76 NATURAL HISTORY BULLETIN 



dered by a regular lateral row of several spines, and a pair of 

 larger peroral spines. 



Papular pores occur mostly singly over most of the dorsal 

 surface of the rays and disk, and sometimes between the mar- 

 ginal plates. In M. pectinatus, the type, they are few and 

 conj&ned to the rays. 



When first established by Perrier this genus included two very 

 small unlike species (M. pectinatus and M. echinulatus) from 

 the West Indies. The former should be the type. These were 

 both incorrectly figured. Later (op cit., 1894) Perrier rede- 

 scribed them in detail, and added a third. They are quite unlike, 

 as shown by the original types, which I have examined, and may 

 represent two diverse genera, owing to the different character of 

 the adambulacral armature, for in the second species there is 

 an inner longitudinal fan of three webbed spinules on each plate. 



As in Rhegaster they may both be the young of other genera, 

 such as Porania, etc. I have studied a considerable series of 

 these and allied species, but they are all small and probably 

 mostly quite immature. Therefore, the status of the type of the 

 genus seems to me rather uncertain. 



At present, it seems best to restrict the genus to those species, 

 which, like the type, have a simple transverse row of adambula- 

 cral spines. Those that have an inner longtitudinal comb should 

 be referred to Pormiiella, or to Rhegaster. 



Marginaster pectinatus Perrier. 



Marginaster pectinatus Perrier, Bull. Mus. Comp. Zool., ix, p. 16, 1881. 

 Nouv. Arch, du Mus., vi, p. 229, pi. i, figs. 4,, 5, 1884. Exp. Sci., 

 Trav. et Talism., p. 167, 1894, redescribed. Sladen, Yoj. C?hall., xxx, 

 p. 366, 1889. Verrill, 1914c, p. 18. 



The type of this species, which is preserved in alcohol in the 

 Mus. Comp. Zoology, has been examined by me. 



The adambulacral plates are small and crowded; each bears 

 an inner or furrow-spine and one or two outer ones, there being 

 but one of these distally, while on the middle part of the groove 

 there may be two, or sometimes three spines, forming a small 

 simple transverse row. These spines are all very small, slender, 

 and crowded, most so near the middle of the rays. 



The interactinal areas contain about ten to twelve plates, 



