RAYS AND BILATERAL SYMMETRY 265 



knowledge, it still seems worthwhile to push our inquiry a little 

 farther. 



Assuming the truth of our hypothesis that the two budding 

 zones of new rays correspond to the region of closure of the 

 hydrocoel ring, the question more specifically put is this : What 

 is the relation of A to this point of closure, and with reference to 

 both the original five rays and the budding zones? In no 

 echinoderm, so far as is known, does the closure of the hy- 

 drocoel ring take place radially. Indeed, in all cases the 

 primary tentacles are already clearly established before the 

 closure. But on our hypothesis it would seem to be necessary 

 to suppose, if we regard A as exactly homod3''namous with the 

 other five rays, that it arose at the exact point of closure ; i. e.^ 

 that closure does take place radially here. 



Now, of course, this may be precisely what happens. 

 There is no inherent impossibility of its being so. But refer- 

 ence of the question to what is known about the embryology of 

 other star-fishes suggests another explanation which has, at 

 least, enough basis of fact to warrant us in presenting it. 



On examining the descriptions and figures of the develop- 

 ment of star-fishes, particularly of Asterina gibbosa and As- 

 terias -pallida^ which have been so fully investigated recently 

 by Ludwig, Bury, MacBride, and Goto, one finds that arm A of 

 Pycnopodia occupies the exact position of the -preoral lobe, or 

 ' larval organ ' of these embryos. And when it is remembered 

 that this organ is retained for a considerable time after meta- 

 morphosis in some species, this coincidence becomes all the 

 more striking. Fig. 8, which we have copied from Ludwig, 

 '82, p. 173, represents diagrammatically the characteristics of 

 Asterina gibbosa shortly after metamorphosis ; and when it is 

 placed alongside our fig. 4, which, it will be remembered, is 

 an actual tracing of a young Pycnopodia with three pairs of 

 accessory arms, the correspondence of which we are speaking 

 is brought out strikingly enough. 



Now, of course, we would not be so rash as to imagine that 

 the arm A has arisen as an actual transformation of the preoral 

 lobe. All we wish to do at present, all we can do, indeed, is 

 to make clear the interesting fact of the correspondence in po- 



