246 



s. GOTO 



is such a bewilderiug number of more or less perfect intermediate stages 

 tliat one is forced to range them all under one head. 



" The most imix)rtant A'ariations occmTing in this species are in respect to 

 dimensions and the size of the marginal plates. Some examples have a 

 more stelLxte for:«, while others verge onto an arcuate pentagonal shape. 

 Again, one extreme has fairly massive superomarginals, which are large up 

 to the tip of the ^-erj blimt ra}-, while in the other extreme, the upper 

 marginals are more umnerous and relatively smaller, and the ti^) of the ra}- 

 is not so blimt. The latter form is not so niunerous as the first, which is 

 t\-pical. The effect of this difference of shape on the relative dimensions is 

 sllo■s^•n in the following table : 



" Table showing effect of sha^x? on relative dimensions in Leptycliaster 

 onomalvs. 



"The numl)er of spinelets on the outer half of the adambulacral plates 

 (i.e., on actinal face) varies slightly, and they are a trifle longer and 

 slenderer in some individuals than in others. The tip of the ray in this 

 s])ecies is rounded and blimt, while it is sharp in jxicificm. This, with the 

 fewer and more transversely placed inferomarginal x^lates of anomcdus, ^^-ill 

 seiTB to distinguish the longer raj'ed individuals from pacificvs. Variation 

 A\ithin narrow limits occm-s in the size of the paxilLi?, in general the ex- 

 amples with more massive marginals having the larger paxilla3. The gi-eatest 

 difference in general facies is caused, however, by the variation in size of 

 marginals refeiTed te above. A slight difference in width causes a varying 

 amount of encroachment upon the abactinal paxillar area. In general the 



