SPICULES OF SYNAPTA AND AUIUCULARIA LARVA. 495 



anchors developed iu couuection with each other ? But to 

 assume that this question is capable of being answered implies 

 that we know the answers to several questions which natur- 

 ally precede it. Grant that the huge anchor-shaft owes its 

 size, shape, and disposition to the factors already mentioned, 

 grant also that the relatively small rod of the incipient plate 

 is small because of the small syncytium concerned, yet we are 

 quite ignorant either as to why the six to ten nuclei should 

 migrate in a bunch from the mass of the syncytium on to the 

 internal side of the shaft, or as to why this cluster of nuclei 

 when pi'oduced should give rise to a separate plate-spicule. 

 Why should not the internal cluster of nuclei merely deposit 

 an outgrowth of the shaft on its internal aspect in the same 

 manner as the gastral actiuoblast of the calcareous sponge 

 deposits the gastral ray on the triradiate basis ? Personally, 

 I am as yet quite unable to suggest solutions to these problems. 



Before concluding I may mention that occasionally no plate 

 is developed in connection with the anchor, the anchor remain- 

 ing solitary (fig. 24). For some reason or other, the migration 

 of nuclei on to the internal surface of the shaft has (presum- 

 ably) not taken place. Whether plate structures are ever 

 developed in Synapta, apart from the anchor, I cannot say 

 for certain. I have observed several fully-formed solitary 

 plates, but, although no signs of disturbance were visible, 

 yet I suspect that the anchors had, in every case, been de- 

 tached. It is, indeed, hard to suppose that the plates would 

 a.ssume the normal shape (which is modified iu connection 

 with the anchor^) when developed by themselves. I have 

 also observed two or three instances of what appear to be 

 young plates developing on their own account, but, again, I 

 cannot be quite certain that they had not become detached 

 from anchor counterparts. Anchors can develop without 

 plates (I think the instances are too numerous to suppose 

 that the plates had merely become detached), but it is 

 very doubtful if normal plates can develop apart from 

 anchors. 



' Tills is conspicuously shown in the abnormalities figured by Herouard (2). 



