JAPANESE ASTEROIDEA. 93 



plates and compactness of paxillge. Two extremes, both from station 4223, 

 will serve to illustrate variability of j)axill8G. Eacli specimen lias 11 = 3é mm. ; 

 in A, paxillEG in neiglibom-liood of madi-eporic body have seven to twelve 

 sx^inelets, occasionally as few as fom- on very small ones; of these never 

 more than one is situated in centre of tabulum and rather more than half 

 the paxillae have no central spinelet at all ; in B the x^axillse ordinarily have 

 from twelve to twenty-two sjpinelets, of which tln*ee to five occupy centre of 

 tabulum, and all or very nearly all have central spinelets, those pasillïe on 

 outer part of ray having so few as one central spinelet.^^ When these two 

 specimens are placed side by side the difference is very striking. Many of 

 the specimens, both small and large, lack a central spinelet to paxillse 

 altogether, thus resembling exactly some Atlantic examples. There is as 

 much if not more difference in the extremes of these specimens than is shown 

 by Sladen's two figaires illustrating paxillae of Ctenodiscus austrcdis and G, 

 •procurators^ while in the extremes of body-form the difference is gi'eater tlian 

 between figm-es 1 and 7, illustrating the same two species. I thought at 

 first that the difference in size of madreiDoric body might furnish a character 

 of some constancy, to separate hrcmsei from crispatus, the latter having the 

 larger body. Tliis character also is very variable in Pacific specimens, some 

 examples having fairly large, others small madreporic bodies. In four rather 

 poorly preserved Atlantic sx^ecimens the madreporic body is more constant 

 and is one and one-half times gi'eater in diameter than in Pacific examples 

 of the same size. 



"The shaft of the paxiUa varies in length. In specimens from very 

 deep water it is longer than in shallow water specimens. 



" The bases of j)axill8e, or the abactinal ' j)lates,' are circular and rather 

 closely placed, usually not quite toucliing. They are largest about one-half 

 r from centre of disk, decreasing in size toward tip of ray and centre of disk. 

 Along mid-radial line where there are no papulae the plates are smaller and 

 usually broadly elliptical. Papulae are not regularly aiTanged, fom to six usual- 

 ly occiuTing about a plate. They are single, and are lacking on a cii'cular 



1) " Atlantic specimeus show the same range of variation." 



2) " Challenger Asteroidea, pi. 30, figs. 4 and 9." 



