I 1 CT> ■ S C£> ■ ^ cn ^ ~ :~~ I — s of <-> Vf r- ■■ i3 ■fai r^ oo 76 1% Aberdeen University Studies : No. 46 Zoological Studies (Fifth Series) University of Aberdeen. COMMITTEE ON PUBLICATIONS. Convener : Professor James W. H. Trail. F.R.S., Curator of the Library. UNIVERSITY STUDIES, General Editor: P. J. Anderson, LL.B.. Librarian to the University. ^RoU of Alumniin Arts of King's College, 1596-1860. P.J.Anderson. —Records of Old Aberdeen^ 1157-1891. A. M. Munro, F.S.A. Scot. Vol. I. —Place Names of IVest Aberdeenshire. James Macdonald, F.S.A. Scot. —Family of Burnett of Leys. George Burnett. LL.D., Lyon King of Arms. —Records of Invercaulii, 1547-1828. Rev. J. G. Mtchie. M.A. —Rectorial Addresses in the Universities of Aberdeen, 1835-1900. P. J. .Anderson. —Albemarle Papers, 1746-48, Professor C. S. Terry, M.A. —House of Gordon. J. M. BuHoch, M.A. Vol. I. —Records of Elgin. William Craraond, LL.D. Vol. L —Avogadro and Dalton. A, N. Meldrum, D.Sc. —Records of the Sheriff Court of Aberdeenshire. David Littlejohn, LL.D. Vol. I. —Proceedings of the Atuitomical and Anthropological Society, 1902-04. — Report on Alcyouaria. Professor J. Arthur Thomson, M.A., and others. .—Researches in Organic Chemistry. Prof. F. R. Japp, F.R.S., and others. —Meminisse Jitvat : with Appendix of Alakeia. Alexander Shewan, M.A. —Blackhalls of that Ilk and Barra. Alexander Morison, M.D. — Records of the Scots Colleges. Vol. I. P. J. Anderson. —Roll 0/ the Graduates, 1860-1900. Colonel William Johnston, C.B., LL.D. — Studies in the History of the University. P. J. .\aderson and others. Studies in the History and Art 0/ the Eastern Provinces of the Roman Empire. Professor Sir W. M. Ramsay, D.C.L.. and pupils. — Studies in Pathology. William BuUoch, M.D., and others. , — Pioceedings of the Anatomical atid Anthropological Society, 1904-06. . — Subject Catalogues of the Science Library and the Law Library. P. J. Anderson. —Records of the Sheriff Court of Aberdeenshire. David Littlejohn, LL.D. Vol. II. — Studies on Alcyonarians and Antipatharians. Prof. Thomson, M.A., and others. — Surgical Instruments in Greek aiui Roman Times. J. S. Milne, M.A., M.D. —Records of the Sheriff Court of Aberdeenshire. David Littlejohn. LL.D. Vol. IIL , — Flosculi Graeti Boreales. Ser. IL Professor John Harrower, M.A. . — Record of the Quaiercentemiry, 1906. P. J. .\nderson. —House of Gordon. J. M. Bulloch. M.A. Vol. U. — Miscellany of the New Spaldmg Club. Vol. IL —Religious Teachers of Greece. James Adam, Litt.D. (Gifford Lectures, 1904-06.) — Science and Philosophy of the Organism. Hans Driesch, Ph.D. (GIff. Lect., 1907.) — Proceedings of the Anatomical arut Anthropological Society, 1906-08. -Records of Elgin. Vol. IL Rev. S. Ree. B.D. — Pigmentation Survey of School Children. J. F. Tocher, B.Sc. — Science and Philosophy of Organism. Hans Driesch. Ph.D. Vol. IL (Giff. Lect.. 190S. —Studies on A Icyonariam and Hydroids. Prof. Thomson, M.A., and others. Ser. Ill .—Publications of Scottish Clubs. Professor C. S. Terry, M.A. —Aberdeen Friars : Red, Black, White, Grey. P. J, Anderson. —Studies on A Icyonariam. Professor Thomson. M.A., and others. Ser. IV. —Records of Old Aberdeen. A. M. Munro. Vol. II. — .Musa Latimi A berdonenses : Poetae Minores. W. K. Leask. M.A. —Bulletins of the .Aberdeen and North of Scot I ami College of Agriculture. Nos. 1-14. —Records of Inverness. Vol. I. W. Mackay and H. C. Boyd. —Zoological Sttulies. Professor Thomson and others. Ser. V. —Subject Catalogue of the Phillips Library. 1900. No. I.- „ No. 2.- „ No. 3.- I90I. No. 4-- ,, No. 5-- 1902. No. 6.- No. 7-- 1903. No. 8.- „ No. 9.- 1904. No. 10.- „ No. II.. ,, No. 12.- 1905. No. 13.. ,, No. 14. „ No. 15.- „ No. i5.- 1906. No. 17.. „ No. i8.. ,, No. 19.. •• No. 20.- „ No. 21.- ,, No. 22. „ No. 23. „ No. 24. 1907. No. 25 „ No. 26.. „ No. 27.- ,, No. 28. „ No. 29.' ,, No. 30. :go8. No. 31. „ No. 32. „ No. 33. „ No. 34.- „ No. 35.. ., No. 36. 1909. No. 37.- No. 3S.- No. 39. „ No. 40.' „ No. 41. No. 42. 1910. No. 43.- „ No. 44- 1911. No. 45.- „ No. 46.- „ No. 47.- Zoolosfical Studies chiefly on Alcyonarians (Fifth Series) By J. Arthur Thomson, M.A. Professor of Natural History |. |. Simpson, M.A., B.Sc. Carnegie Fellow R. N, Rudmose Brown, D.Sc. Sophia L. M. Summers, M.A., B.Sc. George Crane, B.Sc. James Ritchie, M.A., B.Sc. ABERDKKN Printed for the University 191 1 CONTENTS. PAGES Note on a Hydracoralline from Rockald. By Professor Thomson ---.-- 1 On a New Pseudaxonid Genus — Bendrogorgia. By Mr. Simpson 3-8 ECHINOIDEA AND AsTEROIDEA FROM THE MeRGUI AeCHIPELAGO AND MOSKOS IsLANDS, LoWER Burma. By Dr. Rudmose Brown --- ........ 9-23 EcHINOIDEA FROM THE KeRIMBA ArCHIPELAGO, PORTUGUESE EaST AfRICA (MOZAMBIQUE). By Dr. Rudmose Brown ---.-........ 24-32 AsTEROiDEA OF PORTUGUESE East Africa, COLLECTED BY Jas. J. SiMPSON. By Mr. Simpson and Dr. Rudmose Brown 33-48 On a New Species of Cactogorg/a. By Mr. Simpson 49-51 Plate 51 Note on a Peculiar Clutch of Blackbird's Eggs and other Abnormalities. By Professor Thomson ....---.- [53]-[55] Note on Eunephthya Glomerata from the F.eroe Channel. By Professor Thomson [57]- [58] Note on Dendrobrachia fallax, a rare and remarkable antipatharian. By Professor Thomson 59-60 Antipatharians from the Indian Ocean. By Miss Summers ----- 61-69 Plate 61 Alcyonakians from the Gulf of Cutch. By Professor Thomson and Mr. Crane - - [71]-[75] The Marine Fauna of the Merqui Archipelago : The Hydroids. By Mr. Ritchie - 1-27 Two PLATEg 27 Hicksonella, a New Gorgonellid Genus. By Mr. Simpson 29-40 Plate - . - - 29 v. — Note on a Hydrocoralline from Rockall. By Professor J. Arthur Thomson, M.A. (Read 20th December 1909. Received 20th December 1909.) I RECEIVED this summer, from a trawler, three pieces of an interesting Hydrocoralline brought up near Eockall, — that lonely granite rock in the North Atlantic (57° 36' N. lat.: 13° 41' W. long.),— 184 miles west of St Kilda, 260 from the North of Ireland, 290 from the nearest part of the mainland of Scotland. As northern records of Hydrocorallinae are few and far between, it is of interest to register this one. The specimens are white flabellate colonies which agree with the description and figures of Sfylaster gemmascens (Esper) given by Professor P. i\Iartin Duncan {Tram. Zool. Soc, viii., 1874, p. 332, pi. i.\. 12 figs.). The diagnosis, quoted by Duncan from ]\[ilne-Edwards and Haime, reads : — "The corallum is subflabelli- form. The branches often coalesce, and the younger are crowded with small granulations, which are irregularly placed between the calyces. The old branches are almost smooth. The calyces are alternate on young branches, and sparingly developed on the old ; they are circular, oval, or deformed, and have projecting margins. There are from twelve to sixteen septa, which are often irregular." The surface of the colony shows the " cups " that are characteristic of Stylasterids. They occur all round the smaller branches, but are chiefly lateral on the larger. Each " cup " consists, as is well known, of the cavity of a nutritive polyp or gastrozooid, surrounded by a circle of twelve or so smaller cavities lodging the tactile dactylozooids. Each cup bears a deceptive resemblance to the calyx of a Madrepore, a resemblance heightened in some cases, notably in Aulojwra, by septa-like ridges extending inwards from the dactylozooids. It is historically interesting to notice that in Martin Duncan's memoir of 1874, where this form is beautifully figured, it is still misinterpreted as a Madreporarian, with wliich it has, of course, nothing whatever to do. To get a general picture of the nature of a Hydrocoralline colony, we have to imagine a much-branched hydrorhiza in which lime is secreted from the tubes instead of a perisarc, so that numerous fine canals are enclosed in a coherent calcareous framework. To this we have to add that the polyps are dimorpliic or triniorphic, — gastrozooids, dactylozooids, and sometimes medusoid reproductive buds. The type of Stylaster gemmascens came from the Indian Ocean The Nortli Atlantic forms identified with the type were dredged by the "Lightning" and the " Porcupine " (530 fathoms). Another record is given by Sars {Fork. Selskahs Ckristianid, 1872, p. 115), from a great depth in the Foldenfjord, Norway. (lasiud separately, Mth February 1910.) VI. — On a new Pseudaxonid Genus — Den drogorgla. By Jas. J. Simpson, M.A., B.Sc, Carnegie Research Fellow, University of Aberdeen. (Read 20th December 1909. Received 20th December 1909.) In 1900 Professor Hickson {Marine Investigations in South Africa — The Alcyonana and Hydrocorallincc, p. 85) described two specimens under the name of Jtincella elonr/ata (Pallas), with the following observations : — " Owing to the very imperfect state of our knowledge of the Juncella group of Alcyonarians, I have considerable hesitation in naming the two specimens of the genus sent to me from the Cape. . . . One of the most characteristic features of the specimen is the great preponderance of triple- star spicules 0'07 mm. in length, but there are also many spicules of the shapes known as double-stars, warted spindles, etc. There are very few spicules of the club-shape which are so characteristic of the species J.juncea and J. gemmacea." The specimens were dredged at Eij Bank, off Algoa Bay, long. 25° 51' 30" E., lat. 33° 58" S. Depth— 25 fathoms. Bottom— dark sand. In 1904 {Alcyonaria of the Cape of Good Hope, part ii., p. 233), while admitting that the triple-stars might be regarded as a character sufficient to distinguish these specimens as a distinct species. Professor Hickson refrained from doing so, but renamed them Juncella elongata (Pall.) var. capensis. Professor Hickson very kindly sent me a small poition of one of his specimens along with various Juncellids, and later, Professor Thomson placed a magnificent specimen, identical with the type, at my disposal. The latter specimen was also found at the Cape. Owing to the inadequate descriptions of Juncella elongata very diverse forms have been, from time to time, ascribed to it, but an investigation of the spicules hitherto undescribed, has shown that the specimens under considera- tion— Professor Hickson's and ours — cannot be referred to the genus Juncella. In fact the specimens do not belong to the Juncellids at all. Before entering into a discussion of the systematic position of our specimen, we shall give a short description supplementary to that which Professor Hickson gave of those specimens wliicli hv referred to Juncella elongata (Pallas) var. capensis, n. Dendrogorgia capensis, n. gen. et sp. A beautiful specimen of a deep-red colour (Fig. 1), massive in appearance, and branched openly in one plane in what is evidently a false dictotomy. The base is wanting. The total height of the colony is 20 cm. and the maximum breadth is 3'5 cm. The diameter of the main stem at the base is 9 mm. Near the base it gives rise to two primary branches with diameters of 8 mm. and 7 mm. respectively. The former di\'ides after a distance of lO'o cm., giving rise to a branch 6 mm. in diameter at its point of origin. The latter branches after a distance of 5 cm. ; the diameter of this secondary branch is 6 mm. Slightly beyond this point the main branch has been contorted and curves towards the secondary branch. All the branches taper very slightly ; one, however, maintains its original diameter throughout and terminates bluntly in a dome. The axis is 5'5 mm. in diameter at the base and tapers to a fine point in the branches ; it is comparatively soft and is easily cut with a knife. The horny part is spongy and the calcareous portion is composed of long smooth or slightly warty ii-regular spicules, quite unlike those of the ccenenchyma. These are longitudinally disposed. The axis is composed of concentric layers, which, however, are not very distinct ; the outermost of these may be detached in flakes, and the actual arrangement of the spicules is there visible. When boiling down the ccenenchyma in strong caustic potash many of the spicules become detached from the axis, and prolonged boiling of the axis disintegrates the greater part of it. The axis is separated from the creuenchyma-proper by a thin horny layer in which spicules identical with those of the axis are embedded; this detaches with the ccenenchyma, but it is undoubtedly a young layer of the axis. The ccenenchyma is almost uniformly thick throughout, being, however, slightly thinner towards the base. The proportion of ccenenchyma to axis is markedly different at the various levels (Fig. 3), but in this connection it is noteworthy that increase in thickness towards the base is due not so much to growth in the ccenenchyma as to growth in the axis. The ccenenchyma is comparatively soft, but at the same time densely spiculose. Canal system (Fig. 2). — Around the periphery of the axis there is a series of minute longitudinal canals all of the same size. Towards the outside of the ccenenchyma, between the polyps, not internal to them, there are also a number of longitudinal canals, and several are also scattered in the ccenenchyma between these two series. The whole of the ccenenchyma external to the inner series is penetrated by a net- work of transverse canals uniting the various longitudinals. The polyps are distributed over the whole of the ccenenchyma (Fig. 3), the actual number at one level being dependent upon the position in the colony. There are no very distinct verruca, due no doubt to the great thickness of the ccenenchyma; the anthocodiae are withdrawn into the feaf Kio. I. Colony, to show gpiioiul liul>it. Kui. 2. Tiaiisviisi' siTtions at tliieo levels, to show tl.i- intiinal ^truuliin;. r--^' ii^'W ■■■:■ it -.'^ iTH Fig. 3. Small portion enlargefl, to sho« the nature and distribution of the verm Fio. 4. Spicules. ccenenchyma, and their position is marked by a slightly elevated wart-like projection with a central pore, around which may be discerned an eight- rayed figure. The spicules of the ccenenchyma (Fig. 4, a, h, and c) consist almost entirely of triple-stars. Of these there are two distinct types — (a) triple-stars with very densely-warted and closely-set heads, and with very short handles ; (6) triple-stars with much longer handles and with the heads covered with openly-set long warts ; (c) a few double clubs also occur. Most of these spicules are red in colour. The spicules of the axis are colourless ; they are very irregular in shape. Some are almost smooth, while others are warty. A few of the characteristic variations in these spicules are shown in Fig. 4 (d-z). The following are some of the measurements ^ in millimetres : — 1. Ccenenchyma (Fig. 4, «-c). (a) Triple-stars with short shafts and densely-warted star portions. 0-076 ; 0-068 ; 0-065 ; 0-038 ; 0-03. (6) Triple-stars with long shafts and witli few simple waits in the star portion. 0-087 ; 0-068 : 0-0C5 ; OO.nS. (c) Double-stars — a few of these occur, but they are usually small and may be undifferentiated triple-stars. 0072 X 0-03 ; 0057 x 0-042 ; 0038 x 0019. 2. Axis (Fig. 4, d-i). 0-152x0-076; 0-114x0065; 0-106x006; 0106x0075. There can be no doubt that our specimens belong (1) to the Order Pseuda.xonia, and (2) to the Family Sclerogorgidre, which is thus defined by Bourne (Lankester's Treatise of Zoology, part ii., "The Anthozoa," p. 25): "The medullary mass forms a distinct axis consisting of closely packed elongate spicules with dense horny sheets. The axis does not contain solenia but is surrounded by longitudinal canals, i.e., by large solenia which are connected with tiie zooid cavities by smaller ramifying solenia." The position of the Gorgonellidffi, in general classification, is a matter of some difficulty, but in a memoir on that family about to be published, we have suggested that forms such as the one under consideration may be annectent between types like Suhcroyor/jia and types like Jinicella, but our present knowledge does not warrant the inclusion of this form in the I The measurement of the triple-stars taken is the maximum, i.e., from the extreme end of one " star " to the extreme end of an .id jacent " stjir." 8 Gorgouellidie. The axis is markedly " sclerogorgic " iu that it consists of individual spicules, different from those in the ccenenchyma, embedded in a horny matrix. The horny substance, in which the spicules lie longitudinally, is deposited in the form of concentric lamime, but the same applies to the genus Suhcrogorijia, as was pointed out by Gray in his original description of this genus, where he says : — " Axis, pale brown, formed of rather loosely concentric fibrous lamimv, containing a large quantity of calcareous matter." The proportion of the horny material to the calcareous is, however, very much greater in the present case than iu Suherogorgia; the nature of the colony is quite different ; there is no trace of two main longitudinal canals larger than the others in the present specimen, and the spicules are very distinctive, both those of the axis and those of the ccenenchyma. For these reasons it seems necessary to establish a new genus, which we propose to call Dendrogorgia, in the Family Sclerogorgidse, Order Pseudaxonia. The following generic diagnosis, based however on a single species, may be given : — Colony very robust, slightly branched approximately in one plane; the branching is a false dichotomy. The axis is " sclerogorgic," and is moderately soft; it does not contain solenia, and is composed of concentric lamin;e, consisting of a horny matrix in which spicules are embedded longitudinally. The spicules of the axis are smooth or warty, and very irregular iu shape. The ccenenchyma is very thick, and is almost of a uniform thickness throughout. It is densely packed with small spicules which are pre- dominately triple-star-shaped ; double-stars also occur. The triple-stars are of two kinds — (1) those with short "shafts" and large, closely tuberculated "heads," and (2) those with long "shafts" and openly-warted "heads." The canal system is very definite; it consists of (1) an inner longitudinal series separating the axis from the cojnenchyma, and (2) an outer longitudinal series which is situated near the periphery. Uniting these there is a dense network of small transverse solenia. The polyps are disposed over the whole cteneuchyma ; there is no distinct separation into verructe and anthocodia- ; they are capable of being completely retracted into the ccenenchyma when a small pore surrounded by an octoradiate structure is to be seen. Zoca/t^y.— Bird Island, E. by N. { N., 5 miles (Cape Colony); depth, 40 fathoms ; bottom — mud. {Issued separately, \9lh Februan/ 1910. 9 II. — Echinoidea and Asteroidea from the Mergui Archipelago and Moskos Islands, Lower Burma. By R. N. Rudmose Brown, D.Sc. (Received 12th October 1909. Read 22nd November 1909.) The material which forms the subject of the present paper was collected by Mr James J. Simpson and myself, during our investigation of the pearl-oyster fisheries of the Mergui Archipelago on behalf of the Indian Government in 1907. Previous to this, the only published record of Echinoids and Asteroids from the Mergui Archipelago was to be found in papers ' by Prof P. ilartin Duncan and Mr W. P. Sladen, founded on material obtained by Dr John Anderson in 1882. From the Moskos Islands several species of Echinoderms are recorded by Dr A. E. S. Anderson in the Surgeon Naturalists Report for the season 1898-1899 (Report of the Marine Survey of India, 1899). The present collections would have been larger if time and opportunity had permitted, for the fauna of the Archipelago is clearly a rich one, but both Mr Simpson and myself had to concentrate all our attention on the pearl oysters and questions immediately relating to that subject : there was, in consequence, all too little time for general collecting. Some of the largest specimens, notably those of Pentaceros superbus, P. lincki, and P. gracilis, we could never hope to have obtained in perfect condition with a dredge ; these we owe to our divers, whom we instructed to bring up anything they could find in addition to pearl and mother-of-pearl oysters. This is probably one of the first expeditions in which a diver has been employed to make zoological collections, and the success of the method, even more in other groups than in Echinoderms, should commend it to others. My acknowledgments are due to Prof. F. Jeffrey Bell for his valuable opinion on certain species ; to Prof. J. Arthur Thomson, who kindly gave me accommodation in his laboratory at Marischal College, Aberdeen ; and to ])r W. S. Bruce, who was of great assistance to me in obtaining literature. ECHINOIDEA. The present collection contains fifteen species, none of which are new. Dr Anderson, in 1882, collected si.\ species, two of which, Temnopleurus toreumaticus (Klein), Agass., and Arachiwides placenta (Linn.), Agasa., are not represented in this collection. The total number of species of Echinoids known from the Mergui Archipelago is, therefore, seventeen. ' " Echinoidea of the Mergui Arcliipelago," by P. M. Duncan and W. P. Sludc-n, Jmir. Linn. Soc. London ZooL, xxi. (1889), pp. ■.•,ir,..nO. "Asteroidea," by W. P. .Sladin, hir. cit., pp. 319-331. 10 The fifteen Echinoids in the present collection are as follows :— 1. Phyllacanthus haculosiis (Lam.), A. Ag. 2. Diadema saxatile, Linn. 3. Echinothrix tm-camm, Eet. 4. Astropyga radiata, Gray. 5. Asthenosmna Ch-ubei, A. Ag. 6. Echimmetra lucunter, de Blainv. 7. Salmads bicolor, Agass. 8. Salmacis Dussiimieri, Agass. 9. Salmacis sulcata, Agass. 10. Salmacis ghhator (Bell). 11. Mespilia glohuhcs, Agass. 12. Laganum depressum, Less. 13. Laganum decagonale, de Blainv. 14. Laganum sp. 15. Lovenia subcarinata. Gray. The most noteworthy species in this collection are Asthenosoma Gruhei and Salmacis glohator. A. Crruiei, of interest in itself as a specimen of a little known species, has an additional claim to notice since its occurrence in the Mergui Archipelago extends to the Indian Ocean the range of a species that was previously known only from the Philippines and Dutch East Indies. Salviacis glohator has been recorded previously from Singapore and the Pacific, but never from the Indian Ocean. 1. Phyllacanthus baculosus (Lam.), A. Ag. A. Ag., Rev. Ech., p. 388.' Locality. — XXIX.,^ High Peaked Island, coral reef. A large specimen, 60 mm. in height and 75 ram. in diameter. Most of the spines are of a very light-green coloui- banded with faint purple. Distributed from the Eed Sea, Zanzibar, ^lozambique, and Mauritius, to Timor and the Philippines. 2. Diadema saxatile, Linn. A Ag., Rev. Ech., p. 274. Localities. — III., Iron Island, shore pools; XIV. and XXIX., Bushby Island and High Peaked Island, coral reefs. ' References are given only to Agassiz's Revision, except in the case of species described since the date of that work. ^ These numbers refer to the Stations. 11 It is noticeable that at Iron Island this species flourishes in rock pools, away from any coral reefs, contrary to its usual habit. Known from the Atlantic, Indian, and Pacific Oceans. 3. Echinothrix turcarum, Ret. A. Ag., Rev. Ech., p. 416. Locality. — VIII., Port Maria, Elphinstone Island, 3 fathoms, fine sand. The spines of this species are but slightly hollow as a rule : in this specimen, however, they are hollow throughout and very nearly approach the spines of E. calamaris. The ambulacral areas are markedly gibbous near the abactinal area. Height, 27 mm. ; diameter, 55 mm. ; longest spine, 56 mm. Distributed in the Indian Ocean and Pacific to Japan and Fiji. 4. Astropyga radiata, Gray. A. Ag., Rev. Ech., p. 420. Locality. — XXXII., Bentinck Island, 29 fathoms, soft mud. A single small specimen, almost bare of spines. Known from Zanzibar and Mozambique to the Philippines. 5. Asthenosoma Grubei, A. Ag. A. Ag., Proc. Amer. Acad., xiv. p. 200 (1879). Locality. — XIV., Bushby Island, 15 fathoms, rock and sand. A single dry specimen in perfect condition, agreeing with the "Challenger " specimen as described and figured by Sladen (" Challemjcr " Reports, iii. p. 82, and pis. -w., xvi., xvii., etc.). It may be identical with Astlienosoma varium from the China and Java Seas, but Grube's description of that species is too vague and general to be of much value to systematists. Not previously known from the Indian Ocean, and apparently only recorded from the Philippines and Dutch East Indies. C. Echinometra lucunter, de Blainv. A. Ag., Key. Kcli., p. 431. Locality. — Vlll., Port Maria, Elphinstone Island, 3 fathoms, fine sand. Two small specimens in perfect condition. The spines are light-green and olive-green respectively (in alcohol), and are tipped with straw colour; a few narrow abrui)tly towards their extremities. Distributed in the Indian and Pacific Oceans. 12 7. Salmacis bicolor, Agass. A. Ag., Rev. Ecli., p. 471. Localities.— XXXllL, Christmas Island, 23 fathoms, sand and mud; IX., Bentinck Island to Courts Island, 12 to 26 fathoms, coral and sand; XXV., Gregory Group, 4 to 14 fathoms, sand and shell ; II., Iron Island, 10 to 25 fathoms, stones and mud. In general appearance these specimens bear a strong resemblance to those wliich I have referred to S. glohator, Bell, but on closer examination they prove to be distinct. The sutural furrows are marked, and the sutural pores conspicuous ; the tubercles are prominent even above the ambitus ; the spines are numerous all over the test, the ambital ones attaining a length of 15 mm. Height, 30 mm. 27 mm. 24 mm. Diameter, 43 mm. 41 mm. 34 mm. Actinostome, 13 mm. 12 mm. 10 mm. Recorded from the Western Indian Ocean, Ceylon, and the Philippines ; and from King Island, Mergui Archipelago, by Dr Anderson. S. Salmacis Dussumieri, Agass. A. Ag., Rev. Ech., p. 473. Localities. — XXXIII., Christmas Island, 23 fathoms, sand and mud; IX., Courts to Bentinck Islands, 12 to 26 fathoms, coral and sand ; XVI., Alligator Rock, 8 to 18 fathoms, rock and sand ; XXX., Fly Island (High Peaked Island), 8 fathoms, sand ; XXV., Gregory Group, 4 to 14 fathoms, sand and shell. A good series of this species, all with the spines complete. Spines spathiform and up to 17 mm. in length at the ambitus ; much shorter, 5 mm., sharper and relatively stouter on the abactinal surface ; on the actinal surface mostly spathiform, especially around the actinostome, and from 5 to 7 mm. in length. Only ambital and actinal spines banded with purple. Height, 19 mm. 19 mm. 19 mm. 11 mm. Diameter, 42 mm. 41 mm. 42 mm. 29 mm. Recorded from the China Seas, the Philippines, and "East India Islands"; and from King Island, Mergui Archipelago, by Dr Anderson; also Mozambique. 9. Salmacis sulcata, Agass. A. Ag., Rev. Ech., p. 476. Localities.— IX., Bentinck to Courts Islands, 12 to 26 fathoms, coral and sand ; XVIII., Ott Paway Island, 10 to 21 fathoms, sand and shell ; XVII., Sir 13 John Malcolm to Charlotte Islands, 18 fathoms, coarse sand; XXV., Gregory Group, 4 to 14 fathoms, sand and shell; XXIV., Cat and Kitten, 12 fathoms, rock and sand. All small specimens, the largest being 38 mm. in diameter. In the largest the spines are tipped with violet ; the smaller ones have them entirely white. Height, 18 and 20 mm. Diameter, 37 and 38 mm. Recorded from the Eed Sea, Mozambique, and Ceylon, to the Philippines and Australia ; and from King Island, Mergui Archipelago, by Dr Anderson. 10. Salmacis globator, Bell. Bell, P.Z.S. (1880), p. 4.31, pi. xli. figs. 2, 3, and 8. Localities. — XVI., Alligator Rock, 8 to 18 fathoms, rock and sand; XXII. Hastings Harbour, St Luke's Island, 15 to 20 fathoms, sand and shell. These two specimens agree in all respects with F. Jeffrey Bell's descrijjtion and figure of S. glohator fi. There can be no doubt as to the identity of the Mergui specimens, and the only difficulty that confronts one is the synonymy of the species (see F. P. Bedford, P.Z.S. (1900), p. 282, pi. xxii.). The primary spines are greenish-white and encircled with narrow bright red bands at irregular intervals ; on the actinal surface they are more numerous and longer, attaining a length of 10 mm. ai'ound the actinostome ; they are flattened at the tips. Height, Diameter, Actinostome, Professor Bell's Salmacis alexandri {S. glohator a) (loc. ril.) is not represented in this collection. Distributed from Singapore to the east coast of Australia, and possibly in the Korean Straits. 1 1 . Mespilia globulus, Agass. A. A^'., Hl-v. Ech., p. 477. Localities. — Vlll., Port Maria, Elphinstone Island, 4 fathoms, sand; XX., High Island, 5 fathoms, sand and rock; XXV., Gregory Group, 4 to 14 fathoms, sand and shell. The median tracts of the ambulacral areas are scarcely bare except in the largest specimens. Not previously recorded from tlie Indian Ucean, but known from ihe Pliilippines, Japan, and the Sandwich Islands. 41 mm. 36 mm. 59 mm. 57 mm, 15 mm. 16 mm, 14 1'2. Laganum depressum, Less. A. Ag., Rev. Ech., p. 518. Localities.— Ylll; Port Maria, Elphinstone Island, 3 fathoms, fine sand; XXXII., Bentinck Island, 29 fathoms, soft mud. The edges of the larger specimens have decided re-entrant angles, especially posteriorly. A young specimen (22-5 mm. long) has the angular outline of the adult, but the edges are straight. Known from Zanzibar and Mozambique to the Philippines and Australia ; and collected by Dr Anderson at King Island, Mergui Archipelago. 13. Laganum decagonale, de Blainv. Peronella decagonalis, A. Ag., Rev. Ech., p. 520. Localities.— XXXIL, Bentinck Island, 29 fathoms, soft mud; XIII., Maria Island, shore pools ; XVIII., Off Paway Island, 10 to 21 fathoms, sand and shell ; XXV., Gregory Group, 4 to 14 fathoms, sand and shell ; XXII., Hastings Harbour, St Luke's Island, 15 to 20 fathoms, sand and shell ; and XLI., Moskos Islands, 12 to 15 fathoms, sand and rock. The largest specimen has a longitudinal diameter of 108 mm. The youngest specimens have an almost circular outline with a bare suggestion of the angles. A young denuded test (31 mm. diameter) from Hastings Harbour recalls Clypeaster humilis, but, I think, is referable to this species. Distributed in the Bay of Bengal and the Western Pacific. Also recorded from Mozambique. 14. Laganum sp. Locality. — XVIII., Off Paway Island, 10 to 21 fathoms, sand and shell. Five small dead and denuded tests were found in the dredge; the largest has a longitudinal diameter of 17 mm. The test is orbicular and somewhat swollen ; there are four genital pores. These specimens seem to be identical with those described and figured by F. P. IJedford as Laganum sp., from Singapore and Malacca {P.Z.S. (1900), p. 285, pi. xxiii. fig. 7, a and h). They may or may not be the young of L. depressum. 15. Lovenia subcarinata, Gray. A. Ag., Rev. Ech., p. 377. Locality. — I., Tavoy Island, 10 fathoms, sand and shell. 15 Three individuals were obtained, of which the largest is 32 miu. long and 24 mm. broad. Its occurrence in the Mergui Archipelago extends the range of the species, which was previously known only from the Philippines to China, Japan, and the Sandwich Islands, and Dutch East Indies. Later it has, however, been recorded from the Mozambique coast. ASTEROIDEA. Xineteen species of Asteroids were collected, including no new species but many new records for the ^Mergui Archipelago, and, in some cases, for the Indian Ocean. These collections have therefore not borne out Dr Anderson's belief of " a reasonable expectation that a number of new species may ultimately be found in the Mergui Archipelago." Dr Anderson ' collected nine species, of which three were new. Of these nine species only four occur in the present collection, or, at the most, five, if we can include Astropecten JTcinprichii, should it prove to be the same as A. niauntuoius. None of Anderson's new species are included in our collection. The total number of Asteroids known from the Mergui Archipelago is therefore 24, or possibly 23. In addition, a species of Cidcita was frequently seen on the coral reefs, but no specimen reached this country. The Asteroids in the present collection are as follows : — 1. Archaster typictui, M. & T. 2. Graspidaster hesperus (M. & T.), Sladen 3. Astropecten mawitianns, Gray. 4. Astropecten zebra, Sladen. 5. Astropecten poli/acanthics, M. & T. G. Luidia fordfer, Sladen. 7. Luiclia macnlota, M. & T. 8. Goniodiscus articulatm (Linn.), Liitkeu. 9 Stellaster incei. Gray. 10. Anthenea jlavcscens (Gray), Perr. 11. Anthenea pentagomda (Lam.), Perr. 12. Peiiiaceros granulosus, Gray. 13. Pentaceros lincki (de Blainv.). 14. Pentaceros superhus, Mobius. 15. Pentaceros gracilis, Liitken. 16. Palmipcs rosaccns (Lam.), Duj. & Hupd. ' Jour. Linn. Soc. Lond. Zool., xxi. (\»CQ), pp. 319-331. 16 17. Fromia viillepm-ella, Gray. 18. Retaster cribrosiis (von Mart.). 19. Echinastcr pwyureus (Gray), Bell. The absence of any species of Linckia is noticeable, for this genus is widely diffused and common in many parts of the Indian Ocean. Three species constitute new records for the Indian Oceau, namely, Craspidaster hesperus, a Pacific species, and Anihenea Jlavescens and Pentaceros granulosus, both Australian forms. Pentaceros gracilis is extremely abundant on the pearl banks of the Mergui Archipelago, and has since been noted in numbers on the pearl banks of the Mozambique coast of Africa ; yet, as far as I can ascertain, it was previous to these discoveries looked upon as an Australian form ; a fact which forcibly illustrates the poverty of our knowledge of the marine fauna of many parts of the Indian Ocean. 1. Archaster typicus, M. & T. Perrier, Revision de Stellerides,' Arch, de Zool. e.\per. et gen., v. (1875), p. 265. Localiti/. — XVI., Alligator Eock, 8 to 18 fathoms, rock and sand. Two specimens in which .K = 44 and 46 and r = 10 and 9 respectively.- In the smaller specimen a single spine appears on one supero-marginal ; otherwise the supero-marginals have no trace of spines. A similar occurrence in this species is noted by Liitken {Vidensk. Medd. (1864), p. 136), and by Sladen (" Challenger " Reports, xx.x., p. 124). This species is also recorded from the Mergui Archipelago by Dr Anderson in 1882. "Widely distributed in Eastern Indian Ocean and Western Pacific. 2. Craspidaster hesperus (M. k T.), Sladen. Sladen, "Challenger" Reports, xx.x. p. 177, pi. xvii. fitrs. 5-7 ; and pi. xviii. tigs. 1-4. Locality. — XXXII., Off Bentinck Island, 29 fathoms, soft mud and sand. Two specimens agree with the young phase described by Sladen. The adpressed spinelets on the infero-marginal plates tend to fall oft' very readily, but their presence is quite evident in both specimens. R = 22 r=6-5. Apparently not previously recorded from the Indian Ocean. Known from Japan to Singapore. ' In the case of species included in Perrier's Eevision, I give no other reference than to his paper as a rule : the synonymy is therein discussed. - All measurements are in millimetres. 17 3. Astropecten mauritianus, Ciray. Perrier, loc. dt., v. p. 279. Localities. — Frequent on pearl banks throughout the Archipehigo, in 5 to 25 fathoms, sand and rock. R^lll r=21-5 ^=81 >-=19. Undoubtedly this species is very closely allied to A. Hempnchii, M. & T., despite de Loriol's belief that the two species cannot be confused (M^m. Soc. Phy. Hist. Nat. Gen., t. xxix., No. 4, p. 74, and pi. xxi.). He lays stress on the absence of supero-marginal spines in the angles of A. Ifemprichii and their presence in A. mauritianus, but this, as Jeffrey Bell has pointed out, is not a reliable character in Astropecten. It will not be surprising if the two species prove to be one and the same. I have named the Mergui specimens after careful comparison with specimens in the British Museum. A new record for the P^astern Indian Ocean, but Dr Anderson collected A. Hemprichii in the Mergui Archipelago. 4. Astropecten zebra, .Sladnn. Slailen, '•Challenger" Reports, xix. p. 212, and pi. xx.wi. figs. .3 ami 4 ; pi. xxxix., figs. 7 to 9. Localities. — IX., Courts Island to Bentinck Island, 12 to 26 fathoms, coral and sand ; XXV., Gregory Group, 4 to 14 fathoms, sand and broken shell. Four specimens, the two larger of which are each in process of regrowing an arm. i2=2G /• = 7-.5 A' =30 r = 8 E = 25 r=7 A' = 16-5 r = 5-5. A slight prominence in the centre of the upper surface of the disc appears as a conical beak in the smaller specimens; it is most prominent in the smallest. Of the supero-marginal plates, as many as tiie eight innermost on either side of each arm may bear spines, but of these the last two are very rudimentary. In the smallest specimen, one or two of the supero-marginals on each side of the median interradial line bear spines. Even in the larger specimens, however, the number of spines is variable, bearing out .Jeffrey Bell's contention as to the absence of value of these in classification (Hornell and Herdman, Ceylon Pearl Oyster Fisheries Report, ii. p. 149). Bell thinks {loc. cit.) that A. zebra and A. Hemprichii are identical. They certainly approach one another very closuly in their characters, but those whicli I'ell figures (from photographs) as^l. Hemjiriihii seem to be A. zehra and apparently were thus named by Sladen. Known previously from Torres Strait and Ceylon. 18 5. Astropecten polyacanthus, M. i T. Perrier, loc. cit., v. p. 275. Locality. — I., Tavoy Island, 8 fathoms, shelly sand and mud. .B = 15"5 r = 4. Breadth of arm at base = 4'5. A single specimen, apparently a young form, in which the spines of the ventral surface are scarcely developed. One supero-marginal on either side of the median interradial line bears a distinct tooth-like spine inclined slightly inwards. The specimen seems to be referable to this species. Eecorded previously from the Eed Sea, Zanzibar, and Mozambique, the Seychelles and Ceylon to Hong Kong, the Fiji Islands, and Port Jackson. 6. Luidia forcifer, Sladen. Sladen, " Challenger " Reports, sxx. p. 258, pi. xliv. fig. 5, and pi. xlv. figs. 5 and 6. Locality. — IX., Between Courts Island and Bentinck Island, 12 to 26 fathoms, coral and sand. A single specimen, not of full growth, but which can be referred to this species. 7i = 30 /• = 5. Collected by Dr Anderson at King Island and at Sir William James Island in the Archipelago, and also known from Torres Strait and the Arafura Sea. 7. Luidia macnlata, M. . 512. Localities. — X., Montepes Bay, sand and mud, 5 to 22 fathoms ; XIII., Pemba Bay, mud, 10 to 20 fathoms. A large and a small specimen. The outline of the former is decidedly pentagonal with rounded edges. The abactinal surface, from the swollen border, is convex but slightly flattened again at the apex. The extremities of the poriferous zones of the petals are a little concave. Height, 11 mm. Long, dia., 51 mm. Trans, dia., 42 mm. „ 7 mm. „ „ 31 mm. „ „ 21 mm. Distribution. — .Japan, Formosa, Malay Archipelago generally, and Indian Ocean including Kea Sea, Ceylon, Mauritius, and Ibo. 14. Clypeaster humilis, Agass. A. Ag., Rev. Ech., p. 510. Locality. — II., Maiyapa Bay, sand, mud, and coral, 10 fathoms. Several young individuals having the characteristic form of the adult. Their slightly pentagonal outline gives them a strong resemblance to certain species of Laganuni. Distribution. — Western Pacific and Indian Oceans. 30 lo. Laganum depressum, Less. A. Ag., Rev. Ech., p. 518. Localities.— 11., Maiyapa Bay, sand, mud, and coral, 10 fathoms ; III., Mtundo Bay, sand, shell, and coral, 6 fathoms ; VI., Kero-Nyuni Bay, sand, 5 to 10 fathoms. Many specimens of various ages. In shape these specimens show many of the variations in outline habitual in this species: the truncated angles give a sub-decagonal or sometimes almost orbicular outline. The greatest transverse diameter is just anterior to the anterior pair of ambulacra in the largest specimen ; in the smaller ones, it tends to run posterior to them or across the apex : this is not in accordance with what A. Agassiz describes (loc. cit.). In the larger specimens, the petals are relatively shorter than in the young, and in the largest specimen of all the margin of the test is distinctly swollen. In this specimen the posterior sides tend to show re-entrant angles, and on one side the test has evidently been damaged and partial regeneration, at least to the extent of the upper and lower surfaces fusing, has taken place. The largest specimen has these dimensions : — Long, dia., 68 mm. Length, ant. pair ambulacra, 17 mm. Trans, dia., 65 mm. Length, odd ambulacrum, 18 mm. Distribution. ■ — Zanzibar, ilergui, Australia, the Philippines, and Fiji Islands. 16. Laganum decagonale, de Blainv. A. Ag., Rev. Ech., p. 520 Locality. — III., Mtundo Bay, sand, shell, and coral, 6 fathoms. A single dead specimen rather more oval than decagonal in outline. Distribution. — Western Pacific and Bay of Bengal. IV. Echinodiscus auritus, Leske. A. Ag., Rev. Ecli., p. 531. Localities. — Kifuki Island and Foomo Island. Only one specimen was collected but it is a very perfect one. The outline, especially posteriorly, is somewhat irregular. The edge on one side anterior to the lunule is very ragged and gives the impression that it has suffered injury and has undergone a certain degree of re- generation. The length of the lunule is 40 mm., the total distance from the edge of the test to the apex is 92 ; this is a longer lunule than generally is found. The anus, however, is as usual in a line with the inner ends of the lunules. 31 The measurements of this tine specimen are worth recording. Height, 13 mm. Long, dia., 158 mm. Trans, dia., 147 mm. Post, petals, 29 mm. Odd petal, 34 mm. Width, porif. zone, 5 mm. Width, inter, porif. zone, 55 mm. Anus from ed;_'e of test, 37 mm. Mouth from anterior edge, 76 mm. Distribution. — Indian Ocean, especially the west, Amboiua, Philippines, and Straits of Macassar. 18. Maretia planulata, Gray. A. Ag., Rev. Ech., p. 570. Localities. — III., Mtundo Bay, sand, shell, and coral, 6 fathoms ; VI., Kero-Nyuui Bay, sand, 5 to 10 fathoms ; X., Montepes Bay, sand and mud, 5 to 22 fathoms. A series of seven specimens of various ages. All the specimens seem to be a little flatter than is usual in this species. The slight anterior ambital grove, never at all prominent, is practically indistinguishable in the largest specimens. The variations in colour (in spirit specimens) common in this species are not noticeable. A uniform straw colour prevails, except in one specimen, in which the lateral petals on one side show a certain amount of violet brown coloration, and the same colour appears on the primary tubercles. The primary spines in this case are banded with faint violet brown. Height, 16 mm. Long, dia., 68 mm. Trans, dia., 54 mm. 12 mm. „ „ 46 mm. „ „ 37 mm. Distribution. — Western Pacific through the Indian Ocean to Zanzibar. 19. Lovenia subcarinata, Gray. A. Ag., Rev. Ech., p. 577. Localities. — VI., Kero-Nyuni Bay, sami, 5 to 10 fathoms; XllL, Pemba Bay, mud, 10 to 20 fathoms. Several fine specimens. The specific characters which distinguish this species from L. elongata in young specimens — the sub-triangular actinostome, the outward slope from the anal system to the abactinal surface, and the triangular actinal plastron — are maintained in tiie older ones. More mature specimens show no departure from the small number of primary tubercles on the abactinal surface in the anterior half of the posterio-lateral mterambulacra. 32 Height, 19 mm. Length, 46 mm. Diameter, 35 mm. 19 mm. ,. 47 mm. „ 34 mm. 14 mm. „ 36 mm. „ 26 mm. 11 mm. „ 29 mm. „ 22 mm. Distrihution. — Only known from the Western Pacific — Japan to the Sandwich Islands — and the Mergui Archipelago. 20. Brissopsis luzonica, A. Ag. A. Ag., Rev. Ech., p. 593. Locality. — V., Namegus Bay, mud and rocks. A solitary injured specimen belongs to this species. The anal plastron is missing. The bare areas on the actinal surface with undiminished width run from the mouth to the sub-anal plastron. Height, 23 mm. Long dia., 45 mm. (approx.). Trans, dia., 38 mm. Distrilmtion. — Western Pacific, from New Zealand to Japan. This record gives a great extension of range. o -1. Schizaster gibberulus, Agass. A. A^., Rev. Ech., p. 612. Locality. — VI., Kero-Nyuni Bay, sand, 5 to 10 fathoms. A single specimen. This species strongly resembles S. canaliferus. The points of difference between the two species enumerated by Agassiz {loc. cit.) are not of great value, for all are very slight. If anything, the test of S. gibberulus is narrower and more arched, the apical system is slightly more anterior, and the bare abactinal surfaces are larger than in S. cancdi/erus. The tubercles of the actinal plastron cannot be said to be " coarser" and " more distinct," but the spines on the anterior half of the plastron are conspicuously spathulate. The species is certainly very closely allied to the Mediterranean S. canali/ents. Height, IS mm. Length, ant. petals, 16 mm. Long, dia., 42 mm. Length, post, petals, 7 mm. Trans, dia., 35 mm. Length, odd petal, 22 mm. Distribution. — Red Sea and Ceylon. {Issued separately, \7lh February 1910.) 33 IV. — Asteroidea of Portuguese East Africa, collected by Jas. J. Simpson, M.A., B.Se. (1907-i9()fi). By Jas. J. Simpson, M.A., B.Sc, and R. N. Rudmose Brown, ixSc. (Received r2th October 1909. Read •22nd November 1909.) The present collections were made on the Mozambique Coast of Portuguese East Africa, between September 1907 and May 1908. The portion of the coast on which faunistie work was done extends from near Cape Delgado in 10° 42' S., to Pemba Bay (Maunhane Point), 12° 58' S. Very little work of this nature has been done on the western shores of the Indian Ocean, so that, although the collections made by the E.I.M.S. " Investigator " have done much to elucidate the fauna of the eastern side, our knowledge of the fauna of the east coast of Alrica is very scanty. The present collections therefore may help in some measure to fill in many gaps in our knowledge, both of the Indian Ocean fauna in general and in the distribution of many tropical forms in particular. The collection of Asteroids consists of fourteen named species, but in addition to these we have described four species which we have at present refrained from naming. The latter include two species of Pentaceros and one of Anthenea ; they are probably all young forms, so that although it has been impossible to refer them to any known species we do not feel justified in describing them as new. A study of different stages of growth, in other species of Asteroids, has convinced us that it is almost impossible to be certain of any species from a young specimen, and we strongly agree with Professor F. Jeffrey Bell that the application of new names to what may be young and immature stages is a course to be thoroughly condemned. A very good example of this is the species Linckia marmorata, which we have here described in some detail, as from the original description, based on a young specimen, it is almost impossible to identify mature individuals. We have, however, given short descriptions of the essential diagnostic features ot' these young forms, as they may, by comparison with others, help to form a series of developmental stages in definite species. We have attempted to show the geo^iraphical range of those species which occur on this coast. Archaster nngulaUcs is here recorded for the first time from the western side of the Indian Ocean ; and Luidia aspera appears to be new for the East coast of Africa. The bathymetrical range in distribution and the nature of the associated bottom has also been recorded when possible, and a comparison made with other records. Numerous observations on colour schemes have been given, and these, e :]4 though often ignored by the cabinet zoologist, are of great significance to the field biologist, and show the futility of basing specific characteristics on colour. Very good examples of this are seen in species of Fentaceros and Culcita We would like here to draw attention to a peculiar association or commensalism which, although it has been observed in Holothurians, has so far escaped notice in Asteroids, viz., that between a species of Fierasfer (as yet unidentified) and Fentaceros lincki. While dissecting specimens of this species of starfish for drying, we were surprised to find occasionally a small Fierasfer aliye; careful dissection showed that these fish, occasionally in pairs, inhabited the stomach of the starfish, but were quite healthy and normal : in fact, we were able to keep them alive for some time in tanks. On placing some of these large starfish in the same tanks, it was possible to watch the Fierasfers passing out and in at the " mouth " of the Fentaceros. This disappearance and reappear- ance continued daily, so that we feel quite in a position to say that it is another example of comraensalisin, similar to that between fish and Holo- thurians, added to our knowledge of "associations." Several very interesting specimens of Linckia dvplax showing difierent stages in the regeneration of lost arms occur in the collection. Especially noteworthy are the comet-forms. We are glad to have this opportunity of expressing our indebtedness to Professor J. Arthur Thomson, for kindly granting us laboratory accom- modation in his department in the University of Aberdeen, as well as for other facilities; and we wish also to thank Professor F. Jeffrey Bell, for allowing us to compare our specimens with tliose in the British Museum, and for the interest he has taken in the collection. For general utility we have adopted the classification given by Sladen in his report on the '• Challenger " asteroids (^'Challenger" Reports, Zoology, vol. xxx.), although a better classification has since been published in Bronn's Thierreich. In accordance with the former, the following table shows the general relationships of the various genera and species reported upon : — Class ASTEROIDEA. Sub-Class Euasieroidea, Sladen. Order I. PHANEROZONIA, Sladen. Family A k c h a s t E k i d ^. Sub-Family Archastekln.*. Archaster angulatus, M. and T. 35 Family A s T i; o p E c t e K i D .«. Sub-Family Astropectenin^. Astropecten Hemprichii, M. and T. „ poll/acanthus, M. and T. Sub-Family Luidiin^e. Luidia maadata, M. and T. „ aspera, Sladen. Family P e N T a G o n A s T E K 1 D ,«, Perrier. Sub-Family Goniodiscin.*:. Stellaster incei. Gray. Family A N T h e n e i d .?:. Anthenea sp. Family P e N T a c e r o t i d ^. Pentaceros lincki, (de Blainv.). „ superbus, Mbbius. „ gracilis, Liitken. sp- Culcita schmideliana (Retz.). „ sp- Order IT. CRYPTOZONIA, Sladen. Family Linckiid.-e. Sub-Family Lisckiin/E. Ophidiaster cylitulricus (Lam.), M. and T. Li/ivkia diphu-, M. and T. „ marmorata (Michelin). Nardoa variolata, Gray. F'amily P r e k a .s r e k i i> .e. Sub-Family Ptekasterin/E. Retailer cribrosvs (vdn Mart.). 36 Archaster angulatus, M. and T. This typically shallow-water species is represented in the collection by four specimens of different size, all captured in the same haul of the dredge in Mtundo Bay, between the islands of Wamizi and Kifuki. Their measurements ^ are as follows : — E=58 r=10 E=55 r = 9o ^=36 r=7 I! = 27 to 29 r=6 R therefore varies from iSr to 5'8r. In the first and largest specimen one arm has evidently been regenerated. This is the first record, as far as we can discover, from the extreme western side of the Indian Ocean. Locality. — Station III., Mtundo Bay (Wamizi Island to Kifuki Island). Previously recorded from — West Australia ; Freemantle ; Port Darwin ; Tijrres Strait ; New Guinea: Philippines; Fiji Is.; Mauritius. Astropecten Hemprichii, M. and T. Several specimens of various ages and slightly different in superficial appearance represent this species. They were obtained in three separate localities, viz., Tunghi Bay, Mtundo Bay, and Montepes Bay. They agree exactly with those of the same species in the British Museum, collected by H.M.S. "Alert" on the Mozambique Coast. One specimen has been regenerating three arms from the disc. Localities. — Station I., Tunghi Bay; Station III., Mtundo Bay (Wamizi Is. to Kifuki Is.); Station X., Montepes Bay. Previously recorded from — Mozambique Coast ; Red Sea ; Mauritius ; Ceylon ; Tuticorin ; Mergui Archipelago. Astropecten polyacanthus, M. and T. In the specimens, which we have identified with this species, the first supero-marginal plate is strongly armed with a vertical spine. The second supero-marginal on each side of the median interradial line is devoid of a spine, while the remainder have the same structure as the first. Sladen draws attention to a specimen in which this characteristic absence is not pronounced. This species is fairly abundant on the coast, and its distribution is ' All measurements are in millimetres. 37 interesting. It is a typically shallow-water species, as the following bathy- nietrical notes will show : — China, . . . \ Beach. Japan, Admiralty Islands, . Port Jackson, . Mergui Archipelago, East Africa, 5 to 25 fathoms. 16 to 25 fathoms. 2 to 11 fathoms, and 6 to 15 fathoms. 8 fathoms. 6 to 12 fathoms. Locality. — Station VI., Kero-Nyuni Bay, near Ras Pekawi. Previously recorded from — Japan; China: Fiji Is.; Admiralty Is.; Port Jackson ; Banda Sea ; Arafura Sea ; Macclesfield Bank : N. W. Australia ; New Zealand ; Andaman Is. ; Mergui Archipelago ; Ceylon ; Seychelles ; Amirante Is. ; Mauritius ; Zanzibar ; and the Red Sea. Laidia maculata, M. and T. This species is represented by two specimens : the first, from Mtundo Bay, has seven arms ; and the .second, which is immature, has five complete arms, and three being regenerated from the disc. The diagnostic character.? in this species seem to be fairly constant, and our specimens agree very well with those previously described. Locality. — Station III., Mtundo Bay (Wamizi Is. to Kifuki Is.) ; Station VI., Kero-Nyuni Bay. Previously recorded from — S. Japan ; Philippines ; Singapore : Malacca ; Macclesfield Bank ; Mergui Archipelago : Tuticorin ; and Mozambique. Luidia aspera, Sladen. This is a very distinct species, and we have no hesitation in referring our specimens to it. It is characterised by the form of the paxilhe and by the armature of the infero-marginal plates. The species is represented by one specimen from Tunghi Hay, and several detached arms which were obtained in a different haul of the dredge near the same place. The following are the measurements of the complete individual : — ;?=145 r=16 i2 = 9r. Sladen, in describing this species, gives Ii-7 5r and R = 8-5r. He also remarks that the number of arms varies from 8 to 10, but in the above specimen the number was 7, so that this feature cannot be regarded as specific. The bathymetrical distribution of this species shows a considerable range. The specimens from the Admiralty Islands were obtained in from 10 to 150 38 fathoms, whereas those on the East Coast of Africa were dredged in 9 to 15 fathoms. ZocaZ%.— Station I., Tunghi Bay. Prevwusly recorded Jrom—VhiU^Tpines ; Admiralty Is.; Macclesfield Bank. Stellaster incei, Gray. This species is represented by a single specimen, iu which E=285 and r=10. The nature ot the bottom on which a certain species exists is always interesting, inasmuch as it has olten a distinct bearing on the individual specimens; but S. incei does not seem to be associated with any particular kind of bottom. At Mergui we found it occurring on — (1) mud and stones ; (2) sand and mud ; (3) coral and sand ; and (4) sand and shell. Its bathymetrical distribution is, however, restricted. The following records are noteworthy :— East Africa, 3 to 15 fathoms; Torres Strait, 6 to 8 fathoms; Arafura Sea, 28 fathoms; Philippines, 18 fathoms; Port Molle, 5 to 11 fathoms; Port Denison, 6 fathoms; Ceylon, up to 40 fathoms. Locality. — Station VI., Kero-Nyuni Bay. Previously recorded from — Korea ; Singapore ; North-West Australia; Port Curtis and Albany Island ; and Mergui Archipelago (in addition to the above). Anthenea sp.? There occur in the collection seven small specimens which we have referred to the genus Anthenea. They are all immature, and as we were unable to identify them with any known species, we have refrained from establishing a new species to include them. The following notes, however, may be ot interest: — 7i' = 4-J /■=18 A = 2-3r R = \l r=18 A=2-3r A' = 41 r=17 A =2-4 7- ii = 39 r=17 A=2-3r A =33 r= 15 A=2-2r A = 29 r=13 A = 2-2»- A = 28 /■ = 12 A=2-3r Tliere are nine marginal plates, excluding the unpaired terminal one ; these are rectangular in shape and bear no pedicellariae. Tlie supero-maiginal plates are tuberculated ; all, except those in the arcs, bear two tubercles, while three or four are not infrequent ; the tubercles are arranged transversely. The infero-marginal plates bear spines only towards the extremities of the arms ; three spines sometimes occur, either clustered or disposed longitudinally, never transversely. •Sit The lophial line is composed of about 17 to 19 lounded plates ; the proximal 5 ot these bear tubercles, while the innermost of each line has a very large tubercle : a distinct pentagonal area is thus enclosed. The arrangement of the plates within this area is irreyular: a few sometimes bear minute tubercles. The madrepore is lozenge-shaped, and lies between the two adjacent plates which form one of the angles of the pentagon. Valvular pedicellaria' occur in a discontinuous row on each side of the lophial line, while a few are also scattered on the interlophial plates and in the pentagonal area. On the ventral surface the granulation is not prominent, and the plates are distinct. A definite line of valvular pedicellariae extends on either side of the ambulacral grove ; these lie usually longitudinally, but occasion- ally oblique. Larger forms also occur, scattered on other plates. The ambulacral armature appears to consist of a single series, but near the apex of the arms a trace of a second series of spines is discernible. Each plate bears four almost equal spines ; these are obtuse, flattened at the tips, and arranged in a palisade. Locality. — Station X., Montepes Bay. Pentaceros lincki,i de Blainville. Locality. — Stations I. to X., Bottom— sand, or sand and rock. Previously recoi'ded from — Mergui ; Tuticorin ; Ceylon ; Mozambique ; and Zanzibar. Pentaceros superbus,' Mobius. Locality. — Stations F. to X., Bottom — sand, or sand and rock. Previously recorded from — Tuticorin ; Mergui ; Sumatra. Pentaceros gracilis,' Luiken. Locality. — Siations I. to X., Bottom — sand, or .■-and and rock. Previously recorded from — Mergui; East Australia; Port Molh- and Port Denison, Queensland. ' These three well-known species are extremely abundant over ;iliiiost the whole coast, and are a distinct menace to pearl-oyster beds. During the period over which my work extended on the coast, more than five thousand of these were brout;ht up in the dredge, while on the shallow reefs thousands may be seen daily at low tide. The colour patterns on all these species, but e-ipecially on f . tincki, are worthy of attention. The general tone is in most ca.se9 blue, but the foUowing variations in the colour of the spines were observed — (1) central spines orange, the others creamy-white ; (2) all the spines vermilion-red : (3) all the spines creamy-white. Another type had bright yellow as a groundwork, while the spines were orange-coloured. These few observations demonstrate tlie futility of basing any siwiific character on coloration in brightly- coloured asteroids. J. J. S. 40 Pentaceros sp. In the collection there are three small specimens which are undoubtedly immature, but which belong to the genus Pentaceros. After a careful comparison of the type specimens of the various species of this genus in the British Museum, we have come to the conclusion that they approach nearest to P. nodulosus, Perrier, but in view of the fact that they are young and immature we would refrain from definitely referring them to this species. The following notes may, however, be useful in illustrating the chief points of similarity and contrast. The present specimens differ from the British Museum specimens on which Perrier founded the species P. nodulosus (Archiv. de Zool. Expir., v., 1876, p. 53) in the following respects: — ^ = 67 r = 24 i?=2-8/- iJ = o9 ;-=21 R=-2-Br E = oB r = 21 P=2-3r Perrier '.'ives E = 2^r. Marginal plates: — Type specimen, . 36 to 38. Present specimens, . 15 to 17. Tubercles on the lophial line : — Type specimen, . Ellipsoidal. Present specimens, . Dome-like. In addition to this we may note that in our specimens there are tubercles, conical or dome-like, on the lateral and supero-marginal plates. On the lateral plates they become larger towards the disc, while, in contrast to this, on the supero-marginal plates they are more prominent towards the tip of the arm. Ill larger specimens of P. nodulosus these lateral lines are not to be seen, and it is possible that those of the specimens before us would disappear with the growth of the individual. In want of evidence regarding the possibility of this, we think it more advisable not to identify our specimens with this species at present. In the living animal the tubercles are light yellow, and the rest of the dorsal surface is dark brown. The position of the madreporite is the same as in the type specimen, but whereas in P. nodulosus it is lozenge-shaped, in the present forms it is triangular to heart-shaped. 41 The spines of the ambulacral grove also show some variation : — In P. nodulosiis — (1) inner series, . 7 to 9 spines on each plate. (2) outer series, . 3 larger spines. In the present specimens — (1) inner series, . 5 spines on each plate. (2) outer series, . 2 larger spines. If our view of the species be correct, the great differences which may obtain between immature and mature specimens are most noticeable. Locality. — Station IX., Matemo Island, Ibo Bay. Pentaceros sp. Another small specimen from Tuughi Bay is too young for specific determination. -R = 32; r = 13'5. The disc is distinctly elevated in the manner of P. lincki. On each lophial line are three or four prominent conical tubercles, culminating in size in the topmost one on the central disc, which is 3 to 4 millimetres in height. The tubercles are closely granulated, and through the granulated surface projects the small sub-acute apex. On the last two or three distal infero- and supero-marginal plates there are small spinous tubercles. This may be a young form of P. gracilis, a species in which we know the development of spines is very variable. On the other hand, however, it may be an immature stage of P. lincki, but any categorical statement would be ill-advised. Locality. — Station I., Tunghi Bay. Culcita schmideliana (lletz.). This very variable species is represented in the collection by a single large dried specimen which has become slightly damaged by damp. Only the dorsal skeleton and fragments of the ventral surface remain ; however, the ventral tesselated plates show well. In life the aboral surface was dark brown willi much lighter brown poriferous areas ; the oral surface was purple, becoming paler towards the mouth ; the sides of the ambulacral grove were yellow, and around the actinostome there were distinct dark blue areas. Referring to this species from Ceylon, Professor Herdman {licp. on Ceylon Pearl Oys. Fish., vol. ii. p. 144) remarks that " this cushion-like starfish . . . shows on the aboral surface a bright orange-coloured pentagon closely / 42 papillated, and with the ambulacral groves running as narrow -red lines out to the angles. On the aboral surface there are short red spines on the well-marked lobed areas, while the surface between has a fine fluffy or velvet-like appearance." Locality. — Station IX., Matemo Island. Previoushi recorded from — Eed Sea; Dar-es-Salaam ; Zanzibar; Mozam- bique; Madagascar; Mauritius; Ceylon; Andaman Is.; and Amboina. Culcita sp.? This large starfish, popularly known as the "leather bun," is almost universally distributed along the coast, though not in such abundance as Pentaceros. The following notes may show the protean nature of the colour schemes in this wonderful genus. It is difficult to say if one or more species are here represented, but the former is more probable. These varieties are, at any rate, discernible : — A. With black tubercles and scattered soft poriferous areas. Ground, yellow, . . Areas, grey. blue, bright yellow, salmon-pink, light grey, bright yellow, grey, brown. greenish-brown, yellow, dark grey, yello wish-brown, dark brown. B. With tubercles but no different coloured poriferous areas. Tubercles black, with the general tone yellow, orange-yellow, brown, or pink. C. With coloured poriferous areas but no tubercles. Ground, yellow, Areas, brown, blue, ... „ grey. „ salmon-pink, . „ brown. One specimen was entirely red. These notes show the extraordinary variability in colour of the highly coloured Asteroids, a fact very often overlooked by the cabinet zoologist, but well known to every field naturalist. This is evidence how little, if any, reliance can be placed on colour as a specific character. 43 Ophidiaster cylindricus (Lam.), M. and T. This is a typical coral-reef species, and is represented in the collection by several specimens of different ages from Montepes Bay. The following notes may prove interesting in a study of development : — ^=146 r=U A' = 104 r=ll ^=92 r= 9 E= f.9 ?■= 8-5. In life the colour schemes of this species are extremely striking, and are by far the most brilliant in the associated fauna. Many specimens are dark red all over ; others are of a bright yellow, with dark red to vermilion blotches on the arms. They are somewhat .slimy to the touch when alive. There is never more than one madrepore plate. Locality. — Station X., Montepes Bay, between Kilalia Island and Sinkori Island. Prenamly recorded from — Muscat (Brit. Mus.) ; Mauritius ; Ceylon ; Kondavi, Fiji Islands ; Moluccas. Linckia diplax (M. and T.). This is apparently one of the commonest Asteroids on the coast. There are numerous specimens from three separate localities, viz., Tunghi Bay, Kero-Nyuni Bay, and from the reefs separating Das Kolas Is. from Matemo Is. ^=175 r=ll 7i'=16r. The number of arms varies from four to six, and two madrepore plates occur on every individual except in one with six rays, which has tliree. There are three comet individuals, in one of whicli 72=112, /■ = 9; in another R = 60, r = 8 ; while in the third li = 89, and ?• = 39. In the largest specimens the small rays are thick in proportion to their length, and comparatively blunt. In the others they are almost normal in proportion. In several individuals one or more arms have been truncated, and regeneration is in progress. The two predominant general colours are brown and blue, with black dots. Many, however, are greenish blue on the aboral surface, and bear minute black dots, which give the whole the appearance of a brandling coral. The oral surface is purplisli-red. Localities. — Station 1., Tunghi Bay; Station IX., Matemo Is.; Station YI., Kero-Nyuni Bay. Previously recordrd from — Mozambique ; Madagascar ; Bourbon ; Mauri- tius ; Ceylon; Christmas Is.; New Caledonia; Fiji; and Friendly Islands. 44 Linckia marmorata (Michelin). (Figs. 1-4.) The specimens which we have referred to this species were at first a little perplexing — in fact, from the original description and figures of L. marmorata it is well-nigh impossible to distinguish the species. The type specimen was described by Michelin in 1844 in the Revue Zoologiqite as Ophidiaster mnrmoratus, thus : — " 0. minima ; radiis quinis, cyliudraceis, violaceis et luteolis ; parte superior! longitudinaliter quinque costatis, transverse rugosis, subtilis- sime granulatis, ad interstitiis varie tribus poris ornata ; parte inferior! granulosa ; canalibus clausis ; marginibus papillosis, duabus lineis par- vulis tuberculis ornatis, taberculo madreporiformi parvulo, rotundo, sulcato, violaceo." In 1845 Michelin again referred to it in the Magazin de Zoologie, p. 21, and also figured it (plate 10) — (" Essai d'une faune de I'ile Maurice "). In 1869, von Martens placed the species in the genus Linckia (Claus von der Decken's Eeise ; Seesterne und Seeigel, p. 130); while in 1875 Perrier (Archiv. de Zool. Expir) revised the species but gave little new specific character to it. He notes, however, that the type specimen had the following measurements : — R = 10 mm., r = 4 mm. The species was therefore established on a very immature specimen. In 1884, Bell resuscitated the species and made a valuable contribution to the specific diagnosis {Zool. Coll. of H.M.S. "Alert," p. 125). On this description, and taking as a basis some specimens in the British Museum which Bell has referred to this species, we have identified these puzzling forms in the present collection. The following notes are given as a contribution towards the study of this species, and should be taken in conjunction with Bell's valuable variation references. The arms are five in number ; they are subconical, being slightly flattened on the ventral surface. The measurements of the different specimens in the present collection are : — /i' = 48 r = 9 i^=38 r = 7 A' =37 ,. = 7 R=37 r = l 72 = 36 r = l i2-34 ?'=6 i.e., R = 5| to 5t''. 45 The following is a description of the largest specimen (Figs. 1-4). The plates are all arranged in definite rows ; they are markedly convex, and the axis is transverse. The following series may be distinguished : — One mid-dorsal series ; two sub-lateral series ; one very broad .sub-ventral series. Several of the plates appear as if composed of segregations of smaller dome-like plates. The median row, or mid-dorsal series, terminates in a larger plate at the junction of the arm and disc. Between these five plates, which are arranged pentagonally, five others are disposed symmetrically, while in the centre of the disc there occurs a single plate of a size similar to the others. The arrangement of the tubercles on the surface of the disc is markedly symmetrical. The sub-lateral rows on opposite sides of adjacent arms are continuous through the angle; this series does not extend to the tip of the arm, but is terminated abruptly by the union of two sub-parallel poriferous areas. The plates of the sub-ventral series are more or less rectangular in shape ; they are densely covered with coarse regular granulations which are interrupted by a furrow, running medianly to half-way from the ventral edge of the plate. The poriferous areas lie in six longitudinal rows, three on each side of the arms ; the average number of pores in each group is about fifteen ; the areas are distinctly larger than the corresponding plates. Uu each side of an arm the upper sub-lateral series of plates does not extend to the tip; consequently the poriferous areas which are otherwise separated by these plates are conlluent in this region. Occasionally these larger poriferous areas are united transversely by small and almost inconspicuous lines, bearing pores. The madrepore is simple and distinct, and recalls the structure of the coral Fun'jia. It is larger tlian the plates of the dorsal surface, and lies in the angle formed by the dorsal row of plates and the upper sub-lateral row. The ambulacral armature is bi-serial ; the inner row consists of large and small alternate spines, four or six on each plate (five may occur) ; they are blunt, sub-triangular in shape, and interlock. The spines of the outer series are larger, more distant, vertical, conical, and occur in pairs or threes on each plate. In both series there is a tendency towards an alternation in these numbers — e.g. (1) inner series, four and six on alternate plates ; (2) outer series, two and three on alternate plates ; but from about halfway from the mouth to the tip of the arm the number tends to diminish and the rows lose their regularity. Occasionally a single 46 Fig. 1. Till of arm, dorsal view, sliowing the disposition of the various series of plates and poriferous areas. Fig. 2. Portion of arm, ventral view, showing the structure and arrangement of the adanibulacral armature. LiNCKIA MAr.Mor.ATA. 47 Fio. 3. Portion olaiin, v.-ntnil surfaci-, sliovvin<; tlic a.lanil.ul;u;ral .irmatii Vir.. 1. Oral Region. LiNl'KIA MAKMOItATA. 48 additional spine may occur near the middle of a plate ; this is more marked towards the distal end. The colour fundamentally is brownish-grey, but the centre of the disc is usually purple. There are usually three or four transverse purple bands on each arm ; these extend in some cases to the ambulacral grove. Locality. — Type specimen : Station II., Wamizi Island. Five other specimens : Station I., Tunghi Bay. Nardoa variolata, Gray. Two specimens from the reefs around Matemo Island represent this species. In the largest Il=Qb, r=ll. Neither specimen is fully mature, and the number of plates on the side of each arm does not exceed 30 or .31, whereas in full-grown individuals 45 is not infrequent. Locality. — Station IX., Matemo Island. Previously recorded from — Eed Sea ; Zanzibar ; Bourbbn ; Mauritius ; Mozambique ; Eodriguez : Ceylon ; and Australia. Retaster cribrosus (von Mart.). This species is very abundant on the coast, and specimens from four different localities are represented in the collection. i? = 38, r= 9; K = 4-2r. 7i' = 37, r=13; B=-ir. Most of the specimens are fully grown, and in some of them the arms taper more than is usual in this species. In some individuals the paxillie spinelets on the dorsal surface are not well developed, but even in young specimens they are never entii-ely absent. The colour of the specimens is rather unusual. Most of them are black, with ashy-white ridges on the dorsal surface ; sometimes these ridges are ruddy or even orange-coloured. One specimen has only four arms, and these are arranged symmetrically. Localities. — Station II., Maiyapa Bay (near Magi-mazizi) ; Station III., Mtundo Bay (between Wamizi Is. and Kifuki Is.) ; Station VI., Kero- Nyuni Bay ; Station IX., Matemo Island. Previously recorded from — Samoa and Philippines to Mergui ; Amboina ; Port Molle and Thursday Island, 4 fathoms (coral); Singapore; Ceylon; Mozambique ; Dar-es-Salaam ; and Zanzibar. {Issued .teparalelii. ISWt Febriian/ 1910.) 49 On a New Species of Cactogorgia. By Jas. J. Simpson, M.A., B.Sc, Caruegie Research Fellow, Natural History Dc-ijartment, University of Aberdeen. (With One Plate.) (MS. received January 24, 1910. Read January 24, 1910.) AmoN"GST the unnamed Alcyonaria in the collection of the Royal Scottish Museum, Edinbui-gli. is a beautiful colonj* belonging to the genus Cacto- gorgia, whicli 31r Eagle Clarke has kindly handed me for identification and description. In 1907 {Trans. Roy. Soc. Edin.) I established the genus Cactogorgia for several small colonies from the Indian Ocean, and referred these to three separate species, viz. celosioicles, alciformis, and expansa. Thomson and M'Kinnon, in Trans. Linn. Soc. (ZooL), 1909, have described another species from the Seychelles under the name of Cactogorgia lamjxis, and the present colony must also be referred to a new species. This we propose to name Cactogorgia agariciformis, n. sp., on account of its ver}^ definite mushroom-shape. It is interesting to note that the inclusion of these two new species has not necessitated anj- change in the original generic diagnosis. Cactogorgia agaricifomiis, n. sp. This species is represented by a single specimen of a slightly orange- yellow colour — that is, after prolonged preservation in alcohol. It has been attached to rock, and the basal disc is overgrown by an encrusting sponge. The colonj' (fig. 1) is 75 cm. in height, and consists of two very distinct parts: (1) a lower, almost cylindrical, stalk, 4-8 cm. long, 7 mm. in diameter at the base and 12 nun. at the top ; and (2) an upper, polj-p-bearing, part, elevated in the centre, circular in outline and expanded horizontal!}', giving the whole colon}- a very distinct mushroom appearance. The breadth of the capitulum is 31 mm., and its maximum height 12 mm. The whole colony is very stiff and rigid, owing to tlie densely interlaced, lar<^e, warty spindles, which are quite visible to the naked e3'e. These are arran;.J^ 5. Cactogorgia agahiciformis, n. sp. Mr J. J. SrMPsoN". 51 The record of tlie locality of this specimen has been unfortunately lost. All the other species have been recorded from the Indian Ocean. In some respects this species approaches C'actogorgia expansa, but it is easily distinguished by the characteristic shape of the colony iind by the architecture of the anthocodial armature. The following table gives a summary of the differences in the anthocodial armature for the different species of C'actogorgia : — Species. " Crown.^' " Point." C. celosioides, Simpson. C. expansa, Simpson. C. akiformis, Simpson. C. lampas, Thomson and M'Kinnon. C. agaridfornm, n. sp. 7-10 rows of curved spindles. About 8 rows of curved spindles. 10-14 rows of curved spindles. 6 rows of horizontal spindles. 22-28 rows of slightly curved spindles. 1 large pair, with occasionally 1 or 2 smaller ones between. 6-8 pairs arranged en chevron. 10-15 spindles only slightly en chevron. About 3 converging pairs of spindles, and between two "points" lies a single spindle. 4 pairs of curved spindles en chevron, with a few scattered between the " points." EXPLANATION OF PLATE. Fig. 1. Colony enlarged almost twice natural size, Fig. 2. Stalk broken across to show the large main canals ( x 6). Fig. 3. Polyps enlarged to show anthocodial armature ( x 35). Fig. 4. Spicules from the stalk ( x 25). Fig. 5. Spicules of the verrucre ( x 25). Fig. 6. Spicules of the anthocodise ( x 40). Fig. 7. Spicules from the aboral surface of the tentacles ( x So). (Issued separately March 10, 1910.) i 53 X.— Note on a Peculiar Clutch of Blackbird's Eggs and some other Abnormalities. By Professor J. Arthur Thomson, M.A. (Read 28th February 1910. Received 28th February 1910.) That the eggs of birds show considerable variability — in size, shape, and colouring — is, of course, very well known, but I thought it might be of use to record an interesting case of variation — especially in size — in a clutch of Blackbird's eggs. I owe the clutch and the data as to the order in which the eggs were laid to my friend Mr R. Hay Fenton, who not long ago very generously presented to the Natural History Museum of Aberdeen University his remarkably fine and almost complete collection of the eggs of British birds. There are four eggs in this Blackbird's clutch, which have the following dimensions — (i.) 38 X 24 mm. (ii.) 33 X 23 mm. (iii.) 28 X 22-5 mm. (iv.) 27 X 21 mm. The largest egg (i.) was laid first, but it did not develop; the ne.xt largest (ii.) followed and was fertile; the third (iii.) was near the normal size; the smallest egg (iv.) was laid last. The third and fourth were quite fresh when the clutch was taken. In his great work on the eggs of the birds of Europe, Dresser notes that the eggs of the Blackbird vary from 28 x 2133 mm. to 32 x 22 35 mm., and quotes as an average of 48 eggs, 28 x 21 mm. ; so that the interest of the clutch exhibited is that the first egg is far above the average, and that the abnormality goes on decreasing egg after egg. There is a progressive variation in a minus direction. I do not wish to make too much of a little thing, but it is of interest to compare a case like this with others where we see at work the mysterious regulative tendency which is characteristic of living creatures. Thus Raymond Pearl recently described a case (Journ. Uxper. ZooL, vi., 1909, pp. 339-351, 1 pi.) in which the first egg laid by a pullet was very abnormal in shape — elongated ovate pyriform. In the subsequent eggs laid there was a quite gradual change of shape, whicli was regulative in character, until finally the eggs were quite normal. Numerous abnormalities in the eggs of the common fowl have been recorded, and the collection now shown illustrates not a few. It is an 54 object-lesson in variability. The most remarkable form which I have to show consists of a small oval giving origin to a twisted tube like the horn of a shorthorn sheep. Such occurrences have been in some measure cleared up by a recent daring experiment made by Pearl and Surface {Science, xxix., 1909, pp. 428-9). In order to determine the nature of the stimulus which induces the making of a shell, they performed an operation on a hen as a result of which the contents of the intestine were made to pass through the shell-secreting part of the oviduct. The interesting result was, that they got curious enshelled masses of various shapes, and they were led to the conclusion that the stimulus which excites the shell-making glands is mechanical rather than chemical in nature, and that the formation of the shell is brought about by a strictly local reflex, and is not immediately dependent upon the activity of other portions of the reproductive system. It may be said that this was to a certain extent known before from a study of what are popularly known as " wind-eggs." These are not true ova, they contain no vitellus. They consist of a mass of albumen the stimulus of which has induced the making of an enveloping shell. It has also been shown that foreign bodies may ascend from the cloaca into the oviduct and become surrounded by a shell. Thus a species of Distomum, which frequently occurs in the bursa Fabricii opening into the cloaca, may pass up the oviduct and be included intact in the albumen of an egg. In an interesting paper on abnormal eggs in fowls, J. Kunstler points out {M6m. Soc. Sci. Bordeaux, in., 1903, pp. 65-72, 7 figs.) that a frequent cause is a lack of tone in the oviduct, the normal peristaltic movements being disturbed in consequence. Thus an egg may return on its path and become surrounded by a second shell, thus resulting in one form of tlie not unfamiliar ovum in ovo. In a paper on " Ovum in Ovo," by F. H. Herrick {American Naturalist, xxxiii., 1899, pp. 409-414, 3 figs.), it is suggested that an abortive egg or egg-fragment may be enshelled and then included within the shell of another egg of larger size. In other cases, what is included has nothing of the nature of an egg about it, though it has been enshelled. Entirely different again are eggs with double or triple yolks, where we. have to deal with a fusion of the albumen in two or more ova, which are treated in the oviduct as one egg and surrounded by a single shell. This process may be sometimes complicated by the inclusion of a third egg of normal size and already covered by a hard shell. So when we speak of an egg within an egg, we may mean one of three or four different things. The problem of the factors which determine the shapes of the eggs of birds has been discussed in a very interesting paper by Prof. D'Arcy W. I 55 Thompson {Nature, 4th June 1908, pp. 111-113). The egg, consisting of a slightly extensible membrane filled with an incompressible fluid is subject to external pressure from the radially contractile oviduct, and an equation for the shell can be worked out. It is pointed out that from the nature and direction of the usual peristaltic wave in the oviduct the pressure will be greatest somewhere behind the middle of the egg ; in other words, the tube is converted for the time being into a more conical form, and the simple result follows that the anterior end of the egg becomes the broader and the posterior the narrower. One may recall how the peristaltic move- ments of the intestine in many animals, such as the rabbit, divide the faecal matter into spherical or oval masses. The object of the present note is to record a case of variation with subsequent regulation, to show in a collection of hen's eggs how large the crop of variations is, and to note the suggestions that have been offered in interpretations of certain not uncommon freaks, such as a trumpet-shaped egg or an egg within an egg. (Issued separately, Wth May 1910.) 57 XI. — Note on Eunephthya glomerata, Verrill, from the Faeroe Channel. By Professor J. Arthur Thomson, M.A. (Read 28th March 1910. Received 28th March 1910.) I HAVE previously recorded in the Proceedings of this Society the occurrence of two interesting Anthozoa from the Faeroe Channel, Primnoa reseda, Linn., and Antipathes larix, Esper, and as it seems worth while continuing, the list as specimens turn up, I wish now to add Eunephthya glomerata, Verrill. I obtained a single specimen from the collections of the " Goldseeker," and I am indebted to Prof. D'Arcy W. Thompson, C.B., for permission to make it the subject of this note. The small colony rises to a height of about 22 mm. with a maximum breadth of 15 mm. A short trunk bears about three dozen polyps in ill- defined groups. There are so many spicules that the whole colony is stiff, and the surface has a glistening, frosted appearance — more or less white in colour. Most of the polyps are 3 to 4 mm. in height, and the upper part shows eight strongly-marked longitudinal ridges and intervening grooves. The ridges taper, bend inwards round the mouth, and are continued into the inturned tentacles. These are curved like a crozier, and their aboral surface is heavily armoured with spicules which cease near the tip. The spicules are clubs and spindles, some compact irregular bodies and a few cruciate forms. The clubs vary greatly in size and appearance, but most of them have enlarged heads, extremely rougli, with stout blunt processes usually directed towards the apex. The following measurements were taken of length and breadth — the latter measured across the club, 0-15 x 006; 0"2 X 0'04 ; 0"3 x O'l mm. The spindles have usually sparse, blunt tubercles. Two common sizes are — 04 x 004 ; 08 x 002 mm. The large and somewhat difficult genus Bunephthi/a, as re-defined by Kiikenthal, includes more or less branched Nephthyid colonies, with thin canal walls not thickly filled with spicules, with polyps without Stiitzbiindel, in groups or singly, retractile or non-retractile, but not divided into distinct calycine and retractile portions. The nearest genus is Gersemia, in which a portion of the polyp is retractile within a distinct calyx. The genus Eunephthya is divided into two main divisions — (1) Alq/oni/orm, with thickened contractile terminal branches, with completely retractile, elongated cylindrical polyps disposed singly ; and (2) Nephthyi/omi, with unthickened, non-retractile terminal bi-anches, with polyps grouped in bundles, club-shaped, and non-retractile. 58 It is evident that our specimen belongs to the Nephthyiform division or sub-genus of Eimephtluja, and to the group which Kiikenthal calls Divaricatse wlomeratffi, including E. glomeraia, Verrill, E. hyaliTia (Danielssen), E. racemosa, Studer, and some doubtful species. I have recorded this specimen for faunistie reasons — it is a characteristic- ally boreal animal previously recorded from deep water of 500 to 600 I'athoms in Northern Seas, e.ff., o.ff Franz-Joseph Land, Spitzbergen, Jan Meyen, and Greenland. It was dredged in the Fseroe Channel, Station 18a, 18th June 1906, from a depth of 355 metres. But the specimen has also some systematic interest. My specimen agrees with Eunephthya glomerata in all respects except that it shows more numerous delicate spindles, more irregularly headed clubs, and among its irregular spicules none that I should call double stars. By making several distinct preparations from different parts of the colony, I have convinced myself that the slight divergence in the spiculation is quite unimportant, and not greater than that separating some other recorded varieties of the species. I strongly suspect that many of the spicules described and figured as double stars are the broken off heads of very much knobbed clubs. In appearance, and in some of its spicules, my specimen very closely resembles Eunephthya racemo&a, Studer, from Newfoundland, and may, I think, be taken as showing that the latter should be included in E. glomerata, Verrill. Kiikenthal has already united with Eunephthya glmiierata quite a number of species — Ammothea lueikeni, Mareuzeller, Nephthya polaris, Nephthya Jlavescens, Nephthya rosea, Drifa islandica, and Gersemiopsis arctica (all due to Danielssen's exaggerated emphasis on minute differences). 1 think E. racemosa, Studer, should be added to the list. Moreover, in the detailed description and numerous figures that Danielssen has given of Drifa hyalina (which, as Kiikenthal shows, is certainly a species of Eunephthya), I find no satisfactory reason for keeping even it apart. Danielssen noted that Nephthya Jlavescens, N. polaris, and N. rosea, which are now iJentihed with Eunephthya glomerata, are viviparous, and he gave a careful account of the embryos found free in the gullet. It is interesting to notice that the specimen under consideration, collected in the month of June, has numerous large free embryos in some of its polyps. Perhaps, as in some other cases, the viviparous habit is au adaptation to life in deep water. [Issued separately, llth May 1910.) (JouRN. R. MiCR. Soc, 1910, pp. U2-113) 59 V. — Note, on Dendrohnwhla fallax Brook, a Rare and Scmarkahk Antipatharian. By Professor J. Arthur Thomson, M.A. (Bead Fcbmarij 16, 1910.) In a collcctidii of Alcyoii;irian.s wliicli 1 received for description from His Soreue Highness the Prince of Monaco, there were four specimens of unusual appearance and puzzling character — with a spinose axis and pinnate tentacles — wliich are undoubtedly refer- able to a remarkable type of Antipatharian, which llrook described in ]8S'.I under the name Dcadrohrachia fallax. Although I have not much that is new to add to Brook's excellent descrijjtion and figures, it may be of interest to record the re-observation of an ex- traordinary type, which seems to have remained unnoticed for more than twenty years. Very unfortunately, three of the specimens were dry, while the fourlli, which was preserved inspirit, had only a few extremely brittle polyps. What are the peculiarities of the type wliich give it a somewhat aloof position among Antipatharians ? 1. The axis is without a central canal. In its younger parts it consists of about five longitudinal ridges or plates, standing out from a thin central stem and showing a distantly dentate margin. As growth goes on, there seems to be an increase in the number of outstanding ridges, and at the same time, by the deposition of successive concentric layers of horn, the deep troughs between the ridges are filled up, and an apijroximation to the ordinary type of Antipatharian axis is thus reached. There is great diversity in the size and .shape of the spines in different parts of the colony. The colour of the axis varies from yellowish-brown to amber-brown. The basal portion, which was ab.sent in Brook's two specimens, is well seen in two cases. 2. The iiolyps are even more remarkable than the axis. They arise laterally, often in sub-opposite pairs, but there may be a con- sidcralile interval, of two lengths or more, between two pairs. In short, they are much more distant than is usual in Anti]iathariaijs. Nor do they, in most cases, stand out at right angles, as Antipa- tharian polyjjs usually do ; they are often appressed to the twig, or form an acute angh> with it. The tentacles are retractile, which is also unusual, so tliat in some cases there is simply a circle of knobs around the prominent oral cone. StUl more striking is the fact that they bear well-developed pinnules, six to seven pairs in the 60 Note on Denclrolrachia fallax Brook. By J. A. Thomson. twelve polyps examined. It was impossible to cut the extremely 1 irittle tissue, and the tentacles split veiy readily up the middle, but in foui- or five cases there seemed no doubt as to the presence of eight. Brook was not able to determine the number. In one case it seemed faii'ly certain that there- were eight mesenteries. The occurrence of eight pinnate tentacles at once suggests an Alcyouariau, and one was reminded that in many cases Alcyouid Alcyonarians grow over the naked axes of Gorgonids so thoroughly that a very deceptive appearance results. There is not, however, anything — such as uncovered tips on the twigs, or hummocking of the ccenenchyma — to lead one to suppose that the specimen is not a iinity. Moreover, there is no hint of Alcyonarian spicules, and the pinnules on the tentacles are much more irregular than in Alcyonarians. One would not indeed consider this possibility, were it not that some cases of the masking of extrinsic axes by Alcyonarians are almost incredilaly deceptive, and were it not tliat the type in question is such a puzzling Antipatharian. That Antipathariaus have affinities with Zoantharians seems probable, and it is interesting to remember that the primitive type Edivardsia has only eight complete mesenteries, that a young Hakampa has eight tentacles, and that some sea-anemones, e.g. Adinodcndron, liave irregularly pinnate tentacles. There seems to be little doubt that Dcndrohrachia fallax is a primitive Anti- patharian, nearer than many to the Zoantharian stock. One would like to see more of it. Brook's two specimens were dredged in 1876 from 425 fathoms off Ascension ; those now exhibited were dredged in 1901 from 219 fathoms off the island of St. Vincent, in the Capj Verde Islands. LOsn.X: PRIXTIJ. Ur «-,I.,.tAM CLOWES *XD SONS. LmrTED OR. AT WCNDMILI. SIEKIT. W., ASI> DUKE STREET. STAMFOKD 6TEEM, JOURN. R. MICRO. SOC. 1910, PI. V. West. Newman imp. INDIAN OCEAN ANTI PATH ARIANS. Li (JouRX. R. MiCR. Soc. 1910. pp. 273-281.) VIII. — Antipatharian.s from the Indian Ocean. By Sophia L. M. Summers, M.A. B.Sc. (Natural History Department, University of Aberdeen.) {Read March 16, 1910.) (Plate V.) The Antipatharians here reported on were collected for the most part at Ibo, in Portuguese East Africa, by Mr. Jas. Siiupsou, M.A. B.Sc, Carnegie Fellow, University of Aberdeen. A few were collected in the Mergui Archipelago by !Mr. R. X. Rudmose- Brown, B.Sc, and Mi: Simpson. The collection includes fourteen species, of which three areaew. The list is as follows. EXPLANATION OF PLATE V. Fig. 1. — Antipathes abies Gray. Arrangement of spines, main axis. „ 2. — A. salicoides sp.n. Arrangement of spines near the tip of the axis. „ 3. — Pteropathes simpsoni s,p. u. Arrangement of spines. „ 4. — Antipathes salicoides s'p.n. Arrangement of spines, main item. „ 5. — Stichopat)ies bispinosa noxn.n. Arrangement of spines. „ 6. — Antipathes abies Gray. Showing forked spines. „ 7. — A. salicoides sp. n. Polyps on main branches. „ 8. — A. tristis Duchassaing. .\rrangement of spines. „ 9. — Cirripatlus iiulica sp. n. Arrangement of spines. „ 10. — Stichopathcs bispinosa. Polj-ps. „ 11. — S. echinulata. Polyps. 62 Transactions of the Society. Family ANTIPATHID.E Yerrill. Siib-faraily Antipathix-e Brook. Section 1. Indivisse. Genus Cirripathes Blainville. Cirripathes indica sp. n. C. pro/diiqua Brook. C. spiralis Linn. Genus SHchopathes Brook. Stichopathcs (^z'rma. (Brook) = Cirnpatlics{^)diversa^ro6k ■=. (?) S. alcocki Cooper. ,S^. Uspiiiosa nom. n. = Cirripathes (1) Jlagellum Brook ; n. = Stichopathes flagellum Eoule. S. echiimlata Brook. S. gracilis Gray. Section 2. Ramosse. Genus Antipathes Pallas. Antipathes salicoides sp. n. A. fiircata Gray. A. tristis Duchassaing. A. lentipinna Brook. A. gracilis Gray. A. abies Brook. Genus Pteropathes Brook. Pteropathes simpsoni sp. n. Cirripathes indica sp. n. Plate V. fig. 9. See C^■rrlpaart. Locality.— Toriwguese East Africa. Previously recorded from the Indian Ocean, Jlolucca, Norway, MediteiTauean, Ceylon, Kurrachee, and the East Indies. Stichopathes diversa (Brook). See CirHpathes (?i di>-ersa Brook, Report on the Antipatharia of the 'Challenger' Expedition, p. 87, pi. 12, lig. 12. A stout spiral, 15 cm. high. The polyps are arranged on the outer side, while the inner is left bare. Tiiey are straw-coloured in spirit and are very flat, with long tentacles and prominent mouth. There are two types of spines, large ones with small ones between 64 Transactions of the Society. them. The large ones are blunt and rough, and those on the outer side of the spu-al are much longer than those on the inner. The small or secondary spines are short, sharp, and triangular. Cooper has described (1909) a new species, Stichopathes alcocki, but his de- scription of it corresponds with Brook's description of Cirripathes diversa, e.g. in the an-angement and form of the spines. Brook could not decide whether his specimen belonged to Cirripathes or Stichopathes, as it was quite devoid of polyps. Cooper remarks on the large size of the central canal. It is probable that S. alcocki Cooper should be united with S. diversa (Brook). The canal is well seen in this specimen, the wall of the corallum being comparatively thin. Locality. — Portuguese East Africa. Previously recorded from Galle, Ceylon. Stichopathes lispinosa norn. n. Plate Y. figs. 5, 10. = C'in-ipathes (1) JIageJhim Brook non = Stichopathes JlageUum Koule. See Brook, Report on the Antipatharia of tbe ' Challenger ' Expedition, p. 87, pi. xii fig. 13. The longer of the two specimens is 130 cm. in length, and is twisted into large spirals with a diameter of 21 cm. The axis is stout and sinuous, and tapers gradually. The diameter at the base is 7 mm. It is covered with thin ccenenchyma, through which the spines project. The polyps are confined to one side of the stem, and are quite different from those of other species. They are flat, with a small mouth overlapjjed by two of the tentacles. A little to the side are two other tentacles, and pushed down the side of the axis is a thu'd pan-. At first sight one would believe the polyps to have eight tentacles. So closely are the mouths situated, that it is difficult to discover to which polyp the tentacles belong. The tentacles are short and thick. The spines are arranged in longi- tudinal rows, nine of which can be seen from one aspect. They are veiy large (about Q-o mm.), and they gi-adually decrease until exactly opposite the longest spines are the shortest (about 0 • 1 mm.). There are secondary spines scattered between the rows. The spines are distinctly papUlose at the tip. Members of a row are about a length apart. The central canal is very large, having a diameter of 1 • 5 mm. The an-angement of the spines and the general characters of the specimen correspond exactly with Brook's description of Cirripaihes {'>) flagellum. As Brook's specimen had no poh'ps, he could not decide whether it should be referred to Stichopathes or to Cinipathes. Zocality.—Fonngixe&e East Airica. Previously recorded from Ceylon. Antipatharians. By Sophia L. M. Summers. 65 Stichopathes echinulata Brook. Plate V. fisj. 11. See Brook, Report on the Antipatbaria of the ' Challenger' Expedition, p. 92, pi. xii. fig. 9. Of several specimens, the longest is 95 cm. It is twisted in irregular spirals. The axis gradually tapers to a fine point. The diameter at the base is 1 mm., and of the central canal 0-5 mm. The polyps are large and crowded on one side of the axis. In some of the specimens they show a peculiar mode of growth, being bunched together at intervals in a sort of hummock. Tlie longest diameter of the polyp is 3 mm. The mouth is raised on a prominence 1 mm. in height. The tentacles are long and thin. The spines are numerous, and are arranged in steep spu-als. Of the vertical longitudinal rows of the spiral, nine can be seen from one aspect. In one part of the stem the spines are short, tri- angular, and much compressed ; at another part they are much longer. Members of a row are about three lengths apart. Localitij. — Portuguese East Africa. Previously recorded from Mauritius. Stichopathes gracilis Gray. See Brook, Report on the Antijiathaiia of the 'Challenger' Expedition, p. 90, pi. xii. figs. 17-19. Two fragments, one 26 5 cm., the other 8 cm. The larger specimen is somewhat spirally coiled. The axis is black and tapering. At the base it is 2 mm. in diameter, at the tip 1 mm. The polyps are on one side of the axis, and are large and circular (2 mm. in diameter) ; the tentacles are short and thick ; the mouth is prominent. The spines vary very much in diflerent portions of the axis. In slender portions of the stem the spines are arranged spirally and in longitudinal rows. They are triangular and com- pressed and stand at riglit angles to the axis. In thicker jiortions the arrangement is less regular. Most are simple and have a sharp apex, but a few are forked at the tip. Localitij. — Portuguese East Africa. Previously recorded fi-om Fiji, Ked Sea, Seychelles, Ceylon. Antipathes salicoides sp. n. Plate V. figs. 2, 4, 7. The whole colony is like a miniature weeping-willow. The main stem, which seems to lie broken, is only 10-5 cm. iiigh. It is 2 mm. in diameter at the base, and gives ofl" long tapering branches from one side, the first of which reaches a length of 32 "5 cm. Some of these branches remain undivided, wliile others give off long slender twigs. The wliole specimen is densely 66 Transactions of the Society. covered with polyps, which are arranged alternately on the branches. They are large and circular; the mouth is prominent; the tentacles are short and thick— the two transverse ones lying a little below the level of the others. The diameter of the largest polyp is about 2 mm. The spines on the main stem are crowded, and do not all point in one direction — an unusual feature. They are arranged neither in rows nor in spirals. On the branches the spines are short, sharp, and triangular. They are arranged in longitudinal rows, seven of which can be counted from one aspect. The members of a row are about one length apart. Locality. — Portuguese East Africa. Antipathes f areata Gray. See Antipathes (^)furcata Brook, in Keport on the Antipatharia of the ' Challenger ' Expedition, p. 104, pi. si. tig. 2. The main stem is broken, and is only 5 cm. in length. It gives rise to a branch which reaches the length of 8 cm., and is then broken. It in turn gives rise to a branch 23 cm. in length. The axis of this branch tapers veiy markedly, and gives off delicate branches on all sides and in no definite order. These are mostly directed upwards. The polyps, which are badly preserved, are elongated in the du-ection of the longitudinal diameter of the axis, and are twice as long as they are broad. They are about 1 mm. in length. They lie close together, but are not crowded. The tentacles are short. The spines are short, triangular, and com- pressed, and are far apart. They are arranged in longitudinal rows, six of which can be seen from one aspect. Locality. — Portuguese East Africa. Previously recorded from IMadeii-a. Antipathes iristis Duchassaing. Plate Y. fig. 8. See Antipathella (?) tristis Brook, Report on the Antipatharia of the 'Challenger' Espeditiou, p. 111. A small delicate colony 5 cm. high. The stem is slender, and gives off delicate branches ii-regularly at right angles to the stem. Anastomoses occur, but are not frequent. The polyps are situated on one side of the axis. They are small, and have short digiti- form tentacles. The polyps are about 1 mm. apart. The spines are sharp and triangular, and are arranged in irregular longitudinal rows, six of whicli can be seen from one aspect. Members of a row are about two lengths apart. Locality. — Portugue.se East Africa. Previously recorded from Guadeloupe, Santa Cruz, ]\Iontserrat, Martinique, St. Lucia, and Barbadoes. Aniipatharimis. Bij Sophia L. M. Summers. 67 Antipaihes lentipinna Brook. See Antipathes (?) hntipinna Brook, Report on the Antipatharia of the 'Challenger' Expedition, p. 103, pi. xi. fig. 19. A shrub -like colony 28 cm. liii,'li, densely branched. The main axis is black and hard, 7 mm. in diameter. In the branches and pinnules the axis is light brown. The polyps are arranged on one side of the axis, and are large (2 mm. in diameter) and crowded together. Here and there small polyjjs are crowded in between the large ones. Branches are given off, usually from one side only. On the main stem the arrangement of spines is very irregular. The axis is flattened out, and the spines are scattered all over it, but on the branches they are arranged in regular spirals and longitudinal rows, five of wliich can be seen from one aspect. The spines are long and triangular, and are very close together. Locality. — Portuguese East Africa. Previously recorded from the Eed Sea. Antipathes gracilis Gray. See AntipathiUa (?) gracilis Brook, Eeport on the Antipatharia of the ' Challenger ' Expedition, p. 113, pi. xi. fig. 8. Two specimens, one 8 • 5 cm. in height, and the other a little over 9 cm. The latter is a delicate colony, with no regular mode of branching. The branches are thin and short, and are given off at right angles. Anastomoses are frequent, but the terminal fronds are free. The polyps are crowded together on one side of the axis. They are circular, 1 ■ 5 mm. in diameter. The tentacles are short and thick, one pair lying sligiitly below the level of the other two pairs. The spines are long and numerous, and are arranged in spirals and longitudinal rows, five of which can be seen from one aspect. Members of a mw arc from two to three lengths apart. Locality. — Portuguese East Africa. Previously i-ecorded from West Indies and Madeira. Antipatlies ahies Gray. Plate V. figs. 1, 6. See Antipathes ahies Brook, Eeport on the .\ntipatlinria of the ' Challenger' ExiKtlition, p. 70, pi. xi. tig. 21. The four specimens show a remarkable difference in the mode of growth. The first specimen, from Five Islands, con.sists of two fragments, mucli-branched and bearuig many polyps. Some of the polyps are very well preserved, and show a prominent mouth raised G8 Transactions of the Society. on a cone, and long tentacles. The polyps are arranged close to- gether on one side of the stem, and are relatively large. The ccenenchyma is thin, and the spines can be seen projecting through it. The spines are very numerous. On the lower part of the stem they are sharp and needle-like, and stand out at right angles to the axis, but on the branches they are small and blunt, and lean towards the axis. The second specimen is 16 cm. in length. The mode of brandling is the same as in the first case. The branches are given off all round. The main axis is thick. The polyps are situated on one side of the axis, and are crowded. They are 0 ■ 5 mm. in the longest diameter. The mouth is slit-like and prominent. The tentacles are much contracted, and are placed in three paii'S round the mouth. The spines are exactly of tlie same type as those of the first specimen. Locality. — Portuguese East Africa. The third specimen branches in one plane and is fan-like. Tlie main axis is thick and flattened out at the base, and gives off alternate branches. The polyps are similar to those of the second specimen, but are far apart, being separated on some parts of the axis by intervals of fully 0 ■ 5 mm. The spines are of tlie same type as in the previous specimens, but are not so long on the stem. Locality. — Portuguese East Africa. The fourth specimen is from Bentinck Island, Mergui. It is similar to the last specimen in being fan-like, but the branches are smaller and more delicate. It is 12 • 5 cm high. The polyps are crowded together on the liranches, and the largest ha\'e a diameter of 0 • 5 mm. The ccenenchyma is very thin. The spines agree with those of the other specimens, but the needle-like spines on the axis are very numerous. It seems that Antipathes abies is a very variable species. Pteropathes siuypsoni sp. n. Plate V. fig. 3. A small colony, 10 ■ .5 cm. in length. The branching is irregular, and there seems to be no main axis. The whole specimen is clothed on one side with whitish polyps, which give it the appearance of being covered with a mould. These polyps are unfortunately so badly preserved that their structure cannot be made out, though in some parts long tentacles are seen. The spines, which are large and numerous, are triangular with a broad base. They are an-anged in very steep spirals ; seven or eight can be seen from one aspect. Towards the tip of the axis the spirals become irregular. The spines also form longitudinal vertical rows, the members of which are about a lengtli apart, and in some places even less Locality. — Mergui Archipelago. Antipatharians. By Sophia L. M. Summers. 69 Literature referred to— ix Addition to that Xoted by Brook (1889) and Roule (1904). 1889. Brook, G. — Antipatharia. ' Clialleiijier ' Reiwrt, .xxxii. 222 pp , 15 pl.s. 1903. Cooper, C. — Antipatharia. In Gardiner's Fauna and Geography of the Maldive and Laccadive ArchipeLigoes, ii. pp. 791-6, 1 pi. 1903. EouiiE, L. — La distrilaition bathvmetrique des Antipathaires C.B. Ass. Franc, 31 Sess., p. 236. 1904. EocLE, L. — Etsultats Gomp. Sci., par Albert I. de Monaco, Fasc. 30, pp. 6-99, 10 pis. 1905. Thomson, J. AETHtrR —Scotia Collections, Scottish Antarctic Expe- dition. Eeport on the Antipatharians. Proc. E)y. Phys Soc. Edinburgh, xvi. pp. 76-9. 1905. Thomson, J. Abthdk, .V J. J. Simpson — Report on the Antipatharia, in Herdman's Ceylon Pearl Oyster Fishery Bt-ports, Eoyal Society, iv. pp. 93-106, 1 pi. 1907. Eoui.E, L.— Sur la valeur morphologique des upines du polypier des Antipathaires. Comptes Eendus, 1907, pp. 1533-4. 1907. Thomson, J. Arthur— Note on a large Antipitharian from the Faroes. Proc. Eoy. Phys. Soc. Edinburgh, xvii., 1 pi. 1907. HiCKSON, S. J. — Alcyonaria, Antipatharia, and Madreporaiia collected by the ' Huxley ' from the north side of the Bay of Biscay, in August 1906. Journ. Marine Biol. Assoc, viii. No. 1. 1909. Cooper, C— Antipatharia of the Percy Sladen Trust Expedition to the Indian Ocean. Trans. Linn. Soc London (Zool.) vii. pt. 4, pp. 301-21, 1 pi. 1909. SiLBBBFELD, E— Japanisclie Antipatbarien. Abhandl. der Math.- Phys. Kl. der k. Bayer. Akad. Wiss., i. Supp. 1, BJ. 7, Abliandlg. pp. 4-30, 2 pis. lOSDnS : PRlNTtn »T WILLIiM CLOKIS ASP lOSS. I.lliirKD OEMI wmDMlLL STEEET, W.. iHP DOK« STIREKT. CT.MrolU. miEET. fcE. 7 11 From ihf Annat.s and IMagazine of Natural History, Ser. 8, Vol. iii.,^;«-!7 1909. Alcyonarians from the Gulf of Cuich. By Prof. J. Arthur Thomson and Mr. George Crane, B.Sc, University of Aberdeen. (Preliminary Note.) In the course of an investigation of tlie sliallow-water fauna of part of the Gulf of Cutch, Mr. James Hornell made a small collection of Alcyonarians which presents some features of interest. The precise district was the coast of Okharaandal, which forms the N.W. extremity of the Kattiawar Peninsula, and Mr. Hornell has called our attention to tiie fact that specimens of Dendronejihtlnja (heltcr known as Spongodes), of Lophoffort/ia, &c. could he collected at low tide. The collection includes eight species, one of which — Aslro- muricea stdlifera — is new. There is also a new variety of a remarkable species of Echinomuricea previously found in the Indian Ocean. The position of the various species may be indicated as follows : — Order ALCYONACEA. Family Alcyonid.* .... (1) Sclerophytum poh/dactylum (Ehrcn- berg). Family Nkphthyidje . . (2) Dendronephthi/a (SpomjoJes) dendro- phi/ia (Wright nud Stiider). (3) Dindrimejyhthya (iypungodes) bre:i- rtima lUurckbiLrdt). [ 362 ] Prof. J. A. Thomson atid Mr. G. Crane on Oi-der AXIFERA. Family Muriceid^ (4) Astromuricea stelHfera, sp. n. (5) Echinomuricea nliginosa, Thomson and Simpson, var. tenerior, noT. Family GorGONID^ .... (6) Lophogorgia lidkem, Wright and Studer. (7) Juncdla juncea, Pallas. Order STELECHOTOKEA, Section Pesnatulacea. Family V1RGULAEID.S . . (8) Virgularia rumphii, Kolliker. (1) Sclerophytum polydactylum (Ehrenberg) is a well- known widespread species, previouslj reported from the Red Sea, Maldives, Gulf of Manaar, China Sea, Zanzibar, British New Guinea. It is characterized by the absence of siphono- zooids, the small size of the autozooids, and the tough fleshy texture. The specimens from the Gulf of Cutch were large, the maximum dimensions being 5 cm. in height by 14 in length and 8 in breadth. (2) Dendronephthjfa [Spongodes) dendrophyta (Wright and Studer), a species of the flattened umbellate type in Kuken- thal's dendrophyta group, previously recorded from Philip- pines and China Sea. It is represented by looselv branched and close-set types of polyparium, as figured by Wright and Studer and by Kiikenthal respectively ; the anthocodi^ show the characteristic eight double rows of curved spicules, 4 or 5 in each row ; a trivial feature, noted by Wright and Studer, namely the occurrence of numerous superficial x -shaped spicules on the branches, is very marked. The specimens were collected in the month of December, and they show abundant reproductive bodies — probably sperm-sacs — up to 0*25 mm. in diameter, attached to the mesenteric bands far below the polyp-stalks. Some specimens show a few small polyp-bearing twigs on the top of the stalk below the foliate branches. (3) Dendronephthya (Spongodes) hreviravia (Burckhardf), a species of the flattened umbellate type in Kiikenthal's forida group, previously recorded from China Sea and Torres Strait. A peculiarly fine specimen has a polyparium 12-5 cm. in height, with diameters of 10-5 cm. and Dcm., with a very short stalk 1 cm. in height, and root-like attachments of about 6 cm. The anthocodiaj show the characteristic eight double rows of spindles in chevron, with 5-7 in each row, [ 363 ] n Alcyonarians fi-om the Gulf of Culch. the uppermost projecting sliglitly. A feature of some interest on several specimens is tlie occurrence of a number of small twigs on the siiort stem portion almost tlown to the level of the stolons, eacli twig bearing two or three polyps. (4) Astromuricea stelUfera, sp. n. — A reddish, fan-like, flexible olony (l-i cm. in heigiit by 28 cm. in breadth in maximum dimensions) with very abundant anastomosis. The axis is dark glossy brown and almost smooth. The coenen- chyma is very rough. The verrucae are crowded on ail sides of the axis; they are cylindrical and their apex is frinq:ed by about a dozen projecting spicules. Tiie antiiocodire are completely retractile witiiin tlie verriicaa ; there is a low, almost iiorizontai, tentacular operculum ; two colourless converging spindles lie on tlie aboral surface of each tentacle, and there is a single or double ring at the base of the tentacles; other- wise there seem to be no spicules in the |)olyps. The spicules of the coenenchyma are (1) irregnlar warty stars and tootiied plates, (2) stout spindles with tuberculate warts, and (3) small irregular bodies — all of a rose-red colour. This species differs from the other members of the genus in many details, e. 1 1910.] ON HTDR0ID3 FROM MERGUI ARCHIPELAGO. 799 [^From the PROCEEBIJfGS OF THE ZoOLOGIOAi SOCIETY OF LONDON, 1910.] rpubiigijed October 1910.] Tlie Marine Fauna of the Mergui Archipelago, Lower Burma, collected by Jas. J. Simpson, M.A., B.Sc, and R. N. Rudmose-Brown, D.Sc, University of Aberdeen, February to May 1907. — The Hydroids. By James Ritchie, M.A., B.Sc, Natural History Department, The Royal Scottish Museum *. (Plates LXXVI. k LXXVII.t, and Test-fig. 79.) The Hydroids were reprosented in the collections brought together in the Mergui Archipelago by Dr. John Anderson in 1882, and described in the Journal of tlie Linnean Society for 1889, by a meagre total of six species, two of which were regarded by Hincks as new. The reasonable anticipation that tiie careful collecting of Dr. Brown and Mr. Simpson would add many species to the Hydroid fauna of this region of the Indian Ocean has been amply fulfilled ; for the present collection, confined to a littoral area, contains representatives of thirty species, only three of which {Campanidaria raridentata and Idia prisiis, both widely distriljuted species, and Aglaophenia crispata, a synonym of Lytocarpus pen- 7iarius,Jide Billard, 1909, p. 329) were included in llincks's list. As only two of the species, being described as new to science, must be regarded as peculiar to the Archipelago, the collection is well fitted to show the relationship which the Hydroid fauna bears to that of neighbouring seas. Leaving out of account an luidetermined species of Plumidaria, nineteen of tlie Mei-gui species have already been recorded from the Indian Ocean, the majority of these occurring oil' Madagascar anil tlie eastern shores of Africa. The remaining ten species, indicated by • Communicated by R. KiKErATRtCE, F.Z.S. + For explanation of the I'lates sec p. B25. [IJ 800 MR. J. RITCHIE ON HYDEOIDS [^f'lj 2-i, asterisks in the following list, have thus been added to the fauna of the Indian Ocean. As to the wider relationships : — Three species are peculiar to the Indian Ocean {Corydendrium. sessile, Hehella crateroides, CalyceUa oligista), leaving twenty- six to be accounted for. Of these, twelve are so widely distributed that tlieir occui-rence, emphasizing the normal nature of the fauna, can give no indication of special affinity : three Ijeing found in the Atlantic and Xorth Pacific Oceans and in Aus- tralasian seas (Opercidwell a lacerta, Plvjmdaria setacea, Anienella secundaria) ; two in the Atlantic and North Pacific Oceans {Ferigonimusrepens, Campanulariararidentaia) ; and seven in the Atlantic Ocean and in Malay- Australian seas {Pennaria disticha, Haleciuni tenellimi, Canipanularia corrugata, Hehella calcarata, ^ertularella polyzonias, Idia pristis, Diphasia digitalis). Of the remaining fourteen, four have been found only in the Atlantic Ocean (Eiulendrium aUemiattim, Cuspidella costaia, Lafoea serraia, Lafo'ea venusta), seven in the Malay- Australian area {Eudendrium generalis, Cladocoryne haddoni, Halecitmi sitnplex, ThyroscypMm vitiensis, Sertularella cylindrica, Sertularella quadrideiis, Sertularia turbinata), one in the ISTorth Pacific Ocean ( Thyroscyphiis regularis), and two in both the last-named areas {Lytocarpus ]3ennarius and L. phoeniceus). The closer relationship is obviouslv with the fauna of the Pacific Ocean (represented by ten distinctive species), and in particular with the Malay-Australian portion of that ocean (represented by seven species). The significance of the apparent Atlantic affinity is minimised when it is understood that of the four species common to it and the Mergui Archipelago, one (Eudendrium attenuatum) is identified with much doubt, while the remaining three are very minute epizoic species, the presence of which in other regions may conceivabh' have been overlooked. It is i-emarkable that of the thiity species recorded, seven- teen were found climbing upon larger Hydroids and one upon a Polyzoon, the majority being minute, habitual epizoa, be- longing in the main to the families Halecidfe, C'ampanularidje, Campanulinidse, and Lafoeidre. The examination of this collection, and of others, leaves with me the di.stinct impression that the epizoic Hydroids are not distributed indisciiminately ujion all types of Hydroid host. Thus, in general, the members of the family Plumularida; remain comparatively free from extraneous gro\\-ths — a fact to be correlated, perliaps, with their possession of nematophores ; while colonies, belonging especially to the famUj' Sertularida;, are occasionally so overgrown that the structures of the host are much obscured. Even distinct pre- ferences for certain species may be observed. Taking, for each epizoic species in this collection, the total numbers of stations at which it was found, and adding those numbers for all the epizoic species, we find that epizoic species were found at a sum total of forty five stations. At thirty-tico of these Idia pt-istis was either the only host, or one of the hosts where more than one species L2J 1910.] FUOM MEUGUI ARCniPELAGO. 801 was infested ; Tht/rosoyphns vitieiisis at nine ; Serttdarella quad- ridens at foui' ; Corydendriam sessile, Eadetulrium atlenuatiim, Lytocarpus phceniceiis each at two ; and Diphasifi digitalis and Plumidaria setacea each at one. This result bears out the impression gained from liaudling the collection, that Idia prisiis is remarkal)!y subject to infestation liy extraneous Hydroid colonies. Frequently its hydrothecic were all but obscured by the attendant growths of Hydroids and Polyzoa, and this although the state of the hydranths and of the cienosarc general]}', indicated that the host colony was in a perfectly healthy condition. The occurrence, in Lafoea veausta (?), of what appears to be a step towards a more intimate symbiosis ; of a canaliculated ccenosarc in the stem and blanches of Sertularella quadrideus ; of the liitherto undescribed gonosomes of Haleciavi simplex and Tlirp-oscyphtis regularis, are worthy of notice. Finally, I would express my thanks to Dr. R. N. Rudmose- Brown and Mr. J. J. Simpson for entrusting this collection to me for examination. List of Species. I. Gymnohlastea. Family Clavid.b. * Corydendriuin sessile, sp. n., p. 802. Family Bougaixvillidj;. * Ferigo7iimus repens Wright, p. 804. Family Eudendhid^. * Evdendriuvi attemmtum Allman (?), p. 804. * „ generalis Lendenfeld, p. 805. Family Corynid^. Cladocoryne haddoni Kirkpatrick, p. 805. Family Pennarid^. Pennaria disticha Goldfuss, var. australis Bale, p. 80C. II. Calyptoblastea. Family IlALECiu.i:. *Ualecium simplex Pictet, p. 807. „ tenellum Hiucks (?), p. 808. Family Campaxularid.*. Campa/nularia corrugata Thorncly, p. 809. „ raridentata Alder, p. 809. Eehdla calcarata A. Aga.ssiz, p. 810. „ crateroides Ritchie, p. 810. Thyroscyphus regularis Jiiderholm, p. 81 1. „ vitiensis Slarktanner, p. 812. [3] 802 MR. J. RITCHIE OX HTDROiDS [May 24, Family Campanulinid^. * Opercularella lacerta Johnston, p. 812. *Cahjcella oligista, sp. n., p. 813. Cuspidella costafa, Hincks, p. 814. Family Lafoeid^. Lafoea serraia Clarke, p. 815. * „ ?;ejiMs«a Allmaii(?), p. 815. Family Sertui-arid.*;. * Seriularella CJ/lind/rica Bale, var. pusilla, n., p. 817. „ polyzonias L., var. cornuta Ritchie (?), p. 81 8. * ,, quadridens Bale, p. 818. Idia jyristis Lamouroux, p. 820. JDiphasia digitalis Busk, p. 821. Sertularia turbinata Lamouroux, p. 821. Family Plumularid^. Flumularia setacea Linn. (?), p. 822. „ sp. indet., p. 822. Antenella sectmdaria Gmelin, p. 822. Lytocarpus pennarius Linn., p. 822. „ phos7iiceus Busk, p. 823. • indicates a species recorded for the first time from the Indian Ocean. Systematic Discussion. I. GYMNOBLASTEA. Family Clavid.e. CORTDENDRIUM SESSILE, sp. n. (Plate LXXVI. figs. 1 & 2.) Teophosome. — Colony small, the largest of the three collected being only 37 mm. in height. The stem is strongly fascicled, 1'5 mm. thick towards the base, and straight. It may divide into main branches, which are beset by numerous smaller, almost equal branchlets, about 1 mm. in length. These spring from two opposite sides, and lie roughly in one plane. The branchlets on any one side are almost equidistant from each other, but their positions bear no regular relation to those of the opposing series. They taper very slightly towards the base, but there is no trace of ringing nor of wi'inkling. The majority of the hydranths spring from tubes on the anterior sui'face of the branclilets, although a few are also sca,ttered on the main branches. The hydranth tubes do not become free from the common fascicle, except rarely and for an extremely short distance ; nor are special hydrophore-like portions developed. Thus the hydranth projects simply from the open mouth of a tube, the adcauline wall of which is adnate. The hydranth tubes [4] 1910.] FROM MERGUI ARCHIPELAGO. 803 are airanged in two series, tlie polyps of one series alternating with those of the other. The aperture faces outwards and upvyards, is slightly elliptical in shape, and has an even margin, indicating the boundary between tlie thicker perisarc .-ind the place where the chitinous coat becomes so thin and delicate tliat it follows the movements of the polyp. The perisarc throughout is covered by a thin coating of extraneous material — diatoms, sand -grains, fragments of sponge-spicules, and such like. The hydranths are much contracted, and in this state exhibit a pyriform body with well-developed hypostome. The tentacles appear to number about forty. GososoME unobserved. Colour, in alcohol, grey. Dimensions : — Diameter of branchlets immediately above origin 0'3I-0-38 mm. „ hyd ran th tube 0-25-0'31 mm. Hydranth, length 0-52-0-98 mm. „ diameter 0-21-0-42 mm. Nematocysts on tentacles, length 5 yu. ,, ,, breadth 3 /x. Locality. Three small colonies, with hydrorhiza embedded in a .«;ponge, were found at St. 35, between Warden Island, llowe Island, and Lyall Island, 15 to 20 fathoms, rock and sand. So little is known of the variation liable to occur in the indi- vidual species of this genus, that it is difficult to fix on definite specific characters. In this case, I have relied mainly on the structure of the tubes from which the hydranths project. In the majority of the species of Corydendrium so far described these are free for a considerable distance belo\v the hydranth ; but here the tubes are adnato up to the end. This gives to the branches a dumpy appearance, unlike the lax growth of other species. In this respect it approaches most closely C corragatum Nutting (1905, p. 941), which is to bo distinguished, however, by its distinct " hydrophore-like structures," which are " usually distinctly corrugated with deep irregular annulations" ; by the presence of a well marked annular constriction near the origin of the branches ; and by the largo size (5 inches) of the colonies. The species, Soleniopsis dendrifonnis, described by me in 1907 (1907, p. 491) as representing a new genu.s, under the erroneous supposition that Corydendrium gave rise to gonophoros with free meiiusre — I had been unable to refer to the magnificent de.scription of Weismann (1883), — obviously belongs to tiie genus Cory- dendrium, as Dr. Stechow has recently pointed out (1909. |>. 9). Corydendrium dendriformis dillers from C. parasiticum in possessing thick, definite stems, with branchlets arranged in pseudo-pinnate fashion, and in having the free portion of tlie tubes from which the hydranths project much shorter than tiiose of C. parasiticum. [5] 804 SIR- J- RITCHIE ON- HYDROIDS [May 2-t, Family B o c G A i s v i l L I D .E. Perigosimus kepess Wright, 1858. Rare examples of an epizoic species occur, wLicli cannot be specifically distinguished from British specimens of P. repens, the simple lax habit of which they exactly assume. There are dLffei-ences between the dimensions of the Mergui and of British examples ; for whOe the former are smaller in height and in the proportions of their hydi-anths and hydrocaulus, in respect of the nematocysts in the tentacles the order is reversed. The compai-ative table which follows shows at a glance the size- relations of the two forms. About twelve tentacles cro^-n each hydi-anth. No trace of gonosome was observed. Dimensions : — ilergui Typical Scottish specimen. example *. Height of colony 3 mm. 6 mm. Diameter of hydrocaulus 0'04 mm. 0'07 mm. Hydi-anth, length 0-17-0-24mm. 0-24-0-34mm. „ greatest breadth 0-08-0-13 mm. 014-0-15mm. Xematocysts of tentacles, length 5"5 /x. 4'5 /x. „ „ breadth 3 /j.. 2'2 /i. Locality. Rare colonies epizoic on Cori/denclrium sessile, from St. 3.5, between Warden Island, Howe Island, and LyaU Island, 1.5 to 20 fathoms, rock and sand. The present record adds P. repeyis to the fauna of the Indian Ocean. It has already been noted from the eastern and western sides of the North, and the western side of the South Atlantic Ocean, from the Mediterranean Sea, and from the eastern and western (Japan: Stechow, 1909, p. 25) sides of the Pacific Ocean. Family Eudendrid^. EuDEXDRiuM ATTExuATUSi Allman (?), 1877. Many poor colonies, lacking any trace of hydranth or gono- some, I refer, with uncertainty, to this species, on account of resemblance in the skeleton. The Mergui specimens attain a somewhat gi-eater length (3 inches) than the original examples ; but the delicate, very slender, non-fascicled stems, with their few branches, and their short hydranth-bearing ramules lying in one plane and set alternately at legular intervals of about 1 mm., are very similar in both cases. Three or four rather irregular annu- lations mark the base of each ramule, and occasionall)' a few odd rings occur irregularly on the ramules and on the stem itself. The regiUar and close alternation of the hydmnth-bearing ramules seems the most evident character of an indefinite species, though a somewhat similar arrangement is observed in E. maldivense Borradaile (1905, p. 838). * Slide of specimen from Loch Carron, 60 fathoms, in my collection. [6] 1910.] FROM MERGUI ADCHIPELAGO. 805 Dimensions : — Stem, diameter 0-20-0-24 mm. Hydranth-beaiing ramules, diameter . 0'll-0-16 mm. Localities. Many colonies from St. 22, Hastings Harbour, 3 to 20 fathoms and shoie, rock and snnd ; and a few poor colonies, paler in colour, from each of Stations 2;i, Five Islands, 8 to 12 fathoms, rock and sand, and mud ; and 25, Gregory Group and Crichtou Island, 4 to 14 fathoms, stones and broken shells, and rock. Recorded originally by Allman from S,S.\V. of Tortugas, at a depth of 60 fathoms, this doubtful species forms an addition to the fauna of Indian seas. EuDENDRiUM GENERALis Lendenfeld, 1885. The colonies belonging to this species are mature, but are only about 1 cm. in height. They bear well- developed female gonophores, oval in shajie, which spring in a wliorl of four or five from the base of a hydianth, as in Kirkpatrick's description and figures (1890, p. 607, pi. xv. fig. 2). The trophosome of this species is hard to distingiiisli from that of E. capillare. The more slender character of the branches of the latter, relied on in part by Kiikpatrick, is of little v.alue, since a considerable amount of variation occurs in both species. On the other hand, the branching of E. capillare is more profuse, and irregular ; also in the Australian species there is, in the smaller branches, distinct alternation, while the whole colony appears more rigid than the lax growtlis of E. capillare. Dimensions : — Stem, diameter 0-14 mm. Hydranth-beai-ing ramule, diameter 0'09 mm. Locality. Creeping in small (juantity upon Etidendrium attenu- aium ? and Idia pristis, both from St. 22, Hastings Harbour, 3 to 20 fathoms and shore, rock and .sand. Until the present record from tlie Indian Ocean, this species has been found only in the Australian region : Port Phillip, near low water (Lendenfeld, 1885, p. 351); Ton-es Straits— 20 miles N.N.W. of Warrior Island, 5J- fathoms; Murray Island, 15 to 20 fathoms (Kirkpatrick, 1890, p. 007). Family C o r y n i d iE. Cladocoryne haddoxi Kirkpatrick, 1890. Except in size (for they are even more diminutive ihan these recorded by Kirkpatrick and Jiiderholm), the Murgui specimens agree with the description an, 1907, and 1908, as Ihdoconlyle coojnri, .syn. Feyinaria australis. var. cooperi) and that from Christmas Island in the paper following. II. CALYPTOBLASTEA. Family H a l e c I D .«. Haleciu.m SIMPLE.X Pictet, 1893. (Plate LXXVII. figs. 10 &. 11.) This verj- rare species is represented by closely woven colonies covering the surface of a Polyzoon which is clustered at the base of colonies of Corydendrium sessile. The presence of the gono- some, so far uudesci-ibed, renders these specimens of particular interest. Trophosome. — To the naked eye the minute colonies are almost invisible, the hydranths being most easily di.scerned as they project from the surface on which the colonies grow, in close groups, resembling clusters of the smaller species of the entoproctan polyzoon, Pedicellina. The stolon strands are of small diameter, but can often be traced foi' a considerable distance. Their courses are com- plicated, however, by the occurrence of offshoots, which, uniting with other stolons, form a network which lies closely upon the substratum, or is interwoven with it. Short, unbranched, unwrinkled hydrocauli arise from the stolon at short distances from each other. Bnsally their diameter is small, but it gradually increases until, at the ba.se of the hj-dro- theca, the girth of the hydrocaulus is about equal to that of the stolon. The hydrotheca; are very small. Their walls, which are well developed, form an inward curve to the margin, so that, instead of being flaring or trumpet-shaped, as in many species of JIalecium, they are rather saucer-shaped. Very clo.se to the margin occurs a ring of bright dots, chitinous prominences on the inner surface of the hydrotheca to which the hydrant h was attached. Here the internal ])rominence.s are more liighiy lieve- lopcd than in any other species I have examined, for they .sometimes rise into pronounced spines 7 /it in length (PI. LXXVII. fig. 11). Occasionally two or even three hydrothecac occur in succes,sion, a. new hydrocaulus arising from within an old hydrotheca. The hydranth is of g:-eat size. When contracted it is about twice as long as the hydrooaidus, but when extended it reaches a length of one millimetre or even more. It con.•^ists of a long [9] 808 ME. J. RITCHIE OX HTDROIDS [May 24, cylindrical neck, the base of whicli is attached to the hydrotheca at the chitinous projections already described. Distally this swells out into a very definite bulb, which is surmounted by a circle of from 18 to 20 tentacles. At the bases of the tentacles a ring of very large bean-shaped nematocysts, 30 /* long hj 6 f/. broad, surrounds the' hydranth. Similar cells occur in the coenosarc of hydrocaulus and stolon. GososoME. — This does not appear to have been described. A gonangium arises from the side of the hydrocaulus, about half-way between the hydrotheca and the stolon. The female gonangiun* is borne on a short stalk, and is roughly calceolate in shape, resembling the female gonangium of H. beanii, except that here the tubular orifice of the upper surface does not occur. All the gonangia I have examined contain ova to the number of about six. The male gonangium, therefore, remains unknown. Dimensions : — Stolon, diameter 0'06-0'07 mm. Hydrocaulus, length 0-18-0-28 mm. Hydrotheca, depth 0-024-0-027 mm. „ diameter at margin O'08-OIO mm. Gonangium, length 0'42-0'50 mm. „ greatest diameter O'l 7-0-21 mm. Hydranth, length extended 1 '05 mm. „ • ,, contracted 0-78 mm. Locality. Epizoic on a polyzoon associated with Corydendrium sessile, from St. 35, between Warden Island, Howe Island, and Lyall Island, 15 to 20 fathoms, rock and sand. Hitherto recorded only from Amboyna, in the Moluccas (Pictet, 1893, p. 22), this record adds the species to the fauna of Indian seas. Halecium TEXELmsi Hincks (?), 1861. The trophosomes are lax in habit, and in all other respects agree with Hincks's species ; but, in the absence of the gonosome, confident identification is impossible, especially since the likeness between the trophosomes of young examples of the Austi-alian H. parvidum Bale and ff. tenelhim appears to be almost perfect (see Markt.-Turner., 1890, p. 218). Dimensions : — Stem, diameter 0-045 mm. Hydrotheca, depth 0-024-0-030 mm. „ diameter at margin 0-099-0-108 mm. The dimensions of the Mergui specimens are considerablv smaller than those of British examples. Localities. Epizoic on Thyroscj/phvs vitiensis from Stations 15 and 16, Ravenshaw Island, Sir John Malcolm Island, and Alligator Rock, 5 to 18 fathoms, rock and sand, or rock and mud. Recorded with some doubt from Eastern Australia bv Bale (1884, p. 65) and Lendenfeld (1885, p. 405), this widely dis- [lOJ l'>10.] FROM MEHGUI ARCHIPELAGO. 809 tiibiited species has been described from the Indian Ocean only by Billard, from Macalonga, Mozambique, 22 metres (1907 6, p. 338). Family Campaxularid.e. Campanularia corrugata Thornely. Except at Station 1, where it was also associated with Sertu- Inrellu guadridens and Diphaiia digitalis, this species always occurred as an epizoon on Idiapristis. Localities. St. 1, east of Tavoy Island and Port Owen, 4 to 12 fathoms, sand and broken shells, and mud; short and stout, strongly ribbed hydrothecse on Diphasia digitalis, and long, cylin- ilrical individuals on Idia pristis. Stt. 15 and 16, Ravenshaw Island, Sir John Malcolm Island, and Alligator Rock, 5 to 18 fathoms, rock and sand, or rock and mud ; occasionally. St. 22, Hastings Harbour, sliore to 20 fathoms, rock and sand ; fairly common. St. 23, Five Islands, 8 to 12 fathoms, rock and sand, an mm. „ breadth at mouth 0-63 mm. Peduncle, diameter 0-20 mm. Gonangium, length 3 mm. ,, breadth 1mm. Localities. A larije bunch of colonies from St. 22, Hastings [13J 812 MR. J. RITCHIE ON HTDROIDS [Maj- 24, Harbour, shore to 20 fathoms, rock and sand. St. 23, Five Islands, 8 to 12 fathoms, rock and .sand, and mud ; two colonies. St. 25, between Warden Island, Howe Island, and Lyall Island, 1 5 to 20 fathoms, rock and sand ; many colonies. This species has already been recorded from the China Sea (Jaderholm, 1896, p. 9), and from Bird Island, Algoa Bay, Cape Colony (Warren, 1908, p. 344, as T. ceqiialis). Thtrosctphus vrriESSis Marktanner-Tumeretscher, 1890. A widely distributed species, occurring at eight of the fourteen stations from which Hydroids were obtained. In this, as in T. regidaris, the h3'dranth is attached to a row of minute chitinous prominences on the inside of the lower portion of the hydrotheca. These are an-anged as in the previous species, but they are moie strongly developed, and are placed on a slightly raised portion of the hydi'othecal wall. Localities. St. 1, east of Tavoy Island and Port Owen, 4 to 12 fathoms, sand and broken shells, and mud ; common. St. 3, French Bay, King Island, and south end of Iron Island, 3 to 8 fathoms, mud and rock, or sand ; common. St. 9, between Bentinck Island and Courts Island, 12 to 26 fathoms, sand and shells ; rare. St. 14, Bushby Island pearling-gi-ound, shore to 21 fathoms, sand and mud; rare. Stt. 15 and 16, Ravenshaw Island, Su" John Malcolm Island, and Alligator Rock, 5 to 18 fathoms, rock and sand, or rock and mud; rare. St. 19, Pave Island and Pink Island, 7 to 9 fathoms, rock and sand ; rare. St. 25, Gregory Group .and Crichton Island, 4 to 14 fathoms, stones and broken shells, and rock ; fairly common. St. 35, between Warden Island, Howe Island, and Lyall Island, 15 to 20 fathoms, rock and sand ; lare. Family Campaxulinid^. Opercclaeella lacerta Johnston, 1847. Identification depends upon trophosome characters alone, the gonosome being absent. I can detect nothing, however, -which could distinguish these specimens from Xorth Sea examples of the above species. It occurs in its simplest form, a creeping stolon, sending up here and there short ringed stalks, on each of which is poised an exceedingly hyaline hydrotheca. Dimensions : — Hydrotheca, depth * 0-21 mm. diameter 0-08-0-O9 mm. Peduncle, diameter 0-04 mm. LocaliUj. jMoskos Islands, 3 to 26 fathoms, rock and sand, or rock and mud ; rare, growing on a stem of Plumularia setacea. Although the range of the species is a wide one, for it has been recorded from the North of Europe, from both sides of the * From top of operculum. [UJ 1910.] FROM MERGUI ARCUIPELAGO. Sl'A North Atlantic Ocean, and from the "Western Pacific (China and Eastern Australia), it has not hitherto been found in the Indian Ocean. Calycella oligista *, sp. n. (Plate LXXVI. figs. 3 & 4.) Trophosome. The colony is epizoic, and consists of a creeping, delicate stolon which lies in close contact with the specimen upon which it occurs. From this stolon liydrothecie spring at irregular intervals. The hydrothecfe are exceedingly minute, delicate, liyaline, and ti'ansparent, without a tr.ace of the smoky tint which charac- terises C. syringa. In shape they are tubular and cylindrical, of approximately the same diameter for the greater portion of their length, although the proximal half is usually slightly bulged, and tapeis, more or less rapidly, to join the stalk. The margin of the hydrotheca, which is very slightly everted, is definite, and is of considerable strength, for it is seldom that it is found in a collapsed condition. It is divided into shallow crenulations, similar to those in C. syringa, each of which sulitends the base of an opercular valve. In no case was a dujiliration of the margin noticed. The operculum is distinct from the liydrothecal wall, and is composed of about twelve to fourteen exccedinglv delicate valves, difficult to distinguish. The cavity of the hydrotheca is separated from that of the stalk by a distinct but delicate diaphragm. The stalk is usually very short, so that the hydro- theca almost rests on the stolon, but in one case (PI. LXXVI. fig. 3) a well developed stalk was observed. This stalk was marked by a few irregular indentation.s, but in other cases the stallcs were smooth, and ringing of a regular nature never occurred. The hydranth is large compared with tlie size of the hj'drotheca. It is attached by a basal disk to tlie hydrotheca! wall immediately above the diaphragm, and is continued upwards ns a cylindricnl body with conical hypostome, surrounded by a whorl of about seven to ten tentacles. The ectoderm is thick, and the tentacles, in contraction, are dumpy and stout. The gonosome was not observed. Dimensions : — Stolon, diameter 30/*. Hydrotheca, depth 120-138 /a. ,, greatest diameter 45-66 /i. Localities. St. 23, Five Islands, 8 to 12 fathoms, rock and .sand, and mud ; very rare, on Idia pristis. St. 35, between Warden Island, Howe Island, and Lyail Island, 15 to 20 fathoms, rock and .sand ; very rare, on Idia pristis and Thyroscyphns vitiensis. Systematic position. — This species differs remarkably in size, coloiar, texture, and in the develojiment of the peduncle and its ringing from C. syringa, although there is a similarity between • AAVyiffros, smallest. Proc. ZooL. See— 1910, No. LIII. 53 [15] 814 MR. J. RITCHIE OX HYDROiDS [May 24, its hydrothecfe and some of the varieties of the latter species (see Broch, 1909 b, text-fig. 22). It approache.s most closely C. nuttingi Hargitt (1909, p. 378), fi-om which it difl'ers in being smaller in size, in being cylindrical instead of tapering from hj'drothecal margin .to base, and in lacking distinctly annulated peduncles, with "annulations occasionally extending some distance (rarely over the entire body) on the thecal walls." I regard the unnamed specimen, mentioned and figured by Miss Thornely (1908, p. 83, pi. ix. fig. 5), from Khor Shinab, 10 to 12 fathoms, in the Soudanese Red Sea, as belonging to this species. CuspiDELLA cosTATA Hincks, 1868. (Plate LXXYII. iig. 8.) To the short and incomplete description given by Hincks the following observations may be added. The hydrotheca is cylindri- cal for the greater part of its length, but near the base it gradually tapers downwards. Frequently the basal portion of the hj-dro- theca lies along the object upon which the epizoon is growing, the distal portion bending upwards at an angle, slightly after the manner of Lafo'ea serrata. Within the bent portion the retracted polyp is generally found. Similar bent hj-drotheca; have been described by Dr. Billard from La Hougue (1904, p. 165). The lines of growth held to be characteristic of the species, and clearly marked in our specimens, are not lines of growth in the ordinary sense, marking the place where new gi-owth has commenced. They indicate, instead, the margins of earlier hydro- thecie, the opercular flaps of which occasionally remain attached and give an appearance of longitudinal fluting to the succeeding portion of the hydrotheca, as is suggested in the terminal seg- ment in Hincks's figures (1868, pi. xl. figs. 5, 5 a). The secondary hydrotheca (produced probably on the regeneration of an entirely new hydr'anth) lies within the primary, and for some distance their walls, though in close contact, remain distinct. The tertiary, when such occurs, lies within the secondary, and so on. Thus the terminal segment is, as Hincks observes, " of thinner material than the rest," for its walls are those of a single hydrotheca, while proximal to the uppermost segment the walls, owing to the telescoping arrangement, are considerably strengthened. It is noteworthy that the proxim.al segment is usually of distinctly less calibre than its successors. The operculum of the hj'drotheca is composed of about 14 or 15 flaps. The h3'di'anth is strongly retractile, retiring on contraction to the proximal third of the hydrotheca. It appears to have about six to eight tentacles, which are closely set with whorls of large oval nematocysts, the whorls being 7-5 jj. apart. The average size of the nematocysts in these whorls is 5-2 p. long by 2 /n bi'oad, but at the tip of the tentacle larger examples occui-, 10-5 /i long by 5 //. broad. Ko gonosome was observed. [16] 1010.] FltCM MERGL'I AnCIilPELAGO. 815 Dimensions : — Hydrotlieca, length 0-7-1-12 mm. „ diameter ^ 0-10-0-14 mm. Length of flaps of operculum 011-0-13mm. Localities. St. 1, east of Tavoy Lsland and Port Owen, 4 to 12 fathoms, sand and broken shells, and muil ; rare hydi-otheca; on Thyroscyplms vitiensis. St. 3, French Bay, King Island, and south end of Iron Island, 3 to 8 fathoms, mud and rock, or sand ; very rare, on Thyroscyphus vitiensis. Famil}- L A F o E I D .E. LafoEa serrata Clarke, 1 879. A widely distributed, but moderately i-are species. Repeated regeneration of the hydranth had furnished one hydrotlieca (from St. 22) with the exceptionally large number of eleven successive inai'gins. Localities. St. 1, east of Tavoy Island and Port Owen, 4 to 12 fathoms, sand and broken shells, and mud ; fairly conmion, on Idia pi-istis. St. 9, between Bentinck Island and Courts Island. 12 to 26 fathom.s, sand and shell; very rare, on Thyroscyphus vitiensis. Stt. 15 and 16, Ravenshaw Island, Sir John Malcolm Island, and Alligator Rock, 5 to 18 fathoms, rock and .sand, or rock and mud ; rare, on Idia pristis. St. 22, Hastings Harbour, shore to 20 fathoms, rock and sand ; fairly common, on Thyro- scyphus vitiensis and Idia pristis. St. 23, Five Islands, 8 to 12 fathoms, rock and sand, and mud ; rare, on Eiidemlrium attenna- iuml St. 25, Gregory Grouji and Crichton Island, 4 to 14 fathoms, stones and broken shells, and rock ; i-are, on Idia pristis. St. 35, between Warden Island, Howe Island, and Lyall Island, 15 to 20 fathoms, lock and sand ; rare, on Idia pristis. LAFoiJA VExrsTA AUman (?), 1877. (Plate LXXVI. figs. 5 A 6.) Various Hydroids have associated with them a small form which I refer, with a query, to this species. The trophosome consists of a creeping hydrorhiza from which spring small c\'lindrical hydrotheciP, set on stout but variable .stalks, and zoned by numerous, well-deKned, and regular corru- gations. These decrease in distiuctne.ss towards the base of the hydrotheea. The margin of the hydrotlieca is everted, .and the stalk bears no regul.ar rings. A markeil and distinctive character exists in the colour of the peiisarc which jiossesses a brownish- smoke tint, .slightly varinble in cleiisity. An exceedingly delic.iti* membrane forms the floor of the iiyilrotliec;i. The hydranth is cylindricid and minute, funiislied with a hemispherical l\vpostouie, surrounded by tentacles v;iiying in number fi-om nine to eleven. The tentacles atipie Co- imp HYDROIDS FROM MERGUI ARCHIPELAGO. p. Z.S 1910. Pl.LXXVr. Jakes Ritchie . del . London 5te*«0£ccpic Co. imp. HYDROIDS FROM MERGUI ARCHIPELAGO. 1910.] FROM MEROUI AIlCIlIPELAaO. 825 LiAMouRorx, 181G. Histoire raiic1l with hvilraiilbs. X 25. .'i. Ca!i/cella u!ii/isla,!ij>.a. IlvdrothfC.l :iMil livilranlli. X 2.'!0. 4. , , .) " " 5. Zafoea vennsta -Mlmail (r). IlvJrotlieca and livdr.iTilli. X «l). C. Plate LXXVII. Fig. 7. Tlii/rogcyplms regtikiris .liideiliolni. Portion of branch with Ij.vilrolhtc* and Koiiangium. X 25- 8. CuspiilcUa costata Hincks. Ilydrolhtxa, showing clmrai-tiT ot' annular iiiai'kingK. X 70. fl. ScHularella ci/liiidrica, mv. pusiHa.tmv. Portion of stoni, with hydro- thecu!. X 40. 10. Rahcium simplex Pictet. Trophosomc and gonosome. X 70. lj_ _ llvdrotht'ca, showing intcniul chitinoiis pro- jections. X 310. 12. Sertularelhi qimdrideiis Palo. Showing variation in rpnioloness of hydrothciM'. {a) Proximal portion of branch. (4) Distal porlioii of same branch. Each X 19. V'Y-c-'j JOURN.R.MICR.S0C.1910.P1.XIII. West, Newman photo-liti JouM. R. Mick. Soc, 1910, pp. 681-1)2. f 29j XIV. — Hicksonella, a New Gorgmiellid Genus. By James J. Simpson, M.A., B.Sc. (Carnegie Fellow, University of Aberdeen). (Read October 19, 1910.) Plate XIII. In my recently published " Eevision of the Juncellid-group of the Gorgonellida-,"* I did not include any notice of tlic unique form described by Professor S. J. Hickson under the name of JunccUa spiralis. Its puzzling and divergent character made a separate discussion advisable.f In the Eevision the Juncellids were divided into three genera, namely, Juncella, Scirpearia, and Nicclla, and an examination of the diagnoses given there, along with the following descriptions, • Proo. Roy. Irieh Acad., No. 7 (1910) pp. 247-386 (19 pis.). t I wish to tako this opportuuity of thanking Professor J. Arthur Thomson for entrusting these forms to me for identification and description, and also tho Carnegie Trust for a grant to cover the cost of the illustrations. EXPLANATION OF PLATE XIII. Fig. 1. — Portion of Hicksonella spiralis, enlarged to show tlie spiral form and also the distribution and nature of the verrucje. x 3. „ 2. — Spicules from tho verruca' of ditto : (n) near base, [b) near tip. ,, 3. — Spicules from tho ccunenchyma of ditto : (n) near base, (6) near tip. ,, 4. — Two views of the same portion of ll.flagcllata sp. n. : (a) polyp-bearing, (b) non-polyp-bearing aspect, x 4. ,, 5. — Spicules of ditto. „ 6. — Two views of the same portion enlarged of U. capcnsis sp. u. : (a) polyp- bearing, (6) non-polyp-beariug aspect. X 4. , 7. — Spicules of H. capensis sp. n. 30 Transactions of the Society. will show the impossibility of linking the specimens now under consideration to any of tliese genera. It will be shown that it is necessary to establish a new genus. As a full description of the Juncclla spiralis type has been given by Hickson,* we may briefly consider some of the most characteristic features before proceeding to any taxonoinic con- sideration. The colonies were all unbranched ; one of them was 220 mm. in length without the base, which had been broken off. The axis was pale brown in colour, and had rings of lime embedded in the horny matrix. The verruca; were all prominent and arranged irregularly on two-thirds of the circumference of the stem.''leaving\ bare track on one side free from verrucas from end to end of the colony. The bare track and verrucye were covered with a dense armature of spicules, and " it is difficult to believe that the verrucse can ever be retracted." The spicules consisted chiefly of irregularly tuberculated plates and spindles. "The spicules are tightly jammed together to form an impenetrable armour. The surface of the verrucse has a distinctly squamate appearance, the plate-like spicules slightly over-lapping." Hicksonella f spiralix g.n. ( = Juncclla spiralis Hickson). Plate XIII. tigs. 1-3. Two specimens referable to this species occur in a collection from Cape Colony. Professor Hickson very kindly sent me a small portion of his type specimen for examination, and one of the present colonies agrees with it ia almost every feature. It is 40 cm. in length, the diameter is 3 • 5 mm. near the base, and about 2 mm. at the tip. The tapering is thus very gradual, in fact at a distance of 7 cm. from the tip there is very little difference in the diameter from that at the base. Another specimen is 20 cm. in length, and has a maximum diameter of 2' 5 mm. at the base. In Loth specimens the basal part is wanting. The axis is densely calcareous, and very hard in the lower portion, so that, except near the apex, the colony is very rigid. It is composed of concentric laminae of the typical Gorgonellid type. There is a distinct, central, more densely calcareous portion, and the surface is marked by a series of minute longitudinal furrows. The diameter near the present base is 1 • 5 mm., but this diminishes gradually to a hair-like fineness at the tip. • The Alcyonaria of the Cape of Good Hope. Part [II., Marine Investiga- tions in South Airica, iii. (1904) pp. 231-33 (4 figs.). t I wish to associate with this new and remarkable genus the name of Pro- fessor Sydney J. Hickson, D.So. F.R.S., who has done so much to elucidate the structure and relationship of Alcyonarians, and to whom we are indebted for the original observations on Hich$onella. Hicksonella, a New Gorgonellid Genus. By J. J. Sivipson. Lil] The ccenenchyma is very thin and densely spiculose ; wlieii dry it is extremely brittle. The canal S3'stem is difficult to determine owing to the thin- ness of the coenenchynia. It was found impossible to detach a portion of the ccEnenchyma from the axis without damaging the canals. Serial sections were made from a decalcified portion of the ccenenchyma, but the I'esults were not very satisfactory. The furrows on the axis suggest a concentric series of canals separating the ccenenchyma from the axis, but it is very doubtful whether one or more of these is larger than the others, and the possession of an outer series of canals is quite uncertain. From end to end of the colony there is a narrow streak devoid of polyps, which participates in the spiral arrangement of the colony (pi. XIII. fig. 1). It occupies between one-fourth and one- third of the circumference of the ccenenchyma. The remaining three-fourths to two-thirds is densely covered with small elongated verrucffi. In the smaller specimen the basal portion, for a distance of 8 cm., is devoid of polyps. The verrucse are long and club-shajied, and somewhat resemble those of Scirpearia Jiagellum ; the terminal part is considerably enlarged. They are about 1'5 mm. in length, and about 0 o mm. in diameter at the widest part. They are very spiculo.se, and tlie surface, when viewed with a lens, reveals a series- of minute hori- zontal, overlapping scales, which recalls the armature of a Caligorgid. At the utmost they are only feebly retractile into the ca>nen- chyma, but their enormous size, the thinness of the cffineuchynia and their dense armature, would seem to preclude the i>ossibility of any great degree of retraction. WTien the tentacles are inturned there is a distinct eight-rayed figure at tlie summit of the verrucae, i J. J. Simpsoii. (_33'? the colony. Towards the base the polyps iliminish in number while on the basal 4 or 5 cm. they are altogether absent. No definite arrangement is discernible, and young polyps occur scattered amongst the older forms. The verrucje are elongated anil club-shaped (])1. XIII. fig. 4a); they are about 1"5 mm. in lengtii and about (I'o mm. in diameter near the apex. They are not retractile into the cceuenchyma, and the swollen terminal part is due to the withdrawal of the antho- cpdia within the verruca. They are densely spiculose, and the scales on the aboral surface of the tentacles form a pseudo- operculum to the partially retracted anthocodia. Fig. 46 shows the characteristic appearance of the portion figured in 4'( from the non-polyp-bearing asjiect. The spicules (pi. XIII. fig. 5) of this species are very character- istic. The following are some of the chief types, with measure- ments, lengtii by breailth, in millimetres : — 1. Thick, mas.sive, wartv .spindles. (O-OaS x 0-015; 0-049 X 0-011; 0-0.'58 X 0-015; 0-038 x 0-011.) 2. Smaller ; usually curved spindles with the warts more developed on the convex side. (0-046 X 0-008; 0-031 X 0-008.) 3. Irregular scales (from the verrucae). (0-031 x 0-011; 0-027 X 0-015.) 4. Aberrant type, resembling crosses. (0-034 x 0-023.) 5. Peculiar, bilaterally svmmetrical type with a sort of thick foliaceous expaii.sion. (0 - 015 x 0 • 015 ; 0 ■ 015 x 0 • Oil ; 0-011 x 0-011.) 6. Birotate forms (like thosi^ in SuhiToi/on/ia verriculata). (0-011 X 0-008; 0-008 X 0-008.) LocalUi/ : — Cape Morgan, N. J AV., 10^ miles ; depth, 77 fulboms. By dredge. Eocks and broken shells (2(). vii. 1901). Hicksonella cape.nidx sp. n. Plate XIII. figs. 0, 7. This species has been establislied to include a very character- i.stic colony 75 cm. in lengtii without tiie basal portion. The diameter at the present base is 3-5 mm.; midway it is 3-25 mm., while near the tip it is 3 mm., so that tlie lain'iing is very slight. The actual tij) is conii-al. The colony is twisted in an irreguhir (i|icm spiral thruugliout its entire length, but in such a way that tlic bare tract is always to the inside. The coeiienchynia is extreiiu'ly lliiii and is dcnsi'ly s])ii'ulose. The axis is cum])os('d dl' concentric lainiiia-, wiiich consist of a horny substant^e iinpregnated with some form of calcanMius niatU^r. It is verv hard, wliite in colour, and the surface is marked by deep longitudinal furrows. It tapers to a hair-like fineness at tlie tip. A 3 34 Transactions of the Society. The polyps are disposed over about three-fourths of the surface of the colony, leaving a bare longitudinal track which is very marked in the lower part, but almost disappears m the upper portion The verrucas are small, elongated and slightly club- shaped ; they are about 2-5 mm. long and 0-25 mm. in diameter, and are much more openly disposed than in the other species (pi XIII. fig. 6a). Fiu. 66 sliows the appearance from the nou- polvp-bearing aspect. "" The tentacles are first infolded and then withdrawn into the upper part of the verrucfe, but the verrucce themselves are not retractile into the coenenchyma. The colour of the ccenenchvma is orange-red, but the verruc« are almost white. The spicules of this species are extremely irregular in^form, but the following types may be distinguished (pi. XIII. tig. 7) :— 1. Short sliglitly warty spindles. (0-07 X 0-015.) 2. Spindles longer and more warty. (O-Oy x 0-038; 0-08 X 0-038.) 3 Spindles still longer and more warty. (0-13 x Q-O.^; 0-14 X 0-046; 0-16 X 0-065; 0-2 X 0-06.) 4. Very irregular forms (probablv developed from [3]). (0-17 X Oil; 0-16 X 0-13.) 5. Flat, UTegular scales. (0-14 X 0-09.) 6. Thick, wartv, almost spherical forms. (U - 06 X 0 - 045 ; 0-053 X 0-034.) 7. Crosses (aberrant). (0-12 X 0-11.) Locality:— T^edi Cliff, S. of Morewood Cave, N.W. | N. 6^ miles. Natal. Depth, 37 fathoms ; bottom, sand and shells. Position of Micksonella. In this connexion the axis stands first to be considered. As mil be seen from the descriptions of the various species, it is com- posed of concentric laminiP ; these laminae consist of a horny sub- stance containing some form of calcareous deposit. It has been impossible so far to determine the nature of this limy deposit, but it is hoped that some solvent may be found to decompose the organic matter, and so enable a microscopic examination of the inorganic residue to be made. UntH this is done, however, deter- mination based on axial structure is impossible. We have, how- ever, in our Eevision of the Juncellids, discussed the possible affinities of several other specimens whose axis is similar to the one under consideration, and the resemblance strongly suggests that the present specimens approach closely to the Gorgoiiellidae, to which family we would therefore temporarily assign them. In the work cited above, after an exhaustive examination of a very large numl)er of .JuncelUds, we suggested an emended classifi- cation, and included in the genus Juncclla only those species whose spicules contained the type known as "clubs." This type of spicule Micksonella, a New Gorgcnellid Genus. Bij J, J. Simpson. I 35j was first described and figured by KiiUiker in his Icones Histio- logies, p. 140, taf. xviii. fig. 46. In the Revision of the Juncellid- group of the Gorgonellidje several figures of this type of spicules are given (fig. 4, a-g). Tiie general form approaches that of the well-known " Indian club ;" there is a distinct smooth median part or liandle, which is surmounted by a few spines. The club-part also Ijcars spines, and the most important characteristic is the fact that these species are all directed away from the shaft, and do not arise perpenilicularly. Some doubt seems to have arisen since tlie time of Kolliker as to the exact nature of the " club," and this dubiety accounts for the original inclusion of the species sjnralis in the genus Juncdla. Hickson (op. cit.) describes a form of s]iicules as a club (pi. viii. fig. 8), but it is a club essentially different from Kdlliker's type. The spiculation as a whole is quite unlike that of any species of Juncdla (op. cit., figs. 14, 23, and 26), so that it is imjjossible to refer the present specimens to that genus. They are even further removed from Scirpearia and Nicdla, so that it seems necessary to form a new genus to include them. Diagnosis of Hidcsondla , g . n. Colony simple, flagelliform, and generally twisted in a mure or less open spiral at least in the ohler colonies. Tlieaxis consists of concentric lamina' which are comi>osed of a horny substance im- pregnated with some form of calcareous matter. It is generally hard, and the surface is marked by longitudinal ridges and furrows. The coenenchyma is extremely thin and densely s])iculose ; it is very brittle, especially when dry. The polyps are disposed in a broad longitudinal liand ; this leaves a narrow bare strip which traverses the whole length of the colony. The verrucas are not retractile into the coenenchyma, and are elongated, slender and slightly club-shaped in the upper portion ; they are covered with minute spicules, which appear like overlapjiing scales; the spicules on the aboral surface of the tentacles form a sort of pseudo- operculum to the partially retracted anthocodia. The spicules vary in th(^ dilferent species, but the following are the cluef types: — (1) Irregularly warted spindles; (2) Hat, smooth, or slightly warty scales ; (3) large irregular forms ; (4) crosses ; and (5) bi-rotate forms. Si'EGiKic Diagnoses. Hidcsonella spiralis (Hickson) = Juncella spiralis Hickson. Colony unbranched ; in the larger forms spirally twisted. The ccenenchyma is thin and densely packed with scale-like spicules ; the axis is composed of concentric lamin* of a horny substance. 36 Transactions of the Society. in which a calcareous deposit is embedded. The polyps are restricted to a region occupying two-thirds to three-fourths of the circumference of the coeneuchyma ; a longitudinal bare tract occupies the remaining part. The verrucse are long and cluli- shaped, and are evidently not retractile into the coeneuchyma ; they are closely packed together, and are covered with minute overlapping, scale -like spicules. The flat thin scales on the aboral surface of tlie tentacles forms a sort of pseudo-operculum to tlie partially retracted. polyp. The chief types of spicules are (1) in the coenenchyma very thick spindles with close-set irregular warts, passing by gradual transitions to almost spherical warty forms ; (2) in the polyps (a) long thick spindles with few long warts ; {h) irregular forms and crosses ; {r) small flat thin scales. Hicksonella flagellata sp. n. Elongated filiform colonies which, at any rate in the younger specimens, have only a trace of a very open spiral structure. Tlie ccenenchyma is extremely tliin and densely spiculose. The axis is horny and calcareous, and is composed of concentric laminae. The polyps are confined to a broad longitudinal band occupying about two-thirds of the circumference. The verrucse are elongated and club-shaped, and are not retractile into the coenenchyma. The spicules are exceedingly minute, and very characteristic. They consist for the most part of (1) thick massive, warty spindles ; (2) peculiar, bilaterally symmetrical forms, w-ith a sort of thick folia- ceous expansion ; (3) bi-rotate forms (like those in the Suberogorgia verriculata) ; and (4) scales. Hicksonella capensis sp. n. Colony elongate, simple, and irregularly twisted in a spiral manner. The coenenchyma is thin, and very spiculose. The axis is compcsed of concentric horny and calcareous laminae. The polyps are disposed in a broad longitudinal band ; they are very long, slender, and slightly club-shaped ; they are covered with small spicules transversely arranged, and are not retractile into the coenenchyma. The spicules consist of («) short, slightly warty spindles ; {b) longer, and more warty spindles : (c) irregular forms ; (fZ) flat, irregular scales ; and {e) crosses. Distribution. — (1) Geographical. All three species were found oH the east coast of South Africa, and therefore their inclusion in the Family (lorgonellidffi does not extend the distribution of that family. n. apiralis. Off Cape Morgan, H.Jlagdluta. Off Cape Morgan. //. capensis. Red CliH', south of Morewood Cave, Natal. Hicksonella, a New Gorgoncllid Genus. By J. J. Simpsmi. (^ZiJ (2) Bathymetrical. This genus, like other Gorgonellids, is represented by shallow-water lorius; M. spiralis was dredged in 36 and 47 fathoms, and H. capensis in 37 fathoms. XOTE OK THE GeNUS. Before leaving Hicksonella, however, it may be of interest to consider some of the more characteristic features, for example : (1) the distribution of the verrucje, and (2) the spiral form. Pro- fessor Hickson (ii. p. 232), has drawn attention to these, and has put forward several suggestions as to their possible origin, so tliat, before taking each in detail, we shall quote his observations in full, and so obviate any possibility of misinterpretation which might result from detached references, "This bare track, i.e. the part devoid of polyps, is seen in some other species of Juncellch. In the description of J. juncea, from the Isle of Bourbon, Milne-Edwards and Haime state that the calices leave some trace of a median coenenchymatous space. Eidley also states that there is a distinct groove in the specimen of -/. juncea obtained b)' the ' Alert.' The squamate armature of the verrucse shows some affinities with the characters of the Primnoidse, but, as the plate-like spicules are so small and there .are no definite opercular plates, its affinities with Juncella are closer. It is noteworthy, however, that in tlie Primnoine genus Calypteriims the calices do not occur on one side of the stock. The track which is free from tlie calices in Calyp)terinus, however, is covered by the overlapping scales of the lateral calices so as to form a tube. These bare tracks on one side of the stock in Juncella sjnrcclis and Calypterimt,s allmani have a certain resemblance to the liare tracks on one side of the smaller branches of some forms of Solenocaulon, and suggest the presence of symbiotic Crustacea. Tliere is no evidence in support of this at present, but it would be worth the trouble of any naturalist, wl\o has the opportunity of dredging in these waters, to note the character of any Alpheidre or other animals that might possibly live with this Juncella. Dr. Gilchrist's note tliat nothing was found around which the stock twisted, is of the nature of a supjiort for the suggestion that th(! spiral form of the larger stocks is associated with the presence of some epizoic animal. We may, for the present, regard the spiral form and the bare track as characters of the species, but if they prove to be mere adaptations to an epizoic animal their importance must be considerably discounted." The question of the nature of tlie bare tract is of more than pas.sing importance, for if, as Professor Hickson suggests, it is due to some epizoic animal, it must be discounted in any qiiestion of a taxonomic nature. A study of the species which he has named along with some others of a similar character, may, liowever. help 38 Transactions of the Society. to suggest another possible explanation. For this purpose let us consider the following species : — 1. Juncella juncea. 2. Scirpearia flagellum. 3. Suherogorgia kdllikeri. 4. Lopkogorgia lutkeni. 5. Juncella trilineata. 6. Scirpearia quadrilineata. 7. Calyptcrinus allmani. Juncella juncea and Scirpearia flagellum, in common with all other Juncellids, except Juncella trilineata and Scirpearia quadri- lineata, liave tlie polyps disposed in two longitudinal series separated by two diametrically opposite longitudinal bare tracts. In Suherogorgia 'kollikeri the polyps arise from all parts of the stem and branches, with the exception of a narrow, often wavy, portion on either side of the cylindrical axis. In Lojjhogorgia lutkeni the polyps are numerous and occur on both sides of the stem and brandies ; the central portion of both the main stem and branches on both surfaces are destitute of polyps. Again, the stem and branches are flattened in the plane of branching and are marked on the flat sides by a narrow winding groove or furrow, which is continued along the flattened surfaces of the secondary branches. The polyp-bearing surface is divided into two lateral liands by means of the two grooves. In Juncella trilineata polyps arise in three different bands, leaving three narrow bare strips, each of which has in its centre a slight rib or keel. In Scirpearia quadrilineata the polyps are grouped in four definite longitudinal series separated by four bare spaces. In Calypteriyius allmani there is a bare tract on one side of the colony, otherwise the polyps occur all round. In Hicksonella the verructe are all prominent, and arranged irregu- larly on two-thirds of the circumference of the stock, leaving a bare tract on one side free from \'errucie for the whole length of the stock. A detailed discussion of the question under consideration with regard to the various species of Juncellids has been given in the " Revision of the Gorgonellids," so that it is only necessary here to recapitulate the conclusions wliicli have been there stated. 1. In all species of Juncellids, so far known, the polyps are disposed in a certain number — two, three or four longitudinal series which are sepaiated by a similar number of bare tracts. This number is constant for the species. 2. There is always a definite number of nutrient canals larger than the others, which are known as the main longitudinal canals. 3. These main canals always correspond in number and position to the bare tracts. Hicksonella, a New Gorgonellid Genus. By J. J. Simpson. [39] 111 Suberofforgia kdllikeri, also, tlie loiicritudinal bare strips denote the position of the two longitudinal canals much larger than the others. The same holds true in the case of Lophogorgia lutkeni. In all the species so far described the ccenenchyina is moderately thick, and it has been possible to verify these points by means of transverse sections, but unfortunately the extreme thinness and the densely spiculose nature of the ccenenchyma in Calgpierinus allmani and Hicksonella have rendered it impassible to determine, so far, whether any main longitudinal nutrient canals exist. Is it not possible, however, that in these species, as in the others cited, the occurrence of a bare tract throughout the entire length of the colony may be an outward manifestation of internal morphological structure, and that it corresponds to the position of a single main longitudinal nutrient canal 1 If this be so, there exists in Juncellids a complete series from one to four main canals, thus : 1. iricksoHella sp. . . . One main canal. 2. Juncellajv.ncea, etc. . . Two main canals. 3. JunccUa triliveata . . Three main canals. 4. Scirpearia quadnlineata . Four main canals. Other features, Imwever, seem to point to the unilateral sym- metry being of morpliological and not of mechanical origin. And in this connexion the spiral nature of the colonies must be con- sidered. It will be seen from the descriptions of the three species given in this papjer that two of these are spirally twisted, and also that the spiral arrangement does not commence at the very base, but at a considerable distance from it. In the third species tlie colonies are all small, and it is possible tliat they, later on, might have developed the spiral form. An analogous case is seen in several Antipatharia in which the polyps have morphologically a unilateral disposition. In the waters around the Mergui Archipelago and also on the reefs on the East Coast of Africa, we have had occa-sion to examine a large uumlier of spirally twisted colonies, and in no case was any form of support found. Now in the majority of these cases and also in the larger colony of Hicksonella spiralis examined, if any rigid support existe, and consequently, in the.se, the axis is less flexible and more brittle. The increase in the thickness of the older part of a colony is due, not to an increase in the thickness of the coenenchyma, but almost entirely to an increase in the thickness of the axis (see tigs. 10, 13, 10, and 43). (5) Spicules. The spicules of this group are characterized by their exlrcme sinallness ; in fact, in no other group of Alcyonaria do we find the prednniiiuinl spicule so minute. Measurements of these with any precision are only possilile with a high magnification. The largest measurements for the group, viz., those in Nicclln dichotoma, are only 0'2o mm. x 0*0() mm. ; but in the genus Juncella the largest are those in Juncella trilineata, which are over 0'070 mm. x 0'038 mm. ; while in 24 Proceedings of the Royal Irish Acadenii/. Scirpearia the largest are those of >S'. anomala, viz., 0"15 mm. x 0"0y4 mm. , while in S. pcdinata the largest are only 0"061 mm. x 02.3 mm. It is not necessary here to enter into the details of all the different kinds of spicules and their variations which occur throughout the group, as this is more fully dealt with under the various species, and the figures given there will, moreover, give a much better idea of these than a lengthy description. At the same time it might be well to define in a general way the various fi/pes which have been described. (a) Clubs. — Fig. 4 (a-fj) gives some idea of the variations of this type. The general shape approaches that of the well-known Indian-club ; and the most important characteristic is the fact that the warts or spines on the club- portion are all directed away from the shaft, the central part of which is smooth. The spines do not arise perpendicularly (see also figs. 14, 23, and 26). (b) Double-chtbs.— Tig. 5 {a-b) shows two variations of this type. They have the shape of what are usually known as dumb-bells. There is a distinct median constriction which may be more or less well defined, and may vary in length as well as in breadth (see figs. 27, 65, 75, and 113). The warts may be large or small, smooth, papillose, or very warty, but all arise perpen- dicularly from the head (figs. 87 and 113), or they may be situated either close together or wide apart ; and, according to which method occurs, the head will be regular or irregular in outline (contrast figs. 63 and 65). The head itself may be hemispherical or slightly conical (contrast figs. 54 and 65). Fig. 6 (a-d) shows characteristic variations of this type as seen in Juncella. (c) (Double- irJt eels or eapstci^is. — Fig. 7 (a and b). This type consists of a cylindrical shaft on which there are two whorls or warts. The ends of this shaft (the hubs) may be either almost smooth or markedly warty. (d) Elongated double-clubs. — This type may be derived from the typical double- clubs, and merges gradually to another form which is sometimes described as double spindles, and this again may pass into the simple spindle. Fig. 8 rt, b, and c show how these merge imperceptibly into one another. In the cases we have described, however, these spicules are all of about the same size, so that the distinction is made chiefly on the basis of the amount of constriction visible and the proportionate length of the " head " to the constriction (see figs. 63 and 65), (e) Sirrifle spindles. — In certain species, e.g. JVicella dichotoma (see fig. 114), there is a type of spicule which may be described as a simple spindle, and which in size contrasts so strongly with the double-clubs that there are no intermediate stages connecting the double-club with the spindle. (See also fig. 118 of Nicella inoniliforme^ Simpson — A Revision of f/ie Gorgonellidat. 25 The different variations which occur in these types will be considered in detail in describing the different species ; but certain generalizations must be briefly referred to here. It is very important when describing spicules from any colony to state precisely from what part of the colony the preparation has been made. New species have been established on slight differences in the size and shape of spicules, and also on the preponderance of one type of spicule over another. With a view to testing the degree of certainty with which this procedure might be justified, we have made different preparations under different conditions from the same colony ; and we now give the results derived from over 500 preparations. (1) The different types of spicules retain their own distinct chai-acteristics, no matter from what level of the colony they may be taken. (2) Spicules from different levels of the same colony or from colonies of different ages show marked deviations in absolute size, but not in proportionate size. (3) Spicules in the coenenchyma alone differ from those in the verrucae alone, e.g. in Scirpearia furcata, the double-club type, with hemispherical ends, is confined to the coenenchyma, whereas the elongated double-club is restricted to the verrucae. This obtains in all specimens examined. It therefore follows that when examining spicules for specific deter- mination the factor of primary importance is the character of the spicules. Ne.xt comes the average size of these spicules; while of no importance whatever is the proportionate numbers of each type, as this depends on the proportion of coenenchyma and verrucae taken for the preparation. If, then, a single preparation be made from a certain part of a colony, and no criterion be given as to the exact age of this portion, subsequent workers will experience great difficulty in making preparations from a similar part. To obviate this difficulty another metiiod may be employed, namely, to take coenenchyma and verrucae from diflerent levels for the single preparation, and so obtain a representative sample of the spicules of the specimen. This method has been found to be of great service in identification, and is the one employed in the preparation of this memoir. Now it has been seen that the disposition of the verrucae is not a constant even in a single specimen, and that its inclusion as a generic character is untenable. If therefore the separati if the specimens of this Juncellid-group of the Gorgonellidae into genera is to be accomplislied, it must be based on the character of the spiculation. E 26 Proceedings of the Royal Irish Academy. If a preparation of spicules be made in the manner described, there should be no difficulty in at once deciding whether or not the type described and figured, as a "club," is present or not. (See figs. 4, 14, 23, and 26.) On the other hand, figs. 114, 115, 116, and 118 give a good idea of the " long- spindle type " and its proportion to the small double-club. Any of the figures given of the various species of Scirpearia — e.g. figs. 27, 31, or 65 — will at once mark these off as quite distinct from the other two types. VI. — Possible Affinities of the Gokgonellidae. In the "Challenger" Report on the Aleyonaria (vol. xxxi.), Wright and Studer divide the (lorgonacea into two large sections : — I. Scleraxouia, and II. Holaxonia. In the Scleraxouia they recognize the Sclerogorgidae as a distinct fam'ily, with the following characters : — " In the representatives of this family a distinct axis is formed of a tissue consisting of numerous closely intercalated elongated spicules, with dense horny shields. The axis is surrounded by longitudinal canals, into which there oj)eii the reticulated coenenchymatous canals uniting the polyps." In the Holaxonia there occurs the family Gorgonellidae, in which " the axis is lamellar and calcareous, but retains its shape after the extraction of the calcareous matter." The nature of the " calcareous matter " is, however, not specified, so that it is very difficult to interpret exactly what may have been the opinion of these authors. In " A Treatise on Zoology," part ii., Bourne divided the Aleyonaria into five large orders as follows : — (1) Stolonifera. (2) Alcyonacea. (3) Pseudaxonia. (4) Axifera. (5) Stelechotokea. The Stolonifera, Alcyonacea, and Stelechotokea are sufliciently distinct, and most certainly have no connexion with the Gorgonellidae, so that any further reference to them would be superfluous. The Pseudaxonia have been described as " Synalcyonadea forming upright, branched colonies. The zooid cavities short ; the zooids embeilded in a coenenchyma containing ramifying solenia and numerous spicules. The Simpson — A Revision of the Gorgonellidae. 27 eoenenchyma clifferentiated into a cortical and medullary portion, the latter containing spicules different from those of the cortex, densely crowded together, and sometimes cemented together to form a supporting aj.is." One of the families of this order — namely, the Sclerogorgidae — is thus defined : — " The medullary mass forms a distinct axis, consisting of closely packed, elongate spicules, with dense horny sheaths. The axis does not contain solenia, but is surrounded by longitudinal canals — i.e., by large solenia — which are connected with the zooid cavities by ramifying solenia." Of the genus Suberogorgia, Gray, in his original description (Proc. Zool. Soc, 18.57, p. 159), says: — "Axis pale-brown, formed of rather loosely concentric fibrous laminae, containing a large quantity of calcareous matter." From .the Pseudaxonia the Axifera are thus differentiated: — " Synal- cyonacea forming colonies consisting of a coenenchymatous rind, investing a horny or calcified axis. The axis may be horny or composed of a calcified horny substance. ... It never contains solenia, and is never formed of fused spicules. The eoenenchyma completely invests the axis, and contains solenia, and calcareous spicules embedded in the mesogloea." Bourne does not include the Gorgonellidae in his scheme of classification ; and as the nature of the calcareous constituent in this family has never been investigated, or even commented upon, it is impossible to say whether they ■■ are Pseudaxonia or Axifera. The time at our disposal has not permitted of a detailed investigation of this very important problem ; but as a contribution to this study the following observations may be useful : — Suberogorfjia — An examination of the axis of a specimen of this genus reveals the following features : — (1) It consists of a horny matrix, in wiiich large irregular spicules are embedded longitudinally. These spicules are easily seen with a strong lens, and appear to be deposited concentrically. (2) The axis after decalcification retains its original shape. (3) Prolonged boiling in caustic potash causes a sligiit disintegration; and the individual spicules may thus be separated. (4) The spicules of the axis are quite different from those of the eoenenchyma. (5) A thin horny layer may be detached from the axis, in wliich the spicules may be seen embedded. Juneella elongata var. capensis. — Hickson (xiii.) described an Alcyonarian from Cape Colony under this name, but at that time llie spicules of this species were unknown . Subsequent study, and a consequent resuscitation of E2 28 Proceedings of the Royal Irish Academy. that old but imperfectly known species, have necessitated the removal of the Cape specimen from this genus. In many respects this colony is unique; and the writer has described it separately as Dendrogorgia (n. g.) capensis, Proc. Eoy. Phys. Soc, Edin., vol. xviii. (1910), p. 62. The following notes on the axis are of great importance in this connexion : — . (1) The axis consists of concentric laminae. (2) These laminae are composed of a horny matrix, in which long irregular spicules are embedded horizontally. (3) The horny substance preponderates in amount over the calcareous matter, so that even with a lens the spicides are not very evident. (4) The spicules differ greatly in size and shape from those of the coenenchyma. (5) Prolonged boiling in strong caustic potash results in a partial separation of these spicules. (6) Wlien the coenenchyma is detached from the axis, a thin, white, trans- parent film is generally found adherent to it. If this he peeled off and placed under a microscope with a one-sixth objective, spicides identical with those of the axis are seen embedded in it. From these facts it is at once evident that the axis in the case of this specimen is distinctly sclerogorgic. It consists of spicules different from those in the coenenchyma embedded in a horny matrix, the individual com- ponents of which are laminae deposited concentrically; and, further, it is possible to separate the outer layer, which is usually detached with the coenenchyma. Juncella jtincea may be taken as a type of gorgonellid axis, and the following are the chief points observed : — (1) The axis consists of a horny substance impregnated with lime. (2) The horny material greatly preponderates over the calcareous. (3) The axis is very hard, and is cut with difficulty. (4) It is impossible to see individual spicules either in a cross-section or a longitudinal section ; but (a) The axis is deposited in the form of concentric laminae. [I) A thin layer is usually found adherent to the detached coenenchyma, (c) This layer may be sepai'ated from the coenenchyma. (5) It contains small spicules not very unlike those of the coenenchyma, but different from them. Simpson — A Revision of the Gorgonellidae. 29 With regard to the coenehchyina in these tlivee groups, the following notes are interesting : — (1) In all of them there is a circle of large canals separating the coenenchyma from the axis. (2) Near the periphery of the coenenchyma there is also a circle of longitudinal canals which communicate directly with the polyps. (3j These two series are united by numerous interlacing transverse solenia. The polyps are very similar in all three groups. There is no definite distinction into verruca and anthocodia. There is a pseudo-operculum formed of small spicules on the aboral surface of the tentacles in all three groups. The polyps are in all cases capable of complete retraction into the coenenchyma. A further point of similarity may be pointed out in the case of Suberogorgia and Juncellids, namely, the possession of a definite number of longitudinal canals in the inner series larger than the others, which deternune the distribution of the polyps. It would be premature to draw any hard and fast conclusions from these few observations ; but it may be consiilered a question wlielher the •three groups taken in the following ordei-, (1) Suberogorgia, (2j DcndnKjorgia capemis, and (3) the Juncellids proper, may not represent a line of evolution. In the first of these the spicules of the axis are large, and there is only a small amount of horny matrix ; in the second the spicules are smaller, and there is a larger proportion of horny material ; while in the last the spicules (if such is the nature of the calcareous matter) are extremely small, and the proportion of horny substance to the calcareous is enormou-sly increased. For the present, and until the exact nature of the limy deposition in the axis of the Gorgonellidae is investigated, it is therefore inadvisable to rank them with the Axifera, and it is more than probable that their affinities are closer to the Pseudaxonia. VII.— Division of the Gorgonelliuae into Genera. Before proceeding to formulate a scheme of classification which may approximate to a natural classification, and whieli will be based on the foregoing considerations, it may be well here to recapitulate the most recent diagnosis of the genera under consideration, and see to what extent each of these may be considered valid. Juncdla. — Tlie colony is simple or branched, the polyps are sometimes small, disposed in two lateral rows, sometimes with well developed and ,S0 Proceedings of the Royal Irish Academy. elongated veiTucae. The coeueuchyma is thick, with an external layer which contains simple and double clubs. Scirpearia. — The colony is simple, with a cylindrical calcified axis and thin eoeneuehyraa. The polyps are seated in two longitudinal rows on each side of the stem. The spicules are double-clubs and spindles. Scirpearella. — The colony is simple or very feebly branched. The axis is calcareous, brittle, smooth or grooved. The polyps are arranged in rows or spirals, retractile, with more or less prominent verrucae. The coenenchyma is moderately thick and finely granular. The spicules are spiny spindles and double-clubs. Ellisella. — The colony is simple or diehotomonsly branched, with a thick coenenchyma, and slightly developed verrucae, which are disponed in two rows on the axis. The coenenchyma contains both doulile-clubs and spindles. Ctenocella. — The colony is branched in one plane, and so that all the simple twigs arise in an ascending order from the upper surface of the stem. The verrucae are short on two sides of the twigs. There are distinct median furrows. The spicules are mostly double-clubs ; those of the polyp-calyces are, according to Eidley, somewhat different from those of the coenenchyma, being longer and provided with two, often three whorls of tubercles. The inner whorls so approach in the middle of the spicules, that the median naked zone, which is characteristic of the spicules of the coenenchyma, is here absent. Nicella. — The colony is upright, branclied, with a thin coenenchyma, and protruding verrucae, which arise perpendicularly, and appear to be terminally truncated. The polyps arise from either side of the stem and branches, leaving a middle space free. The spicules form a cortical layer of small double-clubs, and an internal layer of long densely warty spindles. An examination of these diagnoses reveals the fact that we have here to deal with three distinct groups. The first of these is represented by the various species of the genus Juncella, and is characterized by the fact that its spicules include simple clubs. The second is restricted to the genus Nicella, and is distinctly separated by the character of its spicules, which include small double-clubs and long, densely warted spindles. The thii'd comprises Ellisella, Scirpearia, Scirpearella, and Ctenocella, which agree in having neither clubs nor long spindles, but whose spicules all include double-clubs. These distinctions may be tabulated thus : — A. Spicules include clubs (Juncella). B. Spicules do not contain clubs — (1) Spicules include extremely elongated spindles (Nicella). (2) Spicules do not contain elongated spindles (Ctenocella, Ellisella, Scirpearia, Scirpearella). Simpson — A Revision of the Gorgonellidac. 31 In view of our previous discussion on the various characters which may be considered of taxonomic importance, we may now take each of these genera in rotation. (1) CtcnoccUa. — Only one species of this genus has so far been described, so that the generic diagnosis given above is a recapitulation of its specific characters. In spiculation it is essentially of the Scirpearia-type ; and the parti'nilar kind of spicules described above is quite characteristic of the group. It corresponds to the elongated double-club, which may appro.ximate to the double-spindle, and eventually to the simple spindle which has been already described. It has been my privilege to examine a large number of colonies of this species (j^ectiimfa), and the only character in which it differs essentially from other genera is its peculiar mode of branching. The secondary and ter- tiary branches ''see figs. 36-41), however, are long, simple, and flagelliform ; and if one of these detached branches be taken for identification, it will at once be referred to the genus Scirpearia. The disposition of the verrucae and the ti/pcs of spicules correspond in every detail with the diagnosis of Scirpearia. Is it justifiable, then, to continue recognizing a genus on the basis of its branching alone, when a part of the same colony may be indisputably referred to another genus ? We prefer to answer this question in the negative, and consequently abolish the genus Ctenocella, and rank the only known species under the name Scirpearia pectinata. (2) Mlisclla. — It will be remembered that KoUiker in 1864 first drew attention to the spicules of this family, and, with the small amount of material at his disposal, separated the genus Juncella into two groups. (1) Those with clubs (J. juneea and J. rjemmacea), and 2) those witliout clubs (/. elongata). Studer(1878)in revising the family limited the generic diagnosis tlius: — (1) Spicules: clubs, and double-clubs (Juncella). (2) Spicules : double-clubs, and spindles — A. Calyces not prominent (EUisella). B. Calyces markedly projecting (Scirpearia). In discussing the question of the nature of the verrucae we pointed out that this character could not be relied upon for even specific determination, so that Studer's groups A and B, or, in other word.s, the genera EUisella and Scirpearia, cannot on this basis be regarded as distinct. In the descriptions of the various species of EUisella which have since been established no further character of generic imjiortance has lieen added, and an examination of tlie generic diagnosis of Elli.sella and Scirpearia, given by Wright and Studer, shows them to be identical. We have examined the 32 Proceedings of the Royal Irish Academy. type specimens of Ellisella, and compared them with authentic species of Scirpearia, and could find no reason for separating them. Hickson (xv, pp. 818-819), in his vahiable contribution to the study of this group, has suggested the abolition of the genus Ellisella and has united the species included under that name to those of the genus Juncella. He, however, divides the species so included into two groups — (1) those with clubs and (2) those without clubs, the former of which, as will be evident, corresponds to Juncella; and the latter, with the exception of J. spiralis, which will be discussed later, to Ellisella as defined by their spiculation. The result of this is that the genus Juncella, which was distinguished by the presence of the clubs amongst its spicules, now includes forms whose spiculation is identical with that of Scirpearia and Scirpearella. The question now resolves itself into, " How are we to distinguish between (1) those species of Juncella whose spicules contain no clubs, (2) Scirpearia, and (3) Scirpearella?" In other words, we have still to find generic characters and separate Ellisella, Scirpearia, and Scirpearella. As the result of an examination of all the type species of Ellisella, Scirpearella, and Scirpearia (with the exception of *S'. flageUum, of which, however, we have seen numerous authentic specimens in the Monaco collection), we are fully convinced that nothing in the spiculation of these types is of sufficient importance to be used as a generic character, so that it is incumbent upon us to examine in detail the other features which have been used as diagnostic. Branching. — Scirpearia is described as simple, Scirpearella as simple or very feebly branched, and Ellisella as simple or dichotomously branched. Now the question of branching, as has been already shown, is of no importance in diagnosis. Specimens otherwise identical are described in this memoir, in which one may be of great length and simple, another elongated and bifurcating, while a third may be of no exceptional height and yet very markedly branched. A very good example of this may be seen in Scirpearia furcata. Contrast (1) the specimens from Mergui — (2) that orginally described by Thomson and Henderson as Scirpearia, sp., and (3) the specimen from Providence Island, all of which are included in this report. The very fact, however, that a species of Scirpearia has been described in which branching occurs shows the futility of relying upon this feature. Nature of the Verrucae. — The question of prominent or non- prominent verrucae has already been discussed, and, as it has implicitly been abandoned by most authors, need not occupy our time here ; but it is essential to pomt out that the omission of this as a generic character almost finally necessitates the abolition of the genus Ellisella. Simpson — A Revision of the Gorgonellidae. 83 We would, therefore, in view of these considerations, put forward llie following emended classification, and proceed to define the various genera in terms of such characters as seem to warrant attention. Family GOEGONELLIDAE.. JUNCELLID GkOUP. Division 1. Spicules include cluhs (Juncella). Division 2. Spicules do not include cluhs — A. Spicules include lon, „ Kent Saville, xxi, p. 92. „ Jlexilis Studer. „ „ Germanos, viii. ., „ Hickson, xv, p. 821. „ fragilis Pddley, xxxiii, p. 347, PI. XXXI, fig. D. ,. „ var., xxxiii. :> „ Thomson and Henderson, xxxix, p. 314. Simpson — A Revision of the Gorf/onellidae. '59 Jwncella fragilis var. rubra Thomson aud Henderson, xxxix, p. 314. „ barbadensis Duch. and Mich., v, p. 22, PI. v, fig. 5. ., „ Wright and Studer, 1, p. 159, PI. xxiv, fig. 14. Juncella juncea. This is a very old species, as may be seen from the Bibliography. Pallas referred it to the genus Gorgonia; but Valenciennes, in 1841, rightly considered it as a Juncellid, and placed it in the genus Juncella ; and in this genus it has remained, and has been so regarded by most authors ; but, in 1859, Gray, for no apparent reason, established the genus Helicella to include it. No one has, however, confirmed his opinion, so that it is unnecessary to discuss its position there. The species is a fairly distinctive one ; but very little positive content has ever been given to it. Ridley, in his Report on the Alcyonaria collected by H.M.S. " Alert," says : — " Neither Milne-Edwards and Haime nor Valenciennes give details full enough to enable the student to identify their species satisfactorily with that of Pallas and Esper. In the ' Alert ' specimens and that figured by Esper the verrucae are closely packed over the cortex. In our specimen, which is about 46 inches (1150 mm.) long by 6 mm. thick at tlie present broken base and 3'5 nnn. thick at the tip, the basal end is almost smooth, the verrucae being either level with the surface or depressed below it ; towards the middle of the length they become projecting until they reach a height of about 1-25 mm. ; they are then adpressed against the surface of the cortex. A distinct median groove is to be traced along most of the stem." This was a most important contribution, and was the first description of the variation in the size of the verrucae, which is such a marked feature in this species, and which lias led to several mistakes in identification. Kolliker, in 1865, first introduced the question of spicules into this species, and gives two figures of these (Tab. xvili, figs. 45 and 46). One of these represents a thick single-club, and the second a double-club. In the many records and short descriptions which occur scattered throughout Alcyouariau literature very little furtlier was added, so that the following short description sums up the chief points upon which the species was identified. The colony is simple and elongated ; the cortex is thick ; the spicules contain clubs and double-clubs ; the verrucae vary in size in the various parts of the colony (Ridley) ; the axis is hard and calcareous ; there are usually two bare streaks in the coenenchyma. Practically no attention was paid to the extraordinary fertility of variation which occurs with regard to all the.se characters, not only in ditterent specimens, but also in different parts of the same specimen; nor 40 Prorre(h'nqs of the lioi/nl Irish Acmlcmij. was there any allowance made for different stages of development. As a result of this, three species— viz., J. fragilis, J. flexilis, and J. harhadensis — were established on what must now be regarded as young colonies of this species. The large number of specimens, which undoubtedly belong to this species which we have been able to examine in detail with respect to the differences on which these three species were based, confirm beyond doubt the opinion of several authors — notably Ridley, Studer,Hickson, and Thomson, that these cannot be regarded as distinct. I give here a short description of these three species, followed by a systematic study of a large number of specimens which may help to give a true estimate of the variability of certain characters and the constancy of others, and so form a basis for a definite specific diagnosis. J. fragilis Ridley. In 1884 Ridley established the species fraf/ilis for two specimens from Queensland with the following characteristics : — Stem long, unbranched, diminishing very slowly to the tip, which may be either clavate or sharp- pointed, flexible, and easily broken. The diameter at the base is 5 mm., at the apex 3-4 mm., except when the apex consists of a fine point. The cortex is thick and creamy-white when dry ; there is no trace of a lateral line in the upper three-fourths. The verrucae are small, about 1 mm. in height, clavate, closely adpressed against the cortex, crowded over all parts ; axis very slender, about 1 mm. in diameter at the base and hair-like at the apex ; near the base it is olive-brown, hard, and beset with longitudinal striae. The cortical spindles are the same as in J. gemmacca. He points out the following differences between this species and J. gemmacea : — (1) The verrucae are small and crowded. (2) There are no lateral lines in the upper three-fourths. (3) The colour is pale creamy-white. (4) The heads of the double stellate spicules are more abundantly tuberculated. Later, in 1887, Ridley referred, with doubt, two colonies from Mergui to this species as a variety. One of these was white or cream-coloured, the other was pale brick-red. He notes that these specimens approach J. juncea, which, he says, is distinguished from J. frdgilis by its greater size, its red colour, its larger and more distant polyp-verrucae, the presence of a space bare of verrucae above the base and by the possession of equal-ended double- stars. These specimens, he says, stand midway between_;wiia'« a,ad fragilis. Simpson — A Revision of the Gorgonellidae. 41 In 1905 Thomson and Henderson referred several fragments from Ceylon to this species. The axis was marked by longitudinal striae. In some the verrucae were nearly 2 mm. in height, and the diameter of the axis was 1 mm. ; in others tlie verrucae were much smaller, and tlie diameter of the axis was 2 mm. The spicules showed some variation from those in Ridley's specimens. Juncella fragilis var. rubra. In the same paper Thomson and Henderson establislied a new variety, namely, nthra, to include a long, flexible, complete colony, which tapered gradually throughout its entire length. There was no trace of a lateral line or groove. The verrucae were numerous and closely adpressed, measuring about 1 mm. in height. From the above descriptions it is evident that this species has no definite specific character. Ridley himself had doubts as to its distinctiveness ; but its " simple " character at once separates it from ./. fjnnmncca. The specimens from Mergui are undoubtedly-/, jimcca; but Ridley practically acknowledges this. I have examined the specimens described by Thomson and Henderson, and although these undoubtedly coincide with the description of J. fragilis, they also agree with young forms of J. juncca. A comparison of these epecimens and Ridley's descriptions, with the numerous colonies of various ages which I was fortunate in obtaining at Mergui, proves beyond doubt that this species was based on young stages ofj.juncea, so that I would suggest the merging of this species into J". /(/new. At the same time the variations in the different characters, as seen in these specimens, are of great interest, and show how difficult it is to be certain of any species on a single or even a few specimens, especially if thuy are young. The cuie.sticju » J Base. Midway. Apex. HO 1 . 156 7-5 6-5 5 5 1 25 25 2 115 5 6-5 4 0-5 3-25 3 114 5-5 5 4-5 3 1-25 2-25 4 110 5-5 7-5 5 4 0-75 2-5 3 110 8 8 4 4-5 1-75 2 6 105 5'5 7'5 6 3-0 1 3 7 101 0-0 ' 4 4 1-25 8 100 + 7 8-5 6 4-5 (0-5) 1-25 9 100 fi 5 4-5 3-5 1-25 2-25 10 91 7 ; 5 4 1-5 2-5 11 90 7 7 6-5 4 1-5 3-25 12 86 6 6 5 3 1-5 2-5 13 83 + 8-5 9 6 5 5 (0-5) 1-5 3 14 82 6 0 4 2 2 2 15 76 6-0 7 5 3 1-75 . 2-5 16 72 5 5 4 2 1-5 2 17 60 7 ' 5-5 4 1-5 2-75 18 64 5 5-5 3-5 2-5 1-25 1-75 19 61 5-5 8 4-5 3 I 25 2-25 20 61 4-5 4 3-0 2 1-25 1-75 21 60 4-5 4-5 4 3 0-75 2 22 57 5-5 6 6 3-5 1 3 23 53 4-5 4-5 3-5 1-5 1-5 1-75 24 52 4-5 4-5 4 2 1-25 2 25 49 7 5 6 5 2-5 2-5 2-5 26 45 3-5 « 5-5 3 0-25 2-75 27 41 3-5 3 2-5 lo 1 1-25 The siipertieial ililleieiiLes hclwoL'ii llio two varieties liavin-j; been noted, we may consider the general morpliology of tlie .species, and see to what e.Ktent the varieties are wortliv of distinction. 46 Proceediiiffs of the Royal Irish Academy. ' Coemnchymn. — Figs. 10 [a, h, and c] and 13 («, h, and c] are given to show the structure of the coenenchynia at the same levels as the corresponding figs. 9 (a, b, and c] and 12 (", h, and c) respectively. The coenenchynia may be differentiated into two distinct regions — (1) an outer superiicial layer, in which the polyps are embedded, and which contains no canals ; and (2) an inner layer, which is intersected by a meshwork of transverse canals. In var. « the superficial layer is mucli thinner than the inner layer ; but in var. b the superficial layer is the thicker of the two. At the tip of tlie colony there is a conical growing point devoid of polyps and having no canals. The thickness of the coenenchyma hardly varies throughout the colony. Canal Systems. — This consists of (1) a longitudinal system and (2) a transverse system. The longitudinal system is composed of two series of longitudinal canals — («) an outer, situated between the two layers of tlie coenenchyma, and with which tlie polyps connect directlj^ ; and (b) an inner, situated between the inner layer of the coenenchyma and the axis. These two systems are united by the transverse canals which penetrate the inner layer of the coenenchyma. The number of canals in each of the two longitudinal series diminishes in number from the base of the colony upwards. This, of course, is natural, since the number of polyps also decreases. In the inner series of longitudinal canals there are two, situated diametri- cally opposite one another, which are much larger than the others. They occur in both varieties ; but in the case of var. b they correspond to the position of the longitudinal bare .spaces. The greater thickness of the coenenchyma in var. /' may e.Kplain the absence of this phenomenon in the latter variety. Axis. — The axis is marked by longitudinal ridges and furrows. The number of these correspond to the number of canals in the inner longitudinal series, and consequently diminishes towards the apex, so that this character is of no taxononiic importance (fig. 11 («, b, and c)). The structure of the axis is very well seen in this species. It is composed of concentric laminae of horn, impregnated with small limy sclerites. To the inside of the canals of the inner longitudinal series a layer may be detached showing the sclerites in situ. The innermost layers are much more densely spiculose than the outer. The diameter of the axis (unlike the coenenchyma) gradually diminishes towards the tip of the colony, where it becomes almost hair-like, and contains very little lime. Simpson — A Revision of the Gorgonellidae. 47 Spicules of J. juncea. The characteristic spicule is the simple club, which has been described already. A few typical variations are also shown in fig. 14. These are also double-stellate forms and double wheels or capstans. The following are some of the measurements, length by breadth, in millimetres : — (1) Cluhs.—0-ll X 0-04; 0-1 x 0-0:55 ; 0-09 x 0-034; 0-085 x 0-032; 0-08 X 0-03; 008 x 0-02. (2) Double Stars.— Q-Ol x 0-05; 0-09 x 0-045; 0-08 x 004; 0-08 x 0-03. DistrHution of J. juncea. (1) Australia. — Port Denison, Queensland, 4 fathoms (as J. juncea and J. fragiUs) ; Dirk Hartog, W. Australia, 45 fathoms ; j\Iermaid Straits, N.W. Australia, 50 fathoms; Torres Straits, 7-11 fathoms. (2) Oft Sombrero Island, West Indies (as J. harhadensis). (3) King Island Bay and elsewhere (Mergui). (4) Ceylon Seas (as J. juncea), Gulf of Manaar (as J. frai/ilis). (5) Bourbon, and between Flat Island and Mauritius (as J. Jiexilis). (6j Ternate (as J. Jiexilis). ■ (7) Maldives (as J. Jiexilis) . (8) Off Table Island, Cocos Group, Auduinans, 15-35 fathoms. Specijic Diagnosis of J. juncea. Colony simple or sub-simple, elongate, sometimes filiform, sometimes very thick : the coenenchyma varies greatly in thickness iu the different specimens, but is constant in each ; this affects tiie external appearance of the colony. The canal system is of the typical Gorgonellid structure, and there are two main longitudinal canals. These; may or may not produce an external impression ; in colonics with a thin coenenchyma their position is denoted externally by two longitudinal bare tracts; but in those with a very thick coenenchyma, no trace of this is to be seen. The polyps are distributed differently in these two types ; iu the former they are disposed in two longitudinal series, in which there is a varying number of rows, which diminish from tlie base upwards; in the latter they are crowded all over the coenenchyma. The verrucae vary greatly in shape in the difl'crent parts of the colony ; near the base tl)cy are low and dome-like, or may even Ijo depressed beneath the surface of tiic coenenchyma ; they gradually increase in size until near the top they are usually sub-conical, directed upwards, and adpressed to the coenenchyma. The axis is hard and dexiblo ; it is composed 48 Proceedings of the Royal Irish Academi/. of concentric laminae, which consist of a horny substance impregnated with some form of calcareous matter. The surface is marked by longitudinal grooves, which correspond iu number to the inner series of longitudinal canals, and therefore diminish from the base upwards. Sometimes two larger than the others are to be seen, and these represent the position of the two main canals. The spicules consist of the usual Juncella types, and include clubs, double-wheels, and double-stars. The colour varies from pure white, through orange, to dark red. XL — Juncella gemmacea. Figs. 15-19. Gorgonia gemmacea A^alenciennes, MSS. dans la Coll. du Mus. Paris. Verrucella „ Milne-Edwards and Haime, xxvi., p. 185, B 2, f. 7. Ju7icella „ KoUiker, xxiii., p. 140, t, 14, f. 4. „ „ AVright and Studer 1., p. 158, PI. xxxiv, fig. Vi. „ „ var. Eidley, xxxii., p. 241. „ „ Studer xxxvii., p. 117. „ „ Thomson and Eussell, xliii., p. 162. Ellisella „ Gray, xii., p. 26. Ellisdla macidata (pars) Wright and Studer, 1. Juiicella elongafa vav. Eidley xxxiii., p. 346. Valenciennes in 1855 established this species to include a specimen iu the Natural History Museum in Paris, under the name of Gorgonia gemmacea (MSS. dans la collect, du Museum Paris). In 1857 Milne-Edwards and Haime leferred the species to the genus Verrucella, and defined it as follows : — " Polypi^roide dont les branches, assez uombreuses et cylindriques, se dichotomosent de loin en loin, et s'ecartent beaucoup entre elles ; les ramuscules terminau.v allonges. Coenenchyme tres-friable, d'un jaune ferrugineux a la surface et blanchatre puis de I'axe. Yerrues caliciferes tres- saillantes, arrondees et dressees centre la tige." They give a very good figure, showing the mode of branching. In 1865 KoUiker removed the species from the genus Verrucella to Juncella, and noted for the first time that " clubs " occurred amongst the spicules just as in J. junceo.. He gives two figures — (1) a club-shaped spicule (woodcut 19, 1); (2) a cross-section of the axis (Pl. xiv, fig. 4). Gray in 1870 referred this species to the genus Ellisella with no appai-ent justification. (See our Historical Xote.) This change, however, was not recognized by any subsequent authors, so that Eidley in 1884 identified some specimens from Queensland, under the name of J. gemmacea, and Simpson—^ Revision of the Gorgonellidae. 49 remarked that the spicules are almost indistinguishable In mi those of J. jioncea, while at the same time he expressed the ojiinion that ./. flcrilvi Studer might not be a distinct species. The position of this species has already been discussed. It would be useless to go in detail into all the records of tliis species; and in fact, it is very difficult to say whether the records of J. juncea and J. gcmnuwea are all correct in identification, as several authors do not mention whether their colonies were simple or branched. Another complication, however, creeps in. When fragments of colonies were examined, is it not possible that a branch of ./. r/i'inmacea might be referred to .7. jiincca, especially when we remember that the question of Inanching is the chief distinction between the two species ? Before going on to discuss the various characters of this species in detail, I would give the following quotation, as it is not only of great interest, but has apparently been overlooked by several authors in then- identification of this species. Ridley (1884) referred a colony from the X.-E. coast of Australia to the species dongata ; but in 1887 (xxxii., p. 241) ho replaced it in the species gemmacca, noting that he had overlooked the fact that J. dongata had no " clubs." He says : — " It will be seen that we probably liave a very variable species before us, colour, form, and size being alike not to be depended on liy themselves. The spiculation is fairly constant, but differs so little from that of the allied forms {J.jimcdla and fmgilis) as to be scarcely a sufficient guide per sc to the recognition of the species." From the fact that the specimen referred to was dichotomovisly branched, I feel justified in recognizing it as J. gemmacca. An interesting feature about this specimen is the fact that when found it had been broken oil' at the base, and the broken part had been overgrown with coenencliynia, so that it had been living free in the water. An analogous state was observed in the case of a specimen of Ids hippuru Linn, in the Littoral Collection from the Indian Ocean. In the Mergui Collection tlierc is a large numljer of specimens of t!iis species, and these are augmented by several from the Indian Ocean Collection from the Indian Museum, Calcutta. By means of these it has been possible to study and compare several characters which arc very variable in a manner which would have been impossible with only a single or even a few specimens. Brnnehing. — The niode of brandling is of tlie nature of a false dichotomy. The large main branches of the colony arc again branched almost in one plane, but the general appearance of the colony is bushy. II 50 Proceedings of the Royal Irish Academy. The distance between branchings, though not constant, seems to increase from the base upwards, so that the longest unbi-anched parts are the terminal twigs. This is more marked in the taller colonies. In young dwarf specimens the relative distances between the origin of the several branches in ascending order is less pronounced, and the branches themselves are pro- portionately thicker. Tliese- latter specimens therefore have a diflerent appearance from the older and more elongated colonies, but must be ranked in this species when we take into consideration the mode of growth, which will be discussed later. Before doing so, however, it will be well to tabulate corresponding measurements in individual colonies, and see how far these give us a clue to the mode of growth. Fortunately we have in our possession intermediate stages which show the different developments during growth from the shortest to the tallest. The following tables may serve to form a basis for such a study. In Table A the measurements ai-e all given in centimetres. The symbol + indi- cates that the exact length is not known, owing to the basis of attachment having been broken off. Tajble a. Specimen. i Height. 1 Length of main stem. Distance between branches. Length of twigs. I. 90 6 3-lb 26 II. SO 12 -f 2-10 28 HI. 76 4 ( 2-10 22 IV. 70 4 + 2-12 22 V. 65 5 + 2-12 16 VI. 62 5 2-11 15 VII. 5.5 4 2-9 16 VIII. 55 2 • 2-14 14 IX. 54 3 + 2-10 13 X. 53 S t 3-10 14 XI. 44 ^ i 2-10 12 XII. 42 3 + 2-7 0 13 Figures 15, 16, and 17 show the branching in the colonies, which have been proportionately reduced. The largest (lig. 10) was 800 mm. in length. Let us consider two colonies which in general build are quite unlike one Simpson — A Revision of the Gorgonellidac. 51 another. The large specimen is from tiie Mergui Archipelago antl tlie smaller is from the Andamans. The former is 40D mm. in height; tlie latter i.s 2^0 mm. in height: the longest twig in the former is 220 mm., that in the latter 70 mm. The greatest distance between branchings in the former is 120 mm.; in the latter it is only 27'o mm. So far, then, the measurements are proportionate; but when we take into consideration the corresponding diameters in the various parts, the ditterence is at once very marked. In the smaller colony the total diameter of the several branches and twigs is greater tlian in the larger ; so that, at first, it is difficult to conceive that the dwarf colony coiUd develop into a colony similar to the larger. Table B. o ■ .a c tjjO Distance between branches. = ■» .: I'M ■s .■ Distance between brandies. Diameter of brandies at origin. Length of terminal twig. 1" Dislanie lietween branches. Diameter of branches at origin. Lengili of terminal twig. 25 70 5 70 9 20 20 5-5 70 9 27-5 20 5-5 65 8 15 6-6 40 5-5 110 7 230 25 G 70 440 45 o'5 120 700 100 G 2i0 22-5 G 40 5-0 55 6 22-5 6 40 . 5 95 4-5 17-5 5-5 50 5 G5 4 12 6-5 Twig 4-5 120 3-5 Twig 6 Twig 3 Table B gives several measurements from three colonies of diHereiit sizes. One large Irrancli has been selected and followed to the lip of llu- colony. The various lengths represent the consecutive di-slunces at whicii blanches arise from it. The first feature wliieh may bo seen from this Table is tlie fact that the distances at which tlie ditlerent branches arise do not increase proportionately from the base upwards. (2) Such increase as exists is more maiked in tlie taller specimens. (8) In the very dwarf colony, the distances actually diminish in the u]ipor hall'. Let us now critically examine the meusureiueuts given in the same Table H 2 52 Proceedings of the Royal Irish Academij. of the diameters at the corresponding parts. (1) The diameters of the branches in the young colony are equal to, and, in some cases, greater than, corresponding measurements in tlie older colonies. The series of measure- ments of twelve specimens given in Table C shows that (1) the length of the main stem varies very little ; (2) there is a distinct tendency towards an increase of length in the younger- branches and twigs in the older specimens. Table C. M AIN StKS BuANCH. Twig. =+H :i =« ■~ ej <«.< C3 n H 1 Height in centimetres, Breadth in centimetres, 31 = a -a O 1 = (5 Thickness c Coenenohym 3I Hi If I. 90 20 9 8 0-5 4 2 1 2 II. 80 35 9 7-5 0-75 3-5 2 0-75 2 III. 75 24 9-5 8 0-75 4 2 1 2 IV. (0 18 9 8 0-5 3-0 2 0-75 2 V. 65 15 7-5 6 0-75 4 2 1 2 VI. 62 24 6 5 0-5 4 2 1 2 VII. 55 30 6-5 5 0-75 4 2 1 2 vm. hh 23 6 5 0-5 4 2 1 2 IX. 54 20 7 5 1 4 2 1 2 X. 53 16 7 5 1 4 2 1 2 XI. 44 20 4-5 3 0-75 4 2 1 2 XII. 42 27 ,S 6 1 4 2 1 2 In Table C we have sought to analyse the nature of the constitution of the diameter — in other words, to find the actual proportion of axis and coenenchyma ; and here several very interesting and useful facts ha\e come to light. (1) The thickness of the coenenchyma at the tip of the twigs is almost a constant ; in reality it is slightly thicker in the smaller specimens. It is noteworthy that the diameter of the axis at this part is negligible, being of a hair-like fineness. (2) Although the total diameter of branches lower down is greater than in the twigs, the actual thickness of the coenenchyma is never greater, aa4,-iu some cases, is actually less. Simpson — A Revision of (he Gorgonellidae. 53 (3) The thickness of the eoenenchyma in the main stem is seldom as great as in the branches or twigs, and is usually from 0-75 to O'o times its thickness. Bearing these facts in mind, let us see if any inference may be made as to the mode of growth, and also as to the possibiUty of these apparently diverse forms being i-eferable to a single species. (1) We find that what obtains with regard to the various measurements in different parts of the same colony also holds good in the corresponding parts of colonies of ditl'erent sizes. (2) It is also obvious tliat increase in thickness in the older parts of a colony is due, not to increase of thickness in the eoenenchyma, but to increase in the diameter of the axis. (3) Increase in length in the younger branches and twigs is not propor- tional to increase in the thickness of the eoenenchyma, but tends rather to the reverse of this situation. We are therefore in a position to conclude that increase in tiie thickness of the eoenenchyma is not proportionate to the age of the colony, but that the eoenenchyma attains to its typical thickness at a very early stage, and that further elongation and consequent thickness are caused more by the gtowth of the axis than the eoenenchyma ; or, in other words, the earlier period of growth consists chiefly in development of the eoenenchyma, while the strengthening of the axis and elongation of the colony come at a later period. Consequently the younger colonies are more bushy and fleshy, and the older colonies have proportionately a greater amount of axis, and are therefore more rigid. The distribution of the polyps, the nature of the verrucae, and the details of the canal-system are exactly similar to those described for the protruding verrucae \a.\\Qiy oi JunccUa juncea. Figs. 18 an long spindles described for the latter species have since proved to be extrinsic. An examination of the type specimen, and also of the spicules of J. racnnosn in the British Museum, has proved beyond question that J. miniacm is not distinct from J. racemosa. This has already been pointed out (Thomson and Simpson, xli.) in connexion with another specimen which occurs in the collection of Littoral Alcyonaria of the Indian Museum, Calcutta (figs'. 20, 21, and 22). In the same collection, but hitherto undescribed, is a small portion of a delicate colony branched in one plane (fig. 21). The l)ase is wanting; and what appears to be the main stem may be only a primary branch which, has been broken off at the point of attachment of an acorn shell. It is 30 mm. in length, and is distinctly crescentic in shape. Five thread-like branches arise from the convex side, and one only from the concave. The longest of these is 55 mm. ; and it is noteworthy tiiat the branches are also curved. They in turn give origin to finer branchlets, which, with very few exceptions, arise from the convex side. Two acorn-shells have become attached to the colony; and these are overgrown with polyp-bearing cocncnchyma : while one of them has given rise to a proliferation of the axis. The coenenchyma is very thin; and it is impossible to discover the nature and number of the main canals. The axis is thread-like, and is impregnated with lime. The polyps are more scattered than in any of the previously described specimens, and stand almost perpendicularly. The arrangement of these is not easily determined. In the finer twigs they occur in two single rows (fig. 22); but the intrusion of young polyps and consequent development tend to obliterate this symmetry, and give an irregular arrangement. The verrucae, when expanded, are cylindrical, and higher than broad ; in this condition an eight-rayed structure is seen at the top. On contiaction they become dome-like, and no trace of the rays is to be seen. The colour of the colony is a pale brick-red throughout. The spicules are of the types characteristic of this species. Locality. — Andamaus. Diagnosis of J. racemosa. Colony delicate, branched in one plane ; the branches tend to arise from one side of the stem, and the branchlets show a similar tendency ; coenenchyma thin ; polyps in the finer twigs and at the tips of the branches are usually disposed in two lateral rows ; in the older parts of the branches 56 Proceedings of the Roi/al frish Academ//. ihey occur irregularly over the coenenchyma. The verrucae, when expanded, are cylindrical, and either stand perpendicularly or are inclined upwards to the stem ; when contracted they are low and dome-like. The spicules are very minute, and are distinctly prickly in appearance (fig. 23). The following types may be distinguished : — (a) Slender clubs, with a distinct smooth middle portion, surmounted by a spiny head, only slightly thicker than the constriction. (b) Short, stumpy clubs, much broader in proportion to their length, and with the spines slightly more divaricate. (c) Douhle wheels, with an elongated hub. (d) Elongated forms, with a distinct smooth constriction, which may be conveniently termed double spiiidles. (e) Occasional qnudriradiafes. if) Needles in the anthocodiae. The following may be taken as typical measurements in millimetres as they occur in all the specimens so far known : — («) 0-076 X 0-031 constriction 0-008 X 0-008. 0-065 X 0-031 j> O-OII X 0-008. 0-065 X 0-023 > 0-008 X 0-008. ih) 0-053 X 0-0.34 J 0-008 X 0-009. 0-053 X 0-031 ) 0-011 X 0-008. 0-049 X 0-031 y 0-008 X 0-008. (c) 0-057 X 0-027 i 0-011 X 0-008. 0-051 X 0-027 i 0-011 X 0-009. 0-046 X 0-027 y 0-008 X 0-008. (d) 0-076 X 0-027 y 0-011 X 0-008. 0-068 X 0-027 y 0-011 X 0-011. 0-061 X 0-031 y 0-008 X 0-008. (e) 0-038 X 0-034 if) 0-1 X 0-04, 0-06 X 0-04 Colour-schemes — Br own throughou t. Vermilion-red. Orange-red, with tips of the verrucae yellow. Pale brick-red. Localities. — Hyalonema-ground, off Japan, 345 fms. (Wright and Studer). Andamans, 120 fms. (Thomson and Henderson). Andamans, „ (Thomson and Simpson). Andamans, „ (as stated above), Simpson — A Revision of the Gorgonellidae. 57 XIII.— Juncella trilineata. Figs. 24-26. Juncdla trilineata Thomsou and Henderson, xxxix., p. 315. In 1905 Thomson and Henderson established this veiy characteristic species, of which the following notes are of particular interest. The specimen was spai-ingly branched. "The polyps arise in three different bands, leaving three narrow, bare strips, each of which has in its centre a bare rib nr keel- Under each bare strip lies a large longitudinal canal." Only one specimen of this remarkable form has so far been described ; it was obtained at Patani, Siam. The spicules are of the Juncella type, and are hardly (listinguishal)lo from those of J. juncca. Very characteristic, iK)wever, is tiie presence of ^Areemain canals situated symmetrically around the axis. The result of this on the external appearance is that there are three longitudinal spaces on the coenenchyma devoid of polyps ; the veiTucae arc therefore disposed in three longitudinal groups, and this arrangement is unique amongst Juncellids. "The polyps, which measure from 1-1 mm. to 15 mm. in height, are arranged in transverse rows of .3-4 ; but many smaller polyps occur wliich break this regularity." For a short distance from the end of the branches the polyps occur in three single rows (fig. 24) ; but passing downwards two, three, four, or more are to be seen, and scattered among these are immature forms, so that all that can be said with regard to the disposition of the polyps is that they occur in three longitudinal groups, the exact number in a transverse row depending on the position in the colony and on its stage of development. The coenenchyma is of the typical Juncellid type ; it consists of an outer non -canal- bearing part and an inner canal-bearing part. These are separated by a concentric series of outer longitudinal canals. In the inner series of longitudinal canals which separate the coenenchyma from tlie axis there are, however, three much larger than the rest ; these corrcsjiond to tlio three longitudinal tracts devoid of polyps and separate ofT the three polyp- bearing ridges (fig. 25). In the Indian Ocean Littoral Collection there occurs a small, simple colony, 55 mm. in heigiit and 2-5 mm. in maximum diameter. Tlie attacli- ment has been broken off, but has evidently not been far from the present base. For a short distance from the base there is a portion devoid of polyi>s. Throughout the remainder of the colony the polyps seem to be arranged in three irregular longitudinal series, and as a cross-section reveals what are evidently the main canals, we feel justified in referring the specimen to this I 58 Proceedings of the Royal Irish Academy. species. The verrucae are about 1 mm. in height, but are not so densely packed as in the tjpe-speeimen (fig. 2-i) ; but this may be due to its immature condition. The colour is orange-yellow, but the verrucae are paler. Locality — Off Ceylon, 34 fms. Diagnosis of J. trilineata. Colony upright, sparingly branched, the branches long and slender. Very characteristic is the presence of three main longitudinal canals and the consequent disposition of the polyps in three longitudinal groups. The verrucae may show the different stages characteristic of the genus. The following types of spicules may be distinguished fig. 26,. {n) Slender clubs, with a distinct central bare portion ; the knobs on the handle stand almost perpendicularly ; the projections on the club-portion arise at a slight angle and are directed downwards. 0-068 X 0-019 length of constriction 0-017. 0-068 X 0-017 „ „ 0-017. 0-068 X 0-019 „ „ 0-015. {h) Clubs, similar to (a), but thicker in proportion to their length. 0-072 X 0-038 length of constriction 0-018. 0-068 X 0-034 „ „ 0-017. (c) Double-stars, with very few large smooth warts at each end. 0-076 X 0-038 constriction 0019 x 0-015. 0-066 X 0-042 0-015 X 0-015, 0-061 X 0-034 0-015 X 0-012 (rf) Double-wheels. 0-065 X 0-034 constriction 0-015 x 0-014, Colour. — Dark red. Localities. — Patani, Siam. Off Ceylon, 34 fms. XIII A. — Appendix to Juncella. " Incertae Sedis." Juncella santae-crucis. 1. Juncella santae-crucis Duch. and Mich., v., p. 21, t. 2, f. 1. 2. Juncella viviinella (?), santae-crucis Gray, xii., p. 29. J. "Polypario stirpe simplici, rigid o ; axe terete, lutescente, gracili; cortice cretaceo, albo ; calj'cibus irregulariter biseriatis, inaequalibus, nempe nunc majoribus nunc duple minoribus ; ore terminali, parvo, radiato." Simpson — A Revision of the Gorgoncllidae. 69 " The polyps are irregularly disposed in a ilouble row on each side of the colony ; there is a median bare space on each side of the two flattened faces. The verrucae are unequal in height, and stand at right angles to the colony ; they are conical in shape ; the summit has a small opening which shows a radiated structure." Duchassaing and Michelntti had, however, only a fragment devoid of base; the breadth was 5-.5 mms., including the verrucae, the longest of which were two mms. in height. 2. Coral simple, rigid ; axis cylindrical, yellowish, slender ; bark cretaceous, white ; cells irregularly disposed in a double row on each edge of the stem, unequal; some twice as large as the others, smooth, terminal, small, and radiated ; lateral area flat and naked, with a central groove. ioca^i^,— Island of St. CroLx (West Indies). Juncella funiculina. Juncdla funiculina Duch. and Mich., v., p. 22, PL vii., ligs. 9 and 8. Colony simple, flexible; polyps in a single series on two sides, small, adpressed to the stem, and directed upwards ; oval opening small, with a radiate structure ; coenenchyma thin, white ; axis yellowish. Local ity. — Guadalo upe. Juncella barbadensis. Juncella barbadensis Duch. and ilich., p. 22, PI. v., figs. 5 and 6. Colony attached, simple, filiform, white; polyps elongated with club- shaped spicules ; verrucae in a single series on each side ; there is a distinct median groove on each liare space. It is larger and ni(in> robust tliaii J. funiculiTui ; the verrucae are larger. Localities. — Barbadoes and Guadaloupe. Juncella calyculata. Gorgonia calyculata Ellis and Solander, vi., p. 9-5. Juncella calyculata Valenciennes, xlvi. Gorgonella calycnlnta KoUiker, xxiii., p. 140. MHsella calyculata Gray, ii., p. 26. Ellis and Solander's description is as follows :— This Gorgon grows in a sub-divided order, having erect, thick branches with truncated papillae. The flesh is ash-coloured without, and purple on the inside, furnished with large, cup-shajjcd mouths, dispo.sed close together in a (|uincunx order, and looking upwards, having polyps wiih eight fringed claws extending themselves from them. TIh; bono is of a dark-brown I '2 60 Proceedings of the Royal Irish Academij. colour and horny nature. This sea-shrub sends forth round white eggs, larger than any of the genus. Locality. — Isle of Bourbon. Juncella hystrix. Juncella hystrix Valenciennes, Comptes Eendus, xli., p. 14. Juncella hystrix Milne-Edwards and Haime, O'orall., i., p. 186. Juncella hystrix Johnson, xix., p. 143. Juncella hystrix Johnson, xviii., p. 506. The only description of any importance of this species is as follows : — Stem delicate. Verrucae markedly projecting. Locality. — Bahia. Juncella vimen. Juncella vimen Valenciennes, Comptes Eendus, xli., p. 14. Juncella vimen Milne-Edwards and Haime, CoralL, i., p. 186. Milne-Edwards and Haime describe this species thus : — Verrucae disposed laterally in such a manner that a large distinct non-polyp-bearing median space is left. Locality. — Isle of Bourbon. Juncella surculus. Juncella surculus Johnson, xviii., p. 506. Juncella surculus Johnson, xix., p. 143. Locality. — Senegal. Juncella laevis. 1865. Juncella laevis Verrill, xlviii., 1865, p. 189. 1870. Juncella laevis Gray, xii., p. 29. Verrill's original description is as follows : — " Corallum tall, simple, subcylindrical, rather slender, diminishing in size both at the summit and near the base, where the polyps become obsolete. Cells adpressed, scarcely prominent, arranged in two broad bands, leaviny a narrmv, median, naked space on each side, along which there is a well-nuirked groove ; they are placed alternately, at a distance of about one-fifth (-2) inch, in about six vertical rows on each side, pro- ducing a quincunx arrangement ; axis slender, cylindrical, calcareous, white, surrounded by about sixteen longitudinal lobes, two of which are larger and correspond with the lateral grooves ; the others to the rows of polyps. Length of the single specimen, imperfect at each end, 20 inches ; greatest diameter, \ (•25) inch. Colour yellowish-brown, in alcohol." Simpson — A Revismi of (he Oorfjonellidne. 61 Locality. — Hong-Kong, China. Gray (xii., p. 29), not having seen the specimen, simply recapitulates the above description. Juncella extans. Juncella extans, Verrill, xlvii., p. 37. " Tall and simple, writh the very prominent verrucae curved inwards, and arranged crowdedly in a baud on each side of the axis, leaving a wide, naked space on each side. Colour wliite. Axis greyish-white, stony, and rigid." Locality. — Fayal, Azores. XIV. — Scirpearia emend. (a) Discussion of the Genns. 1830 Scirpearia, Cuvier, i. p. 319. 1878 Scirpearia, Studer, xxxiv., p. 660. 1887 Scirpearia, Studer, xxxv., p. 67. 1901 Scirpearia, Studer, xxxvii., p. b'l. 1889 Scirpearia, Wright and Studer, 1., p. Ixv. 1889 Scirpearella, Wright and Studer, 1., pp. Ixv and 1,J4. 1855 Ctenocella, Valenciennes, xlvi., p. 14. 1857 EUisella, Gray, x., p. 287. This genus was establislied by Cuvier iu 1830 to include Pennatula mirabiiis, but the following note may be interesting : — Milne-Edwards ami Haime (Hist. Nat. Corall., 1. 0. 214) say: "The Alcyonarian described ami figured by Cuvier under tlie name Pennatula mirabiiis seems to be very little connected with Virgidaria mirabiiis, as some have suggested. It has a slender stem attenuated at the two extremities, and bearing at each side a simple series of widely separated polyps. Cuvier formcil of it the genu.s Scirpearia, which has been adopted by Elirenberg. Lamarck placed it in his genus Funiculina, near Pavouaria, under the name of Funint/ina ci/llntlrita. Fleming thought that tlie species was not distinct from Virgularia ; and Blainville affirmed Lliat it was nothing but a (Jorgonia. None of ihcse opinions seem to me admi.ssible. It is too imperfectly known tu have a place assigned to it iu a scientific classification of corals." 1 11 1878 Studer resuscitated tlie genus, and gave the following diagnosis : — "Colony simple or branched; axis cylindrical, bnniy, and calcareous; coeuenchyma thin; calyces projecting; in two longiludinal rows on the ^^i(lcs of the stem and branches ; spicules double-clubs and spindles." This, then, must lie our starting-point in generic determination. 62 Proceedings of the Royal Irish Academy. In 1901 he re-united under the name of Seirpearia all the Gorgonellids with a simple, flagelliform colony which have large verrucae in the form of clubs, and whose spicules are double-clubs and spindles. The coenenchyma is thick and the colony is bilaterally symmetrical. The polyps are disposed on two sides of the axis. Wright and Studer in 1889 give the following diagnosis : — " Colony simple with a cylindrical calcified axis and thin coenenchyma. The polyps are seated in two longitudinal rows on each side of the stem. The spicules are double-clubs and spindles. The genus may include Seirpearia viirahilis Cuvier and Viminella Jtagellmn Gray." It must be remembered, however, that in the same memoir they separated off the genus Scirpearella as follows: — "Colony simple or very feebly branched. Axis calcareous, brittle, smooth, or grooved. Polyps arranged in rows or spirals, retractile, with more or less prominent verrucae. Coenenchyma is moderately thick and finely granular. The spicules are spiny spincUes and double-clubs. We have already shown, however, that such a distinction cannot be said to obtain, and have already proposed the uniting together of Seirpearia, Scirpearella, EUisella, and Ctenocella, and have given an emended diagnosis. In this emended sense, then, we now proceed to classify specimens with these characters into different species. Before doing this, however, it might be well to give the following list of the various species which have, from time to time, been referred to the genus under consideration, under the names Seirpearia, Scirpearella, Ctenocella, and EUisella :— Seirpearia fiagellum. Seirpearia fureata. Seirpearella profunda. Scirpearella gracilis. Scirpearella rubra. Scirpearella indica. Scirpearella atirantiaca . Scirpearella alba. Scirpearella divisa. EUisella maculata. EUisella calarmcs. EUisella coccinea. EUisella elongata. Ctenocella pectinata. Simpson — A Revision of the Gorgoncllidae. 63 Some of these — e.g., jia^idlum and elatu/ata — have, at times, appeared under other generic names; but these will be discussed later under the species in question. {h) Classification of the Species. In formulating a scheme of classification for these different species, and also th(i large numlier of undescribed specimens which I have before me for identification, two courses were available, either (1) to describe every individual specimen, and name it on account of certain differences which may or may not be inherent, or (2) to study the group as a whole, tabulate all the points of difference in the various specimens, eliminate all variations, such as occur in the same colony, reject all environmental modifications, and arrange the specimens around some central type. The latter plan has been adopted in the present work ; and for this reason it has been necessary to aliolish several of the previously described species, not on account of their absolute identity with formerly described species, but on account of the differences wiiich obtain in these different forms having proved to be not greater than differences appearing in an individual specimen. A very good example of this is seen in the case of Scirpearia furcata. Such a procedure has been possible in the case under consideration only on account of the large number of specimens which it has been my privilege to examine ; and it is more than probable that when a larger mass of material is available, it may still be possible to diminish the number of species in this report. The characters on which the present classification are based are the following : — (1) the number of main longitudinal canals, (2) the nature of the spiculation, and (3) the nature of the branching. These, of course, are not all of equal value; imt a reri/ rigifi .ir^tnmtion may be made into two classes based on tiio number of nuiin longituilinal canals. It has been found that in this group specimens have either (ur> or four main canals. The nature of the branching wIumi it comes to be a question of "simple or branchcii," as we have already pointed out, is of little value except in certain well-defined species. This is very evident in such colonies as those described under Scirpearia furcata. On the other hand, the very character- istic mode of branching seen in Sciri)C'iria pectiiwta woulil seem to justify its inclusion as a specific character. Scirpearia andamanensis and Scirpearia ramosa are also worthy of consideration in this respect. 64 Proceedings of the Royal Irish Academy. The nature of the spiculatioii is also a character on which great reliance may be placed as a specific deteriniuaut : and, in the case of Scirpcaria, it has proved to be of great value. Very good examples of this may be seen in the great contrast between the spicules of Scirpcaria jn-qfimdn and Scirpcaria alba, or between Scirpcaria flagcllwm and Scirpcarifi ramosa, or between Scirpearia furcata and Scirpcaria thomsoni. By means, then, of a combination of these characters, it has been possible to arrange the numerous specimens which have been examined into certain fairly definite groups. It will be seen that in the great majority of cases each group is represented by a single species ; but where possible we have suggested affinities. It seems preferable, however, to designate these at present as groups rather than as species, although the latter procedure must also be used for reference. It is unnecessary to enter into the details of each group here, as that is much better left over until the various specimens are discussed; but we submit the following classification : — SCIEPEAEIA. Division 1. — Main Longitudinal Canals, two in number: — (a) profunda-group. (6) pectinata-group, (c) elongata-group, (d) flagellum-group, (e) thomsoni-group, (I/) alba-group, (ff) aurantiaca-group, (h) furcata-group, (i) andamanensis-group, (j) ramosa-group, (k) ceylonensis-group, {I) maculata-group, . Scirpearia profunda emend. Scirpearia hicksoni n. sp. Scirpearia verr^icosa n. sp. Scirpcaria anomala n. sp. Scirpearia pcctinata emend. Scirpearia clongaia emend. Scirpeariu Jlagellmn emend. Scirpearia thomsoni n. sp. Scirpcaria alba emend. Scirpearia aurantiaca emend. Scirpcaria furcata emend. Scirpcaria andamanensis n. sp. Scirpearia ramosa n. sp. Scirpearia ceylonensis n. sp. Scirpearia tnacidata emend. Division 2. — Mai7i Longitudinal Canals, FOUK in number : — (a) quadrilineata-group, . Scirpcaria quadrilineata n. s.p. Simpson — A Revision of the Gorgonellidae. 65 Profunda-group. This group is characterized by the enormous size of the spicules. The two chief types which occur are : — (1) Double-clubs with almost hemispherical ends, and (2) Elongated double-clubs, which approach double-spindles and even spindles. Four species may be recognized : — , 1. Scirpearia profunda Wright and Studer emend. 2. Scirpcaiia hicksoni n. sp. 3. Scirpearia verrucosa n. sp. ■i. Scirpearia anovuda n. sp. The following differential diagnosis of the spicules of these four species may be useful : — Scirpearia profunda (emend.). In this species the spindle-type predominates over the elongated double- club. The spindles are massive, very warty, and irregular in outline (fig. 27). Typical measurements are 0"122 x 0057 ; 0'114 x 0 049 ; and a more slender type 0106 x 0034; 009 x 0034. The double-clubs have almost hemispheri- cal ends, and have practically no constriction, 0'084 x 0046 ; 008 x 0053. Scirpearia hicksoni n. sp. The spicules of this species are very regular in outline ; they are covered with slightly papillose warts ; and the elongated double-dubs have extremely blunt ends Fig. 31). (1) double-clubs :— 008 x 005; 0-075 x 005. (2) Elongated double-clubs :— Oil x 0045; 0-085 x 0-035. Scirpearia verrucosa n. sp. In this species the spicules are very irregular in outline ; they arc covered with long papillose warts, which are widely separated. The ends of the elongated double-clubs and doulile-spindles are markedly ])oiiited, and have the form of elongated cones (fig. 33). (1) Double-dubs:— 0-0!>5 x 005; 007 x 004. (2) Elongated double-clubs .—014 x 004; 0-11 x 002. Scirpearia anomala n. sp. The spicules of this species aie not den.sdy covered with warts, and the warts themselves are only slightly papillose. Tlie ends of the elongated double-clubs and double-sjundlt's arc markedly conical (fig. 35). (1) Double-dubs:— 0-061 x 0042; 006 x 0-04. 2) Elongated double-du 1 s .O'lo x 0-034; 0-095 x 0-046. 66 Proceedings of the Royal Irish Academy. XV. — Scirpearia profunda (Wright and Studer). Fig, 27. Scirpearella profunda Wright and Studer, 1., p. 155, PI. XXXI., fig. 2 ; Ph xxxii., figs. 1 and 1« ; PL xxxiv., fig. 7. Scirpearella gracilis Wright and Studer, 1., p. 156, PI. xxxi., figs. 1 and In ; PI. XXXIV., fig. 6. Scirpearella rubra Wright and Studer, 1., p. 157, PI. xxxiv., fig. 5. ScirpeareJla moniliformc Thomson and Henderson, xL, p. 82. We have examined the type specimens in the British Museum of these three species, and have come to the conclusion that they cannot be regarded as distinct. S. rubra is undoubtedly the same as S. gracilis ; but S. profunda difters in that it is branched. When we take into consideration, however, the great length of the tiagelliform branches of S. profunda, we are quite in a position to conceive the longest fragment of S. gracilis as a portion of a branch of a much larger colony than that formed by S. profunda. These two species, as Wright and Studer observe, " were taken at the same haul of the dredge from a depth of 130 fms. ; although, no doubt, closely related forms, there seem sufficient differences to justify their being for the present treated as distinct." The spiculation is essentially the same in all three species ; and the nature and distribution of the verrucae show variations not greater than those in other cases of specimens undoubtedly belonging to the same species. For this reason we feel justified in merging the three species under the earliest name, S. profunda. The following are the chief characteristics of the different tj-pes : — 6". pi-ofunda. — The colony is feebly branched. The axis is calcareous, brittle, and of a circular outline, with some spiral grooves; it is formed of several concentric calcareous layers, which easily peel off. The polyps are in irregular spirals on the stem and branches, from 2 mm. to 3 mm. apart, but closer to one another towards the termination of the branches. The older verrucae are more conical than the younger ones. When fully retracted they are oblong conical. The coenenchyma is moderately thick and finely granidar. The colour in spirits is a whitish-brown. Loc'ditij. — "Challenger" Station 177, off the Xew Hebrides; depth, 130 fms.; bottom, volcanic sand. e recognized — (a) slender and very warty, and (b) more massive spindles. In addition to these, the most deliuile type is the large double-club ; tliese have very massive warty ends, and practically no constriction, and some have more hemispherical heads tlian the K 2 68 Proceedings of the Roifal Irish' Academy. others (c). These three types may be regarded as characteristic ; but other forms occur — e.g. irregular forms {d} : double-wheels (e), crosses {g). The irregular forms (d) show variations which might be regarded as departures from double-clubs or from the massive spindles, and may be intermediate between the types ^ij and (c;. In the same way those represented by (/) may be looked upon as annectant forms between types {e) and {g). The crosses (g) show great variation. The following measurements in millimetres may be taken as t}'pical : — (a) Spindles, slender and very warty : 0-106 x 0-034 ; 0-10:] x 0-30 ; 0-09 X 0-03-4. (J) Spindles, very warty and massive: 0-122 x 0-057; 0-118 x 0-057; 0-114 X 0-049. (c) Double-clubs, with massive warty ends and practically no constriction ; 0-084 X 0-046; 0-072 x 0-046; 0-08 x 0053. {d} Irregular forms : 0095 x 0-053 ; 0-095 x 0-046 ; 0-076 x 0053. (e) Double-wheels, a few : 0072 x 0-034; 0-057 x 0-027. (/) Crosses: 0-11 x 0076; 0-084 x 0-061; 0-061 x 0-034. amended Specific Diagnosis. The colony is simple and feebly branched ; in the latter case the branches are long and tlagelliform. The axis is calcareous and brittle ; it is composed of concentric layers ; the surface is marked by longitudinal grooves ; some- times two of these are deeper than the others. The polj'ps are disposed in two longitudinal series ; this arrangement may be obscured in the older parts ; and then the disposition may simulate a spiral. Near the base four rows may occur in each series ; but this number diminishes in the younger parts, so that near the tip there is only a single row, alternately on opposite sides. The verrucae when partially retracted are conical, but when more fully withdrawn are nipple-like. The canal-system is typical. The coenenehyma is moderately thin and finely granular. The spicules are characterized by the presence of large, thick, warty spindles longer than the large double- clubs. XYI. Scirpearia hicksoni, n. sp., figs. 28-31. It has been found necessary to establish a new species for two portions of what must have been a very long, simple colony ; they are not continuous, however; and judging from tlie difference in the diameter of the axis in the two parts an intermediate piece of considerable length must have been lost. The base is wanting, and this must also have been some distance from the Simpson — A Revision of the Gorgonellidae. 69 present basal part, so that the colony wlieii complete must have been of great length. The lower of the two parts under examination is 18 cm. in lencrih, the upper part, which bears the tip of the colony, is 15 cm. The axis at the present base is 2-5 mna. in diameter, and tapers after 18 cm. to 2 mm. In the upper portion the axis tapers from 1-5 mm. to a fine point. Thus we see that the part of the colony having an a.xis varying from 2 mm. to 1-5 mm. is wanting; and this at the lowest estimate cannot have been less than 18 cm.', so that, without taking into account the basal part, the colony could not have been less than 50 cm. In all probability the total length would have exceeded 70 cm., so that we are dealing with a very long, simple fiagelliform colony. The surface of the coeneuchyma is coarsely granular, and, especially on the verrucae, there are numerous ridges formed by aggiegalions of spicules (cf. Huhcrogorijia ornata, Thomson and Simpson). The coenenciiyma proper is extremely thin ; but the large size of the verrucae renders this feature less evident. The general colour of the colony is brick-red ; but wliere the anthocodiae are not retracted they appear as wiiite specks on the tips of the verrucae. The polyps have a very characteristic arrangement; but this cannot be regarded as specific, as it is only superficial, and may have been caused during the process of killing. In the lower portion of the colony about one-third of tlie surface is bare ; and the verrucae seem to arise in the same plane on either side, and are continuous with it (fig. 28). This, of course, causes a crowding on the other two-thirds. On the side diametrically opposite the above bare space there is also a tract devoid of polyps (fig. 29). In tiie upper portion this arrange- ment is still visible ; towards the tip of the colony, however, the polyps seem to be distributed all round the coenenchyma ; but a trace of the bilateral arrangement is still discernible (fig. 30). The verrucae are large and have the form of truncated cones ; they stand perpendicularly to the coenenchyma. Tlie largest are 4 mm. in height and 2'5 mm. in diameter at the base; but towards tlie tip of the colony they are only 2'5 mm. in height and 1 mm. in diameter. The larger of the verrucae are markedly conical ; Imt llie younger forms are very much flattened owing to the contraction of the lliin walls ; they are then less definite in position ; and many iiave their tips eitlior incurved or directed upwards. When partially retracted they have a very marked eight- rayed structure at the summit, and show eight to twelve longitudinal ridges formed by segregations of spicules. 70 Proceedings of the Royal Irish' Academi/ . The anthocodiae are verj' ininute ; the tentacles are short and white, and bear one row of pinnules. They are first infolded, and then the tip of the verruca is introverted. The canal system is typical ; and the two main canals are evident in a cross- section, corresponding to the two bare spaces. There is no inequality in their size, so that we are justifled in concluding that the apparent arrange- ment of the polyps is due to contraction while killing. The axis is cylindrical and very densely calcareous ; it is composed of concentric laminae. Near the base it is dark brown in colour ; but in the younger part it is of a golden-yellow hue. The surface is mai'ked with irregular longitudinal striae which correspond to the inner series of canals. There is no suggestion of two depressions larger' than the others. The spicules of this species are very characteristic (fig. 31). They are very regular in outline, and are covered with warts, which are slightly papillose at the summit. The elongated double-clubs are extremely blunt at the ends. The following are the chief types, with measurements, length by breadth, in millimetres : — {a) Large double-clubs with a short broad constriction. The ends are almost hemispherical ; the warts are few in number, large and papillose: O'OS x 0-05; 0-075 x 0055; 0-075 x 0-05. (h) Smaller double-clubs very similar to the above : 0-06 x 003 ; 0-05 X 0-025. (c) Elongated double-clubs with round ends. In some of these the constriction is very marked, while in others it is hardly visible, so that this type passes through double-spindles to simple-spindles. They are covered with few, large, papillose warts: O'll x 0045; 0-11 X 0-035; 0-1 X 0-04; 0085 x 0035; 0-07 x 0-02. Zoca?i not very calcareous, and the surface has only very iiidefuiite longitudinal striae. The spicules (fig. 33) of this species are extremely characteristic ; they are covered with long papillose warts, which are for the most part widely separated, and so give a very irregular outline to the spicules. The ends of the elongated double-clubs and double-spindles are markedly pointed, and have the form of elongated cones. Tlie following are the chief types, with measurements, length by breadth, in millimetres : — (fl) Double-clubs with a short, broad constriction, with almost hemispherical ends and with large, slightly papillose warts : 0-095 x 0-05 ; 0-09 X 0-045; 0-08 x 0-05; 0-07 x 0-04. (b) Elongated double-clubs approaching double-spindles ami even- spindles ; these have markedly conical ends ; the constriction may be more or less definite ; and they are covei-ed with relatively distant, long, papillose warts: 0114 x 004; 0-13 x 0-035; 0-11 X 0-03; 0-11 X 0-02. Locality. — Andamans Sea, 55 fms. XVIII. — Scirpearia anomala n. sp. Figs. 34 and 35. This species has been establislieil for a small, complete, simple colony in the Littoral Collection of the Indian Museum. It is 17 cm. in length, attached to a piece of decayed shell which is overgrown with I'olyzoa and worm-tubes. The diameter of the colony near the base is 1-75 mm. ; midway it is 2 mm., while near the apex it [is 15 mm. ; so that there is only a slight gradation. 72 Proceedings of the Royal Irish Academtj, The coenenchyma is moderately thin aud finely granulav ; the general colour of the colony is orange-yellow; but the yerrucae are reddish. The polyps are confined to two longitudinal, lateral tracts, separated by two bare spaces. Near the base of the colony, and also in the younger part near the tip, there is a single i-ow of polyps in each series ; but in the middle part there are two irregular rows, owing to crowding and the interposition of young polyps. The verrucae, when only partially retracted, are cylindrical, elongated, and slightly turned towards the coenenchyma fig. 346). Near the base (fig. 34a! and the tip fig. 34c: they are almost completely retracted, and then appear as low warts, and may even be sunk within pits in the coenenchyma. The great majority of tlie verrucae are directed upwards, but some are turned downwards. When expanded they are about 1'25 mm. in height and 1 mm. in diameter at the base. The canal system is typical ; the two main longitudinal canals are only slightly larger than the others. The axis is cylindrical, tapers only slightly, and is calcareous. The colour varies from brown to yellow ; the surface is marked by faint longi- tudinal striae. The spicules of this species (fig. 35) are very characteristic. They consist of double-clubs, double-spindles, and some which approach spindles. They are not densely covered with warts ; while the warts themselves are only slightly papillose. The following are the chief types, with measurements, length by breadth, in millimetres : — (ff) Small double-clubs, with almost hemispherical ends, and irregularly covered with small papillose warts aud with a short, broad con- striction: 0-061 X 0-042; 0-06 x 0-04. (h) Slightly elongated double-clubs very openly warted and with relatively blunt ends: 0-11 x 0-06; 0-095 x 0-046; 0-099 x 0-049. (c) More elongated double-clubs, approaching double-spindles and even spindles. The ends are markedly conical, and the constriction is more or less definite : 0015 x 0-034; 0-08 x O'OS. Locality. — Andamans. Pectinata Gkouf. This group is easily distinguished by the character of the spiculation, but also, and more readily, by its unique type of branching. Simpson — A Revision of the Gorgonelliilae. 73 XIX. — Scirpearia pectinata enieiul. Figs. 36-45. Keratophyton seba Thesaurus, t. Ill, p. 193, PL cv., fig. 19. Gorgonia pectinata Pallas, xxvii., p. 224. » ,, Pallas, xxviii., p. 179. Lamarck, xxiv., t. 11, p. :!20, et 2n(l edil., p. 498. Pterogorgia „ Dana, cxi., p. 652. CtenoceUa , „ Valenciennes, xhi., p. 14. ., ,, Milne-Edwards and Hainie, xxvi., t. 1, p. 185. „ „ Eidley, xxxiii., p. 348. „ „ Studer, xxxvii., p. 119. GorgoncUa „ Kolliker, xxiii., p. 140, PI. xviii., fig. 41. This species, as we have already pointed out, is the sole representative of the genus formerly known as CtenoceUa ; .so that the diagnosis of that genus in the early records summarizes the specific characters. Valenciennes, in establishing the genus (Comptes Eendus, t. xli., p. 14), gave the following generic diagnosis : — " Le sclerobase s'allongeant en baguettes droites et pectinees d'un .seul cote de la tige priiicipale." Milne-Edwards and Haime in 1857 refer to the genus as follows : — " Polypi^roide s'allongeant en baguettes droites et pectinees d'un seul c6t^ " ; and also : " Polypieroide dont la tige et les branches sent cylindriques et ressemblent beaucoup aux Juncelles. Sclerenchyme sub-verruqucux. Couleur jaune-rougeatre. " Zoca^i^^.— Mers de I'Inde." Wright and Studer (1., p. Ixvi) gave tlu; fnllipwing diagnosis: — " The colony is branched in one plane ; and so tliat all the simple twigs arise in an ascending order from the upjicr surface of the stem. The verrucac are short on two sides of the twigs. There ar» distinct median furrows. The spicules are warty double-clubs ; those of the polyp-calyces are, according to Ridley, somewhat different from those of the coenenchyma, being longer and pi'ovided with two, often three, whorls of tubercles. Tlio inner whorl so approach in the middle of the spicules that the median naked zone which it. characteristic of the spicules of the coenenchyma is here absent." With regard to the "Alert" specimens, Ridley says: — "The front and back of the two main (outer) branches are bare of polyps for from one-third to half their length from their origin. Tlie verrucae arc but slightly prominent on the outer branches. The colour is pale salmon." Localities, — Warrior Reefs, Torres Straits, 12 fathoms. L 74 Proceedings of the Royal Irish Academy. Of the spicules he says : — " The verrucae spicules show a modification of the same type as those of the general cortex, being only more elongated than those, and bearing two or sometimes three distinct whorls of tubercles, besides a few median terminal ones on each half of the spicule ; the two inner whorls almost meet in the middle, so as to obliterate the median bare zone, which is characteristic of the cortical spicules." While working on the coast of Lower Burmah I was fortunate in obtaining a large number of this very interesting species ; and these have formed the basis of a somewhat detailed study. The following table gives a few of the measurements of some of these ; and notes have been added where it was considered necessary. Taken in conjunction with the various paragraphs which follow, it may serve to elucidate the more important characters of this species. 1^ n saqouviq " itjEtniJci JO is tC o rf CO <£> r» 00 00 CO sdij naa.ttjaq iC CO Tj< "^ ■^ ;d ■^ iO o ■9'! ji'puaadg ., .2-3 00 - QJ O 0) is 2 „ eat ry b aris O 00 0 0 0 o o 3 O to & 0 o «s "S ^ o oo »o 00 ID .re 1^ 00 in t« \ts > + •< i^ 00 Ci r^ o t^ CI ••*« r* ■* •^. c^ CO CO (M -f CO eo n CO .C A Ph a ^ is lO ^ + O r- 00 oo t* OO «n ■^ CO CO o ■* eo N CO g i i| oo to r- r- r- ^ to *» < ^ »o .o + S bo J §" C-l CO C-l o CO t'- ■<• eo ■* •uaonaadg t-J ^ > >■ >• t— t > !d H L2 76 Proceedings of the Royal Iri ah 'Academy. The following notes on some of the aberrant specimens may serve to give an idea of the inherent specific character : — II. One of the jnimary branches has been broken off after a distance of 28 cm. ; but the branch which arises nearest that point has developed twigs on the inner side, and has so continued the general development as if primary. VII. One of the primary branches, along with the first two secondaries which arose from it, has been broken off; but the fourth has taken its place, and continued the regular development of tertiaries just as if they were secondaries. IX. One of the primary branches is only feebly developed, and has six short slender secondaries. The second secondary has developed tertiaries after the manner of a primary. X. A similar mode of development to that described for II. has taken place in this specimen. Bratuihhig. — The branching of this unique type is extremely characteristic. The main stem is usually very short, and gives rise to two branches dichoto- mously ; these arise at varying angles in the different specimens. In some they lie almost horizontally (fig. 36) ; in others they are inclined at 45° (fig. 37), or even 60° (fig. 38), to the horizontal. In a typical specimen these primary branches give rise to secondaries on the upper inner aspect in a symmetrical manner, giving a distinct comb-like arrangement. The angle at which these arise is very characteristic. When colonies are preserved in spirits or dried, they usually contract, so that the secondary branches overlap on either side (fig. 39) ; but a study of these, when immediately taken from the water, shows that this does not occur when growing freely. In this condition all the secondary branches stand vertically, and arise from the primary branches at an angle complementary to that at which the primary branches arise from the main stem. Thus if the primary branches are horizontal, the secondary aiise at right angles ; if the primary branches arise at an angle of 60 from the main stem, the secondary branches come off at an angle of 30''. Stages between these are of course not infrequent. Fig. 3 shows the habit of a colony in the contracted condition, while figs. 36-38 show different angles of origin. Secondary complications sometimes occur in the branching, but it is noteworthy that these tend to follow the type already described. For example, in several specimens one of the primary branches has been broken off ; but the secondary branch which arose at this point has developed tertiary branches in a manner analogous to the primary branch (fig. 40). Occasionally SiMPisoN — A Revision of the Gorgonellidae. 77 also the majority of the secondary branches may be only feebly developed, but one may give rise to a large number of tertiaries. Sometimes, for no apparent reason, tertiaries may arise from the secondary liranches; but in all cases these arise on the inner side and ascend vertically, thus maintaining the specific type of branches (fig. 41). Main Caiuds. — In every tertiary and secondary brancli there are two large canals running from end to end; these correspond with the bare portions of the coenenchynia, and are consequently in the plane of branching. In dried specimens their position is usually denoted by a groove due to the collapse of the canal walls. In young colonies and in the upper part of large colonies these secondary canals unite with the canals in the primary brancli, one on either side ; but towards the base of older colonies they do not all unite ; but the last three to ten may run parallel in the primary branches, and so pass into the main stem, where as many as twenty may be visible (figs. 42 and 43). Distrihution of polyps. — In no case do polyps occur on the main stem. On the primary branches tliey are restricted to the outer aspect, i.e., the side diametrically opposite the one from which the secondary branches arise. On the secondary branches they are disposed on the two inner surfaces — i.e., the surfaces in the plane of branching are bare (fig. 44). In tin- upper lialf of the secondary branches, however, the polyps may encroach on the bare spaces, and appear as if distributed all over tiie coenenchynia. Nature of the, verrucae. — In the younger parts of the colony the verrucae are low ami dome-like ; but in the older portions they seem to become smaller, and in the lowest parts may appear as pit-like depressions. Fig. 1 shows the structure of an expanded polyp. Spicules. — The spicules of this species might be said to consist almost entirely of double-clubs, or, at any rate, of double-clubs and double-spindles (fig. 45). It is possible to group these into several distinct types which )iiiiy show an evolution-series. It is noteworthy, iiowever, that all are practically of the same length, so that it is improbable that they are dilferent stages in development. The following groups, with their measurements, length by bieadth, in millimetres, may be distinguisheil : — {a) DuuhU-clvhs with hemispherical ends and a narrow Ijaie const riction definitely marked oil': 0057 x 0038; 005:'. x 0-05:< ; 0053 x 0-034. (6) DouUe-clubs with the "heads" much more open than in (a), i.e., there is a distinct whorl of warts on either side of the constriction, and the "hub" is very warty: 0057 ■< 0038; 0057 * 00:i4 ; 0057 X 0031. 78 Proceedings of the Royal B'ish- Academy. (c) Douhlc-chihs with still more open ' heads," i.e., one whorl of warts on either side of the constriction, and the " hub " with only about three warts. These approach fZoiiJ^e-M'/tce/.s : 0*057 x 0-038; 0-057 x O-OSi; 0-053 X 0-031. [dy More slender double-chd)s with a proportionately longer constriction, and with no definite arrangement of the warts of the " heads," which might be termed .divaricate : 0-053 x 0-031; 0-053 x 0-027. (e) More elongated doxMe-duhs which approximate to double-spindlcs. The warts are large, but have no definite arrangement : 0-061 x 0-023 ; 0-057 X 0-023 ; 0-057 x 0-021. (/) DoMe-spindles not markedly warty (in some there is hardly any constriction): 0-057 x 0019; 0-057 x 0-017; 0-057 x 0-015. {(j) The type figured as {g) is evidently a developmental form of one of the other types : 0-046 x 0-023 ; 0-046 x 0-022. A small portion, about 20 cm. long, of a primary branch of what has evidently been a large colony occurs in the Littoral Collection of the Indian Museum: twenty-seven secondary branches arise from it; all are simple except one which is dichotomously branched ; the longest is 17 cm. in length. The surface of the coenenchyma is granular ; the thickness attains a maximum of 1 mm. The polyps are disposed irregularly ; on the primary branch there is one distinct bare tract, with a fairly deep groove, the other is not so evident ; on the secondaries it is almost impossible to detect a bare streak. The verrucae when expanded are slightly adpressed to the coenenchyma ; when retracted they are low and dome-like. They are very small, being about 1 mm. in diameter at the base, and varying from 0-5 to 1 mm. in height. The canal system is typical of the species. The axis is cylindrical and yellowish ; it is composed of concentric laminae, and there is a distinct white core which is more calcareous than the outer laminae. There is a slight trace of grooving on the surface. The anthocodiaie and spicules agree in every detail with those described for the species. Locality. — An damans. Note. — This specimen is described in the table given in the Indian Ocean Littoral Alcyonaria Eeport (Thomson and Simpson) as specimen M. Simpson — A Revision of the Goi-gonellidae. 79 Distrihvtion : — Indian Ocean (Pallas). Seas of the Moluccas (Lamarck). India and China (Gray). Cuba (British Museum Collection of H. Christy). Ott' North- West Cape, West Australia, 3-4 fathoms (Studer). Torres Straits (Studer). Cuba (Ridley). Warrior Eeef, Torres Straits, 12 fathoms (Ridley). Mergui Archipelago, Burma. Andamans (Ind. Mus. Litt. Coll.). Elongata Group. This group is easily differentiated from the others by means of its spicu- lation. The spicules are characteristic and very minute. XX. — Scirpearia elongata (figs. 46-48). Gorgonia elongata Pallas, xxviii., p. 179. Gorgoiiia elongata Esper, vii., t. Iv. Gorgonia eloTigata Lamarck, xxiv., t. ii., p. 220, 2nd cd., p. 499. Gorgonia elongnfa Dana, iii., p. 664. Jmicella elongata Valenciennes, xlv., p. 182. Juncella elongata Valenciennes, xlvi., p. 14. Gorgonia elongata Ellis and Solander, vi., p. 96. Juncella elongata Milne-Edwards and Hainie, xxvi., i., j). 1S7. Juncella elowjata Kolliker, xxiii., p. 140. Ellisella elongata Gray, x., p. 287. Ellisella elongata Gray, xi., p. 481. Ellisella elongata Gray xii., p. 25. Ellisella eoccinea Gray, x., p. 287. Ellisella eoccinea Gray, xi., p. 481. Ellisella eoccinea Gray, xii., p. 26. Nee. Juncella elongata Hick.son, xiii., p. 85. Nee. Juncella elongata Thomson and Henderson, xl., p. HI. This is a very old species, but one whicii has caused more trouble lo systematists than any other in the group, owing to the fact that tlie spicules have hitherto never been investigated. The descriptions, based on a few superficial characters, are so vague tliat on those alone it is possible to identify almost any branching Gorgonella with this species. 80 Proceedings of the Roi/al Irish Academy. It is very doubtful if the long list of synonyms given here were in all eases correctly identified ; but in the absence of the specimens themselves, it is better to retain them until definite information on this point is forth- coming. While examining the Alcyonaria in the Museum of the Eoyal College of Surgeons, London, I came across a beautiful specimen labelled Gorgonm dviHjatti (Eeg. Ko. 184), belonging to the Hunterian Collection, of which the following description occurred in the catalogue: — " It consists of a short, broad stem, from which seven main branches arise ; these, after proceeding about 6-7 inches, give off a branch which proceeds upwards nearly parallel with the main stem, and about equal to it in thickness. The crust is of a vermilion colour ; and the polyp-cells are very numerous and arranged in alternate rows, especially towards the free extremities of the branches, which are all more or less flattened. The axis is of a light yellow colour, and of a small size in comparison with the crust." EaMtat. — West Indies. As this is the oldest authentic specimen bearing the specific name elongate, I have considered it advisable to resuscitate this old species, give it some positive content, and regard this specimen as the type. For this purpose. Dr. Burne has supplied me with a beautiful photograph of the colony and also a sketch drawn with a "camera lucida," on which fig. 46 is based. Preparations of the spicules have "also been made for the first time, and fig. 48 gives the chief types which occur. In the collection of Gorgonellids in the British Museum, there is a very delicately branched colony which Gray referred to the species Ellisella coccinea, established by him in 1857, with the following diagnosis : — " Coral furcately branched ; branches sub-cylindrical, \evj long, virgate ; bright scarlet." The spicules of this specimen are identical both in types and measurements with those of the specimen in the Hunterian Collection ^figs. 47 and 48), and an examination of the general habit of the two colo- nies will at once render it obvious that they cannot be regarded as distinct. Both the specimens are from the " West Indies." The type specimen (fig, 46j is almost 1 metre in height, and is complete. There is a large spreading basis of attachment from which a very thick stem about 12 mm. in breadth arises. The branching commences almost at the very base. One of the primary branches is 41 mm. in diameter ; but the secondary branches, at a considerable distance from this, have a bi-eadth of 4 mm. ; about the midiUe of the colony the smaller elongated branches are 3'o mm. in diameter, and at 8 cm. from the tip they are 2 mm. in diameter. There is considerable anastomosis in the lower part. Simpson — A Revision of the Gorgonellidtc. 81 The branching is distinctly dichotomous, and the branches enclose an acute angle ; this is also very marked Ellisella coccinca. The coenenchyma is very thin, and in the dried state extremely brittle ; it is densely spiculose. The canal system is not easily recognized, owing to the fact that V>oth the specimens are very old, and have been presei'ved in a dry condition; but it is still possible to detect two large longitudinal canals. Their position is, however, very marked externally. The polyps are disposed throughout tlie whole colony in two very definite longitudinal series, separated by very wide and distinct bare tracts, which, in the lower region, are depressed and furrow-like. In the older branches there are four to six rows of polyps in each series ; these are situated in what appears to be diagonal arrangement. In the younger branches and twigs the number diminishes to two, and eventually to a single row situated laterally and irregularly alternating. The verrucae are slightly elevated, with the oral aperture dircelcd upwards, but they are very much shrivelled, owing to desiccation. The axis is typically Gorgonellid in structure, and is very hard, especially in the lower portions. The fact, however, that the specimens are dry rendeis the axis harder and more brittle. The spicules of this species are extremely characteristic and very minute. They consist of (1) small double-clubs with closely set, almost smooth warts; (2) double-clubs with more irregular heads; (3) small, slender, elongated double-clubs; and (4, spindles. (Sec figs. 48 and 4Sa.) The following are some of the measurements, length by breadth, in mm.:— (1) 0068 X 0042; 0-065 x 0-042; 0-053 x 0-038. (2) 0-061 X 0-03 ; 0-057 x 0-025 ; 0-057 x 0-03. (3) 0-061 X 0-023 ; 0061 x 0-019 ; 0-057 x 0-015. (4) 0-06 X 0-023 ; 0-058 x 0'015. Locality — West Indies. Flagellu.m-Group. This is a very distinct group, and is characterized chiefly by the nature of the spicules. These are remarkable for the great length of the constriction, the open disposition of the warts, and tlie almost smooth nature of llie latter. M 82 Proceedings of the Royal Irish Academy. XXI. — Scii'pearia flagellum emend. Figs. 49-60. 1863. Juncella fiagellum Johnson, xviii., p. 505. 1864. „ „ „ xix., p. 142. 1870. VimineUa „ Gray, xii., p. 29. 1881. Scirpearia „ Studer, xxxvi., p. 558. 1891. „ ochracea Studer, xxxvi., p. 559. 1901. „ fiagellum Studer, xxxviii., p. 53, PI. ix., figs. 1-3; PI. xi., figs. 10 and 11. 1901. „ ochracea Studer, xxxviii., p. 53, PI. ix., figs. 4-6. 1909. „ fiagellum Thomson and Kussell, xliii., p.. 163, PI. 8, fig. 2. This is a very old species, aud was originally referred to the genus Juncella. In 1 870 Gray assigned it to his new genus Viminella ; but with no apparent reason, and without giving any further specific content. He, however, gives as a synonym, J. caians Verrill ; but as this was based on purely external characters, it is extremely doubtful whether much stress can be laid on the identity with the latter species. We have for this reason excluded it from the list of synonyms. In 1901 Studer rehabilitated the species, and gave a description of the spicules and, also, very good figures of the colonies. He referred the species to the genus Scirpearia. He, however, established another species — namely, ochracea, which caianot now be regarded as distinct from that under consideration, and which we therefore give as a synonym. With regard to Juncella flagelltnn, Johnson, in establishing the species, says : — " I have ventured to assign this coral to the genus Juncella. Valenciennes, a naturalist for whom I entertain the highest respect, considers it to be the Scirpearia mirahilis of Cuvier. There is, however, so much doubt as to what the coral so named by the illustrious Frenchman really is, that I hesitate to ascribe mine to that species — the more especially as it clearly falls within the definition of the genus Juncella' (as it appears in the " Histoire Xaturelle des Coraillaires " of Milne-Edwards, vol. 1., p. 186), forming a member of the section of Gorgonellaceae, which is made up of Gorgouiad corals, having a smooth bark and a sub-lithoid axis, containing so much carbonate of lime as to effervesce in muriatic acid. From J micella jwncea Esper and J. vimen Val. (species found at the Island of Bourbon) it ' The iutroduction of the study of spicules has, however, removed it from the genus Juncella, from the fact that it contains no club-shaped spicules. Simpson — A Revision of the Gorgonellidae. 83 would seem to be distinguished by the large size of tlie cup-bearing papillae ; from J. elongafa, a Mediterranean species, by its being simple, not branched." The original description of the species is as follows : — " Simple, elongated, slender, flexible, slightly twisted on its own axis, and tapering upwards. Bark calcareous, white, smooth, and impuncturate, enveloping a hard, grey axis, which has a somewhat polished surface, marked with straight striae. The axis is highly charged with carbonate of lime. The coral is, quadrangular in section, and has on each of the two narrower sides two series of closely set papillae, having the eight-lobed orilices of polyp cells at their apices. These papillae are obpyriform or ovate ; and in dried specimens they are turned upwards and adpressed to the stem. Near the base of laige specimens the papillae are in three somewhat irregular rows- The other two sides of the stem are free from papillae; but there is a slightly elevated line along the middle. The base spreads out to a moderate extent upon the object to whicli it is attached. The spicula of which the liark is composed are tul^erculated staves, two or tliree times as long as broad, the tubercles having a tendency to collect at the extremities. " The longest example of this coral wlueh 1 have seen measured about 7 feet in length; and it was without its basal portion. The greatest thickness was three-eighths of an inch ; the largest papillae were the tentli of an inch in length, and about tlie same across. In another example, 5 feet in leugtli, the base spread out to the size of a shilling ; and the papillae commenced about 3 inches above this basal expansion. The smallest specimen that has occurred was 31 inches long; this is in the British Mu.seum. In the collection of that establishnaent there is a large stone, with numerous specimens of this coral' upoii it, alongside examples of Caligorgia verticillai-is Gray {Primnoa verticillaris Milne-Edwards). These were brouglit from St. Michael's, one of the Azores, and presented to the Museum by Mr. McAndrew." Studer (xxxviii.) adds the following note with regard to the " Monaco " specimens : — The colonies are long and flexible, and attain a length of 650 mm. The polyps are club-shaped, slightly inturiied towards the axis; they are arranged on two sides of the stem ; in the lower part in several rows ; but towards the tip in a single row, alternating on the two sides. The spicules are spindles and double-clubs. Their dimensions are 0'0C7 ^ 0-015 mm.; (10(31 X 0-0154; 0-056 x 0015; 0-067 x 0-025. The colour varies from whitish yellow to red. • It is, of course, doublfiil whether thise me rcdUy 7. c/oHjo/d, os it wouUl bo imposfiWe todtrific their specific or even generic position by a supcrticiul exumination. M2 84: Proceedings of the Royal Irish Academy. Locality. — To the east of Graciosa, Azores, 454 metres. To the east of Pico, Azores, 318 metres. With reference to S. ochracea Studer (xxxviii.) makes the following observations : — This species is more delicate than S. flagellum. The axis is calcareous; white ; rigid near the base, flexible near the tip. The polyps occur on two sides of the stem ; towards' the base in two irregular rows, but merging into only one row on each side. They have the shape of cylindrical warts or truncated cones, and stand almost perpendicularly to the coenenchyma. They are 2 mm. in height, and about 2 mm. in diameter at the base. The spicules are very like those of S. flagellum ; they consist of double-clubs, with large warts and spines at the two ends : sometimes of a yellowish ochre, sometimes of a white colour. They are slightly larger than those of S. flagellum. The colour of the colony is yellowish brown to orange. Taking into consideration what has already been seen with regard to variation in the group, we see no reason for separating this off as a distinct species. Locality. — To the east of Pico, Azores, 318 metres. We have examined a beautiful, whip-like colony, 37 cm. in lengtli, from Naples,' which we refer to this species. The diameter near the base is 2 mm. ; but near the tip it is only 1 mm. It gradually tapers upwards, but the terminal 25 cm. are almost uniform in thickness throughout. The coenenchyma is very thin and finely granular ; the surface is marked by longitudinal ridges and furrows, which are the outward expression of the internal canals ; two of these are much deeper than the others. The general colour of the colony is reddish orange, but the tips of the verrucae are distinctly more reddish. The lower 2'5 cm. of the stem are devoid of polyps ; this is followed by two opposite longitudinal bare tracts which persist to the tip of the colony. On the other two sides the polyps are disposed in a single row in each series. This gives the colony a very markedly bilateral appearance. The verrucae are cylindrical, tall, and narrow. They average 2 mm. in height and 1 mm. in diameter. They stand sometimes in opposite pairs ; but the more common arrangement is alternate. The polyps on the same side are separated by distances of about 3'5 mm The verrucae are longitudinally striated; and the summit has a very definite eight-rayed structure. They stand almost ' This specimen was given to me for identification by Professor J. Arthur Thomson, who suggested that it might be incorporated in this memoir. Simpson — A Revision of the Gorgonellidae. 85 perpendicularly iu iikaiiy cases ; Imt more frequently they are slightly turned towards the stem (fig. 49). A very noteworthy featuie in tliis connexion is to be observed. The colony has been broken and preserved in two portions. In the longer upper part the polyps are nearly all directed towards the tip ; while in the lower part they are almost all turned downwards. Taking these points into consideration, it may be inferred that the verrucae have power of rotation through 180° both longitudinally and vertically, or, in other words, the anthocodiae may take up any position on the surface of a liemisphere whose radius is the length of a polyp. The canal system is well developed, but there are very few canals, owing to the small number of polyps which occur on the colony. The two main canals are very large ; and their position is indicated on the surface of the coenenchyma by two very delicate depressions on the axis ; also by two grooves larger tlian the others. The axis is hard and densely calcareous ; it is yellow in colour ; and the surface is marked by longitudinal ridges and furrows. Attached to the colony is a young bivalve (probably I'leria macroptera). The spicules of tliis specimen consist of the following types (fig. 50) : — (1) Double-clubs with a long, narrow constriction, and with almost hemispherical ends. Tbe warts are irregularly disposed, are few in number, and are almost smooth: 0-07 X 0-03; 0-0G5 X 0-03; 0-065 X 0 023. (2) More elongated double-clubs witli the same characteristics, and with blunt ends. A noteworthy feature about this specimen is tlie fact that there are very few double-spindles or types with conical ends. Locality. — Naples. We have also referred to this species a specimen in the Cape Collection. The spiculation is typical ; and tlie only dilferenco is the very close disposition of the verrucae. We liave shown, however, that this is a character iu which the species shows great variability. It is a very characteristic colony, growing on a piece of branching coral (like Lophohelia) (fig. 51). It is 9 cm. iu length, and bears one branch (which has been broken) at a distance of 2 cm. from the base. The coenenciiyma is thin and coarsely granular. The general colour of the colony is creamy-white. The lower 2-5 cm. of the m^un stem and also the part of tlie branch which is present (1-5 cm.) are devoid of polyps. On the remainder of the main stem the verrucae are disposed on two sides, and alternate almost regularly. They have the form of flattenetl domes, and give the sides of the colony a 86 Proceed in [I s of the Royal Irish Academy. very undulating appearance (fig. 52). Their bases meet in the middle line. The tips of the verrueae have a very definite eight-rayed structure. Fig. 53 was made from a longitudinal section through the colony to show the attachment of the strong retractor muscles of the anthocodiae. The canal-system is typical, but the inner portion of the coeuenchyma, that is to say, the portion between the two longitudinal series of canals, is very minute. The axis is pale yellow in colour, and very flexible ; the surface is marked by indistinct longitudinal striae. The spicules (fig. 54) of this specimen are typical of the species, but are on the whole larger and broader. Locality. — Buffalo Eiver, East London, N., 15 miles, 310 fathoms. Bottom, coral and mud. In tlie Cape Collection there are also a number of small young colonies, which are extremely interesting, and which are undoubtedly young forms of this species. The longest of these is 7'5 mm. and the smallest 3'5 cm. in length. They have all the same general appearance, and maintain the relative proportions throughout, so that a short description of one colony will give the essential characters (fig. 55). All are attached to pieces of rock, coral, or shell. The stem is about 1 mm. in diameter near the base, and only very slightly less at the tip. The coenenchyma is very thin, and finely granular ; the general colour of the colonies is a bright orange-yellow. The polyps are disposed in two longitudinal series ; and although the two median bare tracts are not well pronounced, the colony has a markedly bilateral appearance. They occur in a single row in each series ; but the interposition of young forms sometimes masks this distribution. They stand sub-opposite or sometimes alternately ; but the young polyps tend to break this otherwise regular structure (fig. 56). The verrueae are elongated and cylindrical ; they are turned towards the stem, and are directed upwards ; their surface is marked by longitudinal ridges and depressions ; the apex when partially closed has a distinct eight- rayed structure ; in many cases the infolded tentacles may be seen projecting around the oval opening. The canal system is well developed ; the canals are distinct but few in number ; the two main canals are large. The axis is cyUndrieal, hard, and very calcareous; the surface is marked by very indistinct longitudinal striae. The spicules (fig. 57) are characterized by the small number and large size of the almost smooth warts and by the very marked constriction in the Simpson — ^4 Revision of the Gorgonellidae. 87 double-clubs. The following are the chief types, with inoasurements, length by breadth, in miUimetres : — (1) Double-clubs with a very long constriction and with almost hemispherical ends. The warts are almost smooth; they are openly disposed and arranged almost in whorls : 0-068 x 0-0.^4 ; 0-061 X 0-03 ; 0-057 X 0-027. (21 Elongated double-clubs passing to double-spindles. There is a very distinct constriction ; and the ends are markedly conical. The warts are not closely set, and are almost smooth : 0-114 x 0-02.3; 0-103 X 0-023 ; 0-095 x 0027 ; 0-095 x 0-023. Irregular forms, crosses, and scales from the tentacles also occur. Locality.— 0''Neil Peak, KW., j W. 9| miles; 90 fatlioms. Bottom, broken shell. To show the varied appearance of the venucae, we have included here three figures of specimens of Scirpearia fla(j<'U ii,m in the Monaco Museum. ■ (See figs. 58, 59, and 60.) XXII. Scirpearia thomsoni, n. sp., figs. 61-63. Juncella elongata Thomson and Henderson, xL, p. 81, PI. i., fig. 10 ; PI. ix., fig- 17. We have no hesitation in establishing this new species for a .specimen which was originally referred to the species JumcUa elongnta by Thomson and Henderson, who were compelletl to base their diagnosis on the very inadequate description of this species which was available at the time of publication of the Indian Ocean Deep Sea Alcyonaria Peport. At that time the spicules of Scirpearia elongata {Juncella clongafa) were unknown; but an investigation of the spicules of an old specimen in the Museunj of the lioyal College of Sui'geons, and the consequent resuscitation of that old but imperfectly known species has caused the necessity of i-emoving the present specimijn. The colony shows several very characteristic features : for example, (1) the nature of the branching, (2) the marked rigidity of the colony, (3j the nature of the verrucae ; but most of all the distinctive character of the spicules, which mark it off as a very definite and new .species. The specimen is 22 cm. iu height, and is branched appro.ximately in one plane. The branching is alnmst dichdtomous; and tlie silhouette of the axis (tig. 61,1 gives the essential features. On the wiiole, the colony is very rigid, owing to the very densely calcareous nature of the axis. The coenen- chyma is moderately tliin, \mi densely spiculose ; the general colour of the colony is salmon-pink. 88 Proceedings of the Royal Irish Academy. " The axis is calcareous, rigid, and brittle ; it is slightly oval in section ; but in the younger portions it becomes quite cylindrical, and tapers till it is thread-like. It shows a very white core surrounded by a brownish cortex." The polyps are disposed in two longitudinal series on opposite faces, each of which consists of from two to four irregularly alternating rows. The verrucae are low and truncate ; when retracted there is a deep depression in the centre which is directed slightly upwards. This gives a very characteristic appearance (fig. 62). They are about 0"4 mm. in height and 1'5 mm. in diameter at the base. The spicules are extremely characteristic, and quite unlike those of any other species (fig. 6.3). They consist of the following types, with measure- ments, length by breadth, in mm. (a) Double-clubs with almost hemispherical heads, and with a relatively long constriction. On either side of the constriction the large warts are arranged in a whorl, while beyond this there is a very warty hub which gives the whole head a very irregular outline : 0 08 X 0 04; 0-07 x 0-035. (b) Smaller double-clubs in which the whorl is not so pronounced : 0-07 X 0-46; 0-07 x 0-042. (c) A peculiar type, which approximate to capstans with terminal warty projections : 0-08 x 0-04 ; 0-07 x 0-021. (d) Elongated double-clubs with a long, narrow constriction, with the inner warts arranged appro.ximately in a whorl, and with more or less elongated and irregularly warted hubs: 0-114 x 0-053; 0-114 X 0046; 0-095 x 0-05. (e) Double-spindles (some of these approach spindles). The ends are almost conical, and are variou.sly covered with very irregular warts which gi\-e the whole a very ragged outline : 0-125 x 0-038, 0-11 X 0-03; 0-1 X 0-027. Locality. — Bay of Bengal, 88 fathoms. , XXIII. Scirpearia alba (Thomson and Henderson), iigs. 64 and 65. Scirpearella alba Thomson and Henderson, xl., p. 82, PI. ix., fig. 15. This species was established for tliree long, incomplete specimens, of a white colour, 28, 411, and 408 mm. in length, with a corresponding diameter at the lower end of 1-75, 2-3, and 1-75 mm. Two of the colonies are unbranched ; but the largest branches at a distance of 251 mm. from the lower end. The axis is cylindrical, hard, brittle, and very 'calcareous, but becomes Simpson — A Revision of the Gorgonellidae. 89 very flexible and filiform near the tip. It is lufirked l>y a iiuiiiber of grooves which iiiii up for a sliort distance, and also l)y a nvnidjor of small protuberances. The stem is oval in section, with a groove on the two flattened surfaces faintly marked in two of the specimens. The verrucae occur in a single roiu on each side of the stem, those of one row alternating with those of the other. They are low and truncated (0'45 mm. in height), laterally compressed, with spreading basis (fig. 64). The diameter is r4 mm. at the base and 0'65 mm. at the apex. The coenenchyma is moderately thick. The spicules of this species (fig. 65) are extremely cliaracteristic. They consist essentially of double-clubs, which are almost as l)road as long, and have a very short but extremely thick median constriction. Their ends are almost hemispherical, and are covered with abundant rugose warts. There are also a few elongated narrow double-clubs, witli more openly-warted heads, and with a longer constrictiut close examination reveals two indistinct, sinuous longitudinal bare tracts. The verrucae are low domes, and scarcely project beyond the coenenchynia (fig. 70). Owing to the extreme thinness of the coonenchyma, the canal system is very ill-defined. The axis is very calcareous, hard, and, in the younger parts, brittle. It is composed of very thick concentric lainiiiac : the surface is faintly and irregularly marked by grooves. Locality. — Ceylon Seas. As was the case with the two species already discussed, the two now described are identical on superficial e.\aininatioii. Let us now proceed to investigate in what respects tiie two groups dilfer. /«)•(•«<« -group. indim-gTow^. Tlic verrucae are separated into The verrucae are separated intn two longitudinal series by two very two longitudinal series by indistinct distinct bare tracts. bare tracts which m.ay even disapjiear near the base. There are two or three longitudinal pije verrucae appear as if distri- rows in each .series. buted in five to seven slightly spiral rows. like. The verrucae are low and dome- ^,^^ ^^^^^^^^ ^^^^^, ^^ ^^^ ^^^^ ^,^j pointed, (2) projecting lodges, (3) low and dome-like, (4), almost level with the coenenchynia. 96 Proceedings of the Poyal Irish Academy. Thus we see that, although superficially they may present very different appearances, when we investigate the various characters nothing of specific moment can be found to obtain. The question of " five to seven slightly spiral rows " resolves itself into two series of two to four rows in which the bare tracts are hardly distinguishable. JuncelM elongate! (Val.) Hickson, xv., p. 821, Hiekson referred some fragments to this species, but expressed doubt as to the identification. He gave the following notes : — One specimen (in three pieces) was 315 mm. in length. The total diameter was 3-5 mm., and the axis 2 mm. in the middle region. Xearer the base the coenenchyma is relatively thin or very thin, and nearer the apex much thicker. The colour of the coenenchyma is pale pink and the verrucae are throughout shallow domes, white hi colour. The verrucae are separated by distinct bare tracts into two longitudinal series. In the portion I examined there were six to seven rows in each series. Other specimens were pale red and orange-red in colour. In the latter, which was 230 mm. in length, the verrucae were scattered and prominent towards the distal end, but there is an almost smooth coenenchyma near the base (fig. 71). The spicules are double-clubs, warted spindles, and a few more elongated spindles, with fewer tubercles arranged in regular rows. The warted spindles and double-clubs vary in length from 0'08 to 0'08.5 mm. Some of the pointed spindles are 0"1 mm. in length. There is evidently a good deal of variation in the shape of the spindles (fig. 72), The colour, the prominence of the verrucae, and the definiteness of pronounced tracts free from verrucae, are also characters in which the species shows much variation. Locality. — S. Nilandu, 2.5 to 30 fathoms (Maldives) Hulule, Male Atoll, 25 to 30 fathoms (Maldives). JYote. — In one specimen Hickson says clubs similar to those in J. juncea occur ; but this probably belonged to that species. In the Littoral Alcyonaria Collection of the Indian Museum there is a portion, 35 cm. in length, of what has evidently been a long flagelliform colony ; both the basal and terminal parts are wanting. The coenenchyma is granular and moderately thick. The diameter is almost constant throughout the part under examination ; it is about 4 mm., while that of the axis is 2 mm. The general colour of the colony is orange- red, but the anthocodiae are white. The polyps are disposed in two longitudinal series separated by two narrow Simpson — A Revision of the Gorgonellidae. 97 bare strips which become more indistinct, but still visible, towards the tip (fig. 73). These are spirally twisted ; but this is, no doubt, due to a general torsion of the colony. In each series the polyps appear in rows diverging from the bare tracts ; this gi\'es a very marked spiral arrangement, but this is also due to torsion. Transversely four or five is a common number in each series. The verrucae are sub-cylindrical and closely adpressed to the stem; the outer insertion is lower than the inner. They arc about 1-5 mm. in height and 0-75 mm. in diameter at the base. When retracted they are sub- conical, and have eigl),t converging lips (fig. 74). The canal system is very definite and typical ; the two main longitudinal canals are extremely large. The axis is yellow in colour, and markedly calcareous; the surface is apparently smooth ; it tapers only slightly in the portion preserved. Locality. — Off Table Island, Cocos Group, Andamans, 15-35 fathoms. When we take into consideration the fact that the great majority of these forms are fragmentary, and also the slight basis on which the genera Juncella, Scirpearia, and Scirpearella were formerly dillerentiated, there is small cause for wonder that the various specimens were referred to one or other of these genera on account of differences which we hope to show are not specific, but only different manifestations assumed by extremely plastic organisms. We have made a very exhaustive study of the spicules in all the forms of which descriptions have been given ; and although these show certain deviations, nevertheless they may be grouped into a number of more or less definite types. Fig. 75 gives a very good representation of the difterent tyjies and deviations therefrom in the case of the spicules in the Indian Museum specimen. Fig. 72 of the spicules of Hickson's Juncc/la clongata has also been added, and a comparison of these two groups should at once indicate the affinities of the.se two apparently different forms. A similar comparison might be made with regard to the others with a like result. If, then, the character of spiculation can be regarded as specific, we should be compelled to unite all these extremely divergent forms into one very variable .species. This procedure may, at first sight, seem rather drastic, as, it may be argued, the dillerent variations occurred not in each specimen but in different specimens. They distinctly show a range of variation which cannot be easily com- prehended within an individual colony. We are, however, fortunately in the possession of a large colony wliieli has the same characteristic spiculation, and which does actually show a range 0 98 Proceedings of the Rojjal Irish Academij. of variation as great as, if not greater than, that recorded for the individual portions hitherto described. We therefore propose to give a faii-ly exhaustive account of this colony, and regard it as the type of the species in its emended form. A beautiful colony of an orange-red colour 52 cm. in height and about 16 cm. in breadth. It is largely branched, approximately in one plane, and in a manner similar to that in Juncella gemmacea. It is complete to the very base ; but some of the branches have been broken off. These are nearly all preserved, however ; and it is possible to piece them together so as to get an idea of the nature of the colony as a whole when living (fig. 76). The main stem has a diameter of 4'5 mm. at the base and 3'.5 mm. at a height of 100 mm. where it has been broken off. Tiie first branch arises at a distance of 18 mm. from the base ; it is 3 mm. in diameter at its point of origin, and tapers gradually to a point ; it is 14.5 mm. in length. The second arises after another 33 mm., and attains a length of 445 mm. ; it is 4 mm. in diameter at its origin, and gives rise to a secondary branch 375 mm. long after a distance of 82 mm. ; the diameter of the latter is 3 mm. at its point of origin. A third primary branch comes off at a distance of 95 mm. from the base, and is 3 mm. iu diameter near its origin ; it is 430 mm. in length, and tapers gradually to a conical point. The coenenchyma is thin and finely granular; it is of a pale yellow colour, but the verrucae are red. Near the base long streaks of red extend longitudinally from the verrucae and interlock, giving a peculiar tessellated pattern (cf. the type specimen of Scirpearella sp. B.). This feature may be seen in other parts of the colony. The polyps are disposed on the branches iu two longitudinal series, separated by two distinct bare tracts, which may be more irregular or even altogether absent. Near the base the verrucae are only slightly elevated, and in many cases hardly project beyond the coenenchyma (fig. 77). Near the origin of the second primary branch there are 3-5 longitudinal rows in each series ; the verrucae are low and dome-like, or in some cases like bluntly truncate cones, having an eight-rayed structure at the summit (fig. 78). About midway on the third primary branch there are 4-5 longitudinal rows in each series ; tlie verrucae are sub-cylindrical and closely adpressed to the stem (fig. 79). Towards the tips of the branches the number of rows of polyps in each series diminishes to two and eventually to one ; the verrucae are sub-cylin- drical or in some cases dome-like (fig. 80). Simpson — A Revision of the Gorgonellidae. 99 Thus we see that this specimen exhibits all the variation phases which are represented in the various specimens previously discussed. The canal system is typical ; the two large main canals corresponding to the two bare tracts are very pronounced. The axis is cylindrical, calcareous, and made up of concentric laminae. It tapers gradually from the base upwards, and is fairly flexible. The coenenchyma is thus of an almost uniform thickness thi-oughout. The surface of the axis is marked by longitudinal striae, the number of which varies in the different parts of the colony. The following are the chief types of spicules (fig. 81), with their measurements, length by breadth, in mm : — {a) Small double-clubs with a narrow constriction, and with openly warted ends: 007U X 0-038; 0-068 X 0046; 0-068 x 0-034. (b) Smaller double-clubs with comparatively few warts on the ends : 0-OGo X 0-034; 0-061 x 003; 0-057 X 0-038. (c) Smaller double-clubs with the ends more den.sely covered with smaller warts: 0046 X 0-023 ; 0-042 x 0-019; 0-038 x 0015. {d) Elongated double-clubs with openly warted ends: O-QS x 0-023; 0-068 X 0-031. (c) Elongated double-clubs with closely warted ends : 0-072 x 0-03 ; 0-068 X 0-027 ; 0-065 x 0-023. (/) Narrower double-clubs, simulating spindles: 0*076 x 0019; 0072 X 0-023 ; 0-065 x 0-019. Locality. — Providence Island, 29 fathoms. In the Cape Collection there is a large numliei of (•Dlcniics which are extremely diverse in external appearance, but all of whicii have essentially the same spiculation. It is absolutely impossible to differentiate these from »S'. furcata, so that I have decided to include them in this species and give a few notes on each specimen, with special reference to the variations. In addition to the more mature colonies, there are a few undoublt-dly young forms, the largest of which is only 50 cm., and the smallest 8-5 cm. in length. All are of a creamy-white colour, and form a striking anncctcnt series, showing the various " types " of verrucac which are undoubtedly only different stages in retraction (fig. 82). Locality. — Hood I'omt, N., 5J miles, 42 fathoms. IJotlom : sand and shells. We shall commence with those forms in which the \orrucae are very small, and gradually pass to those in which tiiey are more expanded, and show that a series exists connecting the most extreme types. O 2 100 Proceedings of the Royal Irish Academy. A beautiful, simple colony of a pale orange colour. It is 17 cm. in length. The base is broken off and the tip is dome-like. The coenenchyma is moderately thick, and is finely granular. The polyps are distributed over the whole of the coenenchyma ; in some parts they appear as if in spirals, but they are in reality in longitudinal rows, the members of which irregularly alternate. Four of these rows may be seen from one aspect. The verrucae are extremely small, and ai-e sunk into pits in the coenenchyma, so as to be almost level with it (fig. 83). The members of one longitudinal series are separated by distances about three to four times the length of the verrucae. The verrucae themselves are somewhat cylindrical, and have a distinctly eight-rayed summit. There is not the slightest trace of a bare tract. The canal system is, however, typical. The two large main canals are quite prominent in a cross-section. This reminds one of the type of Jimcella juTicea with non-projecting verrucae. The axis is lamellar, densely calcareous, and very hard ; the surface is indefinitely marked by longitudinal atriations. Locality. — Off and east of Cape Morgan, 36 fathoms. Bottom : stones. A long, simple, flagelliform colony, 50 cm. in length. The diameter near the base is 5 mm. ; near the tip it is 4 mm. The coenenchyma is thick, being slightly over 1 mm. throughout. The general colour of the colony is a dull orange-red. The polyps are distributed in two longitudinal series separated by two narrow bare tracts ; there are four to seven alternating rows in each series. The verrucae are small and are closely adpressed to the coenenchyma, being sunk in pits so as to be almost level with it. They are very much retracted, however, and there is every reason to believe that when expanded they would be more than double their present length (fig. 84). The members of one row irregularly alternate with those of the adjacent row. The canal system is well developed, and is clearly seen in the thick coenenchyma ; the two main canals are very large. The axis is flexible, but very hard and densely calcareous ; it is about 2"5 mm. in diameter near the base. The surface is marked by distinct longitudinal striae. Loculity. — Umhlangakulu River mouth, N.-W. by N., 7^ miles ; 50 fathoms. Bottom : sand, shell, and sponge fragments. An almost complete colony, 24 cm. in length, of which only the base is wanting. This specimen is extremely interesting, as it shows to what extent the polyps may be extruded in this species. Simpson — A Revision of the Gorgonellidae. 101 The diameter of the stem is 4 mm. near the base, but diminishes gratlually to 3 mm. near the tip. The coenenchyraa is thick ; the general colour of the colony is pale yellow ; but the tips of the verrucae and tlie anthocodiae are white. The present specimen agrees in detail with llie last, except in the nature of the verrucae (cf. figs. 84 and 85). Locality. — Off and oast of Cape Morgan, .36 fms. Bottom : stones. A beautiful, complete, simple colony, 22-5 cm. in length ; the coenenchyma is moderately thick and densely granular ; the general colour is a bright orange-yellow ; but the tips of the verrucae and the anthocodiae are wliite, and there are also white streaks throughout the coenenchyma. The polyps are distributed in two lateral, longitudinal series ; the bare median tracts are faiily well delined. The number of rows in each series varies from two to four. The verrucae are sub-cylindrical, directed upwards, and adpressed to the stem (fig. 86). Ttie member.s of two adjacent rows alternate with one another so that the tip of one verruca is on a level with Llie base of the next liigiier in tlie adjacent row. The verrucae are about l-2o mm. in height and 075 mm. in diameter. Near the base they are much smaller, more distant, and a few are even sunk into pits in the coenenchyma. The anthocodiae are white ; the tentacles are short, but have a dense aboral armature. The canal system is typical and well developed ; tlie two main canals are easily seen when a piece of the coenenchyma is detaciied. The axis is slender, flexible, but very calcareous ; tlie surface is marked by longitudinal striae. Luadity.—\] mhloii Kivcr moutli, N. by W. half W., >i\ miles, 4.5 fms. Bottom : sand, shells, and hard ground. A small, complete colony, 15"5 cm. in height ; is almost identical with the last specimen. The following dillerences may be noted : — (1) The colour is almost brick-red. (2) The verrucae are slightly smaller and are more adpressed to the coenenchyma. (iJnlh these differences are probably due to greater retraction and to the fact that the colony iUself is smaller.) Locality. — Umhloti lliver mouth, N. by W. half \V., Sj miles, 40 fms. Bottom: sand, shells, and hard ground. In the Littoral Collection of tlie Indian Museum tiiere are four filiform colonies which have the characteristic spiculalion of H. furailn, U) which species we have therefore assigned them. They differ considerably in 102 Proceedings of the Royal Irish Acddemy. external appearance, so that the following notes and figures (figs. 88 and 90) give some idea of the fertility of variation. Let us commence with those in which the verrucae are most contracted. A long, simple filiform colony, 82 cm. in length, and having a maximum diameter of 2-5 mm. The coenenehyma is finely granular, and only 0'25 mm. in thickness near the base. The colour of the colony is a pale orange-yellow ; but the tips of the polyps are reddish. The verrucae are small and wart-like ; when retracted they are sunk into the coenenehyma, and show an octoradiate structure (figs. 88rt and 886). The polyps are disposed in two longitudinal series, with two or three transverse, irregular rows in each series. No polyps occur on the lower basal part of the colony. The polyp-bearing areas are separated by two bare tracts, in one of which there is a distinct furrow, caused by the collapse of one of the main canals ; the position of the other main canal is clearly visible owing to the extreme thinness of coenenehyma. The verrucae are about 1 mm. in diameter. The canal system is well marked and is quite typical of the group. The canals themselves are all very large. The axis is cylindrical, dark brown at the base, where it is about 1 5 mm. in diameter ; and pale yellow in the upper portion, where it is hair-like in fineness. The surface is marked by indistinct longitudinal furrows and ridges. Fig. 89 shows the chief types of spicules. LocoJity. — Off Malabar Coast, 36 fms. A small, complete, simple colony, 27 em. in length, from the Andamans, also occurs in the Indian Museum Littoral Collection. It is of a creamy- white colour, and is almost imiform in thickness throughout ; it agrees in detail with the last specimen from the Malabar Coast, except that each transverse row has only one polyp or occasionally two polyps. The axis is of a straw colour throughout. The spicules are identical with those described for the other specimens. Locality. — Andamans. Two slender colonies, of a creamy-white colour, with projecting wart-like verrucae. The smaller colony is complete, and is 39 cm. in length ; it is 2"25 mm. in diameter near the base (without verrucae), and about the middle of the colony ; the basal portion which is present is 47 cm. in length. The diameter at the base is 2-75 mm., while at the broken end it is 4 mm. The coenenehyma is granular, and moderately thin ; it is creamy-white in colour. Polyps do not occur for a considerable distance from the base ; thereafter tjhey are separated into two longitudinal series by two sinuous depressions Simpson — ,1 Revision of the Gorqonellidae. 103 (fig. 90&) ; the two serit-s approach so closely togetiier as to appear as if merged into one, covering the whole of the coenenchyina, especially towards the middle of the colony. The verrucae are low, broad, and mound-like; they are sometimes 2'5 mm. in diameter at tlie base. Towards tiie base (fig. 90rt) they are almost level with the coenenchyma ; while near the tip they are often sunk into depressions in tlic coenenchyma (fig. 90f). The oj^ening is circular, sometimes elongated, and has eight li]is amuiid it, giving a very definite pattern. The canal system is typical ; the two main canals are large. The axis is brown at the base, but yellow in tlie younger part ; it is composed of concentric laminae, and is markedly calcareous. Tliere are definite longitudinal striae, especially in the older part. Fig. 91 shows the predominant spicular types. Locality. — Ofi' Malabar Coast, 36 fms. Amongst the Alcyonaria collected by the writer in Ihe Mergui .Arclu- pelago, Burma, there are six specimens which undoubtedly belong to tiiis species. Five of these are long and llagdlifcirm, and represent a series in development ; they are very slender, and taper only slightly from base to apex. The following measurements will serve to illustrate tlie most salient features : — Total Diameter of Diameter Diameter Specimen. length of colony of colony of axis Colour. colony. at biise. midway. nt base. I. 27 cm. l*7o mm. ]•.') mm. 1 mm. Crcamy-wbito, bnt yellowish towardh t>a«e. II. 42 cm. .3 mm. 2'it this niimber decreases both towards the base and the apex, in each of which two or even one is the common number. Young forms occur scattered throughout the other veirucae, and the distribution is then very difficult to determine. The verrucae are very minute and wart-like ; when retracted, they are slightly sunk into the coenenchyma, and present a distinct eight-rayed figure which simulates a pseudo-operculum. In some cases they protrude slightly, and give the surface of the colony a faintly undulating appearance. The axis is composed of concentric laminae, and is markedly calcareous ; the surface varies in colour from black, through brow-n to pale yellow, according to its age. It tapers only very slightly. The canal system is well developed ; even in these slender specimens a cross-section, when viewed with a hand-lens, reveals the two longitu- dinal series. The part of the coenenchyma between these two series is very small compared with the outer non-canal-lsearing part. The two main canals are extremely large in proportion to the others ; and to this is dixe the very obvious longitudinal depressions even in the non-polyp- bearing part. Localitij. — Mergui Archipelago, Burma. Scirpearia furcata var. robusta. Figs. 92-96. We have examined two characteristic colonies, one from the Indian Museum Littoral Collection and one from the Mergui Collection. These exhibit certain differences from the other specimens ot furcata, but for the present we would consider them as a variety of furcata. The colony in the Indian Museum Littoral Collection is complete with its basis of attachment ; it is 20 cm. in height and 3 cm. in breadth, and consists of a main stem from which a branch of 6'5 cm. in length arises at a distance of 6 cm. from the base (fig. 92). A second branch arose 1 cm. from the first ; but this has been broken at the point of origin. The diameter of the main stem near the base is 4 mm. ; near the tip it is o-o mm. The two branches seem to arise in planes perpendicular to one another. The stem and branch are cylindrical. The coenenchyma proper is finely granular and thin, never attaining a Simpson — .4 Revision of the Gorgonellidae. 105 thickness of over 1 mm., biit about OS mm. near the base. Near the tip of the main stem it has been rubbed off. The general colour of the colony is brick-red. On superficial examination the polyps appear to be distributed over the whole of the coenenehyma : but a minute inspection reveals a disposition in two longitudinal series sepai-ated by a sinuous line in the lower portion : this is more marked in the upper half and in the branch where a distinct depression is visible. No polyps occur on the basal 1'5 cm. The verrucae are large and dome-like ; they are about 2 mm. in diameter and r25 mm. in height. There is a trace of an eight-rayed structure at the ■summit (fig. 9.3). They vary very little in the different parts of the colony. Tlie canal system is t}'pical ; the two large main canals are very distinct. The axis is cylindrical, very calcareous, and gives great rigidity to the colony; it is composed of concentric laminae. The colour varies from brown in the lower portion to pale yellow near the tip. The diameter near the base is over 3 mm. ; it does not taper very markedly until it approaches the tiji. The surface is marked by indistinct longitudinal striae. Tlie spicules (fig. 94) consist of double-clubs and elongated double- spindles, which in some cases approached the spindle type. The following are the chief types, with measurements, length by breadth, in millimetres : — (a) Double-clubs with a short constriction and with the warts somewhat regularly disposed : 0-08 x 0-04 ; 0-073 X 0-046 ; 007 X 0042. (h) Smaller double-clubs with the warts nearest to the constriction arranged in a whorl : 0-06 X 0-034; 0-04.5 X 0025. (c) Elongated double-spindles with irregular di.sposed warts: 01 X 0-035; 0-095 X 0-03; 0-09 x 003; 008 x 0-025. {d) Spindles (like type (c), but with no constriction): 0-09 X 0-025; 0-08 X 0-02. Types (c) and {d) are more abundant in the verrucae. Very characteristic is the occurrence of a large number of conically shaped elongated double- clubs and spindles. Locality. — Andamans. Another very characteristic, complete, simple colony, 17 cm. in length, occurs in the Mergui Collection. Externally it recalls the projecting- verrucae type of Jvncdln jv.ncm ; but the nature of the spiculation prechulcs this possibility. The disk of attachment is present. The diameter at the base, without verrucae, is 2 mm.; it increases in thickness very markedly. P 106 Proceedings of the Royal Irish Academy. so that near the middle of the colony it is 5-5 mm. 'nicluding verrucae) ; from this position to the tip it decreases, so that midway it is only 3 5 mm., while the apex itself is distinctly pointed (fig. 95). The coenenchyma is finely granular, and, except near the base and towards the tip, it is very thick. About the middle of the colony, where the diameter of the axis is O'To mm., the coenenchyma is 2 mm. in thickness. The colour of the colony is creamy-white. The verrucae are dome-like; but the oral opening is directed slightly upwards ; they are about 1 mm. in height and 1 mm. in diameter at the base. The colour is markedly flattened throughout its entire length ; on each of the two flattened surfaces there is a very deep groove ; these separate the polyps into two longitudinal series. In each series there is a varying number of polj^s ; near the base there are four transverse rows ; towards the middle of the colony there are five ; while from this point the number diminishes, so that near the apex there is a single row in each series. Young forms occur amongst these, however, and break the fundamental symmetry. The canal system is well marked ; the two main canals corresponding to the two longitudinal grooves are very large ; in this and other respects it is characteristic of the group. The axis is very slender ; at the base it is only slightly over 1 mui. in diameter ; from this it tapers gradually to an almost hair-like fineness at the tip. It is black in colour near the base, but passes through pale brown to yellow near the apex. The spicules (fig. 96) are almost identical with those in the previous specimen, both in types and measurements. Locality. — Mergui Archipelago, Burma. XXVI. Scirpearia andamanensis n. sp. Figs. 97-101. This new species is established for a very distinctive specimen in the Littoral Collection in the Indian Museum. The colony is 17 cm. in height and 9 cm. in maximum breadth ; it is laxly branched in one plane. The branches arise in an irregular and sub- alternate manner, and are considerably elongated. The basis of attachment is broken off at what is e\idently a short distance from the actual base. (The colony is shown complete in fig. 97.) The stem and branches are cylindrical, and taper very slightly. The coenenchyma is about 1 mm. in thickness ; and this is almost constant Simpson — A Revmon of the Gorgonellidae. 107 throughout the colony, being slightly thinner in the older portions. Tlie surface is finely granular. The general colour of the colony is ochreous-yellow ; but the tip.s of the verrucae and the anthocodiae are white. The polyps are distributed in two longitudinal series situated laterally — that is, on the aspects perpendicular to the plane of ramification. In each series there are from three to four irregular rows. The two bare spaces are quite distinct,- and only here and there are median depressions to be seen. The verrucae vary considerably according to the stage of retraction. When expanded they are mammilliform, are directed upwards, and adpressed to thi^ coenenchyma. This is well seen near the tips of some of the smaller liranclies where the coenenchyma is relatively thicker and where tliey are depressed into the coenenchyma (fig. 98). When partially retracted they are wart-like or sometimes like short truncated cones standing perpendicular to the coenenchyma (fig. 99). When still further retracted tliey appear as small rounded projections or may be even sunk beneath the surface of the coenenchyma (fig. 100). In all stages an eight-rayed figure is discernible. Tliey are about 1 mm. in diameter, and may attain a height of over 1 mm. Two large main canals corresponding in position to the Ijare tracts are plainly visilile in a cross-section. The small canals ai'e very numerous owing to the large number of the polyps in a transverse row. The axis is cylindrical and calcareous. It is about 3 mm. in diameter at the base, but gradually tapers to an almost liair-like fineness. It is composed of concentric laminae. The surface is greenish-brown in colour, Init towards the centre it is whiter owing to the greater amount of calcareous matter; there are indistinct longitudinal striae. The spicules (fig. 101) are pale yellow or colourless; they consist of the following types, of which the measurements in millimetres are given : — (a) Large double-clubs, witli almost hemisplierical ends, and a very sliort median constriction : O'OT X 00:55 ; GOT X 0-0;5 ; OOGo X 0-04 ; 0-06 X 0-04. ib) Smaller double-clubs with more openly warted lieads and a longer constriction : 0-045 X 0-03. (c) Elongated double-clubs with comparatively few irregularly distributed warts: 0-08 x 0023; 0-08 x 0-02; 0-OG X 0-015. ((/) Spindles (these may be modifications of type (r'; in whidi I be ccm- striction is not visible) : 0-09 X 0-02. Locality. — Andamans. P2 1 08 Proceedings of the Royal Irish Academy. XXYII. Scirpearia ramosa n. sp. Figs. 102-104. In the Littoral Collecriou of the Indian Museum there occurs a very- beautiful and characteristic branched colony for which it has been necessary to establish a new species. The mode of branching, the nature of the verrucae, and the distinctive character of the spicules, are all features of great importance. The colony is complete with its basis of attachment ; it is 14 cm. in height and about 9o cm. in maximum breadth, and is branched irregularly in one plane. The majority of the branches arise at nearly right angles ; they are long, and may ascend for a considerable distance without giving rise to finer twigs. They vary very little in diameter throughout their entire length (fig. 102). They are flattened in the plane of ramification, so that a cross-section is elliptical. The diameter of the main stem is 2*5 mm., but some of the branches are 3 mm. in their longer and about 2 mm. in their shorter diameter. The coeneuchyma has a very granular surface ; it is 1 mm. in thickness towards the tip of the branch, but considerably less in the older parts where the axis is thicker. The colour of the colony in spirit is yellowish-red ; but the verrucae are of a more decided reddish tint, and streaks of red pass indefinitely from them, and gradually merge into the general tone of the coenenchyma. When dry the whole colony is almost ochreous yellow. The polyps occur on the branches, but not on the main stem ; they are distributed in two distinct series on the sides, or non-flattened aspects, of the branches ; but occasionally they encroach on the flattened surfaces. There are thus two very distinct bare longitudinal zones. The verrucae have the appearance of very low truncated cones, and are almost crater-like ; they hardly project beyond the coeneuchyma. This is due to their great contractility, as is evident from the shrunken appearance. They are about 0-5 mm. in height and 2 mm. in diameter at the summit. The oral opening is very large ; it is circular in outline, and the eight retracted tentacles apparently form a pseudo-operculum (flg. 103). The canal-system is typical of the group ; the two main canals are very large, and correspond to the bare tracts. On several of the branches there is a distinct longitudinal furrow indicating their exact position. The axis is cylindrical, and is composed of definite concentric laminae ; a cross-section shows lines radiating from the centre to the circumference. The outer more horny portion is brown in colour, but the more calcareous central part is white. The surface is marked by longituiliual ridges and furrows, the number of which varies according to the portion of the colony examined. Two Simpson — A Revision of the Gorgonellidae. 109 of the furrows, larger and deeper than tlie others, correspond in position to the two large main canals. The spicules consist essentially of double-clubs; but these may be elongated and narrow, and with so short a constriction as to appear like warty spindles. The warts are large and close-set. The spicules are either pale yellow or colourless. In the coenenchyma there are only double-clubs, with warty, hemispherical heads, and a short constriction. The following measure- ments, in millimetres, are typical : — 0-07 X 0-05 ; 0-07 X O-O-i."). 0-05 X 0-03 ; 0-04 x 002.5. The spicules of the polyps are, on the whole, longer and narrower than those of the coenenchyma. They are (1) Double-clubs, witli warty, slightly elongated heads, and with a slmrt constriction : O'OO X 0-02; 008 X 0-02; 0-06 X 0-025. (2) Tiiicker double-clubs, more like those of tlie coenenchyma : 0-08 X 0-035. (3) Warty spindles (occasionally a constriction is discernible) : 0-07 X 0-02. Lucality. — Andamans, 20 fms. Specific Bingnosis. Colony branched in one plane; most of the branches arise almost perpendicularly, but soon turn upwards ; they are flattened in the plane of ramification, vary very little in thickness throughout their entire lengtli, and terminate bluntly. The polyps are distributed for the most pait on the non-iiattened aspects of the branches, and stand perpendicularly ; the verrucae, when retracted, have the form of low, truncated cones, and may even appear almost level with the coenenchyma. The oral opening is closed by the inturned tentacles, wliicli tluis form a pseudo-operculum. The spicules consist essentially of [I) broad double-chibs, with a slmrt constriction and almost hemispherical ends; (2) elongated, broad double- clubs, with very ruuiuled ends, and witli the same character as tiie previous type; and (3) longer and narrower doulde-clubs, which may appro.ximate double-spindles, and eventually spindles. XXVllI. Scirpearia ceylonensis n. sp. Figs. 105-107. Among tlie Alcyonaria collected by I'rofessor lierdman in Ceylon is a beautiful i)ranched specimen which was not described in the general report. It has been found necessary to establish anew species to include it. 110 Proceedings of the Rojial Irish Academy. The total height of the colony is 31 cm. ; it consists of a main stem 30 cm. in height, from which four branches arise, all on one side. The first arises at a point 4 cm. from the base ; and the others after 2-5, 3, and 10 cm. consecutively. The lowest branch is broken, and is 13 cm. in length, but was evidently much longer ; the others are 5-5, 10, and 15 cm. respectively (fig. 1051. The main stem after the origin of the first branch and all tlie branches are markedly flattened in the plane of ramification. The diameter of the main stem near the base is 2 mm., and its greatest breadth in the flattened portion 3 mm. The branches vary considerably in thickness. The colour of the colony is a pale orange-yellow ; but the verrucae are more reddish. The coenenchyma has a very granular surface ; it is nearly 1 mm. in thickness in the branches, but thinner in the older parts, where the axis is proportionately larger. The polyps occur on the branches and also on the main stem, except on the portion below the origin of the first branch ; they are distributed in two longitudinal series on tlie sides of the branches ; the flattened aspect is broad, and quite devoid of polyps. In each series this is a single row ; but over- crowding or the interposition of young forms sometimes obliterates the symmetry (fig. 106). The verrucae, when retracted, are low, truncated cones, and often show very distinct wrinkling ; they project very little beyond the coenenchyma. Many of the anthocodiae are only partially withdrawn ; and the infolded tentacles appear to form a cone ; on further retraction their bases foi'm a horizontal pseudo-operculuni, and the verrucae present a very shrunken appearance. The tentacles are eventually quite covered up by the inturued sides of the verrucae. The canal system is typical and well defined ; the two main canals, corresponding in position to tlie bare tracts, are large ; and a depression is sometimes visible owing to a collapse of the walls. The axis is thin, cylindrical, composed of concentric laminae, and markedly calcareous. It is yellow in colour; and the surface is striated, two grooves slightly larger than the others being seen in some places. The spicules (fig. 10*7) consist of the following types, with measurements, length by breadth, in millimetres. («) Double-clubs, with a short constriction and with almost hemispherical heads, very irregular in outline, covered with few large warts: 0-08 X 0-042; 0-076 x 0-046; 0076 x 042. (i) Elongated double-clubs, with rounded ends, and openly-warted : 0-08 X 0-038 ; 008 x 0-034. Simpson — A Revision of the Gorgonellidm. 1 1 ] (c) More elongated double-clubs, merging into double-spindles. The warts on these are sometimes disposed in wliorls : 0084 x 0-027 ; 0082 X 003; 0082 x 0-026. From these measurements it will be seen that there is very little difference in the lengths of the various types, but that the breadths diminish proportionately more than the lengths. Intermediate forms also occur. Locality. — Off Galle, Ceylon. XXIX. — Scirpearia maculata. Figs. 108 and 109. EUisdla. maculata Studer, xxxiv., p. 629, Taf. iv., fig. 27 {a, h, and c). Elliaella maculata (pars) Wright and Studer, 1., p. 160, PI. xxxiv., fig. 9. MliseUa calamus Studer, xxxiv., p. 660, Taf. v., fig. 28 (a, b, c, d, and c). Ellisella calamus Eidley, xxxiii., p. 348. It is with considerable hesitation that we still recognize this species as distinct. It has been impossible, however, to examine the type specimen of the species ; but we have seen a Banda specimen in the Britisli Museum, of which Professor Bell has sent me a photograph (fig. 108). The other specimen, from the Torres Straits, described in the " Challenger" Report, has proved, on examination of the spicules, to be Juncclla gcmmacea. There can be no doubt, however, that Ellisella calamus is the same as Ellisella maculata, since in spiculation they are identical, and the macroscopic characters on which they are separated are only \ariational differences. This will be evident from the following description. Studer, in describing E. maculata says : — The stem is cylindrical, forked, divided into only a few long cylindrical branches. The colony is 5 em. in height ; the diameter of the stem is 5 mm., that of a branch 3 mm. One of the branches is 13 cm. in length. The stem and branches are co\'ered with verrucae, which hardly project; these occur laterally, on the thicker branches, in sevei'al rows, leaving a narrow, shallow median space, which disappears in the twigs. The verrucae have a circular opening. The spicules are (1) double-clubs, 0-095 mm. in length ; and (2) a few warty spindles, 0'084 mm. long. The colour of the coenenehyma is orange-red ; the verrucae are dark red. Locality. — Mermaid Straits, North-West Australia, 50 fms. In separating A'. c«te?i«s fromii'. maculata he gives the following diagnosis of the former : — Simple, rod-like, cylindrical stem. The length of the largest specimen is 80 cm. The maximum diameter is 2 mm. The axis is horny and 112 Proceedings of the, Roi/al Irish Academ;/. calcareous, with alternate horny and limy rings, flexible, yellowish. The cortex is fairly thick. The verrucae project as pointed cones only in the upper portion. They occur on the sides of the stem in quincunx, in several rows, leaving a narrow, shallow, smooth space, which gradually becomes narrower till it disappears in the terminal portion. The spicules are like those of maculata, namely, spiny double-clubs and spindles (0-06). Locality. — Mermaid Straits, 50 fms. Eidley (xxxiii., p. 348), in identifying a specimen in the "Alert" Collection with E. calamus, gives the following notes : — A specimen 9 inches (225 cm.) long; incomplete. The colour is dark brick-red. The fusiform spicules were almost twice as long as those of Studer's specimen. He says nothing of the dimensions of the double-clubs. Locality.— Vort Denison, Queensland, 4 fms. The following notes from the " Challenger " specimen in the British Museum (fig. 108) may be of interest : — The fragment is 50 mm. in length, and has a diameter varying from 3'5 mm. at the base and 2 mm. near the tip. The coenenchyma is about 1 mm. in thickness throughout ; the canal system is typical of the group, and there are two distinct main canals which define two longitudinal bare spaces, although Wright aud Studer refer to only " a very narrow median groove." The polyps are disposed in two longitudinal series ; but a torsion of the whole colony has resulted in a false spiral appearance. The verrucae are small and dome-like ; some are adpressed to the stem; while others are almost retracted within the coenenchyma. The axis is of the typical Juncellid structure. "Wright and Studer thus define the spicules, of which the chief types are shown in fig. 109 : — " The spicules consist of (1) salmon-coloured spindles, 012 x 0^04 mm.; 0-08 X 0-02 mm. (2) sherry-coloured double-clubs: 0-1 x 0-06 mm.; 0-06 X 0-04 mm. (3) Xeedles; 006 x 0-02 mm. Locality. — Banda Islands. Note. — Fig. 110 of the Torres Straits specimen of Juncclla gcmmacea, which was originally described as Ellisdla maculata, has been added here to illustrate convergence in the group, and show how futile it is to attempt to separate Juncellids into genera without an examination of the spicules. XXX. Scirpearia quadrilineata n. sp. Figs. 111-113. It has been found necessary to establish this new species to include a very distinctive specimen in which the most predominant feature is the presence Simpson — A Revision of the Gorgonellidac. 1 1 3 of four main longitudinal canals, and Ihe consequent distribution of llie verrueae in four longitudinal series. The colony is complete, simple, and flagelliform ; the basis of attachment is conical, covered with coenenchyma, and spread over a piece of rock. The total height of the colony is 35 cm. ; the diameter at the base is 4-5 mm. ; near the tip it is 2 mm. A very noticealjle feature in the general appearance of tlie colony is the fact that it is markedly square in section. The coenenchyma is thin ; near the base it is 0'75 mm. in thickness ; but near the tip it approaches 1 mm. Around the periphery of the axis there is a system of longitudinal canals, of which /o^o- are markedly lai-ger than the others; these are arranged symmetrically, equidistant from one another; and thus forming the corners of a square (fig. 111). No outer system of longitudinal canals was visible ; but the coenenchyma isso thin that these may be easily overlookiid. It is extremely difficult to cut through the coenenchyma without damaging it, so that it is quite possible that these are present. The polyps are disposed in a very characteristic fashion. They are grouped in four definite longitudinal series, separated by four bare spaces which correspond in position to the four main canals (fig. 112). Each series consists of a single row ; but near the middle of the colony they are somewhat crowded, and give an appearance of two rows, duo in great part to displacement. Near the l)ase and towards the tip they are more openly arranged, but always in four series. The verrueae are low and dome-like, and have a maximum height of 0-5 mm. Towards the tip of the colony and near the base they tend to become almost level with the coenenchyma; while tl»e extreme basal portion is quite destitute of polyps. When partially closed they show a very distinct eight-rayed figure. The anthocodiae are very small, and arc all retracted w^ithin the verrueae. The axis is made up of concentric laminae ; it is extremely limy and very hard ; the colour of the outside is brown, but the core is white ; the surface is marked by faint longitudinal striae. Near the base the diameter is .S mm. ; but towards the tip it becomes almost hair-like and less limy. The spicules are quite distinctive. We have figured six types (fig. 113). (ff) Double-clubs with very densely warted and regular heads; the constriction is very short; and the warts are symmetrically arranged : O'OG x O'O.j ; O'OTO x 0040 ; O-OTG x 0-046. 114 Proceedings of the Royal Irish Academy. {b) Smaller double-clubs, with a longer constriction, with more open heads, and with the warts less symmetrically arranged : 0'06 X O'Oo-i ; 0-05 X -031 ; 0-049 x 0-0:27. (c) Elongated double-clubs, tending towards double-spindles, with rounded blunt ends: 0-091 X 0-038 ; 0-0S7 X 0-03-1 ; 0-083 X 0-03-i. {d) Elongated double-spindles, with pointed ends, and with a definite constriction: 0-118 x 0-03-4; 0-114 x 0-31; 0-114 x 0-023; 0-103 X 0-023. (e) Long spindles with a hint of constriction: 0-125 X 023; 0-114 X 0-031. (/) Shorter spindles also with a hint of constriction: 0-095 x 0019; 0-087 X 0-015; 0-016 x 0-023. We have little hesitation in defining («;, (&), and (c) as distinct types ; but it is just possible that (/) might develop into (c) or (d) according as increase with growth was greater in length or in breadth. So many of each kind occur, however, that we feel justified in defining them as separate for the present, at any rate, until more is known with regard to their growth. The colour of the coenenchyma is a bright orange-red — but the tips of the verrucae are more reddish. Zocaliti/. — Laccadives, 30-40 fathoms. Diagnosis, colony simple ; spicules contain double-clubs and double- spindles, with transitions to spindles. The coenenchyma is thin, and contains fotir main longitudinal canals. The verrucae are disposed in four definite longitudinal series, separated by four bare tracts, which correspond in position to the four main canals. The colony is markedly square in section. XXXI. Genus Nicella emend. (a) Discussion of the Genus. This genus was established by Gray in 1870 (Cat. Lith. Brit. JIus., p. 40) in the following terms : — Coral fan-like, in one plane, branched ; branches forked, rather diverg- ing. Bark smooth, brown. Polyp cells cylindrical, truncated, diverging from the stem at nearly right angles , mouth open. Axis calcareous, white solid. To this genus he refers a specimen under the name Kicclla mauritiana, and gives as a synonym his previous Scirpcaria dicliotoma (P.Z.S., 1859, 481-2). Kidley (xxix., p. 130) identified a specimen from Mauritius under the name Nicella dichotoma, and made the following observation on the spicules : "There is a dense cortical layer of small double-heads and a subjacent Simpson—^ Revision of the Gorgoncllidae. 115 layer of longer densely tuberculate spindles, having a l)are median space more or less clearly indicated." "Wright and Studer, with these facts as a basis, give the following diagnosis : — "The colony is upright, branched, with a thin coenenchyma and protruding verrucae, which arise perpendicularly, and appear to be terminally truncated. The polyps arise from either side of the stem and branches, leaving a middle space free. The spicules form a cortical layer of small double-clubs and an internal layer of long densely warted spindles." The following species have been from time to time referred to this genus : — -A^. dichoiovia (Gray). N. manritiana (Gray). JV. Inxa Whitelegge. N. flahdlata (Whitelegge). N. reticulata Thomson and Simpson. N. pustulosa Thomson and Simpson. An examination of the type-specimen of Scirpearella moniliforme Wright and Studer, in the Collection of the British Museum, has revealed the fact that this species should be included in the genus Nicella. Thomson and Henderson also referred Verrucella flaldlata Whitelegge to this genus, so that the generic diagnosis has been emended to include these forms. Thomson and Simpson (xli., p. 2G7) referred a specimen in the Littoral Collection of the Indian Museum to this genus under the name Nicella pustulosa, with tlie following reservation : — "It is with some hesitation that we refer this type to the genus Nicella. It is a matter of no small difficulty to distinguish between Nicella, Gorgonella, and Verrucella. . . . " Our specimens approach Nicella in several respects, though agreeing with none of the described species; and as the positive characters of the other genera are absent, we feel justified in making a new species to include these forms." The present study of this genus has, however, convinced me that the presence of the abnormally large spindles is a character which cannot be overlooked; so that, while still acknowledging the specific rank of the specimens under consideration, I would suggest their withdrawal from the genus Nicella, but until a revision of the species of Verrucella and Gorgonella has been made I would not hazard an oj)inion on their generic position. With Q2 1 1 6 Proceedings of the Royal Irish Academy. regard to the spicules of this group, we also made the following observation : — " Distinctions based on spicules alone are very unsatisfactory in this group (Gorgonella and Verrucella), because the spieulation varies at different levels ; and transition forms are so numerous and varied that it is sometimes almost impossible to distinguish between double-spheres, double-stars, and double- clubs, each in turn passing gradually to double-spindles. In Verrucella . . . there are double-stars ; in Gorgonella . . . double-spheres occur." As I have elsewhere pointed out, I doubt very much the validity of these two genera, on the present spicular distinction, but await a revision of the known species for a solution of the difficulty. (b) Classification of tlie Species with emended Diagnoses. On this basis four species may be recognized, and are included in this report. These are : — iV. dichotoraa Gray. N. flahellata (Whitelegge). N. reticulata Thomson and Simpson. N. monUifwme (Wright and Studer). The following short specific diagnoses may prove useful : — Nicella fiMbellata. The colony is branched in one plane ; the smaller branches tend to arise from one side of the larger. The coeuenehyma is moderately thin, and often presents a ridged appearance due to segregations of spicules. The polyps are disposed in two longitudinal series; in the younger part they occur in a sinuous row on either side of the branch ; but in the older portions they are more numerous and may encroach slightly on the median bare spaces. The venucae vary in shape and size according to the stage of retraction ; when expanded they are prominent, and show an eight-i"ayed figure at the summit ; when retracted they appear as low conical warts, and there is no trace of an octo-radiate structure. The axis is composed of concentric laminae ; and the surface is marked by longitudinal striae. The spicules consist of (1) small double-clubs, {2} small double-wheels, (3) elongated double-clubs, (4) long, massive, bluntly terminating double- spindles, (5) long, slender simple-spindles. (See fig. 115.) Nicella reticulata. Colony branched in one plane, with abundant anastomosis ; ilie branches and twigs are \ery slender, so that the colony is extremely reticulate and fiabelliform. The coenenchyma is thin and finely granular. The polyps are SiMi'SON — A RciHsion of tlie Gor(joneUi(l(ie. 117 disposed mainly in two longitudinal series; but deviations from this type occur, owing in some cases to overcrowding, in others to the anastomosis. The verrucae are usually low and dome-like. Tlie spicules consist of (1) small double-clubs and elongated double-clubs, and ('!) lung duuble-spiudles and simple-spindles. Tiiese two sets are quite distinct ; but the spindles are not so disproportionate in length to the double-clubs as in most other species. Nicdla moniliforme. Colony simple or feebly branched, slender, filiform, and of almost uniform diameter throughout ; polyps disposed in two longitudinal series, near the tip in one row, but in the older parts in two or more indefinite rows in each series. The spicules are very characteristic. They include small (li,iiil)le- clubs and elongated slender double-clubs : also spintlles of two kinds (1) long, slender, ninny spindles, and (2) loiuj, thick, denseli/ warlcd spindles. The spindles are sometimes more than twice as long as the typical double- clubs. (See figs. 117 and 118.) XXXI 1. Nicella dichotoma Gray. Fig. 114. Scirpearia dichotoma Gray, xi., p. 481. Nicella mauritiana (Gray), xii., p. 40, fig. 12. non Media mauritiaiM Studer. Nicella dichotoma Ridley, xxix., p. 130. Nicella dichotuma Thomson and llussell, xliii.,p. IGl, I'l. vii., tigs. 1 and 5. Nicella laxa Whitelegge, xlix., p. 319, Tl. xvii., figs. 30-33. This species was established by Gray in 1859 under llie name of Scirpcarm dichotoma. He defined it tlius : — " Coral fan-like, in a single plane, irregularly dichotomous. Gells cylindrical, elongateil, truncated, in a row on each side of the branches, sub-alternate." Z(3ca/%. —Mauritius. In 1870 ho formed another species, Nicella mtmritiana, while he gave as a ^ywitwywx Scirpearia dichotoma. Since this new species is the same as the older dichotoma, it was unnecessary to give it a new name, although he referred it to a new genus, so that the newer name must give way to the older. The description of Nicella mam-itiana is as follows : — " Coral fan-like, dichotomously branched ; stem cylindrical, longitudinally striated; bark thin, pale brown; cells elongate, cylindrical, longer than the diameter of the stem, ascending, truncated at the tip, placed rather irre- gularly, sub-alternate (rarely sub-opposite) on each side of the stem an.l branches; axis pale greyish-brown." ZotW%.— Mauritius. Eidley in 1882 re-identilied the species, and dcscribeil some specimens 118 Proceedings of the Royal Irish Academy. from Mauritius, giving some positive, additional characters. One of his specimens was 340 mm. in height, and 240 mm. in maximum diameter. He says : — " The shape of the verrucae varies considerably according as to whether they are open or closed ; in the former conditiou they are rectangular at the apex, while in the latter they appear conical with rounded apices. The basal diameter may vary from V2b mm. to 2'25 mm. when closed. The spicules consist of a dense cortical layer of small double-heads and a sub- jacent layer of longer densely tiiberculate spindles having a bare median space more or less strongly indicated. The colour is variable, (1) ochreous yellow to a dull flesh colour, (2) dirty white." In 1897 Whitelegge established a new species under the name Nieella laxa with the following characters : — The colony is feebly branched ; the branching is lateral and in one plane. The axis is laminate and calcareous. The coenenchyma is thin, and when viewed with a lens presents a series of minute ridges, forming a network of raised lines, which are lighter in colour and consist of double-club spicules. The polyps are confined to the sides of the stem and branches in a single row on each side. The verrucae are large, alternate, and stand nearly at right angles ; they are divided at the summit into eight lobes. The spicules consist of (1) short double-clubs with smooth or warty tubercles: 01 X 0-05 mm.; O'OV X 0-03 mm.; 0-05 x 0-02 mm.; (2) fusiform spindles with rather obtusely pointed ends and a spiny tuber- culated surface : 0-25 x 0-06 mm.; 0-2 x 0-05 mm.; 01 x 0-03 mm. Many of both kinds are a little flattened. The colour is a light mouse-grey. He says: — "This species differs from N. dichotoma by its smaller and more distant polyps and by its lax method of branching." We have already seen that neither of these two characters is of much taxonomic importance ; and, taking into consideration Eidley's observations on the size of verrucae in diflerent stages of retraction, we do not feel justified in ranking this as a distinct species. At the same time Gray gives a very good figure of his N. mav.ritiana, and the branching there is almost identical with that figured by "Whitelegge. In Gray's figure also the distribution of the polyps varies in different parts of the colony, so that while in some branches they are more closely packed, in others they are quite as distant as in Whitelegge's figure. The spicules are identical with those described by Eidley ; and the network of ridges described by Whitelegge, though not given in Gray's description, are unmistakably present in his figure. We therefore see no reason for rarking N. laxa as a separate species. Thomson and Eussell, 1909 (xliii., p. 161, Plate vii., figs. 1 and 5) describe some specimens as follows : — Several colonies of chestnut-browu to umber-brown colour. The largest is Simpson— .1 Revision of the Gorgonellidae. 1 19 20 cm. in heiglit by 8 cm. in maximum breadLli, and consists of a main stem, with lateral branches, which are again repeatedly branched. For the most part the branching is in one plane ; but this is not rigorously adhered to. On the main stem of one of the larger specimens there is a curious gall-like swelling from which branches arise on all sides. The stem is 4 mm. in thickness at its base, and gradually tapers to 2 mm. at the ends of the branches. The axis is light brown in colour, and verj' calcareous. On the surface of the general coenenchyma, and on the verrucae there are irregular wavy longitudinal ridges, producing a characteristic bark- like appearance. Under the low-power microscope the textui'e seems finely arenaceous. The verrucae are very prominent, rising more or less perpendicularly to a height of 2 mm. They occur on all sides of the stem; l>ut in tlie upper parts of the branches a bilateral arrangement is well dotiued. At Ihe apex of the verrucae there is an indication of eight lobes, from which the tentacles here and there project. Another specimen, the basal part of a large colony, branches in a some- what irregular fashion, and not rigidly in one plane. The verrucae are much less bilateral, especially near the base of the colony. Examination of the spicules shows that this may be referred to N. dkhotoma. Locality. — Salomon A, 65 fathoms ; Salomon B, 60-120 fathoms. XXXIII. Nicella flabeUata (Whitelegge). Fig. 115. Verrucella flabdlata Whitelegge, xlix., p. 319, Plate xvil., figs. 34-37. Nicdla flabcllata Thomson and Henderson, xl., p. 80. This species was established by Whitelegge for a specimen from Funafuti, but was tlien included in the genus Verrucella. Thomson and Hendersun, in identifying a specimen from the Indian Ocean with this species, concluded that it should really be referred to the genus Nicella : and in this we thoroughly agree. The spiculation is quite distinctively Nicellid in character ; and, as these authors point out, the actual shape of the verrucae matters little in a generic diagnosis. As a matter of fact, the natuie of the verrucae, as .shown in the figure given by Whitelegge, is intermediate between that in N. dichotoma and the Indian Ocean specuuon. The notes following may serve to indicate the chief specific characteristic. The colony is branched in one plane ; the branches show a tendency to arise from one side. The axis is densely calcareous and is striated. A noteworthy feature is the presence of two distinct grooves corresponding in position to the two main canals. 120 Proceedings of the Roi/al Iris// Acculemj/. The polyps occur in a siunous row on each side of the younger branches ; on the stem and on the older portions of the branches they are more numerous, and encroach on the two bare, flattened surfaces, always leaving a slight median depression free. Those on opposite sides alternate. The verrucas may be slightly prominent or may appear as low conical warts. When partially retracted, they show an eight-rayed figure ; but when fully withdrawn, this is not evident. An average height may be taken as 1 mm. The coenenchyma is of medium thickness, and may have ridges on the surface. The canal system is the typical Juncellid. The si)icules are essentially of two types, viz., small double-clubs and long thick double-spindles. The double .spindles are about four times as long as the small double-clubs. There are, however, in addition to these two types : — (1 ) some small double-wheels, with elongated warty hubs ; (2) elongated double-clubs ; (3) long slender spindles with practically no constriction. Very small short rods and spiny spindles occur in the tentacles. The colour of the Funafuti specimen was yellowish-white : that of the Indian Museum specimen was ochreous yellow and brownish-white. XXXIV. — Nioella reticulata Thomson and Simpson. Fig. 116. Nicdla reticulata Thomson and Simpson, xli., p. 266, Plate iv., fig. 5 ; Plate VIII., fig. 12. This species was established Ijy Thomson and Simpson (xli., p. 266) for specimens in the Indian Museum Littoral Collection. We have considered it advisable, liowever, to recapitulate the original description for the sake of completeness. A typical colony measures 27 cm. in height by 16 cm. in maximum breadth, and is attached by a very much broadened expansion. It consists of a main stem, only 2 cm. long, and measuring 4'5 mm. in diameter. At the distal end of the main stem four branches arise, two suli-opposite and two at slightly different levels, but all very close together. These divei'ge at varying angles, the two lower being almost horizontal, the other two also in the same plane of ramification. These ramify irregularly in one plane and anastomose freely, forming a large, almost semicircular, flabelliform mass, with very irregular meshes. The coenenchyma is tliin and compact, and presents a glistening arenaceous appearance. The colouring is very peculiar, being generally reddish-brown in the lower part of the colony, but gradually merging into slaty grey in the upper parts. Patches of grey appear throughout the red in some of the colonies, and vice versa ; while one colony from the Laccadives is almost uniformly of a brick-red colour. The surface bears longitudinal Simpson — .4 Revision of the Gorgonellidue. 121 furrows, which are simious, and sometimes almost spirally twisted ; one being generally deeper than the others. These extend into the secondary branches, and even into one side of the twigs, the number diminishing with the size of the branches. The axis is very calcareous and cylindrical in form. It is composed of concentric laminae, and has an ahnost olive-green colour at tlie base, gradually merging into a pale yellow in the smaller branches. The polyps are disposed essentially in two longitudinal series; but deviations occur in several places, due sometimes to the anastomosis and sometimes to overcrowding. They are chiefly lateral on the main stem or primary branches ; in the secondary branches they are arranged almost all round. On the finer branches and twigs they occur for tiie most part on two sides ; but this rule is broken occasionally by the occurrence of polyps on all the four sides. The verrucae are dome-like, but slightly flattened on the twigs. They are separated by intervals of about 1 mm. in the branches ; but their bases touch on the branchlets and give an undulating appearance. They measure about 0-5 mm. in lieight and 1 mm. in diameter. "^^Hien the verruca closes over the retracted polyp, an eight-rayed star is formed by the eight lobes of the wall. The anthocodiae are very minute and are completely retractile ; the spicules are arranged transversely on the tentacles. The spicules of the coenenchyma consist of small doable-clubs, elongated double-clubs, double-spindles, and simple-spindles. The double-spindles and simple-spindles in this species are not so markedly disproportionate as in most other species ; but their distinctive character justifies their inclusion in the genus Nicella. The following are a few of the more common types, witli measurements in millimetres : — (a) Double-clubs, with smooth warts : 0-05 X 0-04 ; constriction 0-02 broad X 0-008 long. 0048 X 0-04 „ 0-02 „ x 0-005 „ (5) Elongated double-clubs, with fewer and more irregular warts : 0-06 X 0-04 ; constriction O'Oo broad X 0-01 long. 0-048 X 0-035 „ 0-02 „ x 0-012 „ (c) Spindles with round warts, and double spindles, having a smooth part in the middle : 009 X 0-025 ; smooth part, 0-02 long. 0-085 X 0-028 „ „ 0-018 „ (rf) Minute crosses, with a very distinct cross, 0-04 x 004. (e) Minute irregular crosses, elongated along one diagonal, with distinct cross, 0-05 x 0-03. R 122 Proceedings of the Royal Irish Academy. Those of the tentacles are short, warty rods : 0-05 x 0-015 ; 0-06 X 0-015 ; 0-6 X 0-015. Localities. — Persian Gulf, 48-49 fms. Laccadives, 30-50 fms. XXXV. — Nicella moniliforme emend. Figs. 117 and 118. Scirpearella nioniliforme Wright andStuder, p. 156, PI. xxxiv., fig. 8. non. Gorgonia moniliforme Lamx., xxv., p. 420. nee. Scirpearella moniliforme Thomson and Henderson, xl., p. 82. nee. Scirpearia monilifm'mis Gray xii., p. .39. This species, as established by Wright and Studer in the " Challenger " Report, is a very distinctive one, based chiefly on the character of the spiculation. The colony may be simple or feebly branched ; the branched type-specimen was 505 ram. in length; and the branch arose at a distance of 215 mm. from the base ; one of the uubranched forms was 325 mm. in length. The colonies are very slender and do not vary much in diameter throughout the entire length. The coenenchyma is thin and coarsely granular. " The axis is very deejDly grooved ; ten grooves can be very easily counted on the older portion of the axis ; but these diminish to two at the apex. These ridges show through the coenenchyma as linear furrows." " The polyps are arranged on the stem, the lower portion in four irregular rows; towards the apex they are alternate and arranged on either side of the stem; while for the first 60 mm. of the stem, counting from the basal disk, they are absent. They are retractile within the well-marked Ijut shallow verrucae ; these latter measure at their base 1 mm. An occasional verruca will be found larger and more elevated than the rest, measuring 1-5 mm. in diameter and the same in height ; these generally are to be found near the summit of the axis." The disposition of the verrucae is in two longitudinal series; and the two bare tracts are marked by distinct furrows larger than the others. Unfortunately Wright and Studer give no figure of the colony itself ; and, as the figure of spicules is somewhat misleading, we have thought it advisable to add to this memoir two figures from the type-specimen in the British Museum (figs. 117 a, b, and c). The colour in spirit is white. The nature of the spicules in this species and also their relative proportions are very striking, and mark it off as distinct. The following four types can easily be identified: — («) long, comparatively slender spindles, covered with coarse spines or small warts ; (b) long, thick spindles, very Simpson — A Revision of the GorgonelUdae. 123 densely waited ; (c) slender double-clubs, with elongated conical ends, and with the constriction more or less marked ; {d) small double-clubs, with almost hemispherical ends and with a definite smooth constriction : aberrant forms, such as crosses, (c) also occur. There are small needles in the anthocodiae. The following measurements, length by breadth in millimetres, will give the relative proportions of these different types (see fig. 118) : [a) Spindles — long, thin spiny or with small warts : 0-2 x 0-034 ; 0-15 X 0-026; 0-13 x 0-02. {h) Spindles — long, thick and densely warted : 0-15 x 0-046 ; 0-13 x 0-042. (c) Double-clubs — slender, with elongated ends, and with the constriction more or less markedly defined: 0-11 x 004.") ; 0-o99 x 0-043; 0-087 X 0-03; O-OOo x 0-02.5. {(1) Double-clubs — with massive ends, and witli a distinct, short, smooth constriction: 0 072 x 0-042; 0-072 x 0-038; 0-057 X 0-038. (e) Crosses— 0-16 x 0-11 ; 0-12 x 0-12. (/) Needles— small (in anthocodiae) : 0-06 x 0-011; 0-04 x 0-02. Locality — Amboina : 100 fathoms. XXXVI. Bathy.metrical Distribution-. The whole group is essentially littoral in its distriliution. The great majority of the specimens hitherto described have been dredged within the hundred-fathom line ; in fact, the only records outside this range are from (1) " Challenger " Station 232, known as the Hyalomma-gvownd off Japan, 345 fathoms; (2) "Challenger" Station 177 off tlie New Hebrides, 130 fathoms; (3) a dredging made by the "Investigator," off the Andamans in 124 fathoms; and (4) off the Azores, 150 and 200 fathoms. At the first of these Juncclla raccmosa and Seirpearia profunda were obtained, at the second Seirpearia profimda, at the third only Juncella racemom, and at the fourth only Seirpearia flayellum. Consequently these are the only three species whioli can lay claim to deep-sea forms ; and it is interesting that all the records of these species are from over 100 fathoms, and also that each has been found in distant localities over this deptli. At the same time it is not improbable that these specimens occurred in deep water at the edge of an almost vertical reef, and that these were merely " escapes " from tlie reef. Such records are not unknown ; and the writer has experienced similar occurrences in the deep water off the almost perpendicular reefs on the east coast of Africa. E2 124 Proceedings of the Royal Irish Academy . Nicella monUifm-me is recorded from oue hundred fathoms, and the only other records from over fifty fathoms are Scirpearia thomsoni and Scii-pearia alba, both from 88 fathoms. The following table will give at a glance the chief records for each of the species in this report : — SPECIES. DEPTHS IN FATHOMS FROM WHICH RECORDED. JunccUa juncca, . 0-10; 4 ; 7-11; 15-35; 25-30; 45; 50 Juncclla gcmmacca, 0-8; 4; 8; 11; 19; 12-20; 32. Juncella racemosa. 120; 345. Juncella trilineata, 34. Scirpearia profunda. 130 ; 345. Scirpearia hicksoni, 36. Scirpearia veniicosa. 50. Scirpearia anomala, ? Scirpearia pectinata, 3-4; 12; 30. Scirpearia clongata, Scirpearia fiagellum, 90; 150; 200; 60-120. Scirpearia thomsoni. 88. Scirpearia alba. 88. Scirpearia aurantiaca, . 30-50; 60-130; 120; 150; 130. Scirpearia fircata. 15. 20 30-40 ; 50-78. Scirpearia andamanensis, ? Scirpearia ramosa, 20. Scirpearia ceylonensis, . ? Scirpearia mamdata. ? Scirpearia C[uadrilineata, 30-60. Nicella dichotoma, ? 60-120. Nicella fiabellata. 45. Nicella reticidata, 30-50; 48-49. Nicella moniliforme. 100. It is quite probable, however, that when more inshore-work is carried on in tropical seas records will be abundant from water of much less depth than that at present given. Eidley in referring to the depths at which Juncclla gemmacea occure gives " between tide marks," and, as has been already pointed out in the " Biological Note," it is no uncommon occurrence on the scattered coral reefs of the Mergui Archipelago to see at low spring tide huge colonies of J. gcmmacca and J. juncca as well as Melitodcs and other Alcyonaria swaying to and fro in the aii'. Simpson — A Revision of the Gorgonellidae. 125 XXXVII. Geographical Distkibution. The great importance of the Geographical Distribution of even a small group of animals, but especially those whose early life is pelagic and whose adult life is sedentary, is becoming more and more evident. Such knowledge, combined with systematic oceanographical observations, may eventually help to solve many problems that at present are a source of great perplexity to the biologist. It is premature to attempt such a distriliution of .luncellids; but in view of the fact that in this memoir a general survey of the group, so far as it is known, has been given, and as the references to localities, especially in the case of the older species, are extremely scattered, the following summary may serve as a basis for a more detailed study when furtiier records are forthcoming. Although doubt may exist as to the specific deterniinaliun of those species added as an appendix to the genus Juncella, it may be useful to include them here, inasmuch as they are in all probability Juncellids. It has been considered inadvisaV)le with the limited records at our dispo-sal to draw any conclusions as to tlie dispersal of these organisms, as to their origin as a part of a littoral fauna, or as to the probability of their being originally indigenous in certain areas. Distribution of the Juncdla-group of Gorgonellids. The Juncella-group of Gorgonellids occurs botli in the Atlantic and Pacific waters, but almost entirely withui the Tropics of Cancer and Capricorn, and also chiefly in the Pacific Ocean. The extreme records North and South are " Off Japan " and " OH' Cape Colony." The following are the chief centres : — (1) Red Sea, (2) Persian Gulf, (3) Laccadives, (4) Maldives, (5) West Coast of India, (6) Ceylon, (7) Andamans, (8) Mergui, (9) Bourbon, (10) Mauritius, (11) Cape Colony, (12) East Indies, (l.S) Japan, (14) East Coast of Australia, (15) West Coast of Florida and in the Atlantic, (16) East Coast of Central America, (17) N.-E. of South America, {lH) Azores, (19) Mediterranean Sea. Genus Juncella. This is the most widely distributed genus in tlie group, and is almost entirely a Pacilic Ocean form. Genus Seirpcaria, This genus is entirely restricted, with the exception of S.flagclhim, so far as the present records show, to the Pacific Ocean. 126 Proceedings of the Royal Irish Academy. Genus Nicella. This genus is entirely restricted to the Pacific Ocean. Let us now illustrate " associations of species " in different localities. (a) Laccadives, (h) Maldives, . (c) Ceylon, {d) Andamans, (e) Mergui, (/) N.-E. Australia, . ((/) Bourbon-Mauritius, Qi) Cape of Good Hope, S. aurantiaca, S. quadnlineata, and N. reticulata. J. juncea, S. furcata. J. gemmacea, J. trilineata, S. aurantiaca, S. ceylon^nsis, S. furcata. J. juncea, J. racemosa. S. hicksoni, S. verrucosa, S. anomala, S. anda- manensis, S. rarnosa. J. juncea, J. gemmacea, S. furcata. J. juncea, J. gemmacea. J. juncea, J. gemmcuxa, N. dichotoma. S. flagellum, S. furcata. Beferervces to various large Collections of Juncellids. " Challenger " Collection. This collection was made by H.M.S. " Challenger," during her cruise round the world, 1873-76. The specimens are deposited in the British Museum, and were described by Wright and Studer in the Zoological Report of the " Challenger" Collections, vol. xxxL, pp. 153-181. Juncella juncea, Juncella juncea, Juncella gemnuicea, . Juncella gemmacea, . Juncella racemosa, Scirpeai-ia maculata, Scirpearia profmda, Sdrpearia profunda, Scirpearia pi'ofunda, Nicella moniliforme, DESCEIBED AS Juncella juncea, var. alha, p. 168. Junt,ella harbadensis, p. 159. Juncella gemmacea, p. 158. Ellisella nuiculata (pars), p. 160. Juruxlla racemosa, p. 159. Ellisella maculata (pars), p. 160. Seirpearella profunda, p. 155. Scirpearella gracilis, p. 156. Seirpearella rubra, p. 157. Sciipearella monilifwme, p. 156. Simpson — A Revision of fhe Gorgonellidae. 127 "Alert" Collection. This collection was made during the Surveying Voyage of H.M.S. " Alert " during the years 1881-82. The Gorgonellids were reported on by Eidley in " The Zoological Collections " of H.M.S. " Alert," 1884, pp. 345-349. DESCRIBED AS . Juncclla jvMcca, p. 345. . Juncella fragilis, p. 347. . Juncella gemmacca, p. 346. . Juncclla elongata, var., p. 346. . Ctcnocclla ixctinata, p. 348. . Ellisclla calamus, p. 348. Ceylon Collection. This collection was made by Professor Herdman in the Ceylon Seas in 1904 while investigating the Pearl Fisheries of the Gulf of Manaar. The type-specimens are deposited in the British Museum, and were reported upon by Thomson and Henderson, " Ceylon Pearl Oyster Eeport," Poyal Society, 1905. Supplementary Eeport, No. sx., Alcyonaria, pp. 311-315. described as Juncclln juncea, Juncclla juncea, Jtincella gemmacea, Juncella gemmacea, . Scirpearia pectiTiata, Scirpearia maculata, Juncclla juncea, Juncella juncea, Juncella juncea, Juncclla juncea, Juncclla trilineatii, Scirpearia furcata, . Scirpearia ficrcata, . Scirpearia aurantiaca, Scirpearia a^irantiaca, Scirpearia ccyloncnsis. Juncclla juncea, p. 314. Juncella gcmnuicm, p. 313. Juncella fragilis, p. 314. Juncclla fragilis, var. rubra, p. 314. Juncella trilineata, p. 315. Scirpearia sp. (?), p. 313. Scirpearella sp. B., p. 312. Scirpearclla aurantiaca, p. 311. Scirpearella divisa, p. 312. {undcscribed). Maldive Collection. 1. This collection was made by Mr. Stanley Gardiner in 1900, and was described by Hickson in "The Fauna and Geography of the Maldive and Laccadive Archipelagoes," vol. ii., part iv. " The Alcyonaria of the Maldives," part iii., pp. 816-823. described as Juncella juncea, . . . Juncella juncea, -^.820. Jun/:ella juncea, Juncella jurwca, Scirpearia furcata, Scirpearia furcata, Scirpearia furcata, Scirpearia furcata Juncclla flexilis, p. 821. Juncella elongata, p. 821. Sciipcaria furcata, p. 822. Scirpearia furcata, var., p. 822, Scirpearclla indica, p. 822. Juncella elongata (Val), p. 821. 128 Proceedings of the Royal Irish Academy. Maldive Collection. II. (described by Thomson and RusseU, 1910). DESCRIBED AS Juncella gemifnacea, . . Jimcella gemmacea. Scirpcaria flagcUum, . . Scirpearia flagellum. Scirpmria aurantiaca,. . Scirpcarella ourantiaca. Nicella dichotoma, . . Nicella dicJiotoma. Monaco Collection. I. The collections made by the Prince of Monaco, during the scientific voyage of the yacht " Hirondelle" in the North Atlantic Ocean, in 1886-88, contain several Gorgonellids. These have been reported npon by Studer, in "Eesultats des Campagnes Scientiiiques du Prince de Monaco," 1901, fasc. XX., pp. 52, 53. DESCKIBED AS Scirpearia JlageUvm, . . Scirpearia flagellum, p. 53. Scirpearia flagellum, . . Scirpearia ochracea, Tp. 53. Indian Museum Deep-Sea Collection. This Collection was made during the cruise of the old E. I. M. SS. " Investigator " in the Indian Ocean. The specimens are deposited in the Indian Museum, Calcutta, and were reported on by Thomson and Henderson, in the memoirs of the Indian Museum, Alcyonaria (1906). Juncella racemosa, Scirpearia profunda, Scirpearia alba, Scirpearia thomscmi, Nicella flabellata, . DESCRIBED AS Juncella miniacea, p. 81, Scirpcarella moniliforme, p. 82. Scirpcarella alba, p. 82. Jimcella clongata, p. 81. Nicella flabellata, p. 80. Indl&n Museum Littoral Collection. This Collection was made during the surveying cruises of the E. I. M. SS. " Investigator" in the Indian Ocean. The type specimens are deposited in the Indian ^Museum, Calcutta. They were reported on by Thomson and Simpson, in the Memoirs of the Indian Museum Alcyonaria, 1909 ; but specific names were given only to a few ; Simpson — .1 Revision of the Gorgonellidne. 129 descriptions of the others were tcabiilated, so that the following list will enable these to be identified. DESCRIBED AS Jufuxlla juncea, . . . E. and F. Juncella gcmmxaceM, . . 0. Juncella triliiicata, . . R. Scirpearia pectinata, . . M. Seirpea7'iu andamaiiensis, . N. Scirpearia, anomala, . . Q. Scirpearia aurantiaca, . . B. Scirpearia fureata, . . H, G, D, I. Scirpearia furcata var. robusta, P. Scirpearia ramosa, . . K. Scirpearia verrucosa, . . C. Scirpearia hicksoni, . . A. Scirpearia quadrilineota, . J. Wood-Mason Collection. This Collection was made by W. J. Wood-Mason in the Indian Ocean. A few of the specimens were described by Thomson and Simpson, but the majority of them were left over for incorporation in this paper. The types are deposited in the Indian Museum, Calcutta. Juncella i-accmosa. Juncella gemmacea. Scirpearia aurantiaca. Scirpearia furcata. Nicella Jiabellata. Mergui Collection. I. This Collection was made by Dr. John Anderson for the trustees of tlic Indian Museum, Calcutta, where the specimens are deposited. They were described by Ridley in the Journal of the Linnean Society, vol. .\.\i., pp. 240-243. DESCRIBED AS Juncella juncea, . . . Juiuxlla fragUis, var., p. 242. Juncella gemmacea, . . . Juncdla gemmacea, p. 241. Scirpearia pectinata, . Ctenocella peetinata, p. 24:3. 180 Proceedings of the Royal Irish Academy. Mergui Collection. II. This Collection was made lay Simpson and Brown in the Mergui Archipelago, Burma, in the spring of 1907. The specimens are deposited in the Natural History Museum, Aberdeen University, and are reported on here for the first time. They include the following species : — JunccUa jimcca. Juncella gemmacea. Scirpearia yectiimta. Scirpearia furcata. Schyearia furcata var. rohusta. Australian Museum Collections. This Collection was made by Mr. C. Hedley for the Australian Museum, where the specimens are deposited. It was reported upon by Whitelegge in the " Memoirs of the Australian Museum XII.," The Alcyonaria, Part ii. (1897 ?), pp. 318-820. DESCRIBED AS Nicella dichotoma, . . . Nicella laxa, p. 318. Nicclla Jlahcllata, . . . Verrucella flaiellata, p. 319. Simpson — A Revision of the Gorgonellidae. 131 BIBLIOGRAPHY.' i. 1817. CuviER, Regne Animal. ii. 1830. CuviEK, Regne Animal. Nouv. Edit., t. iii., Paris, iii. 1846. Dana, J. D., Zoophytes, United States Exploring E.xpedition. Philadelphia, iv. 1861. DUCHASSAING et Michelotti, Mdmoire sur les Coraillaires des Antilles. Mem. E. Acad. Sci. Torino, 2 ser. xix. V. 1866. DUCHASSAING et Michelotti, Supplement, ibid., xxviii., pp. 97-206. 11 Plates, vi. 1786. Ellis and Solander, The Natural History of many curious and uncommon Zoophytes. London, vii. 1797. EsPER, E. J., Die Pflauzenthiere in Al)bildniigen uebst Bes- chreibungen. Niirnberg, 1791-1797. 4 vols, viii. 1896. Germanos, N. K., Gorgonaeeeu von Ternate. Abh. Senck. Nat. Ges., xxiii., Heft, i., pp., 145-184. 4 Plates, ix. 1857. Gray, J. E., Description of new Genera of Gorgoniadae. Proc. Zool. Soc, pp. 158, 159. X. 1857. Gray, J. E., Synopsis of the Families and Genera of Axiferous Zoophytes or Barked Corals. Proc. Zool. Soc. xi. 1859. Gray, J. E., Description of some new Genera of Lithophytes. Proc. Zool. Soc, xxvii., pp. 479-486. xii. 1870. Gray, J. K, Catalogue of the Lithophytes or Stony Corals in the Collection of the British Museum, London, xii.ft. HiCKSON, S. J., The Cambridge Natural History, vol. i., pp. 329-364. xiii. 1900. HiCKSON, S. J., Alcyonaria and Hydrocorallinae of the Cape of Good Hope. Part 1, Marine Investigations in South Africa, i., No. 5, pp. 67-96. 6 Plates, xiv. 1904. HiCKSON, S. J., The Alcyonaria of the Cape of Good Hope. Part ii., Marine Investigations in South Africa, vol. iii., pp. 211-239. 3 Plates. XV. 1903. HiCKSON, S. J., Alcyonaria of the Maldives, Fauna and Geogr. Maldives and Laccadives, ii., pt. 4, pp. 807-826. Plate LXVII. ' The Roman numerals correspond to the numbers given in the text. " S2 132 Proceedings of the Royal Irish Academy. xvi. 1899. HiLES, IsA I.. The Gorgonacea collected by Dr. Willey. Willey's Zool. Eesults, pt. 11, pp. 195-206. 2 Plates, xvii. 1899. HiLES, ISA L., Report on the Gorgonacean Corals collected by Mr. J. Stanley Gardiner at Funafuti. Proc. Zool. Soc, London, pp. 46-52. 4 Plates, xviii. 1863. Johnson, J. Y., Description of a new species of Juncella. Proc. . Zool. Soc, London, pp. 505, 506. xix. 1864. Johnson, J. Y., Description of a new species of flexible Coral belonging to the genus JunceUa obtained at Madeira {J. flcujellum). Ann. Mag. Nat. Hist., 3 ser. xiv., p. 142. X.X. 1870. Kent, Saville, Trans. Eoy. Micro. Soc, p. 92. Plate xvii., fig. 38. xxi. 1893. Kent, Saville, The Great Barrier Eeef of Australia, xxii. 1897. Klunzinger, C. B., Die Korallthiere des Eothen Meeres. Part 1. Die Alcyonarieu und Malacodermen, p. 59. 4 plates, xxiii. 1865. Kolliker, A., Icones Histiologicae, pp. 131-142. 8 Plates, xxiv. 1816. Lamarck, Histoire Naturelle des Animaux sans Vertebres- vol. ii., Paris. XXV. 1816. Lamouroux, Histoire des Polypiers coralligenes flexibles. Caen, p. 419. xxvi. 1857. Milne-Edwaeds and Haime, Histoire ISTaturelle des Corail- laires ou Polypes proprement dits. 3 vols., 1857-1860. xxvii. Pallas, Charakteristik der Thierpflanzen. pp. 224-6. xxviii. 1766. Pallas, Eleuchus Zoophytorum. xxix. 1882. EiDLEY, Stuart 0., Contributions to the knowledge of the Alcyonaria. Part I., Ann. Mag. Nat. Hist., 5 ser. ix., pp. 125-1-33. 1 Plate XXX. 1882. KiDLEY, Stuaet 0., ibid., Part ii., loc. cit., 6 ser. ix., pp. 184-193. xxxi. 1883. EiDLEY, Stuart 0., The Coral-fauna of Ceylon, with descrip- tions of new species. Ann. Mag. Nat. Hist., 5 ser. xi., pp. 250-262. xx.\ii. 1887. liiDLEY, Stuart 0., Eeport ou the Alcyoniid and Gorgoniid Alcyonaria of the Mergui Archipelago, collected for the Indian Museum. Journ. Linn. Soc, London, xi., pp. 223- 247. 2 Plates. Simpson — A Revision of the Gorgonellidae. 183 xxxiii. 1884. Ridley, Stuakt O., Zoological collections of H.M.S. "Alert," London. xxxiv. 1878. Studer, Th., Febersicht der Anthozoa Alcyonaria welche wahiend der Reise S.M.S. -'Gazelle" uni die Erde gesammelt wurden. Mouatsber., Akad. Wiss., Berlin, pp. 632-688. 5 Plates. XXXV. 1887. Studer, Th., Versiich eines Systeme.'; der Alcyonaria, Arcli. Naturges., liii., pp. 1-74. 1 Plate. xxxvi. 1891. Studer, Th., " Note Pr(^liniinaire sur les Alcyonairesprovenant des Campagnes du Yacht I'Hirondelle." 2 Partie Mem. Soc. Zool. de France, vol. iv., Paris. xxxvii. 1894. Studer, Th., "Alcyonarien aiis der Sanunluiig des Natur- historisclien Museums in Liibeek," Mitteilungen der Geog. Gesell., und des Naturhist. Museums in Liibeek, 2 ser., Heft, 7 und 8, pp. 103-128. xxxviii. 1901. Studer, Th., " Alcyouaires provenant des Campagnes de rHirondelle." Eesultats Campagnes scientifiques du Prince de Monaco. Fasc. xx., pp. 04. 1 1 Plates, xxxix. 1905. Thomson, J. A., and Henderson, W. L>., Alcyonaria; Ceylon Pearl Oyster Fisheries Report. Roy. Soc, London, pp. 271- 328. 6 Plates, xl. 1906. Thomson, J. A., and Henderson, W. D., The Alcyonarians collected by the R.l.M.SS. " Investigator " in the Indian Ocean. The Alcyonarians of the Deep Sea. Calcuttai pp. 79-83. xli. 1909. Thomson, J. A., and Simpson, J. J., The Alcyonarians collected by tlie R.l.M.SS. " Investigator " in the Indian Ocean. The Littoral Alcyonarians. xlii. 1909. Thomson, J. A. (unpublished). xliii. 1909. Thomson, J. A., and Russell, E. S., Alcyonarians of Sladen Trust Expedition. Trans. Linn. Soc. (Zool.) xiii. (1910), pp. 139-164. xliv. 185-. Valenciennes, mss. Coll. du Janlin ties Plantes, Paris, xlv. 1855. Valenciennes, Abstract of a Monograph on the Family Gor- gonidae. Ann. Mag. Nat. Hist., vol. 16, ser. 2, pp. 177-183. xlvi. 1855. Valen'CIENNES, E.xtrait d'une Monographie de la Famille des Gorgonides de la class des Polypes. Comptes Reudus, xli., pp. 7-15. 134 Proceedings of the Royal Irish Academy. xlvii. 1864. Verrill, A. E., List of the Polyps and Corals (sent by the Museum of Comparative Zoology to other Institutions in exchange), with annotations. Bull. Mus. Comp. Zool., Harvard, i., No. 3, pp. 29-40. xlviii. 1865. A'eerill, A. E., Synopsis of Polyps and Corals of N. Pacific Expedition, Part ii., Alcyonaria. Proc. Essex Instit., vol. iv., pp. 148-196. Plates v. and vi. xlix. 1897. Whitelegge, Th., The Alcyonaria of Funafuti, Part ii. Mem. Australian Mus., iii., pt. 5, pp. 307-320. 2 Plates. 1. 1889. Wright, E. P., and Studer, Th., Eeport on the Scientific Eesults of the Voyage of H.M.S. " Challenger." Zoology, vol. xxxi., Alcyonaria. DESCRIPTION OF PLATES. Fig. 1. Polyp of Scirpearia peetinafa enlarged ( X 25) to show structure. 2. Portion enlarged of Monaco specimen to illustrate the motility of the polyps. 3. Cross-section of a Juncellid axis to show the concentric laminae and the " ridges and depressions." 4. a-(/. " Clubs " of JuiiceUa, (e-g) showing characteristic variations. 5. a and b. Two kinds of double-clubs. 6. Variation forms of double-clubs. 7. Double-wheels or capstans. 8. Transition from the elongated double-club (a), through the double- spindle (b), to the simple spindle (c). 9. Three portions of var. a. of Juncella juncea to show the disposition and nature of the polyps, (a) near the base of the colony, (b) midway, (c) near the tip. 10. Cross-sections of var. b. of J. juncea to show the internal structure. The levels of (a), (b), and (c) correspond to those of fig. 9. 11. Three views of the superficial appearance of the axis in J. juncea. The portions shown are from the parts of the colony given in fig. 9. Simpson — A Revision nf tlie Gorgonellidae. 135 Fig. 12. Portions of J. juncea var. b. to show tlio distriliution and nature nf the verrucas, (a) near the base, {h) midway (non-polyp-bearing aspect), (c) near the tip. 13. ./. juncea var. b. Cross-sections at the three levels given in fig. 12 to show the internal structure. 14. Spicules of J. juncea. 15. 16, and 17. Three colonies of t/MwccZ/a (7«m»i«ceS'. pectinata. 46. Colony of Sciipearia elongata in the Museum of the Royal College of Surgeons, London (from a photograph suppUed by Dr. Burne). 47. Spicules of the Royal College of Surgeons specimen of Scirpearia elo'iigata. 48. Spicules of the British Museum specimen of Scirpearia eloru/ata. 49. Scirpearia fiagellum. Portion of Naples specimen enlarged ( X 6) to show the nature and distribution of the verrucae. 50. Spicules of Scirpearia flagelhim (Naples specimen). 51. Colony (nat. size) of Scirpearia jlagellum (Cape). 52. Portion of colony (fig. 51) to show the nature of the verrucae. 53. Longitudinal section through the portion of Scirpearia fiagellum shown in fig. 52 to show the internal structure and the attachment of the strong retractor muscles. 54. Spicules of Scirpearia fiagellum (fig. 51 specimen). Simpson—^ Revision of the Gorgonellidae. 137 Fig. 55. Youug colony of Scirpearia Jktgdliun (nat. size) (Cape). 56. Portion of colony (fig. 55) enlarged (xl2) to show the nature and distribution of the polyps. 57. Spicules of Scirpearia fiagdlum (fig. 55). 58. Portion of a Monaco specimen to show the distribution of the verrucae (X4). 59. Same as 58) „„ „ _„>• different specimens. 60. Same as 08) 61. Scirpearia thomsoni u. sp. Silhouette of axis to show the natui-e of the branching. 62. Scirpearia thomsoni n. sp. Portion enlai-ged ( X 6) to show the disposition and nature of the verrucae. ■63. Spicules of Sciipearia thomsoni n. sp. 64. Scirpearia alba. Two portions enlarged ( X 5) to show the nature and distribution of the ^-eirucae at different levels, («) near tip, {b) near the base. 65. Spicules of Scirpearia alba. 66. Scirpearia aurantiaca. Portion enlarged ( X 5) near the middle of the colony to show the nature of the verrucae. 67. Scirpearia aurantiaca. Portion enlarged ( X 5) near the tip of the colony to show the nature of the verrucae. 68. Spicules of Sciipearia aurantiaca. 69. Scirpearia furcata. Two views of the same portion of the t}-pe specimen of Scirpearia sp. (?) enlarged (x5) to show the nature and distri- bution of the verrucae. 70. Scirpearia furcata. Part of type specimen of Sdrpearella sp. B. 71. Scirpearia furcata. Part of type specimen of Juncella tlongata (Hickson). 72. Scirpearia furcata. 'Spicules of type specimen of Juncella clongata (Hickson). 73. Scirpearia furcata. Two views of the same part of a colony from the Indian Collection ( x 5) to show the nature and distribution of the verrucae 74. r(il}ii of Scirpearia fircata. '75. Spicules of Indian Collection specimen of Scirpearia furcata. 13S Proceedings of the Royal Irish Academy. Fig. 76. Silhouette of axis of "Providence" specimen of Scvtpearia furcata (t n- s.). 77. Scirpearia furcata. Small portion of main stem of " Pro\idence " specimen to show the verrucae. 78. Scirpearia furcata. Two \'iews near the base of the second primary branch of the " Pro\idence " specimen. 79. Scirpearia furcata. Two views midway on the third primary branch of the " Providence " specimen. 80. Scirpea.ria furcata. Two views near the tip of the third primary branch of the " P^o^"idence " specimen. 81. Spicules of the " Providence " specimen of Scirpearia furcata. 82. Sciipearia furcata. Complete colony (nat. size) of a young specimen in the Cape Collection. 83. Scirpearia furcata. Portion enlarged I'x 8) of a Cape specimen to show the low nature of the verrucae. 84. Scirpearia furcata. Portion enlarged (x4) of a Cape specimen to show the nature of the verrucae. 85. Scirpearia furcata. Portion enlarged f X 5) of a Cape specimen to show the nature of the verrucae. 86. Scirpearia furcata. Portion enlarged (x5) of a Cape specimen to show the distribution and uatui'e of the verrucae. 87. Spicules of a Cape specimen of Scirpearia furcata. 88. Sciipearia furcata. Two views of a portion near the middle of a colony in the Indian Collection to show the distribution of the verrucae, («) polyp-bearing aspect, (6) non-polyp-bearing aspect. 89. Spicules of ludiau Collection specimen (Fig. 88) of Scirpearia furcata. 90. Three views from a specimen of Scirpearia furcata in the Indian Collec- tion to show the distribution and nature of the verrucae at different levels, («) near base, [b] midway, (c) tip. 91. Spicules of Scirpearia furcata. (Specimen fig. 90.) 92. Scirpearia furcata, var. rohista. Colony (nat. size) to show the general habit and the distribution of the verrucae. 93. Scirpearia furcata, var. rohista. Portion enlarged ( X 5) near the base to show the nature of the verrucae. 94. Spicules of Sciipearia furcata, var. robusta. (Andamans specimen.) Simpson — A Revision of the Gorgonellidae. 139 Fig. 05. Scirpearia furcata, var. rohusta. Three portions enlarged (x 5) to show the proportions of the different parts and also the nature and distribution of the verrucae, («) near base, (h) midway, (c) near tip. 96. Spicules of Smyearia furcata, var. rohxista. (Mergui specimen.) 97. Scirpearia andnmaneTisis, u. sp. Colony (nat. size) to show the mode of branching and the general habit. 98. Scirpearia andamancnsis, n. s. Portion near the tip of a branch enlarged ( X 6) to show the nature of the verrucae when shghtly retracted. 99. Scirpearia andamancnsis, n. sp. Portion of a branch enlarged ( x 6) to show the nature of the verrucae when partially retracted. 100. Scirpearia andamancnsis, n. sp. Portion near the base enlarged ( X 6) to show the nature of the fully retracted verrucae. 101. Spicules of Scirpearia andamanensis, n. sp. IQ'l. Scirpearia ramosa, u. sp. Colony (nat. size) to show the mode of branching and the general habit. lOo. Scirpearia ramosa, n. sp. Portion enlarged (xG) to show the nature of the verrucae. 104. Spicules of Scirpearia ramosa, u. sp. 105. Scirpearia ceylonensis, n. sp. Colony one-half nat. size to show the mode of branching and the general habit. 106. Scirpearia, ceylonensis, n. sp. Portion enlarged (x5) to show the disposition and nature of the verrucae. 107. Spicules of Scirpearia ceylonensis, n. sp. 108. "Challenger" specimen of Scirpearia maculata from Banda. (From a photograph supplied by Prof. Bell.) 109. Spicules of the " Challenger " specimen of Scirpearia mMtdata. 110. Fragment of Juncella gemmacea, originally described as EUisclla nuicidata. 111. Transverse section through Scirpearia quadrilineala,\\. sp., to show tin- structure of the coenenchyma and the position of the four main canals. 1 12. Two portions of Scirpearia quadrilincata, n. sp., slightly enlarged (X 1 J) to show the distribution and nature of the verrucae at diflerciit levels, (a) near the tip, (J) near the base. 140 Proceedings of the -Royal Irish Academy. 113. Spicules of Scirpearia guadrilinea.ta, n. sp. 114. Spicules of Nicella dichotoma. 115. ST^icnies oi Mcella flabellafa. 116. Spicules of Nicella. reticulata. 117. Three portions of Nicella moniliforme, enlarged (x5) to show the difference in the distribution and nature of the verrucae at the various levels, (a) near the base, (J) middle of the colony, (c) near the tip. 118. Spicules of Nicella moniliforme. rroc. R. I. Acad., Vol. XXVIII., Sect. I!. Plate I. ^}iM^. Fiff. 5. Fig. 2. l-i»f. 7. ,yAA ;/.--^V r^^^ //'A\ f>?^^'^^ i'''::^t i:'-^^ /A Flu. 1. .8 L,V) L^ ^y 'm "KV s^ .'.A 0^ Fiff. O. Fie. M. Simpson — Gorgonellidae. Proc. R. I. Acad., Vol. XXVIIL, Sect. 15. Plate 11. -— • -7" '■^•-'ii ;-;-^Vi:A'.^ #;:*r■■'i^■'^'■■^^"■••••■.iV^^-l vmmmm mil0 b Her. » Fig. 10. b. ~ a Fig. i:«. Simpson— Oorgouellidae. Proc. R. I. Acad., Vol. XXVIII., Sect. B. Plate 111. Mk &■ Fig II. Kig. 1.5. Simpson— Gorgonelliilae. V\K. lO. Proc. R. I. Acad., Vol. XXVIII., Sect. B. Plate IV. Kis i; Fiif 20 FIs. 1». '1!' fe---'i''^v, . f ■ Ct^.;i'--./..'-fe;,'.» Fi,r. 18. S1.MPS0N — Gorgoiiellidae. Proc. E. I. Acad., Vol. XXVIII., Sect. B. Plate V, M\. .Wis ■-*• -T^ 1^ ^^' a b. If Ftx --1 rlc -O. vv"- H-A IIP fe^M ^^ Proc. R. I. Acad., Vol. XXVIII., Sect. B. Plate VI. Fig. 31. Fisf. 29- Vtg. .11. lis. :i:i. i-i(t •■»■•* Simpson — GorgouelliJae. Proc. E. I. Acad., Vol. XXVIII., Sect. B. Plate VII. fig 36. ••is .1?. 1-iff. 38 I iff .lO Vlg. 40 FIgr. .11 SiMPSox — Gorgonellidae. Proc. E. I. Acad., Vol. XXVIII., Sect. B. Plate VI 11. i: its i;:s.' <\. 'UP Q>°0 KIk i:» Ch A <^ FlK I* Simpson— Goigonellidae. Proc. R. I. Acad., Vol. XXYIIL, Sect. B. Plate IX. Fie. *9 Simpson — (iorgonellidae. rig -lO Proc. R. I. Acad.. Vol. XXYIIL, Sect. B. Plate K. Fiff. 51 I- is- 52 Fiff. .^.1 Fig. .5.* Figr. 54 V\s- iB Simpson — Gorgonellidae. Pi-oc. R. I. Acad., Vol. XXVIII, , Sect. B. Plate XL ft' ■ ..I ■■A Y Fiff. U3 Fi(j. Fig. 68 Fig. 69 I: -I IS; J Fig. ;o .•V ^?^ I'iff ?> ■-^,1^- ^'iV^'V e/ v^ if ^WV -.fy ■;# M 1 ^ -i^^. i:«^ te l¥ Fiff. ?2 Fiff. J.'l m- a ^^s FlR. 71 kBf^ iVfl P^C^& iL< u^ ^^ ^^nB Fi|f. J.i Simpson— Gorgonellidae. Proc. R. I. Acad., Vol. XXVIII., Sect. B. Plate XIII. fe ^a) I *"iS S6 Mm Mm (h) TO ■ I 1 Ki"; s:i i i ■ -*T- ' •;■' *. u ; M :>> 1 1 s^- f -lit:?:-!! ?'.i---.'5 .if. -rjaS-S'! v.: 't « ^0M i^e. ^^jT' ec' J i- (^) Fis It (a)V..v • v^ V '!( (tr Fiff. HO FlK HI Simpson — Gorgonellidae. . '.■■Sri^.N,..-- ■-■ Proc. R. I. Acad., Vol. XXVIII., Sect. B. riate XIV. mm v^ '% Fi:;. OI Simpson— Goigoiiellidae. • .1 V. ^ ^•i. r I I ! a "..^ ■^ ^■^f 1-IK !M» Proc. Pi. T. AL-ad., Vol. XXVIII., Sect. P.. Plate XV, Fig. OH b Fig. »l Fig. OO SiMi'SON — GorgDiiellidiie. a. Fijt »•» Proc. B. I. AcacL, Vol. XXVIII., Sect. B. Plate XVI. Hg. 9; jii; ■ Simpson — Gorgonellidae. Ftg. 10:1 Proc. R. I. Aead., Vol. XXVIIL. Sect. B. PlatL- XVII. ^i^ Fig. 109 Fig. I08 Simpson— Gorgonellidae. Proe. R. I. Acad., Vol. XXVIII., Sect. B. TLATK XVITI. a Kijf. Ill Fi^. Iiu ,.,. i ?\ i.'.' . .'•'1 ■•'i^. II I Kig. Il.> SiMi'soN — Gorgonellidae. FIk 11.1 Proc. E. I. Acad., Vol. XXVIII., Sect. B. Plate XIX. Fig. lie .^?«^r^^=^''^'T^^^'^- ^ - ^ iH^j^.^^Sj^^i^ CI Simpson — GorgonelliJae. ALCYONARIANS OF THE "THETIS" EXPEDITION. By PROFESSOK J. ARTHUR THOMSON, M.A., ANIJ MISS DORIS L. MACKINNON, B..Sc. Curiieyie Scholar, Utdversity of Aberdeen. ALCYONAKIA. By J, Arthur Thomkon and Doius L. MAfKiNNON. (Plates Ixi-lxxjcii.) The collection of Alcyoiiarians niude by the " Tlielis " includes thirty-five species, of which thiitcen are new, — alargepioportion. When the specimens are arranj^ed in systematic order, it is seen at a glance that tlie collection has a very detinile character, namely, the proportionately lar<;e numher of Isidie {e.g., six species of Mopsea, four new) and of Prinmoidre (e.g., five new species of Plumnrella). These Isid and Prininoiil colonies are extraordinarily beautiful, and their supeificial resemblance to laTge fronds of Sertularians is striking. It is a matter for grati- fication that numerous specimens of some of tlie new species have been preserved, so that it was possible to make sure that one was not dealing with iiulividual variations. Another feature of the collection is the frequent occurrence of encrusting colonies of Alci/onium (Erythropoflium) ineinbra- iiaceuin and A. replaiis described b}' Kiikenthal, which grow over Gorgonid axes in a very misleading fashion. It is not too much to say that an unprejudiced observer, wlio had not seen tlie real state of affairs exposed in unmistakeable cases, would de.scribe these Gorgonid axes encrusted with Enjlhropodiuvi as peculiar Axifera. LIST OF SPECIES ^ Order I. STOLONIFERA, Hickson. Family CORNULARIIDiE. Clavulariajlava, Hickson. Order II. ALCYONACEA, VerrilhK'ro. p»rte). Family A LC YO N I D jE. Alcyonium (Erythropodiuia) membra naceum, Kiikenthal. ,, „ reptans, Kiikenthal. * „ etheridgei, sp. uov. ^ Those that aru new are marked with an aateriik. 1^4 " THETIS " SCIKNTIFIC RESULTS. Family NEPHTHYlDiE. Lithophyinm flabeUuni (Q.G.). * Dendromphthya ivailei, sp. nov. Order III. PSEUDAXONIA, G. von Koch. Family MELITODID^. Mopsella daviyera, Ridley. „ lexlifortnis (Lauiarck). Psilacabnria yracUlinui, Ridley. Parisis amtralis, Wright and Studer Order IV. AXIFERA, G. von Koch. Family ISID^. Mopsea dichofoma (Linne). „ encrinula (Lamarck). * ,, auslrali^, sp. nov. * „ Jlabelhim, sp. no%-. '' ,. eUgaiis, sp. nov. * „ u-hUd-'.ggti, sp. nov. Acantkoi-lsH'ihdhnn, Wriglit and Studer. Family PRIMNOIDiE. Siachyodes studeri, Vers!uys(,?. regulari,, Wrigl.t and Studer). ^Aniphilaphis plumacea, sp. nov. ^ liumardla Uvvis, sp. nov. * „ ihelis, sp. nov. * ,, corrnscans, sp. nov. * „ tilicoiiles, sp. nov. ^ » ,, rerslnysi, sp. nov. Primno^lla auslralasirn, Gray. ,, flagellnm, Studer. „ . distans, Studer. *Caligorgia laevis, sp. nov. Family GORGOXID.E. Lf.plogorgia, sp. (?) Family GORGONELLID^ Ctenocella pectinata (Pallas). ALCYONARIA— J. A. THOMSON' AN'D DOFUS L. MACKIXXOV Uo Order V. STELECHOTOKEA, Bourne. Seel ion A 8 I P H O N A C K A . J-'nmi/;/ TKLEHTiDAi. Telexto trichostemnia (Dana). nrhorea, Wright and Stu.ler. Sf.cfion P E N N A T Q L A C K A. Frtmihj kophobe(j:\ixoxid^. Kophobdemnon schmeUzii{^oi\ikvi), = Sclerobe'emnon schmellzii, Kolliknr. Family FTEROEIDID^ Godeffroyia elegans, Ktilliker. Sarcophyllum aiistrale, Kolliker. Of the thirtj'-fivR species in the collection, the following thirteen have been ])reviously recorded from Australian seas ; — Clavularia Jlava, Hickson. Mopsella chivigera, Ridli^y. „ textiformi.t (Lamarck). Psilacabaria graciVima, RiiUej'. Parisis (fiistrah's, Wright and Sluder. Mppsea dichotoma (I.inne). ., encrinT-la (Limarck). Acanthoifis ^ffiihnlhtm, Wright ami Studer. Primnoella Jlagefhi,m, Stud er. Primnoella anstralasia, Gray. Ctenocella pectinala, Pallas. Tehsto trichogtemma (Dana). Sarcophyllum austral", Kolliker. The most striking of the rectrils of flisiriliution are the following: — Primnoella JlngeUnm, Studfir, prc\ionsly reported hy Hickson from Australijin seas, was originally found in ilie .Smith Atlantic, off Patagonia ; Primnoella ilialnns, .Stiiih-r, was previously found off the West Indies and ISrazil ; lip)olyps are spindles and clul>s of slender form, with a few projecting warts. Tiieir dimensions in mm. are 306 x 034; -27 x 068 ; 204 x 017 ; 17 x 017. The colour of the colonies is greyish-brown to dark-brown ; the polyps are a darker shade of the same colour. Tn many respects this species comes very near to llickson's Alcyoniwn pjirpiireunt'K sHickson— The Alcyonaria of the Cape of Good Hope, part ii.— Marine Investigations in Sonth Africa, iii., 190-1, pp. 215-217, pi. vii., fig. l.pl.ix., fig. 18. 150 " THETIS SCIENTIFIC RESULTS. Our species differs from Hickson'.s in the absence of the char- acteristic purple colour, the presence of a distinct stalk-portion, the sparser distribution of the polyps, the larger size of the polyp spicules, and the simpler character of the coenenchyiua spicules, tiie double-clubs in Alcyoniutnpurpureumhemg figured with at leHst four whorls of warts. The powdered appearance of the surface of the coeuenchyma in our specimens is exceedingly characteristic. Locality. — Manning Bight, 22 fathoms. Famt/y NEPHTHYIDAE. Genus LITHOPHYTUM, Forskal. LITHOPHYTUM FLABELLUM (Q.G.) Lithophylnm JlabeUum, Kiikenthal, Versuch. einer Revision der Alcyonarien, ii. Die Familie der Nephthyiden, i. Theil. Zool. Jahrb., xix., 1903, p. 111. A single catkin of this species was found among some broken fragments in the pi-esent collection. Locality. — Coast of New South Wales. Previouslj' recorded from Zanzibar, Solomons, Tumbatu, New Ireland, Carteret Harbour, and Egniont. Genus DENDRONEPHTHYA, Kiikenthal. {Spongodes in part). DENDRONEPHTHYA WAITEl, sp. 7iov. (PL Ixii., tig. 4 : PI, Ixv., tig. 2 ; PL ixvii., fig. 3.) Station 25. Several compact, bushy colonies, from 5 to 7 cm. in height, represent this interesting species. The specimens are all intact, and show great uniformity in their mode of growth. From a luxuriantly-rooting basal attachment rises a somewhat flaccid stalk, which just above the roots has a diameter of 10 mm., but swells out to IG mm. higher up. At a height of about 2 cm. the stem is surrounded by foliaceous expan.sions bearing polyps along their edges. Above this, numerous small branches come off, and the stem finally divides into two main branches, which ramify extensively. The stem tends to be exposed on one side in its upper reaches, as there is a much sparser occurrence of branches on one side than on the other. ALCYONARIA — J A. THOMSON AND DOKIS L. MACKINNON. 151 Over the branching side of the colony the polyp-bundles form a dense covering. The polyps occur in bundles of about six, in typically " glomerate" arrangement. The polyp-stalks are very long, l-5-2'5 nnii. The poiyp-hends are al)Out IS min. and make an angle willi tlieir stalk that may be either a right angle or obtuse. The .Stiitzbiinilel is very weaklv developed ; it consists of about four almost smooth spindles. In contrast with the long polyp-stalk the Stiitzbiindel seems quite inconspicuous, and this gives a Slereacanl/iiaAike appear- ance to the polyp. The armature of the polyp consists of eight longitudinal rows of closely apposed pairs of curved spindles: there are about six to eight pairs in a row, of which the uppermost pair is the most strongly developed. One or l)oth of this uppermost pair may project slight!}' above the jiolyp-hoail. Between the adjacent "points" thus formed, a few small spindles lie without regular ariaugeinent. Uidike the spicules of the StUtzljiindel and polyp-stalk, wluch are yeHow-brown smooth spindles with scarcely an}' warts, the spicules of the |)oIyp-head are colourless, bent spindles, with more numerous, rather prominent warts. Tiie following measurements were taken of length and breadth in millimetres: — (1) spicules of polyp-stalk, 1-37 x •08.'); 765 x •035 ; -425 x -017 ; (2) polyp-spicules, -76.5 x -051 ; -544 x -034 ; •34 X -017 ; •255 x 025. There is a median row of very small, transverseij'-disposed spicules along eacli tentacle. Tiie s]iiculesof the cortex are, in the upper part of tlie colony directed more or less transversely. The}- are long spindles (2'97 x ^17 mm. ; 1^78 x -lO mm.; 5 27 x -34 mm.) with few, small warts. Ill the branches these superficial sjiindles are smaller, and take a longitudinal direction. Below the point where the ring of foliaceous branches comes oft, the character of the cortical spiculation changes. There are still a few spindles, but the great majoi ity of the spicules are small opaque, white stars, tri- and quadri-radiates and irregular spiny forms. '204 x '17 mm. ; •119 X -102 ram. ; 102 x -085 mm. In the canal-walls "of the lower part of the colony a few similar star-like spicules are to be found. In the upper readies of the canal-walls no trace of spicules could he found. The colour of the stem is browni>h gray with a reddish tinge on the branches ; the roots are dark gray, and the polyps and their stalks are red-brown. In certain features this species approaches very closely Dendroneplilhyu maxima, Kiikenthal, particidarly in the form of the long-stalked polyps with their weakly-developed StiitzbUiidel, and in the remarkable smoothne.ss of the cortical spindles. But our species shows laiger polyps, a greater number of spicule pairs in the longitudinal rows on the polyps, and these spicules are not smooth as in D. maxima, nor do they show any 152 " THETIS " SCIBNTIFIC RESULTS. tendency to form clubbed ends. Fui thermore, the spicules on the tentacles are not arranged in a double row "en chevron." The spindles of the corte-K are very much larger than in D. mdrima ; the star.s and irregular spiny bodies of tlie hase are similar iu both species, but the canal-walls in our specimens do not show any of the numerous, disc-shaped concretions characteristic of D. maxima. Order III. PSEUDAXOXIA, G.i-ouKoch. Family MELITODIDAE. Genus M O P S E L L A, Gray. MO PS ELL A CLAVIGERA, Ridley. (Plate Ixviii., fis;. 9.') Stations 34, 40, 42 and 48. Mopsella davigera, Ridley, Report ZooL Coll. H.M.S. " Alert," 1884, p. 360, pi. xxxvii., fig. b. pi. xxxvii., figs. a-a'". The largest specimens have lengths of 18 era., SS-o cm., 30 cm., and 53 cm. respectively. The branching is generally dichutonious and is strictly in one plane. The branches arise from the nodes, which are very markedly swollen. In the largest specimens they have a diameter of as much as 18 mm. There is no anastomosis. The polyps are chiefly confined to one surface, over which they are disposed irregularly in slightly projecting verrucse. Into these the polyps are in the most cases retracted. The polj'p armature consists of eight triangular points of three or four couveiging pairs of delicate spindles ; below these are about two rows of similar spicules disposed horizontally. A double row of minute spicules, " en chevron," occurs along the middle line of each tentacle. The spicules of the cceneiichyma are exactly like those described by Ridley, i.e. — (I) orange-coloured fusiform shapes, rather coarsel}' tuberculate, swollen, tapering, with occasional " Hlattkenlen" at one end— -18 x -034 mm.; -15 x 035 mm. (2) "Blattkeulen" of lemon-yellow colour, in shape like small pointed tiowels, with a short handle, and a blade formed of about two lancet shaped lamiiije — -076 x 034 mm. ; 051 x OS? mm ; 042 x 025. With regard to the distribution of the ccencmhyma spicules, it may be remarked that the orange-coloured spindles generally ALCYONARIA — J. A. THOMSON AND U0R1^S 1.. MACKINNON. 1^3 form tlif lower strntuin, over the surface of wliicli llie yellow clubs are disposed in a thiu layer, which may cover the entire colony, or may he connned to a hroad circle round each i">lvp- opening. Locality. — Eleven miles east of Broken Bay, 30-40 fms. Previously recorded frcjm Port Curtis, Queen.sland, .j- 11 fms, Port Molle, 14 lm,s , Thursday Island, Torres Stiaits, 4-6 fms.; Dirk Hartog Island, West Australia (Studer). MOPSELLA TEXTIFORMIS, Lamarck. (I'lrtte Ixiii., tigs. 4 and 5.) Mopsella lexti/orinis, Ridley, Report Zool. Coll. II.M.b. "Alert," 1884, pp. 358-360. The strong basal portion of what must have heen a large colony. It is about 23 cm. in height, and the diameter near the base i.s as much as 3 cui. The length of the inter-nudcs varies from 4 to 9 mm., and of the nodes 4-6 mm. The colour of the axis is rose. At the liase where it is weathered and the longi- tudinal canals are exposed, the axis has a very remaikable labyrinthine pattern. The main stem gives off several strong branches in one plane, and these .seem to have given off numerous secondary branches strikingly slender in contrast. There his been abundant anastomosis'. Most of the crenenchyma has been worn off, but here and there patches ivmain of a bright yellow colour. In legard to spicules the specimen agrees well with Ridley's description, but very few of the characteristic dentate " Hlatt- keulen " have long shafts. In these spicules the shafts are usually orange and the heads lemon yellow, The nodes contain the usual smooth rods. Locality. — Lord Howe Island. Previously recorded from many Australasian .Stations. Genu.s P S I L A (J A I! A K I A, RiUI"!/. PSILACABARIA (JRACILLI .\l A, liid/e,/. Psilacabariagracillima, Ridley, Kcp. Zool. Cull. li.M..S. "Alert," 1884, pp. 361-365, pi. xxxvii., figs, d-d", pi. xxxviii., figs. f-{". Stations 34, 36, 48 A number uf broken pieces seem referable to Psilacaba/'iii graciliima, Ridley. The chief characteristics of this tpecies may 154 "THETIS SCIENTIFIC RESULTS. be summarised : — dichotomous biauching approximately at right angles and in one plane ; absence of anastomosis ; the axis of the internodes hard, white, not striated; low, rounded, tubercular polyps arranged in a rather irie<;ular manner, in spirals, and oppoisite ; chief forms of spicules. — (1) lurge cylinders (-28 to -31 X 087' mm.), fusiform to sub-clavate, with thick axis, slightly tapering to round-pointed ends, provided with strong tubercles with tendency to arrangement in whorls ; (2) smaller spindles (•21 to -25 X -053 to 071 mm.), more or less curved, with sharp ends and roughened tubercles. The colour of the specimens is yellowish-brown. Locality. — South Coast of New South Wales. Previously recorded from Port Molle, Queensland, 12-30 fathoms ; Port Darwin, 8-12 fathoms ; East Australia, 42 fathoms. Genus P A R I S I S, Verrill. • PARISIS AUSTRALIS, Wright and Studer. (Plate Ixx.) Parisis austraUs, Wright and Studer, Chall. Rep., Zool., xxxi., 1889, pp. 183-184, pi. xli., fig. 5. Stations 13, 34, 44, 47, 48. This species was founded on two fragments, much overgrown by an encrusting sponge. In the present collection there are numerous specimens, many of tliem almost entire, so that we are able to amplify the original description. The largest colony is 35 cm. high, with a spread of 18 cm. From a slightly encrusting calcareous base arises the cylindrical stem, with an average diameter of 6 mm. which soon begins to give off branches. These arise on the sides, alternately and in one plane. Higher up all distinction between main stem and branches is lost. The upper part of the colo^iy consists of a close tangle of equal sized branches, 3-3-5 mm. in diameter, which divide dichto- mously or give off short branches quite irregularly, but invari- ably in one plane. The tendency is for all the branches to bend upwards at the tip, and the axillary angle is 45"-60°. Though the branches often overlap one another, or even entwine slightly, there is never any anastomosis. The surface of all the colonies is more or less encrusted with a siliceous sponge, which often entirely obscures the underlying stcuctuie, and produces by its numerous projecting spicules a curious brown, velvety surface. Beneath this is the hard pave- ment-like surface-layer of the Parisis, consisting of cream-white coenenchyma spicules. XlCTONARIA — J. A. THOMSON AND DORIS L. MACKINNON. 155 In tlie lower part of the stem tlie calcareous and Iiornj' joints have approximately equal lengths of about 2 5 mm. ; hut higher up the internodes greatly ])redominate, 3-4 mm., while the nodes are reduced to mere constrictions, -5 mm. in length. The intpr- nodes are white, and bear .somewhat distant longitudinal grooves. On the younger colonies and smaller branches the arrangement of the polyps is bilateral, but on the larger branches they may occur all round. In one specimen where they were less obscured by sponge-growth than elsewhere, the jiolyps had a height of 75 mm. with a diameter of 1 mm. : they projected from the branch as low, rounded warts. The spicules are exactly like those described by Wright and Studer — i.e., massive warty spindles, some almost as broad as long, and approaching splieres. The warts are high and prom inent and finely sculptured. The following measunniients were taken of length and breadth in millimetres. — -261 x 17 ; -25 x •1 ; -2 X 16; 17 x -1. This species is distinguished fiom Parisis /'riUicosa, Verrill — (1) by tlie_branches coining oil' at angles of 45^ 60", instead of at approximately right angles ; (2) by the polyps occurring all round the branches instead of being strictly bilateral ; (3) by the smaller size, and greater slenderness of the spicules. Locality. — South Coast, New South Wales. Previously recorded from Station 163 B, off Port Jackson, 35 fathoms. Order IV. AXIY^IX, G. von Koch. Family 1 SI DAE. Genus M O P S E A, Lamouroux. MOPSEA DICHOTOMA, FAnnl (Plate Ixvii. tig. 1.) Morpsea dichotoma, Wright and Studer, Cliall, Hep., Zool., xxxi., 1889, pp. 41-42, pi. ix., Hg. 10. Stations 48, 47. 53. The largest specimen is an almost complete lyre-shaped colony, rising from a slightly encrusting calcareous base to a iieighl of 22-5 cm. The main stem, 3 mm. in diameter near the base, divides to form two equal branclii s at a height of 2-5 cm. These two main branches give rise along one side to a number of 156 " THETIS " SCIENTIFIC RESULTS. secondary branches which ran parallel to one another. As the.>e secondary branches are nearly as thick as the main branch from which they spring, the eftect of a repeateii dichotomy is produced, an effect that is heightened by the tendency of the main branch to bend outwards after each brancli is given off, so that its course describes a, series of shallow curves. The secondary bianclies rise straight upwards and may remain unbranched throughout their length, or may divide dichotoniously- Branching is strictly in one plane. The polyps are arranged in close spirals all over the larger colonies, though one young specimen shows the polyps in a single alternating row on each side of the branch. The polyps are club-shaped, with truncated mouths ; they are duected upwards, and bent towards the stem. lu the largest specimen the upper parts of the polyps Imve all been rubbed off, which produces a dece(itive appearance of very short, truncate polyps with their mouths directed outwards from the stem. There is a considerable variety in the spicules (PI. l.xvii. fig. 1). (1) Tiiere are somewiiat flattened, curved spindles, produced on the conve.x side into a number of sharp, prominent teeth. The following measurements were taken of lengtli and breadth in millimetres. --238 x -102 ; -187 x 085 ; 1 18 x -068 ; -102 x 051. (2) .Spicules of similar form to (1), but simpler and with only a few small warts.— 153 x 051 ; 136 x -053 ; 102 x 034 mm. (3) Small " Blattkeulen " with a very small shaft, bearing sometimes a few warts, and supporting a group of sharp, blade- like projections.— -102 x -085 ; -09 x -068 ; -051 x -032 mm. (4) Small irregular bodies and " capstans." — 085 x "085 ; -102 x 0-85 ; 0-68 x Uol mm. The colour of the colonies is pale-brown to cream. Previously recorded from the Indian Ocean, and Port Jackson, New South Wales, 35 fathoms. MOPSEA ENCRINULA, Lamarck. Mopsea eiicrinula, Wriglitnnd Studer, Chall. Bep., Zool., xxxi., 1889, pp. 43-44, pi. vii., figs 1, 1% l^, pi. ix., fig 11. Stations 34, 44, 47. A few incomplete specimens in the present collection agree closely >vith the description of Mopsea eiiorinula (Lamarck), given liy Wiight and Studer. Brauching is pbmie like in one plane ; the club-shaped polyps, 1 mm. long, cover the brarches aid twigs in a close spiral, and are incurved towards the stem. Many of the spicules are yellow and white flattened s|)indles, approaching the " lancet-shaped plates " of Wright and Studer's description. ALCyONAUIA— ,T. A. THOMSOX AM) DORIS I, MAOKINVOX. 157 These bear nuniBrous high, projeetina; warts, wliicli tend to be raoie numerous towards one siiie of tlie spicuh' tlian the other, especially if the spicule is curved, wlien tiie oon\ex side is always much the stronger toothed ; length !j3- itreadtli in mm. 187 x Ool ; •15 X 06 ; 12 X 05. There are also scales with very iriegular margins and numerous spiny warts. — 102 x -082 mm. ; 063 x •0+ mm.; 02 x 02 mm. tiiirdly, there are small irregular bodies.— 025 x -025 ; 051 x 025. Tlio colour of the specimens [s orange-hrown : the polyps yellowish-white. Localihj. — Eleven miles east of Broken Hay. Pre\ioiisly recorded from Australia (North-west Coast), 50 fathoms; Station 162, off Kast Moncoeur Island, Bass Strait, 3S fathoms. MOPSEA AUSTRALTS, sp. nw>. (Plate Ixiv. tigs. 1 and 2., pi. Ixvii. tig. 5.) One specimen has the basal portion almost intact, tne othets are brOKen, branching pieces of various lejigths. All are in the dried condition. The mode of branching, ;ind, in fact, the whole general appearance of the colony is very similar to Mopsea dichotoma ( I,inn6). The pol3'ps are arranged in close spirals round the branches. They are small (S-TS mm. in length), club-shaped, and, in the dry state at any rate, are verj' closely |iressed to the suiface of the branch, .so that their mouths are hidden. Their armature consists of indistinct longitudinal rows of transversely arranged, slightly overlapping spicules ; there are from fifteen (o eighteen of these in the abaxial rows. The spicules of this species are i-niallfcr than those of M . ilicfio- tonia, ami much less spiny. The following types occur — (1) rather bioad, flattened, almost scale-like spicules witli relatively prominent teetli round their edges, and with a few warts over the surface. These spicules vary in shape from Hatlened spindles lo rough oc-shapes and ovals. The following nu asuremenis were taken of length and breadth in millimetres. — 187 x 034; 153 X 068 ; 119 X 051 ; 102 x 068. (2) Stout spindles and some clubs with prominent wart->. — •15:^ x '034 mm. ; I 19 x 017 mm. ; •102 X 017 mm. (3) Capstans and small, irregular bo'lies.— •051 X 034 mm. : 034 x 034 mm. ; -068 x 017 mm. ; O.'M x •017 mm. The colour of tiie dried specimens is deep cream. It must be admitted that Mapitea dicholoma (Liime). M. elongata, Roide, and the species at present umler discussion, are 168 " THETIS " SCIENTIFIC KESULTS. vei'3' closely related. The only noteworthy difteieuce, as it seems to us, lies in tlie character uf the spicules. Eoule lays emphasis on the mode of branchiug, which he finds to be more sparse in M. elouffatu than in M. dichotoina, and with a stiong tendency to the formation of long, simple branches. Our new species, M. austialis, agrees with Roule's description of the branching of J/, elongala, but it also agrees with the undoubted specimens of M. dichotoma in the present collection — so closely, indeed, that it was at first mistaken for that species. We do not think that the mode of branching can here be safely used as a specific distinction. Perhaps the same is true in regard to the details of spicule-forni. Locality. — Eleven miles east of Broken Bay. MOPSEA FLABELLUM, sp. nov. (Plate Ixiii. figs. 1-3 ; pi. Ixvii. tig. 6 ; pi. Ixxi.) Stations 34, 44. This species is based on one complete colony and a number of pieces. The complete specimen rises from a slightly encrusting cal- careous base to a height of 245 cm. Branching begins at a ht-ight of 5 5 cm. an I is very luxuiiaut ; the branches are con- fined almost exclusively to one plane, and there is a strong tendency to dichotomy ; they are slender throughout, and do not tapei- much ; the stouter branches have a diameter 2-5 mm., and the twigs of almost 2 mm., near their tips. Near the base the diameter of the axis is 4 mm. Here the ccBnenchyma has been worn ofl, and the bulk of the stem is made up of the horny joints which are about 2 mm. long, the calcareous joints being reduced to •75 mm. ; in some cases they are quite overlapped by the horny joints. Higher up the cal- careous joints have lengths of 3-4 mm., and the horny joints -25- ■5 mm. The colour of the horny joints is brown ; that of the calcareous joints varies from cream-white near the base of the colony to orange in the twigs. The calcareous joints bear marked longitudinal tlutings. The branches arise from the calcareous joints. The polyps occur in close-wound spirals all over the stem and branches In the j'oungest twigs the spir.tl is wider but nowhere is there any trace of a bilateral arrangement. The polyps are club-shaped, with .somewhat truncate mouths which are pressed against the cortex of the branch. The average length of a polyp is 1 mm. The calyces are armoured with about eight rather indefinite longitudinal rows of overlapping spicules, ALCYONARIA— J. A. TIIUMSON AND DORIS L. MACKINNON. 159 fifteen to eighteen in a row. These spicuh-s are flat, yellow scales circular, Sshaped and irreguhir ; they are smootli, oi' bear a few simple warts, and their margins are deeply dentate or serrate. Their dimensions in millimetres are, '255 .v -136 ; ••2*i4 x -102 ; -17 x ■068 ; -153 x 1.53. A low eight-rayed operculum is formed by similar scales. The spicules of the ccenenchyma are yellow spindles and cylinders (-187 x 035; .17 x -017; -136 x 051 mm.). They have relatively few, but large warts. There are also small irregular bodies (085 x -051 ; -051 x -034 ; 008 x 05 mm ). The genera! colonr ofthec lonyis orange-brown, the polyps are rather lighter. MOPS E A ELEGANS, sp. nov. (Plate Ixiv. figs. 3 and 4 ; pi. Ixviii. tig. 5; pi. Ixxii.) Stations 34, 41, 42, 4(, 48. " Several beautiful golden brown colonies, with rich dichulo- mous, almost parallel, brandling. The largest is 34 cm. in height, with a spread of about 5 cm. ; the branches Inive an average diameter of 2 mm. The axis shows the usual white calcareous internodes with tiue longitudinal Ihitiiig and short amber-coloured noiles. There is some variety in tlie origin of the branches. In ceitain cases the dichotomy is precisely at the node ; in other cases the calcareous internode foiins immediately below the node a slight shelf from which a new branch arises with a horny node ns the tirst joint. The cccnenchyma is very thin, and, wiili the excep- tion of a median line on each surface, is in great part hidden by the numerous polyps. These occur in alternating rows on each side, here and there encroaching on the free median spaces. Most of the branches show two rows on each side in their upper reaches, but in the lower parts of the colony three or four rows are often seen. The polyps are somewhat club- shaped, 0 75-1 mm. in height by about 0-5 mm. in maximum breadth. They project at an acute angle to llu; axis, but liie upper parts are incurved. In the upper reaches there are about twenty on each side in a centimetre. The superficial spicules are Hat, often oval scales, with tlie following dimensions in millimetres. — 073 x '060 ; 052 x 052 ; 0 38 X 0 55. Each scale shows a nucleus, often excentric, with fine ridgfs radiating from it. Om- maigin «( the scale is entire, the other bears more or less deep indentations, and the part of the scale to this side of the nucleus is studded with enmll warts. The whole might be compared to a ctenoid fish s^ale. ai,d they overlap one another, the overlappi d part being the smooth portion. ALOYONAKIA ".r. A. THOMSON ANII DOKls i. MACKINNON. ll'il of the stem is missing. The stem has a maximum diajiielei of -1 mm. ; the average diHiueter of llie larger brandies is 2 mm , and of the twigs, 1 mm. Anotlicr specimen, al>o imomjilete, lias a height of 16 cm., witli a span of 11 cm. The ioweupart of llio stem is overgrown Ijy a sponge. Polyps occur here and there on the branches, but tiiey are mainly coiihned to the twigs, along each siile of which they are arranged alternately in a single row. Frequently this arrange- ment becomes irregular, the polyps encroach on the middle line, or a double row may he formed along each side. Tlier<- arc ahoul fifteen polyps to 1 cm. in each row. The polyps are small, -S-'To mm. in height, chibslia|)t'd, trun- cate, and turned upwards towards the tip of the twig. They are covered with closely-fii ting, elongated spicules arranged transversely, which cither interlock by means of their dentate margins, or overlap one another ; the abaxial rows are composed of about sixteen such spicukis arranged in an imbricate manner. Similar spicules form a low, eight-rayed operculum. The Sjiicules of tlie ctenencliyma are like those of the polyps, but follow the loTigituiliiial diiection of the stalk and form a sort of pavement over its suiface, adjacent spicules being clcselj' interlocked by their teeth. The calcareous internodes of the axis bear fine longitudinal ridges, which are dentate at the upper and lower ends of the joint. The spicules are small, colourless, somewhat flattened spindles and lancet-shaped plates, fieqnently curved, and hearing numerous sharp, tooth-like warts, which are usually more strongly di'velopedon oneside of tliespicule than the other. The following measurements wei'e taken of length and breadth in millimetres:— -238 x 085; -17 x 008; ■]o6 x 051. There are also numerous smaller spindles anil forms approach- ing clubs:— 153 X V2r, ■ ll-J x 035 ; -085 x 031 ; -008 x 017 mm. The colour of the colonies is pale-brown to cream-white, the axis occasionally tinged with pink. Locality. — Eleven miles east of IJroken Hay. Genus A C A N T II I ) I S I S, Wriyhl and Slttder. ACANTHOISIS FLAHELLUM, Wriyht and Sltider. (Plate Ixii., Hgs. 1 and 2.) Acanthoisis flabelluin, Wright and .Studer, Chall. Ue|)., Zool., xxxi., 1889, pp. 45 46, pi. viii., figs. 1, la., lb., pi. ix., fig. 12. Stations 22, 47 and 53. Some very fine fan shaped specimens of an orange-brown colour agree with Wright and Studer's description of Acanthoiais 162 "THETIS SCIENTIFIC RESULTS. flahellum, lliough there appears to be a stronger tendency to anastomosis than in the Challenger specimens. The height of the largest colony is 24 cm., with a width of 16 cm. across the expanded portion. A small, broken piece of a colony is of a brown colour with the polyps tending to encroach on the middle line of the branch in- stead of being s'rictly bilateral iu arrangement. The spicules here are colourless, and rather smaller than in the orange specimens. Previously recorded from Port Jackson, 30-35 fathoms. Fannly PRIMNOID^. Genus STACHYODES, Wright and Studer. STACHYODES STUDERI, Versluys. Stachyodes regularis, Wright and Studer, Uhall. Rep., Zool., xxxi. 1889, p. 55, pi. xi., figs. 2, 2» ; pi. xx, fig. 3. Stachyodes studeri, Versluys, Die tjorgoniden der Siboga Ex- pedition, ii. Die Primnoidse, 1906, pp. 94-96, figs. 112-117. Stations 15, 42, and 44. Tlirfe incomplete specimens 11 cm., 23 cm. and 38 cm. in length respectively. On the most slender specimen the polyps occur in whorls of eight to nine ; on the largest there are as many as ten to eleven in a whorl. Previously recorded from Kermadec Islands, 600 fathoms ; Celebes Sea (Siboga), 1080 and 11 6.5- 1264 M. Genus A M P H I L A P H I S, Wright and Studer. AM PHIL APHIS PLUM ACE A, sp. nov. (Plate Ixv., fig. 3 ; pi. Ixviii., tig. 3 ; pi. Ixxiv.) Stations 22, 40, 44. This delicate and graceful form bears a certain resemblance to an uncurled ostrich plume. Branching is approximately in one plane, and the brandies and twigs show a strong tendency to sweep togetlier in long, drooping curves. Occasionally the branches come off like the barbs along tlie shaft of a feather, but more generally the branching is dichotomons, or quite irregular. ALCYONARIA— J. A. THOMSON AND DORIS L. MACKINNON. 103 Three of the four specimens are practically intact. The largest has a height of 18-5 cm. witli a diameter near the hase of .3 mm., the corresponding dimensions of the smallest are 8-5 cm. and 2 mm. The coenohyma is very thin, and allows the dark, bronze-liko axis to sliine throngh. The pol}-ps occur very rarely in two lateral rows; generally they are anaiigfd in a spiral, which becomes closei- in the uppir part of the colony. The polyps are 1-1-5 mm. in length. They stand out markedly from the branch at an angle of 45"-60". They are armonied witli eight close-set longitudinal rows of overlapping " ctenoid " scales : there iire from eight to twelve in a row on the abaxial side, and a rather smaller number on the adaxial. The uppermost scale of each row is more strongly developed than the rest, and its upper edge is somewhat reflexed, so that it stands out from the operculum. In this way a sort of collar or circum-operculum is formed. Above this eight triangidar scales form a well-defined, conical operculum. The dimensions of the " ctenoid " scales in millimetres are •316 X I8I7 -255 X -272 ; -204 x -loS. Their free edge is entire or crisply waved . around tl:e well-marked nucleus numerou.s warts are grouped, and the clear border rouml the exposed portion of the scale hears strongly-marked radiating ridges. The opercular scales are isosceles triangles with a strong con- cavity to the outer surface, and a corresponding ridge internally, which extends for about three-quarters the length of the scale. The surface is elaborately warted, and the narrow clear margin bears ridges. The following measurements were taken of length and breadth in millimetres. — -415 x 204; 34 x 221. The spicules of the cceiieiichyma are approximately circular scales, with a central nucleus, round which wans are grouped. The border is clear, with at most a few low ridges (-187 x -17 mm. ; -136 X 119 m.m. ; •112x112 mm.). The colour of the specimens is fawn to brown. This species differs from Amiihil aphis regularis, Wright and Studer, in the following points:— (1) its much more slender build, a'ld bushier, more luxuriant branching; (2) the almost invariable arrangement of the polyps in .spiials, and their much denser crowding ; (3) the smaller size of the polyps, and (4) the greater number of scales in the abaxial rows of tiie calyx. From Amphilaphis ahielina , Studer, it is distinguished:— (1) by the close spiral arrHngement of the polyps; and (2) by showing no tendency fur the branches to come oft at right angles from the Stem. Locality. — Eleven miles E. by N. of 13arrenjo6y, 30-40 fathoins. ALCYONARIA — J. A. THOMSON AND DORIS L. MACKINNON. 103 ai-e smaller than those described for P. /if.uua, ami :ire suioolhcr and of more simple outline. The chii-f point of ditterence, liow- ever, lies in the mode of branching, whiuii in F. penaa is close and typically foiither like. Versluys speaks of as many as twenty-three twigs on each side of a hianch in a stretch of 5 cm.; these twigs, moreover, are rn-ver longer tlier on the thicker branches. The usual number of whoilsin a length of 3 cm. is twenty-two to twenty-four ; the lenglli uf a polyp is about 1 mm. Tln' mouths are directeil distally. The only complete longitudinal rows of calyx scales are the abaxial and the al)axial-lateral ; iIkmo are never more than nine overlapping scales in the abaxial rows ; thenuudjerin theubaxiul lateral is more variable, but is usuallj' aliout six. Tlie adaxial and adaxial-lateral rows are very incomplete, usually consisting of two to three scales. The opercular scales are triangular and pointed, the two abaxials being slightly larger than the others. ALCroNARIA— J. A. THOMSON AND DORIS L. MACKINNON 17$- Family GORGONELLID^. Genus CTENOCELLAi", Valenciennes. OTENOCELLA PECTINATA, Pallas. (Plate Ixxxi.) Ctenocella pectinata, Pallas, Elenchus Zooptiytoium, 17G6, p. 179. Ctenocella peclinalo, Ridley, Zool. Coll. H.M.S. "Alert" ISS't p. 348. Ctenocella pectinata, Wright and Studer, Chall. Rep., Zool.. xxxi.. lS89,p Ixvi. , Stations 34, 42, 44, 4V. There are several very perfect colonies of typical lyre-shape ; they are all in the dried condition. The height of the largest specimen is 50 cm., with a width of 42 cm. From a slightly encrusting base rises a cylindrical stem, 7 mm. in diameter, which soon forks. The two branches thus formed diverge at an angle^of 45°, and give oiF, along the upper side only, a series of parallel, ascending twigs. The space between two twigs is about 8 mm. One or two of these lesser branches may be stronger than the rest, and may give off numerous ascending twigs in turn, or fork dichotoniously. More generally, the branche.-; are simple, from 4-12 cm. in length, and with an almost uniform diameter of 2-3 mm. The axis is light brown, and strongly furrowed. The ccen- enchyma is thin, but very compact and smooth ; in many places a distinct median furrow can he made out. The polyps have, on the twigs, an irregular bihiteral nrrHnge- ment, but on the larger branches they occur all over the surface. They are very numerous, about 05 mm. apart, and are all retracted' into very low, wart-like verrucas. The spicules include the following types : — colourless double clubs — 068 X •034 mm.; 051 x 034 mm.; a few crosses — 051 x ■051 mm.; -034 x 0-34 mm.; and more elongated forms approach- ing double-spindles, with scarcelv any "waist" — '085 x •OJSmm.; •06S X 0-25 mm. The colour of the specimens is creara-wliite to ycllowisli. Locality. — Eleven miles east of Broken Bay. Previously recorded from Indian Ocean, seas of the Moluccas, India and China, Cuba, Australia (Warrior Reefs, Torres Strait, 12 fathoms ; off N. W. Cape, W. Australia, 3-4 fathoms). ^o In his " Revision of the Gorgonellidic " (Croc. Roy. Irisli Academy, 1910, p. 319), Mr. .). J. Simpson has shown convincingly that the genus. Ctenocella shoiilii be merged in Srirjiearia. 174 " THETIS " SCIENTIFIC RESULTS. Order Y. STELECHOTOKEA, Bourne. Family TELESTID^. Genus TELESTO, Lamouroux. TELESTO TRICHOSTEMMA, Dana. Teleslo Irichoslemma, Wright and Studer, Cliall. Rep., Zool., xxxi., 18S9, pp. 264, -265. ■ • Station 44. Three branching specimens, much overgrown with Hydroids, etc. The largest specimen is 19-5cm. in height, with an average diameter of 2'.5 mm. The polyps are about 5 mm. apart ; they are 2o-3 mm. in height, and project at an angle of 45° to 60°. The spicules are as described by Wright and Studer, i.e., "elongated spindles with strong lateral spines, often provided with lateral branched processes, or forked at one extremity." The colour of the specimens is yellowish-brown. Previously recorded from Torres Strait, .3-11 fathoms; Fiji Islands; Maldives. TELESTO ARBOREA. Wright and Studer. (Plate Ixvii., fig. 2.) Telesto arborea, Wright and Studer, Chall. Rep., Zool., xxxi., 1889, pp. 262-264, pi. xxxix., figs. 1, la. Telesto arborea, Thomson and Henderson, Proc. Zool. Soc, 1906, pp. 431-435. Station 42. There are a few small, much broken species in the present collection. The lateral polyps have dimensions of 4 x 2 mm. or 5 X 25 mm. The colour is dark brown. The spicules are as described by Wright and Studer ; we give a figure of them here, as there is apparently none in any previous report of this .species. Previously recorded from Arafura Sea, 49 fathoms ; Zanzibar (Kokotoni Harbour, 5 fathoms, and Wasin Channel, 10 fatlioms); Maldives. ALCYONARIA .1. A. THOMSON AND DOKIS 1,. MACKINNON. 175 Family KOPHOBELEMNONII)^. Genus K O P H O B E L E M N O N , A'oWj/ter. KOPHOBELEMiNON SCHxMELTZII {KMiker). tSclerobelemnonschmelizii, Kolliker, Anatom. Systeiiiat. Besclireib. Alcyonarien, 1 Abtli. Die Pennatuliclen, 1872, p. 312, pi. xxi., figs. 184a, 184n, IS.o. Station 2.5. Three specimens of a chocolate-brown Kophobdemnon. The colonies are cylindrical, 13-1 4 cm. high, the polyp-bearing portion rather thicker (7-9 mm.) than the stalk (5-G mm.) ; there is no terminal bladder ; the polyps are arranged in rather irregular longitudinal rows, leaving only a small bare strip on the pro- rachidial side. The tentacles of the polyps are without spicules. The upper part of the polyps is much poorer in spicules than the lower. Siphonozooids, small, brown, wart-like, in very numerous longitudinal rows. A.\is thick, with a tendency to ([uadransiular cross-section in the lower ])art of the colony. Spicules biscuit- shaped, flat discs and oi-shapes, with a few waits. Thej- are very sparse in the cuti.s of the stalk, numerous in the club portion lound the siphonozooids and on the polyps, absent in the interior. This descri|>tion agrees with that of Sclerohcleianon schmellzii, Kolliker. Thomson and Simpson have shown' ' that it is impossible to maintain a hard and fast line between the genern Kophobelemnon and ScUrobelemon, since there e.xist species sucli as Ko/)/iobelemiion biirgeri, Herklots, and K. intermedium, Thom.son and Simpson, partaking of the characters of both. We therefore keep this species in the older genus Kopliobelemnoti. Locality. — Off Newcastle, 24-48 fathoms. Previously recorded from Formosa. Family PTE ROEI D I D^. Genus GODEFFROYIA, Kiilliker. GODEFFROYIA ELfilGANS, Kolliker. Godeff'royia elegana, Kolliker, Anatom. Systemat. Beschreib. Alcyonarien, 1 Abth. Die PennatuUden, 1872, p. IIC, figs. 63-6."). Stations 22, 54. Kolliker gives the following statement of the generic characters — "Small, delicate sini-pens, of the same type of growth us Pier- 11 Thomson an90 p. lHt>. 1115 EXPLANATION OK PLATE LXI. Fig L —Alci/onm>n(Erythropodium)reptans, Kiikenthal, growing on axis of Primnoella aiislralasice, Gray, x 5. Fia. 2. — Polyp of Alcyonium etheridyei, sp. nov. x 20. Fig. 3. — Lobe of colony of the same with expanded polyps, x 2. MEM. AUSTR. MUS. tV. Hl.ATK XLI. U. DAVIDSON, Del. KXPLANATION OF PLATE LXII. Fig. 1. — Acanthoisis flabelhim, Wright :iud Studer. x 2. Fig. 2. — Detail of tlie same, x 20. Fig. 3. — Very young colony of Alcyonium etheridgei, sp. nov. x 2. Fig. 4. — Dendronephthya taaitei, sp. nov. Coniplote colony Nat. size. MEM. AUSTR. MUS. IV. I'r-ATE LXII. (i DAVinsON, Del. EXPLANATION OF PLATE LXilL Fia. 1. Small portion of colony of Mopsea flnbellum, sp. nov. Nat. size. Fir/. 2. — Axis of same, x 3. Fig. 3. — Enlargement of polyps, x 12. Fi". 4. — Spicules of MopsMa textiformis, Lamarck. Fig. 5. A.xis of same, with a few patches of coenencliyraa. Nat. size. MEM. AUSTR. MUS. IV. I'LATK 1.XI1I. G. DAVIDSON, Del. EXPLAXATION OF PLATE LXIV. Fig. 1. — Mopsen iiustralis, sp. nov. Nat. size. Fig. 2. — ,, „ Enlarged tip of a dried branch, x 15. Fig. 3. — - „ deyaiis, sp. nov. Small portion of a colony. Nat. size, Fig- 4- — ,, „ . Enlargement of stem and polyps, x 25. MKM. AUSTK. MUS. IV. Pl.*TK LXIV. 2 1 hi / f;. riAVIDSllN, Del. EXPLANATION OF PLATE LXV. Fig. L — Polyp of Caligorgia Icevix, sp. nov. Fig. 2 — ,, Detidronephthya waitei, sp. nov. Fig. 3. — ,, A mphilaphis pluniacea, sp. nov. Fig. 4. — „ Pliunarella corruscans, sp. nov. Fig. 5. — ., Plumarella Jilicoides, sp. nov. MENi AUSTR. MUS. IV. Plate I.XV, r f\ G. DAVIDSON, Del. EXPLANATION OF PLATE LXVI. Fig. 1. — Polyp of Pluniareila Icevis, sp. nov. Fit,'. 2. — ,, Mopsea whiteleggei, sp. nov. Fig. 3.- —Spicules of the same. Fig. +. — Polyp of Phiinarella versluysi, sp. nov. Fig. 5. — „ „ thelis, sp. nov. Mem. austr. mts. iv. Pl.ATK LXVl. wf"^: r^ G. DAVIDSON, Del. EXPLANATION OF PLATE LXVIL Fig. 1. — -Spicules of Mopsea dichotoma, Linne. Fig. 2. — ,, Teleslo arbarea, Wright and Studer. Fig. 3. — ,, bendronephthya tvaitei, sp. nov., (a) cortex (b) ba.se, (c) Stiitzbiindel, (d) polyp. Fig- -t. — ,, Alcyonium etheridgei, sp. nov. Fig. 5 — ,, Mopsea australis, sp. uov. Fig. 6. — ,, „ Jlabdlum, sp. nov. MEM. ACSTR. MU? IV. PHTK LXVII, ^^^7^ ^OC" r-'V -r' p-j^ v^^M li. DAVIOSON, Del. EXPLANATION OF PLATE LXVIIL Fig. L — Spicule;! of Plumarella Jilicoides, sp iiov. ,, versluysi, sp. nov. Ainphilaphis plumacea, sp. nov. Plumarella Icevis, sp. nov, Mopsea elegans, sp. nov. Plumarella thetis, sp. nov. Caligorgia l(Bvis, S|>. nov. Plumarella corruscans, sp. nov Mopsella clavitjera, Kidlev. Fig. 2 Fig. 3.— Fig. 4— Fig. 5.— Fig. 6 — Fig. 7.— Fig. 8 — Fig- 9.— MBM. AUSTR. MUS. IV. Platk I.XVIII. ■i "•* ^^^Qp.)^^' ' i- '^«i^^ ''\ -e*^' »"•»" I l\ "^ G. DAVIDSON. Uel. EXPLANATION OF PLATE LXIX. Ahyoniiim etheridgei, sp. nov. Nat. size. MEM. AUSTR, MUS. IV. Platk LXIX. M t ^ 1 EXPLANATION OF PLATE LXX. Parisis australis, Wright and Studer. | nat. size. MEM. AUSTR. MIJS. IV, Plate LXX. EXPLANATION OF PLATE LXXI. Mopsella flahellum, sp. nov. Nat. size. MEM. AUSTR. MUS. iv. Plate LXXI EXPLA-NATION OF PLATE LXXII. Mopsea elegans, sp nov. Nat. size. MEM. AUSTK. MUS. IV. HLATt LXXII. RXPLANATION OF PLATE LXXIII. Mopsea whUeleggei, sp. no v. Nat. size MEM. AUSTR. MUS. IV. Platk LXXIII. EXPLANATION OF PLATE LXXIV. 4mphilaphis phiniacea, sp. nov. Nat size. MEM. AUSTR. MUS. IV, Plate LXXIV, EXPLANATION OF PLATE LXXV. Plumarelli Icevis, sp. nov. Half nat. size. MEM. AUSTR. MUS. IV. Pl-ATK LXXV. EXPLANATION OF PLATE LXXVI. flumarella thetix, sp. nov. Nat. sizp. MEM. AUSTR. MUS. IV. Plate LXXVL EXPLANATION" OF PLATE LXXVIL Pluniarella cumiscaiis, sp. nov. Hull' n«t. size MKM. AUSTR. MUS, IV. Pi.vrK LXXVIl EXPLANATION OF PLATE LXXVIII. Flttniarella JilicoiiliiS, S|i. nov. Nat. size. MEM. AUSTR. MUS. IV. Plate LXXVIIl. EXPLANATION OF PLATE LXXIX Plumarella versb'ysi, sp nov. Nat. size. MEM. AUSTR. MUS. IV, Plate LXXIX. EXPLANATION OK PLATE LXXX. Caligorgia kevis, sp. nov. Nat. size. MEM. AUSTR. MUS. IV. Plate LXXX. EXPLANATION OF PLATE LXXXJ. Ctenocella pectinata, Pallas. Half iiat. size MBM. AUSTR. MUS. IV. Plate LXXXI. EXPLANATION OF PLATE LXXXII Sarcophyllum auslrale, Kolliker. Nat size Mem. austr, mus. iv. Pr.ATF. I.XXXIT. } PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY QL mi C6S925 Biological & Medical Zoological studies chiefly on alcyonaiians i5M