76 W. WOODLAND. 



development of the triradiate even before the three minute 

 needles secreted by the sextet have united to form an aggre- 

 gate or compound spicule (Study I^ fig. 32), and therefore 

 long before any stresses in the sponge-wall could possibly 

 have produced any such modification. Therefore, if this 

 larger ray of the young triradiate is the large "posterior" 

 ray of the adult spicule, it seems to me that the inheritance 

 of this particular form of spicule is proved. But the fact is 

 that this larger ray of the young triradiate does not, by any 

 means, always become the '• posterior" ray of the adult 

 spicule, though it often does (probably for the reasons 

 assigned in Study I, pp. 263 — 265). Observations of the 

 Sycon oscular rim show that the young triradiates are dis- 

 posed very irregularly, and only exceptionally does the large 

 ray point exactly towards the base of the sponge. As I have 

 stated in Study I " the large ray is often found pointing as 

 much as 80° from the downward vertical line," and more 

 recent observations prove that the large ray may not 

 uncommonly lie in exactly the opposite direction to that in 

 which the majority of the basal rays of the adult spicules lie; 

 in short, the large ray of the young triradiates may point in 

 any direction, whereas that of the adult triradiates generally 

 points basall3^ These facts seem to me to show that the 

 large ray of the young Sycon triradiate is not the inherited 

 large " posterior " ray of the adult spicule, since, were it so, 

 the cause which is capable of thus reproducing in the young 

 spicule a structural characteristic only directly producible 

 in the adult spicule should be equally capable of determining 

 the appi'opriate disposition of that structural characteristic. 



As an alternative to the inheritance hypothesis I can only 

 suggest that the above feature of the Sycon triradiate is of 

 the same order of form-modifications as the clubbed extremi- 

 ties of the triradiates of Clathrina clathrus, the gastral 

 ray spikelets of C. cerebrum, etc. — modifications produced 

 ontogenetically without reference to the economy of the 

 organism, and possibly mechanical or crystallomorphic in 

 nature. Assuming this to be the explanation of the large 



