H 1918 Commonwealth of Australia Department of Trade and Customs FISHERIES Biological Results of the Fishing Ex-periments carried on by the F.l.S. "Endeavour," 1909-14. H. C. Dannevig, Commonwealth Director of Fisheries VOL. IV. Published by Direction of the Minister for Trade and Customs, Hon. J. A. Jensen, M.H.R. Sydney, 1916-1918 1 CONTENTS. PART L— Published 2nd June, 1916. Page. I. — Report on the Sea-Lilies, Starfishes, Brittle-Stars and Sea- Urchins obtained by the F.I.S. " Endeavour " on the coasts of Queensland, New South Wales, Tasmania, Victoria, South Australia, and Western Australia. By Hubert Lvman Clark. Plates I.-XLIV " . . . . " . . 1 PARTS II.-IIL— Published 8th August, 191(5. II. — Report on the Polychseta obtained by the F.I.S. '' Endeavour " on the coasts of New South Wales, Victoria, Tasmania, and South Australia. Part 11. By William B. Benhani. Plates XL VI.-XL VIII " 125 III. — Report on the Gephj'rean Priapulus obtained by the F.I.S. " Endeavour " in Australian Waters. By William B. Benham. . . . . . . . . . . . . . . 103 PART IV.— Published 31st October, 1916. IV. — Report on some Fishes obtained by the F.I.S. " Endeavour " on the coasts of Queensland, New South Wales, Victoria, Tasmania, South and South-western Australia. Part lY. By Allan R. McCulloch. Plates XLIX.-LVIII 167 PART v.— Published 18th May, 1918. V. — Report on the Cephalopoda obtained by the F.I.S. "Endeavour" in the Great Australian Bight and other Southern Australian Localities. By S. Stillman Berry. Plates LIX.-LXXXVIII. 201 PART VL— Published 12th October, 1918. VI. — Title Page, Contents and Index . . . . . . . . . . 303 LIST OF CONTRIBUTORS. Page. Benham, William B. — Report oil the Polychsfita obtained by the F.I.S. " Emleavour " on the coasts of New South Wales, Victoria, Tasmania, and South Australia. Part II. .. .. .. .. ..125 Report on the Gephjrean Priapulus obtained by the F.I.S. "Endeavour" in Australian Waters,. .. .. .. I(j3 Berry, S. Stillman— Report on the Cephalopoda obtained by the E.I.S. "Endeavour" in the Great Australian Bight and other Southern Australian Localities. . . . . . . . . . . . . . . 201 Clark, Hubert Lyman — Report on the Sea-Lilies, Starfishes, Brittle-Stars and Sea- Urchins obtained by the F.I.S. " Endeavour " on the coasts of Queensland, New South Wales, Tasmania. Victoria, South Australia, and Western Australia . . . . . . 1 McCuUoch, Allan R.— Report on some Fishes obtained by the F.I.S. " Endeavour " on the coasts of Queensland, New South Wales, Victoria, Tasmania, South and South-western Australia. Part IV.. . 167 ^>^\'< I. Report on the Sea-Lilies, Starfishes, Brittle-Stars and Sea-Urchins obtained by the F.I.S. " Endeavour " on the coasts of Queensland, New South Wales, Tasmania, Victoria, South Australia, and Western Australia. HUBERT LY^IAN CLARK. Curator of Echinoderms, Museum of Comparative Zoology, Cambridge, Mass., U.S.A. Plates i.~xuv. : Text Figs. 1—11. SEA LILIES, STARFISHES, ETC. — CLARK. REPORT ON THE SEA-LILIES, STARFISHES, BRITTLE-STARS AND SEA-URCHINS. I. — Introduction. The collections dealt with in this report consist of the Echinoderms, excepting the Holothurians or Sea-cucumbers {beche-de-mer), or in technical terms the " Actinogonidiate Echinoderms," obtained by the Federal Fisheries Investigation Ship " Endeavour " along the Australian coast south of Lat. 24°S. at depths of less than three hundred fathoms, during the years 1909-1914 inclusive. The Holothurians were by mutual agreement sent elsewhere for study. The collections were sent to me in five instalments and have included altogether one thousand and sixty-one specimens representing one hundred and fifteen species, of which thirty- nine seem to be new to science and fifteen others are now recorded from Australia for the first time. Others, previously known from northern Australia are now listed from stations south of the tropics, and the extension of ranges in this way is oftentimes very great. Up to the present time (1 July, 1915), the list of Actinogonidiate Echinoderms known from Australia was about three hundred and twenty, but owing to numerous cases of incorrect identifications and of inaccurate locality labels, it is impossible to give exact figures. There is no doubt, however, that including the large additions made by the " Endeavour," the known Echinoderm fauna of Austraha includes more than three hundred and fifty valid species, aside from the Holothurians. The geographical distribution of these species is very inter- esting. At least two hundred are endemic, or at least have not yet been reported from elsewhere, while most of the remainder are East Indian species, ranging southward along the eastern coast of the continent to southern Queensland or even to Port Jackson, New South Wales. In his very interesting and useful paper on the " Recent Crinoids of Australia, "i Mr. Austin H. Clark has recognised the existence of a North Australian and a South Australian subregion. Port Jackson on the east coast and Shark Bay on the west serving as the approximate boundary between the two. Study of the Echinoderms as a whole (including the Holothurians), however, 1. A. H. Clark— Mem. Austr. Mus., iv., 15, 1911. 4 " ENDEAVOUR SCIENTIFIC RESULTS. indicates that such a division of the AustraUan marine fauna fails to bring out the interesting contrasts between the east and west coasts, and I would suggest the recognition of three AustraUan subregions as follows : — 1. East Australian subregion, extending from about 35°S. Lat. to 15°, merging northward in the East Indian or, more properly, Indo-Pacific fauna of the Torres Strait district. Of the more than four hundred Australian Echinoderms now known, about sixty-five, or some 16% belong exclusively to this fauna. Some of the characteristic species are the following : — Crinoids. Comatula cratera. Ptilometra mulleri. Compsometra loveni. Asteroids. Mediaster monacanthus. Anthenea acuta. Astrogonium duheni. Ophiurans. Conocladus amhlyconus. Ophiarachnella ramsayi. Ophiura multispina. Echini. Prionocidaris australis. Centrostephanus rodgersii. Clypeaster australasioe. Holothurians. Tceniogyrus australianus. Molpadia dissimilis. Cucumaria mirabilis. Some of the species characteristic of this fauna no doubt range south of Lat. 35° and even reach Port Phillip, Victoria, but such cases are exceptional and do not invalidate the recognition of the subregion. The remainder of the more than two hundred Echinoderms known from the eastern coast of Australia, north of Shoalhaven Bight, New South Wales, are Indo-Pacific forms, many of which find a centre of abundance in Torres Strait. The distance southward to which they range differs greatly in different species, but it is evident from the " Endeavour " collection that most of SEA LILIES, STARFISHES, ETC. — CLARK. 5 them may range as far south as Wide Bay, Queensland, and some are known to reach Port Jackson, New South Wales. 2. South Australian subregion, including Tasmania and the mainland coast south of 35°S. Lat. It merges on the east coast with the East Australian subregion, and some of its characteristic species reach Port Jackson. This is the typical Australian marine fauna and includes about one hundred and twenty species or approximately 30% of the known Echinoderms. More remarkable, however, is the fact that the great bulk of this fauna is endemic or at least is not at present known from elsewhere. Apparently not more than a dozen of the Echinoderms known from the southern coasts of Australia range north of Lat. 28° or east or west of the Australian region. A very few (of which Ophionereis schayeri is a conspicuous example) occur in New Zealand. One {Echinus horridus) is known from South America and South Africa and perhaps more will ultimately be found with that range. One starfish {Luidia maculata) has the very unusual range of encircling the Australian continent and reaching China and Southern Japan on the north. The following are some of the most characteristic species of this remarkable fauna : — Cbinoids. Comanthus trichoptera. Ptilometra rnacronema. Compsometra incommoda. ASTEKOIDS. Nectria ocellata. Tosia aurata. Pseudolinckia rhysa. Plectaster decanus. Coscinasterias dubia. Ophiukans. Astroconus australis. Ophiactis resiliens. Ophiothrix spongicola. Echini. Goniocidaris tubaria. Microcyphus annulatus. Amblypneustes grandis. Holopneustes porosissimus . Heliocidaris erythrogramma. 6 " ENDEAVOUR SCIENTIFIC RESULTS. HOLOTHURI ANS . Leptosynapta dolabrifera. Pentacta spinosa. The western limit of this subregion is not easy to determine. The little work the " Endeavour " did in the Great Australian. Bight revealed a very interesting fauna and suggests that many of the Echinoderms of the South Australian subregion do not extend their range west of 125°E. Long. At any rate, there is no doubt that once Cape Naturaliste is passed, the Echinoderm fauna undergoes an appreciable change, although a few species from the south (as Ptilometra macronema) range northward to Geraldton, Western Australia. 3. West Australian subregion, extending northward from Lat. 35°8. ; the northern limit camiot be given as we know almost nothing of the echinoderm fauna between Lat. 20° on the north-west coast to Port Darwin, a distance of more than a thousand miles. Some eighty -five species of Echinoderms are now known from West Australia ; of these ten or a dozen are species of the South Australian subregion and rather more than thirty are species from the north, relatively few of which are, however, truly East Indian. More than forty species, or half the fauna, are peculiar to the subregion, so far as we know, and many of these are very remarkable forms, often representing peculiar genera. The following are some of the characteristic species : — Crinoids. Comatula etheridgei. Petasometra helianthoides. Neometra gorgonia. Asteroids. Nectria ocelli f era. Oreaster nodulosus. Culcitaster anamesus. Linckia tyloplax. Ophiurans. Conocladus microconus. Astrogymnotes catasticta. Ophiocreas adhcerens. Ophiothrix smaragdina. sea lilies, starfishes, etc. — clark. 7 Echini. Phyllacanihus magniflcus. Centrostephanus tenuispinus. Peronella aphnostina. HOLOTHURIANS. Caudina tetrapora. Colochirus axiologus. It is useless to speculate on the origin or significance of this West Australian fauna until we know the fauna of the north-western coast. The region between Shark Bay and Torres Strait should be thoroughly explored, and, as soon as possible, that between MelviUe Island and Timor. When this vast area is as well known as South-eastern Australia we shall be able to form an intelligent opinion concerning the West Austrahan subregion. We shall know then, too, whether we ought to recognise a North AustraHan subregion, and if so, whence its fauna came. There are some very characteristic species from that district, but they do not come within the scope of this report. So far as the present evidence goes, it seems that the Western Austrahan fauna is a derivative of that of the South Aus- tralian subregion with a considerable element from the tropical Indian Ocean. The characteristic South Australian fauna is clearly derived from that of the east coast with at least a small addition from a southern fauna common to South Africa and southern South America. The exact extent and significance of this southern element is one of the most interesting and least known features of the South Australian fauna. There is little question that the East Australian fauna is derived directly from the Indo -Pacific and is still closely associated with the fauna of the East Indies and the Pacific Islands. Whether any of the east coast fauna has been received through Torres Strait or whether it has all come around the eastern end of New Guinea, or whether it is largely of southern origin and has spread northward and eastward, are still questions to be answered, but I see very little evidence indeed in favour of the latter view. In the above discussion, I have obviously omitted all reference to other groups of animals, believing I could best do my part in the solution of the problem by presenting simply the evidence offered by the group I am studying. I believe the Echinoderms are a very satisfactory group by which to test possible lines of migration, the great antiquity of the phylum, the relatively small number of genera and species- 8 " ENDEAVOUR " SCIENTIFIC RESULTS. and the general distinctness of lines between the classes and orders, all tending to make a correct interpretation of the distribution at least hopefully feasible. Of course, we are just beginning to approach the problem and do not as yet know all of its factors, but the possibiUty of a solution cannot be fairly denied. The collections made by the "Endeavour" have certainly thrown a flood of light on the composition of the Australian Echinoderm fauna, and hence bring us that much nearer to a knowledge of its origin. For the honour done me in entrusting these valuable collections to me, and for the privilege of studying them, I desire to express my thanks to the Minister of Trade and Customs of the Commonwealth of Austraha, Honourable Frank Gwynne Tudor, and to the former Comptroller- General for Trade and Customs, Nicholas Lockyer, Esq., I.S.O., and, of course, to Mr. Robert Etheridge, Curator of the Australian Museum. Since my correspondence has been entirely with Mr. Etheridge, I cannot forbear the expression of my great appreciation of his uniform courtesy, consideration and patience. To collaborate with him in scientific work is indeed a great pleasure. II. — Description of the Genera and Species. CRINOIDEA. The Crinoids collected by the " Endeavour " are of the greatest interest. There are two hundred and sixteen specimens, representing sixteen species, eight of which seem to be new to science, and ten genera, of which three have not previously been recorded from Australia. More than haK the specimens (one hundred and forty-four) represent a single well-known species, while five species are represented by only a single specimen each, and five others by only two each. Of the eight species previously known to science, all have been recorded from Australia before, indeed seven of them are characteristic of the Australian fauna and only two have been reported outside this region. The " Endeavour " collection extends, often very considerably, the known range of four of these eight species. Undoubtedly the magnificent new stalked crinoid from off the south-eastern corner of the continent is the most notable form taken, but the new species of Oreometra and Cosmiometra are very fine. The number of species of Crinoids now recorded from Austraha is sixty-six, not including those taken in depths exceeding three hundred SEA LILIES, STARFISHES, ETC. — CLARK. 9 fathoms nor those known from Sahul Bank (south of Thnor) ; of the sixty-six, thirty-seven, or more than half, are known only from Australia. Family PENTACRINITID^. Genus Metacrinus, P. H. Carpenter. Metacrinus cyaneus,^ sj). nov. (Plate I.) Stem rather stout, apparently not exceeding 400-500 mm. in length, and ranging from 5 to 8 mm. in diameter in the different specimens ; pentagonal in cross section, with shghtly rounded angles ; immediately beneath the calyx, the sides are grooved and the angles sharp, but after five or six nodes, the grooves and angles have virtually disappeared. Cirrus sockets confined to the nodals, transversely oblong, the shorter diameter markedly less than the height of the nodal, one-fourth of which lies below the sockets. Nodals not at all produced at angles, nor otherwise conspicuous, but occasionally a low swelling is indicated on the rounded angle. Internodals 7-15, but rarely fewer than 9 ; the uppermost becomes fused with the nodal on the lower half of the stem ; the upper and lower margins are prettily crenulated along the suture, except on the lowest internodes ; their radial faces are plane (except on uppermost internodes) while the internodal angles may bear low, rounded swellings (scarcely large enough to be called tubercles), though these are often entirely wanting ; the internodals in some specimens are of nearly uniform height, but generally alternate internodals are higher than those between ; this difference may be very marked, especially on the uppermost internodes ; along the midradial line of the upper internodes, there is a conspicuous pit between each pair of stem-segments (nodals and inter- nodals alike), but this vertical series of pits becomes indis- tinct at about the ninth or tenth internode and lower down completely disappears ; the lower internodes have a smooth, shining, porcelain-like surface, but the upper ones lack this entirely . Cirri 50-65 mm. in length (eight to ten times the stem- diameter), the longest with 60-64 segments ; basal segments low (not half as high as wide), gradually becoming higher, but never becoming even approximately as high as wide ; terminal ■claw scarcely longer than the preceding joint, slightly curved : 1. Kvaj/eo?=dark blue, dusky, in reference to the colour of the stem. 10 '■ ENDEAVOUR " SCIENTIFIC RESULTS. there are more or less evident median tubercles on a dozen or fifteen of the terminal segments ; there tend to be two on each segment but they are very low and small as a rule. Cirri transversely oval in cross-section at base, but soon become cylindrical and distally more or less laterally compressed. Fully developed cirri have a smooth, polished surface, as usual. Basals rather prominent, somewhat produced downward over the angles of the stem, broadly in contact laterally. Radials two or three times as wide as high, with a marked proximal angle (between the basals) and a distinctly concave distal margin. IBr normally 6 (1+2, 4+5), ^ but often 7 (1+2, 4+5), occasionally 5 (1+2) and rarely 4 with no syzygy. Axillary triangular, almost as high as wide, lateral angles truncate, distal angle somewhat rounded. Pinnules on second, third and fifth segments and on sixth when seven are present. IIBr 6-11, usually 9 or 11, not rarely 7, but 6 only in one instance ; a syzygy between 3 and 4 ; in one instance the syzygy is between 2 and 3 instead. Illfr almost always present, 9-21, usually 13 or 15 ; a syzygy between 3 and 4. IVBr often present, 16-27, usually 19 or 21 ; more generally present on the outer, than on the inner IIIBr series. Arms 45-60 (57 in ho lo type), with about 140 brachials after the IVBr series ; articular tubercles slightly indicated ; dorsal surface smooth ; beyond the IIIBr axillary the brachials have somewhat flaring distal margins, more or less spinulose at first, but becoming smooth near arm tip ; distally the arms are somewhat compressed ; the arms of the ho lo type are more than 200 mm. from the radial to the tip. Oral surface of arms with nmnerous granules and little plates, which are particularly crowded along the margins of the food-groove. Pinnules rather stout, four-sided, somewhat flattened ; each segment is more or less concave or hollowed out on both the proximal and distal sides ; this is very marked on distal side of first pinnule but on the second is nearly wanting, the distal margin of the latter being compressed and keel-like ; subse- quently the concavity of the distal and proximal sides of the pinnules becomes a constant feature ; the oral surface of each pinnule is provided with granules which guard the food- groove ; the terminal half of each pinnule is thus covered orally by a double, interlocking series of pointed granules. First pinnule (that on IBr 2) about 20 mm. long with 17-18 segments ; second somewhat longer ; third a little longer than 1. Not counting the radials themselves, of course. SEA LILIES, STARFISHES, ETC. CLARK. 1 I second ; fourth (that on IIBr 4) about 25 mm. long, with 22-23 segments ; fifth about as long ; sixth a Httle shorter ; succeeding pinnules decrease in length very gradually, becoming about 10 mm. long, with 15 or 16 joints,; this size continues to the terminal part of the arm, where, about fifty segments from the tip, it drops quickly to 2 mm. and less, with only three segments or even fewer ; there are half a dozen or more intermediate pinnules with 12, 10, 8 and 6 joints, but the transition is quite abrupt as is usual in the genus. Colour (in alcohol or dried from alcohol) either uniformly light bluish-grey, or crown pale yellow or whitish, and stalk deep greenish-blue in marked contrast. The holotype (in alcohol) is uniformly whitish with a distinct orange tinge, but only a small part of the stalk is attached to the magnificent crown. The stalk of this specimen, preserved dry, is deep greenish-blue. Unfortunately there are no notes on the colour in life, of any of the specimens, but it is not likely that the bluish shade in the stalk is the result of preservation. It is possible that the yellowish crowns are the result of preserva- tion in alcohol. There are seven specimens. This superb crinoid is well distinguished from all the pre- viously described members of the genus by the large number of segments in all the various numerical series ; thus the number of internodals (usually more than 9 and often 14) is the maximum for the genus, the number of cirrus-joints is nearly the maximum, the number of radials is often the maximum, and the segments of both the IIBr and the IIIBr series approach the maximum number. In certain particu- lars, M. rotundus, Carpenter, seems to be the nearest relative of M. cyaneus, but it has decidedly fewer cirrus-joints, and the distal brachials are not at all flaring, the arms being quite smooth; moreover the radials and the members of the IBr series are stouter, smoother and of quite a different shape in M. rotundus. As the only stalked crinoid as yet known from the coasts of Austrahai, Metacrinus cyaneus is of unusual interest and will certainly rank as one of the " Endeavour's" most noteworthy discoveries. The geographical isolation of M. cyaneus is remarkable, the Kermadec Islands, nearly 2500 miles to the north-east being the nearest locality where Metacrinus has been taken. Locs. — Eastern Slope, Bass Strait, about 200 fathoms. This specimen has been selected as the holotype. South-east of Wilson's Promontory, Victoria. 1. Sahul Bank, south of Timor, is scarcely Australian. 12 " ENDEAVOUR " SCIENTIFIC RESULTS. Twenty and a half miles S., 19° E. of Cape Everard, Victoria, 90 fathoms. One good specimen from an unknown locality, but the appearance leaves little room for doubt that it is from the station south-east of Cape Everard, Victoria. Family C0MA8TERID.E. Genus Comatula, Lamarck. COMATULA CRATERA,! SJ). nOV. (Plate II., fig. 1.) Disk of holotype, 13 mm. across ; arms 175 mm. or more in length, 4 mm. in transverse diameter at base, 5 mm. at 20 mm. distance from base, and 3| mm. at 50 mm. from base. Centrodorsal pentagonal, 4 mm. in diameter ; its sides are straight or more or less concave ; it is flush with the radials, all of which are fully exposed. Cirri none in the holotype and eight other specimens ; in three specimens, a single small cirrus is present in each ; one specimen has two and one has three such cirri, each with 16-19 joints. The second radial is more than twice as high as the first and half again as wide ; the third (axillary) is triangular, nearly twice as wide as high. Brachials (beyond the tenth) broadly triangular, the pinnule- bearing end three times as wide as the other ; the distal margin is slightly concave and projects as a faint articular tubercle where the brachial is widest ; the lowest brachials are nearly equally wide on the two sides and have the articular tubercle more marked on their proximal margins, but these tubercles are never conspicuous. Along the median line of the arm is a shght keel ; it is better marked on some brachials than on others and the degree of development difl^ers in differ- ent specimens. Syzygies between br. 1 and 2, and 3 and 4 and then at intervals of 5-7 joints to about br. 30, after which the usual interval is 5, though 4 or 6 occur frequently. Pin- nules rather stout but becoming long and slender distally ; Pi about 20-25 mm. long, with some 45 segments, of which not more than 12-15 make up the small terminal comb ; suc- ceeding pinnules successively shorter to Pg which is about 16 mm. long and has some 25 segments ; terminal combs confined to first three pairs of pinnules ; basal segments on 1. K/)aT6/jo'= strong, stout, in reference to the unusually thick arms. SEA LILIES, STARFISHES, ETC. — CLARK. 13 all pinnules, much wider than high, and somewhat com- pressed, especially at distal margin ; on P2, P3, Pa, and P(„ this compressed distal margin becomes a conspicuous project- ing keel on segments 2 and 3 ; beyond Ps, the pinnules gradually lengthen and become more slender until they may exceed 20 mm. in length and have 35 segments. Mouth radial in position, not very close to margin of disk, with equally developed food-grooves running to all the arms. There is no plating of disk or of the oral surface of arms and pinnules, but the membrane covering the disk is, when fully dried, seen to be filled with calcareous granules of very small size. Colour (in alcohol or dried), light fawn-colour with or without purple markings ; perhaps in life all would show these purple markings more or less clearly ; when well- developed they appear as longitudinal stripes one on each side of each arm, with a branch runnning up each pinnule ; distally the stripes fade away altogether ; proximally the stripes on the inner side of a pair of arms unite at the tip of the axillary from which they arise, while those on the outer side broaden out and cover the radials except for a narrow median area ; even the centrodorsal may be purplish ; the disk and oral surface of the arms are yellowish, quite yellow when dry. Whether this fine comatulid should be considered distinct from C. Solaris is, of course, a matter of opinion. The locality is some four hundred miles further south than C. Solaris has been recorded and the depth is considerably greater than any published for that species. In view of these facts, it seems to me that the absence of cirri and reduction of the centrodorsal plate, combined with the very different terminal combs of the oral pinnules and the absence of plating on disk and oral surface of the arm bases, warrant the recognition of the form by a different name. In C. Solaris the terminal comb of Pi consists of 35-40 segments and occupies nearly half the pinnule ; the shape of the individual teeth is, more- over, very different from what is seen in C. cratera. The comatulid taken by the " Alert " at Port Molle, Queensland, referred by Carpenter to C. Solaris, but specially discussed because of the absence of cirri and the unplated disk, is very possibly a specimen of C. cratera. Lac. — Eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. Fourteen specimens. 14 " ENDEAVOUR ' SCIENTIFIC RESULTS. Genus Comanthus, A. H. Clark. COMANTHUS PERPLEXUM,^ Sf. nOV. (Plate III., fig. 2.) Centrodorsal well developed, nearly 5 mm. in diameter, slightly concave ; around its margin is a series of about 15 eirrus-sockets, 7 or 8 of which bear more or less developed cirri ; the largest of these are about 13 mm. long and consist of 16-19 segments, of which the terminal 3 or 4, aside from the claw, carry very insignificant elevations or ridges on the proximal dorsal margin. IBr series, 3, all visible, the axillaries broadly triangular, not in contact interradially. IIBr 4 (3+4), low, broad, smooth and practically without articular tubercles; distal margins distinct, shghtly flaring, especially on IIBr.2 ; the whole series is only 5 mm. long and is nearly 4 mm. wide. IIIBr series 2, except on one IIBr series, where one arm has a IIIBr series 4 (3+4) and the other arm is undivided. There are 4 IVBr series present, all 4 (3+4) and there are thus 44 arms ; these are about 100 mm. long and have approximately 100 segments beyond the last division ; there is a syzygy between 3 and 4, and a second ten to fifteen segments beyond ; after that the syzygial interval is 5, with remarkable uniformity. Brachials quadri- lateral but except near tip of arm, the pinnular side is much wider than the one opposite ; distal margins flaring, thickened and minutely rough but not spinulose. Pinnules rather slender; Pj about 17 mm. long, with 45 segments, of which only the terminal 10 form the comb ; P2 shorter with about 35 segments, 10 in the comb ; P3 remarkably short and weak, scarcely 5 mm. long and consisting of only 18-20 segments, the terminal comb of 7 segments is well-marked ; none of the subsequent pinnules have a comb, but they gradually increase in length to 10 mm. or more, still consisting, however, of about 20 segments ; none of the segments are conspicuously spinu- lose, though the terminal ones are more or less so, but the basal segments are notably triangular. Disk wanting. Colour (dry), uniformly pale fawn ; oral surface of arms and pinnules brownish-black. There is no doubt that this species is very near C. briareus (Bell), but after comparison with numerous specimens, I am unwilling to refer it to that species. The colour is naturally the most striking difference, but, of course, is the least 1. Per pZea;ws= puzzling, in reference to the doubt whether it is really a distinct species. SEA LILIES, STARFISHES, ETC — CLARK. 15 significant ; nevertheless it is noteworthy that all recorded specimens of C. briareus and all of the thirty specimens at hand are blackish-brown, brownish-black, or purplish-black ; one recorded from Western Australia by A. H. Clark has lighter lines on the arm. Aside from colour, the flaring distal margins of the low radials and of the brachials in C. perplexum are quite different from those of C. briareus, while the smoothness of the brachials and pinnules is noticeable, as contrasted with the rough arms of C. briareus. I have carefully compared the single specimen with the descriptions of C. weberi and C. rolula of A. H. Clark, at the kind suggestion of that well-known authority on the group, but the cirri show at once that it is not C. weberi, and the characters of the centrodorsal, the cirri and the brachials seem sufficient to distinguish it from C. rotula. The geographical isolation of this interesting comatuhd is remarkable for no near relative has been taken south of Port Molle and C. briareus is not kno^vn from south of Port Denison, some six hundred miles to the north. Log. — Eleven miles south by east of Ballina, New South Wales, 27-28 fathoms. COMANTHUS PLECTROPHORUM,^ Sp. TIOV. (Plate IV., fig. 1.) Disk about 30 mm. in diameter ; arms rather more than 100 mm. long. Centrodorsal about 10 mm. in diameter and more than 2 mm. thick ; its bare central area is very rough, slightly concave and about 6 mm. across. Cirrus -sockets in three crowded and irregular horizontal series. Cirri LV.-LX., 29-37, usually about 32 ; some proximal segments, say 4-9, are cylindrical and longer than thick, 5 is particularly long ; beyond 10 the distal dorsal margin projects sHghtly, the segments become compressed and a marked dorsal keel is formed, which is most fully developed on the four or five segments preceding the penultimate ; seen from the dorsal side this is more of a tubercle than a keel, but the lateral aspect is very keel-like. Radials entirely concealed ; IBr 1 wide and low, in contact for about three-fourths of their height ; radial axillaries, very low and wide, at least three times as wide as high, not at all in contact with each other. IIBr series 4 (3+4), in one instance only, 2, well separated from each other externally ; internally IIBri are more or less in contact. IIIBr series 1. 7r\yKTpoi'=a, spur-f 0o/96'w=to bear, in reference to the notable spurs on the pinnules. 16 " ENDEAVOUR " SCIENTIFIC RESULTS. 4 (3+4), in two instances only, 2 ; in the dry specimen the distal margins of this series and the preceding are thickened and shghtly flaring and there is Httle indication of synarthriai tubercles, but in the alcohohc specimen while very slight synarthriai tubercles are indicated the joints are all smooth and indistinct. IVBr series, when present, 4 (3+4) ; not present at all in the dry specimen, but four such series are present in the alcoholic. Arms 40-44, terete ; brachials soon become very markedly triangular, with thickened, flaring, roughened distal margins, but distally they again resume a quadrilateral form and the distal margins are not peculiar. A syzygy occurs between 3 and 4 and then at a variable distance of 8-16 segments ; after the second syzygy there may be a second interval of 8-10 segments, but as a rule syzygies become frequent and very regularly spaced, in one specimen at intervals of 5 segments and in the other at 6 ; that is, in one specimen each syzygial pair is followed by 3 normal segments and then another syzygial pair, while in the other specimen, 4 normal segments intervene. Pinnules long and slender, especially the oral pinnules, of which the first exceeds 30 mm. in length and consists of more than 70 segments ; succeeding pinnules decreasing in length to P4 which is only about 15 mm. long and contains only 20-25 segments. Comb of lowest pinnules consists of about 12-15 segments ; the combs persist as far as P4 on which the comb consists of 7 or 8 segments. Basal segments of all pinnules, except at very tip of arm, much wider than long ; on the basal and middle portions of the arm, 3-12 of these widened segments are conspicuous for their flaring, spinulose margins, which are prolonged on the aboral side into remarkable spinulose spurs ; these spurs are, of course, largest and most fully formed on the basal segments (excepting the lowest two) of the lowest pinnules ; passing distally along the arm, as well as along each pinnule, these spurs become less and less evident and finally disappear. Disk quite fully plated, especially along the food-grooves ; the position of the mouth cannot be determined in these specimens. Colour (of alco- holic specimen) hght brown ; cirri somewhat fighter than arms ; on drying, the colour becomes uniformly fight fawn or very pale buff. This is a very well characterised and remarkable Comanthus, related to C. bennetti and C. pinguis, but easily distinguished from either, by the curious pinnule segments. The small number of arms and the colour will further distinguish it from C. bennetti, while the shape of the arms, especially at SEA LILIES, STARFISHES, ETC. — CLARK. 17 base, separate it from C. pinguis. From the South AustraUan species C. trichoptera, the size, the cirri and the pinnules all distinguish it without difficulty. Loc. — East of Flinders Island, Bass Strait, 100-300 fathoms. Two specimens. COMANTHUS SPANOSCHISTUM,! Sp. nov. (Plate IV., fig. 3.) Disk 5-12 mm. in diameter ; arms 30-80 mm. long. Centro- dorsal large and flat, 5 mm. in diameter in adults, and about 1 mm. thick. Cirri, in one or in two imperfect series, about XXIV., 15-20 ; the longest are 10-15 mm. long ; segments 4-7 or 3-8 are cylindrical and longer than wide, segment 4 (or 5) may be nearly twice as long as thick ; distally the segments are a trifle compressed and the distal margin dorsally is elevated to form a low tubercle, which, at least on the pen- ultimate segment, is spiniform. Radials hidden ; IBr 1 low and wide, more or less in contact ; radial axillary triangular, high and pointed, the height about two-thirds of the width. IIBr series, not often developed; when present, invariably 4 (3-J-4) ; arms usually 10, in one specimen 13, in another 14, and in a third 19; No IIIBr series is developed. Brachials at first quadrilateral, but soon becoming triangular ; distal margins becoming more and more flaring and overlapping until near tip of arm where the brachials become quadrilateral again and the margins hardly flare. Syzygies between 3 and 4 and then at an interval of 7-11 segments ; after that the syzygial interval is usually 5, but may be only 3. Pinnules long and slender ; Pi is over 10 mm. long and consists of about 30 segments, of which about 15 make up the comb ; P2 is somewhat smaller and P3 is only 8 mm. long and consists of 20 segments, 9 in the comb ; P^ is 8 mm. long, but has no comb ; succeeding pinnules become more slender and increase in length to about 10 mm. ; basal segments of lower pinnules, more or less triangular with greatly flaring, spinulose margins which tend to project like rough spurs on the aboral side of the largest segments, much as in the preceding species {C. plectrophorum). Disk, except around tip of anal tube, smooth and naked ; there are calcareous nodules in the anal tube. Mouth interradial. Colour, in alcohol and dry, more or less yellowish, with either a green or a brown cast ; one 1. (r7ro»/o'?=rare4-ff;;^<<rTo'9=divided, in reference to the scarcity of arm divisions. 18 " ENDEAVOUR ' SCIENTIFIC RESULTS. specimen is very dull purplish with yellow cirri ; others are yellowish more or less clouded with purplish ; in the 19-armed individual the dull purplish predominates. Twenty-three specimens. The specimen from off Noosa Head, Queensland, is notice- ably more slender than the others and the cirri are somewhat more compressed but these differences are very shght and no more than one might expect in view of the wide separation of Noosa Head from the Tasmanian waters where this species seems to be so common. It is no doubt nearly related to C. trichoptera, but the scarcity of division series and the absence of the peculiar axillaries characteristic of that species, would seem to preclude regarding the two as identical. Many of the specimens, even small ones, have the genital pinnules immensely swollen with the reproductive cells ; in some cases only the pinnules near the tip of the arm are unaltered. The specimens which have dates on the labels were taken in March and April, 1914, and thus it is evident that the species breeds in late summer and early faU. It is possible, of course, that it has a prolonged breeding period. Locs. — East of Babel Island, Bass Strait, 60-70 fathoms and 50-80 fathoms. North-east of Babel Island, Bass Strait, 100-170 fathoms. Twenty miles east of Babel Island, Bass Strait, 65 fathoms. East of Maria Island, Tasmania, 78 fathoms. East of Flinders Island, Bass Strait, 70-100 fathoms. Eastern Slope, Bass Strait, 70-120 fathoms. Off Noosa Head, southern Queensland, 16 fathoms. Family ZYGOMETRID^. Genus Zygometra, A. H. Clark. Zygometra elegans (Bell). Antedon elegans, Bell, Rep. Zool. Coll. H.M.S. " Alert," 1884, p. 162. Zygometra elegans, A. H. Clark, Smiths. Misc. Coll., 1., 1907, p. 348. The occurrence of this comatuhd as far south as Sandon Bluffs, New South Wales, is an interesting extension of its range, since it is not recorded from a lower latitude than that of Port Curtis, Queensland. The three specimens are all of SEA LILIES, STARFISHES, ETC. CLARK. 19 large size (arms nearly 100 mm. long), but show considerable diversity of colour. One has 38 arms, the greater part of which are light brown, but basally all are heavily shaded with purple, the joints in particular being very dark ; there are 35 cirri with about 54 segments ; they are, at base, dorsally pale brown, orally deep purple, but chstally the pui'ple becomes very gradually dominant until it includes all of both surfaces. The second has at least 36 arms (several are missing) which are light brown with more or less of a purplish cast ; basally three of the rays are very deep purple, the other two being decidedly lighter ; there are more than 30 cirri with 47-53 segments, light brown at base, but becoming purple distally. The third specimen, from off Noosa Head, Queensland, has only 29 fawn-coloured arms (dry), with a faint pvirpHsh cast, the joints distinctly darker than the segments ; there are, in marked contrast, 27 dull purple cirri, pale brownish at base dorsally (as in the centrodorsal), with 45-47 segments. In this dry specimen the IIIBr series is 2 except in one instance of 4 (3-)- 4) but in the other specimens it is common to have the IIIBr series 4 (3+4), although 2 seems the usual number. The IIBr series, and the IVBr series, when present, are invariably 4 (3+4). The only VBr series noted is 2. Locs. — Eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. Off Noosa Head, southern Queensland, 16 fathoms. Zygometra microdiscus {Bell). Antedon microdiscus. Bell, Rep. Zool. Coll. H.M.S. "Alert," 1884, p. 163. Zygometra microdiscus, A. H. Clark, Smiths. Misc. Coll., 1., 1907, p. 348. This is a small, dry specimen of uniformly pale brown colour, with only 22 arms. There are 8 IIBr series, one of which is 2, the rest 4 (3+4) ; there are 4 IIIBr series, all of which are 4 (3+4). The centrodorsal is low, conical, with 25-30 cirri, the longest having more than 40 segments. P2 has 27 segments and is relatively large, but is less than 20 mm. long. The occurrence of this species near the Capricorn Islands is quite to be expected, but nevertheless extends the known range of the species far to the south. The present specimen is naturally referred to Carpenter's Antedon multi- radiata, but I think A. H. Clark is quite right in regarding that species as a synonym of Z. microdiscus. Loc. — Thirteen miles north-east of North Reef, Capricorn Group, off Port Curtis, Queensland, 70-74 fathoms. 20 " ENDEAVOUR SCIENTIFIC RESULTS. Family CALOMETRIDiE. Genus Oreometea, A. H. Clark. Oreometra pericalles,! sf. nov. (Plate III., fig. 1.) Disk wanting ; centrodorsal 6 mm. in diameter. Cirri all missing, but cirrus-sockets 17, their ventral portion well on to the radials ; centrodorsal slightly convex, smooth. IBr 1 and 2 (the axillary) united by syzygy ; radials in contact laterally, but IBr 1 and 2 clearly separated from the adjoin- ing series ; each of these segments is about three times as wide as high and 2 has a low, obtuse distal angle ; there is a cylindrical, blunt, spine-like tubercle (sometimes a pair) on the proximal margin of 1 and another on the distal margin, in the median line ; there are 1-3 somewhat similar but flattened projections on each anterior margin of the axillary. IIBr series 4 (3-f4), or once, 3 (2+3) ; segments 1-3 have the spine-like tubercle at the median, distal margin prominent, while the axillary may have one somewhat smaller, on each side ; an additional similar tubercle may be developed in connection with the well-marked distolateral angles. Lower brachials quadrilateral with conspicuous distolateral angles on the pinnular side ; the median spine-like tubercle is also often developed ; after the seventh or eighth, the brachials tend to become triangular and the distal margins tend to flare, but there is no overlapping. Syzygy between 2 and 3 on every arm, with a IIBr series, but between 3 and 4 on arms arising direct from the radial axillary ; second syzygy is somewhere between the eighth and fourteenth segments ; after that the syzygial interval is five or six segments. Pi somewhat prismatic, about 7 mm. long, of about 15 segments, curved inward, flexible, but not at all flagellate, rather stout throughout ; the first segment is notably en- larged, but not the second ; P2 erect, and spine-Hke, not at all prismatic, about 13 mm. long, of 15 nearly cylindrical segments, with spiny distal margins, all but the terminal and 2 lowest, longer than thick ; P3 very similar, perhaps a Httle longer ; P4 similar, perhaps a little shorter than P3 ; P5-P10 similar, but successively shorter and with fewer segments, which have less and less spiny distal margins ; Pi 1 of 11 segments and about 8 mm. long ; succeeding pinnules dis- tinctly prismatic and gradually increasing to about 10 mm. in length. Side and covering plates well developed. Arms about 75 mm. long. 1. 7re/)(A:r(\X)J<f=very pretty, in reference to both form and colouration. SEA LILIES, STARFISHES, ETC. — CLAEK. 21 Colour, of the single dry specimen, purple and white ; centrodorsal, dorsal surface of brachials and of pinnules, and scattered spots elsewhere, white or whitish ; IBr series dirty whitish with a purple tinge ; IIBr series, a few of the lowest brachials, a conspicuous hne along each side of each arm, and oral surface of pinnules, at least on basal portion, more or less deep purple. This very handsome species was at first referred to Zygo- metra on account of the syzygy in the radial axillary. Mr. A. H. Clark kindly suggested comparing it with Calometra and the examination of the pinnules indicated relationship to that genus. I think there can be no doubt, however, that this specimen is a representative of the recently described genus Oreometra and it is quite near the type-species, 0. marice, A. H. Clark. The differences in the IBr series and in the pinnides, however, forbid referring it to that species. The occurrence of a syzygy between brachials 2 and 3 is note- worthy ; Mr. Clark says nothing about syzygies in either one of his two published descriptions of 0. marice. He does say, however, that the connection between IBr 1 and 2 is "an exceedingly close synarthry," while I have been convinced after a careful examination of the distal face of IBr 1 that the joint is certainly not a synarthiy in the present species and I do not see why it is not a syzygy. If it is a syzygy or a pseudosyzygy, it seems to me the genus is a connecting form with the Zygometridse, if not actually a member of that family. The pinnules obviously count against any such view. The cirri I have not seen. Loc. — Thirteen miles north by west of Double Island Pointy Queensland, 25-26 fathoms. Family HIMEROMETRID^. Genus Amphimetra, A. H. Clark. Amphimetra crenulata [Carpenter). Antedon crenulata, Carpenter, Jo urn. Linn. Soc, Zool., xvi., 1882, p. 507. Amphimetra crenulata, A. H. Clark, Smiths. Misc. Coll., Ix., 10, 1912, p. 16. This is a fine adult individual with 28 arms ; it had about 25 cirri, the longest with 45 segments. The colour (dry) is uniformly brownish-white. The species was not previously known from south of Port Curtis, Queensland. Loc. — Thirteen miles north by west of Double Island Point, Queensland, 25-26 fathoms. 22 " ENDEAVOLTR " SCIENTIFIC RESULTS. Family COLOBOMETRID.E. Genus Oligometra, A. H. Clark. Oligometra thetidis, H. L. Clark. Oligometra thetidis, H. L. Clark, Mem. Austr. Mus., iv., 11, 1909, p. 522. These two specimens are uniformly yellowish in alcohol and show no noteworthy peculiarities. Loc. — Eastern Slope, Bass Strait. Oligometra zebra, i sp. nov. (Plate II., fig. 2.) Disk wanting ; arms about 70 mm. long. Centrodorsal, 3 mm. in diameter, markedly concave. Cirri in a single, crowded, fairly regular marginal series of about 20, 12-13 mm. long, each with 27 or 28 segments ; all segments wider than long ; basal ones nearly square but distally each successive segment becomes more com- pressed, ventrally rounded, and dorsally transversely ridged ; transverse ridges never conspicuous, but more evident on the last four segments (not counting the claw), especially on the penultimate, where it forms a well-marked opposing spine. IBr series not peculiar, the synarthrial tubercle, however, being quite distinct ; the axillary is nearly twice as wide as high ; the adjoining rays are well separated from each other. Arms 10 ; brachials at first quadrilateral, becoming triangular, and then near tip of arm again quadri- lateral ; synarthrial tubercle of 2 and 3 well-marked ; neither distal margins nor disto lateral angles peculiar, though the latter are evident enough, especially on terminal half of arm ; general contour of arm quite smooth. Pi about 7 mm. long, of 18 segments, not at all rigid or otherwise pecuHar ; the basal 6 or 7 segments are wider than long while the succeeding are squarish, or a little longer than wide ; the oral margins of the lower segments are barely produced into one or two minute spines, while the aboral margins are inconsi)icuous, flattened and a trifle roughened ; segments 8-12 are somewhat prismatic and the remaining segments are distinctly flattened. Pg is 9 mm. long, with 18-20 segments ; it is very similar to Pi but is evidently stouter and is clearly the largest pinnule on the arm ; the oral margins of 1. Zebra, a Latinised form of the native name of the well-known African quadruped ; in reference to the colouration of the arms. SEA LILIES, STAHFISHES, ETC. CLARK. 23 the basal segments are smooth, while the aboral margins are more or less produced as flattened, spinulose projections ; the distal margins of all the segments, except the first five or six are finely spinulose, but none of these characteristics are at all conspicuous. P3 is much like P2, but a little smaller, while P4, which is probably the shortest pinnule on the arm, is only a trifle shorter than P3 ; succeeding pinnules become very slender, 8-9 mm. long and with 22 or more segments. On the basal half of the arm, the pinnule -segments, except the basal two or three, have distinctly, but finely spinulose aboral edges and distal margins, but on the outer half of the arm they seem to be quite smooth. Colour (dry) of centrodorsal, basal portion of the cirri, all arm-segments and pinnules, uniformly pale buff or brownish- white ; cirri gradually become dull purple at tip, the lighter shade being confined more and more completely to the ventral side of the segments and finally is crowded out altogether ; arm joints rich reddish-purple, in sharp and handsome con- trast to the fight segments ; oral surface of pinnules very- dark, almost black. One specimen. This seems to be a well-marked species, the long cirri, the flexible first pinnule, and the absence of conspicuous pro- cesses on the basal pinnule-segments combining to distinguish it at once from the other species of the genus. Loc. — Eleven miles east-south-east of Clarence River mouth, New South Wales, 35-36 fathoms. Family THALASSOMETRIDiE. Genus Ptilometra, A. H. Clark. Ptilometra macronema {Miiller). Comatula macronema, Miiller, Monatsb. K. Preuss. Akad., 1846, p. 179. Ptilometra macronemu, A. H. Clark, Smiths. Misc. Coll., 1., 1907, p. 358. These ten specimens are all typical examples of this ■characteristic South Austrafian species. Locs. — Great Austrafian Bight, about 131° E., 62 fathoms. South-east of Ffinders Island, South Austrafia, 37 fathoms, Sanders Bank, off Kangaroo Island, South Austrafia, 28 fathoms. Forty miles west of Kingston, South Australia, 30 fathoms. 24 " ENDEAVOUR " SCIENTIFIC RESULTS, Ptilometra mulleri, a. H. Clark. Ptilometra mulleri, A. H. Clark, Proc. Biol. Soc. Washington, xxii., 1909, p. 41. This large series of Ptilometras (one hundred and forty- five specimens) shows no tendency to intergrade with the preceding species and fully justifies Mr. Clark's separation of the two forma The occurrence of this species oft' the mouth of the Clarence River, New South Wales, extends its known range northward about 150 miles. It is interesting to note that no Ptilometras were taken by the " Endeavour " in Tasmanian or Victorian waters, though the genus has been reported from Port PhiUip. None of the material in the present collection throws any Ught on the disagreement between Mr. A. H. Clark and myself with reference to the young of Ptilometra. We are each equally sure of being right, and no doubt will continue to feel so until the life history of one of the members of the genus is fully known. It seems to me possible that Bell's Antedon wilsoni, which I have never seen, and my Himerometra 'pcedofhora are the same species, but I am perfectly sure the latter is not a Ptilometra and has no near relationship to that genus. It is- to be hoped that some Australian zoologist will before long investigate carefully the hfe history of the comatulids, so accessible from Sydiiey or Melbourne. Lacs. — Eleven miles east-south-east of Clarence River mouth, New South Wales, 35-36 fathoms. Eight miles east of Sandon Bluff's, New South Wales, 35-40 fathoms. Six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Between Port Stephens and Newcastle, New South Wales,. 22-60 fathoms. Shoalhaven Bight, New South Wales, 15-45 fathoms. South-eastern coast of Australia. Genus Cosmiometra, A. H. Clark. COSMIOMETRA DASYBRACHIA,! Sp. TIOV. (Plate IV., fig. 2.) Arms 20, about 90 mm. long. Centrodorsal about 4 mm. high and scarcely 5 mm. in diameter at base, markedly 1. ^a<Ti)9=rough, downy+y3/j«xtwv = the arm, in reference to the- fine roughness of the arms, distally. SEA LTLIES, STARFISHES, ETC. — CLARK. 25 conical ; cirrus-sockets in ten vertical series of 3 or 4 each (about 35 altogether), the series separated radially by a rather broad area, with finely roughened surface. Cirri 45 to 50 mm. long, with 68-74 segments, of which the fifth or sixth is longest (much longer than thick) and all but the basal fifteen have the distal dorsal margin produced into what looks like a sharp spine, when seen from the side ; distal segments, including this dorsal spine twice as wide as high. Radials completely concealed except for the small disto- lateral corner which appears in the interradial angles of the calyx. IBrl crescentic, the proximal margin strongly convex, the distal correspondingly concave ; excepting only the distal median portion, the whole surface is rough and spinulose, especially the lateral margins. IBr2 (axillary), rhomboidal with anterior margins, sHghtly concave ; all margins finely serrulate and lateral surfaces rough and spinulose ; synarthrial tubercle of 1 and 2, smooth and rounded, low but distinct. IIBr series, 2, very similar to the IBr series, but the lateral spinulose surfaces are much less marked, especially on the axiUary and are nearly wanting on the inner side. First two brachials are similar but are relatively much longer and narrower and have no spinulose lateral areas, though the lateral margins are very serrulate and rough. The division series and these two brachials have strongly flattened sides ; the division series are in closely appressed contact, but the lower brachials are much less so, and probably in life hardly touch. Brachials 3-10 quadri- lateral, but subsequent segments become triangular, the distolateral angle being more and more prominent ; beyond the middle of the arm the segments become quadrilateral again and nearly as long as wide ; proximally both proximal and distal margins are everted and rough, but beyond the 12-14 joint only the distal margin is conspicuous ; even near the tip of the arm, however, the distal margin is not very flaring and it never overlaps the following segment. Beginning at the 35th, or thereabouts, the distal dorsal surface of each segment is rough with minute spines and these soon cover the whole median area of the arm ; near the tip, this area is a very narrow band. Pi about 11 mm. long, of 25 smooth segments, the six basal ones much wider than long, somewhat axe-head shaped, much flattened dorsally ; remaining segments longer than wide, or at least as long, the eighth segment being the longest. Pg similar, but only 10 mm. long and with only 22 segments. P3 similar, but a little shorter. P4-P10 more prismatic, only •26 " ENDEAVOUR " SCIENTIFIC RESULTS. 5-6 mm. long, with 12-14 segments, which have the distal margins, and a sharp dorsal keel, finely serrulate ; remaining pinnules similar, but becoming much more slender, longer (8-10 mm.), and with more segments (17-18) ; the sharp, serrate dorsal keel is well-marked. Side and covering plates, highly developed. First syzygy between 3 and 4 ; in eleven arms of one specimen, but in only one arm of the holotype, a second occurs between 5 and 6, and the third is somewhere in the region between segments 18 and 27 ; in the other arms, the second syzygy is generally not until after the twentieth segment ; distally the syzygial interval is at first 6-8, but near the tip of the arm it decreases to 4 or 5. Colour (in alcohol) yellowish-brown; when dry, the shade is much lighter. While this species resembles both C. komachi and C. helene in the general character of the arms, the much more numerous cirrus joints, the longer pinnules and the markedly conical centrodorsal seem to prove it is quite distinct. The genus was not previously known from Australian waters. Loc. — East of Flinders Island, Bass Strait, 70-100 fathoms. Two specimens. Family ANTEDONID^. Grenus Compsometra, A. H. Clark. COMPSOMETRA INCOMMODA {Bell). Antedon incommoda, Bell, Ann. Mag. Nat. Hist., (6), ii., 1888, p. 404. Compsometra incommoda, A. H. Clark, Mem. Austr. Mus., iv., 15, 1911, p. 792. Loc. — South-east of Flinders Island, South Australia, 37 fathoms. Two specimens. Compsometra loveni (Bell). Antedon loveni, Bell, Proc, Zool. Soc, 1882, p. 534. Compsometra loveni, A. H. Clark, Proc. Biol. Soc. Washing- ton, xxi., 1908, p. 131. None of the specimens of Compsometra are especially note- worthy, but they give added weight to the correctness of Mr. A. H. Clark's opinion that there are two distinct species in the seas of south-eastern Australia. To judge from the present collection, as well as from previous records, C. incommoda would seem to be the more southern form, occurring from Port Jackson, New South Wales, southward and westward SEA LILIES, STARFISHES, ETC. — CLARK. 27 to Flinders Island, South Australia, while C. loveni occurs from Port Jackson northward to Claremont Island, Queens- land. It is true that C. loveni is recorded from Port Phillip, Victoria, but further proof of its occurrence there is desirable. Perhaps the specimen in the British Museum is the victim of a misplaced label or a slip of the pen. Locs. — Twenty miles north-north-east of Double Island Point, southern Queensland, 30 fathoms. One specimen. Thirteen miles north by west of Double Island Point, Queensland, 25-26 fathoms. One specimen. ASTEROIDEA. The Starfishes, which the "Endeavour" collected, may fairly be counted as one of the most interesting lots which has been brought together by any one vessel since the " Challenger's " famous voyage. It is not so large as some that have been gathered by the " Albatross," but its value is disproportionate to its size, since it adds so much to our knowledge of the Australian marine fauna, three-fourths of the species taken being remarkable for one reason or another. The collection consists of two hundred and sixty-five speci- mens representing forty-four species, but fifteen of the species are represented by a single individual each ; only four species are represented by more than twenty specimens each and of none are there twenty-five. As a natural consequence there is little opportunity to discuss growth changes or even individual diversity. Of the forty-four species, twenty-one are new to science, four are new to Australia and nineteen were previously known from that continent. But even among these nineteen species, there are a number whose occurrence in the " Endeavour " collection adds important data to our knowledge of their distribution. There is one new genus and several of the new species compel modifications of the diagnoses of the genera in which they are placed. The occurrence of the genera Lonchotaster, Mimaster, Mediaster, Calliaster, Ogmaster, Pteraster, Zoroaster, Pedicellaster and Odinia in Austrahan waters is of great interest ; several of them have not previously been known in that quarter of the globe, and at least three of them are characteristic of the North Atlantic, where all, save Calliaster and Ogmaster occur. The geographical distribution of the species is of much interest, as it reveals a very marked difference between the eastern, southern and western coasts of the continent. Of 28 " ENDEAVOUR " SCIENTIFIC RESULTS. the forty-four species, one from Lord Howe Island and one from Masthead Island, were not collected by the " En- deavour " but are from shore collections made by naturahsts not connected with that vessel. Of the remaining forty-two, twenty-three were taken only in southern Australian waters, that is, south of Lat. 37° and east of Long. 128°, or in other- words, on the coasts of South Australia, Victoria and Tas- mania ; of these twenty-three, nine were taken in the Great Australian Bight and three, all new to science, only there. There are eleven species from the eastern coast, of which ten were not found south of Port Jackson, while one was found at Shoalhaven Bight ; only two of the eleven are new foi'ms, the other nine being Avell-known species of tropical Australia. Only three species, all previously known, were found on both, the eastern and southern coasts. There are four species from the western coast only, north of Cape Naturahste, all of which, are very distinct and highly characteristic forms. Of one species, specimens are present from Western AustraUa and from Tasmania, but the identity of the two is not perfectly estab- lished. There are no species in the " Endeavour " collections represented by specimens from ail three coasts, but a few are known with that extensive range, such as Luidia mactdata, Coscinasterias calamaria and AllosticJiaster polyplax ; the two latter are distinctty southern species but the Luidia is a tropical form and its occurrence on all sides of the continent is most interesting. Two species, Or easier gracilis and Ansero- 'poda rosacea occur on both the east and west coasts, but not on the south ; like the Luidia they are tropical species. It is probable that Plectaster decanus is a southern species occurring on all three coasts, but the evidence is as yet incomplete. About one hundred species of Starfish had been recorded from Australia, including Lord Howe Island and Tasmania, prior to 1915, but of these some twenty are of very doubtful, authenticity and ten others, while vahd species, are of very unlikely occurrence in Australia. The " Endeavour " col- lections include twenty-five additions to the fist, thus making up for most of those which must be stricken from the full list. It may not be out of j^lace to add that the Museum of Com- parative Zoology collection contains sixteen additional species from Torres Strait, and we are therefore safe in affirming that more than a hundred valid species of starfish are known from Australian waters, in less than three hundred fathoms. Of these at least seventy, or more than two-thirds, are peculiar to Australia. SEA LILIES, STARFISHES, ETC. — CLARK. 29 Family LUIDIID^. Genus Luidia, Forbes. LUIDIA FORFICIFER, Slttden. Luidia forficifer, Sladen, Chall. Rep., Zool., xxx., 1889, p. 258. This individual seems to be undoubtedly conspecific with the one taken by the " Thetis " in Newcastle Bight, New South Wales, at about the same depth. Both specimens differ irom Sladen's types, which were taken north of Australia, in the absence of the pedicellariae along the ambulacral furrows ; in the New South Wales material, the pediceUarise occur only close to the mouth. It is possible that more and better material will show this difference to be quite inconstant, but on the other hand, direct comparison of specimens from the northern coast of Australia with specimens from the south-east coast may demonstrate that they represent two distinct species. Loc. — Six miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. LrriDiA maculata, Muller and Troschel. (Plate V.) Luidia maculata, Miiller and Troschel, Sys. d. Ast., 1842, p. 77. The occurrence of this East Indian species in the Great Australian Bight is particularly noteworthy. All of the six specimens have seven rays. Logs. — East of Eraser Island, Queensland, 25-30 fathoms. Eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. South-east of Flinders Island, South Australia, 37 fathoms. Family ASTROPECTINID^. Genus Astropecten, Gray. AsTROPECTEN PECTiNATUS, Sladen. Astropecten pectinatus, Sladen, Journ. Linn. See, Zool., xvii., 1883, p. 251. Most of these twenty -one specimens are large adults (largest with R = 125 mm.) but half a dozen are still immature, with R only 35-40 mm. There are no specimens of intermediate size. In colour there is considerable diversity ; the specimens from 30 " ENDEAVOUR " SCIENTIFIC RESULTS. Oyster Bay, Tasmania, are uniformly gray, those from the south coast of AustraHa are pale yellow, about half the adults are pale fawn-colour, and the remaining specimens are red- brown. Both the " Thetis " and the " Endeavour " collec- tions show that this is a common and widely distributed species, apparently characteristic of south-eastern Australia. The careful working out of its growth changes would be of the greatest value to students of starfishes, as it would throw much light on the validity of specific characters in the large, cosmopolitan genus to which it belongs. It is to be hoped that some Australian zoologist will soon undertake the work. Logs. — Eighteen miles south-west of Lady Elliott Island, Queensland, 18 fathoms. Eleven miles south by east of Ballina, New South Wales, 27-28 fathoms. Six miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. Shoalhaven Bight, New South Wales, 15-45 fathoms. Oyster Bay, Tasmania. Southern coast of Australia. Great Australian Bight, 80-120 fathoms. AsTROPBCTEN PREissii, Mulhr and Troschel. Astropecten preissii, Miiller and Troschel, Arch. f. Naturg.,. ix. i., 1843, p. 119. Four of the rays in this individual are regenerating, while the fifth is broken in two ; the latter, however, has no part lacking, and measures 150 mm. in length. Since r=only 20 mm., J?=7J r, whereas in Miiller and Troschel's type R=5^ r. In other particulars, however, this specimen agrees well with Miiller and Troschel's description, and since their type is said to have come from south-western Australia, it seems best to refer this individual to their species. Loc. — South-east of Flinders Island, South Australia, 37 fathoms. Genus Lonchotaster, Sladen. LONCHOTASTER MAGNIFICUS,! Sp. nOV. (Plate VI., fig. 1-2.) R=155 mm. ; r=50 mm. ; R=3r. Br=45-50 mm. ; R =3-3.5 br. Disk very large, flattened but elevated in the central area and along the radii ; rays broad, flat, tapering 1. Magnificus=su\)eTh ; in reference to its notable size and appearance. SEA LILIES, STABFISHES, ETC. — CLAEK. 3L unifomily to a blunt point. Disk and rays, wdthin area bounded by superomarginal plates, covered by small (little exceeding 1 mm. in diameter of the slightly capitate top), crowded paxillae, each of which bears numerous (50-100) short, slender, dehcate, somewhat glassy spinelets ; paxillse without definite arrangement, except along sides of rays,, where series parallel to the width-diameter of the supero- marginal plates may be distinguished ; there are usually 2, but may be 3 such series adjoining each superomarginal plate. Madreporic body very large, 14 mm. across, carry-ing about 50 paxillae Avhich tend to conceal it verj^ completely ; these paxillse are lower and less perfectly formed than those on the disk itself ; the outer margin of the madreporite is 9 mm. from the inner margin of the superomarginal plates. The latter are 46 or 47 in number on each side of each ray ; those near the interradial angle are about 2 mm. long by 6 mm. wide, but as one passes distally length increases and breadth decreases and near the middle of the arm they are 3 mm. long and 4 mm. wide ;. the length never quite equals the width ; all of the superomarginals are densely covered by minute spinelets like those of the paxillse ; there are no spines or enlarged spinelets anywhere on the abactinal surface. Terminal plate small, sUghtly swollen, smooth, pentagonal- cordate, with rounded angles. Infero marginals correspond in number and position with the superomarginals but are much larger and project con- spicuously beyond them, except near tip of ray ; the inter- radial ones are 2 mm. long by 9 mm. wide, while those near the middle of the arm are 3 mm. by 6 mm. ; their covering consists of a close coat of short, flattened, blunt spinelets, much coarser than those on the superomargmals. Fasciolar channels between marginal plates neither pecuhar nor con- spicuous. Actinal intennediate plates wanting at tip of ray ; the series adjoining the adambulacrals begins at about the tenth inferomarginal from tip ; a second series begins at about the 17th-20th inferomarginal from tip ; a third at about the 30th ; the last becomes irregular and difficult to follow before reaching the mouth. The remainder of the large, actinal inter- radial areas are covered by similar but smaller plates forming parallel, radial series between adambulacral and infero- marginal plates ; these radial series are fairly regular ; usually a distinct unpaired series of about 5 plates occupies the midradial lines, but reaches little more than half-way to the inferomarginals. All the actinal plates are covered, like the inferomarginals, with short, flattened, rough spinelets, \\dder and blunter on central area of each plate than along its- margins. There are no spines or enlarged spinelets anywhere. "32 " ENDEAVOUR " SCIENTIFIC RESULTS. Adambulacrals about 57 on each side of each furrow, dis- tinctly longer than wide, except near arm-tips ; furrow margin with 8 conspicuous spines, the adoral and aboral ones being distinctly smallest, the others about 3 mm. long ; outside of this series is a well-spaced line of 3-5 shorter and more slender spines and external to them is an irregular line of 4-6 small, slender spinelets ; a few extra spinelets may occur on the outer <x)rners of the plates, but there is no tendency to form pedi- cellarial groups. No pedicellarise are to be seen. Oral plates large, swollen, covered by flattened spinelets similar to those on the actinal plates, but distinctly larger ; oral spines coarse and heavy, but neither number nor arrangement is peculiar. Colour, light dirty grayish, more or less yellow on various irregular areas. This individual is much larger than either of the species, collected by the " Challenger," and for which the genus was established, is known to be, and it is not impossible that it will prove to be an adult specimen of L. forcipifer, which was taken south of Australia in 1950 fathoms. This seems to me highly improbable, however, for the difference in the form, in the paxillse and in the matter of pedicellariae are very marked. Nevertheless I think there can be Httle doubt that the two forms are congeneric. Loc. — Great Australian Bight, 80-120 fathoms. Genus Psil aster, Sladen. PsiLASTER ACUMiNATUS, Sladen. Psilaster acuminatus, Sladen, Chall. Rep., Zool., xxx., 1889, p. 225. Except in size, these four specimens show little diversity ; in the smallest R=42 mm.; in the largest R = 100 mm. The actinal interbrachial areas differ considerably in size in the different specimens ; in the one from the Eastern Slope, Bass Strait, they are relatively much larger than in Sladen's figure, which answers well for the other specimens. All four indi- viduals are somewhat rubbed, and the two largest are in rather poor condition ; in none do I find any trace of spines on the superomarginal plates. Logs. — Eastern Slope, Bass Strait, 80-200 fathoms. South of Gabo Island, Victoria, 200 fathoms. Great Australian Bight, 80-120 fathoms. SEA LILIES, STARFISHES, ETC. CLARK. 33 Family GONIASTERID^. Genus Mimaster, Skiden. ]\IlMASTER GRACILIS,! Sp. nOV. (Plate VII., fig. 1-2.) R=60 mm.; r=18 mm.; R=3.3 r. Br=18 mm.; 11=3.3 br. Br at middle of ray, 8 mm. ; at tip, 3 mm. Disk large, very flat in holotype, but much thicker and more elevated in other specimens ; rays narrowing rapidly from the rather broad base, at first flattened but more or less cyhndrical near tip. Abactinal surface of disk and rays covered by rounded or substellate plates, which, near centre of disk and along the radial lines are quite closely crowded ; each plate bears a paxilla crowned with 10-12 (or more) slender spinelets exceeding the length of the paxilla stalk ; the appearance of preserved specimens abactinally depends primarily on whether these spinelets were drawn together into a cylindrical bundle on each paxilla, when the starfish died, or whether they re- mained more or less flaring ; the spinelets are so delicate, they are frequently more or less broken. Madreporic body large, about 5 mm. in diameter, fully exposed ; its distal margin is about 2 mm. from the superomarginal plates. The latter are 50 or more in number, very small, scarcely larger than some of the abactinal plates ; each carries an ordi- nary paxilla. As a rule, they are above and opposite corre- sponding infero marginals but here and there they alternate with them ; often the series is not easy to distinguish. In a small specimen, R=27 mm., the superomarginals are rela- tively much larger ; particularly in the interbrachial arc they are well-marked oblong plates, longer than wide, and forming a very regular series. Terminal plate moderate, roughly dumb-bell shaped, being deeply notched both distally and proximally ; apparently, in life, it is covered by minute spinelets, but in the preserved specimens, except the youngest, these are all rubbed oft'. Infero marginals of about the same number as supero- marginals or somewhat fewer, larger and forming a more perfect series ; their paxillae are larger and carry longer spinelets, but they are by no means conspicuous. Actinal intermediate plates wanting at tip of ray ; the series adjoining the adambulacrals begins at about the tenth inferomarginal plate from tip ; a second series begins at about the seven- teenth plate and a third at about the twenty-second ; at the L (?raa7is=slender ; in reference to the very narrow rays. 34 " ENDEAVOUR " SCIENTIFIC RESULTS. base of the ray there are 8 such series. Aetinal interradial areas large, covered by the transverse series of squarish, scarcely overlapping plates ; these plates have the paxilla stalk reduced to an oval knob, but the spinelets are long, numerous and flaring. Adambulacral plates more than 70, wider than long ; those nearest the mouth are about 2 mm. wide by a millimeter long ; furrow margin with 3, or usually 4, slender, blunt, somewhat flattened spines of which the middle one or two are longest and may be 2 mm. long ; outside of this series are three or four parallel series of similar but smaller spines ; the whole plate is thus fairly well covered by 15-20 spines, the marginal ones largest, the outermost like those on the aetinal plates. There are no pedicellarise. Oral plates conspicuous, swollen ; each bears a curved ridge 5 mm. long, carrying a double series of spines ; these ridges lie with concave surfaces together, so that although in contact distally and proximally, they are well separated at the middle ; the spines at the distal end are like the spines on the aetinal plates but become longer and stouter proximally and at the tip of the jaw are larger than any of the adambulacral spines. Colour, light yellowish-brown or dirty yellowish. Five specimens. The previously known species of this genus are from the North Atlantic, from the Pacific, west of the southern point of America, and from the Moluccas. The discovery of this fourth species south-east of Australia is therefore of great interest. There is little reason to doubt that it is a Mimaster, but it is easily distinguished from the previously known species by the shape of the rays, and by the adambulacral armature. In the appearance and position of the madreporic body, the Australian species resembles M. cognatus, the South American species ; in the North Atlantic species M. tizardi, the madreporic body is concealed by paxillse. The genus is a perplexing one and its real relationships are still obscure. Locs. — South of Gabo Island, Victoria, 200 fathoms. This specimen has been selected as the holotype. Between Gabo Island, Victoria, and Disaster Bay, New South Wales, 50-100 fathoms. Genus Nectria, Gray. Nectria ocellata, Perrier. Nectria ocellata, Perrier, Arch, de Zool. exp. et gen., v., 1876, p. 4. This is a nice series of twenty-two specimens, the smallest having R = 1 7 nnii . and the largest, R = 1 30 mm. In the latter SEA LILIES, STARFISHES, ETC. — CLARK. 35 r=35 mm. and br=40 mm. Comparison of these two specimens brings out the interesting fact that there is ahnost no increase, during growth, in the number of the big paxilh- form ossicles ; in the small specimen there are 7 in the midradial line of each arm, and about 8 in each interradial area ; the whole dorsal surface is almost completely covered hy them. In the big specimen, there are only 7 or 8 well formed ones on each radial line, while in the interradial areas the number is not more than 8 or 10 ; along the sides of the basal part of each ray nev\' ones have formed, but they are smaller than the others and merge into the ordinary abactinal plates. But in the large specimen, the ossicles are widely separated from each other, the abactinal skeleton and papular areas being fully exposed. In the small specimen, there are only 12 marginal plates in each series on each ray ; they are closely joined with no trace of intermarginal papulae. Even in the large specimen, however, I fail to find any intermarginal papulae. Locs. — North-east of Cajje Pillar, Tasmania, 55-80 fathoms. Oyster Bay, Tasmania, 60 fathoms. Off Babel Island, Bass Strait, 50-60 fathoms. Between Davenport and Launceston, Tasmania. Forty miles west of Kingston, South Austraha, 30 fathoms. Great Australian Bight, 80-120 fathoms. Nectria ocellifera [Lamarck). Asterias ocellifera, Lamarck, Anim. sans Vert., ii., 1816, p. 553. Nectria ocellifera, Gray, Ann. Mag. Nat. Hist., vi., 1840, p. 287 {oculifera, lap. cal.). "^rhese five specimens are probably from the same lot as those in the Western Australian Museum at Perth, by means of which it was possible to clear up the difference between N. ocellata and N. ocellifera.'^ There can be no doubt, I think, that the two species are perfectly distinct, and the " Endeavour " collections seem to confirm the view that their areas of distribution do not overlap. The specimens of N . ocellifera at hand are all small, having R only about 40 mm. Loc. — Between Geraldton and Cape Naturahste, Western Australia. 1. H. L. Clark— Rec. West Austr. Mus., i., 1914, p. 139. 36 " ENDEAVOUK " SCIENTIFIC RESULTS. Genus Nymphaster, Sladen. As regards the generic name, it is to be regretted that Koehler still persists in using Dorigona in place of Nymphaster^ Sladen, in 1889, Verrill, in 1899, and Fisher more recently, have pointed out that Dorigona is not one of Perrier's genera, but dates back to Gray. Its type is either D. reevesii, Gray or D. longimana (Mobius) ; no other species can be the type since these are the only species mentioned by Gray, who, however, designated no type. It has been shown, and, so far as I know, never questioned, that D. reevesii, Gray is synonymous with Goniodiscus capella, Miiller and Troschel, 1842. Now von Martens constituted a subgenus, Ogmaster, for G. capella in 1865, a year before Gray proposed Dorigona. All later writers have agreed that G. capella is entitled to generic rank and, including Koehler himself, use Ogmaster capella as the correct combination. Hence Dorigona longimana (Mobius) must be the type of Dorigona, and I am quite unable to see why Sladen should have dehberately replaced Gray's name with one of his own coining, Iconaster. So far as I can see Iconaster is a pure synonym of Dorigona, which is a monotypic genus with Astrogonium longimanum, Mobius, as its type. It is evident then that A^ymphaster and Dorigona are not in any way equivalent terms. Koehler proposes to restrict Nym- phaster to the forms in which the superomarginal plates are separated by a single median row of abactinal plates and use Dorigona for the forms in which the superomarginal plates are in contact dorsally. It is, of course, obvious that Dorigona cannot be used in any such way. Moreover Verrill, in 1899, proposed Nereidaster for the species with separated supero- marginals, and restricted Nymphaster to the species which Koehler would call Dorigona. This he was perfectly justified in doing, for Sladen designated no type for the genus in which he included five species. Clearly Dorigona, Koehler, is a perfect synonym of Sladen's Nymphaster as restricted by Verrill, and has no close relationship to Dorigona, Gray. It is to be hoped, therefore, that our able and honoured French colleague will no longer reject Nymphaster and will confine Dorigona to Mobius' remarkable species. Nymphaster pentagonus,i sp. nov. (Plate VIII., fig. 1-2.) R = 60 mm. or thereabouts ; r=18 mm. ; R=3.3 r. Br= 10 mm. ; R=6 br. Br at middle of ray, 6 mm. ; at tip, 1. Pentagonus =-vi\th. five angles, or, as usually understood, five-sided » in reference to the nearly straight-sided area of abactinal plates. SEA LILIES, STARFISHES, ETC. CLARK. 37 apparently about 2 mm. Breadth at proximal margin of first superomarginals which meet=length of first three supero- marginals (comiting from interradius). Disk flat, rather large ; rays moderately long and quite narrow, all broken at tip, but one seems to be nearly complete, quadrangular in cross-section with very sHghtly rounded angles, the width shghtly exceeding the thickness, at all points. Abactinal plates entirely confined to disk, forming an almost perfectly pentagonal area, the sides of which are so sHghtly concave as to appear ahnost straight ; there are 650-700 plates in this area but most of these are verj^ small and he along the boundaries formed by the superomarginal plates ; the largest abactinal plates are near the centre of the disk in the inter- radii and along the midradial lines. At the centre of the disk and in the rachal areas where the papulae occur, the plates are more or less stellate or substellate basally, the upper surface being roimded pentagonal or subcircular ; each plate carries 15-25 coarse granules, nearly spherical and not at all rough, those of the marginal circle not being appreciably different from the rest ; near the superomarginal plates, the abactinal plates are so small and crowded, the granules tend to form a continuous coat with those on the superomarginals. The latter are apparently rather more than 20 on each side of each ray, and excej^ting the basal 4 pairs are in complete contact in the midradial line ; they are very flat, not at all tumid, and in the interbrachial arcs overhang the infero- niarginals ; eight superomarginals make up each side of the disk-pentagon ; they are wider than long, the middle pair being each 5 mm. A\ide by 2.5 mm. long ; at the middle of the arm the superomarginals are each about 2 mm. long, 2 mm. wide, and 1.5 mm. thick ; all the superomarginals are covered by a uniform coating of granules similar to those on the abactinal plates ; these are largely rubbed ofl' in the present specimen ; on the entire surface of each of the middle pair of superomarginals, there are about 200 granules. Inferomarginals coincide in number and j)Osition with the superomarginals (except when regeneration has occurred), but owing to the presence of the ambulacral groove they are much narrower ; at the middle of the arm they are about 2 mm. long, 1 mm. wide and 1 mm. thick ; hke the superomarginals, they are clothed with a close coat of granules, each one of the middle pair carrjang about 150. Actinal intermediate plates wanting on rays ; the series adjoining the adambulacral plates, extends only to the fourth inferomarginal and contains 10-12 j^lates ; •outside of it is a series of 6-8 plates, then a series of 5-6, then 38 " ENDEAVOUR " SCIENTIFIC RESULTS. a series of 2 or 3, and then a single plate, against the infero- marginals, completes the covering of the area ; all these plates, tend to overlap inwardly and proximally ; all are l.earers of granules like those on the inferomarginal plates. Adambu- tacral plates about 40, at first wider than long, but rapidly decreasing in breadth and becoming twice as long as wide • distal to middle of arm there are about 7 adambulacrals to about 5 inferomarginals ; the furrow margin is at first nearly straight, but soon becomes convex and then the convexity becomes more and definitely restricted to the adoral end ; distal to the middle of the arm, these convexities on plates of opposite sides tend to meet across the furrow, the bridges, thus formed alternating with conspicuous pores where the plates are not in contact ; proximally there are 6 or 7 blunt, somewhat angular, more or less flattened spines, about half a millimeter long on the furrow margin of each plate ; th& middle spines are longest, the proximal and distal shortest ; on the plates distal to arm-middle there may be 8 or 9 spines in the series, but they are decidedly shorter ; outside the marginal series is a parallel set of 3-5 much smaller, thicker spines or high granules, and external to it a parallel series of 5-6 granules like those on the actinal plates ; additional similar granules may occur between the two external series or outside the outer one ; the adambu- lacrals may thus be about as well covered with granules as any of the actinal plates. Oral plates neither conspicuous nor peculiar, only the proximal spines enlarged and they not very greatly. Pedicellariae none. Colour gray, more yellow beneath ; nearly white where the granules are rubbed off. The markedly pentagonal abactinal area, the flat and over- hanging interradial superomarginals, the adambulacral arma- ture and the general proportions remind one very strongly of N. albidus, from the tropical Atlantic, but the adambulacral spines are angular and flattened in the Australian species and eight superomarginal plates, instead of six, make up the disk boundary on each side. None of the species from the Indian Ocean and Philippine Islands seem to be any nearer than N. albidus to this South Australian species. Loc— Great Australian Bight, 129° 28' E., 250-450 fathoms. One specimen. SEA LILIES, STARFISHES, ETC. — CLARK. 39 Genus Mediaster, Stimpson. Mediaster atjstr alien sis, 1 sp. nov. (Plate ix., fig. 1-2; Fig. 1-3.) R=82 mm.; r=30 mm.; R=2.67r. Br=30 mm.; R=2.67 br. Br at middle of ray=ll mm. ; at tip, 2.3 mm. In the largest specimen, R=100 and r=40 mm. Disk large and flat ; arms wide at base but tapering rapidly to a nearly cylindrical tip. Abactinal plates tabulate, more or less paxilli- form (Fig. 1) especially at centre of disk and along midradial Fig. l.xlO. Fig. 2.xl0. Fig. 3.xlO. lines ; small and numerous ; the largest and most paxilli- form are crowned with a marginal series of 12-15 angular, blunt spinelets or coarse granules and within this circle 3-8 similar but smaller granules ; the smallest and least paxilliform plates adjoin the superomarginals in the interradial areas and are crowned with a little group of 3-5 angular granules ; the coarseness or fineness of the paxillar granules shows great diversity in different individuals and even in different parts of the same disk ; thus in one specimen, plates proximal to the madreporite carry 10-20 coarse, well-spaced granules, while plates of the same size, distal to the madreporite, carry 25-40 smaller, more rounded and more crowded granules. Occasionally one or more of the granules is replaced by a 2-jawed pedicellaria ; in the holotype, these pedicellarise are very numerous and are so much higher than the granules (about twice as high), that they are quite conspicuous ; they are also quite symmetrical and frequently entirely replace the granules on the top of the plate (Fig. 2) ; in other specimens, pedicellariae are exceedingly rare, while in still others, though common, they are small and 1. Australia-{-ensis, expressing locality. 40 " ENDEAVOUR SCIENTIFIC RESULTS. asymmetrical ; they occasionally replace one or more of the marginal granules. Papulae numerous, small, single, occurring scattered more or less irregularly among the plates ; occasion- ally a regular series of 6 surrounds a large plate. Madreporite small, about equal to 2 or 3 ordinary abactinal plates, rounded triangular, only half as far from centre of disk as from disk margin. Superomarginal plates about 40 on each side of each ray (48 in the largest specimen), those in interradial region nearly twice as wide as long, more nearly square at middle of arm and then becoming much wider than long again near tip ; all are closely covered with granules, similar to but more rounded, and the marginal ones at least, distinctly coarser, than those on the largest abactinal plates ; there are about 100 on each of the superomarginals near the interradius ; occasionally pedi- cellarise replace granules. Terminal plates small, distinctly longer than wide, not deeply notched distally. Infero marginals apparently one less on each side of each ray than superomarginals, at least in the larger specimens ; the number may be the same in the two series, but in many cases the series do not correspond exactly along the middle of the arm ; the granular covering of the inferomarginals is hke that of the other series. Actinolateral plates in 8-9 series ; the first (adjoining the adambulacrals) extends from the oral plates to the tenth inferomarginal from the tip of the arm ; the next series extends to the twentieth inferomarginal from arm-tip ; the next series only reaches the eighth or ninth inferomarginal, counting from interradial mid-line ; remaining series irregular and of few small plates ; all the plates carry coarse, angular granules like those on the abactinal plates, but no pedicellarise occur ; at least I have seen none. Adambulacral plates squarish or rather wider than long, about 60 (in the holotype), the inner margin at first straight but becoming decidedly convex distally ; furrow margin with a series of 5 or 6 flat, blunt spines, rather crowded, nearly equal or the most adoral shortest (in the figure [Fig. 3 J the spines are too abruptly truncate and are too widely spaced) ; externally is a second series of 4 shorter, thicker, more angular spines ; a third series of 3 or 4 still shorter and thicker spines are borne on the outer margin of the plate ; between this series and the second is a rather conspicuous bare space ; the spines of the third series are scarcely to be distinguished from the angular granules on the actinolateral plates. Oral plates neither conspicuous nor peculiar ; the oral spines, especially at tip of jaw, are, however, svery heavy, more or less prismatic and 2-3 mm. long. Colour, SEA LILIES, STARFISHES, ETC. CLARK. 41 in alcohol, brownish-yellow ; dry, pale brown. Seven specimens. This fine species is obviously near to Mediaster ornatus, Fisher, from the Hawaiian Islands, but differs in the wider, shorter marginal plates and in the adambulacral armature. No other member of the genus is very closely alhed. As regards the internal anatomy, the Austrahan species is a typical Mediaster, the internal abactinal ossicles being very evident and rudimentary superambulacral plates also being present. Logs. — East coast of Fhnders Island, Bass Strait, 40 fathoms. This specimen has been selected as the holotype. East of Babel Island, Bass Strait, 65-70 fathoms. Oyster Bay, Tasmania. Mediaster monacanthus,i sp. nov. (Plate X., fig. 1-2 ; Fig. 4.) R = 110 mm. ; r=37 mm. ; R=3 r. Br=37 mm. ; R = 3 br. Br at middle of ray =19 mm. ; at tip =5 mm. Disk large, flattened or shghtly elevated ; arms wide at base, tapering rather evenly to the blunt tip. Abactinal plates on disc and base of rays large, low, hemispherical, widely spaced, becoming smaller and more and more crowded near the supero- marginals ; they are connected with each other by heavy radiating ossicles, in the spaces between which the papulae arise in groups of 3-15. Each plate is covered by a coat of low, spherical granules ; there are about 50-75 on a plate 4-5 mm. in diameter ; there are no granules on the membrane or on the ossicles between the plates. In the holotype, and some other specimens, most of the disk plates bear at the centre a pointed tubercle or low, blunt spine, about a millimeter in diameter and nearly or quite as high ; in some specimens few or no plates have this spine, so it may be an indication of complete maturity. Distally on the rays and near the margins of the disk, the plates become smaller and crowded, so the granular covering is nearly or quite continuous. Madreporite very small, only about 2 mm. across, situated twice as far from the margin as from centre of disk. Anus conspicuous near centre of disk because surrounded by 10-20 large, smooth, elongated granules of diverse size and shape. 1. y«oVos=single+ tt'/ca»'0«= a thorn or spine ; in reference to the single prominent spine on some abactinal plates. 42 " ENDEAVOUR " SCIENTIFIC RESULTS. Superomarginal plates nearly or quite vertical in position,, and hence scarcely to be seen from above, 30-32 in number, the interradial ones much higher than long, covered with granules similar to those on the abactinal plates ; there are about 200 granules on one of the big interradial supero- marginals. Terminal plate small, nearly circular, consider- ably swollen. Inferomarginals very similar to supero- marginals in number, position, size and form ; they have the same granular covering. In the holotype and in some other specimens intermarginal plates of irregular size, shape and position occur in considerable numbers, but this is not a constant feature, since in some individuals such plates appear to be rare if not altogether wanting. Ac tino lateral plates in about 7 series ; the jfirst, adjoining the adambulacrals, extends nearly or quite to the arm-tip, the second goes about the three-fourths of the way, and the third more than half ; all the actinolateral plates are covered by granules distinctly coarser than those on the abactinal surface. Adambulacral plates about 60, longer than wide, with a fairly straight margin ; on this margin is a series of 4 sub- equal, squarish, blunt spines, about 2 mm. long (occasionally a fifth, adoral, much smaller spine occurs) ; outside this series is a second, consisting of 3 or 4 shorter, but thicker spines, and external to these is a series of 4 much smaller Fig. 4.x 10. spines or angular granules, which do not much exceed the- granules on the actinal plates. Oral plates not conspicuous, or pecuhar ; even the oral spines are not strikingly larger than those on the adjoining adambulacrals. No pedicel- larise were detected anywhere. SEA LILIES, STARFISHES, ETC. — CLAKK. 43^ Colour (dry or in alcohol), brown, darkest on the bare membrane of the disk, hghtest near the tips of the rays. This is a very remarkable Starfish, one of the most inter- esting of the " Endeavour's " captures. It stands midway between Nectria and Mediaster having the heavy, widely separated abactinal plates of the former, but with these plates not at all paxilhform. It is very different from typical Mediaster in its general appearance abactinally, but actinally is very similar to M. australiensis, save that the spines and granules are much coarser. Supplementary internal abactinal plates and rudimentary superambulacrals are present. The ampullae are double. It might be made the type of a new genus between Mediaster and Nectria but there are already so many monotypic genera, one may well be avoided here, since the diagnosis of Mediaster, as at present used, does not reqiiire any essential modification to admit M. monacanthus. This specific name is not altogether a happy choice, since many specimens fail to show the single spinelet on the abactinal plates, so conspicuous in the holotype ; but as the name has been in vogue as a manuscript name for several years, and specimens so labelled have been sent out from Sydney, it seems best to retain it. It is a httle remarkable that this conspicuous Starfish was taken at only one station. Ten specimens. Loc. — Six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Genus Tosia, Gray. TosiA AURATA, Gray. Tosia aurata, Gray, Proc. Zool. Soc, 1847, p. 80. This is a fine series of this handsome pentagonasterid ; the longest has R=73 ; in the smallest R=40 ; in the least pentagonal r=.60R, and in the most pentagonal, r=.68R. One of the specimens from Oyster Bay, Tasmania, with R = 63 mm. is perfectly tetramerous. Eighteen specimens. Locs. — Oyster Bay, Tasmania, 20-40 fathoms. Isle de Phoque, Oyster Bay, Tasmania, 27 fathoms. Mainwaring Cove, west coast of Tasmania, 50 fathoms. South-east coast of South Austraha. South coast of South Australia. 44 " ENDEAVOUR " SCIENTIFIC RESULTS. Genus Pentagonaster, Gray. Pentagonaster stibarus, H. L. Clark. Pentagonaster stibarus, H. L. Clark, Rec. West Austr. Mus., i., 1914, p. 136. No doubt this specimen was taken with the types ; it adds nothing to our knowledge of the species ; there are, however, no abactinal tubercles, but R only=33 mm. By an un- fortunate slip of the pen in the original account of this species, in discussing the relative size of the marginal plates in the genus, I wrote swperomarginal for m/eromarginal, and it is so printed (p. 138, 1. 22), confusing the sentence. Loc. — Between Greraldton and Cape Naturahste, Western AustraHa. Genus Calliaster, Gray. Calliaster spinosus,! sp. nov. (Plate xi., fig. 1-2.) R = 140 mm. ; r=45 mm. ; R=3 r. Br=4() mm. R = 3.5 br. Br at middle of ray, 18 mm. ; at tip, 5 mm. Disk large, somewhat elevated at centre and along radii ; arms abruptly narrowed and tapering gradually to the blunt tip. Abactinal surface covered by moderately large stellate plates, the central portion of which is swollen into a flattened hemi- sphere ; each of these rounded centres is surrounded by a circle of coarse granules, which more or less conceal the papulae arising singly between the plates, and is crowned by a group of coarser granules, the central one of which is dis- tinctly the biggest ; occasionally, though not commonly, one of these crowning granules is replaced by a large 2-jawed pedicellaria fitting into a special depression in the surface of the plate ; each jaw of the pedicellaria is broadly spatulate, rounded at the tip and, when open, lies back in the depression, so as to seem flush with the surface of the plate. The " crowning " granules differ from the " marginal " granules, not merely in being, as a rule, larger, but in that they are more or less spherical, while the marginal ones are very flat. Along the midradial line of each ray, there is a series of plates which extend from the terminal plate, which is large and somewhat swollen (4 mm. long, 3 mm. wide proximally 1. (S'pmo5MS= provided with spines ; in reference to the spiny actinal surface. SEA LILIES, STAHFISHES, ETC. — CLAEK. 45 and 3 mm. thick), to within 15-20 mm. of the centre of the disk ; distally these plates are much like the other abactinal plates and are about 2 mm. in diameter ; near the middle of the arm and proximal to it they become compressed and are much wider than long (3-4 mm. wide by 1.5-2 mm. long), but the 4-6 most proximal are again more or less nearly circular ; these circular proximal plates are the largest of the disk and are very much swollen ; the most proximal, which is largest, is 5-6 mm. in diameter and 3 mm. or more high ; on these big proximal plates, the central granule becomes developed into a huge spine, 2 mm. in diameter ; only one of these persists in the specimen at hand ; that is on the fourth plate of one of the radial series and measures 3 mm. in height. The median radial plates are accompanied nearly, and sometimes quite, to the terminal plate, by a series of similar, but smaller and less conspicuous plates, without spines. At the centre of the disk, and in each interrachus, in line \\'ith the most proximal of the big radial plates, are enlarged plates, which apparently bore spines. The circular madreporite, 4 mm. across, lies just distal to one of these large interradial plates. Superomarginal plates 25-26 on each side of each ray, more or less bare and smooth, and distinctly swollen ; each plate is surrounded by a series of flattened granules : the plates in the interbrachial arc bear a vertically elongated cluster of spherical granules, among the lower ones a stout spine ; on the first supero- marginal (adjoining the interradial hne), this spine is about 6 mm. long and a milhmeter thick at base ; each succeeding plate has fewer granules, and the spine becomes stouter and stands nearer to the centre of the plate ; at the middle of the arm, the spine may be 2 mm. in diameter and there are virtually no granules on the plate ; distally the spines become steadily smaller ; in the present specimen none are left distal to middle of arm ; occasionally one, or even two, of the granules on the plate are replaced by the big, characteristic pedicellarise. Inferomarginal plates one more in number than the superomarginals and hence not corresponding exactly in position, except at each end of the series ; very similar in appearance but with more granules (especially at oral end, in interradial arc) and (to judge from the single one remaining) with more slender spines. Actinolateral plates wanting on distal half of ray ; the first series (adjoining adambulacrals) begins at about the ninths or eighth, inferomarginal (counting from interradial hne) ; the second series begins at the third inferomarginal and a third series at the second ; the rest of the large actinal area is 46 "endeavour ' SCIENTIFIC RESULTS. occupied by about ten similar plates ; these plates are more quadrilateral than the abactinal plates and differ in being less swollen, much more completely covered with granules and in having the central granule replaced by a more or less con- spicuous spine, 1-5 mm. long, 1-2 mm. in diameter, and bluntly pointed, clavate or chisel-shaped at tip ; these spines are smallest near the inferomarginals and largest near the adoral adambulacrals ; the granules are coarser than on the abactinal surface and there is less distinction between " crowning " and " marginal " granules ; many granules are replaced by the large, characteristic pedicellariae. Adambu- lacrals about 55, wider than long or squarish, quite similar to the actinolaterals in having marginal granules on the adoral, outer and aboral sides, a group or series of somewhat coarser granules on the outer end and a large spine at or near the centre ; these spines may exceed 5 ram. in length ; on the first two plates (near mouth) and occasionally elsewhere, a second spine, similar but much smaller, stands just behind the large one ; on the furrow margin of the plate is a series of 7-8 crowded, compressed, blunt spines of which the four or five middle ones are sub-equal, about 2 mm. long, while the aboral, and the adoral (1 or 2) are markedly shorter and more pointed ; on the adoral, marginal corner of the plate is a very large pedicellaria, of the same type as those elsewhere, but with jaws of very different shape, 1.5-2 mm. long, slenderly spatu- late and slightly asymmetrical, bluntly pointed. Oral plates not particularly large or prominent ; each has 7 or 8 large, furrow spines, the proximal ones very large and prismatic ; each plate bears on its surface two large spines hke those on the adjoining adambulacral plates. Colour (dry), pale yel- lowish-brown. This is the most notable and the best characterised of all the " Endeavour's " Starfishes, not previously known to science. The adambulacral armature with big pedicellariae and the spiny actinal interradial areas distinguish it at once from any other member of the genus, although it seems to be nearest the Phihppine species C. corynetes, since it has pedi- cellariae hke those of that species, spines on the actinal plates and an adambulacral armature of the same type. It is readily distinguished from C. corynetes, however, by the absence of spines on the terminal plate and by the granulose abactinal plates. The diagnosis of Calliaster as given by Fisher^ will require considerable modification to admit this fine new species from Austraha (and the Phihppine form, too), for 1. Fisher— Bull. U.S. Nat. Mua., Ixxvi., 1911, p. 171. SEA LILIES, STARFISHES, ETC. — CLAKK. 4/ there are abactinal pedicellariae, the plates are not bare centrally and the actinal intermediate plates have very con- spicuous spines. The vahdity and the sharp definition of the genus are not, however, in any way aflfected. Loc. — Eastern Slope, Bass Strait, 80-200 fathoms. Genus Ogmaster, von Martens. Ogm ASTER CAPELLA {Muller and Troschel). Goniodiscus capella, Miiller and Troschel, Sys. d. Ast., 1842, p. 61. Goniaster {Ogmaster) capella, von Martens, Arch. f. Naturg., xxxi. i., 1865, p. 359. Ogmaster capella, Sladen, Chall. Rep., Zool., xxx., 1889, p. 261. It is exceedingly interesting to find this Uttle-known East Indian species so far south as North Reef, Queensland. Loc. — Thirteen miles north-east of North Reef, 70-74 fathoms ; thirty-eight miles north-east of North Reef, Capricorn Group, off Port Curtis, Queensland, 50-75 fathoms. Five specimens. Genus Stellaster, Gray. Stellaster incei. Gray. (Plate xii., fig. 1-2.) Stellaster incei, Gray, Proc. Zool. Soc, 1847, p. 76. About half these specimens are dark brick red, of varying degrees of intensity, while the others are pale brown or dull yellowish ; the difference is not associated with locality. Regardless of the ground colour, certain individuals have, in marked contrast, a number of the actinolateral plates, ad- joining the adambulacral plates near the mouth, deeply stained with reddish-purple. All the specimens are dry. There is more or less diversity also in the number of tubercles on the abactinal surface, the extremes being 10 and 100, and their size and arrangement is equally diverse. The largest specimen has R nearly 90 mm. Twenty-three specimens. Locs. — Thirteen miles north-east of North Reef, Capricorn Group, off Port Curtis, Queensland, 70-74 fathoms. Wide Bay, Queensland, 18-23 fathoms. Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. 48 " ENDEAVOUR " SCIENTIFIC RESULTS Eleven miles south by east of Ballina, New South Wales, 27-29 fathoms. East-south-east of Clarence River mouth, New South Wales, 35-36 fathoms. Stellaster inspinosus,! sp. nov. (Plate xiii., fig. 1-2.) R = 60 or 75 mm. ; r=23 or 27 mm. ; R=2.6 r. Br at middle of ray, 11 or 14 mm. Very similar to 8. incei, but differing sharply in several features. The marginal plates, though of essentially the same number, encroach more on both surfaces than in S. incei ; as a result, the median radial series of abactinal plates is accompanied by lateral series only to about the fourth (from interradius) superomarginal and there are practically no papulae on the distal half of the arm, while in *S'. incei the lateral series extend to about the eighth plate and the papulae occur far out on the arm ; on the lower surface, there are only about 10 actinolateral plates in each area, aside from the series adjoining the adambulacrals, while in S. incei there are 12-20. The pedicellariae on the adambu- lacral plates are strikingly taller and more slender in S, inspinosus than in *S'. incei, while the big actinal spine is much smaller. More important is the fact that there are no spines on the inferomarginal plates, save that on some arms, the fifth (rarely the sixth) inferomarginal has a very small, flat spinelet, about .75 mm. long, on its outer distal corner. Finally, the granulation is finer in S. inspinosus than in 8. incei and the abactinal tubercles are fewer, 10 in the smaller specimen, but only 2 in the larger. Colour (dry), very pale brown. This Starfish is of particular interest not merely because it seems to replace *S'. incei on the west coast of AustraHa, but because it is apparently a connecting hnk between the genera Stellaster and Ogmaster. Fisher^ thinks " there is no particular similarity between Stellaster and Ogmaster," but I am unable to agree with him. Indeed, I should be inchned to say there is no important difference between the two. Ogmaster is only a Stellaster which has lost its spines and granulation, and this new West Australian species, *S'. inspinosus, is quite intermediate in that 1. I nspinosus=vf ithout spines; in reference to the absence of spines of the inferomarginal plates. 2. Fisher— Bull. U.S. Nat. Mus., Ixxvi., 1911, p. 169. SEA LILIES, STARFISHES, ETC. CLARK. 49 the granulation is finer than in other Stellasters and the marginal spines are gone ; the abactinal spines too are much reduced in number and even the actinal, adambulacral spine is somewhat smaller than in other species. If Ogmaster is not a close relative of Stellaster, it is hard to see what its correct position is. Loc. — Between Geraldton and Cape Naturahste, Western AustraUa. Two specimens. Grenus Anthenea, Gray. Anthenea acuta. Perrier. Goniodiscus acutus, Perrier, Ann. Sci. Nat., Zool., (5), xii., 1869, p. 280. Anthenea acuta, Perrier, Arch, de Zool. exp. et gen., v., 1876, p. 91. This is an excellent series, with R ranging from 40 to 120 mm., and the vertical diameter of the disk from 12 to 50 mm. The growth changes are well shown, as well as considerable individual diversity. Twenty -fo in: specimens. Logs. — Off Fraser Island, Queensland, 32-33 fathoms. Off the Queensland coast. Twenty miles north-north-east of Double Island Point, Queensland, 30 fathoms. Thirteen miles north by west of Double Island Point, Queensland, 25-26 fathoms. Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. Eleven miles south by east of Ballina, New South Wales, 27-29 fathoms. Eleven miles east-south-east of Clarence River mouth. New South Wales, 35-40 fathoms. Eight miles east of Sandon Bluffs, New South Wales, 40 fathoms. Family OREASTERID^. Genus Oreaster, Muller and Troschel. Oreaster australis, Lutken. Oreaster australis, Liitken, Vid. Med., 1871, pp. 252, 263. These are all dry adults, illustrating not only the great individual diversity in abactinal spines, but also the diversity caused by differences in collapse and contraction when drying. 50 " ENDEAVOUR " SCIENTIFIC RESULTS. The proportion of R to r shows little diversity, ranging only from 2.3 to 2.5. The largest specimen has R = 155 mm.; r=60 mm. ; br at middle of ray, 30 mm. In colour, these specimens are either tawny brown or deep brick red ; one of the red ones has the central abactinal area hght yellowish- gray in sharp contrast. Sixteen specimens. Locs. — Eighteen miles south-west by south of Lady ElHott Island, Queensland, 18 fathoms. Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. Oreaster gracilis, Liitken. Oreaster gracilis, Liitken, Vid. Med., 1871, pp. 260, 264. The fact that this specimen comes from the same station as all but one of the specimens of 0. australis makes me hesitate to designate it by a different name, yet it is obviously different. R=155 mm. ; r=48 mm. ; br at middle of ray=22 mm. R = 3 r. Not only are the rays thus much longer and more slender proportionally than in 0. australis, but the meshes of the abactinal skeleton are noticeably finer, and hence the papular areas are much smaller. The colour is hght yellowish-gray. I think there is no doubt that the specimen is 0. gracilis but I beheve it may ultimately be shown that 0. gracilis is only a form of 0. aus'tralis, the pecuharities of which are emphasised in drying. Loc. — Eighteen miles south-west by south of Lady ElKott Island, Queensland, 18 fathoms. Genus Asterodiscus, Gray. Asterodiscus truncatus, Coleman. Asterodiscus truncatus, Coleman, Mem. Austr. Mus., iv., 14, 1911, p. 699. This is a fine series of this notable species, but there is nothing to add to Coleman's careful description, save that the species grows to a size much larger than his types. His ho lo type has R = 103 mm., while his figured specimen has R = 138 mm. The smallest of the " Endeavour " specimens has R = 83, while the large specimens have R = 150-165 mm. SEA LILIES, STARFISHES, ETC. CLARK. 51 It is virtually impossible to distinguish the marginal plates, except partially in the small specimens. In the large speci- mens papulae extend over onto the actinal surface Nine specimens. Locs. — Six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Eastern slope, Bass Strait, 80-200 fathoms. Coast of Victoria. Great AustraUan Bight, 80-120 fathoms. Family LINCKIID^. Genus FiioivnA, Ch-ay. Fromia polypora,! sp. nov. (Plate xiv., fig. 1-2.) R = 86 mm. ; r=18 mm. ; R=4.7 r. Br=-19 mm. ; R = 4.5 br. Br at middle of ray, 14 mm. ; at tip, 4 mm. Disk moderate^ large, not specially flattened ; arms more or less flattened, but cyhndrical near tip. Whole abactinal surface covered by more or less circular plates about one millimeter in diameter, some larger, some smaller, with no definite arrangement whatever ; scattered everywhere are the numerous rather large, single papulae ; the plates are covered by poly- gonal, more or less flattened granules, 10-30 to. each plate, the marginal ones smallest, the central ones noticeably, but not conspicuously larger. Superomarginal plates about 37, not at all conspicuous, but the proximal ones are much larger than the abactinal plates. Infero marginals of about the same number as the superomarginals and distally of about the same size and alternating with them ; proximally they are directly beneath them and are considerably smaller. All the marginal plates are covered by granules Hke those of the abactinal surface ; there are papulae scattered among them. Terminal plate rather large, wider than long, covered by granules. Madreporite small (2 mm. across), rounded triangular, half way between centre of disk and margin. Actinolateral plates in three series ; that adjoining the adambulacrals extends to the fourth inferomarginal from tip of arm ; the next extends to the ninth ; the third extends to the eleventh or thereabouts, but is distally irregular and 1. TroXwTro'^jo 9= having many ways through ; in reference to the numerous papular pores. 52 " ENDEAVOUR " SCIENTIFIC RESULTS. broken up ; all the plates resemble those of the abactinal surface. Papulae occur everywhere except between the adambulacrals, and on distal half of ray, between the adam- bulacral series and the adjoining actinolateral series. The granulation on the actinolateral plates near the mouth is coarser, more angular and more widely sjiaced than elsewhere. Adambulacral plates more than 70, wider than long ; furrow margin with 3 (distally 2) stout, prismatic, blunt spines, rather more than a millimeter long ; back of this series is a second of 2 or rarely 3, short, stout, angular spinelets or granules ; external to them are 5-8 smaller, angular granules similar to those on the actinolateral plates ; except near the mouth, these are usually in two series with 1-3 granules in the inner and 4 or 5 in the outer. Oral plates not conspicuous or peculiar, the marginal and surface spines being essentially hke those on the adambulacral plates, though longer, more pointed and more sharply prismatic. No pedicellariae any- where, unless pairs, trios and quartettes of low, angular spinelets, more or less appressed to each other, on some of the plates near the mouth, are to be so interpreted. Colour (dry), dirty brownish- white. Three specimens. The form and size of this Fromia, in connection with its comparatively uniform granulation and its adambulacral armature, easily distinguish it from the other members of the genus. I do not feel certain that the Western Austrahan speci- men is conspecific with those from Tasmania ; it has R=65 mm.; r=16mm. ; br=16mm. ; br at middle of ray 11 mm., and at tip 3 mm. ; hence R is only 4 r, and only 4 br ; more- over, the arms are httle flattened and taper uniformly to the slender tip, while there are, as a rule, 3 spines in the second adambulacral series. These differences are so shght, it is sufficient to mention them here and await more abundant material. I am further constrained to do this because there is a Fromia in the Museum of Comparative Zoology collection, from Westernport, Victoria, the gift of Mr. J. Gabriel, which is certainly very near to the present species. In it R = 65 mm., and r and br are 16 mm., but br at middle of ray is 13 mm., and at tip is nearly 4 mm. ; hence the individual is much stouter in appearance than either the Western Austrahan or Tasmanian specimens ; the granulation and adambulacral armature also are noticeably coarser and the papulae, especi- ally on the actinal surface, seem larger ; the terminal plate and adjoining marginal and abactinal plates are perfectly smooth, but it is possible that the granulation has been accidentally rubbed off. A good series of Fromias from the western and southern coasts of Australia, and from Tasmania, SEA LILIES, STARFISHES, ETC. — CLARK. 53 will be necessary before we can be sure that we are dealing with only a single species. Locs. — East of Maria Island, Tasmania, 78 fathoms. Between Geraldton and Cape Naturahste, Western Aus- traha. Grenus Ophidiastee,, L. Agassiz, Ophidiaster gonfertus,! 5^. nov. (Plate XV., fig. 1-2.) R = 81 mm. ; r=-9 mm. ; R = 9r. Br=10 mm. ; R=8 br. Br at middle of ray =10 mm. ; at tip=6 mm. Disk small ; rays 5, long and cylindrical. Abactinal plates in 3 regular series, about equal in size to marginals. Papular areas in 8 regular series, one on each side of ray, on actinal surface ; each area contains 15-20 papulae. All abactinal and mar- ginal plates more or less conspicuously tumid. Entire animal clothed in a densely granular skin ; on papular areas, there are 30-40 granules, or even more, to each square milli- meter, but on convexities of abactinal plates, there may be only 10-15, though there are often more. Madreporite single, 2-3 mm. in diameter. Terminal plate scarcely more than a miUimeter in diameter, bare. Adambulacral plates short and numerous ; on the furrow-margin are 2 blunt, rounded, but flattened spines, about 1 mm. long ; the aboral is a trifle the smaller as a rule ; the tips of these adambulacral spines are separated from each other by granules such as cover the surface of plates ; near the mouth each plate carries an oval, blunt, flattened spine about a milHmeter long and rather more than its own length from the furrow-series ; distally, the.se subambulacral spines often occur only on every other plate and the transition from being on every plate to being on every other plate may be abrupt ; in such cases the spines near the mouth seem crowded. Oral plates not peculiar, small and completely buried in granules. Pedicellariae more or less abundant everywhere, but particularly near the mouth and on the papular areas ; they are of the well-known type occurring in 0. germani, 0. lorioli, etc. ; the whole organ is about .75 mm. long by .25 mm. wide and each jaw has 3-5 teeth. Colour (dry), yellow-brown, 1. t7ore/ert««= crowded ; in reference to the tendency of the subambu- acral spines to be very near together proximally. 54 " ENDEAVOUR " SCIENTIFIC RESULTS. Although not taken by the " Endeavour," this specimen was included in the collection, that a description and figures might be published. I have also had for comparison, and have used in drawing up the above description, two specimens also from Lord Howe Island, in the Museum of Comparative Zoology collection, received from the Austrahan Museum several years ago. As these bore the manuscript name O. confertus, it has seemed well to retain it. The species is near 0. germani, but has only a single madreporite and the arms are much longer. In my report on the " Thetis " Echinodermatai I, with some hesitation, referred an Ophidiaster from Lord Howe Island to Perrier's species, but in view of these additional specimens, I am now inchned to think these is little doubt that 0. germani is quite distinct. The character, to the presence of which my hesitation was due, is not, however, a constant one, no two of the specimens being alike in the arrangement of the subambulacral spines, although they are, in all cases, numerous and regularly arranged near the mouth. Loc. — Lord Howe Island, South Pacific Ocean. Genus Pseudophidiaster,^ gen. nov. Linckiidse with three conspicuous series of abactinal plates and eight series of papular areas, at base of arms, the actinal series tending to disappear distally, near the tip of the pointed arm. Body-wall thick and leathery with little or no rigidity. Adambulacral armature in a single series. Actinolateral plates in several series at base of arms ; no papular areas between them. Pedicellarise present. Madreporite very large. Ampullae single, large, with numerous calcareous plates in their walls. No superambulacral plates. Type-species. — Pseudophidiaster rhysus, sp. nov. This remarkable genus does not differ structurally from Ophidiaster in any very fundamental points, but in general appearance the difference is very great. Owing to the lack of rigid abactinal skeleton, dry specimens are very much wrinkled and the longitudinal series of plates are made unnaturally conspicuous ; in alcohoHc specimens, longitudinal furrows are not conspicuous, but numerous transverse grooves 1. H. L. Clark— Mem. Austr. Mus., iv., 11, 1909, p. 529. 2. \lrevBn'i=ia\se-\-ophidiaster ; in reference to its resemblance to that genus not being apparent. SEA LILIES, STAKFISHES, ETC. — CLAEK. 55 are evident, especially on the sides of the arms. The super- ficial appearance of alcoholic specimens of Pseudophidiaster is more like that of Phataria than like that of Ophidiaster, but the latter genus seems to be structurally more nearly allied. Pseudophidiaster rhystjs,i sp. nov. (Plate xvi., fig. 1-2.) R = 150mm. ; r=25 mm. ; R=6r. Br=27 mm. ; R=not quite 6 br. Br at middle of ray, about 15 mm. ; at tip, 3-4 mm. Disk large, considerably elevated, the median radial series of plates extending outward from it in equally elevated ridges ; rays therefore more or less distinctly trigonal at base, but tapering quite uniformly to a cyHndrical, bluntly pointed tip. Whole animal clothed in a thick skin, bearing a dense coat of crowded granules, and conceahng almost completely the skeletal plates ; on papular areas, there are 60 or more granules per square millimeter, but on convexities and eleva- tions of disk and base of rays, the granulation is mvich coarser, sometimes only 4-10 per square millimeter. Median radial series of abactinal plates very conspicuous proximally, but becoming less and less distinct distally ; on each side is a much smaller series which, beyond the middle of the arm, becomes irregular and difficult to distinguish ; then comes the series of superomarginals which are quite well marked except near tip of ray ; inferomarginals as large or larger, and generally easily followed to very tip of ray, since they form the ventro- lateral margin of the arm. Papular areas rather small, each with 20-25 papulae, forming eight regular series, one on each side of arm being actinal in position, below the inferomarginal plates ; distally the series are less regular and the actinal do not reach the tip of the arm. Madreporite very large and conspicuous, 10 mm. long by 8 mm. wide. Terminal plate very small and inconspicuous. Actinolateral plates in three regular series proximally, with a few plates indicative of a fourth series ; the third series disappears about one-third of the distance from mouth to arm tip ; the second extends beyond the middle of the arm ; the series adjoining the adambulacrals apparently extends nearly to the arm tip. Adambulacral armature of 2 flattened, smooth spines nearly 2 mm. long, with blunt, rounded tips, placed close together, the aboral the slightly smaller of 1. ^i;ffo9= wrinkled, shrivelled ; in reference to the appearance of dry- specimens. 56 " ENDEAVOUR " SCIENTIFIC RESULTS. the two ; the adambulacral plates being short and over- lapping, the spines are crowded close together in a very regular and unbroken marginal series ; the actinal surface of the plates is covered by the same coarse granulation, as the actinolaterals bear ; at the very tip of the arm, there occur scattered subambulacral spines, Hke those of Ophidiaster, while close to the mouth, 3 or 4 plates in each series, carry short heavy spines of the same nature, 1-2 mm. long, and 1-1.5 mm. thick at tip ; occasionally there are two such spines on a plate. Oral spines similar to the adambulacrals but somewhat heavier and more prismatic. Pedicellariae present in very diverse numbers, of the usual Ophidiastrine type, each about .75 mm. long, the valves nearly or quite .50 mm. wide and with smooth valves, which are often low and wide ; in one specimen they are fairly abundant on the abactinal surface, but in the others only a very few occur, careful search reveaUng one here and there on the papular areas ; I have found none on the actinal surface. Colour (in alcohol), purplish-brown, €ach arm with three broad, indistinct cross-bands of a hghter shade ; lower surface pale grayish ; dried, the specimens become pale, dingy fawn-colour. This remarkable linckiid is not a handsome Starfish but this is, in part certainty, the effect of preservation. The leathery nature of the body-wall, the form of the rays, the Jiuge niadreporite, the absence of superambulacral plates, the presence of pedicellariae and the calcareous plates of the ampullae combine to make it a very unusual member of the family. The tendency to develop subambulacral spines on the oldest (near mouth) and youngest (near arm-tip) adambu- lacral plates like those of Ophidiaster are suggestive of a phylogenetic connection with that genus and perhaps we shall not be far from the truth if we called Pseudophidiaster a deep-water ophidiasterid. The smallest specimen before mehasR=58mm. ; r=8mm. ; R = 7r; in other particulars, it is not noteworthy. Locs. — Great AustraHan Bight, 80-120 fathoms. Thirty-six miles N. 58° E. of Cape Wickham, King Island, Bass Strait. South of Gabo Island, Victoria, 200 fathoms. •Oyster Bay, Tasmania, 60 fathoms. SEA LILIES, STARFISHES, ETC. CLARK. 57 Family ASTERINID^. Genus Asterina, Nardo. ASTERINA ATYPHOIDA,! Sp. nOV. (Plate xvii., fig. 1-2.) Remarkably similar to A. modesta, Verrill, from Panama. On the upper surface the only difference is in the coarser granulation of the plates in the Australian specimens. Actin- aUy the difference is somewhat more marked ; the adambu- lacral plates almost never have more than one subambulacral spine, while in A. modesta two such spines are often, if not usually, present ; the actinal intermediate plates near the margin have only a single spine while in A. modesta there are two and occasionally three. The largest specimen of A. atyphoida is 19 mm. across; R = 10 mm., and r=8.5 mm. Colour (dry), Hght brown or whitish. Four specimens. The resemblance between these South Australian Asterinas and specimens of A. modesta from Panama is very extra- ordinary, whether one regards it as indication of genetic relationship or as a case of parallehsm. In either case, the degree of difference is very slight in view of the fact that several other species of Asterina occur in the vast intervening distance between the homes of the two species. It is obvious- Ij^ impossible, without far more data and specimens than are now available, to determine what the relationship between them really is. Locs. — Fifteen miles north-west of Cape Jervis, South AustraUa, 17 fathoms. This specimen has been selected as the ho lo type. Off Cape Marsden, Kangaroo Island, South AustraHa, 17 fathoms. Asterina leptalacantha,^ sp. nov. (Plate xviii., fig. 3-4.) R = 18mm. ; r=8 mm. ; R=2.25 r. Br=9 mm. ; R= 2 br. Br at middle of ray =7 mm. ; at tip, 3 mm. Disk and rays moderately arched, but general appearance flattened ; margin formed by the conspicuous infero marginals. Abacti- nal plates as usual in Asterina, more or less markedly 1. (iTi;0o9=the Latin modestus-\-eico'i=iovrQ.; in reference to the resemblance to A. modesta, Verrill. 2. X67rT«\fc'os=fine, slender-f (4^01/^0= thorn, spine; in reference to i;he verj' fine spines on the inferomarginal plates. 58 " El«iDEAVOUR " SCIENTIFIC RESULTS. imbricating and with a proximal notch or concavity for the subtending papula ; there are no small supplementary plates ; across the base of the ray one may count 10-12 series of plates but only 8 or 6 distally above the superomarginals ; at base of ray there are 5-7 series of papulae but only three (or two) extend beyond the middle of ray. Except for a few minute spinelets around the anus and around the madreporite, all the abactinal plates, but 10 or so in each interradial angle, are perfectly bare and smooth ; the interradial plates referred to each carry 6-10 exceedingly slender spines, .75 mm. long. Superomarginal plates small and perfectly bare. Infero- marginals about 24 on each side of ray, very conspicuous, each one bearing a tuft of 12-15 (or more) exceedingly slender spines 1-1.5 mm. long. Ac tino lateral plates in two complete series, while a third runs as far as the twentieth (from inter- radius) inferomarginal ; there are 5 or 6 plates in a fourth series, and a few additional plates in the interbrachial arc ; except the three most adoral plates of the first series, which are perfectly bare, each of these plates carries a single, sharp, relatively stout spine nearly a milhmeter long ; in each interradial area except one, back of the oral plates, is a small bit of naked, uncalcified skin. Adambulacrals about 26, all but the two most adoral, corresponding to an infero- marginal with the intervening actinolaterals ; armature consists of a furrow series of 5 (4-6) spines, the middle one longest, adoral and aboral ones shortest, and a single promi- nent subambulacral spine on the surface of the plate. Oral plates rather large, with only 5 marginal spines, the terminal much the largest, and a single big spine on the surface. Madreporite small, triangular, nearer centre of disk than margin. Colour (dry), pale yellow-brown. This very interesting Asterina, while not collected by the " Endeavour," is included in the collection sent me, and certainly deserves description. It belongs in the section of the genus to which Verrill has recently given the name Patiriella, the type of which is A . regularis from New Zealand. But there is little superficial resemblance between A . leptalacantha and A. regularis and if Asterina is to be broken up, one would hardly expect them to fall into the same section. The new AustraHan species cannot be mistaken for any hitherto described, the relatively long arms, the bare abactinal plates, the remarkable inferomarginals and the spinulation of the actinal surface forming a very distinctive group of characters. Loc. — Masthead Island, Queensland, collected by Mr. A. R. McCulloch, December, 1913. SEA LILIES, STARFISHES, ETC. CLARK. 59 Genus Anseropoda, Nardo. Anseropoda rosacea (Lamarck). (Fig. 5-6.) Asterias rosaceits, Lamarck, Anim. sans Vert., ii., 1816, p. 558, par. 3. Anseropoda rosacea, Fisher, Bull. U.S. Fish Comm., 1906, p. 1089. Evidently this is a more widely distributed species than was supposed, since it is now known from Japan, the Indian Fig. 5.x40. Fig. 6.xl5. Ocean, the western coast of Australia, and the eastern coast of Queensland. No doubt it will soon be reported from the East Indies. The specimens in the " Endeavour " collection have R = 85- 115 mm. ; two ha ve 1 5 and two have 1 6 rays. They 60 " ENDEAVOUR " SCIENTIFIC RESULTS. are pale, dirty grayish, or brown, and show no trace of any of the original colours. Four specimens. Logs. — Off Fraser Island, Queensland, 32 fathoms. Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. Family ECHINASTERID^. Genus Henricia, Gray. Henricia hyadesi {Perrier). Cribrella hyadesi, Perrier, Miss. Sci. Cap-Horn, vi., Zoologie, Echinod., 1891, p. KIOO. Henricia hyadesi, Fisher, Bull. U.S. Nat. Mus., Ixxvi., 1911, p. 293. Although these specimens show considerable diversity in proportions and in spinulation, in view of the well-known inconstancy of characters in the genus, I do not feel justified in recognising more than one species, and that one I, for the present, identify with Perrier 's species from southern South America. I cannot go as far as Ludwig and consider all the Antarctic and sub-Antarctic Henricias as a single species, for I doubt if Studer's H. pagenstecheri is identical with Perrier's H. hyadesi, and I am quite prepared to learn that the Austrahan and South American Henricias are distinct. But until more material is available, it is futile to discuss the matter. It is worthj'^ of note that all the Australian specimens are from water relatively deep for the genus. The largest specimen has R = 80 mm. ; in another R = 75 mm. ; br==18 mm. ; in a third, R = 67 mm. ; br=9 mm. Usually the arms taper to a slender tip, but in one specimen they are rather uniformly stout. One specimen from the Great Austrahan Bight has 6 arms but does not seem to differ otherwise from the rest. All the specimens have lost their original colours, and are now some shade of yellowish-brown. Eleven specimens. Logs. — East of Maria Island, Tasmania, 78 fathoms. Off Babel Island, Bass Strait, 50-60 fathoms. South of Gabo Island, Victoria, 200 fathoms. Great Austrahan Bight, 80-150 fathoms. SEA LILIES, STARFISHES, ETC. CLAKK. 61 Genus Echestaster, Muller and Troschel. ECHIXASTER ACANTHODES,^ Sp. nOV. (Plate xix., fig. 1-2; Fig. 7.) R=70 mm. ; r=12 mm. ; R = 6 r. Br=14 mm. ; R=5 br. Br at middle of ray=9 mm. ; at tip, 3.5 mm. Disk moderate, slightly arched ; arms terete, tapering very uni- formly to the narrow but blunt tip. Abactinal skeleton coarse and rather heavy ; each plate bears a large sharp spine, .5-1 mm. high. Papulge confined to abactinal surface ; the areas in six irregular longitudinal series, each area ^\ith few (2-6) papulae. Adambulacral armature consisting of 3 spines in a vertical series, the smallest high up in ambulacral furrow, the second, Httle larger, just below, and the third, much larger, on the furrow-margin ; except on the 12-16 plates nearest the mouth, there occurs on the actinal surface of every other plate, a conspicuous subambulacral spine, equal to, or larger than, that on the furrow-margin ; back of this Fig. 7.x3. spine, on the actinolateral or more Hkely the inferomarginal plate, is another still larger spine, about 2 mm. high. Oral plates with no surface spines and marginal spines united by membrane, which reaches more or less nearly to their tips, and extends outwards along the rays uniting more or less evidently the furrow-margin spines of the first few adam- bulacral plates. Beyond the disk, each spine is more or less clothed in membrane. Madreporite very small, about half- way between disk-centre and margin. Colour (dry), light brown. Three specimens. The general appearance of this species is more Hke that of the West Indian Echinasters than hke that of its East Indian congeners. The conspicuous spines and the total lack of 1- a/cni'^wCH9=full of thorns ; in reference to the spiny surface. 62 " ENDEAVOUR ' SCIENTIFIC RESULTS. actinal papulae make it an easy species to recognise. The smallest of the three specimens has R=30 mm., and the more or less isolated papulae are in about 4 irregular series. There is only 1 prominent actinal spine for each 3 adambulacral plates, until near tip of arm. The membrane between the adambulacral plates has distinct non-calcified actinal areas about .35 mm. wide by .25 mm. long. Locs. — Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. Thirteen miles north-east of North Reef, Capricorn Group, off Point Curtis, Queensland, 70-74 fathoms. EcHiN ASTER ERiDANELLA, MulUr and Troschel. (Plate XX., fig. 1-2 ; Plate xxi., fig. 1-2.) Echinaster eridanella, Miiller and Troschel, Sys. der Ast., 1842, p. 24. There is a bare possibility that the little specimen with 5 arms from twenty-five miles south-east of Double Island Point, Queensland, R = 25 mm., is not this species, but except for the number of arms, I see no occasion for questioning it. The species is abundant on the reefs of northern Queensland, and it is quite possible that it reaches southward to twenty- five miles south-east of Double Island Point. Of the speci- mens figured, those from Bo wen, Queensland, and from Murray Island, Torres Strait, are typical of the species ; the 4-rayed specimen from Mauritius is exceptional and is, moreover, poorly preserved. Three specimens. Locs. — Bowen, Queensland. A specimen with seven rays. Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. A small specimen with five rays. Echinaster glomeratus,i sp. nov. (Plates xxii.-xxiii. ; Pig. 8.) R = 135 mm. ; r=25 mm. ; R=5.5 r. Br=25 mm. ; R = 5.5 br. Br at middle of ray =18 mm. ; at tip, 5 mm. Disk large, moderately elevated as are the bases of the broad, steadily tapering arms. Madreporite very small ; in holo- type, only about 7 mm. from centre of disk but nearer the margin in other specimens. Abactinal skeleton very coarse 1. Glomeratus= grouped ; in reference to the arrangement of the abactinal spines. SEA LILIES, STARFISHES, ETC. CLARK. 63 and heavy, forming a wide-meshed network with rather large papular areas ; wherever two or more plates meet is an elevation bearing a cluster of 2-12 stout, sharp spines, .5-1 mm. high ; the papular areas contain 12-15 or more papulae ; neither spine-clusters nor papulae show any serial arrange- ment. Actinal surface somewhat flattened ; adambulacral plates very numerous ; their armature shows some diversity ; well up in the furrow is a sharp, slightly curved spine, below which is a second larger and somewhat blunter and straighter one ; below the latter is, generally, a similar but sHghtly larger spine, about 3 mm. long, which stands on the furrow Fig. 8.x5. margin and just outside of it, on the actinal surface of the plate is a similar but much shorter spine ; diversity comes from the fact that there may be two or even three spines, instead of one on the furrow margin and they may stand side by side or in an oblique row, extending aborally from the margin ; the marginal and furrow-spines are more or less encased in membrane. The actinolateral plates, outside the adambu- lacrals, seem to correspond to them in number, and each one bears a series (transverse to long axis of ray) of 2-3 sharp spines. Papulse are abundant on the actinal surface, but do not occur between the actinolaterals and adambulacrals, nor on the actinal interradial areas. Oral plates not pecuHar, but the lateral spines are not concealed by the thin membrane which unites them. Colour (in alcohol), bright yellow-brown; (dry) similar, but duller. Strictly speaking this species belongs in Henricia rather than in Echinaster since the spinelets are in groups and not isolated, but the general appearance is so unlike Henricia that it seems absurd to put it in that genus. It is quite unlike any previousty described Echinasterid and will be easily recognised at once. The smallest of the six specimens at hand has R = 75 mm. ; it is not, however, notably different 64 ENDEAVOUR SCIENTIFIC RESULTS. from the larger specimens, save that the actinal skin is thicker and unites and conceals the spines along the furrow more fully. Loc. — Off Cape Marsden, Kangaroo Island, South Australia,, 17 fathoms. ECHINASTER SUPERBUS,^ Sf. nov. (Plates xxiv.-xxv. ; Fig. 9-10.) R = 110 mm. ; r=15 mm. ; R = 7 r. Br=20 mm. ; R = 5.5 br. Br at middle of ray =14 mm. ; at tij), 4 mm. Disk small, not elevated ; rays nearly terete, tapering uniformly to tip, somewhat flattened actinally. Abactinal skeleton pretty well concealed in the thick skin, but forming a rather fine mesh-work, with relatively small papular areas, which show no serial arrangement. Papul?e large and numerous. Abactinal spines numerous, isolated, diverse in form, 1-2 mm. high, .5-1.75 mm. thick; few have a simple, pointed or blunt, tip ; usually it is clavate to a greater or less degree, chisel- shaped with one or more notches, or with 5-6 low, rounded summits. Madreporite moderately large, very distinct and quite near centre of disk. Adambulacral plates numerous. Fig. 10. x8. Fig. 9.x4. about twice as wide as long, very well defined, separated from each other by distinct membranous spaces half as long as the plates ; each plate carries, high up in the furrow, a small, curved pointed spine, immediately below which is a 1. Superbtis= very fine ; in reference to its notable form and size. SEA LILIES, STARFISHES, ETC. — CLARK. 65 longer, straighter, stouter one ; close to this, but on the furrow-margin is a large flattened spine, 2-2.5 mm. long and about as wide as the length of the plate ; its outer side is more or less deeply furrowed (at least near tip) and its summit is a chisel-shaped or clavate edge ; approximately every other spine is narrower, smoother and more nearly pointed, but there is considerable diversity. On the actinal surface of a very few plates, on the basal part of the ray, occurs a single, low pointed tubercle or spine. Outside the series of adambu- lacral plates occurs a very conspicuous series of spines, the largest on the Starfish ; there are 2 of these to every 5 adambu- lacral plates ; they are 2 mm. high or more, and their con- spicuous chisel-shaped or clavate tips are frequently more than 2 mm. wide ; outside this series, and virtually forming the boundary of the actinal surface, is a series of similar but smaller and more numerous spines, large ones often alternating with small ones as in the adambulacral series. Papulae occur between these two actinal series of spines. Oral plates and spines not pecuHar or notable. Colour (dry), deep purplish- brown, tips of the spines Hghter but apparently only by abrasion. This is another remarkable Echinaster, unhke any previously known, and making the fourth notable species known only from Austraha, besides the long-knowTi East Indian species, E. eridanella and African E. vestitus. It is interesting that with one exception, each of these six species occurs on a different part of the coast ; the present species E. superbus comes from the far north-west of the continent ; two hundred miles further south, at Port Hedland, E. vestitus and the very different E. arcystatus have been taken ; over two thousand miles of coast separate these three species from E. glomeratus of South Austraha ; another gap of some one thousand five hundred miles separates that sharply distinct species from E. acanthodes of southern Queensland ; and just overlapping the latter's area is the southern hmit of E. eridanella, the common species of northern Queensland. Of course further collections and studies may show that these geographical lines of division do not really exist in nature but their apparent occurrence seems worth noting. Unhke the various forms of Echinaster in the West Indian region to which different names have been given, these AustraUan species are very obviously- different from each other and it is difficult to conceive of their intergrading. Loc. — Broome, Western Austraha. 66 " ENDEAVOUR " SCIENTIFIC RESULTS. Genus Plbctaster, Sladen. Plectaster decanus {Mailer and Troschel). (Plate xxvi., fig. 1-2.) Echinaster decanus, Miiller and Troschel, Arch. f. Naturg., ix.i., 1843, p. 114. Plectaster decanus, Sladen, Chall. Rep., Zool., xxx., 1889, p. 535. This remarkable species never having been figured, it seems desirable to figure it here, and the Port Jackson specimen is shown on PI. xxvi., fig. 1-2. Init,R = 115mm., andr=27mm. ; R=4 r. The specimens from Shoalhaven Bight, New South Wales, are smaller and not so well preserved, but they appear to have been placed in formalin and then dried, and hence are more nearly the natural colour. In them the papula areas are bright reddish-purple, while the skeleton and spinelets are Ught yellowish in sharp contrast ; on the actinal surface, there is a considerable suffusion of reddish-purple, especially interradially, near the mouth. The Port Jackson specimen is yellowish-brown, the papular areas darker than the skeleton. Three specimens. Locs. — Shoalhaven Bight, New South Wales, 15-45 fathoms. Off George Head, Port Jackson, New South Wales. Family SOLASTERID^. Genus Crossaster, Miiller and Troschel. Crossaster multispinus,! sp. nov. (Plate xviii., fig. 5-6.) Rays 11. R=40 mm.; r=20 mm.; R=2r. Br=10 mm. ; R=4 Br. Abactinal skeleton an open mesh-Avork of narrow, rather dehcate ossicles, leaving large papular areas, in which, however, the papulae are few, large, isolated. Pseudopaxillae rather crowded on disk, but more widely spaced on arms ; each bears a tuft of 8-15 long, slender, acicular spines, exceeding the stalk in length. Supero- marginals insignificant. Inferomarginals few, widely spaced, conspicuous, about 16 on each side of an arm ; they are greatly compressed at summit into a narrow vertical ridge, carry 25 (more or fewer) sharp, spinelets, most of which are shorter than those on the abactinal pseudopaxiUse. Actinal 1. Multispinus=-with. many spines ; in reference to the armature of the adambulacral and oral plates. SEA LILIES, STAKFISHES, ETC. CLABK. 67 intermediate plates, few and scattered, with 6-12 spine lets ; actinolateral plates very few, small and scattered, with very few spinelets. Adambulacral jjlates as wide as long, or wider ; the furrow margin bears 8 or 7 (proximally), 6 or 5 (distally), slender, not sharp, membrane-united spines, the middle ones longest, 1.5-2 mm. long ; actinal surface of plate with a transverse, slightly curved series of 7-9 similar spines, 2-3 mm. long. Oral plates with a marginal series of 14 similar spines, the proximal longest and largest, 3-3.5 mm, long ; actinal surface of each plate with a longitudinal series of 8-10 similar spines, near inner margin. Madreporite small and inconspicuous, half way between centre of disk and margin. Colour (dry), dull brown ; other specimens are hght yellowish -brown . Although all four specimens have 11 rays each, the three from Tasmania are obviously different from the holotype, in their hghter colour, more conspicuous skeleton, more slender rays, and thinner and less noticeable membrane covering and uniting the spines, both abactinally and actinally. In view of the weU-known diversity in all such particulars shown by the common, northern species of Crossaster (C. papposus), I think there is httle reason to consider these differences of any special significance, especially since the Tasmanian specimens are poorly preserved. The large number, and the slenderness of the adambulacral and oral spines readily distinguish the present species from any of those previously known. Logs. — Between Gabo Island, Victoria, and Disaster Bay, New South Wales, 50-100 fathoms. This specimen has been selected as the holotype. South-east of Bruni Island, Tasmania, 150-230 fathoms. Genus Ptbraster, Muller and Troschel. Pteraster tetracanthus,! sp. nov. (Plate xviii., fig. 1-2.) R=30mm. ; r=24mm. ; R = 1.25r. General form almost perfectly pentagonal, moderately thick. Abactinal surface in such poor condition it is not possible to determine exactly the number of paxillar spines ; paxillse somewhat numerous and paxillar spines long, apparently about 6. Spiracles 1. TeT/3a/cai/6os=having four spines; in reference to the armature of the adambulacral plates. 68 " ENDEAVOUR " SCIENTIFIC RESULTS. small and seemingly widely scattered. Actinolateral spines only 5 mm. long, hence actinolateral membranes cover a relatively small part of actinal surface, and actinal inter- mediate areas are large and densely covered with groups of slender, membrane -united spinelets. Adambulacral spines 4, the innermost very small, the outermost about 4 mm. long, united by a thin membrane, which retreats little (say 25%) between the spines, and which connects with the actinolateral spine. Actinolateral membrane reaches practically to tips of spines, with httle retreat between. Oral plates with promi- nent central keel and very stout suboral spines (one on each plate), 4 mm. long, sharply triangular near tip, but truncate and more or less succulate ; oral spines 4, of which the first (proximal) is much larger than the next, about 3 mm. long, truncate and flat ; no membrane connects these spines. Colour (dry), dull grayish-brown. This specimen is the first representative of the genus to be taken anywhere near Australia. Judging by the keys to Pteraster which have been published one would expect either P. dance, Verrill, or P. stellifer, Sladen, to be near allies, but in reahty the present species is quite unhke either of these. The large actinal interradial areas, the narrow actinolateral mem- brane and the characters of the oral armature combine to give P. tetracanthus a very distinctive facies. Loc. — South-east of Cape Everard to south of Gabo Island, Victoria, 90-150 fathoms. Family ZOROASTERID^. Genus Zoroaster, Wyville-Thomson. Zoroaster macracantha,i gp ^q^; (Plate xxviii., fig. 1-2.) Very similar to Z. spinulosus, Fisher, from the Hawaiian Islands, but differing in a few details. The plates of the medioradial series are the largest on the arm and each succeed- ing series is made up of smaller plates ; the superomarginals are thus larger than the inferomarginals. Two series of actinolateral plates are confined to the actinal surface at the base of the arm. The plates of the medioradial series are provided with large central tubercles (bosses) on which are articulated stout sharp spines 5 mm. or more in length ; on the basal part of the arm such a spine occurs on nearly every 1. ^aKp6<i=^ong-\-uKni'0a=a, thorn, or spine ; in reference to the long abactinai spines. SEA LILIES, STARFISHES, ETC. — CLARK. 69 medioradial plate, but distally only on every other plate ; there is never more than one to a plate ; nearly all these spines are missing in the specimens at hand, but the tubercles are con- spicuous and enough spines are left to show their size. The superomarginals bear similar but smaller and more slender spines. Pedicellarise both large and small are present every- where, very large ones, 1.5-2 mm. long, occurring on the medioradial plates. Adambulacral armature essentially as in Z. spinulosus but pedicellarise appear to be much more abundant. Colour (in alcohol), yellowish-brown. Three specimens. It is not without some hesitation that I separate this Zoroaster from the Hawaiian species, but until something is known of the fauna of the vast area between southern Aus- traHa and the Hawaiian Islands, it seems best not to assume the identity of species in the two regions. Apparently the AustraHan Zoroaster is conspicuously more spiny than its Hawaiian congener, and since the medioradial plates and spines seem to furnish a tangible character, I have determined to give it a different name. The largest specimen has R = 160 mm. ; r=14 mm. ; R = 11.5 r. The primary disk-plates have prominent bosses like those on the medioradial plates and apparently, in life, bore similar conspicuous spines. Zoc— Great Australian Bight, 129° 28' E., 250-450 fathoms. Family PEDICELLASTERID^. Genus Pedicellaster, Sars. Pedicellaster reticulatus,! sf. nov. (Plate xxvii., fig. 3-4.) Rays 6, unequal, three larger and three smaller ; the middle one of the smaller trio much smaller than the others, being about 22 mm. long, while they are over 30 mm. R=37 mm. ; r=5 or 6 mm. ; R = 6 or 7 r. Br=5 mm. ; R = 7 br. Br at middle of ray, 3.5 mm. ; at tip, 1.5 mm. Disk small ; arms relatively broad for a Pedicellaster. Abactinal skeleton made up of slender rods and plates forming a large-meshed net- work ; on each arm there are about 4 longitudinal series of these meshes, with 20-25 meshes in each series ; at many 1. Beticulatus=-with meshes like a net; in reference to the wide- spaced skeleton. 70 " ENDEAVOUR " SCIENTIFIC RESULTS. points, there are only 3 series, a median, and one lateral on each side. Each mesh contains one or more papulae, but in the dry specimen, it is difficult to determine their presence. Each plate of the skeleton bears a single, slender, pointed spine, 1-1.5 mm. long ; also numerous small pedicellarise, which may be either on the plate itself or on the spine or on both ; often the spines, particularly along the actinal margin, have large wreaths or half -wreaths of pedicellariae. Adambulacral plates, 15 to 5 marginals, at base of arm ; armature of 2 spines, in a transverse series ; they are about 1.5 mm. long and subequal or the outer a trifle the longer ; the outer is also a trifle more adoral. There are a very few, scattered, small pedicellarise just within the margin of the ambulacral furrow, which is notably broad and has the tube-feet in two distinct and well- separated series. Oral plates not large, but projecting dis- tally so as to slope sharply into the mouth ; each one bears 3 smooth spines ; one on the actinal surface, which is similar to, about as long as, but stouter than the adambulacral spines; a, second is still stouter and blunter and stands at the proximal end of the plate ; while the third is much shorter than either of the others, is situated at the proximo-lateral corner of the plate and extends straight out across the ambulacral furrow ; it carries a few small pedicellarise. Madreporite large (2 mm. across), conspicuous, near margin of disk. Colour (dry) ; skeleton and spines, white ; membranes and feet, brown. This species is quite unUke the Atlantic members of the genus but bears considerable resemblance to P. improvisus, Ludwig, from the Galapagos Islands. It difters noticeably in the armature of the oral plates, however, as well as in some other details (number of papulae, length of spines, etc.). The inequality of the arms suggests that schizogony occurs, but it is possible that the condition of the present specimen is accidental. Loc. — East of Maria Island, Tasmania, 78 fathoms. Family ASTERIIDiE. Genus Allostichaster, Verrill. Allostichaster polyplax {Muller and Troschel). Asteracanthion polyplax, Miiller and Troschel, Arch. f. Naturg., x. i., 1844, p. 178. Allostichaster polyplax, Verrill, Harriman Alaska Exped., Starfishes, 1914, p. 363. It is a httle odd that this well-known Austrahan Starfish should have been taken only at Oyster Bay, Tasmania ; not SEA LILIES, STARFISHES, ETC. CLARK. 71 all are labelled thirty fathoms and it is possible they were not all taken at that depth, nor at the one spot. One has only 4 arms, having divided very recently ; a second has 6 arms, with indication of a seventh ; a third has 9 arms ; the re- mainder have eight. None are notably large or small. Seven specimens. Loc. — Oyster Bay, Tasmania, 30 fathoms. Genus Cosmasterias, Sladen. COSMASTERIAS DYSCRITA,! Sp. nOV. (Plate xxix., fig. 1-2.) Rays 5. Il==67mm. ; r=ll mm. ; R=6 r. Br=15mm. R=4.5 br. Br at middle of ray =12 mm. ; at 20 mm. from disk-centre, 17 mm. ; at tip, 2.5 mm. Disk small, somewhat arched ; rays shghtly contracted at base, then increasing quickly to maximum breadth, and then tapering steadily to the narrow tip. Abactinal skeleton thick, moderately heavy, with small and distinct papular areas ; those on disk con- tain 3-7 papulae, while the larger ones along sides of ray have 6-8. Disk plates without definite arrangement, bearing 2-10 low capitate spinelets, about .5 mm. high and .25 mm. thick at tip ; there are also a very few big jelipedal (as Verrill calls them) pedicellarise, nearly a millimeter long scattered here and there, besides numerous minute pedicellarise of the ordinary type ; neither spines nor pedicellariae are in the least crowded at any point. Madreporite not very large but quite conspicuous, being situated at the centre of a large, elevated plate, half way between disk-centre and margin, surrounded by a circle of 12-15 spinelets. Abactinal plates of rays similar to those of disk but arranged in 13 series at base of arms, a medioradial series with a dorsolateral series on each side, then the superomarginals, the inferomarginals, and 3 series of actinolaterals ; of the latter, the series next the adambulacrals extends less than one-third of the ray, the second reaches to about the middle of ray or a httle beyond, and the third goes to within 15-20 mm. of the tip. All the plates above the actinolaterals carry capitate spinelets and pedicellarise like the disk-plates, usually only 2-3 spinelets on a plate ; the actinolaterals each carry one, or often two spines, the largest on the Starfish ; near the mouth they are 2 mm. long, blunt, somewhat flattened and nearly a milhmeter wide ; on the 1- ^uo-/f^(To^=liard to determine ; in reference to the uncertain relation- ships of the species. 72 " ENDEAVOUR " SCIENTIFIC RESULTS. outer series of actinolaterals these spines are somewhat smaller and intergrade with similar spinelets on the inferomarginals. Papulge and small pedicellariae occur between the second and third series of actinolaterals. Adambulacral plates numerous, about 13 to 5 adjoining actinolaterals at base of arm ; armature diplacanthid, the spines 1.5-2 mm. long, the inner ones pointed, the outer blunt and slightly flattened ; between the two series occur numerous very large pedicellarise, those near the mouth being distinctly felipedal ; pedicellarise both large and small occur also within the ambulacral furrow. Oral plates, each with 4 spines, an inner proximal at the tip of the jaw, 2 mm. long, moderately stout and pointed ; an outer proximal at the corner of the jaw, much smaller and extending outwards across the furrow ; and two actinal spines behind the inner proximal, similar to it, but longer and stouter and forming with it a longitudinal series of 3 ; there are also numerous small pedicellarise and one or more of the big felipedal ones. Tube-feet in 4 distinct, not greatly crowded series. Colour (dry), brownish-yellow. This Starfish is undoubtedly a Stichaster in the old broad- sense, but Verrill has well shown, what others have suspected, that Stichaster was a heterogeneous and unnatural assemblage. The present species shows some affinity to Asteracanthion luridum, Philippi {=Asterias sulci f era, Perrier) of southern South America, for which Sladen made the subgenus Cosmas- terias. Verrill has raised the group to generic rank, in- cluding certain species which Sladen considered Stichasters- and I think his action is correct and wise. The genus is characteristic of southern South America and that vicinity, so that the discovery of a species which seems to belong to it, in deep water south-east of Australia is of considerable interest. The Australian species is easily recognised by its general facies, short, capitate, scattered spines and adambulacral and. oral armature. Loc. — South of Gabo Island, Victoria, 200 fathoms. Genus Coscinasterias, Verrill. COSCINASTERIAS CALAMARIA {Gray). Asterias calamaria, Gray, Ann. Mag. Nat. Hist., vi., 1840, p. 179. Coscinasterias calamaria, Perrier, Exped. Sci. du " Travail- leur" et du "Talisman," Echinod., 1894, p. 106. This individual has 11 arms, of which 3 are very small; for the others, R=60 mm. There are no large pedicellarise any-^ SEA LILIES, STARFISHES, ETC. — CLARK. 7a where so far as I can discover, in which particular this speci- men agrees with individuals from Westernport, Victoria. Loc. — Sanders Bank, Kangaroo Island, South Australia, 28 fathoms. CosciNASTERiAS DUBiA, H: L. Clark. (Plate XXX., flg. 1-2; Fig. 11.) Coscinasterias dubia, H. L. Clark, Mem. Austr. Mus., iv., 11, 1909, p. 532. This is a much better series of specimens than those col- lected by the " Thetis " on which the species was based. They are as a rule very well preserved and range in size from R=22 mm. to R = 138 mm. The pedicellariae are much more numerous, particularly along the ambulacral furrow, than in the " Thetis " specimens, and the membrane connecting the marginal spines is not distinguishable ; it is probable that its conspicuousness in the " Thetis " specimens was due to their poor preservation. The second series of actinolateral spines is often provided with wreaths of pedicellariae. As growth changes, one notes that in the smallest specimen while there are 5 fairly well-marked series of plates on the abactinal surface, above the inferomarginals, half or more of the supero- marginals carry no spines ; there is only one series of actino- lateral spines and this is confined to the basal haK of the arm ; the inferomarginal spines are relatively long and narrow, not at all widened at the tip ; large pedicellarise are relatively few and are virtually wanting along the ambulacral furrow. Fig. ll.xlS. In the largest specimen, on the other hand, one notes that while all the superomarginals carry conspicuous spines, there 74 " ENDEAVOUR " SCIENTIFIC RESULTS. are still only 5 series of abactinal spines, but the adradial series is very irregular, so that the broad space between the medioradial series and the superomarginals, bears relatively few and considerably scattered spines ; there are two series of actinolateral plates and spines which extend more than three-fourths of the arm-length, and the plates of the inner series often bear 2 spines ; the inferomarginal spines are very conspicuous, 5-7 mm. long and 2 mm. wide at tip, thick, heavy and deeply grooved on the actinal side at tip ; the characteristic large, slender pedicellarise are exceedingly abundant, especially along the ambulacral furrows and about the mouth. The small specimens (dry) are brownish-yellow, but the largest (dry) is dull purplish-gray abactinally with a decided red tinge, that suggests the possible natural colour. Thirteen specimens. Locs. — Oyster Bay, Tasmania, 20-40 fathoms. East of Babel Island, Bass Strait, 65-70 fathoms. Shoalhaven Bight, New South Wales, 15-45 fathoms. South of Gabo Island, Victoria, 200 fathoms. Southern coast of South Australia. COSCINASTERIAS GEMMIFERA {Pemer). (Plates xxxi.-xxxii.) AsteracantJiion gemmifer, Perrier, Ann. Sci. Nat., Zool., (5), xii., 1869, p. 237. Coscinasterias gemmifera, Fisher, Ann. Mag. Nat. Hist., (7), xvii., 1906, p. 574. This species was described from a specimen in Paris, re- ported to have come from ChiH. Sladen, in the " Challenger " Report, listed a specimen, without comment, from the reefs at Kandavu, Fiji ! Now comes this third specimen from Tasmania ! So far as I can judge from Perrier's description and Sladen's " key," the fine, large ll-rayed individual, figured herewith, represents C. gemmifera but I do not beHeve the differences between Perrier's species and Gray's C. cala- muria are worth much. If C. gemmifera really occurs in Chili and Fiji and C. Calamaria really is found at Mauritius (as reported), we have a very notable case of distribution, which deserves critical study. Loc. — Oyster Bay, Tasmania, 20-40 fathoms. SEA LTLIES, STARFISHES, ETC. CLARK. 75 Family BRISINGIDiE. Genus Odinia, Perrier. Odinia australis,! sp. nov. (Plate xxvii., fig. 1-2.) Rays 14. R = 100 mm. ; r=8 mm. ; R = 12.5 r. Br= "2.5 mm. Br where arm is widest (about 13 mm. from centre of disk) 5 mm. ; at middle of arm, about 2 mm. Disk dis- tinctly elevated above arm-bases, thickly plated, the plates without any definite arrangement but bearing minute pointed spinelets and countless pedicellarige ; papulae very numerous, but usually, if not always single. Bases of arms, connected by marginal plates for only 2-3 mm. beyond the disk. Arms narrow at base, then suddenly swollen and often about 10 mm., gradually but quickly returning to the same width as at base ; as far as they are swollen, they are covered by plates, spinelets, papulae and pedicellariae like the disk, but beyond the sweUing a thin skin crowded with minute pedicellariae is their only covering. Along each side of arm is a series of vertical plate- groups (apparently not more than 3 plates to a group distally), occurring at intervals of 2-3 mm. or about 3 adambulacral plates ; on the swollen basal portion of arm these series extend to medioradial line and the two of opposite sides may unite to form a very narrow band across the arm, but these bands are often incomplete, much bent or otherwise irregular ; they are faintly indicated distal to the swollen region but only for a short distance. The first 2 or 3 of these bands carry no spines, only acute, conical spinelets dorsally hke those on the neighbouring abactinal plates, but on the third or fourth band, the lowest plate (adjoining the adambulacral) carries a spine about a milhmeter long ; on succeeding bands this spine increases in length and at the middle of the arm and beyond is 4 mm. long ; on the fifth or sixth band a second spine, just above the first appears and it too increases in length but never quite equals the first ; on the seventh or eighth vertical group (corresponding to the " bands ") a third spine appears but this remains distinctly shorter than the other two ; in life, these spines are apparently covered with a loose membrane, which is densely clothed with, minute pedicellarise, but after capture, probably owing to the rough journey in the trawl, much of this membrane is commonly wanting. Adambu- lacral plates rather longer than broad ; at base of arm, their armature consists of two spines, a very slender, acicular spine on the aboral end of the plate, \^ath 2-3 minute pedicellarise 1. AuMralis=so\itheTn ; in reference to its being the first species of the genus to be taken in the Southern Hemisphere. 76 " ENDEAVOUR " SCIENTIFIC RESULTS. near the tip, extending inwards horizontally directly across the ambulacral furrow, but not quite as long as the width of the furrow, and a stout, flattened spine, 2-3 mm. long, with widened chisel-like tip, situated on the margin or actinal surface of the plate ; this spine is more or less clothed on its outer side with a membrane densely covered with minute pedicellarise. Beyond the middle of the arm the acicular furrow spine disappears, and the big marginal spine becomes smaller and smaller and steadily more acicular until at tip of arm, it is very minute or perhaps wanting. Oral plates each with 3 short (scarcely a millimeter long), blunt spines, one on the inner corner, one (the largest) on the outer comer, and the smallest at the aboral, radial corner ; all carry a few minute pedicellarise. Madreporites ap- parently 4, very small and with few, short, wide furrows ; they are more or less elevated and near margin of disk, but are difficult to make out in an undried specimen. Colour (alco- hoHc or dry), dingy yellowish- white. Two specimens. This is a notable species, one of the "Endeavour's" most interesting captures, since it is the first member of the genus to be taken in the Southern Hemisphere and except the recent- ly discovered Hawaiian species, the only species not from the Atlantic. The genus must henceforth be recognised as cos- mopohtan, for the present form can hardly be considered generically different from the others although its specific characters are well-marked. The occurrence of several madreporites is an unusual feature and the adambulacral furrow spine is also characteristic. Finally it is worthy of remark that no other species of Odinia has been taken in water as shallow as 200 fathoms. Loc. — South-east of Cape Everard, Victoria, 200 fathoms. OPHIUROIDEA. The Ophiurans collected by the " Endeavour " are not an extraordinary lot in any particular, but they give added weight to the view that the AustraHan Ophiuran fauna is relatively poor. Up to the present about one hundred and fifteen species of Ophiurans have been recorded from Aus- tralia in three hundred fathoms or less, but fully half of these are known only from the far northern and north-western coasts of the continent, and are East Indian rather than Australian. Whitelegge lists only thirty -four Ophiurans from Port Jackson and neighbourhood, and six of these are from SEA LILIES, STARFISHES, ETC. CLARK. 77 ■water four hundred fathoms or more in depth, while half a dozen others are of doubtful authenticity. The " Endeavour" collection contains three hundred specimens of twenty-six species, of which five are new to science and three have not previously been recorded from Austraha. There are several other species, previously known from the Torres Strait region which are now recorded for the first time from southern Queensland or New South Wales. Undoubtedly the best feature of the collection is the notable series of Astroconvs australis, but all the Trichasterids and Gorgonocephahds are of great interest. The complete absence of ten or a dozen well-known Port Jackson brittle-stars is rather remarkable. Of the one hundred and twenty-three Ophiurans now known from Austraha, at least sixty are not known from elsewhere. Family OPHIOMYXID^. Genus Ophiomyxa, MulUr and Troschel. Ophiomyxa AUSTRALIS, Lutken. Ophiomyxa australis, Liitken, Add. ad. Hist. Oph., pt. iii., 1869, p. 45. These specimens range in disk-diameter from 15 to 25 mm. Several have the arms more or less distinctly banded, indicating a diversity of colour in life similar to that shown by the West Indian species, 0. flacirda. Fourteen specimens. Locs. — Twenty miles east-south-east of Tasman Head, Tasmania, 80-85 fathoms. East of Maria Island, Tasmania, 78 fathoms. North-east of Cape Pillar, Tasmania, 80 fathoms. Forty miles west of Kingston, South Austraha, 30 fathoms. Genus Ophiobyrsa, Lyman. Ophiobyrsa rudis, Lymun. Ophiobyrsa rudis, Lyman, Bull. Mus. Comp. Zool. Harvard, v., 1878, p. 131. This is a good series of this relatively httle-known species, but they exhibit little diversity. The disk-diameter ranges from 9 to 27 mm. The colour is yellow-brown of diverse shades, the arms often with narrow darker bands. Seven specimens. Locs. — East of Maria Island, Tasmania, 78 fathoms. 78 " ENDEAVOUR " SCIENTIFIC RESULTS. East of Babel Island, Bass Strait, 65-70 fathoms. OfiF Gabo Island Victoria, 200 fathoms. Between Gabo Island, Victoria, and Disaster Bay, New South Wales, 50-100 fathoms. Off Twofold Bay, New South Wales, 30 fathoms. Genus Asteronyx, Muller and Troschel. AsTERONYX LOVENi, Muller and Troschel. Asteronyx loveni, Miiller and Troschel, Svs. d. Ast., 1842, p. 119. The occurrence of this cosmopolitan species south-east of Australia is notable since its known range southward is thus extended over two thousand miles. The specimens are all small, 10-20 mm. disk-diameter, and, so far as I can see, do not differ from North Pacific specimens of the same size. Six specimens. Locs. — East of Babel Island, Bass Strait, 65-70 fathoms. Off Gabo Island, Victoria, 200 fathoms. South-east of Cape Everard, Victoria, 200 fathoms. Genus Euryale, Lamarck. EuRYALE ASPERA, Lamarck. Euryale aspera, Lamarck, Anim. sans Vert., ii., 1816, p. 538. This species was taken by the " Challenger " in Torres Strait, but the present collection extends the known range southward over a thousand miles. All the specimens are dried adults, the disk-diameter only ranging from 25 to 40 mm. There is considerable diversity in colour, however, for the specimen from Wide Bay is orange above, paling to yellow beneath ; one from off Double Island Point is deep orange- brown ; another is brown and another pale brown ; the specimen from Port Curtis is very pale brown ; all the other specimens are dull purple. Fourteen specimens. Locs. — Twenty-five miles south-east of Double Island Point, Queensland, 33 fathoms. Wide Bay, Queensland, 18-23 fathoms. Port Curtis, Queensland, 14-16 fathoms. SEA LILIES, STAEFISHES, ETC. CLAHK. 79 Genus Ophioceeas, Lyman. OpHIOCREAS PHANERUIVI,! S'p. UOV. (Plate xxxiii., fig. 1-2.) Disk 18 mm. in diameter ; arms 380 mm. long ; arms about 20 times the disk-diameter. Disk and arms encased in a smooth, thick skin, which even when dry, shows no sign of calcareous plates or granules and more or less conceals the radial shields and arm-joints. In young specimens the skin is, of course, much thinner and when dry conceals very little. Radial shields narrow and convex, not closely approximate in pairs ; in young specimens, they are flat and are in rather closely joined pairs. Genital shts small (about 3 mm. long), deeply sunken. Tentacle-pores quite smaU ; first 2 pairs on arm, naked ; succeeding pores, all guarded by a pair of short, flattened, rough spines ; where largest (at about middle of arm) the inner spine is 2 mm. long, .80 mm. wide, pointed and prickly for more than half its length ; the outer is 1.5 mm. long by .45 mm. wide and neither as flat, nor as pointed, nor as rough as the inner. The skin of the side of the arm en- cases the basal half of the spines and so forms a thin but con- spicuous wall along each side of the lower surface of the arm ; this occurs in many if not aU species of Ophiocreas. Basal arm-tentacles not encased in cuticular tubes. Teeth only 6, triangular and very thick ; no tooth papillae or oral papillae in the adult; in young specimens, a very few tooth papillae are present. Colour (in alcohol or dry), deep red-purple ; young specimens somewhat hghter. Fourteen specimens. This species can be distinguished at a glance from any Ophiocreas, of which I know, by the remarkably short and flattened tentacle-scales or arm -spines and the small number of teeth. Although the young specimens (disk=7 mm.; arms about 90-100 mm.) were taken much further north than any of the adults, I see no reason to feel any doubts as to their identity, the short arm-spines are so characteristic. The chief growth-changes appear to be the thickening of the skin, the entire loss of tooth-papillae and the separation and convex thickening of the radial shields. In a specimen with disk- diameter 13 mm., these changes are complete. Locs. — Twenty miles off Cape Barren, Cape Barren Island, Tasmania, 70 fathoms. East of Flinders Island, Bass Strait, 80-300 fathoms. Eastern Slope, Bass Strait, 70-100 fathoms. Eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. 1. 0rti/6/3o'?= manifest, evident ; in reference to the species being so- well characterised. 80 " ENDEAVOUR " SCIENTIFIC RESULTS. Ophiocreas sibog^, Koehler. Ophiocreas Sibogce, Koehler, Ophiures de I'Exped. du Siboga, i., Ophiures de Mer Profonde, 1904, p. 165. It is not without some hesitation that I refer these seven specimens to this East Indian species, after comparing them with a cotype of 0. sibogce. The most obvious difference is in colour, the cotype being uniformly yellow and the AustraHan specimens reddish flesh-colour with a more or less heavy purple cast. This difference, how- ever, must be disregarded since Koehler gives the colour of 0. sibogce as " reddish-purple violet," not altogether unhke the Australian material, apparently. Other differ- ences, somewhat intangible, though noticeable when the specimens are side by side, I beheve may properly be re- ferred to individual diversity. The " Endeavour " speci- mens are all adult and are in fine condition ; the largest has the disk 25 mm. across and the arms 610 mm. long. This is, I believe, much the largest specimen of Ophiocreas yet recorded. Although this species and the preceding (0. phanerum) were both taken east of Flinders Island, 80- 300 fathoms, I do not infer from this record that they occurred in the same haul of the trawl, but of course it is possible they did. They are so obviously unhke in their arm-spines (tentacle-scales) and teeth, that they cannot be confused, even if taken together. In regard to the use of the generic name Ophiocreas, I regret to see that Doderlein and Matsumoto treat it as a synonym of Asteroschema. While it is probable that the old line of division between the two genera is not by itself valid, I think it is practically certain that we are dealing with two natural groups which ought to be kept separate. I therefore propose to continue the use of Ophiocreas for species like 0. phanerum and 0. sibogce, and of Asteroschema for species like oligactes (Pallas), even though I am not yet prepared to define either genus accurately. Locs. — East of FUnders Island, Bass Strait, 80-300 fathoms. Great Australian Bight, 129° 6' 30" E., 200-300 fathoms. Family GORGONOCEPHALID^. Genus Asteroporpa, Liltken. AsTEROPORPA AUSTRALiENSis, H. L. Clark. Asteroporpa australiensis, H. L. Clark, Mem. Austr. Mus., iv., 11, 1909, p. 547. This is a nice series of this interesting species, the disk- diameters ranging from about 7 to nearly 18 mm. ; in the SEA LILIES, STARFISHES, ETC.- — CLARK. 81 largest specimen the arms apparently exceed 75 mm. There is little diversity shown in colour or in any other feature. Ten specimens. Locs. — South of Babel Island, Bass Strait, 92-160 fathoms. East of Babel Island, Bass Strait, 65-70 fathoms. Off Gabo Island, Victoria, 200 fathoms. South of Gabo to south-east of Cape Everard, Victoria, 70-80 fathoms. Genus Conocladus, H. L. Clark. CoNOCLADUS AMBLYCONus, H. L. Clark. Conocladus amblyconus, H. L. Clark, Mem. Austr. Mus., iv., 11, 1909, p. 549. This interesting series, with disk-diameter ranging from 12-40 mm. is not quite so diversified in size as that taken by the " Thetis," but in the number and distribution of the rounded tubercles on disk and arm-bases, it shows more diversity. The colour is also notable, some specimens being almost white, and others purplish-brown, while the remainder are fawn-colour. There is no indication of any intergrada- tion with any other species, the distinctive characters being perfectly plain in all the twenty-one specimens. Locs. — Between Port Stephens and Newcastle, New South Wales, 22-60 fathoms. Shoalhaven Bight, New South Wales, 15-45 fathoms. Off Twofold Bay, New South Wales, 30 fathoms. Conocladus oxyconus, H. L. Clark. Conocladus oxyconus, H. L. Clark, Bull. Mus. Comp. Zool. Harvard, lii., 1909, p. 132. The individuals from Shoalhaven Bight, New South Wales, are very similar to the originals from off Port Jackson, and call for no special comment. The specimen from South Austraha was taken with a specimen of the next species {Astroconus australis) and owing to the locality was at first identified as Astroconus. Critical study has, however, con- vinced me that if Conocladus oxyconus and Astroconus aus- tralis are to be maintained as distinct species, this South Australian individual must be called Conocladus. But C. 82 " ENDEAVOUR " SCIENTIFIC RESULTS. oxyconus is not otherwise known from south or west of Shoal- haven Bight, and I must confess that the occurrence of this specimen taken off Cape Wiles, South Austraha, with an undoubted Astroconus has shaken my faith in the distinctness of the two forms. I should like to believe that the specimen under discussion was wrongly labelled, but I cannot see the shghtest reason for such a belief. The question of the relation- ship of Astroconus and Conocladus is thus reopened and will be discussed further under the next species, but it is worth noting here that up to the present time only eight specimens of C. oxyconus are known and two of these are from unknown locahties, though there is reason to think they are from near Port Jackson ; of the other six, two are known to be from off Port Jackson, one is from Coogee, two from Shoalhaven Bight and one from south of Cape Wiles, South Austraha. Of forty specimens of C. amblyconus now known, all are from New South Wales, while all known specimens of Astroconus are from Bass Strait southward and westward. Evidently Conocladus is the more northern genus and C. oxyconus is much the rarest of the three species. Lacs. — Shoalhaven Bight, New South Wales, 15-45 fathoms. Two specimens. Fifteen miles south of Cape Wiles, South Austraha. Grenus Astroconus, Doderlein. Astroconus australis (Verrill). Astrophyton australe, Verrill, Bull. U.S. Nat. Mus., iii., 1876, p. 74. Astroconus australis, Doderlein, Abh. math-phys. Klasse K. Bayer Akad. Wissensch., ii. Suppl.-Bd., 5, 1911, p. 37. This is a remarkable and highly interesting series of this little known species. The fifty-five specimens range in size from young ones, with disks 11-12 mm. in diameter to large adults, whose disks are 45-50 mm. across. Two are perfectly tetramerous. The colour is notably diversified, ranging from nearly uniform pale gray or almost white, through pale gray prettily marked with dull red-purple, to dark gray heavily marked with purphsh ; or the ground colour may be yelloAvish-brown, hght or dark, either with- out markings of red-purple, or prettily marked with that shade ; many individuals have the arms very regularly banded with the red-purple, but in others there are only SEA. LILIES, STARFISHES, ETC. — CLARK. 83 a few dots or no markings at all. Diversity is equally marked in other features. The distance of the first arm- fork from the distal end of the radial shields ranges from .25 of the radial shield length to 1.10 ; or the same fact may be expressed in another way by saying there are from 9 to 15 arm-joints in the undivided arm-base. The number of arm- divisions ranges from 6 to 10 or more. Most of the specimens show the radial shields well separated, and the disk has distinct interradial areas, but there is a tendency on the part of some individuals to have the radial shields united in pairs by thick, concealing skin and in such cases the interradial areas are much reduced ; thus the condition, characteristic of Conodadus, is approached. Further resemblance to Cono- cladus is brought about in certain specimens by the large size of the disk spines. In typical A. australis, these spines are about a millimeter high, conical, and the thickness at base about equals the height ; they are present chiefly on the ' radial shields and are often confined to them ; in many speci- mens they are smaller and more numerous, and rarely they are reduced to low, inconspicuous tubercles. But in some cases, notably in some South Austrahan specimens, the spines are 1.5 mm. high and correspondingly stout. In these cases, too, the spines on the basal parts of the arms are larger than usual and the resemblance to Conocladus may be quite marked. Nevertheless none of the specimens have the radial shields completely united into five disk-wedges, with very much reduced interradial areas as in Conocladtis, nor do any have spines on disk and arm-bases really as big as in that genus. For these reasons one can, without difficulty, separate speci- mens of the two genera, but I am not sure that the separation is not an arbitrary rather than a natural one. I have previ- ously expressed the beUef that Conocladus is derived from Astroconus, while Doderlein thinks the reverse to be the case. Which of us is correct can only be determined by a study of growth-changes in much younger specimens than any at present available. But the youngest specimens in the present series are not at all Conocladus-Vike, the radial shields being narrow and widely separated, the interradial areas large and the disk spines few and small. It is quite possible that further study and more abundant material will show that Conodadus oxyconus is only an extreme form of Astroconus australis, and not a separate species. In that case the genus Conocladus will need a new name, i.e., the species amhlyconus and microconus, for the name Conocladus will have to replace Astroconus as the generic name of VerriU's Astrophyton australe, since it has the sarhe type and antedates Astroconus by two years. In any case, whether C. oxyconus is specifically different from 84 " ENDEAVOUR " SCIENTIFIC RESULTS. Astroconus australis or not, it does not seem to me possible to draw a subfamily distinction between them, as Doderlein does, and place one in the Astrochelinse and the other in the Gorgonocephahnse. If these two groups are worth dis- tinguishing as subfamihes, I think Conocladus must certainly be put in the Gorgonocephahnae. iocs. — Thirty-five miles south-east of Bruni Island, Tas- mania, 150-230 fathoms. North-east of Cape Pillar, Tasmania, 50-80 fathoms. Bay of Fires, Tasmania, 40 fathoms. Twenty miles east of Babel Island, Bass Strait, 35 fathoms. Between Devonport and Launceston, Tasmania. Eastern Slope, Bass Strait, 70-100 fathoms. Off Gabo Island, Victoria, about 200 fathoms. Thirty-six miles N. 58° E., of Cape Wickham, King Island, Bass Strait. South-east coast of Australia. Coast of Victoria. Forty miles west of Kingston, South Austraha, 30 fathoms. Off Murray River mouth. South Australia, 17 fathoms. Sanders Bank, Kangaroo Island, South Australia, 28 fathoms. Fifteen miles south of Cape Wiles, South Australia. Fifty miles south of Cape Wiles, South Australia, 75 fathoms. South-east of Flinders Island, South Austraha, 37 fathoms. Fifteen miles south of St. Francis Island, South Austraha, 30 fathoms. Genus Astrodendrum, Doderlein. ASTRODENDRUM PUSTULATUM,! Sf. nOV. (Plate xxxiv., fig. 1-2.) Disk diameter, 40 mm. ; arms, divided about 8 times, approximately 225 mm. long, measured from mouth ; one arm measured as follows, by divisions, the figures, however, being only approximate : from mouth to first fork, 20 mm. ; second division, 12 mm. ; third, 13 mm. ; fourth, 30 mm. ; 1. Pustulatus=h.a,\\na blisters or pustules ; in reference to the pimple like elevations on the disk. SEA LILIES, STARFISHES, ETC. — CLARK. 85 fifth, 30 mm. ; sixth, 27 mm. ; seventh, 37 mm. ; eighth, 30 mm. ; and ninth, 25 mm. Radial shields distinct, well separated, but in pairs ; distance between distal ends of a pair, 6 mm. ; width of shield at distal end, 3 mm. ; distance between distal ends, across an interradius, 12 mm. Disk covered by a smooth, thick skin without calcareous plates, even at margin ; all over the surface, but chiefly along the radial shields, are scattered minute, pimple-hke elevations, not at all crowded, but where thickest there may be 7 or 8 in a space 3 mm. square; large parts of the surface, especially near margin have none. Oral surface covered by the same smooth, thick, non-calcified skin, but with no pustules. Madreporite single, small, but distinct. Genital slits only 5 mm. long, but very distinct. Teeth and teeth -papillae indistinguishable, rather flat, about three times as long as wide. Mouth-papillae none. First tentacle pore of arm without any protecting scale ; succeeding pores of basal portion of arm with two small peg-like scales ; on first division of arm, there are three scales to each pore, the inner one nearly a milUmeter long and .50-. 66 mm. thick, the second about as long, but more slender, the outer distinctly smaller in every way ; all three have rough, prickly tips when dry. Distally the outer spine disaj^pears and then the second, while the inner one becomes transformed into the character- istic hook. The half-circles of booklets are very narrow, con- sisting of only 2 series of granules ; they are incomplete on the second arm-division and on the first are represented only by httle patches ; on the third and distal thereto they are com- plete. Colour (in alcohol) nearly uniform reddish flesh- colour ; the pustules and tentacle -scales are somewhat darker. This species is easily distinguished from either of the other species in this restricted genus by the characteristic disk- covering and the tentacle-scales. It is so utterly unlike any other known Australian euryalid that there is no possibility of confusion with any of them. Loc. — East of Fhnders Island, Bass Strait, 100-300 fathoms. Genus Astrothamnus, Matsumoto. ASTROTHAMNUS RUGOSUS,! Sp. nOV. (Plate XXXV., fig. 1-2.) Disk 20 mm. in diameter ; arms 5, approximately 140 mm. long, but the tips are so closely spirally coiled, it is not possible to measure them exactly. Disk covered by coarse, rounded 1. Eugosus=TOugh. ; in reference to the rough character of the disk- covering. 86 " ENDEAVOUR " SCIENTIFIC RESULTS. granules .30-. 90 mm. in diameter, among which are a few insignificant, minute plates ; the granules are low, hemi- spherical, with rough, or prickly surface. Radial shields approximate in pairs, 10-12 mm. long by 2-2.5 mm. wide distally, more or less concealed. Arms sharply distinct from disk, the dorsal surface and sides covering by alternating half-circles of glassy hook-bearing, and opaque, smooth granules ; the half-circles are not at all sharply defined under a lens but nevertheless give an appearance of fairly uniform annulations. Genital sHts only about 2 mm. long, but quite wide. Madreporite small but very distinct, just outside the mouth frame ; in the holotype, there are two, not in adjoining interradii, but one is evidently smaller and out of place. Oral surface and lower surface of arms, uniformly and finely granulated. Teeth, teeth-papillae and mouth-papillse numer- ous, similar, spiniform ; teeth somewhat larger in every way than the others. First tentacle-pore of arm with 3 or 4 short, stout, spinulose tentacle-scales ; second with 6 or 7, third with 8, succeeding pores with 8, 9 or 10, nearly to middle of arm, when the number begins to drop rapidly and the scales: become more and more Hke the hooks on the lower ends of the half-circles of hooks, until at the tip of the arm the tentacle- scales are no longer distinguishable. Colour (in alcohol), brownish-yellow ; when dry, considerably Ughter. Four specimens. This is a very distinct species, remarkable for the large number of tentacle-scales. The disk covering too is character- istic, obviously different from that of any of the previously known species. The genus is an East Indian one, reaching Sagami Bay, Japan, on the north and now shown to reach; Tasmania on the south. The specimens from off Cape Everard,. Victoria, are only 9 mm. across the disk and the maximum number of tentacle-scales seems to be 7. Locs. — East of Flinders Island, Bass Strait, 80-300 fathoms.. South-east of Cape Everard, Victoria, 200 fathoms. Family OPHIACANTHID^. Genus Ophiacantha, Miiller and Troschel. Ophiacantha heterotyla, H. L. Clark. Ophiacantha, heterotyla, H. L. Clark, Mem. Austr. Mus., iv., 11, 1909, p. 542. This is a little individual, not 3 mm. across the disk, but there seems no doubt of its identity. Loc. — Between Devonport and Launceston, Tasmania. SEA LILIES, STARFISHES, ETC. — CLARK. 87 Family AMPHIURID^. Genus Ophiactis, Lutken. Ophiactis luteomacflata, H. L. Clark. Ophiactis luteomaculata, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xxv., 1915. These specimens were all taken from the large spines of a Sea-urchin (Prionocidaris australis) and as a rule, they were more or less hidden in Barnacles (Acorn-shells) growing thereon. The species was hitherto known only from the holotype taken at the Murray Islands, Torres Strait, in a large sponge. None of the present series is quite so large as the original but they show little structural divergence. The smallest specimens have 6 arms and are clearly reproducing by fission, but the adults are symmetrically 5-rayed. All the specimens are dry and their colours are dull but the adults have the arms pinkish and more or less distinctly banded. The radial shields are not of a different colour from the rest of the disk. An excellent specific character is a dark band or spot at the middle of the larger arm -spines ; all these specimens show it. Loc. — Eleven miles south-east of Ballina, New South Wales,, 27-29 fathoms. Eleven specimens. Ophiactis resiliens, Lyman. Ophiactis resiliens, Lyman, Bull. Mus. Comp. Zool. Harvard,, vi., 1879, p. 36. One of the individvials from off Kingston, South Austraha, (disk-diameter, 6 mm.) has all the upper arm-plates broken into 2 or 3 pieces. The other specimens, with disk-diameters of 3.5 and 5.5 mm. respectively, are not pecuhar. It may be noted, however, that the species has not previously been known from west of Bass Strait. Three specimens. Locs. — Forty miles west of Kingston, South Australia, 30- fathoms. Off Cape Martin, South Australia, 21 fathoms. 88 " ENDEAVOUR " SCIENTIFIC RESULTS. Ophiactis symbiota,! sp. nov. (Plate XXX vi., fig. 1-2.) Disk 6.5 mm. in diameter ; arms about 26 mm. long. Disk covered by a coat of 150-200 rather coarse, flat, rounded scales, the largest of which are near centre and the smallest near interradial margins ; between these plates or on their corners or edges are low, conical tubercles which increase in length near the interradial margins, becoming short sharp spinelets ; these tubercles and spinelets are not very numerous, some plates having none, while the largest plates have only 5-6 around them. Radial shields small, about 1 mm. long by .5 mm. wide, well separated within but sometimes in contact distally. Upper arm plates, fan-shaped, truncate proximally, strongly convex distally, with lateral corners more or less rounded or truncate. Interbrachial areas below covered like the interradial margins of disk, with rather small scales and numerous sharp spinelets. Oral shields pentagonal, widest proximally, disto-lateral sides concave, distal side shortest ; madreporite very much larger than other four and more iso -pentagonal. Adoral plates large, twice as long as broad, curved, meeting broadly within. Teeth about 6, squarish ; below the lowest is a very small, triangular dental papilla. Oral papillae apparently 2 on a side but the proximal may be the scale of the inner oral tentacle-pore ; distal papillae nearly circular, about .20 mm. in diameter ; proximal, a trifle smaller and more angular. First under arm -plate relatively large, longer than wide, in contact with second, thus broadly separating the adoral plates ; succeeding plates tetragonal, wider than long, with distal margin shghtly con- vex and lateral margins concave, becoming, beyond the middle of arm, pentagonal with a proximal angle and longer than wide ; first 5 or 6 barely in contact. Side arm plates rather large meeting above and below at middle of arm and beyond ; each carries 4, or usually 3, opaque, stout, blunt spines ; upper- most longest, about 1 mm. long ; lowest, shortest, about .60 mm. long. Tentacle-scale single, moderately large, at side of lower arm-plate. Colour (dry), brownish- white. This interesting Ophiuran, I found on a specimen of Com- anthus plectrophorum, clinging closely to the oral surface of the disk. Its position leaves Uttle doubt in my mind that it was commensal with the Crinoid in life, as a number of species of Brittle-stars are known to be. It is utterly unlike 0. 1- (TW/i/3<oTos= living with ; in reference to the apparently symbiotic iiabits. SEA LILIES, STARFISHES, ETC. — CXAKK. 89 resiliens in its general appearance and I at first thought it to be an Ophiolebes, but the mouth parts seem to be undoubtedly those of an Ophiactis and I therefore have referred it to this genus without dissection. The oral surface is rather markedly like Ophiactis abyssicola but the disk-covering is utterly unhke that species, and is indeed the distinguishing specific character. The short arms, few stout arm-spines and pair of oral papillae are additional features worthy of emphasis. Loc. — East of Fhnders Island, Bass Strait, 100-300 fathoms. Genus Amphiura, Forbes. Amphiura constricta, Lyman. Amphiura constricta, Lyman, Bull. Mus. Comp. Zool. Harvard, vi., 1879, p. 22. These specimens are all smaller than Lyman's type and as a rule have only 4 or 5 arm spines, but on a few basal arm joints of the largest specimen (disk diameter, 4 mm.) there are 6 spines. The type locality was Port Jackson, New South Wales, and there is a specimen in the Museum of Com- parative Zoology from Westernport, Victoria. The " Endea- vour " specimens thus extend the known range considerably to the westward. Three specimens. Locs. — Forty miles west of Kingston, South Australia, 30 fathoms. Off Cape Martin, South Australia, 21 fathoms. Family OPHIOTRICHIDiE. Genus Ophiothrix, Muller and Troschel. Ophiothrix aristulata, Lyman. Ophiothrix aristulata, Lyman, Bull. Mus. Comp. Zool. Harvard, vi., 1879, p. 50. These specimens range from 5 to 18 mm. in disk-diameter and make a fine series of this handsome Brittle-star. The colour ranges from nearly white to rosy pink; the spines are always white or colourless Two of the large individuals are beautifully though sparsely spotted with bright brown. There is a broad longitudinal stripe on the upper side of the arms, in one large specimen deep pink, but usually pure white. Jt seems to be lacking in the specimen from Oyster Bay, yO " ENDEAVOUR " SCIENTIFIC RESULTS. Tasmania, and this individual is also peculiar in the large number of disc scales and the short, thorny disk spines ; the disk-diameter, however, is only 6.5 mm., and I beheve the peculiarities are associated with youth. The smallest speci- men is from east of Maria Island, Tasmania, and still retains the large, transparent central plate of the disk, known to be characteristic of the very young of some other species of Ophiothrix ; its disk spines are quite thorny and relatively rather short. Nineteen specimens. Locs. — Twenty miles south-east by east of Tasman Head, Tasmania, 80-85 fathoms. East of Maria Island, Tasmania, 78 fathoms. Entrance to Oyster Bay, Tasmania. East of Babel Island, Bass Strait, 65-70 fathoms. Off Gabo Island, Victoria, 200 fathoms. Ophiothrix c^spitosa, Lyman. Ophiothrix coespitosa, Lyman, Bull. Mus. Comp. ZooL Harvard, vi., 1879, p. 53. These specimens range in disk-diameter from 1.5 to 6 mm., and in colour from bright reddish-pink to purplish-gray^ brownish-yellow and nearly white. The disk covering shows much diversity also, ranging from a uniform covering of minute thorny, somewhat stellate stumps, which even conceal the radial shields, to one where the thorny stumps are very largely replaced by rough spinelets and the radial shields are more or less bare ; there are many intermediate stages. The upper arm plates and the arm spines are more uniform, but not perfectly so. I cannot see that there is any correlation between the characters mentioned and the locality whence the specimen came. The seven specimens from off BaUina, New South Wales, were aU found on the abactinal surface of a Sea-urchin {Prionocidaris australis) ; they are all young and are bright reddish-pink, except two which have pale pink arms and a light coloured disk hardly tinged with pink. In view of the well-known instability of characters in the genus, and the present multiplicity of species, I think it best to regard all the specimens as 0. coespitosa and leave it for some future worker in Australia to determine whether J am right. Forty-one specimens. SEA LILIES, STARFISHES, ETC. — CLARK. 91' Locs. — Twenty-three miles east-south-east of Round Hill Head, Queensland. Nine miles east of Fraser Island, Queensland. Eleven miles south-east of BalHna, New South Wales, 27-29' fathoms. Eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. Forty miles west of Kingston, South Australia, 30 fathoms. Sanders Bank, Kangaroo Island, South Australia, 28^ fathoms. Fifteen miles north-west of Cape Jervis, South Australia, 17 fathoms. Ophiothrix fumaria, Mailer and Troschel. Ophiothrix fumaria, Miiller and Troschel, Sys. d. Ast., 1842, p. 113. This is a well-preserved individual with disk 9 mm. across^ agreeing in all essentials with a specimen of 0. fumaria identified by Mr. Lyman many years ago, which was pur- chased in Hamburg and was reported to have come from the East Indies. The type locality of the species is stated by Miiller and Troschel to be unknown. Loc. — Seven miles south-south-east of Double Island Point, Queensland. Ophiothrix nereidina {Lamarck). Ophiura nereidina, Lamarck, Anim. sans Vert., ii., 1816^ p. 544. Ophiothrix nereidina, Miiller and Troschel, Sys. d. Ast.^ 1842, p. 115. Although small, only 4 mm. across the disk, these arfr typical examples of this beautiful species. They are chiefly of interest in that they extend the known range of the species- many hundreds of miles to the south. Two specimens. Loc. — Nine miles east of Fraser Island, Queensland. '92 " ENDEAVOUR " SCIENTIFIC RESULTS. Ophiothrix spongicola, Stimpson. Ophiothrix spongicola, Stimpson, Proc. Acad. Nat. Sci. Philadelphia, vii., 1855, p. 385. These are typical specimens of this handsome and charac- teristically Australian species. It seems to me quite possible, however, that 0. jumaria is really the proper name for this species, Miiller and Troschel's type having very probably been brought to Paris from Port Jackson, but never having had a locality label. Koehler's careful description of it and his figures apply very closely to this species, more closely than to the specimen Mr. Lyman has called 0. jumaria. In the ultimate revision of the genus, this matter should be given critical attention. Three specimens. Locs. — Shoalhaven Bight, New South Wales, 15-45 fathoms. Fifteen miles north-west of Cape Jervis, South Australia, 17 fathoms. Genus Ophiothela, Verrill. Ophiothela hadra, H. L. Clark. Ophiothela Jmdra, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xxv., 1915, p. 284. This is another species, previously known only from the holotype taken at the Murray Islands, Torres Strait. The present series shows that the original was immature, and the arms are not unusually short and stout in adults, though the general form is rather robust. The largest specimen at hand is 3 mm. across the disk and the arms are 10-11 mm. long. Only the smallest specimens are as hght as the holotype, aU the larger specimens being prettily variegated with dull deep bluish, pink and pale cream-colour ; the centre of the disk, the sides of the arms and the arm spines are pink ; but there is some diversity. There is also some diversity in the armature of the disk, the larger interradial spine-like granules being often wanting. The oral shields are very small, nearly circular, white ; adoral plates larger than oral shield, but quite similar in form and colour, perfectly smooth, in close contact with each other. There are 5 arm spines at base of arm. The knobs on the upper arm plates are conspicuous. AU of these specimens were taken from the primary spines of Sea-urchins [Prionocidaris australis) ; most of the smaller ones are reproducing by fission and have only 2, 3 or 4 rays, but the adults are symmetrically 6-rayed. Loc. — Eleven miles south-east of Ballina, New South Wales, 27-29 fathoms. SEA LILIES, STARFISHES, ETC. — CLARK. 9^ Family OPHIOLEPIDID^. Genus Amphiophiura, Matsumoto. Amphiophiura collet ai, sp. nov. (Plate XXX vii., fig. 1-2.) Disk 17 mm. in diameter; arms all broken, but one is nearly 50 mm. long and indicates a probable length of 55 mm. or more. Disk covered by a coat of 21 large, flat, smooth scales and the 10 radial shields, between which, in the angles, are much smaller, more or less triangular scales ; many of these little scales and some of the larger show a narrow, marginal border as though cemented down to the neighbouring plates ; primary plates very distinct, about 2.5 mm. across, but not conspicuous. Radial shields, 3-4 mm. long and nearly or quite as broad, irregularly rounded triangular, smooth, the inner ends separated by a large scale, but the distal inner corner of one distinctly overlapping the corre- sponding corner of the other for more than a milUmeter. Basal upper arm plate about twice as broad as long ; second and third at least three times as broad as long ; the succeeding 7 or 8 plates are tetragonal, wider than long, broadly in contact, with distal margin convex and much wider than proximal ; beyond the tenth joint the upper arm plates tend to become triangular, as long as broad, and separated from each other, the condition they show on distal half of arm. Interbrachial areas below covered by the oral shield, a large marginal plate, wider than long and 5 smaller, somewhat swollen plates, between them. Genital plates large, more than a milhmeter wide at the upper end, bearing an arm -comb of about 25 low, triangular papillae, placed closely side by side ; there are no papillae on the arm plates. Oral shields oval, sharply pointed proximalty, 4 mm. long by 3 mm. wide ; adoral plates, short, three times as long as broad, meeting fulty and closely within ; oral plates larger than adorals, only slightly swollen. Teeth only 4, triangular. Oral papillae only 3 on a side, borne wholly on the oral plates ; the innermost is Uke the teeth, but narrower, while the other two are very wide and notched like the molar teeth of a dog. Oral tentacle-pore (the second or distal one, of course) opens entirely outside mouth slit, at distal end of oral plate ; its inner side is guarded by 4 papillae, its outer by 6. First under arm plate large, octagonal, much wider than long ; second longer and narrower, hexagonal, about as long as wide ; 1. *:oA,X)yTo'9= glued together; in reference to the appearance of the- disk plates. 04 " ENDEAVOUR SCIENTIFIC RESULTS. succeeding plates wider than long, more or less tetragonal, broadly in contact. Beyond the eighth or ninth joints the under arm plates are nearly or quite separated from each other, pentagonal with proximal angle, much wider than long ; they keep this form to the end of the arm. Side arm plates rather large, but not conspicuous, well separated at base of arm, but broadly in contact distally ; each carries 3 small arm spines, a minute, bluntly pointed, peg-like one on upper distal corner of plate (often wanting distally) and a pair of somewhat larger, more flattened spines, close to the lower distal corner of plate ; except on one or two basal joints, none of the arm spines are as long as one-third the length of the side arm plates. Tentacle-pores rather larger ; the basal ones have 3 or 4 scales on each side, but the number soon dwindles to one on the under arm plate and one on the side arm plate and near the middle of the arm, the former dis- appears. Colour (dry), yellowish-white. This rather handsome species is well characterised by the peculiar oral papillae, the arrangement of the arm spines, the shape of the papillae in the arm-comb and the arrangement of the disk plates. In all these particulars, as well as in the shape of the under arm -plates it differs from A. ornata, which is one of its near alhes, but except in the arrangement of the disk scales it is no nearer to A. undata, which is one of its nearest allies geographically. Its longer arms and very different arm-comb will distinguish it at once from A. cteno- phora (H. L. Clark) taken by the " Thetis " off the coast of New South Wales. Loc. — East of Babel Island, Bass Strait, 60-80 fathoms. Family OPHIODERMATrDiE. Grenus Pectinuba, Forbes. Pectestura dyscrita, H. L. Clark. Pectinura dyscrita, H. L. Clark, Mem. Austr. Mus., iv., 11, 1909, p. 534. This series of specimens has led me to question very much whether P. dyscrita can properly be distinguished from P. anchista, the Japanese species. But after comparing specimens of the same size, I have decided not to unite them at present. The arm spines in P. anchista are thinner, flatter and more pointed than in P. dyscrita. If similar specimens of Pectinura should hereafter be found in the Phihppines or in the East Indies, it would probably result in including the SEA LILIES, STARFISHES, ETC.- — CLARK. 95 Japanese and Tasmanian forms under the same name. The New Zealand species, P. cylindrica, is undoubtedly a very near relation but seems to be a smaller form, with arm spines and arm plates somewhat different. Lyman's P. arenosa from Bass Strait and Bell's P. assimilis from Port Jackson seem to be quite distinct, but it must be granted that the five species form a very homogeneous group. The " Endeavour " specimens show^ plainly that in P. dyscrita at least the relative as well as the actual arm-length increases markedly with age ; in the smallest specimen the disk-diameter (D) is 10 mm., the arm length (A) 40 mm., and hence A=4D, while in another D = 12, A=60, D=5A, and in the largest specimen, D=20, A =140, D = 7A. There is some diversity in colour indicated ; the arms are often banded with brownish, or dirty whitish, and in one case there are 5 or 6 transverse bands, dorsally, of rose-purple. The colour in life is probably white, dull yeUow, or pale brown, with rose-purple markings, spots or bands. Fourteen speci- mens. Locs. — North-east of Cape PiUar, Tasmania, 80 fathoms. East of Maria Island, Tasmania, 78-128 fathoms. East of Babel Island, Bass Strait, 65-70 fathoms. East of Flinders Island, Bass Strait, 80-300 fathoms. Off Gabo Island, Victoria, 200 fathoms. FamHy OPHIOCHITONID^. Genus Ophionereis, Liitken. Ophionereis schayeri (Muller and Troschel). Ophiolepis schayeri, Miiller and Troschel, Arch. f. Naturg., x.i., 1844, p. 182. Ophionereis schayeri, Liitken, Add. ad Hist. Oph., pt. 2, 1859, p. 110. This is a very small specimen, only 4.5 mm. across the disk, but it is not otherwise notable. Loc. — Between Devonport and Launceston, Tasmania. ECHINOIDEA. The collection of Sea-urchins made by the "Endeavour" strengthens the evidence in support of the opinion formed by study of the "Thetis" collection, that Echini are relatively unusually common in the waters of Southern Australia. Considering the fact that there are rather more than one 96 " ENDEAVOUR " SCIENTIFIC RESULTS. thousand four hundred species of Ophiurans known and that only twenty-six occur in the " Endeavour " collection, it is remarkable that twenty-nine species of Echini occur, with fewer than five hundred species known. Of these twenty- nine species, five are here described for the first time and five others are recorded for the first time from Australia. The most remarkable specimen is undoubtedly the adult indi- vidual of Echinus horridus, a species previously known from off the coast of ChiH, but the fine series of Prionocidaris australis, the handsome new Coelopleurus and the new Maretia are worthy of particular mention. Up to the present time about sixty species of Echini have been listed from Australia, but wrong iclentifications have been so numerous and mis- leading, the real number of species is very uncertain. Probably, including the ten here added to the list, there are at least sixty Australian Echini occurring in less than three hundred fathoms, thirty-five of which are pecuHar to AustraUa. Family CIDARID^. Genus Phyll acanthus, Brandt. As regards the generic name Phyllacanthus, I have given fresh and careful attention to Brandt's paper, and I see no ground for rejecting his subgenera. Lambert and Thieryi have argued very strongly that these groups are invahd, but there seems to me a fundamental error in the argument which quite vitiates it. They assume that Brandt is revising and classifying all Echini, and hence \vhen he gives three subgenera under the genus Echinus they assume that all the species of Echinus are to be placed in these three sub- genera, and, of course, one would then liave to be a synonym of Echinus. But such is not Brandt's intention. He is simply giving a classification of the forms noted by Mertens in his voyage ; occasionally other species are mentioned for illustrative purposes, but there is no attempt to group all the known Echini. Under the circumstances, then, I do not see Avhy Brandt's subgenera are not perfectly valid, certainly as much so as any genera of Echini that were proposed prior to 1846. Particularly in the case of Phyllacanthus, Brandt's course is clear; he had a large cidarid from the Bonin Islands to deal with ; its peculiarities were obvious and are fairly well stated ; for it he proposed a subgenus of Cidarites which he called Phyllacanthus, at the same time designating the species as P. dubia. The fact that he was planning for a full 1. Lambert and Thiery— Bull. Soc. Sci. Nat. Haute Marne, vi., 23, 1909. SEA LILIES, STARFISHES, ETC. — CLARK. 97 classification of Echini (as he states in a footnote) and so suggests two sections for the subgenus and mentions certain of Lamarck's species of Cidarites in connection with them, does not so far as I can see affect the status of Phyllacanthus at all. It seems to me clear that the only generic name, except Cidaris {==^ Cidarites), which may be used correctly for dubia and imperialis, v\^hether they are identical, as is generally beheved, or not, is Phyllacanthus. Phyllacanthus imperialis {Lamarck). Cidarites imperialis, Lamarck, Anim. sans Vert., iii., 1816, p. 54. Phyllacanthus imperialis, Brandt, Prodrome, 1835, p. 268. The test of this individual is about 45 mm. in diameter and the longest spines are 60 mm. long and 5 mm. thick. It is un- doubtedly an example of Tenison-Woods' P. parvispina, but I am not prepared to recognise the form by name yet. A good series of specimens from the Murray Islands in Torres Strait are all typical P. imperialis, so far as I can see, but all the New South Wales material I have examined has the slender spines of P. parvispina. The amount of available material from other regions is not sufficient to enable me to decide as yet on the worth of this character. Loc. — Shoalhaven Bight, New South Wales, 15-45 fathoms. Genus Prionocidabis, A. Agassiz. Pbionocidaris australis {Ramsay). Phyllacanthus australis, Ramsay, Cat. Echinod. Austr. Mus., 1885, p. 44. Stephanocidaris bispinosa, A. Agassiz, Rev. Ech., pt. 1, 1872, p. 160 {nomen nudum) ; pt. 3, Sept., 1873, p. 393 (non Cidarites bispinosa, Lamarck, as described and figured by de Loriol, Mem. Soc. Sci. Nat. Neuchatel, v., May, 1873, p. 33). Prionocidaris agassizii Doderlein, Abh. Senck. Naturf. Gcs., xxxiv., 1911, p. 242. This is a remarkable series of a very notable species. Doderlein {loc. cit.) has shown very clearly that the genus Stephanocidari<i A. Ag. is a pure synonym of Goniocidaris a,nd hence carmot be used for thiis species. He shows further that 98 " ENDEAVOUR " SCIENTIFIC RESULTS. Mr. Agassiz's application of the name bispinosa to this species cannot be maintained since de Loriol's restriction of Lamarck's name to the species Mr. Agassiz called annulifera has several months' priority over Mr. Agassiz's selection. The latter being left thus without a name, Doderlein courteously pro- posed to call it agassizii. When at the Austrahan Museum in November, 1913, I was surprised to find that Ramsay's type of PhyllacantJius australis is a fine example of the Sea-urchin Mr. Agassiz and I have called Stephanocidaris bispinosa. It is not strange that no one has suspected it hitherto, for Ramsay gives virtually no description but rehes on three photographs to make clear the specific characters. The whole nomenclatural tangle is now fairly well cleared up, and apparently all living students of the recent Echini are agreed as to the proper use of Lamarck's names for cidarids, even though we still disagree as to generic divisions. The present series of P. australis ranges from 27 to 73 mm. in test-diameter, horizontally (h.d.) ; the specimen with h.d.= 27 mm., has primary spines 67 mm. long or 2.5 h.d., while in the big individual they are only 60-65 mm. long or less than h.d. Evidently 65-70 mm. is the full length of the spine for this species, and growth ceases when that length is attained, though the test may continue to grow indefinitely. The primary spines when young are purple and smooth ; they soon develop sharp teeth along the sides, and with this development the spine becomes somewhat flattened ; at least the oral side is more or less flattened, the aboral side tending to become carina te ; the spine is thus almost or quite triangular in cross-section, where widest ; as the lateral teeth increase in size numerous similar but smaller teeth arise, particularly along the aboral side ; with this change in form of the spine there goes an equally great change in colour ; the teeth are greenish and as they increase in number the spines appear to be less and less purple ; as the spine matures it becomes coated with a loose, colourless or whitish calcareous coat which completely masks the remaining purple colour ; then Sponges, Bryozoa, Brachiopods, Lamellibranchs, Worm- tubes and Barnacles (but particularly Barnacles) proceed to encrust the adult spines, making them the habitations of small Brittle-stars and Crustaceans. Many of the old, short actinal primaries (but not the small ones around the mouth) become very flaring at the tip in old individuals ; thus a spine 3.3 mm. in diameter near the middle may be almost 5 mm. across the obliquely truncate and almost flat tip. The white spots on the collar of the primaries appear as soon as the collar is formed ; they are always visible, but the distinctness SEA LILIES, STARFISHES, ETC. — CLARK. 99 differs much in different individuals and on different spines ; occasionally they merge into more or less complete longitudinal lines, but this is unusual ; they are often well developed on the neck of the spine above the collar. The actinal primaries have the collar very wide, and a thick, somewhat serrate " cap " formed from the outer layer of the spine. The abactinal system is larger than the actinostome, but the difference shows some diversity ; in the smallest specimen the actinostome is little more than 8 mm., while the abactinal system is nearly 12 ; in the largest specimen the measure- ments are approximately 25 and 32 ; but in a specimen of average size the measurements are only 18 and 20. The secondary spines show remarkable diversity of colour ; in several specimens they are bright red, with those of the ambulacra more or less greenish at both base and tip, and those of the interambulacra with very slightly greenish tips. In other cases they are reddish with purple tips and a longi- tudinal purplish stripe, or they are dull flesh-colour with red longitudinal stripe. In the largest specimens, all the second- aries are more or less greenish, those of the interambulacra with darker longitudinal stripe and pink or purplish tip, while those of ambulacra are more nearly uniformly dark greenish. Evidently no specific character can be drawn from the coloura- tion of the secondary spines ! The occurrence of globiferous pedicellarise shows an equal diversity ; in many specimens they are rare or wholly wanting, but in the larger specimens they are generally present, at least in the interambulacra, and in several of the biggest specimens they are excessively abundant and very con- spicuous, as they are white or pale greenish in colour. The stalk is 3 mm. long (more or less) and has a very conspicuous limb. The head is about a millimeter long and the valves seem to be like those of P. baculosa. I have hitherto laid much stress on the absence of globiferous pedicellarise in " Stephano- cidaris bispinosa, A. Agassiz," but evidently the two specimens previously seen by me were misleading. I can no longer doubt that this species and baculosa are congeneric and that I have been wrong in counting the latter a Phyllacanthus. Since as already shown the name Stephanocidaris must be abandoned, I follow Doderlein in using Prionocidaris, of which P. pistillaris {—baculosa) is the type, for the group I have hitherto called Stephanocidaris, the diagnosis being modified to permit the inclusion of baculosa. Thirty-four specimens. Locs. — Off Eraser Island, Queensland. ]00 " ENDEAVOUR " SCIENTIFIC RESULTS. Twenty-five miles south-east of Double Island Point, Queensland. Eleven miles south-east of BalHna, New South Wales, 27- 29 fathoms. Genus Cidaris, Leske. CiDARIS CONFERTA,! Sp. nOV. (Plate xxxviii., fig. 1-4.) Test Httle flattened ; vertical diameter about .60-. 70 of horizontal ; coronal plates 7 or 8 ; areolae large, nearly circular, only the lowest 2 or 3 confluent ; median interambu- lacral area with vertical suture quite distinct and 3-4 series of tubercles on each side of it ; ambulacra about one-fourth of interambulacra in width ; poriferous zones not sunken, very narrow, the pores close together ; median ambulacral area with a double series of tubercles on each side, inner much smaller ; vertical suture not very distinct ; pores large, nearly circular, only shghtly oblique. Abactinal system about .46 of horizontal diameter of test, nearly circular and clearly defined, flat or slightly elevated, well covered with small secondary spines ; genital plates large and nearly square, outer margin a little convex and inner equally or more concave, with pores near distal margin ; ocular plates tri- angular or with inner angle truncated when plate is fully insert, as wide as high or wider, with pores very close to distal margin ; in four specimens oculars I, V and IV are insert, while in the largest specimen all are insert, though IV is very narrowly so ; anal system about one-half diameter of abac- tinal system, covered by a marginal series of 8-10 rather large plates and 12-15 (or more) smaller ones within ; all abactinal and anal plates are fairly covered by rather coarse tubercles ; there are 2-10 on the anal plates, 30 or more on the oculars, and 50 or more on the genitals. Actinostome small, only about .37 of horizontal diameter of test, not at all sunken, closely covered by stout plates, 12-15 in each interambu- lacrum and about a dozen pairs in each ambulacrum. Pri- mary spines when fully grown 60-70 mm. long, and 2.5-3.5 mm. thick, cylindrical and tapering rather abruptly to a bluntly pointed tip ; in the largest specimen they are distincly com- pressed near tip ; except when still immature, they are very closely covered by low, rounded granules, which are so closely crowded they do not form longitudinal series as is usual ; the collar is very low, about 1-1.5 mm. high, very finely striate 1. Confertus=cTovrded ; in reference to the densely crowded granules on the primary spines. SEA LILIES, STARFISHES, ETC. — CLARK. 10 1 longitudinally ; neck smooth, shining, porcelain-like ; as the spine gets older, the smooth, shining neck comes to occupy- more and more of the spine, so that in old spines only the distal half or two-thirds shows the crowded granules clearly ; actinal primaries, smooth, flattened, not at all serrulate, with very broad collars ; secondaries, narrow, flat and blunt. Pedicellariae much as in Cidaris affinis, Phil. ; large globi- ferous pedicellarise very infrequent, wanting in some speci- mens ; tridentate pedicellarise rather numerous ; small globi- ferous somewhat less so. General colour of test yellowish- brown or dull reddish ; secondary spines whitish or yellowish, the larger ones about the primaries with a longitudinal stripe on the outer side, of brownish, greenish or yellowish tint ; this stripe may be very faint and is often wanting, but is probably characteristic of the living animal ; secondaries often tipped with yellowish or reddish ; primary spines pure white (more or less soiled and encrusted distally) with yellow-brown collar ; in the smallest specimen many of the primaries are encircled by two narrow, widely separated, rather faint rings of dull rose-red. The smallest specimen is 25 mm. in diameter, the largest 45 mm. ; the latter is 33 mm. high. Five speci- mens. At first glance these specimens look very much hke Cidaris cidaris of Europe or C. abyssicola of Florida, but on more careful examination the relationship to C. afflnis becomes more evident. I am no longer willing to separate affinis from cidaris generically, for in large series of specimens it is difficult to separate them even specifically. I have therefore aban- doned the genus Tretocidaris in the sense in which I formerly used it, and I am the more ready to do this because I think the " splitting " of genera has been already carried too far in the Cidaridse, and we need now to do a little consolidating and redefining. The occurrence of the present species off South-eastern AustraHa is very interesting, for no nearly related species has been taken by the " Challenger," " Siboga," " Valdivia " or " Albatross " in their various collections in the Indian and Pacific Oceans. The " Challenger " species, C. bracteata, recentlyfoundoff Western AustraHa, is as near an ally as any, but the difference in the primary spines is very marked. Indeed, the crowded condition of the granules on the primaries of C. conferta, and the lack or faint indication of longitudinal series in their arrangement, readily distinguish the present species from any cidarid with which I have compared it. iocs.— Eastern Slope, Bass Strait, 80-200 fathoms. South of Gabo Island, Victoria, about 200 fathoms. 102 " ENDEAVOUR " SCIENTIFIC RESULTS. Genus Goniocidahis, L. Agassiz and Desor. GoNiociDARis CLYPEATA, DoderUin. Goniocidaris clypeata, Doderlein, Arch. f. Naturg., li. i., 1885, p. 82. It is only after much hesitation that I refer these Httle Echini to this Japanese species, but were they from Sagami Bay, Japan, I should do so without question. All are small, and only eight show the characteristic clypeate spines, but there is at least one of these from each of the stations hsted below. All the specimens agree in a greenish colouration, or at least in having a greenish tinge, particularly on the actinal primaries and in lacking the big globiferous pedicellariae characteristic of the two following species. The largest specimen is 20 mm. in horizontal diameter and comes from off Cape Pillar, Tasmania. I have compared it with a specimen of the same size from Japan, and while there are obvious differences, they are so trivial that I cannot find a single good character, nor any tangible combination of characters, which would justify calling the two specimens by different specific names. The " Siboga" took, in the East Indies, south of the equator, a cidarid which de Meijere called G. hirsutispinus but which I beheve to be G. clypeata, so that the occurrence of this species in both Japanese and Tasmanian waters is not so improbable as at first appears. Fourteen specimens. Logs. — Off Port Davey, Tasmania, 88 fathoms. Near Storm Bay, Tasmania. North-east of Cape Pillar, Tasmania, 80 fathoms. Twenty miles east of Maria Island, Tasmania, 128 fathoms. Oyster Bay, Tasmania, 20-40 fathoms. East of Babel Island, Bass Strait, 65-70 fathoms. Goniocidaris geranioides (Lamarck). Cidarites geranioides, Lamarck, Anim. sans Vert., iii., 1816, p. 56. Goniocidaris geranioides, Agassiz and Desor, Ann. Sci. Nat., Zool., (3), vi., 1846, p. 337. These specimens are nearly all small, only four exceeding 25 mm. in diameter. The largest, 37 mm. in diameter, is a good example of the species and shows the distinguishing characters well. The two other adults (30 and 32 mm.) are less typical, but are undoubtedly G. geranioides. Many of the SEA LILIES, STARFISHES, ETC. — CLARK. 103 smaller specimens approach G. tuharia and the very young are distinguishable only with some difficulty. The differences between the preceding species, G. clypeata, and young G. geranioides are also very slight, but they seem to be constant ; in the Tasmanian specimens of G. clypeata there are no large globiferous pedicellarise, while these are generally very com- mon in G. geranioides, and the actinal primaries are more or less greenish, which is not the case in G. geranioides. The growth changes in G. geranioides, as revealed by the present series are very interesting. Young specimens have either long, slender and often perfectly smooth primaries {i.e., without thorns and prickles) or they are short, moderately stout and with a few coarse thorns. As the individual matures, the spines become more and more stout and the abactinal ones become expanded at the tip ; they are rough, but not spiny, the coarse thorns which may have been present in youth, wearing down more or less markedly. Meanwhile the miliary tubercles and spines encroach more and more on the median ambulacral area, until the adult ladder-Uke arrangement of transverse ridges is perfected. As a result of these changes the mature specimens (30 mm. and over) look very unlike the young. Thirty -one specimens. Lacs. — Oyster Bay, Tasmania, 20-40 fathoms. North-east of Cape Pillar, Tasmania, 80 fathoms. Seven miles north-east of Cape Pillar, Tasmania, 50-60 fathoms. Near Storm Bay, Tasmania. Off Port Davey, Tasmania, 88 fathoms. South-west of Rocky Point, Tasmania. Near Mainwaring Cove, Tasmania, 50 fathoms. Forty miles west of Kingston, South Austraha, 30 fathoms. Off Cape Marsden, Kangaroo Island, South Australia, 17 fathoms. South of St. Francis Island, 35 fathoms. GoNiociDARis TUB ARIA (Lamarck). Cidarites tuharia, Lamarck, Anini. sans Vert., iii., 1816, p. 57. Goniocidaris tuharia, Liitken, Vid. Med., 1863 (1864), p. 137. This is a very good series of this well-known and character- istic species. The eighty-eight specimens range in size from 10 to 50 mm. (horizontal diameter). The primary spines seem to reach their maximum length, 40-50 mm., 104 " ENDEAVOUR " SCIENTIFIC RESULTS. when the individual is about half grown (test diameter, 25 mm.), after which they may increase greatly in thick- ness, spininess and general bizarre appearance ; they may decrease in length, probably from attrition, for the largest specimen in the present series has no primarj'^ 25 mm. long. This specimen, by the way, comes from the same station (and bears the same registered number), as the largest and most typical example of G. geranioides ; in colour it is somewhat like the latter, as well as in the character of the primary spines, but the test is that of a perfectly typical G. geranioides. I confess to being uncertain as to the true interrelationships of the Australasian species of Goniocidaris. It will be noticed that all the specimens of G. clypeata taken by the " Endeavour " are from southern and eastern Tasmania ; those of G. geranioides are from southern and western Tasmania and South Austraha ; while the range of G. tubaria is from northern New South Wales southward and westward to western South Austraha, barely touching Tasmania. The specimen listed from Oyster Bay, Tasmania, is, however, a perfectly typical G. tubaria. Unfortunately we know very little of G. umbraculum from New Zealand, and only three old, bare tests are available for my study. After ex- amination of the evidence I can secure, I am inclined to think that we shall ultimately consider G. tubaria and G. geranioides as one species (for which the latter name will be used) but separate subspecies, while the form here hsted as G. clypeata will prove to be identical with G. umbraculum, which will in turn be considered a sub-species of G. geranioides. The Aus- tralasian forms of Goniocidaris would then hst as follows : — G. geranioides (Lamarck). — Southern Tasmania and west- ward to the Great Austrahan Bight. G. g. tubaria (Lamarck). — Southern coasts of continental Australia, from northern New South Wales to central Western Australia. G. g. umbraculum (Hutton). — Eastern Tasmania and New Zealand. I think G. umbraculum is undoubtedly distinct from the Japanese G. clypeata, but am unable to give any satisfactory differences at present. Locs. — Eleven miles south-east of Ballina, New South Wales, 27-29 fathoms. Six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Shoalhaven Bight, New South Wales, 15-45 fathoms. Eastern Slope, Bass Strait, 80-200 fathoms. Twenty miles east of Babel Island, Bass Strait, 65 fathoms. SEA LILIES, STAEFISHES, ETC. CLARK. 105 Off Babel Island, Bass Strait, 50-60 fathoms. Oyster Bay, Tasmania, 30 fathoms. Twenty-four miles south-south-east of Eagles Nest Rock, Victoria, 15 fathoms. Forty miles west of Kingston, South Australia, 30 fathoms. Off Cape Marsden, Kangaroo Island, South AustraHa, 17 fathoms. South of St. Francis Island, South Australia, 35 fathoms- Genus Histocidaris, Moriensen. HiSTOCiDARis ELEGANS [A. Agassiz). Porocidaris elegans, A. Agassiz, Proc. Amer. Acad., xiv., 1879, p. 198. Histocidaris elegans, Mortensen, " Ingolf " Ech., pt. 1, 1903, p. 22. This is an interesting quartette, each one having some point of special interest. The smallest is 20 mm. in horizontal diameter (h.d.) and 12 in vertical (v.d.) ; there are only 6 coronal plates in each series ; the primary spines are 50-58 mm. long, and while finely spiny have no coarse thorns ; the test, secondaries and collar of the white primaries are dull, deep gray-purple. The next specimen is 38 mm. h.d. and 25 v.d. ; there are 7 or 8 coronal plates; the primary spines are wonderfully fine, 100-130 mm. long, 3 mm. thick, and very thorny near base, becoming finely spiny near tip ; the colour is like that of the small specimen. The third specimen is remarkable for the size and shape of the test and the number of coronal plates ; h.d. = 71 mm., and v.d. = 61, so that the height is .85 of h.d. ; there are 12 coronal plates and excessive number for a recent cidarid ; the primary spines are all broken or missing (except actinally), but it is evident that none were thorny ; the test is deep yellow-browTi, the secondaries, pedicellariae, etc., dirty whitish ; the general condition of the specimen is poor though the test is perfectly whole. The fourth specimen is remarkable only because of its large size, h.d. = 78 mm., and v.d. =58; v.d. = .75 h.d. ; there are 10 or 11 coronal plates ; the primary spines are as a rule broken, but a number are whole, all are coarsely thorny and the longest probably exceeded 100 mm. It would seem that in this species, as in the preceding species of Goniocidaris, the primary spines reach their full length, 100-130 mm. when the test is about half grown, and that subsequent changes are 106 " ENDEAVOUR " SCIENTIFIC RESULTS. relatively sUght, except as accident or attrition may decrease the length. With reference to the generic name of this species, I think Lambert and Thiery are right in objecting to any recent species being assigned to the genus Porocidaris ; it is very unlikely that any now known really belong there ; certainly none show the distinctive generic feature. On the other hand I think it still worse to put these characteristic recent forms into Pomel's ill-defined and heterogenous group Plegiocidaris, as my esteemed French colleagues do, and I believe the best plan is to follow Mortensen in creating a special genus Histo- cidaris. I differ from him, however, in considering purpuratus as much a member of Histocidaris as is elegans, and hence I am obhged to reject his proposed genus Poriocidaris. In- cluding purpuratus, Histocidaris is one of the most natural genera yet recognised in the family. Locs. — Eastern Slope, Bass Strait, 80-200 fathoms. East of Fhnders Island, Bass Strait, 100-300 fathoms. Family CENTRECHINID^. Genus Astropyga, Gray. ASTROPYGA RADIATA {Leske). Cidaris radiata, Leske, Add. ad Klein, 1778, p. 52. Astropyga radiata, Gray, Ann. Phil., xxvi., 1825, p. 426. The finding of this conspicuous Sea-urchin so far to the south of its previously known range is one of the " Endeavour's " many interesting discoveries. The " Siboga " took specimens near Ceram, north-west from New Guinea, but the genus has not hitherto been found in Torres Strait or anywhere on the Australian coast. The present specimens are 110 and 145 mm. in diameter ; the larger is dull dark red, the primaries, especially the actinal ones, with a greenish cast ; the smaller is reddish-white with actinostome, genito-ocular ring and outer half (or more proximally) of each abactinal interambulacral plate, rose-red ; the abactinal blue spots show up plainly on the smaller specimen, but are much less distinct on the larger and darker one. The actinal primaries of the smaller speci- men are prettily, transversely banded. Two specimens. Loc. — Twenty -five miles south-east of Double Island Point, ■Queensland, 33 fathoms. SEA LTLIES, STARFISHES, ETC. CLARK. 107 Genus Ch^todiadema, Mortensen. Ch^todiadema granulatum, Mortensen. Chcetodiadema granulatum, Mortensen, Vid. Med., 1903, p. 1. This is another East Indian species, not previously known from south or east of Torres Strait, whose range the " Endea- vour " has found to extend to southern Queensland. The " Challenger " took some small specimens west of Torres Strait, which were recorded as Astropyga pulvinata. The present specimen is larger than any hitherto recorded, measuring 115 mm. h.d. Its colour is also different, the abactinal surface being brownish-ohve and the primary spines pink ; orally, test and spines are whitish, the primaries with a green tinge ; the abactinal blue spots are conspicuous and are distributed as described by Mortensen. Loc— Twenty miles north-east of Double Island Point, Queensland, 29-30 fathoms. Family ECHINOTHURID^. Genus Ar^osoma, Mortensen. Ar.^osoma thetidis {H. L. Clark). Asthenosoma thetidis, H. L. Clark, Bull. Mus. Comp. Zool. Harvard, lii., 1909, p. 134. Arceosoma thetidis, A. Agassiz and H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xxxiv., 1909, p. 176. These specimens are 115-180 mm. h.d. and have the general appearance of those collected by the " Thetis." It is too bad that there is no knowledge of where most of these speci- mens were taken, for the range seems to be quite restricted. x\pparently it is the only echinothurid in the shallower coastal waters of Australia. Log. — South of Gabo Island, Victoria, about 200 fathoms. Nine specimens. Also eight specimens from an unknown locahty. Family ARBACIIDiE. Genus Coelopleurus, L. Agassiz. COELOPLEURUS AUSTRALIS,! Sp. nOV. (Plate xliii., fig. 1.) Test 40 mm. h.d. and 24 mm. v.d. ; height therefore is .60 h.d. Coronal plates 11 or 12 in a column, but only the 1. Australis= southern, in reference to its being the most southern representative of the genus. 108 " ENDEAVOUR " SCIENTIFIC RESULTS. lowest 7 or 8 carry primary tubercles and spines ; interambu- lacral areas abactinally only about 8 mm. broad. Ambulacrall plates 12, each with a large primary tubercle ; ambu- lacra 13 mm. wide at ambitus, where interambulacra are about 11, but only 8 mm. abactinally where interambulacra are 8. Primary tubercles large and smooth, without crenula- tion or perforation, of course. Abactinal system about 11 mm. across; oculars all exsert ; anal system, 5.5 mm. across, covered by 4 similar, triangular plates. Actinostome some 16 mm. across, bare, except for the ring of buccal plates, close around the mouth which bear numerous, ophicephalus pedi- cellariae. Primary spines, when fully developed, 85 mm. long, distinctly curved, tapering to a blunt point, which is nearly always broken off ; collar about 11 mm. long or about .13 of spine-length ; it is very slightly further advanced on the dorsal keel of the spine than on the lower surface ; on each side of the keel on the collar are two sKght but distinct longi- tudinal ridges, and there are six such parallel ridges on the lower side of the collar ; on most of the primaries above the ambitus the collar is much longer, .30-. 45 of spine-length, or the uppermost spine or two is all " collar," i.e., has no smooth,, polished surface. Actinal primaries very fiat, thin, sUghtly curved, abruptly truncate, about half collar. Secondary spines, cylindrical or terete, stout and blunt, 4-5 mm. long. Pedicellariae abundant; the valves of the ophicephalous are not at all constricted as in C. floridayius, but on the other hand are not so wide and blunt at tip as in C. longicollis. Colour of test much as in C. floridanus and C. longicollis, light brownish, or dirty cream-colour, with ocular plates, outer borders of genital plates, outer portion of abactinal interambulacral plates, and a narrow, zigzag, median ambulacral line, vermil- lion red ; inner portion of abactinal interambulacral plates more or less lavender or rose-purple (according to dryness and conditions of lighting) ; secondary spines abactinally reddish- brown ; uppermost primaries and collars of the others very light gray-brown with, often, a greenish tinge ; in mature primaries the polished portion is light apple-green adjoining the collar, on upper surface, but verj^ quickly passes into bright red (not abruptly, however) ; at the very tip of the spine may be one or two rather broad transverse bands of Ught green ; lower surface of primaries, pure white ; actinal primaries white, with collar pale grayish or gray-brown. This is a very typical Coelopleurus and closely related to C. floridamis and C. longicollis, but it differs from both in the character of the dorsal interambulacra. The difference is marked when specimens are compared, but it is less easy to SEA LIX,[ES, STAKFISHES, ETC. — CLAEK. 109 'describe. Perhaps it may be most easily stated thus : in C. floridanus and C. longicollis, the genital plates are about the same width as the bare interambulacral space, while in C. australis they are about the same width as the entire inter- ambulacrum. In C. australis, the outer half of the upper •interambulacral plates, which lack a primary tubercle, is well covered by a couple of secondary and 10-20 mihary tubercles ; in C. longicollis there are very few such tubercles on those plates ; in C. floridanus, the tubercles are numerous, but are very small, so the surface of the outer end of the plate is granular, rather than tuberculous. The Austrahan species resembles C. longicollis in the large size of the dorsal ambu- lacral tubercles ; they are even larger than in that species. It is also like C. longicollis in the characters of fully developed primary spines, but the collar is lower and the colouration is totally different. Coelopleurus has been taken at the Kei Islands and at New Britain, but its occurrence in Bass Strait extends the range of the genus over two thousand miles at least. -Lac. — Eastern Slope, Bass Strait, 60-112 fathoms. Family ECHINIDiE. Genus Echinus, Linne. Echinus horridus, A. Agassiz. (Plate xxxix. ; Plate xl., fig. 1-2.) Echinus horridus, A. Agassiz, Proo. Amer. Acad., xiv., 1879, p. 203. The specimen from south of Gabo Island, Victoria, is not only the most remarkable Echinoid in the collection, but is one of the most extraordinary Sea-urchins which has ever been taken. It has a horizontal diameter at ambitus of 80 mm. and a vertical height of 115 mm. ; there are 33 or 34 coronal plates and 61-62 ambulacrals, in each column ; the abactinal system is 12 mm. across (.15 h.d.) and all the ocular plates plates are exsert ; the anal system is nearly 5 mm. across, and is covered by about 160 small plates, many quite granule- like, among which the flat, circular suranal is very prominent ; the spines are nearly all missing ; those which remain are small secondaries or the bases of broken, bright red primaries ; the actinostome is very small, only 11 mm. across, or less than .15 h.d. I have compared this specimen with the fragments of a slightly larger specimen of E. horridus from off southern 110 "endeavour" scientific results. Chili, a cotype of Mr. Agassiz's species. The only difference of importance is in the covering of the anal system, which in the Chilian specimen consists of fewer and larger plates. This may prove to be a good specific character, but only more abundant material can determine the point, and meanwhile the Australian specimen may best be considered as E. horridus, the resemblance in colour, shape and character of the test and spines being so very marked. The specimen from east of Fhnders Island, Bass Strait, is very much smaller, and is of great interest as indicating that the extraordinary increase in length of the vertical axis occurs only after the individual is half grown. The horizontal dia- meter of this specimen is 41 mm. while the vertical is 32 ; v.d. is thus a little more than three-fourths h.d. ; the abactinal system (8 mm.) and actinostome (7.5 mm.) are relatively a little larger than in the adult, but are notably small. There are 23 coronal plates, indicating that while the height of the test is increasing over 250 per cent., the number of coronal plates increases less than 50 per cent. ; obviously the increase in height is due to marked increase in size of the coronal plates rather than to great increase in number. The colour of this smaller specimen is hke that of the adult, reddish with bright red primary spines. The specimen from off Maria Island, Tasmania, is obviously young. It is partly broken, and lacks the abactinal system entirely. It is 17 mm. in diameter, with an actinostome 7 mm. across and has 15-16 coronal plates in each column. There are no spines left. The tuberculation of the distinctly reddish test is much coarser relatively, if not actually than in the older specimens ; in particular there are several secondary tubercles on each interambulacral plate which are very nearly equal to the primary tubercle in size. It is pos- sible that this specimen is not a young E. horridus, but when locality and depth are considered it seems highly probable that it is. The developmental stages of this species would certainly be of extraordinary interest, and it is to be hoped that at some future day an oceanographic survey of the region between Tasmania and New Zealand and between the latter country and Chili may bring to light the necessary material. Three specimens. Logs. — South of Gabo Island, Victoria, 100-250 fathoms. East of Flinders Island, Bass Strait, 100-300 fathoms. East-north-east of Maria Island, Tasmania, 125-180 fathoms. SEA LILIES, STARFISHES, ETC. CLARK. Ill Genus Parechinus, Mortensen. Parechintjs notius,! 5^. nov. (Plate xli., fig. 1-3.) Test 28 mm. h.d. and 18 v.d. ; height, therefore, is .64 h.d. Coronal plates thick, rough, though not actually sculptured, 20 in a column, each with a primary tubercle whose areola is about half the height of the plate (in midzone) and on each side of which are two secondary tubercles, not much smaller than primary ; the 5 tubercles form a horizontal series ; above midzone on a few plates the secondary tubercles are somewhat more numerous and less regularly arranged, but the uppermost 4 or 5 plates have only 3, 2, 1 or 0 ; inter- ambulacra about 12 mm. broad at ambitus. Ambulacral plates 22, but little lower than the interambulacral, each (in midzone) with a primary and 1 large secondary tubercle ; the latter is near the inner end of the plate and disappears abactinally ; ambulacra 7 mm. wide at ambitus ; poriferous zones narrow (1 mm. wide), the arcs of pores vertical. Abactinal system small and compact, 7.5 mm. across ; all oculars exsert but I nearly in ; anal system large (4 mm. across) nearly circular, covered by numerous (about 30) plates, among which the nearly circular suranal is very distinct and much the largest ; each genital plate carries 3 or 4 secondary tubercles, each ocular, 1. Actinostome rather large, 11 mm. across ; gill shts very slightly indicated ; buccal membrane thin and perfectly bare, except for the 5 pairs of primordial ambulacral plates, which are very near the mouth, but do not form a closed ring ; they are densely covered with ophi- cephalous pedicellarise. Primary spines 7-8 mm. long, abruptly and not sharply pointed, each with about 10 longitudinal ridges, low, rounded and close together ; secondary spines similar but not so long and often somewhat swollen at tip ; mihary spines very few, most of the relatively few mihary tubercles bearing pedicellariae. Globiferous pedicellarise fairly common, very similar to those found on the New Zealand form {albocinctus) of P. magellanicus ; the blade is narrow, slightly widened near tip, with a long and conspicuous end- tooth and a smaller and more slender tooth on the left side, below tip ; base of valve wider than in Mortensen's figure of P. albocinctus ; tridentate pedicellariae very rare or wanting, for I failed to find one ; ophicephalous pedicellariae very abundant ; the valves are broadly rounded at tip and not at all constricted. Colour of test, very pale brown ; spines white. 1. »'oT*os=southern ; in reference to its geographical position. 112 " ENDEAVOUR " SCIENTIFIC RESULTS. I have been much puzzled by this Sea-urchin. It does not look Hke any of the many specimens of P. magellanicus which I have examined, but I am sure it is very nearly related to that species. Indeed it is difficult to point out any im- portant distinguishing features. Perhaps the two most tangible differences are the narrow poriferous areas, with vertical arcs of pores, in P. notius and the noticeable scarcity of mihary tubercles ; in P. magellanicus miliary tubercles are abundant and the poriferous areas are moderately wide with the arcs of pores more or less oblique. The anal system of P. notius looks quite different from that of P. magellanicus, as it is nearly circular and is covered by much more numerous plates. The presence of only one lateral tooth on the globi- ferous pedicellarise, instead of one on each side as in P. magellanicus, and the white primary spines of P. notius are also worthy of mention, though perhaps not reliable specific characters. It may also be mentioned, though I do not lay any weight on the fact, that in P. magellanicus the buccal plates bear numerous tridentate pedicellarise, while in P. notius these plates are covered with ophicephalous. The fact that all oculars are exsert in P. notius is interesting and may ultimately prove to be a good specific character. The general appearance of abactinal system, spines, pedicellarise and buccal membrane is very much that of an Amhlypneustes, but the test is very different, as are the buccal plates. It is a pity the " Endeavour " did not secure an abundance of specimens of Parechinus, for the Australasian species of the genus sorely need elucidation and this single specimen only adds to the difficulties ! Loc. — South of Gabo Island to south-east of Cape Everard, Victoria, 70-80 fathoms. Genus Gymnechinus, Mortensen. Gymnechinus epistichus, H. L. Clark. Gymnechinus epistichus, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xxxiv., 1912, p. 289. This specimen resembles the holotype closely, but differs in two points ; genital 3 (as not infrequently happens in this genus) is completely shut out from the periproct, and the primary spines, especially the actinal ones, are a decidedly more distinct red. The horizontal diameter of the test is 22 mm. Loc. — Twenty-five miles south-east of Double Island Point, Queensland. SEA LILIES, STARFISHES, ETC. — CLARK. 113 Family TEMNOPLEURrD.E. GJenus Salmacis, L. Agassiz. Salmacis oligopora, 1 sy. nov. (Plate xlii., fig. 1-2 ; Plate xliii., fig. 2.) Test 60-81 mm. h.d. ; 29-39 mm. v.d. ; height therefore just less than haK the horizontal diameter. Veiy similar to *S'. dussumieri and *S'. erythracis, but differing in at least three particulars ; relative number of ambulacral and interambu- lacral plates in each column, form of base of valves of globi- ferous jDedicellarise, and colouration. The number of coronal plates in the specimens at hand runs from 28 to 34, while the number of ambulacral plates in a column is from 43 to 55 ; thus the number of ambulacral plates is only .52-. 62 more than the number of coronal plates ; in S. dussumieri and S. erythracis it is .70-. 87 more. The base of the valves of the globiferous pedicellarise is rather higher than wide and the upper lateral corners are elongated to a much greater degree than in any other species of Salmacis, and flare outwards slightly. Finally the colouration is utterly unlike that of S. dussumieri and quite difl'erent from that of S. erythracis. The test is pale brownish ; the miliary spines and pedicellarise are white, the secondary spines are white or pale greenish or tipped with greenish, and the primaries are greenish-white with 1-3 broad, indefinite bands of olive-green ; the actinal primaries may be white without any green or they may be broadly tipped with olive-green ; sometimes there is also a band of oHve green near middle of spine. There is no hint of red or violet anywhere. There are too few specimens of S. dussumieri and S. erythracis at hand to enable me to determine satisfactorily the relationship of the three species to each other. They agree in the peculiar characters of the ambulacra and abactinal system and so form a well marked section of the genus, but it is possible that the characters by which I have separated the three forms from each other are of less significance than I suppose. One of the specimens of S. oligopora from off Sandon Bluffs, New South Wales, has a remarkably deformed abac- tinal area ; the anal and genital plates are sunk much below the abactinal part of the ambulacra, and the latter are bent abruptly downwards and inwards to them ; thus the vertical diameter of the test at the distal portion of jjenital I is 39 mm., 1. oX4'yov=few-}-7ro'/)os-=a way through, a pore; in reference to the relatively small number of ambulacral pores. 114 " ENDEAVOUR " SCIENTIFIC RESULTS. while at the third plate of ambulacrum IV it is 46 mm. ; the sixth or seventh ambulacral plate is about on a level with the genital pores. Another specimen is remarkable in that four genitals are developed as madrepo rites ; to judge from the position of the anus they are genitals 1, 2, 4 and 5 ; each genital plate except 2 has two genital pores, so there are nine such pores altogether. In neither of these exceptional specimens is there any ocular insert. The same is true of the largest normal specimen, but there are 6 specimens in which ocular I is more or less broadly insert. Eight speci- mens. Logs. — Oyster Bay, Tasmania, 20-40 fathoms. This speci- men has been selected as the holotype. Six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. Salmacis sph^roides {Linne). Echinus sphceroides, Linne, Sys. Nat., 10th ed., 1758, p. 664. Salmacis sphceroides, Loven, Ech. Linn., 1887, p. 69. These specimens range in diameter from 40 to 80 mm.; the height ranges from .65 to .75 h.d. The colouration is excep- tionally red, the actinal primaries in particular being bright, almost vermilion, red with several bands of white ; abacti- nally, however, the primaries, even in the reddest specimen, are green or greenish at base. Seven specimens. Log. — Twenty miles north-north-east of Double Island Point, Queensland, 29-30 fathoms. Salmacis virgulata {L. Agassiz and Desor). Salmacis virgulata, L. Agassiz and Desor, Ann. Sci. Nat., Zool., vi., 1846, p. 359. The specimens without definite locality are large, typical examples of this species, showing little or no trace of the horizontal sutural furrows of S. alexandri : they measure 75 and 90 mm. h.d. respectively. The specimen from off Double Island Point, Queensland, is much smaller (only 48 mm. h.d.) and has the sutural furrows well marked ; its most striking feature, however, is the colouration of the primary spines, which at first glance seem to be whitish with SEA LILIES, STARFISHES, ETC. — CLARK. 115 a broad indistinct band of purple around the middle ; the specimen is in poor condition, however, and I think the colour of the spines has been altered ; I beheve they were, as usual, greenish at base, violet or purple distally and with an indistinct light tip ; in the process of preservation the green tint has been nearly completely obhterated. The occurrence of typical S. virgulata on the coast of southern Queensland necessitates the acceptance of Doder- lein's view that S. alexandri is at best only a subspecies of S. virgulata, and I am the more ready to do this because the Museum of Comparative Zoology has now a typical specimen of S. alexandri from Badu Island, Torres Strait. Obviously the geographical ranges of the two forms are by no means distinct. Three specimens. Logs. — Off the coast of Queensland. Twenty-five miles south-east of Double Island Point, Queensland. Genus Microcyphus, L. Agassiz and Desor. MiCROCYPHUS ANNULATUS, Mortcnsen. Microcyphus annulatus, Mortensen, Siam Ech., 1904, p. 101. This is a typical example of this species, 11 mm. in •diameter. Loc. — Seven miles north-east of Cape Pillar, Tasmania, 50-60 fathoms. Genus Amblypneustes, L. Agassiz. Amblypneustes grandis, H. L. Clark. Amblypneustes grandis, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xxxiv., 1912, p. 329. These are all large specimens, 50-63 mm. in diameter, with the height ranging from .60 to .76 h.d. In colouration there is considerable diversity. The specimens from an unknown locality are light brown, with the small spines whitish and the primaries greenish-brown, tipped with whitish ; in one specimen the greenish-brown is confined to the basal part of the spine, especially on the actinal surface. The speci- mens from Flinders Island, Bass Strait, are similar but more green ; the test is olive-greenish and the basal part of the primaries is more green than brown ; in one specimen the primaries are nearly all white. The specimen from between Gabo Island, Victoria, and Disaster Bay, New South Wales, 116 "endeavour" scientific results. is similar to the latter specimen, the primaries being almost wholly white, though the test is distinctly greenish. The South Australian specimen is the most peculiar, for \\'hile the test is greenish and not very unusual, the primary spines are all pale reddish with light tips ; actinally the hght shade occupies more of the spine than the reddish tint. All the specimens show a zigzag band (similar to that of A. formosus) in the abactinai interambulacra, indistinct when the test is dry but very evident when the surface is moistened ; the markings differ obviously from those of A. formosus, how- ever, by being darker than the intervening angles. Doderleini has recently reviewed anew the species of Amhlypneustes, adding another species and a variety to the already perplexing list. He is inclined to doubt the vaHdity of both the present species and the next, considering them forms of A. grisens, but I am unable to agree with him. I confess, however, that I do not believe we have yet reached the truth in regard to the species of this interesting genus. We really have not as yet any clear idea as to the limits of variation in either form, tuberculation or colour, in even a single species, and we zoologists who are thousands of miles from the restricted home of the genus, are thus working largely in the dark in our efforts to identify specimens. Some day a zoologist on the ground, or with a careful and enthusiastic collector on the ground, will secure sufficient material to enable him to work out the actual natural limits of each species. Seven specimens. Locs. — East coast of Flinders Island, Bass Strait. Between Gabo Island, Victoria, and Disaster Bay, New South Wales, 50-100 fathoms. Off Cape Marsden, Kangaroo Island, South Australia, 17 fathoms. Also two specimens from an unknown locahty. Amblypneustes pachistus, H. L. Clark. Amhlypneustes pachistus, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xxxiv., 1912, p. 327. These individuals range in size from 31 to 45 mm. h. d.; the height ranges from a little over .80 h.d. in the smallest to a little less than .90 h.d. in the largest. The colour is the same in all : test, pale brown when dry, more nearly olive 1. Doderlein — Fauna Siidwest-Australiens, iv., 1914, pp. 460-475. SEA LILIES, STARFISHES, ETC. CLARK. 117 when moist ; small spines, pale cream-colour or whitish ; large spines, deep olive-green or greenish-bro\^'n, more or less extensively tipped with whitish. This is essentially the colouration of the holotype and is apparently the character- istic colouration of the species, which seems to me a perfectly valid one. Four specimens. Locs. — Oyster Bay, Tasmania, East coast of Fhnders Island, Bass Strait. Forty miles west of Kingston, South Australia, 30 fathoms. Genus Holopneustes, L. Agassiz and Desor. HoLOPNEusTES INFLATUS, A. Agassiz. Holopneustes inflatus, A. Agassiz, Bull. Mus. Comp. Zool. Harvard, iii., 1872, p. 56. The smallest individual is 30 mm. h.d. and 23 mm. high, while the largest is 72 mm. h.d. and 65 mm. high. All three specimens are light coloured, the spines both large and small being pale gray or nearly cream-colour. The test is decidedly darker, at least when wet. It is interesting to note that in this species, as in the preceding and in Echinus horridus, the relative height increases with age ; in the present case the increase is from .77 h.d. to a trifle over .90. Four specimens. Locs. — North-east of Cape Pillar, Tasmania, 80 fathoms. Oyster Bay, Tasmania. ' East coast of Flicders Island, Bass Strait. Holopneustes porosissimus, L. Agassiz and Desor. Holopneustes porosissimus, Agassiz and Desor, Ann. Sci. Nat., Zool., vi., 1846, p. 364. This individual is 40 mm. in diameter and 38 mm. high ; the height is thus .95 h.d. The ambulacra are 13 mm. wide and the interambulacra 11.5 mm. The test is dull purple and the spines are dull purpUsh-red, without any green (as is usual) near the base. Loc.—Ofi Cape Marsden, Kangaroo Island, South AustraUa. 118 '"endeavour" scientific results. Family STRONGYLOCENTROTID^. Genus Heliocidaris, L. Agassiz and Desor. Heliocidaris erythrogramma {Valencienes). Echinus erythrogrammus, Valenciennes, Voy. Venus, Zooph., 1846, pi. vii., fig. 1. Heliocidaris eurythrogramma, Agassiz and Desor, Ann. Sci. Nat., ZooL, vi., 1846, p. 371. The specimens from Oyster Bay, Tasmania, are typical examples of the species, about 50 mm. in diameter, dark purpUsh in colour, becoming dull green with a purplish cast when dry. The specimen from Flinders Island, Bass Strait, is 68 mm. in diameter, with the test very pale brownish and the spines hght olive green, as in H. tuherculata. The short- ness of the primaries and the absence of any arcs of pores with more than 7 pairs show it must be referred to H. erythro- gramma. I do not feel satisfied, however, that there are really four species of Heliocidaris on the southern and western coasts of Australia, as Doderlein affirms in his recent revision of the genus,i but I think it quite possible that these four forms will prove to be either a single species or perhaps two. I may here protest against Doderlein's action in putting the Peruvian Echinoid, Coenocentrotus gihhosus, into Heliocidaris. The differences in ambulacra, abactinal system, coronal plates and spines far outweigh the single resemblance in globiferous pedicellarise. If the latter is such a weighty character, the species ought to be put in the family Temno- pleuridse, for the globiferous pedicellarise are surprisingly similar to those of Salmacis and Temnopleurus ! On the other hand, I think Doderlein is right in refusing to include the Japanese species crassispina in Heliocidaris. It is better left in Strongylocentrotus, if one is unwilling to recognise the genus Anthocidaris for its reception. Four specimens. Locs. — Oyster Bay, Tasmania, 30 fathoms. East coast of Flinders Island, Bass Strait. Family CLYPEASTRID^. Genus Clypeaster, Lamarck. Clypeaster AUSTRALASIA [Gray). Echinanthus australasice, Gray, Proc. Zool. Soc, 1851, p. 34. Clypeaster australasice, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xlvi., 1914, p. 32. The smallest of these individuals is 46 mm. long, 42 mm. wide and 14 mm. high ; the largest is 90 x 80 x 30 mm. The 1. Doderlein — Fauna Siidwest-Australiens, iv., 1914, pp. 475-487. SEA LILIES, STARFISHES, ETC. — CLARK. 119 widest specimen is .91 as wide as long ; the narrowest is .82. In the highest specimen the vertical diameter is .41 of the length ; in the flattest it is .29. The colour is some shade of brown, usually with a greenish tinge. Several of the speci- mens are shghtly asymmetrical in one way or another. Ten specimens. Locs. — Six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Shoalhaven Bight, New South Wales, 15-45 fathoms. East of Babel Island, Bass Strait, 65-70 fathoms. Clypeaster telurus, H. L. Clark. Clypeaster telurus, H. L. Clark, Rec. West Austr. Mus., i., 1914, p. 166. The unique holotype of this species was taken on the coast of Western Austraha almost directly across the Aus- tralian continent from Fraser Island, Queensland, but a prolonged critical study of these Clypeasters fails to show any character or combination which will distinguish them from it, except that they do not have the posterior inter- radial margin depressed appreciably. They are all larger than the holotype, the largest measuring 145 x 135 x 20 mm. ; as a species character, it seems to be true that the breadth exceeds .90 of the length, while the height is less than .15. There are about 400 primary tubercles to a square centimeter of the aboral surface. The unpaired anterior petal is 39 mm. long and has 52 pore-jjairs. The periproct is 8-12 mm. from the posterior margin, and this considerable distance, taken in connection with the thin, almost fragile, test, and the form of the petals, makes it easy to distinguish this species from C. humilis, which is its nearest relative. Tridentate pedicel- larise, with valves .25-. 75 mm. long, are common on the oral surface. The colour of these specimens is more or less purplish-brown. Loc. — Off Fraser Island, Queensland. Four specimens. Clypeaster virescens, Doderlein. Clypeaster virescens, Doderlein,, Arch. f. Naturg., li.i., 1885, p. 102. This individual is 105 mm. long, 95 mm. wide and 19 mm. high. The colour is yellow-brown with a distinct greenish tinge orally. Although none of the Japanese specimens at 120 "endeavour" scientific results. hand are quite so large, comparison with them fails to reveal a single character by which this specimen can be distin- guished from them. It is hard to beHeve that a Japanese species, not yet known from the East Indies or northern Austraha, should occur in Shoalhaven Bight, New South Wales, but this specimen seems to prove it. There is no chance for a misplaced label, for in this case the metal number is not only wired on to the test, but the wire is firmly rusted in ! I have tried to convince myself that the specimen is an aberrant example of C. australasice, but if that is the case the characters by which species in the genus are at present separated, are simply worthless, and we might as well call all Australian and all Japanese Clypeasters by a single name ! I have also tried to find at least one character which would serve to distinguish this Austrahan specimen from the Japanese species, but I have not succeeded, and I am therefore obliged to record it as C. tnrescens. Loc. — Shoalhaven Bight, New South Wales, 15-45 fathoms. Family LAGANID^. Genus Peronella, Gray. Peronella lesueuri {Agassiz). Laganum lesueuri, Agassiz, Mon. Scut., 1841, p. 116. Peronella lesueuri, A. Agassiz, Rev. Ech., pt. 1, 1872, p. 148. These are notably large individuals, and it is a pity that there is no record of either the locality or the depth. The largest one is 137 mm. long, 124 mm. wide, and only 12 mm. high ; the petaloid area is nearly .60 of test length, but the marginal thickness of the test is less than .03 of the length ! The smallest specimen is 100 mm. long. There is little diversity in shape ; the narrowest specimen has the width .86 of the length, while in the widest it is .96. Loc. — Off the coast of Queensland. Nineteen specimens. Family FIBULAR IIDiE. Genus Echinocyamus, Leske. Echinocyamus platytatus, H. L. Clark. Echinocyamus platytatus, H. L. Clark, Mem. Mus. Comp. Zool. Harvard, xlvi., 1914, p. 63. There is no doubt of the genus of these small pieces, but the reference to E. platytatus involves more uncertainty. I think it safe, however, to consider them that Victorian sp«^f^ies. Loc. — Between Devonport and Launceston, Tasmania. SEA LILIES, STARFISHES, ETC. CLARK. 121 Family SPATANGIDiE. Genus Maeetia, Gray. Mahetia peloria,! sy. nov. (Plate xliv., fig. 1-3.) Length of test, 76 mm. ; breadth, 65 mm. ; height, 18 mm. ; abactinal system 33 mm. from anterior margin ; mouth, 12 mm. Avide by 4 mm. long, with a very prominent hp, its anterior margin 20 mm. from test margin ; periproct, 9 mm. in diameter, on the vertical posterior end of test ; subanal fasciole, distinct but narrow, its area 24 mm. wide by 4 mm. high at middle ; bare sternal area nearly 30 mm. wide posteriorly. Unpaired petal indistinct, virtually wanting. Anterior petals very imperfect, but the posterior series of pore- pairs can be distinguished for about 25 mm. from apex. Posterior petals fairly well developed, about 35 mm. long and 7 mm. wide just proximal to the very shghtly narrowed tip; there are rather more than 20 pore-paii's on each side of the petal. Anterior ambulacrum, although not petaioid, is plainly indi- cated as a gradually widening area, covered only by mihary spines (or tubercles) ; proximally it is very narrow ; on each side is a series of 12-15 simple tube-feet (or pores), at first close together but near margin 3-4 mm. apart. Genital pores 4, large, but close together and concealed by the dense growth of mihary spines which covers the whole abactinal system and adjoining plates. Remainder of abactinal surface covered closely by more or less curved, slender mihary and secondary spines, pediceUarise and scattered primaries ; the latter are nearly aU broken, but when complete are 15-20 mm. long and taper to a very slender tip ; proxunally they are hollow ; they are more or less distinctly cm-ved near the base. In each antero-lateral interambulacrum there are about 35 primaries above the ambitus ; each is attached to a large, perforated tubercle placed in a deeply sunken areola ; in each postero -lateral interambulacrum there are some 50 similar primaries, and in the posterior interambulacrum there are about 25, but these latter are as a rule smaUer than the primaries of the lateral areas. Actinal surface largely free from spines and tubercles of any considerable size, except laterally and near anterior L 7r6\a'/j(os = huge, immense; in reference to the size, remarkably large for a Maretia. 122 " ENDEAVOUR " SCIENTIFIC RESULTS. margin. In the antero -lateral interambulacra the margin is densely clothed with secondary spines and 15-20 rather small primaries ; there are also a few secondaries close to the mouth. In the lateral (or postero-lateral) areas the margin is densely coated with secondary and mihary spines and inwardly with 7 or 8 obhque series of large primaries, 4-5 in each series. Within the subanal fasciole there is on each side a group of 10-12 primary spines, and below it is a triangular area (with apex towards mouth) well covered with secondary and mihary spines. All the rest of the oral surface seems bare, but is in reality sparsely covered with minute miliary spines and pedicellariae. The hp, however, bears a dense cluster of secondary spines, and there are smaller and less dense clusters on each side of the mouth, while similar spines are borne on the peristomal plates themselves. Phyl- lodes conspicuous, though in each one of the posterior pair there are only 7 of the ambulacral " brushes." Peristome covered by about 50 plates. Periproct covered by about 60 plates, each with 1-4 miliary spines. Pedicellariae abundant, of 4 kinds, each of which has 3 valves to make up the head ; the valves of the globiferous pedicellariae are about 1 mm. long, the terminal third or more encased in the glandular tissue ; these valves resemble those of Lovenia elongata as figured by Koehler^, but the blade narrows steadily to the truncate tip which terminates in 4 rather short but distinct teeth, lying practically in the same plane ; the valves of the tridentate pedicellariae are about .80 mm. long and resemble those of Spaiangus ; they are strongly curved, meeting only at tip ; the base is higher than wide, the blade is less than half the length, and its own width (which is greatest distal to its middle) is less than half its own length ; its margin is smooth, becoming finely serrate near tip ; the valves of the ophicephalous pedicellariae are only about .40 mm. long ; they resemble very closely those of Maretia elevata as figured by Doderlein^ ; the valves of the triphyllous pedicellariae are very small, about .20 mm. long, and are borne on the end of a very long neck ; they resemble very closely those of Metalia maculosa as figured by Koehler^. The triphyllous pedicellariae seem to occur every- where on the test, the ophicephalous are mainly if not alto- gether abactinal, the tridentate occur on both the upper and lower surfaces, but seem more abundant actinally, and the globiferous are actinal, occurring chiefly near the mouth. 1. Koehler — Indian Mus. Spat., 1914, pi. xix., fig. 27. 2. Doderlein — Valdivia Ech., 1906, pi. xlviii., fig. 6a-d. 3. Koehler — Indian Mus. Spat., 1914, pi. xix., fig. 70. SEA LILIES, STARFISHES, ETC. CLARK. 123 Colour, in alcohol or dry, pale purplish gray, the muscles around the bases of the primary spines, deep brown ; small spines nearly white when dry ; primary spines more or less pale purple or hght gray ; ambulacral " brushes " around mouth, nearly black. Although resembling Maretia ovata (Leske) {=planulata, auct.) in form of test and in its generic characters, this species is easily distinguished by the longer and narrower posterior petals, the less complete anterior petals and the greater number of abactinal primarv^ tubercles, particularl}- on posterior half of test. The primary spines are somewhat shorter, and those of the actinal surface are scarcely half as- numerous as in M. ovata. The pedicellariae also offer some striking differences between the two species. The occurrence of the genus in Bass Strait is very interesting, for it is essentially a shallow water, tropical group, and has been reported previously from only as far south as Port Jackson, New South Wales. While no depth is recorded for the station where 31. peloria was taken, there is good reason for believing it was in the neighbourhood of 100 fathoms, for there is a station recorded "' twenty-five miles south-west of Cape Everard, Victoria, 83-98 fathoms." In the Dutch East Indies, M. ovata was taken by the " Siboga " at one station in 83 fathoms, but as a rule M. ovata occurs from low water mark to about 20 fathoms. The largest specimen pi M. ovata I have found recorded is 65 mm. long, but this is unusual, and most specimens are from 30 to 50 mm. Loc. — Twenty-six miles south-west of Cape Everard, Victoria. Three specimens. EXPLANATION OF PLATE I. Jletacrinus cyaneus, sp. nov. Photograph of the holotype from the Eastern Slope, Bass Strait, about 2(M) fathoms. Three-fourths natural size. M ':*■ BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate I. G. Nelsox. Cambridge, U.S.A.. phott EXPLANATION OF PLATE II. Fig. L — Comatula cratera, sp. nov. Photograph of the holo- type from eight miles east of Sandon Bluffs, New South Wales, 35-40 fathoms. Six-sevenths natural size. Fig. 2. — OUgometra zebra, sp. nov. Photograph of the holo- type from eleven miles east-south-east of Clarence River mouth, New South Wales, 35-36 fathoms. Natural size. 4 BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate II. E. A. Briggs, Austr. Miis., photi EXPLANATION OF PLATE III. Eig. 1. — Oreometra pericalles, sp. nov. Photograph of the holotype from thirteen miles north by west of Double Island Point, Queensland, 25-26 fathoms. Natural size. Fig. 2. — Comanthus perplexum, sp. nov. Photograph of the holotype from eleven miles south by east of Balhna^ New (South Wales, 27-28 fathoms. Natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate III. 1 E. A. Briggs, Austr. Mus., photo. EXPLANATION OF PLATE IV. Fig. \.~-rComanthus plectrophorum, sp. nov. Photograph of the holot ype from east of Flinders Island, Bass Strait, 100-300 fathoms. Natural size. Fig. 2. — Cosmiometra dasybrachia, sp. nov. Photograph of the holotype from east of FHnders Island, Bass Strait, 70-100 fathoms. Natural size. Fig. 3. — Comanthus spanoschistum, sp. nov. Photograph of the holotype from east of Babel Island, Bass Strait, 60-70 fathoms. Natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate IV. ..^^^^i!s:--&f^ E. A. Briggs, Austr. Mns., jjhoto. EXPLANATION OF PLATE V. Luidia maculata, ]\Iuller and Troschel. Photograph of a specimen from the coast of New South Wales. Abactinal view. One-fourth natural size. BIOL. RESULTS " ENDEAVOUR," Vol. TV. Plate V H. Barnes, Junr., Austr. Mus., photo. EXPLANATION OF PLATE VI. Fig. 1. — Lonchotaster magnificus, sp. nov. Photograph of the holotype from the Great AustraUan Bight, 80-120 fathoms. Abactinal view. Two-fifths natural size. Fig. 2. — Aetinal view of same speciiiien as Fig. L BIOL. RESULTS "ENDEAVOUR," Vol. IV Plate VL E. A Beiggs, Austr. Mus., photo. EXPLANATION OF PLATE VII. Fig. 1. — Mimaster gracilis, sp. nov. Photograph of the holo- type from south of Gabo Island, Victoria, 200 fathoms. Abactinal view. Natural size. Fig. 2. — Actinal view of same specimen as Fig. L BIOL. RESULTS "ENDEAVOUR," Vol. IV Plate VII. "%:?>v4^^ E. A. Briggs, Austr. Mus., iDhoto. EXPLANATION OF PLATE VIII. Fig. L — Nymphaster pentagonus, sp. nov. Photograph of the holotype from the Great Austrahan Bight, 129° 28' E., 250-450 fathoms. Abactinal view. Natural Fig. 2. — Actinal view of same specimen as Fig. 1. BIOT.. RESULTS "ENDEAVOUR," Vol. IV Plate VIII. ^ '4^s.")«:^., E. A. Briggs, Austr. Miis., photo. EXPLANATION OF PLATE IX. Fig. 1. — Mediaster australiensis, sp. no v. Photouraph of the holotype from the east coast of Flinders Island, Bass Strait, 40 fathoms. Abactinal view. Five-eighths natural size. Fig. 2. — Actinal view of same specimen as Fig. L BIOL. RESULTS "ENDEAVOUR," Vol. IV. Plate IX. H. Barnes, Junr., Austr. Mus., jAoto. EXPLANATION OF PLATE X. Fig. L — Mediaster motiacantlms, sp. iiov. Photograph of the holotype from six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Abactinal view. One- half natural size. Fig. 2. — Mediaster monacanthus, sp. nov. Photograph of the paratype from six miles east of Cape Hawke, New South Wales, 47-50 fathoms. Actinal view. One- half natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate X. ' ''^'s:^ H. Barnes, Junr., Austr. Mus., photo EXPLANATION OF PLATE XL Fig. L- — Calliaster spinosus, sp. nov. Photograph of the holo- type from the Eastern Slope, Bass Strait, 80-200 fathoms. Abaetinal view. One-half natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XI .f^ E, A. Briggs, Austr. Mus., photo. EXPLANATION OF PLATE XII. Fig. 1. — Stellaster incei. Gray. Photograph of a specimen from eleven miles south by east of Ballina, New South Wales, 27-29 fathoms. Abactinal view. Nine- tenths natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XTl 'M.' H. Barnes, Junr., Austi-. Mus., photo. EXPLANATION OF PLATE XIII. Fig. 1. — Stellaster inspinosus, sp. nov. Photograph of the holotype from between Geraldton and Cape Natural- iste, Western Austiaha. Abactmal view. Natural size. Fig. 2 — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XIII. E. A. Briggs, Aiistr. Muss., photo. EXPLANATION OF PLATE XIV. Fig. 1. — Fromia polypora, sp. nov. Photograph of the holo- type from east of Maria Island, Tasmania, 78 fathoms. Abactinal view. Two-thirds natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL RESULTS " ENDEAVOUR," Vol. IV. Plate XIV. E. A. Briggs, Austr. Mus., photo. EXPLANATION OF PLATE XV. Fig. 1. — Ophidiaster confertics, sp. no v. Photograph of the holotype from Lord Howe Island, South Pacific Ocean. Abactinal view. Three-fourths natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XV. H. Barnes, Junr., Austr. Mus., photo. EXPLANATION OF PLATE XVI. Fig. 1. — Pseudophidiaster rhysus, sp. no v. Photograph of the holotype from the Great Austrahan Bight, 80-120 fathoms. Abactinal view. One-half natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS "ENDEAVOUR," Vol. IX Pi ATE XVr. E. A. Briggs, Austr. Mus., jjhoto. EXPLANATION OF PLATE XVII. Fig. 1. — Asterina atyphoida, sp. no v. Photograph of the holotyjje from fifteen miles north-west of Cape Jervis, South Austraha, 17 fathoms. Abactinal view. Four times natural size. Fig. 2. — Actinal view of same specimen as Fig. I. BIOL. RESULTS "ENDEAVOUR," Vol. IV. Plate XVII. E. A. Briggs, Austi-. Mus., photo EXPLANATION OF PLATE XVIII. Fig. L — Pieraster tetracanthus, sp. no v. Photograph of the holotype from south-east of Cape Everard to south of Gabo Island, Victoria, 90-150 fathoms. Abactinal view. Natural size. Fig. 2. — Actinal view of same specimen as Fig. L Fig. 3. — Asterina leptalacantha, sp. nov. Photograph of the holotype from Masthead Island, Queensland. Abac- tinal view. Twice natural size. Fig. 4. — Actinal view of same specimen as Fig. 3. Fig. 5. — Crossaster multispinus, sp. nov. Photograph of the holotype from between Gabo Island, Victoria, and Disaster Bay, New South Wales, 50-100 fathoms. Abactinal view. Natiu-al size. Fig. 6. — Actinal view of same specimen as Fig. 5. BIOL. RESULTS "ENDEAVOUR," Vol. IV Plate XVIII. E. A. Briggs, Au.str. Mus., photo. EXPLANATION OF PLATE XIX. Fig. 1. — Echinaster acanthodes, sp. nov. Photograph of the holotype from t\\enty-five miles south-east of Double Island Point, Queensland, 33 fathoms. Abactinal view. Six-sevenths natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XIX H. Barnes, Junr., Austr. Mus., photo. EXPLANATION OF PLATE XX. Fig. L — Echinaster eridanelki, Aliiller and Troschel. Photo- graph of a specimen from Murray Island, Torres Strait. Abactinal view. Natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. EIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XX. H. Barnes, Junr., Austr. Mus., photo EXPLANATION OF PLATE XXI. Fig. L — Echinuster eridanella, 3iuller and Troschel. Photo- graph of a specimen with four rays from Mauritius. Abactinal view. Natural size. Fig. 2. — Echinaster eridanella, Miiller and Troschel. Photo- graph of a specimen with seven rays from Bowen, Queensland. Abactinal view. Natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXI. H. Barnes, Junr., Austr. 'Shis., photo. EXPLANATION OF PLATE XXII. Echinaster glomeratus, sp. nov. Photograph of the holotype from off (Upe Marsden, Kangaroo Island, South Australia, 17 fathoms. Abactinal view. 'I^wo- thirds natural size. BIOL RESULTS " ENDEAVOUR," Vol. IV. Plate XXII. H.' Barkes, Junr., Austr. Mus., photo. EXPLANATION OF PLATE XXIII. Echimister ylomcrattis, sp. nov. Photograph of the holotype from off Cape Marsden, Kangaroo Island, South AustraUa, 17 fathoms. Actmal view. Two-thirds natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV Plate XXIIL H. Barnes, Jimr., Austr. Mus., photo. EXPLANATION OF PLATE XXIV Echinaster superhus, sp. uov. Photograpli of the holotype from Biuoine, Western Australia. Abactinal view. Four-fifths natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXIV, H. Barnes, Junr., Austr. Mus., photo. EXPLANATION OF PLATE XXV Echinaster superbus, sp. nov. Photograph of the holotype from Broome, Western Australia. Actinal vicAV. Four-fifths natural size. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXV H. Barnes, Juiir., Austr. Mus., photo. EXPLANATION OF PLATE XXVI. Fig. 1 . — Plectaster decanus (Miiller and Troschel). Photograph of a specimen from off George Head, Port Jackson. New 8outh Wales. Abactinal view. One-half natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS "ENDEAVOUR," Vol. IV. Plate XXVI H. Barnes, Junr., Aiistr. Mns., photo. EXPLANATION OF PLATE XXVII. Fig. 1. — Odinia australis, sp. nov. Photograph of the holo- type from south-east of Cape Everard, Victoria, 200 fathoms. Abactinal view. Natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. Fig. 3. — Pedicellaster reticulatus, sp. nov. Photograph of the holotype from east of Maria Island, Tasmania, 78 fathoms. Abactinal view. Twice natural size. Fig. 4. — Actinal view of same specimen as Fig. 3. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXVII. E A. Briggs, Austr. Mus., photo. EXPLANATION OF PLATE XXVIII. Fig. I. — Zoroaster macracantha, sp. no v. Photograph of the holotype from the Great Australian Bight, 129° 28' E., 250-450 fathoms. Abactinal view. About one-half natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXVIII. E. A. Beiggs, Austr. Mus., photo. EXPLANATION OF PLATE XXIX. Fig. L — Cosmasterias dyscrita, sp. nov. Photograph of the holotype from south of Gabo Island, Victoria, 200 fathoms. Abactinal view. Natural size. Fig. 2. — Actinal view of same specimen as Fig. I. BIOL, RESULTS "ENDEAVOUR," Vol. IV. Plate XXIX. E. A. Briggs, Austr. Mus., photo. EXPLANATION OF PLATE XXX. Fig. L — Coscinasterias dubia, H. L. Clark. Photograph of a specimen from the southern coast of South Austraha. Abactinal view. One and one-half times natural Fig. 2. — Actinal view of same specimen as Figw L BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXX. H. Babnes, Junr., Austr. RIus., photo. EXPLANATION OF PLATE XXXI. Coscinasterias gemmifera, Perrier. Photograph of a specimen froii) Oyster Bay, Tasmania, 20-40 fathoms. Abac- tinal view. One-half natural size. BIOL. RESULTS "ENDEAVOUR/' Vol. IV. Plate XXXI •/^^'^:-' H. Barnes, Junr.. Austr. Mus., photo. EXPLANATION OF PLATE XXXII. Coscmasterius gernmifera, Perrier. Photograph of a specimen from Oyster Bay, Tasmania, 20-40 fathoms. Actinal view. One-half natural size. BIOL. RESULTS " ENDEAVOIJR," Vol. TV. Plate XXXIl H. Barxes, Junr., Austr. Mus., photo. EXPLANATION OF PLATE XXXIII. Fig. L — Ophiocreas phanerum, sp. nov. Photograph of the holotype from twenty miles off Cape Barren, Cape Barren Island, Tasmania, 70 fathoms. Abactinal view. Two-thirds natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR/' Vol. IV. Plate XXXllL E. A. Briggs, Austr. Mus., photo EXPLANATION OF PLATE XXXIV. Fig. 1. — Asirodendrum pustulatum, sp. no v. Photograph of the holotype from east of Fhnders Island, Bass Strait, 100-300 fathoms. Abactinal view. Natural size. Fig. 2. — Actinal view of same specimen as Fig. L BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXXIV E. A. Briogs, Austr. Mus., photo. EXPLANATION OF PLA^J^E XXXV. Eig. 1. — Aslrotluimnus rugosus, sp. nov. Photograph of the h()l()t>i)e fioiii east of Flinders Island, Bass Strait, 80-,']0() fathoms, i^bactinal view. One and one- third times natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXXV -•---rf. ".-ii;**'' E. A. Br.iGo.s, Aiistr. Miis., photc EXPLANATION OF PLATE XXXVI. Fig. 1. — Ophiactis symbioia, sp. nov. Photograph of the holo- type from east of Flinders Island, Bass Strait, lUO- 300 fathoms. Abactinal view. Four times natural Fig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS " ENDEAVOUR/' Vol. IV. Plate XXXVl. E A. Briogs, Anstr. Mus., j^hoto. EXPLANATION OF PLATE XXXVII. Fig. 1. — Amphiophiura colleta, sp. nov. Photograph of the hole type from east of Babel Island, Bass Strait, GO- SO fathoms. Abactinal view. One and five-eighth times natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. BlOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXXVll E. A. Briggs, Austr. Mns., i:)hoto. EXPLANATION OF PLATE XXXVIII. Fig. 1. — Cidaris conferta, sp. no v. Photograph of the holo- type from the Eastern Slope, Bass Strait, 80-200 fathoms. Abactinal view. Two-thirds natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. Fig. 3. — Side view of same specimen as Fig. 1. Fig. 4. — Basal third of a large spine from the specimen shown in Fig. 1. Four times natural size. BIOL. RESULTS "E^DEAVOUB/' Vol. IV. Plate XXXVlIt. E. A. Beiggs, Austr. Mus., lAoto. EXPLANATION OF PLATE XXXIX. Echinus horridus. A. Agassiz. Photograph of an adult speci- men from south of Gabo Island, Victoria, l(X>-2o(> fathoms. Side view. One and one-third times natural size. BlOL. RESULTS " ENDEAVOUR," Vol. IV. Plate XXXlX E. A. Briggs, Austr. Mus., photo. EXPLANATION OF PLATE XL. Fi,.. \.-~Echinus horrklus, A. Agassiz. Photograph of an KK) J^f'rir" from south of Gabo Island, Victoria, J 00-250 fathoms. Abactinal view. One and one- fourth tmies natural size. Fig. 2.— Actinal view of same specimen as Fig. 1. BiOL. RESULTS " ENDEAVOUR,'' Vol. tV. Plate XL. E. A. Briggs, Austr. Mus., pboto. EXPLANATION OF PLATE XLI. Fig. 1. — Parechinus notins, STp.nov. Photograph of the holo- type from south of Gabo Island to south-east of Cape Everard, Victoria, 70-80 fathoms. Abactinal view . Twice natural size. Fig. 2. — Actinal view of same specimen as Fig. 1. Fig. 3. — Side view of same specimen as Fig. 1. BIOL. RESULTS "ENDEAVOUR," Vol. IV. Plate XLI. E. A. Beigos, Austr. Miis., photo. EXPLANATION OF PLATE XLII. Yia 1 —.Sahmtcis oligopora, sp. iiov. Photograph of the holo- ° type from Oyster -Bay, Tasmania, 20-40 fathoms. Abactinal view. Natural size. Eig. 2. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS "ENDEAVOUR," Vol. IV. Plate XLII. E. A. Bkiggs, Austr. Mi;s., i>hoto. EXPLANATION OF PLATE XLIII. Fig. 1. — Coelopleurus australis, sp. no v. Photograph of the holotype from the Eastern Slope, Bass Strait, 60-112 fathoms. Side view. Natural size. Fig. 2. — Sahnacis oligopora, sp.nov. Photograph of the holo- type from Oyster Bay, Tasmania, 20-40 fathoms. Side view. Natural size. BIOL. RESULTS " ENDEAVOUR," Vol. TV. Plate XLIII. E. A. Beiggs, Aiislr. Mus., photo. EXPLANATION OF PLATE XLIV. Fig. 1. — Marelia peloria, sp. no v. Photograph of the para- type from twenty-six miles south-west of Cape Everard, Victoria. Abactinal view. Natural size. Fig. 2. — Side view of same specimen as Fig. 1. Fig. 3. — Actinal view of same specimen as Fig. 1. BIOL. RESULTS "ENDEAVOUR," Vol. IV Plate XLIV. '-4f^ •CtS^ II. Report on the Polychseta obtained by the F.I.S. " Endea- vour " on the coasts of New South Wales, Victoria, Tasmania and South Austraha. PART 11. BY WILLIAM B. BENHAM, M.A., D.Sc, F.R.S., Professor of Biology at the University of Otago, New Zealand ; Hutton Medallist, New Zealand Institute. Plates XL\I-XLVIII. 128 " ENDEAVOUR " SCIENTIFIC RESULTS. Having now completed the examination of the worms obtained during the cruise of the " Endeavour," it will be convenient to summarise the results of my work, which will entail to some extent a repetition of the introductory remarks in Part I. There it was noted that the " Challenger " obtained fifteen species of Polychaetes in the area covered by the cruise of the " Endeavour," of which twelve were new to science. The present collection contains forty -one species, including a fragment of a species of Sigalion, an indeterminable frag- ment of a species of Phyllodocid, and a species of Halodora, to which I have been unable to assign a name. Of these forty- one, five were obtained previously in this region ; seventeen species hitherto recorded in various other parts of the world are now added to the southern Australian fauna ; and I have found it necessary to make sixteen new species, one or two of which may, however, be merely varieties of already known forms ; and to erect one new genus for a Nereid, namely Cheilonereis. The " Challenger " obtained the fifteen species from four hauls ; the " Endeavour " secured these forty -one species in twenty-six hauls, but it must be borne in mind that in the latter cruise the primary object was an economic one in connection with the Fisheries, and it is highly satisfactory that so many worms, many of them of small size, should have been noted and so carefully preserved by the late Mr. C. T. Harrisson, the Biologist on board. In many instances a single individual of a species was received by me and, therefore, presumably the only one captured. We have a very small but fairly representative sample of the fauna of these deeper waters, for the majority of the families is repre- sented by one or more genera ; the Polynoinse and Eunicidae being well represented. There are some interesting omis- sions, however, for instance no member of the sub-family Aphroditinse were included, which is rather unexpected, as they are deep-water forms of large size. The families Syllidse, Ariciidae, Spionidae being small worms might easily be overlooked in sorting out the contents of the dredge, and burrowing as they do for the most part in mud or others living amongst the CoraUine Algse near shore are not urmaturally absent. One would perhaps have expected more TerebelHdse, of which only on& individual was obtained, and also members of the allied famihes. POLYCH^TA. BENHAM. 129 II. — Local Distribution. It will be useful to tabulate the localities at which the species were obtained so as to compare the differences between the various stations. The majority come from off the east coast of Tasmania, and this preponderance is no doubt due to the fact that more hauls were taken here than further north, at any rate to judge from the material submitted to me, though it is quite likely that in many cases no Polychaetes were obtained in some of the later hauls. I have given the depth at which the worms were found where that information was supphed, or have included information derived from other records. It will be noted that the depth varies from 17 to 1200 fathoms ; only two came from a greater depth than 1000 fathoms — Asychis victorioe from 1100 fathoms and Nephthys macrura from 1200 fathoms. Ihe majority came from between 50 and 100 fathoms. " Endeavour " Collection. Other Records. TASMANIA. Ten miles north of Circular Head. Nereis kerguelensis No depth given. . 10-100 fathoms (Moore) Eunice hassensis ,, Eunice pycnobranchiata „ Oyster Bay. Eunice bassensis 20-40 fathoms Lumhriconereis gulielmi 26 fathoms Storm Bay. Eunice pycnobranchiata No depth given Stauronereis australiensis ,, 38 fathoms (M'Intosh) South Cape. Chloeia incrmis 75 fathoms Breaksea Island, Port Davey. Eunice bassensis No depth given East of Maria Island. Eunice pycnobranchiata 78 fathoms Glycera tesselata 78 fathoms Lepidonotus willeyi 78 fathoms Physalidonotus rugosus 78 fathoms East-north-east op Maria Island. Protvlopsis palliata 100 fathoms 130 ENDEAVOUB SCIENTIFIC RESULTS. " Endeavour" Collection. Other Records. BASS STRAIT. East Coast of Flinders Island. Eunice bassensis No depth given. .38 fathoms (M'Intosh) Eunice pycnobrnnchiata ,, . .38 fathoms (M'Intosh) Oenoyie hastvelH ,, Glycern tesselata „ 31-638 fathoms (Moore) Nereis arenaceodentata „ . .surface (Moore) Physalidonotus rugosus ,, Halodora, sp „ Babel Island. Eunice pycnobranchiata 50-80 fathoms Hyalinoecia tubicola 50-80 fathoms Lumbriconereis sphceroce- phaln 70 fathoms Hesione splendida .50-80 fathoms Polynoe platycirrus 50-80 f ith.oms Physalidonotus rugosus 70 fathoms Sabellastarte inponica 60 fathoms Spirobranchus latiscapus. . . .50-80 fathoms East-nohth-east of Babel Island. Nepkthys inacrura 1200 fathoms North-east of Babel Island. Eunice pycnobranchiata 100-170 fathoms Twenty miles east of Babel Island. Lumbriconereis gulielmi 65 fathoms Eupompe australiensis 65 fathoms 8 fathoms (M'Intosh) Physalidonotus lurritus 65 fathoms VICTORIA. Gabo Island. Eunice siciliensis 75-200 fathoms Polynoe platycirrus 200 fathoms Physalidonotus laevis 200 fathoms PhysaUdo n otus paucibra n - ^ chiatus 200 fathoms Harmothoe etheridgei 200 fathoms Twenty-five miles south-west of Cape EveiIard. Physalidonotus laevis No depth given SotJTH-SOUTH-WEST OF .MoUNT CaNN. Physalidonotus rugosus 55-70 fathoms Lysarete australiensis 55-70 fathoms Cape Nelson. Asychis victor ire 1100 fathoms SOUTH AUSTRALIA. Forty miles west of Kingston. Lepidonotus hedleyi 30 fathoms Ceratoneteis falcaria 30 fathoms Scion e harrissoni 30 fathoms Eleven miles south-east of Cape Martin. Owenia fusiformis 21 fathoms ■Cape Jervis. , Eunice pycnobranchiata 17 fathoms POLYCH.^TA. — BENHAM. 131 " Endeavour " Collection. Other Records. Kangaroo Island. Eunice ba-9-sensis 40 fathoms Notopygos labiatus 40 fathoms Cape Wiles. Sigalion, sp. incert 100 fathoms South of St. Francis Island. Thnlanessa ocvlata 35 fathoms Ophelia dannevigi 35 fathoms Notomastus eisigi 35 fathoms One hundred and forty-two mixes south-south-west of St. Francis Island. Sternasfis scvtala No depth given. .55-260 fathoms (Moore) Great Australian Bight. Notopygos labiatuts No depth wiven Scalisetosvs anstralienftis . ... ,, Cheilonereis peristomialis . . .250-450 fathoms PhvUodocid, gen. et sp. indet. . .No depth given NEW SOUTH WALES. Between Port Stephens and Newcastle. Eunice pycnobranchiata 22-fiO fathoms Oenone hnsvelli 22-60 fathoms III. — Geogkaphical Distribution. I have also tabulated the distribution of the species. Apart from the five widely distributed species {Eunice siciliensis, Hyalinoecia tuhicola, Sternaspis scutata, Owenia fusiformis, and Hesione splendida) the relation of the Austra- lian fauna is distinctly A\ith that of the Pacific ; a very few are Antarctic or Subantarctic, such as Nephthys macrura and Nereis kerguelensis, yet both these enter the Pacific area, as they have been recorded from New Zealand. The most pecuUar discontinuous distribution is that pre- sented by Nereis arenaceodentata, hitherto found only off the coast of Massachusetts. It is, however, a small form and might readily be overlooked or regarded as an immature individual of some larger species. Under each species its wider distribution, if any, is dealt with in the text. It is interesting to note the great difference between the Australian and New Zealand Polychaete fauna. Of the series of species here recorded only six have been hitherto found in the latter seas, apart from the widely distributed species Oivenia fusiformis, Hyalinoecia tuhicola, and Sternaspis - scutata. 132 ENDEAVOUR SCIENTIFIC RESULTS, The Polychaete fauna of the Pacific is but little known r Ehlers has described that of New Zealand, I have given a preliminary account of the Kermadec fauna, but otherwise it is only from the Northern Pacific that we have much infor- mation— that of the Philippines by Grube, of Japan by Marenzeller, Izuka and Moore ; and of the west coast of North America by Moore, Johnson and Treadwell, references to whose papers will be found in the text. d (^ a-. i o ^ s ^■5 *"i Elsewhere. H > c /2-t3 i5?- z;n Hesione splendida X Ind. Oc; Red Sea ; N. Polyiioe plntyclrrus X X X Tnd. Oc. [Pacific Leptdonotus hedlcyi X X affin. Ind. Oc. affin. Japan Phy.mlido)iotus rugosus . . . X • 1 ■ X affiu X X X ; P turrituti. . ... P. paucibranchiatns Hnrmothoe efheridgei X 1 Scalisetosits australie.nsi/i . . X affiii. N. Pacific Thalancssa oculata 1 ^ N. Pacific Sigalion, sp. incert ! X affin. Pacific Eupompe (iufitralie>}sis. . . .| X Torres Strait Halodora, sp. incert X Nephthys macrura X 1 X Subantarctic X X X Notopygos labiatus Stauronereis australiensis .. X 1 [0. ; Pacific Eunice siciUensis X Med • Red Sea- Ind X X X X X X E. pyoiobranchidta Hynlinoecia tuhicola X X Atlantic ; Pacific Lumhricon ereis Kphoeroce- pluda X X L. quliehni X affiii. Japan Oenone hasivflli . X X X Lysurete mtstraliensis Nereis kerguelensis X X Subantarctic N. nrenaceodentala X East coast N. America [Pacific Cheilovereis peristomialis . X X affin. Japan, N.E. Ceratonereis fcdcaria X Ceylon [Pacific X Oicenin fusiforrnis X X Arctic ; Med.; Pacific Scione harrissoni X X X Notomastns eisigi Ophelia dannevigi X Sabellastarte japonica X Japan X Japan Protulopsis palliata X Ceylon [Pacific Slernaspis saifnta X i X Med. ; North Sea ; N. POLYCH^TA. — BENHAM. 13H IV. — Description of the Genera and Species. Family NEREIDiE.i Genus Nereis, Cuvier. Nereis kerguelensis, M'^Intosh. Nereis kerguelensis, M'Intosh, Chall. Rep., Zool., xii., 1885, p. 225. Id., Ehlers, Polychaeten, Hamburger magal- haenische Sammelreise, Hamburg, 1897, p. 65 ; Id., Ehlers, Neuseeland. Annelid., 1907, p. 11. Id., Gravier, Deuxieme Antarct. Exped. Francaise, — Annelides Polychetes, 1911, p. 76. Three complete individuals, one of which is a female filled with eggs though still in the " nereid " stage. The worms are larger than those recorded by Ehlers from South Georgia. The largest is 80 mm. by 5 mm. and 8 mm. over the parapodia, with 78 segments. A second, 65 mm. by 6 mm. and 8 mm. with 75 segments ; the third is 48 mm. by 4 mm. and 6 mm. with 70 segments. There are one or two discrepancies between the figures given by M'Intosh and Ehlers, and certain slight differences between these and my own observations. The peristomial cirri, which are relatively long in the " Challenger " figure, are in these specimens short as shown by Ehlers. So, too, there are differences in proportion in the tentacles, which M'Intosh shows as narrow and Ehlers as stout and short ; here, too, the present specimens agree with Ehlers'. I have in a previous paper dealing with N. australis^ pointed out that in one and the same species there is a great range of variation in the relative length of these cirri which appears to be independent of locality or age or. of methods of preserva- tion. Other features being in accord, these proportions are of less importance. Loc. — Ten miles north of Circular Head, Tasmania. Distributiofi. — New Zealand ; Kerguelen Island ; South Georgia ; Falkland Islands ; South Shetlands ; and other localities hereabouts. 1. In accordance with the International Rules of Nomenclature the usual title of this family, Lycoridae, must be replaced by Nereidae. 2. Benham — Report on Polychaeta, Subantarctic Islands of New Zealand,. 1909, p. 23fi. 134 '■ ENDEAVOUR " SCIENTIFIC RESULTS. Nereis arenaceodentata, Moore. (Plate xlvi., figs. 1-3.) Nereis arenaceodentata, Moore, Proc. Acad. Nat. Sci. Phila- delphia, 1903, p. 720. A single complete specimen of this small species is present. Its colour is pale pinkish brown in spirit. The body is broadest anteriorly, and tapers backwards ; the parapodia are long, increasing in length posteriorly ; their length is as great as the width of the body. The worm measures 22 mm. in length by 2.25 mm. across the body, and 3.25 mm. over the parapodia anteriorly ; but posteriorly these latter numbers are 1.25 mm. and 3 mm. respectively. Although the present specimen agrees in all essentials with Moore's species, yet, since that has been recorded only from the Atlantic coast, and the occurrence of the same species in two such widely different and dissimilar areas is so rare, if not unknown amongst Annehds, it is desirable that an account of this individual should be put on record and illustrated, so that any doubts about the identity may be set at rest, or at least the material provided for criticism. These notes and figures were, I need scarcely say, made before I became acquainted with Moore's account. The prostomiuiii is broader than long, \\ hich may be due in part to the protrusion of the pharynx, stretching the base. The eyes are large, the anterior pair larger than the posterior. The anterior pair as usual rather more widely separated than the posterior, to which they are closely placed. Moore shows them to be of the same size and in contact. The tentacles are short,,less than half the length of the pros- tomium. The palps are very large, and broad. Of the peristomial cirri the posterodorsal, that is the upper- most, is longest, reaching to the back of the fourth chaeti- gerous segment ; the dorsal anterior, nearly as long, reaching to the anterior margin of this segment ; the two ventral much shorter, scarcely reaching to the end of the second segment. In this respect there is a difference, for what it may be worth, from Moore's account and figure. The parapodia are all ahke, relatively long ; the various lobes pointed, distinctly separated and somewhat divergent. POLYCHJETAv — BENHAM. 135 The notopodial chaetophoral lip is unusually long and ■distinct. The dorsal cirrus reaches only about half way along the dorsal ligule ; there is no enlargement at its base. The chsetse are all colourless, as are also the acicula. In the 13th foot, the dorsal chaetae are about fourteen in number, long homogomph spinigers, arranged in a fan-shaped way. The supra-acicular bundle in the ventral lobe consists of five similar chsetse, A^ith four heterogomph falcigers, which have a long and knife-like appendix, hooked terminally, and densely setose. (PL xhi., fig. 23.) The sub-acicular bundle contains the same two kinds of chaetae. All the spinigers and all the falcigers are alike throughout the worm. This specific name was evidently derived from the nature of the denticles on the buccal region, which is covered with small conical paragnaths forming a broad band of approxi- mately equal width all the way round. Amongst them, how- ever, are a few rather larger ones arranged in longitudinal rows at intervals on the sides and under surface, which do not appear to correspond to the areas VI., VII., VIII. . The maxillarj^ region presents groups of small denticles, of about the size of the larger buccal kind, which are arranged as follows : — I. A rectangular patch formed of four transverse rows of 2, 2, 3, 4. II. An elongated oval group of numerous densely massed denticles. III. A patch of five or six transverse rows forming a squarish group. IV. A broad, oval patch of densely massed points. It is only in the details as to the number of denticles in the various patches that difi'erences exist between this and Moore's specimens (see his Figs. 5-6, PL xL). It is recognised that slight differences may exist in this matter without affecting the identity of the species ; a few more or a few less denticles in a group is a common variation. Loc. — East coast of Fhnders Island, Bass Strait. Distribution. — Wood's Hole, Massachusetts. 136 " ENDEAVOUR " SCIENTIFIC RESULTS. Genus Ceratonereis, Kinberg. Ceratonereis falcaria, Willey. (Plate x\\±, figs. 4-10.) Ceratonereis falcaria, Willey, Ceylon Pearl Oyster Fisheries,, pt. iv., Suppl. Rep.,, XXX., — Polychseta, 1905, p. 272. A single specimen of a small worm seems to be identical with Willey 's species. Its length is 20 mm., its breadth 1.25 mm. Its diameter is nearly uniform for the 60 segments of which it consists. The parapodia are short, less than half the diameter of the body ; the dorsal cirrus is very long. The prostomium is longer than its basal breadth and the preocular region is narrow and deeply grooved dorsally in the median line, giving the impression of a deep frontal incision, which, however, is scarcely existent in this indi\adual (Fig 4). The tentacles are short, about half the length of the prosto- mium. The palps are relatively large. The four eyes are large, nearly equal in size, and the two on each side are placed in nearly a straight fore and aft line, so that the four form a square. The tentacular cirri appear to be short and stout, the longest reaches only to the back of the first chsetigerous segment ; the next only touches the hinder margin of the peristomium. They are smooth. The parapodia have four lobes nearly equal in length' short and rounded. The no to podium and neuropodium are not very distinctly separated. The notopodial ligule is rather smaller than the middle ligule (or lip of dorsal chsetophore). The neuropodial ligule is stouter than the upper one. The dorsal cirrus is more than twice the length of the uppermost ligule (PI. xlvi., fig. 5). The posterior feet differ a httle from the anterior (PI. xlvi.^ fig. 6). The upper hgule is much shorter than the middle,, which is now the longest of the four lobes. The chsetge of the notopodium exhibit the remarkable change that Willey describes in his account of the species, that is, the homogomph spinigers of the anterior feet (PI. xlvi., fig. 7) are, in the posterior feet, replaced by a couple of thick brown homogomph falcigers, of AVhich the appendix POLYCH-ETA. — BENHAM. 137 is very short, brown, and bluntly bidentate ; the shaft is transversely striated with a deep articular cup (PI. xlvi., fig. '8). In the anterior feet the dorsal bundle contains only a few (2-4) homogomph spinigers, the appendix long, with extremely fine denticulations. The change in the character of these chsetae occurs on or about the 22nd segment. The ventral supra-acicular bundle contains also four homo- gomph spinigers, two of which have long and two shorter appendices. Below are four heterogomj^h falcigers, the shaft of which is thick and striated ; the appendix is moderately long, setose along the edge, and with a feeble terminal hook (PI. xlvi., fig. 9). The sub-acicular bundle contains four heterogomph spinigers, with a short appendix, and two or three heterogomph falcigers. In all the spinigers, dorsal and ventral, anterior and posterior, the appendix is the same in form though it may differ in length in the same bundle. In the first two feet, which as usual have only the neuro- podial lobe, all the chsetse are spinigers ; the upjjer bundle has two homogomphs and two heterogomphs with a short appendix ; those of the lower bundle are heterogomphs. The jaws. — The maxilla has nine blunt denticulations, and like the rest is pale brown. The paragnaths are very pale, rounded cones (PI. xlvi., fig. 10). The formula is :— I. Absent. II. 5 or 6, forming an ovol group. III. Absent. IV. A transverse curved group of about 8 denticles. Remarks. — At first I supposed that this worm was different from C. falcaria, as it diflers in three points from the few details given by Willey : firstly, in the segment at which the change of chaeta? begins, which is the 17th in the Ceylon worm ; secondly, in the absence of denticles in the area III. ; and, thirdly, in the number of denticulations on the maxilla, which Willey states to be only six. It may very likely prove that some of these features are variable. The denticles, for instance, in my species are very pale, and their absence in III. may be due to their having fallen away ; while 138 " ENDEAVOUR " SCIENTIFIC RESULTS. as observations on species of Eunice tend to show, the segment at which the change of chsetae occurs may probably be quite inconstant. hoc. — Forty miles west of Kingston, South Austraha, 30 fathoms. Distribution. — Ceylon. Genus Cheilonereis, gen. nov. The peristomium is produced laterally and ventrally to- form a large, widely extending, collar-like lip {Cheilos=^\i^) ; the base of the notopodium is raised into a lamelhform expan- sion exceeding in height the rest of the parapodium through- out the greater part of the body. The type species is Nereis cyclurus, Harrington. i Cheilonereis peristomialis, sp. nov. (Plate xlvi., figs. 11-18 ; Plate xlvii., figs. 19-22.) A single entire specimen of a large species which, like N. cyclurus, Harrington, is characterised by the great develop^ ment of the lower and lateral portions of the peristomium to form a great mobile collar-like lip, which partly covers even the protruded pharynx. It is a broad worm with greatly enlarged lamelliform bases of the notopodia, within the cirri, which with the lamelbform dorsal ligules," give it, at the first glance, a likeness to a Phyllodocid. It is a female filled with eggs, which enter the parapodial lobes. The length is 130 mm. for 116 segments ; the greatest breadth is at about the 25th segment, where it is 10 mm. across the body and 13 mm. over the feet ; its height is 8 mm. The diameter increases from 4 mm. at the peristomium to this breadth and then decreases to 8 mm. at about the mid- body, and thence to 2 mm. at the hinder end. The prostomium is rather broader than long ; the two pairs of eyes are nearly in line, close to the margin, but the posterior eye is hidden below the greatly developed anterior margin of the peristomium. The tentacles are slender, about half the length of the prostomium, and scarcely reach beyond the tips of the palps. The palps are large, with a small terminal joint, and spring from the lateral region about midway along the prostomium. 1. Harrington— Trans. New York Acad. Sti., xvi., 1898, p. 219. POLYCH^TA. — BENHAM. 13t> The peristomium is nearly twice the length of the next segment, dorsally smooth but ventrally pleated lengthwise ; whereas the sides are infolded in this individual, they are, in specimens from New Zealand, much longer than the ventral region, and expanded to form a hood or collar which projects outwards and forwards as a longitudinally folded membrane, evidently very mobile in life. Its appearance is well seen in the figure given by Johnson^ (PI. iv., fig. 46) of N. cyclurus from American waters. The peristomial cirri are short and approximately of equal length, the uppermost is longest, reaching to the 2nd chaeti- gerous segment. The parapodia are relatively long, especially in the posterior region of the body. The dorsal cirrus, on the anterior feet, is rather more than twice the length of the dorsal ligule (PI. xlvi., fig. 17), but by about the 35th foot it only exceeds it by a little (PL xlvi., fig. 18), and in the 50th ancl following it only about reaches to its apex. This relation is retained for the remainder of the body (PI. xlvii., figs. 19-20). The actual length of the cirrus decreases, it is not merely relatively shorter. The base of the notopodium just w ithin the cirrus is, already in the 15th segment, quite noticeably elevated, and by about the 20th is elevated and compressed laterally to form a lamella. In the 30th this becomes much higher and continues to do so posteriorly. Meanwhile the dorsal ligule, which is at first a rounded lobe (PI. xlvi., figs. 14-15), soon becomes pointed, elevated and compressed to form a lamella, so that the two structures, over the greater part of the body, unite to form a conspicuous lamella on the upper edge of which the dorsal cirrus is implanted in a slight notch (PI. xlvi., fig.. 18). The entire lamella is vascular and glandular, and especially the ligule. In the 9th foot (PI. xlvi., fig. 16) the middle hgule is a rounded, conical lobe, shorter than the dorsal, Avhich diverges upwards from it. The ventral ligule is still shorter than the middle and even less than the lips of the ventral chsetophore. As wiU be seen from a comparison of the figures of successive feet a line taken along the outer margin of the hgules, which is at first nearly vertical comes to slope downwards and inwards, forming a greater and greater angle with the vertical till in the hinder feet it is about 70°. 1. Johnson — Proc. Boston Soc. Nat. Hist., xxix., 1901, p. 400. 140 " ENDEAVOUR " SCIENTIFIC RESULTS. The chsetae are as follows : — The notopodial bundle consists of very long homogomph spinigers. The neuropodial supra- acicular bundle contains homogomph spinigers and rather fewer heterogomph faleigers (PI. xlvi., fig. 13), while in the sub-acicular bundle there are only a great number of hetero- gomph faleigers (PL xlvi., fig. 12). The spinigers have a long delicate appendix of the same length and structure throughout the body. The appendix of the faltate chaeta is rather long, and finely setose. There is a slight difference in the form of the articular cup of these chsetse in various j3arts of the foot and in different segments, as may be seen in the figures. The first and second parapodia have ventral bundles only ; all the appendices are broken off in the first foot, but in the second some of the sub-aciculars are heterogomph spinigers, with a shorter appendix than in the homogomphs. The third foot also has similar chsetse in the sub-acicular bundle, and the appendix of the falcate is longer than in the posterior feet. The jaws. — The maxillse are brown, with 7-8 denticulations on the curved exposed region, and a further 7-8 on the straight basal region. The dental formula of the paragnaths, which are shown in Fig 21, is :— I. One or two rather large conical denticles in a longitudina 1 row, with, in some individuals, 2-3 small ones at their side. II. An irregular rectangular patch of about 16 large denticles in three transverse rows. III. A transverse patch of large and small denticles in three rows, the large ones in the second and third rows. IV. An obUquely rectangular patch of 16 rather large denticles in two oblique rows (on the outer side of the maxilla). V. Absent. VI. An irregular rounded group of 2-3 curved rows of rather large denticles. The group lies further forward than the rest of the proximal or buccal series. VII- ; VIII. Forms a band of numerous denticles, those along the anterior margin (8-10 in number) larger than the rest, which are quite small and more numerous, and irregularly disposed in 2, 3 or 4 rows. The large ones do not extend up to the group VI. POLYCH.^TA. — BENHAM 141 Remarks. — This species is closely allied to the Eastern Pacific species. Harrington found C. cyclurus in association with Hermit-crabs. The present species differs from it in the form of the parapodia both in the atokons and in the epitokous forms. In the atokoiis condition the dorsal ligule has a rounded instead of the sharply pointed tip of C. peri- stomialis, and the notch in which the dorsal cirrus arises is much deeper, owing to the greater height of the basal lamella. The dental formula also, if I understand Harrington's account rightly, differs in that the denticle in V. is present. His account, however, is very obscure. He writes thus :- — " (5) Median dorsal proximal teeth =tMo points, prominent, in an ill-defined ring on proximal basal somite, which is slightly strongest at ventral side." It may be that these " two points " are the two groups of larger denticles in the compartment VI. The dental formula given by Izuka for C. shishidoi agrees closely with that in the present species. It appears to me that the shght differences between C. cyclurus and the Japanese species C. shishidoi, Izuka, ^ are insufficient for specific distinction. Moore has recorded C. cyclurus from Vancouver Island- and from the C'alifornian coast. ^ Ramsay^"" also describes it from the N.E. Pacific. Harrington's specimens came from Puget Sound. The two species are evidently closely alhed, and it is here that the trinominal nomenclature would be useful ; it has not been introduced into Polychset literature, but it well might be. Thus we should have C. cyclurus cyclurus from American waters, C. cyclurus shishidoi from Japan, and C. cyclurus peristomialis from Australasian waters. It is perhaps rather dangerous to suggest a new generic title for any Nereid, judging from the disuse of the many names suggested by Kinberg and Malmgren, but the extra- ordinary development of the lower lip differs entirely from anything hitherto described, so far as I can, ascertain from the Hterature accessible to me here. 1. Izuka — Journ. Coll. Sci. Imper. Univ. Tokyo, xxx., 1912, p. 177. 2. Moore— Proc. Acad. Nat. Sci. Philadelphia, 1908, p. 343. 3. Moore— Loc. cit., 1911, p. 246. 3a. After the MS. had been sent to the Editor I came across Ramsay's account of C. cyclurus (Proc. Zool. Soc, 1914, p. 237), wherein he suggests that the extraordinary development of the peristomium was almost worthy of generic recognition. He draws attention to the " curious homogom))h jalcate setae " which occur in the notopodial bundle of the parapodia in hinder part of the worm. I find in my MS. account, written some 10 years ago or more, that these occur also in our New Zealand species. Ramsay also discusses the relation of this species to Alitta virens. 142 " ENDEAVOUK " SCIENTIFIC RESULTS. There are, it is true, resemblances to Alitla virens, Sars, as may be seen by consulting the memoirs of Malmgren,i Ehlers- and TurnbuU-^ but from this species the present differs not only in its collar-like lip but in the form of the parapodium, especially in the much greater development of the dorsal lamella and in the proportions of this to the other parts of the foot ; in the greater length of the dorsal cirrus as well as in other details which may be merely specific. Although A. virens, as well as some other species such as N. lamellosa, Elilers-^ has the ventral surface of the peristo- niium marked by furrows, so that it resembles C. jieristomialis when the lip is folded fan-hke over the movith, yet in none of them is there any evidence that it can be spread outwards as it is here. Even Turnbull, who studied A. virens alive, says nothing of this ; indeed his and other figures show that it is not collar-like. ioc— Great Australian Bight, Long. 129° 28' E., 250-450 fathoms. Distribution. — I have specimens from New Zealand waters, described in MS. some years ago, which agree precisely with the above. One of them is in the Heteronereid phase, which may as well be described here. The Epitokous Condition. In the epitokous state this worm consists of 130 segments and measures 140 mm. Its greatest breadth is 20 mm. at the 15th segment, including the para podia, or 10 mm. over the body alone. The change in the parapodia occurs at the 27th segment, when the various foliaceous outgrowths appear as small structures, which gradually increase in size till the mid-body and then dwindle again and disappear at about the 85th, so that the worm is apparently not completely in the heteronereid condition (PI. xlvii., fig. 22). In the 40th foot the dorsal lamella is higher, more promi- nent than in the atokous state, and overhangs the base of the dorsal cirrus, so that the notch becomes a deep bay. The dorsal chsetse are replaced by the usual natatory bristles. The dorsal, middle and ventral ligules have undergone little change, but the ventral chsetophoral lip has developed a large semicircular folium, as has also the base of the ventral cirrus. 1. Malmgren — Nordislia Hafs-annulaten, 1805, p. 183. 2. Ehler.s — Die Borstenwiirmer, 1864-68, p. 559. .3. Turnbull — Trans. Connecticut Acad., iii., 1876, p. 266. 4. Ehlers — Loc. cit., p. 564. POLYCHiETA - — BENHAM. 143 The first specimens obtained in Ne\\ Zealand waters were found in the topmost Avhorls of a gastropod shell inhabited by a Hermit-crab, but later others were met with free in the Family GLYCERID.E. Genus Glycera, Savigny. Glycera tesselata, Grube. (Plate xlvii., figs. 23-25.) Olycera tesselata, Grube, Arch. f. Naturgesch., xxix., 1863> p. 41. Id., Ehlers, Die Borstenwiirmer, 1864-68, p. 654. Id., Treadwell, Bull. U.S. Fish Comm., xx., 2, 1903, p. 201. Id., Moore, Proc. Acad. Nat. Sci. Phila- delphia, 1903, p. 464 ; Id., Moore, Loc. cit., 1908, p. 348; Id., Moore, Loc. cit., 1911, p. 300. Id., Izuka, Journ. Coll. Sci. Imper. Univ. Tokyo, xxx., 1912, p. 241. Five individuals, from the seas around Tasmania, represent this non-branchiate species with long pointed anterior para- podial hps and two shorter rounded posterior lips, with long filamentous pharyngeal papillae, and a terminal eyespot on the prostomium. One specimen measures about 50 mm. in length, for it is twisted and, therefore, its actual length impossible of measure- ment. Its breadth is 3 mm. and the number of segments about 100. The prostomium is short, with about 10-12 annulations ; its length about twice its basal width, and as great as the length of two normal segments. It is, however, bent backwards in this individual and covers 7-8 segments, which are much shorter than the later ones. At its tip it carries the usual two pairs of short tentacles, and between them, on the ventral surface, is a black pigment spot, which, however, is absent in one specimen from Maria Island, Tasmania. The body segments are biannulate dorsally, the length of those in the anterior region about l-5th the diameter of the body, but in the hinder half of the worm l-4th this diameter. The colour is in most cases a uniform yellowish brown, but in two specimens from Maria Island the anterior segments and the prostomium are speckled with brown pigment. 144 " ENDEAVOUR " SCIENTIFIC RESULTS. The parapodia are, with the exception of the anterior, of equal length, about l-3rcl the width of the body. The propor- tions of the length of a parapod to its height is about as 3 : 2. The two anterior lips are long, narrow, and pointed, the ventral shghtly longer than the dorsal (PI. xlvii., fig. 23). The two posterior lips are separated by a distinct, rather deep, incision ; they are rounded and much shorter than the anterior lips. The ventral is rather longer and of less height than the dorsal. The dorsal cirrus is nearly spherical, a little removed from the foot up the side of the body, with a deep constriction at its base. The ventral cirrus is conical, its tip reacliing as^ far outward as the posterior lip. The chsetse are colourless, the acicula golden. The dorsal chfette are capilliform, about 9-10 in number,, with upturned tips and a narrow denticulated and striated flange. The ventrals are homogomph spinigers, 5 in the supra-acicular and 8 in the sub-acicular bundle ; the appendix has a narrow striated and denticulated flange on one or both sides ; the edge seems to have a row of minute dots along it, which, under high raagnification, are seen to be minute prominences. The dorsals have the same ornamentation. The first two parapodia are without the dorsal cirrus ; the three or fovn* anterior feet have both the posterior and anterior lips pointed. The feet gradually increase in size till the 8th, after which they remain uniform. The pharynx, in this specimen everted, is short and wide ; its length is 5 mm., its breadth 4 mm. at the anterior end. Its base overlaps the anterior feet. In a soft specimen the everted pharynx has a length of 7 mm., a terminal breadth of 3 mm., and resembles M'Intosh's figure of G. amboinensis. Its surface is covered with closely set, uniformly distributed- long filamentous papillae, giving it a velvety appearance, These papillae have a subterminal oval disc at one side of the apex, which appears sucker-hke in that it is slightly hollowed out (PI. xlvii., figs. 24-25). The apex of the papilla carries a few stiff" sensory hairs. The subterminal oval disc is known from the work of Gravier,i but neither in this paper nor in that of Jourdan^ nor Oppenheim"^ are these terminal hairs shown, though the sense cells and nerve fibres are traced out. 1. Gravier— Bull. Sci., xxxi., 1800, p. 422. 2. Jourdan— Ann. Sci. Nat., Zool., (7), xiii., 1892. 3. Oppenhcini — Proc. Anier. Acad. Arts and Sci., xxxvii., 1902, p. 553. POLYCH^TA. — BENHAM. 145 The usual four black jaws, surrounded at their base by a circle of soft stout conical processes, lie at the apex of the pharynx. Remarks. — The occurrence of a Mediterranean species in these waters appears at first rather remarkable. But we may bear in mind that Eunice siciliensis and Hyalinoecia iubicola are instances of the same wide distribution. More- over, the present species has already been recorded from the North Pacific and from the coasts of California and Japan by Moore. It appears to me probable that G. amhoinensis, M'lntosh,^ is synonymous with this. His specimen Avas in a poor state of preservation, and apart from his statement that the pharyngeal papillae are " conical '" it fits well with this. Logs. — East coast of Flinders Island, Bass Strait. Off Maria Island, Tasmania, 78 fathoms. Distrilmtion. — Mediterranean ; Atlantic ; North Pacific. Family OWENIID.^. Genus Owenia, delle Chiaje. OwENiA FUSiFORMis, delle Chiaje. Owenia fusijormis, delle Chiaje, Descr. notom. animal, invert. Siciha citeriore, 1842. (PI. 175, figs. 1-6.) Ammochares ottonis, Grube, Beschr. neue oder wenig bek- annter Anneliden, 1846, p. 164, Owenia filiformis, Claparede, Mem. Soc. Phys. Geneve, xix., 1868, p. 446. Owenia brachycera, Marion, Revue Sci. Nat. Montpellier, iv., 1875. Owenia tenuis, Haswell, Proc. Linn. Soc. N.S. Wales, vii., 1883, p. 633. A shelly tube, 22 mm. in length and 4 mm. in diameter, containing a fragment of a worm measuring 16 mm. by 2 mm., and consisting of 8 chaetigerous segments without head or anal end, must be attributed to this species, which as Ehlers- has pointed out is of practically cosmopolitan distribution. 1. M'Intosh— Chall. Rep., Zool., xii., 1885, p. 345. 2. Ehlers — Die Polychseten des magellan. u. chilen. Strandes, 1901, p. 193-. 146 " ENDEAVOUR " SCIENTIFIC RESULTS. The above synonymy has been taken from St. Joseph's- useful work in which the European Polychsetes are reviewed and the references gathered together. I have changed the name of the family, which has hitherto borne the title Ammocharidee since the time when the genus^ was termed Ammochares, Grube ; but as modern zoologists agree in resurrecting delle Chiaje's generic name, the name of the family according to the rules of zoological nomenclature must be derived from its type genus. Loc. — Eleven miles S. 74° E. of Cape Martin, South Australia, 21 fathoms. Distribution. — From Greenland and Davis Strait to the Mediterranean ; ChiH ; Phihppines ; Japan ; and New Zealand. Family TEREBELLID.E. Genus Scione, Malmgren. SCIONE HARRISSONI,2 Sp. UOV. (Plate xlvii., figs. 26-31.) A single individual in its tube of small shells and shelly fragments, Polyzoa and mud so loosely held together that it falls to pieces when the worm is removed. The worm measures 34 mm. in length, with an anterior diameter of 3 mm., which posteriorly decreases to I mm. It is ill-preserved, soft and readily ruptured. There are 43 distinct segments, but the mid-body is so soft, being partially decayed, that it is not easy to be sure of the number of segments here. The hinder segments, however, are well preserved, and the last 25 carry rather high chsetigerous papillae. The colour is a dull brown, but the dorsal surface of the branchial segment is deep reddish brown. It is a female filled with eggs. There are 15 notopodial bundles of capilliform chsetae, each bundle inserted in a small prominence. The first bundle is in the 4th segment. 1. St. Joseph— Ann. Sci. Nat., Zool., (8), v., 1898, p. 397. 2. So named after the kite Mr. C. T. Harrisson, Federal Covernment Biologist, on board the ill-fated " Endeavour," to whose care we owe the excellent preservation of these Polycha?tes. • polych.i;ta. — benham. 147 The ventral shields are 13 in number ; anteriorly they are transverse, rectangular ; posteriorly, square. The genus is characterised by three cutaneous lateral folds at the anterior end belonging to as many segments. The prostomium is produced into a dorsal thin upper lip, with an undulating margin, behind which is the tentacular ridge which extends downwards at each side nearly to the lower hp (Pi. xlvii., figs. 26-28). The tentacles appear short, but were so intertwined that it was impossible to measure them. There are no eyes or pigment spots. The peristo- mium is produced laterally and ventrally into a large thickish fold, broader laterally, where its edge is reflected ; narrow ventrally, allowing the rectangular lower lip to be seen (PI. xlvii., figs. 26-27). The second segment and the next are fused together. I follow Ehlers in this analysis of the region as given for Terrebella {Scione) cefrafaA On what is probably the 3rd segment is the single pair of arborescent gills, \\ hich spring from a nearly vertical wall in front of the first chsetigerous segment (PI. xlvii., figs. 26-28). From the hinder margin of this branchial segment rises the second lateral cutaneous flap, which extends downwards to the level of the lower edge of the tori uncinigeri of the follow- ing segments, but does not extend across the ventral surface. The third fold, which springs from the hinder portion of the first chsetigerous, is of less width but of the same extent as the second. The arborescent gill has a thick stem Avhich bears a few short thick branches, which in their turn divide again into a niimber of short twigs, so that the gill itself appears as a mass of densely aggregated branches. In front of the base of each gill is a small conical prominence projecting forwards, and carrying at its apex a narrow cylindrical process, which is inclined outwards (PI. xlvii., fig. 28). (Are these the representatives of a second, anterior, pair of gills ?) The uncini commence on the second cheetigerous segment as usual, and the first six are uniserial ; but on the 7th and on the eight following segments they have an " alternate " arrangement. Those in the abdomen, however, are again uniserial, all the points being turned in the same direction. ]. Ehlers— Florida Anneliden, 1887, p. 248. 148 '^ENDEAVOUR " SCIENTIFIC RESULTS. The uncini are all alike, bvit those of the abdomen are smaller. Seen from the side, each appears to have three small teeth on the back of the main tooth, but from above it is seen that there are three rows of testh (PI. xlvii., figs. 29- 30). The capilliform chaetae have curved points with a flange on each side, of which that on the convex side is much broader and longer than the other (PL xlvii., fig. 31). Remarks. — This worm agrees with Malmgren's^ diagnosis of Scione, and though he makes no mention of the cutaneous folds in his generic diagnosis he describes them for the species S. lobata. Log. — Forty miles west of Kingston, South Australia, 30 fathoms. Family CAPITELLID.E. Genus Notomastus, Sars. NOTOMASTUS EISIGI, SJ). YIOV. (Plate xlvii., figs. 32-35.) A single specimen, somewhat curled and imperfect pos- teriorly, was enclosed in the remains of a tube of sand and shelly fragments. The incomplete worm measures about 36 mm. in length. The thorax, which measures 12 mm. in length, consists of the usual twelve segments, bearing only capillary bristles. The diameter increases from 2 mm. at the peristomium to 3 mm. at the sixth segment and then decreases to 2.5 mm. at the twelfth. The prostomium is small, rounded, with a minute terminal globular knob, and is retractile to some extent within the peristomium. The peristomium is nearly twice the length of one of the following segments, which are very glandular and bi-annulate, the skin being cut uj) into the areas by secondary fissures, as in the European species. (These two facts are not shown in the figures.) The peristomium is without chsetse, and I am unable to find any on the following segment. If they are present they must be very 1. Malmgren — Nordiska Hafs-annulaten, 1865, p. 383. Dr. Leiper (Ann. Mag. Nat. Hist., 1908, p. 468) includes this in his list of Polychaete generic names that are pre-occupied ; it will do no harm to retain it for the present. POLYCH.^TA. — BENHAM. 149 minute, for when rolling the worm over till the following segments are in profile so as to show the bristles, I can see none on the second segment. Possibly they are embedded in the glandular tissue, as in the clitellum of an earthworm at maturity. Each segment after the second bears two widely separated tufts of capillaries ; all are ahke, both ventral and dorsal, slightly curved, finely and simply pointed, with a small flange on each side, of which one is broader than the other, but in both the oblique striations are very faint. In the third and fourth segments they are only few in number and quite short, Avhile in the fifth and following they are more numerous and longer. The ventral surface of the last seven thoracic segments is traversed by a narrow band-like ridge, about r-6th of the diameter of the body, and bounded on each side by a narrow furrow. This ridge is continued, though in a less marked manner, on the abdomen. The abdominal region is marked off from the thoracic by its smaller diameter ; the segments are no longer than those anteriorly, but the skin is smooth, so far as one can judge in the ill-preserved, soft and rather damaged condition, for the sand and shelly fragments contained in the intestine cause the body wall to rupture on handhng. The abdomen is about 2 mm. in diameter, of uniform width over the 20 segments which remain, and I can see no annulations. The first abdominal segment, though having all the appear- ance of the others, is provided with capillaries in the dorsal and ventral bundles (PI. xlvii., fig. 33). Each of the following segments carries the usual notopodial and neuropodial tori uncinigeri, in which there is a single row of hooded hooks. The neuropodial torus extends up the whole side of the body as a distinct ridge, and ends dorsally in a free simple " gill." (PI. xlvii., fig. 32.) Above this is the rounded " lateral organ," and on the dorsal surface the short notopodial torus, which is not at all prominent and nearly meets its fellow in the median line as in N. lineatus. The hook (PI. xlvii., fig. 34) has a long shaft curved like the chaeta of an earthworm ; it is bent gently at its proximal end, and terminates distally in the usual hook, with an accessory tooth on its upper surface. The whole hook is almost en- tirely enclosed in a large nearly circular hood, which is in reality double (PL xlvii., fig. 35). 150 " ENDEAVOUR " SCIENTIFIC RESULTS. Remarks. — This differs from the diagnosis of the genus- Notomastus as given by Eisigi in the presence of capillaries on the first abdominal segment. It differs from Eunotomastus- of M'lntosh^ in having only 12 thoracic segments and in the presence of capillaries only in the anterior abdominal' segments. I hesitate to form a new genus, as my literature is not suffi- ciently modern for me to ascertain whether such a divergence is already known. Loc. — South of St. Francis Island, South Australia, 35 fathoms (with Thalanessa oculata and Ophelia dannevigi). Family OPHELIID^. Genus Ophelia, Savigny. Ophelia dannevigi,^ sp. nov. (Plate xlviii., figs. 36-37.) A single specimen, 20 mm. in length with a breadth of 1.25 mm. across the ventral surface, and a height of 2 mln. The body contains 32 chsetiggrous segments, and has the usual spindle shape. The prostomium carries at its apex a finely pointed conical tentacle with a pair of nuchal organs at its base, appearing as little brown streaks. The mouth is at the level of the first chsetse. The anus is surrounded by a circlet of 16 subcylindrical tentacle-like organs, eight on each side, arising from a common basal membrane. In addition there is a couple of ventral lancet-shaped stouter processes (PI. xlviii., fig. 37.) The ventral surface is traversed by a broad furrow, marked with transverse muscular bands, and bounded by a well-defined rounded ridge on each side. This " sole " is continued on to the head where, however, the ridges are absent. At first narrow, it attains its full width at about the tenth segment, and retains this till near the hinder end, where it again narrows, terminating at the base of the ventral anal lobes. 1. Eisig — Die Capitelliden, Naples Monograph, 1887, p. 807. 2. M'Intosh— Chall. Rep., Zool., xii., 1885, p. 388. 3. Named after the late Mr. H. C. Dannevig, Commonwealth Director- of Fisheries. POLYCH^TA. BEXHAM. ISl! The notopodia and neuropodia are distinct lobes, except in the anterior segments, ^here they are very small. Each lobe is a short cyhndrical structure A\ith definite hps, and on the anterior face is a small round-topj}ed papilla. There is no ventral cirrus (PI. xlviii., fig. 36.) Parapodial gills are present after the tenth segment, except on the last three ; each has the usual fonn — simple, tongue-shaped and grooved. The first and second parapodia are very inconspicuous and carry but few chsetse : I counted three in the former and five or six in the latter. In the remainder they are numerous, all are alike, capilliform, of a shiny brownish tint. The last three segments of the body decrease in diameter successively and abruptly, so that from the side they appear to be telescoped into one another, with a projecting posterior angle, where the chsetse arise ; these are very long on thes-e segments, and reach beyond the anal tentacles. The body may, then, be analysed as follows : — Total, 32 segments, of which the first ten and the last three are Avithoiit gills, so that there are 19 gill-bearing segments. Remarks. — It presents a good deal of similarity to O. negleda, Schneider, as described by St. Joseph, ^ but he gives eighteen anal tentacles in addition to the ventral lobes, and the number of segments is one less. Loc. — South of St. Francis Island, South Austraha, 3o fathoms (with Thalanessn ocvlafa and Noiomastits dsigi). Family MALDANID.^. Genus Asychis, Kinherg. ASYCHIS VICTORI.^, SI). IIOV. (Plate xlviii., figs. 38-45.) Praxilla abyssorum, MTntosh, Chall. Rep., Zool., xii., 1885, p. 408. A fragment, consisting of the head and seven chsetigerous segments, measures 53 mm. in length by 4 mm. in diameter. The prostomium slopes downwards to the ventral surface, is convex dorsally with a broad median flat ridge or platform occupying about half its width (PI. xhnii., fig. 38) ; this is transversely striated and bounded on each side by a deep 1. St. Joseph— Ann. Sci. Nat., Zool., (8), v., 1898, p. 369. 152 " ENDEAVOUR " SCIENTIFIC RESULTS. furrow, nearly longitudinal, which, however, curves outwards anteriorly and ceases before it reaches this end (PI. xlviii., fig. 40). The anterior end is rounded, and its thin margin is somewhat "upcurved. Along each side of the prostomium is an upstanding lamella, whose thin upper edge is entire, and posteriorly is continuous with the margin of the peristomium. The prostomium is dorsally overhung by a transverse flap of the peristomium, which is also entire. If this flap be lifted up the prostomial platform or flattened ridge is seen to extend below it for about half the length of the peristomium ; this concealed portion is equal in length to the exposed portion of the platform. The peristomium is biannulate, the posterior annulus being about half the length of the anterior, which is traversed by a horizontal furrow on each side dividing it into an upper and a lower portion, which latter is prolonged into the marginal lamella of the prostomium (PI. xlviii., fig. 38). The ventral surface of the head is flat, and the mouth is a longitudinal slit widened posteriorly into a nearly circular aperture, which lies at about the middle of the anterior annulus of the peristomium (PI. xlviii., fig. 39). The second segment and the four following are biannulate ; the anterior annulus is chfetigerous and about twice the length of the posterior. The segments gradually increase in length, so that the distance between the 6th and 7th bundles of chaetae is about twice that separating the 2nd and 3rd bundles, the actual distances being 9 mm. and 5 mm. respectively. Each of the segments 4, 5, 6 and 7 has a lateral glandular cushion along its entire length, which reaches up the side as far as the notopodium, but on the 8th this glandular cushion is limited to a quadrate area round the torus uncinigerus. The first chsetigerous segment, which succeeds the peri- stomium, has only dorsal capillary chsetse, arranged in a vertical fan-shaped bundle protruding from a two-hpped notopod. In the following segments there is, in addition, a ventral torus uncinigerus, in which the hooks are arranged in a single row. This torus is very short in the third segment, rather longer in the fourth, and attains its full development in the fifth, where it is nearly half the height of the body. POLYCH.^TA.— BENHAM. 153 The number of chsetse in the first and second bundles as well as the vertical length of the bundle are greater than in the succeeding segments ; the length of the second bundle being twice that of the sixth, but, in compensation, the height of the torus has increased. There are two kinds of chaetse in the notopodium (PL xlviii., fig. 41) — (a) long, brownish capilhforms, with a very narrow flange and attenuate apex which is biserrulate (PI. xlviii.^ fig. 44) ; and {b) much shorter, rather thinner, colourless, biflanged bristles, whose apex is also very long, but without serrulations ; the flange on one side is much wider than on the other (PI. xlviii., fig. 45). These smaller chaetse lie anteriorly to the longer. I counted in two bundles 20 of the longer and 12 of the shorter sort. The hooks (PI. xlviii., figs. 42-43) have a nearly straight shaft with a long terminal tooth, Avith three or four suc- cessively smaller teeth on its back, and the usual tuft of delicate filaments projecting upwards in front. Remarks. — There is, I think, reason to suppose that Praxilla abyssorum, M'Intosh, is identical with this species. The figures in the " Challenger " Report (PI. xlvi., figs. 10, 11) certainly disagree with the genus Praxilla in the form of the hea^l and in other details. The figure is not very detailed, owing possibly to the state of contraction of the worm, but it shows the biannulate peristomium and the horizontal furrow on its side. It is true that MTntosh states that the uncini do not begin till the 5th segment ; that again is possibly due to its state of preservation, for the tori in these anterior segments are quite short and the hooks few. The figure of the hooks, however, agrees precisely with that found in the present species. It appears, moreover, that he found two kinds of dorsal chsetae in the first four segments at any rate, for he writes : — " Some of the bristles are furnished with a distinct wing, while others are simple slender tapering structures." Another fact weighs with me, viz., that the worm was obtained in nearly the same region, " midway between the Antarctic region and Australia, in 1950 fathoms." The form of the head, as well as certain other features, agree with Ehlers' account^ of Asychis (Maldane) amphiglypta from South Georgia. At first he placed it in the genus Mal- dane, but in his account of the New Zealand Polychsetes^ it was included in the genus Asychis. 1. Ehlers — Polychseten Hamburger mai^alhaenische Sammelreise, Ham- burg, 1897, p. IFQ. 2. Ehlers — Neusceland. Annelid., 1907, p. 2fi. 154 " ENDEAVOUR " SCIENTIFIC RESULTS. I have duplicates of specimens which I had sent to Ehlers from Waiheke, near Auckland, New Zealand, and noticing the resemblance between the Victorian specimen and Ehlers' account, I examined the New Zealand species. I find that the second or smaller sort of dorsal chsetse are totally unlike those figured by him for the A. amyliiglypta from South Georgia, Avhere he describes them as acicular in form. In the Waiheke specimens, however, they are transparent bristles, with a Avide flange which is not traversed by the shaft and has the striations palmately arranged, instead of pinnately — that is, they radiate from the end of the shaft. These chsetse are situated anteriorly to the longer capillaries, not pos- teriorly as the acicular-like chseta? in A. amphiglypta are stated to be. It is possible that Ehlers overlooked these smaller chsetse, which only just project through the skin. When mounted in Canada balsam the flange is scarcely visible, and then they appear bluntly pointed ; again, the capillaries, which he represents as " jointed," are, in the New Zealand as well as in the Victorian species, finely biserrulate along the fine apical point. The present Morm is quite distinct from the New Zealand species and from A. amphiglypta. I have placed it in Kinberg's genus Asychis, since it appears to agree better with that than with any of Malmgren's genera, but as I have but Httle recent literature on the family, I have described the external features somewhat in detail. Loc. — Thirty miles south of Cape Nelson, Victoria, IKH) fathoms. Family STERNASPID.^. Genus Sternaspis, Otto. Sternaspis scutata, Ranzani. Thalassema scutata, Ranzani, Isis, xi., 1817, p. 1457. Sternaspis thalassemoides, Otto, Nova Acta, Acad. C.L.C., X., 1821, p. 619. Sternaspis princeps, Selenka, Chall. Rep., Zool., xiii., 1885, p. 6. One specimen, a good deal shrunken, with a length of 12 mm. and a width of the anterior chsetigerous region 4.5 mm. POLYCH^TA. — BENHAM. 155 I have compared it with specimens from Naples and can find no essential differences ; neither proportions of body Tior of the shield, neither in the colour nor configuration of the latter, neither in the niimber of bundles of chsetae round its margin, nor in the number of those in the rows in anterior region of the body. The only instance, so far as I know, of the capture of a member of the genus in these southern waters is that recorded by Selenka in his " Challenger " Report in which he describes under the title of aS'. princeps specimens from deejo water east of the East Cape of New Zealand. A very brief account is given, accompanied by a figure which does not show the features used in his diagnosis. In that account two characters are mentioned — the existence of " about forty " bundles of chsetae around the shield, and the presence of an oblique ridge across the shield dividing it into two unequal areas. But this is true also of *S'. scutafa from Naples. The number of bristle-bundles varies, apparently with the size, for in a small specimen there are 28-30 bundles, and in a larger one I find 38 bundles. I have specimens of Siernaspis from two widely separated parts of the New Zealand coast — from Akaroa, in the South Island, 6 fathoms, and from the mouth of Akitio Kiver in the North Island, 20-36 fathoms. The former are smaller than the latter, but I can find no difference between them and the Naples specimens, comparing similar-sized individuals. I am compelled, therefore, to suppose either that there are two species of Sternaspis in our New Zealand waters, or that 8elenka's species is a synonym — which is the view I take.^ Loc. — One hundred and forty-two miles west of Kingston, South Australia, 30 fathoms. Distribution. — European seas ; Japan ; North Pacific^ ; New Zealand. Family SABELLID^. Genus Sab ellast arte, Kroyer. Sabellastarte japonica, Marenzeller. Laonome japonica, Marenzeller, Sud-japan. Annelid., 1884, p. 16. The species must be removed from the genus Laonome, Malmgren, in which the thoracic chsetae are of two distinct forms, to that of Sabellastarte, in which there is but one form , though this occurs in two sizes in a bundle.-^ 1 . I give further details about the«e New Zealand specimens in Trans. N.Z. Inst., vol. xlviii. 2. Moore— Proc. Acad.. Nat. Sci. Philadelphia, 1903, p. 487. 3. St. Joseph— Ann. Sci. Nat., Zool., (7), xvii., 1894, p. 249. 156 " ENDEAVOUR " SCIENTIFIC RESULTS. The species is one of the giants in the family. Three indi- viduals were received by me ; two were still within the tube, the third fortunately had been removed and was in consequence well preserved. I may add a few notes in supplement of the diagnostic description by Marenzeller. The worm measures 168 mm. in length over all, of which the gills occupy 45 mm., but this is not their true length, for they are much contracted and twisted. In a second indi- vidual, measured in the tube, the body is 126 mm. long and the gills less twisted add 60 mm. to this. The gills are, thus, approximately half as long as the body. The thorax is 1 1 mm. in length, 10 mm. in width, and 7 mm. in height. In this individual there are only seven chsetigerous segments, the first segment being without any chaetae ; the uncini commence with the dorsal bristles on the second segment. Another specimen exhibited a still greater degree of almormality, for it has only four bundles of chaetse with the uncini below, although the ventral gland shields and intersegmental furrows indicate the normal eight segments. In this case the 1st, 3rd, 4th and 8th are without chaetse. The abdomen contains more than 200 segments, with an anterior diameter of 10 mm. ; it is flat ventrally, much arched dorsally, but further backwards the arching decreases till in the hinder half of the body the dorsal surface is nearly as fiat as the ventral. The diameter is pretty uniform throughout, till close in front of the hinder end, when it rather suddenly tapers, the end being rounded. As Marenzeller finds, the fjsecal groove does not cease at the hinder end of the thorax, but is traceable as an ill-defined shallow furrow along the dorsal surface forwards, in a curved line towards the median dorsal depression, Avhich lies just behind the collar-gap anteriorly. The groove would probably disappear in soft specimens ; it is recognisable in this well- preserved individual owing to the shght irregular foldings of the skin. A similar but narrower furrow symmetrical with it exists on the left side. The general colour of the worm is pale brown with a purplish tint ; on the thorax this tint is produced by minute dots of pigment, which are especially aggregated above the notopodia, thus producing a broad purphsh band, which extends on to the posterior face of the collar lobes and on to the branchial base. The gills themselves are uniformly pale brown, without any purphsh tint, as also are the edge and anterior face of the collar. T-OLYCH^ETA. — BENHAM. 157 Ventrally the gland shields are darker than the neigh- bouring parts ; the region between them and the chaetigerous papillae on the abdomen being almost white. The gills are symmetrical, the ventral margin of the base of each is inrolled in a plane sjjire which describes more than a circle. The radioles (gill filaments) really arise in a single series, as Quatrefages^ pointed out for iS'. indica, Sav., but some have broader outer edges than others, and thrust them out of series so that there appears to be an inner and an outer rank. There are at least 130 radioles on each gill. The tip is naked for a distance equal to about twice the length of the upper pinnae. 2 The dorsal tentacles are 15 mm. in length, the so-called "palps" or buccal lobes are folded, directed forwards, and about as long as the ventral lobe of the collar. The collar has a deep median cleft separating a right and left ventral lobe, which is triangular with rounded tip, over- arching the first segment. The lateral portion, which is continuous with this, is low, shghtly reflected, and decreases in height as it passes upwards to the level of the notopodia ; here a deep incision separates it from a dorsal lobe, which is directed forwards and inwards, at right angles to the former, to meet, but not to unite with, its fellow in the mid-hne anteriorly. The tube is a rich chestnut brown in colour, of tough horny material ; the lower end is paler, transparent and flexible ; the upper end opaque, much darker, and coated with fine grey mud. The lower end being softer is crumpled and twisted, so that the true length is difficult to determine, but it is about 220 mm., with a diameter of 12 mm. at the upper end. As to the capilliform chsetae of the thorax, the figures given by Marenzeller exhibit some shght differences from my observations, which, however, may be due to his artist. Thus the flange on the shorter sort (PI. iii., fig. 4B) is some- what narrower than in the chaetae examined by myself, and the more slender sort (PI. iii., fig. 4A) is represented as having a second flange, which is not always the case in the present specimens ; while here the flange is broader than Marenzeller's figure indicates. 1. Quatrefages — Hist. Nat. Annel., ii., 18fi5, p. 432. 2. I follow M'Intosh and Willey in the use of these terms, though it is unfortunate that thev use " radiole "' in a different sense from Grube. l58 '• ENDEAVOUR " SCIENTIFIC RESULTS. These small differences, however, are of little importance when the general agreement is so close. Remarks. — It seems useless to attempt to identify Baird's^ " Sabella grandis " from which the gills were absent, and almost as useless to speculate as to whether the present species is or is not identical with Spirographis australiensis, Haswell,- from Port Jackson, for little precise anatomical detail is given. Both, however, are large forms, whatever genus they may belong to. It is possible that the large size and the inrolled gill-base led Haswell to place it in that genus. The former is said to have come from New Zealand, but no Sabellid of this size (six and a-half inches without the gills) has been met with on these shores since that date. Loc. — Off Babel Island, Bass Strait, about 60 fathoms. Distribution. — Japan. Family SERPULID/K. Genus Spirobranchus, Blainville. Spirobranchus latiscapus, Marenzeller. (Plate xlviii., figs. 46-50.) Pomatostegus latiscapus, Marenzeller, Siid- Japan. Annelid., 1884, p. 22. Id., Moore, Proc. Acad. Nat. Sci. Phila- delphia, 1904, p. 173. " ? Pomatoceros strigiceps," M'Intosh, Chall Rep., Zool., xii., 1885, p. 173. Id., Treadwell, Bull. U.S. Fish. Comm., xxiii., 3, 1906, p. 1179. I have transferred the species to the genus Spirobranchus on accoimt of {a) the wide V-shaped ventral area on the thorax, and {b) the form of the abdominal chsetae, which are not " sickle " shaped as in Pomatostegus, but " trumpet " shaped,^ and (c) the larger number of denticulations on the uncini.^ It was Ehlers who fii'st pointed out these differences between the two genera, which previously had been dis- tinguished by the form of the operculum. Ehlers, Maren- zeller, St. Joseph and others have relegated this feature to a 1. Baird — Jour. Linn. Soc, viii., 1S65, p. KiO. 2. Haswell— Proc. Linn. Soc. N. S. Wales, ix., 1885. p. 073. 3. Ehlers— Florida Anneliden, 1887, p. 299. 4. Pixell— Trans. Linn. Soc, Zool., xvi., 1913, p. 78. rOLYCH^TA. — BENHAJI. 159 secondary position, and it is this feature of the present species which led to its being placed in the genus Pomatostegus, for the discs are in two or more tiers and there are no outgrowths from the surface of the opercular disc such as occur usually in Spirobranchus, but, on the other hand, it is calcareous instead of being of horny material. There are one or two features in which the jaresent worm appears to differ from that described by Marenzeller — (a) in the extent of the basal denticulated area on the collar chsetae, and (6) in the presence of a distinct lateral notch in the collar. A comparison of the worm with that briefly described by M'Intosh under the title '' I Pomatoceros strigiceps " from New Zealand seas leads me to identify that worm withMaren- zeller's. It is certainly not Pomatoceros strigiceps, JNIorch, which is very common along our shores, and has been identified by Ehlers as P. coeruleus, Schmarda.^ The fcAV details given by jM'Intosh and his figures of the chsetse agree with the present species. The figure of the side view (PL Iv., fig. 4), though rather obscure in details, in that the outline of the collar lobes are not definitely shown, when read in association with his account of it, renders it very evident that the " peculiar region having the aspect of a partially closed fan," Ij'ing in front of the dorsal lobe of the collar, corresponds with \Ahat I term the " latero-dorsal lobe " in my account below. It seems desirable to add a short account of the worm with some figures to supplement Marenzeller's. The dimensions agree almost precisely with those given by him (PL xlviii., fig. 46). The operculum carries 3, 4 or 5 discs, which are circular, pale pinkish in colour, calcareous, with a smooth upper face, which is nearly flat or slightly concave ; there are neither outgrowths nor raised radial ridges on the surface. The vertical surface between the successive discs is feebly ridged longitudinally. In one specimen, possessing five discs, the terminal one is produced to form a smooth cone (PL xhaii., fig. 49). (Has this been worn away in other specimens ? Does this correspond to the characteristic spines of Spirobranchus ? It is A^orth noting that in some species such as »S'. maldivensis, Pixell,- the disc is unarmed and flat.) 1. Ehlers— Neuseeland. Annelid., 1904, p. 67, and 1907, p. 30. 2. Pixell— Loc. cit, p. 84. 160 " ENUEAVOim " SCIENTIFIC RESULTS. The branchial base has its ventral margin inrolled in a plane spire for about a circle ; the right and left bases arc connected by a membrane which reaches to the level of the origin of the filaments or radicles, of which there are 30 on each side, rather thick and arising m a smgle series. The collar is on the ventral surface directed forwards so as to cover the gill bases ; it is entire here, though in the median line it is folded, which gives it the appearance of an incision (PL xlviii., fig. 47). On each side close to the dorsal surface, there is an incision separating a dorsal triangular lobe, directed forwards, which is posteriorly continuous with the thoracic membrane or " palhum." Partially hidden by this dorsal lobe is a separate " latero- dorsal " lobe, slightly folded lengthwise, which is directed forwards and extends slightly beyond the free edge of the ventral lobe (PI. xlviii., fig. 43). this I take to be M'Intosh's fan-shaped region. The first bundle of chsetse (collar chsetse) is widely separated from the following dorsal bundles. It is situated at the upper end of an oblique tongue-shaped ridge, seen through the transparent " palhum " owing to its faint purple colour. It runs upwards from the level of the torus uncinigerus of the next segment, towards the apex of the dorsal collar lobe ; it appears to be a support to this lobe or to the pallium (PI. xlviii., fig. 48). The tori uncinigeri are so arranged on the thorax that their ventral margins reach successively lower levels, thvis a wide triangular area is outlined, with its open base forwards and the apex situated on or before the last thoracic segment. Gland shields occur within this area (PI. xlviii., fig. 47). The collar chsetse are about six in number, and have the usual shape (PI. xlviii., fig. 50), though the basal denticulated region is much more extensive than is shown in Marenzeller's figure. In all the other features there is agreement witji that species. Loc. — Off Babel Island, Bass Strait, 50-80 fathoms. Distribution. — Japan ; Hawaii. POLYCH.^TA. — BENHAM. 161 Genus Protulopsis, St. Joseph. Protulopsis palliata, Willey. (Pi. xlviii., figs. 51-55.) Protulopsis palliata, Willey, Ceylon Pearl Oyster Fisheries, pt. iv., Suppl. Rep., XXX., — Polychaeta, 1905, p. 316. A single female specimen of a worm in its tube. The worm measures 21 mm. excluding the gills, which are 6 mm. in length, and the thorax 6 mm. ; the diameter is 2 mm. More or less cylindrical anteriorly, the body becomes flat at the hinder end, where the long capillary bristles project from each side to a distance equal to the width of the body. There is a dorsal gland at this hinder end. There is little to add to Willey 's account ; the fact that the tori uncinigeri commence in the fourth cha^tigerous segment is confirmed. There are two curious structures on the base of the gill- support, one on the dorsal, the other on the ventral face. Dorsally, the low transverse dorsal lobe of the collar seems to be continued forwards over the base of the gill as an irregu- lar folded membrane as far as the origin of the radioles ; this membrane is broadly adherent to the gill-base (PI. xlviii., fig. 53), and I take it to be " the wide lappet, which is rolled upon itself," described by Willey as occurring in the Ceylon species. He further suggests that it " is probably able to follow the branchial spire to its termination," a statement which I don't quite understand. If, however, I am correct in ascribing the present worm to Willey's species, this " lappet " certainly cannot do this ; it is by its adherence limited to the gill-base. On the ventral face, there is on each side, well removed from the collar, a fold of skin nearly semicircular in outline which I do not remember to have seen figured or described previouslv (PI. xlviii., fig. 52). The ventral lobe of the collar is very greatly developed, forming a deep pouch in the preserved material with its opening forwards. Its connection laterally with the palhum or thoracic membrane is rather complex, and I have found it difficult to analyse in the small specimen, in which, too, the membranes very readily tear when touched. I have given figures, which I beheve to be accurate representations of their relations. 1(32 " E^•l)EAVouI^ " scientific results. I may remark that the distinction of the genus or sub- genus Protulopsis rests on a very uncertain feature. St. Joseph! separates it from Protula owing to the difference in the form of the abdominal bristles, which in the older genus are " sickle shaped," whereas in the new sub-genus they are less curved ; at the same time he points out in a note on p. 338 in referring to Filograna, that in the case of these dehcate chsetse immersion in alcohol for some time tends to alter the curvature, to diminish it, so that a sickle form may straighten out, and it is easy then to confuse the chsetse. However, I have retained the name as Willey's species is easily dis- tinguished. Loc. — East north-east of Maria Island, Tasmania, 100 fathoms. Distribution. — Galle, Ceylon. 1. St. Joseph — Ann, gci, Nat., Zool., (7), xvii., 1894, p. 263. Plate XLV. will xot be PoBLiSHELf. 162 ENDEAVOUR SC'IEMTIFIC RESULTS. I may remark that the distinction of the genus or sub- genus Protulopsis rests on a very uncertain feature. St. Joseph^ separates it from Protula owing to the difference in the form of the abdominal bristles, which in the older genus are " sickle shaped," whereas in the new sub-genus they are less curved ; at the same time he points out in a note on p. 338 in referring to Filograna, that in the case of these dehcate chaetse immersion in alcohol for some time tends to alter the curvature, to diminish it, so that a sickle form may straighten out, and it is easy then to confuse the chaetse. However, I have retained the name as Willey's species is easily dis- tinguished. Loc. — East north-east of Maria Island, Tasmania, 100 fathoms. Distribution. — Galle, Ceylon. 1. St. Joseph— Ann, Sci, Nat., ZooL, (7), xvii., 1894, p. 263. EXPLANATION OF PLATE XLVI. Nereis arenaceodentata, Moore. Fig. 1. — Head (x 8). The peristomial cirri are removed from the left side of the figure. Fig. 2. — The fifteenth i^arapodium ( X 35). The chsetse are omitted. Fig. 3. — A sub-acicular chseta from the neuropodium (X 720). Ceratonereis falcaria, Willey. Fig. 4. — Head ( X 40). Camera outline. Fig. 5. — The thirteenth parapodium (x 70). Fig. 6. — A parapodium posterior to the 21st segment with the characteristic dorsal chsetae ( x 70). Eig. 7. — A notopodial chseta from an anterior segment (X 720). Fig. 8. — A notopodial chseta from the 22nd and subsequent segments (x 720). Fig. •). — A neuropodial chseta ( X 720). Fig. 10. — The maxillary region of the pharynx (enlarged). Cheilonereis peristomialis, sp. nov. Fig. 11. — Head from the side (x 4) drawn from a specimen from New Zealand in which the peristomial lip is fully spread out. Fig. 12. — A sub-acicular < hseta from the neuropodium (X 250). Fig. 13. — A supra-aciciilar chseta from the neuropodium (X 250). Fig. 14. — The second parapodium ( x 10). Fig. 15. — The third parapodium (x 10). Fig. 16. — The ninth parapodium (x 10). The lamella is commencing to develop. Fig. 17. — The twentieth parapodium (x 10). Fig. 18.— The thirty-fifth parapodium (xlO). The dorsal lamella and the dorsal ligule are now at their maxi- mum development. BIOL. RESULT8 ' ENDEAVOUR," Vol. IV Ptate XL VI. W. li. Benham, del. EXPLANATION OF PLATE XLVII. Cheilonereis peristomialis, sp. nov. Fig. 19. — The fiftieth parapodium (x 10). Fig. 20. — The ninetieth parapodium (x 10). Fig. 21. — The everted pharynx (enlarged). Fig. 22. — A parapodium of a female Heterqnereid phase ( x 8) from a New Zealand specimen. Glycera tesselata, Grube. Fig. 23. — A parapodium { x 35). Fig. 24. — A pharyngeal papilla (x 250). Fig. 25. — The tip of another papilla, side \dew showing the terminal diiic ( x 250). Scione harrissoni, sp. nov. Fig. 26. — Anterior end from the side (x 10). Somewhat diagrammatic ; the tentacles are removed with the exception of one ; they were broken away in order that the structure of the head might be seen clearly ; the giU has been cut short. Fig. 27. — Ventral view (x 10). Fig. 28. — ^Dorsal view (x 10). The left gill is represented diagrammatically, as it had been broken in examin- ing the head. Fig. 29. — A thoracic uncinus (x 360). Fig. 30. — An uncinus from above (x 720). Fig. 31. — A thoracic capiiliform chseta (x 360). Notomastus eisigi, sp. nov. Fig. 32. — Side view of greater part of the worm ( x 4). Fig. 33. — Dorsal view of last two thoracic and anterior abdo- minal segments ( x 8). The circular marks on the ab- dominal segments are the " lateral organs," outside which are the angular " gills " at the upper end of the neuropodial tori. The notopodial tori are close together dorsally. Fig. 34. — An abdominal chseta (x 45). Fig. 35. — View of a chaeta from above (free hand) showing the double wing enclosing the hook. BTOL. RESULTS " ENDEAVOUR " Vol. IV. Plate XLVIT. W. B. Benham, cleL EXPLANATION OF PLATE XL VIII. Ophelia dannevigi, sp. nov. Fig. 36. — A parapodiuin and gill, the latter cut short (much enlarged). Fig. 37. — Ventral view of the posterior end ( x 20). The chsetse and gills are cut short on the left of the figure. Asychis victorice, sp. nov. Fig. 38 — Side view of the fragment (x 2). Fig. 39. — Ventral view of the anterior end (x 2). Fig. 40. — Dorsal view of anterior end ( x 2). Fig. 41. — A notopodial bundle of chsetse (x 35). The long capilliforms are cut short. Fig. 42. — One of the neuropodial chsetae ( x 35). Fig. 43. — The apex of a neuropodial chseta ( x 250). Fig. 44 — A portion of a capilhform chseta near its apex (X 360). Fig. 45. — One of the shorter flanged, finely pointed chsetse from a notojoodial bundle (x 250). Spirohranchus latiscapus, Marenzeller. Fig. 46. — Entire animal ( x 2). Fig. 47. — Ventral view of the anterior end ( X 4), showing the V-shaped furrow in which the ventral glands he, and behind them the transverse ventral portion of the jjaUivnn or thoracic membrane ; the gills have been cut short. Fig. 48. — Side view ( x 4) to show the latero-dorsal lobe of the collar. The gills and operculum are now omitted. Fig. 49. — Operculum ( x 6) in its fullest stage of develop- ment with a conical terminal jjlate. Fig. 50. — One of the collar chsetse (x 250). The tip is broken. Protulopsis palliata, Willey. Fig. 51. — Side view of the anterior end ( x 6). The gills cut short and the collar somewhat cUagrammatically treated. Fig. 52. — Ventral view of the collar ( x 6), sho\\"ing the great development of the ventral lobe of the collar, which forms a pouch, the base of which overhangs the first segment. On the gill-base is a curious upstanding semicircular fold. Fig. 53. — Dorsal view (x 6). The pallium or thoracic mem- brane on the left side is cut away. The peculiar anterior continuation of the dorsal lobe of the collar (?) is seen passing along the gill-base to which it is adherent. Fig. 54. — An uncinus, greatly enlarged. Fig. 55. — An abdominal chaeta (x 250). BIOL. RESULTS ' ENDEAVOUR," Vol. IV. Plate XLVIII. VV. B. Benham, del o3 III. Report on the Gephyrean Priapulus obtained by the F.I.S. " Endeavour " in Australian Waters. BY WILLIAM B. BENHAM, M.A., D.Sc, F.R.S., Professor of Biology at the University of Otagq, New Zealand ; llutton Medallist, New Zealand Institute. REPORT ON THE GEPHYREAN PRIAPULUS. CLASS GEPHYREA. Order PRIAPULOIDA. Family PRIAPULID.E. Genus Priapulus, Lamarck. Priapulus caudatus, Lamarck. Pria'pulus caudatus, Lamarck, Anim. sans Vertebres, iii., 1801, p. 467. Priapulus caudatus, Lamarck, var. antarcticus, Miohaelsen, Jahrb. Hanibm'g. Wiss. Anstalt., vi., 1889, p. 10. A single specimen of this species was received, but without any indication of the locality at which it was obtained. It has already been recorded from Antarctic seas, and the variety antarcticiis differs from the specific type only in the ventral interruption of the circles of pre-branchial papilla\ The same feature has been noted in some of the varieties from the northern hemisphere, which were originally regarded as distinct sjjecies by Ehlers, so that the value of this feature in the southern forms is of no weight as a varietal character. The matter has been discussed by Shipley, i Fischer,^ and Storikow,^ so that there is no need for me to do more than put the occurrence on record. But it appears that Priapulus has not hitherto been obtained in the southern hemisphere so far north as this must have been, for the " Endeavour " did not go into the Antarctic circle. Loc. — Not recorded. Distribution. — Greenland ; Spitzbergen ; North Sea ; Magellan Strait ; South Georgia ; Falkland Islands ; Ker- guelen ; Cape Adare. 1. Shipley — " Southern Cross " Collections, Gephviea, 1902, p. 284. 2. Fischer — Abhandl. aus. d. Gebiete d. Naturwiss., xiii., 1895, p. 22 ; Id., Fischer, Hamburg. Sammelreise magalhaens., Cephyrea, 1895, p. 0. 3. Storikow — Zoo). Anzeig., xxv., 1902, p. 1.55. IV. Keport on some Fishes obtained by the F.I.S. " Endeavour " on the coasts of Queensland, New South Wales, Victoria, Tasmania, South and South-western Austraha. PART IV BY ALLAN R. Mcculloch, Zoologist, Australian Museum, Sydney. Plates XLix-Lvin ; Text Fig. 1-2. FISHES. MCCULLOCH. 169 REPORT ON THE FISHES. Pabt 4. A CONSIDERATION of the outcome of the investigations of the " Endeavour " under the direction of the late Mr. Harald C. Dannevig proves that his contentions favouring the estabhsh- ment of the trailing industry^ in Austraha were well founded. In spite of much adverse criticism, he maintained the opinion that quantities of marketable fish were obtainable within reasonable distance of market centres, basing his arguments upon a practical knowledge of the catches of the " Endeavour" at numerous stations between Bowen on the Queensland coast, and Gerald ton in Western Australia. Some of his Reports upon the work have been pubUshedi, but he was engaged upon an analysis of the investigations, particularly those around the south and south-eastern coasts, when he was lost with his ship and all her company. Largely owing to the influence of his work, however, and imder the skilful management of Mr. David G. Stead, the Govermnent of New South Wales has now estabhshed three trawlers, and considers the project sufficiently profitable to warrant its extension in the near future. Representatives of other States are likewise looking favourably uj^on the trawling industry, so that it is possible that Mr. Dannevig 's hope of an abundant supply of trawled fish forming a valuable addition to the food-supply of several Austrahan cities may soon be reahsed. Owing to the depletion of the staff of the Australian Museum, caused by the enhstment of several of its members, the difficulty of handling the large collection of fishes made by the " Endeavour " has been greatly increased. For the same reason it has not been possible to devote so much time to their study as formerly, which will account for the brevity of the following Report. I am indebted to Mr. Stead for assistance in procuring fresh specimens of several of the rays described in the following pages. I also owe much to my assistants, Mr. Ellis Troughton and Mr. Frank McNeil for their wiUing help in dealing with the specimens referred to herein. 1. Reports by the Director of Fisheries — Commonwealth Parliam-entarv Report, 1909. No. 69. — F. 15398. 1910, No. 70.— F.16668 1910, No. 47.— F.13794 1910, No. 61.— F.16559 Notes on Australia's Fisheries, with a summary of the Results obtained by the F.I.S. " Endeavour," Melbourne, 1913 (JDept! Trade and Customs). 170 "endeavour" scientific re&ults. FamUy ALOFJIBM. Genus Alopias, Eafinesque. Alopias vulpinus, Bonnaterre. Alopecias vulpes, McCoy, Prodr. Zool. Vict., Dec. ix., 1884, pi. Ixxxviii. The head of a large example, with the pectoral fins attached, was picked up by the trawl of the " Endeavour " at a depth of 200 fathoms. It had been bitten off cleanly from the body by some larger animal, probably a Killer Whale, Orca gladiator, Bonnaterre, and was but httle decomposed when secured. It had the general appearance of McCoy's illustration, quoted above, though the upper profile of the head was very gibbous, rising from the snout in a strongly convex curve. Loc. — Off Cape Everard, Victoria. Genus Urolophus, Miiller & Henle. Urolophus, Miiller & Henle, Sitzb. Akad. Wiss. Berhn, ii., 1837, p. 117 {U. cruciatus) ; Id., Garman, Mem. Mus. Comp. Zool., xxxvi., 1913, p. 405. Trygonoptera, Miiller & Henle, Plagiost., 1838-41, p. 174 {T. testacm, Miill. & Henle) ; Id., Garman, Loc. cit., p. 409. The presence or absence of a dorsal fin has been considered of sufficient importance to distinguish Trygonoptera and Urolophus, but this character is apparently variable even within the hmits of a species. It is usually present in T. testacea, but an adult example in the AustraHan Museum having all other characters of that species is without any trace of it ; T. viridis, on the other hand, generally lacks a dorsal fin, but it is well developed in several specimens which apparently belong to this species. Variation of this character in young examples of Urolophus has already been noted by Waite^, and his observations are borne out by the " Endeavour " material. The species of Urolophus differ from Trygonoptera in having a shorter and thicker tail, but in the absence of other differ- entiating structures this can scarcely be regarded as of generic value. 1. Waite— Mem. Austr. Mus., iv, pt. 1, 1899, p. 43. FISHES. MCCULLOCH. 171 Key to the Australian species. a. Tail short, its length less than its distance from the mouth. Urolophtjs. h. Back Avith dark transverse and longitudinal markings. cruciatiis. hh. Back nearly uniform. avrantiacus aa. Tail longer than its distance from the mouth. Trygonoptera c. Internasal valve with a broad fringe on its posterior margin. Nostrils with broad lobes posteriorly. d. Tail without lateral folds. testaceus. cc. Posterior margin of the internasal valve with a narrow fringe or lobules. Nostrils without broad posterior lobes. e. Papillae behind lower jaw fewer, widely spaced. /. Disc shghtly broader than long, back uni- formly coloured. viridis ff. Disc much broader than long, back with darker cross-bars. expansus ee. Papillae behind lower jaw numerous, close together. Back white speckled. bucciUentus Urolophus cruciattjs, Lacepede. Banded Stingaree. Raja cruciata, Lacepede, Ann. Mus. d'Hist. Nat., iv., 1804, pp. 201, 210, pi. Iv., fig. 2. Leiobatus cruciatus, Blainville, Bull. Soc. Philom., 1816, p. 121. Urolophus epMppiatus, Richardson, Voy. " Erebus and Terror," Fishes, 1845, p. 35, pi. xxiv. Urolophus cruciatus, Giinther, Brit. Mus. Cat. Fish., viii. 1870, p. 485 ; Id., Dumeril, Hist. Nat. Poiss., i. 2 1865, p. 626 ; Id., Macleay, Proc. Linn. Soc. N. S Wales, vi., 1881, p. 378 ; Id., Johnston, Proc. Roy Soc. Tasm., 1882 (1883), p. 141, and 1890 (1891), p 39 ; Id., McCulloch, Zool. Res. " Endeavour," i., pt 1, 1911, p. 14 ; Id., Garman, Mem. Mus. Comp. Zool. xxxvi., 1913, p. 408. I have carefully compared twenty-two specimens of this species with numerous U. aurantiacus of similar size, and fail to find any differences between them other than colour- marking. The dark bands and spots of U. cruciatus are very constant, however, being similarly disposed in all my speci- mens, though somewhat variable in degree of development. 172 " ENDEAVOUE SCIENTIFIC RESULTS. In a fine example 242 mm. wide the markings are almost exactly similar to those of Richardson's figure, though the contrast between the fight and dark parts is more striking ; they are also well defined in two young examples and nineteen foetal specimens. Waited has noted that foetal specimens possess a small dorsal fin. It varies considerably in development in my nineteen specimens ; a few have a distinct, though smafi fin in advance of the spine ; in others it is represented by a mere fleshy tubercle, while in most it is wholly wanting. Locs. — East of Flinders Island, Bass Strait. Off Port Davey, Tasmania, 88 fathoms. Near Mainwaring Cove, west coast of Tasmania, 50 fathoms. Bay of Fires, Tasmania, 40 fathoms. Investigator Group, South AustraUa, 37 fathoms. Urolophus aurantiacus, Muller & Henle ? Yellow-hacked Stingaree. (Plate xHx.) ? Urolophus aurantiacus, Miiller & Henle, Plagiost., 1841, p. 173, pi. Ivi. ; Id., Schlegel, Faun. Japon., Pisces, 1850, p. 307. Urolophus aurantiacus, Bleeker, Nederl. Tijdschr. Dierk., ii., 1865, p. 68. Urolophus cruciatus, Waite, Mem. Austr. Mus., iv., pt. 1, 1899, p. 43, and Mem. N.S. Wales Nat. Club, 2, 1904, p. 10 (not U. cruciatus, Lacepede). Breadth of the disc 1.01-1.06 into the length from the tip of the snout to the end of the ventral fins. Tail, measured from its end to the middle of the vent, 1.2-1.3 in the length from the same point to the tip of the snout. Interocular space a little greater than the narro\\est interspiracle width, and 1.1-1.3 in the preocular portion of the head. Internasal width 2.2-2.6, and width of mouth 1.4-1.7 in the preoral portion of the head. Disc more or less angular in front, rounded laterally. The snout forms a distinct triangular projection in some speci- mens, but is a mere tubercle in others which scarcely breaks the contour of the anterior profile. Anterior margins of pectorals almost straight from the snout to the broadly rounded exterior angles ; postero-lateral margins convex, forming obtuse angles with the inner margins. Margins of ventrals convex, the angles rounded. 1. Waite — Mem. Austr. Mus., iv., pt. 1, 1899, p. 43. FISHES. — MCCULLOCH. 173 Eyes small, much smaller than the spiracles, their length 2 §-3 in the interocular space. Postero-exterior angles of the intemasal lobe separated from the hinder margin as small, rounded tubercles ; posterior margin a httle sinuous, and fimbriate. NostrUs large, confluent with the mouth, and separated by a thick median frenum. Teeth pavement-hke in the female ; each has an elongate flattened spine in the male. A median papiDa behind the lower jaw, which is sub- divided into two to four lobes ; there is another simple or bifurcate one on each side of it, and one near each angle of the jaws. Tail depressed, its width between the ventral fins sub- equal to that of the mouth ; its length from the middle of the vent is usually a httle less than that from the same point to the mouth, but it is longer in young examples. The spine is inserted at or shghtly in advance of the middle of its length ; it is depressed, with serrated edges. Caudal fin originating directly behind the end of the spine above, below the middle of its length on the lower surface ; its A\adth is subequal to that of the intemasal space, but is somewhat variable. Colour. — Rich ochreous yellow above when fresh, some- times stained \\dth pink towards the edges of the disc. An indefinite brown mechan dorsal stripe may extend from the eyes to the dorsal spine. White below, piurphsh towards the edges. In preservative, the disc becomes grey. Described from eleven specimens 245-387 mm. in total length, most of Avhich were collected by the " Thetis " Expedition. Foetal examples differ only in having the body rather more circular, and the caudal fin ahnost black. U. aurantiacus was original^ described from a specimen collected by Langsdorff at the Goto Islands, Japan^. It was united with U. cruciatus by Dumeril,^ who has been followed by most later writers, but it appears to constantly differ in its colour-marking, though a careful comparison of several specimens of both species fails to reveal any struc- tural differences. Bleeker has identified specimens from Port Jackson as U . aurantiacus, while those described above appear to be ver}^ similar to Miiller and Henle's figure of that species. Loc. — Off the Mamiing River to Port Kembla, New South Wales, 20-84 fathoms. 1. U. aurantiacus has been omitted from recent papers and catalogues of Japanese fishes. It is perhaps identical with U. fuscus, Garman. 2. Dumeril— Hist. Nat. Poiss., i. 2, 1865, p. 626. 174 " ENDEAVOUR " SCIENTIFIC RESULTS. Urolophus testaceus, Milller & HenU. Common Stingaree. (Plate L.) Trygonoptera testacea, Miiller & Henle, Plagiost., 1841, p. 174, pi. Ivii. ; Id., Diimeril, Hist. Nat. Poiss., i. 2, 1865, p. 629 ; Id., Waite, Mem. Austr. Mus., iv., pt. 1, 1899, p. 43-44 ; Id., Waite, Mem. N.S. Wales Nat. Club, 2, 1904, p. 10 ; Id., Stead, Ed. Fish. N.S. Wales, 1908, p. 119 ; Id., Fowler, Proc. Acad. Philad., 1907 (1908) p. 419 ; Id., Garman, Mem. Mus. Comp. Zool., xxxvi., 1913, p. 410. Urolophus (Trygonoptera) testaceus, Giinther, Brit. Mus. Cat. Fish., viii., 1870, p. 486. Urolophus testaceus, Macleay, Proc. Limi. Soc. N.S. Wales, vi., 1881, p. 379 ; Id., Ogilby, Cat. Fish. N.S. Wales, 1886, p. 6 ; Id., Lucas, Proc. Roy. Soc. Vict., (2), ii., 1890, p. 46. Trygon testacea, Zietz, Trans. Roy. Soc. S. Austr., xxxii., 1908, p. 292. Trygonoptera mullerii, T. henlei, and T. australis, Stein- dachner, Sitzb. Akad. Wiss. Wien, hii., 1866, pp. 479- 480, pi. vi., fig. 4-5, and pi. vii. Breadth of the disc ahnost equal to the length from the tip of the snout to the end of the ventral fins. Tail, measured from its end to the middle of the vent, 1.03-1.1 in the length from the same point to the tip of the snout. Interocular space subequal to the narrowest interspiracle width, and 1.1- 1.2 in the preocular portion of the head. Internasal width 2.8-3, and width of mouth 2.04-2.3 in the preoral portion of the head. Snout angular, but not produced beyond the margin of the disc. Anterior margins of the pectorals shghtly sinuous or almost straight from the snout to the broadly rounded exterior angles ; postero -lateral margins convex, their junc- tion with the inner margins rounded. Eyes of moderate size, their length 2.5-2.7 in the interocular space. Inner margin of spiracle with an obtuse angular pro- jection. Nostrils with large free lobes posteriorly which overhang the angles of the mouth. Internasal valve with a broad free fringe of flattened tentacular processes, which are about fourteen in number on each side ; the postero-extemal FISHES. MCCULLOCH. 175 angles form small papillae which are hidden in the mouth. Teeth pavement -like, each with a horizontal ridge in the female, which is elevated into an obtuse spine in the male. A fimbriated lip behind the upper jaw, the median fimbrise being largest. A median, bilobed papilla behind the lower jaw, with one or two smaller ones on each side. Tail depressed without lateral folds, its width between the ventral fins equal to that of the mouth ; its length from the middle of the vent is a little less than that from the same point to the end of the snout. The spine is inserted at about the middle of its length, and is depressed with serrated edges. Dorsal fin small, but well developed, placed directly before the spine. Caudal narrow, its wddth equal to about two- thirds of the intemasal space ; it originates below the hinder fourth of the spine on the upper surface, and extends forward as a ridge below almost or quite to beneath the dorsal. Colour. — Uniform cinnamon brown above when fresh, including the eyes and tail. The extreme edges of the disc and caudal fin are white. Lower' surfaces white, with a broad pale brown margin to the pectorals and ventrals. Larger specimens are darker brown, with the intramarginal portions of the disc and caudal fin blackish. The above description is based on two examples 157 and 225 m^m. wide, the smaller of which is from Queensland and the other from Port Jackson. The figure is prepared from a female specimen 244 mm. wide taken at Botany Bay by Mr. J. H. Wright. A large specimen 322 mm. wide differs from the others in lacking a dorsal fin, and in having the tail somewhat shorter, its length to the middle of the vent being 1.1 in the length from the same point to the end of the snout. Its colour is coffee-brown above and white below, with the edges of the disc broadly margined with brown. It is apparently not distinct from U. testaceus, having all the other characters of that species. Locs.- — This is the common Sting-ray of shallow water in New South Wales. It is also recorded from Queensland, Victoria and South Australia. The " Endeavour " collection includes two small examples from five miles south-west of Boomerang Hill, Fraser Island, Queensland ; 15 fathoms. I have also examined three smaU specimens 146-184 mm. wide from Port Hacking, New South Wales, which were collected by Mr.