MAUREEN E. DOWN Zjorocallida, New Order, and Doraster constellatus, New Genus and Species, with Notes on the ^oroaster idae Echinodermata: Asteroidea) A SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY ? 1970 NUMBER 64 SERIAL PUBLICATIONS OF THE SMITHSONIAN INSTITUTION The emphasis upon publications as a means of diffusing knowledge was expressed by the first Secretary of the Smithsonian Institution. In his formal plan for the Insti- tution, Joseph Henry articulated a program that included the following statement: "It is proposed to publish a series of reports, giving an account of the new discoveries in science, and of the changes made from year to year in all branches of knowledge not strictly professional." This keynote of basic research has been adhered to over the years in the issuance of thousands of titles in serial publications under the Smithsonian imprint, commencing with Smithsonian Contributions to Knowledge in 1848 and continuing with the following active series: Smithsonian Annals of Flight Smithsonian Contributions to Anthropology Smithsonian Contributions to Astrophysics Smithsonian Contributions to Botany Smithsonian Contributions to the Earth Sciences Smithsonian Contributions to Paleobiology Smithsonian Contributions to Zoology Smithsonian Studies in History and Technology In these series, the Institution publishes original articles and monographs dealing with the research and collections of its several museums and offices and of professional colleagues at other institutions of learning. These papers report newly acquired facts, synoptic interpretations of data, or original theory in specialized fields. Each publica- tion is distributed by mailing lists to libraries, laboratories, institutes, and interested specialists throughout the world. Individual copies may be obtained from the Smith- sonian Institution Press as long as stocks are available. S. DILLON RIPLEY Secretary Smithsonian Institution SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY NUMBER 64 Maureen E. Downey Zorocall ida, N e w Order, and Doraster constellatus, New Genus and Species, with Notes on the Zoroasteridae (Echinodermata: Asteroidea) SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON 1970 ABSTRACT Downey, Maureen E. Zorocallida, New Order, and Doraster constellatus, New Genus and Species, with Notes on the Zoroasteridae (Echinodermata: Asteroidea). Smithsonian Contributions to Zoology, 64: 1-18. 1970.?During a study of the Asteroidea of the Caribbean and Gulf of Mexico, a new genus of the family Zoroaster- idae was discovered. A general survey of the family revealed apparent relationships with the fossil Calliasterellidae and prompted the conclusion that both of these families were incorrectly placed in the order Forcipulatida. A new order, the Zorocallida, is therefore erected for the Zoroasteridae and Calliasterellidae. The new genus, Doraster, is described, with the type-species D. constellatus. Several other observations on the classification of the Zoroasteridae are made, including the synonymizing of three species of Zoroaster found in the Atlantic. Official publication date is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1970 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 - Price 30 cents (paper cover) Maureen E. Downey ZorOCallida, N e w Order, and Doraster constellatus, New Genus and Species, with Notes on the Zoroasteridae (Echinodermata: Asteroidea) As every asteroid systematist realizes, the order Forci- pulatida has been a catchall for starfishes which could not conveniently be placed in other orders. An arti- ficial group, of obviously polyphyletic origin, the for- cipulates should be accorded more attention than they have received from systematists concerned with the present unsatisfactory classification of the Asteroidea. Tortonese's (1958) new order, Euclasteroidea, erected to accommodate those manifest misfits, the Brisingidae, was a decided advance toward the reorganization of the Forcipulatida. It is hoped that the removal of the Zoroasteridae and Calliasterellidae from this order, proposed in this paper, will further elucidate the definition of the f orcipulates. The Zoroasteridae are the only living family heretofore referred to this order which do not have crossed pedicellariae; whose skele- ton does not form a reticulate network; and in which the mouth plates are deeply sunken in the actinostome. More important, they are also the only ones with superambulacral plates?a very unlikely character for a supposedly advanced group like the Forcipulatida. Their removal leaves only the Heliasteridae and the very large polyphyletic family Asteriidae (and prob- ably some of the fossil Uractinina) in the forcipulates. With Euclasterida and Zorocallida removed, the way should be clear for a resorting and reassessment Maureen E. Downey, Department of Invertebrate Zoology, National Museum of Natural History, Smithsonian Institu- tion, Washington, D.C. 20560. of the Asteriidae and a better understanding of their relationships. The Forcipulatida have been defined as follows: Pedicellariae of basal piece and two valves, either straight or crossed. Skeleton reticulate, forming a network of rectangu- lar or very irregular meshes. Podia in two or four rows, always with a suckered disc; with simple ampullae. Papulae on all surfaces. Paxillae never present. Spines and tubercles not occurring in groups or bundles. Madreporite always on aboral surface. Median carina frequently present. No conspicuous marginals. Aboral ossicles usually arranged in regular rows. Adambulacral plates generally short and crowded. Mouth plates frequently inconspicuous and sunken in actinostome; with ambulacral plates prominent ex- cept in Brisingidae. Generally stellate, with small disc and long tapering arms; arms five to many. Most of the statements in the above definition, drawn from a number of sources (Fisher, 1911, 1928; Hyman, 1955; Spencer and Wright, 1966), are either qualified in some way, or totally untrue for some asteroids here- tofore included in the order. Table 1 shows the more simple and straightforward definitions which result 1 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY TABLE 1.?Comparison of major characteristics of Euclasterida, Zorocallida, and Forcipulatida Characteristic Pedicellariae Skeleton Podia Papulae Paxillae Spines and tubercles Median carina Marginals Aboral ossicles Adambulacral plates Mouth plates Arms Euclasterida Always crossed. Weakly developed, not reticulate, of thin plates or transverse arches widely separated. In two rows, with large suckered disc; ampullae simple. Wanting or confined to aboral disc and arm bases. Never present. Singly, never in groups of bundles. Never present. Never present. In irregular transverse arches. Not compressed. Expanded, neither ambs nor adambs predominant; not sunken in actinostome. Always more than five. Zorocallida Always straight. Very regular longitudinal and transverse rows of imbricating plates. In four rows, becoming two distally; with large suckered disc proximally, pointed distally; ampullae bilobed. None below inferomarginals. Present only in certain fossil forms. Same. Always present. Present and well-defined. In regular longitudinal and transverse rows. Compressed. Adambs project into sunken actinostome. Never more than five. Forcipulatida Straight or crossed. Reticulate, more or less open meshwork. In two or four rows, with large or small suckered disc; ampullae simple. On all surfaces. Never present. Same. Sometimes present. If present, inconspicuous, mostly irregular. Mostly irregular. Compressed. Ambs predominant; sometimes, sunken in actinostome. Five to many. from separating the Euclasterida and Zorocallida from the remaining Forcipulatida. ZOROCALLIDA, new order DIAGNOSIS.?Disc more or less domed, with a definite arrangement of enlarged primary plates; median ca- rina on arms, with all arm plates in rather compact and imbricating longitudinal and transverse series; mouth-frame deeply sunken in actinostome; alternating carinate and noncarinate adambulacral plates; at least some tubefeet large and conical, with tiny suckered disc. CONTENTS.?Families Zoroasteridae (recent) and Calliasterellidae (fossil). DISCUSSION.?A new order, the Zorocallida, is here proposed to accommodate the Paleozoic-Mesozoic fam- ily Calliasterellidae and the recent family Zoroasteri- dae. This new order suggests a common root with the Forcipulatida and certain of the Valvatida. Like the forcipulates, the Zorocallida have pedicellariae con- sisting of a basal piece and two valves. They have both the quadriserial and biserial tubefoot arrangement, while the forcipulates may have either arrangement. In both orders, the adambulacral plates are com- pressed, and spines and tubercles occur singly, never in groups or bundles. A median carina, always present in the Zorocallida, is present also in some Forcipula- tida. The Zorocallida share with many of the Valvatida the orderly arrangement of primary disc plates, com- pact longitudinal and transverse series of arm plates, similar mouth-frame, and the presence of superambu- lacral plates. It seems possible to theorize that the Zorocallida are more ancient than the Valvatida and Forcipulatida, with which they share certain basic characters, and perhaps they may provide a clue to the descent of the valvate and forcipulate starfishes from a common origin. The strong resemblance between the Zoroasteridae and the Middle Devonian-Upper Cretaceous family Calliasterellidae was noted by C. Wyville Thomson (1873), the author of the first species described in the Zoroasteridae. In his decription of Zoroaster ful- gens, he called attention to the affinities of this species NUMBER 64 FIGURE 1.?a, Calliasterella mira (Trautschold). Note the so-called "odontophore" displaced toward the right, and the space to the left of it b, Calliasterella mira. Note the paired carinate and noncarinate adambulacral plates. (From Schondorf, 1910.) with Arthraster dixoni Forbes, an Upper Cretaceous starfish from the lower chalk of Balcome pit near Amberley, Sussex. Examination of Doraster, the new genus described herein, confirms beyond a doubt that these two families are indeed very closely related; Doraster corresponds almost plate for plate with Cal- liasterella mira (Trautschold) in dorsal aspect (the only one exposed with plates in situ), as far as can be determined from photographs published by Schondorf in 1909 (Figures la, b, 2a), and from a plaster cast of part of the type, obtained through the courtesy of Dr. A. N. Soloviev of the Paleontological Institute in Moscow. I hestitate to place these two asteroids in the same subordinate taxon, however, especially as Schon- dorfs careful and detailed study does not agree in certain details from my own observations of his photo- graphs and of Doraster. For example, Schondorf s re- construction of the mouth-frame of C. mira is mostly imaginary, as the mouth is not fully exposed in the material he worked with. In fact, Schondorf himself, in his discussion of the mouth-frame of C. mira, im- plies that his reconstruction is based on Viguier*s (1878) figure of the mouth-frame of Asterias glacialis ( =Marthasterias glacialis), an assumption he justifies by certain similarities between the ambulacral plates of C. mira and the ambulacral pieces of the mouth- frame figured by Viguier. The alternating carinate and noncarinate adambula- cral plates and the large conical tubefeet with small suckered discs are characteristics unknown in the fossil Calliasterellidae, but their presence can be inferred from a reexamination of Schondorfs photographs of C. mira. Scattered in the matrix of the specimen are numerous, mostly dissociated, plates. Among them Schondorf quite correctly identified several large cari- nate plates as adambulacrals, but he failed to notice that, in nearly every case, a smaller noncarinate plate lay next to them (Figure 16). In view of the condition of this character in the Zoroasteridae, I believe these smaller noncarinate plates are also adambulacrals which alternate with the larger carinate adambulacrals, just as in the recent Zoroasteridae. Given these simi- SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY FIGURE 2.?a, Schondorf s reconstruction of Calliasterella mira. b. Schondorf s reconstruction of cross section of arm of Calliasterella mira; compare with c, cross section of arm of Doraster consttllatus. larities, it is only reasonable to suppose that the tube- feet were also similar (although it is unlikely that we shall ever know with certainty), especially in view of the postulated relationship with the Valvatida. As my experience with fossil asteroids is limited, it would be best not to get deeply involved in a discus- sion of the similarities between other Calliasterellidae and Zoroasteridae, but a few further comments on the striking resemblance between Calliasterella mira and Doraster seem called for. Schondorfs reconstruction of C. mira shows a single, centrally located interradial plate at the edge of the disc (Figure 2a), which he calls an odontophore; however, careful examination of his photographs shows this plate is actually dis- placed to the right of the interradial center (Figure la) , and there is a hiatus between it and the next plate to the left, thus indicating the possibility that there were originally a pair of plates in this position, as is the case with Doraster. I also question Schondorfs interpre- tation of what he calls "spines" below the adradial plates on the sides of the arms; his description of them sounds more like small actinolateral plates, and again, the photographs (Figure la) help to support this view. Thus the reconstruction of the cross section of the arm (Figure 26) by Schondorf is probably incorrect. Incidentally, the species described by Kesling and Strimple (1966) as Calliasterella americana, while superficially resembling C. mira, seems a much more "primitive" animal; the so-called adambulacral plates are completely different, and there are definitely no small plates between the "adradials" and the "adam- bulacrals." Also, the type and placement of the spines is completely different. Characters of the Recent Zoroasteridae Enlarged disc plates, in a definite arrangement of primary radials, interradials, central, and perhaps additional plates, is a very important character in this family, as in the order. So are carinal plates along the NUMBER 64 mid-arm and longitudinal and transverse series of arm plates, in most cases decreasing in number of longi-rows from proximal to distal. Alternate carinate and noncarinate adambulacral plates are present in all recent Zoroasteridae, if we eliminate the very doubtful genus Prognaster. (See page 14.) Superambulacral plates are present in all members of this family, although they are present only proxi- mally in most genera, and in Zoroaster are much reduced. Large, duck-billed pedicellariae are a distinctive feature of many of the recent Zoroasteridae. No other family has these huge curved pedicellariae, with one valve longer and broad-topped. They have a decidedly bird-headed look, more like avicularia than pedicel- lariae. They are very large and numerous in Dor aster, slightly smaller and less curved in Zoroaster, shorter and heavier in Myxoderma, like Doraster in Cnemid- aster (the genus perhaps nearest to Doraster in many ways), blunter and more equal-valved in Bythiolophus, and frequently absent in Mammaster. Small straight pedicellariae are numerous in all these genera. There are no pedicellariae in Pholidaster, and they are un- known in Prognaster (if, indeed, Prognaster exists). Tubefeet, changing from straight with a large suckered disc within the peristomial cavity to stout and conical with a very small suckered disc, in four rows proximally, becoming two rows distally, are another prominent characteristic shared by all living Zoroasteridae. All the members of this family are regularly five- rayed, with long slender, tapering arms, and a relatively small disc. The bathymetric range of the family is from 100 to 2000 fathoms, and they extend from about 50? N to about 40? S, with the greatest number of species occurring in the tropical Indo-Pacific. Key to the family ZOROASTERIDAE (Modified from H. L. Clark, 1920) A. Disc plates conspicuously enlarged, more or less convex, bare, or at least not covered with spinules; disc and upper part of arms skin-covered. B. Disc plates distinctly stellate, not much swollen Dortuttr BB. Disc plates decidedly swollen, more round than stellate. C. Adradials absent; 3 rows of actinolateral plates Mammasttr CC. Adradials present; 4-5 rows of actinolateral plates CnemidasUtr AA. Disc plates plane, or, if convex, closely covered with spinules; no covering of skin on dorsal surface. B. Carinals smooth and bare; surrounded by series of skin-covered squamules; no pedicellariae Pholidaster BB. Carinals not surrounded by series of skin-covered squamules; both small straight and large duck-billed pedicellariae present. C. First superambulacral plate modified to form a conspicuous buttress connect- ing first two ambulacrals with body wall. D. Adradials present; superomarginals not over lapping carinals Myxodtrma DD. Adradials absent; superomarginals strongly overlapping carinals Bytkiolophus CC. First superambulacral plate not modified to form a conspicuous buttress be- tween ambulacrals and body wall Zoroastsr Doraster, new genus With the characters of the type-species, D. constellatus, new species. ETYMOLOGY.?Greek, dora=skin, hide, and aster= star, (m.); specific name: Latin, con-stellatus = with stars, starry (m.). Doraster constellatus, new species FIGURES 3-11 DESCRIPTION.?The plates of the disc dorsum are large, flat, stellate, smooth, skin-covered. The cen- tral plate usually bears a stubby tubercle and a similar SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY t FIGURE 3.?Dor aster constellates, side view (interradius). Note fused interradial adradials. Holotype (USNM El 1352). FIGURE 4.?Doraster constellatus (specimen cleared with chlorox). Small supplementary platelets visible between carinal and adradial plates. Para type (USNM El 1353). NUMBER 64 TABLE 2.?Measurements of specimens of Doraster constellatus Specimen Oregon 3583 Alaminos 21/ 68-A-13 Oregon 382 A B C D ? F Oregon II 10602 Albatross 2376 A B C D Oregon 2780 A B Albatross 2376 A B C D Albatross 2377 A B Albatross 2396 Albatross 2376 A B Albatross? Preservation dry dry dry dry dry dry dry dry dry dry dry dry dry dry dry ale ale ale ale ale ale ale ale ale ale Major radius R(mm) 80 121 122 * 105 * 95 110 183 66 119 109 118 239 * 180 134 110 87 134 136 57 66 46 102 Minor radius r(mm) 12 15 12 14 11 11 13 12 23 9 11 16 16 38 18 17 21 13 12 18 17 9 9 8 15 Height of disc (mm) 9 10 12 14 11 11 12 11 23 11 20 16 20 31 13 22 22 14 14 17 21 7 11.5 7 18 Height of arm at third carina (mm) 6 9 7 11 9 8 8 8 14 7 10 10 15 20 11 14 11 9 7 11 16 5 6 5 10 Number of interradial actinolateral rows 5 4-5 3 3 3 3 3 3 5 2 3 3 3 4 4 3 4 3 + + + 2 + + 3 Number of terminal actinolateral rows 2* 1 ? * 0 * 0 0 0 0 0 0 1 0 0 0 0 0 + + + 0 + + 0 Number of carinals 52* 73 69 * 80 * 60 81 99 50 73 60 72 100 * + + + + + + + + + + *A11 arms broken or regenerating. +Skin-covering obscures plates. tubercle sometimes occurs on other primary disc plates. The anus, surrounded by small spinelets, is located in one of the angles of the central plate. Surrounding the centrodorsum are 5 primary interradial plates, and between and underneath them are 5 radial plates on which the 5 pointed lobes of the centrodorsum rest; just outside and partly overlaying the primary interradials are 5 large primary radials; a pair of interradial plates, entirely concealed by skin and the overlapping primary radials, lie between and beneath the primary radials and 4 enlarged imbricating (in large specimens, fused) adradials (Figure 3) which occupy the interradial angle. Small pedicellariae and tiny granuliform platelets (Figure 4) occur between the plates of the disc. The madreporite is slightly smaller than the disc plates, raised, covered with irregular channels, and located between a primary 378-414 O?70 2 interradial and the enlarged interradial adradials; it actually rests on a pair of concealed interradials. The first carinal is enlarged, about the same size as the interradial adradials. The interradial gonads are in grape-like clusters, short, hardly extending into the arm; they are attached interadially 1 on either side of the enlarged 1st super- ambulacral plates. The jaw (Figures 5, 6) is one solid piece (made up of 2 pairs of adambulacral plates), projecting into the peristome. Behind it are 4 pairs of adambulacrals joined across the interradius by a tooth-and-socket arrange- ment strongly reminiscent of the way the ambulacral plates are joined at the top of the ambulacral groove. Internally, the front of the jaw is flaired like the cari- nate adambulacral plates and bears, on each side, 3 subequal spines completely covered with tiny straight SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY FIGURE 5.?Doraster constellatus, oral view from outside. Paratype (USNM El 1353). pedicellariae, which thus form a wreath around the peristome. External to these 3 pairs of oral spines is a single pair of long sharp spines. The next 2 pairs of adambulacral plates behind the jaw bear each a single pair of long sharp spines, and the 2 pairs behind them bear 4 to 6 similar spines. The 7th and 8th pairs of interradial adambulacrals are separated by a large duck-billed pedicellaria. The first 2 pairs of ambulacral plates are also strongly fused and greatly swollen. The ambulacrals curve sharply upward under the disc and thus support the disc in a domed position. The 2 fused pairs are buttressed against the body wall by a solid ridge formed of the first 2 (enlarged) superambulacral plates. A pore for the oral tubefoot occurs in the angle between the fused ambulacrals and the buttressing superambulacrals. Superambulacral plates of two kinds occur, alternately, but do not continue to the end of the arm; rather, they diminish in size and eventually disappear. A straight, triangular super- ambulacral plate alternates with a thin, flaired super- ambulacral plate which terminates in a broad flat disc resting on the ambulacral plates. They gradually be- come similar, small, and granuliform, and do not con- tinue much beyond the middle of the arm. The tube- feet are in 4 rows proximally, 2 rows distally. Around the mouth, they are straight and conspicuously annu- lated and terminate in a large suckered disc, but beyond the confines of the disc, they are stout and conical and taper to a tiny disc, indicating that the original, more "primitive" condition in this order was with a large suckered disc, while the pointed tubefoot is a functional adaption. On the arms, the carinal plates and 2 rows of plates on either side (adradials and superomarginals) are covered with skin, devoid of spines or spinules, but may have small pedicellariae, especially proximally. The carinals may have a central, non-spine-bearing tubercle. Adradial plates overlap carinals and superomarginals. NUMBER 6 4 FIGURE 6.?Doraster constellatus, oral view from within disc. This specimen shows also the skin covering the dorsal surface of the arms, and the wreath of small pediccllariae around the mouth. Paratype (USNM 10742). Carinals and adradials about equal in size, somewhat cruciform, wider than long; superomarginals slightly smaller, lobed or, in large specimens, somewhat trian- gular, tapering toward actinal surface. On large speci- mens, an irregular row of tiny granuliform platelets (Figure 4) occurs between the carinals and the adra- dials; a few also occur sporadically on the disc. Proximally, there are 5 rows of plates between the superomarginals and the adambulacrals covered with small sacculate spinules; the row adjoining the supero- marginals is the inferomarginal row. About halfway out on the arm, one row of plates between the infero- marginals and the adambulacrals drops out, then fur- ther out, another, until at the end of the arm only the carinals, adradials (by this time very small), supero- marginals, and inferomarginals remain. The row of plates adjoining the adambulacrals is very narrow, longer than broad, and only the edge of the plate may be visible. Most of the actinolaterals carry a long, flat- tened, appressed spine, directed upward and attached to a prominent tubercle. All of these arm plates are in regular longitudinal and transverse rows; however, the latter do not correspond to the adambulacrals. Papulae occur singly or in groups of up to 5 between the plates; none occur below the inferomarginals. The terminal ocular plates are cordiform, covering the arm tip, with 3 or more coarse terminal spines. Adambulacral plates are of two alternating sorts, carinate and noncarinate (Figure 7) ; the carinate adambulacrals project strongly into the ambulacral groove and bear 4 movable spines on a transverse row of tubercles. The spines are stout and rounded at the FIGURE 7.?Doraster constellatns, cleaned specimen, showing the carinate and noncarinate adambulacral plates. Paratype (USNM E11353). FIGURE 8.?Doraster constellatus, showing the large duck-billed pedicellariae typical of the Zoroasteridae. Holotype (USNM El 1352). NUMBER 6 4 11 ?! P \ V, hi ' * G ? . * ki '"' V ? * \ \ ft K ?i * H? , v ? ? Ik > ; ? J T , ?CV . n ?s . - W k ? ? ' ' ^*?' > s ? ^ ' ? * ? . I , ^ . n ? -111 * 1 Hf if 9 - \ >?B FIGURE 9.?Doraster constellatus, dorsal aspect. Holotype (USNM E11352). FIGURE 10.?Doraster constellatus, ventral aspect. Holotype (USNM El 1352). 12 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY point of attachment, tapering rapidly to a point; they may be curved. The 1st or 2nd spine usually bears a large duckbilled pedicellaria (Figure 8), and there are numerous small straight pedicellariae. The noncarinate adambulacrals bear 3 to 5 small spinelets on their outer surface and do not project into the furrow; they may also bear small straight pedicellariae. SIZE RANGE.?R=42 mm to 220 mm; r = 7 mm to 32 mm. Average R / r = 1 0 / l . R = major radius, from center of disc to tip of arm. r=minor radius, from center of disk to interbrachial margin. MATERIAL EXAMINED.?Oregon station 3583, 09? 16'N, 81?37'W, 280 fms., May 1962, 1 specimen; Oregon station 382, 29?11.5'N, 88?07.5'W, 190-210 fms., June 1951,6 specimens; Oregon station 2780, 11? 36'N, 62?52'W, 215-230 fms., April 1960, 2 specimens; Alaminos station 21/68-A-13, 27?37.5'N, 95?20'W, 350 fms., November 1968, 1 specimen; Oregon II sta- tion 10602, 07?46'N, 54?35'W, 299 fms., May 1969, 1 specimen; Albatross station 2376, 29?03'N, 88?16'W, 324 fms., February 1885,8 specimens; Albatross station 2377, 29?07'N, 88?08'W, 210 fms., February 1885, 4 specimens; Albatross station 2396, 28?34'N, 86?48'W, 335 fms., March 1885, 1 specimen; Albatross station unknown, winter cruise 1885, 1 specimen. LOCATION OF TYPE-MATERIAL.?Holotype (USNM El 1352) and 22 paratypes in United States National Museum; 1 paratype in British Museum; 1 paratype in Museum of Comparative Zoology at Harvard. Mammaster sigsbeei (Perrier, 1880) In view of the fact that some specimens of Doraster constellatus have previously been misidentified as Mammaster sigsbeei (Perrier), it may be helpful to include here a brief redescription of M. sigsbeei. DESCRIPTION.?The arrangement of plates on the disc is definite and unmistakable: a centrodorsal plate, next, 5 interradial plates, then 5 larger radial plates. In each interradial arc are 2 large somewhat crowded- looking plates, and over the base of each arm are 3 large tumid plates. All these plates are raised, tumid, round, somewhat bare (although they may have small spinelets around the edge). The madreporite is small and inserted between the 2 interradial arc plates and the interradial primary plate. The jaws bear 3 transverse rows of 2, 2, and 3 to 4 long acute spines, and inside the mouth, not visible without dissection, are 2 more pairs, the central pair short and blunt, and the outer pair curved away from the jaws like cow's horns, both pairs covered with small pedicellariae. Both small and large straight pedicel- lariae are present in the ambulacral groove and on the dorsal surface, especially on the disc. The number of rows of tubefeet diminishes from 4 to 2 about half to three quarters of the way down the arm. Only the carinal plates and 1 row of plates on each side extend all the way to the end of the arm. The terminal plate is as broad as long, with an indentation FIGURE 11.?Doraster constellatus, dorsal aspect of cleaned specimen. Paratype (USNM El 1353). NUMBER 64 13 on the proximal dorsal side. The carinals are over- lapped on each side by the plates of the next adjoining row. These 3 rows of plates are comparatively bare, with only a few spinules on the distal edges of the plates, and are probably covered by a thin skin. The next row of plates is covered with tiny spinules, is over- lapped by the adradials, and in turn overlaps the simi- lar row of plates below it. Both of these rows diminish and disappear before reaching the end of the arm, the lower row first, then the upper. Next come 2 rows of narrow elongate plates which disappear distally before the rows above. Each of the plates in these 4 spinose rows bears 1 larger spine on a turbercle. The adambulacral plates are alternately carinate and non- carinate, with every other one projecting strongly into the furrow and bearing 4 stout movable spines in a transverse row. These spines are frequently curved and the 1st or 2nd often bears 1 or 2 large pedicellariae. The adambulacral plates between do not project into the furrow, and have no furrow spines, but bear 2 to 4 somewhat flattened spines distally side by side in pairs. The difference between the alternating adam- bulacrals becomes less distinct distally. JUVENILE.?A specimen from Combat station 450 has R=9mm. The centrodorsal plate is by far the largest plate and bears a blunt tubercle. There are only 2 rows of plates on the arms beyond the carinals proximally and only 1 row distally. There is no ap- preciable alteration of the adambulacral plates. There are only 2 rows of tubefeet the whole length of the arm. The primary plates of the disc, except for the centrodorsal plate, are not appreciably raised and tumid. There are no pedicellariae. MATERIAL EXAMINED.?MCZ 876, cotype, 1 dry, off St. Kitts, 208 fms., Blake collection; MCZ Ex. 877, cotype, 1 dry, 28?42TST, 88?40'W, 321 fms., Blake col- lection; MCZ 4048, 8 dry, Cuba, Old Bahama Chan- nel off Punta Alegra, 195-230 fms., Atlantis stations 2981B, 2982B, 2982C, March 1938; MCZ 4049, 7 dry, west end of Old Bahama Channel, 235-260 fms., At- lantis stations 2983, 2983A, March 1938; MCZ 4053, 1 dry, Cuba, off Puerto Tanama, 295 fms., Atlantis station 3371, April 1938; MCZ 4054, 4 dry, Cuba, Old Bahama Channel off west end of Cayo Romano, 245- 255 fms., Atlantis stations 3387, 3388, April 1939; MCZ 4057, 3 dry, Cuba, off Bahia de Matanzas, 285 fms., Atlantis station 3483, May 1939; MCZ 4050, 4 dry, south end of Santoren Channel, southeast of Cay Sal Bank, 250 fms., Atlantis station 2985; MCZ 4051, 3 dry, Nicholas Channel south of Cay Sal Bank, 280- 300 fms., Atlantis station 2987, March 1938; MCZ 4056, 1 dry, Nicholas Channel south of Cay Sal Bank, 325 fms., Atlantis station 3443, May 1939; MCZ 4055, 13 dry, Cuba, off Caibarien, 245-265 fms., Atlantis stations 3431, 3434, 3435, 3436, 3437, 3438, April 1939; MCZ 4052, 1 dry, Cuba, off Bahia Matanzas, 170-255 fms., Atlantis station 3000, March 1938; MCZ 880, cotypes, 2 wet, same as MCZ 876; USNM E3967, 1 dry, Smithsonian-Johnson expedition station 23, 18?32'N, 66?21'W, 260 fms., 1933; Combat sta- tion 450, 10 dry, 23?59'N, 79?43'W, 350 fms., July 1957; Oregon, 5 dry, no station, Florida Keys, 200 fms.; Oregon station 2775, 1 dry, 11?35'N, 62?37/W, 220-230 fms., April 1960. The Genera and Species of Zoroasteridae The characters of the other genera of Zoroasteridae are firmly established by the original authors and by H. L. Clark (1920) and W. K. Fisher (1919a, b). I see no need to redescribe any of them here. A list, how- ever, of the genera and species and their known dis- tribution follows. GENERA AND SPECIES Zoroaster Thomson, 1873 fulgens Thomson, 1873 actinocles Fisher, 1919 ada mi Koehler, 1909 alfredi Alcock, 1893 angulatus Alcock, 1893 barathri Alcock, 1893 carinatus Alcock, 1893 carinatus philippinensis Fisher, 1916 evermanni Fisher, 1905 evermanni mordax Fisher, 1919 gilesii Alcock, 1893 hirsutus Ludwig, 1905 longispinus Ludwig, 1905 (?juv.) macracanthus Clark, 1916 magnificus Ludwig, 1905 microporus Fisher, 1916 ophiactis Fisher, 1916 ophiurus Fisher, 1905 orientalis Hayashi, 1950 perarmatus Clark, 1920 planus Alcock, 1893 DISTRIBUTION North Atlantic, 120-1600 fms. Aleutians, 1217 fms. Bay of Bengal, 569 fms. Bay of Bengal, 1300-1380 fms. Indian Ocean, 597-705 fms. Bay of Bengal, 1520 fms. Andaman Sea, 130-250 fms. Borneo, 415 fms. Southern California, 216? 510 fms. Washington to California, 275-815 fms. Andaman Sea, 400-500 fms. Acapulco, Mexico, 1878 fms. Eastern Tropical Pacific, 780-1320 fms. Australia, 250-450 fms. Gulf of Panama, 1671 fms. Moluccas, 700 fms. Celebes, 834 fms. California, 1059 fms. Japan, 105-475 fms. Peru, 536 fms. Laccadive Sea. 1200 fms. 14 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY GENERA AND SPECIES spinulosus Fisher, 1906 fentm Sladen, 1889 (?juv.) Bythiolophus Fisher, 1916 acanthinus Fisher, 1916 Cnemidaster Sladen, 1889 wyvillei Sladen, 1889 nudus Ludwig, 1905 squameus Alcock, 1893 zea Alcock, 1893 Dor aster, new genus constellatus, new species Mammaster Perrier, 1894 sigsbeei (Perrier, 1880) Myxoderma Fisher, 1905 sacculatus Fisher, 1905 platyacanthus Clark, 1913 platyacanthus rhomaleum Fisher, 1919 Pholidaster Sladen, 1889 squamatus Sladen, 1889 distinctus Fisher, 1919 Prognaster Perrier, 1891 (genus doubtful) grimaldii Perrier, 1891 (? seepage 14) DISTRIBUTION Hawaii, 328-558 fins. New Guinea, 1070 fms. Celebes, 559 fms. Philippines, 761-1089 fms. Gulf of California, 998 fms. Laccadive Sea, 1043 fms. Indian Ocean, 597-1200 fms. Caribbean, Gulf of Mexico, 190-350 fms. Caribbean, 200-450 fms. West coast of North America, 284-916 fms. Lower California, 284 fms. Oregon to California, 223- 550 fms. Philippines, 100 fms. Banda Sea, 140 fms. Azores, 1568 fms. The Atlantic Species of Zoroaster In the course of this study, I examined critically the Atlantic species of the genus Zoroaster. Unfortunately, the types of many species from other parts of the world were not available to me, and a complete revision of the genus was neither possible nor, at this point, par- ticularly necessary. It is appropriate, however, to in- clude here my conclusions regarding the species found in the Atlantic. The characters used to distinguish the species of Zoroaster are, on the whole, unsatisfactory. Degree of development of the adradial plates is more a function of growth than a firm taxonomic character; presence of a central spine on certain plates is untrustworthy, because of the weakness of attachment (presence or absence of a tubercle is not necessarily indicative of presence or absence of a spine) and also because this, too, may be a function of growth. Large duck-billed pedicellariae are lacking in Pholidaster and sometimes in Mammaster, but their number and placement are variable within the species of Zoroaster. The armature of the actinolateral plates apparently is also variable within a species. As with most asteroids, the number of furrow spines is a dependable character. Armature of the disc plates, degree of ensacculate spinulation, and tumidity of disc plates are mostly useable but less dependable specific characters. The ratio of minor to major radius may prove useful and if so, I would be inclined to separate Perrier's longicaudus from the rest of the Atlantic Zoroasters. Long slender arms also characterize ackleyi and seem to set it apart from trispinosus, fulgens, and diomedeae; however, gradations between ackleyi and these last three "species" lead one inevitably to the con- clusion that R/r is an environment-dependent factor. Size and placement of papular pores should, perhaps, be given more attention, but are probably a function of R/r. The synonymizing of diomedeae, ackleyi, and tri- spinosus (and possibly longicaudus) with fulgens has the unfortunate effect of rendering H.L. Clark's (1920) key to the species of Zoroaster practically use- less. With the types, however, of many of the species he listed unavailable to me for study and comparison, I feel it would be unwise to attempt a revised key here. As this is mostly a deep water family, I am sure that further synonymizing among the Zoroasteridae, par- ticularly in the genus Zoroaster, is inevitable with fur- ther study on a worldwide basis. H.L. Clark (1920) indicates that Prognaster longi- cauda Perrier is a Zoroaster; in this he is quite correct. It is probably Zoroaster fulgens. Neither Clark nor myself, however, had an opportunity to examine Prog- naster grimaldi Perrier, the type species of the genus. P. grimaldi is known only from the type specimen (Azores, 1568 fms.), and I suspect, as did Clark, that it, too, is a Zoroaster, probably fulgens. The only char- acter on which Prognaster is separated from Zoroaster is that the former has all adambulacral plates alike, the latter has the usual zoroasterid alternate carinate and noncarinate adambulacrals. Adambulacral plates alike would be a very important and unusual character in this family, but I doubt that it really exists in Per- rier's specimen. Perrier's figures in his Monaco report (1896) show an unusually wide area between the carinate adambulacrals and the figures look, in every respect, like Zoroaster fulgens; Perrier, however, states that he did not observe two kinds of adambulacral plates. As he had only one specimen, he was probably reluctant to dissect it sufficiently to determine this character. NUMBER 64 15 Zoroaster fulgens Thomson, 1873 Zoroaster ackleyi Perrier, 1883. Zoroaster diomedeae Verrill, 1884. Zoroaster trispinosus Koehler, 1895. Zoroaster bispinosus Koehler, 1909. DESCRIPTION.?A 5-lobed centrodorsum, 8 primary radials (in overlapping pairs) and 5 large inter- radials, plus five large first carinals, make up the disc dorsum. The anus is between the centrodorsum and 2 primary radials, which are single rather than being paired like the other 6. The anal pore is surrounded by small flattened spines. Small pedicellariae may be scattered over the disc. The madreporite is small, flat, channeled, and located just distal to a primary inter- radial. Each carinal plate on the arms has 2 lobes on each side and overlaps the carinal plate proximal to it. Transversely, the carinals overlap the adradials and the superomarginals overlap both the adradials and the inferomarginals. Below the inferomarginals are 6-2 rows of actinolateral plates (6 near the disc, be- coming 2 by the end of the arm). Each actinolateral plate bears a long, slender, appressed spine, directed upward, and mounted on a horseshoe-shaped tubercle, plus numerous small sacculate spinelets. The infero- marginals are similarly armed. The superomarginals and carinals frequently (but not always) bear a stout erect central spine or tubercle, and these plates, as well as the adradials, are covered with small spinelets. The carinals are broader than long proximally and longer than broad distally. The superambulacral plates are very small and do not extend beyond about the middle of the arm. The tubefeet are in 4 rows proximally, becoming 2 rows less than half way out on the arm. They are stout, coni- cal, terminating in a small suckered disc (except for those actually within the peristomial cavity, which have large suckered discs), and the ampullae are double. The adambulacral plates are alternately carinate and noncarinate; the carinate adambulacrals bear 5 spines, 2 long slender furrow spines with many small straight pedicellariae (a large duck-billed pedicellaria may re- place the many small ones) and 3 outer spines which are similar to the 3 small spines of the noncarinate adambulacrals; these small spines do not bear pedicellariae. The jaw, of 4 fused adambulacral plates, bears 2 short stout oral spines on each side in a row along the oral edge of the jaw, covered with small straight pedi- cellariae, and behind them are 2 pairs of long acute spines. DISCUSSION.?Careful examination of the three so- called species of Atlantic Zoroasters reveals that they belong to one rather variable species. H. L. Clark (1920) reviewed the Zoroasteridae very thoroughly and gave keys to the genera and species. He stated, quite correctly, that Zoroaster trispinosus Koehler is a synonym of Z. fulgens; Z. bispinosus Koehler is a lapsis calami for Z. trispinosus. The other species names which he listed from the Atlantic as having once been considered species of Zoroaster have been put into other genera in the family (Z. sigsbeei=Mammaster sigsbeei; Z. longicauda=Prognaster longicauda). P. longicauda, however, is definitely a Zoroaster and must be returned to that genus. Without more material, it is not possible to decide if this is a distinct species of Zoroaster or another synonym of Z. fulgens. The types of Zoroaster fulgens could not be located; however, a specimen from Wyville Thomson's own collection, identified by him and now in the British Museum (Faroe Channel, 500 fms.) has been exam- ined, as have the types of Z. ackleyi and Z. diomedeae. In addition, several hundred nontype specimens of the three so-called species have been examined and com- pared. Presence, absence, or degree of development of the superomarginal spines, and relative width of the carinal plates are the sole characters on which the Atlantic Zoroasters have been separated. These charac- ters, vague and unsatisfactory to begin with, prove extremely variable throughout the range of the genus in Atlantic waters. As the oldest name, Zoroaster ful- gens has priority, and Z. ackleyi and Z. diomedeae are junior synonyms. On the whole, Zoroaster fulgens from the northern part of its range tends to be more spinous and more robust; also, it is generally found below 1000 fathoms. In the southern part of its range, it is more compact of skeleton, less spinous, and the arms, because of the more compact skeleton, seem to be more slender; it is found in waters as shallow as 200 fathoms. MATERIAL EXAMINED.?USNM 9066, Albatross sta- tion 2052, 39?40'N, 69?21'W, 1098 fms., August 1883, 7 specimens (syntypes of Zoroaster diomedeae) ; USNM 26258, Albatross station 2043, 39?49'N, 68?- 28'W, 1467 fms., July 1883, 1 specimen (syntype of Z. diomedeae) ; USNM 14050, Albatross station 2043, see above, 6 specimens (syntypes of Z. diomedeae); 16 SMITHSONIAN CONTRIBUTIONS TO ZOOLOGY USNM 14049, Albatross station 2095, 39?29'N, 70?- 58'W, 1342 fms., September 1883, 40 specimens (syn- types of Z. diomedeae); USNM 14048, Albatross sta- tion 2035, 39?26'N, 7O?O2'W, 1362 fms., July 1883, 73 specimens (syntypes of Z. diomedeae); USNM 9017, Albatross station 2035, see above, 7 specimens (syntypes of Z. diomedeae); USNM 9019, Albatross station 2051, 39?41'N, 69?20'W, 1106 fms., August 1883, 5 speci- mens (syntypes of Z. diomedeae); USNM 9223, Alba- tross station 2084, 40?16'N, 67?05'W, 1290 fms., September 1883, 1 specimen (syntype of Z. diome- deae) ; USNM E4944, Dry Tortugas, Florida, collected by P. Bartsch, 13 specimens; USNM 14274, Albatross station 2077, 41 'WN, 66?02'W, 1255 fms., September 1883, 1 specimen; USNM 14275, Albatross station 2084, see above, 2 specimens; USNM 7864, Albatross station 2208, 39?33'N, 7l?16'W, 1178 fms., July 1884, I specimen; USNM 11265, Martha's Vinyard, Massa- chusetts, 1390 fms., 2 specimens; USNM 14830, Alba- tross station 2684, 3 9 ^ 5 ^ , 70?54'W, 1106 fms., July 1886, 1 specimen; USNM 12041, Albatross station 2571, 40?09'N, 67?09'W, 1356 fms., September 1885, II specimens; USNM 11263, Albatross station 2562, S g 0 ^ ^ , 71?25'W, 1434 fms., August 1885, 54 speci- mens; USNM 11265, Albatross station 2564, 3 9 ^ 2 ^ , 71?23'W, 1390 fms., August 1885, 80 specimens; USNM 11262, Albatross station unknown, 1885, 10 specimens; USNM 15544, Albatross station 2732, 37?27'N, 73?33'W, 1152 fms., October 1886, 6 speci- mens; USNM 17999, Albatross station 2196, 39?35>N, 69?44'W, 1230 fms., August 1884, 1 specimen; all of the above specimens were identified by A. E. Verrill as Zoroaster diomedeae. USNM E5696, South of Dry Tortugas, Florida, col- lected by W. Schmitt, August 1932, 1 specimen; USNM 18507, Albatross station 2655, 27?22'N, 78?- 07'W, 338 fms., May 1886, 5 specimens; USNM 10422, Albatross station 2394, 28?38'N, 87?02'W, 420 fms., March 1885, 2 specimens; USNM 18450, Alba- tross station 2396, 28?34'N, 86?48'W, 335 fms., March 1885, 2 specimens; USNM 18504, Albatross station 2659, 28?32'N, 78?42'W, 509 fms. May 1886, 1 spec- imen; USNM 18505, Albatross station 2380, 28?- 02'N, 87?43'W, 1430 fms., March 1885, 1 specimen; USNM 18506, Albatross station 2381, 28?05/N, 87?- 56'W, 1330 fms., March 1885, 1 specimen; USNM 38243, Fish Hawk station 7281, 24?13'N, 81?58'W, 304 fms., February 1902, 1 specimen; Alaminos sta- tion 3C/68-A-7, 27?42'N, 87?44/W, 1500 fms., July 1968, 1 specimen; Oregon station 1537, 24?29'N, 83?27'W, 212 fms, June 1956, 2 specimens; Oregon station 2574, 26?34'N, 89?53'W, 1450 fms, July 1959, 4 specimens; Oregon station 3561, 16?35'N, 80