A. E. Verrill— Revisioyi Genera and Species of Starfishes. 149 



Y. Calliaster proposed by Gray, in 1840, for the single type- 

 species, C. Childreni^ is very distinct from the genera already named, 

 not only on account of the spinose plates of both surfaces, but also 

 by reason of its very different adambulacral spines. 



VI. Hosia. When Gray established this genus, in 1840, he 

 referred to it only H. flavescens. Perrier (1876) has redescribed the 

 types of this species and refers them to two distinct species of true 

 Anthenea (Gray, 1840). Therefore Hosia becomes a s3nionym of 

 the latter. In 1847 and 1866, Gray added another species {H. spi- 

 nidosa) to Hosia, but according to Perrier (1876), who reexamined 

 Gray's type, this species belongs to a different genus. Fie referred 

 it to his section C of Pent ago7i aster. It has spinulose marginal 

 plates, and also valvular pedicellariae. It is probably an immature 

 species of Tosia, or of some closely related genus. 



VII. The names proposed by Miiller and Troschel, in 1842 (Syst. 

 Ast.), are next in order. 



Astrogonium, as stated above (p. 145), was formed by uniting four 

 of Gray's genera. It thus became a composite group without any 

 definite type, and not very different from the original group called 

 Goniasier (1st section) by Agassiz. In 1847 and 1866, Gray applied 

 the name to a more limited group, including A. granularis (Retz.), 

 which is nearly allied to Tosia, together with species now referred 

 to Odontaster. 



If it were to be iised at all in the modern system, it should be 

 restricted to the group containing A. granularis. But as it was 

 an artificial group and should have had no real status originally, 

 it should properly drop out of use except as a synonym of Gray's 

 genera. 



By Sladen (1889) Astrogonium was restricted to Gray's genus 

 Pentagonaster=Stephanaster Ayres. 



Perrier (1894) has used it improperly for a very different group, 

 {=^ Pseudarchaster + Aphroditaster Sladen), including several deep-sea 

 species, none of which were known to M. and Troschel, nor to Gray. 



VIII. GoniodisGus. M. and Troschel (1842) constituted this 

 genus by reuniting five of Gray's genera, together with forms un- 

 known to Gray. Perrier (1894) has very judiciously restricted the 

 name to those species that have stellate or 6-lobed abactinal plates, 

 included in it by M. and Trosch., such as G. cuspidatus (Lam.), and 

 in this sense it should be adopted. 



The genera proposed subsequently to those already mentioned 

 have given rise to no great confusion and therefore need not be dis- 

 cussed here. 



