly, and supported at their anterior corners by rudimentary excavate hinge-plates in the form of two spurs extending inwards towards the centre line of the valve: these enclose an imperfect hinge-trough; cardinal process small. The brachi- dium rises from the floor of the valve near the middle as two ventrally directed deviating lamellae which bifurcate slightly at their extremities: outer sides of the lamellae with two short accessory processes, representing the anterior portions of the descending branches: the lamellae extend backwards as low convergent plates forming a pseudo-septum as far as the imperfect hinge- trough. . REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 31 Dimensions: Length Breadth Thickness Figured specimen (pl. III, fig. 7) 10°3 11 5°6 mm. Others. South Africa 10-7 10°5 Baha sae Umkomaas Q°4. 10°2 ee » 9°3 LOFT 4°4 » 8:6 9°3 4°3 Habitat: Umkomaas, Natal; Port Elizabeth, Algoa Bay; Durban, Natal; East London, etc. Remarks: ‘Though originally described from Mauritius, there are several specimens from such places as Port Elizabeth, Durban, etc., both in the British Museum (Nat. Hist.) and in private hands, which conform closely with Lamarck’s description. As Lamarck’s type was never figured, it seems desirable that illustrations should be given here of the South African specimens upon which my conclusions as to identity are based. Among the specimens preserved in the Zoological Department of the British Museum are seven from Umkomaas, Natal (J. H. Ponsonby coll. 1901.9.23.61-7), unfortunately without precise details as to depth, etc. ‘I'wo other and larger specimens together with several juveniles are also in the same Institution, and are labelled simply ‘S. Africa. J. H. Ponsonby. 99.4.14.3765-3771'. These two sets, as well as numerous others in private collections, including specimens in my own collection from Durban, Natal, and Port Elizabeth, Algoa Bay (both ex. H. McClelland, 1922), have been used in the study of the species. A specimen from the British Museum set marked ‘S. Africa’ has been selected to illustrate the exterior features (pl. III, fig. 7) and both this and others have been used for the interior details. The punctae in the test of this species differ from those of M. striata in being more oval and on the whole less numerous (pl. III, fig. 10). Judging from the examination of about 15 specimens the number ranges 120-220 per sq. mm. at the middle of the ventral valve. ‘The most prevalent numbers, however, are 160-190. The dimensions of the punctae are a little variable according to the presence or absence of strong growth-lines: they average 40-45 25-30 p on outer surface, and 15-20 X 15-20 w on inner surface. Large spicules occur in the pallial sinuses, but are scanty in the cirri. Victor Sganzin (*%p. 12) in his reference to the Terebratula pisum Lamk. says it is very small and extremely rare: it is found at a great depth in Tombeau Bay, Mauritius, and dredging is necessary in order to obtain it. The Terebratula natalensis of Krauss (®p. 33, pl. 2, figs. 11a-c) appear to belong to the species under review. His figure (fig. 11b) of the hinge-processes and brachial-support is remarkably accurate except that he appears to have overlooked the accessory ledges on the outer sides of the deviating lamellae. In his description he speaks of the shells being generally white in colour, more rarely flushed with red, and notes the sulcate character as well as the finely striated or almost entirely smooth condition of the surface of the valves. He records the species as living in great numbers at Natal Point (= Durban) in a depth of some fathoms, on stones, Cardita variegata, Arca kraussi, etc. 32 ANNALS OF THE SOUTH AFRICAN MUSEUM With regard to the species described and figured by Kiister (**p. 36, pl. 2b, figs. 4-7) as the 7. natalensis of Krauss, there is a little uncertainty as to absolute identity. His figures seem to suggest that he may have had before him an example of the Terebratula capensis Ad. and Rve. (= Kraussina deshayesi Dav.). The costation shown in the figures is stronger than in M. pisum (Lamk.). Sowerby (*°pp. 345-6), in his remarks on the species, says, ‘also found at Sydney by Mr. Jukes’. This is undoubtedly an error: the Sydney citation probably refers to the Megerlina lamarckiana (Dav.), which was unknown at that time. The Terebratula algoensis G. B. Sow., was described and figured in 1847 (7p. 362, pl. 71, figs. g1-2) from a single bleached ventral valve in the British Museum (N. H.), Zoological Department, labelled ‘Algoa Bay; J. S. Bowerbank’. I have examined the specimen, which bears a registration number (rather indistinct), viz. 42.12.19.26, on the interior of the valve, and consider it to be a somewhat irregularly-grown valve of M. pisum (Lamk.). The number of punctae per sq. mm. is about 160, and the dimensions are; outer, 45-50 X 25-30 w; inner, 15-20 X 15-50 p. | In the Journal of Conchology for 1901, E. A. Smith (*%p. 116) records Kraussina atkinson (T. Woods)—a Tasmanian species—for Algoa Bay, Cape Colony (Brit. Mus., J. H. Ponsonby). I have examined the specimens in question—four in mumber B.M. 1900.6.13.5-8)—and find them to be un- doubted juveniles of M. pisum (Lamk.). The number of punctae per sq. mm. ranges from 160 to 182, and the size of the pores on the outer surface is 30-40 X 25-30 p. In the ‘Kraussina atkinson (T. Woods), judging from three specimens in my own collection from Long Bay, S. Tasmania (the type locality), the pore- density per sq. mm. is much greater, being 255-264; and the pores are practi- cally circular and measure: outer, 30-35 X 30-35 »; Inner, 25-30 X 25-30 u. They are distinctly visible under a lens. I do not consider atkinsom a true Kraussina as it differs in shell-characters, and in its cardinalia and brachidium. Davidson’s figures (%pl. 21, figs. 5-6) are, unfortunately, not quite accurate. The shells in my possession show the sulcate type of folding and two of them have rather indistinct costae on their outer surface. ‘The species seems to be passing from a costate to a smooth stage. The ventral valve has a submeso- thyrid beak, a fairly large foramen, incomplete, and bordered anteriorly by small triangular deltidial plates; no dental plates; pedicle-collar deep, sessile, vertically striated, and slightly free anteriorly. In the dorsal valve the cardinalia consists of two upstanding socket-ridges on the inner sides of which are slight plates or buttresses which descend to the floor of the valve without meeting in the median line. These descending buttresses extend forward and converge about half-way down the valve, leaving a triangular trough below the apex in which is seen the two scars of the dorsal pedicle muscles. ‘There is little or no cardinal process. Under each of the dental-socket brackets a slight cavity is present. About the centre of the valve arises the rudimentary brachidium in the form of two lamellae directed outwards and ventrally. The oe REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 33 conjunct bases of these lamellae extend backwards to a point which is em- braced by the ends of the converging buttresses of the cardinalia, and the whole process presents the appearance of a bifurcated septum. ‘The upper extremi- ties of the lamellae are slightly forked, but bear no accessory processes on their external faces. With the exception of the latter, the whole structure is essen- tially that of Megerlina lamarckiana, or at least an early stage thereof, but whether atkinsoni should be placed in Megerlina or not is doubtful until more specimens are examined. It should certainly be removed from Kvaussina. I have recently received from Dr. Barnard a very small, smooth, white, sulcate shell, showing rather strong growth-halts. The foramen is incomplete: the deltidial-plates imperfect. The specimen was dredged off Cape Natal (Durban) in 62 fathoms by the Pieter Faure, and was identified by G. B. Sowerby as Araussina atkinson. Its museum registered number is A. 5604. For the present, I am inclined to regard it as a juvenile Megerlina pisum. Note. The interior details of the Terebratula pisum Lamk. for Mauritius are unknown and I have based my conclusions as to the identity of the South African specimens on outward appearance only. If further examples are obtained at Mauritius showing differences in cardinalia and brachidium the South African species should be known as Megerlina natalensis Krauss. Megerlina capensis (Adams and Reeve) (Pl. III, fig. 8). ? Terebratula natalensis Krauss in Kiister, Conch. Cab., von Mart. and Chem., Bd. VII, I, 1844 and 1848, p. 36, pl. 2b, figs. 4-7. (Plate published 1844; text 1848.) Terebratula capensis Adams and Reeve (non Gmelin), Voyage of H.M.S. ‘Sama- (anoeLO50; p. 71, pl. 21, fig. 4 Gn colour). Kraussia deshayesu Davidson, Proc. Zool. Soc., 1852, p. 80, pl. 14, figs. 20-1 (in colour). Terebratula (Kraussia) deshayesu Dav. L. Reeve, Conch Icon., vol. XIII, 1861, pl. 9, figs. 35a, b. ? Terebratula (Kraussia) pisum Lamk. L. Reeve, Conch. Icon., vol. XIII, 1861, pl. 9, fig. 36b (non 36a). Kraussina deshayestt Davidson, ‘Mon. Rec. Brach.’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. pt. II, 1887, p. 122, pl. 20, figs. 31, 31a and 31b (= same figures asin, 1652, but in black: fig. 31a — fig. 20; fig. 31b,= fig. 21): Description: Shell small, subovate, valves almost equal in depth; dorsal valve sulcate, ventral broadly cardinate, type of folding sulcate. Surface of both valves costate, some bifurcated and intercolated costae; concentric growth- lines moderate: test finely punctate. Colour yellowish, suffused with crimson. Beak suberect: foramen large, submesothyrid, incomplete; deltidial plates discrete, small, trigonal. In ventral valve, pedicle-collar sessile, slightly free in front; hinge-teeth without dental plates; interior sparsely tuberculate, with 34. ANNALS OF THE SOUTH AFRICAN MUSEUM a fringe of stronger tubercles just within the margin. In the dorsal valve the cardinalia and brachidium are essentially the same as in M. striata and pisum, but the accessory processes on outer sides of divergent lamellae are represented by slight curved ridges; interior of valve with rows of pustules radiating from the beak, increasing in size progressively, and ending in strong spines just within the margin. Dimensions : Length Breadth Thickness Figured specimen (1) 7-4 6-7 3-1 mm. (Pl. III, fig. 8). Others (2) 8-3 7°6 Bannan (3) 8-5 8-2 38, Habitat: Cape of Good Hope, 120 fathoms. Remarks: I have seen four specimens of this species from the Cape, all of which are in the British Museum (three in the Zoological Department and one in the Geological Department). One of these specimens (No. 1) is now refigured (pl. III, fig. 8). It was originally figured by Reeve (*8pl. IX, fig. 35b) and is part of the Lombe Taylor collection in the above Institution, registered as 74.12.11.386. Reeve also figured another specimen (No. 2 above) from the Cuming collection in the British Museum (*8pl. IX, fig. 35a). The species was first described as Terebratula capensis in 1850 by Adams and Reeve (’pl. X XI, fig. 4) from a specimen dredged by the ‘Samarang’ in 120 fathoms at the Cape of Good Hope. The figure (presumably natural size) shows a shell 13:3 X 14:2 mm. in size. I have not been able to trace the original example. Davidson, in 1852 (pl. XIV, figs. 20-1) figured and described the species as Kraussia deshayestt as there was already a K. capensis (Gmelin) (a synonym of K. rubra Pallas). He gave the locality as “Korea. Coll. Cuming’. The figure is repeated by Davidson in ‘Recent Brachiopoda’ (pl. XX, figs. 31, 31a, 31b, in black), with the habitat corrected as follows: ‘Dredged by Sir Edward Belcher off the Cape of Good Hope, in a depth of 120 fathoms.’ This specimen is in the Davidson Collection, Geological Depart- ment, British Museum (N.H.) registered as B. 12402, and measures, according to the figure (fig. 31) 10°6 X 9°5 mm. The shells of this species have rather large ovate pores, evenly spread over the costae and grooves, and ranging from 224 to 280 per sq. mm. about the middle of the ventral valve. The size of the pores (externally) is 40-50 xX 30-50 py. The South African Museum has one perfect shell and a dorsal valve which I attribute to this species. ‘The specimens were dredged with M. striata off East London in 32 fathoms by the Pieter Faure and are registered as A. 5663. The perfect shell is longitudinally oval and measures: 8-3 X 7-6 mm. It agrees closely with Davidson’s specimen (pl. 20, fig. 31) in the geological department of the British Museum (Nat. Hist.) and that of pl. III, fig. 8 of this memoir. It is suffused with red. ‘The interior of the odd dorsal valve is papillose: the brachium is broken. REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 35 Subfamily MAGELLANINAE Beecher, 1893. Genus Terebratella d’Orbigny. neazg. Cok. Ac. Sc. Paris, XXV, p. 260: Terebratella rubiginosa Dall Terebratella sp. Dall, Amer. Journ. Conch., vol. 6, 1870, p. 122, pl. 6, fig. 4. Terebratella suffusa (non Reeve) Dall, Amer. Journ. Conch., vol. 7, 1871, p. 65. Terebratella rubiginosa Dall, Amer. fourn. Conch., vol. 7, 1871, p. 65. Dall, Proc. Acad. Nat. Sct. Philad., 1873, p. 135. Terebratella(?) rubiginosa Dall. Davidson, ‘Mon. Rec. Brach.’, pt. 2, 1887, puoi pl. 16, fic. 19 (after Dall). Terebratella rubiginosa Dall, ‘Annot. List of Recent Brach.’, Proc. U.S. Nat. Mus., vol. 57, 1920, pp. 372-3. Remarks: According to Dall (p. 372) the type locality for this species is Simons Bay, Cape of Good Hope. ‘The solitary specimen is in the United States National Museum. Dall reports that, ‘The species is entered in the early Smithsonian register with a large number of mollusks collected by Stimpson at the above locality during the Ringgold and Rodgers exploring expedition’, and he thought there was no reason to doubt its having formed part of that collection. It is unfortunate that more is not known of this important member of a typically southern genus. ‘The original specimen has been well described by Dall and Davidson and from their statements it would appear that there is some peculiarity in the septal attachment of the loop. It is to be hoped that more specimens will be obtained in dredgings in Cape waters. The nearest allied species seems to be the Terebratella enzenspergert Blochmann from Kerguelen Is. (“Gauss’ and ‘Challenger’ Expeditions). ‘This species was originally referred to Terebratella dorsata (Gmelin) by Davidson (°pp. 44-5, pl. 4, fig. 4), a South American species. In a joint report on the ‘Siboga’ brachiopods published in 1937°8 reference was made to ? Terebratalia sp. from the region of the Sulu Islands, Moluccas, dredged at 275 metres. Some time afterwards I received a number of other species which had been mislaid and among them I found further and more adult examples of the species in question and coming from the same station and from another near by at 522 metres. They had been referred to Terebratella dorsata by a previous worker, but are not that species, nor Terebratella sanguinea (Leach) a New Zealand form, though coming near to the latter in general appearance, but smaller. They appear to be a new species of Terebratella. This occurrence of a member of the essentially Southern Hemisphere Magellaninae in this neighbourhood is of particular interest from the point of view of distribution. The form is now being closely studied. 36 ANNALS OF THE SOUTH AFRICAN MUSEUM 4. Conclusions In this memoir the following six new species are described from South African waters: Crania roseoradiata, Gryphus capensis, Terebratulina meridionalis, Megathiris capensis, Kraussina crassicostata and Megerlina striata. ‘Vhree of these forms have been misidentified previously, three are quite new to the South African marine fauna-list, Megathiris being the most noteworthy. No doubt further dredgings would add others. Amplified descriptions are given of some of the previously known South African brachiopods. The South African brachiopod fauna is not a large one, but is, nevertheless, of particular interest. ‘There are fifteen species belonging to nine genera, viz. Crania (1 sp.), Lingula (2 spp.), Agulhasia (1 sp.), Terebratulina (2 spp.), Gryphus (1 sp.), Megathiris (1 sp.), Kraussina (3 spp.), Megerlina (3 spp.), and Terebratella (1 sp.). There are also a few rather doubtful records of other forms. The peculiar genus Agulhasia is confined to South African waters, though it is said to have Cretaceous representatives in Europe. Kvaussina, as delimited in this memoir, is almost restricted to the Cape, the only exception being a species—Kraussina gardineri Dall—found in the northern Indian Ocean, in 123-153 fathoms on the Saya de Malha Banks. This primitive genus is probably an old one, but nothing is known of its geological history. The related genus, Megerlina, has three species at the Cape, viz. striata (150 fathoms), pisum (150 fathoms), and capensis (120 fathoms), one of which, pisum, is also recorded from a great depth at Mauritius; one representative, the genotype, M. lamarckiana (17-110 fathoms), in south-east Australia; one, M. atkinsoni, from Tasmania (Long Bay, 10 fathoms); and one species, M. davidsoni, found in the crater at St. Paul Island, south Indian Ocean (tide level to 10 metres), where also occurred Liothyrella wintert (Blochmann). From Mauritius are also recorded Lacazella mauritiana Dall, allied to the northern L. mediterranea (Risso); ?Gryphus cernica (Crosse) from the stomach of a fish dredged at 80 fathoms; and Liothyrina (?Gryphus) sp. Blochmann. At the Island of Réunion a single shell resembling the Mediterranean-East Atlantic Mihlfeldtia truncata was found on a crustacean at 200 fathoms: it was described under the name of Morrisca gigantea by Deshayes in 1863. Dall'é places this tentatively in the genus Pantellaria and as possibly synonymous with the West African (407 fathoms) Pantellaria (olim Miihlfeldiia) echinata (F. & O.), which he also cites for the Cape of Good Hope, 224 fathoms (Jeffreys’s coll. U.S. Nat. Mus.), as well as New South Wales (Angas coll. U.S. Nat. Mus.). ‘This form requires further study. In an earlier part of this memoir I referred to the presence in the British Museum (Nat. Hist.) of an immature example of Miihlfeldtia truncata labelled ‘South Africa. J. H. Ponsonby coll. 1900’. From the Saya de Malha Banks (123-153 fathoms) are recorded Gryphus sp. indet. Dall, which has affinities with the northern forms G. vitreus and G. sphenoideus; Kraussina gardinerr Dall (as stated above) and ?Aetheza (olim Hemt- REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 37 thyris) sladeni (Dall). The genotype of Aetheza is Terebratula gualtert Morris from the Oamaruian—Eocene to Miocene—of New Zealand. The recent Hemi- thyris colurnus Hedley from the east coast of Australia (100-250 fathoms) may belong to this genus. It may be of interest to refer briefly to the known brachiopod fauna from certain places to the south-east of South Africa, especially the Marion Island district including Prince Edward Island and Crozet Island; Kerguelen Island and Heard Island; and St. Paul Island. The Marion Island group is situated on an elongated submarine ridge of less than 1,000 fathoms in depth and is separated from South Africa by sea less than 2,000 fathoms deep. From this neighbourhood are recorded the following species: ‘Waldheima’ (?Magellania) kerguelenensis Dav., with two specimens of Platidia anomioides (Sc. & Ph.) attached, off Marion Is. 100 fathoms; Terebratulina septentrionalis Dav. (?Couth.) off Marion Is. 150 fathoms; Platidia anomioides, five examples, off Prince Edward Is., depth ?; and Liothyrella moseleyi (Dav.), off Crozet Is. 210 fathoms. Kerguelen and Heard Islands are situated on a similar submarine elevation of less than 1,000 fathoms deep and separated from the Marion Island group by sea of over 2,000 fathoms deep. From here are recorded Tereoratella enzenspergeri Bloch. (= T. dorsata Dav. non Gmelin), Royal Sound, Kerguelen, 20-30 fathoms and Observatory Bay, Kerguelen; ‘Waldheimia’ (? Magellania) kerguelenensis, Balfour Bay, Kerguelen, 20-60 fathoms, and Observatory Bay, Kerguelen, 5-6 fathoms; and the same species with three examples of Tegu- lorhynchia pyxidata (Dav.) and ?Gryphus sp. (= Terebratula uva Dav. non Brod.), two fragments of a dead shell, off Heard Island 150 fathoms. St. Paul Island lies on a small elevation of the sea-floor less than 1,000 fathoms deep and separated from the Kerguelen group by sea over 1,000 fathoms deep. The species recorded from here are Liothyrella wintert (Bloch.), 371 fathoms, and Megerlina davidson (V élain), low water in crater. The new observations in this memoir provide further material for the study of former land connections in southern regions. Little advance, however, can be made in this study until further researches are carried out in the case of the Tertiary brachiopods of many areas. The absence of Tertiary marine rocks in South Africa adds to the difficulty. There is still much to be done in connexion with the clearer definition of certain genera and species recorded from the island groups named above. The ‘Waldheima’ kerguelenensis Dav., for example, possesses cardinalia of a peculiar type, judging from Davidson’s figures, assuming these to be correct Reelolil, fie. 85° pl. +X, fig.,16).. Pichler (**pl. XLIII;, fig..16) shows a different type of cardinalia in a somewhat younger Kerguelen example of Magellania presumed by him to be Davidson’s species. The ?Gryphus sp. is probably a new species and not related to the Liothyrella uva (Brod.) of South America and the Falklands. It may indeed have close affinity with the Gryphus capensis Jackson of the Cape. 38 ANNALS OF THE SOUTH AFRICAN MUSEUM The Terebratulina septentrionalis of Davidson from Marion Island may be closely allied to the 7. merzdionalis Jackson of the present memoir. The Platidia anomioides from Marion Island and Prince Edward Island is another problem. It is apparently a wide-spread northern form (fossil and recent) of the ancient Tethys and has spread west through the ancient Strait of Panama to California. ‘The occurrence off Marion Island, if authentic, would suggest a derivation from the Eastern Tethys, via the East African coast and so to Marion Island, when the latter was connected by shallower seas with South Africa, probably in Tertiary times. In connexion with the above there is the interesting occurrence of the Tethyian genus Megathiris at the Cape, and Lacazella at Mauritius and Gryphus sp. at Saya de Malha Banks. With regard to the genera Araussina and Megerlina in South African waters, these may have spread from the ancient India-Africa isthmus across the Indian Ocean and have been distributed by the Mozambique and Agulhas currents from the east coast round the Cape. There is also to be considered the occurrence of Megerlina at St. Paul ielend South Indian Ocean and off south-east Australia. There are many other problems, but until the various points raised above are clarified, it would be hazardous to dogmatize on the origin of the South African brachiopod fauna. So far as present evidence is concerned, it might be pointed out that the peculiar brachiopod fauna of South Africa seems to show little affinity with that of Antarctica, though the presence of Crania and Lingula seem to link it with the Australian region. More information is required regarding the duration of the free-swimming larval stage of these marine sedentary animals before we are in a position to answer questions as to former connections between the different oceans and different land-masses. If the free-swimming stage is long enough, it is possible for species to be transported fair distances by the aid of ocean currents. The inarticulate brachiopods are interesting from this point of view. Brachiopoda are essentially dwellers on the continental shelves and the majority inhabit waters less than 166 fathoms in depth: they are also influenced by the nature of the sea-bottom; Lingulids, especially, being mud and sand dwellers and restricted to the littoral region in warm waters. Lingula is rare below 7 fathoms, and is usually found living in burrows in low water in estuaries and bays, indicating a preference for habitats more or less freshened by river waters. The pelagic larvae of Lingula, possessing a mouth and functional stomach and therefore self-sustaining, have been met with occasionally. ‘They have thus considerable capacity for transportation over deep oceans in surface currents. Yatsu, in his memoir of 1902,°° says larvae of Lingula ‘anatina of Japan may live in the free-swimming stage in aquaria for not longer than two months. J. H. Ashworth, in 1915,°’ gave an interesting account of the larvae of REVISION OF SOME SOUTH AFRICAN BRACHIOPODA 39 Lingula (presumed to be anatina) obtained by him in the southern part of the Red Sea and in the Indian Ocean about 4° south of Colombo. He also refers to earlier records of Lingula larvae near Zamboanga, Mindanao, Philippines; off Japan; and off South Burma. In addition, he cites the occurrence of larvae, presumed to be of Lingula, off the west coast of Africa. There is a species recorded from here, the Lingula parva E. A. Smith 1871, taken at Whydah, Dahomey. The researches of Yatsu and Ashworth seem to suggest that the free-swimming stage of Lingula is about five or six weeks. 5. Bibliography LIST OF PRINCIPAL REFERENCES 1. P.S. Pallas, 1766, Miscel. Zool., p. 182, pl. 14, figs. 2-11. J. H. Chemnitz, 1785, Neues systematisches Conchylien-Cabinet, vol. VIII, p. 94, pl. 77, figs. 703a, b, c, and p. 78, pl. 76, figs. 688a, b. 3. G. B. Sowerby, 1847, Thesaurus Conchyliorum, vol. 1, p. 346, pl. 68, figs. 12-14, and p. 362, pl. 71, figs. g1-2. 4. J.B. Lamarck, 1819, Animaux sans Vert., vol. VI, p. 245. (Text written by Valenciennes, owing to Lamarck’s blindness.) 5. I. Davidson, 1880, “Report on the Brachiopoda dredged by H.M.S. “‘Challenger’’ during the years 1873-1876’. Voyage of ‘Challenger’, Zool. vol. I. 6. F. Krauss, 1848, Die Stidafrikanischen Mollusken: Brachiopoda, pp. 32-3, pl. II, figs. 10-11. 7. A. Adams and L. Reeve, 1850, ‘Mollusca, in Zool. of the voyage H.M.S. ‘““Samarang”’ ’ Brachiopoda, pp. 71-2, pl. XXI, figs. 1-5. 8a. T. Davidson, 1852, Proc. Zool. Soc., p. 80, pl. XIV, figs. 20-1. iN) 8b. id. 1852, Ann. Mag. Nat. Hist., ser. 2, vol. 9, p. 369. g. W. H. Dall, 1870, Amer. Journ. of Conchology, vol. VI, pp. 122, pl. VI, fig. 4. 10. id. 1871, id. vol. VII, p. 65. 11. W. King, 1871, Ann. Mag. Nat. Hist., ser. 4, vol. 7, pp. 109-12, pl. XI, figs. 1-7. 12. P. Fischer and D. P. Oehlert, 1892, Brachiop. Mission Scient. du Cap Horn (1882-1883), PP. 330-1. 13. E. A. Smith, 1901, Journ. of Chonchology, vol. X, p. 116. 14a. F. Blochmann, 1906, Zool. Anzeig., vol. XXX, pp. 690-702 and 824, figs. 1-3. 14b. id. 1908, ert. fiir wissen. Kool., Bd. XC, pp. 596-644, pls. XXXVI-XL. 15. J. Allan Thomson, 1918, “Brachiopoda: Australian Antarctic Exped., 1911-1914’, Sci. Rep., ser. C, vol. 4, pt. 3, 75 pp. 4 pls. and map. 16. W.H. Dall, 1920, Proc. U.S. Nat. Mus., vol. 57, pp. 261-377. 17. J. W. Jackson, 1921, Ann. Mag. Nat. Hist., ser. 9, vol. 7, pp. 40-9. 18. L. Reeve, 1862, ‘Monogr. of Crania’, Conch. Icon., pl. I, figs. 2-3. 19. T. Davidson, 1886-1888, “Monogr. of Recent Brachiopoda’, Trans. Linn. Soc., ser. 2, vol. IV, Zool. Part I, pp. 1-73, pls. 1-13, 1886; Part II, pp. 75-182, pls. 14-25, 1887; Part III, pp. 183-248, pls. 26-30, 1888. 20. F. W. Hoeninghaus, 1828, Beitr. zur Monogr. d. Gattung Crania, p. 3, fig. 3. 21. C. Hedley, 1906, Proc. Linn. Soc., New South Wales, vol. XXXI, p. 467. 22. L. Reeve, 1860, ‘Monogr. of Terebratula’, Conch. Icon., pl. IV, fig. 14. 23. J. W. Jackson, 1918, Geol. Mag., Dec. 6, vol. 5, pp. 73-9. 2 Be a H. Crosse, 1873, Journ. de Conchy., vol. XXI, p. 285, and vol. XXII, p. 75, pl. L §- 3- 25. W. eB oe 1910, Trans. Linn. Soc., Lond., 2nd ser. Zool. vol. XIII, pt. 3, pp. 439-41, pl. 26. 26. W.B. Alexander, 1914, Records of West Aust. Museum, vol. I, pp. 239-40. 27. J. A. Thomson, 1926, Ann. Mag. Nat. Hist., ser. 9, vol. 18, p. 525. ANNALS OF THE SOUTH AFRICAN MUSEUM id. 1927, ‘Brachiopod Morphology and Genera (Recent and Tertiary)’. N.Z. Board of Science and Art, Manual No. 7, vi and 338 pp., 2 pls., 103 text-figs. C. E. Beecher, 1893, Trans. Conn. Acad. Sci., vol. IX, pp. 376-91. J. W. Jackson, 1916, Geol. Mag., Dec. 6, vol. 3, pp. 21-6. J. A. Thomson, 1916, Geol. Mag., Dec. 6, vol. 3, pp. 496-505. E. E. Deslongchamps, 1884, Bull. Soc. Linn. Normandie, ser. 3, vol. 8, pp. 210, 243, pl. 7, ITPER The id. 1884, Etudes Crit., 1, pp. 121, etc., pl. 19, fig. 7, p. 159, pl. 19, fig. 11. E. Suess, 1859, Sztzungsb. k. Akad. Wissensch. Wien., Bd. XX XVII. T. Davidson, 1861, Ann. Mag. Nat. Hist., ser. 3, vol. 8. H. C. Kiister, 1844 and 1848, Conch. Cab., by Mart. and Chem., Bd. VII, I, p. 36, pl. 2b, figs. 4-7 (plates 1844; text 1848). G. B. Sowerby, 1846, Thes. Conch., vol. I, pt. VII, pp. 345-6, pl. 60, figs. 37-9. id. 1847, Thes. Conch., vol. I, pt. VII, p. 362, pl. 71, figs. 91-2. L. Reeve, 1861, ‘Monogr. of Terebratula’, Conch. Icon., pl. UX, fig. 35b. D. P. Oehlert, 1887, in Fischer’s Manuel de Conchyliologie. C. Schuchert, 1893, The American Geologist, vol. XI, No. 3. C. Schuchert and Clara M. Le Vene, 1929, ‘Brachiopoda’, Fossilium Catalogus, by J. F. Pompeckj, p. 24. E. E. Deslongchamps, 1864, Recherches sur l’ Organisation du Manteau chez les Brachiopodes Articulés, etc., Paris and Caen. Victor Sganzin, 1843, Mém. Soc. du Mus. d’hist. Nat. de Strasb., vol. 3, pt. 2. A. Philippi, 1887, Die tertiaren und quartaren Versteinerungen Chiles, 4to, Leipzig. D. P. Oehlert, 1888, L’ Annuaire Géologique Universel, Tome IV. H.,von Ihering, 1903, Anales del Museo Nacional de Buenos Aires, ser. III, vol. I. B. Faujas de Saint-Fond, 1799, Hist. de la mont. St.-Pierre. A. d’Orbigny, 1847, Pal. frang. Ter. cret., 1V. J. de Morgan, 1883, Bull. Soc. Zool. de France, tom. VIII. P. Fischer and D. P. Oehlert, 1891, Exped. Scient. du ‘Travailleur’ et du ‘Talisman’, 1880- 1883, Paris. J. G. Jeffreys, 1858, Ann. Mag. Nat. Hist., ser. 3, vol. 2. F. Sacco, 1902, I Brachiopodi dei terreni tertiarit del Piemonte e della Liguria, Torino. T. Davidson, 1870, Geol. Mag., vol. vii, pl. X XI, figs. 6-8. W. H. Turton, 1932, The Marine Shells of Port Alfred, S. Africa, ‘Brachiopoda’: pp. 259-60, pl. LXX, figs. 1836, 1838, 1842 and 1843. z P. Eichler, 1911, Die Brachiopoden der Deutschen Stidpolar-Expedition, 1901-1903. N. Yatsu, 1902, ‘On the Development of Lingula anatina’, Journ. Coll. Sci. Imp. Univ. Tokyo, vol. XVII, art. 4, 112 pp., 8 pls. J. H. Ashworth, 1915, “On Larvae of Lingula and Pelagodiscus (Discinisca)’, Trans. Roy. Soc. Edinb., vol. LI, pt. I, pp. 45-69, pls. [TV and V. J. Wilfred Jackson and G. Stiasny, 1937, The Brachiopoda of the Siboga Expedition, Mon. XXVII, p. 20. > Lhe ANNALS OF THE SOUTH AFRICAN M USEUM are issued i in parts at : eee : 1 - intervals as material becomes available. As far as possible each volume is devoted exclusively to a particular subject (Zoology, ey yee ‘Two or more volumes may be in course of 4 publication concurrently. Ea Most of the Geological and Palaeontological papers are issued i in mals with fic: Js 4 Geological Survey of the Union of South Africa. Se - Some volumes and parts are out of print, and others are only sold as parts of a set, or volume, a respectively. ‘The prices of parts published prior to 1940 have been increased. Out of print: Vols. I, I, V (Part 9), VII, VIII, IX (Part I), XII (Part 3 XXII, XXIV a (Part 2), XXXI (Parts 1, » 2, 3) Vol. 4 Li joe III. 1903-1905 Zoology .. .. aa oclaae pa ea IV. 1903-1908 Palaeontology .. peak o- V. 1906-1910 Geology, Palaeontology, Zoology, Anthropelogy ae VI. 1g08-1910 Zoology .. UE ONey ees : IX. 1911-1918 Botany .. a ce ary _ (excl. Part 1) X. $«Igi1-1914 Zoology ; eg Rigi tes XI. 1911-1918 Zoology . te cd ee ae XII. 1913-1924 Palaeontology ind Cee oe .. (excl. Part 7) XIII... 1913-1923. Archaeology and Zoology ..—_—.. 2 eee XIV. 1915-1924 Zoology .. a Be oh a 7 ae eee XV. 1914-1916 Zoology .. se cs cece So apes a MOVIL 9919-1999 . Botany. 3. pe oS 5 a eee XVII. 1917-1920 Zoology: .. ... «. ats nec ee XVIII 1921 Zoology oa oe XIX. 1924-1925 Zoology .. a ae as ne ae ‘ XX. 1924-1926 Zoology .. ae ig aes = Ne ae: XXI. 1925-1927 Zoology .. He Se ere ee ae ee XXII. 1925-1926 Zoology ae oe 5 aos XXIV. 1929-1938 ‘Anthropology and Ethnology ses ~ » (excl. Part 2); XXV. 1927-1928 Zoology .. ae wets: Gla 292 Giguere XXVI.- 1928 Zoology .. ve wes pie ee a eae se ae XXVII. 1929 Anthropology .. fe We eho XXVIII. 1929-1932 Palaeontology .. Se SE es XXIX. 1929-1931 Zoology .. =t XXX. 1931-1935 Zoology .. aie INDEX of papers, authors, and subjects, published j in Vols L-Xxx. See XXXII. 1934-1950 Palaeontology a foes a (Pare 4 ‘cams XXXII. 1935-1940 Zoology .. na en ee : : - ees XXXII. . 1939 Zoology .. .. he en me Pent tii XXXIV. 1938 Zoology .. a ne XXXV. Reserved for conclusion of i broeeaph i in Vol. XXXIV. XXXVI. 1942-1948 Zoology ia. oe ae BSN ec ae XXXVII. 1947- ~— Archaeology Sit. i? a ee XXXVIII. 1950 Zoology .. Spe ae a Cobre MAMIX:.f952 4; Zoology’ 2:00) ue. 4 at wie ine Gee ie XL. 1952- Botany... se a4 ce ae (Part 1) — | Copies may be obtained from— The LIBRARIAN, Soutru ArricAn Museum, Cape Town, ny Pw 0 OW ND KD ww wD ww OO 1) U9 eB OD G9 oA GO OO OO OO oe 09 mt OO oR ip ONnNW KF NK cere the Geological and Palaeontological parts, which are obtainable from the ge GOVERNMENT PRINTER, PRETORIA. COMOMAMTOFCOOMIOVFGOMRO.OMAMOGAIDOADMOS onN000. Mi ea sch - 7. rg f | sees . . frees ANNALS OF THE SOUTH ‘AFRICAN MUSEUM VOLUME XLI PART EH, containing : — . On South African Coccinellidae species incertae. By H. ANDREAE, D.Sc. The genus Anobium Thunb. By H. ANDREAE, D.Sc. — . Notes on the Ethiopian Pentatomidae. IV. A new genus and species from the Cape Province. By D. Leston, F.Z.S., F.R.E.S. (With one text-figure.) . Notes on the Ethiopian Pentatomoidea. X. Some specimens from southern Africa in the South African Museum, with a note on the remarkable pygophore of Elvisura irrorata Spin. and description of a new species of Piezodorus Fieber. By D. Leston, F.Z.S., F.R.E.S. (With 17 text-figures.) . Studien tiber siidafrikanischen Meloiden. By Dr. Z. Kaszas. DEC 1 4 1953 ) Sa LIBRARN ISSUED OCTOBER 1953 PRICE 7s. 6d. PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY THE RUSTICA PRESS LIMITED, COURT ROAD, WYNBERG, CAPE a 3 i a : EE timate ot, Rah” Daa oe ye wis SAGES TS py ast pan eS re 2. On South African Coccinellidae, species incertae. By H. ANDREAE, D.Sc., Hon. Curator of Coleoptera, South African Museum. Coccinella gibba ‘Thunb., Nov. Spec. Ins. 1, 1781, p. 13, f. 14, from the Cape, is cited in W. Junk’s Coleopterorum Catalogus, pars. 118, p. 67, as Epilachna incertae sedis. Crotch has already remarked (Revis. Cocc. 1874, p. 9): ‘Little or no reliance can be placed on the localities’ in this publication. Yet there is a South African species which agrees with Thunberg’s description in so far that all the characters given are within its limits of variation, although exactly the same combination is not found in the specimens in the South African Museum collection. ‘The main points are: Light brownish red, metasternum black, prothorax with a median black spot, elytra with six black spots each, two basal, three median, forming an arcuate band, one towards the apex near the suture; 5-8 mm. The figure shows no spot on the prothorax and a straight band narrowed in two places, widened in three. Out of sixteen specimens which I refer to gibba, six have a spot on the pro- thorax, one has the median spots (3, 4, 5) connected, one has the normal spots 6 and 8 missing (8 alone is missing in 8), the metasternum alone is black or fuscous in 7, in the others either some abdominal segments are infuscated at the base or the whole underside is red. 5°5-7-5 mm. Cape: Douglas; Transvaal: Potchefstroom, Pretoria, Pietersburg, Lydenburg, Zoutpansberg. In the Transvaal specimens, spots 1, 2 or I, 2, 5, sometimes also 6, 7 or 6, 7, 8, are connected, the elytra showing a design like EL. hirta ab. caesarea Sic., but head and prothorax are red. J. Weise describes (Deutsche Ent. Keitschr. 1888, p. 83) a form from the Kwango, northern Angola, in which spots 1 and 2 are also connected, forming a dentate band from one humeral callus to the other; he stated that EL. hirta Thunb. var. invalida Muls. gradually changes into var. insidiosa Muls., therefore he gave no names to the intermediate forms. FE. gibba ‘Thunb. apparently differs from var. invalida Muls. only by elytral spots 3, 4, 5 connected, 6 and 8 missing, central spot on the prothorax and metasternum black. As it is a very variable species, the definition should be amended. Light brownish red, prothorax with or without median black spot, elytra red witha black spots (2, 3,2, 1), 1, 2 or 1, 2, 5 or 3, 4, 5, sometimes also’6, 7 or 6, 7, 8, connected, 8 or 6 and 8 may be missing. Underside red, metasternum mostly black or fuscous, abdominal segments often infuscate at the base. 41 VOL. XLI. PART 2. DEC 9 1993 42 ANNALS OF THE SOUTH AFRICAN MUSEUM E. gibba Thunb. therefore comprises the specimens with red head and pro- thorax so far placed in ab. discors Muls. and caesarea Sic. (both have head and prothorax black), and znvalida Muls. with connected elytral spots and dark metasternum. As E. gibba (p. 13) has preference, E. hirta (p. 23) becomes a subspecies. Key to the South African forms 1” Head and prothorax red. subsp. gibba 2” Elytra red with black markings (2, 3, 2, 1), (2, 3, 2) or (2, 3, I). 3” Some elytral spots connected, metasternum mostly black or fuscous, abdominal segments often infuscate at base. gibba f. typ. 3’ All the spots free, underside red, only sides of metasternum and spots on abdominal segments sometimes infuscate. ab. invalida 2’ Elytra black with free or connected yellowish markings (1, 2, 2, 1), basal and apical lunulate, others rounded or subquadrate. 4” Markings 1, 2 or I, 2, 3 or I, 2, 3, 1 connected. ab. peringueyt 4’ All the markings free. ab. insidiosa 1’ Head and prothorax black. subsp. hirta 5’ Elytra like 2’. 6” Markings 1, 2, 3 free or connected, 4, 5, 6 free. hirta f. typ. (Ab. 12-verrucata F. may belong here; description not available.) 6’ Markings 4, 5 forming a transverse band, the others free. ab. guttatofasciata 5’ Elytra like 2”. 7” Spots 1, 2, 5 connected, the others free, 8 (and 6) sometimes missing. _ab. caesarea 7’ All the spots free, or others than 2 and 5 connected. ab. discors Of the other species described from the Cape in the same paper and not identified so far, C. trinotata, p. 11, fig. 11, may be identical with or closely allied to Martinella justitiae Gorh. from Natal; the descriptions agree fairly well, except that in the type of justztvae the suture is partly black and the three spots are united into one patch. Of the two specimens in the South African Museum, the male agrees completely with the description while the female has no black suture, only a triangular patch with rounded angles and concave sides indicating that it is formed from three spots. So far no specimen with free spots has been recorded. C. oculata, p. 14, fig. 18, and C. lunata, p. 19, fig. 28, are probably not from South Africa; Junata might be a Cyrtocaria (Madagascar). In Germ. Mag. Ent., IV, 1821, p. 183, Wiedemann described C. erythrochila from the Cape, collected by the Rev. Hesse, Gape Town, which has not been recorded again. The author compares it with two species of Exochomus, auritus (flavipes) and haemorrhoidalis, but points out that the scutellum is much larger and the fine punctation of the elytra almost striate. The coloration is very peculiar: aeneous, sides of prothorax and elytra with a broad, reddish-yellow margin. To this species I refer a specimen in the collection of the South African Museum from Jakkalswater, Bushmanland, coll. R. M. Lightfoot, Oct. 1911, ON SOUTH AFRICAN COCCINELLIDAE 43 although it is smaller (3 mm. instead of 2 lines) and the last abdominal segment as well as the legs (except the posterior side of the femora) is testaceous. These differences I take to be sexual, the type being a female and the present specimen a male. It is a Hyperaspis allied to H. pumila Muls. but without yellow spot on the disc of the elytra, the testaceous margin twice as broad at the shoulder, widened towards the apex and terminated by the apical patch which is just indicated by a slight widening of the rounded end. 3. The genus Anobium Thunb. By H. ANDREAE, D.Sc., Hon. Curator of Coleoptera, South African Museum. In Novae Insectorum Species, 1, 1781, p. 8, Thunberg established his genus Anobium with six species, all recorded from South Africa, three from other countries also; this caused some trouble and was not accepted by other entomologists. Fabricius in Mantissa Insectorum, 1787, p. 35, mentions only two as synonyms, the others he could obviously not identify, and in the Junk catalogue, part 23, 1910, three are mentioned, the others remain species dubiae. Thunberg’s description: “Antennae clava perfoliata, triarticulata. Thorax marginatus, rotundatus. Corpus convexum, lineari-oblongum. Elytra flexilia.’ looks rather vague, but it actually applies only to some Cleridae, subfam. Corynetinae,* and even here the genera Yenerus Cast., Tarsostenus Spin., Cory- netinus Reitt. and Opetiopalpus Spin. are excluded; only seven genera with together twenty-three species described from South Africa and some unnamed species in the South African Museum’s collection were left. Under these circumstances it was not difficult to identify all Thunberg’s species. 1. A. ruficolle Vhunb., l.c., p. 8, is now Necrobia ruficollis F., as stated by Fabricius and Junk, Catalogus; a cosmopolitan. 2. A. capense Thunb., l.c., p. 9, is now Prosymnus capensis (Thunb.), so far unknown. Three unnamed specimens in the museum’s collection agree with the description, except the base of the antennae, joints 1-6, which are rufo-testaceous; this is unimportant, the rufous or testaceous base of antennae has also been overlooked in No. 4 and No. 6. The fasciae of the elytra are formed by white appressed hairs, elsewhere the upper side bears unusually long and strong black bristles. The terminal joints of maxillary and labial palpi are elongate triangular, and the femora are deeply grooved on the under side. The species shows all the important characters of Prosymnus Cast., the different vestiture of the elytra alone would not justify the establishing of a new genus. Length 3-5 to 5 mm. Apparently rare. Cape: Cape Town, coll. J. C. Bridwell; Kalk Bay, coll. R. M. Lightfoot; Bredasdorp, coll. H. Fry. * In Nova Acta Soc. Sc. Upsala, VII, 1821, p. 174, Thunberg described Dasytes opacus and D. rufipes, and transferred his Anobium coeruleum and A. viride to the genus Dasyies, citing the transfer of coeruleum to Dasytes by Fabricius (Syst. Eleuth., II, 1801, p. 75) and accepting it, although both species are really Cleridae and have nothing to do with Dasyies. Dasytes rufipes Thunb., so far not identified, is according to the description certainly different from Anobium rufipes Thunb. and probably a Dasytes. 44 THE GENUS ANOBIUM THUNB. 45 3. A. bifasciatum Thunb., l.c., p. 9. Now Thricera bifasciata Thunb., syn. T. 4. A. bifasciata Hintz. The two descriptions agree very well, but Thunberg’s figure (rather poor and not agreeing with the description) has probably so far prevented the identification. Cape: widely distributed but rare. Described by Hintz (Deutsche Ent. Zeitschr., 1902, p. 403) as T. bifasciata from Dunbrody, two spec., probably coll. Father O’Neil; by Gorham (Proc. Zool. Soc. Lond., 1905, 2, p. 274) as T. bicinctella from Port Elizabeth, two spec., coll. Dr. H. Brauns. In South African Museum coll., one spec. from Dunbrody, apparently coll. by Father O’Neil, no date, and two from Knysna, Oct. 1916, coll. L. Peringuey. I found one at Parow near Cape Town, 28/8/47, and one on the farm Tierhoek, Piquetberg Mts., 19/10/47. The var. tricolor Hintz is recorded from Dunbrody (Hintz), 2 spec.), Port Elizabeth (Gorham, 2 spec.) and East London (South African Museum coll., Oct. 1912, R. M. Lightfoot, one spec.). viride Thunb., l.c., p. 9, now type of the genus Notostenus Spin. (Clerites IT, 1844, p. 89). A well-known species, frequent on flowers of the arum lily. Thunberg’s statement: ‘Habitat in Africa & India Oriental’ is a mistake, the species is South African. One hundred and ten specimens in the South African Museum coll., mostly from the western Cape, but also from Port St. Johns and Durban. 5. A. rufipes Thunb., l.c., p. 10, is Necrobia rufipes De Geer. This was stated by Fabricius (Mant. Ins., 1787, p. 35), but not taken over by Schenkling in Junk, Catalogus, part 23; a cosmopolitan. 6. A. coeruleum Thunb., l.c., p. 10, is a Dolichopsis, not a Notostenus, as stated in Junk, Catalogus, part 23, p. 136. A very variable species, colour from golden green to violaceous blue, elytra different from head and prothorax or upper side unicolorous, antennae with articles 2 to 4 (sometimes 5 also) testaceous, rufous or dark fuscous above, rufous below, prothorax fairly convex or rather flat, punctation uniform and rather close or sparse on the disc, especially in the anterior half, length 2-5 to 5mm. The extreme forms may easily be mistaken for different species, but the intermediate forms make any division impossible. Frequent from Cape Town to the Cedarbergen and Zwartbergen, on flowers of various plants, especially Ficoideae, Selaginaceae and Compositae. Dolichopsis cyanella Gorh. (Trans. Ent. Soc. Lond., 1878, p. 155) is the same species and therefore a synonym. 4. Notes on the Ethiopian Pentatomidae. IV. A new genus and species from the Cape Province. By D. Leston, F.Z.S., F.R.E.S. (With one text-figure.) Carvalhocoris gen. nov. MobDERATE size, oval; base of venter tuberculate; mesosternum with a low but distinct carina of even height throughout; odoriferous aperture large and oblique, continued by a long, smooth and gently curving ‘horn’ towards and almost reaching the anterior border of the mesosternum at its furthermost point; evaporating area distinct, its lateral border raised; rostrum reaching to posterior trochanters, second segment longest, third widest and subequal to the fourth. 2 Carvalhocoris nigra gen. et sp. nov. § 46 NOTES ON THE ETHIOPIAN PENTATOMIDAE 47 Head breadth to length ratio as 3: 2; greatly deflected, sides sinuate, apex widely rounded, margins not reflexed; interlobar sutures straight; upper surface slightly convex. Pronotum with anterior margin broadly emarginate, anterior border between and behind the eyes smooth and impunctate; lateral margins reflexed, more so anteriorly, rounded gradually but slightly, entire; postero-lateral angles rounded, obtuse, rather raised; posterior border gently emarginate, postero- internal angles about 150°. Scutellum long, broadly rounded at apex, sides narrowing just before the middle; disc raised slightly, posterior portion deflected gradually downwards. Apex of corium reaching caudally as far as the apex of the scutellum; membrane dark, almost obscuring the few parallel veins; reaching to apex of abdomen. The whole upper surface densely punctate, rugosely so on the scutellum, connexivum punctate. Tibiae flat and canaliculate externally. Antennae short, ratios of segments 6: 10: 11: 14: 13; fourth and fifth segments slightly thicker than second and third; ‘first not surpassing apex of head. Head, thorax and abdomen beneath punctate but less so than the upper surface. Type: Carvalhocoris nigra sp. nov. This genus should be placed in Pentatominae, tribe Pentatomini; it pertains to most of the genera placed in Menidaria Distant. From Amphimachus Stal it can be distinguished by the shorter scutellum, from Aegaleus Stal by the absence of impunctate fasciae on the scutellum and from Menida Motsch. by the smaller ventral spine. This last genus has been used as a dumping ground for many species and is a composite one as at present used; it is possible that certain species included in it should be placed within Carvathocoris. Carvalhocoris nigra sp. nov. Shining black except for the antennae, these flavous, darker apically. Head above with a fine narrow flavous margin anteriorly; anterior margin of pro- notum narrowly flavous but irregularly so in the centre; lateral margins of the pronotum narrowly flavous; corio-pronotal junction externally flavous. Connexivum flavous with contiguous black spots either side of the sutures. Under side flavous with dark brown punctures; tibiae and femora flavous spotted with brown, heavily so on the latter. A linear series of three large black spots, one on each thoracic pleurite centrally. Ventral spiracles brown. Length, 26mm.; maximum pronotal width, 9 4 mm. Type: @ Prieska, Cape of Good Hope, October 1887 (now in South African Museum, Cape Town). Paratypes: 4 Aliwal North, Cape Province, December 1922. @Q same, 4,350 ft., January 1923. (Both collected by Turner and now in British Museum I am happy to be able to name this genus in honour of Dr. J. C. M. Carvalho, the eminent Brazilian hemipterist and explorer. My thanks are tendered to the Director, and Dr. Hesse, of the South African Museum, for submitting a small collection of their specimens to me for determination. 5. Notes on the Ethiopian Pentatomoidea. X. Some specimens from southern Africa in the South African Museum, with a note on the remarkable pygophore of Elvisura irrorata Spinola and description of a new species of Piezodorus Fieber. By D. Lesron, F.Z.S., F.R.E.S. (With 17 text-figures.) By courtesy of the Director and Dr. A. J. Hesse, I have examined a collection of previously undetermined Shieldbugs belonging to the South African Museum, Cape Town. Most of the specimens were taken at localities within, or adjacent to, the Union. The literature contains not a few references to the fauna of South Africa; the pioneer workers were Palisot de Beauvois, Thunberg and Germar. The first important work is that of Stal (1865); in the first volume of Hemiptera Africana he collated all the earlier records and placed them in recognizable modern genera. Later, Stal (1876) brought his classification up to date and the broad outlines of generic and higher taxonomy in use to-day are entirely due to this great Swedish entomologist. It is only in later works that localities are given with sufficient precision to form a basis for zoogeography, and these works include Wallengren (1875) ; Distant (1892, 1898); Schouteden (1912); Schumacher (1913); Hesse (1925, 1935) and Leston (1952a, 19525). Besides these, new species from the area are described in papers by Bergroth, Distant, Montandon, Jensen-Haarup, Schouteden and Leston, as well as in the catalogues of Westwood, Dallas and Walker. Solely to aid further knowledge of the distribution of species, the following list is given. | BRACHYPLATIDAE Lest., 1952a (Plataspidae) Gelastaspis brown Kirk. 1902. Kirkaldy. Entomologist. xxxv, p. 166 Umtali, Southern Rhodesia (A. Bodong) CypnipAE Billberg, 1820 Legnotus tibialis Stal 1853. Stal, Ofv. Vet. Ak. Forh, x, p. 222 M’fongosi, Zululand, iv—v, 1934, xii-li, 1934-5, v, 1935 (W. E. Jones) Legnotus melaleucus (Thunb.) 1783. Thunberg, Nov. Ins. Sp., ii, p. 50 Kamieskroon, Namaqualand, ix, 1930 (South African Museum staff) 48 NOTES ON THE ETHIOPIAN PENTATOMOIDEA 49 PENTATOMIDAE Leach, 1815 SCUTELLERINAE Leach, 1815 SCUTELLERINI Leach, 1815 The nomenclature and classification of the Scutellerinae have recently been investigated (Leston, 1953a, 19530); as a result, the groups Scutellerin: Leach, Sphaerocorini Stal and Elvisurini Stal are now treated as subtribes within Scutel- lerini. All three subtribes occur in the Union of South Africa. Elvisura trrorata Spinola 1837. Spinola, Essai Hem. p. 359 Karkloof (now: Kloof), Natal, x11, 1915 (Bell Marley) This was a female, but through the courtesy of the British Museum (Nat. Hist.) authorities I have been able to examine the genitalia of a male of this rarity. The genus is distinct from all other Scuiellerinae in that the scutellum is distinctly keeled and the bugs have a marked resemblance to Buprestid beetles. The male genitalia are shown in figures 1-4. The pygophore is large with a well-developed caudal lip: the cephalad margin is remarkable for the presence of a pair of large, flat, rectangular and freely movable flaps attached immediately within the margin; they appear to be adherent throughout their length but with a root-like attachment at their outer corner. The flaps are very thin but sclerotized, with a pale band along the proximal border; the distal border is fringed by a few fine hairs (not shown in the figure). In work now in progress on the Angolan Pentatomid fauna a pair of movable pieces has been found on the internal-lateral margins of the pygophore in Crollius Dist.; they were triangular and large so that, when brought together in the resting position, they completely covered the pygophoral opening. Dr. R. I. Sailer of the U.S. National Museum has informed me, in Uitt., that he has discovered similar processes on the pygophore of other Podofini, including Podops Lap.; he proposes to call them hypopygeal appendages. It is probable that the hypopygeal appendages are homologous in the Podopid genera with the flaps noted in Eluisura irrorata; thus it is proposed to adopt this term for the structures noted in FE. zrrorata, but the difference in position of the appendages in the two groups must, of course, be stressed; it is not a major difference presenting insuperable difficulties in homology. A difficulty is met, however, in assessing its taxonomic import; is it a primitive character? The little known of the life-histories of Podops and of Elvisura suggests quite different ecological roles so that structural convergence is hardly to be considered seriously. The theca of E. irrorata (figs. 2 and 3) is strongly sclerotized and short; it is quite impossible for the vesica and conjunctival appendages to be retracted (a constant character in the Scutellerint). Ventrally there projects a pair of appen- dages, third conjunctival appendages in Singh-Pruthi’s terminology; they are sclerotic except towards the base, where they fuse, and are thus freely movable. On the dorsal surface of the theca project the fused first and second conjunctival 50 ANNALS OF THE SOUTH AFRICAN MUSEUM appendages. As in all other Scutellerini, the second is more sclerotized while the first is membraneous except for the tips. The first conjunctival appendages are apically bifid and terminally sharp. The principal difference to be noted in the Elvisura trrorata Spinola 1. Pygophore from above. H, hypopygeal appendages; p, posterior margin. 2. Aedeagus, ventral. B, basal plates; T, theca; V, vesica; a, b, c, first, second and third conjunctival appendages. 3. Aedeagus, lateral. S, supravesical process. 4. Paramere, lateral. Scale = 0:5 mm. appendages of Elvisura from those of other genera is the large size of the common basal area of the first and second conjunctival appendages. The third pair of appendages are covered, sparsely, by minute points. The vesica exhibits the usual basal convolutions and well-developed ejacula- tory reservoir while the gonopore is oblique and large: these are all common characteristics of the Scutellerint but, in addition, the external part of the vesica NOTES ON THE ETHIOPIAN PENTATOMOIDEA i supports a rigid, densely sclerotized, thin and rough-edged vertical lamina. This structure has also been noted in Steganocerus multipunctatus and (Leston, 19536) termed the supravesical process. It is flecked in E. zrrorata by a few minute broad-based spines and whereas in S. multipunctatus the supravesical process is somewhat thick and terminated at its highest point forwards by a bifurcation, in £. irrorata it is thin and simple anteriorly, albeit very ragged at the margins. As in most Scutellerint, the actual path taken by the ductus seminis remains untraced due to the density and complexity of the vesica; the basal plates have not been studied nor figured except merely to show their attachment. The parameres are simple hook-like pieces with a ridge of hairs; they take a form common to many Scutellerint. Thus the entire male genitalia of FE. zrrorata is of the usual Scutellerinine pattern as described elsewhere (Leston, 1953); it differs in possessing (1) a supravesical process similar to that found in Steganocerus and (2) a pair of movable processes on the pygophore which are tentatively homologized with the hypo- pygeal processes of certain genera of Podopini. Cryptacrus comes (Fabr.) 1803. Fabricius, Syst. Rhyngotorum, p. 130 Amatongas Forest (between Umtali and Beira); the specimen is of var. apicalis Dist. PacHycorin! Dallas, 1851 (Tetyrinz) Deroplax circumducta (Germar) 1837. Germar, Rev. Ent. (Silbermann), v, p. 190 Vryburg, Bechuanaland, 1904 (Jones); Otjitundua (160 miles north-west of Outjo), South West Africa, 11, 1926 (S. Afr. Mus. Exp.). The South West African specimen is close to D. nigropunctata (Stal), but Iam unable to distinguish between these two species. ‘They must remain distinct pending an examination of Stal’s type. Deroplax silphoides (Thunb.) 1783. ‘Thunberg, Nov. Ins. Sp., ii, p. 29 Pretoria, Transvaal, x, 1944 (R. Wilson). This specimen is similarly marked to a long series in the British Museum from Nyasaland. It is referred to the variable D. silphoides. Figs. 5-8 depict the range in variation; it is considered undesirable to add further to the number of named varieties of this species. The male genitalia have been examined (figs. 9-12); they are similar to the genitalia of D. circumducta figured elsewhere (Leston, 1953) and quite unlike the genitalia of Scutellerin. ‘The pygophore encloses a wide proctiger which is marked by a pale band near the base. The aedeagus, shown laterally in fig. 9, encloses a long filimentous vesica of the ‘penisfilum’ type in Baker’s terminology (Baker, 1931). Surrounding the vesica is a long, finger-like extenstion of the conjunctiva. The first and second conjunctival appendages only are present 52 ANNALS OF THE SOUTH AFRICAN MUSEUM (the third is always pre- sent in Scutellerini); they are, as usual, joined basally. The first pair is long and membraneous, terminating in a lightly sclerotized point; the second is biramous, sclerotized, with one arm produced into a filiform process extending to the apex of the vesica. The parameres are shown in two views (figs. 11, 12); the head bifurcates and is closely spinose. On the dorsum of the parameres is a tubercle carrying an armature of bristles while the basal area of the tubercle is similarly armed. It shouldbe noted that figures g and 10 are somewhat dia- grammatic. Deroplax silphoides (Thunb.). 5-8. Variation in a variety from Nyasaland. AMYOTINAE Lest., 1953a (Asopinae) Dorycoris pavoninus (Westw.) 1837. Westwood, Cat. Hemipt., coll. Hope, i, p. 39 Stellenbosch, iv, 1932 (P. v. Heerden). The specimen is of a form near to var. fuscosus. PopoPiNnaE Dallas, 1851 (Graphosomatinae) This subfamily includes the two tribes Podopint Dallas and Tarisim Stal (Graphosomin auctt.). How far it is a natural assemblage remains for future investigation to elucidate, but while the Podopini are reasonably homogeneous the Tarisini are certainly not so and probably many of the species are Pentatominae. Bolbocoris rufus (Westw.) 1837. Westwood, op. cit., i, p. 12 Kamanyab, S.W.A., ii, 1925 (S. Afr. Mus. Exp.); Kaross (30 miles NW. of Kamanyab), S.W.A., ii, 1925 (S. Afr. Mus. Exp.). Despite the widespread NOTES ON THE ETHIOPIAN PENTATOMOIDEA 53 distribution of two of its species the genus Bolbocoris is centred in South Africa, with at least five species in the Transvaal. PENTATOMINAE Leach, 1815 This enormous subfamily is overdue for revision, but until this has been attempted the Stalian groups are retained. co a Deroplax silphoides (Thunb.) 9. Aedeagus, lateral. t, theca; co, conjunctiva; a, b, first and second conjunctival appendages; d, ductus seminis; v, vesica; d.e, ductus ejaculatorius. 10. Pygophore, terminal. p, proctiger; r, right paramere. 11, 12. Two views of right paramere. PENTATOMINI Leach, 1815 Menida lythrodes (Germar) 1837. Germar, Rev. Ent. (Silbermann), v, p. 175 Letaba, Transvaal, v and xii, 1945 (E. C. G. Bedford). ‘From orange-tree —fed on red-scale and on orange.’ 54 ANNALS OF THE SOUTH AFRICAN MUSEUM Menida decoratula (Stal) 1853. Stal, Ofv. Vet. Ak. Forh., x, p. 222 Kaoko Otavi, S.W.A., iii, 1926 (S. Afr. Mus. Exp.). Menida distanti Horvath 1892. Horvath, Term. Fiizet, xv, p. 258 Louis Trichardt, ‘Transvaal, i, 1928 (R. F. Lawrence); Bulawayo, Southern Rhodesia, v, 1917 (R. W. E. Tucker); Mtunzini, Zululand, vi, 1940. Agonoscelis puberula Stal 1853. Stal, op. cit., x, p. 216 Outjo, S.W.A. (S. Afr. Mus. Exp.); Potgietersrust, Transvaal, iv, 1934 (R. F. Lawrence); Upington, Cape Province, vii, 1936 (S. Afr. Mus. staff). Agonoscelis odendaal Dist. 1910. Distant, Ann. Mag. Nat. Hist., (8), vi, p. 95 Salisbury, Southern Rhodesia, v, 1913. Agonoscelis erosa (Westw.) 1837. Westwood, op. cit., i, p. 33 Outjo, S.W.A., 1, 1926 (S. Afr. Mus. Exp.). 7 Amphimachus circumflexus (Stal) 1855. Stal, op. cit., xi, p. 182 Florida, Transvaal, x, 1918 (R. W. E. Tucker). No locality was given by Stal; the majority of specimens seen come from the Transvaal. Antestia lymphata Kirk. 1909. Kirkaldy, Cat. Hemiptera, i, p. 129 Salisbury, Southern Rhodesia. This specimen has been reported upon else- where (Leston, 1952c); it is distinct from the coffee-bugs and these have been removed to Antestiopsis Lest. A. lymphata Kirk. appears to be extremely rare; the British Museum possesses only Dallas’s original type and paratype, described by him as Pentatoma maculata. ‘The South African Museum’s specimen is the first to have a precise locality. Gen. Piezodorus Fieber 1861. Fieber, Europ. Hem., \xx, p. 329 1905. Jakovlev, Rev. russe Ent.., v, p. 142 (Pausias) Pausias, type Piezodorus martin Put., is only distinguishable from Pzezodorus in that the spiracles are pale, whereas in the latter they are dark; P. hessez has pale spiracles, but in all other respects it is a true Piezodorus. ‘These conditions are best satisfied, as are the zoogeographical factors, by sinking Pauszas (Syn. Nov.). When dealing with a large number of specimens from many parts of Africa it has been found that the usual taxonomic characters: length of ventral spine, shape of mesosternal carina, antennal ratios, etc., give little help in separating species. The separation of P. pallescens, purus, and hybnert can probably only be based on the genitalia of the males, but the degree of variation in the parameres NOTES ON THE ETHIOPIAN PENTATOMOIDEA 55 needs further investigation. It appears that the three might best be considered as facets of a single polytypic species; the zoogeography has not helped in this group and ecological factors are probably the cause of the different ‘species’. Key to the African (south of the Equator) species The only other species reported from the area is P. bequaerti Schout. from the Belgian Congo. 1. Spiracles pale. Length over 11 mm. hesset Spiracles dark. Length less than 11 mm. 2 2. Posterior tibiae usually flattened. Second antennal segment always longer than third. purus Posterior tibiae not flattened. Second antennal segment equal to or shorter than third. 3 3. Parameres as in fig. 16. hybnert Parameres more elaborate. pallescens Prezodorus purus (Stal) MORAG Sel, Clos Gtis, 3G 0, OI Kaoko Otavi, $.W.A., ii, 1926; Warmbad (12 miles SE. of Zesfontein), Kaokoveld, S.W.A., ii, 1925; Outjo, S.W.A., i, 1926; Kamanyab, $.W.A., iii, 1925; Kaross (30 miles NW. of Kamanyab), S.W.A., ii, 1925 (all S. Afr. Mus. Exp.). Keimoes and Riemvasmak, Gordonia, Cape Province, vii, 1925 (K. H. Barnard); Aughrabies Falls, Kenhardt Div., Cape Province, v, 1934 (R. F. Lawrence); Murraysburg District, Cape Province, ili, 1931 (Mus. staff). The determination of Prezodorus species is a matter of some difficulty; the species are variable and the Oriental hybnert (Gmel.), now spread over much of East and Central Africa, has been confused with purus by authors. The genitalia are very similar throughout the genus. Fig. 13 shows the aedeagus of purus laterally; the conjunctival appendages are very large and almost entirely membraneous; it has been found difficult to homologize them with the appendages in other groups. The head of the paramere (fig. 14) enables this species to be distinguished from both Aybneri and pallescens (Germ.). The latter is also a common species in South Africa. Piezodorus hesset sp. nov. ®. Pale yellow-brown, a red band across the pronotum between the postero- external angles, ante-ocular area of head red; a black metallic line just within and parallel with the lateral paraclypeal margins. Hemielytra with a brown speck on the apex of the median margin of the clavus. Antennae indefinitely reddish. Venter pale yellow-brown, spiracular eminences pale dirty-yellow. Pronotum and head punctate, the former with its anterior margin raised and rounded, its lateral margins straight, reflexed and yellow. Elytral texture thin, semi-transparent. Membrane glassy, colourless. Antennal segments 0-42, 1-00, 56 ANNALS OF THE SOUTH AFRICAN MUSEUM 1:16, 1-11 and 1-01 mm. long respectively. Ventral spine stout, reaching to before the intermediate coxae. @ Length 12-4 mm. Maximum pronotal breadth 6-2 mm. 6 Length 11-8 mm. Maximum pronotal breadth 6:2 mm. Holotype 9: Otjikondo (40 miles WNW. of Outjo, S.W.A., 1, 1925 (S. Afr. Mus. Exp.). Piezodorus purus Stal. 13, aedeagus, lateral. 14, head of paramere, lateral. Piezodorus hessei sp. nov. 15, metathoracic gland opening, and evaporatorium. Piezodorus hybneri (Gmel.). 16, head of paramere, lateral. Piezodorus hesset sp. nov. 17, head of paramere, lateral. Paratypes: Three ¢ ¢ same locality as holotype; one 9 Kaoko Otavi, Kaokoveld, S.W.A., ili, 1926; one ¢ Otshu, Hoarusib River (50 miles W. of Kaoko Otavi), S.W.A., ii, 1926; one 3 Kaross (30 miles NW. of Kamanyab), S.W.A., ii, 1925 (all S. Afr. Mus. Exped.). Holotype and paratypes in the South African Museum except for one para- type presented to the British Museum (Nat. Hist.), and two retained in the author’s collection. NOTES ON THE ETHIOPIAN PENTATOMOIDEA 57 Gynenica capenert Lest. 1953. Leston, Rev. zool. bot. Afr. (in press) M’fongosi, Zululand, i, 1935 (W. E. Jones). ‘This species appears to be con- fined to Natal and Zululand. Gynenica marginella Dallas 1851. Dallas, List. Hemipt. Brit. Mus., i, p. 181 Resolution, Grahamstown, Cape Province, i-iv, 1928 (Miss Walton). Stenozygum alienatum (Fabr.) 1803. Fabricius, Syst. Rhyngotorum, p. 173 Wankie, Southern Rhodesia, xi, 1923 (C. W. Tyler). Boerias brunnea Jensen-H. 1931. Jensen-Haarup, Ent. Medd., xvii, p. 325 Knysna, Cape Province, 1, 1931 (K. H. Barnard). The previously known localities for this species are given by Leston (19526); it is confined to the south- west of the Cape Province. Boerias maculata ( Dist.) 1910. Distant, op. cit., (8), vi, p. 87 Warrenton, Cape Province, iv, 1931 (J. T’.). Not previously recorded from the Cape Province, but known from Southern Rhodesia, the Transvaal, and Basutoland. Boerias rubrocincta (Dist.) 1910. Distant, op. cit., (8), vi, p. 87 Pretoria, Transvaal, v, 1927 (S.M.); Smithfield, O.F.S., 1910 (Kanne- meyer), ‘On orangia’. ‘These two records considerably extend the published distribution. Distant described the species from Natal. Boerias victorin (Stal) 1856. Stal, op. cit., xiii, p. 194 [nec Jeannel, 1913] Keurbooms River, Knysna, Cape Province, i, 1931 (K. H. Barnard). Durmia haedula (Stal) 1865. Stal, Hemipt. Afric., i, p. 149 Louis Trichardt, Transvaal, i-1i, 1928 (R. F. Lawrence). Durmia tomentwentris (Germ.) 1837. Germar, Rev. Ent. (Silbermann), v, p. 168 M’fongosi, Zululand, i-ii, 1935 (W. E. Jones). Carbula litigatrix Kirk. 1909. Kirkaldy, Cat. Hemipt., 1, p. 88 M’fongosi, Zululand, 1, 1935 (W. E. Jones); Cayimaeis (Caimaiais, 25 miles SE. of Zesfontein), Kaokoveld, $.W.A., iii, 1925 (S. Afr. Mus. Exp.). 53 ANNALS OF THE SOUTH AFRICAN MUSEUM Veterna sanguineirostris (Thunb.) 1822. Thunberg, Hemipt. Rostr. Cap., 11, p. 4 1890. Distant, C.R. Soc. ent. Belge, xxxiv, p. lv (mimica) 1892. id., Naturalist in Transvaal, p. 250 (patula) I can find no constant characters upon which to separate V. patula Dist. and V. mimica Dist. from this species; they are weak but geographical subspecies; sanguineirostris centred in the Cape Province, patula in the Transvaal, Natal, and Southern Rhodesia, and mimica in the Belgian Congo. (Syn. Nov.) Subspecies sanguineirostris Somerset West, Cape Province, vili-ix (A. J. Hesse). Subspecies patula Pretoria, Transvaal, xi, 1930. Diploxys fallax Stal 1865. Stal, Hemipt. Afric., i, p. 129 M’fongosi, Zululand, i, 1935 (W. E. Jones). Coponia waterbergensis (Dist.) 1902. Distant, Ann. S. Afr. Mus., ii, p. 252 Louis Trichardt, ‘Transvaal, i-ii, 1928 (R. F. Lawrence). The specific distinc- tions are obscure in this genus but the specimen before me has been compared with Distant’s type. Undoubtedly C. cornuta (Dist.), C. thoracica (Dist.) and C. waterbergensis are very closely allied, and their differences, based on the pro- notal angles, are scarcely specific. | Lerida punctata (P. de Beauv.) 1805. Palisot de Beauvois, Ins. Afr. Amer., p. 84 M’fongosi, Zululand, 1, 1935 (W. E. Jones). Farnya versicolor (Dist.) 1881. Distant, Proc. 5); ram 1 Mannchen aus NW. Cape Province, Namaqualand, Bowesdorp (zwischen Garies und Springbok), Sept. 1941, leg. Museum Staff (Monotypus) in der Sammlung des South African Museum. Diese Art ist wegen ihrer ganz eigenartigen Farbe der Fliigeldecken sofort zu erkennen und sieht keiner anderen Jselma-Art ahnlich. Morphologisch steht die neue Art in der Gruppe von JI. flavipennis Haag-R. und JI. namaqua Pér., von welchen sie aber durch die vollkommen abweichende Farbe der Fliigeldecken, die sehr grobe Skulptur des Vorderkorpers und durch die Fihlerform des Mannchens leicht zu unterscheiden ist. Ich benenne diese interessante neue Art zu Ehren des Herrn Kollegen Dr. A. J. Hesse im S.A. Museum, Cape Town. Iselma csikw sp. nov. Gestreckt, Vorderkérper braunschwarz, Fliigeldecken in der Mitte heller, an den Seiten und neben der Naht breit dunkler braun, die ganze Ober- und Unterseite mit anliegenden, sehr dichten, grauweissen Haaren bedeckt, so dass der Koérper grau erscheint. Kopf langoval, stark gewolbt, Schlafen lang und parallel, hinten halbkreisf6rmig abgerundet; die Punktierung gleichmassig, ziemlich fein und dicht, die Zwischenraume zwischen den Punkten etwa so gross wie die Punkte selbst. Die Behaarung kurz, hinten nach vorne und vorne nach innen gerichtet. Halsschild viel langer als breit, in der Mitte am breitesten, STUDIEN UBER SUDAFRIKANISCHEN MELOIDEN e 65 nach vorne starker und fast gerade verjiingt, nach hinten ziemlich parallelseitig und nur vor dem sehr scharfen und dicken Basalrand ausgeschweift. Die Mitte nur leicht und die Scheibe beiderseits kaum merkbar verflacht. Die Skulptur fein aber sehr dicht, die Behaarung anliegend und verschiedenartig gerichtet. Fltigeldecken lang parallelseitig, Schultern abgerundet etwas vortretend, die Skulptur besteht aus dusserst dicht stehenden, raspelartigen, in der Quere stel- lenweise zusammenfliessenden Punkten. Die Behaarung sehr dicht und anlie- gend, schrag nach aussen gerichtet. Fihler fadenformig, die Glieder ziemlich gleichformig, am Ende kaum dinner, das 3. Glied das langste, doppelt so lang wie das 4. und auch langer als das Endglied. Die Glieder von 4. an werden stufenweise kiirzer, so dass das 10. merklich kiirzer ist als das 4. Bezne einfach, Tarsen mit langen Gliedern, Schienen am Ende mit langen goldgelben Haaren, Tarsen unten ebenfalls goldgelb behaart. Die Enddorne der Hinterschienen kurz und dick, ihr Ende breit und schrag abgestutzt, der innere Enddorn etwas langer. Unterseite sehr dicht mit raspelartigen Punkten besetzt, vollkommen anliegend grauweiss behaart. Lange: g mm. Breite: 3 mm. 1 Weibchen aus Ost-Afrika, Tanagebiet, Witu (Monotypus) in der Samm- lung des Ungarischen Naturwissenschaftlichen Museums in Budapest. Diese Art ist durch mehrere Eigenschaften sehr charakteristisch. Sie besitzt das nordlichste Vorkommen in der Gattung und morphologisch steht sie in der Nahe von I. quadrimaculata Borchm., doch steht sie weit von den bekannten Arten. Eine ahnliche Farbung und Behaarung des Korpers, sowie eine ahnliche Fihlerform kommt bei den ubrigen Arten nicht vor. Ich erlaube mir diese neue Art Herrn Direktor E. Csiki, dem verdienstvollen ungarischen Coleopterologen zu widmen. Im folgenden gebe ich eine Bestimmungstabelle der bisher bekannten Arten der Gattung Iselma Haag-R. 1 (40) Die Enddorne der Hinterschiener viel kiirzer als das 1. Glied der Hintertarsen. 2(15) Der ganze Korper schwarz, selten mit metallischem Schimmer. 3 (12) Der ganze Korper tiefschwarz, héchstens die Fliigeldecken mit schwachem Metall- schimmer. 4 (5) Fliigeldecken am Ende zugespitzt, schwarz behaart, Kopf und Halsschild eingedriickt (nach Haag-Rutenberg). 1. hirsuta Thunb. 5 (4) Fliigeldecken am Ende abgerundet. 6 (7) Ober- und Unterseite mit ausserst langen und dichten, vollkommen schwarzen Haaren besetzt, die Behaarung ist doppelt, d.h. es sind langere, mehr aufstehende und dazwischen kiirzere aber kaum anliegendere Haare vorhanden. Fliigeldecken an der Basis neben den Schultern stark eingedriickt, der Eindruck reicht schrag weit nach hinten. Halsschild in der Mitte und beiderseits neben der Mitte stark eingedriickt. L.: 15 mm. NW. Cape Province (Nieuwoudtville). . 2. lanuginosa sp. nov. 7 (6) Die Behaarung, besonders die anliegenderen Haare viel kiirzer, die langeren Haare viel sparlicher, der Eindruck nur an der Schulter deutlich, nicht nach hinten verlangert, die Eindriicke des Halsschildes viel schwacher, oder sogar vollkommen fehlend. 8 (9) Kopf flach und breit, hinten abgestutzt, Schlafen mit gerundet rechtwinkligen Beulen, ohne Mandibeln scheinbar quadratisch. Halsschild beiderseits neben der Mitte sehr schwach eingedriickt, grob und dicht punktiert. Fliigeldecken sehr dicht, 66 9 (8) 10 (11) II (10) 12 (3) 13 (14) 14 (13) 15 (2) 16 (23) 17 (18) 148) (7) 19 (20) O (19) 21 (22) ANNALS OF THE SOUTH AFRICAN MUSEUM vorne etwas zusammenfliessend punktiert, die anliegende Behaarung dunkelbraun und neben der Naht etwas seidenglanzend. L.: 9-12 mm. Cape Province. -» . 3. Ursus Thunb: Kopf langlich, Stirn mehr gewolbt, hinten vollkommen breit abgerundet, ohne Mandibeln langer als breit. Fliigeldecken mit deutlicherem Erzschein, dichter behaart, auch hinten mit in der Querrichtung zusammenfliessender Punktierung. Grosser; letztes Abdominalsegment beim Mannchen in der Mitte etwas ausgeschnit- ten und die Seiten beiderseits etwas eckig aufgebogen. 3. Glied der Fiihler beim Mannchen breiter und auch etwas langer als das 4. Kopf breiter und flacher, hinten weniger abgerundet. L.: 13 mm. Cape Province (Giftberg, Van Rhyns- dorp). . . . 4. simillima sp. nov. Kleiner; letztes Abdominalsegment beim Mannchen abgestutzt, Oberflache breit verflacht, Seiten beiderseits ohne Ecke. Das 3. Fihlerglied beim Mannchen schmaler und etwas ktirzer als das 4. Kopf schmaler und langer, hinten breiter abgerundet, Stirn gewolbter. L.: 7-9 mm. Cape Province (Giftberg, Van Rhyns- dorp). . . . 5. analis sp. nov. Fliigeldecken mit ausgesprochenem Metallschimmer, die Unterseite, besonders aber das Abdomen ebenfalls metallisch. Kopf und Halsschild, sowie die Brust meist auch mit metallischem Schimmer. Halsschild langer als breit, mit scharfem und dickem Basalrand, beiderseits neben der Mitte der Scheibe mit je einem flachen, langlichen Eindruck; Kopf ziemlich gross, viel breiter als die Halfte des Halsschildes. Die anliegende Behaarung des Korpers braungelb, die aufstehenden Haare besonders am Vorderkorper sehr lang und schwarz, die Fliigeldecken mit Bronzeglanz. Oberflache der Fliigeldecken schwach eingeflacht. L.: 10-11, 5 mm. Cape Province (Clanwilliam). . 6. planidorsis Pér. Halsschild kugelig, nach allen Richtungen gewolbt, so breit wie lang, die Basis nur schwach gerandet, Scheibe in der Mitte und beiderseits neben der Mitte ein- gedriickt. Kopf sehr klein und flach, nur halb so breit wie der Halsschild. Die anliegende Behaarung dunkelbraun, die aufstehende schwarz. Fligeldecken blaugriin, Unterseite blaulich, Kopf und Halsschild mit griinlich-kupferigem Schimmer. L.: 7-9, 5 mm. Cape Province (Clanwilliam, Worcester). . 7. metallescens Per. Korper schwarz, Fliigeldecken aber einfarbig rotgelb, aelbror braungelb, oder gelbrot mit schwarzen Flecken, oder schwarz mit gelbroten Makeln. Fliigeldecken mit gelber, oder rotlicher Zeichnung, oder gelb mit schwarzen Flecken. Fliigeldecken schwarz, die Seiten von den Schultern bis zum Nahtwinkel aber sehr breit rot gesAumt. Kopf und Halsschild ausserordentlich grob punktiert, die Punktierung der Fliigeldecken auch sehr grob und viel dichter als die des Hals- schildes. Halsschild ohne anliegende Behaarung, nur sehr lang schwarz behaart, Fliigeldecken mit schwarzer, anliegender Behaarung und nur hie und da mit einigen langen, aufstehenden Haaren. Letztes Abdominalsegment des Mannchens ausgeschnitten. L.: 15 mm. N.W. Cape Province Ne ae Bowesdorp.) . 8. hessei sp. nov. Fliigeldecken gelbrot mit schwarzen Flecken, oder schwarz mit seibed Makeln. Fliigeldecken gelbrot mit dunkler, breiter Naht, welche sich nach hinten erweitert, oder die dunkle Farbung erweitert sich oft so stark, dass nur die Schultern und je eine lange Makel am Seitenrand hinter den Schultern gelb bleibt. Halsschild sehr lang, fast zylindrisch gewolbt, sehr sparlich und ziemlich fein punktiert. Die anliegende Behaarung der Fliigeldecken gelbrot. L.: 6-11 mm. S.W. Afrika (Abachaus, 48 km. NW. von Otjiwarongo). . . . g. hobohmi Kasz. Fliigeldecken gelbrot mit schwarzen Flecken, die Naht héchstens am Ende dunkel. Beine und Fiihler dunkel, héchstens die Hinterschienen gelbrot. Fliigeldecken gelbrot, mit je 2 grossen, der Lange nach miteinander verbundenen schwarzbraunen Flecken; nur die Basis, das Ende, die schmale Naht und der Seitenrand, sowie die Mitte an den Seiten zwischen beiden Flecken gelb. Halsschild sehr dicht und ziemlich fein punktiert, die Mitte stark eingedriickt. Kopf breit und gewolbt. Die 22 (21) 23 (16) 24 (25) 25 (24) 26 (35) 27 (28) 28 (27) 29 (34) 30 (33) 31 (32) 32 (31) 33 (30) 34 (29) 35 (26) 36 (37) 37 (36) 38 (39) STUDIEN UBER SUDAFRIKANISCHEN MELOIDEN 67 Oberflache hellgelb dicht behaart. Fiihler des Mannchens am Ende des 4. und 5. Gliedes dick. L.: 12 mm. Ost-Afrika (Lukuledi). . 10. qguadrimaculata Borchm. Schienen und Tarsen, sowie die Fihler gelbrot. Fliigeldecken gelbrot, am Ende aber mit je einem grossen, ovalen, schwarzen Fleck, so dass die Naht im hinteren Finftel auch schwarz ist. Kopf viel schmaler, Halsschild viel sparlicher punktiert, Fliigeldecken hellgelb, Vorderkorper braun und Unterseite ganz anliegend gelbgrau behaart. L.: 12 mm. Ost-Afrika (Barikiva). . . . It. ttbialis Borchm. Fliigeldecken einfarbig gelbrot, rotgelb, hdchstens mit ausserst schmaler dunkler Naht, oder der ganze Korper einfarbig graubraun. Der ganze Korper einfarbig graubraun. Die Behaarung sehr dicht und anliegend grauweiss, Halsschild lang, mit scharfem Basalrand, Kopf schmal und lang, mit parallelen und langen Schlafen, hinten aber breit abgerundet. Das 3. Fiihlerglied sehr langgestreckt, fast doppelt so lang wie das 4. L.: 9g mm. Ost-Afrika (Tana- gebiet, Witu). Neil ie L2H CSUKI SD. MON: Korper schwarz, die Fliigeldecken aber gelbrot, oder rotgelb. Fliigeldecken nur mit einfarbiger, schwarzer Behaarung. Meist grosse Arten. Fliigeldecken mit dunkler Naht. Fihler stark und lang, etwas gesagt, Halsschild oval, dicht und grob punktiert. Die schwarze Behaarung der Fliigeldecken kurz. L.: 11-14 mm. (nach Haag-Rutenberg und Péringuey). Cape Province, Cape Division. . . . 13. rufipennis Haag-R. Fligeldecken einfarbig gelb oder rot, die Naht nicht dunkel. Grosser, 11-15 mm. Fliigeldecken gelb, Halsschild breit, stark gewolbt, kaum langer als die grésste Breite. Die Mitte, sowie die Scheibe beiderseits neben der Mitte des Halsschildes flach eingedriickt. Oberflache des Halsschildes sehr grob punktiert. L.: 13-14 mm. Cap der guten Hoffnung. . . 14. flavipennis Haag-R. Halsschild ausserst grob punktiert, die Mittellinie der Scheie unpunktiert und etwas aufgewolbt, die Scheibe beiderseits hie und da mit einigen (oder nur mit je einem) Spiegelfleck. Fliigeldecken starker punktiert und an den Seiten lederartig gerunzelt. L.: 14-16 mm. NW. Cape Province (Namaqualand). 15. namaqua Pér. Fligeldecken rot. Halsschild schmaler und langer, sehr grob aber sparlich punktiert, Kopf ebenfalls schmal und sehr lang, Wangen parallel, Kopf hinten halbkreis- formig abgerundet. Fliigeldecken grob punktiert und die Punkte meist in Quer- richtung zusammenfliessend. L.: 11-15 mm. N.W. Cape Province (Namaqua- land). . . . 16. brunneipes Haag-R. Kleiner, 9 mm. nicht tiberragend. Halsschild ziemlich rundlich, stark punktiert, dazwischen fein gerunzelt, die Mitte und beiderseits die Scheibe schwach verflacht. Kopf ziemlich breit und kurz, Schlafen hinten sich plotzlich verschmalernd. Fliigeldecken sehr dicht und grob punktiert, zusammenfliessend. L.: 7-9 mm. Cape Province (Clanwilliam). . . . 17. boschimana Pér. Fliigeldecken hell behaart, oder die Haare am Grunde schwarz, am Ende aber gelb, so dass die Behaarung von oben gesehen hell erscheint, oder zwischen den gelben Haaren sind auch schwarze Haare untermischt. Fuhler stark gesagt. Oberseite doppelt behaart, die lange, aufstehende Behaarung ist schwarz, die kiirzere einfarbig gelbrot, die Basis der Haare auch gelbrot; Kopf und Halsschild mit dichter und anliegender, gelber Behaarung. Halsschild langlich und die Scheibe beiderseits flach eingedriickt. Die Punktierung der Fliigeldecken ziemlich sparlich. L.: 10-12 mm. Cape Province (Grahamstown, Willowmore). . 18. optata Pér. Fihler einfach, fadenformig, nicht gesagt, Halsschild kiirzer, meist rundlich. Die kurzere Behaarung der Fliigeldecken am Grunde schwarz, am Ende aber gelb, so dass die Behaarung von oben gesehen einfarbig gelb erscheint. Halsschild in der Mitte der Scheibe und auch beiderseits eingedriickt. Kopf verhaltnissmassig klein und langlich, die Punktierung der Fliigeldecken sehr dicht. Grosser, die Behaarung mehr braunlich. L.: 8-g mm. Cape Province (Cape Town, Van Rhynsdorp). . . . 1g. rubripennis Lap. 68 ANNALS OF THE SOUTH AFRICAN MUSEUM 39 (38) Halsschild scheibenformig, die Mitte und die Scheibe beiderseits meist ebenfalls flach, der Kopf viel kiirzer, rundlich und breiter. Die Punktierung der Fliigeldecken etwas feiner und dichter. Die Behaarung mehr gelblich. Kleinere Art. L.: 6-8 mm. N.W. Cape Province (Namaqualand). . . . 20. pallidipennis Haag-R. 40 (1) Der innere Enddorn der Hinterschienen etwas langer als das lange 1. Hintertarsen- glied. Der aussere Enddorn kaum halb so lang als der innere. K6rper schwarz, Flugeldecken gelb. 41 (42) Halsschild langlich oval, die Mittellinie erhaben und lang, Kopf mit anliegender gelber Behaarung. L.: 7 mm. (nach Péringuey). NW. Cape Province (Namaqua- land). - » . 21. okiepana Pér. 42 (41) Halsschild rundlich, kugelig, die Mittellinie nur in der Mitte etwas erhaben, oder vollkommen flach. Kopf ohne gelbe Behaarung. Fliigeldecken blassgelb, gelb behaart, hie und da mit einigen schwarzen, aufrecht stehenden Haaren. L.: 6-8 mm. N.W. Cape Province (Bushmanland). . . . 22. longispina Kasz. Gen. 4 SB. ee 2 \ \ i, F AN 3 ie ~~ 7 RY» Saeed > are ee +e —— = ¥ Sela ye MA ca cae ee ete. Gta we rv oe LA ila. avee” uth. on é Say Sd - ; bh ‘ : oa YN | at ae Bet aie oe aTve.* ij \ > Es ST re Sere: v ae 2 Se aes ei ‘ oe " e a, = if eS he ISSUED APRIL 1954 PRICE 15s. aie * is PRINTED FOR THE a. _ TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY THE RUSTICA PRESS (PTY.) LIMITED, COURT ROAD, WYNBERG, CAPE 7. South African Pycnogonda. By K. H. Barnarp, D.Sc., F.L.S. (With 34 text-figures. ) This paper is the result of an examination of the South African Museum material comprising mainly the collection made by the Cape Government trawler s.s. Pieter Faure, together with a moderate number ‘of specimens obtained from time to time by shore-collecting. The history of our knowledge of South African Pycnogonids has been given by Flynn (1928). Gordon (1932) has redescribed three species from the type material. Since then Helfer (1937), Barnard (1946) and Stock (1951) have described material from South Africa south of lat. 15° S. Flynn (1928) gave a list of the then-known South African species, including a Mozambique record. The list omits Parapallene merstraszi which Flynn himself records in his paper; but on the other hand it includes Pallenopsis crossland: which neither Flynn nor any other author has recorded from South Africa. Not counting ‘Pallenopsis sp. Loman’ (1923) the list contains thirty-six species. A second paper by Loman, also published in 1923, and containing two new records (one from Angola), was not mentioned by Flynn. In 1937 Helfer described three new genera and species; in 1946 Barnard one new genus and ten new species; and in 1951 Stock one new genus and two new species. In the present paper three of Loman’s species and records, one of Flynn’s species, and one of Helfer’s are regarded as synonyms. The list is appreciably longer than Flynn’s, and contains fifty-three species. All the accepted families are represented in the fauna. Of the newly recorded constituents it is interesting to note the occurrence of the genera Pipetta and Austrorapius. The genus Queubus Brnrd., of uncertain affinities, is placed near but not in the family Pycnogonidae. No examples of ‘extra-legged’ (10 or 12 pairs) species have been found in South African waters. These remarkable Pycnogonids are mostly inhabitants of the Antarctic, though examples have also been found in the Caribbean Sea (see: Hedgpeth, 1947, Smiths. Misc. Coll., cvi, pp. 9 sqq.). The early stages of some species are parasitic in the zooids of Hydroids, causing the formation of galls. An example from South Africa has been noted by Warren (1908, Ann. Natal Mus., i, p. 293) and referred to by Flynn (1928, P- 5): Collectors and others should be warned against the use of cotton wool as a plug in tubes or bottles to prevent shaking while in transit, because no other animals, not even Crustacea, seem to get their claws so inextricably entangled SI mor, XLi, PART 3. AUC Iso 82 ANNALS OF THE SOUTH AFRICAN MUSEUM in it. A plug of soft paper should be interposed between the animals and the cotton-wool plug in a tube. List of South African Species Colossendeidae Colossendeis colossea Wilson , macerrima Wilson Pipetia capensis Brnrd. Rhopalorhynchus kréyert. Wood-Mason Nymphonidae Nymphon capense Hodgson ? = phasmatodes Bohm 5, australe Hodgson 5, signatum Mobius », adistensum Mobius 5, ptlosum Mobius syn. bipunctatum Flynn » comes Flynn » affins Stock 5, natalense Flynn 5, macroctenatus Brnrd. 5 erenatiunguis Brnrd. » settmanus Brnrd. Pallenidae Callipallene sp. Pseudopallene gilchristt Flynn Parapallene spinosus (Mobius) af nierstraszi Loman a calmam Flynn e hodgson: Brnrd. ¥ algoae Brnrd. Metapallene dubitans (Hodgson) syn. Procidella gibber Helfer Pallenoides magnicollis Stock Pallenopsis oscitans (Hoek) 5, brevidigitata Mobius 5 intermedia Flynn syn. Pallenopsis sp. Loman, 1923, P. fluminensis Loman, 1923, non Kroyer 5 capensis Brnrd. », (Rigona) crossland: Carp.* be » ovalis Loman Hannoma typica Hoek * See p. 116, SOUTH AFRICAN PYCNOGONIDA 83 Phoxochilidiudae Phoxochilidium capense Flynn Anoplodactylus aculeatus Mobius ms pelagicus Flynn ‘Pallene’ lappa Bohm Endeidae Endeis clipeatus (Mo6bius) 5, mollis (Carp.) Ammotheidae Béhmia chelata (Bohm) 5, tuberosa Mobius Achelia quadridentata (Hodgson) 5, brevicauda (Loman) Nymphopsis cuspidata (Hodgson) syn. abstrusus Loman Kyphomia setacea Helfer ? Ammothella setacea Ascorhynchus ornatum (Helfer) Austroraptus thermophilus Brnrd. T anystylidae Tanystylum ornatum Flynn Discoarachne brevipes Hoek Pycnogonidae Pycnogonum cataphractum Mobius $5 forte Flynn si microps Loman 45 portus Brnrd. 3 pusillum Dohrn Incertae sedis FT. Queubus jamesanus Brnrd. Key to the South African Familtes Body extended, crurigers distinctly separated (except Pallenopsis subg. Rigona, where they are contiguous). Proboscis directed forward (except Hannonia). A. No chelifers (in adult). No auxiliary claws. Colossendeidae B. Chelifers present. 1. Ovigers present in both sexes, with compound spines (fig. 3d). Ocular tubercle not, or not far, in front of 1st crurigers. a. Palps present, 5-jointed. Nymphonidae b. Palps absent, or knob-like, or 3-4-jointed. Pallenidae 2. Ovigers present in g only, with simple spines. Phoxochilidiidae C. Chelifers and palps absent. Auxiliary claws present. Endeidae Body more or less compact, segmented or the segmentation suppressed, crurigers usually narrowly separated. Proboscis directed forwards, or more or less downwards, or bent under body. 84 ANNALS OF THE SOUTH AFRICAN MUSEUM A. Palps present. Ovigers present in both sexes, without apical claw. 1. Palps 6-10-jointed. Ammotheidae 2. Palps 4-6-jointed Tanystylidae B. Palps absent. Ovigers sometimes present in § only, with apical claw Pycnogonidae Fam. CoLOSssENDEIDAE 1908. Loman, Siboga Exp. monogr., xl, p. 21 and conspectus facing p. 19 (subfam. Colossendeinae.) 1913. Schimkewitsch, Zool. Anz., xli, p. 614. 1913. Bouvier, 2me Exp. Antarct. Franc., p. 37. 1933. Calman and Gordon, Proc. Roy. Soc. Lond., B, cxili, p. 113. Legs 8-12. Body of 4-6 segments, fused or free. Proboscis large, immovable. Chelifers absent (usually) in adult. Palps long, usually 8-10-jointed, arising from knob-like processes. Ovigers in both sexes, arising from knob-like processes, 10-jointed, with apical claw, last 4 joints with non-serrate (usually) spines. Bases of palps and ovigers approximate or contiguous. Legs long, tarsus usually of fair length, tarsus and propodus unarmed or not armed with strong spines, without auxiliary claws. Genital apertures on all legs (2nd coxal joint) in both sexes. Key to the South African Genera Octopodous 1. Body segments fused. Crurigers not widely separated. Very large species. Colossendeis 2. Body segmented. Crurigers widely separated. a. Abdomen elongate. 2nd coxa longer than Ist or 3rd. Pipetta b. Abdomen minute. All three coxae subequal Rhopalorhynchus Gen. Colossendeis Jarzinsky 1881. Hoek, Rep. H.M.S. ‘Challenger’, iii; pp. 28, 61. 1919. | Bouvier, loc: cit.) p, 53: 1917. id., Res. Scz. Camp. Monaco, fasc. 51, p. 7. 1932. Gordon, “Discovery Rep., vi, p. 11. 1936s.) id Austral Antarct. Exp. (C2 ee. ps 7 1944.) 1d! Bb eAUNE ZZ. Antarct: Hap... B. ve pte lor Octopodous. Proboscis more or less evenly cylindrical, or constricted basally. Key to the South African Species Last 3 joints of the legs together much shorter than two-thirds length of 2nd tibial joint (brevitarsal group). 1. Proboscis stout, constricted basally. Palps much longer than proboscis, 4th joint shorter than 2nd. colossea 2. Proboscis slender, evenly cylindrical. Palps a little longer than proboscis, 4th joint longer than end. macerrima ¢ a —— eo SOUTH AFRICAN PYCNOGONIDA 85 Colossendeis colossea Wilson 1881. Wilson, Bull. Mus. Comp. Zool. Harv., viii, p. 244, pl. 1, fig. 1., pl. 3, figs. 5-7. Seateettioek, loc. cit., pp. 61, 147, pls, figs. 1, 2;+pl.°10, figs. 1-5 (gigas)... 1917. Bouvier, loc. cit., p. 13, pl. 1, fig. 2, pl. 2, fig. 1 (coloured) (references). 1923. Calman, Rec. Ind. Mus., xxv, p. 266. 1928. Flynn, Fish. Mar. Biol. Surv. Rep., vi, Spec. Rep. 1, p. 7. 1944. Gordon, loc. cit, p. 11. 1953. Stock, Temminckia, ix, p. 311. Proboscis stout, constricted in basal third. Palp much longer than proboscis, 4th joint shorter than and, 6th about twice as long as 5th; 7th, 8th and oth subequal. Femur of legs subequal to ist tibia, much longer than 2nd tibia. Length of body (excl. proboscis) 21 mm., proboscis 33 mm., legs 240 mm. (Hoek: 32°5, 47-5 and 301 mm. resp.). Reddish (see Bouvier’s coloured figure). Localities: off Cape Point, 660-755 fathoms (Flynn); off Cape Point, 650- moeagniath. (S. Afr. Mus.); off Gneka River, 43 fath., and 29° 44 S. 31° 20’ E., 46 fath. (Flynn). Distribution. Southern Indian Ocean, East Indian Sea, off west coast of South America, Japan, North and Central Atlantic. Colossendeis macerrima Wilson foam Visor, loc: cit-)'p. 246, pl. 1; fig. 2, pl. 3, figs: 9-12, pl. 5, fig. 32: 1881. Hoek, loc. cit., pp. 64, 147, pl. 8, figs. 3-7 (leptorhynchus) far ye0) Bouvier, loc. cit., p. 10, pl. 1, fig. 1 (coloured), pl: 3, figs. 1) 2. maze Calman, Toc. cit.,‘p: 267." ' 1928. Flynn, loc. cit., p. 7 (macerrime, and p. 6 macerriama, typ. err.). 1953. Stock, Temminckia, ix, p. 307, fig 17 e-h. Proboscis slender, nearly evenly cylindrical, often slightly upturned in distal third. Palps only a little longer than proboscis, 4th joint much longer than 2nd, 5th and 6th subequal, 7th and 8th subequal, gth usually a little longer than 8th. Femur of legs subequal to Ist tibia, much longer than and tibia. Length: body 17 mm., proboscis 34 mm., legs 155 mm. Reddish. Localities: off Cape Point, 660 fath. (Flynn); off Cape Point, 700-800 fath. (S. Afr. Mus.); off Gneka River, 43 fath. (Flynn). Distribution. As for colossea. Remarks. Much less abundant than colossea in Cape waters. Flynn has remarked that the shallow-water records of both species need confirmation. Gneka* River is about midway between Keiskamma Point and East London. * Spelt Nieca in the s.s. Pieter Faure log-books and Rep. Gov. Biologist 1901 chart; Ncera on sheet 5 of the topographical map issued by the Irrigation Dept. 1939; also sheet SE 35/24 Port Elizabeth, 1:500,000. 1950. 86 ANNALS OF THE SOUTH AFRICAN MUSEUM Gen. Pipetta Loman 1904. Loman, Tijdschr. Ned. Dierk. Ver. (2) viii, p. 264. TQOG,) 1d loc vent piney. 1914. Hodgson, Zool. Anz., xlv, p. 159. 1915. id., Ann. Mag. Nat. Hist. (8) xv, p. 141 (English version of 1914 paper). 1927. id., D. Stdpol. Exp., xix (zool. xi), p. 314. Octopodous. Body segmented, very slender. Crurigers slender, very widely separated. Ocular tubercle conical, no eyes. Proboscis elongate, very slender. Abdomen elongate, slender. Palps slender, 8-jointed. Ovigers 10-jointed, with small apical claw. ‘Legs slender, no auxiliary claws. Two species hitherto known: P. weberi Loman 1904 from the East Indies, 2,081 metres; and P. australis Hodgson 1914 from the Antarctic (‘Gauss’ winter quarters), 2,450 metres. Calman and Gordon (1933, loc. cit., p. 114) exclude this genus from the Colossenderdae; Hedgpeth (1947, Smiths. Misc. Coll., cvi, p. 5) includes it with a question mark. Pipetta capensis Brnrd. Bic. 1946. Barnard, Ann. Mag. Nat. Hist. (xi) 13, p. 60. Body rod-like, slender, with scattered microscopic granules. Crurigers separated by increasingly wide intervals, the end pair slightly the longest, 4th pair shortest; 1st with a low conical elevation on dorsal apex, 2nd-4th each with a strong upstanding spiniform process. Ocular tubercle high, slender, conical, no eyes. Proboscis proximally slightly swollen, distally very slender. Palps slightly longer than proboscis, 1st joint short, swollen, 2nd longest, apically swollen, with a few setae, 4th long, about two-thirds length of 2nd, with a few setae, 3 spine-setae and a longer one on apex, 5th-8th joints short, sparsely setose. Oviger 1st joint short, swollen, 2nd a little more than twice the grd, 4th one and a half times the 2nd, 5th short, 6th longest, apically clavate, 7th-1oth joints short, more or less swollen, each with a few (2-4) blunt denticles on inner margins, apical claw small. Legs very slender, 3rd coxa with a strong upstanding slender spiniform process; femur with a few long setae and a spine-seta on dorsal apex; Ist tibia slightly shorter than femur, 2nd tibia slightly shorter than ist tibia, both with scattered setae; tarsus about half length of propodus, the latter slender, slightly curved, with fine setae on lower margin, claw feebly denticulate on lower margin. Body, tip of proboscis to tip of abdomen, about 7-5 mm. (specimen broken in half); proboscis 3:5 mm., abdomen 1:75 mm., femur 2 mm., Ist tibia 1°75 mm., 2nd tibia 1-5 mm. Locality: off Cape Point, N. 81° E. 32 miles, 460 fath. 1 specimen (S. Afr. Mus.). SOUTH AFRICAN PYCNOGONIDA 8 7 t C Fig. 1. Pipetta capensis Brnrd. a. dorsal view. 5. lateral view of ocular tubercle, from left side. c. palp. d. oviger. e. 2nd tibia, tarsus and propodus of leg. 88 . ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. Resembles weber: in having the tarsi long proportionately to the propodi (not very short as in australis), but differing from both species in having spiniform processes on the 2nd-4th crurigers. Gen. Rhopalorhynchus Wood-Mason 1873. Wood-Mason, 7. Asiat. Soc. Beng., xlii, p. 171. 1908. Loman, loc. cit., p. 23. Octopodous. Body segmented, slender, rod-like. Crurigers short, stout, widely separated. Ocular tubercle short, conical, eyes present. Proboscis elongate, fusiform. Abdomen very short. Palps slender, 9-jointed. Ovigers slender, 10-jointed, with small but stout unguiform apical claw. Legs slender, without auxiliary claws. Rhopalorhynchus kréyeri Wood-Mason igus? 1873. Wood-Mason, loc. cit., p. 171, pl. 13. 1923. Calman, Rec. Ind. Mus., xxv, p. 268, fig. 1 (references and synonymy; the date of publication of Haswell’s tenuissima is 1885). 1953. Stock, Temminckia, ix, pp. 279, 280, fig. 1b (chart). Fig. 2. Rhopalorhynchus kréyeri Wood-Mason. Dorsal view, with lateral view of ocular tubercle from right side. Proboscis with a sharp denticle medio-dorsally at about the middle of its length. Ocular tubercle rather abruptly constricted above the eyes (cf. Carpenter, 1907, fig. 28). Oviger inner margins of 7th-1oth joints with numerous feebly biserrate spines (in several rows), roth joint in addition with a stout spine distally opposing the small unguiform claw (as in Calman’s figure; apparently overlooked by Loman, 1908, pl. 15, fig. 216). Legs with coxal joints subequal, femur distally clavate in 9, in ¢ more slender and less conspicuously clavate, 1st tibia slightly shorter than femur, 2nd tibia slightly shorter than Ist, tarsus about two-thirds propodus, claw strong, shorter than tarsus (cf. Loman, pl. 15, fig. 220). SOUTH AFRICAN PYCNOGONIDA 89 Proboscis 7:5, femur g, Ist tibia 8-5, and tibia 7-5 mm. Localities: off Hood Point, near East London, 42 fath.; off Port Shepstone, Natal, 24 fath.; and off Cape Natal (Durban), 85 fath. 399 (S. Afr. Mus.). Distribution. Maldive Is., Ceylon, Indian Seas, Andaman Is., East Indies, Torres Straits, Queensland. Remarks. ‘The position of the denticle on the proboscis corresponds with Loman’s fig. 215, not with Calman’s fig. 1 a, 6. Calman refers to the variation in the claw of the legs, but does not suggest that slender and strong claws are characteristic of the g and Q respectively. Fam. NyYMPHONIDAE 1913. Schimkewitsch, Zool. Anz., xli, p. 607. 1913. Bouvier, 2me Exp. Antarct. Franc., p. 65. 1932. Gordon, ‘Discovery’ Rep, vi, p. 24. Legs 8 or 10. Body segmented. Proboscis immovable. Chelifers well developed, usually conspicuously chelate. Palps 5-jointed (7-jointed in Paranymphon Caullery). Ovigers in both sexes, 10-jointed, with apical claw, last 4 joints usually with denticulate spines (8-jointed, without denticulate spines in Paranymphon). Legs with or without auxiliary claws. Genital apertures on all legs in 9, on last 3 in g. Gen. Nymphon Fabr. 1794. Fabricius, Ent. Syst., 1v, p. 417. 1888. Sars, Arch. Naturv. Christ., xii, p. 352 (Chaetonymphon). 1908. Loman, Szboga Exp., monogr., xl, p. 37. 1915. Calman, Brit. Antarct. (Terra Nova) Exp., zool. III, p. 28. 1923. Loman, Swed. Antarct. Exp., 1, 2, p. 10. 1932. Gordon, loc. cit., p. 26 (revision of Antarctic species). 1937. Giltay, Mem. Mus. Roy. Hist. Nat. Belge. (2) fasc. 9, p. 83 (key to 7 species on west coast of Africa, Morocco to Cape of Good Hope). 1944. Gordon, B.A.N.Z. Antarct. Res. Exp., B, v, pt. 1, p. 17 (key to Antarctic and Subantarctic species). 1951. Stock, Mem. Inst. Roy. Sci. Nat. Belge. (2) fasc. 43, p. 5. 1953. id., Temminckia, ix, p. 282. Octopodous. Proboscis subcylindrical. Ocular tubercle at base of neck, never in front of bases of ovigers, eyes usually present. Finger and thumb of chelifer with more or less slender spiniform denticles. Oviger with apical claw, last 4 joints with denticulate spines. Legs with or without auxiliary claws. Remarks. A large genus, the Antarctic and Subantarctic species of which have been usefully analysed by Gordon. gO ANNALS OF THE SOUTH AFRICAN MUSEUM Key to the South African Species I. Auxiliary claws vestigial or absent. A. Auxiliary claws vestigial. Main claw slender. Oviger modified; denticu- late spines on last 4 joints 23-37. B. Auxiliary claws absent. 1. Finger and thumb of chelifer apically crossing, each with 50-65 spines. Palp and joint longest. a. Neck short, crurigers narrowly separated. i. Crurigers less than half their own width apart. 3 oviger modified, spines 48-50. ii. Crurigers about their own width apart. g oviger not modified. a. Spines on oviger 54-63(66). B. Spines on oviger 44. b. Neck long, crurigers widely separated. Spines on oviger 25. 2. Finger and thumb of chelifer not crossing, with very numerous and fine spines (300 +). and, 3rd and 4th joints of palp subequal. Spines on oviger 46. II. Auxiliary claws well developed. A. ‘Tarsus at least half length of propodus, sometimes a little longer than propodus. 1. Finger and thumb of chelifer apically crossing; former with many more (twice as many) spines than latter. a. Legs feebly setose. i. Neck short. Tarsus shorter than propodus. Oviger with 36-38 denticulate spines. ii. Neck long. Tarsus longer than propodus. Oviger with 40-42 denticulate spines b. Legs strongly setose. Neck short. Tarsus shorter than propodus. Oviger with 23-26 spines. 2. Finger and thumb not crossing, an approximately equal number of spines on both. a. Finger and thumb each with less than 25 spines. Oviger with 34-36 spines. 6. Finger and thumb each with 45 or more spines. Oviger with 39 spines. B. ‘Tarsus only one-third length of propodus. Finger and thumb of chelifer with an equal and moderate number of long spines, and long curved setae on the setose pad. Oviger with 51-55 spines. Nymphon australe Hodgson australe capense comes affinis natalense microctenatus signatum distensum pilosum crenatiunguis angolense setimanus 1902. Hodgson, br. Mus. Rep. ‘Southern Cross’ Crust., p. 257, pl. 40. 1902. Mobius, D. Tiefsee Exp. ii, p. 181, pl. 26, figs. 1-6. (Chaetonymphon altioculum.) 1913. Bouvier, loc. cit., p. 73, figs. 25-31 (stylops). 1915. Calman, loc. cit., p. 36 (synonymy). 1920) )) Elymmn, loc. jet. .p.) 16: 1932. Gordon, loc. cit., p. 59, figs. 25 d, 26 b. 1938. id., Austral. Antarct. Exp: @) AUG. p.i13. 1944. id., loc. cit., p. 24 and var. caecum, p. 25, fig. 5 a-d. Body compact, neck short, crurigers separated by a distance varying from a quarter to equality with their own width. Proboscis subcylindrical. Ocular SOUTH AFRICAN PYCNOGONIDA QI tubercle high, conical, eyes subterminal. Abdomen reaching to end of 1st coxa of last leg, clavate or pyriform. Sparsely setose on cephalic lobe and crurigers. Chelifer scape longer than proboscis, hand subequal to scape, finger and thumb longer than palm, nearly straight, apically curved and crossing, each with 38-46 closely set spines. Palp with and joint longest, 4th and 5th joints subequal. Oviger in § modified, 5th joint longest, distal half strongly expanded, 6th joint also swollen, last 4 joints with 23-37 denticulate spines. Legs rather short, setose, 2nd coxa slightly longer than 1st + 3rd, femur shorter than either tibia, Ist tibia longer than and, tarsus longer than propodus, claw slender, half length of propodus, auxiliary claws vestigial. Body including abdomen 7-8 mm. Locality: off Cape Point, 1,190 fath. (Flynn). The locality of another specimen, from a depth of only 46 fathoms, is regarded by Flynn as doubtful. Distribution. Circumpolar (antarctic), nearest locality to South Africa: Bouvet Island. Remarks. The s.s. Pieter Faure obtained no specimens referable to this species. Nymphon capense Hodgson ? 1879. Bohm, MB. Ak. Wiss. Berlin, p. 173 (phasmatodes). 1908. Hodgson, Tr. Roy. Soc. Edinb., xlvi, p. 169, pl. 1, figs. 2, 2a. 1928. Flynn, loc. cit., p. 4 (remarks on synonymy). fez Gordon: Ann. Mas. Nat. Hist.) (10), 1x,\p: 117, figs. 11, 12 (redes- cription). ne32,) id., loc. Cli., pp. 27-32, 34 (in key), 62. Body stout, neck short, base of oviger large, occupying most of the space between Ist cruriger and anterior cephalic expansion. Crurigers distinctly but narrowly separated. Ocular tubercle stout, low, rounded, eyes distinct. Proboscis bottle-shaped, apically rounded. Abdomen slightly longer than last crurigers, oblique (at about 45°). Chelifer scape a little longer than proboscis, hand longer than scape, finger and thumb apically curved and crossing, with 60-65 closely set spines, in 2 or 3 alternating sizes. Palp with 2nd joint longest, 3rd, 4th and 5th progressively shorter. Oviger ¢ with 3rd joint inflated, 4th and 5th subequal, rather short and stout, 5th apically dilated, 6th deeply excavate on one side with long setae protruding from the cavity, last 4 joints with 48-50 denticulate spines (Hodgson: Byea2, 00,024, Gordon, (12,,12, 11, 19), claw: with 13-15, spinules. In; 2 normal (see: comes). Legs slender, feebly setose, 2nd coxa equal to 1st + 3rd, femur shorter than Ist tibia, and tibia longer than 1st (7, 8 and 10-7), tarsus and propodus equal or latter a little longer, claw two-thirds to three-quarters the length of propodus, no auxiliary claws. Body (¢ 2) 2:7-2°8, proboscis 1-3-1-6, 3rd leg femur 2-6-3, Ist tibia 3-4-4, 2nd tibia 4-4-6 mm. (Gordon). Q2 ANNALS OF THE SOUTH AFRICAN MUSEUM Locality: off Dassen Island, 35 fath. 99 and ovig. 3d (Hodesos South- west of Cape of Good Hope, 50 fath. (Béhm). Remarks. Flynn thought that Hodgson’s species was synonymous with Boéhm’s. Béhm’s specimen was a @ with possibly abnormal ovigers. Hodgson’s description was based on ovigerous gj as well as 99, and the type material has been re-examined and figured by Gordon, who does not, however, discuss the possible synonymy. Giltay (1937) and Stock (1951) both accept the synonymy. The species is a small one. Nymphon comes Flynn Figs. 3, 5¢ 1929. Flynn, loc. cit., p. 14, figs. 4-6. Body stout, neck well marked, of medium length, cephalic segment subequal to rest of body (incl. abdomen). Crurigers separated by distances not quite equal to their own width, their length subequal to median width of segments. Ocular tubercle shortly conical, in juvenile (body 5 mm.) with 2 apical points (as seen from in front) which gradually disappear leaving a blunt point in adult, eyes distinct. Proboscis stout, expanded in middle, apically rounded. Abdomen about as long as last crurigers, slightly clavate, horizontal. A few minute setules on hind margins of segments, and on crurigers. Chelifer scape slightly longer than proboscis, hand about equal to scape, finger and thumb longer than palm, apices curved and crossing, inner margins with more or less regularly alternating long and short spine-teeth, about 50 in number. Palp rather stout, 2nd joint longest, 3rd, 4th and 5th progres- sively shorter. Oviger in ¢ 4th joint slightly longer than 5th, slightly curved, 5th slightly curved at base, apically expanded on ventral surface with a semi- circlet of strong spines, last 4 joints with 54-63 (66) denticulate spines resp. 17-21, 14-16, 11-12 (Flynn: 15), and 12-14. In 9 4th joint slightly longer than 5th, both nearly straight. Legs moderately slender, ist tibia a little longer than femur, 2nd tibia one and a half to one and a third times as long as femur (Flynn says 2nd tibia is twice the femur and one and a half times Ist tibia, his measurements are nearly in agreement, but his figures are not), tarsus sub- equal to, or slightly shorter than propodus, claw half to two-thirds as long as propodus, no auxiliary claws. Femur and Ist tibia in 3 not appreciably swollen distally. Femoral cement glands in ¢ (ventral) minute, very numerous, at least 70. Feebly setulose, more numerous setae on Ist and and tibiae. Body @ incl. abdomen 12, abdomen 1°5, proboscis 6, 2nd leg 67 mm. (femur 15, Ist tibia 17, 2nd tibia 20 mm.). Body dg incl. abdomen 10-5, abdomen 1°5, proboscis 5, 2nd leg 59 mm. (femur 12, Ist tibia 14:5, 2nd tibia 18 mm.). Localities: False Bay, 22 fath. 1 g (Flynn); off Cape Infanta, 46 fath. 4 99 2 non-ovig $d; off Cape Seal, 80 fath. 1 juv.; off Cove Rock (East London), 43 fath. 1 non-ovig ¢ (S. Afr. Mus.). SOUTH AFRICAN PYCNOGONIDA 93 Remarks. In addition to the discrepancy in the leg shown in fig. 6, Flynn figures the oviger as g-jointed (fig. 5). Although reaching a larger size than Flynn’s § specimen (the East London g is approximately the same size as the type), these specimens seem to belong to comes. There is one point of difference: the length of the claw of the legs. a Cc d b Fig. 3. Nymphon comes Flynn. a. dorsal view. 5b. front views of ocular tubercle of juvenile, 5 mm. in length (top), intermediate stage, and adult (bottom). c¢. oviger g 5th joint, ventral view above, external view below. d. oviger g distal joints with spine and claw further enlarged. In the type it is about one-third the propodus, but in all the present specimens, including the juvenile, it is at least a half, usually two-thirds the propodus. This is a much larger species than capense (phasmatodes), even the type being nearly three times as big; but it is difficult to find (from Hodgson’s description) morphological differences. ‘The denticulate spines on the last 4 joints of the oviger are given by Hodgson as resp. 15, 12, 11, 12 in number; and he describes the teeth on the finger and thumb of the chelifer as being of three sizes. 94 ANNALS OF THE SOUTH AFRICAN MUSEUM Nymphon affinis Stock 1951. Stock, loc. cit., p. 5, figs. 1-6. This species, described from a single 2 5-8 mm. in length, is very close to comes, and Stock himself suggests that it may be the @ of the latter species of which only the ¢ has hitherto been known. Stock relies on the following differences in establishing a separate species. Finger and thumb of chelifer with alternating larger and smaller spines; Stock considers that Flynn would certainly have mentioned this feature if it had been present in comes. The 5th joint of palp is relatively longer: five- sevenths the length of 4th, instead of half. The different number of spines on the oviger: 15, 11, 10, 8 = 44, as against 66. Little reliance can be placed on the first difference, as Flynn’s descriptions are not always very detailed. Variation would probably be found in the relative lengths of the joints of the palp and legs if a large number of specimens of different sizes were available. For the present, affinis may stand, characterized by the 44 spines on the oviger. The 5 mm. juvenile specimen of comes has 54 spines on the oviger. Locality: WLiideritzbucht, 8 fath. (Stock). Nymphon natalense Flynn 1926) Blymn; loc. jcit., p. 11s fig.9: Body slender, neck long and narrow, cephalic segment a little longer than rest of body (incl. abdomen). Crurigers separated by distances about twice their own width, their length considerably greater than median width of segments. Ocular tubercle shortly conical, eyes distinct. Proboscis rather stout, expanded in middle, apically rounded. Abdomen shorter than last crurigers. Chelifer scape slightly shorter than proboscis, hand longer than scape by nearly one-third (details of finger and thumb not given). Palp 2nd joint longest, 3rd, 4th and 5th progressively decreasing. Oviger 5th joint longest, apically expanded in g, denticulate spines on last 4 joints resp. 9, 6, 5, 5. Legs long and slender, 2nd coxa very long, a little more than twice as long as ist and about half the length of ist tibia, 1st and 2nd tibiae subequal, tarsus shorter than propodus, claw about three-quarters the length of propodus, no auxiliary claws. Feebly setose. Body 3:6, proboscis 1-2, leg 18-7 mm. (femur 3:2, rst and 2nd tibiae 4 mm.). Locality: off Port Natal (Durban), pelagic, 1 damaged 3 (Flynn). Remarks. ‘There are several discrepancies between the description of the legs, the measurements, and the figure. I have seen no specimens. - SOUTH AFRICAN PYCNOGONIDA 95 Nymphon microctenatus Brnrd. Fig. 4 1946. Barnard, Ann. Mag. Nat. Hist., (xi), 13, p. 60. 9—Body moderately stout, neck moderate, base of oviger separated from ist cruriger and occupying the whole of the rest of the neck behind the anterior cephalic expansion. Crurigers separated by about their own width, or rather less, their length subequal to median width of segments. Proboscis very stout, H ee reomrars Cc Fig. 4. Nymphon microctenatus Brnrd. a. dorsal view of cephalic segment with proboscis and basal joint of chelifer. 5. lateral view of ocular tubercle. c. 4th and 5th joints of palp. d. inner view of chela (spines much more numerous than shown). nearly evenly cylindrical, apex truncate. Ocular tubercle low and rounded, eyes distinct. Abdomen not extending beyond last crurigers, clavate, hori- zontal. Glabrous. Chelifers widely separated at base, scape slightly longer than proboscis, hand about as long as scape, palm short, finger and thumb both slightly curved downwards, their apices following the same curve, i.e. not crossing, margins with an enormous number (estimated at 300-400) of slender closely set spinules 96 ANNALS OF THE SOUTH AFRICAN MUSEUM like a very fine comb. Palp long and very slender, almost glabrous, 2nd, 3rd and 4th joints subequal, 5th a little more than half 4th joint, gently curved. Oviger 5th joint longer than 4th (subequal to 3rd + 4th), 4th with slight swelling proximally on inner side, 5th straight, 6th with a rather prominent spine on inner apex, last 4 joints with 46 denticulate spines (14, 11, 10, 11) on inner margin, spines with 6 (4 major) denticulations, claw with 18-19 spines. Legs, coxae and femur moderately stout, tibiae distinctly more slender, 2nd coxa a little longer than 1st + 3rd, femur a little shorter than Ist tibia, and tibia considerably longer than Ist, tarsus distinctly but not much shorter than propodus, claw about half the length of propodus, no auxiliary claws. Coxae and femur glabrous, Ist tibia very sparsely setose, 2nd tibia sparsely and finely setulose. Body incl. abdomen 7-5, abdomen 1, proboscis 3, 2nd leg coxa 3:5, femur 10, Ist tibia 11-5, 2nd tibia 20 mm. Locality: off Buffalo River (East London), 195 fath. 1 9 (S. Afr. Mus.). Remarks. ‘The specimen is broken in pieces, but all essential parts are present. Nymphon signatum Mobius Fig. 54 1902. Mobius, D. Tiefsee Exp., ii, p. 178, pl. 24, figs. 1-7. 1932. Gordon, Ann. Mag. Nat. Hist. (10) ix, p. 104, fig. 5 (redescribed). 1932. id., loc. cit., pp. 28-9 (in Table iii), 35 (in key). Body moderately stout, neck stout and moderately long, base of oviger occupying two-thirds of its length, cephalic segment widely expanded. Crurigers separated by about half their own width, their length less than median width of segments. Proboscis stout. Ocular tubercle higher than wide, apex bluntly pointed or feebly bifid. Abdomen extending slightly beyond last crurigers, horizontal or slightly oblique. Glabrous. Chelifer with a few setae dorsally and apically, chela at least as long as scape, finger and thumb apically curved and crossing, former with 60-64 spines, latter with about 35 longer spines, the two series of spines approximately the same length. Palp 2nd joint longest, 3rd-5th progressively shorter, 4th ++ 5th subequal to 3rd. Oviger in ¢ 4th joint slightly shorter than 5th, both curved, 5th with apical lobe bearing a semicirclet of spines (cf. comes); in 9 Ath joint longest, 4th and 5th joints straight; last 4 joints with 36-38 (11-13, g-11, 7-9 and 8-10) denticulate spines, each with 5 major denticulations, claw two-thirds roth joint, with 10 spinules. Legs moderately long and slender, femur and Ist tibia distally swollen in 3, femur shorter than Ist tibia, 2nd tibia longer than Ist tibia, with spines on its apex, tarsus nearly as long as propodus, latter with numerous spines on lower margin, claw nearly half length of propodus, auxiliary claws well developed. Feebly setose, especially in 9, scattered setae on femur, more numerous on Ist s- ® SOUTH AFRICAN PYCNOGONIDA 97 and and tibiae. Femoral cement glands in ¢ (ventral) 17-20, the distal ones much smaller than the proximal ones. Body incl. abdomen 8:5, proboscis 4:5, abdomen 1-5, 2nd leg femur 9g, ist tibia 11, 2nd tibia 15 mm. an py a ON aag iy Le. as stn MEER Ly, hy Lae: enn PH Hp Pippen rm Desh se iulaniauusaner es g Fig. 5. Chela of: a. Nymphon signatum Mobius, juv. 5. Nymphon distensum Mobius. c. Nymphon comes Flynn. Localities: Agulhas Bank, 100 metres (Mobius); False Bay, 19-27 fath. I ovig. 6; 10 miles off Cape St. Blaize, 1 9; off Gt. Fish Point, 49 fath. 3 ovig. gg, I non-ovig 3, 1 immature J; off Glendower Beacon (Port Alfred), 66 fath. 1 non-ovig. ¢ (S. Afr. Mus.). Remarks. With the holotypes of this species and distensum before her, Gordon pointed out certain small differences, but remarked that intermediate forms 98 ANNALS OF THE SOUTH AFRICAN MUSEUM might be found to unite them. The small amount of material in my hands does not confirm this, although two of the differential characters mentioned by Gordon must be discarded. Spines are present in all cases on the apex of and tibia; evidently in the holotype they had been rubbed off. Differences in the number of spines on ventral margin of propodus are not reliable. The two species are distinguished by their build, length of neck, and relative lengths of tarsus and propodus (tarsus < propodus in szgnatum, tarsus > propodus in distensum). The 3 from False Bay is particularly heavily built, spines on finger and thumb of chelifer resp. 50 and 30, denticulate spines on oviger (only one is complete) 45. A small specimen, body incl. abdomen 2°75 mm., from off Umhlangakulu River (Natal), 50 fath., may belong to this species. The neck and distance separating the crurigers as in signatum, but abdomen oblique, ocular tubercle rather high and distinctly bifid apically. Second coxa a little longer than ist + 3rd, distinctly swollen apically. Oviger with 32 denticulate spines (10, 8, 6, 8). Chelifer hand with finger strongly curved, with fine apical point curving in the opposite direction, with 38 spines, thumb with 28 spines. Tarsus shorter than propodus, claw scarcely one-third length of propodus, auxiliary claws two-thirds main claw. Nymphon distensum Mobius Fig. 50 1902. Mobius, loc. cit., p. 179, pl. 25, figs. 1-6. 1928. Flynn, loc. cit., p. 10. 1932. Gordon, loc. cit., p. 105, fig. 6 (redescribed). 1932. id., loc. cit., pp. 28-9 (in Table iii), 35 (in key). Body rather slender, neck slender and rather long, base of oviger occupying less than half its length, cephalic segment not widely expanded. Crurigers separated by distances less than their own width (except the last 2 pairs), their length subequal to median width of segments. Proboscis stout. Ocular tubercle rather high, conical, more or less constricted above eyes, apex more or less distinctly bifid. Abdomen very slightly longer than last cruriger, clavate, horizontal. Glabrous. Chelifer scape subequal to proboscis, hand as long as scape, finger and thumb apically curved and crossing, former with 60-70 (holotype 9, fide Gordon: 60-64), latter with 30-33 longer spines, the two series of spines of approximately equal length. Palp 2nd joint longest, 3rd-5th progressively shorter, 4th + 5th subequal to 3rd. Oviger in ¢ 4th and 5th joints subequal, both curved, 5th with apical lobe bearing semicirclet of strong spines (cf. comes); in @ 4th and 5th joints subequal, both straight; last 4 joints with 40-42 (13-16, 9-11, 8-9 and g-10) denticulate spines, each with 5 major denti- culations, claw two-thirds to three-quarters length of 1oth joint, with about 15 spines. SOUTH AFRICAN PYCNOGONIDA 99 Legs long and slender, femur and 1st tibia distally swollen in 3, 2nd coxa equal to or a little longer than 1st + 3rd, femur shorter than 1st tibia, 2nd tibia much longer than (one and one-third) 1st tibia, apex with spines, tarsus a little longer than propodus (Gordon: tarsus three-quarter length of propodus in 4th leg, but figure of 1st leg shows tarsus a trifle longer than propodus), claw about half length of propodus, auxiliary claws one-third, or a little more, main claw. Feebly setose, especially in 9, a few scattered setae on 1st and 2nd A tibiae. Femoral cement glands in 3 (ventral) 12-15. Fig. 6. Nymphon pilosum Mobius. a. ocular tubercle, front (above) and lateral (below) views. b. femur g. ¢. oviger § 5th joint. d. spines from distal joints of oviger, normal and abnormal (or worn). Body incl. abdomen ovig. g 11, 2 13:5, abdomen J 1°5, 2 2, proboscis 3 5, 2 6, and leg 3 64, 9 92 mm. (femur J 11°5, 9 20, Ist tibia J 15, 9 23, 2nd tibia 3 25°5, 2 37°5 mm.). Localities: Agulhas Bank, 154 metres, 1 2 (Mobius); off Gneka River, 43 fath. 1 2 (Flynn); off Buffalo River (East London), 195 fath. 3 gd (1 ovig.), 12; off Cape Morgan, 87 fath. 1 2; off Cape Seal, 37 miles, 80 fath. 1 juv.; Cape Point N. 71° E. 18 miles, 230 fath. 1 9 (S. Afr. Mus.). Remarks. Holotype @ redescribed by Gordon. The few specimens I have examined all agree with distensum as described by Gordon, and for the present this species may be retained as distinct from szgnatum. 100 ANNALS OF THE SOUTH AFRICAN MUSEUM The largest 9, whose measurements are given above, came from the Cape Point locality. Flynn’s specimen appears to be aberrant in having the tarsus shorter than propodus (cf. szgnatum) as in Mobius’s figure, which, as Flynn says, cannot be relied upon for strict accuracy. N. arabicum Calman 1938 seems to be very similar to distensum. Nymphon pilosum Mobius Fig. 6 1902. Mobius, loc. cit., p. 179, pl. 24, figs. 8-12. 1928. Flynn, loc. cit., p. 8, figs. 1, 2 (bzpunctatum). Body moderately stout, neck short and stout, base of oviger occupying whole space between Ist cruriger and cephalic expansion. Crurigers separated by about their own width or rather less, their length a little less than median width of segments. Proboscis moderately stout, usually widest in proximal half and slightly tapering to a rounded apex. Ocular tubercle rather high, columnar, slightly expanded above owing to the large eyes, surmounted by a transverse ridge which usually shows two little points. Abdomen distinctly longer than last crurigers, cylindrical or slightly clavate, horizontal. A few setae on hind margins of cephalic and following two segments. Chelifer sparsely setose, subequal to or slightly longer than proboscis, hand subequal to scape, finger and thumb apically curved and crossing, former with about 40 spines, latter with about 26 larger spines, the two series of spines of approximately equal length. Palp 2nd joint longest, 4th and 5th sub- equal or the latter slightly the shorter, 4th + 5th subequal to 3rd. Oviger in ¢ 4th joint slightly curved, with slight swelling proximally on inner side, 5th slightly longer than 4th, curved, apically swollen, setose, especially on apex ventrally; in 9 4th and 5th joints subequal, straight; last 4 joints with 23-26 (g, 5-6, 4-5 and 5-6) (one specimen with 10, 6, 5, 6 = 27), stout and strongly denticulate spines, with 3 major denticulations, claw with 6-7 spines. Legs moderately stout, femora and 1st tibiae not distally swollen in d, 2nd coxa subequal to, or in the juv., slightly longer than, 1st + 3rd coxae, femur shorter than Ist tibia, which is shorter than 2nd, tarsus distinctly though not much shorter than propodus, claw not quite half length of propodus, auxiliary claws about half length of main claw. A few setae on the coxae, femur with short and longer outstanding setae, tibiae with numerous long outstanding setae. Femoral cement glands in ¢ large, about 5 or 6 (4-7) on low knobs or ridges on ventral surface. Body incl. abdomen 7, abdomen 1-75, proboscis 3, 2nd leg 38 mm. (femur 6, 1st tibia 7, 2nd tibia 9-5 mm.). The non-ovigerous ¢¢ are a little smaller than the largest 9. Localities: Agulhas Bank, 154 metres 1 9 (Mobius); off Port Natal (Durban), tow-net, 5 gd; 4 9° (Flynn); off Cape St. Blaize, 85-90 fath., 4 non-ovig. 3d, SOUTH AFRICAN PYCNOGONIDA IOI 6 99; off Cape Seal, 80 fath. 1 9; off Cape St. Francis, 75 fath. 1 non-ovig. 3; Someon oy 2 20.1 1) 200math, 12) (S. Ate, Mus.) 20% )567 S..919 11 1, 339 metres (Fisheries Survey, 1948). Remarks. There would seem to be little doubt that bipunctatum is synonymous; at least it is strange that Flynn did not compare his specimens with pilosum; possibly being taken in a tow-net and therefore presumed to be pelagic uncon- sciously created an impression of specific distinctness. The ocular tubercle is characteristic. he stout spines on the last joints of the oviger are also very striking. EAA Fig. 7. Nymphon crenatiunguis Brnrd. a. chela with denticles further enlarged. 6. apex of claw of oviger 69. c. chela of a g 2°25 mm. in length. Owing to their pilosity the legs are often much obscured by foreign matter; some of the specimens have examples of the stalked barnacle Scalpellum agulhense Brnrd. attached to the legs; and one specimen has one of the barnacles standing upright immediately behind the ocular tubercle. Nymphon crenatiunguis Brnrd. Bic. 7 1946. Barnard, loc. cit., p. 60. Body moderately stout, neck short, base of oviger contiguous with first cruriger and extending nearly to beginning of cephalic expansion. Crurigers separated by less than their own width, their length subequal to (a trifle 102 . ANNALS OF THE SOUTH AFRICAN MUSEUM shorter than) median width of segments. Proboscis nearly cylindrical, apex rounded. Ocular tubercle rather low, conical, in front view with 2 little points surmounting the eyes. Abdomen slightly longer than last crurigers, clavate, oblique (at about 45°). Glabrous. Chelifer scape subequal to proboscis, hand shorter than scape, palm oblong, widening slightly distally, finger and thumb relatively short and stout, former subequal to palm, apices not crossing, finger with 20(9)-22(¢) closely set spiniform denticles, thumb with 18(Q)-17(¢) more slender and slightly more widely spaced denticles. Palp and joint longest, 4th + 5th joints a little longer than grd, 5th longer than 4th. Oviger in ¢ 5th joint longer than 4th, gently curved, with apical lobe bearing strong recurved spines; in 2 4th and 5th subequal; last 4 joints with 34-36 denticulate spines (Jj 13, 7, 7,7; 214, 9, 6, 6 and 13, 9, 7, 7), Spines stout with 3 major denticulations. Claw with about 15 denticles, which distally become successively shorter and form a rounded and minutely crenulate (apparent) apex, the true apex being acute and minute. This form is found on both ovigers of the g and of the 9, though in one oviger of the latter the distal denticles are sharper and more prominent than in the figure (the apex thus appears more serrate than crenulate). Legs moderately stout, 2nd coxa subequal to ist + grd, femur shorter than ist tibia, 2nd tibia longer than Ist, tarsus and propodus subequal, claws about one-third propodus, auxiliary claws strong, at least half main claw. Glabrous except the 2nd tibia which is sparsely and minutely setulose. Femoral cement glands in ¢ 4-5 low knobs or tubercles on ventral surface. Body incl. abdomen ¢ 4, 2 3, abdomen 3:75, 9-5, proboscis 3 1°75, 9 1°5, leg femur ¢ 4°5, 2 4, Ist tibia ¢ 5, 9 5, 2nd tibia ¢ 7-5, 9 6-5 mm. Locality: off East London, 85 fath. 1 non-ovig, g, 1 9 (S. Afr. Mus.). Remarks. Distinguished by the hand of chelifer, which resembles in general that of grossipes Hoek 1881 and multtdens Gordon 1932. As in these two species the 5th joint of the palp is longer than 4th; but the tarsus and propodus of the legs are subequal, and the spines on the last joints of oviger are different (in shape from those of grossipes, and in number from those of multidens). The apex of claw of oviger seems to be unusual if not peculiar. One non-ovigerous ¢, body incl. abdomen 2:25 mm., from off Durnford Point, Zululand, go fath., agrees with the above described specimens in most characters, including the peculiar claw of the oviger, but has the hand of the chelifer sightly different. The finger is subequal to palm, but both it and the thumb are more slender, the former with 15, the latter with 14 denticles (fig. 7c). The femur has only 3, but very distinct, cement gland knobs. The 5th joint of oviger has the apical lobe with recurved spines, and the denticulate spines on last 4 joints number 12, 9, 7, 6. Nymphon angolense Gordon ? 1923. Loman, Goteb. K. Vet. Samh. Handl., xxvi 6 (Medd. Goteb. Mus. Zool. Avd. no. 22), p. 5. (N. gracillimum, non Calman.) SOUTH AFRICAN PYCNOGONIDA 103 19p220 Gordon, loc. ieit., (ps 77, figs. 30, 37. no374)/) Giltay, loc, cif., p:. 88. LOG.) tock) loc. cit.,\p.) 7. Neck long and slender, base of oviger contiguous with rst cruriger. Crurigers separated by more than their own width, their length greater than median width of segments. Proboscis cylindrical, apex rounded. Ocular tubercle low, rounded. Abdomen slightly longer than last crurigers, elevated at about 45°. Chelifer scape subequal to proboscis, hand slightly longer than scape, palm slender, finger and thumb slightly longer than palm, slender, apices crossed, about 55 spines on finger, 45 on thumb. Palp 2nd joint longest, 3rd three-fifths end, 4th half 3rd and 5th slightly longer than 4th. Oviger in ¢ slender, 5th joint longer than 4th, straight, with a few hook-like spines on apex; in 2 4th and 5th joints relatively short; last 4 joints with 39 denticulate spines (13, 10, 8, 8) (Stock: ¢ 13, 9, 7, 7). Claw with 14 spinules. Legs slender, 2nd coxa slightly longer than 1st + grd, femur shorter than Ist tibia, 2nd tibia longer than ist, propodus slightly longer than tarsus, auxiliary claws at least two-thirds length of main claw. Body incl. abdomen 7, proboscis 1-8-1-9, femur ¢ 6-8, 2 7, ist tibia ¢ 9-2, 9 8-4, and tibia 12 mm. Localities: off Elephant Bay, Angola (approx. 13° 10’ S. 12° 45’ E.), 73-91 metres (Gordon); Tiger Bay, and Elephant Bay, 100-110 metres (Stock) ; off Port Alexander, 60 fath. (Loman). Remarks. Gordon set out several differences between angolense and the holotype of gracillamum, but not the number of spines on the oviger. This is not included in Calman’s original description but is given, on Calman’s authority, Dylon (loc) cit. p. 5, tootnote)) as'39 (12) 846; 7) and) 30 (12,7, 5,16). Momams specimens showed 31 (11, 6, 6, 8). and 28 (10, 7, 5, 6). All these numbers are less than those in angolense. Gordon made no comment on Loman’s record, but it seems more probable that the Port Alexander specimen (‘verstitimmelten’, Loman) should be identified with angolense than with an Antarctic species, in spite of the difference in number of the spines on the oviger. Port Alexander lies between the two localities recorded by Stock. Nymphon setimanus Brnrd. Fig. 8 1G46.,, Barnard: loc: (cit... OL. Body moderately stout, neck long (cf. distensum), base of oviger contiguous with first cruriger and occupying one-third to nearly one-half of neck. Crurigers separated by rather less than their own width, their length subequal to median width of segments. Proboscis stout, nearly cylindrical, apex rounded-truncate. Ocular tubercle low, scarcely if at all higher than basal width, in lateral view 104. ANNALS OF THE SOUTH AFRICAN MUSEUM conical, in front view apically truncate, with two minute points above the eyes. Abdomen extending a little beyond last crurigers, somewhat elongate, oblique (at about 45°). Glabrous. Chelifer scape a little shorter than (or scarcely as long as) proboscis, hand subequal to scape, finger and thumb much longer than palm, slender, curved, apically crossing, each with 13-15 equally long spines, setose pad well developed with numerous very long, almost spiniform setae on inner surface. Palp and and 3rd joints subequal, or the latter a trifle the longer, 5th shorter than 4th, 4th + 5th longer than 3rd. Oviger 4th and 5th joints subequal, last 4 joints eae ~ > 7 Fig. 8. Nymphon setimanus Brnrd. Inner view of chela 9. with 51-55 denticulate spines (15, 13, 12, 11 or 18, 12, 12, II or IQ, 12, IT, 13), each with about 6 major denticulations, claw with about 13 denticles, the distal ones separated by crenulations, the apical point scarcely extending beyond last denticle (cf. crenatiunguis). Legs moderately stout, 2nd coxa equal to or a trifle longer than 1st + 3rd coxae, femur only a little shorter than ist tibia, and tibia not much longer than ist, tarsus short, about one-third propodus, claw short, subequal to tarsus, auxiliary claws two-thirds main claw. Almost glabrous, even on 2nd tibia only a few minute and scattered setules. Body incl. abdomen 6, proboscis 2, abdomen 1-25, 2nd leg 2nd coxa 2:5, Ist + 3rd coxae together 5, femur 6, Ist tibia 6-5, 2nd tibia 7 mm. Pinkish, the palps and ovigers and a band around apex of femur pale whitish (Pt. Elizabeth Mus. and U.C.T. specimens). Localities: Bird Island Passage (Algoa Bay), 10-16 fath. 1 9; off East London, 32 fath. 1 9; off Port Shepstone (Natal), 24 fath. 1 9 (S. Afr. Mus.); shelly beach, Cape Recife (Port Elizabeth Mus.); main channel, Knysna harbour; Bushmans River mouth (Algoa Bay) (Univ. Cape Town Ecol. Surv. 1947 and 1950). SOUTH AFRICAN PYCNOGONIDA 105 Remarks. Distinguished from all the other South African species by the short tarsus. Neck shorter than in andamanense Calman 1923. ‘The hand of chelifer bears some resemblance to that of maculatum Carp. 1910. Fam. PALLENIDAE 1908. Loman, Szboga Exp. monogr., xl, p. 40, and conspectus facing p. 19 (subfam. Palleninae). 1909. Schimkewitsch, ool. Anz., xxxiv, p. 6. mons) ic, ibid., xli, p. 610. 1944. Gordon, B.A.N.Z. Antarct. Res. Exp., B, v, pt. 1, p. 36. 1947. Hedgpeth, Smiths. Misc. Coll., cvi, p. 4. Octopodous. Body segmented (or last two segments fused). Proboscis immovable. Chelifers present, chelate. Palps absent, or represented in ¢ by a I-4-jointed rudiment. Ovigers in both sexes, 10-jointed, with or without apical claw, distal joints with a single row of serrate, acute, or oval spines. Legs with or without auxiliary claws. Genital pores on all legs in 9, on last two legs in g. Eggs in several masses. Key to the South African Genera I. Ocular tubercle at hinder end of cephalic segment. A. Legs with auxiliary claws. Palps absent. 1. Oviger without apical claw. Finger and thumb of chelifer serrate or spinose. Callipallene 2. Oviger with apical claw. Finger and thumb of chelifer smooth, with small dentiform lobe. Pseudopallene* B. Legs without auxiliary claws. 1. Oviger with apical claw. Palps absent. Parapallene 2. Oviger without apical claw. a. Palps reduced to a single knob-like joint. Metapallene b. Palps absent. Pallenoides II. Ocular tubercle at front end of cephalic segment. Palps absent or reduced to tubercles. A. Legs with auxiliary claws. Oviger without apical claw. Pallenopsis B. Legs without auxiliary claws. Oviger with apical claw. Hannonia Gen. Callipallene Flynn 1836. Johnston, Mag. Xool. Bot., i (4), p. 380. (Pallene; nom. preocc. Megerle, 1823.) 1908. Loman, loc. cit., p. 42. (Pallene.) | 1929. Flynn, Mem. Queensl. Mus., 1x, p. 252, footnote. 1948. Correa, Pap. Avulsos Dept. Zool. S. Paulo, ix, 1, p. 1 (key to species). *The presence or absence of auxiliary claws is not specifically mentioned in the single South African species referred to this genus. See p. 107. 106 ANNALS OF THE SOUTH AFRICAN MUSEUM 1952. Stock, Beaufortia, no. 13, pp. 1-14, figs. 1-27. (European species. ) 1953. id., Boll. Mus. Civ. Venezia, vi, 2, p. 179. Cephalic segment with neck. Last 2 body segments often fused. Proboscis short, without setae around mouth. No trace of palps. Chelifer scape I-jointed, finger and thumb finely denticulate. Oviger 10-jointed, without (usually) apical claw, spines on distal joints oval or round (not pointed), smooth or finely serrated, 5th joint in ¢ with apical process. Legs with auxiliary claws. Callipallene sp. Fig. g General appearance, length and shape of neck similar to fig. 7 of Stock 1952; all sutures between body segments distinct. Crurigers separated by Q b | Cc d Fig. 9. Callipallene sp. a. hand of chelifer. 6. propodus and claw of leg. c. 5th joint of oviger g. d. distal joints of oviger 3. intervals subequal to their own width; length slightly less than median width of segments. Ocular tubercle conical. Abdomen obliquely upturned at 45-60°. Body and crurigers glabrous. Chelifer scape stout, hand turned downwards, inner margin of finger minutely and indistinctly serrulate, of thumb with about 10 conspicuous blunt spiniform teeth. Oviger 5th joint longest, in ¢ with outstanding apical process bearing a seta, the 4 distal joints with 38-46 (10, 10, 9, 9,—12, 12, 10, 12) bluntly rounded spines, the margins of which appear to be quite smooth, no apical claw. SOUTH AFRICAN PYCNOGONIDA 107 Legs, femur in @ swollen, rst tibia shorter than femur and shorter than and tibia, which is slightly longer than femur, propodus gently curved, with 4 large spines proximally, auxiliary claws strong, two-thirds length of claw, simple (not pectinate at base); glabrous except for one or two fine setae on each joint. Body incl. abdomen 1°5, 2nd leg 7-5 mm. Localities: off Walker Point, 47 fath., and off Cape St. Blaize, 40 fath. (S. Afr. Mus.). Three specimens (2 jg, 1 mutilated 9 with developing ova). Remarks. As these specimens show no outstanding features by which they can be separated from the other species of the genus, I think no specific name should be given to them until further material is available. Gen. Pseudopallene Wilson 1804. Latreille, Nouv. Dict. Hist. Nat., xxiv, p. 137. (Phoxichilus.) 1878. Wilson, Trans. Connect. Ac., v. 1902. Stebbing, Knowledge, xxv, p. 187. (Phoxichilus, non. Latr.) 1909. Schimkewitsch, loc. cit., pp. 6, 7, 8. 1910. Hodgson, Schultze Reise, iv, p. 225 (in key). 1917. Bouvier, Res. Sci. Camp. Monaco, li, p. 28. nego. Elynn, loc. 'cit..p. 23: 1951. Hedgpeth, Smiths. Misc. Coll., cvi, p. 4. 1953. Stock, Temmuinckia, ix, p. 204. [not Pseudopallene Bouvier 1913 = Austropallene Hodgson 1914] Proboscis. with setae around mouth. No trace of palps. Chelifer scape 1-jointed, hand short and swollen, finger and thumb not serrated or spinose, but with dentiform projections. Oviger 10-jointed, with serrated apical claw, 5th joint in g with apical projection. Legs without auxiliary claws. Remarks. After considerable argument as to the validity of Latreille’s name, the name Pseudopallene seems to have been generally accepted. Pseudopallene gilchristt Flynn 1Q2O ge ly iilWlOC- Cit. 4) 29 pe. 113. Body slender, segmented. Crurigers widely separated. Neck elongate. Ocular tubercie low, rounded. Abdomen shorter than last crurigers. Chelifer scape shorter than proboscis, opposable margins of finger and thumb each with a small rounded lobe. Legs slender, elongate, minute spines on Ist and 2nd tibiae, claw more than half length of propodus (no auxiliary claws ?). Body 3:6-3:9, femur 3°8-4°8, 1st tibia 4°5-5°6, 2nd tibia 6-6-8 mm. Locality: off Port Natal (Durban), surface (tow-net) (Flynn). Remarks. Flynn makes no mention of auxiliary claws; presumably they are absent, in conflict with the generic definition. No examples of this species are in the South African Museum collection. 108 ANNALS OF THE SOUTH AFRICAN MUSEUM Gen. Parapallene Carp. 1892. Carpenter, Sc. Proc. R. Dublin Soc., n.s., vii, p. 553- 1908. Loman, loc. cit., pp. 40, 42. 1909. Schimkewitsch, loc. cit., pp. 7, 9. 1929. Flynn, Mem. Queensl. Mus., ix, p. 258. 1937. Calman, Ann. Mag. Nat. His. (10), xx, p. 530. 1938. id., John Murray Exp., v, p. 156. 1953. Stock, Temmuinckia, ix, pp. 282, 297, fig. 2b (chart). Body segmented. Proboscis constricted in middle. Chelifer scape 1-jointed, hand swollen, finger and thumb fitting closely together when closed, margins not denticulate. Palps completely absent. Oviger 10-jointed, in 9 4th joint longest and 5th without apical process, in g 5th longest and with or without apical process, distal 3 or 4 joints with a row of biserrate (usually, sometimes obscurely serrate or simple) spines, with apical claw. Legs without auxiliary claws (except in challengeri). Genital pores on last 2 legs in 3, on all legs in 9. Key to the South African Species 1. Each cruriger with conspicuous pointed process dorsally. a. 3rd coxa with conical process of ventral apex. spinosus b. 3rd coxa with semicirclet of strong spines on ventral apex. calmani 2. Crurigers without processes (or only a minute tubercle on each). a. Neck long. i. 3rd coxa with a single spine on ventral apex. Feebly serrate spines on 7th-ioth joints of oviger 19, 16, 16, 16. algoae ii. Strongly serrate spines on 7th-1oth joints of oviger 10, 9, 7, 9. nierstrazi b. Neck short. 3rd coxa with semicirclet of fine spinules on ventral apex. Feebly serrate spines on 7th-1oth joints of oviger 8-9, 8-9, 7, 7-6. hodgsoni Parapallene spinosus (Mobius) Fig. 10 1902. Mobius, D. Tiefsee Exp., ili, p. 188, pl. 28, figs. 8-12. (Anoplodactylus s.) Crurigers well separated by about their own width, length subequal to or slightly longer than median width of segments; each with an upstanding conical process on dorsal apex. Neck moderately constricted, not defined by a collar at base. Proboscis cylindrical, apically truncate. Ocular tubercle high, conical, apex recurved and shortly bifid or narrowly truncate (acute in juv. with body length 7-8 mm.), eyes of uniform size. Abdomen in 9 not longer than last crurigers, in ¢ slightly longer, rather stout, nearly erect. Glabrous. Chelifer scape stout, with a spine on upper apex and a fringe of spine-setae on inner surface, hand curved downwards and inwards so that finger lies ventral to thumb, palm subglobular, setose on inner surface around base of finger and thumb, finger longer than thumb, inner margins of both entire. Oviger 4th joint longest, in 2 8th-1oth joints with 7, 4-6, and 6 spines on inner margin, spines moderate, simple or sometimes obscurely notched, or apically somewhat hooked, claw with a few (8-9) spines on inner margin and 3 on both SOUTH AFRICAN PYCNOGONIDA 109 inner and outer margins apically, the last pair adpressed to base of spiniform apex (but these somewhat variable); in g 7th-roth joints with the spines slightly stronger than in Q but not stout (claw missing in the only adult 3 specimen). Legs rather stout, 3rd coxa with a conspicuous conical process on ventral apex, femur about one and a half times 1st tibia, with 3 conical tubercles on dorsal apex, the medio-dorsal one much larger than the dorso-lateral ones, Ist tibia with 4 conical tubercles on dorsal apex, the dorsal pair larger than the b Fig. 10. Parapallene spinosus (Mobius). a. inner and front views of left chela. b. lateral view of 3rd coxa of leg. c. propodus and claw of leg. d. gth and toth joints with claw of oviger Q. e. oviger of juvenile, body length 8 mm. lateral ones (but not as large as the medio-dorsal one on femur), also a small medio-dorsal tubercle near base, and one near middle on each lateral surface; ond tibia a little longer than femur; propodus without heel, with 4 strong spines proximally on lower margin, followed by smaller ones, upper apex projecting over base of the stout claw. Very short and fine spinules ventrally on apices of 2nd and 3rd coxae; on the other joints in these spinules become more numerous distally, especially on 2nd tibia, but to the naked eye the legs appear glabrous or the tibiae very finely hispid; in ¢ however there are a few distinct but slender spines on femur, several on Ist tibia, and more on 2nd tibia. Base of chelifers to base of abdomen 2 10, abdomen 2-2:25, proboscis 3, and leg 38 mm. (femur 10°5, Ist tibia 6-5, 2nd tibia 12). Localities: St. Francis Bay, Agulhas Bank, 10 metres (Mobius); off East London, 52 fath. 1 9, 1 juv.; off Glendower Beacon (Port Alfred), 66 fath. I not quite adult g; False Bay, 17-33 fath. 2 99, 2 juv.; Table Bay, 22 fath. 1g (S. Afr. Mus.). I10 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. The conical process of the 3rd coxa is very conspicuous and forms an easy mark of identification. See further under calmanz. Loman (1908, loc. cit., p. 64) suspected that this species should not be included in Anoplodactylus on account of the number (Qg) of joints in the oviger. The South African Museum specimens confirm this. They are obviously examples of spinosus, although Mébius seems to have ignored the position of e ‘ . Fig. 11. Parapallene calmani Flynn. a, 6. lateral and ventral views of 3rd coxa of leg. c¢. distal joints of oviger 3 with roth joint and claw further enlarged. d. distal joints of oviger 2. e. femur and 1st and 2nd tibiae of leg ¢. the ocular tubercle and makes no mention of it as he does in the case of Anoplodactylus aculeatus. He had only one specimen and rashly assumed that the 2 would be without ovigers. One of the South African Museum specimens (body length 8 mm.) is particularly interesting because it has an oviger closely resembling M6bius’s figure 12; a figure is given here showing that it is not quite fully developed, with the distal joints not completely demarcated and without spines on inner margins. P. spinosus is a larger species than calmant. Parapallene calmani Flynn Big 00 1928. Flynn, loc. cit., p. 16, figs. 7,8. In general the description given for spinosus will apply to calmani. Chelifer with spine on upper apex. The 3rd coxa is not conically produced, but has a SOUTH AFRICAN PYCNOGONIDA Tete) semicirclet of about 8-9 strong spines on ventral distal margin. The oviger, both in ¢ and 9, has stronger spines on the inner margins of 7th-r1oth joints. Localities: off East London, 47 fath. (Flynn); Algoa Bay, 10-17 fath. 3 non-ovig. §g, 1 9; off Glendower Beacon (Port Alfred), 66 fath., 1 non-ovig. 6 (S. Afr. Mus.). Remarks. The presence in South African waters of two species both with conical projections on the crurigers makes the identification of Flynn’s species a little doubtful. Flynn might have overlooked the apical spines on the 3rd coxa, and his fig. 8 gives no indication of them; on the other hand he could scarcely have overlooked conspicuous conical processes and his figure, from the viewpoint from which it is drawn, would surely have given some indication Fig. 12. Parapallene algoae Brnrd. a. cephalic segment with chelifers. 6. lateral view of ocular tubercle. c. distal joints of oviger 9. of projections if they had been present. I therefore assign the specimens with a semicirclet of spines on 3rd coxa to calmani; and the species with conical processes on 3rd coxa is obviously spinosus (Mobius). Flynn says the 4th joint of the 10-jointed oviger is the longest; his fig. 7 shows seven joints beyond the longest joint, making an 11-jointed oviger. This is probably a draughtsman’s error. Parapallene algoae Brnrd. Bic 1946. Barnard, loc. cit., p. 61. Body rather slender. Crurigers without processes, well separated by at least their own width, length greater than median width of segments. Neck well marked, only half as wide as apical width of cephalon, not defined by a ie? ANNALS OF THE SOUTH AFRICAN MUSEUM collar. Proboscis cylindrical, apically truncate. Ocular tubercle high, conical, apex, acute, eyes of uniform size. Abdomen shorter than last crurigers, nearly vertical. Body and crurigers microscopically hispid. Chelifer as in calman, but without spine on upper apex of scape. Oviger (2) 4th joint longest, 7th-1oth joints with numerous spines: 19, 16, 16, 16, spines biserrate apically, claw finely serrate on inner margin except at base. Legs (2) 3rd coxa with a single short but prominent spine on ventral apex, dorsal apex of femur and ist tibia each with a semicirclet of 4-6 spines, Ist tibia distinctly shorter than femur, 2nd tibia one and a half to one and three- quarter times as long as Ist tibia, propodus as in calmam but lower margin with 7-8 large spines, 3 or 4 of the proximal ones being the largest, claw relatively longer than in calmam. Femur minutely hispid, ist tibia finely spinulose-setose, 2nd tibia more strongly so, the longest (but none as long as width of tibia) spine-setae forming a row or fringe along each side. Base of chelifers to base of abdomen 15, abdomen 2, proboscis 4:5 mm.; in S. Afr. Mus. specimens all legs severed from bodies, longest 80 mm. (femur 19'5, Ist tibia 14, 2nd tibia 20). Localities: Algoa Bay, off Gt. Fish Point, and off Cape Morgan, 32-87 fath. 4 99 (S. Afr. Mus.); Plettenberg Bay, 30 fath. (Port Elizabeth Mus.). Parapallene hodgsont Brnrd. Fig. 13 1946. Barnard, loc. cit., p. 61. Crurigers well separated by about their own width, length subequal to median width of segments, each with a very small tubercle on upper surface near apex. Neck constricted, but short, without collar at base. Proboscis cylindrical, apically truncate. Ocular tubercle high, conical apex acute or bifid, eyes of uniform size. Abdomen as long as last crurigers, oblique. Body and crurigers glabrous. Chelifer as in calmani but without spine on apex of scape. Oviger (9) 4th joint longest, 7th-10th joints with rather stout spines resp. 8-9, 8-9, 7, 7-6, spines simple or feebly serrate subapically, claw very feebly serrulate on inner margin. Legs (2) 3rd coxa with semicirclet of fine spinules on ventral apical margin, femur longer than Ist tibia, 2nd tibia longer (about one and one-third) than femur, propodus as in calmani but spines on distal part of lower margin stronger. Femur slightly hispid, 1st and 2nd tibiae more conspicuously so, 2 (or 3) small spiniform tubercles on upper apex of femur and st tibia. Base of chelifers to base of abdomen 8-5, abdomen 1-5, proboscis 3 mm.; and leg 48 mm. (femur 12, Ist tibia 10, and tibia 16-5 mm.). Localities: off East London and Hood Point, 47-52 fath. 3 99, 1 juv. (S. Afr. Mus.). SOUTH AFRICAN PYCNOGONIDA 113 Remarks. The feebly serrate or simple spines on 7th-10th joints of oviger prevent these specimens from being identified with merstraszi; also the neck is shorter. Parapallene merstrasz1 Loman 1908. Loman, loc. cit., p. 44, pl. 9, figs. 122-127. 1928. Flynn, loc. cit., p. 18. 1938. Calman, loc cit., p. 158, fig. 7. 1953. Stock, loc. cit., pp. 297, 299, fig. 3 a (chart). Fig. 13. Parapallene hodgsoni Brnrd. a. posterior view of cruriger. 5b. lateral view of ocular tubercle, with hind view of apex. c. distal joints of oviger 9. Locality: 29° 35’ S., 31° 14’ E. (Natal), 25 fath. (Flynn). — Distribution. East Indies. Remarks. Calman examined a syntype which had the ocular tubercle truncate with 2 points (fig. 7 A); this might be regarded as a malformation. He also described the spines on the coxae and following joints but without stating whether they were dorsal or ventral; presumably they were dorsal. The spines on 7th-1oth joints of oviger are strongly serrate (Loman, fig. 124). Gen. Metapallene Schimk. 1909. Schimkewitsch, loc. cit., pp. 7, 11, and table I. 1910. Hodgson, loc. cit., p. 235 (Heteropallene). 1938. Helfer, SB. Ges. naturf. Fr. Berlin, 1937, p. 172 (Procidella). 1952. Stock, Bull. Inst. Roy. Sci. Nat. Belge, xxviii, 14, p. 4 (remarks on Procidella) . 1953. id., Beaufortia, 1V, No. 35, p. 38 (remarks on Procidella). It4 ANNALS OF THE SOUTH AFRICAN MUSEUM Body robust, segmented, crurigers not widely separated. Cephalic segment broad, distally expanded, neck short. Proboscis short, without setae around mouth. Abdomen short. Ocular tubercle at base of neck. Chelifer scape 1-jointed. Palp reduced to a single joint. Oviger 10-jointed, 5th joint with or - without apical process in 3, distal joints with a single row of denticulate spines, no apical claw. Legs without auxiliary claws. Remarks. At Liideritzbucht, S.W.A., three specimens of a small Pycnogonid have been collected and each has been made the type of a new species by three separate authors. Although one suspects that these three specimens all belong to one and the same species, the question is into what genus this species should be put. Two new genera have been proposed. Heteropallene Hodgson must fall into synonymy, because both Hodgson and, a year earlier, Schimkewitsch named Pallene languida Hoek (1881, p. 79, pl. 12, figs. 1-5) as the genotype of their genera. M. languida is based on a 3 with the palps reduced to mere knobs; Hodgson’s specimen, sex not stated, had rudiments of palps ‘each a slightly curved joint’ ; in Procidella based on a 9 Helfer said each palp was reduced to a single conical joint, but this is really the undeveloped oviger (see his figure) which he stated was absent. In Hodgson’s specimen the 5th joint of the oviger ¢ has no apical process, whereas in M. languida it has. In view of these ambiguities, the two earlier Liideritzbucht specimens can be only provisionally included in Metapallene, with Procidella as a synonym. The third specimen has recently been adequately described as Pallenozdes magnicollis (see infra). Stock (1953) considers that Procidella gibber may be identical with M. dubitans, but that as it was founded on a juvenile specimen it should be ignored. Metapallene dubttans (Hodgson) Fig. 14a 1910. Hodgson, loc. cit., p. 226, fig. 4 (distal joints of oviger) (Heteropallene d.). 1938. Helfer, loc. cit., p. 172, fig. 5 (Procidella gibber). 1953. Stock, loc. cit., p. 38, fig. 3 (Procidella gibber). Body very robust, crurigers narrowly separated, with a few small setae distally, the posterior pair very short and almost completely fused, with the short abdomen embedded between them (Hodgson) or freely projecting (Helfer’s figure). Ocular tubercle low, broad, near hind margin of cephalic segment (Helfer’s figure); eyes small (Hodgson), without pigment (Helfer). Chelifer stout, scape and hand setose, finger and thumb denticulate. Oviger 4th and 5th joints longest, 5th without apical process (Hodgson’s figure), 7th-ioth joints with 45 (12, 10, 11, 12 in Hodgson’s figure) serrate spines; no apical claw. Legs sparsely setose; propodus with 2 large spines proximally, claw very stout, no auxiliary claws. SOUTH AFRICAN PYCNOGONIDA Length 1-5 mm. (Hodgson), 4:4 mm. (Helfer). Locality: Liideritzbucht (Hodgson, Helfer). Hodgson’s specimen was collected by Dr. L. Schultze; Helfer’s specimen was collected 18 Dec. 1903, therefore presumably also by Dr. Schultze who was in Liideritzbucht in December 1903 (1908. Schultze, Reise, i, Introduction, p. vi). Fig. 14. a. Procidella gibber Helfer 2 (after Helfer). 5, c. Pallenoides magnicollis Stock, 3 with propodus of leg and distal joints of oviger (spines omitted) further enlarged (after Stock). Gen. Pallenoides Stock 1951. Stock, Mem. Inst. Roy. Sct. Nat. Belge, (2) fasc. 43, p. 8. 1952. id., Bull. Inst. Roy. Sct. Nat. Belge, xxvili, 14, p. 4. Differs from Metapallene in having no trace of palps. Oviger without apical Proboscis with a constriction near base, and setae process on 5th joint J. around mouth. Pallenoides magnicollis Stock Fig. 14), ¢ 195 Us otock, loc. eit. p. 11, figs..7-13: Compared with Metapallene dubiians the following points may be noted. Abdomen short, embedded between the last crurigers as in Hodgson’s description (thus differing from Helfer’s Procidella). Ocular tubercle farther forward than in Helfer’s figure; eyes distinct. Oviger with 48 (14, 11, 10, 13) serrate spines on 7th-1oth joints. Length 1-75 mm. (3). Locality: Liideritzbucht, 8 fathoms. ‘Mercator’ (Stock). 116 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarks. ‘The locality and the close resemblance of this specimen to the specimens described by Hodgson and Helfer, lead one to suspect that all three specimens belong to one species. However, Stock has given a good recognizable description of a species, whereas it must be admitted that dubitans is a species inquirenda. Provisionally therefore I follow Stock. Re-examination of Hodgson’s and Helfer’s specimens may help, but abundant fresh material is really required to solve the problem satisfactorily. Gen. Pallenopsis Wilson 1881. Wilson, Bull. Mus. Comp. Zool. Harv., viii, p. 250. 1902. Mobius, D. Tiefsee Exp., iti, p. 184. 1908. Loman, loc. cit, p. 65. 1913. Bouvier, 2me Exp. Antarct. Fr., p. 107. 1915. Calman, Terra Nova Exp., iii, p. 41. 1916. Loman, Zool. Med., ii, p. 15. 1923. Calman, Rec. Ind. Mus., xxv, p. 279. 1927. Hodgson, D. Stidpol Exp., xix, p. 334. 1932. Gordon, ‘Discovery’ Rep., vi, p. 87. 1933. Stephensen, Medd. om Gronland, \xxix, 6, p. 21. 1947. Hedgpeth, Smiths. Misc. Coll., cvi, pp. 3, 4, 6. 1953. Stock, Temminckia, ix, pp. 281, 288, fig. 4 (chart). Body segmented, or segments fused. Ocular tubercle on front margin of cephalic segment overhanging base of proboscis. Chelifer with scape 2-jointed, or undivided, finger and thumb moderate, when closed meeting or gaping. Palps reduced to tubercles. Oviger 10-jointed in g, in 9 rudimentary, some of the joints sometimes fused; apical joint with numerous spine-setae, or a single row of non-serrate spines, no apical claw. Legs with auxiliary claws large or small, occasionally absent. Genital pores on last two legs in g, on all legs in 9. Key to the South African Species A. Finger and thumb of chelifer meeting when closed. 1. Body segmented. a. Auxiliary claws strong. i. 2nd tibia one and one-sixth to one and a quarter as long as Ist tibia. Legs (at least in 3) conspicuously setose. intermedia ii. 2nd tibia one and one-third to one and a half as long as Ist tibia. Legs not conspicuously setose. capensis 6. Auxiliary claws weak. brevidigitata 2. Body unsegmented or partly segmented (Rigona). a. All segments fused. tst tibia with a few spine-setae, some arising from small tubercles. ovalis b. 3rd and 4th segments fused. 1st tibia with numerous digitiform processes bearing spine-setae. [crosslandi|* B. Finger and thumb of chelifer slender, curved, gaping when closed. oscitans * P. crosslandi Carpenter (1910. 7. Linn. Soc. Lond., xxi, p. 257, pl. 27, figs. 10-20. Sudanese Red Sea) is included because it was listed by Flynn (1928, p. 6), although neither Flynn nor, so far as I can discover, any other author has recorded it from South African waters. - — _ — = . a aE SOUTH AFRICAN PYCNOGONIDA Lyle]. Pallenopsis intermedia Flynn Figs. 15a, 16 ? 1923. Loman, Ark. Zool., xv, 9, p. 10. (Pallenopsis sp.) 1923. id., Medd. Goteb. Mus., no. 22 (Goteb. K. Vet. Handl., xxvi), p. 3 ( fluminensis, non Kréyer). 1928. Flynn, loc. cit., p. 20, figs. 10-12. Body segmented. Crurigers well separated, but by spaces less than their own width. Ocular tubercle conical, ending in a sharp point, anterior eyes Fig. 15. a. Pallenopsis intermedia Flynn. b, c. Pallenopsis brevidigitata Mobius, with toth joint of oviger. d. Pallenopsis ovalis Loman. larger than posterior pair. Proboscis cylindrical, apically rounded. Abdomen directed more or less obliquely upwards. Spiniform setae on dorsal surface of body, scattered or more or less in paired groups of 1 or 2 on each segment, shorter and more numerous along ends of crurigers, 9 usually less setose than dg; abdomen with 2 rows of spines (often rubbed off). Chelifer scape divided into 2 subequal portions, with spiniform setae on distal margin of each; finger and thumb simple, closely fitting, finger longer than thumb, glabrous (without spinose pad). Palpal tubercles distinct. Oviger 10th joint as long as gth, oval, apically blunt, with numerous spine- setae; in g 5th joint distinctly longer than 4th, and 6th conspicuously swollen. Legs 2nd coxa subequal to 1st + 3rd, femur and Ist tibia subequal or the latter slightly the shorter, 2nd tibia slightly longer than femur; propodus cylindrical, without heel and not projecting over base of claw; auxiliary claws large, at least half length of main claw; short spiniform setae on all 118 ANNALS OF THE SOUTH AFRICAN MUSEUM joints, longer setae on 1st and and tibiae arranged chiefly in 3 fringes (one mid-dorsal and one along each side), some of the longest setae slightly pinnate, dg more conspicuously setose than 9. Cement duct on ventral surface of femur short but distinct in ovig. ¢. Body incl. abdomen ¢ 13, 2 15, proboscis 35, 2 6, 2ndleg g 56 (femur 13, Ist tibia 12, and tibia 14), 9 67 mm. (resp. 17, 15, 18). Amber-coloured, darker lines on sides of legs. Localities: Sebastian Bay, Walker Bay, Cape Barracouta and Cape Infanta, 24-40 fath. (Loman); False Bay, off Cape Infanta, and off Gneka River (see footnote, p. 85), 23-43 fath. (Flynn); False Bay and Agulhas Bank as far east as Cape St. Francis, and in the East London area, 20-90 fath. (S. Afr. Mus.). Remarks. In the s.s. Pieter Faure collection there are no specimens from the area between Cape St. Francis and the East London area (of which the most westerly locality is Gneka River mouth) in spite of the very considerable amount of trawling done in and around Algoa Bay. On the trawling grounds on the Agulhas Bank between Cape Infanta and Cape St. Francis this is the commonest Pycnogonid. For this reason it seems certain that the specimens identified as fluminensis Kroyer (a Brazilian species) by Loman (1923) should really be assigned to intermedia. | Ovigerous gg were found in September, October, December, February, April, May and July; thus breeding probably occurs throughout the year. Apparently very like vanhéffen Hodgson (1927. loc. cit., p. 336, fig. 9) from the Antarctic, but the latter has a slender proboscis. Loman (1923. Ark. Zool.) did not give the size of his specimen from Cape Point Lighthouse; presumably it was collected in the littoral zone like Discoarachne brevipes and Nymphopsis abstrusus (= cuspidata). It may perhaps be a specimen of this species. | Pallenopsts capensis Brnrd. 1946. Barnard, loc. cit., p. 62. Body segmented. Crurigers widely separated (cf. brevidigitata, fig. 155), each longer than median width of segment and about one and a half times as long as their own width. Ocular tubercle high, bluntly conical, anterior eyes slightly larger than posterior pair. Proboscis cylindrical, apically rounded. Abdomen horizontal or slightly oblique, about as long as last cruriger. Body with 1 or 2 short spine-setae in middle of each segment, and a few similar ones on ends of crurigers. Chelifer scape longer than proboscis, divided into 2 subequal portions, hand as in intermedia, finger glabrous. Palpal tubercles conspicuous. Oviger as in intermedia, in g 5th joint only slightly longer than 4th, 6th conspicuously swollen. Legs 2nd coxa subequal to 1st + 3rd, femur longer than rst tibia, end tibia longer than femur, propodus without heel and not projecting over base of SOUTH AFRICAN PYCNOGONIDA 119g claw, lower margin with 4 or 5 large spines proximally and 4-6 shorter ones distally, auxiliary claws strong, half length of main claw. Fine and short spine-setae on apices of coxal joints, scattered over femur and Ist tibia, more numerous on 2nd tibia. Cement duct on ventral surface of femur inconspicuous in ovig. 3. | Body incl. abdomen ¢ 18, 2 16, proboscis g 6:5, 2 6, 2nd leg g 98 (and coxa 9, femur 25, Ist tibia 23, 2nd tibia 33), 9 75 mm. (resp. 7°5, 19, 17, 23). Fig. 16. Pallenopsis intermedia Flynn. a. distal joints of oviger g. 0b. chelifer. c. one of the longest tibial setae. d. propodus and claws of leg. Localities: ‘Table Mountain 8. X E#E., 58 miles, 190 fath. 1 ovig. ¢; Cape Point N. 16° E. 10 miles, 85 fath. 1 2; Cape St. Blaize N. X E. 73 miles, 125 fath. 1 non-ovig. ¢ (2nd leg 84 mm.). (S. Afr. Mus.) Remarks. A 9 specimen, off Tugela River mouth, 200 fath., resembles the above specimens in all characters mentioned, but is much more slender, resembling in fact the 9 brevidigitata with which it was taken. Body incl. abdomen 10:5, 2nd leg 72 mm. (femur 17, Ist tibia 17, 2nd tibia 21); width of femur and 2nd tibia resp. 1 and -75 mm., as compared with 2 and 1-5 mm. in the above @ capensis. Except for the auxiliary claws and the terminal joint of oviger the specimen would be identified as brevidigitata. 120 ANNALS OF THE SOUTH AFRICAN MUSEUM Pallenopsis brevidigitata Mobius Fig. 150, c 1902. Mobius, loc. cit., p. 185, pl. 27, figs. 7-13. 1928. Flynn, loc. cit., p. 19, fig. 9. 1938. Calman, John Murray Exp., v, p. 160. Body segmented. Crurigers widely separated, at least twice as long as their own width, width across 2nd pair subequal to body length (excl. abdomen). Ocular tubercle short, conical, on extreme front of cephalon. Proboscis cylindrical, apically rounded. Abdomen horizontal, not extending as far as end of last crurigers. Dorsal surface glabrous. Chelifer scape divided, 1st portion longer than (but not ‘nearly twice as long as’) and, hand as in intermedia, finger glabrous. Palpal tubercles small. Oviger in 9 4th and 5th joints subequal, 6th slightly expanded, more so in J, 1oth joint slender, apically tapering, with a single row of large spines on lower margin. Legs 2nd coxa at least twice as long as 3rd, femur and ist tibia subequal, and tibia longer, propodus without heel and not projecting over base of claw, with large spines proximally on lower margin, followed by a comb-like series of smaller spines, claw three-quarters length of propodus, slender, auxiliary claws very short. Fine and short spine-setae, chiefly in 3 rows, on femur, Ist and 2nd tibiae. Cement duct on ventral surface of femur inconspicuous in ovigerous ¢. Body incl. abdomen @ 9:5, proboscis 5, 2nd coxa 5, femur 16, Ist tibia 15, and tibia 18 mm. Localities: 29° 44'S. 31° 20’ E. (off Durban), 46 fath. (Flynn); off Gwayang River, Mossel Bay, 31 fath. 1 damaged 9; off Tugela River mouth, 200 fath. 1 damaged 2; off Cape Natal, 54 fath. 2 ovig. 3d, 2 99 (S. Afr. Mus.). Distribution. Off Dar-es-Salaam, 404 metres; Zanzibar area, 421-457 metres. Remarks. All the South African Museum specimens are more or less damaged. Calman refers to several discrepancies in M6bius’s figures. Another inaccuracy occurs in fig. 7 which shows an extra (tibial) joint in the 2nd legs. Pallenopsis (Rigona) ovalis Loman Fig. 15d 1908. Loman, loc. cit., p. 68, pl. 10, figs. 137,)138: 1923. Calman, loc. cit., p. 284, fig. 11. 1928. Flynn, loc. cit., p. 23. 1953. Stock, loc. cit., fig. 3 @ (chart). Body unsegmented. Crurigers contiguous at their bases. A more or less distinct median longitudinal rib on cephalon. Ocular tubercle short, conical, SOUTH AFRICAN PYCNOGONIDA 121 anterior eyes scarcely larger than posterior pair. Proboscis cylindrical. Abdomen pointing obliquely upwards or nearly vertical. Dorsal surface glabrous. Chelifer with scape undivided, hand as in intermedia, finger with a spinose pad on basal half (cf. alcocki Calman, loc. cit., fig. 9 c). Palpal tubercles distinct. Oviger in ¢ 5th joint slightly shorter than 4th, 6th slightly swollen, 10th joint as in intermedia. Legs 1st coxa with 2 small spiniferous projections on upper apex, 2nd coxa distinctly longer than 3rd, femur and 1st tibia subequal or the latter slightly € Fig. 17. Pallenopsis oscitans (Hoek). a. tarsus, propodus and claw of leg. 6. hand of chelifer. c. distal joints of oviger @. shorter, both with a small spiniferous process on upper apex, 2nd tibia longer than femur or Ist tibia and more slender; propodus, claw and auxiliary claws as in intermedia. Numerous short spine-setae on joints, and some longer ones arising from small tubercles on ventral surface of femur conspicuous in ovigerous <. Body incl. abdomen 6, proboscis 3, femur 5°5, Ist tibia 5, 2nd tibia 6 mm. Another specimen, of which the body is damaged, has femur 7:5, rst tibia 7, and and tibia 8 mm. Localities: 10 miles SE. of East London, 47 fath. 1 ovig. ¢; Algoa Bay, 25 fatto ous. atc. Mus): Distribution. East Indies, Andaman Is., Ceylon. 122 ANNALS OF THE SOUTH AFRICAN MUSEUM Pallenopsis oscitans (Hoek) Fig. 17 1881. Hoek, Rep. H.M.S. ‘Challenger’, iii, p. 89, pl. 13, figs. 1-5. (Phoxichili- dium o.) Body segmented. Crurigers well separated (by about their own width), each slightly longer than median width of body. Ocular tubercle low and rounded, eyes not traceable. Proboscis cylindrical, slightly tapering distally, apically rounded. Abdomen elongate, slender, slightly clavate, oblique. A few fine and rather long setae on the crurigers and abdomen. Chelifer with scape a little longer than proboscis, 2nd joint slightly shorter than ist, hand nearly as long as 2nd joint, palm short, finger and thumb distally slender, curved, when closed gaping, apices crossed. Palpal tubercles moderately large but inconspicuous. Oviger in 9 5th joint a little longer than 4th, 6th not swollen, apical joint slender, with long fine setae. Legs 2nd coxa subequal to ist + 3rd, femur very slightly longer than 1st tibia, 2nd tibia longer than femur, propodus without heel but with a very large spine basally, followed by smaller ones on lower margin, apex not pro- jecting over claw, claw strong, reaching almost to base of propodus and forming with the large spine on the latter a prehensile ‘chela’, auxiliary claws short and weak. Fine setae on all joints, becoming more numerous on femur and ist and 2nd tibiae but not forming dense fringes as in intermedia, nor a thick fur. Body incl. abdomen 11°5, abdomen 4:75, proboscis 5, 2nd leg 47 mm. (2nd coxa 4, femur 11, Ist tibia 10, 2nd tibia 14 mm.). Locality: off Gape Point, N. 86° E., 43 miles, go0-1,000 fath. 1 9 (S. Afr. Mus.). Distribution. 38° 25’ N., 35° 80’ W. (Azores), 1,675 fath. Remarks. Although the proboscis is more like that of pzlosa as described by Hoek, and the legs are somewhat more hairy, there is little doubt that this specimen should be identified as oscztans; the apical joint of the oviger and the propodus of the legs, with its enormous basal spine, correspond with Hoek’s figures. The ocular tubercle shows no trace of eyes, but the specimen has suffered desiccation. Differs from longirostris Wilson in the apical joint of the oviger, and propodus; and from tritonis Hoek (syn. holti Carp.) in the same features and also in the relative lengths of palm and fingers of chelifer (see: Carpenter, Fish. Irel. Sct. Invest., 1905, pl. 1, figs. 3, 4, 6, holti; I have not seen Hoek’s 1883 paper). P. calcanea Steph. 1933 differs in having the ovigers closer together, a much larger ocular tubercle, a shorter abdomen, a different apical joint on the oviger, propodus of legs with 2 moderate-sized basal spines on a distinct heel, and no auxiliary claws. SOUTH AFRICAN PYCNOGONIDA 123 Pallenopsis sp. Gordon 1932. Gordon, loc. cit., p. 91, fig. 45. A young specimen from 35° 14’ S., 6° 49’ E., pelagic, which Stephensen (loc. cit., p. 24) suggests may be the same as his calcanea (loc. cit., p. 21, fig. 5) from Greenland. Gordon’s record scarcely comes within the South African area. Gen. Hannonia Hoek 1881. Hoek, Rep. H.M.S. ‘Challenger’, iii, p. 92. 1891. Sars, Norw. N. Ail. Exp. Pycnogonida, p. 6. 1902. Pocock in Lankester, Encycl. Brit., 1oth ed. Arachnida (type of a separate family). 1904. id., Q. 7. Micr. Sci., n.s. xlviii, p. 225 (reprint of 1902 article). 1904. Loman, Zool. Jahrb. Abt. Syst., xx, p. 385 (systematic position). 1905. Cole, Ann. Mag. Nat. Hist., (7) xv, pp. 408, 410 (systematic position). 1908. Loman, loc. cit., pp. 15, 16, synopsis facing p. 19 (systematic position). 1909. ‘Thompson, Cambr. Nat. Hist., iv, p. 533 (systematic position). 1927. Calman, Trans. Zool. Soc. Lond., xxii, p. 410 (systematic position). 1947. Hedgpeth, Smiths. Misc. Coll., cvi, p. 4. | Fig. 18. Hannonia typica Hoek. a. dorsal view. 6. lateral view, with tubercle (? vestigial palp) further enlarged. c. ocular tubercle of adult, from right side. d. chelifer of juvenile (above) and adult (below). e. oviger g with distal joints further enlarged. f. spine from leg. Body stout, segmented, crurigers narrowly separated. Proboscis on narrow stalk, swollen (sac-like), deflexed and bent beneath body. Chelifers rudi- mentary, 2-jointed, more or less chelate. Palps absent (see remarks). Ovigers 124 ANNALS OF THE SOUTH AFRICAN MUSEUM in both sexes, 10-jointed, not modified in g, with apical claw. Legs stout, tarsus very short, no auxiliary claws. Genital pores on last 2 legs in 4, on all legs in 9. Eggs in a single cake-like mass held by both ovigers. Femoral cement glands 3 seemingly absent. Remarks. An endemic South African genus with one species, whose systematic position has been the subject of much discussion. The presence of the minute papilla, between the bases of the chelifers and ovigers, discovered by Calman (1927) is herewith confirmed. I have not, however, seen any indication in my specimens of its being 2-jointed. It is present in both sexes, lies just below a small spiniferous process on the side of the cephalic segment, and has a subapical spinule. It occurs in a juvenile specimen, 2-5 mm. body length (as measured below), but is neither more nor less feebly developed than in adults. Calman’s suggestion that this papilla represents the last vestige of a palp would seem to bring the genus into an intermediate position in the series Béhmia-Rhynchothorax-Pycnogonum, where Loman in his conspectus (1908) placed it. Calman did not accept Bouvier’s suggestion that Hannonia was an Ammotheid. Hedgpeth (1947) places it in the Pallenidae. Hannonia typica Hoek Fig. 18 1691. Eloek; loc. Jeit..4p. 92, pli 14. fiesme-ise 1904. Loman, loc. cit., p. 383, pl. 14, figs. 12-15. 1910. Hodgson, Schultze. Reise, iv, p. 227. 1O23 ss Momany Aghw00l-..%V. Os Ong 1927. Calman, loc. cit., p. 410. Body compact, the posterior margins of the segments forming transverse arched ridges, each with a median boss. Crurigers separated by less than half their own width, their length subequal to or a little longer than median width of segments, each with a knob on dorsal apex. Ocular tubercle moderately high, bluntly conical, eyes distinct. Proboscis very much swollen beyond the comparatively narrow basal stalk. Abdomen extending beyond last crurigers, clavate, bent slightly downwards. Short spinules on the trans- verse ridges and bosses, crurigers, abdomen, top of ocular tubercle, and front and sides of cephalic segment. Chelifers never quite vestigial, but variable, the chela sometimes well developed and apparently with mobile finger. Ovigers similar in both sexes, but in ¢ the distal joints rather stouter and more spinose, the spines minutely setulose. Legs stout, 2nd coxa slightly longer than either ist or 3rd, on last 2 legs in 3 with a process on ventral apex, more conspicuous on 3rd than on 4th leg, femur slightly longer than either of the tibiae, which are subequal, tarsus narrower than 2nd tibia, propodus without heel, with numerous small but no SOUTH AFRICAN PYCNOGONIDA 125 strong spines on lower margin, claw about one-third, or a little less, length of propodus. Numerous spines on the joints, arising from slightly raised tubercles, particularly distinct on dorsal surface of 2nd tibia, and especially in juveniles. Femoral cement glands § seem to be absent; neither Loman (1904, p. 384) nor myself have found any. Body (base of chelifers to base of abdomen) 5, proboscis 4, abdomen 2, 2nd leg femur 2-5, 1st and 2nd tibiae 2 mm. Localities: Cape Town (Hoek, Loman); Liideritzbucht (Hodgson); Port Natal (Durban) (Loman); Saldanha Bay and Melkbos Strand (Table Bay) littoral, Mossel Bay, 20 fath. (S. Afr. Mus.). Remarks. Not nearly so common as Discoarachne brevipes; in the course of a considerable amount of shore-collecting at Sea Point (Table Bay) I have never found a specimen there, although on the other side of Table Bay (Melk- bos Strand) it is moderately common. Two ovigerous $4 were collected at Saldanha Bay in September. Fam. PHOXICHILIDIIDAE 1908. Loman, Szboga Exp. monogr., xl, p. 62, conspectus facing p. 19. (Phoxichilidae.) 1913. Schimkewitsch, Zool. Anz., xli, p. 611. 1947. Hedgpeth, Smiths. Misc. Coll., cvi, p. 4. Octopodous. Body segments free or fused. Chelifers well developed, chelate. Palps rudimentary, reduced to tubercules or absent. Ovigers in 3 only, 5-9-jointed, with simple spines. Legs usually with auxiliary claws. Key to South African Genera 1. Cephalic segment short, without neck. Oviger 5-jointed. Phoxichilidium 2. Cephalic segment with distinct neck; ocular tubercle near front margin. Oviger 6-jointed. Anoplodactylus Gen. Phoxichilidium M.-Edw. 1836. Johnston, Miscell. Zool., i, Mag. Zool. & Bot., i, p. 378. (Orithyia, preocc. Fabr. 1708.) 1840. Milne-Edwards, Hist. Nat. Crust., (Roret’s Suite a Buffon. Crust.), lll, Pp. 535. 1881. Hoek, Rep. H.M.S. ‘Challenger’, iii, p. 31 (part). 1908. Loman, Szboga Exp. monogr., xl. pp. 63, 64. * Body segments free. Ocular tubercle on anterior portion of cephalic segment, in advance of Ist crurigers. Chelifers with scape undivided, finger and thumb gaping when closed. Palps completely absent. Ovigers ($ only) 5-jointed. Legs with minute auxiliary claws. Genital pores on last 2 legs in 3, on all legs in 9. 126 ANNALS OF THE SOUTH AFRICAN MUSEUM Phoxichilidium capense Flynn. 1928. Flynn, loc. cit., p. 27, figs. 15, 16. Body stout, especially the cephalic segment and ist pair of crurigers. Crurigers narrowly separated (the 3rd and 4th pairs more widely separated). Ocular tubercle obtusely conical, eyes distinct. Abdomen rather short and stout, erect. Proboscis stout, apically truncate. Chelifer rather stout, as long as proboscis, hand (palm) longer than broad, finger and thumb curved, inner margins entire. Legs (2) stout, 2nd coxa with strong conical projection (bearing the genital pores) on ventral apex, femur longer than either of the tibiae, which are sub- equal (text; according to the measurements and figure the 1st tibia is slightly the longer), propodus with strong spines on heel, lower margin with about 12 spines, claw strong, auxiliary claws minute. Body 3°75, proboscis 3:18, femur 5:17, Ist tibia 4°31, 2nd tibia 3-75 mm. (Flynn). Locality: Hout Bay, west coast of Cape Peninsula, depth not given (Flynn). Known from a single 9. Gen. Anoplodactylus Wilson 1821. Say, 7. Ac. Nat. Sct. Philad., ii, p. 59. (Anaphia.) 1878. Wilson, Amer. 7. Sct. (3), xv, p. 200. 1908. Loman, loc. cit., p. 71. 1908. Norman, 7. Linn. Soc. Lond., xxx, p. 202. (Anaphia.) 1912. Loman, Bull. Inst. ocean. Monaco, no. 238, p. 7. (subgen. Halosoma Cole). 1923. Calman, Rec. Ind. Mus., xxv, p. 285. 1927. id., Trans. Zool. Soc. Lond., xxii, pp. 405, 407. Body segments free. Ocular tubercle on front margin of cephalic segment overhanging base of proboscis. Chelifer with scape undivided, finger and thumb either stout and meeting when closed, or slender and gaping. Palps completely absent. Oviger 6-jointed, ultimate and penultimate joints feebly developed, without strong spines, no apical claw. Legs with or without auxiliary claws. Genital pores on last 2 legs in 4g, on all legs in Q (but see Calman, 1927). Femoral cement glands ¢ on femur usually single, tubular (numerous and cribriform in cribellatus Calman, 1923). Remarks. Wilson’s genus is said to be the same as Anaphia Say, but this synonymy has not been generally adopted. Mobius’s South African species spinosus has been transferred to Parapellene. p Key to the South African Species 1. Hand of chelifer stout, finger and thumb short, meeting when closed. aculeatus 2. Hand of chelifer slender, finger and thumb long, slender, gaping when closed. _pelagicus SOUTH AFRICAN PYCNOGONIDA | edly | Anoplodactylus aculeatus Mobius 1902. Mébius, D. Tiefsee Exp., ili, p. 188, pl. 28, figs. 1-7. Body slender. Crurigers separated by intervals not greater than their width, their length subequal to median width of segments. Ocular tubercle conical, sharply pointed, eyes distinct. Abdomen short, vertical. Chelifer scape stout, longer than proboscis, with conical process on dorsal apex, hand stout, finger and thumb short, stout, meeting when closed. Fig. 19. Anoplodactylus pelagicus Flynn. a. dorsal view. 6. lateral view of cephalic segment. c. hand of chelifer. d. 3rd-6th joints of oviger g with 5th and 6th joints further enlarged. e. 2nd coxa of leg with genital pore. f. cement gland on femur. g. tarsus, propodus and claw of leg. Legs long, 2nd and 3rd coxae with a spiniform process on ventral apex, femur longer than ist tibia, both joints with a spiniform process on dorsal apex, ist tibia also with a spine at end of basal third of its length, 2nd tibia longer than femur (subequal in fig. 1), sparsely setose, propodus with 4 large spines on basal heel and 8 smaller ones on lower margin, claw strong, no auxiliary claws. ; Body, excl. abdomen, about 8 mm. (according to Mébius’s fig. 1). 128 ANNALS OF THE SOUTH AFRICAN MUSEUM Locality: Agulhas Bank, 126 metres, 2 99 (Mobius). Remarks. ‘This species, based on the female only, is very like insignis (Hoek) 1881, based on the g, from Bahia, S. America. Both have projections on the chelifer scape, 2nd and 3rd coxae, femur and Ist tibia; but insignis has also a process on the Ist coxa, and the crurigers are a little more widely separated. Cf. also insignis subsp. bermudensis Cole (1904. Proc. Boston. Soc. Nat. Hist., xxxi, p- 325, pl. 20, and pl. 22, figs. 21-29). Anoplodactylus pelagicus Flynn Fig. 19 1920. Flynn; loc, cits, ps25, fie. 4e Body slender. Crurigers separated by intervals greater than their width, their length greater than median width of segments. Ocular tubercle low and flattened or rounded on top, eyes small, unpigmented. Abdomen not very long, erect. Chelifer scape slender, extending a little beyond tip of proboscis, hand with long slender finger and thumb, curved and gaping when closed, inner margins with a few spine-setae. Oviger 3rd joint longer than 2nd, both elongate, 5th somewhat swollen, with a patch of setae distally and 3 small recurved denticles on inner margin, 6th shorter than 5th, ovate, setose on inner margin. Legs long, slender, ventral apex (with genital pore) of 2nd coxa of last 2 legs ¢ slightly more prominent than on first 2 legs; femur and 1st and 2nd tibiae subequal, 2nd tibia in the $.A.M. specimen not so much more slender than Ist tibia as is shown in Flynn’s figure, and the propodus not so noticeably thicker, 2 strong spines on basal heel and about 18 small spines on inner margin, claw strong, auxiliary claws very small (scarcely visible except when claw is seen in dorsal or ventral view); tibiae finely and sparsely setose. Cement gland on dorsal surface of femur tubular, not very long. Body, excl. abdomen, ¢g 2°5, proboscis 1-5, femur (also 1st and 2nd tibiae) 2:75 mm. Flynn’s measurements are slightly greater, except the leg measure- ments of ¢. Localities: off Port Natal (Durban), pelagic (Flynn); Cape Point NE. X E 2 E. 28 miles, 300 fath. 1 ¢ (S. Afr. Mus.). Remarks. Flynn refers (loc. cit., p. 3) to the capture of 7 specimens in a tow- net, but in the absence of further data it does not follow that they were from the surface or even pelagic: the tow-net may have been attached to the beam of the trawl. The single specimen in the South African Museum was taken together with “‘Alcyonarians’ (s.s. Preter Faure log-book, P.F. no. 18159). ‘Pallene’ lappa Bohm 1879. Bohm, MB. K. Ak. Wiss. Berlin, p. 182. Ibo, Mozambique. 1881. Hoek, loc. cit, p. 31. SOUTH AFRICAN PYCNOGONIDA 129 1910. Hodgson, loc. cit., p. 225. 1912. Loman, Bull. Inst. ocean. Monaco, no. 238, p. 6. 1926. Flynn, loc. cit. p. 4. Hoek and Hodgson considered that B6hm’s specimen was immature, with not fully developed ovigers. Loman, with whom Flynn agrees, included it in Halosoma, a subgenus of Anoplodactylus. Loman gives as one of the characters of Halosoma: ovigers 6-jointed; but Flynn says B6hm’s specimen has 7-jointed ovigers. The locality is, strictly speaking, outside (12° S.) the South African area, but the record is included here as Flynn has referred to it. Fam. ENDEIDAE 1908. Norman, 7. Linn. Soc. Lond., xxx, p. 231. 1932. Gordon, ‘Discovery’ Rep., iii, p. 93. See generic characters. Gen. Endeis Phil. 1843. Philippi, Arch. Naturg. Jahrg., ix, p. 175. 1902. Mobius, D. Tiefsee Exp., iii, p. 195. (Phoxichilus, non Latr.) 1902. Stebbing, Knowledge, xxv, p. 187. (Chilophoxus.) 1908. Loman, Siboga Exp. monogr., xl, p. 77. (Phoxichilus, non Latr.) 1908. Norman, loc. cit., p. 231. 1915. Calman, ‘Terra Nova’ Exp. Rep., iii, p. 48. no2gid-. iec. Ind. Mus., xxv, p. 200: Octopodous. Body segmented.. No chelifers. No palps. Ovigers only in J, 7-8-jointed, without apical claw, with simple spines. Legs with auxiliary claws. Key to the South African Species 1. Crurigers narrowly separated. Propodus of legs apically produced over base of claw. clipeatus 2. Crurigers well separated. Propodus of legs not apically produced. mollis Endets clipeatus (Mo6bius) Fig. 20 1902. Mobius, loc. cit., p. 196, pl. 30, figs. 6-10. (Phoxichilus clipeatus [sic]; clypeatus on plate.) 1920. 5 Elynn, loc. cit., p. 20: Body rather stout, cephalic segment with front margin feebly bilobed, with a minute chitinous point on each lobe in the ¢ specimen (possibly representing the remnants of the chelifers). Crurigers narrowly separated, about as long as median width of segments. Ocular tubercle short, conical. Abdomen about as long as last crurigers, apically notched. Proboscis stout, swollen in middle, apex truncate. Body and crurigers glabrous. Whole surface of body and appendages closely and minutely ‘pitted’ (? glandular). 130 ANNALS OF THE SOUTH AFRICAN MUSEUM Oviger 8-jointed, but articulation between 7th and 8th joints obscure, 6th joint with a variable number of more or less recurved spines on inner margin. Legs, femur and 2nd tibia subequal, 1st tibia slightly shorter, propodus apically produced in a cylindrical process over base of the claw, claw and auxiliary claws strong. Glabrous. Femoral cement glands about 40 in a single row. a F b Fig. 20. Endeis clipeatus (Mobius). a. tarsus, propodus and claw of leg. b. oviger 3. c. 6th and 7th—8th joints of same further enlarged. Body (3) incl. abdomen 4, proboscis 3, 2nd leg 20 mm. (femur 5, Ist tibia 4°25, 2nd tibia 5:25 mm.). Yellowish, a red line down middle of body and red longitudinal lines on legs. Localities: St. Francis Bay, shallow water (Mébius); off west coast of Cape Peninsula (Flynn); St. James, False Bay, littoral. 1 juv.; west coast of Cape Peninsula, littoral; False Bay, 10 fath. and Algoa Bay, 52-63 fath. 2 99,1 ¢ (S. Afr. Mus.). | Endeis mollis (Carp.) Rig om 1894. Carpenter in Herdman’s Ceylon Pearl Fish. Rep. Suppl. Rep., xiii, p. 182, pl. figs. 1-7. : 1907. id., Trans. Linn. Soc. Lond. (2) zool. xii, p. 98. 1923. Calman, loc. cit., p;,203,, fig.. 16. 1927. id., Trans. Kool. Soc. Lond., xxii, p. 408. 1938. id., John Murray Exp., v. p, 160. 1951. Stock, Mem. Inst. Roy. Sct. Nat. Belge. (2) fasc. 43, p. 17, figs. 23, 24. SOUTH AFRICAN PYCNOGONIDA 131 Body rather slender, cephalic segment with 2 rounded lobes in front over base of proboscis (possibly representing the chelifers). Crurigers separated by more than their own width, slightly longer than median width of segments. Ocular tubercle short, conical. Abdomen slightly longer than last crurigers, oblique. Proboscis stout, swollen in middle, apex truncate. Body and crurigers glabrous. Whole surface of body and appendages closely ‘pitted’ (? glandular). Fig. 21. Endeis mollis (Carp.). a. dorsal view. 6. inner view of 4th—7th joints of oviger 3. c. outer view of 6th joint of same. d. propodus and claw of leg. Oviger 7-jointed, on a basal knob, penultimate joint swollen, inner margin on inner side (as coiled up in situ) with 2 recurved spinules, outer surface with a broad lobe bearing 3 spinules, apical joint with 1 spinule on inner margin and 2-3 on apex. Legs, femur slightly longer than Ist tibia, smooth, cylindrical, with one fairly strong spine-seta and 2 smaller setae on dorsal apex, 2nd tibia slightly longer than femur, 1st and 2nd tibiae with a few scattered setae, propodus apically truncate, not produced over base of claw. Femoral cement glands about 23, in a single row. Body (tip of proboscis to tip of abdomen) ¢ 7:5, 2nd leg 18 mm. (femur 4°5, Ist tibia 4, 2nd tibia 4-75 mm.). 132 ANNALS OF THE SOUTH AFRICAN MUSEUM Locality: off Durnford Point, Zululand, 13 fath. 1 non-ovig. 3g, 1 9 (S. Afr. Mus.). Distribution. Ceylon, Maldives, Arabian coast, Indian coasts, Nicobars, Christmas Island, Tonga Is. Remarks. Apparently resembles the typical form. Calman records some variations in the shape of the femur, number of cement glands, etc. Fam. AMMOTHEIDAE 1908. Loman, Szboga Exp. monogr., xl, conspectus facing p. 19 and Pp. 49: 1909. ‘Thompson, Cambr. Nat. Hist., iv, p. 534. 1913. Schimkewitsch, Zool. Anz., xli, p. 612. 1947. Hedgpeth, Smiths. Misc. Coll., cvi, p. 4. Octopodous. Body extended or compact, segmented or the segmentation more or less obsolete. Proboscis more or less movable, directed forwards or downwards. Chelifers reduced, chelate in juvenile, but rudimentary in adult. Palps 6-10-jointed (Nymphonella 17-20). Ovigers in both sexes, (9)-10-jointed, with or without apical claw, with simple or serrate spines. Legs usually with auxiliary claws. Genital pores on all legs in 9, on last 2 legs in g. Eggs in several masses. | Key to the South African Genera I. Proboscis fusiform or subcylindrical, directed forwards or downwards. A. Palp 7-jointed. Béohmia B. Palp 8-jointed. Achelia C. Palp g-jointed. 1. All body segments fused. Crurigers narrowly separated. Nymphopsis 2. Body segments, at least the anterior ones, distinct. Crurigers widely separated. a. Legs with auxiliary claws. Ocular tubercle very elongate. Kyphomia 6b. Legs without auxiliary claws. Ainigma II. Proboscis flask-shaped, curving downwards. Austroraptus Gen. Béhmia Hoek 1881. Hoek, Rep. H.M.S. ‘Challenger’, ii, p. 24. 1902. Mobius, D. Tiefsee Exp., iii, p. 189. 1908. Loman, loc. cit., conspectus facing p. 19. 1947. Hedgpeth, loc. cit., p. 4 (in list of genera). Body stout, segmented, arched. Crurigers narrowly separated. Proboscis elongate, conical, inserted ventrally and bent under cephalic segment. Chelifers rudimentary, 2-jointed, chelate, inserted ventrally and invisible from above. Palps 7-jointed. Ovigers 10-jointed, with apical claw, not modified in g. Legs stout, tarsus very short, no auxiliary claws. Genital pores on all legs in 9, on last leg in g. Pale non-spinose (or almost so) areas on dorsal surface of all femora in g. Eggs carried in a single cake-like mass. SOUTH AFRICAN PYCNOGONIDA 133 Remarks. An endemic South African genus, with two species. Flynn (loc. cit.) claimed to have found genital pores on the dorsal surface of the 2nd coxa of all the legs; this position would be distinctly unusual, in fact unique, if it were correct. My own observations disprove this; the genital pores are in the normal ventral position, and on the last leg only (as in Pycnogonum). Key to the South African Species 1. Smaller species. Legs not more than twice as long as body. Tubercles distinct, but moderate. Cephalic segment parallel-sided or widening in front. Femur 2 (2) or 3 (g) times as long as wide. chelata 2. Larger species. Legs about four times as long as body. Cephalic segment narrowing in front. Femur (¢) four and a half to five times as long as broad. tuberosa Béhmia chelata (Bohm) Higes22 1879. Bohm, MB. Ak. Wiss. Berlin, p. 192, pl. 2, figs. 5-5d. (Pycnogonum c.) Roar © Eloek, loc. cit., p. 24. 1902. Mobius, loc. cit., p. 189, pl. 28, fig. 15 (front of cephalic segment). 1909. Schimkewitsch, Xool. Anz., xxxiv, p. 3, fig. 1 A and C (chelifer and palp). 1915. Calman. ‘Terra Nova’ Exp. Rep. ool., iii, p. 13 (oviger mentioned). ig208 2 blynun, loc. cit., p. 30, fig. 17. Body strongly arched, median portion of each segment raised into a rounded boss, similar to the ocular tubercle, the boss on 4th segment smaller than the others, all sharper and more pronounced in ¢ than 92; cephalic segment nearly parallel-sided in 2 with somewhat prominent rounded antero-lateral corners, in ¢ sides diverging forwards, antero-lateral corners prominently and subacutely produced. Crurigers narrowly separated, the last pair a little more widely separated; dorsal apices tending, especially in 4g, to form small nodiform bosses. Ocular tubercle conical apically rounded in Q, subacute in 4, eyes not conspicuous. Proboscis inserted ventrally, base wide, quickly tapering to a subacute apex, projecting downwards or bent under cephalic segment. Abdomen projecting beyond last crurigers, clavate, deflexed. Whole surface reticulate, with short spinules arising from conical bases, including ocular tubercle. Chelifer scape stout, spinose, chela rather slender, but well developed and apparently functional (finger mobile). Palp 4th joint much the longest, 5th small, 6th and 7th rather slender, more so in ¢ than Q, spinose. Oviger 5th joint longer than 4th, last 4 joints with rather strong spines on inner margins, mostly in pairs, resp. about 8, 5, 4, 5 pairs, sometimes fewer in g, inner apex of 10th joint in g with a single unguiform process resembling but smaller than the apical claw; alike in both sexes except for this latter feature. Legs stout, stouter in 9 than 3, decreasing in length backwards, the 3 coxae subequal in length, but 2nd somewhat stouter in 2 owing to the genital pores, 134. ANNALS OF THE SOUTH AFRICAN MUSEUM on last leg in § a little longer than either of the other two; femur longer than 1st tibia on all legs in 9, subequal to it in g or on the Ist leg slightly shorter, about twice as long as wide in 9, about 3 times in g, 2nd tibia subequal to femur in Q, slightly longer in 4, tarsus and propodus together a little shorter than Ist tibia, propodus without heel, with several short but stout spines along lower margin, claw about half length of propodus. Surface reticulate like that of the body, and spinulose. At dorsal apex of 2nd coxa of all legs in both sexes Fig. 22. Béhmia chelata (B6hm). a. abnormal double ocular tubercle. 6. chelifer. c. palp. d. oviger g with apex of roth joint and claw further enlarged. e. apex of 10th joint and claw of oviger 2 f. dorsal apex of 2nd coxa ($9) of leg. g. ventral apex of 2nd coxa of last leg ¢. h. dorsal surface of femur of 3 with pale patch (‘cement gland’). a small pale spot. Femoral cement gland in ¢ an elongate oval pale spot on dorsal surface on each leg. Eggs in a single cake-like mass. Body (from front margin of cephalic segment to base of abdomen) g@ 5:5, proboscis 2°5, abdomen 3 2, @ 1°5, 1st leg (without claw) approx. g 10, 2 8-5 mm. Amber-coloured, or reddish-brown, sometimes bicoloured dark and light, with dark bands on legs (see under Remarks). Localities: (loc. ? Bohm); St. Francis Bay, shallow water (Mobius); off Ball Point, Cape Town (Flynn); False Bay, littoral to 12 fath.; off Cape St. Blaize, off Knysna, and off Cape St. Francis, 25-70 fath; off Gt. Fish Point, Keiskamma, East London and Cape Morgan, 33-77 fath.; off Umtwalumi and Umhloti River mouths (Natal), 25-40 fath.; off Tugela River and Durnford Point (Zululand), 13-47 fath. 3¢, 29, juv. (S. Afr. Mus.). SOUTH AFRICAN PYCNOGONIDA 135 Remarks. On two points my observations differ from those of Flynn. The more important is the position of the § genital pores. Although Flynn had both gg and 99 he did not notice that the pale spot on dorsal apex of 2nd coxa occurred in both sexes. It is obviously not a genital pore. When the integument of this joint is treated with a clearing reagent the pale spot is seen to be due to feeble chitinization; elsewhere the integument is strongly but not uniformly chitinized, producing a reticulate pattern (fig. 22f). Further proof that these pale spots are not openings is shown by the fact that spines may be present on them as elsewhere on the integument. The true genital pore in g is a very definite opening, in the normal position on the ventral surface of the 2nd coxa, and it occurs only on the last leg. On the upper surface of the femur of all the legs is another pale patch, oval in shape and very conspicuous, but which occurs only in the g. Flynn calls this the cement gland, and it probably is; but I have failed to find the longitu- dinal slit-like opening described by him (cf. also under tuberosa). The paleness of the patch is due to lack of chitinization, as in the case of the spot on 2nd coxa; and there are minute spinules and even a few spines on its surface, if not centrally at least marginally. Some small circular structures (fig. 22h) may be the actual cement glands, but I cannot detect any individual or communal pores. P The chelate termination of the oviger was mentioned incidentally by Calman (1915), but he did not state that it is found in the male only. This species is common on branching Alcyonarians (Gorgonia, Melitodes, Spongioderma, Villogorgia, Wrightella, etc.) but I have also found it in the littoral zone among Hydroids and Polyzoa. With such a typically warm-water distribution (Zululand southwards), Flynn’s record ‘Ball Point, Cape Town’ is a little surprising, if this locality © really is in Table Bay. I have not been able to trace Ball Point on any chart. I have not taken Béhmia chelata either at Sea Point or at Melkbos Strand (both in Table Bay), but it is fairly common at low tide at St. James, False Bay. It is, however, a more frequent inhabitant of Alcyonarians, and nearly every piece of ‘Fan-coral’ (Gorgonia flammea) which I have examined in a fresh state, has one or more of these little Pycnogonids clasping its branches. Coloration. As a rule the animal is uniformly coloured, irrespective of the colour (white, yellow, pink, red) of the Alcyonarian on which it is found. Two lots, however, from the Keiskamma-Cape Morgan area are bicoloured and very striking. The ground colour is the usual amber or yellowish-brown with dark umber-brown on parts of the body and legs as follows: on the body the dark colour occupies the whole upper surface except the antero-lateral corners of the cephalic segment, ocular tubercle, and a patch behind it, the 4 dorsal tubercles, and the apical bosses on the crurigers; the abdomen has a dark streak on each side; on the legs the 1st coxa is dark, dorsally only in juvenile, but also ventrally in adult. The 2nd coxa is dark dorsally with a slight suffusion ventrally in adult; the femur and Ist tibia each with a dark 136 ANNALS OF THE SOUTH AFRICAN MUSEUM band in distal half on dorsal surface only. The pale spots on the 2nd coxae ($2) and the cement gland (¢) remain pale. One of these lots is recorded as being found on a ‘yellow Alcyonarian’, and the piece preserved with the Pycnogonid is Wrightella fragilis. In the preserved state this Alcyonarian is white with the nodes yellow or brownish, but whether the two colours are distinct in life is not recorded. Aberration. One ¢ (Somerset Strand, False Bay) has the ocular tubercle divided into two tubercles placed transversely; one of the eyes seems to be se pre nerqh CHAM abyAMEEUL At 4 MMe phe veWeiiertyatyye geen SUE ECT SP PEE er pre iyrtre ties Fig. 23. Béhmia tuberosa Mobius. Dorsal surface of femur of g showing pale patch (‘cement gland’), with a portion of the latter further enlarged; semidiagrammatic, minor reticulation omitted. present, but the others could not be traced. For rarity of abnormalities in Pycnogonids see Gordon, 1932. ‘Discovery’, Rep., vi, p. 131. Bohmia tuberosa Mobius Fig. 23 19025 J Mobius: loc) cit... p. 160, ple 2o esanlo ml Ae 1909. Schimkewitsch, loc. cit., p. 3, fig. 1 B and D (chelifer and palp). g—similar to chelata, but the ocular tubercle, antero-lateral processes of cephalic segment, median bosses on segments and those on apices of crurigers all very prominent and sharply conical, the median boss on 4th segment lower than the others. Sides of cephalic segment somewhat convergent. Surface reticulate and spinulose as in chelata. Chelifer and palp as in chelata, but in the latter the 7th joint is shorter than the 6th. Oviger as in chelata, but the pairs of spines on last 4 joints more numerous, resp. about 10, 8, 6,9; and roth joint without an unguiform process on lower apex. SOUTH AFRICAN PYCNOGONIDA 137 Legs longer and more slender than in chelata; 2nd coxa a little longer than either 1st or 3rd, femur a little longer than Ist tibia, about five times as long as broad, 2nd tibia a little longer than femur, rather slender and tapering so that tarsus is very little narrower than its apex, tarsus and propodus two and a half to three times in length of and tibia, propodus without heel, with numerous spines on lower margin, claw a little more than one-third (but not quite half) length of propodus (Mobius, fig. 14: half length of propodus). Surface reticulate and spinulose, with pale spot on dorsal apex of 2nd coxa of all legs. Femoral cement gland an elongate-oval pale patch on dorsal surface of femur in all legs. Genital pore on ventral apex of 2nd coxa of last leg only. Body (front margin of cephalic segment to base of abdomen) 6, proboscis 4, abdomen 3, 2nd leg (without claw) approx. 23 mm. (femur 6, Ist tibia 5, and tibia 7 mm.). Reddish-brown. Localities: Agulhas Bank, 154 metres 1 g (Mobius); off East London, 250-300 fath. 1 non-ovig. ¢ (S. Afr. Mus.). Remarks. The ‘Valdivia’ specimen was a ¢ (dorsal patches) 7 mm. in length including abdomen. There is an obvious discrepancy in the length of the claw on the legs in Mobius’s two figures; as fig. 14 shows a claw half length of propodus and therefore not so very different from that of the specimen above described, one must regard the long slender claws in the picture of the whole animal (fig. 13) as fanciful. With a body length of 7 mm. incl. abdomen, the 2nd leg of the ‘Valdivia’ specimen was 13 mm. long; the present specimen measuring 9 mm. incl. abdomen has the 2nd leg 23 mm. long. The remarks on the pale coxal spot and femoral patches and the position of the ¢ genital pore made under chelata apply here also. With the greater length of femur, the femoral patch is more elongate. The patch itself is paler than the rest of the integument, but there is a still paler line down the centre of it, as seen under a moderate magnification. This is not due to a slit-like opening, but to a double series of minute transverse ‘marks’ (fig. 23). Their real structure is doubtful, but they appear to be in the substance of the chitin as they are not obliterated by rubbing a needle-point over them either on the external or internal surface. (Schimkewitsch, loc. cit., p. 4, called them: Internal trabeculae.) Scattered over the pale area are a number of minute ‘glands’, smalier and more numerous than those in chelata. These can be removed by rubbing the internal surface. Gen. Achelia Hodge. ? 1838. Costa, Fauna Regn. Napoli. Arachn. Trach., p. 7. (Phanodemus.) 1864. Hodge, Rep. Brit. Assoc., xxxii, Not. misc., p. 102 and Ann. Mag. Nat. Hist. (3) xiii, p. 114. 1881. Hoek, Rep. H.M.S. ‘Challenger’, iii, p. 26. 1881. Dohrn, F. Fl. Golf. Neapel., iii, pp. 133, 225, 227, 228 (Ammothea, non Leach). 138 ANNALS OF THE SOUTH AFRICAN MUSEUM 1891. Sars, Norw. N. Atl. Exp. Pycnogonidae, p. 120. (Ammothea, non Leach.) 1908. Loman, Szboga Exp. monogr., xl, pp. 10, 11 and synopsis facing p. 19. (Ammothea, non Leach.) 1913. Bouvier, 2me Exp. Antarct. Franc., pp. 45, 46, 138. 1915. Calman, ‘Terra Nova’ Exp. Zool., ii, p. 56. 1917. Bouvier, Res. Sct. Camp. Monaco, li, p. 38. (Ammothea subg. Achelia.) 1927. Hodgson, D. Stidpol Exp., xix (zool. xi), p. 344. 1932. Gordon, ‘Discovery’ Rep., vi, p. 110. 1938. Calman, John Murray Exp., v, p. 160. 1938. Gordon, Austral. Antarct. Exp., C II, 8, p. 22. Body compact, segmented, but segments more or less fused. Crurigers contiguous or narrowly separated. Ocular tubercle near front margin of cephalic segment. Proboscis large, fusiform, directed forwards. Chelifer with single-jointed scape, and terminal more or less bifid joint, which in juvenile forms a small but functional chela. Palp 8-jointed, more or less geniculate, and and 4th joints longest. Oviger 10-jointed, in both sexes, 7th-1oth joints each with 1 (or 2) pinnate or doubly-serrate spines, without apical claw. Legs moderate, tarsus small, propodus strongly spinose on lower margin, claw strong, auxiliary claws moderate or strong. Genital pores on all legs in 9, on last two legs in @, at apices of conical processes. Femoral cement gland ¢ single, opening near dorsal apex. Remarks. Since Leionympha Mobius was shown to be the same as Ammothea Leach, the small, compact, shallow-water species formerly included in the latter genus, are now put into Achelia. Bouvier (1913 and 1917), however, uttered a warning against regarding this name as stable, suggesting (as did Dohrn) that Phanodemus was the earliest generic name for these forms. Key to the South African Species 1. Abdomen extending beyond last crurigers. quadridentata 2. Abdomen not extending beyond last crurigers brevicauda Achelia quadridentata (Hodgson) Fig. 24 1910. Hodgson, Schultze. Reise, iv, p. 223, figs. 2, 3 (palp, oviger). Body compact, widest in front, narrowing posteriorly, cephalic segment large, segments 3 and 4 fused, and abdomen apparently not marked off by a suture. Crurigers contiguous, each with a conical tubercle on upper apex. Proboscis as long as body with abdomen, broadly fusiform. Ocular tubercle on front margin of cephalic segment, conical, eyes distinct. Abdomen extending to end of 1st coxa of last leg. Chelifer scape undivided, the terminal joint representing the chela feebly bifid. Palp geniculate, 2nd and 4th joints longest, distal joints setose. Oviger SOUTH AFRICAN PYCNOGONIDA 139 5th joint longest, 7th, 8th and gth joints each with one doubly-serrate (pinnate) spine, 10th joint with 2 similar spines, one laterally and one apical; alike in the two sexes, but that of 3 rather stronger. Legs without processes, 1st coxa with a small setiferous tubercle on posterior part of distal margin dorsally, with a smaller one below it, propodus with 3 strong spines proximally on lower margin, and 2 moderately strong ones distally, auxiliary claws about half length of main claw. Genital pores ¢ at apices of rounded processes on ventral apices of 2nd coxae of last 2 legs; in 9 Fig. 24. Achelia quadridentata (Hodgson). a. dorsal view (4th joint of palp foreshortened). b. lateral view of 2nd coxa of last leg g. c. apex of femur with cement gland g. d. propodus and claws of leg. e. oviger g with distal joints further enlarged. on all legs, but not on projections. Femoral cement gland 3 small, behind the dorsal apical spine. Body incl. abdomen 1-5, proboscis 1-5 mm. Localities: Liideritzbucht (Hodgson); St. James, False Bay, littoral. 1 ovig. 6, 1 9 (K. H. Barnard, 1912); Table Bay, 5 fath. (University Cape Town Ecol. Surv.). Remarks. ‘The above description of the St. James’s specimens covers that given by Hodgson, and I think the identification of these specimens is correct. Hodgson’s specimens had 4 proximal spines on the propodus; the present specimens have only 3. 14.0 ANNALS OF THE SOUTH AFRICAN MUSEUM Achelia brevicauda (Loman) 1904. Loman, Zool. Fahrb. Abt. Syst., xx, p. 376, pl. 14, figs. 1-4 (3) Body compact, widest in front (but not so wide as in quadridentata), antero- lateral angles of cephalic segment subacutely angular (Loman: a strong conical tubercle), segments 3 and 4 fused, abdomen not marked off by a suture. Crurigers contiguous, a double setiferous tubercle (or two separate tubercles) at each antero-distal and postero-distal corner dorsally. Proboscis nearly as long as body, fusiform. Ocular tubercle on front margin of cephalic segment, rounded or bluntly pointed apically, ies distinct. Abdomen short, extending to end of last crurigers. Chelifer, palp and oviger as in quadridentata (cf. fig. 24). The 6th joint of oviger has a strong recurved spine proximally, as in A. Jangi Dohrn (1881. pl. 5, fig. 2). Fig. 25. Nymphopsis cuspidata (Hodgson). Legs without prominent processes, Ist coxa with 2-3 setiferous conical tubercles on both antero- and postero-distal corners dorsally, distal corners of 2nd and 3rd coxae each with 1-2 much smaller and inconspicuous tubercles, femur with 2 major spines and an apical one on upper margin, propodus with 3 strong spines on lower margin proximally and 2-3 less strong ones distally, auxiliary claws about half length of main claw. Genital pores ¢ on prominent ventral processes on last 2 legs. Femoral cement gland ¢ small, opening behind the apical spine (cf. fig. 24 c). Body incl. abdomen 1-25, proboscis 1, 3rd leg without claw 3:75 mm. Localities: Port Elizabeth. 1 non-ovig. 3 (Loman); Sea Point, Cape Town. 1 non-ovig. ¢ (K. H. Barnard, 1914). Remarks. ‘The pinnate spines on the distal joints of the oviger are described as ‘oak-leaf spines’ (Loman, et auct.) and are figured by Loman as more or less of this shape. Both in this and the previous species, however, the general outline is ovate, the basal serrations or pinnae being much longer than the distal ones. Even Dohrn’s figures (1881) do not do full justice to these spines. SOUTH AFRICAN PYCNOGONIDA I4!I Gen. Nymphopsis Hasw. 1885. Haswell, Proc. Linn. Soc., N.S.W., ix, p. 1025. 1908. Loman, loc. cit., p. 49. 1920. Flynn, Pap. Proc. Roy. Soc. Tasman. for 1919, p. 83. 1O22 Loman, Ak... <00l-,, xvA.Oy De 7. 1928. id., Tijdschr. Ned. Dierk. Ver. (3), i, Afl. 2, p. 39. 1929. Flynn, Mem. Queensl. Mus., ix, p. 256. no92))) Gordon, loc: cit. p: 129: Body compact, segments completely fused (the cephalic segment distinct in denticulata Gordon 1932). Crurigers separate. Ocular tubercle a short distance Fig. 26. Nymphopsis cuspidata (Hodgson). a. lateral view of ocular tubercle and dorsal spiniferous processes. b. chelifer of immature example. c. apex of scape and degenerate chela of adult. d. palp. e. oviger . behind front margin of cephalic segment. Proboscis large, fusiform, directed forwards. Abdomen with 2 pairs (genotype armata) or 4 pairs of spinose processes. Chelifer with 1-jointed (genotype armata) or 2-jointed scape, apex more or less cup-like, into which the small terminal joint can be withdrawn, in adult the latter is a feebly bifid joint, but in juvenile is a functional chela. Palp g-jointed, and and 4th joints longest. Ovyiger 10-jointed, in both sexes, distal joints with long spine-setae, but no serrate spines (except denticulata Gordon), without apical claw. Legs moderate, tarsus small, propodus strongly spinose on lower margin, claw strong, auxiliary claws well developed, or feebly, or absent. Genital pores on all legs in 9 on short projections; in ¢ on long curved processes on last 142 ANNALS OF THE SOUTM AFRICAN MUSEUM 2 legs. Femoral cement gland ¢g single, opening near dorsal apex. Eggs in several packets. Remarks. ‘The distribution of this curious and easily recognized genus embraces Australia, East Indies, Japan, Venezuela, Falkland Islands and South Africa. The body, abdomen, chelifers and legs are armed with spiniferous processes and/or branched or pinnate spines. Nymphopsis cuspidata (Hodgson) Figs. 25-27 1910. Hodgson, Schultze. Reise, iv, p. 221, fig. 1 (juv.). (Ammothea c.) 1923. Loman, loc. cit., p. 7, fig. D (g). (N. abstrusus.) Body together with the crurigers broadly oval in outline, in mid-dorsal line 3 elevated spiniferous processes. Cephalic segment with a spinose tubercle, sometimes more or less divided, on each antero-lateral corner. Crurigers longer than median width of body, contiguous, each with 2 spinose tubercles dorsally, one on antero-distal corner and a much larger and upstanding one on postero-distal corner. Proboscis shortly fusiform. Ocular tubercle high, cylindrical, tapering above the eyes to a conical point. Abdomen long, slender, with 4 pairs of spiniferous processes dorso-laterally, of which the gnd and 4th pairs are the largest, 2 small spines on apex, and 1-2 laterally. Chelifer scape 2-jointed, 1st joint with a spiniferous process on dorsal apex, end joint with 3 such processes of which the middle one is the largest, and 2-3 on distal margin, hand of chela in juv. with 2 spines, finger and thumb curved, inner margins with minute spaced denticles, in adult chela degenerate, represented by a feebly bifid non-spinose joint. Palp with a few long spines on 4th joint, and numerous long spines on 5th-gth joints. Oviger with long spines on 6th-8th joints, 9th joint with a single spine, roth joint with 2 spines. Legs with spiniferous processes arranged mostly in pairs or 2 series dorso- laterally, most numerous on ist and 2nd tibiae; Ist coxa with a stout spiniferous upstanding conical tubercle (like the postero-distal one on cruriger) medio- dorsally; 2nd coxa with ventral apex moderately protuberant on all legs in 9, © produced in a cylindrical process on first 2 legs in $ with long spine-setae all around its apex; produced in a long curved process on last 2 legs in § with long spine-setae only on the surface facing away from the body, the apex (with genital pore) minutely spinulose; femur slightly longer than 1st tibia, and latter slightly longer than and tibia; propodus slightly curved, with spaced single spines on upper margin, lower margin with stout spines of which the 3 proximal ones are very strong and prominent, claw strong, auxiliary claws stout, about half length of main claw. Femoral gland ¢ lying along dorsal surface of femur immediately under the cuticle, and opening between 2 simple spines near apex. SOUTH AFRICAN PYCNOGONIDA 143 Body and abdomen each about 1:5, proboscis 3, leg (without claw) 7-5 mm. Localities: Liideritzbucht (Hodgson); Cape Point Lighthouse (Loman); Sea Point, Cape Town, littoral (K. H. Barnard, 1912, 1914, 1919); False Bay, 12 fath.; Port Elizabeth, East London (S. Afr. Mus.). Remarks. ‘There can be no doubt that Loman’s specimen is the adult of Hodgson’s species, who stated that both his specimens were immature. As Hodgson said, the spinose nature of the animal was sufficiently characteristic to justify describing a new species even on such material. Fig. 27. Nymphopsts cuspidata (Hodgson). a. spine from cruriger and dorsal surface of Ist coxa. 6. pedunculate spine from leg. c¢. lateral view of 2nd coxa of leg 9. d. apex of 2nd coxa of Ist and and legs g. e. 2nd and 3rd coxae and femur of 4th leg g, with apex of 2nd coxa, and femoral cement gland further enlarged. £. propodus and claw of leg. He correctly remarked that the ‘branched’ spines were really cuticular prolongations bearing one or more movable spines. They should not be called ‘compound spines’ (Loman) or “pinnate spines’ (Flynn), at least not in this species; in korotnewi Schimk. the armature of the legs does appear (Loman, 1908. loc. cit., pl. 13, fig. 179) to be composed of pinnate spines. The spiny armature of the animals collects much foreign matter; unencum- bered specimens are rare, and dirty ones are difficult to clean. Loman described the proboscis as ‘slender’, but this adjective is scarcely applicable to the present specimens. The 9 from False Bay has the proboscis shorter than the chelifers and quite invisible in dorsal view; also the medio-dorsal processes on the body are very much elevated, their height (excluding the apical spine) almost equal to the length of the abdomen. 144 ANNALS OF THE SOUTH AFRICAN MUSEUM Recent collecting (1946, 1947) by the Zoology Department of the University of Cape Town shows that this species is fairly common in shallow water in some places in False Bay. Gen. Kyphomia Helfer 1938. Helfer, SB. Ges. Naturf. Fr. Berlin, 1937, p. 179. 1953. Stock, Beaufortia, iv, no. 35, p. 41. Body compact, last two segments fused. Crurigers widely separated. Ocular tubercle near front margin of cephalic segment, slender, elongate. Chelifer with well-developed scape, and rudimentary 2nd joint. Palp g-jointed, 2nd and 4th joints longest. Oviger 10-jointed, with a few serrate spines, no apical claw. Legs moderate, tarsus small, propodus spinose on lower margin, claw strong, auxiliary claws as long as main claw. Remarks. Stock (1953) states that Helfer’s types of this species have been lost, but from its likeness to an East Indian species he is convinced that Helfer’s species should be included in the genus Ammothella (1900. Verrill, Tr. Conn. Ac. Sci., X, p. 581, as subgen. of Ammothea: palp 1o- instead of 8-jointed). Kyphomia setacea Helfer Fig. 28a 1938. Helfer, loc. cit., p. 179, fig. 9 a-c. Body covered with small warts. Crurigers about equal to median width of body. Proboscis ovoid. Ocular tubercle longer in 9 than in 4, eyes pigmented. Abdomen slender, clavate. Chelifer with 1-jointed scape, 2nd joint knob-like with 2 apical points. Oviger with a few setae and serrate spines on last 4 joints (2-3 serrate spines on last joint). Legs with numerous long and short setae, femur in 2 swollen, in 3 nearly cylindrical, auxiliary claws as long as main claw. Femoral gland ¢ opening at apex of a slender process at end of femur. Length 3 mm., span 6-7 mm. Locality: Agulhas Bank, 35° 19’ S. 20° 12’ E., 126 metres (Helfer). Gen. Ainigma Helfer 1938. Helfer, loc. cit., p. 181 Body compact, segmented. Crurigers widely separated. Ocular tubercle near front margin of cephalic segment, high. Chelifer with well-developed scape and rudimentary 2nd joint. Palp g-jointed, 2nd and 4th joints longest. Oviger 10-jointed, apical joints with small serrate spines, terminal joint non- serrate. Legs moderate, tarsus one-third length of propodus, which is non- spinose, claw as long as propodus, no auxiliary claws. SOUTH AFRICAN PYCNOGONIDA 145 Remarks. Hedgpeth (1947. Smiths. Misc. Coll., cvi, p. 4) accepts and places this genus in the Ammotheidae. Stock (1952. Bull. Inst. Roy. Sci. Nat. Belge, XxVili, 14, p. 5; and 1953. loc. cit. infra) considers it a synonym of Ascorhynchus. A species of Ascorhynchus (A. glaber Hoek, 1881) was collected by H.M.S. Challenger between the Cape and Kerguelen (46° 46’ S. 45° 31’ E., 1,375 fath.). a b Fig. 28. a. Kyphomia setacea Helfer, 2 with distal joints of leg. 6b. Aznigma ornatum Helfer, lateral view, dorsal view of cephalic segment, distal joints of oviger, and distal joints of leg. (After Helfer.) Ainigma ornatum Helfer Fig. 285 1938. Helfer, loc. cit., p. 181, fig. 10 a-f. 1953. Stock, Beaufortia, vi, no. 35, p. 41, fig. 6. (Ascorhynchus 0.) 1953. id., Temminckia, ix, p. 304 (in key to species). Body covered with small warts, each segment with a conical medio-dorsal process. Crurigers about equal to median width of body, each with a conical upstanding process. Proboscis ovoid. Abdomen slender, clavate. Chelifer with 1-jointed scape, 2nd joint knob-like. Length 2:88 mm. Locality: Agulhas Bank, 35° 19’ 8. 20° 12’ E., 126 metres (Helfer). 146 ANNALS OF THE SOUTH AFRICAN MUSEUM Gen. Austroraptus Hodgson 1907. Hodgson, Nat. Antarct. Exp., 11, p. 54. 1915. Calman, ‘Terra Nova’ Exp. Rep., iii, p. 62. 1927. Hodgson, D. Stidpol Exp., xix (zool. xi), p. 349. 1932. Gordon, ‘Discovery Rep., vi, p. 114. 1944. id., B.A.N.Z. Aniarct. Res. Exp., v, p. 57- Body compact, segments fused or only indistinctly demarcated. Crurigers narrowly separated or contiguous. Proboscis flask-shaped, stout at base, narrowing to a point, directed or curved downwards. Chelifers small, chelate, rudimentary. Palp 6-jointed, the apical joint inserted laterally on the 5th and d Fig. 29. Austroraptus thermophilus Brnrd. a. dorsal view (palp bent downwards). 5b. lateral view. c. ventral view of proboscis. d. 1st and 2nd coxae of 4th leg ¢. projecting forwards or upwards (the apical joint: may be tripartite). Oviger 10-jointed, in both sexes, apical joints with simple non-serrate spines, no apical claw. Legs moderate, tarsus short, auxiliary claws moderate, small, or minute. Remarks. Previously known only from the Antarctic (3 species); the dis- covery of a species in the considerably warmer waters of South Africa is noteworthy. Appears to differ from the species of Tanystylum with 6-jointed palps only in the shape of the proboscis. Austroraptus thermophilus Brnrd. Figs. 29, 30 1946. Barnard, Ann. Mag. Nat. Hist. (xi) 13, p. 62. Body compact, all the segments fused, smooth; cephalic segment broad with a tubercle at each antero-lateral corner. Crurigérs contiguous, their SOUTH AFRICAN PYCNOGONIDA 147 length greater than median width of segments, each with a tubercle on antero- distal and postero-distal corners dorsally, a little more prominent in g than 9. Ocular tubercle conical, rather broad and low, eyes distinct. Proboscis markedly flask-shaped, broad proximally, rapidly narrowing to a downwardly curving apex. Abdomen elongate, extending beyond end of Ist coxa of last leg, horizontal. Chelifers vestigial, composed of a single conical projection with 2 apical setules, less prominent in 9 than in g. Palps robust, curving downwards, 4th Fig. 30. Austroraptus thermophilus Brnrd. a. ocular tubercle from right side. 6. palp. c. oviger g. d. femur and tst tibia of leg g. ¢. apex of femur with cement gland ¢. f/f. distal joints of leg. joint longest, angular in middle of upper margin, a pale or semitransparent oval spot indicating a gland (but orifice not traceable), 5th joint conical, setose, 6th ovate, inserted laterally on 5th and projecting forwards and upwards, setose. Oviger in § with 4th and 5th joints gently curved, distal joints feebly armed, 8th and oth each with a simple spine on inner margin, toth joint with 2 apical spines; in 9 similar but shorter and weaker, 4th and 5th joints straight. Legs 1st coxa with the antero-distal and postero-distal corners produced in a tubercle, the anterior one larger than the posterior and carrying I or 2 spines, both tubercles are more prominent in J than in 9; 2nd coxa with lower apex slightly but not prominently gibbous, genital pores in @ on all legs, in 3 only on last two; femur with upper apex produced, Ist and 2nd tibiae subequal (and subequal to femur) with low rounded spiniferous tubercles arranged more or less in 2 longitudinal dorso-lateral series; propodus with 3 large spines proxi- mally on lower margin, auxiliary claws well developed, nearly half length of 148 ANNALS OF THE SOUTH AFRICAN MUSEUM main claw. Femoral cement gland ¢ single, elongate, lying immediately below the cuticle, opening on distal surface of the apical projection. All the tubercles on the cephalon, crurigers and Ist coxae, and the nodulosities on the legs are microscopically hispid. Body incl. abdomen 3 (trunk and abdomen subequal), proboscis 1-25, leg without claw 8 mm. Localities: False Bay, 23 fath. 1 9; off Cape St. Blaize, 34 fath. 1 ovig. 3g; off Nanquas Peak (east of Algoa Bay), 63 fath. 2 non-ovig. g¢ (S. Afr. Mus.). Remarks. Distinguished from the genotype polaris by the contiguous crurigers, and from all the three Antarctic species by the longer abdomen and more markedly flask-shaped proboscis, shorter ocular tubercle, shorter tubercles on Ist coxae, and the distinct though not prominent nodulosities on the tibiae; the auxiliary claws also are stronger. Fam. TANYSTYLIDAE 1913. Schimkewitsch, ool. Anz., xli, p. 613. 1947. Hedgpeth, Smiths. Misc. Coll., cvi, p. 4. Octopodous. Body compact, segments fused or segmentation visible only laterally. Crurigers narrowly separated or contiguous, or fused. Proboscis large, fusiform, directed forwards. Chelifers absent in adult, sometimes rudiments in juvenile. Palps 4-6-jointed. Ovigers in both sexes, 10-jointed, without compound spines. Legs stout, tarsus very short, auxiliary claws present. Genital pores on all legs in 9, on last 2 in g. Eggs in several masses. Key to the South African Genera 1. Palp 6-jointed. Tanystylum 2. Palp 5-jointed. Discoarachne Gen. 7 anystylum Miers 1879. Miers, Philos. Trans. Roy. Soc. Lond., clxviii, p. 213. 1932. Gordon, Ann. Mag. Nat. Hist. (10), x, p. 87. 1Q924) Wd. Discovery mivep.. vino lle Body subcircular, unsegmented. Crurigers narrowly separated or contiguous, or fused. Proboscis stout, movable, directed forwards. Palp 6-jointed. Key to the South African [incl. Tristan d’ Acunha] species 1. Crurigers fused. ornatum 2. Crurigers separate. [ pfeffert | Tanystylum ornatum Flynn 19262) Plynn) loc, cit),;p. 33) es. 20, om. Crurigers fused, the divisions between them indicated by superficial markings; each cruriger with a pair of knobs distally on dorsal side. Ocular tubercle moderately high, rounded, surmounted by a short pointed process. Proboscis ovoid. Abdomen clavate, oblique. SOUTH AFRICAN PYCNOGONIDA 149 Chelifers represented by a pair of small setose processes. Palp 4th joint longest, 2nd and 6th subequal. Oviger in ¢ 4th joint curved, 5th longest, 8th at right angles to 7th, toth with 2 apical spines; in @ (as figured by Flynn) Ath and 5th joints subequal, 6th-9th joints each with 1 spine on outer apex, gth also with 2-3 spines on inner apex, 10th joint with 2 apical spines. Legs 1st coxae ending in 2 or more lobes dorsally, the three coxae subequal, femur and tibiae subequal, propodus without heel. Body with proboscis, excl. abdomen 1-8 mm. Locality: off west coast of Cape Peninsula, 25-30 fath. (Flynn). Remarks. ‘The oviger as represented in fig. 21a as that of a 9 seems rather remarkable, especially as regards the 10th joint with its 2 stout opposing spines (cf. Discoarachne brevipes 3). It may be noted that the latter species was taken in the same locality together with the TYanystylum specimens. Gen. Discoarachne Hoek 1881. Hoek, Rep. H.M.S. ‘Challenger’, iii, p. 74. 1901. Cole, ool. Jahrb. Abt. Syst., xv, pp. 243 sqq. 1904. Woman, ibid., xx, p. 383. 1908. id., Siboga Exp. monogr., xl, pp. 14 sqq. and synopsis facing p. 19. Resembling Yanystylum but palp 5-jointed. Genital pores in ¢ on last 2 legs, in 2 on all legs. Femoral cement gland ¢ a single pore near dorsal apex. Remarks. Loman (1904) erroneously stated that the crurigers were fused, and that the ¢ oviger was g-jointed. The latter statement was corrected by him in 1923 (loc. cit. infra). The genus is sometimes regarded (1909. Thompson, Cambr. Nat. Hist., iv, p- 535) as a subgenus of Tanystylum. Discoarachne brevipes Hoek Fig. 314, b leo) wiloek, loc. cit. pp. 74, 105, pl. 7, figs. 8-12. TOO Cole, loc. cit. p:) 243, pl..13. 1904. Loman, loc. cit., p. 379, pl. 14, figs. 7-11. 1908. id., loc. cit., p. 46 (incidentally mentions ‘commensals’ [? Rotifers] on the palps). 1910. Hodgson, Schultze. Reise, iv, p. 227. nee welkoman, Ark. Zool, xv79, p- 6, fig. C. nOZee Hibynn loc: cit...p4)35- 1951. Stock, Mem. Inst. Roy. Sci. Nat. Belge (2) fasc. 43, p. 3. Body oval, crurigers distinct and not fused though separated only by clefts. Proboscis somewhat fusiform. Ocular tubercle low, rounded, eyes distinct. Abdomen extending to end of last crurigers, rather slender, clavate, horizontal. Chelifers represented in young specimens by 2 small setiferous processes, sometimes 2-jointed (Loman, 1904, fig. 7a); in older specimens these tiny 150 ANNALS OF THE SOUTH AFRICAN MUSEUM stumps may be retained, but as often as not no trace of the chelifers is found. Palp slender, with first 2 joints short, 3rd longest, often swollen, 4th short. Oviger in § with 8th joint attached at right angles to 7th, roth with 2 apical and rather stout spines simulating a minute chela; in 9 normal, roth joint longer than oth, with several apical spines. b Gu a Fig. 31. a, b. Discoarachne brevipes Hoek, dorsal view, and goth and roth joints of oviger <. c, d. Pycnogonum forte Flynn, dorsal view with 3rd right leg to same scale (coxae in dorsal view, remainder of leg in postero-lateral view). Legs stout, the three coxae subequal, femur somewhat contorted, tarsus much narrower than end tibia, propodus with 3 stout spines at base but no distinct heel. Very minutely hispid. Femoral cement gland ¢ single, near the apex on dorsal surface. Body (tip of proboscis to tip of abdomen) 4-4°25 mm. Localities: Cape ‘Town, littoral (Hoek, Cole, Loman, Flynn); off west coast of Cape Peninsula, 25-30 fath. (Flynn); Cape Point Lighthouse, littoral (Loman); False Bay (probably littoral) (Hodgson); Port Natal (Durban) (Loman); Liideritzbucht (Stock); Port Nolloth, Saldanha Bay, Table Bay, west and east coasts of Cape Peninsula, Kleinmond (mouth of the Palmiet River, Cape Province), Plettenberg Bay, East London, all littoral (S. Afr. Mus.). Remarks. This is a very common Pycnogonid on both sides of the Cape Peninsula among weeds, Hydroids, anemones, etc., between tide marks. SOUTH AFRICAN PYCNOGONIDA I51 Ovigerous jg have been collected in February, July and September to November; probably breeding occurs throughout the year. Stock’s record is presumably based on the ‘Mercator’ collections, although he makes no mention of this species in the systematic part of his paper. The specimens collected by Dr. Schultze, and recorded by Hodgson, came from False Bay, not from Liideritzbucht. ; Fam. PycNOGONIDAE 1908. Loman, Siboga Exp. monogr., xl, conspectus facing p. 19. 1909. Thompson, Cambr. Nat. Hist., iv. p. 539. 1913. Bouvier, 2me Exp. Antarct. Franc., pp. 46, 149. 1927. Calman, Trans. Zool. Soc. Lond., xxii, pp. 408, 410. 1947. Hedgpeth, Smiths. Misc. Coll., cvi, p. 5 (in list of families and genera). Decapodous (Pentapycnon) or octopodous. Body segmented or the segmen- tation more or less suppressed. Proboscis projecting forwards or more or less bent under body. Chelifers reduced or absent, if present either chelate in adult or only injuvenile. Palps usually absent. Ovigers present in ¢ only, 6-9-jointed, with apical claw, but no spines. Legs without auxiliary claws (sometimes vestiges). Genital pores on last leg only in both sexes. Eggs in a single cake- like mass. Remarks. Several genera which are in one respect or another aberrant have been associated with the genus Pycnogonum in this family. Gen. Pycnogonum Briin. 1764. Briinnich, Entomologia, p. 84. 1881. Dohrn, F. Fl. Golf. Neapel, ii, p. 202. 1902. Mobius, D. Tiefsee Exp., ii, p. 193. 1908. Loman, loc. cit., pp. 5-9, conspectus facing p. 19 and p. 34. 1913. Bouvier, loc. cit., pp. 150, 156. coz one ice Ant...Scls Nat Zool. (10), Vs ps 113: Octopodous. Body with segments 3 and 4 often close together and junction difficult to trace, or fused. Crurigers separate or contiguous. Proboscis movable, stout. Chelifers absent (but see: P (?) claudum Loman, 1908). Palps absent. Ovigers in § only, sometimes absent even in ovigerous specimens. Legs with auxiliary claws vestigial or absent. Genital pores on 2nd coxa of last leg in both sexes, in 9 on dorsal surface, in ¢ on ventral surface. Remarks. Dohrn (loc. cit., p. 203) stated that the genital pores in 9 were dorsal, and this applies to his puszllum; but in his nodulosum he described them as being ventral, albeit near the posterior margin. Key to the South African Species I. No auxiliary claws. A. Junction of 3rd and 4th segments not evident, these two segments more or less fused. Crurigers contiguous. 152 ANNALS OF THE SOUTH AFRICAN MUSEUM 1. 2nd tibia at least half length of 1st tibia. Proboscis cylindrical. a. 4 medio-dorsal tubercles (only 1 on cephalic segment). cataphractum b. 5 medio-dorsal tubercles (2 on cephalic segment). Sorte 2. and tibia scarcely one-third length of 1st tibia. Proboscis tapering. portus B. 3rd and 4th segments quite distinct, and all crurigers separate. Ocular tubercle very small. microps II. Small auxiliary claws present. pusillum Pycnogonum cataphractum Mobius 1902. ) Mobius; loc cit, ps 194, plaso, teenie Integument reticulate. Body stout, cephalic segment wider than following segments, all of which are (according to the figure) distinct; 4 medio-dorsal conical tubercles, one on each segment. Crurigers contiguous, each with a conical boss or process on dorsal apex. Ocular tubercle conical, rounded, shorter than the tubercle behind it, eyes distinct. Proboscis stout, somewhat narrowed in distal half, apex truncate. Abdomen extending slightly beyond last crurigers, apically truncate. Legs with conical or digitiform processes, one on dorsal apex of Ist coxa, 4-5 on femur, 1-2 on ist tibia, 1 on 2nd tibia, tibiae subequal, each slightly shorter than femur, claw one-third length of propodus (drawn too long in the figure), no auxiliary claws. Smaller specimen: body 9, proboscis 4, abdomen 1 mm.; larger (incom- plete) specimen; proboscis 5 mm. Locality: St. Francis Bay, shallow water (Mo6bius). Remarks. P. tumulosum Loman, 1908, East Indies, has somewhat similar nodose legs, but has 2 medio-dorsal tubercles on cephalic segment and none on 4th segment. The smaller of the ‘Valdivia’ specimens had eggs attached to the ventral surface but no ovigers. Pycnogonum forte Flynn Ries ouch 19205) Flyin loc. (cits) ps Sila ese To ml@n(eyE Very close to cataphractum. Third and 4th segments fused, last pair of crurigers fused; 5 medio-dorsal, more or less digitiform, tubercles of which 2 are on cephalic segment and one on each of the other segments. Abdomen longer, extending considerably beyond last crurigers. Legs in holotype less conspicuously tubercular than in cataphractum. In the present specimen the legs are conspicuously tubercular; a conical tubercle on upper apex of Ist coxa, 1 or 2 low rounded tubercles dorsally on both 2nd and grd coxae; on femur 2 rows of rounded tubercles increasing in size distally, with a large conical tubercle between the apical pair; Ist tibia with a large tubercle near base followed by 2 rows of rounded tubercles; 2nd tibia similar, but the tubercles less prominent; propodus feebly granulose on upper surface, finely setulose on lower margin, claw about one-third length of propodus, SOUTH AFRICAN PYCNOGONIDA 153 auxiliary claws absent. Genital pores in present specimen on dorsal surface of 2nd coxa of last leg. Total length 10:5 mm., proboscis 5 mm., abdomen 2 mm. (Flynn’s specimen resp. about 10°3, 3°8, 1°5). Localities: off Gneka River*, 1 2 (Flynn); off Great Fish Point, 49 fath. 1 9 (S. Afr. Mus.). Remarks. ‘This species is so extraordinarily like cataphractum that, when more material is available, it will probably be found to be conspecific. Fig. 32. Pyncnogonum portus Brnrd. Dorsal view; lateral view, with process further enlarged; lateral view of leg, with dorsal surface of femur, and tarsus and propodus further enlarged. Mobius may possibly have overlooked the 2nd tubercle on the cephalic segment, but however that may be the longer abdomen in forte seems to be a definite differential character, unless it be sexual. Flynn’s figure gives the impression of shallow depressions between the crurigers and the body proper; in the present specimen they are much deeper. And though the 4th segment is fused with the grd the suture is visible dorsally. The present specimen has a very small and indistinct granule on the hinder slope of the ocular tubercle. Pycnogonum portus Brnrd. Fig..92 1946. Barnard, Ann. Mag. Nat. Hist., (xi) 13, p. 62. Integument reticulate. Body stout, cephalic segment very broad, following segments rapidly decreasing in width, junction of 3rd and 4th difficult to trace dorsally; 5 medio-dorsal tubercles, 2 on cephalic segment, one on each of the other segments, each tubercle ornamented with a number of minute conical * See p. 85 footnote. 154. ANNALS OF THE SOUTH AFRICAN MUSEUM or digitiform processes. Near hind margin of cephalic and of and segment 2 conical tubercles on either side of the medio-dorsal tubercle, and on 3rd segment one similar tubercle flanking the medio-dorsal one; several additional minute granules or projections, resembling those on the medio-dorsal tubercles, scattered over the integument. Crurigers contiguous, each with an ornamental tubercle on upper apex. Ocular tubercle rather low, rounded, eyes distinct. Proboscis stout at base, rapidly tapering to a narrowly truncate apex. Abdomen extending beyond last crurigers, ornamented like the medio-dorsal tubercles and with a small tubercle on dorsal apex. Legs stout, nodulose and granulose with granular and digitiform outgrowths similar to those on body, Ist coxa with a tubercle on dorsal apex, femur rather strongly convex on proximal ventral surface and on distal dorsal surface, on the latter several outgrowths of various shapes, some bearing a spinule, and also a large prominent tubercle on apex, Ist tibia a little shorter than femur, with a low tubercle or boss proximally on dorsal surface, and numerous out- growths, 2nd tibia very short, scarcely one-third length of Ist tibia, tarsus spinulose on lower surface, propodus slightly tapering, granulose, with spinules on lower margin, claw about one-third length of propodus, no auxiliary claws. Genital pores on dorsal surface of 2nd coxa of last leg in 9; not traceable on the ovigerous specimen which is presumably a 6. Total length 3-5 mm., body plus abdomen 2:75 mm. Locality: Port Elizabeth, littoral. 1 9, 1 ovigerous § (University of Cape Town Ecological Survey, 1936). Remarks. ‘The integumentary ornamentation, and the shortness of the 2nd tibia are the noteworthy features of this species. Remarkable also is the complete absence of ovigers in the specimen which carries a flat cake-like mass of eggs. The eggs are loosely cemented to the ventral surface of the body, and the specimen is presumably a dg, especially as no distinct and large genital pores, as are easily observed in the second specimen, can be traced. So far as I am aware the absence of ovigers in ovigerous specimens had not been observed except by Mobius (1902, loc. cit.) in his description of cata- phractum. Are the ovigers never developed in some species? Or does a male before he is fully developed (structurally and perhaps also sexually) have to carry as best he can the packet of eggs which a female may dump upon him? Pycnogonum microps Loman 1904. Loman, Zool. Jahrb. Abt. Syst., xx, p. 378, pl. 14, figs. 5, 6 (9) Integument reticulate and granulose. Body stout, the segments especially 3rd and 4th distinct, cephalic segment not greatly wider than the others; 4 medio-dorsal tubercles, two (a smaller anterior and a larger posterior one) on cephalic segment, and one on each of segments 2 and 3. Crurigers almost as long as their segments, not contiguous, each with a small tubercle near dorsal SOUTH AFRICAN PYCNOGONIDA 155 apex. Ocular tubercle very small, eyes minute. Proboscis very stout, cylindrical, apex truncate. Abdomen extending slightly beyond ends of Ist coxae of last legs, apically truncate. Legs stout, granulose but without tubercles or bosses; 2nd tibia nearly as long as Ist, no auxiliary claws. Body 3, proboscis 1:5, abdomen -5 mm. Locality: IUlovo or Isipingo (Natal), littoral. 1 Q (Loman). Remarks. Quite different from the other South African species in the form of the body. Loman’s figure 6, showing the body in profile, does not seem to correspond with fig. 5, as it shows only the smaller medio-dorsal tubercle on the cephalic Fig. 33. Pycnogonum pusillum Dohrn. Lateral view; portion of integument; rst leg. segment, and a transverse ridge culminating in the median boss on each of the following segments. The intersegmental sutures seem to be wrongly indicated. Pycnogonum pusillum Dohrn Fig. 33 1881. Dohrn, Fauna Flora Golf. Neapel., iii, p. 207, pl. 16, figs. 4-8. 1953. Stock, Bull. Inst. Roy. Sci. Nat. Belge, xxviill 14, pp. 1-3, and 5, fig. 1. Integument with numerous, but scattered granules crowned with minute spinules. Body stout, segments distinct, cephalic segment not wider than the others; 3 medio-dorsal prominences, one each on the cephalic and the two following segments. Crurigers not contiguous, but angle of divergence between and and grd greater than that between the others; each with a prominence near dorsal apex. Ocular tubercle as large as the body prominences, eyes well developed. Proboscis very stout, slightly tapering to the truncate apex. Abdomen extending slightly beyond end of 1st coxae of last legs, apically rounded. No ovigers. I 56. ANNALS OF THE SOUTH AFRICAN MUSEUM Legs stout, granulose like the rest of the integument, without tubercles or bosses, but dorsal apex of femur somewhat gibbous; and tibia nearly as long as Ist; auxiliary claws small but distinct. Genital pore not traceable. Body 2, proboscis -75, abdomen -5 mm. Locality: False Bay (Cape), 8 fath. (Univ. Cape Town Ecol. Surv., Aug. 1952, one specimen). Distribution. Gulf of Naples (Dohrn); Santa Maria Bay, Angola (13° 25’ S.). (Stock.) Remarks. After its original description seventy years elapsed before this species was rediscovered; and then far away from its previously known habitat. It is strange that a couple of years later another example should be collected, and even farther afield. That is assuming that the identification of the present specimen is correct. ‘There is no justification for a n.sp.; the single specimen can be compared only with pusillum, and only one feature requires comment. Dohrn used the words ‘betrachtlicher Hocker’ and ‘betrachtlicher Buckel’ in describing the projections on the segments and crurigers, but unfortunately omitted to indicate these projections in his figure (as Stock pointed out). I would use a similar term, boss, projection or prominence, in preference to ‘tubercle’ for the present specimen. Stock, however, used the word ‘tubercle’, and his figure gives the impression that these prominences are in fact small and well- defined tubercles. Dr. Stock, to whom a sketch of the Cape specimen was sent, considers that it should be identified as pusillum; and states that the ocular tubercle and body prominences are subject to considerable variation in the genus Pycnogonum. INCERTAE SEDIS Gen. Queubus Brnrd. 1946. Barnard, Ann. Mag. Nat. Hist. (xi) 13, p. 63. Octopodous. Body distinctly segmented, anterior part of cephalic segment produced considerably in front of first crurigers. Proboscis movable, elongate, slender, tapering. Chelifers and palps absent. Oviger ¢ 10-jointed, with apical claw (2 unknown). Legs strong, but of good length proportionately to body, end coxa longer than either ist or 3rd, tarsus short, propodus stout, with heel and strong spines on lower margin, claw strong, no auxiliary claws. Femoral pores on ventral apex of and coxa of last 2 legs. Eggs carried in a single mass. Remarks. Resembling Pycnogonum in the absence of chelifers and palps, but differing in the armature and subchelate shape of the propodus of the legs. The absence of the 2 makes the position and relationships of the genus uncertain. Pigrogromitus Calman, 1927, is another genus with some remarkable features and of uncertain position. ‘The name of the present genus is taken from the same passage in Shakespear. SOUTH AFRICAN PYCNOGONIDA 157 Queubus jamesanus Brnrd. Fig. 34 1946. Barnard, loc. cit., p. 63. g — In general resembling Pycnogonum crassirostre, but legs relatively longer. Integument smooth. Hind part of cephalic and 4th segments, and anterior part of 2nd and 3rd segments raised into a prominent conical boss. Crurigers Fig. 34. Queubus jamesanus Brnrd. Dorsal view 3 (gap between 3rd and 4th crurigers drawn too wide, abdomen foreshortened and drawn in true length alongside); lateral view, with oviger, 6th and roth joints of latter further enlarged; tarsus, propodus and claw of leg. separated by half to three-quarters their own width (gap between 3rd and 4th in the figure is too wide), their length less than median width of segments. Ocular tubercle large, conical, not very high, eyes distinct. Proboscis long, narrowing rapidly, the distal two-thirds slender, cylindrical. Abdomen deflexed, clavate, with pointed apex. Oviger 10-jointed, 4th and 5th joints longest, 5th curved, 6th ovately expanded, inner margin with a strong bifid spine and a simple spine, 7th-1oth joints with minute spinules, apical claw short. Legs smooth, 2nd coxa longer than either 1st or 3rd, but not quite equal to the other two combined, femur subequal to the three coxae combined, Ist and and tibiae subequal, each shorter than femur, tarsus strongly spinulose on 158 ANNALS OF THE SOUTH AFRICAN MUSEUM — lower surface, propodus curved, with heel carrying 6-7 strong spines, with other spines on lower margin decreasing in size, lower apex angularly promi- nent, spinulose, claw strong, reaching to the heel on propodus, no auxiliary claws. Base of proboscis to base of abdomen 4 mm., proboscis 3 mm., leg without claw 9 mm. Locality: St. James, False Bay, low-tide. 1 ovig. 3 (K. H. Barnard, 1912). Remarks. The single mass of eggs was carried by the right oviger only, the left being folded up alongside. Ol ise e The ANNALS OF THE SOUTH AFRICAN MUSEUM are issued in parts at irregular intervals as material becomes available. As far as possible each volume is devoted exclusively to a particular subject (Zoology, Botany, etc.). Two or more volumes may be in course of — publication concurrently. Most of the Geological and Palaeontological papers are issued in conjunction with the _ Geological Survey of the Union of South Africa. Some volumes and parts are out of print, and others are bal sold as parts of a set, or vee respectively. The prices of parts published prior to 1940 have been increased. Out of print: Vols. I, II, V (Parts 1, 2, 9), VII, ve IX vee = XII (Part e XU, XXIV (Part 2), XXXI (Parts 1, 2, 3). im Vol. . III. 1903-1905 Zoology .. os Sais - <5 (apg eee IV. 1903-1908 Palaeontology .. ae V. 1906-1510 Geology, Palaeontology, Zoology, Anthropology VI. 1g08-1910 Zoclogy = IX. 1911-1918 Botany (esl Part 1) X. &«Ig1i-1914 Zoology a ois XI. sgt1-1918 Zoology a3 A oe XII. 1913-1924 Palaeontology sind Geology .- (excl. Part 7) XT. 1913-1923 Archaeology and Zoology Loe tte 5 Ae ere XIV. 1915-1924 Zoology oe Bre ge Meee XV. 1914-1916 Zoology . 4 x See as Se XVI. 1917-1933 Botany : an XVII. 1917-1920 Zoology .. J XVIII “1921 Zoology. .. XIX. 1924-1925 Zoology .. ae =a — : XX. 1924-1926 Zoology .. ae se See ae se s XXII. 1925-1927 Zoology .. ate Re a eee ae XXIII. 1925-1926 Zoology ee Be a XXIV. 1929-1938 Arthenpolos ana Eshnology . (excl. Part 2) XXV. 1927-1928 Zoology .. oe : oo as eee XXVI. 1928 Zoology = > oe ae ae a | XXVIT,. 1929 Anthropology eee as Se oie “6 XXVIII. 1929-1932 Palaeontology os eS sai pee XXIX. 1929-1931 Zoology .. a ae see vee 1G 13 0) 10 ‘6 6 ~ : 5 ° ‘ COMOMMOCODOMIO0O DOO ANNALS SOUTH AFRICAN MUSEUM - VOLUME XLI - PART IV, containing :-— 8. Les Paguristes des cétes occidentales et méridionales d’ Afrique. Par JACQUES Forest. (Avec planche IV et 70 figures dans le texte.) ISSUED JUNE 1954. PRICE 155. PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY THE RUSTICA PRESS (PTY.) LIMITED, COURT ROAD, WYNBERG, CAPE aee Me mee at Pena Ra aid 8. Les Paguristes des cotes occidentales et méridionales d’ Afrique. par JAcquEs ForREstT Muséum National d’Histoire Naturelle, Paris (Avec planche IV et 70 figures dans le texte) Ie INTRODUCTION Sur les cinquante-et-une espéces ou varietés de Paguristes mentionnées par Alcock dans son inventaire des Pagurides du monde (1905), trois seulement appar- tenaient a la faune ouest-africaine et il n’y en avait qu’une qui fut signalée pour PAfrique du Sud. Encore dans ces quatre cas, ne connaissait-on que les types! On aurait pu croire a cette époque que le genre Paguristes, florissant dans lIndo- Pacifique et dans la mer des Antilles, n’était que sporadiquement et fort pauvrement représenté dans |’Atlantique africain. La description, depuis lors, de trois nouvelles espéces ne modifiait guere les conclusions précédentes: il ne s’agissait toujours que de quelques individus. Aujourd’hui, aprés avoir dépouillé un matériel abondant et d’origine variée, il faut bien constater que la pauvreté de la faune ouest et sud-africaine en Paguristes n’était qu’apparente: il y a peut-étre en certain point de ce littoral plus d’espéces et plus d’individus appartenant a ce genre qu’en aucune autre région du monde. Aux sept espéces connues, s’ajoutent maintenant neuf especes nouvelles. Paguristes mauritanicus dont on ne connaissait que le type a été récolté en plus de vingt-cing stations et les collections en renferment des centaines d’exemplaires. La liste des espéces n’est d’ailleurs certainement pas close, car bien des régions et bien des niveaux restent inexplorés. A considérer la carte de repartition (fig. 69) il semblerait que c’est autour de Dakar que les Paguristes sont les plus nombreux et les plus variés; il est peu probable en effet que des dragages opérés en d’autres points puissent ramener un échantillonnage de faune pagurienne plus intéressant que celui du 18 janvier 1952, en Baie de Rufisque: les quelques centaines de Pagures récoltés—de petite taille puisqu’ils occupaient un volume total de moins d’un litre- appartenaient a quatorze espéeces différentes dont quatre de Paguristes. On doit cependant tenir compte du fait que cette région de Dakar, depuis la cote jusqu’a une profondeur de 100 a 120 métres, est l’une des mieux connues du littoral ouest- africain; il ne faudrait pas conclure, au vu des vastes blancs qui figurent sur notre carte de répartition, que les Paguristes sont totalement absents des cotes du Cameroun ou du Sud-ouest africain par exemple. Ces blancs signifient surtout que l’on se trouve en présence de zones pratiquement inexplorées ou des recherches ultérieures permettront sans doute de retrouver des espéces signalées ailleurs ou révéleront des formes nouvelles. Si on considére d’autre part que beaucoup de Paguristes vivent a 159 VOL. XLI. PART 4. AUR S 1994 160 ANNALS OF THE SOUTH AFRICAN MUSEUM des profondeurs assez grandes—Alcock estimait que 30% des espéces de ce genre se trouvent a un niveau inférieur a 100 brasses (180 métres environ)—on peut s’attendre a en découvrir de nouvelles dans |’Atlantique africain lorsqu’on aura la possibilité de draguer en-dessous de 120 meétres, limite de la plupart des récoltes effectuées a ce jour. Le matériel provenant de profondeurs plus grandes est fort peu abondant et provient pour la plus grande part des campagnes du ““Travailleur’”’ et du ““Talisman”’. La grande majorité des Paguristes intertropicaux d’Afrique occidentale ont un point commun: a l’exception de P. oculatus var. rubropictus, variété d’une espéce méditerranéenne et de P. marocanus qui vit a une profondeur plus grande que les autres, tous présentent une anomalie de lappareil génital femelle; lorifice sexuel est unique et situé sur la coxa gauche; il n’y a pas trace d’orifice droit. Dans le genre Paguristes nous n’avons pour l’instant retrouvé cette disposition de l’appareil génital femelle que chez des espéces de la Mer Rouge; ceci n’est pas le seul point commun entre les Paguristes d’Afrique occidentale et ceux de Mer Rouge dont la systématique embrouillée par Nobili est entierement a revoir: P. hispidus, P. micro- phthalmus et P. rubrodiscus—les trois espéces aux écailles oculaires trés allongées— sont plus proches de P. jousseaumei, décrit par Bouvier d’Aden et de Suez, que d’aucune autre espéce ouest-africaine et cette constatation s’ajoute a celles que nous avons eu l’occasion de faire a propos d’autres Paguridae. Certains Diogenes du Golfe de Guinée existent aussi en Mer Rouge, ou y sont représentés par des formes © voisines (Forest, 19524, p. 13). Chez d’autres Paguridae appartenant ceux-la a la sous-famille des Eupagurinae, il n’existe aussi qu’un orifice génital femelle, situé sur la coxa de la troisieme patte thoracique gauche comme chez les Paguristes; il s’agit des genres Sympagurus et Parapagurus. Il est intéressant de noter que dans ces deux genres, que lon considére en général comme bien éloigné des Paguristes, le male posséde également deux paires de pléopodes en général modifiés en gonopodes (voir p. 163). | | Au point de vue biogéographique on remarque qu’il existe sur la cOte occi- dentale d’Afrique, entre les Tropiques, un groupe de Paguristes dont on ne trouve aucun représentant ni au nord ni au sud, ni en une autre région du monde. _L’examen de la carte de répartition (fig. 69) montre que le genre est représenté par dix espéces entre !’Equateur et le Tropique du Cancer. Parmi ces dix espéces il en est quatre que I’on retrouve au sud de l Equateur, P. oculatus var. rubro-pictus (si la détermination d’Odhner est exacte), P. mauritamicus, P. hispidus et P. virilis. Une cinquieme espéce, la seule qui serait propre a cette région, est fort proche de P. virilis: c’est P. skoogt. Ainsi, dans le cas ot |’abondance et la diversité des Paguristes au large de Dakar ne seraient pas seulement dues a une meilleure connaissance de ce point du littoral, on pourrait admettre que la céte sénégalaise constitue le centre de dispersion des espéces intertropicales ou tout au moins la région ou le genre a rencontré les con- ditions de vie les plus favorables. Les Paguristes d’Afrique occidentale paraissent rechercher les fonds coquilliers, caillouteux ou vaseux, de préférence aux fonds de sable fin. On ne les rencontre guére dans la z6ne intercotidale, mais surtout a partir de quelques métres jusqu’a cinquante métres environ. I] semble que, pour une espéce au moins, il y ait une PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 161 relation entre la latitude et la profondeur a laquelle on la trouve: Paguristes mauri- tanicus a été capturé entre 5 et 30 métres en A.O.F. et entre 50 et 100 métres au Congo. P. marocanus qui n’appartient pas au groupe intertropical proprement dit a été dragué entre 115 en 290 metres, mais il n’est connu que par un trés petit nombre d’exemplaires et il est possible qu’on le rencontre a d’autres niveaux. Pour l’Atlantique africain intertropical, nous avons disposé d’un matériel abondant dont la base était constituée par la collection du Muséum d’Histoire Naturelle, avec notamment les spécimens recueillis par le ““Travailleur’’ et par le “Talisman”. Nous avons également trouvé de nombreux Paguristes parmi les Pagures rassemblés au cours des derniéres années par les zoologistes de l'Institut Frangais d’Afrique Noire et dans la collection qui nous a été confi¢e par I’Institut Royal des Sciences Naturelles de Belgique (Croisiére du “Mercator” et Expédition “M’Bizi’’). D’autres échantillons ont été récoltés par MM. E. Postel et R. Sourie de Dakar. La répartition géographique des espéces a été établie en tenant compte également des spécimens dragués au large de la Gold Coast par R. Bassindale, lesquels seront étudiés en détail avec les autres Pagures de méme provenance. Nous n’avons pas rencontré de difficulté particuliére dans l’étude des Paguristes de cette région et ceci surtout parce que nous avions un assez grand nombre de specimens de la plupart des espéces. Toutes—a l’exception de P. difficilis Forest qu'il est assez malaisé de distinguer de P. mauritanicus Bouvier sans examiner les pléopodes sexuels du male—sont relativement faciles a déterminer a l’aide de caractéres bien apparents. Notons cependant qu’ils sont presque tous revétus d’une abondante pilosité constituée par de longs poils plumeux; ces poils retiennent de fines particules de vase entre leurs barbules et cachent les régions du corps qui sont précisément les plus caractéristiques; aussi est-il nécessaire le plus souvent de nettoyer les spécimens a déterminer a l’aide d’une brosse fine. Nous avons été embarrassé par le cas de Paguristes skoogi Odhner: c’est la seule espéce déja décrite dont nous n’ayons pas examiné de spécimens. Le type, contraire- ment a ce que nous espérions, ne se trouverait ni a Stockolm ni a Géteborg. Notre P. virilis en est certainement fort proche, mais ne répond pas exactement a la description d’Odhner, ce qui nous interdit d’identifier les deux espéces pour Pinstant. Les Paguristes d’Afrique du sud—a l’exception du type de P. gamianus de H. Milne-Edwards conservé au Muséum de Paris—nous ont été communiqués par le Dr. K. H. Barnard, et appartiennent au South African Museum ou au Département de Zoologie de l’Université de Cape Town. Ces espéces nous réservaient des difficultés bien plus grandes que les précédentes. Nous avons tout d’abord hésité a établir l’espéce nouvelle décrite ici sous le nom de P. barnardi, en raison de limpossibilité de la distinguer du P. gamianus H. Milne-Edwards d’aprés l’aspect de la carapace et des appendices céphaliques. Cependant des différences marquées dans la pilosité et l’ornementation des pattes ambulatoires et dans la taille, ainsi que le nombre assez grand de spécimens dont 162 ANNALS OF THE SOUTH AFRICAN MUSEUM nous disposions et que l’absence de formes intermédiaires, nous ont finalement amené a penser qu'il s’agissait réellement de deux espéces distinctes. Une seconde difficulté s’est présentée avec le spécimen décrit a la suite de P. gamianus sous le nom de Paguristes sp. C’est un individu male d’une taille supér- ieure a celle des plus grands P. gamianus, dont il différe sur un certain nombre de points. Ces différences sont-elles liées a l’Age du spécimen en question? Quoi qu’il en soit il ne nous a pas paru opportun d’en faire une nouvelle espéce. Enfin, si nous décrivons sans hésitation P. macrotrichus sp. nov. d’aprés un spéci- men femelle déterminé—a tort—par Stebbing comme P. gamianus, nous ne sommes pas absolument certains que le Paguristes male plus petit, et d’aspect assez différent, qui figure ici sous le nom de P. agulhasensis sp. nov., n’appartienne pas a la méme espece. I] subsiste ainsi, dans la systématique des Paguristes sud-africains, quelques points douteux, qui concernent des spécimens isolés. Ils ne seront éclaircis que lorsqu’on aura récolté un matériel plus nombreux, condition essentielle d’une systématique valable, surtout lorsqu’on a affaire a des animaux aussi variables que des Paguridae. Ces espéces ne présentent guere d’affinités avec celles que l’on trouve entre les Tropiques. En ce qui concerne leur répartition géographique le fait le plus frappant qui apparait lorsqu’on considere la figure 70, c’est la distribution des deux especes parentes, P. gamianus et P. barnardi: la premiére est surtout representée sur la cote occidentale d’Afrique du Sud ou s’exerce l’influence des courants froids, alors que la seconde n’a été trouvée que sur la céte sud, baignée par le courant chaud des Aiguilles. On posséde peu de renseignements écologiques sur les espéces sud-africaines ; la plupart des échantillons proviennent de la région de la zone intercotidale ou des faibles profondeurs, sauf le P. agulhasensis dragué par 55 metres et le P. macrotrichus, ramené de 90 meétres. En raison de importance que présentent—comme nous lavons indiqué plus haut—les pléopodes sexuels males dans la systématique du genre Paguristes nous avons fait précéder l’étude proprement dite des espéces ouest-et sud africaines d’un chapitre consacré a la morphologie et a la physiologie de cet appareil génital externe que nous avons comparé a celui de quelques Anomoures. La courte diagnose du genre Paguristes est suivie d’un tableau de détermination valable pour les espéces qui nous intéressent. La description de chacune d’elles est suivie de quelques remarques sur sa position systématique et sur sa répartition. Elles ont été classées suivant l’ordre géographique nord-sud et ouest-est. Les tailles minima et maxima ont été données pour chaque échantillon: nous avons choisi comme dimension la longueur de la carapace (l.c.) mesurée depuis la pointe du rostre jusqu’au bord postérieur de la carapace, ce qui permet des comparaisons beaucoup plus significatives que la longueur totale. Les abréviations utilisées sont p. 1, p. 2, p. 3, etc. . . . pour les chélipédes et les appendices thoraciques suivants, et pl. 1, pl. 2, pl. 3, etc. . . . pour les appendices abdominaux. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 163 II. Les APPENDICES SEXUELS DES Paguristes MALES L’un des caractéres les plus remarquables des Paguridae est la régression accen- tuée des appendices abdominaux, uropodes exceptés. Chez la plupart cette régres- sion se traduit par la réduction, plus forte chez les males que chez les femelles, des appendices situés sur le coté gauche et par la disparition complete de ceux de droite. A cété de ces formes chez lesquelles est réalisé le type pagurien parfait, il en existe d’autres qui sont encore pourvues d’une ou de deux paires d’appendices dans la région antérieure de abdomen. Les plus primitifs a cet égard sont des Pagurinae, les Paguristes, dont il est plus particuliérement question ici, et les Paguropsis. Dans ces deux genres les femelles ont conservé des pl. 1 pairs et les males des pl. 1 et des pl. 2 pairs. Aucun autre représentant de la sous-famille ne présente plus de pléopodes symétriques, mais chez une espéce australienne, T7rizopagurus strigimanus (Miers) 11 existe, sur le deuxieme somite abdominal du male, un petit appendice droit différant seulement du gauche par la taille plus réduite et par la plus grande variabilité de lexopodite (Forest, 1952c, p. 8, fig. 8). Chez les Eupagurinae la persistance d’appendices pairs sur l’abdomen peut revétir divers aspects. Dans certains genres le male posséde des pl. 1 pairs, alors que la femelle en est dépourvue, c’est le cas de Tomopaguropsis. La formule inverse est observée chez Nematopagurus, Pylopagurus, Pagurodes, Ceratopagurus chez lesquels la femelle seule a des pléopodes symétriques, ceux de la premiere paire. Enfin chez Parapagurus, Sympagurus et Xylopagurus, si la femelle n’a que des pléopodes impairs ou si méme ceux-ci ont disparu, le male est encore pourvu de pl. 1 et de pl. 2. Lorsque les pl. 1 subsistent seuls, que ce soit chez le male ou chez la femelle, ce sont des appendices d’un type assez simple et peu varié: un article basilaire assez fort, un article distal en général foliacé, bordé de longues soies; leur role dans la reproduction n’apparait pas nettement, bien qu’on les désigne souvent sous le nom de pléopodes sexuels. Par contre lorsqu’il existe a la fois des pl. 1 et des pl. 2 pairs chez le male, il s’agit, a quelques exceptions prés,! d’organes tres différenciés constituant les éléments d’un appareil copulateur dont la structure permet d’inter- préter le fonctionnement et le rdle. Les Paguristes, objets de la présente étude, ont des appendices males batis sur le méme type, mais cette uniformité recouvre une grande diversité dans les proportions des régions homologues et dans ’ornementation. On est ainsi amené, en systémati- que, a accorder aux pl. 1 et aux pl. 2 des Paguristes une valeur identique a celle qu’on attribue aux premiers appendices sexuels des Brachyures. Morphologie et Physiologie Nous avons choisi comme type d’appareil génital externe male de Paguristes celui de P. oculatus Fabricius, espéce bien connue, d’assez grande taille et tout a fait représentative du genre au point de vue qui nous intéresse maintenant (fig. 1). 1 Quelques espéces appartenant au genre Sympagurus présentent une variabilité extréme des pléopodes pairs. Chez Sympagurus bicristatus A. Milne-Edwards par exemple en rencontre des males pourvus des deux pl. 2 ou du gauche seulement. D’autres n’ont plus de pl. 2, mais deux pl. 1. Les cas extrémes sont l’absence compléte de pléopodes pairs et trés rarement la présence simultanée des pl. 1 et des pl 2. Ces appendices sont constitués par deux articles simples au plus, ils ne sont parfois représentés que par un bourgeon. 164 ANNALS OF THE SOUTH AFRICAN MUSEUM La premiére paire de pléopodes s’insére immédiatement en arriére du dernier segment thoracique. Les deux appendices sont assez rapprochés, rabattus en avant sur la région sternale thoracique, logés entre les coxae de la derniére paire de péréiopodes. Ils se composent (fig. 2) de deux articles 4 demi-soudés: T’article basilaire assez court a section sub-triangulaire et l’article distal foliacé, enroulé de telle fagon qu’il forme une profonde gouttiére ouverte vers le plan sagittal. Lorsque Fig. 1. Paguristes oculatus Fabricius: ¢ coxae des p. 5 et gonopodes. animal est examiné par la face ventrale cet article distal parait donc constitué par deux lames: l’une, que nous qualifierons d’inférieure (/.2.), visible en totalité au premier plan; la lame supérieure, en contact avec le thorax, est en partie cachée par la précédente. La lame inférieure est allongée, épaissie et plus fortement calcifiée dans le voisinage du bord libre qui est garni de longs poils dirigés vers avant, légérement spatulée dans la région antérieure; son bord distal est armé de nom- breux crochets cornés recourbés vers l’extérieur. La lame supérieure est découpée en deux lobes: un lobe distal triangulaire (/.d.) et un lobe interne (J. znt.) trés arrondi couvert de longs poils. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 165 On observe chez les adultes, au moment de la reproduction, a l’intérieur méme des deux appendices, de nombreuses masses arrondies qui ont Vaspect de petites oranges épluchées. Chacune de ces masses est constituée par un agrégat de cellules a gros noyaux qui représentent sans doute des glandes dont la sécrétion intervient dans la fixation des spermatophores sur la femelle. Les pl. 2 (fig. 6) sont plus longs et plus gréles que les pl. 1; ils s’insérent beau- coup plus loin sur abdomen et bien plus latéralement. Ils se composent d’un trés court article basilaire (coxa rudimentaire), d’un second article long et mince, presque rectiligne, obliquement rabattu vers le plan sagittal (basis) et d’un article allongé, foliacé, enroulé sur lui-méme et formant ainsi une gouttiére ouverte vers Pextérieur. Cet article parait résulter de la soudure partielle de endopodite pro- prement dit et d’une piéce homologuable a V’appendix masculina de Thalassina et peut-étre a appendix masculina et au stylamblys des Natantia supérieurs (Eucyphidea). La région antérieure de chaque appendice a un bord libre €paissi, couvert de longs poils dirigés vers l’avant; elle est logée a lintérieur de la gouttiere du pl. 1 corres- pondant chez la plupart des Paguristes examinés. D’apres la structure de ses constituants, on peut imaginer que le fonctionnement de l’appareil génital externe des Paguristes est le suivant: au moment de la copulation qui n’a jamais été observée dans ce genre” mais qui est sans doute comparable a celle des Ecrevisses, décrite par plusieurs auteurs, la face ventrale du male viendrait en contact avec la face ventrale de la femelle. Les premiers pléopodes du male sappliquent contre les coxae de la quatrieme paire de pattes thoraciques de la femelle. C’est ici qu’interviendraient les minuscules crochets qui arment le bord antérieur des premiers pléopodes sexuels: fixés aux coxae de la femelle, dont la région antérieure forme une créte anguleuse garnies de denticules, ces crochets maintiendraient les appendices males en position. Les spermatophores sont émis au niveau de l’ouverture postérieure de la gouttiére des pl. 1; Varticle distal des pl. 2, jouant le rdle de balai, les pousse a l’intérieur de la gouttiére puis les expulse au niveau des orifices de la femelle ou ils adhérent aux longs poils qui couvrent les pieces sternales et la base des appendices thoraciques. Le dispositif décrit ci-dessus est moins parfait mais comparable a celui qu’on observe chez le Cancer pagurus § dont l’article distal des pl. 1 enroulé sur lui-méme ne forme plus simplement une gouttiére mais un canal complet, a l’intérieur duquel se déplace l’extrémité antérieure du second pléopode comme un piston dans un cylindre, pour reprendre la comparaison de Balss (1944). L’appareil genital externe chez quelques Anomoures L’anatomie comparée de l’appareil génital externe male ne serait pas un élément négligeable dans l’étude de |’évolution des Reptantia, mais ne permettrait pas a elle seule de situer les divers groupes les uns par rapport aux autres. Ce n’est pas dans le but de placer exactement les Paguristes parmi les Paguridae, ni les Paguridae parmi les Anomoures que nous allons comparer les appendices sexuels males de ? Cavolini (1787) a décrit, mais de facon extrément peu précise, la copulation chez des Pagures qui étaient peut-étre des P. oculatus. ANNALS OF THE SOUTH AFRICAN MUSEUM 166 N “EAS HHATEN NN if Ne (( Un KK Pléopodes sexuels droits chez quelques Anomoures g: en haut, pl. 1, en bas, pl. 2. 2, 6, Paguristes oculatus Fabricius; 3, 7, Sympagurus gracilipes A. M.-Edw.; 4, 8, Galathea squamifera Leach; 5, 9, Thalassina anomala Herbst. l.i., lame inférieure; L.int., lobe interne; 1.d., lobe distal; ex., exopodite; end, endo- Figs. 2-9. podite; a.m., appendix masculina; a.i., appendix interna. a PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 167 Paguristes oculatus a ceux d’autres formes. Cette comparaison montrera simplement les rapports qui existent a ce point de vue entre les Paguristes, les Eupagurinae, les Pylochelidae et les Galatheidae, et a quel point un autre Anomoure, Thalassina, peut- étre considéré comme primitif en ce qui concerne la structure de l’appareil génital externe. Les figures 3 et 7 représentent les pl. 1 et pl. 2 droits d’un Eupagurinae, Sympagurus gracilipes A. Milne-Edwards. Les premiers appendices se composent comme chez P. oculatus de deux articles. Le distal est foliacé mais ne présente qu’un léger enroule- ment si bien qu’il n’est plus possible de parler de lames supérieure et inférieure. Par ailleurs, les bords de l’article forment une courbe continue, si bien qu’il n’y a plus de limite précise entre les régions que nous avons désigné sous le nom de lobe distal et de lobe interne chez Paguristes oculatus. Quant aux pl. 2 ils sont assez trapus et formés de trois articles comme chez les Paguristes. On observe en plus un exopodite réduit a un bourgeon qui est d’ailleurs absent chez un autre Eupagurinae, Parapagurus pilosimanus. Les Pylochelidae dont nous n’avons pas de spécimens males en collection semblent presenter une assez grande hétérogénéité des pléopodes pairs males. Les figures données par Boas (1926, fig. 12 D, E et fig. 13 D, E) montrent qu’il existe de notables différences entre les Mixtopagurus et les Pylocheles. Si les pl. 1 sont dans les deux cas assez voisins et nettement plus petits que les pl. 2, ceux-ci sont pourvu chez les premiers d’un exopodite qui manque chez les seconds. D’autre part l’article distal des pl. 2 présente chez Pylocheles trois lobes séparés par de profondes encoches qui paraissent correspondre a celles, moins marquées, qu’on observe sur le pourtour de Particle distal chez P. oculatus. Chez Galathea squamifera Leach (Galatheidae) on retrouve Penroulement accentué de la région distale des pl. 1 (fig. 4). Les pl. 2 (fig. 8), plus longs et plus forts que les pl. 1, ont une région distale trés élargie, légerement enroulée, et résultant aussi de la fusion partielle de deux articles. Il existe un exopodite rudimentaire comme chez Sympagurus gracilipes. Si les différences les plus importantes entre les pléopodes sexuels males de P. oculatus et ceux des quelques Anomoures que nous avons jusqu’a maintenant passés en revue, concernent surtout les proportions des différentes régions des appendices, la fusion plus ou moins complete de certains lobes et la présence chez certains d’un court exopodite, dans tous les cas ’appareil destineé a faciliter la fixation des spermato- phores sur la femelle fonctionne a peu prés dans les mémes conditions. I] se compose d’une partie droite et d’une partie gauche, qui interviennent sans doute simultané- ment mais indépendamment lune de l’autre: les spermatophores issus de chacun des deux orifices sexuels male sont véhiculés par les appendices situés du cété corres- pondant. Chez Thalassina anomala (Herbst) les deux premiéres paires d’appendices abdominaux sont modifiés en gonopodes chez le male, mais on n’a plus ici un appareil double: les quatre pléopodes forment un appareil unique. Chacun des pl. 1 (fig. 5) est constitué par une piéce résultant de la cohalescence de plusieurs articles dont les limites sont encore visibles. I1 comprend une lame principale bordée de longues soies sur son bord externe, un lobe distal saillant et arrondi, un 168 ANNALS OF THE SOUTH AFRICAN MUSEUM lobe antéro-interne couvert de petits poils crochus constituant un rétinacle. Les pl. 2 (fig. 9) ont une structure plus compliquée que chez les formes précédentes: endopodite comprend une piéce basilaire allongée, déprimée, sur laquelle s'articule, en avant, un flagelle sub-cylindrique constitué par de courts articles plus ou moins cohalescents, et une piéce de méme longueur que la portion basilaire et également déprimée, que Boas désigne sous le nom d’appendix masculina. Un lobe interne situé dans la région proximale de l’appendix masculina et qui forme rétinacle serait suivant le méme auteur a homologuer a l’appendix interna des Eucyphidea. L’exopodite multi-articulé est sub-cylindrique et égal aux deux tiers environs de lendopodite; son article proximal est beaucoup plus long que les suivants. Dans leur complexité les pl. 2 de Yhalassina anomala male paraissent bien plus proches de ceux de certains Astacura, Astacus astacus L. par exemple (voir Balss 1941, p. 172, fig. 230) que de ceux des Paguridae. Nous avons dit qu’il s’agissait d’un appareil unique: en effet, au moment de la copulation, les deux pl. 1 maintenus en contact par les rétinacles forment une seule gouttiére dont le bord postérieur s’ouvre au niveau des orifices sexuels qui sont trés rapprochés. Les spermatophores sont poussés a l’intérieur, puis a l’extérieur de cette gouttiére par les extrémités des pl. 2 accolés eux aussi l’un a l’autre par leurs rétinacles. Utilisation des appendices sexuels males pour la détermination des Paguristes Chez les Brachyures, les premiers pléopodes des males adultes se sont révélés comme d’excellents caractéres et permettent souvent de distinguer avec une facilité relative des espéces par ailleurs fort voisines. Nos observations nous ont montré qu’il en était de méme pour les Paguristes: dans ce genre, les premiers et aussi les seconds appendices abdominaux présentent, d’une part, une grande constance de forme, de proportions et d’ornementation chez les males adultes d’une espéce donnée, et d’autre part, des differences caractéristiques d’une espéce a lautre dans de nombreux cas. La technique d’utilisation des pl. 1 et des pl. 2 males dans la systématique des Paguristes est trés simple. Les appendices sont désarticulés a la base et détachés sous la loupe binoculaire a l’aide d’une aiguille lancéolée; ils sont ensuite immergés pendant quelques instants dans l’alcool absolu, qu’ils proviennent d’animaux con- servés dans l’alcool a 70° ou dans l’eau formolée. On les monte enfin a la résine mastic dans la position qu’ils occupent sur le Pagure lorsqu’on |’examine par la face ventrale. Dans beaucoup d’espéces les pl. 1 présentent un enroulement trés accentué de l’article distal si bien que la lame inférieure est enticrement par dessus et cache une partie de la lame supérieure; il est donc nécessaire de retourner la préparation pour voir la totalité du lobe distal et du lobe interne. Lorsque l’enroulement est plus faible; ce qui est le cas pour plusieurs espéces sud-africaines, l’article distal s’étale entre lame et lamelle et son contour est entiérement visible du méme cété de la préparation. Les seconds pléopodes ont également une région distale plus ou moins enroulée en hélice. I] n’est pas toujours possible d’étaler parfaitement cette région en pressant PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 169 sur la lamelle, et il arrive que les différents articles des appendices n’occupent pas une position absolument comparable d’une préparation a l’autre. Les dessins ont été exécutés avec la face la plus pileuse de la région distale par dessus. Les figures 42 a 54 qui représentent les pl. 1 des Paguristes males des cétes occi- dentales et méridionales d’Afrique montrent la diversité d’aspect de ces appendices et les points sur lesquels ils different: présence ou absence de crochets sur le bord Figs. 10-12. Trois états du développement du premier pléopode sexuel chez Paguristes gamianus H. Milne-Edwards: 10, jeune ¢ a carapace de 4 mm.; jeune ¢ a carapace de 5°5 mm.; 12, g adulte 4 carapace de 12 mm. antérieur de la lame inférieure, proportions de cette lame, forme du lobe interne. Les pl. 2 sont sans doute moins différenciés mais dans de nombreux cas la portion distale de l’endopodite a une forme caractéristique. Il faut noter que les pléopodes sexuels ne sont pas toujours de bons caractéres, et que P. gamianus ne parait guére pouvoir étre distingué de P. barnardi par les pl. 1. Pour que les comparaisons prennent toute leur valeur il est nécessaire qu’elles portent sur des appendices complétement formés; il est probable que la longueur par rapport a celle du corps, la forme et l’ornementation des pléopodes ne sont acquises qu’a la maturité sexuelle. C’est ce que montrent les figures 10 4 12 qui représentent, au méme grossissement, trois états de développement du pl. 1 droit 170 ANNALS OF THE SOUTH AFRICAN MUSEUM chez P. gamianus. Le premier appartient a un individu a carapace de 4 mm., le second est celui d’un spécimen a carapace de 5:5 mm., le dernier enfin est celui d’un adulte a carapace de 12 mm. Entre les deux extrémes, alors que le rapport des tailles est de 3 a 1, le rapport des longueurs des appendices est de 5 a 1. III. SysTEMATIQUE PAGuRIDAE Dana Pagurinae Ortmann Gen. Paguristes Dana Rostre variable, parfois trés réduit mais toujours présent. Ecailles oculaires bien développées, plus ou moins écartées. Ecailles antennaires fortes, denticulées. Flagelle antennaire de longueur moyenne ou court. Endopodite des maxillules avec un petit appendice latéral externe. Flagelle de ?endopodite des premiers maxilli- pédes dirigé vers l’intérieur et non vers l’avant. Maxillipedes externes rapprochés ala base. Chélipedes de méme forme, le plus souvent sub-égaux, a ongles générale- ment cornés, a doigts mobiles dans un plan horizontal. Dactyle des quatriéme pattes thoraciques terminal, ne formant pas une pince avec le propode. 13 paires de branchies (pas de branchies sur le dernier segment thoracique). Chez le male une paire de pléopodes, modifiés en gonopodes, sur chacun des deux premiers segments abdominaux. Sur chacun des trois segments suivants un seul pléopode—le gauche—dont la rame externe est bien développée et la rame interne rudimentaire. Chez la femelle, une paire de pléopodes uniramés sur le premier segment abdominal; sur les trois segments suivants, un pléopode gauche aux deux rames bien développées, plus fort que chez le male; sur le cinquiéme segment un pléopode gauche de méme type que ceux du male. Souvent un repli de Pabdomen prenant naissance en arriére du dernier pléopode biramé et formant chambre incubatrice. Les Paguristes ont une trés large répartition; on en rencontre dans la plupart des mers chaudes ou tempérées, depuis la zone littorale jusqu’a des profondeurs de plusieurs centaines de metres. Si les Paguristes d’Afrique du sud paraissent former un groupe a part, dont les représentants n’ont pas, jusqu’a present, été signalés en d’autre régions, rien ne prouve cependant quil y ait un cloisonnement géographique et que certains d’entre eux ne seront pas retrouvés dans des localités beaucoup plus septentrionales. C’est la raison pour laquelle nous avons rassemblé les espéces des cdtes occidentales et méridionales d’Afrique dans un tableau unique qui permettra de les distinguer les unes des autres. Tableau de détermination des Paguristes des cétes occidentales et méridionales d’ Afrique. 1. Ecailles oculaires uni—ou, rarement, bidentées .. os Ecailles oculaires a bord antérieur armé de 3 dents au moins. .. : 2. Flagelles antennaires de méme longueur ou plus courts que les atiloawles semilenese Flagelles antennaires nettement plus longs que les pédoncules oculaires Hes eo Out 10. Palys 12. 13. 14. 15. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 171 . Diamétres des cornées sensiblement égal a la moitié du diameétre maximum des pédoncules oculaires (fig. 25). Afrique du Sud, littoral. .. a i Paguristes engyops (p. 204). Diamétre des cornées légérement inférieur au diamétre maximum des pédoncules oculaires (fig. 14). Cdtes du Sahara, 115-290 metres. ore oH a Paguristes marocanus (p. 175). Chélipédes paraissant finement granuleux; le gauche beaucoup plus fort que le droit. Une large tache rouge sur la face interne de la main. Maroc, A.O.F., jusqu’a 150 metres. Angola (?). Paguristes oculatus var. rubro-pictus (p. 172). Chélipédes tuberculés ou épineux, sub-égaux. .. ay. Ss sk ih Be sais Sige Face interne du propode des pattes ambulatoire revétue de poils plumeux trés serrés dissimulant compléetement l’ornementation de cette face. Région supérieure du dactyle des mémes appen- dices armée seulement de fines épines cornées, cachées sous une frange de poils extrémement épaisse (fig. 39, et pl. IV). Afrique du Sud, littoral... ae Paguristes barnardi (p. 208). Pilosité des pattes ambulatoires beaucoup moins forte, laissant apparaitre l’ornementation du tégument. .. sie a: at Ba Sc Ae a aye = 33 is, Os Ecailles oculaires modérément écartées. Bord supérieur du carpe, du propode et du dactyle des p. 2 avec de fortes dents a base calcifiée, a pointe cornée. Des epines cornées sur la face interne du propode de ces appendices (fig. 38, et pl. IV). Afrique du Sud, jusqu’a 24 metres. Paguristes gamianus (p. 200). Ecailles oculaires trés écartées. Des dents beaucoup moins fortes sur les p. 2. Pas d’épines cornées sur la face interne du propode de ces appendices. .. i on oH ¥: Sesuke ny Ecailles antennaires avec 4 a 5 dents sur le bord interne (fig. 27). Afrique du Sud, 90 métres. Paguristes macrotrichus (p. 211). Ecailles antennaires avec une seule dent sur le bord interne (fig. 26). Afrique du Sud, 55 métres. Paguristes agulhasensis (p. 205). Ecailles oculaires beaucoup plus longues que larges._ .. a aE aE ang Sigs Ecailles oculaires sensiblement aussi longues que larges. Se uss Ay ig Bay pli Pédoncules antennulaires a peine plus ihe que les a oa oculaires (fig. 19). Sénégal, 6-15 metres. sin é Be Paguristes rubrodiscus (p. 189). Pédoncules antennulaires Lyne tn les yeux de ie moitié de la longueur de leur dernier article au moins. wi ie ay ig ahs oe sh ua as Be sc SLO: Pédoncules antennaires a peu pres de méme longueur que les pédoncules oculaires (fig. 20). Guinée frangaise, 8-10 metres. .. ; : os . .Paguristes microphthalmus (p. 191). Pédoncules antennaires dépassant les yeux de la moitié de ia longueur de leur dernier article. (fig. 22). Liberia, Congo, 12-15 metres. si bie Se . .Paguristes hispidus (p. 196). Ecusson céphalothoracique sensiblement aussi long que large. aH Mi ae ‘ot SOR Ecusson céphalothoracique nettement plus long que large. Be si an be Be LG: Pédoncules antennaires et oculaires de méme longueur. Angola, 72 métres. . Paguristes staaee e 197). Pédoncules antennaires n’atteignant pas tout a fait la base des cornées (fig. 21). Guinée fran- caise, Congo, 8-36 metres. Le m si ae oe .. Paguristes virilis (p. 193). Rostre dépassant nettement les pointes latérales (fig. 15). Sénégal, Mauritanie, 5-50 metres. reser Saget (p. 176). Rostre atteignant au plus l’alignement des pointes latérales. .. spe se an Propode des p. 2 a peine deux fois plus long que haut et a bord supérieur trés fortement denté (fig. 33). Sénégal, 8-40 metres. i bes .. Paguristes oxyacanthus (p. 186). Propode des p. 2 au moins deux fois et demie athe long que haut et a bord supérieur armé de dents aigués mais petites. aie oe Be ae whe me ae se AiG ti Main des chélipédes ovale, bord antérieur des pléopodes 1 du ¢ inerme (fig. 44). Mauritanie, Sénégal, Congo, 5-85 metres. .. .. Paguristes mauritanicus (p. 179). Main sub-triangulaire, bord antérieur ais pléopore I di na armé de crochets (fig. 45). Sénégal, 8-10 metres. ae ; 2 Paguristes difficilis (p. 183). 172 ANNALS OF THE SOUTH AFRICAN MUSEUM Paguristes oculatus Fabricius var. rubro-picitus A. Milne-Edwards et E. L. Bouvier Figs. 13 et 28 Paguristes maculatus var. rubro-picta, A. Milne-Edwards et E. L. Bouvier, 1892, p. 207. Paguristes oculatus var. brunneo-pictus, A Milne-Edwards et E. L. Bouvier, 1900, p. 165, pl. VI, fig. 1. Paguristes oculatus var. brunneo-pictus, ‘T. Odhner, 1923, p. 6. Materiel examiné: 1°. Les spécimens récoltés par le ‘““Travailleur’ et par le ‘““Talisman”’ et décrits par A. Milne-Edwards et E. L. Bouvier (types de la variété). 2°. 1 g, l.c. 13°5 mm., Cap de Naze (région de Dakar), 28-30 m., févr. 52, Cremoux coll. 3°. 1 9 lc. 4 mm., Ile Bissagos, 120 m., “‘Président Théodore Tissier’, 1936, Station 718. Description: L’écusson céphalothoracique (fig. 13) est toujours notablement plus long que large et représente plus de la moitié de la longueur totale de la carapace. Le rostre long et aigu dépasse de beaucoup les dents latérales et atteint a peu prés le milieu des écailles oculaires. Les pédoncules oculaires sont légerement renflés a la base et au sommet; le gauche, en général un peu plus grand que le droit, a une longueur comprise entre les 3/4 et les 4/5 de celle de la région pré-cervicale. Les cornées sont échancrées postérieurement. Les écailles oculaires sont largement écartées, leur moitié distale est triangulaire, l’extrémité est acuminée et le bord antéro-latéral entier. Les pédoncules antennulaires atteignent ou dépassent légeére- ment le bord postérieur des cornées. Les pédoncules antennaires arrivent au tiers ou au quart distal des pédoncules oculaires; le deuxiéme article est armé d’une épine au bord antéro-interne et de deux longues épines a l’angle antéro-externe: le sommet bidenté de l’écaille antennaire n’atteint pas tout a fait l’extrémite distale du dernier article; son bord interne présente quatre a neuf dents acérées, et son bord externe, deux ou trois dents. Le chélipéde gauche est en général plus long et plus gros que le droit: la largeur de la main gauche est égale aux 3/4 de sa longueur environ, alors que la main droite est deux fois plus longue que large. La face supérieure du carpe, du propode et du dactyle des deux chélipédes est couverte de nombreux petits tubercules coni- ques a pointe cornée, uniformément répartis, et qui, sur les bords des articles, prennent laspect de faibles dents cornées. Les pattes ambulatoires dépassent légérement les chélipédes. Les p. 2 (fig. 28) ont un mérus a bord inférieur armé de denticules aigus; le bord supérieur du carpe est défini par une rangée de sept 4 neuf dents cornées, celui du propode par sept a dix dents, et celui du dactyle par une ligne des denticules dont la taille diminue réguliérement de l’arriére vers l’avant; le bord inférieur et la face interne du pro- pode sont faiblement spinuleux et le bord inférieur du dactyle est armé de longues et fines épines cornées. Les p. 3 présentent, aux mémes endroits que les précédentes, des dents beaucoup plus réduites. PAGURISTES D’ AFRIQUE L’OUEST ET DU SUD 173 N — =m | Cro Figs. 13-17. Région antérieure de la carapace et appendices céphaliques antérieurs: Paguristes oculatus var. rubro-pictus A. M.-Edw. et E. L. Bouvier; 14, P. marocanus A. M.-Edw. et 13; E. L. Bouvier; 15, P. fagei Forest; 16, P. mauritanicus Bouvier; 17, P. difficilis Forest. 174. ANNALS OF THE SOUTH AFRICAN MUSEUM Les femelles possédent un orifice sexuel sur chacune des coxae des p. 3; un vaste repli de abdomen qui prend naissance en arriére du troisiéme pléopode impair, recouvre la plus grande partie de le ponte. La lame principale des pl. 1 du male s’élargit distalement en spatule; son bord antérieur est armé de nombreuses et fortes épines en crochets, disposées en plusieurs rangées dans la région médiane. Le lobe interne est séparé par une profonde encoche aigué du lobe distal qui n’atteint pas tout a fait l’apex de la lame inférieure. La coloration des régions calcifiées est d’un blanc-rosé plus ou moins maculé de rouge. Les principales marques persistantes sont des taches d’un rouge pourpre mélé de violet, situées sur les faces interne et externe du mérus des chélipédes prés de l’articulation du carpe; il y a aussi une large tache sur la face interne de la région palmaire, en arriére du doigt mobile. Les p. 2 et les p. 3 présentent deux larges anneaux rouge, l’un sub-distal, ’autre proximal, sur le dactyle, et un anneau proximal sur le propode. La pilosité de cette variété est plus forte que celle de la forme typique. Les chélipédes sont recouverts d’une pubescence qui prend laspect d’une courte frange sur les bords externes de la main. Remarques: Le Paguristes oculatus Fabricius est avant tout une espéce méditer- ranéenne, mais on |’a signalé aussi dans |’Atlantique entre la céte méridionale du Portugal et le Gap Mazagan. Les spécimens récoltés au large de la cote occidentale du Maroc et du Soudan par le “Travailleur’”’ et par le ““Talisman”’ ont été décrits sous le nom de P. oculatus var. brunneo-pictus par A. Milne-Edwards et E. L. Bouvier, mais huit ans auparavant, dans des observations préliminaires sur ce matériel, ces auteurs avaient propose pour la variété le nom de rubro-picta qui—la loi de priorité s’appliquant ici sans qucun doute—doit étre substitué a celui de brunneo-pictus. Un male, beaucoup plus grand que ceux du “Travailleur” et du ““Talisman’’, dragué au sud de Dakar et vivement coloré, répond a la description de la variété, ainsi qu’une petite femelle capturée au large des [les Bissagos. Ces spécimens ne présentent que peu de differences morphologiques avec ceux de Méditerranée. Les pléopodes sexuels en particulier, qui permettent de distinguer facilement des Paguristes fort voisins, sont ici identiques dans les deux cas. La distinction ne repose en fait que sur la pigmentation (et encore ne s’agit-il que de taches colorées supple- mentaires ou plus intenses), sur la pilosité plus forte, sur la taille plus faible, et sur une gracilité plus grande des p. 2 et p. 3 dans la variété. Ces caractéres sont pro- bablement étroitement liés aux facteurs externes; profondeur, conditions physico- chimiques du milieu, etc. ... Gependant nous n’avons pas rencontre d’intermédiaires entre la forme typique et la variété, et nous continuerons a les distinguer tout au moins provisoirement. | Si les spécimens signalés par Odhner sous le nom de P. oculatus var. brunneo- pictus appartiennent bien a cette espéce, elle s’étendrait vers le sud jusqu’a l’Angola. PAGURISTES D’AFRIQUE L’ OUEST ET DU SUD 175 Paguristes marocanus A. Milne-Edwards et E. L. Bouvier Figs. 14, 29, 42, 55 Paguristes marocanus, A. Milne-Edwards et E. L. Bouvier, 1891, p. 152. Paguristes marocanus, A. Milne-Edwards et EK. L. Bouvier, 1892, p. 207. Paguristes marocanus, A. Milne-Edwards et E. L. Bouvier, 1900, p. 167—70, pl. XXIII, fig. 1-6. Matériel examiné: 1 3, l.c. 3°5 mm., “Talisman’’, 8.7.83, devant le Cap Bojador, PouNect 17. 6, W., 130 m. (type dessinc). 1 g,1.c. 4 mm., “Talisman” 13.7.83, au nord du banc d’Arguin, 21° 51’ N. et 19° 48’ W., 115-140 m. (type). Les autres spécimens mentionnés par les auteurs ne figurent pas dans la collection. I 4g, Le. 3 mm., cdtes du Sahara, 1884, de Cuverville coll. Description: L’écusson céphalothoracique (fig. 14) est tres allongé et deux fois plus long que la région postérieure de la carapace; sa largeur représente les 4/5 environ de sa longueur. Le rostre aigu est aussi saillant que les dents latérales dont il est séparé par deux profondes concavités. Les pédoncules oculaires sont assez larges et renflés aux extrémités; leur longueur représente les 5/8 environ de celle de Pécusson céphalothoraciaque. Les cornées sont tres grandes. Les écailles oculaires sont écartées, longuement lancéolées, et ne présentent aucune indentation sur le bord antéro-latéral. Les pédoncules antennulaires dépassent les yeux de la moitié de leur dernier article environ. Les pédoncules antennaires atteignent les cornées; le deuxieme article est armé d’une seule épine a chacun de ses angles antérieurs. L’écaille antennaire atteint le milieu du dernier article; son extrémité est bidentée, et elle présente deux dents sur son bord interne qui est concave et trois dents sur son bord externe; le flagelle est a peine plus long que les pédoncules oculaires. Les chélipédes sont sub-égaux. Le carpe est aussi long que le dactyle et plus long que le bord palmaire interne. La main est deux fois plus longue que large. Les bords interne et externe de la face supérieure du carpe sont marqués, le premier par quatre dents assez fortes, le second par une rangée irréguliere de petits tubercules. I] existe aussi des tubercules sur la face supérieure de la main, trés arrondis et peu nombreux dans la région médiane, un peu plus denses et un peu plus saillants sur les cotés et notamment sur le bord palmaire interne qui est armé de quatre ou cing petites dents. Les pattes ambulatoires p. 2 (fig. 29) et p. 3 dépassent largement les chélipedes; elles sont gréles et leurs deux derniers articles sont peu déprimés latéralement. Les bords supérieur et inférieur du mérus des p. 2 sont spinuleux; le bord supérieur du carpe est armé de six dents acérées et celui du propode de cing dents; le rapport des longueurs du propode et du dactyle est légerement supérieur a 2/3. Les p. 3 sont inermes a l’exception du bord supérieur du carpe qui présente l’épine distale habituelle; la coxa est nettement plus longue et le mérus nettement plus court que pour p. 2- L’article proximal des pléopodes 1 du male (fig. 42) n’est que légérement plus court que l’article distal. La lame inférieure de ce dernier est prés de quatre fois 176 ANNALS OF THE SOUTH AFRICAN MUSEUM plus longue que large, le bord antérieur est armé de huit ou neuf courts crochets; le lobe interne atteint et le lobe distal dépasse apex de la lame inférieure. Nous n’avons pas pu examiner de femelle de cette espéce. Les spécimens étudiés sont tous décolorés et d’une teinte blanchatre uniforme. D’aprés A. Milne-Edwards et E. L. Bouvier “la couleur dans l’alcool tire sur le rose-chair, avec quelques taches blanchatres’’. La pilosité générale est assez faible; cependant de trés longs poils garnissent le bord antérieur des écailles oculaires et ’ornementation des chélipédes est en grande partie dissimulée par les poils qui les recouvrent et qui sont salis par la vase. Remarques: Cette espéce au nom assez mal choisi puisque tous les spécimens connus ont été capturés bien au sud des cétes marocaines? ne risque d’étre confondue avec aucun des autres Paguristes de la cote occidentale d’Afrique. La région pré- cervicale beaucoup plus longue que la région postérieure, le bord frontal aux trois dents aigués aligneées, les écailles oculaires aigués, a bord antérieur entier et frangé de longs poils, la briéveté des flagelles antennaires permettent de lidentifier rapide- ment. A en juger par les spécimens examinés, et qui sont des males adultes dont le plus grand a une carapace de 4 mm., c’est une espéce de petite taille, Pune des plus petites du genre. C’est aussi une espéce qui vit a une assez grande profondeur, puisque le ““Talisman”’ l’a capturée sur des fonds de 115 a 290 m. Le fait qu’on ne ait signalée qu’au large des c6tes du Sahara ne signifie pas que sa répartition eéographique est aussi restreinte. Le matériel récolté au-dela des fonds de 120 m. est en effet pour linstant relativement peu abondant. Le Paguristes qui parait présenter le plus d’affinités avec P. marocanus appartient a la faune sud-africaine: c’est P. engyops Barnard qui vit dans la zone littorale. Nous verrons les rapports qui existent entre les deux espéces a la suite de la description de la seconde. Paguristes fager Forest Figs. 15, 30, 43, 57 Paguristes fager Forest, 1952, p. 260, fig. 3. Matériel examiné: De nombreux individus § et 9 de 1:5 a 7 mm. (l.c.) devant Thiaroye, baie de Rufisque, dragage sur fond coquillier par 10 a 15 m., 18-1-52, Paraiso coll. (Syntypes). 2 det 3 9, lic. 3-5 a 5-5 mm.: Cap Blanc, 50 m., 1907, Gruvel coll. 1 9, Le. 1-8 mm.: chalutier “Cabellou’’, au large du Sénégal, 12 a 15 m., 10-1-41, Monod coll. 3 6 et 16 9, lc. 3a 6 mm.: chalutier “Vers l’Horizon’’, au large du Sénégal, 8 m., fond caillouteux a grosses Ascidies, 21-1-41, Monod coll. 6 g et 2 9 ovigeres, l.c. 2 4 5 mm.: Gorée, filets a langouste, 23, 24-8-50, Paraiso coll. 6 g et 9, M’Bao (Sénégal), dragage a un mille de la céte, 5-14 m., 7-2-51, Delais coll. 3 Milne-Edwards et E. L. Bouvier l’ont sans doute ainsi baptisée parce que l’un des spécimens se trouvait dans une coquille de Stnistralia marocana, PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 177 Figs. 18-22. Région antérieure de la carapace et appendices céphaliques antérieurs: 18, Paguristes oxyacanthus Forest; 19, P. rubrodiscus Forest; 20, P. microphthalmus Forest; 21, P. virilis Forest; 22, P. hispidus A. M.-Edw. et E. L. Bouvier. 178 ANNALS OF THE SOUTH AFRICAN MUSEUM 10 get 12 2 dont 8 ovigéres, l.c. 2 4 7 mm., chalutier “Maid Honour’, 8° 38’ N., a 8° 42’ N., 8 a 12 m., 1 au 8-3-48, Cadenat coll. Description: Les régions pré-et post-cervicale de la carapace sont sensiblement de méme longueur; la premiere (fig. 15) est un peu plus longue que large, le rapport de ses dimensions étant de 5/4 environ. Le rostre, large et en angle obtus, dépasse légérement Valignement des deux dents latérales. Les pédoncules oculaires, un peu plus courts que le bord frontal, sont renflés aux extrémités, et tout particuliére- ment dans la région proximale chez les jeunes individus. Les cornées atteignent le milieu du dernier article du pédoncule antennulaire et l’extrémité du pédoncule antennaire. Les écailles oculaires sont grandes, larges, trés rapprochées et leur bord antérieur est armé de cing a huit denticules. L’angle externe du deuxiéme article des antennes est uni-ou bidenté. Les écailles antennaires atteignent au maximum le milieu du dernier article du pédoncule; leur extrémité est plus ou moins nettement bidentée, et elles sont armées de deux ou trois épines sur le bord interne légérement concave et d’une épine sur le bord externe. Les chélipédes sont égaux et de méme forme. Le bord inférieur du mérus forme une créte denticulée. La main est deux fois plus longue que large, le doigt mobile est un peu plus long que le bord palmaire interne. Un léger hiatus subsiste entre les doigts lorsque ceux-ci sont fermés. La face supérieure des trois derniers articles porte de nombreux tubercules coniques, peu élevés et a pointe cornée, plus aigus vers les bords latéraux. Sur la face supérieure de la main, en arriere de l’articulation du dactyle, existe une pro- tubérance tuberculée plus ou moins développée. Les pattes ambulatoires, p. 2 et p. 3 dépassent les chélipédes. Les p. 2 (fig. 30) ont un mérus a bord inférieur non denticulé, un carpe a bord supérieur armé d’une rangée irréguliére de dents cornées assez fortes qui se prolonge sur le propode sous la forme d’une ligne de dix a douze dents cornées plus petites. Le dactyle a une section circulaire et son diamétre a la base est compris quatre a cing fois dans sa longueur; le bord supérieur et le bord inférieur sont marqués par une rangée d’épines minuscules largement espacees. Cet article est a peu pres de méme longueur que le précédent. Les p. 3 sont peu différentes des p. 2 dans leur forme générale, mais elles sont inermes a l’exception d’une épine distale sur le bord supérieur du carpe et d’une rangeée de tres petites épines sur le bord inférieur du dactyle. La femelle ne posséde qu’un orifice sexuel situé sur la coxa de la p. 3 gauche, et ne présente pas de repli de abdomen en arriére du troisieme pléopode impair. Les pléopodes 1 du male (fig. 43) ont une lame inférieure a moitié distale élargie en spatule et bordée de fines épines recourbées vers l’extérieur. Le lobe interne est séparé du lobe distal qui dépasse l’apex de la lame inférieure par une encoche arrondie. La coloration de fond des régions calcifiées est d’un blanc jaunatre ou orange. Les écailles oculaires, la base des pédoncules oculaires et une étroite région sous les cornées sont rouge-orange, alors que la partie moyenne des pédoncules oculaires, les pédoncules antennulaires et antennaires, et l’endopodite des maxillipédes externes, sont d’un bleu plus ou moins intense. Les chélipédes sont maculés de rouge-orange. La face interne du mérus présente une coloration constante: la région antérieure PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 179 est blanche, le reste d’un rouge-orange présentant un maximum d’intensité a la limite des deux zones.. Les deux premieres paires de pattes ambulatoires présentent également de grandes zones rouge-orange qui couvrent en particulier la moitié proximale du propode et du dactyle. Remarques: Il n’y a pas de raisons de comparer Paguristes fagei a l'une plutét qu’a autre des espéces de la cote occidentale d’Afrique. I] posséde tout un ensemble de caractéres, et notamment une pigmentation qui le distinguent de tous les autres représentants du genre. C’est ainsi que le bleu intense des pédoncules oculaires, des antennules et des antennes permet d’identifier au premier coup d’oeil les exem- plaires frais, méme de trés petite taille. Au bout de quelques années dans l’alcool, il se produit une décoloration presque totale, mais on peut, dans la plupart des cas, reconnaitre la ligne brisée qui sépare la zone foncée de la zone claire sur la face interne du meérus des chélipédes. D’autres caractéres stables et d’observation facile sont utilisables par ailleurs: la forme de la carapace et surtout celle du bord frontal avec le rostre qui dépasse légérement les dents latérales dont il est séparé par des concavités peu profondes, les écailles oculaires courtes, larges et rapprochées, plutét quadrangulaires que triangulaires. Les pléopodes 1 du male sont aussi d’un type trés particulier et ne présentent quelque ressemblance qu’avec ceux du P. oculatus. P. fagei est un espece de petite taille: les plus grands exemplaires ont une carapace de 7 mm., et chez les plus petites 2 ovigéres cette région du corps n’a pas plus de 2-5 mm. Des ¢ de 2 mm. de carapace ont déja des pléopodes sexuels de méme forme que chez les individus plus agés. Une partie des spécimens observés étaient a l’intérieur de coquilles de Gastéropodes: Nassa, Fusus, Turritella, Tritonalia, etc. ..., mais la plupart étaient logés a Pintérieur de colonies de Bryozoaires dont chacune présente un aspect bien particulier: la masse centrale abritant le Pagure se prolonge latéralement par deux longues cornes formant balancier et présente dorsalement une créte digitée. La répartition géographique de la nouvelle espéce, telle qu’on peut se la repré- senter actuellement, ‘s’*étend du Cap Blanc (22° N. environ), au Sud de Dakar (8° 38’ N.). Elle a été récoltée principalement entre 5 et 15 m. de profondeur, mais Péchantillon du Cape Blanc a été dragué a 50 m. Paguristes mauritanicus Bouvier 1906 Figs. 16, 31, 44, 56 Paguristes mauritanicus Bouvier, 1906 a, p. 186, fig. 1; 1906 b, p. 96, fig. 1; 1906 ¢, p- 199. Matériel examiné: 1 3, \.c. 7 mm.: au large de Novak Chott, 16-24 m., Gruvel coll. (Type). I 9, lc. 4°5 mm., 1 g l.c. 6-5 mm. et 1 9 l.c. 4 mm.: Baie de Hann (Dakar), 6-1o m., sable coquillier grossier, 1950, R. Sourie coll. 180 ANNALS OF THE SOUTH AFRICAN MUSEUM a7), Figs. 23-27. Région antérieure de la carapace et appendices céphaliques antérieurs: 23, Paguristes gamianus H. M.-Edw.; 24, Paguristes sp. (Table Bay); 25, P. engyops Barnard; 26, P. agulhasensis sp. nov.; 27, P. macrotrichus sp. nov. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 181 Nombreux individus 3 et 2 de 3 a 8 mm. (l.c.) dont plusieurs ovigéres: devant Thiaroye, baie de Rufisque, 10-15 m., sable coquillier, 18-1-52, Paraiso coll. 1 9, l.c. 3 mm., 1 J, lc. 5 mm., M’Bao (Sénégal), dragage a un mille de la céte, 5-14 m., 7-2-52, Delais coll. 1 g, l.c. 6 mm.: “‘Mercator’’, dragage entre les Iles Tamara et Roume (Iles de Los), 8-10 m., 10-12-36. ieee, 4mm... 9 les 2-5;mimt 1) 2) less imm:, et 29 ovigeres, ics 4) et 5 mm., “M’Bizi’, St. 144, 26 milles SSW. Cap Lopez, 95-100 m., 10-3-49. 1 2 ovigere, l.c..5 mm.:. “M’Bizi’, St. 142,:29 milles'S.. Cap: Lopez, 50 m.; 9-3-49- 2 6, lc. 4:5 et 8-5 mm.: “M’Bizi’, St. 26, 25 milles SW. Pointe de Banda, 85 m., 6-9-48. I 9 juv., l.c. 2°5 mm., 1 9, l.c. 6 mm.: “M’Bizi’, St. 8,15 milles SW. Moita Seca, 50 m., 3-8-48. 48 spécimens récoltés en 15 stations entre 14 et 44 m. par R. Bassindale au large de la Gold Coast. Le nombre, la taille et le sexe des individus provenant de chaque station figureront dans |’étude d’ensemble des Pagures récoltés par R. Bassindale. Description: La région postérieure de la carapace est de méme longueur ou un peu plus courte que l’écusson céphalothoracique; celui-ci (fig. 16) a une largeur sensiblement égale aux 6/7 de sa longueur; les aires latérales sont faiblement spinu- leuses. Le rostre, en angle obtus, n’atteint pas l’alignement des dents latérales. Les pédoncules oculaires sont renflés aux extrémités; le rapport de leur longueur a celle de la région précervicale est compris entre 3/4 et 2/3. Les écailles oculaires sont largement écartées; leur région distale est obliquement tronquée, et le bord antéro- latéral est armé d’une épine apicale et de deux ou trois petites dents plus petites. Les pédoncules antennulaires dépassent les yeux de la moitié de la longueur de leur dernier article au moins. Les pédoncules antennaires atteignent le milieu des cornées; le deuxieme article est armé de deux trois épines a l’angle antéro-externe qui est assez saillant; l’écaille antennaire, dont l’extrémité bidentée arrive au quart ou au tiers proximal du dernier article, est armée de une ou deux dents sur son bord interne légérement concave et de deux ou trois dents sur son bord externe. Le chélipéde droit est en général un peu plus fort que le gauche, mais de méme forme. Le carpe, de méme longueur que le dactyle, est plus long que le bord palmaire interne. La main est a peu prés deux fois plus longue que large, son épaisseur maximum est comprise 2-4 fois environ dans sa longueur. La face supéri- eure du carpe est armée de cing a six fortes dents sur le bord interne et de dents plus petites irréguli¢érement disposées le long du bord externe. La face supérieure du propode et du dactyle est couverte de tubercules coniques peu élevés qui ont plutét l’aspect de gros granules chez les grands spécimens. Ces tubercules sont un peu plus aigus sur le bord externe de la main et ce sont cing ou six fortes dents cornées qui marquent le bord palmaire interne. Les pattes ambulatoires de la premiere paire (fig. 31) dépassent largement les chélipédes. Le bord supérieur du carpe est armé de cing ou six dents, Le bord 182 ANNALS OF THE SOUTH AFRICAN MUSEUM supérieur du propode d’une douzaine de dents; cet article est a peu prés trois fois plus long que haut, alors que la longueur du dactyie représente a peu pres sept fois sa hauteur a la base. Le rapport des longueurs du dactyle et du propode, mesurés le long de leur bord supérieur, est de 5/3 environ. Les p. 3 sont un peu plus gréles et un peu plus longues que les p. 2; elles sont inermes a l’exception d’une épine distale au bord supérieur du carpe. Les femelles n’ont qu’un orifice sexuel qui s’ouvre sur la coxa de la p. 3 gauche. Chez les femelles adultes, un large repli de Pabdomen qui prend naissance en arriére du troisiéme et dernier pléopode biramé recouvre cet appendice et le précédent et’ éventuellement, une grande partie de la ponte. Les pléopodes 1 du ¢ (fig. 44) ont une lame inférieure quatre fois plus longue que large, a bord interne légérement concave, et dépourvue de crochets sur le bord antérieur; sur les préparations, le lobe interne, trés large et trés arrondi déborde largement la lame inférieure et se rattache au lobe distal trés saillant par une concavité réguliere et peu profonde. Des exemplaires qui n’ont séjourné qu’un mois dans l’alcool, présentent la coloration suivante: les régions calcifiées ont un fond blanchatre; il y a une large tache brun-rouge en avant de la carapace, tout prés du bord frontal; les écailles oculaires, la base des pédoncules antennaires, et les pédoncules oculaires tout entiers sont d’un brun-rouge intense. Les chélipédes sont brun-violacé, les doigts et les tubercules blancs. Les deux premiére pattes ambulatoires portent de larges anneaux alternativement clairs et foncés. Sur les échantillons qui ont séjourné plus longtemps dans l’alcool, les marques colorées s’éclaircissent et tendent a devenir rose. C’est la coloration des pédoncules oculaires et des pattes ambulatoires qui subsiste le plus longtemps. Les poils plumeux trés denses qui couvrent la région frontale dissimulent lornementation de cette région, mais laissent voir une grande partie des pédoncules oculaires. Les appendices thoraciques sont aussi fortement pileux notamment sur les faces supérieure et inférieure, cependant les doigts des chelipédes restent toujours visibles. Remarques: Le Paguristes mauritanicus n’était connu jusqu’a présent que par Punique individu ¢ décrit par Bouvier. Les nombreux spécimens dont nous dis- posons maintenant montrent qu’il ne s’agit pas d’une espéce rare. C’est de la région de Dakar que proviennent la plus grande partie des échantillons, dragués depuis la cote jusqu’a une profondeur de 20 4 25 m. Les individus capturés au sud de cette région et jusqu’a l’embouchure du Congo, proviennent d’un niveau inférieur, de 50 a 100 m., et présentent une différence de coloration dans les pédoncules oculaires, lesquelles ont une moitié proximale bleutée, et une moitié distale orange. Les autres caractéres, et notamment la forme des pléopodes du 3, ne permettent pas de con- sidérer ces spécimens comme appartenant a une forme différente. L’espéce la plus voisine est le P. difficilis sp. nov. décrit plus loin, dont Paspect est fort peu différent, et qui ne peut guére étre distingué que par la forme des chélipédes et surtout par celle des pléopodes 1 du mle, trés caractéristiques. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 183 Les coquilles abritant les P. mauritanicus appartiennent aux genres les plus divers, Bullia, Cancellaria, Clavatula, Gibbula, Marginella, Mesalia, Nassa, Terebra, Trophon, Turritella, etc. C’est en Baie de Rufisque, devant Thiaroye, et par 10 a 15 m. qu’ont été capturés les plus nombreux spécimens de P. mauritanicus. Ce dragage a ramené en outre d’autres espéces de Paguridae: Petrochirus pustulatus juv., Pagurus pectinatus juv., Pseudopagurus granulimanus, Diogenes ovatus, D. pugilator var. intermedius, Eupagurus sculptimanus et E. souriei, et enfin des spécimens de trois autres espéces nouvelles de Paguristes: P. faget, P. rubrodiscus et P. oxyacanthus. Paguristes difficilis Forest. Figs. 17, 45, 61 Paguristes difficilis Forest, 1952, p. 259, fig. 1. Matériel examiné: 8 § de 4 a7 mm., 2 9 de 7 mm.: Baie de Hann, 8-10 m., sable coquillier grossier, Janvier 1950, Sourie coll. Type. A Description: L’écusson céphalothoracique (fig. 17) est a peu prés de méme longueur que la région postérieure de la carapace et un peu plus long que large; ses aires latérales sont faiblement spinuleuses. Le rostre obtus et arrondi n’atteint pas tout a fait Palignement des dents latérales. Les pédoncules oculaires renflés aux extrémités, ont une longueur égale aux 5/6 de la région pré-cervicale; les cornées sont profondément échancrées postérieurement. Les écailles oculaires sont largement écartées; leur moitié distale est triangulaire; l’épine distale est suivie de deux ou trois petites dents sur le bord antéro-latéral. Les pédoncules antennulaires dépassent les yeux de la moitié de la longueur de leur dernier article. Les pédoncules antennaires atteignent le milieu des cornées; le deuxieme article est armé de trois dents a angle entéro-externe qui est assez saillant; la pointe de Pécaille antennaire atteint le tiers proximal du dernier article, ses bords latéraux sont armés de deux ou trois dents; le flagelle antennaire est égal aux 3/4 de la longueur de la carapace. Les chélipedes sont égaux. Le carpe est un peu plus court que le dactyle dont la longueur représente les 2/3 environ de la longueur totale de la main. Celle-ci est sub-triangulaire et présente son maximum de largeur pres de la base; son épaisseur maximum est comprise 2-7 fois dans la longueur. Le bord interne de la face supéri- eure de carpe est défini par une rangée de quatre fortes dents cornées, et le bord externe par des dents plus nombreuses mais plus petites. Le propode et le dactyle ont une face supérieure couverte de nombreux tubercules arrondis pourvus d’une trés petite pointe cornée. Le bord interne de la main porte une rangée de quatre fortes dents. Le propode des pattes ambulatoires de la premiére paire est armé de douze a quinze fortes dents cornées sur son bord supérieur; le bord supérieur du dactyle porte également des petites dents cornées sur toute sa longueur. Le rapport des longueurs du dactyle et du propode est de 3/2 environ. Les p. 3 sont inermes, exception faite de la petite dent distale au bord supérieur du carpe. 184 ANNALS OF THE SOUTH AFRICAN MUSEUM Figs. 28-33. Deuxiéme patte thoracique gauche, face interne, dénudée: 28, Paguristes oculatus var. rubro-pictus A. M.-Edw. et E. L. Bouvier; 29, P. marocanus A. M.-Edw., et E. L. Bouvier; 30, P. faget Forest; 31, P. mauritanicus Bouvier; 32, P. rubrodiscus Forest; 33, P. oxyacanthus Forest. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 185 Les femelles n’ont qu’un orifice sexuel sur la coxa de p. 3 gauche. Un large repli abdominal qui prend naissance en arriére du troisitme pléopode impair recouvre les appendices précédent, et, éventuellement, une partie de la ponte. Les pléopodes 1 du male (fig. 45) ont une lame inférieure a peine deux fois et demie plus longue que large qui, sur les préparations, recouvre presque complétement le lobe interne. Chez les adultes le bord antérieur de cette lame est armé de petites épines recourbées vers l’extérieur. Le lobe distal est étroit et séparé du lobe interne par une profonde concavité. La région antérieure de la carapace, les écailles oculaires, la base des pédoncules antennaires sont roses. Les pédoncules oculaires sont orange-clair. Les chélipédes sont de la méme teinte avec les doigts blancs. Les pattes ambulatoires portent de larges anneaux d’un rose plus foncé. Les poils qui couvrent la région frontale et les appendices thoraciques sont assez denses mais ne dissimulent pas completement l’ornementation du tégument. Remarques: Cette espéce est la plus difficile a caractériser parmi les Paguristes d’Afrique occidentale. En effet, elle présente une grande ressemblance dans |’aspect de la carapace, des appendices céphaliques antérieurs et des pattes ambulatoires, avec. P. mauritanicus. Si la région pré-cervicale est nettement moins allongée chez P. difficilis, si P. mauritanicus a de plus des pédoncules oculaires un peu plus gréles, ce sont la des caractéres assez variables et sur lesquels il serait dangereux de baser la détermination. Par contre la forme des chélipédes différe sensiblement d’une espéce a autre; chez P. mauritanicus, le main, subovale, présente son maximum de largeur vers le milieu de l’article et le rapport de |’épaisseur maximum sur la longueur a une valeur de 2-4 environ. Chez P. difficilis, la main plus finement granulée est sub- triangulaire et présente son maximum de largeur un peu en avant de l’articulation du carpe; elle est aussi moins renflée: le rapport de son épaisseur maximum sur sa largeur ayant une valeur de 2-7 environ. Lorsqu’il s’agit de spécimens frais la dis- tinction peut se faire d’aprés la pigmentation des pédoncules oculaires: de teinte claire chez P. difficilis, d’un rouge-brun tres foncé chez P. mauritanicus. Mais pour les males, le meilleur moyen de distinguer les deux espéces est d’examiner les pl. 1. Chez P. difficilis, d’un rouge-brun trés foncé chez P. mauritanicus. Mais pour les males, le meilleur moyen de distinguer les deux espéces est d’examiner les pl. 1. Chez P. difficilis la lame inferieure est beaucoup plus large et, chez l’adulte, armée de crochets sur le bord antérieur. Chez les plus petits exemplaires étudiés les crochets sont absents, mais la forme générale est le méme que chez les individus plus Aagés et Paspect bien différent de ce qu’on observe chez les P. mauritanicus au méme stade. Les dix spécimens que nous avons entre les mains nous sont parvenus sans coquilles, mais quelques-uns présentent l’aplatissement dorso-ventral et l’étirement des plaques sternales thoraciques qui caractérisent les pagures vivant dans des coquilles a étroite ouverture, Conus ou Marginella par exemple. Paguristes difficilis n’a été récolté que dans une seule localité, au large de Dakar, sur les mémes fonds que P. mauritanicus: le méme dragage a ramené trois spécimens de cette derniére espéce. 186 ANNALS OF THE SOUTH AFRICAN MUSEUM Paguristes oxyacanthus Forest Figs. 18, 33, 47, 58 Paguristes oxyacanthus Forest, 1952, p. 261, fig. 5. Matériel examiné: 1 9 juv., l.c. 7 mm.: Baie de Hann (Dakar), 8-19 m., sables coquilliers grossiers, 7-1-50, Sourie coll. 1 Q juv., lc. 6-5 mm.: Baie de Hann, 12 m., cailloux et coquilles, 17-6-50, Sourie coll. 3 9, lc. 9°5 a 11 mm., 2 G juv., lc. 5:5 et 7 mm.: devant Thiaroye, Baie de Rufisque, dragage sur fond coquillier, 10-15 m., 18-1-52, Paraiso coll. Type. I g, l.c. 12°5 mm.: sud de Gorée, 40 m., Février 52, Postel coll. Description: L’écusson céphalothoracique est a peu prés aussi long que la région postérieure de la carapace; le rapport de sa longueur a sa largeur est de 7/6 environ; sa surface présente un relief tourmenté, avec une dépression transversale trés marquee en arriére du bord frontal, avec des tubercules saillants et de nombreuses spinules sur les aires antérieures et latérales; le rostre en angle obtus n’atteint pas tout a fait Palignement des dents latérales. Les pédoncules oculaires sont renflés a la base et, plus faiblement, au niveau des cornées qui présentent une échancrure postérieure aigué; leur longueur est égale aux 4/5 de celle de la région pré-cervicale. Les écailles oculaires sont écartées; elles sont armées d’une €pine terminale et de trois petites dents sur le bord antéro- latéral. Les pédoncules antennulaires dépassent les yeux de la moitié de leur dernier article. Les pédoncules antennaires n’atteignent pas la base des cornées; l’angle antéro-externe saillant du deuxiéme article est pourvu de quatre dents groupées par deux. L’écaille antennaire atteint le tiers proximal du dernier article, son extrémité est bifide, et ses bords latéraux portent trois fortes dents, celles du bord interne étant groupées dans la moitié proximale; le flagelle est égal aux 2/3 de la carapace. Les chélipédes sont é¢gaux et fortement épineux; le bord inférieur de la face interne du meérus est défini par une douzaine de denticulations assez fortes. Le carpe est plus long que le dactyle et celui-ci un peu plus long que le bord palmaire interne. La face supérieure du carpe et du propode est réguli¢érement convexe de avant vers l’arriére, mais parait plutét concave que convexe en section trans- versale: en effet, les bords latéraux sont armés de trés fortes dents a pointes cornées légérement inclinées vers avant et vers lextérieur: le bord externe du carpe est armé de huit dents, le bord interne de quatre dents trés fortes; le bord externe du propode est défini par une rangée de dents assez nombreuses, mais plus petites; par contre, le bord interne ne porte que trois dents trés développées. Le reste de la face supérieure de ces articles est couvert de tubercules coniques de plus petites tailles. La face externe du dactyle c’est-a-dire celle qui est orientée vers le plan sagittal de lanimal—est large et plate; elle est limitée en haut par une ligne de dents cornées assez forte, en bas, par des tubercules coniques cornés peu saillants; il existe une troisieme ligne longitudinale intermédiaire constituée par de petites épines cornées. — Les pattes ambulatoires p. 2 et p. 3 dépassent largement les chélipédes. Les p. 2 (fig. 33) ont un merus dont le bord inférieur est spinuleux. La région supérieure PAGURISTES D’ AFRIQUE L’OUEST ET DU SUD 187 du carpe est armée de dix fortes dents cornées irréguli¢érement disposées. Le bord supérieur du propode est défini par une ligne de neuf dents un peu plus petites. Toute la face interne de cet article est couverte de petits tubercules coniques a pointe cornée. La face interne du dactyle porte une rangée supérieure et une rangée inférieure de petites spinules cornées. Les deux articles distaux sont assez fortement déprimés latéralement; le propode n’est pas beaucoup plus de deux fois plus long que haut et le rapport des longueurs du dactyle et du propode est de 5/4 environ. Les p. 3 different peu des p. 2 dans la forme de leurs articles, mais le bord supérieur du carpe ne porte qu’une épine distale; il n’y a pas de fortes dents sur le bord supérieur du propode mais toute la face interne de cet article est couverte de tubercules coniques moins forts que sur p. 2. Les femelles n’ont qu’un orifice sexuel, lequel s’ouvre sur la coxa de p. 3 gauche; un vaste repli de la peau qui prend naissance en arricre du troisieme pléopode impair recouvre tous les pléopodes antérieurs et par conséquent la totalité de la ponte. Les pléopodes 1 du male ont une lame inférieure pres de quatre fois plus longue que large, a bord latéraux sinueux et paralléles a bord antérieur faiblement denticulé. Le lobe interne plus large que la lame principale est séparé du lobe distal trés proémi- nent par une encoche arrondie peu profonde. Tous les spécimens observés présentent une coloration caractéristique. Les regions calcifiées sont d’un blanc-rosé, maculé de rouge carmin. Cette teinte est particulierement intense sur la carapace, ou elle forme deux taches en arriére des concavités du bord frontal, sur la base des pédoncules antennaires, dans les régions proximales des quatre derniers articles des p. 2 et des p. 3, et dans la région antérieure de la face interne du mérus. Les pédoncules oculaires et antennulaires sont blancs avec des bandes longitudinales vermillons. La région antérieure de la carapace, les appendices céphaliques antérieurs, et les pattes thoraciques sont fortement pileux. L’extrémité antérieure des pédoncules oculaires émerge d’un revétement assez dense de poils plumeux. Remarques: Cette espéce se rapproche du P. mauritanicus Bouvier et du P. difficilis Forest par la forme et les proportions de la carapace et des appendices céphaliques antérieurs. Elle s’en distingue par la coloration, par la taille et Pacuité des dents et épines des chélipédes et des pattes ambulatoires p. 2 et p. 3, par la proportion des articles de ces appendices: le rapport de la hauteur a la longueur du propode des p. 2 en particulier est 4 peine inférieur a 1/2 alors qu’1il est au plus égal a 1/3 dans les deux autres especes; les plépodes 1 du ¢ ont une lame principale beaucoup moins large que chez P. difficilis, et a bord antérieur denticulé, alors que cette région est inerme chez P. mauritanicus. P. oxyacanthus est une espéce assez grande si on la compare aux autres Paguristes de cette région: les 2 de 5:5 et 7 mm. (I.c.), draguées en Baie de Rufisque le 18-1-52, présentent des caractéres juvéniles: la premiére paire de pléopodes (pl. 1) est peu développée, les pléopodes impairs sont petits et les trois premiers ont une rame interne réduite, et il n’y a pas de repli abdominal en arriére de pl. 3. Au contraire 188 ANNALS OF THE SOUTH AFRICAN MUSEUM chez les 2 de plus grande taille, ce repli est trés développé et recouvre non seulement la région antérieure de l’abdomen mais les coxae des pattes ambulatoires. 2 2 de Figs. 34-36. Deuxieme patte thoracique gauche, face interne, dénudée: 34, P. microphthalmus Forest; 35, P. hispidus A. M.-Edw. et E. L. Bouvier; 36, P. virilis Forest. 6-5 et 7 mm. (l.c.), présentent des caractéres intermédiaires et notamment un repli abdominal peu développé. Les spécimens qui nous sont parvenus avec leurs coquilles étaient logés dans des Mesalia, des Murex, et des Turritella. ‘Tous proviennent de la région de Dakar, d’une profondeur comprise entre 8 et 40 m. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 189 Paguristes rubrodiscus Forest Figs. 19, 32, 46, 59 Paguristes rubrodiscus Forest, 1952, p. 261, fig. 6. Matériel examiné: 1 9, 1.c. 7 mm., et 1 9 ovigeére, l.c. 7-5 mm.: Gorée, Aott 1946. Type. I g, lc. 9 mm.: Anse Bernard (Dakar), sable et cailloux, 6-10 m., 6-5-50, Sourie coll. 1 dg, lc. 10 mm.: Baie de Hann, vers le banc de Bel Air, 12 m., cailloux et coquilles, 17-6-50, Sourie coll. I 9 ovigére, l.c. 7 mm.: Gorée, filets a langoustes, 23,24-8-50. I individu adulte (non extrait de sa coquille): Gorée, dragage sur fond rocheux, 15 m., 28-12-51, Delais coll. peeeee 45a Gimm.. 209 Veo 5-5cet 6smm~.,) 1 2 juve, lc. 4, mm: ; devant Thiaroye, Baie de Rufisque, dragage sur fond coquillier, 10-15 m., 18-1-52, Paraiso coll. Description: L’écusson céphalothoracique (fig. 19) a une longueur sensiblement égale aux 5/4 de sa largeur et aux 4/7 de la longueur totale de la carapace; elle est faiblement spinuleuse en arriere des pédoncules antennaires. Le rostre est aigu et atteint l’alignement des dents latérales dont il est séparé par deux profondes con- cavités. Les pédoncules oculaires sont renflés a la base et s’élargissent au niveau des cornées qui sont assez petites et ne sont pas échancrées postérieurement. Les écailles oculaires sont plus longues que larges, rapprochées et armées de quatre a cing indentations dans la région antérieure. Les pédoncules antennulaires dépassent les pédoncules oculaires du quart ou du tiers de la longueur de leur dernier article. Les pédoncules antennaires sont plus courts et n’atteignent pas les cornées; leur deuxiéme article présente deux ou trois épines a chacun des angles antérieurs; les écailles antennaires sont longues, elles arrivent au tiers distal du dernier article pédonculaire, elles ont une extrémité bidentée et présentent trois ou quatre dents irréguliérement disposées sur le bord interne qui est légérement con- cave et deux ou trois dents au bord externe. Les flagelles antennaires ont une longueur égale au 3/4 de la région antérieure de la carapace. Les chélipédes sont égaux et de méme forme. Le carpe est allongé et sa face supérieure est armée de dents cornées acérées dont les plus fortes sont disposées en deux rangées longitudinales, une externe de six, l’autre interne de quatre. La main est deux fois plus longue que large, le dactyle est nettement plus long que la région palmaire. La face supérieure porte des tubercules coniques peu proéminents sauf sur le bord palmaire interne qui est marqué par quatre trés fortes dents cornées crochues dirigées vers l’avant. Il n’y a pratiquement pas de hiatus entre les doigts lorsqu’ils sont en contact par leurs ongles cornés. Les deux premiéres paires de pattes ambulatoires dépassent légérement les chélipédes. Le mérus des p. 2 (fig. 32) est armé de cing 4a six petites dents sur son bord inférieur. Le bord supérieur du carpe est défini par une rangée irréguliére de neuf fortes dents cornées, le bord supérieur du propode par six a huit dents cornées 190 ANNALS OF THE SOUTH AFRICAN MUSEUM un peu plus petites. Les p. 3 sont un peu plus longues que les p. 2, leurs articles sont inermes, a l’exception du carpe dont le bord supérieur présente une ou deux dents distales. Pour ces deux paires d’appendices le dactyle est un peu plus long que le propode et a une section circulaire. La femelle ne posséde qu’un orifice sexuel situé sur la coxa de la troisiéme patte thoracique gauche. Immédiatement en arriére du troisieme pléopode impair on observe chez les femelles adultes un petit repli du tegument qui est frangé de longs poils et recouvre la base de cet appendice. Les pléopodes 1 du male (fig. 46) ont une lame inferieure trés allongée, a bords latéraux rectilignes et paralléles sur presque toute leur longueur, a bord antérieur arrondi et inerme. Le lobe interne beaucoup plus large que la lame inférieure est séparé du lobe distal par une encoche étroite et profonde. Les régions calcifiées sont d’un blanc-rosé, maculé de rouge. Il y a en particulier deux taches rouges persistantes de chaque cété de la région pré-cervicale de la carapace, en arriere des zones épineuses. Les pédoncules antennulaires portent une bande longitudinale rouge sur toute la longueur de leur face supérieure et de leur face inférieure. La pigmentation la plus caractéristique est celle de la face interne du mérus des chélipédes qui est marquée dans sa région antérieure par une tache rouge arrondie, toujours visible apres un séjour de plusieurs années dans Valcool. La plus grande partie de la région antérieure du corps disparait sous un revéte- ment de longs poils plumeux, particulicrement denses sur les aires latérales de la carapace, sur les pédoncules et les écailles oculaires, sur les écailles antennaires, et sur les face supérieure et inférieure des pattes thoraciques. Remarques: Le Paguristes rubrodiscus présente quelques affinités avec P. hispidus A. Milne-Edwards et E. L. Bouvier et avec P. microphthalmus Forest qui ont comme lui des écailles oculaires plus longues que large et contigués, des cornees petites et peu ou pas échancrées en arriére; cependant tout un ensemble de caracteres ne laisse guere la possibilité de le confondre avec lune ou l’autre de ces formes: région antérieure de la carapace bien plus longue que large, rostre étroit et aigu, pédoncules antennulaires a peine plus longs que les pédoncules oculaires, alors que dans les deux autres espéces, ces appendices dépassent les cornées des 2/3 de leur dernier article. La présence de la tache circulaire rouge sur la face interne du merus des chélipédes permet d’ailleurs de Videntifier rapidement. Les trois espéces désignées ci-dessus sont apparentées aux Paguristes de la Mer Rouge et l’aspect général du P. perspicax Nobili (décrit comme variété de P. jousseaume: Bouvier) est fort voisin de celui de P. rubrodiscus, les différences portant sur des caractéres nombreux mais, dans l’ensemble, peu marqués. La taille des P. rubrodiscus adultes examines est comprise entre 4:5 et Io mm. (l.c.); les Q adultes ne présentent qu’un court repli de l’abdomen en arriére de pl. 4. Quelques-uns des spécimens étaient logés dans des coquilles de Clavatula ou de Turnitella, et un d’eux dans une coquille enti¢érement recouverte de Balanes; tous ont été récoltés dans la région de Dakar, entre 6 et 15 m. de profondeur. 9 ot PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 1QI Paguristes microphthalmus Forest Figs. 20, 34, 48, 60 Paguristes microphthalmus Forest, 1952, p. 260, fig. 4. Matériel examiné: 3 3, l.c. 4:5 a 8 mm., 1 @ ovigere, l.c. 6 mm.: “‘Mercator’’, dragage entre les iles Tamara et Roume (Iles de Los), 8-10 m., 10-12-36. Type. Description: La région pré-cervicale représente prés des 2/3 de la longueur totale de la carapace; le rapport de sa largeur sur sa longueur est de 5/6 environ. Le rostre, en angle obtus, atteint presque l’alignement des deux dents latérales. Les pédoncules oculaires sont deux fois plus larges a la base qu’au niveau des cornées dont l’échancrure postérieure n’est qu’a peine indiquée. Leur longueur est légére- ment inférieure aux 3/4 de celle de la région pré-cervicale. Les écailles oculaires sont longues, assez étroites, et contigués; leur bord antérieur est tronqué et armé de trois longues épines suivies d’une petite dent sur le bord externe. Les pédoncules antennulaires dépassent les cornées des 2/3 de leur dernier article. Les pédoncules antennaires sont a peu pres de méme longueur que les pédoncules oculaires; le deuxiéme article a un angle antéro-externe peu saillant armé de deux ou trois dents. L’écaille antennaire a une extrémité bidentée qui dépasse de peu la base du dernier _article; son bord interne, concave, porte cinq ou six dents, le bord externe deux ou trois dents. Les chélipédes sont égaux et assez courts; la face interne du mérus dont le bord inférieur est armé de six ou sept dents a une largeur maximum sensiblement égale au 3/4 de la longueur. Le carpe est un peu plus long que le bord palmaire interne et celui-ci est de méme longueur que le doigt fixe. La face supérieure du carpe présente de fortes dents cornées dont la plupart sont disposées en deux rangées: Pune de six le long du bord externe, l’autre de quatre ou cing le long du bord interne. La main est couverte de tubercules coniques, peu saillants, sauf sur le bord palmaire interne marqué par cing dents cornées. Les deux paires de pattes ambulatoires sont un peu plus longues que les chéli- pedes. Les p. 2 (fig. 34) ont un mérus dont le bord inférieur est marqué par quelques spinules. La région supérieure du carpe porte une rangée de six ou sept fortes dents, et une seconde rangée plus externe de six autres dents qui n’occupent que la moitié distale; le bord supérieur du propode est défini par une rangée de douze dents assez fortes, réguli¢rement espacées, qui se prolonge jusqu’ a la moitié distale du dactyle sous la forme de denticules de taille décroissante. Le rapport des longueurs du dactyle et du propode, mesurées le long du bord supérieur, est de 4/3 environ. Les p. 3 sont inermes a l’exception d’une faible denticulation sur le bord supérieur du mérus et d’une dent distale au bord supérieur du carpe. Le rapport des longueurs du dactyle et du propode est voisin de 3/2. Les femelles n’ont qu’un orifice sexuel, qui s’ouvre sur la coxa de p. 3 gauche. Le repli de abdomen qu’on observe chez elles en arriére du troisitme pléopode impair, est peu développé. Les pléopodes 1 du male (fig. 48) ont une lame inférieure quatre fois et demie a cing fois plus longue que large, a bord antérieur arrondi. Une rangée de minus- 192 ANNALS OF THE SOUTH AFRICAN MUSEUM cules épines assez espacées s’étend du milieu du bord latéral externe au tiers interne du bord antérieur. Le lobe interne trés développé est s¢paré par une encoche aigué du lobe distal, lequel dépasse largement la lame inférieure. Les exemplaires observés sont d’une teinte blanchatre a peu prés uniforme. Une tache plus foncée sans doute rouge a Vorigine subsiste cependant sur la face interne du mérus des chélipédes, dans sa région antérieure. Gaillerol.m Figs. 37-41. Deuxiéme patte thoracique gauche, face interne, dénudée: 37, Paguristes agul- hasensis sp. nov.; 38, P. gamianus H. M.-Edw.; 39, P. barnardi sp. nov.: 40, P. engyops Barnard; 41, P. macrotrichus sp. nov. PAGURISTES D’ AFRIQUE L’OUEST ET DU SUD 193 Les poils qui couvrent les régions latérales de la carapace et les appendices céphaliques antérieurs ne sont pas assez denses pour cacher complétement le tégument. Par contre les chélipédes et les deux paires de pattes ambulatoires p. 2 et p. 3 disparaissent en grande partie sous un épais revétement de poils plumeux. Remarques: Par Vaspect général de la région antérieure du corps, par la forme allongée et la contiguité des écailles oculaires, P. microphthalmus Forest se rapproche de P. hispidus A. Milne-Edwards et E. L. Bouvier et de P. rubrodiscus Forest. La longueur relative des pédoncules antennaires et oculaires permet de distinguer facilement les trois espéces: dans la premiére, les pédoncules antennaires et ocu- laires sont a peu prés de méme longueur, dans la seconde, les pédoncules antennaires dépassent les yeux de la moitié de la longueur de leur dernier article et dans la troisiéme, ils n’atteignent pas les cornées. Par l’aspect de la région antérieure du corps P. microphthalmus est aussi apparenté a P. jousseaume: de la Mer Rouge; les différences portent sur la forme et l’ornementation des chélipédes et des deux premieres pattes ambulatoires. Les quatre spécimens qui constituent le type de cette espéce ont été dragués dans les parages des iles de Los, par 8 a 10 m. de profondeur. Paguristes virilis Forest Figs. 21, 36, 49, 63 Paguristes virilis Forest, 1952, p. 262, fig. 7. Matériel examiné: 1 g et 2 9, l.c. 5-5 mm.: Le Roume (I. de Los), sable vasard, BOC URUS ig eca 7.5 mim. et 1 P ovigere, lc. 5-5 mm.: Mercator’, dragage entre les tiles Tamara et Roume (I. de Los), 8-10 m., 10-12-36. ‘Type. I g, Lc. 6 mm. et 1 9 ovigére de 7 mm.: “M’Bizi”’, station 174,13 milles W. de Banana, vase, 30-35 Mm., 3-4-49. 12 spécimens provenant de 5 stations entre 11 et 28 m. au large de la Gold Coast. Ces spécimens seront étudiés avec l’ensemble de la collection R. Bassindale. Description: L’écusson céphalothoracique est a peu prés de méme longueur que la région postérieure de la carapace, sensiblement aussi large que long et couvert de petites saillies spinuleuses, plus nombreuses sur les aires latérales. Le rostre obtus est nettement plus court que les dents latérales. I] n’y a pas de limite définie entre le bord frontal et les cotés de la carapace. Les pédoncules oculaires sont assez étroits dans leur région moyenne mais renflés aux extrémités; ils sont un peu plus courts que la région antérieure de la carapace. Les écailles oculaires sont aussi longues que larges, assez écartées et leur bord antérieur est quadridenté. Les pédoncules antennulaires dépassent les yeux de la moitié de leur dernier article. Les pédoncules antennaires plus courts n’atteig- nent pas les cornées, angle antéro-externe saillant de leur deuxiéme article est armé de quatre dents, leur bord interne de deux dents; |’écaille antennaire, qui atteint le milieu du dernier article pédonculaire, a des bords latéraux rectilignes dont chacun 194 ANNALS OF THE SOUTH AFRICAN MUSEUM est armé de cinq ou six fortes dents régulicrement espacées. La longueur du flagelle est égale aux 3/4 de celle de la carapace environ. Les chélipédes sont égaux. Le carpe et le dactyle sont de méme longueur et plus long que le bord palmaire interne. Le rapport de la largeur a la longueur de la main est de 3/5 environ. La face supérieure de ces articles est couverte de tubercules coniques peu saillants sauf sur les bords latéraux du carpe et sur le bord palmaire interne ou ils prennent l’aspect de fortes dents a pointe plus ou moins émoussée. Les doigts entrent en contact sur toute leur longueur lorsqu’ils sont fermés. Les deux paires de pattes ambulatoires p. 2 et p. 3 dépassent les chélipédes de toute la longueur de leur dactyle. Les p. 2 (fig. 36) ont un mérus dont le bord supérieur seul présente quelques denticulations; le bord supérieur du carpe est armé d’un nombre trés variable de dents a pointes cornées; il y a sur l’article suivant sept a onze dents plus faibles. Le dactyle est trés gréle, son plus grand diamétre est compris neuf fois environ dans sa longueur; il se termine par un ongle trés petit. Le rapport des longueurs du dactyle et du propode est de 5/3 environ. Les p. 3 different des p. 2 en ce que leur mérus est réguli€rement denticulé le long du bord supérieur alors que les autres articles sont inermes. Les pattes sont plus gréles encore que les précédentes: le dactyle est dix fois plus long que large et pres de deux fois plus long que le propode. La femelle ne posséde qu’un orifice génital, sur la coxa de p. 3 gauche. En arriére du troisieme pléopode impair biramé, un court repli de ’abdomen coiffe la partie postérieure de la ponte. Les pléopodes 1 du male (fig. 49) sont trés développés. Le lobe interne n’est séparé du lobe distal que par une simple fente et la lame supérieure prend ainsi Paspect d’un triangle tres allongé. Les specimens observés ne préesentent pas de coloration caractéristique. La pilosité est assez forte mais les longs poils plumeux qui couvrent les bords de la carapace, les appendices céphaliques antérieurs et les pattes ambulatoires ne sont pas assez denses pour dissimuler complétement le tegument. Par contre les chélipédes disparaissent sous un épais manchon de poils. Remarques: Paguristes virilis appartient au groupe de Paguristes dont les écailles oculaires sont courtes et ont un bord antérieur denticulé. Quelques caractéres stables et bien apparents permettent de le distinguer des espéces déja deécrites: région antérieure de la carapace aussi large que longue, rostre trés bas, écailles antennaires a bords rectilignes avec des dents aigués nombreuses et réguli¢rement espacées. C’est sans doute avec Paguristes skoogi qu’il présente les plus grandes affinités; nous n’avons malheureusement pas pu savoir ou sont conserves les spécimens décrits sous ce nom par T. Odhner en 1923. D’aprés la description originale et la figure qui l’accompagne, deux caractéres permettraient de distinguer les deux espéces: chez P. skoogi les pédoncules antennaires sont aussi longs que les pédoncules oculaires et les écailles antennaires sont armées de trois dents au bord externe, et de quatre a cinq dents au bord interne, alors que, chez P. virilis, les pédoncules antennaires n’atteignent pas les cornées et les bords externe et interne des écailles antennaires sont armés de cing ou six dents. On peut encore relever des PAGURISTES D’ AFRIQUE L’OUEST ET DU SUD 195 Figs. 42-54. Premier pléopode sexuel droit chez les Paguristes g des cdtes occidentales et méridionales d’Afrique: 42, P. marocanus A. M.-Edw. et E. L. Bouvier; 43, P. fagei Forest; 44, P. mauritanicus Bouvier; 45, P. difficilis Forest; 46, P. rubrodiscus Forest; 47, P. oxyacanthus Forest; 48, P. microphthalmus Forest; 49, P. virilis Forest; 50, P. hispidus A. M.-Edw. et E. L. Bouvier; 51, P. engyops Barnard; 52, Paguristes sp. (Table Bay); 53, P. agulhasensis sp. nov.; 54, P. barnardi sp. nov. ie 42 0) /0051147./40)< 205) 52.) X) 1551 54.) 6 22), ues autres, X95. 196 ANNALS OF THE SOUTH AFRICAN MUSEUM différences dans la taille des pédoncules oculaires, dans l’ornementation et les pro- portions des chélipédes et des deux premiéres paires de pattes ambulatoires. Il faut noter que les pléopodes 1 du male de P. virilis ont une forme bien caractéristique et sont prés de deux fois plus longs que chez les autres Paguristes de méme taille. Chez les femelles ovigéres examinées le repli abdominal est peu développé et ne recouvre que la base du troisieme pléopode biramé. Les coquilles a Pintérieur desquelles étaient logés quelques-uns des spécimens examinés appartiennent aux genres Murex et Clavatula. P. virilis a été décrit d’aprés du matériel provenant d’une part des iles de Los, d’autre part, d’un point beaucoup plus méridional, puisque situé au large de lembouchure du Congo, mais ces deux localités, géographiquement fort éloignées, présentent un caractére commun: les eaux renferment une grande quantité de vase en suspension et les fonds sont vaseux. En ce qui concerne la profondeur, remarquons que les P. virzlts du Congo ont été pris par 30 a 35 m., alors que ceux des I. de Los ont été dragués sur des fonds de 8 a 10 m. Paguristes hispidus A. Milne-Edwards et E. L. Bouvier Figs. 22, 35, 50, 62 Paguristes hispidus, A. Milne-Edwards et E. L. Bouvier, 1892, p. 208; 1900, p. 170, pl. XXIII, fig. 7-10. | Matériel examiné: 1 3, 4 mm (type) et 1 2 3-5 mm., “Commandant Parfait’, Monrovia (Liberia). 1g, 6 mm., “M’Bizi’, St. 38, 4 milles W. Moita Seca, sable et vase, 12-15 m., 5-10-48. Description: A. Milne-Edwards et E. L. Bouvier ont donné une description détaillée du type, un male dont la carapace mesure 4 mm. Comme le spécimen de M’Bizi est légerement différent, probablement en raison de sa taille plus grande, il est utile de donner une description le l’espéce qui, portant sur plusieurs spécimens, aura ainsi une validité plus grande. | La région antérieure de la carapace n’est que légérement plus longue que la région postérieure: le rapport de la largeur de l’écusson céphalothoracique a sa longueur est de 7/8 environ; ses aires latérales sont spinuleuses. Le rostre est beau- coup moins saillant que les dents latérales. Les pédoncules oculaires sont gréles, fortement renflés a la base et légérement au niveau des cornées; celles-ci sont petites et ne présentent pas d’échancrure postérieure; la longueur des pedoncules représente les 2/3 environ de celle de Vécusson céphalothoracique. Les écailles oculaires sont contigués et trés allongées; leur bord antérieur est armeé de quatre longues dents aigués. Le bord antérieur des cornées atteint le quart proximal du dernier article du pédoncule antennulaire et le milieu du dernier article du pédoncule antennaire. Le deuxiéme article de ce dernier présente deux ou trois dents aigués sur la saillie antéro-externe. L’écaille antennaire dont l’extrémité atteint la base du dernier article, est armée, en plus des deux dents apicales, de trois ou quatre dents sur le bord interne et d’une dent sur le bord externe. PAGURISTES D’AFRIQUE L’ OUEST ET DU SUD 197 Les chélipédes sont égaux et de méme forme. La face supérieure de la main est sub-triangulaire; elle présente sa plus grande largeur—soit les 5/8 de sa longueur— au niveau du quart proximal et se rétrécit progressivement jusqu’a l’extrémité des doigts. Le dactyle est nettement plus long que le bord palmaire interne. Un large hiatus subsiste entre les doigts lorsque les ongles sont en contact. La face supérieure du carpe, peu bombée, est limitée vers l’intérieur par une rangée de six dents cornées et vers l’extérieur par quelques dents moins fortes. La face supérieure de la main est couverte de tubercules coniques peu serrés dont la plupart sont disposés en lignes longitudinales. Le bord palmaire est marqué par quatre ou cing dents un peu plus aigués. Les pattes ambulatoires, trés gréles, dépassent largement les chélipedes. Le bord supérieur du carpe de p. 2 (fig. 35) est armé de sept épines. Le propode et le dactyle sont inermes; la longueur du premier ne représente que les 2/3 de celle du second. Les p. 3 sont également inermes, a l’exception du bord supérieur du carpe pourvu d’une épine distale; leurs articles et surtout le propode et le dactyle, sont plus gréles encore que ceux des p. 2. La femelle examinée (a carapace de 3°5 mm.) n’a qu’un orifice sexuel sur la coxa de p. 3 gauche. Les trois premiers pléopodes impairs de ce spécimen présentent une disposition particulicre. Ils sont rapprochés les uns des autres et groupés dans le quart antérieur de Pabdomen. Un quatriéme pléopode, uriramé, est situé loin en arricre. Les pléopodes I du male (fig. 50) ont une lame inférieure un peu plus étroite a Pextrémité antérieure qu’a la base, le bord antérieur de cette lame est inerme. Le lobe interne est trés développé et séparé par une profonde encoche du lobe distal fort saillant. Les spécimens examinés présentent une coloration générale blanchatre. La pilosité est assez peu développée si on la compare a celle des autres Paguristes de la cote occidentale d’Afrique. Les poils, bruns, assez longs, mais peu serrés, retiennent cependant la vase et cachent en partie le tégument. Remarques: Par leurs pédoncules antennaires qui dépassent les cornées de la moitié de leur dernier article, par la gracilité de leurs pattes ambulatoires et l’absence de dents ou d’épines sur leurs deux derniers articles, ces Paguristes de petite taille (le 3S de 4 mm. et la 2 de 3:5 mm. sont adultes) ne risquent pas d’étre confondus avec une autre espéce. On ne peut guére leur trouver d’affinités qu’avec P. microphthalmus et P. rubrodiscus comme nous l’avons signalé a la suite de la description de ces deux formes. L’espéce n’était connue que par les deux spécimens capturés par le ‘““Com- mandant Parfait” au large du Liberia a une profondeur non précisée, décrits par A. Milne-Edwards et E. L. Bouvier et par des spécimens de Gold Coast et du Dahomey signalés par Balss. Le male dragué par l’Expédition ‘“‘M’Bizi” provient d’une localité beaucoup plus méridionale, située au large du Congo, et d’une profondeur de 12 a 15 m. Paguristes skoogt Odhner Paguristes skoogt, Odhner, 1923, p. 6. pl. 1. Parmi les Pagurisies des cOtes occidentales ou méridionales d’Afrique que nous avons examinés, aucun n’était identifiable a cette espéce, tout au moins si l’on se 198 ANNALS OF THE SOUTH AFRICAN MUSEUM rapporte a la description qui en a été faite par Odhner. D’autre part, il ne nous a pas été possible de découvrir ou étaient conservés les spécimens en constituant le type. La description qui figure ci-dessous a été établie d’apres celle d’Odhner. Description: L’écusson céphalothoracique est un peu plus long que large et représente un peu plus de la moitié de la longueur totale de la carapace. Le rostre, court et obtus, n’atteint pas l’alignement des dents latérales. Les pédoncules ocu- laires sont cylindriques: leur longueur est égale aux 3/4 de celle de la région anté- rieure de la carapace. Les écailles oculaires sont assez écartées, un peu plus longues que larges et armées de trois dents sur le bord antéro-latéral, en arriére de l’épine distale. Les pédoncules antennulaires dépassent les cornées de la moitié de leur dernier article; les pédoncules antennaires sont aussi longs que les pédoncules oculaires. Les écailles antennaires atteignent le milieu du dernier article des pédon- cules; elles présentent quatre ou cinq dents sur le bord interne et trois dents sur le bord externe. La longueur du flagelle est légerement supérieure a la moitié de celle de la carapace. Le chélipéde droit est un peu plus long que le gauche. Le carpe et le propode sont recouverts de petites épines sans pointes cornées; le bord interne de chacun de ces articles est armé de quatre a cinq dents plus fortes. Les doigts sont en contact sur toute leur longueur. 3 Les pattes ambulatoires p. 2 et p. 3 dépassent les chélipédes; le rapport des longueurs du propode et du dactyle est égal a 2/3 environ. Les pattes de la premiére paire (p. 2) sont armées de dents sur le bord supérieur du carpe et du propode et de petites épines localisées sur les bords supérieur et inférieur du dactyle. Les pattes suivantes (p. 3) ne présentent qu’une épine distale sur le bord supérieur du carpe. Paguristes skoogt, comme la plupart des espéces ouest-africaines est assez fortement pileux. Les régions ou les poils sont les plus longs et les plus denses sont les aires branchiales de la carapace, la face supérieure des chélipédes, les faces supérieure et inférieure des pattes ambulatoires. Remarques: Le dessin et la description de Paguristes skoogi donnés par Odhner, suffsants pour distinguer cette forme des seules espéces d’Afrique occidentale con- nues jusque la, P. hispidus, P. oculatus et P. mauritanicus, présentent quelques impréci- sions qui rendent difficile la comparaison avec les autres Paguristes recemment deécrits. L’espéce la plus proche est sans doute Pagauristes virilis Forest, caractérisée par un écusson céphalothoracique trés court, des écailles antennaires a bords latéraux réguliérement dentés, des pattes ambulatoires a dactyle long et gréle, et des pleopodes 1 trés allongés chez le male. Les caractéres suivants, décrits ou figurés par Odhner, nous obligent a con- sidérer pour l’instant, P. skoogi comme une espéce distincte de P. wirilis: Pécusson céphalothoracique est un peu plus long que large, les pédoncules oculaires ont une longueur égale aux trois quarts de celle de la région antérieure de la carapace, les écailles antennaires sont armées de trois dents sur le bord externe et non de cing ou six, les pédoncules antennaires sont aussi longs que les pédoncules oculaires et non nettement plus courts. PAGURISTES D’ AFRIQUE L’OUEST ET DU SUD 199 Ik yy y Uy, 59 4 Figs. 55-68. Deuxiéme pléopode sexuel droit chez les Paguristes $ des cétes occidentales et méridionales d’Afrique: 55, P. marocanus A. M.-Edw. et E. L. Bouvier; 56, P. mauritanicus Bouvier; 57, P. fagei Forest; 58, P. oxyacanthus Forest; 59, P. rubrodiscus Forest; 60, P. micro- phthalmus Forest; 61, P. difficilis Forest; 62, P. hispidus A. M.-Edw. et E. L. Bouvier; 63, P. virilis Forest; 64, P. engyops Barnard; 65, P. gamianus H. M.-Edw.; 66, P. barnardi sp. nov. 67, P. agulhasensis sp. nov.; 68, Paguristes sp. (Table Bay). Figs. 55, 60, 62, 67, x 35; 58, 65, 68, x 22. Les autres, x 28. 200 ANNALS OF THE SOUTH AFRICAN MUSEUM L’examen des six exemplaires d’Odhner, s’il est possible de les retrouver, ou de spécimens provenant de la méme région que le type (Port Alexander, Angola, 72 m. de profondeur) et la comparaison des pléopodes du male, qui, chez P. wirilis, sont si caractéristiques, confirmera la validité de cette derniére espéce ou permettra de Pidentifier a Paguristes skoogi. Paguristes gamianus (H. Milne-Edwards) Figs) if 14 125/29. 90)05 et ply ave Pagurus gamianus, H. Milne-Edwards, 1836, p. 283; 1837, p. 235. Paguristes gamianus, Stimpson, 1858, p. 74. Paguristes rosaceus, Barnard, 1947, Pp. 3753 1950, p. 420, fig. 78 e-g. et non Paguristes gamianus, Stebbing, 1910, p. 351. 29 > oo Balss}) TO12)) pl On, tiga o: yr) 29 oy) Stebbing, 1920, p.- 257. 9 9 », Barnard, 1950, p. 418, fig. 78 a—d. Matériel examiné: 2 3, l.c. 12 et 15 mm.; 2 9, Lc. 11 et 14.mm.: Cap de Bonne- Esperance, Raynaud coll. (Type). 1 9, lc. 10 mm.; 2 9 ovigéres, l.c. 11 et 11-5 mm.: Buffels River, 23-2-40. 3 9, lc. 10 42 14 mm.; 1 9 ovigere, l.c. 11 mm.: Hondeklip Bay, 8-2-40. — 2 9, lic. 10 et 11 mm.: Groen River, 7-3-40 1 9, lc. 10 mm.: Zout River, 19-4-40. 1 Q juv., lc. 6 mm.: Langebaan Lagoon, 16-7-46; 1 9 ovigere, l.c. 11 mm.: Langebaan Lagoon, 26-4-48. 1 9, lc. 7 mm.; 1 9 ovigére, l.c. 10 mm.: Langebaan Lagoon. 1 9, lc. 13°5 mm.: Table Bay, dragage par 17 m. 1g, l.c. 8 mm.: St. James (False Bay), 23-2-33. 1 9, lc. 16 mm.: False Bay, dragage par 24 m., 22-2-52. 1g, 1l.c.8mm., 1 9, l.c. 16 mm.: False Bay, dragage par 11-12 m., 5-3-52. I g juv., l.c. 6 mm. (avec un orifice sur la coxa de la troisieme patte thoracique droite); 4 9 juv., Le. 4°5.45:°5mm.; 2 9, l.c. 6:5 et 7 mm.: Cape Hangklip, 8-1-40. 4 2 ovigeres, l.c. 7 a 11°5 mm.: Hermanus, 30-6-39. 2 9 juv., l.c. 4 et 5 mm.; 2 9 juv., l.c. 5 mm.: Danger Point, 5-7-39. 2 Ole: 7et 7.5 mm) e325 mm.) Dancertone 2 9, l.c. 7 et 8 mm., 1 9, I.c. 7-5 mm.: Keurbooms River, Plettenberg Bay (Type de P. rosaceus Barnard). Tous ces échantillons—a l’exception du premier qui est le type de H. Milne- Edwards et qui est conservé au Muséum de Paris—proviennent des collections du Département de Zoologie de |’Université de Cape Town et du South African Museum. Ce matériel nous a été aimablement communiqué par le Dr. K. H. Barnard, Directeur du South African Museum. Description: L’écusson céphalothoracique est un peu plus long que la région postérieure de la carapace et le rapport de sa largeur sur sa longueur est égal a 4/5 environ. Le rostre aigu dépasse nettement les dents latérales. Les pédoncules PAGURISTES D’ AFRIQUE L’OUEST ET DU SUD 201 oculaires sont renflés a la base et plus faiblement au niveau des cornées qui sont assez petites et présentent une échancrure postérieure. Leur longueur représente les 5/9 environ de celle de la région pré-cervicale. Les écailles oculaires largement écartées, sont lancéolées et ne présentent pas d’indentations sur le bord antéro- latéral. Les pédoncules antennulaires sont a peu pres de méme taille que les pédon- cules oculaires. Les pédoncules antennaires n’arrivent pas tout a fait a la base des cornées; leur second article est armé de trois dents aigués a langle antéro-externe qui est assez saillant. L’extrémité bidentée de l’écaille antennaire atteint ou dépasse de peu le milieu du dernier article; son bord interne, sensiblement rectiligne, est armé de trois dents assez peu saillantes; son bord externe présente au moins une dent dans la région antérieure. Le flagelle est presque aussi long que la région antérieure de la carapace. Les chélipédes sont égaux et de méme forme. Le carpe est de méme longueur ou légérement plus court que le dactyle et plus long que la région palmaire. La main est un peu plus de deux fois et demie plus longue que large. Le carpe, le pro- pode et le dactyle, sont couverts par dessus de dents épineuses a pointe cornée, particuliérement développées sur le bord interne des articles. Les pattes ambulatoires sont assez courtes et trapues: elles dépassent légére- ment les chélipedes. Les p. 2 (fig. 38 et pl. ITV) ont un meérus a bord inférieur spinu- leux; le bord supérieur du carpe porte des dents cornées qui existent aussi sur toute la face interne du propode et sont particulicerement développées sur le bord supérieur de cet article ou on observe une rangée de huit dents a trés longue pointe cornée, plus fortes que sur le carpe. Cette rangée se prolonge sur le dactyle sous la forme de dents de plus en plus petites qui se réduisent a des épines cornées dans le voisinage de l’ongle. Le propode est a peine deux fois et demie plus long que haut. Le dactyle a une hauteur a la base comprise cing fois environ dans sa longueur. Le rapport des longueurs du premier et du second de ces articles est sensiblement égal a 3/4. Les p. 3 different des pattes précédentes en ce que leurs dents épineuses, qui occupent les mémes emplacements, sont beaucoup moins fortes. Les femelles possédent un orifice sexuel sur chacune des coxae des p. 3; un repli de abdomen qui prend naissance en arriére du troisieme pléopode impair biramé recouvre les appendices ovigéres. Les pl. 1 du male (fig. 12) ont un article distal en général peu enroulé; la région désignée sous le nom de lame inférieure chez P. oculatus n’est pas nettement séparée du lobe distal. Les pl. 2 (fig. 65) ont une portion distale a peine plus courte que la portion proximale. La région antérieure de la carapace, les pédoncules oculaires, les antennes, les chélipédes et les pattes ambulatoires sont rouge-rosé ou carminé et souvent maculés de pourpre; la moitié distale des dactyles des p. 2 et p. 3 est blanche. Cette colora- tion disparait rapidement dans l’alcool et la plus grande partie des spécimens examinés sont uniformément jaunatres. La pilosité est assez forte sur les chélipédes et sur les deux paires.d’appendices suivantes, mais les poils plumeux fasciculés ne forment pas de revétement continu et ne dissimulent pas complétement l’ornementation du tégument. 202 ANNALS OF THE SOUTH AFRICAN MUSEUM Remarques: Le type de Paguristes gamianus H. Milne-Edwards est représenté par 2 get 2 2 adultes provenant du Cap de Bonne-Espérance. Le spécimen signalé par Balss (1912), et ceux de Stebbing redécrits par K. H. Barnard, different considérable- ment du type et appartiennent beaucoup plus probablement a l’espéce décrite plus loin sous le nom de P. macrotrichus. Par contre le Paguristes rosaceus que K. H. Barnard a décrit sans avoir pu le comparer au type de l’espece de H. Milne-Edwards est bien a mettre en synonymie avec celle-ci. Nous décrirons plus loin sous le nom de Paguristes sp. un spécimen apparenté a Paguristes gamianus mais en différant cependant sur quelques points assez importants: extrémités bidentées des écailles oculaires, forme des chélipédes, proportions des deux derniers articles des pattes ambulatoires, ornementation des pléopodes sexuels. Nous indiquerons a la suite de la description de ce Paguristes sp. les raisons qui nous font hésiter a le considérer comme appartenant a une nouvelle espéce. Parmi les Paguristes provenant d’Afrique du Sud et que nous avions déterminés comme P. gamianus figuraient un certain nombre de spécimens qu’il était impossible de distinguer de la forme typique par l’aspect de la carapace et des appendices céphaliques antérieurs mais qui présentaient une pilosité fort différente des chéli- pedes et des deux paires de pattes ambulatoires. Aprés épilation, la forme et l’ornementation des dactyles de celles-ci se sont également révélées nettement différentes; nous avons été ainsi amené a séparer ces individus de Vespeéce de H. Milne-Edwards et a les décrire sous le nom de Paguristes barnard: sp. nov. Les échantillons de P. gamianus proviennent surtout de la zone intercotidale des cétes ouest et sud-ouest d’Afrique du Sud, depuis Buffels River, jusqu’a la région de Cape Town. L’un d’entre eux provient cependant d’un point situé beaucoup plus a Vest, de Plettenberg Bay. Quelques spécimens parmi les plus grands ont été dragués entre 11 et 24 m. de profondeur. Paguristes sp. Figs. 24, 52, 68 Matériel examiné: 1 3, l.c. 20 mm., Table Bay (céte occidentale d’Afrique du Sud, dragage par 17 m. Description: L’écusson céphalothoracique est un peu plus long que la région postérieure de la carapace; le rapport de sa largeur et de sa longueur est égal a 6/7. Le rostre ne dépasse pas l’alignement des dents latérales. Les pédoncules oculaires, renflés 4 la base ont une longueur égale aux 2/3 environ de celle de la région pré- cervicale. Les cornées sont petites et légérement échancrées postérieurement. Les écailles oculaires sont triangulaires et ont un sommet bidenté. Les pédoncules antennulaires sont un peu moins longs que les pédoncules oculaires. Les pédoncules antennaires sont plus courts encore; le second article a un angle antéro-externe fortement saillant et armé de trois dents aigués. L’écaille antennaire atteint le tiers distal du dernier article pédonculaire, son bord interne est armé de quatre fortes dents, son bord externe de trois dents. Le flagelle est aussi long que la région antérieure de la carapace. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 203 Les chélipédes sont égaux et de méme forme. Le carpe est un peu plus court que le dactyle dont la longueur représente plus des 2/3 de celle de la main; celle-ci est sub-triangulaire et plus de deux fois plus longue que large. La face supérieure des trois derniers articles est couverte de fortes dents a pointe cornée, moins nom- breuses sur les doigts ou on observe deux bandes longitudinales inermes. Les pattes ambulatoires sont un peu plus longues que les chélipedes. Les trois derniers articles des p. 2 sont armés par dessus de fortes dents cornées acérées. La face interne des deux derniers est également épineuse, on observe en particulier une rangée longitu- dinale sub-médiane de sept dents sur la face interne du propode. Cet article est a peu prés trois fois plus long que haut et d’un tiers plus court que le dactyle. Celui-ci a une hauteur maximum comprise six fois dans sa longueur. Les pattes de la paire suivante présentent des proportions voisines et des dents cornées implantées dans les mémes régions mais moins fortes. Les pl. 1 du male (fig. 52) ont une lame inférieure a bord antérieur arrondi armé de six ou sept crochets, séparée du processus distal trés long par une incisure bien marquee. L’unique exemplaire ne présente pas de coloration caractéristique. Les trois premieres paires de pattes thoraciques présentent une pilosité assez forte constituée par des touffes de longs poils, plumeux a l’extrémité, particuliére- ment nombreux sur les régions de la face supérieure des chélipedes proches du bord externe, sur la région supérieure du carpe et sur les régions supérieure et inférieure du propode et du dactyle des deux paires ambulatoires. I] n’y a pas de revétement continu de poils plumeux trés serrés sur la face interne du propode de ces appen- dices et les dents épineuses situées sur cette face sont toujours bien dégagées. Remarques: Le spécimen deécrit ci-dessus présente des affinités avec Paguristes gamianus, tout en en différant sur un certain nombre de points. Comme il s’agit d’un individu de taille nettement supérieure a celle des plus grands P. gamianus observés, il est possible que ces différences soient liées a Age et nous avons préféré ne pas le décrire comme espéce nouvelle. Les ressemblances avec Vespece de H. Milne- Edwards portent sur Vaspect général, sur les proportions de la carapace et sur lornementation des trois premiéres paires de péréiopodes. Paguristes sp. se distin- gue de P. gamianus par la briéveté du rostre, par les écailles oculaires dont l’extré- mité antérieure est bidentée, par la forme des chélipedes dont la main est sub- triangulaire et dont le dactyle est trés long, par les proportions des deux derniers articles des pattes ambulatoires p. 2 et p. 3: le propode de ces appendices a une hauteur comprise trois fois environ dans sa longueur: le dactyle une hauteur maximum comprise six fois environ dans sa longueur, alors que ces rapports valent respectivement 2°5 et 5 chez P. gamianus. Un dernier caractére morphologique rend probable l’appartenance de Paguristes sp. a une espéce distincte: les pl. 1 g présentent, sur le bord antérieur de la lame inférieure des crochets bien développées qui n’existent pas chez P. gamianus. Quant aux pl. 2, la forme de leur région distale est particuliére (fig. 68). I] faut noter que ce Paguristes a été capturé en méme temps qu’une femelle de P. gamianus typique dans la région de Table Bay. 204 ANNALS OF THE SOUTH AFRICAN MUSEUM Paguristes engyops Barnard Figs. 25, 40, 51, 64 Paguristes engyops, Barnard, 1947, p. 375; 1950, p. 421, fig. 78 h-z. Materiel examiné: 3 g, l.c.2 a3 mm., 6 Q ovigeres, l.c. 2:25 a 3 mm., Buffels Bay, False Bay, littoral, dans des coquilles de Trochidae (Type). I 4, l.c. 7 mm., Langebaan Lagoon, 26-4-48. 4 9, le. 3 a 7-5 mm., 1 @ ovigére, l.c. 5-5 mm., Langebaan Lagoon, récolte intercotidale. Description: La longueur de l’écusson céphalothoracique est égale aux 7/6 environ de sa largeur et aux 3/5 de la longueur totale de la carapace. Le rostre aigu atteint ’alignement des deux pointes latérales dont il est séparé par deux fortes concavités. Les pédoncules oculaires larges a la base s’amincissent progressivement jusqu’a la région cornéenne qui est légérement renflée; leur longueur représente les 4/7 environ de celle de la région pré-cervicale; les cornées ne sont pas échancrées postérieurement. Les écailles oculaires sont rapprochées, assez grandes, lancéolees, uni- ou, rarement, bidentées a l’extrémité avec parfois une épine vers le tiers distal du bord interne qui est fortement convexe. Les pédoncules antennulaires dépassent les yeux de la moitié environ de la longueur de leur dernier article. Les pédoncules antennaires atteignent la base des cornées, leur second article a un angle antéro- externe aigu mais peu saillant. L’écaille antennaire courte et large atteint la base du dernier article pédonculaire, son bord interne est rectiligne et inerme ou faiblement denticulé, son bord externe convexe et armé de deux ou trois fortes dents. Le flagelle antennaire est un peu plus court que les pédoncules oculaires. Les chélipédes sont sub-égaux chez les femelles et chez les jeunes males. Chez les males adultes le droit est un peu plus fort que le gauche. Le carpe est un peu plus long que le dactyle. La longueur de ce dernier représente les 2/3 environ de la longueur totale de la main. Il y a quelques tubercules peu saillants sur la face supérieure du carpe, plus forts sur le bord interne. La région palmaire est aussi pourvue de quelques tubercules, plus développés également sur le bord interne. Les doigts sont couverts de petits tubercules arrondis. Une forte dent au milieu de la face interne du doigt fixe se loge dans une concavité bien marquée de la face correspondante du doigt mobile. Les pattes ambulatoires sont trapues; elles sont nettement plus courtes que les chélipédes et comme eux sont faiblement armées: des dents peu nombreuses et émoussées existent sur le bord supérieur du carpe et du propode des p. 2 (fig. 40), alors que le dactyle n’est armé par dessous que de quelques spinules cornées placées en arriére de l’ongle. Les p. 3 sont inermes a l’exception d’une dent distale sur le bord supérieur du carpe et de quelques spinules sur les faces interne et inférieure du dactyle. Dactyle et propode des deux paires sont sub-égaux. Les femelles possédent une paire d’orifices génitaux, et un trés court repli abdominal en arriére du troisiéme pléopode impair. Les pl. 1 du male (fig. 51) ont une lame inférieure spatulée dans la région distale et armée sur le bord antérieur de PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 205 huit a dix crochets. Le lobe interne est presque aussi proéminent que le processus distal dont il est séparé par une profonde encoche. La pilosité est constituée par des touffes de poils plumeux trés denses sur les écailles oculaires, plus ou moins abondants sur les trois premiéres paires d’appendices thoraciques. Remarques: Si la description originale du Paguristes engyops différe quelque peu de celle que nous donnons ci-dessus, c’est que le type est constitué par des individus de trés petite taille. Les individus plus agés ont en particulier une région antérieure de la carapace moins allongée, des pédoncules oculaires plus gréles, des cornées plus renflées que les jeunes. I] est intéressant de noter que l’espéce la plus voisine de P. engyops est P. maro- canus des cétes marocaines. Tous deux sont de petite taille, ont une carapace et un bord frontal d’aspect voisin, des cornées non échancrées postérieurement et dépig- mentées dans l|’alcool, des écailles oculaires 4 peu prés de méme forme, des flagelles antennaires courts et épais, des pléopodes males de méme type. Elles sont cependant bien distinctes: la réduction des cornées chez P. engyops, la forme des chélipédes, les proportions des pattes ambulatoires sont autant de caractéres qui interdisent toute confusion. I] existe aussi entre les deux espéces des différences écologiques considérables. Les spécimens connus de P. marocanus, malheureusement fort peu nombreux, ont été dragués sur des fonds de 115 a 290 m., alors que l’un des spécimens de P. engyops au moins provient de la zone intercotidale, les autres étant seulement qualifiés de “‘littoraux”’ sur les étiquettes. Les plus petits males observés (2 mm. de carapace) ont déja des pléopodes du type adulte, et les plus petites femelles (2:25 mm.) sont déja ovigéres, portant 3 ou 4 oeufs dont le plus grand diamétre est compris entre o-8 et 1-2 mm. Paguristes agulhasensis sp. nov. Figs. 26, 37, 53, 67 Matériel examiné: 1 3, l.c. 5-5 mm., Agulhas Bank, dans une coquille de Fusus rubrolineatus, 55 metres. (Type.) Description: La longueur de Vlécusson céphalothoracique est légérement supérieure a la largeur et sensiblement égale aux deux tiers de la longueur de la carapace. Le rostre, large et aigu, dépasse de peu |’alignement des pointes latérales. Les pédoncules oculaires, épais et sub-cylindriques, sont égaux aux deux tiers de la région antérieure de la carapace. Les écailles oculaires qui sont trés écartées ont une région antérieure triangulaire se terminant par une pointe trés aigués. Les pédoncules antennulaires sont de méme longueur que les pédoncules oculaires. Les pédoncules antennaires n’arrivent pas tout a fait au niveau des cornées, langle antéro-externe du second article est fort saillant et armé de trois dents aigués. L’écaille antennaire dont la pointe dépasse le milieu du dernier article pédonculaire a un bord interne concave armé d’une dent proximale, et trois dents aigués sur le _ bord externe. Le flagelle est un peu plus court que l’écusson céphalothoracique. 206 ANNALS OF THE SOUTH AFRICAN MUSEUM Les chélipédes sont sub-égaux. La longueur du dactyle est sensiblement la méme que celle du carpe et représente les 2/3 environ de la longueur totale de la main. La face supérieure du carpe et du propode est ornée de quelques tubercules a sommet corné mais peu aigu sauf sur le bord palmaire interne, lequel est marqué par quatre fortes dents. Sur la moitié proximale du dactyle dans le prolongement du bord palmaire existent des dents cornées aigués. Les p. 2 (fig. 37) portent sept ou huit petites dents cornées irréguliérement implantées dans la région supérieure du carpe et une rangée de sept dents un peu plus fortes sur le bord supérieur du propode; on observe sur le bord supérieur du dactyle des denticules de taille décroissante, réduits a leur pointe cornée dans le voisinage de Vongle, et, sur le bord inférieur, une douzaine de spinules cornées. Le rapport des longueurs du propode et du dactyle est sensiblement égal a 3/4, et la hauteur maxi- mum du dernier article est comprise cinq fois et demie dans sa longueur. Les p. 3 sont inermes a l’exception du bord supérieur du carpe pourvu d’une dent distale et des régions supérieure et inférieure du dactyle armées de spinules cornées. Les régions les plus pileuses sont la région rostrale, les écailles antennaires, les aires branchiales de la carapace et les appendices thoraciques. Les poils présents sur le propode et sur le dactyle des trois premieres paires de péréiopodes sont dépourvus de barbules, alors que sur les autres régions on observe une forte propor- tion de soies plumeuses. Remarques: La description ci-dessus est celle d’un unique spécimen male, de petite taille mais possédant des pléopodes sexuels bien développés. I présente quelque ressemblance dans l’aspect général avec Pagurisies gamianus et avec le spécimen décrit plus loin sous le nom de P. macrotrichus sp. nov. Paguristes agulhasensis se distingue des Paguristes gamianus de méme taille par la forme du rostre plus large, moins aigu et moins saillant, par les pédoncules oculaires plus épais, par les écailles oculaires plus écartées, par les écailles antennaires armées d’une seule dent proximale sur le bord interne et non de trois ou quatre, et enfin, caractére qui est peut-étre le plus net, par le présence de dents moins nombreuses et beaucoup moins fortes sur le propode et le dactyle des pattes ambulatoires; ces deux articles présentent d’ailleurs une pilosité moins dense et constituée par des poils simples alors qu’on observe une forte proportion de poils plumeux chez P. gamianus. I] faut noter que les premiers pléopodes sexuels du spécimen étudié (fig. 53) semblent avoir atteint leur taille et leur forme définitive ce qui permet de croire que nous avons affaire a un adulte, alors que chez les P. gamianus de méme taille, soit 5°5 mm. de carapace, ils sont plus courts et n’ont pas acquis leur plein développe- ment. La structure de ces appendices ne permet cependant pas de distinguer les deux espéces. La ressemblance de P. agulhasensis et du spécimen femelle décrit plus loin sous le nom de Paguristes macrotrichus nov. sp. est sans doute plus grande notamment en ce qui concerne les denticules implantés sur les deux derniers articles des pattes ambulatoires p. 2 et p. 3. P. macrotrichus présente cependant plusieurs caractéres qui ne permettent guére de l’identifier 4 P. agulhasensis pour l’instant et en particulier un rostre plus court et arrondi, des écailles oculaires plus petites et plus écartées PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 207 d U Paguristes oculatus type P.oculatus rubro-=pictus P,marocanus “Pp. fagei Vid P,mauritanicua Pecitiiciiitis P,oxyacanthus P.,rubrodiscus P.microphthalmus P.virilis P,hispidus P,skoogi P.gamianus P,engyops P,agulhasensis P,barnardi no fe O. . P,macrotrichus O =) & ® A Vv A ss) O eal XK MM =| | r | O S @ @ @ Fig. 69. Le genre Haguristes sur les cOtes occidentales et méridionales d’Afrique. Les espéces chez lesquelles les femelles ne possédent qu’un orifice génital sont représentées par des signes carrés. 208 ANNALS OF THE SOUTH AFRICAN MUSEUM encore, des écailles antennaires armées de plusieurs dents sur leur bord interne, une zone allongée dépourvue de tubercules ou de dents sur les doigts des chélipédes et enfin une pilosité plus forte, qui sur le propode et le dactyle des pattes ambulatoires est constituée par de longs poils plumeux. Paguristes barnardi sp. nov. Figs. 39, 54, 66, et pl. iv. Paguristes rosaceus (pars), Barnard, 1947, p. 375. Matériel examiné. 1 3, l.c. 6 mm., 1 ¢ juv., le. 4 mm., 1 9 juv., Le. 4°5 mm.: Kleinmond, 19-3-309. I g, l.c. 10 mm.: Kleinmond, 20-3-39. 5 d, lc. 8a 13 mm., 1 Q, lc. 5-5 mm.: Cap Agulhas, 29-9-39. (Syntypes.) 4 4, le. 5:5 a 7°5 mm., 1 Q ovigere, l.c. 6 mm.: Arniston, 23-11-39. 5 6, lc.6a10mm.: Still Bay, 10-11-39. (Syntypes.) I 4, l.c. 6-5 mm., 3 d juv., l.c. 3a 4 mm., 1 9 juv., l.c. 4 mm.: Knysna, 6-4-30. 2 4, l.c. 7 et 10 mm., 1 ¢ juv., l.c. 5 mm., 1 9 ovigere, |.c. 5 mmesoy Sune 2:5 et 3-5 mm.: Storms River, 2-4-39. (Syntypes.) I Q, lc. 5 mm., 1 Q, l.c. 6-5 mm.: Jeffries Bay, 1-6-30. 1 Q, l.c. 7 mm.: Port-Elisabeth. Tous les échantillons appartiennent a |’Université de Cape Town ou au South African Museum; ils ont été récoltés dans la zone intercotidale. Description: En ce qui concerne la carapace et les appendices céphaliques antérieurs la description de P. gamianus peut s’appliquer a P. barnardi. Les différences entre les deux espéces portent surtout sur les trois premiéres paires de péréiopodes et sur la pilosité. Dans l’espéce nouvelle on observe les caractéres suivants: Le carpe des chélipédes est un peu plus long que le dactyle et sa face supérieure est armée de dents cornées plus nombreuses que chez P. gamianus. Le carpe des p. 2 (fig. 39) est aussi plus épineux. Le dactyle des p. 2 et des p. 3, un peu plus long que le propode, est peu déprimé latéralement; son bord supérieur ne porte pas de fortes dents a base calcifiée, mais de nombreuses spinules cornées de tres petite taille. Comme chez P. gamianus les pl. 1 ¢ (fig. 54) sont peu enroulés, l’encoche qui sépare le lobe interne du lobe distal est en général plus profonde que dans l’espéce de H. Milne-Edwards, mais ce caractére n’est pas assez net pour qu’on puisse lui accorder une valeur discriminatoire. Par contre la portion distale du dernier article des pl. 2 ¢ (fig. 66) est bien plus courte que la portion proximale, alors que les deux régions sont sub-égales chez P. gamianus. Les régions calcifiées des spécimens les moins décolorés sont teintées de rose. Le carpe et le propode des chélipédes ont une face supérieure bordée latérale- ment d’une frange trés dense de longs poils plumeux et les dents cornées présentes sur cette face sont en grande partie dissimulées par un revétement trés épais de poils plumeux plus courts. La face supérieure du carpe, les faces supérieure et inférieure du propode et du dactyle des deux paires d’appendices suivantes sont également garnies d’une épaisse frange de longs poils plumeux qui cachent les dents et spinules PAGURISTES D’AFRIQUE L’OUEST ET DU SUD 209 présentes sur ces régions (pl. iv). Comme sur les chélipédes, il y a sur la face interne du propode un feutrage continu de poils courts. Remarques: L’examen des Paguristes figurant dans les collections du South African Museum et de |’Université de Cape ‘Town sous le nom de P. gamianus nous a amené tout d’abord a en séparer un spécimen qui sera décrit sous le nom de P. macrotichus sp. nov. Les autres peuvent étre rassemblés en deux groupes: le premier comprend les individus identifiables au véritable P. gamianus de H. Milne-Edwards et en particulier le type de P. rosaceus Barnard. Le second est constitué par d’assez nombreux spécimens fort proches de P. gamianus par l’aspect de la carapace et des appendices céphaliques mais présentant des différences marquées dans l’ornemen- tation des dactyles des pattes ambulatoires et surtout dans la pilositeé. La constance de ces différences, absence de formes intermédiaires et certaines considérations biologiques nous ont décidé a faire du second groupe une espéce nouvelle que nous sommes heureux de dédier au Dr. K. H. Barnard. Les caractéres permettant de distinguer P. barnardi de P. gamianus ont été indiqués ci-dessus. L’aspect du dactyle de la premiére ou de la seconde patte ambulatoire montre immédiatement a quelle espéce on a affaire: chez P. gamianus cet article est déprimé latéralement et armé par dessus de fortes dents a base calcifiée, bien visibles en raison du peu de développement de la pilosité (pl. 1V en bas); chez P. barnard:, sa section est sub-circulaire et il présente par dessus et par dessous une frange trés fournie qui couvre complétement les régions supérieure et inférieure et cache la totalité des spinules cornées (pl. IV en haut). I] faut noter que P. gamianus est de plus grande taille que P. barnardi. Le tableau ci-dessous donne la taille moyenne (I.c.) en millimétres pour les différentes catégories dindividus: 3 juvéniles, ¢ adultes, 2 juvéniles, 2 adultes (ovigéres incluses), et 9 ovigeres. gd juv. adultes Q juv. Qadultes Qovig. g¢ + Q adultes P. gamianus 3) 10-7 5:2 9:7 9-9 9-8 P. barnardi 3:8 6-4 39 Oyen a) 5:8 7-8 Le nombre d’individus examinés, quarante-quatre pour la premiére espéce, trente-quatre pour la seconde, est suffisant pour que les chiffres obtenus aient une signification; il faut retenir en particulier la différence de taille 4 la maturité sexuelle: la plus petite 9 ovigére mesure 7 mm. chez P. gamianus et 5 mm. seulement chez P. barnardt. : Dans l’ensemble la répartition géographique des deux espéces présente aussi des différences: les échantillons de P. gamianus proviennent pour la plupart des cétes ouest de l’Afrique du Sud alors que les P. barnardi ont été capturés sur la céte sud. Sur les 13 localités ou ont été récoltés les P. gamianus, 12 se trouvent a l’ouest et au nord de Danger Point, cap situé dans la région occidentale de la céte sud. Sur les 8 localités de P. barnardi, 7 sont a est de ce cap. Sur la figure 70 les courants marins ont été schématiquement représentés, ainsi que les points ou ont été récoltés P. gamianus et P. barnardi. La premiere espéce a surtout été récoltée sur les cdtes baignées par les courants froids (courant du Benguela) et la seconde, sur la céte sud, la ou le 210 ANNALS OF THE SOUTH AFRICAN MUSEUM courant chaud des Aiguilles exerce son influence. La région ou les deux espéces se rencontrent l’une et l’autre est celle du Gap et de Danger Point qui voit aussi la confluence des eaux chaudes et froides. En ce qui concerne le niveau de récolte, notons que tous les P. barnard: se trouvaient dans la zone intercotidale comme beau- coup de P. gamianus mais que quelques-uns de ceux-ci ont été dragués entre 11 et 24 m. CAPE TOWN O} ™~ NG Fig. 70. Détail de la distribution des Paguristes sur les cotes d’Afrique du Sud. Pour l’expli- cation des signes voir fig. 69. Courants chauds — — -— et courants froids ———. PAGURISTES D’AFRIQUE L’OUEST ET DU SUD QI Paguristes macrotrichus sp. nov. Figs. 27, 41 Paguristes gamianus, Stebbing, 1910, p. 351. .F ah Balss, 1912, p. 91, fig. 2. s i Stebbines 19205 py 2577. y se Barnard, 1950, p. 418, fig. 78, a-d. Matériel examiné: 1 9, l.c. 7 mm., région de Natal, 90 metres, identifiée par Stebbing comme P. gamianus H. Milne-Edwards. (Type.) Description: La longueur de lécusson céphalothoracique est légerement supér- ieure a sa largeur et représente les deux tiers environ de la longueur totale de la carapace. Le rostre arrondi atteint ’alignement des pointes latérales. Les pédon- cules oculaires sont sub-cylindriques, assez épais, faiblement dilatés au niveau des cornées qui présentent une échancrure postérieure. Leur longueur est égale aux 3 de celle de la région pré-cervicale. Les écailles oculaires sont petites, triangu- laires dans la région distale, et trés écartées. Les pédoncules antennulaires sont un peu plus longs que les pédoncules oculaires, ceux-ci dépassant les pédoncules antennaires du tiers de leur longueur. Les écailles antennaires sont longues; leurs bords latéraux sont armés de trois a cing dents irrégulicrement développées. Les flagelles antennaires ont la méme longueur que l’écusson céphalothoracique. Les chélipédes sont égaux. Le carpe est plus court que le dactyle et celui-ci a une longueur égale aux deux tiers de la longueur totale de la main. La face supér- ieure de ces trois articles porte des tubercules sauf dans les régions voisines des bords de contact des doigts. Il y a cing dents corneées sur le bord interne du carpe, quatre sur le bord palmaire interne, et d’autres plus nombreuses mais plus petites et plus aigués sur les régions externes des doigts. Les p. 2 (fig. 41) sont armées de quelques dents cornées sur la face supérieure du carpe, de sept dents plus réguliéres sur le bord supérieur du propode et d’une rangée de denticules de taille décroissante sur le bord supérieur du dactyle: le bord inférieur de cet article est également armée d’épines cornées assez longues. Le rapport des longueurs du propode et du dactyle est sensiblement égal a trois quarts. Les p. 3 sont inermes a l’exception du bord supérieur du carpe armé d’une épine distale et de la face interne du dactyle qui présente deux rangées longitudinales de spinules cornées, lune prés du bord supérieur l’autre prés du bord inférieur. Un repli abdominal recouvre les trois premiers pléopodes du spécimen étudié qui est une femelle non ovigére. La pilosité est surtout constituée par de longs poils fins et plumeux assez denses mais ne dissimulant pas l’ornementation du tégument. Les régions les plus pileuses sont: le bord frontal, les écailles oculaires et antennaires, les trois premiéres paires de péréiopodes et particulicerement les régions supérieures de leurs trois derniers articles ou l’on observe des faisceaux de longs poils insérés a la base des dents et des tubercules ou dans des dépressions du tégument. Remarques: Le spécimen étudié ci-dessus a été déterminé par Stebbing, décrit et figuré par Barnard sous le nom de Paguristes gamianus. 11 appartient probablement DP. ANNALS OF THE SOUTH AFRICAN MUSEUM a la méme espéce que les spécimens décrits par Balss sous le méme nom. En fait, il s'agit d’une forme bien différente de l’espéce de Milne-Edwards, comme en témoigne la comparaison des figures 23 et 27 pour la région antérieure de la carapace et les appendices céphaliques antérieurs, et 38 et 41 pour les pattes ambulatoires. L’éco- logie des deux espéces est aussi bien différentes: P. macrotrichus a été capturé par go m. (Stebbing) et 150 m. (Balss) de profondeur, alors que les véritables P. gamianus proviennent de la zone intercotidale ou d’une profondeur n’excédant pas 24 m. P. macrotrichus parait plus proche du spécimen deécrit plus haut sous le nom de Paguristes sp. qui est soit un P. gamianus trés agé, soit une espéce nouvelle. A ne considérer que l’aspect de la région antérieure du corps, 11 parait hasardeux d’identi- fier les deux spécimens a une méme espéce. Les pédoncules oculaires sont bien plus eréles chez Paguristes sp. qui, par ailleurs, a des écailles oculaires bidentées, mais il faut faire preuve de beaucoup de prudence dans l’utilisation de caractéres présentés par des individus de taille trés dissemblable et, qui plus est, de sexes différents. La comparaison des autres régions du corps ne peut trancher la question. En effet si les chélipédes sont d’une forme voisine, les pattes ambulatoires paraissent moins fortement armées chez P. macrotrichus: chez celui-ci on observe seulement sur la face interne du propode de peu nombreuses et minuscules spinules alors qu’il existe en cet endroit de fortes dents cornées chez Paguristes sp. La différence dans la denticu- lation est ici bien plus accentuée qu’entre les plus petits et les plus grands spécimens de P. gamianus. La comparaison entre P. macrotrichus et P. agulhasensis sp. nov. figure a la suite de la description de ce dernier. Parmi tous les Paguristes d’ Afrique du Sud que nous avons eu l’occasion d’étudier, Punique spécimen femelle de P. macrotrichus est celui qui provient de la localité la plus orientale (Natal) et de la plus grande profondeur. BIBLIOGRAPHIE Alcock, A., 1905. Catalogue of the Indian Decapod Crustacea in the collection of the Indian Museum. Pt. 2. Anomura, fasc. 1. Pagurides, i-xi, 1-197, 16 pl. Balss, H., 1912. ‘Paguriden.’ Wiss. Ergebnisse der deutsche Tiefsee-Exp. “Valdivia”, Iena, 20, no. 2, 87-124, pl. 7-11. , 1921. ‘Crustacea VI: Decapoda Anomura und Brachyura.’ Beitrége zur Kenntnis der Meeres- fauna Westafrikas, Hamburg, 3, no. 2, 13-67. , 1940-44. “Decapoda: Morphologischer Teil.” (H. G. Bronns: Klassen und Ordnungen des Tierreichs, Leipzig, 5 [1], 7. Buch, 22-660). Barnard, K. H., 1947. ‘Description of new species of South African Decapod Crustacea with notes on synonymy and new records.’ Ann. Mag. Nat. Hist. (11) 13, 361-92. » 1950. ‘Descriptive Catalogue of South African Decapod Crustacea (Crabs and Shrimps).’ Ann. S. Afr. Mus., 38, 1-837, 154 figs. Boas, J. E. V., 1880. ‘Studier over Decapodernes Slaegtskabsforhold.’ K.D. Vid. Selsk. Skr. 6 R. nat. og. math. Afd., Kobenhavn, J, 2, 26-210, 7 pl. , 1926. ‘Zur Kenntnis Symmetrischer Paguriden.’ Biol. Meddel. K. Danske Vid. Selsk, 5, 6, 1-52, 25 fig. PAGURISTES D’ AFRIQUE L’ OUEST ET DU SUD 213 Bouvier, E. L., 1892 ‘Etude sur les Paguriens recueillis par le Dr. Jousseaume sur les cétes de la Mer Rouge.’ Bull. Soc. Philom. Paris, 1891-92, sér. 8, 4, 50-5. , 1906 a. ‘Sur les Crustacés Décapodes marins recueillis par M. Gruvel en Mauritanie.’ Bull. Mus. Nat. Hist. Nat., Paris, 12, no. 4, 185-7, fig. 1. , 1906 b. Mission des Pécheries de la céte occidentale d’ Afrique, no. 7, 95-97- , 1906 c. Act. Soc. Linn., Bordeaux, J (61), 198-200. Brocchi, M., 1875. ‘Recherches sur les organes génitaux males des Crustacés Décapodes.’ Ann. Sc. Nat. Kool. (6), 2, 1-131, pl. 4-10. Cavolini, 1787. Memoria sulla generazione dei Pesci e dei Granchi, Naples. Chevreux, M. et Bouvier, E. L., 1892. “Voyage de la Goélette “‘Melita’’ aux Canaries et au Sénégal, 1889-90, Paguriens.’ Mem. Soc. Zool. France, 5, 83-144, pl. I-IV. Fabricius, J., 1775. Systema Entomologiae. Forest, J.. 1952 a. ‘Remarques sur les genres Diogenes Dana et Tr Aalnennts Henderson... .’ Bull. Inst. Roy. Sci. Nat. Belgique, 28, no. 11, 1-15, 15 fig. , 1952 6. ‘Notes préliminaires sur les Paguridae des cétes occidentales d’Afrique. II. Diagnose sommiaire de six espéces nouvelles appartenant au genre Paguristes Dana.’ Bull. Mus. Nat. Hist. Nat., Paris, 2e sér., 24, no. 3, 257-62, figs. » 1952 c. ‘Contributions a la revision des Crustacés Paguridae. 1. Le genre Trizopagurus.’ Meém. Mus. Nat. Hist. Nat., Paris. sér. A, Zool. 5, no. 1, 1-40, 25 figs. Krauss, F., 1843. Die Siidafrikanischen Crustaceen, Stuttgart, 1-68, pl. 1-4. Miers, E. J., 1881. ‘On a collection of Crustacea made by Baron Hermann Maltzam (sic = Maltzan) at Gorée Island, Senegambia.’ Ann. Mag. Nat. Hist., London, (5) 8, no. 45-47, 204-20, 259-81, 364-77, pl. 13-16. Milne-Edwards, A., et Bouvier, E. L., 1891. ‘Sur les modifications que présentent les Pagures suivant le sens de l’enroulement de la coquille qu’ils habitent.’ Bull. Soc. Philomath. Paris, sér. 8, 3, 151-3. ———, 1892. “Observations préliminaires sur les Paguriens recueillis par les Expéditions du ‘““Travail- leur” et du “Talisman”’.’ Ann. Sci. Nat. Zool., sér. 7, 13, 185-226. , 1900. “Expéditions du ““Travailleur” et du ‘““Talisman’’.’ Crustacés Décapodes 1. Paris, 1-396 B2nple Milne-Edwards, H., 1836. “Observations zoologiques....’ Ann. Sci. Nat. Zool., Paris, sér. 2, 6, 257-88, 2 pl. Monod, Th., 1933. “Sur quelques Crustacés de l’Afrique occidentale.’ Bull. Com. Et. Hist. Scient. A.O.F., 15, no. 2-3, 456-548, figs. Nobili, G., 1906 a. ‘Faune carcinologique de la Mer Rouge. Décapodes et Stomatopodes.’ Ann. Sc. Nat. Zool., Paris, 1, 1-347, pl. I-XI. , 1906 6. ‘Mission J. Bonnier et Ch. Pérez (Golfe Persique 1901). Crustacés Décapodes et Stomatopodes.’ Bull. Sci. France-Belgique, Paris, 40, 13-159, pl. II-VII. Odhner, Th., 1923. ‘Marine Crustacea Podophthalmata aus Angola und Siid-Afrika gesammelt von H. Skoog 1912.’ Géteborg K. Vet. Samh. Handl., 27, no. 5, 1-39, figs. Rathbun, M. J., 1900. “The Decapod Crustaceans of West Africa.’ Proc. U.S. Nat. Mus., Washington, 22, no. 1199, 271-316. Schmitt, W., 1926. ‘The Macruran, Anomuran, and Stomatopod Crustaceans collected by the American Museum Congo Expedition, 1909-1915.’ Bull. Amer. Mus. N. H., New York, 53, 1-67, g pl. Stebbing, T. R. R., 1g10. ‘General catalogue of South African Crustacea.’ Ann. S. Afr. Mus., 6, 281-593, pl. XV—-XXII. , 1920. “South African Crustacea.’ IX. ibid., 17, pt. 4, 231-72, pl. XVIII-XXVII. Stimpson, W., 1858. ‘Prodromus descriptionis animalium evertebratorum, quae ... VII. Crustacea Anomoura.’ Proc. Acad. Sci. Philadelphia, 63-90. Hain: ste Pa aie Rhine ae ai: ald) Bie Cee hk : 4 { ~ “4 rel , , ny ‘ 3 Aten Y ' ; 4 ] ; R os . - : 2 Hy Anh : i e ‘ i \ 5 ob ‘ % ce " ; PR A ee a ie ” geet tic tat vate ke ane Oe? Ae CR Rae iy i i Ay ! ‘ ee ip ‘ oaty a aed ‘i 4 py ‘ ‘ AS & a ae, fi w nw ro. « ‘ p a, , Rohe so ‘Pat eco te onan eck WANs co Wd ees om | ne AG). : 5 BY otha iby ao y ¥ A Ls % 4 iT y 5 Yi 1 —< ) Ae its % bes AOS baat EEG Vas Delt ess ty my i fe: mit) sh ‘a i i - * i a rt any: , F e rt k : P-.) jalan’ robs ; 4 99 » at diaphragm aperture: 0:06 — 0:08 mm. NEW SPECIES OF HYDROZOA FROM SOUTH AFRICA PEPE Length of nematotheca plus stalk: 0:05 — 0:06 mm. Female gonotheca, length: 0-50 — 0°83 mm. - ss max. width: 0-14 — 0:29 mm. Male gonotheca, length: 0-49 — 0°56 mm. . i max. width: 0-18 — 0-24 mm. In the paratypes some of the measurements show a slightly larger variation, as follows: Length of hydrothecal pedicel, abcauline: 0-20 — 0:29 mm. ee 55 fy ,, to diaphragm aperture: 0-15 — 0:25 mm. Length of hydrotheca, abcauline: 0:06 — 0-12 mm. Max. width of hydrotheca, at margin: 0-10 — 0-15 mm. Length of nematotheca plus stalk: 0:03 — 0:06 mm. Holotype in Zoology Department, University of Cape Town, no. FB 131 B, from dredgings in False Bay, at 5-8 metres. Numerous colonies on Antenella africana, some bearing gonophores. Paratypes: TB 1 B, FAL 78S, FAL 169 X, FAL 217 N, in Zoology Depart- ment; FAL 78 S also in S.A. Museum. Localities. False Bay, from 4 to 18 metres. Table Bay, from 19 to 20 metres. Remarks. Kygophylax cornucopia differs from other members of the genus in the stolonial nature of the colony, absence of coppinia, position of nematophores, length of hydrothecal pedicel and asymmetry of diaphragm. It is included in the genus VO aese “oun ——— MALO SS fesreens eeee: are ©, SECS Was eS N xe O OE mS si og Fig. 26. Lerneopoda galei Kroyer. a, 9; b, 3 of L. bidiscalis (redrawn from Wilson after Kane). Clavella pagelli (Kroyer). c, 2; d, 3. Remarks. In the present specimen the ovisacs are rather stout and not quite as long as trunk, but may not be fully developed. The 3 here figured belongs to another species, but shows the generic characters. Gen. Clavella Oken 1815. Oken, Lehrb. Nat. (3), 1, p. 358. 1822. Blainville, 7. de Phys., xcv, p. 438 (Lernaeomyzon). 1832. Nordmann, Mikr. Beitr. wirbell. Th., 2, p. 53. (Anchorella, non Cuvier). 1910. Stebbing, Ann. S. Afr. Mus., vi, p. 561 (Lerneomyzon). 1915. Wilson, Proc. U.S. Nat. Mus., xlvii, p. 666 (key to species). 1918. Leigh-Sharpe, Parasitology, xi, p. 118. 1920.,, id., 7. Mar. Biol. Assoc. Plymouth, n-s., xu, p. 332 (excretory system). 292 ANNALS OF THE SOUTH AFRICAN MUSEUM 1925. id., Parasitology, xvii, p. 194 (revision of British species). 1932. Wilson, Bull. U.S. Nat. Mus., no. 158, p. 513. 1933. Saby, Proc. Zool. Soc. Lond., 1933, 4, p. 857 (uncinata, anatomy). 1934. Gurney, ibid., 1934, 2, pp. 196, sqq. (uncinata, development). 1939. Yamaguti, loc. cit., pp. 553-6. 1939. Poulsen, Vid. Medd. Dansk naturf. Foren., cii, p. 223 (uncinata, biology). —Cephalothorax long, bent backwards at an angle with axis of trunk, with or without dorsal carapace. Trunk ovoid, without posterior processes, abdomen, or anal lamellae; an unpaired genital process often present. Antenna I 3-jointed. Antenna 2 uniramous. 2nd maxillae fused, usually short, sometimes absent. g—Cephalothorax and trunk folded together, forming an unsegmented ovoid mass, hind end rounded, not projecting below bases of 2nd maxillae. Antenna 1 2-jointed. Antenna 2 uniramous, 3-jointed. Maxilla 1 bipartite. No anal lamellae. Key to the South African species Cephalothorax @ distinctly longer than trunk. pagelli 2. Cephalothorax 2 about same length as trunk. denticis Clavella pagelli (Kroyer) eo Olena. 1863. Kroéyer, Naturh. Tidsskr. (3), ii, p. 295, pl. 16, fig. 3 (Anchorella p.). 1865. Heller, ‘Novara Crust., p. 242, pl. 24, fig. 6 (A. canthart). 1910. Stebbing, loc. cit., p. 526 (Lerneomyzon p.). 1913. Scott, T. & A., Brit. Parasit. Copep., p. 224, pl. 69, figs. 1-3 (can- thart). 1915. Wilson, loc. cit., pp. 668, 669, and 671, 672 (in key) (canthart and pagellt). 1924. Brian, Parasitol. Maurit., p. 52 (Clavellodes c.). —Cephalothorax distinctly longer than trunk, curved backwards, with dorsal carapace, expanded into 2 lobes where it joins the trunk. Trunk ovoid, usually longer than wide and width usually greater than dorso-ventral depth, but sometimes subspherical, a small knob-like genital process. 2nd maxillae very short, fused, forming a lobe between the lobes of the cephalothorax, with a clavate bulla ending in a brush of minute bristles. Ovisacs about as long as cephalothorax, cylindrical, slightly tapering distally. g—See figure 26 d. Dorso-ventral depth greater than ‘head to tail’ length. 9 cephalothorax 3-4 mm., trunk 2 mm., ovisacs 3 mm., g (major diameter) 0-5 mm. Locality. Cape: on Cantharus bleekeri (sic., now Pachymetopon blochi) (Heller) ; on gills of White Steenbras (Pagellus lithognathus) (S. Afr. Mus. several 99 and g<). Distribution. European seas, on Cantharus and Pagellus. Remarks. T. & A. Scott include pagelli Kréy. with a ?, and adopt Heller’s specific name. Wilson keeps the two separate in his rather unsatisfactory key. SOUTH AFRICAN PARASITIC COPEPODA 293 There would seem to be little doubt that canthart is a synonym of pagelli. Among the present lot those specimens with subspherical trunk have the ovisacs shorter and more distinctly tapering than the others. Clavella denticis (Kroyer) 1863. Kroyer, loc. cit., p. 296, pl. 16, fig. 4 (Anchorella d.). 1665. Heller, loc. cit., p. 243. 1910. Stebbing, loc. cit., p. 562 (Lerneomyzon d.). 1915. Wilson, loc. cit., pp. 668, and (in key) 672. —(Characters from Wilson’s key.) Cephalothorax slender, linear, about same length as trunk; latter ovoid, considerably longer than wide, narrowed anteriorly, posteriorly truncate. Genital process present. Ovisacs thick, ellipsoidal. 2nd maxillae? Locality. Cape, on Dentex rupestris (Heller). Distribution. European seas. Gen. Clavellopsis Wilson 1915. Wilson Proc. U.S. Nat. Mus., xlvii, p. 686 (key to species). 1939. Yamaguti, loc. cit., pp. 556-62. 1950. Kirtisinghe, Parasitology, xl, p. 84. 9—Cephalothorax usually shorter and thicker than in Clavella. Trunk squat, often wider than long, sometimes with posterior processes, but no abdomen or anal lamellae; an unpaired genital process present. Antenna I 4-jointed. Antenna 2 biramous, endopod 1-jointed, exopod 2-jointed. 2nd maxillae short, broad, fused, with folds of skin or swellings, with apical bulla. g—Cephalothorax and trunk at right angles, latter strongly arched dorsally, unsegmented, hind end projecting below bases of 2nd maxillae. Antenna 1 3-jointed. Antenna 2 biramous, each ramus I-jointed. 1st maxillae tripartite. Remarks. Differs from Clavella in general body form, 1st and 2nd antennae, 2nd maxillae in 9, and 1st maxillae in J. Wilson (p. 688) gives a reference to his own original description of Clavella robusta; but the paper is not quoted in his bibliography (p. 721), and seems to have been omitted from the Zoological Record. Key to the South African species 1. No posterior processes. fallax 2. Two dorsal and 2 ventral posterior processes. hostilis Clavellopsis fallax (Heller) Fig. 27 a-c. 1865. Heller, ‘Novara’, Crust., p. 241, pl. 24, figs. 4, 5. 1924. Brian, ‘Parasit. Maurit’., p. 53. 1943. Heegaard, Ark. Zool., xxxiv, A 18, p. 32, figs. 88-94. —Cephalothorax longer than trunk, with dorsal carapace, and a lobe on either side at junction with trunk. Trunk dorso-ventrally flattened, broader than long, front and hind angles rounded, genital process about }-4 length of trunk. 2nd maxillae short, fused, lobulate, bulla mushroom-shaped. Ovisacs stout, cylindrical, apically rounded. 294. ANNALS OF THE SOUTH AFRICAN MUSEUM g—Dorso-ventral depth about equal to ‘head to tail’ length. —Cephalothorax 2°5, trunk 1:5 X 2, ovisacs 2°5-3°5 mm.; ¢ I mm. Localities. On gills of White Steenbras (Pagellus lithognathus) (loc. unrecorded) ; East London, in mouth of Pagrus nasutus. (S. Afr. Mus, 99, 33.) Distribution. Mediterranean, on Dentex vulgaris. Remarks. Kurz (1877. ertschr. Wiss. Zool., xxix, p. 393, pl. 25, figs. 5, 6, 29, 51, 52) has described C. sargi which, according to Wilson’s key (loc. cit., p. 687), differs in having a cylindrical genital process 4 length of trunk instead ) K Weave Fig. 27. Clavellopsis fallax (Heller). a, 2 from Pagellus; 6, 2 from Pagrus, with ¢ attached to head; c, 6. Clavellopsis hostilis (Heller), d, e, dorsal and ventral views 9. of a minute one, and smooth instead of wrinkled 2nd maxillae. Yamaguti (loc. cit., 1939, p. 556, pl. 48, figs. 144-148) records sarg: from Japanese Sparid fishes. : In the South African specimens from Pagellus the thoracic and maxillary lobes vary in distinctness; and the genital process in the specimens from Pagrus is larger and more pointed than in those from Pagellus. Clavellopsis hostilis (Heller) Big 2700 e. 1865. Heller, ‘Novara’ Crust., p. 243, pl. 24, figs. 7, 7 a. 1915. Wilson, loc. cit., pp. 669 and (in key) 702 (? Brachiella). 1924. Brian, ‘Parasitol. Maurit’., p. 51 (Clavella h.). —Cephalothorax curved ventrally, about as long as trunk. Head without carapace. Trunk broader than long, tumid, with dorsal median concavity; 2 dorsal and 2 ventral posterior processes, of about equal length, the 2 dorsal ones far apart, the 2 ventral ones approximate; a small genital process. Antenna I 4-jointed. Antenna 2 biramous. 2nd maxillae short, entirely fused, wrinkled, with small bulla. Ovisacs stout, apically rounded. SOUTH AFRICAN PARASITIC GCOPEPODA 295 Cephalothorax and trunk each 1:25 mm., width of trunk 2 mm., ovisac 2-5 mm. Locality. Natal, on gill-rakers of Umbrina robinson (S. Afr. Mus. 1 9). Distribution. Mediterranean, on Umbrina cirrhosa (Heller). Remarks. The present specimen appears to belong to Clavellopsis (1st and 2nd antennae), and is certainly very similar to Heller’s species. No ¢ is present, and Heller had no g; and as Wilson remarks the 3 is necessary for an exact generic determination. Wilson’s suggestion to transfer Heller’s species to Brachiella seems rather a strange one. C. appendiculata Kirtisinghe 1950, on Chirocentrus, has similar posterior processes. Gen. Charopinus Kroyer 1864. Kroyer, Naturh. Tidsskr., (3), ti, p. 361. (Nomencl. Preuss. Ak., gives date 1863; Marschall, and Neave give 1864.) 1913. Scott, T. & A., Brit. Paras. Copep., p. 188. 1915. Wilson, Proc. U.S. Nat. Mus., xlvii, p. 652 (key to species). 1928. Leigh-Sharpe, Parasitology, xx, p. 276. 1932. Wilson, Bull. U.S. Nat. Mus., no. 158, pp. 511 and (in key) 617, 610. 1939. Yamaguti, loc. cit., p. 552. 1946. Capart, Bull. Mus. Hist. Nat. Belg., xxii, no. 10, pp. 1-6, figs. —Cephalothorax elongate, often flexed backwards, head usually without carapace. Trunk swollen, pear-shaped, flattened dorso-ventrally, with a pair of posterior processes dorsal to the ovisacs; no genital process, abdomen or anal lamellae. Ovisacs long, cylindrical. Antenna 1 indistinctly 4-jointed. Antenna 2 biramous. 2nd maxillae elongate, either fused at their tips with a bulla of varying shape, or each one apically enlarged into a chitinous bar, or knob, or processes. g—Cephalothorax separated by a constriction from the trunk, which is segmented, with anal lamellae. Antenna 1 indistinctly 4-jointed. Antenna 2 biramous. Remarks. Distinguished from Brachiella by the (usually) more elongate 2nd maxillae in 9, and the segmented J. T. & A. Scott (loc. cit., pp. 190, 191) describe the posterior processes as situated ‘ventrally’, and on pl. 55, figs. 1 and 5 label ventral view as ‘dorsal’. Parasitic in the nasal passages or spiracles, or on the gills, of sharks and skates. Charopinus dubius 'T. Scott Fig. 28 1900. Scott, T., 18th Ann. Rep. Fish. Board Scotl., p. 130, pl. 7, fig. 15. EGUg) COLL. & As loc! Cit., p:, 190, pl. 55, He. 5: 1915. Wilson, loc. cit., p. 654 (in key). 9—Cephalothorax slightly shorter than trunk, which is dorso-ventrally flattened, pear-shaped in dorsal view, concave dorsally; posterior processes slightly shorter than trunk. Maxilliped with very small apical spine instead of a 296 ANNALS OF THE SOUTH AFRICAN MUSEUM claw (cf. Wilson, 1915. loc. cit., pl. 41, fig. 113. Brianella corniger). and maxillae longer than trunk, firmly joined at their tips to a large boat-shaped chitinous plate. Cephalothorax 5 mm., trunk 6 mm., width 5 mm., posterior process 5 mm., 2nd maxillae 7 mm., ovisacs 9-10 mm. Locality. Port Elizabeth, on a skate (1 Q sent by Dr. van Hille, Rhodes University, 1948). Distribution. Scotland. Remarks. The ¢ of this species is unknown. Fig. 28. Charopinus dubius T. Scott. a, 9; 6, end view of bulla; c, antenna 1; d, antenna 2; e, maxilliped. Gen Brachiella Cuvier 1830. Cuvier, Régne Anim., ed. 2, iil, p. 257. 1905. Miculicich, Zool. Anz., xxviii, pp. 599 sqq- 1913. Scott, T. & A., Brit. Paras. Copep., p. 203. 1915. Wilson, Proc. U.S. Nat. Mus., xlvii, p. 698 (key to species). 1928. Leigh-Sharpe, Parasitology, xx, p. 25. 1932. Wilson, Bull. U.S. Nat. Mus., no. 158, pp. 520 and (in key) 618, 619. 1933. Saby, Proc. Kool. Soc. Lond., 1933, 4, p. 873, figs. (merluccu and obesa, anatomy). 1939. Yamaguti, loc. cit., p. 566. 1951. Gnanamuthu, Spolia lumpr Chondracanthus merluccti Parabrachiella australis Lepeophtheirus insignis Cecrops latreillea Orthagoriscicola muricatus Philorthragoriscus serratus Pennella filosa 3 orthagorisct Achtheinus dentatus Lernanthropus paradoxus Brachiella sp. Achtheinus dentatus Lerneopoda galei Cardtodectes medusaeus Brachiella macrura] Ancistrotos ostracionis Brachella macrura Clavella pagelli Clavella pagellt Clavellopsis fallax [Caligus ligusticus] [Clavellopsis fallax] Caligus pelamydis Clavella denticis Achtheinus pinguts Lepeophtheirus plotose [Argulus dartevellet | [Caligus mauritanicus | Argulus alexandrensis| Trebius caudatus Lepeophtheirus sp. [Caligus mauritanicus | [ Lepeophtherrus longipes | [Sceaenophilus tenuis] | Polyrhynchus sciaenae| [ Lernanthropus gislert | [ Brachiella sciaenophila] Perissopus dentatus Strabax monstrosus Pandarus armatus Achtheinus dentatus >» pingurs SOUTH AFRICAN PARASITIC COPEPODA Seriola lalandet Caligus lalandet Serranus goliath see Epinephelus Sharks unspecified Alebion carchariae Perissopus dentatus Dinemoura producta ms latifolia Achtheinus dentatus Skate unspecified Charopinus dubius Sparus see Acanthopagrus Sphyraena commersont Argulus belones [Caligus affinis] Pandarus cranchu Sphyrna species Pandarus armatus 5 cranchi [ Nemesis robusta] Squalus acanthias Caligus coryphaenae Achtheinus dentatus Stegostoma fasciatum Pandarus cranchit Temnodon see Pomatomus Tetrodon hypselogeneion Caligus tetrodontis Lepeophtheirus brachyurus Thunnus thynnus [Pennella filosa| Trigla species Medesicaste penetrans Umbrina robinson Clavellopsis hostilis Aiphias gladius [Pennella filosa| 9 -« Crassicornis | Keus [Argulus zei| Kygaena see Sphyrna Mammals Balaenoptera acutorostrata Pennella crassicornis » Species 5, balaenopterae FREsH WATER Amphibians Tadpoles Dolops ranarum Fishes Barbus sp. Lernaea barbicola Barbus gunningt and swierstrae Dolops ranarum Cichlidae [Chonopeltis inermis] [ Lernaea] [ Lamproglena] Eutropius Dolops ranarum Gnathonemus Chonopeltis inermis Heterobranchus [Dolops ranarum] Huro salmonoides (imported) Dolops ranarum Hydrocyon [ Argulus] [ Lates Dolops ranarum| Marcusenius Chonopeltis inermis 312 ANNALS OF THE SOUTH AFRICAN MUSEUM [ Micropterus dolomieu (imported) [ Polypterus Protopterus Sandelia capensis Serranochromis thumbergi Synodontis melanostictus Tilapia mossambica and other species Achtheres microptert | Lernaea haplocephala| | Dolops ranarum| Argulus capensis [ Lamproglena| Chonopeltis inermis Dolops ranarum [ Lernaea] 11. The Breeding and Growth of Hymenosoma orbiculare Desm. (Crustacea, Brachyura). By G. J. BRoEKHUYSEN, PH.D., Department of Zoology, University of Cape Town. (With 13 text figures.) INTRODUCTION The growth and reproduction of the South African shore crab Cyclograpsus punctatus was described earlier (Broekhuysen, 1941). The present paper deals with the Crown crab Hymenosoma orbiculare.. Whereas Cyclograpsus lives on the upper part of the shore and is common among broken rocks, Hymenosoma lives on the lower part of the shore and is restricted to quiet areas where fine sands and mud accumulate. It is thus common in lagoons and estuaries. Cyclograpsus can withstand exposure to air for considerable periods and prefers this, but Hymenosoma either buries itself in damp sand covered by a shallow layer of Fig. 1. Hymenosoma orbiculare Desm. Large specimens from the Sand Vlei Estuary, natural size. Left: male; right: female. Byes) 314 ANNALS OF THE SOUTH AFRICAN MUSEUM water or moves down the shore with the ebbing tide and often extends below tide marks. Thus Barnard (1950) has recorded it from 45 fathoms. Neverthe- less there are certain similarities between the two crabs: both tolerate a very wide range of temperature and salinity. The ecological niche occupied by H. orbiculare shows much similarity to that occupied by Carcinus maenas of the Northern hemisphere (Broekhuysen, 1936). Carcinus, however, seems to prefer a more solid bottom than HAymenosoma, although the species was very common in the Wadden Sea, N.W. of Holland, where the bottom is sandy. Hymenosoma orbiculare Desm. has been described by Barnard (1950). The adult ¢ and 9 are illustrated in fig. 1; and in fig. 2 an example of the first crab (post-larval) stage. During the course of the estuary survey carried out by the University of Cape Town, the species has been recorded from the following localities: Lambert’s Bay, Steenberg’s Cove, Berg Fig. 2. Hymenosoma orbiculare Desm. River Mouth, Saldanha Bay, First crab (post-larval) stage. A Megalopa stage Langebaan Lagoon, Mil- has not been observed. nerton Estuary, Hout Bay, Sand Vlei Estuary, False Bay, Klein River Lagoon, Breede River, Great Brak River Mouth, Knysna Lagoon, Sundays River Mouth, Bushman River Mouth, Keiskamma River Mouth, The Haven, Port St. Johns, Durban Bay, St. Lucia Bay, Kosi Bay, estuaries near Inhambane (Portuguese East Africa). Balss recorded it from Liideritzbucht and Barnard from Olifants River Mouth, so that its total dis- tribution is from South West Africa around the Cape to Portuguese East Africa (see fig. 3). Also recorded from Zanzibar (Lenz, 1905). Hymenosoma geometricum Stimpson was described from False Bay. Barnard (1950) considers this to be merely a deep-water variety of H. orbiculare. ‘The identity of the two species will be discussed further in the present paper. MATERIAL AND METHODS The bulk of the material was collected in Sand Vlei Estuary at Muizenberg between the Vlei Bridge and the Foot Bridge (see fig. 4). During the course of the investigation the estuary was subjected to wide variations in salinity BUD) THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. (E961 ‘€ “TTX “snpy DIN ‘UUp edTIFY YING Jo wun JowYyoA[OJ OY.L, corte ‘Aeq Wos,q) ‘paqoa]]09 us0q sey a4vjNI1G0 "FT 2IdYM WOJJ SUOTLIS YY JO BWIOS SuTMOYs eoUyY UJoYINOG jo deur Vv & “sty 3o9€ oVE ofE o0€ o8TS 09 oVT ofS 007 SVHINOV AdVvD UOISIUSY fieg sfiasyar . B4AIqQGOY, ejuejuj adep | : | usuidoas, \ieg mys. \ HLIGVZITI LUOd S : edoy I {ld aava__7p009 40 adep id jaquymsqiuig GSoVE | AWa \ aS1W4 UMOJSUOWIS-2 coultazteio! CZ fawwoy Bsaquazini eg jnopy-—” jeesyopno—- NMOL AdVI—y UuoJsaupW WEA 9 T1GV.L (= Polsiyy iOd ae 2 puowuiayy¥—s % oce WN}eMwWA—L>—: adVO osuidisi., A 2 NVEuna-s Yes 2 -- do ent er" See o JONVYO $ . yan oa oe of ‘tuo i “NA Meares Ch ey . 09% | “S| gla } | SINOUVW ODNIYNOT is @ ce ce see eeeece em we ereed oes S2ee wees ceresseoce ae ce ceesc-f c ‘ ovZ| sxzt © YBNDTANVSOW_S 00 08S o9T oT ofS o0Z __ ofl o9 | uiog Jagueg SNuUeWIS} diyysuey sdeg Soe. “—NAOL ddWO . ~ UIBJUOJAOZA F\u00ge 7 upeqasuey \ fied eyuepies Jaysousajed A $,8419qQUI91S as) Hote Y heg sjsaquey y InoZ “Y UsOUO heg dyjyapuoy —Yy Sjajjng c—HLOTION LYOd o0f <32> 8% "S] S412] 9U1S wmnaziivaant| : 09f VWIldsAV LSAM HLNOS | obz AWE SIATVAA ssouy 822% adey Jobl 08! oF! ANNALS OF THE SOUTH AFRICAN MUSEUM 316 HLNOW ISA OGNWS “Avg asjeq OjuT Ssutusdo 1a, A puvg jo yyNOPY 9, b “Sty fo \/ ee Y a LS e i es e e THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 317 indicated by a limited number of salinity determinations which have been tabulated in Table I. Table I Salinities determined in the Sand Vlei Estuary during the course of the investigation (see for stations fig. 4). The figures with an asterisk were bottom sample readings. Station || March| April May June July | August | September | December | January I = |8'31-9'0!) 7°4-7°6 | 3°3-31°8 | 075-58 = 33°7-34°3 = rae 2 ay 8-4. | F2-T5) | 28-2570 104-577 | 0-5-0'6* |.8:4-32°3 | 24-2 20°1 3 8-9 79 | 7°2-7°5 | 0°8-25:2 | 0-4-5°3 0:65) 0:0-21-2 21°4 4 g°2 3°3-8°2 | 2:0-23'1 1°5—-28°9 21°5 4°8* 5 2°8 3°3 The salinites given in the above table are in 9/9. Random samples of crabs were collected at regular intervals from February 1947 to February 1948. The total was 1,494 females and 867 males. After sexing, the maximum width of the carapace was measured. The degree of maturity was judged from the shape of the abdomen, and the adults were dissected to determine the developmental stage of the gonads. Berried females were recorded and the developmental stage of the developing embryos noted. During the course of the investigation a considerable number of Hymenosoma were kept in the laboratory. They were in glass dishes with a little water. Sand and shell fragments were provided as a substratum. These crabs were under constant observation. THE REPRODUCTIVE STAGES OF THE MALE A total of 752 males collected from Sand Vlei had a carapace breadth exceeding 14 mm. By dissecting, three stages could be distinguished in the reproductive system: (a) gonads undeveloped or small; (5) gonads developing; (c) gonads fully developed. The monthly changes in the percentage of crabs which fell into these three categories is shown in Table II and text-figure 5. Table IT Developmental Stages of the Gonads of Males from Sand Vlei during the period February 1947 to February 1948 (figures given are percentages). 318 ANNALS OF THE SOUTH AFRICAN MUSEUM Date Gonads undeveloped | Gonads developing | Gonads well developed | Total number of crabs February I 38 61 94 March 5 26 69 175 April 4 13 83 78 May 10 24 65 78 June 4 26 70 23 July ) 28 72 25 August 2 37 61 41 September - - - I October ~ - - 3 November 67 2 31 52 December 21 54 26 39 January 8 4! 51 143 No records are available for September and October as the water-level in the estuary was so high that collecting was difficult. It will be seen that most of the males had fully developed gonads between February and August. In November (and possibly October) they had small or resting gonads and in December to January they were growing. It therefore appears that during spring and early summer most males were unable to copulate as their gonads were not mature. The breeding-season was in late summer and in winter, and covered a period of eight months. It must be stressed that these conclusions refer only to crabs in Sand Vlei Estuary and UNDEVELOPED Ej DEVELOPING UU] WELL DEVELOPER PERCEN TAGE Re ad 8 SS = a, OE a a ees Ge Gy ne ey = 2 ae ie eg OS ae Fa) bE SS Sea ee Osan Fig. 5 The development of the gonads in males. During the months September and October too few males were obtained to make observations possible. differences may well occur in crabs inhabiting other localities. Thus’ ‘of 25 mature males caught at Lambert’s Bay in October 1947, 4 per cent had un- developed gonads, 12 per cent developing gonads, and 84 per cent well- developed gonads. As will be seen later, the develop- mental cycle of the male fits in rather well with that of the female. THE REPRODUCTIVE STAGES OF THE FEMALE A total of 1,314 females with a carapace width exceeding 14 mm., were examined and of those which were not in berry, THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 319 687 were dissected and the state of the gonads recorded in the same three categories as was used for the males, namely: (a) ovary undeveloped or small; (bd) ovary developing; (c) ovary well developed. A total of 327 females were in berry, so that this represents a fourth phase in the reproductive cycle. In order to determine whether these berried females had resting or active gonads, 245 of them were dissected and their condition recorded under the same categories as before. The results are summarized in Tables III and V and text-figures 6 and 7. The upper part of text-figure 6 shows the percentages of berried females, while the lower part gives the percentages of dissected females not in berry with gonads in one of the three different developmental stages. Table II The activity of the reproductive organs of female crabs at Sand Vlei during the period February 1947 to February 1948. | Total number Date Females not in berry | Females in of crabs ] berry examined scree | eee) f Number of Undeveloped | Developing Well- females larger || Percentage of ovaries ovaries developed than 14.mm. | females in ovaries dissected | berry % % % | % 12°2. 747 71 28 I 93 I 136 ee 96 4 oO 112 oO 152 253 100 O O 47 ) 67 15°4 100 O ) 37 O 40 6°5 68 30 I 47 O 50 27°5 3 26 71 35 52 67 6-6 II 56 33 9 33 27 19°20.6 44 22 BB 9 64 25 18-7 17 ) 83 6 85 39 30°7 ia 33 66 12 78 54 12°8 13 25 63 8 79 39 26°8 8 25 67 12 76 51 23°9 33 33 oe 3 94 54 13°10 10 20 70 10 55 22 13°11 25 33 42 12 20 25 25711 54 31 15 52 30 7 16°12 64 29 7 41 10 49 g't. 48 93 6 I go 4 243 23°1 96 4 Oo 82 O 97 Table III and text-figure 6 show that there is a definite periodicity in the reproductive cycle of the female. The breeding-season is in the winter and the spring. During summer the ovaries are in a resting stage, but by May development starts to take place. 320 ANNALS OF THE SOUTH AFRICAN MUSEUM 80 ee : aaa (r 0g a FEMALE S IN BERRY 3 °F ia Hl 20} Te! | e . e s . e = e ° Ld s } | Le © *.*.1 DEVELOPMEN- :."+).1 TAL STAGES -"*:.1 OF OVARIES Pore a Ue epee iinnin PERCENTAGES Faults | UNDEVEL. OVARIES c===| DEVELOPING OVARIES r | (TT) WELL DEVEL. OVARIES | Fig. 6 The development of the gonads in females. The upper half gives the percentages of females in berry, while the lower half represents the percentages of dissected, not in berry, females with gonads in different developmental stages. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 321 While the investigation was carried out plankton obtained in plankton netting in the Sand Vlei Estuary was also examined for the presence of larval stages. The results are given in Table IV. Table IV The occurrence of Hymenosoma zoeae in Sand Vlei Estuary. | | Date June July | August September | October | November | December Quantity some |some-many; some many many few-some few As could be expected the presence of zoeae coincided with the months of the year that females were in berry. As the breeding-season falls mainly in the winter, it coincides with the months of heavy rain and consequently Sand Viei fills up with rain water. This eventually causes a strong current which flows into the sea so that large numbers of zoea larvae are lost to the estuary. As will be mentioned later the outflowing water will have a salinity too low for the zoeae to remain alive. Table V Analysis of samples of mature females from other localities. Total number of females Not in berry In berry Lambert’s Bay 226 = 100% yh = Bue 142)— 6307, October 1946 of these: 18% with undeveloped ovaries. 35% with developing ovaries. 48% with well-developed ovaries. Lambert’s Bay 246 = 100% TO2— OG December 1946 Ofir7- 72% with undeveloped ovaries. 54. = 22% 15% with developing ovaries. 14% with well-developed ovaries. Hermanus 160 = 100% 142 = 88% LOne—s iy, September 1947 of137: 2% with undeveloped ovaries. 37% with developing ovaries. 61% with well-developed ovaries. Berg River Mouth || 94 = 100% 36 = 38% 58 = 62% September 1949 of 24: 54% with well-developed ovaries. * An additional 74 (46%) had still traces of hatched eggs. 322 ANNALS OF THE SOUTH AFRICAN MUSEUM Although there is a possibility that the tide sometimes may carry zoeae larvae into the estuary, those swept out will outnumber those swept in. The maintenance of the Hymenosoma population in the estuary must be in great part due to the limited number of zoeae which hatch when the mouth of the estuary is blocked by a sandbank, or when the mouth is still open while very little water flows out and the tide enters at high water. The possibility of crabs invading the estuary from False Bay cannot be completely ruled out. As has been mentioned previously in addition to the crab material which was collected at Sand Vlei some was collected from other localities. Hymeno- soma was extremely common in a small blind lagoon at Lambert’s Bay, and in October and December 1946 material was collected and examined. In September 1947 material was collected from the large Hermanus lagoon and in September 1949 crabs were collected and examined from the Berg River Mouth. The results are given in Table V. (See previous page.) Many crabs were also kept in the laboratory. The salinity of the water in which these were kept varied from 2°7°/9) to 30:9°/o9. This variation is not abnormal because under natural conditions the species occurs in places where large variations in salinity do take place. The temperature varied between 13° C. and 15° C. In seventy-one cases these crabs extruded ege-batches. In Table VI these have been tabulated according to the months in which they occurred. Table VI Egg-batches extruded by 34 crabs kept in the laboratory. Month Number of | Percentage of total Month Number of | Percentage of total egg-batches | number of egg-batches egg-batches | number of egg-batches January 2 2°8 July II 15'5 February 4 5:6 August 15 21°3 March I 1-4 September 7. 9°9 April I 1°4 October 13 18°3 May I 1-4 November 8 II3 June 4 56 December 4 56 The information given in Table V indicates that the breeding-season for Hymenosoma females from Lambert’s Bay, Hermanus, and the Berg River Mouth is similar to that from the Sand Vlei Estuary crabs. Table VI shows that under the stated laboratory conditions, egg-laying could take place in any month of the year, but mainly occurred from July to December. This also agrees with what was found at the Sand Vlei Estuary (see text-figure 6). THe RELATIONSHIP BETWEEN THE GONAD ACTIVITY OF THE Two SEXES If text-figures 5 and 6 are compared, it is clear that the short resting-period of the gonads of the males (from October to December) coincides with the THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 323 last half of the breeding-period of the females. This may be expected, because copulation during the egg-carrying period must be down to a minimum. A similar state of affairs was found to exist in Cyclograpsus punctatus (Broekhuysen, 1941). NUMBER OF EGG-BATCHES EXTRUDED BY EACH FEMALE DURING A SINGLE BREEDING-SEASON In order to ascertain whether [===] UNDEVELOPED one female produces more than E=j DEVELOPING A : (00 WELL DEVELOPED one batch of eggs in a breeding- A total of 245 females in berry were dissected and the results 100 season, berried females were dissected and the state of the gonads recorded. If only one ied egg-batch was produced, the gonads of berried females could 60 : be expected to be in a resting e aes stage; if, however, more than one AG O me are tabulated in Table VII which © Pal athe peace = bi 9 9 ts is graphically expressed in text- eS POLS OI? : figure 7. Fig. 7 Developmental stages of the gonads in females Table VII in berry. State of development of ovaries of berried females. % females with % females with % females with Total number of Date undeveloped ovaries | developing ovaries | well-developed ovaries crabs May 04 6 fe) 16 June 24 77 O 13 July 39 41 20 54. August 39 42 19 62 September 53 31 16 51 October 83 17 Oo 12 November 54 43 3 37 Table VII and fig. 7 show that in May most of the berried females had undeveloped or resting gonads. During the rest of the breeding-season (June to September) the ovaries were in various stages of development while the crabs were carrying developing embryos. This is strong evidence that more 324 ANNALS OF THE SOUTH AFRICAN MUSEUM than one batch of eggs is produced per breeding-season. As shown below (Table VIII) these results were confirmed by observations on crabs kept in the laboratory. Table VIII The interval between the hatching of one batch of eggs and the laying of the next in between consecutive moultings in laboratory cultures. Interval in days \ Number of cases || Interval in days | Number of cases || Interval in days | Number of cases 9 2 17 3 36 I 10 I 18 2 39 I II 2 19 I 42 I 12 2 21 2 43 2 14 2 32 2 48 I 15 2 33 2 Average interval 23 days From Table X the average incubation time of eggs kept under the stated laboratory conditions appears to be 43 days. The average interval between batches was 23 days (see Table VIII). ‘The total time from the attachment of one egg-batch until the attachment of the next during the breeding-season was therefore about 66 days. The breeding-season during the period of the investigation lasted from May until October covering approximately 150 days. Although information obtained on crabs kept in the laboratory under controlled conditions may differ to some extent from what is the case for crabs under natural conditions, the possibility of two to three batches of eggs in one breeding- season may be provisionally accepted. THE DEVELOPMENT OF THE EGGs IN THE EGG-BATCHES In the eggs of the marine Brachyura the following ten developmental stages can be distinguished by external examination of the living eggs: Stage I. No segmentation has occurred and no external cleavage can be detected. Stage II. The first cleavages have taken place. Stage III. A considerable number of cleavages have taken place. Stage IV. The first indication of invagination has become visible. Stage V. A distinct division between a yolk-free and a yolk-containing part can clearly be seen. This stage covers the whole period between the first indication of the forming of the germ-layers and the formation of the eye-pigment. Stage VI. The first indication of the eye-pigment is externally visible. Stage VII. The first indication of the chromatophores which will form pigment bands has become visible. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 325 Stage VIII. The larvae are clearly pigmented, a fair amount of yolk remains, but is becoming reduced in quantity. Stage IX. The yolk has nearly disappeared, the egg-shells are rupturing, and the larvae are emerging. Stage X. Only dead eggs and empty egg-shells remain. Fig. 8. Hymenosoma orbiculare Desm. Left: first zoea stage; right: last zoea stage. As in Cyclograpsus punctatus (Broekhuysen, 1941) it is the first zoea stage which escapes from the female abdomen (fig. 8, left). During the period May 1947 to the first week of January 1948 a total of 323 ege-batches of berried females were examined and the stage of development of the embryos recorded. The results are given in Table IX. Individual variation in the development of one batch was slight and this was also found to be the case in other marine crabs (Broekhuysen, 1936, 1941). The duration of each stage was not the same and the laboratory experiments showed that the duration from stage I to beginning of stage [IV was approximately the same as from stage IV to beginning of VII and stage VII to IX. In text- figure g stages I to III have been lumped together and the same has been done for IV to VII and VII to IX. 326 ANNALS OF THE SOUTH AFRICAN MUSEUM Table 1X The development of attached embryos during the 1947 breeding-season expressed in percentages. Dead | Total number Date®|| Stage ds| “Al |) HT | Va OVE SEE | VTE eggs of crabs Yo |, Yo yor il feral azo: |e: |) ou ||) ven ||) ol eran a 27°5 3 97) S25 4) 19 Wag 6 fo) 6 oO oO 32 6-6 co) 1G al eie2e| arora! 4d 6 fo) 6 ) fo) 17 20°6 Oo CONmm Tf ot Qe aoe Gotha Oo Oo o | 56 9 18-7 O o | 16 B25 S20 25 fo) O 32 30°7 O Oe i A | eI sg Oy HO as fo) Oo 2 42 12°8 13 3 13 30) tO 3 ste) oO oO 30 26:8 5 5 | 21 5 24 A 5) 5, || 24 Por aes 3 38 23°9 4 Anil aLA: 4S 6 6 Oo | 49 2 (0) 2 51 13°10 17 17 25 8 8 O 8 8 oO (o) 8 12 13°11 22 DE 50. Wy 21 fo) O ) ) fo) ) 9 25°11 Oo 3 6 fo) 8 6 6 | 69 ) 3 36 16-12 O 20 13207732017 40 O Oo Oo Oo oO 5 gl Oo 0) Oo fo) fo) Odi 10.490 (0) O 10 Fig. g indicates that the majority of females produced two and perhaps three batches of eggs during the breeding-season of 1947 at Sand Vlei. This, there- fore, confirms what has been said on page 324. Fig. 9 also indicates a con- siderable individual variation. INCUBATION TIMES OF EGGS KEPT UNDER LABORATORY CONDITIONS As has been mentioned previously crabs were kept in the laboratory while the field-work was in progress. The salinity of the water in which the crabs were kept was known at the beginning and the water-level was marked on the glass. At intervals distilled water was added to compensate for evaporation. At the end of the experiment (in some cases lasting over a year) the water in some of the dishes was titrated and the salinity determined. In this way a rough estimate of the salinity conditions during the experiment was obtained. As will be shown further on, the developing eggs of Hymenosoma orbiculare are rather euryhaline and the limited salinity changes of the water in the dishes will probably not be very important. The glass dishes containing the crabs were kept in a controlled temperature room. Most of the time the temperature was 13° C. but sometimes the temperature went down to 12° and up to 15° for a short time. The development of 43 batches of eggs extruded by crabs kept under these conditions was studied. The incubation time of all 43 batches was determined with an accuracy of two days. The results are given in Table X. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 32°77 27 MAY IS NOVEMBER 6 JUNE 28 NOVEMBER r: JUNE is Ii6 DECEM 18 JULY BER 12 AUGUST : 9 JANUARY 26 AUG. a . a a ae : OCTOBER il Ww Wik DEVELOPMENT STAGES SEAT aane B fe) PERCENTAGES Fig. 9 The development of the attached embryos during the 1947 breeding-season expressed in percentages and three groups of development stages each of which takes about the same time. (The observations for 30 July are not graphed as they are practically the same as for 18 July.) 328 ANNALS OF THE SOUTH AFRICAN MUSEUM Table X Incubation time of eggs under laboratory conditions. Maximum Maximum temperature salinity range Incubation time in days Average range (Op ss 12—13 1529 49 49 12—I4 14——29 43, 44, 47 45 15—29 42, 43, 44, 44, 44, 44, 44, 45, 45, 47, 48, 48 45 16—29 44 44 I2—I5 14—29 40 40 15—29 43, 45 44 13—14 14——29 43 43 I15—29 39, 49, 41, 41, 42, 42, 42, 44, 44 42 16—29 38, 39 39 13—I5 I4——I9 41, 42 42 15 41 41 I5—29 37, 38, 40, 41, 42, 43, 45 41 14 15—29 38 38 Average of all observations 43 The figures in the above table show that 43 days was the average incubation time of the eggs of H. orbiculare at a temperature between 12° C. and 15° C. The figures also suggest that the more constant the temperature, the shorter the incubation time. A greater amount of variation of the temperature seems to increase the incubation time, as was also found to be the case for Cyclograpsus punctatus (Broekhuysen, 1941, p. 344). In July 1947 a female with a carapace width of 18-7 mm., kept in a glass dish without any sand or shell fragments, extruded eggs which got attached to the hairs of the pleopods in the normal way. The presence of a soft sub- stratum, therefore, seems not to be essential for a normal attachment of the eggs in this species. In this respect H. orbiculare seems to be different from C. punctatus and C. maenas (Broekhuysen, 1936, 1941) in both of which the extruded eggs did not get attached if sand and shell fragments were not supplied. In laboratory cultures there were twelve cases of two batches of eggs being produced which developed normally although copulation had taken place only once, i.e. before the first batch of eggs. In two cases, three batches were produced after one copulation; the eggs of the first two batches developing normally while those of the third batch developed abnormally and never produced zoeae. There were four cases where three batches were extruded after one initial copulation and the eggs of all three batches developed normally into zoeae. This shows that a female H. orbiculare can produce as many as three normal batches of eggs in succession although fertilized only by one initial copulation. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 329 In some cases, however, the eggs of the third batch were not properly fertilized. The fact that one copulation suffices for the fertilization of more than one batch of eggs has also been found to hold good in other Brachyura (Gosse, 1852; Williamson, 1900; Churchill, 1917-18; Duncker, 1934; Broekhuysen, 1936, 1941). In all the cases where small immature female crabs were put into the glass dishes and reared apart from males, the eggs developed only into an irregular cleavage stage and then died. In one instance four egg-batches were produced by such a female during a period of thirteen months. In this case none of the eges developed further than into an irregular cleavage stage. In some cases females which had just moulted when caught were put into a glass dish. After some weeks these females moulted again, but as no male was in the dish, refertilization could not take place. Nevertheless the egg-batches produced later contained fertilized eggs, showing that the moulting process had not affected the spermatozoa stored in the female. These cases also show that copulation is not essential for extrusion of eggs. The time between copulation and egg-extrusion varies considerably as can be seen from Table XI. ; Table XI The interval between moulting and the next egg-extrusion in the crabs kept in the laboratory. Copulation took place after the moult No copulation took place Date of Date of Date of Interval in Date of Date of moulting copulation egg-laying days moult ing egg-laying Interval 14°3 | 17°3 315 77 25°3 3°6 59 30°4. 30°4. about 54 24:6 1°4 6:8 126 22°5, 22°5 57 43 6-4 1°7 85 255 ois 78 73 13°4 6-7 83 19°6 21-6 29°7 39 27°4. 19°6 52 21°6 21°6 27°83 66 27°4. 10°8 104 3°38 a 7°10 64 28-4. 25°7 87 3°38 ? 22°10 79 6:5 about 57 2:7 11°8 11°8 8-10 57 12°5 10°8 89 11°8 11°8 26-11 106 155 4°7 49 22°8 22°8 about 84 155 18-7 63 1511 Q°Il II‘ll 28°1 78 19°5 18°8 90 Q'Il II‘1l 12°2 94. 26°5 BNioy| 65 I5‘11 nigh 5:2 81 26°5 23°9 118 I-12 OG) Q°2 69 46 48 60 | 10°6 10°8 60 Average 71 13°6 20°8 67 17°6 7c 142 4°7 11°9 68 Average 80 330 ANNALS OF THE SOUTH AFRICAN MUSEUM As shown later, copulation coincides with moulting of the female. If the two parts of ‘Table XI are compared it will be seen that there is a difference of nine days between the average intervals, and copulation therefore seems to have a slight stimulating effect as regards egg-extrusion. During the laboratory experiments several observations of the method whereby females with hatching eggs aided the zoea’ larvae to escape from underneath the abdomen were made. On such occasions females stood on their pereiopods and bent their abdomens backwards. The zoeae then poured from beneath the abdomen into the surrounding water. COPULATION In Hymenosoma orbiculare copulation takes place between a hard male and a soft, newly moulted female. In this respect this species behaves in the same way as C. maenas and many other crabs. Copulation is preceded by an embrace of the female by the male, some time before she actually moults. The male holds the female underneath him with his pereiopods, but the female is in the normal position (i.e. dorsal side up). Actual copulation was observed several times in the laboratory. The following notes were taken on one of the occasions: I1°11. °47 Female recently moulted, leathery. Add male from dish 12. After a few seconds the male mounts the female. ‘This male removed and substituted by male from dish 15. The latter, after a few seconds, approaches the female which is busy digging into the sand. Male is first on top of female, but then gets underneath the female. He then turns over on to his legs and in doing so keeps female underneath him and turns her over on to her back and copulation commences. In the limited number of copulations observed, the process lasted more than half an hour and probably longer. After copulation had occurred, the female was kept embraced by the male for a considerable time, sometimes more than a day. No actual experiments were carried out with a view to establishing the presence or absence of sex recognition. The general impression, however, was that sex recognition was poorly developed if present at all. The procedure in H. orbiculare seemed very similar to that suggested by Broekhuysen (1937) for C. maenas, i.e. the seasonal periodicity in copulation is only caused by the seasonal moulting act in sexually mature females, when the female due to its soft condition cannot evade or resist a male trying to copulate. These remarks, however, are tentative and more experimental work is necessary. GROWTH In order to obtain information on the rate of growth, the number of moults, the time required for hardening after moulting, the existence of sexual dimor- phism and the average maximum age, all the crabs collected in Sand Vlei estuary were sexed and measured. The index of size used was the width of the carapace in millimetres. Crabs which were just about to moult, or had just moulted were noted. As has been mentioned a fairly large number of crabs THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 331 were kept in captivity in some cases for periods of over a year. The growth of those which were caught while still small was recorded in order to supplement the data obtained in the field. MOuLTING As shown earlier, moulting not only marks a stage of growth, but in the present species it also controls the possibility of successful copulation. Moulting is thus essential to the animal but is also a dangerous period, for the crab is practically defenceless and open to attack by predators, including its own species. It has been noted that males copulate with soft, just moulted females and this must decrease the danger of soft females being attacked by males, considerably to the advantage of the females. In some cases females which moulted in the laboratory were killed by males in the same dish. If this was not an abnormal occurrence due to captivity, it means that the moulting female is not altogether immune to attacks by males. Hymenosoma orbiculare and also the other species of Brachyura where males copulate only with soft females seem to have some advantage over species in which copulation occurs between two hard crabs. The process of moulting in H. orbiculare is identical to what takes place in other Brachyura and has been described for Carcinus maenas and Cyclograpsus punctatus by Broekhuysen (1936 and 1941). One interesting difference between the hardening of the new shell of C. punctatus and that of H. orbiculare was found. While H. orbiculare always remained submerged during the process, it was found that in C. punctatus it is essential that the crab should only be submerged part of the time for the hardening process to proceed normally. The duration of the hardening process was observed in the laboratory. For 52 crabs between 8 and 22 mm. carapace-width at a temperature between 12 and 15° C., the average time was four days and the extremes two and nine days. Unfortunately the records do not permit one to determine the effect of temperature or the size of the crab on the duration of the hardening process. The increase in size after moulting was recorded for 99 crabs kept in captivity under fairly constant temperature. These records include a certain number of observations on crabs kept in very low salinities and some which were kept in water with a salinity of over 3 5/00: As the increase in size of crabs kept in abnormal salinities was roughly the Same as the increase for crabs in normal salinities, the records were combined. Also, no significant difference was found between males and females. A summary of all the records gave the following results: Size range Number of crabs | Percentage increase in carapace width 5—8 mm. 15 22% 9—12 mm. 18 24.% 13—16 mm. 48 15% 17—22 mm. 18 11% 332 ANNALS OF THE SOUTH AFRICAN MUSEUM Thus also in H. orbiculare the rate of increase decreases as the crab gets bigger and older. PERIODICITY IN MOULTING As females only copulate after moulting, it was thought possible that mature females would show a distinct periodicity in moulting, correlated with the breeding-season. In Table XII percentages of moulting females and males, and females in berry have been combined. Only crabs larger than 14 mm. have been considered. ‘The information is expressed graphically in fig. to. Table XII % of moulied | % of females | Total number % of moulted Total number Date Females in berry of females males of males February 1947 3 ; I 136 5 108 March 37 oO 119 14 184 April 35 a) 40 5 78 May 21 27 117 5 ZoMe June 19 52 52 13 31 July 8 81 93 13 30 August 4. 78 go 2 43 September 13 04. 54 (0) I October 5 55 22 oO 3 November 13 27 99 6 53 December 4 10 49 5 39 January 1948 4 2 338 4 159 From Table XII and fig. 10 it can be seen that mature females can and do moult during the whole of the year, but that there is a definite maximum in moulting activities during the months March, April and May. Fig. 10 also shows that this increase in moulting precedes the breeding-season. The observations available for the males seem to indicate that there is no clear maximum for the moulting in this sex. In fig. 11 the comparable figures for crabs kept in the laboratory have been graphed. If the two figs. 10 and 11 are compared it will be noticed that they differ little for the females but quite distinctly for the males. The females kept in captivity also show a sudden increase in the number of moulting crabs in March, April and May, the same as was found in the crabs at Sand Vlei. For the males the number of observations of captive crabs is limited but they do show a clear maximum of moulting in March. This was not found in the crabs at Sand Vlei. It should be stressed that fig. 11 is based on crabs kept under rather uniform conditions and may therefore be more accurate in a comparison of the behaviour of females to males. In any case it is interesting to note that in these THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 333 laboratory crabs, the moulting activity of the males precedes. the maximum moulting activity of the females, in other words the males have already moulted and become hard when the females moult and are ready for copula- tion. In the field where conditions are not uniform this adaptation apparently becomes less striking. SEXUAL DIMORPHISM In Hymenosoma orbi- culare there is relatively little external difference between the two sexes except for the shape of the abdomen, and the 2ABS NUMBER OF MOULTING CRA MONTHS Big Tt The number of moultings taking place in the 100 --~MOULTED FEMALES 904 ——MOULTED MALES 74 ! *+t++ FEMALES IN vA 804 BERRY jr, (e) PERCENTAGES ev bh o Oo Fig. 10 The percentages of (a) moulted or moulting females, (6) moulted or moulting males and (c) females in berry, plotted against the months of the year. The crabs were collected at random at the mouth of Sand Vlei. development of the pleopods, which are adapted to carry- ing eggs in the female. The chelae, however, are some- what bigger and broader in the case of the male (fig. 1). In order to determine whether there was a _ con- sistent difference in size be- tween males and females of the Sand Vlei population, a total of 1,417 females, of which 316 were females in berry, and 856 males were measured. The results are given in Table XIII and different months in crabs kept in captivity. graphed in fig. 12. 334 ANNALS OF THE SOUTH AFRICAN MUSEUM —---— ALL FEMALES MALES pH FEMALES WITH OVA NUMBER OF CRABS 6 ree RRM Saraue Rll Sm ~ (5,5.6.7.8. 10.11.12.13.14,15,6.17.18.19.20.22. 24. 26. CARAPACE -WIDTH IN MILLIMETRES Fig. 12 The distribution of all the crabs measured over classes differing one millimetre. The size is expressed in carapace-width in mm. (females with ova means females in berry). THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 335 Table XIII | The carapace width in millimetres of crabs from Sand Vlei Carapace Females (those in | Females in berry | Males| Mean carapace Mean carapace width berry included ) only width for females width for males 5 mm. ~ - I 5 I - 2 6 - - I 7 I - I 8 I - I 9 ~ ~ I 10 I ~ I II I - 2 12 19 a 5 13 66 12 28 14 254 12 44, 16 18 15 381 21 105 16 241 24 132 17 126 35 118 18 108 ais 94. 19 IOI 64 84 20 103 52 86 21 43 22 60 22 20 12 39 23 2 I 36 24. 2 I 10 25 = < 3 26 - = 2 Total 1,471 316 856 The mean for females is 16 mm. and for males 18 mm. The difference is very small and was tested statistically. The t-test of significance showed that at the 1 per cent level of significance the estimated ¢ was 2-819 and the 5 per cent level 2-074. The calculated value for t was 0-136. There was therefore no significant difference in size between the two sexes in the population of H. orbiculare at Sand Vlei at the time of the investigation. The shape of the curves in fig. 12 may suggest the presence of two year-classes in both females and males. THe RATE oF GROWTH Since crabs can only increase in size when they moult the rate of growth is dependent on the increase at a moult and the frequency of the moults. As has been mentioned before the experiments in the laboratory did not reveal any significant difference between males and females in the increase in size after a moult. The increase after moulting is mentioned on page 331. As regards the number of moults, or rather the duration of the interval between two successive moults, it is difficult, if not impossible, to obtain information on crabs living under natural conditions. The only information available, there- 336 ANNALS OF THE SOUTH AFRICAN MUSEUM fore, comes from crabs reared in the laboratory where conditions were not quite natural. This may or may not have affected the duration of the intervals between successive moults. In Table XIV the available information has been tabulated. The females not in berry and those in berry as well as the sexes have been kept separate. The crabs have been divided into five size-classes. Table XIV Interval between successive moults in days, in crabs reared in the laboratory. (Extremes are given in brackets.) FEMALES MALES Not in berry In berry SLC GR ee mm. Average Number Average Number Average Number duration of cases duration of cases duration of cases 5 40(18—58) 5 30(18-40) 5 5-8 33(25-45) 11 35(22-57) 4 Q-12 52(30-go) 12 37(38-96) By. 13-16 89(59-139) II 219(216—221) 2 32(81-214) 4 17-22 128(42-183) 5 174(109-221) II 66(51-79) 4 This table reveals some interesting points: (a) From the table it appears that the interval between successive moults, even in crabs of less than 5 mm., was considerable. For the smallest sizes this may be partly due to the young crabs taking some time to settle down to laboratory conditions after their capture. (b) The interval between two successive moults even for crabs of the same size- class varied a lot. (c) The table also shows that the intervals in the case of females in berry were considerably longer than those for females not in berry and males. The number of observations on males, however, may be too small to enable any positive conclusions, but they suggest that where the female growth can be affected adversely by breeding, this may not be the case for the males. This point will be mentioned again later. . (d) The interval gets longer as the crab gets older and therefore bigger and in this respect H. orbiculare behaves in the same way as other Brachyura. In order to determine the development of the crabs at Sand Vlei, size/ frequency graphs were plotted for each month of the year and for both sexes. These are given in fig. 13. While every attempt was made to obtain random samples, there is no doubt that a high proportion of the smallest size groups, particularly those under 7 mm., evaded capture. To this extent the samples and the graphs constructed from them are biased. 9 THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. Ly TOTAL aA NUMBER OF CRABS és 4 99 138 FEB: 1947 OF CRABS NUMBER a) (Mil) MALES | SEX? E41) FEMALES i 86 178 99 392 JAN. 1948 (5 6 8 10 12 14 16 I& 20 22 24 26 28 CARAPACE WIDTH IN MILLIMETRES Fig. 13 Size/frequency of males and females for the different months. The crabs were all collected at the Sand Vlei mouth. 337 338 ANNALS OF THE SOUTH AFRICAN MUSEUM Even so the following important points are shown: (a) In 1947 a considerable number of very small H. orbiculare made their appearance in the catches in September. These were the new generation of the year. In addition there were a considerable number of female crabs of 16 to 22 mm. carapace-width which were probably over 1 year old. Hardly any males were caught. In October the position had changed very little. In November a distinct group of female and male crabs of 12 to 18 mm. had appeared. It is unlikely that these consisted of crabs of the same season as they appeared too big for that. It is impossible to decide where these crabs came from, although the possibility of an invasion into the estuary from the shallow sea along the shore of False Bay cannot be ruled out. A certain number of crabs hatched in 1946 were still present, but by December most of these seemed to have vanished. (b) If we examine fig. 13 we shall see that in November and December the curves for the males and the females more or less coincided. In January, February and March the males gained more and more so that by April there was a considerable difference in size between the males and the females, the former being the largest. In May, however, this difference suddenly disappeared due to a sudden increase in the size of the females. It should be remembered that March, April and May are months in which the females showed an increase in their moulting activity (see figs. 10 and 11), prior to the beginning of the breeding-season. During June, July and August the males again gradually gained on the females. This was the time when most females had egg-batches and therefore were not in a condition to moult. This did not affect the males (see Table XIV) which apparently continued to increase in size. (c) In all months except April, there appeared to be more females than males and in September nearly all the males from the older generation had disappeared. During the course of the investigation a number of very small Hymenosoma crabs were collected and reared in captivity. Some of the results of this part of the investigation have been combined in Table XV. Table XV Some small Hymenosoma crabs reared in captivity at 13° to 15° C. Increase in size in Number of Salinity in parts Sex millimetres moults Time in days* per thousand 3 Rae 2 55 35 3 A BONG eas 3:5 5 230 35 — 36 3 4 to 5 — 16 5 329 36 — 36 y 4 to5— 13 4 169 31 — 35 Q 6-8 — 14:6 4 230 | 5— 14 2 4 to5 — 13 4 252 Ops * The number of days covers the period from the date of moult at which the smallest size mentioned was reached until the moult at which largest size mentioned was reached. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 339 The above table indicates that crabs of about 14.mm. are more than 8 months old. It should be remembered, however, that these crabs were kept under unnatural conditions and fig. 13 seems to indicate that the crabs grow faster under natural conditions. The size at which the crabs of the Sand Vlet population became adult In males it was not easy to decide macroscopically when the crab had become mature. In females, however, the change in shape of the abdomen is an obvious indication of maturity. Moreover the development of the ovaries can usually be seen without the aid of a microscope. Extrusion of eggs is the surest sign of maturity having been reached. During the investigation the smallest females with ‘adult’-shaped abdomen and the largest females with ‘juvenile’-shaped abdomen were recorded and it soon became evident that a lot of variation eccurred. Some females with a carapace-width as small as 12:2 mm. had ‘adult’ abdomens, while others as large as 17 mm. still had somewhat ‘juvenile’ abdomens. The smallest female carrying a batch of eggs was 12:6 mm. in the Sand Vlei population. Dissections showed that in one instance a female as small as 12-2 mm. had well-developed ovaries, indicating that she was mature. Dissections also revealed that some males of 12:3 mm. and 12:5 mm. had well- developed genital tubes, indicating that they were mature. In one instance a female of 15-5 mm. was collected that had a rather immature abdomen but dissection revealed well-developed ovaries. The shape of the abdomen, therefore, is not always decisive. As will be seen later different environments may affect the minimum size at which maturity is reached. Although the above data shows that H. orbiculare females from Sand Vlei could become mature at 12-13 mm., they were usually mature at 13-14 mm. carapace-width. This is confirmed in fig. 12 in which among others the size/ frequency of berried females has been plotted. SEx-RATIO When all the sexable crab samples from Sand Vlei were added together it was found that 1,471 or 63-2 per cent were females and 856 or 36:8 per cent were males. It is interesting to note from fig. 13 that this predominance of females over males was specially marked in the January population. From then on the difference between males and females decreased gradually until in April the situation was such that there were more males than females. From April onwards the females again became predominant. No satisfactory explanation can be given for the large predominance of females over males. It is very unlikely that the males were overlooked during collecting. The increased predominance of females after May may be partly due to the fact, noticed during collecting of samples, that the females usually do not expose themselves so readily as the males and that, therefore, the chances of being swept away to the sea by the current are greater for the males, 340 ANNALS OF THE SOUTH AFRICAN MUSEUM DiIscussION Although the fact that H. orbiculare is often found in estuaries indicates that the species is euryhaline, there seems to be a certain minimum salinity tolerance, as these crabs do not occur in waters with a constant very low salinity. In order to estimate this minimum salinity limit, crabs were kept in water of a low salinity and under controlled temperature conditions. These experiments showed that within a temperature range of 12°-15° C., a salinity of 5 per thousand interferes with the normal development of the eggs. Eggs extruded under these conditions did not develop beyond the first cleavage stages. However, when females carrying embryos in an advanced stage of development were kept under the same conditions, the embryos did hatch sometimes but the resulting zoea larvae died immediately. A salinity of 0-2-1-4 per thousand and a temperature of 12°—14° C. had the same effect on development. Although salinities of 1-5 per thousand at a temperature of 12°-15° C. seriously affect the early development of this species if the crabs are exposed to these conditions long enough, it apparently has little or no effect if the crabs are only exposed to it for a short time. This is shown by the figures in Table I. From this table it can be seen that during part of June, July and August 1947 the salinity at the mouth of Sand Vlei fell well below 4 per thousand but this apparently did not affect the development of the larval stages in the egg, although it may have affected hatching zoeae and eggs freshly extruded. As the salinity in estuarine sands changes slowly, even when the water above is almost fresh, the habit of the females to dig into the surface layer of the sandy substratum may have saved the eggs from any adverse influence due to too low a salinity. During the low-salinity experiments in the laboratory, several females extruded eggs. Low salinities, therefore, apparently do not prevent females from extruding eggs. Nothing definite can be said regarding the effect of prolonged exposure to high salinities, but there are strong indications that although the species may survive and even develop under these conditions, it has some effect. Barnard (1950) in his Monograph on the South African Decapoda mentions a variety of H. orbiculare from deeper water in False Bay, which at first sight seems to be specifically distinct from the normal orbiculare. Although Barnard stresses the difference he does not consider it necessarily another species. During the course of this investigation the present author had an opportunity of examining fairly large numbers of these deep-water Hymenosoma from False Bay. In addition to the extensive granulation, mentioned by Barnard, there was a striking difference in the size when compared with H. orbiculare from the mouth of Sand Vlei. In Table XVI the sizes of these deep-water crabs are compared with those of the Sand Vlei population. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 341 Table XVI Comparison of crabs dredged below 12 fathoms in False Bay with those from Sand Vlei estuary: False Bay (dredged) Sand Vler Females Maximum carapace-width 10 mm. 24mm. Average carapace-width 7-9 mm. 18 mm. Smallest 9 with mature abdomen 6—7 mm. 13-14 mm. Smallest berried female 6 mm. 12°6 mm. Males Maximum carapace-width 12 mm. 26 mm. Average carapace-width 7-6 mm. 18 mm. The total number of deep-water crabs available was 96 99 and 85 gg. The differences in size between the two populations are very striking. If the deep- water crab is the same species as H. orbiculare which populates the mouth of Sand Vlei, the difference must have been caused by external conditions. The factors which may be concerned are: (a) constant high salinity in the case of the False Bay crabs, and a lower average salinity, but varying tremendously, in the case of the Sand Vlei population ; (b) considerable depth in the case of the False Bay crabs, and relatively shallow water in the case of the Sand Vlei crabs; (c) a less variable temperature in the case of the False Bay crabs. Without more detailed field-work and experiments, it is not possible to be certain which of these factors is or are the more important. There is some evidence, however, that salinity may be important. In 1948 during an investi- gation of the ecology of St. Lucia Estuary on the Zululand coast (Day, Millard and Broekhuysen, 1954) it was found that salinities in this estuary were very high (34-53 per thousand). Specimens of H. orbiculare were collected and measured, and it was found that females were becoming mature at 5 and 6 mm. carapace-width and one female of 4 mm. had a ‘mature’ abdomen. Several berried females were only 5 and 6 mm. These crabs were therefore comparable, at least as far as size, with those from False Bay. Although few measurements are available the Hymenosoma crabs inhabiting Langebaan Lagoon also seem to be of small size. The smallest mature female measured was 10 mm., while a female of 7 mm. carrying an egg-batch was collected in Saldanha Bay into which Langebaan Lagoon opens. The salinity in this lagoon is near the salinity of normal sea-water and therefore fairly high. It seems, therefore, that high salinity is at least one factor which decreases the size of the mature crabs. It also increases the size of the chelae in the males and causes excessive granulation. There are also indications that there 342 ANNALS OF THE SOUTH AFRICAN MUSEUM is a difference in breeding habits between the deep-water Hymenosoma from False Bay and those from Sand Vlei. When the Hymenosoma results are compared with those for Cyclograpsus punctatus (Broekhuysen, 1941) it is evident that there is a great deal of similarity between the two species. Both are winter-breeders. In C. punctatus breeding covers the period May to November and in H. orbiculare the period is from June to November. In both cases the females produce several egg-batches in the one breeding-season. Whereas C. punctatus has a second, minor breeding-season in the summer this is not evident in H. orbiculare. In both species the gonads of the male show a periodicity in their activity which is adapted to the rhythm in the females. There is also a great similarity as regards the incubation period of the eggs of the two species. The incubation time of the eggs of C. punctatus, at a constant temperature of 16°5° C., is little over a month, while at a tem- perature of 12°-15° C. it took the eggs of H. orbiculare 38 to 48 days to develop and hatch. In Cyclograpsus copulation takes place between two hard crabs while in Hymenosoma it only takes place between a hard male and a freshly moulted female. In C’. punctatus no seasonal difference between the moulting periods of the two sexes was observed, while in Hymenosoma orbiculare such a difference did seem to occur. This is probably related to the fact that in the latter copulation only takes place when the female has recently moulted. In neither species is there a significant difference in size between the two sexes. In C. punctatus from the shore of False Bay the majority attained an age of two to three years. The present investigation indicates that the majority of H. orbiculare die in their second year. In both species there were more females than males, although this appeared to be much more pronounced in the case of Hymenosoma. SUMMARY (1) Hymenosoma orbiculare is a crab which occurs in the shallow waters of estuaries as well as in deeper water along the shore of South Africa and Portu- guese East Africa. Most of the material on which this paper is based was collected in the False Bay area between February 1947 and February 1948. (2) The population, showed a definite periodicity in the activity of the gonads and the breeding-season appeared to be the winter months. (3) Hymenosoma females extruded two or three egg-batches during the breeding-season.. (4) The development of the eggs was followed in the field and in the laboratory. (5) Females will only allow males to copulate after the female has moulted and is still soft. (6) Although moulting may occur at any time of the year, the females showed a sudden increase in moulting a few months before egg-extrusion. THE BREEDING AND GROWTH OF HYMENOSOMA ORBICULARE DESM. 343, Under laboratory conditions males also showed a periodicity, which seemed to be adjusted to that of the females. Under natural conditions this was not so evident. (7) Evidence is given that during the breeding-season males increase more rapidly in size than the females. This may be due to the fact that the majority of females are carrying eggs and therefore do not moult during that period. (8) The percentage increase in size of the crab after each moult shows a eradual decrease as the crab grows larger. (9) The size at which the female crabs became mature was established for the Sand Vlei population for the period of the investigation. In crabs from other localities it was found that the size varied according to environment. (10) There was evidence that the majority of crabs, in the population studied, reached an age of somewhat over one year, after which a heavy mortality occurred. (11) There are the usual Brachyuran sexual differences between the male and the female Hymenosoma orbiculare; but there is no significant difference between the sizes of the two sexes. (12) Of all the crabs collected and sexed 63-2 per cent were females and 36:8 per cent males. (13) The influences of low and high salinities are discussed and the life history of H. orbiculare is compared with that of C. punctatus. ACKNOWLEDGEMENT I have pleasure in thanking Professor J. H. Day for his valuable criticism. REFERENCES Barnard, K. H., “Descriptive Catalogue of South African Decapoda Crustacea (Crabs and Shrimps).’ Ann. S. Afr. Mus., XX XVIII, 1-864, 1950. Broekhuysen, G. J.,“On Development, Growth and Distribution of Carcinides maenas (L.).’ Arch. Neerl. Zool., 11, 257-399, 1936. ——, ‘Some notes on Sex Recognition in Carcinides maenas (L.).’ Arch. Neerl. Zool., 111, 156-64, 1937- ——., ‘The Life-History of Cyclograpsus punctatus, M. Edwards: Breeding and Growth.’ Trans. Roy. Soc. S. Af., XXVIII, 4, 331-6, 1941. Churchill, E. P., ‘Life History of the Blue Crab’. Bull. U.S. Bur. Fisher., 36, 1917-18. Day, J. H., Millard, N. A., Broekhuysen, G. J. B., “The Ecology of South African Estuaries, Part IV, The St. Lucia System.’ Trans. Roy. Soc. S. Af, XXXIV, 1, 129-56, 1954. Duncker, H., ‘Gefangenschaftsbeobachtungen an Sesarma cinerea Milne Edwards.’ Zool. Jahrb. Syst., 66, 285-90, 1934. Gosse, P. H., ‘On the Sloughing of the Spider-Crab (Maia squinado).’ Ann. Nat. Hist., X, 210-12, 1852. Williamson, H. C., ‘II. Contribution to the Life-History of the Edible Crab (Cancer pagurus Linn.), 18th.’ Rep. Fisher. Board Scotl., 3, Sci. Invest., 77-143, 1900. The ANNALS OF THE SOUTH AFRICAN MUSEUM are issued in parts at irregular a intervals as material becomes available. As far as possible each volume is devoted exclusively to a particular subject (Zoology, Botany, etc.). Two or more volumes may be in course of ‘publication concurrently. Most.of the Geological and Palaeontological papers are issued in conjunction with the Geological Survey of the Union of South Africa. Some volumes and parts are out of print, and others are only sold as parts on a set, or volume, respectively. The prices of parts published prior to 1940 have been increased. Out of print: Vols. I, II, V (Parts 1, 2, 9), VII, Vill, IX (Part 1), XII (Part ?s XXII, ‘g XXIV (Part 2), XXXI (Parts 1, 2, 3). Vol. 5 ree Hf ea III. 1903-1905 Zoology .. en Peas He api IV. 1903-1908 Palaeontology. .. 157-114, 3 Length of deutomerite .. 418 is 443 54.2-300 - Width of deutomerite .. 274 # 257 i 357-214. ed Ratios for species determination based on average figures: Length of Protomerite Total Length Width of Protomerite Width of Deutomerite i 8 (Oye lf @ 2 The greatest variation is in the length of the deutomerite (242 microns) and the most constant feature is the width of the protomerite (43 microns variation). General Shape. The deutomerite is more or less oval in shape but it is slightly pointed at the posterior end. The greatest width is usually towards the posterior end in adults but in younger individuals the anterior is usually the widest part. The protomerite is usually two-thirds as long as it is wide. The Epicyte. ‘This is very thin and mature sporonts are very easily ruptured even when handled very gently. This feature is not so noticeable in immature specimens. The usual longitudinal ridges and circular myonemes are present. The nucleus occurs anywhere in the deutomerite. It is fairly large and averages about 50 microns in diameter. It contains 4 or 5 nucleoli which are rather characteristic as they stain only slightly darker than the rest of the nucleus in 400 ANNALS OF THE SOUTH AFRICAN MUSEUM Heidenhain’s iron haematoxylin and are often difficult to see. Nevertheless, when viewed in section under an oil immersion lens, they appear as very definite structures with definite outlines. They are often clumped together at one side of the nucleus. (Fig. 11.) Apart from the nucleus there is, in the deutomerite, a most characteristic extra-nuclear body. ‘This is usually about a quarter the size of the nucleus and is spherical with a very smooth surface. Occasionally it is broken into two, three Fic. 7. Gregarina sandoni n. sp., outlines from life. a-c. from normal host. e. from starved host. d. the only individual in a well-fed host. x 76. or more smaller bodies and may be elongated instead of spherical. The body stains fairly darkly in haematoxylin but not at all by the Feulgen method. No similar structure was found in the other species nor was any reference found to it, or any similar structure, in the literature. (Typical examples have been inserted in the outline drawings, fig. 7a—-e.) The interlocking device is similar to that described for G. fastidiosa. In this species the syzygy is fairly firm and the primite and satellite are not easily separated with normal handling. The epimerite is of the normal type found in this genus. The whole structure is inside a cell of the gut-wall against which the protomerite is applied. The con- FOUR NEW SPECIES OF GREGARINES FROM CAPE PENINSULA 401 tents of the cell, including the nucleus, break down and the effect on the surrounding cells is the same as that described for G. fastediosa. Few specimens were obtained in whole mounts with the epimerite still intact; in fact, this structure seems to be exceptionally easily shed and this may explain why so many immature sporonts are found free in the gut with the early formation of syzygy. Cysts are oval and are white or light buff in colour. They are found in the faeces and hind-gut. The average dimensions for the shortest and longest diameters are 330 and 500 microns respectively. The cysts range from 282 x 420 to 420 X 670 microns. The largest cyst recorded here was very much bigger than the average size but most of the cysts did not vary very much from this average. Cysts dehisced in three days after they were removed from the gut. They did not dehisce when exposed to the atmosphere but only in the damp chamber. Dehiscence was by 7 or 8 spore ducts most of which seemed to be operational. However, when a cyst was artificially ruptured just prior to dehiscence it was found to contain 20 ducts. It would seem that all ducts are not extruded simultaneously and, after the first 7 or 8 ducts are extruded the pressure inside is too low to force the others out. Spores are dolioform and are extruded in long chains. They are very regular in size and are 7-1 microns long and 3°55 microns wide (measured under an oil immersion lens), exactly twice as long as broad. No success was obtained in attempts to make them exsporulate by putting them in fluid from the mesenteron and hepatic caeca. An interesting example of the dependence of the gregarines upon the nutrition of their host was discovered when a specimen of M. capensis was starved for a month and then opened. The mature sporonts were well below average in size and much thinner, especially the deutomerites. Their average dimensions were: Total length ve se .. 464 microns Length of protomerite eA: Ss Width of protomerite .. aa TOA an Length of deutomerite .. a 400 z Width of deutomerite .. 55 ES 5 One of these starved individuals is illustrated in figure 7e. The endo- plasm of these starved gregarines was less dense than normal. Cysts of normal size and shape were discovered but their contents appeared to be coagulated into dense lumps leaving the rest of the cyst transparent. These cysts did not dehisce and no spores were formed in them. Figure 7d was the only parasite in an apparently well-fed host and provides an interesting contrast to 7e. 402 ANNALS OF THE SOUTH AFRICAN MUSEUM Gregarina impetuosa n. sp. Figs. 8-10, 11d. All specimens of this species were found in the anterior mesenteron of the host. It is a small gregarine with a well-marked septum, at the position of which there is a slight constriction of the epicyte. b d The mature sporonts are _ bi- associative and the cephalonts bear a knob-shaped epimerite on a short stalk. Most characteristic of this species is its activity; the sporonts glide forwards across a slide covered in gut debris at about 500 microns per minute, pushing the débris aside as they move. This activity made measuring and drawing difficult but it was possible on rare occasions when the gregarines were slowed down by some exceptionally dense obstruction. Mature Sporonts (fig. 8). These lie free in the mesenteron and are usually associated in pairs. They are unaffected by the 0°75 per cent saline and remain alive in Fic. 8. it for long periods. Gregarina impetuosa n. sp., outlines from life, The following dimensions were nuclei inserted from sections. x 100. taken from the primite only: Average Medium _ Range Total length ai .. 2QI microns 300 microns 357-228 microns Length of protomerite .. 57 ss 57 53 72-A2 x Width of protomerite .. 66 ie 71 ys 86-43 re, Length of deutomerite .. 234 a 243 fs 286-185 es Width of deutomerite .. 126 Hs 114, Bi 171-71 KS Ratios based on averages: Length of Protomerite Total Length Width of Protomerite Width of Deutomerite Eo? 1.9 The feature that shows the least variation is the length of the protomerite (28:7 microns) while the greatest variation is in the length of the deutomerite (101 microns). FOUR NEW SPECIES OF GREGARINES FROM CAPE PENINSULA 403 Fic. 9. Gregarina impetuosa n. sp., showing early development of karyosomes and effect on parasitized cells. This shows the youngest parasite discovered. d. deutomerite. e. epimerite. k. karyosomes or nucleoli. n. nucleus. p. protomerite. x 800. Fic. 10. Gregarina impetuosa n. sp., a later stage. Lettering as in fig.9. X 570. 4.04. ANNALS OF THE SOUTH AFRICAN MUSEUM General Shape. ‘The deutomerite is approximately oval in shape and the greatest width is usually towards the posterior end, though not always. The deutomerite has a slight waist in the middle. The protomerite is only slightly wider than it is long. There is usually a papilla on the anterior end where the epimerite was attached. The epicyte is fairly thick and the gregarines are seldom, if ever, ruptured with normal handling. The usual longitudinal ridges and circular myonemes are present. The nucleus (fig. 11d) occurs in any position in the deutomerite and is, on the average, 34 microns in diameter. In the mature sporont there are from 6 to 8 nucleoli which stain darkly with haematoxylin though a few cases were Fic. 11. Nuclei stained in Heidenhain’s iron haematoxylin showing ‘karyosomes’ or nucleoli. a. G. fastidiosa. b. G. gibbsi. c. G.sandoni. d. G. impetuosa. Sa FOUR NEW SPECIES OF GREGARINES FROM CAPE PENINSULA 405 observed where some of the nucleoli in a single nucleus stained a much lighter colour than the others. In the youngest cephalont (fig. 9) there was only one nucleolus; this specimen was the smallest cephalont encountered in any of these four species. The nucleolus must divide up at a very early stage as in a somewhat larger but still very young cephalont it had already divided up into five (fig. 10). The epimerite was only seen in sections and all cephalonts examined were very young. It was knob-shaped and borne on a very short stalk and the whole structure was inside a cell of the gut-wall. The interlocking device is similar to that described for Gregarina fastidiosa. In this species the syzygy is very firm and the pairs are not easily separated by normal handling. Cysts. Although well over 100 specimens of M. capensis were examined no cysts of this species, and, consequently, no spores were discovered. Although the mode of dehiscence of the cysts and the shape of the spores of this species are not yet known, the known features, such as the knob-shaped epimerite and the biassociation of the sporonts well before cyst formation, have led to this species being placed in the genus Gregarina Dufour. Diagnostic differences between Gregarina sandont and Gregarina impetuosa Gregarina sandom Gregarina impetuosa (a) Large gregarine. Small gregarine (see average dimensions). (6) Very sluggish. Very active. (c) Nucleus with 4 or 5 nucleoli. Nucleus with 6 to 8 nucleoli. (d) Nucleoli stain only slightly darker Nucleoli stain much darker than than rest of nucleus. the rest of the nucleus. (e) Extra-nuclear body present. No extra-nuclear body. This paper was originally written in 1943 and presented in part fulfilment of the degree of M.Sc. at the University of Gape Town. Thanks are due to Dr. H. Sandon, then Senior Lecturer in Zoology, for help and guidance, and to Dr. A. J. Hesse of the South African Museum for his help in identifying the cockroaches. A set of slides has been deposited in the South African Museum. REFERENCES Bhatia, B. L., 1938. Fauna of British India, Protozoa (Sporozoa). Bhatia, B. L., and Setna, S., 1924. ‘On some new cephaline gregarines.’ Parasitology, XVI, pp- 279-88. Bush, S. F., 1928. ‘A study of the gregarines of the grasshoppers of Pietermaritzburg, Natal.’ Ann. Nat. Mus., V1, pp. 97-169. Gibbs, A. J., 1946. ‘Stylocephalus ingeri, sp. nov., a cephaline gregarine found in the gut of Gonocephalum arenarium (Coleoptera).’ Trans. Roy. Soc. S. Afr., XXXI (2), pp. 169-80. Sprauge, V., 1941. “Studies on Gregarina blattarum with particular reference to the chromosome cycle.’ Illinois Biol. Monogr., XVIII, No. 2. Watson, M. E., 1916. ‘Studies in gregarines.’ Illinois Biol. Monogr., II, No. 3. Watson-Kamm, M., 1922. ‘Studies in gregarines.’ Jllinois Biol. Monogr., VII, No. 1. 17. Finf Dactylispa-Typen Péringuey’s. 165. Beitrag zur Kenntnis der Hispinae (Coleopt. Chrysomelidae). Mit einer Abbildung. Von ErtcH UnMann, Stollberg-Sachsen. Wenn man die siidafrikanischen Dactylispa-Arten nach den Beschrei- bungen Peringuey’s bestimmen will, so steht man vor einer schwer lésbaren Aufgabe. Seine Beschreibungen reichen bei weitem nicht mehr aus, weil, wegen der Fille der Hispinae-Arten des siidafrikanischen Raumes, eine Betrach- tung der Objekte nach Gesichtspunkten erfolgen muss, von denen Péringuey nichts wissen konnte. Es war damals, 1898 und 1908, nicht vorauszusehen, wie gross die Artenzahl sein kénnte. Auch heute wissen wir noch nicht, wie viele Arten eine genaue Durchforschung Siidafrikas liefern wird. Herrn Dr. H. Andreae vom Siidafrikanischen Museum in Kapstadt ist es zu. danken, dass wir heute tiber die Dactylispa-Arten Péringuey’s gut unter- richtet sind. Sein ausserordentlich grosses Entgegenkommen erméglichte mir das Studium von Typen, und wo das nicht anging, hat Herr Kollege Andreae durch genauestes eigenes Studium zur Klarung der gestellten Fragen beige- tragen. Seine Studien haben es erméglicht, dass die Arten Péringuey’s in meinem Schliissel der Dactylispa-Arten Afrikas, der in einer belgischen Zeit- schrift erscheinen soll, mit ziemlicher Sicherheit untergebracht werden konnten. Ihm und dem Siidafrikanischen Museum in Kapstadt sage ich hiermit nochmals meinen herzlichsten Dank. Dactylispa capicola (Péring.) 1898. Ann. S. Afr. Mus., i, p. 121. (Hispa c.) Die Untersuchung des Materials ergab folgendes: ‘Bei capicola sind ausser dem eigentlichen Typ (mit rotem Zettel) zwei solche Stiicke vorhanden.’ Eins davon habe ich gesehen aus Natal: Durban. ‘Der Typus ist fein weisslich behaart. Auf den Decken stehen die Haare zwischen den Punkten; oft ver- schmutzt und schwer sichtbar.’ Péringuey: ‘nearly glabrous’. Die Decken sind aufrechtstehend behaart. ‘Der Halsschild ist mit Ausnahme der Flachen anliegend behaart, wenig dicht, die Haare am Seiten- und Hinterrand nach der Seite, am Vorderrande nach vorn gelagert, nur zwischen den Vorderrand- dornen halb aufgerichtet. Die Stirn ist niedergedriickt und langsrunzlig. Die 407 408 ANNALS OF THE SOUTH AFRICAN MUSEUM Halsschildflachen sind gross, sie nehmen etwa ein Viertel bis ein Fiinftel der Oberflache ein.’ Seitenbewehrung deutlich 2,1. ‘Fithlerglied 1 (von der Seite gesehen) vorn leicht abgeschragt, die untere Kante etwas vorgezogen, die Aussenecke (von oben gesehen) leicht vorspringend. Der Unterschied zwischen den Seitenranddornen und den Spitzenranddornen der Decken ist deutlich, aber viel schwacher als bei gracilis.”. Die Seitenbewehrung ist halb so lang wie der Halsschild breit. ‘Der Abstand der Punktreihen voneinander ist grésser als der Abstand der Punkte in den Reihen, deshalb treten die Zwischenraume in der Langsrichtung mehr hervor. Rippen kann man das eigentlich nicht nennen.’ Péringuey: ‘elytra costulate’. Bei dem Stiick von Durban ist die Oberflache der Decken so stark skulp- tiert, dass zwischen den einzelnen Dornen in der Langsrichtung tatsachlich erhabene Rippenteile erscheinen. Dactylispa congrua (Péring.) IGOG:> loo. cits 123: (Espa) Typus: Natal, Durban. Der Typus ist gut erhalten, es fehlt ihm aber die rechte Decke. Farbung: Kopf, Halsschild, Schildchen und Unterseite schwarz, Fihler gelbbraun, 1. und 2. Fihlerglied schwarz, 3. schw4rzlich-braun, Beine gelbbraun, Decken dunkelbraun mit schwachem Bronzenschimmer, mit schwarzen Dornen. Behaarung: Innenseite der Augen mit hellem Haarsaum, Wangen deutlich behaart, Halsschild auf der Scheibe und auf dem Abfall neben den Vorder- dornen goldgelb behaart, die beiden kleinen glatten Flachen kahl; Decken deutlich behaart, jeder Punkt mit einem goldgelben Harchen vom Durch- messer eines Punktes. Stirn zwischen den Fihlern mit kurzem Kiel, in der Mitte etwas eingedriickt, hinten mit kurzer, schwacher Furche, etwas schmaler als eine Augenbreite, vom glatten Hals fein abgesetzt, nur wenig hoher als dieser. Fihler bis hinter die Schultern reichend, schlank, etwas schmaler als die halbe Stirn, zur Spitze nur schwach verdickt. Glied 1 etwas kraftiger als die anderen, 2 schmiéler, ellipsoidisch, 3 so lang wie 1, schmaler als 2, zylin- drisch, 4-5 zylindrisch, jedes kiirzer als 3, 6 kurz, so lang wie 2, 7-11 eine schwache Keule bildend, 7 konisch, etwas so lang wie 5, 8-10 schwach quer, jedes so lang wie 6, 11 eif6rmig, um die Spitze langer als 10. Alle Glieder fein behaart, die Keule pubeszent. Halsschild auf der Scheibe dicht punktiert, mit sehr feinem Mittelkiel, mit je einer kleinen, runden, kahlen Flache beiderseits, dahinter mit seichtem Quereindruck. Neben den Vorderranddornen mit recht kleinem Borstenkegel, Hinterecken mit vorgezogenem Borstenkegel. Beweh- rung schlank, Vorderranddornen mit fast gleichen Aesten, diese kirzer als eine Halsschildbreite, Seitendornen 2,1, das vordere Paar so lang wie die Vorderdornen, mit kurzem, breiten Stiel, der freie Dorn entfernt, etwa halb so lang wie einer der Seitendornen. Antebasalrand fein. Schildchen dreieckig, schagriniert. Decken glanzend, hinter den Schultern wenig eingezogen, ohne BEITRAG ZUR KENNTNIS DER HISPINAE 409 betonte Zwischenraume II, IV, VI, VIII. Punktreihen gut zu verfolgen, 7. hinter dem Schulterdorn 3 beginnend, g. und 10. in der Mitte vereinigt. Eindruck schwach. Dornen ziemlich lang, schlank, ohne verdickte Basis. Naht hinter den schlanken Schliessdérnchen fein bedornt. Raum I mit einer Reihe von Zusatzdornen; Raum II mit II 1-5, I1 5 diinn; IV mit IV 1,3,5 und einigen diinnen Zusatzdornen; VI auf der Schulter mit den Achseldornen 1,2,3, dann mit VI 2,3,4; VIII mit VIII 4 und 5. Letzterer mit verstarkter Basis, vorn und hinten mit einem Zusatzdorn. Randdornen der linken Decke 17, die langsten kaum langer als die langen Scheibendornen, nach dem Hinter- winkel zu und zur Naht allmahlich kiirzer. Jede Klaue mit einem Basalzahn- chen. 4. mm. Zur Beschreibung Péringuey’s: ‘thorax pubescent, narrowly grooved longitudinally in the centre. ...’ Die vermeintliche Langsgrube ist der Mittel- kiel. ‘elytra . . . glabrous’ stimmt nicht. ‘(elytra) having on each side three dorsal rows of moderately long spines’ ist recht summarisch, vielleicht meint er mit der 1. Reihe die Dornen des II. Raumes, mit der 2. Reihe die des IV. Raumes und mit der 3. Reihe die Dornen des VI. Raumes mit VIII 5. ‘a few short spines along the suture’ ist richtig beobachtet. Dactylispa gratula (Péring.) 1898. loc. cit., p. 123. (Hispa g.) Herr Kollege Dr. Andreae, der den Typus untersucht hat, stellt das Folgende fest (von mir etwas geadndert). Die Art ist durch die Bildung der Stirn besonders ausgezeichnet. Die Stirn hat zwei weitlaufig punktierte und behaarte Hocker, die in ihrer ganzen Lange durch eine glatte, dreieckige, vorn zu beiden Seiten des Stirnkieles flach auslaufende Grube getrennt sind. Die deutlich, aber nicht hoch abgesetzten Hinterrander der Hocker schliessen ein stumpfwinkliges Dreieck ein, das ebenso glatt und glanzend ist wie der Hals und mit ihm in einer Ebene liegt, aber durch eine feine Linie abgetrennt ist. Fiihler ebenso lang und schlank wie bei D. pubicollis Chap. Alle Glieder langer als breit, das 1. noch langer als bei D. pubicollis, etwas vor der Mitte am breitesten, zur Spitze schwach verjiingt, am Vorderrande weniger scharf abgestutzt, etwas langer als das 3. Glied 3-6 an Lange allmahlich abnehmend, 7 deutlich langer als 6, 7-10 an Lange langsam abnehmend, 11 um zwei Drittel langer als 10, aber etwas schmaler, etwa zweimal so lang wie breit. Hlalsschildscheibe matt, dunkel, Bewehrung und Flachen dunkel-gelbbraun, kurz pubeszent, nach meiner Lesart (Uh.) in der Mitte mit sehr feinem Langskiel (‘very faintly grooved longitudinally in the middle’ bei Péringuey). Seitenflachen des Halsschildes klein und schmal, deutlich gewolbt, glatt und glanzend. Sie heben sich von der sonst punktierten, matten, dunklen Oberflache scharf ab. Sie umschliessen eine fast halbkreisf6rmige Flache. Schildchen mit ovaler, tiefer, an der Basis scharf begrenzter, zur Spitze hin flach auslaufender Grube, die den 410 ANNALS OF THE SOUTH AFRICAN MUSEUM gréssten Teil des Schildchens einnimmt. In den Punkten der Decken stehen feine, farblose Haare, etwa so lang wie der Durchmesser eines Punktes. Bei schlechter Beleuchtung kaum erkennbar. Naht im basalen Drittel mit kurzen Dornen, in der Spitzenhalfte mit feinen, nach hinten geneigten Dérnchen von etwa 0,05 mm. Lange. Raum I mit kurzen Dornen im mittleren und Spitzen- drittel; Raum II mit starken Dornen, die zwei ersten basalen und der letzte apikale kirzer; Raum IV kurz hinter der Mitte und am Absturz mit einem Dorn, der ebenso stark ist wie die Dornen des II. Raumes; and der Basis, vor und zwischen den starken Dornen mit je einem kurzen Dorn; Raum VI mit starken Dornen, die in der Basalhalfte langer als die des IJ. Raumes sind; Raum VIII links mit einem kleinen Dorn kurz hinter der Mitte, einem starken am Absturz (VIII 5), rechts mit je einem starken hinter der Mitte und am Absturz, je einem kleinen vor dem ersten starken, unmittelbar vor und hinter dem zweiten (VIII 5), und vor der Spitze. Rand links mit 13 Dornen, so lang wie die von VI, aber dinner, rechts mit 17. In der Basalhalfte sind die Dornen langer, die Spitzenranddornen nach der Naht zu stark verkirzt. Hinter der Schulter messen sie 0,4 mm., am Nahtwinkel 0,1-0,15 mm. 4 X 2mm. Natal (Malvern), G. A. K. Marshall. Uber die Scheibendornen schreibt Péringuey: ‘having two dorsal series of moderately long dark spines, set at some distance from one another [nach meiner Deutung sind das die Dornen von II und VI], with two shorter spines on the posterior part,. between the first and second rows [bezieht sich auf die Bedornung von IV]; the suture is also spinulose [wie oben].’ Dactylispa inanis (Péring.) 1898. loc. cit., p. 124. (Hispa 1.) Typus: Natal, Malvern, VI, 1897 (7893). Dem Typus fehlt die rechte Decke, sonst ist er gut erhalten. Gelbbraun, schwarz: Augen, auf dem Hals ein Fleck hinter jedem Auge, ein Saum um jede Halsschildflache, ein kleines Fleckchen in jeder Ecke der Schildchenbasis, das scheinbar auch ein wenig auf den Hinterrand des Halsschildes tibergreift (vielleicht dort nur durchscheinend), die Dornen der Deckenscheibe mit Ausnahme von Achseldorn 1, ein Streifen auf der Schulterflache innen, ein Fleck auf dem II. Zwischenraum an der Basis, am Hinterwinkel einige Rand- dornen (angedunkelt) und die Seiten der Hinterbrust. Behaarung: diinn, wenig auffallend, hellgelb, ein Saum am Innenrande der Augen, die Mitte der Scheibe des Halsschildes, aber undicht, jeder Punkt der Decken mit einem diinnen Harchen. Fiihler ziemlich dicht behaart. Stirn in der Mitte ein- gedriickt, mit kurzen Kiel zwischen den Augen, vom glanzenden Hals abgeschniirt, allmahlich zu ihm hinabgewolbt. Filer bis hinter die Schultern reichend, kraftig, zur Spitze nur ganz allmahlich verdickt. Glied 1 gross, stark, ellipsoidisch, Glied 2 fast kuglig, etwa ein Drittel so lang wie I, Glied 3 so lang wie 1, aber schmiler, zylindrisch, wenig langer als 4, Glied 4-6 einander BEITRAG ZUR KENNTNIS DER HISPINAE 411 ahnlich, von abnehmender Lange, 7 schwach konisch, ein wenig verdickt, so lang wie 4, 8-10 unter sich fast gleich, etwas langer als breit, 11 wenig langer als 10. Halsschild in der Mitte der Scheibe gewolbt, punktiert, beiderseits stark niedergedriickt, mit feinem Mittelkiel, beiderseits mit gewélbter, glatter Querflache, der niedergedriickte Teil unpunktiert, unbehaart, Ante- basalrand fein, davor flach quergefurcht. Vordere Borstenkegel klein, auf dem Vorderrande neben den Vorderdornen, hintere Borstenkegel an der Spitze der rechtwinkeligen Hinterecken. Bewehrung stark, kraftig. Vorderast der Vorderdornen kiirzer als der Hinterast. Seitenbewehrung 2,1, der Seiten- rand dort konvex erweitert, der mittlere Dorn etwas langer als der vordere, der freie Dorn etwa halb so lang wie der mittlere. Schildchen dreieckig, mit abgerundeter Spitze. Decken parallel, Raum II, IV, VI teilweise schwach rippenformig. Punktreihen regelmdssig, die intrahumeralen an der Spitze nach aussen gebogen. Dornen krdaftig, mit verdickter Basis. Naht mit sehr kleinen Zahnchen, die nach dem Hinterrande zu etwas grésser werden, Schliessd6rnchen etwas grdsser. Raum I vor der Mitte mit einem einzelnen Dérnchen; Raum II mit IT 1-5 (II 2 grésster Scheibendorn, II 5 klein); IV mit IV 1,3,5; VI mit 3 ziemlich gleichen Achseldornen, mit VI 2,3,4; VIII mit VIII 2,3 (beide klein), 4 und 5 (Doppeldorn). 3 Basaldérnchen und ein Schildchen-Zusatzdérnchen. 17 Randdornen, die vom Hinterwinkel bis zur Naht an Lange abnehmen, die Spitzenranddornen recht kurz. Klauen- zahnchen undeutlich. Zur Beschreibung Péringuey’s: ‘faintly grooved longitudinally in the middle’, muss ‘carinate’ heissen. Nach der Beschreibung der Dornenverteilung auf den Decken ist die Art nicht zu erkennen. Péringuey meint, es seien 2 Reihen Dornen da, das waren die auf Raum II und IV, ‘with two isolated ones in the second interval’ bezieht sich auf IV 3 und IV 5; ‘two set close to one another, etc.’ bezieht sich auf den Doppeldorn VIII 5. Dactylispa pretiosula Péring. 1908. Ann. S. Afr. Mus., 5, p. 337- 2 Typen. Malvern, Oct. 22, 1907. Beide Typen sind sehr gut erhalten. Sie sind auf ein Stiick Karton geklebt. Ich wahle das grdéssere, rechte Stiick als Lectotypus (Abb. 1). Farbung: Glanzend, gelbbraun, schwarz: Augen, Umegebung der drei glatten Halsschildflachen, Randdornen des Aussen- winkels und die Dornen der Deckenscheibe mit jeder Basis, auf der Scheibe selbst die Schulter auf der Innenflache, die Schulterkehle, Aussenwinkel um VIII 5 herum und einige Langsstreifen auf den Zwischenstreifen, z.B. ein Fleckchen hinterm Schildchen auf der Schildchenreihe, die Naht (ange- dunkelt), der III. Zwischenraum im 2. Zwischenstreifen, auf dem 3. Zwischen- streifen der V. Raum im Anschluss an die Schulterschwarzung, eine Querbinde zwischen den Dornen II 5 bis VIII 5 (das Schwarz zuweilen stellenweise mit schwachem Metallschein). Angedunkelt ist der glatte Hals und die beiden 412 ANNALS OF THE SOUTH AFRICAN MUSEUM Ie Ire a) rar = PAA i - wie SNES ers Oey a = ee, oer is) x oR D <) PORE Siete Ay reat Onin 3 ; < S = a ee . ~ "Mee. Deo as EMS sera Bare Ries = OO igs Ae ean Soe eae as = 5 CN ge, = —¢ Aes ox ; ee ee “ioe 7D Ay o a) y foe ~ 3 Sees Oy ey c 4S ? g ' (se) ESS WEBB be Ta (PI J " ee A a in BS lis, se, Dactylispa pretiosula Péring. Lectotypus. B = Basaldérnchen. S = Schliessd6rnchen an der Naht. Z = Schildchen-Zusatzd6rnchen, 1,2,3, = Achseldornen. 1,2,3, VI 2 = Schulterdornen. II, IV, VI, VIII = gerade (Zwischen) rdume. Von den Dornen sind auf der rechten Decke in ihrer wahren Grésse gezeichnet: II 224-5 Vite Sens OW bee OWN leo ora a Vie Aas BEITRAG ZUR KENNTNIS DER HISPINAE 413 Basalglieder der Fiihler und noch einige Stellen auf den Decken. Die helle und die dunkle Farbung sind nicht scharf abgegrenzt. Unterscheide: Zwischenraume oder Raume liegen zwischen den einzelnen Punktreihen; Zwischenstreifen oder Streifen liegen zwischen den Rippen, zwischen Naht und I. Rippe (II. Raum), oder zwischen I. und II. Rippe (das ist zwischen II. und IV. Raum) und so weiter. Behaarung. Innenrand der Augen mit hellem Haarsaum, Scheibe des Halsschildes mit einzelnen Harchen, jeder Punkt der Decken mit einem langen, gekriimmten, hellen Harchen. Auf dem Halsschild sind die Harchen sicher zum Teil abgerieben. Fihler fein behaart. Stirn mit den gewdlbten Augen in einer Flucht verrundet, erstere zwischen den Fihlern mit kurzem Kielchen. Kopf vom glatten Hals abgeschniirt, Stirn zum Hals hinabgewoélbt. Hals mit mattem Fleck in der Mitte (Stridulationsapparat?). Fiihler die Schultern erreichend, diinn, Glied 7-11 nur unmerklich verdickt. Glied 1 von der iiblichen Starke, 2 ellipsoidisch, langer als breit, 3-5 zylindrisch, unter sich fast gleich, 3 so lang wie 1, 6 wie die vorhergehenden, etwas kiirzer als 5, 7 so lang wie 5, 8-10 doppelt so lang wie breit, 11 um die Spitze langer als to. Halsschild auf der Scheibe niedergedriickt und dicht gerunzelt-punktiert, mit glattem Mittelstreifen (Kielchen) und einer schragen, ovalen, gewdélbten Flache beiderseits. Antebasalrand fein. Vordere Borstenkegel klein, unweit der Vorderdornen, hintere Borstenkegel an den spitz vorspringenden Hinter- ecken. Bewehrung kurz und krAaftig, der freie Dorn etwas kiirzer als die anderen, unter sich fast gleich langen Dornen. Schildchen breit, dreieckig, mit schwachem Eindruck. Decken hinter den Schultern nicht eingezogen, zur Spitze etwas verbreitert, mit regelmassigen Reihen und Raumen. 9g. und Io. Reihe in der Mitte kurz vereinigt. Schildchenreihe und gerade Raume rippenartig, der VIII aber nur schwach konvex. Naht mit einem Paar senk- rechter Schliessdérnchen und einer Reihe winziger Zahnchen; Raum I mit einem einzelnen Doérnchen hinter der Mitte; II mit II 1,2,4,5; IV mit IV 1,3,5; VI auf der Schulter mit den Achseldornen 1,2,3, dann VI 2,3,4,; VIII mit VIII 4 und dem Doppeldorn VIII 5 auf erhabener Basis. Randdornen beiderseits 14, die des Seitenrandes so lang wie die Dornen der Scheibe, am Hinterwinkel und bis zum Nahtwinkel von abnehmender Lange, die nahe am Nahtwinkel recht kurz. Eindruck deutlich. Klauenglied mit einem Zahnchen an jeder Klaue. Lectotypus 3,5 x 2 mm. (vor der Spitze), Paratypoid etwas kirzer und schmdler, 34 x 14 mm. Es ist nicht ausgeschlossen, dass D. pretiosula Péring. und D. hirsuta Gest. identisch sind. Die ausfiihrliche Beschreibung Gestro’s passt ganz auf unsere Art, vielleicht haben unsere beiden Typen auf den Decken etwas mehr schwarze Streifen. Eine Synonymierung muss aber bis zum Vergleich der Typen unterbleiden. Bei D. hirsuta gibt es gewisse Abweichungen, auf die ich schon in friiheren Arbeiten hingewiesen habe. Man weiss nicht, ob sie individuell oder artbedingend sind. Vielleicht handelt es sich hier um einen Formenkreis einander sehr ahnlicher Arten. 414 ANNALS OF THE SOUTH AFRICAN MUSEUM Zur Beschreibung Péringuey’s: (1) ‘Pallide straminea’, die Typen sind nachgedunkelt. (2) ‘fere glabra’, nur auf den Vorderk6érper zu beziehen. (3) ‘Caput punctulatum’, die Stirn ist nicht punktuliert, aber schagriniert. (4) ‘striga’, bezieht sich auf den kurzen Mittelkiel des Halsschildes. (5) ‘inter- stitiis vix distincte albido setulosis’, bei x 22 ist die Behaarung deutlich. , The ANNALS OF THE SO UTH AFRICAN M USEUM | are teued ¢ in oe at {hee 3 _intervals as material becomes available. As far as possible each volume i is devoted exclusively to a particular subject (Zoology, Botany, etc.). Two or more volumes: may bei in course oo Sa publication concurrently. _ 2 ae ae fe j _ Most of the Geological and Palacontoldpieal papers are issued in conjunction with the a Geological Survey of the Union of South Africa. oe tae eae Some volumes and parts are out of print, and others are only sold as parts ofa ‘set, or volume, . a respectively. The prices of parts published prior to 1940 have been increased. 5 ae Out of print: Vols. I, II, V a I, 2, 9), VII, VII, IX (Part 1), XII (Part 7)s Xx XXIV (Part 2), XXXI (Paris 1, 2, 3). : Vol. III. 1903-1905 Zoology .. ve i ae Nis a Peis IV. 1903-1908 Palaeontology : cree V. 1906-1910 Geology, Palaeontology, Zoclogy, Anthropology NMI. 1908-1910 Zoology IX. 1g11-1918 Botany .. as te ats 2. ek Part ) X. «g11-1914 Zoology : we a ae = XI. s911-1918 Zoology ate Sc 3s ae XII. 1913-1924 Palaeontology and Ghotney af .. (excl. Pare: 7) XIII. 1913-1923 ee rn and Zoology an «a eee XIV. 1915-1924 Zoology : ra : XV. 1914-1916 Zoology XVI. 1917-1933 Botany a