iiippiiiiiiiiiip^ >:Mi;'t \CZ-~C HARVARD UNIVERSITY LIBRARY MUSEUM OF COMPARATIVE ZOOLOGY GIFT OF '^sWlc^o^ '\,\'M>% MEMOIRS OF THE MUSEUM OF COMPAEATIVE ZOOLOGY AT HARVARD COLLEGE VOL. LV CAMBRIDGE, MASS., U. S. A. IPvinteJ) tor tbe /IDuseum 1938 MEMOIRS OF THE MUSEUM OF COMPAKATIVE ZOOLOGY AT HARVARD COLLEGE VOL. LV CAMBRIDGE, MASS., U. S. A. printeC* for tbe /iDuseum 1938 c^ /IDemotrs of tbc /iDuseum ot domparative Zooloas AT HARVARD COLLEGE Vol. LV ECHINODERMS FROM AUSTRALIA AN ACCOUNT OF COLLECTIONS MADE IN 1929 AND 1932 WITH 28 PLATES (16 COLORED) AND 64 TEXT FIGURES BY Hubert Lyman Clark CAMBRIDGE, U.S.A. printeo for tbe /iDuseum 1938 FEB 9 1938 CONTENTS Introduction Crinoidea . Coniasteridae Comatella . Capillaster Comatulella Valida Comatula Comanthina Comantheria Comanthus Zygometridae Zygometra . Himerometridae Heterometra Amphimetra Mariametridae Lamprometra Colobometridae Petasometra Austrometra Oligometrides Calometridae . Neometra Aporometridae Aporometra Antedonidae . Compsometra Dorometra Monilimetra Asteroidea Astropectinidae Astropecten Luidiidae Luidia Archasteridae Archaster Goniasteridae Nectria Tosia Pentagonaster Goniodiscaster Styphlaster Iconaster Anthenoides Page 3 12 14 14 15 15 16 18 21 23 26 29 29 33 33 35 35 35 36 36 40 40 41 41 41 41 45 45 46 47 57 60 60 71 71 74 74 77 77 78 79 SO 88 91 91 IV CONTENTS Stellaster Asterodiscus Notes on the Subfamily Antheneinae Anthaster Pseudanthenea Pseudoreaster Gymnanthenea Anthenea Non-Australian Antheneas Australian Antheneas Oreasteridae . Protoreaster . Pentaceraster Culcita . Ophidiasteridae Austrofromia Nardoa Linckia . Bunaster Leiaster Ophidiaster Hacelia Tamaria Pseudophidiaster Asteropidae . Petricia Asterinidae Asterina Disasterina Manasterina Paranepanthia Patiriella Nepanthia . Parastcrina Anseropoda Echinasteridae Echinaster Plectaster . Metrodiridae Metrodira . Pterasteridae Retaster Asteriidae Coscinasterias Astrostole . Allostichaster Smilasterias Uniophora Paqe 94 101 101 102 103 103 105 110 114 116 130 130 131 132 132 132 132 133 134 136 137 139 139 142 142 142 143 143 156 157 159 163 169 179 181 183 183 186 187 187 188 188 189 189 191 193 195 196 CONTENTS Ophiuroidea Ophiomyxidae Ophiomyxa Trichasteridae Astrobrachion Euryale Gorgonocephalidae Asteroporpa Astroconus Astrocladiis Astrochalcis Ophiacanthidae Ophiacantha Ophiothaiima Aniphiuridae Ampliiura Ophicentrus Ophionephthys Amphipholis Amphistigma Amphiodia Amphioplus Ophiactis Ophiodaphne Ophiotrichidae Ophiotlirix Lissophiothrix Macrophiothrix Ophiotrichoides Placophiothrix Ophiopteron Ophiomaza Ophiocnemis Ophiothela Ophiogyrana Ophiochitonidae Ophionereis Ophiocomidae Ophiocoma Ophiomastix Ophiarthrum Ophiodermatidae Ophiiirodon Opliiarachna Pectinura Ophiochasma Ophiaraclinella Cryptopelta Page 196 201 201 202 202 202 204 204 205 206 208 208 208 211 213 213 236 240 243 244 247 249 256 270 271 271 275 278 305 309 317 317 318 319 320 321 321 328 328 337 339 339 339 341 343 345 346 354 VI CONTENTS Ophiodyscrita Ophiolepididae Ophiomisidium Haplophiura Amphiophiura Ophiura Dictenophiura Ophioteichus Ophiolepis . Ophioplocus Echinoidea Cidaridae Goniocidaris Prionocidaris Phyllacanthus Centrechinidae Centrechinus Centrostephanus Echinothuriidae Asthenosonia Temnopleuridae Temnopleurus Salmacis Temnotrema Mespilia Microcyphus Amblypneustes Holopneustes Echinidae Pseudechinus Tripneustes Nudechinus Strongylocentrotidae Echinostrephus Pachycentrotu.s Heliocidaris Echinometridae Parasalenia Echinometra Clypeastridae Clypeaster Hesperaster Arachnoifles Laganidae Laganum Peronella Fibulariidae Fibularia Paqb 355 357 357 357 358 358 359 360 364 365 366 369 369 369 372 376 376 377 378 378 380 381 383 385 388 389 389 391 395 395 397 398 402 402 403 404 407 407 408 410 410 411 414 417 417 417 421 421 CONTENTS Echinocyamus Scutellidae Echinodiscus Echinoneidae Echinoneus Nucleolitidae Apatopygus Cassidulidae Echinolampas Palaeopneustidae Archaeopneustes Hemiasteridae Protenaster Hypselaster Moira Spatangidae Brissopsis Rhynobrissus Brissus Eupatagus Gonimaretia Maretia Breynia Lovenia Echinocardium Holotliurioidea Cucumariidae Cucumaria Pentacta Leptopentacta Pseudocolochirus Pentathyone Thyone Orbithyone Staurothj'one Actinocucumis Pseudociicumis Phyllophorus Lipotrapeza Urodemas . Psolidium Psolus Holothuriidae Stichopus Labidodemas Holothuria Actinopyga Molpadiidae . Page 422 423 423 424 424 425 425 428 428 429 429 429 429 43) 432 433 433 433 435 436 437 437 438 439 440 441 444 444 445 453 456 458 461 473 474 477 481 485 494 497 503 507 511 511 516 516 538 538 vm CONTENTS Molpadia Paracaudina Aphelodactyla Synaptidae Chondrocloea Leptosynapta Protankyra Chiridota Taeniogyrus Trochodota Local Lists Lord Howe Island New South Wales Queensland Northern Territory Augustus and Champagay The Broome Region The Perth Region Albany, W. A. Coasts of South Australia Tasmania Bibliography Index Explanation of Plates Islands and Mctoria Page 538 540 542 545 545 548 552 555 557 557 559 559 562 563 564 566 567 569 571 571 572 574 582 593 ECHINODERMS FROM AUSTRALIA ECHINODERMS FROM AUSTRALIA INTRODUCTION Unusual opportunities for studying and collecting echinoderms on the coasts of Australia in recent years warrant the publication of this account of the material accumulated. In 1929, a liberal grant from the Carnegie Institution of Washington, supplemented by a generous gift from the Australian National Research Council and aid from the Museum of Comparative Zoology, enabled me to spend nearly six months in active collecting in AustraUan waters. Landing at Darwin, N. T., June 13, we carried on work in that vicinity until the end of July, both by shore collecting and by dredging in shallow water. Mrs. Clark accompanied me and Mr. Arthur A. Livingstone of the Australian Museum, Sydney, joined us at Darwin, remaining with us for some four months. Mrs. Clark made water color sketches from Ufe of many of the more interesting echinoderms and frequently aided in the shore collecting. Mr. Livingstone devoted himself to assisting me in every possible way and I cannot speak too highly of his invaluable coopera- tion. I do not know how to adequately express my appreciation of, and my gratitude for, the indefatigable assistance of these two fellow-workers. But work at Darwin would have been seriously handicapped without the cooperation of the hospitable residents of that community. The chairman of the Northern Territories Commission, Mr. James Horsburgh, most generously gave us the use of his residence, on the very shore of Port Darwin, for a laboratory, and a more attractive and satisfactory place for such work is not to be found on the northern coasts of Austraha. In every way possible, Mr. Horsburgh helped us, and showed his unfaihng interest in our work throughout our stay of more than six weeks. To Colonel Robert H. WeddeU, the Resident Commissioner at Darwin, our thanks are also due, for such aid as he could give. Shore collecting in the immediate vicinity of Darwin is very poor and it was necessary to make frequent visits to East Point, six miles out from town, in order to determine what the normal intertidal and low-tide fauna of the region really is. Mr. Jack Wetter was engaged to pro\dde transportation for these excursions and soon became a constant and invaluable helper in many 4 memoir: museum of comparative zoology ways. Through him we were enabled to get a motor boat and thus could carry on dredging in Port Darwin. Without this we should have left Darwin with a meagre and inaccurate idea of the echinoderms which occur there. My obhgations to Mr. Wetter are therefore very great and very real and I am glad to thus express my appreciation of his services. To Mr. and Mrs. F. A. K. Bleeser, our most sincere gratitude is due for much helpful advice and countless favors, evidence of a friendship which has enriched life ever since. Thanks are due, and here gladly extended to the numerous dwellers at Darwin who helped us with gifts and loans of specimens, and suggestions as to good collecting grounds. Especially we thank Mrs. Gordon, our kind hostess at the Victoria Hotel, Mr. P. W. Wilson, Mr. and Mrs. Joseph Hardy, Mrs. Jessie Littlefield, and Mr. E. J. Foster. From Darwin, we travelled westward to Broome, by the Western AustraUaa government's boat, the "Koolinda." During this journey, we had the opportunity of visiting two ports which appear on an ordinary map to give promise of being good collecting grounds. The first of these is Wyndham, at the head of Cam- bridge Gulf, where it is not even necessary to land to reaUze its utter unfitness as a home for echinoderms. The amount of mud in the water of Cambridge Gulf is simply amazing and the absence of suitable reefs and shores is perfectly obvious. The second port of call was Derby, at the head of King Sound. The entrance to the Sound is marked by attractive islands and gives promise of good collecting but the vicinity of Derby is as hopeless as Wyndham, barren mud flats and mangrove swamps stretching away east and west as far as eye can see. Coming out of King Sound and rounding Cape Leveque, one enters a very different area and the clear water and numerous shore reefs as Broome is ap- proached give promise of an abundant marine fauna. This promise is not behed ! Broome is bej^ond doubt a paradise for a marine zoologist. The clear water, the firm bottom of Roebuck Bay, the reefs and tide- pooLs at Entrance Point and Gantheaume Point, the very great rise and fall of the tide (exceeding tliirty feet in September), combine to provide conditions extraordinarily favorable to animal hfe. Sponges, worms, echinoderms (especial- ly holothurians), crustaceans (especially crabs), mollusks, and ascidians abound not only in number of individuals but in diversity of genera and species. Shore collecting anywhere, at almost any time except high water, is more or less re- warding, but at very low water surpasses description, while dredging constantly yields a wealth of forms of amazing variety. We had planned to stay a month at Broome — we stayed two — and left with the keenest regret, well conscious that we had not begun to exlaaust the possibihties of the region. CLARK: AUSTRALIAN ECHINODERMS 5 We met with the same kindness and cordial hospitality at Broome to which Darwin had introduced us. Everyone was willing or eager to help and it is impossible even to mention all those to whom our thanks are due. Captain A. C. Gregory, whose extensive pearl fisheries make him a well-known figure on the northwest coast, generously provided us, without cost, a building equipped with electricity and running water, located almost at liigh water mark and very convenient for the nearby collecting grounds. To express here, as I gladly do, sincere thanks for this indispensable help is but a meagre return. I wish to thank particularly also, Mr. Reginald A. Bourne, master of the pearling lugger "Bonza" who became greatly interested in our work and rendered invaluable assistance in the matter of dredging and visiting more distant places. Thanks to the "Bonza" we collected at several points on both sides of La Grange Bay and even as far west as Wallal. Mr. Bourne also made it possible for us to visit pearUng vessels at work and enUst the aid of their divers in bringing up specimens from the bottom where they were gathering shell. The Resident Commissioner, Colonel W. 0. Mansbridge, gave me some valuable sea stars from the Lacepede Islands which were secured by pearlers while working in that vicinity, an area we could not visit. Among the many to whom thanks ought to be expressed are Captain Beresford E. Bardwell, Mr. W. F. Clarke, Mrs. W. H. Milner, Mrs. S. V. Ogilvie, and Mr. and Mrs. Stanley Pigott. During the weeks we were at Broome, we were able to make a most inter- esting and profitable visit to the lighthouse at Cape Leveque where we were hospitably entertained for three days by the keepers, Mr. and Mrs. Newman and Mr. and Mrs. Robson, to whom our thanks are here cordially tendered. The success of this excursion was largely due to the hearty cooperation of Mr. Jack Tytherleigh, who took us in his car, and to Mr. and Mrs. Stanley Dewar, then Uving at Beagle Bay, at whose suggestion the visit to the Cape was made. Thanks are due Mrs. Dewar also for her help in the shore collecting. We left Broome, October 1, for Perth and en route stopped at Onslow, where a few hours hunting along the shore yielded no echinoderms; at Cossack, which looked much more promising and is reported to be an excellent collecting ground ; at Carnarvon, where conditions looked far from favorable; and at Geraldton, where a full day's work along the shore both north and south of the jetty yielded a number of interesting echinoderms. We reached Fremantle, October 8, and next day, at Perth, entered upon the contacts with Australian Museums and scientific men, which have proven so pleasant and so rewarding. The generosity and cheerful helpfulness of my colleagues in the Commonwealth have been so 6 memoir: museum of comparative zoology great and so delightful that any adequate expression of gratitude is out of the question, but I must bear testimony to them as enthusiastically as I can. At Perth, where I had planned to stay but two days, Mr. L. Glauert of the Museum and Dr. E. W. Bennett of the University of Western Australia made three weeks pass all too quickly. Shore collecting at Point Peron and Rottnest Island, dredging in the area between Fremantle and Garden and Rottnest Islands, and a two days excursion to Bunbury and Koombana Bay, yielded a surprising number of echinoderms, including several new to science. To add to these treasm'es, Mr. Glauert permitted me to search the Museum shelves and Dr. Bennett laid before me the accumulations in the Biological Laboratory and Museum at the University, and either as gifts or loans everything of interest was unreservedly added to my store. Besides these two bountiful friends. Dr. N. T. M. Wilsmore, Professor of Chemistry at the University, and Mr. Duncan Swan and Mr. D. L. Serventy, students in the Biological Department, made me their lasting debtor. It is a pleasure to offer my very sincere thanks to the Honorable P. Collier, Premier of Western Australia, for a pass on the railroads of that state for myself and an assistant, and to Mr. F. Aldrich, Chief Inspector of Fisheries, for the use of the Department's boat for my dredging excursions. Without this generous help, so cordially given, I could not have accomplished half as much during my delightful three weeks. Leaving Perth, by the famous transcontinental train, on October 21, we enjoyed the three days journey across Western Australia and the extraordinary Nullarbor Plain, into South Australia, to the pleasant city of Adelaide. A week's stay here was made profitable through the cordial assistance of Mr. Herbert M. Hale, who enabled me to enjoy the very unusual experience of col- lecting living echinoderms (at Port Willunga) in the morning and fossil echini (at Blanche Point, Maslin Bay) in the afternoon. Mr. Hale also loaned me many important specimens which were needed for further study, and enabled me to examine at the Museima much valuable material both recent and fossil. From Adelaide we went on to Melbourne, where Mr. J. A. Kershaw of the Victorian National Museum and Dr. Frederic Chapman, the Commonwealth Palaeontologist, were most cordially helpful in enabling me to see much im- portant material. Mr. Kershaw very kindly sent to Cambridge as a loan for critical study a number of interesting echinoderms. From ^lelbourne, I went to Hobart, where Mr. CUve Lord, Director of the Tasmanian Museum, whose recent death is such a loss to Australian scientific circles, did everytliing in his power to make my brief visit deUghtful and profitable. Thanks to the generous CLARK: AUSTRALIAN ECHINODERMS / cooperation of the Chairman of the Sea Fisheries Board, Col. J. E. C. Lord, a day was spent dredging in the estuary of the Derwent and I was thus enabled to see a fauna as different as can be from that of the vicinity of Broome. Professor T. T. Flj'nn of the University, now at Queens University, Belfast, Ireland, was most generous in presenting me with much valuable material which he had himself collected at several points on the Tasmanian Coast. To all of these colleagues, I offer my best thanks. Our next stop was at Sydney, where ten very profitable and delightful days were spent in collecting in Port Jackson, at Long Reef and at Gunnamatta Bay, and in studying material at the Australian Museum. Dr. C. Anderson, Director of the Museum, did everything possible to aid me and to demonstrate the hos- pitality, wealth and efficiency of the great institution of which he is the honored and capable head. My debt to him and to the splendid and most friendly staff of the Museum can never be adequately set forth — had I the honor to be one of their own number, I could not have been given more or better aid. My par- ticular debt to Mr. Arthur A. Livingstone, the Curator of Echinoderms, has al- ready been mentioned. Mr. Melbourne Ward was most generous in placing at my disposal his motorcar and motorboat during my stay in Sydney. As he also took an active part in the collecting done, my obligation to him is very great. It is a pleasure to express here my heartiest thanks to all my colleagues at Sydney. As our return to the LTnited States was via China and Japan, it was possible to spend a day in Brisbane, and so visit the Queensland Museum and enjoy the hospitaUty of Mr. Heber A. Longman, the Director. ^Vhile the collection of echinoderms is not extensive, there is much noteworthy material on display, especially a series of Oreasters from the northern Queensland coast. Like his colleagues at the other Australian Museums, Mr. Longman was more than will- ing to loan material and assist my work in every possible way. I offer him my very sincere thanks. After my return to Cambridge in 19.30, an expedition to Australia was planned at the Museum, for the purpose of adding to our collections of the land fauna, particularly vertebrates and insects. It was not intended to include the marine fauna, but for various reasons, which need not be discussed here, it was decided that I should join the party in Sydney, in March 1932. Thanks to a generous grant made by the Milton Fund of Harvard University, it was possible for me to carry out three projects which were not possible in 1929, and which seemed quite essential to a proper rounding out of my studies on Australian echinoderms. These projects were, visits to Lord Howe Island, in the Tasman 8 memoir: museum of comparative zoology Sea, where the southernmost coral reefs of the world occur, and to Port Essing- ton, on the Coburg Peninsula of the Northern Territory, where the first British settlement on the northern coast of the Continent itself was attempted; and a month of dredging at Broome, to exploit the very rich echinoderm fauna found there in 1929. At Lord Howe Island, thanks to the courtesy of the Trustees of the Austra- lian Museum, and the characteristic cooperation of Dr. Anderson, I had the as- sistance once more of Mr. Arthur A. Livingstone. During the three weeks, which we spent on the island, Mrs. Clark, Mr. Livingstone and I, often aided by vol- unteer assistants, collected all of the 18 species of echinoderms previously known from the Lord Howe reefs and beaches, and 36 others, including several new to science. A brief discussion of the echinoderm fauna of the island will be found on pp. 559 of this report. Particular thanks are due to Mr. Robert Bax- ter, Mr. and Mrs. Reginald Kirby and Miss Kama Birmingham for their very kind, and often indispensable help in our work at Lord Howe. The main purpose of the visit to Port Essington was to examine the littoral echinoderm fauna of the Coburg Peninsula and see whether it is similar to that of Darwin and equally poor, or whether the poverty of the fauna at Darwin is due to some local conditions. It was of course necessary to use Darwin as the base from which this excursion could be made, and we were extraordinarily fortunate in the combination of circumstances which gave us the required amount of time between our arrival on the monthly boat from Sydney and the departure of the bimonthly Western Australian steamer for Broome; which offered the best possible condition of the tide (for that time of year); and which enabled us to secure the services of Captain Jack Hayles of the motorship "Maroubra" whose hearty interest and cooperation in our work could be exceeded only by his knowledge of the coast and ability as a seaman. To him I am glad to offer my sincere thanks for making the excursion well worth the trouble and expense. My friend, Mr. F. A. K. Bleeser of Darwin went with me and proved an inval- uable assistant in the collecting as well as a most congenial companion. We made three landings on the peninsula during favorable conditions of the tide — one near the lighthouse at Cape Don, one at Smiths Point, at the entrance and on the eastern side of Port Essington, and one at Coral Bay on the western side. Best of all we had several hours at Allaru Island, a small islet lying well off shore to the west of Port Essington. At none of these points was the collecting notable in itself, but the similarity to that in the vicinity of Darwin was so obvious, there was no doubt left in my mind that the comparative poverty of the fauna CLARK : AUSTRALIAN ECHINODERMS 9 at Darwin is not due to local conditions but is characteristic of the whole coast of the northern territory. The month of dredging at Broome was all that had been expected and more, the chief drawback being the weather which was very different from that we had so thoroughly enjoyed in August and September, 1929. The winter days of June were regrettably short, allowing only about eleven hours of daylight in which we could work. There was much overcast sky and some rain, and on many days there was too much wind to permit dredging very far off shore. This prevented a greatly desired visit to the Lacepede Islands, northeast of Broome, but Mr. Bourne and other local authorities do not think that conditions at those islands differ in any essential way from those along the coast, or that there is any reason to expect any differences in the fauna. The pleasure and success of the month's work were very largely due to the hearty cooperation of Mr. R. A. Bourne, master of the "Bonza," who had done so much to help me in 1929. In addition to fulfilhng his contract with me satisfactorily in everj^ detail so far as weather permitted, Mr. Bourne provided at his own expense, a diver and the necessary pump and accoutrements to assist us. While this was an advantage in some ways, and the diver brought up many desirable specimens including species we did not and probably could not hope to get with a dredge, it involved a very great waste of time, for a diver necessarily works very slowly, covers a small area and sees only relatively large and noticeable forms. Mr. Bourne and his admirable crew of three aboriginals ("abos," for short) Paddy, Ramy and McKinna, were so assiduous in dredging that we covered a great deal of ground when not delayed by the diver. We left Broome each Monday, as early as wind and weather permitted and returned with our sjjoil late Saturday evening. If there is a more fascinating sport than dredging under pleasant conditions on good bottom, I do not know what it is, and those weeks working up and down nearly three hundred miles of coast, east and west of Broome, will always be looked back upon as among the most thrilling of my life. My thanks to Mr. Bourne are most deep and sincere. The friends we had made in 1929 were most cordial in helping us in every way, and it is a pleasure to thank them again. It may appropriately be mentioned here that Captain Beresford E. Bard- well in 1933 made a trip to the northeast of Broome, going as far as Augustus Island and the Port George Mission. In return for a little financial help from the M.C.Z., Captain Bardwell sent us a notable collection of echinoderms, including a number of species which I did not secure, and throwing much light on the fauna east of King Sound. This collection is included in the present report and is 10 memoir: museum of comparative zoology particularly discussed on pp. 566. My heartiest thanks are here tendered Captain Bardwell for this timely and important assistance. After leaving Broome early in July, there was little opportunity for further collecting. Our first stop on the southward journey, that at Port Hedland, per- mitted several hours ashore and as the tide was favorable, it was possible to gather a few interesting echinoderms. But our other ports were made late in the day and no further chance for shore work occurred before reaching Perth. There I was again the recipient of courtesies, loans and gifts from Messrs. Glauert, Bennett and Wilsmore, and also from Professor G. E. Nichols of the Zoological Department of the University, who was indefatigable in his service and unceasing in his kindness. The winter weather was not propitious for collecting and our single dredging trip was badly handicapped by wind and rain. The hospitality of Mr. and Mrs. Joseph Hardy, whose friendship we had made at Darwin in 1929, was so generous and delightful during our two weeks stay in Perth, it cannot be passed over in silence, but words of thanks are quite inadequate, especially since the memory of it is so unspeakably saddened by the lamentable death of our young hostess only a year later. From Perth, all the collections made at Darwin and on the Coburg Penin- sula, as well as the large amount of material gathered at Broome, were sent to Sydney, and there, by the courtesy of the Australian Museum, they were re- packed with the Lord Howe Island collections and with loans from the Museum, and all were sent in one shipment to Cambridge. For this great assistance, I take pleasure in thanking the Trustees of the AustraUan Museum, Dr. Anderson and the very efficient and kind packer, Mr. Henry S. Grant. The entire amount of material thus accumulated from my two visits to Australia consists of 11,484 specimens, representing 422 species of 184 genera. Of this, 9,647 specimens were collected by Mrs. Clark, Mr. Livingstone and myself, aided of course by many others, particularly by Mr. Bourne at Broome. The balance is made up of 955 gifts and 700 loans from our bountiful Australian colleagues and friends and the 182 sent by Captain Beresford E. Bardwell of Broome, as already mentioned. In the following pages this notable mass of material is treated first systematically by classes, and then geographically by areas visited. In genera containing representatives of more than one species, previously known forms are treated first, then the new species; in each group the species are arranged alphabetically. The source of the material is mentioned under each species. Wlien no other source or collector is mentioned, it is to be understood that we were the collectors. It is hoped that a subsequent report CLARK: AUSTRALIAN ECHINODERMS 11 may give a full account of the echinoderm fauna of Australia as at present known, with a discussion of its apparent history and relationships. I wish to express here my deep sense of obUgation, and my sincerest thanks, to the Carnegie Institution of Washington, the National Research Council of Austraha, and the Committee in charge of the Milton Fund of Harvard Uni- versity for the financial help which made the collecting of the material and the preparation of colored plates possible; to Mr. George R. Agassiz for the finan- cial help, which made possible the colored plates and prompt pubUcation; to my wife, Frances L. S. Clark, and to Eugene N. Fischer, for their patience and skill in making the colored illustrations for this report so satisfactory; to Mrs. Clark further for much help in the preparation of the manuscript, particularly the index, bibliography and faunal lists; to Mr. F. P. Orchard for his admirable efficiency in making the necessary photographs; to Mrs. EUzabeth Grundy, Mrs. Carmen Witter and Mrs. Anna O'Connor for their skill and care in tjrping the manuscript; to Mrs. Marjorie H. Pattee for invaluable help in connection with checking references and catalog numbers, and tabulating statistics; and partic- ularly to Dr. Thomas Barbour, Director of the Museum of Comparative Zool- ogy, whose generous interest and unfaihng sympathy have made it possible for me to enjoy the extraordinary opportunities I have been given, and to prepare unhindered by routine duties, this voluminous report. 12 memoir: museum of comparative zoology CRINOIDEA The collection of crinoids contains 910 specimens representing 20 genera and 32 species. Of the genera, two are new so far as the name is concerned, but one is proposed for a species described many years ago whose generic position has hitherto been a puzzle. Of the 32 species only 7 are new and of these 4 make up the new genus of Antedonidae discovered at Broome. It is not strange that the number of new species is small, for much attention has been given to AustraUan and East Indian crinoids during the past thirty years. The new genera are: Aporometra, type Himerometra paedophora H. L. Clark. There are two other species in the genus, one of which is described beyond. Monilimetra, type M. nomima sp. nov. There are three other species in the genus, all found at Broome and hitherto undescribed. The 7 new species fall into three genera as follows: Petasometra brevicirra W. A., False Cape Bossut. variegata N. T., Darwin. Aporometra occidenialis W. A., Bunbury Monilimetra bicolor W. A., Broome. lepta W. A., Broome. nomima W. A., Broome. poecila W. A., Broome. Crinoids are common, even abundant, in many places on the tropical coasts of Australia but below latitude 27° few species occur on either the eastern or western sides of the continent, or on the southern coast. Nor are these few species of the temperate coasts notable for either size or color, for they are small and only one is at all showy. Hence thej^ make up an inconspicuous part of the marine fauna. On some parts of the northern coast, however, and on the north- ern part of the Great Barrier Reef, comatulids are abundant and conspicuous, even in shallow water and along shore near low water mark. But they are seldom found where the water contains any silt or there is any deficiency in either salinity or oxygen. Consequently they occur in abundance and in all their full beauty only where the water is kept in the best possible condition by strong tidal or other currents. They are therefore apt to be more or less local in their distri- CLARK: AUSTRALIAN ECHINODERMS 13 bution. Most comatulids live a sessile life, moving about little or not at all, but a few of the smaller species (especially in the Antedonidae) are good swimmers if sufficiently stimulated. While some species occur, often in abundance, on open, hard, sandy bottoms, where they attach themselves to any object, rock, shell or alga, fixed to the bottom, most of the larger forms occur in sheltered recesses under or among rocks near and below low water mark, or among sponges, gorgon- ians and stony corals. Of the 32 species in the present list, 21 occur in the Broome region, and 14 were found only there. At Darwin, we found 8 species but 6 of these occur also at Broome and at least 5 occur on the Barrier Reef. Only 5 species are at hand from the western Australian coast. Obviously Broome is a particularly favored spot but this is quite to be expected in view of the purity of the sea-water and the great rise and fall of the tides. The literature dealing with Australian crinoids is almost wholly the work of Austin Hobart Clark, with minor contributions by Gislen (1919) and the present writer (1909, 1914, 1915, 1916, 1921, 1928, 1932). Mr. Clark's contri- butions to our knowledge of the East Indian fauna is equally valuable and his monograph on the "Siboga" comatulids is a "sine qua non" for any worker on the crinoids of the Indo-Pacific region. His recent magnificent monograph on the Comasteridae (1931) brings order out of chaos in a masterly way and lays a foundation for all future work with that perplexing family, which is not likely to be essentially altered in years to come. In the study of the present collection I have enjoyed the inestimable ad- vantage of frequent consultation with Mr. Clark either by personal conference in Cambridge and Washington or by correspondence. His open-minded readi- ness to see my side whenever we disagreed has made it possible for us to discuss debatable points in a very satisfactory manner, and his extraordinary knowledge not only of the animals themselves but of the Hterature dealing with them, has enabled him to save me from oversights and blunders. Of course, Mr. Clark cannot fairly be held responsible for any errors that may be found in the present report but for whatever merit it may possess much of the credit should be given to him. It is a pleasure to offer here my sincerest thanks to my friend of more than twenty-five years, for his ready and generous assistance. In collecting crinoids at the Murray Islands in 1913, I soon learned that their reactions to chemicals were very different from those of other echinoderms, and the experiences of 1929 and 1932 have confirmed the impressions formed at that time. No narcotic has been found which will cause them to relax, as Epsom 14 memoir: museum of comparative zoology salts do for other echinoderms, but no real ei'fort has been made to find such a substance as it is quite unnecessary. All of the comatulids which I have taken alive kill quickly and satisfactorily, without autotomy or distortion, in strong alcohol. If it is desired to have them die with the disk exposed, they should be lifted carefully from the sea water and plunged into alcohol deep enough to cover them, in a pan or flat dish large enough to permit the full expansion of the arms. It does not much matter whether they be plunged with oral side down or up, but it is more convenient to seize the centrodorsal with thumb and finger (or forceps) and put them in the dish of alcohol, mouth down. They should be held down for a few moments till all movement has ceased, after which the arms can be arranged as desired. If the specimens are to be dried, as is often desirable for convenience in packing and in subsequent storage and study, a httle corrosive sublimate should be put in the alcohol. After thorough saturation with such poisoned alcohol the specimen may be dried as convenient by placing, mouth down, on a sloping board. For the best results, the killing should be done very soon after capture. Comatulids are surprisingly sensitive to stale sea-water and specimens left in a bucket for even an hour or two may begin to go to pieces, shedding first the distal pinnules and then the tips of the arms; finally the middle and basal parts of the arms break up. Once this process begins it is difficult to stop it, and such individuals will often go to pieces when placed in alcohol. COMASTERIDAE COMATELLA MACULATA Adinometra maculata P. H. Carpenter, 1888. "Challenger" Comat., p. 307. Comatella maculata A. H. Clark, 1908a. Smithson. Misc. Coll., 52, p. 207. This widespread species occurs at Broome but is not very common. We also found it at Cape Leveque but only secured two small specimens. Some of the specimens taken at Broome are of very large size, the arms being 100 mm. or more in length; in the largest specimen, which has 20 arms and in the dry con- dition is nearly or quite black, the longest arms are 150 mm. long. The cirri are XXIII, 18-22. The smallest specimen has 17 arms, 50-60 mm. long, while one a trifle larger has only 13 arms; in this latter specimen the cirri are XX, 16-17. Most of the specimens have more than 23 cirri, but the number of seg- ments is only 16-20. The smaller specimens give evidence of a dark green tint CLABK: AUSTRALIAN ECHINODERMS 15 which becomes yellowish distally on the arms. The cirri are, in all cases, more or less purple at least distally. The material at hand consists of the following 10 specimens. Western Australia : Broome, under rocks near low water mark, Entrance and Gantheaume Points, August and September, 1929. 1 specimen. Broome, dredged, June, 1932. 2 small specimens. Cape Leveque, under rocks, August 1929. 2 specimens. Capillaster multiradiata Asterias multiradiata Linne, 1758. Sys. Nat., ed. X, p. 663. Capillaster multiradiata A. H. Clark, 1909. Vid. Med., p. 134. This crinoid is characteristic of the East Indian region, ranging only (so far as we yet know) to the Maldive Islands on the west and the Carolines on the east, to Formosa on the north and tropical Australia on the south. There are numerous Australian records but the immediate vicinity of Broome furnished all of the ten specimens at hand. There multiradiata occurs under or among rocks near low water mark and is also occasionally dredged on the shoals, but it is a relatively uncommon and inconspicuous form. The most noticeable of the present series is one with 16 arms, most of which are about 75 mm. long; there are four II Br series and two III Br. A smaller individual, with most of the arms broken, has only 13, and there are but two II Br and one III Br series. The largest specimen has 42 arms about 100 mm. long; the cirri are XXIV, 24-26. In other specimens the cirri range from IX to XXIII and the number of segments from 15 to 26. Most of these specimens of multiradiata appear gray, fawn-color or hght brown dorsally, much darker orally; the pinnules are some- times finely spotted with silvery white. One specimen is a dull, dark purple, and the cirri in most specimens are more or less purplish near the tip; in other cases they are oUve-green. Comatulella brachiolata Comatvla brachiolata Lamarck, 1816. Anim. s. Vert., 2, p. 535. Comatulella brachiolata A. H. Clark, 1911. Amer. Jour. Sci. (4) 32, p. 1,30. This little comatulid is one of the characteristic marine animals of south- western AustraUa. We met with it only at Bunbury but there is a specimen at 16 memoir: museum of comparative zoology hand from Cottesloe Beach. Dredging in Koombana Bay, one or two miles west of the breakwater in 5-8 fms., Oct. 26, 1929, yielded a number of specimens and my field notes say: This "Comatula was largely a rosy-red in hfe, dorsally more or less variegated — the cirri a very bright red, more scarlet, but there is not a sharp contrast between cirri and the calyx." "Rather a rigid thing, came up in dredge, opened out flat, attached by cirri and even pinnules also, to sea- weeds. Killed usually without a tremor!" (The killing was done by plunging at once into strong alcohol). The 1 1 specimens at hand were taken as follows : Western Australia: Cottesloe Beach. Property of Perth Museum. 1 fine speci- men. Bunbury. Collected and presented by Naturalists Club, Bunbury; drift on Back Beach, January, 1930, 1 badly damaged specimen. Bunbury, Koombana Bay, 5-8 fms., October 26, 1929, E. W. Bennett and H. L. Clark leg. 9 fine specimens. Validia rotalaria ComatuJa rotalaria Lamarck, 1816. Anim. s. Vert., 2, p. 534. Comatula (Validia) rotalaria A. H. Clark, 1918. "Siboga" Uiist. Crin., p. 6. This strikingly handsome crinoid is one of the characteristic species of tropical Australia. We met with it while dredging near Broome in 1929 and again in the same region in 1932 but did not take it elsewhere. It has been reported from as far south as Port Molle, Queensland (20° 30' S.) and Baudin Island, Western Australia (14° 08' S.). The specimens taken at Broome there- fore extend the known distribution considerably to the southwest. They range from an immature individual with 22 arms, of which 13 are (or were) 30-50 mm. long and 9 are 5-16 mm., to a fine adult also with 22 arms, but having them 75-125 mm. long, nearly all over 100. The smaller specimen has 2 cirri with about 15 segments, but the larger has no trace of cirri or of cirrus sockets. An- other small specimen has 20 arms, 40-70 mm. long (mostly about 60) but has the cirri VIII, 14-18, interradial in position as in Comatula pcctinata var. pur- purea. Both the small specimens of rotalaria show the fleshy knobs on the disk so often conspicuous in larger specimens, but curiously enough neither shows the least similarity in arm arrangement to Comatula etiwridgei A. H. Clark, wliich is now beUeved by Mr. Clark to be the young of rotalaria and the types CLARK: AUSTRALIAN ECHINODERMS 17 of wliich had the longer arms about 70 mm. Apparently not all rotalarias pass through an eiheridgei stage or else they pass through it while still very small. Some of the specimens dredged at Lagrange Bay in 1932 are superb individuals with 30 or 31 arms, 140 mm. long. In specimens witli 29 arms or more, there are 6 arms on each ray (or on all but one) with the III Br series always on the inner side of the II Br axillary. The color of the alcohohc specimens is a grayish brown lightest on the centrodorsal and on the backs of the arms, darkest orally. The dry specimens are very different : the calyx and basal part of arms is yellow-brown with a dis- tinctly greenish cast, passing more or less gradually into a dull rose red or deep pink; in one specimen only the distal part of the arms is distinctly pink but in all of the others, the general appearance of the whole specimen, excepting only the calyx and oral face of disk, is strikingly deep pink. In the young specimens, the cirri are a very light brown or fawn color. In hfe, the colors are very different and the following extracts from my field notes may be quoted as showing the impression made on the collector when he first meets with this splendid comatuHd. In 1929, I wrote: "No cirri. 20 arms; yellow or greenish on centrodorsal and out on base of arms dorsally, but pinnules silvery dorsally, more or less yellow-tipped; sutures dark, more or less purplish; oral surface very dark. Exquisite." On June 7, 1932, we were dredging in 4-5 fms. at the entrance to Lagrange Bay and took several specimens of rotalaria. My field notes say: "A superb creature marvellously variegated with black and white orally, greenish- yellow aborally. No cirri. 20 or more arms. . . . Dry specimens show complete change of color." It seems to me that the differences between rotalaria and the ten-armed species of Comatula are sufficiently important to warrant treating the sub- genus proposed by A. H. Clark as a genus. Nothing is gained by using a sub- generic group for this very distinct species, and I have therefore called it Validia rotalaria. The 14 specimens at hand were taken as follows : Western Australia : Broome, on Pearl Shoal, 5-7 fms., September 12, 1929. 1 fine adult. Broome, east of Pearl Shoal, 7-8 fms., September 14, 1929. 3 fine adults. Off entrance to Lagrange Bay, 4-5 fms., June 7, 1932. 7 very fine adults. Broome, June, 1932. 1 adult and 2 small specimens. 18 memoir: museum of comparative zoology COMATULA PECTINATA Asterias pedinata Linne, 1758. Sys. Nat., ed. X, p. 663. Comatula pectinaia Dujardin and Hupe, 1862. Nat. Hist. Zooph., p. 203. This is undoubtedly the most abundant crinoid on the coasts of northern AustraUa. Under suitable conditions it swarins over large sponges and rock fragments; on one sponge brought up in our small dredge, we counted over two hundred individuals of all sizes and colors. The largest specimens have the arms opposite the mouth over 125 mm. long, with those adjoining the mouth about 75-80 mm. In small individuals this inequality of the arms is not nearly so marked and is often hardly evident. The diversity of color is very great at Broome; the most common color is red or purplish of some shade. My field notes say: "Almost black to almost white; deep purplish common; also common is whitish or pinkish with red hues on calyx and arms, and cirri red, at least at tip. Dry specimens brick red or darker." On another occasion, I wrote: "One bright crimson; one deep purple with dorsal side of arms, bright yellow. A juve- nile specimen was bright scarlet, each pinnule just tipped with white; and it became bright yellow in alcohol. Several were hght grayish with two dark Unes on back of each arm; cirri yellowish; dried, hght red brown. Some specimens dull orange or orange-brown." The individuals of this species which have the cirri confined (usually in pairs but not rarely singly) to the interradial angles of the centrodorsal were called Alecto purpurea by J. Miiller nearly one hundred years ago (1843) and the name has persisted ever since. Even as late as 1931, the species Comatula purpurea (J. Miiller) is recognized by the very best authority on crinoids, Mr. A. H. Clark. In 1919, Gislen, studying material from northwestern Australia, decided that purpurea was no more than a variety of pectinaia and a few years later (1923) I expressed my sympathy with this view but considered more data necessary before we could be sure what the relation between the two forms really is. The abundance of these comatulids at Broome has enabled me to secure such data, and I am satisfied that purpurea is not properly even a variety. That is, I do not find any reason to beheve it breeds true or is confined to any locahty or habitat where pectinata does not occur. Where Uving conditions are "hard" (judged, of course, by human standards) all of the peclinatas may be small and have the cirri characteristic of purpurea but where conditions are favorable, as at Broome, the two forms occur together and in about equal num- CLARK: AUSTRALIAN ECHINODERMS 19 bers and connecting links are not rare; these usually take one of two forms — they may be typical purpurea except for having three cirri at one or more inter- radial angles of the centrodorsal or they may be peciinata with the cirri lacking from 3 or 4 of the radial sides of the centrodorsal. Of course, such specimens may be arbitrarily assigned to one form or the other, but it seems to me clear that the difference is not morphological. For convenience in referring to the form with only 10 (or fewer) interradially placed cirri, we may use the phrase "forma purpurea." It should be noted, however, that this is not a depauperate form or pecuhar in any other particular. It occurs commonly with typical pectinata and shows the same diversity in size, form, and color. The failure to find pectinata in any form at our collecting grounds on the Coburg Peninsula or at Darwin is puzzling though I have no doubt it is due to the large amount of silt in the water. It is not impossible that in September and October, before the rains begin, the water may be clearer and this common comatuUd might be found in certain favored localities. In support of this theory are three very small specimens with arms about 15 irun. long or less, which we dredged in 3-6 fms., on a bottom of sponges, gorgonians, etc., near the Shell Islands and near Channel Island, at Darwin. In hfe they were bright colored (variegated, red, yellow and white) but they are now pale brown or dingy white. One of these has been identified by Mr. A. H. Clark as a young pectinata and I see no reason to hesitate in calhng the others the same. The only possible reason for doubting the identification is that we failed to find any specimens large enough to make it certain, and we did secure the adults of three other comasterids near the Shell Islands, all of which undoubtedly pass through a 10-armed Coma- tula stage. For the present, therefore, I list pectinata from Darwin only with question. The material at hand consists of 78 specimens, of which 47 are referred to pectinata and 31 to "forma purpurea." The locaUties follow: C. pectinata forma typica. ?Northern Territory: Darwin, near Shell and Channel Islands, July, 1929. 3 very young specimens, too immature to permit certain identification. Western AustraUa: Broome, various localities, low tide mark to 8 fms., June, 1932. 36 specimens. Broome, Pearl Shoal, 5-8 fms., August, 1929. 5 speci- mens. Near False Cape Bossut, September, 1929. 2 specimens. 20 memoir: museum of comparative zoology "Northwestern Australia, April, 1915. Capt. Due leg." Gift of E. W. Bennett. 1 specimen. C. pectinata forma purpurea. Queensland: Great Barrier Reef, Eagle Island (north of Cooktown). T. T. Flynn leg. 2 young specimens. Western Austraha: Broome, various locaUties, August and September, 1929. 17 specimens. Broome, Cable Beach, September, 1929. Frances L. S. Clark leg. 3 specimens. Broome, dredged at various locaUties, 4-8 fms. 8 specimens. Near False Cape Bossut, September, 1929. 1 specimen. COMATULA SOLARIS Lamarck, 1816. Anim. s. Vert., 2, p. 533. This large ten-armed comatuhd is one of the most striking of the crinoids found on the coast of Northern Australia. Its great size, robust structure, and diversity of coloring combine to attract unusual attention. The first one seen by us was creeping along the hard, sandy bottom of Roebuck Bay in an effort to overtake the rapidly outgoing spring tide, September 21, 1929. The body, with oral side up, was raised some three inches above the sand, the weight being sustained on the tips of the ten arms. Progress was by no means slow but not rapid enough to keep up with the ebbing .sea. WTien picked up and placed in a bucket, taken to the laboratory and killed in alcohol, this individual showed no fragihty or tendency to break off its arms, which at that time were about 200 mm. long; in its present diy condition the body is 25 mm. acro.ss and the arms only 175 mm. long; there are about 20 cirrus sockets with but 6 cirri, of some 18 segments. The color of this specimen when living was as follows: centrodorsal and cirri, yellow; arms basally (on the dorsal side) of the same color but this becomes increasingly Ughter distally passing through cream-color to almost white at the arm-tips; pinnules yellowish or cream-color with few or many brown bands; oral surface of pinnules and arms, brown; disk and oral pinnules a bright brown in decided contrast. A smaller specimen taken a few days later was very different in color and in its present dry state is nearly uniform hght brown. A very hand- some specimen with arms about 130 mm. long was dredged in 4 fms. off Cape Villaret, southwest of Broome. There are 16 cirri and these, with the centre- CLARK : AUSTRALIAN ECHINODERMS 21 dorsal and adjoining arm bases are dull purple; the rest of the dorsal surface is almost white while the pinnules and the oral surface are deep blackish brown in striking contrast. These three were the only examples of solans met with in 1929 but in June, 1932, we found it not rarely in our dredging, and there are 10 of these specimens at hand. The largest are superb creatures with the arms 210-230 mm. long and 6 mm. or more wide, without the pinnules. They reveal great diversity in color- ing, though now chiefly brown of some shade, or deep purple. My field notes say that one specimen was notable for having "the pinnules on basal part of arm orally Ught blue." In another the distal pinnules terminated in yellow. Such bright colors are not indicated in the preserved specimens. One individual is so much smaller than the others, it deserves a word of comment. The arms are only 50-70 mm. long and the cirri are XXII, 15-17; the color is light brown more or less variegated with a darker shade. COMANTHINA BELLI Adinomefra belli P. H. Carpenter, 1888. "Challenger" Comat., p. 334. Comanthina belli A. H. Clark, 1911. Amer. Jour. Sci. (4) 32, p. 130. This notable comatuhd is characteristic of the northern and northwestern coasts of Australia from the northern end of the Barrier Reef (the Murray Islands) to Shark Bay and the Abrolhos Islands in the west. It is beyond ques- tion the most beautiful marine animal found in those waters, its large size, great number of arms and extraordinary diversity of color making it very con- spicuous. We took but a single specimen at Darwin but it is very common at Broome, where it often occurs near low water mark or even in pools left by the ebbing spring tides. To give any adequate account of its beautiful appearance would involve a description of nearly every specimen, which is, of course, out of the question. Suffice it to say that large individuals have more than 100 arms and up to at least 150; one specimen has apparently more than this, possibly nearly 200 but anything like an accurate count is impossible without wrecking the specimen. In Ufe the big individuals are fully 300 mm. across, with the disk exceeding 40 (the calyx, of course, much less) and the arms 125-140 mm. long. In the largest specimen there is a VII Br 2 series on one ray; if every ray di- vided six times, there would, of course, be 320 arms, but it is a familiar fact that in all multibrachiate crinoids, after the first two or three forkings, the division 22 memoir: museum of comparative zoology series become very irregular. Several young individuals were taken with 30-55 arms which are 50 to 100 mm. long; the most remarkable of these is a specimen with 33 arms, 60-75 mm. long (in life), which when living had the calyx, disk, and base of arms bright lemon yellow but distally the pinnules became, gradually not abruptly, a bright turquoise blue — a most exquisite and, so far as our col- lecting shows, a unique color form. The diversity of coloration shown by belli surpasses that of any other echino- derm with which I am acquainted. Shades of red are the least common and the most fugacious. One notable specimen is almost uniformly black, while in strik- ing contrast is an individual with the arm bases pale olive variegated with cream-white, the centrodorsal, the interradial areas of the calyx and the distal half of the arms cream-white ; many arms have broad but rather indefinite bands of dusky near the tips. Bright lemon yellow, orange-brown, deep brown, and green individuals are frequently seen and one beautiful specimen was (and to some extent still is) a fine light blue; a broad whitish band on the middle part of each arm fornis a wide white circle separating the markedly blue arm bases from the Ughter blue tips. Green indi\'iduals are seldom unicolor but show several shades variously intermingled. One feature of the coloration is remarkably con- stant — the cirri are pure white in life. Even the darkest colored specimens show this character, making it a convenient "recognition mark" — for the collector! Unfortunately in preserved specimens, the cirri generally get more or less stained and are likely to take on the color of the adjoining arm-bases. But in most of the dry specimens, they show a very considerable contrast to the rest of the animal. A very small dry crinoid from Cable Beach, Broome, with 15 arms, 7 of which are about 45 mm. long, is referred with some doubt to this species. The cirri are XIV, 10-12 and are pure white, in contrast to the very pale brown of the arms. There are three II Br 4 and one III Br 2 and the basal segments of the lower pinnules show indications of the projecting ridges which are such a striking feature of adult belli. But similar projections are to be found in young Coman- theria briaretis and it is by no means easy to see any marked or constant difference between young individuals of the two species. Besides this 15-armed specimen from Cable Beach, there are three young comatulids which have given me a great deal of trouble. Mr. A. H. Clark has been so good as to critically examine the largest of these (100-125 mm. across) and thinks it is undoubtedly a "young specimen of Comanthina belli undergoing adolescent autotomy." The other two are smaller and differ to some extent in general appearance, but after prolonged CLARK: AUSTRALIAN ECHINODERMS 23 comparison with young belli and young briareus, it seems best to consider the three identical and include them all under the present species. The 29 specimens at hand were taken as follows : Northern Territory: Darwin, near Shell Islands, 5-8 fms., July 25, 1929. One adult specimen. Western Austraha: Broome, August and September, 1929. 15 specimens; 1 young one of doubtful identity. Broome, Cable Beach, September 2, 1929. Frances L. S. Clark leg. 2 specimens. Broome, Cable Beach, September 2, 1929. Frances L. S. Clark leg. 1 very young specimen ; identity not certain. Broome. From Perth Museum. 1 specimen. Broome, June, 1932. 7 specimens; 1 young one of doubtful identity. Broome, Entrance Point, July 4, 1932. 1 young specimen. False Cape Bossut, September, 1929. 1 young specimen; identity dubious. COMANTHERIA BRIAREUS Antedon briareus Bell, 1882. Proc. Zool. Soc. London, p. 534. Comantheria briareus A. H. Clark, 1911. Fauna Slid west- Austral., 3, p. 437. This crinoid is very common in the Broome region; we found it at Cape Leveque, and along the coast from there to Cape Bossut at least, it occurs wher- ever conditions are suitable. How far west and south it ranges is still open to question. It is recorded from "?Vicimty of Perth (2 specimens)" (A. H. Clark, 1911, p. 454) and a specimen in the Perth Museum (No. 9681) is said to be from "Bald Island, 28-32 fms." I am more than skeptical as to these labels. Bald Island is east of Albany on the southern coast of Austraha and I do not beheve that any Comantherias occur there. Moreover, the specimen is an old one, dry and in poor condition, and no number or label is attached to it. The two speci- mens Usted by Mr. Clark from the "vicinity of Perth" were in the Michaelsen and Hartmeyer collection and it is very probable that they too suffer from a misplaced label. If actually collected by the German zoologists, I think they were taken in Shark Bay, wliich would then be the southern hmit for this species on the West- ern AustraUan coast. That briareus occurs in Shark Bay is not wholly improbable 24 memoir: museum of comparative zoology but until there is further evidence than is as yet available, I think we may con- sider the vicinity of Cossack (Port Walcott) as about the southwestern Umit of its range. East of Cape Leveque there are records from the region of the Holo- thuria Banks, and there is one small specimen in the present collection from Darwin. As found at Broome, briareus is a very perplexing comatuhd. The diversity in size, number of arms and color is very great, and there is no apparent correla- tion with differences in habitat. Most of the specimens were dredged and not rarely a number of specimens of diverse size and appearance came up in a single haul. But both large and small specimens of dull and of brightly colored forms were found along shore just below low water mark. The smallest specimens at hand are 50-60 mm. across while the largest has an expanse of 225; in this big indixadual the anterior arms measure about 130 mm. but the posterior are scarcely 80. The number of arms ranges from 21 to 65; there are a number of specimens with 40-57 arms but fully three-fourths have fewer than 40 (21-38). At first, it seemed to me probable that these indi\'iduals with fewer than forty arms repre- sented a second species, rotula A. H. Clark, but the more I have studied my ma- terial the more the conviction has grown that it is impossible to recognize a second species on that basis. The diversity in color suggested that there might be a useful distinction possible along that line but this proves out of the question. The largest and finest specimens from Broome are dry and are a more or less uni- form greenish-yellow dorsally; in life, these were yellow or green of some shade dorsally, with black and white pinnules in handsome contrast ; in some specimens (notably the largest), the yellow does not cover the whole dorsal surface but oc- cupies a very broad longitudinal band on the radials and all the brachials. Many of the small specimens are essentially like this and there is no question in my mind that these greenish and yellowish individuals must all be referred to a single species regardless of size and number of arms. On the other hand, many small specimens are now (and were in hfe) uniformly brown of some shade; the brown maj' be reddish, purplish, or grayish, and hght or very dark; one small specimen is very hght almost fawn-color but m life was "clear yellow brown" — not however with any trace of green. A few specimens are clearly intermediate between the brown and the greenish groups having the dorsal side brown, more or less spotted or marked with yellowish-green. I am, therefore, convinced that all my material must be referred to briareus with the comment that the tendency at Broome is toward the handsome yellow and green shades. If rotula is a valid species, it is not represented in my material. CLARK: AUSTRALIAN ECHINODERMS 25 Two specimens taken at Broome in 1929 are of a uniformly light gray color. One has 28 arms, 50-80 mm. long, with only 3 III Br 2 and 5 III Br 4 (no further divisions) ; the other has 33 arms, 60-90 mm. long, with only 5 III Br 2 and 8 III Br 4. These were at first referred to Comanthus timorensis, but the character of the arms and cirri has led me, after discussing them with Mr. A. H. Clark, to refer them to hriareus. They are, of course, young and the species characters (one might even say the generic characters) are not yet fixed. The 47 specimens which I am referring to hriareus were collected at the following points: Northern Territory: Darwin, near Shell Islands, about 5 fms., July 24, 1929. 1 young specimen with only 18 arms; having 4 II Br 4 and 4 III Br 2. Western Australia: Cape Leveque, August, 1929. 1 small brown specimen with 43 arms. Broome, Cable Beach, September 2, 1929. Frances L. S. Clark leg. 3 small specimens. Broome, August and September, 1929. 10 adults and young. Broome, June, 1932. 31 adults and young. Loc. ? Labelled "Bald Island. 28-32 fms." 1 small adult with 35 arms. Loaned by the Perth Museum. COMANTHERIA PERPLEXUM Comanthiis perplexum H. L. Clark, 1916. "Endeavour" Ech., p. 14. Coinantheria perplexum A. H. Clark, 1931. Crinoid Monog., 1, pt. 3, p. 506. There is a very fine specimen, from the AustraUan Museum, of this species, hitherto known only from the holotype, taken in 15 fms. off Norwest Islet, Capri- corn group, Queensland, in December 1930, by Livingstone and Boardman. The holotype was from off Ballina, N. S. W., 49-51 fms. The present specimen is a very fine one with 38 arms, 100 nmi. long; there are only a few cirri, VI, IS. The disk was lacking in the holotype but is present in this individual. It is 17 mm. across, yellowish-gray in color, plump, with a large anal tube and covered irregu- larly and not closely with whitish papillae about .30 mm. in diameter. As in the holotype, the dorsal surface of this specimen is light fawn color, the oral surface somewhat darker. 26 memoir: museum of comparative zoology COMANTHUS PARVICIRRA Aledo parvicirra J. Muller, 1841. Monatsb. Preuss. Akad. Wiss., p. 185. Comanthis parvicirra A. H. Clark, 1908. Smithson. Misc. Coll., 52, p. 203. It is a puzzling fact that no comatulids which may be referred to this species were taken by us on the northern or western coasts of Australia. It is a wide- spread species in the Indo-Pacific region and has been recorded several times from Western Australia. There are several specimens in Mjoberg's collection from Broome and southwestward which Mr. A. H. Clark assigns to parvicirra (see Clark, 1931, pp. 636 and 664-666) although Gislen identified them as Comasters. But all the Comanthus which we found at Broome seem best referred to samoana and timorensis. The only comatuHds in the collection at hand which may well be assigned to parvicirra are 7 very small individuals from Port Curtis, Queensland, and these are much too young for certain identification. The largest is about 40 mm. across and has 17 arms; the cirri, X, 12-13, are as distinctive as any feature, and I have little doubt that this is a young parvicirra. The other 6 specimens are very small, 15-20 mm. across and have but 10 arms each. The only reason for calUng them parvicirra is their association with the larger individual. Comanthus samoana A. H. Clark, 1909. Proc. U. S. Nat. Mus., 37, p. 30. This comatulid is now known from a considerable number of locahties on the tropical coast of Australia and among the western Pacific islands. Its occur- rence, therefore, at Broome is not surprising but is none the less interesting. It is not, however, common there for in 1929 we took it only at or near False Cape Bossut and in 1932 only one adult specimen was taken. The specimens at hand are typical and call for Uttle comment. The smallest is about 40 mm. across, has 18 arms and the cirri are XVII, 11. There are 3 specimens with 20 arms each, the smallest about 50 mm. across. The larger specimens have 22, 23, 27, and 32 arms respectively; the largest is about 150 mm. across. The cirri range from XIX, 11-12 to XXVI, 15-16. The color is pale brown in the smallest specimen and purpUsh-brown or yellow brown in the other young ones. The larger specimens are more or less uniformly dull, dark brown, with a purphsh or grayish cast. CLARK: AUSTRALIAN ECHINODERMS 27 The 8 specimens at hand were taken as follows : Western AustraUa: False Cape Bossut, September 8, 1929. 3 adults and 2 young. Broome, June, 1932. 1 adult and 2 young. COMANTHUS TIMORENSIS Aledo timorcnsis J. Muller, 1841. Monatsb. Preuss. Akad. Wiss., p. 186. Comanthus iimorensis A. H. Clark, 1931. Crinoid Monog., 1, pt. 3, pp. 159 and 603. Probably the commonest multibrachiate crinoid on the tropical coasts of Australia, this fine species shows almost as great diversity in color and appear- ance as does Comanthina belli. Adult specimens are not hard to recognize but young and half-grown specimens are easily confused with parvicirra. The very young specimens listed below are included under timorensis because they appear to be young Comanthus of some species and the present species is most likely to be the one to which they belong. No specimen of Comanthus under 80 mm. across can be identified specifically with certainty unless it is one of a series which connects up with unmistakable adults. The smallest specimen at hand is from Darwin and is only about 10 mm. across, probably 15 mm. in life; there are but 10 arms and there are no division series. The cirri, IX, 9, are short and weak and quite hke those of young timorensis. Small specimens from Broome, 30-40 m m . across have 11-13 arms and a very good specimen from Darwin about 70 mm. across has 16 arms with 6 II Br 4 series; the cirri are typical, XI, 11. My field notes say of this indi\adual: "Very active, both creeping and swinmaing. A hand- some Ught brown and white." The large specimens offer no features of special interest, but it may be men- tioned that a particularly good one from Darwin with 32 arms has no fewer than 9 III Br 2 series. From Cape Leveque are two large specimens with 48 and 52 arms and no cirri; in the larger one every division series is 4 (3 +4). A very large specimen from Broome has only 43 arms but they are about 140 mm. long, and it has cirri XV, 10-13. One specimen from Broome, about 150 mm. across in life, has but four rays; there are 33 arms and every division .series is 4 (3 +4). As for color, little need be said; preserved material is all brown of some shade, ranging from very light in the little individuals to very dark, almost black in some of the larger specimens. But in hfe the colors were diversified and often very handsome. One taken at Darwin was "black with each pinnule tipped with 28 memoir: museum of comparative zoology yellow-green." At Broome my notes say "usually dark but range from green at one extreme to yellow-brown at the other." There are 28 specimens in hand from the following localities: Queensland : Capricorn group, Norwest Islet. 1 poor adult. Loaned by Australian Museum. Northern Territory: Darwin, near jetty, 7 fms., mud, July 1, 1929. 1 very young specimen. Identification ? Darwin, near Shell Islands, 5 fms., July 25, 1929. 1 very fine adult. Darwin, near Leper Station, 3-5 fms., May 25, 1932. I young specimen. Western Australia: Cape Leveque, August, 1929. 3 adults. Broome, August and September, 1929. 9 adults and 4 very young. Broome, June, 1932. 7 adults and 1 very young. COMANTHUS TRICHOPTERA Comahila trichoptera J. Muller, 1846. Monatsb. Preuss. Akad. Wiss., p. 178. Comanthus trichoptera A. H. Clark, 1909. Zool. Anz., 34, p. 363. It was an interesting experience to find this little comatulid, typical of the temperate waters of southern Australia, living among the corals at Neds Beach, Lord Howe Island, with many echinoderms characteristic of the Great Barrier Reef. It seems to have become well established there but does not grow to a large size. The largest taken were about 125 mm. across but most were under 100 mm. The arms are noticeably more slender than in large specimens from Port Jackson. In life the Lord Howe specimens are "variegated white and brown of various shades, mostly yellowish." They resemble very closely specimens taken in November, 1929, at Port Willunga, South Austraha. The cirri are small with usually 13 or 14 segments, seldom 15-17, but the specimens from Port Willunga are not essentially different in this particular, though the number of segments will probably average one or two more in the South AustraUan material. In his key to the species of Comanthus, subgenus CenoHa, (1931, p. 530) Mr. Clark says of trichoptera: "cirri with about 20 segments," in contrast with "cirri with 14-17 segments" in allied species. My material shows 13-17 segments as the rule, only in the largest individuals of trichoptera do I find 20 and occasionally 21. CLARK : AUSTRALIAN ECHIN0DERM9 29 The number of arms in the Lord Howe specimens is usually under 30, the largest having 27 or 28, but there is one with 30 and one with 33. The largest specimen in the present series is from Port Jackson and has 31 rather stout arms, about 75-80 mm. long. A much smaller specimen from Port Willunga has 37 arms. From Bunker's Bay, W. A., there is a specimen with rather stout arms nearly 90 mm. long but there are only 21, while another specimen from the same lot has 28 arms though they are Uttle over 50 mm. long. The occurrence of trichoptera at Lord Howe Island extends the range of the species several hundred miles to the east. It was, therefore, interesting to find a specimen among the echinoderms from Dongarra, W. A. given to me by Professor E. W. Bennett, extending the known range several hundred miles to the north- west. The 71 specimens at hand were collected at the following places: Lord Howe Island, Neds Beach, among corals, 18 specimens. New South Wales : Port Jackson, Bottle and Glass Rocks, November 27, 1929. 2 large adults. : No definite locality. 1 adult. Loaned by Australian Museum. South Australia: Port Willunga, November 2, 1929. 10 specimens. Western Australia: Albany, under stones below low tide mark, January 10, 1929. E. W. Bennett leg. 8 specimens young. Cape NaturaUste, east side. Bunker's Bay, under stones, January 15-17, 1930. E. W. Bennett leg. 26 specimens, adult and j'oung. Bunbury, Koombana Bay, 5-8 fms., October 26, 1929. 2 specimens, young. Cottesloe Beach. 3 specimens, adult and young. Dongarra. E. W. Bennett leg. One specimen, young. ZYGOMETRIDAE Zygometra comata A. H. Clark, 1911b. Mem. Austral. Mus., 4, p. 762, footnote. This is a common species at Broome and a considerable series was brought home. Mr. A. H. Clark gave me great assistance in determining them, certain individuals being very puzzUng. The number of arms ranges from 11 in a speci- men 70 mm . across, to 25 in a specimen 115 mm. from tip to tip; the largest 30 memoir: museum of comparative zoology specimens are 175-200 mm. across and have from 16 to 24 arms. The colors are yellow and purple but the relative amounts of the two, their distribution and the shades show infinite variety. The shades range for the yellow from almost white to deep buff and for the darker color from Ught grayish-purple to a very deep, rich, reddish purple. The color arrangements are too diversified to warrant description but it is common to have the arms banded either broadly or nar- rowly; often the pinnules only are banded and the dorsal side of the arms is prettily mottled or simply unicolor. It is hard to see how and where the line is to be drawn between this .species and punctata which was described the following year. So far as the number of segments in the cirri is concerned, the series of specimens from Broome shows that there is a considerable range but only a few large specimens show as many as 25-27, while Mr. Clark says (1918, p. 59) "more than 25 (usually about 30) cirrus segments" in comata. Most of the present specimens have 20-23 which accords with the "less than 25" given for punctata (1918, p. 60) but not with the "(18-21)" which is added. None of the Broome specimens have fewer than 20 segments. Although not previously recorded from northern Australia, this comatulid is very common at Broome. Like all the Zygometras of that region, it delights in the clean shallow water and hard sandy bottom of Roebuck Bay. At extreme low water in the greatest spring tides (as in September, 1929) a large part of the water ebbs out of Roebuck Bay and vast areas of a clean, firm gray sand are ex- posed. On this bottom are countless patches or isolated plants of coralline and other algae, besides partly buried rock fragments or shells scattered abundantly about, well separated from each other, and each forming the nucleus of a little animal community seeking shelter from the pitiless exposure to which the unusual ebb of the tide subjects them. On almost every alga or other projection, a coma- tuUd will be found (usually but one) and the vast majority of these are Zygome- tras and chiefly comata. It is a hardy species, enduring this exposure to the sun, transferring to a bucket and transportation to the laboratory without damaging itself in any way. Left by the retreating tide, the coraatulids fall relaxed on their sides, the arms closed on the disk and pointing to the water that has abandoned them, but when the tide returns, enduring for a time the washing back and forth of the now coming, now going wavelets, they soon find the water deep enough to enable them to sit erect, expand their arms and renew their normal life. Seeing this flower-Uke expanding of the wilted comatulids is one of the most interesting pleasures of watching the incoming tide. CLARK: AUSTRALIAN ECHINODERMS 31 The 39 specimens of comata at hand were all taken in or near Roebuck Bay- in 1929 and 1932, excepting two which were secured in the vicinity of Lagrange Bay in 1929. Zygometra elegans Antedon elegant Bell, 1884. "Alert" Rep., p. 162. Zygometra elegans A. H. Clahk, 1907. Smithson. Misc. Coll., 50, p. 348. This is another very common comatuUd at Broome, occurring under the same circumstances as comata, with large specimens of which small specimens of elegans might be confused. It is, however, a larger but more delicate species, with more arms and many more cirrus segments. It is hard to see why Gislen (1919, p. 19) treated elegans as merely a variety of microdiscus but I suspect (from his descrip- tion) that the individual which he describes as microdiscus was really a large elegans with the maximum number of arms and of cirrus segments. The large series of specimens at hand from Broome show well the following differences between the two species : Division series: in elegans, III Br 2, with few series beyond, and those often 2. in microdiscus, III Br 4, and the many subsequent series almost always 4. Cirri: in elegans XX-XXX, 34-56, usually under 45. in microdiscus XXVII-XLIII, 35-70, usually over 45. Lowest pinnule in elegans not nearly so flagellate as in microdiscus and smaller in every way. Color in elegans is more varied on the whole, combinations of hght grays, buffs, fawn-color and purple of various shades being usual while unicolor specimens are rare; in microdiscus, deep purples are prevalent and unicolor specimens are common, but handsomely striped and variegated specimens are by no means rare, and are very striking. The largest specimen has 51 arnis about 130 mm. long and the XX cirri have 45-56 segments. A very fine specimen from Darwin has only 36 arms but they are nearly 150 mm. long; the cirri have about 40 segments. In hfe, according to my field notes on this specimen, "the color was distinctly dull purple and wliite; disk whitish or pale cream color; pinnules tipped with white and also white on dorsal side." The smallest specimen at hand has 31 arms about 45 mm. long; all of the III Br series are 2. The cirri have 36 segments. 32 memoir: museum of comparative zoology While occurring with comata on open sandy bottom, this species also occurs among rocks, and is frequently dredged on bottoms covered with algae, sponges and other marine growths. The 48 specimens before me are from the following places : Queensland: Capricorn group, off Norwest Islet, 6 fms., December, 1930. Livingstone and Boardman leg. 3 specimens. Loaned by Australian Museum. Northern Territory: Darwin, near Shell Islands, 3-6 fms. on sponge and alcyona- rian bottom, July 2, 1929. 1 very fine adult specimen. Western Australia: Broome, Cable Beach. September 2, 1929. Frances L. S. Clark leg. 2 fine adults. Broome, on hard sandy bottom, August and September, 1929. 19 specimens. Broome, dredged at various points, 3-8 fms., June, 1932. 23 specimens, adult and young. Zygometra microdiscus Antedon microdiscus Bell, 1884. "Alert" Rep., p. 16.3. Zygometra microdiscus A. H. Clark, 1907. Smithson. Misc. Coll., 60, p. 348. This truly magnificent comatulid is by no means rare in the Broome region but it is not so common as either of the preceding species and prefers deeper water. Most of the specimens taken were dredged in 5-8 fms. but a very few were found on the hard sandy bottom of Roebuck Bay during the extreme low tides of September, 1929. The color is primarily a reddish-purple or even deep claret with the dorsal surface of the arms more or less yellow, but the amount of yellow varies enormously; at one extreme are individuals which are practically all purple, really unicolor, (and they are not rare) while at the other are those in which the yellow predominates to such an extent that only the tips of the cirri and more or less of each pinnule distally are purple. The most beautiful speci- mens are yellow or nearly white, finely speckled more or less profusely with purple. In preserved material, even that which is very carefully prepared, the yellow shades become buff or light brown and the purple becomes dingy and often distinctly brown. Naturally such museum material fails to give any adequate idea of the beauty of the living animal. In Hfe, the finest specimens are over 300 mm. across. CLARK : AUSTRALIAN ECHINODERMS 33 In the present series the most interesting are the largest and smallest. The latter taken in Lagrange Bay, has but 19 arms 20-25 mm. long; there are 12 division series, all 4 (3 + 4) ; the cirri are XIII, 25-29 and relatively very long, more than half as long as the arms; the color is uniformly deep purple. In spite of its small size, this is quite a typical microdiscus. The largest specimen, which in its present dry condition is very nearly 300 mm. across was taken in Roebuck Bay. It has about 110 arms but it is impossible to count the exact number with- out serious damage to the specimen. My field notes say that in life it was "deep claret with dorsal side of arms yellow; some young arms are tipped with white;" at present it is a deep purple, and only distally do the arms show that they were yellow dorsally in life; they are now a light brown dorsally near the tips. The dark purple cirri are XLIII, 52-70, and are 50-55 mm. long; a few of them are cream-color near the base, and one of them is forked at the tip. The 25 specimens of microdiscus at hand were taken as follows: Western AustraHa : Broome, August and September, 1929. 6 specimens, 1 very large. Lagrange Bay, 5-8 fms., September, 1929. 5 specimens, 1 very young. Broome, June, 1932. 14 specimens, showing great variety in size and color. HIMEROMETRIDAE Heterometra crenulata Antedon crenulata P. H. Carpenter, 1882. Jour. Linn. Soc. London, 16, p. 507. Heterometra crenulata A. H. Clark, 1918. "Siboga" L'nst. Crin., p. 77. Our first capture of this strikingly handsome comatulid came at the close of a long day's dredging outside Roebuck Bay. Our last haul was made in 5-8 fms. on Pearl Shoal and contained half a dozen of what seemed at the moment the most lovely comatulid I had ever seen. The arms were a pure milk white, the cirri bright rose color in sharp contrast. The somewhat rigid arms remained curved outward Uke the petals of a flower and the old book name "sea-lily" was most appropriate. We subsequently dredged other specimens, whose coloration was so different they were not even suspected of being the same species. In 1932, we met with Heterometra often and the diversity of color led to the supposition 34 memoir: museum of comparative zoology that we were collecting three different comatuUds. It was a great surprise to discover on critical study here in the Museum that they all represent a single and long known species. According to coloration, three forms may be recognized but with sufficient material, it is obvious that they intergrade completely. At one extreme are the milk white forms with rose colored cirri already mentioned. When preserved in alcohol or dried, these lose much of their beauty, the white becoming dingy or tinged with purple and oftentimes small spots or blotches of a purplish shade appear. At the other extreme are the individuals, usually large ones which are deep red purple or dark crimson with few or no lighter markings ; these commonly have a coarser and more rugged structure than the white ones. The third group comprises the more or less handsome individuals with banded arms ; these may be white, or at least very light, as to ground color with the arms banded with red or purple of some shade, or they may be red or purple with whitish bands and markings. The number, width and shade of armbands show endless diversity. The cirri are always red or purple of some shade, ranging from Hght rose to a deep red violet. One very small comatuhd, apparently a Heterometra, and hence almost surely this species, was taken in 1932. There are but 13 arms, about 15 mm. long; the division series are 4 (3 + 4) but the cirri are only X, 13; the color is very pale brown, the cirri with a very slight tinge of violet. The number of arms in the present series of crenulata ranges from 16 to 30 but the great majority of individuals with arms over 60 mm. long have 20-25. The cirri are XX-XXXII, 34-46, but small individuals often have fewer segments. The largest specimens have the arms 75-90 mm. long. There is great diversity in the development of the projections on the segments of the basal pinnules; often they may best be designated as spines but in other cases they must be characterized as "wings;" light colored individuals seem on the whole to tend towards wings, while the dark colored ones are more spinous; the correlation is, however, very imperfect. The 50 specimens at hand were taken as follows : Queensland: Capricorn group, Norwest Islet, 6 fms., December, 1930. Boardman and Livingstone leg. 2 specimens. Loaned by Australian Museum. Western Australia: Broome, on or near Pearl Shoal, 5-8 fms., September, 1929. 18 specimens, adult and young. Broome, dredged at various points in 5-8 fms., June, 1932. 30 specimens, adult and young. clark : australian echinoderms 35 Amphimetra jacquinoti Comatula jacquinoti J . Muller, 1846. Monatsb. Preuss. Akad. Wiss.. p. 178. Amphimetra jacquinofi A. H. Clark, 1918. "Siboga" Unst. Crin. p. 85. This stiff and rather ungainly comatuhd is by no means common at Broome but we met with it several times while dredging, and one specimen was found at extreme low water in September, 1929. These individuals are all adult with the ten arms 110-130 mm. long. The cirri are XV-XXX, 28-36. The color in hfe was deep reddish purple, becoming very dull on drying; the arms are more or less marked or banded with white or yellow, but the location and amount of the Ught color show great diversity; sometimes it is confined to the pinnules, or to the dorsal side of the tips of the arms, or to regenerating portions of injured arms. The 10 specimens in the present collection were taken as follows: Western AustraUa: Broome, about one inile off jetty, August 27 and September 5, 1929. 2 specimens. Broome, off Gantheaume Point, 2-4 fms., August 30, 1929. 3 specimens. Broome, 5-8 fms., June, 1932. 5 specimens. MARIAMETRIDAE Lamprometra gyges Antedon gyges Bell, 1884. "Alert" Rep., p. 160. Lamprometra gyges A. H. Clark, 1913. Proc. Biol. Soc. Washington, 26, p. 144. This many-armed comatulid is conmion on the tropical coasts of Australia and its diversity of color and of arm length make it a very perplexing species, and I am not at all convinced that several different species are not included under Bell's specific name. On the other hand the supposed distinction between gyges and protectus (Lutken) is based on a very variable character and the Une is hard to draw. For the present, however, regardless of their differences, I am including all of my Mariametridae under the one name, hoping that at some future time critical study of more material may satisfy my present doubts. The material before me consists of 24 specimens ranging from a young one with only 19 arms, 20-25 mm. long and with cirri XV, 14-15, up to fine adults 36 memoir: museum of comparative zoology with 45-51 arms, 70-100 mm. long, and with the cirri XXXVIII-XLVIII, 26-30. Most of the specimens are gray or brown of some shade and apparently unicolor, but a few of the lighter colored ones seem to have had a broad whitish band on the arms in life. Unfortunately these Lamprometras did not attract my interest at the time of their capture sufficiently to receive the attention they deserved and my field notes are confused and inadequate. It is greatly to be regretted that the very rich collecting at Broome and the enthusiasm and industry of Mr. Bourne and his crew often swamped me with specimens to be cared for, and left time for only very meagre notes. The 24 Lamprometras at hand were taken as follows : Northern Territory: Darwin, near Shell Islands, 3-6 fms., sponge and alcyonarian bottom, July, 1929. 5 specimens, young. Western Australia: Broome, August and September, 1929. 6 specimens, adult. False Cape Bossut, September, 1929. 1 specimen, adult. Broome, June, 1932. 12 specimens, adult, large and small. COLOBOMETRIDAE Petasometra brevicirrai sp. nov. Centrodorsal thin, discoidal with a broad flat dorsal pole, about 3 mm. across; cirrus sockets in a single closely crowded marginal row. Cirri XXI, 18-22, about 9 mm. long, somewhat recurved (fig. 1); cirrus segments sub-equal in length but, when seen from below, 4 and 5 are much the widest, 3 and 6 some- what narrower, 2 and 1 are about equal or 1 is narrower; beyond 6 the segments decrease slightly but steadily in width until the terminal segments are no wider than long; the dorsal transverse ridge, characteristic of the genus, is present near the middle of each segment but is not well marked until after the fifth and never becomes spiniform until the penultimate segment on which the opposing claw is fairly well developed. Radials concealed; I Br well developed, the width about two and a half times the length; axillaries large, broadly pentagonal, the width not twice the height; there are four II Br 4 (3 +4) series and one II Br 2; the division series are all well rounded and except for the proximal half of the first elements, distinctly separated from each other. There are 15 arms, 45-60 mm. long; the first brachials are ' brevis = short + cirrus, in reference to the characteristic cirri. CLARK: AUSTRALIAN ECHINODERMS 37 noticeably rounded, somewhat higher externally than internally where they are in contact for one-half their length or more ; second brachial very similar and the two together are somewhat higher than 3 and 4 which are united by syzygy; brachials 3-7 are as high on one side as the other but from the eighth on they are more wedge-shaped though they never become nearly triangular; the second syzygy occurs after an interval of about 13 brachials and the following after about 8. Figs. 1 and 2. Petasometra brevicirra: 1, a cirrus; 2, pinnules pi — ps. x 8. Pa absent ; Pd about 8 mm. long, rather stout, with about 20 segments, the distal portion is terete and slender but by no means flagellate or even delicate; the basal segments are about as wide as long and relatively large but diminish rapidly in size. Pi (fig. 2) is very similar, a trifle longer perhaps, with about 23 segments. P2 is similar but a little smaller. P3 is distinctly smaller with about 15 segments but the transition from P2 to P3 is not abrupt or striking; subsequent pinnules not essentially different but ultimately become longer and more slender; as they increase in length the number of segments rises again to exceed 20. The margins of the pinnule segments are uniformly smooth. Color (of dry specimen) Ught purpUsh-brown, the joints between the bra- chials conspicuously darker; distally the dorsal surface of some arms is evidently purple; centrodorsal dark purphsh brown; cirri at base dorsally, deep buff but 38 memoir: museum of comparative zoology passing into purplish-brown on ventral side and distally; a broad band of dull dark purple on each side of each cirrus, near the dorsal side, becomes narrow dis- tally and fades away at the penultimate segment; disk light colored, except the anal cone which is very dark. Holotype, M. C. Z. No. 916, from False Cape Bossut, Western Australia, September 9, 1929. The peculiarities of this unique specimen were not noted at the time of cap- ture, and no special field notes are connected with it. It is apparently a Petaso- metra but the short cirri with few segments and the short, rather rugged pinnules set it apart from the previously known species, while its whole appearance is strikingly different from the following species. It seems strange that in all the collecting of June, 1932, no further examples of Petasometra were taken. Petasometra variegatai sp. nov. Centrodorsal thin, discoidal with a broad slightly concave dorsal pole, over 3 mm. across; cirrus sockets in a single closely crowded marginal row. Cirri about XXIII, 22-25, 11 mm. long, markedly recurved (fig. 3); cirrus segments very short and wide proximally but becoming narrower and even compressed distally, with little change in length; the basal segments are twice as high as long but the distal ones are not much higher than long; the dorsal transverse ridge is present from the second segment on but it is never very conspicuous and even the opposing claw may be rather small, though it is usually well-marked. Radials concealed; I Br well developed but low, the width about four times the length; axillaries also low, the greatest height hardly one-third of width, the anterior angle rounded and the adjoining sides slightly concave; there are seven II Br 4 (3-1-4) series and three II Br 2; there are nine III Br 4 (3-1-4), one III Br 2 and one III Br 5 (4-1-5) series. There are thus 31 arms, which are relatively short and stout ; all are more or less damaged but they were apparently not over 80 mm. long (perhaps 100 mm. in life) ; the division series are all rather rugged and stout, more or less in contact and the II Br series is a Uttle inclined to be swollen; the brachials are wide and low about twice as wide as long; beyond the seventh or eighth they become wedge-shaped; the first syzygy is between 3 and 4, the second between 11 and 12, and subsequently at intervals of 6-8 brachials. Pa absent; Pj not more than 10 mm. long, rather stout with 20-23 segments; distal portion tapers rapidly to a terete but by no means slender tip; basal seg- ' variegatus = variegated, in reference to the handsome black and white coloration. CLARK : AUSTRALIAN ECHINODERMS 39 ment large, oblong, wider than high; second more nearly square and considerably smaller; third and following segments successively smaller and becoming cylin- drical; margins all smooth. Pj (fig. 4) is very similar and scarcely longer. P2 is similar but a trifle smaller and with fewer segments. P3 is distinctly smaller and P4 much smaller and with only about a dozen segments; succeeding pinnules small Figs. 3 and 4. Petasometra variegata: 3, a cirrus; 4, pinnules, pi — ps. x 8. but soon increasing in length and number of segments until distally there are again more than 20 in each pinnule but they are all small. Color (of alcohoUc specimens) very dark brown, almost black, variegated with cream-color; many arms are almost white dorsally, at least on distal half, but not the pinnules; many cirri are white dorsally and ventraUy but blackish on sides. Holotype, M. C. Z. No. 917, from near Shell Islands, Darwin, N. T., 3-6 fms., sponge and alcyonarian bottom, July 15, 1929. A second specimen was taken at the same time and place but is somewhat smaller, has only 19 arms, 70-80 mm. long, with eight II Br 4 series and one II Br 2. The first syzygy is between brachials 3 and 4, but the next is far out on the arm beyond the thirtieth brachial, and there are very few all together. The 40 memoir: museum of comparative zoology cirri are XX, 22-24, short, thick and strongly recurved. This paratype is dry and the color is dull gray brown, variegated and spotted with a light yellow-brown ; there is much less of the light color than in the holotype. My field notes say of these specimens that "they seemed almost black with dorsal side of arms cream-white. Very handsome. Died on way home and never relaxed, hence are poor specimens." In several points this handsome species suggests Antedon darae Hartlaub and I have tried to convince myself that it should be called by that name, but the type of that East Indian form was much larger than the Darwin specimens, yet had only 12 arms. It also had much longer cirri and the colors and color pattern were totally different. It seems best, therefore, to keep the Australian species separate. Austrometra thetidis Oligometra thetidis H. L. Clark, 1909. Mem. Austral. Mus., 4, p. 522. Austrometra thetidis A. H. Clark, 1918. "Siboga" Unst. Crin., p. 111. This little comatulid is represented in the present collection only by three dry specimens belonging to the Australian Museum which were taken in De- cember, 1929, west southwest of Gabo Island, N. S. W. ; they were trawled in 70 fms. by Captain K. Moller, who presented them to the Museum. They are somewhat smaller than the types but are otherwise similar; one is cream color, one is suffused with a reddish-violet tint and the third is distinctly light red- dish-violet with indistinct small areas of cream color. Oligometrides adeonae Comatula adeonae Lamarck, 1816. Anim. s. Vert., 2, p. 535. Oligometrides adeonae A. H. Clark, 191.3. Smithson. Misc. Coll., 61, no. 15, p. 37. This species has already been recorded from Broome and vicinity by Gislen (1919) who has discussed his specimens with great care, so it is not necessary for me to go into details. We did not find adeonae at Darwin or at Cape Leveque, perhaps because local conditions are not favorable. It is one of the commonest of the small comatulids in the Broome region and we met with it constantly. It was particularly common and easy to collect (along with Zygometra comata), at the extremely low September tides in 1929, on the hard-sand bottom of Roe- buck Bay. My field notes say: "Very lovely: deep red-purple, with or without white." "Deep red- violet and white." "Sometimes apparently deep yellow, with or without white on basal pinnules." Alcoholic specimens keep their colors CLARK: AUSTRALIAN ECHINODERMS 41 very well. Dry specimens are duller and show no white but instead are various shades of yellow and pale brown. Practically all gradations may be found between the purple and the yellow individuals. The most interesting specimen at hand is one with 11 arms, a single II Br 2 series being present; there are no other notable peculiarities, except perhaps that it is the most conspicuously yellow of all the specimens preserved. Another notable specimen, with arms 75-80 mm. long, has one arm which gives off a branch at almost right angles, in the vicinity of the 18th brachial; just one or two brachials further on the main arm forks and three brachials further the left hand arm forks again. This individual therefore has, distally, 13 arms. The largest specimens have arms about 90 mm. long and the cirri are XXX- XXXII, 28-30, while the smallest individuals have arms about 35 mm. long and the cirri XVI-XIX, 16-21. The 61 specimens before me were taken as follows: Western AustraUa: False Cape Bossut, September, 1929. 3 specimens. Broome, August and September, 1929. 40 specimens. Broome, Jime, 1932. 18 specimens. CALOMETRIDAE Neometra gorgonia A. H. Claek, 1914. Records W. A. Mus., 1, p. 125. Through the kindness of Mr. Glauert one of the paratypes of this striking species is before me, apparently either No. 5 or 6 of Mr. Clark's Hst. All the arms are broken but more than 36 and probably 40 were present; there are two IV Br series. There are 10 cirri (all broken) and 13 sockets on the margin of the large flat centrodorsal. The locality where this and all other known specimens of qorgonia were taken, is between Fremantle and Geraldton, 80-120 fms. APOROMETRIDAE, fam. nov. aporometra' gen. nov. Arms 10, relatively short, with comparatively few brachials and pinnules, but numerous syzygies. Centrodorsal low hemispherical completely covered with cirrus sockets and two or more marginal series of cirri. Cirri rather numerous, up to XXV, with very numerous short, wide segments, stout at base but tapering 1 6.-Kopo% = wilh no way through, difficult to find the way, hence puzzling, + metra, the widely used termi- nation for comatulid genera. 42 memoir: museum of comparative zoology gradually to a slender, somewhat flattened, recurved or loosely coiled tip; there are no spines, ridges or tubercles, except a very insignificant "opposing tubercle" on the penultimate segment. Disk thinly but more or less completely plated. First pinnule large, conspicuous and rather stiff, but with only about 10 segments; second pinnule much smaller and flexible, sometimes the smallest pinnule on the arm. Genotype, Himerometra paedophora H. L. Clark. The discovery at Bunbury of a second species of this peculiar genus has led to critical study of its characters by both Mr. Austin H. Clark and myself. We had previously disagreed radically as to its status, but are now in complete agree- ment. I am very grateful to Mr. Clark for his careful attention to this matter and especially appreciate his frank acknowledgment of the validity of the genus and the necessity of making a new family for its reception. So peculiar are the characters of this family that Mr. Clark says (in litt.) "trying to figure out its proper systematic position has been one of the most perplexing puzzles I ever tried to solve." He has finally concluded that Aporo- metra is nearer to Notocrinus than to any other genus, that each is unique in its own family, and that the two families form a separate suborder, the Noto- crinida. "The most satisfactory disposition of the Notocrinida would seem to be to consider it as a suborder within the Oligophreata, equivalent to, and on the same basis as the Comasterida, Mariametrida and Tropiometrida." It needed but slight comparison of the specimens from Bunbury with those from the coast of New South Wales to show that they are not the same species, but the possibihty that one or the other was identical with the comatulid from Victoria described by Bell in 1888 as Antedon wilsoni required consideration. Thanks to the kindness of Mr. D. Dilwyn John, an exchange was effected with the British Museum by which two cotypes of wilsoni came to the M. C. Z. Com- parison of these with the other material at hand was made independently by Mr. A. H. Clark and myself and we quite agree that there are at least three species of Aporometra now known from the southern coasts of Australia. These may be distinguished from each other as follows: Key to the Species of Aporometra A. Longest cirrus segments half again as long as broad. (Port Philip, Victoria) wiboni A^. Longest cirrus segments not longer than broad. Cirri with 25-35 (usually 28-30) segments of which most are about as long as broad. (New South Wales) paedophora Cirri with 39-61 (usually 40-50) segments of which even the longest are distinctly broader than long. (Southern Western Australia) occidentalia CLARK : AUSTRALIAN ECHINODERMS 43 Aporometra occidentalism sp. nov. Centrodorsal hemispherical or rounded conical, often more or less flattened, with cirrus sockets arranged in about 10 colimins of 2-4 each. Cirri about XX, 39-61 but usually 40-50, 20 mm. or more in length when fully developed nearly or quite equalling the arms; basally the segments are about twice as wide as long but distally, tho they become somewhat flattened, the width is relatively less; there are no projecting spinelets, tubercles or ridges distinct enough to be mentioned, except a minute opposing tubercle on the penultimate segment; the cirri taper very noticeably on the distal half and are frequently recurved or even coiled at the tip. Radials barely visible, the distal border slightly concave; IBri about three times as broad as long; axillaries broadly rhombic about twice as wide as long Arms 10, some 25-28 mm. long. First brachials about twice as long exteriorly as interiorly, approximately twice as broad as the external length; second brachials similar but with distal borders more oblique; brachials 3 and 4 united by syzygy; next two brachials about twice as broad as long but with oblique ends; the following brachials similar, commonly alternating with syzygial pairs; distally the brachials are distinctly broader than long. Pi is 7.5 mm. long with about 14 segments, smooth, somewhat flexible, tapering evenly from a rather stout base to a slender but not delicate tip ; fourth and following segments about 3 times as long as thick. ?„ is similar but Uttle more than half as large and has only 10 segments. P- is only 3 imii. long, with 11 segments, slightly longer than broad; the pinnule is rather stout basally, the third, fourth and fifth segments supporting a large ovoid gonad; the following pinnules to Ps are similar but sHghtly longer; after Fu there is no evident gonad; distal pinnules 3.5 mm. long with 12-14 segments, the terminal ones being minutely rough with recurved spinelets and tubercles. Color in life as described beyond. Holotype, M. C. Z. no. 964, from Koombana Bay, Bunbury, W. A., 5-8 fms. Perhaps the most rewarding feature of the excursion made by Professor E. W. Bennett and myself to Bunbury in October, 1929, was the discovery that this little comatulid is abundant in that vicinity. It was, therefore, possible to secure enough material to completely vaUdate the genus and establish the species. (Michaelsen and Hartmeyer had secured three specimens in Koombana Bay in ' occidentalis = western, in reference to its geographical occurrence on the Australian coast. 44 memoir: museum of comparative zoology 1905 but when A. H. Clark reported on their collection of crinoids, he considered these little comatulids as "very young" Ptilornetra macronema, and listed them accordingly.) WThile dredging in 5-8 fms. in Koombana Bay, October 26, 1929, one or two miles west of the breakwater, Bennett and I found comatulids very common, and easily recognized three species; the largest proved to be Comatulella brach- iolata, a much smaller and more delicate species was Compsometra incommoda, and the third and most common, was the present form. We preserved a large number of specimens, of which 148 are at hand. In alcohol they have undergone practically no change of appearance but the dry ones are paler and more rigid, of course. The color of the oral surface is always brown, usually very dark, the disk being marked more or less with whitish; the dorsal side of arms and calyx is brown, ranging from a distinctly yellowish, to a deep, somewhat purplish shade ; the cirri are often purplish, brownish or even reddish but are generally very light and often almost white, in rather marked contrast to the arms, which they some- times nearly approximate in length or even exceed. These cirri are efficient "holdfasts," admirably adapted to the bottom on which the comatulids were living, for this was covered, more or less, with algae (chiefly Cystophora) and other vegetation (chiefly Cymodocea). The dredge came up full of these plants and the various comatulids (not to mention other echinoderms) were well tangled with them and with each other by means of the long, recurving cirri of this species. It is worthy of note that we did not find Ptilometra at Bunbury — nor at any other point on the Australian coast, which seems a bit odd. The smallest specimens of Aporometra which we took at Bunbury were hardly 20 mm. across; the longest of the dozen cirri are about as long as the arms (say 8-10 mm.) and have some 35 segments. The largest specimens have the arms 25-30 mm. long and the longest of the 20-25 cirri are but little shorter; they have more than 50 segments of which the terminal dozen are very small. There are only 20-25 pinnules on each side of the longest arms. The most interesting biological fact about this comatulid is that it is vivi- parous; that is to say, the eggs are not shed from the genital pinnules but undergo their development in and on them. The original specimens from which the species pacdopliora was described, carried pentacrinid larvae in various stages of growth borne on the pinnules. Those individuals were collected in late summer or early fall (February and March). The specimens taken in Koombana Bay in late spring have the genital pinnules with large eggs and young embryos, not yet penta- crinids. Each pinnule seems to have 4 or 5 eggs but usually if development is CLAEK: AUSTRALIAN ECHINODEEMS 45 well under way there are but 3 or 2 embryos. I suspect that only one penta- crinid develops on each pinnule, the other eggs and embryos serving as nourish- ment for its growth. The whole life history could be easily worked out at Bun- bury during a summer, I have little doubt, and would be a very interesting and valuable study. It is possible that the work could be done at Fremantle, for besides the Bunbury material, I have 3 small specimens of this comatulid labelled as taken near Garden Island in 1929, which were given to me by Professor Bennett. ANTEDONIDAE COMPSOMETRA INCOMMODA Antedon incommoda Bell, 1888a. Ann. Mag. Nat. Hist., (6), 2, p. 404. Compsometra incommoda A. H. Clark, 1911b. Mem. Austral. Mus., 4, p. 792. This Uttle comatulid is found all along the western coast of Australia as far north at least as Geraldton. The specimen taken there is a rather small but quite typical indi\ddual which we found in a little tide-pool on the rocky shore north of the town. It seemed a most unlikely place for a comatuHd for there was little animal life in those particular tide-pools. The 59 specimens at hand were taken at the following points : Western Australia: Geraldton, tide-pool north of jetty, October 7, 1929. 1 speci- men, young. Dongarra. E. W. Bennett leg. et don. 9 specimens, young. Cottesloe, North Beach. L. Glauert leg. et don. 1 specimen, young. Fremantle, off Garden Island, 7 fms., October, 1929. 8 speci- mens. Bunbury, Koombana Bay, 5-8 fms., October 26, 1929. 6 specimens. Bunbury, Basalt Reef, September 28, 1930. E. W. Bennett leg. et don. 1 specimen, young. East of Cape Naturaliste, Bunkers Bay, January 15-17, 1930. E. W. Bennett leg. et don. 17 specimens, mostly young. Albany, under stones below low water mark, January 10, 1929. E. W. Bennett leg. et don. 16 specimens, young. 46 memoir: museum of comparative zoology compsometra loveni Antedon loveni Bell, 1882. Proc. Zool. Soc. London, p. 534. Compsometra loveni A. H. Clark, 1908. Proc. Biol. Soc. Wash., 21, p. 131. There are 79 specimens of this well-known comatulid in the collection at hand, including both adult and young. They were all dredged in Port Jackson, near Middle Head, in 4-5 fms., on November 21, 1929. DOROMETRA PARVICIRRA Antedon panicirra P. H. Carpenter, 1888. "Challenger" Comat., p. 204. Dorometra panicirra A. H. Cl.\rk, 1917. .Jour. Wash. Acad. Sci. 7, p. 128. This dehcate little comatuUd is apparently not rare in Port Darwin, as we met with it in both 1929 and 1932. There is great diversity of color, ranging in life from "a handsome light brown and white" (in the dry specimen the brown has become a rather faint gray-brown) to "fine black, white and cream-color . . . the finely variegated arms defied description." Another is described as "dull blackish with not very evident yellowish markings on pinnules and arms;" this specimen is now dull purple with cirri and pinnules and irregular markings on arms, nearly white. Specimens preserved in alcohol are yellowish-brown or pale brown with cirri and pinnules whitish. In spite of this wide diversity in color, all of the Dorometras at hand seem to represent but a single form and the resem- blance to Carpenter's Antedon parvicirra from the Philippine Islands is so close, it seems best to refer them to that species. The largest specimen, the dry, light-colored one mentioned above, has the arms 50-60 mm. long and 10 mm. across the widespread pinnules; the cirri are about XXX, 12-14, and the pinnules are those characteristic of the species; the first small, slender with 10-12 segments, the second somewhat longer and the third much longer and larger; the tips are broken on the pinnules so that it is not possible to give the exact number of segments, but the second has at least 13 and the third at least 18. A small Dorometra from an island north of Cooktown, Queensland, was given to me at Hobart, Tasmania, by Professor T. T. Flynn, and a critical exam- ination shows no good ground for not calling it parvicirra. The range of this comatulid is thus extended far to the south, adding one more species of echin- oderm to the rich Barrier Reef fauna. CLARK: AUSTRALIAN ECHINODERMS 47 The 6 specimens of parvicirra at hand are from the following localities : Queensland: Eagle Island, Great Barrier Reef, north of Cooktown. T. T. Flynn leg. et don. 1 specimen. Northern Territory: Darwin, near Shell Islands, 3-6 fms., July, 1929. 4 speci- mens. Darwin, near Leper Station, 3-5 fnxs., bottom covered with coralline algae, broken coral, sponges and alcyonarians. May 25, 1932. 1 specimen. MONILIMETRA ' gen. nov. Similar to Dorometra in most characters but distinguished by the more or less monihform first pinnules, which are shorter and much more slender than the second and third. Genotype, Monilimetra nomima sp. nov. This Uttle group of species, known as yet only from the vicinity of Broome presents a curious analogy to the West Indian genus Coccometra, as has been pointed out to me by A. H. Clark. For Coccometra contains three closely related species known as yet only from the northwestern part of the West Indian region and they are characterized by the moniliform first pinnule. But Coccometra and the related Thysanometra of the northeastern part of the East Indian region differ strikingly from Monilimetra in the much longer and more slender cirri and first pinnules. The new genus is, I beheve, much nearer to Dorometra than to the Thysanometrinae. There seem to be four quite different forms of Monilimetra m the Broome region, which may, for the present at least, be considered distinct species. It is not improbable that more abundant material will show a greater intergradation than the present available specimens indicate. The four species may be dis- tinguished as follows: Key to the Species of Monilimetra A. Cirrus segments 4-6, not conspicuously difFerent from the others, rather stout, slightly compressed, not twice as long as thick; color of arms dorsally, brown of some shade or yellowish or whitish, often more or less variegated with purple or dusky and light shades. ' monile = a necklace + metra, the widely used termination for comatulid names, in reference to the moniliform first pinnule. 48 memoir: museum of comparative zoology B. Normal mature cirri short, with 13-15 segments, the basal ones not much compressed; pinnules unicolor or at any rate not handsomely banded. Fourth pinnule very large, nearly or cjuite equalling the third; first pinnule con- spicuously moniliform with more than 23 segments; color brown of some shade, sometimes variegated, sometimes very light nomima P^ourth pinnule obviously smaller than third; first pinnule moniliform only in basal half, with fewer than 23 segments; color light, faintly mottled with dusky, some- times with a broad longitudinal blackish stripe on dorsal side of arms lepta B^. Normal mature cirri with 16-18 segments, all rather short, compressed though not markedly so, proximally; pinnules handsomely banded purple and white poecila A^ Cirrus segments 4-6 elongated, more than twice as long as the diameter of the nearly cylindrical middle; color purple and white; cirri pure white (except sometimes basally); arms usually deep purple but may be white and only purple on basal pinnules .... bicolor MONILIMETRA NOMIMA ^ Sp. nOV. Centrodorsal low hemispherical, not quite 3 mm. in diameter, well covered with cirrus sockets except at the pole ; those nearest the pole are very small while those near the margin are largest. Only 3 cirri are present and all of these are broken, but one lacks only the terminal claw; this one (fig. 5) has 13 segments, short, wide and, except basally, distinctly compres.sed ; segments 4-6 are longest but are not much longer than distally wide; opposing claw small and incon- spicuous. (As shown in two paratypes, the terminal claw is not conspicuous or pecuhar, but normally curved.) Radials concealed; I Br low, its width about 4 times its length; axillaries more or less triangular, with slightly concave sides, their length not quite equal to their width; the synarthrial joints between them and the I Br, and also those of the first brachials with them are not at all close. There are 10 arms of equal size; all are broken but they would not exceed 35 mm. in length, and the number of brachials was probably not more than 70 (counting syzygial brachials as one) ; the first brachials are low and twice as high on the outer margin as on the inner where they are rather markedly in contact ; the width is about equal to twice the height of the outer margin; the second brachials are somewhat higher and the third and fourth (united by syzygy) together equal the second; the succeeding three or four brachials are about twice as wide as high and nearly quadrangular; after that they become more wedge-shaped but never become triangular; syzygies occur between 9 and 10 and then at irregular but usually rather short intervals. ' f 6/ji/io$ = conventional, in reference to the fact that this species shows best the typical characters of he genus. CLARK: AUSTRALIAN ECHINODERMS 49 Pi (fig. 6) is about 12 mm. long, very slender, with about 23 segments, of which the basal six to twelve are about as wide as long but beyond that the length increases and the moniUform character of the pinnule disappears near the Figs. 5 and 6. Monilimetra nomima: 5, a cirrus (lacking terminal claw); 6, pinnules pi — ps. x 8. tip. P2 is about 14 mm. long and has some 25 segments but only 2 or 3 at the base are as wide as long, and distally they are twice as long as wide. P3 is much longer and stouter, apparently about 25 mm. long (the terminal portion is too much curved and twisted for measurement) with more than 30 segments. P4 is 50 memoir: museum of comparative zoology about equally stout but is not so long, although it has about 30 segments. Pe is much smaller and shorter with about 20 segments. The following pinnules are somewhat smaller but soon increase in length though they remain very slender. Beginning with the third pinnule, the outer distal corner of each segment projects as a minute spine and the distal margin begins to be slightly serrate; this spininess of the distal margin of the pinnule segments becomes very marked near the middle of the arm but decreases again on the distal pinnules. Color (dry) : brown, lightest on centrodorsal and arm-bases and again dis- tally, darkest on dorsal side of brachials 8-20, where it is a deep purple-brown; the change of shade is very gradual, and nowhere abrupt. Beginning at the very base of each ray, a light yellowish-brown line, ill-defined and soon broken into irregular patches, runs out on the dorsal side of each arm; there are also minute specks of this light shade on each side of this Une. Pinnules and cirri, light brown or even pale brownish-white. Arms distally, somewhat banded as each segment is brown with a broad light margin. Holotype, M. C. Z. No. 953, dredged at Broome, W. A., June, 1932. There are 4 paratypes and 2 other specimens which I am referring to this species. None of them is Uke the holotype in color. One is uniformly cream color dorsally and is probably bleached. Another appears to be uniformly yellow- brown, but under a lens, faint markings corresponding to those on the holotype are barely distinguishable. The other two paratypes are light reddish-brown or fawn-color, variegated with a darker purplish-brown; they are, however, in poor condition and were apparently stained, probably by contact with other echinoderms in the collecting or preparing of the specimens. A fifth specimen, somewhat similar to these but much hghter in color and a little smaller and more deUcate in structure, is referred to this species but is not considered a paratype. There are XXXIII cirri with 13-15 segments, but they seem more slender and much le.ss compressed than in the paratypes, none of which, however, have re- tained enough of their evidently very fugacious cirri for adequate comparison. As regards the pinnules, the proportions are just as in the types but the measure- ments and actual number of segments are considerably less than in the holotype. Thus Ps has fewer than 20 segments and is only 10-15 mm. long, and P4 is dis- tinctly smaller but is much more hke P3 than like P5. I take these pecuUarities of the pinnules to be youthful characters, which would have soon been outgrown. Another specimen referred to novmna is very young with arms only 10-12 mm. long. The pinnules are all very slender and with few segments and the cirri are about XXV, 11. The chief reason for referring this very juvenile specimen to CLARK: AUSTRALIAN ECHINODERMS 51 nomima is its resemblance to the much larger atypical fifth specimen just dis- cussed, the peculiarities of which I beUeve are due to immaturity even though the arms are fully 30 mm. long. Compared with the Comasteridae and Zygometridae, the Antedonidae are rare in the Broome region and were met with only now and then, not more than half a dozen times during each visit. The present species is undoubtedly the com- monest and was taken at least four times. Two specimens were taken in Roebuck Bay in August, 1929, and early in September two more were dredged near False Cape Bossut. In 1932, three specimens were secured at different times in dredg- ing between Broome and Cape Villaret. MONILIMETRA LEPTAI Sp. nOV. Centrodorsal low hemispherical, about 2 mm. in diameter, the bare pole rough with minute cirrus sockets; cirri XXXV, 15-17, crowded and more or less strongly recurved (fig. 7). Segments 4-6 are longest and least compressed but the length is scarcely equal to twice the diameter at middle ; distally the segments become shorter and wider and moderately compressed; opposing claw low and inconspicuous; terminal claw small but sharp and curved. Radials concealed; I Br low, its width at least 4 times its length; axillaries more or less triangular or low pentagonal; the lateral margins may be considered as blunt points on a triangle, or as very short sides of a pentagon; the width of the axillary is considerably greater than its length. Synarthrial joints not very close and synarthrial tubercles iiLsignificant or wanting. There are 10 approxi- mately equal arms, about 35-40 mm. long; the number of brachials probably exceeds 80; the character of the brachials and position of the syzygies is essen- tially as in the preceding species. Pi (fig. 8) is about 6 mm. long, with some 20 segments, of which only the basal 6 or 7 are wider than long, so the moniliform character of the pinnule is not conspicuous. P2 is similar but is longer and stouter with about the same number of segments but only about three basal ones are wider than long. P3 is conspicu- ously longer, at least 15 mm., and stouter, with 23 or more segments, all but the two or three basal ones much longer than wide. P4 is distinctly smaller with about 15 segments and Ps is similar but smaller still. Subsequent pinnules not ' XcTTTOS = delicate, in reference to the slender arms and generally fragile structure. 52 memoir: museum of comparative zoology peculiar. Spininess of distal margins of pinnule segments more or less evident under a lens but not at all conspicuous. Color (dry) : cirri, pinnules and sides of arms very pale brown, nearly white; dorsal surface of each arm with a broad dark brown or blackish stripe; in alcohol the color is not essentially different. Disk plump and relatively large, pale gray- ish, more brown on the anal papilla and dark along the margins of the food grooves. Figs. 7 and 8. Monilimetra lepta: 7, a cirrus; 8, pinnules, pi — ps. x 8. Holotype, M. C. Z. No. 950, from Broome, W. A., dredged in 5-8 fms. June, 1932. There are five paratypes, also dredged in the vicinity of Broome in June, 1932. Apparently all of the specimens were taken at the same time and place. They agree well in structural details and are distinctly more slender than the preceding species, nomima. In color they show some diversity and not one has the dark stripe on the arms. Three are uniformly very light, a dull pale cream color on the dorsal side; one of these has the cirri white with no markings, while in the second a few distal .segments of the largest cirri have a dusky blotch; the third has the cirri dusky or purplish-gray. The two other specimens have more or less dusky or hght brownish-purple blotches on the brachials, on each side, but CLARK: AUSTRALIAN ECHINODERMS 53 these are not conspicuous; in the larger, the cirri are nearly white, but in the smaller they are almost lead-color. A seventh specimen, which may properly be called a paratype, was taken on Pearl Shoal in 1929; the arms are all broken and the cirri are nearly all gone but I do not hesitate to refer it to the present species as the pinnules correspond com- pletely and such cirri as are present agree with those of lepia. The color, a uni- form light shade not quite white, is also appropriate. This species is the most Dorometra-hke of any of the four Monilimetras at Broome. The monihform character of the first pinnule is by no means striking and the resemblance of the comatuhd to Dorometra is obvious. MONILIMETRA POECILA' Sp. nOV. Centrodorsal low hemispherical, more than 2 mm. in diameter, the small bare area at the pole without evident cirrus sockets; cirri XXXIII, 16-18, crowded and more or less recurved (fig. 9) ; segments short and wide, the first three are wider than long, the fourth is about as wide as long, the fifth and sixth are evi- dently longer than wide ; the following segments are about as long as the distal width; opposing claw low and insignificant; terminal claw small but sharp and curved ; while the cirri are flattened on the sides they are not markedly compressed. Radials concealed; I Br low, its width about 4 times its length; axillaries low pentagonal, lateral margins very short, width greater than length. Synarthrial joints not at all close, the lower brachials in particular being very noticeably separated, except at the very margin. Arms 10, rather unequal, longest exceeding 50 mm.; several are regenerating distally; the number of brachials exceeds 140 on the longest arm. The form of the brachials and the position of the syzj^gies is essentially as in nomima. Pi (fig. 10) is about 8 mm. long, with some 20 segments, of which the basal are as wide as long and even distally the length exceeds the width very little, yet owing to the lack of constrictions between the segments the monihform character of the pinnule is not so striking as might be. P2 is similar but is longer and stouter; the number and relative proportioas of segments is about the same. P3 is con- spicuously larger in every way, nearly 20 mm. long with about 24 segments. P4 is very similar to P3 and Uttle if any smaller. P5 and subsequent pinnules decrease uniformly in size to about Pi 1 or P 12 after which the length increases a little but 'ttoikiXos = variegated, in reference to the strikingly handsome coloration. 54 memoir: museum of comparative zoology the stoutness does not. Spininess of distal margins of pinnule segments more or less evident on P4-P10 but is not very conspicuous. Color (dry) : variegated, whitish, yellow and purple; ground color of centro- dorsal, cirri, base of arms and pinnules, whitish, of dorsal side of arms, yellow ; Figs. 9 and 10. Monilimetra poecila: 9, a cirrus; 10, pinnules pi — ps. x 8. on each cirrus segment (excepting the most proximal and the two terminal) is a conspicuous purple spot, and often these spots tend to run together dorsally, making an incomplete (sometimes complete) girdle around the segment; each pinnule has 8-10 purple bands of somewhat indefinite position and completeness; each basal brachial has a large purple spot on each side dorsally but beyond the eighth, the spots tend to coalesce more or less into an indefinite band of dull purple or dusky; the regenerating portions of the arms are a uniform hght yellow CLARK: AUSTRALIAN ECHINODERMS 55 without markings but the pinnules borne thereon are banded; disk variegated whitish and purple. Holotype, M. C. Z. No. 956, dredged in 5-8 fms. between Cape Vilaret and Broome, W. A. June, 1932. There are three paratypes but all are more badly broken than is the holotype. One of these, taken at the same time and place, is similar in size and structure so far as can be determined, but is much darker in color, the yellow on the arms is pale, the purple is much more abundant and the spots have coalesced into broad irregular bands. The other two paratypes show the same contrast in color but it is even more marked ; the hghter one is nearly white (possibly the yellow has been bleached in the alcohol in which it is preserved), variegated with a dull reddish- purple (the shade may have been altered in alcohol) and with the pinnules less frequently and conspicuously banded; the darker specimen is deep red-purple with a whitish mid-dorsal line on the I Br series and the pinnules purple and white banded as usual. This unusually lovely comatulid is very sensitive to handhng and only the holotype has any arms still attached to the calyx. Two of the paratypes were taken September 5, 1929, "at extreme low water, far south of the jetty" at Broome. My field notes go on to say, "ten long, slender, very graceful arms; finely variegated with purple and white ; very delicate and went all to pieces in pail, breaking the arms up into little bits." When the two specimens taken in 1932 were dredged the present holotype was very little damaged, so it was possible by plunging it at once into strong alcohol to get a very good specimen, even the cirri remaining in place. Obviously this fine species cannot be considered at all common in the Broome region. MONILIMETRA BICOLOR^ Sp. nOV. Centrodorsal low, hemispherical, about 2 mm. in diameter, almost com- pletely covered with cirri, which are about XXXVIII, 15-17, crowded and more or less recurved (fig. 11) ; segments 4-6 elongated, more than twice as long as the diameter of the nearly cylindrical middle ; succeeding segments longer than wide and increasingly compressed; opposing claw sharp and rather prominent ; terminal claw sharp but not strongly curved. Radials concealed; I Br low, its width at least 4 times its length; axillaries ' bicolor = of two colors, in reference to the characteristic coloration. 56 memoir: museum of comparative zoology low pentagonal with the lateral margins very short and the width much greater than the length. Synarthrial joints very open at least as far as the ninth or tenth brachial. Arms 10, approximately equal, about 35-40 mm. long; the number of brachials exceeds 80; the character of the brachials and position of the syzygies is essentially as in nomirna. Figs. U and 12. Monilimetra bicolor: 11, a cirrus; 12, pinnules pi — ps. x 8. Pi is 5 or G mm. long with some 20 segments, the basal ones short and wide, the distal somewhat longer than wide ; on the whole the pinnule is quite monili- form. Po is longer and less moniUform but has about the same number of seg- ments. P3 is much longer and stouter, 12-15 mm. long with about 25 segments, all but the basal 2 or 3 much longer than wide. P4 is almost as long and stout as P3, while Pe is markedly smaller and Pio is the smallest of the pinnules. Spininess of distal margins of pinnule segments, not very marked but with a lens it is evi- dent on the basal and middle joints of pinnules 3-10. Color (dry) : calyx and basal part of arms, deep-purple ; pinnules (except some of the larger basal ones) Ughter and distally quite Ught, so the arms appear distinctly Ughter near tips; cirri pure white in striking contrast, the very basal segments, however, are deep purple. CLARK: AUSTRALIAN ECHINODERMS 57 Holotype, M. C. Z. No. 946 from 5-8 fms., between Cape Villaret and Broome, W. A., June, 1932. There are three paratypes as follows : Near False Cape Bossut, September, 1929. 1 specimen like the holotype, but all cirri and most pinnules are wanting. Near entrance to Roebuck Bay, June, 1932. 1 specimen somewhat larger and darker than the type (pinnules are dark) ; all cirri wanting; arms about 50 mm. long. Pearl Shoal, 7 fms. September 26, 1929. 1 superb specimen. Arms about 60 mm. long, cream-white, but pinnules on basal 10-18 mm. of each arm, red-purple in sharp contrast ; cirri pure white. This species is very close to nomima and might perhaps be considered a color form of that species but the difference in the cirri seems to be constant and the coloration of the present form is so striking it seems better to regard it as a dis- tinct species. It is evident that it cannot be considered common in the Broome region. ASTEROIDEA The collection of sea-stars contains 4270 specimens, representing 46 genera and 104 species. Of the genera 3 are new, one being simply a subdivision of the old genus Anthenea, while each of the other two is based on a single specimen of a sea-star so very unusual as to necessitate a new generic group. Of the 104 species, no fewer than 29 represent hitherto undescribed species, of which a dozen belong to the family Asterinidse and half a dozen others represent the old genus Anthenea, in a broad sense. The new genera are : Styphlaster, type S. notabilis sp. nov. Monotypic. Gymnanthenea, type Anthenea globigera Dod. There is a second species at the Abrolhos Islands, W. A. Manasterina, type M. longispina sp. nov. Monotypic. The 29 new species are scattered through 16 genera, as follows: Astropecten pidcherrimus W. A., Broome. Luidia hexactis W. A., Montgomery Reef, Collier Bay. Archaster laevis W. A., Broome. Goniodiscaster acanthodes W. A., Broome. Goniodiscasier bicolor W. A., Broome. 58 memoie: museum of comparative zoology Styphlaster notabilis Anthenoides dubius Gymnanthenea laevis Anthenea acanthodes Anthenea crassa Anthenea elegans Anthenea obesa Anthenea polygnatha Bunaster variegatus Asterina alba Asterina heteractis Asterina lutea Asterina perplexa Disasterina spinulifera Manasterina longispina Paranepanthia rosea Patiriella brevispina Patiriella nigra Nepanthia magnispina Nepanthia tenuis Nepanthia variabilis Parasterina occidentalis Echinaster varicolor Astrostole insular is W. A., Broome. W. A., Broome. W. A., Abrolhos Islands. Q., Port Curtis. Q., Port Curtis. W. A., Broome. W. A., Geraldton. W. A., Broome. W. A., Bunkers Bay. Lord Howe Island. Lord Howe Island. W. A., Broome. Lord Howe Island. W. A., Broome. W. A., Rottnest Island. W. A., Rottnest Island. W. A., Bunbury. Lord Howe Island. W. A., Augustus Island. W. A., Broome. W. A., Broome. W. A., Point Peron. W. A., Broome. Lord Howe Island. As shown by the above list Lord Howe Island is the home of 5 of the new species, and Port Curtis, Queensland of 2; all the remaining 22 are Western Australian. Captain Bardwell secured 2 of these at or near Augustus Island, while the vicinity of Broome is the home of 13. The coast from Geraldton southward to Bunkers Bay, with its outlying islands (Abrolhos, Rottnest) provides the type localities for 7 species. Sea-stars are the most generally observed and collected of Australian echino- derms. The more or less striking form and the frequently briUiant colors make even a pearl-shell diver stop and look, while fishermen of all sorts are apt to bring them in, now and again, as a part of their spoil. The clear waters and firm sandy bottoms in the vicinity of Broome make that part of the Australian coast the richest collecting ground for a student of sea-stars, no less than one-third (35) of the present list having been taken in that region. The Barrier Reef area also has CLARK: AUSTRALIAN ECHINODERMS 59 a notable sea-star fauna but is not so well represented in the present list since we did not do any collecting there. On the southern coast of the continent the aster- oid fauna is less abundant but is notable for many characteristic genera and species, of which only a few appear in the present work. Western Australia between Geraldton and Albany has a very coiLsiderable number of sea-stars, many of which are characteristic of that coast. In all regions, the smaller species live beneath rock fragments or in crannies and holes in the rocks while the larger ones commonly lie more or less exposed on the bottom. A few species frequent wharf piles or lie buried just below the surface on a sandy bottom, but none will be found where the water is brackish or befouled with mud. The literature dealing with Australian sea-stars is fairly extensive. Sladen's monumental work on the "Challenger" Asteroidea (1889) is of course invaluable. Bell's report on the "Alert" collection (1884) suffers from that writer's well- known idiosyncrasies. My reports on the "Thetis" Echinoderms (1909), the echinoderms of the Western Australian Museum (1914), the "Endeavour" echinoderms (1916), the Echinoderms of Torres Strait (1921), "Some Echino- derms from West Australia" (1923), and the echinoderms of the South Australian Museum (1928) all contain more or less material concerning sea-stars. Fisher's admirable and always trustworthy monographs (1906, 1911, 1919, 1928, 1930) and Doderlein's^ very helpful accounts of Semon's Torres Strait sea-stars (1896) and of the "Siboga" asteroids (1917, 1920, 1935) are indispensable. Doderlein's small monographs on Anthenea (1915) and Oreaster (1916) are also valuable. In very recent years, Livingstone has published a number of excellent papers (1930- 1936) which warrant the belief that Australian sea-stars now have a master, whose work will be worthy of the rich and fascinating field into which he has entered. In the study of the present collection, I have had the very great advantage of constant correspondence with both Fisher and Livingstone. The latter has seen nearly all of the collections and the former has examined many of the more inter- esting or perplexing specimens. The unfailing interest and helpful suggestions of these colleagues have been of incalculable value and have no doubt saved me from errors. If any serious blunders are detected in the present account of Australian sea-stars, it will be the fault of the writer only and in spite of my colleagues' cheerful readiness to help. Through Mr. Livingstone, the AustraUan Museum, has generously loaned me material of the greatest importance for my attempts to elucidate the perplexing genera Anthenea and Nepanthia. ' The recent death of Dr. Doderlein is a great blow to all students of echinoderms, for the care with which his work was done and his sound judgment made his work invaluable to his colleagues even when we could not accept his conclusions. 60 memoir: museum of comparative zoology In collecting sea-stars, little care need be used in the handling, save in the case of Luidia. In this genus, there is a tendency, when roughly or even carelessly handled, to constrict muscles at the base of each arm which autotomously sever the arm from the disk; once this process begins it is almost impossible to check it. But if the specimen can be placed in a basin of fresh sea-water before the impulse starts, narcotization with Epsom salts (MgS04) is usually easy, after which killing can be done without danger of injury. All sea-stars, so far as my experience goes, are very susceptible to Epsom salts, after the use of which, they may be placed in alcohol or formalin for kiUing. They may be left in alcohol indefinitely but never in formaUn more than a few hours, at the most. If they are to be dried, the for- mahn (about 4% of the commercial fluid) should contain some corrosive subli- mate (HgClo) — not too much or a whitish film may form on the specimen; something less than 1% HgCl.> is ample.' Better specimens of many sea-stars, particularly those with arms of considerable bulk, may be prepared if instead of being narcotized they are killed by being transferred directly from sea-water to fresh-water; they may later be placed in the corrosive-formahn until completely saturated therewith and then dried as rapidly as possible in the shade or in an oven, not too hot. The colors of sea-stars are lamentably evanescent and it is impossible to retain them in Museum material, except when they are dull and of little interest. AlcohoHc specimens always bleach more or less but specimens killed quickly in corrosive-formalin and then dried rapidly and thoroughly, may retain a considerable amount of pigment and will hold for years more or less resemblance to the living animal. ASTROPECTINIDAE ASTROPECTEN GRANULATUS Plate 1, fig. 1 MiJLLER and Tkoschel, 1842. Syst. Ast., p. 75. This is the common sea-star at Broome, occurring in considerable numbers on the sandy-mud bottom of Roebuck Bay, near normal low water mark and further out. Most of the specimens seen were of moderate size, with R = 50-75 mm. but many larger ones were taken with R = 100-118 mm.; the smallest at • Dr. R. T. Jackson has prepared beautiful dried specimens, using simply a very dilute solution of cor- rosive sublimate. His account of his method should be consulted (1930). CLARK: AUSTRALIAN ECHINODERMS 61 hand, with 10 superomarginals in each series, has R = 7 mm., r = 3 mm., and br = 3.5 mm. while in the largest with 45 superomarginals, R = 118 mm., r = 22 mm., and br = 25 mm. The change of form from youth to age is therefore very great, the ratio of R to r and br undergoing a marked alteration; in youth R = 2.3r and 2 br while in the full grown adult R = 5.4r and 4.75br'-. But the adult ratios are approximately reached while granulatus is still less than half grown; specimens with R = 50 mm. have very nearly the proportions of adults. It is of interest to note that while the arm increases in length from 7 to 50 mm., i.e., 7x, the number of superomarginals has increased only from 10 to 30, i.e., 3x; when the arm reaches 118 mm., i.e., 17x the youthful length, there are only 45 superomarginals; the arm-length has thus inci'eased about four times as fast as the number of supero- marginals. In life, the dorsal surface of this sea-star is a deep bluish-gray, very similar to the color of tlie sandy mud in which it lives; this color disappears completely in preserved specimens. The lower surface and marginal spines are pure white ; the pedicels are cream-colored. In young individuals when R = 12-15 mm., the tips of the arms are dusky; as the arms increase in length tliis dusky area forms a band across the tip; when R = 20 mm., a second dusky area may appear and form another band across the arm. These bands are brown in color when fully formed and brown areas now appear, in many individuals, on the disk in the inter- radii. As the specimens increase in size, these bands and areas break up into irreg- ular blotches and spots occupying the same general positions near the middle of the arm and in the interradii. They reach their maximum development when R = 50-60 mm. ; after that they begin to disappear and in specimens with R = 100 mm., they are poorly defined or lacking. Specimens dredged on a bottom of ordinary sand in 5-7 fms. were more yellowisli-gray or brownish than those found at and near low watermark around the jetty at Broome. In the intertidal zone, this Astropecten was easily detected when the tide was out by the star- shaped figure in the sand, caused presumably by the water currents among the paxillse of the dorsal surface; hence the size as well as the form of the sea-star was indicated though the animal itself was wholly buried. Some years ago, I recorded Astropecten granulatus from South Africa (1923, Ann. So. Air. Mus., 13, p. 250), the identification being based on a dozen speci- mens with R = 7.5-38 mm., and I even affirmed that Sladen's species monacanthus ' It must be constantly borne in mind that the form and proportions of sea-stars are often modified to a notable degree by the method of preservation; diy specimens are greatly affected by the rapidity or slowness of drying. This is particularly true in Astropecten. 62 memoir: museum op comparative zoology was the same thing. Now that a large amount of material is available, it is per- fectly clear that monacanthus is a distinct species and the two can be easily dis- tinguished. Comparison of the South African specimens with the material from Broome indicates that it is quite different and further study suggests that it is nearer to A. notograptiis Sladen than to any species hitherto described. Hence granulahis should be deleted from the list of South African sea-stars. The 53 specimens of granulatus at hand were all taken in or near Roebuck Bay, mostly in the vicinity of the jetty at Broome. Astropecten hartmeyeri DoDERLEiN, 1917. "Siboga" Ast.: Astropecten, p. 156. The material upon which this species is based was collected at Shark Bay, Western Austraha. It is not surprising therefore that it proves to be a common sea-star in the vicinity of Broome, but it is interesting to note that the range extends .still further to the northeast as far as Augustus Island, where Captain Bardwell collected 3 fine specimens. The largest of those taken at Broome has R = 82 mm. and is therefore nearly twice as large as the type material, but a still larger specimen is at hand, in which R = 110 mm., r=18 mm. and br = 20 mm. The arms are thus relatively more slender than in Doderlein's material but this is what would be expected, as shown by the measurements given for A. granulatus (See p. 61). This very large, dried specimen belongs to the Western Australian Museum and is from the coast of that State but the exact locality if not known. The color is very dark, an apparently uniform brown without any markings, quite different from the hand- somely marked smaller specimens, but, except that the spinulation is much coarser than in the Broome material, the resemblance of this big individual to the smaller ones is surprisingly close. The largest Broome specimen is the darkest of that series but is as clearly marked as the smaller ones; the smallest (R = 10- 14 mm.) are very hght, except for the darker markings. A single specimen, taken between Broome and Wallal, with R = 36 mm., is very dark and the markings are difficult to make out; such an individual might easily grow into a dark, unicolored adult. The specimens from Augustus Island arc noticeably different from those taken at Broome in the much wider rays; R = 60 mm., while r and br are each about 15 mm. A Broome specimen with R = 65 mm. has r and br only 12-13 mm. CLARK: AUSTRALIAN ECHINODERMS 63 and this is the usual ratio in the adult Broome material, but some of the smaller specimens have R only = 4 r. The color of the Augustus Islands specimens also sets them apart, for while the dark markings are evident, the ground color is dull and there is a dingy pink tinge evident, particularly on the lower surface. This may however be artificial due to some chance element in their preservation, for they are admirably dried. Whether hartmeyeri is really distinct from zebra Sladen seems to me rather doubtful but the only specimens of zebra available for comparison are too young to be of real service. Larger collections from Northeastern Austraha are necessary for a final decision. The 31 specimens of hartmeyeri at hand come from the following places, all on the coast of Western Austraha. Augustus and Champagay Islands, October, 1933. Capt. Beresford E. Bardwell leg. 3 specimens. Broome, 5-8 fms., September, 1929. 2 specimens, adult. Broome, Pearl Shoal, 5-8 fms., September, 1929. 4 specimens, young. Broome and vicinity, 5-8 fms., June, 1932. 20 specimens, adult and young. Between Broome and Wallal, 1930. R. A. Bourne leg. 1 small adult. Loaned by the Australian Museum. Locahty unknown. 1 very large adult. Loaned by the Western Austra- lian Museum. ASTROPECTEN MONACANTHUS Sladen, 1883. Jour. Linn. Soc. London: Zool., 17, p. 263. Sladen's types were only half-grown but Doderlein (1917, p. 150) has given a good account of this sea-star with excellent figures. It is a widespread species of the Indian Ocean, and has been taken previously on the northwestern coast of Austraha. It is not very common in the vicinity of Broome and was not taken above the lowest tide mark or near the jetty, but there are 13 specimens at hand, of which 4 young ones were dredged in September, 1929, near, if not in, Roebuck Bay, one half-grown individual (a loan from the AustraUan Museum) was taken by Mr. R. A. Bourne, in 1930 between Broome and Wallal, and 8 were dredged in June 1932 in 5-8 fms., southwest of Broome; the last lot range from very young ones with R = 7 mm. to an adult with R = 48 nma., the largest specimen yet re- corded. In this specimen r = 11 mm. and br= 12.5 mm., so that R is a httle more than 4r and a trifle less than 4br. A rather marked change in form occurs after 64 memoir: museum of comparative zoology R = 16-18 mm. Prior to that the arms are pointed, their width at base is more than one-third R and the sides are straight; thereafter the arms become blunt, the width at base is proportionately less but near tip is more, and the sides curve inward distally. The color in life (grayish, more or less evidently blotched with dusky or brown, particularly near the middle of the arms or distal thereto) is so much like that of granulatus that the two species were confused during collecting and only after they were dried and examined with a lens were the very obvious differences between the two species noticed. The opinion I expressed some years ago (1923, Ann. S. Afr. Mus., 13, p. 250) that the two species are identical is quite wrong. The difference in the spinulation of the inferomarginals is not emphasized by Sladen but Doderlein's figures (1917, pi. XVII) bring it out fairly well. The flat, rounded scale-like spines which cover the inferomarginals in monacanlhus are recognizable even in specimens in which R = only 7 mm. and are distinctive in specimens with R = 13 mm. The conspicuously flattened and widened outer adambulacral spines of monacanthus are also distinctive after R = 12 mm., and in large specimens (R = 40-50 mm.) the difference in the shape of the arms is ob- vious. ASTROPECTEN POLYACANTHUS MiiLLER and Troschel, 1842. Syst. A.st., p. 69. We met with this species only at Port Jackson, from which place it has long been known. There 2 specimens, with R = about 60 mm., were dredged near Middle Head, November 21, 1929. In Ufe the coloration was striking; the pax- illar area of the dorsal surface was deep purple, while the oral surface was more or less red-orange, with the margins pale yellow; the marginal plates were more or less purplish dorsally but faded out into yellow orally. These fine colors have entirely disappeared in the dry specimens. Mr. Melbourne Ward has generously sent me 2 unusually large and striking specimens labelled: "Beaches. Lindeman Island. Aug. 1934." This locahty is on the Barrier Reef not far from Port Mackay, Queensland. One of the specimens has R = 105 mm. (or more) and r only 21 ; it is thus a little larger than Miiller and Troschel's type-specimen (see Doderlein, 1917, p. 135) and the rays are a little narrower. The color is a rather bright brown with the unusually long marginal spines whitish; many of the inferomarginal spines are 10 to 1 1 mm. long and little more than a miUimeter wide, quite flat but sharply pointed. The other specimen CLARK: AUSTRALIAN ECHINODERMS 65 is a little smaller and the color is a gray-brown with the spines white. The mar- ginal spines are as conspicuous as in the larger specimen and even more bristling; several are 12 mm. in length; they are as a rule somewhat less flattened than in the other indi\'idual. On the whole these specimens resemble more closely one in the M. C. Z. collection from Zanzibar, with R = 85 mm., than they do the Port Jackson specimens, with their shorter, wider rays. Astro PECTEN preissii MuLLER and Troschel, 1S43. Arch. f. Naturg. 9, 1, p. 119. Southwestern Australia is the type locality for this fine species but the range extends eastward along the southern coast at least as far as St. Vincent GuK. Michaelsen and Hartmeyer found it near Fremantle and also near Albany, W. A. Doderlein (1917, p. 162) suggests that the latter specimens represent a local form which he calls albanicus, but some of our specimens from Fremantle correspond very closely with the description given, though they are less like the figure. Only further collecting in southwestern Australia can determine whether the wide- armed form is a constant variety but I am inclined to doubt it. Between Fremantle and Garden Island this sea-star is quite common and we secured a good series ranging from R = 6 mm. to R = 155 mm. In the very small specimens the rays are of course notably short and R = only 2r; when R = 32 mm., the proportion is R = 4r; at 70 mm. R = 5r; at 126 mm., r = 6r; and at 155 mm., R = 7r. In the last mentioned specimen, there are 78 superomarginals, and a number of them scattered here and there bear small erect spines as in Doderlein's largest specimen. In life, there are two color forms but as the colors fade completely after drying and all assume the usual "museum-color," no other difference can be detected. The possibihty of this being a secondary sex character deserves investi- gation. Most commonly the color is yellowish-brown or brownish-orange, with the distal ends of the inferomarginal plates orange. In striking contrast, the other form is rich \aolet. Both are unicolor without markings. The material in hand consists of 40 specimen>s from the following localities: Western Australia: Shark Bay, 2 fms., December, 1923. 1 adult. Loaned by the Western Austrahan Museum. Between Fremantle and Garden Island, 2-3 fms., October, 1929. 30 specimens, young and adult. Locality doubtful. 9 specimens, young. 66 memoir: museum op comparative zoology astropecten schayeri DoDERLEiN, 1917. " Siboga" Ast. : Astropecten, p. 60. It is a matter of great interest to find in the Victoria National Museum at Melbourne, a specimen of this remarkable Astropecten. It is of almost exactly the same size and appearance as the type but is in somewhat better condition. The group of actinolateral plates in each interradial area is very striking as are the very numerous adambulacral and oral spines. Doderlein's opinion that this is the most primitive of known Astropectens seems to me quite justifiable. The unique holotype was taken in "Tasmania" and is now in the Berlin Museum. The Melbourne Museum specimen bears the label "Portland, Victoria. Presented by Goldstein, 19.4.79." The resemblance of this species to A. pedinatus Sladen from the southeastern coasts of Australia is striking, and one is at first inclined to think them identical, but in addition to the important diiTerence in the actinolateral plates, which Doderlein stresses, the difference in the large spines of the inferomarginal plates is very great, the adambulacral armature is different and the madreporite is more nearly marginal in schayeri. Astropecten triseriatus MiJLLER and Troschel, 1843. Arch. f. Naturg. 9, 1, p. 118. On October 14, 1929, while dredging between Fremantle and Garden Island, we took an arm fragment of an Astropecten which was conspicuously different from preissii, the common Astropecten of that area. The superomarginal plates bore many spines in striking contrast to the unarmed superomarginals of preissii. The dorsal surface of the fragment was grayish, the sides and lower surface cream-color or yellowish. The fragment is about 50 mm. long, 20 mm. wide at one end, 15 mm. at the other. Comparison with a specimen of triseriatus loaned by the Western Australian Museum, in which R = 80 mm., shows that the frag- ment is part of the middle of the arm of a triseriatus in which R = about 1 10 mm. The type locality for the species is Southwestern Australia but it seems to be rare and few specimens are known. Doderlein (1917, p. 126) records one, with R = 97 mm., from Northwestern Australia which he compares with the type, a much smaller individual. The specimen loaned by the museum at Perth is in good condi- tion, though one ray is regenerating. R = 80 mm., r= 16, R = 5r. Nearly all the superomarginal plates have 3 erect spines and some have 4. This individual was found at South Beach, Fremantle, in November, 1932. CLARK: AUSTRALIAN ECHINODERMS 67 ASTROPECTEN VAPPA MiJLLER and Troschel, 1843. Arch. f. Naturg., 9, 1, p. 119. The type of this species is a very small specimen in the Berlin Museum with R = 23 mm. Of the 7 specimens at hand, 4 are of approximately the same size, but 3 are much larger. One has R = 40 mm. and is a fairly typical example of the species. Together with a smaller specimen it bears a label with a note on the color in life but there is no locality given. The note on color is written on the back of an AustraUan Museum label and reads : "Pale gray dorsally, more or less cream orally and marginally, with interradial superomarginal spines tipped with bright brown-orange." There is little doubt that these specimens were dredged near Middle Head, Port Jackson, November 21, 1929. The species has long been known from the New South Wales coast though unfortunately misidentified by both Whitelegge (who called it triseriatus) and myself (who called it pectinatus), as Doderlein long ago pointed out (1917, p. 125). My accounts of pectinatus in the "Thetis" Report (1909) and in the "Endeavour" Report (1916) undoubtedly refer to vappa; the only excuse that can be offered is that vappa was quite inade- quately known prior to 1917, when Doderlein's masterly revision of the genus Astropecten was published, bringing order out of chaos and making all students of sea-stars his lasting debtors. Besides the two specimens from Port Jackson there are three young individ- uals from the southwestern coast of Au.straUa, given to me by Professor E. W. Bennett. One is very well preserved but lacks a locality label. The others are obviously from beach wrack and are labelled: "Drift on Middleton Beach, Febru- ary, 1929. E. W. Bennett." Middleton Beach is near Albany, W. A. The type locality for vappa is Southwestern Australia and it has been recorded from Shark Bay, W. A. by Doderlein (1917). Mr. Melbourne Ward has kindly sent two specimens of vappa bearing the label: "Sand flats. Lindeman Island, Cumberland Group, Queensland. July- October, 1934." This is much the most northern point at which vappa has been taken. These indi\'iduaLs are about half grown, with R = 60-68 mm. The color seems to have undergone little change; the upper surface is variegated hght and dark gray; the lower is cream-white. The superomarginal spines are relatively few and very small, but the inferomarginals are very long (often 6-7 mm.) and conspicuous. 68 memoir: museum of comparative zoology astropecten velitaris Von Martens, 1865. Arch. f. Naturg., 31, 1, p. 360. This East Indian sea-star is not uncommon in water 5-8 fms. deep, near Roebuck Bay and southwestward along the coast. Nearly all of the 17 specimens dredged in September, 1929 and in June, 1932 are adults and several are larger than any specimens yet recorded. The series ranges in size from R=14 mm. (R = 4r) to R = 96 mm. (R = nearly 7r). The large specimens thus show a much greater length of ray proportionately than has previously been accredited to the species. Even the smallest individual with R only 14 mm. has the erect spines on the first superomarginal plates conspicuous while the occurrence of spines on the second superomarginal plate is common in specimens with R = 60 mm. or more. In no case however are more than 4 spines present in a single interbrachial arc. In the living animal the arms are relatively wide and flat, or about equal to one- fourth R, but the proportion alters when the specimens are killed (particularly if subsequently dried) and in some large specimens R = 6 br. This is a handsome species when living as the following notes on the color in hfe indicate. "Youngest specimens greenish-gray, irregularly blotched with a dusky shade; just a tinge of purple on the superomarginal plates and a few supero- marginal spines showing orange at base. Adults, oral surface and inferomarginal spines white, but there is more or less of an orange area on the basal half of each spine. Upper surface fawn-color with two (sometimes only one) irregular blotch- like cross-bands of a dark shade near middle of each arm or distal thereto. Superomarginal plates, terminal plate and an interradial line or blotch extending inward on disk from margin two-thirds of the way to center, purple or violet. Superomarginal spines, orange, tipped with purple. In the largest specimens, the colors are less bright." AsTROPECTEN PULCHERRIMUS' Sp. nOV. Plate 1, fig. 2 R = 62 mm. (in hfe about G8) ; r=16 mm. (in hfe, about 20); br=17.5 mm. (in hfe, about 22). In life R is about 3.5r and rather more than 3br but in the dry holotype R nearly equals 4r and 3.5br. In a small paratype (also dry) with R = 46 mm., the arms are a little narrower and R = 4br. Rays taper gradually to a ' pulcherrirnijs = most beautiful, as this is certainly the lovehest member of the genus. But alas, the beauty is ephemeral! CLARK: AUSTRALIAN ECHINODERMS 69 rather wide and rounded tip ending in a moderate ternunal plate having a dis- tinct median depression. Superomarginals 28, those in the interbrachial arc more than 4 mm. high, not 2 mm. long (i.e. parallel to the longitudinal axis of arm) and approximately 2 mm. wide, but width merges in height too gradually to make an accurate statement possible; distally the plates become lower and occupy more of the upper surface of arm, the eighteenth for example being only about 2 mm. high and 2 mm. long but nearly 3.5 mm. wide; superomarginals and terminal plate covered with granules, of which those at the upper end of the interbrachial plates are largest (about .30 mm. in diameter) while those on the distal and ter- minal plates are scarcely half so large; there is no indication of superomarginal spines. Paxillae numerous and densely crowded; the larger ones in the interradial areas have about 16 marginal papillae, short and thick, and about 8 granules on the top; at base of arm there are 14-16 longitudinal series of paxillae, and the number does not decrease greatly towards the arm tip but the indi\'idual paxillae become smaller and smaller. No madreporite is visible. Interradial areas below with only 2 plates on each side of the midline; each of these carries 8-10 spinelets most of which are flattened and widened at tip. Inferomarginal plates about 28 in number, their size and position corresponding well with the superomarginals; they are covered closely with short, flat, scale-hke spinelets, widened and truncate or rounded at tip; at the outer end of each plate is a single stout spine, which on the proximal plates is flat and truncate but be- comes less flat and more pointed distally; at the base of the arm, these spines exceed 3 mm. in length and are about .75 mm. wide at tip; below this terminal spine but along the aboral margin of the plate (except on the first plate where it is nearly central) is a widely spaced series of 3 or 4 (rarely 5) much smaller but some- what similar spines; beginning about the fifth or sixth plate one of the scalehke spinelets near the base of the uppermost of this series becomes enlarged and a little further out a second one also, so that there is a trio of flat spines at the outer aboral corner of the plate; distally the aboral spinelets disappear, the aboral margin near the middle of the arm having 2, then 1 and then none of them; at the tip of the arm there are simply 2 reduced but relatively large spines on the aboral outer corner of plate. Adambulacral armature consisting of about ten or a dozen spines, 2 mm. long (more or less), flattened and widened at the tip to a greater or less degree; one is somewhat curved and set edgewise on the furrow margin ; while a little back of it are 2 others, one on each side; back of them and more widely separated are 2 others with conspicuously widened tips; then follow, on the surface of the plate 70 memoir: museum of comparative zoology 2 or 3 more with very wide tips and behind them, several much smaller and less conspicuous spines. Oral plates covered with about 4 series of thick blunt spines, of which the innermost are longest and largest; the marginal spines are much more slender than those on the median crests. Color in life extraordinarily brilUant for an Astropecten; the lower surface, marginal spines and most of the superomarginal plates are pure white ; 6-9 supero- marginals, in each series, bright rose-red in sharp contrast; these red plates are arranged in three (or only two) well separated groups on each side of each arm, the distalmost consisting of a single plate or sometimes two. Paxillar surface of arms bright brownish-orange, with irregular groups of blackish spots in the inter- radii and 3 or 4 smaller ones scattered along the arms. The dried specimen is a uniform light brown with marginal spines and most superomarginal plates still lighter; in each interradius is a definite group of spots of a distinctly purple shade. Holotype, M. C. Z., no. 3123, from Lagrange Bay, northwestern Australia, in about 5 fms. June 10, 1932. This extraordinary sea-star was brought up by our diver, Mr. Norman AMiitworth, who reported that he found it lying on the bottom beside a sponge. Later, between Lagrange Bay and Broome we dredged a smaller specimen, with R = 46 mm., showing essentially the same coloration. A few days later a third specimen was secured in an extraordinary manner. A Japanese diver whom we knew as "Wan" brought up a fine large specimen of Stellasler princeps, a sea-star of sufficient rarity to make additional specimens very welcome. Turning it over to look at the handsome purple and white oral surface, we were amazed to find an Astropecten of this species half-enclosed in the Stellaster's stomach! It was carefully rescued from its living death but the disk and basal part of the arms were already partially digested and only the distal portion of three rays were uninjured. This individual had R = 55 mm. more or less and the arms are narrow as in the small paratype, but this may be due to the unusual way in which it met its death'. No other specimens were found. Although there can be little doubt that this very unusual Astropecten be- longs to the Monacanthus-group of Doderlein, it is easily distinguished from both granulatus and rnonacanthus by the armature of the inferomarginal plates, par- ticularly the flat, truncate marginal spines, and the aboral series of spines which those plates carry. In life of course, this species can never be confused with any other. ' That it was dead when found by the Stellaster is a possibihty not to be ignored but its appearance when secured makes this seem unlikely. CLARK: AUSTRALIAN ECHINODERMS 71 LUIDIIDAE LUIDIA AUSTRALLAE DoDERLEiN, 1920. "Siboga" Ast.: Luidia, p. 266. While the distinction between this species and the following seems rather trivial yet since it is recognizable, apparently constant and associated with different geographical areas, it may properly be recognized in the name. The material at hand consists of 7 typical specimens and an arm fragment, and 2 rather small specimens which seem to represent a color variety. Typical australiae is deep yellow or buff, more or less blotched and variegated with dull, dark green. Large specimens sometimes have the whole or nearly all of the dorsal surface dark. Mr. Livingstone has written me of a specimen he secured at Lord Howe Island during the Christmas hohdays, 1932, with R = 245 mm. "Abactinal surface wholly black, actinal surface and spines cream-coloured." It "has a trace of very dark green." "It is closer to a black than to a green to my mind but may be it could be called blackish green." This is of course a very large and doubtless old specimen. The one which we took at Lord Howe in April, 1932, is not quite so large and the buff and dark shades are about equal in extent. The specimens from the \'icinity of Fremantle have the dark and hght shades very much mixed but the dark predominates and has a distinct greenish tinge. All of the specimens of australiae at hand have 7 arms but usually one (or more) of them is regenerating. There is considerable diversity in the relative size of disk and arms, the ratio ranging from R = 6r to R = 8r. While dredging near Garden Island, off Fremantle July 22, 1932, we took two small Luidias which were so different in color from any australiae I have seen that I suspected they might represent an undescribed species but comparison with other Luidias fails to show any character except color by which they can be dis- tinguished from small australiae. In life, the ground color was very hght, almost cream-color while the blotches and markings of the upper surface were chocolate- brown, with no hint of green. The dry specimens after three years are little changed save that the brown has faded to some extent. For convenience, they may be designated as "forma brunnea." Each has 7 arms; in one R = 82 mm., in the other R = 110 mm. A single arm of a small Luidia, showing the characteristic large paxillee near the tip which distinguish australiae is in the collection; it was taken at Hamehn 72 memoir: museum of comparative zoology Bay, near Augusta, W. A., January 20, 1930 by Professor E. W. Bennett, and measures 80 mm. long by 12 mm. wide at base. Aside from this fragment and the 2 specimens of "brunnea" just described, the material of australiae at hand is as follows : Lord Howe Island. Southeast end of lagoon beach, buried in sand, its position in- dicated by the radiate depression, Aprils, 1932. 1 large adult. Western Australia : Between Fremantle and Garden Island, 2-3 fms., October 14, 1929. 2 small adults. Rottnest Island, Thompsons Bay. 1 small adult. Loaned by the Western Australian Museum. Rottnest Island, 1934. Captain Beresford E. Bardwell leg. et don. 2 small adults. LUIDIA maculata MiJLLER and Troschel, 1842. Syst. Ast., p. 77. It was a great surprise to discover that the Luidias taken at Broome were obviously different from those taken on the eastern and western coasts of Aus- traUa. Of course, the most obvious difference is that they have S arms, whereas all the other Austrahan specimens available have but 7. It was then noted that they lack the large paxillae near the tips of the arms, so characteristic of aus- traliae, and it thus seemed probable that they should be referred to the Asiatic species and not to the Australian. Comparison with specimens from Hong Kong and Ceylon showed this to be the case. Of the 6 specimens at hand, 1 was taken Sept. 14, 1929 in 7-8 fms. of water, on the south side of Roebuck Bay. Although brought up by a diver and handled with great care it slowly but persistently broke off all Init one of its arms. In life R = 225 mm. and the following color notes were taken: "Cream color with most of disk light gray; arms marked with light bluish gray and dark greenish- gray ; more or less of a line of ochre-yellow paxilla; along middle of ray and also on disk. Each pedicel white with a longitudinal orange-stripe on outer side." After eight years the dry specimen shows but very little change, except of course that the pedicels are completely dried up. The other 5 specimens were taken in 1932, 2 near Broome, the other 3 on June 17, in Beagle Bay, by our Japanese diver, Wan. The largest specimen with R = 230 mm. was wrecked in the dredge but the second Broome specimen is in CLARK: AUSTRALIAN ECHINODERMS 73 perfect condition, showing form and color as well as the day it was taken. Of the Beagle Bay specimens 2 were "light gray irregularly mottled with yellow-brown along midradial hne, and with blackish laterally and terminally; marginal spines and lower surface white. Feet pale yellow with an orange longitudinal stripe." The third specimen was "uniformly deep slate gray above." Six of the arnxs were regenerating and those youthful tips were dark brown. In the dry specimen the contrast between these regenerating portions and the rest of the animal is still very marked. Whether the orange stripe on the pedicels is always present in maculata my notes fail to make certain. Even more regrettable is my failure to note whether it is present or wanting in australiae. LUIDIA HEXACTIS' Sp. UOV. Plate 17, fig. 1 Rays 6. R = at least 140 mm. ; r = 15 mm. ; br = 15 mm. R = rather more than 9r or br. Disk relatively small and arms slender. Paxillse in 5 longitudinal series of close set squares with sides 1.5 mm. long, on each side of ray; between these lateral series are 5 somewhat irregular series of more polygonal and smaller pax- illae; near the tips of the arms, the paxillse are still smaller and less regularly ar- ranged ; the paxillfe are closely covered with very fine granules none of which are at all spine-like; the arms, which are fairly high at base (8 or 9 mm.) but flattened on the upper side, are thus quite smooth. Madreporite just distinguishable close to the interradial margin. Inferomarginal plates very short and wide; there are about 7 in 10 mm. of arm-length while their width is about 5 mm.; the surface is well covered by a longitudinal series of 2-6 spines and a large number of very much smaller spines or spinelets ; the largest spine is more than 2 mm. long, wide at base but rapidly attenuate to an acute tip; it is situated on the outer end of the plate and close to it but below is generally a second spine of almost equal size; the remaining spines, when present, are well spaced, and on the basal half of the arm, one, often the smallest is at the inner end of the plate; the spinelets are very slender and acute and most of them are less than a millimeter in length, often much less. Adambulacral armature simple, with 2 large furrow spines and 2 on surface of plate; outer furrow spine 2-3 mm. long, strongly compressed, straight, tapering to a blunt point; inner spine not so long, curved, markedly acute; the 2 spines on ' ej =six + a.KTts =ray, in obvious reference to the number of rays. 74 memoir: museum of comparative zoology surface of plate are situated side by side, 2 mm. long (or less), tapering to a slender point ; back of them is a large pedicellaria, often accompanied by a much smaller one; occasionally there are 3 pedicellariae on a plate; pedicellariae made up of two long, slender jaws with 1 or 2 or several slender, sharp spinelets at their base. Actinolateral plates only 2 on each side, but each carries a very large pedicellaria. Oral plates with 7-9 long marginal spines, 5-7 long spines on the crest, and a few small spines or pedicellariae on the inner end; usually 2 pedicellariae composed of 4 or more short spinelets on the side of each oral plate, deep in the mouth. Color of oral surface very light, probably white or cream-colored in life; dorsal surface Ught buff blotched and marked with large irregular areas of dull, dark greenish-gray. Apparently the color in life was much the same as, and very similar to that of L. maculata. Holotype, M. C. Z. no. 3134, collected "on a sandbank near Montgomery Reef, more or less in Collier Bay, between Cockatoo and Augustus Islands," northwestern AustraUa, by Captain Beresford E. Bardwell, in October, 1933. There is with it a note by the collector "This was half dead and badly broken when discovered." This is a very well marked species for while it seems to belong to Doderlein's Quinaria-group, it is very unlike the only 6-armed species, penangensis, in its paxillse and dorsal surface, and is not near any of the 5-armed species. It seems to be nearest to forficif era but aside from the difference in the number of arms, the difference in the armature of inferomarginal and adambulacral plates is conclu- sive. The possibility that hexactis is simply a 6-rayed maculata is debarred by the differences in the pedicellariae and in the armature of the inferomarginal plates. ARCHASTERIDAE Archaster typicus MiJLLER and Troschel, 1840. Monatsb. Preuss. Akad. Wiss., p. 104. There are six good specimens at hand of this well-known East Indian sea- star. They were given me by Mr. Melbourne Ward who collected them at Linde- man Island, on the Barrier Reef, near Port Mackay, Queensland. This is by far the most southern station at which the species has been taken. The present speci- mens are of moderate size (R = 60 mm., more or less) and closely resemble indi- viduals of similar size from the Caroline Islands. CLARK: AUSTRALIAN ECHINODERMS 75 Archaster laevis^ sp. nov. Plate 17, fig. 2 R = 105 mm. ; r = 13 mm. ; br = 14 mm. ; hence R = about 8r or br. In life and in the better-preserved specimens, the disk and arms are high (in the holotype, 10 mm.) and the aboral surface is perfectly flat except beyond the middle of the arms where it gradually decUnes to the terminal plate. Paxillae very close set, the tops sharply hexagonal, excepting some large ones on disk, particularly near madreporite, and a series of small circular ones adjoining the superomarginal plates on each side of each arm; largest paxillae of disk each have 40 or more, angular, truncate, but often sUghtly convex, granules on the surface and about 30 similar ones around the margin; there is no essential difference between the marginal and the surface granules; near the disk margin and on the arms, the paxillae are much smaller ; on each arm is a median series which is conspicuous the paxillae-tops are so short and wide, less than a milhmeter long and fully 2 mm. wide; there are about 25 granules on such a top; on each side of the median series are five longitudinal series of hexagonal paxillae and a sixth less regular series of smaller, nearly circular ones bearing only 6-10 granules; the hexagonal paxillae carry about 10-15 (or more) granules of which only 1-6 are central, the others marginal. No pedicellariae have been detected. Superomarginal plates about 53 in each series, placed vertically along the sides of the rays except near the tip where they encroach more and more on the upper surface, the distalmost pair sometimes coming in contact in the midline at the base of the moderately large terminal plate; at the base of the arms, the superomarginal plates are 6 mm. high and 2 mm. wide (i.e. long.) ; there are no spines on the superomarginal plates but they and the terminal plates are closely covered with coarse granules Uke those borne on the paxillae of the distal part of the arm. Madreporite large, 4 mm. in diameter, and 5 mm. from the disk margin. Inferomarginal plates of about the same number as the superomarginals, but their width (corresponding to height of superomarginals) is considerably less, those at base of arm being about 5 nmi. wide; plates completely covered with wide, flat spinelets, having rounded tips; on outer margin of plate are 2 short (less than 2 mm.), fiat, wide, round-tipped spines, which project very httle from the side of ray but are somewhat appressed to it (on some plates, particularly at tip of arm, only one spine is present) ; below these 2 inferomarginal spines are 1 faem= smooth, in reference to the close-set and even paxillse, which give a notably smooth dorsal surface. 76 memoir: museum of comparative zoology several (1-4) somewhat similar but smaller spines which intergrade with the spinelets covering the plate. Adambulacral plates with an armature of 6-8 (usually 7) spines, a rather slender but blunt furrow spine, 2 similar but somewhat stouter marginal spines and 2 pairs of surface spines ; tlie inner pair is blunt but not much flattened, the outer is larger and quite flat and the adoral is distinctly the largest spine on the plate. Oral plates conspicuous with about 9 moderately stout blunt spines on the furrow margin, about 4 on the posterior margin and a double series on the crest with 8-10 similar but shorter spines. Color in hfe: "Oral surface, inferomarginal spines and madreporite cream white ; upper surface chiefly yellow-brown variegated with light fawn-gray which becomes almost cream- white near tips of rays; an indistinct dull blue-gray line 1.5 mm. wide along the middle of each arm, extending onto disk but not clear to center." The dried specimeas have lost their distinctive and handsome colora- tion; and are now a dull hght yellow brown very indefinitely variegated with darker; in some cases the darker shade predominates; in a particularly well- preserved dry specimen the light yellowish brown shade includes most of the sea- star but the margins of the rays near the l)ase and scattered groups of supero- marginal plates are a dull reddish with a hint of purple; these reddish areas are ill-defined and merge into the yellow-brown ground color. Specimens dredged near Fremantle showed essentially the same coloration as those from Lagrange Bay but the tints were slightly different; the fawn-gray was almost a dull light purple. Holotype, M. C. Z. no. 3135, from the vicinity of Broome, W. A., 5-8 fms. That this handsome Archaster is nearly related to A . angulatus admits of no doubt, but the smooth tesselated aboral surface caused by the crowded, truncate prismatic granules of the paxilla^ gives it a very characteristic appearance quite unlike that of any specimens of angulatus available for comparison. It is possible however that carefully preserved specimens of angulatus would more nearly re- semble laevis and that the latter is only an extreme form of the older species. Whether there are color differences of importance remains to be determined as no information as to the color of angulatus is available. This fine sea-star is not uncommon in shallow water, on the sandy coast of northwestern Australia but we did not find it near low tide mark or in tide-pools. It was a surprise to meet with it again near Fremantle, while dredging between that port and Rottnest Island, but no differences worthy of note have been detected between the specimens taken in these widely separated areas. CLARK: AUSTRALIAN ECHINODERMS 77 The form of dry specimens, especially large ones, is apt to undergo con- siderable change in the process of preservation; the paxillar field of the rays be- comes much narrower and sinLs below the level of the superomarginals especially in the midline, and the upper ends of the superomarginals themselves are drawn inward by this shrinking so that those plates no longer form a vertical wall at sides of the rays but he rather considerably on their aboral surface. There is httle diversity of form among U^^ng specimens but young ones of course have the rays relatively shorter; thus the smallest specimen at hand, with R = 23 mm. has r and br about 5 mm. so that R is only equal to 4.5 r or br. In adults there is some difference in the slenderness of the arms; thus the largest indi\adual, with R= 122 mm. has br= 18-20 mm. so that R is little more than 6 br, but the great majority have R in excess of 7 br. The 26 specimens at hand are from the following places : Western AustraUa: Vicinity of Lagrange Bay, 5-7, fms., on bottom of sand and shells, September 11, 1929. 2 specimens. Vicinity of Roebuck Bay and southwestward 5-8 fms. June, 1932. 20 specimens, adult and young. Between Fremantle and Rottnest Island, 8-10 fms. October 19, 1929. 4 specimens. GONIASTERIDAE Nectria multispina H. L. Clark, 1928. Rec. S. Austr. Mus., 3, p. 375. It was a real disappointment not to see Nectria alive or even freshly killed, but all of the sea-stars of that genus in the present collection were given to me by Professor E. W. Bennett. One of them is an indubitable specimen of this South Austrahan species, with R fully 75 mm. It is in fairly good condition but is some- what waterworn. It is "one of many" found in "drift on the sand beach at Middleton Beach," near Albany, W. A., "about Christmas, 1929." The color when collected was "(1) orange (2) red"; presumably this means that the lower surface was orange and the upper red; there is no trace of these colors now. It is interesting to learn that the range of this well-marked species extends so far west as Albany. 78 memoik: museum of comparative zoology Nectria ocellata Perkier, 1876. Arch. Zool. Exp., 5, p. 4. Professor Bennett has also given me 2 specimens of this Nectria, which were taken by Mr. D. L. Serventy, "in a single haul of otter-trawl (trawler "Bonthorpe") in 90 fms., 33° 15' S x 126° 22' 15" E, on February 23, 1930." This locality is in the western end of the Great Bight and is the fui'thest west I beheve that this species has been taken. In life these two specimens are recorded by Mr. Serventy as "chestnut-orange" but they are now so completely bleached as to be a dirty white. One has R = 70 mm., r = 22 mm. and br = 24 mm., so R is about 3 r or br. In the smaller specimen, R = 54 mm., r = 20 and br= 17, so R = 3br but not quite 3r. TOSIA AURATA Gray, 1847. Proc. Zool. Soc. London p. SO. The only s])ecimens of this characteristic AustraUan sea-star in the present collection are 2 dredged in Ralph Bay, Hobart Harbor, Tasmania, in 2-3 fms., November 15, 1929. In Ufe the coloration was much hke that of the more com- mon australis; variegated with fawn color and brown above, grayish fawn and brown below; they are now "museum color." The larger has R = 49 mm., the smaller, 43 mm.; R = 1.5r in both. ToSIA AUSTRALIS Gray, 1840. Ann. Mag. Nat. Hist. (1), 6, p. 281. The only specimen at hand of this conunon sea-star which is a typical aus- tralis was dredged on the west side of Hobart Harbor, Tasmania, in 2-3 fms., Nov. 15, 1929. It is small with R only 19 mm. and r about 13. In hfe the color was deep brown above, variegated with fawn color, and cream-color below vari- egated with brown; these shades are all quite lost now. All of the remaining Tosias are from Western Australia, 6 from the vicinity of Fremantle and 3 from Bunbury and all represent the form astrologorum; those from Bunbury are somewhat nearer typical australis. These were dredged in 5-8 fms., in Koombana Bay, October 26, 1929; they were "variegated light and dark brown," and are all rather small. The largest individual secured anywhere was CLARK: AUSTRALIAN ECHINODERMS 79 found under a stone at Point Peron and was quite different in color from the others ; it was variegated with dark and Ught graj^, the Ught shade with a bluish tint; it has R = 31 mm., r = 20 mm.; 9 of the distalmost superomarginals are very large but the tenth is replaced by 3 small plates. The 5 remaining specimens were found on the wharf-piles of an old jetty at Garden Island. It was a great surprise to me to find so rigid and inert a sea-star active enough to climb a wharf pile and able to chng closely to its surface. It was a severe blow to the idea conceived a priori that Tosias must Uve lying flat on the bottom! In the National Museum at Melbourne there is a remarkable Tosia australis of the typical sort which is perfectly tetramerous with only 24 superomarginal plates. Pentagonaster dxtbeni Gray, 1847. Proc. Zool. Soc. London, p. 79. Among the rocks at Point Peron, W. A., in October, 1929, four specimens of a Pentagonaster were taken wliich may properly be referred to this species, since Gray gives Western Australia as the type locaUty for diibeni. It is interesting to note however that all of these specimens from Point Peron have the arm-tips wide and rounded and the terminal superomarginal on each side enlarged and swollen in varying degrees, whereas Gray's figure (1866, pi. 3, fig. 2) shows httle indication of these features. On the other hand a specimen from South AustraUa in the M. C. Z. is almost exactly Uke Gray's figure while specimens from Port Jackson are intermediate some approaching the Point Peron specimeas rather closely. It is evident that in this species as in the genotype, P. pulchellus from New Zealand, there is a tendency towards the development of two extremes — one having relatively unenlarged superomarginals distally, while in the other the distal superomarginals are greatly enlarged and swollen. As no correlation has yet been discovered between these extremes and their habitat or distribution varietal names are hardly worthwhile, although they would be as valid as the form "astrologoruni" of Tosia australis, which represents a similar tendency to enlargement, or at least swelHng, of superomarginal plates. The specimens of diibeni from Point Peron range in size from R = 26 mm. to R = 38 mm.; in all r is about equal to one half R. The largest has but 32 supero- marginals, as there are 4 superomarginals on only 2 rays and on those rays only on one side. In a second specimen there are 44 superomarginals; on only one side of one ray is the number as small as 3; on the other side of that ray are 6 supero- 80 memoie: museum of comparative zoology marginals but three are abnormally small and 2 are irregularly placed; on another ray, a large superomarginal is crowded out of the left-hand series onto the upper surface of the arm and separates the terminal plate from the aboral plates of the ray ; this specimen shows less conspicuous abnormalities on the oral surface. The two remaining specimens are smaller and both have the normal number of 40 superomarginals. This is a very pretty sea-star in life as the upper surface is bright colored and the separate aboral plates are outlined in fawn-color, cream-color or white. Two of the Port Peron specimens were yellow-orange, one was bright vermilion red and one, the largest, was a deeper red. They were found under rocks, or in plain sight, on the sides of rocks. GONIODISCASTER AUSTRALIAE Plate 8, fig. 2 ToRTONESE, 1935. Boll. Mus. Univ. Torino, 45, Series 3, no. GO, p. 3. This is one of the most beautiful of the many fine sea-stars of northwestern Australia and is not at all uncommon in the vicinity of Broome, but it was not found in less than 5 fms. of water. Owing to its striking coloration it is often brought up by the divers of the pearling fleet. Of the 16 at hand, the smallest has R = 33 mm., r = 10 and br=10, therefore R = 3.3r or br; the largest has R = 92 mm., r = 35 and br = 34, so that R = 2.6r or br. But the rays are sometimes much more slender; in a specimen with R = 67 mm., R = 3.3r and 3.75br. Obviously, there is more or less diversity in form but this is not so striking as the difference in color shown by individuals quite similar in size and form. Most commonly, the characteristic color is a deep rose-pink, particularly on the marginal plates ; usually in adult specimens the abactinal surface develops more or less green pig- ment; in some individuals however this pigment is gray and not at all green and may completely mask the pink. A specimen taken June 14, 1932, north of Broome is described in my field notes thus: "a lovely gray with black tips to arms and an indefinite blackish area at center of disk ; hints of green in interradial areas. No pink. Lower surface hght cream. Dried pink!" The colors disappear rapidly after drying, the green and gray being particularly evanescent; the pink fingers moi'e or less for weeks and months; some specimens still show, after five years, dis- tinct indications of it on the marginal plates and especially at the arm tips. While this species is undoubtedly near to scaber (Mobius), I think Tortonese CLARK: AUSTRALIAN ECHINODERMS 81 is correct in considering it distinct. Even those specimens with the shortest, widest arms have them longer and narrower than in scaber and the supero- marginals are definitelj- more numerous. There is considerable individual diversity in the numi^er of pedicellariae present. Curiously enough Tortonese makes no reference to the following species {coppingeri) which is undoubtedly the nearest relative of australiae. Probably he considered Bell's species identical with forficulatus Perrier, as does Doderlein (1935), an opinion with which I cannot agree, as shown below. It is unfortunate that Doderlein (1935) introduces a new name, australis, for a Goniodiscaster, allied to porosus Koehler, which he calls a "nov. forma" and diagnoses in 5 words! It is very different from australiae Tortonese. The latter name was pubhshed in July, Dciderlein's in October, but as they are differently spelled both can be used if it is necessary. Goniodiscaster coppingeri Pentagonasier coppingeri Bell, 1884. "Alert" Rep., p. 128. Goniodiscaster coppingeri H. L. Clark, 1909. Bull. M. C. Z., 52, p. 110. Mr. Melbourne Ward dredged near Lindeman Island, Great Barrier Reef, July-September, 1934, 2 large individuals of this apparently rare sea-star, which he has kindly sent to me. In them R = 80-85 imn., r = 26-27 mm.; hence R = 3r; br is not equal to r. These specimens resemble so closely Bell's description, save in one respect, that I have no doubt of their identity, especially since one of his specimens came from Port Curtis. They are exactly like the specimen from Mt. Adolphus Island, Torres Strait, figured in my Torres Strait Report (1921). The one particular in which these specimens differ from Bell's description is in the minor radius. Bell says that in his larger specimen r = 46 mm. and br = 24 mm. If that were correct, R = 2r, which would make the form of his specimens quite out of keeping with his description. Moreover he says that in the smaller specimen R = 82, r = 26 mm. Now one of the specimens at hand is exactly this size and br also is 26 mm. Thus r and br are equal which is approximately true in very many sea-stars. There seems no doubt that the figures 46 and 24 are both typographical errors or shps of the pen. Probably each should be somewhere between 28 and 32, for the long narrow arms, R 3-3. 5r, is a striking feature of the species.^ 1 Since the above was written Mr. Dilwyn John of the British Museum has very kindly remeasured Bell's specimens. He confirms the measurements of the smaller but states that in the larger, r 33 mm. not 46. He also enclosed a sketch showing where on the arm "br = 24 mm.," which indicates that according to my way of measuring br = 30 mm. or more, as would be expected. 82 memoir: museum of comparative zoology Doderlein (1935) inclines to believe that coppingeri is a synonym of for- ficulatus Perrier. To this I cannot agree. Koehler (1910) has given figures and a detailed description of Perrier's species and one of the largest of Koehler's specimens is at hand (M. C. Z., No. 2094). Fisher (1919, p. 325) has discussed some of the details of this specimen and doubts if it is a typical /or/lcM^a/ws, but this does not matter for the present discassion, since its pecuUarities have nothing to do with the main differences between Perrier's and Bell's species. These dif- ferences are in the shape of the rays and in the number of superomarginals. Bell's species is in these particulars very near the preceding (australiae) and I have compared specimens of that species, having R = 33 mm. and 48 mm. respec- tively, with the forficulatus mentioned above which has R = 40 mm. (No. 2094) . The differences are obvious. The AustraUan specimens have a much smaller disk and longer, narrower arms, with narrower, more numeroas superomarginals. Thus in the forficulatus with R = 40 mm., the breadth of the arm at base is 15 mm. (R = 2.66br), and there are 12 superomarginals on each side of each ray with a thirteenth indicated. In the Australian specimens, R 3-3. 5br, and there are 15 or 16 superomarginals in the smaller specimen, 18 or 19 in the larger. The granulation of all the plates in the two Australian species, regardless of size, is distinctly less coarse than in forficulatus and the armature of the adambulacral plates is made up of shorter and less heavy spines, particularly in the surface series. There seems no doubt that forficulatus approaches the group of Gonio- discasters with, shorter wider arms, such as granuliferus (Gray) and porosus (Koehler), while the Australian species belong rather to a group with arms approaching in form to those of Paragonaster. The type of the genus, pleyadella (Lamarck), is intermediate between the two groups, each of which might be treated as a genus if the number of species involved were large enough to make such a division convenient. As for the differences between coppingeri and australiae, they are obvious enough when specimens are side by side. The west coast species has a much rougher dorsal surface with many coarse granules and small pointed tubercles on the aboral plates and even on the superomarginal plates. These are not very conspicuous in Tortonese's published figure but are well shown in liis original photograph, of which he sent a copy to Fisher who has kindly given it to me. In most well preserved specimens, the rough, almost "prickly" upper surface contrasts markedly with the almost smoothly granular surface of coppingeri; in one of the specimens of the latter from Lindeman Island, a few low tubercles CLARK: AUSTRALIAN ECHINODERMS 83 are present on many superomarginals but they are not nearly so numerous or noticeable as in australiae. Orally the two species are so much alike it is doubtful if a coastant difference can be found. The specimens of coppingeri at hand indicate that the color in life was darker than that of australiae, an almost brick red. Information on this point is greatly to be desired. GONIODISCASTER FORAMINATUS Goniodiscus foraminatus Doderlein, 1916. Zool. Jahrb. : Syst., 40, p. 41.5. Goniodiscaster foraminatus Doderlein, 1935. "Siboga" Ast. : Oreasteridae, p. 79. An interesting young Goniodiscaster loaned by the museum at Perth (No. 9018) may well be referred for the present to this species, the type-locality for which is Shark Bay. In this specimen R = 35 mm. while r and br are approxi- mately haK as much. The present "museum-color" is very Ught, cream-color orally and pale yellowish-brown above. The resemblance to Doderlein's figm'es is very close. In only one particular does this young individual fail to quaUfy for the Shark Bay species; the pedicellariae are not at all numerous or conspicuous. Since this may be either a youthful character or an individual pecuUarity, we are justified in considering tliis specimen a young foraminatus. It was collected by Mr. K. Barker at Garden Island, off Fremantle, in 1915. Miiller and Troschel's species G. seriatus, with which Doderlein considers Gray's Pentaceros granulosus synonymous, comes from southwestern AustraUa but the Garden Island specimen is quite certainly not a young example of that species. The form and proportions and the granulation of the aboral plates are very different from those characters in seriatus and these differences can hardly be associated with the youthfulness of the specimen. Goniodiscaster integer Livingstone, 1931. Rec. Austr. Mus., 18, p. 135. Five little sea-stars taken at Lindeman Island, Barrier Reef, in 1934 and sent to me by Mr. Melbourne Ward, apparently represent this species but I am not satisfied as to whether it is really distinct from pleyadella. However until the contrary is shown, the name integer may conveniently be used for the form occur- ring on the eastern coast of Queensland, since adults are easily distinguished. A typical adult is readily told from an adult pleyadella by the tapering rays, 84 memoie: museum of comparative zoology smaller, less sunken papular areas, and more irregular, coarse granulation. But young individuals of pleyadella to a size of R = 32 mm. bear little resemblance to the adults, as Doderlein's (1896) figures show, and apparently the two species, if species they be, cannot be certainly distinguished from each other until R = 40-50 mm. The present specimens have R = 6, 10, 15, 26 and 31 mm. respectively. The 3 largest show the 5 primary radial tubercles plainly but they are most conspicu- ous in the individual with R = 26 mm. This specimen is also much the lightest colored, a brownish cream-color; the others are Ught brown. There is nothing of importance to add to Livingstone's careful description. GONIODISCASTER ACANTHODES' Sp. nOV. Plate 5, fig. 2 Rays 5. R = 73 mm., r = 37 mm., br = 35 mm. ; R = 2r or 2br. Aboral surface sUghtly arched, covered with numerous plates of moderate or small size, hidden by a coat of granules and pedicellariae ; most of the plates also bear one or more rounded tubercles of diverse sizes, the largest more than a miUimeter in diameter and fully as high; on each ray there are 3 parallel series of these tuberculated plates, the median being shut off from the terminal plate by the meeting of supero- marginals, while the lateral series drop out somewhat sooner. Pedicellariae fairly numerous, square-jawed, longer than wide or high, very unevenly dis- tributed among the granules. Granules of very unequal size, more or less rounded. Papulae few and small on the disk becoming much more numerous and larger on the rays where, in groups of 4-7, they form 3 (or 5) (5 or 7 basally) ill-defined series, not sharply distinguished from each other. Superomarginal plates 11 (or 12) on each side of each ray; the interradial pair lie nearly flat on aboral surface of disk and measure about 10 mm. wide by 4 mm. long but the inner end is clearly longer than the outer. Distally the plates are longer and not so wide, and he more on the side of the ray; the tenth is only 7 mm. wide while its outer side is 5 mm. long and 3 mm. high. The plates are covered with a very close smooth coat of granules, smallest on the lateral margins and coarsest at the inner end; in addition, the plates carry 1-12 rounded tubercles of very diverse sizes, hke those on the aboral plates, irregularly distributed but chiefly near the inner end. Terminal plate relatively small, covered by granules like those of the superomarginals. ' i.Kav6udris =Jull oj tliortts, in reference to the numerous tubercles on the aboral surface. CLARK : AUSTRALIAN ECHINODERMS 85 Oral surface flat, the actinal intermediate areas rather large, covered by a closely matched pavement of 50 or more polygonal plates, 2-5 mm. in diameter; the granules and pedicellariae covering these plates are much longer and less crowded than those of the upper surface ; the innermost series of actinal plates extends out to the seventh inferomarginal but the adjoining series only reaches the fifth. Inferomarginal plates correspond in number and position (not exactly) with the superomarginals, but there is an additional very small inferomarginal (or usually 2 of them) just below the terminal plate where the tip of the ray turns up; there is no superomarginal to correspond with these; granular coat of infero- marginals close and even but coarser than that of the upper series — not so coarse however as that of the actinal intermediate plates; a few small tubercles may be found on one or more of the distalmost inferomarginals. Adambulacral armature in three distinct series; innermost of 5 (4-6) strongly compressed spines (2 mm. long, more or less) with rounded tips, the middle ones longest, the outermost very small; the second series consists of 2 (or distally only 1) very broad, stout spines (.75-1.50 mm. wide) with truncate (or rounded) tips; the outermost series consists of 2 or 3 short, thick spinelets, not so high or so big as those of the second series; this outermost series tends to be irregular in its distribution and may be lacking on some plates. Pedicellariae, with jaws much longer than wide, occur on many adambulacral plates, often being placed on the adoral margin between the first and second series of spinelets. Oral plates with a heavy armature; about 10 very stout prismatic or squarish spines with rounded tips crowd each margin; back of them is a second series of 6, even stouter, but shorter and less prismatic spines; on the inner crest of each plate is a series of 4 or 5 much smaller, very short spines. Color in Ufe very striking; marginal plates, tubercles and coarsest granules bright brick-red, the rest of the aboral surface bluish-gray; lower surface yellow- ish. All the colors disappear rapidly in preserved material. The dry holotype is a dull reddish-brown above, a Hghter and less reddish color below; the rays are Ughter than the disk aborally. Holotype, M. C. Z. no. 3154, dredged southwest of Broome, in 7-8 fras., June, 1932. The selection of a holotype has been a difficult matter owing to the great diversity in the development of tubercles and spinelets on the aboral surface. Orally there is individual difference in details, particularly in the size, propor- tions and number of the spines in the outermost adambulacral series, but no essential modifications have been noted. The differences aborally may be 86 memoir: museum of comparative zoology grouped under four heads. 1. Tubercles on superomarginal plates: these, so evi- dent in the holotype, are lacking in most specimens, the superomarginals being as smoothly coated as can be; only 3 specimens altogether have these tubercles. 2. Tubercles on aboral plates: the diversity in these is surprising and does not seem to be correlated in any way with size; they are largest and most spherical in the holotype ; in a somewhat smaller specimen they are nearlj^ as large but are conical and more numerous and stand out with great distinctness from the gran- ules which cover the plates; in a third specimen, they are somewhat conical but on most plates there is only one or one is conspicuously bigger than the others; in a fourth specimen, they are small but numerous, many plates having 10-15 nearly spherical tubercles in a not very compact group; other specimens have very few tubercles and these confined to the plates of the disk and the carinal plates at the base of the rays; the smallest specimen (R = 37 mm.) has very few tubercles and a specimen with R = 42 mm. has none at all on either aboral plates or superomarginals. 3. In about half the specimens, a coaspicuous tubercle notice- ably larger than its fellows Ls found on the large carinal plate at the base of each ray; in no specimen are these symmetrically developed; usually they are evident only at the base of three or four rays; in one case, a specimen with R = 47 mm., all 5 are present but one is accompanied by a second tubercle nearly as large; when most conspicuous these tubercles may be 2 nun. high and more than 2 mm. in diameter at base ; ordinarily they are bluntly conical but they may be rounded, or terminate in 2 or 3 sUghtly indicated blunt points. 4. While the holotype and many other individuals have the rays wide and rounded at tip, in half a dozen individuals the rays are distinctly pointed, and other specimens intergrade be- tween the extremes. In spite of this interesting diversity, I have no doubt as to the specimens representing a single species. All were collected Isj' me and showed the striking and characteristic coloration in life which distinguishes this handsome sea-star. Owing to the difference in the matter of tubercles on the superomarginal plates, some specimens of acanihodes fall into one section of Doderlein's recently pub- lished (1935, p. 77) key to the species of Goniodiscaster while others fit into the alternative section. The holotype runs down at once to G. scaher Mobius but comparison with the figures given shows it is not that species, the number of superomarginals and their granular coating being obviously different. Individuals of the present species, lacking superomarginal tubercles run down to seriatus M. & T. or (jranuliferus Gray, but again comparison with the figures given by Doderlein shows that the Australian species is quite distinct. The remarkable CLARK: AUSTRALIAN ECHINODERMS 87 constancy in number of superomarginals in acanthodes is worthy of special mention; the smallest specimens have 11 or 12, the largest has 11 normally de- veloped and a twelfth just forming next to the terminal plate, and other large specimens have 12, with exceptionally a thirteenth just forming. All of the 24 specimens at hand were dredged on a firm, sandy bottom in 7-8 fms. of water in the vicinity of Broome, in June, 1932. Oddly enough we did not meet with the species in 1929. GONIODISCASTER BICOLOR' Sp. nOV. Plate 5, fig. 1 Rays 5. R = 80 mm., r = 40 mm., br = 40 mm. R = 2rorbr. In all details of structure like those specimens of the preceding species (acanthodes) which lack tubercles on the superomarginal plates; no details of tuberculation, spinulation, granulation, form, size or proportion, have been discovered by which the two forms can be separated. The granulation in bicolor is a little finer, especially on the marginal plates, when specimens of like size are compared side by side, but the difference is so slight it cannot be expressed in words or figures so as to help m determining to which species an unidentified specimen belongs. No specimens of bicolor however have any conspicuous tubercles on the superomarginal plates. The coloration of the present form is so strikingly different, and in each species the coloration shows so little diversity, that it never even occurred to me when collecting them that the two were very closely alUed. Not until the dried material was unpacked in Cambridge did I realize how very much alike the two forms are. The present species (bicolor) has the disk and bases of the rays uni- formly gray with no hint of blue in it; the remainder of the ray is a very fine rose-red, quite different from the red of the marginal plates in acanthodes. Madre- porite pinkish and a few dots near center of disk deep rose-red. This lovely sea-star was much less common than acanthodes and only 7 specimens were taken. The first were found in Lagrange Bay but others were taken nearer to Broome. Unfortunately my field notes do not make clear whether the two species were ever taken together. The smallest specimen of bicolor taken has R = 58 mm. The largest is the holotype (M. C. Z. No. 3156) whose measure- ments are given above. ' bicolor = having two colors, in reference to the striking coloration. 88 memoir: museum of comparative zoology The objections to basing species of sea-star on color are so obvious as to re- quire no amplifying, but the present case is so extraordinary I have been at a loss as to what the best course might be. To designate one as a color phase of the other involves the difficulty of determining which is the original "species" and which is the "form"! That one is the male and the other the female of a single species is conceivable, and deserves investigation when fresh material is available, but no such remarkable case of sexual dimorphism in color is known at present among echinoderms, and seems highly improbable in this case. It seems best therefore to give each form a specific name, in spite of the difficulty of distinguish- ing museum material. As a matter of fact, the dry specimens of bicolor cannot all be distinguished beyond question from the corresponding material of acanthodes but as a rule the interradial superomarginals are slightly lighter colored than those on the arms while this is not the case in acanthodes. STYPHLASTER ' gen. nov. Allied to Goniodiscaster but readily distinguished by the coarse, polygonal, flat-topped granules of diverse sizes which cover the papular areas of the aboral surface, contrasting markedly with the large spherical granules which cover the convex skeletal plates. Marginal plates notably convex and covered with very coarse, spherical granules; the distal inferomarginals extend out laterally beyond the superomarginals so that they are conspicuously visible from above and form the margin of the distal half of each ray. Adambulacral armature in 4 crowded series of very heavy spines, the furrow series stout and similar to the others but not so large; outermost series, small, prismatic and truncate. Pedicellariae ap- parently wanting. Genotype, Styphlaster notabilis sp. nov. This is a well-marked genus, for while its relationship to Goniodiscaster is obvious enough, the general facies is quite unlike any member of that genus. The whole abactinal surface is well-arched, considerably above the marginal plates, but the basal part of each ray is conspicuously arched above the inter- radial areas as well as above the somewhat flattened central pentagon of the disk itself. ' aTv4>\6s =rough, rugged + acrrrip =a star, in reference to the unusually rough and rugged appearance of the aboral surface. CLARK: AUSTRALIAN ECHINODERMS 89 Styphlaster notabilis' sp. nov. Plate 17, figs. 3-4 Rays 5, short and stout, high at base. R =57 mm., r =28 mm., br = 26 mm. R is thus about twice r or br. At tip, each ray is wide and rounded; 10 mm. from tip, the width is 18 mm., the height, 8 mm. ; nearer tip the height is a full 9 mm., while at base, say 20 mm. from disk-center, it is 15 or 16 mm. At center, the disk is 13 or 14 mm. thick. Aboral plates rhombic, polygonal or rounded, rather sharply defined, and markedly convex; the center of each plate is occupied by a cluster of large spherical granules, the marginal ones smallest; one of the central ones is often conspicuously larger than the others and may become a considerable tubercle oftentimes higher than thick; on several of the carinal plates, in each series, these central tubercles become stout erect spines, but they are erratic in distribution and very unequal in size; the largest are about 2 mm. high, 1-1.5 mm. in diameter and very blunt. Margins of plates, and papular areas, covered with flat-topped polygonal granules of diverse size and form. Madreporite distinctly pentagonal, wider than high; width slightly more than 5 mm., lieight slightly less; distal angle 11 mm. from superomarginal plates. No pedicellariae have been detected. Superomarginals only 8 on each side of each ray, rather conspicuously con- vex and covered along the margins with flat-topped polygonal granules, and elsewhere with large spherical granules, particularly large on the upper end of the interradial plates, where they are a millimeter or more in diameter; at the tip of the ray the superomarginals he largely on the upper surface of the ray, the last pair, which are about as wide as long, being broadly in contact, and separat- ing the small terminal plate from the aboral surface of ray. Actinal intermediate plates about 42, in three quite distinct series; back of the oral plates is a single tetragonal or pentagonal plate and from this there extends, adjoining the adambulacrals, a series of 13 rounded pentagonal or roughly circular plates, progressively smaller distally, as far as the seventh inferomarginal; a second series of 5 or similar plates extends to the third in- feromarginal, and a third series of two or three plates (4 or 5 plates altogether) fill the remaining space; all of these actinal plates are somewhat convex and covered with granules similar to those on the inner end of the inferomarginals; no pedicellariae have been detected. ' 7iota6!7i;s= notable, of obvious significance. 90 memoir: museum of comparative zoology Inferomarginals 11 on each side of each ray; these plates are much more flattened than the superomarginals and nos. 4-7 exceed in size those of the upper series; all, except the two distalmost, which are very small, are wider than long; the first 5 correspond in position with the superomarginals ; the sixth is less distal, the seventh and eighth underlie the seventh of the upper series, while the ninth, tenth and eleventh, underlie the big eighth superomarginal; beginning with the fourth the inferomarginals are visible from above and the seventh and eighth are particularly conspicuous, extending out at least 2 mm. beyond the upper series. (This is not due to flattening in the drying of the specimen, which has undergone practically no change since it was collected except the slight fading and duUing in coloration.) Adambulacral armature very stout and crowded; each plate carries four series of spines; marginal (or furrow) series consists of 4 subequal spines, 2 mm. long, less than a millimeter thick, rounded at tip, the middle pair nearly cylin- drical, the others distinctly flattened; second series, 3 very stout, rounded, prismatic, spines with rounded tips, the middle spine slightly the largest; third series, 3 smaller, more prismatic and more truncate spines, the middle one much the largest; outermost series, 3 or 4 (rarely 5) small prismatic, truncate spinelets irregularly arranged to cover the rest of the plate. Armature of oral plates, a direct continuation of the adambulacral armature; hence each pair of plates is densely crowded with about 35 stout spines, corresponding perfectly in arrange- ment with those of the adambulacral plates. Color of dry specimen, a light reddish-brown with a lavender cast; areas between the marginal plates and between those of the aboral surface very light, almost cream-color, except on the swollen basal portion of each ray; madreporite brownish-yellow; orally the reddish-brown of the marginal plates is evident but the interradial areas and the oral and adambulacral armature is more yellowish- brown. Aly field notes describe the color in life as follows: "Coloration indis- tinctive. Lavender-brown; disk and inferomarginals, which show clearly on distal part of arm, are more brown; superomarginals more lavender. Madreporic plate, lavender-white. Sutures between all abactinal plates, whitish. Lower surface, flesh-color to reddish, the interradial inferomarginals being quite flesh- red, though dull. Adambulacral armature, creamy-white." Holotype, M. C. Z. no. 3158, taken near entrance to Roebuck Bay in 5-7 fms., June 27, 1932. This notable sea-star was obviously a novelty and we dredged repeatedly at and near the place where we found it in the hope of secur- ing more specimens, but in vain — the holotype is still unique. CLARK: AUSTRALIAN ECHINODERMS 91 ICONASTER LONGIMANUS Astrogonium longimanum MoBius, 1859. Neiie Seesterne, p. 7. Iconaster longimanvs Sladen, 1889. "Challenger" Ast.. p. 261. The occurrence of this remarkable sea-star on the northwestern coast of Australia is of great interest. We did not meet with it at any point but the West- ern Australian Museum has a fine specimen (No. 2982) taken at Broome. In it R = 61 mm. and r=18 mm. It agrees in all essentials with 2 larger specimens from "Adolphus Island near Cape York" received many years ago at the M. C. Z. from Ward's Natural Science Establishment, Rochester, N. Y. These were col- lected by Mr. Henry A. Ward during his visit to Torres Strait and the Great Barrier Reef in 1896. In the larger specimen, R= 110 mm., r =28 mm. A photograph of a living specimen, taken by Mr. Archie Embury at Nor- west Island, Capricorn Group, Queensland, Ues before me, accompanied by the followmg notes on the color in life, which deserve pubhcation: "Arms and marginal plates Ught orange-brown. Interradial areas reddish-brown. On each side of the median series of radial plates, a cream-white line, not reaching to center of disk. Lower surface, buffy-yellow." Mr. Melbourne Ward has kindly sent me 2 small specimens of Iconaster from Lindeman Island, Barrier Reef, which he collected in 1934. They are dry and Ught gray-brown in color. One is a symmetrical individual with the 5 rays, 35 mm. long, while the other is a smaller but perfectly tetramerous specimen, not quite symmetrical howe^•er; one ray is only 25 mm. long and has 12 superomar- ginals on each side; opposite it, the ray is 27 mm. long and has 13 superomarginals on each side ; if the short ray is considered anterior, the ray at its right is also 27 mm. long but has 12 superomarginals on its anterior side and 13 on the pos- terior; the fourth ray is 28 mm. long with 13 superomarginals on each side. Anthenoides dubius^ sp. nov. Plate 17, figs. 5-6 Rays 5, slender and flat. R = 18 mm., r = 6 mm., br = 5 mm.; hence R = 3r and 3.6 br. Aboral surface covered by a thin, rather sparsely granulated skin; granules apparently beneath the skin as in Stellaster; aboral plates obscured except on rays where a single median series of squarish or oblong plates extends 1 dubius = doubtful, in reference to the uncertainty as to whether this probably immature form belongs in this genus. 92 memoir: museum of comparative zoology nearly to tip; the last 3 superomarginals are in contact and separate these radials from the terminal plate; beyond the fifth superomarginals the radial plates are very minute. Single small papulae occur at each corner of the radial plates and of the first 3 superomarginals. A few low, oblong pedicellariae quite small but relatively of large size are scattered on the distal portions of the disk. On each basal radial (or the first carinal) plate is a relatively large bluntly rounded tubercle, higher than thick. Madreporite relatively large, circular, a little more than its own diameter from the interradial superomarginals, in line with, and between, two of the radial tubercles. Superomarginal plates, 12 on each side of each ray, the terminal 3 com- pletely in contact ; excepting the last 2, their length and breadth are nearly equal, the height rather less; the twelfth is much shorter than wide while the penulti- mate is intermediate between it and the tenth; the first is much wider on the inner margin than on the outer ; the skin which covers the aboral plates seems to extend over the superomarginals and a few scattered, very minute granules may be seen with a lens, but there are no spines or tubercles. Terminal plate relatively very large, 1.5 mm. long, 1 mm. wide at base, narrower at tip where it carries 4 horizontally projecting short, thick spinelets, 2 from the upper margin and one from each lateral margin. Actinal interradial areas with 12-14 well defined plates, arranged thus: a very small unpaired triangular plate close to the oral plates; distal to this a pair of relatively large, irregularly hexagonal plates, wider than long; next a trans- verse series of 3 plates much wider than long, and then 2 plates much wider than long, adjoining each of the interradial pair of inferomarginals ; laterally distal to these on each side, in contact with either the first or .second inferomarginal or both, is a much smaller elongated plate, and beyond this there may be a still more minute plate; all of the.se actinal plates carry more or less scattered minute gran- ules, and the larger ones also bear an incipient pedicellaria or a conspicuously larger granule; the largest plates may carry both a pedicellaria and a granule. Inferomarginals correspond exactly in number, size and position with the superomarginals; they are quite smooth and carry no spines or spinelets; on the distal margin of some of the distal plates, one or two minute granules may be detected. Adambulacral plates more than 30 on each side; the proximal ones are large, about as long as wide, with a marginal series of 5 subequal, flattened, round-tipped spines; distally the plates become very small and the number of spines reduced to 2 or, at last, to 1 ; on the surface of the proximal adambulacrals there is a series of 3 or 4 minute granules forming a distinct line parallel to the CLAEK: AUSTRALIAN ECHINODERMS 93 margin, but distally this line is gradually reduced to 2 granules and ultimately to 1 or none. Oral plates conspicuously swollen; each has a marginal series of 5 or 6 relatively large spines, the innermost at tip of jaw much the largest, about .75 mm. long; there is a single large granule on the surface of each plate, near center, and there are much smaller granules present, particularly near the median crest of the jaw. Color of upper surface pale gray; lower surface lighter but not quite white, although under a lens it looks so. My field notes say of the color in hfe of other specimens: "light olive gray, variegated with darker, the terminal plate and a distinct band on distal half of arm, dark; lower surface pure white except terminal plates which are dark. On preservation grays become browns." The paratj^pes now range from nearly white to very pale yellowish-brown; there are no mark- ings whatever nor do the terminal plates stand out in any contrast. Holotype, M. C. Z. no. 3160 from Broome. Gift of Captain Beresford E. Bardwell, June 1932. There are 4 paratypes dredged in June, 1932 near Broome, considerably smaller than the holotype but agreeing with it in all essentials. The skin of the aboral surface is scarcely indicated so that the outhnes of the plates are easily seen, the papulae are more conspicuous, the granules very minute. There are no tubercles on the basal radials but in the 2 larger specimens, the coming develop- ment of such tubercles is plainly indicated. The terminal plate has only a single median spine projecting out from the distal end; probably the presence of 2 in the holotype is al)normal. The lower surface is quite free from granules and the number of plates is of course more or less considerably reduced. In 1929, a very small pentagonal sea-star was dredged in the vicinity of Lagrange Bay, in 5-7 fms., which at the time was quite unidentifiable but which seems to be almost surely a very young specimen of this new species. It is 6 mm. in diameter with R = 3.5, r = 2.5, more or less. There are about 70 aboral plates, symmetrically arranged, 15 at center, and 10 or more in each radius; 3 or 4 pairs of papulae are also evident in each radius ; there are 2 superomarginals present on each side of each ray, the first relatively very large, the second much smaller; a relatively huge terminal plate with 3 projecting spines at the tip, completes each ray. Orally, 30 inferomarginals, 40 adambulacrals, 5 pairs of orals and 10 actinal intermediate plates cover the surface ; 2 spines at the tip of each oral plate are disproportionately big, and the same is true of the orals themselves, as well as the interradial inferomarginals. There are no granules on either surface. This is a very perplexing sea-star and as even the holotype is probably still 94 memoir: museum of comparative zoology quite young, its affinities are uncertain. It was at first beUeved to be simply the Juvenal form of Stellaster incei but in the large series of that species, taken at Broome and at Lindeman Island, there are a number of young individuals as small as the holotype of the present species and even smaller. Comparison shows at once that they are certainly not identical. The young Stellasters have wider arms, a more arched disk, more numerous and more granular aboral plates, fewer superomarginals, and the inferomarginal spines more or less well developed. Orally however the resemblance is extraordinary but Stellaster has many more granules and more numerous adambulacral and oral spines. The possibility that dubius is the young of Ogmaster or Iconaster is promptly negatived by compari- son with specimens of those genera. That it is really an Anthenoides seems im- probable owing to the long slender arms but the superficial resemblances between these specimens from Broome and young Anthenoides peircei Perrier from Bar- bados are so striking, it cannot be considered improper to let dubius remain in the same genus with the West Indian sea-star until more nearly adult material is available. Stellaster incei Plate 2, figs. 1-2 Gray, 1847. Proc. Zool. Soc. London, p. 76. This is one of the common sea-stars at Broome and 44 specimens lie before me, all but four from 5-8 fms. in the vicinity of Broome and southwestward to Lagrange Bay; the exceptions are 3 specimens from Lindeman Island, Queens- land, and a large adult from Augustus Island, sent on by Captain Bardwell. The specimens from Broome undoubtedly fall in the incei group, according to the latest revision of Stellaster, a careful study by Doderlein (1935, pp. 86-101). But they show great diversity among themselves in form and spinulation and lead one to doubt very seriously whether the forms to which names are given by the pre-eminent European authority on sea-stars, are really worthy of such recognition. This is not the place to take up the question in detail but some notes on the specimens at hand may be of value. In size the forty Broome specimens range in length of R from 14 to 72 mm. The form of these two extremes is the same, R = 2.3 r, and this may be considered characteristic of this group of specimens young and old; the extremes of form are R = 2r and R = 2.6r. The smallest specimen has an incipient tubercle on each of the five basal radial plates (i.e. the first plate of the carinal series), and these 5 CLARK: AUSTRALIAN ECHINODERMS 95 tubercles are commonly the first to appear and the largest of all, but it is not often that all 5 are present, symmetrically placed and uniformly developed. In some specimens they are wanting, notably in an individual with R = 35 mm., which has no tubercles at all; in this case, the plates which should bear these primary tubercles are easily distinguished and I am unable to satisfy myself that they never bore tubercles; it is not beyond question that these were accidentally knocked off in the dredge, when the specimen was taken, but careful examination of this and other specimens makes me feel this is very unUkely. After R = 40 mm. tubercles develop in more or less considerable numbers and quite irregularly; there is also great diversity in size and sharpness; they rarely exceed 2 mm. in height and may be that much in thickness at base; commonly they are more slender and they are very generally pointed but in some individuals they are quite blunt; the number ranges from a dozen or even fewer to more than 300; while the largest number naturally occurs on the largest specimen there is no close correlation between the number and size — a specimen with R = 55 mm. has only about 45 tubercles and most of these are very small. An important feature of these Broome Stellasters is the presence, in three of the largest specimens, of spines on the superomarginal plates; in all cases, they are low and pointed and irregularly distributed; in one case they are mostly on the distal plates while in the others they are chiefly in the interradii; in the largest specimen two often occur on a single plate, one near the upper, the other near the lower, margin. As Doderlein considers the absence of spines on the superomar- ginals one of the main features of the equestris group, in which he places incei, their occurrence in these Broome specimens is particularly interesting. It leads me to beheve that St. tuberculosus v. Martens, of which only a single specimen — from an imknown locality — is known (figured by Doderlein), is simply a speci- men of incei, hke these from Broome, with which it agrees in size and form. It is of further interest that the three specimens with superomarginal spines have 2 spines on many of the inferomarginal plates, especially in the interradii, but some large specimens without superomarginal spines show this same pecuUarity. In the smallest incei from Broome, single inferomarginal spines are just beginning to show on the first pair (interradial) of plates, but are well-developed on the follow- ing seven; in all other specimens they are present on all the inferomarginaLs. In life Stellaster incei is a handsome sea-star but the colors fade rapidly and disappear more or less completely. At Broome the coloration is surprisingly con- stant in shades, but very diverse in details. Young individuals are light brown above and cream-white below; as they grow, some, and at last many, aboral plates 96 memoir: museum of comparative zoology become outlined with dark brown; in some specimens, there is more or less varie- gation of light and dark brown; as a rule, the larger the specimen, the more con- spicuous are the dark shades and markings; some individuals have an olive or gray cast but fundamentally the color is brown. While the specimen is still very small, R = 18 mm., the pair of actinal intermediate plates, right behind the oral plates become dusky purple and soon the color has become deep purple and the dark color has appeared in other plates, particularly in the series adjoining the adambulacral plates. In larger specimens, these colored plates are a deep purple- brown and even the median inferomarginal plates are dusky-purplish. The number of dark actinal intermediate plates in any specimen, and indeed in any area, is subject to great diversity, so that no two specimens in a small lot will be exactly alike; each plate is always either wholly white or wholly dark, but as many as 19 plates in one area may be dark. Very early, an orange spot appears at the base of each interradial, inferomarginal sjiine; soon the oral spines begin to show orange, and as the animal grows the extent and intensity of the orange increases greatly. The adult incei is thus brown above, more or less marked with darker, and white below, with few or many interradial plates dark purplish- brown, and much intense orange on oral, adambulacral and (basally) inferomar- ginal spines; pedicels, including suckers, white. After drying, specimens become "museum color" but in well preserved individuals, the dark plates of the actinal areas are detectable by their dull brown-orange color, and an orange tint may still be noted in the adambulacral armature. One of the specimens taken at Broome in 1932 is notable for being sym- metrically hexamerous. It is adult with R =60 mm. and r =30. On one ray, at the distal end of the ninth superomarginal on the left hand side (seen from above) , there is an incipient secondary arm, as though the ray would fork at that point; the new ray has 3 superomarginals on each side, but those of the inner side are very small as is the terminal plate; there are two inferomarginal plates, one on each side; the ambulacral furrow does not divide to enter the new arm. On two other rays, at the distal end of the eleventh superomarginal, there is a distinct indication of similar abnormal growth but actual development of additional superomarginals has not occurred. The specimen taken by Captain Bardwell at Augustus Island differs ob- viously from specimens of like size from Broome in the much longer arms, for R nearly equals 3r. It thus resembles very closely the form which Doderlein figures (1935, p. 99, pi. XXVI, figs. 2, 2a) as S. equestris forma gracilis, originally named by Mobius as St. gracilis and based on a specimen from China. But I fail to CLARK: AUSTRALIAN ECHINODERMS 97 detect any other character, save arm-length, by which it can be distinguished from the Broome material, and as it seems to have had the same coloration judging from the present appearance, particularly of the actinal intermediate areas, it seems only reasonable to consider it a very long-armed incei. This decision is confirmed by one of the 3 specimens which Mr. Ward has sent from Lindeman Island. This individual has R=68 mm., r = 24 mm.; hence R =2.8r±. The actinal surface shows 36 of the orange-brown plates characteristic of incei. In practically every detail this specimen is identical with the one from Augustus Island. The other 3 specunens from Lindeman Island are very young, the smallest Stellasters I have ever seen. In the smaller R =9 and r =4 mm., hence R =2.25r. There are 6 superomarginal plates on each side of each ray and an incipient seventh adjoins the terminal plate in each series. The 5 primary radial tubercles are visible but very low. There are 7 inferomarginal plates in each series; most of them show no trace of an inferomarginal spine but on 1, 2 or 3 plates in each series a minute spine is present near the distal margin of the plate. The other specimen has R = 10.5 and r =4.5 mm., hence R = 2.33r. There are 7 superomar- ginal plates on each side but the seventh is distinctly smaller than any others. The 5 primary radial tubercles are barely evident. There are 7 inferomarginal plates in each series; the first are much the largest and have no marginal spines but most of the succeeding plates have the spine on the distal margin distinctly developed. Stellaster inspinosus Plate 3, figs. 1-2 H. L. Clark, 1916. "Endeavour" Rep., p. 48. This is the characteristic Stellaster of the western coast of Australia and is common near Fremantle. We dredged many specimens between that port and Garden Island in 2-3 fms. Their striking and constant coloration, so different from that of incei, was obvious evidence that they were not that species but the presence of at least a few inferomarginal spines in most of the specimens led me to suppose they might not be inspinosus. E.xamination however of the 34 specimens at hand, all adult with R = 70-90 mm., and comparison with a cotype of inspino- sus, shows plainly that all are representatives of that species. It is a little strange that no young or half-grown specimens were secured, for scores of specimens were dredged. 98 memoir: museum of comparative zoology Misled by my careless failure to refer to the convex disk plates, and by the presence of inferomarginal spines in his specimens, Doderlein (1935, p. 89) has described and figured this species under the name St. gibbosus. He published a very brief and hardly descriptive reference to gibbosus in October, 1916, but as my account with figures was published June 2, 1916, there seems to be no doubt that the less accurate name survives; the aboral plates are more or less gibbous but the inferomarginals are usually not entirely lacking in spines though these are commonly small, and poorly and irregularly developed. The coloration in life seems to be very constant but unfortunately disappears more rapidly and even more completely after preservation than does that of incei. The upper surface of the living sea-star is a dull brick red while the lower surface is a reddish white with a variable number of actinal intermediate plates deep red, nearly claret, and the inferomarginals, brown with a claret cast ; adam- bulacral and oral plates and spines, yellowish, sometimes almost orange, but pedicels pure white. The series of actinal plates adjoining the adambulacrals is quite uniformly red and in many cases, most, or possibly all, of the other actinal plates are also red ; this is particularly likely to be true of large individuals. The largest of the present series has R = 87-90 mm. and is completely hexam- erous; it is fau'ly but not entu'ely symmetrical, some of the rays being longer than the others; as three are broken at the tip, it is impossible to determine what their lengths in life may have been. Stellaster princeps Plate 4, figs. 1-2 Sladen, 1889. "Challenger" Ast., p. 323. It was a great surprise and pleasure to find this magnificent sea-star present at Broome, where we secured a number in 1929, of which 6 are at hand, and 17 in 1932. Several were brought up by pearl-shell divers, or by our own divers (in 1932), but most of the specimens were dredged in 5-11 fms. between Roebuck and Lagrange Bays. After its discovery by the "Challenger" near Booby Island in Torres Strait and Sladen's description in 1889, this Stellaster lay almost forgotten until Living- stone's discovery of a specimen from "Western Australia" in the Australian Mu- seum. His description and figures (1932) threw valuable additional light on its CLARK: AUSTRALIAN ECHINODERMS 99 characteristic features, but he most magnanimously and unselfishly made no reference to the fact that he was present and enjoj'ed the experience of the redis- covery of the species at Broome m August, 1929. That I described (1914) a specimen from Port Hedland, W. A., belonging to the Western AustraUan Museum, as a new species under the name of Stellaster megaloprepes, reflects no credit on my perspicacity and I am quite unable to account for my making no reference to princeps at that time. It was not until I saw the type of megaloprepes again in 1929 at Perth, after having handled princeps at Broome, that I recognized the identity of the two. Curiously enough, in his recent admirable revision of Stellaster, Doderlein (1935) has completely ignored my species, which is just as well, as it has no validity whatever. Sladen, Livingstone and I (in the description of megaloprepes) have given so many details as to structural features and spinulation as to make any detailed discussion of them here quite unnecessary but it will be of interest to tell some- thing of the diversity shown by the fine series of 23 specimens at hand. They range in size from R = 63 to R = 150 mm., the last very much the largest Stellaster ever recorded; there are several others nearly as large. The form does not show great diversity but there is gome difference in the relative length of the rays; as a rule R = 3r and 3br, more or less, but in very large specimens it is nearer 2.6r. In Ufe the disk is high, the vertical diameter about one fourth R; in preserved specimens, the height depends very largely on the rapidity with wMch the specimens are killed and preserved; the longer the process, the flatter the speci- men, as a rule. There is great diversity in the spinulation. In the smallest specimen (R = 63 mm.), the 5 primary tubercles on the basal radial plates are very con- spicuous, 4 mm. high, 2 mm. thick at base but sharply pointed; on the disk there are about 30 very much smaller spines and on the basal part of each ray (less than the basal half) there are 3-5 carinal, conical tubercles, not quite 2 mm. high, and about 20 very much smaller tubercles irregularly placed but tending to form 2 series on each side of the carinal row. There are no spines or tubercles on the superomarginals. The inferomarginal plates each bear 2 spines, the upper and inner, much the larger, about 4 mm. long; the distal half dozen plates bear only one spine each. Adambulacral plates (near mouth) have 6 rather slender furrow spines, and 2 large, flattened spines on the surface; of these the adoral is more than 2 mm. high and its breadth is fully half as much; the distal spine is much smaller. There are many relatively large pedicellariae on the actinal inter- mediate plates, especially on the series adjoining the adambulacrals. Oral plates, 100 memoir: museum of comparative zoology with 7 marginal spines, innermost largest, compressed, rounded at tip, outer- most small and pointed; on surface of each plate near the proximal end is a very large flat spine (somewhat larger than on the adambulacrals) set transversely across the plate. As growth proceeds the number of aboral tubercles increases both relatively and actually but there is considerable diversity in their size, form and distribu- tion; in some cases they are short and thick, with blunt tips, in others they are much more spiniform with very sharp tips; in some indi\dduals the carinal series does not extend outward beyond the base of the arm, and may be very incom- plete even there; usually it runs out about half the length and occasionally it extends clearly almost to the arm-tip. The 5 primary radial tubercles generally lose their preeminence as growth proceeds and in adult specimens can be made out only with careful examination; occasionally however they continue larger and more conspicuous than their fellows. In the largest specimen, there are more than 600 tubercles on the aboral plates and even on the interradial superomar- ginals small pointed tubercles occur; these are chiefly single at the upper end of a plate, but occasionally are near the lower end; now and then two occur on the same plate and on one first superomarginal there are 3 tubercles in a vertical series. Tubercles on the superomarginals are however quite unusual and occur only in the largest specimens. Livingstone (1932) has pointed out how numerous the inferomarginal spines become (5-7 on a plate) and how they are arranged, in the interradii, in two diagonal series, all of which is well exemplified in the big specimen in hand; the largest inferomarginal spines are 8 mm. long by 2.5 mm. wide. Sladen's (1889) description of the adambulacral and oral plates and the actinal intermediate areas fits the present specimens very well, but the big adambulacral spines are fully 6 mm. long. The coloration of princeps in life is one of its most characteristic features. The upper surface is olive-gray with dark gray or dark brown (almost black) markings; particularly conspicuous are a group of irregular spots distal to the middle of each arm, and scattered spots on the disk; the lightness or darkness of the ground color shows some individual diversity and the markings of course are not exactly ahke in any two specimens; as a rule, the larger specimens are darker and more spotted or marked. The oral surface is a clear ivory white, while a circular area, of which the mouth is the center, is a bright \'iolet or purple ; the pedicels are pure white in contrast to the ivory-white of the adambulacral arma- ture. The diameter of the purple area shows some diversity but its circular form is constant and there is no tendency for the interradial areas as a whole to be CLAEK: AUSTRALIAN ECHINODERMS 101 purple. Unfortunately the colors are very fugacious; soon after death olive-gray becomes brown and begins to fade, so that in a few hours the characteristic color is gone. In dry specimens the purple color persists for several weeks but it ultimately disappears completely. Dry specimens at the present time are more or less uniformly "museum-color." In most of the larger specimens the disk both above and below is discolored probably from the oily secretion of the hepatic organs. In some individuals, especially the younger ones, the dark markings on the rays persist in an orange-brown shade. ASTERODISCUS TRUNCATUS Coleman, 1911. Mem. Austral. Mus., 4, p. 699. A single small specimen (R = 70 mm.) of this interesting sea-star was taken by Mr. D. L. Serventy in a productive dredge-haul, Feb. 23, 1930, by the trawler "Bonthorpe," in 90 fms. at the western end of the Great Australian Bight, 33° 15'S X 126°22'15"E. It was presented to me by Professor Bennett. Although the aboral surface is badly rubbed by the rough handUng in the dredge, the speci- men is particularly interesting as the smallest individual of the species yet taken. Unfortunately there are no data as to the color in life; the dry specimen is light gray, somewhat darker aborally. Notes on the Subfamily antheneinae Among the most conspicuous and generally common sea-stars of tropical AustraUa are the members of this subfamily, and the diversity of their color and size is so bewildering that the division into well-defined genera and species is a most perplexing task. The great German zoologist, Doderlein, brought order out of what was almost chaos when he published (1915) his revision of the genus Anthenea, but the large amount of material collected in 1929 and 1932 and the considerable series in the Australian Museum^ necessitate a new attack on the problem. The arrangement and keys here offered are of course based on Doder- lein's work and it is hoped may prove as useful, but there is no expectation that they are complete or in any way the "last word" on the subject. In the end, ' The entire lot was most generously sent to me in Cambridge so that I could compare them with our collection. There have thus been available to me 249 specimens of Anthenea and its nearest allies. 102 memoir: museum of comparative zoology some Australian zoologist must be the authority, who, through the advantage of large numbers of living specimens from various points on the tropical and sub- tropical coasts of that continent, may find the key to the real inter-relationships of these fascinating sea-stars. Without any pretence of deUmiting the subfamily, the following 5 genera are here discussed. Key to the Genera of Antheneinae A. Dorsal surface and marginal plates covered with a closely crowded coat of flat-topped polygonal granules; some (often many) abactinal plates carry a large smooth tubercle, often with one or two smaller tubercles adjoining it closely or it may be replaced by 3-7 similar but smaller and closely crowded tubercles Anthaster A'. Dorsal surface not as above. B. Superomarginal plates of interbrachial arcs with 1-3 (usually 3) spines on the outer margin; many actinal intermediate plates with a tuft of spines instead of the usual pedicellaria; only 3 adambulacral spines in the furrow series Pseudanfhenea B'. Superomarginal plates with granules or tubercles or both, but no spines; no spines on actinal intermediate plates; more than 3 adambulacral spines in furrow series. C. Only 2 series of adambulacral spines, outer end of adambulacral plates being bare and smooth (except on distal third of arm in some Pseudoreasters). Some aboral plates, often many, particularly in the carinal series with big hemispherical tubercles, 2-6 mm. in diameter; superomarginal plates more or less vertical, with numerous coarse granules, the upper end not forming a conspicuous part of the aboral surface of the sea-star Pseudoreaster Tubercles on aboral surface much less conspicuous, usually not 2 mm. in diameter; superomarginal plates with upper half bare and smooth, forming a marginal border for aboral surface, with a few (often only 2 or 3) small tubercles or coarse granules on the lower half Gymnanthenea CK Adambulacral spines in at least 3 series Anthenea ANTHASTER This well-marked genus was established by Doderlein for the following species and, as yet, no other is known. Anthaster valvulatus Oreaster valvulaius MiiLLER and Troschel, 1843. Arch. f. Naturg. 9 (1), p. 115. Anthaster valvulatus Doderlein, lOl."!. .Tahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. (27) .30. Wlien Doderlein established his new genus, only a single specimen of this interesting sea-star was known — the original type on which Miiller and Troschel based the species, but a few years ago (H. L. Clark, 1928) it was learned that on CLARK: AUSTRALIAN ECHINODERMS 103 the coast of southern AustraUa, Anthaster is by no means rare. Now it appears that it ranges around the southwestern corner of the continent and as far north at least as Cottesloe, for there are 3 specimens at hand, loaned by the Perth Museum, 1 from Long Reach Bay, Rottnest Island, and 2 from North Beach, near Cottesloe. Apparently the species is not common and lives in moderately deep water. The specimens from North Beach are in poor condition and were apparently from beach "wrack," but the one from Rottnest is in excellent condi- tion and bears a label saying that the "colour" (presumably in life) was "brick- red." Pseud ANTHENEA This genus was also established by Doderlein for a single species: PSEUDANTHENEA GRAYI Anthenea grayi Perkier, 1876. Arch. Zool. Exp., 5, p. 94. Pseudanihenea grayi Doderlein, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 26. Perrier based his species on a single specimen in the British Museum from an unknown locaUty. Apparently it has never been met with since. PSEUDORE ASTER This genus was proposed by Verrill for the following species but has re- ceived scant attention. Its validity having been called in question by Doderlein (1915, p. 49) even after some of its characters had been pointed out by Fisher (1911, p. 174), the availability of a large number of specimens, showing great diversity in size and tuberculation, has led to a careful reexamination of its status, which results in the belief that it is fairly entitled to recognition. No second species has as yet been found but it is not at all unlikely that it may be discovered in the marvellous marine Ufe of the northwestern coast of Australia. Attention should be called here to the curious slip of the pen by which Doder- lein (191.5, p. 49) gives "Protoreaster" as Verrill's name for this genus. As he had correctly quoted both Verrill and Fisher on the preceding page, and as he subse- quently (1916, p. 420) estabhshes a genus Protoreaster for Oreaster nodosus (L.) and two closely alUed species of that genus, without any reference to his use of the 104 MEMom: museum of comparative zoology name the previous year, it is obvious that the latter is simply a lapsus calami to which his attention was never called. It cannot therefore invalidate in any way his later use of the name. PSEUDOREASTER OBTUSANGULUS Plate 6 Asterias obtusangula Lamarck, 1816. Anim. s. Vert., 2, p. 556. Pseudoreaster obtusangulus Verrill, 1899. Trans. Conn. Acad., 10, p. 148. This is one of the commonest and most characteristic sea-stars of the north- western coast of Australia from Port Hedland (where Michaelsen and Hartmeyer procured their only specimens; see Doderlein, 1915, p. 49) northeastward to Augustus Island, at least. There is only a single specimen in the Australian Mu- seum and that is from the "coast south of Broome." Apparently no specimens are in the Western Australian Museum. It is not unreasonable to suppose therefore that the genus is confined to the strictly northern coast of Western Australia. The present series consists of 26 specimens — 4 adults from Augustus Island (B. E. Bardwell leg., October, 1933), 8 adult and young from False Cape Bossut (Sep- tember, 1929), and 13 adult and young from the Broome region (September, 1929 and June, 1932), plus the small adult belonging to the Australian Museum, al- ready mentioned. The smallest specimen has R =42 mm., r = 20, while the largest has R = 115 mm., r=56; this ratio of R=2r seems to be remarkably constant regardless of size or locality. The aboral surface is elevated so that the form is more like Oreaster than in any species of Anthenea. The vertical diameter in life may be .40R and even in some dried specimens it is .33R; most individuals however become nearly flat when dried. The diversity in tuberculation of the aboral sur- face is very great. The young individuals have the big tubercles scarcely 2 mm. in diameter and they occur only on the carinal series of plates; occasionally this condition persists into adult life (thus a specimen with R=80 nam. has but 3 tubercles, exceeding 3 mm. in diameter, one at the left hand side of the tip of one ray and 2 at the right hand side of the distal half of another ray) but usually the big tubercles are fairly numerous, though very irregular in distribution, in speci- mens with R = 50 mm. or more; an extreme case is an individual with R = 70 mm., which has 83 big tubercles distributed thus: 18 on one ray (3 on distal supero- marginal plates) , 14 on the next ray, 20 on the next (2 on distal superomarginals) , CLARK: AUSTRALIAN ECHINODERMS 105 13 on the next (2 on distal superomarginals) and IS on the next (1 on a distal superomarginal). The occurrence of the big tubercles on the superomarginals is by no means uncommon but it is not characteristic. More common is the reverse condition where the coarse granulation of the superomarginal plates is repeated on some of the distal aboral plates, replacing the big tubercles. Indications of this occur in the great majority of adults but the extreme case is a specimen with R = 76 mm. in which the big tubercles are confined to the carinal series and only those tubercles nearest the disk exceed 3 mm. in diameter; but there are 77 aboral plates on the distal portion of the rays which bear clusters of coarse tubercles similar to those on the superomarginals. The oral surface shows little change in growth from the smallest to the largest specimen and there is surprisingly little individual diversity. The color in life is often bright red, with the superomarginal granules distally, black, and the big tubercles brown or red. Some specimens are not so bright colored, and my field notes, from which the following items are quoted, show that there is a great deal of diversity: "As a rule the color was varied and variegated with dull grays, brown, reds and purples." "In color, no two are alike; all are variegated with browns, ranging from deep red-brown to brown-orange; often interbrachial areas have large blotches of a very dark color; orally cream-color with interradial areas, at least the inferomarginals, red, orange, brown or dull red violet, more or less." "The handsomest specimen is grayish with much red of several shades, especially large red blotches in interbrachial areas." Another was "mottled browns and greens, rather dark." The bright colors fade quickly after death particularly if the spechnen is dried. The reds become more and more orange and often disappear entirely but most dry specimens retain more or less of the orange tint and a few are distinctly light dull reddish. GYMNANTHENEA' gen. nov. Genotype, Anthenea globigera Doderlein, 1915. Jahrb. Nassau. Ver. Naturk, Wiesbaden, 68, p. (29), 50. Although Doderlein seems to consider his globigera rather close to australiae, owing to the bare upper ends of the superomarginal plates, it is really quite well set apart from that species and other Antheneas by its adambulacral armature which, as Doderlein points out, is much like that of Pseudoreaster. Indeed the ' yvjivos =nafcfd+ Anthenea, in reference to the bare aboral surface of the superomarginal plates. 106 memoir: museum of comparative zoology oral aspect of Gymnanthenea is rather markedly like that of this otherwise quite different genus. As a second species, closely allied to globigera, occurs at the Abrolhos Islands, it seems best to place them in a genus by themselves, hence the formation of Gymnanthenea. Key to the Species of Gymnanthenea Dorsal tubercles more or less numerous; ilorsal pedicellariae few, small, inconspicuous; dorsal surface covered witli a thin skin globigera Few or no dorsal tubercles; large pedicellariae conspicuous on aboral plates, especially on adradial series; 2-5 tubercles but no pedicellariae on superomarginal plates; dorsal surface covered with a thick, dark skin laevis Gymnanthenea globigera Plate 11, fig. 2 Anthenca globigera Doderlein, 1915. .Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. (29), 50. This species was based on a single specimen from Turtle Island, Northwest- ern Australia, collected by Michaelsen and Hartmeyer. It proves to be one of the commonest sea-stars in the vicinity of Broome, especially in shallow water along shore. Young individuals occur frequently under rocks in tide pools while the larger specimens are common in water 1-5 fms. deep. It is a well-characterized form and shows comparatively little diversity. The series at hand ranges from R = 6.5 mm., r = 3.5 mm., to R = 85 mm., r=40 mm.; hence R=2r in all except the very youngest individuals. In the youngest, there are no tubercles at all and the superomarginal plates have only very fine granules, but in a specimen with R = 12 mm., the 5 basal radial tubercles are evident; when R = 20 a number of additional tubercles are present in the carinal series on each ray ; a small tubercle and coarse granules are present on each superomarginal, on the outer side; the upper surface, like that of most of the aboral plates, is relatively smooth and flat. With increased growth, the number of tubercles and granules increases, until the adult condition is reached when most of the dorsal surface, excepting on the distal part of the rays, is well covered with tubercles, granules and small pedicellariae; in typical specimens, the distalmost radial plates and the last 6-8 adradial plates on each side are quite bare but have a fine shagreen-like surface; the upper side of the distalmost superomarginals may have a similar surface. In life, young specimens are cream-color or pinkish-white, with marginal CLARK: AUSTRALIAN ECHINODERMS 107 plates distally pink; then dark mottlings appear on the rays and by the time R= 30-40 mm. great diversity of color is evident. Individuals are often bright orange-red, but owing to the fact that the lines between the different species of Anthenea were very vague in my mind during the field work and the number of specimens taken was too great to permit of each one being properly noted as to color and other characters, it is impossible now to say whether any particular colors or shades are associated in any characteristic way with this species. To the best of my recollection orange, pink and red shades were common and in adults merged with browns and grays, in bewildering dissimilarity. The dry specimens are for the most part more or less "museum color" but some individuals give distinct indications of having been variegated with orange or red; the disk is often darker than the distal part of the arms. The 55 available specimens of globigera are from the following localities: "North Australia": 1 specimen with R =59 mm., identified by me (1928, p. 384) as " Anthenea flavescens." This specimen has nothing to do with flavescens but is undoubtedly globigera and raises the question as to where on the coast of the Northern Territory this specimen was taken. Loaned by the South Australian Museum. "Northern Territory": 1 young specimen dry and broken, also loaned by the South Australian Museum. Here again it would be very interesting to know just where on the coast of the Northern Territory this specimen was taken. The identification admits of no question. Port Darwin. Messrs. Christy and Godfrey leg. et don. 1 specimen, adult in excellent condition but bleached to a uniform brownish cream-color. Not typical, for the granules on the superomarginal plates are smaller and somewhat more numerous than they should be, and the outer series of the adambulacral armature usually contains 3 and sometimes 4 spines. It is not rare to find 3 spines in a series on Broome specimens but in no case is the number frequent, and 4 have not been found on a single plate. Loaned by the Australian Museum. (G 3813). Port Darwin, 1935. 1 specimen, young (R=48 mm.). Loaned by the Australian Museum. (J 5894). 108 memoir: museum of comparative zoology Western Australia: Augustus Island, October 1933. Capt. Beresford E. Bard- well leg. 8 typical specimens, small adults, and young. Broome, August and September, 1929. 21 specimens, adult and young, but mostly young ones taken in tide pools or near jetty. Broome, June 1932. 20 specimens, adult and young, but mostly adults dredged in Roebuck Bay and south- westward. False Cape Bossut, September, 1929. 1 adult. "Estuary of Swan River, near Fremantle." 1 small adult, with exceptionally thick skin, loaned by the Australian Museum. The locality seems to me highly improbable. Gymnanthenea laevis' sp. nov. Plate 19, figs. 4, 5 Rays 5. R = 70 mm., r =34 mm. R=2r. Upper surface elevated, particu- larly on the basal portion of each ray where the vertical diameter is fully 20 mm. whereas the distal part of the ray is flattened and only 9-10 mm. thick. Upper surface covered with a smooth, thick skin, in which the large groups of papulae are evident and the big pedicellariae are conspicuous. Tubercles wholly wanting, except near the base of each ray, where the first of the carinal plates carries an obvious tubercle about a millimeter in diameter but scarcely a millimeter high; proximal or distal to this there is usually a smaller tubercle and similar small tubercles may occur on one or both sides; althogether about 15 of these small tubercles can be made out with a lens. Pedicellariae, low, horizontal, with jaws 1-3 mm. long; as they are nearly white in contrast to the skin, they are very conspicuous; there are 15-20 on the basal portion of each ray but confined to the disk; while often on the adradial series of plates, they are more usually on the adjoining series, and a few are even further from the carinal plates. Superomarginals form a conspicuous, smooth margin to disk and rays. The first ones are nearly 5 mm. wide by 4 mm. long but little more than 2 mm. high; on the vertical surface are 2 (or 3) tubercles, about a miUimeter in diameter, conspicuous because of their light color, and below them are a very few small granules. Distally the plates are somewhat smaller, but even the twelfth, at ' teems = smooth, in reference to the thick smooth skin and absence of tubercles. CLARK : AUSTRALIAN ECHINODERMS 109 the tip of the ray, is fully 2 mm. square and the pair of marginal tubercles is evident; granules do not occur beyond the fourth plate. Madreporite 5 mm. long by 4 wide, only 7 mm. from center of disk. Oral surface with large, skin-covered actinal interradial areas, the outlines of the plates obscured by the skin but each plate conspicuously indicated by its large pedicellaria, or a group of granules or both. Inferomarginal plates large and conspicuous, the middle pair, 5 mm. wide and 3.5 mm. long; they correspond well in number and position with the superomarginals, but at the tip of the arm, adjoining the very small terminal plate, an extra inferomarginal is intercalated. First inferomarginals bear 15-20 (or more) tubercles and a big pedicellaria conspicuous because of their light color; the pedicellaria lies across the plate and the larger tubercles are between it and the superomarginal; at the extreme outer end of the plate are some small granules; the largest tubercles are smaller than those on the superomarginal; on succeeding inferomarginals the armature is similar, but there are often two pedicellariae ; distally the granules disappear, and the number of tubercles falls to 9, 4, 3, 2 and 1 on the distalmost plates. The series of actinal plates adjoining the adambulacrals extends out to the twelfth inferomarginal; the first two or three each carry a huge pedicellaria with or without one or two tubercles; distally the number of tubercles increases to 3 or 4 and the pedicellaria decreases in size; rarelj^ a plate lacks a pedicellaria. The next series of actinal plates extends to the eighth inferomarginal and con- sists of about 15 plates; the first one is hke the first ones in the adjoining series wholly occupied by the huge pedicellaria but on subsequent plates the tubercles are more numerous and the pedicellaria is more often wanting; the remaining plates of the actinal intermediate areas, 35-40 in number, usually each carry a large pedicellaria and half a dozen small tubercles but not rarely the pedicellaria is wanting, especially on the smaller plates. Adambulacral armature very simple, in two series; a marginal or furrow series of half a dozen rather slender spines, the median ones longest, the outer- most on each side smallest ; on the surface of each plate close to the furrow series are two (sometimes 3 near base of arm) subequal, stout spines, side by side, expanded and rounded at tip; outer half of plate perfectly bare. Oral plates small, each armatured by marginal and surface spines like two adambulacral plates; marginal spinelets at apex of jaw relatively large and conspicuous, but the most proximal of the surface quartet is the smallest while the distalmost is the largest spine let on the whole oral surface. Color of dry specimen, aborally, purplish-brown with tubercles and pedi- 110 memoir: museum of comparative zoology cellariae whitish in more or less marked contrast; madreporite brown; orally, wood-brown with tubercles, spinelets and pedicellariae conspicuously hghter. Holotype, M. C. Z. no. 2772, from Western Austraha, Abrolhos Islands, "Wallaby Group, shore collecting." Percy Sladen Expedition. W. J. Dakin leg. et don. This specimen was identified by me (1923, p. 239) as Anthenea globigera Doderlein, but now that a large series of that species is available for study, it seems best to give this very exceptional specimen specific rank. Only further collecting at the Abrolhos can demonstrate the propriety of such action. ANTHENEA Geay, 1840. Ann. Mag. Nat. Hist., 6, p. 279. Genotype, Anthenea chinensis Gray, 1840, I.e. = Asterias pentagonula Lamarck, 1816, Anim. s. Vert., 2, p. 554. Since the majority of the species of this fine genus occur on the Australian coast it has seemed desirable to prepare a key to all the known species.' Doder- lein (1915) did this admirably but the large collections of Australian material available to me have led me to somewhat different conclusions in regard to the status and relationships of some of the species, and it seems best to state them here. The ehmination of Pseudoreaster and Gymnanthenea leaves Anthenea a more homogeneous group. The species are seldom sharply defined but it is hoped that the following key will prove as useful to future workers as Doderlein 's has been to me. I am recognizing as vahd, 19 species and 1 variety of which 14 species occur on the coasts of Austraha, chiefly in the tropics. Key to the Species of Anthenea N.B. Owing to the great diversity in body form shown by the more common species of Anthenea due in part to differences in age, size and environmental conchtions, and in part to methods of preparation and the care used therein, it must not be expected that this key will prove a satisfactory guide in every instance. Similar diversity in number, size and form of spines, tubercles, granules and pedicel- lariae will also give much tiouble but it is hoped, nevertheless, that normal, adult specimens can be satisfactorily traced to the proper species. Senescent specimens of acuta are very perplexing, and of course, young individuals of any species with R = 35 mm. or less, will always give difficulty. It is doubtful whether such young individuals can be identified correctly without comparison with considerable series of specimens. ' 1 have not included Anthenea mexicana A. H. Clark (1916), as the single specimen on which it is based is only a half grown individual without distinctive characters and said to be from Mexico, a locality which certainly requires confirmation. CLARK : AUSTRALIAN ECHINODERMS 111 .. Arms relatively long, R= 2r, more or less (1.90-2.45 r in most eases, but in some senes- cent specimens only 1.5 r), narrowed distally, and commonly more or less pointed (at one-tenth R from tip, width is about .15 R or even less but may be .18-20). B. Each of the 5 basal radial plates with a conspicuous pointed tubercle; many distal superomarginals and some inferomarginals bear a spine-like pointed tubercle . . . rudis B^ Not as above. C. Pedicellariae extraordinarily abundant on oral surface, many of small size; infero- marginals with G-15 or even more; actinal intermediate plates with the usual pedi- cellaria, often accompanied by 2 or 3 smaller ones poh/giiatha C . Pedicellariae not excessively abimdant, orally, rarely more than 3 or 4 on an infero- marginal plate. D. Disk elevated, covered with a smooth, thin skin, with few pedicellariae or small spinelets; big blmit spinelets on many aboral plates; superomarginals interradially low and wide with few (1-8) granules, but distally higher than wide, with more granules obesa D'. Not as above. E. Superomarginal plates in interbrachial arc more or less horizontal forming a conspicuous part of the aboral surface. F. Aboral plates and inner portion of superomarginals bare. All adradial plates with pedicellariae flavescens Only a few adradial plates with pedicellariae flavescens var. nitda F . Aboral plates more or less covered with spinelets, tubercles and pedicellariae; superomarginals (at least the lower half) well covered with granules. G. Aboral surface with numerous bluntly pointed tubercles, forming 5-9 distinct longitudinal series on each ray; inferomarginal plates closely covered with small granules, and only a single small pedicellaria (2 mm. =>= ) or often none elegans G^. Aboral surface diverse; inferomarginals with coarse granules, at least at outer end and usually 2 or more large pedicellariae. Aboral tubercles numerous and coarse; upper half of superomarginals bare on each side with a few coarse granules, often in a narrow, or even a single, vertical series crassa Aboral tubercles few, scattered, relatively insignificant; superomarginals rather uniformly covered with granules though they may be fewer and coarser at upper end (See also pentagon ida) aspera E^. Superomarginal plates in interbrachial arc more or less vertical or so small they form an inconspicuous part of the aboral surface. (See also elegans). H. Pedicellariae very large both aborally (with jaws up to 1.7 mm. wide) and orally (with jaws up to 4 mm. wide), and correspondingly con- spicuous crudelis iV'. Pedicellariae numerous but not exceptionally large. J. Disk more or less conspicuously elevated, the reticulate nature of its skeleton often distinctly evident in big specimens; superomarginal plates well covered with granules to upper margin acuta 112 memoir: museum of comparative zoology J^. Disk usually not much elevated, or reticulated; upper end of supero- marginals more or less bare. Superomarginals low, wide, with tubercles only on lower lialf ; aboral tubercles low and more or less flattened australlae Superomarginals high, narrow, with a large granule near top; below it a swollen pair or trio and the rest of the plate (lower half) more or less covered with smaller, but often coarse, granules; aboral tubercles high, not flattened (See also sihogae) conjungens A^. Arms short and rounded at tip, form often quite pentagonal; R = l.G-1.8r; at one-tenth R from tip, width is .20 R or more. K. Aboral plates near arm-tips not much enlarged or conspicuous; aboral surface with few (or more numerous) tubercles or large spinelets, but commonly with very numerous pedicellariae and small spinelets, (specimens with R = C5 mm. or less may have quite smooth aboral plates and few pedicellariae). Aboral tubercles few, irregularly scattered, seldom capitate; superomarginal plates with coarse granules, often only a few on inner end, the coarsest at outer; coarsest granules of inferomarginal plates also at outer end; rays often somewhat pointed. peniagonula Aboral tubercles more or less numerous, especially on rays, somewhat capitate; marginal plates rather uniformly covered by a coat of small granules; rays wide and rounded regalis K^ Aboral plates near arm -tips enlarged and more or less conspicuous; aboral surface usually with more or less numerous tubercles. (See also regalis). L. Enlarged aboral plates near arm-tips, each with several (2-8) large granules or small tubercles. M. Whole aboral surface covered with big pointed tubercles, 9-13 series on each ray acanthodes M''. Aboral surface with relatively small, blunt or capitate tubercles. O. Aboral tubercles numerous, small, in about 9 parallel series on each ray. . . godeffroyi O^ Aboral tubercles fewer, larger, in not more than 5 series on each ray. Dorsal surface with few, small spinelets but many pedicellariae sibogae Dorsal surface with numerous more or less capitate spines and few pedi- cellariae merloni I}. Enlarged aboral plates near arm-tips, each with 1 large granule or tubercle (some- times 2). Tubercles of aboral surface and of superomarginals few and coarse. . . .tuberculosa Tubercles of aboral surface and of superomarginals, small, more like large gran- ules viguicri Although I have inckided in the above key, these 20 nominal forms, it is my candid opinion that several are not vahd. 1 have no confidence in either flavescens or its variety 7iuda and beheve they are based on young, probably unidentifiable, specimens of some of the other species', but as I cannot prove Pcssibly Gymnanthenea globifera (Dod.) CLARK: AUSTRALIAN ECHINODERMS 113 this, it is necessary to leave them in the key. The species acanthodes, crudelis and godeffroyi are each based on a single specimen and are open to question on the grounds of hybridity or extreme individual diversity. They can only be accepted as valid when more specimens are found. The species crassa and aspera occur at Port Curtis together and it is possible they represent the two extremes of a very variable species. On the other hand, connecting links may be consid- ered as hybrids. Only further collecting at Port Curtis and observations on the living animals can clear up the possible doubt. The group of Antheneas with short, wide, rounded arms, having enlarged tuberculated aboral plates near the tip, is a puzzling one and only much more material can decide whether there is more than one species. As the oldest name, tuberculosa will stand, but I am very dubious whether mertotii and viguieri can be maintained. The probability of sibogae being valid is better, but as stated above godeffroyi rests on a single specimen and may prove to be only a variant, and acanthodes may be a hybrid! The two species pentagonula and regalis are not very clearly separable and may possibly be identical. Laj'ing aside these doubts for the present, except for flavescens and )iuda, the IS species of Anthenea show the following distribution. On the southeastern coast of Asia and perhaps in the northern East Indies is pentagonula, the geno- type. To the west are found rudis, from the Mergui Archipelago to the Persian Gulf, and regalis, coasts of India and Ceylon. In the southern East Indies and on the coast of northern Australia occur sibogae and its near relatives mertoni, tuberculosa and viguieri. (The related godeffroyi is also supposed to be Austra- lian). To the southeast are found aspera, a^assa and acanthodes on the coast of northern Queensland, while further south, even reaching Tasmania (apparently) is acuta. No Antheneas (s.s.) are known from the long coast between Darwin and Cape Leveque but southwestward from the latter point are found at least four species australiae, conjungens, elegans and polygnatha, while from still further south off Geraldton, W. A., comes obesa. The locaUty whence crudelis comes is unknown. Since the genera Anthaster, Gymnanthenea and Pseudor- easter are exclusively Australian, it is evident that the tropical and subtropical coasts of Australia have proved a very favorable area for the speciation of An- theneinae. \Miile it is hoped that the key given above will prove of practical value in enabling collectors and workers to distinguish the different species of Anthenea, it is obvious that the sequence of species therein is in no way natural. It is not possible in the light of our present knowledge to arrange a natural sequence; 114 memoir: museum of comparative zoology hence in the following pages, the 20 forms will be divided into 2 groups, the non- Australian and the Australian. In the first group the species will be discussed alphabetically and in the second, the known species will be discussed alphabetical- ly, before the new forms are described. non-australian antheneas Anthenea crudelis Anthenea australiae var. crudelis Doderlein, 1915. Jahrb. Nassau. Ver. Naturk. Wies- baden, 68, p. 53. Although Doderlein considers the unique specimen on which this form is based, a variant of australiae, it seems to me better to treat it as a distinct species, since it is from an unknown locality. The huge and relatively few oral pedicellariae, the markedly pointed rays, and the few tubercles and numerous big pedicellariae of the aboral surface, distinguish it readily from all Australian Antheneas. Anthenea flavescens Hosia flavescens Gray, 1840. Ann. Mag. Nat. Hist. 6, p. 279. Anthenea flavescens Perrier, 1876. Arch. Zool. Exp., 5, p. 92. Anthenea flavescens var. nuda Doderlein, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 42. All of the specimens hitherto referred to this species and variety are so small that I doubt whether their actual specific relationship will ever be satis- factorily established. The figures given by Doderlein represent two species not at all closely related (evidenced by the number and character of the supero- marginal plates but even more strikingly by the difference in the form and length of the rays and the number, arrangement and surface of the aboral plates). The smallest specimen of pentagonula at hand (in the AI. C. Z. series from Hong Kong) has R = 33-34 mm. and hence is considerably larger than the type of nuda (R = 27 mm.) but there is httle doubt in my mind that Doderlein's specimen (from "Hahnahera Strasse") is a young pentagonula. The individual which my German colleague refers to Gray's species flavescens is also very young CLARK: AUSTRALIAN ECHINODERMS 115 (R=only 37 mm.) and its identity is quite uncertain. Future workers should refrain from referring Antheneas to flavescens unless they have adult material. And the distinctive characters of such adults are as yet unknown! Anthenea pentagonula Asterias pentagonula Lamarck, 1816. Anim. s. Vert., 2, p. 5.54. Anthenea pentagonula Perkier, 1876. Arch. Zool. Exp., 5, p. 90. This seems to be a common sea-star at Hong Kong, or at least it was 75 years ago, for the M. C. Z. has 24 specimens taken by Captain Putnam in 1861. There were originally many more but they have gone to museums in all parts of the world in exchange for more desired material. Young individuals have the arms relatively long and somewhat pointed and the short wide arms of the adult are not well-marked until R =75 mm. As the "Albatross" took no specimens in the Philippines and the "Siboga" secured no adults in the Dutch East Indies, -pentagonula does not appear to have ranged far from the mainland coast of Asia. The largest of the M. C. Z. specimens has R =90-95 mm. Anthenea regalis KoEHLER, 1910. Ech. Indian Mus. : Asteroidea, p. 82. Tliis species seems to be confined to the coasts of India and Ceylon. A speci- men in very poor condition with R = 100 mm., was secured by me on the beach at Madras, India, in August, 1932. It agrees well with Koehler's figure. A much smaller specimen from Ceylon, in excellent condition, with R=47 mm., is also in the M. C. Z. collection. It agrees well in body form with the adult specimens but has not yet developed nearly so many tubercles; this is true also of Doder- lein's Ceylon specimen with R =51 mm. It is unfortunate that Doderlein has referred very young Antheneas from the Phihppines and from Torres Strait to this Indian species. Fortunately however he gives excellent photographs of these young individuals from which one may feel quite sure that neither is regalis; nor can they both well be referred to the same species; note the striking difference in the size and number of the papulae. It seems to me possible that the smaller specimen may be a young sibogae but I have no suggestion as to the identity of the larger. 116 memoir: museum of comparative zoology Anthenea rudis KoEHLER, 1910. Ech. Indian Mus. : Asteroidea, p. 86. This species is based on two Antheneas from the Mergui Archipelago, which are too young for any accurate determination. There are 3 specimens at hand, all obviously young, which seem to belong here. The smallest belongs to the Austral- ian Museum (J437) and has R = 14 mm. ; the big pedicellariae of the oral surface indicate an Anthenea and the probability that it is rudis rests on the conspicuous presence of 3 large radial tubercles; the locality whence it came is unknown but Mr. Livingstone thinks it is non-Australian. The other 2 specimens are in the M. C. Z. collection; one with R =31 mm. is from the Persian Gulf and the inter- brachial arc is almost exactly as Doderlein figures it in a Ceylon specimen; the other, with R =35 mm., is from Ceylon and differs from the smaller individual in having a conspicuous small tubercle (or very large isolated granule) on each superomarginal plate. Doderlein and Koehler refer to such tubercles in their descriptions but they do not appear in the figures given. Although all the known specimens of rudis are small (the largest is Doderlein's with R=40 mm.), the species seems to be very well marked and worthy of recognition. The known range is from the Mergui Archipelago to the Persian Gulf. australian antheneas Anthenea acuta Goniodiscus acuttis Perrier, 1869. Ann. Sci. Nat. (5), 12, p. 280. Anthenea acuta Perrier, 1876. .\rch. Zool. Exp., 6, p. 91. Although Doderlein (1915) was unwilling to use this name because Perrier's description was so short and indefinite, the universal use of it, during 60 years, for the common Anthenea of the New South Wales coast justifies the continued use unless there is some reason why such usage is really incorrect. In this case, as in scores of others, the original description is not distinctive when all of the now known Antheneas are considered, but I find nothing in it which is inappropriate to some specimens at least of the New South Wales species. Moreover the name is particularly appropriate to this species, since the rays are as a rule more acute than in most Antheneas; certainly for none of the now known species is the name CLARK: AUSTRALIAN ECHINODERMS 117 as generally appropriate. Doderlein has renamed the species (1915, p. 53) as A. australiae var. sidneyensis, regarding it as only a variety of the west coast species which he describes from Fremantle. The two have certain points in common but are, 1 think, really quite distinct. Thanks to the AustraUan Museum, there are at hand from that institution, 39 specimens of acuta, making with the 11 in the M. C. Z., 50 specimens available for study. I greatly regret not having seen Uving or fresh material, nor are there any notes at hand on the color in life. The present series of dried specimens shows great diversity in size, form and tuberculation. The smallest has R = 16 mm., r = 8, while the largest have R = 120-130 mm. and r = about 70 ; one specimen with R about 120 mm. has R= about 1.6r (the condition of the specimen precludes exact measurements), while a specimen with R =54 mm. has such long rays that R =2.45r. When well preserved, large individuals of this species have a notably high disk, with the vertical diameter as much as 40-50 mm. when R = 120-130 mm. But many specimens when dried are very flat, and all young individuals with R less than 20 mm. have the disk but Uttle elevated. As growth proceeds however the tendency is for the lesser radius (r) and the vertical diameter to increase more rapidly than R, so that full grown specimens have a flat, high body and relatively short, more or less triangular rays. There is great diversity in the number and size of tubercles and spinelets; very young specimens have but few and they are small; as growth proceeds there are divergent tendencies; in one hne tubercles become more evident and greatly predominate, while on the other hand the spinelets increase more markedly and tubercles are relatively infrequent or even almost wanting. The reticulation of the aboral skeleton which is often marked in large specimens first begins to be evident when R = 75-80 mm. but there are many large adults which do not show it at all. Nothing has been re- corded, so far as I know, of the color in hfe. The 39 AustraUan Museum specimens at hand are from three points on (or off) the Queensland coast from Fraser Island southward; from five New South Wales locaUties (30 specimens from Port Jackson) ; and 1 large individual from Tasmania. One specimen is labelled "Port Chalmers, New Zealand" but as Anthenea is not included in Mortensen's recent careful revision of New Zealand sea-stars, verification of this label is necessary. The specimen is a small but typi- cal acuta, and of course, Anthenea might occur on the New Zealand coast but hardly, one would think, at Port Chalmers. 118 memoir: museum of comparative zoology Anthenea aspera DoDERLEiN, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 35. There are 5 adult specimens of this fine species in the Australian Museum all from Port Curtis, Queensland; the labels on two give the additional information, "12 fms." The smallest specimen has R = 85 mm., r = 47 mm., R = 1.8r; in the largest specimen, R = at least 125 mm. but the tips of the arms are so up-curved, it is impossible to give an accurate measurement, r = 70 mm., hence R = 1.8r as in the smallest. In other specimens R exceeds l.Sr sUghtly and in no case is it less. In the types R = 1.72 and 1.74 r but so much depends on the way that R is measured this difference is not important ; if careful allowance is not made for the bending up of the arms at the tips, the present specimens have R = 1.7 r rather than 1.8 r. Besides the relatively long and pointed arms, the large and conspicu- ous superomarginal plates, usually closely covered with granules, and the aboral surface more or less closely covered with numerous pedicellariae, spinelets and pointed tubercles, distinguish this species. In the largest specimen the disk is elevated and the aboral skeleton somewhat reticulated as in acuta. An Anthenea from Port Denison, Queensland, with R only 65 mm. and the color a dull brownish purple, is referred to aspera with some hesitation, as the upper ends of the superomarginals are not well covered with granules, and in this particular the resemblance to crassa is notable; on the other hand the rays are narrow and pointed and the covering of the disk is more like aspera. As the speci- men is still immature, its identity cannot be positively determined. The color of the dried specimens of aspera shows some diversity. The largest two are dark brown aborally with the superomarginals and the oral surface yellow-brown in marked contrast. Of the other specimens, the smallest is (as already stated) dull brownish-purple and the next larger is a dull light brown with a distinctly purpUsh or reddish cast. The remaining two are brown with little contrast between the upper and lower surfaces. According to a note on the label of one of these specimens the color in fife was "Irregular mottling of brown and sage green; actinal surface yellow." Anthenea australiae DoDERLEiN, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 52. Among the Antheneas dredged at Broome in 1932 are 3 typical examples of this well-marked species. In the largest R = 85-93 mm., r = 40, R = 2.25 r± ; the CLARK: AUSTRALIAN ECHINODERMS 119 second specimen has R = 60-70 mm., r = 30; in the smallest, R = 47 mm. and r is about one-half as much. They agree almost perfectly with Doderlein's descrip- tion and figures, and the color is now a typical "museum-color" or as Doderlein puts it "gelbUch braun." It is greatly to be regretted that there are no field notes with these specimens to show what their color in life was. They were not dis- tinguished when dredging from the numerous specimens of conjungens and other species which we were constantly taking. Doderlein refers to Fremantle as the type locaUty for this species, and says there were many specimens. He adds that there was one small specimen from Shark Bay. It seems strange that we failed to take a single Anthenea in the vicinity of Fremantle either in 1929 or 1932. Furthermore there were none at the Museum of the University, nor did I note any in the Western AustraUan Museum. None have been sent to me by Professor Bennett. There is however a specimen in the M. C. Z., presented by Professor W. J. Dakin, which was taken "off Fremantle," a smaller specimen was dredged by Professor Dakin off the Abrolhos Islands. E\'idently the species belongs to the west coast fauna and is not com- mon as far to the northeast as Broome. Its occurrence at Fremantle, which is probably near the southern limit of its range, is apparently either very local or perhaps seasonal. Anthenea conjungens Doderlein, 1935. "Siboga" Ast.: Oreasteridae, p. 107. Doderlein bases this species on a single specimen in the Hamburg Museum from "Australia" — no more definite locality is known. It proves to be one of the commonest sea-stars at Broome and a fine series of 24 specimens was taken during September, 1929 and June, 1932. There are also at hand, 2 adult speci- mens from the Australian Museum, which are labelled as from "Northwestern AustraUa." The smallest individuals have R =23-25 mm. and r = 12-14 mm.; these little ones are notably well characterized by the fact that each aboral plate, with few exceptions, carries at its center a single, relatively large, rounded tubercle; now and then, a pedicellaria replaces or accompanies the tubercle and occasionally, the tubercle is replaced by a pedicellaria with a large granule on each side. In the smallest specimen, many plates on the distal portion of the arms are quite bare. One of the larger specimens is exactly the size of Doderlein's unique holotype (R = 75, r = 36 mm.) but is less flattened, the arms at middle being 13 mm. high; the vertical diameter of the disk is 19 mm.; this individual 120 memoir: museum of comparative zoology corresponds very closely in all details with the description and figure of the holo- type. The largest specimen has become much flattened but even were that not so, the rays are notably short and wide; R = 125 mm., r = 65 and br = 65; R = 1.9 r or br; the rays are thus nearly triangular with the sides about 65 mm. long. At the other extreme is a half grown specimen with R = 65 mm., r = 27 and br = 27; R = 2.4 r or br; considering the rays as triangles, the sides are 40 mm., the base only 27. On the whole, conjungens is one of the least variable species of Anthenea, except perhaps in color. As already stated, the failure to recognize the number of species occurring at Broome and hence to distinguish between them has led to a regrettable untrustworthiness in my notes. As far as I can remember, there was great diversity of color among these common Antheneas, with shades of orange or red, purple and brown predominating. The dry specimens show much diversity as they are, ranging from pale brown to bright brown in the young, and from light brownish to orange or to purplish in the larger ones, usually with irregular dusky blotches. Anthenea godeffroyi DoDEELEiN, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 45. This species rests on a single specimen with R = 79 mm., r = 44 mm. While closely allied to both merloni and sibogae, the parallel series of very small tuber- cles gives a very different aspect to the dorsal surface. When the original descrip- tion was pubhshed, the type of the species was supposed to be from Samoa, but in his latest publication Doderlein (1935) expresses the belief that it is from Aus- traUa, as it bears the same catalogue mmiber (in Museum Godeffroy) as the type of conjungens, which is undoubtedly from Australia. As Anthenea is not known from east of the Great Barrier Reef region^, it is quite unlikely that it occurs at Samoa. Moreover the obvious relationship of godeffroyi to the following species (mertoni, sibogae and tuberculosa) makes it highly probable that the unique holotype came from tropical Austraha. It will not be surprising if it turns out to be an extreme variant of merioni, a species of which we still know all too few specimens. ' Records from New Zealand, Fiji and western coast of Mexico are highly improbable and require confirmation. CLARK: AUSTRALIAN ECHINODERMS 121 Anthenea MERTONI Plate 7, figs. 1-2 KoEHLER, 1910a. Abh. Senckenb. Nat. Ges., 33, p. 268. This is a fine, well-marked species, originally described from a single adult specimen from the Aru Islands. Doderlein (1915) records a somewhat smaller specimen in the Berlin Museum from the "Gazelle" collection in INIermaid Strait on the northwestern coast of Australia. I cannot avoid the feeling that there is some mistake about this locality, for among all the scores of Antheneas collected in 1929 and 1932 along the northwestern coast, both east and west of Broome, no individual was taken which could possibly be referred to rncrtoni or one of its near allies, so I find it hard to believe that it occurs in Mermaid Strait or on any part of the Australian coast, west of Cape Leveque. There are at hand however half a dozen specimens of mertoni which demon- strate its occurrence in the Gulf of Carpentaria, and as far west as Darwin. In the Australian Museum collection are 4 specimens; 3 from Mapoon on the eastern side of the entrance to, and 1 from Pearce Island, Sir Edward Pellew Group, far down in, the Gulf of Carpentaria. The specimens from Mapoon have numer- ous more or less capitate aboral spines, and the distal aboral plates on the arms have few and large tubercles. In 2 of the Mapoon specimens R = 65-70 mm., r = 37-40, so that R = 1.7 r±, but the third is much smaller with R = 50 mm. and r = 26, so that R is nearly equal to 2 r; moreover the rays are narrower and more pointed than in the larger specimens, as is to be expected in young individuals. The specimen from Pearce Island is in very poor condition, badly water worn, but it is a large individual with R = 83 mm. and r = 50. At Darwin, we were so fortimate as to dredge near Channel Island, in July, 1929, a very fine specimen of this Anthenea. Mrs. Clark made a color sketch of it as soon as possible, which was fortunate, as the present "museum color" gives no idea of its handsome appearance in life. The upper surface was gray, with large, irregular blackish-brown blotches; the oral surface was Ught reddish- buff. In structural details and appearance, this specimen corresponds very closely to Koehler's tj^pe specimen, which was a little larger with arms a little longer. The Darwin specimen has R = 85, r = 50 mm. Among all the scores of Antheneas subsequently collected in the Broome region, we never saw one that resembled this specimen in either form, tuberculation or color. We dredged near Shell Island at Darwin however a very small Anthenea, with R = 10 mm., r = 5.5 mm., 122 memoir: museum of comparative zoology which I believe is a young mertoni, although of course, it does not yet show any distinctive features. In April, 1936, Mr. Livingstone sent me 2 young specimens from Darwin acquired by the Australian Museum in 1935. Anthenea sibogae DoDERLEiN, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 47. Among the specimens in the Australian Museum collection is an Anthenea (with R = 65 mm.) from Albany Passage, northern Queensland, 9-12 fms., which I am referring to this species with some misgiving. Another specimen of the same size, but with wider more rounded rays, as in mertoni, is also in the collection and is labelled Thursday Island. The two specimens are so aUke in the form and covering of the superomarginal plates that I feel sure they are the same species, and in this character of the upper marginal series they resemble Doderlein's figure of sibogae so closely, and differ so evidently from the much coarser granu- lation of mertoni that it seems best to refer them to sibogae. But it must be re- membered that sibogae is as yet known from only a single, large specimen, in which R = 102 mm. ; the arms are strikingly slender, though short, and the inter- brachial arcs notably flattened. If these features are constant, the species is very well characterized but the specimens at hand make one dubious as to their constancy, and if they are not constant, it is doubtful whether sibogae and mertoni can be maintained as distinct species. In this connection, it is important to note that the Antheneas Usted in my Torres Strait Report (1921, p. 29) as tuberculosa Gray are not that species (unless mertoni and sibogae are both synonyms thereof, as I believed in 1921 and still think is by no means improbable) but are evidently identical with the Australian Museum specimens which I am here calling sibogae. One of these Torres Strait specimens is young and is figured by me in its natural colors, which are bright and varied — very different from the Darwin mertoni or indeed from any other Anthenea I have ever seen. On account of this striking color difference, and the differences in superomarginal plates and aboral tuberculation, it seems wisest to let the three species stand but much more material is needed before their validity can be considered demonstrated. Anthenea tuberculosa Gray, 1847. Proc. Zool. Soc. London, p. 77. The diagnosis of this species was published in June and in September again in the Annals and Magazine of Natural History (1847, 20, p. 198). Only Clark: Australian echinoderms 123 a single specimen was in Gray's hands at the time and that was from Port Essington, Northern Territory, whence naval officers sent many a notable sample of the Australian fauna home to the British Museum. Later on (1866) Gray published a good figure of this individual and he does not seem to have had any other available specimen. Yet Perrier (1876, p. 88) refers to "echan- tillons desechfe" as "les types" of Gray and describes under the name tuber- culosa a sea-star which is apparently not Gray's species, quite a different thing to which Doderlein has since (1915, p. 52) given the name australiae. Apparently the true tuberculosa is a rare species, for D5derlein has seen but a single specimen, which he (1935, p. 106) records as taken by the Siboga at Pulu-Jedan in the Aru Islands; excellent figures are given. There are no speci- mens in the museums at either Sydney or Perth. In the M. C. Z. collection there is an Anthenea from an unknown locality, which I am referring to this species because of its very marked resemblance to Doderlein's figures. It has R = 67, r = 38 mm. (R = 1.75 r), and is perfectly flat aborally with a vertical diameter of 16 mm. The granulation, or better, the tuberculation of the supero- marginals is even coarser and more sparse than in mertoni and the distal aboral plates of the arms seldom have more than one large tubercle. This specimen probably came into the M. C. Z. collection from Ward's Natural Science Estab- lishment, Rochester, N. Y.; Mr. Ward made a trip to Torres Strait in 1896 and collected much marine material, some of which subsequently came to the M. C. Z. and this Anthenea was probably acquired then. Although it is easily dis- tinguished from all the specimens of mertoni and sibogae that I have seen, it is not unlikely (as already stated) that the three nominal species are in reality identical, as I held in my Torres Strait Report (1921, p. 29). Anthenea viguieri DoDEELEiN, 1915. Jahrb. Nassau. Ver. Naturk. Wiesbaden, 68, p. 34. This species was established to provide a name for a sea-star misidentified by Muller and Troschel, and later by Perrier, and by Viguier. The specimen is in the Berlin Museum and has been figured by Doderlein. A specimen, not quite so large, is in the M. C. Z. from Restoration Island, northern Queensland. It was secured from Ward's Natural Science Establishment, Rochester, N. Y., in 1896. There are no specimens in the AustraHan Museum. Whether the species is really valid or only an extremely non-tuberculated form of tuberculosa remains to be demonstrated. 124 memoir: museum of comparative zoology Anthenea acanthodes' sp. nov. Plate 18, fig. 2 R = 97 mm., r = 55 mm., R = 1.76 r. Disk high, 40 mm. in vertical diameter. Arms correspondingly arched at base but becoming flattened rather abruptly 35-40 mm. from the tip. Aboral surface of disk and basal part of arms, covered with big bluntly pointed tubercles or low, stout spines; on the basal part of arms these form 9-13, more or less distinct, nearly parallel series. Distal part of arms covered with irregular tuberculated plates as in the mertoni group; the tubercles on these plates are coarse and rounded hke those on the .superomarginal plates. The latter are rather few (14 or 15 on each side of a ray), vertical in position, much higher than long, conspicuously covered, except at margins, with coarse tubercles, the uppermost largest. Inferomarginal plates correspond in size and position; the outer end and distally the whole plate carries coarse tuber- cles ; in the interbrachial arc the inner portion of the plate is closely covered with small tubercles like those of the actinal intermediate plates. Pedicellariae on marginal plates (in both series) few and small, often wanting altogether. Oral surface not peculiar in any way but typically Anthenea-hke. Adambulacral armature stout but not remarkably so. Color of dry specimen, brown, lightest orally, darkest on aboral surface of rays. Holotype, AustraHan Museum No. J5367, from Port Curtis, Queensland. This is the most sharply defined species of Anthenea, occurring in Australia. It looks as though it might conceivably be a hybrid between crassa and mertoni but I do not believe for a moment that this is the case. Anthenea crassa^ sp. nov. Plate 18, fig. 1 R about 115 mm. (along the adambulacral furrow, the arms are 125 mm. long, on the aboral surface, about 105 mm.); r = about 55 mm., R = 2.1 r; at base of arm, br is rather more than r. Aboral surface low and somewhat flattened in holotype but in some specimens it is considerably arched; one with R = 90 ' o-KavdosSris = full of thorns, in reference to the remarkably spiny aboral surface. ^ crassus = coarse, in reference to the large tubercles and granules of the aboral surface and supero- marginal plates. CLARK: AUSTRALIAN ECHINODERMS 125 mm. has v.d. about 30 mm. In this specimen a certain amount of reticulation of the aboral skeleton is evident but no other specimen shows it. Aboral surface of disk and rays covered with coarse tubercles or stout blunt spines, which, on the basal part of the rays form about 9 more or less definite series ; in the holo- type, which has the heaviest tubercles (often over 2 mm. high and more than 2 mm. in diameter at base) the series are not very well defined but can be dis- tinguished along the median radial line. In all the paratypes, the tubercles are smaller though they may be 2 mm. high. Pedicellariae aborally are very few in the holotype and some other individuals, but in several specimens they are numerous. Superomarginal plates about 15 in number, lying more or less on the aboral surface in the interbrachial arcs but becoming more vertical distally; in the specimen with the high disk, they are nearly vertical even in the interbrachial arc. Distally they are pretty well covered with coarse granules but basally the margins are more bare and the uppermost granules are much the largest and form a single vertical series of 2-4. Inferomarginal plates more numerous (16 or 17) and not corresponding exactly in position with the upper series; they are closely covered with granules, coarsest at the outer end, and carry several (2-6) pedi- cellariae. Oral surface not peculiar but the adambulacral armature is exception- ally heavy, all the spines, especially the marginal series, being more than usually stout. Color of dry holotype, dark brown but some of the paratypes are more yellow-brown. Holotype, Australian Museum No. J5368, from Port Curtis, Queensland. There are 7 paratypes, all from Port Curtis, where this species evidently replaces acuta of the more southern coast. The smallest has R=70 and r=35 mm.; the tubercles are of course much smaller than in the holotype but other- wise the resemblance is marked. One individual with R = 95-100 mm. is notable as having the longest arms relatively of any Anthenea examined; r is only a little more than 40 mm., hence R =2.4-2.5 r. The narrowness of the arms is indicated by the fact that 40 mm. from the mouth br is only 35 mm. An Anthenea from Port Curtis, with R = 105 mm. has given much difficulty in its identification but owing to the position, character, and armature of the superomarginal plates, it is here referred to crassa, with the suggestion that it is a hybrid between crassa and aspera, the two common Antheneas at Port Curtis. The color (brown aborally, with the marginal plates and oral surface, yellow- brown in rather sharp contrast) is like that of the large specimens of aspera, and the attenuate, somewhat pointed rays, are also Uke that species. Moreover there are numerous pedicellariae and few spinelets on the aboral surface as in 126 memoir: museum of compaeative zoology aspera. But because of the superomarginal plates, it cannot be placed in that species without qualification. The possibility of its being a hybrid seems reason- able. Anthenea elegans' sp. nov. Plate 18, fig. 4 R = 120 mm., r = 60, R = 2 r; br (at 60 mm. from mouth) 52 mm. Disk moderately elevated, v.d. =32 mm. Nearly all of the paratypes are flat or only a little convex. Disk covered with small pedicellariae, spinelets and tubercles, rounded or bluntly pointed ; if pedicellariae are very numerous, the spinelets are rather few but if pedicellariae are few the spinelets are very numerous. In the holotype, the tubercles are large, up to 2 mm. in height and in diameter; they are arranged in 9-1 1 well separated series on each arm and between them the pedicel- lariae are excessively numerous. Paratypes not essentially different but the tuber- cles are smaller and more crowded and there are not so many series on each arm. Superomarginal plates about 18, much wider (in the interbrachial arc) or higher (distally) than long, well covered with granules, largest on convexity of the plate, smallest close to the margins ; in young specimens the marginal portions of each plate are quite bare and the uppermost granules are enlarged; distally, in adults, some or many granules are enlarged; especially on upper half of each plate. In many cases, the superomarginals are nearly vertical even in the interbrachial arcs but they always form a conspicuous border to the disk and rays. Many aboral plates near arm tip are somewhat enlarged and carry several tubercles and gran- ules as in the mertoni group, but the character is well-marked only in big adults. Inferomarginal plates notably large, uniformly and closely granulated, with a single small pedicellaria or none (in the holotype, more commonly, none) ; on a few plates at the middle of the interbrachial arcs, the granules at the outer end are notably enlarged and in some j^oung indi\'iduals the same is true; in most adults however, this does not seem to be characteristic. Oral surface as usual in the genus, but the adambulacral armature is very well developed and the series of spines back of the marginal series contains 3, and often 4, more or less flattened spines. Color (dry) more or less dull purple-brown, but lightest on marginal plates and near arm-tips. Holotype, M. C. Z. No. 3192, from Broome, Western Australia, 5-8 fms., June, 1932. ' elegans = handsome, in contrast to crassa, in reference to its more beautiful form and finer appearance. CLARK: AUSTRALIAN ECHINODERMS 127 The holotype is much the largest specimen but there is a good series of 16 paratypes with R ranging from 50 to 100 mm. On the whole there is relatively little diversity in the form, or in tuberculation of the aboral surface, but there is a good deal of diversity in the position of the superomarginal plates; the degree to which they form a conspicuous part of the dorsal surface depends in part on the care with which the specimens were prepared, including the rapidity or slowness of the drying. But it also depends in some degree on the size of the individual and the convexity of the disk. This species is nearly as common at Broome as conjungens but adult speci- mens of the two species are easily distinguished. Individuals with R less than 50 mm. are more easily confused. My field notes however give no clue as to the colors in life of elegans; presumably they were as diversified as in other species. The smallest specimen at hand is a curious monstrosity, as one of the ambu- lacra forks 6 mm. from the mouth and gives rise to a sixth ray slightly smaller than the other five; it has 13 superomarginals on each side whereas the other rays have 15. Anthenea obesa' sp. nov. Plate 19, fig. 1 R = 100 mm.; r = 48 mm.; R = more than 2 r; br at 48 mm. from mouth, 37 mm.; vertical diameter of disk about 33 mm. Aboral surface covered with a smooth, thin skin, with very few pedicellariae or small spinelets but with about 230-250 coarse, blunt or truncate tubercles sparsely and irregularly distributed over the disk and the proximal portion of the arms; these tubercles are 2-3 mm. high and up to 2.5 mm. in diameter; they tend to form 2 parallel series on each ray. Superomarginal plates about 16 on each side of an arm, very low, (longer than high) on basal half of arm but becoming higher than long distally; in the interbrachial arc the plates are scarcely visible from above and there is usually but a single small tubercle on each plate; sometimes however a second small tubercle, or several granules may be present ; distally 5 or 6 coarse granules occur on each plate and the plates themselves form a conspicuous margin to the ter- minal half of the arms. Inferomarginal plates moderately large, quite uniformly covered with coarse granules. Oral surface as usual in Anthenea but adambula- cral spines in all three series are unusually short and stumpy. Color in Ufe is ' obesus = fat, in reference to the high disk and arms. 128 memoik: museum of comparative zoology recorded as "scarlet or crimson-red;" the dry specimens are light reddish-brown above, very much lighter, almost a reddish-white, below. Holotype, Western Australian Museum, No. 4920, dredged bj' the "En- deavour" off Geraldton in 29 fms. There is a single paratype of this strongly marked species taken at the same time and place as the one described above. It is somewhat smaller (R = 83 mm.) and there are fewer aboral tubercles, but there are a good many more granules on the interbrachial superomarginal plates. Anthenea polygnatha' sp. nov. Plate 18, fig. 3. Plate 19, figs. 2-3 R = 125 mm.; r = 60 mm. R = 2 r or more; br at 60 mm. from mouth, only 40 mm. These are the measurements of the holotype but this Anthenea seems to be dimorphic and a specimen of the other form has R = 115 mm.; r = 60 mm.; hence R is a little less than 2 r; br at 60 nun. from the mouth is 50 mm. or more. In the long armed form, the interbrachial arc has a nearly straight margin 40-50 mm. long while in the other, the straight portion of the interbrachial arc is only 20-25 mm. and even then it is not really straight but slightly concave. The 3 smaller specimens at hand, with R = 78, 93 and 97 mm. all have wide rays and curved interbrachial arcs. Of 2 larger specimens, one with R = 130 mm. is dis- tinctly narrow-armed while the other with R = 127 mm. is more of the broad- armed form, though it has interbrachial arcs that are nearly straight for 40 mm. It is possible that this difference in form is associated with sex and the matter deserves investigation whenever fresh material is available. Aside from this difference in form, the 7 specimens at hand show little diversity. The aboral surface is covered with hundreds of low spine-like tubercles, smallest in the small specimens, largest in the holotype. There is more or less evident a tendency to have these tubercles form radiating parallel series running out from the center of the disk. On the arms these series are usually quite evident, more so in the smaller specimens than in tlie larger. In all the specimens, the tubercles along the sides of disk and arms (i.e. just above the superomarginal ' TToXi/s = ma/ji/ + yi>ados = jaw, in reference to the very large number of pedicellariae on the oral surface. CLARK: AUSTRALIAN ECHINODERMS 129 plates) tend to be longer and more conspicuous than elsewhere, and this is a rather good distinguishing mark when specimens are mixed with conjungens, elegans and australiae. There is some individual diversity as to the degree to which the tubercles are pointed ; in some specimens they are quite sharp. Superomarginal plates small, in the interbrachial arc almost as long (or high) as wide, becoming larger distally but never forming a conspicuous part of the aboral surface. Their position ranges from nearly horizontal in one arc of the holotype which is a flat, slowly dried specimen, to perfectly vertical in most of the others. There is little doubt that in hfe, they are normally nearly or quite vertical in the interbrachial arcs, and are but little visible from above. They are more or less covered with granules ; on the lower half the granules are small and cover the plate well but on the upper portion the granules are larger and the uppermost is almost a tubercle, above and on each side of which the plate is quite bare; besides the granules each plate bears 1-5 pedicellariae of diverse sizes. Inferomarginal plates considerably larger than those of the upper series, completely covered with small granules and pedicellariae; the number of pedi- cellariae on each plate ranges from 4 up; in the holotype, many of the plates carry 15-20 pedicellariae. Actinal intermediate plates with numerous pedicellariae, as many as 4 occurring on some plates and rarely does a plate carry only a single large pedicellaria. Adambulacral armature in three series as usual but the spines of the second series are unusually long and flattened; even in the smallest speci- men this feature is noticeable. Color in life purple or violet or orange; as in the other species of Anthenea at Broome, there seems to be no constancy or dis- tinctiveness in the color. The dried specimens still show some traces of purple or orange, but in general they are "museum color." Holotype, M. C. Z. No. 3196, from Broome, Western Australia, 5-8 fms., June, 1932. Of the 6 paratypes, the 2 largest were presented to me in 1929 by Colonel W. 0. Mansbridge,,who informed me that they came from the Lacepede Islands, north of Broome. We did not meet with the species that year but in June of 1932, we took the holotype and the other paratypes in our dredging south and south- west of Roebuck Bay. The excessive development of pedicellariae, the long adambulacral spines and the increased length of the aboral tubercles above the superomarginal plates combine to make this species easy to recognize in spite of its dimorphic tendency. 130 memoik: museum of comparative zoology OREASTERIDAE Protoreaster nodulosus Pentaceros nodulostis Pehrier, 1876. Arch. Zool. Exp., 5, p. 53. Protoreaster nodulosus Doderlein, 1916. Zool. Jahrb.: Syst., 40, p. 420. Although hitherto considered a rare species, this is a common sea-star in the Broome region and specimens of very diverse size and almost equally diverse color were constantly taken during our dredging. It was also found at or above low water mark during the extreme tides of September, 1929. The smallest speci- men has R = 40 mm., r = 16, R = 2.5 r ; in the largest individual at hand R = 155- 160 nun., and r = 63 mm., so the proportions are essentially the same. In a per- fectly tetramerous specimen from the Lacepede Islands, R = 90 mni. and r = 30, hence R = 3 r, and in a hexamerous specimen also from the Lacepedes, R = 85 mm. and r = 35, hence R is not quite equal to 2.5 r. On the whole, the form, propor- tions and tuberculation show rather notable constancy. The two largest speci- mens however show interesting details of tuberculation; in the larger the 5 big hemispherical radial tubercles which are so distinctive a feature of nearly all individuals are practically wanting in two radii and in the other three are in- conspicuous, not nearly so large as the adjoining tubercle which begins the characteristic carinal series; in the other specimen, the large tubercles of the carinal series are very irregular in form, number and arrangement, the big radial tubercles are wanting or displaced, and there are 5 large interradial tubercles; on one arm the proximal tubercles of the carinal series are about as usual, on two, they are wanting and on two, they are abnormally numerous, of diverse sizes and crowded out of position. In this latter specimen, there are notable irregularities in the distal inferomarginal series. My field notes describe the color of a number of specimens but the diversity is so striking it has no significance ; in general it may be said that the aboral sur- face is usually brown, gray or green of some shade, with the papular areas a different shade from the plates and the papulae dark; the large tubercles usually stand out in a distinctive shade or color; in one extreme specimen they were purphsh-black with the rest of the aboral surface cream-white. The oral surface is more constant ; it is fundamentally cream color but shades of the aboral side pass over onto it more or less by way of the interbrachial inferomarginals, and in highly colored specimens the central portion of each actinal intermediate plate may be quite evidently tinted with some shade of color. In all specimens, of all CLARK: AUSTRALIAN ECHINODERMS 131 sizes, the adambulacral armature is white or at least cream-color, and the pedicels are white (more or less translucent, of course) with the terminal suckers bright hght violet. It is curious how constant this color of the pedicels is. There are 25 specimens of nodulosus at hand, of which the 2 non-pentamerous specimens from the Lacepede Islands, mentioned above were gifts from Colonel W. O. Mansbridge; 5 adults were taken at False Cape Bossut in September, 1929; 2 large adults were collected at Augustus Island by Captain Bardwell in October, 1933, and the remainder, including half a dozen very young specimens, were taken at or near Broome, in 1929 or in June, 1932. Pentaceraster australis Oreaster australis LttxKEN, 1871. Vid. Med., 23, p. 253. Pentaceraster australis Doderlein, 1916. Zool. Jahrb.: Syst., 40, p. 433. Mr. Melbourne Ward has kindly sent a small Oreasterid with R = 42 mm. which seems to be the young of this species, common at several points on the Queei^sland coast. Li\'ingstone's (1932) careful description of the growth stages and his admirable figures leave no doubt as to the identification, but I have not hitherto seen so small a specimen. Pentaceraster gracilis Oreaster gracilis Lutken, 1871. Vid. Med., 23, p. 260. Pentaceraster gracilis Doderlein, 1916. Zool. Jahrb. : Syst., 40, p. 437. In 1929, Colonel W. O. Mansbridge of Broome, presented me with an unmis- takable specimen of this fine species taken at the Lacepede Islands, north of Broome. One of the very few disappointments which we met with in our collect- ing that year was the failure to find even one specimen of this rare and much- desired form. The disappointment was deepened in 1932 when a month of very extended and intensive dredging failed to reveal a specimen. The few specimens hitherto known have come from the coast of Queensland. There are none in the AustraUan Museum and only 3 in the M. C. Z. ; one of the.se, a small one with R = 133 mm., is from an unknown locaUty but the 2 large ones, with R = 200-235 mm., are from Warrior Reef, in Torres Strait, where they were secured in 1896 by Mr. Henry A. Ward. The specimen at hand from the Lacepede Islands has R = 220 mm., r = 100 mm. The color of the dry specimen is a dull purplish- brown, which suggests that the color in life was a brick-red. 132 memoir: museum of comparative zoology culcita novaeguineae MiJLLER and Troschel, 1842. Sys. Ast., p. 38. Among tlie valuable discoveries which Captain Bardwell made during his collecting at Augustus and Champagay Islands, none surprised nie more than that of Culcita of which he sent 5 specimens. They are so similar to those in the M. C. Z. from Torres Strait, that the identification seems beyond question but the extension of range is very remarkable, as I can find no records for Culcita nearer than Torres Strait, 1200 miles or more to the east. The specimens at hand, in their dry and flattened condition range from R = 70 to R = 115 mm. All are markedly pentagonal and all but one have the sides nearly straight; in one, the concavity of two sides is as much as 15-20 mm. In the smallest specimen the marginal plates are very conspicuous, but in none of the others are they visible; there are 12 on each side of the pentagon in the upper series and 14 or possibly 16 in the lower; the inferomarginals near the tip of the arm are ill-defined. This smallest specimen is quite brown but the others are more definitely gray; all have more or less variegation or tinting with yellowish or brownish. OPHIDIASTERIDAE AUSTROFROMIA POLYPORA Fromia polypora H. L. Clark, 1916. "Endeavour" Ech., p. 51. Ausfrofromia polypora H. L. Clark, 1921. Torres Strait Ech., p. 48. Mr. Glauert has kindly loaned me a small specimen of this still little known species, taken at Rottnest Island, and belonging to the Western Australian Mu- seum. It is considerably distorted but R is about 65 mm. and r is almost one- fifth as much; at base, the arms are 12 mm. in diameter but they taper rather rapidly to a blunt tip about 5 mm. across. The specimen is now museum color with a faint reddish tinge but there is no indication of what the color was in life. Nardoa pauciforis Linckia pauciforis von Martens, 1866. Arch. f. Naturg. 32, pt. 1, p. 69. Nardoa pauciforis Sladen, 18S9. "Challenger" Ast., p. 412. The range of this species is extended far to the southward by its discovery near Mackay, Queensland. Mr. Ward has sent a very typical specimen with CLARK: AUSTRALIAN ECHINODERMS 133 R = 110-115 mm., bearing the label "Coral reef, Seaforth Island near Lindeman Island, Cumberland Group, Queensland, July, 1935." LiNCKIA GUILDINGII Gray, 1S40. Ann. Mag. Nat. Hist., 6, p. 285. The discovery of this tropicopolitan sea-star at Lord Howe Island was not very surprising but to find it among the species occurring at Broome was most unexpected. Still more interesting is the color in life of the specimens taken at Broome. The first specimens taken were brought up by our diver, Norman Whitworth, from 4-5 fms. in Lagrange Bay; there were 3 lying exposed on the sandy bottom and Whitworth said he saw no others, yet oddly enough one has 4, one has 5 and one has 6 arms. The 4-armed one has R = 195 mm. on the longest ray but only 155 on two others and the fourth was bitten (or broken) off close to the disk but was healed and beginning to regenerate; the arms are 15-18 mm. in diameter near base and taper evenly to the tip; there are 3 madreporites, sub- equal and normal, 1 in an interradial area on the disk, the other 2 in the adjoining interradius, on the side of the base of a ray. The 5-armed individual is more un- symmetrical in arm-length yet it has only a single madreporite. Beginning to the right of the madreporite, the length of R is 175, 165, 178, 102 and 84 mm. and the diameter of the arms near base is 13-14 mm. The 6-armed specimen has 2 madre- porites; the ray between them has R = 100 mm.; the following rays (to the right) have R = 165, 185, 125, 155 and 180 mm. In color these 3 specimens were ahke; they were "very blue, with a greenish cast," but distally the arms had more of a violet tinge. The blue was not so deep a shade as in L. laevigata. On drying, the color became light blue but soon changed to gray with a violet tinge which becomes more or less red-violet distally and orally; there are also distinct red- violet patches, 1-5 mm. across, on the oral surface. There is no record of blue Linckias in the West Indies. Off Cape Vilaret, in about 5 fms., Whitworth brought up a young guildingii, which was violet, not at all blue; it dried reddish but soon became violet-buff, more nearly violet on disk. A small adult specimen (R = 100 mm.) in the M. C. Z. from Moraine Cay, Bahama Islands, is very distinctly violet, even in its present dry condition and must have been much the same color in life as this Cape Vilaret specimen. The latter has 2 small madreporites, one almost rudimentary, and the rays measure 87-112 mm. in length, 6.5-8 mm. in diameter. Mr. Bourne gave me another specimen of guildingii, which he collected near Broome and dried ; 134 memoir: museum of comparative zoology it is much distorted, but has a single small madreporite, and 5 arms, of unequal length, 2 bitten (or broken) off near the disk; the most perfect arm is about 170 mm. long and 17 in diameter; the color of this specimen is a light brown, similar to that of the West Indian specimens at hand. Unfortunately Mr. Bourne did not recall the color in life. Another specimen from Broome has been loaned by the Perth Museum; it is now a rather bright yellow brown especially on the oral side; there are 2 madreporites with a ray between them, and 6 arms, 70-170 mm. long. The single specimen from Lord Howe Island is rather small but quite sym- metrical, with 5 arms, and 1 madreporite ; R = 90-95 mm. It was given me by Mr. Ivor Maidment who picked it up on Middle Beach; he reports the colors as "brown above, almost red below, along furrows." It is now however in poor con- dition and quite bleached, having been obviously exposed to rain and sun. In the Australian Museum tliere is a fine 6-rayed specimen from Lord Howe, having R = 100 mm., more or less. BUNASTER VARIEGATUS' Sp. UOV. Plate 22, fig. 1 Rays 5. R = 23 mm., r = 6, br = 6, hence R nearly = 4 r or 4 br. Aboral surface well covered by the superomarginal plates of which there are about 16 or 17 in each series, and a single carinal row of 16 or 17 similar plates; the superomarginal plates increase in size distally, so that the most distal are considerable larger than the ones near disk; this increase in size distally is not so evident in the carinal plates; on the basal half of the arm, there is, on each side of the carinal plates, a series of 10-12 very small plates; all of the aboral plates lie close together and the narrow areas between are closely covered with minute granules like scales, set more or less on edge. No pedicellariae, or "ball and socket" plates, have been detected on the holotype or any other specimen either aborally or orally although prolonged search with magnifications up to 90 diameters has been made. Papulae are present singly on the arms in most of the angles between superomarginal and carinal plates. Terminal plate large, nearly circular, swollen. Inferomargmal plates 18 or 19 in number, largest near middle of arm, smallest basally, the distal ones more or less circular. Intermarginal plates 8 or 9 extending out to middle of arm. Intermarginal papulae relatively large and conspicuous, at least distal to intermarginal plates. ' variegatus, in reference to the diversified coloration. CLARK : AUSTRALIAN ECHINODERMS 135 Actinal intermediate plates in two series but the outer only extends to about the tenth inferomarginal. There are 12-15 large single papulae, in a well spaced series, below the inferomarginals. Adambulacral armature as in other Bunasters, 2 furrow spines, short, rounded, subequal, on each adambulacral plate and a large subambulacral spine, almost as wide as long, so the series is quite crowded. Oral plates with 3 marginal spines but none on surface. Color of disk and tips of arms rose-red (purplish-red or dull brown m para types), base of arms cream-white with more or less brown variegation; remainder of arm variegated with brown, reddish and white; oral surface whitish more or less variegated with brown. There are 5 specimens which show this rather pretty coloring. The other 8 are bleached, to a uniform pale yellow-brown in .3 cases but more nearly white in the others. Apparently the variegated specimens were dried either without being placed in alcohol or before it had affected the color perceptibly; the others were probably bleached in alcohol and later dried. Holotype, M. C. Z. No. 3214, from Bunkers Bay, Western AustraUa, in a shallow intertidal pool. E. W. Bennett leg. et don. This Bunaster is almost exactly the same size as the type of B. lithodes Fisher, with which it has been critically compared. The differences are obvious: the wide (almost circular) subambulacral spines in a crowded series are very different from the narrow, well-spaced spines of lithodes; the outer actinolateral series of plates is much shorter and the individual plates smaller than in the Philippine species; the granulated areas between the skeletal plates are much narrower and less extensive in the Australian species, the papulae are single and there are no pedicellariae ; the color is notably different, more rose or purple in variegatus, but whether this difference is of any value, only large series of speci- mens can tell. In the Ught of the material now at hand, there is no doubt that the Bunasters taken by Professor W. J. Dakin at the Abrolhos Islands, which were recorded as lithodes (H. L. Clark, 1923, p. 241) are really immature variegatus. Comparison of one of them with a variegatus of similar size indicates this clearly. Besides the holotype there are 12 specimens of Bunaster in the present collec- tion, all of which, except the very young, with R less than 12 mm. may be con- sidered paratypes. They were taken as follows : Western AustraUa: Rottnest Island, near Bathurst Point, December, 1929 and January 1930. Miss Glauert and L. Glauert leg. 7 speci- mens, 1 adult and 6 young (2 in natural colors appar- ently). 136 memoir: museum of comparative zoology Same locality but taken in 1931. 4 specimens, young, 1 in natural colors, apparently. Bunkers Bay, "shallow intertidal pools with loose rocks and a little muddy sand; small clusters of a short brown alga in pools." 1 adult specimen (holotype) and 1 very simi- lar but not so large. Leiaster leachii Plate 9 Ophidiaster leachii Gray, 1840. Ann. Mag. Nat. Hist., 6, p. 284. Lciasfcr leachii de Loriol, 188.5. Mem. Soc. Phy.s. Hist. Nat. Geneve, 29, No. 4, p. 40. On April 21, 1932, one of our last days at Lord Howe Island, Mr. Robert Baxter, whose knowledge and ready assistance had been invaluable during our stay, brought in a superb sea-star, with R =275 mm. (diameter of arms less than 20 mm.), which he had "after half an hour's work, dug out of the coral rock near shore on the South Reef." The color was striking, a brownish-buff, irregularly blotched and spotted with crimson; the big madreporite was red-violet. It was new to him as well as to us, though there was no doubt of its being a Leiaster. A few days later, in looking over sea-stars in the Australian Museum, a Leiaster from Lord Howe was seen which appeared to be L. speciosus v. Mart. Comparison of my Lord Howe specimen with the specimen of speciosus in the M. C. Z., secured at the Murray Islands in 1913 (See H. L. Clark, 1921, p. 74) raised the question as to the validity of von Marten's species. Specimens of Leiaster are unfortunately very rare (there are only 3 in the M. C. Z.) but after consideration of those I have seen in life and in museums, and of the published descriptions and figures, I must confess I can find no difference that is trustworthy between speciosus and leachii, and the latter name unfortunately has priority. Owing to the length and somewhat flaccid character of the arms, Leiaster leachii undergoes considerable change of form when preserved; the shrinkage of the Lord Howe specimen in drying is about 10% in length of arms with no corresponding decrease in diameter; tlie arms are thus shorter and stouter than in life. A few small pedicellariae are present on the distal part of the arms. The color is now a uniform dark and dingy Iirowu-red becoming a light red, reddish- CLARK: AUSTRALIAN ECHINODERMS 137 yellow and finally dull yellowish distally. But the specimen was dried hastily and under adverse conditions, as we were packing up to leave the island. Mr. Livingstone has written me that during the Christmas holidays of 1932 he took "another beautiful Leiaster" at Lord Howe, with R = 226 mm. The color in life was "surface covered with blotches of orange, magenta and crimson, the last two colours predominating. The colouring is extremely vivid and most strikingly beautiful." It is interesting that this specimen has the same type of coloration as that of the specimen taken in April, whereas the Murray Islands specimen was uniform crimson. One can but wonder whether the unicolored and particolored forms are really one species! Ophidiaster armatus KoEHLER, 1910a. Al)h. Senckenb. Nat. Ges., 33, p. 277. Among the sea-stars sent me from Lindeman Island, near Mackay, Queens- land, by Mr. Melbourne Ward in 1934, are two specimens which after much hesitation, I refer, for the present, to this species. The large one has R = 66 mm. and hence is a little larger than the largest of Koehler's types. The breadth of the arm at base is nearly 9 mm., so the proportions are very similar to those of armatus. But the arms are more slender and tapering and the groups of papulae are much less noticeable than in Koehler's figures. However the specimen in hand is dry, while Koehler's photograph is obviously of a specimen not yet dried. No pedicellariae were found in the types of armatus; there are many in the Linde- man Island specimen. The color of armatus was deep yellowish-violet with 2 or 3 faint, irregular yellowish bands on the distal part of the arms. The present dry specimen is an almost uniform deep brown-violet with the faintest possible indication of banding on the arms, visible only in particularly good light. The smaller specimen from Lindeman Island has R = 38 mm. It is a dull light brown, more or less evidently variegated with a darker more purple shade; on the arms these darker blotches might be considered as indefinite bands; oral surface somewhat lighter. The series of papular groups on the oral surface of base of arm, which are very evident in the larger specimen, can scarcely be detected in this small one, which is thus much like a Tamaria. The projecting pointed tubercles on the marginal, and some aboral plates at the tip of the arm are evident enough but are not at all conspicuous. It seems better to call these 138 memoir: museum of comparative zoology specimens armahis than to attempt to differentiate them as a distinct species. The specimen in the M. C. Z. from the CaroUne Islands which I (1921, p. 82) referred to armatus is certainly not this species. It is too young for satisfactory identification but is probably granifer Ltk. Ophidiaster confertus Plate 10, figs. 2, 3 H. L. Clark, 1916. "Endeavour" Rep., p. 53. This is certainly one of the characteristic reef animals at Lord Howe, even though it occurs also (so Mr. Livingstone tells me) on the New South Wales coast. It was the first sea-star that collecting on the South Reef flat near Mt. Lidgbird yielded and subsequent collecting showed it was common all over that extended area. It also occurred at Neds Beach, indicating that it is distributed all around the island. It lives more or less exposed among coral heads and rock fragments so it is quite conspicuous. There are 25 specimens in the series at hand; they fall naturally into two groups of 6 young ones and 19 adults. The young ones have R = 10-23 mm. In the smallest, there is a single papula at each point where any four of the adam- bulacral, marginal and aboral plates meet each other; hence there are 8 longi- tudinal series of 9-11 very distinct single papulae along the arm. In the next larger specimen (R = 12 mm.) there are 2 papulae at each point, except basally distally and orally; in a specimen with R = 13 mm., there are 3 papulae at some points. A specimen with R = 17 mm. is at this stage but the one with R = 23 mm. has 4-7 papulae in the aboral groups and 3 in those on the oral surface. In color, the young specimens under R = 20 mm. are very light, little pigment having developed but the one with R = 23 mm. is light orange brown. The adults range from R = 58 to R = 155-160 mm. There is some diversity in the stoutness of the arms : one specimen has R = 97 and br = 12, hence R = 8 br, and the largest specimen has br = 18, hence R is about 8.5 br, but another speci- men with R = 100 mm. or less has br 15 mm. or more, so that R is only about 6 br. It might be mentioned here that in very young specimens R = 4 br. In coloration, confertus is never mottled or variegated, but unicolor, save for the tendency to a lighter tint on the oral side near the ambulacral furrow. Most individuals are a deep tawny yellow or orange brown, but some are lighter. CLARK: AUSTRALIAN ECHINODERMS 139 ranging to a dull yellow orange; or darker, almost a real brown. In the dried specimens the range is from a rather dingy wood-brown to a very distinct (almost chrome) yellow. Alcoholic specimens are a light wood-brown. Hacelia helicosticha Ophidiaster helicostichus Sladen, 1889. "Challenger" Ast., p. 405. Hacelia helicosticha H. L. Clark, 1909. Bull. M. C. Z., 52, p. 111. A very large example of this fine species, with R = 150-155 mm., is at hand, loaned by the Australian Museum. It is a perplexing specimen because the actinal papulae are poorly developed and only here and there at isolated spots can one count ten (or even nine!) longitudinal series of papulae. But the shape of the rays, associated with the additional actinolateral plates, leaves no doubt of the genus. The specimen bears only the label J69, but Mr. Livingstone tells me it is from northwestern Australia and was collected by Mr. F. J. CJibbons. Tamaria fusca Gray, 1840. Ann. Mag. Nat. Hist., 6, p. 28.3. A single very young ophidiasterid is at hand from the Great Barrier Reef Expedition St. XVI, ^ 2 niile west of North Direction Island, 20 fms., stony bot- tom. It has been identified by Livingstone as representing Gray's species and I see no reason for questioning the fact. It must be added however that ophi- diasterids as small as this (R = 10-11 mm.) cannot be identified with certainty unless associated with a series of larger specimens. Tamaria megaloplax Linckia megaloplax Bell, 1884. "Alert" Ech., p. 126. The untangUng of the snarl in which this species was enmeshed seems to have been accomplished satisfactorily by Livingstone, though I cannot accept all of his conclusions as to the synonymy. The growth changes in this species are more extraordinary than in any other sea-star, except perhaps Culcita or some of the multiradiate forms. Young individuals of megaloplax, having R 140 memoir: museum of comparative zoology less than 45 mm. are conspicuously prickly, most of the marginals and many aboral plates being capped with a pointed tubercle or spine. As growth proceeds, these tubercles are more and more resorbed. But the rate of resorption differs greatly in different individuals; in one specimen at hand from Broome, given me by Mr. R. A. Bourne, with R only G2 mm., the process is complete — not a tubercle is left; in another specimen, from the Lacepede Islands, given me by Colonel W. O. Mansbridge, with R =70 mm. many tubercles on the aboral plates, and most of those on the marginals are still evident though greatly reduced; in an adult specimen, with R rather more than 100 mm., which we took at Broome in June, 1932, the tubercles are mostly resorbed but there are some still present on the inferomarginals. Diversity in color is also rather considerable. Young specimens seem to be variegated light and dark browns, with the oral surface light. Adult specimens are very handsome — cream color orally and variegated aborally with pale gray and brown of at least two well-marked shades (See H. L. Clark, 1921, pi. 8, fig. 1, as T. tuberifera). Besides the three specimens mentioned above a very small Ophidiasterid, with R = 12 mm., from Lindeman Island, sent by Mr. Ward, may possibly repre- sent this species. It is obviously different from the young fusca listed above, as the skeletal plates each bear one or more large granules, which indicate that each plate will soon be provided with such a tubercle as characterizes megalo'plax. On one or two points, I have to differ from my friend and colleague, Living- stone, in regard to this species. He says (1932, June, p. 3G9) of the specimen with R=58 mm., from Holothuria Bank, which he figures: "It is obviously the only specimen of megaloplax from Holothuria Bank handled by Bell." A speci- men in the M. C. Z. received from the British Museum is one of Bell's original specimens and was labelled by Bell himself as "Linckia megaloplax." Livingstone has examined this specimen but says of it (1932, Feb., p. 260) that he thinks it should be referred to Koehler's species hirsuta and he figures it under that name. After long and careful examination of the available material and Livingstone's figures, I am convinced that hirsuta Koehler is based on a young specimen of megaloplax; or at least, specimens from northwestern Australia, which answer to Koehler's description and figures are undoubtedly to be referred to megaloplax and not to an Andaman Island species. The only known Andaman Island speci- men, Koehler's type of hirsuta, is of course immature; it is possible that when adults are taken they will prove to be different from megaloplax. Meantime I include hirsuta in the synonymy of Bell's species. CLARK : AUSTRALIAN ECHINODERMS 141 Tamaria TUMESCENS Ophidiaster tumescens Koehler, 1910a. Abh. Senckenb. Nat. Ges., 33, p. 277. Tamaria tumescens H. L. Clark, 1921. Torres Strait Ech., p. 94. Tamaria propetumescens Livingstone, 1932, .June, Rec. Austr. Mus., 18, p. 369. Among the dry sea-stars from the Lacepede Islands which Colonel Mans- bridge so generously gave me at Broome in 1929 was a Tamaria, unlike any species known to me, but we failed to take any living specimens in all our col- lecting. In 1932, Mr. Bourne gave me a small specimen of the same thing, with all the arms more or less missing. On June 16, 1932 in Pender Bay, Wan, our diver, brought up a living individual and afterwards in our dredging, north and south of Roebuck Bay, we secured a number of specimens. In the Perth Museum, is a fine specimen from Broome, kindly loaned me by Mr. Glauert; it has R =95 mm. but has lost all trace of its natural color and is now a Hght brown. In life, tumescens is a very lovely sea-star, bright old-rose color with the papular areas Ught gray brown and the sides of the ambulacral furrows yellowish. This series of 15 specimens shows conclusively that Livingstone's prope- tumescens is but a form of Koehler's species; each was based on a single small specimen, half grown or less. The Broome Tamarias range from R= about 40 (the arm tips are missing) to R = 98 mm. In this large adult, r and br each = about 15 mm.; hence R=6.5 r or br as in Livingstone's specimen; in the text Koehler gives R = 7.5-8 r but his figures indicate R about equals 6 r. The Broome specimens show little diversity R ranging from 6 to 7 r. There is more diversity in the presence or absence of spines on the inferomarginals and the extent of the adradial series of plates, while the arrangement of the aboral plates, especially on the disk, is very diversified. Usually inferomarginal spines are wanting and the adradial series are short, but the specimen from the Lacepede Islands has many spines and long adradial series, and a Broome specimen with R =87 mm. has numerous inferomarginal spines, very irregularly distributed, and the adra- dial plates extend to the middle of the arm. So far as the present series shows there is no correlation between the presence of spines and the size of the specimen. This is one of the most pleasing sea-stars in the Broome fauna, and dried specimens may retain more or less of their rose color for years provided they have never been in alcohol. It lives exposed on the sandy sea-bottom and like most ophidiasterids, it is very rigid and inert. 142 memoir: musexjm of comparative zoology pseudophidiaster rhy8us H. L. Clark, 1916. "Endeavour" Rep., p. 55. Professor Bennett has sent a specimen of this sea-star taken Feb. 23, 1930, in the Great Austrahan Bight, by Mr. D. L. Serventy, on the trawler "Bou- thorpe," in 90 fms., lat. 33° 15' 0" S., long. 126° 22' 15" E. According to Mr. Serventy's notes it was "dark purple" in Ufe. It is now a dull gray, quite different from the fawn color of the types. It is a rather small specimen with R only about 100-110 mm. ASTEROPIDAE Petricia obesa Plate 10, fig. 1 H. L. Clark, 192.3. .Jour. Linn. Soe.: Zool., 35, p. 241. This remarkable sea-star is one of the most strikingly colored of echinoderms. When Uving it is brilliant scarlet with the small madreporite and the papulae almost black. The big interradial pedicellariae apparently close tightly when the animal is captured and being completely buried in the thick skin, do not show at all. On being dried, after killing in formalin and corrosive-subUmate, a large specimen (R = 70 mm.) kept its color very well and now, six years later, is a brilUant yellowish red. A smaller specimen (R = 54 mm.) is now a reddish yellow. The difference is apparently due to the much thinner skin in the smaller in- dividual. A third specimen (R = 75 mm.) is an orange-brown, but this one was not preserved until several days after collection and then was placed in methy- lated spirits and subsequently dried. The largest and smallest specimen each have the 10 conspicuous interradial pedicellariae on the dorsal side, but the third specimen has only 9; none of these specimens show any other pedicellariae any- where. The largest specimen was taken at Bunkers Bay, in January, 1930, and was generously sent me by Professor Bennett. The other two were taken by Professor Bennett and myself at Point Peron, October 23, 1929. The larger one, lying on the bottom in G-8 ft. of water, by its brilUant color attracted our attention from a low bluff overlooking some reefs and rocks a few yards from shore. Mr. Bennett kindly dove for it and presented it to me — permanent trophy of a never-to-be- CLARK: AUSTRALIAN ECHINODERMS 143 forgotten day! Never having seen a living specimen of the eastern species, vernicina, and having handled but very few museum specimens, I do not know how important the differences between the two species really are. So far as present material is concerned, they are quite distinct. ASTERINIDAE ASTERINA ANOMALA H. L. Clark, 1921. Torres Strait Ech., p. 95. It was a pleasant surprise to find this odd little sea-star, known hitherto only from a few specimens taken at Mer, at the northern end of the Barrier Reef, very common at Lord Howe Island. A series of 369 specimens is at hand, regarding which certain statistics may be of interest and possibly, value. As regards size, the species is always small, with R = 10-12.5 mm. at the most. As for form, it is rare indeed to find a symmetrical specimen, so persistently autotomous are these little sea-stars. One 4-rayed individual with R = 12.5 mm. seems to have attained stabihty as it is approximately symmetrical. Of individuals with more than 4 rays, not a perfectly symmetrical one has been noted though several ap- proach it having two groups of equal rays; but the difference between the two groups is not hard to see. Normal mature rays are narrower and higher in life than in preserved material which tends to become quite flat. The number of raj^s ranges from 2 to 8, but curiously enough, excepting only 2, which of course indi- cates a fragmentary specimen, 5 is the least common number and is really very infrequent, while 7 is the most common and is found in a few more than half the 369. The average number of rays in the whole lot is almost exactly 7, if the 4 2-rayed fragments are omitted. One-fourth have 8 rays, but 10% have 6, 8% only 4, and almost 5% have only 3. Only 5 specimens are pentamerous and one of these has a ray spht so that there are apparently 6 rays but only 5 jaws. As a rule a madreporite is wanting but one is not rarely found and two or three may be present. Truly the name anomala is appropriate for this Asterina. In color, anomala is rather constant (See H. L. Clark, 1921, pi. 7, fig. 8) but some individuals are darker than others and a few show small deep red blotches aborally. At Cape Leveque, northwestern Austraha, where two specimens were taken in August, 1929, the red was predominant, covering the disk and base of 4 normal rays; the median part of these rays was deep green and the margins 144 memoir: museum of comparative zoology were rusty, as were the 2 or 3 young rays growing out from one side. Preserved material ranges from pure white to a rather deep brown. Like other Asterinas this species usually occurs on the under side of rock fragments and in hollows and crevices in dead coral or rock. But at Neds Beach, Lord Howe, on April 6, 1932, many were found in coralline algae on the upper surface of stones, cHnging awkwardly among the branches, a habitat so ill- adapted for a flat sea-star like Asterina one suspects that only abnormal condi- tions of some sort led to it. This little sea-star has a wide distribution on the coast of tropical Australia. It was originally taken at Mer at the northern end of the Barrier Reef; as it proves to be very common at Lord Howe, it is probable that it is to be found anywhere between those two points where local conditions permit, but of course its small size, protective coloring and secretive habits cause it to be easily overlooked. The 382 specimens in the present series are from the following points : Lord Howe Island: April, 1932. 369 specimens, adult and young. AustraUan Museum coll. 2 adults, 1 with 7, 1 with 8 rays. Northern Territory : Darwin, dredged near Shell Island, 4-5 f ms. 5 small adults, 1 with 4 rays, 3 with 7, 1 with S. Color, variegated green and greenish- white, with no red or rust-color. Western AustraUa: Cape Leveque, August, 1929. 2 small adults, 1 with 6, 1 with 7 rays. Broome, August, 1929. 2 small 7-rayed adults so recently divided that only 3 or 4 arms are at all evident. Broome, June 1932. 2 young specimens, 1 with 6, 1 with 7 rays. Asterina burtonii Gray, 1840. Ann. Mag. Nat. Hist., 6, p. 289. It is rather remarkable that we met with this widespread East Indian and East African species but once and that was not until June, 1932, when a single small specimen (R = ll mm.), bright orange- vermilion above and below was dredged near the entrance to Roebuck Bay in 5-8 fms. Its rays are long and narrow, both r and br equalling but 5 mm. Another specimen is at hand, loaned by the Western Australian Museum. It has R = 17 mm. and was taken at Bernier Island, Shark Bay, W. A. \^^lile somewhat stouter than the Broome specimen, CLARK: AUSTRALIAN ECHINODERMS 145 it is not essentially different. The species has furthermore been recorded from the Abrolhos Islands (H. L. Clark, 1923, p. 243), so that the occurrence in Shark Bay is quite to be expected. The only other record for the Australian Coast which seems vahd is that from the Murray Islands, at the northern end of the Great Barrier Reef (H. L. Clark, 1921, p. 96). ASTERINA CORONATA FASCICULARIS Plate 12, fig. 1 Fisher, 1918. Ann. Mag. Nat. Hist. (9), 2, p. 110. The discovery that this is the common sea-star of Darwin and the neighbor- ing coast was one of the most interesting results of our collecting there in 1929. It was the first sea-star found and proved to be one of the very few echinoderms that occurred along the shore near town. Later collecting showed that it was of general occurrence along the coast but was most common at East Point. It seems to prefer very shallow water for all of the few specimens dredged were small. Tide pools left with little or no water at low tides were its favorite habitat. It lives, like other Asterinas, on the underside of rock fragments and hence wherever such fragments were exposed at low tide, coronata was a possible "find." While the number of rays is almost always 5, in the present series of 108 speci- mens, there are two with but 4 rays, two with 6 and one with 7. None of these specimens shows any trace of autotomy and it is evident that asexual reproduc- tion does not normally occur in this species. The 4-rayed specimens are symmet- rically tetramerous; in one R = ll mm., in the other 23. The smaller 6-rayed individual is perfectly symmetrical with R = 12 mm. and no madreporite present, but the larger specimen is not symmetrical for while five rays are about 20 mm. long, the sbcth is only 17, and there are 2 madreporites, each in an interradius, but not near the extra ray. The 7-rayed specimen is small and not noticeably unsymmetrical but R ranges from 8 to 9 mm. and when seen from the oral side, one ray is evidently shorter and narrower than the rest. Compared with a normal pentamerous specimen of similar size, the rays are rather wider and more rounded distally but the difference is not marked. There is only one madreporite and that is small and squarely in a radius. In size, the present series ranges from R =3.5 to R =33 mm. The ratio of R to r is commonly about 2 to 1. In the 7-rayed specimen the ratio is only about 1.5 to 1 but in the larger 4-rayed individual it is 2.5 to 1. Of course the exact 146 memoir: museum of comparative zoology ratio is affected in the preparation of specimens; interradii shrink more in drying than do the radii; hence in most large pentamerous specimens R =2.25r or there- abouts. Very young specimens of course have rather shorter rays with R about equal to 1.6r. There is the gi-eatest diversity in the number of elevated spiniferous aboral plates. In one specimen with R=21 mm. there are no such plates conspicuous but on each ray several can be detected; they are however low and bear no large spinelets. A somewhat smaller specimen has 1-3 enlarged plates on each ray and on the disk, but they are small and low. At the other extreme is an individual with R =25 mm. having 30-45 conspicuous, elevated, spiniferous plates on each ray. Most of the full grown specimens with R exceeding 30 mm. have 20-35 such plates on each ray but in one individual with R =30 there are only 10-15. I am convinced that no specific or varietal character, on which reliance can be placed, is to be found in this noticeable feature taken by itself. The same is true as regards the oral and adambulacral spines; the number of adambulacral spines is typically 6-8, very commonly 7, but often 8 in full grown individuals; the oral spines are usually 8 on each plate but there may be as few as 6 and in some big individuals there are 9. As for the actinal intermediate spinelets there are commonly 4 or more on each plate, but occasionally a plate is seen with only 2 or 3 ; 2 is unusual, but 3 is not a rare number. This large series of specimens, with the diversity it shows in spinulation naturally throws some light on the varieties proposed by P'isher (1919, p. 413). As I have seen nothing I can call pedicellariae, the variety euerces is not affected by the Darwin material, which for the most part falls readily into variety fascicularis, as would be expected, since one of the types of that form is from Port Essington. But it seems clear that not much rehance can be placed on the number of adambulacral and oral spines, and certainly not on the number and con- spicuousness of the enlarged aboral plates. The questions therefore arise whether the variety cristata can be maintained and whether it and fascicularis are suf- ficiently constant to warrant separating them from typical coronata. The one feature which seems to warrant such a grouping is found in the armature of the actinal intermediate plates. In the groups of 3 or more, often 5 or 6, spinelets found in fascicularis, we have a very evident difference from the 2 or 3 spinelets of coronata and cristata. Between these two latter forms, the difference in the enlarged aboral plates is striking, in the few specimens at hand, but it may be that a large series of Japanese specimens will show as much diversity in this particular as do the specimens from Darwin. The two specimens at hand from CLARK: AUSTRALIAN ECHINODERMS 147 the Caroline Islands are alike in this feature but are rather distinctly unlike in the actinal intermediate spinelets; the type of cristata has 2, or often 3, rather short, blunt, diverging spinelets on each plate, while a somewhat smaller speci- men has 3, rarely 2 or 4, longer, acute, clustered spinelets, much as in fascicular is. A specimen from Zanzibar is like the Japanese one on the aboral surface but orally is peculiar in that several actinal intermediate plates just back of the orals carry straight "combs" of 5 or 6 slender acute spines, which become groups of 4 or 3 on the more distal plates. In a somewhat larger Japanese specimen such a comb, with 6 spinelets, is present on a single plate in one area. As a result of all these comparisons, I am led to conclude that the varieties named by Fisher may well be recognized until abundant material from Japan, the Carolines, and the East Indies shows that their distinguishing features are not reliable. Should such material show as great constancy in the actinal spinulation as is shown by the Darwin material, it will not be surprising li fascicularis proves to be a valid species, characteristic of tropical Australia, perhaps confined to the coast of the Northern Territory. The probability is rather great that the Zanzibar form will ultimately prove to be distinct — at least as a recognizable variety. It is curious that this sea-star should prove to be so common in the vicinity of Darwin and yet not be found at any great distance either east or west of that place. At Quail Island, about 35 miles to the west we found 7 specimens, agreeing in all details with those from East Point on the other side of Port Darwin. It was a great disappointment not to find even one specimen in Port Essington, a type locality, nor at any other point where we collected on the Coburg Peninsula. The species is not known from Torres Strait or on the Barrier Reef, nor has it yet been detected at any point on the coast of Western Australia. The colors in Hfe of this interesting little sea-star are attractive and somewhat varied, and the following extracts from my field notes are pubhshed, since pre- served material undergoes great change. AlcohoUc specimens become almost completely bleached to a dingy white, while dried material becomes either plain "museum color" or some shade quite foreign to the living tints. Several large specimens are now a dull blackish-brown aborally and one is very distinctly dull violet, a shade not noted in Uving specimens. "The normal coloration is mottled olive-greens, light and dark, with more or less dark dull red, usually in irregular blotches. One specimen was very largely bright rust-red, over most of the dorsal surface," and retained this color well, for a time, after drying. "Most specimens lose the green shades on preservation be- coming brown or gray." "Some specimens occur with no trace of green dorsally; 148 memoir: museum of comparative zoology these are more or less fawn-color mottled with brown and have a distinct red tinge." "Most specimens have red markings but the shade may be very deep; in a few cases it was replaced with black. A common feature is a blotch of car- mine at the base of each arm; in one specimen this was nearer vermihon." "Pig- ment is obviously very soluble in alcohol as the colors are dissolved out rapidly making the alcohol bright orange-red and leaving the specimens pale museum- color." The 108 specimens of fascicularis at hand were taken as follows: Northern Territory: Darwin, East Point, June and July, 1929. 87 specimens, adult and young. Darwin, near Shell Islands, 3-6 fms., July, 1929. 5 speci- mens, young, 1 hexamerous. Darwin, Talc Head, July, 1929. 2 specimens, adult and young. Darwin, West Point, June, 1929. 4 specimens, adult and young. Darwin, near laboratory, June 19, 1929. 2 specimens, young. Darwin, near Leper Station, 3-5 fms., May 25, 1932. 1 speci- men, young. Quail Island, tide pools, July 7-9, 1929. 7 specimens, adult and young. ASTERINA INOPINATA Livingstone, 193.3. Rec. Austral. Mus., 19, p. 3. This is a very well marked species, admirably described and figured by Liv- ingstone; no reference is however made to color. In a letter, Livingstone says that "the color varies shghtly. Sometimes the species is a uniform cream; some- times blotched here and there with unevenly placed patches of dark green." I think these notes must be based on specimens in alcohol for Professor W. J. Dakin said that the large specimen which he gave me was collected by himself at Long Reef and was "blue and white" in hfe. The specimens which I took at the same place a few days later showed the diversity of coloration to which Mr. Livingstone refers, but the ground color was a real white, not cream color, and when any pigment was present, as was usually the case, it was distinctly bluish and not green; in seme cases it was dusky but in others it was a definite, though CLAEK: AUSTRALIAN ECHINODERMS 149 rather dark, blue. Except for one individual from Tasmania, in the Australian Museum, all the known specimens of inopinata are from the coast of New South Wales, between the Tuggerah "Lakes" on the north and Shell Harbor on the south. On November 28, 1929, we had the good fortune to take 11 specimens, during low tide, from the under surface of rocks, on Long Reef. They ranged from 13 to 30 mm. in diameter, and the lesser radius was very nearly equal in life to the greater, but some individuals were very distinctly pentagonal, especially as soon as they had contracted. The blue and white coloration is so unusual that the species is unmistakable in life. The specimen given me by Dr. Dakin is the largest yet found as, even in its dried and somewhat contracted condition it is 35 mm. across; as even now R = 20 mm., the diameter in Ufe was fully 40 mm. None of the specimens at hand are small enough to show the remarkable interra- dial sUts described by Livingstone in very young individuals but some of the smallest specimens show a distinct interradial notch. The 13 specimens of inopinata before me were taken at the following points: New South Wales: Colloroy, Long Reef. W. J. Dakin leg. et don. 1 large adult. Colloroy, Long Reef, November 29, 1929. 11 specimens, adult and young. Shell Harbor. May 4, 1932. 1 small specimen. ASTERINA SCOBINATA Livingstone, 1933. Rec. Austral. Mus., 19, p. 1. When in Hobart in November, 1929. I was the recipient of many interesting echinoderms from Professor T. T. Flynn of the University and among them are 3 specimens of this interesting Asterina, so well described and figured by my AustraUan colleague. The smallest has R only about 7 mm. but it is not essen- tially different from the two larger specimens in which R = 14-15 mm. In one of these, the disk is notably high, the arms quite pointed and the color a gray-brown while in the other the whole aboral surface is somewhat flattened, the arms are much more rounded at the tips and the color is a very light yellow-brown. It is probable that these differences are wholly due to differences in preservation of the material. The small specimen was taken with the light-colored adult and is quite similar to it in form and tint. All 3 specimens show more or less clearly 5 pairs of remarkable, small, uncalcified spots on the actinal intermediate areas, which are 150 memoir: museum of comparative zoology apparently characteristic of the species, but their position and distinctness seem to admit of considerable diversity. Evidently this sea-star occurs widely distributed on the Tasmanian coast for the gray-brown specimen was taken by Professor Flynn at "low tide, Wyn- yard," northwestern Tasmania, while the others bear the label "Eagle Hawk- Neck, Tasmania" (which is on the southeastern coast), January 27, 1928, V Irwin Smith, collector. With some hesitation, 1 also refer to this species, 2 Asterinas which 1 col- lected in the Derwent estuary at Hobart, on that notable November 15, 1929. My field notes made at the time read: "Asterina sp.? — 2 specimens — Deep brown above, cream-color below in sharp contrast. Under rocks at tide line, with exigua but obviously much rarer." Unfortunately the exigencies of collect- ing on that trip prevented careful preservation of these specimens; they are ir- regularly contracted and flattened, and the spinelets of the oral surface are relaxed and prostrate quite unlike the condition of the other specimens. More- over the actinal non-calcified spots are very inconspicuous or wanting altogether in some areas and the aboral surface has the plates more numerous, more crescen- tic and more densely covered with minute spinelets. Hence these Hobart speci- mens have a different facies from the others and I have hesitated about their identification. But the more I have compared them with the other Tasmanian Asterinas in hand which seem to be scobinata without doubt, the more I am con- vinced they had best be called by the same name, as they fall within a moderate range of variation. In one specimen R = 13 mm. while the other is smaller; both have very broad rays (7-8 mm.) but this is in part, at least, due to their not very satisfactory preservation. Only further collecting can demonstrate whether their color in life as given above is characteristic of the species ; they are now ordinary "museum color." Asterina alba' sp. nov. Plate 22, fig. 7 Rays 5. R = 9 mm., r = 4.5-5 mm. R not quite 2r. Form distinctly stellate with the interbrachial arcs rather acute. Aboral plates distinctly imbricated radially, much less so or not at all interradially ; each plate carries 2 or more very minute spinelets, scarcely visible to the unaided eye; these spinelets are very ' albus = white, in reference to the color in Ufe. CLARK: AUSTRALIAN ECHINODERMS 151 short as well as slender and are usually in a single linear series of 2-7, but on a few of the larger plates, they may be in double series of about the same length ; supero- marginals squarish, with a central group of 3 or 4 of the very minute spinelets; inferomarginals, with projecting tufts of spinelets, form the ray margins as usual. Actinal intermediate plates relatively few; about 10 large ones, back of the oral plates, and 15 diagonal series (of 6 decreasing to 2) on each side of each ray; each large plate has 1 relatively stout spine (or sometimes 2) at its center while the remaining plates have usually 2 smaller spines. Adambulacral armature, a furrow series of 4 slender but blunt spines and a subambulacral series of 3 spines set obliquely on the plate; these may be subequal or the one nearest the furrow, or the middle one, distinctly largest, and bigger than any spine in the furrow series. Oral plates each with a marginal series of 7 or 8 close-set, slender spines, the innermost longest, and one large spine on the surface. Color in life, almost pure white, but the larger individuals show, under a lens, traces of orange-yellow and a few scattered patches of dusky or purple, on the aboral surface. Specimens in alcohol are whitish, but dried specimens range from pure white to yellow-brown. Holotype, M. C. Z. No. 3239 from under surface of a rock fragment, Neds Beach, Lord Howe Island, April, 1932. There are 69 paratypes, chiefly from Neds Beach and the southern reef-flat near Mt. Lidgbird, but several were found on the reef north of Goat Island. They occurred with anomala but were always easily distinguished by the color and the pentamerous symmetry. They range in size from R = 2.5 mm. up to the holo- type, which is much the largest specimen taken. One specimen, 10 mm. across, is hexamerous but, although not perfectly symmetrical, shows no evidence of autotomy. All the larger specimens show, under the leas, one or more non- calcified spots in the actinal intermediate areas, similar to those of A. scobinata. But in the Lord Howe species, these spots show great diversity in number and position, and are always very minute; in the holotjqje there seem to be a pair well separated from each other in each area some distance back of the oral plates, but in other specimens, they are wanting in one or more of the areas, often in all but one, and their position differs in different areas; sometimes 3 occur in one area, one or two of which are near the adambulacral plates. In most of the small specimens, these curious areas seem to be lacking. It is unhkely that their presence has any definite significance. This species is easily distinguished by its regular pentamerous but markedly radiate form, the absence of spinelets of any considerable size on the aboral sur- 152 memoir: museum of comparative zoology face, the adambulacral armature, and particularly by the single large spinelet on the surface of each oral plate. Asterina heteractis^ sp. nov. Plate 22, fig. 5 Rays 7, narrow, high, relatively long and well separated. R = 7 mm., r = 4 mm., R = 1.75r. Disk flat. Aboral plates, of irregular shape and only imperfectly imbricated. Each plate carries a small tuft of 3 or 4 low blunt spines, not at all thorny. Papulae relatively large, in 4 series on the basal half of each arm. Inferomarginal plates fairly conspicuous, each with a horizontal comb of 4-7 relatively long spinelets which may be a little thorny near tip. Madreporite exceedingly small with only 6-8 pores, at the base of a ray on one side of the median line. Actinal intermediate plates few in 3 series on each side of each ambulacral furrow ; only the innermost extends to the tip of the arm while the outermost is very short. Each plate carries 1-3 (usually 2) short, sharp spinelets; when 2 or 3 are present one is often distinctly larger than its fellow (or fellows). Adambula- cral armature in 2 series as usual; on the furrow margin is a group of 3 (or 4) short, blunt, relatively large spinelets and on the surface of the plate is a similar but somewhat smaller trio set at an angle to the furrow series. Oral plates small, each with 4 relatively large opaque marginal spines, the innermost not very much larger than the outer ones ; on the surface of each plate is a similar spine (or sometimes 2). Color in life, uniform salmon pink above, whitish below, with no indication of green or red anywhere; the dry specimen is almost uniformly yellowish- white. Holotype, M. C. Z. no. 3258, from the under surface of a rock fragment on sand above low water mark at Neds Beach, Lord Howe Island, April, 1932. Collected and donated by Miss Kama Birmingham. Besides the holotype, there are 3 other Asterinas from Lord Howe which belong to this species. One is a small, nearly white, 6-rayed specimen found on the under side of a rock fragment; as 3 rays have R=4-5 mm. and the op- posite 3 have R about 2 mm., it is evident that autotomy has taken place. The other 2 specimens were dredged in the lagoon near Goat Island; they showed no trace of green or red but were uniformly light orange-colored in life; that ' ertpos = other than usual + 6.ktls = a ray, in reference to the number of arms. CLARK: AUSTRALIAN ECHINODERMS 153 tint however was lost as soon as they were dried and they are now very pale brown or nearly white. Each has but 3 arms (R = 7 mm.) and is obviously the result of autotomy so recent that the growth of the new arms has not yet begun. That this little Asterina is closely related to anomala is evident, but it is apparently quite distinct. The most obvious difference is in the spinulation of the aboral plates; in anomala the spinelets are much more numerous, somewhat glassy and thorny at the tip; while in heteractis they are few, opaque and rather blunt without thorns. On the actinal plates of anomala are 3-5 thorny spinelets, while in heteractis, the spinelets are fewer and not thorny. Both on the adam- bulacral and oral plates there are more and larger spines in anomala than in heteractis, even in specimens of the same size. How much weight may be given to the marked difference in the form of the rays and in coloration, only much more material of heteractis can determine. Asterina lutea' sp. nov. Plate 12, fig. 2 Rays 5, broad and rounded, the interbrachial areas only slightly concave with a distinct notch in the interradial line. R = 20 mm., r= 16 mm. R = 1.25r. Disk moderately elevated, the vertical diameter being about 6 mm. Aboral surface covered with numerous imbricated plates, largest in the proximal in- terradii and least imbricated distally in the same areas; there are about 15-17 longitudinal series of plates covering each radius, with the papulae conspicuous between them; the large papular areas on each ray are rounded diamond-shape, 15-16 mm. long and 11-12 mm. wide. Aboral plates with more or lesss evident series of minute spinelets which curve about the lower or distal side of each pap- ula; on the interradial and lateral plates where there are no pores, these curved series of spinelets are reduced to minute tufts which disappear entirely on the insignificant superomarginal plates. Margin of body formed as usual in Asterina by inferomarginals which carry tufts of numerous, relatively long but exceed- ingly slender spinelets. Madreporic plate small, triangular, only 3 mm. from anus. Actinal intermediate areas with numerous plates arranged in about 10 series parallel to the ambulacral furrow; those adjoining the furrow are largest and 3 or 4 extend to the tip of the arm; the outer series become smaller and ' luteus = orange-colored, in reference to the color in life. 154 memoir: museum of comparative zoology shorter until the outermost is made up of only 3 or 4 plates. Each plate carries a group of spinelets, slender and sharp; on the larger plates, they form a very definite line or comb of 5-7, the middle ones distinctly longest, the outermost shortest ; on the small plates near the inferomarginals the spinelets form a minute tuft of 3 or 4. Adambulacral armature in two fanhke series of relatively long and very slender spines; the furrow series is made up of 7 (rarely 8) spines, while the subambulacral series contains 6 sUghtly swollen ones; the two series are remarkably ahke save for the evident, though slight difference in size. Oral plates conspicuous, the inner margin of each being elevated to form an evident ridge; marginal spines 9, long and slender, the innermost large and relatively stout, about half as large again as the distal ones which are similar to those of the furrow series. A slightly curved transverse series of 4 or 5 small spines occurs near the middle of each oral plate and distal to this is an insignificant group of 2 or 4 minute spinelets. Color in Ufe bright orange-vermilion, quite uniform over both upper and lower surfaces. Sometimes the color is not so bright and might be called dull yellow. In one case the orange was obscured by a grayish tint except around the margins, but when this specimen was dried it became as orange as the others. Individuals under 10 mm. in diameter have little or no pigment but are nearly or quite white; the color however develops rapidly after that. On preservation, either in alcohol or by drying, the orange color is retained for a longer or shorter time but ultimately it disappears and the specimens become nearly white in alcohol or, if dried without immersion in alcohol, "museum color" of a lighter or darker shade; often the radial areas, in dry material, are distinctly dusky, unUke the interradii and margins; in other cases, the margin is distinctly hghter than the disk and basal portion of the rays; in one specimen the latter area is still quite definitely orange. Holotype, M. C. Z. no. 3261, from under surface of a rock, near low water mark. Entrance Point, Broome, August, 1929. This was a relatively common sea-star near low water mark at Entrance and Gantheaume Points, Broome. We also took it at Barred Creek to the north of Roebuck Bay and Captain Bardwell found one specimen at the Champagay Islands or Augustus Island. It lives like other Asterinas on the under surface of rock fragments but its bright color makes it more conspicuous than most members of the genus. It belongs in the same section of Asterina with nuda H. L. C. from the Murray Islands and orthodon Fisher from Hong Kong. While obviously near these two species, it is recognizable by certain distinctive features. CLARK: AUSTRALIAN ECHINODERMS 155 Compared with the type of nuda, the present species differs markedly in color in life, has smaller and more numerous skeletal plates, most noticeable in the aboral interradii and the actinal interradial areas, and the aboral plates are far more spiniferous and not at all smooth and shining. Compared with orthodon, the Western AustraUan form may be distinguished by the adambulacral and oral armature; the form and proportions of the furrow series, and the marginal series on the oral plates is markedly different in the two species. Mortensen (1934, p. 9) says that the specimen of orthodon which he studied was of a "hght pink colour, when received" which indicates a marked difference in color in life from lulea. There are 55 specimens of lutea at hand ranging in size from young ones with R = 5-6 mm. up to a very large individual with R = 20 mm. and r = 20. The ratio of R to r shows considerable range in the preserved specimens, from R = 1.15r to R = 1.4r. In life however there is less diversity, as the form is quite pentagonal with the sides usually a trifle concave or distinctly convex. Autotomy apparently never occurs and non-pentamerous variants are rare. There isoneperfectlytetram- erous young individual, a square with 4 equal sides about 8 mm. long. Another young individual, about 15 mm. across is perfectly hexamerous with nearly straight sides, while an adult just twice as large (R = 15 mm.) is also perfectly hexamerous, but its sides are deeply concave (r = 10-11 mm.). AsTERiNA perplexa' sp. nov. Plate 22, fig. 4 Rays 5, with form distinctly stellate. R = 14 mm., r=10 mm., R = 1.4r. Disk about 5 mm. in vertical diameter. Aboral plating as usual but coarse ; there are only 16-18 plates in the carinal series from center of disk to terminal plate; each plate carries a short comb or small tuft of minute sharp spinelets but many of these have been knocked off in handUng the dry specimen. Papulae numerous and very large, the perforations in the aboral skeleton being .30-.40 mm. in diameter. Margin of inferomarginal plates very conspicuous with long tufts of very slender spinelets. Madreporite very small, not much more than 2 mm. from anus. Actinal intermediate areas with relatively few and rather large plates, each of which bears a pair of long and very slender, acute spines, longer as a rule than • perplexus = puzzling, in reference to the resemblance to two or more quite diverse species. 156 memoir: museum of comparative zoology the diameter of the plate. Adambulacral armature in the usual two series; the furrow series with 4 (or sometimes 5) long very slender spines, the subambulacral series with 4 (or 3) equally long spines. Oral plates with rather conspicuous me- dian ridge on each side of which are 4 long and slender spines; the marginal spines 7-9, slender and delicate, the innermost not specially enlarged. Color in life, "distinctly white but with indefinite orange-yellow areas abor- ally; orally pure white." The dry specimen is very pale brown, lightest on distal part of rays and along margin, decidedly darker and more dusky on disk and basal part of rays. Holotype, M. C. Z. no. 3266, found clinging closely to under side of a rock fragment at Neds Beach, Lord Howe Island, April 6, 1932. When taken this individual was supposed to be the adult oi A. alba and even now the aboral surface is strongly suggestive of the same thing. But orally the differences are so great a single glance is sufficient to prove they are not closely related. On the other hand, the resemblance to inopinata is great, particularly on the oral surface, but the more stellate form, the huge papulae, the long, slender actinal spinelets, and the color in life offer great obstacles to a close relationship. Again comparison with lutca suggest an alliance but after all this is probably superficial and not real. It is unfortunate that no other specimens were found; probably the normal habitat of the species is in somewhat deeper water than is accessible for collecting without a dredge, and where the nature of the bottom precludes the use of that invaluable aid. DiSASTERINA SPINULIFERA^ Sp. nOV. Plate 22, fig. 6 Rays 5, relatively narrow and high, but the strong contraction of dorsal muscles is evident and has led to the distal part of the arms being recurved; it is hard to say how much this contraction has altered the normal form. R = 7-8 mm. , r = about 4 ; hence R is nearly or quite 2r ; arms taper but Httle, br at base is 4 mm. and near tip it is fully 3 ; height of arm at base, fully 3 mm. Aboral plates numer- ous, very crowded, bare; on basal portion of arms, they are strongly imbricated but this is certainly due in part to the contraction already mentioned; distally and along the sides of the arms, the imbrication is much less marked and probably wanting in a normally contracted specimen. If there is any skin or membrane • spinuliferus = bearing little spines, in reference to the very slender spinelets on some aboral plates. CLARK: AUSTRALIAN ECHINODERMS 157 over the aboral plates in life, it has disappeared in drying; nearly all the plates are perfectly bare, but low down in each interradial area, above the marginal series, there are half a dozen or more which carry one or more long exceedingly slender spinelets, easily overlooked without a lens; near the madreporite are 2 or 3 plates which bear similar but much shorter spinelets. Papulae apparentlj' few and irreg- ularly scattered, not conspicuous. Superomarginal plates rather large in a fairly distinct series, but the inferomarginals form the actual margin of the body; they are small, project but little, and carry a linear series of numerous, minute spine- lets which are long and slender on the proximal plates but become very short near the tip of the arms. IMadreporite small but easily seen. Actinal intermediate plates as usual in the genus, small, each with a single large sacculate spinelet, and forming 5 or 6 parallel series along each side of the ambulacral furrow; several of the series extend to the tip of the ray. Just back of the oral plates, the non-calcified area characteristic of the genus is e\ddent, its distal boundary being sharply defined by a single large plate. Adambulacral armature made up of a furrow series of 3 or 4 slender spinelets and a large subambulacral spine. Oral plates each with a large spine on the sur- face and a marginal series of 4 slender acute spinelets not much larger than those of the furrow series on the adambulacrals. Color in life not noted; dried, it is a dingy white. Holotype, M. C. Z. no. 3267, from Broome, northwestern Australia, June, 1932. This little sea-star was probably dredged in 5-8 fms. of water. Unfortu- nately its characteristics were not conspicuous enough to attract attention when collected, and it was only when the Asterinas from Broome were being critically studied several years later that its unique character was detected. The genus Disasterina has been carefully reviewed by Livingstone (1933, pp. 5-11) who recognizes 3 species, all occurring on the Barrier Reef or the neighboring Queens- land coast. The present species is readily distinguished by the character of the aboral surface, especially the spiniferous plates in the interradii. MANASTERINAi gen. nov. Similar to Disasterina, but with the aboral skeleton less rigid, the plates on the basal part of the rays lightly or not at all in contact ; many plates carry a single ' fiavds = loose, flaccid, + Asterina, in reference to the lack of rigidity in the aboral skeleton. 158 memoir: museum of comparative zoology minute spinelet ; near the tips of the arms, the plates form 7-9 longitudinal series and many of them carry long slender spines ; actinal intermediate areas without a non-calcified spot back of the oral plates, though the actinal plates themselves in that region are not very closely in contact (in the dry specimen) . Genotype, manasterina longispina sp. nov. As at present understood, Disasterina is a homogeneous group of four small species, so it seems better to make the present larger aberrant form the type of a new genus rather than to stretch the boundaries of the older one to include it. Manasterina longispina' sp. nov. Plate 21, figs. 1-2 Rays 5, form markedly stellate, with moderately deep but rounded inter- brachial arcs. R = 24 mm., r = 12 mm., hence R = 2r. Arms 12 mm. wide at base and 5 mm. close to the bluntly pointed tip. Aboral plates very numerous, ir- regularly rounded, hghtly convex, of very diverse form and size; except for 2 or 3 near the madreporite, none are as much as a millimeter in diameter; on the distal half of the arms, the plates are closely in contact in 7-9 distinct longitudinal series but proximally (except near the margin and in the interradii) these become confused and the plates themselves are not completely in contact so that naked skin shows between them often; many plates overlap others at one point or another but there is scarcely any real imbrication. Many plates are perfectly bare but a considerable number have near the center a single minute spinelet and near the madreporite, which is distinct though not large, are several plates with 2-5 spinelets each. Near the tips of the arms, the spinelets become conspicuously elongated and may exceed a millimeter in length. Superomarginal plates small, forming an inconspicuous series but inferoinarginals very conspicuous, each bearing a tuft of 2-4 very slender spinelets, 1.5-2 mm. long; this marginal fringe is very noticeable particularly on the distal half of the arms. Papulae appar- ently few and irregularly scattered; they are not easily detected in the dry specimen. Actinal intermediate areas relatively small, with 8 or 9 series of plates back of the orals but only 4 (or 3) extend to the arm-tips; each plate carries one long, ' longispina = having long spines, in reference to the notably long spines of the marginal and actinal plates. CLARK: AUSTRALIAN ECHINODERMS 159 slender, acute, sacculate spine, which hes pointing outward to the margin; back of the oral plates there are indications that the plates are not in close contact but there is no non-calcified area such as is found in Disasterina. Adambulacral plates with a furrow series of 4 (distally, 3) slender spines, the median pair com- monly much longer than the other two; a single large subambulacral spine is present on each plate, resembUng closely the actinal spines but somewhat larger; the subambulacral spines of the most proximal adambulacral plates are the largest spines on the sea-star and measure as much as 2 mm. in length. Oral plates with a single big spine on the surface, just Uke the subambulacrals, and a marginal fringe of 5 spines, of which the innermost is longest, the next is a Uttle smaller and the 3 outer ones are much smaller. Color in Ufe, aborally "reddish-pink" and orally "pale lemon yellow." The dry specimen is nearly white below, light yellowish-brown (museum color) above on disk and bases of arms; dull pinkish on distal part of arms and marginal fringe ; where skin shows between plates it is nearly black. Holotype, M. C. Z. no. 3268, from Rottnest Island, Western AustraUa, 1934, Captain Beresford E. Bardwell leg. et don. Captain Bardwell wrote me in regard to this remarkable sea-star that "it was very soft and fleshy when found" and "about 33^ inches" across. As it is now less than 2 inches across, the shrinkage in drying has been extraordinary, quite in keeping with the statement that it was very soft and fleshy. It was living "attached to a flat sandstone rock" "on pure sand bottom." No other specimen was seen. This is one of Captain Bardwell's most interesting discoveries among echinoderms, and were I not so strongly averse to mingling emotion with science, it would surely bear his name. As it is I can only express here again my very great appreciation of his work as a collector and my sincere gratitude for his generous help. There is no danger of confusing this sea-star with any previously known. Paranepanthia grandis Nepanthia grandis H. L. Clark, 1928. Rec. S. Austr. Mus., 3, p. 393. Asterinopsis grandis Livingstone, 1933. Rec. Austr. Mus., 19, p. 15. The discovery that this species ranges to the coast of Western AustraUa was one of the surprises of collecting at Point Peron in October, 1929. A very large individual was found there on the under side of a rock fragment near low-water 160 memoir: museum of comparative zoology mark. It was closely appressed to the rock and owing to its dull fawn-color might have been easily overlooked. It underwent little change of color in drying, or subsequently, but it shrank fully 10% in size; even so it is the largest specimen yet recorded as R = 60-65 mm. and r = 28. We met with no other specimens but in January, 1930, Professor E. W. Bennett found 2 individuals at Bunkers Bay, immediately east of Cape Naturaliste, which he has generously sent to me. One of these, found in "drift on sand beach," is the smallest specimen yet recorded as R = onIy 7 mm. The other is a small and typical adult with R = about 35 mm. Both of these Bunkers Bay specimens are a light grayish-brown, quite different in appearance from the brighter yellow brown of the large in- dividual from Point P^ron. Livingstone's (1933, p. 15) transfer of my Nepanthia grandis to Asterinopsis Verrill (1913, p. 480) was quite justified, but unfortunately Verrill's genus is a heterogeneous group, for which he selected as type one of Lamarck's species, 'penicillaris. Mortensen (1933, p. 258) has recently pointed out that this name probably lacks vaHdity in which case Asterinopsis must also be invalid. How- ever this may be, Verrill included in his genus both West Indian and Australian forms which are certainly not congeneric. As a beginning therefore in clearing up the confusion, I suggest that the Australian and certain East Indian species be segregated under the name Paranepanthia Fisher. The large individual of grandis taken at Point Peron seemed so similar to the figure of Fisher's type of platydisca, that it was only by careful comparison of the specimens themselves, the essential differences could be determined. The most important of these are obviously the very much shorter paxillar spinelets and the smaller and far more numerous aboral plates of platydisca. The spinelets in grandis are nearly 3 times as long as in platydisca and make a correspondingly more conspicuous tuft. Moreover in the Philippine species in a series of aboral plates reaching the margin at middle of interbrachium, aljout 26 or 27 plates are present, while at the middle of ray about 17 or 18 can be counted; in the speci- men of grandis however which is considerably larger than Fisher's unique type, only 22 or 23 plates are present in the interradial series and only 12-14 at middle of arm. There are similar differences in the adambulacral and oral armature in the two species for grandis is a coarser, higher and heavier sea-star than the Philippine form. The madreporite in platydisca is nearly triangular, sharply defined, with well marked furrows on its surface, whereas in grandis the madre- porite is ill-defined and not easy to detect. There can be no doubt that the two species are quite distinct. CLARK: AUSTRALIAN ECHINODERMS 161 Livingstone (1933, p. 16) has already expressed his opinion that grandis is a Paranepanthia but could not see any satisfactory distinction between that genus and Asterinopsis — a name which must, as already pointed out, be aban- doned. Besides plolijdisca and gj-andis, I should include in Paranepanthia, brach- iata (Koehler) as Fisher suggests, pediccllaris Fisher and the two following species. Paranepanthia praetermissa Asterinopsis praefcnuissa Livingstone, 1933. Rec. Austr. Mus., 19, p. 14. While collecting at Bottle and Glass Rocks, in Port Jackson, in November, 1929, with Mr. Melbourne Ward, two specimens of this interesting sea-star, so recent 1}^ described by Livingstone, were found. While the relationship with grandis is obvious, the flatter disk, wider and more rounded rays and very much smaller and more numerous aboral plates distinguish praetermissa easily. The larger of the 2 specimens at hand has R = 32 mm. while the smaller one is about one-third as large (R = ll mm.). Paranepanthia rosea' sp. nov. Plate 22, fig. 8 Rays 5, form markedly stellate, rather fiat. R = 7 mm., r = 5 mm., hence R = 1.4r; br = 4-4.5 mm. Aboral plates of rather uniform size, in regular longi- tudinal series on the arms, the 3 median series rather distinctly set out from the rest ; imbrication of plates not very conspicuous but evident along the middle and upper series of plates on the rays. Each plate bears a group of numerous, minute, diverging, glassy spinelets which cover its surface completely. Papulae rather numerous and more or less conspicuous (according to size and condition of speci- men) ; usually 6 rows are evident on the basal part of each ray but the distance to which they extend distally shows considerable diversity. Superomarginals, 2 or 3 times as large, forming the margin of the animal; each one is well covered by the large number of minute spinelets which it bears. Madreporite small (.60- .70 mm. across) but distinct, only about 1.5 mm. from center of disk. Actinal intermediate plates in longitudinal series parallel to adambulacral furrow; only the one adjoining the furrow extends to the arm-tip; the others fall short by increasing distances until the sixth or seventh consists of only 2 or 3 ' rosea, in reference to the pretty pink color, often retained in dry specimens. 162 memoir: museum of comparative zoology minute plates near the interradial line; each plate is fully covered by a rather dense cluster of minute spinelets like those on the aboral plates but obviously longer. Adambulacral armature made up of a furrow series of 6 (rarely 7) slender spines of which the median are longest, the outermost on each side, small- est ; on the surface of the plate is a cluster of 8 or more very slender spines inter- mediate in size between those of the furrow series and the spinelets of the actinal plates. Oral plates rather small but with a conspicuous marginal series of 7 spines (the innermost largest) on each plate; on its surface is a cluster of 6-8 much smaller spines similar in size and appearance to the subambulacrals. Color in life rose-red, more or less variegated with cream-color; often the rose-red is little interrupted but in some cases, the cream-color predominates quite markedly; oral surface is always cream color. Dry specimens range (ac- cording to methods of preservation) from the natural rose-red to ordinary "museum color" or to nearly white. Holotype, M. C. Z. no. 3272, from the cove on the northeast corner of Rott- nest Island, Western Austraha, October 19, 1929. This pretty little sea-star, so different from any of the other Asterinidae of the west coast, is common at Rottnest Island and it was also found at Point Peron. It lives closely attached to the underside of rock fragments in shallow water. Of the 13 specimens which we collected at Rottnest, the smallest is hardly 9 mm. across while the largest have R nearly 8 mm.; of these one has 6 rays but there is no indication of autotomy; a smaller specimen (R = 5.5 mm.) apparently has only 4 rays but on the oral surface a minute fifth ray is evident. Besides the material which we collected, there are at hand 9 specimens which were taken at Rottnest by Captain B. E. Bardwell in 1934, 2 collected by Mr. G. Bourne at the same island in 1931 and 7 also from Rottnest, loaned by the Western Austrahan Museum. The pair collected by Mr. Bourne are hexamerous — probably they are selected specimens, preserved because they are 6-armed; one is full grown (R = 8 mm.), the other is much smaller (R = 5.5 mm.). Adult pentamerous specimens show considerable diversity in form due to differences in length of the minor radius and the breadth of arm; thus a specimen in the Western Austrahan Museum with R = 8 mm. has r and br 5.5-6 mm. while one of those taken by Captain Bardwell having R = 8 mm. has r and br 4.5-5 mm. and hence appears much more stellate. There is no Asterinid, of which I know, with which this species can be confused. Although much smaller than the other species of Paranepanthia, the resemblance to young specimens of grandis is so obvious, there seems no doubt that rosea is congeneric with that species. CLARK: AUSTRALIAN ECHINODERMS 163 Patiriella calcar Asterias calcar Lamarck, 1816. Anim. s. Vert., 2, p. 557. Patiriella calcar Verrill, 1913. Amer. Jour. Sci., 35, p. 484. This fine and very distinctive sea-star was met with several times on the coast of New South Wales, but the finest specimen secured was taken at Hobart, Tasmania, November 15, 1929, where it was found above low tide mark but under water, exposed on the surface of a rock in a sheltered nook. It was nearly four inches across (R = 45 mm. in the dry specimen) and the color was "bright orange, with disk and interradial areas variegated with greenish and brown ; central part of disk cream white with a brown spot at very center; lower surface cream-color." Of 3 specimens which we took at Long Reef, CoUoroy, November 28, 1929, the smallest (R = 20 nmi.) has 9 equal rays, and of 3 small individuals found at Gunnamatta Bay on November 26, one, with R = 15 mm., has but 7 rays. We found a small specimen on November 27, at Bottle and Glass Rocks, Port Jack- son, which has 8 rays but one is distinctly shorter than the other 7. Mr. W. Heron of Coffs Harbor, N. S. W., kindly gave us in May, 1932, 34 small specimens of calcar from that vicinity, 2 of which have 9 rays and 1 has 7. These Coffs Harbor specimens have the rays more slender and attenuate than any of those from further south, but this might be associated with the method of kiUing and drying. Patiriella exigua Asterias exigua Lamarck, 1816. Anim. s. Vert., 2, p. 554. Patiriella e.vigiia Verrill, 1913. Amer. Jour. Sci., 35, p. 484. This httle sea-star is very abundant on the main coral reef on the western side of Lord Howe Island, particularly at the northern end and even more so at the southern end where it adjoins Mt. Lidgbird. We also found it common at Gunnamatta Bay, N. S. W. and at Hobart, Tasmania. A single specimen was secured at Long Reef, Colloroy, N. S. W. but probably it is also abundant there at suitable times. Altogether there are 2022 specimens at hand, of which 1959 are from Lord Howe. The 32 specimens from Gunnamatta Bay are all peutamerous and small, the largest less than 20 mm. across. The 30 specimens from Hobart are a little larger, several being 21-23 mm. across, and one is perfectly tetramerous, with sides 16 164 MEMom: museum of comparative zoology mm. long; this last was a gift from Professor T. T. Flynn. The Hobart specimens were found on the under side of rocks above low water mark, as is usually the case. The color was "varied green or brown above with more or less red, never much; greenish or greenish-blue below." The notable harmony with the back- ground has been well emphasized by Mortensen (1933, p. 253) as seen by him in South Africa. At Lord Howe Island, this sea-star is so abundant, large collections were made at the two extremes of the reef to see if any local differences existed in the proportion of specimens having more or fewer than 5 rays. The distribution of exigua at Lord Howe is peculiar; not a single specimen was found on the eastern side of the island, not even at Neds Beach, where conditions for collecting echino- derms are the best of any one area on the island shore; on the western coast, exigua proved to be abundant on that coral reef area, abutting on Phillip Point, which shuts in North Bay, and also at the extreme southern end of the main reef where it abuts on Mt. Lidgbird. In the latter area, it was most abundant near high water mark on the rocks along shore. This peculiar distribution leads me to believe that exigua is a recent arrival at Lord Howe having been brought over from Port Jackson on foul ship bottoms. The exiguas living at North Bay were prevailingly light buff and greenish but often were variegated with shades of green and lacked the buff. Those on the south reef were distinctly darker, varie- gated light and dark green often with some orange; one specimen was predomi- nantly orange. In both regions, the sea-stars were found on the under surface of rocks between tide-marks, but along the shore near Mt. Lidgbird they were almost equally common on the exposed surfaces of the rocks. In 1913, I found exigua very common at Erub, an island in Torres Strait (see H. L. Clark, 1921, p. 97) and examined nearly 600 specimens to learn the percentage of non-pentamerous variants; the result showed that 92% were pen- tamerous, 6% were hexamerous, 1.5% were tetramerous and 3 specimens had 7 rays. Conditions at Erub were very favorable apparently and the sea-stars grew to a large size, the largest 35 mm. across, as compared with a maximum of 25 mm. at Lord Howe. (A specimen loaned by the Australian Museum is 28 mm. across). It occurred to me at Lord Howe, that comparison with the data from Erub would be interesting, so 800 specimens of exigua were collected at North Bay and 1100 along the shore near Mt. Lidgbird. The result is interesting but not extraordi- nary. At North Bay, 2.5 % of the specimens were non-pentamerous, 2 % ( 17 speci- mens) being hexamerous and .5% (4 specimens), tetramerous. At the southern end of the reef, a trifle more than 2% are non-pentamerous, exactly one in a CLARK: AUSTRALIAN ECHINODERMS 165 hundred being hexamerous; 1 specimen has 7 rays, 5 have 5.5 (i.e. one ambulacral furrow is forked at a greater or lesser distance from the mouth) and 6 have but 4 rays. It is odd that there are twice as many hexamerous individuals at North Bay as at the southern reef flat but the numbers involved are too small to make the apparent fact significant. The difference between the Lord Howe population as a whole however, and that at Erub is perhaps striking enough to be important. Evidently variation is much more general at Erub than at Lord Howe, both 6 and 7 rayed specimens being much more common at the northern island. This might perhaps be reasonably construed as evidence that the Lord Howe population is younger and living under less favorable conditions and therefore variants are promptly ehminated; thus supporting the view that it has been introduced from Port Jackson in the past century. PaTIRIELLA GUNNIl Asterina gunnii Gray, 1840. Ann. Mag. Nat. Hist., 6, p. 2S9. Patiriella gunii Verrill, 1913. Anier. Jour. Sci., 35, p. 484. A large series of this characteristic Australian sea-star is at hand consisting of 375 specimens, ranging in diameter from 3 mm. to 75. Of the 375, 90% (342) have 6 rays, 5% (19) have 7 rays, while 8 specimens have 8 rays and 6 specimens have but 5. The color in hfe shows extraordinary diversity. The 385 specimens come from the following places : Western AustraUa: Dongarra, E. W. Bennett leg. et don. 4 very young speci- mens. Garden Island, near Fremantle, October 14, 1929. 1 small adult. Rottnest Island, Point Vlaming, February, 1930, Swan and Drummond leg. E. W. Bennett don. 1 very small 7 rayed specimen, 3 mm. across. Found on seaweed. Rottnest Island, October, 1929. 1 very small specimen. Rottnest Island, 1934. Captain B. E. Bardwell leg. et don. 10 specimens. Point Peron, October, 1929. 108 specimens. Bunbury, October 26, 1929. 6 specimens. Bunkers Bay, January, 1930. E. W. Bennett leg. et don. 174 specimens of all sizes. 166 MEMom: museum of comparative zoology Ellenbrook Beach, January, 1930. E. W. Bennett leg. et don. 5 specimens. Hamelin Bay, January, 1930. E. W. Bennett leg. et don. 4 specimens. Exact locality, date and collector uncertain but undoubtedly collected between Point P^ron and Hamelin Bay. 51 specimens, adult and young. South Australia: Port Willunga, November 2, 1929. 1 small adult. New South Wales: Locality ? Mr. S. Larnach leg. et don. 2 young specimens. Colloroy, Long Reef, November 28, 1929. 7 adults. Patiriella brevispina' sp. nov. Plate 22, figs. 2-3 Similar to gunnii aborally, though the granules on the plates seem a trifle lower and more angular when dry specimens are compared side by side. On the oral surface all spines shorter and stouter than in typical gunnii and there are generally fewer to a plate. Actinal intermediate plates with very short spines, shorter than the plate that carries them and often not twice as long as wide ; on most of the plates there are two such stumpy spines but near the oral plates, there is only a single spine to a plate and the large plates immediately distal to the orals may have none. Adambulacral plates with a furrow series of 2 short, flat- tened, truncate spines, much stouter basally than at tip; on the first 2 or more plates there are 3 such spines and in large specimens (R = 40 mm. or more) as many as 6 or 8 proximal adambulacrals have 3 spines; on the surface of each plate is a single stout subambulacral spine, often wider at tip than at base and only 2-4 times as long as wide. Oral plates with a similar spine on the surface, at their distal end and a marginal series of 4 (in large specimens, 5) stout blunt spines, the innermost largest. Color in life "deep purple, or brownish-crimson oi- more nearly brown (but with surprisingly little difference) with deep orange tube- feet." Dry specimens unicolor, faded but with a distinctly purplish tint especially on the distal part of the rays. The largest specimen, which was uniform deep purple in life, now has the disk and basal part of rays dull orange and the distal margins and the distal part of the rays light violet. Holotype, M. C. Z. no. 3305, dredged in 5 8 fms. in Koombana Bay, Bun- bury, W. A., October 26, 1929. 1 brevis = short, in reference to the remarkably short actinal spines. CLARK: AUSTRALIAN ECHINODERMS 167 Our attention was first attracted to this sea-star by its handsome purple color quite different from anything we had noted in the hundreds of gunnii which we had seen at Point Peron and along shore at Bunbury. When we found that all of the Patiriellas dredged in Koombana Bay agreed in this striking coloration, our interest was quickened. Subsequently, Professor Bennett found the same form at Bunkers Bay and Ellen Brook Beach, which he kindly sent me with some addi- tional specimens of large size from Bunbury. On November 2, 1929, while collect- ing along shore during low tide, at Port Willunga, South Australia, I found a very fine specimen of this sea-star about 80 mm. across. No specimens from further east have come under my notice. The material in hand consists of 20 specimens ranging from 17 to nearly 90 mm. in diameter. There are 2 specimens with 7 rays among those dredged in Koombana Bay. There is room for considerable doubt whether this form is really a vaUd species or merely a variety of gunnii, a species which shows very great diversity in color and considerable in spinulation. But the more I have studied the material at hand, the more I feel convinced that brevispina is consistently and constantly distinct. Of course direct study of the question at Bunbury or some equally good locality may demonstrate that my conviction is ill-founded. The 20 specimens of brevispina at hand are from the following places: Western Austr aha: Bunbury, Koombana Bay, 5-8 fms., October 26, 1929. 14 specimens, small adults and young. Bunbury; gift of Bunbury Naturalists Club through Pro- fessor E. W. Bennett, 3 large adults, in very poor condi- tion. Bunker's Bay, January, 1930. E. W. Bennett, leg. et don. 1 large adult ; very poor. Ellen Brook Beach, January, 1930. E. W. Bennett. 1 small adult. South AustraUa: Port Willunga, November 2, 1929. 1 large adult; very fine. Patiriella nigra' sp. nov. Plate 21, figs. 3-4 Very similar to P. regularis of New Zealand, so that structural differences on which any weight can be put are diflScult to find. Only two or three are worth ' niger = black, in reference to the extraordinary color of the living animal. 168 memoir: museum of comparative zoology mentioning and it is quite possible that these are inconstant and worthless. There is little well-preserved material of rcgularis available (although a cotype of ^'errill's and 6G other specimens are in the M. C. Z. collection); the best is an adult in alcohol and some small, dry adults collected in "New Zealand" and given to the Museum by Mr. Stanley L. Larnach in 1931. These specimens permit a careful comparison with the type of nigra which is in perfect condition, and the following differences are tangible; in the interradial areas the aboral plates in regularis carry relatively few spinelets (3-12) and this group rapidly becomes crescentic as one passes inward from the margin, and subtends a large papula; in nigra, these plates carry large dense groups of spinelets (12-15 or more) which cover the area closely and do not become crescentic until well up on the disk; again, the large inner marginal spines on the oral plates are distinctly flatter and more abruptly truncate in nigra than in any specimen of regularis I ha\'e seen; finally the 2 furrow spines in nigra are more nearly equal, more enlarged at base and more acute at tip then in regularis. These differences taken by themselves alone would not warrant the estab- lishment of even a named variety but in connection with the locality where taken and the extraordinary color in life, it is justifiable to use a distinctive name. Color in hfe, black, "coal black" according to my field notes; it was certainly not a dark green or greenish-black nor a deep purplish-black, but simply a dead black; the lower surface was dusky. After killing in corrosive-formalin and dry- ing, the lower surface became museum color around the margins but suffused with green inwardly from a definite boundary about 5 mm. from the margin; the aboral surface is dark green for 10-1.5 mm. inward from margin and then becomes rather abruptly brown-orange. Holotype, M. C. Z. no. 3311, from under surface of a rock fragment near western margin of reef-flat at Neds Beach, Lord Howe Island, April 22, 1932. On the evening of April 21, 1932, at a social gathering, a resident of Lord Howe, herself an enthusiastic collector of marine animals (particularly shells), asked me if we had found "the black starfish." Incredulously I replied that I had never seen or heard of such an animal l)ut she assured me that she had found them at very low tides, although they seemed to be quite rare; they were to be found clinging to the underside of rock fragments and were "quite black all over." Mr. Livingstone and I discussed the matter later and agreed that she must have something else in mind than a sea-star. But the very next day, while mak- ing what was expected to be my last visit to Neds Beach, I turned over a rock fragment near the western edge of the reef-flat and there was a beautiful specimen CLARK: AUSTRALIAN ECHINODERMS 169 of a perfectly black sea-star! The next day three of us worked for three hours on the same and adjoining parts of the reef-flat but in vain, no second specimen was secured. And the following day we bade adieu to Lord Howe. At the AustraUan Museum, the following week, Livingstone showed me three specimens of Patiriella from Lord Howe Island, but they had all been pre- served in alcohol and hence are simply museum color; there is no means of know- ing what the color in life may have been. According to Farquhar (1895, p. 197), the New Zealand species regnlaTis shows great diversity in color, "greenish-gray," "yellow, orange, purple, dull green variegated with yellow," "dark purple and many shades of blue." Of course if further collecting at Lord Howe reveals speci- mens of Patiriella which are not black, the status of nigra will be very dubious, yet the crowded spinulation of the interradial areas may warrant its retention as a variety. The holotype is an almost perfect pentagon with R = 27-28 mm. and r = 21; the vertical diameter is 10 mm. Thanks to Mr. Livingstone, one of the specimens from Lord Howe in the AustraHan Museum is at hand. It has R = 35 mm.; the spinulation of the interradii is dense and it may properly be considered an example of nigra. Nepanthia belcheri Asterina {Nepanthia) belcheri Perkier, 1876. Arch. Zool. Exp., 5, p. 240 (.320 of reprint). Henricia heteradis H. L. Clark, 1909. Mem. Austral. Mus., 4, p. .530. Nepanihia polypla.r Doderlein, 1926. K. SAenska Vet.-Akad. Handl. (3) 2, no. 6, p. 20. The determination of the systematic position of the Nepanthias which were taken at Lord Howe Island, and of specimens in the Austrahan Museum, loaned for identification, has caused me a great deal of perplexity. The best solution — the only one really satisfying — is to call them all a single species and apply to them the name belcheri Perrier, vahdated in 1884 by Bell. Having presumably examined the types in the British Museum, Bell assigned to belcheri a specimen from Port Jackson, thus fixing a definite locality for the species. As he does not suggest, in his exasperatingly brief reference, that his specimen had other than 7 arms (the number given by Perrier as characteristic of the species) but does state that it had 4 madreporites, Bell has added but very little to our knowledge of the species. Doderlein based his species polyplax on a single specimen of Nepan- thia from "Rockhampton" (presumably Queensland) which had 6 rays and 6 madreporites. This striking symmetry seemed to him sufficient ground for the estabhshment of a new species. But material from Lord Howe Island shows that no reliance can be placed on the number of either arms or madreporites. Doder- 170 memoir: museum of comparative zoology lein's admirable figures lead me to identify his species with the one occurring at Port Jackson and Lord Howe Island. Placing this typical Nepanthia in the genus Henricia is the most absurd mistake of which I have been guilty in all my work on Echinoderms and is of course inexcusable. It can be accounted for only by my unfamiliarity with Nepanthia and overfamiliarity with Henricia in 1909. This sea-star is one of the characteristic echinoderms of Lord Howe where it occurs on both sides of the island under rock fragments. Symmetrical specimens are rare; nearly all show evidence of autotomy to a greater or less degree. The color in life is an inconspicuous dark dull green indistinctly speckled orally with reddish. The color is fugacious and dry specimens are usually museum-color; occasionally however a specimen shows a marked purplish rose-color which per- sists for years. It is a literal truth that no two of the 56 specimens at hand, nearly all from Lord Howe Island (4 being on loan from the Australian Museum), are exactly alike in number, size and form of arms, and in number of madreporites. Appar- ently 6 is the normal number of arms as 27 specimens show that number, while only 8 individuals have 7, 7 have 5, 3 have but 4, 1 has 8, 1 has but 3, and 1 is a single arm isolated but obviously in process of growth to become a "comet" such as are common in the autotomous reproduction of Linckia guildingii. Perfect "comets" in belcheri are by no means rare; one of the most typical consists of an arm 12 mm. long and 5.5 mm. wide, with 5 little arms about 4 mm. long radiating symmetrically from the oral end. Another very symmetrical specimen consists of an arm 18 mm. long, from the oral end of which radiate 4 arms 10 mm. in length. The 3-rayed specimen is obviously a hexamerous specimen so recently di\'ided by autotomy that new rays have not yet become distinctly differentiated though growing tissue is evident. The 8-armed specimen has 2 arms 16-18 mm. long and 5, 11-12 mm., while the eighth arm, projecting only 4 mm. from the disk margin is obviously a supernumerary bud crowding in between two normal arms. As already stated symmetrical specimens are rare ; there are 2 perfectly symmetrical pentamerous specimens, with R = 10 and 17 mm. respectively; the smaller has 2, the larger, 4 madreporites. A symmetrical hexamerous specimen with R = 30-32 mm. is the largest specimen of belcheri that we found at Lord Howe. It is not perfectly symmetrical however for, though there are 6 properly placed madre- porites, one is twice as large as the others and is divided into two nearly distinct plates. A 7-armed specimen with R = 22-25 nmi. is almost perfectly symmetrical as there are 7 madreporites of equal size located one in each interradius. CLARK : AUSTRALIAN ECHINODERMS 171 Aside from the number of rays and madreporites this species is easily dis- tinguished from brevis by the fewer and coarser spines making up the oral and adambulacral armature; in belcheri, the larger ones are thickened at tip and blunt — which is not the case in brevis. Specimens of bcJchcri from the mainland coast of Australia do not seem to be common. One was secured by the Great Barrier Reef Expedition and was as- signed by Livingstone (1933, p. 262) with frank skepticism to "{?brcvis)." He makes no reference to poJyplax Doderlein and probably had not at the time of writing seen that author's figures. There are at hand 2 specimens loaned by the Australian Museum which were taken at the mouth of Lane Cove River, Port Jackson, many years ago and are listed by Whitelegge (1889, p. 201) as "Patiria crassa Gray." Years later, I (1925, p. 2) recorded a specimen in the Stockhohn Museum from Lane Cove River and corrected Whitelegge's identification of the species, pointing out that the Lane Cove River specimens were identical with those from Lord Howe Island, which I carelessly persisted in calling "Henricia." The Stockholm specimen has 7 nearly equal arms, about 35 mm. long; it is prob- ably one of Dr. Ramsay's original series. Excepting the recently secured material from Lord Plowe Island, all of the specimens at hand are old ones and the spinelets are more or less rubbed off, especially on the disk and basal part of rays, but aside from this artificial loss of spinelets the appearance of the two largest individuals indicates that after R = 30 mm. (or thereabouts) the spinelets are gradually replaced aborally by fewer, shorter and coarser, pointed outgrowths of the plates giving them a rougher coat and leaving the aboral surface of the animal relatively bare and not at all "vel- vety." Even on the oral surface a change to fewer and shorter spinelets is evi- dent, but in the interradii and on the sides of the rays it is not very noticeable. In order to summarize our present knowledge of Nepantliia, and set out the characters which apparently distinguish the various species, including the new ones described beyond, the foUowmg key is offered as a beginning towards a satisfactory understanding of the genus. Key to the Species of Nepanthia A. Arms long; R = 5-6r =^. A small supplemental plate proximal to each papula; furrow spines, S-10; R a little more than 6r maculaia No conspicuous supplemental plates proximal to papulae; fiurow spines, 5 or 6; R rather less than 6r tenuis 172 memoir: museum of comparative zoology A^. Arms short ; R = 3^r ± . B. Rays 5, swollen at base, then rather abruptly smaller; madreporite single suffarcinata B". Not as above. C. Rays 6 or 7 (rarely 5 or 8), short and thick in adults; madreporites 2-8; aiitot- omy occurs very generally; adambulacral armature with about 5 furrow spines belcheri^ C Rays typically 5 (occasionally 6, very rarely 7); madreporite usually single, 2 or 3 in multiradiate individuals; autotomous only when young; furrow- spines more than 5. D. Aboral plates densely clothed with minute spinelets, almost like velvet; innermost oral spine not notable. Aboral skeleton with numerous, short, wide crescentic plates, not ar- ranged in very regular longitudinal series except near tip of arms. Adambulacral and oral armature of numerous very slender, acute spines brevis Aboral skeleton with fewer, longer, narrower, more crescentic plates in very regular series; adambulacral antl oral armature of fewer, coarse, blunt spines variabilis D'. Aboral plates with fewer, coarser spinelets, not at all velvety; innermost spine conspicuously large, flattened and acute magnispina Of the above 7 supposedly valid and properly defined species, 5 are in the present collection — 2 previously known species from northeastern and eastern Australia and 3 hitherto undescribed species from the northwestern coast. On the western coast Nepanthia seems to be replaced by Parasterina. Sladen's species suffarcinata like joubini is non- Australian. No specimens of either species have come under my eyes. Nepanthia brevis Asterina (Nepanthia) brevis Perkier, 1876. Arch. Zool. Exp., 5, p. 241 (321 of reprint). Nepanthia brevis Sladen, 1889. "Challenger" Ast., p. 387. This is the best known of the species of Nepanthia as it has been taken re- peatedly in Torres Strait and on the coast of northern Queensland. As it has been figured by Bell (1884) and by myself (1921), there is little occasion for discussion of it here. Specimens in the Australian Museum, kindly sent to me for study, show considerable diversity in form and in spinulation but in view of the fact that these characters are profoundly affected by the condition of the specimen when killed (whether contracted or relaxed and to how great a degree) and by the ' Koehler's species joubini from Cochin China is so near to belcheri, no way of distinguishing them is available. Koehler himself makes no reference whatever to belcheri. CLAKK: AUSTRALIAN ECHINODERMS 173 method of killing and preservation, it would be absurd to try and distinguish species on such trivial details. It may be well however to record certain devia- tions from typical for the use of future students. A normal adult specimen of brei'is has 5 arms and 1 madreporite; R =40 mm. * withrabout one-third as much; the arms are blunt with the width near middle about one-fifth R and near tip, practically the same ; at base the width is about equal to r. Some variations from this normal form are as follows: a specimen from Port Curtis, Queensland, has R =30 mm. ± and r and br, at all points (except the very tip of arm) nearly or quite 9 mm. On the other hand a specimen in the Australian Museum (J 5151) has arms 45-50 mm. long with r = 15, and br = 14, 10 and 7 mm. at base, middle and near tip. Most specimens have 5 arms and a single madreporite, but two of the series at hand have 6 arms and a very small and irregular specimen has 7; of the two with 6 arms, one has 4 madreporites and the other 3, while the little 7-armed individual has 3 very small ones. One of the 6-armed specimens has tapering and somewhat pointed arms, but it is hard to tell how much this is due to the method of drying. The most constant character of brevis seems to be the long and very numerous slender spines making up the adambulacral and oral armatures. The discovery of breids at Darwin was one of the interesting results of our collecting there as the range is tlius extended far to the west. The 4 specimens dredged at the Shell Islands are very typical but the little 7-armed Nepanthia picked up on the beach at East Point is certainly an extreme variant, though there is no good reason, save the number of arms, for declining to call it brevis. The 20 specimens of this species at hand range from R =12 to R =55 mm. They come from the following stations: Queensland: Port Curtis, Rat Island. Ward and Boardman leg. 4 specimens, loaned by the Australian Museum. Port Denison. 4 specimens, loaned by the Australian Museum. At, and in vicinity of Thursday Island, 1928. Melbourne Ward leg. 5 specimens. Loaned by Australian Museum (nos. 5147- 5151). Northern Territory: Darwin, near Shell Islands, 2-3 fms., July, 1929. 4 speci- mens, 1 adult. Darwin, near East Point, washed up on beach. 1 very young specimen with 7 rays. Locality unknown. 2 specimens without labels or numbers. Loaned by Aus- tralian Museum. 174 memoir: museum of comparative zoology Nepanthia magnispina' sp. nov. Plate 20, figs. 1-2 Rays 5, but only 3 are normal, 2 at one side being very small, growth having only just begun; whether this is the result of autotomy or of an accident is not clear. R = 18, 28 and 32 imn.; r = 9; br = 9-ll mm.; arms high at base, about 7.5 mm. Aboral skeleton made up of more or less crescentic plates, about as wide as long, arranged in fairly definite longitudinal series; they are relatively narrower and longer than in brevis; the spiniferous ridge is narrow, more or less curved, and carries a dozen or more short, sharp, glassy spinelets; the surface is thus only sparsely covered and is not in the least "velvety." Inferomarginal plates rela- tively large and conspicuous forming a very definite margin to the rays. No madreporite is present but its absence is probably accidental. Actinal intermediate plates in 5 or 6 series at base of ray but the two outer- most series contain only 1 or 2 plates each ; the next series contains 7 or 8 plates, the next has 35 plates and reaches nearly to the arm-tip; the innermost has as many plates as there are adambulacrals and accompanies that series to the tip of the arm. All actinal plates carry a tuft of short, sharp, divergent, glassy spinelets. Adambulacral plates with a furrow series of 4 or 5 divergent spines, which are neither slender nor stout, but the middle 3 are about as long as the plate margin and have opaque blunt tips; back of the furrow series is a second similar series of shghtly smaller spines and on the surface of the plate are a few (1-6) much smaller spinelets. Oral plates each with 5 or G marginal spines similar to the largest furrow spines but the innermost is abruptly largest, 1.5-2 mm. in length, flattened but quite acute. Color, the usual yellowish-brown of museum material; color in Ufe was not recorded. Holotype, M. C. Z. no. 3320, from Augustus (or Champagay) Island, northern Western Austraha, October, 1933. Captain Beresford E. Bardwell leg. This was the only Nepanthia secured by Captain Bardwell and was sup- posed to be the species common at Broome but even a hasty examination showed it could not be placed there. It was then thought to be bi-evis but it is even more different from that species, particularly in the adambulacral and oral armature. It is therefore necessary to give it a distinctive name and await further material. ' magnus = big -f spinus = a spine, in reference to the innermost oral spine, which is unusually con- spicuous. CLARK : AUSTRALIAN ECHINODERMS 175 Nepanthia tenuis^ sp. nov. Plate 20, fig. 3 Rays 5, slender, terete and blunt, flattened orally. R = 64 mm., r and br = 11 mm.; hence R is a little less than 6 r or br. Aboral surface covered with very numerous small plates ; on the disk and along the median area of each ray most of these plates are crescentic and subtend a papula but elsewhere they are rounded quadrangular; there is no small plate proximal to each papula; along the sides of the arms the plates form more or less regular longitudinal series. All plates are densely covered with minute, short, glassy spinelets, easily rubbed off in handhng the dry specimen. Papulae large and numerous extending in several series to the tip of the arm. Madreporite single, large, about as far as its own diameter from the center of the disk. Both supero- and inferomarginal plates are small and form no very well- defined boundary to the aboral surface. Actinal intermediate plates in about 7 series of which the 3 innermost extend nearly or quite to the tip of the arm; the series adjoining the ambulacral furrow is made up of much the largest plates; all actinal plates are densely covered with spinelets. Adambulacral arm- ature of a furrow series of 5 or 6 relatively thick, blunt spines, a Uttle longer per- haps than the adambulacral plate itself; back of that is a second and usually a third series of similar but slightly smaller spines, while on the outer end of the plate are some spinelets hke those on the actinal plates. Each oral plate with a marginal series of 10-12 long spines, similar to the furrow spines but longer; on the surface of the plate is a large compact group of a dozen spines or more. Color in hfe: "hght gray with an indefinite whitish area on distal half of arm, with irregularly scattered small flecks of deep green; there is a reddish tinge to the gray and there are scattered white plates on basal half of arm. Lower surface white with a few scattered flecks of dull green along ambulacra." The dried specimen is uniformly museum-color. Holotype, M. C. Z. no. 3321, from Roe Bank, 4 fms., outside Roebuck Bay, Broome, W. A., sand, shells and nulUpore, June 20, 1932. This fine Nepanthia, more hke maculala, the genotype, than any of the other species is readily distinguished by the character of the dorsal skeleton and the ' ieraujs = slender, in reference to the .arms which are so much more slender than in the common Nepan- thia of the same region. 176 memoir: museum of comparative zoology adambulacral armature. Fortunately Sladen (1889) was wise enough to give figures of these features as they are in maculala, so that in spite of the rarity of that species it is possible to understand easily how different from it the West- ern Australian species is. That ie7iuis is not common at Broome is evident from the facts that we did not meet with it at all in 1929, there are no specimens in the Museums at either Perth or Sydney and in 1932 we only secured 2 specimens besides the holotype. These are both much smaller; one has R = 4.5 mm. while in the other it is about 38. Nepanthia variabilis^ sp. nov. Plate 10, figs. 4-5. Plate 20, figs. 4-5 Rays 5, rarely 6 in adults, rather stout, more or less tapering. R = 60 mm., r and br = 13; hence R = about 4.5r or br. \^^lile these measurements of the holotype are typical there is considerable diversity on both sides of this normal form. Aboral skeleton as in brevis, and other typical members of the genus — crescentic plates on median area of arms, with rounded quadrilateral plates in longitudinal series on the sides. Spiniferous ridges on the plates well developed and well covered with little spinelets but owing to the breadth and depth of the areas between the ridges the aboral surface is not smooth or velvety. Papulae large and numerous, all over sides and top of arms. Madreporic plate single, rather large, nearer to center of disk than to margin; when there are 6 arms, there are 2 or 3 madreporites, but in small individuals (R less than 15 mm.) it is usually hard to find any madreporite at all. Marginal plates small but inferomarginals are easily distinguishable and form a definite boundary to the oral surface. Actinal intermediate plates in about 7 series at very base of ray but only 4 series extend to middle of ray and of these only the innermost goes clear to the tip of the arm. Each actinal plate is closelj' covered with a dense coat of small spinelets. Adambulacral armature, much as in tenuis — a furrow series of 5 or 6 relatively stout, blunt spines, a similar slightly smaller series back of it, in which the individual spines are com- monly stouter (at least near base of arm) than those of the furrow series; the remainder of the plate is covered by a third series of spines, or by little spinelets or by both. Each oral plate has about 8 marginal spines, much like those in the ' variabilis = marked by diversity, in reference to the remarkable diversity in form and color which this sea-star exhibits. CLARK: AUSTRALIAN ECHINODERMS 177 furrow series of the proximal adambulacrals but longer; on the surface of the plate is an elongated cluster of more or less similar spines. Color in life, verj' diversified and often brilliant; the dry specimens are usually museum-color but many show blotches of dusky and these are quite conspicuous in some of the smaller specimens. Holotype, M. C. Z. no. 3323, from Broome, W. A., August, 1929. The diversity of this sea-star is really extraordinary. As regards form, differences in the slenderness of the rays, in the degree to which they taper, in the bluntness or acuteness of the tip, in their relative height at base and even in their number lead to remarkably unlike-looking specimens. In color, the diver- sity is even more notable, so great as to be almost unbelievable. A few extracts from my field notes will make this evident: "Variegated, white, Ught orange- brown and liver-brown;" "variegated greens and whitish, rather like brevis;'' "bright red, adambulacral spines black, just the very tips almost white, tube- feet white with tips almost black;" "light fawn color with a very few (about 5) irregular small blotches of bright dark brown, oral and adambulacral spines deep green;" "disk and base of raj^s aborally bright light violet, madreporite pale green, remainder of rays variegated whitish and light brown with a few widely scattered ii-regular small blotches of crimson; oral surface near margin pale fawn color but near mouth and grooves the tufts of spinelets are rusty ; oral and adambulacral spines deep green;" "variegated whitish and light brown (some blotches are black brown) with irregular blotches of bright rose-color, lower sur- face fawn and rust-color, oral and adambulacral spines green;" "variegated white and red-violet, oral surface pale pink-brown (almost salmon) with oral and adambulacral spines and adjoining spinelets deep blue-green in sharp contrast;" "white mottled with black especially on disk and base of arms with a very few inconspicuous blotches of purple-brown, madreporite cream-color;" "bright orange-red with center of disk vermilion, distal portion of arms varie- gated with cream-color and vermilion, oral surface, orange, oral and adambulacral spines more or less brown distally;" "almost rose-red." Most of the above specimens were collected along shore in Roebuck Bay, August 26, 1929, but a few were dredged. In their present dry condition there is still considerable diversity but the majority are museum color of a darker or lighter shade; some show distinct blotches of dusky or dull reddish, some have a reddish or purplish tinge and one small speciiiien is a rather bright rose-red aborally. Growth changes are interesting. All specimens with R = 13 mm. or less are 178 memoir: museum of comparative zoology either symmetrically hexamerous (3 specimens) or liave obviously undergone autotomous division (4 specimens); of the latter, 3 have 2 large and 4 small arms and one has 2 large and 3 smaller arms. Another specimen is almost a "comet" having one arm with R = 16 mm. and .5 additional arms with R =7-12 mm. All other specimens with R = 15 mm. or more are symmetrical pen- tamerous specimens except a few that are obviously regenerating one or more rays lost by accident of some kind, and three individuals which each deserve a word of comment; one is a perfectly tetramerous individual (with R=30mm.); one is a hexamerous adult with R = 42-52 mm. ; and one is a hexamerous specimen in which R =30 mm. in two adjoining arms and 23-25 mm. in the other 4, un- doubtedly a case in which a hexamerous adult is developing after autotomy and interesting as showing that hexamerous specimens do not necessarily arise, as might be supposed, from the 3-armed half (after autotomy of a 5-armed specimen) giving rise to 3 additional arms instead of 2. Mortensen (1933, p. 276) has given it as his opinion that "the idea that species normally 6-armed and self- dividing as young transform into 5-armed, non-dividing adult, is a mistake — for Asteroids as well as Ophiuroids." In my opinion, my good friend is mistaken — the present species of Nepanthia seems to me to be unquestionably one in which young hexamerous individuals give rise through autotomy to symmetrical, non-dividing pentamerous adults. Occasionally hexamerous adults arise but they are the rare exception. Another growth change that deserves some comment is in the abundance and distribution of the spinelets on the aboral plates. Orally the changes are less noticeable. In all young specimens, until the arms are 20-25 mm. long the larger aboral plates are well-covered, though not densely so, with the usual minute spinelets and the smaller plates have a more or less evident tuft of the same. As growth proceeds however and size increases, we find two divergent tendencies. In some individuals, usually (but not always) those with wide, blunt arms, the spinelets seem to increase but little in number and hence occupy only the raised portion of the expanding plates, and at the same time, they in- crease in size particularly in thickness at base; the surface of the adult thus becomes rougher, with the individual plates increasingly conspicuous. On the other hand in some individuals, usually with slender tapering arms, the spinelets increase greatly in number and cover the plates large and small so completely that the surface assumes a "velvety" appearance and feeling. The difference between the two extremes is striking but as it is not associated with other dis- tinguishing characters and as there are many intermediate specimens, it can be CLAHK: AUSTRALIAN ECHINODERMS 179 regarded only as "individual diversity." It suggests however that two allied but quite distinct species might arise very easily and naturally by some change in environment that led to a keener struggle for survival. This is one of the common and characteristic sea-stars of the Broome area. We did not find it at Cape Leveque but it occurs at False Cape Bossut and in Lagrange Bay. There are 51 specimens at hand, the largest having R =64-67 mm. Parasterina crassa Pafiria ? crassa Gray, 1S47. Proc. Zool. Soc. London, p. S3. Parasterina crassa Fisher, 1908. Smitlison. Misc. Coll., 52, p. 90. On October 7, 1929, we spent a most interesting day at Geraldton, W. A., and among the echinoderms collected along shore were two small Nepanthia-like sea-stars, which prove to be young examples of Gray's species. They show a dis- tinct difference in form, for one has R = 28 mm., br = 7, R = 4br, while the other has R = 25 mm., br = 8, R = 3br. Otherwise they agree well, particularly in the considerable number of enlarged swollen, aboral plates characteristic of the species. We did not meet with another specimen of crassa during our further col- lecting on the Western Australian coast but Mr. Glauert kindly loaned me 6 specimens from the Perth Museum and Professor Bennett has sent 6 specimens. Of the Perth Museum specimens, a 6-armed one from Garden Island is most notable; it is a large individual (R = 50 mm. ±) in poor condition; one ray is dis- tinctly smaller than the others. Of the other specimens 4 were found at Cottesloe Beach and the fifth was found on the piles of an old jetty at Fremantle. Of the specimens sent by Professor Bennett, 3 are from Dongarra and 2 deserve special comment; one is perfectly tetramerous and has R about 23 mm.; the other has the five arms, 33-35 mm. long and 11 mm. thick, very blunt, the aboral surface almost completely covered by the enlarged swollen plates ; no other specimen is comparable in this particular. Of the other 3 specimens, 1 is from jetty piles at Rockingham, the other 2 have no locality label. One of these however is notable for its unusually heavy, stout structure and large size; R = 64 mm., r and br, 17 or 18 mm., R = 3.7r or br. The vertical diameter of the disk and base of rays is 14 mm. The enlarged, swollen aboral plates are numerous but not excessively so. Unfortunately none of the specimens at hand have any 180 memoir: museum of comparative zoology notes as to the color in life. The smaller specimen of the two from an unknown locality sent by Professor Bennett has a very definite dull brownish-red color and my recollection of the specimens found at Geraldton is that they were red-brown. Parasterina troughtoni Livingstone, 1934. Rec. Austral. Miis., 19, p. 179. Among the sea-stars loaned me by the Perth Museum are two specimens of this recently described species. One is a trifle larger than Livingstone's unique holotype (R = 16-18 mm.) but the other is twice as big (R = 32-34 mm.). The species characters are well marked, particularly in the larger specimen. The holotype was taken at Albany, W. A., but there is no locahty known for the two specimens in hand, though they are supposed to be from Western AustraHa. Both specimens are typical museum color. Parasterina occidentalism sp. nov. Plate 21, fig. 5 Rays 5, long and slender. R = 29 mm., r and br = 7.5 mm. R = almost 4r or br. Rays wide, with oral surface flat, sides nearly or quite parallel and tips very bluntty rounded. Aboral surface closely covered with thick plates, each plate densely covered with a coat of minute, short, truncate spinelets; papulae rather large so placed that there is one at each corner of every plate, but sunken almost out of sight so closely placed are the plates; aboral surface looks and feels very smooth, so crowded are both plates and spinelets; plates are of two sorts, rhombic more or less equilateral, or longer than wide (in the interradial areas and along sides of ray nearly to tip) and irregularly diamond-shape wider than long (distal portion of arm and thence forming a median band narrowest at disk) . Disk itself covered by irregularly polygonal plates not noticeably different from those at base of arms. Terminal plate rather large and projecting but covered by short, minute spinelets. Madreporite small, triangular, sunken, half-way between disk center and margin. Marginal plates noticeable, if not conspicuous, in a uniform double series at the margin of the disk and arms. Superomarginal plates unusually large and ' occidenlalis = of the west, in reference to its being characteristic of the western coast of Austrah'a. CLARK : AUSTRALIAN ECHINODERMS 181 noticeable, corresponding in size, form and position with the inferomarginals. Only one series of actinal intermediate plates extends to the tip of the ray, the plates corresponding in number and position with the adambulacrals and also with the inferomarginals; on basal third of ray, a second series of much smaller plates, 16 or 17 in number, extends as far as the twelfth inferomarginal ; a third series of 7 or 8 similar plates extends to the fifth inferomarginal, and half a dozen moi'e fill up the rest of the intermediate area. Adambulacral armature made up of a furrow series of 4 or 5 subequal, rather stout spines with bluntly rounded tips, and a subambulacral group of 10-12 similarly thick and blunt but much shorter spines on the surface of the plate; the inner 4 or 5 of this group are some- times arranged so that they seem to form a second series back of the furrow series but this is rarely e\-ident. Each oral plate with a marginal series of half a dozen spines subequal or the innermost longest, similar to but larger than the adambulacral furrow-series; in large specimens the inner spines become almost prismatic at tip; on the surface of each plate is a longitudinal double series of smaller spines, about 5 in the outer row and 2 in the inner. Taken altogether both adambulacral and oral plates are pretty well covered with spines. Color in life, dull rose-red (in other specimens, a more crimson-red is marked), completely lost after drying. Holotype, M. C. Z. No. 3332, from a small cavern under a big rock at Point Peron, W. A., October, 1929. Although we did not meet with this species again, Mr. Glauert has loaned me 5 specimens and Professor Bennett has given me two, which however have no data with them and are in rather poor condition. Of the Perth Museum speci- mens, 2 in very poor condition are from Shag Rocks, Penguin Island, while the others are from near Fremantle. From Cottesloe Beach is a good graj^-brown specimen, with R nearly 40 mm., the arms tapering to a blunt point. A fine specimen from Garden Island, with R = 55-56 nun. and br = 1 1 (hence R = 5br) is painted brick-red, presumably as in Ufe. The largest specimen of all is from North Beach, Fremantle, and has R = 65 mm., and br = 14; the color is yellowish- brown. This species is very well defined and cannot be confused with any other now known. Anseropoda rosacea Asfcricw rosaceiis Lamarck, 1816. Anim. s. Vert., 2, p. 558. Anseropoda rosacea Fisher, 1906. Bull. U. S. Fish Comm., p. 1089. 182 memoir: museum of comparative zoology ^\^lile at Broome in 1929, a local shell collector showed me a fragment of a specimen of this sea-star and said it came from Eighty Mile Beach. Although we sought eagerly to find a specimen, we were not so fortunate. In 1932 however, on the afternoon of June 20th, while dredging in 4 fms. on Roe Bank just outside Roebuck Bay, we took 25 specimens. Oddly enough we did not again meet with the species. My field notes say: "All of the specimens are more or less damaged though some of the largest are nearly perfect. The color underwent no change in preservation; the upper side is variegated gray like the coarse sand on which the animal Uves; the lower side is creamy white with more or less rust color along the ambulacral grooves." On drying, the gray tints tend to become brown but otherwise there is very little change. Obviously the specific name selected by Lamarck has no reference whatever to color but only to a fancied resemblance in form, to a rose. Of the 25 specunens, one is too badly damaged to permit counting the rays but of the other 24 plus a small specimen loaned by the Australian Museum 1 has 11 rays 1 has 13 1 (the smallest) has 14 1 has 14} 2i as one furrow forks! 7 have 15 10 have 16 4 have 17 Evidently less than 16 is the typical number of rays (at Broome) but with nearly three times as many having fewer than having more. The average for the 25 specimens is only 15.4 rays. The smallest specimen has R =30-35 mm. while the largest has R = 115-120 mm. As a rule, R = 1.12-1. 16r. The chief diversity shown is in the form of the tips of the rays, wUch range from rather acutely pointed to broadly rounded. The terminal plate is usually quite conspicuous and often the tip of the ray is tipped up in such a way as to make the plate more than normally noticeable. The specimen loaned by the Australian Museum is the most perfect one I have seen. It is only 75-80 mm. in diameter and the margin curves downward, so that the form is surprisingly like a certain type of Japanese umbrella. There are 15 rays and the terminal plate is a conspicuous tip to each one. Mr. R. A. Bourne, master of the "Bonza," took this specimen in 1930, between Broome and Wallal. CLARK: AUSTRALIAN ECHINODERMS 183 ECHINASTERIDAE ECHINASTER ARCYSTATUS H. L. Clark, 1914. Rec. W. Austral. Mus., 1, p. 148. We did not meet with this species but Professor Bennett has kindly given me a fine specimen from "a weedy reef" at Bunkers Bay, southwestern Western Australia taken by him in January, 1930. The arms are more than usually short and stout; R=87 nmti. but br = 19, so that R is only about 4.5br. The color is brown, the lower surfaces and the coarse aboral network Ughter than the ground color of the upper surface. ECHINASTER SUPERBUS H. L. Clark, 1916. "Endeavour" Rep, p. 64. While working in Pender Bay, June 16, 1932, our diver, Wan, brought up what still seems to me one of the most remarkable sea-stars I have ever seen as it was certainly one of the handsomest. More careful examination showed it to be a very large specimen of this species, of which the colors in life had not previously been known. The description in my field note-book is as follows: "The upper sur- face is a fairly light yellowish-gieen, with center of disk and five large blotches, arranged longitudinally on each ray and diminishing m size distally, a distinctly darker shade; on these darker areas the extraordinary spines are rose-color, elsewhere they are cream- white; lower surface, cream-white; feet white; oral armature lavender, this color extending distally a little way on the adambulacral spines; papulae pale brown; more or less greenish dorsally." Wan reported that he found this beautiful sea-star "on top of a rock." We did not meet with another specimen but two from the Perth Museum, kindly loaned by Mr. Glauert, are at hand. One of these (2985) from an unknown locality with R = 143 mm. is only a trifle smaller than the Pender Bay specimen in which R = 145 mm., r=24 and br =27.5. The color is a fairly uniform Ught reddish-brown with a lavender cast. In the Pender Bay specimen the spines are shorter, stouter, more numerous and more extraordinarily clavate than in no. 2985; in the latter, a rather conspicuous bare area runs longitudinally along each side of the series of big actinal spines which parallels the adambulacral furrow. These bare strips were rather evident in the unique holotype (see H. L. Clark, 1916, pi. XXV) but are not evident in the Pender Bay specimen nor in the second specimen from the Perth Museum. 184 memoir: museum of comparative zoology The latter, though much smaller (R = 110 mm., as in the holotype), resembles the big Pender Bay specimen in its spines, more nearly than it does no. 2985. It was taken at Broome in 1914 and has been varnished, probably in a futile effort to preserve the remarkable coloration; the present color is a dark reddish-brown, the tips of the spines brownish-yellow. The Pender Bay specimen has lost all traces of green or white but many aboral spines still retain an evident rose-red color. ECHINASTER VARICOLOR' Sp. UOV. Plate 11, fig. 1 Rays 5, stout, terete, bluntly pointed. R = 160 mm., more or less; r = 16, br= 22-24; R=l()r, but only about 7br. Aboral skeleton irregularly reticulate but even in dried specimens not conspicuous, bearing many low spines, 1-1.5 mm. high, .5-1 mm. thick at base; there are 10-15 of these spines to each square centi- meter, but they are least numerous proximally and most numerous near the arm tips; they are bluntly rounded or truncate and often concave or pitted at tip. No marginal plates are distinguishable. Madreporite very small, only 2 mm. across, about 8 mm. from center of disk. Oral surface similar to aboral but the spines are somewhat longer. Papulae are numerous to within 5 mm. of the ambulacral furrow. Adambulacral armature at middle of arm consists of a single large marginal spine about 3 mm. long, sometimes tapering to a blunt point but more often strongly flattened into a chisel-like tip more than a millimeter wide, usually at right angles to the furrow; within the furrow attached to the base of the big spine is a rather short and blunt spine even its tip lower than the surface of the actinal plates; on many plates a small subambulacral spine, similar to those on the actinal plates, is present back of the big spine ; at t he base of the arm this spine is present on every adambula- cral plate but at the middle and distally its occurrence is irregular. The lowest row of actinal plates (nearest the furrow) carries spines rather evidently larger than the others and at some places on some of the arms this series of actinal spines is noticeable, but it is irregular in its development and is never con- spicuous. Oral plates with 3 or 4 big, clavate spines on the free margin, below which deep in the mouth are two small spines ; on the surface of each plate there may be a small spine but this is frequently lacking. Color in life deep buff, spotted with red-purple; dried specimen uniform bright red-brown becoming yellow-brown along the ambulacral furrows. ' varicolor = of diverse colore, in reference to the different color forms noted in living specimens. CLAKK: AUSTRALIAN ECHINODERMS 185 Holotype, M. C. Z. no. 3339 from sandy bottom, southwest of Broome, 5-8 fms., June, 1932. The diversity of color shown in life by this fine Echinaster is so remarkable that I supposed I had 3 species until careful study showed that aside from color there are no characters whatever by which specimens can be distinguished. The first specimens taken were from far south of Broome and were buff, spotted with red-purple; while there was individual difference in the number and size of the spots, there was no question as to the identity of the specimens. The following week, while working to the north of Broome, Wan, the diver, brought up a fine, large Echinaster, uniformly violet in color, though somewhat yellowish along the ambulacral grooves. Subsequently specimens intermediate between the spotted and unicolor forms were taken and it became probable that only one species was involved. Finally, in Roebuck Bay, we took an Echinaster of rather smaller size, uniformly bright red in color without trace of buff or violet; several speci- mens were ultimately secured. On being dried, the three color forms became uni- color, the shade ranging from yellow-brown, through brown, to deep red-brown; some specimens are a very bright red brown, but there is no correlation between the color in fife and the present shades. In all specimens the adambulacral armature and more or less of the adjoining actinal surface is yellowish or brown- ish-yellow. All of tlie specimens have 5 arms but several have met with mishaps and are regenerating lost rays; the worst case has one arm normal, about 115 mm. long, a second arm about 100 mm. long gives some indication of having regenerated from close to the disk, a third arm has 30 mm. of the old arm base and 35 mm. regenerated distally, a fourth has 20 mm. of the old arm and 30 of the new, while the fifth arm lost close to the disk has just begun its regeneration. Whether marauding fishes are the cause of such mishaps one can only guess. No very small Echinasters were taken — the smallest has R = 86 mm. while in the largest R = 190 mm. The proportion of major and minor radius ranges from R = 8 to lOr, and the slenderness of the arms ranges from R = 6 to 8.5 br. It is worth noting that br seems to always exceed r, often very considerably. This sea-star is fairly common in the Broome region but we did not meet with it in 1929, probably because it does not occur along shore and we did com- paratively Uttle dredging. In our extensive dredging and use of a diver in 1932, we secured 26 specimens. For the convenience of future field workers the 3 color forms may be designated as forma maculata, forma violacea and forma rubra. The relationship of varicolor to luzonicus is perfectly obvious but the dif- 186 memoir: museum of comparative zoology ferences in life are even more so. Aside from size and color however, it is difficult to specify any point in which there is a clear cut distinction. When dry specimens of the same size are compared, it is evident that the spinelets are larger and less numerous, the skeletal network more open, in varicolor. Near the tip of the ray in luzonicus, the subambulacral spinelets form a very definite series close to, and parallel with, the adambulacrals, and often a series of actinal spinelets runs along just outside the subambulacrals. In varicolor there are no continuous series of spinelets outside the adambulacrals, the subambulacrals and actinal spinelets being scattered and irregular in their linear arrangement. The adambulacral spines in varicolor are more flattened and much wider at the tip than in luzonicus but this is a difference of degree only. To sum up, it may be said that luzonicus is a small species with R = 75-100 mm., usually with 5, but often with 6 or even 7 arms, reproducing autotomously; the rays are relatively slender, covered with numerous small spinelets; distally the subambulacral and adjoining actinal spines tend to form longitudinal series parallel to the adambulacrals ; adambula- cral spines often somewhat flattened at tip; color rusty-red speckled with black- ish, or often the blackish color spreads over the whole dorsal surface; there is no violet or purple in the coloration. On the other hand varicolor is a very large, strictly 5-rayed species, not autotomous, with R = 150-190 mm.; the rays are stout but terete, covered with coarse spinelets, relatively less numerous than in luzonicus; on the distal portion of the arms, the spinelets do not form definite longitudinal series parallel to the adambulacral spines; the latter are markedly compressed and widened at tip particularly on the basal part of the arm; color deep buff spotted with purple, or uniformly violet or bright red, with no black or dusky spots or shade. The uniformly violet specimens suggest the possibihty of identity with Echinasier purpureus of East Africa and Mauritius but the shape of the arms and the absence of a wide bare strip parallel to the ambulacral furrow are adequate differences. Plectaster decanus Echincwter decanus MitLLER and Troschel, 1843. Arch. f. Naturg., 9 (1), p. 114. Plectaster decanus Sladen, 1889. "( 'hallenger" Ast., p. 535. Of course I regretted not having the opportunity to see this odd sea-star alive but fortune did not favor me. Professor Bennett has however sent me a good specimen with R=90 mm. =*= taken by Mr. D. L. Serventy, February 23, 1930, while on the trawler "Bonthorpe" in the western end of the Great Austra- CLARK: AUSTRALIAN ECHINODERMS 187 lian Bight, 33° 15' S x 126° 22' 15" E, 90 fms. Mr. Serventy has attached a color note to the specmien reading ''Red and purple patches," wliich I construe as meaning that in life the skeletal parts are red and the large papular areas purple. METRODIRIDAE Metrodira sxjbulata Plate 11, fig. 3 Gray, 1840. Ann. Mag. Nat. Hist., 6, p. 282. Near the entrance to Roebuck Bay, in 4-7 fms., this pecuHar sea-star was not rare and we secured 16 specimens, ranging from R = 16 to R =72 mm. As the largest previously known specimen seems to be the one described and figured by Koehler (1910, p. 284) which had R= 60-65 mm., these larger specimens are of particular interest. The 16 specimens show no Uttle diversity in the slenderness of the arms, in the degree to which their aboral surface is flattened and in the size and number of the aboral spinelets. At one extreme is a specimen with slender, terete arms, in which R =55 mm., r =8 mm., and br =7 mm., and at the other is an individual with broad, flattened arms, in which R=60 mm., r = 10 mm. and br = 10 mm. Hence the relative proportions range from R=6-7r and 6-8 br. In the slender-armed specimens, the aboral spinelets are small, usually single, about 10-12 in a space 3 mm. square while in the broad-armed specimens the spinelets are larger, often in transverse series of 3 or 4, and 15-20 in a 3 mm. square. The two forms intergrade completely. A specimen with R = 68 mm. is pei'fectly tetramerous. The most interesting thing about the finding of Metrodira was the discovery of its unexpectedly handsome colors in life. Museum specimens are so universally yellow-brown, pale-brown, dirty yellowish or nearly white that it has never been suspected that the colors are very unUke any known Echinasters. My field notes describe the hving animal thus: "A most attractive sea-star ; pure white beneath; light gray with dull blue-gray blotches; each papula surrounded by a rust- colored ring." The gray has a distinctly bluish tint so that the effect is of a blue and white sea-star with small orange-brown circles on the upper side. Besides the 16 specimens taken at Broome, there are 3 other specimens at hand. One is a loan from the Perth Museum and was taken at Canarvon in 1927 by Dr. Ehrenreich. Aside from the remarkable extension of the species range to 188 memoir: museum of comparative zoology the southwest, this specimen is of more than usual interest because of its short wide arms: R= 54-56 mm. while r and br = 10; hence R=5.5 r or br. However the specimen was evidently dried slowly and the width of the arms is thus un- naturally increased a little. The other 2 specimens show an equally notable extension of the range of Metrodira to the southeast, for Mr. Ward took them at Lindeman Island in "July-Sept., 1934." They are further notable for large size and very slender arms ; R = 82 mm. in the larger specimen and br is only about 9 mm. ; hence R = 9br. These specimens, now dry, are a dingy, but rather dark brown, very different from the other specimens, but no morphological differences have been found. PTERASTERID.4E Retaster insignis Sladen, 1882. .Jour. Linn. Soc. London, 16, p. 200. The appearance of this sea-star in life is so unusual, so different from museum specimens, that the taking of several just outside Roebuck Bay on June 13, 1932, in 5-8 fms. was a most interesting surprise. The brilhant color and the extraordinarily slimy surface were quite unexpected, aside from the fact that the species was previously known only from the East Indian region, Torres Strait and eastern coast of Austraha. Additional specimens subsequently taken in the same vicinity showed the same remarkable characters. Altogether 13 specimens were secured ranging from R =42 to R =80 mm. The ratio of R to r shows little diversity; it is least in the smallest specimen where R=2.2r or 2br while in a specimen with R =68 mm. the other extreme is reached, as R =2.7r or 2.4br. One specimen with R=80 mm. and one with R=60 mm. have 6 rays each but are not perfectly symmetrical, one arm (or two) being somewhat smaller than the others. The color in life is given as follows in my field notes: "One was bright vermilion red, one deep vermiUon while others had net work red but areas greenish yellow." Most of the specimens were more or less completely bright red. In addition to this very striking color, the appearance was made even more remarkable by the fact that every specimen was covered with a layer of per- fectly transparent colorless jelly making the specimens slimy and unpleasant to handle. On one specimen this layer was 6 mm. or more in thickness, so that none of the spines which cover the animal projected beyond it. The jelly had CLARK: AUSTRALIAN ECHINODERMS 189 little consistency however and disappeared quickly as the specimens were handled and being preserved. It left no trace behind in the dry specimens. But it may have a protective function in life. • The Perth Museum has loaned me a specimen of this species from Shark Bay, notable not only for the locality so far to the southwest of Broome but also for its large size; with R =83 and br =41.5 mm., it is distinctly the largest speci- men yet recorded. Mr. Ward has sent from Lindeman Island, Queensland, the smallest specimen which has yet been taken. It is a perfectly symmetrical pentamerous individual with R = 14 mm. and r and br scarcely 6; it is interesting to find the ratio identical with that of adults. The species characters are well shown even in so young a specimen. ASTERIIDAE COSCINASTERIAS CALAMARIA Asterias calamaria Gray, 1840. Ann. Mag. Nat. Hist., 6, p. 179. Coscinasterias calamaria Perkier, 1894. "Travailleur et Talisman" Stell., p. 106. This sea-star is widespread in the non-tropical waters of Australia, as it is common at Lord Howe Island on the east and at Rottnest on the west, and apparently at all suitable places between along the southern coasts of the con- tinent. The number of rays ranges from 7 to 11, but large symmetrical speci- mens, with R exceeding 100 mm., commonly have 11. However autotomy and regeneration are so continually taking place that even very large specimens often have one or more small rays. The specimens at hand from Hobart, Tasmania, are particularly large and fine. They are like the one called gemmifera (Perrier) in the "Endeavour" Report (H. L. Clark, 1916, p. 74) but after careful com- parison of many specimens of all sizes from various places, 1 am convinced that the Tasmanian specimens are not essentially different from other Australian material and all should be called calamaria. Incidentally I may add that I ha^•e compared specimens of similar size from Mauritius, Western Australia, South Austraha, New South Wales, Lord Howe Island and New Zealand, and find no characters by which to satisfactorily distinguish them. The number of pedi- cellariae and the distribution of the large ones is very diverse and this gives a different facies in extreme cases. Thus a Mauritius specimen has few pedicel- lariae, no large ones, while a New Zealand specimen has a great many large ones and few small ones so that at first sight they seem quite unlike. New Zealand 190 memoir: musexim of comparative zoology specimens show much diversity among themselves in spinulation as well as in pedicellariae. The color in life of calnmaria shows some diversity. Quoting my field notes we find that at Hobart: "Largest was brownish-yellow, the huge circles of pedi- cellariae on the dorsal spines (which swell up around and even above the spine tip, so that it looks like a small zoanthus) dull bluish. Other specimens were olive-greenish with spines blue and lower surface cream-color." At Lord Howe: "Color very varied usually blue or bluish variegated with brown of several shades (rarely reddish) and olive-greenish. Some specimens show almost no blue. Colors fugitive and blue especially fugacious." At Port Willunga, S. A.: "Brown and dull bluish with many large spines bright, deep blue." At Rottnest and Point Peron: "Colors very variable and fugacious; a common color is dull blue with disk and more or less of the ray-bases, bright brown, but greenish shades instead of blue, are frequent; variegation with bluish, brown and grayish shades is the rule." The 63 specimens at hand are from the following widely separated localities : Lord Howe Island: April, 1932. 5 specimens, 4 very small. New South Wales: Port Jackson, Bottle and Glass Rocks, November 27, 1929. G specimens, young and unsymmetrical. Gunnamatta Bay, November 26, 1929. 4 specimens, 1 large adult and 3 half grown. Shell Harbor, May 4, 1932. 6 specimens, young and un- symmetrical. Tasmania: Hobart, estuary of the Derwent, November 15, 1929. 3 specimens, 2 very large and symmetrical. South Australia: Port Willunga, November 2, 1929. 1 specimen, small but symmetrically 10-armed. Western Australia: Bunkers Bay, January, 1930. E. W. Bennett leg. et don. 4 specimens, adult and young. Point Peron, October, 1929. 5 specimens, young. Rottnest Island, October, 1929. 19 specimens, young and very young; one only 8 mm. across, with 7 arms. Rottnest Island, 1934. Captain B. E. Bardwell leg. et don. 7 specimens, small adults and young. Garden Island, July, 1932. 2 specimens, very young. Cottesloe Beach, July, 1932. 1 specimen, young; 4 rays, and 5 buds of rays just starting on opposite side of disk. CLARK: AUSTRALIAN ECHINODERMS 191 ASTROSTOLE INSULARIS^ Sp. IIOV. Plate 8, fig. 1 Rays 7, moderately wide, bluntly pointed and flattened in dry specimen, more nearly cylindrical in life. R =77 mm. ; r = 10; br =8 at base but 10-12 mm. near middle of arm. R = nearly Sr and 9.5 br. Disk of moderate size with a firm skeleton and small interspaces. Aboral skeleton of arm consists of 5 series of stout plates, a carinal and a dorsolateral on each side, below which are the stout superomarginals, the largest of the skeletal plates. Carinal plates roughly pentagonal or diamond-shape with rounded or truncate angles, closely united with each other in the midUne; every other one (distally every third or fourth) carries a stout bluntly pointed spine, 2-3 mm. long, with a heavy wreath of pedicellariae about its base; dorsolateral series of small irregular plates, a very variable number of which carry a single small wreathed spine. Superomarginal plates irregularly rounded triangular, wider than long, broadly in contact with each other at the uppei- end, in contact with the inferomarginals by a narrow outer end and still more narrowly in contact at the upper, distal corner with the small dorsolaterals; every second or third plate carries a large wreathed spine like that on the carinals; a considerable area of each plate has a distinct shagreen surface. Madreporite single, 3 mm. across, 5 mm. from center of disk. Inferomarginals large, longer than wide, forming (with the spines they bear) a conspicuous margin to the arm; each plate has an oblique ridge, the lower end of which is more distal than the upper; on this ridge is a pair of somewhat flattened spines, the upper one 4-5 mm. long, a millimeter or more in width, with a bluntly pointed or rounded tip; the lower spine is about 3.5 mm. long, less than a millimeter wide and has a square cut tip. Actinal plates small but distinct on basal two-thirds of arm, widely spaced so that each plate is in close contact with lower end of ridge on inferomarginal plate; it carries a flattened square-cut spine somewhat smaller than the lower inferomarginal as though it were a third mem- ber of that series; the uppermost of the 3 spines (i.e. the upper inferomarginal) carries a cluster of forcipiform pedicellariae, on its upper sm-face near the middle or at its base. Adambulacral plates short and crowded ; each carries a pair of narrow chisel- shaped spines, about 2.5 mm. long, the inner a trifle shorter and distinctly nar- rower than the outer; the tips of the latter are a trifle widened. Oral plates small, ' insularis = pertaining to an island, in reference to the type locality at Lord Howe. 192 memoir: museum of comparative zoology each with 2 spines, a large one at the inner end and a smaller more pointed one on the outer corner. There are 4 contiguous pairs of post-oral adambulacral plates, each carrying one large spine on its margin. Pedicellariae relatively few; the forcipiform are confined to the clusters on the aboral and marginal spines; few if any are scattered on the surface; forfici- form pedicellariae of diverse sizes are scattered about on the aboral surface especially near the base of the rays; none are very large or conspicuous. On the oral surface, a very few large forficiform pedicellariae with jaws over a milhmeter long occur in, or close to the ambulacral furrows, near the mouth, and a number of small ones are present on the restricted actinal intermediate areas; distally there are virtually no pedicellariae. Color in life variegated brown and blue; the brown has a gray tinge and the blue has a slight tendency towards green, but the large marginal spines are a very bright blue ; madreporite pale brown or yellowish. The colors are very fugacious and dry specimens have lost all their beauty. In a specimen larger than the holotype, the large aboral and marginal spines are quite red in contrast to the general museum color. Holotype, M. C. Z. no. 3362, from Lord Howe Island, April, 1932. This handsome sea-star is one of our most interesting discoveries at Lord Howe. It lives with and like Coscinasierias calamaria on the reef flat near low tide level, under and among fragments of coral rock. It passes easily for a 7-armed form of that species, with unusually blue coloration. It is closely related to A. rodoJphi of the Kermadecs with which it agrees in the scarcity of pedicella- riae but it is readily distinguished from that species by its more open skeleton and fewer spines; thus in rodolphi, 20 mm. of the carinal ridge or of the inferomar- ginal plates includes about a dozen plates while in insidaris only about 8 will be included; again in rodolphi the adambulacral and inferomarginal spines have much wider and more expanded tips than in insularis; most important of all, is that the dorsolaterals in rodolphi are far better developed than in insularis of the same size, being more numerous, larger and stouter and more generally spine- bearing. From scabra of New Zealand, insularis is much more obviously different; the fewer pedicellariae, of all sorts, the much larger spines, the much better de- veloped spiniferous actinal plates (almost or quite wanting in scabra with R = 75 mm. or less), combine to give the Lord Howe species a very different facies. From the Astrostoles of the eastern Pacific, platei and paschae, insularis is easily dis- tinguished by the number and form of the rays and accompanying details of spinulation and pedicellariae. Perhaps I may well add here that in my descrip- CLARK: AUSTRALIAN ECHINODERMS 193 tion of paschae (1920, p. 105) I have made the mistake of interpreting the actinal spiniferous plates as being actinal lobes of the inferomarginals, and have stated that there are 3 inferomarginal spines in each series; actually the arrangement is as described above for insularis. Besides the holotype which is a small adult, there are at hand 3 paratypes. One of these is an apparently mature but somewhat asymmetrical adult, with R ranging from 70 to 110 mm. The dorsolateral series are much better developed on the basal part of the arm and the inferomarginal spines are much wider at tip, where they are also more or less channeled; the superomarginal spines have also taken on the same character. The madreporite is big and divided into a larger and a smaller part but they are in close contact. The outer series of adambulacral spines has the tips noticeably widened, though not quite so much so as in rodolphi. Pedicellariae, as in the holotype, conspicuous by their absence! A second speci- men is a young individual with R = 35 mm. The dorsolaterals are small and very rarely spiniferous. The other aboral spines are relatively very large. There is no trace of an actinal series of plates and hence only 2 spines occur in connection with each inferomarginal plate. Pedicellariae are even more infrequent than in the adults. The third specimen is a freak; it consists of a disk about 13 mm. in diameter and one arm about 50 mm. long. Growing out from the disk are G very young arms, with their lengths ranging from 8 to 16 nmi. Apparently an Astro- stole with 7 normal arms 50 mm. long lost 6 of them (whether autotomously or not, who can say?) and was replacing them by typical regeneration when the specunen was taken. Allostichaster polyplax Astcracantliiou polyplax Miller and Troschel, 1844. .\rch. f. Naturg., 10 (1), p. 17S. Allostichaster polyplax Verrill, 1914. Harrinian Alaska E.xp. Starfishes, p. 363. Tliis is one of the few sea-stars conmion to New Zealand and Austraha. Its range, hke that of Coscinasterias calamaria, extends from New Zealand, New South Wales and Tasmania westward to Rottnest Island. Unlike calamaria however, polyplax does not occur at Lord Howe, at least it has not yet been found there. Comparison of individuals of the same size from New South Wales, South Austraha and Western Austraha show no character or trend to indicate more than a single form. Judging from the small series of specimens at hand, in South Austraha 95% of the specimens which are not the result of recent autotomy have 8 arms; in Western Austraha, 90 7o; in New South Wales, 81%; and in 194 memoir: museum of comparative zoology Tasmania, only 77%. But I do not see any taxonomic significance in these figures even supposing large series of specimens showed them to be reliable. My field notes on this secretive httle sea-star may be of some interest. While at Perth, I wrote: "Not common, but found at Rottnest and at Point Peron — color in life brown with a markedly greenish cast, cream-color under- neath. Colors very fugacious and specimens become yellowish with more or less orange mottUng on death, and even these shades change or disappear." At Bunbury, I noted that the color was "hght greenish-brown becoming more orange- brown along ray-margins." At Port Willunga, S. A., my notes say it is "quite common. All (?) have 8 arms; browner than those on west coast, not so evidently green. One large specimen quite different as it was an orange-brown, not at all bright however, but with no hint of oUve or green." At Hobart, I found that polyplax was common; "color usually brownish and red but some are greenish as at Perth and Port Willunga. The small specimens have red more evident than in adults. Common in dredge hauls, on or in shells, crannies of any sort." The present series of 106 specimens contains many part-specimens for autotomy is repeated constantly, and perfectly symmetrical specimens are rarely seen; these commonly have R =20 nun. or more. The smallest specimen at hand has 3 arms, each about 8 mm. long; the broken surface has healed but no new arms have begun to bud. In the same lot is a little sea-star with 2 arms, side by side, 10 mm. long and 6 arms opposite to them each about 5 mm. long. Speci- mens with 7 arms are rare, but one from Hobart has 5, 18 mm. long and 2, 15 mm., while another from the same place has 4 large (12-13 mm.) and 3 small (8 mm.) arms. The largest specimens are from Bottle and Glass Rocks, Port Jackson ; the largest of these has, on one side, 4 arms with R = 36 mm. and op- posite to them are 4 with R = 12-15 mm. Of these Port Jackson specimens, 4 of maximum size, have 4 (or 3) large and 4 (or 5) small arms, indicating that asexual reproduction by autotomy continues long after maturity is reached. The 106 specimens of polyplax at hand come from the following places: New South Wales: Colloroy, Long Reef, November 28, 1929. 3 specimens. Port Jackson, Bottle and Glass Rocks, November 27, 1929. 5 specimens, including several large adults. Shell Harbor, May 4, 1932. 8 specimens. Tasmania: Hobart, estuary of Derwent, November 15, 1929. 27 specimens; adult and young. South Australia: Port Willunga, November 2, 1929. 28 specimens, mostly small adults. CLARK: AUSTRALIAN ECHINODERMS 195 Western Australia: Middleton Beach, Albany, January, 1929. E. W. Bennett leg. et don. 5 specimens. Bunkers Bay, January, 1930. E. W. Bennett leg. et don. 23 specimens. Bunbury, October 25, 1929. 1 specimen. Point Peron, October, 1929. 1 specimen. Rockingham, Cymodocea beds, 4-5 feet of water, February 7, 1932. E. W. Bennett leg. et don. 1 small specimen. Rottnest Island, October 19, 1929. 2 specimens. Cottesloe Beach, E. W. Bennett leg. et don. 2 small 7-rayed specimens. Smilasterl\s irregularis H. L. Clark, 192S. Rec. S. Austral. Mus., 3, p. 402. Among the interesting echinoderms which were loaned by the National Museum at ]Melbourne is a symmetrical pentamerous sea-star of small size which may be considered the second known specimen of this species. The identification is by no means assured but as the holotype was nearly twice as large as the present specimen (in wliich R=25 mm., and br 5 mm.), too exact a correspon- dence should not be expected. Certainly in the characters of the oral surface,the identity is very close. The chief unhkeness is in the character of the aboral skeleton, which is delicate, with regularly arranged longitudinal series of papular areas, wider than long. Until more and better material is available, it seems better to call this Victorian specimen irregularis than attempt to differentiate it as a new species. If the type of irregularis were at hand for comparison, the question might be settled satisfactorily but as it is in Adelaide, my Australian colleagues must do the comparing. The present specimen was taken at San Remo, Westernport, Victoria by ]\Ir. G. Coghill, January 28, 1909. I may be permitted to add that there are in the M. C. Z. collection a number of small asteroids, in poor condition, which were dredged at Westernport and in Port PhiUp by Mr. J. Gabriel, who presented them to us, with other imidentified material, in 1915. Some of these are large enough and in good enough condition to assure their identity with the specimen from San Remo but they make me even more doubtful whether that specimen is really irregularis and not rather a distinct species. For the present, the matter must be left in tliis unsatisfactory condition. 196 memoir: museum of comparative zoology Uniophora granifera Aster/as granifera Lamarck, 1816. Anim. s. Vert., 2, p. 560. Uniophora granifera Bell, 1881. Proc. Zool. Soc. I^ondon, p. 497. It was disappointing not to find Uniophora at either Port Willunga or Ho- bart, but that failure gave an added thrill to finding a specimen at Bottle and Glass Rocks, Port Jackson, November 27, 1929, when Mr. Melbourne Ward took me there for a morning's collecting. This individual was a bright, deep carmine in life, shaded on depressed areas with dusky. It is now light brown with the spherical tubercles of all sizes very light, almost a brownish-white. It has R =55 mm. and r = 14; hence R=4r. As a typical example of the species, it is very satisfactory but as the species was already well known from Port Jackson, it naturally throws no light on the distribution or taxonomy of this still imper- fectly known genus. OPHIUROIDEA The collection of ophiurans contains 3482 specimens representing 45 genera and 132 species. Of the genera 7 are new but nearly half of them are the result of segregation from the big unwieldy genus Ophiothrix, 3 fairly well-defined groups. The remaining species listed under Ophiothrix are a heterogeneous group which needs further study and subdivision. Of the 4 genera, actually new morpholog- ically, it is interesting to note that 2 are from Lord Howe Island (though one occurs also on the Queensland coast) one is from the historic Port Essington and one is from Broome. Of the 132 species no fewer than 54 are undescribed. This seems like an un- reasonably high percentage of novelties but when one considers the richness and diversity of the Australian fauna and how little attention the ophiurans have received, it is not so surprising. That there are 14 species to be added to the already unwieldy genus Amphiura is to be regretted and emphasizes not only the ubiquity of the group but the sore necessity for its prompt and careful revi- sion. It is remarkable that so many Amphiuras occur while only one Amphipholis and one Amphiodia were taken, and each is a wide-ranging species. The new genera are : Ophiothauma, type heptactis sp. nov. Monotypic. Amphistigma, type minuta sp. nov. Monotypic. LissoPHiOTHRix, type delicata .sp. nov. Monotypic. CLARK : AUSTRALIAN ECHINODERMS 197 Macrophiothrix, type Ophiura longipeda Lamarck. 21 species. Placophiothrix, type Ophiothrix melanostida Grube. 8 species. Ophiodyscrita, type acosmeta sp. nov. Monotypic. Ophioteichus, type parvispinum sp. nov. 2 species. The 54 new species, belong to only 25 genera. They are as follows, the type locality being also shown. Euryale euopla Astroco7ius occidentalis Astrodadus granulatus Ophiacantha ameleta tenuispina Ophiothauma heptactis Amphiura acrisia bidentaia bradiyadis cataphes diacritica dolia leucaspis magnisquama micra multiremula nannodes phrixa plena stidacantha Ophiocentrus fragilis Ophionephthys decacantha tenuis Aviphistigma minuta Amphioplus didymus sienaspis Ophiadis acosmeta brevis fuscolineata laevis Lissophiothrix delicata W. A.: "Bald Island, east of Albany." W. A.: "North Beach near Perth." Q. : Lindeman Island near Mackay. Celebes; Siboga Sta. no. 117. Q. : Port Curtis N. T. : Coburg Peninsula, Port Essington. W. A. : Broome W. A.: Broome W. A. : Broome N. S. W. : Port Jackson Q. : Whitsunday Passage N. S. W. : Port Jackson W. A.: Lagrange Bay N. S. W.: off Botany, 33-56 fms. W. A. : Broome N. S. W.: CoUoroy, Long Reef W. A. : Rottnest Island W. A. : Broome W. A. : Bunkers Bay W. A. : Broome N. S. W. : off Port Jackson, 120 fms. W. A. : Broome W. A. : Broome Lord Howe Island W. A.: Broome N. T.: Darwin W. A.: Dongarra N. T. : Darwin W. A. : Broome W. A. : Bunbury W. A. : Broome 198 memoir: museum of comparative zoology Macrophiothrix callizona cahjptaspis lampra scotia spinifera Ophiotrichoides irregularis pulchra Placophiothrix albolineata Ophiogymna lineata Ophionereis hexadis stigma tigris Ophiocoma occidentalis Ophiomastix notabilis Ophiarachna megacantha Pcctinura nigra Ophiarachnella paudgramda rugosa Cryptopelta callista Ophiodyscrita acosmeta OphioteicJms multispinum parvispinum Ophiolepis unicolor W. A. : Broome W. A. : Broome N. S. W.: Port Jackson W. A. : Broome W. A.: Broome Q. : Port Curtis W. A. : Broome Lord Howe Island Q. : west of Low Islands, 6-8 fms. N. T. : Darwin W. A. : Cape Leveque Q. : Northwest Islet W. A. : Point Peron W. A.: Cape Leveque Q. : off Double Island Point, 33 fms. W. A. : Bunbury Q. : off North Direction Island, 19 fms. W. A. : Broome W. A. : Broome W. A. : Broome Q. : Lindeman Island Lord Howe Island W. A. : Broome In several cases, varieties have required recognition; of these the following 5 are here named for the first time. Ophiactis savignyi var. lutea Ophionereis semoni var. nigra var. badia Ophiurodon cinctum var. pulchelhwi Ophiarachnella ramsayi var. pulchra N. T. : Quail Island W. A. : Point Peron W. A. : Dongarra W. A. : Cape Leveque W. A. : Rottnest Island As shown by the above list Lord Howe Island is the home of only 3 of the new species but 2 of those represent new genera. The coast of Queensland fur- nishes type localities for 9 species and that of New South Wales for 6. From the Northern Territory, there are 4 species and a variety, while the coasts of Western Australia are the home of 31 new species and 4 new varieties of which no fewer than 20 claim Broome for their type locality. CLARK: AUSTRALIAN ECHINODERMS 199 Brittle-stars are obviously the most abundant echinoderms of the Australian coast. Owing to the small size and secretive nature of many species they are easily overlooked and the average collector who is not interested in the group never realizes how many are passed over. The larger forms such as the more common species of the Ophiocomidae, Ophiodermatidae and Ophiotrichidae living as they do under rock fragments or among corals, and being more or less active in their movements, attract the attention of most "along-shore" collectors, but they make up only about one-fourth of Australia's ophiuran fauna. Naturally condi- tions at Broome, so favorable for sea-stars are equally suitable for brittle-stars, and about one-half of the 132 species in the following account were taken in the Broome region. Of course the Barrier Reef area is abundantly populated with ophiuroids but as we did no collecting there the following list gives a very inade- quate idea of that fauna. On the southern coast of the continent brittle-stars are much less abundant than in the tropics but some very interesting forms occur there; that fauna is of course quite inadequately represented in the present collec- tion. Western Australia between Geraldton and Albany has a large and varied ophiuran population well represented in our collection by some 30 species, of which about one-fourth are here described for the first time. Brittle-stars occur under the most diverse environmental conditions. Many are to be found only among corals, particularly the gorgonians, and large sponges, especially those with a branching habit, often carry great numbers of ophiurans especially of the genera Ophiactis and OphiothrLx (and its allies). Curiously enough certain sponges are completely free from ophiurans, presumably because of some secretion obnoxious to the echinoderms. Very few ophiurans, except the multibrachiate forms like Euryale, live exposed freely on the sea- bottom. Some Uve buried just below the surface on sandy bottoms, while others have become adapted to a subterranean Ufe and are found only deeply buried in sand or even in mud. Such forms have in many cases become strikingly adapted to such a life, an increase in the length of arms and a decrease in the calcareous matter covering the body being frequent. But the best hunting ground for brittle- stars is among corals and coralline algae, particularly in the dead basal portions, and under rock and coral fragments in such areas. Many small forms can be se- cured only by placing quantities of such material in pails or basins with just enough sea-water to cover it. With the passage of a few hours, the oxygen in the water apparently becomes deficient and all the small animals come out of their hiding places and are easily secured by the collector. 200 memoir: museum of comparative zoology The literature dealing with Australian ophiurans is not extensive. Lyman's (1882) great work on the "Challenger" Ophiuroidea is of course absolutely essen- tial. Koehler's various reports are invaluable particularly those on the "8iboga" littoral ophiurans (1905), on the Ophiuroidea taken by Michaelsen and Hart- meyer in Western Australia (1907), the "Albatross" Phihppine material (1922) and on Mortensen's great collections (1930). Doderlein's works on the Euryalids (1911, 1927) are essential for that group. Bell's report on the "Alert" echino- derms (1884) cannot be ignored and my various reports on Australian collections (1909, 1914, 191G, 1921, 1923 and 1928) contain considerable additional material. The death of Koehler in 1931 was a great blow to all students of Echinoderms and it is to be regretted that at the present moment there is no outstanding student of the Ophiuroidea. In collecting brittle-stars, little care need be used except in the case of the long-armed Amphiuridae. These have a deplorable habit, especially if they are fully mature sexually, of shedding the disk, and of course if this happens in the dredge, the chances of its recovery are small. If it is psssible to handle the ani- mals carefully from the time they are first discovered, shedding of the disk is very rare. The breaking of the arms is of course the most frequent mishap ; with dredged material it is very general except in the forms with short stout arms and even here it is frequent. The only remedy is great care in handling. Some species resent any handling and break off the arms on the sliglitest provocation, but this is rare. Brittle-stars are very easily affected by Epsom salts (MgSOi) and once they are completely narcotized can be handled with impunity; if they are to be dried after kilhng the arms should be arranged while still flexible in whatever position they are desired. With specimens of considerable size, it is a convenience to have the arms arranged, as far as possible, parallel to each other all pointing in one direc- tion. In methods of killing and preservation, brittle-stars may be treated like small sea-stars (see p. 60) but it is an interesting fact that their colors (excepting bright reds) are much less fugacious than in the Asteroidea. Many brittle-stars keep their colors very well in alcohol, and color patterns are very generally retained. In formahn the colors usually are altered but little but the specimens themselves are quickly damaged and often ruined by the action of the fluid on the tissues. Consequently it is better not to leave brittle-stars in formalin more than a few hours at the most. CLARK : AUSTRALIAN ECHINODERMS 201 OPHIOMYXIDAE Ophiomyxa australis Plate 13, figs. 1-2 LuTKEN, 1869. Add. ad Hist. Oph., pt. 3, p. 45. This active brittle-star was met with at Lord Howe Island but was not com- mon, only 2 specimens being seen. They were found hidden in crannies in the coral at Ned's Beach, and are about half grown; disk, 1.5 mm. in diameter, arms, some 80 mm. long. In life one was "red brown with a purplish cast above, dark- est on disk, rather light on distal part of arms; scattered gray flecks on disk; o or 6 narrow, gray bands on arms, faint and widely spaced, most evident distally; lower surface of arms flesh-red." The other specimen was "oHve, yellow-brown and yellow; lower surface of arms, yellow-brown; gray flecks on disk and variega- tion on arms; no definite bands; no red." In their present dry condition, the first specimen is deep brown above, with gray flecks and markings evident; a red tint is still visible on lower surface of arms; arm-spines tend to be very light, almost white on distal half of arms. The second specimen is a lighter, more oUve, brown and of course shows no trace of red. A small brittle-star with disk 6 mm. across and arms all broken, found in "a conglomerate boulder, taken by trawler "Goonamba," about lat. 33° 44' S., long. 151° 88' E. (about 16-18 miles northeast of South Head, Port Jackson, about 15 miles from land), 75-80 fms.. May, 1924" and presented to the Australian Mu- seum by C. W. Mulvey, is apparently a young Ophiomyxa, but it is in too poor condition to permit positive identification. The National Museum at Melbourne has kindly loaned an Ophiomyxa "dredged in Port Phillip by J. A. Kershaw," which is a young example of australis with disk about 12 mm. across. At Port Willunga, South Australia, I secured a very large adult of this spe- cies, the largest I have ever seen. The disk was about 35 mm. across and the arms 140 mm. long or more; in its present dried condition the disk exceeds 30 mm. In life it was "deep oUve green above, dull flesh-red and orange (rather brightly colored) beneath, especially around mouth; arms indefinitely banded hght and dark distally." We did not meet with Ophiomyxa at any point on the Western Australian coast nor has it been included in any of Professor Bennett's collections. 202 memoir: museum of comparative zoology TRICHASTERIDAE ASTROBRACHION ADHAERENS Ophiocreas adhaerens Studer, 1884. Abh. K. Preuss. Akad. Wiss. Berlin, p. 54. Mr. Melbourne Ward has kindly sent me a very fine specimen of a young simple-armed trichasterid which he dredged in 9 fms. in Whitsunday Passage, off Lindeman Island, Queensland. It measures 6 mm. across the disk and the much coiled and twisted arms are well over 100 mm. long. The color is a rather bright yellowish-brown. Comparison with a cotype of Studer's species (which specimen, by the way, has 6 arms, but only 5 jaws) of about the same size fails to reveal any point of difference to which weight can be given. Comparison with a cotype of Ophiocreas phanerum H. L. C. is of no value, the specimen of phanermn being so large no real comparison is possible. I am inclined to agree with Morten- sen (1924, p. 99-101) that phanerum and consiridum Farquhar are identical and both probably synonyms of Studer's adhaerens. But a good series of specimens from 6-18 mm. in disk diameter will be needed before we can feel sure of the matter. Doderlein's estabhshment of the genus Astrobrachion for these forms seems justifiable. A very small Astrobrachion with disk only 2 mm. in diameter and arms 20-30 mm. long, belonging to the Australian Museum, may properly be assigned to this species for the present. It was taken in 15 fms. at Port Molle, Queensland. Euryale' aspera Lamarck, 1816. Anim. s. Vert., 2, p. 538. The tropical coasts of Australia seem to be a congenial home for this remark- able "basket-fish" which ranges as far south as Double-Island Point on the Queens- land Coast and at least as far as Cape Villaret in Western Australia. It is appar- ently not common at Darwin as a single small individual dredged there near the Leper Station, May 25, 1932, was the only one we secured; it is now (dry) 3 mm. across the disk and has arms 15-20 mm. long, which branch 5-7 times. It is somewhat more pigmented than similar small specimens from Broome, but it was 'I cannot agree with Doderlein's (1911, p. 9, 10) line of reasoning which necessitates calling thLs genus Euryala. The names Euryale, Gorgonocephalus and Astrophyton were never truly synonymous. None was a monotypic group or had a type designated. Lyman as first reviser in 1880 was quite within his rights in delimiting the groups as he did and it is unnecessary to alter the names. CLARK: AUSTRALIAN ECHINODERMS 203 taken on a bottom more or less covered with coralline algae, sponges, etc., very different from the habitat at Broome, where all of the very young individuals taken were living on gorgonians while the adults live free and unprotected on the hard sand bottom. The young are pure white at first; then the disk and basal part of arms show a reddish tinge and the tips of the big spines on the distal ends of the radial shields are quite pink. Later, when disk is 6-10 mm. across, red- purple pigment becomes conspicuous around the margins of the disk and in irregular patches on the arms. Many adults still retain a more or less purple color but there is great diversity. My field notes at Broome say: "dark wine-red or brownish-red, disk more purple than arms;" "dull yellowish disk, rose-color on basal part of some arms; "fawn-brown, variegated;" "usually considerable red- dish orally;" "dark grayish with no trace of red or purple." The adults are 25-40 mm. across the disk while the arms range up to at least 250 mm. In well prepared specimens, the length of the arms is 6-8 times the disk diameter. It is not un- likely therefore that living adults have a feeding area 500-600 mm. across. One specimen from Broome is perfectly tetramerous the disk being a square with sides, 12-13 nrmi. long. The 40 specimens at hand are from the following localities : Queensland: Port Curtis. 1 specimen, very young but well pigmented. Loaned by AustraUan Museum. Lindeman Island, July-September, 1934. 10 specimens, very young. Melbourne Ward leg. et don. Northern Territory: Darwin, dredged near Leper Station, May 25, 1932. 1 very young specimen. Western Australia: Broome, August and September, 1929. 13 specimens, adults. Broome, June, 1932. 15 specimens, very young and adult. EURYALE EUOPLA' Sp. nov. Plate 23, fig. 1 Similar to aspera in its general structure and characters, but with much heav- ier and coarser spines on the disk and proximal arm divisions. The first fork of the arms is distinctly further from the margin of the disk than in adult aspera and apparently there are fewer subsequent divisions; owing to the condition of the ' euoTrXos = well-armed, in reference to the very heavy spinulation of disk and basal part of arms. 204 memoir: museum of comparative zoology specimens, which are dry, it is impossible to reach a positive conclusion on this point. In large specimens of aspera, the first fork is about 8-15 mm. from the ends of the radial shields; in the specimens of euopla, it is 15-30 mm. In aspera, when the disk and arm spiniform tubercles are unusually large, they may measure as much as 4 mm. high and 1.5 mm. in diameter at the base, the tip being smaller and often sharp-pointed. In euopla, these spiniform tubercles are often 5-7 mm. long, 2 mm. or more in thickness and blunt or markedly clavate at tip. The num- ber of these clumsy-looking spines is undoubtedly fewer than in aspera but it is difficult to put the difference into figures; it is probably safe to say that in speci- mens of the same size there are not more than two-thirds as many in euopla as in aspera. Holotype, Western Australian Mus. no. 9683, from Bald Island, east of Albany, W. A. There are two specimens of this striking Euryalid, which agree in all essen- tials. Both are yellow-brown in color with the center and interradial areas of the disk considerably darker than the radial shields and the sides of the arms. They were taken, apparently at the same time and place. The holotype is 30 mm. across the disk and 10-12 mm. across the base of each arm which fork at least 6 or 7 times, and probably 8 or 9. The paratype is 25 mm. across the disk and the base of each arm is 8-10 mm. wide. The occurrence of Euryale on the southern coast of Western Australia is most surprising, as no specimen of the genus has hitherto been reported from south of lat. 20° on the west coast and 26° on the east. It is hard to believe that there is not some mistake about the locality label, but in any case there is little question that these two specimens should not be referred to the widespread species aspera. It is not incredible that they are extreme representatives of that form but until connecting links are found the name euopla may well be used for them. GORGONOCEPHALID.AE ASTEROPORPA AUSTRALIENSIS H. L. Clark, 1909. Mem. Austral. Mus., 4, p. 547. There are 10 nice examples of this species at hand, presented by the Austra- lian Museum, ranging in disk diameter from 5 to 17 mm. They were trawled in 70-75 fms. of water, off Cape Everard, Victoria, March 6, 1929. CLARK: AUSTRALIAN ECHINODERMS 205 ASTROCONUS AUSTRALIS Astrophyton australe Verrill, 1876. Bull. U. S. Nat. Mus., 3, p. 74. Astroconus amtraUs Doderlein, 1911. Abh. IVIath.-phys. Klasse K. Bayer. Akad. Wiss., II Suppl. Abh. 5, p. 37. There is at hand an interesting young but quite typical example of this characteristic AustraUan species loaned by the Australian Museum. It is the smallest specimen on record, only 6 mm. across the disk, with arms 20-25 mm. long. The general tint is cream-color but there are light brown marks on the disk and rings on the arms of the same shade. The label reads: "On brown sea-weed (on hook and line), Banks Strait, off FUnders Island, N. E. Tasmania, 7 fms." Astroconus occidentalism sp. nov. Plate 23, fig. 2 Disk 32 mm. in diameter, with 5 arms exceeding 100 mm. in length, forking at least half a dozen times and probably more; width of arm at base about 10 mm., its height about one-half as much. Disk mostly covered by 5 radiating wedges, between which in the interradii are narrow sunken areas; each wedge consists of two indefinitely outUned radial shields only slightly separated from each other by a shallow depression, broadest distally. Wedges covered by a pavement of circular or elliptical flat plates, half a miUimeter in diameter or less, surrounded by and mixed with, granules or small plates of diverse size — many extremely small; distal portion of wedges and interradial areas between covered by small plates and granules, the difference being that the plates are flat, the granules more or less hemispherical; the plates are not numerous enough any- where to form a smooth pavement. Scattered all over the disk, quite irregularly in the interradial "valleys" as well as on the radiating wedges are scores of more or less bluntly conical tubercles, the largest of which are 1.5 mm. in diameter and almost as high, but there is great diversity in size; they are not smooth but finely ridged and wrinkled in an irregular way. On the distal end of each wedge the tubercles become more uniform in size and arrangement ; they are a millimeter in diameter or less and arranged in 3 or 4 transverse series forming ridges separated from each other by rather deep valleys. These tubercled ridges continue out on the arm; there are only 2 or 3 before the arm forks about 5 mm. from the disk. ' occidentalis = western, in reference to its being the representative of the genus on the western coast of AustraHa. 206 memoir: museum of comparative zoology On the amis from the first fork to the second or third or a Uttle further on some branches, the tubercled ridges continue as incomplete rings on the top and sides of the arm; at first there are 10-15 tubercles on each ring but the number de- creases distally and they disappear near the middle of the arm, though the rings themselves continue far out and can be distinguished almost to the very tip of the arm. Between the ridges is a fine but not smooth coating of minute plates and granules. Entire lower surface covered by a fine, granular coat, coarsest in the inter- radial areas. Tentacle pores small; first pair well within disk, with no tentacle- scales. Each succeeding pore is more or less concealed by a slight ridge on its adoral side, which carries 5 or 4, short peg-Uke, flattened spines, terminating in 3 long, glassy teeth or thorns; the ridges themselves merge into the tubercled ridges of the upper surface and sides of the arm. Each mouth angle carries a large number of teeth and oral papillae of diverse sizes, carried irregularly on the sides as well as at the tip of the jaw. Genital slits about 4 mm. long, near the upper margin of the interbrachial area, on each side, close to the arm. Madre- poric plate single, well defined, close to the mouth frame in one interradial area. Color purphsh-brown lightest on tubercles and ridges, the valleys and de- pressions contrastingly darker; there are also irregular dark markings on the disk. Orally the contrast between light and dark is conspicuous; the inter- brachial areas, mouth frame and arm bases are brownish yellow with dull purple markings; on the arms these purple markings are short transverse bars, arranged regularly along each side but the regularity of arrangement disappears after the first fork. Holotype, Western Australian Museum no. 116-37, from North Beach, near Fremantle, W. A. This is a very notable gorgonocephalid obviously related to Astroconus auslralis but easily distinguished by the regular tuberculated ridges on the arms, which give it a very distinctive facies. Presumably it replaces australis on the western coast of the continent. ASTROCLADUS GRANULATUS^ Sp. nOV. Plate 23, fig. 3 Disk 11 mm. in diameter; arms 35 mm. or more in length, branching 8-10 times; first fork about 4 mm. from disk margin at distal ends of radial shields; ' granulatus = TOUgheued by granular elevations, in reference to the granular surface of the disk. CLARK: AUSTRALIAN ECHINODERMS 207 arm 3 mm. wide before forking. Disk decagonal, with 5 concave interradial sides, and 5 nearly straight sides where arm bases are in contact with radial shields; latter about 5 mm. long by 1.5-1.75 imn. wide, nearly parallel in pairs, strongly convex but not at all sharply defined; interradial areas shghtly, center of disk markedly, depressed in tlie dry specimen. Whole surface of disk covered with a nearly uniform, but not at all crowded, coat of rounded granules, coarsest on radial shields, finest at center of disk. Arms evenly covered on dorsal surface and sides with a nearly uniform granulation hke that of the disk, coarsest near base of arm and becoming more and more fine distally; on the distalmost branches, alternating rings of coarser and finer granulation can be distinguished but they are ill-defined and can scarcely be made out until after the fifth or sixth fork of the arm. Whole lower surface covered with a fine, low granulation. Tentacle pores very small, with 6 pairs before the first fork of the arm. Tentacle scales are found between the first and second forkings of the arm; at first there are 2 or sometimes 3, but distally there are more commonly 3; they are very small and their bases scarcely form a distinguishable ridge, while the glassy thorns at the tip are very minute. On the distalmost branches, the tentacle scales are relatively much larger making the tips of the arms very rough in the dry specimen — no doubt very "sticky" and chnging in life. Madreporite single, small, wider than long, on the soft interbrachial area just outside the mouth frame. Each mouth angle occupied by a group of teeth, dental papillae and oral papillae, not sharply dis- tinguishable but much smaller distally than at the center of mouth. Color uniformly rich red-brown; there is no clue to the color in hfe. Holotype, M. C. Z. no. 4899, from Lindeman Island, Great Barrier Reef, Queensland. July-September, 1934. Melbourne Ward leg. et don. This seems to be a very well marked species characterized by the granular covering without tubercles or spines and the very smooth oral surface without any tentacle scales until after the first arm fork. Besides the holotype, there is at hand a very small gorgonocephaUd, with disk only 5-6 mm. across, which Mr. Ward dredged in 9 fms. off Lindeman Island in 1929, and gave to the AustraUan Museum. This specimen is very pale cream-color, the radial shields are con- spicuous in parallel pairs and the granulation is very fine, but I think there is no doubt it is a young individual of granulatus. 208 memoir: museum of comparative zoology astrochalcis tuberculosus KoEHLER, 1905. "Sil)oga" Oph. Litt., p. 1.30. A huge gorgonocephalid in the Australian Museum (J 5294) is certainly an Astrochalcis and probably represents this species; but it is strongly contracted, the arms being rolled inward so as to completely conceal the oral surface, and is now very dry, so that any real study of the specimen is impossible without wrecking it. It is fully 75 mm. across the disk and the arms are 30 mm. wide at base. The ground color is a Ught purplish-brown but it is irregularly spotted and marked with a dirty cream-color and the arms distally become the same tint. The plating of disk and arms is quite smooth but there are rather numerous tubercles, which are so low as to be mere convexities, 1.5-3 mm. across. This specimen was taken in September, 1928, in Albany Passage, northern Queensland, on a gravelly bank, in 9-12 fms. OPHIACANTHIDAE Ophiacantha clavigera KoEHLER, 1907. Fauna Siidwest- Australians : Oph., p. 247. In Koombana Bay, Bunbury, W. A., one of the original localities for this species, we took 11 specimens, October 26, 1929. The largest is 3.5 mm. across the disk and hence is much larger than any of Koehler's 7 specimens; the arms are about 14 mm. long. The smallest specimen is only 1 mm. across the disk. Growth changes in this species are remarkable, for the disk spinelets tipped with 2 or 3 glassy teeth gradually lose those tips and become changed into little sugar-loaf shaped tubercles. In the largest specimen the uppermost arm-spine at base of arm is not merely clavate but is actually forked at tip. The color of clavigera ranges from indistinctive light brownish to white, but the distal ends of the radial shields are rather conspicuously white. At Broome in June 1932, a small ophiacanthid was dredged which it seems best to refer to this species. As the disk is only 2 mm. across and the arms 12-14 mm. long, it is obviously too young for certain identification, but the gen- eral appearance, the disk covering, the white-tipped radial shields, and the arm-spines are much Uke those of clavigera. As this species has been taken in Cockburn Sound near Fremantle, the occurrence in the vicinity of Broome would indicate a distribution coinciding with that of some other Western Aus- traUan echinoderms. CLARK: AUSTRALIAN ECHINODERMS 209 Ophiacantha DISCOIDEA Ltoan, 1879. Bull. M. C. Z., 6, p. 57. The Australian Museum loaned 2 unidentified Ophiacanthas which prove to represent this species. One was collected at the Murray Islands in 1907 by Hed- ley and McCulloch while the other has no data but was probably taken at the same place. The Murray Island specimen has the disk 2.5 mm. across and is in excellent condition. The other is a trifle larger but all the arms are more or less broken. Both are dry, and bleached to a very pale brownish white. Ophiacantha heterotyla H. L. Clark, 1909. "Thetis" Ech., p. 542. There are a number of unidentified Ophiacanthas from the Australian Mu- seum to be referred to this species. Only one calls for any comment. It was taken by Mr. Melbourne Ward in 1929, off Gordon, D'Entrecasteaux Channel, Tasmania, in 5 fms. It is unusually large (4 mm. across the disk) and dark colored, the upper surface of disk a distinct brown, the lower surface and arms more of a gray brown ; at the center of each mouth shield and of each of the first two lower arm plates is a nearly circular blackish spot with ill-defined boundaries. Some specimens from off Port Jackson show the same feature but less con- spicuously; they are themselves lighter colored. Besides the specimen from Tasmania, the material at hand consists of 17 specimens taken as follows : New South Wales: 15 miles northeast of South Head, Port Jackson, 75-80 fms. May, 1924. C. W. Mulvey leg. et don. 4 specimens, 2 adult. 2.5-4 miles off Botany Bay, 33-56 fms. Trawler, "Goonam- bie." McNeil and Livingstone leg. 13 specimens, adult . and young. Ophiacantha ameleta^ sp. nov.^ Ophiacantha da.l