490 W. WOODLAND. 



these conspicuous strands of protoplaem joining the extremi- 

 ties of the arms with the handle is due to the presence of 

 these arms and not vice versa, as I suggested when I first 

 observed the scleroplasm associated with the fully-formed 

 spicule (7). The strands are not muscular and exercise no 

 " tractive action" which can account for the recurved shape 

 of the armS; i.e. they do not exert an active pull producing 

 this effect, though it is possible that they have a slight 

 passive influence in this connection. 



The DEVELorMENT op the Spicule in Synapta digitata. 



As might be anticipated, the development of the plate-and- 

 anchor spicules of Synapta digitata proceeds on very 

 much the same lines as that of Synapta inh aureus, but 

 there is one difference which, though slight, is yet remarkable 

 on account of its striking incomprehensibleness. In the last 

 section I described the plate of the spicule as arising in the 

 form of a rod disposed transversely to the length of the 

 anchor-shaft (figs. 16, 18, e.g.). In S. digitata the plate 

 also arises as a rod, but curiously enough it is disposed 

 parallel with the anchor-shaft and not transversely to 

 it (figs. 17, 19). I have observed scores of young plates of 

 both species of Synapta, but I have never observed a single 

 exception to this rule, that the plate-rod of S. inhajrens is 

 disposed at right angles to, and the plate-rod of S. digitata 

 parallel with, the length of the shaft. 



Theoretical Considerations. 



The distribution and disposition of the plate- and-anchor 

 spicules in the body-wall of Synapta are subjects which first 

 demand our attention, llespecting the former there is little 

 to say, since the spicules are uniformly spread over the entire 

 area of the body-wall. However, it is worth remarking that 

 the production of these relatively few but huge spicules in 

 Synapta involves the concentration of numerous scleroblasts 



