JAPANESE ASTEROIDEA. 



137 



The number of marginal plates is certainly of some systematic 

 importance, but in my opinion only in a subordinate degree. 



The relative extent to which the superior and inferior marginals 

 project laterally has been taken hold of by some observers for 

 distinguishing species. But this also appears to me to be hardly 

 of systematic importance. Tt is true that in certain species the 

 inferomarginals are always more projecting than the supero- 

 marginals ; but in the present species, while the inferomarginals 

 stand prominently outward in some specimens, they are flush 

 with the superomarginals in others. 



Again, the relative breadth of the paxillar area and the 

 superomarginals is, in my opinion, only of a very secondary im- 

 portance for specific distinction ; much depending in this case on the 

 contraction of the different muscles when the specimen was killed. 



Most characteristic of all are the armature of the supero- 

 marginals, the lateral spines of the inferomarginals, the adam- 

 bulacral spines, and the armature and form of the mouth-plates. 

 The paxillae are also characteristic of species, and the position 

 and form of the madreporite may also be taken into account. 



