26o RITTER AND CROCKER 



pect that evidences of growth in the mesench3^matous tissue 

 surrounding these parts would appear at this time ; but such is 

 not the case. The external perihasmal canal, a.f.c.^ appears 

 more extensive here than it really is, owing to a break in its 

 outer wall ; but no evidence of special growth in connection 

 with it is recognizable. 



QUESTIONS RAISED BY THE FACTS PRESENTED, AND AN 

 EFFORT TO INTERPRET THEM. 



The final interpretation of ray multiplication in Pycnopodia 

 in relation to the structure of the larva is obviously impossible in 

 the absence of knowledge of the embryology of the species. 



Nevertheless the facts at hand raise several questions, answers 

 to which are at least suggested when they are studied in the light 

 of what is known about the development of other echinoderms, 

 particularly species closely related to Pycnofodia. The most 

 interesting of these questions are the following : (i) Have not 

 the two definite positions at which rays are produced relation 

 to the place of closure of the hydroccel ring of the larva ; or, 

 perhaps more precisely, to the two ends of the hydroccel before 

 it becomes a closed ring? (2) Is the definitel}'' expressed axis, 

 A IV, of bilateral symmetry the same as the axis of sym- 

 metry of the larva? (3) What is the relation of ray A to the 

 five rays of the asterid ground plan? 



Bury, '89, summarized, for the entire class of Echinoderms, 

 all that was then known about the place of closure of the h}'- 

 drocoel ring. According to this summary, the closure always 

 occurs interradially, but the particular interradius differs for 

 different groups, four of the five original interradii being the 

 place for some one or more groups. For all the astertds, so far 

 as known, t/ic position is the interraditis 5-/ relative to the 

 hydrocosl itself. But it is well known that in most star-fishes 

 (A. Agassiz, '77 ; Ludwig, '80 ; Bury, '95 ; MacBride, '96, and 

 Goto, '97) a rotation of the antiambulacral disc takes place dur- 

 ing metamorphosis in such a way that, following Ludwig, who 

 designates the primary tentacles of the hydroccel with Arabic 

 numerals, and tlie incipient antiambulacral raj^s with Roman 

 numerals, hydroccel tentacle 5 comes into relation with anti- 



