WEST INDIAN STARFISHES 107 



swelling where the conical tubercles would have appeared later. 

 In some the dorsal plates are entirely covered with granules, but 

 in others the central area is naked, the amount of naked surface 

 increasing with age, but not regularly so. The papular pores in- 

 crease in number with age and cover more and more of the 

 median radial areas and the central area of the disk, but these 

 areas have no sharp boundaries. 



The type was from off South Carolina, in rather shallow water. 

 This species was taken by the Albatross at twelve or more sta- 

 tions in the West Indian region, in 21 to 50 fathoms, and by the 

 Blake at three stations, in 84 to 129 fathoms (all young). 



This, or an allied species, has been recorded from Brazil, under 

 the names G. ciispidattis or G. semilunatus, which properly be- 

 long to the Indo-Pacifie species. I have seen no Brazilian speci- 

 mens. 



Several variations (Nos. a-h), due largely to age, have been 

 described by the writer. (See Revision Genera, pp. 151-155, 

 1899.) Details of the distribution are also given in that article, 



PUnthaster dentatus (Per.) Fisher, 



Pyrenaster dentatus (Per., p. 242, 1884). Verrill, 1899, p. 167 = PUnthaster 

 dentatus Fisher, 1911b, p. 165. (Notes on type.) 



This species was recorded by Perrier (1884, pp. 170, 242), as 

 from Blake stations, Nos. 100, 111, 260, in 250 to 1,500 fathoms, 

 and by me (1899) from 478 to 1,639 fathoms. 



Also recorded by Perrier from Blake station 264, in 41 fathoms. 

 It is evidently a deep-sea species, of a deep-sea group. This last 

 record is typographically erroneous, Blake station 264 was in 416 

 fathoms, gray ooze off Grenada. Perrier (1884) gives the depth 

 correctly on page 180. (Table of localities, but erroneously in 

 other places in the same report.) 



Subfamily Mediastebin^ Verrill. 



Mediasterince Verrill, Kevision Genera, p. 178, 1899; op. cit., 1914a, p. 294. 



This subfamily was proposed in 1899, for Mediaster and sev- 

 eral allied genera, closely related to GoniasterimE, but having the 

 dorsal plates in the form of parapaxillae, besides other characters. 



Since its establishment large numbers of related new genera 

 and species have been discovered by several deep-sea explora- 



