BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) 12 JAN 1967 ZOOLOGY VOL. 12 1964-1965 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) LONDON: 1967 DATES OF PUBLICATION OF THE PARTS No. i ..... 16 September 1964 No. 2 ..... 2 April 1965 No. 3 15 April 1965 No. 4 ..... 21 April 1965 No. 5 ..... 30 April 1965 No. 6 28 May 1965 No. 7 10 June 1965 No. 8 6 August 1965 No. 9 ... . . 6 August 1965 PRINTED IN GREAT BRITAIN BY ADLARD & SON LIMITED BARTHOLOMEW PRESS, DORKING CONTENTS ZOOLOGY VOLUME 12 PAGE No. i. Polyzoa from West Africa. Notes on the genera Hippoporina Neviani, Hippoporella Canu, Cleidochasma Harmer and Hippoporidra Canu & Bassler (Cheilostomata, Ascophora). By P. L. COOK (Pis. 1-3) I No. 2. Freshwater Gastropod Mollusca from Ethiopia. By D. S. BROWN (Pis. 1-3) 37 No. 3. The British Museum (Natural History) expedition to East Nepal 1961-62. Introduction and lists of localities. By J. G. SHEALS and W. G. INGLIS (Pis. 1-7) 95 No. 4. A preliminary revision of the Indo-Pacific Alosinae (Pisces : Clu- peidae). By P. J. P. WHITEHEAD 115 No. 5. A critical review of the marine Nematode genus Euchromadora De Man, 1886. By JOHN W. COLES 157 No. 6. The feeding mechanisms and preferred foods of three species of Pycnogonida. By WILLIAM G. FRY (Pis. 1-5) 195 No. 7. A review of the Elopoid and Clupeoid fishes of the Red Sea and adjacent regions. By P. J. P. WHITEHEAD 225 No. 8. Acanthodrilidae and Eudrilidae (Oligochaeta) from Ghana. By R. W. SIMS 283 No. 9. The Cichlid fishes of Lake Nabugabo, Uganda. By P. H. GREENWOOD 313 Index to Volume 12 359 POLYZOA FROM WEST AFRICA NOTES ON THE GENERA HIPPOPORINA NEVIANI, HIPPOPORELLA CANU, CLEIDOCHASMA HARMER AND HIPPOPORIDRA CANU & BASSLER (CHEILOSTOMATA, ASCOPHORA) P. L. COOK BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. i LONDON: 1964 POLYZOA FROM WEST AFRICA NOTES ON THE GENERA HIPPOPORINA NEVIANI, HIPPOPORELLA CANU, CLEIDOCHASMA HARMER AND HIPPOPORIDRA CANU & BASSLER (CHEILOSTOMATA, ASCOPHORA) BY P. L. COOK Department of Zoology, British Museum (Natural History) Pp. 1-35 ; Plates 1-3 ; 8 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. i LONDON: 1964 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. This paper is Vol. 12, No. i of the Zoological series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals. Trustees of the British Museum (Natural History) 1964 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued September 1964 Price Fifteen Shillings POLYZOA FROM WEST AFRICA NOTES ON THE GENERA HIPPOPORINA NEVIANI, HIPPOPORELLA CANU, CLEIDOCHASMA HARMER AND HIPPOPORIDRA CANU & BASSLER (CHEILOSTOMATA, ASCOPHORA) By PATRICIA L. COOK CONTENTS Page 1. INTRODUCTION .......... 3 2. Hippoporina Neviani ......... 4 3. H. pertusa (Esper) ......... 5 4. H. americana (Verrill) ........ 6 5. H. lacrimosa sp. n. ......... 7 6. H. acuta sp. n. .......... 8 7. Hippoporella Canu ......... 8. H. gorgonensis Hastings ........ 9 9. H. pusilla (Smitt) 10 10. Cleidochasma Harmer . ........ n 11. C. porcellanum (Busk) . . . . . . . . n 12. C. contractum (Waters) . . . . . . . . 14 13. C. oranense (Waters) ..... ... 17 14. C. brancoense (Calvet) ......... 19 15. C. rotundorum (Norman) ...... 20 16. C. granulosum (Canu & Bassler) and C. spiculiferum (Canu & Bassler) 21 17. Hippoporidra Canu & Bassler .... .22 18. NOTES ON THE ASSOCIATION OF GASTROPODS AND PAGURIDS (HERMIT- CRABS) WITH ENCRUSTING ANIMALS .... • 23 19. DEVELOPMENT AND STRUCTURE OF THE ZOARIUM IN Hippoporidra . 24 20. H. edax (Busk) 26 21. H. janthina (Smitt) ......... 27 22. H. littoralis sp. n. .......•• 23. H. senegambiensis (Carter) ..... .29 24. H. picardi Gautier ..... 31 25. ACKNOWLEDGMENTS ......••• 32 26. SUMMARY ......••••• 32 27. REFERENCES ...... • • 33 i. INTRODUCTION POLYZOA from the north and north-western coasts of Africa and from the neighbour- ing islands, have been described ; but there has been hitherto no systematic collection from the coast of southern Senegal and the Gulf of Guinea. The collections here 4 PATRICIA L. COOK described consist of the " Calypso " Collection I, from Senegal and the Bay of Biafra, and Collection II, from the Cape Verde Islands ; the Marche-Marchad Collections, from Senegal ; and the Achimota Collection, from Ghana (see Cook, 1964 : 44). The holotypes of new species, together with representative specimens of all the material described, are deposited in the British Museum (Natural History). Material was treated with eau de javelle for examination of calcareous parts, and decalcified and stained to show chitinous parts. The following measurements were made where possible : Length of zooecium . . Lz Width of operculum . . lop Width of zooecium . . Iz Length of avicularium . La Length of orifice . . Lo Length of mandible . . Lm Width of orifice . . lo Length of ovicell . . Lov Length of operculum . . Lop Width of ovicell . . lov Specimens in the British Museum, Zoology Department, are referred to by their registered numbers, thus : 1911.10.1.1134, and in the Palaeontology Department thus : B 1620. Species assigned by various authors to Hippoporina belong to three distinct genera. These were defined by Harmer (1957 : 977 and 1038), and their synonymies may be summarized as follows : — 1. Hippoporina Neviani 1895, type-species Cellepora pertusa Esper (= ? Hippo- diplosia Canu, 1916, type-species H. verrucosa Canu, chosen by Hastings, 1930. If Hippodiplosia verrucosa is congeneric with Hippoporina pertusa, Hippodiplosia is a synonym of Hippoporina, but this is very doubtful). 2. Hippoporella Canu, 1917, type-species Lepralia hippopus Smitt ( = Hippo- porina, type-species as used by Canu, 1917, Lepralia hippopus Smitt ; and Hippoponella1 Canu & Bassler, 1920, type-species Lepralia hippopus Smitt). 3. Cleidochasma Harmer, 1957, type-species Gemellipora protrusa Thornley (= Hippoporina Canu & Bassler, 1920, not Neviani, type-species Lepralia porcellana Busk). The genus Hippoporidra is similar in character to Cleidochasma, but is here confined to species with cortical zooecia and a special type of interzooecial avicula- rium (see p. 22), which encrust gastropod shells. 2. HIPPOPORINA Neviani Hippoporina Neviani, 1895 : 109. Harmer, 1957 : 977- Hippodiplosia Canu : Osburn, 1952 : 338. TYPE SPECIES. Cellepora pertusa Esper, 1796 : 149. 1 1 agree with Hastings (1930 : 724) and Osburn (1952 : 348) in regarding Hippoponella as an absolute synonym of Hippoporella. The second type-species of Hippoporella later chosen by Canu & Bassler (1920 : 377), Hippoporella perforata, was not included in the original introduction of the genus by Canu, but is congeneric with H. hippopus. Examination of Hincks's figured specimen of Lepralia hippopus (1880, pi. 33, figs. 8, 9, 1911.10.1.1134, Northumberland coast, Norman Coll.) which has all the characters of Smitt's description and figures, shows that, contrary to Canu & Bassler's diagnosis of Hippoponella, the ovicell is not closed by the operculum. POLYZOA FROM WEST AFRICA DIAGNOSIS. Zoarium encrusting. Zooecia with marginal pores and frontal pores, but with the area proximal to the orifice without pores ; lateral and distal septula present. Orifice rounded with a pair of proximal lateral denticles forming a broad sinus, lateral peristome frequently slightly raised. Adventitious avicularia often present. Ovicell hyperstomial, porous, not closed by the operculum. 3. Hippoporina pertusa (Esper) (Text-fig. IA) Cellepora pertusa Esper, 1796 : 149, pi. 10, figs. 1,2, Recent, Mediterranean. Hippodiplosia pertusa (Esper) Hastings, 1930 : 724, pi. 17, fig. 118, Galapagos, 4-15 fath., Gorgona, Panama. Osburn, 1952 : 340, pi. 40, figs. 5-8, S. California, 17 and 23 fath., Mexico and Galapagos. MATERIAL. Achimota Coll. Stn. 47, Dredge haul No. i, 4.1.51, 44 m., i4"V. Stn. 117, Agassiz trawl No. 2, 5.iv.5i, 64 m., 320. British Museum, Galapagos, 1929.4.26.142, " St. George " Coll. Pt. Vincente, FIG. i. Hippoporina, treated with eau de javelle. A. H. pertusa, zooecium and ovicell. Ghana, Achimota Coll. I4V. B. H. americana, zooecium and ovicell. Nigeria, 1959.2.20.4. California, 65 ft., 1962.6.16.12, and Gulf of Mexico, 28° 58' N., 89° 9' W., 1961.11.2.35, Cheetham Coll., and many other specimens. H. pertusa is not present in the " Calypso " or Marche-Marchad Collections. 6 PATRICIA L. COOK DIAGNOSIS. Hippoporina with large, wide zooecia. Avicularia usually absent. Ovicells large, wide, pores frequently irregular. Dimensions. Lz 0-70 mm., Iz 0-60 mm., Lo 0-225 nun., lo 0-25 mm., Lov 0-35 mm., lov 0-50 mm. No avicularia have been seen in any of the material examined. There are 3-4 lateral and 4-5 distal septula, the latter either scattered or in a row near the base of the distal wall. H. pertusa has a wide geographical distribution extending from the Arctic to California, Florida and Suez (see Hastings, 1930, and Osburn, 1952). The west African specimens show little variation and are very similar to material from Galapagos and the Gulf of Mexico. 4. Hippoporina americana (Verrill) (Text-fig. IB) Lepralia americana Venill, iSysa : 415, pi. 7, fig. 4, Long Island Sound to Beverley, Mass., low-water to 30 fath. (not pi. 7, fig. 5 = Cryptosula pallasiana (Moll.), see Osburn, 1912 : 241). Hippodiplosia americana (Verrill) Hastings, 1930 : 725, pi. n, fig. 61, Balboa, Panama. Marcus 1937 : IOI» pl- 20» fi§s- 54A> B> Brazil, Bahia de Santos, 10-20 m. Osburn, 1952 : 339, pi. 40, fig. 4, Mount Desert, Maine to N. Carolina, and Gulf of California to Costa Rica and Galapagos, 10-30 fath. MATERIAL. British Museum, Lagos, Nigeria, 1959.2.18.2, Hill Coll. Port Harcourt, Nigeria, 1959.2.20.3, 4, 6 (on wood, gastropod and oyster shells respectively) ; Freetown, Sierra Leone, 1959.2.20.5 (on Balanus amphitrite amphitrite Darwin), Stubbings Coll. W. Africa, 1959.2.22.1, Frewen Coll. Balboa 4, Panama, 1929.4. 26.147, " St. George " Coll. Santos, Brazil, 1948.2.16.29, Marcus Coll. H. americana is not present in the " Calypso " Collections, the Marche-Marchad Collections, or the Achimota Collections. DIAGNOSIS. Zoarium encrusting, sometimes plurilaminar. Peristome raised laterally, a large umbo frequently present proximally to the orifice. Avicularia often absent, when present with acute mandibles, directed centrally and distally, sub-rostral cavity swollen, with a complete bar. Dimensions. Lz 0-38 mm., Iz 0-30 mm., Lo 0-12 mm., lo 0-12 mm., Lm 0-05 mm., Lov 0-20 mm., lov 0-225 mm. There are 3-4 lateral and 4 distal septula, about one-third of the way up the distal wall. The avicularia appear to arise secondarily, as in several cases the swollen sub-rostral chamber, when developing or broken, shows the frontal pores beneath it (see Text-fig. IB). The bar supporting the mandible is slender, and is frequently broken even in slightly worn specimens. The ovicell is often very lightly calcified frontally, and the pores may be very large and irregular. H. americana appears to be tolerant of a wide range of temperature and of slightly lowered salinity, the west African localities listed above being associated with river estuaries and saline lagoons. POLYZOA FROM WEST AFRICA 7 5. Hippoporina lacrimosa sp. n. (Text-fig. 2A) HOLOTYPE. B.M. 1963.4.16.4, Achimota Coll., Stn. 47. MATERIAL. Marche-Marchad Coll. I, 3gE, Baie de Seminoles, Gore"e, 8.xii.53, 38 m., growing over the convex surface of Cupuladria sp. 3gA. Coll. II, 25F, S.W. du Cap. Manuel, Dakar, 20.11.56, on Cellepora sp. 250. Achimota Coll. Stn. 47, Dredge haul No. i, 4.1.51, 44 m., 148, Paratypes, and Stn. 48, as above, No. 2, 44 m., 46J. Stn. no, Agassiz trawl, No. i, 4.iv.5i, 40 m., 48L. Stn.in, as above, No. 2, 43 m., 4gB. Stn. 112, as above, No. 3, 43 m., 6oJ. Stn. 132, as above, 2.v.5i, 44 m., 4oD and 42,]". H. lacrimosa is not present in the " Calypso " Collections. B 0 FIG. 2. Hippoporina, treated with eau de javelle. A. H, lacrimosa, zooecia with avicularia, and ovicell. Ghana, Achimota Coll. 148. B. H. acuta, zooecium with avicularia, and ovicell. Ghana, Achimota Coll. 5gD. DIAGNOSIS. Zoarium encrusting. Small lateral peristomial tubercles and sub- oral umbo frequently present. Avicularia sometimes paired, sub-rostral cavity 8 PATRICIA L. COOK slightly swollen, with a delicate bar. Rostrum and mandible broadly spathulate, directed proximally. Ovicell porous, prominent, occasionally extending over two distal zooecia. Dimensions. Lz 0-50 mm., Iz 0-35 mm., Lo o-n mm., lo 0-125 mm., Lm o-io mm., Lov 0-20 mm., lov 0-27 mm. The distal communication pores of H. lacrimosa differ from those of the other species as they are so low on the wall that they may pass through the angle between the distal and basal walls. In these cases, secondary calcification round the pores produces an appearance very similar to that of dietellae. 6. Hippoporina acuta sp. n. (Text-fig. 2B) HOLOTYPE. B.M. 1963.4.16.5, Achimota Coll. Stn. 35. MATERIAL. Achimota Coll. Stn. 35, Dredge No. i, 2i.xii.5o, 37 mm., 590 (on Jullienella foetida}, Paratypes. Stn. X, Axim, Hospital reef, 7.1.51, 68J (with 68E Hippoporidra littoralis see p. 28). " Calypso " Coll. I. Stn. 114, devant baie 8. Antonio, 8.vii.56, 50 m., C 15 O. H. acuta is not present in the " Calypso " Collection II from the Cape Verde Islands, or in the Marche-Marchad Collections. DIAGNOSIS. Zoarium encrusting. Zooecia similar to those of H. lacrimosa. Avicularia sometimes paired, lateral to the orifice, sub-rostral cavity very swollen ; with a complete bar, directed proximally and centrally. Mandible triangular, acute, curved. Ovicell porous, less prominent than in H. lacrimosa. Dimensions. Lz 0-40 mm., Iz 0-30 mm., Lo o-n mm., lo 0-12 mm., Lm o-io mm., Lov 0-20 mm., lov 0-32 mm. H. acuta resembles the figure of Hippothoa reversa (Verrill i875b : 41, pi. 3, fig. i, Gay Head, New England), but differs in the inward orientation of the avicularia. H. acuta and H . lacrimosa are very similar but no form intermediate to the two very different types of avicularia has been seen. The two species also differ in the position of the distal septula, which in H. acuta are like those of H. americana ; and in the wider, less prominent ovicells of H. acuta. The oral denticles of H. acuta are slightly more proximal in position than those of H. lacrimosa. Material of both species in the Achimota Coll. encrusts the Foraminiferan, Jullienella foetida Schlumberger, and thus belongs to the silty-sand community of Buchanan (1958) and Bassindale (1961). 7. HIPPOPORELLA Canu Hippoporella Canu, 1917 : 36. Osburn, 1952 : 348. Hippoponella Canu & Bassler, 1920 : 379. TYPE-SPECIES. Lepralia hippopus Smitt, 1868 : 20, 127, pi. 26, figs. 99-105, Spitzbergen and Finmark. DIAGNOSIS. Zooecia small, frontal pores rare. Orifice with oral spines and a pair of small proximal-lateral denticles, proximal margin slightly curved. Small POLYZOA FROM WEST AFRICA adventitious avicularia and occasional vicarious avicularia present. Ovicells imperf orate, not closed by operculum. 8. Hippoporella gorgonensis Hastings (PL i, fig. i, Text-fig. 3A) Hippoporella gorgonensis Hastings, 1930 : 723, pi. 12, figs. 62-72, pi. 17, figs. 119, 121, Galapagos Coiba, and Gorgona, shore to 15 fath. Marcus, 1937 : 99. pi- J9» ngs- 53A> B, Brazil, Bahia de Santos, 20 m. Soule, 1961 : 20, Gulf of California, 1-17-5 fath. HOLOTYPE. B.M. 1929.4.26.137, Gorgona. MATERIAL. Achimota Coll. Stn. 35, dredge haul No. i, 2i.xii.5o, 37 m., 590. Stn. 72, as above, No. 3, 24.1.51, 38 m., 6iO. Stn. no, Agassiz trawl No. i, 4.iv.5i, 40 m., 488. Stn. 133, as above, No. 3, 2.v.5i, 51 m., 45!!. Stn. K, on trawl debris, i mile offshore, 2 miles beyond Densu, 2.iii.49, 360 and 44H. British Museum, Galapagos, 1929.4.26.130, 131, 134-136, 139, 141. Gorgona, 1929.4.26.132, 133, 137. Coiba, 1929.4.26.161. Taboga, 1929.4.26.138. Jicaron, 1929.4.26.140, " St. George " Coll. Santos Brazil, 1948.2.16.30 and Guariya, Brazil, 1948.2.16.78, Marcus Coll. H. gorgonensis is not present in the " Calypso " or the Marche-Marchad Collections. a 0-2mm FIG. 3. Opercula in Hippoporella. A. H. gorgonensis, Ghana, Achimota Coll. 44H. B. H. hippopus, Kola peninsula, 1899.7.1.1430. c. H. multidentata, Siboga Coll. Sulu archipelago, 620. D. H. pusilla, Cape Verde Islands, 1926.10.1.19. DIAGNOSIS. Zoarium encrusting, plurilaminar. Zooecia with marginal pores but no frontal pores, and 2-4 lateral and distal septula, the latter low on the distal wall. Orifice with 4 oral spines and a pair of minute proximal-lateral denticles, proximal border nearly straight ; surrounded by lateral and proximal umbos. Avicularia adventitious, frequent, variable, usually paired, one each side of the orifice, occasionally larger, variously orientated. Ovicell hyperstomial, not porous, frequently with a spiked umbo on the front. io PATRICIA L. COOK Dimensions. Lz 0-30 mm., lz 0-25 mm., Lo 0-08 mm., lo 0-08 mm., Lm 0-04 mm., Lov 0-12 mm., lov 0-20 mm. The young zooecia at the growing edge may be without avicularia or umbos surrounding the orifice, they consequently have a very different appearance to those in older parts of the colony (see Hastings, 1930 : 723). The opercula of the west African specimens appear to be of the type with curved lateral sclerites figured by Hastings (1930, pi. 12, fig. 72) ; they are very similar to those of H. hippopus (Kola Peninsula, 1899.7.1.1430, Busk. Coll.). H. gorgonensis is very similar to specimens of H. multidentata (see Harmer, 1957 : 1090, pi. 73, figs. 9-12), from Ceylon (1936.12.30.41^) and India (1936.12.30.117), but opercula of H. multidentata from the Sulu Archipelago (Siboga Coll. 626), do not have any lateral sclerites (see Text-fig. 30). H. gorgonensis has 4 oral spines and H. multi- dentata has 6. When an ovicell is present the number of spines is reduced to 2 in H. gorgonensis and 4 in H. multidentata. Avicularia are present on the ovicelled zooecia of H. gorgonensis, but absent in H. multidentata. Harmer noted that the ovicells of H. gorgonensis had a " more complete frontal wall " than those of H. multidentata. H. gorgonensis has not been previously reported from the eastern Atlantic. 9. Hippoporella pusilla (Smitt) (Text-fig. 30) Discopora albirostris forma pusilla Smitt, 1873 : 70, pi. 12, fig. 233, Florida, 9-60 fath. Cellepora sexspinosa Waters, 1899, pi. 3, fig. 12, Madeira (no description). Norman, 1909 : 311. Lepralia watersi Calvet, 19060 : 216 ; 1907 : 412, pi. 27, fig. n, Cape Verde Islands, 20-25 m. Holoporella pusilla (Smitt) Waters, igiSa : 22, pi. 2, figs. 7-9, Cape Verde Islands, io fath. MATERIAL. " Calypso " Coll. II. Stn. 26, 15° 16' 30" N., 23° 47' 31" W., i8.xi.59, 50-65 m., ile Sao Tiago, C65l. Stn. 42, Punto do Anciao, 20.xi.59, 15-30 and 60 m., Cio4E. Stn. 75, 16° 04' 20" N., 22° 58' io" W., 25.xi.59, 45 m., Cio6F. Stn. 86, 16° 36' 20" N., 22° 52' 05" W., 26.xi.5g, 25 m., C7iE. British Museum, Cape Verde Islands, 1926.10.1.19, Waters Coll. Museo do Seminario, Funchal. i slide, Type of C. sexspinosa Waters. Hippoporella pusilla is not represented in the Marche-Marchad or Achimota Collections. DIAGNOSIS. Zoarium encrusting, plurilaminar. Zooecia larger than those of H. gorgonensis or H. multidentata. Orifice lepralioid with a large sub-oral umbo, a pair of distinct oral denticles and 6 long oral spines. One or two small avicularia near the orifice, with pointed mandibles, directed distally. Occasionally other adventitious avicularia present. Ovicell small, finely tuberculate, raised above the orifice and wide open frontally. Dimensions. Lz 0-45 mm., lz 0-30 mm., Lo o-io mm., lo o-io mm., Lm 0-06 mm., Lov o-io mm., lov 0-15 mm. Calvet tentatively suggested that his L. watersi was synonymous with C. sexspinosa, and Waters, in describing material from the Cape Verde Islands stated that L. watersi was identical with both C. sexspinosa and H. pusilla. The type of C. sexspinosa is a very young colony without ovicells, but shows the long oral spines. Harmer POLYZOA FROM WEST AFRICA n (1957 : 1097) considered that the ovicells of H. pusilla were unlike those normally found in Hippoporella, but the ovicells of Waters's Cape Verde Islands specimens (1926.10.1.19) are similar to those of H. multidentata. They are small and very open frontally, and significantly narrower than those of H. gorgonensis. In ovicelled zooecia, only one pair of spines is present, thus differing from H. multidentata. The large umbo below the orifice is not accompanied by lateral processes, as in H. gorgonensis. H. pusilla differs from both H. gorgonensis and H. multidentata in the form of the operculum and Hastings (1930 : 724) remarked on the differences between the opercula of these two species. Those of H. pusilla are the larger, with a nearly straight proximal edge, and with a pair of very short lateral sclerites, ending distally in strong protuberances which articulate with the oral denticles. These protuberances give the operculum the characteristic notched appearance figured by Waters (igiSa, pi. 2, fig. 8), and see Text-fig. 3D. 10. CLEIDOCHASMA Harmer Cleidochasma Harmer, 1957 : IO38- Soule, 1961 : 18. TYPE SPECIES. Gemellipora protrusa Thornley, 1905 : 119, pi. 7, Ceylon. DIAGNOSIS. Zoarium plurilaminar, encrusting or erect. Orifice cleithridiate, with a pair of well-developed condyles. Oral spines frequently present.1 Marginal and a few frontal pores present, lateral and distal uniporous septula, or multiporous rosette plates or dietellae present. Interzooecial and adventitious avicularia present. Ovicell hyperstomial, with a frontal area, not closed by the operculum. ii. Cleidochasma porcellanum (Busk) (PI. i, fig. 4, PI. 2, figs, i, 2, Text-figs. 4A-E) Lepralia porcellana Busk, 1860 : 283, pi. 31, fig. 3, Madeira. Lepralia cleidostoma Smitt, 1873 : 62, pi. n, figs. 217-219, Florida, 30-120 fath. Hippoporina porcellana (Busk) Hastings, 1930 : 722, Taboga, Gorgona, Galapagos, 1-30 fath., synonymy. Sakakura, 1935 : 23, pi. 4, fig. 6, Pleistocene, Japan. Marcus, 1937 : 9^, pi. 19, fig 51, Brazil, Santos, 20 m. ; 1938 : 217, text-fig. 17, St. Helena, 27-55 m., synonymy and distribution. Osburn, 1952 : 344, pi. 41, figs. 1-3, southern California to Callao, Peru. Hippoporina simplex Canu & Bassler, 1930 b : 49, pi. 6, figs. 3, 6, Tunisia. Hippoporina bassleri Calvet, 1931 : 84, pi. 2, figs. 33A-C, Cape Verde Islands. Cleidochasma bassleri (Calvet) Harmer, 1957 : I042. pi- 71. n8s- 5» l6> East Indies, synonymy and distribution. Cleidochasma porcellana [sic] (Busk) Soule, 1961 : 18, Gulf of California, 1-45 fath. LECTOTYPE. Chosen here, B.M. 1899.7.1.1726, Madeira, Busk. Coll. LECTOPARATYPES. 1899.7.1.1727, 1728. MATERIAL. " Calypso " Coll. I. Stn. 3, 13° i' N., 17° 24' W., 15^.56, 51-55 m., C54F. Stn. 25, 4° 36' 5" N., i° 31' W., 24^.56, 50 m., CioH. Stn. 26, 4° 37' N., o° 50' W., 24^.56, 100-90 m ., Ci7B (on shell). Stn. 45, o° 25' N., 9° o' E., 8.vi.56, 73 m., C48E on 48 A and D, Discoporella spp., and C55C on 55F, C. brancoense, and 55E, Hippoporidra senegambiensis. Stn. no, off Principe, 7.vii.56, CgL (on shell). 1 Oral spines are apparently absent in C. protrusum and C. rotundorum, but are present in the other species described here. 12 PATRICIA L. COOK Stn. 114, devant B. de Sao Antonio, 8.vii.56, 50 m., Ci5B. Stn. P2, entre Pte. da Mina et I. S. Ana, 4-V.56, 8-10 m., C8H (on shell). Stn. P3, devant Cais de S. Ana, 26.vi.56, ii m., CigA (on shell). Coll. II. Stn. 24, 15° 16' 34" N., 23° 47' 44" W., 55-60 m., 95E, looE. Stn. 26, 15° 16' 30" N., 23° 47' 31" W., 50-65 m., i2iN, I25N. Stn. 65, 15° 37' N., 25° 69' W., 150-120 m., 670. Marche-Marchad Coll. I. Cap Matakong, Guine'e Ise., 4 I, on shell with C. oranense and other species. Achimota Coll. I. Stn. 126, Agassiz trawl No. 3, I2.iv.5i, 16 m., 37G (on shell). Stn. K, on trawl debris, i mile offshore, 2 miles W. of Densu R., 2.iii.49, 4 fath., 361 and 44E (on shell). Coll. II. Chorkor, seine net, i.ii.So, 22N (on worm tubes and shell). British Museum, Madeira, 1899.7.1.1726, 1727, 1728, 2352, 2353, Busk. Coll. Madeira, 1910.11. 1.IH2A & C (with C. contractum), 1115 ; Santa Cruz, 115 fath., 1910.11.1.1116 ; Grenada, 170 fath., 1910.11.1.1113 ; Havana, 1910.11. i.iii4A, Norman Coll. John Adams Bank, off Brazil, 1899.7.1.3347, Busk. Coll. Bahia, 1899.5.1.800, Hincks Coll. Gulf of Mexico, 1932.3.7.87, Canu & Bassler Coll. Gulf of Mexico, 1959.8.20.9, Lagaaij Coll. South pass of Mississippi R., 28° 58' N., 89° 8' W., 1961.11.2.25, and Campeche Bank, 1961.11.2.22, 23, 24, Cheetham Coll. Gorgona, 1929.4.26.126, " St. George " Coll. Galapagos, 1933.12.10.26, Canu & Bassler Coll. Tizard reef, China Sea, 1889.8.21.65. Alexandria, 1963.8.2.37, O'Donoghue Coll. Ceylon, i936.i2.30.34A, Thornely Coll. Philippines, 1931.12.30. 114, 115, Canu & Bassler Coll. Siboga Coll., all specimens listed by Harmer (1957 : 1042). 0-2mm FIG. 4. Variation in shape of orifice and position of avicularia in Cleidochasma porcellanum. A. Campeche Bank, Gulf of Mexico, 1962.11.2.22. B. Ceylon, 1936.12. 30.34A. c. Bay of Biafra, " Calypso " Coll. C55C. D. Madeira, 1899.7.1.1726. E. Tizard Reef, 1889.8.21.65. Museu do Seminario, Funchal, i slide in Waters Collection. Naturhistoriska Riksmuseet, Stockholm. No. 284, W. of Tortugas, 16.1.1869, 42 fath. ? Type of Lepralia cleidostoma. POLYZOA FROM WEST AFRICA 13 DIAGNOSIS. Zoarium encrusting, plurilaminar. Zooecia with marginal pores and porcellanous, smooth or mammillate frontal, 3-4 lateral and distal septula. Orifice with 3 oral spines, cleithridiate, very variable in shape. Operculum with 2 lateral sclerites. Avicularia frequently paired, with a complete bar, situated at or below the level of the sinus, directed outwards and upwards, mandible acute. Ovicell with a small area on the proximal border. Dimensions. Lz 0-45 mm., Iz 0-30 mm., Lo 0-14 mm., lo o-n mm., Lop 0-14 mm., lop o-io mm., Lav. 0-09 mm., Lm 0-05 mm., Lov 0-15 mm., lov 0-19 mm. The appearance of C. porcellanum varies considerably with the nature of the frontal wall, the position, dimensions and proportions of the orifice, the size and position of the avicularia, and the age and degree of calcification of the ovicells. The frontal wall was described by Busk as " uneven, bossed, granular, polished, porcellanous ". Busk's colonies show two types of frontal, although the other zooecial characters are the same. In 1899.7.1.1726 the frontal is hyaline and smooth, in 1899.7.1.1727 and 1728 it is mammillated, and greatly resembles the figure of H. bassleri given by Calvet. Mammillate f rentals are also present in 1961.11.2.25, Gulf of Mexico, 1899.7.1.2352 and 2353, Madeira, and 1889.8.21.65, Tizard Reef, but in the majority of specimens the zooecia are smooth. The variation in the shape of the orifice is very great, and several forms have been recognized. Some orifices are elongated, with a small triangular sinus, others are rounded with a wide, shallow sinus (see Text-figs. 4A-E). However, these extremes may occur within one zoarium (e.g. 1936. 12. 30. 34 A), and intermediate forms exist throughout the material. Furthermore, there appears to be no positive correlation of orifice type with locality or depth. C. porcellanum thus differs from the 3 species of Hippoporidra described below (pp. 28-32), where the different forms of the orifice are exclusive and are correlated with other distinguishing characters. The orifice may be surrounded by one proximal and a pair of lateral raised umbos (1899.7.1.1728 and Achimota Coll. 22N). Hastings (1930 : 722) expressed doubt as to the validity of the position of the orifice with relation to the distal wall as a specific character. The orifice is near the distal edge of the zooecium in Siboga Coll., Japan, 53 fath., Owston Coll. 2X ; 1911.10.1.1112, Madeira ; 1899.5.1.800, Bahia ; and 1931.12.30.114, Philippines. It is appreciably removed from the distal edge in Siboga Coll., Torres Straits, Haddon Coll. 173 ; Achimota Coll. 44E ; 1936.12.30.34, Ceylon and 1933.12.10.26, Galapagos. Waters (1899 : 10, pi. 3, figs. 16, 17) mentioned the large variation in the opercula, which may also be seen in the figures of Smitt (1873, pi. n, figs. 217, 218, 219) and Canu & Bassler (1920, text-fig. H4B-E ; 1929, text-fig. 1326-11). The size and position of the avicularia varies considerably (see Text-fig. 4A-E). In many specimens (or even parts of one colony) they are minute or absent, in others they are larger, with elongated triangular mandibles, which may be curved or flat. Their position varies, from one close to the orifice to one near the borders of the zooecium, and from level with the poster to below the sinus, sometimes within one specimen (e.g. 1936.13. 3O.34A pt., Ceylon). Occasionally some zooecia have a I4 PATRICIA L. COOK large avicularium, with a mandible similar to that of C. late-rale Harmer (1957 : 1044, pi. 71, figs. 9-11), e.g. 1911.10.1.1116 Santa Cruz, and 1961.11.2.22, Campeche Bank. Near the growing edge the ovicells are conspicuous and hyaline, with a distinct, thinner, frontal area, and frequently with paired channels on the proximal lateral border which may mark the former position of the oral spines. At this stage, the ovicell may be faintly striated, with lines of slightly thicker calcification below the surface which reflect light differentially. Later the surface becomes finely tuberculate and finally the ovicell is progressively immersed by heavy secondary calcification. Striated ovicells were figured by Smitt (1873, pi. n, fig. 217) who, however, also showed an older, immersed ovicell (fig. 218). Smitt's specimen has been re-examined and a photograph of one of the young striated ovicells is given on pi. 2, fig. i for comparison with fig. 2. Striated ovicells are present in the following specimens : — Achimota Coll. 44E ; " Calypso " Coll. II Ci25N ; 1911.10.1.1112, Madeira ; 1899.9.7.1.3347, John Adams Bank ; and Siboga Coll., Ceylon, Thornely Coll. 55. Norman (1909 : 305) first equated L. cleidostoma Smitt with L. porcellana Busk, and Calvet (1931 : 84) stated that the avicularia of H. bassleri were similar to those of the other two species and remarked " Ce sont des formes voisines ". Harmer (1957 : 1042) established a very wide range of variation for C. bassleri, but all the Pacific and Indian Ocean records he included had been placed in H . porcellana by Marcus (1938 : 217) who distinguished H. bassleri by its more constricted operculum and regularly mammillate frontal. Neither of these distinctions appears now to be specific. C. bassleri was originally described from the Cape Verde Islands and C. porcellanum from Madeira, and thus both species also have similar eastern- Atlantic type-localities. The conclusion is inevitable that C. porcellanum (Busk) varies in a similar degree to, and indeed includes, C. bassleri (Calvet). The distal septula in C. porcellanum are small and slit-like, close to the base of the wall. The basal lamina of the growing edge is frequently ridged vertically on either side of each pore, forming a chamber. This development may be seen particularly well in " Calypso " Coll. CgsE. Thus the distal " chambered pore " of C. porcellanum resembles both the sunken rosette plates of C. protrusum and the true dietellae of C. contractum. A similar variation in position and relationships of the distal pores is found in Hippoporina (see p. 8). 12. Cleidochasma contractum (Waters) (Text-fig. SA) Lepralia contracta Waters, 1899 : n, pi. 3, figs. 4-6, Madeira. Norman, 1909 : 306, pi. 41, figs. 5, 6, Madeira. Perigastrella contracta (Waters): Hastings, 1930 : 722, pi. n, fig. 60, Galapagos, 10-30 fath., Gorgona. Marcus, 1937 : 9^, pi- 20, figs. 52A, B, Brazil, Santos, 20 m., synonymy and discussion. Hippoporina contracta (Waters) : Osburn, 1940 : 428 ; 1952 : 346, pi. 41, figs. 4-5, Ecuador to Gulf of California. Marcus, 1955 : 298, Espirito Santo, 35 m. Cleidochasma contracta [sic] (Waters) Soule, 1961 : 19, Gulf of California, 1-40 fath. POLYZOA FROM WEST AFRICA 15 The type-material of C. contractum is preserved in the Museo do Seminario, Funchal, Madeira (see Norman, 1909 : 306). MATERIAL. " Calypso " Coll. I. Stn. 4, 12° 55' N., 17° 33' W., i6.v.56, 65-75 m., €36 (on sandy worm-tubes with Rhynchozoon sp.). Stn. 17, 5° N., 5° 28' 30" W., 2i.v.56, 27 m., C56E (on Jullienella foetida} . Stn. no, i° 2' 45" N., 7° 17' 34" E., 7.vii.56, 25-40 m., G. Frere N.E., C32A (on stem of alga). Stn. 114, devant B. de Sao Antonio, 8.vii.56, Ci5C. Stn. T5, a 1' E. Ste Diego, 7.V.56, nunes a pied, C26A. Coll. II. Stn. 14, 14° 53' 43" N., 23° 31' 24" W., 17.^.59, 25-30 m., C68G. Stn. 24, 15° 16' 34" N., 23° 47' 44" W., i8.xi.59, 55-60 m., CgiH, €970, and Stn. 25, 15° 16' 30" N., 23° 47' 31" W., i8.xi.59, 50-65 m., Ci2iD. Marche-Marchad Coll. II. 3 A, no information (on Murex shell). 0'4mm FIG. 5. Cleidochasma, treated with eau de javelle. A. C. contractum, zooecia, with large spathulate, and small avicularia, and ovicell, Ghana, Achimota Coll., C, F. B, c. C. rotundorum, B. Ovicell. c. Zooecium with basal wall of broken ovicell, Madeira, 1911.10.1.1664. Achimota Coll. I (all specimens on /. foetida). Stn. 35, dredge haul No. i, 2i.xii.5o, 37 m., 596. Stn. 45, as above, No. 5, 22.i.5i, 22 m., 62K. Stn. 47, as above, No. i, 4.1.51, 44 m., 1411. Stn. 48, as above, No. 2, 46L. Stn. 72, as above, No. 3, 24.1.51, 38 m., 6il. Stn. 73, as above, No. 4, 41 m., 58G. Stn. no, Agassiz trawl, No. i, 4.iv.5i, 40 m., 48N. Stn. in, as above, No. 2, 43 m., 490. Stn. 112, as above, No. 3, 6oL. Stn. 131, as above, No. i, 2.v.5i, 37 m., 4iM, 43M. Stn. 132, as above, No. 2, 44 m., 4oF and 42L. Stn. 133, as above, No. 3, 51 m., 45N. Specimen C, F, Achimota. Coll. II. 23Q, Seine net, Chorkor shore, i2.v.5o. British Museum, Madeira, i9ii.io.i.ni2A (with C. porcellanum] , 1156 and 1160 (on clinker), and Havana, 80 fath., 1911.10.1.1114, Norman Coll. St. Vincent, 16 PATRICIA L. COOK Kingstown Harbour, 1931.5.2.5 pt., Totton Coll. Galapagos, 1929.4.26.189, " St. George " Coll. Campeche Bank, 24° 4' N., 91° 50' W., 26 fath., 1961.11.2.16, and S. Pass of Mississippi R. 28° 4' N., 58' N., 89° 9' W., 1961.11.2.15, Cheetham Coll. Mauritius, 1934.10.6.24. Museo do Seminario, Funchal, i slide, Type of L. contracta Waters. DIAGNOSIS. Zoarium encrusting, plurilaminar. Zooecia with marginal pores only, dietellae present. Orifice with 4-8 oral spines and a strongly beaded rim, with a pair of strong denticles marking a rounded sinus. Avicularia with a bar, very variable in size and occurrence, occasionally absent from areas of the colony. Usually one or two rounded avicularia, or one large sub-oral or lateral avicularium, with a spathulate mandible. Ovicells hyperstomial with a large semi-circular area on the front. Dimensions. Lz 0-41 mm., Iz 0*33 mm., Lo o-n mm., lo o-io mm., Lm 0-03-0-23 mm., Lov 0-18 mm., lov 0-20 mm. The type-species of Perigastrella Canu & Bassler (1917 : 68) is " Lepralia labiata Boeck, 1861 ". Boeck (1862 : 49-50) did not describe this species, which was introduced by Smitt (1868 : 27, 175). Osburn (1940 : 428) gave reasons for separat- ing L. contracta from Perigastrella, and placed the species in Hippoporina (Canu & Bassler not Neviani, see p. 4). Harmer (1957 : 823 and 1025) emphasized that L. labiata and L. contracta were not congeneric. The variability in appearance of C. contractum is principally dependent upon the development of the avicularia. At the growing edge of a colony the zooecia frequently have only a small sub-oral umbo, and avicularia are absent, as Waters noted (1899 : n). The lateral peristome is sometimes slightly raised laterally. Where conditions of growth are crowded, or where the colony is plurilaminar and secondary calcification has taken place, the orifices are immersed and there is a large number of avicularia, some of which may obscure the orifice altogether. The specimens from Havana (1911.10.1.1114) and Mauritius (1934.10.6.24) have a large avicularium directed distally alongside the orifice in the same manner as that of the specimen from Madeira figured by Waters (1899, pi. 3, fig. 4). Examination of the figured slide shows that this type of avicularium is associated only with fertile zooecia. Specimens from the Achimota Coll. have a large avicularium either lying across the proximal edge of the orifice or pointed distal-laterally, whereas many of those from the Cape Verde Islands (" Calypso " Coll. II) have only small rounded avicularia and a sub-oral umbo. However variable the superficial appearance of the colonies, C. contractum may always be recognized by the beading round the distal edge of the orifice, and by the large semicircular area on the front of the ovicell. The occurrence of distinct dietellae in C. contractum was noted by Hastings (1930 : 739, expl. of pi. n, fig. 60). The communications in Cleidochasma are very variable ; in C. protrusum (Thornely), (Ceylon, 1906.12.3.11 and 1936.12.30.63, Thornely Coll.), and C. areolatum (Canu & Bassler), (Philippines, 1931.12.30.106, 107, Canu & Bassler Coll.), they consist distally of a pair of large multiporous rosette plates, surrounded by ridges of calcification, similar to those of C. oranense and POLYZOA FROM WEST AFRICA 17 C. brancoense (see p. 19) ; but in C. porcellanum they approach a chambered structure like that of C. contractum (see p. 14). The number of oral spines in C. contractum is variable. Four are present in most of the material from west Africa and the Gulf of Mexico. The zooecia of the Galapagos, Kingstown harbour and Mauritius specimens, however, have 5-8 spines ; while Marcus (1937) figured 7 spines in Brazilian, and Norman (1909) 6-7 in Madeiran, material. Whereas colonies from the Cape Verde Islands, Senegal and the Bay of Biafra encrust stones and shells, those from the coast of Ghana almost exclusively encrust the Foraminiferan Jullienella foetida (see p. 8). Distribution. Madeira, Cape Verde Islands, west Africa, Brazil, West Indies, Tortugas, Florida, Long Island Sound, Woods Hole, Panama, California, Colombia, Galapagos. 13. Cleidochasma oranense (Waters) (PI. 2, fig. 3, PI. 3, fig. 2, Text-fig. 6A-B) Lepralia oranensis Waters, igiSb : 101, pi. 12, figs. 11-13, Oran, "zone coralligene " 54 fath.; Petit Tahou, Liberia. Canu & Bassler, ig28b : 33, pi. 3, figs. 8-10, Cap Blanc, 20-30 m. Gemelliporella oranensis (Waters) Barroso, 1925 : 177, text-figs, i, 2, Oran, 62-65 " brazas " (= fathoms). LECTOTYPE, chosen here, Manchester Museum, slide from Oran, 54 fath., Waters Collection. MATERIAL (specimens with ovicells* marked with an asterisk). " Calypso " Coll. I. Stn. i, 21° 05' N, 17° 14' W, io.v.56, 43-45 m., C5iD* (on barnacle), and C52C (on shells and Schizzamina sp., a Foraminiferan). Marche-Marchad Coll. I. Konakrey, Guine'e Ise., iD*, 21*. Sud de Goree, 24.11.53, 40-41 m., nE. Cap Matakong, 4A (on Pecten shell, encrusting base and erect branches). S.W. Madeleines, i5.ix.53, 48 m., 25A*, and 9.i.54, 45-46 m., 27C*. Sud de presque Tile de Cap Vert, i8.ii.54, 95 m-> 33^. Coll. II. S.E. des Madeleines, i5.ix.53, 48 m., 2A. Baie de Gore"e, 50-100 m., 7G. Large de Gor6e, 5.vii.55, 50 m., 8D, 27E. S.W. Cap Manuel, 20.ii.56, 50 m., I3A. Au large de Saloum, 9. in. 55, 50 m., I7C. S. Baie de Goree, i8.ii.54, 95 m., 3oA. S.W. Madeleines, 9-i.54, 47.5 m., 316* (unilaminar erect expansions), and i5.ix.53, 48 m., 4oC*. Coll. III. Either S.W. Madeleines, i5.ix.53, 48 m., or S. de Gore"e, i3.xi.53, 34-37 m., i6A. Large de Gore"e, 50 m., Stn. 55.vii.5A, i8C. S.W. Cap Manuel 50 m., Stn. 56.iii.2oD, 26A ; and Stn. 56.ii.2oB, 256. " Gerard Frdca " Dragage i, 18.11.54, 23D ; Dr. 5, 276. No information, 2iC, 24C. Achimota Coll. I. Stn. 47, Dredge haul No. i, 4.1.57, 44 m., I4E. Stn. 62, as above, No. 2, i8.i.5i, 30 m., 8716. Stn. 117, Agassiz trawl No. 2, 5-iv.5i, 64 m., 32D, & Coll. II, I3A. Stn. 132, as above, 2.v.5i, 44 m., 5oA. Coll. II. 23R, Chorkor shore, seine net, I2.V.50 (worn). British Museum, Cap Blanco, west Africa, 11-35 fatn-> i92I-5-23-7 and 1927.10.30.11, Metzelaar Coll. west Africa, 1959.2.16.1. Manchester Museum. Waters Coll. 5 slides from Oran, the lectotype, labelled i8 PATRICIA L. COOK " material sent by Canu ", is probably the specimen figured by Waters on pi. 12, fig. ii. 2 slides from Jullien Coll., Petit Tahou, Liberia. C. oranense is not present in the "Calypso " Collection II, from the Cape Verde Islands (see p. 20). DIAGNOSIS. Zoarium with an encrusting base and erect branches composed of superposed layers of zooecia. Zooecia with large marginal pores and occasional small frontal pores. Orifice with 4 oral spines, cleithridiate, with large condyles, proximal edge of anter straight. Operculum globose distally with a pair of lateral sclerites. 2 distal and 1-2 lateral rosette plates. Avicularia large, interzooecial, numerous, directed distally, with an acute rostrum and complete bar. Mandibles long and curved. Ovicells hyperstomial, but immersed, eventually almost hidden by secondary calcification, with a small area on the proximal frontal border. Polypide with 15 tentacles (Barroso). Dimensions. Lz 0-50 mm., Iz 0-30 mm., Lo 0-19 mm., lo 0-13 mm., Lav 0-40 mm., Lm 0-35 mm., Lov 0-20 mm., lov 0-25 mm. The zoarium of C. oranense has an encrusting base, from which grow erect, 0-2 mm FIG. 6. Opercula and mandibles in Cleidochasma. A, B. C. oranense. A. Operculum. B. Mandible, Ghana, Achimota Coll., 32D. c. D. C. brancoense. c. Operculum, Bay of Biafra, " Calypso " Coll., C55F. D. Mandible, Cape Verde Islands, " Calypso " Coll., POLYZOA FROM WEST AFRICA 19 irregular branches which occasionally anastomose. The colony is frequently large (maximum height observed 60 mm.), and the branches are hollow, or in some cases grow round an algal or hydroid core. The structure of each branch is plurilaminar, secondary layers being budded from the base of the branches ; these usually grow in the same direction as the primary layer (cf. C. brancoense below). The operculum has been figured by Waters and Barroso ; it is semi-globose distally, and usually golden-brown in colour. Barroso also figured the pair of large distal rosette plates. In C. oranense the distal wall of each zooecium is concave, and thus the rosette plates are closer together than those of C. brancoense (see below). The avicularian chambers communicate with the neighbouring zooecia by lateral septula, but do not reach the basal wall of the zoarium. Each zooecium usually has an avicularium on one distal-lateral wall, but occasionally it may be absent ; the avicularia seem to arise above the marginal pores, after the zooecial walls have been laid down at the growing edge. Ovicells do not appear to have been described before in C. oranense. When young, they are large, wide and prominent, with a small area on the proximal border. They are rapidly immersed by secondary calcification, and even in slightly older parts of the colony they are inconspicuous, the obscured distal part of the operculum being the only sign of the presence of a fertile orifice, which does not otherwise differ in any way from those of other zooecia (see pi. 2 fig. 3). C. oranense is known only from the north and west African region, from Oran to the coast of Ghana. It has not been reported from Madeira or the Cape Verde Islands, and is absent from the Bay of Biafra (cf. C. brancoense below). 14. Cleidochasma brancoense (Calvet) (PI. 2, fig. 4, PI. 3, fig. i, Text-fig. 6c-D) Lepralia brancoensis Calvet, igoOa : 159 ; 1907 : 410, pi. 27, figs. 6-9, Cape Verde Islands, ilot Branco, 110-180 m. Waters, igiSa : 4 (listed only). MATERIAL (specimens with ovicells* marked with an asterisk). " Calypso " Coll. I. Stn. 45, o° 25' N, 9° i' E, 8.vi.56, 73 m., C55F* (with C. porcellanum C55Q. Coll. II. Stn. 24, 15° 16' 34" N, 23° 47' 44" W, i8.xi.5Q, 55-60 m., Cgil, Cg4A, C97A*. Stn. 26, 15° 16' 30" N, 23° 47' 31" W, i8.xi.5g, 50-65 m., (ile Sao Tiago), C65L*, Ci27F*. Stn. 31, no details, C66R. C. brancoense is not present in the Marche-Marched or Achimota Collections. DIAGNOSIS. Zoarium encrusting with erect, plurilaminar, irregular anastomosing solid branches formed by spiral growth. Zooecia with large marginal and occasional frontal pores. Orifice with 2 oral spines, cleithridiate, with a pair of large condyles forming a subtri angular sinus. Operculum with a pair of lateral sclerites. 2 distal and 1-2 lateral rosette plates. A pair of small rounded avicularia with a complete bar and rounded or triangular mandibles, level with the distal border of the orifice, usually directed inwards and slightly proximally, frequently present, also on the fertile zooecia. Occasional large interzooecial avicularia, with a complete bar and elongated spathulate mandibles present. Ovicells hyperstomial but immersed, 20 PATRICIA L. COOK becoming obscured by secondary calcification, with a small area on the proximal frontal border. Dimensions. Lz 0-60 mm., Iz 0-45 m., Lo 0-26 mm., lo 0-16 mm., Lop 0-25 mm., lop 0-15 mm., La 0-15 mm., La (interzooecial) 0-50 mm., Lm (interzooecial) 0-40 mm., Lov. 0-25 mm., lov. 0-37 mm. The zoarium of C. brancoense is very similar in appearance to that of C. oranense, but differs in the method of growth of the erect branches, which are solid. The secondary layers may arise randomly on the primary layer and develop in any direction from these foci, some branches when broken show a spiral structure similar to that of Hippoporidra senegambiensis (Carter) (see p. 25). The distal wall of each zooecium is convex, with a strong central ridge ; thus the pair of rosette plates are widely separated, unlike those of C. oranense. The large interzooecial avicularia are sporadic in occurrence, often being absent over areas of the colony. Calvet's material had no ovicells ; those in the " Calypso " specimens are larger and wider than those of C. oranense, and are even more rapidly obscured by secondary calcification ; only a few show a small area on the proximal border. C. oranense and C. brancoense are closely related species ; the distribution of C. brancoense is confined to two widely separated areas, the Cape Verde Islands and the Bay of Biafra, in neither of which C. oranense has been found. 15. Cleidochasma rotundorum (Norman) (PI. i, fig. 2, Text-fig. 5B, c) Celleporajanthina (Smitt) Waters, 1899 : 14, pi. 3, figs. 1-3, Madeira (not Hippoporidra janthina, see p. 27). Canu & Bassler, 1920 : 615, text-figs. I85A-E. Cellepora rotundora Norman, 1909 : 311, pi. 42, figs. 8, 9, Madeira, 70 fath. Hippotrema rotundora (Norman) Canu & Bassler, ig28a : 141. LECTOTYPE, chosen here, specimen in the Museo do Seminario, Funchal, figured here on pi. i, fig. 2. LECTOPARATYPES, remaining specimens on the same slide. MATERIAL. British Museum, Madeira, 70 fath., 1911.10.1.1664, Norman Coll. C. rotundorum is not present in the "Calypso " Collections, the Marche-Marchad Collections or the Achimota Collections. DIAGNOSIS. Zoarium encrusting, plurilaminar or discoidal. Zooecia with marginal pores and 2-3 rows of frontal pores, the area proximal to the orifice non- porous. 1-2 lateral and 2 distal multiporous rosette plates present. Orifice rounded, with a pair of distinct denticles forming a shallow, rounded sinus ; peristome raised laterally and a small proximal umbo sometimes present. Avicularia adventitious, frequently paired, beside the sinus, directed distally, with a complete bar and triangular mandible. Ovicells prominent, globose, finely and regularly tuberculate, with a very large triangular membranous frontal area. Dimensions. Lz 0-45 mm., Iz 0-30 mm., Lo 0-13 mm., lo 0-13 mm., Lav 0-09 mm., Lov 0-20 mm., lov 0-27 mm. POLYZOA FROM WEST AFRICA 21 C. rotundorum somewhat resembles C. mirabile Harmer (1957 : 1045, pi. 71, figs. 15, 17, 18, text-fig. 113 ; 6ooD, Borneo Bank, 59 m.), which also has globose, tuberculate ovicells with large frontal areas, and similar avicularia. C. rotundorum differs in the less elongated shape of its orifice and in the absence of oral spines. The zoarium of C. mirabile is selenariiform, and the basal surface has characteristic deep depressions (see Harmer, 1957 : 1045). C. rotundorum may perhaps assume a selenariiform habit, but the lectotype and the specimens in the Norman Collection are all encrusting. Some colonies in both collections grow over small lamellibranch shells, almost covering both surfaces and thus producing conical, free colonies. These may be those described as " disk-shaped " by Waters. The remainder of the specimens encrust calcareous tubes of the serpulid worm, Ditrupa arietina1 (O. F. Muller). Waters's figure of the ovicells of C. rotundorum shows the extensive frontal areas more clearly than that of Norman, where they are drawn in foreshortened view. Only one ovicell in Norman's Madeiran specimens has the membranous covering in position, in all the others it is no longer present, and in many the ovicell is broken, only the basal wall remaining (see Text-fig. 50). The frontal pores are more extensive in this species than in any other Cleidochasma, but are not found in the central sub-oral area, which is tuberculate and often umbonate. Waters mentioned the umbo, but both he and Norman illustrated the frontal as uniformly porous. Canu & Bassler (1927 : 21, 31 and i928a : 141) erected the genus Hippotrema for forms with tremopores, including L. rotundora, and, as type-species, chose Lepralia edax forma janthina Smitt (see p. 27), which was also originally figured with a porous frontal. Unfortunately Canu & Bassler did not figure specimens of Hippotrema janthina, but referred only to their earlier text-figures of the " Cellepora janthina group of Waters " (1920 : 615, text-figs. 185 A-E). These figures are copies of Water's (i85A-c) and Norman's (i85D-E) illustrations of C. rotundorum, not of H. janthina (Smitt). Osburn (1952 : 354) pointed out that the frontal of H. janthina was not a tremocyst, but that the rows of sub-marginal pores were " carried up " in the calcification to produce an appearance of tremopores. Examples of this type of calcification may also be seen in Cleidochasma spiculiferum (Canu & Bassler), which was also originally referred to Hippotrema (see below), and in the " cortical zooecia " of Hippoporidra (see p. 22). 16. Cleidochasma granulosum (Canu & Bassler) and C. spiculiferum (Canu & Bassler) Hippoporidra and Cleidochasma have many characters in common. Hippoporidra is here restricted to species with " cortical zooecia " and a special type of interzooecial avicularium, which encrust gastropod shells (see also Lagaaij, 1952 : 147). Osburn (1952 : 356-7) included Hippoporidra granulosa Canu & Bassler (ig3oa : 43, pi. 8, figs, i, 2 ; Galapagos, 40 fath.) and Hippotrema (?) spiculifera Canu & Bassler (i93oa : 43, pi. 8, figs. 3-5, Galapagos, 40 fath.) in Hippoporidra. Both species have 6 oral spines, a character absent in Hippoporidra, and, although 1 Notes on the identification of this worm and of another associated Polyzoan, Bugula ditrupae Busk, from Madeira, are given by Ryland (1960 : 94-96). 22 PATRICIA L. COOK H . granulosa encrusts shell, neither have any association with Gastropod or hermit- crab, nor do they possess cortical zooecia or interzooecial avicularia. Thus neither can be referred to Hippoporidra as denned here. Specimens of H. granulosa (Galapagos, 40 fath., 1933.12.10.29, " part of type material ", and 1933.12.10.10, Canu & Bassler Coll.) have a pair of small distal rosette plates, an elongate, cleithridiate orifice, with an operculum (like that figured by Osburn, 1952, pi. 42, fig. 14) which greatly resembles that of Cleidochasma porcellamim. A specimen of Hippotrema (?) spiculifera (Galapagos, 40 fath., 1933.12.10.6, Canu & Bassler Coll.) has a pair of distal rosette plates and an orifice very like that of C. rotundorum. The numerous ovicells (which were not found by Canu & Bassler) also resemble those of C. rotundorum and C. mirabile, being large, globose, finely tuberculate, and having a large triangular membranous frontal area. The avicularia are adventitious, and the long curved mandible is directed proximally. Both H. granulosa and H. spiculifera are here referred to Cleidochasma. 17. HIPPOPORIDRA Canu & Bassler Hippoporidra Canu & Bassler, 1927 : 21, 31 ; 1929 : 418, text-fig. 163. Lagaaij, 1952 : 147. TYPE SPECIES, Cellepora edax Busk, 1859 : 59, pi. 9, fig. 6 ; pi. 22, fig. 3, Pliocene, Suffolk. DIAGNOSIS. Zoarium normally encrusting Gastropod shells, frequently (perhaps exclusively) those inhabited by hermit-crabs ; plurilaminar, nodular, branched, massive. Autozooecia with marginal pores and a few frontal pores, and with dietellae. Orifice hippoporine, variable, with a pair of strong denticles. Oral spines absent. Nodules and branches formed by cortical zooecia, with 3-4 rows of frontal pores and reduced orifices. Avicularia small, frontal, and small and large interzooecial, with triangular and spathulate mandibles. Ovicells hyperstomial, with a frontal area, not closed by the operculum. Cortical zooecia. Large zooecia (Lz twice that of autozooecia), recumbent in the primary layer, erect in subsequent layers. Frontal apparently pierced by tremopores but in fact, with 2-3 rows of " carried up " pores (see above). Orifices reduced, surrounded by large peristomial tubercles. Polypides not found, zooecial cavity lined by a thick yellow membrane and filled with fibrous and granular material. These cortical zooecia form the base and core of the nodules and branches. Interzooecial avicularia. Avicularia in which the sub-rostral chamber is very large, with dietellae, and a frontal wall with both marginal and frontal pores. In the primary layer a small avicularium is centrally placed and variously orientated. In the later, superposed layers, where these avicularia form part of nodules or branches, the erect nature of the sub-rostral chamber reduces the extent of the frontal, and the avicularium may appear to be adventitious. The largest avicularia are only found as part of the nodules or branches. The sub-rostral chamber is large and very convex, and the mandibles are very variable in shape. The muscles fill the sub- rostral chamber. Each species of Hippoporidra displays a wide individual range of variability, but some characters seem to be fairly constantly correlated, and it is upon these that POLYZOA FROM WEST AFRICA 23 specific status is based in the descriptions below. It must however be stressed, that until the breeding and development of living colonies have been observed, together with a full analysis of the association with Gastropod and/or hermit-crab, the separation of the records and specimens below into species, is of necessity arbitrary, and may not reflect actual relationships. 18. NOTES ON THE ASSOCIATION OF GASTROPODS AND PAGURIDS (HERMIT-CRABS) WITH ENCRUSTING ANIMALS A number of animal species are known to live upon the shells of Gastropods, whether these are inhabited by the Mollusc or by hermit-crabs. Stechow (1921 : 29-31) described two encrusting coelenterates, Janaria mirabilis and Hydrocorella africana, with branched growths. Specimens of H. africana examined (Coelenterate section register, 1957.4.26.22, False Bay, S. Africa), measure 13 mm. across, but have no branches, and are botryoidal. A very fine lamina of the Coelenterate grows out around the orifice of the shell, as in Hippoporidra (see p. 25). Growth of a similar lamina was also noted by von Marten (1877 : 183) in an un-named Polyzoan associated with a hermit-crab from South Africa (see p. 32). Hesse & Doflein (1943 : 533, text-fig. 596), described a spiny Coelenterate, Hydractina sodalis, encrusting a shell inhabited by Eupagurus sp. H. sodalis is similar to Palythoa senegambiensis Carter (1882 : 418, pi. 16, fig. 2A-c, 3A, B) which has 4 erect arms. Other associations of Coelenterates with hermit-crabs were noted by von Marten (1876 : 20). Botryoidal colonies of the Ctenostomatous Polyzoan Alcyonidium nodosum were described by O'Donoghue & de Watteville (1944 : 428, pi. 16, figs. 17, 18) and O'Donoghue (1957 : 92), encrusting a Gastropod, Comminella papyracea from S. Africa. Specimens of A. nodosum (1963.3.20.12, Port Nolloth and Lamberts Bay, S. Africa, O'Donoghue Coll.) show that the large zooecia forming the nodules some- what resemble the cortical zooecia of Hippoporidra (see p. 22). They are about twice the length of the autozooecia, and appear to be filled with granular material. Plurilaminar, erect irregular branches are present in A. polypylum Marcus (1941 : 63, pi. 8, fig. 27) encrusting Murex from Brazil, but the description does not mention groups of enlarged zooecia. The membraniporan Cheilostome, Antropora tincta (Hastings) produces pluri- laminar erect branches (up to 50 mm. in length) when encrusting gastropod shells inhabited by hermit-crabs (see Osburn, 1950 : 54, pi. 4, fig. 7 ; pi. 29, figs. 7-8, S. California to Galapagos). The original colonies described by Hastings (1930 : 708) encrust shell fragments and stones, and although plurilaminar, have no erect outgrowths. In Conopeum commensale Kirkpatrick and Metzelaar (1922 : 983, pi. i, fig. 1-9, pi. 2, fig. 14) from west Africa, which is associated with hermit-crabs, the zoarium is plurilaminar without either nodules or erect branches (Cook MS). Several Polyzoan species, all with branched zoaria and all associated with hermit- crabs, have been recorded from west Africa. Unfortunately their zooecial characters were not described, and although it is very probable that they are synonymous with Hippoporidra senegambiensis, H. picardi or H. littoralis this cannot be established beyond doubt. The records are as follows : — 24 PATRICIA L. COOK (a) Keruniella valdiviae and Cellomma keruniformis , from the mouth of the Congo, associated with Eupagurus pollicaris Say var. alcocki Balss. ; see Stechow, 1921 : 31 and Doflein, 1914 : 350, text-fig. 3020. Fig. 3020 shows Eupagurus varians Bened. with a similar branched growth covering the shell inhabited by the hermit-crab. (b) Eschara perosa Kirchenpauer MS, Studer (1889 : 28), on a Mollusc shell. The colony had cylindrical branches and was from a sandy substrate off west Africa (4° 40' N, 9° 10' W, with Cupuladria sp.). Studer (p. 54) also listed Eschara sp. from S. Africa (33° 59' S, 17° 52' E, 91-5 m.) growing on Fusus mandarinus inhabited by Eupagurus ungulatus Studer. This specimen was figured (pi. 23, fig. SA) with two stout branches, at 180° to one another, extending laterally from the orifice of the shell. It is very similar in appearance to the figure of " Schizopodrella " sp., associated with Pagurus corallinus given by Schmitt (1931, text-fig. 43), which had in addition a median branch (see p. 25). Roger & Buge (1948 : 461-470) examined several species of fossil and Recent Polyzoa encrusting shells, by means of X-ray photography. They concluded that in the fossil specimens the association was between Polyzoan and Gastropod, not Polyzoan and hermit-crab. However, Buge (1957 : 320-323), having examined a large number of Recent specimens from west Africa using the same methods, concluded that it was possible that the association was between Polyzoan and Pagurid. He also considered whether the association might be primarily with the Gastropod, and secondarily with the Pagurid, but inclined to the former theory. The lectotype and other specimens of H. edax have been examined by means of X-rays. No positive evidence of association with hermit-crabs has been found in the fossil specimens, but of 200 Recent specimens of H . edax and H. senegambiensis examined 120 were inhabited by hermit-crabs. In young colonies the proportion is even higher ; 41 out of 50 specimens of H . senegambiensis (Achimota Coll. 87AI) had hermit-crabs present. In no case has a Mollusc been found, all the remaining shells being apparently empty. A few shells from west Africa examined from the Mollusca Section of the British Museum had no Polyzoan encrustation and were inhabited by the original Gastropod. Thus it appears that the association in Recent species is undoubtedly between Polyzoan and Pagurid. In the lists of material below, those shells which can be seen to be inhabited by a hermit-crab are marked with an asterisk, thus : — nA*. 19. DEVELOPMENT AND STRUCTURE OF THE ZOARIUM IN HIPPOPORIDRA Some of the larger shells indicate that there may be initial settlement of several larvae and that the final zoarium is formed by the most successful colony ; but, in the great majority of shells, the zoarium appears to be the result of the settlement of a single larva. The following specimens, all of H. senegambiensis, are particularly instructive : — Marche-Marched Coll. I, 6D, III, gA ; Achimota Coll. I, 44!, 87AI, II, gA. In one case (44!) the ancestrula is visible ; the orifice is rounded, with very little sinus. In all these specimens the colony starts from a point near the orifice of the shell, and, where a hermit-crab is present, above the flexure of the larger chela. POLYZOA FROM WEST AFRICA 25 The colony first develops as a unilaminar sheet over the shell, producing cortical zooecia and interzooecial avicularia. Nodules of plurilaminar growth, with cortical zooecia at their centre, and groups of ovicells in the hollows between them, are evident, often before the shell is completely covered by the primary layer of zooecia. Where the shell is of the long-spired Turritella type, the first branch is developed early, also before the shell is completely covered ; it is almost invariably at an angle approaching 180° to the long axis of the shell, so that it grows in the direction away from the spire. Subsequent branches tend to develop in a median position in a plane at 90° to the spire of the shell. On short-spired shells or those where several barnacles have also settled, the growth is nearly always nodular and branches are rarely formed, and there seems to be some correlation between shell-type and zoarial form. The erect branches are themselves nodular, with groups of cortical zooecia, and secondary branching occurs in large zoaria. The cortical zooecia apparently form the core of each branch, each group budding further individuals, while the surrounding autozooecia grow up and cover the protuberance thus formed. Sections of branches show the large central cortical zooecia, and the small zooecia of the covering layers, which appear to develop spirally in some zoaria (cf. Cleidochasma brancoense, p. 20). Some very large colonies are almost globular, and appear to be the result of plurilaminar growth between the branches, which becomes so thick that they are obscured. The zoarium is usually extended as a lamina surrounding the shell-orifice and forming a cowl over the carapace of the hermit-crab. This is particularly well- developed in some specimens of H. picardi (Marche-Marchad Coll. Ill, nA) where the lamina has also developed ventrally to the Pagurid and formed a tube. In one of these zoaria the tube has become so extended that its orifice is apparently too small for the Pagurid to pass through. The pattern of zooecia which is seen on the basal side of the free lamina has led some authors to presume that it marks the position of the original shell orifice and that the shell itself has been eroded by the Polyzoan (see Busk, 1859 : 59» and Carter, 1882 : 417). Both X-ray photographs and broken specimens show that the shell is not eroded and is present even in large colonies, where there is considerable plurilaminar growth. For example, H. senegam- biensis from Loanda, 1896.7.30.1, is a broken zoarium in which the shell may be seen embedded in from 30-50 layers of zooecia. The pigmentation of the zoaria of H. janthina was noted by Smitt (1873 : 64) and Osburn (1952 : 355). Carter (1882 : 417) described the pigmented areas of H. senegambiensis in detail, but examination of the plentiful material now available has shown that their distribution is not constant. Carter found that the cortical zooecia were pigmented and the surrounding depressed fertile areas were not. Zoaria with this type of pigmentation, and also with the reverse arrangement are present in H. senegambiensis from the Marche-Marchad Coll. Ill, I4A. A section through a branch shows that the areas of pigment alternate in position throughout the growth of the colony. The very large interzooecial avicularia are usually present at the base of the nodules and branches, surrounding the groups of ovicells, which are always found 26 PATRICIA L. COOK in the hollows between the cortical zooecia, as noted by Carter (1882 : 417). The frontal area is completely membranous in young ovicells of all species, but it rapidly becomes calcined, resulting in a semicircular plate with a small distal, membranous pore. 20. Hippoporidra edax (Busk) (PI. 3, figs. 5-7) Cellepora edax Busk, 1859 : 59, pi. 9, fig. 6, pi. 22, fig. 3, Pliocene (Gedgravian) , Suffolk ; 1861 : 154, pi. 34, figs. 3, 3A, Coast of Devon. Duvergier, 1924 : 46, pi. 6, figs. 5-10, Helvetian, Miocene, France. Lepralia edax (Busk) Moore, 1937 : 2O2> on shell containing Eupagurus cuanensis, Fleshwick Bay and N. of Kitterland, 18 fath., S.W. coast of the Isle of Man. Hippoporidra edax (Busk) Lagaaij, 1952 : 147, pi. 15, fig. 13, pi. 16, fig. 6, Pliocene, Netherlands, synonymy. Buge, 1957 : 32°- pi- IJ> n§- 2> pi- I2> n§s- 3~6» Redonian, western France. LECTOTYPE. B. 1620, B.M. Palaeontology Department. MATERIAL. British Museum, Palaeontology Dept. B. 1620 and D. 60463, Coralline Crag, Pliocene, Britain. Zoological Dept., Coast of Devon, 1899.7.1.1410, Busk Coll., fig'd Busk, 1861. Guernsey, 1899.7.1.141^, 1963.4.16.1, Busk Coll. 1887.7.29.49 (chitinous parts). 1911.10.1.1143, Norman Coll. Plymouth, 1899.5.1.1517, Hincks Coll. Scilly, 35 fath., 1963.3.30.273, Harmer Coll. Port Erin, Isle of Man, 1963.12.30.1*, Eggleston Coll. Sedgwick Museum, Cambridge. C5 1007-8, Coralline Crag, Orford Castle, Suffolk. C5IOO9-I4, Orford. C5ioi5-26, Hall pit, Sudbourne, Suffolk. C35829-33, Ged- grave, Suffolk. C5I395, Sudbourne, Suffolk. H. edax is not present in collections from west Africa. Diagnosis. Zoarium encrusting Gastropod shells plurilaminar, nodular. Zooecia with marginal and a few frontal pores. Orifice with a fairly wide, rounded sinus. Cortical zooecia with more numerous frontal pores and reduced orifice. Small acuminate and large acute interzooecial avicularia present, with doubly constricted bar. Ovicell with frontal area. Dimensions. Lz 0-33 mm., Lz (cortical) 0-60 mm., Iz 0-30 mm., Lo o-io mm., lo 0-08 mm., Lo (cortical) 0-02 mm., lo (cortical) 0-02 mm., Lav o-io mm., Lov 0-13 mm., lov 0-14 mm. Lagaaij gave his reasons (1952 : 140, 148) for regarding Cellepora parasitica Michelin, 1847 as a synonym of H. edax, rather than of C. parasitica Busk, 1859. The dome-shaped zoarium described as Lepralia edax by Waters (1885 : 297) from the River Murray Cliffs, Australia, may have belonged to Conescharellina. Duvergier (1924 : 46) first noted that the zooecia composing the zoaria of C. edax were of two kinds, and he described the irregularly porous frontal of the cortical zooecia. The Recent specimens from S.W. Britain and the Channel Islands have been included here with the Pliocene material which they resemble more than the closely related H. janthina and H . senegambiensis. The orifices of H. edax have a smaller, deeper sinus than those of H. janthina, and the interzooecial avicularia are all acuminate, not spathulate, as in H. senegambiensis. The zoaria of H. edax are also apparently POLYZOA FROM WEST AFRICA 27 always nodular, without erect branches, although this may be more a function of the type of shell available than a specific character (cf. H. littoralis p. 28). The zoarium of the lectotype is nodular, measuring 15 mm. x 20 mm. The majority of the zooecia are small with few or no frontal pores. The cortical zooecia are considerably larger with small pits indicating frontal pores. Few avicularia are present, but the same types are present in larger numbers in the specimens from Guernsey and Devon. Ovicells are present in the lectotype (cf. Lagaaij, 1952 : 148) ; they are almost hidden on the face of the shell nearly concealed by the slide, but may be seen by tilting the mount. These ovicells are incomplete, the frontal wall no longer being present, but they so greatly resemble broken ovicells of Recent specimens that it is reasonable to assume that they, too, possessed a similar delicate frontal area. The specimens in the Sedgwick Museum also have broken ovicells, and zoaria from the series 035829-33 show the frontal pores of the cortical zooecia particularly clearly. FIG. 7. Hippoporidra, treated with eau de javelle. A. H. littoralis, zooecium and cortical zooecium, Ghana, Achimota Coll., 76. B, c. H. senegambiensis. B. Zooecium and inter- zooecial avicularium. c. Ovicell, Ghana, Achimota Coll., 14!. D. H. picardi, zooecium and interzooecial avicularium, Senegal, Marche-Marchad Coll. I, 33H. 21. Hippoporidra janthina (Smitt) Lepralia edax forma, janthina Smitt, 1873 : 64, pi. n, figs. 224, 225, off Florida. Lepralia maculata Ulrich & Bassler, 1904 : 423, pi. 115, figs. 8a, b, and 9, pi. 118, fig. 7, Calvert Formation, Plum Point, 3 miles S. of Chesapeake Beach, Miocene. Cellepora minuta Canu & Bassler, 1923 : 182, pi. 25, figs. 10-13, Miocene, N. Carolina, and Pliocene, S. Carolina. Hippotrema janthina Canu & Bassler, ig28a : 141 (not Cellepora janthina Waters = Cleidochasma rotundorum see p. 20). Hippoporidra janthina (Smitt) Osburn, 1952 : 355, pi. 45, figs. 13-15, Gulf of California to Lower California, 2-60 fath. Hippoporidra edax (Busk) : Soule & Duff, 1957 : IJ3> Pleistocene, California. MATERIAL. British Museum, Gulf of Mexico, 28° 58' N., 89° 9' W., 1961.11.2.14, Cheetham Coll. H. janthina is not present in the collections from west Africa. 28 PATRICIA L. COOK DIAGNOSIS. Zoarium encrusting Gastropod shells, nodular or branched. Zooecia slightly larger than those of H. edax, with marginal pores and 2 rows of frontal pores. Orifice with very shallow wide sinus. Small avicularia frequent, larger acuminate avicularia rare. Ovicell wide, with rounded frontal area. Dimensions. Lz 0-45 mm., Iz 0-35 mm., Lo 0-13 mm., lo 0-13 mm. H. janthina is obviously very closely related to H. edax, being distinguished by the wider shallower sinus and the greater frequency of frontal pores. At present, all western Atlantic records are grouped here under H. janthina. Soule & Duff stated that H. edax had not been "reported as living from the Pacific coast"; there is evidence (Cook MS), that a living form closely related to H. janthina has been recorded from Mazatlan, Mexico. 22. Hippoporidra littoralis sp. n. (PI. i, fig. 3, Text-fig. 7A) HOLOTYPE. B.M. 1963.4.16.3, Achimota Coll. Stn. C. MATERIAL. Achimota Coll. I. Stn. C, Winneba shore, rock face, 22.xi.4g, yB,* Paratypes. Stn. U, Tenpobo, No. i reef, 2i.xi.49, 67F, and Coll. II, 4A, 17.1.49. Stn. X, Axim, Hospital reef, 7.1.51, 68E. British Museum, Malacostraca Section, Apam, Ghana, with Pseudopagurus granulimanus (Miers) biafrensis (Monod), 1958.5.9.116-7.* H. littoralis is not present in the " Calypso " or Marche-Marchad Collections. DIAGNOSIS. Zoarium plurilaminar, nodular. Zooecia with marginal and at most a few scattered frontal pores, frequently with none. Orifice with anter narrower than poster, with a pair of denticles and sub-oral umbo large, constant. Cortical zooecia larger than autozooecia, orifice much smaller, surrounded by large blunt tubercles. Small rounded and semi-circular interzooecial avicularia present, none very large, often extremely rare. Ovicells with frontal area. Inhabits littoral and shallow waters. Dimensions. Lz 0-28 mm., Iz 0-30 mm., Lz (cortical) 0-45 mm., Iz (cortical) 0-40 mm., Lo 0-13 mm., lo 0-13 mm., Lo (cortical) 0-07 mm., lo (cortical) 0-05 mm., Lov 0-15 mm., lov 0-18 mm. It is possible that H. littoralis may be found to be a form of H. senegambiensis related to growth on short-spired shells (with hermit-crabs), of shallow waters. Until it can be discovered whether the ecological conditions inherent in the Pagurid association are able to affect zooecial as well as zoarial characters, the features in which these specimens differ from H. senegambiensis and H. picardi must rank as specific. The orifice is constantly wider proximally than distally, even in the cortical zooecia. This is a condition approached by some zooecia of the primary layer in H. senegambiensis, but not to the same degree, or so constantly. This type of orifice is also similar to that of H. janthina, from which H. littoralis differs in the form of the zoarium, the scarcity of frontal pores, and in the type of avicularia present. H. littoralis is confined to the shore of Ghana. Forest (1956 : 338, text-fig. 2) POLYZOA FROM WEST AFRICA 29 gave the distribution of the hermit-crab, Pseudopagurus granulimanus biafrensis as intertidal, and the association of H . littoralis may always be with this and similar littoral species. It may be compared with the deeper water association between Diogenes ovatus, Turritella shells and H, senegambiensis. 23. Hippoporidra senegambiensis (Carter) (PL 3, figs. 3-4, Text-figs. 7B-c, SA-D) Cellepora senegambiensis Carter, 1882 : 416, pi. 16, figs. IA-V. LECTOTYPE. Chosen here B.M. ig634.i6.2A, Guinea, West Africa.1 LECTOPARATYPES. 1963.4.16.26, C. MATERIAL. " Calypso " Coll. I. Stn. 6, 10° 19' N., 16° 33' 40" W., i6.v.56, 73-60 m., C2, specimen 4. Stn. 7, 9° 40' N, 13° 35' 5" W, I2.V.56, 8m., C4C*. Stn. 17, 5° N., 5° 28' 30" W., 2I.V.56, 27 m., C56H and C36A*, with Diogenes ovatus Miors. Stn. 19, 5° 2' 30" N., 5° 24' 40" W., 2I.V.56, 21-27 m., C4iA. Stn. 45, o° 25' N., 9° E., 8.vi.56, 73 m., C55E. Marche-Marchad Coll. I. Konakrey, Guinee Ise, II. Baie de Goree, 1954, 46-48 m., 6D. Sud de Goree, 24.xi.53, 40-41 m., nM. A Test de Goree, 6.iv.54, Coll. III. " Gerard Freca ", Dragage i, i8.ii.54, 23j ; 5, I5F ; 7, 8A. Sud de Goree, i3.xi.53, 33-35 m., gA* ; 34-37 m., I7A*. Devant 1'Anse de Bernard, Dakar, 12 m., loA*. Baie de Goree, I4A. Either S.W. Madeleines, i5.ix.53, 48 m., or Sud de Goree, I3.xi.53, 34-37 m., I7A*. Devant le " Castel ", Goree, 15 m., 22A*; 33-34 m., 27.xi.53, 296 ; 24K*, no information. Achimota Coll. I. Stn. K, on trawl debris i mile offshore, 2 miles W. of Densu R., 2.iii.49, 4 fath., 44!. Stn. 47, dredge haul No. i, 4.i.5i, 44 m., 14!. Stn. 62, dredge haul No. 2, i8.i.5i, 30 m., 87IA*. Stn. 63, as above, 36 m., 87IIA. Stn. 71, as above, 24.1.51, 30 m., 73A*. Stn. 98, as above, 14.111.51, 25 m., 36*. Stn. 112, Agassiz trawl No. 3, 4.iv.5i, 43 m., 6oS and Coll. II, loA*. Stn. 130, Agassiz trawl, No. 2, 26.iv.5i, 32 m., 706* and Coll. II, gA*. Stn. 131, as above, No. i, 2.v.5i, 37 m., 43U. Stn. 132, as above No., 44 m., 5oF. Coll. II. Stn. 133, as above, No. 3, 51 m., 7G. Stn. 117, as above, No. 2, 5.iv.5i, 64 m., I3F. Stn. 126, as above, No. 3, i2.iv.5i, 16 m., nA (W). Forest Coll. No. 3*, M'Bizi, Stn. 26, 1948-49, avec Eupagurus alcocki Balss (W). No. 4, Dakar, iv.1951. No. 5*, devant Joal, Senegal, 25 m., with Diogenes ovatus Miors. No. 6*, Baie de Rufisque, Goree, 18.1.52, 10-15 m., with D. ovatus and Paguriste fagei Fort. No. 7*, Entre Rufisque et Goree, 26.iii.53, 25 m., with D. ovatus. British Museum, Guinea, west Africa, 1963.4.16.2. Guinea coast, 1904.7.16.1, 2, 3, 4 (W). Cape Blanco, 1922.9.12.1*. Loanda, Angola, S. of da Costa, 1896.7.30.1 (W). Locality ? 1851.11.14.82, Belcher Coll. 1 Carter (p. 419) gave the location of his specimens as " Liverpool Free Public Museum ". Mrs. N. McMillan, of the City of Liverpool Museums, informs me (in litt. 29.7.1963) that " they must be presumed lost in the bombing ". The specimens in box 1963.4.16.2 were labelled as " Type specimens of Cellepora senegambiensis Carter " by Kirkpatrick, but none of them is that figured on pi. 16, fig. i, by Carter. 30 PATRICIA L. COOK DIAGNOSIS. Zoarium usually with long branched rugose arms. Zooecia small, with marginal pores and a few scattered frontal pores. Orifices of primary layer with a wide shallow sinus, orifices of subsequent layers usually with a rounder, deeper sinus. Cortical zooecia large, with 3-4 rows frontal pores, and very small orifices with large peristomial tubercles. Small adventitious avicularia frequent, large and small interzooecial avicularia with triangular and spathulate mandibles. Ovicells with frontal area, occasionally tuberculate. Dimensions. Lz 0-30 mm., Iz 0-25 mm., Lz (cortical) 0-70 mm., Iz (cortical) 0-50 mm., Lo o-io mm., Iz o-io mm., Lo (cortical) 0-08 mm., lo (cortical) 0-05 mm., Lm 0-04-0-15 mm., Lov o-ii mm., lov 0-15 mm. The orifices of the zooecia of the primary layer differ from those later zooecia in that their sinuses are wide and shallow, like those of H. janthina. They are never as wide as those of H. littoralis, but in a few colonies the primary type persists in a large number of the zooecia which form the plurilaminar branches. These zoaria are also marked by the predominance of the triangular type of inter zooecial avicularium and are labelled " W " in the lists of material above. Colonies of Hippoporidra sp. from Fiji (1862.2.4.13) and the Philippines (1882.7.29.48 and 1899.7.1.3518, 3519) also have orifices with a wide sinus and avicularia with triangular mandibles, which are, however, much shorter. 0-2mm FIG. 8. Mandibles in Hippoporidra. A-D. H. senegambiensis. A. Acute form, Senegal, Forest Coll. No. 3. B, c. Spathulate forms, Ghana, Achimota Coll. 14!. D. Rounded form, Ghana, Achimota Coll. 14!. E-F. H. picardi. E. Small sinuate form. F. Large, beaked sinuate form, Senegal, Marche-Marchad Coll. I, 33H. POLYZOA FROM WEST AFRICA 31 The structure of the zoarium has been described above (p. 24). In a few cases the branched arms are not developed (Cape Blanco, 1922.9.12.1 and Forest Coll. No. 6), but in each of these, there has been a heavy settlement of barnacles upon the shell during the development of the Polyzoan. The large interzooecial avicularia show a great deal of variation. The principal types are spathulate, semicircular and triangular, but there is some intergradation between these shapes. The young ovicell has a large membranous area frontally. Calcification reduces this first with a bar proximally above the orifice, and finally by progressive closure to a small circular distal area. In some zoaria the ectooecium is raised into tubercles distally and laterally, but generally the ovicell is smooth. H. senegambiensis is distributed round the west African coast from northern Senegal to the Bay of Biafra, and has been found off the coast of Angola. It has not been reported from Madeira, or the Canary or Cape Verde Islands. In character H. senegambiensis is intermediate between, but distinct from, both H. littoralis and H. picardi. The great majority of the colonies are associated with shells of Turritella annulata Kiener, which are inhabited by the Pagurid Diogenes ovatus. T. annulata is found off the Ghana coast from 10-64 m-> an(i is the dominant species of the silty-sand community or Turritella zone (see Bassindale, 1961 : 492, 500), which extends from 14-5-36 m. in depth. D. ovatus has a bathy metric distribution of 12-40 m. (see Forest, 1956 : 338, text-fig. 2). 24. Hippoporidra picardi Gautier (Text-figs. 70, SE-F) Hippoporidra picardi Gautier 1962: 254, text-fig. 22, Gulf of Salonika, Agean Sea, 8o-ioom. MATERIAL. " Calypso " Coll. I. Stn. i, 21° 05' N., 17° 14' W., io.v.56, 45-43 m., CiF. Stn. 6, 10° 19' N., 16° 33' 40" W., i6.v.56, 76-60 m., C2A, specimens i, 2* and 3. Marche-Marchad Coll. I. Sud de presque Tile de Cap Vert, i8.ii.54, 95 m-> 3oD and 33!!. Coll. III. As above, 46-50 m., lA*. Baie de Goree, 4A. 5610, IBC, 176-200 m., 6A*. Port Etienne, Mauritanie, vii.1953, 37-38 m., 7A. Au large de Goree, Stn. 556, 95 m., nA*. " Gerard Freca ", 2-15, 160-140 m., I2A ; N.W. Parar des Almadies, 170-145 m., I3A ; Dragage 5, i8.ii.54, igA ; 4, as above, 97-98 m., 28B. Forest Coll. Sud de Goree, 25.iii.54, 95 m., No. i*, with Dardanus arrosor (Herbst) and Pagurus cuanensis Thompson, No. 2. British Museum, Cape of Good Hope, 1842.12.2.29, Belcher Coll. H. picardi does not occur in the Achimota Collection. DIAGNOSIS. Zoarium like that of H. senegambiensis. Zooecia larger and more recumbent. Frontal of autozooecia with 2-4 rows of pores. Orifice large, with well-developed denticles and a very small, rounded sinus. Orifices of cortical zooecia reduced in size. Interzooecial avicularia with triangular, spathulate and sinuate mandibles, some very large. Ovicell large with a rounded frontal area. 32 PATRICIA L. COOK Dimensions. Lz 0-45 mm., Iz 0-30 mm., Lz (cortical) 0-70 mm., Iz (cortical) 0-50 mm., Lo 0-14 mm., lo o-n mm., Lo (cortical) o-io mm., lo (cortical) 0-06 mm., Lm 0-04-0-15 mm., 1m 0-05-0-25 mm., Lov 0-20 mm., lov 0-20 mm. H. picardi differs from both H. senegambiensis and H. littoralis in its larger zooecia, with more porous frontal, and in the shape of the orifice. The largest interzooecial avicularia, although closely related to those of H. senegambiensis with a rounded mandible, differ in their sinuous distal edge (see Text-figs. SE-F). The ovicells, however, show no significant differences, except their generally larger size. Von Marten (1877 : 183) mentioned a Polyzoan with erect arms growing on a shell inhabited by a hermit-crab from 50 fath. off the South African coast (33° 59' S., 17° 52' E.), but did not give any further description. H. picardi is now known from the more northerly west African localities, and from the Cape of Good Hope. It is a deep-water form, and it is interesting that the first, and only previous record, is not from the Mediterranean proper, where Hippoporidra does not occur (see Gautier, 1962 : 255), but from deep water in the Aegean. Gautier did not give the name of the hermit-crab associated with his specimen. The two species of crab named in the Forest Collection both extend in distribution to a depth of 100 m. P. cuanensis has the more northerly range, extending to the south coasts of Britain and Ireland. 25. ACKNOWLEDGMENTS I am indebted to Dr. A. B. Hastings (British Museum (Natural History)), first for her interest and encouragement during this work, and second, for the opportunity to use her collection of references on Pagurid associations, upon which section 18 is based. My thanks are also due to Dr. A. Andersson (Naturhistoriska Riksmuseet, Stockholm) and Dr. D. E. Owen (Manchester Museum), for the loan of specimens ; to Mr. J. V. Brown and Mr. P. Green for the photography, and to Mr. A. C. Wheeler for the X-ray photographs. Finally I am grateful to Dr. J. P. Harding (Entomo- straca section), for his help and criticism. 26. SUMMARY The synonymies of three genera, Hippoporina, Hippoporella and Cleidochasma' are given. Four species of Hippoporina are described ; H. pertusa, here found from Ghana, has a very wide geographical distribution ; H. americana, also widely distributed, is here recorded solely from waters of low salinity in Nigeria. H. lacrimosa and H. acuta, spp. n., are forms in which the avicularia differ from those of species previously described. Hippoporella gorgonensis is recorded for the first time from the eastern Atlantic, and is briefly compared with H. multidentata, a species from the Pacific and Indian Oceans. H. pusilla, here occurring in the eastern Atlantic from Madeira and the Cape Verde Islands only, is also described. Five species of Cleidochasma are described. C. porcellanum, which exhibits great variation in the character of the orifice, is a circum-tropical species, here considered to include C. bassleri ; C. contractum, another widely distributed species, is tolerant of colder waters. Ovicells are described for the first time in C. oranense and C. brancoense, species with large, erect zoaria. C. rotundorum has not been reported from west POLYZOA FROM WEST AFRICA 33 Africa, but is found at Madeira, and has been previously confused with one of the species here included in the genus Hippoporidra. Hippoporidra is re-defined, and two species previously assigned to it, H. granulosa and H. spiculifera, are referred to Cleidochasma. Some associations between Pagurids, Gastropods, and encrusting organisms are briefly listed and the characters peculiar to the Hippoporidra- association described. Detailed examination, including X-ray photography, of the large amount of material available, has shown that the Gastropod shell is not eroded by the Polyzoan, and that the Mollusc is absent, so that the association, in Recent specimens at least, is between Polyzoan and Pagurid. H. edax and H. janthina, neither of which is west African, are briefly defined and described. The occurrence of the three west African species of Hippoporidra appears to be correlated with the species of Gastropod shell available, and with the bathymetrical distribution of the Pagurids with which these are associated. H. littoralis is an intertidal and shallow-water species, H. senegambiensis occurs in deeper waters, and H. picardi, which has been previously reported only from the Aegean, inhabits even greater depths, up to 200 m. 27. REFERENCES BARROSO, M. G. 1925. Notas sobre briozoos del Mediterraneo. Bol. Real. soc. esp. Hist. Nat., 25 : 177-184. BASSINDALE, R. 1961. On the Marine Fauna of Ghana. Proc. zool. Soc. Lond., 137, 4 : 481-510. BOECK, A. 1862. Forelobige Bemaerkninger angaaende de ved de sydlige og vestlige Kyster af Norge forekommende Polyzoer. Fork. Vid.-Selsk. Christiana Aar 1861 ; 49-50. BUCHANAN, J. B. 1958. The bottom fauna communities across the continental shelf off Accra, Ghana. Proc. zool. Soc. Lond., 130 : 1-56. BUGE, E. 1957. Les Bryozoaires du Neogene de 1'ouest de la France. M6m. Mus. Nat. hist. Nat., n.s. 100, 6 : 1-436. BUSK, G. 1859. A Monograph of the Fossil Polyzoa of the Crag. Mon. Palaeont. Soc. Lond. : i-xiii, 1-136. 1860. Zoophytology. Catalogue of the Polyzoa collected at Madeira. Quart. J. Micr. Sci., 8 : 280-285. 1861. Descriptions of New or imperfectly known Polyzoa No. i. Ibid., n.s. i : 153-156. CALVET, L. igo6a. Note preliminaire sur les Bryozoaires du Travailleur et du Talisman. Butt. Mus. Paris, 12, 3 : 154-166. I9o6b. Deuxieme note. Ibid., 12, 4 : 215-222. 1907. Bryozoaires. Expeditions Scientifiques " Travailleur " et " Talisman ", 1880-1883, 8 : 355-495. Monaco. 1931. Bryozoaires provenant des Campagnes Scientifiques du Prince Albert ler de Monaco. Res. Camp. Sci. Monaco, 83 : 1-152. CANU, F. 1916. Bryozoaires du S.W. de la France. Bull. Soc. geol. Fr., s6r. 4, 15 : 320-334. — 1917. Danmark Ekspeditionen til Groenlands nordostkyst 1906-1908, Bryozoa, par G. R. Levinsen. Rev. crit. Paleozool., 21, i : 35-37. CANU, F. & BASSLER, R. S. 1917. A Synopsis of American Early Tertiary Cheilostome Bryozoa. Bull. U.S. nat. Mus., 96 : 1-87. - 1920. North American Early Tertiary Bryozoa. Ibid., 106 : v-xx, 1-879. — 1923. North American Later Tertiary and Quaternary Bryozoa. Ibid., 125 : ii-vii, 1-302. 1927. Classification of the Cheilostomatous Bryozoa. Proc. U.S. nat. Mus., 69 : 14:1-42. 34 PATRICIA L. COOK CANU, F. & BASSLER, R. S. ig28a. Fossil and Recent Bryozoa of the Gulf of Mexico Region. Ibid., 72, 14 : 1-199. - 19280. Les Bryozoaires du Maroc et de Mauritanie. Mem. Soc. Sci. nat. Maroc., 18 : 7-85. - 1929. Bryozoa of the Philippine Region. Bull. U.S. nat. Mus., 9, 100 : ii-xi, 1-685. - i93oa. The Bryozoan Fauna of the Galapagos Islands. Proc. U.S. nat. Mus., 76, 13 : 1-78. - 19300. Bryozoaires marins de Tunisie. Ann. Sta. oceanogr. Salammbo, 5 : 1-91. CARTER, H. J. 1882. Remarkable forms of Cellepora and Palythoa from the Senegambian coast. Ann. Mag. nat. Hist., (5) 9 : 416-419. COOK, P. L. 1964. Polyzoa from west Africa. I. Notes on the Steganoporellidae, Thalamoporellidae and Onychocellidae (Anasca, Coilostega). Ann. Inst. Oceanogr. (Calypso, VI), 41 : 43-78. DOFLEIN, F. 1914. Das Tier als Glied der Naturganzen. Tierbau und Tierleben Bd. 2. Leipzig and Berlin. DUVERGIER, J. 1924. Deuxieme note sur les Bryozoaires du Neogene de 1'Aquitaine. Act. Soc. Linn. Bordeaux, 75 : 5-50. ESPER, E. J. C. 1794-7. Forsetzungen der Pflanzenthiere in Abbildungen nach der Natur mit Farbenerleuchtet nebst Beschreibungen, Theil, i : 1-230. (Pp. 1 17-168 were published in 1796). Niirnberg. FOREST, J. 1956. Sur une collection de Paguridae de la Cote de 1'Or. Proc. zool. Soc. Lond., 126, 3 : 335-367- GAUTIER, Y. V. 1962. Recherches ecologiques sur les Bryozoaires Chilostomes en Mediterran£e Occidentale. Theses presentees a la Faculte des Sciences de I'Universite d' Aix -Marseille, 91 : 9-434- HARMER, S. F. 1957. The Polyzoa of the Siboga Expedition part 4, Cheilostomata Ascophora II. Rep. Siboga Exped., 28d : 641-1147. HASTINGS, A. B. 1930. Cheilostomatous Polyzoa from the Vicinity of the Panama Canal. Proc. zool. Soc. Lond., 1929 : 697-740. HESSE, R. & DOFLEIN, F. 1943. Das Tier als Glied des Naturganzen. Tierbau und Tierleben, Ed. 2, Bd. 2. Jena. HINCKS, T. 1880. A History of the British Marine Polyzoa : i-cxli, 1-601. London. KIRKPATRICK, R. & METZELAAR, J. 1922. On an Instance of Commensalism between a Hermit Crab and a Polyzoon. Proc. zool. Soc. Lond., 1922 : 983-990. LAGAAIJ, R. 1952. The Pliocene Bryozoa of the Low Countries. Meded. Geol. Sticht.,C5, 5 : 1-233- MARCUS, E. 1937. Bryozoarios Marinhos Brasileiros I. Bol. Fac. Filos. Cienc. S. Paulo, Zool., 1 : 5-224. - 1938. Bryozoen von St. Helena. Vidensk. Medd. Dansk naturh. Foren., 101 : 183-252. - 1941. Sobre Bryozoa do Brasil. Bol. Fac. Filos. CiSnc. S. Paulo, Zool., 5 : 3-208. - 1955- Notas sobre Briozoos marinhos Brasileiros. Arq. Mus. nac., 42 : 273-324. MARTEN, E. von. 1876. Ueber einige neue Beitrage der afrikanischen Gesellschaft und der "Gazelle ". S.b. ges. Naturf. freunde Berlin, 1876 : 18-22. 1877. Ueber ein auf todten Schneckenschalen angesiedeltes Bryozoon. Ibid., 1877 : 183. MICHELIN, H. 1847. Iconographie Zoophytologique. Description des Polypiers fossiles de France : 249-328. (For dates of publication of the parts of this work see Lagaaij, 1952 : 215.) MOORE, H. B. 1937. Marine Fauna of the Isle of Man. Proc. Lpool. biol. Soc., 50 : 5-293. NEVIANI, A. 1895. Briozoi neozoici di alcune Localita dTtalia. Parte Prima. Boll. Soc. Rom. Stud. Zool., 4 : 109-123. NORMAN, A. M. 1909. The Polyzoa of Madeira and neighbouring islands. /. Linn. Soc. (Zool.}, 30 : 275-341. POLYZOA FROM WEST AFRICA 35 O'DoNOGHUE, C. H. 1957. Some South African Bryozoa. Trans, roy. Soc. S. Africa, 35, 2 : 71-95- O'DONOGHUE, C. H. & DE WATTEViLLE, D. 1944. Additional notes on Bryozoa from South Africa. Ann. Natal Mus., 10, 3 : 407-432. OSBURN, R. C. 1912. The Bryozoa of the Woods Hole Region. Bull. U.S. Bur. Fish, 30 : 205-266. - 1940. Bryozoa of Porto Rico. Scientific Survey of Porto Rico and the Virgin Islands., 16, 3 : 321-486. New York. - 1950. Bryozoa of the Pacific Coast of America, Part, i, Cheilostomata Anasca. Rep. Allan Hancock Pacific Exped., 14, i : 1-269. - 1952. Op. cit., Part 2, Cheilostomata- Ascophora. Ibid., 14, 2 : 271-611. ROGER, J. & BUGE, E. 1948. L' Association Cellepore-Gastropode dans les faluns de la Touraine. Bull. Soc. geol. Fr., s£r. 5, 17 : 461-470. RYLAND, J. S. 1960. The British Species of Bugiila (Polyzoa). Proc. zool. Soc. Lond., 134, i : 65-105. SAKAKURA, K. 1935. Pliocene and Pleistocene Bryozoa from the Bozo Peninsula, i. Bryozoa from off the Dizodo Beds. /. Fac. Sci. Tokyo Univ. Geol., 4 : 1-48. SCHMITT, W. L. 1931. Shelled Invertebrates of the Past and Present. Smithsonian Sci. Ser., 10. SMITT, F. A. 1868. Kritisk Forteckning ofver Skandinaviens Hafs-Bryozoer, 4. Ofvers Vetensk-Akad. Fork. Stockholm, 1867, 24 Bihang : 3-230. - 1873. Floridan Bryozoa, Part. 2. K. svenska VetenskAkad. Handl., 9, 4 : 3-83. SOULE, J. D. 1961. Results of the Puritan- American Museum Expedition to Western Mexico, 13. Ascophoran Cheilostomata (Bryozoa) of the Gulf of Mexico. Amer. Mus. Novit., No. 2053 : 1-66. SOULE, J. D. & DUFF, M. M. 1957. Fossil Bryozoa from the Pleistocene of Southern California. Proc. Calif. Acad. Sci., 4, 29 : 87-146. STECHOW, E. 1921. Neue Gruppen skelettbildenden Hydrozoen. Verhandl. deutsch. Zoolog. Gesell., 26 : 29-31. STUDER, T. 1889. Die Forschunsgreise S.M.S. Gazelle, 1874-1876. 3, Zool. &> Geol. : 1-322. THORNELY, L. R. 1905. Report on the Polyzoa, in, W. A. Herdman. Rep. Pearl Oyster Fisheries, Gulf of Manaar, 4, Suppl. Rep., 26 : 107-130, London. ULRICH, E. O. & BASSLER, R. S. 1904. Bryozoa, in, Systematic Paleontology of Maryland. Maryland Geological Survey, Miocene : 404-429, Baltimore. VERRILL, A. E. i875a. Brief Contributions to Zoology from the Museum of Yale College, 32. Results of Dredging Expeditions off the New England Coast in 1874. Amer. J. Sci. Arts, 9 : 411-415. Op. cit. 33. Ibid., 10, : 36-43. WATERS, A. W. 1885. Chilostomatous Bryozoa from Aldinga and the River Murray Cliffs. Quart. J . geol. Soc. Lond., 37 : 309-347. - 1899. Bryozoa from Madeira. /. R. micr. Soc., 1899 : 6-16. Some Collections of the Littoral Marine Fauna of the Cape Verde Islands. J. Linn. Soc. (Zool), 34 : 1—45. Some Mediterranean Bryozoa. Ann. Mag. nat. Hist., (9) 2 : 96-102. PLATE i Hippoporella, Cleidochasma and Hippoporidra FIG. i. Hippoporella gorgonensis Hastings. Ghana, Achimota Coll., 3&G. Zooecia, those in the centre of the colony with ovicells. x 32. FIG. 2. Cleidochasma rotundorum (Norman). Madeira, Museo do Seminario, Funchal, lectotype. Zooecia with ovicells. x 32. FIG. 3. Hippoporidra littoralis n. sp. Ghana, Achimota Coll., 76. Zooecia, treated with eau de javelle. x 36. FIG. 4. Cleidochasma porcellanum (Busk). Tortugas, Naturhistoriska Riksmuseet, Smitt Coll., No. 284. (? type of Lepralia cleidostoma Smitt). Zooecia and ovicells (see also PL 2, fig. i). X5o. Bull. EM. (N.H.) Zool. 12, i PLATE i PLATE 2 Ovicells in Cleidochasma FIG. i. C. porcellanum (Busk). Tortugas, Naturhistoriska Riksmuseet, Smitt Coll., No. 284. Striated ovicell (see also PL i, fig. 4). x 113. FIG. 2. C. porcellanum. Ceylon, iQ36.i2.3o.34A pt., Thornely Coll. Zooecia and ovicells with secondary calcification, x 54. FIG. 3. C, oranense (Waters), west Africa, 1921.5.23.7. Zooecia, avicularia and ovicell (o*). X54- FIG. 4. C. brancoense (Calvet). Bay of Biafra, " Calypso " Coll., C55F. Ovicell (o*). X6o. Bull. B.M. (N.H.) Zool. 12, i PLATE 2 * *• j&tf 1 PLATE 3 Cleidochasma and Hippoporidra FIG. i. Cleidochasma brancoense (Calvet). Cape Verde Islands, " Calypso " Coll., CgyA. Part of a branch, showing the large interzooecial avicularia ; treated with eau de javelle. X4-2. FIG. 2. C. oranense (Waters). Ghana, Achimota Coll., 320. Part of a branch, treated with eau de javelle. X 8.5. FIGS. 3-4. X-ray photographs of Hippoporidra senegambiensis (Carter), 5ma, 4oKV, 120 seconds. FIG. 3. Senegal, Marche-Marchad Coll. Ill, loA. Showing position of gastropod shell (s) and chela of hermit-crab (c). x 1.7. FIG. 4. Guinea, west Africa, 1963.4.16.2 B, C, Lectoparatypes. Showing laminae of Polyzoan growing out from orifice of gastropod shells (/). X 1.7. FIGS. 5-7. X-ray photographs of H. edax (Busk), 5ma, 4oKV, 120 seconds. FIG. 5. Guernsey, 1963.4.16.1, Busk Coll. x 1.7. FIG. 6. Plymouth, 1899.5.1.1517, Hincks Coll. x 1.7. FIG. 7. Britain, Pliocene, Coralline Crag, 61620, lectotype. x 1.5. Bull. B.M. (N.H.) Zool. 12, t PLATE 3 PRINTED IN GREAT BRITAIN BY THOMAS DE LA RUE & COMPANY LIMITED LONDON 1. 0, FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA r<*f ' APk 196; ^ D. S. BROWN BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 2 LONDON: 1965 FRESHWATER GASTROPOD MOLLUSC A FROM ETHIOPIA BY D. S. BROWN Medical Research Council, c/o Section of Experimental Taxonomy, Zoology Department, British Museum (Natural History) Pp. 37-94 ; Plates 1-3 ; 47 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 2 LONDON : 1965 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical Series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. This paper is Vol. 12, No. 2 of the Zoological series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals. Trustees of the British Museum (Natural History) 1965 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued April, 1965 Price Twenty-one Shillings FRESHWATER GASTROPOD MOLLUSC A FROM ETHIOPIA By D. S. BROWN SYNOPSIS Gastropod molluscs were collected by the author between May and September 1962 from a variety of freshwater habitats in Ethiopia. Twenty-eight forms are represented in the collection of which a total of eleven have not been previously recorded from Ethiopia or have been recorded with doubt by previous authors. Two forms in each of the genera Bulinus and Ancylus appear to be new. It is considered that seven of the forms recorded from Ethiopia for the first time belong to the African component in the fauna, whereas only three are possibly of palaearctic origin. It is concluded that the wide distribution and diversity of some of the palaearctic forms justify the importance attached by Bacci (1951) to the palaearctic element in the molluscan fauna of the Ethiopian Plateau. INTRODUCTION A COLLECTION of freshwater molluscs was made by the author between May and September 1962 in the course of a survey to determine the distribution of potential intermediate hosts of human schistosomiasis in Ethiopia. Facilities provided by the Haile Selassie I, University College of Addis Ababa made it possible to visit many regions of Ethiopia, and a wide variety of aquatic habitats between sea level and 3,000 metres (10,000 ft.) were examined. An account of the Gastropoda collected during the survey is given in the present paper. Four new forms of Basommatophora are described and a total of ten do not appear to have been pre- viously recorded from Ethiopa. Prosobranchia are relatively poorly represented in the collection ; no specimens of the genera Cleopatra, Gabbia or Lanistes were found. This can be only partly attributed to the high altitudes at which most of the work was carried out (over 1,500 m.), as the latter genus is present in Lake Tana (1,900 m.). The arrangement of families and subfamilies is that employed by Wenz and Zilch1. As taxonomy within several families of African freshwater molluscs is in a state of flux at the level of the genus and below, a generic arrangement has been selected from existing accounts of the African fauna. Genera are used in the manner of Mandahl-Barth (1954), and Bulinus and Biomphalaria are classified in accordance with the species-group system proposed by this author (i957a, b). Ferrissia Walker is used in accordance with past practice although according to Dr. B. Hubendick (in litt.) true Ferrissia do not occur in Africa. Identifications have been based on shell and anatomy, and information is given that is relevant in comparison to existing descriptions which are amplified in some cases. Wherever possible, the forms in the present collection have been assigned to previously described taxa and where there are differences to the descriptions of previous authors these are discussed. Two forms of Bulinus, and one form of Ancylus are apparently new but have not been named as it is evident that revision is necessary 1Wenz, W. Handbuch der Palaeozoologie, Berlin. 6. Gastropoda Pt. i Prosobranchia 1938-1944. Pt. 2. Euthyneura (by A. Zilch) 1959-1960. 4o D. S. BROWN throughout these genera before the status of these Ethiopian forms can be properly established. One purpose of the present paper is to bring the material to the atten- tion of malacologists who may undertake works of revision in the future. In comparison to other parts of Africa a relatively large number of travellers have made collections of molluscs in Ethiopia. Bacci (1951) gives a comprehensive catalogue of records of Ethiopian Mollusca up to 1941 and proposes many cases of synonymy. Synonymies given in the present paper are primarily local and the references included represent either original locality records or nomenclatural variations. Works in which names have been cited without change are omitted. Localities are denned in relation to towns or other landmarks which are all listed at the end of the paper with map references. In the text localities are grouped into provincial areas in order to make possible a rapid assessment of distribution by reference to Map i in which approximate provincial boundaries are shown. The routes followed during the survey are also indicated on the map. The majority of habitats in which molluscs were found were situated on the Ethiopian plateau between altitudes of approximately 1,800 m. (6,000 ft.) and 2,700 m. (9,000 ft.). A general account of the geography and climate of the country is given by Ayad (1956) ; the distribution and nature of aquatic habitats is described elsewhere (Brown, in preparation). The number of specimens obtained in each sample is shown between brackets, and an asterisk indicates that only empty shells were found. The collection is deposited in the Section of Experimental Taxonomy, British Museum (Natural History). Specimens have also been deposited in the World Health Organisation Snail Identification Centre, Charlottenlund, Denmark. These collections are referred to in the text as (B.M.N.H.) and (W.H.O.) respectively. It is a pleasure to record my thanks to Mr. J. MacFarlane and Dr. A. Tj0nneland (Haile Selassie I University College, Addis Ababa) for their ready assistance. I am grateful to the Trustees of the British Museum (Natural History) for the provision of working facilities, and to Dr. C. A. Wright for accommodation in the Section of Experimental Taxonomy. Dr. G. Mandahl-Barth has been generous of his time, I am indebted to him for much stimulating discussion and the opportunity to examine the collection of the World Health Organisation Snail Identification Centre. Dr. J. Burch has kindly allowed me to quote some of his unpublished observations on the chromosomes of Bulinus. Thanks are due to Mr. D. Claugher and Miss J. Lines for the care they have devoted to the preparation of photographs and radulae. Family VIVIPARIDAE BELL AMY A Jousseaume Bellamya unicolor (Olivier) Cyclostoma unicolor Olivier, 1804 : 39, pi. 31, fig. 9 ; Paludina unicolor Martens, 1866 & 1867 ; Vivipara unicolor ; Jickeli, 1874 ; Viviparus unicolor ; Thiele, 1933 ; Bacci, 1940 ; Piersanti, 1940 ; Viviparus (Bellamya) unicolor ; Bacci, 1943 ; Bellamya unicolor unicolor ; Bacci, 1951 ; Bellamya unicolor ; Mandahl-Barth in Ayad, 1956. Paludina abyssinica Martens, 1866 : 97, pi. 3, fig. 7, & 1867 ; Vivipara abyssinica ; Jickeli FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA Eritrea MAP i. Ethiopia. Boundaries of Provinces indicated by broken lines ; routes followed in making the present collection indicated by solid lines. 42 D. S. BROWN 1874 ; Viviparus unicolor abyssinicus ; Piersanti, 1940 ; Bellamya unicolor abyssinica; Bacci, 1951. LOCALITIES. Arussi : 5 km. approx. S. of Adamitullo (between Lake Zwai and Bulbulla River) (8)*; Lake Abyata, east shore (i)*. Begemeder : Lake Tana at Gorgora (130). Shell Lake Tana (pi. i, i & 2). Although the apices are eroded in the larger specimens many shells are moderately short spired having the length of the spire approxi- mately equal to that of the aperture. The peripheral angulation is acute in young specimens, but is indistinct in large shells. The shoulder angulation is scarcely visible in both big and little specimens. Microsculpture on the apical whorls consists of coarse spiral ridges, which subsequently become finer and wavy and are crossed by irregular transverse growth ridges. In a representative shell there are 5 spiral ridges at the beginning of the second whorl, and 16 at the beginning of the third whorl. The outer lip of the aperture is thickened and coloured black-brown in many of the larger shells, in others this dark band is present a short distance from the outer lip and further back there may be one or two more according to the size of the shell. All the bands are usually situated in the ultimate whorl. Similar bands have been observed in Bellamya from other parts of Africa and it is likely that growth takes place in several stages in such populations. There appears to be sexual dimorphism in size : 19 of 23 specimens of more than 20 mm. length are female ; 3 of n specimens between 15 and 20 mm. are female. Largest shell : L = 27-7 mm.; ML = 13-7 mm.; W = 19-4 mm. Arussi : near Lake Zwai (pi. i, 3). All the shells are distinguishable from Lake Tana specimens of the same size by the more acute peripheral and shoulder angula- tions, and the relatively sharp apices. Microsculpture on the lower whorls is similar to that of the Lake Tana specimens, but the apical whorls are smooth, probably as a result of weathering. Sub-fossil in a lacustrine deposit. Largest shell : L = 21-5 mm. ; ML — 10-3 mm. ; W = 16-5 mm. ; 6| whorls. Anatomy (Text-figs. 1-5) The male genital system (Text-figs, i and 2) closely resembles that of Viviparus (= Bellamya) unicolor as described by Rohrbach (1937). A crescentic testis with a narrower anterior end is situated on the right side of the roof of the mantle cavity close to the right side of the rectum. The ureter is attached to the adjoining edges of the testis and the rectum. After emerging from near the posterior end of the testis, the vas deferens passes beneath connective tissue across the floor of the mantle cavity to the left side, where it widens and runs forward beneath a branchial fold as the prostate " gland ". After becoming narrower the vas deferens passes into the right tentacle, at the tip of which is the male aperture. The uterus occupies a similar position to the testis in the male, with the vagina projecting as a short tube from the floor of the mantle cavity. Great variation was observed in the number and the size of the shelled embryos present in large females FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 43 of approximately the same size. Occasionally only a single young snail was present' but more frequently a few large ones were arranged in a single row (Text-fig. 3) . In some specimens many young snails were arranged in two or three rows ; the greatest number of young observed in such a uterus being 28 (Text-fig. 4) . Radula (Text-fig. 5) The length of the median cusp of the central tooth is more than half the breadth of the cusp. There are 3-7 lateral cusps on each side of the median cusp, and in addition the upper lateral edges of the central tooth may bear a few notches. met Mantle cavity of male exposed Female of shell length 27-2 mm., FIGS. 1-5. Bellamya unicolor from Lake Tana. Fig. i. by turning mantle to right. Fig. 2. Testis. Fig. 3. left side of mantle removed and the remainder turned to the right exposing uterus, embryos numbered 1-6. Fig. 4. Uterus of female 21-5 mm. shell length with wall removed to expose young snails and embryos. Fig. 5. Cusp bearing parts of two representative central teeth from radula. 44 D. S. BROWN REMARKS. In the shape of the shell, and the presence of a short median cusp on the central tooth of the radula, these specimens conform to the Bellamya unicolor group described in Uganda by Mandahl-Barth (1954). If the apices were not eroded some of the larger shells from Lake Tana (pi. i, i) would approach the elongated shape of B. abyssinicus (Martens) in which the spire is approximately equal to i^ times the aperture length. In addition to the original specimens from South Ethiopia, von Martens' species appears to be known only from Lake Tana (Martens, 1867 and Piersanti, 1940), but the range of variation in the present sample from this lake indicates that the form is not separable from B. unicolor. As appears to be the case in other parts of Africa, Bellamya has only been found in or near large bodies of water in Ethiopia. The sub-fossil shells from the vicinity of Lake Zwai probably represent a population that existed when the floor of this part of the Rift Valley was more extensively covered by fresh water than it is at the present time (Mohr, 1961) ; there is a close resemblance between these shells and those described by Bacci (1940) from a nearby locality. Thiele (1933) records living B. unicolor from near Lake Zwai, and it is possible that living populations remain to be discovered in some of the Rift Valley Lakes. Rohrbach (1937) found specific anatomical differences between B. unicolor from the Nile at Cairo, and B. capillatus from N. Rhodesia. The male genital system of the Lake Tana specimens resembles that of the specimens from the Nile described by Rohrbach, but only a few of the Lake Tana females have a large number of embryos arranged in two or more rows as described by this author, who found up to 86 relatively small embryos in B. unicolor in contrast to 25-40 larger ones in B. capillatus. Though the number of embryos in the Lake Tana females is generally even less than that found in B. capillatus by Rohrbach, the arrangement in several rows when a large number is present, considered in conjunction with the anatomy of the male system, is evidence that this population is related to B. unicolor of the Nile. The small number of embryos present in the uterus may represent an adaptation to local conditions and it is probable that the anatomical and ecological distinctions between B. unicolor and B. capillatus will be found to be less clear than was thought by Rohrbach. DISTRIBUTION. Bellamya is known in Ethiopia only from Lake Tana, near Lake Zwai, Lake Abyata (Bacci, 1943), and an unspecified locality in the southern part of the country (Martens). Many species which are probably conspecific with B. unicolor have been described in other parts of Africa ; the range of the genus extends down the basin of the River Nile to the Mediterranean and as far west as Senegal, the southern limit of distribution appears to be in Zululand. Family AMPULLARIIDAE PILA Roding Pila speciosa (Philippi) Ampullaria speciosa Philippi, 1849 : 18, and in Martin & Chemnitz, 1851 : pi. n, fig. 2. LOCALITY. Sidamo : 40 km. approx. W. of Dolo on Filtu road, near Ganale River, (i).* FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 45 Shell The dimensions of the single empty shell are : L = 50 mm. ; ML = 38 mm. ; W = 45-5 mm. REMARKS. This species is known from Somalia and the Kenya-Somalia border (Bacci, 1951 and Pain, 1961), and it is not surprising that its range extends for at least a short distance up the Ganale River into Ethiopia. Family VALVATIDAE VALVATA Miiller Valvata sp. Valvata nilotica Jickeli, 1874 : 223, pi. 7, fig. 29 ; Bacci, 1940 ; Valvata nilotica nilotica ; Bacci, 1951 ; Valvata nilotica var. scioana ; Pollonera, 1888 : 62 ; Valvata nilotica scioana ; Bacci, 1940. Valvata sp. Mandahl-Barth in Ayad, 1956. LOCALITIES. Choa : 72 km. N. of Addis Ababa boundary of Debra Markos road, (17) ; 0-5 km. S. of Debra Berhan, (50) ; Lake Wonji, (10). Shell (pi. i, 4-9) Two specimens are illustrated (pi. i, 6 and 8) which represent the extremes of variation in the rate of descent of the whorl within a single sample (72 km. N. of Addis Ababa on Debra Markos road). In the Debra Berhan sample (pi. i, 4 and 5) the range of the ratio L/W is o-82-i-oo (10 shells). The umbilicus of shells from Lake Wonji (pi. i, 7 and 9) is distinctly smaller than in the other samples. Fine transverse ribs are more or less regularly spaced on the apical whorls, where there is also a fine spiral sculpture. Largest shell : L = 4-8 mm. ; ML = 3-0 mm. ; W = 6-0 mm. ; 3! whorls. REMARKS. If Valvata of the Egyptian Nile are regarded as distinct from European species, it is possible that V. saulcyi Bourguignat 1853 should take precedence over V. nilotica Jickeli. As a detailed comparison between specimens of Valvata from Ethiopia, Egypt and Europe has not yet been made, a specific identifi- cation of the specimens in the present collection is impossible. Variety scioana was described on the basis of three specimens from near Debra Berhan, in which the spire was considered to be less depressed, and the umbilicus more restricted, than in Jickeli's illustration of V. nilotica. Bacci (1940) observed that the majority of fossil specimens of Valvata from the vicinity of Lake Zwai were similar to var. scioana, and most of the shells in the present collection resemble this form in possessing a relatively rapidly descending whorl. The shells from Debra Berhan have even higher spires than the shells of scioana described by Pollonera : the mean ratio L/W is 0-90 (10 shells), compared to ratios of 0-66 (var. scioana) and 0-63 (V. nilotica} calculated from the dimensions given by Pollonera and Jickeli respectively. Restriction of the umbilicus does not appear to be necessarily 46 D. S. BROWN associated with increase in the rate of descent of the whorl as suggested by Pollonera. Both low-spired and the high-spired shells from the north of Addis Ababa (pi. i, 6 and 8) have a large umbilicus, but it is restricted in all the specimens from Lake Wonji. Valvata revoili Bourguignat, 1889, is listed with reservation by Bacci (1951), and is considered to be a terrestrial species by Dr. G. Mandahl-Barth (pers. comm.). DISTRIBUTION. Valvata sp. appears to have a wide range at moderately high altitudes in Ethiopia, occurring in both still and slow-flowing waters ; it is recorded by Ay ad from south of Addis Ababa (Akaki) and from the Blue Nile near Lake Tana. This genus reaches the southern limit of its distribution in North Africa ; several species have been described from the lower course of the Nile and V. tilhoi Germain was described from the vicinity of Lake Chad. Family THIARIDAE MELANOIDES Olivier Melanoides tuberculata (Muller) Nerita tuberculata Muller, 1774 : 191 ; Melania tuberculata ; Blanford, 1870 ; Morelet, 1872 ; Jickeli, 1874 ; Bourguignat, 1872, & 1833, & 1885 ; Pollonera, 1898 ; De Rochebrune & Germain, 1904 ; Neuville & Anthony, 1908 ; Kobelt, 1910 ; Melania (Melanoides} tuber- culata ; Thiele, 1933 ; Bacci, 1940 & 1941 ; Melanoides (Thiara) tuberculata ; Piersanti, 1940, & 1941 ; Melanoides tuberculata ; Ayad, 1956 ; Verdcourt, 1960 ; Melania tuberculata var. costata Bourguignat, 1864 : pi. 15, figs. 3 & 10 ; Pollonera, 1898. Melania dembea Reeve, 1860 : pi. 23, fig. 161. Melania curvicosta Pantanelli, 1897 : 204. LOCALITIES. Arussi : 5 km. approx. S. of Adamitullo (between Lake Zwai and Bulbulla river), (6) *; Lake Zwai, N.W. shore, (i) ; River Awash, S. of Nazareth, (52). Eritrea : 20 km. approx. W. of Asmara on Massawa road. Sidamo : Lake Margherita, E. shore, (13).* Wallo : Lake Haik, (2). Shell Spiral ribs are present on the lowermost part of the shell, and tubercles are generally better developed than in the forms described by Mandahl-Barth (1954). In specimens from the River Awash, south of Nazareth, conspicuous transverse ribs are formed by the fusion of the large tubercles. REMARKS. Specimens from near Saati (between Asmara and Massawa), with thick transverse ribs, were identified by Pollonera (1898) as var. costata Bourguignat. The specimens described above from the River Awash also resemble this form. No specimens were found in Lake Tana, the type locality of Melania dembea Reeve, a form with moderately well developed transverse ribs that does not appear to differ significantly from M. tuberculata. DISTRIBUTION. M. tuberculata is widely distributed in Ethiopia from near sea-level to an altitude of about 6,000 feet (Lake Tana) on the plateau. Its range includes most of Africa, excluding the Niger and Congo basins, and extends into the Arabian Peninsula and Asia. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 47 Family PHYSIDAE PHYSA Draparnaud Physa sp. LOCALITY. Choa : River Awash, below Koka Dam, (50). Shell (pi. 2, i) The shell is moderately shiny ; many specimens have a light-coloured band a short distance from the outer lip. Fine transverse corrugations extend across the whorl, and are interspersed with short ones which are arranged in spiral rows on some parts of the surface. Anatomy (Text-figs. 6-9) The distribution of pigment in the mantle (Text-fig. 6) forms light spots on a grey background, with small dark spots that are most numerous near the mantle edge. From beneath the anterior mantle edge projects a narrow flap that continues around the visceral hump with long finger-like projections from the posterior part. The ovotestis consists of many acini scattered in the digestive gland. There are numerous blunt projections from the seminal vesicle region of the hermaphrodite duct (Text-fig. 7). Three regions may be distinguished in the female duct (Duncan, 1958) ; the proximal oviduct i is straight and narrow, and this leads to the junction between a highly convoluted oviduct 2 and the wide oviduct 3 which terminates in a short vagina. After leaving the junction with the hermaphrodite duct the vas deferens receives numerous prostatic diverticula, each of which is branched several times. A blunt papilla projects into the lumen of the preputium from the junction between the penis sheath and the preputium (Text-fig. 8) ; the tip of the penis usually reaches as far as the tip of the papilla, but the penis is considerably shorter in a few specimens. A preputial gland may be seen as a lightly pigmented oval area on the outer wall of the preputium, on the opposite side to the insertion of the retractor muscle. Internally the gland bulges into the lumen of the preputium, filling it almost completely. The junction between the male and female systems (Text-fig. 9) is similar to that described by Duncan (loc. cit.) for Physa fontinalis. The vas deferens (vd) arises from the hermaphrodite duct (hd) before this duct enters the albumen gland, but the junction between the hermaphrodite duct, oviduct i, and the lumen of the albumen gland is surrounded by albumen gland tissue. Two fertilisation pockets project from the base of oviduct i. REMARKS. In both shell and anatomy the Ethiopian Physa resemble the southern European species P. acuta Draparnaud as described by Slugocka (1913) ; similar Physa have been examined (W.H.O.) from Kenya and Southern Rhodesia. These shells also closely resemble P. borbonica Ferussac originally described from Mauritius. The elucidation of the relationships of the Physa in East Africa is a matter for future investigation. DISTRIBUTION. Physa has not been reported previously from Ethiopia, but its occurrence might be expected as the genus is present in Egypt and Kenya. Cases of introduction of species of Physa in other parts of the world are known, and as Physa sp. was abundant in the single locality where it was found, it is likely that the D. S. BROWN snail has been recently established in Ethiopia and will become more widely distributed in the future. pr 2 mm 2 mm rm ov 1 1 mm FIGS. 6-9. Physa sp. from below Koka Dam. Fig. 6. Dorsal view of whole animal with visceral hump removed, pigmentation shown on anterior part of mantle only. Fig. 7. Distal genital organs. Fig. 8. Copulatory organ, penis and papilla shown by dotted lines and wall of preputium opened to show preputial gland. Fig. 9. Junction of male and female systems, position of albumen gland shown by dotted lines. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 49 Family LYMNAEIDAE LYMNAEA Lamarck Lymnaea truncatula (Miiller) Buccinum truncatulum Miiller, 1774 : 130 ; Limnaea truncatula? ; Jickeli, 1874 ; Bourguignat, 1883 ; Pollonera, 1898 ; Germain, 1904 ; Connolly, 1928 ; Lymnaea truncatula ; Wright & Brown, 1962 ; Galba truncatula ; Piersanti, 1941. Limnaea peregra? ; Jickeli, 1874. LOCALITIES. Arussi : 50 km. W. of Shashamanne on Soddu road, (4). Begemeder : 19-97 km. N. of Gondar on Asmara road, 6 samples, (100) . Eritrea : 19 km. E. of Teramni on Decamere road, (i) ; 66 and 101 km. S. of Asmara on Addis Ababa road, (6) ; River Toquor near Mekerka, (i). Kaffa : 5 and 9 km. W. of Kombi, (2). Choa : 2-150 km. N. of Addis Ababa boundary on Debra Markos road, 5 samples, (15) ; Lake Wonji, (2) ; 29 km. W. of Addis Ababa on Lekemti road, (4). Wollo : Kombolchia village on Dessie road, (2) ; 8 km. N. of Dessie on Asmara road, (5). Harar : 71 km. on Carsa road from junction with Dire Dawa- Harar road, (20) . Shell (pi. 2,5) There is a close resemblance to specimens from near Debra Markos described by Wright & Brown (1962). Largest shell : L = 11-7 mm. ; ML = 6-2 mm ; W =6-0 mm. Several speci- mens in this sample are conspicuously malleated (pi. 2, 5). Radula In the 12 radulae examined the endocones of the first few lateral teeth are relatively smaller than in the lateral tooth illustrated by Wright & Brown (loc. cit.). In some lateral teeth the meso- and ectocones are elongated giving the teeth a bicuspid appearance which resembles the condition in L. truncatula from Aden (Wright, I963b), though the cusps are shorter in the Ethiopian specimens. REMARKS. The specimen of L. peregra? from the River Toquor near Mekerka illustrated by Jickeli (1874) closely resembles L. truncatula which was collected from the same locality during the recent expedition. However it is possible that living populations of European Lymnaea other than L. truncatula will be discovered in Ethiopia. DISTRIBUTION. The majority of previous records of L. truncatula are from northern Ethiopia, but it has also been found in a sub-fossil condition as far south as the Nargi Plain, near Lake Stephanie (Piersanti, 1941). The present records suggest that this species is widely distributed at moderately high altitudes on both sides of the Rift Valley, and living populations may also occur at lower altitudes as is the case in Tanganyika (Mandahl-Barth, pers. comm.). Lymnaea natalensis (Krauss) Limnaeus natalensis Krauss, 1848 : 15, pi. 5, fig. 15 : Blanford, 1870 ; Lymnaea natalensis ; Mandahl-Barth in Ayad, 1956 ; Wright & Brown, 1962. Limnaeus natalensis exsertus Martens, 1866 : 101, pi. 3, figs. 8 & 9 ; Limnaea natalensis var. exserta ; Jickeli, 1874 ; Limnaea exserta ; Bourguignat, 1883 ; Pollonera, 1898. 50 D. S. BROWN Limnaea orophila Morelet, 1868 : 87, pi. 7, fig. 4 ; Limnaea natalensis var. orophila Jickeli, 1874. Limnaea caillaudi Bourguignat, 1883 : 89, pi. 10, figs. 100 & 101 ; Pollonera, 1898 ; Connolly, 1928 ; Radix caillaudi ; Bacci 1941. Limnaea acroxa Bourguignat, 1883 : 90, pi. 10, fig. 94. Limnaea alexandrina Bourguignat, 1883 : 92, pi. 10, figs. 95 & 96 ; De Rochebrune & Germain, 1904 ; Limnaea exserta var. alexandrina ; Pollonera, 1898. Limnaea raffrayi Bourguignat, 1883 : 93, pi. 10, figs. 97 & 98 ; Pollonera, 1898. Limnaea aethiopica Bourguignat, 1883 : 94, pi. 10, figs. 92 & 93 ; Neuville & Anthony, 1908. Limnaea africana " Riippell " Bourguignat, 1883 : 95, pi. 10, fig. 99 ; Neuville & Anthony, 1908. Limnaea graveiri Bourguignat, 1885 : 23, fig. 6. Limnaea soleilleti Bourguignat, 1885 : 24, fig. 7. ? Radix pereger ; Piersanti, 1940. LOCALITIES. Arussi : Lake Zwai, (n). Begemeder : 308 km. S. of Asmara on Gondar road, (15) ; Lake Tana at Gorgora, (6). Choa : Tributaries of the River Awash at Akaki and Dukem, (12) ; River Awash below Koka Dam, (12); 22-318 km. N. of Addis Ababa on Asmara road, 4 samples, (75) ; 21 km. E. of Wolisso, (10). Eritrea : 22-83 km. S. of Asmara on Gondar road, 4 samples, (90) ; 15 km. approx. N.W. of Asmara on Cheren road, 2 samples, (30) ; River Toquor near Mekerka, (60) ; 3 km. E. of Teramni on Decamere road, (5) ; 66 and 101 km. S. of Asmara on Addis Ababa road, (30). Harar : 10 km. W. of Harar on Dire Dawa road, (30) ; Lake Aramyia, (50) ; 43 km. on Carsa road from junction with Dire Dawa-Harar road, (30). Kaffa : 26 km. W. of Sokuru, (6) ; 9 and 22 km. W. of Assendabo, (60) ; 13 and 14 km. W. of Jimma on Bonga road, (18) ; 34 and 38 km. N. of Jimma on Agarro road, (27) ; 4 km. E. of Jimma on Addis Ababa road, (24). Sidamo : Lake Margherita, (40) ; Neghelli, (40). Tigre : 0-5 km. W. of Adowa on Gondar road, (4) ; 222 km. S. of Asmara on Addis Ababa road, (5). Wollo : 26 km. S. and 0-5 km. N. of Kombolchia on Dessie road, (20) ; i km. W. of Batei, (23) ; i and 30 km. N. of Dessie on Asmara road, (80). Shell (pi. i, 10 and n) Five of the largest shells from all the samples were measured ; the extreme values of the mean ratio L/ML are 1-27 and 1-51, and the values for individual shells range between 1-23 and 1-54. Largest shell (northern Choa) : L = 24-7 mm. ; ML = 18-7 mm. ; W = 14-5 mm. Although there is a wide variation in the shape of the spire and the aperture through the whole series of samples, many of the samples have a distinctive appearance. A relatively high proportion of long-spired samples were obtained from Kaffa and Harar Provinces. Spiral rows of corrugations and depressions are present in many samples, but the extent to which this microsculpture covers each shell varies, and it may be confined to the apex only. Microsculpture is lacking from many of the shells collected in Harar, and in none of them is it well developed. Anatomy Both the intensity and arrangement of mantle pigmentation is very variable. A few unpigmented spots may be surrounded by a dense dark background, or at the other extreme pigmentation may be reduced to a fine network. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 51 The genital system resembles that of specimens from Go j jam Province, Wright & Brown (1962). Some specimens from every sample were dissected, and in all of them the distal part of the prostate gland was swollen and a large internal fold was present. Radula The endocones and mesocones of the lateral teeth of several radulae (14 radulae from 3 samples examined) are more widely separated than in the teeth illustrated by Wright & Brown (loc. cit.). REMARKS. Haas (1936) recognised 6 subspecies of L. natalensis in Africa, but Hubendick (1951) found no clear geographic pattern in the distribution of variation in shell and anatomical characteristics. However, Mandahl-Barth (1954) divided L. natalensis into 3 subspecies, of which caillaudi included certain E. African forms and all the species described from Ethiopia by Bourguignat. Specimens from the Victoria Nile, which are distinguished from L. natalensis by characteristics of the shell and prostate gland, are identified by the same author as L. exserta Martens, which was originally described from the highlands of Ethiopia. According to the proportions given by Mandahl-Barth the approximate values of the ratio L/ML vary between 1-13 (L. natalensis nyansae] and 1-50 (L. exserta}. Many of the specimens in the present collection have the proportions of L. natalensis caillaudi, but as a con- siderable number conform to the subspecies natalensis natalensis and others to L. exserta, the material is not assigned to a subspecies. None of the Ethiopian specimens dissected has a prostate gland like that described for exserta from the Victorian Nile by Mandahl-Barth, and the identity of the Uganda form with Martens' species must be considered doubtful. All the shells of Lymnaea obtained from Lake Tana, the type locality of L. caillaudi, closely resemble Bourguignat's description of this species and they are distinctive (pi. 2, n) in com- parison to all the other samples. Radix pereger recorded by Piersanti (1940) appears to represent a small shell of L. natalensis. DISTRIBUTION. As is the case in many other parts of Africa L. natalensis is widely distributed and abundant in Ethiopia, occurring in a variety of habitats including shallow fast-flowing streams and lakes. Family PLANORBIDAE Subfamily BULININAE BULINUS Miiller Bulinus truncatus species group Buli nus truncatus sericinus (Jickeli). Bulinus truncatus (Audouin, 1827) sensu Connolly, 1928 : 177. Physa contorta Michaud, 1829 sensu Martens, 1869 : 99 ; Blanford, 1870 : 476 ; Bourguignat, 1883 : 126 ; Isidora contorta ; Pollonera, 1898 : 12 ; Neuville & Anthony, 1906 : 274, fig. 9. Isidora brocchii Ehrenberg, 1831 sensu Pollonera, 1898 : 12. Physa natalensis? Kuster, 1841 sensu Martens, 1869 : 214 ; Bulinus natalensis ; Connolly, 52 D. S. BROWN 1928 : 175, figs. 3 & 4 ; Bacci, 1940 : 455 ; Piersanti, 1941 : 276, fig. 40 ; Physa natalica ; Bourguignat, 1883 : 126. Bulinus (Diastropha) raymondianus var. porrecta (Martens, 1874) sensu Piersanti, 1941 : 275, fig. 39- Isidora sericina Jickeli, 1874 : 194, pi. 7, fig. n ; Bulinus (Bulinus) sericinus ; Mandahl-Barth, 19575 ; Bulinus sericinus ; Wright & Brown, 1962. Isidora schackoi Jickeli, 1874 : 197, pi. 7, fig. 12 ; Bulinus schackoi ; Connolly, 1928 ; Isidora schackoi mut. minima Pollonera, 1898 : 12. Physa coulboisi ? Bourguignat, 1888 sensu Neuville & Anthony, 1908 : 266, fig. 5. Physopsis africana (non Krauss, 1848) Neuville & Anthony, 1908 : 266, fig. 5. LOCALITIES. Arussi : Lake Zwai, (150) ; 50 km. approx. from Shashamanne on Soddu road, 2 samples, (100) ; River Awash, S. of Nazareth, 2 samples, (16). Begemeder : 38 km. N. of Gondar on Asmara road, (100) ; 66 km., 77 km., 97 km. and 116 km. N. of Gondar, (50) ; 279 km. N. of Gondar, (4) ; Lake Tana at Gorgora, (2). Choa : Lake Biete Mengest, (150) ; Lake Hora, (150) ; Lake Bishoftu, (20) ; swamp N. of Debra Zeit town, (100) ; Akaki River at Akaki, (40) ; Moggio River at Moggio, (5) ; River Awash on Addis Ababa-Jimma road, (2) ; 21 km. E. of Wolisso on Jimma road, (10) ; 10 km. W. of Ambo on Lekemti road, 2 samples, (20) ; Lake Wonji, (20) ; 16 km.-i85 km. N. of Addis Ababa on Debra Markos road, 8 samples, (350) ; 67 km. and 130 km. N. of Addis Ababa on Asmara road, (40) ; 12 km. N. of Debra Sina, (2) ; 96 km. N. of Robi, (6). Eritrea : 22 km.-83 km. S. of Asmara on Gondar road, 4 samples, (90) ; 3 km.-38 km. N.W. of Asmara on Cheren road, 7 samples, (120) ; 66 km. S. of Asmara on Addis Ababa road, (20) ; River Toquor near Mekerka, (37). Harar : Lake Aramyia, (300) ; 17 km. (Langhei village) and 43 km. on Carsa road from junction with Dire Dawa-Harar road, (350). Kaffa : 26 km. approx. W. of Sokuru on Jimma road, (2) ; 22 km. W. of Assendabo on Jimma road, (25) ; 4 km. E. of Jimma on Addis Ababa road ; (20) ; 4 km. W. of Jimma on Bonga road, (40) ; 38 km. N. of Jimma on Agarro road, (2). Sidamo : Lake Margherita, east shore, (40). Tigre : 167 km. and 200 km. S. of Asmara on Addis Ababa road, (60) ; 341 km. S. of Asmara, (2) ; Lake Ashangi, N.E. shore, (100). Wollo : i km. W. of Batei, N. of Kombolchia road, (30) ; i km. within N. boundary of Dessie on Asmara road, (20) ; 8 km. and 30 km. N. of Dessie on Asmara road, (60) ; Lake Haik, S. shore, (130). Shell (Text-fig. 10) Because of the great variation in shell shape it is necessary to consider separately several more or less distinct forms. In many of the populations occurring on the FIG. 10. Bulinus truncatus sericinus and Bulinus sp. (truncatus group), shell outlines drawn by camera lucida. a-t, B. truncatus sericinus. a-c, Choa, 72 km. N. of Addis Ababa on Debra Markos road, d, Eritrea, reservoir 4 km. N.W. of Asmara on Cheren road, e, Begemeder, 38 km. N. of Gondar on Asmara road, f, Choa, 12 km. N. of Debra Sina on Asmara road, g-i, Eritrea, River Toquor near Mekerka. j, Eritrea, 38 km. N.W. Asmara on Cheren road, k, m, Wallo, Lake Ashangi. 1, Sidamo, Lake Margherita. n, o, Arussi, 5 km. S. of Adamitullo. p, t, Choa, Lake Hora. q, r, Choa, Lake Biete Mengest. s, Choa, swamp N.W. of Debra Zeit town, u-w, Arussi, Bulinus sp. from Lake Awasa. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 53 W 54 D. S. BROWN plateau the shell is narrower than is usual in other parts of the range of B. truncatus in N.E. Africa, and in some populations the spire is unusually elongated (Text-fig. IDC and e). Further characteristics, developed to a variable extent, are the relatively great length of the ultimate whorl, and the somewhat shouldered appearance of the apical whorls. Some samples contain a wide range of shell shapes (Text-fig. 10 a-c, g-i), but others are uniform. The kind of shell described above is common at moderately high altitudes in Ethiopia, but in the same regions populations with small, widely umbilicate, short-spired shells occur (d and j), and also a few populations reaching a relatively large maximum size (f). Maximum individual size varies widely between populations, generally the largest shells do not exceed 15 mm., and many populations appear to be full grown at less than 10 mm. Exceptionally large specimens were present in a reservoir near Asmara. Largest shell : L = 21-5 mm. ; ML = 12-9 mm. ; W = n.8 mm. ; 5 whorls. Transverse ribs, arranged regularly on the upper whorls, are present in at least some shells in every sample, and are overlain by flaps of periostracum in many speci- mens. A faint spiral sculpture is occasionally present. The columella is slightly twisted in some samples. Four samples of particular interest from the plateau region are described below. 1. Eritrea : River Toquor near Mekerka (Text-fig. 10, h-j). Type locality of Isidora schackoi and /. sericina Jickeli, 1874. Text-figure 10 illustrates the variation in the length of the spire. Largest shell : L = 10-7 mm. ; ML = 7-3 mm. ; W = 6-6 mm. ; 4 whorls. Ratio L/ML ranges between 1-15 and 1-50 (37 shells). 2. Begemeder : 38 km. N. of Gondar (pi. 2, 2 and Text-fig. loe). Exceptionally long-spired shells, some of which have the surface of the ultimate whorl conspicuously malleated. Largest shell : L = 13-3 mm. ; ML = 8-0 mm. ; W = 6-8 mm. ; 4$ whorls. Ratio L/ML ranges between 1-51 and 1-80 (20 shells). 3. Tigre : Lake Ashangi (Text-fig. lorn). Relatively short-spired shells with rounded apertures. Fine, irregular transverse ribs are present, and faint spiral sculpture is present on the apex. 4. Wollo : Lake Haik. An exceptionally small form. Largest shell : L = 8-4 mm. ; ML = 6-5 mm ; W = 5-5 mm. ; 3^-3^ whorls, (sample of approx. 100). Populations that are similar to some of those of the plateau were found at lower altitudes south of Addis Ababa : between Addis Ababa and Debra Zeit ; between Shashamanne and Soddu, and in the River Awash south of Nazareth. These localities, with the exception of the River Awash, are elevated in relation to the floor of the Rift Valley. The populations occurring in the Bishoftu Crater Lakes at Debra Zeit, and in the Rift Valley Lakes, are distinctive and are described below. 5. Choa : Bishoftu Crater Lakes (Text-fig. 10, p-r and t). Collections were made in Lakes Biete Mengest, Hora, and Bishoftu (see Mohr, 1961 for terminology). All the shells are relatively wide, with a more or less evenly rounded aperture that has an obtuse angle above. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 55 Largest shell : L = 14-2 mm. ; ML = 9-8 mm. ; W = 10-5 mm. ; 4 + whorls, (L. Biete Mengest). Ribs are well developed in both the shell and the periostracum, and spiral sculpture is exceptionally well developed in the Lake Hora population. In a nearby swamp (W. of Debra Zeit town), the shells have longer spires and narrower apertures (Text-fig. IQS). 6. Lakes Zwai and Margherita (Text-fig. lok and 1). The shells have relatively large apertures compared to those of the plateau populations. Anatomy The mantle and genital organs of B. truncatus sericinus from near Debra Markos were described by Wright & Brown (1962). Mantle pigmentation, like shell shape, is often characteristic of a particular population, but the value of pigmentation as a taxonomic character is reduced by the great variation that may be observed in samples from a small area. No characteristics in the gross morphology of the genital organs have been observed that can be correlated with the various shell shapes in the present collection, but it is likely that the size at which sexual maturity is reached in each population is related to the maximum size achieved in that habitat. Wholly aphallic specimens (which were not present in the samples from near Debra Markos) are present in several of the samples, but the aphallic condition does not appear to be correlated with either shell shape or geographical distribution. Of 20 samples examined, 7 contain aphallic individuals. The proportions of aphallic specimens found range between 8 per cent. (Lake Bishoftu, 25 specimens), 92 per cent. (Lake Zwai, 25 specimens), and 100 per cent. (Lake Margherita, 10 specimens). Radula (Text-figs. 19-21) More or less distinctive characteristics are present in each sample of radulae that has been examined, and it is likely that a detailed analysis would reveal a variation comparable to that observed in the shell. The central tooth usually bears two cusps, but in one radula the entire row of centrals bears a single median cusp. The mesocone of the lateral teeth is occasionally bluntly pointed, but generally approaches an arrow-head shape which is very obvious in the Lake Ashangi specimens (Text-fig. 21). One or two interstitial cusps (Wright & Brown, loc. cit.) may be present on the lateral teeth. Transition between the lateral and marginal teeth usually takes place between teeth 6 and 8 with the development of additional cusps between the endo- and the mesocone, and lateral to the ectocone (Text-figs. 19, 20). The meso- cone can be recognised in many of the marginal teeth, but up to nine more or less similar cusps may be present on the median side of these teeth, and a variable number of small cusps is present on the lateral side. Folds or ridges, which appear to correspond to the " fluting " described by Stiglingh et al (1962), are present on the endo- and mesocones of the lateral teeth of many specimens. Folding is particularly well developed on the cusps of the lateral teeth of specimens from 38 km. N. of Gondar, which have broadly based endocones of a characteristic shape (Text-fig. 19). REMARKS. Apart from locally described species and varieties, and a single shell identified by Neuville & Anthony as B. coulboisi, shells of Bulinus (Bulinus) from 56 D. S. BROWN Ethiopia have been identified as either Egyptian or South African species by previous authors. Mandahl-Barth (1960) and Wright & Brown (1962) regard the locally described forms as synonymous with B. sericinus (Jickeli), belonging within the B. truncatus group. The present collection provides information about the variation of the Bulinus truncatus group which is of value in the interpretation of the inter- relationships between the Ethiopian forms and their relationship to the Bulinus of other parts of Africa. Only 37 moderately sized specimens were collected in the type locality of B. sericinus and B. schackoi, but the series contains sufficient variation in spire height to encourage the view that elongated shells of the schackoi type may be considered as part of the continuous variation of a population of B. sericinus (Wright & Brown, loc. cit.). However, although some other samples in the present collection contain a more or less complete intergradation between the sericinus and schackoi types of shell, many large samples contain a comparatively restricted range of variation. Sample 2 described above consists almost entirely of long-spired schackoi shells, of which the range of variation of L/ML scarcely overlaps that of any other popula- tions. Such distinctive populations may be produced either by genetic characteris- tics of the snails, or ecophenotypic effects of the habitat, or both factors acting together. The occurrence of small schackoi snails together with small sericinus in the same habitat, where they have presumably experienced the same environmental influences, is evidence that there is some genetic difference between the forms. It has long been appreciated that there is a close resemblance between the shells of some Ethiopian Bulinus, and those of B. natalensis (Krauss) of South Africa. Specimens from northern Ethiopia have been identified as natalensis (Martens, 1869 and Bourguignat, 1883), as have examples from in or near the Rift Valley Lakes (Connolly, 1928, Bacci 1940, Piersanti, 1941). Connolly divides specimens from the Bishoftu Crater Lakes (Bishoftu and Hora Harsadi) and Lake Zwai into two forms, one of them probably conspecific with natalensis and the other somewhat different, which occurred together in one lake (Bishoftu). Specimens similar to those described by Connolly were collected during the recent expedition in Lakes Zwai, Bishoftu, and Biete Mengest (=Hora Harsadi of Connolly). In addition, a further population has been found (Lake Hora, see 5 above) with exceptionally well developed spiral sculpture. All the lateral radula teeth examined in Ethiopian specimens have arrow-head shaped cusps, in contrast to the South African Bulinus (Bulinus) which are placed in the Bulinus tropicus group (including natalensis Krauss), having sharply pointed cusps (Mandahl-Barth, I957b, and 1962). How- ever, a close relationship between Ethiopian and South African forms is not precluded, as a form of B. tropicus from Transvaal which possesses arrow-head shaped cusps has been described by Stiglingh et al. (1962) which is considered by Mandahl-Barth (pers. comm.) to represent B. natalensis (Krauss) and to belong within the truncatus group. It is thus a matter for conjecture whether certain populations of Bulinus (Bulinus} occurring in Ethiopian lakes are related most closely to South African Bulinus with arrow-head shaped cusps, to populations of the B. truncatus group occurring FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 57 at relatively low levels in N.E. Africa, or to B. truncatus sericinus of the Ethiopian plateau. Dr. J. B. Burch (in. litt.) has found differences in chromosome number between several populations of the B. truncatus group in N.E. Africa (see under Bulinus sp. below), and it is to be hoped that such studies will be of value in elucidat- ing the relationships discussed above. At present, the inclusion of the varied forms in B. truncatus sericinus is provisional. DISTRIBUTION. B. truncatus sericinus is widely distributed in Ethiopia above an altitude of about 1,500 m. (5,000 ft.), and has been found at 2,945 m. (9,600 ft.) (38 km. N. of Gondar). This subspecies is regarded as including Bulinus from Western Aden Protectorate (Wright, 1963^, but the records of B. sericinus from other parts of Africa probably refer to other forms of Bulinus. Bulinus sp. LOCALITIES. Arussi : 5 km. approx. S. of Adamitullo (between Lake Zwai and Bulbulla River) (6)*. Sidamo : Lake Awasa (400). Shell (pi. 2, 4 and Text-figs. lou-w) Most of the shell length is contributed by the ultimate whorl, the apical whorls forming a small sharp spire. The aperture is large, more or less evenly rounded, and has an obtuse upper angle that is frequently more than 90 degrees. As a result of the steep curvature of the upper part of the whorls the suture is deep. Fine transverse ribs are present, but there are no free flaps of periostracum. Spiral sculpture is well developed on the apices of many shells, and is occasionally present on some parts of the ultimate whorl. Fresh shells are pale straw-coloured or almost white. Largest shell : L=i3-8 mm.; ML=io-5 mm.; W=io-4 mm.; four and three- eighths whorls. Anatomy Out of 25 specimens dissected, 9 are aphallic and 6 lack penis sheaths. Radula (Text-fig. 22) The lateral teeth have arrow-head shaped cusps but the radula teeth appear to be significantly smaller than in specimens of the same size from plateau populations of B. truncatus sericinus. REMARKS. Living snails have a characteristic blue-grey bloom ; approximately 10 per cent, of those collected lacked dark pigment and appeared to be albinos. Dr. J. B. Burch (1964) finds that the haploid chromosome number of specimens collected in Lake Awasa is 18, in contrast to 36 in B. truncatus from Egypt and Iran, and 72 in laboratory bred stock of B. truncatus sericinus collected by Dr. C. A. Wright in Western Aden Protectorate. Burch (1960) has also reported a haploid number of 18 in the Bulinus tropicus group. The low chromosome number of the Lake Awasa Bulinus, and the similarity of the shells to sub-fossil shells from the vicinity of Lake Zwai (Text-fig. 10), are facts that in conjunction suggest that the form may be an ancient one now restricted in range. 58 D. S. BROWN Bulinus africanus species group B. africanus ovoideus (Bourguignat) Physopsis ovoide aBourguignat, 1879 : 16 ; B. (P.) africanus ovoideus ; Mandahl-Barth, 1957 : pis. 17 & 18. LOCALITIES. Begemeder : 14 km. N. of Medhanie Alem on Asmara road (27) ; Lake Tana at Gorgora (21) ; 8 km. N. of Gorgora (5). Kaffa : 44 km. E. of Jimma town boundary on Addis Ababa road (80). Shell The outstanding characteristics of each sample are described below. 1. Medhanie Alem (pi. 2, 9, 12). The spire is relatively short and sharp and the surface of the ultimate whorl is evenly curved with no trace of a shoulder. Spiral microsculpture is very well developed on the apex, and extends nearly to the aperture in several specimens. Where the microsculpture extends to the lowest whorl, the nodules become elongated and scarcely elevated from the surface, giving the appearance of rows of striae. Largest shell : L= 17-7 mm. ; W = i3-imm.; ML=n-4mm.; 4^-5 whorls. 2. Lake Tana (pi. 2, 8, u). The spire is relatively longer than in the preceding sample but the apical whorls are less exserted and not sharply pointed. The whorls are flattened near the suture so that a blunt shoulder is formed. Spiral microsculpture similar to that described above is present, but is less distinct. All the larger shells have a thickened ridge at approximately the same distance from the aperture, indicating that growth has been temporarily interrupted. Largest shell : L^n-ymm.; ML— 9-1 mm. ; W=7>8 mm. ; 4! whorls. 3. Kaffa (pi. 2, 13-15). The larger specimens are narrow with rounded apical whorls and comparatively long spires. The shell colour is not brown as in the preceding samples but whitish to pale horn. The aperture lengths and shell widths are smaller than in Medhanie Alem shells of the same length over the whole of the size range, and more whorls have been completed at the same length. Micro- sculpture similar to that of the preceding samples is present. At, or near, the aperture of all but the smallest shells is a light coloured band which marks a temporary cessation of growth. Largest shell : L = 16-2 mm. ; ML — 11-4 mm. ; W = 9-3 mm. ; 5 whorls. Anatomy In all the samples the mantle is similarly pigmented with distinct dark blotches (Text-fig. 18). The penis sheath is longer and broader than the preputium, the range of PS/PP lying between 1-07 and 1-73 (20 specimens dissected). Radula (Text-fig. 24) Transition between the lateral and the marginal teeth is marked by the develop- ment of a small cusp between the endocone and the mesocone, which usually takes place at the loth tooth (12 radulae examined). REMARKS. The shells collected on the north shore of Lake Tana show some resemblance to Bulinus globosus. Similar shells collected by Dr. N. Ayad near Bahr Dahr on the south shore were identified as this species by Mandahl-Barth (in FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 59 Ayad, 1956). However, the copulatory organs of the specimens from Gorgora are similar to those from near Medhanie Alem, the shells of which resemble B. africanus ovoideus, and it is probable that this species is present on both the north and south shores of the lake. With regard to the large number of apparently suitable habitats that were searched in which no B. africanus were found, it may be concluded that populations are widely separated. This kind of distribution would favour the development of distinct local characteristics as described above in the case of the shell. DISTRIBUTION. The present records from Begemeder appear to be the most northerly for B. africanus in East Africa. B. africanus ovoideus appears to have a wide range on the Ethiopian plateau but is not abundant ; it is widely distributed in Kenya and Uganda and extends into the Congo and Tanganyika (Mandahl-Barth, Bulinus ugandae Mandahl-Barth Bulinus (Physopsis) globosus ugandae Mandahl-Barth, 1954 : I:[4» fig3- 55a~h '• B. (P.) ugandae Mandahl-Barth, 1957. LOCALITY. Sidamo : Lake Margherita, east shore (16). Shell (pi. 2, 6, 7 and 10) The majority are moderately high-spired, with rounded whorls that are more or less shouldered near the suture. Largest shell : L = 14-7 mm. ; ML = 10-9 mm. ; W = 11-3 mm. ; 4^ whorls. The upper angle of the aperture is obtuse and frequently a right-angle ; the inner edge of the aperture is folded narrowly over the umbilicus which is open. Truncation of the columella is so slight in many specimens that it resembles the oblique twisting occasionally present in Bulinus (Bulinus). Irregular transverse ridges of varying size are scattered over the whole shell surface, and the second whorl of some specimens bears spiral ridges which cross the transverse ridges forming a poorly denned network of elevations and pits. In some specimens there is a very fine spiral microsculpture over the rest of the shell surface formed by rows of shallow ridges and furrows but never consisting of well defined nodules as in B. africanus ovoideus. Anatomy (Text-fig. 17) In contrast to B. (P.) africanus ovoideus the mantle lacks definite patches of pigment and is evenly grey in appearance (Text fig. 17). On the inner surface there is a long fold between the rectum and the kidney extending as far back as the posterior end of the kidney, and a large fold along the whole length of the surface of the kidney. As all the copulatory organs are everted the length of the penis sheath can only be estimated as equal to or slightly longer than the preputium. Radula (Text-fig. 23) Transition between the lateral and the marginal teeth takes place earlier than in the preceding species, at about the 8th or gih tooth (4 radulae examined). REMARKS. Although columellar truncation and spiral microsculpture are almost absent in several shells, the presence of a long renal fold confirms that these specimens 60 D. S. BROWN belong to the Bulinus africanus species group. There is a close resemblance to shells of B. ugandae from Kosti (Sudan) and Kisumu (Kenya) figured by Mandahl-Barth (i957b, figs. 4od and 4ib), and it is not surprising that the range of this species should extend into south-west Ethiopia. Comparison with B. abyssinicus (Martens) from Somalia (Mandahl-Barth, iQ57c) shows that in general appearance the Lake Margherita specimens are closer to B. ugandae. However, as the original description states that spiral microsculpture is completely absent from the apex of B. ugandae, the presence of some spiral microsculpture in the Lake Margherita shells might be taken to indicate a close relationship with B. abyssinicus. This is not an important point of resemblance though, because as ugandae was originally described as a subspecies of B. globosus they are probably closely related, and the presence of slight spiral sculpture in ugandae is not surprising as it is frequently well developed in globosus. DISTRIBUTION. It is likely that B. ugandae occurs elsewhere in Ethiopia but the species is probably rare as only a single colony was found during several days collecting on the shore of Lake Margherita. Five specimens with eroded apices collected in Lake Stephanie and identified by Piersanti (1941) as B. (P.) africanus may represent B. ugandae. The range of this species is confined to north-east Africa including Uganda, Tanganyika, Kenya and Sudan. Bulinus forskali species group Bulinus forskali (Ehrenberg) Isidoraforskalii Ehrenberg, 1831 : 20 ; Jickeli, 1874 ; Pollonera, 1898 ; Neuville & Anthony, 1908 ; Physa forskalii ; Martens, 1869 ; Morelet, 1872 ; Bourguignat, 1883 ; Neuville & Anthony, 1906 ; Pyrgophysa forskalii ; Neuville & Anthony, 1905 ; Bulinus forskali ; Connolly, 1928 ; Haas, 1932 ; Bulinus (Pyrgophysa) forskalii ; Bacci, 1941 ; Bulinus (Bulinus} forskalii '. Mandahl-Barth, 19575 & 1960. Physa forskalii var. elatior Martens, 1866 : 100. Bulinus (Isidora) forskali var. cylindrica Piersanti, 1941 : 275, fig. 38. Pyrgophysa scalaris (non Dunker, 1845) Neuville & Anthony, 1905. LOCALITIES. Arussi : Lake Zwai, N.W. shore (15) ; i km. S. of River Awash on Nazareth-Asella road (15). Begemeder : Lake Tana at Gorgora (5). Sidamo : Lake Margherita, east shore (50) ; army barracks at Neghelli (2) ; artificial pool 10 km. approx. S. of Uachile Wells (50). Wollo : 62 km. S. of Kombolchia on Addis Ababa road (20). Shell (pi. 3, i and 2) Few of the shells in any sample exceed 8-5 mm. in length, at which size about 5^ whorls have been completed (Text-fig. 27). Largest shell : L = 10-6 mm. ; ML = 4-3 mm. ; W = 3-5 mm. ; 6| whorls (Lake Tana). The aperture occupies less than half of the total length of the shell length in specimens of more than 3 mm. shell length (Text-figs. 25 and 26). The shoulder on the upper whorls is acute in many specimens, and frequently bears a distinct carination (pi. 3, 2). Irregular growth ridges of varying size are present, and there are regularly spaced ribs on at least the second and third whorls of most shells. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 61 Anatomy (Text-figs, n and 28) The copulatory organ (Text-figs, na-c and 28) is small in comparison to that of B. scalaris, with a penis sheath that is narrower than the preputium and less than twice its length in most specimens. In the Lake Margherita sample the ratio PSyPP ranges between 0-95 and 2-42 with a mean of 1-57 (14 specimens). REMARKS. Bulinus fischerianus (Bourguignat), the only species of the Bulinus forskali group originally described from Ethiopia, is provisionally placed within the synonymy of the B. scalaris (Dunker) in the present paper (see below). As the only other members of the forskali group described in north-east Africa are B. forskali (Ehrenberg) and two other Egyptian forms belonging within the synonymy of this species (Mandahl-Barth, igsyb), it remains to compare the present Bulinus with B. beccarii (Paladilhe, 1872) which has recently been recorded by Wright (1963^ from the Western Aden Protectorate. Shells of the Ethiopian specimens here identified as B. forskali have the transverse ribs and carination far more strongly developed than do specimens of B. beccarii from Western Aden Protectorate. Wright described differences in the shell and radula between beccarii and the other members of the B. forskali group, and concluded that the part played by beccarii as an intermediate host of Schistosoma haematobium suggested a relationship with B. senegalensis or B. cernicus. There is no evidence that the Ethiopian Bulinus in question act as an intermediate host of S. haematobium, and the waters in which they were found are still in contrast to the slowly flowing water inhabited by beccarii. As no significant differences have been observed between shells of Ethiopian forskali and specimens from Egypt and Kenya, it is concluded that the Ethiopian populations form part of the extensive north-east African range of this species. DISTRIBUTION. B. forskali is recorded from relatively low altitudes in southern Ethiopia by Martens (1866), Haas (1932) and Piersanti (1941), and there are several records in eastern Eritrea also from low altitudes. The species is recorded from Lake Zwai (Connolly, 1928), and from southern Choa (Neuville & Anthony, 1908). The present records go some way towards connecting these scattered areas of distribution, and suggest that the species has a widespread distribution in the south extending at least as far north as the River Awash at an altitude of about 5,000 feet. B. forskali appears to be absent from the vicinity of Addis Ababa and Asmara, but its presence in southern Wollo Province and Lake Tana suggests that there is also an extensive range of scattered distribution on the plateau. B. forskali is widely distributed in Egypt and Africa south of the Sahara, but does not appear to have been recorded at such high altitudes outside Ethiopia. Bulinus scalaris (Dunker) Physa scalaris Dunker, 1845 : 164, 1853 : pi. 2, figs. 5 & 6. Physa fischeriana Bourguignat, 1856 : 146, pi. 2, figs. 1-3 ; Bourguignat, 1883 ; Physa fischeri ; De Rochebrune & Germain, 1904. LOCALITIES. Begemeder : 4 km. N. of Gorgora to 248 km. S. of Asmara, 10 samples (230). Kaffa : 44 km. E. of Jimma town boundary (17). 62 D. S. BROWN 18 63 Shell (pi. 3, 3) Maximum individual size is greater than that of B.forskali in most of the samples. Largest shell : L = 12-5 mm. ; 6 whorls. The spire is relatively less narrow than that of B. forskali, the rate of increase in the number of whorls is slightly slower (Text-fig. 27), and the length of the aperture tends to be longer (Text-figs. 25 and 26). There is no shoulder on the lower whorls, but the upper whorls of many specimens are shouldered, though usually less acutely than in B. forskali. In a few specimens a slight carination is present. Regularly spaced ribs which are less well defined than in B. forskali are present on the upper whorls, and fine spiral sculpture is present in several samples. Anatomy (Text-figs, n, 13 and 28) The copulatory organ (Text-figs. lie and f) is larger than that of B. forskali, with a penis sheath that is wider than the preputium and at least twice its length in the majority of specimens. In a sample from 4 km. N. of Gorgora the ratio PS/PP ranges between 1-68 and 5-00 with a mean of 3-01 (19 specimens). The epiphallus (Text-fig. 13) is not as long as that depicted by Mandahl-Barth (1957^ p. 87), but from the length of the everted penis it is evident that the epiphallus is capable of extending at least twice the length of the sheath. The vagina is everted to some extent in all of the specimens examined ; this condition was first recorded in B. scalaris by Mandahl-Barth (1960) and appears to be characteristic of fixed specimens of this species. REMARKS. Wright (ig63a) points out that shells of B. scalaris (including specimens in the type series) are occasionally somewhat shouldered, and this is also the case in the present collection. It is thus impossible to contrast B. scalaris with B. forskali as having " not the faintest trace of a shoulder angle " (Mandahl-Barth, i957b). However, B. scalaris can readily be distinguished from B. forskali by the shape and size of the copulatory organ, and the shape of the shell. Bulinus fischerianus (Bourguignat) is tentatively included in the synonymy above because of the close resemblance between Bouguignat's figure of this species and many of the shells of B. scalaris in the present collection. DISTRIBUTION. The shell recorded from the River Moggio by Neuville & Anthony (1905) as scalaris and later (1908) figured as forskali, closely resembles the former species, which may thus occur at relatively low altitudes in southern Choa. FIG. ii. Copulatory organs of a B. forskali from Lake Margherita, ; b. B. forskali from Lake Zwai ; c. B. forskali from 62 km. S. of Kombolchia ; d. Bulinus sp. (forskali group) ; e. B. scalaris from 47 km. N. of Medhanie Alem ; f. B. scalaris from 4 km. N. of Gorgora. All specimens between 8-3 and 9-8 mm. shell length. FIG. 12. Bulinus sp. (forskali group), epiphallus seen by transparency through wall of sheath. FIG. 13. B. scalaris from 4 km. N. of Gorgora, wall of sheath dissected away to expose epiphallus. FIG. 14. Ancylus sp. from 15 km. N. of Dilla, pseudobranch. FIG. 15. A ncylus fluviatilis from 45 km. N. of Gondar, pseudobranch. FIG. 16. Burnupia caffra from 59 km. N. of Addis Ababa on Debra Markos road, dorso-lateral view of pseudobranch. FIG. 17. Bulinus ugandae from Lake Margherita, outer surface of mantle. FIG. 18. Bulinus africanus ovoideus, outer surface of mantle. D. S. BROWN 0 • 025 mm FIGS. 19-24. Radula teeth. FIGS. 19-21. Bulinus truncatus sericinus. FIG. 19. 38 km. N. of Gondar, central, i, 6, 7, and 8. FIG. 20. River Toquor near Mekerka, central, i, 6, 7, and 8. FIG. 21. Lake Ashangi, i. FIG. 22. Bulinus sp. (truncatus group) from Lake Awasa, central, i, 10, 23 and 24. FIG. 23. Bulinus ugandae from Lake Margherita, central, i, 9 and 15. FIG. 24. B. africanus ovoideus from 5 km. W. of Assendabo. central, i, 9, 12 and 26. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 65 However, other shells figured by these authors from Goro Gomoto, a nearby locality, resemble forskali more closely, and as populations of forskali are known from the south and east of this area (see records for Arussi), the Neuville & Anthony records are provisionally regarded as representing B. forskali. Two widely separated groups of populations of B. scalar is have been found which may be part of a scattered distribution in small temporary habitats throughout the plateau of Ethiopia. The record of B. fischeri by De Rochebrune & Germain (1904) from the S.E. side of the Rift Valley may also represent this species. Elsewhere in Africa B, scalaris is known to have a scattered distribution in Angola, Congo, N. Rhodesia, Kenya and Uganda. Bulinus sp. LOCALITY. Begemeder : 4 km. N. of Medhanie Alem (approx. 322 km. S. of Asmara on Gondar road) (100). Shell (pi. 3, 4) Similar to that of B. scalaris apart from the relatively deeper suture. Largest shell : L = 12-5 mm. ; ML = 5-4 mm. ; W = 4-9 mm. ; 6| who Is. Anatomy (Text-figs, n and 12) The copulatory organ (Text-fig, n) is intermediate in size between that of B. forskali and B. scalaris but resembles that of forskali in the relationship between the dimensions of the penis sheath and the preputium. The ratio PS/PP varies between 0-75 and 1-57 with a mean of 1-04 (10 specimens) , and the sheath is narrower than the preputium. A relatively short ephiphallus can be, seen through the thin wall of the penis sheath (Text-fig. 12). The vagina is not everted as in B. scalaris. REMARKS. The characteristics of the shell and the copulatory organ are constant in a large number of specimens, but as only a single population was found it can only be provisionally regarded as representing a new form within the B. forskali group. It is an interesting possibility that the population is a hybrid one between scalaris and forskali, as in other parts of their range these species are found together with no signs of interbreeding (Mandahl-Barth, I957b). If this does prove to be a distinct form of Bulinus there is sufficient resemblance between the shell and that figured by Bourguignat to justify the use of the name fischerianus. The type locality of fischerianus was not precisely defined but it probably lies on the northern plateau of Ethiopia or Eritrea as Lake Tana is given by Bacci (1951) as the locality of other material attributed to the same collector ( Verreaux) . Germain (1921) and Haas (1936) regarded Bulinus fischerianus, B. beccarii and B. cernicus as synonyms of B. forskali, while Mandahl-Barth (ig57b) considered cernicus to be a distinct species with fischerianus and beccarii tentatively included as synonyms. As Wright (1963^ has recognised beccarii to be a distinct species in Western Aden Protectorate it is necessary to consider the relationship between beccarii and fischerianus. The status of fischerianus, which is an older name than either cernicus or beccarii was not discussed by Paladilhe who compared beccarii only to Egyptian forms of Bulinus. Direct comparison between the descriptions and 66 3-0 2-5 ML 2-0 1-5 2-20 1-76 1-32 UD 0-88 0-44 25 D. S. BROWN 26 length length 5-0 6-7 8-4 10-0 11-7 27 4- O length 5-0 6-7 8-4 10-0 11-7 • 28 length 3-3 5-0 6-7 8-4 10-0 11-7 29 5-0 6-7 8-3 10-0 11-7 *£' ML 1-72 2-15 2-58 3-01 3-44 3-87 4-30 4-73 5-16 D FIGS. 25-29. Aperture length (ML) plotted against shell length (L),+ =Bulinus forskali from 62 km. S. of Kombolchia ; • =B. scalaris from 4 km. N. of Gorgora. FIG. 26. Aperture length (ML) plotted against shell length (L), + =B. forskali from Lake Margherita ; • =B. scalaris from 47 km. N. of Medhanie Alem. FIG. 27. Number of whorls (Wh) plotted against shell length (L), -f =B. forskali from 62 km. S. of Kombolchia; o=B. forskali from 10 km. S. of Uachile Wells ; • =B. scalaris from 47 km. N. of Medhanie Alem. FIG. 28. Length of copulatory organ (PS + PP) plotted against shell length (L), + =B. forskali from Lake Margherita ; • =B. scalaris from 3 km. N. of Gorgora. FIG. 29. Shell diameter (D) plotted against umbilicus diameter (UD), -f ----Gyvaulus convexiusculus from Tafwa, near Dahla, Western Aden Protectorate, collected and det. C. A. Wright ; • =G. costulatus from i km. S. of Debra Berhan. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 67 figures of fischerianus and beccarii is hampered by the difference in size of the figured specimens, and one important feature distinguishing beccarii, the relatively large lowest whorl could be regarded as merely a juvenile character as the figured specimen of beccarii is only 4 mm. long. However, the manner of growth appears to be significantly different in the two forms, as at a length of 4 mm. beccarii has completed 4-^ whorls, whereas at 8 mm. length, fischerianus consists of only 5 whorls. Subfamily PLANORBINAE PLANORBIS Miiller Planorbis planorbis parenzani Bacci Planorbis (Tropidiscus) planorbis parenzani Bacci, 1940 : 456, pi. i, fig. i. LOCALITY. Arussi : 5 km. approx. S. of Adamitullo (between Lake Zwai and Bulbulla River) (50).* Shell (pi. 3, 15 and 16) Largest specimen : 0=5-3 mm. ; UD = 2-3 mm. ; H = 2-0 mm. ; 4 whorls. REMARKS. These specimens are similar to sub-fossil shells described by Bacci (1940) in the maximum individual size and the rate of widening of the whorl, but differ in being relatively thicker and possessing a less acute basal angulation. This difference is seen most clearly in the aperture which is far more rounded in the present specimens. The chief characteristic distinguishing P. planorbis parenzani from P. planorbis philippi Monterosato is the relatively acute basal angulation in the former. As this characteristic is not well defined in the present shells, which like Bacci's specimens were collected from the plain that formed part of an ancient bed of Lake Zwai, it is doubtful whether the separation of the two forms is justified. DISTRIBUTION. No living specimens have been found of this Planorbis which appears to be closely related to the small forms of Planorbis planorbis that occur in N. Africa and the near East. ANISUS Studer Anisus natalensis (Krauss) species group Anisus natalensis (Krauss) Planorbis natalensis Krauss, 1848 : 83, pi. 5, fig. 9 ; ? Planorbis natalensis ; Blanford, 1870 ; Gyraulus (Anisus) natalensis ; Mandah-Barth in Ayad, 1956 ; Anisus natalensis ; Wright & Brown, 1962. Planorbis abyssinicus Jickeli, 1874 : 215, pi. 7, figs. 2ia-c ; Bourguignat, 1883 ; Pollonera, 1898 ; Germain, 1904 ; Neuville & Anthony, 1908 ; Connolly, 1928 ; Planorbis abyssinicus var. gravieri Germain, 1904 : 353. LOCALITIES. Arussi : Lake Zwai, N.W. shore (30) ; 30-60 km. W. of Shasha- manni on Soddu road, 4 samples (187). Begemeder : 66-97 km. N. of Gondar on Asmara road, 3 samples (220) ; Lake Tana at Gorgora (16). Choa : 22-318 km. N.E. of Addis Ababa on Asmara road, 5 samples (153) ; 32-185 km. W. of Ambo (50); Lake Wonji (15); 21 km. E. of Wolisso (21). Eritrea : 39 km. S. of Asmara on Gondar road (6) ; 8 and 38 km. N.W. of Asmara on Cheren road (145) ; 144 km. S. of Asmara on Addis Ababa road (193) ; 19 km. E. of Teramni on Decamere road (150). Harar: 43 and 46 km. on Carsa road from junction with Dire Dawa — Harar road (60). Kaffa : Between Kombi and Jimma, 4 samples (120) ; 4 km. W. of 68 D. S. BROWN Jimma on Bonga road (2) ; 34 km. N. of Jimma on Agarro road (3). Sidamo Wondo (i) ; 30 and 45 km. south of Adolla (138) ; Lake Margherita (67). Tigre Lake Ashangi, N.E. shore (10) ; 0-5 km. S. of Adowa on Gondar road (i). Wollo Lake Haik, S. shore (50). Shell The majority of specimens resemble those from near Debra Markos described by Wright & Brown (1962), but there are relatively few shells in the present collection with as large and rounded an aperture as the shell illustrated in the earlier paper. Largest shells : D = 7-0 mm. ; UD = 3-3 mm. ; H = 1-8 mm. (Lake Tana at Gorgora). D = 7-0 mm. ; UD = 3-5 mm. ; H = 1-8 mm. (Eritrea, 19 km. E. of Teramni) . Anatomy Anatomical characteristics are similar to those previously described (Wright & Brown, loc. cit.). The anus opens near the tip of the pseudobranch, which is a simple lobe lacking any ridges. The ovotestis consists of up to 30 acini, which are arranged in two irregularly alternating rows. The spermathecal duct is stout, and may be wider than the vagina at the point of junction. There are up to 18 primary prostate gland lobes several of which may be subdivided. In many samples the copulatory organ is similar to that previously described, having the penis sheath slightly longer and about half as wide as the preputium, but in a few samples (e.g. Lake Zwai) the sheath is relatively shorter and narrower. The penis is similar to that of Debra Markos specimens in appearance, but a few specimens have now been observed in which there is some evidence that the cuticularised part of the penis is formed in a manner similar to that described by Hubendick (1958) for Gyraulus. The question of the homology of the terminal part of the penis of A . natalensis with the stylet of Gyraulus species can only be decided when the development of the penis has been studied in detail. REMARKS. Three differing accounts of the anatomy of Anisus natalensis have been recently published (Azevedo et al. 1961, Wright & Brown 1962, and Demian 1962). No specimens from the type locality (Umgeni Valley, Natal) appear to have been studied, but Wright & Brown examined specimens from Transvaal in which the penis was found to have a long cuticularised portion similar to that of Ethiopian specimens. As observations on the copulatory organ indicate that there may be more than one form of Anisus with this kind of penis in Ethiopia, the present specimens are provisionally referred to the A . natalensis species group. It should be noted that the specimen of natalensis illustrated by Krauss is far thinner with a more flattened aperture than in any Ethiopian specimen, or that illustrated by Azevedo et al. from Mozambique. Specimens of A . natalensis from Tanganyika described by Demian possess a very small terminal penial stylet which is different to that of the Ethiopian and Transvaal specimens. A small stylet has also been described in A . misellus Morelet from Angola (Wright, i963a). Another condition is described by Azevedo et al. (loc. cit.) in specimens from Mozambique which have a prostate gland consisting of a single lobe and a short penis with a terminal opening that lacks any stylet. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 69 DISTRIBUTION. Snails of the A. natalensis group are widely distributed in Ethiopia and are often abundant ; a population was found at 2,890 (9,400 ft. approx.) in northern Choa. The presence of a similar penis in South African specimens indicates that this form has a widespread range in Africa. GYRAULUS Charpentier Gyraulus costulatus (Krauss) Planorbis costulatus Krauss, 1848 : 83, pi. 5, fig. 8 ; Planorbis costulatus var. ; Jickeli, 1874 ; Gyraulus costulatus ; Mandahl-Barth in Ayad, 1956 ; Wright & Brown, 1962. Planorbis stelzneri Martens (non Dohrn), 1869 : 212. Planorbis aethiopicus Bourguignat, 1883 : 128 ; Jickeli, 1874 : pi. 7, fig. 23. Caillaudia angulata Bourguignat, 1883 : 129 ; Jickeli, 1874 : pi. 7, fig. 22 ; Piersanti, 1941. Planorbis gibbonsi Nelson, 1878 sensu Neuville & Anthony, 1908 ; Connolly, 1928. LOCALITIES. Begemeder : 97 km. N. of Gondar on Asmara road (6). Choa : Between n km. N.E. of Addis Ababa and i km. S. of Debra Berhan on Asmara road, 5 samples (230) ; 32-59 km. N. of Addis Ababa on Debra Markos road, 4 samples (39) ; 10 km. W. of Ambo (8) ; 21 km. E. of Wolisso (2) ; tributaries of River Awash S. of Addis Ababa at Akaki, Dukem & Moggio (57). Sidamo : 17 and 30 km. S. of Adolla (70). Tigre : 0-5 km. S. of Adua on Gondar road (4). Wallaga : 22 km. W. of Guder on Lekemti road (22) ; 20 km. E. of Lekemti on Addis Ababa road (2). Wollo : i and 8-5 km. N. of Dessie (90). Shell (pi. 3 ; 5 and 6, 10, n) The relationship between diameter and umbilicus diameter for part of a sample from i km. S. of Debra Berhan is shown in Text-fig. 29. The ratio D/UD ranges from 3-06 to 3-93, with a mean of 3-40 (20 shells). The whorls are rounded in small shells, but the majority of medium and large specimens possess an angulation that is more or less acute and may bear a projecting fringe of periostracum. The angulation may be situated close to the underside of the shell, so that the lower surface is flattened, or nearer to the midline in which case the lower surface is rounded. In one sample (8 km. approx. N. of Dessie) the angulation is very slightly developed even in large specimens (pi. 3, 10 and u). Transverse ribs of varying size are present, and may be covered with brown bands of periostracum. A very fine spiral sculpture is present in a few samples. Four empty shells from Adowa (pi. 3, 5 and 6) have a rapidly widening whorl, and an acute angulation situated in the middle of the whorl. The height of the aperture is less than its width. A further specimen of this form, which resembles Caillaudia angulata Bourguignat, was obtained in Choa (59 km. N. of Addis Ababa, on the Debra Markos road). Anatomy (Text-figs. 30, 31) Some parts of the anatomy of G. costulatus have been previously described for specimens from Mozambique (Azevedo et al, 1957) and Angola (Wright, I963a), but in neither of these accounts is the pneumostome region adequately described. In a well relaxed specimen (Text-fig. 30) the pulmonary siphon (PN) is a spout-like projection of the thickened ventral edge of the pneumostome, the left edge of which passes posteriorly for a short distance along the floor of the mantle cavity, overlying 70 D. S. BROWN the rectum as a short median rectal ridge (MED, following the terminology of Schutte & Van Eeden, 1959). The pseudobranch lobe bears a dorsal lamella (UL), that corresponds to that described by Schutte & Van Eeden, but, with the free edge lying on the right side of the pseudobranch lobe, its position is reversed com- pared to that in Biomphalaria pfeifferi. As a result of this, the anus (A), which opens near the base of the attachment of the dorsal lamella, is hidden by this lamella in G. costulatus. From the free margin of the dorsal lamella at the level of the anus, a ridge passes posteriorly and overlies the rectum as the lateral rectal ridge (RR). The genital organs (Text-fig. 31) are similar to those of specimens from Angola and Mozambique, but not more than three lobes of the prostate gland are present even in large specimens (Wright records 9 in Angolan specimens). The penial stylet resembles that described by Wright. Radula In all of 5 radulae (9 km. approx. N. of Bessie) the cusps of unworn central teeth are not short and rounded as figured by Azevedo et al. Lateral teeth have three cusps and towards the I2th tooth in each transverse row an additional cusp is developed lateral to the ectocone. The upper part of the i2th tooth is relatively broad compared to that of the preceding teeth and the subsequent marginal teeth bear 5 or 6 cusps, among which that representing the mesocone of the lateral teeth is usually the biggest. REMARKS. Ethiopian Gyraulus costulatus can readily be distinguished from Gyraulus collected in Aden Protectorate and identified as G. convexiusculus Hutton by Wright (1963^. The whorl of the Aden specimens widens less rapidly than in G. costulatus, and the umbilicus is wider (Text-fig. 29) ; this difference in shell shape is reflected in the far more elongated genital system of G. convexiusculus. There is considerable variation in shell shape but this does not lend itself to arrangement in subspecific categories as is possible for G. costulatus in Uganda and adjacent territories (Mandahl-Barth 1954). No Ethiopian specimens were collected from lakes which were the kind of habitat in which well defined forms were found by Mandahl-Barth. A few specimens resembling Caillaudia angulata Bourguignat were obtained from two localities. As no other Gyraulus were found in these places, and no shells of similar shape are present in the other samples of G. costulatus, it is possible that Bourguignat's species is distinct from G. costulatus. The anatomy of the single living specimen that was obtained is similar to that of G. costulatus. G. angulata was described from one of Jickeli's figured specimens of Planorbis costulatus var. (PI. 7, fig. 22), the other of which (pi. 7, fig. 23) was named P. aethiopicus by Bourguignat who considered it to be strikingly different to P. costulatus of Krauss. Krauss figured a shell that is unusually flattened with a carination and coarse ribs, whereas Jickeli's specimen is relatively thick with an almost evenly rounded whorl and finer ribs. The synonymy of the Ethiopian forms with G. costulatus cannot be established with certainty until large samples are examined from near the type locality in Natal. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 71 Planorbis abyssinicus Jickeli is listed as a subspecies of G. costulatus by Bacci (1951), but careful examination of the description and figure of the shell shows that the species resembles Anisus natalensis far more closely. Planorbis ehrenbergi Beck (=P. cornu Ehrenberg, nom. nud.) is recorded from near Addis Ababa by Neuville & Anthony (1908) with some reservation. This species appears to be a Gyraulus and is otherwise known from the Egyptian Nile. The shell (Roth, 1856, pi. 2, figs. 6-9) is not angulated, and the aperture is relatively rounded in a manner very similar to the specimen illustrated in pi. 3 (figs. 10, u) of the present paper. If this is a valid species, it is likely as Neuville & Anthony remark, that it does occur in Ethiopia. The present specimens are referred to G. costulatus, pending the clearer definition of the forms of Gyraulus occurring in Africa, on both conchological and anatomical grounds. DISTRIBUTION. G. costulatus is widely distributed in Ethiopia and is often abundant in small slow-flowing waters ; a population was found at an altitude of 2,800 m. (9,150 ft.). Further south in Africa this species has an extensive range that includes Natal and the Congo. LENTORBIS Mandahl-Barth Lentorbis junodi (Connolly) Hippeutis junodi Conolly, 1922 : 121 ; Segmentina (Hippeutis) junodi Conolly, 1925 : 200, pi. 4, fig. 30 ; Lentorbis junodi ; Mandahl-Barth, 1954. LOCALITY. Choa : 62 km. S. of Kombolchia (18). Shell The shells closely resemble the type series (B.M.N.H.) of which two specimens are illustrated by Wright (i963a), and conform to descriptions given by Mandahl- Barth (1954) and Azevedo et al (1961). In all the shells there is a spiral micro- sculpture of fine wavy lines which are much finer than the longitudinal striations mentioned by Azevedo et al (loc. cit.). The colour has been described by previous authors as red-brown to light orange, but the present specimens are very pale horn coloured, or almost colourless. Largest specimen : D=4'2mm.; H=i-6mm.; 4^ whorls. Anatomy (Text-figs. 32) The structure and arrangement of the genital organs (Text-fig. 32) agrees with the description given by Mandahl-Barth (1954), except in respect of the prostate gland which bears a larger number of lobes (18 instead of 12) in the present material, the spermatheca which has a long instead of a short duct, and the penis which is not coiled in the upper part of the sheath. As described by Azevedo et al (loc. cit.) the lobes of the prostate appear to open into a separate prostatic duct, and the vas deferens opens at the tip of the penis. The ovotestis consists of approximately 20 acini arranged in two alternating rows. REMARKS. There is a close resemblance between the type series of L. junodi and two of the original specimens of L. benguelensis (Dunker) apart from the presence of traces of septa in junodi already noted by Wright (i963a). This author con- 72 D. S. BROWN trasted the copulatory organ of his specimens of L. benguelensis, in which the sheath and preputium are nearly equal in length and width, with the copulatory organ of L.junodi in which the sheath is about twice as long and narrower than the preputium according to Azevedo et al (1961). These authors account of the copulatory organ of L. junodi is confirmed in the present specimens, but further anatomical study is desirable to establish the distinctness of these two species with certainty. DISTRIBUTION. Lentorbis does not appear to have been previously recorded from Ethiopia. Mandahl-Barth (1954) drew attention to the great distance separating L. junodi in Uganda from the previously known range confined to Mozambique, and it seems likely that the apparently discontinuous range of Lentorbis merely reflects the ease with which specimens are overlooked. SEGMENTORBIS Mandahl-Barth Segmentorbis (Segmentorbis) angustus (Jickeli) Segmentina angusta Jickeli, 1874 : 220, pi. 7, fig. 24 ; Segmentina angusta? ; Blair in Ayad, 1956 ; Segmentorbis angustus ; Mandahl-Barth, 1954. LOCALITIES. Arussi : Lake Zwai, N.W. shore (i). Choa : 64 km. S. of Kombol- chia on Addis Ababa road (i). Eritrea : 15 km. N.W. of Asmara on Cheren road (4) ; River Toquor near Mekerka village (4). Kaffa : 22 km. approx. W. of Assendabo on Jimma road (2) ; 34 km. N. of Jimma on Agarro road (2). Shell Two sets of septa are present in all the specimens obtained from the type locality (River Toquor, near Mekerka), and there are traces of a third set in one shell. Up to five basal septa are present in shells from other localities. All the specimens from near Mekerka have the centre of the upper surface slightly depressed, but in the specimen from Lake Zwai this surface is flat. Spiral sculpture that is finer and less regular than that of L. junodi is present on some shells. Largest shell : Lake Zwai) D =5-0 mm.; H— 1-7 mm.; 4^ whorls. Anatomy (Text-fig. 33) The pneumostome region is similar to that of L. junodi with the anus opening at the left edge of an anal lobe that lacks ridges. The genital organs of a specimen from the type locality are illustrated in Text-fig. 33. There are between 20 and 30 slender acini, arranged in two alternating rows, composing the ovotestis. An extremely thin part of the hermaphrodite duct precedes the seminal vesicle region, which is large in comparison to that of L. junodi, and bears filamentous processes. Similar genital organs are present in a specimen from Kaffa Province, which has even longer filamentous processes from the seminal vesicle region and about 40 acini in the ovotestis. REMARKS. There is a close resemblance between the shells of specimens from near Mekerka and those described by Jickeli, and although this author does not mention internal septa in his description, at least one septum is visible on the basal surface near the aperture in pi. 7, fig. 24C. Between i and 5 sets of septa have been recorded in specimens of S. angustus from Uganda (Mandahl-Barth, 1954) and Mozambique (Azevedo et aL 1961). FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 73 met FIGS. 30, 31. Gyraulus costulatus from i km. S. of Debra Berhan. Fig. 30. Anterior part of animal viewed from left side, mantle cut on left side and raised to expose pulmonary siphon and pseudobranch. Fig. 31. Genital organs. FIG. 32. Lentorbis junodi from 62 km. S. of Kombolchia, genital organs. FIG. 33. Segmentorbis angustus from River Toquor near Mekerka, genital organs. 74 D. S. BROWN According to these authors the duct of the spermatheca is approximately four times as long as the sac. The kind of spermatheca present in Ethiopian specimens is described by Mandahl-Barth for S. planodiscus (Melville & Ponsonby). However, as noted by this author, the sexual condition of the individual snail may effect the shape of the spermatheca and it is possible that the same factor accounts for the differences between the present specimens and those described by Azevedo et al (loc. cit.). These differences lie in the number of acini in the ovotestis, the number and arrangement of the lobes in the prostate (only four 3-branched lobes were observed in material from Mozambique), and the greater size of the seminal vesicle region in the Ethiopian specimens. DISTRIBUTION. Like the preceding species, S. angustus is easily overlooked, and is probably more widely distributed and abundant in Ethiopia than the records indicate. The species has a wide distribution in Africa, extending at least as far south as Mozambique. Segmentorbis (Acutorbis) kanisaensis (Preston) Segmentina kanisaensis Preston, 1914 : 265, pi. 18, figs. 17-19 ; Segmentorbis kanisaensis ; Mandahl-Barth, 1954. LOCALITY. Choa : 62 km. S. of Kombolchia (i). REMARKS. The occurrence of this species in Ethiopia is to be expected as it was originally described from the Nile at Kanisa and has a wide range extending to Natal (Connolly, 1939) and Angola (Wright, BIOMPHALARIA Preston Biomphalaria sudanica species group Biomphalaria sudanica (Martens) Planorbis sudanicus Martens, 1870 : 35. Planorbula alexandrina (non Planorbis alexandrinus Ehrenberg, 1831) Pilsbry & Bequaert, 1927 : 131. Afroplanorbis boissyi tanganikanus (Bourguignat, 1888) Bacci, 1941. Planorbis bozasi De Rochebrune & Germain, 1904 a : 141 ; 1904 b : pi. i, figs. 2-4. Planorbis abyssinicus (non Planorbis abyssinicus Jickeli, 1874) Piersanti, 1941 : fig. 37. LOCALITIES. Arussi : Lake Zwai, N.W. shore (78) ; Lake Abyata, E. shore (15)*. Sidamo : Lake Awasa, E. shore (2)* ; Lake Margherita, E. shore (34). Shell (PL 3, figs. 7-9, 12-14, Text-fig. 34) The shells from Lakes Zwai, Awasa and Margherita have a flattened upper surface with a blunt angulation at the periphery. Beneath the angulation the surface of the whorl is more or less flattened, and there is a basal angulation close to the suture. The aperture is markedly asymmetrical in outline, being wider than it is high in the majority of specimens. The range and mean of ratio UD/H for shells between 9-0 and 10-9 mm. diameter are shown in Text-fig. 34. Coarse irregular transverse ridges are present, and where these are crossed by spiral grooves a nodular micro- sculpture is formed. Largest shell (Lake Zwai) : D=i5 mm.; UD=57 mm.; H=47 mm. ; 5^ whorls. (Lake Abyata) : D = n-8 mm.; UD=374 mm.; H=4«4 mm.; whorls. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 75 In the Lake Abyata specimens (pi. 3, figs. 12-14) the upper surface of the shell is slightly curved and the peripheral angulation is not so sharp. The width of the whorl does not increase so rapidly, the shape in cross section is almost symmetrical, so that the aperture is relatively small and almost as high as it is wide. The range and mean of ratio UD/H are included in Text-fig. 34. Anatomy Animals were obtained only from Lakes Zwai and Margherita. The penis sheath is a little longer or shorter than the preputium ; PS/PP in 10 specimens from Lake Zwai ranges from 0-78-1-04 mm. Radula The cusps of the lateral teeth are triangular (Lake Awasa, 12 radulae), or some- what arrow-head shaped (Lake Margherita, I radula). In marginal teeth where the ectocone is present it is usually a single cusp, but this may have a serrated edge and up to three small cusps have been observed. REMARKS. It will be seen in Text-fig. 34 that although the three samples of B. sudanica lie at the upper limit of the variation of UD/H, there is no marked discontinuity from the series of samples of B. pfeifferi rueppelli, one of which (No. 26) has a greater mean ratio UD/H than the sample of B. sudanica from Lake Abyata (No. 25). A large number of radulae are available only for B. sudanica from Lake Zwai, and although a single ectocone cusp is more frequently present on the marginal teeth in these specimens, there are no consistent differences to the radulae of the population of rueppelli mentioned above, or another (No. 24) which also has a high mean ratio UD/H. Both these populations are from near the floor of the Rift Valley close to the River Awash, so that although in general appearance they resemble the plateau form of rueppelli, it is possible that some genetic exchange has taken place with the lake populations of B. sudanica. Pilsbry & Bequaert (1927, p. 131) state that they have seen specimens of Planorbula alexandrina from Lake Abaja (Lake Margherita), and it is of interest that the radula of a specimen from this lake has cusps approaching the arrow-head shapecharacteris- tic of the Biomphalaria alexandrina species group (Mandahl-Barth, ig57a). The copulatory organ resembles that of B. sudanica, having the penis sheath shorter or not much longer than the preputium. Planorbis bozasi described from Lake Shala closely resembles B. sudanica from Lake Zwai. No Biomphalaria were found in Lake Shala during the recent expedition and it is unlikely that living molluscs occur in the lake at the present time because of the high alkalinity. It is evident from the figure and the dimensions (diameter =14-5 mm.) that "Planorbis abyssinicus" recorded by Piersanti from Lake Stephanie is not Jickeli's species, which resembles Anisus natalensis, but is probably B. sudanica. DISTRIBUTION. Biomphalaria sudanica is restricted to certain lakes in the southern part of the Ethiopian Rift Valley ; it occurs in Sudan, Uganda, Congo and N. Rhodesia, and has recently been found in Ghana (Mandahl-Barth, pers. comm.). 76 D. S. BROWN Biomphalaria pfeifferi species group Biomphalaria pfeifferi rueppelli (Dunker) Planorbis ruppellii Dunker, 1848 : 42 ; Martini & Chemnitz, 1856 : 41, pi. 5, figs. 10-12 ; Martens, 1869 ; Blanford, 1870 ; Morelet, 1872 ; Jickeli, 1874 ; Nevill, 1878 ; Bourguignat, 1883 ; Pollonera, 1898 ; De Rochebrune & Germain, 1904 ; Neuville & Anthony, 1908 ; Connolly, 1928 ; Giovannola, 1939 ; Planorbis (Biomphalaria} ruppellii ; Piersanti, 1940 ; Piersanti, 1941 ; Biomphalaria ruppellii ; Mandahl-Barth in Ayad, 1956 ; Biomphalaria rueppelli ; Wright & Brown, 1962. Planorbula alexandrina (non Planorbis alexandrinus Ehrenberg, 1831) Pollonera, 1898 : n. Planorbis biossyi (non Planorbis biossyi Potiez & Michaud, 1838) Jickeli, 1874 : 213 ; Pollonera, 1898 : ii ; Giovannola, 1937 : X5^ ; Planorbis biossyi var. asmarica Satta, 1936 : 193. Planorbis adowensis Bourguignat, 1879 : n ; 1888 : pi. i, figs. 1-4 ; Germain 1904 ; Neuville & Anthony, 1906 ; Germain, 1921 ; Giovannola, 1939 ; Planorbis herbini var. adowensis ; Pollonera, 1898 ; Biomphalaria adowensis ; Bacci, 1941. Planorbis herbini Bourguignat, 1883 : 101 (Jickeli, 1874 : pi. 7, fig. 18) ; Pollonera, 1898 ; Germain, 1904 ; Connolly, 1928. Planorbis cecchii Pollonera, 1887 : 2. Planorbis bridouxi (non Planorbis bridouxianus Bourguignat, 1890) Neuville & Anthony, 1908 : 253, fig. 2 ; Germain, 1931 : 366. Planorbula boccardi Pollonera, 1898 : n, figs. 22-25. LOCALITIES. Arussi : 5 km. approx. S. of Adamitullo near Bulbulla River (i)* ; River Awash S. of Nazareth (5) ; i km. S. of River Awash on Asella Road (150). Begemeder : 3 and 8 km. N. of Gorgora on Gondar Road (14) ; 5 and 66 km. N. of Gondar on Asmara road (24) ; i, 14 and 115 km. N. of Medhamie Alem on Asmara road (115). Choa : Lake Hora (23)* ; Lake Bishoftu (10) ; S. of Akaki (80) ; N. of Dukem, 2 samples (135) ; Awash river on Addis Ababa- Wolisso road (60) ; 21 km. E. and 30 km. W. of Wolisso (65) ; i-io km. W. of Ambo, 4 samples (300) ; 32-177 km. N. of Addis Ababa on Debra Markos road, 5 samples (460) ; 11-51 km. N. of Addis Ababa on Asmara road, 4 samples (540) ; 222 (Robi village) and 264 km. N. of Addis Ababa on Asmara road (250) ; 98 km. N. of Robi village on Asmara road (52). Eritrea : 22-83 km. S. of Asmara on Gondar road, 6 samples (280) ; 2 km. W. of Adi Ugri on Arezza road (8) ; 3 km. E. of Teramni on Decamere road (6) ; 66 and 70 km. S. of Asmara on Addis Ababa road (225) ; River Toquor near Mekerka village (214). Harar : 10 km. W. of Harar on Dire Dawa road (no) ; 9 km. E. of Harar on Jijiga road (38) ; 8 km. E. of Aramyia village (8) ; 46 km. on Carsa road from junction with Dire Dawa-Harar road (300) and 3 km. approx. W. of junction (10). Kaffa : 9 km. W. of Kombi-22 km. W. of Assendabo on Addis Ababa-Jimma road, 8 samples (180) ; 4 and 14 km. W. of Jimma on Bonga road (190) ; 34 and 38 km. N. of Jimma on Agarro road (130). Sidamo : Storage dams at Neghelli, 3 samples (238). Tigre : 0-5 km. S. of Adowa on Gondar road (180)*; 222 km. S. of Asmara on Addis Ababa road (7). Wallaga : 9 km. W. of Guder on Lekemti road (8). Wallo : i km. within N. boundary of Dessie on Asmara road and 30 km. N. of town (280) ; 0-5 km. W. of Kombolchia on Dessie road (14) ; 10 km. approx. E. of Kombolchia on Batei road (400) ; i km. W. of Batei N. of Kombolchia road (169) ; 26 km. S. of Kombolchia on Addis Ababa road (30). FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 77 Shell The shells are generally thicker, with a narrower umbilicus and less flattened upper surface than in B. sudanica. The largest specimen is 15-8 mm. in diameter and consists of 4! whorls, i.e. approximately one whorl less than in a sudanica shell of the same size. A spiral microsculpture consisting of rows of short ridges and nodules is present on the lower and upper surfaces of some specimens in the majority of samples. The relationship between D, UD, and H was found to vary in the same sample with the size of the shell, i.e., UD/H tended to increase with D as in two samples from near Debra Markos, Gojjam Province, described by Wright & Brown (1962). The mean and range of the ratio UD/H for 25 samples is shown in Text-fig. 34 ; a limited size group of D has been selected in order to minimise the effect of variance of UD/H with D, on the variance between different samples. Anatomy Some details of the gross anatomy of B. pfeifferi rueppelli are given by Wright & Brown (loc. cit.), and the only difference observed in the present material is that the internal muscle pillars of the preputium are variable in their relative size instead of being constantly asymmetrical. The proportions of the copulatory organ of a sample from 10 km. W. of Harar (Text-fig. 34, 3) with a narrowly umbilicate shell are : mean PS/PP = 0-83 ; range 0-59-1-00 (10 specimens). In a samples from 34 km. N. of Jimma (Text-fig. 34, 23) with a widely umbilicate shell : mean PS/PP == 0-80 ; range 0-73-1-16 (10 specimens). Radula As was the case in specimens from near Debra Markos the degree of subdivision of the ectocone of the marginal teeth (mistakenly called lateral teeth, Wright & Brown loc. cit., p. 297, 1. 34) is variable. Four cusps are present in 4 out of 15 radulae examined, but in 13 radulae at least some marginal teeth have an undivided ectocone. REMARKS. There are such conspicuous differences in shell shape between many of the populations of B. pfeifferi rueppelli in Ethiopia that the existence of many synonymous records is not surprising. The range of variation of UD/H (Text-fig. 34) of several samples overlaps only 'slightly or not at all, but there is a more or less continuous variation in the mean ratio UD/H between the most narrowly and the most widely umbilicate populations. No differences have yet been observed in the anatomy or radulae of samples representing the extreme limits of this variation, and there appears to be no justification for recognising more than a single form of the B. pfeifferi group in the plateau region of Ethiopia. Mandahl-Barth (1960) has proposed that the use of subspecies in this group should be abandoned, but it seems useful to retain a subspecific category rueppelli for the B. pfeifferi of the Ethiopian Plateau (and possibly also the Arabian Peninsula), where a higher proportion of narrowly umbilicate populations reaching a small maximum size seems to occur in comparison to the rest of the range of the species. After examining the original specimens, Bacci (1943) concluded that Planorbula alexandrina and P. boccardi recorded from Eritrea by Pollonera were not significantly different. The figures of the latter species represent a frequently occurring form of D. S. BROWN rueppelli. It is highly probable that the two single shells of P. boissyi recorded by Jickeli and Pollonera are relatively widely umbilicate shells of rueppelli, to which species Giovannola (1939) refers his own record (1937) of P. boissyi and that of Satta. As pointed out by Mandahl-Barth (igsya), P. herbini and P. cecchii are probably narrowly umbilicate and juvenile shells respectively of rueppelli, but with regard to the occurrence of B. sudanica in the Rift Valley of Ethiopia it now appears that P. bozasi belongs to this species. The specimens of P. bridouxi illustrated by Neuville & Anthony represent a narrowly umbilicate rueppelli with a widely flared aperture. 1-3- 1-2 - 1-1 i 1-O- DD, 'H " 0-9- 0-8- 0-7 - 0-6 - 34 1234567 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Localities FIG. 34. Range and mean of the ratio UD/H (umbilicus diameter/height) of Biomphalaria shells between 9-0 and 10-9 mm. diameter. Localities 1-24 and 26 B. pfeifferi rueppelli, 25, 27 and 28 B. sudanica. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 79 List 1. Eritrea ; 32 km. W. of Asmara on Cheren road. 2. Choa ; Lake Hora. 3. Harar ; 10 km. W. of Harar. 4. Tigre ; 0-5 km. S. of Adowa. 5. Wallo ; i km. N. of Dessie boundary. 6. Harar ; 46 km. on Carsa road. 7. Eritrea ; 83 km. S. of Asmara on Gondar road. 8. Eritrea ; River Toquor near Mekerka. g. Choa ; 32 km. N. of Addis Ababa on Debra Markos road. 10. Eritrea ; 101 km. S. of Asmara on Addis Ababa road. 11. Eritrea ; 39 km. S. of Asmara on Gondar road. 12. Harar ; 9 km. E. of Harar. 13. Choa ; 35 km. N. of Addis Ababa on Asmara road. 14. Choa ; 10 km. W of Ambo on Lekemti road. of localities 15. Kaffa ; 26 km. W. of Sokuru on Jimma road. 1 6. Sidamo ; army barracks Neghelli. 17. Choa ; 59 km. N. of Addis Ababa on Debra Markos road. 18. Choa ; 4 km. W. of Ambo on Lekemti road. 19. Choa ; 42 km. N. of Robi on Asmara road. 20. Kaffa ; 9 km. W. of Assendabo on Jimma road. 21. Begemeder ; i km. N. of Medhanie Alem on Asmara road. 22. Begemeder ; 14 km. N. of Medhanie Alem on Asmara road. 23. Kaffa ; 34 km. N. of Jimma on Agarro road. 24. Choa ; outflow below Koka Dam. 25. Arussi ; Lake Abyata. 26. Arussi ; i km. S. of River Awash on Asella road. 27. Sidamo ; Lake Margherita. 28. Arussi ; Lake Zwai. Particular attention must be given to P. adowensis as this form has been frequently discussed in the literature. Bourguignat states (1883, p. 101) that P. adowensis is " remarquable par sa forme globuleuse, et par la rapidite de sa croissance spirale ", but the specimen illustrated (1888) from Lake Tanganyika shows these characteristics to a lesser degree than does the shell illustrated as P. bridouxianus, which closely resembles the specimens from near Adowa in the present collection and that illustrated by Mandahl-Barth (i957a). Germain (1904) accepts that both bridouxianus and adowensis are narrowly umbilicate forms, traces a continuous gradation between sudanica and bridouxianus, and suggests geographical ranges for some of the intermediate forms. Giovannola (1939) also finds a geographical pattern in distribution of shell shape, contrasting B. rueppelli of Eritrea with the relatively narrowly umbilicate P. adowensis occurring in the vicinity of Harar. However, Neuville & Anthony (1908) do not recognise adowensis as a distinct species and draw attention to the great variation occurring in a limited area and the existence of mixed populations. The present collection supports this view, also held by Mandahl-Barth (loc. cit.), that there is no clear geographical separation in the ranges of the two forms, and it appears likely that narrowly umbilicate shells occur in all parts of the range of B. pf. rueppelli. It is also evident that when a large number of populations are considered there is a gradation between the extremes of shell shape. However, individual populations may exhibit only a limited part of the range of shell shape, 8o D. S. BROWN and this may also be so in groups of populations, as shown by the present samples from near Harar which are all relatively narrowly umbilicate as was observed by Giovannola. Specimens of " adowensis " from near Adowa were found to have the penis sheath much shorter than that of rueppelli by Mandahl-Barth. No marked difference was found between the copulatory organs of two samples representing the extremes of shell shape in the present collection (see Anatomy, above). DISTRIBUTION. The range of B. pfeifferi extends over most of Africa south of the Sahara ; B. rueppelli has been recorded from many parts of this range outside Ethiopia but it is here concluded that only the Biomphalaria of the Ethiopian plateau, possibly together with those of the Arabian Peninsula, merit recognition as the subspecies B. pfeifferi rueppelli. Family FERRISSIIDAE BURNUPIA Walker Burnupia caffra (Krauss) Ancylus caffer Krauss, 1848 : 70, pi. 4, fig. 13 ; Burnupia caffra ; Walker, 1923 ; Burnupia sp. cf. caffra ; Brown, 1961. Burnupia (P abyssinica Jickeli) ; Mandahl-Barth in Ayad, 1956. LOCALITIES. Begemeder : 8 km. N. of Gorgora (30). Choa : n km. and 67 km. N. of Addis Ababa on Asmara road (47) ; 36 km. and 59 km. N. of Addis Ababa on Debra Markos road (12) ; 32 km. W. of Wolisso (2). Sidamo ; 17 km. and 30 km. S. of Adolla (45). Shell The shells resemble a small sample from Addis Ababa (Brown, 1961), in possessing a well developed apex that is turned to the right and situated in the posterior third of the shell. Shell shape varies considerably in the larger samples that have now been obtained ; the extreme values of the ratio H/L in the sample from Begemeder are 0-38 and 0-48. The shape of the shell base also varies within samples, it may be an evenly rounded broad oval, or narrow and irregularly flattened. Radial sculpture may be hardly detectable and confined to the edge of the shell, or well developed over much of the surface and visible at a relatively low magnification (x 10). Anatomy (Text-fig. 16) The pseudobranch (fig. 16) is divided into basal and upper lamellae, each of which may be folded transversely. The rectum passes through the posterior part of the basal lamella and the anus is not covered by the upper lamella. The genital organs are similar to those described by Brown (1961). Characteristic features are, the few large lobes of the ovotestis, the absence of protuberances on the seminal vesicle region of the hermaphrodite duct, and the long lobes of the prostate gland. Larger specimens than those previously available have now been dissected. Up to ii terminal lobes are present in the ovotestis (8 were previously described), and FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 81 two additional lobes are sometimes present in the prostate gland giving a total of 10 terminal lobes. A dilatation of the vas deferens, preceding the junction with the penis sheath, has been observed in all specimens. All of 20 specimens dissected are euphallic. The "carrefour " Brown (loc. cit.) situated at the base of the albumen gland duct proximal to the junction between the hermaphrodite duct, uterus and vas deferens, appears to correspond to the " fertilisation sac " in Ancylus fluviatilis (Duncan, 1960). Radula (Text-fig. 37) The central teeth are bicuspid. Inner lateral teeth have three large cusps and a small additional lateral cusp that increases in size towards the radula margin. Marginal teeth bear up to six small sharp cusps. A radula from the Begemeder sample is unusual (Text-fig. 37) in having asymmetrical cusps on the central tooth, relatively large additional ectocones on the lateral teeth and bicuspid endocones on the first lateral teeth of one side. Grooves on the endocones of the lateral teeth are particularly marked in all the radulae seen of this sample. REMARKS. In the larger samples the extent of variation is such that selected shells from each sample could be considered to belong to different species. As it is likely that many of the existing specific diagnoses within the genus Burnupia are founded on trivial differences in shell shape and ornamentation, the Burnupia in this collection are regarded as a single form. Connolly (1939) gives the range of B. caffra as Natal and Cape Province, and the species is recorded from the Congo by Pilsbry and Bequaert (1927). The occurrence of similar specimens in Ethiopia suggests that B. caffra has a very wide range in Africa. It is also likely, considering the variation present in the Ethiopian samples, that the mixed populations mentioned by Pilsbry & Bequaert containing B. caffra and other species of Burnupia were composed of only a single form. However, it is remarkable that shells similar to caffra have not been found in East Africa ; specimens figured by Mandahl-Barth (1954) as B. stuhlmanni (Martens, 1897), and B. crassistriata (Preston, 1911), differ clearly from the Ethiopian specimens, as do the paratypes of B. kempi (Preston, 1912, B.M.N.H.). Two species of Ancylus described by Jickeli (1874) from a locality near Asmara, Eritrea, have been regarded as closely related to Ancylus fluviatilis (Miiller) by Walker (1914), but Mandahl-Barth (in Ayad, 1956) raises the question that one of these species may be Burnupia by identifying specimens from near Asmara and Addis Ababa as Burnupia ? abyssinica (Jickeli). This record was unfortunately overlooked by Brown (1961) but Dr. Mandahl-Barth now considers that these specimens are "true Ancylus" (in litt.). As the specimens figured by Jickeli have the apex turned to the right far less acutely than do any Ethiopian Burnupia, and as no specimens of this genus were found in Eritrea during the recent expedition it is almost certain that Jickeli 's species is not a Burnupia. DISTRIBUTION. B. caffra appears to have a widespread range on the Ethiopian plateau N.W. of the Rift Valley, and also occurs at similar altitudes on the S.E. side of this valley. This species does not appear to have been recorded from the 82 D. S. BROWN surrounding parts of N.E. Africa, but it probably has an extensive range as it was originally described from Natal and has been recorded in the Congo. 36 35 40 FIGS. 35-40. Radula teeth. Figs. 35, 36. Ancylus fluviatilis. Fig. 35. 10 km. N. of Debra. Berhan, central. Fig. 36. River Toquor near Mekerka, central, 3, 29, 31 and 38. Fig. 37. Burnupia caffra from 8 km. N. of Gorgora, i, central, i. Fig. 38. Ancylus sp. from 15 km. N. of Dilla, central, i, 3, 14, 23 and 27. Fig. 39. Ferrissia clessiniana ? from 17 km. S. of Adolla, central, i, 8 and 16. Fig. 40. Ferrissia isseli ? from 89 km. W. of Guder, central, i, 8 and 16. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 83 FERRISSIA Walker1 Ferrissia clessiniana (Jickeli)? Ancylus clessinianus Jickeli, 1882 : 366 ; Ferrissia clessiniana ; Walker, 1914 : figs. 9-11. Ancylus sp. Blanford, 1870. LOCALITIES. Arussi : 55 km. S. of River Awash on Asella road (45). Choa : 93 km. W. of Addis Ababa on Jimma road (i). Harar : 46 km. on Carsa road from Junction with Dire Dawa-Harar road (12). Sidamo : 4 km. S. of Wondo on Dilla road (30) ; 60 km. S. of Argereselam on Neghelli road (2) ; 17 km. S. of Adolla on Neghelli road (20). Shell The outline of the base is relatively narrow, wider in front than behind and more or less straight-sided (Text-figs. 46, 47). In one sample from Sidamo (S. of River Awash on Asella road) there is a gradation between this shape and an irregular oval outline. The apex is situated to the right of the longitudinal axis at about two- thirds of the longest dimension from the anterior margin ; in 10 shells this distance varies between 0-66 and 0-75 of the total length. The anterior surface is convex and the posterior surface concave ; the sides are almost flat and concave near the margin. Radial sculpture is absent except on the apex, where fine ribs are visible at a magnification of x 50. None of the specimens has a septum. In two of the samples an unusually large size for Ferrissia in Africa is reached. Largest shell (Sidamo, 17 km. S. of Adolla) : 4-3 mm. ; W = 2-5 mm. ; H = 1-2 mm. Sidamo : S. of River Awash : L = 5-2 mm. ; W = 3-7 mm. ; H = i -7 mm. Anatomy All 10 specimens dissected are aphallic. Radula (Text-fig. 39) The central tooth is bicuspid. Three or four cusps are present on the median lateral teeth, and additional small cusps are developed on the lateral sides of the teeth towards the margin of the radula. The marginal teeth bear about 8 slender cusps that are nearly equal in size. REMARKS. Four specimens recorded as Ancylus sp. by Blanford from near Mai Wahiz in Tigre Province appear to be the only Ferrissia previously collected in Ethiopia. Blanford compares these specimens to the Indian species A. verruca Benson, and they are considered to be a Ferrissia by Walker (1914) and Gwatkin (in Walker, loc. cit.). The present specimens resemble F. clessiniana (Jickeli), originally described from near Alexandria, in the large individual maximum size and elongated basal outline ; other species of Ferrissia known to occur in Egypt (figured by Walker, loc. cit.) have a distinctly more rounded outline. As all the specimens of F. clessiniana? were collected from submerged vegetation, xThe genus Ferrissia is here used in accordance with past usage, but it is probable that true Ferrissia do not occur in Africa (see Introduction). 84 D. S. BROWN including narrow stems and grasses, the possibility that their elongated shape may be partly due to the influence of the substratum should be considered. However, specimens collected by Dr. C. A. Wright at Tafwa, near Dhala (Western Aden Protectorate) from stones also have an elongated shape similar to that of F. clessiniana. There are consistent differences between the radulae of F. clessiniana ? and the Ferrissia described below which has a rounded outline. DISTRIBUTION. F. clessiniana? appears to be widely distributed in the southern half of Ethiopia, occurring in streams containing thick vegetation. Ferrissia isseli (Bourguignat) ? Ancylus isseli Bourguignat, 1866 : 214, pi. 33, figs. 13-18 ; Ferrissia isseli ; Walker, 1914. LOCALITY. Wallaga : 89 km. W. of Guder (22). Shell (Text-fig. 45) The outline of the base is ovate and more or less symmetrical, with the sides not straight for more than a short distance. The apex is blunt and rounded, situated slightly to the right of the longitudinal axis at about two-thirds of the longest dimension from the anterior margin. This distance varies in 10 shells between 0-64 and 0-73 of the total length. The anterior surface is gently convex and the posterior surface slightly concave or nearly straight. Lateral surfaces are slightly convex and not at all flared towards the margin as in the preceding species. Radial sculpture consists of very fine ribs confined to the apex. None of the specimens has a septum. Largest shell : L = 3-2 mm. ; W = 2-3 mm. ; H = 1-5 mm. Anatomy Ah1 of the 10 specimens dissected are aphallic. Radula (Text-fig. 40) As in the preceding species the central tooth is bicuspid, but the shape of the upper parts of the lateral and marginal teeth, and the arrangement of the cusps is conspicuously different. The median side of the upper part of each lateral tooth is elongated and bears one large cusp (which may represent the fused ectocone and mesocone), that is set apart from the other cusps. There is relatively little change in the form of the teeth until close to the margin of the radula. Marginal teeth bear 4-5 widely spaced cusps of unequal size. REMARKS. The shells are relatively higher than those of the Ferrissia species described by Mandahl-Barth (1954), and resemble the shells from Alexandria identified as F. isseli and F. pallaryi by Walker (1914). As F. pdllaryi was described from a sample of only three specimens, it is preferable to refer the Ethiopian specimens tentatively to the older name. The broader basal outline of F. isseli? compared to that of F. clessiniana? could be attributed to the fact that all specimens of the former species were collected from stones. However, the radulae provide strong evidence that the two forms are distinct. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 85 Family ANCYLIDAE ANCYLUS Miiller Ancylus fluviatilis species group Ancylus fluviatilis Miiller Ancylus fluviatilis Miiller, 1774 : 201 ; Blanford, 1870. Ancylus abyssinicus Jickeli, 1874 : 223, pi. 7, figs. 27 & 28 ; Bourguignat, 1883 ; Pollonera, 1898 ; De Rochebrune & Germain, 1904 ; Pseudancylus abyssinicus ; Connolly, 1941. Ancvlns hamacenicus Bourguignat, 1883 : 84, (= A. compressus Jickeli, nom. mid. 1874, pi. 7, fig. 26). LOCALITIES. Begemeder : 5 km. and 45 km. N. of Gondar on Asmara road (6) and (78). Choa : 10 km. N. of Debra Berhan on Asmara road (4) ; 47 km. N. of Addis Ababa on Debra Markos road (2). Eritrea : 18 km. W. of Asmara on Cheren road (18) ; River Toquor near Mekerka village (55). Tigre : 222 km. S. of Asmara on Addis Ababa road (20) . Shell (Text-figs. 41-43) There is great variation in the shape of the shell base, the relative proportions of length and height, and the shape of the apex. In a series of shells from the River Toquor near Mekerka, Eritrea (Text-fig. 41), the ratio W/L varies between 0-74 and 0-83, and H/L between 0-37 and 0-51. In the Tigre sample the apices are blunter, and one shell is almost circular in outline (Text-fig. 42). The shells from 3okm.N. of Gondar are characterised by very large curved apices, which overhang the posterior margin (Text-fig. 43). The apex is sculptured with a variable number of radial ribs, that are coarse in comparison with those of the succeeding species. A few ribs (primary ribs) extend over the rim of the apical depression into the hollow of the depression, and additional ribs are developed in increasing numbers towards the periphery of the shell. On the apex primary ribs are usually larger than later formed ribs, but a short distance from the apex all ribs are of similar size, and near the periphery the radial sculpture consists of fine wavy ridges. In some shells the sculpture does not extend beyond the ridge marking the end of the first growth stage. Fine circular sculpture connects the primary ribs on the apex, and coarser more or less regular circular ridges are present on the rest of the shell surface. Shells differ in the number and relative size of the primary ribs, and in the rate at. which later formed ribs are intercalated between them. In some shells approximately 40 ribs are present in the vicinity of the apical depression, but in others a smaller number (15 or less) of larger ribs is present. Each sample has a more or less distinctive type of radial sculpture ; both the extreme forms described above are present in the Mekerka sample, although the form with large primary ribs is the more frequent. Anatomy (Text-fig. 15) The pseudobranch contains deep transverse folds, one of which usually covers the anus (Text-fig. 15). Apart from the presence of 5-6 lobes of the prostate gland instead of 3, the genital organs resemble those described by Lacaze-Duthiers (1899). Wright (1963^ 86 U. S. BROWN comments that in specimens from Aden the flagellum appears to have a separate opening into the preputium instead of opening into the penis sheath as described by Lacaze-Duthiers. This observation was based on a specimen in which the penis sheath was everted into the preputium ; in other Aden specimens that have been dissected the structure is like that described by Lacaze-Duthiers, as is also the case in Ethiopian specimens. Radula (Text-figs. 35 and 36) The central tooth usually bears one small cusp. Median lateral teeth bear a single large cusp, but an ectocone becomes progressively well developed towards the edge of the radula, and is sub-divided into 3 or 4 small cusps on the lateral edge of the marginal teeth. Cusps are absent from the outer marginal teeth. In one longitudinal row of marginal teeth an endocone is present, and in a radula of a specimen from 10 km. N. of Debra Berhan the central tooth bears two small cusps (Text-fig. 35). In certain specimens the cusp on the central tooth is extremely small, and this probably accounts for Jickeli's failure to find this cusp in A. abyssinicus. Wright (1963^ illustrates a bicuspid central tooth in A. fluviatilis from Aden. REMARKS. The extent of variation in shell shape in the series (55 specimens) from the River Toquor near Mekerka, suggests that A. abyssinicus Jickeli and A. hamacenicus Bourguignat, which were originally described from this locality, are synonymous. In view of the probability that the shape of ancylid shells is affected by the substratum on which they grow, the relatively narrow shape of hamacenicus may be a modification induced by growth on narrow plant stems. All of the original nine specimens of hamacenicus were found on vegetation hanging in the water, whereas abyssinicus was collected from stones. Jickeli observed a close resemblance between small specimens of A. abyssinicus and A . fluviatilis from Europe, but recognised significant differences between fully grown shells, and in the radulae. Walker (1914) is of the opinion that both Jickeli's species are closely related to A . fluviatilis , but Connolly (1941) records A. abyssinicus from the Yemen as a distinct species. Wright (1963^ identifies further specimens from Western Aden Protectorate as A . fluviatilis, and refers both Connolly's records to this species. The present collection contains a wide variety of Ancylus shells resembling A. fluviatilis, and it would be desirable to compare this variation in shell shape with that occurring in Europe, and to assess the significance of differences in conjunction with anatomical information. A detailed study of this kind is beyond the scope of the present work, but it appears that the differences described by Jickeli between A. abyssinicus and A. fluviatilis are not significant. Of the outstanding shell forms in the present collection, that with a large hooked apex (Text-fig. 43) is similar to the specimen of A. pileolus Fer. illustrated by Roth (1856, pi. 2), but the flattened rounded form (Text-fig. 42) does not appear to have been described in Europe. It is likely that detailed study will reveal differences between A . fluviatilis in its European range and the isolated groups of populations in North Africa and the Arabian Peninsula which are probably remnants of a formerly more extensive range. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 87 Two forms described below appear to be distinct from the populations included above in A. fluviatilis. It is possible that distinct African forms of Ancylus also exist in Algeria (Bourguignat, 1864). DISTRIBUTION. A. fluviatilis has an extensive range on the central and northern parts of the N.W. massif of the Ethiopian plateau ; the lowest altitude at which specimens were found was 2,240 m. (7,400 ft.). This species reaches the southern limit of its range in N. Africa and the Arabian Peninsula. Ancylus sp. LOCALITIES. Choa : 30 km. W. of Wolisso (55) ; 89 km. W. of Guder (22) ; 33 km. E. of Bako (62). Kaffa : 28 km. N. of Jimma on Agarro road (35). Sidamo : 15 km. N. of Dilla (27). Shell (pi. 3, 17 and 18) The outline is broadly oval with more or less evenly curved sides. The apex is depressed, and situated at about two thirds of the longest dimension from the anterior margin. In 10 shells from 30 km. W. of Wolisso this distance varies between 0-65 and 0-69 of the total length. The blunt apex is not at all hook-shaped apart form the slightly projecting posterior rim of the apical depression. The anterior surface is evenly curved, the posterior surface is more or less straight, except at the apex and near the margin. Radial sculpture is well denned only on the apex, where 60-80 fine ridges extend to near the apical depression. Circular sculpture consists of irregular fine ridges. Largest shell : L = 7-0 mm. ; W = 5-7 mm. ; H = 2-7 mm. ; ratio H/W = 0-47. Anatomy (Text-fig. 14) The pseudobranch is a simple lobe, with usually only a trace of the transverse folds that are better developed in A. fluviatilis. This appears to be a consistent difference between the two forms, but it must be borne in mind that the appearance of the pseudobranch depends very much on the degree of contraction of the organ. No significant differences have been observed between the genital systems of the two forms. Radula (Text-fig. 38) There appear to be differences to A . fluviatilis in the shape of both the upper and the basal parts of the teeth. The ectocone is relatively well developed on the first lateral tooth, and subdivision of the ectocone begins between the i8th and 2oth teeth, instead of at about the 25th tooth in a radula of A. fluviatilis of approximately the same shell length. REMARKS. Ancylus sp. is characterised by the small maximum individual size, the depressed apex that is not hooked, and the fineness of the radial sculpture. Shells of similar shape are present in one sample of A. fluviatilis (Text-fig. 42), but these specimens reach a greater maximum size, and the apical sculpture is relatively coarse. The distribution of Ancylus sp. is confined to S.W. Ethiopia, and appears to be isolated from that of A . fluviatilis. DISTRIBUTION. Ancylus sp. has a scattered distribution in S.W. Ethiopia occurring in small stony streams. 88 D. S. BROWN 46 47 FIGS. 41-44. Dorsal and lateral shell outlines drawn by camera lucida. Figs. 41-43. Ancylus fluviatilis. Fig. 41. River Toquor near Mekerka. Fig. 42. 222 km. S. of Asmara on Addis Ababa road. Fig. 43. 45 km. N. of Gondar on Asmara road. Fig. 44. Ancylus ashangiensis sp. n. from Lake Ashangi, paratypes, B.M.N.H. FIGS. 45-47. Dorsal, lateral and posterior outlines. Fig. 45. Ferrissia isseli ? from 89 km. W. of Guder. Figs. 46, 47. Ferrissia clessiniana. ? Fig. 46. 17 km. S. of Adolla. Fig. 47. 46 km. on Carsa road from junction with Dire Dawa-Harar road. Ancylus ashangiensis sp. n. LOCALITY. Wollo : E. shore of Lake Ashangi (14).* D. S. Brown, 6.9.62. Shell (pi. 3, 19 and Text-fig. 44) The outline of the base is narrow, tapering posteriorly with almost straight sides. The high and sharply pointed apex is slightly turned to the right and situated at about three-quarters of the longest dimension from the anterior margin (in 10 shells this distance varies between 0-72 and 0-86 of the total length). The tip of the shell is produced beyond the apical depression in a lip which is turned upwards so that the apical depression lies horizontally. The anterior surface is slightly convex becoming flattened towards the apex and the anterior margin ; the posterior surface is shallowly concave. Radial sculpture on the apex consists of moderately strong ribs which become less well defined peripherally and do not reach the shell margin. Circular sculpture of fine ridges is present on the apex and there are coarse irregular growth ridges elsewhere. Holotype (B.M.N.H. coll. No. 1963920) : L = 4-8 mm. ; W = 2-3 mm. ; H = 2-2 mm. Paratypes, B.M.N.H. and Naturhistoriska Museet, Goteborg. (Gen.-Kat. 1963- "934) • FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA 89 REMARKS. Empty shells were abundant at the waters edge, and at distances of up to several hundred metres from the present shore. The poor development of the peripheral sculpture may in part be due to abrasion of the shells. There is variation in the relative width, but all the shells possess the distinctive sharp apex with a more or less horizontal apical depression, and a relatively narrow basal outline compared to that of A. fluviatilis. Of the two species of Ancylus previously described from Ethiopia A . hamacenicus Bourguignat is the narrower, but is not nearly so laterally compressed as ashangiensis. A. hamacenicus (= compressus Jickeh) was collected together with A. abyssinicus by Jickeli who acknowledged (1874, p. 223) that it might be a shell form produced by individuals living on plant stems. Both these species have been discussed above and are regarded as belonging to A. fluviatilis. The fact that all the Lake Ashangi specimens have the characteristic narrow shape is against the possibility that this may be merely a response to substratum, and supports the view that the populations are significantly distinct from the other forms of Ancylus described in this paper. Of the Ancylus known from other parts of N. Africa, A. peraudieri Bgt. and A. strictus Morelet (Bourguignat, 1864) resemble A. ashangiensis in their basal outline but differ considerably in the shape of the apices. DISCUSSION The present collection contains ten forms which are either new or have not been previously recorded from Ethiopia. Pila speciosa Lentorbis junodi Physa sp. Segmentorbis kanisaensis Bulinus ugandae Ferrissia isseli? Bulinus sp. (truncatus group) Ancylus sp. (fluviatilis group) Bulinus sp. (forskali group) Ancylus ashangiensis Physa sp., Ancylus sp. and A. ashangiensis may be regarded as having palaearctic origins ; the seven remaining forms belong to the African fauna. Ethiopia lies within the North East African Province of the zoogeographical regions of Africa proposed by Pilsbry and Bequaert (1927) on the basis of the molluscan fauna. Bacci (1951) considered that the prevalence of palaearctic forms on the Ethiopian plateaux justified the distinction of an Ethiopian Sub-Province from a Somali Sub-Province. The aquatic gastropod species considered to have palaearctic relationships by Bacci are listed below together with those in the present collection : Bacci (1951) Present Collection Theodoxus africanus Valvata nilotica scioana Valvata sp. Physa sp. Radix pereger Galba truncatula Lymnaea truncatula Ancylus abyssinicus A. hamacenicus Ancylus fluviatilis Ancylus sp. go D. S. BROWN The following species are known only as empty or sub-fossil shells. A ncylus ashangiensis Hydrobia gortanii Tropidiscus planorbis parenzani Planorbis planorbis parenzani Zoogeographical significance cannot be attached to the present record of a single population of Physa as a similar species occurs, probably as a result of introduction by human agency, further south in Africa. Significant additions to Bacci's list of palaearctic forms have thus been made only in the genus Ancylns. On the other hand, of the additional forms belonging to the African component in the fauna, four were found on the plateau : L. junodi, S. kanisaensis, Bulinus sp. (forskali group) and Ferrissia isseli? and three : Pila speciosa, Bulinus, ugandae and Bulinus sp. (truncatus group) at lower altitudes. Burnupia caff r a is another species of African affinities that occurs on the plateau and was not known to Bacci. The present collection increases the number of forms of freshwater gastropod molluscs of African rather than palaearctic affinities known to occur in Ethiopia, but it may be concluded from the wide distribution of L. tnmcatula and the genera Valvata and Ancylus, together with the diversity of forms within the Ancylus fluviatilis species group, that these are long established representatives of the palaearctic fauna on the Ethiopian plateaux. SUMMARY A collection of freshwater Gastropoda comprising 28 forms, 10 of which have not been previously recorded from Ethiopia, is described. Ancylus ashangiensis sp.n. belonging to the Ancylus fluviatilis species group is described. REFERENCES AYAD, N. 1956. Bilharziasis survey in British Somaliland, Eritrea, Ethiopia, Somalia, Sudan and Yemen. Bull. Wld. Hlth. Org. 14 : 1-117, IO n§s- AZEVEDO, J. F. DE, MEDEIROS, L., & FARO, M. 1957. O§ nioluscos de agua doce do Ultramar Portugues. 2. Moluscos do Sul do Save (Mozambique). Estudos, Ensaios e Documentos, 31 : 1-116, 34 pis. Ministerio do Ultramar, Lisbon. AZEVEDO, J. F. DE, et al. 1961. Os moluscos de agua doce do Ultramar Portugues. 3. Moluscos de Mozambique, ibid : 88 : 1-394, 62 pis., 62 figs. BACCI, G. 1940. Molluschi fossile dell 'antico fondo del Lago Zwai. Ann. Mus. Stor. nat. Genova, 60 : 454-458, 2 figs. - 1943. Nuovo contribute alia conoscenza della malacofauna dell' Africa Orientale Italiana. ibid. 61 : 120-140, 2 figs. - 1951. Elementi per una malacofauna dell'Abissinia e della Somalia. ibid. 65 : 1-144. BLAXFORD, W. T. 1870. Observations on the Geologv and Zoology of Abyssinia. London. BOURGUIGNAT, J. R. 1864. Malacologie de L'Algerie. vol. 2. Paris. - 1883. Histoire malacologique de 1'Abyssinie. Ann. Sci. nat. 15 : 47-162, 4 pis. - 1888. Iconographie malacologique des animaux mollusques fluviatiles du lac Tanganika. Paris. BROWN, D. S. 1961 A description of Burnupia sp cf caffra (Krauss) Gastropoda, Ancylidae) from Ethiopia. Ann. Mag. nat. Hist, series 13, 4 : 377-382, 6 figs. FRESHWATER GASTROPOD MOLLUSCA FROM ETHIOPIA gi BROWN, D. S. 1964. Observations on the distribution and ecology of freshwater gastropod Mollusca in Ethiopia. Contributions from the Biology Department, University College, Addis Ababa : in press. BURCH, J. B. 1960. Chromosome Numbers of schistosome vector snails. Z. f. Tropenmed. u. Parasit. n (4), 449-452. - 1964. Cy to logical studies of Planorbidae (Gastropoda : Basommatophora). I. The African subgenus Bulinus s.s. Malacologia 1 : In Press. CONNOLLY, M. 1928. On a collection of land and freshwater mollusca from Southern Abyssinia. Proc. zool. Soc. Lond. 1928 : 163-184, i pi., 4 figs. - 1939. A monographic survey of the South African non-marine Mollusca. Ann. S. African Mus. 33 : 1-660, 19 pis., 58 figs. - 1941. South Arabian non-marine Mollusca. British Museum (N.H.) Expedition to South-West Arabia, 1937-38, 1 : 17-42. DEMIAN, E. S. 1962. Anisus oasiensis n.sp. A new planorbid species from Egypt. Ark. Zool. Series 2, 15 : 149-162, 9 figs. DUNCAN, C. J. 1958. The anatomy and physiology of the reproductive system of the fresh- water snail Physa fontinalis (L.). Proc. zool. Soc. Lond. 131 : 55-84, 13 figs. GERMAIN, L. 1904. Sur quelques mollusques terrestres et fluviatile rapportes par M. Ch. Gravier du Desert Somali. Bull. Mus. Hist. nat. Paris, 10 : 344-359, i fig. - 1919. Sur les Limnees africaines appartenants au groupe du Limnaea (Radix) natalensis Krauss. ibid., 179-186. - 1921. Faune malacologique des Isles Mascareignes. Angers. GIOVANNOLA, A. 1937. Schistosomiasi intestinale da 5. mansoni nel Harar e sua transmissione con il Planorbis boissyi. Riv. Parassit. 1 : 158-162, i fig. - 1939. Ospiti intermedi dello Schistosoma mansoni in Africa Orientale Italiana. ibid. 3 : 139-144, pis. i, 2 figs. HAAS, F. 1932. Die Kobeltsche Bearbeitung det vin C. V. Erlanger in N. O. Afrika gesammel- ten Mollusken — Berichtungen und Nachtrage. Senckenbergiana, 14 : 173-185. - 1936. Binnen-mollusken aus Inner-Afrika. Abh. senckenberg. naturf. Ges. 431 : 1-156, 8 pis., 17 figs. HUBENDICK, B. 1951. Recent Lymnaeidae. Kungl. svenska veten. Handlingar, 3 : 1-222, 5 pis., 369 figs. - 1958. The development of the penial stylet in Gyraulus (Moll. Pulm.). Ark. Zool. Series 2, 11 : 427-429, 4 figs. JICKELI, C. F. 1874. Fauna der Land-und Susswasser Mollusken Nort-Ost-Afrikas. Nova Acta Leop. Carol., 37 : 1-352, n pis. LACAZE-DUTHIERS, H. DE. 1899. Des organes de la reproduction de I'Ancylus JJuviatilis. Arch. Zool. exp. Gen., 7 : 33-120, 44 figs. MANDAHL-BARTH, G. 1954- The freshwater molluscs of Uganda and adjacent territories. Ann. Mus. Afr. centr. Tervuren, 32 : 1-206, 96 figs. - i957a. Intermediate hosts of Schistosoma. African Biomphalaria and Bulinus. Bitll. Wld. Hlth. Org. 16 : 1103-1163, 15 pis., 17 figs. - i957b. Intermediate hosts of Schistosoma. African Biomphalaria and Bulinus. 2. ibid. 17 : 1-65, 25 pis., 8 figs. - 19570. La validita di Bulinus (Physopsis) abyssinicus (Martens), 1'ospite intermedio di Schistosoma haematobium in Somalia. Accad. naz. Lincei, series 8, 23 : 478-481, i p]., i fig. MANDAHL-BARTH, G. 1960. Intermediate hosts of Schistosoma in Africa — some recent information. Bull. Wld. Hlth. Org. 22 : 565-573, 2 pis., 2 figs. - 1962. Key to the identification fo East and Central African freshwater snails of medical and veterinary importance, ibid. 27 : 135-150, 32 figs. MARTENS, E. VON. 1866. Ueber einige afrikanische Binnenconchylien. Malakozool. Bl. 13 : 91-110, 2 pis. - 1867. Ueber einige Muscheln de oberen Nilgebeites. ibid. 14 : 17-20. 92 D. S. BROWN MARTENS, E. VON. 1869. Ueber einige abyssinische Schnecken. ibid. 16 : 208-215. MOHR, P. A. 1961. The geology, structure, and origin of the Bishoftu Explosion Craters. Bull. Geophysical Observatory, University College Addis Ababa, 2 : 65-101, 13 figs. MORELET, A. 1872. Notice sur les coquilles terrestres et d'eau douce recueillies sur les cotes de 1'Abyssinie. Ann. Mus. Stor. nat. Genova, 3 : 180-201. NEUVILLE, H. & ANTHONY, R. 1905. Premiere liste de mollusques d'Abyssinie (collection Maurice de Rothschild). Bull. Mus. Hist. nat. Paris, 2 : 115-117. - 1908. Recherches sur les mollusques d'Abyssinie. Ann. Sci. nat. gth series, 7 : 241-340, 3 pis., 35 figs. PAIN, T. 1961. Revision of the African Ampullariidae species of the genus Pila Roding 1798. Ann. Mus. Afr. centr. Tervuren, series in 8vo, 96 : 1-25, i pi. PIERSANTI, C. 1940. Molluschi del lago Tana e delle zone finitime. In Missione di studio al Lago Tana, 3 : Chimica e biologia, 233-241, n pis. - 1941. Mollusca. Missione biologica Sagan-Omo, 12 (6) : 263-281, 12 pis. PILSBRY, H. A. & BEQUAERT, j. 1927. The aquatic molluscs of the Belgian Congo with a geographical and ecological account of Congo malacology. Bull. Amer. Mus. nat. Hist., 53 (2) : 69-602, 67 pis., 93 figs. POLLONERA, C. 1888. Molluschi dello Scioa e della valle dell'Havash. Bull. Soc. malac. ital. 13 : 49-87. 1898. Molluschi terrestri e fluviatile dell' Eritrea raccolti dal Generale Boccard. Boll. Mus. Zool. Anat. comp. Torino, 13 : 1-13, i fig. ROCHEBRUNE, A. T. DE & GERMAIN, L. 1904. Mollusques rccuellies par la Mission du Bourg de Bogas. Mem. Soc. zool. Fr., 27 : 5-29. ROHRBACH, F. 1937. Oekologische und morphologische untersuchungen an Viviparus (Bellamva) capillatus Frauenfeld und V. (B.) imicolor Olivier. Arch. Molhiskenh. 69 : 177-218, 20 figs. ROTH, J. R. 1856. Spicilegium molluscorum orientalium annis 1852 et 1853 collectorum. Malakozool Bl. 3 : 17-58, 2 pis. SCHUTTE, C. H. J. & VAN EEDEN, J. A. 1959- Contributions to the morphology of Biomphalaria pfeifferi (Krauss). 2. Internal anatomy. Ann. Mag. nat. Hist, series 13, 2 : 136-156, 22 figS. SLUGOCKA, M. 1913. Recherches sur 1'appareil genital des Gasteropodes pulmones du genre Phvsa. Rev. suisse Zool., 21 : 75-109, 2 pis. STIGLINGH, I., VAN EEDEN, J. A. & RYKE, P. A. J. 1962. Contributions to the morphology of Bulinus tropicus. Malacologia, 1 : 73-114, 5 figs. THIELE, J. 1933. Die von O. Neumann in Abissinien gesammelten und einige andere afri- canischen Landschnecken. Sitz. Ges. Naturf. Freunde Berlin, 280-323, 2 pis. WALKER, B. 1914. Notes on the Ancylidae of N. Africa. Nautilus, 27 : 113 — 117, i pi. WRIGHT, C. A. ig63a. The freshwater gastropod Mollusca of Angola. Bull. brit. Mus. nat. Hist. (Zool.), 10 : 447-528, 16 pis., 126 figs. ig63b. The freshwater gastropod molluscs of Western Aden Protectorate, ibid. 10 : 259-274, 2 pis., 12 figs. WRIGHT, C. A. & BROWN, D. S. 1962. On a collection of freshwater gastropod molluscs from the Ethiopian Highlands, ibid. 8 : 287-311, 2 pis., 32 figs. FRESHWATER GASTROPOD MOLLUSCA EROM ETHIOPIA ABBREVIATIONS USED IN FIGURES AND TEXT an anus pa ag albumen gland pg agl cut edge of duct within albumen gland pm bp brood pouch pn br branchial ridge PP cm columellar muscle pr ct ctenidium PS dl dorsal lamella r fp fertilisation pocket rm H shell height rr hd hermaphrodite duct sv L shell length t 1m left anterior shell adductor muscle UD It left tentacle ur ML aperture length ut mp mantle process va met cut edge of mantle vd med median rectal ridge vl ov 1-3 oviducts 1-3 W p penis $ papilla preputial gland posterior shell adductor muscle pulmonary siphon preputium length prostate penis sheath length rectum retractor muscle lateral rectal ridge seminal vesicle testis umbilicus diameter ureter uterus vagina vas deferens ventral lamella shell width male aperture LOCALITIES REFERRED TO IN TEXT References are to the East African Metre Grid i : 500,000 G.S.G.S. 4355, and for Harar only G.S.G.S. 1401 (First Edition, 1960). It should be noted that the spelling of place names is variable, and that many old town names are being replaced by modern Amharic names. Adamitullo Adi Ugri Adolla Adowa Agarro Akaki . Ambo . Argereselam Asmara Assendabo . Bahar Dar . Batei . Bishoftu (DebraZeit) Bonga Carsa . Cheren Debra Berhan Debra Markos Debra Zeit Decamere Dessie Dilla . Dire Dawa Dollo . HCT 3269 Dukem .... HDE 5472 HFL 4147 Filtu JBN 9666 HCE 6650 Ghion (Wolisso) . . . HDD 5344 HFE 5166 Gondar .... HEJ 9894 HCR 0568 Gorgora .... HEJ 8154 HDE 4280 Goro Gomoto . . HDE 5030 approx. HDD 3992 Guder HDD 3191 HCL 1219 Harar DJ 6028 HFL 6496 Hagere Hiwot (Ambo) HCR 6656 Jimma .... HCR 2748 HEC 9182 Kombolchia . . . HEF 4425 HEF 7037 Kombi .... HDD 0804 Koka HDE 8237 HCP 6002 Lake Abaya (L. Margherita) DJ 3045 Lake Abyata . . . HCT 2741 HFT 0243 Lake Aramyia . . . DJ 4539 HDM 2269 Lake Ashangi . . . HEM 2190 HDS 2442 Lake Awasa . . . HCL 1882 HDE 6267 Lake Biete Mengest . . HDE 6469 HFL 6669 Lake Bishoftu . . . HDE 6269 HEF 3331 Lake Guder (Lake Hora) HCK 9008 Lake Haik . . . HEF 3852 DJ 3160 Lake Hora Harsadi, or Orsedi JBJ 5362 (Lake Beite Mengest) 94 D. S. BROWN Lake Hora . . . HDE 6471 Nego . . . HDE 7530 approx Lake Margherita . . . HCK 6607 River Awash (S. of Nazareth) HDF 0229 Lake Shala .... HCT 2525 River Ganale Doria .. . JBJ 3384 Lake Wonji .... HDD 4773 River Toquor (at Mekerka) Lake Zwai .... HCT 4169 Saati .... HFT 8825 Lekemti .... HDH 9505 Shashamanne . . . HCL 2198 Massawa .... HFU 1018 Soddu .... HCK 2857 Medhanie Alem . . . HFD 6802 Sokuru .... HCR 9173 Mekerka .... HFT 3910 Teramni .... HFL 4260 Moggio (Mojo) . . . HDE 7851 Uachile Wells . . . HBT 7202 Nazareth .... HDE 9445 Wolisso .... HDD 5344 Neghelli .... HBU 3910 Wondo .... HCL 0130 PLATE i FIG. i. Bellamya unicolor from Lake Tana, large specimen, (x i«5). FIG. 2. Bellamya unicolor from Lake Tana, small specimen. ( x 3). FIG. 3. Bellamya unicolor from 5 km. S. of Adamitullo, sub-fossil. (X3). FIGS. 4, 5. Valvata sp. from i km. S. of Debra Berhan. ( x 7-5). FIGS. 6, 8. Valvata sp. from 72 km. N. of Addis Ababa on Debra Markos road. ( x 7-5). FIGS. 7, 9. Valvata sp. from Lake Wonji. ( x 7-5). FIG. 10. Lymnaea natalensis. 38 km. N. of Jimma on Agarro road. (X3). FIG. ii. Lymnaea natalensis from Lake Tana at Gorgora. ( x 3). Bull. B.M. (N.H.) Zool. 12, 2 PLATE i PLATE 2 FIG. i. Physa sp. from outflow below Koka Dam. FIG. 2. Bulinus truncatus sericinus from 38 km. N. of Gondar. FIG. 3. Bulinus truncatus sericinus from Lake Hora. FIG. 4. Bulinus sp. from Lake Awasa. FIG. 5. Lymnaea truncatula from 19 km. N. of Gondar. FIGS. 6, 7, 10. Bulinus ugandae from Lake Margherita. FIGS. 8, ii. Bulinus africanus ovoideus from Lake Tana. FIGS. 9, 12. Bulinus africanus ovoideus from 14 km. N. of Medhanie Alem. FIGS. 13-15. Bulinus africanus ovoideus from 5 km. W. of Assendabo. All figures are x 3. Bull. B.M. (N.H.) Zoo/. 12, 2 PLATE 2 PLATE 3 FIG. i . Bulinus forskali from Lake Tana. FIG. 2. Bulinus forskali from Lake Margherita. FIG. 3. Bulinus scalaris from 4 km. N. of Gondar. FIG. 4. Bulinus sp. from 4 km. N. of Medhanie Alem. FIGS. 5, 6. Gyraulus costulatus from 0-5 km. S. of Adowa on Gondar road. (xy-5). FIGS. 7-9. Biomphalaria sudanica from Lake Zwai. FIGS. 10, ii. Gyraulus costulatus from 9 km. N. of Dessie. (xy-5). FIGS. 12-14. Biomphalaria sudanica from Lake Abyata. FIGS 15, 16. Planorbis planorbis parenzani from 5 km. S. of Adamitullo, sub-fossil. ( x 7-5). FIG. 17. Ancylus sp. from 28 km. N. of Jimma on Agarro road, lateral, posterior and dorsal views. FIG. 1 8. Ancylus sp. from 15 km. N. of Dilla, lateral, posterior and dorsal views. FIG. 19. Ancylus ashangiensis sp.n. from Lake Ashangi, lateral, posterior and dorsal views. Holotype B.M.N.H. coll. no. 1963 920. All figures are x 3 unless otherwise stated. Bull. BM. (X.H.) Zoo/. 12, 2 PLATE 3 PRINTED IN GREAT BRITAIN BY THOMAS DE LA RUE & COMPANY LIMITED LONDON 0. THE BRITISH MUSEUM (NATURAL HISTORY) EXPEDITION TO EAST NEPAL 1961-62 INTRODUCTION AND LISTS OF LOCALITIES J. G. SHEALS AND WILLIAM G. INGLIS BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 3 LONDON : 1965 THE BRITISH MUSEUM (NATURAL HISTORY) EXPEDITION TO EAST NEPAL 1961-62 INTRODUCTION AND LISTS OF LOCALITIES BY J. G. SHEALS AND WILLIAM G. INGLIS Pp. 95 - 114 ; 7 Plates ; i map BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 3 LONDON : 1965 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical Series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. This paper is Vol. 12, No. 3 of the Zoological series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals. Trustees of the British Museum (Natural History) 1963 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued April, 1965 Price Twelve Shillings THE BRITISH MUSEUM (NATURAL HISTORY) EXPEDITION TO EAST NEPAL 1961-62 INTRODUCTION AND LISTS OF LOCALITIES By J. G. SHEALS & WILLIAM G. INGLIS INTRODUCTION As a complex transitional zone the Himalayan region is of outstanding biogeo- graphical interest. Extending for 1,500 miles from Upper Burma in the East to Afghanistan in the West, this system of Tertiary mountains forms an almost unbroken chain between the Indo-Gangetic Plain and the high plateaux of Central Asia. Zoogeographically the elevation is responsible for a remarkable southerly extension of the Palaearctic zone resulting in a peculiar confluence of Indian, Indo-Malayan, Mediterranean and Northern Palaearctic elements, while floristically the range forms a unique corridor between tropical India and arid Central Asia enabling plants of the temperate Sino-Japanese region to migrate westwards to the border of the African- Indian desert. A detailed geographical account is given by Burrard & Hayden (1907). The system is not a single mountain chain but a complex series of nearly parallel ranges separated by valleys and high plateaux. Three longitudinal groups can be readily recognised. The Great Himalaya, an inner range on which lie the highest peaks; the Lesser Himalaya, a series of ranges rising to approximately 15,000 ft. and an outer system of foothills, the Siwaliks, contiguous to the Tarai forest and the plains of India. Although their upheaval was probably accompanied by a further elevation of the inner ranges, the Siwaliks are more recently formed and, in many places, a unique system of longitudinal valleys or Duns is found between these foothills and the intermediate ranges. The Himalayas can also be conveniently divided into four East- West sections and of these the least known yet possibly the most interesting biologically is the Central or Nepal Himalaya. This section, approximately 500 miles in length lies almost entirely within the present political boundaries of Nepal but geographically is denned as that portion of the system lying between the Tista and Kali rivers. The major peaks include Kanchenjunga, Everest, Makalu and Dhaulagiri and the general trend of the Great range in this section is from East to North of West except near the eastern extremity where the long Singalila ridge extends abruptly Southwards to the plains from the Kanchenjunga massif. Each section of the Himalayas has certain characteristic features. In Nepal the dun valleys are well marked and there is a tendency for the great peaks to stand in clusters or rows, each aggregation being known to the Nepalese as a Himal, e.g. Langtang Himal and Jal Jale Himal. The five principal rivers of the Nepal Himalaya are the Karnali, Rapti, Bagmati, Gandak and the Kosi, all affluents of the Ganges. The Karnali and Rapti systems drain the western area between the Garhwal border and Dhaulagiri, and the great basin of the Gandak occupies the central zone between the Annapurna and Langtang Himals. The Bagmati rises in the lesser ranges and g8 J. G. SHEALS AND W. G. INGLIS drains the Central Nepal Valley, while the Kosi, one of the most important of all the Himalayan rivers, drains the region east of Gosainthan. Its principal tributary, the Arun, cuts an impressive gorge through the Great Range a few miles East of Mount Everest, and its most easterly affluent, the Tamur, drains the South-western face of Kanchenjunga and the Western slopes of the Singalila ridge. PREVIOUS BIOLOGICAL EXPLORATION The isolationist policy adopted by Nepal until 1947 was so rigidly enforced that Landon (1928) was able to present a list of all the Europeans who had visited the country during the period 1881-1925. During these 45 years about 153 persons visited Nepal for military, official and antiquarian purposes and 55 as guests of the Maharaja. Although there was no general prohibition against Asiatics, with one notable exception even the most privileged Europeans were forbidden to travel outside the Tarai or, in upland Nepal, more than a few miles beyond the narrow limits of the Kathmandu Valley. Despite these restrictions, however, significant biological contributions were made during the igih century. The early botanical work, culminating in the publication of Don's Prodromus Florae Nepalensis (1825) and Wallich's Tentamen Florae Napalensis (1824-26), began with the visit of Francis Buchanan (later Hamilton), botanist and physician, to Kathmandu in March 1803. Buchanan stayed in the Valley for about a year and later (1807-1814) directed a great botanical survey of north and west Bengal, detaching Asiatic members of his staff to collect economic plants from the Nepal Himalaya (Prain, 1905, Burkill, 1953). Shortly after the appointment in 1816 of Edward Gardner as first political Resident at Kathmandu, Nathanial Wallich, then in charge of the East India Company's Botanic Garden at Calcutta, sent two collectors to work under the Resident's direction. They appear to be those who sent the first seed of Rhododendron arbor eum to England in 1818 (Burkill op. cit.). Wallich himself spent the year 1821 in Kathmandu but like Buchanan was hampered by restrictions on his movements although his collectors penetrated deep into the mountains along the pilgrims' road to Gosainkund. David Don, who curated the private collections of Aylmer Bourke Lambert, based his Prodromus on Buchanan's Kathmandu material, part of this collection having been presented to Lambert in 1806, but he also studied Nepalese plants sent to Lambert by Wallich. A substantial part of Lambert's herbarium, including the Buchanan collection, was purchased by the British Museum in 1842. A number of mosses collected by Gardner, Buchanan and Wallich are described in W. J. Hooker's Musci Exotici (1818, 1820). Gardner & Wallich also presented a number of birds, mammals and many insects from Nepal to Major General Thomas Hardwick. Some of these Hardwick described, others he presented to the British Museum and to the Zoological and Linnean Societies. Hardwick returned to England in 1823 and assisted Gray (1830-35) with the first volumes of the projected Illustrations of the Indian Fauna (Kinnear 1925, Sawyer 1953). Brian Houghton Hodgson was Assistant Resident, and later Resident, at Kathmandu during the period 1820-1843. Diplomat, oriental scholar and EAST NEPAL EXPEDITION, 1961-62 99 antiquarian, he has been described as one of the greatest and least thanked of all our British Residents and one of the greatest labourers in the sphere of Indian research and record (Hunter 1896, Landon 1928). His contributions to zoology were no less distinguished. He early appreciated the zoogeographical significance of the Himalayas and recognised a faunal zonation associated with altitudinal differences. During his service at Kathmandu he laid down the foundations of Himalayan mammalogy in upwards of 70 scientific papers. His interests also extended to ornithology and it is probably in this field that his greatest zoological contributions were made. After leaving Kathmandu he continued his researches in retirement at Darjeeling, returning to England in 1858. The most important part of his collections of specimens and drawings was presented to the British Museum (Gray 1863) and the smaller part originally presented to the East India Company's Museum was merged with the main collection in 1881. Hodgson's many zoological papers are catalogued by Hunter (1896) and amongst the early reports on his fish and reptile collections are those of Cantor (1839) and Gunther (1860, 1861, 1864). Sir Joseph Hooker's remarkable exploration of East Nepal is described in his famous Himalayan Journals first published in 1854. It is surprising that this visit to Nepal was sanctioned but some light is thrown on the circumstances by Hooker's recollections in Hunter's Life " I owe it entirely to his (Hodgson's) personal influence with the late Sir Jang Bahadur that I was permitted to travel in Eastern Nepal over ground never before or since traversed by any European and to visit the jealously guarded passes of the Nepalese Tibet frontier ". Hooker left Hodgson's house at Darjeeling on 27th October 1848 and with a party of 56, including a guard of 6 Nepalese soldiers and 2 officers and " Mr. Hodgson's bird and animal shooter, collector and stuff er ", crossed the Singalila ridge near Tanglu on 5th November : Hooker's route in Nepal can be fairly easily traced, although problems of translitera- tion render the identity of a few localities problematical. Crossing the " Sakkiazung ridge " (Mahabharata Lekh) near Malinge he descended to the Hima Khola which he followed to its junction with the Tamur. He then made his way up the valley to Walungchung Gola passing through " Mywa Guola ", which, from his description must certainly be the trading village now called Dobhan (sometimes spelt Dumuhan) . After exploring the region of the " Wallanchoon Pass " i.e. Tiptala Bhanjyang (16,740 ft.) he followed the Yangma Khola to Yangma Village and thence to the Tibetan frontier again at Khangla Deorali (18,315 ft.). Traversing the high ridge South of Yangma he then explored the Ghunsa Khola as far as " Kambachen " (Khamachin). On 5th December he began his southward trek, and crossing the Simbua Khola near Chairam, passed through " Tchonboong " (? Phungphung) and Sablakhu. Finally he followed the upper reaches of the Kabeli and on i5th December recrossed the Singalila a few miles North of Phalut. After exploring Sikkim and Assam, Hooker returned to Britain in 1851 and in conjunction with Thomas Thompson produced the first volume of the projected Flora Indica (1855) but the series was subsequently abandoned owing to lack of financial support. Much of Hooker's bryophyte material is described by Mitten (1859, 1860). During the remainder of the igth century almost the only contributions were ioo J. G. SHEALS AND W. G. INGLIS those made by the members of the Residency Staff, although a collector of the Indian Museum appears to have visited Kathmandu in 1871. In the preface to his Lepidoptera Indica Moore (1890) acknowledges the collection of numerous specimens of moths from Nepal made by Major General G. Ramsay, Resident during the period 1852-67, while the Residency Surgeon from 1876-1878, John Scully, published on birds (1879) and later on the Chiroptera (1887). During the latter part of the igth century several major compilations on the flora and fauna of the Indian sub-continent appeared (e.g. Hooker's Flora of British India, 1872-1897 and the first monographs of the Fauna of British India) but in relation to the Nepalese species these were substantially dependent on the collections made before 1850. Boulenger, Annandale, Wall & Regan (1907) reported on a collection of fish, amphibia and reptiles made in the vicinity of Kathmandu in 1906 by R. Hodgart, a collector of the Indian Museum, while I. H. Burkill, a botanist, also of the Indian Museum, visited Kathmandu at the end of 1907. In his notes on the journey Burkill (1910) includes a list of plants and gives an account of the cultivation and zonation of vegetation in the Valley and on the route between Raxaul and Kathmandu. In 1912 the Bombay Natural History Society initiated the Mammal Survey of India with the primary object of obtaining topotypes for systematic studies and to investigate the variation and distribution of the fauna in the sub-continent. Collections in Nepal were made during August 1920-March 1921 by the British Envoy, Lt. Col. R. L. Kennion, with the assistance of the Society's collector N. A. Baptista. The localities covered were the Tarai, the Duns (Rapti Valley), the Kathmandu Valley and the basin of the Gandak up to 50 miles North of Kathmandu. A scientific report on this collection was made by Hinton & Fry (1923). A further collection in the districts west of Kathmandu was made by Baptista during the year May 1922-May 1923 and a supplementary report was written by Fry (1925). Collections of birds made during this time by Kennion and Baptista were of value in confirming the status and distribution of a large number of species (Prater 1928). Hay (1934) observes that the serious botanical exploration of Nepal did not commence until 1928 when the Prime Minister, Maharaja Chandra Sham Sher, appointed Major Lall Dhwoj to collect seeds and specimens. Dhwoj collaborated with the Forest Research Station, Dehra Dun, in a brief exploration of S.W. Nepal in the Spring of 1929 (Parker 1932) and subsequently collected in Central and East Nepal. Many hundred sheets of specimens were received by the British Museum (Natural History) and to him is credited the introduction of several fine plants including Gentiana ornata. Dhwoj died in 1931 and was succeeded by Professor K. Sharma who made extensive and valuable collections during the period 1931-1937. Valuable plant collections were also received at this time from Lt. Col. F. M. Bailey, Envoy at Kathmandu during the period 1935-38. An eminent Lepidopterist, Bailey made collections in the Tarai and in the Valley, while the Nepalese workers whom he trained brought in botanical and zoological material from West Nepal and from the country North of Kathmandu (Bailey 1951). EAST NEPAL EXPEDITION, 1961-62 101 Zoological exploration after the second world war appears to have been initiated by S. Dillon Ripley, an ornithologist who visited the Valley in April and May 1947. Later that year another ornithologist, B. E. Smythies, received permission to make a fortnight's tour of the pilgrim trail to Gosainkund with the stipulation that he did not go within 5-6 miles of the Sacred lakes (Smythies 1948). This almost unprecedented concession foreshadowed a more liberal attitude by the Nepalese Authorities and, apart from certain minor restrictions, after the beginning of 1948 the barriers to foreign exploration were gradually removed. As a result, since 1949 a large number of expeditions have visited Nepal. The majority of these have been primarily mountaineering, although many have had secondary biological objectives, for example Tilman's expedition to the Langtang and Annapurna Himals (Tilman, 1952) and the Californian Himalayan Expedition to Makalu (Houston & Long, J955 '< Leviton, Myers & Swan, 1956). The British Museum (Natural History) and . Royal Horticultural Society jointly sponsored expeditions to West Nepal in 1952 and 1954 (Williams, 1953 ; Sykes, 1956). These expeditions were mainly concerned with flowering plants although both made small but valuable entomological and other zoological collections. Additionally, biological studies in the Nepal Valley have been continued by Mrs. D. Proud (1949, 1952, 1955) and by Herklots (1962). THE 1961-62 EXPEDITION The project was planned with a view to obtaining botanical, entomological and other zoological material from the Eastern part of the Nepal Himalaya, a compara- tively little-known area but potentially interesting by virtue of its position along the confluence of distinctive biogeographical zones. The greater part of the work was carried out in the Dhankuta province between the Arun River in the West and the Kanchenjunga Massif and the associated Singalila Range in the East. The botanical programme was directed mainly towards the Cryptogamic flora, particularly the Bryophyta, while the primary concern of the zoologists was an investigation of the Micro-arthropoda and Nematoda of soil, forest litter and allied habitats. As the collection of phanerogamous plants was not a major objective the choice of season for the field work was less restricted. Although greater populations of insects might be expected during the main flowering season from April to August a great part of this season coincides with the monsoon period when travel in the hills is difficult and hazardous owing to swollen torrents. Moreover, during these months the high humidities might severely limit the efficiency of the desiccation funnels used for collecting micro-arthropoda. Consequently the field work was planned to begin at the end of the monsoon and the collections were actually made between iyth September 1961 and 20th March 1962. However, there was still a certain conflict of requirements. The botanist, and to a certain extent the entomologist, sought a wide cover of the area under investigation and their collecting techniques did not preclude continuous movement, but the zoologists required a permanent shelter to house their bulky apparatus. Also the zoologists' interests were centred largely around the forests of Rhododendron and Evergreen Oaks at altitudes where J. G. SHEALS AND W. G. INGLIS Sketch map of East Nepal : scale 16 miles to the inch (approx.). EAST NEPAL EXPEDITION, 1961-62 103 the population of winged insects would almost certainly be low after mid November. To meet these varied requirements it was planned to acquire some type of building at an altitude of about 5,ooo ft. in the Taplejung region, and to use this as the nucleus of a base camp from which excursions could be made to all parts of the Eastern region. There are two practical routes from India to the Taplejung area. The eastern route leads from Darjeeling over the Singalila at Sandakphu or Phalut and thence by way of Memeng and Anbung to the middle reaches of the Tamur. The southern approach is from Jogbani, Bihar State, via Dharan Bazar and Dhankuta. For a small party with a moderate amount of equipment the Darjeeling route is certainly preferable, for mechanical transport can be used as far as the crest of the Singalila and from this point the 5-6-day march is not severe, but for a large party with bulky stores the road and rail transport problems are less complicated on the southern route. Early in the planning stage the Commanding Officer of the H.Q. British Gurkha Lines of Communication at Dharan offered his full co-operation and it was resolved to use Dharan as an initial base where porters could be recruited and the bulked stores split into man-loads. The stores and equipment, weighing approximately 3^ tons, were taken by sea to Bombay and thence by rail, via Calcutta, to Jogbani and the final 30 mile road trip to Dharan was made in trucks belonging to the Brigade of Gurkhas. The Expedition left Dharan for the hills with 119 porters on iyth September 1961. The success of the plan hinged largely on securing the use of a permanent building and a small reconnaissance party was fortunate in finding a suitable rest house or Dharam Sala at Sanghu, a small village at an altitude of about 6,000 ft. on the slopes of the Maewa Khola Valley. The village elders were co-operative and an added attraction was a level grass plot immediately in front of the Dharam Sala which would accommodate a large base camp. Sanghu was very near the geographical centre of the Dhankuta province and although the land in the immediate vicinity of the village was intensely cultivated, extensive forests of Evergreen Oak and Rhododendron could be reached within about two to four hours. The main party arrived at Sanghu on 26th September and the porters were paid off the following morning. The ground floor of the Dharam Sala was equipped as a kitchen and the first floor as a laboratory-cum-messroom ; tables, shelves, and racks for the soil extraction apparatus being cheaply made in bamboo. The attic space provided additional storage for canned foods and equipment. Water supplies at Sanghu were always adequate but initially some difficulty was experienced in securing food supplies for the permanent staff of field assistants. This problem was solved by the energetic efforts of the Sirdar and after the October harvest supplies were always adequate. Kerosene for lighting and cooking was always available but rather expensive at about 8/- per gallon. Communication with Dharan, mainly for the purpose of sending and receiving mail, was maintained by runner. During the whole time of the Expedition Sanghu base was manned by some British members of the Expedition except during the period 4th-3ist December, and during the last 2 weeks of January io4 J- G. SHEALS AND W. G. INGLIS when subsidiary bases for the exploration of the Arun and Tamur Valleys were set up at Tumlingtar and Dobhan respectively. Health problems were few, blood- sucking insects did not constitute a problem even at low altitudes but leeches were troublesome until the end of the monsoon and were common up to altitudes of 12,000 ft. It was found that leeches detached themselves immediately when given a short spray with a piperonyl butoxide/pyrethrum aerosol preparation. Sanghu (27° 21' N, 87° 33' E) is at about 6,000 ft. on the south side of one of the terminal branches of the Maewa Khola near the Western end of the main valley system. The river runs West to East and rises on the slopes of a North-South- running ridge, the Milke Danra, which is about 11-12,000 ft. high in this area. The valley is steep-sided and narrow so that the village is somewhat enclosed by high ground particularly to the South and West. Sanghu, not marked on some maps, lies east of Tamrang and west of Tembe. It is not a compact group of houses and the name applies to the south side of the valley between two major streams. The Dharam Sala, which was used by the Expedition, and the school building lie on one of the very few flat areas of ground, just beside the main East-West track running from Taplejung in the East to Chainpur in the West crossing the ridge by a pass — the Milke Bhanjyang — at about 11,000 ft. The entire valley side is carved into small paddy fields, up to about 7,500 ft., and is very densely cultivated. On the high ridges behind the camp is a belt of Evergreen Oak forest from 7,500 ft.-9,300 ft. and above this, Rhododendron forest extends about to the top of the ridge which is bare of large bushes or trees. The area produces large quantities of oranges (Santala), which — with bananas — grow up to about 6,000 ft. The main food crop is rice, with some potatoes and other vegetables of minor importance. Domestic animals are cattle, water-buffalo, sheep, pigs and hens. Because of the intensity of the farming the only uncultivated land below 7,500 ft. tended to be in the gullies and on very steep slopes, although the walls of the paddy fields afforded a minor habitat of some interest. The river to the west of Sanghu (between Sanghu and Tamrang) passes, at one point, through a narrow gorge bounded by steep cliffs about 200 ft. high. The flat area of sandy soil at the bottom of the gorge, bordered on the west by the river and the east by a steep bank covered by mixed vegetation, afforded an isolated natural area unaffected by cultivation. Dobhan (27° 22' N, 87° 37' E) lies at the junction of the Maewa Khola and Tamur River about half-a-day's march East of Sanghu at an altitude of about 2,500 ft. It is a major trading centre with a number of permanent shops. There is a small, flat plateau lying North-South in the junction of the two rivers both of which are crossed by chain bridges. The soil is very light and sandy with a certain amount of mica. Tumlingtar (27° 18' N, 87° 13' E) is an extensive flat plateau lying at about 1,800 ft. between the Arun River and the Sabhaya Khola at their junction. The plateau is bounded by steep cliffs, about 150-200 ft. high and is extremely dry with only very few pools of stagnant water so that most water for use on it must be carried up from the rivers. The soil on the plateau and the surrounding areas is poor, sandy, and red EAST NEPAL EXPEDITION, 1961-62 105 in colour with a high mica content. The edges of the plateau and the steep hills bordering the valley systems are marked by deep and extensive erosion gullies. The cliffs on the west side, bordering the Arun River are very steep while those on the east are less vertical with a fairly extensive flood plain between their bottom and the Sabhaya Khola. As at Sanghu and Dobhan every available piece of land is cultivated. The most obvious feature of the entire area was this extensive cultivation. Virtually every scrap of land which could be used for crops was used with a great reduction and, in some localities, elimination of any natural habitat. BOTANY Over 5,000 botanical specimens were collected. Special attention was paid to bryophytes and lichens and together these constituted over 75% of the material collected. About 300 phanerogams and a similar number of pteridophytes were also dried. The conditions in the main areas investigated were much more moist than had been expected and the profusion of bryophytes generally was extremely gratifying. In particular the rich hepatic flora in the immediate vicinity of Sanghu base provided scope for intensive study. The rocks and shaded stream banks in gullies bore a profusion of bryophytes and bamboo thickets in these gullies harboured many unusual mosses of the family Sematophyllaceae. Small bryophyte species were plentiful on north-facing walls of rice terraces and were also abundant on the trunks of trees and on moist ground in the Evergreen Oak forests at about 7,500-9,000^. Above this altitude, in the Rhododendron forests, foliaceous lichens, particularly Anaptychia and Usnea species, were common. The dry Tumlingtar plateau, which was visited in December, provided a sharp contrast with the Sanghu area. Here hepatics were uncommon but in shaded habitats mosses were plentiful although limited to a few species. The steep cliffs above the Sabhaya and Arun Kholas supported a remarkable flora of crustaceous lichens. On the open cliffs there were also patches of xerophytic Selaginella and in the less exposed areas colonies of Fissidens species were not uncommon. A camp was made for a few days in the Mewa Khola gorge above Dongen. In this remote and unspoiled area the dense forests (predominantly oak) were extremely rich localities and would have provided scope for several weeks' collecting. Unfortunately inclement weather made a longer stay impossible. Of particular interest was the profuse flora of smaller mosses and hepatics carried by the branches of shrubs. Amongst the hepatics the smaller Lejeuniaceae were prominent on the bare parts of twigs and stems and also on fern fronds. Daltonia species were also common and this was the only locality where species of Distichophyllum were found. In February a short visit was paid to the Tarai forest immediately south of Dharan. A few mosses were collected but at this time conditions were much drier than in the hill forests and Cryptogams were not abundant. LOCALITIES SANGHU BASE CAMP AREA. 27° 21' K, 87° 33' E. Alt. 5,500-7,000 ft. 2.x.6i- -4.xii.6i, 1.1.62-16.1.62, 11.11.62-19.11.62. io6 J. G. SHEALS AND W. G. INGLIS Numerous habitats in immediate vicinity of camp site, including walls, rocks, banks of streams and tracks, shallow gullies, stems of shrubs, decaying wood, tree stumps and bamboo grooves. Ficus religiosa : roots, trunks and upper branches. Rice terraces : standing crops, stubble and terrace walls. Small wood immediately above camp. Small ravine by Tamrang track : rocks and scrub by river. Gully torrents by Tembe track : banks, rocks, bamboo groves and overhanging trees. Castanopsis forest : on and amongst trees and in clearings. Scrubland above Castanopsis forest : various habitats including shallow ravines. SANGHU OAK FOREST AREA. 27° 19' N., 87° 32' E. Alt. 7,800-9,300 ft. 2.x.6i, 5.xi.6i, 28.xi.6i, 2g.xi.6i, i.xii.6i. Forest, mainly Quercus lamellosa and Q. lineata : forest floor, decaying wood, on shrubs and trees, rocks and along tracks. Scrubland below forest. MILKE BHANJYANG. 27° 19' N., 87° 31' E. Alt. 9,300-13,000 ft. 4.xii.6i, 29.xii.6i. Rhododendron forest : forest floor, rocks, on trees and along tracks. Downland above timber line : along tracks. DOBHAN. 27° 22' N., 87° 37' E. Alt. 2,500 ft. 28.i.62-3.ii.62. Low plateau between Tamur Khola and Maewa Khola : wooded cliffs at river junction and along East bank of Maewa Khola. East bank of Tamur Khola : sheltered slopes, damp cliffs and gullies. MEWA KHOLA GORGE CAMP AREA. 27° 33' N., 87° 36' E. Alt. 8,000- 9,000 ft. 22.1.62-26.1.62. Dense mixed evergreen forest : forest floor, rocks, walls, on trees and along tracks. Base of 1,000 ft. rock cliff : numerous habitats. West low cliff near river. Smaller collection from various habitats along the route from Dobhan to the Mewa Khola camp, 20.1.62-21.1.62 and 27.1.62-28.1.62. DHARAN. 26° 49' N., 87° 16' E. Alt. c. 1,000 ft. 26.ii.62-27.ii.62. Sal forest : on trees and in ditches. TUMLINGTAR. 27° 18' N., 87° 13' E. Alt. 1,800 ft. 7.xii.6i-26.xii.6i. Plateau : ponds, bamboo groves and isolated trees. East bank of Sabhaya Khola : wooded cliffs and gullies. West bank of Sabhaya Khola : wooded cliffs, rocks, gullies and sandy shore. South bank of Hinwan Khola : cliffs and along Chainpur track. East bank of Arun Khola : wooded cliffs and gullies. CHAINPUR. 27° 17' N., 87° 19' E. Alt. c. 4,000 ft. 6.xii.6i, 7.xii.6i-28.xii.6i. Isolated trees and along tracks. EAST NEPAL EXPEDITION, 1961-62 107 ENTOMOLOGY Over 21,000 insects were collected and although special attention was paid to the Diptera a wide range of material belonging to other orders, including representatives of the Orthoptera, Dermaptera, Ephemeroptera, Odonata, Neuroptera, Anoplura, Trichoptera, Hemiptera, Isoptera, Lepidoptera, Coleoptera and Hymenoptera, was obtained. With exception of the Odonata, which were particularly abundant over the paddy fields until late October the population of winged insects was not high. The field work began in late September and ended in early February and during this period most of the plants were in seed and failed to attract insects. Moreover, cold winds and frequent frost had by late November resulted in a further reduction of the population in exposed places. These circumstances stimulated a thorough exploration of the deeper ravines and other sheltered localities and an unusually extensive range of small Diptera was collected. Much of this material was collected by sweeping the dense vegetation bordering streams in deep gullies. Had the Expedition been in the field during the flowering season the time would have been almost fully spent in pursuit of the larger flies visiting the blooms and less attention would have been paid to the unexpectedly rich microfauna of the ravines. Although a systematic study of the material has not yet been made a preliminary examination has revealed interesting affinities with the fauna of the mountainous regions of Burma, Indo-China and Northern Thailand. In localities up to an altitude of about 9,000 ft. palaearctic species were generally few and mostly those with a cosmopolitan distribution. An interesting exception were the coprophagous Sphaeroceridae (Borboridae) which were represented by palaearctic species even at altitudes as low as 1,000 ft. LOCALITIES SANGHU BASE CAMP AREA. 27° 21' N., 87° 33 'E. Alt. 5,500-7,000 ft. 2.x. 61-4. xii.6i, i. i. 62-16. i. 62, n.ii. 62-19. ii. 62. Traps containing fermenting bananas and pears. Fermenting fruits of the Bhor Tree (Ficus religiosd) on camp site. Deep river gorge on Tamrang track : banks and mixed vegetation, dry and spray-splashed rocks. Standing rice crops and rice stubble. Herbaceous plants and flowering shrubs along tracks and in gullies. Shallow ravine below camp site : splashed rocks and banks of stream. Castanopsis forest : amongst trees and in clearings. Scrubland above Castanopsis forest : amongst Lycopodium sp. and shrubs. Blossoms of Prunus sp. Caked mud in dried pools. Gully torrents by Tembe track : banks, rocks and amongst overhanging trees. Small wood immediately above camp. SANGHU OAK FOREST AREA. 27° 19' N., 87° 32' E. Alt. c. 7,800-9,300 ft. 2.xi.6i, 26.xi.6i, 23.xi.6i-4.xii.6i, 11.1.62-14.1.62. io8 J. G. SHEALS AND W. G. INGLIS Forest, mainly Quercus lamellosa and Q. lineata : amongst trees in clearings and along tracks. Scrubland below forest amongst Lycopodium sp. CHAINPUR. 27° 17' N., 87° 19' E. Alt. c. 4,000 ft. 6.xii.6i. Around Ficus religiosa and on grassy plateau. TUMLINGTAR. 27° 18' N., 87° 13' E. Alt. 1,800 ft. 8.xii.6i-25.xii.6i. Traps containing fermenting banana at camp site. Sandy West shore of Sabhaya Khola : pile of fermenting millet grain, leaves, human excrement and amongst evergreen shrubs. West bank of Sabhaya Khola : grassy and rock slopes. Plateau : bare sandy ground and flowering crop of Guizotia abyssinica. East shore of Arun Khola, west of Tumlingtar : dwarf bamboo and evergreen shrubs in ravine, dried up streams and rock strewn mud flats. DOBHAN. 27° 22' N., 87° 37' E. Alt. 2,500 ft. 23.1.62—3.^62. East bank of Tamur Khola : ferns, mixed vegetation and amongst overhanging trees in deep ravine ; mixed vegetation on sheltered slopes. East bank of Maewa Khola : rice stubble, arid slopes and sheltered slopes with mixed vegetation. Tamur Khola : spray-splashed rocks. Level plateau between junction of Tamur Khola and Maewa Khola : flowering shrubs, grasses and mixed vegetation. ZOOLOGY Although the zoological programme was concerned primarily with the Micro- arthropoda and Nematoda of soil and forest litter some general collecting was also undertaken and approximately 500 spiders and harvestmen, 800 mites (parasitic and free-living), 30 " other " arthropods, 392 fish, 37 amphibia, 48 reptiles, and 32 mammals were taken. Free living nematodes were collected by the Baermann technique in which small muslin bags of soil are immersed in water held in a glass funnel. The animals move out of the soil into the surrounding water and eventually accumulate in the funnel stem. Soil samples were also taken from arable land and dried for subsequent examination for cyst-forming plant-parasitic species. In addition some helminth parasites of vertebrates were collected. Much has been written on extraction techniques for the investigation of soil and litter Micro-arthropoda (e.g. Macfadyen 1953) but for the fauna of organic soils and forest litter the only practical device for expedition purposes is some form of desiccation funnel. In this type of apparatus which was originally devised by Berlese (1905), the soil or litter is suspended on a sieve over a funnel and as the material dries the animals move downwards and eventually drop into a tube of preservative. Many recent modifications incorporate certain refinements which are important for quantitative work but almost all of these are unsuitable for expedition purposes as they either require elaborate laboratory facilities or utilise samples too small to yield sufficient material for systematic studies. EAST NEPAL EXPEDITION, 1961-62 109 Although desiccation funnels have also been used with some success for studies of mineral soil populations there is considerable evidence that in such soils, and particularly in those containing a high proportion of particles belonging to the clay fraction, certain animals either do not respond to the stimulus of desiccation or perish before making their escape (Sheals 1957), and better results are often obtained with a flotation technique. A suitable apparatus is described by Raw (1955). Essentially the process consists of eliminating stones and large plant fragments by washing and sieving and then flooding the fine soil with a magnesium sulphate solution of specific gravity 1-12 in a flotation vessel known as the Ladell. The soil particles are dispersed in the solution by a stream of air led in from below, and, after settling, the animals and plant debris on the surface of the solution are collected by decanting on to a suitable sieve. A further separation of animals from plant debris is possible but this part of the process cannot be carried out without good laboratory facilities. The bulk of the Micro-arthropoda was collected by means of a simple desiccation technique. Plastic funnels, 8^ in. in diameter, were used and the material under examination was carried on a circle of metal gauze 6^ in. in diameter with a regular square mesh of 3 mm. Fifty-two funnels were set up on a bamboo framework in the Dharam Sala and a small battery of 8 funnels, used for the examination of dung, was accommodated in a canvas shelter. Paraffin pressure lamps were suspended over the indoor funnel assembly and the soil and litter samples were left on the funnels for at least 10 days. A series of mineral soil samples from grazed turf plots and rice terraces was examined by means of a flotation technique based substantially on the apparatus described by Raw (op. cit.) but modified for operation with limited facilities. The initial washing process was carried out with a 7 -mesh sieve into plastic bowls, using jets of water from a hand-operated pump, and the soil was dispersed in the flotation vessel with a stream of air from a motor car foot pump. The finest sieve used had a regular square mesh of 152 (A (100 mesh B.S.S.). No attempt was made to separate animals from the plant debris in the " float ", consequently many extracts are " dirty " and sorting is likely to be tedious. Nevertheless the use of flotation appears to have been justified, for a preliminary examination has revealed the presence of lightly sclerotized Acari, forms which are often difficult to obtain with desiccation methods. In addition to obtaining material for systematic studies the micro-arthropod and nematode surveys were undertaken with a view to gathering ecological and zoo- geographical data. In this connection it was felt that quantitative assessments would be useful and the numbers of individuals in the samples are being recorded. Population comparisons however will of necessity be made on a proportional basis for, owing to the difficulty of standardizing samples of the various soils and litter studied and the errors inherent in the extraction techniques, absolute numbers have little meaning. Many of the habitats studied were sampled repeatedly, not so much with a view to detailed phenological studies, but rather to obtain an adequate series of developmental instars and to obtain some replication for comparisons. no J. G. SHEALS AND W. G. INGLIS LOCALITIES SANGHU BASE CAMP AREA. 27° 21' N., 87° 33' E. Alt. 5,500-7,000 ft. 2.x.6i-4.xii.6i, 1.1.62-16.1.62, 1.11.62-17.111.62. Various habitats including mosses, soil and detritus on boulders, around trees and in small gullies. Standing rice crops and walls of terraces. Rice stubble ; soil and litter. Cattle dung. Ficus religiosa : tree-hole debris and soil round roots. Surface and deeper soil of grazed turf plots (examined for micro-arthropods by both flotation and desiccation-funnels techniques). Castanopsis forest : tree trunks, bark, leaf litter, surface and deeper soil, decaying wood and mosses. Scrubland above Castanopsis forest : soil and litter. Small wood immediately above camp-site. Deep river gorge on Tamrang track : soil, litter, plant root systems, soil round roots of ferns and clinging to rocky cliffs. SANGHU OAK FOREST AREA. 27° 19' N., 87° 32' E. Alt. 7,800—9,300 ft. i6.x.6i, 26.x.6i, 17.1.62. Dense evergreen oak forest (Quercus lamellosa and Q. lineata dominating) : litter, surface and deeper soil, bark, mosses, liverworts and lichens on boulders and trees, decaying wood, tree-hole debris. Forest clearing : soil of grazed mossy turf and moss alone. MILKE BHANJYANG. 27° 19' N., 87° 31' E. Alt. 9,300-13,000 ft. i2.x.6i, i7.xi.6i, 22. xi. 61-24. xi.6i, 2.xii.6i, 14.1.62 ; 17.1.62, 28.11.62. Rhododendron forest : litter ; surface and deeper soil ; mosses, liverworts and lichens on trees and boulders ; decaying wood. Alpine turf above forest : soil examined to a depth of 9 in. with both flotation and desiccation funnel techniques for micro-arthropods and replicates examined for nematodes in two sub-samples, surface — 5 in. and 5 in. -9 in. Mosses, lichens and liverworts in same general area, on stones and boulders. Top of Milke Danra North of Milke Bhanjyang : grass and soil, surface soil round roots of rhododendrons, surface litter, moss, hard frozen soil (for nematodes only). TOPKE GOLA. 27° 3' N., 87° 35' E. Alt. 12,500-13,000 ft. 27.x.6i-28.x.6i. Small coniferous forest : litter, soil, tree-hole debris. Rhododendron forest : litter, soil, decaying wood, mosses and organic soil on large boulder, tree-hole debris. Juniper scrub (Juniperus (?)squamata) : litter and soil. Alpine turf : surface soil. SURKENAGI. 27° 14' N., 87° 50' E. Alt. c. 8,000 ft. io.xi.6i. Rhododendron / Oak forest : litter, surface soil, deeper soil, tree-hole debris, soil of turf in clearing. EAST NEPAL EXPEDITION, 1961-62 in MEMENG. 27° 12' N., 87° 56' E. Alt. c. 5,000 ft. u.xi.6i. Organic soil in rock crevices and soil of riverside turf. SOLE. 27° 10' N., 87° 56' E. Alt. c. 8,000 ft. i2.xi.6i. Evergreen oak forest : litter, surface soil, tree-hole debris. SANDAKPHU. 27° 06' N., 88° 01' E. Alt. 12,000 ft. I3.xi.6i. Rhododendron / conifer scrub : surface soil. Dwarf Rhododendron copse : surface soil. DHARAN. 26° 49' N., 87° 16' E. Alt. c. 1,000 ft. 2i.xi.6i, 27.ii.62. Sal forest : surface soil, decaying wood, moss on bank of stream. Rice stubble : surface soil. PHUSRE. 26° 50' N., 87° 17' E. Alt. c. 3,000 ft. 22.xi.6i. Moss on dry bank, litter in scrubland on side of track. MULNGHAT. 26° 56' N., 87° 19' E. Alt. c. 1,000 ft. 23.xi.6i. Grazed turf : surface soil. Moss and litter under bushes near stream bed. POPTI LA ROUTE. 27° 47' N., 87° 21' E. Alt. 10,800 ft. 2i.xii.6i. Rhododendron I Conifer forest : litter, soil, mosses on trees and boulders, decaying wood, tree-hole debris. Car doer inum giganteum : soil around roots. Evergreen Oak forest : litter, soil and decaying wood. NUM. 27° 33' N., 87° 18' E. Alt. c. 5,000 ft. i5.xii.6i. Rice stubble : surface soil. Grazed turf : surface and deeper soil. Castanopsis forest : litter and surface soil. DHOJE. 27° 26' N., 87° 13' E. Alt. c. 6,500 ft. i5.xii.6i. Grazed turf : surface soil. Ungrazed grassland : surface soil. DHANRGAON. 27° 28' N., 87° 15' E. Alt. c. 7,000 ft. i5.xii.6i. Mixed woodland, mainly oak : litter. Mosses and soil on old stone walls. TUMLINGTAR. 27° 18' N., 87° 13' E. Alt. c. 1,800 ft. 2i.xii.6i-25.xii.6i. Immediate area of camp site : rotting banana on ground, grass roots and soil, soil round bamboo, moss, rotting wood. Hinwan Khola : damp grazed turf surface soil, soil from beneath rotting maize straw and straw itself. Sabhaya Khola : shallow dammed lagoon with fish trap and very heavy growth of algae, boulders along edge of river. DOBHAN. 27° 22' N., 87° 37' E. Alt. 2,500 ft. 25.i.62-i.ii.62. Taplejung bank of Tamur Khola : damp, muddy soil, from grassy area and also under trees ; bottom litter and soil in small wood ; mosses. Plateau between junction of Tamur and Maewa Kholas : surface soil and grass roots, surface litter beneath and soil from around roots of scattered bushes. ii2 J. G. SHEALS AND W. G. INGLIS Tamur Khola about 1-2 hours march South of Dobhan : forest bottom litter, turf and mosses. ANBUNG. 27° 16' N., 87° 42' E. Alt. c. 2,500-3,000 ft. 31.1.62-2.11.62. Pinus roxburghii woodland : litter. Deep gully : plant debris and soil. Mixed woodland, mainly Castanopsis : litter and soil. Ridge above Hangpang, Alt. c. 6,000 ft. : Scrubland, mainly Pinus roxburghii : litter and soil. Junction of Tamur and Kabeli Kholas : forest bottom litter, soil and mosses. HATIA. 27° 44' N., 87° 21' E. Alt. c. 7,500 ft. 23.xii.6i. Evergreen Oak forest : litter and surface soil. WALUNGCHUNG GOLA. 27° 41' N., 87° 47' E. Alt. c. 12,700 ft. 2.ii.62. Rhododendron I conifer forest : litter, soil, moss on rocks. Coniferous forest : bottom litter. SELAP. 27° 38' N., 87° 49' E. Alt. 10,000-10,300 ft. 4.11.62. Rhododendron forest : litter, and moss on rocks. Rhododendron / Berberis scrub : litter. LUNGTHUNG. 27° 33' N., 87° 48' E. Alt. c. 6,400 ft. 5.11.62. Rice stubble : surface soil. HELLOK. 27° 31' N., 87° 48' E. Alt. 6,100 ft. 5.11.62. Mixed forest, mainly oak : litter and surface soil. DHANKUTA. 26° 59' N., 87° 21' E. Alt. c. 4,000 ft. 3.iii.62. Coniferous forest : litter. HILE. 27° 01' N., 87° 19' E. Alt. c. 6,000 ft. 3.iii.62. Soil from cultivated terrace, no plant cover. Grazed damp turf : surface soil. CHITRE. 27° 07' N., 87° 24' E. Alt. c. 7,000 ft. 4.iii.62. Mixed forest, mainly oak : surface litter. PERSONNEL J. G. Sheals, zoologist, leader first phase ; W. G. Inglis, zoologist, leader second phase ; R. L. Coe, entomologist, deputy leader ; A. H. Norkett, botanist, K. H. Hyatt, zoologist ; T. Le M. Spring Smyth, administrative officer ; and B. P. Sharma, liaison officer. ACKNOWLEDGEMENTS Thanks are due to H.M. Government of Nepal for permission to visit the area, and the financial support received from the Trustees of the British Museum, the Percy Sladen Memorial Fund, the Godman Fund, the Mount Everest Foundation, the Royal Society, Messrs. I.C.I. Ltd., the John Lewis Partnership, the late J. Spedan Lewis Esq., and Messrs. Glaxo Laboratories Ltd. is gratefully acknowledged. Throughout the planning and operational stages a great deal of advice and assistance EAST NEPAL EXPEDITION, 1961-02 113 was given by H.E. the Ambassador and the staff of the British Embassy Kathmandu, and special thanks are due to the Military Attache, Lieut. -Colonel J. O. M. Roberts, M.C., M.B.E. and his successor Lieut. -Colonel C. G. Wylie. Invaluable assistance with the many problems arising during the movement of equipment across India was given by the staffs of the Offices of The United Kingdom High Commissioner in Bombay and Calcutta. In this connection particular thanks are due to Mr. L. S. Widdows of Calcutta. The facilities provided at the H.Q. British Gurkha L. of C. Dharan for storage, load-splitting, assembling porters and for forwarding mail and freight were invaluable, and thanks are due to the Command- ing Officer, Brigadier G. S. N. Richardson, D.S.O., the Officers and N.C.O.s for their generous co-operation and hospitality. A number of meteorological instruments were kindly provided on loan by the Royal Geographical Society and thanks are due also to a large number of companies for gifts of their products. REFERENCES BAILEY, F. M. 1951. Notes on butterflies from Nepal. /. Bombay nat. Hist. Soc. 50 : 64-87 and 281-295. BERLESE, A. 1905. Apparecchio per raccogliere presto ed in gran numero piccoli artropodia. Redia 2 : 85-89. BOULENGER, G. A., ANNANDALE, N., WALL, F. & REGAN, C. T. 1907. Reports on a collection of Batrachia, Reptiles and Fish from Nepal and the Western Himalayas. Rec. Indian Mus. I : 149-158. BURKILL, I. H. 1910. Notes from a journey to Nepal. Rec. bot. Surv. India 4 : 59-140. 1953. Chapters on the history of Botany in India. I. From the beginning to the middle of Wallich's service. /. Bombay nat. Hist. Soc. 51 : 846-878. BURRARD, S. G. & HAYDEN, H. H. 1907. A sketch of the geography and geology of the Himalaya mountains and Tibet. Calcutta : Superintendent of Government Printing. CANTOR, T. E. 1839. Spicilegium Serpentium Indicorum. Proc. zool. Soc. Lond. 1839 '• 31-34 and 49-55- FRY, T. B. 1925. Bombay Natural History Society's Mammal Survey of India, Burmah and Ceylon. Report No. 37(a) Nepal. /. Bombay nat. Hist. Soc. 30 : 525-530. GRAY, J. E. 1863. Catalogue of the specimens and drawings of mammals, birds, reptiles and fishes of Nepal and Tibet presented by B. H. Hodgson Esq. to the British Museum. (2nd Edition.) London : Trustees of The British Museum. GUNTHER, A. 1860. Contributions to a knowledge of the reptiles of the Himalaya mountains. Proc. zool. Soc. Lond. 1860 : 148-175. 1861. List of the cold blooded Vertebrata collected by B. H. Hodgson Esq. in Nepal. Proc. zool. Soc. Lond. 1861 : 213-227. - 1864. Reptiles of British India. 2 vols. London: Taylor & Francis. HAY, T. 1934. Plants of Nepal. /. R. hort. Soc. 59 : 459-462. HERKLOTS, G. A. C. 1962. Some plants from Chandragiri Ridge. Part i. /. R. hort. Soc. 87 : 18-23. HINTON, M. A. C. & FRY, T. B. 1923. Bombay Natural History Society's Mammal Survey of India, Burmah and Ceylon. Report No. 37. /. Bombay nat. Hist. Soc. 29 : 399-428. HOUSTON, R. C. & LONG, W. 1955. The California Himalayan Expedition to Makalu. Sierra Cl. Bull. 40 : 1-17. HUNTER, W. W. 1896. Life of Brian Houghton Hodgson, British Resident at the Court of Nepal. London : Murray. KINNEAR, N. B. 1925. The dates of publication of the birds, in Gray and Hardwicke's " Illustrations of Indian Zoology " with a short account of General Hardwicke. Ibis. 1925 : 484-489. U4 J. G. SHEALS AND W. G. 1NGL1S LANDON, P. 1928. Nepal. 2 vols. London : Constable. LEVITON, A. E., MYERS, G. S. & SWAN, L. W. 1956. Zoological results of the Californian Himalayan Expedition to Makalu, Eastern Nepal, i. Amphibians and reptiles. Occasional papers of the Nat. Hist. Mus. of Stanford Univ. i : 1-18. MACFADYEN, A. 1953. Notes on methods for the extraction of small soil arthropods. /. anim. Ecol. 22 : 65-77. MITTEN, W. 1859. Musci Indiae Orientalis : an enumeration of the mosses of the East Indies. /. Linn. Soc. (Bot.) 3 (Suppl.) : 1-171. — 1860. Hepaticae Indiae Orientalis : an enumeration of the Hepaticae of the East Indies. /. Linn. Soc. (Bot.) 5 : 89-128. PRAIN, D. 1905. A sketch of the life of Francis Hamilton (once Buchanan) sometime superinten- dent of the Honourable Company's Botanic Garden, Calcutta. Calcutta : Bengal Secretariat Press. PARKER, R. W. 1932. List of plants collected in West Nepal. For. Bull. Dehra Dun. 76 : 1-9. PRATER, S. H. 1928. Fauna of Nepal. In Landon, P. Nepal, i : 279-334. PROUD, D. 1949. Some notes on the birds of the Nepal Valley. /. Bombay nat. Hist. Soc. 48 : 695-719. 1952. Further notes on the birds of the Nepal Valley. Ibid. 50 : 667-670. 1955- More notes on the birds of the Nepal Valley. Ibid. 33 : 57-78. RAW, F. 1955. A flotation extraction process for soil microarthropods. In Kevan, D. K. McE. Soil zoology, pp. 341-346. London : Butterworths. SAWYER, F. C. 1953. The dates of issue of J. E. Gray's " Illustrations of Indian Zoology ". /. Soc. Bibl. nat. Hist. 3 : 48-55. SCULLY, J. 1879. Contribution to the ornithology of Nepal. Stray Feathers 8 : 204-368. 1887. Chiroptera of Nepal. /. Asiat. Soc. Beng. 56 : 233-259. SHEALS, J. G. 1957. The Collembola and Acarina of uncultivated soil. /. Anim. Ecol. 26 : 125-134. SMYTHIES, B. E. 1948. Some birds of the Gandak Kosi watershed including the pilgrim trail to the sacred lake of Gosain Kund. /. Bombay nat. Hist. Soc. 47 : 432-443. SYKES, W. R. 1955. 1954 expedition to Nepal. Part i. /. R. hort. Soc. 80 : 538-544. 1956. 1954 expedition to Nepal. Part 2. Ibid. 81 : 6-14. TILMAN, H. W. 1952. Nepal Himalaya. Cambridge : University Press. WILLIAMS, L. H. J. The 1952 expedition to Western Nepal. /. R. hort. Soc. 78 : 323-337. PLATE i The Expedition Base Camp at Sanghu. The rest house (left) facing the tents was equipped as a laboratory-cum-messroom. . B.M. (N.H.) Zool. 12, 3 PLATE i 11 .«• • \ t • v.. PLATE 2 Sanghu Base Camp. Part of the battery of " outdoor " funnels used for the extraction of arthropods from dung, etc. Bull. B.M. (N.H.) Zool. 12, 3 PLATE 2 PLATE 3 Sanghu Base Camp. Flotation extraction : decanting the " float " on to a 100 mesh sieve. Bull. B.M. (N.H.) Zool 12, 3 PLATE 3 PLATE 4 The Tamur River below Dobhan, looking North. Bull. B.M. (N.H.) Zool. 12, 3 PLATE 5 Maewa Khola : evergreen oak forest above Sanghu Base Camp. Bull. B.M. (N.H.) Zoo/. 12, 3 PLATE 5 PLATE 6 Milke Danra : downland and rhododendron forests. Bull. B.M. (N.H.) Zoo/. 12, 3 PLATE 6 PLATE 7 Jaljale Himal at Topke Gola. Bull. B.M. (N.H.) Zoo/. 12, 3 PLATE 7 PRINTED IN GREAT BRITAIN BY THOMAS DE LA RUE & COMPANY LIMITED LONDON A PRELIMINARY REVISION OF THE INDO-PACIFIC ALOSINAE (PISCES: CLUPEIDAE) P. J. P. WHITEHEAD BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 4 LONDON : 1965 A PRELIMINARY REVISION OF THE INDO-PACIFIC ALOSINAE (PISCES: CLUPEIDAE) BY P. J. P. WHITEHEAD Department of Zoology, British Museum (Natural History' Pp 115-156 ; 13 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 4 LONDON: 1965 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical Series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. This paper is Vol. 12, No. 4 of the Zoological series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals. Trustees of the British Museum (Natural History) 1965 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued April, 1965 Price Sixteen Shillings A PRELIMINARY REVISION OF THE INDO-PACIFIC ALOSINAE (PISCES : CLUPEIDAE) By P. J. P. WHITEHEAU CONTENTS Page INTRODUCTION* . . . . . . . . . . 117 SYSTEMATIC CHARACTERS ......... no THE ALOSINAE OF THE INDO-PACIFIC Hilsa kelee .......... 129 Hilsa ilisha ... ....... 134 Hilsa reevesii .......... 139 Hilsa macrura . . . . . . , . . . 141 Hilsa toli ........... 143 Gudusia chapra .......... 148 Gudusia variegata ......... 150 THE WKST AFRICAN <;EXUS Etlnnalosa ...... 153 ABSTRACT Hilsa Regan and Gudusia Fowler are the only two Indo-Pacific genera of the clupeid sub- family Alosinae. New systematic characters are examined and the genus Hilsa is divided into the subgenera Hilsa (H. kelee) and Temialosa (H. ilisha, II. reevesii, II . macrura and //. toli). Macrura van Hasselt is not a senior synonym of Hilsa. Two species of Gudusia are recognized (G. chapra and G. variegata}. The position of the West African genus Etlnnalosa Regan is examined ; on present evidence, Ethmalosa is considered intermediate between the Indo-Pacific and the Atlantic Alosinae. INTRODUCTION The last revision of the Indo-Pacific Alosinae was that of Regan (1917), although Fowler (1941) listed all Indo-Pacific species and gave extensive synonymies. The more important species, such as Hilsa ilisha, have been fairly well studied, but others are little known and are poorly represented in museum collections. The material available has, however, enabled me to correct certain errors in previous synonymies and, using characters not employed before, to define the species more precisely. At the same time these characters have renewed the problem of the true phyletic relationships, both within the Indo-Pacific group, and between them and the Eastern Atlantic and New World groups. Such features as the development of a cleithral lobe or the presence of a striated fronto-parietal wedge appear in clupeid genera which otherwise are not closely related, and until the significance of such n3 P. J. P. WH1TEHEAD characters within the whole family can be reviewed, their importance relative to other characters within the Alosinae can only be guessed at. For this reason, and also because of the lack of specimens in some species, this study can represent no more than a preliminary revision. Regan (loc. cit.) treated as a single group those clupeid genera which have a dis- tinct notch in the middle of the upper jaws, stating that " all these fishes appear either to be migratory, entering rivers to spawn in fresh or brackish water, or are permanently fluviatile." Regan then divided this group into those genera in which the epibranchial gillrakers near the angle of the arch fold down over the gillrakers of the ceratobranchial ; and those in which the gillrakers near the angle of the arch lie roughly in the same plane. In the first section he placed the genera Alosa, Caspialosa, Brevoortia, Pomolobns and Ethmidium ; and in the second section he placed the gizzard shads and the genera Hilsa, Gudusia and Ethmalosa. Svetovidov (1952) placed all these genera in the subfamily Alosinae, except the gizzard shads, which he removed to a separate subfamily, the Dorosomatinae. Regan's gillraker character would seem to be a useful means of dividing the Alosinae into two tribes. It is supported by a difference in pelvic finray number, the Indo-Pacific genera having 8 rays, and the New World, Eastern Atlantic and Mediterranean genera having 7 or 9. However, the West African genus Ethmalosa appears to be as closely allied to the Atlantic Alosinae as it is to those of the Indo- Pacific. This is discussed in more detail later. Both Hilsa and Gudusia contribute to important fisheries, and in Bengal, Hilsa ilisha is the most popular of all marine fishes (Hora 1954). The Hilsa ilisha fisheries of India have received considerable attention among Indian workers, and much of the available biological knowledge has been summarised by Pillay and Rosa (1957). H. ilisha, and probably all other species, is anadromous, but land-locked populations exist (e.g. in Chilka Lake — see Mitra and Devasandarum 1954). Gudusia on the other hand is purely fluviatile. Hilsa has a wide distribution, from Natal to China, but Gudusia is restricted to the rivers of India and Burma. Regan (loc. cit.} used the form of the fronto-parietal ridges to divide his seven species of Hilsa into two groups. I have here followed Fowler (1934, 1941) in using this character to support a subgeneric division within Hilsa. Further characters are discussed which reinforce such a division. The following classification has been adopted here : — Subfamily Alosinae (Shads) Genus Hilsa Subgenus Hilsa (H. kclee] Subgenus Tenualosa (H. ilisha, H. reevesii, H. toll, H. macrura) Genus Gudusia (G. chapra, G. variegata) Note on measurements used. LENGTH : standard length used throughout. HEAD LENGTH : the longest measurement, from premaxillary symphysis to posterior border of suboperculum, occasionally to posterior border of operculum. A REVISION OF THE INDO-PACIFIC ALOSINAE 119 MAXILLA LENGTH : from premaxillary symphysis to maxilla tip. PREORBITAL DISTANCE : this measurement includes the eye, i.e. premaxillary symphysis to posterior border of eye. CAUDAL LENGTH : length of unsealed portion of lower lobe. SYSTEMATIC CHARACTERS (a) Pronto-parietal ridges. The two forms of fronto-parietal ridge found in Hilsa are shown in Figure IA and B. In the Hilsa form (IA) the wedge-shaped ridges are exposed and bear numerous longitudinal striae. In the median area between the two wedges, the f rentals are covered by a thin layer of skin. The supra-occipital is exposed for some distance anteriorly, between the posterior tips of the frontals ; posteriorly it is raised into a median, transverse ridge. ABC S.OCC. FIG. i. Fronto-parietal striation patterns (dorsal view of posterior part of head, skin removed). A. " Hilsa " pattern (180 mm. fish, syntype of Clupea durbanensis — H. kelee). B. " Tenualosa " pattern (210 mm. fish, H. ilisha). c. " Gudusia " pattern (150 mm. fish, G. variegata). s. occ. supra-occipital, par. parietal, fr. frontal. In the Tenualosa form (IB) there is no wedge-shaped striated fronto-parietal ridge, although the lateral margins of the frontals in this area may form one or two longi- tudinal grooves. The entire region is covered by a thick layer of skin, but the lateral margins of the frontals are sometimes visible. The supra-occipital is partly exposed anteriorly, as in the Hilsa form ; posteriorly it bears a longitudinal ridge, flanked by two smaller longitudinal ridges. In Gudusia (Figure ic) the fronto-parietal ridges are striated and exposed and resemble those of the Hilsa form. However, the striae are not continued forward along the frontals to the same extent as in the Hilsa form. lao P. J. P. WHITEHEAD The difference between the two subgenera of Hilsa in fronto-parietal ridge form is in fact greater than that which separates the clupeid genus Sardinella from either Harengula (Western Atlantic) or Herklotsichthys (Indo-Pacific). In these genera the fronto-parietal character appears to hold genuine phyletic significance (White- head, I964a), and it is therefore possible that the two groups of Hilsa might be more correctly allocated to separate genera also. For the time being, however, and until this character can be studied in other clupeid genera, I have preferred to maintain a subgeneric division only. (b) Maxillary bones In the genus Hilsa there are two supra-maxillae lying along the dorsal margin of the maxilla. The second or posterior supra-maxilla is expanded and paddle-shaped posteriorly and pointed and shaft-like anteriorly. There is some variation in the shape of the expanded portion (smaller than in other species in H. ilisha and H. reevesii), but there is no trenchant difference in size or shape between the two subgenera. The first or anterior supra-maxilla, however, is slightly deeper in Hilsa (Hilsa) than in Hilsa {Tenualosa) . The expanded (posterior) portion of the maxilla is similar in general shape in both subgenera, but is rather longer in H. ilisha and H. reevesii. In the subgenus Hilsa (Hilsa), however, there are four to six longitudinal ridges on this expanded portion (Figure 2A), but these are absent in Hilsa (Tenualosa) (Figure 2B and c). Occasion- ally species of the latter subgenus may have one or more fine grooves in the maxilla, but these are never developed into ridges. H. macrura has a much shorter maxilla than have other species (Figure 2B), the tip not passing the vertical through the pupil of the eye. In Gudusia the maxilla and supra-maxillae closely resemble those of Hilsa (Tenu- alosa}, especially H. (T.) toll. (c) Cleithral lobe. In clupeid fishes the tissue covering the anterior margin of the cleithrum may be expanded in places, thus giving the gill cavity an irregular posterior outline. In some genera, for example in Sardinella, there is a fleshy lobe at the postero-ventral angle of the gill opening, just above the base of the pectoral fin. This lobe, which occurs in some genera of the Dussumieriidae, I have elsewhere termed the cleithral ' flap ' (Whitehead, 1963), but ' lobe ' is perhaps a better term. In the subgenus Hilsa, the cleithral lobe is well developed, breaking the posterior outline of the gill opening and projecting sufficiently to cover the underlying fila- ments of all but the first gill arch (Figure 3A) . In front of the cleithral lobe there is an oblique groove. The final two branchiostegal rays cover this groove. In species of the subgenus Tenualosa, the cleithral lobe projects little into the gill cavity and in most cases barely interrupts the cleithral outline (Figure 36). In FIG. 2. Maxillary bones in species of Hilsa (right side). Arrow indicates vertical through eye centre. A. Hilsa kelee (syntype of Clupea durbanensis, 180 mm.). B. Hilsa macrura (160 mm. specimen, Sarawak), c. Hilsa ilisha (210 mm. specimen, Calcutta). s. max. i anterior supra-maxilla. s. max. 2 posterior supra-maxilla. A REVISION OF THE INDO-PACIFIC ALOSINAE s.max.2 s.max.1 B P. J. P. WHITEHEAD o.db. cl.l. A REVISION OF THE INDO-PACIFIC ALOSINAE 123 H. macrura the cleithral lobe is rather more developed, but it never projects into the gill chamber to the same extent as in H. kelee. In H. macrura and other species of Hilsa (Tenualosa) the groove in front of the cleithral lobe is either very shallow or absent. This character is less apparent in juvenile fishes, and in small specimens of Hilsa (of about 50 mm.) the cleithral lobe is similar and poorly developed in both subgenera. In Gudusia the cleithral lobe is slightly developed, and resembles that of H. macrura, i.e. it is intermediate in size between Hilsa (Hilsa) and Hilsa (Tenualosa). (d) Gill filaments. In the subgenus Hilsa, the gill filaments of the outer hemibranch of the first gill arch are very short, barely half the length of those of the inner hemibranch. In species of Hilsa (Tenualosa), the filaments of the outer hemibranch are more than a half the length of those of the inner hemibranch (Figure 3 A and B). In juveniles, the outer filaments in species of Hilsa (Tenualosa), tend to be rela- tively shorter, but the distinction between the two subgenera is normally obvious, and especially if comparative material is available. Gudusia resembles Hilsa (Tenualosa), having the filaments of the outer hemibranch more than half the length of those of the inner hemibranch. (e) Gillrakers. In the subgenus Hilsa, the gillrakers on the second, third and fourth arches are curled outwards (Figure 46) . At the tip of the raker there is a small knob which, in situ, rests against that of its neighbours so that the raker tips are held in line. The inner (convex) margin of each raker bears a series of short spines or serrae (about 100 on each raker). Along the outer margin (concave) there is a thin flange of skin, expanded on some rakers, narrow on others. The lower gillrakers of the first arch in the subgenus Hilsa are longer than the corresponding gill filaments, and in fishes over 80 mm. standard length they exceed eye diameter (three-quarter eye diameter at 50 mm.). In the subgenus Tenualosa the gillrakers on all arches are straight or very slightly curved (Figure 4A). The serrae are slightly fewer than in Hilsa and there is no flange of skin along the outer margin of the raker. The tips of the rakers are pointed, except in H. macrura. The lower gillrakers of the first arch in Hilsa (Tenualosa) are as long as those in Hilsa (Hilsa) except in the case of H. macrura (Figure 40), where they are about half the length of the corresponding gill filaments (less than the diameter of the pupil at 150 mm. standard length). In Gudusia the gillrakers are straight or slightly curved and resemble those of H. ilisha in form and length. FIG. 3. Cleithral profile and gill arches (right side, operculum removed). A. Hilsa kelee (syntype of Clupea durbanensis, 180 mm.). B. Hilsa ilisha (210 mm. specimen, Calcutta). cl. I. cleithral lobe. o. db. outer hemibranch of ist arch. i. db. inner hemibranch of ist arch. 124 A P. J. P. WHITEHEAD B mm. FIG. 4. Gillrakers from lower part of first arch in species of Hilsa. A. Hilsa ilisha (215 mm. fish). B. Hilsa kelee (syntype of Clupea durbanensis, 180 mm.), c. Hilsa macrura (155 mm. fish). (f) Opercular bones. In Gudusia the lower margin of the operculum rises more steeply than in either Hilsa (Hilsa) or Hilsa (Tenualosa) (Figure 5A and B). If the line of the lower oper- culum margin is produced posteriorly, it intersects the dorsal body profile near or in front of the dorsal origin in Gudusia, but behind this point in Hilsa. In Gudusia tho suboperculum is in consequence less rectangular. Within Hilsa (Tenualosa) there is some variation in the shape of the operculum and suboperculum. In H. macrura the latter is almost rectangular and the lower border of the operculum is near the horizontal. H. toll and H. ilisha resemble H. kelee (Figure SA). In H. reevesii the operculum and suboperculum are broader than in other species, the length of the suboperculum (upper exposed margin) being contained less than twice in the height of the operculum (more than twice in all other species of Hilsa, but not in Gudusia}. A REVISION OF THE INDO-PACIFIC ALOSINAE 125 op B s.op FIG. 5. Opercular bones in Hilsa and Gudusia. Dotted arrow indicates line projected along upper (exposed) margin of sub-operculum. A. Hilsa kelee (syntype of H. durbanensis, 180 mm.). B. Gudusia variegata (150 mm. fish). op. operculum. s. op. sub-operculum. (g) Pseudobranchiae. Pseudobranchiae are present in all genera of the Alosinae. In Hilsa there are slight interspecific differences in the shape of the pseudobranch, and one of these is sufficiently marked to be commented on. The pseudobranch in H. ilisha and H. reevesii differs from that of all other species of Hilsa in having a slightly longer base and a more attenuated appearance (Figure 6A). Although the longest pseudobranchial filaments are as long as those in other species, filament length decreases more gradually posteriorly (cf Figure 6A and B). The difference between these two forms is not great, but is easily apparent when a comparison can be made. In addition, H. ilisha and H. reevesii have a distinct groove below the base of the pseudobranch, into which the tips of the gillrakers of the first arch fit. A groove is absent in H . toll and H . macrura, but occurs in H. kelee. The pseudobranch in Gudusia is attenuated and resembles that of H. ilisha. A groove is present. THE ALOSINAE OF THE INDO-PACIFIC While Hilsa kelee is sufficiently distinct from the remaining species of Hilsa to be separated subgenerically (or even generically) , Gudusia combines characters from both of the subgenera, but at the same time possesses its own distinctive features. The question arises whether the degree of affinity between the three taxa, Hilsa (Hilsa}, Hilsa (Tenualosa) and Gudusia, is equal or whether one has diverged further from the other two. The distinctive characters of each are shown in Table I. 126 P. J. P. WHITEHEAD Hilsa (Hilsd) is distinguished from Gudusia on seven counts, and from Hilsa (Tenualosa) on five, while the latter is distinguished from Gudusia on four counts. But obviously some characters have more phyletic importance than others. The fronto-parietal character links Hilsa (Hilsa} with Gudusia, but in other respects Gudusia would appear to have evolved as a fluviatile offshoot of Hilsa (Tenualosa), the increase in scale number being linked with this change of habitat. If Gudusia is recognised as a separate genus, then there is good reason for separating Hilsa A B FIG. 6. Pseudobranch shape in species of Hilsa (right side, with thickened lower edge shown uppermost). A groove present along lower edge in both species. A. Hilsa ilisha. B. Hilsa kelee. A REVISION OF THE INDO-PACIFIC ALOSINAE 127 (Tenualosa) at generic level too. Alternatively, each of the three taxa might be accorded subgeneric rank. Much depends on the importance accorded to the development of the fronto-parietal sculpture. Under these circumstances, I have preferred to leave Gudusia as a separate genus and to retain the Tenualosa forms within Hilsa for the time being. Fowler(i958) proposed a new subfamily, the Gudusiinae, defined to include Gudusia alone, but this does not seem to be justified. TABLE I Hilsa (Hilsa) Hilsa (Tenualosa) Gudusia Curled gillrakers. No fronto-parietal Small and numerous striated wedge. scales. Large cleithral lobe. Steeply rising operculum lower border. Ridged maxilla. Fluviatile habit Short outer hemi- branchon first arch. Diagnosis. Indo-Pacific clupeid fishes with a distinct notch in the middle of the upper jaw and with the gillrakers of the epibranchial of the first arch not folding over those of the ceratobranchial. Gillrakers present on inner face of epibranchial of third arch. Pseudobranch present. Body compressed, with pre- and post-pelvic scutes, but no pre-dorsal scutes. Two supra-maxillae, the posterior one expanded ; no hypomaxilla. Operculum smooth or with a few fine vertical striae. Anal fin moderate 18-29 rays, of which the last two are normal and not enlarged. Pelvic fins present, 8-rayed, not reduced in size ; pelvic origin a little behind dorsal origin. No alar scales on caudal fin. Purely riverine (Gudusia) or marine and anadromous (Hilsa). From Natal to China. KEY TO THE GENERA AND SUBGENERA i Scales large, 40-50 in lateral series ; lower edge of operculum as in Figure 5A ; marine, anadromous ......... Hilsa Regan a Fronto-parietal ridges exposed, bearing many longitudinal striae (Figure IA) ; exposed part of expanded portion of maxilla with 4-6 longitudinal ridges (Figure 2A) ; cleithral lobe prominent (Figure 3A) ; gill filaments of outer hemibranch on first arch not more than half length of inner hemibranch ; gillrakers on second, third and fourth arches curled outwards (Figure 46) subgenus Hilsa b Fronto-parietal ridges covered by skin, a few or no striae present (Figure IB) ; maxilla smooth on exposed part of expanded portion (Figure 2B) ; cleithral lobe not prominent (Figure 36) ; gill filaments of outer hemibranch on first arch more than half length of those on inner hemibranch ; gillrakers on second, third and fourth arches straight or very slightly curved (Figure 4 A and c) subgenus Tenualosa n Scales small, 80-120 in lateral series ; lower edge of operculum steeply inclined as in Figure 5 B ; fluviatile. . . . . . . Gudusia Fowler 128 P. J. P. WHITEHEAD Hilsa Regan Hilsa Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 303 (type Paralosa durbanensis Regan, ex. Durban). Paralosa Regan, 1916, Ann. Durban Mus., 1 (3) : 167 (type Clupea durbanensis Regan) (non Bleeker 1872, see below). Tenualosa Fowler, 1934, Proc. Acad. nat. Sci. Philad., 85 : 246 (type Alosa reevesii Richardson ex China). Macrura Fowler, 1941, Bull. U.S. nat. Mus., No. 100 : 626 (type Clupea kelee Cuvier) (see note on Macrura van Hasselt). Indo-Pacific alosinid fishes with moderate scales, 40-50 in lateral series, 12-20 in transverse series ; scales adherent. Suboperculum rectangular, or subrectangular, its junction with the operculum, if produced, cutting the dorsal profile in the middle of the dorsal base or behind (Figure 5A). Branchiostegal rays 5-6. Gillrakers numerous, 60-220, slender. Dorsal rays 17-20, pelvic 8, anal 17-22. Ventral scutes, 16-19 pre-pelvic and 11-15 post-pelvic. Two subgenera and five species recognised here, from coasts and rivers from Natal to China. Note on synonymy. Fowler (1941) believed Macrura van Hasselt 1823 to be the earliest available generic name for those clupeid fishes related to the ' Keelee ' of Russell (1803, p. 75, pi. 195), i.e. kanagurta, ilisha, toll, etc. Smith (1949, p. 90) followed Fowler's example, as also did Whitley (1948 and 1953) and Munro (1955). Indian workers, on the other hand, have generally placed these species in Hilsa Regan 1917. Regan (1916) proposed Paralosa for those Indo-Pacific species placed by Bleeker (1872) in Alosa. Later, (Regan 1917) he created Hilsa to replace his monotypic Paralosa, presumably on grounds of homonymy, although this is not stated ; Paralosa Bleeker 1872 is in fact a junior synonym of Sardinella (Whitehead i964a). Since the works of Fowler, Smith and Whitley are quite extensively used, and since the original indication of Macrura by van Hasselt is not easy to locate, it is worth citing the relevant passage here. Macrura is first mentioned in a letter from van Hasselt to Temminck published in the Algemeene Konst- en Letter-Bode, 1, No. 20, i6th May, 1823, and titled " Uittreksel uit een' Brief van Dr. J. C. van Hasselt aan den Heer C. J. Temminck." p. 329 " De Koelee (?) van Russ. Tab. 195 is hier in groote hoeveelheid, en hierbij voegt zich eene, welke ik heb doen afbeelden en den naam Macrura gegeven heb "* The Koelee is the Keelee of Russell (1803). No figure accompanies the text (but see below, p. 142) . The use of a capital initial letter for Macrura does not necessarily indicate a generic name, since capitals are occasionally, and for no apparent reason, used elsewhere in the text (Clupea Melostoma for example). A translation of this letter, published the following year (1824) in the Bull. Sci. nat. geol. (Ferussac), 2 : 89-92, confirms that Macrura was intended as a specific name. *" The Koelee (?) of Russ. Plate 195 is here in large numbers, and to that must be added one which I have had figured and have given the name Macrura." A REVISION OF THE INDO-PACIFIC ALOSINAE 129 p. 92 " Dans le genre Clupea Lin. mes collections sont plus riches ; j'ai divers individus du Cl. melastoma Schn., et 1'espece represented par Russell, pi. 195, se trouve ici en quantite ; il faut y joindre une autre espece que j'ai fait dessiner sous le nom de macrura." Macrura van Hasselt should not, therefore, appear in the synonymy of Hilsa. Subgenus Hilsa As denned in key. A single species recognised here, H. (Hilsa) kelee. Hilsa kelee (Cuvier) Clupea kelee Cuvier, 1829, Regne animal, ed. 2, 2 : 320 (on Keelee Russell, 1803, Fishes of Coromandel, 2 : 75, pi. 195 upper figure : type locality, Vizagapatam) . Macrura kelee : Fowler, 1941, Bull. U.S. nat. Mus., No. 100 : 627 ; Smith, 1949, Sea Fishes S. Africa : 90 ; Munro, 1955, Mar. F -water fish. Ceylon : 24 ; Fowler, 1956, Fishes of the Red Sea : 69 ; Fourmanoir, 1957, Mem. Inst. Sci. Madagasc. Serv. oceanogr., 1 : 8, fig. i. ?Clupeonia blochii Valenciennes, 1847, Hist. Nat. Poiss., 20 : 353 (type locality, Tranquebar). Hilsa blochii : Fowler, 1926, /. Bombay nat. Hist. Soc., 30 (4) : 3 ; Idem, 1938, List. Fish. Malaya : 27. Alosa brevis Bleeker, 1848, /. Ind. Arch., 2 : 638 (type locality, Bima, Sumbawa Island) ; Idem, 1872, Atlas Ichth. Ind. Nderland., 6 : 116. Hilsa brevis : Fowler, 1928, Mem. Bishop Mus., 10 : 30. Macrura brevis : Munro, 1958, Papua New Guinea agric. /., 10 (4) : 116. Alausa kanagurta Bleeker, 1852, Verh. Bat. Gen., 24 : 13, 34 (type locality, Batavia, Muntok, East Indies) ; Idem, 1865, Ned. Tijdschr. Dierk., 2 : 35, 176. Alosa kanagurta : Bleeker, 1872, Atlas Ichth. Ind. Nedrland., 6 : 114, pi. 265, fig. 5 ; Bean & Weed, 1912, Proc. U.S. nat. Mus., 42 : 590 ; Hora, 1924, Mem. Asiatic Soc. Bengal, 6 : 480 ; Suvatti, 1937, Index Fish. Siam : 9. Clupea kanagurta : Day, 1878, Fishes of India, pt. 4 : 640, pi. 162, fig. 4 ; Idem, 1889, Fauna British India, Fishes, 1 : 377 ; Pillay, 1929, /. Bombay nat. Hist. Soc., 32 (2) : 355 ; Tirant, 1929, Serv. oceanogr. Peches Indochine, 6e note : 118, 120 ; Hardenberg, 1931, Treubia, 13 (i) : in. Clupea (Alosa) kanagurta : Weber & de Beaufort, 1913, Fishes Indo-Aust. Archipelago, 2 : 67 ; Chabanaud, 1926, Serv. oceanogr. Peches Indo-Chine, ie note : 8. Hilsa kanagurta : Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 304 ; Fowler, 1934, Proc. Acad. nat. Sci. Philad., 86 : 86 ; Idem, op. cit., 87 : 90, fig. 8 ; Suvatti, 1937, Index Fish. Siam : 12 ; Fowler, 1938, List Fish. Malaya : 27 ; Bertin, 1940, Bull. Mus. Hist. nat. Paris, (2) 12 : 281 (Bleeker cotype) ; Blegvad, 1944, Danish Sci. Invest. Iran, pt. 3 : 63 ; Misra, 1947, Rec. Ind. Mus., 45 (4) : 390 ; Quereshi, 1957, Agric. Pakistan, 8 (2) : 107. Hilsa ganagurta : Chu, 1931, Biol. Bull. St. John's Univ., 1 : 14 (misspelt). Harengula kanagurta : Paradice & Whitley, 1927, Mem. Queensland Mus., 9 : 76, pi. 12, fig. i ; McCulloch, 1929, Aust. Mus. Mem., 5 : 39 ; ?Wu, 1929, Contr. Biol. Lab. Sci. Soc. China, 5 (4) : 17, fig. 13 (see note on distribution, below). Alausa ilisha : (non Hamilton-Buchanan) Bleeker, 1852, Verh. Bat. Gen., 24 : 33 ; Kner, 1865, Reise Novarra, Fische : 331. Clupea ilisha : Giinther (part), 1868, Cat. Fish. Brit. Mus., 7 : 445 ; Day (part), 1878, Fishes of India, pt. 4 : 640 ; Idem (part), 1889, Fauna Brit. India, 1 : 376. Alausa brachysoma Bleeker, 1853, Nat. Tijdschr. Ned. Ind., 5 : 527 (type locality, Padang, Sumatra) (non Sardinella brachysoma Bleeker). Alosa brachysoma : Bleeker, 1872, Atlas Ichth. Ind. Neerland., 6 : 115, pi. 262, fig. 5. Hilsa brachysoma : Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 305. 130 P. J. P. WHITEHEAD Alosa chapra : Giinther, 1866, Fishes of Zanzibar : 123. Alosa malayana Bleeker, 1866, Ned. Tijdschr. Dierk., 3 : 294 (type locality, Java, Sumatra) ; Idem, 1872, Atlas Ichthy. Ind. Neerland., 6 : 114, pi. 265, fig. 4. Clupea platygaster Giinther, 1868, Cat. Fish. Brit. Mus., 7 : 448 (on Bleeker's Sumatra specimen of Alausa brachysoma in British Museum) ; Regan, 1914, Trans, zool. Soc. London, 20 (6) : 276. Clupea (Alosa) platygaster : Weber & de Beaufort, 1913, Fishes Indo-Austr. Archipelago, 2 : 66, fig. 24 ; Chevey, 1932, Inst. oceanogr. Indochine, ige note : 9. Alosa platygaster : Roxas & Martin, 1937, Dept. Agric. Comm. Manila Tech. Bull., 6 : 21 ; Herre, 1953, List. Philipp. Fishes : 71. Clupea durbanensis Regan, 1906, Ann. Natal Gov. Mus., 1 (4) : 4, pi. 4 (type locality, Durban Bay) ; Idem, 1908, op. cit., 1 (3) : 242 ; Gilchrist & Thompson, 1908, Ann. S. Afr. Mus., 6 : 268 ; Gilchrist, 1913, Marine Biol. Rep. S. Africa, No. i : 59. Paralosa durbanensis : Regan, 1916, Ann. Durban Mus., 1 : 167 ; Gilchrist & Thompson, 1917, Ann. Durban Mus., 1 (4) : 297. Hilsa durbanensis : Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 305 ; Barnard, 1925, Ann. S. Afr. Mus., 21 (i) : in ; Fowler, 1925, Proc. Acad. nat. Sci. Philad., 77 : 195 ; Barnard, Ann. S. Afr. Mus., 21 (2) : 1017 ; Fowler, 1934, Proc. Acad. nat. Sci. Philad., 86 : 411 ; Idem, 1935, op. cit., 87 : 365. Note on synonymy. Following Regan (1917), many modern authors have either doubtfully cited or ignored Clupea kelee Cuvier, 1829 (name without description in footnote, p. 320 in 2nd edition), but Fowler (1941) and Smith (1949) give Cuvier's name priority over kanagurta Bleeker. Cuvier states, however, that his C. kelee is based on Russell's figure of Keelee. Cuvier's name is thus a valid indication, being " a bibliographic reference to a previously published description, definition or figure " (Art. 16 (a), International Code of Zoological Nomenclature 1961). Giinther (1868, p. 445) included Alausa toll Cantor in his synonymy of Clupea ilisha. The smaller of the two Cantor specimens listed by Giinther is H. kelee ; the larger is probably H. toll (but see p. 135). The remaining specimens listed by Giinther under Clupea ilisha are H. kelee. Alosa malayana Bleeker, as Weber and de Beaufort (1913) point out, seems to have been based on a juvenile specimen of H. kelee. Since eye diameter shows negative allometry with standard length, the maxilla appears to be shorter in smaller fishes, barely reaching to eye centre. I have followed Fowler (1941) and Day (1878) in placing Clupeonia Uochii Valen- ciennes in this synonymy, but have done so tentatively since Valenciennes based his description on the doubtful Clupea sinensis Bloch (see notes on the synonymy of H. toll). Fowler (1941) followed Regan (1917) in separating H. durbanensis and H. brachy- soma from H. kelee, distinguishing the first by its shorter head, and the second by its deeper body. In Figure 7A and B, head length and body depth are plotted (as percentages of standard length) for the specimens listed under Study Material. The figure shows that there is considerable variation (possibly sexual) in body depth, but that this cannot be ascribed to regional variation. In head length, the Durban specimens (i.e. H. durbanensis} have shorter heads than do the Gulf of Aden fishes, but the remaining Indian Ocean specimens are intermediate. On the basis of head length, the Gulf of Aden fishes might be recognised as distinct, but in all other 40 34 32 30 V -o ~D O CO 34- co A REVISION OF THE INDO-PACIFIC ALOSINAE 131 A C*5^) i 1 A o o • X A 38 _ ft 0 0 x 36 — X 5 Q_ B 36 H O O O o ° 32-1 30- 28- 0 100 200 Standard length (mm.) FIG. 7. Body depth (A) and head length (B) as percentages of standard length, plotted against standard length for specimens of Hilsa kelee. • Durban, Kenya coast, Zanzibar. ® Gulf of Aden, x East Indies. A India. 132 P. J. P. WHITEHEAD morphometric and meristic measurements they are inseparable from the rest. I have found no other means of distinguishing H. durbanensis and H. brachysoma, and therefore recognise a single Indo-Pacific member of the subgenus Hilsa. DESCRIPTION. Based on 24 specimens over 50 mm. listed under Study Material. In percentages of standard length : body depth (29-2) 33-3-41-5 (Figure 7 A), head length 27 -1-36-3 (Figure 76) ; snout length (5-7) 6-4-8-5 ; eye diameter 7 -2-9 -5, maxilla length 12-0-16-4 ; operculum, height 15-9-20-6, width 6-9-10-0 ; pectoral length 17-8-21-0, pelvic length 10-3-12-1, anal base 16-7-17-5 ; pre-dorsal 41-7-50-5, pre-pelvic 49-8-57-5. FIG. 8. Hilsa kelee (from Day, Fishes of India, modified). Body strongly compressed, its depth a little greater than head length. Snout a little smaller than eye diameter, pre-orbital length (i.e. including eye) equal to post- orbital length. Lower jaw included when mouth firmly shut ; maxilla length equals pre-orbital length, extending to below middle or posterior part of eye ; exposed portion with longitudinal ridges ; two supra-maxillae. Pseudobranch shape as in Figure 6B ; a groove present below base of pseudobranch. Dorsal with low scaly sheath ; dorsal origin just nearer to snout tip than to caudal base ; pelvic origin below 3rd-5th branched dorsal rays, usually nearer to caudal base than to snout. Pectorals not reaching pelvic base ; in larger fishes, a groove, to receive upper edge of pectoral ; no axillary pectoral scale. Pelvics about three-fifths of pectorals, nearer to pectoral base than to anal origin ; axillary scale present, about three-quarters length of fin. Anal with low scaly sheath ; anal origin slightly nearer to caudal base than to pelvic base. Caudal fin just longer than head, lower lobe longer than upper. No teeth. Cutaneous sensory canals of head branching over suborbitals, pre- operculum, operculum, and extending onto scales behind head. Adipose eyelid with vertical slit exposing three-quarters of pupil. Dorsal surface of head with thin cutaneous covering, numerous fronto-parietal striae exposed, as shown in Figure IA. A REVISION OF THE INDO-PACIFIC ALOSINAE 133 Cleithral lobe prominent (Figure 3A). Gill filaments of outer hemibranch on first arch about half length of those of the inner hemibranch (Figure 3A). Gillrakers of second to fourth arches curled outwards, about 100 short serrae along inner margin of each raker (Figure 46) ; gillrakers on lower part of first arch longer than corres- ponding gill filaments. Opercular bones as in Figure 5A. Dorsal iv 13-14, pelvic i 7, anal iii 17-19. Pre-pelvic scutes 16 (Regan gives 16-18), post-pelvic scutes 12-13 (Regan 11-13), total 28-30. Scales in lateral series 42-45, 13-14 in transverse series (after Regan). Gillrakers fine and numerous, about 100-150 on lower part of first arch. COLOUR IN ALCOHOL : Back and upper parts of head brown, flanks silvery. A dark humeral blotch followed (in some specimens) by seven or eight smaller black blotches. Tips of anterior dorsal rays dusky, caudal tips faintly dusky. DISTRIBUTION : Natal, East African coast, Gulf of Aden, coasts of India, Burma and Siam. The record from Amoy (Wu 1929) was perhaps based on a juvenile of H. reevesii, as also a Cambodia record (Chevey and Le Poulain 1940). SIZE : Up to 220 mm. (Regan). TYPE MATERIAL : The type of Hilsa kelee is Russell's figure of Keelee. Although a poor drawing by modern standards, it is adequate to establish the identity of the species. I have found no reference to any specimens of H. kelee on which Russell might have based his drawing. The type of Alosa brevis Bleeker was apparently lost shortly after its description (Bleeker 1872). The holotype of Clupea platygaster Giinther, a syntype of Alausa kanagurta Bleeker, and the syntypes of Clupea dur- banensis were examined and are listed below. Study Material. i fish, 130 mm., purchd. of Mr. Franks (1862.2.4.10). i fish, 132 mm., purchd. of Mr. Damon (1866.8.14.100). i fish, 98 mm., Bleeker Collection (1867.11.28.25). 5 fishes, 83-117 mm., Madras, coll. F. Day (1889.2.1.1979-83). i fish, 117 mm., Orissa, coll. F. Day (1889.2.1.1986). i fish, 89 mm., Kurrachee, coll. F. W. Townsend (1898.12.24.59). i fish, 48 mm., Sabaki mouth, Kenya (1955.1.18.1). 4 fishes, 129-133 mm., Aden, coll. Fraser-Brunner (1962.3.26.202-205). 5 fishes, 65-91 mm., Jibuti, coll. Fraser-Brunner (1962.3.26.206-210). i fish, 103 mm., Alausa brachysoma Bleeker HOLOTYPE of Clupea platygaster Giinther), Padang, Sumatra, purchd. of Dr. Bleeker (1867.11.28.24). i fish, 132 mm., SYNTYPE of Alausa kanagurta Bleeker, no locality (1867.11. 28.26). 1 fish, 90 mm., " Harengula (Paralosa] zeylanica " of Sale Catalogue (1879), Ceylon? (Rijksmuseum, Leiden, No. 7495) (unpublished Bleeker name). 2 fishes, 169 and 180 mm., SYNTYPES of Clupea durbanensis Regan, Durban (1905.6.8.19-20). 1 fish, 76 mm., Durban, Natal (1919.4.1.1). 2 fishes, 67 and 180 mm., Durban, Natal (1919.9.12.1-2). 134 P. J- P- WHITEHEAD i fish, 116 mm., Durban, Natal (1915.7.6.2). (Dry specimens) i fish, 102 mm., Pinang (Alausa toll from Cantor's collection) (1860.3.19.439). i fish, 130 mm., Zanzibar, Playfair Collection (1867.3.9.371). Subgenus Tenualosa As defined in the key, p. 127. Four species recognised here. KEY TO THE SPECIES of Tenualosa i Caudal lobes as long as head ; pseudobranch rather attenuated and with groove below (Figure 6A). a Upper (exposed) border of suboperculum contained more than twice in depth of operculum ; scales 45-48 in lateral series ; coasts and rivers from Arabia to Burma ......... H. (Tenualosa) ilisha b Upper (exposed) border of suboperculum contained less than twice in depth of operculum ; scales 42-45 in lateral series ; coasts and rivers of China H. (Tenualosa) reevesii n Caudal lobes longer than head ; pseudobranch not attenuated, without ventral groove (Figure 6B). a Maxilla short, not reaching eye centre ; suboperculum almost rectangular, its upper (exposed) border almost horizontal ; cleithral lobe small but present H. (Tenualosa) macrura b Maxilla longer, reaching eye centre or beyond ; suboperculum with rounded posterior margin, upper (exposed) border more steeply inclined, as in Figure 5A ; cleithral lobe barely apparent, as in Figure 3B . . H. (Tenualosa) toll H. macrura, with very long caudal lobes and a short maxilla, is fairly distinctive, but H. ilisha, H. toli and H. reevesii are extremely alike, especially in juvenile stages. In examining stuffed specimens, where the pseudobranch character cannot be checked and where the caudal lobes are often damaged, it is difficult to distinguish between H. toli and H. ilisha. The difference between these two species in scale counts and in the lengths of the maxilla, head and caudal lobes might perhaps be ascribed to mere geographical variation. But the difference in pseudobranch shape, and the presence or absence of a groove below it, appear to be sufficiently consistent for the two species to be kept separate. H. reevesii, on the other hand, may yet prove to be only a subspecies of H. ilisha when more specimens are available. Hilsa ilisha (Ham. Buch.) Clupanodon ilisha Hamilton-Buchanan, 1822, Fishes of the Ganges : 243, 382, pi. 19, fig. 73 (type material from : Ganges estuaries, Patua, Goya Rarra, Calcutta, Dhasa). Alausa ilisha : Cantor, 1850, /. Asiatic Soc. Bengal, 18 : 1282 ; Bleeker, 1852, Verh. Bat. Gen., 24 : 33 ; Kner, 1865, Reise Novarra, Fische : 331. Clupea ilisha : Day, 1878, Fishes of India, pt. 4 : 640, pi. 172, fig. 3 ; Idem, 1889, Fauna Brit. India, Fishes, 1 : 376, fig. 115 ; Lloyd, 1907, Rec. Ind. Mus., 1 : 221 ; Tirant, 1929, Serv. ocdanogr. Peches Indo-Chine, 6e note : 118. A REVISION OF THE INDO-PACIFIC ALOSINAE 135 Clupea (Alosa) ilisha : Steindachner, 1896, Ann. Hofmus. Wien., 11 : 228. Hilsa ilisha : Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 306 ; Fowler, 1934, Proc. Acad. nat. Sci. Philad., 85 : 246 ; Shaw & Shebbeare, 1937, /. Asiatic Soc. Bengal, 3 : 13 ; Berlin, 1940, Bull. Mus. Hist. nat. Paris, (2) 12 : 281 ; Blegvad, 1944, Danish Sci. Invest. Iran, pt. 3 : 63, fig. 28 ; Misra, 1947, Rec. Ind. Mus., 45 (4) : 389 ; Pillay, 1952, Proc. Indo-Pac. Fish. Counc., 3 (Section 2 s 2/8) ; Pillay, 1954, /• Asiatic Soc. Bengal, 20 (i) : 69 ; Idem, 1957, Indian J. Fish., 4 (2) : 345 ; Quereshi, 1957, Agric. Pakistan, 8 (2) : 104, fig. 6a ; Khalaf, 1961, Fishes of Iraq, Baghdad : 17 ; Mahdi, 1962, Fishes of Iraq, Baghdad : n ; Pillay & Rosa, 1963, FAO Fish. Biol. Synopsis No. 25 : i, figs, i (adult), 3 (embryos), 4-6 larvae and juveniles). Macrura ilisha : Fowler, 1941, Bull. U.S. nat. Mus., No. 100 : 633 ; Idem, 1956, Fishes of Red Sea, 1 : 69. Tenualosa ilisha : Munro, 1955, Mar. F-water Fishes Ceylon : 25. Clupea palasah Cuvier, 1829, Regne animal, ed. 2, 2 : 320 (on Palasah Russell, 1803, Fishes of Coromandel, 2 : 77, pi. 198 (type locality, Vizagapatam) ; Gunther, 1868, Cat. Fish. Brit. Mus., 7 : 445- A lausa palasah : Valenciennes, 1847, Hist. Nat. Poiss., 20 : 432 (part, i.e. Ganges and Malabar specimens only) ; Jerdon, 1849, Madras J. Lit. Soc., 15 : 345 ; Day, 1865, Fishes of Malabar : 235- Clupea (Alosa) palasah : Steindachner, 1896, Ann. Hofmus. Wien, 11 : 229. Note on synonymy. The description of Clupea ilisha given by Hamilton-Buchanan (1822) must be presumed to refer to this species, even though the figure shows a fish with the short maxilla (not reaching posterior eye border) and a suggestion of the longitudinal ridges on the maxilla characteristic of H. kelee. However, the latter species can be excluded since Hamilton-Buchanan emphasises the presence of pectoral axillary scales, which are absent in H. kelee ; also, the dorsal view of the head shows no fronto-parietal striae. Hamilton-Buchanan distinguished his species from the Palasah of Russell (1803, p. 77, pi. 198) because of a very slight difference in fin-ray counts and because the Palasah apparently lacked pelvic axillary scales ; such pelvic axillary scales are, however, present in all species of Hilsa. Since Russell also describes a pectoral axillary scale, and since neither Russell nor Hamilton- Buchanan show the long caudal lobes characteristic of H. toll and H. macrura, it must be presumed that both were referring to the present species. Bleeker (1852) similarly made a distinction between Palasah and this species, mainly on pectoral length ; but the pectorals in H. ilisha vary in length (in my material), sometimes reaching the pelvic base and sometimes falling far short of this point. As mentioned earlier, much of the material listed by Gunther (1868, p. 446) under Clupea ilisha can be referred to H. kelee. There is, however, a specimen (skin) from Cantor's collection (225 mm., Ikan Truboh) which may be H. toll ; there are only 40 scales in lateral series (H. ilisha 45-48) but the (damaged) caudal lobes seem to be too short. Unfortunately the diagnostic pseudobranch character cannot be checked. Three of the four stuffed specimens listed by Gunther (p. 445) under Clupea palasah are H. ilisha ; one Gangetic specimen is missing. Of the fishes mentioned by Valenciennes under A lausa palasah, two specimens from the Ganges and one from Malabar are H. ilisha, but a specimen from Bombay 136 P. J. P. WHITEHEAD and one from Pondicherry are H. toll. DESCRIPTION. Based on 24 fishes, 99-325 mm. standard length from the coasts of India, and from the Persian Gulf and the River Tigris (see list of study material) . In percentages of standard length : body depth 31-0-39-5, head length 28-6-33-5 > snout length 5-8-7-3, eye diameter 4-6-6-8, maxilla length 12-2-14-5 ; operculum, height 13-6-15-7, breadth 7-5-9-6 ; pectoral length 18-3-21-5, pelvic length 10-8- 12-8, caudal length 25-7-33-2 ; pre-dorsal distance 47-0-52-0, pre-pelvic distance 51-0-55-0. In relation to standard length, eye diameter shows negative allometry, and there is an indication that pectoral and pelvic lengths may show positive allometry. FIG. 9. Hilsa ilisha (from Day, Fishes of India, modified). Body strongly compressed, its depth variable, a little greater than head length. Snout greater than eye diameter in fishes over about 100 mm., less than eye diameter below this ; pre-orbital length (including eye) two-thirds of post-orbital length. Lower jaw included when mouth firmly shut ; maxilla length just less than pre- orbital length, extending almost to posterior border of eye, exposed portion without longitudinal ridges but occasionally with faint longitudinal striae ; two supra- maxillae. Pseudobranch attenuated, as shown in Figure 6A ; a groove present below border of pseudobranch. Dorsal with low scaly sheath ; dorsal origin just nearer snout tip than base of caudal ; pelvic origin below ist-4th branched dorsal rays, just nearer to caudal base than to snout. Pectorals variable, sometimes reaching pelvic base ; axillary scale present, half length of pectoral. Pelvic fins about three-fifths length of pectorals, nearer to pectoral base than to anal origin ; axillary scale present, about half length of fin. Anal with scaly sheath ; anal origin equidistant between pelvic base and caudal base ; anal longer than in H. reevesii, its base greater than distance snout tip to anterior margin of pre-operculum. Caudal fin equal to or just shorter than head length, lower lobe longer than upper. No teeth. Cutaneous sensory canals of head branching over suborbitals, pre- A REVISION OF THE INDO-PACIFIC ALOSINAE 137 operculum, operculum, and extending onto scales behind head. Adipose eyelid with vertical slit exposing three-quarters of pupil. Dorsal surface of head thickly covered by skin, no fronto-parietal striae, but lateral margins of frontals may show one or two longitudinal ridges through skin (see Figure IB). Cleithral lobe not prominent, barely interrupting cleithral outline (Figure 36). Gill filaments of outer hemibranch on first arch half to three-quarters length of those of inner hemibranch (Figure 36). Gillrakers on all arches straight or slightly curved, not curled outwards, tips pointed, as long as corresponding gill fila- ments in adults (Figure 4A). Bones of opercular series resembling those of H. kelee (Figure 5 A). Dorsal iv-v 14-16, pelvic i 7, anal ii-iii 18-20, branchiostegal rays 5. Pre-pelvic scutes* (14) 15-19, post pelvic scutes (10) 11-15 (I6), total (26) 27-32 (33) (612 fishes). Scales in lateral series (39-44) 45-49 (49) (443 fishes), in transverse series 17-20 ; exposed portion of scales with numerous longitudinal striae, edge of scale pectinated, especially in larger fishes ; minute scales covering caudal, except along posterior border. Gillrakers fine and numerous, 120 (young) to 200 on lower part of first arch (after Regan). Trunk vertebrae 12-13, caudal vertebrae without haemal spines (10) 11-12 (13), caudal vertebrae with haemal spines 20-22 (23), total vertebrae (44) 45-46 (114 fishes). The regressions of various body measurements on standard length for Hooghly river and Chilka Lake specimens were calculated by Pillay (1957), who showed sig- nificant differences between samples from these two localities in five non-meristic characters. Pillay and Rosa (1963), summarising earlier studies, stated that " each major river system, the Chilka Lake and the Saurashtra Coast have their own stocks of hilsa and there is very little, if any, intermingling among them". The differences between these stocks involve meristic as well as non-meristic characters, and in addition, fishermen report differences in taste (dependent probably on fat content). The skull of H. ilisha has been described and figured by Moona (1959), the pharyn- geal pockets by Kapoor (1955), the swim-bladder by Srivastava (1955), blood characteristics by Pillay (1954, 1958), and the branchial skeleton by Khanna (1961). COLOUR : " Silvery, shot with gold and purple ; no spots in the adult, but a row of them along the upper third of the body in the immature " (Pillay & Rosa) ; in alcohol, body brown, sometimes still silvery, upper surfaces darker, fins hyaline. Bionomics, fishery, etc. : H. ilisha has considerable economic importance in Indian waters (especially in the Bay of Bengal) and in Pakistan and Burma. Following a symposium in 1952, current information on the Hilsa fisheries was assembled (/. Asiatic Soc., 20 (i) : 1-79 (1954) ) by the Hilsa Sub-Committee of the Indo-Pacific Fisheries Council. The many papers dealing with the bionomics, life history and fishery of H. ilisha published since then have been usefully summarised by Pillay and Rosa (1963). The esteem in which the ' hilsa ' is held in Bengal is reflected in the many references * Scute, scale and vertebral counts are from Hooghly river and Chilka Lake specimens examined by Pillay (1957) ; rare counts placed in parenthesis. My material lies well within these ranges. 138 P. J. P. WHITEHEAD to its quality and flavour in Sanskrit and Bengali literature,* while Day (1878) and other writers testify to the popularity of the ' sable fish ' amongst Europeans. SIZE : Females attain a larger size than males (Pillay & Rosa, 1963). On the Godavari, Chacko and Ganapati (1949) recorded females of 356-600 mm., and males of 300-432 mm. On the same river, Pillay and Rao (in Pillay & Rosa, 1963) record the smallest mature male as 256 mm., and the smallest mature female 370 mm., but in the Hooghly such mature fishes were smaller (160-170 mm. and 190-200 mm. respectively — see Pillay, 1958). DISTRIBUTION: Persian Gulf (Tigris, Euphrates, Lake Hammar), West Pakistan (Indus), west coast of India, Ceylon, Bay of Bengal, deltaic area of Burma, coastal waters of Cochin China (see Pillay and Rosa, 1963). TYPE MATERIAL : The type of Hilsa ilisha is Hamilton-Buchanan's figure (PI. 19, fig. 73). It was based on a juvenile, which might account for the rather short maxilla. But the anal branched ray count (19) distinguishes it from H. toll, as also does the short caudal. The absence of fronto-parietal striae, and the presence of a pectoral axillary scale rule out H. kelee. Paratypes of Alausa palasah are listed below. Study material. 3 fishes, 115-131 mm., Tigris (1875.1.14.11-13). 2 fishes, 208-290 mm., Calcutta (1889.2.1.1962-3). 9 fishes, 30-115 mm., Orissa (1889.2.1.1964-9) + (3 unregistered). 7 fishes, 52-101 mm., Madras (1889.2.1.1970-5) + (i unregistered), i fish, 212 mm., Canara (1889.2.1.1976). 3 fishes, 68-79 mm-, Sind (1889.2.1.1977-8) + (i unregistered). 1 fish, 323 mm., Bombay (1889.2.1.2022). 6 fishes, 97-146 mm., Sittang R. (1891.11.30.396-401). 2 fishes, 297-323 mm., Ganges, PARATYPES of Alausa palasah Valenciennes (M.N.H.N.Paris No. 3685-6) i fish, 195 mm., Malabar, coll. Dussumier (M.N.H.N.Paris 4976). i fish, 325 mm., Ormara, Meknam Coast (1899.5.8.11). 6 fishes, 99-127 mm., Basra (1920.3.3.178-82) + (i unregistered). 25 fishes, 63-135 mm., Padma R., Bengal (1923.6.30.1-10) + (15 unregistered). i fish, 108 mm., Bengal (1934.10.17.11). (Dry specimens) i fish, 350 mm., R. Tigris (1875.1.14.14). i fish, 360 mm., R. Tigris (1875.1.14.15). fi fish, 370 mm., India (1861.4.2.2). i fish, 410 mm., Madras (1883.11.26.81). fi fish, 340 mm., Suttapore, Ganges (1848.2.1.66). |i fish, 300 mm., Suttapore, Ganges (1848.2.1.65). *For example, hilsa is described as matsyaraja (king of fishes), and elsewhere it is said, Illisah jitapiyusah (Hilsa surpasses nectar) — see Hora 1954. ^Clupea palasah of Giinther (Catalogue, p. 445). A REVISION OF THE INDO-PACIFIC ALOSINAE 139 Hilsa reevesii (Richardson) Alosa reevesii Richardson, 1846, Ichthy. China Japan : 305 (type locality, China seas). Alausa reevesii : Valenciennes, 1847, Hist. Nat. Poiss., 20 : 437 ; Sauvage, 1881, Bull. Soc. Philom. Paris (ser. 7) 5 : 107. Clupea reevesii : Giinther, 1868, Cat. Fish. Brit. Mus., 7 : 446 ; Peters, 1880, Monatsh. Akad. Wiss. Berlin : 926 ; Giinther, 1889, Ann. Mag. nat. Hist. (ser. 6) 4 : (219) 229 ; Elera, 1895, Cat. Fauna Filip., 1 : 583 ; Rutter, 1897, Proc. Acad. nat. Sci. Philad.: 64. Hilsa reevesii : Regan, 1917, Ann. Mag. nat. Hist., (ser. 8) 19 : 306 ; Chu, 1931, Biol. Bull. St. John's Univ., No. 1:13; Fowler, 1934, Proc. Acad. nat. Sci. Philad., 85 : 246 ; Kimura, 1935, /. Shanghai Sci. Inst., (sect. 3) 3 : 104 ; Mori, 1952, Mem. Hyogo Univ. Ag., 1 (3) : 32. Macrura reevesii : Fowler, 1941, Bull. U.S. nat. Mus., No. 100 : 630. Alausa palasah : (non Cuvier) Richardson, 1846, Ichthy. China Japan : 306 ; Cantor, 1849, /. Asiatic Soc. Bengal, 18 : 282 (based on Richardson's specimen). ?Harengula kanagurta : Wu, 1929, Contr. Biol. Lab. Sci. Soc. China, 5 (4) : 17, fig. 13. PClupea (Alosa) thibaudeani : Durand, 1940, Notes Inst. oceanogr. Indochine, 36 : 6, fig. i ; Chevey & Le Poulain, 1940, Trav. Inst. oceanogr. Indochine, 5° mem.: 19. PClupea (Alosa) kanagurta : Chevey & Le Poulain, 1940, Trav. Inst. oceanogr. Indochine, 5e mem. : 19. Note on synonymy. Fowler (1941, p. 630) included Clupea palasah Giinther in this synonymy, but as noted earlier, three of Giinther's four specimens are H. ilisha and the fourth is missing. Durand (1940) compared his Clupea (Alosa) thibaudeani with H. kanagurta, stating that his fishes had a broader operculum and more gillrakers (operculum breadth 1-5 in height, gillrakers 290). But Durand's figure shows a well-defined pectoral axillary scale, an upper suboperculum border rising verj^ steeply and there is enough detail on the dorsal part of the head to suggest that fronto-parietal striae were absent. It seems likely that Durand's specimens belong to Tenualosa, and the high gillraker count, low scale count (40-42) and low dorsal ray count (17) point to H. reevesii ; the broad operculum seems to confirm this. A series of black spots are shown on the flanks, but his specimens were 122-240 mm. and were probably juveniles. The two references to kanagurta in the synonymy (from Amoy and Cambodia) are tentatively placed here, mainly on geographical grounds. DESCRIPTION. Based on five fishes 150-500 mm. standard length from Hong Kong and China (see list of Study Material). Measurements as in H. ilisha. In percentages of standard length : body depth 28-8-33-9, head length 27-4-31-7 ; snout length 6-8-7-5, eye diameter 3-6-6-5, maxilla length 11-2-13-3 ; operculum, height 12-2-15-0, breadth 8-2-9-5 ; pectoral length 18-0-18-8, pelvic length 9-6-11-5, caudal length 28-0-31-2 ; pre-dorsal distance 49-6-50-7, pre-pelvic distance 50-0- 51-7. In relation to standard length, eye diameter shows strong negative allometry ; negative allometry may occur also in body depth and snout length in fishes up to 150 mm. in length. Body strongly compressed, its depth a little greater than head length. Snout greater than eye diameter in fishes over about 150 mm. ; less than eye below this ; pre-orbital length (including eye) about two-thirds of post-orbital length. Lower i4o P. J. P. WHITEHEAD jaw included when mouth firmly shut ; maxilla length just equal to pre-orbital length, extending just to posterior border of eye, exposed portion without long- itudinal ridges but with many fine longitudinal striae in large fishes ; two supra- maxillae. Pseudobranch attenuated, as in Figure 6A ; a groove present below lower border of pseudobranch, more pronounced in larger fishes. Dorsal with low scaly sheath ; dorsal origin about equidistant between snout tip and base of caudal ; pelvic origin below anterior unbranched dorsal rays, equidistant between snout tip and caudal base. Pectorals not reaching pelvic base ; axillary scale present, two- fifths length of pectoral. Pelvic fins half length of pectorals, nearer to pectoral base than to anal origin ; axillary scale present, just less than half length of fin. Anal with scaly sheath ; anal origin equidistant between pelvic tips and caudal base ; anal base shorter than in H. ilisha, less than distance from snout tip to anterior margin of pre-operculum. Caudal fin a little longer than head length, lower lobe longer than upper. No teeth. Cutaneous sensory canals of head branching over sub-orbitals, pre- operculum, operculum, and extending onto scales behind head. Adipose eyelid with vertical slit exposing whole pupil. Dorsal surface of head thickly covered by skin, no fronto-parietal striae, but lateral margins of frontals may show one or two longitudinal ridges through skin (see Figure IB). Cleithral lobe not prominent, scarcely breaking cleithral outline (Figure 36) . Gill filaments of outer hemibranch on first arch half to three-quarters length of those of inner hemibranch (Figure 36). Gillrakers on all arches straight or slightly curved, not curled outwards, tips pointed, as long as or slightly shorter than corresponding gill filaments. Bones of opercular series resembling those of H. ilisha but operculum broader, suboperculum longer, and junction between two rising more steeply ; upper (exposed) border of suboperculum less than twice in height of operculum. Dorsal iv 13-15, pelvic i 7, anal iii 15-17. Branchiostegal rays 5. Pre-pelvic scutes 17*, post-pelvic 14-15, total 31-32. Scales in lateral series 40 (Fowler) 42-45, in transverse series 16-17 (Regan) 1 exposed portion of scales with numerous longi- tudinal striae, edge of scale pectinated in larger fishes ; minute scales on caudal, except along posterior border. Gillrakers fine and numerous, 150 (young) to 250 on lower part of first arch (after Regan). COLOUR : In alcohol, upper surfaces brown, flanks silvery or golden, no spots on upper part of flank in larger fishes, but a faint series in a specimen of 65 mm. " Extremite des pectorals largement rembrunie (en eau formolee) " (Chabanaud 1936). SIZE : Up to 575 mm. standard length. DISTRIBUTION : China Seas, Shanghai, Kiukiang, Hong Kong ; Philippines (Elera's records) ; Korea (Fusan) rare (Mori 1952) ; Cambodia (Durand 1940). TYPE : A mounted skin, as listed under Study Material. * Regan (1917) records 18 + 13-14 for the same material, evidently counting as a pre-pelvic scute the small scute between the bases of the pel vies. A REVISION OF THE INDO-PACIFIC ALOSINAE 141 Relationship to H. ilisha. H. reevesii and H. ilisha are evidently closely related, the former replacing the latter along the coasts of China. On the basis of my (admittedly meagre) material I am keeping the two separate since further small differences besides that of oper- culum shape have emerged from analysis of available data. In H. reevesii the pelvic fins are set slightly nearer the snout and, possibly as a result, the anal origin is equidistant between the pelvic tips and the caudal base, (pelvic base and caudal base in H. ilisha}. In addition, the anal has slightly fewer branched rays in H. reevesii (15-17 ; cf 18-20) and the base of the anal is noticably shorter. Thus, in H. ilisha, the anal base is almost as long or longer than the pectorals and equal or greater than the distance snout tip to anterior pre-operculum margin ; in H. reevesii, the anal base is about two-thirds pectoral length and, measured from snout tip, reaches the posterior border of the eye (juveniles) or a little beyond (adults) . I have included Elera's Philippine records, although these do not appear to have been validated since. If the records of H. ilisha from Cochin China are correct, then the Indo-Malayan Archipelago is not the boundary between these two species. Study Material. i fish, 445 mm., China (1934.3.5.1). i fish, 365 mm., China (1884.2.26.75). i fish, 500 mm., Shanghai (1895.5.31.24) i fish, 65 mm., Amoy, China (1928.6.22.1). i fish, 234 mm., Hong Kong (I939-3-234)- 1 fish, 150 mm., China (coll. Reeves) (1963.8.20.1). 2 fishes, 570-575 mm., Kiu Kiang (1888.3.23.44-45). (Dry specimens) i fish, 295 mm., HOLOTYPE of Clupea reevesii, China (coll. Reeves) (1963.8.20.2). i fish, 140 mm., China (coll. Reeves) (1963.8.20.3). Hilsa macrura (Bleeker) Alausa macrurus Bleeker, 1852, Verh. Bat. Gen., 24 : 31 (on macrura Kuhl & van Hasselt 1823) (type locality, Batavia, Java). Alosa macrurus : Bleeker, 1861, Versl. Meded. Akad. Wet. Amsterdam, 12 : 64 ; Idem, 1868, op. cit., (ser. 2) 2 : 294 ; Idem, 1872, Atlas Ichth. Ind. Neerland, 6 : 113, pi. 264, fig. 4. Clupea macrura (Kuhl & van Hasselt) : Bleeker, 1852, Verh. Bat. Gen., 24 : 31 (name in synonymy) ; Gunther, 1868, Cat. Fish. Brit. Mus., 7 : 448 ; Vinciguerra, 1926, Ann. Mus. Civ. Stor. nat. Genoa (ser. 3) 10 : 619 ; Tirant, 1929, Serv. oceanogr. Peches Indo-Chine, 6e note : 119, 174 ; Hardenberg, 1931, Treubia, 13 (i) : in. Clupea (Alosa) macrura : Weber & de Beaufort, 1913, Fish. Indo-Aust. Arch., 2 : 65 ; Chevey, 1932, Inst. oceanogr. Indo-Chine, ige note : 9. Hilsa macrura : Regan, 1917, Ann. Mag. nat. Hist. (ser. 8) 19 : 307 ; Fowler, 1934, Acad. nat. Sci. Philad., 85 : 246 ; Idem, 1938, List Fish. Malaya : 27. Macrura macrura : Fowler, 1941, Bull. U.S. nat. Mus., No. 100 : 632. i42 P. J. P. WHITEHEAD Note on synonymy. As noted below (p. 145), Bleeker's original description of Alausa macrurus was based entirely on a figure prepared by Kuril and van Hasselt, stated by Bleeker to be 310 mm. (Kuhl and van Hasselt are known to have made their drawings life-size wherever possible). A Kuhl and van Hasselt specimen of 420 mm. in the Leiden Museum is thus too large to have been the fish from which the drawing was made. Dr. Boesman informs me (in litt.} that this fish is labelled " Clupea macrura ", but that there is also a second (old) label " Clufi. palasah Cuv.". He considers the specimen to be Hilsa toll and makes the suggestion that from this specimen a second drawing of Clupea macrura may have been made by Kuhl and van Hasselt, and that it was this second drawing that Valenciennes saw and identified (correctly) as toll. Whatever the truth in this, it seems certain that no Kuhl and van Hasselt specimen of Hilsa macrura exists, and that Bleeker should be accorded authorship of this name. The type of Alausa macrurus Bleeker 1852 is not the Kuhl and van Hasselt drawing on which Bleeker based his description, because this figure was never pub- lished (a further reason for considering Bleeker the true author of this name). It is therefore necessary to select a neotype from amongst Bleeker's material. Bleeker (1853, p. 502) records his first specimen from Batavia, a fish of 342 mm. total length. This fish can be identified with a specimen in the Bleeker collection at the Leiden museum (the larger of two, RMNH 7112) which is 245 mm. standard length and 337+ mm. total length (caudal tips damaged). This specimen is here designated neotype. Dr. Boesman has examined this fish and confirmed its identity for me. The specimen from Bleeker's collection supposed by Giinther (1868, p. 448) to be the type of Clupea macrura appears to be lost. Bertin (1940) listed a supposed cotype of Alausa macrurus Bleeker in the Paris Museum. This fish is Hilsa macrura but probably derives from a later collection by Bleeker than that from which the neotype has been selected. DESCRIPTION. Based on 2 fishes, 141 and 151 mm. standard length from Sarawak ; and one fish 181 mm., a Bleeker specimen (see list of Study Material). In percentages of standard length : body depth (33-7) 377-38-0, head length 24-0-25-4 ; snout length 4-9-5-4, eye diameter 5-5-5-8, maxilla length 8-1-8-8 ; operculum, height n-4-12'9, breadth 5-8-6*4 ; pectoral length 19-9, pelvic length 9-9-11-5, caudal length 46-3-46-5 ; pre-dorsal distance 46-5-48-8, pre-pelvic distance 49-7-51-3. Regan (1917) implies that caudal lobes show positive allometry in re- lation to head length. Body strongly compressed, head length about two-thirds of body depth. Snout less than eye diameter ; pre-orbital length (including eye) almost equal to post- orbital length. Lower jaw included when mouth firmly shut ; maxilla length three- quarters of pre-orbital length, reaching to eye centre ; exposed portion smooth, without longitudinal ridges or striae ; two supra-maxillae. Dorsal with low scaly sheath ; dorsal origin a little nearer to snout tip than to base of caudal ; pelvic origin below 3rd-4th branched dorsal rays, equidistant between snout tip and caudal base. Pectorals not reaching pelvic base ; axillary scale present, almost three- quarter pectoral length. Pelvic fins about half length of pectorals, a little nearer A REVISION OF THE INDO-PACIFIC ALOSINAE 143 to pectoral base than to anal origin ; axillary scale present, three-quarters length of fin. Anal with low scaly sheath ; anal origin equidistant between pelvic base and caudal base. Caudal fin almost twice head length, lower lobe a little longer than upper. No teeth. Cutaneous sensory canals of head branching over suborbitals, pre- operculum, operculum, and extending onto scales behind head. Adipose eyelid with vertical slit exposing whole pupil. Dorsal surface of head covered with thick skin, no fronto-parietal striae, but lateral margins of frontals may show one or two longitudinal ridges visible through skin. Cleithral lobe intermediate between that of H. kelee and H. ilisha, to some extent interrupting cleithral outline. Gill filaments of outer hemibranch of first arch three-quarter length of those of inner hemibranch. Gillrakers on all arches straight or slightly curved, with a small distal knob (Figure 40) , half length of corres- ponding gill filaments. Operculum bones as in H. kelee but suboperculum more rectangular, and the line of its (exposed) border with the operculum nearer the hori- zontal (if projected, cutting the dorsal profile far behind the dorsal fin). Dorsal iv-v 15, pelvic i 7, anal iii 17, branchiostegal rays 5. Pre-pelvic scutes 17 (Regan 16-18, but see footnote p. 140), post-pelvic 14, total 31. Scales in lateral series 45, in transverse series 14-15 (Regan) ; minute longitudinal striae along posterior border of scale ; minute scales covering caudal lobes except along posterior border. Gillrakers fine and numerous, 60-80 on lower part of anterior arch (Regan). COLOUR : In alcohol, dorsal surfaces brown, flanks silvery or golden, a faint dark humeral spot visible ; fins colourless except for faint dark edge to caudal. SIZE : 350 mm. (Regan). DISTRIBUTION : Java, Sumatra, Bankalis, Borneo, Singapore (Weber and de Beaufort, 1913). According to the latter authors " it forms the object of a very important fishery at the mouth of some rivers in Borneo, Malacca and Sumatra ". TYPE : Neotype, a fish 245 mm. standard length (total length ca 337 mm., but caudal tips damaged), Reg. No. RMNH 7112, in the Bleeker collection at the Rijks- museum van Natuurlijke Historic, Leiden, identified as Bleeker's first specimen from Batavia (see above, p. 142). Study Material. i fish, 151 mm., Sarawak (1868.6.9.2). i fish, 141 mm., Sarawak (1895.2.28.72). i fish, 181 mm., East Indies, coll. Bleeker (M.N. H.N.Paris No. 2039). Regan (1917) records a specimen of 350 mm. but I have been unable to find this fish. Hilsa toll (Valenciennes) Alausa toli Valenciennes,* 1847, Hist. Nat. Poiss., 20 : 435 (type locality, Coromandel, Pondi- cherry) ; Cantor, 1849, /. Asiatic Soc. Bengal, 18 : 1281. *Following Bailey (1951) I have throughout cited Valenciennes as sole author of volume 20 of the Histoire Naturelle des Poissons, 1847. 144 p- J- p- WHITEHEAD Alosa toli : Bleeker, 1872, Atlas Ichth. Ind. Norland., 6 : 113, pi. 265, fig. 4. Clupea toli : Giinther, 1868, Cat. Fish. Brit. Mus., 7 : 447 ; Day, 1878, Fishes of India, pt. 4 : 641 ; Idem, 1889, Fauna of Brit. India, 1 : 377 ; Duncker, 1904, Mitt. Naturhist. Mus. Hamburg, 21 : 186 ; Pillay, 1929, /. Bombay nat. Hist. Soc., 33 : 355; Tirant, 1929, Serv. ocSanogr. Peches Indo-Chine, 6e note : 119 ; Hardenberg, 1931, Treubia, 13 (i) : no ; Idem, op. cit., 15 (3) : 231. Clupea (Alausa) toli : Martens, 1876, Preuss. Exped. Ost-Asien, 1 : 405. Clupea (Alosa) toli : Weber & de Beaufort, 1913, Fishes Indo-Aust. Arch., 2 : 64 ; Chevy, 1932 (Aug. 25th), Inst. ocdanogr. Indo-Chine, ige note : 9. Sardinella toli : Jordan & Evermann, 1902, Proc. U.S. nat. Mus., 25 : 166 ; Jordan & Richard- son, 1909, Mem. Carnegie Mus., 4 : 166. Hilsa toli : Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 306; Hora, 1923, /. nat. Hist. Soc. Siam, 6 : 174 ; Chu, 1931, Biol. Bull. St. John's Univ., No. i : 14 ; Herre & Myers, 1937, Raffles Mus. Bull., No. 13 : 12 ; Smith, 1945, Bull. U.S. nat. Mus., No. 188 : 44 ; Misra, 1947, Rec. Ind. Mus., 45 (4) : 390 ; Quereshi, 1957, Agric. Pakistan, 8 (2) : 104, fig. 7. A lausa palasah : Valenciennes, 1847, Hist. Nat. Poiss., 20 : 432 (part, i.e. Bombay and Pondi- cherry specimens). ?Clupea macroura (Kuhl & van Hasselt) Valenciennes, 1847, Hist. Nat. Poiss., 20 : 437 (name only ; misspelt). Alausa argyrochloris Valenciennes, 1847, Hist. Nat. Poiss., 20 : 440 (type locality : He de France) (the single Dussumier specimen, not the two Quoy & Gaimard specimens — see below) . Alausa ctenolepis Bleeker, 1852, Natuurk. Tijdschr. Ned. Ind., 3 : 74 (type locality, Batavia, Muntok, Singapore). Clupea chapra : (non Ham.-Buch.), Giinther, 1868, Cat. Fish. Brit. Mus., 7 : 447 ; Beavan 1877, F -water Fish. India : 118. Alosa ilisha : (non Ham.-Buch.), Bleeker, 1874, Ned. Tijdschr. Dierk., 4 : 148. PClupea ilisha : Giinther (part), 1868, Cat. Fish. Brit. Mus., 7 : 445. Hilsa sinensis : Fowler, 1930, Proc. Acad. nat. sci. Philad. (1929) : 592, 598 ; Idem, 1938, List Fish. Malaya : 28. Tenualosa sinensis : Munro, 1955, Mar. F-water Fishes Ceylon : 24. Macrura sinensis : Fowler, 1941, Bull. U.S. nat. Mus., No. 100 : 631 ; Chu & Tsai, 1958, Quart. J. Taiwan Mus., 2 (1-2) : 104. Note on synonymy. Fowler (1930) identified this species with Clupea sinensis Bloch 1795 and later (Fowler 1941) included it in the synonymy of Clupea sinensis Linnaeus. But Linnaeus' description of C. sinensis, which appears for the first time in the tenth edition of the Systema Naturae, is too vague for a positive identification to be made (Valenciennes (1847) comments on this and concludes that Clupea sinensis Linnaeus, together with Clupanodon sinensis Lacepede " doit etre raye" de nos catalogues ichthyologiques "). Of possible Chinese clupeids with moderate anal fins (i.e. about 16 rays) the gizzard shads can be eliminated since a filamentous last dorsal ray is not mentioned ; the dussumieriids Dussumieria and Etrumeus have too many branchiostegal rays (6 in C. sinensis} and Spratelloides has too few anal rays (16 in C. sinensis) ; Sardinops and Clupea are possible, but Linnaeus states that C. sinensis is similar to Clupea harengus " sed latior ". Daubenton (1787, p. 202) translates this as " sa largeur est plus considerable ". It is the possibility that depth rather than overall size was meant which prevents exclusion of either Sardinella or Herklotsichthys from consideration. A REVISION OF THE INDO-PACIF1C ALOSINAE 145 Herklotsichthys (Harengula auct.) and Sardinella are possible since both have the appearance of truncated posterior branchiostegal rays, a character stressed in Linnaeus' description. In both Herklotsichthys and Sardinella, certain species or individuals either lack, or have feebly developed teeth (" os edentulum " in C. sinensis) . On the other hand, neither Herklotsichthys nor Sardinella attain the size of Hilsa. Within Hilsa itself, H. reevesii is also a possibility. But since Clupea sinensis does not appear amongst the Linnaean type specimens at Uppsala University listed by Lonnberg (1896) and in view of the inadequacy of the original description I have here preferred to use Valenciennes' name toll. C. sinensis L. of Bloch (1795) does not help to identify Linnaeus' fish. Bloch's figure shows a fish with a black mark at the base of the anterior dorsal rays, which is characteristic of some species of Sardinella. There is also a black border to the dorsal and the caudal, such as is found for example in S. melanura or S. sindensis. Valenciennes (loc. cit.) doubted Bloch's identification, believed Bloch's fish to have been different from Linnaeus', and called it Clupeonia blochii [placed by Day (1878) and Fowler (1941) in the synonymy of H. kanagurta (i.e. H. kelee]. Day (loc. cit.), however, quotes Prof. Peters (in. litt.}, who had examined a dried specimen and believed it to be the model for Bloch's figure, stating " I am of (the) opinion C. sinensis Bloch is C. toll, Cuv. and Val.". Peters also examined this specimen for Giinther, who considered it identical to Hamilton-Buchanan's Clupanodon ilisha (Giinther, 1868, p. 446). Unfortunately, most of Gunther's ' ilisha ' material is clearly H. kanagurta (see p. 130). I do not know whether this specimen is still extant, but the shortness of the maxilla and indeed of the whole head in Bloch's figure, as well as the distinctive black marks on the fins, are much more in accord with Herklotsichthys or Sardinella than with Hilsa. It can be noted, however, that Valenciennes (loc. cit. p. 436) describes black dorsal and anal borders in his Alausa toll (based on Dussumier's notes). The operculum is too narrow for H. reevesii and the caudal is too short for H. toll. The Clupea sinensis (or Hareng de la Chine) of Bonnaterre (1788), Daubenton (1788), Ray (1788), Gmelin (1789), and Walbaum (1792) were based on Linnaeus' description ; those of Martens (1876) and Schneider (1801) were based on Bloch. A further problem concerning this synonymy is the identity of Kuhl and van Hasselt's macrura, a drawing of which both Valenciennes and Bleeker saw. Valen- ciennes (1847, p. 437) believed this fish to be his Alausa toll. However, Bleeker (1852, p. 32) maintained Kuhl and van Hasselt's name, showing that Valenciennes' toll had a longer head. In the Atlas, Bleeker (1872*) definitely separated the two species (on head length, jaw length and scales). Fowler (1941) tentatively placed Clupea macroura (K. v.H.) of Valenciennes in his synonymy for H. toli, while still accepting Bleeker 's macrura as distinct. I have followed the same course, for reasons given in more detail under H. macrura (p. 142). The single Dussumier specimen of Alausa argyrochloris Valenciennes is Hilsa toli, but the two Quoy and Gaimard specimens from Mauritius are Sardinella, probably S. dayi Regan. Bertin (1940) was therefore wrong to assume the latter to be para- *I have here accepted the Atlas dating as reconstructed by Mees (1962). I46 P. J. P. WHITEHEAD topotypes of H. toli. Valenciennes' Bombay and Pondicherry specimens of Alausa palasah are also H. toli. FIG. 10. Hilsa toli (from Day, Fishes of India, modified). DESCRIPTION. Based on n fishes, 83-396 mm. standard length from the coasts of India and the Indo-Malayan Archipelago (including one of Bleeker's two specimens of Alausa ctenolepis, 420 mm. total length — see list of Study Material). In percentages of standard length : body depth 32-6-39-0, head length 24-8-28-0 ; snout length 5-3-6-5, eye diameter 4-3-7-3, maxilla length 11-2-12-5 '> operculum, height 12-2-14-3, breadth 6-2-7-3 (8-6) ; pectoral length (16-2) 19-2-21-2, pelvic length (9-3) 10-3-13-1, caudal length 33-3-37-6 ; pre-dorsal distance 44-0-49-8, pre- pelvic distance 49-2-54-5. In relation to standard length, eye diameter shows strong negative allometry, and both pectoral and pelvic fins show positive allometry. In these specimens, smaller fishes have slightly deeper bodies. Body strongly compressed, its depth greater than head length. Snout greater than eye diameter in fishes over about 120 mm., less than eye diameter in juveniles ; pre-orbital length (i.e. including eye) four-fifths of post-orbital length (equal in fishes of 100 mm. or less). Lower jaw included when mouth firmly shut ; maxilla length equal or (in large fishes) greater than pre-orbital length, extending beyond posterior eye border ; exposed portion without longitudinal ridges but occasionally with one or two faint longitudinal striae ; two supra-maxillae. Dorsal with low scaly sheath ; dorsal origin just nearer to snout tip than to caudal base ; pelvic origin below 4th~7th branched dorsal rays, equidistant or just nearer to caudal base than to snout. Pectorals not reaching pelvic base ; axillary scale present, half length of pectoral. Pelvic fins about half length of pectorals, just nearer to pectoral base than to anal origin ; axillary scale present, about two-thirds length of fin. Anal with low scaly sheath ; anal origin just nearer to caudal base than to pelvic base. Caudal fin a little longer than head length, lower lobe longer than upper. No teeth. Cutaneous sensory canals of head branching over sub-orbitals, pre- operculum, operculum, and extending onto scales behind head. Adipose eyelid with vertical slit exposing all or three-quarters of pupil. A REVISION OF THE INDO-PACIFIC ALOSINAE 147 Dorsal surface of head thickly covered by skin, no fronto-parietal striae, but lateral margins of frontals may show one or two longitudinal ridges visible through skin. Cleithral lobe not prominent, barely interrupting cleithral outline. Gill fila- ments of outer hemibranch on first arch half to three-quarters length of those of inner hemibranch. Gillrakers on all arches straight or slightly curved, not curled outwards, tips pointed, as long as corresponding gill filaments in adults. Bones of opercular series as in H. kelee. Dorsal iv-v 14-15, pelvic i 7, anal iii 15-17, branchiostegal rays 5. Pre-pelvic scutes 17 (Regan, 17-18), post-pelvic scutes 12-13 (Regan, 11-13), total 29-30. Scales in lateral series about 40, 14-15 in transverse series ; exposed portion of scales with numerous longitudinal striae, edge of scale pectinate, especially in larger fishes ; minute scales covering caudal, except along posterior border. Gillrakers fine and numerous, 70-95 on lower part of first arch (after Regan). The swim-bladder of H. toll has been described by Nayak and Bal (1955), and the food and feeding habits by Chacko (1949). COLOUR : " Silvery, shot with yellow and purple, a dark shoulder spot in young " (Day, 1878). In preserved material (alcohol), the back is brown, the flanks silver or golden, and the shoulder spot is very faint in the smaller specimens. In two small fishes (83 and 108 mm.) the caudal lobes are edged in brown ; in the remainder, the fins are hyaline. SIZE : 460 mm. DISTRIBUTION : India, Malay Peninsula, Singapore, Pinang, East Indies, Siam, Formosa, China (Fowler, 1941). TYPE MATERIAL : Bertin (1940) listed for this species only the holotype and paratypes of Alausa argyrochloris Valenciennes (discussed above). Dr. Blanc in- forms me that there are in fact two (dry) Valenciennes specimens in the Paris Museum (460 mm. ex Pondicherry, coll. Leschenault, No. 3939 ; and 440 mm. ex Bombay, coll. Roux, No. 3940), and has kindly examined them for me. Head lengths (24-3 and 25-0 per cent, of standard length respectively) confirm that the specimens are H. toll and not H. ilisha. Unfortunately, the caudals are both damaged (21-6 and 25-0 per cent, of S.L.) and the scale counts (39 and 39) are likely to be a little low. Since the locality and collector of the first fish (No. 3939) are those first mentioned by Valenciennes, this specimen is here designated lectotype. Study Material. *2 fishes, 83-108 mm., (Waterhouse collection — no locality) (1858.8.15.68-9). i fish, 121 mm., HOLOTYPE of Alausa argyrochloris (M.N. H.N.Paris No. 2738). 1 fish, 312 mm., HOLOTYPE of Alausa ctenolepis Bleeker, no locality (1867.11.28.23). |i fish, 216 mm., Pondicherry, coll. Belanger (M.N. H.M.Paris No. 3687). 2 fishes, 122-124 mm., Orissa (1889.2.1.1984-5). 2 fishes, 276-396 mm., Bombay (1889.2.1.2018-19). 2 fishes, 110-163 nun., Bombay (1889.2.1.2020-21). *Gunther's Clupea chapra (Catalogue, p. 447). 148 P. J. P. WHITEHEAD |i fish, 314 mm., Bombay, coll. Dussumier (M.N. H.N.Paris No. 3684). (Dry specimens) ? i skin, 225 mm., Ikan Truboh (Cantor collection) (1860.3.19.438). Gudusia Fowler Gudusia Fowler, 1911, Proc. Acad. nat. Sci. Philad., 63 : 207 (type, Clupanodon chapra Hamilton-Buchanan) . Indo-Pacific alosinid fishes with very small scales, 80-100 in lateral series, 27-35 in transverse series. Scales adherent. Sub-operculum subrectangular or cres- centic, its junction with the operculum, if produced, cutting the dorsal outline near or in front of the dorsal origin (Figure 55). Branchiostegal rays 6. Gillrakers numerous, 200 or more on lower part of first arch. Dorsal rays 14-16, pelvic 8, anal 20-24. Ventral scutes, 18-20 pre-pelvic and 8-10 post-pelvic. Two species recognised here, from the rivers of India and Burma. 1 Depth less than 40 per cent, of S.L.; head more than 28 per cent.; anal iii 19-22 G. chapra 2 Depth more than 40 per cent, of S.L. ; head less than 28 per cent. ; anal iii 22-26 G. variegata Gudusia chapra (Ham. Buch.) Clupanodon chapra Hamilton-Buchanan, 1822, Fishes of the Ganges : 248, 383, (type locality* Upper Ganges). Alosa chapra : Gray, 1832-34, Illustr. Indian Zool. Hardwicke, 2 : pi. 92, fig. 2. Alausa chapra : Valenciennes, 1847, Hist. Nat. Poiss., 20 : 440. Clupea chapra : Day, 1869, Proc. zool. Soc. London : 385 ; Idem, 1878, Fishes of India, pt. 4 : 639, pi. 161, fig. i ; Idem, 1889, Fauna British India, Fishes, 1 : 375. Gudusia chapra : Fowler, 1911, Proc. Acad. nat. Sci. Philad., 63 : 207 ; Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 307 ; Fowler, 1938, List. Fish. Malaya : 26 ; Idem, 1941, Bull. U.S. nat. Mus., No. 100 : 634. Gadusia chapra : Chaturvedi, 1959, Indo-Pac. Fish. Counc. Occ. Paper 60/2 : 2 (misspelling). Clupea indica Gray, 1832-34, Illustr. Indian Zool. Hardwicke, 2 : pi. 19, figs. i-2x ; Giinther, 1868, Cat. Fish. Brit. Mus., 7 : 444 ; Beavan, 1877, F -water Fish. India : 118. PClupea champil Gray, 1832-34, Illustr. Indian Zool. Hardwicke, 2 : pi. 91, figs. 5-6. PPellona champil : Valenciennes, 1847, Hist. Nat. Poiss., 20 : 324 (on Gray). Alausa microlepis Valenciennes, 1847, Hist. Nat. Poiss., 20 : 439 ; Bleeker, 1853, Verh. Bat. Gen., 25 : 146. Clupea suhia Chaudhuri, 1912, Rec. Ind. Mus., 7 : 436, pi. 38, fig. i. Note on synonymy. Giinther's description of Clupea chapra seems to refer to a species of Hilsa (scales 42, 13 post-pelvic scutes, anal 19) ; the two specimens listed by Giinther are H. toll. Alausa champil Cantor is based on two (dry) specimens which are in fact Kowala thoracata Valenciennes. Clupea champil Gray is based on a drawing which Cantor (1850) believed to be an " indifferent copy " of one in Hamilton-Buchanan's dupli- cate (and unpublished) series, labelled by Buchanan " Clupea champil B." The drawing can only doubtfully be identified with Gudusia chapra. Some Indian workers have referred to Gadusia, but this is incorrect. •(•Valenciennes' Alausa palasah (1847, p. 433). A REVISION OF THE INDO-PACIFIC ALOSINAE 149 DESCRIPTION. Based on 10 fishes, 92-140 mm. standard length from Allahabad (Ganges) and Gowhatty (see list of Study Material). In percentages of standard length : body depth 32'5-4O-o (see also Figure 13), head length 28-2-30-2 ; snout length 5-0-6-3, eye diameter 7-3-8-5, maxilla length 10-4-12-6 ; operculum, height 13-2-13-6, breadth 6-8-7-3 ; pectoral length 20-2-21-2, pelvic length 11-3-11-7, caudal length 30-0-30-3 ; pre-dorsal distance 48-6-52-5, pre-pelvic distance 49-0-51-7. In relation to standard length, eye diameter and maxilla length show negative allometry. FIG. 1 1 . Gudusia chapra (from Day, Fishes of India, modified) . Body strongly compressed, its depth greater than head length. Snout less than eye diameter ; pre-orbital length (including eye) just less than post-orbital length. Lower jaw included when mouth shut ; maxilla length four-fifths of pre-orbital length, extending to below eye centre or posterior rim of pupil ; exposed portion without longitudinal ridges or striae ; two supra-maxillae. Pseudobranch attenu- ated, with groove below, as in Hilsa ilisha (Figure 6A). Dorsal without low scaly sheath ; dorsal origin more or less equidistant between snout tip and base of caudal ; pelvic origin below unbranched dorsal rays or just in front, a little nearer to snout tip than to caudal base. Pectorals almost reaching pelvic base ; axillary scale present, one-third to one-quarter length of fin. Pelvic fins half length of pectorals, nearer to pectoral base than to anal origin ; axillary scale present, about one-third length of fin. Anal fin without scaly sheath ; anal origin equidistant between pelvic base and caudal base ; anal base slightly shorter than in G. variegata, about three- quarters to four-fifths pectoral length and usually shorter than the distance snout tip to posterior margin of pre-operculum. Caudal fin just less than head length, lower lobe longer than upper. No teeth. Cutaneous canals of head branching over sub-orbitals, pre-operculum, operculum and onto scales behind head. Adipose eyelid with vertical slit exposing three-quarters of pupil. Dorsal surface of head covered by skin, but fronto-parietal striae exposed (Figure ic). Cleithral lobe not prominent, scarcely breaking cleithral outline. Gill fila- 150 P. J. P. WHITEHEAD ments of outer hemibranch of first arch half to three-quarters length of those of inner hemibranch. Gillrakers on all arches straight or slightly curved, not curled outwards, about as long as corresponding gill filaments in adults. Bones of oper- cular series as in Figure 5B, the upper (exposed) margin of the suboperculum rising at a steep angle (to cut dorsal profile before dorsal origin if projected). Dorsal iv 11-13 (first unbranched ray minute), pelvic i 7, anal (ii) hi 19-22, branchiostegal rays 6. Pre-pelvic scutes 17-19, post-pelvic 10 (n), total 27-29. Scales in lateral series 75-100, 27-34 in transverse series (Regan) ; exposed edge of scale smooth ; minute scales covering caudal except along posterior border. Gillrakers fine and numerous, 200 or more on lower part of first arch (Regan). The pharyngeal pockets of G. chapra have been described by Kapoor (1954, 1957), the morphology of the swim-bladder by Srivastava (1956), and the structure of the alimentary tract by Srivastava (1957). COLOUR : In alcohol, back brown, flanks silvery or golden, a series of faint black spots along upper flank ; caudal edge black. SIZE : 140 mm. (up to 8 inches — Day, 1889). DISTRIBUTION : " Freshwaters of rivers and tanks in Sind and throughout India as far south as the Kistna river ; absent from the Malabar coast and Madras " (Day, 1889). Study Material. i fish, 85 mm., Cachar (Assam) (1867.2.14.36). i fish, 122 mm., Gowhatty (1889.2.1.1952). i fish, 102 mm., Lahore (1889.2.1.1953). 1 fish, 76 mm., Goalpara (1889.2.1.1954). 2 fishes, 69-76 mm., Brahmaputra (1889.2.1.55-56). 5 fishes, 37-105 mm., Orissa (1889.2.1.57-61). 17 fishes, 46-140 mm., Allahabad, Ganges (1934.10.17.1-10) (7 unregistered). 2 fishes, 41-49 mm., Assam (1963.8.23.1-2). 2 fishes, 99-111 mm., India (1963.8.23.3-4). (Dry specimens) i fish, 148 mm., Ganges (1848.2.1.67). i fish, 151 mm., no locality, India House collection (1860.3.19.676). Gudusia variegata (Day) Clupea variegata Day, 1869, Proc. zool. Soc. London : 623 (type locality, Irrawaddy river, Burma) ; Idem, 1878, Fishes of India, pt. 4 : 639, pi. 161, fig. 4 ; Idem, 1889, Fauna Brit. India, Fishes, 1 : 375 ; Vinciguerra, 1890, Ann. Mus. Civ. Stor. nat. Genova, (ser. 2) 9 : 350 ; Lloyd, 1907, Rec. Ind. Mus., 1 : 221 ; Jenkins, 1910, Rec. Ind. Mus., 5 : 138. Gudusia variegata : Regan, 1917, Ann. Mag. nat. Hist., (8) 19 : 308 ; Myers, 1924, Amer. Mus. Nov., No. 150 : i ; Prashad & Mukerji, 1929, Rec. Ind. Mus., 31 (3) : 209 ; Fowler. 1941, Bull. U.S. nat. Mus., No. 100 : 635. DESCRIPTION. Based on a single fish, 155 mm. standard length from Bassein (Burma) . A REVISION OF THE 1 N DO-PAC I F I C ALOS1XAE 151 In percentages of standard length : body depth 41-6, head length 27-7 ; snout length 5-9, eye diameter 6-4, maxilla length 11-5 ; operculum, height 13-3, breadth 6-9 ; pectoral length 18-2, pelvic length 11-3, caudal length 30-1 ; pre-dorsal distance 49-5, pre-pelvic distance 52-2. FIG. 12. Giulusia variegata (from Day, Fishes of India, modified). Body strongly compressed, its depth one and a half times head length. Snout less than eye diameter ; pre-orbital length (including eye) three-quarters of post- orbital length. Lower jaw included when mouth shut ; maxilla length almost equal to pre-orbital length, extending just beyond eye centre ; exposed portion without longitudinal ridges or striae. Pseudobranch attenuated, with groove below as in Hilsa ilisha (Figure 6A). Dorsal with very small scaly sheath ; dorsal origin about equidistant between snout tip and base of caudal ; pelvic origin below unbranched dorsal rays, a little nearer to snout tip than to base of caudal. Pectorals not reaching pelvic base ; axillary scale present, about half length of fin. Pelvic fins a little over half length of pectorals, nearer to pectoral base than to anal origin ; axillary scale present, almost half length of fin. Anal fin with very low scaly sheath ; anal origin equidistant between pelvic base and caudal base ; anal base exceeds pectoral length and exceeds distance snout tip to posterior border of pre-operculum. Caudal fin a little greater than head length, lower lobe longer than upper. Teeth absent, except minute teeth on tongue. Cutaneous sensory canals of head branching over suborbitals, pre-operculum, operculum and onto scales behind head. Adipose eyelid with vertical slit exposing three-quarters of pupil. Dorsal surface of head covered by skin, but fronto-parietal striae exposed (see Figure ic). Cleithral lobe not prominent, scarcely breaking cleithral outline. Gill filaments of outer hemibranch of first arch half to three-quarters length of those of inner hemibranch. Gillrakers on all arches straight or slightly curved, not curled outwards, about as long as corresponding gill filaments. Bones of opercular series as in Figure 50, the upper (exposed) margin of the sub-operculum rising at a steep P. J. P. WHITEHEAD angle (to cut dorsal profile before dorsal origin if projected) ; sub-operculum almost crescentic. Dorsal iv 12 (first unbranched ray minute), pelvic i 7, anal iii 22, branchiostegal rays 6. Pre-pelvic scutes 19, post-pelvic n, total 30. Scales in lateral series 90, 32 in transverse series (Regan) ; exposed part of scale with a single prominent vertical striation ; edge of scale pectinated ; minute scales covering caudal except along posterior border. COLOUR : In alcohol, brown on back, flanks golden, a series of brown spots along upper flank, some expanded vertically, those behind dorsal extending right across back ; dark spot at base of posterior dorsal rays ; caudal tips colourless (" tipped with black " — Day 1869). SIZE : 155 mm. (7 inches, Day 1869). DISTRIBUTION : Rivers of Burma. 42 40 oo 38 £.36 <U O 34 CO 32 40 60 80 100 120 Standard length (mm.) 140 160 FIG. 13. Body depth as a percentage of standard length, plotted against standard length for species of Gudusia. X G. variegata • G. chapra A REVISION OF THE 1NDO-PACIFIC ALOSINAE 153 Relationship to G. chapra. The differences separating this species from G. chapra are small, and it is possible that, when more specimens are available, G. variegata will be regarded merely as an eastern form or subspecies of G. chapra. The difference in body depth may perhaps be due to allometric growth, since the specimen of G. variegata is the largest examined. In Figure 13 body depth as a percentage of standard length is plotted against stan- dard length for 30 specimens of G. chapra and i of G. variegata. The points are rather scattered, but the graph suggests that the body is relatively deeper in fishes below 50 mm. and also in those above 120 mm. Other differences between G. variegata and G. chapra are the smooth-edged scales in the latter and perhaps colouration. THE WEST AFRICAN GENUS ETHMALOSA As noted earlier, the West African genus Ethmalosa resembles the Indo-Pacific Alosinae (Hilsa and Gudusia), and differs from the Atlantic, Mediterranean and New World Alosinae, in two characters, Regan's gillraker character and pelvic fin ray count (8 ; cf 7 or 9). However, the weight attached to these two characters as indicators of probable phyletic relationships must first be viewed in the light of possible zoogeographical relationships. In general, the clupeoid fauna of West Africa is poor in both species and genera when compared with that of either the New World or the Indo-Pacific. Thus, there are no West African gizzard shads (Dorosomatinae) , or round herrings (Dussum- ieriidae), and only a single engraulid genus, the widespread Engraulis, with a single species close to the North Atlantic E. encrasicolus (Whitehead 1964!)). In the subfamily Clupeinae, records of Sardina pilchardus have probably been based on one or two stray fishes from the North Atlantic. Sardinella, however, with two species, S. aurita Valenciennes and S. maderensis (Lowe), is principally an Indo-Pacific genus. 5. maderensis (for which S. eba [Valenciennes], S. granigera Valenciennes and S. cameronensis Regan are probably synonyms) occurs only off the West African coast and in the Mediterranean, but S. aurita is found along the Atlantic coast of America, in the Mediterranean, and also in the Western Pacific (Japan, Philippines). S. aurita is, however, replaced in the Indian Ocean by 5. longiceps Valenciennes. Harengula, a Western Atlantic genus, is not represented in West Africa ; H. rouxi Poll is a species of Sardinella (Whitehead I964a). The subfamily Pellonulinae shares no genera with the Indo-Pacific. Finally, in the subfamily Pristigasterinae, the genus Ilisha is shared both with the Indo-Pacific and with the Western Atlantic. But Tucker (1954) believed the West African species, /. africana (Bloch), to be more closely allied to the species and genera of the Atlantic coast of South America than to any Indo-Pacific form. He felt that there might one day be a case for recognising the distinctness of the West African /. africana from the Indo-Pacific forms by the creation of a new subgenus. Thus, only two West African clupeid genera are shared with the Indo-Pacific. But they are also the only two Western Atlantic clupeid genera which are found in West Africa ; and at species level both show closer links with the Western Atlantic 154 P- J- 1' VH-1 I THREAD than with the Indian Ocean. On zoogeographic grounds, therefore, Ethmalosa would be expected to show less affinity with the Indo-Pacific than with the New World genera. Against this argument, there is further morphological evidence linking Ethmalosa with the Indo-Pacific Alosinae. The fronto-parietal striation pattern is identical to that found in Hilsa kelee (but it also resembles that in Brevoortia). Perhaps more significant, the rather characteristic gillrakers on the upper parts of all arches and the lower parts of the third and fourth arches in Ethmalosa may well represent merely an extreme form of the curled gillrakers found in Hilsa kelee. The pseudo- branch in Ethmalosa also resembles that of H. kelee. On the other hand, the maxilla in Ethmalosa is smooth (but ridged in Brevoortia} and the cleithral lobe is little developed (as in Brevoortia, Alosa). On present evidence therefore, the question must remain open. Ethmalosa may have been derived from the Western Atlantic, or from the Indian Ocean, or it might represent an independent relict of a once widespread alosinid fauna. Until this problem can be solved, a tribal division in the Alosinae would be unrealistic. ACKNOWLEDGEMENTS It is a pleasure to record my thanks to Dr. M. Boesman, of the Rijksmuseum van Natuurlijke Historic (Leiden) for his help in the problem of Hilsa macrura. The manuscript was read by Dr. P. H. Greenwood, to whom go my sincere thanks for suggestions and criticisms. Finally, I must express my gratitude to Dr. M. Blanc of the Museum National d'Histoire Naturelle, Paris, who so willingly lent me Valen- ciennes' specimens and other material, and kindly undertook the measurements on type specimens of H. toll. REFERENCES BAILEY, REEVE M. 1951. The authorship of names proposed in Cuvier and Valenciennes' " Histoire Naturelle des Poissons". Copeia, No. 3 : 249-251. BERTIN, L. 1940. Catalogue des types de poissons du Museum National d'Histoire Naturelle. Bull. Mus. Hist, nat., (2) 12 (6) : 244-322. BLEEKER, P. 1852. Bijdrage tot de Kennis der Haringachtige visschen van den Soenda- Molukschen Archipel. Verh. Bat. Gen. Batavia, 24 : 1-52. - 1853. Diagnostische beschrijvingen van nieuwe of weinig bekende vischsoorten van Batavia. Nat. Tijdschr. Ned. Ind., 4 : 451-516. - 1872. Atlas Ichthyologique des Indes Orientates Neerlandaises, 6 : 61-140. BLOCK, M. E. 1795. Naturgeschichte der AusldndischenFische, 9 : 1-192^18.325-432. BONNATERRK, J. P. 1788. Tableau encyclopediqtic et methodique, Ichtliyologie, i-lvi, 1-215, 102 pis. CANTOR, T. 1850. Catalogue of Malayan fishes. J . Asiatic Sac. Bengal, 18 : 983-1443. CHABANAUD, P. 1936. Les poissons de Mer exotique aux Halles centrales de Paris — Kemarques sur diverses especes, notamment Solea senegalensis Kaup. et Hilsa reevesi H.B. Bull. Mus. Hist. nat. Paris, 8 (i) : 61-64. CHACKO, P. I. 1949. Food and feeding habits of the fishes of the Gulf of Manaar. Proc. Ind. Acad. Sci., 29B : 83-97. CHACKO, P. I. & GANAPATI, S. W. 1949. On the bionomics of Hilsa ilis/ia (Ham.) in the Godavari river. /. Madras Univ., 18 : 16-22. CHEVEY, P. & LE POULAIN, F. 1940. La peche dans les eaux donees du Cnmboge. Trav. Inst. oc6anogr. Indochine, 5° mem.: 1-193. CUVIER, G. 1829. Regne Animal, (edition 2) 2 : 1-406. A REVISION OF THE INUO-PAC1FIC ALOSINAE 155 DAUBENTON, L. J. M. 1787. Encyclopedic Methodique, Histoire Naturelle, 3 (Poissons) : i-lx, 1-435. DAY, F. 1869. On the freshwater fishes of Burma. Proc. zool. Soc. London : 614-623. - 1878. The fishes of India, Williams & Norgate, London, 778 pp. — - 1889. The fauna of British India — Fishes, Taylor & Francis, London, 548 pp. DURAND, J. 1940. Notes sur quelques poissons d'especes nouvelles ou peu connues des eaux douces cambogiennes. Notes Inst. ocianogr. Indochine, 36 : 1-40. FOWLER, H. W. 1930. Notes on Japanese and Chinese fishes. Proc. Acad. nat. Sci. Philad., 81 : 589-616. - 1934. Descriptions of new fishes obtained 1907 to 1910 chiefly in the Philippine Islands and adjacent seas. Proc. Acad. nat. Sci. Philad., 85 : 233-367. - 1941. Contributions to the biology of the Philippine Archipelago and adjacent regions. Bull. U.S. nat. Mus., No. 100 (13) : 1-879. - 1958. Some new taxonomic names for fishlike vertebrates. Notul. Naturae, Philad., No. 310 : 1-16. GMELIN, J. F. 1789. Systema Naturae Linnaeus, 1 (3) : 1033-1516. GUNTHER, A. 1868. Catalogue of the fishes in the British Museum, London, 7B : 273-512. HAMILTON-BUCHANAN, F. 1822. An account of the fishes of the Ganges, Edinburgh : 1-405, pis. 1-59. HORA, S. L. 1954. Proverbs and popular sayings concerning hilsa fish current in Bengal. /. Asiatic Soc. (Sci.), 20 (i) : 19-27. KAPOOR, B. G. 1954. The pharyngeal pockets and its associated structures in Gadusia chapra (Ham.). Curr. Sci., 23 : 162-163. - 1955. The anatomy and histology of the pharyngeal organ in Hilsa ilisha (Ham.). /. zool. Soc. India, 6 : 167-172. 1957. The pharyngeal pockets of Gadusia chapra (Ham.). Jap. J. Ichthyol., 5 : 132-135. KHANNA, S. S. 1961. The hyobranchial skeleton of some fishes. Ind. Jour. Zoot., 2 (i) : 1-55. LONNBERG, E. 1896. Linnaean type specimens of fishes in the Royal University of Uppsala. Bihang K. Svenska Vet. Akad. Handlingar., 22 (4) : 3-45. MARTENS, E. 1876. Die Preussiche Expedition nach Ost-Asien, Berlin, 1 : 1-412. MEES, G. F. 1962. A preliminary revision of the Belonidae. Zool. Verhandl. Leiden, 54 : 1-92. MITRA, G. N. & DEVASANDARUM, M. P. 1954. On the hilsa of the Chilka Lake with a note on the occurrence of hilsa in Orissa. /. Asiatic Soc. (Sci.), 20 (i) : 33-40. MOONA, J. C. 1959. The skull of Hilsa ilisha Hamilton. Sci. Cult., 24 (2) : 97-98, figs, i, 2. MORI, T. 1952. Checklist of the fishes of Korea. Mem. Hyogo Univ. Agric., 1 (3) : 1-228. MUNRO, I. S. R. 1955. The marine and freshwater fishes of Ceylon. Canberra, Dep. of External Affairs : 1-349. NAYAK, K. P. D. & BAL, D. V. 1955. The air-bladder and its relation with the auditory organ in Hilsa toli (Cuv. & Val.). /. Univ. Bombay N.S., 23B (5) : 53-66. PILLAY, T. V. R. 1954. Morphological and serological characters of the hilsa, Hilsa ilisha (Ham.) with special reference to racial investigations. /. Asiatic Soc. (Sci.), 20 (i) : 69-74. - 1957. A niorphometric study of the populations of hilsa, Hilsa ilisha (Hamilton) of the river Hooghly and of the Chilka Lake. Indian J. Fish., 4 (2) : 344-386. - 1958. Morphological and physiological characters of the blood of hilsa, Hilsa ilisha (Hamilton) of the river Hooghly. Proc. Indian Acad. Sci., B47 (3) : 155-162. PILLAY, S. R. & ROSA, H. 1963. Synopsis of biological data on hilsa, Hilsa ilisha (Hamilton) 1822. FAO Fisheries Biology Synopsis No. 25, Rome. RAY, P. A. F. 1788. Histoire Naturelles des Animaux, Paris, pp. 1—710. REGAN, C. T. 1916. Fishes from Natal, collected by Mr. Romer Robinson. Ann. Durban Mus., 1 (3) : 167-170. 1917- A revision of the clupeoid fishes of the genera Pomolobus, Brevoortia and Dorosoma, and their allies. Ann. Mag. nat. Hist., (8) 19 : 297-316. 156 P. J. P. WHITEHEAD RUSSELL, P. 1803. Description and figures of two-hundred fishes collected at Vizagapatam on the coast of Coromandal, W. Bulmer & Co., London, 2 : 1-85. SCHNEIDER, J. G. 1801. Systema Ichthyologiae Blochii : 1-584. SMITH, J. L. B. 1949. The sea fishes of Southern Africa, Central News Agency, S. Africa, 550 pp. SRIVASTAVA, P. N. 1955. Features of histological interest in the air-bladder and pneumatic duct in Hilsa ilisha (Ham.). Curr. Sci., 24 : 96. - 1956. Morphology and histology of the air-bladder of Hilsa ilisha (Hamilton) and Gadusia chapra (Hamilton) and their connection with the internal ear. Proc. nat. Inst. Sci. India, 22B : 28-33. - 1957. Primitive features in the alimentary canal of Gadusia chapra (Hamilton). Curr. Sci., 27 : 144-145. SVETOVIDOV, A. N. 1952. (Clupeidae). Tabl. anal. Faune U.R.S.S. N.S. No. 48 2 (No. i) : I-33I- TUCKER, D. 1954. The " Rosaura " Expedition, 4 : Report on the fishes . . . Bull. Brit. Mus. nat. Hist. (Zool.), 2 (6) : 163-214. VALENCIENNES, M. A. 1847. Histoire Naturelle des Poissons, Paris, 20 : 1-472. WALBAUM, J. J. 1792. Petri Artedi sued Genera Piscium (Ichthyologiae 3) : 1-723. WEBER, M. & DE BEAUFORT, L. F. 1913. The fishes of the Indo-Australian Archipelago, E. J. Brill, Leiden, 2 : 1-404. WHITEHEAD, P. J. P. 1963. A revision of the recent round herrings (Pisces : Dussumieriidae) . Bull. Brit. Mus. nat. Hist. (Zool.), 10 (6) : 305-380. - ig64a. Herklotsichthys Whitley 1951 to replace Harengula Valenciennes 1847 for Indo- Pacific species (Pisces : Clupeidae). Ann. Mag. nat. Hist., (13) 6 : 273-284. - ig64b. New data extending the range of the ' antitropical ' anchovy genus Engraulis into the tropics (in Russian), Zool. Zh., 43 (6) : 879-888. WHITLEY, G. P. 1948. New sharks and fishes from Western Australia. Pt. 4. Aust. Zool. Sydney, 11 : 259—276. - 1953. Studies in ichthyology No. 16. Rec. Aust. Mus., 23 : 133-138. Wu, H-W. 1929. A study of the fishes of Amoy, Pt. i. Contr. Biol. Lab. Sci. Soc. China, 5 (4) : 1-90. PRINTED IN GREAT BRITAIN BY THOMAS DE LA RUE & COMPANY LIMITED LONDON T * A CRITICAL REVIEW OF THE MARINE NEMATODE GENUS EUCHROMADORA DE MAN, 1886 JOHN W. COLES BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) ZOOLOGY Vol. 12 No. 5 LONDON: 1965 A CRITICAL REVIEW OF