60 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



Genus CRASPIDASTER Sladen. 



Craspidaster Sladen, 1889, p. 175. Type, ArcJiaster hesperus Miiller and 

 Troschel. 



CRASPIDASTER HESPERUS (Muller and Troschel). 



Plate 9, fig. 3. 



Archaster hesperus Muller aud Troschel, 1840, p. 104 ; 1842, p. 65. 

 Stellaster sulcatus Mobius, 1859, p. 11, pi. 5, figs. 1 aud 2. 

 Craspidaster hesperus Sladen, 1889, p. 177, pi. 17, figs. 5-7 ; pi. 18, figs. 1-4. 

 Astropecten macer Sluiter, 1889, p. 297 ; 1895, p. 53. 



Sladen has given figures and a careful description of this species 

 and has noted the principal variations. 



The largest of the Philippine specimens has R=52 mm., r=12 mm., 

 11=4.33 r; superomarginal plates 38; furrow spines, 6-7. In this 

 specimen the little thumblike subambulacral spinelet described and 

 figured by Sladen in his Hongkong specimen is not present, and 

 the fringe of spinelets along the inner edge of the inferomarginal 

 plates is found much farther along ray than indicated by Sladen. 

 It is variable even in the Philippine examples; in some it persists 

 nearly to the end of the ray (station 5209) while in an example from 

 station 5375 the fringe is rudimentary beyond the limit of actinal 

 intermediate plates. 



Anatomical notes. — The midradial line and center of disk are with- 

 out papulae. Here the plates are nearly circular or with 5 or 6 faint 

 scallops by which the plates overlap slightly. Elsewhere the plates 

 have 5 or 6 short lobes b}^ which they are in connection with neigh- 

 boring plates. This arrangement is unlike that of Prionaster and 

 Goniopecten^ in which the plates are independent. The papulae are 

 usually in sixes about each plate. In the family Astropectinidae 

 all gradations between independent circular or elliptical plates and 

 stellate imbricated ones are found. 



The stomach is spacious and from the dorsal part arise five radial 

 diverticula, each of which, dividing, gives rise to two hepatic coeca 

 which extend along ray to about the seventh superomarginal. Thus 

 the coeca arise from a common tube, and not independently as in 

 Prionaster. The ventral portion of the stomach, which is very ex- 

 tensible, is not separated from the aboral portion by any sharp line 

 of division. Its radial portions contain small gastropods and pele- 

 cypods. The intestinal coecum is represented by a single large sac- 

 like diverticulum lying in an interradius on the dorsal surface 

 of stomach, its extremity nearly reaching the marginal plates. I 

 can not determine whether this connects with the stomach, but a 

 minute intestine passes upward in the interior of a slight epiproctal 

 cone. A lumen is present, and presumably the intestine opens by a 

 tiny pore, although I have not been able to see it. The paxillae are 



