238 VERRILL 



a single comb-like row of six to twelve, on some of the plates. The 

 peractinal plates bear a smaller double group of similar spines, mostly 

 six to ten in number. The spines on the actinal surface of the adam- 

 bulacral plates stand in two rows, or often only one row distally, 

 increasing regularly in size and length to the one or two unpaired 

 ones at the edge of the groove. These are rather large, slightly 

 clavate, and obtuse. The intra-ambulacral spine is much smaller and 

 shorter, acute or subacute, not very slender. The two apical peroral 

 spines are rather stout, subacute ; the epiorals and adorals are similar. 



Taken at Dutch Harbor, Unalaska, Yakutat, Fox Cape, Sitka, etc. 

 (Harriman Expedition). Common. 



Recorded by Dr. Fisher (as eschrichtii) from numerous localities 

 between Bering Straits and Yakutat, Alaska, from the shore to 86 

 fathoms, and on the Asiatic side to Kamchatka and the Kuril Islands. 



Dr. Fisher records specimens of this species carrying eggs, taken 

 June 8, 10, and 14, 1906; and one carrying young up to 4 mm. in 

 diameter with eight pairs of ambulacral feet to each ray ; taken by 

 the " Corwin," in Bering Sea. (See Fisher's pi. lxviii, fig. i.) 



Probably this subspecies has formerly been confused with H. san- 

 guinolenta, which it much resembles, superficially. 



It can best be distinguished from H. sanguinolenta by the larger 

 and decidedly transversely oblong inferomarginal ossicles and their 

 clusters of spines, of the same form. These plates in H. sanguino- 

 lenta are usually smaller, scarcely oblong, and bear smaller divergent 

 clusters of spines. This is also, in most cases, a stouter species, with 

 shorter rays and much larger disk. Other differences appear on 

 closer examination, especially in the dorsal skeleton and marginal 

 plates, and in the much narrower and simpler rows and longer adam- 

 bulacral spines. 



It is so unlike H. levittscula that there is no need to compare the 

 two, in detail. 



This subspecies is, apparently, quite unlike H. eschrichtii Miiller 

 and Troschel, as already remarked. The latter was described as 

 having the proportions i : 4, or about those of ordinary specimens of 

 typical H. sanguinolenta, from which it differs less than var. pec- 

 tinata, or subsp. miliaris. I consider it nearly the same as var. pec- 

 tinata. 



One specimen, taken with the type of H. tumida, has nearly the 

 same spinulations as the latter, both above and below, and yet does 

 not have the oral region inarched. It has, however, a somewhat 

 smaller and less turgid disk. Its radii are 8 mm. and 20 mm. ; ratio, 

 1:2.5. 



