THE president's ADDRESS. 39 



well marked characters appearing again and again, in a sporadic 

 fashion, in widely separated groups, and it is impossible to 

 arrange our material in such a way that this shall not take 

 place. I do not, of course, refer to cases of convergence due to 

 adaptation to similar conditions of life, which are common 

 enough in sponges as in other animals. The characters in ques- 

 tion are again evidently non-adaptive. Take, for example, the 

 distribution of what are termed trichodragmata amongst the 

 Tetraxonida. The trichodragma is a bundle of short, hair-like 

 spicules, apparently all originating in one and the same mother- 

 cell. They occur scattered through the soft tissues of the sponge 

 without any relation to other structures, and it is impossible, at 

 any rate in the present state of our knowledge, to assign to them 

 any useful function. They sometimes occur in a whole group of 

 evidently closely related species and have in several cases been 

 used as the chief distinguishing character of genera. In the 

 genus Cinachyra, on the other hand, they occur in only a single 

 one of the twenty species recognised. They occur, again, in 

 such widely separated families as Stellettidae, Tetillidae and 

 Desmacidonidae, and, wherever they are met with, they exhibit 

 an extraordinary uniformity of structure. A precisely similar 

 phenomenon is observable in the case of the small monaxon 

 spicules known as microxea amongst the Calcarea. 



Now we know that mutations are discontinuous in their mode 

 of origin and that the same mutation may occur again and 

 again. Discontinuity in the present distribution of trichodrag- 

 mata and microxea almost certainly implies discontinuity in 

 origin. Probably these spicules have arisen suddenly, and on 

 many occasions, as the result of some unknown change in the 

 constitution of the germ-plasm. It is, no doubt, more a 

 chemico-physical than a biological phenomenon, and that, I 

 believe, is true of all mutations. 



Let us now consider briefly some examples of what are prob- 

 ably to be interpreted as mutations in the case of megascleres. 

 It is well known that derivatives of the primitive tetract not 

 infrequently exhibit meristic variations as regards the number 

 of their rays. It is very doubtful whether all of these can be 

 regarded as mutations, for in many cases we find great varia- 



