262 RITTER AND CROCKER 



formulated above. The difficulty is that ray II does not appear 

 simultaneously with I, but at an earlier time. It should be 

 noted, however, in this connection, that all the information we 

 have is to the effect that there is but very slight difference in 

 the time of origin of the five original tentacles ; and it is quite 

 probable, therefore, that more attention to this point will pro- 

 duce surer evidence of their successional appearance from be- 

 hind forward. 



From Ludwig, '82, we get some very significant indirect evi- 

 dence on this point. Although he gives no information about 

 the time of origin of the primary tentacles with reference to one 

 another, he does point out the order in which the first pairs of 

 ambulacral feet appear on the different tentacles. Thus he 

 shows (p. 149), that they appear on 3 = IV, first, then follow 

 2 and 4 = III and V, and conclude with i and 5 = II and I ; 

 and 3 he says has two pairs before i and 5 have undergone 

 any changes from their primitive tentacular state. In other 

 words, in Astcj'ina^ after the very inception of the primary ten- 

 tacles, the order of which we do not know, the development of 

 the -posterior rays of the radial water system proceeds consider- 

 ably in advance of those of the anterior ones, the order being: 

 IV in the lead; III and V next in order and together; and 

 lastly II and I, also together. How easy and natural the de- 

 duction, then, that for Pycjiopodia it would be rx and Ix next ; 

 rx^ and Ix^ next, and so on ! 



We, of course, know too much now about the time-order of 

 origin of embryonic structures to permit us to place any reliance 

 on its phylogenetic significance when considered independently 

 of the functional value of the parts concerned. But in a case 

 like the present, where the organs are so entirely alike, both 

 structually and functionally, as are the rays of a star-fish, we 

 may safely suppose that any constant difference in the time of 

 their origin and rate of development is of phyletic value. 



The conception of ray multiplication in Pycnopodia thus ar- 

 rived at is not in accord with that held by some students of the 

 echinoderms. Thus, Cuenot, '91, has stated in a comprehen- 

 sive and at the same time concise way, what might be called 

 the "decentralization theor}-," under which he would consider 



