rK fjJl MARINE BIOLOGICAL LABORATORY. Received Accession No mi mi Given by V^oMj^v^Ojl^ Lct^^^^^a^^ \*Uo book on pamphlet is to be removed from the lab- oratory without the permission of the Trustees. Place, . -. . THE VOYAGE OF H.M.S. CHALLENGER. ZOOLOGY-VOL. XXXII. > . REPORT tyc.t SCIENTIFIC RESULTS OK THE VOYAGE OF H.M.S. CHALLENGER DURING THE YEARS i 8 7 3-7 6 UNDER THE COMMAND OF Captain GEORGE S. NARES. R.N., F.R.S. AND THE LATE Captain FRANK TOURLE THOMSON, R.N. PREPARED UNDER THE SUPERINTENDENCE OF THE LATE Sir C. WYVILLE THOMSON, Knt., F.R.S., &c. REGIUS PROFESSOR OF NATURAL HISTORY IN THE UNIVERSITY OF EDINBURGH DIRECTOR OF THE CIVILIAN SCIENTIFIC STAFF ON BOARD AND NOW OF JOHN MURRAY, LL.D., Ph.D., &c. ONF. OF THE NATURALISTS OF THE EXPEDITION Zoology— Vol. XXXII. IPubltsfoed bp ©mer of Der jflajestp's ©obraiment PRINTED FOR HER MAJESTY'S STATIONERY OFFICE AND SOLD BY LONDON :— EYRE & SPOTTISWOODE, EAST HARDING STREET, FETTER LANE EDINBURGH :— ADAM & CHARLES BLACK DUBLIN :— HODGES. FIGGIS. & CO. 1889 Price Twenty-five Shillings. The Editor of the Challenger Reports will be greatly obliged to Authors sending him copies of separate papers, or references to works in which the Challenger discoveries are referred to, or the observations of the Expedition are discussed. This will greatly facilitate the compilation of a complete Bibliography, and the discussion of the results of the Expedition, in the final Volume of the Series. Letters and Papers should be addressed — JOHN MURRAY, Challenged Oee/ce, 32 Queen Street, EDINBURGH. PRINTED BY NEILL AND COMPANY, EDINBURGH, FOR HER MAJESTY'S STATIONERY OFFICE. % rYY CONTENTS. I. — Report on the Antipatharia collected by H.M.S. Challenger during the years 1873-1876. By George Brook, F.L.S., F.R.S.E. (The Manuscript was received in Instalments between 29th December 1888 and 22nd July 1889.) II. — Supplementary Report on the Alcyonaria collected by H.M.S. Challenger during the years 1873-1876. By Professor Th. Studer, M.D. ( The Manuscript loas received 28th December 1888 and 1st March 18S9.) III. — Report on the Deep-Sea Keratosa collected by H.M.S. Challenger during the years 1873-1876. By Professor Ernst Haeckel, M.D., Ph.D., Hon. F.R.S.E. (TJie Manuscript ivas received I4th March 1889.) EDITORIAL NOTES. Tins Volume contains Parts LXXX., LXXXL, and LXXXII. of the Zoological Series of Reports on the Scientific Results of the Expedition. Part LXXX. — About a year ago Mr. George Brook, Lecturer on Comparative Embryology in the University of Edinburgh, was induced by me to undertake the examination and description of the small group of Antipatharia collected during the Expedition. Since the collection was placed in his hands Mr. Brook has devoted nearly the whole of his time to the investigation. The accompanying Report, which may be regarded as a Monograph, shows what large additions have been made to our knowledge of the morphology of this interesting group of deep-sea animals. The Report consists of 222 pages of letterpress, illustrated by 15 litho- graphic plates and numerous woodcuts. Part LXXXI. — This is a Supplementary Report by Professor Studer on a few species of Alcyonaria, which were found in his possession too late to be included in the Report on the Alcyonaria, by Professors "Wright and Studer, published in January last (Part LXIV. Vol. XXXI.). The Report includes a summary of the geographical and bathymetrical distribution of the Alcyonaria, and consists of 31 pages of letterpress, with 6 chromo-lithographic plates. Part LXXXII. — In this Part Professor Ernst Haeckel makes his fourth contribution to the Challenger series of Zoological Reports, and here deals Vlii THE VOYAGE OF H.M.S. CHALLENGER. with a group of Deep-Sea Keratosa. These are interesting from the symbiosis with Hydroids, from the mass of the pseudo-skeleton, consisting of the materials of deep-sea deposits, from the primitive forms described under the name Ammoconidoe, and the difficulties which were experienced by many naturalists in determining the true nature of these remarkable organisms. The Eeport extends to 92 pages, accompanied by 8 chromo-lithographic plates. This volume concludes the Zoological Series of Eeports on the Scientific Eesults of the Expedition, with the possible exception of a few supplementary notes to some of the memoirs and Professor Huxley's Eeport on the genus Spirula, which may appear as an appendix to the concluding Summary Volume. The first volume of these Zoological Eeports was issued in 1880, and the others have appeared at short intervals up to the present time. During the same period there have been issued from the press two Botanical Volumes, one volume of Physical and Chemical Eesearches, and three volumes of the Narrative of the Cruise with appendices. A second Physical and Chemical Volume will be published within the next two months ; the volume on Deep- Sea Deposits will follow next year, and the whole work will be completed in a Summary Volume. As the Biological Series of Special Eeports may now be said to have been completed, there is appended a list of these arranged, first, according to the subject matter, and, second, according to the contents of each volume. Of Zoological Eeports there are nine Eeports (in 12 Parts) dealing with the Vertebrata ; one Eeport (in 3 Parts) dealing with the Tunicata ; thirteen Eeports (in 17 Parts) dealing with the Molluscoidea and Mollusca ; fourteen Eeports (in 16 Parts) dealing with the Arthropoda ; five Eeports (in 7 Parts) dealing with the Echinodermata ; five Eeports (in 6 Parts) EDITORIAL NOTES. ix dealing with the Vermes ; fifteen Reports (in 18 Parts) dealing with the Ccelenterata ; and three Reports (in 3 Parts) dealing with the Protozoa. Of Botanical Reports there are three Reports dealing with the Botany of Oceanic Islands, and one Report on the Marine Diatomace^e. These Biological Reports have been issued, whenever ready, and without any reference to systematic arrangement of the subjects treated of, in eighty- seven separate parts, and as thirty-four volumes of Reports, the whole, together with the illustrations, being bound up in forty-two large quarto volumes. These volumes contain 24,700 pages of letterpress, 2600 quarto lithographic and chromo-lithographic plates, with many maps and numerous woodcuts. The Biological Reports have been contributed by sixty-two separate authors, forty-two of whom are resident in the United Kingdom, India, and the British Colonies, seven in Germany, three in the United States, two in Holland, and one in each of the following countries, viz., France, Russia, Sweden, Norway, Denmark, Switzerland, Belgium, and Italy, so that nearly all civilised countries have taken part in the pro- duction of this great work. The cosmopolitan or international nature of the undertaking becomes still more evident, if account be taken of the large number of naturalists and others, in different parts of the world, who have in various ways assisted the several authors and the editor while carrying on their investigations. For my own part, I desire now to record my indebtedness, and to convey my thanks to all the contributors, and to all those who have, by the loan of specimens, books, and manuscript, by information and advice, or in any other way, assisted me in carrying to a successful issue the biological work con- nected with these reports on the Scientific Results of the Expedition. All the type specimens referred to in the Zoological Reports have been, or will be, within the next few months, placed in the British Museum. It is somewhat remarkable that although the various Challenger Collections X THE VOYAGE OF H.M.S. CHALLENGER. have been transported to and from so many distant parts of the world, they have as yet met with no accident. The Botanical Collections are in the National Herbarium at Kew. John Murray. Challenger Office, 32 Queen Street, Edinburgh, 6th September 1889. -LIST OF THE CHALLENGER ZOOLOGICAL REPORTS ARRANGED IN SYSTEMATIC ORDER. Vertebrata : — Human Skeletons (part xxix. vol. x., and part xlvii. vol. xvi.). Seals (part lxviii. vol. xxvi.). Bones of Cetacea (part iv. vol. L). Marsupialia (part xvi. vol. v.). Birds (part viii. vol. ii.). Anatomy of Petrels (part xL vol. iv.). Anatomy of Spheniscidse (part xviii. vol. vii.). Development of Green Turtle (part v. vol. i.). Fishes (part vi. vol. i., part lvii. vol. xxii., and part lxxviii. vol. xxxi.). Tunicata : — Tunicata (part xvii. vol. vi., part xxxviii. vol. xiv., and part lxxvi. vol. xxvii.). MOLLUSCOIDEA AND MoLLUSCA : — Brachiopoda (part i. vol. i.). Polyzoa (part xxx. vol. x., part 1. vol. xvii., and part lxxix. vol. xxxi.). Cephalodiscus (part lxii. vol xx.). Phoronis (part lxxv. vol. xxvii.). Cephalopoda (part xliv. vol. xvi.). Pteropoda (part lviii. vol. xix., part Ixv. vol. xxiii., and part Ixvi. vol. xxiii.). Nudibranehiata (part xxvi. vol. x.). Marseniadre (part xli. vol. xv.). Heteropoda (part lxxii. vol. xxiii.). Scaphopoda and Gasteropoda (part xlii. vol. xv.). Polyplacophora (part xliii. vol. xv.). Lamellibranchiata (part xxxv. vol. xiii.). Anatomy of Deep-Sea Mollusca (part lxxiv. vol. xxvii.). Arthropoda : — Pelagic Hemiptera (part xix. vol. vii.). Pycnogonida (part x. vol. iii.). Brachyura (part xlix. vol. xvii.). Anomura (part lxix. vol. xxvii.). Macrura (part lii. vol. xxiv. ). Schizopoda (part xxxvii. vol. xiii.). Stomatopoda (part xlv. voL xvi.). Cumaeea (part lv. vol. xix.). Phyllocarida (part lvi. vol. xix.). Isopoda (part xxxiii. vol. xi., and part xlviii. vol. xvii.). Amphipoda (part lxvii. vol. xxix). Cirripedia (part xxv. vol. viii., and part xxviii. vol. x.). Copepoda (part xxiii. vol. viii.). Ostraeoda (part iii. vol. i.). Echinodermata : — Holothurioidea (part xiii. vol. iv., and part xxxix. voL xiv.). Echinoidea (part ix. vol. iii.). Ophiuroidea (part xiv. vol. v.). Asteroidea (part li. vol. xxx.). Crinoidea (part xxxii. vol. xi., and part Ix. vol. xxvi.). Vermes : — Myzostomida (part xxvii. vol. x., vol. xx.). Annelida (part xxxiv. vol. xii.). Gephyrea (part xxxvi. vol. xiii.). Nemertea (part liv. vol. xix.). Entozoa (part lxxi. vol. xxiii.). and part Ixi. EDITORIAL NOTES. XI OcELENTERATA : — Siphonophorse (part lxxvii. vol. xxviii.). Depp-Sea Medusa? (part xii. vol. iv.). Hydroida (part xx. vol. vii., and part lxx. vol. xxiii. ). Corals (part vii. vol. ii.). Eeef Corals (part xlvi. vol. xvi.). Actiniaria (part xv. vol. vi., and part lxxiii. vol. xxvi.). Antipatharia (part lxxx. vol. xxxii.). Alcyonaria (part lxiv. voL xxxi., and part lxxxi. vol. xxxii.). Ccelenterata — continued. Pennatulida (part ii. vol. i.). Calcarea (part xxiv. vol. viii.). Hexactinellida (part liii. vol. xxi.). Tetractinellida (part lxiii. vol. xxv.). Monaxonida (part lix. vol. xx.). Keratosa (part xxxi. vol. xi.). Deep-Sea Keratosa (part lxxxii. vol. xxxii.). Protozoa : — Radiolaria (part xl. vol. xviii.). Foraminifera (part xxii. vol. ix.). Orbitolites (part xxi. vol. vii.). II.— LIST OF ZOOLOGICAL VOLUMES OF THE REPORT, WITH THE CONTEXTS OF EACH. Volume I. (1880) contains: — Part I. — Brachiopoda. By Thomas Davidson, F.R.S., F.L.S., F.G.S., V.P.P.S. ' Part II. —Pennatulida. By Professor Albert v. Kolliker, F.M.R.S., Hon. F.E.S.E. Part III.- — Osteacoda. By G. Stewardson Brady, M.D., F.R.S., F.L.S. Part IV.— Cetacea, Bones of. By Professor William Turner, M.B. (Lond.), F.R.SS. L. & E. Part V. — Green Turtle, Development of the. By William Kitchen Parker, F.B.S., F.L.S., F.Z.S. Part VI. — Shore Fishes. By Albert Giinther, M.A., M.D., Ph.D., F.E.S., V.P.Z.S., F.L.S. Volume II. (1881) contains : — Part VII. — Corals. By Professor H. K Moseley, M.A., F.E.S., F.Z.S., F.L.S. Part VIII— Birds. By P. L. Sclater, F.E.S., F.L.S., and others. Volume III. (1881) contains : — Part IX. — Echinoidea. By Alexander Agassiz. Part X.— Pycnogonida. By P. P. C. Hoek, Assist. Zool. Lab., Leyden. Volume IV. (18S2) contains: — Part XL- — Petrels, Anatomy of the. By W. A. Forbes, B.A., F.L.S., F.G.S., F.Z.S. Part XII. — Deep-Sea Medusa. By Professor Ernst Haeckel, M.D., Ph.D. Part XIII. — Holothurioidea. First Part.— The Elasipoda. By Hjalmar Theel. Volume V. (1882) contains: — Part XIV. — Ophiuroidea. By Theodore Lyman. Part XVI. — Marsupialia. By Professor D. J. Cunningham, M.D., F.R.S.E., F.E.C.S.I. Volume VI. (1882) contains : — Part XV. — Actiniaria. By Professor Richard Hertwig. Part XVII. — Tunicata. Part I. — Ascidise Sim- plices. By Professor W. A. Herdman, D.Sc, F.R.S.E., F.L.S. Volume VII. (1883) contains: — Part XVIII. — Spheniscid-e, Anatomy of the. By Professor Morrison Watson, M.D., F.R.S.E., F.Z.S. Part XIX. — Pelagic Hemiptera. By F.Buchanan White, M.D., F.L.S. Part XX. — Hydroida. First Part. — Plumularida;. By Professor G. J. Allnian, M.D., LL.D., F.R.SS. L. & E., M.R.I.A., V.P.L.S. Part XXI. — Orbitolites, Specimens of the Genus. By W. B. Carpenter, C.B., M.D., LL.D.. F.R.S., F.G.S., V.P.L.S. Volume VIII. (1883) contains : — Part XXIII. — Copepoda. By G. Stewardson Brady, M.D., F.R.S., &c. Part XXIV— Calcarea. By X. Polejaeff, M.A., of the University of Odessa. Part XXV. — Cirripedia. — Systematic Part. By P. P. C. Hoek, Leyden. Xll THE VOYAGE OF H.M.S. CHALLENGER. Volume IX. (1884) contains : — Part XXII— Foraminifera. By H. B. Brady, F.E.S., F.L.S., F.G.S. (One vol. text and one vol. plates.) Volume X. (1884) contains : — Part XXVL— Nudibranchiata. By Dr. Rudolph Bergh. Part XXVIL— Myzostomida. By Professor Lud- wig von Graff. Part XXVIIL— Cirripedia.— Anatomical Part. By Dr. P. P. C. Hoek. Part XXIX. — Human Skeletons. First Part. — The Crania. By Professor William Turner, M.B., F.R.SS. L. & E. Part XXX. — Poltzoa. Part I. — Cheilostomata. By George Busk, F.R.S., V.P.L.S., &c. Volume XL (1884) contains : — Part XXXL— Keratosa. By N. Polejaeff, M.A. Part XXXIL— Crinoidea. Part I.— Stalked Crinoids. By P. H. Carpenter, M.A, D.Sc. Part XXXIII. — Isopoda. Part I.— Genus Serolis. By F. E. Beddard, M.A, F.R.S.E., F.R.M.S., F.Z.S., M.B.O.U. Volume XII. (1885) contains : — Part XXXIV. — Annelida Polych;eta. fessor W. C. M'Intosh, F.R.S. By Pro- Volume XIII. (1885) contains : — Part XXXV. — Lamellibranchiata. By Edgar A. Smith, F.Z.S. Part XXXVI. —Gephyrea. By Professor Emil Selenka. Part XXXVII. — Schizopoda. By Professor G. 0. Sars. Volume XIV. (1886) contains : — Part XXXVIII. —Tunicat a. Part II.— Ascidise Compositse. By Professor W. A. Herdman. Part XXXIX. — Holothurioidea. Second Part. By Dr. Hjalmar Thdel. Volume XV. (1886) contains :— PartXLI. — MARSENiADiE. By Dr. Rudolph Bergh. Part XLII. — Scaphopoda and Gasteropoda. By Rev. R. Boog Watson, F.L.S. Part XLIII. — Polyplacophora. By Professor Alfred C. Haddon, M.A., M.R.I. A. Volume XVI. (1886) contains :— Part X LI V.— Cephalopoda. By William Evans Hoyle, M.A. (Oxon), M.R.C.S., F.RS.E. Part XL V.— Stomatopoda. By Professor W. K. Brooks. Part XL VI.— Reef Corals. By John J. Quelch, B.Sc. (Lond.). Part XLVII. — Human Skeletons. Second Part. By Professor Sir William Turner, Knt, LL.D., F.R.SS. L. & E. Volume XVII. (1886) contains :— Part XLVIIL— Isopoda. Part II. By F. E. Beddard, M.A, F.R.S.E., &c. Part XLIX.— Brachyura. By Edw. J. Miers, F.Z.S., F.L.S. Part L. — Polyzoa. Part II.— Cyclostomata, Ctenostomata, and Pedicellinea. By George Busk, F.R.S., V.P.L.S., &c. Volume XVIII. (1887) contains :— Part XL. — Radiolaria. By Professor Ernst Haeckel. (Two vols, text and one vol. plates.) Volume XIX. (1887) contains :— Part LIV.— Nemertea. By Dr. A. , A. W. Hubrecht, LL.D., C.M.Z.S. Part LV.— Cumacea. By Professor G. O. Sars. Part LVI. — Phyllocarida. By Professor G. O. Sars. Part LVIII. — Pteropoda. Part I. — Gymnoso- niata. By Paul Pelseneer, D.Sc. Volume XX. (1887) contains : — Part LIX. — Monaxonida. By Stuart 0. Ridley, M.A., F.Z.S., and Arthur Dendy, B.Sc, F.Z.S. Part LXI. — Myzostomida (Supplement). By Professor L. von Graff. PartLXII. — Cephalodiscus dodecalophus. By Professor William C. M'Intosh, M.D., LL.D., F.R.S. Volume XXI. (1887) contains : — Part LIII. — Hexactinellida. By Professor F. E. Schulze. (One vol. text and one vol. plates.) Volume XXII. (1887) contains : — Part LVIL— Deep-Sea Fishes. By Dr. Albert Giinther, M.A., M.D., Ph.D., F.R.S. EDITORIAL NOTES. X1U Volume XXIII. (18S8) contains :— Part LXV. — Pteeopoda. Part II. — Thecosomata. By Dr. Paul Pelseneer. Part LXVI. — Pteropoda. Part III. — Anatomy. By Dr. Paul Pelseneer. Part LXX. — Htdroida. Second Part. By Pro- fessor G. J. Allnian. Part LXXI. — Entozoa. By Dr. 0. v. Linstow. Part LXXII. — Heteropoda. By Edgar A. Smith, F.Z.S. Volume XXIV. (1888) contains :— Part LII. — Crustacea Macrura. By C. Spence Bate, F.R.S., F.L.S. (One voL text and one voL plates.) Volume XXV. (18SS) contains : — Part LXIIL — Tetractinellida. By Professor W. J. SoUas. Volume XXVI. (1888) contains:— Part LX. — Crinoidea. Part II. — Coniatulse. By Dr. P. H. Carpenter. Part LXVIIL— Seals. By Professor Sir William Turner. Part LXIIL — Actiniaria (Supplement). By Professor Richard Hertwig. Volume XXVII. (1888) contains:— Part LXIX. — Anomura. By Professor J. K. Henderson. Part LXXIV. — Anatomy of Deep-Sea Mollusca. By Dr. Paul Pelseneer. Volume XXVII. — continued. Part LXXV. — Phoronis buskii. By Professor W. C. M'Intosh, F.R.S. Part LXXVL— Tunicata. Part III. By Pro- fessor W. A. Herdnian, F.L.S. Volume XXVIII. (1888) contains:— Part LXXVII. — Siphonophor^e. By Professor Ernst Haeckel. Volume XXIX. (1888) contains :— Part LXVIL— Amphipoda. By Rev. Thomas R. R. Stebbing, M.A. (Two vols, text and one voL plates.) Volume XXX. (1889) contains:— Part LI. — Asteroidea. By W. Percy Sladen, Sec. L.S. (One vol. text and one vol. plates.) Volume XXXI. (1889) contains : — Part LXIV. — Alcyonaria. By Professor E. P. Wright, M.D., and Professor Th. Studer, M.D. Part LXXVIIL— Pelagic Fishes. By Dr. A. Giinther, F.R.S. , &c. Part LXXIX. — Supplementary Report on the Polyzoa. By A. W. Waters, F.L.S., F.G.S. Volume XXXII. (1889) contains : — Part LXXX. — Antipatharia. By George Brook, F.L.S., F.R.S.E. Part LXXXI. — Supplementary Report on the Alcyonaria. By Professor Th. Studer. Part LXXXII. — Deep-Sea Keratosa. By Pro- fessor Ernst Haeckel. III.— BOTANICAL VOLUMES, WITH THEIR CONTENTS. Volume I. (1885) contains : — Present State of Knowledge of various Insular Floras, being an introduction to the first three parts of the Botany of the Challenger Expedition. By W. B. Hemsley, A.L.S. Part I. — Botany of the Bermudas and various other Islands of the Atlantic and Southern < Iceans. — The Bermudas. By W. B. Hemsley, A.L.S. Part II. — Botany of the Bermudas and various other Islands of the Atlantic and Southern Oceans. — St. Paul's Bocks, &c. By W. B. Hemsley, A.L.S. Part III. — Botany of Juan Fernandez, South- eastern Moluccas, and the Admiralty Islands. By W. B. Hemsley, A.L.S. Volume II. (1886) contains :— Part IV. — Diatomace.e. By Conte Abate Fran- cesco Castracane. THE VOYAGE OF H.M.S. CHALLENGEB. ZOOLOGY. REPORT on the Antipatharia collected by H.M.S. Challenger during the Years 1873-76. By George Brook, F.L.S., F.R.S.E., Lecturer on Comparative Embryology in the University of Edinburgh. PREFACE. In June 1888 I was invited by Dr. John Murray to prepare a Report on the Antipatharia obtained by the Challenger Expedition, aud the majority of the specimens were soon afterwards placed in my hands for identification. Probably no group of marine animals has received so little systematic attention during recent years as the Antipatharia, and our knowledge of their morphology has hitherto been confined to a partial study of two or three species. The descriptions of genera and species alike have been almost entirely based on skeletal characters, and prior to the Challenger Expedition probably few specimens were in existence, the polyps of which were well preserved. During the recent "Blake" and "Hassler" expeditions, a number of new and interesting species were brought to light by the United States Coast Survey, which have been described by Pourtales, but no attempt has been made to rearrange the species already known in the light of these newer investigations. It soon became apparent that the Challenger collection, although not large, offered exceptional facdities for a morphological study of the group, many of the more impor- tant features of which have hitherto escaped notice. I have, therefore, taken this opportunity of attempting a partial revision of the Antipatharia, and have endeavoured, so far as opportunity would allow, to place the classification on a more natural basis. In the time at my disposal I have been unable to visit many of the more important Continental collections, and have relied chiefly on a comparison of the collections of the (ZOOL. CHALL. EXP. PiET LXXX. — 1889.) L111 a ii THE VOYAGE OF H.M.S. CHALLENGER. British Museum, and of the Zoological Museum at Copenhagen, for the identification of species described by the earlier authors, the majority of which are very imperfectly characterised. The British Museum collection is very extensive, and contains represen- tatives of over 40 species, including most of the types described by Gray. Unfortu- nately very few of the specimens are preserved in spirit, and as the polyps of a large number of species are still unknown I have been unable to suggest the generic position of several of them. Owing to the very imperfect description of the majority of the forms already known, and the consequent confusion in nomenclature which exists, it has appeared necessary to extend the scope of the present Report so as to include an account of the whole of the known Antipatharia. I trust that the additional informa- tion which I am enabled to give of all the species that have come under my notice, sixty-eight in all, will prove sufficient for their future identification. Undoubtedly a study of the fine collections of the Museum of Comparative Zoology at Harvard College would have rendered my work both easier and much more complete, but unfortunately this has been impossible. Professor Alexander Agassiz, to whom I am indebted for many acts of kindness, informs me that the specimens were handed to Professor Verrill for examination several years ago, and that they have not yet been returned. The following pages include descriptions of 16 genera, 98 species and 4 varieties, of which 11 genera, 41 species and 2 varieties are new. Representatives of 9 genera, 19 species and 1 variety are included in the Challenger collection, all of which are new. The types of the remaining new species are all in the British Museum, with the' excep- tion of three; two of these are preserved in the Zoological Museum at Copenhagen, and the third is in the Museum of Comparative Zoology at Harvard College. The fact that all the Challenger species are new is largely accounted for by the circumstance that nearly all the specimens were obtained in regions hitherto unexplored, or from which no Antipatharia had previously been recorded. The collection is remark- ably deficient in littoral forms, whilst on the other hand quite a number of species have been shown to inhabit abyssal depths, a fact which has hitherto been unknown. The abyssal species present many new and interesting features, and are characterised by a type of dimorphism which is apparently not found in any other Zoantharia. In the limited time allowed for the completion of my Report I have only been able, in addition to the systematic portion of the work, to prepare a preliminary account of the anatomy of the Antipathinas. The structure of the Schizopathime will form the subject of a future paper, which I hope to finish before the end of the year. I desire to acknowledge gratefully the privileges afforded me by Dr. Albert Giinther of the British Museum, Professor Chr. F. Liitken of Copenhagen, Mr. Charles Stewart of the Museum of the Royal College of Surgeons, London, and Mr. Moore of the Liverpool Free Museum, all of whom have rendered me every facility for an investiga- tion of the collections under their charge. REPORT ON THE ANTIPATHARIA. ill My hearty thanks are due to Dr. Anton Dohrn of the Naples Zoological Station for specimens and information concerning the distribution of the Mediterranean species, and to Professor F. Jeffrey Bell for valuable assistance during my study of the British Museum specimens. I am also indebted to Professor E. Perceval Wright and Mr. W. Percy Sladen for much friendly advice during the progress of my work. The drawings have been chiefly made under my supervision by my assistant Mr. F. G. Binnie, but Plates XI. and XII., together with a number of the histological figures, have been drawn by Mr. James T. Murray. In conclusion I desire to express my warmest thanks to Dr. John Murray for having given me the opportunity of investigating such a little-known group of animals, and for his many acts of kindness and assistance during the progress of my Report. My thanks are also due to Mr. James Chumley of the Editorial Staff for the trouble he has taken with my proof-sheets. TABLE OF CONTENTS. Bibliography, Critical, Review of Literature, General Morphology, The homologies of the mesenteries, Complete and incomplete mesenteries, Dimorphism, Colony formation, . Ccenenchyma, Skeleton formation, Origin and arrangement of spines, Retrogressive development, Classification, Table of Genera and Species, Description of Species, Geographical Distribution, . Bathymetrical Range, Anatomy of the Antipathin^:, Mesenterial filaments, General conclusions, Systematic Index, Page 1 5 35 52 59 60 62 64 65 65 67 71 77 79 179 187 191 214 216 219 BIBLIOGRAPHY. 1. 1605. Cldsius, C, Exoticorum, libri decern. ; Antw., 1605. 2. 1617. Pon,£, J., Monto Baldo descritto; Venet., 1617. 3. 1623. Bauhini, C, Pinax theatri botanici ; Basil., 1623. 4. 1648. Aldrovandus, U., Museum metallicum ; Bonon., 1648. 5. 1697. Boccone, P., Museo di Fisica e di Esperienze ; Venet., 1697. 6. 1699. Moeison, Robt., Plantarum historiae universalis oxoniensis, t. iii.; Oxon., 1699. 7 1700. Tournefoet, J. P. de, Institutiones rei herbaria?, t. iii. et Corollarium ; Paris, 1700. 8. 1713. Petiver, J., Gazopbylacii naturae et artis, Decades x. ; Lond., 1713. 9. 1724-1726. Valentyn, Fr., Bescbryving van Oud en Nieuw Oostindien, t. v.; Amsterd., 1724-26. 10. 1725. Marsigli, L. F., Histoire physique de la mer ; Amst., 1725. 11. 1733. Linnaeus, C. v., Hortus Cliffortianus ; Amst., 1733. 12. 1739. Breyn, J., Prodromi rariorum plantarum, Fasc. ii. ; Gedani, 1739. 13. 1750. Rdmphius, G. E., Herbarium Amboinense, t. vi.; Amstelod., 1750. 14. 1760. Seba, A., Locupletissimi rerum naturalium thesauri, &c, t. iii.; Amstelod., 1760. 15. 1765. Donati, V., (SulTAntipathe delFAdriatico), Giornali d'ltalia spettante alia Sc. nat. sci., t. i. p. 54 (cf. Nardo, 61). 16. 1766-1768. LiNNiEus, C. v., Systema naturae, Ed. 12 (cf. also Ed. 10). 17. 1766. Pallas, P. S., Elenchus Zoophytorum, 8vo, Hagae Com.; 1776. 18. 1768. Boddaert, Lyst der Plantend. CoraaL (Dutch translation of 17, with addition of plates). 19. 17S6. Ellis and Solander, Natural History of many curious and uncommon Zoophytes, 4to ; Lond., 1786. 20. 1787. "Wilkens and Herbst, Charakteristik der Thierpflanzen (translation of 17, with additions and plates). 21. 1788. Esper, E. J. C, Die Pnanzenthiere, 4 parts and 2 volumes of plates, 4to; Niirnberg, 1788-1830. 22. 1792. Brtjguiere, J. G., Encyclopedie methodique (Hist. nat. d. vers.), t. i., 4to; Paris, 1792. 23. 1816. Lamarck, J. B., Histoire naturelle des animaux sans vertebres, t. ii.; Paris, 1816. 24 1816. Lamotjroux, J. V. F., Histoire generale des Polypiers coralligenes flexibles, 8vo ; Caen, 1816. 25. 1821. Lamotjroux, J. V. F., Exposition methodique des genres de l'ordre des polypiers, 4to; Paris, 1821. 26. 1824. Lamourodx, J. V. F., (Bory de Saint- Vincent et E. Deslongchamps), Encyclopedie methodique, t. iv. Zoophytes, 4to ; Paris, 1824. 27. 1826. Risso, Histoire naturelle des principales productions de l'Europe nuridionale, t. v. ; Paris, 18^6. 28. 1832. Gray, J. E., On the Animal of Antipathes. Proc. Zool. Soc. Lond, pt. ii. p. 41. 29. 1834. Ehrexberg, C. G., Die Korallenthiere des Rothen Meeres; Berlin, 1834. 30. (?) Bertolini, A., Amenitatio Italica (cf. Nardo, 61). 31. 1834. Blainville, H. M. D. de, Manuel dActinologie, 1 vol. 8vo, and atlas of plates ; Paris, 1834-37. 32. 1842. Gray, J. E., Synopsis of the British Museum, 1842 Ed., p. 135. 33. 1843. Nardo (Savaglia, n. gen.), Atti 5 Congresso d. sc. ital. in Lucca (cf 61). (zool. chall. exp.— part lxxx. — 1889.) Llll 1 2 THE VOYAGE OF H.M.S. CHALLENGEE. 34. 1846. Dana, J. D., Eeport on Zoophytes (United States Exploring Expedition), 1 vol., 4to, and atlas of plates; Philadelphia, 1346-49. 35. 1849. Haisie, J., Note sur le polipieroide d'un Leiopathes glaberrima. Ann. d. Sci. Nat., ser. 3, t. xii. pp. 284-7. 36. 1850. Milne-Edwards and Haime, British Fossil Corals. Pcdxontogr. Soc, vol. i., 1850, Introd. p. lxxiii. 37. 1851. Milne-Edwards and Haime, Monographic des Polypiers fossiles des terrains pafeozoiques (Distrib. method, d. 1. classe d. polypes, pp. 175-6). Archives du Museum, t. v. ; Paris, 1851. 38. 1857. Gray, J. E., On the Animal and Bark of Antipathes. Proc. Zool. Soc. Land., 1857, p. 113, 1 pi. 39. 1857. Gray, J. E., Description of New Genera of Gorgoniacte. Proc. Zool. Soc. Land,, 1 857, p. 158, 1 pi. 40. 1857. Gray, J. E., Synopsis of the Families and Genera of Axiferous Zoophytes or Barked Corals. Proc. Zool. Soc. Lond., 1857, pp. 278-294, 1 pi. 41. 1857. Milne-Edwards, H., Histoire naturelle des Coralliaires, t. i.; Paris, 1857. 42. 1860. Gray, J. E., Notice of New Corals from Madeira. Ann. and Mag. Nat. Hist., ser. 3, vol. vi. p. 311. 43. 1860 and 1866. Duchassaing and Michelotti, Memoire sur les Coralliaires des Antilles. Mem. Acad. Torino, ser. 2, t. xix., 1860; also Supplement, ibid. t. xxiii., 1866. 44. 1864. Lacaze Duthiers, H. de, Mcrnoire sur les Antipathaires (gen. Gerardia). Comptes rendus, t. lix. p. 86; Ann. d. Sci. Nat. (Zool.), ser. 5, t. ii. pp. 169-239, and 6 pis. (Abstr. Ann. and Blag. Nat. Hist., 1864, pp. 241-243.) 45. 1864-1865. Lacaze Ddthiers, H. de, Memoire sur les Antipathaires (gen. Antipathes). Comptes rendus, t. lix. pp. 192-195; Ann. d. Sci. Nat. (Zool.), ser. 5, t. iv. pp. 1-61, and 4 pis. (Abstr. Ann. and Mag. Nat. Hist, 1864, p. 198.) 46. 1865. Verrill, A. E., Classification of Polyps. Proc. Essex. Inst., 1865; Amer. Journ. Sci. and Arts, vol. xl., 1865; Ann. and Mag. Nat. Hist, 1865, pp. 191-197. 47. 1866. Kolliker, A., Icones Histiologies, Abth. II. Hft. 1 (D. Bindesubstanz d. Ccclenteraten), 4to, 10 plates and 13 figs.; Leipzig, 1866. 48. 1866. Verrill, A. E., Keview of the Polypi of the East Coast of the United States. Mem. Boston Soc. Nat. Hist, vol. i. (1866-^69) pp. 1-45. 49. 1868. Gray, J. E., Description of New Genera and Species of Alcyonoid Corals in the British Museum. Ann. and Mag. Ned. Hist., ser. 4, vol. ii. pp. 441-445. 50. 1868. Heller, C, Die Zoophyten und Echinodermen d. Adriat. Meeres; Wien, 1868. 51. 1868. Pourtales, L. F. de, Contribution to the Fauna of the Gulf Stream at Great Depths. Bull. Mus. Comp>. Zool., Harvard Coll., No. 6, 1867, No. 7, 1868, pp. 103-142. 52. 1868. Verrill, A. E., Critical Beuiarks on Halcyonoid Polyps in the Museum of Yale College. Amer. Journ. Sci. caul. Arts, vol. xlv., 1868, pp. 411-416; Ibid., vol. xlvi., 1868, pp. 143-144. 53. 1869. Verrill, A. E., Notes on the Radiata of the Museum of Yale College. Trans. Connect Acad., vol. i., 1869, pp. 247-596. 54. 1870. Duchassaing, P., Eevue des Zoophytes, et des Spongiaires des Antilles, 8vo ; Paris, 1870. 55. 1871. Lutken, C. F., Antipathes arctica, en ny Sortkoral fra Polarhavet. Oversigt o. d. K. D. Vid. Selsk. Forhandl., 1871, pp. 18-26, 3 figs.; Ann. and Mag. Nat. Hist, ser. 4, vol. x. pp. 77-83, and figs. 56. 1871. Pourtales, L. F. de, Catalogue of the Museum of Comparative Zoology at Harvard College, No. iv. Deep-Sea Corals; Cambr., Mass., 1871. 57. 1872. Verrill, A. E., Eevised List of Stony Corals and Antipathacea described in Dana's Eeport on Zoophytes (Appendix to Dana's " Coral Reefs "). 58. 1874. Pourtales, L. F. de, Zoological Eesults of the "Hassler" Expedition (Corals). Cat. Mus. Comp. Zool., pt. viii. Dcep-Sea Corals, pp. 33-50, and 4 pis. 59. 1875. Haeckel, E., Arabische Ivorallen; Berlin, 1875-76. 60. 1877. Kldnzinger, C. B., Die Korallthiere des rothen Meeres, Abth. i., Alcyonarien u. Malakodermen, 98 pp. and 8 pis. ; Berlin, 1877. REPORT ON THE ANTIPATHARIA. 3 61. 1877. Nardo, G., SuLYAntipathe deH'Adriatico. Atti 1st. Ven., ser. 5, t. iii. pp. 673-678. 62. 1878. Koch, G. v., Mittheilungen iiber Ccelenteraten ; Zur Pliylogenie der Antipatharia. Morphol. Jahrb., Bd. iv., SuppL, pp. 74-86, and 1 pi. 63. 1878. Koch, G. v., Anatomie v. Isis Neapolitana, n. sp.; Ibid., pp. 112-127, and 1 pi. 64. 1878. Pourtales, L. F. de, Report on Corals of the " Blake " Expedition. Ball. Mus. Comp. Zobl., Harvard Coll., vol. v. pp. 197-212, and 1 pi. 65. 1879. Studer, T., "Gazelle" Korallen, Abth. II., Anthozoa polyactinia. Monatsber. d.k. -prmtss. Akad. d. Wiss. Berlin, 1878 (1879), pp. 524-550. 66. 1879. Agassiz, A., Letter (No. 3) to Carlisle P. Patterson (on the dredging operations of the "Blake"). Bull. Mus. Com}}. Zobl., Harvard Coll., vol. v. pp. 289-302. 67. 1879. Haackhs, W., Zur Blastologie der Korallen. Jenaisclte Zeitsehr., Bd. xiii. pp. 269-320, and 1 pL 68. 1879. Haacke, W., Ueher d. System u. Stammbaum d. Korallen Classen. Zool. Anzeiger, 1879, pp. 261-262. 69. 1880. Carter, H. J., Report on Specimens dredged up from the Gulf of Manaar. Ann. and Mag. Nat. Hist., ser. 5, vol. v. pp. 437-457, and 2 pis. 70. 1880. Carter, H. J., On the Antipatharia with reference to Hydradendrium spinosum. Ann. and Mag. Nat. Hist., ser. 5, vol. vi. pp. 301-305; also further remarks, Ibid., pp. 395-397. 71. 1880. Pourtales, L. F. de, Zoological results of the "Blake" Expedition to the Caribbean Sea, &c, pt. vi. Report on Corals and Antipatharia. Bull. Mus. Comp. Zobl., vol. vi. No. 4, Anti- patharia, pp. 113-118, 1 pi. 72. 1882. Hertwio, R., Report on the Actiniaria. Challenger Reports, Zoology, vol. vi. (Zool. Chall. Exp., pt. xv.). 73. 1884. Andres, A., Fauna u. Flora d. Golfes v. Neapel, Monogr. ix., Die Actinien; Leipzig, 1884. 74. 1884. Carus, J. V., Prodromus Faunae Mediterranese, pt. i. Ccelenterata ; Stuttgart, 1884. 75. 1884. Keller, C, Abstammungsverhaltnisse der Pflanzenthiere. Kosmos, 1S84, pp. 120-132. 76. 1886. Koch, W., Neue Anthozoen a. d. Golf. v. Guinea, 36 pp., 5 pis; Marburg, 1886. 77. 1886. Krtjkenberg, C. F. W., Fortgesetzte Untersuchungen ii. d. Skeletine (Gorgonidse and Antipathes). Zeitsehr. f. Biol, Bd. xxii. pp. 241-260. 78. 1888. Agassiz, A., Three Cruises of the "Blake," 2 vols., Cambr., Mass., 18S8 (Bull. Mus. Comj,. Zool., vols. xiv. and xv.). 79. 1889. Brook, G., Preliminary Remarks on the Homologies of the Mesenteries in Antipatharia and other Anthozoa. Proc. Roy, Sot: Edin., vol. xvi. pp. 35-39. 80. 1889. Brook, G., On a New Type of Dimorphism found in certain Antipatharia. Proc. Roy. Soe. Edin., vol. xvi. pp. 78-83. CRITICAL REVIEW OF LITERATURE. Our knowledge of the Sclerobasic or Black Corals may be said to date from the publication of the Elenchus Zoophytorum of Pallas in 1776, previous authors, including Linnaeus, not having recognised their essential difference from the Gorgonidae. Several species were already known and figured in the works of Rumphius, Seba, Marsigli, and others, but these, with the exception of two or perhaps three, first received a place in the binomial system at the hands of Pallas. Two forms are included in the 10th edition of the Systema naturae under the genus Gorgonia, viz., Gorgonia abies, Linnasus, and Gorgonia spiralis, Linnaeus, whilst Gorgonia senea, Linnaeus, is perhaps synonymous with the former species. The work of Pallas is remarkable for its clearness and precision; in it we first find clearly stated those characters which still are of ordinal value amongst the Antipathidae. It is true that the chief characters of his new genus Antipatlies, new rather by definition than in name, rested on points of skeletal structure, and little was known at that time, nor indeed until quite recently, about the structure and organisation of the polyps. Nevertheless the words of Pallas — " Stirps cornea; extus scabra, attenuata; cortice gelatinoso," define the Antipathidae as clearly to-day as they did a century ago, and separate them from all other known Zoantharia. This is all the more remarkable when it is remembered how much the horny skeleton of Hydroids and Actinozoa is subject to variations within limited groups; how even species of the same genus may differ in this respect. Yet with regard to the Antipathidae the structure of the skeleton alone is sufficient to separate them from other forms. Had subsequent investi- gators been more careful in following the characters laid down by Pallas, much confusion and error might have been avoided. The possession of a spinose horny axis has proved to be such an essential character of all Antipathidae, that those which have been described as smooth have either been erroneously so described, or do not belong to the family at all. At a time when nothing was known of the structure of the polyps, it was only natural that the species of Antipathes should be regarded as closely allied to Gorgonia, to which indeed their dendritic sclerobasic axis seems very closely allied. The similarity 6 THE VOYAGE OF H.M.S. CHALLENGER. between the mode of growth of those fan-like forms included in Milne-Edwards' division Rhipidipaihes and some species of Rhipidogorgia, Echinogorgia, &c. , is most marked, and has led Esper and others into numerous errors of identification. In the absence of polyps, the nature of the ccenenchyma and the presence of spines form the only reliable characters at the present day by which the Antipathidae may be identified. Ellis and Solander in speaking of Antipathes cupressina, Pallas ( = Gorgonia abies, Linnaeus), put the matter very clearly, so far as the differences were known at the time. They say, "Linnaeus has classed this species under Gorgonia, to which it is very nearly allied ; but the flesh of this tribe is so remarkably gelatinous and the whole bone or hard part is so covered with sjiines, which even are to be distinguished in the inner lamina?, that there is sufficient reason for making of it another genus." There is one point in which Pallas was mistaken, viz., the nature of the ovaries. These he regarded as external chitinous bodies, not always present, which are nowT known to be parasitic structures. He says : — " Ovaria,, calyces cornea stirpi insidentes, subturbinati," and thought the presence of these calyces to indicate an affinity between Antipathes and Sertularia ; with this exception, however, the characters of his genus Antipathes are remarkably clear and accurate. Pallas (17) describes ten species in all, one of which, Antipathes orichalcea, as he himself suspected, does not belong to the group. Two species are from the Mediterranean, viz., Antipathes fcenicidacea and Antipathes dichotoma, the latter described from- a species figured by Marsigli, and of which Pallas had not seen a specimen. The other seven are all from the Indian Ocean, a general term of which it is at present difficult to define the limits. Antipathes cupressina is evidently the Gorgonia abies of Linnaeus, and Antipathes spiralis is the Gorgonia spiralis of the 10th edition of Linnaeus' Systema naturae, the Gorgonia abies, var. spiralis, of the 12th edition. The five remaining species, viz., Antipathes ericoides, Antipathes pennaeea, Antipathes myriophylla, Antipathes fiabellum, and Antipathes clathrata, appear to be chiefly founded on types described and often figured in the Herbarium Amboinense of Rumphius. These have not all been identified by subsequent authors, but all apparently conform to the ordinal characters. Twenty years later Ellis and Solander (19) described six species of Antipathes, three already recorded by Pallas, and three new ones, viz., Antipathes ulex from Batavia, Antipathes subpinnata from the Mediterranean, and Antipathes alopecuroides from South Carolina. The latter species, probably owing to imperfect definition and the absence of a figure, has not since been identified, although there appears every probability that it must have been met with in one or other of the American Exploring Expeditions. The same authors give a figure of the polyp of Antipathes spiralis, Pallas, or at least of the tentacles and oral cone, the basal portion of the polyp being omitted. The polyp was REPORT ON THE ANTIPATHARIA. 7 shown to have six tentacles arranged in a radiate manner around the mouth. The drawings were made from a dried specimen which was first soaked for some time in water, and imperfect though they are, form the only drawings of the polyp of this species with which I am acquainted. A form described under the same name by Pourtales is, as will be shown later, generically distinct. The drawings of Ellis brought out for the first time an important point of difference between Antipathes and Gorgonia, namely, that the polyps of the former have only six instead of eight tentacles. More recently this numerical difference has been shown to be accompanied by important structural differences, but until within the last few years the exact bearing of these points on the systematic position of Antipathes has not been understood. Next in point of time follows Esper (21), who in his beautiful work Die Pflanzen- thiere described aud figured ten species. This author's descriptions, though long, are often indefinite, but as a rule his figures are good. Three of the species described appear new to science, viz., Antipathes larix from the Mediterranean, and Antipathes virgata and Antipathes reticulata, probably from the East Indies. Esper's species Antipathes glaberrima is the Savaglia of Donati and the Italians, and probably forms a considerable part of the " Black Coral " of commerce. In three cases where Esper thought to have obtained species described by Pallas, viz., Antipathes famiculacea, Antipathes flabellum, and Antipathes clathrata, he describes and figures specimens of decorticated Gorgonidas and not the true Antipathes. All three forms described by Pallas have a spinose sclerenchyma, whereas those described by Esper are all smooth. The same remark applies to his new species Antipathes ligulata, which has a smooth axis, and, as first suggested by Dana, is probably a decorticated Gorgonid. Antipathes compressa, Esper, is founded on the base of some large species. Dana says that Esper's figure agrees with the base of his Antipathes arborea, whilst Gray suggests that Esper's species may be the base of Antipathes myriophylla. In any case the name should be dropped, having no specific value, and its retention only adds to the confusion of the croup. Esper does not describe Antipathes ericoides, but gives a figure of it, and remarks that there are many forms allied to Antipathes myriophylla, Pallas, of which Antipathes ulex, Ellis and Solander, is one, and Antipathes ericoides, Pallas, another. The latter, however, does not seem so closely related to Antipathes myriophylla as Esper would have us suppose. Finally his species Antipathes paniculata appears to be founded on a variety of Gorgonia abies, Linnseus {Antipathes cupressina, Pallas), as was first sug- gested by Lamarck. Dana, however, points out that it differs in the relative development of the lateral branches. There is a fine specimen of this form in the British Museum, which seems to differ from Antipathes abies (Linn.) Gray, in possessing stronger lateral paniculate branches, but in other respects agrees with the earlier type ; thus, at most, it can only be regarded as a variety. Bruo-uiere (22) in 1792 gave a synopsis of the species already known, and described 8 THE VOYAGE OF H.M.S. CHALLENGER. a new one, Antipathes triquetra, from Manila. Unfortunately this species is based on specimens which were too imperfect to give specific characters, though they certainly appear to have differed from the species described by Pallas. The species has not been identified by subsequent authors, and is omitted from the work of Milne-Edwards. Its thick triangular stem, with the angles twisted spirally, should make this form easily recognised, even in the absence of particulars of the mode of branching ; I am in the meantime, however, obliged to include it amongst the species dubise. In 1816 Lamarck (23) described six species of Antipathes, but appears to have been particularly unfortunate in his identifications. His Antipathes mimosella is probably the same as Antipathes ulex, Ellis and Solander, and the latter name has priority. On his own admission, Lamarck's Antipathes scoparia is synonymous with Esper's Antipathes vifgata, whilst Antipathes pyramidata is not an Antipathes at all, having a smooth and somewhat vitreous axis, and probably comes under Verrill's genus Iridogorgia. A fine specimen in the British Museum collection is labelled Iridogorgia pyramidata. Lamarck gave the name Antipathes radians to Esper's Antipathes fceniculacea, which he showed to be different from the true Antipathes fceniculacea, Pallas ; the species as already stated does not belong to this order. Antipathes corticata, Lamarck, is a distinct form figured by Haeckel in his Arabische Korallen, but neither author gives us a detailed description. The remaining species, viz., Antipathes lacerata and Antipathes pectinata, may be distinct also, but Lamarck's descriptions are very unsatisfactory, and neither form has been identified by subsequent investigators, so that for the present both must be included amongst the species dubise, the definitions being insufficient for identification. Lamouroux (24) next added two species to the list, both of which have been accepted by subsequent investigators. It appears probable, however, that his Antipathes pinnatifida, which has since been observed by Studer amongst the Corals of the " Gazelle " Expedition, may prove to be a variety of Antipathes ulex, Ellis and Solander. There appears to be very great variation amongst specimens of the Antipathes myrio- phylla and Antipathes ulex type, and at present it seems difficult to distinguish between those points which are of specific value, and others which only represent individual varia- tions. Owing to the limited number of specimens which I have been able to compare, it has been impossible to decide with certainty, but seeing that no two specimens of this type appear alike, I have preferred temporarily to consider all varieties of one form, for which it seems necessary to retain the name Antipathes ulex, Ellis and Solander. In the description of Antipathes boscii, Lamouroux makes no reference to the occurrence of spines, a fact which may perhaps have led Gray to include this species in his genus Leiopathes. Verrill has more recently described a specimen, which he considers referable to this species ; it was obtained by Agassiz off South Carolina, the original habitat. In this specimen the spines are apparently well marked on all parts of the sclerobasic REPORT ON THE ANTIPATHARIA. 9 axis. A small specimen in the University Museum at Copenhagen, which was received through the Museum of Comparative Zoology, agrees with Verrill's description, and is the one on which my description has been based. Lamouroux's description is too indefinite to enable one to decide with certainty whether this specimen agrees with his type. Lamouroux's Exposition Methodique (25) contains no new information regarding the Antipathidse, and so far as that section of the work is concerned, simply reproduces the descriptions and figures from Ellis and Solander's work. In 1824 he published in the Encyclopedie Methodique (26) a synopsis of the forms already known. Twenty-six species are described, of which one is new to science. This (Antipathes eupteridea) has since been identified by Pourtales from Martinique, the original locality, and judging from the mode of branching, may possibly prove, when the polyps are better known, to belong to the subfamily Schizopathinge. At any rate it appears to have a type of branching which so far as is known at present is not found amongst the Antipathinse. The same work gives a rather fuller description of Antipathes boscii, but again no reference is made to the spines. Lamouroux mentions that he has received fragments of this species from the He cle Re, off the West Coast of France. Risso (27), two years later, only recorded one species of Antipathes {A. larix, Esper) as the result of his researches on the Mediterranean shores of France, but another form which he named Eunicea Antipathes may also have belonged to the group, though it appears impossible to decide at present. In 1832 Gray (28) contributed a note on the Animal of Antipathes, the purport of which will be discussed in connection with a later paper of his on the same subject. Blainville (31), in his Manuel dActinologie (1834-1837), instituted a new genus, Cirrhipathes, for the reception of Antipathes spiralis, Pallas, and another simple form which he named Cirrhipathes Sieboldi ( = Palmijuncus anguinus, Rumphius?), but of which he gives no description. This genus was proposed by Blainville in consequence of Ellis' observations on the form of the polyp of Antipathes spiralis, Pallas, and for the reception of forms which, like it, have an unbranched sclerobasic axis. The systematic value of this genus as defined by Blainville will be considered later when discussing the merits of the classification adopted by Milne-Edwards. In 1842 Gray (32) proposed to separate those species of Antipathidse having a smooth sclerobasis, and include them in a new genus Leiopathes. The name occurs in the 1842 edition of the Synopsis of the British Museum, but he does not appear to have defined the genus until the publication (40) of his Revision of Axiferous Zoophytes in 1857. In addition to the smooth nature of the axis, Gray calls attention to the presence of spicules in the coenenchyma, a character which he considered linked this genus more closely to the Gorgoniclae. I propose to refer to the subject again when considering the arrangement suggested by Mdne-Edwards. (ZOOL. CHALL. EXP. — PART LXXX. — 1889.) ^^ 2 10 THE VOYAGE OF H.M.S. CHALLENGER. Dana, in his classic work on the Zoophytes (34), gives a short resume of the species already described by Pallas, Esper, Lamarck, and others, and describes and figures two new species. One of these, Cirrhipathes anguina (Cirripathes Sieboldi, Blainv. ?), he regarded as the Palmijuncus anguinus of Rumphius, and adopts Blainville's name as a synonym. The other form, Antipathes arborea, is very closely allied to Antipathes dichotoma, Pallas. Dana was the first to recognise the true relationship of the Antipathidee and their close affinity to the Actiniaria. On page 574 of his work he says : — The Antipathidaes " like the Gorgonidae secrete a corneous axis, but are placed amongst the Actinoidea as the tentacles have the naked character peculiar to this suborder, and the polyps closely resemble those of Madreporse in appearance and habit. The existence of genital lamellae within the visceral cavity is not yet proved ; as this is the deciding character, the propriety of the present arrangement cannot be considered fully established." Dana's work contains figures of both his species with the polyps in situ. These bring out several new points tending to remove the Antipathidae still further from the Gorgonidae. In the first place, his figures show the undoubted naked character of the tentacles, a feature which may have been presumed from Ellis' drawing of the polyp of Antipathes spiralis, Pallas, but which he now placed beyond doubt. It was clear from his figures that the tentacles of Antipathes, as well as those of Cirrhipathes, are not allied either in number or in form to those of the Gorgonidae. A further point which seems to have escaped comment, but which, nevertheless, is of considerable importance, is that Dana's figures first brought out a difference in tbe mode of arrangement of the individual zooids on the axis. In his unbranched species, Cirrhipathes anguina, the axis is comparatively stout, and the zooids are distributed all around the stem as in Juncella and many Gorgonidae. In the branched form Antipathes arborea, on the other hand, the branches and branchlets are relatively slender, and the zooids are distributed in single longitudinal series, usually with their oral surfaces all turned in one direction. In short, Dana's figures are the first, and' for a considerable time remained the only ones, which gave any adequate idea of the appearance of a living colony of an Antipatharian. In 1849 Jules Haime (35), the colleague of Milne-Edwards, described, under the name of Leiopathes lamarcM, a form which had previously been confused with Leiopathes glaherrima (Esper). Gray's second note on the Animal and Bark of Antipathes (38) appeared in 1857. In the earlier one already referred to he had described the appearance of the polyps of a form which he believed to be identical with Antipathes dichotoma, Pallas. His specimen was sent from Madeira in a dry state, and showed on the minute branches, at irregular intervals, a number of red pellucid tubercles. These on maceration in water proved to be the polyps, provided with six tentacles, but in other respects supposed to agree with those of Gorgonia. He also stated that "minute, pellucid, oval bodies, which REPORT ON THE ANTIPATHARIA. 11 are perhaps similar to the irregular papillary spicules fouud in the bark of Gorgonia, are scattered through the bark of this species of Antipathes, and the axes of its smaller branches are minutely tubular." Gray's observations on the polyp are no advance on the information already supplied by Ellis, excepting that he doubts the existence, in the species studied, of the cup-like oral aperture with a crenate margin figured by that author. I am, however, induced to discuss them at greater length on account of the questionable identity of the species referred to, and also because this form possibly constituted the type of his new genus Leiopathes. In his second paper Gray states that the species formerly observed " has been separated from the others of the genus because the surface of the axis is smooth and not covered with a number of minute, uniform cylindrical spines like the true Antipathes, and has been called for that reason Leiopathes" evidently referring to his note of 1842 already mentioned. He then goes on to describe the appearance of the " bark " of a long simple-stemmed Antipathes from the Seychelles, which he regarded as a new species allied to Antipathes spiralis, Pallas, " if more than a very fine straight specimen of that species." The ccenenchyma is stated to contain flakes of a substance insoluble in strong hydrochloric acid or caustic potash, and supposed to be siliceous. This paper is illustrated by a plate, from which I have been enabled to identify the specimen now in the British Museum collection. It apparently belongs to Cirrhipathes anguina, Dana, and although dry, shows the same arrangement of polyps (and spines ?) as figured by that author. The identity of the species referred to in the earlier paper is not so certain. Undoubtedly it is not Antipathes dichotoma, Pallas, as Marsigli, from whose work Pallas took his description, not only notes the presence of spines, but figures their arrangement both near the base of the stem and on a more slender pinnule. I have been unable to find any specimen of Antipathes dichotoma, Pallas, in the British Museum collection, but am disposed to think that a specimen of Antipathes glaberrima, Esper, the locality for which is not stated, may be the species referred to by Gray. In this species the axis is perfectly smooth and glossy in the older portions of the colony, and the ccenenchyma has been stated sometimes to contain, or rather have adhering to it, the spicules of various Axifera, sponges, &c. Later in the year Gray (40) contributed to the same journal a, Synopsis of Axiferous Zoophytes, in which he included the Antipathidse. He divides the axiferous zoophytes into three suborders, in the third of which, Ceratophyta, the Antipathidse form the first family. He describes three genera, Leiopathes, Antipathes (with a subgenus, Cirrhi- pathes), and Sarcogorgia. Under the genus Leiopathes he includes two species, viz., Antipathes glaberrima, Esper, and Antipathes boscii, Lamouroux. The species of which he described the polyps in 1832 and then named Antipathes dichotoma, Pallas, he evidently now regards as Antipathes glaberrima, Esper, and queries the two as synonymous, and in referring to his original note quotes the name glaberrima instead of dichotoma. 12 THE VOYAGE OF H.M.S. CHALLENGER. He makes no comment on the change, which is to be regretted, because it is still doubtful what species he selected as the type of his genus Leiopathes. Subsequent authors have accepted Antipatkes glaberrima, Esper, as the type, and this view seems probable. I am unable to understand on what grounds he included Antipathes boscii in the genus Leiopathes, the essential character of which is that the axis is smooth instead of spinose. Presumably Gray had seen or thought he had seen a specimen of the species, or it should have been included in the special list, given at the end, of those species which had not come under his notice. It seems highly probable that his words, " Lamouroux's figures represent the bark forming small masses between the branchlets, as I have observed it on the Madeira specimen," refer to the supposed Antipathes dichotoma, Pallas, rather than to any specimen of the true Antipathes boscii which had come under his uotice. Under the genus Antipathes, Gray describes eight new species, one (Antipathes gracilis) a simple form belonging to the subgenus Cirrhipathes, all the others being more or less branched. All his diagnoses are very imperfect, and in nearly every case insufficient for identification. Fortunately the types of these forms, nearly all of which appear distinct, are in the collection of the British Museum, and have thus been accessible for reference. In the section devoted to the description of species, I have added to Gray's original diagnoses such particulars as seem of use in identification. Under the name Antipathes spinescens he appears to have included two forms of the " bottle brush ".type, which differ considerably in general appearance and may be specifically distinct. For the type not conforming to Gray's description I have suggested the name Antipathes spinescens, var. minor. A species from Madeira which is referred to Antipathes subpinnata, Ellis and Solander, certainly does not belong to that species. Gray seems to have been doubtful on the subject, and admits that his specimen does not agree with Ellis and Solander's figure, and adds, " I had originally described it as distinct under the name of Antipathes Wollastoni." I have been unable to find any printed description of Antipathes wollastoni, and it does not appear to be described in any of his most numerous papers, a list of which is kept at the British Museum. It may be, however, that Gray only means that he had prepared a description, but that before printing it in the Synopsis under consideration, he decided that it was not a new form. In any case the specimen still bears the name Antipathes Wollastoni in Gray's handwriting, and as the species in question is quite distinct from the type of Ellis, I have retained that name for it. Five other species observed by Gray are referred to : Antipathes tdex, Ellis and Solander, Antipathes myriophylla, Ellis and Solander, Antipathes abies (Linn.), Gray, Antipathes larix, Esper, and Antipathes reticulata, Esper, respectively. The first four are probably correctly identified, but the last named does not agree with Esper's figure and description. It is very slender, lacks the characteristic short stiff secondary pinnules, and the reticulum is not constructed on the same plan. The apical portions of the colony are free, and the REPORT ON THE ANTIPATHARIA. 13 spines are not nearly so prominent as in Esper's species. The specimen in question forms part of a collection from the West Indies purchased from Scrivener. Another species described from the same collection was named Antipathes atlantica by Gray. I find no specimen bearing that name in the collection, but several agree with Gray's characters. All approach his supposed Antipathes reticulata, Esper, in form, and a careful comparison has led me to believe that all belong to one species, though there is considerable difference in the relative thickness of the branches. I have therefore retained the name Antipathes atlantica, Gray, for these specimens, and include Antipathes reticulata, Gray, non Esper, as a synonym. I find no specimen in the British Museum collection bearing the name Antipathes pluma, Gray. There are, however, two or three specimens which agree with Gray's characters, one without locality, one from St. Helena, and one more recently received from the East Indies. As the result of a comparison of these specimens with fragments of Antipathes pennacea, Pallas, from the Paris Museum, I am led to suppose the two forms to be identical. There is, however, a considerable variation in the length of the pinnules in different specimens ; in some, certain pinnules become elongate and pinnate, whilst in others all remain simple. There appears, as far as I could ascertain, no sufficient variation in the spines to afford constant characters. I have therefore regarded Gray's Antipathes plmna as synonymous with Antipathes pennacea, Pallas. The genus Sarcogorgia must have been included in the Antipathidse by an oversight, unless Gray regarded Antipathes and Gorgonia as members of one order, which appears possible. Gray's original description of Sarcogorgia phidippus occurs in the Proceedings of the Zoological Society for 1857. Professor E. P. Wright, to whom I have referred in the matter, informs me that he has not seen Gray's specimen, but, judging from the figure, he is of opinion that it may be identical with Spongioderma verrucosa (Kolliker), one of the Briareidas. Gray points out that " some of the smooth species referred to the genus Antipathes by Esper, as Antipathes foeniculacea, Antipathes clathrata, and Antipathes ligulata, are evidently the axes of some species of Gorgoniadse that have lost their bark ; " he should also have included Antipathes flabellum, Esper, in the same category. His definition of the genus Antipathes is based on an examination of dry specimens, and it is evident that he failed to grasp the true generic, or, as one might now call them, ordinal characters. He says : — " Bark fleshy, with imbedded, large and small brown (siliceous) plates, easily deciduous. Axis simple or branched, horny, covered with numerous close-set, sub-cylindrical spines." Gray evidently regarded the genus Anti- pathes as closely allied to Gorgonia, and as such, probably possessing siliceous plates or spicules within the ccenenchyma. He combats Dana's view that Antipathes is nearly allied to the Actiniaria, basing his opposition on his studies of the dried polyps of Leiopathes glaberrima and Cirrhipathes anguina. He adds, " I am aware that the tentacles do not appear to be pinnated, when they are examined after they have been 14 THE VOYAGE OF H.M.S. CHALLENGER. dried, but this is the case with the animals of all the Gorgoniadse I have examined under similar conditions. The pinnae appear to be permanently withdrawn under such circumstances." The first volume of Milne-Edwards' Histoire Naturelle des Coralliaires (41), contain- ing an account of the Antipatharia, appeared in the autumn of 1857, so that the author could not have seen Gray's papers on the subject communicated to the Zoological Society in the same year. I am not aware which was published first, but think it probable that this volume may have been issued before the publication of Gray's Synopsis, which occupies the last pages of the 1857 volume. In any case it is unfortunate that two important and independent revisions of the order should have been published in the same year, and that no attempt has since been made to bring their results into harmony. Milne-Edwards, acting on the suggestions of Dana, forms a new suborder of the Zoantharia, Zoantharia sclerobasica or Antipatharia,, for this group. He is of opinion that the spinose nature of the axis insisted on by Pallas and others as an essential character, though usual, is not universal, and cannot be regarded as of ordinal value. Milne-Edwards adopts the genera Cirrhipathes, Blainville, and Leiopathes, Gray, and proposes to further subdivide the original genus of Pallas, by the establishment of three new genera. A glance at the subjoined table will show the arrangement proposed : — simple — rod-like, ...... Cirrhipathes, Blainville. Sclerobasic axis - branched and - \ chitinous; surface of large "branches ' free. rough ; branchlets smooth ; ccenenchyma containing siliceous elements, Antipathes (Pall.) emend. coalescent, ( tufts, Arachnopathes, n. gen. and disposed <. in form of (a fan, Rhipidipathes, n. gen. Leiopathes, Gray. vitreous, . II yalopathes, n. gen. Milne-Edwards admits that with the little knowledge then available of the morphology of the Antipathidse, it would be difficult to establish natural genera with certainty, and the arrangement proposed in the above table is given mainly with a view to aid in the determination of species. It will be seen that the generic name Antipathes is retained in a modified sense. Those forms having, or supposed to have, a vitreous sclerenchyma are separated under the name Hyalopathes ; those having a chitinous axis and showing more or less complete fusions amongst the branches are allocated to two new genera. Arachnopathes includes those forms having the branchlets more or less collected into tufts, and Rhipidijxtthes those which, like A. flabellum, have a fan-like growth, and have the branches and branchlets confluent. Thus those species which remain to constitute the genus Antipathes, in sensu Milne- Edwards, may be shortly REPORT ON THE ANTIPATHARIA. 15 described as branched Antipathidse having a spinose chitinous axis, presenting no fusions between any parts of the corallum. No new species are recorded, and those previously described are allocated to one or other of the above genera, as best could be done from the descriptions available. With the exception of Duchassaing and Michelotti, subsequent investigators have not followed Milne-Edwards in his subdivision of the genus Antipathes, partly on account of the admittedly unsatisfactory state of our knowledge of the group, and partly owing to the fact that it has been generally felt that genera founded on the structure and arrange- ment of the axis alone could have little value in a natural system. Pourtales especially has preferred, pending a fuller study of the polyps themselves, to regard the Antipathidse as consisting of only a single genus, Antipathes, those species having a simple axis being included in a subgenus, Cirrhipathes. As no new genera have been proposed since the publication of Milne-Edwards' work, it may be well to consider here the systematic value of the genera therein defined, in the light of the information brought forward in the present monograph. I do not, of course, for a moment suppose that the information which has been obtained concerning the structure of the zooids of some twenty species will prove sufficient for an adequate classification of the group, particularly when it is remembered that we have absolutely no information on the structure of the zooids in the majority of the species described. Still I believe it to be sufficiently complete to be of service in the present instance. In the first place it may be stated generally that although in those forms studied, a similarity in the structure of the zooids is frequently associated with a similarity in the type of branching of the corallum, this is by no means always the case. With regard to the more or less frequent occurrence of fusions between different portions of the corallum, a point which will be discussed in detail later, it may here be stated that the evidence available at present appears to show that such fusions are not constant in all the species of ti genus, and are therefore of no generic value. Indeed amongst the species referable to the Rhipidipathes type of Milne-Edwards, there are at least three well-marked types of polyp. In other cases where the fusions are slight, and more accurately defined as adherences, the feature is more or less accidental, and probably not even of specific value. Turning now to the genera adopted by Milne-Edwards, it will be well to consider first those described by Blainville and Gray. Cirrhipathes, Blainville. — This genus was constituted by Blainville in consequence of Ellis' observations on the polyp of Antipathes spiralis, Pallas, and for the reception of those forms in which the sclerobasic axis remains simple. At the time Ellis' figures of the polyps of Antipathes spiralis were the only reliable ones extant for any species of Antipathes, and further investigation has shown that the form of polyp assigned to this genus by Blainville is, so far as his definition goes, not confined to it, but is shared by 16 THE VOYAGE OF H.M.S. CHALLENGER. branched as well as unbranched types. Thus, as Milne-Edwards points out, the simple elongate character of the axis is the only feature, in the absence of a fuller knowledge of the polyps, which separates this genus from the other Antipathidas. This is the character which has been regarded as generic by subsequent investigators. Antipathes spiralis, Pallas, constitutes the type of the genus, but unfortunately we have at present no certain knowledge of its polyps, excepting such as may be surmised from the drawings in Ellis and Solander's Zoophytes. These, which only include the mouth and tentacles, appear to represent a rounded polyp with the tentacles arranged in a radiate manner — the form of polyp indeed which one has been accustomed to regard as typical of the Antipathidae. Whether the polyps were arranged in a single row along the axis as in typical Antipathes, or all around the axis as in Cirrhipathes anguina, Dana, is uncertain, as Ellis gives no information on the point. An especially interesting feature of the drawings, and one which has given rise to frequent comment, is the curious cup-shaped mouth with a crenate margin, a form of oral aperture which does not appear to be shared by the species subse- quently studied by Lacaze Duthiers and G. v. Koch. From a study of allied forms I am inclined to believe that this is a natural feature of the species, somewhat exaggerated, and not an altogether artificial appearance, as some have supposed. At any rate, in Cirrhipathes propinqua we have a type of oral cone, which with a little exaggeration (possibly in Ellis' case due to maceration of previously dried specimens) would agree fairly well with the drawings referred to. Pourtales (71) in 1880 described and figured a species which he regarded as possibly identical with Antipathes spiralis, Pallas — a form which he had previously looked upon as a spiral variety of Cirrhipathes deshonni, Duchassaing and Michelotti. The polyps as described and figured by Pourtales are quite unlike those of any species known at the time. The tentacles are long, fleshy, finger-like processes which do not usually shrink much in spirit and are evidently non-retractile. The polyps appear alternately large and small, and are arranged on one side of the stem only. By a comparison of the drawings and description of this form with specimens of Cirrhipathes spiralis from the East Indies (the original habitat), I have convinced myself that, irrespective of the structure of the polyps, the two forms are distinct. Pourtales indeed was doubtful of their identity, but had no means of comparison at the time. The only other species of Cirrhipathes previously described of which any account is given of the polyps is the Cirrhipathes anguina of Dana, a form which he regarded as probably identical with Palmijuncus anguinus, Rumphius {Cirrhip>athes Sieboldi, Blain- ville). This is a species having rounded polyps with radiately arranged tentacles. The polyps are not alternately large and small as in Pourtales' species, but subequal and disposed all around the axis instead of in linear series. Haeckel has since figured a .similar arrangement in Antipathes corticata, Lamarck. Between Dana's type and that of Pourtales there is a marked difference both in structure and arrangement — a difference prob- ably sufficient to be of generic value. The question now arises, does Antipathes spiralis, REPORT ON THE ANTIPATHARIA. 17 Pallas, belong to either of these types, and if so to which 1 Ellis' figure evidently agrees more closely with Cirrhipathes anguina than with Antipathes spiralis, Pourtales, the chief difference to be noticed being in the shape of the mouth. In Oirrhipathes spiralis (Pallas), the mouth is represented as widely open, and consisting of a cup-like portion with a crenate margin, whereas in Cirrhipathes anguina, Dana, the figure shows the mouth contracted and the oral surface only slightly elevated. As previously stated, a new form (Cirrhipathes propinqua), which I have been enabled to study, agrees fairly well with the figure in Ellis and Solander's work, and in this species the polyps are distributed all around the stem. The other species included in this genus, those described by Gray, and Duchassaing and Michelotti, have been defined from a study of dry specimens, so that no information is available as to the structure of their polyps. From this general survey of the question, however, it will be seen that Blainville's chief character for his new genus, the simple filiform character of the corallum, has no generic value, as species having widely different types of polyp may agree in this respect. In subdividing the genus it becomes necessary to retain the name Cirrhipathes for that section of it to which the type species, Cirrhipathes spiralis (Pallas), belongs. Unfortunately it is impossible to decide the question with certainty at present. The species in question is not found amongst the material collected during the Challenger Expedition, and none of the specimens of it which I have seen are preserved in spirit. From a study of the British Museum collection I have, however, been able to throw a certain amount of light on the question, which, if not sufficient for absolute accuracy, lends a considerable air of probability to the arrangement here adopted. I find that amongst the spiral species in that collection there are one or two specimens in which the polyps are imperfectly preserved. In these cases the zooids appear as rounded elevations on the ccenenchyma, and are distinctly visible all around the stem. The specimens, although more slender than that figured by Ellis, appear to agree with the description, and the fact that they came from the Indian Ocean, makes it still more probable that these are really the Cirrhipathes spiralis (Pallas). I have thus been led to include Cirrhipathes propinqua, n. sp., Cirrhipathes anguina, Dana, and Cirrhipathes spiralis (Pallas), in the modified genus Cirrhipathes, and to include Antipathes spiralis, Pourtales, and one or two other forms in a new genus, for which the name Stichopathes is proposed. Those species of which I haA*e been unable to obtain any information regarding the type, and mode of distribution of the polyps, can only be assigned a proper place when the necessary information is obtained. Leiopathes. — Gray's definition of this genus (40) is as follows : — "Axis smooth, polished, branched, forked. Bark soft, deciduous, deliquescent, sometimes forming (when dry) smooth transparent masses at the fork of the branchlets. " This definition differs in two points from his earlier notes on the type species (28). He omits all reference to the presence of spicules in the ccenenchyma, though whether intentionally or by accident (ZOOL. CHALL. EXP. PAKT LXXX. 1889.) Llll 3 18 THE VOYAGE OF H.M.S. CHALLENGER. I am unable to say. His published papers certainly lay stress on the fact that the ccenenchyrna of Antipathes glaberrima, Esper, which has generally been accepted as his type, contains spicules, and it was not until some years later that Lacaze Duthiers pointed out that these are in reality foreign to it. Milne-Edwards includes this character in the definition given in his work. Gray on the other hand gives in his later definition a character not included in the original, namely, that the " bark " when dry sometimes forms smooth transparent masses at the forks of the branches. In any case this is of minor importance, but so far as I can ascertain, it is characteristic rather of Antipathes boscii, Lamarck, than of Leiopathes glaberrima (Esper). Gray includes the former species in his new genus, whilst Milne-Edwards does not. I have not seen any specimen in the British Museum which appears referable to Antipathes boscii, and for the reason already given I have followed Verrill in regarding it as a spinose species. Milne-Edwards includes three species, viz. : — Leiopathes lamarcki, Haime, Leiopathes glaberrima (Esper), and Leiopathes compressa (Esper) ; that is to say, all those forms which have been described as possessing a smooth sclerenchyma, but in other respects agreeing with other Antipathidge. Leiopathes lamarcki, Haime, is the Savaglia of Douati and the Italians, and is the species which Lacaze Duthiers (44) has shown to differ so essentially from the Antipathidas that it has been necessary to establish a new family, Savagliidge (Gerardidad), for its reception. Antipathes compressa, Esper, cannot be considered to rank as a species ; the type specimen only consisted of the base of some large form, which may or may not have been one of the Antipathidas. Dana suggests that the base of his Antipathes arborea agrees with the figure of Antipathes compressa, Esper, whdst Gray compares it to the base of Antipathes myriopliylla. I think, however, that in this reference Gray does not refer to Antip>athes myriopliylla, Pallas, but to a large virgate species in the British Museum, which also bears that name on the label. There is no resemblance to Antipathes compressa, Esper, in any of the specimens of Antipathes myriopliylla, Pallas, which have come under my notice. Finally, Antipathes glaberrima, Esper, in spite of its name, is not a smooth form, as was first shown by Lacaze Duthiers, and afterwards confirmed by Pourtales and others. The stem and main branches are smooth and polished, but the younger branchlets all bear distinct though somewhat distant spines. Thus both generic and specific names are misleading, and the genus Leiopathes, so far as Gray's definition goes, is not a good one. I have, however, been enabled to study the polyps of Antipathes glaberrima, Esper, a species which was included amongst the material kindly supplied to me from the Naples Zoological Station, and I find that this species differs structurally from any with which I am acquainted, and possesses characters sufficiently distinctive to demand its allocation in a separate genus. It thus becomes necessary either to retain the genus Leiopathes in an amended form or to establish a new one. As it is very desirable to avoid, wherever REPORT ON THE ANTIPATHARIA. 19 possible, the creation of synonyms I have decided to retain the generic name Leiopathes for this species, though I do so with considerable hesitation. In the first place the name has been associated only with such characters as are not of generic value, characters indeed which are deceptive or false. Then again, though it is generally accepted that Gray intended Antipathes glaberrima, Esper, as the type of his genus, we have no type specimen to refer to, and he is by no means clear on the subject. It is only because Antipathes glaberrima, Esper, has the greater portion of the axis smooth, and that its ccenenchyma has been found frequently to have spicules adhering to it, that the species can be made to agree with Gray's definition, and in this respect a dry specimen of Savaglia lamarchi would fulfil the conditions equally well. One result of this investigation is to bring out clearly the fact that so far as we know at present there are no species of true Antipathidse in which the sclerenchyma is entirely without spines, the only species in which the axis is smooth throughout being Savaglia lamarcki, and this on morpho- logical grounds has been removed to another family. Hyalopatlies.— -This genus was proposed by Milne-Edwards, who gives the following short definition :—" Axe sclerobasique rameux, lisse et d'un aspect vitreux." He was led to separate the forms included in this genus on account of the semi-hyaline aspect of the sclerenchyma, which he supposed to be associated with a difference in chemical composition. Three species are included in the genus, all described by Lamarck, viz.: — Hyalopatlies pyramidata, Hyalopatlies pectinata, and Hyalopatlies corticata. The first named, Milne-Edwards' type, has since proved not to belong to the Antipathidse, and is now arranged in Verrill's genus Iridogorgia as Iridogorgia pyramidata. I find a fine specimen in the British Museum collection. The sclerenchyma is semitransparent, smooth, and undoubtedly has a vitreous aspect. With regard to the second species, Hyalopatlies pectinata, Lamarck only gives a very imperfect description, and so far as I am aware, it has not been identified by subsequent investigators. I find nothing in his description to give one the idea that the axis differs from the typical chitinous one of Antipathes, indeed he does not mention the colour of the axis, as is the case in his description of Antipathes pyramidata, but Milne- Edwards may have examined the type specimen. Lamarck describes the spines as few, but if they are present at all the species may belong to the Antipathidre, although it would scarcely conform to the definition of the genus Hyalopatlies. With the little information obtainable at present, I have been compelled to include this form amongst the species dubiae. The third species, Hyalopatlies corticata, has since been observed by Haeckel, who gives a figure of a living colony in his Arabische Korallen. In the short description which Haeckel gives of this figure, he speaks of the sclerenchyma being black with a glassy aspect and as also being regularly spinose as in the Antipathidas generally. These researches, though probably not sufficient to enable us to assign to this species its natural 20 THE VOYAGE OF H.M.S. CHALLENGER. position, at any rate show that it cannot belong to the genus Hyalopathes as defined by Milne-Edwards. The axis is not smooth, neither is it seini-hyaline. With regard to the glassy appearance of the black sclerenchyma, it may be stated that this is a character of frequent occurrence amongst the Antipathidse, and one which has no generic value. The same appearance was noted by Marsigli in his description of the species which we now know as Antipathes dichotoma, Pallas. I have also observed it in several species of Cirrhipathes and other forms, and, so far as could be ascertained, this aspect is not constant in all the specimens of a species, but appears to depend to some extent on age. So far as can be ascertained from Haeckel's figure and description, this species differs from any other with which I am acquainted in the arrangement of the zooids on the corallum. In other branched forms the zooids are almost invariably arranged in a single linear series on the branches and branchlets. In Haeckel's figure the corallum is represented as very thick in proportion to its length, and the branches do not appear to taper much. The zooids are relatively small, radiate in outline, and not confined to one aspect of a branch. They are a considerable distance apart, and apparently distributed irregularly at any point around the axis. The thickness of the corallum and the shape and arrangement of the zooids recall the characters of the genus Cirripathes as modified in the present Report, but the corallum in this species is branched. On account of the fact that we are stUl ignorant of the structure of the zooids and the arrangement of the spines, I have temporardy included this species amongst a group of others which are too imperfectly known at present to have a definite position assigned to them. In accordance with this arrangement I propose to drop the generic name Hyalopathes, which should be regarded as a synonym of Iridogorgia, Verrill. Of the three species comprised in it, the type species is not an Antipatharian ; another which does not possess the generic characters is probably closely allied to Cirrhipathes, whilst we know too little of the third to assign it any definite position. The remaining species of the old genus Antipathes, Pallas, are divided by Mdne- Edwards into three genera, the characters of which depend solely on the mode of branching, and the presence or absence of fusions between adjoining parts of the corallum. He retains the name Antipathes for those forms which appear to be without fusions, and divides those in which fusions are frequent into two sections. The name Arachnopathes is suggested for those in which the corallum forms a more or less thick confused mass, the branchlets of which are distributed all around the branches, that of Rhipidipathes for the flattened fan-like forms, in which the extension of the colony takes place chiefly in one plane. It has been admitted on all sides that genera of Antipatharia based solely on skeletal structures must be more or less artificial, and Milne-Edwards himself anticipated that further researches would necessitate a modification in his arrangement. Unfortunately we do not yet possess all the information necessary for a proper elucidation of the REPORT ON THE ANTTPATHARIA. 21 group. It will, however, be interesting to ascertain what precise bearing the new facts here brought forward have on the arrangement of Milne-Edwards from two points of view. First it will be instructive to ascertain whether in the light of more recent researches it will be possible to retain the generic names given by Milne-Edwards, and secondly we may arrive at some conclusion as to how far a particular mode of branching appears to be ef generic value. Fur this purpose it appears more convenient to consider the more restricted genera first. Arachnopathes. — The following is Milne-Edwards' definition of his genus Arach- nopathes : — " Axe sclerobasique se divisant en une multitude de branches tres greles que se dirigent en divers sens et se soudent entre elles aux points de rencontre, de facon a constituer des reseaux dont la reunion forme une touffe arrondie. Tissu sclerobasique noir et opaque." He includes only two species, viz.: — Arachnopathes ericoides (Pallas), and Arachnopathes clathrata (Pallas). If I have been correct in my identification of the former species, its mode of branching is precisely that described by Milne-Edwards, and the whole corallum lacks the apparent flatness shown in Esper's plate. A small fragmentary specimen in the British Museum collection may be the Antipathes clathrata of Pallas, but if so does not show the marked difference in thickness between the branches and branchlets to which Milne-Edwards refers. This specimen agrees with the former in consisting of a thick dense mass of branchlets all fused into one firm network, but there is not the same marked spiral arrangement of the branchlets as in the former species. A third form here described as new {Arachno- pathes aculeata) has precisely the same thick matted corallum as the two former species, but in this case the branchlets, although frequently collected into groups, are chiefly confined to one margin of the branches. These three species undoubtedly have a peculiar form of branching in common, and one, too, which is not found in any other described species, so far as I have been able to make out from the frequently scanty descriptions available. The polyps are not knowm in any of the species, so that whether in this case a particular type of polyp is associated with this peculiar corallum, I am unable to say at present. A much branched type in the Challenger collection {Antipathella contorta) is certainly in some respects closely allied to Arachnopathes clathrata (Pallas). It shows the same marked contrast between the thickness of the branches and the innumerable needle-like branchlets which spring from them, as is figured by Morison (6). In the Challenger species, however, there appears to be no regular fusion between the slender branchlets in the manner indicated by Morison. The polyps of this species do not appear to differ in any important respect from those of other members of the genus Antipathella. I have not seen a specimen which I could with certauity refer to Arachnopathes clathrata (Pallas), and, so far as I am aware, it has not been identified by subsequent investigators. A specimen in the British Museum, which agrees fairly well with the original description, does not show such a marked 22 THE VOYAGE OF H.M.S. CHALLENGEE. contrast in thickness between the branches and branchlets as is figured by Morison, and as Pallas refers to this figure, it is to be presumed that his type specimen agreed with it. The species Arachnopaihes panicidata, Duchassaing and Michelotti, and Arachno- patlies columnaris, Duchassaing, referred to the genus Arachnopaihes by the authors, appear to me to have no place there. The former species, judging from the figure, is a form allied to Antipathes gracilis, Gray, and is one of the fan-like species, in which the branches are less confluent than in typical RMpidipath.es. Arachnopaihes columnaris, of which Pourtales has given us a photograph, has a similar corallum and polyp (?) to Antipathes larix, Esper, and has been provisionally referred to Parantipathes, n. gen. Fusions occur occasionally between the branches, but it must be remembered that in this type the stem is simple, and therefore the typical Arachnopathes form, brought about chiefly by fusion between branchlets belonging to adjacent branches, cannot occur. The reticulum to which Duchassaing refers forms a tube for a parasitic worm, and its presence is therefore neither generic nor specific, but depends on the presence of the parasite. In the systematic portion of this Report I have temporarily retained the name Arachnopatlies as generic, in order to link together the three species Arachnopathes ericoides, Arachnopathes clathrata, and Arachnopathes aculeata, until more detailed information is obtained regarding them. Rhipidipathes. — Milne-Edwards gives the following definition of this genus : — " Polypier sclerobasique dont les branches s'^talent sur un meme plan en forme d'evantail et se soudent entre elles aux points de contact, de facon a constituer un re'seau." At the time only two species had been described which were considered referable to this genus, viz. : — Antipathes fiabellum, Pallas, and Antipathes reticulata, Esper. Gray in a paper published about the same time (40), and other authors more recently, have described quite a number of species which possess in a more or less marked degree the characters referred to. One of the Challenger species {Aphanipathes cancellata) shows a closer and more regular network than Antipathes fiabellum, Pallas. Starting with this species as the one in which the network is most complete, one may trace this character through a number of forms in which it is less and less marked until finally the original feature has entirely disappeared. Such a series might include A. cancellata, A. fiabellum, A. hypnoides, A. reticulata, A. gracilis, A. paniculata, D. and M. (non Esper), and A. tristis. To begin with, the reticulum is formed by bridges of sclerenchyma which pass across from branch to branch, giving a more or less rectangular network, the sides of the meshes being subequal in thickness. In other forms a similar result is obtained by fusion between pinnules of adjacent branches and a general confluence of the stouter portions of the corallum. In A. hypnoides one begins to find the terminal pinnate branchlets free, and not showing the fusions so abundant in other parts. From this type onwards in the series there is a gradual increase in the size of the terminal fronds in which fusions do not occur, until in REPORT ON THE ANTIPATHARIA. 23 A. panicidata, D. and M., whole paniculate branches are without adhesions, and the fusion affects only the stem and stronger branches. Finally we have in A. tristis, Duehassaing, a type in which actual confluence of parts has ceased to exist, and where, as Pourtales assures us, the fusions described by Duehassaing are more properly to be considered merely as adherences. It only requires one step further to reach such types as A. pedata, Gray, on the one hand, and A. myriophyUa, Pallas, on the other. Both types are fan-like, the former relatively simple with elongate pinnules, the latter quite as complex as A. Jlabellum, Pallas, but entirely without fusions. It will thus be seen that the presence of fusions between certain parts of the corallum is not a reliable character for generic purposes. This fact will be brought out still more prominently if we now consider the structure of the polyps of some of these forms, and enquire whether in any case species not possessing Rhipidipathan characters have a type of polyps found also in that group, and vice versa. I have not been able to study the polyps of Antipathes Jlabellum, the only specimens available being dry, as in most of the species referred to. In Anti- pathella assimilis, n. sp., the form of the reticulate corallum is almost identical with that of Antipathes reticulata, Esp. The polyp of this species is rounded or oval and is provided with six tentacles, two of which, those at each extremity of the mouth, are usually, though apparently not always, inserted at a lower level than the other four. According to Pourtales the zooids of Antipathes tristis have a similar form and arrangement of the tentacles. This type of polyp is by no means confined to species presenting fusions between different parts of the axis, but is seen typically in Anti- pathes subpinnata, Ellis and Solander, and other laxly pinnate types. Aphanijiathes cancellata, n. sp., has quite a different form of polyp — a type which Pourtales has termed sessile. The polyp is oval and so short that it is almost hidden amongst the spines of the sclerenchyma, which often project through the ectoderm in spirit specimens, as is figured by Pourtales in the case of Antipathes humilis. Here again this type of polyp is by no means confined to the species of the genus Rhip)idip>at]tes, but is common to Aphanipathes sarothamnoides, n. sp., and a number of non-reticulate species from the West Inches described by Pourtales. It is true that in Rhipidipathes jlabellum, and also in two or three new species which are probably allied to it, the reticulum is formed in a different manner to that of either Aphanipathes cancellata or Antipathella assimilis, and at present we know nothing of the polyps of these types. Unless, however, they should ultimately prove to have a form of zooid unlike any yet described, the generic name Rhipidipathes ceases to have any systematic value. I have not retained it here, because it would be necessary to use it in a restricted sense, and in the absence of further information on the subject it seems advisable so far as possible to refer all species of which the zooids are not known to the genera with which they seem to have most in common. 24 THE VOYAGE OF H.M.S. CHALLENGER. Antipathes. — Under this genus Milne-Edwards classes all forms of Antipathidae which have not been assigned a place in other genera. It is needless to say that it includes a medley of species, united only by one common character,— the absence of fusions between adjoining parts of the corallum. So far as the mode of branching is concerned, we find almost every variety from simply pinnate forms, to the most complex. One need only mention Antipathes pennacea, Pallas, Antipathes subpinnata, Ellis and Solander, Antipathes larix, Esper, Antipathes abies (Linnaeus), Antipathes rnyriophylla, Pallas, Antipathes dichotoma, Pallas, and Antipathes virgata, Esper, to show the endless variety. At the present time some forty or fifty species might be named, all of which would agree in the plan of branching with one or other of the species enumerated above. Just as the presence of fusions in a colony appears to be of no generic value, so I con- ceive the absence of them to be a minor feature, of value perhaps for specific purposes (though even then not always), but of no value generically. Of the species which would come under this genus as defined by Milne-Edwards, I find some with dimorphic zooids and others without ; whilst in the latter section alone I find amongst the comparatively few species examined at least four distinct types of zooids. Evidently then it is impossible to adopt Milne-Edwards' modification of this genus, and in attempting to found new ones I have relied chiefly on the form and structure of the zooids in the various species observed. Gray (42) in 1860 described two additional species from Madeira, viz., Antipathes (Cirrhijjathes) setacea and Antipathes gracilis. I have been unable to find the type of Cirrhipathes setacea in the British Museum collection, but the type of his var. occi- dentalis is preserved there, and as this appears to differ from any form previously described I have here raised it to the rank of a species. In the same year the first part of Duchassaing and Michelotti's Memoir on the Corals of the Antilles (43) appeared, and in this and the concluding portion published in 1866 four new forms are described, Cirrhipathes desbonni, which has since been met with by Pourtales, and three species of Antipathes. Of these Antipathes americana is probably a distinct species, whilst Arachnopathes particulate/, seems closely allied to Antipathes atlantica, Gray, and Antipathes dissecta is equally closely related to Antipathes glaberrima, Esper. Lacaze Duthiers was the first to study the structure of the polyps of the Antipathidse, and in 1864 published the first of two important memoirs on the morphology of the group. The earlier memoir (44) treats of the structure and relations of Leiopathes lamarcki, Haime. The author was enabled to study living specimens collected by the coral fisher- men off the coast of Algeria, and for the first time to make out the structure and affinities of the polyp of this form which had not previously been observed. He shows how in different states of preservation the whole or various parts of the colony have received different names. REPORT ON THE ANTIPATHARIA. 25 Lamarck made two species — one of the naked sclerobasic axis [Antipathes glaberrima), and another of a specimen covered with ccenenchyma [Gorgonia tuberculata). Isolated polyps have been named Palythoa denudata and Zoanthus sp. The leading points of Lacaze Duthiers' investigation will be found under the genus Savaglia [infra, p. 51), and I will here only mention those bearing on the systematic position of the species. In the first place this form, for which Lacaze Duthiers creates the new genus Gerardia, has not a spinose axis nor does its ccenenchyma contain spicules peculiar to it, but only those which reach it from other forms and become adherent. It thus differs from Antipathidse in the absence of spines. The sclerenchyma is thin and horny, and is primarily secreted around the axis of some other form, usually one of the Muriceidse, so that the mode of branching is not characteristic of the Gerardia, but of the particular species on which it becomes parasitic. In older specimens where the stems and branches extend beyond the Gorgonid basis, its growth becomes bushy. The polyps have twenty-four tentacles arranged in two rows of twelve each, and each tentacle corresponds to an interseptal chamber as in true Actiniaria. There is also a system of canals in the ccenenchyma, bringing the whole of the polyps of a colony into communication. In his second memoir Lacaze Duthiers (45) gives an account of the structure of Antipathes subpinnata and Antipathes larix, which have only six tentacles. Here too he confirms the surmise of Dana, that the Antipathidse are closely related to the Actiniaria. The mesenteries bearing reproductive organs, which Dana supposed to exist, are described and figured by Lacaze Duthiers from living or fresh specimens. The mesenteries are, however, unecpially developed. He describes six : two principal ones bearing reproductive organs placed in a line parallel with the branchlet on which the polyp is placed dividing it into two similar halves, and four others less fully developed and destitute of reproductive organs, two in each lateral portion of the polyp. No sections were made, but so far as could be made out no system of canals in the ccenenchyma similar to those of Gerardia are visible externally, though Lacaze Duthiers supposed them to be present. On account of these researches of Lacaze Duthiers, and the interesting morphological points which they bring out, the Sclerobasic Zoantharia have been divided by Yerrill (46) and others into two suborders: — GERAEDiDiE, with twenty-four tentacles and mesenteries, . genus Gerardia. Antipathid.e, with only six tentacles, . . . genus Antipathes, tyc. In 1868 Heller (50) recorded two species as occurring in the Adriatic, viz., Anti- pathes subpinnata, Ellis and Solander, and Leiopathes glaberrima (Esper), both forms already known from the Mediterranean. Verrill, in a review of the-Polypi of the East Coast of the United States (48), mentions only two species, Antipathes boscii, Lamouroux, collected by Agassiz, near Charleston, and Antipathes alopecuroides, Ellis and Solander, the latter on the authority of Elks. (ZOOL. CHALL. EXP. PART LXXX. — 1889.) LM i 26 THE VOYAGE OF H.M.S. CHALLENGER. In another paper on the Halcyonoid Polyps in Yale Museum, Verrill (52) describes a new Gorgonid, Paramuricea cancellata, founded on a species of Dana's, which he regards as probably equivalent to Antipathes Jlabellum, Esper (non Pallas), but in a later part of the same review says that he has a species of Echinogorgia, which agrees exactly with Esper's plate of Antipathes Jlabellum. In 1869 the same author described a new species, Antvpathes panamensis, from Panama Bay. Pourtales in a number of papers published between 1867 and 1880, dealing chiefly with the Coral Fauna of the Caribbean Sea and the Gulf of Mexico as explored by the "Hassler" and "Blake" Expeditions, describes altogether eighteen species. Of these eleven are described as new and the remaining seven are referred to Old World types or to forms previously described from the AntUles. In 1867 and 1868 (51) three species were described as new. Two of these, viz., Antipathes filix and Antipathes humilis, have zooids of a type which Pourtales terms " sessile." The zooids are very short, so much so that in spirit specimens the spines of the sclerenchyma project through the zooids in all directions. The third species, Antipathes tetrasticha, has alternate double rows of branchlets, and small elongate polyps. Two other species are also mentioned and partly defined. These appear in a later paper (56) to have been identified as Antipathes dissecta, Duchassaing and Michelotti, and Antipathes lenta, n. sp., although Pourtales does not refer to his former descriptions. In his account of the Corals of the " Hassler " Expedition (58) four other species are recorded, two of which are new. Of these, Antipathes abietina seems allied to Antipathes filix, Pourtales, whilst Antipathes Fernandezii has apparently a similar polyp to Antipathes tetrasticha, and perhaps belongs to the same genus. The other species recorded are Cirrhipathes desbonni, Duchassaing and Michelotti, and Antipathes columnaris (Duchassaing). It appears, however, from a later paper by the same author (71) that the form referred here to Cirrhipathes desbonni is a spiral species differing essentially from that type, and was afterwards referred by Pourtales to Cirrhipathes spiralis, Pallas. The second identification appears to me to be equally erroneous, as the polyps in the species referred to are distributed in a single longitudinal series, whereas in Cirrhipathes spiralis (Pallas), so far as I can ascertain, they are arranged subspirally all around the axis. The species referred to by Pourtales, which is particularly abundant in the West Indian Seas, is the one which I have selected as the type of the new genus Stichopathes. The Corals of the " Blake " Expeditions are recorded in two papers. The earlier one (64) includes descriptions of two species of Antipathidse not previously observed by Pourtales. These were referred to Antipathes myriophylla, Pallas, and Antipathes tristis, Duchassaing. In the later paper (71) Pourtales corrects an error referring to his identification of Antipathes myriophylla, Pallas. It appears that the specimen referred REPORT ON THE ANTIPATHARIA. 27 to this species is really a larger and more densely branched form of Antipathes filix, Pourtales. In this, his last work on the subject, five new species are described, and a sixth, Antipathes eupteridea, Lamouroux, was met with apparently for the first time since the species had been described by Lamouroux from a Martinique specimen over fifty years previously. Amongst the new forms described, Antipathes salix, Antipathes rigida, and Antipathes thyroides have a similar polyp to that of Antipathes humilis, Pourtales, and the two former are probably only varieties of the same species. Antipathes picea appears to possess a type of polyp not previously described. The oral cone forms a prominent rounded knob, the mouth apparently being very small in spirit specimens, and the tentacles, arranged somewhat in pairs, are flattened and have a crenate margin. Finally, Antipathes tanacetum agrees precisely with Antipathes picea in the mode of branching, but has much more elongate spines. The polyps in this specimen were too badly preserved to show the arrangement of parts. The true Cirrhipathes desbonni, Duchassaing and Michelotti, is also recorded. Pourtales is the first and indeed almost the only author who has given us figures of the arrangement of spines in all the species described. His plates also include figures of the polyps of seven species. In 1871 he advocated the removal of Gerardia lamarcki from amongst the Anti- patharia, and suggested that it should be included amongst the Zoanthidse as the type of a new subfamily. He then argued that the polyps of Gerardia differ in no particular from those of the Zoanthidse in the arrangement, number, or shape of the tentacles, and even agree with that group in the habit of encrusting the derm with small foreign bodies. He at the same time pointed out that this genus has no other relationship with the Antipathidse than the property of secreting a horny axis. He calls attention to the fact that the genera of Antipathidse as at present defined are based solely on the solid parts, and adds : — " It has seemed to me, however, that two distinct types of polyps could be distin- guished, the one well circumscribed, flower-shaped, symmetrically radiate, with long tentacles ; the other so elongate longitudinally that the radiate shape is quite indistinct, the six tentacles being disposed in pairs at some distance from each other." At the same time he points out that amongst the few species examined there appeared to be no connection between the form of the polyp and the general shape of the corallum. Throughout all his papers Pourtales uses the name Antipathes as the sole generic designation, but in the last of the series makes an attempt to use the difference in the shape of the polyps, and in the disposition and form of the spines, to draw characters for a revision of the group. He calls attention to the fact that there are at least two different types of spines, and that these are usually associated with a different form of polyp. In the one type the 28 THE VOYAGE OF H.M.S. CHALLENGER. spines are cylindrical, generally densely set, and sometimes collected in tufts, as in Antipathes humilis. They are frequently unequal on the two sides of the pinnule, being longer on the side occupied by the polyp. In the second type the spines are triangular and compressed. In this case the spines are disposed regularly in quincuncial order around the pinnules, ' generally, though not always, disposed spirally, and showing no tendency to elongation in the neighbourhood of the polyps. According to Pourtales' experience those species having triangular spines have polyps with longer tentacles than those in which the spines are cylindrical, and the polyps are usually more regular in shape. " In a very few instances the tentacles are found retracted, as figured by Lacaze Duthiers ; in most cases they are simply contracted, and in many species they are probably not retractile at all." Pourtales was also of opinion that the shape, size, and arrangement of the spines probably afford reliable specific characters. He is indeed the first author who has given us reliable information on these points in the description of new species. Unfortunately Pourtales was not spared to complete the much needed revision which he contemplated, but the foregoing account of his views on the subject will show that the lines on which he proposed to work are, in the main, those which have been adopted in the present monograph. Gray in 1868 (49) described a new species of Cirrhipathes (C. Jiliformis) from Australia, but as in other species described by this author, the polyps were not observed, and the description is insufficient for specific purposes. The type is in the collection of the British Museum, as well as three or four other specimens of the same species collected more recently. Duchassaing (54), in his final review of the Zoophytes and Sponges of the Antilles, describes shortly four new species. Two of these, Arachnojxtihes columnaris and Rhipidipathes tristis, have since been observed by Pourtales, who has figured a specimen of the former, and also the arrangement of the spines in both species. The other species described by Duchassaing, viz., Antipathes taxiformis and Antipathes melancholica, are very imperfectly described ; the latter is probably allied to Antipathes dissecta, Duchassaing and Michelotti. Liitken in 1871 (55) described a new form, Antipathes arctica, taken from the stomach of a shark captured off the coast of North Greenland. This is the first and only record (with the questionable exception of Antipathes boscii already referred to) of the occurrence of any species of Antipatkidse north of the Mediterranean and the southern States of North America. Liitken in a footnote mentions that Sir Wyville Thomson had informed him that specimens referable to this group had been collected during the expeditions for the exploration of the deep water around the British Islands. I am, however, unable to find any reference to them in Sir Wyville Thomson's published works, or in the zoological results of the various expeditions referred to. REPORT ON THE ANTIPATHARIA. 29 Haeckel, in his Arabische KoraHen (59), gives a figure of a living colony of Antipathes corticata, Lamarck, but unfortunately only gives a few words in explanation of the plate. He also refers to Gerardia lamarclri, and copies Lacaze Duthiers' figures. No new species are described. Klunzinger (60) in 1877 recorded the occurrence of Cirrhipathes anguina, Dana, in the Red Sea, and also called attention to another species, Antipathes isidis- plocamos, first observed in the Red Sea by Ehrenberg. Fragments of the stem are figured, and Klunzinger discusses the probable identity of a specimen, figured as Antipathes compressa by Esper, with this species. The specimen is, however, too imperfect to admit of proper identification, and may be the base of any of the larger species already described. It may nevertheless prove to be a distinct species, but in the absence of more detailed information it must temporarily be included amongst the species duhise. Studer (65) in 1878 gave notes of two species of Antipathidse collected during the " Gazelle" Expedition, viz., Antipathes fcenicidum, Lamarck (I Antipathes jbeniculacea, Pallas), off West Australia and Mermaid Straits, in 45 to 50 fathoms, and Antipathes pinnatifida, Lamouroux, Mermaid Straits, in 50 fathoms. He also mentions the occurrence of broken portions of a large stem, from 900 fathoms, too fragmentary for identification. G. v. Koch (62), in a paper on the Phylogeny of the Antipatharia, first gives an account of the structure of Antipathes larix, and calls attention to the fact that the pinnules arise at right angles to the stem and are disposed in six longitudinal rows. The polyps are placed in a single row on the superior surface of each pinnule, and are elongated in the direction of the pinnules. The mouth is situated on a conical or cylindrical projection from the peristome, and its aperture is usually oval, with the longer axis directed trans- versely. He describes ten mesenteries in the oesophagus which are unequally developed, only two being complete. These correspond with the long axis of the polyp, and divide it into two symmetrical halves. Four others, not so fully developed, are placed two on each side, so that each chamber corresponds with a tentacle. Four others still more rudimentary " kauni in den Magenraum hereinragende Scheidewande sind so angeordnet, dass sie der Langsachse zunachst stehen." All the mesenteries consist of a thin hyaline layer of connective tissue, which is clothed on both sides by entoderm. In the base of the polyp there is a longitudinal septum having a dilation at its free extremity, in the cavity of which the sclerobasic axis is contained. This is surrounded by an epithelium from which it is derived, and which is probably a portion of the ectoderm. Koch next describes what he regards as a new genus (Gephyra) of Zoantharia, which appears to link closely the Antipatharia with the Actiniaria. The polyps of this species, named Gephyra dohrnii, have eighty or more tentacles which can be retracted, as in many Actiniae and in Gerardia. They are found singly or in colonies, parasitic on s 30 THE VOYAGE OF H.M.S. CHALLENGER. the axis of Isis elongate/,, Esper, and other forms, and surround their bases with a black horny mass which becomes separated from a portion of the ectoderm and is homologous with the axial skeleton of Antipatharia. The polyps of a colony are united to each other by basal processes, but have no true ccenenchyma. Koch admits that this species appears at first glance to belong to the true Actiniae, but believes it to differ on account of the presence of a chitinous base secreted by the polyps, which is not distinguishable from that of the Antipatharia. In his discussion of the phylogenetic relations of the Antipatharia, Koch calls attention to the fact that Lacaze Duthiers first showed their relationship to the Hexacoralla through Gerardia, but that on account of the peculiarity of the skeleton of the Antipathidae, the latter still remained an isolated group. Koch thinks he has added a link in Gephyra, which by its skeleton unites the Antipathidae with the Hexacoralla, and proposes to derive Antipathes from skeletonless Hexacoralla similar to the Actiniae of the present day. The polyps in their anatomical and histological structure are quite like many small Actinias, but in Gephyra and Antipathes there are simplifications, particularly in the musculature, to be attributed to a reduction in the size of the polyps, and their union into colonies. The tentacles for the same reasons have become fewer, but scarcely altered. The mouth and oesophagus show no marked variation in Gephyra, Gerardia, and Antipathes. In Antipathes only two mesenteries are fully developed, the remaining eight being more or less rudimentary, but by their position and kind of degeneration we may judge that the ancestor of Antipathes had six tentacles and the same number of antimeres. Gerardia has twenty-four mesenteries and the same number of tentacles, whilst Gephyra approaches the Actiniae closely in the number of tentacles and mesenteries. Gerardia stands alone in having a network of canals uniting the polyps, but Gephyra would in this respect approach it more closely were it shown that the colony results from budding. Koch suggests the following as the most probable stages in the phylogenetic history : — 1. Soft Actiniae which secrete for their support a horny substance by means of the basal ectoderm. 2. Those situated on a slender cylindrical base surround it and enclose it with horny matter. 3. The polyps, by budding, form a colony. Axis ceases to be solely secreted around some foreign substance, and now grows independently beyond the limits of supporting substance. 4. Degeneration sets in in certain parts. 5. The axis becomes entirely independent. With increase in number, the polyps become reduced in size, and connected with this is a reduction in the number of mesenteries and tentacles. REPORT ON THE ANTIPATHARIA. 31 It may be stated here that Gephyra, which has not only been found in the Mediterranean, but also off the N.W. coast of France and off the Irish coast, has been included by Andres amongst the true Actiniaria, as a member of the genus Sagartia, as modified in his monograph of the Naples Actiniaria, Haacke (67) in a review of the morphological relations of the Anthozoa, discusses the position of the Antipatharia from data supplied by the researches of Lacaze Duthicrs and v. Koch. As will be shown later, the species on which these observations are chiefly founded, viz., Antipathes larix, can scarcely be considered typical of the group. On this account the diagram of Haacke illustrating the arrangement of the mesenteries (sarcosepta) in Antipathes is misleading, and as a matter of fact the six mesenteries first described by Lacaze Duthiers, although unequally developed, are all complete in the usual sense in which that term is applied. This subject will be discussed in greater detail when considering the morphology of the group. Carter in 1880 described (69) under the name Hydradendrium spinosum, a species dredged in the Gulf of Manaar, which was supposed to belong to the Hydractiniidse. The species in question, of which good figures are given, has a branched chitinous axis densely covered with spines. The polyps were not observed. It appears from a later paper by the same author (70) that the Rev. A. M. Norman suggested that the species in question was probably an Antipathes, and Carter admits its close resemblance to Antipathes ulex, Ellis and Solander. He then discusses the affinities of Antipathes, and its probable relation to the Hydractiniidae. Surely the researches of Lacaze Duthiers in 1866 showed that this could not be the case, whilst v. Koch in 1878 demonstrated that Antipathes larix, Esper, possesses all the essential characters of true Zoantharia. The question need not be discussed here further than to state that through the kindness of Mr. Moore of the Liverpool Museum I have been enabled to examine a specimen of Carter's species, which undoubtedly is very closely allied to Antipathes idex, Ellis and Solander. There are, however, certain differences between this form and specimens of Antipathes ulex which I have seen, but whether these are of specific value I am unable to decide at present. In the meantime I have regarded Carter's species as distinct. In the Report on the Challenger Actiniaria (72) R. Hertwig discusses incidentally the phylogenetic relations of the Antipatharia. A new family of Actiniaria is established for certain interesting Actinians, particularly those described by Moseley : — Actinia abyssicola and Actinia gelatinosa, together with others related to them. The following are the characters : — Amphianthid.e. — Hexactiniae, which are attached to the axial skeletons of Gorgonidas, with shortened sagittal and elongated transverse axis ; transverse axis lying parallel to the axial skeleton of the Gorgonia; circular muscle mesodermal; the principal septa only are perfect and sterile. In addition to the species included in the Challenger collection, Hertwig supposes the Actinia s. Catherine, Lesson, and Gephyra dohrnii, v. Koch, 32 THE VOYAGE OF H.M.S. CHALLENGER. to belong to the same family. R. Hertwig points out that v. Koch's researches on Antipathes larix, Esper, already referred to, show an interesting connection between the form of the polyp in that species and the peculiar elongation of the body in the Amphi- anthidfe. In Antipathes larix " the body is elongated in the direction of the skeletal axis, and the transverse axis of the animal thereby appears lengthened, whilst the sagittal axis is shortened. . . . Two pairs of septa, which correspond to the oral angles, are sterile, and consequently comport themselves like directive septa, whilst the two remaining pairs, lying in the prolonged transverse axis, bear reproductive organs, and are therefore best termed accessory septa." In conclusion he thinks it probable that the Amphianthidsfi bring about the transformation of the Actiniaria to the Antipatharia. It is, however, necessary to determine whether the paired arrangement of the septa and the presence of the directive septa can be demonstrated in Antipatharia, and whether the sagittal and transverse axes have the same direction in both groups. The evidence which I have been enabled to obtain on these points will be found in the chapter devoted to morphology. Andres, in his work on the Actiniaria of the Gulf of Naples (73), discusses the position of Gephyra dohrnii, v. Koch, which he places in his genus Sagartia, as modified in the monograph in question. G. v. Koch saw in the sclerenchymatous membrane of the base of this species a mode by which the axial skeleton of Antipatharia could be produced, and suggested Gephyra as a bridge from the Actiniaria to the Antipatharia. Andres points out that the relation is, so to speak, physiological and not phylogenetic. He calls attention to the similarity in this respect of two deep-sea Actinians described by Moseley, viz.: — Actinia abyssicola, found on the stems of Mopsea, and Actinia gelatinosa on those of Gorgonia. At present data are wanting to enable us to decide whether in such cases there is a true morphological affinity or only a parallelism of function. With regard to the systematic value of the power of the " base " to secrete a basal membrane, the following quotation is of interest : — " Forse non Ion tana parente e la Phellia nummus che abita pure acque profonde, secerne abbondante muco solidificabile, ed ha macchie marginali alterne chiare e scure. II carattere abbracciante (amplectens) della base ha poca importanza ; perche gli animali talora aderiscona a corpi piatti con la base allargata come un'altra attinia qualsiasi." In 1886 Koch (76) described a new species of Antipathes from the Gulf of Guinea [Antipathes squamosa), which is allied to Antipathes spinescens, Gray, if not identical with it. A new species in the Challenger collection [Antipathes cylindrica) has also a similar " bottle-brush " type of corallum. In a recent review of the results of the cruises of the " Blake," Agassiz (78) refers to the species of Antipatharia described by Pourtales, calling attention to the frequency and wide bathymetric range of certain forms, such as Stichopathes pourtalesi (= Antipathes spiralis, Pourtales) and Antipathes cohimnaris. With reference to the latter species, s REPORT ON THE ANT1PATUAR1 A. 33 Agassiz calls attention to the fact first noted by Pourtales, that in certain species of Antipathidse the small twigs on the stem and branches of the corallum may be modified so as to form a tube serving for the protection of parasitic annelids (Ma/rphysa, &<•.). A similar action of parasitic annelids has been noticed in Lophohelia, Stylaster, AUopora, &c. Agassiz, in an earlier account of certain dredging operations of the "Blake" (66), noticed that the Antipathidse, like certain Gorgonidse (especially Riisea), showed a lirighl bluish phosphorescence when coming up in the trawl. {ZOOL. CHAM,. EXP.— PART L.XXX. 1889.) ''' '' GENERAL MORPHOLOGY. In the present section I propose to give a general outline of the structure of the various genera of Antipatharia, more especially with regard to the form of the zooid and the number and relative development of the mesenteries. On account of the previous want of information on the subject, it will also be interesting to consider generally the bearing of these preliminary observations on the relationship of the Antipatharia to other Zoantharia. The Antipatharia are all colonial in habit, and the corallum is usually fixed by a dilated base to some foreign object. Some species are parasitic, and it seems probable that in the ancestors of the Antipatharia this feature was more frequent than in living types. The production of a colony from the original oozooid developed from an ovum has not been studied, but it is probable that this takes place by budding. The colonies consist of blastozooids united together by ccenenchyma, the whole of the soft tissues being supported on a central horny axis. The zooids of the Antipathida; are all constructed on the same plan, but amongst the genera examined there is a gradual specialisation in one direction, corresponding to a physiological division of labour, and finally resulting in dimorphism. In an ideal case the zooid is more or less rounded in outline, and consists of a short hollow cylinder projecting beyond the surface of the ccenenchyina for no great distance. The peristome usually bears a central conical projection, on the surface of which the mouth is situated. The tentacles are six in number; they are arranged radially in zooids with a rounded outline, but in many cases there is a more or less well-marked arrange- ment in longitudinal rows, due to elongation of the zooid in one direction. The body-wall is not separated from the peristome by a well-marked "margin," as in typical Actiniaria, and passes imperceptibly into the ccenenchyma. The oral cone often shows on its inner margin a number of crenate folds, which are partially everted portions of the stomodseum, each fold occupying the space between two mesenteries. In other cases the surface is quite smooth. The mouth leads through a short stomoda?um into the ccelenteron. The 36 THE VOYAGE OF H.M.S. CHALLENGER. stomodseum is supported by a number of mesenteries 1 (sareosepta) which incompletely divide the ccelenteron into a number of radial chambers. The mouth is rarely round, most frequently it is slit-like, the long axis being placed at right angles to the axis of the branch on which the zooid is situated. The axis corresponding with the elongation of the mouth will be spoken of as the " sagittal " axis, that at right angles to it as " transverse." I have not been able to note any well-marked siphonoglyphe in the various species examined, though, on account of the small size of the zooids, these would be difficult to make out in preserved material, even should they exist. The tentacles arise from the margin of the peristome, or some form the peristome and others form the body-wall. In the Antipathinse there are always six tentacles present. These are simple finger-like outgrowths of the ccelenteron. In the Schizopathinae each zooid bears only two tentacles, but in this subfamily the zooids are dimorphic, and three zooids correspond morphologically with an unspecialised zooid of the ordinary type. In Dendrobrachia the tentacles are branched. The mesenteries, on account of their different relative development, may be conveniently considered under two heads, viz., "primary" and "secondary." This division is, however, artificial, and is merely used for convenience of description. In all Antipathidse the primary mesenteries are six in number, and are well developed. The secondary mesenteries are developed in a varying degree in different genera. Of those already examined the number is six or four, or the series may be entirely wanting. The number, arrangement, and comparative length of the mesenteries may be best studied by means of a series of horizontal sections commencing at the oral surface.. The number of mesenteries in the oral cone generally differs from that in the lower section of the ccelenteron. The arrangement in the various genera will be best understood by reference to the diagrams. o Leiopathes. This genus may be conveniently taken first on account of the fact that, as the mesenteries are present in a multiple of 6, it is more directly comparable than other genera with the regular Hexactinise. In Leiopathes glaberrima the mouth is somewhat elongated in the sagittal axis, whilst the zooid is elongated in the transverse axis. In neither case, however, is the elongation so pronounced as in some other forms. A horizontal section through the upper portion of the oral cone (fig. 1) shows that 1 I have used the term mesentery in the present Report in preference to septum or sarcoseptum, because, whatever objection there may be to its use in reference to the Anthozoa, it has within that group a well-defined meaning. It is solely applied to the soft radiating partitions passing from the body-wall to the stomodreum, which imperfectly divide the ccelenteron into chambers. In addition to these in Antipathida- the individual zooids are imperfectly separated from one another by means of vertical mesoglceal partitions, to which the terms septa and sareosepta might be considered equally applicable, but which would not come within the meaning of the term mesentery. REPORT ON THE ANTJPATHAUI A. 37 twelve mesenteries are present, enclosing twelve interseptal chambers of the ccelenteron. In this region the mesenteries are so disposed that an interseptal chamber is situated at each end of the sagittal axis of the mouth, and the middle of each wall of the mouth is also bordered by an interseptal chamber. In other words, both the sagittal and median transverse axes correspond with interseptal chambers instead of mesenteries. The mesenteries are here all equally developed, and consist of a delicate plate of mesoglcea lined on each side by entoderm. Each quadrant of the oral cone contains three mesen- teries, two complete interseptal chambers, and half each of a chamber in the sagittal and another in the median tranverse axis. For convenience of description, the mesenteries have been numbered from 1 to 12 in fig. 16, which is diagrammatic. A few sections lower down, the arrangement just described becomes changed. The mesenteries numbered Fig. 1. Fig. 2. Fig. 3. Figs. 1-3. — Horizontal sections of Leiopathes gldberrima. Fig. 1. — Section through the upper portion of the oral prominence, passing also through the two pairs of lateral tentacles. Fig. 2. — Section near the base of the oral prominence, showing the lumen of each of the four lateral tentacles opening circumoral sections of the ccelenteron. Fig. 3. — Section in a plane beneath the insertion of the sagittal tentacles, showing the six primary mesenteries projecting ccelenteron ; those in the transverse axis are most important. into the into the 4 and 9 in fig. 16 lose their connection with the wall of the oral cone, and become lost, and those numbered 3 and 10 now become gradually more important, and change their position so as to occupy the median transverse axis. This arrangement is represented in fig. 2, which is taken from a horizontal section passing through the base of the lateral tentacles. A little further down four other mesenteries, viz., numbers 2, 5, 8, and 11, lose their connection with the outer wall, and after persisting for a short distance as projections from the wall of the stomodseum, ultimately disappear. In the lower portion of the stomodseum (fig. 3) only six mesenteries remain, namely, those numbered 1, 3, 6, 7, 10, and 12 in fig. 16. These are the mesenteries referred to as "primary." The transverse primary mesenteries are most important, and bear the reproductive organs as well as a well-developed convoluted mesenterial filament. The other primary 38 THE VOYAGE OF H.M.S. CHALLENGER. mesenteries are somewhat shorter, and on account of the fact that the point where they leave the stomodseum is nearer to the body-wall, have a shorter horizontal course ; they bear a rudimentary mesenterial filament, but are devoid of reproductive elements. The section of the ccelenteron corresponding to a tentacle will be seen from the foregoing account to vary in different parts. Two tentacles correspond to the sagittal axis, one being situated at each end of the stomodseum (these are " lateral " as regards the position of the zooids on a pinnule). In these cases the section of the coelenteron corresponding to a tentacle, that for example between mesenteries 1 and 12, is an enlargement of one of the twelve interseptal chambers present in the oral cone. In the species under consideration these tentacles are usually longer and thicker than the others, and frequently extend horizontally in spirit specimens, whilst the other four stand up vertically. The remaining four tentacles each cover a section of the coelenteron corresponding to two and a half chambers in the oral cone. When the mesenteries numbered 3 and 10 come to occupy the median transverse axis, one tentacle corresponds to the space between mesenteries 1 and 3, another to that between 3 and 6, and so on. Cirripathes ( Cirrhipathes). In the genus Cirripathes, as restricted in the present Eeport, the zooids are distributed all around the axis, and are never found in single linear series, as in Stich!» further subdivided by the interposition of another pair of mesenteries, giving six in all. It therefore appears as if, within the family, three pairs represent the more primitive arrangement, and that those genera possessing five pairs, as well as Leiopathes with six, have been elaborated from it. This view appears also to be supported by our knowledge- of the ontogeny of the Hexactinia?. Neither Sagartia dohrnii nor Savaglia lamarcki appear to offer any assistance in a solution of the question. The former has been shown by Andres to be a true Actinian. The fact that Sagartia dohrnii secretes a horny basal membrane, which may become tubular, appears to have no phylogenetic value. Many true Actiniaria have the same power. Savaglia has nothing in common with Antipathidte beyond the possession of a branched lamellate sclerenchyma, which, however, is always primarily parasitic, as in Amphianthidse, but which may extend beyond the limits of the foreign basis. The zooid, so far as its structure is known, belongs to the true Actinian type, and has no similarity whatever to the zooid of Antipathidas. The only essential point on which it differs from colonial Actiniaria appears to consist in the fact that the coenenchyma possesses a series of interzooidal canals, one of which opens into the base of each interseptal chamber. It appears probable that some such communications must also exist between the zooids of certain Zoanthidoe (e.g., Epizoanthus stellaris, R. Hertwig). The spinose horny axis, which, excluding Savaglia, is peculiar to the Antipatharia, is related through Dendrobrachia to that of certain Gorgonacea. In certain portions of the axis of Dendrobrachia the sclerenchyma is rugose, with a spinose margin. In Acanthoisis, Wright and Studer, the axis consists of short calcareous nodes and more elongate horny internodes. The internodes are rugose with a dentate margin, giving an appearance very similar to the axis of Dendrobrachia. In Gorgonella, Val., the axis is horny and rugose, but without the dentate margin. Thus whilst the spinose sclerenchyma of Antipathidae appears to be linked to that of Actiniaria through Savaglia, it on the other hand is linked to that of certain Gorgonacea through Dendrobrachia. The sclerenchyma of Leiopathes glaberrima too, the stem and main branches of which are always smooth and glossy, appears intermediate between that of normal Antipathidas on the one hand and Savaglia on the other. On this account it appears probable that a truer knowledge of the systematic position of the AutipathidaB is more likely to be obtained by a study of the zooids than by a study of the sclerenchyma. Finally, a few points in which the Antipatharia resemble certain other Zoantharia may be indicated ; some of them have been already mentioned. A general resemblance between Savaglia and Zoanthidse is most marked, but it is as yet uncertain whether their mesenteries are arranged on the same plan. The arrange- ment in Zoanthidee is most peculiar, and a renewed study of the arrangement in Savaglia is very desirable. The Amphianthidse, as R. Hertwig has already pointed out, bear a general resemblance 70 THE VOYAGE OF H.M.S. CHALLENGER. in form to certain Antipathidse. In both families an elongation in the transverse axis is often marked. In Amphianthidse, however, the stomodseum is usually elongated in the transverse axis, so that the siphonoglyphes, supported by the directive mesenteries, almost touch one another. In Antipathidse the elongation in the stomodseum, when such occurs, usually takes place in the sagittal axis. It appears probable that the elongation in the transverse axis, characteristic of the Amphianthidse, and also found though not always well marked in Antipathidse, does not afford a character of phylogenetic value. In Dendrobrachia fallax we have a form with branched tentacles, thus showing that the type of the tentacle once thought to be characteristic of Alcyonaria may occur also in Antipatharia as well as in Actiniaria. CLASSIFICATION. It will be well to consider briefly the value of the present collection for a rearrange- ment of the Antipatharia. In the past, owing to the lack of information concerning the organisation of the zooids, it has only been possible to make use of skeletal characters, the mode of branching, and the size, shape, and arrangement of the spines in the definition of both species and genera. Even Pourtales, who has figured the zooids of several of the West Indian forms, and recognised two or three distinct types, considered it premature to attempt any rearrangement of the group based on the information at hand. I consider, however, that the information brought forward in the present Report, although, undoubtedly, not sufficient for a complete reorganisation of the group, throws sufficient light on the subject to indicate at any rate the lines on which future classification must be based, and a partial revision has been attempted in consequence. This seems the more justifiable on account of the relatively large number of species of which I have been enabled to make a microscopical examination of the zooids. Including the species now described as new, and supposing the synonymy here adopted to be correct, the list of species referable at present to the Antipatharia may be fixed at 98, and of many of these we have at present only the most meagre information, so that in many cases it is impossible to decide whether the list might not be still further reduced. Previous authors have only given us information concerning the structure of the zooids in three species, viz., Savaglia lanxarcki, Antipathella subpinnata, and Parantipathes larix. G. v. Koch has more recently given us a more detailed account of the structure of Parantipathes larix, including a more accurate description of the number and position of the mesenteries. The other two species have not been examined by subsequent investigators, and Lacaze Duthiers' account of Antipathella subpinnata is very incomplete. Of the 97 species referred to I have been enabled to study the structure of the zooids of 22, viz., 16 Challenger species, 4 from the Mediterranean (including Parantipathes larix and Antipathella subpinnata), and 2 in the British Museum collection. An examination of these species, the detailed results of which I hope to publish in due course, has led me ■ to the conclusion that a proper arrangement of the Antipatharia can only be completed when we possess accurate information concerning the morphology of all the species. A marked contrast between the Alcyonaria and the Zoantharia is found in the fact that, i'2 THE VOYAGE OF H.M.S. CHALLENGES. whereas the number of tentacles and mesenteries is constant in the former order, it is most variable in the latter. In this respect the Antipatharia agree with other Zoantharia, at least to this extent that there is a variation in the number of mesenteries present in the various species already examined. This distinction has necessarily led to a difference in the characters selected as being of generic value in the two great sections of the Anthozoa. I am thus inclined to think that the method adopted by E. Hertwig in the classification of the Actiniaria will be likely to yield the most reliable results if applied to the Antipatharia also. The characters available for a classification of the Antipatharia are more limited than might be at first supposed. The sclerenchyma is apparently always chitinous, and is more or less spinose in all the species described, excepting Savaglia lamarcki, which constitutes the only known species of the Savagliidse. The mode of branching has been generally admitted to give no characters of generic value, indeed, so far as can be made out at present, this feature is sometimes not even of specific value. For instance, in Antipathes picea and Antipathes tanacetum, described by Pourtales, the two forms are said to be precisely similar in the mode of branching, the species being con- sidered distinct on account of differences in the spines. The value of the mode of branching for generic purposes has already been partly discussed when considering the value of the genera proposed by Milne-Edwards. More need not be added at present. Further investi- gation may, however, show that some of the genera here defined (e.g., Aphanipathes) may bear subdivision, and that in this case distinct types of branching may yield characters of value. With regard to the form, size, and mode of distribution of the spines, it may be stated that three distinct types, viz., cylindrical, triangular, and knobbed, have been observed ; their value as an aid to classification does not at present seem clear. Un- doubtedly, as Pourtales has already pointed out, the form of the polyp frequently bears a definite relation to the form of the spines, but this is not invariably the case, and the three types are closely linked together by intermediate forms. The cylindrical type may become compressed and shortened, whilst in the other direction Antipathes Jilix, Pourtales, lorms a link between those forms having simple cylindrical spines and others, such as Aphanipathes pedata and Aphanipathes cancellata, in which the knobbed feature is most marked. We have thus to fall back on the structure of the zooids and ccenenchyma to supply the chief characters, and these are precisely the features which have hitherto received the least attention. The characters to be considered of ordinal value depend to a great extent on whether the Savagliidse are to be included in the Antipatharia. Undoubtedly Savaglia lamarcki has little in common with the Antipathidse beyond the possession of a con- tinuous and branched horny sclerenchyma and a non-spiculate ccenenchyma. Its zooid has the typical Actinian structure, and the system of canals in the ccenenchyma are, so far as is known at present, without parallel in the Antipathidse. A horny sclerenchyma, such as that of Savaglia, is by no means confined to the Antipatharia. The colonial REPORT ON THE ANTIPATHARIA. 73 Is 'anthidaj amongst the Actiniaria have a similar axis, which in some cases, too, becomes tubular. In Savaglia the colonies are larger and more important, and apparently the sclerenchymatous sheath may be continued beyond the foreign substance which at first forms its support, but in both cases the growth is at first parasitic. In the Zoanthidas it apparently never ceases to be so. On the other hand the zooid of Savaglia does not appear related to that of the Zoanthidse. R. Hertwig and Erdmann regard the alterna- tion of macro- and microsepta as the most prominent feature of the latter group, and this condition does not obtain in Savaglia. I have not been able to make sections of this species, and am unable to add anything to the researches of Lacaze Duthiers. Temporarily I have retained it amongst the Antipatharia, but it must be regarded as a genus quite apart, and one which may ultimately be included in the Actiniaria. With the exception of Savaglia the most constant feature of the Antipatharia is the presence of spines on the axial horny sclerenchyma. This is a feature which, so far as I know, is only shared, and then in a modified manner, by one other genus of Anthozoa, viz., Acanthoisis, Wright and Studer. In this genus, the only one amongst the Gorgonidse which appears to approach the Antipatharia in this respect, the axis consists of alternating calcareous nodes and horny internodes ; the internodes have the surface raised in ridges, which are dentate.1 Next in importance, and with the additional exception of the Dendrobrachiidaa (also at present limited to one species), we may consider the simple nature of the tentacles and the absence of a sphincter muscle, as a necessary result of which the tentacles cannot be covered by the upper portion of the body-wall. The former feature was until recently supposed to separate sharply the Zoantharia from the Alcyonaria. We now know, however, that certain families of Actiniaria (Sarcophianthidse and Thalassianthidse) have the tentacles branching or bushy, and thus approach the Alcyonarian type. Dendrobrachia, too, amongst the Antipatharia has pinnate tentacles. The presence of a sphincter muscle, though frequent in the Actiniaria, is not a constant feature. The group includes all grades of differentiation in this respect. Further, and again with the exception of Savaglia, the arrangement of mesenteries in the Anti- patharia is constant, and in all genera yet described (? Dendrobrachia) they may be reduced to one type. The characters which I have considered of generic value refer chiefly to the form of zooid and the number and relations of the mesenteries. The latter have been found to vary from twelve to six in the Antipathidse. The following is the arrangement at present proposed : — 1 Wright and Studer, Challenger Alcyonaria, Zool. ('hall. Exp., pt. lxiv. (vol. xxxi.) p. 45, pi. viii. tigs. 1, In, 16. (ZOOL. CHALL. EXP. PART LXXX. 1889.) Llll 10 74 THE VOYAGE OF H.M.S. CHALLENGER. ANTIPATHARIA, Milne-Edwards. Antipathaeea, Dana. Ceratophyta (pars), Gray. Colonial Zoantharia, possessing a continuous horny sclerobasic axis, which consists of thin concentric lamella? usually enclosing a central canal. The horny axis is usually more or less branched, and is spinose in all known genera excepting Savaglia. The ccenenchyma consists of the fused bases of the zooids ; it is always thin, and never con- tains any spicules proper to it. The sclerobasis may be independent or parasitic ; in the latter case the sclerenchyma forms a sheath around some foreign body. The colony is generally fixed by a basal dilation of the sclerenchyma, but in some cases this is replaced by an elongate flattened hook-like base ending in a point. The sclerench)'ma is probably a secretion of the ectoderm. Family I. Savagliidj, n. n. Gerardidx, Verrill. Antipathes (pars), Auctt. Antipatharia the zooids of which possess 24 simple tentacles and 24 mesenteries, and are connected together by a ccenenchyma possessing a series of interzooidal canals opening into the base of each antimere. Sclerenchyma parasitic, not spinose. The tentacles may be completely covered in contraction by the anterior portion of the body- wall. The zooids have the typical Actinian structure. Savaglia, Nardo, 1843. (= Gerardia, Lacaze Duthiers, 1864.) Family II. AntipathidyE, Verrill (emend.). Antipatharia in which typically the individual zooids have six simple tentacles, but in dimorphic genera three individuals, each possessing only two tentacles, are morpho- logically equivalent to one zooid of the ordinary type. The tentacles are contractile, but apparently never "retractile." Sclerenchyma rarely parasitic, always spinose, and usually possessing a central canal. The ccenenchyma is not generally traversed by a number of interzooidal canals, but the individual zooids are brought into communication by a stolon-like lateral outgrowth of their ccelentera. Six primary mesenteries are always present, two of which occupy the transverse axis and bear the reproductive organs. There may be six or four secondary mesenteries which are always short, or these may be absent altogether. This family corresponds to the genus Antipathes as instituted by Pallas. REPORT ON THE ANTIPATHARIA. 75 Subfamily Antipathin.e, Brook. Zooids not dimorphic, each possessing six tentacles, which may he radiately arranged, or in two rows of three each. There is a tendency for the transverse axis of the zooid to become much elongated in the direction of the horny basis. This subfamily corresponds to the family Antipathidse of Vcrrill, and includes all the species of Antipathida; previously described, of which the zooids are known. It includes the following genera : — Girripaihes (Blainv.), emend. Stichopathes, n. gen. Leiopathes (Gray), M.-Edw. (erne ml.). Antipaihes (Pall.), emend. Antipaihella, n. gen. Aphanipaihes, n. gen. Tylopathes, n. gen. Pteropathes, n. gen. Parantipathes, n. gen. Subfamily Schizopathin.-e, Brook. Zooids dimorphic, each with two tentacles. Of the three individuals morphologically comparable with the unspecialised zooid of Antipathinaj, two are reproductive (gono- zooids) and one is nutritive in function (gastrozooid). The following genera belong to this group : — Scliizopatlies, n. gen. Bathypathes, n. gen. Taxipathes, n. gen. Cladopathes, n. gen. Family III. Dexdeobeachiib^, Brook. Antipatharia the zooids of which have branched " retractile " tentacles. The scleren- chyma is apparently without a central canal and is distinctly spinose. In the younger portions of a colony the sclerenchyma forms irregular plate-like longitudinal ridges varying in number, the angles between which are filled up by a further secretion of sclerenchyma, so that ultimately the axis becomes cylindrical as in the Antipathidas. Dendrobrachia, n. gen. In seeking to establish new genera I have, with one exception, made a microscopic exam- ination of the polyps of the type species. There are, however, so many species of which the polyps are not known that it is impossible at present to refer all to their jDroper position. Structural characters have not previously been taken into consideration, and authors have 7