A. E. Verrill — Mevision Genera and Species of Starfishes. 159 



The typical species appears to be destitute of pedicellariae, but 

 small, high, pincer-like pedicellarise, with spatulate blades or chisel- 

 shaped blades, occur in many of the allied species more recently 

 described (e. g. Perrieri, Vincenti, hcesitatis). 



In many of the species of this group all the dorsal marginal plates, 

 except the last or two last pairs, are separated by one or more rows 

 of abactinal plates. But in several deep-sea species (Perrieri, etc.) 

 three to five distal pairs of marginal plates are in contact medially. 

 In some species different individuals have been found to vary, in this 

 respect, from those having only one pair joined, to those with three 

 or four pairs joined. 



Considerable variations also occur among the species, in respect to 

 the character and arrangement of the adambulacral spinules, in the 

 number and arrangement of the papular pores, and in the form of 

 the abactinal j^lates. 



Most of these characters are not sufficiently constant nor important 

 for generic divisions, but may well afford grounds for dividing the 

 group into convenient sections. (See p. 180.) 



Sladen, in his great work on starfishes, included in his genus Pen- 

 tagonaster not only those that are here separated as typical Goniaster, 

 but also others that apparently belong to Mediaster and Iloplaster, 

 besides some that belong perhaps to undescribed genera. 



To Mediaster I refer three of his new species : viz. P. Japonicus, 

 P. Patagonicus, and P. arcuatus. But as the existence of internal 

 connecting ossicles between the abactinal plates has not been ascer- 

 tained for either of these species, this reference is based on the 

 general appearance and on the character of the plates, spinules, and 

 pedicellariae. 



His P. lepidus appears to be a true Hoplaster. It has odd inter- 

 radial marginal plates and all the plates are spinulose. P. gibbosus 

 Perrier also appears to be generically distinct, as well as P. inter- 

 jnedhis and P. dentatus. If these forms be eliminated, the genus 

 becomes more homogeneous and better capable of definition, though 

 it still remains an extensive group. 



Pei'rier, in his later works, has generically separated numerous 

 species that he formerly referred to Pentagonaster, such as Mosaster, 

 Odontaster, and the forms that he refers to Porigona. But some 

 of the other species that he has described as belonging to this genus, 

 especially P. intermedius and P. dentatus, also appear to be worthy 

 of generic distinction. 



In this article, I have constituted several new genera to include 

 some of these peculiar forms, hitherto referred to Pentagonaster, 

 together with some new species. 



