EASTERN TROPICAL PACIFIC ASTEROIDEA. 97 



C. Carinal plates of ray smooth and bare surrounded by series of skin-covered squamules. 



Pholidasler. 

 CC. Carinal plates not surrounded by sorios of skin-covorrd squamules. 



D. A conspicuous buttress (a specialized supcranibulacral plate) connects first two anibulacral 

 plates with body-wall; supcranibulacral plates present. 



E. Adradiai plates present; superomarginals not overlapping carinals Myxoderma. 



EE. Adradiai plates wanting; superomarginals strongly overlapping and dominating the 



carinals Bylhwlophus. 



DD. Superambulacral plates wanting, and hence no buttresses between ambulacral plates and 

 body-wall Zoroaster. 



Maxoinaster. 



Perrier, 1894. Travailleur et Talisman Stell., p. 125. 



Type-species: Zoroaster sig.'^beti Perkier, 1880. Comp. Eend. 91, p. 436. 



This genus is still monot3T)ic, for the allied East Indian species all differ in 

 the character of the marginal plates at the base of the arms Alcock (1893. Ann. 

 Mag. Nat. Hist., ser. 6, 11, p. HI) considers tlu-ee of his eight Zoroasters con- 

 generic with sigsbeei, but he failed to note this particular feature which seems 

 of no Uttle importance. Fisher makes no reference to it and hence it does not 

 seem to have attracted his attention. His work has shown that Alcock's three 

 species are Cnemidasters but he is not sure that Mammaster can be maintained 

 as distinct from that genus. Manunaster is e\adently not common as I have 

 found no published records of its occurrence other than those in the Blake 

 Report — one station near St. Christopher in 208 fms. and one in the Gulf of 

 Mexico is 321 fms.; seven specimens altogether. The largest of the Blake 

 specimens had R about 75 mm. Perrier selected a much smaller specimen 

 (R = 49 mm.) for his description and does not refer to the large one. 



Cnemidaster. 



Sladen, 1889. Challenger Ast., p. 423. 



Type-species: C. wynllii Sladen, 1889. Loc cit. MonotJ^lic when established. 



The collections made by the Albatross in the Far East have enabled Fisher 

 to revolutionize our knowledge of this genus, which was based on a single very 

 inunature specimen. Fisher had seven specimens at his disposal and has been 

 able to make clear some of the growth-changes. These bring out the impor- 

 tant fact that the Zoroasters described under the names nudus, squameus, and zea 

 are aU congeneric with wyvillii. Indeed the four species are so closely related, 

 it is exceedingly difficult to separate them properly. The following key is the 

 result of studying cotypes of nudus and squameus, and the descriptions and 



