Fisher — Notes on the Systematic Position of Starfishes. 5 



cceca with long subdivisions, so that eacli ray appears to have from 

 six to ten separate ccBca of unequal length ; tube feet with well- 

 developed sucking disks; lateral abactinal plates not cruciform 

 nor regularly imbricated 



Radiaster [Radiaster elegans Perrier, R. tizardi 

 (Sladen) and R. notabilis (Fisher)]. 



2. Gonads consisting of numerous tufts extending in a radial series near 

 the superomarginal plates for over half the length of the ray; 

 hepatic coeca two, not appearing multiple on account of long sub- 

 divisions as in the preceding; tube feet with very small sucking 

 disks; lateral abactinal plates distinctly four-lobed, regularly 

 imbricated 



Mimastrella gen. nov. [Genotype Mimastrella cognata = 



Mimaster cognatus Sladen].* 



SoLASTER AND Crossaster. — Tliese two genera have been 

 united by most recent writers,! although in practice it is not 

 very difficult to recognize them. Since new species of the 

 Crossaster type are continually being described, it is becoming 

 more and more desirable to keep them separate. A good 

 differential character is the presence in Crossaster of a complete 

 membraneous interbrachial septum between the internal inter- 

 radial calcareous dorsoventral pillar and the margin. The 

 pillar arises from the mouth plates and passes upward, its point 

 of union with the abactinal skeleton being usually marked by a 

 smooth spineless area. In Crossaster papposus between this 

 calcified buttress and the margin there is a definite septum 

 separating the gonads of adjacent rays, while in Solaster endeca, 

 S. borealis and S. abyssorum the pillar is present, but not the 

 membraneous septum; as a result the gonads of adjacent rays 

 are not separated, and the ccelom is continuous. My recently 

 described Solaster scotophilus has a. complete membraneous septum 

 and the outward habit of Crossaster papposus . It must therefore 

 be classified as a Crossaster. 



' Mimaster can not be used for this group because when described the genus was 

 monotypic. The genotype, M. tizardi, being congeneric with Radiaster elegans, the 

 name Mimaster becomes strictly a synonym of Radiaster. 



+ For some of the reasons for uniting them see "Asteroidea of the North Pacific," 

 p. 329. 



