^ V ■ ^ ' lit' >*- .t >■ vq i.,/>»<; '**-''ii.rf V /' ^^V ■ • ^i*^^-' N -- V T.>, vV' '.t"-*x:'K ^«' " '^ -'i .^Sfi ^v? J--:.>. :-:^i^rv.l^.^^^ jjp. ,>*.. V-' >' .;.^s:'l >^'i'm. -mif* '.t„. DISCOVERY REPORTS VOLUME XII Cambridge University Press Fetter Lane, London Neiv Tork Bombay, Calcutta, Madras Toronto MacmiUan Tokyo Maruzen Company, Ltd All rights reserved DISCOVERY REPORTS Issued by the Discovery Committee Colonial Office, London on behalf of the Government of the Dependencies of the Falkland Islands //sr ^ VOLUME XII CAMBRIDGE AT THE UNIVERSITY PRESS 1936 PRINTED IN GREAT BRITAIN BY WALTER LEWIS, M.A., AT THE UNIVERSITY PRESS, CAMBRIDGE ■^r^ CONTENTS COAST FISHES. PART I By J. R. Norman Introduction . Cape Verde Islands West Africa . South Africa Ascension Island Tristan da Cunha GouGH Island . Appendix — Exocoetidae POLYCHAETE WORMS. II (published 31st January, 1936) By C. C. A. Monro, M.A. Introduction Station List List of Species Geographical Distribution Systematic Account THE SOUTH ATLANTIC (published 9th December, 1935) page 3 4 7 35 54 57 58 58 page 61 6i . 69 72 • 76 References 195 ECHINOIDEA AND OPHIUROIDEA (published 12th March, 1936) By Th. Mortensen Introduction page 201 Echinoidea 209 Ophiuroidea 236 Index 347 Plates I-IX following page 348 THE BIRDS OF THE SOUTH ORKNEY ISLANDS (published 28th February, 1936) By R. A. B. Ardley, R.N.R. Introduction page 351 Systematic Account 352 List of Literature • 3'76 Plates X-XI I following page ^'id LARVAE OF DECAPOD CRUSTACEA (published 2nd September, 1936) By Robert Gurney, D.Sc. Introduction page 379 Part I. Stenopidea 379 Part II. Amphionidae 392 Part HI. Phyllosoma 400 LIST OF PERSONNEL December, 1936 DISCOVERY COMMITTEE The Rt. Honourable The Earl De La Warr {Chairman) Sir S. F. Harmer, K.B.E., Sc.D., F.R.S. {V ice-Chairman), British Museum J. O. Borley, O.B.E., M.A., Fisheries Adviser G. H. Creasy, Colonial Office Representative Vice-Admiral Sir Percy Douglas, K.C.B., C.M.G. Rear-Admiral J. A. Edgell, C.B., O.B.E., R.N., Admiralty Sir J. Fortescue Flannery, Bt., M.I.C.E. {Consulting Naval Architect) Dr S. W. Kemp, F.R.S. H. G. Maurice, C.B., Ministry of Agriculture and Fisheries Sir J. Middleton, K.C.M.G., K.B.E. J. M. Wordie, M.A., Royal Geographical Society Secretary: F. H. Harper, M.B.E. Technical Officer: H. Horsburgh, A. M.I.C.E., Crown Agents for the Colonies Shipping Officer: E. A. Nattriss, O.B.E., Crown Agents for the Colonies Accountant: G. Bryant Clerical Assistants: Miss B. M. Borley Miss W. J. Hope Miss M. E. Jeffries Miss D. B. Meggitt SCIENTIFIC STAFF Director of Research N. A. Mackintosh, A.R.C.S., D.Sc. H. F. P. Herdman, M.Sc. J. W. S. Marr, M.A., B.Sc. J. E. Hamilton, M.Sc. G. W. Rayner, B.Sc. E. R. Gunther, M.A. F. D. Ommanney, A.R.C.S., Ph.D. A. J. Clowes, A.R.C.S., M.Sc. T. J. Hart, D.Sc. G. E. R. Deacon, B.Sc. A. H. Laurie, M.A. Curators: Miss H. E. Bargmann, Ph.D. Miss D. M. E. Wilson Clerical Assistant: Miss S. M. Isaacson MARINE EXECUTIVE STAFF R.R.S. 'Discovery //' Lieut. L. C. Hill, O.B.E., R.N.R., Captain Lieut. R. Walker, R.N.R., Chief Officer Sub-Lieut. T. H. B. Gates, R.N.R., Third Officer Lieut. H. Kirkwood, R.N.R., Second Officer Sub-Lieut. V. A. J.-B. Marchesi, R.N.R., Fourth Officer Eng. Lieut. -Comdr. W. A. Horton, R.N. (Retd.), Chief Engineer A. N. Porteous, Second Engineer R. G. Gourlay, Third Engineer J. R. Strong, M.R.C.S., L.R.C.P., Surgeon R.R.S. ' William Scoresby Lieut. C. R. U. Boothby, R.N.R., Captain Lieut. R. C. Freaker, R.N.R., Chief Officer Lieut. A. F. Macfie, R.N.R., Senior Second Officer Sub-Lieut. A. S. Marshall, R.N.R., Second Officer B. Dales, Chief Engineer J. F. Warnock, Second Engineer [Discovery Reports. Vol. XII, pp. 1-58, December, 1935.] COAST FISHES PART I. THE SOUTH ATLANTIC By J. R. NORMAN Department of Zoology, British Museum (Natural History) ^/^/a CONTENTS Introduction page 3 Cape Verde Islands 4 West Africa 7 South Africa 35 Ascension Island 54 Tristan da Cunha 57 Gough Island 5^ Appendix — Exocoetidae 5^ I COAST FISHES PART I. THE SOUTH ATLANTIC^ (INCLUDING THE CAPE VERDE ISLANDS, WEST AFRICA, SOUTH AFRICA, ASCENSION ISLAND, TRISTAN DA CUNHA AND GOUGH ISLAND) By J. R. Norman Department of Zoology, British Museum (Natural History) (Text-figs. I -1 5.) INTRODUCTION THE collection of coast fishes^ obtained by the R.R.S. 'Discovery' and the R.R.S. ' William Scoresby ' is so large that it has proved necessary to divide the report on these fishes into three sections, of vv^hich the present one is the first. The second part ^ A certain number of fishes were also obtained by the expedition in New Zealand. Of these, some were collected by the R.R.S. 'Discovery II'; others were presented by the N.Z. Fisheries, Ltd., WelHngton, and were selected from specimens in the fish market ; and a small series of freshwater fishes was obtained through the courtesy of the New Zealand Marine Department. These New Zealand fishes will form a very valuable addition to the collections of the British Museum, but it has not seemed necessary to deal with them in detail here. A list of the species represented is given below, and references to nearly all these are to be found in the check-list of New Zealand fishes published by Phillipps in 1927 {N.Z. Marine Dept., Fisheries Bull., No. i). Trachiinis novae- zelandiae, Richardson Dactylopagrus macropterus (Schn.) Parapercis colias (Schn.) Hemerocoetes pauciradiatus, Regan Thyrsites atun (Euphrasen) Gobiomorphiis gobioides (Cuv. and Val.) Favonigobius lateralis (Macleay) Tripterygion variiim (Schn.) Helcogramma tnedium (Giinth.) Genypterus hlacodes (Schn.) Agonostomus forsteri (Schn.) Seriolella brama (Giinth.) Pterygotrigla picta (Giinth.) Chelidonichthys kumu (Lesson and Garnot) Lepidotrigla brachyoptera, Hutton Lopho?iectes gallus, Giinth. Pelotretis flavilatus, Waite Peltorhamphus novae-zeelandiae , Giinth. Rhombosolea retiaria, Hutton Rhombosolea leporina, Giinth. Meuschenia convexirostris (Giinth.) ^ The term "coast fishes" was suggested by Regan (1914, Rep. Brit. Antarct. {'Terra Nova') Exped., 1910, Zool., I, p. 24) to include not only the littoral forms but also fishes that may occur at no great distance from the coasts in water down to two or three hundred fathoms deep, and are not pelagic or bathypelagic. The true oceanic fishes have been dealt with in a previous report (Norman, 1930, Discovery Reports, 11, pp. 261-370, pi. ii, 47 text-figs.). Geotria australis, Gray Eptatretus cirr/iatus (Schn.) Mustelus manazo, Bleek. ? Raja nasiita, Miill. and Henle Trygon brevicaudata, Hutton Argentina elongata, Hutton Retropinna retropinna (Richardson) Galaxias fasciatiis, Gray Galaxias attenuatus (Jenyns) Prototroctes oxyrhynchus, Giinth. Gonorhynchus gonorhynchus (Linn.) Chlorophthalmiis nigripinnis, Giinth. Conger conger (Linn.) Ariosoma habenata (Richardson) Macrtironus novae-zelandiae (Hector) Physiciiliis bachus (Schn.) Hoplopteryx affinis (Giinth.) Zeus faber, Linn. Poly prion oxygeneios (Schn.) Caesioperca lepidoptera (Schn.) Usacaranx lutescens (Richardson) 4 DISCOVERY REPORTS will include the important collections made during the trawling survey of the Magellan- Falkland Islands region, and will also deal with the fishes of the coast of Chile. The fishes of South Georgia, the South Orkneys, the South Shetlands, and those of the Antarctic Continent, will form the subject of the last section of the report. On her first commission the ' Discovery ' made brief calls at Ascension and Tristan da Cunha on her way to the south, and a few fishes were obtained at both these islands. In July 1927 she called at the Cape, and while the ship was in dock at Simon's Town Mr E. R. Gunther and Mr F. C. Fraser were able to make a trip lasting about five days in one of the South African commercial trawlers (S.T. 'Richard Bennet') and to preserve a valuable collection of deep-water fishes obtained with the trawl. On the homeward voyage in 1927 the 'Discovery' made several hauls off the coasts of Angola and French Congo, and off Annobon, obtaining a very interesting series of fishes, several of which have proved to be new to science. After leaving the Gulf of Guinea a call was made at the Cape Verde Islands, where a few fish were collected. In June 1927 the 'William Scoresby ' was at Gough Island, and one or two fishes were obtained from here. All the text-figures accompanying this report are the work of Lieut.-Col. W. P. C. Tenison, D.S.O., who has prepared the drawings with his customary care and skill. The author takes this opportunity of expressing his sincere thanks to the members of the Discovery Committee for permission to study these collections and to undertake the preparation of this report. CAPE VERDE ISLANDS Examples of only five species were collected here, but three of these are of particular interest. BELONIDAE Belone ardeola (Cuv. and Val.). Belone ardeola, Cuvier and Valenciennes, 1846, Hist. Nat. Poiss., xvni, p. 425. Belotie trachiira, Cuvier and Valenciennes, 1846, t.c, p. 456; Troschel, 1866, Arch. Naturgesch., xxxii (i), p. 234; Gunther, 1866, Cat. Fish., vi, p. 235; Fowler, 1919, Proc. U.S. Nat. Mus., LVi, pp. 196, 217, fig. I. ? Belone depressa, Poey, 1856-8, Mem. H.N. Cuba, n, p. 296. Belone depressa, Gunther, 1866, t.c, p. 235. Belone lovii, Gunther, 1866, t.c, p. 236. Tylosurus ardeola, Jordan and Evermann, 1896, Bull. U.S. Nat. Mus., XLvn (i), p. 713; Ever- mann and Marsh, 1902, Bull. U.S. Fish. Comm., xx (1900), p. 99. Belone (Tylosurus) ardeola, Metzelaar, 1919, Trap. Atlatit. Vissch., pp. 29, 218. Strongylura ardeola, Nichols and Breder, 1928, Zoologica, N.Y., vni, p. 423; Breder, 1932, Carnegie Inst. Washington, Publ. 435, p. 7, figs. ? Strongylura longleyi, Breder, 1932, t.c, p. 12, pi. ix, text-figs. St Vincent. 2. ix. 27. Hand line: 2 specimens, 360, 370 mm. Length of head (to tip of upper jaw) 2f to nearly 3 J in that of fish (without caudal). Diameter of eye rather more than interorbital width and i^ to 2 in postorbital part of CAPE VERDE ISLANDS 5 head. 7 to 9 gill-rakers on lower part of anterior arch, no to 130 scales from occiput to origin of dorsal fin. Dorsal 13-17. Anal 18-21. Origin of pectoral usually nearer to last ray of anal than to head, sometimes equidistant from them. Caudal peduncle strongly depressed. Hab. West Indies ; Azores ; Cape Verde Islands ; Tropical West Africa ; St Helena ; Ascension. The presence of gill-rakers and the comparatively slender jaws place this species in the genus Belone, Cuv.^ On comparing three examples from the Cape Verde Islands, including the type of B. lovii, Giinther, with six from the West Indies {B. depressa, Giinther), I am unable to find any essential differences, and there is little doubt that the same species is to be found on both sides of the Atlantic. I have followed American authors in identifying this species with B. ardeola of Cuvier and Valenciennes, originally described from Martinique, but a re-examination of the types of this and other species of Gar-fishes described by the French authors is badly needed. The form recently described by Breder as Strongylura longleyi is said to differ from Belone ardeola only in the longer head, of which the depth is less than the width, and the somewhat larger eye. Of the West Indian specimens examined by me some appear to be referable to Breder's species, but I am of the opinion that the study of a large series of specimens would reveal this to be, at the most, a subspecies of B. ardeola. I have also examined twelve examples of B. trachura from Ascension and St Helena, and find that these differ from both West Indian and Cape Verde Islands specimens of B. ardeola only in the slightly higher number of dorsal and anal rays and perhaps in the larger number of predorsal scales. The eye is a little smaller than that of the Cape Verde Islands specimens, but agrees very well with some of the West Indian examples. On the whole, I think it best to recognize two subspecies: B. ardeola ardeola from the West Indies, Azores and Cape Verde Islands;^ and B. ardeola trachura from Ascension and St Helena. The measurements of head and eye, and the fin-ray and scale counts, are shown in the accompanying table. West Indies (6) Cape Verde Islands (3) Ascension and St Helena (12) Head in length of fish Eye in postorbital part of head Number of predorsal scales Number of dorsal rays Number of anal rays 2|t0 3 if to 2 (2t^) 112-114(130) (12)13-14(15) (16) 18-19 2f to 31 4 to if (2) 110-112 (129) 13-14 18(19) 2| to 3I (li)l|t0 2 (no) 120-130 (14) 15-17 19-21 1 See Regan, 1911, Ann. Mag. Nat. Hist. (8) vii, p. 332. ^ Examples presumably belonging to this subspecies have been recorded from Senegambia (Rochebrune) and St Thome (Osorio). 6 DISCOVERY REPORTS CARANGIDAE Decapterus punctatus (Cuv.). Norman, 1935, Ami. Mag. Nat. Hist. (10) xvi, p. 254. St Vincent. 2. ix. 27. Hand line: i specimen, 222 mm. Hab. Atlantic coast of America, from Cape Cod to Brazil, occasionally farther north ; Cape Verde Islands ; Tropical West Africa. POMACENTRIDAE Glyphisodon hermani, Steind. Glyphidodon {Parma) hermani, Steindachner, 1888, SitzBer. Akad. Wiss. Wien, xcvi (i), (1887), p. 59, pi. iii. St Vincent. 2. ix. 27. Hand line: i specimen, 205 mm. This species is new to the British Museum collection. I have nothing to add to Steindachner's excellent description and figure. SCORPAENIDAE Scorpaena laevis, Troschel.' 1866, Arch. Naturgesch., xxxii (i), p. 206. St Vincent. 2. ix. 27. Hand line: i specimen, 175 mm. Depth of body 3 in the length, length of head 2\. Depth of head at origin of dorsal fin 1 1^ to I ^ in its length. Snout about as long as eye, diameter of which is 4I to 5 in length of head and equal to or a little greater than interorbital width. Scales on cheek and on opercular region, but no obvious pores. A deep occipital pit. Praeorbital spines strong, the anterior with a secondary spine at its base; 3 well-developed spines on suborbital ridge. Supraorbital tentacle absent in the smaller specimen, little larger than the tentacle above the anterior edge of the eye and about \ diameter of eye in the larger; a number of other membranous processes on head in the larger specimen. Length of palatine band of teeth about \ diameter of eye. 9 gill-rakers on lower part of anterior arch. Scales cycloid ; about 37 to 40 in a longitudinal series above lateral line, 4 or 5 between last soft-ray of dorsal and lateral line ; breast scaled ; a number of membranous processes scattered over body, especially on the back and in the region of the lateral line. Dorsal XI-XII 9 ; third and fourth spines longest, 2\ to 2| in length of head. Anal III 5; second spine stronger and a little longer than third. Pectoral with 19 rays, extending to above vent or a little beyond, its length^ i^ to if in that of head ; base broad, the lowermost ray inserted level with the root of the pelvic spine and well in advance of uppermost. Greyish brown, with irregular darker markings, of which those on the fins tend to form bars ; inner surface of pectoral with large dark brown or black spots, those near the base of the fin uniting to form cross-bars. Hab. Cape Verde Islands. In addition to the small specimen collected by the Discovery Expedition, there is a much larger one in the collection of the British Museum, 295 mm. in total length, 1 See also p. 30 of this report. ^ Measured from upper angle. CAPE VERDE ISLANDS 7 which has been included in the above description. This species is very close to S. senegalensis, Steindachner/ of which I have examined a single specimen from the Gold Coast, 280 mm. in total length. The Cape Verde species appears to have a shorter, deeper head, with a somewhat shorter snout, and the spines on the head are generally stouter and less acute. The supraorbital tentacle is smaller and less branched, the pectoral fin is a little shorter, and there are minor differences in the coloration. Fowler ^ has given a detailed description of a specimen, 295 mm. in length, from the Cape Verde Islands, identified by him as S. senegalensis, but it seems probable that his fish is referable to S. laevis. Both species are readily distinguished from S. porcus, Linn., S. scrofa, Linn., and S. tisttilata, Lowe, by the smooth scales, scaly breast, and other Fig. I . Scorpaena laevis. X 5. characters. S. plumieri, Bloch, from the Atlantic coast of tropical America, is closely related to S. laevis and S. senegalensis, but has the axil of the pectoral fin jet black with a few pure white spots, and there are other differences in coloration. Further, the eye is somewhat smaller and the spinous dorsal fin lower. DACTYLOPTERIDAE Dactylopterus volitans (Linn.). St Vincent. 2. ix. 27. Hand line: i specimen, 300 mm. WEST AFRICA Fishes were obtained from five stations, three off the coast of Angola, one off the French Congo, and one off Annobon, Gulf of Guinea. None of the specimens are from a depth of more than 100 m. Altogether nearly 300 specimens were collected from these stations, representing about 50 species, of which 9 have proved to be new to science. 1 1881, Denkschr. Akad. Wiss. Wien, XLiv, p. 31, pi. iv. - 1919, Proc. U.S. Nat. Mus., LVi, p. 214. 8 DISCOVERY REPORTS CARCHARINIDAE Mustelus laevis, Risso. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m.: i male specimen, 700 mm. SQUATINIDAE Squatina oculata, Bonaparte. 1840, Icon. Faufie Ital. (28), fig.; Lozano Rey, 1928, Fauna Iberica, Feces, p. 494, pi. vii. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: 2 male specimens, 380, 560 mm. These speciinens clearly belong to this Mediterranean species, which is also found on the southern and eastern coasts of Spain and has been beautifully figured by Lozano Rey in his monograph of the Selachians of Spain. S. fimbriata, Miiller and Henle, of the Mediterranean, has been regarded as synonymous with S. oculata, but there can be little doubt that it is identical with S. aculeata, Cuvier. Regan^ recorded a large specimen of Squatina from Lagos as S. africana, Regan, a Natal species, but a re-examination of this specimen shows that it belongs to S. oculata. The two species are very closely related, but S. oculata has smaller spiracles than S. africana, with fringed anterior margins, the lobes of the caudal fin more acutely pointed, and the dorsal fins narrower and more acute. There are also differences in the coloration. The large example re- corded by Metzelaar^ from Goree is clearly referable to 5. oculata. The following represents a synopsis of the European and African species of Squatina : L A mid-dorsal series of enlarged denticles in the adult, forming a row of sharp spines along the back aculeata, Cuy.[= fimbriata, M. and Yi.] IL No mid-dorsal series of enlarged denticles in the adult. A. Dermal denticles not carinate, a large patch on lower surface between pectoral fins; distance from anterior angle to posterior end of base of pectoral \ or nearly \ the extreme length of the fin squatina, 'Linn. { = angelus,'Dnm.'\ B. Dermal denticles 3-7 carinate, none on lower surface between pectoral fins; distance from anterior angle to posterior end of base of pectoral a little more than \ the extreme length of the fin. 1. Width of spiracle equal to or less than diameter of eye, its anterior margin fringed; lobes of caudal fin (in adult) more or less acutely pointed oculata, Bonap. 2. Width of spiracle greater than diameter of eye, its anterior margin not fringed; lobes of caudal fin (in adult) more obtusely pointed africana, Regan TORPEDINIDAE Torpedo torpedo (Linn.). Torpedo narce, Gunther, 1870, Cat. FisJi., vni, p. 449; Pellegrin, 1914, Ami. lust, oceanogr. Monaco, vi (4), p. 8. Narcacion torpedo, Garman, 1913, Mem. Mus. Camp. ZooL, xxxvi, p. 306. Torpedo torpedo, Lozano Rey, 1928, Fauna Iberica, Feces, p. 518, pi. ix, fig. i. 1 1915, Ann. Mag. Nat. Hist. (8) xv, p. 124. '' 1919, Trop. Atlant. Visscli., p. 191. WEST AFRICA 9 St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 4 specimens, 180-260 mm. Hab. Mediterranean and neighbouring parts of the Atlantic. RAJIDAE Raja miraletus, Linn. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 2 male specimens, 435. 440 mm. Raja sp. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: one empty egg- capsule. SYNODONTIDAE Saurida parri, Norman. 1935, Proc. Zool. Soc, p. 126, fig. 15. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 2 specimens, 37, iismm.i St. 279. ID. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m. : 12 specimens, 80-125 mm. Hab. Off the coasts of Angola and French Congo. MURAENIDAE Muraena unicolor (Delaroche). St. 271. 30. vii. 27. Off Elephant Bay, Angola. Shore collection: i specimen, 355 mm. This fish is in rather poor condition, but appears to belong to this species. Muraena sp. St. 283. 14. viii. 27. Off Annobon, Gulf of Guinea. Large dredge, 18-30 m.: 14 specimens, 40-120 mm. GADIDAE Bregmaceros maclellandi, Thompson. Norman, 1930, Discovery Reports, 11, p. 339. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : i specimen, 51 mm. ZEIDAE Zeus faber, Linn. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: 3 specimens, 240-290 mm. SERRANIDAE Epinephelus goreensis (Cuv. and Val.). St. 271. 30. vii. 27. Elephant Bay, Angola. Hand line, 18 m.: 2 specimens, 265, 345 mm. Epinephelus aeneus (Geoffr.). St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m. : i specimen, 600 mm. 1 The larger of these is the holotype. 10 DISCOVERY REPORTS Neanthias accraensis, Norman. 193 1, Ann. Mag. Nat. Hist. (10) vii, p. 354, fig. 2. St. 279. 10. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m.: i specimen, 130 mm. This species was known previously only from the Gold Coast. In life the belly is said to be silvery and the upper part of the body darker. The head is ornamented with 3 or 4 broad oblique yellow bands. The median fins are mottled with yellow, and the soft dorsal is fringed with red. Rhegma guineensis, sp.n. St. 283. 13. viii. 27. Off Annobon, Gulf of Guinea. Large dredge, 18-20 m.: i specimen, 29 mm. Very close to R. thaimoshim, Gilbert, from the Pacific coast of Panama, but without supraocular tentacles. Depth of body about 3 i in the length, length of head 2\. Snout shorter than eye, diameter of which is 4 in length of head and more than twice the Fig. 2. Rhegma guineensis ?■ Holotype. X4. interorbital width. Maxillary extending to a very little beyond posterior margin of eye; lower jaw a little projecting; no distinct canine teeth in jaws, but some of the anterior teeth apparently somewhat enlarged. A single broad spine on the praeoper- culum; two broad, flat spines embedded in the operculum. About 5 gill-rakers on lower part of anterior arch. 48 (?) scales in lateral line. Dorsal VII 20. Anal III 16 or 17. Pectoral extending nearly to above first anal spine, a little shorter than head. Brownish above, rather paler below ; a round dark spot, a little smaller than the eye, on the operculum ; sides of head with traces of two dark lines ; dorsal, anal and caudal fins blackish ; other fins paler. One other species of this genus has been described, namely Caribrhegma gregoryi, Breder, from the Glover Reef off the coast of British Honduras. In this species supra- ocular tentacles are present, and the dorsal has 15 soft-rays, the anal 12. The two flat spines on the operculum said to be characteristic of Caribrhegma are to be found also in Rhegma thaumasium, and the other differences mentioned by Breder seem to be only of specific importance. 1 The course of the lateral line shown in the figure is doubtful, as most of the scales are missing in this region. WEST AFRICA ii CHILODIPTERIDAE Genus Synagrops, Giinther Melanostoma (non Schiner), Steindachner and Doderlein, 1884. Type M. japonicum, Steind. and Doderl. Synagrops, Giinther, 1887. Type Melanostoma japonicum, Steind. and Doderl. Parascombrops, Alcock, 1889. Type P. pellucidiis, Alcock. Hypoclydonia, Goode and Bean, 1895. Type H. bella, Goode and Bean. MaccuUochina, Jordan, 1922. Type Synagrops serratospinosa. Smith and RadclitTe. Key to the species^ I. Fin-spines all without serrae. [Synagrops.] A. About 30 scales in lateral line; depth 3I to more than 4 in length; eye 3 to 3 1 in head; anal II 7. 1. Maxillary to below centre of eye, 2I to 2^ in head bellus (Goode and Bean) 2. Maxillary to below anterior margin of pupil, 2J in head japonicus (Steind. and Doderl.) B. About 40 scales in lateral line; depth 3^ to 3^ in length; eye 3I to 3I in head; anal II 9 microlepis, sp.n. II. Spines of pelvics and sometimes anterior spines of dorsal and anal with serrae on outer edges. {Parascombrops.} A. Only pelvic spines with serrae; snout nearly as long as eye; depth 3! to 4 in length philippinensis (Giinth.) B. Anterior spines of dorsal and anal as well as spines of pelvics with serrae; snout much shorter than eye; depth 3 to 3I in length serratospinosus. Smith and Radcl. Through the courtesy of the United States National Museum, I have been able to examine an authentic specimen of S. japonicus from Suruga Bay (' Albatross ', No. 51434, 100 mm.). I have also seen 4 specimens of S. bellus from the Dry Tortugas, Florida, received through the kindness of Dr W. H. Longley, and have included a redescription of that species here. 5. japonicus and the two species of the subgenus Parascombrops have recently been redescribed by Fowler," and full synonymies will be found in his paper. Synagrops bellus (Goode and Bean). Hypoclydonia bella, Goode and Bean, 1895, Ocean. Ichth., p. 236, fig. 237; Jordan and Evermann, 1896, Bull. U.S. Nat. Mus., xlvh (i), p. 11 15, fig. 475. Depth of body 3! to 3 f in the length, length of head about 3. Snout a little more than I as long as eye, diameter of which is 3 to 3 J in length of head and a little greater than interorbital width. Maxillary extending to below middle of eye, length 2^ to 2^ in that of head ; lower jaw projecting. Upper jaw with broad bands of minute villiform teeth, which are separated at the symphysis, and with a pair of strong, slightly curved canine teeth anteriorly. Lower jaw with a marked concavity anteriorly on each side of the symphysis ; anterior part of jaw with bands of minute villiform teeth, and with a 1 Synagrops natalensis, Gilchrist, from Natal, has been very briefly described, but is probably identical with S. japonicus. 2 1930, Bull. U.S. Nat. Mus., c (10), p. 136. 12 DISCOVERY REPORTS pair of symphysial canines of moderate size ; sides of each mandible with 4 to 6 strong canines, of which the two hindermost are largest and about as large as those of the upper jaw; some minute teeth between and posterior to the lateral canines; a triangular patch of villiform teeth on the vomer and a band of similar teeth, tapering behind, on each palatine. Praeoperculum finely serrated along hinder and lower edge ; upper part of operculum with two feeble, flat spines. Gill-rakers of moderate length and rather stout; 13 to 15 on lower part of anterior arch. Scales thin, cycloid; 29 (?) in a longi- tudinal series ; lateral line high up on body, with a gradual curve which nearly follows the outline of the back. Dorsal IX, I 9 ; interspace between the two fins f to f diameter of eye. Anal II 7; first spine about ^ as long as second, which is 4 to 4I in length of head. Pectoral with 16 or 17 rays, length f to | that of head. Pelvic I 5 ; spine without serrae; origin a little in advance of pectoral base. Caudal forked; caudal peduncle about 2 J times as long as deep. A black blotch on upper part of spinous dorsal fin. Hab. Atlantic coast of tropical America, in deep water. Described from 4 specimens, 90-133 mm. in total length, from the Dry Tortugas, Florida. This species proves to be very close to S. japonicus, but appears to have a somewhat larger mouth. It is possible that the two species will eventually have to be united. Fig. 3. Synagrops microlepis. Holotype. x2. Synagrops microlepis, sp.n. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m.: 13 specimens, 33-60 mm. (holotype, 50 mm.) Closely related to the preceding species. Depth of body 3^ to 3I in the length, length of head 2| to 2f . Snout about f as long as eye, diameter of which is 2,1 to 3! in length of head and very little greater than interorbital width. Maxillary scarcely extending to below middle of eye, length about 2| in that of head ; lower jaw strongly projecting; concavity in the mandible very shallow. Upper jaw with bands of minute villiform teeth, separated at the symphysis by a rather broad interspace, on either side of which are one or two strong, curved canines, sometimes depressible; lower jaw anteriorly with bands of villiform teeth and with one or two pairs of small symphysial canines, laterally with 3 or 4 curved canine-like teeth on each side, continued posteriorly by a short narrow band of minute teeth ; a V-shaped patch of small teeth on the vomer and a narrow band of similar teeth on each palatine. Praeoperculum with denticulations WEST AFRICA 13 along the lower margin and on lower part of hinder edge ; a few denticulations at angle of praeopercular ridge ; upper part of operculum with two divergent ridges, terminating posteriorly in thin, flat spines. Gill-rakers slender, of moderate length; 13 or 14 on lower part of anterior arch. Scales cycloid; about 40 in lateral line. Dorsal IX, I 10; the two fins nearly contiguous ; fifth spine longest, length 2^ to 2^ in that of head. Anal II 9; first spine very short, second 2f to 3I in length of head. Pectoral with 16 or 17 rays ; extending to above origin of anal or beyond, length | to f that of head. Pelvic I 5 ; spine without serrae ; origin below or very little in advance of pectoral base. Caudal peduncle about twice as long as deep. Brownish, with a number of minute dark dots on back and upper parts of sides, which tend to form a dark band on either side of base of dorsal fin, the two bands uniting behind to form a dark patch on upper surface of caudal peduncle ; a narrow dark line on either side of base of anal fin and a narrow dark vertical bar at base of caudal ; lateral line dusky ; dorsal and caudal fins with small dark dots ; an indistinct blackish area on distal part of spinous dorsal ; other fins pale. LATILIDAE Latilus semifasciatus, Norman. 193 1, Ann. Mag. Nat. Hist. (10) vn, 356, fig. 3. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: i specimen, 215 mm. This species was known previously only from the holotype, 300 mm. in length, from the Gold Coast. In life there is said to be a "lustrous blue" shade along back at base of dorsal fin. Yellow is present in the neighbourhood of the mouth and orbit. POMADASIDAE Pomadasys jubelini (Cuv. and Val.). Pristipoma jubelini, Steindachner, 1870, SitzBer. Akad. Wiss. Wien, Lx (i), p. 675, pi. ii. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 4 specimens, 295-345 mm. St. 279. 10. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m. : 12 specimens, 35-80 mm. Hab. Tropical West Africa. SCIAENIDAE Otolithus macrognathus (Bleeker). Steindachner, 1870, SitzBer. Akad. Wiss. Wien, lx (i), p. 690, pi. vii. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 2 specimens, 445, 520 mm. Hab. Tropical West Africa. Otolithus senegalensis, Cuv. and Val. Steindachner, 1870, t.c, p. 687, pi. vi. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m.: i specimen, 395 mm. Hab. Tropical West Africa. 14 DISCOVERY REPORTS Sciaena aquila (Lacep.). [? Sciaena hololepidota (Lacep.).] St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 3 specimens, 275-295 mm. Sciaena angolensis, sp.n. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 4 specimens, 240-260 mm. (holotype, 240 mm.). Depth of body 3f to 3I in the length, length of head about 3. Snout shorter than eye, diameter of which is 3! to 3 1 in length of head and about i\ times interorbital width. Snout obtuse and rounded ; jaws equal anteriorly; maxillary extending to below middle of eye or a little beyond ; 2 pairs of open pores beneath the lower jaw ; teeth in villiform bands anteriorly, tapering laterally to a single series ; those of outer row in upper jaw Fig. 4. Sciaena angolensis. Holotype. x i. enlarged, becoming canine-like anteriorly; those of inner row of lower jaw moderately enlarged. Margin of praeoperculum with some minute denticulations along posterior edge; angle rounded, with some larger and more distinct teeth; operculum with two weak, flat spines. 9 or 10 gill-rakers of moderate length on lower part of anterior arch. Scales finely ciliated ; 47 to 49 in lateral line, 5 or 6 from origin of dorsal to lateral line ; lateral line tubules with 3 to 5 short branches. Dorsal IX, I 29-30 ; spines slender, fourth longest, about f length of head; soft dorsal scaled only at base. Anal II 7; first spine very short ; second stronger than those of dorsal fin, length more than \ that of head. Pectoral about | length of head. Pelvics shorter, f length of head. Caudal pointed, middle rays f to | as long as head. Silvery; back darker; dusky patches on opercular region ; a dark spot superiorly in axil of pectoral ; dorsal fins with blackish margins ; pectorals, anal and caudal more or less dusky. Apparently most nearly related to S. aquila (Lacepede), differing mainly in the some- what larger scales, larger eye, pointed caudal fin, etc. Umbrina ronchus,Valenc. Unibrina canariensis, Steindachner, 1867, SitzBer. Akad. Wiss. Wien, lvi (i), p. 638, pi. vi, fig. i Umbrina ronchus, Steindachner, 1882, Denkschr. Akad. Wiss. Wien, xlv, p. 8. WEST AFRICA 15 St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: 4 specimens, 280-370 mm. Hab. West Africa. SPARIDAE Dentex maroccanus (Cuv. and Val.). Steindachner, 1894, Notes Leyden Mas., xvi, p. 13. St. 271. 30. vii. 27. Elephant Bay, Angola. Hand line, 18 m.: 2 specimens, 250, 265 mm. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m. : 4 specimens, 180-205 mm. Hab. West Africa. Dentex macrophthalmus (Bloch). Giinther, 1859, Cat. Fish., i, p. 370. St. 271. 29. vii. 27. Elephant Bay, Angola. Seine net, 5-0 m.: 2 specimens, 60-88 mm. Hab. Mediterranean and adjacent seas ; West Africa. Dentex filosus, Valenc. Steindachner, 1868, SitzBer. Akad. Wiss. Wien, lvii (i), p. 975. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m. : i specimen, 295 mm. Hab. Coast of Algiers and West Africa ; South-east Africa. Dentex cuninghami, Regan. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: 3 specimens, 255-320 mm. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m.: i specimen, 120 mm. Hab. Angola. Pagellus mormyrus (Linn.). St. 271. 29. vii. 27. Elephant Bay, Angola. Seine net, 5-0 m. : 2 specimens, 50-60 mm. Box boops (Linn.). St. 271. 29. vii. 27. Elephant Bay, Angola. Seine net, 5-0 m.: 10 specimens, 1 15-135 mm. Diplodus rondeleti (Cuv. and Val.). St. 271. 29. vii. 27. Elephant Bay, Angola. Seine net, 5-0 m.: 6 specimens, 80-no mm. MAENIDAE Pterosmaris melanurus (Cuv. and Val.). Smarts melanurus, Steindachner, 1881, Denkschr. Akad. Wiss. Wien, XLiv, p. 26, pi. ii, fig. 2. St. 271. 29. vii. 27. Elephant Bay, Angola. Seine net, 5-0 m.: i specimen, 100 mm. Hab. Tropical West Africa. KYPHOSIDAE Kyphosus incisor (Cuv. and Val.). Jordan and Evermann, 1898, Bull. U.S. Nat. Mus., xlvh (2), p. 1386. St. 271. 29. vii. 27. Elephant Bay, Angola. Seine net, 5-0 m. : 3 specimens, 285-295 mm. Hab. West Indies; Brazil; West Africa (Angola). i6 DISCOVERY REPORTS The above specimens appear to be identical with one in the British Museum col- lection from Rio de Janeiro, so that, like K. sectatrix (Linn.), this species is found on both sides of the Atlantic. It differs from K. sectatrix in having a longer anal fin, with 13 soft-rays, longer pelvics, smaller scales above the lateral line, and a different colora- tion. K. gallvei, Cunningham, from St Helena, is doubtfully distinct from K. sectatrix, which has been recorded from Madeira and the Canary Islands on the eastern side of the Atlantic. LABRIDAE Julis atlantica, (Giinther). Coris atlantica, Giinther, 1864, Cat. Fish., iv, p. 197. St. 283. 13. viii. 27. Annobon Island, Gulf of Guinea. Hand line: i specimen, 250 mm. Hab. Cape Verde Islands ; West Africa. In life the body is said to be coloured a brilliant ultramarine blue, with some green patches ; red about the cheeks and at base of dorsal fin. TRICHIURIDAE Trichiurus lepturus, Linn. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 2 specimens. 225, 960 mm. GOBIIDAE Periophthalmus barbarus (Linn.). 9. viii. 27. French Congo. Shore collection: 9 specimens, 65-155 mm. Gobius (Gobius) angolensis, sp.n. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 37 specimens, 37-100 mm. (holotype, a male, 95 mm.). Body moderately compressed, depth 4 to 4J in the length, length of head 3 to 3^. Breadth of head usually rather less than its depth, which is | to f of its length. Snout short and obtuse, as long as or rather shorter than eye, diameter of which is 3 j to about 4 in length of head ; interorbital space very narrow. Jaws equal anteriorly ; maxillary extending to below anterior part of eye ; no distinct canine teeth ; tongue truncate or slightly bilobate. Head naked, except for 3 or 4 scales on upper part of operculum; nape scaled, with a shallow median groove ; body covered with ciliated scales, 27 to 29 in a longitudinal series, 7 or 8 (occasionally 9) in a transverse series between second dorsal and anal fins; breast scaled. Cutaneous papillae of head well developed, the principal series arranged as in the accompanying diagram. Dorsal VI, I 11-12; the two fins narrowly separated ; second or second and third rays of first dorsal more or less prolonged and filamentous in males, as long as or longer than head ; second dorsal and anal elevated. Anal 11-12. Caudal obtusely pointed, longest rays about as long as head. Pectoral a little shorter than head. Length of pelvics about § the distance from their base to the anal fin. Yellowish brown, with two broad dusky patches on back and sides, one below each dorsal fin ; usually a black spot or vertical bar at base of caudal ; first dorsal and anal dusky ; second dorsal and caudal with dark bars separated by narrower pale (yellow in life) interspaces ; pectorals and pelvics dusky. WEST AFRICA 17 This species is most nearly related to G. niger, Linn., of which G. jozo, Linn, is probably a synonym, and to G. roulei, De Buen. From both species it may be dis- tinguished by the larger head, fewer scales in a longitudinal series (27-29 instead of 33-39)' the long filamentous rays of the first dorsal fin in the male, the number and arrangement of the cutaneous papillae on the head, particularly in the nuchal and infraorbital series, and by the coloration. Further, in G. niger and G. roulei the nape is completely or almost completely without scales. 1 From G. maindroni, Sauvage, from Senegal, Sierra Leone and Niger, it may be distinguished by the larger head and eye and by the coloration. Fig. 5. Gobius {Gobiiis) angolensis. Holotype. Ixi. Fig. 6. Diagrammatic view of head of Gobius (Gobius) angolensis, showing the arrangement of the series of cutaneous papillae. Gobius, sp. St. 283. 13. viii. 27. Off Annobon, Gulf of Guinea. Large dredge, 18-30 m.: 2 specimens, 17, 18 mm. Acentrogobius koumansi, sp.n. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m. : 44 specimens, 30-100 mm. (holotype, 100 mm.). St. 279. 10. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m. : 6 specimens, 90-105 mm. 1 An example of this species was sent to Dr Fernando de Buen of Madrid, who has made a special study of the Gobies of this subgenus, and I am greatly indebted to him for his opinion as to its relationships. D .XH DISCOVERY REPORTS Body compressed, depth 4^ to 4! in the length, length of head 3 (young) to 3^. Breadth of head much less than its depth, which is about f of its length. Snout short, rounded, shorter than eye, diameter of which is 2f (young) to 3I in length of head ; interorbital space very narrow, the eyes being nearly contiguous in the young. Mouth oblique, the jaws equal anteriorly or lower a little prominent; maxillary extending to below middle of eye ; both jaws with several rows of teeth, those of the outer row en- larged, those of the innermost row a little enlarged; teeth of outer row of upper jaw stronger than those of lower, forming more or less distinct curved canines ; no canines Fig. 7. Acentrogobiiis koiimansi. Holotype. xi. Fig. 8. Diagrammatic view of head of Acentrogobhis kotimansi, showing the arrangement of the series of cutaneous papillae. in lower jaw ; tongue truncate or slightly bilobate. Head naked except on upper surface behind eyes and on lower part of cheek, where there is a row of 3 or 4 scales ; body covered with finely ciliated scales ; 25 to 27 in a longitudinal series, 7 or 8 in a transverse series between second dorsal and anal fin; breast scaled. Cutaneous papillae on head well developed, arranged as in the accompanying diagram. Dorsal VI, I 16-17; the two fins narrowly separated; second, third and fourth rays of first dorsal sometimes a little prolonged, longest nearly as long as head (males ?); second dorsal and anal elevated. Anal 18-20. Caudal pointed, the middle rays longer than the head. Pectoral without free, silk-like rays, nearly as long as head. Pelvics extending nearly to vent. Yellowish brown, with about 5 dark spots along each side and usually with traces of a row of similar spots along back just below dorsal fins ; fins more or less dusky distally ; indistinct greyish bars on spinous dorsal. WEST AFRICA 19 According to Koumans' synopsis of the genera of Gobiinae (1931) this species must be placed either in Oxyiirichthys or in Acentrogobiiis.^ It resembles certain species of Oxyiirichthys, especially in the large mouth, number of dorsal and anal rays, pointed caudal, etc., but differs in the form of the dentition. It fits much better into Acentro- gobitis, as defined by Koumans, and the diagnosis of that genus may be amended for its reception. It appears to be related to A. schlegeli (Giinther), from which it may be readily distinguished by the different dentition, the larger eye, and the greater number of dorsal and anal rays. BLENNIIDAE Blennius velifer, sp.n. St. 271. 30. vii. 27. Elephant Bay, Angola. Shore collection. 17 specimens (12 males, 5 females), 25-50 mm. (holotype, a male, 50 mm.). Depth of body 4f to 5 in the length, length of head about 4. Snout obtuse, the anterior profile steep in males, less abrupt in females. Diameter of eye 3I to more than Fig. 9. Blennius velifer . X3. 4 in length of head and greater than interorbital width. Maxillary extending to below anterior part or middle of eye ; two large canines in lower jaw and usually two smaller ones in the upper jaw. A small simple tentacle at each anterior nostril; no orbital tentacles; no occipital tentacles or filaments, but males with a well-marked occipital 1 I am greatly indebted to Dr F. P. Koumans for his kindness in examining a specimen of this species, and for his opinion as to its probable systematic position. He has also examined an example of Gobius angolensis. 3-2 20 DISCOVERY REPORTS crest. Dorsal XII 14-15; distinctly notched; commencing above upper angle of gill- opening and ending just before the caudal; spinous portion higher than soft portion in males, rather lower than soft part in females. Anal 16-20. Caudal rounded. Pectoral as long as or longer than head, not or scarcely extending to above origin of anal. Males with traces of dark cross-bars on sides, and with some large dark spots on caudal region of body ; a round black spot immediately behind the eye ; spinous part of dorsal chocolate brown ; anal with a narrow dark submarginal band, each ray tipped with white ; females with numerous dark spots on hinder part of body, and with irregular cross-bars anteriorly ; a dark spot behind the eye as in males ; anal fin spotted with brown. This species appears to be most nearly related to B. trigloides, Cuv. and Val., and B. bufo, Lowe. Blennius, sp. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Net (4 mm. mesh) attached to back of trawl, 64-65 m.: I specimen, 23 mm. This is perhaps the young of B. ocellaris, Linn., but appears to have a deeper body and rather fewer rays in the dorsal fin (XI 13). BROTULIDAE Brotula barbata (Schn.). Regan, 1915, Ann. Mag. Nat. Hist. (8) xv, p. 128. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m.: 2 specimens, 510, 560 mm. SCORPAENIDAE Scorpaena canariensis (Sauvage). St. 274. 4. viii. 27. Oft' St Paul de Loanda, Angola. Large otter trawl, 64-65 m.: 2 specimens, 60, 130 mm. See description below. Scorpaena angolensis, sp.n. St. 271. 29. vii. 27. Elephant Bay, Angola. Large fish-trap, 20 m.: i specimen, 80 mm. See description below. Revision of the Score aenidae of the Mediterranean and neighbouring parts of the atlantic The difficulty experienced in identifying the Scorpaenids collected by the 'Dis- covery' has led me to undertake an examination of all the species occurring in the Mediterranean and in the eastern Atlantic. Previous authors have found it difficult to ascertain the hmits of the genera found in the Atlantic, and my own work has shown quite clearly that, in order to arrive at any definite conclusions concerning the sub- division of the family, a thorough revision of the species of the world will be necessary. It follows, therefore, that the arrangement adopted here is a tentative one, and although I have succeeded in obtaining a satisfactory idea as to the species found in the region under consideration, the grouping of these into genera is open to criticism. WEST AFRICA 21 Key to the Genera I. 30 to 31 vertebrae; dorsal with 15 spines, anal with 7 or 8 soft-rays; spines on upper surface of head feeble; no distinct suborbital ridge; interorbital space flat; head densely scaled; pectoral with narrow base, some of its rays branched ... ... Sebastes. II. 24 to 26 vertebrae; dorsal with 11 to 13 spines, anal with 5 soft-rays; a more or less distinct suborbital ridge, generally armed with spines. A. Dorsal continuous, though somewhat notched, the penultimate spine at least I as long as the last; second anal spine generally longer than third; interorbital space always more or less concave, its width less than diameter of eye. 1. Dorsal with 13 spines; palatines without teeth ... ... ... Scorpaenodes. 2. Dorsal with 12 (occasionally 11 or 13) spines; palatines with teeth. a. Pectoral with all the rays simple, and with narrow base; 13 or 14 gill-rakers on lower part of anterior arch; head more or less scaled ... ... Pontinus. b. Pectoral with some of the rays branched (except in very young), and with broad base. a. Gill-rakers of moderate length, 16 to 22 on lower part of anterior arch; lower rays of pectoral free from membrane for at least -3- of their length, and upper part of hinder edge of fin truncate ; suborbital ridge smooth or with one small spine. Helicolenus. fi. Gill-rakers rather short and stout, 9 to 12 on lower part of anterior arch; lower rays of pectoral free from membrane only at their tips, and hinder edge of fin evenly rounded; suborbital ridge generally armed with spines Scorpaena. B. Dorsal deeply notched, the penultimate spine only about 3 as long as the last; second anal spine much shorter than third; interorbital space flat, its width greater than diameter of eye ; pectoral with 20 or more rays, some of them branched ; dorsal with 12 spines; lateral line a naked groove with prominent tubes ... ... Setarches. Genus Sebastes, Cuvier 1829, R. Anim., ed. 2, n, p. 166. Type Perca norvegica, Ascan. Eusebastes, Sauvage, 1878, Nouv. Arch. Mus. H.N. Paris (2) i, p. 115. Type Sebastes septen- trionalis, Gaimard. As now restricted, this genus includes only two species in the North Atlantic: S. marimis (Linn.) and S. viviparus, Kr^yer. These have been regarded by many authors as identical, but, according to Swenander,i Jensen,"^ Duncker,^ Saemundsson'' and others, the species are valid. Genus Scorpaenodes, Bleeker 1857, Nat. Tijdschr. Ned.-Ind., xni, p. 371 ; 1876, Versl. Akad. Wet. Amsterdam (2) ix, p. 296. Type Scorpaena polylepis, Bleeker. Scorpaenodes africanus, PfafF. 1933, Vid. Medd. Dansk not. For., xciv, p. 311, fig. 13. Hab. Senegal. Known only from the holotype, 65 mm. in total length, from Dakar. 1 1906, Kgl. Norske Vidensk. Selsk. Skr. (1905), No. 9, p. 7. 2 1922, Vid. Medd. nat. For. Kjobenhavn, lxxiv, p. 105. 3 1927, in Grimpe and Wagler, Tierwelt Nord- u. Ostsee, Lief, x. Tail xii. Heft 2, p. 2. * 1932, in Joubin, Faun. Ichth. Atlant. Nord, x, figs. 22 DISCOVERY REPORTS Genus Pontinus, Poey 1858, Mem. Hist. Nat. Cuba, 11, p. 172. Type Pontinus castor, Poey. Sebastoplus, Gill, 1863, Proc. Acad. Nat. Sci. Philad., p. 208. Type Scorpaena kuhli, Bowdich. Poey described two species of the genus Pontinus, P. castor and P. pollux, both probably shallow-water forms from the West Indies. I have not examined specimens of either species, and follow Goode and Bean in regarding them as congeneric with Scorpaena kuhli, Bowdich, which is the type of Gill's genus Sebastoplus. The genus does not occur in the Mediterranean, and there are only two species in the eastern Atlantic, which may be distinguished as follows : I. 5 or 6 series of scales between last soft-ray of dorsal and lateral line, 9 to 11 on cheek below suborbital ridge ... ... ... ... ... ... ... ... ... kuhli. II. 3 series of scales between last soft-ray of dorsal and lateral line, 6 or 7 on cheek below suborbital ridge accraensis. Sebastes nigropunctatus, Giinther, from St Helena, is a Pontinus, and may be dis- tinguished from both the above by the form of the dorsal spines, none of which are elongate, and by the coloration. Several species have been described from the western Atlantic, but these are not represented in the British Museum collection.^ Pontinus kuhli (Bowdich). Scorpaena kuhlii, Bowdich, 1825, Excur. Madeira, p. 123. Sebastes kuhlii, Lowe, 1839, Trans. Zool. Soc, 11 (3), p. 176; Lowe, i860, Hist. Fish. Madeira, p. 1 15, pi. xvii; Gunther,i86o, Cat. Fish., n, p. 102; Steindachner, 1867, SitzBer. Akad. Wiss. Wien, LVi (i), p. 671; Capello, 1881, Mem. Acad. Sci. Lisboa, XLVi (N.S. vi, pt. i), p. 11; Vaillant, i888, Exped. Sci. ' Travailleur' et ' Talisman', Poiss., p. 370; Vinciguerra, 1893, Atti Soc. Ital. Sci. Nat. Milano, xxxiv, p. 312; Collett, 1896, Res. Camp. Sci. Monaco, x, p. 13. Sebastes filifer, Valenciennes, 1843, in Webb and Berthelot, Canaries (Ichth.), p. 21, pi. ii, fig. 2.^ Sebastoplus kuhlii, Gill, 1863, Proc. Acad. Nat. Sci. Philad., p. 208. Sebastes {Sebastichthys) filifer, Sauvage, 1878, Nouv. Arch. Mus. H.N. Paris (2) i, p. 118. Pontinus kuhlii, Goode and Bean, 1895, Ocean. Ichth., p. 253. Pontinus filifer, Goode and Bean, 1895, t.c, p. 254. Depth of body about 3 in the length, length of head 2 J to 2|. Snout longer than eye, diameter of which is ^\ to 5 in length of head and if to more than twice interorbital width. Two strong, backwardly curved spines on the praeorbital; suborbital ridge with 3 or 4 spines ; 5 praeopercular spines, the uppermost strongest and with a smaller spine at its base, the lowermost sometimes wanting. Gill-rakers of moderate length, the longest less than \ diameter of eye; 13 or 14 (including rudiments) on lower part of anterior arch. Scales spinulose and ciliated; 5 or 6 series between last soft-ray of dorsal and lateral line; breast scaled. Dorsal XII 9-10;^ generally the second and 1 Sebastes nematophthalmus , Giinther (i860. Cat. Fish., 11, p. 99) was described from two specimens, a large stuffed one believed to have come from Mauritius, and one in spirit, 165 mm. long, which formed part of a collection made by Sir R. Schomburgk in the West Indies. These appear to represent distinct species. I cannot find any trace of the "long, slender, tapering filament above the posterior angle of the orbit" in the West Indian specimen, which is an undoubted Pontinus. 2 I am indebted to Dr W. H. Longley for notes upon the type of this species In the Paris Museum. 3 The last ray of both dorsal and anal fins in all Scorpaenids is cleft nearly to its base, and has con- sequently been counted as two rays by many workers. WEST AFRICA 23 third spines much longer than the remainder, but sometimes only the second or only the third elongate; longest spines if to 2| in length of head. Anal III 5. Pectoral with 16 or 17 rays, the lower ones a little thickened and free from membrane at their tips; fin extending to a little beyond vent. Pale brownish, the upper parts of the sides spotted with darker brown, the spots sometimes tending to form irregular longitudinal rows ; dorsal and sometimes the caudal fin spotted with brown ; other fins nearly uniform. Hab. Coast of Portugal ; Madeira ; Canaries ; Azores ; etc. In the British Museum 10 specimens, 210-320 mm. in total length. Sebastes {Sebastichthys) bibroni, Sauvage,^ described from a single specimen 208 mm. in total length from Sicily, seems to belong to this genus. Dr W. H. Longley has kindly examined the type in the Paris Museum for me, and reports that the pectoral rays (18) are all simple and that there are 13 gill-rakers (including rudiments) on the lower part of the anterior arch, the longest being more than \ the diameter of the eye. The length of the first dorsal spine is 13 mm., the second 36 mm., the third 34 mm., and the fourth 25 mm. It is possible that this species will prove to be identical with Ponthms kuhli, to which it is obviously very closely related, but I have hesitated to unite the two forms, especially in view of the fact that P. kuhli has not been recorded from the Mediterranean. Pontinus accraensis, sp.n. Closely related to P. kuhli. Depth of body about 3 in the length, length of head a little more than 2. Snout very little longer than eye, diameter of which is about 4 in length of head and 2| times the interorbital width. The spines above the anterior angles of the orbits are directed upwards and outwards instead of posteriorly; the second praeopercular spine is considerably stronger, this being generally very in- conspicuous in P. kuhli, and the fourth has three points. 12 gill-rakers on lower part of anterior arch. 3 series of scales between last soft-ray of dorsal and lateral line, 6 or 7 on cheek below suborbital ridge. Dorsal XII 9; only the second spine elongated, length 2f in that of head. Anal III 5. Pectoral with 17 rays. The dark spots on the upper part of the body are more distinct, and there is a row of spots along the lateral line ; the caudal as well as the soft dorsal fin is ornamented with small dark spots. Hab. Accra, Gold Coast. Only the holotype known, a specimen 215 mm. in total length collected and pre- sented to the British Museum by Dr F. R. Irvine in 1930. Genus Helicolenus, Goode and Bean 1895, Ocean. Ichth., p. 248. Type Scorpaena dactyloptera, Delaroche. There appear to be four species of this genus in the Mediterranean and Atlantic, one of which {H. maderensis) occurs only off the American coast, and another at the Cape. The other two may be distinguished as follows : I. Diameter of eye af to 4 in head (in specimens of 145 to 230 mm.); only one or two pairs of spines on occipital region; body uniformly coloured or with irregular dark markings on upper parts of sides dactylopterm. 1 1878, Nouv. Arch. Mus. H.N. Paris (2) i, p. 116, pi. i, fig. 3. 24 DISCOVERY REPORTS II. Diameter of eye 3f to 4 in head (in specimens of 63 and 103 mm.); 3 pairs of spines on occipital region; body mottled with dark brown ... ... ... ... inicrophthalmus. Helicolenus dactylopterus (Delaroche). ? Scorpaena malabarica, Schneider, 1801, in Bloch, Syst. Ichth., p. 194. Scorpaena dactyloptera, Delaroche, 1809, Ann. Mus. H.N. (Paris), xiii (77), p. 337, pi. xxii, fig. 9; Risso, iSio, Ichth. Nice, p. 186; Risso, 1826, Hist. Nat. Europ. Merid., in, p. 369; Smitt, 1893, Scand. Fish., i, p. 154, fig. 43; Holt and Calderwood, 1895, Sci. Trans. R. Dublin Soc. (2) v, p. 409, pi. xlii, fig. i ; Holt and Byrne, 1908, Fisheries Ireland Sci. Invest., 1906, V, p. 9, pi. i; Duncker, 1927, in Grimpe and Wagler, Tierwelt Nord- u. Ostsee, Lief. X, Teil XX, Heft 2, p. 4, fig. 2. Sebastes imperialis, Cuvier, 1829, R. Anim., ed. 2, 11, p. 167; Cuvier and Valenciennes, 1829, Hist. Nat. Poiss., iv, p. 336; Lowe, 1839, Trans. Zool. Soc, 11 (3), p. 175; Lowe, i860, Hist. Fish. Madeira, p. 171, pi. xxiv. Sebastes kuhlii, Valenciennes, 1843, in Webb and Berthelot, Canaries {Ichth.), pi. ii, fig. i. Sebastes dactylopterus, Giinther, i860. Cat. Fish., 11, p. 99; Moreau, 1881, H.N. Poiss. France, II, p. 317, fig. 117; Steindachner and Doderlein, 1885, Denkschr. Akad. Wiss. Wien, XLix, p. 201 {S. hilgendorfi, Doderlein MS.]; Carus, 1889-93, Pi'odr. Faun. Medit., 11, p. 638; Collett, 1896, Res. Camp. Sci. Monaco, x, p. 12; Roule, 1907, Arch. Zool. exper. gen., (4) VI, Notes et Revue, p. xv; Jaquet, 1907, Bull. Inst, ocean. Monaco, cix. Helicolenus dactylopterus (part), Goode and Bean, 1895, Ocean. Ichth., p. 249. Scorpaena {Helicolenus) dactyloptera, Fage, 1918, Rep. Danish Ocean. Exped. 1908-10, 11, A 3, p. 102. Depth of body 2y to 3^ in the length, length of head 2| to 2|. Snout shorter than eye, diameter of which is af to 4 in length of head and 2 to 2I times interorbital width. Praeorbital spines feeble; suborbital ridge generally smooth, but sometimes a very small spine below posterior edge of eye; 5 praeopercular spines; a spine above the front of each orbit and 3 or 4 above its posterior angle ; one or two pairs of spines on the occipital region. Maxillary with a patch of scales in the centre. Gill-rakers of moderate length, the longest j to | diameter of eye; 16 to 18 on lower part of anterior arch. Scales spinulose and ciliated; about 5 series between last soft-ray of dorsal and lateral line; breast scaled. Dorsal XII (occasionally XIII) 12 (-14); third or fourth spines generallv longest, 2^ to nearly 3 (large specimens) in length of head. Anal III 5. Pectoral with 19 rays, the 2 uppermost simple, the next 8 or 9 branched, and the lowermost 8 or 9 simple and sometimes somewhat thickened; fin extending to above vent or a little beyond. Pale yellowish brown (red or pinkish in life), paler below; sometimes with some dark markings forming irregular bars on upper parts of sides, generally more prominent in the young; pharynx blackish or dark brown. ^ Hab. Mediterranean and adjacent parts of the Atlantic, northwards to Scandinavia, southwards to the Cape Verde Islands; Azores; Japan (?). In the British Museum numerous specimens, 115-230 mm. in total length. There are 4 specimens from Japan in the British Museum collection, 145-230 mm. in total length, but all are in somewhat poor condition. After careful comparison I am unable to detect any important differences between these specimens and some of 1 For a detailed account of the variation in the coloration of this species see Holt and Byrne (1908). WEST AFRICA 25 equal size from the Atlantic. Helicolemis dactylopterus is said to be common on the western side of the Atlantic, mainly between latitudes 30° and 40° N, but, judging from a single specimen of 122 mm. length, taken by the 'Albatross' in the North-West Atlantic (40° 00' 15" N, 70° 55' 30" W) at a depth of 136 fathoms, this should be a distinct species, distinguishable by the smaller eye, which is about 3^ in length of head and about twice the interorbital width, less deeply concave interorbital region, and by the constant presence of a small spine on the suborbital ridge. The scales may be a little larger in the American form, but the specimen examined is in a poor state of preservation. The Western Atlantic species will stand as H. maderensis, Goode and Bean, a somewhat unfortunate name, as it has been clearly shown that this form does not occur at all on the eastern side of the Atlantic. H. maculatus (Cuv. and Val.), from the Cape, is closely related to H. dactylopterus, but may be readily distinguished by the rather longer and more numerous gill-rakers (21 or 22), larger scales, and by the coloration. H. dactylopterus is usually found in fairly deep water, chiefly between 100 and 400 fathoms, but young and half-grown individuals have been recorded from lesser depths. The maximum size of the species appears to be about 24 inches. Helicolenus microphthalmus, Norman. 1935, Proc. Zool. Soc, p. 612, fig. I. Hab. OflP Saltburn, Yorks ; 30 fathoms. Known only from the types, 63 and 103 mm. in total length. Genus Scorpaena, Linnaeus 1758, Syst. Nat., ed. 10, p. 266. Type Scorpaena porcus, Linnaeus. There are numerous species of this genus in all tropical and temperate seas, presenting considerable variation in squamation and in the armature of the head. It seems certain that a careful revision of all the species would reveal definite characters for the subdivision of the genus, and even the few species dealt with here represent a somewhat heterogeneous assemblage. It is probable that the species with the breast naked (Scor- paena) will have to be separated from the remainder, and it is very doubtful whether the species described by Koehler as echinata (? =-cristtdata, Goode and Bean) should be included. 1 The presence or absence of an occipital pit, originally believed to be of primary importance, proves to be of little value as a generic character, and within the genus every gradation exists between a deep quadrate pit on the occiput and a very shallow depression or none at all. There appear to be nine species of Scorpaena in the Mediterranean and adjacent parts of the Atlantic, of which one was originally described by Sauvage in 1878, but has not since been recognized, and another is described below as new to science. 1 See p. 32 of this report. 26 DISCOVERY REPORTS Key to the Mediterranean and Eastern Atlantic Species I. Breast naked; scales on head not visible, completely embedded in the skin. A. Scales smooth, but with crenulate margins, small; 6 or 7 between last soft-ray of dorsal and lateral line; spines on suborbital ridge small or wanting parens. B. Scales ciliated, sometimes spinulose, larger; 3 to 5 between last soft-ray of dorsal and lateral line; spines on suborbital ridge well developed. 1. Occipital pit present; base of pectoral broad, the lowermost ray inserted level with root of pelvic spine and more or less in advance of uppermost. a. Pectoral with 18 (occasionally 17 or 19) rays; distance from origin of dorsal to anterior edge of occipital pit equal to or a little greater than eye ; snout shorter than eye, which is 3 to 4^ in head. a. Head 2^ to 2i in length; pores on head numerous; supraorbital tentacle generally of moderate size or small ... ... ... ... ... ustulata. jS. Head 2 J to 2 J in length; pores on head fewer; supraorbital tentacle larger, -3 to f eye angolensis. b. Pectoral with 19 or 20 rays; distance from origin of dorsal to anterior edge of occipital pit il to more than twice eye; snout longer than eye, which is 4I to 6| in head scrofa. 2. No occipital pit; base of pectoral rather narrow, the lowermost ray inserted a little above level of root of pelvic spine and about opposite uppermost ... caiiariemis. H. Breast fully scaled; always some scales visible on opercular region, and often on cheek also. A. Scales all cycloid ; a well-developed occipital pit. 1. Depth of head at origin of dorsal i\ to li in its length; supraorbital tentacle, when present, small; pectoral extending to vent or a little beyond, li to if in head laevis. 2. Depth of head at origin of dorsal i J in its length; supraorbital tentacle large, § to f eye; pectoral extending nearly to origin of anal, i\ in head senegalensis. B. Scales spinulose and ciliated; no occipital pit. 1. Pectoral with 15 or 16 rays; 2 or 3 rather feeble spines on posterior part of suborbital ridge; head 2i to 2f in the length madiirensis. 2. Pectoral with 21 or 22 rays; 7 or 8 strong spines on suborbital ridge, which is prominent; head 2 J to 2 J in the length echinata. Scorpaena porcus, Linnaeus. 1758, Syst. Nat., ed. 10, p. 266; Bloch, 1785, Nat. ausl. Fische, ni, p. 5, pi. clxxxi; Risso, 1810, Ichth. Nice, p. 187; Giinther, i860. Cat. Fish., 11, p. 107; Carus, 1889-93, Prodr. Faun. Medit., II, p. 640; Roule, 1907, Areh. Zool. exper. gen. (4) vi, Notes et Revue, p. xvii; Jaquet, 1907, Bidl. Inst, ocean. Monaco, cix; Page, 191 8, Rep. Danish Ocean. Exped. 1908-10, II, A 3, p. 103. Cottus massiliemis, Forskal, 1775, Descr. Anim., p. 24. Scorpaena massiliensis, Lacepede, 1802, Hist. Nat. Poiss., iii, pp. 258, 269. Scorpaena fasciata, Costa, 1850 (?), Faun. Napoli, H. Pesci, Scorpaena, p. 3, pi. iv. Depth of body 2 J to nearly 3 in the length, length of head 2^ to 25. Snout as long as or rather shorter than eye, diameter of which is 4 to 5 in length of head and greater than interorbital width. No visible scales on head, but a number of pores. A deep occipital pit. Spines on praeorbital strong, each with a single point, but those on suborbital ridge feebly developed or wanting. A well-developed supraorbital tentacle, WEST AFRICA 27 and a few membranous processes scattered over head. Length of band of palatine teeth about § diameter of eye. 11 or 12 gill-rakers on lower part of anterior arch. Scales smooth, with the hinder margins crenulate; 6 or 7 series between last soft-ray of dorsal fin and lateral line ; breast naked. Dorsal XII 9 ; third, fourth and fifth spines longest, 2^ to 3^ in length of head. Anal III 5; third spine as long as or longer than second in adults, second spine longest in immature specimens. Pectoral with 16 or 17 rays, extending to above vent or a little beyond. Reddish brown, variously mottled with darker and dotted with deep black, the dots sometimes margining the darker areas ; in smaller specimens the fins decorated with irregular dark spots, blotches and cross-bands ; often a black blotch on hinder part of spinous dorsal ; pectoral spotted and marbled with dark brown or black, some larger dark spots in the axil. Hub. Mediterranean and adjacent parts of the Atlantic, straying northwards to the British Isles. In the British Museum numerous specimens, 38-260 mm. in total length. Scorpaena ustulata, Lowe. ? Scorpaena notata, Rafinesque, 1810, Car. N. Gen., p 33. Scorpaena ustulata, Lowe, 1841, Proc. Zoo!. Soc, vin (88 and 89), p. 36; Giinther, i860, Cat. Fish., u, p. no; Bellotti, 1888, Atti Soc. Ital. Sci. Nat. Milano, xxxi, p. 213, pi. iv, fig. i ; Carus, 1889-93, Prodr. Faun. Medit., u, p. 641; Moreau, 1891, H.N. Poiss. France, Suppl. p. 26; Collett, 1896, Res. Camp. Sci. Monaco, x, p. 10, pi. iv, fig. 15; Roule, 1907, Arch. Zool. exper. gen. (4) vi. Notes et Revue, p. xxi; Jaquet, 1907, Bull. Inst, ocean. Monaco, cix; Fage, 1918, Rep. Danish Ocean. Exped. 1908-10, n, A 3, p. 103. Scorpaena porcus, Costa, 1850 (?), Faun. Napoli, II. Pesci, Scorpaena, p. 2, pi. iii. Scorpaena tenerijfea, Jordan and Gunn, 1898, Proc. Acad. Nat. Sci. Philad., p. 345. Depth of body 2f to 3 in the length, length of head 2\ to 2\. Distance from origin of dorsal to anterior edge of occipital pit equal to or a little greater than diameter of eye. Snout blunt, shorter than eye, diameter of which is 3 to 4 in length of head and about twice the interorbital width. No visible scales on head, but numerous pores, which in preserved specimens give parts of the head a pustulate appearance. A deep occipital pit. Spines on praeorbital strong, the anterior with two points ; 4 well-developed spines on suborbital ridge. Supraorbital tentacle generally developed, sometimes small or even absent ; its length generally j to i diameter of eye ; sometimes a few membranous processes on head. Length of band of palatine teeth about \ diameter of eye. 10 to 12 gill-rakers on lower part of anterior arch. Scales spinulose and ciHated ; 3 or 4 series between last soft-ray of dorsal and lateral line; breast naked; a few membranous processes sometimes present on body. Dorsal XII 9 (XIII 8 in one specimen) ; third to fifth spines longest, 2 to 2\ in length of head. Anal III 5; second spine longer and stronger than third. Pectoral with 18 (occasionally 17 or 19) rays, extending to above origin of anal or not quite as far; base broad, the lowermost ray inserted level with root of pelvic spine and more or less in advance of uppermost. Reddish brown; nearly uniform or variously mottled with darker and paler on body and fins ; often dotted with black ; nearly always a black blotch on hinder part of spinous dorsal fin. 4-2 28 DISCOVERY REPORTS Hob. Mediterranean and adjacent parts of the Atlantic; Azores. In the British Museum numerous specimens, 65-205 mm. in total length. Scorpaena angolensis, sp.n. Closely related to S. ustulata, but length of head 2J to 2^ in that of fish. Diameter of eye 3^ to 4^ in length of head and if to nearly twice interorbital width. Pores on head rather less numerous. Anterior praeorbital spine single or with a very small secondary spine at its base. Supraorbital tentacle larger, branched, its length ^ to f diameter of eye ; membranous processes well developed on head and others on parts of body, especially in the region of the lateral line. Length of palatine band of teeth f or I diameter of eye. Dorsal XII 9 ; fourth and fifth spines longest, about twice in length of head. Anal III 5. Pectoral with 18 rays. Fig. 10. Scorpaena angolensis. St. 271. x i|. Hob. Coast of Angola. In addition to the specimen collected by the Discovery Expedition, there are two more in the British Museum collection, 120 and 155 mm. in total length, found among the unidentified Scorpaenids. The larger of these has been selected as the holotype. S. angolensis is very closely related to S. ustulata, and has probably been mistaken for that species. In addition to the diff'erences mentioned in the description, if examples of equal size are compared, S. angolensis will be seen to have a more slender body and rather larger mouth. Scorpaena scrofa, Linnaeus. Linnaeus, 1758, Syst. Nat., ed. 10, p. 266; Bloch, 1785, Nat. aiisl. Fische, ni, p. 10, pi. clxxxii; Risso, 1810, ichth. Nice, p. 188; Costa, 1850 (?), Faun. Napoli, II. Pesci, Scorpaena, p. i, pi. ii; Lowe, i860, Hist. Fish. Madeira, p. 105, pi. xvi; Giinther, i860, Cat. Fish., n, p. 108; Moreau, 1881, H.N. Poiss. France, u, p. 310, fig. 116; Day, 1887, Proc. Zool. Soc, p. 342; Carus, 1889-93, Prodr. Faun. Medit., 11, p. 639; Roule, 1907, Arch. Zool. exper. gen. (4) vi. Notes et Revue, p. xix; Jaquet, 1907, Bull. Inst, ocean. Monaco, cix; Holt and Byrne, 1908, WEST AFRICA 29 Fisheries Ireland Sci. Invest., 1906, v, p. 26, fig.; Fage, 1918, Rep. Danish Ocean. Exped. 1908-10, II, A 3, p. 103; Metzelaar, 1919, Trop. Atlant. Vissch., p. 285. Scorpaena barbata, Lacepede, 1802, Hist. Nat. Poiss., iii, p. 259. Scorpaena lutea, Risso, 1810, Ichth. Nice, p. 190; Risso, 1826, H.N. Europ. Merid., iii, p. 371; Roule, 1907, t.c, p. XX. Scorpaena scrofa var. obesa, Lowe, i860, t.c, p. 105 ; Goode and Bean, 1895, Ocean. Ichth., p. 245. Scorpaena scrofa var. histrio, Lowe, i860, t.c, p. 106. Depth of body 2§ to 3I in the length, length of head 2^- to 2f . Distance from origin of dorsal to anterior edge of occipital pit i\ times to more than twice diameter of eye. Snout pointed, longer than eye, diameter of which is 4I (young) to 6| in length of head and equal to or as much as i^ times interorbital width. No visible scales on head, and comparatively few pores. A somewhat shallow occipital pit. Praeorbital spines strong; two smaller intermediary spines between the principal ones ; 4 well-developed spines on suborbital ridge. Supraorbital tentacle very variable in size, its length from \ to | diameter of eye; sometimes altogether wanting; membranous processes variously developed on chin, praeorbital, edge of praeoperculum, and on other parts of head. Length of band of palatine teeth | to f diameter of eye. 10 to 12 gill-rakers on lower part of anterior arch. Scales ciliated; 4 or 5 series between last soft-ray of dorsal and lateral line; breast naked; usually a number of membranous filaments and pro- cesses on body, especially in the region of the lateral line and on the back. Dorsal XII 9 (occasionally XIII 8); third or fourth (sometimes fifth) spines longest, if to 2| (young) in length of head. Anal III 5 ; second spine subequal to or a little longer than, and stronger than third. Pectoral with 19 or 20 rays, extending to above vent or not as far; base broad, the lowermost ray inserted level with root of pelvic spine and in advance of uppermost. Coloration very variable ; generally reddish brown or yellowish, the head, body and fins marbled and spotted with darker brown; in smaller specimens the markings on the fins tend to form irregular cross-bars; often a large black blotch on hinder part of spinous dorsal fin. Hab. Mediterranean and adjacent parts of the Atlantic, ranging southwards to Madeira and beyond, and straying northwards to the British Isles. In the British Museum numerous specimens, 140-480 mm. in total length. This is a very variable species, particularly in the coloration, the size of the supra- orbital tentacle, and to some extent in the size of the scales. It is possible that the examination of a large series of examples would lead to the recognition of more than one form. As a general rule, specimens from the eastern Mediterranean appear to have the supraorbital tentacle better developed than those from its western end, but in examples from Madeira the size of the tentacle varies considerably. Scorpaena canariensis, Sauvage. Sebastes [Sebastichthys) canariensis, Sauvage, 1878, Notiv. Arch. Mus. H.N. Paris (2) i, p. 117, pi. i, figs. I, 2. Pontinus canariensis, Goode and Bean, 1895, Ocean. Ichth., p. 255. Depth of body 3 in the length, length of head 2|. Snout about as long as eye, diameter of which is nearly 4 in length of head and i^ times interorbital width. No 30 DISCOVERY REPORTS visible scales on head, but numerous small pores. No occipital pit. Two strong prae- orbital spines, and 3 or 4 on suborbital ridge. Supraorbital tentacle about h diameter of eye; other membranous processes at anterior nostril, on praeorbital, and on edge of praeoperculum ; a few small processes on upper surface of eyeball. Palatine band of teeth very narrow, its length about f diameter of eye. 10 gill-rakers on lower part of anterior arch. Scales ciliated ; 3 series between last soft-ray of dorsal and lateral line ; breast naked ; no membranous processes on body. Dorsal XII 9 ; third and fourth spines longest, about twice in length of head. Anal III 5; third spine a little longer than second. Pectoral with 18 rays, extending to beyond origin of anal; base rather narrow, the lowermost ray inserted a little above level of root of pelvic spine and about opposite uppermost. Pale yellowish brown, with indistinct, narrow, oblique, greyish stripes Fig. II. Scorpaena canariensis . x i. following the series of scales above the lateral line; fins all yellowish; a small dark spot on the membrane between the bases of the sixth and seventh dorsal spines, another between the seventh and eighth, and another between the third and fourth soft-rays. Hab. Canary Islands ; off the coast of Angola. This species does not appear to have been recognized since originally described by Sauvage, and I have some doubt whether the examples from Angola are really referable to it. Dr W. H. Longley has sent me some notes on the type of S. canariensis (185 mm.) in the Paris Museum, but, although there appear to be some minor differences between this and the specimens described above, I do not think it advisable to give the latter a new name without actual comparison. Scorpaena laevis, Troschel. 1866, Arch. Nati/rgesch., xxxii (i), p. 206. ? Scorpaena senegalensis. Fowler, 1919, Proc. U.S. Nat. Mus., lvi, p. 214. Hab. Cape Verde Islands. A description and figure of this species has been given on p. 6 of this report. WEST AFRICA 31 Scorpaena senegalensis, Steindachner. 1881, Denkschr. Akad. Wiss. Wien, XLiv, p. 31, pi. iv. Very closely related to S. laevis, but depth of head at origin of dorsal i| in its length. Snout longer than eye, diameter of which is 5^ in length of head and about equal to interorbital width. Spines on head generally more acute. Supraorbital tentacle larger and much branched, its length | to | diameter of eye. Dorsal XII g. Anal III 5. Pectoral with 19 rays, extending nearly to above origin of anal, its length (measured from upper angle) 1 1 in that of head. Hab. Coasts of tropical West Africa. In the British Museum 2 specimens, 125 and 280 mm. in total length, from the Niger and the Gold Coast. Scorpaena madurensis, Cuv. and Val. 1833, Hist. Nat. Poiss., ix, p. 463. Sebastes maderensis, Lowe, 1839, Trans. Zool. Soc, 11 (3), p. 175; Giinther, i860. Cat. Fish., u, p. 102; Lowe, i860. Hist. Fish. Madeira, p. 177; Steindachner, 1867, SitzBer. Akad. Wiss. Wien, LVi (i), p. 673; Collett, 1896, Res. Camp. Sci. Monaco, x, p. 15; Collett, 1897, Arch. Naturv. Christian., xix. No. 7, p. 4; Kolombatovic, 1904, Hrvat. Naravosl. Driist. Ghjsnik, XV, p. 186; Roiile, 1907, Arch. Zool. exper. gen. (4) vi. Notes et Revue, p. xvi; Jaquet, 1907, Bull. Inst, ocean. Monaco, cix; Fage, 1918, Rep. Danish Ocean. Exped. 1908-10, II, A 3, p. 102. Scorpaena rubellio, Jordan and Gunn, 1898, Proc. Acad. Nat. Sci. Philad., p. 344. Depth of body 2f to 3 in the length, length of head 2\ to 2f . Snout as long as or shorter than eye, diameter of which is 3I to 3! in length of head and i| to if times interorbital width. Head with numerous small pores; cheeks and opercular region with visible scales. No occipital pit. 2 strong praeorbital spines, but only 2 or 3 spines on hinder part of suborbital ridge. A small supraorbital tentacle sometimes present ;^ a few small membranous processes on head. Length of band of palatine teeth a little more than I diameter of eye. 10 to 12 gill-rakers on lower part of anterior arch. Scales spinulose and ciliated ; 5 series between last soft-ray of dorsal and lateral line ; breast scaled; no membranous processes on body. Dorsal XII 9 or 10; fourth to sixth spines longest, about twice in length of head. Anal III 5; second spine longer and stronger than third. Pectoral with 15 or 16 rays, extending to above vent or beyond; base broad, the lowermost ray inserted level with root of pelvic spine and in advance of uppermost. Brownish, with four irregular darker cross-bars, the first just behind the head, the last on the caudal peduncle; dorsal, anal and caudal fins spotted and blotched with dark brown, the caudal with a broad transverse bar of the same colour; pectoral spotted and barred with brown, and with some small white spots in the axil. Hab. Mediterranean and adjacent parts of the Atlantic; Azores. In the British Museum 17 specimens, 62-140 mm. in total length. 1 This is particularly well developed in two examples from Cyprus, which in all other respects appear to be exactly similar to examples of similar size from other parts of the Mediterranean and from Madeira. 32 DISCOVERY REPORTS Scorpaena echinata, Koehler. 1896, Ann. Univ. Lyon, xxvi, p. 478, pi. xxvii, figs. 4-6. Scorpaena cristulata, Holt and Byrne, 1908, Fisheries Ireland Sci. Invest., 1906, v, p. 20, pi. ii. Depth of body af to 3^ in the length, length of head 2^ to 2^. Snout about as long as eye, diameter of which is 4I to 4! in length of head and twice or nearly twice width of interorbital space, which is very shallow. Top of head, cheeks and opercular region with visible scales. No occipital pit. Two blunt praeorbital spines ; about 7 or 8 spines on suborbital ridge, which is prominent. No supraorbital tentacle, but some small filamentous processes scattered over the head. Length of band of palatine teeth about f diameter of eye. 11 or 12 gill-rakers on lower part of anterior arch. Scales spinulose and ciliated ; 6 or 7 series between last soft-ray of dorsal and lateral line ; a few fila- mentous processes along lateral line; breast scaled. Dorsal XI-XII 9-10; fourth to sixth spines longest, 3^ to 3^ in length of head; soft dorsal largely covered with scales. Anal III 5; second spine longer and stronger than third. Pectoral with 21 or 22 rays, extending about to above vent ; the lower rays much thickened ; base broad. Pelvics not nearly reaching vent. Yellowish brown (reddish in life); uniform or with irregular patches of black on body; a black area covering greater part of spinous dorsal, and another on soft dorsal ; anal sometimes with a large black blotch ; pectoral with a large dusky area in the centre. Hab. Deep water off the west and south-west of Ireland and in the Bay of Biscay. In the British Museum 6 specimens, 330-510 mm. in total length. It is possible that this species will prove to be identical with S. cristulata, Goode and Bean, from off the coast of Georgia, U.S.A., as suggested by Koehler himself, but I have hesitated to unite two species from different sides of the Atlantic without actual comparison of specimens. I have not been able to examine examples of the American species, and, judging from the published description and figure (which exhibit certain discrepancies), cannot find any definite characters to separate this from S. echinata, although it is possible that the scales will prove to be larger in the latter. Meanwhile, I think it better to regard the two species as distinct. ^. capensis, Gilchrist and von Bonde,^ which Barnard'^ doubtfully places in the genus Sebastosemus, Gill, is clearly related to Scorpaena ecJmiata and S. cristulata, but has thirteen dorsal spines. Further, the eye is larger, the interorbital space much narrower, the head less heavily armed, the maxillary broader, and the pectoral fin shorter. In spite of the difference in the number of dorsal spines, there can be little doubt that the two northern Atlantic species and that from the Cape are congeneric, and it seems probable that this character has considerably less value in the differentiation of genera than has generally been supposed. 1 1924, Rep. Fish. Mar. Biol. Surv. S. Afric, in (1922), Spec. Rep. No. VII, p. 18. ^ 1927, Ami. S. Afr. Mus., xxi, p. 909. WEST AFRICA 33 Genus Setarches, Johnson 1862, Proc. Zool. Soc, p. 176. Type Setarches guentheri, Johnson. Only one species of this genus occurs in the Atlantic, namely, S. guentheri, Johnson, from Madeira. In the British Museum 3 specimens, 220-290 mm. in total length, including the type of the species. PLATYCEPHALIDAE Platycephalus gruveli, Pellegrin. 1905, Bull. Soc. zool. Fr., xxx, p. 138, pi. iii, fig. i; 1914, An?i. Inst, oceanogr. Paris, vi (4), p. 81, figs. 14, 15. St. 279. 10. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m. : 2 specimens, 135-140 mm. Hab. Tropical West Africa. TRIGLIDAE Lepidotrigla cadmani, Regan. 1915, Ann. Mag. Nat. Hist. (8) xv, p. 128. St. 272. 30. vii. 27. Off Elephant Bay, Angola. Large otter trawl, 73-91 m. : 6 specimens, 125-240 mm. St. 279. 10. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m. : 16 specimens, 125-165 mm. Hab. Off the coasts of Angola and French Congo. This species was known previously only from the types, five specimens 130-170 mm. in total length, from Lagos. BOTHIDAEi Eucitharus linguatula (Linn.). Norman, 1930, Discovery Reports, \i, p. 359. St. 272, St. 274, St. 279. Off Angola and French Congo. 9 specimens, 28-212 mm. Arnoglossus imperialis (Rafin.). Norman, 1930, t.c, p. 360. St. 272, St. 274, St. 279. Off Angola and French Congo. 3 specimens, 75-90 mm. The following additional specimens have come to light since the publication of my previous report: St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Net (4 mm. mesh) attached to back of trawl, 64-65 m. : 3 specimens, 29-30 mm. Bothus podas (Delaroche). Norman, 1930, t.c, p. 362. St. 271, St. 299. Angola and Cape Verde Islands. 4 specimens, 38-73 mm. ^ Some of the Heterosomata have been dealt with in a previous report. For the sake of completeness, the species are listed again here. 34 DISCOVERY REPORTS SOLEIDAE Solea (Dicologlossa) chirophthalmus, Regan. Norman, 1930, t.c, p. 363. St. 274. Off Angola. 2 specimens, 145, 210 mm. 3 additional specimens, 245-255 mm., from the same locality. CYNOGLOSSIDAE Cynoglossus (Areliscus) lagoensis, Regan. 1915, Ann. Mag. Nat. Hist. (8) xv, p. 129. St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Large otter trawl, 64-65 m.: 2 specimens, 455. 510 mm. Hab. Coast of Angola. Symphurus nigrescens, Rafin. Norman, 1930, t.c, p. 363. St. 274, St. 279. Off Angola and French Congo. The following additional specimen has come to light since the publication of my previous report: St. 274. 4. viii. 27. Off St Paul de Loanda, Angola. Net (4 mm. mesh) attached to back of trawl, 64-65 m. : I specimen, 48 mm. Fig. 12. Chirolophius kempi. Holotype. x |. LOPHIIDAE Chirolophius kempi, sp.n. St. 279. 10. viii. 27. Off Cape Lopez, French Congo. Large otter trawl, 58-67 m.: i specimen (holotype), 140 mm. Head nearly as broad as long, length (measured to lower angle of gill-opening) about \ that of fish. Snout nearly twice diameter of eye, which is about 7 in length of head and a little less than interorbital width. Praemaxillaries with two series of depressible teeth anteriorly and one series of small fixed teeth laterally ; teeth in lower jaw in about three series; two teeth on each side of vomer. A pair of obtuse divergent WEST AFRICA 35 spines on each side of the snout; each supraorbital ridge bearing 2 or 3 very blunt spines ; humeral spine simple, truncate or rounded posteriorly. First ray of spinous dorsal more than | as long as head, reaching beyond base of third ray when laid back, with a complicated terminal flap; second ray broken off; third and fourth progressively shorter, with pairs of short lateral branches ; fifth and sixth developed, connected by membrane basally. Soft dorsal with 8 rays, anal with 6, pectoral with 15 or 16. Greyish brown ; lower surface of pectoral rays blackish. This is the first species of the genus to be described from the Atlantic. Of the Indo- Pacific species it appears to approach most nearly to C. moselyi, Regan, from deep water north of New Guinea. I have much pleasure in naming this interesting fish after Dr Stanley Kemp, F.R.S., Director of Research of the Discovery Expedition. SOUTH AFRICA About 170 specimens were obtained in South African waters, representing 54 species. A few of these were collected at St. 90 (Basin of H.M. Dockyard, Simon's Town, False Bay), from the loth to the 12th of July, 1926, but the majority were obtained during the voyage made by Mr E, R. Gunther and Mr F. C. Fraser on the ' Richard Bennet ' from the 6th to the loth of July, 1927. A number of stations (A-Q) were made during this trip, and, in order to avoid needless repetition, the details of these stations are given below, only the letter referring to the particular station appearing under the different species. All the fishes were taken with the commercial otter trawl. Nearly all the species represented are to be found in Dr K. H. Barnard's excellent monograph on the Marine Fishes of South Africa, ^ and a reference to this work is given in each case. TrawHng stations of the Cape Trawler ' Richard Bennet ' 34° 00' S, 17° 58' E. 210-173 m. 34° 2' S, i7°4i'E. 311 m. 33°58'S, i7°44'E. 302 m. 33° 53' S, 17° 38' E. 310 m. 33° 57' S, 17° 29' E. 310-375 m. 34° 6' S, i7°42'E. 311 m. 34° 2' S, 17° 44' E. 311-292111. 34° 4' S, 17° 36' E. 292-402 m. 34° 8' S, 17° 33' E. 402-? 548 m. 33° 48' 30" S, 17° 35' E. 274-301 m. 33°44'S, i7°32'E. 301-311111. 33" 42' S, 17° 29' E. 311-402111. 33° 48' S, 17° 29' 30" E. 402-235 m. 33°48'S, i7°3o'E. 329 m. 33°42'S, i7°32'E. 329 m. 3o°47'S, i7°33'E. 311m. A. 6. vii. 27. B. 7. vii. 27. C. 7. vii. 27. D. 7. vii. 27. E. 7. vii. 27. F. 8. vii. 27. G. 8. vii. 27. H. 8. vii. 27. J. 8. vii. 27. K. 9. vii. 27. L. 9. vii. 27. M. 9. vii. 27. N. 9. vii. 27. 0. 10. vii. 27 P. 10. vii. 27 Q. 10. vii. 27 1 1925-7, Ann. S. Afr. Mus., xxi, 1065 pp., 37 pis., text-figs. 5-2 36 DISCOVERY REPORTS MYXINIDAE Eptatretus hexatrema (Miiller). Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 16. St. 90. lo-ii. vii. 26. Large gauze fish-trap, 10 m.: 2 specimens, 460, 610 mm. SCYLIORHINIDAE Scyliorhinus (Scyliorhinus) capensis (Miill. and Henle). Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 40. St. B. I female specimen, 570 mm. Scyliorhinus (Halaelurus) regani, Gilchrist. Barnard, t.c, p. 42. St. A. I male specimen, 540 mm. This species, which was not previously represented in the British Museum collection, appears to be most nearly related to S. natalensis, Regan, and S. polystigma, Regan. &^^^|S'£€^^^'^:^>^;5H^1^P^ Fig. 13. Scyliorhinus (Halaelurus) regani. x J. Scyliorhinus (Apristurus) saldanha, Barnard. Barnard, t.c, p. 44. St. J. I male specimen, 430 mm., 2 females, 385, 415 mm. These specimens agree fairly well with Barnard's description of S. saldanha. This species is apparently very close to S. profundorum, Goode and Bean, from the North Atlantic, and S. indicus, Brauer, from the Indian Ocean. CARCHARINIDAE Mustelus canis (Mitchill). Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 30. St. D. I young female specimen, 290 mm. (removed from a specimen, loio mm. long). SQUALIDAE Squalus acanthias, Linn. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 47. St. D. 3 embryos with yolk-sacs, 175-195 mm. SOUTH AFRICA 37 Squalus acutipinnis, Regan. Barnard, t.c, p. 48. St. A. I male specimen, 410 mm.; i female, 415 mm. St. O. I male specimen, 665 mm.; i female, 575 mm. Spinax spinax (Linn.). Etmopterus spinax, Barnard, t.c, p. 49. St. J. 2 male specimens, 330, 335 mm. Spinax granulosus, Giinth. Etmopterus granulosus, Barnard, t.c, p. 49, pi. ii, fig. S. St. J. 2 male specimens, 345, 370 mm. Spinax lucifer (Jordan and Snyder). Etmopterus lucifer, Barnard, t.c, p. 50. St. J. I male specimen, 410 mm. TORPEDINIDAE Narcobatus nobilianus (Bonap.). Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 89. St. ?. 2 female specimens, 440, 500 mm. RAJIDAE Raja alba, Lacep. Raia marginata, Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 65, pi. iv, fig. i. St. ?. I female specimen, 605 mm. (width of disc 465 mm.). Raja smithi, Miill. and Henle. Barnard, t.c, p. 66, pi. iv, fig. 4. St. N. I female specimen, 210 mm. (width of disc 150 mm.). Raja barnardi, sp.n. St. A. I immature male specimen, 375 mm. (width of disc 210 mm.). See description below. Raja caudaspinosa, Von Bonde and Swart. St. J. 2 young female specimens, 200, 220 mm. (width of disc 100, 105 mm.). St. N. I immature female specimen, 485 mm. (width of disc 270 mm.). Raja leopardus, Von Bonde and Swart. St. J. I young male specimen, 175 mm. (width of disc 95 mm.); i young female, 135 mm. (width of disc 75 mm.). Revision of the South African Species of the Genus Raja The difficulty of identifying the specimens of this genus obtained by the S.T. ' Richard Bennet ' led me to suggest to Dr K. H. Barnard of the South African Museum that it might be of interest to revise our knowledge of the South African species of Raja. Several European rays have been recorded from the Cape, but in very few cases 38 DISCOVERY REPORTS has an actual comparison of specimens from the two regions been made. In addition, Dr C. von Bonde and Mr D. B. Swart have published a report upon the skates and rays collected by the S.S. ' Pickle ', in which five new species of the genus Raja are described. Unfortunately, nearly all these species are based upon very small specimens, and no indication is given by the authors as to their relationships with previously known forms. The description of new species of this genus upon the basis of young individuals only is to be regretted, as these fishes change considerably with age, and until a complete series of stages becomes available it is quite impossible to refer such species to their correct place in the system. Through the kindness of Dr von Bonde I have been able to examine type material of R. parcomaciilata, R. albalinea, and R. leopardus, now preserved in the collection of the Government Marine Survey of South Africa: the types of R. caiidaspinosa and R. diirbanensis, said to be in the same collection, cannot unfortunately be found. Dr Barnard has been good enough to send to the British Museum as a loan all the specimens of Raja in the collection of the South African Museum, including three stuffed examples exhibited in the public galleries, as well as the type of R. spinacidermis. For this courtesy, and for the kindly interest that he has shown in this revision, I take this opportunity of offering my warmest thanks. The arrangement of the species adopted here still remains more or less tentative, and further material of most species, including as far as possible examples of all stages of both sexes, will be required before it will be possible to arrive at any definite con- clusions concerning the South African members of this difficult genus. Key to the South African Species I. Terminal parts of lateral line tubules on lower surface pigmented, appearing as small blackish spots and streaks; only one enlarged spine on back in adult; anterior margins of disc emarginate; vent nearer end of tail than tip of snout; length of snout about 4-J in width of disc. ''^«'«- II. No pigment spots or streaks on lower surface; vent about equidistant from tip of snout and end of tail or nearer the former; length of snout 4I to more than 6 in width of disc. A. Disc never completely spinulose; eye + spiracle i§ to nearly 3 in length of snout. I. Never more than one row of spines along middle of disc. a. Snout abruptly narrowed into a long sharp point; upper surface of disc quite smooth ; 40 to 46 rows of teeth ... ... ... ... • • • • • • alba. b. Snout not abruptly narrowed; upper surface of disc more or less spinulose in parts. a. Large buckler-like spines often present in mature females, mostly absent in males; 36 to 44 rows of teeth; tail with i (males) or 3 to 5 (females) rows of enlarged spines ... ... ... ... ■ • • • • • ■ ■ • rhtzacantnus. j3. No large buckler-like spines in either sex. * Width of disc f to * of total length of fish; eye + spiracle 2 to 2% in length of snout, which is 5 j to 6 in width of disc. I 26 to 28 rows of teeth; tail with only one row of enlarged spines in both sexes; no ocelli on pectoral fins ... ... ••• ••• ••• smithi. SOUTH AFRICA 39 ft 44 to 50 rows of teeth; tail with 3 to 5 rows of enlarged spines in both sexes ; a rounded ocellus on each pectoral fin ocellifera. ** Width of disc scarcely | of total length of fish; eye + spiracle 2| in length of snout, which is 4^ in width of disc; 40 to 42 rows of teeth ... barnardi. 2. 3 or more rows of spines along middle of disc (except in young); usually a triangular patch of enlarged spines on the shoulder; width of disc about f of total length of fish. a. 32 to 42 rows of teeth; interorbital width less than longitudinal diameter of eye; eye + spiracle if in length of snout ... ... ••• caudaspinosa. b. 50 to 80 rows of teeth; interorbital width equal to or greater than longitudinal diameter of eye; eye + spiracle 2 to 2f in length of snout kopardus. B. Upper surface of disc completely covered with close-set, fine, setiform spinules; no enlarged spines (except in young); eye + spiracle 3^ in length of snout; 60 rows of teeth (in adult) spinacidermis. Raja batis, Linnaeus. 1758, Syst. Nat., ed. 10, p. 231; Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 70, pi. iv, fig. 3; Clark, 1926, Fisheries, Scotland, Sci. Invest., 1926, i, p. 50, pis. xxxi, fig. b, xxxii, xxxiii, figs, a and b. Raia stabuliforis, von Bonde and Swart, 1923, Rep. Fish. Mar. Biol. Surv. S. Afric, in (1922), Spec. Rep. v, p. 12. Disc broader than long, its width about f of the total length ; anterior margins more or less undulated and deeply emarginate ; outer angles nearly rectangular. Vent rather nearer to end of tail than to tip of snout. Snout acutely pointed, its length about 4^ in width of disc. Interorbital width less than diameter of eye + spiracle, which is about 2f in length of snout. Internasal width about 2i in praeoral length of snout. Teeth more or less flat ; about 52 rows. Upper surface mainly smooth, but with some small scattered spinules, chiefly on snout, anterior margins of pectorals and middle of back ; 2 or 3 praeocular and i or 2 postocular spines ; a single large nuchal spine ; tail with a median series of about 21 strong spines, alternately larger and smaller, and with an irregular series of 4 or 5 spines on each side ; 3 spines between the dorsal fins. Lower surface rough on snout, but otherwise smooth. Upper surface brownish, with a few irregularly arranged darker spots ; lower surface greyish ; terminal parts of lateral line tubules pigmented, appearing as small blackish spots and streaks. Hab. Coasts of Europe, from Iceland and Scandinavia to Madeira ; Mediterranean ( ?) ; South Africa. Described from a single mounted female specimen, 680 mm. in total length (480 mm. across disc), from off Cape Point, 100 fathoms. It is possible that the South African form described here will eventually prove to be distinct from the European R. batis, but, as I have only seen a single stuffed example, I have hesitated to separate the two at present. Comparison with a European example of similar size suggests that the South African form may have a narrower interorbital region and perhaps a larger eye, but since artificial eyes have been inserted in the specimen accurate measurements are impossible. The arrangement of the spines on the tail appears to be different, the disc is rather more spinulose, and the enlarged nuchal spine has no counterpart in R. batis from Europe. 40 DISCOVERY REPORTS Raja alba, Lacepede. 1803, Hist. Nat. Poiss., v, p. 661, pi. xx, fig. 1; von Bonde and Swart, 1923, t.c, p. 5. Raja marginata, Lacepede, 1803, t.c, p. 662, pi. xx, fig. 2; Regan, 1908, Ann. Natal Mus., i, p. 242; Barnard, 1925, t.c, p. 65, pi. iv, fig. i; Clark, 1926, t.c, p. 47, pis. xxviii, xxix, XXX figs, a and b, xxxi fig. a. Disc broader than long, its width about f of the total length; anterior margins undulated ; outer angles pointed. Vent about equidistant from tip of snout and end of tail. Snout abruptly narrowed into a long sharp point, its length 4§ to 5I in width of disc. Interorbital width about equal to diameter of eye + spiracle, which is 2 (young) to nearly 3 in length of snout. Internasal width if (young) to 2 in praeoral length of snout. Teeth with long conical points in the middle of the jaws, more obtuse and with short points laterally; 40 to 46 rows. Upper surface quite smooth; one praeocular and generally one postocular spine ; no nuchal or scapular spines, and no median spines on disc; tail with a median series of 1 1 to 18 spines, extending forward to posterior end of base of pelvic, and with a lateral series on each side. Lower surface with small spines on the snout and along the anterior edges of the pectorals. Upper surface uniformly brownish or more or less spotted with white; lower surface white, the tail and margins of pectorals and pelvics often brownish or blackish, especially in the young. Hab. Coasts of Europe, from the English Channel to the Mediterranean ; coasts of northern and north-western Africa ; South Africa. Described from 12 specimens, 270-605 mm. in total length (205-465 mm. across disc), from Kalk Bay, Simonstown, Agulhas Bank, Cape St Blaize, and off Bird Island, Natal. Comparison of South African with European material shows that as a general rule the snout is a little longer in examples from Europe, but this character appears to be subject to considerable variation and I am unable to detect any other differences of importance. Clark has shown that Lacepede's young black-bordered R. marginata is identical with the white-bellied adult, R. alba, described by the same author. In accordance with Article 28 of the International Rules, the name alba (p. 661) takes precedence of marginata (p. 662). Raja rhizacanthus, Regan. Raja capensis {non Gmelin), Miiller and Henle, 1841, Plagiost., p. 151; Dumeril, 1865, Hist. Nat. Poiss., I, p. 540, pi. xii, figs. 11 and 12; Sauvage, 1891, Hist. Nat. Poiss. Madagascar, p. I ; von Bonde and Swart, 1923, t.c, p. 4. ? Raja capensis, Kner, 1869, Reise 'Novara', Zool. i, 5. Fische, p. 419. Raia rhizacanthus, Regan, 1906, Ann. Natal Mus., i, p. 3, pi. iii. Raia clavata, Barnard, 1925, t.c, p. 64, pi. iv, fig. 2. Disc broader than long, its width f to ^. of the total length ; anterior margins more or less undulated; outer angles pointed. Vent equidistant from tip of snout and end of tail or a little nearer to the former. Snout with a short, obtuse, triangular projection, its length 5j to more than 6 in width of disc. Interorbital width equal to or rather greater than (a little less than in young) diameter of eye + spiracle, which is 2\ to SOUTH AFRICA 41 about 2f in length of snout. Internasal width i^ to if in praeoral length of snout. Teeth pointed in males, at least in middle of jaws, blunt in females ; 36 to 44 rows (fewer in young). Upper surface of disc and tail in young with small scattered spines, chiefly confined to the snout, interorbital region, anterior parts of pectorals and the sides of the tail ; in adults the disc and tail are more or less covered with small spinules ; mature females sometimes with large round "bucklers" bearing claw-like spines scattered irregularly over the upper and lower surfaces, which are nearly always absent in males ; 2 praeocular and 3 postocular spines in the young, the numbers being reduced or the spines disappearing altogether in the adults ; young with a pair of scapular spines, disappearing in the adults ; a median series of 27 to 45 spines, extending anteriorly to beyond the scapulary region in the young, but scarcely beyond the pelvic region in adults ; tail in females with one or two lateral series of spines. Lower surface smooth in the young, but adults with some small spines, especially on the snout. Brownish or greyish, with or without darker and paler markings; young generally with a dark ocellated spot, sometimes circular, sometimes oblong, near the middle of the base of each pectoral ; lower surface pale, occasionally with some dark patches. Hob. South Africa, from Walfish Bay to the coast of Natal ; Madagascar. Described from 13 specimens, 135-840 mm. in total length (90-620 mm. across disc), from Kalk Bay, False Bay, Agulhas Bank, off Cape St Blaize, and off Bird Island, Natal. This species is closely related to R. clavata, the thornback ray of European seas, but the two appear to be distinct. In R. clavata the upper surface is entirely spinulose, even in the newly hatched young, whereas in the African species the adults are never completely covered with spinules. Comparison of specimens of equal size shows that in R. clavata the spinules are always more closely set. Further, in R. clavata the tail is constantly rather longer, the vent being distinctly nearer to the tip of the snout than to the end of the tail, and there is a difference in the shape of the snout. The four large specimens from near Cape St Blaize and west of Cape Point, sent to the British Museum in 1900 by Dr Gilchrist, and identified by Dr G. A. Boulenger as R. batis, prove to belong to this species. Raja smithi, Miiller and Henle. 1841, Plagiost., p. 150, pi. xlix, fig. i ; Barnard, 1925, t.c, p. 66, pi. iv, fig. 4. Disc broader than long, its width f to f of the total length ; anterior margins a little undulated; outer angles obtusely pointed. Vent nearly equidistant from tip of snout and end of tail. Snout a little produced, its length 5f to 5! in width of disc. Inter- orbital width equal to or a little less than diameter of eye + spiracle, which is 2 to about 2f in length of snout. Internasal width i| to if in praeoral length of snout. Teeth rather widely spaced, those in the middle of the jaws pointed in both sexes ; 26 to 28 rows. Upper surface with small four- or five-rooted spinules on snout, interorbital region, anterior, posterior and outer parts of pectorals, and on the middle of the back, the last being more numerous in adults; no enlarged ocular spines; young with 14 to 16 median spines on the tail, and adults with 4 or 5 additional median spines on the 42 DISCOVERY RI':PORTS back behind the nuchal region ; sides of tail with several series of very small spinules. Lower surface quite smooth. Upper surface more or less uniformly brownish or greyish ; lower surface white, sometimes with irregular black blotches and with black margins to the posterior part of the disc ; lower surface of tail black. Hab. South Africa. Described from 3 specimens, 210-520 mm. in total length (150-330 mm. across disc), including the type of the species (a dried skin). Barnard regards R. eatoni, Giinther, from Kerguelen Island, as a local variety of R. smithi, but, although the two species are clearly related, they seem to be distinct. R. eatoni has a longer and more pointed snout and there are obvious differences in the spination. Raja ocellifera, Regan. 1906, Ann. Natal Mus., i, p. 2, pi. ii; 1908, t.c, p. 242; Barnard, 1925, t.c, p. 67. Disc broader than long, its width | to f of the total length ; anterior margins more or less undulated ; outer angles rounded or obtusely pointed. Vent a little nearer to tip of snout than to end of tail. Snout with a short, obtuse, triangular projection, its length 5I (young) to 6 in width of disc. Interorbital width greater than diameter of eye, but less than that of eye + spiracle, which is 2 (young) to 2f in length of snout. Internasal width i^ to 1 1 in praeoral length of snout. Teeth with sharp points in males (but often much worn), more or less obtuse in females ; 44 to 50 rows. Upper surface of disc and tail smooth, except for a few small spinules on tip of snout, on rostral ridges, and on anterior margins of pectorals, these spinules stronger in males ; 2 to 4 praeocular and 2 or 3 postocular spines; usually i to 3 median nuchal spines; young with a pair of scapular spines ; a median series of spines commencing on posterior part of body and extending on to tail, commencing further forward in females than in males ; in young of both sexes the series commences immediately behind the suprascapular region ; tail with one or two series of spines on each side. Lower surface quite smooth. Upper surface brownish, with or without small darker spots ; a large bluish-black, white-edged ocellus near the middle of the base of each pectoral ; lower surface vmiformly pale. Hab. South Africa, from False Bay to Natal. Described from 12 specimens, 125-490 mm. in total length (88-340 mm. across disc), from False Bay, Agulhas Bank, off Cape St Blaize, Algoa Bay, and the coast of Natal, including the types of the species. This species is closely related to R. miraletiis from the Mediterranean and the west coast of Africa, which has, however, a somewhat longer tail, a longer snout (4^ to 5 J in width of disc and 2\ to 2| times eye + spiracle), narrower interorbital region (equal to or less than, only occasionally greater than, diameter of eye), and there are only 38 to 42 rows of teeth. In R. miraletus the ocellus is nearly circular, whereas, in R. ocellifera this tends to be horizontally ovate. In some respects the two specimens obtained by the ' Discovery ' off the coast of Angola approach the African form, but should, I think, be referred to R. miraletus. SOUTH AFRICA 43 The specimens from near Cape St Blaize, sent to the British Museum in 1900 by Dr Gilchrist, and identified by Dr Boulenger as R. miraletus, prove to belong to R. ocelli/era. Raja barnardi, sp.n. Disc a little broader than long, its width scarcely f of the total length; anterior margins a little undulated; outer angles smoothly rounded. Vent a little nearer to tip of snout than to end of tail. Snoutwith a rather short, obtuse, triangular projection, its length 4I in width of disc. Interorbital width about equal to diameter of eye; eye + spiracle 2| in length of snout. Inter- nasal width 2^ in praeoral length of snout. Teeth more or less pointed in middle of jaws; 40 to 42 rows. Upper surface of disc and tail mainly smooth, but a large patch of spinules on anterior part of each pectoral, and some scattered spinules on snout, inter- orbital region, middle of back and hinder parts of pectorals; 2 praeocular and 4 post- ocular spines; 3 median nuchal spines, with a smaller one on each side; a single median spine above the suprascapulary region; 2 scapular spines; a series of 24 median spines extending from just behind the suprascapulary region to the first dorsal; anterior part of tail with a some- what irregular series of spines on each side ; edges of tail with numerous small spinules. Lower surface quite smooth except at edges of snout. Upper surface brownish, with traces of small pale spots; lower surface uniformly pale. Hob. Ofl^ Cape Town. Fig. 14. Raja barnardi. Holotype. X; Described from a single male specimen, 375 mm. in total length (210 mm. across disc), from 34° S, 17° 58' E, at a depth of 210-173 m.; obtained by the Discovery- Expedition. Raja caudaspinosa, von Bonde and Swart. ? Raia albaliuea, von Bonde and Swart, 1923, t.c, p. 6, pi. xx, fig. i. Raia caudaspinosa, von Bonde and Swart, 1923, t.c, p. 8, pi. xxi, fig. i; Barnard, 1925, t.c, p. 66. Disc rather broader than long, its width about f of the total length ; anterior margins a little undulated ; outer angles obtusely pointed. Vent much nearer to tip of snout 6-2 44 DISCOVERY REPORTS than to end of tail. Snout with a very small and obtuse projection, its length 6 in width of disc. Interorbital width less than diameter of eye; eye + spiracle if in length of snout. Internasal width 2j in praeoral length of snout. Teeth mostly flat, but some in middle of jaws more or less bluntly pointed ; (32 to 36) 40 to 42 rows. Upper surface of disc mainly smooth, but a large patch of stellate-based spinules on anterior part of each pectoral ; a group of spines on the snout ; a series of 9 spines above each orbit and spiracle, and a pair between the spiracles ; 2 median nuchal spines and a pair of scapular spines ; a series of median spines extending from just behind the suprascapulary region to the first dorsal, and a lateral series of smaller spines on each side ; on the tail there is an additional series at each edge, making 5 rows in this region. Lower surface quite smooth. Upper surface more or less uniformly brownish grey. Hah. South Africa, from off Cape Town to the coast of Natal. Described from a single female specimen, 485 mm. in total length (270 mm. across disc), from 33° 48' S, 17° 29' E, at a depth of 402-235 m. ; obtained by the Discovery Expedition. The unique holotype was a female, 346 mm. in total length (172 mm. across disc). It seems probable that R. alhalinea represents the young of the species described above, but in the absence of examples of intermediate size it is impossible to confirm this. The following description of the young stages is based upon the very small type of R. albalinea (no mm.) and two somewhat larger examples (200, 220 mm.) obtained by the Discovery Expedition: Disc subcircular, its width about | the total length; anterior margins very little undulated. Length of snout about 6 in width of disc. Interorbital width equal to or rather less than diameter of eye ; eye + spiracle i J to about 1 1 in length of snout, 30 to 34 rows of teeth. Upper surface of disc more or less covered with scattered spinules in the smallest example, but in the larger these tend to be more strongly developed on the anterior parts of the pectorals; 2 praeocular and 2 or 3 postocular spines; 2 or 3 median nuchal spines and 2 or 3 scapular spines; a median series of 22 to 27 strong spines extending from just behind the suprascapulary region to the first dorsal ; in the type this series is continuous with the nuchal spines ; in one of the larger specimens there are 2 or 3 spines on each side of the median row on the disc ; tail with several series of spinules laterally, those at edge larger. Lower surface quite smooth. Upper surface pale brownish grey, with some rather indistinct and nearly horizontal white lines near the edges of the pectoral fins. Described from 3 specimens, 110-220 mm. in total length (55-105 mm. across disc). Raja leopardus, von Bonde and Swart. Rata quadritnaculata (non Risso), von Bonde and Swart, 1923, t.c, p. 5; Barnard, 1925, t.c, p. 70, pi. iv, fig. 5. Raia leopardus, von Bonde and Swart, 1923, t.c, p. 7, pi. xx, fig. 2; Barnard, 1925, t.c, p. 74. Rata lintea, Barnard, 1925, t.c, p. 72. Raia naevus, Barnard, 1925, t.c, p. 72. Disc broader than long, its width about f of the total length ; anterior margins more or less undulated, except in young, deeply notched in adult males ; outer angles broadly SOUTH AFRICA 45 I'ounded. Vent a little nearer to tip of snout than to end of tail. Snout with a rather short, obtuse, triangular projection, its length 4I to more than 6 in width of disc. Interorbital width equal to (young) or greater than diameter of eye, but always less than that of eye + spiracle, which is 2 to 2f in length of snout. Internasal width if to 2^ in praeoral length of snout. Teeth pointed in mature individuals of both sexes, but often worn quite flat; 50 to 80 rows. Upper surface of disc and tail mainly smooth, but with some scattered small, often stellate-based spinules, chiefly on snout, anterior parts of pectorals and sides of tail; larger spines all with stellate bases; a series of spines above each orbit and spiracle, and a pair between the spiracles; young with 2 to 4 median nuchal spines, and 2 or 3 scapular spines ; in adults there is a triangular patch of spines on the nucho-scapulary region ; young with a series of 25 to 27 median spines extending from just behind the suprascapulary region to the first dorsal; these are gradually reduced during growth, being represented by rather obtuse spines in a mature male, and are absent altogether in large females; i or 2 lateral series on each side of the median line of the back (except in young), persisting in the largest specimens in which the median series has disappeared, and 2 series, with some irregularly arranged additional spines, on each side of the tail. Lower surface rough on snout and (in adults) the anterior edges of the pectorals; otherwise smooth. Upper surface brownish or greyish, sometimes with numerous round dark spots, chiefly obvious in the young; sometimes traces of pale, dark-edged ocelli, and occasionally a very faint naevus-like ocellus near the middle of the base of each pectoral ; lower surface uniformly pale or with some irregularly shaped but more or less symmetrically arranged greyish or blackish patches on pectorals and pelvics. Hab. South-western Africa, off Cape Peninsula and Saldanha Bay; coast of Natal. Described from 14 specimens, 110-975 ^^^^- in total length (58-625 mm. across disc), from off Dassen Island, Table Bay and Cape Point, and from the coast of Natal, in- cluding the types of the species. I have little doubt, after examining all the available material, that the forms described by Barnard as quadrimacnlata and naevm represent the same species, the description of the former being based upon very large female specimens. The types of R. leopardus are both very small (no, 180 mm. in total length; 58, 95 mm. across disc), but are almost certainly the young of the species described above. The mounted specimen, 740 mm. in total length (485 mm. across disc), identified by Barnard as R. lintea, appears to belong to this species. This species has been confused with R. naeviis of European seas, but the two are quite distinct. In naeviis the snout is rather shorter and blunter, its length 5 (young) to about 6| in width of disc ; the interorbital width is equal to or rather less than diameter of eye ; and there are only 54 to 60 rows of teeth. Further, comparison of specimens of similar size shows the spination to be different, the small spinules being much more numerous and better developed in the European species. Also, R. leopardus lacks the large ocellus which is so characteristic of R. naeviis. 46 DISCOVERY REPORTS Raja spinacidermis, Barnard. Raia spinacidermis, Barnard, 1923, Ann. S. Afr. Mus., xiii, p. 440; Barnard, 1925, t.c, p. 73, pi. iv, fig. 6. ? Rata durbanensis, von Bonde and Swart, 1923, t.c, p. 11, pi. xxii, fig. i; Barnard, 1925, t.c, p. 69. ? Raia plutonia, Barnard, 1925, t.c, p. 68. Disc broader than long, its width about f of the total length ; anterior margins scarcely- undulated ; outer angles broadly rounded. Vent very little nearer to tip of snout than to end of tail. Snout pointed but not produced, its length 4! in width of disc. Inter- orbital width a little greater than diameter of eye + spiracle, which is 3^ in length of snout. Internasal width 2 in praeoral length of snout. Teeth in middle of jaws slightly pointed ; 60 rows. Upper surface of disc and tail wholly covered with closely-set, fine, setiform spinules, which are larger and closer together on the tail than elsewhere; no enlarged spines. Lower surface of disc smooth; tail, except the median line of the basal part, spinulated like the upper surface. Upper surface pale slaty grey, becoming a little darker towards the hinder margins of the pectorals and distinctly darker on pelvics ; lower surface similarly and as deeply coloured as upper surface. Hab. South Africa. Described from the single type specimen, a female, 600 mm. in total length (510 mm. across disc), believed to be from off Cape Point in deep water. This species appears to be most nearly related to the European shagreen ray, R. fullonica, Linnaeus. It seems probable that the two very young specimens, 120 and 190 mm. in total length (68 and 100 mm. across disc), from south of the Agulhas Bank and from off Cape Point, identified by Barnard as R. plutonia, belong here. In the smaller of these there is a median series of spines on the disc and tail, but these are already disappearing in the larger specimen: there are also some spines above the orbits and spiracles, and one or two pairs of scapular spines. DOUBTFUL SPECIES Raja montagui, Fowler (= R. maculato, Montagu nee Shaw). Raia maculata, Barnard, 1925, t.c, p. 71. This species has been recorded by Bleeker (i860) and Pappe (1866), but it is probable that the true montagui does not occur in South Africa and that their specimens should be referred to some other species. Raja parcomaculata, von Bonde and Swart. 1923, t.c, p. 9, pi. xxi, fig. 2. The type specimen from Natal examined by me is only 181 mm. in total length (60 mm. across disc). SOUTH AFRICA 47 CHIMAERIDAE Chimaera africana, Gilchrist. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 95. St. ?. I female specimen, 625 mm. This species is well distinguished from C. monstrosa, Linn., by the absence of a distinct anal fin, as well as by the shorter pectoral fins, black coloration, etc. The lateral line in the present specimen is distinctly sinuous, and not straight as described by Barnard. Further, with regard to the cephalic branches of the lateral line, on one side of the head the opercular and malar branches arise together from the suborbital, whereas, on the other side they are united to form a common branch for a very short distance. It is not unlikely that C. africana will eventually prove to be identical with C affinis, Capello (= C. pliimbea. Gill, and C. abbreviata, Gill), which has been described from large specimens taken on both sides of the Atlantic. The caudal filament perhaps becomes shorter with age. ARIIDAE Galeichthys feliceps, Cuv. and Val. Boulenger, 1911, Cat. Fresh-water Fish. Africa, 11, p. 381, fig. 295. 29. vii. 27. Simon's Town, i specimen, 150 mm. Found while draining the dry dock. CONGRIDAE Congermuraena albescens, Barnard. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 189, pi. ix, fig. i. St. E. I specimen, 725 mm. SCOMBRESOCIDAE Scombresox saurus (Walbaum). Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 259, fig. 16. St. K. I specimen, 355 mm. (from the stomach of Merluccius capensis). MACRORHAMPHOSIDAE Notopogon macrosolen, Barnard. Barnard, 1925, Atin. S. Afr. Mus., xxi, p. 279, pi. xi, fig. 3. St. G. I specimen, 220 mm. St. M. 1 specimen, 250 mm. St. O. I specimen, 263 mm. MACRURIDAE Coryphaenoides (Paramacrurus) fasciatus (Giinther). Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 340. St. A. 6 specimens, 260-345 mm. St. B. and H. 4 specimens, 130-240 mm. St. J. 9 specimens, 390-650 mm. 48 DISCOVERY REPORTS Coryphaenoides (Oxygadus) braueri, Barnard. Barnard, t.c, p. 342, pi. xiii, fig. 5. St. J. 12 specimens, 220-405 mm. (the largest example has the tail broken). Malacocephalus laevis (Lowe). Barnard, t.c, p. 344. St. B. 4 specimens, 510-660 mm. (three of the examples have the tail damaged). Lionurus leonis, Barnard. Barnard, t.c, p. 349, pi. xiii, fig. 6. St. O. I specimen, 360 mm. Lionurus sp. St. J. I specimen, 245 mm. MERLUCCIIDAE Merluccius capensis, Casteln. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 320, pi. xii, fig. 5. St. B. I specimen, 173 mm. St. C. 3 specimens, 405-530 mm. St. P. 5 specimens, 640-850 mm. Barnard notes that "there still seems room for doubt as to whether the Cape Hake is really distinct from the northern Atlantic M. vtilgaris". I have carefully compared examples of both species, and find that, in addition to the somewhat larger scales, the Cape form may be readily distinguished by the larger eye, greater number of gill-rakers (13 or 14 instead of 7 or 8), and longer pectoral fin. GADIDAE Lepidion capensis, Gilchrist. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 324, pi. xiii, fig. i. St. J. 6 specimens, 310-500 mm. Lepidion natalensis, Gilchrist. Barnard, t.c, p. 324. St. J. I specimen, 408 mm. A Synopsis of the Species of Lepidion In order to satisfy myself as to the systematic position and nomenclature of the two species found at the Cape, I have been led to examine all the specimens of this genus in the British Museum collection, and have prepared a brief synopsis of the group. Genus Lepidion^ Lepidion, Swainson, 1838, N.H. Fishes etc, i, p. 318. Type Gadus lepidion, Risso. Haloporphyrus, Giinther, 1862, Cat. Fish., iv, p. 358. Type Gadus lepidion, Risso. ^ According to the International Rules, this is not invalidated by Lepidia, Savigny — a genus of worms. SOUTH AFRICA 49 Key to the Species^ I. Snout li to 1 1 times eye, which is 4! to 5^ in head; barbel longer than eye. A. Depth 4 to 4J in length; pectoral about i| in head. 1. Dorsal 4 + 52-56; 15 or 16 scales between first dorsal fin and lateral line ... guentheri. 2. Dorsal 4+60; about 18 scales between first dorsal fin and lateral line ... oidema. B. Depth more than 5 in length; pectoral about iJ in head; dorsal and anal with deep black margins, the black area covering greater part of fin posteriorly ... ... natalensis. II. Snout as long as or shorter than eye, which is zf to 3f in head ; barbel generally shorter than eye. A. Filamentous dorsal ray much longer than head; dorsal 4 or 5 + 52-62; 155 or more scales in lateral line. 1. About 155 to 180 scales in lateral line, about 13 to 16 between first dorsal fin and lateral line; caudal peduncle 3^ to 4 times as long as deep. a. Dorsal 4 (5) + 52, anal 46-48; scales 1 55-1 60/1 3 or 14 ... ... ... lepidion. b. Dorsal 4+56-62; anal 49-54; scales 180/15 or 16 eqiies. 2. About 220 to 250 scales in lateral line, about 18 to 20 between first dorsal fin and lateral line; caudal peduncle i\ to nearly 3 times as long as deep. a. Last ray of second dorsal nearly above that of anal; eye 3!, pectoral if in head; dorsal 5 + 60, anal 52 ... ... ... ... ... ... ... ... inosimae. b. Last ray of second dorsal posterior to that of anal; eye 2f to 35, pectoral if to if in head; dorsal 5 + 52-56, anal 46-50. a. Caudal peduncle ig to twice as long as deep; filamentous dorsal ray not broad and compressed ... ... ... ... ... ... ... ... capensis. /3. Caudal peduncle 23 to nearly 3 times as long as deep ; filamentous dorsal ray broad and compressed ... ... ... ... ... ... ... ensiferus. B. Filamentous dorsal ray much shorter than head; about 140 scales in lateral line; dorsal 5 + 50; caudal peduncle about 2i times as long as deep ... ... ... ... nwdestus. Lepidion guentheri (Giglioli). Haloporphyriis lepidion {noii Risso), Johnson, 1862, Ann. Mag. Nat. Hist. (3) x, p. 166; Giinther, 1862, Cat. Fish., iv, p. 358. Haloporphyrus guentheri, Giglioli, 1880, Nature, XXI, p. 202; Vinciguerra, 1883, Ann. Mus. Civ. stor. nat. Genova, xviii, p. 558; Giinther, 1887, Deep-Sea Fish. 'Challenger' , p. 90, pi. xviii, fig. A; Carus, 1889-93, Prodr. Faun. Medit., 11, p. 576. Lepidion guentheri, Goode and Bean, 1895, Ocean. Ichth., p. 370. Hab. Mediterranean and adjacent parts of the Atlantic. In the British Museum 2 specimens, 500 and 610 mm. in total length. Lepidion oidema (Tanaka). Haloporphyrus oidema, Tanaka, 1927, Fig. Descr. Fish. Japan, XLI, p. 796, pi. clxxi, fig. 472. Hab. Deep water off Misaki, Sagami Prov., Japan. The type was 345 mm. long. This species may prove to be identical with the pre- ceding one. 1 Microlepidium , Garman, is distinguished by the longer lower jaw, higher number of rays in the first dorsal fin, absence of filamentous dorsal ray, much larger number of pyloric appendages, elc. There are two species: M. verecundum (Gilbert) and M. grandiceps, Garman. so DISCOVERY REPORTS Lepidion natalensis, Gilchrist. Lepidion natalensis, Gilchrist, 1922, Rep. Fish. Mar. Biol. Surv. S. Afric, 11 (1921), Spec. Rep. HI, p. 62; Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 324. Depth of body about 5^ in the length, length of head nearly 4. Snout i| times as long as eye, diameter of which is 5 in length of head and about equal to interorbital width. Maxillary extending to a little beyond middle of eye; barbel very slightly longer than eye. 9 gill-rakers on lower part of anterior arch. Dorsal 5+58 ;^ filamentous ray if times as long as head. Anal 55. Pectoral about i| in length of head. Pelvic 7-rayed; not reaching vent, longest ray f length of head. About 20 pyloric caeca. Fig. 15. Lepidio?! natalensis. xj. Pinkish grey; dorsal and anal fins with deep black margins, the black area becoming broader behind and covering the greater part of the posterior parts of the fins ; caudal blackish ; pectorals and pelvics dusky. Hab. Coasts of south-east Africa. In the British Museum a single specimen, 408 mm. in total length. Lepidion lepidion (Risso). Gadus lepidion, Risso, 18 10, Ich. Nice, p. 118. Lotta lepidion, Risso, 1826, H.N. Europe, ni, p. 218. Lepidion rissoi, Swainson, 1838, N.H. Fishes etc., 1, p. 319. Lepidion rubescens, Swainson, 1839, N.H. Fishes etc., 11, p. 300. Haloporphyrus lepidion, Giglioli, 1880, Nature, xxi, p. 202; Vinciguerra, 1883, Ann. Mus. Civ. stor. nat. Genova, xvni, p. 554, pi. iii; Giinther, 1887, Deep-Sea Fish. 'Challenger', p. 91; Cams, 1889-93, P^odr. Faun. Medit., u, p. 576; Goode and Bean, 1895, Ocean. Ichth., p. 370, fig- 323- Hab. Western Mediterranean. In the British Museum a single specimen, 253 mm. in total length, from Nice. Lepidion eques (Giinther). Haloporphyrus eques, Giinther, 1887, t.c, p. 91, pi. xviii, fig. B; Holt and Caldervvood, 1895, Sci. Trans. R. Dublin Soc, v (11), p. 446, pi. xxxix, figs, i, 2; Koehler, 1896, Ann. Univ. Lyon, XXVI, p. 487; Liitken, 1898, Danish Ingolf Exped., u, i. Ichth. Res., p. 30, pi. iv, fig. 7. * Gilchrist gives 8 rays in the first dorsal, but this is probably an error. SOUTH AFRICA 51 Lepidion eques, Goode and Bean, 1895, Ocean. Ichth., p. 371; Collett, 1905, Rep. Norweg. Fish. Mar. -Invest., 11 (3), p. 69; Koefoed, 1926, Rep. Set. Res. 'Michael Sars' N. Atlant. Exped. 1910, IV (i), ZooL, p. 124, fig. 50. Haloporphyrtis lepidion van eques, Roule, 19 19, Res. Camp. Sci. Monaco, Lii, p. 78. Hab. Eastern Atlantic. In the British Museum several specimens, up to 350 mm. in total length, including the types of the species. Very closely related to, or perhaps identical w^ith L. lepidion. Lepidion inosimae (Giinther). Haloporphyrus inosimae, Giinther, 1887, t.c, p. 92, pi. xx, fig. B. Hab. Inosima, Japan. In the British Museum 4 specimens, 212-305 mm. in total length — types of the species. Lepidion capensis, Gilchrist. Lepidion capensis, Gilchrist, 1922, Rep. Fish. Mar. Biol. Surv. S. Afric, 11 (1921), Spec. Rep. in, p. 61 ; Barnard, 1925, Ann. S. Afr. Mtis., xxi, p. 324, pi. xiii, fig. i. Hab. South Africa. In the British Museum 7 specimens, 310-500 mm. in total length. Lepidion ensiferus (Giinther). Haloporphyrus ensiferus, Giinther, 1887, t.c, p. 92, pi. xix, fig. A. Hab. Off the mouth of the Rio Plata. In the British Museum 4 specimens, 265-350 mm. in total length — types of the species. Lepidion modestus (Franz). Haloporphyrus modestus, Franz, 1910, Abh. K. Bayer. Akad. Wiss. Miinch., Suppl. iv, Abh. i, p. 28, pi. iv, fig. 13. Hab. Yokohama, Japan. Only the type known, 340 mm. in total length. TRACHICHTHYIDAE Hoplostethus mediterraneus, Cuv. and Val. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 362. St. J. I specimen, 165 mm. Hoplostethus atlanticus, Collett. Collett, 1889, Bull. Soc. zool. Fr., xiv, p. 306; Goode and Bean, 1895, Ocean. Ichth., p. 189. St. J. 9 specimens, 210-430 mm. This species is readily distinguished from the preceding by the relatively smaller eye, smaller scales, indistinct abdominal scutes, higher number of dorsal rays, etc. It was not included by Barnard in his South African monograph. 7-2 DISCOVERY REPORTS ZEIDAE Zeus capensis, Cuv. and Val. Barnard, 1925, Ann. S. Afr. Mus., xxi, p. 373, pl- xvi, fig. 3- St. A. 2 specimens, 190, 210 mm. Pseudocyttus maculatus, Gilchrist. Barnard, t.c, p. 376. St. J. 7 specimens, 210-445 mm. Neocyttus rhomboidalis, Gilchrist. Barnard, t.c, p. 377. St. J. 3 specimens, 100-105 mm. AUocyttus verrucosus Gilchrist. Barnard, t.c, p. 378, pl. xvi, fig. 4. St. J. 19 specimens, 100-395 mm. CARANGIDAE Trachurus trachurus (Linn.). Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 531, pl. xxiii, fig. i. I specimen, 380 mm., presented by Messrs Irvine and Johnstone. BRAMIDAE Brama raii (Bloch.). Barnard, 1927, Ann. S. Afr. Mm., xxi, p. 594, pl- xxiv, fig. 3. St. D. I specimen, 520 mm. St. G. 2 specimens, 500, 570 mm. St. ?. I specimen, 610 mm. SCIAENIDAE Umbrina capensis, Pappe. Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 578, pl. xxiii, fig. 4- I specimen, 360 mm., presented by Messrs Irvine and Johnstone. SPARIDAE Dentex rupestris, Cuv. and Val. Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 714. St. 90. II. vii. 26. Large gauze fish-trap, 10 m.: i specimen, 68 mm. Dentex argyrozona, Cuv. and Val. Barnard, t.c, p. 717. I specimen, 340 mm., presented by Messrs Irvine and Johnstone. Pachymetopon blochi (Cuv. and Val.). Norman, 1935, Ann. S. Afr. Mus., xxxii, p. 12, fig. 3. St. 90. 11-12. vii. 26. Hand line and large gauze fish-trap, 10 m.: 2 specimens, 85, 200 mm. SOUTH AFRICA 53 Sparus globiceps (Cuv. and Val.). Barnard, t.c, p. 685. I specimen, 350 mm., presented by Messrs Irvine and Johnstone. Diplodus rondeleti (Cuv. and Val.), var. capensis, Smith. Barnard, t.c, p. 691. St. 90. 10. vii. 26. Hand line, 10 m.: i specimen, 300 mm. Pagrus laniarius, Cuv. and Val. Barnard, t.c, p. 694, fig. 24. I specimen, 390 mm., presented by Messrs Irvine and Johnstone. CLINIDAE Clihus taurus, Gilchrist and Thompson. Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 858. 29. vi. 27. Simon's Town, i specimen, 122 mm. Found while draining the dry dock. BROTULIDAE Bidenichthys capensis, Barnard. 1934, Ajin. Mag. Nat. Hist. (10) xni, p. 234, fig. 3. St. 90. II. vii. 26. Hand net, 1-2 m.: i specimen, 45 mm. This interesting little fish is new to the British Museum collection. OPHIDIIDAE Genypterus capensis (Smith). Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 887, pi. xxxv, fig. 5. St. D. 4 specimens, 440-475 mm. St. P. 3 specimens, 700-770 mm. This species is closely related to the Australian and New Zealand G. blacodes (Schn.), but may be recognized by the smaller eye. SCORPAENIDAE Helicolenus maculatus (Cuv. and Val.). Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 907. St. C. 4 specimens, 304-360 mm. St. ?. I specimen, 350 mm. Scorpaena (??) capensis, Gilchrist and von Bonde.^ Sebastosemus capensis, Barnard, t.c, p. 910. St. J. 2 specimens, 370, 400 mm. 1 See note on p. 32 of this report. 54 DISCOVERY REPORTS TRIGLIDAE Chelidonichthys capensis (Cuv. and Val.). Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 940, pi. xxxiv, fig. 3, fig. 28 c. St. A. 2 specimens, 405, 435 mm. COTTUNCULIDAE Cottunculoides inermis (Vaillant). Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 923, pi. xxxiv, fig. i. St. J. I specimen, 267 mm. TETRODONTIDAE Tetrodon honckeni, Bloch. Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 970, pi. xxxvi, fig. 6. 12. vii. 27. Simon's Town Dockyard. 2 specimens, 160, 195 mm. LOPHIIDAE Lophius piscatorius, Linn. Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 999. St. A. I specimen, 380 mm. St. B. I specimen, 150 mm. ASCENSION ISLAND 28 specimens were collected at this locality, representing 12 species. SERRANIDAE Paranthias furcifer (Cuv. and Val.). St. I. 16. xi. 25. Medium rectangular net, 16-27 m- ^ specimen, 72 mm. CARANGIDAE Trachurops crumenophthalmus (Bloch). St. 2. 17. xi. 25. Shore collection — rock pools: 2 specimens, 200, 225 mm. POMACENTRIDAE Glyphisodon saxatilis (Linn.). St. 2. 17. xi. 25. Shore collection — rock pools: 7 specimens, 22-165 mm. Pomacentrus leucostictus, Miill. and Trosch. St. I. 16. xi. 25. Medium rectangular net, 16-27 ^■'- 2 specimens, 60, 69 mm. LABRIDAE Thalassoma ascensionis CQuoy and Gaim.). St. 2. 17. xi. 2^. Shore collection — rock pools: i specimen, 105 mm. ASCENSION ISLAND SS BLENNIIDAE Rupiscartes atlanticus (Cuv. and Val.). St. 2. 17. xi. 25. Shore collection — rock pools: i specimen, 140 mm. Salariichthys textilis (Quoy and Gaim.). St. 2. 17. xi. 25. Shore collection — rock pools : 6 specimens, 23-60 mm. MUGILIDAE Myxus (?) curvidens (Cuv. and Val.). St. 2. 17. xi. 25. Shore collection — rock pools: 2 specimens, 47, 100 mm. SCORPAENIDAE Scorpaena scrofina, Cuv. and Val. St. I. 16. xi. 25. Medium rectangular net, 16-27 ^■'- 2 specimens, 44, 48 mm. BALISTIDAE Balistes vetula, Linn. St. I. 16. xi. 25. Medium rectangular net, 16-27 ^■' i specimen, 70 mm. Melichthys piceus (Poey). St. 2. 17. xi. 25. Shore collection — rock pools: i specimen, 175 mm. ANTENNARIIDAE Antennarius multiocellatus (Cuv. and Val.). St. I. 16. xi. 25. Medium rectangular net, 16-27 ^^•'- ^ specimen, 9 mm. Another specimen, 70 mm. long, was presented by Mr L. W. Shaw. In comparison with the fishes of St Helena, which have been dealt with by Giinther, Melliss, Cunningham, and Clark, ^ among others, those of Ascension have been some- what neglected. The island was visited by Osbeck,- who listed 9 species, several after- wards utilized by Linnaeus. More than a hundred years later the ' Challenger ' made a small collection here, which was reported upon by Giinther,^ who also recorded 3 further species in the following year.* Nichols and Murphy ^ published a note on Balistes vetula from Ascension, and, finally, in 1919 Fowler^ recorded about a dozen species, and also described a new species of Abudefdiif (= Glyphisodon), which may also occur at St Helena. 1 For references to these papers see, Fowler, 1919, Proc. U.S. Nat. Mus., LVI, p. 217. 2 1765, Reise Ost.-Ind. China, pp. 385-96. 3 1880, Shore Fish.'ChaHenger', p. 5. * 1881, Ann. Mag. Nat. Hist. (5) viii, pp. 430-40. ^ 1917, Copeia, No. 39, p. 2. ^ T.C., pp. 217-27. S6 DISCOVERY REPORTS St West West South Species Helena Indies Brazil Africa! Africa Carcharinus obscurus (Le Sueur) X X X X Atlantic Lycodontis moringa (Cuv.) X X X . . Tylosurus caribbaeus (Le Sueur) . X . Belone ardeola, Cuv. & Val. X X , X . Exocoetus ba/iiensis, Ranzani X X X X Atlantic Aulostomus maculatiis, Val. X X X . My ripristis jacobus, Cuv. & Val. X X X . Holocentrutn adscerisionis (Osbeck) X X X X , Epinephelus aeneus (Geoffr.) . . . X . Mediterranean Epinephehis adscemionis (Osbeck) X X X X X Paranthias furcifer (Cuv. & Val.) , X X Pacific coast, trop. America Rypticus saponaceus (Schn.) X X X X X Priacanthus cruentatiis (Lacep.) X X . X X Pacific Apogott axillaris, Val. X . , , Malacanthus plumieri (Bloch) , X X , , Decapterus sanctae helenae (Cuv. & Val.) X X X X Caranx lugubris, Poey X X X X? Atlantic and Pacific Caranx hippos (Linn.) X X X X X Atlantic and Pacific Trachurops crumenophthalmus (Bloch) . X X X X Atlantic and Pacific Trachinotus glaucus (Linn.) X X? X X Mediterranean Upeneiis martinicus, Cuv. & Val. . X X Diplodiis argenteus (Cuv. & Val.) . X X Chaetodon sanctae-heletiae, Giinth. X . Pomacanthus pan/ (Bloch) , X X Glyphisodon saxatilis (Linn.) X X X X Pacific coast, trop. America Glyphisodon ascensionis (Fowler) X? . Chromis marginatus (Cast.) X X Pomacetitrus leiicostictus, Miill. & Tr. X X X X.? Thalassoma ascensionis (Quoy & Gaim.) X . Harpe nifa (Linn.) X X X Pseiidoscarus giiacamaia (Cuv.) . X X X Teuthis hepatus (Linn.) X X X X Blennius cristaius, Linn. , X X X > < Rupiscartes atlanticus (Cuv. & Val.) X X X X Pacific coast, trop. America Alliens textilis (Cuv. & Val.) X X X Ophioblennius wehbi (Val.) , X X X Mugil cephalus, Linn. X X X X Mediterranean ; Pacific coast of America Myxus curvidens (Cuv. & Val.) X X Echeneis naucrates, Linn. X? X X X X All warm seas Scorpaena plumieri, Bloch , X X . Scorpaena scrofina, Cuv. & Val. X , , , Dactylopterus volitans (Linn.) , X X X X Atlantic Bo thus mellissi, Norman X , . Alutera scripta (Osbeck) X X X Pacific (?) Canthidermis niaculatus (Bloch) , X X X Pacific (?) Balistes vetula, Linn. X X , X Melichthys piceus (Poey) X X X , . Lactophrys tricornis (Linn.) X X X X . Antennarius multiocellatus (Cuv. & Val.) X X • Including Madeira, Canaries, Cape Verde Islands, etc. ASCENSION ISLAND 57 In addition to the earlier collections made by H.M.S. 'Challenger' (1873-76), Mr T. Conry (1881), and Dr A. McCloy (1908), the British Museum has received two larger and valuable lots from Ascension in recent years, one presented in 1927 by Dr J. J. Simpson of the Liverpool Public Museum, the other presented in 1932 by Colonel S. T. Haley. With this material available, it seems desirable to draw up a provisional list of the fishes recorded from the island, and to indicate in the form of a table the distribution of the various species. A glance at this table shows at once that the fauna, like that of St Helena, is predominantly West Indian and Brazilian in char- acter. Of the 49 species recorded, 27 or 28 occur also at St Helena, and doubtless others will be found to be common to the two islands. A certain number of the species also appear to occur on the coast of West Africa, but many of the records from this region are unreliable. TRISTAN DA CUNHA Examples of 3 species were obtained here, including a fine specimen of a new species of Decapterus, which has been described and figured by me elsewhere. CARANGIDAE Decapterus longimanus, Norman. 1935, Ann. Mag. Nat. Hist. (10) xvi, p. 255, fig. i. St. 4. 30-31. i. 26. Hand line, 40 m.: i specimen (holotype), 470 mm. CHILODACTYLIDAE Acantholatris monodactylus (Carmichael). Chaetodon monodactylus, Carmichael, 1818, Trans. Linn. Sac, xn, p. 500, pi. xxiv. Chilodactylus carmichaelis, Cuvier and Valenciennes, 1830, Hist. Nat. Poiss., v, p. 360; Kner, 1869, Reise 'Novara', Zool. i, 5. Fische, p. 90, pi. v, fig. i. Chilodactylus monodactylus, Regan, 1913, Ann. Mag. Nat. Hist. (8) xi, p. 466. This species was redescribed and figured by Kner from St Paul Island, in the same latitude as Tristan da Cunha but 4500 miles distant. There are in the British Museum collection two small specimens from Tristan da Cunha presented by the South African Museum, and another larger one from the same locality collected by the Shackelton- Rowett Expedition ('Quest'). This species is quite distinct from the Chilean Acantho- latris gayi (Kner), of which the 'Challenger' obtained two fine examples from Juan Fernandez. 1 SCORPAENIDAE Sebastichthys capensis (Gmelin). Barnard, 1927, Ann. S. Afr. Mus., xxi, p. 908. St. 4. 30-31. i. 26. Hand line, 40 m.: i specimen, 340 mm. Also obtained from Gough Island ('Scotia' and 'Quest'). ^ The Chilean form is well described by Cuvier and Valenciennes (1833, H.N. Poiss., ix, p. 489) as Cheilodactylus carmichaelis, and it was subsequently figured by Valenciennes (1850, in Cuvier, R. Anim., Discip. Ed., Poiss. pi. xx.xi, fig. 2). 58 DISCOVERY REPORTS GOUGH ISLAND Examples of only 2 species were obtained from this locality. CHILODACTYLIDAE Acantholatris monodactylus (Carmichael). St. WS 123. 8. vi. 27. Hand line, 47-72 m. : i specimen (head only), no mm. BOVICHTHYIDAE Bovichtus diacanthus (Carmichael). Callionymiis diacanthus, Carmichael, 1818, Trans. Linn. Soc, xii, p. 501, pi. xxvi. Bovichthys diacanthus, Giinther, i860. Cat. Fish., 11, p. 249; Regan, 1913, Ann. Mag. Nat. Hist., (8) XI, p. 467; Regan, 1913, Trans. R. Soc. Edinh., xlix, pp. 239, 256, pi. ix, fig. 5. St. WS 123. 9. vi. 27. Shore collection: 3 specimens, 200-245 mm. Originally described from Tristan da Cunha, this species was taken by the ' Scotia ' at Gough Island. Regan has shown that it is distinct from the Chilean form, B. chilensis, Regan. It is represented at the Island of St Paul by a closely related species, B. veneris, Sauvage. APPENDIX A certain number of flying fishes were obtained by the Expedition, and the specimens are being studied by Dr Anton F. Bruun of the Marinbiologisk Laboratorium, Copen- hagen, who has undertaken a general revision of the Exocoetidae of the Atlantic. He has kindly furnished me with the following identifications of specimens from areas covered by the present report. EXOCOETIDAE Oxyporhamphus micropterus (Cuv. and Val.). St. 289. 23. viii. 27. 3° 04' 45" N, 16° 52' W. Hand net: 2 specimens, 182, 189 mm. St. 294. 25. viii. 27. 4° 33' 15" N, 16° 52' 45" W. Hand net: i specimen, ca. 163 mm. Exocoetus volitans, Linn. II. xi. 25. 1° 04' S, 12° 50' W. Flew on board: i specimen, 204 mm. 17. iv. 33. 3° 21' S, 8° 37' W. On deck: i specimen, ca. 176 mm. Exocoetus obtusirostris, Giinth. 26. X. 25. 16° 19' N, 18° 24' W. Flew on board: i specimen, ca. 176 mm. 25. X. 25. 17° 30' N, 18° 16' W. Flew on board: i specimen, 212 mm. Cypsilurus cyanopterus (Cuv. and Val.). 6. V. 32. 19° 06' S, 38° 39' W. On deck: i specimen, ca. 167 mm. Cypsilurus lineatus (Cuv. and Val.). 24. X. 25. 21° 00' N, 18° 05' W. Flew on board: i specimen, ca. 427 mm. [Discovery Reports. Vol. XII, pp. 59-198 January, 1936] POLYCHAETE WORMS. II By C. C. A. MONRO, M.A. Assistant Keeper in the Department of Zoology, British Museum CONTENTS Introduction P^S^ ^' Station list ^^ List of species ^9 Geographical distribution 7^ Benthic species 7^ Pelagic species 7" Systematic account 7^ Key to the families 7^ Amphinomidae 79 Aphroditidae "i Polynoidae °4 Sigalionidae i°3 Phyllodocidae i09 Alciopidae i^S Tomopteridae i20 Typhloscolecidae I22 Syllidae 124 Nereidae ^34 Nephthydidae i39 Glyceridae i4' Sphaerodoridae I44 Eunicidae ^45 Eunicinae i45 Onuphidinae i4" Lumbrinereinae iS3 Ariciidae ^59 Cirratulidae 161 Spionidae 161 Chaetopteridae 162 Chlorhaemidae 102 Opheliidae 164 Maldanldae 167 Sabellariidae 17° Amphictenidae 172 Ampharetidae i73 Terebellidae 176 Amphitritinae i77 Thelepinae i°t Polycirrinae 103 Trichobranchinae loS Sabellidae 187 Serpulidae 19^ Incertae sedis i93 References i95 POLYCHAETE WORMS. II By C. C. A. Monro, m.a. Assistant Keeper in the Department of Zoology, British Museum (Text-figures 1-34) INTRODUCTION THE Polychaeta studied in this report were collected by the staff of the Discovery Committee in the R.R.S. 'Discovery' and 'Discovery II', in the 'William Scoresby', and at the Marine Biological Station, South Georgia. There follows a list of the stations. Those made by the ' Discovery' and the ' Discovery II ' are given first and have no letters prefixed to their numbers; those of the 'William Scoresby' follow, and have the prefix WS, and lastly those of the Marine Biological Station are preceded by MS. The symbols in the list of stations representing the various kinds of gear used are explained both in my earlier report (1930) on the Discovery Polychaeta and in the Station Lists published in this series of Reports. STATION LIST R.R.S. 'DISCOVERY' St. 17. 4. iii. 26. 46 miles N 46° E of Jason Light, South Georgia. 500-250 m. Gear N 70 V. St. 27. 15. iii. 26. West Cumberland Bay, South Georgia; 3-3 miles S 44° E of Jason Light, iiom. Gear DL. Bottom: mud and rock. St. 28. 16. iii. 26. West Cumberland Bay, South Georgia; 3-3 miles S 45° W of Jason Light. 168 m. Gear DC. Bottom: mud. St. 30. 16. iii. 26. West Cumberland Bay, South Georgia; 2-8 miles S 24° W of Jason Light. 251 m. Gear DLH. Bottom: mud and stones. St. 39. 25. iii. 26. East Cumberland Bay, South Georgia. From 8 cables S 81° W of Merton Rock to 1-3 miles N 7° E of Macmahon Rock. 179-235 m. Gear N 4-T. Bottom: grey mud. St. 41. 28. iii. 26. i6i miles N 39° E of BantT Point, South Georgia. Gear N 70 V. St. 41 A. 265-150 m. ; St. 41 B. 265-150 m.; St. 41 d. 240-150 m. ; St. 41 e. 150-100 m. St. 42. I . iv. 26. Off mouth of Cumberland Bay, South Georgia. From 6-3 miles N 89° E of Jason Light to 4 miles N 39° E of Jason Light. 120-204 r"- Gear OTL. Bottom: mud. St. 45. 6. iv. 26. 27 miles S 85° E of Jason Light, South Georgia. 238-270 m. Gear OTL. Bottom: grey mud. St. 53. 12. V. 26. Port Stanley, East Falkland Island. Hulk of 'Great Britain'. 0-2 m. GearRM. St. 100. 2. X. 26. 33° 20' to 33° 46' S, 15° 18' to 15° 08' E. 260-310 m. Gear TYF. St. 114. 12. xi. 26. 52° 25' S, 9° 50' E. 0-5 m. Gear N 100 H. St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia. From 4-1 miles N 54° E of Larsen Point to 1-2 miles S 62° W of Larsen Point. 230-250 m. Gear OTL, NCS-T. Bottom: grey mud. 62 DISCOVERY REPORTS St. 124. 18. xii. 26. 53° 45' 30" S, 36° 32' 30" W. 0-5 m. Gear N 100 H. St. 128. 19. xii. 26. 53° 38' 30" S, 37° 08' W. loom. Gear N 100 H. St. 130. 20. xii. 26. 54° 06' S, 36° 23' W. 77 m. Gear N lOO H. St. 133. 20-21. xii. 26. 53° 45' 30" S, 35° 46' 30" W. o-s m., 50 m., 100 m. Gear N 100 H. St. 136. 21. xii. 26. 54° 22' S, 35° 21' W. 0-51x1. Gear N 100 H. St. 137. 22. xii. 26. 54° 19' 30" S, 35° 03' 30" W. 132 m. Gear N 100 H. St. 138. 22. xii. 26. 54° 17' S, 34° 47' W. 77 m., 155 m. Gear N 100 H. St. 139. 22-23. xii. 26. 53° 30' 15" S, 35° 50' 45" W. 0-5 m. Gear N 100 H. St. 142. 30. xii. 26. East Cumberland Bay, South Georgia. From 54" n' 30" S, 36° 35' W to 54° 12' S, 36° 29' 30" W. 88-273 m. Gear NCS-T. Bottom: mud. St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia. From 54° 04' S, 36° 27' W to 53° 58' S, 36° 26' W. 155-178 m. Gear NCS-T. Bottom: green mud and sand. St. 149. 10. i. 27. Mouth of East Cumberland Bay, South Georgia. From 1-15 miles N 76^° W to 2-62 miles S 11° W of Merton Rock. 200-234 m. Gear NCS-T. Bottom: mud. St. 151. 16.1.27. 53° 25' S, 35° 15' W. 1000-750 m. Gear N 70 V. St. 156. 20.1.27. 53° 51' S, 36° 21' 30" W. 200-236 m. Gear DLH. Bottom: rock. St. 160. 7. ii. 27. Near Shag Rocks, South Georgia, 53° 43' 40" S, 40° 57' W. 100-50 m., gear N 70 V; 177 m., gear DLH. Bottom: grey mud, stones and rock. St. 164. 18. 11. 27. East end of Normanna Strait, South Orkneys near Cape Hansen, Coronation Island. 24-36 m. GearBTS, NCS-T. St. 167. 20. 11. 27. Off Signy Island, South Orkneys, 60° 50' 30" S, 46°i5'W. 244-344 m. Gear N 7-T. Bottom: green mud. St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 61° 25' 30" S, 53° 46' W. 342 m. Gear DLH. Bottom: rock. St. 175. 2.111.27. Bransfield Strait, South Shetlands, 63° 17' 20" S, 59° 48' 15" W. 200 m. Gear DLH. Bottom: mud, stones and gravel. St. 182. 14. iii. 27. SchoUaert Channel, Palmer Archipelago, 64° 21' S, 62° 58' W. 278-500 m. Gear NCS-T. Bottom: mud. St. 190. 24.111.27. Bismarck Strait, Palmer Archipelago, 64° 56' S, 65° 35' W. 93-126 m. Gear DLH. Bottom: stones, mud and rock. St. 268. 25. vii. 27. 18° 37' S, 10° 46' E. 73-0 m. Gear N 100 B. St. 274. 4. viii.27. Off St Paul de Loanda, Angola. From 8°4o'i5"S, i3°i3'45"E to 8°38'i5"S, 13° 13' E. 64-65 m. Gear OTL. Bottom: grey mud. R.R.S. 'DISCOVERY IT St. 334. 4.11.30. 53° 43' S, 36° 51' W. iio-om. Gear N 70 B. St. 362. 25. ii. 30. 56° 04' S, 29° 15' W to 56° 03!' S, 29° 20' W. 97-0 m. Gear N 100 B. St. 363. 26. 11. 30. 2-5 miles S 80° E of SE point of Zavodovskl Island, South Sandwich Islands. 329-278 m. Gear DLH. St. 366. 6. ill. 30. 4 cables S of Cook Island, South Sandwich Islands. 322-155 m. Gear DLH. Bottom: black sand. St. 368. 8. 111. 30. Douglas Strait, Southern Thule, South Sandwich Islands, i mile N of Twitcher Rock. 653 m. Gear DLH. Bottom: black mud. St. 371. 14. ill. 30. I mile E of Montagu Island, South Sandwich Islands. 99-161 m. Gear OTL, N7-T, N4-T. STATION LIST 63 Gear St. 373. 19. iii. 30. 58° 00' S, 33° 44' W. 275-0 m. Gear TYFB. St. 374. 20. iii. 30. 57° 55' S, 37° 30' W. 270-0 m. Gear TYFB. St. 395. 13.V. 30. 48°26f'S, 22° 10' W to 48°26i'S, 22° 06 J' W. 1500-160001. N450 H. St. 399. 18. V. 30. I mile SE of SW point of Gough Island. 141-102 m. Gear DLH. St. 404. 24. V. 30. 35° 34' S, i5°ooJ' E. loi-om. Gear N 100 B. St. 405. 4. vi. 30. 33° so}/ S, 15° 46' E to 34° 16' S, 15° 02' E. 1200-0 m. Gear TYFB. St. 407. 12. vi. 30. 35° 13' S, 17° s°¥ E to 34° 57' S, 17° 48' E. 220-0 m. Gear TYFB. St. 413. 2i.viii. 30. 33° 13'S, i5°46i'E. 550-350 m., 350-0 m. Gear TYFB. St. 419. 30. viii. 30. 36° 29' S, 18" i6i' E to 36° 29' S, 18° 15^' E. 84-0 m. Gear N 100 B. St. 446. 9. X. 30. 36° 14' S, 16° 09I' E. 106-0 m. Gear N 100 B. St. 448. 10. X. 30. 39° 03' S, 16° iif E. i6i-om. Gear N 100 B. St. 449. 11-12. X. 30. 42°3o|' S, 15° 14J' E. 150-0 m. Gear N 100 B. St. 450. 12-13. X. 30. 44° 57I' S, 12° 57i' E to 44° 56i' S, 12° 54' E. 150-0 m. Gear N 100 B. St. 451. 13-14. X. 30. 47° 19I' S, 11° 05' E. 170-0 m. Gear N 100 B. St. 453. 16-17. X. 30. 54° 05I' S, 3° 57J' E to 54° 07' S, 04^ 03' E. 165-0 m. Gear N 100 B. St. 454. 17. X. 30. 53° 42' S, 4° 42' E. 192-0 m. Gear N 100 B. St. 455. 18. X. 30. 53''55i'S, 4°47'E. ii6-om. Gear N 100 B. St. 456. 18. X. 30. I mile E of Bouvet Island. 40-45 m. Gear DLH. St. 458. 19. X. 30. 7 miles S 50° W of Cape Circumcision, Bouvet Island. 357-377 m. Gear DLH. St. 459. 19. X. 30. 55° 09J' S, 2° 00' E. 183-0 m. Gear N 100 B. St. 460. 20-21. X. 30. 56° 46' S, o°4if' W. 155-0 m. Gear N 100 B. St. 461 c. 21-22. X. 30. 56° 44' S, 2° 22' W. (No particulars of depth or gear given on label with specimens.) St. 474. 12. xi. 30. I mile W of Shag Rocks, South Georgia. 199 m. Gear DLH. St. 514. 26. xi. 30. 55° 51' S, 35° 32' W. 155-0 m. Gear N 100 B. St. 527. ii.xii. 30. 54° 09^' S, 34° 29V W. i22(-o)m. Gear N 450 H. St. 533. 16. xii. 30. 59° 36' S, 42° 34' W. 165-0 m. Gear N 100 B. St. 567. 3.1.31. 66° 45' S, 89° 24' W. 140-0 m. Gear N 100 B. St. 569. 4.1.31. 68°4oi'S, 96°2i'W. 137-0 m. Gear N 100 B. St. 575. 8.1.31. 67°53i'S, 9i°23'W. 97-0 m. Gear N 100 B. St. 579. 10. i. 31. 66° 41I' S, 79° 10' W. 180-0 m. Gear N 100 B. St. 588. 13.1.31. 66° iii'S, 7i°5oi' W. 460-150 m. Gear N 100 B. St. 590. 14.1.31. 65° 2oi' S, 73° 30J-' W. ii50-i40om. Gear TYFH. St. 591. 14.1.31. 64° 51^' S, 74° 22|' W. 122-0 m. Gear N 100 B. St. 599. 17.1.31. 67° 08' S, 69° 06^' W. 203 m. Gear DLH. St. 600. 17.1.31. 67° 09' S, 69° 27' W. 501-527 m. Gear OTL. St. 619. 19. ii. 31. 59° 33' S, 43° 07J' W. 1 14-0 m. Gear N 70 B. St. 652. 14. iii. 31. Burdwood Bank, 54° 04' S, 61° 40' W. 171-169 m. Gear OTL. St. 701. 16. X. 31. 14° 39-3' N, 25° 517' W. 242-0 m. Gear TYFB. St. 702. 17. X, 31. 10° 59-3' N, 27° 03-8' W. 236-0 m. Gear TYFB. St. 704. 19. X. 31. 3° 377' N, 29° 14' W. 231-0 m. Gear TYFB. St. 705. 20. X. 31. 0° 03-4' N, 30° 36-8' W. 150-0 m. Gear TYFB. 64 DISCOVERY REPORTS St. 707. 22. X. 31. 6° 44' S, ss'^ 33' W. 182-0 m. Gear TYFB. St. 708. 23.x. 31. 10° 20-6' S, 34° 547' W. 208-0 m. Gear TYFB. St. 709. 24. X. 31. 14° 01-4' S, 36° 307' W. 216-0 m. Gear TYFB. St. 710. 26. X. 31. 21° 45' S, 39° 50' W. 294-0 m. Gear TYFB. St. 713. 29. X. 31. 31° 37-1' S, 45° 00' W. 200-0 m. Gear TYFB. St. 714. 30. X. 31. 35° 09-5' S, 47° 00' W. 246-0 m. Gear TYFB. St. 716. i.xi. 31. 42" o8-8'S, 51^35' W. 212-0 m. Gear TYFB. St. 718. 3. xi. 31. 47° 27-2' S, 55° 10-2' W. 262-0 m. Gear TYFB. St. 724. 16. xi. 31. Fortescue Bay, Magellan Strait. 0-5111. Gear NS. St. 929. 16. viii. 32. 34° 2i' S, 172° 48' E to 34° 22-2' S, 172° 49-8' E. 58-55 m. Gear OTL. St. 934. 17. viii. 32. 34° II-6' S, 172° 10-9' E to 34° 11-4' S, 172° 10-3' E. 98-0 m. Gear DRL. St. 935. 17. viii. 32. 34° 11-5' S, 172° 08-5' E to 34° 11-9' S, 172° 08-5' E. 84-0 m. Gear DRL. St. 936. 18. viii. 32. 35° 03-5' S, 172° 58-2' E to 35° 05-4' S, 172° 587' E. 50-0 m., gear DC; 50-57111., gear N4-T. St. 937. 18. viii. 32. 35° 187' S, 173° 08-2' E. 48-0 m. Gear DC. St. 938. 18. viii. 32. 35° 30-6' S, 173° 19' E. 37-0 m. Gear DC. St. 939. 18. viii. 32. 35° 49-6' S, 173° 27' Etc 35^ 51-6' S, 173° 28-9' E. 87-87111. GearN4-T, DC. St. 941. 20. viii. 32. 40° 51-4' S, 174" 48-2' E to 40° 55-8' S, 174° 467' E. 128-0 m. Gear DRL. St. 1 148. 9. iii. 33. 63° 52' S, 0° 24-9' W. 0-5 m. Gear N 100 H. R.R.S. 'WILLIAM SCORESBY' St. WS 4. 30. ix. 26. 32° 45' S, 18° 10' E. 45-47 m. Gear DL. St. WS 20. 28. xi. 26. 53° 52' 30" S, 36° 00' W. 190 m. Gear N 100 H. St. WS22. 30. xi. 26. 53° 38' S, 35° 35' W. 82 m. Gear N 100 H. St. WS25. 17. xii. 26. Undine Harbour (North), South Georgia. 18-27 m. GearBTS. Bottom: mud and sand. St. WS26. 18. xii. 26. 53° 33' 15" S, 37° 45' 15" W. 96 m., 192 m. Gear N 100 H. St. WS27. 19. xii. 26. 53° 55' S, 38° 01' W. 107 m. Gear N 100 H. St. WS33. 21. xii. 26. 54° 59' S, 35° 24' W. 130 m. Gear N 100 H. St. WS35. 21-22. xii. 26. 55° 13' 15" S, 34° 59' W. 0-5 m. Gear N 100 H. St. WS38. 22-23. xii. 26. 54° 01' S, 35° 14' W. 106 (-53) m. Gear N 100 H. St. WS39. 23. xii. 26. 54° 08' S, 35° 43' W. 87 ?m. Gear N 100 H. St. WS 40. 7. i. 27. 55° 09' S, 35° 58' W. 175-100 m. Gear N 70 V. St. WS 44. 8. i. 27. 55° 06' S, 36° 57' W. 750-500 m., 1000-750 m. Gear N 70 V. St. WS45. 8.1.27. 54° 38' 30" S, 37° 30' 55" W. 0-5 m. Gear N 100 H. St. WS 53. 11-12. 1. 27. From 54° 03' 30" S, 38° 35' W to 53^ 29' S, 37° 13' 45" W. 0-5 m. Gear N 100 H. St. WS54. 12.1.27. 53° 29' S, 37° 13' 45" W. 500-250 m. Gear N 70 V. St. WS55. 12.1.27. 53° 15' 30" S, 37° 13' 45" W. 164 m. Gear N 100 H. St. WS 79. 13. iii. 27. 51° 01' 30" S, 64° 59' 30" W. From 51° 00' S, 65° 00' W to 51° 03' S, 64° 59' W. 132-131 m. Gear N 7-T. Bottom: fine dark sand. St. WS 83. 24. ill. 27. 14 miles S 64° W of George Island, East Falkland Island. From 52° 28' S, 60° 06' W to 52° 30' S, 60° 09' 30" W. 137-129 m. GearOTC. Bottom : fine green sand and shell. STATION LIST 65 St. WS 84. 24. iii. 27. 7I miles S 9° W of Sea Lion Island, East Falkland Islands. From 52° 33' S, 59° 08' W to 52° 34' 30" S, 59° 11' W. 75-74 m. Gear OTC. Bottom: coarse sand, shell and stones. St. WS 85. 25. iii. 27. S miles S 66° E of Lively Island, East Falkland Islands. From 52° 09' S, 58° 14' W to 52° 08' S, 58° 09' W. 79 m. Gear OTC, N 7-T, N 4-T. Bottom: sand and shell. St. WS 90. 7. iv. 27. 13 miles N 83° E of Cape Virgins Light, Argentine Republic. From 52° 18' S, 68° 00' W to 52° 19' 30" S, 67° 57' W. 82-81 m. Gear OTC. Bottom: fine dark sand. St. WS 177. 7. iii. 28. 54° 58' S, 35° 00' W. 97-0 m. Gear N 100 B. St. WS200. 2i.iv. 28. 59° 05' S, 46° 32' W. 93-0 m. Gear N 100 B. St.WS2ii. 29.V.28. 50° 17' S, 60° 06' W. i6i-i74m. Gear N 4-T, NCS-T. Bottom: green sand. St. WS 212. 30. V. 28. 49° 22' S, 60 ' 10' W. 242-249 m. Gear N 4-T, NCS-T. Bottom: green sand, mud and pebbles. St. WS213. 30. V. 28. 49° 22' S, 60° 10' W. 249-239 m. Gear NCS-T. Bottom: green sand, mud and pebbles. St. WS214. 31.V. 28. 48° 25' S, 60° 40' W. 208-219 m. Gear NCS-T, DC. Bottom: fine dark sand. St. WS215. 31.V. 28. 47° 37' S, 60° 50' W. 219-146 m. Gear DC, NCS-T. Bottom: fine green sand. St. WS 216. i.vi. 28. 47° 37' S, 60° 50' W. 219-133 m. Gear N 7-T. Bottom: fine sand. St.WS2i9. 3.vi.28. 47° 06' S, 62° 12' W. ii6-ii4m. Gear NCS-T. Bottom: dark sand. St. WS 220. 3.vi. 28. 47° 56' S, 62° 38' W. 108-104 m. Gear NCS-T. Bottom: brown sand. St. WS221. 4. vi. 28. 48° 23' S, 65° 10' W. 76-91 m. Gear OTC. Bottom : brown sand, mud, pebbles, large stones and shell. St. WS 223. 8. vi. 28. 49° 13' S, 64° 52' W. ii4-ii4m. Gear OTC. Bottom: coarse brown sand and shell. St. WS 225. 9. vi. 28. 50° 20' S, 62° 30' W. 162-161 m. Gear OTC. Bottom: green sand, shell and pebbles. St. WS 226. 10. vi. 28. 49° 20' S, 62° 30' W. 144-152 m. Gear NCS-T. Bottom: green sand. St. WS228. 30. vi. 28. 50° 50' S, 56° 58' W. 229-236 m. Gear OTC, N 4-T, NCS-T. Bottom: shell and coarse white sand. St. WS 229. I. vii. 28. 50° 35' S, 57° 20' W. 210-271 m. Gear N 4-T. Bottom: fine green sand. St. WS231. 4. vii. 28. 50° ID'S, 58° 42' W. 167-159 m. Gear N 4-T, NCS-T. Bottom: fine green sand. St. WS234. 5- vii. 28. 48=52' S, 60° 25' W. 195-207 m. Gear OTC, NCS-T. Bottom: fine green sand. St. WS236. 6. vii. 28. 46° 55' S, 60° 40' W. 272-300 m. Gear DC, N 4-T, NCS-T. Bottom: dark green sand and mud. St. WS237. 7. vii. 28. 46° 00' S, 60° 05' W. 150-256 m. Gear N 4-T, NCS-T. Bottom: coarse brown sand and shell. St. WS 239. 15. vii. 28. 51° 10' S, 62° 10' W. 196-193 m. Gear OTC. Bottom: coarse dark sand. St. WS 243. 17. vii. 28. 51° 06' S, 64° 30' W. 144-141 m. Gear OTC, N 4-T. Bottom: coarse dark sand. St. WS 244. 18. vii. 28. 52° 00' S, 62° 40' W. 253-247 m. Gear N 7-T. Bottom : fine dark sand and mud. 66 DISCOVERY REPORTS St. WS 245. 18. vii. 28. 52° 36' S, 63° 40' W. 304-290111. GearN4-T. Bottom: dark green sand, madrepore, pebbles and shell. St. WS246. 19. vii. 28. 52° 25' S, 61° 00' W. 267-208 m. Gear OTC, N7-T, N4-T. Bottom: coarse green sand and pebbles. St. WS247. 19. vii. 28. 52° 40' S, 60° 05' W. 172 m. Gear DLH. Bottom: rock. St. WS248. 20. vii. 28. 52° 40' S, 58° 30' W. 210-242 m. Gear OTC. Bottom: fine green sand, pebbles and shell. St. WS 249. 20. vii. 28. 52° 10' S, 57° 30' W. 166 m. Gear DLH. Bottom: fine brown-green sand, shell and stones. St. WS351. ii.i. 29. 54° 21' 30" S, 34° 59' W. 750-500 m. Gear N 70 V. St. WS408. 26. ii. 29. 53° 50' S, 62° 10' W. ii2-om. Gear N 70 B. St. WS411. 14. iii. 29. 52° 08' S, 52° 35' W. loo-om. Gear N 70 B. St. WS536. 24.1.31. 56° 28' S, 27° 21' W. 102-0 m. Gear N 100 B. St. WS537. 25-26.1.31. 56° 10' S, 25° 35' W. 67-0 m. Gear N 70 B. St. WS541. 28.1.31. 57°5ii'S, i9°5i*' W. 102-0 m. Gear N 100 B. St. WS 544. 29. i. 31. 60° 59' S, 17° 50' W. 146-0 m. Gear N 100 B. St. WS545. 30.1.31. 61° 51' S, 17° 15' W. 124-0 m. Gear N 100 B. St. WS547. 30.1.31. 62° 40' S, 17° 02' W. 154-0 m. Gear N 100 B. St. WS548. 31.1.31. 64° 07' S, 15° 38' W. 106-0 m. Gear N 100 B. St. WS549. 31.1.31. 65° 17' S, 15° 33' W. 128-0 m. Gear N 100 B. St. WS550. I. 11. 31. 66°sii'S, i5°24' W. i2i-om. Gear N 70 B. St. WS551. i.ii. 31. 68° 17I' S, 14° 26V W. i2i-om. Gear N 70 B, N 100 B. St. WS 552. 2. ii. 31. 68° 53' S, 13° 03' W to 68° 50' S, 13° 03' W. (Depth not recorded on label.) Gear N 100 B. St. WS 555. 6. ii. 31. 60° 27' S, 19° 36' W. 174-0 m. Gear N 100 B. St. WS 564. 21. ii. 31. Moltke Harbour, South Georgia. Shore collection. St. WS576. 17. iv. 31. 51° 35' S, 57° 49' 45" W. 34-24 m. Gear OTC. Bottom: sand. St. WS582. 30. iv. 31. 53° 42' 30" S, 70° 55' W. 12 m. Gear NH. St. WS583. 2. V. 31. 53° 39' S, 70° 54' 30" W. 14-78 m. GearBTS. Bottom : sand and stones. St. WS648. 22. vi. 31. 15° 19' 30" S, 75° 13' W. iii-om. Gear N 100 B. St. WS742. 5.ix. 31. 38° 22' S, 73° 41' W. 35 m. GearBTS. St. WS748. 16. ix. 31. 53° 41' 30" S, 70° 55' W. 30o(-o)m. GearBNR. St. WS750. 18. ix. 31. 52° 12' S, 67° 19' W. 95(-o)m. GearBNR. St. WS752. 19-20. ix. 31. 51° 20' S, 63° 17' W. i6o(-o)m. GearBNR. St. WS755. 2i.ix. 31. 51° 39' S, 57° 39' W. 75(-o)m. GearBNR. St. WS 756. 10. X. 31. 50° 54' 39" S, 59° 58' W. 118-90 m. Gear OTC. Bottom: gravel, mud and sand. St. WS758. 12. X. 31. 48° 31' S, 61° 19' W. ii2(-o)m. GearBNR. Bottom: rock. St. WS762. 16. X. 31. 43° 50' S, 65° 01' 51" W. 67-65 m. Gear OTC, N7-T. Bottom: sand and mud St. WS 763. 16. X. 31. 44° 14' S, 63° 28' W. 87-82 m. Gear OTC. Bottom: mud and sand. St. WS 764. 17. X. 31. 44° 38' 15" S, 61° 58' 30" W to 44° 38' 45" S, 61° 49' 30" W. 106 m., gear DC; 1 10-104 i^-' g^^i" OTC. Bottom: fine green sand. St. WS765. 17.X.31. 45° 07' S, 60° 28' 15" W. ii3-ii8m. Gear OTC. Bottom: mud and sand. STATION LIST 67 St. WS766. 18-19. X. 31. 45° 13' S, 59° 56' 30" W. 545 m. Gear NCS-T. Bottom: fine dark green sand. St. WS 770. 21. X. 31. 46° 03' S, 66° 34' W. 95 (-0) m. Gear BNR. Bottom: brown-grey clay. St. WS771. 29. X. 31. 42° 41' 45" S, 60° 31' W. 90 m. Gear DC, NCS-T, N 7-T. Bottom: dark green sand. St. WS772. 30. X. 31. 45° 13' 22" S, 60° 00' 15" W. 309-153 m. Gear NCS-T, N 7-T, N4-T. St. WS 773. 31 X. 31. 47° 28' S, 6o°5i'W. 291m., gear DC; 291-29S m., gear OTC. Bottom: green sand and mud. St. WS 774. I. xi. 31. 47° 08' S, 62° 02' W. 139 m. Gear DC. Bottom: dark green sand and mud. St. WS776. 3.xi. 31 46° 18' 15" S, 65° 02' 15" W. 110-99 m. Gear OTC, DC. Bottom: green mud and sand. St. WS777. 3.xi. 31. 45° 56' S, 66° 24' W. 98-99 m. Gear OTC. Bottom: mud and sand. St. WS 781. 6. xi. 31. 50° 30' S, 58° 50' W. 148 m. Gear NCS-T, OTC. Bottom: dark green sand and mud. St. WS 7S2. 4. xii. 31. 50° 29' is" S, 58° 23' 45" W. 141 m. Gear DC. 50° 27' 45" S, 58° 29' 45" W. 141-146 m. Gear OTC. Bottom: green sand. St. WS783. 5. xii. 31. 50° 02' 45" S, 6o°io'W. 155 m. Gear DC. 50° 02' 45" S, 60° 14' Wr 155-159 m. Gear OTC. Bottom: rock, mud and sand. St. WS784. 5. xii. 31. 49° 47' 45" S, 61° 05' W. 170-164 m. Gear N 7-T. Bottom: dark green sand. St. WS785. 6. xii. 31. 49° 23' 45" S, 62° 41' 15" W. 150-146 m. Gear OTC, N 7-T. Bottom: dark green sand. St. WS786. 7. xii. 31. 49° 07' S, 63° 55' W. 133 m. Gear DC. Bottom: dark sand. St. WS787. 7. xii. 31. 48° 44' S, 65° 24' 30" W. io6-iiom. Gear OTC. Bottom : coarse brown sand. St. WS788. 13. xii. 31. 45° 07' S, 64° 54' W. 82-88 m. Gear OTC. Bottom: grey mud and sand. St. WS795. 18. xii. 31. 46° 14' S, 60° 24' W. 157-161 m. Gear OTC. Bottom: sand. St. WS796. 19. xii. 31. 47° 53' 30" S, 63° 32' 30" W. io6-ii3m. Gear OTC. Bottom: coarse brown sand. St. WS797. 19. xii. 31. 47° 47' 43" S, 64° 07' 30" W. iii-ii4m. Gear OTC. Bottom: stones. St. WS798. 20. xii. 31. 47° 32' S, 65° 02' W. 49-66 m. Gear NCS-T. Bottom: pebbles, shell and sand. St. WS801. 22.xii.31. 48° 26' 15" S, 61° 28' W. 165m. Gear OTC, NCS-T. Bottom: dark sand. St. WS803. 5.1.32. 50° 33' 45" S, 62° 05' 30" W. 174-186 m. Gear OTC. St. WS804. 6. i. 32. 50° 22' 45" S, 62° 49' W. 150-143 m. Gear OTC. 50° 21' 15" S, 62=53' W. 143-150 m. Gear N 7-T. Bottom: gravel and sand. St. WS805. 6.1.32. 50° 10' 15" S, 63° 29' W. 148 m. Gear NCS-T. Bottom: coarse dark sand. St. WS 807. 7. i. 32. 49° 50' 30" S, 65° 03' W. 125-126 m. Gear OTC. Bottom: dark sand. St. WS808. 8.1.32. 49° 40' 15" S, 65° 42' W. iio-io7m. Gear NCS-T. Bottom: brown- green sand. 68 DISCOVERY REPORTS St. WS809. 8. i. 32. 49° 28' 15" S, 66° 29' W. 107-10401. Gear NCS-T. Bottom: brown sand. St. WS811. 12. i. 32. 51° 24' 30" S, 67° 53' W. 96-9801. Gear OTC. Bottom: sand and stones. St. WS813. 13.1.32. 51° 35' 15" S, 67° 16' 15" W. 106-102 m. Gear OTC. Bottom: dark sand. St. WS 814. 13. i. 32. 51° 45' 15" S, 66° 40' W. 111-118 m. Gear OTC. Bottom: coarse dark sand. St. WS 824. 19. i. 32. 52° 29' 15" S, 58° 27' 15" W. 146-137 m. Gear OTC. Bottom: green sand and shell. St. WS 825. 28-29.1.32. 50° 50' S, 57° 15' 15" W. 135-144 m. Gear OTC. Bottom: sand, mud and shells. St. WS832. i.ii. 32. 50° 49' S, 67° 55' W. 75-0 m. Gear N 70 B. St. WS834. 2.11.32. 52° 57' 45" S, 68° 08' 15" W. 27-38 m. Gear OTC, NCS-T. Bottom: dark brown and grey stones, mud and sand. St. WS836. 3.11.32. 53° 05' 30" S, 67° 38' W. 64 m. GearBTS. St. WS 837. 3. 11. 32. 52° 49' is" S, 66° 28' W. 98-102 m. Gear OTC, N 7-T, N 4-T, NCS-T. Bottom: coarse dark green sand and pebbles. St. WS 839. 5. 11. 32. 53° 30' 15" S, 63° 29' W. 403-434 m. Gear N4-T. Bottom: fine sand and mud. St. WS840. 6.11.32. 53°52'S, 6i°49' i5"W. 368-463 m. Gear OTC, N 4-T. Bottom: green- grey sand. St. WS841. 6.11.32. 54° 11' 45" S, 60° 21' 30" W. 109-120 m. Gear OTC. Bottom: stones and shell. St. WS847. 9.11.32. 50° is' 45" S, 67°S7'W. 51-56 m. Gear OTC. 50° 18' 45" S, 67° 44' 00" W. 56-84 m. Gear NCS-T. St. WS 848. 10.11.32. 50° 37' 30" S, 66° 24' W. ii5-ii7m. Gear OTC. Bottom: dark green sand. St. WS 849. 10. 11. 32. 50° 56' 45" S, 64° 58' W. 137-137 m. Gear OTC. Bottom: dark sand. St. WS851. 11.11.32. si°39'3o"S, 62°oi' is"W. 22i-i97m. Gear OTC, N 7-T. Bottom: stones. St. WS852. 21. Hi. 32. 44° 12' 30" S, 64° 13' W. 86-88 m. GearBTS. St. WS8s6. 23.111.32. 46° 35' S, 64° 11' W. 104-104 m. GearBTS. St. WS863. 28. Hi. 32. 49° 05' S, 64° 09' W. I2i-ii7m. GearBTS. St. WS866. 29. Hi. 32. 50° 37' 45" S, 64=15' W. 137-144 m. Gear OTC. St. WS867. 30. Hi. 32. 51° 10' S, 64° 15' 30" W. 150-147 m. GearBTS. St. WS869. 3i.iH. 32. 52°i5'3o"S, 64°i3'4s"W. i87(-o)m. GearBNR. St. WS871. i.lv. 32. 53° 16' S, 64° 12' W. 336-341 m. GearBTS. St. WS877. 4. Iv. 32. 52° 35' 30" S, 61° 04' W. 3So(-o)m. GearBNR. MARINE BIOLOGICAL STATION, SOUTH GEORGIA St. MS 15. 17. 11. 25. East Cumberland Bay, South Georgia, 3 miles SW of Merton Rock to 2J miles NNW of Dartmouth Point, no m. Gear DS. St. MS 62. 24. ii. 26. East Cumberland Bay, South Georgia, i cable E to 3! cables S of Hobart Rock. 31 m. Gear BTS. LIST OF SPECIES 69 St. MS 64. 24. ii. 26. 1-8 miles SExS of King Edward Point Light, East Cumberland Bay, South Georgia. 7-15 m. Gear DS. St. MS 65. 28. ii. 26. East Cumberland Bay, South Georgia. i-6 miles SE of Hobart Rock to I cable N of Dartmouth Point. 39 m. Gear BTS. St. MS 68. 2. iii. 26. East Cumberland Bay, South Georgia. 1-7 miles S^E to 8 J cables SE X E of Sappho Point. 220-247 m. Gear NRL. Euphrosyne arctia, Johnson Euphrosyne maorica, Augener Aphrodite longirostris, Kinberg Aphrodite talpa, Quatrefages LIST OF SPECIES Family AMPHINOMIDAE Chloeia inermis, Quatrefages Family APHRODITIDAE Laetmatonice prodiicta, Grube berg- Harmothoe magellatuca (Mcintosh) Harmothoe exanthema (Grube) Harniothoe exanthema, Grube, % stromi, var.nov. Harmothoe brevipalpa, Bergstrom Harmothoe brevipalpa, Bergstrom, var. ciliata, var.nov. Hartnothoe benthophila, Ehlers Harmothoe ernesti, Augener Harmothoe spinosa, Kinberg Harmothoe spinosa, Kinberg, var. lagiscoides, Willey Harmothoe (Barrukia) cristata (Willey) Eunoe anderssoni (Bergstrom) Family POLYNOIDAE Eulagisca corrieiitis, Mcintosh Hermadion magalhaensi, Kinberg Polynoe antarctica, Kinberg Lepidametria gigas (Johnson) Hololepida australis, n.sp. Halosydna patagonica, Kinberg Antinoe antarctica (Bergstrom) Euphionella patagonica, gen. et sp.nov. Macellicephala mirabilis, Mcintosh EiicroHta mollis (Mcintosh) Eucranta villosa, Malmgren, var. notialis, var. nov. Polyeunoa laevis, Mcintosh Sigalion ovigerum, Monro Leanira quatrefagesi, Kinberg Psammolyce semiglabra, n.sp. Phyllodoce longipes, Kinberg Phyllodoce patagonica (Kinberg) Phyllodoce madeirensis, Langerhans Phyllodoce boiversi, Beixham Eulalia magalhaensis, Kinberg Eulalia picta, Kinberg Genetyllis polyphylla (Ehlers) Alciopa cantrainii (Delle Chiaje) Vanadis antarctica (Mcintosh) Vanadis formosa, Claparede Vanadis crystallina, GreefF Vanadis violacea, Apstein Tomopteris carpenteri, Quatrefages Tomopteris planktonis, Apstein Family SIGALIONIDAE Sthenelais litnicola zealandiae var.nov. (Ehlers), var. novae- Family PHYLLODOCIDAE Eteone sculpta, Ehlers Lopadorhynchiis krohnii (Claparede), var. simplex, Monro Lopadorhynchus uncinatus, Fauvel Pelagobia longicirrata, GreefT Mystides notialis, Ehlers Family ALCIOPIDAE Greeffia oaluiensis, Mcintosh Callizona angelini (Kinberg) Callizonella bongraini (Gravier) Torrea Candida (Delle Chiaje) Family TOMOPTERIDAE Tomopteris cavallii, Rosa Tomopteris septentrionalis, Quatrefages 70 Sagitella lobifera, Ehlers DISCOVERY REPORTS Family TYPHLOSCOLECIDAE Travisiopsis benhatni, n.sp. Syllis prolixa, Ehlers Syllis sclerolaema, Ehlers Syllis hrachychaeta, Schmarda Trypanosyllis gigantea (Mcintosh) Trypanosyllis taeniaeformis (Haswell) Pionosyllis comosa, Gravier Nereis (Eiinereis) hardyi, Monro Nereis cricognatha, Ehlers Nereis callaoana, Grube Nereis jacksoni, Kinberg Nephthys dibranchis, Grube Nephthys serratifolia, Ehlers Glycera capitata. Oersted Goniada eximia, Ehlers Ephesia antarctica, Mcintosh Eunice frauenfeldi, Grube Eunice pemiata (O. F. Miiller) Eunice australis, Quatrefages Diopatra punctifera, Ehlers Diopatra sp. Diopatra neapolitana, Delle Chiaje Rhamphobrachium ehlersi, Monro Omiphis conchylega, Sars Oniiphis iridescens (Johnson) Onuphis dorsalis (Ehlers) Scoloplos marginatus (Ehlers) Family SYLLIDAE Pionosyllis nutrix, n.sp. Eiisyllis kergtielensis, Mcintosh Amblyosyllis granosa, Ehlers Autolytus charcoti, Gravier Autolytus simplex, Ehlers Polybostrichus sp. Family NEREIDAE Nereis eugeniae (Kinberg) Nereis kerguelensis, Mcintosh Platynereis magalhaensis, Kinberg Leptonereis loxechini (Kinberg) Family NEPHTHYDIDAE Nephthys macrura, Schmarda Nephthys squamosa, Ehlers Family GLYCERIDAE Glycinde armata (Kinberg) Family SPHAERODORIDAE Family EUNICIDAE Onuphis aiicklandensis, Augener Hyalinoecia tubicola (O. F. Miiller) Lumbrinereis magalhaensis, Kinberg Lumbrinereis heteropoda, Marenzeller Lumbrinereis cingulata, Ehlers Lumbrinereis near impatiens, Claparede Auseneria tentaculata, Monro Nino'4 Jalklandica, n.sp. Drilonereis filum (Claparede) Family ARICIIDAE Haploscoloplos kerguelensis (Mcintosh) Family CIRRATULIDAE Cirratidus cirratus (O. F. Muller) Cirratulus filiformis, Keferstein Cirratulus antarcticus, Monro Family SPIONIDAE Polydora natrix, Soderstrom Family CHAETOPTERIDAE Chaetopterus variopedatus (Renier) Family CHLORHAEMIDAE Stylarioides kerguelarum (Grube) Flabelligera affinis, M. Sars Stylarioides swakopianus, Augener LIST OF SPECIES 71 Travisia kerguelends, Mcintosh Ammotrypane scaphigera, Ehlers Family OPHELIIDAE Ammotrypane breviata, Ehlers Ophelia hipartita, n.sp. Family MALDANIDAE Clymene (Isocirrus) yungi (Gravier) Axiothella antarctica, Monro Maldane sarsi, Malmgren, var. atitarctica, Arwidsson Lumbridymenella robusta, Arwidsson Asychis amphiglypta (Ehlers) Sahellaria {Phragmatopoma) antipoda, Augener SaheUaria {Phragmatopoma) moerchi (Kinberg) Family SABELLARIIDAE Idanthyrsus armatiis, Kinberg Pectinaria ehlersi, Hessle Ampharete kerguelends, Mcintosh Amage sculpta, Ehlers Phyllocomus crocea, Grube Family AMPHICTENIDAE Family AMPHARETIDAE NeosabelUdes elongatus (Ehlers) Amphicteis philippinarum, Grube Amphicteis guimeri (Sars), var. antarctica, Hessle Family TEREBELLIDAE Amphitrite kerguelends, Mcintosh Atnphitrite afftnis, Malmgren, var. antarctica, var.nov. Leaena abranchiata, Malmgren, var. antarctica, Mcintosh Leaena collaris, Hessle Nicolea chilends (Schmarda) Polymnia nebulosa (Montagu) Neoleprea streptochaeta (Ehlers) Loimia medusa, Savigny Pista mirabilis, Mcintosh Thelepus setosus (Quatrefages) Thelepus cincitmatiis (Fabricius) Streblosorna bairdi (Malmgren), var. antarctica, var.nov. Polycirrus kerguelends (Mcintosh) Poly cirrus hesslei, Monro Hauchiella tribuUata (Mcintosh) Octobranchus antarcticus, n.sp. Sabella oatesiana, Benham Potamilla antarctica (Kinberg) Oridia limbata (Ehlers) Serpiila venuicidaris, Linnaeus Vermiliopsis notialis, Monro Loandalia aberrans, gen. et sp.nov. Family SABELLIDAE Chone diineri, Malmgren Euchone pallida, Ehlers Family SERPULIDAE Spirobranchus latiscapiis (Marenzeller) INCERTAE SEDIS The total number of species, exclusive of a Polybostrichus and a Diopatra to which I have not given a name, is 159. There are two new genera, a Polynoid Euphionella, and Loandalia, a curious tropical West African form which I am unable to assign to any of the known families. The number of new species is eight and there are six new varieties. 72 DISCOVERY REPORTS GEOGRAPHICAL DISTRIBUTION BENTHIC SPECIES Although the present collection was obtained from positions as widely separated as off Peru, West Africa and New Zealand, much the greater part of the benthic material came from two areas, the South Georgia area which for the present purposes includes South Georgia, South Sand^\■ich Islands, South Orkneys, South Shetlands, Palmer Archipelago, Bellingshausen Sea, and the Falkland Islands area which includes the Falkland Islands and between the Falkland Islands and South America. The colder water Polychaete fauna of the South Georgia area is not the same as that round the Falklands, although there is considerable overlapping. The bentliic species fall into the following groups according to the areas in which they were found : I. SOUTH GEORGIA, SOUTH SANDWICH ISLANDS, BELLINGSHAUSEN SEA Laetmatonice producta Harmothoe magellanica Harmothoe spinosa Harmothoe (Barrukta) cristata Eunoe anderssoni Antinoe antarctica Macellicephala mirabilis Eucranta mollis Polyeunoa laevis Phyllodoce patagonica Genetyllis polyphylla Syllis brachychaeta Pionosyllis mitrix Eusyllis kerguelensis Amblyosyllis granosa Aiitolytus charcoti Nereis kerguelensis Leptonereis loxechini Nephthys macrura Glycera capitata Eunice pennata Diopatra sp. RJiamphobrachium ehlersi Onuphis conchylega Lumbrinereis magalhaensis Augeneria tentaculata Scoloplos marginatus Haploscoloplos kerguelensis Cirratulus cirratus Cirratulus antarcticus Cirratulus filiformis Polydora natrix Stylarioides kerguelarum Ammotrypane breviata Clymene [Isocirrus) yungi Axiothella antarctica Maldane sarsi var. antarctica Asychis amphiglypta Ampharete kerguelensis Amage sculpta Phyllocomus crocea Neosabellides elongatus Amphicteis gunneri, var. antarctica Amphi trite kerguelensis Leaena abranchiata, var. antarctica Leaena collaris Pista mirabilis Thelepus cincinnatus Streblosoma bairdi var. antarctica Polycirrus kerguelensis HauchieUa tribuUata Octobranchus antarcticus Potamilla antarctica Oridia limbata Euchone pallida Serpula vermicularis Vermiliopsis notialis GEOGRAPHICAL DISTRIBUTION 73 The following additional species were Discovery Report: Paramphinome atistralis Euphrosyne arctia Aphrodite alta Hermadion ferox Hermadion magalhaensi Eulagisca corrientis Eunoe opalina Harmothoe crosetensis Harmothoe kerguelensis Harmothoe curviseta Antinoe setobarba Eteone sculpta Eteone aurantiaca Eteone rubella Phyllodoce bowersi Phyllodoce longipes Aiistrophylbim charcoti Eulalia magalhaensis Eulalia anomalochaeta Eulalia picta Pionosyllis comosa Pionosyllis maxima Trypanosyllis gigantea Autolytus gibber Grubea clavata Syllis prolixa Syllis brachycola Nereis typhla Plafynereis magalhaensis Ephesia antarctica Onuphis notialis Thelepus setosus Lysilla loveni, var. macintoshi The total number of benthic species is 123. recorded from this area in my previous (1930) Ltimbrinereis antarctica Scoloplos mawsoni Pygospio dubia Nerine sp. Paraonis gracilis Phyllochaetopterus sp. Tharyx epitoca Tharyx sp. Flabelligera affinis Flabelligera pennigera Flabelligera mundata Brada villosa Brada mammillata Scalibregma inflatitm Capitella capitata Notomastus latericeus Notomastus lineatus ? Travisia kerguelensis Travisia kerguelensis, var. gravieri Kesiin abyssoruni Rhodine intermedia Lumbriclymenella robusta Clymene kerguelensis Nicomache sp. Sternaspis scutata Melinna cristata Terebella ehlersi Pista corrientis Neoleprea streptoehaeta Lanicides vayssieri Artacama proboscidea Terebellides minutus Terebellides longicaudatus recorded by me from the South Georgia area II. OFF THE FALKLAND ISLANDS AND BETWEEN THE FALKLAND ISLANDS AND SOUTH AMERICA Euphrosyne arctia Aphrodite longirostris Harmothoe magellanica Harmothoe brevipalpa Harmothoe ernesti Harmothoe exanthema Harmothoe spinosa Harmothoe spinosa var. lagiscoides Eulagisca corrientis Hermadion magalhaensi Polynoe antarctica Hololepida australis Halosydna patagonica Antinoe antarctica Euphionella patagonica Eucranta mollis Eucranta villosa var. notialis Polyeunoa laevis Leanira quatrefagesi Phyllodoce longipes Phyllodoce patagonica Phyllodoce bowersi 74 DISCOVERY REPORTS Eulalia magaUiaensis Eulalia picta Eteone sculpta Mystides notialis Syllis prolixa Syllis sclerolaema Trypanosyllis gigantea Eusyllis kergueletms Autolytus charcoti Autolytus simplex Nereis hardyi Nereis eugeniae Nereis kerguelensis Platynereis magalhaensis Leptonereis loxechini Nephthys serratifolia Nephthys macrura Nephthys squamosa Glycera capitata Goniada eximia Glycinde armata Ephesia antarctica Eunice frauenfeldi Eunice pennata Onuphis conchylega Onuphis iridescens Onuphis dorsalis Lumbrinereis tnagalhaensis Lumbrinereis cingulata Lumbrinereis near impatiens Augeneria tentaculata Ninoe falklandica Drilonereis filum Haploscoloplos kerguelensis Cirratulus cirratus Cirratulus antarcticus Chaetopterus variopedatus Flabelligera affi?iis Travisia kerguelensis Ammotrypane scaphigera Ammotrypane breviata Lunibriclytnenella robusta Asychis amphiglypta Sabellaria (Phragmatopoma) ?noerchi Idanthyrsus armatus Pectinaria ehlersi Phyllocomus crocea Amphicfeis gunneri var. antarctica Amphitrite kerguelensis Amphitrite affinis var. antarctica Nicolea chileiisis Polymnia nebulosa Pista mirabilis Thelepus setosus Thelepus cincinnatus Sabella oatesiatia Potamilla antarctica Serptila vermicularis The following additional benthic species were recorded from this area in my previous (1930) Discovery Report: Syllis brachycola Pista corrientis Syllis variegata Neoleprea streptochaeta Lumbrinereis tetraura Polycirrus kerguelensis Aricia michaelseni Polycirrus hamiltoni Travisia olens Polycirrus hesslei Axiothella antarctica Bispira magalhaensis Clymenella minor Salmacina dysteri, var. falklandica The total number of species recorded by me from this area is 94. There are forty-nine species common to both the South Georgia and the Falkland Islands areas, and these constitute 40 per cent of the total number of the species from the South Georgia area and 52 per cent of the species from the Falkland Islands area Harmothoe exanthema Harmothoe brevipalpa, var. ciliata Harinothoe spinosa Polynoe antarctica Lepidametria gigas Antinoe antarctica Nereis eugeniae III. STRAIT OF MAGELLAN Chaetopterus variopedatus Amphitrite affinis, var. antarctica Neoleprea streptochaeta Thelepus setosus Polycirrus hesslei Potamilla antarctica Serpula vermicularis GEOGRAPHICAL DISTRIBUTION IV. BOUVET ISLAND 75 Harmothoe spinosa Harmothoe spinosa, var. lagiscoides Harmothoe (Barnikia) cristata Harmothoe hrevipalpa Genetyllis polyphylla Lumbrinereis magalhaensis Nephthys macrura V. GOUGH ISLAND Nereis callaoana Serpula vermicularis The following additional species were recorded from this area in my previous (1930) Discovery Report : Syllis brachycola Serpula loveni Chaetopterus variopedatus Diopatra punctifera Diopatra neapolitana Lumbrinereis heteropoda VI. OFF WEST AND SOUTH-WEST AFRICA Stylarioides swakopianus Loandalia aberrans The following additional species were recorded from this area in my previous (1930) Discovery Report: Hermodice carunculata var. didymobranchiata Eurythoe complanata Chloeia viridis Notopygos megalops Malmgrenia micropoides Antinoe epitoca Eupanthalis tubifex Polyodontes mortenseni Euthalanessa dendrolepis Leanira incisa Phyllodoce octilata Eulalia viridis Leocrates diplognathus Syllis variegata Ceratonereis vittata Nephthys lyrochaeta Glycera i esse lata Goniada congoensis Eunice siciliensis Eunice vittata Eunice longicirrata Nicidion edentulum Lumbrinereis africana Lumbrinereis coccinea Drilonereis filum Staurocephalus rubrovittatus Prionospio africana Chaetopterus variopedatus Phyllochaetopterus socialis Cirratulus afer Pycnoderma congoense Maldane decor ata Ozvenia fusiformis Sternaspis scutata, var. africana Amphicteis gunneri, var. japonica Loimia montagui Hypsico?nus torquatus Vermiliopsis glandigerus Vermiliopsis richardi, var. fauveli Ophelia bipartita VII. OFF CHILE Thelepus setosus Chone duneri VIII. OFF PERU 76 DISCOVERY REPORTS IX. OFF NEW ZEALAND Euphrosyne maorica Aphrodite talpa Chloeia inermis Sigalion ovigerum Psammolyce semiglahra Sthenelais limicola, var. novae-sealandiae Phyllodoce madeiremis Trypanosyllis taeniaeformis Pionosyllis comosa Amblyosyllis granosa Nereis cricognatha Nereis jacksoni Nephthys dibranchis Eunice aiistralis Onuphis aucklandemis Hyalinoecia tubicola Sabellaria {Phragmatopoma) antipoda Amphicteis phiUppinarmn Nicolea chilensis Spirobranchus latiscapus PELAGIC SPECIES The following list gives a rough indication of the localities and the minimum and maximum depths at which the pelagic species were obtained : Harmothoe benthophila. South of Cape Verde Island; 236-0 m. Lopadorhynchus krohnii, var. simplex. South of Cape Verde Island; 236-0 m. Lopadorhynchus uncinatus. East of La Plata; 246-0 m. Pelagobia longicirrata. South Georgia; 174-0 m. and 1000-750 m. Alciopa cantrainii. Off Pernambuco ; 182-0 m. Vanadis antarctica. South Georgia, Bouvet Island, Burdwood Bank; 0-5 m. and 1 150-1400 m. Vanadis formosa. Off Cape Verde Island, South Africa, St Paul Rocks, Rio Grande and La Plata; loi-o m. and 242-0 m. Vanadis crystallina. South of Pernambuco and off Bahia; 208-0 m. and 216-0 m. Vanadis violacea. Off South Africa; 84-0 m. and 550-350 m. Greeffia oahuensis. Off South Africa; 550-350 m. Callisona angelini. Off South Africa; 1200-0 m. Callizonella bongraini. Off South Georgia; 174-0 m. Torrea Candida. East of La Plata; 246-0 m. Tomopteris carpenteri. Off South Georgia and Bouvet Island; 0-5 m. and 550-250 m. Tomopteris planktonis. Off Pernambuco, north of Rio de Janeiro, east of La Plata ; 182-0 m. and 249-0 m. Tomopteris cavallii. North-east of Bouvet Island; 170-0 m. Tomopteris septentrionalis . Off St Helena, South Africa, Bouvet Island, South Georgia, Bellingshausen Sea, north of Falkland Islands, off Pernambuco and La Plata; 73-0 m. and 265-150 m. Sagitella lobifera. Off South Georgia and between South Georgia and Gough Island; 1 150-1400 m. and 1 500-1 600 m. Travisiopsis benhami. South Georgia and Bellingshausen Sea; 97-0 m. and 1000-750 m. Loimia medusa juv. Off South Africa; 106-0 m. and 350-0 m. SYSTEMATIC ACCOUNT As in the descriptive part of this report I have made no diagnoses of the famihes, I give a key, which with slight modifications is that of Fauvel and Gravely. I. Head usually well developed. All segments of the body alike except those near the mouth and the terminal segment, or pygidium ... ... ... ... ... ... ... 2 — Body often divided into several distinct regions. Head small, either not markedly de- veloped or profoundly modified. Feet almost always simple, the ventral rami being often in the form of transverse ridges (tori) or pinnules bearing hooks or uncini. Branchiae generally limited to a definite region 16 KEY TO FAMILIES 77 2. Elytra on a certain number of feet, the rest bearing cirri 3 — No elytra ^ 3 . Compound bristles present Sigalionidae — No compound bristles ... ... ... ... ••• ••• ••• •■■ ••• ••■ * 4. Never more than one segment bearing a dorsal cirrus intercalated between two elytrigerous segments. Thread glands present in the feet Polyodontidae — In the hinder region of the body either all segments carry dorsal cirri or there are at least two cirrigerous segments intercalated between two elytrigerous segments (except in Lepidastheniella). No thread glands 5 5. Eyes stalked, rarely sessile. A single tentacle. Facial tubercle conspicuous Aphroditidae — Eyes sessile. Three tentacles (except in Macellicephala, Bylgia and Iphione). Facial tubercle absent or rudimentary .. . ... ... ... ... ... ... Polynoidae 6. A fan-shaped group of broad flattened bristles (paleae) on all segments . . . Chrysopetalidae Dorsal paleae absent ... ... ... ... ■-. ... •■• ••• ••• ■■■ 7 7. Prostomium with two divergent tentacles and flanked by long cirri inclosing acicula. Feet biramous without cirri or bristles Tomopteridae — Prostomium without tentacles. Feet uniramous. Bristles present 8 — Prostomium with tentacles. Feet biramous ... ... ... ... ... ... ... 10 8. Prostomium distinct, conical. Foliaceous dorsal and ventral cirri. A few acicular bristles Typhloscolecidae — Prostomium indistinct ... ... ... ... ... ... ... ... ... ••• 9 9. Body papillated and bearing in addition transverse rows of large spherical capsules Sphaerodoridae — No such papulation. Cirri globular Pisionidae 10. Prostomium small. Five tentacles. Caruncle usually present. Gills well developed. Pharynx unarmed Amphinomidae — Prostomium large... ... ... ... ... ... ... ... ... ... ... 11 1 1 . Pharyngeal armature complex. Upper jaws composed of a number of denticulated plates EUNICIDAE — Pharyngeal armature simple or absent. Tentacles not more than three ... ... ... 12 — Tentacles more than three ... ... ... ... ... ... ... ... ... 14 12. Palps simple. Pharynx armed with a single tooth or a crown of denticles, and followed by a barrel-shaped muscular gizzard. Feet uniramous ... ... ... ... Syllidae — Palps biarticulate, sometimes absent. No gizzard. Pharynx armed or unarmed ... ... 13 13. Dorsal cirri of moderate length, not moniliform. Pharynx with a pair of toothed jaws and almost always with numerous small paragnaths. Feet usually biramous ... ... Nereidae — Dorsal cirri long, moniliform. Pharynx cylindrical, unarmed or armed only with stylets (except in A/ocra/za). Feet biramous and sesquiramous ... ... ... Hesionidae 14. Prostomium conical, annulated, ending in four small tentacles. Pharynx papillated and armed with at least four teeth. Feet biramous ... ... ... ... Glyceridae — Prostomium more or less rectangular with four small tentacles. Feet biramous, with lamellae and a sickle-shaped gill between the rami. Pharynx usually with a pair of jaws and rows of papillae. Bristles simple ... Nephthydidae — Feet with foliaceous cirri, without sickle-shaped gill, usually uniramous 15 15. A pair of enormous globular eyes at the sides of the head ... ... ... Alciopidae — Eyes small, normal Phyllodocidae 16. Body divided into distinct regions 23 — Body not divided into distinct regions ... 17 17. Segments numerous. Without anal branchiae and ventral shield 18 3-2 78 DISCOVERY REPORTS — Body short, stout. Segments few. Filiform anal branchiae and a large ventral shield bordered with stiif bristles Sternaspididae i8. Palps long, tentacle-like 19 — No tentacle-like palps ... ... ... ... ■■• ■•• ■•• ••• ••• ... 22 19. Two long tentacular palps on prostomium ... ... ... ... ... ... ... 20 — One or more pairs of palps inserted on the anterior segments. Branchiae simple, filiform, inserted above the feet. Capillary and usually acicular bristles. Prostomium conical without processes Cirratulidae 20. Two palps and two groups of branchiae retractile into a buccal funnel. Chaetae of first feet prolonged forwards to form a cephalic cage. Body thickly papillated ... Chlorhaemidae — Two long grooved palps not retractile into the mouth. No cephalic cage ... 21 21. Palps without suckers. Pedal lamellae erect. Dorsal branchiae cirriform. Hooded hooks present Spionidae — Palps with sucker-like papillae. Without branchiae. Prostomium spoon-shaped Magelonidae — Anterior cirri flask-shaped or frilled. Lateral branchiae filiform. Several kinds of bristle Disomidae 22. A single median tentacle. Dorsal cirri and foliaceous dorsal branchiae. Capillary bristles and hooded hooks Paraonidae — Prostomium with or without two short tentacles. Parapodia more or less conspicuous. Capillary bristles and forked bristles. No hooks Scalibregmidae — Prostomium blunt, without appendages or with a crown of laciniated lobes. No branchiae. Ventral tori with many rows of minute uncini. Tube sandy Oweniidae — Prostomium with a keel or bordered cephalic plate. An anal plate or an anal funnel with cirri. No branchiae. Dorsal bristles capillary, ventral sigmoid hooks ... Maldanidae 23. A terminal branchial tuft with numerous filaments bearing secondary processes. Pro- stomium indistinct. Uncini ventral in the thoracic region, dorsal in the abdominal region. Tube membraneous or calcareous ... ... ... ... ... ... ... ... 32 — Without terminal branchial tuft ... ... ... ... ... ... ... ... 24 24. Large flattened chaetae (paleae) forming an operculum closing the tube 31 — Without opercular bristles ... ... ... ... ... ... ... ... ... 25 25. Prostomium conical or blunt, without processes. Branchiae on many segments 28 ■ — Prostomium more or less distinct. One pair of tentacle-like palps or numerous tentacular filaments ... ... ... ... ... ... ... ... ••• ... ... ... 26 26. Prostomium with or without two small tentacles. Two long grooved palps. Two to three markedly dissimilar regions, the anterior short with uniramous feet bearing special chaetae in 4th chaetiger. Posterior notopods erect. Uncini pectinate Chaetopteridae — Without tentacles. A cephalic veil and numerous tentacular filaments. Ventral tori with pectinate uncini ... ... ... ... ... ... ... ... ... ... ... 27 27. Tentacular cirri retractile into the mouth. Prostomium distinct. Three to four pairs of branchiae inserted on the first segments ... ... ... ... ... Ampharetidae — Tentacular cirri not retractile. Prostomium indistinct. Branchiae arborescent or rarely subulate, inserted on the first segments. Sometimes absent .. . ... ... Terebellidae 28. With uncinigerous tori 30 — Without uncinigerous tori 29 29. Serrated capillary bristles and acicular hooks. Feet and branchiae conspicuous and erect on the back in the abdominal region ... ... ... ... ... ... ... Ariciidae — Capillary bristles only. Feet without lobes. Branchiae lateral, ligulate. Prostomium pointed; conical ... Opheliidae 30. Prostomium blunt. Anterior region abranchiate: middle with dorsal, arborescent, non- retractile branchiae: often an achaetous and abranchiate caudal region ... Arenicolidae AMPHINOMIDAE 79 — Prostomium conical. Anterior region abranchiate; posterior region with or without simple branchiae; or the branchiae may be branched and retractile into lateral pouches. In the abdominal region, dorsal and ventral tori with sigmoid hooks ... ... Capitellidae 31. An operculum of an anterior row of large golden paleae. Caudal region very small and foliaceous, with hooks at the base. Two pairs of anterior branchiae. Tube of sand, conical, free Amphictenidae — Two large, opercular stalks bearing a crown of paleae. A narrow achaetous and abranchiate caudal region. Branchiae dorsal and numerous. Fixed, sandy tubes, often in masses Sabellariidae 32. Without operculum and without thoracic membrane. Tube membraneous or mucous Sabellidae — Usually with an operculum. Thoracic membrane present. Tube calcareous Serpulidae Family AMPHINOMIDAE Notopodial bristles in transverse rows across the back ... ... ... ... Euphrosyne Notopodial bristles lateral ... ... ... ... ... ... ... ... Chloeia Genus Euphrosyne, Savigny Body oval and segments few. The prostomium which bends over the front end of the body is partly dorsal and partly ventral. Caruncle with three parallel longitudinal lobes. An unpaired tentacle. Lateral tentacles on the ventral surface. Notopodial bristles in transverse rows across the back. Branchiae ramified and similarly disposed in trans- verse rows of trunks behind the dorsal bristles. Paired dorsal cirri on each side. Five pairs of branchial trunks per segment Euphrosyne arctia Six pairs of branchial trunks per segment Euphrosyne maorica Euphrosyne arctia, Johnson. Monro, 1930, p. 34, fig. 4 a-e. Occurrence. St. WS 228 (9); WS 231 (2); WS 246 (4); WS 871 (3). Specific characters. Length about ii mm. Number of chaetigers between 17 and 21. Mouth reaches to anterior border of 5th chaetiger. Caruncle high and superficially divided into three longitudinal lobes. It reaches back to the 6th chaetiger. Branchiae begin on the ist chaetiger and are arranged in transverse rows of five trunks on each side. These trunks usually have four branches and end in small pointed tips. The lower of the two dorsal cirri lies between the 2nd and 3rd most dorsal branchial trunks. The dorsal bristles consist of (i) smooth bifid bristles, (2) ringent bristles, (3) a few bristles apparently intermediate between the " ringents " and the smooth " bifids ". The ventral bristles are simple bifids of two sizes. Remarks. Neither in this nor in the earlier Discovery collections (Monro, 1930) was an example of this species obtained above the loo-m. line. Euphrosyne maorica, Augener (Fig. i). Augener, 1924, p. 259, fig. i a-d. Occurrence. St. 935, New Zealand (i). 8o DISCOVERY REPORTS Specific characters. A small species with a length of about lo mm. for 25 chaetigers. The present specimen measures 5 mm. by 2 mm. at the widest part for 21 chaetigers. There are two pairs of eyes, one ventral just in front of the buccal lobes and one dorsal on the prostomium. The caruncle reaches to the 6th chaetiger and the median tentacle is about one-third of its length. The gills begin on the first chaetiger. In a normal segment there are six branchial trunks on each side. Each trunk is very richly branched and the tufts end in narrow unexpanded tips. The lower of the two dorsal cirri lies between the 4th and 5th most dorsal branchial trunks. The dorsal bristles consist of "ringent" bristles (Fig. i) and smooth bifid bristles. The ventral bristles are smooth bifids of two sizes. Augener could not in his specimens see any serrations on the short arm of the dorsal ringent bristles. In this specimen they are distinct. Moreover, Augener gives the position of the lower dorsal cirrus as between the 2nd and 3rd most dorsal branchia. I find it between the 4th and 5th, and this agrees with Augener 's figure. ■05mm Fig. I. Euphrosyne maorica. Ringent bristle. Genus Chloeia, Savigny Body oval. Caruncle a long plaited crest with marginal folds. Branchiae pinnate; dorsal cirri single; anus terminal. All bristles usually more or less bifurcated. Chloeia inermis, Quatrefages. Benham, 1916A, p. 390, figs. 6-1 1. Augener, 1924, p. 258. Occurrence. St. 939, New Zealand (i juv.). Specific characters. There is no colour pattern, and except for the purple median tentacle and dorsal cirri the body in spirit is more or less without colour. The gills begin on the 5th chaetiger. The chaetae show no trace of serrations. Both in the dorsal and ventral bundles they may be {a) perfectly smooth, {b) may exhibit a minute obso- lescent spur. The ventral bristles are thinner than the dorsal, especially those in the middle of the neuropod which are long, hair-like and extremely fine and also have an incipient spur. The present specimen is a very young example of the species and measures 8 mm. by 3 mm. at the widest part for 23 chaetigers. Up to the middle of the body dorsal bristles of the two types already described are found, and in addition there are a few bristles with small but complete spurs, but in the hinder region the dorsal bristles appear all to be smooth. In the neuropod some of the ventral bristles of the first few chaetigers have an obsolescent spur; otherwise they are smooth. The long, slender, capillary type of ventral bristle is not present behind about the loth chaetiger. . Aphrodite Laetmatonice APHRODITIDAE 8i Family APHRODITIDAE Dorsal bristles smooth ... Dorsal bristles harpoon-shaped ... Genus Aphrodite, Linnaeus A thick dorsal felting covers the elytra. Dorsal bristles of two kinds: (i) protective spines usually projecting through the dorsal felting, (2) very long and slender bristles. Ventral bristles stout and with slightly curved tips. They are arranged in three tiers. Eyes, when present, sessile. Dorsal bristles stout, needle-like, projecting above the dorsal felting. Median tentacle very long Aphrodite longirostris Dorsal bristles slender, not projecting above the dorsal felting. Median tentacle short Aphrodite talpa Aphrodite longirostris, Kinberg (Fig. 2 a, h). Kinberg, 1857, p. 4, pi. i, fig. 3 b-h. Occurrence. St. WS 798 (2). Description. This species has been elaborately figured by Kinberg, who neverthe- less fails to convey its characteristic appearance. The palisade of long dark brown I MM Fig. 2. Aphrodite longirostris. a. Dorsal view. b. Lower pinnate bristle, second foot. needle-like spines (Fig. 2 a) projecting upwards and backwards through the dorsal f eking but not meeting across the back, is unlike that of any other Aphrodite that I have seen. The dorsal spines and the ventral bristles are dark brown and the dorsal felting is greenish grey. The dorsal capillary bristles are covered with mud, but when freed from this show a slight iridescence. The larger specimen measures 75 mm. by 31 mm. at the widest part for about 35 chaetigers. The shape of the body is like that of an 82 DISCOVERY REPORTS A. aculeata but rather wider and deeper relatively to the length. Except that I see no eyes, Kinberg's figure (3 B) of the head well represents that of these specimens. The prostomium is rounded and devoid of ocular peduncles. There is a very long and slender median tentacle without a distal enlargement : it is about three times as long as the head and two-thirds of the length of the palps. Its slightly stouter tentaculophore is about as long as the head. There is a large, laterally compressed, papillated facial tubercle. The tentacular cirri are short, being only about two-thirds of the length of the median tentacle. The tentacular segment, as usual, carries only notopodial bristles. The elytra are figured by Kinberg : their surface is covered with a network of very fine lines and they carry a few minute papillae. The dorsal felting is very solid and compact and has a thickness of about i mm. The dorsal spines are very long, relatively twice as long as those oi A. aculeata, and sharp, and show about 18 mm. of length above the dorsal felting. They make a formidable palisade above the back but leave uncovered a narrow path down the middle. The silky, capillary dorsal bristles appear to be quite smooth, and are thickly covered with mud. The ventral bristles of the first two feet (2nd and 3rd chaetigers) are of three types: (i) the upper are stout dark brown bristles with slightly curved ends ; (2) the middle are rather long bristles with most of the shaft smooth but pinnate towards the slender tip ; (3) the lower are rather like the middle bristles, but much smaller and with strongly developed pinnae over most of their length (Fig. 2 i). In the middle feet the ventral bristles consist of smooth, strong chaetae with slightly hooked ends. They are not bearded. In the hinder neuropods there are the usual denticulated bristles similar to those figured by me (Monro, 1930, fig. 5/-?) for A. aha. In the middle feet the ventral cirri reach to the end of the foot, in the posterior feet they are longer. Remarks. This species is characterized by the long median tentacle and by the palisade of dorsal spines. A. aiistralis has prominent, long dorsal bristles, but they are flattened, relatively delicate structures, very difl^erent from the sharp spines of this species. Of the species of Aphrodite from high southern latitudes A. echidna, Quatrefages of Mcintosh, has a very short median tentacle and short sharp spines projecting through the dorsal felting; A. alta, Kinberg, has dorsal bristles with slender hooked tips which do not penetrate the dorsal felting. Aphrodite talpa, Quatrefages (Fig. 3). Quatrefages, 1865, i, p. 196, pi. vi, figs. 2-4. Fauvel, 1925, p. 140, fig. 4 a-l. Non Ehlers, nee Benham, Augener, Fauvel (191 7). Occurrence. St. 936, New Zealand (3); 939, New Zealand (i). Specific characters. This species is characterized by the slenderness of the dorsal bristles which are entangled for most of their length in the dorsal felting, but have their ends lying obliquely along the back. Moreover, the narrowing of the hinder end into a kind of tail which is apparent in most members of the genus to some extent, is here APHRODITIDAE 83 carried very much further. The anterior four-fifths of the body is broadly oval, but the hinder fifth consists of a narrow caudal prolongation (Fig. 3). The largest specimen measures 50 mm. by 25 mm. at the widest part for 40 chaetigers and the smallest measures 17 mm. by 7 mm. for 35 chaetigers. There is a very small median tentacle, half the length of the head. There is a pair of rounded ocular areas each of which carries two minute black dots, which I take to be eyes. The dorsal bristles are long, rather slender, smooth and with delicate curved tips. They are covered with mud and lie obliquely along the back and with the naked eye are almost impossible to distinguish from the rest of the felting. This separates the present species from the much commoner A. australis, Baird, in which the dorsal bristles are much stouter, more abundant and more prominent. The extent to which the dorsal bristles are entangled with the felting is variable. They may be almost covered by it, or they may lie for the most part loosely above it. The ventral bristles are arranged in three rows in the usual manner. Towards the end they taper suddenly into fine, sharp- looking points, and the narrow tip is usually covered with hairs. In the first and second feet there are the usual hastate bristles and twisted bipinnate bristles and in the hinder region there are found hastate, denti- culated, bipinnate and spinous bristles as in other species. Remarks. I am satisfied that these specimens do not belong to A. australis and they agree well enough with Fauvel's redescription of the type of A. talpa, Quatrefages. Fauvel, however, makes no mention of the caudal prolongation which is a noticeable feature in the present specimens. They are very close to the specimen from the Palmer Archipelago attributed by me (1930, p. 37) to A. alta, Kinberg. They can be distin- guished by the presence of the tail, which in A. alta is very little developed. Fig. 3. Aphrodite talpa. Ventral view. Genus Laetmatonice, Kinberg A median tentacle, beneath which is a large papillated facial tubercle. No lateral tentacles. Eyes on short peduncles. 15-20 pairs of elytra. Dorsal felting either absent or slightly developed. Dorsal bristles harpoon-shaped. Ventral bristles bifurcated with a row of stiff hairs at the tip. Laetmatonice producta, Grube. Gravier, 191 1, p. 80. Fauvel, 1923, p. 38. Monro, 1930, p. 39. Augener, 1932 a, p. 13. Occurrence. St. 363 (3); 474 (10). 84 DISCOVERY REPORTS Specific characters. Dorsal felting absent. May attain a size of 1 8 cm. by 3 J cm. with about 50 chaetigers. 18-20 pairs of elytra. Harpoon-shaped bristles with five to six teeth. Slender bipinnate ventral bristles confined to the first four chaetigers. There may be as many as 50 stiff hairs below the tip of the bifurcated ventral bristles. L.filicornis is a much smaller species with only 15 pairs of elytra. Both Mcintosh and Fauvel state that the slender ventral bipinnate bristles of the anterior region extend to the first five chaetigers. Working, it is true, with rather poor material I have myself failed to find any behind the 4th chaetiger. Family POLYNOIDAE 1. With no lateral tentacles... ... ... ... ... ... ... ... Macellicephala — With lateral tentacles 2 2. 1 2 pairs of elytra .. . ... ... ... ... ... ... ... ... Euphionella — 15 pairs of elytra ... ... ... ... ... ... ... ... ... ... ... 3 — 18 pairs of elytra ... ... ... ... ... ... ... ... ... ... ... 8 — Elytra numerous. Body long, vermiform ... ... ... ... ... ... ... 9 3. Body short, about 40 segments, more or less completely covered by elytra ... ... 4 — Body short, between 40 and 50 segments, not covered by elytra in hinder region ... Hermadion — Body long and vermiform, elytra confined to anterior region ... ... ... Polynoe 4. Lateral tentacles inserted subterminally. A facial tubercle present .. . ... ... Eulagisca — Lateral tentacles inserted ventrally. No facial tubercle ... ... ... ... ... 5 5. Upper ventral bristles slender, capillary ... ... ... ... ... ... ... 6 — Upper ventral bristles moderately stout ... ... ... ... ... ... ... 7 6. Upper ventral bristles capillary with a long spinous region and a hair-like tip ... Antino'e — Upper ventral bristles ending in a minutely bifid tip... ... ... ... ... Eucranta 7. Ventral ramus having either bidentate bristles only, or both bidentate and unidentate bristles Harmotho'e — Ventral ramus having unidentate bristles only ... ... ... ... ... Eunoe 8. Lateral tentacles inserted subterminally. Dorsal bristles slender and heavily pectinated ... Halosydna 9. With both a facial tubercle and an occipital flap ... ... ... ... ... Hololepida ■ — With neither facial tubercle nor occipital flap .. . ... ... ... ... ... ... 10 10. Lateral tentacles inserted terminally ... ... ... ... ... ... Lepidatnetria — Lateral tentacles inserted ventrally ... ... ... ... ... ... ... Polyeunoa Genus Harmothoe, Kinberg Lateral tentacles inserted ventrally. Prostomium with lateral peaks. Fifteen pairs of elytra more or less covering the whole back. Dorsal chaetae stouter than the ventral which are usually bidentate, but may be both bidentate and unidentate. Subgenus Barriikia. Harmothoe (Barrukia) cristata 3- Dorsal bristles with bearded tips Dorsal bristles without bearded tips Dorsal bristles of two distinct types Dorsal bristles of one type Ventral bristles usually all bidentate Ventral bristles partly bidentate and partly unidentate 4. Elytral tubercles few, confined to a small area near scar of attachment Harmothoe benthophila 3 4 7 Harmotho'e magellanica POLYNOIDAE 8s — Elytral tubercles numerous, spread over most of the surface ... ... ... ... 5 5. Smaller elytral tubercles bollard-shaped ... Harmothoe ernesti — Smaller elytral tubercles hook-shaped ... ... 6 6. Hinder border of elytra usually carrying a few large conical vesicles. About 37 segments Harmothoe spinosa — Hinder border of elytra usually carrying long curved spines. About 40 segments Harmothoe spinosa, \a.r. lagiscoides 7. Elytra with large vesicles 8 — Elytra without large vesicles ... ... ... ... ... ... ... ... ... 9 8. Elytral vesicles pear-shaped with an apical papilla ... ... ... Harmothoe exanthema — Elytral vesicles globular, without an apical papilla . . . Harmothoe exanthema var. bergstromi 9. Elytra with only a few cilia at the hinder border Harmothoe brevipalpa — Elytra ciliated over most of the surface ... ... Harmothoe brevipalpa, var. ciliata Harmothoe magellanica (Mcintosh). Lagisca magellanica, Mcintosh, 1885, p. 82, pi. xiii, fig. 5; pi. xviii, figs. 3-4; pi. viiA, figs. 1-2. Harmothoe magellanica, Bergstrom, 1916, p. 280, pi. iv, figs. 1-3. Monro, 1930, p. 54. Augener, ig^^a, p. 18. Occurrence. St. 123 (2); 156 (i juv.); WS 27 (4); WS 225 (4); WS 228 (i); WS 239 (4); WS 244 (5); WS 246 (3); WS 249 (i); WS 764 (i); WS 776 (i); WS 803 (i); WS 825 (5); WS 840 (i); WS 852 (10); WS 871 (numerous). Specific characters. There may be about 15 segments behind the last pair of elytra. There are often small dark spots on the elytra which are smooth except for a small patch of minute tubercles near the umbilicus. The dorsal bristles are stout, numerous and lightly striated ; the ventral are long and clearly bidentate, and the upper ventral bristles have the toothing continued much farther down the shaft than the rest. Remarks. As the hinder end of the body is left uncovered by the elytra in this species, it would perhaps be preferable to refer it to Lagisca rather than to Harmothoe. Harmothoe exanthema (Grube). Bergstrom, 1916, p. 287, pi. iii, fig. 5. Occurrence. St. 53 (3); WS 582 (i); WS 762 (2); Puerto Bueno, Sarmiento Channel, 13 m. (2 juv.). Specific characters. About 40 chaetigers. Anterior pair of eyes at the sides of the head. The elytra do not quite cover the hinder extremity. They are provided with small, conical tubercles and also at their hinder border with a few large pear-shaped vesicles surmounted by a long papilla. The dorsal bristles are strongly pectinated and the ventral are both unidentate and bidentate. This species is allied to H. ernesti, Augener, and seems to live in shallow water. Harmothoe exanthema (Grube), var. bergstromi, var.nov. (Fig. 4 a-c). Occurrence. St. WS 221 (i); WS 583 (i); WS 834 (5). Varietal characters. The pyriform tubercles with an apical papilla are absent and the elytra (Fig. 4 a) carry in addition to numerous small acuminate tubercles (Fig. 4 b) 4-2 86 DISCOVERY REPORTS a number of gigantic soft globular vesicles (Fig. 4 c) without an apical papilla. As in the stem-form the ventral bristles are both bidentate and unidentate, the former being more numerous in the middle of the bundle. The specimens from St WS 834 present a remarkable appearance, for the whole surface of about the hinder third of the scales is a mass of gigantic round vesicles with a granular structure, each attached to the scale by a narrow base. The specimen from St. WS 221 has much fewer and smaller vesicles, and the only remaining scale on the specimen from St. WS 583 has a few large vesicles. All these structures are at once separable from the much smaller, harder, pyriform and papillated tubercles in the stem- form. The largest of the present specimens measures 30 mm. by 5 mm. without the feet for 40 chaetigers. ■DIMM IMM 02 MM a. Elytron. a C Fig. 4. Harmothoe exanthema, var. bergstrotni. b. Acuminate tubercles. c. Globular vesicles. Remarks. Augener, who has seen the type, states that Polynoe vesiculosa, Grube, from the Magellan region is the same as Harmothoe exafithema, and as I regard the present examples as distinctly separable from the latter species I have established a new variety. I rather suspect that Ehlers (1897, p. 14 and 1901, p. 42) had before him examples both of this form and of exanthema, and included them both under the name vesiculosa. Harmothoe brevipalpa, Bergstrom (Fig. 5). Bergstrom, 1916, p. 277, pi. ii, fig. i; pi. iv, figs. 4-7. Augener, 1932Z), p. 100. Harmothoe {Evarnella) impar, var. notialis, Monro, 1930, p. 58, fig. 13 a-d. Occurrence. St. 399 (6); WS 229 (3). Specific characters. A small species measuring about 15 mm. by 2 mm. without the feet for 34 chaetigers. There is a typical harmothoid head with the anterior pair of POLYNOIDAE 87 Fig. 5. Harmothoe brevipalpa. Head from above. eyes in the middle of the lateral surfaces of the prostomium. The palps are normal, being about twice as long as the head (Fig. 5). The elytra have a few small papillae at the external border and are rather sparsely dotted with small conical tubercles. The dorsal bristles are numerous and strongly pectinated. The upper ventral bristles are elongate, slender, bidentate and with a long spinous region: more ventrally they become shorter and more ex- panded distally. The middle ventral chaetae are bidentate. The bristles at the base of the neuropod are unidentate and with a short spinous region. Remarks. Augener has rightly pointed out that the new variety of H. impar described by me in 1930 differs from Bergstrom's H. brevipalpa only in the absence of the great reduction in size of the palps which Bergstrom found in his specimen and treated as an important specific differential. Augener further holds the view that the reduction of the palps in Bergstrom's specimen was an accidental condition due to loss followed by regeneration. As a number of specimens have been found since Bergstrom's original description which agree with his type except in the matter of the palps, I accept Augener's conclusion. My H. impar var. notialis therefore becomes a synonym of H. brevipalpa. Harmothoe brevipalpa, Bergstrom, var. ciliata, var.nov. (Fig, 6). Occurrence. St. WS 583 (i). Varietal characters. The variety differs from the stem-form in that the outer and posterior borders of the elytra are furnished with a fringe of long cilia. Moreover, cilia are not confined to the marginal fringe but are found over a great part of the surface of the scale (Fig. 6), especially in the area adjoining the fringe. In other respects the variety and the stem-form are indistinguishable. Remarks. This variety is based on a single complete specimen measuring 13 mm. by 3 mm. for 35 chaetigers. Harmothoe benthophila, Ehlers. Fauvel, 1923, p. 68, figs. 24 h-o. Occurrence. St. 702 (i). Specific characters. Prostomium bilobed without frontal peaks. Median tentacle longer than the palps, lateral tentacles shorter than the head. Ten to eleven pairs of elytra covering the back; they are large, soft, transparent and carry a few papillae. Dorsal bristles of two sorts: (i) short and curved with rows of scales and tips ending in two blunt points, (2) very long straight bristles with spirally arranged scales and tips also ending in two blunt points. Fig. 6. Harmothoe brevipalpa, var. ciliata. Elytron. 88 DISCOVERY REPORTS The upper ventral bristles are long, slender and spinous with curved bidentate tips ; the lower ventral bristles are shorter and broader; they are almost smooth and have bidentate tips. There is a caudal appendage. The present specimen measures 4 mm. by I mm. without the feet for about 24 chaetigers. Harmothoe ernesti, Augener.'' (Fig. 7 a-d). Augener, 1931, p. 281, fig. 2 a-f. Occurrence. St. WS 811 (i); WS 834 (6). Specific characters. Body rather elongate for a Harmothoe. Number of segments between 35 and 41. An average specimen measures 25 mm. by 3 mm. without the feet. The lateral tentacles are short, about as long as the head, one-third as long as the median tentacle and tentacular cirri, and one-fourth as long as the palps. The anterior pair of eyes is placed in the middle of the head at its lateral edges. The elytra leave the •5MM 01 MM •5 MM Fig. 7. Harmothoe ernesti. a. Elytron. c. Vesicle. b. Tubercles. d. Ventral bristle. posterior extremity uncovered. They are fringed (Fig. 7 a) and carry numerous small bollard-shaped tubercles with thick cuticular caps (Fig. 7 b) and a few large egg-shaped or urn-shaped vesicles (Fig. 7 c). These large vesicles tend to be situated at the hinder margin of the scales especially in the posterior region of the body. In the front region they are distributed more widely over the surface. The dorsal bristles are stout and strongly pectinated except at the tip, which is smooth. Just above the neuropodial aciculum the apex of the chaeta-sac is produced into a finger-shaped process. The ventral bristles (Fig. 7 d) are rather slender, with numerous frills and a bidentate tip. The second tooth lies well below the terminal tooth. POLYNOIDAE 89 Remarks. I regard these specimens as rather doubtfully belonging to Augener's species from the Abrolhos Bank off Brazil. The elytral vesicles appear to be similar, but in Augener's species they are more numerous. The present specimens are probably southern representatives of Hartnothoe impar, Johnston. Harmothoe spinosa, Kinberg. Ehlers, 1913, p. 438, pi. xxvi, figs. 1-12. Bergstrom, 1916, p. 284, pi. ii, figs. 5-6; pi. iii, figs. 1-4. Occurrence. St. 371 (7); 456 (numerous); WS 27 (3); WS 225 (2); WS 239 (i); WS 244 (4); WS 583 (i); WS 762 (3); WS 764 (i); WS 782 (3); WS 784 (2); WS 801 (2); WS 811 (i); WS 824 (2); WS 82s (5); WS 834 (2); WS 837 (2); WS 867 (4); MS 65 (I). Specific characters. There are 37 segments. The elytra are mottled with reddish brown and the back and feet are banded to a variable extent with brown markings. There are numerous, minute, conical, acuminate tubercles on the elytra, and in addition the hinder border may be provided with several large, vesicular tubercles. These are often absent. Furthermore, the external edge of the elytra may or may not be furnished with a fringe of cilia. The dorsal bristles may be strongly pectinated or almost smooth. The ventral are toothed and as a rule bidentate. Specimens with unidentate ventral bristles are, however, found, and also intermediate specimens in which one or two faintly bidentate bristles may be seen in neuropods bearing unidentate bristles. Remarks. Almost the only constant specific character that I can find after examining a large series of this species is the presence of the numerous, minute, conical acuminate tubercles on the elytra. The large haul of specimens from off Bouvet Island yielded a number of examples with unidentate ventral bristles. Harmothoe spinosa, Kinberg, var. lagiscoides, Willey. Willey, 1902, p. 265. Gravier, 191 1, p. 92, pi. vi, figs. 64-69. Harmothoe lagiscoides, Bergstrom, 1916, p. 282, pi. ii, figs. 2-3. Augener, 1932 a, p. 15. Occurrence. St. 456 (8); WS 764 (6). Varietal characters. The number of segments is about 40, and the pigmentation is more intense than in H. spinosa. A few of the terminal segments are left uncovered by the elytra, and the hinder end is noticeably more tapered than in the stem-form. The hinder border of the elytra is furnished with a few long, acuminate spines. Remarks. Bergstrom and Augener have raised Willey's var. lagiscoides to specific rank. In my earlier Antarctic report I included it under H. spiftosa. I have now come to the conclusion that its characters are sufficiently constant to merit at any rate varietal status. Subgenus Barrukia, Bergstrom As Harmothoe, except that the dorsal bristles mostly have bearded tips. Harmothoe (Barrukia) cristata (Willey). Gattyana cristata, Willey, 1902, p. 268, pi. xliv, figs. 1-4. Barrukia cristata, Bergstrom, 1916, p. 297, pi. v, figs. 7-9 and 14. Occurrence. St. 363 (4); 456 (2). 90 DISCOVERY REPORTS Specific characters. There is a row of median dorsal pads running the whole length of the body. The elytra are papillated and have clavate tubercles with crenate tops. The dorsal bristles mostly have bearded tips and the ventral are unidentate with two short rows of teeth on the shaft. Remarks. The specimens from St. 363 are distinctive in having in the hinder part of the elytra a patch of relatively gigantic spinous pustules which are variable in shape and size. Genus Eunoe, Malmgren As Harmothoe, but with unidentate ventral bristles. Eunoe anderssoni (Bergstrom). Harmothoe anderssoni, Bergstrom, 1916, p. 286, pi. iii, fig. 6; pi. iv, figs. 8-10. Monro, 1930, p. 57. Occurrence. St. 123 (3); WS 33 (2); MS 62 (i). Specific characters. A small species measuring about 10 mm. by i mm. without the feet for 34 chaetigers. There is a typical harmothoid head and two pairs of rather large eyes. The anterior pair are on the sides of the head in the middle, and the posterior are dorsal but at the extreme lateral edges. The elytra have both marginal and sub- marginal papillae and both large and small tubercles with irregular jagged tips. The dorsal bristles are numerous and pectinated : the upper neuropodial bristles are rather elongate and slender with a long spinous region showing scales on the blade as well as a toothed edge. From above dov^nwards the bristles show an increasingly shorter spinous region, and the lower ventral bristles are expanded distally and have a very short spinous region showing a toothed edge only. All the neuropodial bristles are unidentate. This appears to be the normal condition, but in the specimens from St. 123 the two or three uppermost bristles in the ventral bundle show distinct traces of a second tooth. Genus Eulagisca, Mcintosh The arrangement of the elytra is as in Harmothoe . The insertion of the lateral tentacles is terminal or subterminal, not ventral. A large facial tubercle is present. The dorsal bristles are stouter than the ventral and lightly pectinated : the ventral are frilled and unidentate. Eulagisca corrientis, Mcintosh. Mcintosh, 1885, p. 91, pi. xiii, fig. 4; pi. viiA, figs. 3-4. Monro, 1930, p. 48, fig. 11 a-e. Augener, 1932 a, p. 19. Occurrence. St. WS 246 (i). Specific characters. This is a large species measuring up to about 80 mm. by 15 mm. without the feet for 36 segments. It is superficially very like a Panthalid. The head is of the lepidonotid type in that the insertion of the lateral tentacles is terminal, but the top of the median ceratophore Hes a little above the lateral ceratophores after POLYNOIDAE 91 the manner of Halosydna. All the appendages are hirsute except the palps, which are covered with small papillae. There is a well-developed conical facial tubercle, and behind the head a nuchal flap or gibbosity. The bristles of the tentacular segment are unusually numerous and well developed. The elytra are arranged as in Harmothoe, and the pseudo- elytrophores are scarcely less prominent than the elytrophores. The dorsal cirrophores are set very low down on the feet and have a prominent lateral expansion. The dorsal cirri are very long, the tips of the bristles only reaching to about half their length. The bristles are as described for the genus. Remarks. I cannot agree with Augener that this species is capable of inclusion under Harmothoe. The genus Eulogisca is very near to Allmmiiella, Mcintosh, from which it difi"ers in the possession of unidentate instead of bidentate ventral bristles and of a facial tubercle. The latter may prove to be only a specific character. Genus Hermadion, Kinberg Fifteen pairs of elytra arranged as in Harmothoe but leaving the hinder end of the body uncovered. There are about 15 segments behind the last pair of elytra. The lateral tentacles are inserted ventrally, but there are no prostomial peaks. The dorsal bristles are numerous, stout, and either smooth or very lightly pectinated, the ventral bristles are unidentate. Hermadion magalhaensi, Kinberg. Fauvel, 1916, p. 423, pi. viii, figs. lo-ii, with synonymy. Occurrence. St. 652 (2); WS 84 (i); WS 576 (i); WS 755 (numerous); WS 841 (2). Specific characters. The elytra are variably coloured with brown and white markings. They are thickly covered with small tubercles. The dorsal bristles are dark brown, stout, upturned, and either smooth or very lightly pectinated: the ventral are unidentate and with well-developed scales. Genus Polynoe, Savigny Body long and vermiform, with numerous segments. Fifteen pairs of elytra confined to the anterior region and leaving a large number of hinder segments uncovered. The lateral tentacles are inserted ventrally. The dorsal ramus is much less developed than the ventral. The dorsal bristles are smooth or lightly pectinated; the ventral are frilled and either unidentate or bidentate. Polynoe antarctica, Kinberg. Kinberg, 1857, p. 23, pi. x, fig. 58. Fauvel, 1916, p. 426. Monro, 1930, p. 53. Harmothoe antarctica, Bergstrom, 191 6, p. 279. Occurrence. St. WS 583 (i); WS 755 (5); WS 869 (5). Specific characters. The elytra are usually punctuated with small dark spots, or they may be bordered with brown. They are quite smooth except for a patch of minute 92 DISCOVERY REPORTS tubercles near the umbilicus. Prostomial peaks are present. The palps are papillated, and the tentacles and cirri carry a few sparse papillae. The dorsal bristles are few in number (four to six) and either smooth or very lightly pectinated. The ventral bristles are short, rather stout, carry rows of spines and are clearly bidentate. Genus Lepidametria, Webster The body is long and vermiform with between 60 and 150 segments and up to 50 pairs of elytra. The lateral tentacles are terminally inserted. The elytra are inserted on segments 2, 4, 5, 7, 9 and on alternate segments up to between the 25th and 30th chaetigers. Behind this the arrangement is irregular. The notopod is represented by an aciculum and sometimes by a few bristles also. Lepidametria gigas (Johnson) (Fig. 8 a, b). Polyno'e gigas, Johnson, 1897, p. 172, pi. vii, figs. 33, 42, 42 a; pi. viii, figs. 48, 48 a, 48 b, 49. Lepidametria gigas, Seidler, 1924, p. 145. Occurrence. St. WS 583 (i). Specific characters. The single specimen measures 70 mm. by 8 mm. including the feet for 88 chaetigers. There are 48 pairs of large elji:ra which completely cover the body. They are mottled with patches of iron grey pigment. The body itself has no colour. The head corresponds to Johnson's figure. It is broad at the base and the anterior pair of eyes is placed laterally at the widest part. The bases of the tentacles are rather elongate. The tentacles, palps and tentacular cirri all extend about an equal distance beyond the end of the head. The ventral cirrus of the first chaetiger is almost as long as the tentacular cirri. The elytra are quite smooth and correspond to Johnson's figure. According to Johnson the elytra are arranged as in Halosydna up to the 33rd chaetiger, i.e. on 2, 4, 5, 7, 9. . .27, 28, 30, 31, 33 ; after that on alternate segments up to the 49th chaetiger and then very irregularly. In this specimen the arrangement is a little different. The elytra are inserted on 2, 4, 5, 7, 9. . .29, 30, and from then onwards on alternate segments to the end of the body except in a few places where this regular arrange- ment is interrupted by two or three elytra being attached to consecutive segments. I have seen no asymmetrical segments with a cirrus on one side and an elytron on the other. The notopod is represented by an aciculum and I see no bristles. The neuropod carries a sheaf of stout bristles with frilled and expanded ends. The tips may be either unidentate or bidentate and both types of bristle occur in the same foot, and also inter- Fig. 8. Lepidametria gigas. a. Upper bristle, first foot. b. Lower bristle from first foot. POLYNOIDAE 93 mediate stages in which the second tooth appears to be abortive. In the front region the bristles are predominantly bidentate and in the hinder region unidentate. The first two chaetigers carry special bristles. In the first foot there are in the upper part of the bundle a few moderately stout bristles with frilled ends and long, whip-like slender tips (Fig. 8 a). The rest of the bundle consists of delicate barbed bristles (Fig. 8 b). In the second foot the bristles are similar to those of the first except that the stout upper bristles are more numerous and the delicate barbed bristles much fewer. Remarks. Johnson's Californian specimens measured up to 165 mm. in length, and the elytra did not extend to the hinder extremity. Moreover, the arrangement of the elytra is a little difi"erent from that given by Johnson, who also makes no mention of the special bristles in the first two segments. Nevertheless, I believe this specimen to be- long to Johnson's species, of which it is probably a young example. The arrangement of the elytra brings it close to my genus Lepidastheniella characterized by the presence of elytra on the 2nd, 4th, 5th, 7th, and on every alternate segment to the end of the body. Ehlers's L. irregularis (Ehlers, 1901, p. 54) is very near, but the elytra are relatively much smaller, more rounded and do not overlap. They present a very different appearance from that of the widely overlapping, more or less reniform or oval structures in the present species. The specimen is stated to have been found as a commensal with a Terebellid which itself occupied an empty Gastropod shell of the genus Valuta. Genus Hololepida, Moore Up to about 120 chaetigers. The head has the lateral tentacles inserted subterminally. There are a large occipital flap or gibbosity and a prominent facial tubercle. The elytra are inserted on segments 2, 4, 5, 7, and on alternate chaetigers up to the 23rd segment : from the 23rd to about the 40th segment they follow an irregular sequence, and from behind about the 40th chaetiger they are found in every segment. The notopodial bristles are fine, smooth capillaries: the upper neuropodial bristles are slenderly lanceolate and delicately denticulated, the lower neuropodial are coarser and have rows of frills and bidentate tips. Hololepida australis, n.sp. (Fig. 9 a-h). Occurrence. St. WS 246 (i); WS 248 (i); WS 824 (i); WS 825 (i). Description. In its general aspect this species is more like a Panthalid than a Poly- noid. It is very large and striking to the eye. None of the specimens is complete and the largest fragment measures 90 mm. by 8 mm. without the feet and 17 mm. with the feet for only 38 chaetigers. The most complete fragment measures 63 mm. by 12 mm. with the feet for 42 chaetigers. Another specimen is broken into three pieces which together measure 95 mm. by 12 mm. including the feet for 59 chaetigers. These three fragments although all apparently belonging to the same individual do not represent an entire animal. The body is elongate, vermiform and flattened dorso-ventrally. In spirit the dorsum is a bluish grey, the feet and the large bolster-like elytrophores being colourless. 94 DISCOVERY REPORTS There are traces of reddish brown pigment on the head, tentacles, dorsal cirri and elytra. The head (Fig. 9 a) is bilobed, and each lobe is roughly triangular. The lobes of the prostomium are continued into the ceratophores of the lateral tentacles, but the cerato- phore of the median tentacle lies a little above the lateral ceratophores as in Halosydna. •IMM f Fig. 9. a. Head from above. b. Elytral tubercles. c. Middle foot. d. Dorsal bristle. Hololepida australts. e. Uppermost ventral bristle. /. Bidentate ventral bristle. g. Intermediate type of ventral bristle. h. Part of ventral bristle highly magnified. At the postero-lateral borders of the head are two pairs of enormous contiguous eyes provided with lenses. The palps reach back to the 5th chaetiger, and the three tentacles and the tentacular cirri are all of about the same length, which is approximately two- thirds that of the palps. Tentacles, cirri and palps are smooth. Below the median POLYNOIDAE 95 ceratophore is a facial tubercle, and the back of the head is covered by a large nuchal flap. The tentacular cirri are carried forward on each side so that they arise a little in front of the head. Between their upper and lower ceratophores and on the inside there is a small cirriform process enclosing an aciculum. Nearly all the elytra are lost. Such as remain are frayed at the edges and damaged. They appear on the following segments: 2, 4, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 26, 29, 30, 32, 35, 38, 39, 40, 42, 43. The most complete fragment has only 42 chaetigers, so that I cannot follow the arrangement in the more posterior region. I suspect that behind the 42nd segment there are elytra on every segment. At any rate a hinder fragment from St. WS 824 has elytrophores on all the segments. The elytra are large, thick, more or less broadly oval, and have a gelatinous appearance. Under the microscope they have a fibrous texture and are dotted with very minute three-pronged tubercles (Fig. 9 b). The feet (Fig. 9 c) are large and triangular. The dorsal ramus is much reduced and is represented by a long projecting process, containing an aciculum, and a small bundle of bristles arising from the anterior face of the neuropod. The ventral ramus is triangular and the anterior lip is produced into a cirriform process inclosing an aciculum. In the notopod the acicular process is postsetal and in the neuropod presetal. Dorsal bristles are absent from the first two chaetigers. The dorsal cirri are long and extend well beyond the tips of the bristles : the ventral scarcely reach to the end of the foot. From about the loth chaetiger backwards there is a white, glandular patch sur- rounding the base of the ventral cirrus. Nephridial papillae are visible from the 6th chaetiger. The dorsal bristles (Fig. 9 d) are long, smooth, very fine capillaries. The ventral bristles are stouter than the dorsal. The uppermost ventral bristles (Fig. 9 e) are long, slenderly lanceolate and towards the hair-like apex have faintly denticulated edges. Below these and occupying most of the neuropod both above and below the acicular process there are moderately stout bristles (Fig. 9/) with bidentate tips, and they have along the blade a row of oblique pockets with plain edges. Between these two types of ventral bristles there are one or two bristles more or less intermediate in type (Fig. 9 g). Their general shape is like that of the upper lanceolate bristles but they carry rows of pockets (Fig. 9 h) like the lower bidentate bristles. Remarks. I was at first inclined to regard these specimens as capable of inclusion within H. magna, Moore, from Alaskan waters, but they show a number of diflFerences which in my opinion justify the establishment of a new species. The arrangement of the elytra in Moore's specimen and mine is similar up to the 32nd segment: Moore gives elytra on 32, 33, 35, 36, 38, 39, 40, 41, etc.: I find them on 32, 35, 38, 39, 40, 42, 43. Moore makes no mention of tubercles on the elytra; in these specimens there are characteristic three-pronged tubercles. Moore states that the two types of ventral bristle are separated in the foot by the aciculum ; in these specimens there are numerous bristles of the bidentate type above as well as below the aciculum. Finally Moore states that dorsal bristles are absent from the first three notopods; in these specimens they are absent from the first two notopods. 96 DISCOVERY REPORTS Genus Halosydna, Kinberg Body oblong. Lateral tentacles subterminally inserted. Pairs of elytra i8 or 21, covering the terminal segments. Dorsal bristles rather slender and strongly pectinated, ventral with rows of frills below the tip, which may be either unidentate or bidentate. Halosydna patagonica, Kinberg. Kinberg, 1857, p. 17, pi. v, fig. 23 a-h. Seidler, 1924, p. 116, with synonymy. Occurrence. St. WS 762 (2); WS 834 (20); WS 837 (i); WS 847 (i). Specific characters. Thirty-six segments and eighteen pairs of elytra. The elytra are fringed and covered with small tubercles in addition to which there are a number of large conical vesicles well figured by Kinberg (fig. 23 H). The dorsal bristles are slender and heavily pectinated, the ventral are rather stout, bidentate and with frilled ends. The neuropod of the ist chaetiger carries special bristles, which are delicate and barbed. Remarks. This species does not apparently penetrate as far south as South Georgia, for it was not represented in the extensive collections from that area made by the 'Discovery' in 1925-7, nor is it reported by Bergstrom from the collections of the Swedish South Polar Expedition. Genus Antinoe, Kinberg Fifteen or sixteen pairs of elytra arranged as in Harmothoe . The lateral tentacles are inserted ventrally and prostomial peaks are present. The dorsal bristles are stout and pectinated ; the ventral are long and slender with elongated spinous regions and hair- like tips. Antinoe antarctica (Bergstrom). Austrolaenilla antarctica, Bergstrom, 1916, p. 291, pi. iii, fig. 8; pi. v, figs, i and 2. Antinoe antarctica, Monro, 1930, p. 66, fig. 18, Occurrence. St. 39 (i); 123 (2); 142 (i); WS 211 (3 juv.); WS 212 (numerous); WS 213 (5); WS 214 (7); WS 229 (numerous fragments); WS 234 (6); WS 236 (numerous fragments); WS 237 (4); WS 244 (5); WS 748 (2); WS 752 (i); WS 758 (i); WS 773 (numerous fragments); WS 784 (i); WS 805 (numerous fragments); WS 839 (2 juv.). Specific characters. The head is very broad and divided by a median groove. The prostomial peaks are not clearly defined. The eyes are very small and sometimes invisible. The median tentaculophore is very large and all three tentaculophores are reddish brown. The elytra have a variable number of small tubercles and on their external margin a few clavate papillae. There are 15 pairs as in Harmothoe. The dorsal bristles are lightly pectinated and the ventral are long, slender, unidentate and end in a bearded or hirsute tip. POLYNOIDAE 97 Genus Euphionella, gen.nov. As Euphione, Mcintosh, setisu Seidler (1924, p. 98), but characterized by the presence of pseudo-elytra {vide Seidler, 1921, p. 90) and of completely smooth ventral bristles. The lateral tentacles are terminally inserted. There are 12 pairs of elytra arranged as in Lepidoiiotm and completely covering the body. The segments having a dorsal cirrus are provided with fan-shaped membranes which Seidler has called pseudo-elytra. Ramose branchiae are present from the 3rd chaetiger on the hinder faces of the feet, and small, globular branchial processes are found on the elytrophores and on the corre- sponding structures in the cirrigerous segments. The dorsal bristles are exceedingly fine barbed capillaries ; the ventral are stout, simple and without teeth or ornamentation of any kind . Genotype: Physalidonotus lobulatus, Seidler. Euphionella patagonica, n.sp. (Fig. loa-l). Occurrence. St. WS 212 (i). Description. The specimen measures 19mm. by 4 mm. without the feet for 25 chaetigers. In spirit there is no colour. The elytra are inserted on segments 2, 4, 5, 7, 9. . .21, 23, as in Lepidonotiis . The head (Fig. 10 a) is rectangular, rather longer than broad. The lateral tentacles are terminal. A median groove is only slightly indicated. I see two pairs of rather indistinct, almost contiguous, eyes at the outer and hinder borders of the head. The lateral tentacles are about half as long again as the head and the median tentacle is about three times as long as this. The tentacular cirri are inter- mediate in length between the lateral and the median tentacles, the dorsal being slightly longer than the ventral. The tentacular segment carries a few bristles. The ventral cirrus of the ist chaetiger is slender and elongate, being about twice as long as the foot. All these appendages are smooth, and have a slight subterminal dilatation. The palps, on the other hand, which are about one-third as long again as the median tentacle, are heavily papillated. The elytra (Fig. 10 b) are large, leathery, firmly attached and cover the whole body. They are roughly slipper-shaped, and so disposed that the part corresponding to the heel overlaps the part corresponding to the toe of the succeeding scale. The elytrophores are narrowly oval structures with their long axis lying across the body. They are at- tached to the elytra a little behind the middle point. The elytra are heavily fringed with long cilia. In the first pair the fringe is absent only from the anterior border, and in the remainder the fringe is confined to the outer and hinder edges. In addition to the marginal fringe of cilia there are a number of shorter cylindrical papillae dotted about on the scale itself in the neighbourhood of the marginal fringe. In the region of the umbilicus the elytra show a transverse thickening or hillock caused by the increase of connective tissue bet\\'een the two layers of cuticle, and on the top of this hillock or crest there are three or four very large conical or capstan-shaped tubercles (Fig. 10 c). Besides these there are two further kinds of tubercle. Between the crest and the hinder MM Fig. 10. Euphionella patagonica. a. Head. b. El3^ron. c. Larger type of echinate tubercle. d. Smaller type of echinate tubercle. e. Disk-shaped tubercle. /. Dorsal surface of two middle segments, showing pseudo-elytra. g. Pseudo-elytron. h. Middle foot. /. Dorsal bristle. k. Ventral bristle. /. Lower ventral bristle. POLYNOIDAE 99 border of the scale there are a few smaller, but still large, roughly capstan-shaped, echinate tubercles (Fig. 10 d) and dotted all over the scale there are numerous small disk-shaped tubercles (Fig. 10 e). These are sparse in the front region of the scale. On the back in the median line there is a longitudinal row of small, very soft pads. The ist chaetiger has a single conical pad. In the 2nd chaetiger the pad shows signs of dividing into two pads lying side by side, for there are two cones arising from a single base. The following four chaetigers have two pairs of pads, one pair behind the other and the members of each pair lying side by side. Behind the sixth foot this arrangement in pairs ceases and the pads are continued in a somewhat irregular fashion to the end of the body. Branchiae are present from the 3rd chaetiger. They are small branching structures lying on the hinder face of the feet on a level with the outer edge of the elytrophores. In addition there are two or three minute globular processes, possibly branchial in function, lying on the front and hinder faces of the elytrophores and of the corresponding structures in the cirrigerous segments (Fig. 10/). These last are provided with low oblong cushions running from the beginning of the foot to a point on a level with the inner edge of the elytrophores. To the inner edge of these cushions there are attached roughly fan-shaped flaps or membranes, the inner edge of which lies up against the sides of the median dorsal pads. They are what Seidler has named pseudo-elytra (Fig. 10 g). They are smooth on the dorsal surface, thin and transparent. On the under side they are provided with six raised ridges apparently muscular in character which exactly overlie corresponding dorso-ventral muscular ridges on the back. In addition the under surface carries about a dozen rows of small vesicles which are apparently glandular. The function of these organs is to me quite unknown. The feet (Fig. 10 h) are triangular in outline. The dorsal cirri are long and slender, extending beyond the tips of the bristles. The ventral cirri are short and stout, barely reaching to the end of the foot. They taper to a point. Both dorsal and ventral cirri are smooth, but the feet, except on their upper surface, and indeed the whole of the under surface of the body, are covered with short clavate papillae with large heads. The notopod forms a small rounded lobe on the anterior face of the foot. It is supported by an aci- culum and carries numerous, exceedingly fine, hair-like, barbed bristles (Fig. 10 i). The triangular neuropod carries a narrow fan-shaped bundle of about 1 5 completely smooth, stout, acicular bristles (Fig. 10 k). Not only is there no trace of ornamentation but there is scarcely any sign of curvature at the tip. In addition to these bristles there are at the base of the bundle one or two shorter bristles with lanceolate heads (Fig. 10 /). In the first foot the neuropod is much reduced and is only slightly larger than the noto- pod. The notopodial bristles are very numerous, but the neuropod carries only three or four bristles similar in type to those of the normal feet but much more slender. Small nephridial papillae are apparent from the 4th chaetiger. The terminal segment carries a single pair of pygidial styles. Remarks. The only species with which, to the best of my knowledge, the present curious form is congeneric is Euphione lobulata, Seidler (1921, p. 89, and 1924, p. 99) loo DISCOVERY REPORTS from Callao. From this E. patagonica differs in the ornamentation of the elytra, in the possession of two or three lanceolate bristles in the neuropod and in several further characters. Seidler's species was based on an anterior fragment, and he attributed the lack of ornamentation on the ventral bristles to loss from wear. The study of the present specimen leads me to the conclusion that this was not so, and that there exists a group of branchiate lepidonotid Polynoidae with pseudo-elytra and smooth ventral bristles. Genus Macellicephala, Mcintosh Body rather stout. Between 17 and 29 chaetigers. Prostomium bilobed, usually with well-developed prostomial peaks. There are no lateral tentacles. 8-13 pairs of elytra. Feet biramous with the dorsal ramus much reduced. Dorsal bristles absent or present in small numbers. They may be spinous or smooth. Ventral bristles long, delicate and transparent. Macellicephala mirabilis, Mcintosh. Mcintosh, 1885, p. 121, pi. xvi, fig. i; pi. xHa, figs. 9-11. Macellicephala sp., Monro, 1930, p. 47, fig. 10 a-b. Macellicephala mirabilis, Augener, 1932 b, p. 102. Occurrence. St. 144 (i). Specific characters. A small species with 17-18 chaetigers and a length of 20- 25 mm. The colour is pink to purple on the back. There are no eyes. The head is markedly bilobed. There are nine pairs of elytra. The dorsal ramus has a few bristles that are either smooth or show just a trace of serration towards the tip. The ventral bristles are long and transparent and lightly serrated on one side only. Remarks. The present specimen is from the same haul as the Macellicephala re- corded by me in my previous Discovery report. It is in a slightly better state of pre- servation than the latter and dorsal bristles are present in a number of the feet. Mcintosh and Augener state that they are smooth : under a high magnification I can see some very fine serrations towards the tip. Genus Eucranta, Malmgren As Harmothoe, except that the upper ventral bristles are of a special kind, being long and slender, with rows of pectinae and minutely bifid tips. Middle ventral bristles bidentate Eucranta mollis Middle ventral bristles unidentate Eucranta villosa, var. notialis Eucranta mollis (Mcintosh). Eupolynoe ?nollis, Mcintosh, 1879, p. 259, pi. xv, figs. 5-9. Eucranta mollis, Bergstrom, 1916, p. 294. Monro, 1930, p. 51. Occurrence. St. 363 (i); WS 211 (i); WS 871 (i). Specific characters. The head is round and the lateral tentacles are inserted ventrally. The prostomial peaks are represented by two minute papillae lying above the large, round anterior eyes. The hinder pair of eyes lies at the back of the head. The POLYNOIDAE ,oi elytra are arranged as in Hormothoe : they are large, soft and smooth except for a patch of small tubercles near the umbilicus. The dorsal bristles are stout and pectinated; the upper and lower ventral bristles are long and slender, with rows of teeth and very delicate forked tips. The middle ventral bristles are frilled and bidentate. Bergstrom saw no transitional bristles of a character intermediate between the two types in the neuropods. I cannot confirm this, for at the top of the sheaf of the middle bidentate bristles there are a few bristles longer and more slender than the rest of the middle bristles and with very delicate bidentate tips which are transitional between the two types. In the lower part of the neuropod the two types of bristle are more abruptly separated. Eucranta villosa, Malmgren, var. notialis, var.nov. (Fig. 1 1 a~h). Occurrence. St. WS 788 (i). Varietal characters. This variety is based on an anterior fragment measuring 9 mm. by 4 mm. without the feet for 25 chaetigers. There is a typical harmothoid head (Fig. II a) with well-developed prostomial peaks. The anterior pair of eyes are almost invisible from above and lie in the middle of the prostomium underneath its lateral edges. The hinder pair lie at the postero-Iateral edges of the head. The median tentacle, the palps and the tentacular cirri are all about equal in length and three times as long as the head. The lateral tentacles are minute and just reach to the end of the ceratophore of the median tentacle. The elytra (Fig. 1 1 b) are fringed on their outer and hinder borders and their surface is dotted not only with cilia (much more sparsely, however, than in Malmgren's figure), but also with small tubercles of variable shape (Figs. 11 c-e). Towards the anterior border the tubercles are smaller than over the rest of the scale. The feet are of the usual harmothoid type with a well-developed notopodial bristle-bundle and a triangular neuropodium. The dorsal bristles are sabre-shaped and strongly pectinated. In the ventral ramus the bristles are long, delicate, slender and carry rows of spines. Only the two uppermost bristles (Fig. 1 1 /) show the minutely bifid apex characteristic of the genus. The following half-dozen bristles are bidentate (Fig. 11 ^), for there is a very slender second tooth below the curved tip : the remainder of the bristles are unidentate (Fig. 1 1 //) and are similar to those of the stem-form. The dorsal cirri are long, reaching well beyond the tips of the bristles: the ventral cirri reach to the end of the foot. Remarks. The only substantial difference between the present southern form and the northern villosa is the presence in var. notialis of the bidentate bristles intermediate between the characteristic bifid bristles at the top of the ventral ramus and the uni- dentate bristles which form the great majority. I have never seen an example of Malmgren's species, but as far as I can gather from the accounts there is no transition between the upper bifid bristles and the unidentate bristles. Genus Polyeunoa, Mcintosh Up to 100 segments and 30 pairs of elytra. The body is elongated and vermiform. The lateral tentacles are inserted ventrally. The first 15 pairs of elytra are attached to DISCOVERY REPORTS the following chaetigers: i, 3, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 25, 28, 31. More posteriorly the arrangement is very irregular. The dorsal bristles are stout, and either smooth or very lightly striated. The ventral bristles are toothed, terminally unidentate, and expanded towards the tip. •QE5MM 5MM 025 MM 02MM i Fig. II. Eucranta villosa, var. notialis. a. Head from above. e. Tubercle from hinder end of elytron. b. Elytron. /. Uppermost ventral bristle. c. Tubercle from front end of elytron. g. Bidentate ventral bristle. d. Tubercle from middle of elytron. h. Unidentate ventral bristle. Polyeunoa laevis, Mcintosh. Mcintosh, 1885, p. 76, pi. xii, fig. 2; pi. xx, fig. 8; pi. viiA, figs. 12-13. Enipo rhombigera, Ehlers, 1908, p. 47, pi. iv, figs. 1-12. Polyeunoa laevis, Bergstrom, 1916, p. 288, pi. iii, fig. 7. Monro, 1930, p. 51. Occurrence. St. 599 (numerous); 600(5); 652(io);WS 33 (i); WS 228(1); WS 246 (numerous); WS 773 (fragments); WS 824 (numerous); WS 825 (numerous); WS 840 (numerous); WS 841 (numerous); WS 871 (numerous); WS 877 (5). SIGALIO'NIDAE 103 Specific characters. There is usually a reddish brown longitudinal dorsal stripe, and transverse markings more pronounced in the elytrophorous segments than in the rest. Occasionally they are both absent, or one may be present and the other absent. Gravier (1911, p. 81) gives a good account of the colour pattern. Tentacles and cirri are smooth. Prostomial peaks are very little developed. The elytra are smooth except for a small patch of minute tubercles. Behind the 31st chaetiger their arrangement is extremely variable, and segments with a cirrus on one side and an elytron on the other are very common. The bristles are as described for the genus. Family SIGALIONIDAE 1. No median tentacle — With a median tentacle ... 2. With a dorsal cirrus on the 3rd chaetiger — Without a dorsal cirrus on the 3rd chaetiger 3. Ventral bristles compound falcigers — Ventral bristles compound spinigers ... Sigalion ... 2 Psammolyce ... 3 . Sthenelais Leanira Genus Sigalion, Audouin and Milne-Edwards Body long and vermiform. Head oval, longer than broad. There is no median tentacle, and the lateral tentacles are reduced to small papillae inserted on the front margin of the head. The ist chaetiger carries dorsal and ventral tentacular cirri, a pair of long palps and on each side two bundles of simple bristles. There is a cirriform branchia on all segments behind about the 5th. The dorsal bristles are simple and denticulated. The ventral bristles are both simple and compound, the latter with short and single-jointed shafts or long and multi-articulate shafts. The elytra are furnished with pinnate papillae on the outer margin. Sigalion ovigerum, Monro (Fig. 12 a-d). Monro, 1924, p. 47, figs. lo-ii. Occurrence. St. 936 (New Zealand) (i). Specific characters. The specimen is an anterior fragment measuring 60 mm, by 2 mm. without the feet for 135 chaetigers. This species is chiefly characterized by the absence of compound bristles with simple single-jointed blades, all the compound neuropodial bristles being multi-articulate. The prostomium (Fig. 12 a) is an oval plate with two pairs of minute eyes, one behind the other. There is a pair of papilliform lateral tentacles lying above the ist chaetiger, and a facial tubercle on the under surface of the head. The dorsal cirrus of the ist chaetiger is shorter than the head and the ventral is a little longer than this. They are about one-fourth of the length of the palps. The branchiae appear as slender papillae on the 3rd and 4th chaetigers and are fully de- veloped by the 5th foot. The elytra are smooth and their posterior border is furnished with about a dozen pinnate papillae (Fig. 12 b) with cyHndrical branches, which are about 10-15 i^^ number. There is no sign of the modification of the elytra into pouches containing eggs recorded in the specimen from Port Jackson. 104 DISCOVERY REPORTS The dorsal ramus (Monro, loc. cit., fig. lo) carries three ctenidia and a long cirriform stylode at its apex. The dorsal bristles are long, slender, capillary and denticulated and have very delicate bidentate tips. The neuropodium has a small papilla probably a bract, on its upper face. Above the neuropodial aciculum there are four kinds of bristles: (i) about half a dozen simple barbed bristles tapering to a point; (2) slender multi-articulate compound bristles (Fig. 12 c) with rows of teeth at the head of the shaft ; (3) one or two compound multi-articulate bristles (Fig. 12 d) with smooth, curiously ex- panded, rounded, fist-shaped apexes to the shaft: in these the manner of articulation between shaft and blade is obscure ; (4) slender multi-articulate compound bristles with normal, smooth tops to the shafts. 5 MM Fig. 12. Sigalion ovigerum. a. Head. c. Multi-articulate bristle with toothed shaft. b. Pinnate papilla from elytron. d. Special multi-articulate bristle. Below the neuropodial aciculum the bristles are all multi-articulate with very long and delicate blades ending like the rest in a beak-like apex. They are thinner than the supra-acicular compound bristles, and the individual articles are longer. The tops of the shafts are smooth. The ventral cirri are subulate and reach to the end of the feet. SIGALIONIDAE I OS Remarks. I know no other Sigalion in which there are no compound bristles with single-jointed blades. Genus Leanira, Kinberg Body long and slender with numerous segments. There is a median tentacle which usually has a ceratophore and a pair of ctenidia. The lateral tentacles are fused with the first foot, which carries dorsal and ventral tentacular cirri, a bundle of simple bristles, a cephalic scoop and a prebuccal lamella. There is a pair of long palps. Cirriform gills are present on all segments except a few in the anterior region. The dorsal bristles are simple capillaries with spiral whorls of teeth. The ventral bristles are compound spinigers with canaliculate blades. Leanira quatrefagesi, Kinberg (Fig. 13). Kinberg, 1857, p. 30, pi. ix, fig. 42 a-e. Ehlers, 1901, p. 59, pi. 5, fig. 8. Monro, 1924, p. 46. Occurrence. St. WS 214 (i); WS 770 (i). Specific characters. Body long, rectangular in section. Neither of the specimens is complete, but that from St. WS 770 is large and measures 185 mm. by 3 mm. without the feet for 140 chaetigers. The head (Fig. 13) is rectangular, broader than long, and towards the hinder border there is some diffuse black pigment, which may be ocular pigment. Beginning in the hinder third of the prostomium there is a kind of raised ridge running forward to the end of the head and continuous with the median tentacle. This I take to be the ceratophore of the median tentacle fused with the head. The median tentacle is spindle-shaped, shorter than the head, and tapers to a fine tip. I see no ctenidia in con- nection with the median tentacle. The lateral tentacles arise from the dorsal surface of the first foot at the junction between the foot and the head and are similar in form to the median. The first foot also carries a dorsal cirrus, reaching back to about the 4th chaetiger, a ventral cirrus about one- third as long as the dorsal, and a bundle of long capillary bristles. The palps are very long, reaching back well beyond the loth chaetiger. At the base of the first pair of feet there is a cephalic scoop and a prebuccal lamella. The elytra are tinged with orange-brown. They are without cilia or papillae, being quite smooth. The cirriform gills appear as small papillae on about the 20th chaetiger and reach their full development about 10 chaetigers further back. The feet are figured both by Kinberg and by Ehlers. Neither of these authors record parapodial ctenidia, and I failed to find any on the material obtained in the Straits of Fig. 13. Leanira quatrefagesi. Head from above. io6 DISCOVERY REPORTS Magellan by the Alert Expedition (Monro, loc. cif., 1924). In the smaller specimen from St. WS 214 they are again not apparent, but in the large fragment from St WS 770 two ctenidia are clearly visible on the upper surface of the feet. Stylodes are numerous, there being about half a dozen in each ramus. The dorsal bristles are of two kinds, the one very fine, long, minutely hispid capil- laries; the other stouter and with spiral whorls of small teeth. It is possible that the former of these two types of bristles is only a more delicate form of the latter, but even under a very high magnification I cannot see whorls of teeth. The ventral bristles are typical of the genus. They are compound, with pointed, canaliculated blades. Remarks. This is the genotype of Kinberg's Leanira. Willey's genus Sthenolepis covers those Leanira which have a median tentacle having a ceratophore and ctenidia ; and Willey wishes to restrict Leanira to those species which lack the ceratophore and tentacular ctenidia. It is noteworthy therefore that Kinberg's genotype probably has a median ceratophore, but whether the prostomial ridge with which the median tentacle is continuous is in fact a ceratophore seems to me to be not yet settled. Anyhow I agree with Horst and Fauvel that Sthenolepis is unjustified. Genus Psammolyce, Kinberg Body long, vermiform. Median tentacle on anterior margin of head. No tentacular ctenidia. Lateral tentacles attached to first foot. A dorsal cirrus on the 3rd chaetiger. Cirriform branchiae on every foot except the first. The back and elytra coated with sand-grains. Dorsal bristles slender, barbed capillaries. Ventral bristles compound falcigers. Psammolyce semiglabra, n.sp. (Fig. i\a-g). Occurrence. St. 936 (New Zealand) (i). Description. The single anterior fragment measures 55 mm. by 7 mm. without the feet for 65 chaetigers. In spirit the dorsal surface is colourless for about the first 20 chaetigers; behind this it is pale brown, the colour deepening from before backwards. The ventral surface is a uniform pale brown except in the neighbourhood of the head. The body is quadrangular in section and somewhat convex dorsally. On the ventral surface there is a deep median longitudinal groove, and on the dorsal surface there is also a slight median longitudinal depression which maybe an artefact. The first three pairs of elytra almost meet in the median line, but over most of the body the elytra cover the deep sides of the body and leave most of the dorsum bare. About the first 30 segments are dorsally almost free from adhesive papillae and sand-grains: further back they are rather sparsely dotted over the dorsum, being most abundant in the region of the apex of the elytra. They increase considerably in density from before backwards. The ventral surface is covered with small, globular papillae. There are no filiform papillae on the ventral surface. The first pair of elytra are lost: they probably concealed the head. The prostomium (Fig. 14 a) is long, soft, and conical and ends in a small median tentacle, about one- third as long as the head. At about its middle and on its lateral edges there are two pairs SIGALIONIDAE 107 of almost invisible eyes lying one beneath the other. A large part of the head is covered by a dorsal fold from the following segment. The first pair of feet carry the lateral tentacles, which are about equal in length to the head, a dorsal cirrus a little longer than the lateral tentacles, a ventral cirrus of about twice this length, and a bundle of long, slender bristles. The palps are about three times as long as the ventral cirrus of the ist •smm 1 •5MM IMM Fig. 14. Psammolyce semiglabra. a. Head from above. e. Unidentate ventral bristle. b. El)rtron from middle of body. /. Shorter bidentate ventral bristle. c. Middle foot from in front. g. Longer bidentate ventral bristle. d. Middle foot from behind. chaetiger. The 2nd chaetiger bears the first pair of elytra, a short branchia and a long ventral cirrus. The 3rd chaetiger carries a pair of dorsal cirri long enough to reach to the tips of the bristles. In these cirri the style is only about half as long as the ceratophore, at the base of which there is a short branchia. io8 DISCOVERY REPORTS The elytra are borne on segments 2, 4, 5, 7, etc., to 27, after which they occur on every segment. They vary somewhat in shape according to their position in the body. In front they are reniform, but except for the first few segments they are semi-circular with the straight edge to the front (Fig. 14 b). At their most dorsal point there is a single club-shaped papillated process. This club-shaped process is absent from the first few scales and is not fully developed before about the 20th chaetiger. Except on its front face the scale is fringed with long adhesive papillae. These are not continuous on the lower border of the scale. This lower border has a wavy edge, and the papillae are gathered into bunches or tufts attached to the tops of the waves and from the emarginate parts of the border papillae are absent. In addition to the marginal papillae the elytra carry numerous adhesive papillae and numerous small grape-like tubercles. The umbilicus is oval. Below each elytron there are a gill and a ctenidium. Above the dorsal ramus of the foot there is a semicircular membrane which may be a second ctenidium, but I have not succeeded in seeing any cilia. The foot is triangular (Fig. 14 c-d). The dorsal ramus consists of an enormous fan- shaped bundle of slender barbed bristles situated on the hinder face of the neuropod. The dorsal bristles project almost as far below the ventral ramus as above it. The ventral bristles are variable, especially in regard to the presence or absence of a second tooth on the blades. The most usual arrangement is as follows. At the top of the neuropod there are usually about half a dozen bristles with rather long and slender blades with unidentate tips (Fig. 14 e). Below these the bristles are bidentate and the blades become rather stouter and shorter from above downwards as far as a short distance below the aciculum (Fig. 14/). Below this and standing a little apart from the rest there is a bundle of much more slender bristles with long and delicate bidentate blades (Fig. 14^). The articulation is heterogomph throughout, and the tops of the shafts are heavily denticulated. At the base of the chaeta-sac there is a tuft of filiform processes that I take to be stylodes. The tapering ventral cirrus reaches to the end of the foot. Remarks. This species is related to P. antipoda (Schmarda), from which it differs in having no filiform papillae on the ventral surface and also in having the lower margin of the elytra wavy with the papillae gathered in tufts. In fact there is to my knowledge no other Psammolyce which has this combination of characters. Genus Sthenelais, Kinberg A long median tentacle with ctenidia at the base. Lateral tentacles reduced to small papillae on the dorsal surface of the first foot. Branchiae on every foot except the first few. Dorsal bristles spinous capillaries. Ventral bristles compound falcigers and also a few simple barbed chaetae. Sthenelais limicola (Ehlers), var. novae-zealandiae, var.nov. (Fig. 15). Occurrence. St. 936, New Zealand (3). PHYLLODOCIDAE 109 Varietal characters. The only difference that I can find between these specimens and a typical example of this species from northern European waters is that in the elytra from the middle of the body, on each side of the opening caused by the indentation of the scale on its outer border, there are two or three fiUform papillae (Fig. 15) not present in the northern form. More- over, the bract above the neuropodium is more flattened and tongue-shaped than in the stem-form and lacks the terminal stylode. Otherwise, in the shape of the head, in the form of the minute elytral vesicles and in the structure of the feet and bristles these specimens are indistinguish- able from the European form. I have not been able to compare the scales from the anterior region of the body, for these are all lost from the present specimens, which are in rather poor condition. They are all incomplete and the largest measures 20 mm. by i mm. without the feet for about 80 chaetigers. ■IMM I Fig. 15. Sthenelais limicola, var. novae-zealandiae. Elytron. 3- Family PHYLLODOCIDAE Tentacular cirri three pairs Tentacular cirri four pairs Tentacular cirri two pairs With a median tentacle ... Without a median tentacle First two tentacular segments fused. Body short and broad... First two tentacular segments separate. Body long and narrow Elongate, benthic animals Short, pelagic animals ... With foliaceous dorsal and ventral cirri With cylindrical dorsal and ventral cirri Mystides ... 2 ... 4 Eulalia ... 3 Genetyllis Phyllodoce Eteone ... 5 Lopadorhynchus Pelagobia Genus Phyllodoce, Savigny Body elongate with numerous segments. Head cordiform or oval. There are four tentacles. Papillae at base of proboscis either diffuse or arranged in longitudinal rows. Four pairs of tentacular cirri distributed over three more or less distinct segments. Feet uniramous or sesquiramous. Dorsal and ventral cirri foliaceous, variable in shape. Bristles compound. 3- Dorsal cirri ear-shaped or semicircular... Dorsal cirri oval, lanceolate or subrectangular With eyes and with an aciculum in the dorsal cirrophore Without eyes and without an aciculum in the dorsal cirrophore With a pedal lobe and bristles in the third tentacular segment Without a pedal lobe and bristles in the third tentacular segment 2 3 ... P. longipes ... P. bower si P. patagonica P. madeirensis 7-2 no DISCOVERY REPORTS Phyllodoce longipes, Kinberg. Monro, 1930, p. 73, fig. 21, with synonymy. Occurrence. St. WS 239 (i); WS 841 (i). Specific characters. The head is cordiform and has a nuchal papilla. The tenta- cular formula is i+JS \-B-^rj. The feet are sesquiramous and have a small ai aJM ^ aciculum running up the broad dorsal cirrophore. The dorsal cirrus is more or less ear-shaped with the long axis at right angles to the cirrophore. The pedal lobe has a small digitiform process at its end, and the ventral cirri are large, foliaceous and pointed. The papillation of the proboscis is unknow^n. The most complete specimen of this species that I have seen is that described by me {loc. at.) in 1930. This measured no mm. by 7 mm. including the feet for i40chaetigers. Phyllodoce patagonica (Kinberg). Atiaitides patagonica, Bergstrom, 19 14, p. 147, fig. 46 a-c, with synonymy. Occurrence. St. WS 177 (i); WS 239 (2 juv.); WS 274 (i); WS 771 (6); WS 784 (i); WS 785 (i). Specific characters. A narrow, elongate species with up to about 200 chaetigers. A nuchal papilla is present and the tentacular formula is i + O \- B ~^r- In the first a\ aN few segments the dorsal cirri are broadly lanceolate, but in the middle and hinder region they are more or less rectangular. The ventral cirrus tapers to a point. The papillae of the proboscis are arranged in 13 rows of which one is mid-dorsal and 12 are lateral. There are about five papillae in the mid-dorsal row and nine or ten in the lateral rows. Phyllodoce madeirensis, Langerhans. Fauvel, 1923, p. 150, fig. 53 d-h. Benham, 1927, p. 74. Occurrence. St. 934, New Zealand (2). Specific characters. The head is cordiform and there is a nuchal papilla. The tentacular formula is i+O \-0 -^,. The dorsal cirri are very variable : they may be oval, lanceolate, asymmetrical, straight or incurved, or subrhomboidal. The ventral cirrus is oval and tapers to a point. The proboscis carries on each side six lateral rows of 10-12 papillae and usually also a median row of four to six papillae. In the present specimens the dorsal cirri are more or less symmetrically lanceolate. There is no median row of papillae. Remarks. I cannot agree that this species is the same as P. patagonica, Kinberg. In the first place in P. patagonica there is a small but distinctly developed pedal lobe with bristles in the third tentacular segment, which is wholly absent in madeirensis; and in the middle and hinder regions the dorsal cirri have a constant and characteristic subrectangular form. PHYLLODOCIDAE m Phyllodoce bowersi, Benham. Benham, 1927, p. 77, pi. A, figs. 27-31. Monro, 1930, p. 72. Occurrence. St. WS 215 (i). Specific characters. The present specimen is small, measuring only 14 mm. by I mm. for 63 chaetigers, whereas Benham 's type measured 60 mm. by 2-5 mm. without the feet for 120 chaetigers. In spirit this is a cream-coloured species. The head is rounded, as broad as long, and there are no eyes. A nuchal papilla is present. The tentacular segments are distinct, and though I have not sufficient material to determine the tentacular formula I can report the presence of bristles in the second and third tenta- cular segments. The dorsal cirri are adherent, transversely elliptical in form and with an outer edge which is a continuous curve. The ventral cirri are large, oval and reach nearly to the end of the lobe. Bristles with long appendix and articular cup with one side produced into a long tapering process. Pharyngeal papillae arranged in six or seven regular longitudinal rows. Genus Eulalia, Oersted Body elongate with numerous segments. Head oval or conical. There are five tentacles. Proboscis smooth or diffusely papillated. Four pairs of tentacular cirri dis- tributed over three segments. Feet uniramous. Pedal cirri foliaceous, cordiform, oval or lanceolate. Bristles compound. Dorsal cirri thin and lanceolate; proboscis papillated E. magalhaensis Dorsal cirri stout and ovate. Proboscis smooth ... ... ... ... ... E. picta Eulalia magalhaensis, Kinberg. Steggoa magalhaensis, Bergstrom, 1914, p. 129, with synonymy. Eulalia magalhaensis, Monro, 1930, p. 75. Occurrence. St. 53 (i). Specific characters. A rather elongate species with up to 350 chaetigers. There are two pairs of eyes behind the median tentacle. The proboscis is diffusely papillated. The tentacular formula is i + O \-B ^ry, and the ventral cirrus of the second tenta- ai aJ\ cular segment is foliaceous. The dorsal cirri are elongate and lanceolate, about three times as long as broad. The ventral cirri are oval and with blunt ends. I find that Eulalia anomalochaeta, mihi (1930), is the same as Pterocirrus hunteri, Benham (1921). Eulalia picta, Kinberg. Natalia picta, Bergstrom, 1914, p. 127, text-fig. 34, with synonymy. Occurrence. St. WS 804 (2). Specific characters. A slender, elongate species with (in spirit) a dark green body and brov^Ti cirri. The tentacular formula is {fide Bergstrom) i + B \-B-^, and the ventral cirrus of the second tentacular segment is stout and ventrally asymmetrical. The proboscis is smooth. The dorsal cirri are thick and ovate. The bristles have the head of the shaft denticulated and short end-pieces. 112 DISCOVERY REPORTS Genus Genetyllis, Malmgren, char, emend. Bergstrom. Body short and rather broad. No median tentacle. Four pairs of tentacular cirri, either slender or spindle-shaped. The first tentacular segment is free of the head, and the first two tentacular segments are fused into a single structure carrying three pairs of tentacular cirri and a pedal lobe with bristles but no acicula. Dorsal cirri large and foliaceous ; the ventral cirri are also foliaceous and curve ventrally upwards and down- wards from behind the foot. The anal cirri are large and cylindrical. Genetyllis polyphylla (Ehlers). Bergstrom, 1914, p. 161, fig. 55, with synonymy. Occurrence. St. WS 27 (4); WS 123 (2); MS 64 (7). Specific characters. All the present specimens are damaged and the largest measures 16 mm. by 2 mm. for 54 chaetigers. The colour is reddish yellow with orange cirri. The head is rounded and longer than broad. There is a pair of large eyes. No nuchal papilla. The tentacular formula is i + 5 \- B -^j {fide Bergstrom). The dorsal cirri are very asymmetrically cordiform with the apex pointing towards the dorsal median line. The ventral cirri are oval with the long axis pointing upwards and down- wards. The bristles have short, rather broad end-pieces. Genus Eteone, Savigny Body elongate with numerous segments. Head triangular, truncated in front. Four tentacles, and on the first segment two pairs of tentacular cirri. The second segment has a foliaceous ventral cirrus, but no dorsal cirrus: it may have a pedal lobe with or without bristles. Dorsal and ventral cirri foliaceous. Proboscis smooth or carrying lateral rows of papillae. Eteone sculpta, Ehlers. Bergstrom, 1914, p. 195, text-fig. 73 a, b, with synonymy. Occurrence. St. WS 755 (i). Specific characters. The present specimen measures 19 mm. by 2 mm. for about 160 chaetigers. There are purple transverse segmental bands across the back and the cirri are yellow. The head is more or less rectangular, slightly longer than broad. The proboscis is smooth. The tentacular formula is O — . The second segment has bristles, but no dorsal cirrus. The dorsal cirri are broadly oval, thick and at their base about as broad as long. In the bristles one side of the articular cup is produced into a long finely denticulated process. Remarks. Augener (1932 a, p. 26) is, I think, justified in making E. rubella, Ehlers, a synonym of this species. Genus Lopadorhynchus, Grube Body short, head broader than long, two eyes and four tentacles. Two pairs of tentacular cirri and sometimes a third rudimentary pair on a single segment, which is PHYLLODOCIDAE 113 fused with the head and devoid of bristles. The first few chaetigers carry simple bristles with hooked ends, the remainder carry compound bristles and usually a few simple bristles. The pedal lobe is rounded and the dorsal and ventral cirri are foliaceous. First two chaetigers very much larger than the rest and carrying a number of stout hooks surrounded by a sinuous collar ... ... ... ... • • • ■ ■ • • • • ^- uncinatus First two chaetigers slightly larger than the rest and carrying small hooks and no collar L. krohnii, var. simplex Lopadorhynchus krohnii (Claparede), var simplex, Monro. Monro, 1930, p. 79, fig. 23 a, b. Occurrence. St. 702 (i). Varietal characters. Up to about lo mm. in length. In the younger specimens eyes are not visible, but in adults a pair is present as in the stem-form. The first two chaetigers have rather stout, cylindrical pedal lobes, about half a dozen simple bristles with hook-like ends, and no ventral cirri. The pedal lobes of the remaining chaetigers are flatter and have oval dorsal and ventral cirri. The bristles are arranged fan-wise and consist of paddle-shaped compound chaetae with oval blades denticulated on one edge. There are no simple hooks behind the second chaetiger. The absence of simple hooks from all chaetigers except the first two is the only substantial difference between the variety and the stem-form. The latter is described by Fauvel (1923, p. 185, fig. 68 a-d) and by Bergstrom (1914, p. 180, fig. 68 a, b). The present specimen is a young one, measuring only 5 mm. by i mm. without the feet for 20 chaetigers. Lopadorhynchus uncinatus, Fauvel. Fauvel, 1923, p. 184, fig. 67 a-g. Occurrence. St. 714 (i). Specific characters. The specimen measures 18 mm. by 2 mm. without the feet for 32 chaetigers. The head is broader than long and there are four tentacles. There are two pairs of tentacular cirri reaching back to the 2nd chaetiger and a third rudimentary pair situated at the base of the second pair. There is a pair of large eyes. The first two pairs of feet are very much larger than the rest, being very stout and rounded in section. They carry a number of large dark brown hooks which are surrounded by a delicate sinuous membrane. There is a small dorsal cirrus, but the ventral cirrus is absent. These first two chaetigers are clearly separated by a constriction from the rest of the body. The normal feet point backwards and consist of a lanceolate pedal lobe with a projecting aciculum, a large rounded vertical lamella, a stout, lanceolate dorsal cirrus and a conical ventral cirrus. The bristles have a fan-shaped arrangement and are all compound except in the 3rd chaetiger which has in addition a few simple bristles. The shafts end in a point, below which there is a kind of notch with which the broad paddle- shaped blade is articulated. This blade has a few delicate denticulations on one edge. Remarks. This specimen agrees in detail with Fauvel 's description. I believe this to be the first record of this species from the South Atlantic. 114 DISCOVERY REPORTS Genus Pelagobia Four tentacles. Two pairs of tentacular cirri, a dorsal and a ventral, borne by a single segment which also carries bristles. The dorsal cirrus of the 2nd chaetiger is reduced or absent. The feet are uniramous and carry slender, cylindrical dorsal and ventral cirri. The feet have a single aciculum and compound bristles with denticulated blades. The proboscis is smooth. Pelagobia longicirrata, Greeff. Fauvel, 1923, p. 192, fig. 72 a-e. Augener, 1929, p. 291, with synonymy. Occurrence. St. 17 (6); 41 a (3); 41 b (6); 41 d (5); 41 e (4); WS 40 (4); WS 44, 1000-75001. (i), and 750-500 m. (5); WS 555 (3). Specific characters. A small species measuring up to about 12 mm. in length with 25 chaetigers. The head is conical and truncated in front. There is a single pair of eyes. Four small thread-like tentacles. The two pairs of tentacular cirri are long and subulate and the first segment has a small pedal lobe with a few bristles. There is no dorsal cirrus on the 2nd chaetiger. The following segments carry slender tapering dorsal and ventral cirri. The pedal lobe is conical and has an aciculum and numerous long, delicate compound bristles. These have the end-pieces denticulated, and often the head of the shaft also. The end- pieces have a delicate border on the side opposite the teeth. Remarks. I have examined the bristles of a random sample of these specimens and in none can I find any denticulation of the head of the chaetal shafts. In this they differ from the northern representatives of the species. P. longicirrata is a very common species in the southern cold-water plankton. Augener found it in the Weddell Sea. Genus Mystides, Theel Small, elongate animals. The head is rounded. There is a pair of eyes and four tentacles. No median tentacle and no occipital papilla. Three pairs of tentacular cirri. The formula is i + 5 - . The feet are uniramous. Dorsal and ventral cirri foliaceous or more or less globular. Bristles compound. Proboscis with soft papillae. Mystides notialis, Ehlers. Ehlers, 1913, p. 457, pi. xxix, figs. 1-4. Occurrence. St. WS 226 (i). Specific characters. The present specimen measures 13 mm. by 0-5 mm. for 52 chaetigers. It corresponds exactly to Ehlers' account. There is a pair of dark eyes and four very slender, transparent tentacles. The first segment is achaetous and carries a pair of tentacular cirri ; the second segment has bristles and two pairs of tentacular cirri, a dorsal and a ventral. All the tentacular cirri are swollen towards the base and are apically slender and filiform. The normal dorsal and ventral cirri are more or less globular. The bristles are compound and have a bifid articulation. The head of the shaft ALCIOPIDAE IIS ... «< ... 4 Alciopa ... 3 Greeffia Vanadis Torrea ... 5 Callizonella . Callizona is scarcely, if at all, denticulated. The end-piece is very slender and rather short. There is a pair of globular anal cirri. Remarks. I find it impossible to decide whether this species is the same as Theel's M. borealis without a direct comparison, and unfortunately Theel's species is not represented in the museum collection. The present species has not been seen since its original description. Family ALCIOPIDAE 1. Bristles all of one kind ... ... ... ... ... ... ... ... ... ... 2 Bristles of more than one kind ... 2. Bristles all long simple swimming chaetae Bristles all long compound swimming chaetae 3. The foot has two cirriform terminal appendages The foot has one cirriform terminal appendage The foot has no cirriform terminal appendage 4. Both long swimming chaetae and short acicular bristles present 5. Long swimming bristles simple... Long swimming bristles compound Genus Alciopa, Audouin and Milne-Edwards Body elongate and transparent. The head ends in a pair of enormous eyes. Five tentacles, of which the median is very small. Proboscis with a crown of papillae, of which the two lateral papillae are only slightly longer than the rest. Three pairs of tentacular cirri, followed by about three chaetigers with small undeveloped feet. In the female the dorsal cirri of the first pair of feet are converted into seminal pouches. Dorsal and ventral cirri are foliaceous. The pedal lobe has no terminal appendage. The bristles are all long, delicate, capillary and simple. Prominent segmental glands are present. Alciopa cantrainii (Delle Chiaje). Fauvel, 1923, p. 203, fig. 76 a-c. Occurrence. St. 707 (i). Specific characters. The characters of this species are those of the genus. Genus Vanadis, Claparede Long, cylindrical, transparent animals. The head ends in a pair of enormous eyes. The proboscis has a crown of long papillae. There are three to five pairs of tentacular cirri. In the female a few of the anterior dorsal cirri are converted into globular seminal pouches. The pedal lobe ends in a single cirriform process. The dorsal and ventral cirri are foliaceous. The bristles are all of a similar kind, very long, fine, transparent and compound. Prominent, deeply pigmented segmental glands are present. 1 . Body colour purplish brown ... ... ... ... ... ... ... V. violacea Body colourless, transparent ... ... ... ... ... ... ... ... ... 2 2. With three pairs of tentacular cirri ... ... ... ... ... ... V.formosa With four pairs of tentacular cirri ... ... ... ... ... ... ... ... 3 3. The proboscis carries a pair of long, lateral cirriform processes ... ... V . crystallina The proboscis is devoid of cirriform processes ... ... ... ... V. antarctica ii6 DISCOVERY REPORTS Vanadis antarctica (Mcintosh). Alicopa antarctica, Mcintosh, 1885, p. 175, pi. xxviii, figs. 2, 3, 4; pi. xxxii, fig. 12. Vanadis antarctica, Benham, 1921, p. 58, pi. viii, figs. 61-63, with synonymy. Occurrence. St. 334 (i); 362 (i); 373 (i); 461c (i); 527 (3); 533 (i); 569 (i); 579 (i); 590 (i); 591 (i); 619 (i); 1148 (i); WS 200 (i); WS 408 (i); WS 411 (i); WS 537 (i); WS 550 (i); WS 551 (2);WS552(i). Specific characters. Four pairs of tentacular cirri, followed by about a dozen seg- ments in which the feet are reduced in size, being smaller than the normal feet of the rest of the body. This area with reduced feet is nearly always pigmented and gives the animal a characteristic appearance of having a pigmented neck. In the female the first two feet have the dorsal cirri converted into seminal pouches. The mouth of the pro- boscis has only rather short papillae and is devoid of the pair of lateral tentacle-like processes present in V. formosa. The pigmented neck is followed by a short unpig- mented area of about three segments, and this again by about three pigmented seg- ments. Behind this the pigment is usually confined to the lateral glands and does not spread over the dorsum. The arrangement of the glands and their accompanying pig- ment is irregular up to about 30th-50th segment, behind which there is a pair of glands in each foot. There are the usual foliaceous dorsal and ventral cirri on the feet, and a lanceolate pedal lobe ending in a projecting aciculum and a small cirriform process. The bristles are very long and delicate and the articulation between shaft and blade very diflicult to see. This is a fairly large species with a breadth of about 5 mm. My most complete fragments measure about 150 mm. in length. Vanadis formosa, Claparede. Fauvel, 1923, p. 205, fig. 77 a-c. Occurrence. St. 404 (2); 407 (2); 701 (i); 704 (i); 705 (i); 713 (2); 714 (i and 2 juv.). Specific characters. Four small digitiform lateral tentacles and a median tentacle. Three pairs of tentacular cirri. The mouth of the proboscis has, in addition to a number of rather short papillae, on each side a long tentacle-like process. In the female the dorsal cirri of the first two feet are converted into large globular seminal pouches. In the male the first foot is reduced to a dorsal and ventral cirrus. The remaining feet have a lanceolate pedal lobe with a long, projecting aciculum, foliaceous dorsal and ventral cirri and numerous very fine compound bristles. The pedal lobe ends in a small cirri- form process. Every chaetiger, except the first three, carries a large dark brown gland. Remarks. In the colder waters of the southern hemisphere the place of this species is taken by V. antarctica (Mcintosh). Vanadis crystallina, Greeff . Fauvel, 1923, p. 206, fig. 77 d, e. Occurrence. St. 708 (2); 709 (i). Specific characters. A small thread-like species with up to about 150 segments. The proboscis carries a pair of long lateral processes. Four pairs of tentacular cirri. ALCIOPIDAE 117 The fourth tentacular cirrus is foliaceous and much larger than the two preceding tentacular cirri. In the female the first foot only has the dorsal cirrus converted into a seminal pouch. About the first five feet behind the tentacular region are small and un- developed and usually lack bristles. A normal foot consists of a lanceolate pedal lobe ending in a small cirriform process, a lanceolate dorsal cirrus and an oval ventral cirrus. The bristles are all very long, fine, compound capillaries. Segmental glands are found on every foot behind the second. Vanadis violacea, Apstein. Apstein, 1893, p. 143, pi. v, figs. 1-4. Occurrence. St. 413 (i); 419 (i). Specific characters. These specimens are fragmentary and the largest has a breadth of 3 mm. without the feet. The body has a less fragile and transparent appear- ance than that of most Alciopids, and recalls rather that of a bottom-living Phyllodocid. The colour is uniform purplish brown. The head extends beyond the eyes. The pro- boscis has a crown of small papillae and is devoid of lateral tentacle-like processes. There are three short stout pairs of tentacular cirri on separate segments. Of these the middle pair is twice as long as the other two in the present specimens. The fourth segment carries a foliaceous cirrus, below which is a minute process which I take to be a rudimentary pedal lobe. Whether this foliaceous cirrus of the fourth segment is to be regarded as a tentacular cirrus or as the first dorsal cirrus of the body is not clear. Apstein gives four pairs of tentacular cirri in his original description and three pairs in his Plankton Expedition paper of 1900. My own inclination is to give only three pairs of tentacular cirri and to treat the foliaceous cirrus of the fourth segment as part of the body region. The bristles begin with the fifth segment, in which the ventral cirrus is somewhat reduced. The feet are stouter than is usual in the Alciopids. Normally there is a lanceolate pedal lobe ending in a small cirriform appendage, a large cordiform dorsal cirrus, and a broadly lanceolate ventral cirrus. The bristles consist of very numerous, long, com- pound capillaries. There are segmental glands both above and below the feet. Remarks. The colour and general aspect of solidity of this species are characteristic. The locality of the type specimen described by Apstein is not known. Genus Greeffia, Mcintosh Body rather massive for an Alciopid and tapered posteriorly. The head does not extend beyond the eyes. There are five tentacles and a pair of enormous eyes. The proboscis carries a pair of long cirriform processes. There are three or four pairs of tentacular cirri. There are no undeveloped parapodia in the anterior region. The feet end in a pair of small, cirriform processes. The dorsal and ventral cirri are foliaceous. There are prominent segmental glands both dorsally and ventrally. The bristles are all slender, compound, capillary, swimming bristles. 8-2 ii8 DISCOVERY REPORTS Greeffia oahuensis, Mcintosh. Mcintosh, 1885, p. 182, pi. xxviii, figs. 5-7; pi. x.xxii, fig. 11; pi. xva, fig. 4. Monro, 1930, p. 82, fig. 25. Occurrence. St. 413 (i). Specific characters. The largest example of this species known to me is a fragment measuring 39 mm. by 4 mm. without the feet for 48 chaetigers. The present specimen, also fragmentary, measures 13 mm. by 4 mm. without the feet for 30 chaetigers. In spirit the colour of the body is pale brown ; the dorsal segmental glands are black and the ventral colourless; the pedal cirri are white. There are three pairs of tentacular cirri. The proboscis carries a pair of long cirriform lateral processes. The foot ends In a pair of subequal cirriform appendages. The dorsal cirri are large, imbricating and cordiform. The ventral are also large, and rounded. The prominent cushion-like segmental glands, which are present both above and below the feet, are not apparent in the first few seg- ments. The chaetae are all slender, compound, capillary swimming bristles. Remarks. This species is distinguished from G. celox (Greeff) by the possession of three instead of four pairs of tentacular cirri. Genus Callizona, Greeff Body elongate, with numerous segments. The head extends beyond the eyes. There are five tentacles and a pair of enormous eyes. The proboscis has a crown of small papillae and is devoid of lateral tentacle-like processes. There are five pairs of tentacular cirri. There are no undeveloped parapodia in the anterior region. The feet end in a single cirriform process. The dorsal and ventral cirri are foliaceous. The bristles are of two kinds: (i) short acicular bristles which may be simple or compound; (2) long, capillary compound swimming bristles. Segmental glands are present. Callizona angelini (Kinberg). Fauvel, 1923, p. 215, fig. 81 d-i. Occurrence. St. 405 (i). Specific characters. A rather large species recorded up to 120 mm. in length. The present anterior fragment measures 2 mm. in breadth without the feet. The five pairs of tentacular cirri are massive and arranged as follows : i-\ 1- . The imbricating dorsal cirri are cordiform anteriorly and more lanceolate farther back. The ventral cirri are oval or lanceolate. In the anterior feet the bristles are mostly rather short, stout and with a minute, slender end-piece; and with them there are a few slender spinigers. These are the extreme types and there is a tendency for one kind to grade into the other. In the middle feet there are numerous, long, slender, compound swimming bristles and below these a single compound acicular bristle with a minute end-piece. Segmental glands are present from about the tenth foot backwards. ALCIOPIDAE 119 Genus Callizonella, Apstein Body slender and elongate. The head, which extends beyond the two great eyes, carries five tentacles. The proboscis has a crown of small papillae. There are five pairs of tentacular cirri. There are no undeveloped feet behind the tentacular segments. The pedal lobe ends in a small cirriform appendage. The dorsal and ventral cirri are foliaceous. The bristles are of two kinds; long, delicate and capillary bristles and short and moderately stout bristles. The long capillary bristles are all simple, and the short acicular bristles may be either simple or compound. Segmental glands are present. Callizonella bongraini (Gravier). Callizona bongraini, Gravier, 191 1, p. 70, pi. iv, figs. 39-43. Benham, 1929, p. 189, pi. i, figs. 11-12. Callizonella bongraini, Augener, 1929, p. 294, fig. 2 a-g. Occurrence. St. WS 555 (3). Specific characters. A small species measuring about i mm. in breadth and usually about 20 mm. in length. There are five pairs of tentacular cirri arranged as follows: iH h - . The ventral cirrus of the third tentacular segment is foliaceous. The feet end II in a small, inconspicuous terminal process, and the dorsal and ventral cirri are foliaceous, more or less ovate. In a normal foot from the mid-body there are about a dozen long, fine, simple capillary bristles and below these a single short, acicular bristle. In the first few feet the bristles are all short and compound. They are of two kinds, and their differences are analogous to the differences between the falcigers and spinigers of the Nereids, to which bristles they have a general resemblance. Whether the simple acicular bristles in the normal feet are compound bristles that have lost their end-pieces it is im- possible to be certain, but I do not believe this to be the case. Segmental glands are present on every foot behind about the tenth. Genus Torrea, Quatrefages Body cylindrical, transparent. The head does not extend beyond the eyes. Five tentacles, the median being reduced to a small tubercle. A pair of enormous spherical eyes. There are three pairs of tentacular cirri and the first two chaetigers are rudi- mentary. In the female their dorsal cirri are transformed into seminal pouches. The proboscis has a crown of papillae and a pair of lateral cirriform processes. The feet have foliaceous dorsal and ventral cirri. The pedal lobe is without any terminal appendage. The bristles are all long, compound, capillary swimming bristles. Segmental glands are present. Torrea Candida (Delle Chiaje). Asterope Candida, Fauvel, 1923, p. 202, fig. 75 a-d. Occurrence. St. 714 (i). Specific characters. The characters of this species are those of the genus. The present specimen is a small anterior fragment measuring 8 mm. by i mm. for 24 chaetigers. I20 DISCOVERY REPORTS Family TOMOPTERIDAE Genus Tomopteris, Eschscholtz The parapodial trunks, i.e. the dorsal and ventral divisions of the feet, are bordered all round by membranous wings or pinnules. In Tomopteris, sensu stricto, a tail is almost alvi^ays absent, and also, in the adult, the first pair of chaetigerous appendages. Hyaline glands are usually present; and there are no rosettes. There are no hyaline glands Hyaline glands are present on the ventral pinnules ... Chromophil gland on a level with the tip of the ventral trunk Chromophil gland below the ventral trunk A large species with the gonad in both rami of the feet A small species with the gonad in the dorsal ramus only ... T. cavallii 2 T. septentrionalis 3 T. carpenteri T. planktonis Tomopteris (Tomopteris) carpenteri, Quatrefages. Augener, 1929, p. 304. Occurrence. St. 124 (i); 128 (i); 133 (6); 136 (2); 138 (3); 139 (6); 160 (i); 374 (16); 459 (2 juv.); 460 (i); 527 (24); 567 (16); 1148 (3); WS 22 (2); WS 26 (3); WS 35 (5); WS 38 (i); WS 39 (I); WS 45 (I); WS 53 (I); WS 53B (i); WS 54 (i); WS 55 (i); WS 536 (5); WS 541 (2); WS 544 (3); WS 545 (i); WS 547 (i); WS 548 (i); WS 549 (2). Specific characters. A large species measuring up to 70 mm. in length for about 35 pairs of feet. The prostomium has no median notch. The neck is short and broad. The cerebral ganglion is transversely elongated and slightly bilobed. There is a single pair of black eyes, often invisible in the adult. The anterior pair of chaetigerous ap- pendages is absent. The second pair is globular at the base and may reach back as far as three-fourths of the length of the body. The feet are conical, and the pinnules oval and rather long. The latter have their origin a short distance before the separation of the foot into its constituent rami. From the third foot backwards there is a conspicuous hyaline gland on the pinnule of the ventral ramus a little above and beyond the apex of the pedal trunk. From the fourth foot there is a very large chromophil gland lying below the apex of the pedal trunk in the ventral ramus. Genital products are found in both rami of the foot. Remarks. It is noteworthy that out of all these specimens there is not one that I recognized as a female. I do not claim to have "sexed" every specimen carefully, but in the adult examples the foot usually contains a flocculent white substance which penetrates into both trunks after the division of the foot into its two rami. This sub- stance appears to be sperm, and I have never seen any ova in the feet. In other species, e.g. in septentrionalis, ova are very conspicuous, and I think I should have seen them if they had been present in carpenteri. Augener records T. platiktotiis from the Weddell Sea and suggests that previous authors have confused it with carpenteri, to which in the arrangement of the pedal glands it is very similar. It is distinguishable by its very much smaller size — it reaches TOMOPTERIDAE a length of only about lo mm. for i8 pairs of feet — by the fact that in carpenteri the pinnules are wrinkled and continued much farther up the feet, and by the presence of genital products only in the dorsal ramus of the foot. In practice it is not easy to dis- tinguish a young specimen of carpenteri from planklonis, of which I have not seen an example in the present collection from south of the Antarctic convergence. Tomopteris (Tomopteris) planktonis, Apstein. Fauvel, 1923, p. 224, fig. 84 9. H; xix, fig. 4- . o -J A. Agassizii, H. L. Clark, 1917. Hawaiian Echini. Echinoneidae. . .Spatangidae, p. 176. ? A. 'Agassizii, H. L. Clark, 1925. Cat. Recent Sea-Urchins Brit Mus., p. 204. St. 55. 16. V. 26. Entrance to Port Stanley, East Falkland Islands, 10-16 m. i young specimen. St. 388. 16. iv. 30. 56° 19' S, 67° 10' W, 121 m. I specimen. St. WS 25. 17. xii. 26. Undine Harbour, South Georgia, 18-27 m. 2 young specimens. 22. X. 34. Weir Creek, Falkland Islands, 3-5 m. 3 young specimens. Although these specimens are quite young, 5-18 mm. long, I have no doubt that the identification is correct. The absence or feeble development of the subanal tube feet affords a marked difference, particularly from A. cavernosiis, in which these tube feet are already well developed in specimens which have just left the marsupium. As for the HEMIASTERIDAE 229 Specimens from Heard Island and the Ross Sea referred by Clark {op. cit., 1925) to A. Agassi zti I think it questionable whether they are actually A. Agassizii, and Clark expressed himself with some reservation on this point. Globiferous pedicellariae were not found in any of these specimens, this type of pedicellaria being thus still unknown in A. Agassizii. This species was hitherto known with certainty only from South Georgia ; the finding of it at the Falkland Islands by the 'Discovery' is thus of considerable interest. Schizaster (Tripylaster) Philippii (Gray) Tripylus Philippii, Gray, 1855. Cat. Recent Echinida Brit. Mus. I, Irregularia, p. 59, pi. v, fig. I. Sdiizaster Pliilippii, A. Agassiz, 1872. Revision of the Echini, p. 612, pi. xxvi, figs. 40, 41. S. {Tripylaster) P/iilippii, Mortensen, 1907. Danish ' Ingolf Exped. Echinoidea, 11, pp. 122, 123. S. (Tripylaster) Plulippii, Mortensen, 1910. Swedish South Polar Exped. Echinoidea, p. 90, pis. xii, figs. 8, lo-ii ; xix, figs. 5, 7, 15-16, 21, 31, 48-49. Tripylaster Pliilippii, H. L. Clark, 1917. Hawaiian Echini. Echinoneidae . . . Spatangidae, p. 177. For other literary references see my work of 1910, loc. cit. St. 51. 4. V. 26. Off Eddystone Rock, East Falkland Islands, 105-115 m. i dead test. St. WS 83. 24. iii. 27. 14 miles S 64° W of George Island, East Falkland Islands, 137 m. I specimen. St. WS 785. 6. xii. 31. 49° 25' S, 62° 37' W, 150 m. i specimen. The specimen from St. WS 83 is of medium size, 60 mm. long, 52 mm. broad. That from St. 51, which is an old, dead test, partly covered with worm tubes, is 77 mm. long, 72 mm. broad, thus distinctly broader than the younger specimen. I see, however, no reason to doubt their identity, such variation in shape being of common occurrence in most species. In the specimen from St. WS 83 I find the pedicellariae of the typical shapes, as described in my work on the Echinoidea of the Swedish South Polar Expedi- tion. The specimen from St. WS 785 is almost totally denuded, only a couple of small tridentate pedicellariae being left. But it is evidently a typical T. Philippii. Amphipneustes Koehleri, Mortensen AnipJiipneitstes Koeiileri, Mortensen, 1910. Swedish South Polar Exped. Echinoidea, p. 94, pis. xi, figs. 2-5, 7, 8, 15, 18; xvii, figs. lo-ii ; xviii, figs. 1-2; xix, figs. 3-4, 8-9, 13-14, 19-20, 23-25, 27. A. Koeiileri, H. L. Clark, 1917. Hawaiian Echini. Echinoneidae. . .Spatangidae, p. 163. A. Koehleri, H. L. Clark, 1925. Cat. Recent Sea-Urchins Brit. Mus., p. 198. St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, no m. Several specimens. St. 42. I. iv. 26. West Cumberland Bay, South Georgia, 120-204 m. 5 specimens. St. 126. 19. xii. 26. 53° 58' 30" S, 37° 08' W, loo(-o) m. i specimen. St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. Several specimens. St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, 155-177 m. i specimen. St. 148. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m. 4 specimens. St. 160. 7. ii. 27. Near Shag Rocks, South Georgia, 177 m. 2 specimens. St. 474. 12. xi. 30. I mile W of Shag Rocks, South Georgia, 199 m. Several specimens. St. WS 33. 21. xii. 26. 54° 59' S, 35° 24' W, 130 m. 4 specimens. 230 DISCOVERY REPORTS None of the present specimens exceed a length of 33 mm. As none of the other specimens known exceed a length of 38 mm. (the type specimen) it is evident that this is about the maximum size of this species, which is, accordingly, noticeably smaller than the other species of the genus Amphipneustes. In my description of this species {op. cit., 1910) I was unable to decide whether young specimens have a fasciole. The large material now at hand, containing specimens of all stages, allows this question to be settled. As was to be expected there is not the slightest trace of a fasciole at any stage of development and growth. In the type specimen {op. cit., p. 96, pi. xi, fig. 18) the labrum was found to reach as far backwards as the middle of the second ambulacral plate. This is a rather unusual con- dition. Generally it reaches only to the middle of the first ambulacral plate ; but it is sometimes more elongate, so as to reach to the middle of the second ambulacral plate. This has nothing to do with age ; the elongate condition of the labrum may be found in quite young specimens and, on the other hand, the short labrum in adult specimens. Sometimes the labrum ends opposite the middle of the first ambulacral plate on one side, of the second on the other side; in such a case the first ambulacral plate on the one side is conspicuously produced posteriorly. The genital pores appear at a size of about 12-15 mm. length. It would appear that the embryos leave the marsupium at a somewhat earlier stage than in A. Lorioli or similis. Amphipneustes Lorioli, Koehler (Plate III, figs. 5-8; Plate IV, fig. 8; Plate IX, fig. 27) Amphipneustes Lorioli, Koehler, 1901. Result. Voyage ' Belgica '. Echinides et Ophiures, p. 12, pis. ii, fig. 12; V, fig. 37; vi, figs. 42, 43. A. Lorioli, Mortensen, 1910. Swedish South Polar Exped. Echinoidea, p. 91, pis. xi, figs. 17, 19; xix, figs. 1-2, 6, 10-12, 17, 22, 26. A. Mortenseni, Koehler, 1912. IP Exped. Antarct. Fran9aise. Echinodermes, p. 176, pi. xv, figs. ii-i7;xvi, figs. 1-5. A. Lorioli, H. L. Clark, 1917. Hawaiian Echini. Echinoneidae . . . Spatangidae, p. 163. A. Lorioli, H. L. Clark, 1925. Cat. Recent Sea-Urchins Brit. Mus., p. 198. St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-335 m. 4 specimens. St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-500 m. 4 specimens. St. 600. 17.1.31. 67° 09' S, 69° 27' W, oft" Adelaide Island, 487-512 m. i specimen. Three of the specimens are females. As shown in Plate III, figs. 5, 6 and Plate IV, fig. 8, there is a conspicuous variation in the development of the petals ; also the general shape of the large specimen (Plate III, fig. 6) is somewhat different from that of the other specimens, it being distinctly higher than the other specimens, and somewhat more flattened on top. It is a curious fact that the marsupia of two of the specimens are quite empty ; the third contains only very few embryos and eggs, three stages of development bemg represented. The largest embryos, which are 3 mm. long and just ready to leave the marsupium, do not show any trace of a fasciole. Pedicellariae have not yet been HEMIASTERIDAE 231 formed; I have found only what appears to be the first rudiment of a globiferous pedicellaria. One of the specimens has the madreporite separated from the right anterior genital plate, as was the case in the type specimen ; in all the other specimens there is no line separating the madreporite from the right anterior genital. All the specimens have three genital pores, the existence of four genital pores in the type specimen being evidently an anomaly. Globiferous pedicellariae, which were hitherto only very imperfectly known in this species, are found well developed. They are of a highly characteristic structure, the valves terminating in a number of quite short teeth (Plate IX, fig. 27), a type not known from any other species of the Amphip7ietistes-Abatus group. As for the other pedi- cellariae I may refer to the description and figures given in my report on the Echinoids of the Swedish South Polar Expedition. I must regard it as almost certain that the A. Mortenseni, Koehler, is not to be main- tained as a species separate from A. Lorioli. The former being known only from a female specimen, the latter from a couple of male specimens, it was of course not easy to point out exactly by which characters the two species difi^ered from each other ; but Koehler thinks the pedicellariae show sufficient diff'erences to prove that the two forms represent diff'erent species. It is remarkable that Koehler could have reached this conclusion. Only tridentate and rostrate pedicellariae were known at that time, and they are abso- lutely identical. In the shape and structure of test the two " species" are exactly alike, apart from the difference due to the different sexes. There only remains the question of the globiferous pedicellariae. In my report on the Echinoids of the Swedish South Polar Expedition (pp. 93-4) I mention having found in A. Lorioli one, very poorly pre- served, globiferous pedicellaria that seemed to be of the same structure as the globi- ferous pedicellariae of A. Koehler i: that is to say with the valves terminating in two long teeth and thus conspicuously different from those here figured. But it is stated that "the valves have in addition to the two large teeth in the point one or two shorter teeth". This might indicate that they are actually different from those of the present specimens. If so, the present specimens are not identical with A. Lorioli, but should be referred to A. Mortenseni, since it is a fact that the globiferous pedicellariae in these forms afford excellent specific characters. But until it is definitely proved that in the forms described respectively as A. Lorioli and A. Mortenseni we have two different types of globiferous pedicellariae, I must believe that both belong to one species, the name of which must be the older, viz. A. Lorioli. I may add that I have had an opportunity of examining the type of A. Mortenseni in the Paris Museum; unfortunately not a single globiferous pedicellaria is left on it. Amphipneustes similis, n.sp. (Plate IV, figs. 1-7; Plate IX, figs. 21-26) St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. i specimen. St. i8o. ii.iii. 27. Schollaert Channel, Palmer Archipelago, 160-330 m. 2 specimens. St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-500 m. 3 specimens. 232 DISCOVERY REPORTS The outline of the test is almost regularly oval; the height varies rather considerably, as seen from the measurements. Particularly the specimen from St. 170 (no. 5 in the table below) is unusually low. The apical system is central, but the greatest height is behind the apical system, the posterior interambulacrum rising as a more or less prominent keel. The periproct is situated on the slightly truncated posterior end, scarcely visible from below. The labrum is rather prominent, the peristomial part of the test rather conspicuously sunken ; the plastron is quite flat. Length mm. Height mm. Width mm. Length of petals mm. Antero- lateral Postero- lateral I 2 3 4 5 72 72 65 61 54 40 44 34 31 25 63 62 57 53 47 28 28 22 24 14 26 25 21 23 14 Q Holotypes Si ^ } Cotypes The petals of the male are rather distinctly sunken, the postero-lateral ones more so than the antero-lateral ones. The frontal ambulacrum distinctly petaloid and almost as much sunken as the antero-lateral petals. In the female the paired petals are transformed into deep marsupia, the width of which varies very conspicuously, as seen in Plate IV, figs. 2, 5. The interambulacra are more or less raised, the posterior one, as stated, forming a keel, which may be quite sharp or more rounded. The sternum is narrow, gently widening towards the posterior end. The madreporite is small. There are usually three genital openings, but the specimen from St. 170 and one of those from St. 182, both females, have only two. In the former specimen the right anterior marsupium is un- developed (Plate IV, fig. 3) ; this specimen accordingly is somewhat abnormal. The spines form a uniform, not very dense covering, among which the numerous, black globiferous pedicellariae in well-preserved specimens stand out very conspicuously on the aboral side. The valves of the globiferous pedicellariae terminate in five long, slender teeth, as in Parapneiistes (Plate IX, figs. 22, 23). The rostrate and tridentate pedicellariae also are of the same type as in Parapneiistes, but there is some variation in the shape of the rostrate form (Plate IX, fig. 26). The colour is a light brownish. The young ones in the same marsupia are in all stages, from the egg to the fully formed young (14 mm.) ready to leave the marsupium. In the female holotype the frontal ambulacrum is somewhat abnormally bent (Plate IV, fig. 2). This is a very perplexing species. It agrees almost completely in every respect with Parapneiistes reductus, Koehler — but there is no trace of fascicles, whereas the fascicles, HEMIASTERIDAE 233 according to Koehler, are particularly well developed in P. reductus, better so than in the species P. cordatus. The fact that the fascioles in Koehler's figure of P. reductus in side view {op. cit., pi. xvi, 14) seem to have been retouched on the photo, whereas figs. 12 and 13 of the same plate, showing the same specimen from above and in end view, do not show any trace of fascioles, might lead one to suggest that some mistake has crept into the representation of the species. Having had an opportunity of re-examining the type specimen in the Paris Museum I can, however, testify to the correctness of Koehler's description and figures. The fasciole has been marked with a black line on the specimen, not on the photo ; but it is quite distinctly seen on the specimen. It might, of course, be suggested that fascioles were originally present in the Discovery specimens, but have been reduced in the course of development. If so, one would expect a fasciole to be developed in the young ones ready to leave the marsupium. There is, however, no trace of a fasciole in the young specimens from the marsupium in the present species. These specimens, it may be added, have the globiferous pedi- cellariae typically developed. The present species forms a most remarkable parallel to P. reductus ; but the character of the fascioles, entirely absent in one and distinctly developed — though very thin — in the other, shows definitely that the two forms belong to two distinct genera. Parapneustes cordatus, Koehler (Plate III, figs. 1-4) Parapneustes cordatus, Koehler, 1912. IP Exped. Antarct. Frangaise. Echinodermes, p. 165, pi. xvi, figs. 15-27. P. cordatus, H. L. Clark, 1917. Hawaiian Echini. Echinoneidae. . . Spatangidae, p. 173. St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-335 ^n- 2 specimens. St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-500 m. 4 specimens. Only a single, apparently male, specimen of this species was taken by the ' Pourquoi- Pas.?'. Koehler's very careful description of this specimen could thus, of course, not be all that was to be desired and in particular the question whether this species is brood-protecting like most other Antarctic Spatangoids had to be left undecided. It is therefore most satisfactory that the 'Discovery' has secured some specimens in fair preservation, which give not only the solution of the question whether the species is brood-protecting, but also other very important information. The specimens, which are all adult, 49-56 mm. in length, agree perfectly with the type specimen in regard to their general shape, only the restriction of the posterior part of the test is sometimes less pronounced. Also in the pedicellariae (particularly those remarkable globiferous pedicellariae with the valves terminating in five long and slender teeth) these specimens agree perfectly with the type. But in regard to the fascioles they differ conspicuously from the type specimen. Whereas this latter had only the peri- petalous fasciole developed, the present specimens also have a latero-subanal fasciole. As a rule the peripetalous fasciole is fairly distinct, but it may be more or less reduced, sometimes more on one side than on the other; the latero-subanal fasciole is generally 5-2 234 DISCOVERY REPORTS distinct on the side of the test ; the subanal part is sometimes quite reduced, sometimes very distinct. There is thus much variation in the development of the fascioles. Both male and female specimens are found in this material, and, as was to be expected, the females have the paired petals transformed into deep marsupia, this species being thus also brood-protecting. The embryos contained in the same marsupium are in all stages of development, from newly laid eggs to fully formed young ones ready to leave the marsupium. The breeding thus appears to go on continuously, i.e. so long as the breeding season lasts ; but of its duration we know nothing. The young ones ready to leave the marsupium are already distinctly elongate, 3 mm. long, and have the peripetalous fasciole quite clearly developed. Pedicellariae have not yet appeared, but the first sphaeridia have been formed. A specimen of 8 mm. length is characteristically elongate (6 mm. broad), almost rect- angular. The circumlateral fasciole is well developed, but the transverse fasciole has only just begun to appear (cf. the development of the fascioles in the young Abatus cavernosiis, as described in my report on the Echinoidea of the Swedish South Polar Expedition, pp. 75-83). The periproct is in this specimen still close to the apical system. Genital pores have, of course, not yet appeared. A couple of globiferous pedicellariae found on the aboral side, and a rostrate pedicellaria found on the posterior end of the test, make the identification of this specimen as a young Parapneustes cordatus quite certain. One specimen, a male, is remarkable in having four genital pores, the madreporite also having a genital pore. The labrum is generally more pointed and narrow than shown in Koehler's pi. xvi, fig. 26; but both forms may occur, as shown in Plate III, figs. 3, 4, and it is not a sexual difference, as I find both the broad and the narrow form to occur in male specimens. Family SPATANGIDAE Echinocardium connectens, Mortensen (Plate IV, figs. 9-10) Echinocardium connectens, Mortensen, 1933. The Echinoderms of St Helena {other than Crinoids). Papers from Dr Th. Mortensen's Pacific Exped. 1914-16, Lxvi (Vid. Medd. Dansk Naturh. Foren., 93), p. 469. St. 299. 4. ix. 27. Tarrafal, San Antonio, Cape Verde Islands, 7-1 1 m. i specimen. This specimen is, at any rate, very closely related to the St Helena species, E. con- nectens, with which it agrees in the character of the frontal ambulacrum, the absence of larger spines (tubercles) above the ambitus, and in the character of the triphyllous pedicellariae. Unfortunately only triphyllous pedicellariae are found and these even are exceedingly scarce. On account of the very scarce and insufficient material hitherto known of £■. connectens, and with only a single specimen in hand of the present form, I do not like to state definitely that it is really identical with the species from St Helena, but I have little doubt about their identity. The specimen is 20 mm. long, 20 mm. broad and 14 mm. URECHINIDAE 235 high, the greatest height being behind the apical system. The frontal end is nearly vertical, the frontal ambulacrum only slightly sunken. The labrum reaches the second adjoining ambulacral plates ; two tube feet are found within the subanal fasciole. There is a rather conspicuous fasciole along each side of the periproct. Since no Echinocardimn has hitherto been recorded from the West African coast (south of Morocco) the discovery of this specimen is of considerable interest. It is most desirable that further material of the species should be obtained in order that its identity may be definitely settled. As it is a shallow-water species which no doubt lives buried in sand or mud, it should not be difficult to collect it in sufficient numbers. Plagiobrissus sp. (young) St. 279. 10. viii. 27. Off Cape Lopez, French Congo, 58-67 m. i specimen. This young specimen, which is only 5 mm. long, seems almost certainly to belong to the genus Plagiobrissus, and quite probably it is Plagiobrissus Costae (Gasco). But it is too young to be definitely identified. From St. 933, 17. viii. 32, 260 m., there is a very young specimen of a Spatangoid which is quite unidentifiable ; the whole oral side is lacking, so that it is not even possible to see whether it is an amphisternous or a meridosternous form. Family URECHINIDAE Plexechinus Nordenskjoldi, Mortensen (Plate IX, figs. 5-7) Plexechinus Nordenskjoldi, Mortensen, 1910. Swedish South Polar Exped. Echinoidea, p. 61, pis. xvii, figs. 1-8; xviii, figs. 5-12. P. A'ordenskjoldi, H. L. Clark, 1917. Hawaiian Echini. Echinoneidae . . . Spatangidae, p. 120. St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, iiom. i specimen. St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. 4 specimens. St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, 155-178 m. 2 specimens. St. 148. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m. i specimen. These specimens in general agree very well with the type, as described in my work of 1910 ; but the oral side can scarcely be said to be " somewhat deepened " in front of the peristome — there is hardly any depression at all to be observed here. Also it can hardly be said that the mouth opening is "almost vertical"; it can at most be said to be dis- tinctly oblique. The largest of the specimens is 19 mm. long, thus a good deal larger than the type. In regard to the pedicellariae it is remarkable that ophicephalous pedicellariae are lacking in all the present specimens ; on the other hand, the tridentate pedicellariae are better developed, reaching a somewhat larger size than in the type (Plate IX, figs. 5-7). In some of the specimens a number of specimens of a Loxosoma are found attached to the spines, mainly on the aboral side. This very interesting little Echinoid is still known only from the vicinity of South Georgia. 236 DISCOVERY REPORTS OPHIUROIDEA Family GORGONOCEPHALIDAE Astrotoma Agassizii, Lyman (Plate V, figs. 1,2; Plate VI, figs, i, 2) Astrotoma Agassizii, Lyman, 1875. Ophiuridae and Astrophytidae. Hassler Exped. 111. Cat. Mus. Comp. Zool., viii, p. 24, pi. iv, figs. 52-56. A. Agassizii, Koehler, 1908. Scott. Nat. Antarct. Exped. Asteries, Ophiures et Echinides, p. 614, pi. xiii, fig. 120. A. Agassizii, Doderlein, 191 1. Japanische und andere Euryalae, p. 100. A. Agassizii, Koehler, 1922. Austral. Antarct. Exped. Echinod. Ophiuroidea, p. 9, pi. Ixxvi, figs. I-II. A. Agassizii, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. 102. A. Agassizii, Fedotov, 1927. Morphologische Studien an Euryalae. Zeitschr. Morphol. Okol. Tiere, ix, pp. 381-4. A. Agassizii, Doderlein, 1927. Indopacifische Euryalae, p. 87. A. Agassizii, Doderlein, 1930. Deutsche Tiefsee-Exped. Ophiuriden. II, Euryalae, p. 372, Taf. i, figs. I, I «. St. 39. 25. iii. 1926. East Cumberland Bay, South Georgia, 179-235 m. 2 large specimens. St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 m. 4 large specimens, 2 small ones. St. 140. 23. xii. 26. StromnessHarbourtoLarsenPoint, South Georgia, 122-136 m. 8 specimens, large and small. St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, 155-178 m. 5 specimens. St. 148. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m. 2 specimens. St. 152. 17. i. 27. 53° 51' S, 36° 18' W, South Georgia, 245 m. 5 specimens. St. 159. 21. i. 27. 53° 52' S, 36° 08' W, South Georgia, 160 m. i specimen. St. 160. 7. ii. 27. Near Shag Rocks, South Georgia, 177 m. 15 specimens. St. 345. 8. ii. 30. 55° 20' S, 34° 47' W, 200-100 m. i specimen. St. 600. 17. i. 31. 67° 09' S, 69° 27' W, 501-527 m. 8 specimens. St. 652. 14. iii. 31. Burdwood Bank, 171-169 m. i specimen. St. WS 83. 24. iii. 27. 14 miles S 64° W of George Island, East Falkland Islands, 137-129 m. I specimen. St. WS 84. 24. iii. 27. 7-5 miles S 9° W of Sea Lion Island, East Falkland Islands, 75-74 m. I specimen. St. WS 85. 25. iii. 27. 8 miles S 66° E of Lively Island, East Falkland Islands, 79 m. 15 speci- mens (young). St. WS93. 9. iv. 27. 7 miles S 80° W of Beaver Island, West Falkland Islands, 133-130 m. 3 specimens. St. WS 108. 25. iv. 27. 48° 31' S, 63° 34' W, 118-120 m. I specimen. St. WS 243. 17. vii. 28. 51° 06' S, 64° 30' W, 144-141 m. i specimen. St. WS 246. 19. vii. 28. 52° 25' S, 61° 00' W, 267-208 m. 3 specimens. St. WS 248. 20. vii. 28. 52° 40' S, 58° 30' W, 210-242 m. 10 specimens. St. WS 818. 17. i. 32. 52° 34' S, 63° 13' W, 278-284 m. 6 specimens. 52° 42' S, 62° 39' W, 312-329 m. I specimen. 52° 53' S, 61° 51' W, 351-367 m. 2 specimens. 52° 29' S, 58° 27' W, 146-137 m. 15 specimens. 50° 50' S, 57° 15' W, 135-144 m. 5 specimens. St. WS 840. 6. ii. 32. 53° 52' S, 61° 49' W, 368-463 m. 2 specimens. St. WS 871. I. iv. 32. 53° 16' S, 64° 12' W, 336-341 m. 4 specimens. St. WS 819. 17. i. 32 St. WS 820. 18. i. 32 St. WS 824. 19. i. 32 St. WS825. 28.1.32 GORGONOCEPHALIDAE 237 The very rich material of A. Agassi zii collected by the ships of the Discovery Com- mittee gives us a much better idea of what a splendid form this Gorgonocephalid is than could be gathered from the specimens hitherto known. The largest of these, collected by the Australasian Antarctic Expedition (cf. Koehler, op. cit., 1922), did not surpass a diameter of disk of 36 mm. — in the present material the largest specimen measures 60 mm. in diameter of disk, others measuring some 45-55 mm. in diameter. In these large specimens the radial ribs are generally very conspicuous, though de- pendent to some degree on the preservation, specimens with the disk somewhat swollen having them much less conspicuous than those with the disk flattened. The latter case particularly holds good of the largest specimen (Plate V, fig. 2), which, indeed, gives much the impression of being senescent. The arms in these large specimens are very powerful, some 11 mm. wide, 13 mm. high, and have a very snake-like appearance. About the length of the arms in these large specimens it is not possible to give any definite statements, because the arms are very much coiled up and in most cases broken. But in a specimen 22 mm. in diameter of disk with the arms uncoiled I find the arm length to be no less than ca. 400 mm. This does not necessarily imply that in the larger specimens the arms are correspondingly longer, for in a specimen of ca. 25 mm. diameter the arms do not much exceed 300 mm. in length. (Koehler states that the arms of his specimen of 36 mm. diameter of disk were more than 320 mm. in length ; evidently they were broken, so that it is unknown by how much they exceeded that length.) It would appear that there is considerable variation in the length of the arms, though not in the arms of the same specimen, such as is characteristic of Asteronyx. In specimens with the disk flattened there is generally a sharp edge along the inter- radii, recalling the belt of marginal plates of Gorgonoceplmlus. There are, however, no conspicuous marginal plates ; on removing the grain-covering one finds the scales here not much larger than the scales underlying the granules of the disk; it is mainly the folding of the skin that produces this sharp edge. The primary plates of the disk are usually distinct in the young specimens, as re- presented by Koehler {op. cit., 1908, pi. xiii, fig. 120); but herein there is a good deal of individual variation. I have opened a couple of specimens in order to see what they feed on. Remnants of Crustaceans, Copepods and what was probably a Hyperiid were found and in one specimen there was a rather large lump of jelly of uncertain origin. It appears that the species feeds on plankton organisms, which it most probably catches with the slender distal part of its arms. The eggs are small and numerous, indicating perhaps a free- swimming larval stage. Astrochlamys bruneus, Koehler (Plate VII, fig. 8) Astrochlamys bruneus, Koehler, 1912. IP Exped. Antarct. Fran9aise. Echinodermes, p. 143, pi. xi, figs. 3, 4, 6, 7, 14, 15. A. bruneus, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. 103. A. bruneus, Doderlein, 1930. Deutsche Tiefsee-Exped. Ophiuriden. II, Euryalae, p. 373, Taf. i, figs. 3-5 (p. 356, fig. le; p. 363, fig. 14 /, m). 238 DISCOVERY REPORTS St. 39. 25. iii. 26. West Cumberland Bay, South Georgia, 179-235 m. 4 specimens. St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, 230-250 m. 2 specimens. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. 3 speci- St. 140. mens. St. 144. St. 148. St. 170. 5. i. 27. Off mouth of Stromness Harbour, South Geotgia, 155-178 m. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m. i specimen. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. i specimen. 4 specimens. Most of the larger specimens of this species are carrying a smaller one on their back (Plate VII, fig. 8), a fact naturally leading to the suggestion that this is a sort of copula- tion, the larger specimen being the female, the smaller the male. And so it actually is, as proved by the examination of the gonads of these specimens. It was possible also to demonstrate that the species is viviparous. The two first specimens opened had their bursae entirely empty ; the third had them packed with eggs, which seemed, however, to be only just fertilized. But the fourth specimen opened settled the question, for it had the bursae filled with embryos, all in the same stage of development. The embryos a A^)^ Fig. 2a-c. Astrochlamys bruneus, Koehler. a. Part of oral side, xS. b, Embryo, showing the five terminal plates, and in the centre the mouth, x8o. c, Arm spine, X45. Fig. 2 d. Arm spine of Astrochlamys sol, n.sp., X45. were small and star-shaped, with an indication of a mouth invagination and the first indication of the skeleton, viz. the terminal plate ; but as yet there was no trace of the ambulacral skeleton or of any other plates (Fig. 2 b). There were some 200 embryos in each bursa, which means that they cannot reach any large size before they leave the mother — in conformity with the small size of the genital slits. The eggs are shed (into the bursae) all at a time, and this appears to be likewise the case with the sperms, none of the males being found to contain ripe sperms in their gonads. This is again in con- formity with the fact that some of the larger specimens carry no male on their back, so that copulation is not going on constantly, as is the case in Amphilycus androphorus, Mortensen, and, to a less degree apparently in Ophiosphaera insignis, Brock, and Ophiodaphne materna, Koehler (cf. Mortensen, Biological observations on Ophiurids. Papers from DrTh. Mortensen's Pacific Exped., LXiii (Vid. Medd. Dansk Naturh. Foren., 93), 1933, pp. 178-88). In these three Ophiurids the male is carried over the mouth of GORGONOCEPHALIDAE 239 the female. Astrochlamys bruneus is the first Ophiurid known to carry the male on the back, as it is also the first viviparous Gorgonocephalid made known. The largest specimen at hand measures 20 mm. in diameter of disk, the arms being CO. 100 mm. long. I give here a sketch of the oral side of this species (fig. 2 a) in order to show the shape of the plates, which is not seen distinctly in either Koehler's or Doderlein's figures. Koehler states that the two first pairs of pores are without papillae, and his pi. xi, fig. 6 (op. cit., 1912) shows it distinctly. Doderlein {op. cit.) states that only the first pore pair is without papillae, and this is clearly indicated in his figure (Taf. i, fig. 4). The apparent contradiction is due to the fact that the first pair of pores in Koehler's figure are the outer mouth pores. But these are covered over by a thick skin which closes up the distal half of the mouth slits. Only when this skin is dissolved do these pores become distinct. An indication of this skin covering the mouth slit is seen on the right side, the upper arm, in Doderlein's Taf. i, fig. 4. Astrochlamys sol, n.sp. (Plate VII, fig. 9) St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. i specimen, attached to a coral-like Bryozoan. Diameter of disk ca. 13 mm., length of arms ca. 50 mm. Arms 10. In its general characters this species closely resembles A. bruneus. Mouth papillae not distinct, being covered by the thick skin that invests the whole animal. First pair of arm pores apparently rudimentary. As in ^, bruneus there are one or two spines (papillae) at the second pore pair, three at the third pair, and thereafter four spines. These latter are slightly different from those of A. bruneus, having fewer thorns (Figs. 2 c-d). The hooks are alike in both ; the hook belts are interrupted in the dorsal median line in the proximal part of the arms. In regard to the plates of the oral region it appears that the buccal plates are somewhat better developed than in A. bruneus. Ventral plates of arms irregularly divided as in bruneus. Not thinking it desirable to remove the specimen from the Bryozoan to which it clings, or to spoil it too much by cleaning off the skin, I cannot give any detailed figures. But the species is so markedly different from the five-armed A. bruneus by reason of its numerous arms that the characters here given should suffice for recognizing it with certainty. Astrohamma tuberculatum (Koehler) Astrothamnus tuberculatus, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures., p. 133, figs. I a-f. Asiroioma tuberculatum, Doderlein, 1927. Indopacifische Euryalae, p. 21. Astrohamma tuberculatum, Doderlein, 1930. Deutsche Tiefsee-Exped. Ophiuriden. II, Euryalae, p. 372, Taf. i, fig. 2 (p. 363, fig. 14 n). St. 190. 24. iii. 1927. Bismarck Strait, Palmer Archipelago. 315 m. 2 specimens. These specimens are 15-16 mm. in diameter of disk, thus somewhat larger than those hitherto recorded (8-10-5 mrn-)- They conform perfectly to the descriptions given by Koehler and Doderlein and do not call for further remarks. 240 DISCOVERY REPORTS Gorgonocephalus chilensis (Philippi) Gorgonocephaliis cMlefisis, Doderlein, 191 1. Japanische und andere Euryalae, pp. 30, 105, Taf. v, fig. 5; viii, figs. I, I «. G. chilensis, H. L. Clark, 1915. Cat. Recent Ophiurans, p. 185. G. chilensis, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. loi, pi. xiv, fig. I. G. chilensis, Doderlein, 1927. Indopacifische Euryalae, pp. 30, 92. G. chilensis, Zirpolo, 1932. Siil Gorgonocephalus chilensis, Lyman. Ann. Mus. Zool. Univ. Napoli, VI, 7, pp. 1-16. Non: Gorgotiocephaltis chilensis, H. L. Clark, 1923. Echinoderm Fauna South Africa, p. 318 { = Gorgonocephalus pectinatus, Mortensen). For the older literature I may refer to Doderlein's work of 191 1 (/oc. cit.). St. 51. 4. V. 26. Off Eddystone Rock, East Falkland Islands, 105-115 m. i large specimen. St. 158. 21.1.27. 53° 48' S, 35° 57' W, South Georgia, 401-41 1 m. 2 specimens, one large, ca. 70 mm. diameter of disk, one small, 20 mm. diameter. St. 160. 7. ii. 27. Off Shag Rocks, South Georgia, 177 m. i young specimen, 9 mm. diameter of disk. St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-335 "i- 3 specimens. St. 576. 17. iv. 31. Falkland Islands, 34-24 m. 2 large specimens. St. WS 72. 5.111.27. 5i°07'S, 57° 34' W, Falkland Islands, 79 m. i specimen, ca. 65 mm. diameter of disk. St. WS 73. 6. iii. 27. 51° 01' S, 58° 54' W, Falkland Islands, 121 m. 2 specimens. St. WS 76. II. iii. 27. 51° 00' S, 62° 02' W, Falkland Islands, 207-205 m. 3 specimens. St. WS 80. 14. iii. 27. 50° 57' S, 63° 37' W, Falkland Islands, 152-156 m. 2 specimens, young, 8 and 15 mm. diameter of disk. St. WS 829. 31. i. 32. 50° 51' S, 63° 13' W, Falkland Islands, 155 m. 4 specimens. This species vi^as not hitherto known to occur as far south as South Georgia and the Palmer Archipelago. The larger of the specimens from South Georgia has almost naked radial ribs, but does not otherwise differ from typical G. chilensis ; in the smaller speci- men, on the other hand, the radial ribs are rather unusually strongly spiny. On the whole, there is much variation in the development of "stumps" on the disk. In the young specimens of G. chilensis the disk is densely covered by a uniform granulation, which is, again, remarkably coarse in the specimen from the Shag Rocks Bank. On the largest of the specimens from St. WS 829 a couple of young ones, 3 mm. in diameter of disk, were found attached. Already at this size the disk is covered by rounded granules, none of the primary plates remaining distinct. In the young specimens the hooks are already present at the base of the arms, but only on the sides, not forming complete belts till beyond the second forking, and in the adult specimens still farther out. The larger specimen from St. 158 is stated in a note to be thus coloured: "Disk whitish, arms salmon-pink, deepest at tips." Two specimens were opened in order to see whether they might be infested with Proiomyzostoma ; nothing of the kind was found. The eggs are small and the gonads exceedingly numerous, much as in Gorgonocephalus eucnemis, etc. ; the presence of young ones on one of the specimens is thus, as in other species, a casual attachment of the TRICHASTERIDAE— OPHIOMYXIDAE 241 young on an adult — not necessarily its parent — and not a case of viviparity or brood- protection. That the specimens from South Africa referred to G. chilensis have nothing to do with this South American species, but represent a distinct species, G.pectinatus, Mortensen, I showed in my paper on the Echinoderms of South Africa (Papers from Dr Th. Mortensen 's Pacific Exped., Lxv (Vid. Medd. Dansk Naturh. Foren., 93), 1933, p. 283). As for the specimens from Kerguelen and New Zealand, likewise referred to G. chilensis, I must reserve opinion until the examination of new and sufficient material allows a definite judgment. The zoogeography of other sub-Antarctic Echinoderms is not in favour of these specimens being identical with the South American G. chilensis. Family TRICHASTERIDAE Astroceras elegans (Bell) Astroschema elegans. Bell, 1917. Brit. Antarct. ('Terra Nova') Exped. Echinoderma, p. 7. Astroceras elegans, Mortensen, 1924. Echinoderms of New Zealand and the Auckland-Campbell Islands. II, Ophhiroidea. Papers from Dr Th. Mortensen's Pacific Exped., xx (Vid. Medd. Dansk Naturh. Foren., 77), p. 107, pi. iv, fig. 3. A. elegam, Mortensen, 1933. Studies of Indo-Pacific Euryalids. Vid. Medd. Dansk Naturh. Foren., 96, p. 53. St. 934. 17. viii. 32. 34° 11' S, 172° 10' E, north of New Zealand, 92-98 m. 2 specimens. These specimens differ from those hitherto known in the arms being of a uniform light brown colour (apart from the white tubercles), not an annulated appearance due to alternating rings of white and brown skin. I do not think this colour difference to be any serious objection to referring them to the species characteristic of the area off the north end of New Zealand, A. elegans. Family OPHIOMYXIDAE Ophiomyxa vivipara, Studer Ophiomyxa vivipara, Studer, 1876. tjber Echinod. a. d. antarkt. Meere. Monatsber. Akad. Berlin, p. 462. Ophioscolex Coppingeri, Bell, 1881. Echinod. Straits of Magellan and coast of Patagonia. Proc. Zool. Soc. Lond., 1881, p. 98, pi. viii, fig. 6. Ophiomyxa vivipara, Ludwig, 1898. Ophiuren d. Sammlung Plate. Zool. Jahrb., Suppl. iv, p. 768. O. vivipara, H. L. Clark, 1915. Cat. Recent Ophiurans, p. 170, pi. 2, figs. 1-2. O. vivipara, Mortensen, 1920. On Hermaphroditism in viviparous Ophiurids. Acta Zoologica, i, p. 10. O. vivipara, H. L. Clark, 1923. Echinoderm Fauna of South Africa. Ann. S. African Mus., xni, p. 313. O. vivipara, Mortensen, 1933. Echinoderms of South Africa. Papers from Dr Th. Mortensen's Pacific Exped., lxv (Vid. Medd. Dansk Naturh. Foren., 93), p. 301. 6-2 242 DISCOVERY REPORTS St. 6. I. ii. 25. Tristan da Cunha, 80-140 m. i specimen. St. 388. 16. iv. 30. 56° 19' S, 67° 10' W, 121 m. I specimen. St. WS 73. 6. iii. 27. 51° 01' S, 58° 54' W, Falkland Islands, 121-130 m. 2 specimens. St. WS 80. 14. iii. 27. 50° 57' S, 63° 37' W, Falkland Islands, 152-156 m. i specimen. St. WS 81. 19. iii. 27. 8 miles N 11° W of North Islands, West Falkland Islands, 81-82 m. 10 specimens. St. WS 84. 24. iii. 27. 7-5 miles S 9° W of Sea Lion Island, East Falkland Islands, 75-74 m. 8 specimens. St. WS 85. 25. iii. 27. 8 miles S 66° E of Lively Island, East Falkland Islands, 79 m. 2 specimens. St. WS 88. 3. iv. 27. 54° 00' S, 64° 57' W, 118 m. 2 specimens. St. WS 93. 9. iv. 27. 7 miles S 80° W of Beaver Island, West Falkland Islands, 133-130 m. 4 specimens. St. WS 243. 17. vii. 28. 52° 00' S, 64° 30' W, West Falkland Islands, 144-141 m. 3 specimens. St. WS 776. 3. xi. 31. 46° 18' S, 65° 02' W, 107-99 m. I specimen. St. WS 804. 6.1.32. 50° 21' S, 62° 53' W, 143-150 m. I specimen. St. WS823. 19.1.32. 52° 14' S, 60° 01' W, 80-95 m. I specimen. St. WS 824. 19. i. 32. 52° 29' S, 58° 27' W, 146-137 m. 10 specimens. St. WS 825. 19. i. 32. 50° 50' S, 57° 15' W, 135-144 m. I specimen. St. WS 848. 10. ii. 32. 50° 37' S, 66° 24' W, 115-117 m. i specimen. In my paper on the South African Echinoderms {op. cit., 1933, p. 302) I remarked that the only difference I could find between the South African and the South American form of this species was that in the South African form there is only one spine at the proximal five or six pore pairs, whereas in the Magellanic form there are two spines already at the third or fourth pore pair. The specimens in the present collection all have two spines from the third or fourth pore pair, which seems to indicate that this dif- ference is reliable, so that it may be possible to distinguish the South African form as a separate variety, var. capensis, n.var. The specimen from off Tristan da Cunha, a locality from which the species had not hitherto been recorded, agrees with the typical South American form in regard to the numbers of spines at the proximal pore pairs. Ludwig {op. at.) proved that Ophioscolex Coppingeri of Bell was identical with Ophiojnyxa vivipara. Having seen the type-specimen of Bell's species in the British Museum, I can confirm the result reached by Ludwig. Ophiomyxa brevirima, H. L. Clark Ophiotnyxa brevirima, H. L. Clark, 1915. Cat. Recent Ophiurans, p. 169, pi. i, figs. 3-4. O. brevirima, Mortensen, 1924. EcJmioderms of New Zealand and the Auckland-Campbell Islands. II, Ophiiiroidea. Papers from Dr Th. Mortensen 's Pacific Exped., xx (Vid. Medd. Dansk Naturh. Foren., 77), p. no. St. 941. 20. viii. 32. 40° 51' S, 174° 48' E, Cook Strait, 128 m. 4 specimens. Referring to the observation recorded in my paper on the New Zealand Ophiurids (p. 113) that two types of breeding occur among the specimens identified as O. brevirima, viz. one having the bursae filled up with a great number of small embryos, the other having only one young in each bursa (a difference tending to indicate that the two types in reality represent two distinct species), it should be mentioned that the present OPHIOMYXIDAE 243 Specimens belong to the type with numerous small embryos in the bursae. The speci- mens, which are not very well preserved, show an indication of dark bands on the arms ; otherwise they are white. Ophioscolex (Ophiolycus) nutrix, n.sp. (Plate VII, fig. 6.) St. 159. 21. i. 27. 53° 52' S, 36° 08' W, South Georgia, 160 m. 2 specimens. St. 160. 7. ii. 27. 53° 43' S, 40° 57' W, near Shag Rocks, South Georgia, 177 m. i specimen. St. WS 42. 7. i. 27. 54° 42' S, 36° 47' W, South Georgia, 198 m. 2 specimens. St. WS 86. 3. iv. 27. 53° 53' S, 60° 34' W, Falkland Islands, 1 51-147 m. i specimen. St. WS 225. 9. iv. 28. 50° 20' S, 62° 30' W, Falkland Islands, 162 m. i specimen. St. WS 228. 30. vi. 28. 50° 50' S, 56° 58' W, Falkland Islands, 229-236 m. 4 specimens. St. WS 824. 19. i. 32. 52° 29' S, 58° 27' W, Falkland Islands, 146-137 m. 2 specimens. St. WS 839. 5. ii. 32. 53° 30' S, 63° 29' W, Falkland Islands, 403-434 m. i specimen. Diameter of disk up to ca. 10 mm., length of arms about three to four times the dia- meter of disk. Both aboral and oral sides of disk with a varying number of small spines, protruding through the thick investing skin; in the specimen figured in Plate VII, fig. 6, they are exceptionally numerous. Radial shields small, but fairly distinct, oval, widely separated — but in some specimens no trace of radial shields is seen. Fig. 3. Ophioscolex nutrix, n.sp. a, Part of oral side, b. Part of dorsal side, x8. c. Side view of distal arm joints, showing the two upper spines transformed into hooks, '30. d. Spicules from stomach wall, X45. Mouth papillae rather irregular and varying in number ; generally the distalmost one or two are somewhat enlarged. Buccal shields conspicuously broader than long, with an obtuse angle within and a straight outer edge. Adoral shields joining within. Ventral plates contiguous in the proximal part of arm, separated beyond the disk; the distal 244 DISCOVERY REPORTS edge is gently convex. The pores have a well-developed tentacle scale (Fig. 3 a). Dorsal arm plates in the basal part of arms fairly regularly divided into two parts (Fig. 3 b). Arm spines two on the joints within the disk; beyond the disk there are three spines. In the proximal part of the arms the lowermost and the uppermost spine enlarged, somewhat flattened and sometimes a little curved. In the distal part the two uppermost spines transformed into hooks (Fig. 3 c). The scales of the disk as usual with a glassy, concentrically striated margin. Rather large mainly three-radiate spicules in the stomach wall (Fig. 3 d). About the colour in life there is no information; the preserved specimens are white. It is evident that this species is nearly related to the South African O. dentatus, Lyman, and it might, indeed, seem questionable whether the small differences to be observed (cf. fig. 3 with the figures of O. dentatus given in my paper on the South African Echinoderms, pp. 310-11, figs. 32-4) are really of suflnicient value for specific distinction. Of this there cannot, however, be the slightest doubt, for I find that the South American form is viviparous and— partly— hermaphrodite, whereas O. dentatus, as well as the closely related O. purpiireus, are not viviparous and have separate sexes. Thus it is certain that the South American form is a quite distinct species, the more interesting as this is the first viviparous species of the genus Ophioscolex to be known. The statement that this species is "partly" hermaphrodite means that some speci- mens are not hermaphrodite, but purely males or females. Thus the species is a facultative hermaphrodite. This is of interest in connection with the fact that the two known viviparous species of Ophiomyxa have separate sexes. It would seem to be a beginning of hermaphroditism that we are witnessing in this primitive Ophiurid, herm- aphroditism in Ophiurids being apparently a specialization acquired in connection with viviparity. As regards the arrangement of the gonads in the hermaphrodite specimens it seems to be the more usual condition that the male gonads are situated at the adradial side, the female gonads along the interradial side of the bursae, but there is no regularity either in the arrangement or the number of the gonads. I have found only one or two young ones at a time in the bursae. No species of Ophioscolex was known till now from the Antarctic or sub-Antarctic region, the "Ophioscolex Coppingeri" of Bell being nothing but Ophiomyxa vivipara (cf. above, p. 242). It is thus a rather surprising fact that no less than two species of the genus, Ophioscolex nutrix and the following species, O. marionis, have been discovered by the ships of the Discovery Committee in the sub-Antarctic region. Ophioscolex marionis, n.sp. St. 1563. 7. iv. 35. 46° 48' S, 37° 49' E, off Marion Island, 113-99 m- i specimen. Diameter of disk 5-5 mm. Arms all broken close to the disk; they are 1-5 mm. broad, somewhat flattened. Disk covered by thick skin, through which some scattered short spines protrude all over the dorsal side, a single one also here and there on the ventral interradii. Apparently no radial shields. The buccal shields are small, rounded tn- OPHIOMYXIDAE 245 angular, adoral shields rather broad, not joining within ; they separate broadly the buccal shield from the first lateral plate. Mouth papillae small, feebly developed. Ventral arm plates distinctly longer than broad, with convex distal edge, joining broadly at least as far out on the arms as they are preserved. Dorsal arm plates very feebly developed, as usual, but covering the whole broad dorsal side between the small lateral plates; they are not divided in two by a transverse line. Arm spines four, short, robust, the lower- most and uppermost ones slightly longer than the two middle ones. One tentacle scale. Colour in alcohol brownish. The species is clearly viviparous; not that I have actually found embryos within it, but the eggs are large, ca. o-z-o-t, mm. and yolky, seven to ten eggs in each gonad; and then it is hermaphrodite, there being one male gonad at the adradial side and one or two female gonads at the interradial side of the bursa. As not a single case is known Fig. 4. Ophioscolex marionis, n.sp. a, Part of oral side, b, Part of dorsal side of arm. The dotted arch-lines represent the outline of the vertebrae seen through the exceedingly thin dorsal plates. X15. of a non-viviparous Ophiurid being hermaphrodite, we may, I think, safely con- clude that this species must be viviparous, as the character of the female gonads also indicates. It may be added that no spicules are found in the stomach wall, such as occur in O. niitrix. That this species is not very closely related to the other Antarctic species of Ophio- scolex, O. mitrix, is evident enough; a comparison of the figures will show the difi^erences in the shape of the buccal shield, ventral and dorsal arm plates, and the mouth papillae, to which may be added the difference in the number of arm spines, one having three, the other four spines, which makes an important difference within this genus. It would appear that the present species is the nearest related to O. quadrispinus, Verrill, from off Nova Scotia ; this species is, however, scarcely sufficiently well known to allow the dif- ferences between the two species to be indicated ; but it seems that O. quadrispinus has no spines on the disk. 246 DISCOVERY REPORTS It is very regrettable that the arms of the single specimen of O. marionis are all broken, so that it cannot be seen whether the upper spines are transformed into hooks in the distal part of the arms, as they are in O. dentatiis-purpureiis and O. nutrix, a character which has an important bearing on the question of the validity of the subgenus or genus Ophiolyciis (cf. my paper on the Echinoderms of South Africa, p. 315). If these spines are not transformed into hooks in O. marionis this species will form a connecting link between Ophiolycus and the typical Ophioscolex. Family OPHIACANTHIDAE Ophiacantha vivipara, Ljungman (Plate VII, fig. 2) Ophiacantha vivipara, Ljungman, 1870. On tvdnne nya arter Ophiurider. Ofvers. Vet. Akad. Handl., 1870, p. 470. Ophiocoma (?) vivipara, Wyv. Thomson, 1877. The Atlantic, 11, pp. 241-4, fig. 50. Ophiacantha vivipara, Ludwig, 1899. Ophiuroideen Hamburger Magalh. Sammelreise, p. 13. O. vivipara, Koehler, 191 2. IP Exped. Antarct. Fran^aise. Echinodermes, p. 138, pi. xi, figs. 1-2, 10. O. vivipara, Koehler, 1917. Echinodermes de Kerguelen. Ann. Inst. Oceanogr., vn, 8, p. 71. O. vivipara, Mortensen, 1920. On Hermaphroditism in viviparous Ophiurids. Acta Zoologica, i, p. 10. O. vivipara, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. 105. O. vivipara, G. A. Smith, 1923. Report on the Echinoderms coll. during the voyage of the ' Quest'. Ann. Mag. Nat. Hist., 9 Sen, xn, p. 368. O. vivipara heptactis. Hertz. 1926. Deutsche Siidpolar-Exped. Ophiuroiden, p. 36. O. vivipara, Koehler, 1927. Austral. Antarct. E.xped. Echinod. Ophiuroidea, p. 12. O. vivipara, Grieg, 1929. Some Echinoderms from the South Shetlands. Bergens Mus. Arbok, 1929, 3, p. 7. For references to the older literature I may refer to Ludwig, op. cit. St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, no m. Several specimens. St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, 179-235 m. 4 specimens. St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 m. Several specimens. St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, 230-250 m. 2 specimens. St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. ca. 15 young specimens. St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, 155-178 m. 4 specimens. St. 148. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m. i young specimen. St. 156. 20. i. 27. 53° 51' S, 36° 21' W, South Georgia, 200-236 m. 2 specimens. St. 159. 21. i. 27. 53° 52' S, 38° 08' W, South Georgia, 160 m. i specimen. St. 160. 7. ii. 27. Near Shag Rocks, South Georgia. 177 m. 5 specimens. St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. 10 specimens. St. 175. 2. iii. 27. Bransfield Strait, South Shetlands, 200 m. 2 specimens. St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 93-130 m. 3 specimens. St. 474. 12. xi. 30. Off Shag Rocks, South Georgia, 199 m. 12 specimens. St. 599. 17. i. 31. 67° 08' S, 69° 06' W, 203 m. I specimen. St. 652. 14. iii. 31. Burdwood Bank. 171-169 m. i specimen. St. 1562. 7. iv. 35. 46° 53' S, 37° 55' E, off Marion Island, 90-97 m. 4 specimens. St. 1563. 7. iv. 35. 46° 48' S, 37° 39' E, off Marion Island, 113-99 m. 6 specimens. OPHIACANTHIDAE 247 St. 1564. 7. iv. 35. 46° 36' S, 38° 02' E, off Marion Island, 108-113 m. Several specimens. St. WS 27. 19. xii. 26. 53° 5s' S, 38° 01' W, South Georgia, 106 m. 10 specimens. St. WS 33. 21. xii. 26. 54° 59' S, 35° 24' W, South Georgia, 130 m. ca. 20 specimens. St. WS 42. 7. i. 27. 54° 42' S, 36° 47' W, South Georgia, 198 m. 2 specimens. St. WS 73. 6. iii. 27. 51° 01' S, 58° 54' W, Falkland Islands, 121 m. 3 specimens. St. WS 80. 14. iii. 27. 50° 57' S, 63° 37' W, Falkland Islands, 152-156 m. 2 specimens. St. WS 81. 19. iii. 27. 8 miles N 11° W of North Island, West Falkland Islands, 81-82 m. 5 specimens. St. WS 84. 24. iii. 27. 7I miles S 9° W of Sea Lion Island, East Falkland Islands, 75-74 m. Several specimens. St. WS 85. 25. iii. 27. 8 miles S 66° E of Lively Island, East Falkland Islands, 79 m. 15 speci- mens. St. WS 88. 6. iv. 27. 54° 00' S, 64° 57' W, 118 m. 8 specimens. St. WS 91. 8. iv. 27. 52° 54' S, 64° 37' W, 191-205 m. i specimen. St. WS 92. 8. iv. 27. 51° 58' S, 65° 01' W, 145-143 m. Several specimens. St. WS 93. 9. iv. 27. 7 miles 80° W of Beaver Island, West Falkland Islands, 133-130 m. 5 speci- mens. St. WS 233. 5. vii. 28. 49° 25' S, 59° 45' W, 185-175 m. 5 specimens. St. WS 234. 5. vii. 28. 48° 52' S, 60° 25' W, 195-207 m. Several specimens. St. WS 248. 20. vii. 28. 52° 40' S, 58° 30' W, Falkland Islands, 210-242 m. Several specimens. St. WS 750. 19. ix. 31. 52° 12' S, 67° 19' W, 95 m. 2 specimens. St. WS 755. 2i.ix. 31. 51° 39' S, 57° 39' W, 77 m. 8 specimens. St. WS 773. 31. X. 31. 47° 28' S, 60° 51' W, 291-296 m. I specimen. St. WS 800. 22. xii. 31. 48° 16' S, 62° 10' W, 139-137 m. 5 specimens. St. WS 804. 6. i. 32. 50° 21' S, 62° 53' W, 143-150 m. 7 specimens. St. WS815. 13.1.32. 51° 52' S, 65° 44' W, 132-162 m. I specimen. St. WS816. 14. i. 32 St. WS817. 14.1.32 St. WS 824. 19. i. 32 St. WS 825. 29. i. 32 52° 10' S, 64° 56' W, 150 m. 2 specimens. 52° 23' S, 64° 19' W, 191-202 m. 14 specimens. 52° 29' S, 58° 27' W, 146-137 m. Several specimens. 50° 50' S, 57° 15' W, 135-144 m. ca. 15 specimens. St. WS 840. 6. ii. 32. 53° 52' S, 61° 49' W, 368-463 m. 15 specimens. St. WS 869. 31. iii. 32. 52° 15' S, 64° 14' W, 187 (-0) m. Several specimens. St. MS 14. 17. ii. 25. East Cumberland Bay, South Georgia, 109-180 m. 6 specimens. St. MS 71. 9. iii. 25. East Cumberland Bay, South Georgia, 110-60 m. Several specimens. In my paper. On Hermaphroditism in viviparous Op/iiurids, I have stated (p. 10) that this species appears to be a protandric hermaphrodite, the old and poorly preserved material then at my disposal making this statement, however, somewhat uncertain. On a renewed examination of the rich material now at hand I find that it is not protandric. Specimens of 5-6 mm. diameter already show gonads of distinctly female character. I think that in one case I have discerned a male gonad in a specimen of 6 mm. diameter of disk, together with distinct female gonads. But otherwise I have found only female gonads in all the numerous specimens of various sizes, from 5 to 18 mm. diameter of disk. (Among the specimens of var. pentactis male specimens are quite common.) This is quite a perplexing case ; one is, indeed, tempted to suggest that the species is usually parthenogenetic. I have re-examined some preparations on which the statements in my paper of 1920 about the hermaphroditism of this species are (partly) based, and find that there is, at least, one clearly hermaphrodite gonad from a specimen of 9-5 mm. dia- 248 DISCOVERY REPORTS meter, whereas the gonads of a specimen of 8-5 mm. diameter appear to be all female. A gonad from a specimen of 14 mm. diameter is purely male. The indications thus are that this species is a proterogynic hermaphrodite. It may be mentioned that there is only one young one, more rarely two, at a time in the bursae, and rarely in more than two to three bursae at a time, in conformity with the fact that rarely more than three to four young ones are found attached to the females. In some cases I found gonads only along the adradial side of the bursae. Some of the specimens from South Georgia are infested by a parasitic organism, probably referable to the Copepod genus Ophioika, described by K. Stephensen from an Ophiacantha from the Bali Sea {Some new Copepods, parasites of Ophhirids and Echinids. Papers from Dr Th. Mortensen's Pacific Exped., lxiv (Vid. Medd. Dansk Naturh. Foren., 93), 1933, p. 205). The parasite does not castrate its host ; I have found young ones in the bursae in a specimen infested with the parasite. A couple of speci- mens from Sts. 170 and 190 are infested by an ectoparasitic Copepod of the genus Cancer illopsis (?), attached on the dorsal side of the arms, close to the disk. M. Hertz {op. at., 1926) has established a variety heptactis for the Kerguelen speci- mens, founded on the fact that an 8-rayed specimen is carrying a 7-rayed young one, all the specimens on the whole being 7-rayed with the exception of only two 8-rayed specimens. The 7-rayed condition thus appears to be hereditary in the Kerguelen form, which is taken to necessitate a separate subspecies for this form. Evidently, however, Dr Hertz has forgotten that Studer ( tjber Echinodermen aus dem antarktischen Meere, ges. a. d. Reise S.M.S. 'Gazelle'. Monatsber. Akad. Berlin, 1876, p. 460) had already established a variety kergtielensis for the Kerguelen form ; if this form then deserves to rank as a subspecies (which I rather doubt), its name must be kerguelensis, the name heptoctis of Hertz falling directly into synonymy therewith. As for the specimens in the present collection 6-rayed and 7-rayed specimens are about equally common. One may find 6-rayed young ones on 7-rayed specimens (whereas the inverse case, 7-rayed young ones on 6-rayed specimens was not observed). Here the number of the arms is decidedly of no classificatory value whatever. As for the 5-rayed specimens, cf. below under the var. pentactis. Ophiacantha vivipara, var. pentactis, n.var. (PlateVII, figs. 3,4) St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. 3 specimens. St. 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 259-354 m. 5 specimens. St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 93-130 m. 3 specimens. St. 599. 17. i. 31. 67° 08' S, 69° 06' W, 203 m. I specimen. In his work on the Echinoderms of the IP Expedition Antarctique Fran9aise ('Pourquoi-Pas?') Koehler has figured (pi. xi, i) a large five-armed specimen under the name of O. vivipara ; he regards this five-armed form as simply identical with the normally 6-7-rayed O. vivipara. Particularly he emphasizes the fact that there are all OPHIACANTHIDAE 249 possible transitions between the forms with the disk covered by a uniform granulation and those with numerous spines on the disk. Although these 5-armed specimens gener- ally have more numerous spines on the disk besides the granules, there is thus no reason for distinguishing them from the typical 6-7-armed O. vivipara ; and also in the other characters they agree, on the whole, with the typical vivipara. Still the matter is not quite so simple. It is undeniable that these 5-armed specimens in general have a much more robust appearance than vivipara (cf. Plate VII, figs. 2-4); also the arms are longer and more robust. This can hardly be due simply to the fact that they have fewer arms than the typical form. Then it is a very noticeable fact that this large 5 -rayed form is not met with among the specimens from the neighbourhood of the Falkland Islands, but only very far south, from the Palmer Archipelago to the Graham Land region. (Koehler's 5-rayed specimens are also all from this southern region.) If it were simply an individual varia- tion of the 6-7-rayed vivipara, it would be hard to understand why such forms should not occur equally commonly also in the Falkland region (the single 5-rayed specimen I have seen from there is only a very young one). The typical vivipara also occurs in the more southern region, together with the 5-rayed form; but whereas the 5-rayed form is of common occurrence in the south, it apparently does not occur farther north. Further, there are among these large 5-rayed specimens several males, whereas no male specimens were observed among the typical 6-7-rayed specimens. None of the female specimens of the 5-rayed form have young ones in the bursae, so it is quite possible that this form is not viviparous — at least there is no proof that it is viviparous. Thus, in my opinion, it is not justifiable simply to identify these specimens with the typical 6-7-rayed O. vivipara. If it is really non-viviparous, it must represent a separate species, but so long as we do not know this for certain, and in view of its resemblance with vivipara in general structure, I think it the safest course for the present to designate it as a variety of O. vivipara. From O. rosea, with which there is much general resemblance, it is distinguished particularly by the outer mouth papillae being simple, not forming a cluster at the outer mouth tube foot. From O. detisispina it diflfers in having both granules and spines on the disk, and in the arm spines being more smooth and not joining in the dorsal median line. Also the shape of the mouth shields is somewhat different (Figs. 5-6). Ophiacantha densispina, n.sp. (Plate VII, fig. i) St. WS 99. 19. iv. 27. 49° 42' S, 59° 14' W, Falkland Islands, 251-255 m. i specimen. St. WS 248. 20. vii. 28. 52° 40' S, 58^30' W, Falkland Islands, 210-242 m. i specimen. St. WS 825. 29. i. 32. 50° 50' S, 57° 15' W, Falkland Islands, 135-144 m. 2 specimens. St. WS 840. 6. ii. 32. 53° 52' S, 61° 49' W, Falkland Islands, 368-463 m. 3 specimens. The type specimen (from St. WS 248) is a large, coarse specimen, 16 mm. in diameter of disk. The arms are all broken, but judging from their size they must have been at least about five to six times the diameter of the disk; they are somewhat flattened, 7-2 25° DISCOVERY REPORTS Fig. 5. Part of oral side of Ophiacantha vivipara, Ljungman (a), and of O. vivipara, vai. pentactis, n.var. (b). x6. Fig. 6. ophiacantha densispina J n.sp. a, Part of oral side, x6. b, Dorsal arm plates, x8. r, Spicules from bursal wall, with prolongations forming spines protruding into the body cavity. Xi35. OPHIACANTHIDAE 251 3-5 mm. broad at the base. The other specimens are smaller and very poorly pre- served. The disk is covered, both on the dorsal and ventral side, by a rather dense coat of coarse spines, 1-1-5 m^n- long; they are smooth, simply pointed in the type specimen, in other specimens more blunt. At most the distalmost part of the radial shields is naked, but there may be a rather distinct naked line proceeding inwards in continuation of the radial shields. There are three to four rather long, slender mouth papillae on each side of the jaw, the outermost not, or only slightly, enlarged, and there are no extra papillae on the distal part of the jaws. The infradental papilla is not at all enlarged, rather smaller than the other papillae, but there are in the type specimen a few small extra papillae placed irregularly at the apex of the jaw. The buccal shields are of a very characteristic shape, almost rectangular, only slightly broader within than without; the adoral plates are small, almost square, sometimes with a more or less conspicuous outward prolongation which separates the buccal shield from the first lateral plate. The ventral plates, which are almost contiguous, have an almost straight proximal edge, the distal edge being convex; they are usually somewhat thickened and elevated in the distal part. A single large, leaf-shaped, more or less pointed tentacle scale. The dorsal plates are almost rhombical, almost contiguous proximally. Arm spines eleven to twelve in the proximal part of the arm, joining in the dorsal median line so as almost to conceal the dorsal plates. They are cylindrical, pointed, rather thorny in the basal part ; they increase very gradually in length upwards, the uppermost ones being as long as five to six arm joints. One of the specimens from St. WS 840 is of a dark brown colour, the other specimens whitish, evidently bleached. This species is viviparous, but appears to have separate sexes. The larger specimens (excepting the type and the specimen from St. WS 99, which are dried and could not be examined as to their sexual character) are males; only a small one, 5 mm. in diameter, from St. WS 825, is a female with young ones in the bursae (I found one large and one very small young one in the same bursa). An interesting anatomical feature is found in this species, viz. that the two bursae at each arm have coalesced above the dorsal side of the arm. They are heavily plated, some of the plates prolonged into irregular spines, turning into the body cavity — to my know- ledge a unique feature (Fig. 6 c). There is undeniably much resemblance between this species and Ophiaca?itha rosea, Lyman, taken by the ' Challenger', also off' Patagonia (St. 308). It is, however, beyond doubt that they are not identical, the main differences being found in the covering of the disk (small, short stumps of O. rosea) and in the mouth papillae, O. rosea having a cluster of distal papillae. Further, the bursae are not coalesced in O. rosea, as I can state, having examined a cotype of that species. It appears that O. rosea is not viviparous. On examining the specimens of O. rosea in the British Museum I found that one from 'Challenger' St. 308 and two specimens from Tom Bay, Patagonia, are infested with Myzostoma, particularly at the bursal slits, one being even wholly within the bursa. 252 DISCOVERY REPORTS Ophiacantha disjuncta (Koehler) Ophiacantha antarctica, Koehler, 1901. Result. Voyage ' Belgica '. Echinides at Ophiures, p. 34, pi. iv, figs. 23-25. Non: Ophiacantha antarctica (Lyman). Ophiodiplax disjuncta, Koehler, 191 1. Brit. Antarct. Exped., 1907-9. Asteries, Ophiures et Echinides, p. 48, pis. vi, figs. 9-1 1 ; vii, fig. 13. Ophiacantha antarctica, Koehler, 1912. IP Exped. Antarct. Fran^aise. Echinodermes, p. 137. Ophiodiplax disjuncta, Koehler, 191 2. Ibid., p. 142. O. disjuncta, Koehler, 1922. Austral. Antarct. Exped. Echinod. Ophiuroidea, p. 15, pi. Ixxviii, figs. 4-5,9-12. O. disjuncta, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. 105. O. disjuncta, G. A. Smith, 1923. Report on the Echifioderms coll. during the voyage of the ' Quest'. Ann. Mag. Nat. Hist., 9 Ser., xii, p. 369. O. disjuncta, Mortensen, 1925. On a small collection of Echinoderms from the Antarctic Sea. Arkiv for Zoologi, xvii A, No. 31, p. 2. Ophiacantha disjuncta. Hertz, 1926. Deutsche Sudpolar-Exped. Ophiuroiden, p. 38, Taf. vii, fig. 5. O. antarctica, Grieg, 1929. Some Echinoderms from the S. Shetlands. Bergens Mus. Arbok, 1929, p. 8. St. 20. 4. iii. 26. 14-6 miles N 41° E of Cape Saunders, South Georgia, 200 m. i specimen. St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, no m. 12 specimens. St. 39. 25. iii. 26. East Cumberland Bay, South Georgia, 179-235 m. 3 specimens. St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 ^- Several speci- mens. St. 123. 15. xii. 26. Off mouth of Cumberland Bay, South Georgia, 230-250 m. 5 specimens. St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. 2 speci- mens. St. 144. 5. i. 27. Off mouth of Stromness Harbour, South Georgia, 155-178 m. Several specimens. St. 148. 9. i. 27. Off Cape Saunders, South Georgia, 132-148 m. i specimen. St. 152. 17. i. 27. 53° 51' S, 36° 18' W, South Georgia, 245 m. 2 specimens. St. 156. 20. i. 27. 53° 51' S, 36° 21' W, South Georgia, 200-236 m. i specimen. St. 159. 21. i. 27. 53° 52' S, 36° 08' W, South Georgia, 160 m. 2 specimens. St. 160. 7. ii. 27. Near Shag Rocks, South Georgia, 177 m. i specimen. St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. Several specimens. St. 175. 2. iii. 27. Bransfield Strait, South Shetlands, 200 m. Several specimens. St. 177. 5. iii. 27. 27 miles SW of Deception Island, South Shetlands, 1080 m. i specimen. St. 180. II. iii. 27. Schollaert Channel, Palmer Archipelago, 160 m. i specimen. St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-335 m. 5 specimens. St. 182. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-500 m. i specimen. St 187. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 259-354 m. 15 specimens. St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 90-130 m. and 315 m. Several speci- mens from both depths. St. 195. 30. iii. 27. Admirahy Bay, King George Island, South Shetlands, 391 m. 15 speci- mens. St. 363. 26. ii. 30. Off Zavodovski Island, South Sandwich Islands, 329-278 m. 8 specimens (i 6-rayed). St. 599. 17.1.31. 67° 08' S, 69° 06' W, 203 m. 10 specimens. St. 600. 17. i. 31. 67° 09' S, 69° 27' W, 501-527 m. 4 specimens. St. WS 42. 7. i. 27. 54° 42' S, 36° 44' W, South Georgia, 198 m. 12 specimens. OPHIACANTHIDAE 253 Some of the specimens are infested by the ectoparasitic Copepod Cancerillopsis, one of them (St. 599) by the entoparasite Ophioika. There cannot be the sUghtest doubt that Koehler's Ophiodiplax disjuncta is identical with the species described by him in 1901 under the name Ophiacantha antarctica. From the descriptions and figures this is indeed quite evident ; it is true the division of the dorsal arm plates was not observed by Koehler in his O. mitarctica, but in a co-type sent me from the Brussels Museum I find them very well developed. No further dis- cussion on the question of the identity of the two "species" is needed. It is curious that Koehler, in recording both O. antarctica and Ophiodiplax disjuncta in his work of 1912 did not notice their similarity. Fig. 7. Ophiacantha disjuncta (Koehler). Part of oral side of two specimens, showing variation in the development of the mouth papillae. X12. The name antarctica, though the older of the two, cannot be used for the species, as there already is the Ophiacantha antarctica (Lyman) originally referred by Lyman to Ophioconis, but quite evidently an Ophiacantha ; thus the name disjuncta has to be used for this species. As for the genus Ophiodiplax, Hertz has rejected it and included the species in the genus Ophiacantha, and I quite agree that the characters of the divided dorsal plates and the more or less conspicuous continuation of disk spinules on to the dorsal side of arms do not afford sufficient basis for making the species the type of a separate genus.There is, however, another character which might be of generic value — the multiplication of the mouth papillae. In some specimens there are a great number of papillae, placed mainly below the normal ones inside the mouth slits (Fig. 7^) ; more generally there are only a few of these supernumerary papillae (Fig. 76), sometimes only one, very rarely none. As pointed out by Koehler in his description of O. antarctica this recalls Verrill's 254 DISCOVERY REPORTS genus Ophiectodia, and if we were to recognize that genus, Ophiodiplax would be a synonym of it. But I do not think that genus sufficiently well founded. One of the specimens from St. 363 is 6-rayed, but otherwise conforms with the typical disjiincta. This species is not viviparous; but the eggs are very large, so that there is hardlythepossibilitythatthelarvaOp///o/)/? 25. are rather broad, about half the length of the disk radius, and contiguous in their distal half. Outside the radial shields there is a rounded, knob-shaped plate which carries one to three small hyaline thorns. The ventral interradii are in one specimen somewhat sparsely covered with rounded scales, in the other almost naked. The buccal shields are rounded triangular, though with a small rounded outer lobe. The adoral plates are of rather unusual shape, with a straight inner side and a very broad outer lobe, widely separating the first lateral plate from the buccal shield. The outermost mouth papilla is small, scale-like, the second long, pointed, spine-like; the third is short, slightly pointed. The infradental papillae of moderate size, a little pointed. First ventral plate rather large, with convex distal edge. The following ventral plates distinctly longer than broad, contiguous, with sides and outer edge forming a slight re-entrant curve. Dorsal arm plates almost circular, not contiguous. Arm spines four, on the pro.ximal joints, very slender and pointed. One small tentacle scale. No indication of particular colour. AMPHIURIDAR 297 Both specimens have the mouth very widely open. The disk is partly loosened from the arms and has contracted a little so as to become somewhat concave. The species appears to be liable to lose its disk. This species appears to be the nearest related to the Japanese species Amphiodia digittila of H. L. Clark (North Pacific Ophiurans, p. 162, fig. 70; Matsumoto, Mono- graph of Japanese Ophiuroidea, p. 199, fig. 54) in which there is likewise a small spiny plate outside the radial shields. That the latter is referred to the genus Amphiodia, the present species to Amp/iiop/m, is no serious objection to regarding them as related, the distinction between these genera not being sharp. As pointed out by Matsumoto the mouth papillae of A. digitida approach those of Amphiopliis, and the present species might perhaps rather be referred to Amphiodia— it all depends on how strictly we limit those genera according to the number of the mouth papillae. The character of the mouth papillae of the present species also recalls Amphioplus acntiis (cf. Fig. 26 a, p. 295), but evidently there is no near relation between these two species. Possibly the specimens in hand are only young ones ; as they have been dried the condition of the genital organs cannot be ascertained. Amphioplus peregrinator, Koehler (Plate VII, figs. 12-15) Amphioplus peregrinator, Koehler, 1912. IP Exped. Antarct. Fran9aise. Echinodermes, p. 135, pi. xi, figs. 5, 11-12. 5. 2. iii. 27. Bransfield Strait, South Shetlands, 200 m. i specimen. ). II. iii. 27. Schollaert Channel, Palmer Archipelago, 160 m. 6 specimens. [. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-335 m. i specimen. >. 14. iii. 27. Schollaert Channel, Palmer Archipelago, 278-500 m. i specimen. J. 16. iii. 27. Fournier Bay, Anvers Island, Palmer Archipelago, 295 m. i specimen. 7. 18. iii. 27. Neumayr Channel, Palmer Archipelago, 259-354 m. 2 specimens. D. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 98-130 m. 6 specimens. There can be no doubt about the identity of these specimens with Koehler's Amphio- plus peregrinator , described {op. cit.) from a single specimen collected by the ' Pourquoi- Pas? '. Some additional information may be gathered from these specimens, the largest of which reach a diameter of disk of c« . 1 1 mm . The incision in the interradial edges of the disk is not a constant character of the species ; on the contrary, the disk, particularly in the larger specimens, often bulges out interradially, and may be considerably swollen. The proximal ventral plates are usually somewhat elevated. The arm spines are often four on the proximal joints in the larger specimens. The primary disk plates are usually very conspicuous, particularly in the young specimens. The specimen from St. 182 shows distinct traces of colour; the ventral side of the arms and the adoral shields are a conspicuous orange, the same colour being also found on the radial shields and the primary plates of the disk. Also in dried specimens these plates may stand out lighter against the general dark grey colour of the disk. In the 13-2 St. 180. St. 181. St. 182. St. 186. St. 187. St. 190. 298 DISCOVERY REPORTS specimen from St. 182 the dorsal side of the arms also shows a faint trace of orange colour. The species has separate sexes, and is not viviparous. The type specimen was taken at Port Lockroy, Palmer Archipelago, 70 m. ; the species is thus as yet known only from the Palmer Archipelago and South Shetlands, 70- ca. 300 m. From St. 182 there are a number of very young Ophiurids, which are not identifiable with certainty; but there are indications that they would have developed into Am- phiurids, and not improbably Amphiura peregrinator . Unfortunately, the intermediate stages are lacking, so that certainty cannot be reached, and it seems, therefore, not desirable to give a description of these young specimens, the general appearance of which is not at all Amphiurid-like. Family OPHIOCHITONIDAE Ophionereis sexradia, n.sp. St. 283. 14. viii. 27. Off Annobon, Gulf of Guinea, 18-30 m. 9 specimens. Diameter of disk of largest specimen 6 mm. ; arms six, about four times the diameter of the disk. Scales of dorsal side of disk very small, but distinct, imbricating; they are slightly larger round the radial shields, which latter are small, about one-quarter the length of the disk radius, oval, separated by several rows of scales. No primary plates discernible. The ventral interradii look as if they were naked in the proximal half; in reality they are scale-covered throughout, as can be substantiated in dried specimens, but the skin is more or less conspicuously dark coloured in this proximal part, which conveys the im- pression that it is naked. Buccal shields spade-shaped, with a short, rounded distal lobe. Adoral plates narrow, but joining within. Usually four equal-sized mouth papillae on each side of jaw. Ventral arm plates somewhat longer than broad, contiguous, with convex distal edge and sides with a re-entrant curve. Dorsal arm plates broadly con- tiguous, somewhat longer than broad, with nearly straight distal edge. The supple- mentary plates of moderate size, but quite distinct. Three slender, equally long arm spines, slightly curved and flattened, of about the length of an arm joint. One large, oval tentacle scale. Colour in alcohol: the disk light greenish, with an elongate brownish spot at the base of each arm. Underside whitish, apart from the above mentioned dark proximal part of the interradii, this colour being evidently due to the stomach shining through the delicate body wall. Arms (and spines) whitish, or cream- coloured, with an irregular band of brownish on every fifth to sixth joint, the inter- vening part dotted with small brownish spots, in the main arranged in two longitudinal series. Although in regard to the configuration of the plates and its general characters this species does not diflPer markedly from the other known species of the genus Ophionereis, it does so by having six arms, all the other species having five arms. Moreover, it appears OPHIOCHITONIDAE 299 to practise self-division, three of the specimens in hand having the three arms in re- generation. This is not known to be the case in any other species of Ophionereis, as, indeed, self-division does not normally occur in five-armed Ophiurids. The fact that only three specimens out of nine show any trace of self-division indicates that this method of propagation is not constant, but evidently occurs only in a certain percentage of the specimens ; still it is common enough to be regarded as a normal feature of the species. In his memoir on the Echinoderms in Michaelsen's Meeresfauna Westafrikas Koehler does not record any species of Ophionereis from the African coast. Evidently he overlooked that Marktanner-Turneretscher {Beschreibiingen neiier Ophiiiriden und Bemerkimgen zu bekannten. Ann. K. K. Naturhist. Hofmuseums, 11, 1887, p. 301) has Fig. 28. Ophionereis sexradia, n.sp. Part of oral side {a) and dorsal side {b). -20. recorded a small specimen of Ophionereis reticulata from " Westafrika". It appears, however, that there must be some error in regard to this specimen. The locality given for it, 0° 7' N, 23"^ 25' W, is about in the middle of the Atlantic Ocean in very great depths. That an Ophionereis reticulata (or any other Ophionereis) should occur here may well be said to be out of the question, and the label of this specimen must evidently be wrong. If Ophionereis reticulata does actually occur off West Africa it is rather strange that it has not hitherto been met with there. At any rate, Ophionereis sexradia is the first species of Ophionereis that has been actually found to occur on the West African coast. Ophionereis novae-zelandiae, n.sp. St. 934. 17. viii. 32. 34° 11' S, 172° 10' E, Cook Strait, New Zealand, 98 m. 1 specimen. Diameter of disk 4 mm. ; arms ca. 30 mm. long, thus some seven to eight times the diameter of disk. Scales of the dorsal side of disk rather coarse, uniform; the primary plates are small, but distinct, and form a fairly conspicuous rosette. The radial shields are small, oval. 300 DISCOVERY REPORTS scarcely one-third of the disk radius. The scales of the ventral interradii of the same size as those of the dorsal side of disk, rather thick. Buccal shields spade-shaped, with rounded corners and a small rounded outer lobe. Adoral plates narrow, joining within, with a conspicuous outer lobe separating the first lateral plate from the buccal shield. Mouth papillae as usual in Ophionereis. First ventral plate squarish, but broader dis- tally. The following ventral plates broadly contiguous, about as broad as long, with convex outer edge and sides with a re-entrant curve. The outer corners not much pro- duced. Dorsal arm plates rather squarish, with sides and distal edge almost straight. Fig. 29. ophionereis novae-zelandiae, n.sp. Part of oral side {a) and dorsal side {b). Arm joints from middle of arm of a larger specimen; dorsal side (c). X20. 22-5. The supplementary plate fairly large and distinct, almost rectangular in the proximal part of the arm. Three slender arm spines, about the length of an arm joint. One large, oval tentacle scale. Colour of dried specimen whitish, with a faint indication of brownish bands on the arms. This species is about intermediate between Ophionereis porrecta, Lyman, and O. aiistraUs (H. L. Clark). From the former it differs notably in the shape of both dorsal and ventral arm plates. The arm spines also are longer and more slender than in O. porrecta, and the colour is lighter. From the South African O. aiistralis it differs in the disk scales being smaller and more uniform ; the shape of the dorsal plates and the supplementary plates is also rather different (cf. Mortensen, Echinoderms of South OPHIODERMATIDAE— OPHIOLEPIDAE 301 Africa, fig. 77, p. 375). On the whole, in spite of the unfortunate fact that only a single specimen is at hand, and we thus do not know whether it is adult or only a young specimen, there can be no doubt that it is a distinct species, the genus Ophionereis being thus represented by two species in New Zealand seas. To the other New Zealand species, Ophionereis fasciata (Hutton), the present species is not closely related. Family OPHIODERMATIDAE Pectinura cylindrica (Hutton) Pectimtra cylindrica, Mortensen, 1924. Ecliinoderms of New Zealand and the Auckland-Campbell Islands. II, Ophiwoidea. Papers from Dr Th. Mortensen 's Pacific Exped., XX (Vid. Medd. Dansk Naturh. Foren., 77), p. 172, figs. 35, 1-2. P. cylindrica, Mortensen, 1925. Ibid., III-V, p. 391. For earlier literary references, see my paper of 1924, he. cit. St. 941. 20. iii. 32. 40° S3' S, 174° 47' E, Cook Strait, New Zealand, 128 m. Numerous specimens. I may recall the fact that the species is viviparous and hermaphrodite (op. cit., 1925). Ophioderma longicauda, var. guineense, GreeflF Ophioderma guineense, Greeff, 1881. Echinodermen beobachtet auf einer Reise nach der Guinea- Insel Sao Thome. Zool. Anzeiger, v, p. 156 O. longicauda, var. guineense, Koehler, 1914. Meeresfauna Westafrikas. Echinoderma, p. 173, pi. ix, figs. 1-3. St. 283. 13. viii. 27. Off Annobon, Gulf of Guinea, 18-30 m. 4 specimens. I quite agree with Koehler that this form from the Gulf of Guinea is not sufficiently different from the West Atlantic and Mediterranean O. longicauda to rank as a separate species. It may even be doubted whether it deserves the rank of a separate variety. But it is not the place here to enter on a detailed study of this question. Family OPHIOLEPIDAE Ophiozonella falklandica, n.sp. St. WS 212. 30. V. 28. 49° 22' S, 60° 10' W, 242 m. ca. 10 adult specimens, and a great number of young ones. St. WS 244. i8. vii. 28. 52° 00' S, 62° 40' W, 253 m. Several specimens. St, WS 818. 17.1.32. 52° 31' S, 63° 25' W, 272-278 m. 4 specimens. St. WS 819. 17. i. 32. 52° 45' S, 62° 27' W, 329-242 m. 4 specimens. St. WS 820. 18.1.32. 52° 53' S, 61° 51' W, 351-367 m. 6 specimens. St. WS 821. 18.1.32. 52° 56' S, 60° 55' W, 461-468 m. I specimen. St. WS 839. 5. 11. 32. 53° 30' S, 63° 29' W, 403-414 m. i specimen. St. WS 871. I. iv. 32. 53° 16' S, 64° 12' W, 336-341 m. I specimen. Diameter of disk of largest specimen 10 mm. ; arms rather robust, scarcely exceeding three times the diameter of disk. Dorsal side of disk covered with coarse, but smooth scales, among which the primary plates are usually very conspicuous ; in the younger specimens there are only some few 302 DISCOVERY REPORTS small plates in the corners between the primary plates, in larger specimens the primary plates are wholly separated by smaller plates. Generally there are two larger plates in the interradii. The radial shields are small, oval, widely separated by a series of two to three plates. The ventral interradii are covered by a varying number of plates, none of which are particularly conspicuous. The genital slits are narrow and short, not reaching beyond the end of the first lateral plate. The buccal shields are slightly irregular, with an acute inner angle; the inner sides concave, the outer edge convex; they may be distinctly longer than broad, or equally long and broad, there being thus a rather considerable variation in their shape. The adoral shields are short, broad distally, narrowing towards Fig. 30. Ophiozonella falklaiidica, n.sp Part of oral side (a). O. amitinum, Studer, 1885. Vbersicht uber die Ophiuriden der' Gazelle' Abh Akad Berlin 1882, p. 16, Taf. ii, fig. 8 a-f. O. amitinum, Ludwig, 1899. Ophiuriden Hamburger Magalh. Sammelreise, p. 4. O. amitinum, H. L. Clark, 1923. Echinoderm Fauna of South Africa. Ann! S. African Mus., XIII, p. 363. (Var. si?iiulans, Mortensen.) 336 DISCOVERY REPORTS O. amitinum, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. 122. O. (uiiititiian, Mortensen, 1933. Echinoderms of South Africa. Papers from Dr Th. Mortensen's Pacific Exped., Lxv (Vid. Medd. Dansk Naturh. Foren., 93), p. 390. (Var. simulans, Mortensen.) St. 91. 8. ix. 26. Off Roman Rock, False Bay, South Africa, 35 m. i specimen (type of var. simulcms, n.var.). St. 159. 21. i. 27. 53° 52' S, 36° 08' W, 160 m. I specimen. St. 160. 7. ii. 27. Near Shag Rocks, South Georgia, 177 m. 12 specimens. St. 474. 12. xi. 30. I mile W of Shag Rocks, South Georgia, 199 m. 10 specimens. St. WS 93. 9. iv. 27. 7 miles S 80° W of Beaver Island, West Falkland Islands, 133 m. 2 specimens. St. WS 99. 19. iv. 27. 49° 42' S, 59° 14' W, 251 m. I specimen. St. WS 210. 29. V. 28. 50° 17' S, 60° 06' W, 161 m. Numerous young specimens. St. WS211. 29. V. 28. Same locality as St. WS 210. Numerous young specimens. St. WS 212. 30. V. 28. 49° 22' S, 60° 10' W, 242 m. Numerous young specimens. St. WS 213. 30. V. 28. Same locality as St. WS 212. Numerous young specimens. St. WS 214. 31. V. 28. 48° 25' S, 60° 40' W, 208 m. Numerous young specimens. St. WS215. 31.V. 28. 47° 37' S, 60° 50' W, 219 m. Several young specimens. St. WS 216. i.vi. 28. Same locality as St. WS 215. Numerous young specimens. St. WS 227. 12. vi. 28. 51° 08' S, 56° 50' W, 320 m. 5 specimens. St. WS 229. I. vii. 28. 50° 35' S, 57° 20' W, 210 m. Several specimens. St. WS 231. 4. vii. 28. 50° 10' S, 58° 42' W, 167-159 m. Several young specimens. St. WS 233. 5. vii. 28. 49° 25' S, 59° 45' W, 185-175 m. Numerous young specimens. St. WS 234. 5. vii. 28. 48° 52' S, 60° 25' W, 195 m. Several young specimens. St. WS 235. 6. vii. 28. 47° 56' S, 61° 10' W, 155 m. Several young specimens. St. WS 236. 6. vii. 28. 46° 55' S, 60° 40' W, 272 m. Several young specimens. St. WS 237. 7. vii. 28. 46° 00' S, 60° 05' W, 150 m. Numerous young specimens. St. WS 244. 18. vii. 28. 52° 00' S, 62° 40' W, 253 m. Several young specimens. St. WS 248. 20. vii. 28. 52° 40' S, 58° 30' W, 210-242 m. i specimen. St. WS 748. 16. ix. 31. 53° 41' S, 70° 55' W, 300 m. 2 specimens. St. WS766. 18. X. 31. 44° 58' S, 60° 05' W, 545 m. 4 specimens, young. St. WS 772. 30. X. 31. 45° 13' S, 60° 00' W, 162-309 m. 5 specimens, young. St. WS 773. 31.X. 31. 47° 28' S, 60° 51' W, 291-296 m. Very numerous young specimens. St. WS781. 6. xi. 31. 50° 30' S, 58° 50' W, 148 m. I specimen. St. WS 782. 4. xii. 31. 50° 28' S, 58° 30' W, 141-146 m. 8 specimens, in poor condition. St. WS 783. 5. xii. 31. 50° 03' S, 60° 10' W, 155-159 m. Numerous young specimens. St. WS 784. 5. xii. 31. 49° 48' S, 61° 05' W, 164-170 m. 7 specimens. St. WS818. 17. i. 32. 52° 31' S, 63° 25' W, 272-278 m. 2 specimens. St. WS 819. 17.1.32. 52° 42' S, 62° 39' W, 312-329 m. I specimen. St. WS 824. 19. i. 32. 52° 29' S, 58° 27' W, 137-146 m. ca. 20 specimens, in poor condition. The very numerous young specimens (by the hundred thousand) bear witness to the excellent food conditions that must exist in Falkland seas ; it seems beyond doubt that this Ophiurid must be a factor of considerable importance in the ecology and economy of these seas. Plate VIII, fig. 2, represents a photo of a live specimen ; it is stated to have the " upper side of disc deep purplish brown, with white plates; arms deep brown above". Some few of the preserved specimens still show distinct traces of this coloration. That this species is not viviparous has already been pointed out by Ludwig {op. cit.) ; I may add that it has, as was to be expected, separate sexes. OPHIOLEPIDAE 337 The specimen from St. 191, False Bay, South Africa, cannot simply be identified with the South American specimens of Ophiocten amitmum. It differs from the latter in the character of the arm comb, in the arms being more distinctly carinate, and in the colour, the arms being distinctly banded, with alternating white and dark, brownish bands. The more important difference is in the arm comb. In the South African form the comb continues downwards, along the genital slit, which it does not do in the South American form; further there is a distinct inner comb, whereas in the South American form there is no such distinct inner comb (Fig. 48 a, b). In my Echinoderms of South Africa {he. cit.) I pointed out the close resemblance between this South African form and the North Atlantic Ophiiira ajfmis, Liitken. As a Fig. 48. Part of dorsal side of Ophiocten amitinum, Lyman {a) and of the var. simulans, n.var. (b). a is from a specimen 6 mm. in diameter of disk, b from a specimen 5'5 mm. in diameter. X22"5. matter of fact, I do not see how they can be distinguished, and I am very much tempted to regard the South African ^^ Ophiocten amitinum" as identical with Ophiura (iffinis. I do not do so here for two reasons; first, because I have not seen any specimen of Ophiocten omitimim from the type locality, off Kerguelen — perhaps the Kerguelen specimens will prove to differ from those from South America and be more like the South African form; and then we do not know O. ajfinis from the West African Coast, unless the Ophiocten africamim of Koehler should prove to be identical with affinis (in my Echinoderms of South Africa, p. 391, I have expressed the opinion that it is more nearly related to the Mediterranean O. Griibei, Heller). In view of these uncertainties I think it preferable for the present to designate the South African form as a variety of Ophiocten amitinum, var. simulans, n.var. 18-2 338 DISCOVERY REPORTS Ophiocten dubium, Koehler Ophiocten dubium, Koehler, 1901. Result. Voyage 'Belgica'. Echinides et Ophiures, p. 20, pi. vi, figs. 40-1. O. dubiutn, Koehler, 191 2. IF Exped. Antarct. Fran9aise. Echinodermes, p. 129. O. dubium, subsp. gaussense, Hertz, 1926. Deutsche Siidpolar-Exped. Ophiuroiden, p. 15, Taf. ii, figs. 4-5. St. 170. 23. ii. 27. Off Cape Bowles, Clarence Island, 342 m. i specimen. St. 190. 24. iii. 27. Bismarck Strait, Palmer Archipelago, 130 m. 2 specimens. These are perfectly typical specimens of this remarkable Ophiocten. The specimen from St. 170 is adult, 9 mm. in diameter of disk. It is a ripe male, full of large testes; this shows the species to have separate sexes and to be, in all probability, non-viviparous. The specimens from St. 190 are young, 3-5 mm. in diameter of disk. I do not think the subspecies gaussense of Hertz valid. The character on which it is based, eight arm spines, instead of nine to ten in the typical form, and somewhat longer than in the latter, is quite unimportant, and no doubt subject to a good deal of variation, as is usually the case in Ophiurids with numerous spines. Since only a very small number of specimens of this species have as yet been recorded, twelve in all, it seems unjustifiable to establish a separate subspecies on this character. Ophiocten bisquamatum, n.sp. St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 n"*- 3 specimens. St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. i specimen. St. 152. 17. i. 27. 53° 51' S, 36° 18' W, South Georgia, 245 m. i specimen. Diameter of disk of largest specimen 3-5 mm. Arms broken, but apparently not more than ca. 12-15 mm. in length. The disk is flattened, but the edge is rather high, rounded, not at all sharp as in most species of Ophiocten. There is usually a large, circular central plate which is somewhat eccentric (one of the specimens has no central plate). The disk is otherwise covered by rather large, imbricating plates, among which the primary plates are not distinct. Along the edges of the plates are found some very small, slightly elevated scales, like very low, nearly flat granules. The radial shields are distinctly broader than long, contiguous distally or wholly separated. The disk scales on the whole are exceedingly thin and delicate, looking dark (transparent) in the middle, whitish along the edges, where they imbricate. The ventral interradii show only some few small plates outside the large, squarish, rounded buccal shields. Mouth papillae four at each side of the jaw, the proximal ones pointed. First ventral plate large, polygonal, with convex distal edge. The second ventral plate contiguous with the first, the third nearly contiguous with the second, the following ones widely separated. All the ventral plates with strongly convex distal edge. Dorsal arm- plates rectangular, with outer edge convex; they are distinctly longer than broad, and broadly contiguous. Three arm spines, the uppermost one the longest, only little longer than an arm joint in the proximal part of the arm. Tentacle scales at the first pore pair one at the adradial, three at the interradial side; the outer one of the three latter is elongate, spine-like, the others more leaf-shaped. The following pores have two, some- OPHIOLEPIDAE 339 times three, spine-like tentacle scales, continuing at least till the middle of the arm, the inner one then gradually becoming smaller and disappearing. The tentacle scales are very like the lower arm spines, and not much shorter, so that it is not easy to tell which are spines and which tentacle scales. The genital slits are wide, with the merest indica- tion of papillae distally. Arm comb consisting only of three or four papillae at the distal edge of the radial shields. No papillae across the base of the arms or along the distal edge of the first dorsal arm plates. Colour of the dried specimens whitish. Fig. 49. Ophiocten bisquamatum , n.sp. Part of oral (a) and dorsal side (b). xzz'^. This is an interesting species, recalling by its large, eccentric central plate O. niegalo- plax, Koehler, with which it also agrees in the shape of the dorsal arm-plates. The small granules of the disk recall O. diibium ; but in the character of the tentacle scales it stands quite apart from the other known species of Ophiocten. On the sexual characters I can give no information. On opening an interradius of the largest specimen I found no genital organs developed, which seems to indicate that none of the specimens are adult. I may take the opportunity here of stating that OpJiiocten megalophx has separate sexes and is not viviparous. Dictenophiura anoidea, H. L. Clark Dictenophiura anoidea, H. L. Clark, 1923. Echinoderm Fauna of South Africa. Ann. S.African Mus., xin, p. 361, pi. xix, figs. 1-2. Ophiura carnea, Hertz, 1927. Deutsche Tiefsee-Exped. Ophiuroiden, p. 69 {Nan: Ophiura carnea, M. Sars). Dictenophiura anoidea, Mortensen, 1933. Echinoderms of South Africa. Papers from Dr Th. Mortensen's Pacific Exped., LXV (Vid. Medd. Dansk Naturh. Foren., 93), p. 388. St. 91. S. ix. 26. Off Roman Rock, False Bay, South Africa, 35 m. 3 specimens. 340 DISCOVERY REPORTS Dictenophiura Skoogi (Koehler) Ophiura Skoogi, Koehler, 1923. Sur quelques Ophiures des cotes de V Angola et du Cap. Goteborg. K. Vet. Vitterhets-Samhalles Handlingar, xxv, 5, p. 11, figs. lo-ii. Dictenophiura Skoogi, Mortensen, 1933. Echinoderms of South Africa. Papers from Dr Th. Mortensen's Pacific Exped., lxv (Vid. Medd. Dansk Naturh. Foren., 93), p. 390, fig. 87 b. St. 279. 10. viii. 27. Off Cape Lopez, French Congo, 58-67 m. 15 specimens. There can be no doubt that these specimens are identical with Koehler's Ophiura Skoogi. Another question is whether this O. Skoogi differs really so much from the North Atlantic O. carnea that it can reasonably be regarded as a separate species. Koehler {op. cit., p. 14) points out quite a number of characters in which O. Skoogi differs from O. carnea. But I do not think a single one of them holds good, except, perhaps, that the radial shields are in general slightly larger in O. Skoogi than in carnea. The only notice- able difference I find is that the lateral plates are more swollen in Skoogi than in carnea, and also that the dorsal arm-plates are somewhat more swollen in the former. In my Echinoderms of South Africa I have given a figure showing that the buccal shields are considerably elongated in O. Skoogi, but this is no constant feature; there is so much variation in the size of the buccal shields that in this respect no reliable difference be- tween Skoogi and carnea (and anoidea) can be found. In the work cited above I have further stated that O. Skoogi differs from both carnea and anoidea in the primary disk plates being "wholly surrounded by small plates", which they are not in the two other species. This is a mistake, partly in Koehler's description, which states that these plates are " separees les unes des autres par une rangee de petites plaques ", partly in Koehler's fig. II, which apparently shows the large plates each surrounded by a circle of smaller plates. They are not so; but all the plates, in the specimens preserved in alcohol, are darker in the centre, a broad edge appearing whitish. This produces the effect seen in Koehler's fig. 11. The same feature is also observable in both carnea and anoidea. There thus remain, as the only characters distinguishing O. Skoogi from carnea, the more swollen dorsal and lateral arm-plates, and in addition the slightly larger radial shields, characters rather insufficient for specific distinction. I think that this form from the tropical coast of West Africa represents merely a variety of the North Atlantic carnea, which latter is recorded from as far south as the Cape Verde Islands. Before, however, the specific value of O. Skoogi is finally decided I think it desirable to have specimens from the north-west coast of Africa for comparison, and for this reason I shall for the present keep it as a separate species. It may be added that this species (or variety) has separate sexes and is not viviparous, as holds good also of O. carnea and anoidea. Amphiophiura Rowetti, G. A. Smith Amphiophiura Rozvetti, G. A. Smith, 1923. Report on the Echitioderms coll. during the voyage of H.M.S. 'Quest'. Ann. Mag. Nat. Hist., 9 Sen, xii, p. 370. St. 20. 4. iii. 26. 14-6 miles N 41" E of Cape Saunders, South Georgia, 200 m. i specimen. St. 27. 15. iii. 26. West Cumberland Bay, South Georgia, iiom. 3 specimens. St. 42. I. iv. 26. Off mouth of Cumberland Bay, South Georgia, 120-204 m. 3 specimens. OPHIOLEPIDAE 341 St. 45. 6. iv. 26. 27 miles S 85° E of Jason Light, South Georgia, 238-270 m. i specimen. St. 126. 19. xii. 26. 53" 58' S, 37" 08' W, South Georgia, 100 (-0) m. i specimen. St. 140. 23. xii. 26. Stromness Harbour to Larsen Point, South Georgia, 122-136 m. 1 specimen (young). St. 156. 20. i. 27. 53° 51' S, 36° 21' W, South Georgia, 200-236 m. 8 specimens (young). St. 363. 26. ii. 30. 2-5 miles S 80° E of SE point of Zavodovski Island, South Sandwich Islands, 329-278 m. I specimen. St. WS 33. 21. xii. 26. 54° 59' S, 35° 24' W, South Georgia, 130 m. 2 specimens. Although G. A. Smith in describing this species omitted to give a figure of the oral side, I can have no doubt that the present specimens belong to that species, the more so as they come from the type locality, off South Georgia. One point alone would seem to be in contradiction to the description given by Smith, viz. the mouth shields, which are stated by Smith to be very large, occupying the whole of the interbrachial area on the Fig. 50. Amphiophiura Rowetti, G. A. Smith. Part of oral side (a) and dorsal side (b); part of arm in side view (c). X 15. ventral surface. The figure of the oral side given here is not in accordance with this statement ; it shows the buccal shields rather small, occupying only the proximal half of the ventral interradius. There is, however, much variation in this respect, some of the other specimens having the buccal shields relatively larger, so as to cover more or less completely the whole ventral interradius ; this depends to a great extent on the state of contraction of the interradius of the specimen (due to food content or the sexual state at the time of capture — and also, of course, to age, the younger specimens having these plates relatively larger). There is no possibility of distinguishing more than one species from the character of the mouth shields. It may be mentioned that none of the spines in the distal part of the arms are trans- formed into hooks, such as are described by Hertz (Deutsche Tiefsee-Exped., Ophiu- roiden, pp. 77-9) for A. concava, Hertz, and A. trifoUiim, Hertz. The specimen from St. 363 (South Sandwich Islands) must, I think, likewise be referred to this species, in spite of the fact that the dorsal arm-plates are somewhat more 342 DISCOVERY REPORTS swollen than in the specimens from South Georgia. As, however, there is only one young specimen, 4 mm. in diameter of disk, from that station, I give this identification with some reserve. This species is viviparous mid hermaphrodite. There is one sausage-shaped testis at the adradial side of the bursa, sometimes also a second smaller one proximal to the large one. There is one female gonad at the interradial side of the bursa, with only one egg developing at a time, and it is almost certain that the development is intra-ovarial ; the developing eggs are large, 0-5 mm. in diameter. Amphiophiura gibbosa, n.sp. St. 175. 2. iii. 27. Bransfield Strait, South Shetlands, 200 m. 3 specimens. Diameter of disk of the largest specimen 6-5 mm., the arms rather stout, ca. 18 mm. in length, thus about three times the diameter of the disk. Disk covered by rather coarse 0 c Fig. 51. Amphiophiura gibbosa, n.sp. Part of oral side {a) and dorsal side {h). Part of arm in side view (c). x 12. scales, among which the polygonal central plate and five primary radial plates are con- spicuous, particularly the latter, which form a wedge between the proximal ends of the radial shields. These primary radial plates, together with the radial shields, are conspicuously elevated, and produce a somewhat gibbous appearance, recalling Ophiosteira Senoiiqid. The radial shields are contiguous in their distal half. In the younger specimens there is a single series of plates in the interradii, in the larger specimen several small plates have been added. No large plate on the edge of the interradii, also the ventral interradii are covered by small plates only, besides the buccal shields, which are fairly large, occupying about half of the interradius. In one case the proximal end of the buccal shield has been separated off as a small plate. Adoral and oral plates (jaws) of about equal size, in the OPHIOLKPIDAE 343 main flat, simple. Mouth papillae usually three to each side of jaw, of the common type. First ventral plate triangular, conspicuously broader than the following ones, which are of the shape usual in this genus ; the three or four proximal ones are contiguous. Dorsal arm plates contiguous in the proximal part of the arms, not much swollen. Arm spines three or four in the proximal part of arms, short, equidistant, none of them transformed into hooks. Arm comb well developed, but the papillae are short, not longer than the arm spines. Papillae along the genital slits rudimentary ; the slits reach to the edge of the disk. Tentacle pores large, with numerous scales or papillae, as usual in the genus. Colour of preserved specimens whitish. The species is viviparous. Probably it is also hermaphrodite, like A. Rowetti; but this I have been unable to ascertain definitely. Not thinking it desirable to spoil the type specimen, the only one preserved in alcohol, I have only opened one interradius from the ventral side, which showed merely the young embryos (with the skeleton just be- ginning to develop) lying as in ^4. Rowetti. The character of the gibbous elevation of the primary radial plates and the radial shields distinguishes this species markedly from A. Rowetti, as well as from the other species oi Amphiopliiura known from the Antarctic seas or elsewhere. The species seems the nearest related to A. Rowetti, of which it may perhaps ultimately prove to be only a variety. Ophiomastus conveniens, Koehler Ophiomastus conveniens, Koehler, 1923. Swedish Antarct. Exped. Asteries et Ophiures, p. 122, pi. XV, figs. 5-6. O. convejiiens, Grieg, 1929. Echinodermata from the Palmer Archipelago, p. 10. Sci. Results Norwegian Antarct. Exped., n, p. 10. St. 181. 12. iii. 27. Schollaert Channel, Palmer Archipelago, 160-33=; "i- ^ specimen. The specimen is a young one, 3 mm. in diameter of disk. It is in full accordance with the description and figures given by Koehler, excepting the fact that the proximal arm joints have only three spines, not four, as had Koehler's specimen. The genital slits have just begun to appear; they are very short, not exceeding one- third the length of the first lateral plate. Whether they will be longer in the adult does not appear from Koehler's description and figures, but the probability is that they remain quite short. On Koehler's statement that the plates of the disk are covered with a very fine and regular granulation it may be remarked that it is the calcareous structure of the plates that is rather coarse, so as to produce a granular appearance; there are no separate granules. No information on the sexual character of this species, whether viviparous or not, or hermaphrodite or not, can be gathered from the single young specimen examined. An interradius was opened, but showed nothing of the genital organs, which had evidently not yet developed. DXII ig 344 DISCOVERY REPORTS Ophiomusium constrictum, n.sp. St. WS871. i.iv. 32. 53° 16' S, 64° 12' W, 336-341 m. 2 specimens. Diameter of disk 5 mm. Arms about three times that length, apparently not very stiff. Disk slightly elevated, covered with rather coarse scales, among which the primary plates are not very conspicuous. Radial shields separated by a wedge of scales, some- what sunken below the level of the surrounding scales. The disk scales are flat, but of a rather coarse structure, as if finely granulated, this granular appearance being more distinct on the radial shields. There is only a single column of fairly large scales in the interradii. The ventral interradii covered with few, rather large, irregular scales. The buccal shields irregular, with the proximal point separated off as a distinct little plate. Fig. 52. Ophiomusium constrictum, n.sp. Part of oral side («) and dorsal side (A); part of arm in side view (c). . : 15. Mouth parts of the usual type. First ventral plate small, rectangular, with the merest indication of a proximal angle. Only the two first ventral arm plates developed, and of the dorsal arm plates at most that of the first arm joint is present and then quite rudi- mentary. The lateral plates are strongly developed, swollen distally, constricted proximally ; this constriction is limited from the swollen part by a rather distinct line and forms a kind of neck, the arms being thus rather distinctly moniliform. There are four rudimentary arm spines, the upper one placed a little apart from the others. Tentacle pores exceedingly poorly developed, only on the first joint are they at all distinct, totally lacking or merely discernible on the second joint. Genital slits very short and narrow, scarcely half the length of the first lateral plate. They are surrounded by some small irregular plates, the one to the adradial side being evidently the distal end of the adoral plate, those to the interradial side of the slit representing the proximal end of the genital scale. Colour of the dried specimens white, the radial shields slightly darker. OPHIOLEUCIDAE 345 This species appears to be the nearest related to Ophiomusium {Ophiomiisa) ultimo. Hertz (Deutsche Tiefsee-Exped., Ophiuroiden, p. 106, Taf. ix, figs. 1-3), in which the arm joints also have a neck-like constriction. Unfortunately, Hertz' figures of this species are not clear, but the much longer arm joints alone show that this East African species (trom off Zanzibar) is quite distinct from the present South American species. Family OPHIOLEUCIDAE Ophiopyren regularis, Koehler (Plate VIII, fig. i) Ophwpyrefi regulare, Koehler, 1901. Result. A'oyage 'Belgica'. Echinides et Ophiures, p. 26, pi. viii, figs. 52-4. O. regularis, Koehler, 1922. Austral. Antarct. Exped. Echinod. Ophiuroidea, p. 36, pi. Ixxxvi, figs. 1-2. St. 159. 21. i. 27. 53° 52' S, 36° 08' W, South Georgia, 160 m. 1 specimen. The single specimen of this species taken by the ' Discovery ' has a diameter of disk of 8 mm., being thus the largest of the few (in all seven) specimens known. The arms are all broken close to the disk, excepting one, of which a piece 8 mm. in length is preserved. A conspicuous feature not mentioned by Koehler is that the underside of the disk is a little concave, looking, indeed, like a large sucking disk. Evidently this means that the species lives attached to the surface of stones, corals and the like, such as is the case with Ophiophycis gracilis, Mortensen (Echinoderms of St Helena, p. 458), and Astro- phiiira (Mortensen, Echinoderms of South Africa, p. 396). The strong development of the tube feet is also in accordance with the suggestion that it lives thus attached. The specimen having one of its interradii broken to pieces, I could examine the gonads. They proved to be purely male in character ; thus it is certain that the species has separate sexes — which makes it probable, but does not necessarily imply, that it is not viviparous (cf. Ophiozonella falklandica). As appears from the figure given here, the specimen differs in some degree from the description and figures given by Koehler. The buccal shields are slightly different, but more so are the ventral plates ; in particular, I do not see these plates divided by a trans- verse line, as shown in Koehler's drawing {op. cit., 1901, pi. viii, fig. 53). However, owing to the faint calcification of the plates in this species (and in the Ophioleucids in general) the outlines of the plates are difficult to make out. The tentacle scales also are rather differently represented in Koehler's and in my drawings; this would, however, seem to be due to some inaccuracy on Koehler's part, the photographic figure given in his work of 1922 agreeing much better with the figure given here than with his figure in the Belgica Ophiurids. The scales at the proximal pores may be rather difficult to dis- tinguish ; they are sometimes hardly more than a raised edge, which is particularly the case on the adradial side of the first pore pair. The spines are of unequal length, the upper much shorter than the lower one. The granules along the edge of the disk do not 19-2 346 DISCOVERY REPORTS form a regular border, as is shown by Koehler (also in the photographic figure). The dorsal side of the disk has evidently been almost completely covered by the granules, only the central plate and a small part of the radial shields remaining bare. Finally the dorsal arm-plates show a characteristic feature not observed by Koehler, the proximal four or five plates having a series of granules along their distal edge (Fig. 53 b). I have not observed transverse lines as shown in Koehler's fig. 52. In spite of the differences here pointed out, I think the present specimen really identical with Koehler's Ophiopvren regulare. With the very scarce material available Fig. 53. Ophiopyren regulare, Koehler. Part of oral side {a) and dorsal side (/)). ,20. it would be unreasonable to lay much weight upon the various minor differences, which are due partly to inaccurate drawing, partly no doubt to individual differences, such as are known to occur to no small extent in these feebly calcified deep-sea forms. It would also be rather strange if two different species of these rare deep-sea forms should occur here in the same region. Plate VIII, fig. I, is a photograph of the living specimen. The colour is stated to have been "disc deep crimson red with a very large pale pentagonal patch in centre. Arms crimson red barred with cream colour." The preserved specimen does not show the slightest trace of this beautiful colour. INDEX Abatus, 224 abnorme (Ophiostigma), 293 aciculatus (Amphioplus), 296 acutispina (Amphiodia), 290 acutus (Amphioplus), 294 affinis (Amphiodia), 288 africamim (Ophiostigma), 293 Agassizii (Abatus), 228 Agassizii (Astrotoma), 236 Agassizii (Sterechinus), 217 Albida (Ophioceramis), 290 albus (Loxechinus), 223 alternans (Amphiura), 279 amitinum (Ophiocten), 335 Amphiodia, 288 Amphiophiura 340, Amphioplus, 294 Amphipholis, 292 Amphipneustes, 229 Amphiura, 266 angularis (Amphiura), 267 angulosus (Parechinus), 223 anoidea (Dictenophiura), 339 antarctica (Amphiura Eugeniae), 285 antarctica (Ophiacantha), 254 antarctica [Ophioceramis), 288 antarctica (Ophiocoim), 254 antarctica (Ophiosteira), 316 antarcticus (Sterechinus), 218 Arbacia, 215 ascensionis (Diadema antillarum), 216 ascia (Amphiodia), 290 asperula (Ophiactis), 262 Astroceras, 241 Astrochlamys, 237 Astrohamma, 239 Astrotoma, 236 Austrocidaris, 212 Belgicae (Amphiura), 279 bidens (Abatus cavernosus), 226 biscutifer (Op/iiokbes), 305 biscutifera (Ophiolebella), 305 bisquamatum (Ophiocten), 338 Bollonsi (Ophiocoma), 260 brevirima (Ophiomyxa), 242 brevirima (Ophiurolepis), 319 bruneus (Astrochlamys), 237 canaliculata (Austrocidaris), 212 capensis (Ophiomyxa vivipara), 242 carinata (Ophiurolepis), 317 carinifera (Opliioglypha), 314 cavernosus (Abatus), 224 Centrostephanus, 217 Chiajei (Amphiura), 287 Synonyms in italics chilensis (Gorgonocephalus), 240 connectens (Echinocardium), 234 constrictum (Ophiomusium), 344 conveniens (Ophiomastus), 343 cordatus (Parapneustes), 233 Costae (Plagiobrissus), 235 crassa (Ophiura flexibilis), 333 cristatus (Ophiomages), 313 Ctenocidaris, 209 cylindrica (Pectinura), 301 da Cunhae (Amphiura), 282 deficiens (Amphiura), 276 densipina (Ophiacantha), 249 Diadema, 216 Dictenophiura, 339 disjuncta (Amphiura microplax), 27 1 disjuncta (Ophiacantha), 252 divisa (Opliioglypha), 329 Doderleini (Ophiogona), 308 dubium (Ophiocten), 338 Dufresnii (Arbacia), 215 Echinocardium, 234 Echinometra, 224 echiniilata (Ophiosteira), 316 elegans (Astroceras), 241 elongatus (Abatus), 227 Eucidaris, 213 Eugeniae (Amphiura), 283 falklandica (Ophiomitrella), 256 falklandica (Ophiozonella), 301 falkhindicus (Ophiochondrus), 259 fragilis (Ophiothrix), 261 frigida (Ophioglypha), 324 gaussense (Ophiocten dubium). 338 Gaussi (Amphiura dilatata), 277 Geliberti (Ctenocidaris), 212 gelida (Ophiurolepis), 318 Genocidaris, 217 gibbosa (Amphiophiura), 342 Gorgonocephalus, 240 gracilis (Amphiura Eugeniae), 279 grandisquama (Amphiura), 269 guineense (Ophioderma longicauda), 301 guineensis (Amphiura grandi- squama), 269 guineensis (Ophiopsila), 260 //e/)/flf//s (Ophiacantha vivipara), 248 hexactis (Ophionotus), 312 Homalophiura, 327 incana (Amphiura), 286 incipiens (Ophioceres), 307 inermis (Ophiacantha angolensis), 255 inermis [Ophiozona), 322 ingrata (Ophiomitrella), 256 ingrata [Ophioripa), 256 inornata (Homalophiura), 327 Joubini (Amphiura), 277 kerguelensis (Ophiacantha vivipara) . 248 Koehleri (Amphipneustes), 229 Koehleri (Ophioperla), 310 Koehleri (Ophiura), 310 laevigata (Ophiogona), 309 Lorioli (Amphipneustes), 230 Loxechinus, 223 lucunter (Echinometra), 224 Ludivigi (Ophioperla), 310 Lymani (Amphiura), 274 Lymani (Ophiuroglypha), 316 maculata (Genocidaris), 217 magellanica (Amphiura), 266 magellanica (Ophionephthys), 291 magellanicus (Notechinus), 220 marionis (Notechinus), 221 marionis (Ophioscolex), 244 Martensi (Ophiurolepis), 321 megaloplax (Ophiocten), 339 megaloplax (Ophiozonella), 303 meridionalis (Ophiura), 330 microplax (Amphiura), 270 monorima (Amphiura), 272 Mortenseni (Amphipneustes), 230 Mortenseni (Amphiura), 279 Neumayeri (Sterechinus), 219 nidarosiensis (Ophiactis), 264 Nordenskjoldi (Plexechinus), 235 Notechinus, 220 Novae-Zelandiae (Ophiactis pro- fundi), 266 novae-zelandiae (Ophiocentnis), 287 novae-zelandiae (Ophionereis), 299 nudipora (Amphipholis), 293 nutrix (Ophioscolex (Ophiolycus)), 243 Ophiacantha, 246 Ophiactis, 262 Ophiocentrus, 287 Ophioceres, 307 Ophiochondrus, 259 Ophiocoma, 260 Ophiocten, 335 Ophioderma, 301 348 DISCOVERY REPORTS Opliiodiplax, 253 Ophiogona, 308 Ophiolebella, 305 Ophiolepis, 308 Ophiomages, 313 Ophiomaria, 310, 315 Ophiomastus, 343 Ophiomitrella, 256 Ophiomusium, 344 Ophiomyxa, 241 Ophionephthys, 291 Ophionereis, 298 Ophionotus, 311 Ophioperla, 310 Ophiopsila, 260 Ophiopyren, 345 Ophioripa, 258 Ophioscolex, 243 Ophiosteira, 314, 315 Ophiostigma, 293 Ophiothrix, 261 Ophiotreta, 255 Ophiozonella, 301 Ophiozonoida, 304 Ophiura, 330 Ophiuroglypha, 315, 316 Ophiurolepis, 317 Parapneustes, 233 Parechinus, 223 partita (Ophiurolepis), 325 patagonica (Amphipholis), 292 paucispina (Ophiolepis) 308 Pectinura, 301 pentactis (Ophiacantha vivipara), 248 peregrinator (Amphioplus), 297 Perrieri (Ctenocidaris), 211 Philippii (Abatus), 228 Philippii (Schizaster (Tripylaster)), 229 picta (Ophiozonoida), 304 Plagiobrissus, 235 platyspina (Ophiopsila), 261 Plexechinus, 235 polaris (Ophiacantha), 254 polita (Amphiura), 279 princeps (Amphiura), 285 protecta (Amphiura angularis), 268 reductus (Parapneustes), 233 regularis (Ophiopyren), 345 relegata (Amphiophiura), 324 resiliens (Ophiactis), 266 resistens {Ophioglypha), 322 roseo-coerulans (Ophiothrix), 262 rotimdus (Ophiomastus), 331 Rouchi (Ophiura), 332 Rowetti (Amphiophiura), 340 Savignyi (Ophiactis), 264 Schizaster, 229 seminuda (Ophiactis), 264 Senouqui (Ophiosteira), 314 serrata (Ophiura), 334 sexradia (Ophionereis), 298 similis (Amphipneustes), 231 simulans (Ophiocten amitinum), 337. Skoogi (Dictenophiura), 340 sol (Astrochlamys), 239 speciosa (Ctenocidaris), 209 spinipes (Amphiura), 267 squamata (Amphipholis), 292 stelliger (Ophiochondrus), 259 Sterechinus, 217 Studeri (Amphiura), 285 tomentosa (Amphiura), 275 tribuloides (Eucidaris), 213 triglochis (Ophiothrix), 261 Tripylaster, 229 tuberculatum (Astrohamma), 239 tuberosa (Homalophiura inornata). 329 tumida (Ophiuroglypha), 317 turgida (Ophiurolepis), 326 Valenciennesi (Ophiacantha), 255 victoriae (Ophionotus), 311 vivipara (Ophiacantha), 246 vivipara (Ophiomyxa), 241 Wallini (Ophiurolepis), 324 PLATE I Fig. I. Austrocidaris canaliculata (A. Agassiz). Adult specimen, carry- ing young ones on the apical system, among the spines. Figs. 2-12. Ctenocidaris speciosa, Mortensen. 2. Specimen infested with Echinophyces ; the spines carrying a number of the bivalve Limopsis sp. Half side view. 3 . Adult normal specimen, with smooth, spade-shaped oral primaries. Oral side. 4. Young normal specimen, side view. Some Limopsis on the spines. 5 . Adult normal specimen, with the spines overgrown by Alcyonidium. Oral side. 6. Denuded test of normal specimen, side view. 7. Same of a specimen infested with Echinophyces. 8. 9. Specimens infested with Echinophyces, side view (8) and aboral side (9). 10-12. Denuded tests of specimens infested with Echinophyces, showing abnormal position of genital pores, at the edge of the peri- stome in the female (10, 12), in the middle of the interambulacrum in the male (11). Figs. 13-15. Eucidaris tribuloides (Lamarck). 13. Denuded test, oral side, of a young specimen from the West Indies. 14. Specimen from Ascension, aboral side. 15. Specimen from Ascension, half denuded. Oral side. All figures natural size. DISCOVERY REPORTS, VOL. Xll PLATE I in Bale Sons A DanitliSOjiL''' Ltmdon PLATE II Figs. 1-4. Sterechinus Neumayeri (Meissner). 1-3. Specimensoflong-spined form, oral side (i,3)andaboralside(2). 4. Large specimen, oral side. Figs. 5-10. Notechinus marionis, n.sp. 5, 6. Adult specimens, oral side (5) and aboral side (6). 7-10. Denuded tests, aboral side (7), oral side (8); side view (9, 10). Figs. H-16. Sterechinus Agassizii, Mortensen. II. Young specimen, oral side. 12, 13. Weil preserved specimens, showing the dense coat of very slender spines. Aboral side (12) and oral side (13). 14-16. Denuded test of very large specimen ; side view (14), aboral side (15) and oral side (16). All figures natural size. N.B. Figure 5 has become blurred in the reproduction, the spines looking as if they were serrate. In reaUty they are quite smooth, as in Fig. II. DISCOVERY REPORTS. VOL. XII PLATE II JohrBiJi' Sons &. Dan.p[!-io-. L"^ L.n.dcT PLATE III Figs. 1-4. Parapneiistes cordatus, Koehler. 1, 3. Male specimen, aboral side (i) and oral side (3). 2, 4. Female specimen, aboral side (2) and oral side (4). Figs. 5-8. Amphipneusies Lorioli, Koehler. 5. Female specimen, denuded; aboral side. 6. Large female specimen, denuded; side view. 7. 8. Male specimen, half denuded; aboral side (7) and side view (8). Fig. 9. Abatus curvidens, n.sp. Type specimen; female, aboral side. Fig. 10, Abatus cavernosus, var. bidens, Mortensen. Aboral side. Figs. 11,12. Abatus cavernosus (Philippi). Inside of test of male (11) and female specimen (12). All figures natural size. DISCOVERY REPORTS, VOL. XU PLATE 111 JohnEifeSuns 4.LaniebioTi.L"^ London PLATE IV Figs. 1-7. Amphipneustes similis, n.sp. I, 2. Female specimen, oral side (i) and aboral side (2). 3. Female specimen, aboral side (Station 170) ; slightly abnormal, the right anterior petal being undeveloped. 4. Male specimen, aboral side. 5. Female specimen (St. 180), aboral side. 6. Male specimen, side view. 7. Female specimen, side view. Fig. 8. Amphipneustes Lorioli, Koehler. Aboral side. Figs. 9, 10. Echinocardium connectens, Mortensen (?). Aboral side (q); side view (10). All figures natural size. DISCOVERY REPORTS, VOL XII PLATE IV inHjIt Slt;^ S-E'^ni-i PLATE V Figs. I, 2. Astrotoma Agassizii, Lyman. Adult specimens, aboral side. Natural size. DISCOVERY REPORTS, VOL XII PLATE V PLATE VI Figs. I, 2. Asirototna Agassizii, Lyman. Adult specimens, oral side. Natural size. DISCOVERY REPORTS. VOL. XII PLATE VI PLATE VII Fig. I. Ophiacantha densispina, n.sp. Type specimen. Aboral side. Fig. 2. Ophiacantha vivipara, Ljungman. Specimen carrying young ones. Figs. 3, 4. Ophiacantha vivipara, wax. pentactis,n. vox. Aboral side. Fig. 5. Ophiomitrella falklandica, n.sp. Aboral side. Fig. 6. Ophioscolex (Ophiolyctis) nutrix, n.sp. Aboral side. Fig. 7. Ophioceres incipiens, Koehler. Abnormal specimen; oral side. Fig. 8. Astrochlamys bruneus, Koehler. Specimens in copulation. Fig. 9. Astrochlamys sol, n.sp. Type specimen. Aboral side. Attached to a colony of Bryozoans. Fig. 10. Amphiura princeps, Koehler. Aboral side. Fig. II. Amphiodia ascia, n.sp. Aboral side. Figs. 12-15. Amphiopliis peregrinator, Koehler. Aboral side (12, 14, 15) and oral side (13). All figures natural size. DISCOVERY REPORTS, VOL Xll PLATE Vll PLATE VIII Fig. I. Ophiopyren regularis, Koehler. Aboral side. Photo from life. Fig. 2. Ophiocten amitinum, Lyman. Aboral side. Photo from life. Fig. 3. Ophiuroglypha Lymani (Ljungman). Aboral side. Photo from life. Figs. 4, 5. Homalophiura inornata (Lyman). Aboral side (4) and oral side (5). Figs. 6, 7. Ophiura Rouchi (Koehler). Aboral side (6) and oral side (7). Figs. 8-13. Ophiurolepis brevirima, n.sp., specimens of various sizes; oral side (8, 10), aboral side (9, 11-13). Fig. 14. Amphioplus acutus, n.sp. Aboral side. Figs. 1-3 enlarged; Figs. 4-14 natural size. DISCOVERY REPORTS, VOL. XII PLATE VIII Figs 1-3. EH Marshall phot F3gs.4-13. V. Huti.phot. pr^li PLATE IX Figs. 1-4. Notechinus marionis, n.sp. 1, Tridentate pedicellaria. x 100. 2, 3. Valves of globiferous pedicellariae, small (2) and large form (3). 2, X 200; 3, X 100. 4. Valve of triphyllous pedicellaria. x 200. Figs. 5-7. Plexechinus Nordenskjoldi, Mortensen. Valves of tridentate pedicellariae. x 100. Fig. 8. Ctenocidaris Geliberti (Koehler). Valve of large globiferous pedicellaria. x 72. Figs. 9-1 1. Abatus cavernosus, var. bidens, Mortensen. Valves of tri- dentate (9), globiferous (10), and rostrate pedicellariae (11). x 100. Figs. 12-16. Abatus elongatiis (Koehler). Valves of tridentate (12), globiferous (two valves in connection) (13), rostrate (14), tri- phyllous (15), and large tridentate pedicellariae (16). x 100. Figs. 17-20. Abatus curvidens, n.sp. Valves of tridentate (17), rostrate (18), and globiferous pedicellariae (19-20). x 100. Figs. 21-26. Amphipneustes stmilis, n.sp. Valves of triphyllous (21), globiferous (side view, 22 ; from the inside, 23), tridentate (24, 25), and rostrate pedicellariae (26). x 100. Fig. 27. Amphipneustes Lorioli, Koehler. Valve of globiferous pedi- cellaria. x 100. DISCOVERY REPORTS, VOL . XII PLATE K Tl, M^r'.ensen del -iDhniiaJtSor.i A i.onn^iofi.L^'l London [Discovery Reports. Vol. XII, pp. 349-376, Plates X-XII, February, 1936.] THE BIRDS OF THE SOUTH ORKNEY ISLANDS By R. A. B. ARDLEY, R.N.R. CONTENTS Introduction page 351 Systematic account 35^ J^feno^/yfes/orrfm, G. R. Gray, Emperor Penguin 352 Pygoscelis adeliae (Hombr. et Jacq.), Adelie Penguin 352 Pygoscelis antarctica (Forster), Ringed Penguin 355 Py^oice/w/Jflpwa (Forster), Gentoo Penguin 357 Eudyptes chrysolophus (Brandt), Macaroni Penguin 358 Mac>-0Hec^M^?^a«?ett5 (Gmelin), Giant Petrel 358 Daplion capensis (Linn.), Cape Pigeon 361 Pagodroma iiivea (Forster), Snowy Petrel 363 Priocella antarctica (Stephens), Silver-grey Petrel 364 Thalassoica antarctica (Gmelin), Antarctic Petrel 365 Pachyptila desolata banksi, Smith, Dove Prion 366 Halobaena caerulea (Gmelin), Blue Petrel 367 Oceatiites oceanicus (Kuhl), Wilson's Storm-Petrel 368 Fregetta tropica melanogaster (Gould), Eastern Black-bellied Storm-Petrel 369 Catharacta skua lonnbergi, Matthews, Brown Skua 370 Catharacta skua maccormicki (Saunders), McCormick's Skua .... 371 Larus dominie anus, Licht, Southern Black-backed Gull 372 Sterna hirundinacea, Lesson, South American Tern 373 Sterna vittata georgiae, Reichenow, Wreathed Tern 373 Phalacrocorax airiceps, King, Blue-eyed Shag 374 Cliionis alba (Gmelin), Sheathbill 375 List of Literature 376 Plates X-XII following page 376 T THE BIRDS OF THE SOUTH ORKNEY ISLANDS By R.A. B. Ardley, R.N.R. (Plates X-XI I ; text-fig. i) INTRODUCTION H E following notes on the birds of the South Orkneys comprise a complete list of all the birds which visit the group for breeding purposes, or which occur as occasional stragglers. A few species of petrel, notably several of the albatrosses, in their ocean wanderings approach quite near to the islands when they are clear of pack-ice, but it is doubtful if they ever come within about ten miles of them. These wanderers of the adjacent seas have not therefore been included. The notes are compiled from observations made during the month of January 1933, when the R.R.S. ' Discovery II ' was engaged in a hydrographic survey of the islands, ^ and during a few days in February 1931, when the ship revisited Coronation Island. There are no truly resident birds in the South Orkneys, for the breeding species all leave the islands in the winter and either become pelagic or move to less rigorous regions. However, occasional individuals of most of the species visit the islands from time to time during the winter. All of the birds are oceanic, with the possible exception of the Sheathbill, and none of them are peculiar to the South Orkneys. In the summer, the bird life is extremely rich, and sixteen species were found to breed in the islands, with two others regarded as possible breeders. These are Eudyptes chrysolophiis and Halobaena caeridea. An excellent account of the birds of the South Orkneys observed during the Scottish National Antarctic Expedition is given by Mr Eagle Clarke in the Ibis for January 1906. The observations of this expedition, however, were almost entirely confined to Laurie Island and the neighbouring islets; the present notes include the whole group. It has not been found necessary in most cases to enlarge on the descriptions of the plumage of such species as are included in the Scotia report, or are already well known. A matter for regret is that hardly any material was brought back from the South Orkneys. This was due partly to the writer's inexperience in skinning and partly to pressure of work, for in only twenty-eight days the entire group was surveyed and the ship's company was kept fully occupied. Eggs of most of the breeding birds were obtained. The breeding-places of sixteen species are shown in Fig. i. My thanks are due to Mr N. B. Kinnear for his assistance and interest, to Mr A. G. Bennett, and to Mr Bruhns of the Argentine Meteorological Station in Scotia Bay for information regarding the bird life there. 1 See J. W. S. Marr, The South Orkney Islands, Discovery Reports x, pp. 283-382, pis. XII-XXV. 352 DISCOVERY REPORTS SYSTEMATIC ACCOUNT Aptenodytes forsteri, G. R. Gray, Emperor Penguin. The Emperor Penguin occurs at the South Orkneys only as an occasional winter straggler. According to Mr Bruhns, the official in charge of the Argentine Meteoro- logical Station in Scotia Bay, one or two birds are seen during the course of nearly every winter. From his description these are mainly aduhs. They only occur when the islands are closely surrounded with pack-ice and are evidently non-breeding birds. In the summer, the bird has apparently not recently been seen at the station, though the islands are often beset with ice for periods during the summer. The Scotia Expedition have a record of a bird, probably belonging to this species, which was seen in Scotia Bay in November 1903, and two were seen in March 1905 (Eagle Clarke, 1906). Emperor Penguins are comparatively common in the ice of the Weddell Sea, and it is probable that there is a fairly extensive breeding ground somewhere within its area. The birds which occur at the South Orkneys are certainly from the Weddell Sea, and travel up on the great ice drift from its western side. In January 1932, several adult birds were seen in pack-ice about 250 miles east-south-east from the Orkneys, but in 1933, when the ' Discovery II ' visited the group, there was no pack-ice in the vicinity. Since the Emperor Penguin is essentially a bird of the pack-ice, it would never be seen in the South Orkneys when open sea surrounds them. Pygoscelis adeliae (Hombr. et Jacq.), Adelie Penguin. (Plate XI, fig. i.) Adelie Penguins are extremely numerous in the South Orkneys, and a rough census taken of all the rookeries in the group resulted in a total of about 1,500,000 nests. This gives a population of about three million. The Scotia Expedition estimated the number at five million (Eagle Clarke, 1906), but it is difficult to estimate the number with any degree of accuracy. There was no evidence either of increase or decrease in the number of nests in the rookeries, and the inference is that the penguin population maintains a fairly even level. At a rough estimate, perhaps half the chicks hatched (about 1,500,000 birds), take the water for the first time each year. Apparently little is known of the normal span of life of penguins, but the mortality rate, even among adults, must be high. Probably in the winter months, as well as in the breeding season, many fall victims to sea-leopards, killer whales, and natural catastrophes. It is doubtful if many perish from starvation, for there is always food in plenty, and the only circumstance which would be likely to involve risk of starvation would be for parties of birds to be stranded on unbroken fields of consolidated pack. This, in the Weddell Sea area where the ice is always moving, would be of rare occurrence. Laurie Island is the chief breeding-place for the birds, and huge rookeries are established on every suitable site. On Weddell and Saddle Islands the majority of birds are P. antarctica, for here the sites are generally steeper. On Powell Island there are large rookeries, but on Coronation and Signy Islands there are only five comparatively small rookeries in all, none of which are to the west of Signy Island. Thus in the South BIRDS OF THE SOUTH ORKNEY ISLANDS 353 Orkneys we see the same curious distribution which was observed in the South Shet- lands, P. adeliae preponderating in the eastern half of the group, and P. antarctica in the western half. The birds come ashore in early October, and egg-laying begins in the last few days 6d INACCESSIBLE ISLANDS. Fig. I. Chart of the South Orkney Islands showing the positions of breeding-places, which are indicated by letters under the place names. Small letters indicate that breeding birds are present, and capitals that large numbers of birds (5000 or more) nest in the localities. A, a. Pygoscelis adeliae. B, b. Pygoscelis antarctica. C,c. Pygoscelis papua. D, d. Macronectes giganteus. E, e. Daption capensis. /. Pagodroma nivea. G. Priocella antarctica. H. Pachyptila desolata banksi. , i. Oceanites oceanictis. k. Fregetta tropica melanogaster . I. Catharacta skua loniibergi. m. Lams dominicanus . n. Sterna hiriindinacea. o. Sterna vittata georgiae. p. Phalacrocorax atriceps. r. Chionis alba. of the month. The staff at the Argentine Meteorological Station stated that every year the date of the first eggs is nearly the same, and that there are always hundreds of eggs to be gathered in the first three days of November. The young are hatched during the first few days of December ; but we did not arrive at the islands until January i, so that the early life of the chicks was not observed. From January i to 2 1 , various rookeries were almost continuously under observation, and the habits of the birds closely conformed with the excellent account given by Dr E. A. Wilson in the report of the National Antarctic Expedition (1901-3). A point which continually struck the observer, and is remarked in nearly all previous reports, 354 DISCOVERY REPORTS was the progressive state of filth, confusion, and smell in the rookeries. In the early days of January the colonies were still comparatively orderly, individual nests were well marked, and the chicks, still fairly small, seemed to be more under the control of the parent birds, who were well able to protect them from the depredations of the many skuas which frequent the rookeries. The young birds grew at an astonishing rate, and by January 5, the "creche" system, whereby parties of from ten to twenty chicks are herded together under the supervision of a few old birds, was beginning to become necessary. The rookeries henceforward lost all semblance of order, becoming pro- gressively filthier and more untidy. The nests were scattered and trampled out, and as the chicks grew larger and individuals wandered from the flocks, the death-rate mounted. Every day more dismembered remains and flattened, trampled corpses of youngsters which had fallen in the race were to be seen. The skuas had growing and ravenous young of their own to feed, and were increasingly vigilant and bold in their attentions. Giant petrels and sheathbills were always in evidence about the rookeries. These birds were never seen to attack a strong and healthy chick ; always they waited until an ailing youngster was almost at the point of death. It is safe to say that no weakly penguin chick has the smallest chance of surviving in an Adelie rookery. If it were lucky enough to escape its natural enemies, it would soon starve from being unable to pester and attract the attention of an old bird and induce it to provide food, or would be trampled beneath the heedless feet of its stronger contemporaries. By January 13 the chicks in the large rookery surrounding EUefsen Harbour in Powell Island were beginning to lose their nestling down. These were reckoned to be from 5^ to 6| weeks old. The chicks at this time assisted one another to pluck the down, with the occasional attention of an old bird, which was given in an absent- minded manner. After January 21 the birds were not watched ashore in the rookeries, but by that date most of the young birds were in their normal first-year plumage, though none were seen to enter the water. In all the nests examined during the early days of January, two chicks were found. No eggs were seen, and addled eggs appear to be a rarity. In this respect P. adeliae differs from some other penguins. Many bones were seen in the nests and the nesting birds probably collect the skeletons of the previous year's dead to add to their usual pile of stones. Adelie Penguins in their choice of nesting sites undoubtedly prefer low, rocky shores, and they were never found on steep rocky slopes and scarps in such positions as might be occupied by Ringed Penguins. With a little practice it is easy to distinguish from seaward the species inhabiting a rookery, even when individual birds cannot be identi- fied. Adelie rookeries have a brick-red appearance, while rookeries of Ringed Penguins are paler and yellowish in colour. This colouring is due to the profusion of excrement about the rookeries, and the difference must result from some small divergence in the general diet of the two species. Another good indication in identification is the character of the site, for a rookery established in a steep and difficult situation is almost certain to contain Ringed Penguins. BIRDS OF THE SOUTH ORKNEY ISLANDS 355 The two species are never found to intermingle in the rookeries, though colonies may be established adjoining one another. The Gentoo Penguin, however, establishes his small settlements among rookeries of P. adeliae, and the two species live in close com- pany on amicable terms. The Adelie Penguins leave the South Orkneys in the early part of April, and probably spend the winter in a pelagic state, moving about among the ice until the following spring, when they again seek their breeding haunts. During the winter a few birds are nearly always present in Scotia Bay, and no doubt parties of birds come ashore to rest, at times, on every landing place in the group. While the ' Discovery II ' was engaged in survey work round the islands several fine opportunities for observing the birds swimming occurred. As the ship cruised slowly along the shores, companies of swimming penguins would often remain close alongside for considerable periods. From the flying bridge, in calm weather when the sea was smooth, their method of swimming was well observed. As stated by Dr E. A. Wilson the flippers alone are used for propellent purposes. The flippers are brought forward to a position almost at right angles to the body, edged horizontally. For the backward stroke, the anterior edges of the flippers are depressed, so that the flippers make an angle of about 30° with the horizontal. They are then swept backward and slightly downward, until they lie at an angle of about 45° with the line of the body with their tips slightly depressed. A momentary pause in the stroke is noticeable here, then the flippers are edged horizontally and swept to the front again for the next stroke. When the birds swim rapidly the pause is so brief as to be hardly noticeable, and the flippers move so fast that it is impossible to follow them, but the same process is evidently carried out. This method of swimming appears to give the birds a gently undulating motion in their progress under water. When they break surface for breathing, or "porpoise", the last down-stroke before surfacing is apparently a strong one, and the head and tail appear to be both raised until the bird comes to the surface, when they are both immediately depressed. Sharp turns are carried out, apparently, with a combined movement of flippers, tail, and feet, the latter acting as a rudder, and for ordinary small alterations of course the tail alone is sufficient for steering. When swimming, penguins carry their feet held out in line with the body, the webs partially folded in so that the claws are nearly closed up on each other. Adelie Penguins often swim on the surface, more frequently when near their rookeries or during rests from feeding. When alarmed they can attain a good speed in the water, and can certainly move at from 10 to 12 m.p.h., at any rate for short distances. Dr Wilson has written such an excellent account of these engaging birds that little more remains to be said concerning their habits. Pygoscelis antarctica (Forster), Ringed Penguin. (Plate X.) In the South Orkneys, Ringed Penguins breed in large numbers on all the islands, and this group ranks with the South Shetlands and South Sandwich Islands as one of the three main breeding grounds of the species. A rough census taken of the group in 356 DISCOVERY REPORTS January 1933 gave a total of about 1,500,000 nests, the birds being about equal in point of numbers with P. adeliae. Although P. adeliae are in the vast majority on Laurie Island, there are several large rookeries of P. antarctica, notably on Cape Robertson and on Ailsa Craig. Saddle, Weddell, and Bruce Islands give nesting sites to large numbers of the birds, and there are several small rookeries on Powell and Fredriksen Islands and the adjacent rocks and islets. Coronation Island, however, is the main stronghold of the species. All along the coasts, wherever rocky outcrops break through the usual ice-cliffs, small colonies of the birds are established, and in Sandefjord Bay and the south-west corner of Coronation Island there are huge rookeries containing at least 200,000 nests. Another large rookery is established on the Robertson Islands, and there are straggling colonies on the Inaccessible Islands. The Ringed Penguin is often found to choose nesting sites in very much steeper and more difficult places than is usual with any other penguin. Sometimes the rookeries are even difficult of access to man, and the toil to which the birds are put in landing on the steep-to shores and climbing to their nests is often enormous (Plate X, fig. 2). On Fredriksen Island, and in some parts of Sandefjord Bay, birds in steep places were sometimes seen to lose their footing and tumble headlong for considerable dis- tances down the rocks. This experience appeared to harm them not at all. From sea- ward, some of the small colonies round the coasts appeared to be placed in even steeper and more precarious sites than those which were visited, this being so in particular on the Inaccessible Islands. The birds breed about three weeks later than P. adeliae, the eggs being laid as a rule in the last few days of November. On January 4, 1933, the nests in the rookeries on Powell and Fredriksen Islands contained eggs and newly hatched chicks in almost equal numbers, none of the chicks being more than about four days old. Each nest had two eggs or young, and few deserted eggs were seen. On January 9 the birds in the Sande- fjord Bay rookeries had reached the same stage, and by January 1 1 almost all the eggs had hatched in this locality (Plate X, fig. i). This retardation in breeding of the birds in the western part of the group may be explained by the circumstance that the Sande- fjord Bay rookeries are mainly more exposed and are thus more likely to be snowed over than the Fredriksen and Powell Island rookeries. On February 15, 193 1, the rookeries at Sandefjord Bay were visited, and at this time the young were almost full grown but had not begun to shed their nestling down, though the first plumage was well developed beneath it. Thus they were nearly a month behind the Adelie Penguins in reaching this stage. A few of the fledglings were killed, and their crops contained remains of Euphausians and Amphipods. The hatching and fledging dates in the South Orkneys seem to be about the same for the South Shetlands. In the rookeries, the Ringed Penguin is very much less tolerant of the presence of sheathbills than is the Adelie Penguin. The latter seems to ignore the sheathbill unless the bird actually disturbs its nest, but it was noticed that Ringed Penguins always chased sheathbills away from their chicks. The two penguins are similar in point of pugnacity, BIRDS OF THE SOUTH ORKNEY ISLANDS 357 but the Ringed Penguin is a more persistent and courageous bird, and will, when enraged, return to the attack of an intruder after being repulsed several times, with complete disregard for danger. These two species are the most engaging of all penguins, and to watch their antics when quarrelling, love-making, attending to their young and travelling between their rookeries and the sea, is a source of the greatest diversion. A practice which is shared exclusively by these two species is that of hauling out on icebergs and detached fragments of ice at sea. Here again the Ringed Penguin will choose most difficult and steep bergs on which to rest, while the Adelie is only seen on low, fairly level- surfaced bergs or flat pieces of sea-ice. Almost anywhere within the range of the Ringed Penguin all accessible bergs are occupied by some of the birds, or bear indications of their visits. Pygoscelis papua (Forster), Gentoo Penguin. (Plate XII, fig. 3.) This bird breeds in the South Orkneys, but in very much smaller numbers than either P. adeliae or P. antarctica. The Scotia Expedition estimated the population of Laurie Island to be 100,000 birds (Eagle Clarke, 1906). In the large Adelie rookery in EUefsen Harbour, small groups of P. papua were found nesting, the total number of nests being about a thousand. Nowhere else in the South Orkneys were rookeries found, and since they apparently only nest in company with P. adeliae in this locality it is very improbable that any rookeries are established on Coronation Island. The only birds seen round Coronation Island were in Sandefjord Bay, where on January 10 about thirty birds were seen ashore near a branch of the large Ringed Penguin rookery. The nests in the EUefsen Harbour rookeries were under observation for several days in the early part of January. These nests showed good discrimination on the part of the building birds in selection of their sites, for nearly all of them were placed in some kind of shelter, under the lee of large rocks or in clefts in the rocky ground. The nests of the Adelies, in comparison, appeared to have been established in quite a haphazard fashion, without regard to any convenience or natural advantage. Usually from three to about twenty nests were associated in each group of P. papua, and the nearest nests of the Adelies were always several feet distant. The contrast in disposition between the two species is remarkable when the birds are observed simul- taneously. The Gentoo is a comparatively quiet and timid bird and when disturbed usually leaves its nest after a few half-hearted attempts at biting. Compared with the Adelie's liveliness and pugnacity, it is a lethargic and stolid bird. The Gentoo also goes to more trouble in the building of its nest, keeps it cleaner and more orderly, and appears to be more solicitous of its chicks. Each pair of parent birds keeps the young in the nest until they are almost ready to fend for themselves. On January 4, 1933, the chicks in the EUefsen Harbour colonies were either newly hatched or only a few days old, and several unhatched eggs were found. Each nest con- tained two eggs or chicks. In the rookeries in the Palmer Archipelago, farther south, the chicks are not hatched until the middle of January, while in South Georgia they are hatched in early December. 358 DISCOVERY REPORTS Thus it is evident that the more rigorous the chmate the later the hatching date, and this rule applies to nearly all Antarctic birds which have a breeding range embracing considerable differences of latitude. In April the Gentoos, like the other penguins, leave the South Orkneys and become pelagic, but occasional stragglers are seen from time to time during the v^^inter. The bird is not a creature of the pack-ice, and is very rarely seen on floes. Probably when approaching the islands the birds wait until the ice is either temporarily clear, or suf- ficiently open for them to make the approach by water. Eudyptes chrysolophus (Brandt), Macaroni Penguin. The Scotia Expedition found five single specimens of E. chrysolophus among penguin rookeries on Laurie Island, and suspected that the bird probably bred in the South Orkneys (Eagle Clarke, 1906). Bennett (1926, p. 312) states that it breeds in all the Dependencies of the Falkland Islands, but there appears to be no record of any rookeries having been found in the Orkneys. During the visit of the ' Discovery II ' a sharp look-out was kept for the bird all round the group, but no sign of a rookery was found. One adult male was captured in the Ringed Penguin rookery at the south end of Fredriksen Island ; this was the only bird seen. Although in view of the diligent search I think it improbable that this species breeds in the Orkneys, it is possible that a small rookery might be established on one of the off-lying islets between the two largest islands, where it might easily be overlooked. As an instance, there is a small rookery of macaronis established right in the middle of the great Pygoscelis antarctica rookery on Deception Island, which has often escaped the observation of people who have visited the Ringed Penguin rookery. In the Scotia Report, Eagle Clarke points to the probability of a breeding ground in the South Orkneys, since the birds captured were young and could hardly have accomplished the rough sea passage of 600 miles from South Georgia. Since this report was published, however, it has been discovered that there are several rookeries of Eudyptes chrysolophus in the South Shetlands. This fact reduces the sea passage to 200 miles, with wind and sea conditions favourable, and it is no longer necessary to assume the existence of a South Orkney rookery. It is unlikely that the bird visits the group during the winter, for it dislikes ice and is never seen when pack-ice is in the vicinity. Macronectes giganteus (Gmelin), Giant Petrel. (Plate XII, figs, i, 2.) The Giant Petrel breeds in considerable numbers on the South Orkneys. On the northern coasts of Laurie Island there are several colonies, and the Scotia Expedition estimated the number of birds at about 5000. On the islets surrounding EUefsen Harbour there is a large colony containing about 600 nests, but the main stronghold of the species is on Signy Island, where their nests are established all along the western coast. A large colony is also present on the slopes above Borge Bay. The nests are all built on low foothills and gently rising slopes near the sea. BIRDS OF THE SOUTH ORKNEY ISLANDS 359 An excellent account of the habits of this species is given by L. H. Matthews (1929), and nothing further remains to be said in that respect. In the South Orkneys, laying takes place about the middle of November, and on January 4 hatching had just commenced in the colony in EUefsen Harbour. This gives an incubation period of about seven weeks. On this date about one in four of the nests contained newly hatched chicks and the majority of the remaining eggs were cracking. By January 14 all the eggs were hatched. The colony at Borge Bay is very straggling, but in their choice of nesting sites the birds are gregarious, and usually from six to twenty nests are associated in groups. On January 17 all the nests here contained chicks. A single parent bird only was usually present at each nest, but sometimes the mate would appear, usually with food for the chick. As usual, when their nests were approached the birds vomited the contents of their stomachs. In the Borge Bay district the contents were found to consist mainly of Euphausians, while in the EUefsen Harbour colony, which adjoins the Adelie rookery, remains of young penguins, and in one case half a large fish, were among the Eu- phausians. The plankton probably came from the stomachs of young penguins. No whalers were working near the South Orkneys in January 1933, and it is probable that the birds found more difficulty than usual in obtaining food. Certainly they kept a very sharp eye on the penguin rookeries, and were ready to attack any dead or ailing chick, advancing upon such an unfortunate at a clumsy waddle, wings half spread and neck outstretched. With regard to colour phases, it was found that at EUefsen Harbour the average proportion of white birds was 10 per cent, and at Borge Bay 9-5 per cent. The average of birds seen about the remainder of the islands was also 10 per cent. This agrees well with the observations of Bennett in the South Shetlands a few years ago, where he found an average of i2| per cent of white birds. The Scotia Expedition, however, found an average of only 2 per cent of white birds in the South Orkneys (Eagle Clarke, 1906). It would be interesting to know if only absolutely white birds were included in this count, for in a number of the birds included in the Discovery percentage, a few feathers were very faintly flecked with grey — a common feature in "white" birds. Eagle Clarke also mentions a number of very dark brown birds seen in the Orkneys. On the present visit, no dark-phase birds at all were seen anywhere in the group. All were in either the white or intermediate phase (Plate XII, fig. 2), and the observations suggest a pro- gressive lightening of the plumage of the South Orkney birds in the last thirty years. L. H. Matthews, in his account of the birds of South Georgia (1929), recognizes three main phases in the plumage of the Giant Petrel — dark, light, and intermediate. Bennett refers to four — " all-browns ", " grey-necks ", " white-necks ", and whites. These phases are referred to by Lowe and Kinnear (1930, pp. 151-4). In my opinion, a better nomenclature would be: (i) dark — uniformly dark brown or brown, (ii) dark inter- mediate— plumage mainly grey-brown, flecked here and there with dirty whitish, grey neck, (iii) light intermediate, generally lighter grey and with more dirty whitish about the plumage than (ii), and with white head and neck, and (iv) white. The gradations of 36o DISCOVERY REPORTS plumage between the dark form and the white are too varied to be classed all together as intermediate as Matthews has done. The true plumage in the dark phase is a rich, very dark brown, which appears almost black in some lights. Lowe and Kinnear are of the opinion that "brown" birds are merely faded specimens of the true dark phase and with this I am in entire agreement. A large series of observations was made during the first two voyages of the ' Discovery II ', and it was repeatedly noted that very dark birds were always in excellent plumage. "All-brown" birds were always drab brown or greyish brown, with a dingy, faded ap- pearance, and in no case was such a bird in good trim. The plumage was invariably to some degree ragged and often there were feathers missing. Probably dark phase birds only remain in good condition for a few months after the moult, after which fading and weathering take place. The same fading occurs in several brown Southern Ocean birds, notably in the case of Thalassoica antarctica. After the moult the brown plumage of these birds is a rich chocolate colour, but later it acquires a drab and faded appearance. A correlation of colour phases with latitude over the South Atlantic Ocean, at any rate in the oceanic range of the species, is clearly indicated by the observations of the ' Discovery II '. North of about latitude 52° S no white birds were seen, and the great majority were dark. Between this latitude and the Antarctic circle, the percentage of white birds increases progressively with at first dark and then light intermediate forms predominating. In summer it would not be exaggerating to say that every degree of latitude southward from the northerly limit of range of the species shows collectively a progressive lightening in the plumage of the birds. In winter there is likely to be little material alteration, for the latitude range of the birds is merely moved bodily northward some five degrees. Observations in winter over the South Indian and South Pacific Oceans did not give such definite results, although more light-plumaged birds were seen in the southerly part of the range. In the Falklands, dark and intermediate phases are equally numerous, the majority of the latter being dark intermediates. Of the South Orkney birds 10 per cent were white, about 75 per cent were light inter- mediates, and the remaining 15 per cent dark intermediates. A considerable number of the light intermediates were very light, being mainly white with light grey-brown specklings. In view of Eagle Clarke's account, it seems probable that in the South Orkneys a white race is evolving. Mr R. A. Falla is engaged on an account of the birds observed on Sir Douglas Mawson's 1929-30 Expedition, and it will be interesting to note his conclusions with regard to the Giant Petrels of Kerguelen and Heard Island. An explanation of the increase in white and light intermediate birds in the far south is suggested by the development of southern whaling during the last twenty-five years. Where whaling is in progress food is abundant and easily obtained, and it is probable that this has brought about an increase in the Giant Petrel population of high southern latitudes. Since colour is correlated with latitude the increase will be mainly among white and intermediate birds, with the result that higher percentages of these forms are BIRDS OF THE SOUTH ORKNEY ISLANDS 361 now to be seen at the South Orkneys. It seems certain that Giant Petrels are com- moner in high south latitudes than they were before whaling commenced. From observations of the chicks in the South Orkney colonies it is evident that the colour phases are inherited. The percentage of chicks in pure white down was about the same as or a little greater than the percentage of white adults. White chicks were almost always seen in the nests of white adults, and in the remaining nests both parent birds were not seen. In only a few cases were both parent birds seen together; those of one pair were both white, the remainder being either both intermediate or intermediate and white. The chicks of intermediate birds were in very light grey or dirty whitish down (Plate XII, fig. i). Each occupied nest in all the colonies visited contained one chick, but there were a number of unoccupied nests, particularly in the Borge Bay district. A possible reason for this was the absence of whalers in Borge Bay, so that fewer birds than usual were attracted to the locality. Addled eggs in all the rookeries amounted to only 2 per cent, which is very low in comparison with other locally breeding petrels and in view of the rigorous climate and the exposed situation of the nests. The iris of the Giant Petrel is generally pale greenish buff, but in the South Orkneys about one in twenty of the birds had pale china-blue irises. This colour has not been seen in South Georgia, nor, apparently, in the Falklands. All the birds with blue irises were found to be light intermediates, but since these comprised 75 per cent of the total, it does not follow that this correlation is exact. Daption capensis (Linn.), Cape Pigeon. The South Orkneys probably form the most populous breeding ground in the South Atlantic for this species. The South Shetlands and South Sandwich Islands give nesting places for many thousands, and a considerable number breed in South Georgia, but there are enormous numbers of nesting birds in the Orkneys during the season. The Scotia Expedition estimated the population of Laurie Island alone at 20,000, and at all the other islands in the group the birds nest in much larger numbers than on Laurie Island. On Weddell, Saddle and Fredriksen Islands many thousands congre- gate, and many parts of the steep coasts of Coronation and Powell Islands are occupied by nesting birds. Signy Island is also well populated. An interesting fact, which holds also with the other open-cliff breeding petrels of the islands {Priocella antarctica and Pagadroma nivea), is that the birds never choose an unprotected southward-facing cliff on which to nest, no matter how favourable the site may appear. This was particularly noticeable on such islands as Saddle and Weddell, where the northern cliffs were dotted with sitting birds, while the south-facing cliffs, although clear of snow, were entirely uninhabited. The rule was found to be invariable, and where colonies were established among groups of rocky cliffs facing in several directions, if nests were present on any face which had a southerly component in its direction, it was always found that some protection, in the shape of a projecting spur of rock, or another cliff opposite, sheltered the nests from southerly winds. 362 DISCOVERY REPORTS In consequence, there are comparatively very few^ clifF-breeding petrels nesting along the south coasts of the islands, and then only where a turn in the coast, or an off-lying islet, affords a sheltered situation. The reason for this circumstance must be that southerly winds are the chief snow- bearing winds in the breeding season, and are more bitter than winds from the northern semicircle. The Scotia Expedition found that laying takes place in the first two weeks of December. Nests in the huge colony on the western side of Fredriksen Island were visited on January 4, and were found to contain hard-set eggs. About one nest in fifteen contained a newly hatched chick. The nests here were small hollows in the rocky ledges of the cliff, lined with a collection of small flat plate-like pebbles from a quarter of an inch to an inch in diameter. On the majority of nests only the brooding bird was present, but at a few both birds were seen. Some of the nests here were only about ten feet above the sea. The birds are usually gregarious, and were nowhere found nesting singly. The colonies are scattered and individual nests may be placed some distance from the rest. On January 9 the nests in the Sandefjord Harbour district contained hard-set eggs and chicks a day or two old in about equal numbers. These very young chicks, almost as soon as hatched, were found to inherit their parents' habit of ejecting the contents of their stomachs at the intruder. In this they showed even greater proficiency than the adults, attaining a longer range and more precision of aim. The chicks are clad in dark smoky grey-brown down, which appears in some lights to have a purplish tinge. There are two lighter circles, almost clear of down, round the eyes, which give the chicks a peculiar spectacled appearance. In the Sandefjord Harbour district, the nests of the Cape Pigeons are established on all the cliffs occupied by Silver-grey Petrels, and the nests of both birds are mixed on the rocky ledges. The Cape Pigeon nesting grounds extend to a lower altitude than those of their neighbours. Some of their nests are only about eight feet above sea-level, and the lower parts of cliffs which are occupied by the two species are in general held more numerously by Cape Pigeons. They have no objection to height, however, and are found at any altitude on the seaward-facing cliffs. The Inaccessible Islands are the main stronghold of Priocella antarctico, and it might be expected that Cape Pigeons would nest there also. None, however, were seen, and since Cape Pigeons are almost always plentiful in the vicinity of their nesting places it is improbable that they breed there. Snowy Petrels are also in the habit of nesting in company with Cape Pigeons and in Borge Bay nests of both birds were found in close proximity. On January 19 the nests in the Cape Pigeon colonies near Borge Bay all contained chicks, ranging from about a fortnight old to newly hatched youngsters. When the birds were sitting on eggs they usually remained on the nest until lifted off, making no movement beyond ejecting their stomach contents. With chicks, however, it was noticed that when approached they assumed a defensive attitude, crouching with wings held BIRDS OF THE SOUTH ORKNEY ISLANDS 363 out from the sides, and facing round to the intruder. In either case they kept up a harsh scolding chatter. The regurgitations were always found to consist of an oily mess almost entirely composed of Euphausians. In the Borge Bay colonies, where the nests were clustered fairly thickly, it was found that about one in four of the nests was unoccupied. These were apparently used as resting grounds by the mates of the sitting birds in the nests near by. The vacant nests seem to indicate that fewer birds than usual were breeding in this locality, owing probably, as with the local Giant Petrels, to the absence of whalers. The proportion of rotten eggs was found to vary in the different colonies, but it averaged about 15 per cent. This is almost certainly due rather to the rigours of the climate than to infertility. The Cape Pigeon is one of the few species which rarely visit the South Orkneys during the winter. In April the birds leave the group and become pelagic. In the South Atlantic their oceanic range in the winter lies some degrees farther north than their summer range, though probably a few birds would be seen almost down to the ice-edge in the winter. They rarely venture over the ice, and it is to this that their absence during the winter at the South Orkneys is in part due. Pagodroma nivea (Forster), Snowy Petrel. The Snowy Petrel breeds in considerable numbers on all the islands in the South Orkneys. It is also one of the commonest winter visitors, for in its oceanic range the bird prefers the vicinity of pack-ice. The nests were usually found in the same localities as those of Daption capensis, and in similar situations, except that the Snowy Petrels often used more sheltered sites on the cliffs. Most of their nests were to some degree protected, either by an overhanging portion of rock, or by being placed in a cranny or corner on the rocky ledges. In several places it was found that favourable nesting sites were provided where the rock face was fissured with diagonal cleavage planes and sheltered ledges were formed. The nest consists of a few small flat stones and scraps of rock debris, and is not usually so neat as the nest of D. capensis. The birds, in nesting, were found to be less sociable than the others of their order, and although in every favourable locality a number of nests were established, the individual nests were usually placed singly and well apart from each other. However, two or three nests were often to be seen on the same ledge. The single eggs are laid in the last few days of November and in early December. On January 4, all the nests examined in the vicinity of EUefsen Harbour contained chicks, ranging from a day or two to perhaps a week old. The chicks are clad in uniform pale pearl-grey down dorsally, and their under surfaces are dull white. Like the young of D. capensis they are proficient in ejecting their stomach contents. One parent bird was present on each of the nests visited here. When approached the birds raised the feathers on the crown, forming quite a noticeable crest, and kept up a harsh chatter similar to the defensive note of D. capensis. When drivenfrom their nests they some- times took wing and wheeled round in the vicinity, uttering a harsh call rather like the 364 DISCOVERY REPORTS sound of a watchman's rattle. This note was sometimes heard on other occasions when birds were flying close to the ship at sea, and it appears to be a scolding note. They were also heard to utter a quiet twittering, and on returning to their chicks made soft clucking sounds. The Snowy Petrel has thus a more extensive vocabulary than most of its relations. On January 8 several nests in Sandefjord Harbour were visited. Of these all but one contained chicks, and in the exception the chick was just hatching. In all, about fifty nests were examined throughout the South Orkneys, and all the Snowy Petrels seen apparently belonged to the one species Pagodroma nivea. Several sitting birds were measured, and it was found that the measurements were fairly regular. Almost certainly no larger birds were seen. Probably P. confusa is confined to the Ross Sea area, if indeed it is a distinct species. Measurements of two nesting birds in Borge Bay were : mm. mm. Bill, length 24 23 „ breadth at base 7 7 Wing 268 265 Tail 107 109 Tarsus 32 30-5 The bill is black, the iris very dark brown, almost black. The plumage is entirely white, except for a few small filaments of feathers just before the eye, which are bluish black. Often when the birds are seen flying it appears certain that there is a patch of grey pigmentation about the under wing-coverts. All the captured birds were examined but no trace of it was found, and it seems that the grey appearance must be an illusion due to shadow. The legs and feet are dark bluish grey, darker on the toes and claws. Priocella antarctica (Stephens), Silver-grey Petrel. (Plate XI, fig. 2.) The Silver-grey Petrel has for some years been suspected of breeding in the South Orkneys, but nests had not apparently been found there before the visit of the 'Discovery II'. The bird breeds in large numbers round the south-western and north-western corners of Coronation Island, where it was estimated that 25,000 nests are established. The nesting grounds seem to be entirely confined to this end of the main group of islands, for although a sharp look-out was kept, nowhere else was there any sign of their breeding. The Inaccessible Islands are, however, the main stronghold of the species. The northern cliffs of all three islands were, on January 25, seen to be dotted with nesting birds from top to bottom, the total number of nests being estimated at not less than half a million. The birds were apparently the only petrels breeding on these islands. On Middle Island, a rookery of Ringed Penguins is established on a steep bluff, and nests of the Silver-grey Petrels were scattered among the penguins. The only previous occasions on which this species has been found nesting was by Nordenskjold's Swedish Expedition, who found the birds breeding on Cape Roque- BIRDS OF THE SOUTH ORKNEY ISLANDS 365 maurel in Graham Land, and by the Australasian Antarctic Expedition of 1911-14 in AdeHe and Queen Mary Land.^ The nests and young appear not to have been described. On January 9 and 10, about forty nests were examined in the Sandefjord Bay district. Of these twelve contained chicks a day or two old and in the remainder the eggs were on the point of being hatched. The date of laying must thus be about the same as that of Daption capensis. The nesting sites and nests are very similar to those of D. capensis, except that generally more sheer and inaccessible locations are chosen. The nest consists of a collection of small flat stones, and the single egg is white, very similar to that of D. capensis but slightly larger. A thin chalky coating was found to be present on all the eggs examined. The chicks are covered in uniform silvery grey down, lighter on the under surface, and they have the blue-grey and rose-pink bills of their parents from birth (Plate XI, fig. 2). At the nests visited only one parent bird was present in most cases, but at a few the pair were seen. On being approached the birds, after the usual manner of petrels, ejected their stomach contents, and uttered a harsh high-pitched chatter, from time to time punctuated with a shrill squeaking note. Measurements of several birds ranged between mm. mm. Bill 46 and 49 Tarsui 3 53 and 55 Wing 330 and 342 The feet, legs and toes are blue-grey, the webs pinkish purple, and the claws grey- black. The upper mandible is mainly blue-grey with patches of pinkish and a black patch near the nares. The tip is grey-black. The lower mandible is rose pink, with a dark grey tip. This species is very similar to the Cape Pigeon in its general habits, and the two birds are very frequently seen in company at sea. Probably they leave their breeding haunts in the South Orkneys at about the same time, and there is no record of Priocella ant- arctica being seen at the islands in the winter. The bird breeds in large numbers on the South Sandwich Islands, particularly in the Southern Thule group, and in considerable numbers on Bouvet. At both these places again they share their breeding grounds with Daption capensis. There are indications that the Silver-grey Petrel also nests on some of the off-lying islands of the South Shetlands, but the South Orkneys may be regarded as the South Atlantic headquarters of the species. Thalassoica antarctica (Gmelin), Antarctic Petrel. Throughout the visit of the ' Discovery II ' not a single Antarctic Petrel was seen in the vicinity of the South Orkneys, and it is certain this bird does not nest in the group. 1 It is believed that nesting places were found in the Enderby quadrant by Sir Douglas Mawson's expedition of 1929-30. 366 DISCOVERY REPORTS Probably the species only occurs when the islands are surrounded by pack-ice, or when ice is in the vicinity, for in its oceanic range the bird was very rarely found to stray more than about fifty miles from the edge of the ice. In February 193 1 several birds were seen at the ice-edge about fifty miles north of Laurie Island, and in December of that year, one at the ice-edge sixty miles north of Coronation Island. In November 1932 birds were observed thirty miles south of Scotia Bay, the ice being then about sixty miles south of the land. The Scotia Expedition saw a number of birds near Saddle Island about the end of March, and doubtless ice was near the islands at that time. The Antarctic Petrel, then, is only a casual visitor. No breeding ground of the species is known in the Weddell area, but it is highly probable that one exists, for a fair number of birds are always to be seen at sea along the Weddell ice-edge. In the Bellingshausen Sea it is possible that the bird breeds on Peter 1st Island. Pachyptila desolata banksi, Smith, Dove Prion. This bird was only found nesting on Signy Island, and since the kind of ground required by the species for breeding is practically confined to Signy, it is probable that it is the only nesting place. The Scotia Expedition did not find the bird on Laurie Island. The nests are made in crannies among the tumbled rocks on the screes, mainly about the sloping parts of the land in preference to the cliffs. In the Borge Bay district there are many flat laminated boulders lying about the moss-clad slopes, and underneath these many nests were found. The nature of the ground usually precluded the digging of burrows, which is the habit of the birds in other localities. The nests are merely shallow hollows in the earth or rock debris in which they are made — a few feathers may form an inadequate lining. In January the slopes on which the birds were breeding were saturated with melted snow, and numbers of nests were very damp. Between January 17 and 21 about 120 nests were examined. Of these, forty contained single eggs, most of which were freshly laid. The most advanced egg was no more than a week incubated. The remaining nests had not yet been laid in, and in at least half of them an old and rotten egg from the previous year was found. That so many eggs should fail to hatch is not surprising when one considers the lateness of the breeding dates. In South Georgia the eggs are laid at least six weeks earlier, and the young do not usually leave the nests until April. By that time winter conditions are prevailing in the South Orkneys, and it is a matter for wonder that any chicks are able to survive. The Prions are not usually seen very near pack-ice and apparently do not like it. It is probable that they wait until Signy Island is clear of pack before they come ashore to breed, but later in the season the ice must often be a source of great inconvenience to them. In most of the nests in which no eggs were found, both parent birds were present in the daytime, but in the nests with eggs only the sitting bird was present. Prions are BIRDS OF THE SOUTH ORKNEY ISLANDS 367 nocturnal in their habits in the vicinity of land; they were never seen outside their nesting holes in the daytime, nor were they observed anywhere close to the shore at sea. At dusk they returned from sea, and flitted about in the vicinity of their nests. At this time the birds in the nests all commenced intermittently to give their low throaty call, and a continuous murmur could be heard all about the colony, leading to the discovery of many hitherto unsuspected nests. This species is evidently the only Prion breeding in the islands, for all the birds seen belonged to it, nor were any of the other kinds recognized at sea in the vicinity. The birds in the South Orkneys are in every way indistinguishable from the South Georgian birds. The eggs are similar, the plumage similar, and series of measurements of birds from the two localities are as follows: South Orkneys South Georgia Bill, length Breadth of bill at base mm. 29 14-5 mm. 30 15 mm. 29 15-2 mm. 30 14-5 mm. 29 13-5 mm. mm 30 — 14-5 14 Tarsus Wing ... 33 191 33 196 32 194 32 198 29-5 198 32 — 198 — Large numbers of the birds on Signy Island fall victims to skuas. Several skuas' nests were found in the vicinity of Borge Bay, and in every case the ground in the vicinity of the nest was thickly strewn with the remains of Prions. The skuas apparently caught them at dawn when they left their nests, or at dusk on their return from sea. Although Prions were never seen near the islands in daytime, in the surrounding seas they are usually very numerous. In February 1931 forty were seen thirty miles south of Coronation Island, and twenty at a position forty miles north of the island. In November at least a thousand were observed between ten and sixty miles south of Scotia Bay, and in January about 1500 between twenty and a hundred miles north of Laurie Island. Almost certainly these were all P. d. banksi. The islands were several times approached when they were surrounded by pack-ice, and on these occasions no birds at all were seen within ten miles of the ice. Halobaena caerulea (Gmelin), Blue Petrel. This species probably deserves a place among the birds of the South Orkneys on the strength of a report by one of the members of the crew of the ' Discovery 11'. Ordinary Seaman A. Jones reported on January 18 that he found two Blue Petrels breeding on the slopes above Borge Bay. The nests were among those of Pachyptila desolata banksi, and in similar sites. He was unable to capture the birds, but obtained an egg, which was unfortunately broken later. It was similar in size to that of P. d. banksi, but more elongate-ovate in form than is usual with the eggs of that bird. He described the birds as having "white-tipped tails and more black about the head than Prions", and there seems little doubt that the report is authentic. Jones had taken a great interest in birds at sea, and was able to identify many species. A diligent 368 DISCOVERY REPORTS search was carried out in the locaUty during the following three days, but no more nests were found or birds seen. The Blue Petrel is fairly frequently met with at sea in the South Atlantic, and is common in the Weddell Sea. In November 1931 three were seen about 300 miles north- west of the Orkneys. The Scotia Expedition captured two about 300 miles south-east of the group. On January i, 1933, twelve were seen between twenty and a hundred miles north of Laurie Island, with Prions. On this day the number of Prions was estimated at 1500, so that if the Blue Petrel nests on Signy Island in numbers proportionate with the numbers of the two species seen at sea, it is not surprising that only two nests were found. On other occasions a few birds were seen between the South Orkneys and Elephant Island, and again north of Elephant Island. Oceanites oceanicus (Kuhl), Wilson's Storm-Petrel. Wilson's Petrels are extremely abundant at the South Orkneys in the summer, being probably even more numerous than the Cape Pigeons. They breed round the coasts of all the islands. The nests, hidden in crannies among the rocks of the cliffs and screes, are difficult to find, and it would be easy to underestimate the population, particularly as the birds are mainly nocturnal. Eagle Clarke, in his account of the birds of the Scotia Expedition, comments on the remarkable regularity of the dates at which this species makes its annual appearance at the South Orkneys. November 11 and 12 were the dates on which the Scotia Expedi- tion saw the first birds. This regularity has been remarked in the South Shetlands also. Usually, in the summer, Bransfield Strait literally teems with birds, but on November 6, 1932, not a single bird was seen there. The next day several were observed north of Elephant Island, and in the following days increasing numbers were seen in the Scotia Sea. On November 20 birds were present in numbers near the South Orkneys. The second week in November is evidently the regular time of the species' annual arrival in this locality. The egg-laying dates appear to be subject to no such regularity. The Scotia Expedi- tion found the first egg on December 1 1 . This must have been a very early laying, for Eagle Clarke states that no chicks had hatched by February 21 of that year. Possibly the egg was unfertile. We found the first egg on January 4, in EUefsen Harbour, but several nests in this locality had not yet been laid in, though sitting birds were present at nightfall. In Sandefjord Harbour, where the birds are very numerous, twenty-four nests were examined, but none of them were laid in on January 10. Between January 17 and 20 two hundred nests were examined on Signy Island, and in about fifty of these freshly laid eggs were found. This lateness in breeding imposes an enormous disadvantage on the birds, and in view of their arrival at the islands a full two months earlier, it is difficult to account for it. That a heavy toll is exacted is proved by the fact that nearly half of the nests contained either rotten eggs from a previous season, or remains of dead young. These old eggs were usually turned out of the actual nest hollow, and lay near by in the cranny. BIRDS OF THE SOUTH ORKNEY ISLANDS 369 The nests are merely shallow depressions in the soil or rock debris in the crannies in which they are made, and the birds show very little intelligence in their choice of sites, except that they always select a hole with a narrow entrance. In one case in EUefsen Harbour, and in five cases in Sandefjord Harbour, the nests were actually made in blocks of ice which had evidently formed among the rocks during the winter. The birds almost certainly return year after year to the same situations, and in these cases had not sufficient initiative to forsake their nests when they found them iced in. The birds were only seen in small numbers near to the coasts in daytime, but in the evenings they were present in large numbers in all the harbours visited, flitting about in flocks over the surface of the water, and apparently feeding. At dusk they were con- tinually flitting round the cliffs and screes, and creeping mouse-like about the entrances of their nesting holes where their mates were on the nests. At this time they frequently utter a loud and characteristic cry, consisting of a harsh call of three notes, the first being about half as long again as the two following. To this the mate responds from inside the nesting hole with a similar call in a slightly lower key, and with the second and third notes more drawn out. This call is extremely loud, and on a still night can be heard at a distance of a mile or more. At times, from inside their nesting holes, the birds make a low whistling note. On several occasions, when in a boat very near to feeding birds flitting about in flocks in the harbours, we heard them utter small mouse-like squeaks, just audible. These notes were accentuated when a number of birds were gathered about a scrap of food. In Sandefjord Harbour on January 10 some birds were observed possibly courting, for among three squatting by a large rock two males were evidently making overtures to a hen. One of the males eventually departed, but pairing was not seen to take place. The Scotia Expedition last saw Wilson's Petrels at the South Orkneys on March 23, but in view of the lateness of their breeding dates it is probable that they do not usually leave the islands until well into April. On April 14, 1930, many birds were seen just north of the South Shetlands, and no doubt these were moving northward. By early May there are no Wilson's Petrels to be seen in the high latitudes of the Southern Ocean, and during the winter they are entirely absent. The majority of them no doubt perform the long migration to the high northern latitudes. Fregetta tropica melanogaster (Gould), Eastern Black-bellied Storm Petrel. This bird visits the South Orkneys in small numbers in the summer, and its breeding grounds are almost certainly confined to Laurie Island. At dusk small numbers of birds were seen, with Oceonites oceamcus, on different occasions in Jessie Bay, Scotia Bay, Wilton Bay and Marr Bay. Two were seen in the daytime in Washington Strait, but although a sharp look-out was kept not a single bird was seen to the westward. No nests were found anywhere but on Laurie Island. The Scotia Expedition found a nest in Uruguay Cove, and Eagle Clarke states that on December 5 it contained a deeply incubated egg and that both birds were present in the nest. This date seems very early indeed, and the fact that both birds were present 370 DISCOVERY REPORTS leads one to suspect that the egg was a well-preserved unhatched specimen from the previous season, for among the southern petrels it is uncommon to find both birds of a pair at the nest after the egg is laid. In the course of our work two nests were found in crannies among rocks on the steep slopes of the eastern side of Wilton Bay, at a height of about 150 feet. In both these the egg was fresh, not laid more than a few days, and a single parent was present in each case. This was on January 28, so that the egg-laying dates must coincide closely with those of O. oceaniciis. One of the sitting birds was captured and preserved. The bill, legs and feet are all black, and the iris very dark brown. Bennett (1926) comments on a peculiar whistling note which he believes this bird to utter in the neighbourhood of its nest. Eagle Clarke also mentions a whistling from the nest. This habit is not, however, confined to this species, for Wilson's Petrels are often heard to make similar sounds. The two birds seem to be very similar in their habits, with one notable exception. Wilson's Petrels nearly always follow in the wake of a ship at sea, while Fregetta tropica melanogaster is usually seen flying ahead of or around a ship and rarely follows the wake. It also appears to be a much more agile bird on the wing, and the following note was made on birds observed at sea between the South Orkneys and South Shetlands : " The flight of these birds was again seen to be most distinctive. They fly rapidly along about four feet above the surface of the sea, and at intervals make a series of short, oblique dashes down to it, striking it with the feet held out and ' skating ' for a distance of three or four feet, raising a considerable splash." The birds are moderately common in the high latitudes of the South Atlantic in summer; and in the Drake Strait, Scotia Sea, and the vicinity of the South Orkneys and South Shetlands, small numbers are recorded as having been observed on almost every day that the ' Discovery II ' was at sea. In the winter they probably migrate to lower latitudes. Bennett found this bird breeding in small numbers on Deception Island, and it is probable that the eastern islands of the South Shetlands also furnish breeding grounds. Catharacta skua lonnbergi, Matthews, Brown Skua. A great deal of confusion has prevailed with regard to the subspecific races of the Great Skua, and in recent years the southern skuas have received considerable attention. The latest analysis by Hamilton (1934) is an important contribution to the subject. A long series of the measurements of mated pairs, from the South Shetlands and South Orkneys, is required to confirm his conclusions. This might prove that the birds of Hamilton's two subspecies interbreed, in which case the subspecific characteristics are no longer valid. On the other hand, if the opposite were proven his case is beyond question. It may be noted that Bennett (1926, p. 319) states that in the South Shetlands he observed that breeding pairs were alike in size. Hamilton was only able to measure thirteen specimens from the South Orkneys, twelve from the South Shetlands, and nine from South Georgia, and he remarks on the BIRDS OF THE SOUTH ORKNEY ISLANDS 371 fact that the greatest size variations are found in the longest locaHty series of measure- ments. Field observations lead me to suspect that the South Orkney skuas vary in size to a much greater extent than is suggested by his figures. A male bird shot in Ellefsen Harbour measured — wing, 419 mm.; culmen, 51 mm.; tarsus, 79 mm. — this was noted to be a " large " bird. Several more were shot in Borge Bay, both "large" and "small", but the measurements were unfortunately lost. Hamilton remarks on the unreliability of field observations alone when the size of birds is in question. In this he is right, but when several birds together are closely observed ashore it is sometimes possible to detect differences in size. Certainly the majority of the South Orkney birds appeared to be "large". Skuas are common in summer all round the group, being probably most numerous in the vicinity of the penguin rookeries. On January 5 nests in the vicinity of Ellefsen Harbour contained hard-set eggs, and on January 8 eggs in the nests in Sandefjord Har- bour had reached approximately the same stage. A great diversity of size and colouring is apparent in the eggs ; the ground colour varying from olive-brown to blue-green. In Borge Bay the breeding operations were much farther advanced, for on January 17 five nests were found, belonging to which were three single chicks and two pairs. The chicks varied in age from a few days to perhaps five weeks, the latter being lanky youngsters with the quills of their first plumage replacing the nestling down. The Scotia Expedition found chicks a week old on January 29 on Laurie Island and it is evident that the breeding dates must be subject to considerable variation. In the Borge Bay district the ground in the vicinity of the nests was littered with the remains of Prions, and it was clear that these formed the staple diet for the local skuas, for there were no remains of any other species. In other parts the birds subsist to a great extent on the eggs and young of penguins. The nests, comfortably lined with lichens and moss, are usually placed on the crests of local eminences, and their immediate vicinity is shunned by other nesting birds. This ensures comparative safety for the skua chicks, which leave the nest very soon after hatching and wander about in the neighbourhood, where they would soon fall a prey to Giant Petrels if any were established close by. The Scotia Expedition found that no skuas remained in the South Orkneys after April, and they were never seen during the winter. It would be interesting to know to which coast they resort during the winter months, for although skuas are sometimes seen at great distances from land, it is highly improbable that the sub-Antarctic skuas spend the winter in a pelagic condition. During the winter cruises of the 'Discovery 11' between South Africa and New Zealand and the ice-edge, only one skua was seen farther than 500 miles from land or south of latitude 45° S. The exception is a solitary specimen seen in 49° S, 120° E towards the end of May. Catharacta skua maccormicki (Saunders), McCormick's Skua. An example of this bird was obtained in Scotia Bay by the Argentine naturalists in 372 DISCOVERY REPORTS November, 1904. During the visit of the 'Discovery 11' no birds were seen in the group, and the normal hmit of range of McCormick's Skua Ues several degrees to the southward. The species no doubt occurs from time to time as a straggler, and the most likely times for its appearance would be during the spring and the autumn. Bennett states that he has seen a few birds in Graham Land, but during the periods that the 'Discovery 11' was in this locality the bird was never definitely identified. Hamilton (1934) states that he did not find the bird in the South Shetlands although formerly it bred there. Larus dominicanus, Licht, Southern Black-backed Gull. This bird breeds in small numbers round most of the coasts of the South Orkneys, and during the entire month we spent at the group a few birds were nearly always in sight from the ship. In all the harbours and straits some were present, and at frequent intervals round the coasts one or two birds would appear and remain in company for short periods. The only area where they were comparatively scarce was along the north coast of Corona- tion Island, where the steep cliffs of the shore allows of no suitable nesting grounds. The Scotia Expedition estimated the summer population of Laurie Island at about three hundred birds, and probably twice that number nest among the other islands of the group. The gulls are the earliest breeders of all the South Orkney birds, for their eggs are laid in the latter half of November. In South Georgia breeding takes place at about the same time. Bennett (1926, p. 318) states that in the South Shetlands the bird breeds six weeks earlier than in the Falklands. This species is practically the only exception in the south to the usual rule that birds with a wide latitudinal breeding range lay their eggs progressively later the higher the latitude of their nesting places. Several nests were found round EUefsen Harbour, but the chicks had all left in early January and none were seen. The brownish chicks are always difficult to see on stony ground, and on the approach of an intruder they often remain quite still, rendering themselves almost invisible. The nests consist of scraps of lichen, feathers and seaweed, and are easy to find even after they have been vacated, on account of the number of limpet shells which always lie scattered around them. All the nests found were on rather flat ground, usually on the top of low slopes behind the beaches. A number of nests were present in the Borge Bay district, and here two chicks were observed. In these birds almost all the down was gone, and they were estimated to be at least six weeks old on January 18. Bennett is of the opinion that the South Shetland gulls are subspecifically different from those of the other parts of the South Atlantic. No noticeable superficial difference is to be observed in the South Orkney birds. The Scotia Expedition found that a few gulls remained throughout the winter. These were possibly induced to stay by the presence of the ship, but the birds would no doubt be able to sustain themselves through the winter. A principal article of their diet is the limpets which are found on the rocks in many parts of the coasts, and these would always be obtainable. BIRDS OF THE SOUTH ORKNEY ISLANDS 373 Sterna hirundinacea, Lesson, South American Tern. The Scotia Expedition found this species breeding on Laurie Island to a total number of from two to three hundred birds. During the visit of the ' Discovery II ' no nests of terns were examined on Laurie Island, and only a few birds were seen, usually singly, at sea round the coast. The localities in which they were found nesting by the Scotia Expedition were not visited, but no doubt the birds still use these breeding places. Sterna vittata georgiae, Reichenow, Wreathed Tern. Both this bird and S. hirundinacea breed in the South Orkneys, and the Scotia Expedition found that the terns breeding on Laurie Island were all of the latter species. It is impossible to differentiate between the two birds by field observation alone during the breeding season, and as only one tern was captured during the visit of the ' Discovery II ', we were not able to identify the species inhabiting some of the colonies. I am of the opinion, however, that most of the terns breeding in the group belong to the species S. V. georgiae, with the exception of the Laurie Island birds, which were not closely observed. A small ternery is established on the low islet forming the south-eastern boundary of EUefsen Harbour. This was visited on January 5, and found to contain about fifty nests. Most of these contained single hard-set eggs, and one newly hatched chick was seen. The nests are shallow depressions in the rubble lying among the rocks, and are lined with small pebbles. A number of terns were seen off Fredriksen Island and in Lewthwaite and Washing- ton Straits. Along the north and west coasts of Coronation Island they appear to be entirely absent, for these coasts provide no suitable nesting sites. Only three birds were seen in Sandefjord Harbour during a stay of several days. Signy Island, however, provides breeding places for perhaps a thousand birds. Two considerable colonies are established on the shores of Borge Bay, the nests being placed high up on the beaches and on the rising ground at the back of them. All the nests examined contained single eggs, and on January 16 these were all about half-set. Thus, with this species again, the breeding dates on Signy Island are later than those of the birds in EUefsen Harbour. A few patches of snow were still lying about the slopes of the land in Borge Bay, and probably this more sheltered area would carry more snow than would the comparatively wind-swept islets about EUefsen Harbour. Consequently the EUefsen Harbour sites would be clear of snow earlier in the summer and this may account for the difference in the breeding dates. The nests in the colonies in Borge Bay were found to be spread over a considerable area of ground. Sometimes two or three nests might be placed within a few feet of one another, but as a rule they were widely scattered. No birds, however, were found nesting in complete isolation. Terns which breed in high latitudes are usually migratory, and it is probable that this species is entirely absent from the South Orkneys during the winter. There appears to be no record of its occurrence at the group during the winter months. 374 DISCOVERY REPORTS Phalacrocorax atriceps, King, Blue-eyed Shag. Blue-eyed Shags breed in colonies on a number of detached rocky islets throughout the South Orkneys. No rookeries were found anywhere on the main islands. They are probably most numerous round Laurie Island, where the Scotia Expedition estimated the summer population to be about 2500 pairs. A large rookery is established on Holmen Gras Islet, just south of EUefsen Harbour, and another on the southernmost of the Inaccessible Islands. On several islets the birds share their rookeries with Ringed Penguins. Since the rocks are usually steep-to, the Ringed Penguin is the only species which would choose to nest on them. In the Palmer Archipelago shags are found sharing their rookeries with Gentoo Penguins. The rookery on Holmen Gras was visited on January 5. The nests all contained two or three chicks, varying in age from a few days to about a fortnight. The chicks are com- pletely bare at birth, the skin of the back being brown-black, and of the underparts pink. After a few days they are covered in brown-grey down. In the chicks the bill is mainly brown, the lower mandible being pinkish bluish at the base, shading into brown at the tip. In the adult birds the bill is dull grey-brown, the lower mandible being paler at the tip. The feet are pinkish, turning into brown-black at the edges of the webs. The birds had lost their nuptial crests at this time, and the caruncles were faded yellow. On January 5 when a survey party was landed on the beach of a small bay on the western side of Fredriksen Island, a flock of shags came swimming slowly in from the northward, approached to within a few yards of the shore and moved off to sea. The flock was in perfect arrowhead formation and the birds were closely packed in the flock. There were about two hundred of them, and as they swam slowly along with only their snake-like heads and necks showing above the water they presented an extraordinary spectacle. It was considered likely that these birds were from some near-by colony, possibly Holmen Gras, on a foraging expedition, and that they were attracted towards the beach by the presence of the boat. On January 26 a rookery on the islet on the western side of Wilton Bay, in Laurie Island, was visited. About 180 nests are established here. The young were all in first plumage and had left the nests. They wandered about the rookery, which was in a very filthy condition, and when disturbed showed little inclination to take to the sea. A few of them swam, but it was evident that they were not yet proficient in the water. These birds were considered to be six or seven weeks old, and at this stage the caruncles are not developed, the eyelids are greyish without a trace of blue, and the legs and feet are grey-brown, with pinkish patches in the webs. The bill is similar to that of the chick. The Scotia Expedition saw only a few birds during the winter. It is suggested that the majority migrate to the Patagonian Islands, for on April 15, 1930, a number of shags were observed near the ship in a position 1 50 miles south-south-east from Cape Horn. These were definitely P. atriceps, and it seems likely that they were Antarctic birds on migration, for it is improbable that birds from Tierra del Fuego would stray so far afield. BIRDS OF THE SOUTH ORKNEY ISLANDS 375 Chionis alba (Gmelin), Sheathbill. (Plate XII, fig. 4.) Sheathbills are common among all the penguin rookeries in the South Orkneys, and even small isolated branches of colonies, containing only a few penguins, are usually found to include at least a pair of the birds. They show no apparent preference for the company of any particular species of penguin, but they are rarely seen far from a rookery. The Scotia Expedition estimated the total population of Laurie Island at two to three thousand birds, and since the number of Sheathbills probably follows closely in proportion with the number of penguins, the population of the whole group must be about twice that number. Our estimate of the population of the group, exclusive of Laurie Island, was about three thousand birds. Nests were found in Ellefsen Harbour on January 4, and on this date all of them con- tained eggs. Most of these had not long been laid, but a few were deeply incubated. In Sandefjord Harbour, on January 9 a number of nests contained eggs in various stages of incubation, but no chicks were found. Two of these nests contained four eggs ; in all the rest two or three were found. In Borge Bay two nests were found on January 18 containing half-set eggs. The Scotia Expedition found chicks on Laurie Island on January 7, and it appears that with this bird, as with some others, the breeding dates are slightly later in the western end of the group than in the eastern part. On January 9 an empty Sheathbill's nest was found in Sandefjord Harbour. Whether this had not yet been laid in or had been robbed it is impossible to say, but the two birds were standing on a rock near by, going through what was evidently a courtship cere- mony. They stood facing one another about two feet apart and, keeping their positions, together performed a series of quick little bows, in a stiff mechanical movement from the hips so that at each bow the tail was elevated. After fifty or more of these bows both birds remained quite still for a moment, when they commenced a further series. This performance was continued for fifteen minutes, when the birds appeared to tire of it and resumed their customary quick inquisitive hopping about the rocks. Pairing was not seen to occur. The nests, usually placed in well-selected holes among jumbled rocks, consist of collections of penguins' tail feathers, remains of egg-shells, small stones, scraps of lichen and miscellaneous rubbish, together with disgorged pellets of indigestible matter. The ground round the actual nest is always strewn with these materials (Plate XII, fig. 4). When the nest is approached, the sitting bird will often commence a harsh chatter, finally fleeing with alarmed cackles. Sheathbills are practically omnivorous, but a great part of their diet in the South Orkneys consists of penguins' eggs, faeces, and young. The latter they usually find dead, but no doubt they would not hesitate to hasten the end of a dying youngster. They have the skuas as rivals, with whom they find it impossible to compete, and probably few young penguins are actually killed by Sheathbills. Mr A. Saunders, the laboratory assistant, saw two Sheathbills attacking an ailing penguin chick on one occasion. 376 DISCOVERY REPORTS The Scotia Expedition found that a few birds remained throughout the winter in the neighbourhood of the ship's winter quarters; and the Argentine meteorologists stated that a few remained in the vicinity of their hut nearly every winter. Except for these birds, which find artificial means of sustenance, it is almost certain that the Sheathbills leave the islands during the winter, for it would be extremely difficult for them to find sufficient food when almost all the other birds had departed. It is suggested that they migrate to the islands and channels off the coast of Patagonia, where they are believed to occur during the winter. Several birds were seen at sea off Cape Horn from the ' Discovery II ' in the middle of April 1930, and two days previously one was in company with the ship, then northward-bound, midway between the South Shetlands and Cape Horn. These were the only occasions on which a Sheathbill was seen far from land during the cruises of the ' Discovery II '. At sea the birds perched about the ship and ffew at a considerable height above the surface of the water. They were never seen to feed, and they probably perform their sea migrations without feeding. LIST OF LITERATURE Bennett, A. G., 1926. A List of Birds of the Falkland Islands and Dependencies. Ibis, pp. 306-33. Clarke, W. Eagle, 1906. Birds of the South Orkney Islands. Ibis, pp. 145-87. Hamilton, J. E., 1934. The Sub-Antarctic Forms of the Great Skua (Catharacta skua skua). Discovery Reports, IX, pp. 161-74. Lowe, P. R. and Kinnear, N. B., 1930. Birds. British Antarctic ('Terra Nova') Expedition Reports, Natural History, iv, part 5, pp. 103-93. London. Matthews, L. Harrison, 1929. Birds of South Georgia. Discovery Reports, i, pp. 563-92. PLATE X Fig. I. Pygoscelis antarctica and chicks. Sandefjord Bay, Coronation Islands, lo. i. 33. The chicks are only about two days old and the nest is still almost intact. Fig. 2. Pygoscelis antarctica climbing to their nests on a steep slope in Sandefjord Bay. The rock here does not always afford good foothold and very great difficulty is experienced by the birds in landing and climbing to their nests in this and similar situations. DISCOVERY REPORTS, VOL. XII PLATE X ■■■■ ^^^^^^^^^Ef ^^^^^^^r iH ^V^^HHHi ' "*"^Hk >t!iiH ^^fc^-.^i^Lf '^' ^^M ^^^V^' ^lEt '^tr-^^^^^f^^^^^^t^lUH Wum ^^^^Rk . ^t^'.^^Kf ■I'^'i^l^^^^^^k^^^^^^^^^^^^^^B^^^I BiMlK^ft|^wn^«IH ^■■^Hh^^^^H Pn ^m^ \ • 1 ^1^' '\\ ^r '''Jl ^i ■(11 CM Q < z c X f- D O U r o Q S PLATE XI Fig. I. Pygoscelis adeliae coming ashore at the rookery on Michelsen Island, Ellefsen Harbour. The birds have just left the sea and are com- paratively clean. The apparently aimless behaviour of the birds is cus- tomary when they are away from their nests. Fig.2. Priocellaantarctica&nd chickin a nest in Sandefjord Bay, io.i.33. DISCOVERY REPORTS, VOL. XII PLATE XI ' ^-^.t'-^-'X - *^-:\'^r- JohnBnlf- Suns 4 tian.rliioTi L*^ London BIRDS OF THE SOUTH ORKNEY ISLANDS PLATE XII Fig. I. Macronectes giganteus and chick. Signy Island, 19.1.33. The bird is a "dark intermediate" and the chick is in light grey down. Fig. 2. Colony of M. giganteus, Michelsen Island, EUefsen Harbour, 4. i. 33. The general lightness in plumage of the nesting birds is illus- trated. Two white birds are in the centre, and all the rest, with the exception of the bird in the left foreground, are "light intermediates". Fig. 3. Pygoscelis papua feeding young, Ellefsen Harbour, 13.1.33. The parent bird has climbed on to the rock in order more conveniently to carry out the operation, the chick being almost as large as itself. Fig. 4. Chionis alba. Ellefsen Harbour, 12. i. 33. Three birds standing in a characteristic statuesque pose on a rock which contains a nesting hole. The approach to the hole is littered with rubbish which appears lighter than the surrounding rock. DISCOVERY REPORTS. VOL Xll PLATE Xll ,^ Sunr. A t.nnirlwttv L''' Loudon Fig 2 R A B Ardlsy phot BIRDS OF THE SOUTH ORKNEY ISLANDS [Discovery Reports. Vol. XII, pp. 377-440, August, 1936.] LARVAE OF DECAPOD CRUSTACEA PART I. STENOPIDEA PART II. AMPHIONIDAE PART III. PHYLLOSOMA By ROBERT GURNEY, D.Sc. CONTENTS Introduction P> XI li2°33Wi 36 90 1-71 0-82 0-71 31-3 2-b5 8-4 4-3 ji between the smallest and largest with regard to the development of antennae and the rudiment of the uropods, but they grade into one another and there is no evidence for a moult separating them into two stages. Measurements of the specimens are given in Table V. Table V. Measurements {in mm.) of Phyllosomas belonging to Panulirus, Form B Antennule Stage Station Length Abdomen Antenna Eye Eye- Body Exopod of Segment I Segment 2 Segment 3 Stalk Eye-stalk maxillipede 2 V •691 7-6 0-5 0-37 0-20 0-20 063 1-3 1-95 3-86 None V 691 93 06 0-49 0-22 0-24 108 1-35 2-5 372 >) V 691 100 06 0-51 027 0-20 i-6s 1-5 30 3-33 ,, V 691 no 0-75 0-57 0-29 0-27 162 1-5 3-3 3-33 ,, VI 690 12-5 0-95 o-6i 0-29 0-31 2-0 1-7 3-7 3-37 )i VI 691 140 i-i 076 0-33 035 2-75 1-8 4-3 3-25 ,, VII 690 170 20 0-92 039 0-41 4-2 19 4-6 3-7 Rudiment VIII 690 20 34 I 06 0-41 051 5-6 2-2 5-7 3-S Longer, with- out setae IX 690 23 5-75 1-24 0-59 059 IO-7 2-2 5-6 3-9 With setae VIII 438 28 53 163 067 067 II-5 24 8-4 3-3 Without setae Stage V? (Fig. 21). Length 7-1 1 mm. Fore-body very narrow, much narrower than hind-body. Hind-body deeply concave behind. Abdomen very small, unsegmented, and without trace of pleopods. The uro- pods are represented by an inconspicuous fold or by a pair of distinct lobes. Peduncle of antennule of three segments. Antenna shorter than peduncle of anten- nule, or, in larger specimens, considerably longer. LARVAE OF DECAPOD CRUSTACEA 413 Fig. 21. Panulirus B, Stage V ?, 7-7 mm. St. 692. 414 DISCOVERY REPORTS Maxillule without palp. Maxillipede i either quite absent, or present as a minute papilla. Maxillipede 2 without exopod. Maxillipede 3 with setose exopod, but without coxal spine. Legs without coxal spine. Leg 4 endopod not fully developed, the dactylus rounded at end, without spines. Leg 5 a minute papilla. Stage VI ?. Length 12-5-17 mm. Peduncle of antenna segmented, the flagellum as long as the antennule. Fig. 22. Pamilirus B, Stage IX ?, 26 mm. St. 689. Maxilla not enlarged distally, without setae. Maxillipede 2 without exopod. Leg 5 longer, unsegmented. Abdomen larger, not segmented. Pleopods represented by simple papillae. Uropods bilobed. Stage VII ?. Length 16-20 mm. Flagellum of antenna longer than antennule. Maxilla with exopod produced backwards, without setae. Maxillipede 2 with rudi- ment of exopod, without setae. Leg 5 larger, in some cases with basal segment distinct. LARVAE OF DECAPOD CRUSTACEA 415 Abdomen segmented dorsally. Pleopods bilobed. Uropods large, biramous, the exopod jointed to basis. Stage VIII ?. Length 20-23 ™rn- Antenna very much longer. Maxilla much larger, but not reaching back as far as maxillipede 2. Maxillipede I finger-like, enlarging at base. Exopod of maxillipede 2 longer, but without setae. Leg 5 reaching to end of somite 5, with basal segment distinct. Abdomen fully segmented ; somite 6 without dorsal spines. Pleopods deeply bilobed ; uropods with both rami jointed to basis. Stage V? Stage VI? Stage VII? Fig. 23. Pmudirtis B, Stages V-VIII?. a. Antennae. b. Maxilla, etc. c. Abdomen. Stage VIII? Stage IX } (Fig. 22). Length 23-26 mm. Maxilla reaching back almost to maxillipede 2, with setae along margin. Maxillipede i with large epipod. Maxillipede 2 with exopod bearing setae. Maxillipedes and legs with gill rudiments and epipods ; one pleurobranch on legs i and 5 and two on legs 2-4. Leg 5 with five distinct segments. Abdomen fully segmented, somite 6 with a pair of large dorsal spines at base of uropods. Pleopods large, with appendix interna. Uropods with small tooth on outer margin of exopod. Panulirus ?, Form D (Figs. 24-27) Fore-body pear-shaped, narrow in front, narrower than hind-body. Hind-body slightly hollowed behind in later stages. Antennule with segments 2 and 3 of peduncle nearly equal, and about half length of segment i . Antenna slender, longer than antennule in earliest stages. A 4i6 DISCOVERY REPORTS Maxillule without palp. Maxilla with exopod setose. Maxillipede 2 without exopod in early stages, later with setose exopod. Maxillipede 3 with setose exopod from earliest stage. Legs with large ventral coxal spines, and large dorsal coxal spines on legs 2 and 3. Leg 5 unbranched, close to abdomen. Dactylus of leg 2 elongated. The position of this Phyllosoma is very doubtful. In general form, and in the lengths of the antennular segments, it agrees with those Phyllosomas which are referred to Panulirus, but in the retarded development of leg 4 it resembles Paliminis. It seems most probable that it belongs to a distinct genus, but I refer it none the less provisionally to -Pamilinis. The series of specimens is not suffici- ently large to give good evidence for the number of stages passed through, but it is possible to distinguish seven, or per- haps eight. As the smallest must be in stage II or III, the total number would be eight or nine. Measurements of the specimens are given in Table VII. The specimen of 7-4 mm. is much more ad- vanced in some respects than that of 8 mm. and may perhaps represent a different species. As will be seen from Table VI, this form was taken at 11 stations, from 11° 25' S, 42° 03' E (off Zanzibar) south- wards to beyond the Cape in 38° 02' S and i7°5o'E. All these specimens appear to be of one species. In addition two small specimens from St. 701 in the Atlantic are included in the table and shown in Fig. 24. They do not belong to the same species, being of rather different shape, and showing some differences in the spines on the legs; but they cannot be referred to either of the Atlantic Fig. 24. Panulirus D, 5 mm. St. 701. species of Panulirus described above. If they are, as they appear to be, closely allied to the eastern form, that would be some evidence that this is a generic rather than a specific type. Genus Jasus, Parker The first larva of the eastern form oi Jasus lalandii has been described by Thomson (1907), Anderton (1907) and Archey (1916), and for the South African form by Gil- christ (1913). In both cases the larva, when first hatched, has a very peculiar form. The LARVAE OF DECAPOD CRUSTACEA 417 antennae are large biramous organs with long setae, by means of which the animal swims actively, while the maxillipedes and legs are coiled up and functionless. Moulting takes place in a few hours, and a normal Phyllosoma appears. To the first stage Gilchrist gave the name Naupliosoma. It is clear that the naupliosoma represents the prezoeal stage which, in this case alone among the Loricata, so far as is known, retains the swimming antenna, as in some Brachyura. If the prezoea can rightly be regarded as a reminiscence of the nauplius, the name naupliosoma is well chosen. Gilchrist suggests that this active stage is retained in order that the surface may be reached more rapidly than a normal Phyllosoma would be able to do. Table VI. List of stations at zv/iich Phyllosomas o/Panulirus, Form D, were taken Station Position Depth No. of Size metres specimens mm. 701 14° 39' N, 25° 52' W 242-0 2 5 1578 11° 25' S, 42° 03' E 500-0 I 4 1576 14° 42' S, 42° 22' E 400-0 I 29 1575 i8^33'S, 4i°35'E 800-550 I 7 1574 2i°45'S, 40°34'E 600-0 I 12-5 1571 27°24'S, 39°2i'E 500-0 I i3"5 441 3i°24'S, 3i°i2'E 180 I 20 1373 3i°i3'S, 3i°49'E 135-0 I 12 1568 34° 48' S, 34° 28' E 1400-0 4 9-16-6 1375 34°3i'S, 26°i9'E 210 I 137 407 35° OS'S, i7°49'E 150-0 2 8,15 1554 38°03'S, i8°4o'E 1500-0 2 5-7, 8-5 Table VII. Measuremetits {in mm.) of Phyllosomas belonging to Panulirus, Fr- ,n D Station 1578 701 1554 1575 '.'^7 1 1554 1568 1568 Length Eye, with stalk Antennular peduncle Segment i Segment 2 Segment 3 Exopod of maxillipede 2 4-15 1-88 j?l S c 5-0 1-95 57 2-2 c B 0 E i 7-4 3-15 0-57 0-35 0-29 8-0 3-0 S II 8-5 3-2 0-59 0-31 0-35 9-0 3-5 0-45 0-22 0-22 II-9 4-4 0-35 0-29 0-27 Station 1373 1574 J IS7V 1375 407 1568 441 1576 Length Eye, with stalk Antennular peduncle : Segment i Segment 2 Segment 3 Exopod of maxillipede 2 12 12-55 4-6 0-65 0-33 0-31 Rudi- men- tary 13-4 4-8 061 0-33 0-33 Rudi- men- tary 137 5-2 o-6i 031 029 Short small setae 15 0-65 0-37 0-39 Short, no setae i6-6 6-2 1-56 0-73 0-71 Small, no setae 20 0-98 0-45 0-43 Small, with 6 setae 29 9-5 1-56 073 071 Large, setose 4i8 DISCOVERY REPORTS It is a matter of great difficulty to determine how many stages there may be in the development oi Josus, since apart from stage I Gilchrist obtained only a single specimen (of 3-8 mm.) of another early stage. Von Bonde (1932) did not add to our knowledge of this species. The Discovery material consists of about fifty specimens all much older than this specimen of Gilchrist's. Taking as a basis specimens of four distinguishable stages from St. 85, it is found that these increase in size by a growth factor of about 1-3, Fig. 25. Panulivus D, 57 mm. St. 1554. whether total length is taken or pre- or. post-labral length, showing that in these parti- cular stages (from 10-4 to 2275 mm.) irrere is no appreciable differential rate of growth. The fifty specimens measured fall into six groups of which the average sizes are : Average size (mm.) Growth factor 10-5 — 13-57 1-29 i6-5 1-22 20-8 1-26 28-6 1-36 35-0 1-23 LARVAE OF DECAPOD CRUSTACEA 419 The growth factor remains about 1-3 or less, and one cannot therefore assume a greater rate of growth for earher stages. If that is so we are forced to conclude that there are as many as thirteen stages, and that Gilchrist's larva of 3-8 mm. is stage 4. Bouvier and Santucci agree in finding nine stages in Palinurus vulgaris, and Stephensen found the same number in Scyllarm arctus. Gilchrist's larva is even less developed than stage II of Palimirus, and it is difficult to believe that it can represent a later stage. In any case Fig. 26. PamiUnis D, 8 mm. St. 407. there seems to be a gap of three unknown stages between this larva of 3-8 mm. and the youngest of the Discovery specimens. Our knowledge of the development of the European forms is not so securely founded that we are entitled to assume a similar course in other genera and species, and having regard to the wide distribution and great size reached by some of them, there may well be a longer period of larval life and a greater number of stages. Consequently I have, for the purposes of this report, accepted a provisional arrangement into thirteen stages, 6-2 420 DISCOVERY REPORTS which has the advantage of drawing special attention to the wide gap in our knowledge of this species. Fig. 27. PamiUrus D, 13-7 mm. St. 1375. a. Antennae. b. Maxilla, etc. c. Abdomen. Characters of the Phyllosoma of Jasvs lalandii. Fore-body as wide as or wider than long. Hind-body very much narrower than fore-body, not hollowed behind. Abdomen narrow. Antenna slender, with long flagellum. Maxillipedes 2 and 3 without exopod, or with minute rudiment of it. Maxillule without palp. Leg 5 closely apposed to abdomen, developing very early and without exopod, or with minute rudiment of it. While the structure of the antennae and the form of the abdomen are those charac- teristic of a genus of the Palinuridae, the absence of the exopod from maxillipede 3 is remarkable. The exopod of maxillipede 2 always develops late and its loss, even in a Palinurid, would not be surprising; but the absence of the exopod in maxillipede 3, seeing that it is functional from the earliest stage in Palimirus and Pamiliriis, is most unexpected. It is absent in all Scyllaridae known. Jasus lalandii (Lamarck) Stage VIII .? (Fig. 28). Length lo-ii mm. Carapace broader than long. Abdomen very small, parallel-sided, unsegmented, with a small spinous process at each posterior angle. Eye-stalk longer than eye, eye and stalk together more than one-third of total body length. Antennule with peduncle of three segments, segments 2 and 3 about equal, and about half length of segment i ; endopod about one-fifth of exopod. Antenna with pe- duncle indistinctly segmented, a small inner spine at position of end of segment 3. LARVAE OF DECAPOD CRUSTACEA 421 Maxillule without palp. Maxilla with distal part not expanded, with four setae. Maxillipede i just traceable as a small swelling. Maxillipedes 2 and 3 without trace of exopod. Maxillipede 3 and legs each with large spine on coxa. Leg 5 fully developed, very long, without exopod. Pleopods traceable under skin. Uropods represented by very small simple folds. Fig. 28. Jasiis lalaiidii, Stage VIII?, 11 mm. St. 254. 422 DISCOVERY REPORTS Stage IX?. Length 12-5-15 mm. Abdomen broadened at base; telson marked off, with large spinous process at each angle. Eye about one-third of body length. Antennule, segments of peduncle the same, endopod and exopod longer. Antenna with peduncle distinctly segmented, with inner spine of segment 3. Maxilla unchanged; maxillipede i a small papilla. Maxillipedes 2 and 3 without exopod. Pleopods present as small simple lobes ; uropods larger, but generally simple. In some cases a distinction of endopod from exopod is traceable. Stage X?. Length 15-17 mm. The only marked difference between this and the preceding stage is that the uropods are distinctly bilobed, but there is an increase in length of the endopod of the antennule and of the antenna which is shown in Table VIII. While the maxilla is generally unchanged, it may, in the largest specimens, show enlargement of the exopod, which may have seven setae. Stage XI ?. Length 19-23 mm. Abdomen stout, the telson narrowing and rounded at end, without spines. Eye less than one-third of body length. Maxilla with exopod greatly enlarged, with about eleven setae, but not reaching back as far as maxillipede 2. Maxillipede i larger, but still a simple process. Maxillipedes 2 and 3 without exopods. Pleopods larger, bilobed ; uropods nearly as long as telson, biramous, but the rami not segmented. There may be a trace of a joint at the base of the exopod. Stage XII }. Length 27-31 mm. This stage differs very little from stage XI except in general growth as shown by the measurements, but may easily be distinguished by the larger, biramous, pleopods with- out trace of appendix interna, and the larger uropods. The exopod of the latter is distinctly jointed, and the outer margin smooth. Telson as before, with smooth margin. Stage XIII ? (Fig. 29). Length 33-37-5 mm. Maxilla with exopod extending back to maxillipede 2. Maxillipede i with large epipod and finger-like process representing endopod. Maxillipede 2 with rudimentary exopod. Maxillipede 3 with small knob representing exopod. Small epipods and gills present on legs in full adult number. Pleopods large, flattened, with appendix interna on each. Uropods slightly longer than telson, the rami jointed to basis, exopod serrated on outer margin. Telson truncated, with small sharp points along margin. This, which is without doubt the last stage before the moult to the Puerulus, has been fully described by Gilchrist. He drew attention to a peculiar elevation dorsally behind LARVAE OF DECAPOD CRUSTACEA 423 the eye, which he interpreted as the rostrum. I do not find this structure as prominent as he figures it, but it is quite possible that it may be the forerunner of the rostrum. Fig. 29. Jasiis lalandii, Stage XIII ?, 37 mm. St. 100 B. Distribution. Table IX shows the distribution of seventy-one specimens. The most northerly station is 268 (18° 37' S) and the most southerly is 448 (39° 03' S) ; but the majority were taken in the region of the Cape and between the Cape and Tristan da Cunha, where the adult is known to occur. Apart from St. 247, which is close to Tristan, it was not taken to the west of that point, nor in the neighbourhood of Gough Island. The area of distri- bution is therefore restricted, and seems to correspond closely with the limit of the subtropical convergence. The smallest individuals, of lo-ii mm., are extremely rare, but it is worth noting that one of them (St. 85) was taken about 900 miles from the nearest 424 DISCOVERY REPORTS land. Unfortunately, as the series of larvae is so incomplete, one can only guess at the age as expressed in the number of the stage, and I am by no means satisfied that the assumption that this larva is in stage VIII is correct. At all events it cannot be older. We know too little about the duration of stages to estimate age in days, but a guess of six weeks may be fairly near the truth. In any case there is evidence here of rapid and extensive travel. Table VIII. Measurements of Phyllosomas 0/ Jasus lalandii Stage Length Antennule Eye Eye- stalk Antenna Pre- Post- Peduncle Exopod Endopod Total labral labral Segment i Segment 2 Segment 3 VIII 10-4 4-4 60 2-15 I I 3-68 0-69 1-65 2-1 2-8 IX I2-9S 5-65 7-3 2-36 I 0-85 4-0 0-84 1-95 2-5 3-2 X 16-85 7-3 9-SS 2-47 I i-i 3-8s 1-2 2-3 3-4 5-6 XI 22-75 9-75 12-95 2-32 I I-I3 3-75 1-67 2-5 3-7 8-1 XII 28-5 12-0 16-5 2-29 I I-I3 39 2-5 3-0 4-7 II XIII 37-5 Specime 15 IS from St. 22-5 1377 2-45 I 1-3 3-35 2-7 30 S-o 17 X 17-8 7-8 100 2-SS I 1-24 41 1-43 2-4 2-8 — X 19-65 8-7 10-9 2-65 I 1-14 4-25 1-45 2-4 3-7 6-2 XI 22-9 9-S 13-5 2-37 I 1-12 3-48 1-4 2-65 4-2 8-4 Measurements in millimetres except for antennule -where proportional lengths of segments are given. Table IX. List of stations at which Phyllosomas of Jasus lalandii were taken Depth metres Number of Size Station Position specimens mm. Puerukis 268 i8°37'S, io°46'E 73-0 I 15 — 80 32°46'S, io°oo' W 30-0 2 11,15 — 81 32°45'S, 8°47'W 650-0 2 15.17 — ■ 82 32°42'S, 2°o5'W 75-0 3 15-17 — 83 32°3o'S, i°23'W 650-0 2 17,21 — 264 33° 06' S, 16° 55' E 80-0 I 15 — 85 33°07'S, 4°3o'E 2000-0 21 10-23 — 86 33°2S'S, 6°3i'E I 000-0 I 14 — 413 33° 13' S, 15° 46' £ 350-0 3 27-34 — 100 33° 33' S, 15° 13' E 250 2 37 3 — — 900-1000 I 36 — — — 35o-4oo(-o) II 28-36 — lOI 34° 01' 8,15° 56' E 850-950 3 32-37 — — — 1310-1410 I 35 — — — 2480-2580 3 35-37 — 102 35° 29' S, 18° 33' E 52 I — — 257 35°oi'S, 10° 18' E 250-0 3 10-20 — 258 35° 03' S, 13° 55' E 320-450 I 16 — 254 35°o4'S, 2°59'E 200-0 I II — 247 37° 20' S, 12° 47' W ioo-ii5(-o) I 31 ■ — 407 35° 08' S, 17° 49' E 150-0 I 34 — — — 275-0 I 34 — — — 800-950 I 33 — 1377 38°36'S, 9°03'E 100 3 17-22 ' — 448 39° 03' S, 16° 11' E 161-0 I 27 LARVAE OF DECAPOD CRUSTACEA 425 Most of the samples in which these Phyllosomas occurred were taken in deep water, but as the hauls were generally continued to the surface, they offer little evidence concerning vertical distribution. A few of the hauls (Sts. 100, loi, 102, 1377) prove, however, that the larvae do reach very considerable depths, even to 2480-2580 m. The specimens from deep water are nearly all in the last stage. Three specimens of the Puerulus stage were caught at 250 m. at St. 100, about 170 miles from land and in an ocean depth of over 3000 m. Fig. 30. Palinurellus'i , 4.7 mm. St. 1580. Genus Palinurellus, Von Martens.? (Figs. 30, 31) St. 1580. 8° 44' S, 41° 50' E. 450-0 m. One specimen. Length 47 mm. Greatest width 3-3 mm. Eye 1-3 mm. ; eye-stalk 075 mm. Fig. 31. Palinwellus}, \"] mm. Dorsal view of head region. 426 DISCOVERY REPORTS Fore-body about as wide as long, with a small but perfectly distinct sharp-pointed rostrum (Fig. 31). Hind-body much narrower than fore-body, not hollowed behind. Abdomen small, distinctly segmented, with bilobed uropods but no pleopods. Antennules with peduncle of two segments, endopod distinct. Antennae with basal part distinct from flagellum and produced outwards into a strong spine. Flagellum nearly reaching end of antennule. Maxillule with small papilliform palp. Maxilla with exopod large, extending behind maxillipede 2, and fringed with setae. Maxillipede i a small papilla. Maxillipede 2 with exopod of fair size but without setae. Maxillipede 3 having large setose exopod. Dactyli of legs short and curved ; each leg with coxal spine and a group of spines at end of merus. Leg 5 about twice as long as abdomen, segmented, with large rudiment of exopod. It is unfortunate that only one specimen, and that a young one, of this Phyllosoma is available, since it is a very remarkable form and the last stage might give some clue to its identity. It is unique in having a well-marked rostrum, and the presence of an exopod on leg 5 is another character which is unusual, and perhaps primitive. The form of the maxillae, presence of exopods on maxillipedes 2 and 3, and presence of a palp on the maxillule, point with certainty to the Palinuridae, and probably to near relationship to Palimirus ; but it obviously does not belong to Palinurus nor Pamdirus. There is a general resemblance to a Phyllosoma described by Santucci (1929) from the Red Sea, and regarded by him as probably belonging to Scyllarides latus, but Santucci does not show a palp on the maxillule and does not mention a rostrum. It is fair to assume that the adult must be a genus possessing a rostrum, ^ which excludes Linupanis and, perhaps, Paliniistiis. There remain only Polinurellus and Puerulus. There is no means whatever for deciding between these two genera, but Palimirellus, with its rela- tively large rostrum and pleopod on somite i , seems to be a primitive genus, and the species P. gmidlachi is recorded from Mauritius and may well occur farther west. I regard it, therefore, as a justifiable speculation to attach this generic name to this Phyllosoma. Genus Scyllarus, Fabricius (Figs. 32-34) For the genus Scyllarus the general characters of the Phyllosoma were established by Dohrn, who described in 1870 the first stage of S. arctus hatched from the egg, and by Hornell (1894), while the complete series of nine stages has been worked out by Stephensen (1923). Stephensen's account is very much more precise and convincing than that of Bouvier or Santucci for Palinurus. The general characters of the genus, as founded on the description of Scyllarus arctus, are as follows : 1 Gilchrist (1916, p. 112) describes a rostrum in the Phyllosoma of Jasus, but it is very doubtful if the structure described can be so interpreted (see above, p. 422). LARVAE OF DECAPOD CRUSTACEA 427 Fore-body much wider than long, and much wider than hind-body. Hind-body not concave behind. Abdomen, in later stages, very broad at base, and forming a direct continuation of the hind-body. Antenna at first very much shorter than antennule, later becoming broad at base, and with large outer pointed process ; flagellum short and broad. Maxillule without palp. Exopod of maxilla with setae in stage i, but without setae in late stages. Maxillipede 2 without exopod, or with exopod rudimentary. Maxillipede 3 without exopod. Leg 2 dactylus not greatly elongated. Leg 5 without exopod. Legs with coxal spines. Pleopods in last stage very narrow, without appendix interna. Telson with large spine on either side. I have not found it possible to make any specific distinctions in the small material available, though in some cases the size of corresponding stages is so different that more than one species is certainly present. Specimens from the eastern Atlantic agree in all respects with Stephensen's description of S. arctiis, and may well belong to that species, though, so far as I know, the adult has not been recorded farther south than Senegal. The only point of interest that arises from this material is the strong evidence that, within the genus Scyllanis, specific differences in the larvae are very small or non- existent. The accompanying figures will show the close agreement between East African species of very small form (Fig. 33) and S. arctiis as figured by Stephensen. Fig. 32 shows what is probably S. arctus in stage VIL Table X. List of stations at which Phyllosomas of Scyllarus were taken Station Position Depth Number of Size metres specimens mm. — 29° 27' N, 15° 7' W 900-0 3 12-16 — i3°2s'N, i8°22'W 900-0 2 12, 15 281 o°46'S, 5°49'E 850-95o(-o) I 12 277 i°44'S, 8°38'E 63-0 I II 276 5°54'S, ii°i9'E 150-0 I 10 270 I3°58'S, ii°43'E 200-0 2 21,27 269 i5°S5'S, io°35'E 6oo-7oo(-o) I 28 711 24°4i'S, 4i°3i'W 290-0 2 13.18 712 28° 02' S, 43° 09' W 224-0 I 5-5 407 35° OS'S, i7°49'E 150-0 I 18 1568 34° 48' S, 34° 28' E 1400-0 I 17-8 1571 27°24'S, 39°2i'E 500-0 I 17 1580 8°45'S, 4i°5o'E 450-0 I 5 1581 7°42'S, 44°i4'E 600-0 I 29 1582 5° 39' S, 46° 22' E i9oo-i85o(-o) I i8-5 1585 o°o6'S, 49°4S'E 500-0 2 6, 22-5 Genus Scyllarides, Gill (Figs. 35-37) The first Phyllosoma of Scyllarides lattis, taken from among the pleopods of the parent, has been described by Santucci (1928). Later larvae have been referred to the genus by 7-2 428 DISCOVERY REPORTS Stephensen (1923), Santucci (1925 b, 1929) and Von Bonde (1932). One of the larvae described by Santucci (1927) as probably belonging to Thenus (pi. iii, fig. i) is actually, Fig. 32. Scvl/arus arctus?, Stage VII. St. 281. as I believe, ScyUandes, while the later stage figured (pi. iii, figs. 2-4) represents another genus, possibly Themis. Santucci (1929) ascribed a Phyllosoma in stage III from the Red Sea to his Thenus series, and identified it as the larva of Scyllarides laUis. While his Thenus was probably, in part, Scyllarides, it is not possible to accept the Red Sea Fig. 33. Scyllarus sp., 5 mm. St. 1580. Stage V 05 mm. St. 1585 Stage VI 13 mm. St. 711 Stage VIII 17-8 mm. St. 1568 Fig. 34. Scyllarus, Stages V, VI, VIII, IX. a. Antennae. b. Maxilla, etc. c. Abdomen. Stage IX 27 mm. St. 270 430 DISCOVERY REPORTS larva as belonging to the same series, or even as a Scyllarid at all. One character alone completely excludes it, namely the possession at this early stage of exopods on maxilli- pedes 2 and 3, for in Scyllarus and Scyllondes the exopod is absent throughout develop- ment from maxillipede 3, and it is absent in the larva of Themis figured by Santucci. Von Bonde (1932) has attributed to Scyllarides eUzabethae a series of larvae taken on the south and east coasts of South Africa, mainly on the ground that this is the only common species in those waters. He distinguishes a series of eleven stages, but the differences in size and structure between some of them is very small indeed. The specimen of 1 1 mm. (his stage VIII) appears to be the last stage, having the gills developed, and very cha- racteristic pointed uropods, and it is very unlikely that there should be three more stages during which the uropods change their shape to the rounded form shown in other figures. It seems much more probable that his material included larvae of more than one species, and that one of them (with rounded uropods, pi. vii) belonged to Scyl- larus. It is to be noted also that the telson spines of his younger larvae are relatively short, as in Scyllarus. Stage I of S. latus is characterized by having the antenna about as long as the an- tennule, and the great length of leg 3, but it is doubtful if these characters are suffi- cient to distinguish the genus from Scyl- larus. APhyllosoma in stage I from Bermuda (Fig. 35) probably belongs to Scyllarides since that genus alone is known from Ber- muda, but its antennae are even shorter than those of Scyllarus arctus, and leg 3 is not strikingly longer than leg 2. Stephensen's larva, which is probably rightly referred to S. latus, is distinguished by the great length of the spines on the telson, and this may be a good character for the genus. If that is so Phyllosoma fiircicaudatum, Bate (1888, pi. xii, D), from St Vincent, no doubt belongs to this genus. It is probably safe to accept long telson spines in early stages, and pointed uropods in later ones, as distinctive for the genus ; but the length of the antenna does not seern to be a reliable character, and in intermediate stages, when the rudimentary uropods are rounded, the two genera may be indistinguishable. Fig. 35. Scyllarides sp., Stage I. Bermuda. LARVAE OF DECAPOD CRUSTACEA Table XI. List of stations at zohich Phyllosomas of Scyllarides icere taken Station Position Depth Number of Size metres specimens mm. 701 276 14° 39' N, 25° 51' W 242-0 5 4 7-14 14 13-8 13-5 13 5°54'S, ii°i9'E 150-0 I 407 35°o5'S, i7°49'E 150-0 3 ^S5S 39°52'S, i8°42'E I 000-0 I 1574 2i°4s'S, 40°34'E 600-0 I 1373 1576 3i°i3'S, 3i°49'E I4°42'S, 42°22'E 135-0 400-0 I I 431 Fig. 36. Scyllarides s^., b mm. St. 701. I 432 DISCOVERY REPORTS I refer to Scyllarides nine specimens from the Atlantic, as shown in Table XI, taken at three stations in the region of Cape Verde Islands, off the west African coast, and near the Cape. The specimens from St. 407 agree so closely with Von Bonde's description that I have little doubt that they belong to S. eUsabethae, but those from Sts. 276 and 701 are larger, and may perhaps belong to S. latiis. Four specimens taken on the east African coast may also be larvae of S. elizabethae. Fig. 37. Scyllarides sp., 7 mm. St. 407. Genus Thenus, Leach.? (Figs. 38-41) General characters of the Phyllosoma. Fore-body pear-shaped, narrow in front in early stages, about i\ times as wide as long; wider than hind-body. Hind-body deeply concave behind. Abdo- men small and narrow. I LARVAE OF DECAPOD CRUSTACEA 433 Antennule with segments 2 and 3 equal, not much shorter than segment i. Antenna short and stout, with strong pointed process on outer side of segment 2 of peduncle, segment 3 very much narrower than segment 2 in late stages. Maxillule without palp. Maxilla without setae on exopod. Maxillipedes 2 and 3 without functional exopods. Legs without coxal spines. Leg 3 with propodus dilated at end. Leg 5 without exopod, but with small rudiment of it in late stages. Fig. 38. Themis sp.?, Stage IV?, 8 mm. St. 691. The fifteen specimens of this form are divisible into five stages, the oldest of which is in the last stage but one. The growth factor for these stages is about i -26 for the oldest, and increases to about 1-5 for the youngest. If this rate of growth is assumed for the unknown younger stages, and if it is assumed that the first larva is about 2-5 mm. long, there will then be three unknown stages, and the youngest of the Discovery specimens will be in stage IV. The stages may be defined as follows : 434 DISCOVERY REPORTS Stage IV? (Fig. 38). Length 8 mm. (pre-labral 4-4 mm. ; post-labral 3-6 mm.): eye and eye-stalk 3-5 mm. Antennule with peduncle of two segments. Antenna very short, bifurcate at end. Maxilla with minute, narrow, apical segment, with or without setae. Of three ap- parently identical specimens two were without setae and one with four setae. Maxilli- pede I absent. Maxillipedes 2 and 3 without exopod. Legs without coxal spines. Leg 5 a small bud. Abdomen small, narrow, unsegmented, without pleopods or uropods. Stage V.^ Length 12-14 mm. Antennule with peduncle of three segments. Antenna a little longer than segment i of antennule, unsegmented or with basal segment marked off; with outer pointed process at end of this basal segment. Maxilla with apical part not expanded, without setae. Leg 5 and abdomen as in stage VI, but uropods indicated by a pair of small folds. Stage VI }. Length 20 mm. Antenna reaching to segment 2 of peduncle of antennule, its peduncle shorter than the flagellum. Maxilla not expanded. Abdomen still small unsegmented. Pleopods represented by small folds; uropods bilobed rudiments. Stage VII ? (Fig. 39). Length 27-28 mm. Ratio of length to width of fore-body i : 1-69; of eye to total length 1:3-4. Lengths of segments of antennular peduncle as 26 : 23 : 23. Antenna with peduncle of three distinct segments, the outer pointed process being on segment 2. Segment 3 narrow and appearing to form part of the flagellum which reaches beyond the peduncle of the antennule. Maxilla without setae, but with apical part widening. Leg 5 with basal segment distinct. Abdomen showing segmentation dorsally. Pleopods small, not bilobed; uropods shorter than telson, the endopod not jointed to stem. Stage VIII ? (Figs. 40, 41). Length 33-36 mm. Ratio of eye to body length i : 3-15. Segment 2 of antenna very much wider than segment 3. Maxilla without setae, widening at end. Maxillipede i a finger-like process, shorter than maxilla. Maxillipedes 2 and 3 with small papillae representing exopod. Leg 5 as long as abdomen, of four segments ; segment 2 with a small bud, perhaps representing exopod. Pleopods bilobed ; uropods longer than telson, the branches jointed to basis. Gills absent. Stage IX? (Fig. 41). Length 41 mm. (British Museum specimen from Congo Expedition). Ratio of eye to body length 1:3-4. LARVAE OF DECAPOD CRUSTACEA 435 Antenna with segment 2 greatly expanded, flagellum extending beyond antennule. Maxilla with exopod expanded and reaching behind maxillipede i, but without setae. Maxillipede i with epipod. Maxillipedes 2 and 3 with rudimentary exopods, without setae. Legs without coxal spines. Leg 5 with small bud representing exopod. Gills present. Pleopods and uropods large ; no spines dorsally on abdominal somite 6. It is naturally impossible to identify this Phyllosoma; but it is evidently closely related to a form described by Stephensen (1923, p. 77) and referred by him to Themis orientalis. The Discovery specimens do not seem to belong to the same species, as Stephensen's figures show strong coxal spines on legs 1-3. Santucci (1926) has de- scribed a larva of 30 mm., also from the Mediterranean, which appears to belong to the same species as my specimens, and this larva he supposes also to belong to Themis orientalis. There are great difficulties in accepting such an identification. In the first place T. oricjitalis is the only species of the genus known, and it is confined to the Indo-Paci- fic region, with the exception of a doubtful record from the Adriatic, and one from Natal (Von Bonde). It seems unlikely that it would have been overlooked in the Atlantic if it is as common as these larvae would show it to be. It is also evident that there must be two distinct species in the Mediterranean of the genus to which these larvae belong. At the same time, our knowledge of the distribution of adult Decapoda is not so exhaustive that such an argument can have much weight. If we knew only the Phyllosomas of Pali- nurus and ScyUarus this Phyllosoma would certainly be regarded as a Palinurid; but it appears that the differences between these two genera cannot be taken as valid for their families. We have already seen that the absence of an exopod from the maxillipedes is not confined to the Scyllaridae, and, if the Phyllo- soma described below as Parribaciis} is in fact a Scyllarid, then a "Panulirid" form of body may also be found in the Scyllaridae. The only character which seems to hold good for separation of the families is the form of the antenna in the last stages, and in the Phyllosoma in question there is a tendency to broadening of the base which indicates a possible relation to the Scyllaridae. For these reasons I accept provisionally the reference of this form to Themis by Stephensen and Santucci, but without conviction of its soundness. 8-2 Fig. 39. Themis sp. ?, Stage VII ?. St. 704. a. Antennae. b. Maxilla, etc. c. Abdomen. (/. End of leg 4. 436 DISCOVERY REPORTS Table XII. List of stations at which Phyllosomas o/Thenus? were taken Depth metres Number of Size Station Position specimens mm. Discovery collection 293 4°i8'N, i6°5i'W ioo-i2o(-o) I 27 704 3° 38' N, 29° 14' W 231-0 I 28 692 2° 02' N, 30° 08' W 350-0 2 13. H 691 o°25'S, 29°56' W 400-0 9 8-33 690 3° 18' S, 3o°oo'W 460-0 2 10,33 British Museum collection — 36° S, 9° E — I 32 — 13° S, 25° w — I 36 — Atlantic ■ — 2 38,43 — Africa: Congo Exped. — I 41 Other localities : Mediterranean (Stephensen, Santucci). Fig. 40. r/(e««^ sp.?, Stage VIII ? 33 mm. St. 691. LARVAE OF DECAPOD CRUSTACEA 437 Genus Parribacus Dana? (Fig. 42) General characters of the Phyllosoma. Fore-body pear-shaped, narrow in front, narrower than hind-body. Hind-body very deeply hollowed behind. Abdomen small and narrow, completely sunk in hollow of hind-body. Segment 2 of antennular peduncle slightly longer than segment 3. Antenna small, basal part widened in late stages. Stage VIII? Stage IX? Fig. 41. Thetius sp.?, Stages VIII-IX?. a. Antennae. b. Maxilla, etc. c. End of leg i, Stage IV?. Maxillule without palp. Maxilla without setae, not widened in any known speci- mens. Maxillipedes 2 and 3 without exopod. Leg 5 fully developed, with setose exopod (in all known specimens), and seated at a distance from abdomen. Pleopods and uropods greatly delayed in appearance. Only three specimens of this Phyllosoma were taken by the 'Discovery II', at the following stations : St. 691. 0° 25' S, 29° 56' V/. One, 15 mm. St. 1576. 14° 42' S, 42" z2' E. One, 20 mm. St. 1582. 5° 39' S, 46" 22' E. One, 37 mm. 438 DISCOVERY REPORTS The specimens from the east coast of Africa are not distinguishable from the Atlantic specimen. Although the oldest of these three specimens is more than twice the size of the smallest one, it diflFers very little in degree of development. The antenna shows a division into flagellum and peduncle, and the latter is rather stout, with a small outer pro- jection (Fig. 42 b). The maxilla is not expanded, and without setae, and max- illipede i is a small papilla. In the speci- men of 20 mm. this appendage is not traceable at all. The abdomen in the two smaller specimens is unsegmented, with- out trace of pleopods, and the uropods are represented in the larger specimen by small folds. In the specimen of 37 mm. the abdomen is still not much more de- veloped, but the pleopods are represented by small simple lobes and the uropods are small and bilobed. This form of Phyllosoma is very re- markable by reason of the large size reached without much progress in devel- opment of antennae, mouth parts and abdomen, while leg 5, which is usually late in developing, is fully formed and provided with an exopod in the smallest specimens known. Phyllosomas of precisely similar form have been described by Guerin, Richters and Spence Bate. Spence Bate's speci- men, of 30 mm., was taken in the West Indies and may well belong to the same species as the one from St. 691. Phyllo- soma guerini, de Haan, also has an exopod on leg 5, but is otherwise very different. A Phyllosoma from China described by Richters (1873, pi. 34, fig. i) seems to belong to the same genus as P. guerini, and there is reason to believe that this genus may be Ibactis (Balss, 1914, p. 81). Richters included the type under discussion in the "Phylicsomes brevicaudes" of Milne-Edwards, and refers them to the Scyllaridae. The general form of the body is so 4mm Fig. 42. Parribacus sp.? a. Specimen of 15 mm. St. 691. b. Antennae of specimen of 37 mm. St. 1582. LARVAE OF DECAPOD CRUSTACEA 439 Strikingly similar to that of Pamdirus that it is difficult to believe that this identification can be correct. The Phyllosomas referred to this group are mainly Indo-Pacific and attain to such extraordinary sizes as 75 mm. (Richters). Having regard to the degree of development of antennae and mouth-parts in larvae of nearly 40 mm. it is probable that there is a long series of later stages during which much change of shape of body and of appendages may take place, and the widening of the antenna seen in the specimen of 37 mm. might lead to a Scyllarid form. Richters figures (pi. xxxiv, figs. 5, 6) two larvae of 61 and 75 mm. which combine the Scyllarid form of antenna with the deeply excavated hind-body of Pamdirus, so that it is clear that the latter character is not confined to the Palinuridae and it seems necessary to accept Richters' reference of this group of Phyllo- somas to the Scyllaridae. So far as concerns the genus to which the Discovery specimens belong only specula- tion is possible, but it must be one with a representative in the western Atlantic. The genus Themis must be excluded if its Phyllosoma is as described above, and of the other genera of Scyllaridae only two species are known from the Atlantic : Ibacus verdi. Bate, and Parribociis ursus (Herbst). The genus Pseudibacus, Guerin, is supposed by Bouvier to be the natant stage of Scyllariis. As Parribacus ursus is known to occur in the Carib- bean Sea and also throughout the Indo-Pacific region (De Man) it is a reasonable guess that this may be the parent of the Phyllosoma in question. KEY TO THE FORMS OF PHYLLOSOMA HERE DESCRIBED a. Maxillipede 3 with setose exopod. i. A small rostrum present Palimrelliis}, p. ^25. bb. Rostrum absent. c. Hind-body not wider than fore-body; lengths of segments of antennular peduncle about as 3 : I : 2 Palinurus, p. 401. cc. Hind-body wider than fore-body ; antennular segments about as 2 : i : i Pamdirus, p. 405. d. Coxal spines absent ; exopod of maxilla without setae to last stage Pamdirus B, p. 411. dd. Coxal spines on maxillipede 3 and leg i ; maxilla with setae Panulirus A, p. 407. ddd. Coxal spines large, on all legs; maxilla with setae ... Panulirus D, p. 415. aa. Maxillipede 3 without exopod, or with small rudiment only. b. Antenna slender, with long flagellum; leg 5 fully developed at early stage Jasus, p. 416. bb. Antenna enlarged at base, flagellum short and broad; leg 5 generally developing late c. Leg 5 with setose exopod Parribacus}, p. ^^l- cc. Leg 5 without setose exopod. d. Abdomen much narrower than hind-body ... ... ... Themis'^, p. 432. dd. Abdomen broad at base, forming continuation of hind-body. e. Telson spines relatively short ; uropods rounded throughout development Scyllarus, p. 426. ee. Telson spines long in early stages; uropods pointed in late stages Scyllarides, p. 427. 440 DISCOVERY REPORTS LITERATURE REFERRED TO Anderton, J., 1907. Observations on New Zealand Fishes. Trans. N.Z. Inst., xxxix, p. 477. Archey, G., 1916. Notes on the marine Crayfish of New Zealand. Trans. N.Z. Inst., XLViii, p. 396. Balss, H., 1914. Ostasiatische Decapoden. II. Die Natantiatind Reptantia. Abh. math. phys. Kl. K.Bayer. Akad. Wiss., Suppl. Bd. 11, Abh. i. Bate, C. S., i888. Crustacea Macrura. Voyage H.M.S. 'Challenger', Zool., xxiv. BoirviER, E. L., 1913. Recherches sur le developpement post-embryonnaire de la Langouste commun, Palinurus vulgaris. J. Mar. Biol. Ass. (n.s.) x, p. 179. Crawford, D. R. and Smidt, W. J., 1923. The Spiny Lobster, Panulirus argus, of southern Florida; its natural history and utilisation. Bull. U.S. Bur. Fish., xxxviii, p. 281. De Man, J. G., 1916. Decapoda of the Siboga Expedition. Eryonidae, Palinuridae, Scyllaridae and Nephrop- sidae. Siboga Exp., xxxix, a 2. Gilchrist, J. D. F., 1913. A free-szvimming Nauplioid Stage in Palinurus. J. Linn. Soc. Lond., xxxii, p. 225. 1916. Larval and post-larval stages of Jasus lalandii (M. Edw.). J. Linn. Soc. Lond., xxxiv, p. 189. HoRNELL, J., 1894. The Phyllosoma or glass crab larva o/Scyllarus. J. Mar. Zool. Jersey, i, p. 69. Nakazawa, K., 1917. On the metamorphosis 0/ Panulirus japonicus. Dobuts Z. Tokyo, xxix, p. 259. RiCHTERS, F., 1873. Die Phvllosomen. Ein Beitrag zur Entwicklungsgeschichte der Loricaten. Z. Wiss. Zool., xxiii, p. 623. Santucci, R., 1925 A. Contributo alio studio dello sviluppo post-embrionale degli Scyllaridea del Mediterraneo. Atti R. Accad. Lincei (6), i, p. 333. 1925 B. Contributo alio studio dello sviluppo post-embrionale degli Scyllaridea del Mediterraneo. II. Scyllarus arctus. III. Scyllarides latus. R. Comit. Talassog. Ital., Mem. cxxi. 1926. Fillosomi di Scillaridi esotici nel Mediterraneo. Mon. Zool. Ital., xxvii, p. 19. 1927. Fillosomi del Mar Rosso con osservazione geografica. Atti 1st. Idrog. Reg. Marina, xi, bis. 1928. II prima stadio post-embrionale di Scyllarides latus. R. Comit. Talassog. Ital., Mem. cxliv. 1929. Ulteriore notisie sulla presenza di fillosomi di Scillaridi esotici nel Mediterraneo. R. Comit. Talassog. Ital., Mem. CLX. Stephensen, K., 1923. Decapoda Macrura. Rep. Danish Oceanog. Expeds. 1908-10 to the Mediterranean and adjacent seas. Thomson, G. M., 1907. Note on the development 0/ Palinurus edwardsi. Trans. N.Z. Inst., xxxix, p. 484. Verrill, a. E., 1922. Decapod Crustacea of Bermuda. II. Macrura. Trans. Conn. Acad., xxvi, p. i . Von Bonde, C, 1932. Post-brephalus development of some South African Macrura. Fish, and Mar. Surv. Union of S. Africa, Rep. 8. Von Bonde, C. and Marchand, J. M., 1935. The natural history and utilisation of the Cape Crawfish, Kreef or Spring Lobster Jasus (Palinurus) lalandii (M. Edw.). Fishery Bull. I. Fisheries and Marine Survey Division, Union of South Africa. DISCOVERY REPORTS Issued by the Discovery Committee, Colonial Ojfice, London on behalf of the Qovernment of the Dependencies of the Falkland Islands Vol. XII, pp. i-viii TITLE-PAGE, LIST OF CONTENTS AND LIST OF PERSONNEL CAMBRIDGE AT THE UNIVERSITY PRESS 1936 Price one shilling net Cambridge University Press Fetter Lane, London New York Bombay, Calcutta, Madras Toronto Macmillan Tokyo Maruzen Company, Ltd All rights reserved PRINTED IN GREAT BRITAIN BY WALTER 'K|?ra/ LEWIS MA AT THE CAMBRIDGE UNIVERSITY PRESS DISCOVERY REPORTS Vol. XII, pp. 1-58 Issued by the Discovery Committee, Colonial Office, London on behalf of the Qovernment of the Dependencies of the Falkland Islands COAST FISHES PART I. THE SOUTH ATLANTIC by J. R. Norman MCDL_E, CAMBRIDGE AT THE UNIVERSITY PRESS 1935 Price five shillings net Cambridge University Press Fetter Lane^ London New York Bombay^ Calcutta, Madras Toronto Macmillan Tokyo Maruzen Company, Ltd All rights reserved PRINTED I N GREAT BRITAIN BY LEWIS MA AT THE CAMBRIDGE UNIVERSITY PRESS DISCOVERY REPORTS Vol. XII, pp. 59-198 Issued by the Discovery Committee, Colonial Office, London on hehalfofthe Qovemment of the Dependencies of the Falkland Islands POLYCHAETE WORMS, II C. C. A. Monro, M.A. CAMBRIDGE AT THE UNIVERSITY PRESS 1936 Price eighteen shillings net Cambridge University Press Fetter Lane, London New York Bombay, Calcutta, Madras Toronto Macmillan Tokyo Maruzen Company, Ltd All rights reserved PRINTED IN GREAT BRITAIN BY WALTER LEWIS MA AT THE CAMBRIDGE UNIVERSITY PRESS DISCOVERY • REPORTS Vol. XII, pp. 199-348, plates 1-IX Issued by the Discovery Committee, Colonial Office, London on behalf of the Qovemment of the Dependencies of the Falkland Islands ECHINOIDEA AND OPHIUROIDEA Th. Mortensen CAMBRIDGE AT THE UNIVERSITY PRESS 1936 Price twenty-seven shillings and sixpence net Cambridge University Press Fetter Lane, London New York Bombay, Calcutta, Madras Toronto Macmilian Tokyo Maruzen Company, Ltd All rights reserved PRINTED 1 N GREAT BRITAIN BY LEWIS MA AT THE CAMBRIDGE UNIVERSITY PRESS :■ or, yy DISCOVERY REPORTS Vol. XII, pp. 349-376, plates X-XII Issued by the Discovery Committee, Colonial Office, London on behalf of the Qovernment of the Dependencies of the Falkland Islands THE BIRDS OF THE SOUTH ORKNEY ISLANDS R. A. B. Ardley, R.N.R. CAMBRIDGE AT THE UNIVERSITY PRESS 1936 Price six shillings net Cambridge University Press Fetter Lane, London New York Bombay, Calcutta, Madras Toronto Macmillan Tokyo Matuzen Company, Ltd AD rights reserved PRINTED IN GREAT BRITAIN BY LEWIS MA AT THE CAMBRIDGE UNIVERSITY PRESS DISCOVERY REPORTS Vol. XII, pp. 377-440 Issued by the Discovery Committee, Colonial Office, London on behalf of the Qovemment of the Dependencies of the Falkland Islands LARVAE OF DECAPOD CRUSTACEA PART I. STENOPIDEA PART 11. AMPHIONIDAE PART III. PHYLLOSOMA Robert Gurney, D.Sc. CAMBRIDGE AT THE UNIVERSITY PRESS 1936 Price nine shillings net Cambridge University Press Fetter Lane, London New York Bombay, Calcutta, Madras Toronto Macmillan Tokyo Maruzen Company, Ltd All rights reserved PRI NTED I N GREAT BRITAIN BY WALTER LEWIS MA AT THE CAMBRIDGE UNIVERSITY PRESS ^■■^■■■■arf r-^, • 'i- ■^' 4 .V>. *• ft^' •f*^ "^mMm. yM% ''f'S'^y^ W^^Ali^? /■l ' A ^ Vfft^c. :^ /. y^A. \ivA^ >;. ;;^4 '^ jft^-t mi ';*fcy.;; '■>\f\* % » ■ ■« 1 -^,