40 BULLETIN 16, UNITED STATES NATIONAL MUSEUM. 



a^. More or less conspicuous spiues or spinules on inferomarginal plates, and also often on superomargi- 



6' Both'^eHL of marginals nearly equally developed, forming a more or less vertical lateral face to 

 ray; at least, the superomargiuals are not small and confined to abactuial surface of ray; the 

 inferomarginals not forming lateral border to ray. 

 c' Inferomarginal plates touching adambulacrals throughout ray; not separated by a series of actinal 



intermediate plates, for a part, or the whole length of ray Astropecten Gray, p. 55. 



€>. Inferomarginals separated from adambulacrals for a part or a whole of the ray by one or more 

 series of actinal intermediate plates. 

 (/'. Superomarginal plates thick and block-like; at least not very thin and confined to side wall 



of ray. 

 c". Actinal interradial areas rather small and intermediate plates few (40 per cent or less of 

 minor radius measured on dorsum; madreporic body never large and concealed by paxillae; 

 inferomarginals never defining ambitus). 

 /'. Marginal plates not tumid, but with special elevations, there being narrow vertical rather 

 shallow fascioles between the plates; actinal spinelets more or less sacculate. 

 (/'. Adambulacral armature forming a very augulated series on the furrow margin, the 

 median spine considerably longer than the rest, and with a flap of integument at tip; 

 all actinal spinelets very sacculate, and marginal spinules inconspicuous; abactinal 



paxillw with stellate, imbricating bases Bathybiaster a Dauielssen and Koren. 



g^. Adambulacral armature forming a curved series, the spinelets more numerous as a 

 rule and central spinelet not enlarged nor bearing a conspicuous terminal flap; actinal 

 spinelets frequently sacculate, and marginal spines or spinules conspicuous; abactinal 

 paxillic with subcircular or subpolygonal bases; not steW&te.. Psilaster Sladen, p. 71. 

 f. Marginal plates distinctly tumid, the fascioles between them obsolete; actinal spinelets 

 not sacculate. 

 g'. Gonads confined to interradial area of disk; not extending into arm; superomarginal 

 spines present or absent; no enlarged actinal adambulacral spines. 

 hK Anal pore present; intestinal coecum consisting of two bi- or trilobate sacs; a very 

 narrow median radial area free from papuhe; adambulacral plates with gently 

 curved furrow margin bearing numerous spinelets in a regular comb, and on actinal 

 surface of plate several parallel rows of similar spinelets; abactinal paxillfe polygo- 

 nal, independent; one or two, more or less, prominent superomarginal spines 

 sometimes present; no odd interradial actinal intermediate plates. 



Pcrsephonaster i Alcock. 



h''. Anal pore probably present; no intestinal coecum; a conspicuous median radial area 



free from papula;; adambulacral plates with an angular furrow margin bearing five 



spinelets, and a few similar ones on actinal surface of plates as in Lcptychaster; 



abactinal plates lobed, imbricating, with low convex tabula; no superomarginal 



spines; odd interradial actinal intermediate plates BlaHasterc Perrier. 



j'. Gonads extending a short distance along dorsal wall of ray, on either side; prominent 

 bristling superomarginal as well as inferomarginal spines; anus present and easily 

 detected; one or two enlarged actinal adambulacral spines. 



Thrissacanthias Fisher, p. 78. 



o IncludesPftoiasterSladen,andprobably //j/as(erDanielssen andKoren. The latter has a long apical 

 cone or funnel, and is evidently a very young individual. 



b Includes Psilasteropsis Fisher. 



« Near, if not identical with this genus, is Bunodaster Verrill, as yet not fully described. The type 

 is B. ritteri, from "California." (See Verrill, American Naturalist, vol. 43, 1909, p. 554, fig. 4.) Pro- 

 fessor Verrill has kindly sent me enlarged photographs of the abactinal and actinal surfaces of the type. 

 The abactinal plates are in the form of low convex parapaxilla;, largest over central portion of disk. 

 The general appearance of the creature is strikingly like Blakiastcr, but I have no details of its anatomy. 

 The aclinal surface is identical with that of Blakiaster, even to the details of the actinal intermediate 

 plates and the presence of the odd interradial actinals. 



