208 s. goto: 



the arm. The plates are much abraded in all my specimens, 

 but they appear to be covered with subcapillary spinelets similar 

 to those of the superomarginals. 



Localitij. — My specimens are from the vicinity of Misaki, two 

 from Okinosé from depths of 480 m. and 1120 m., the other from 

 Yodomi from a depth of 160 m. 



Three dried specimens in S. C. 



Dr. W. K. Fisher thinks that this species possibly belongs to 

 his genus Thrissacanthias, in which the gonads extend some way 

 into the rays, while in Persephonaster they are confined to the 

 interradii. The dried condition of my specimens does not enable 

 me to decide this point. 



Persephonaster misakiensiSf n. sp. 



(PL n, figs. 23-33.) 



I have two specimens of this species in alcohol. Both are of 

 good size and look externally very much like Astropecten ; but the 

 presence of the anus as well as certain features in the armature 

 of the body make it necessary to refer them to Persephonaster ^ 

 which is distinguished from Astropecten chiefly by the presence of 



1) I may here express my opinion parenthetically that so far as can be judged from the 

 published descriptions, the distinction between FMonaster and Persephonaster appears to be 

 vague and uncertain. The former genus was set up by Sladen [ '84, p. 81], to whose work one 

 must refer back for the original diagnosis and the descriptions of typical species. The genus 

 Persephonaster was created by Wood-Mason and Axcock [ '91, p. 430] and more precisely defined 

 by Alcock [ '93, p. 81]. As remarked by Ludwig [ '97, p. 117] in connection with his description 

 of FMonaster subine^inls of the Mediterranean, the presence of the superambulacral plates is a 

 character which does not agree with Sladen's definition of the genus ; but that aside from this and 

 a certain peculiarity in connection with the genital orgiins, the Mediterranean species can .nost 

 naturally be brought in that genus. In fact PMonaster suhinermis and Plut, abyssîcola Ludwig 

 [ :05, p. 30] appear to me to be intermediate forms between the typical species of the two genera 

 in question. Ctenophoraster haioaiiensis Fisher [:06, p. 1015] seems to be nearly allied to 

 Persephonaster misakiensis which is intermediate in character between the former and the typical 

 Persephonaster species. 



