A. E. Yerrill — Revision Genera and Sjyecies of Starfishes. 203 



Asterodon Per. has a pair of dentary spines on each jaw.* Type, 

 A. singularis (M. and T.). Peru and Chili. 



Goniodon Per. has some of the distal marginal plates enlarged. 

 A pair of recurved spines on each jaw. Type, G. dilitatus Per., 

 New Zealand. 



Gnathaster Sladen (restr.) has the marginal plates regularly de- 

 creasing. One recurved spine on each jaw. Type, G. j^&dicellaris 

 Per.= G. meridionalis (t. Leip.), Cape Horn. 



As thus limited and defined Gnathodon is identical with Odontas- 

 ter Ver., of earliea* date. The type cited, however, is not a typical 

 Odo7itaster. 



When Sladen established his genus Gnathodon he included in it 

 all the known forms belonging to the three divisions proposed by 

 Perrier. He did not designate any particular species as the type. 

 His personal studies and detailed descriptions and figures were de- 

 voted to three Antarctic species, all of which would go in the genus, 

 as restricted by Perrier. 



Bell, 1893, combined all the known forms under the name Odon- 

 taster Ver., which is the earliest generic name in the group. The 

 latter was based by me on a single species {O. hispidus), which is 

 one of a group of species having very spinulose paxillse and plates, 

 and is apparently not strictly congeneric with O. m,eridio7ialis Smith 

 (= (9. i:)edicellaris Per.), which Perrier cites as the type of Gnatho- 

 don, 1894. 



It seems to me unnecessary, therefore, to consider Gnathodon 

 (Perrier, restr.) a synonym of Odontaster. 



Since Odontaster was originally established for a more restricted* 

 group, Sladen's name was not originally truly synonymous with it. 

 Therefore it may well be restricted to one of the other subdivisions 

 included in it by him, as Perrier has done imperfectly. 



Another division may be established for G. elongatus SI., and G. 

 miliaris Gray, in which the abactinal skeleton consists of pseudo- 

 paxillse, or low rounded plates covered with granules, while granules 

 also cover the actinal and marginal plates, thus giving them nearly 

 the same appearance as species of Tosia and allied genera, for 

 which, indeed, some of them were formerly mistaken and described 

 (under Aslrogoniiim by Gray, and others ; and under Pentagonaster 

 by Perrier). 



* Leipoldt, 1895, figures an abnormal specimen of A. singularis in which two 

 jaws have each but one median tooth (pi. xxi, fig. 7c). 



