THE PRESIDENT'S ADDRESS. 69 



In Stelletta vestigium, on the other hand, the cladi are reduced 

 to the merest vestiges, and some, if not all of them, may com- 

 pletely disappear, while the shaft remains greatly elongated 

 and forms practically the entire spicule (figs, la Id). Possibly 

 the simple " oxeote " spicules of this and allied species (fig. 8) 

 have arisen in this manner. 



An altogether different line of evolution from the primitive 

 tetraxon archetype appears to have given rise to the typical 

 oxeote spicules (figs. 9, 10) of the monaxonellid division of the 

 Tetraxonida. Here two of the four rays of the primitive 

 tetraxon have probably entirely disappeared, while the remaining 

 two have become extended in a straight line with one another. 

 In the typical " stylote " (fig. 11) and " tylostylote " (fig. 12) 

 spicules probably only a single ray persists, so that the so-called 

 organic centre is situated at one end instead of in the middle. 

 In many species the oxea, styles or tylostyles become ornamented 

 with sharp spinose excrescences (fig. 13). 



In most of the cases which we have so far considered it is 

 easy to see that we are dealing with adaptive modifications. 

 The orthotriaene, dichotriaene, protriaene, anatriaene and 

 discotriaene are all obviously well suited for the fulfilment of 

 their specialised and differentiated functions, and the evolution 

 of these forms is more or less readily explicable in accordance 

 with the well-known principle of the natural selection of 

 favourable variations. The origin of the linear spicules of the 

 monaxonellid forms by complete suppression of two or three of 

 the rays of the primitive tetraxon is, perhaps, not so easy to 

 account for as is that of the triaene series from the same 

 starting-point. In both cases the determining factor was 

 probably, in the first instance, the development of a radially 

 arranged canal-system, requiring a corresponding radial arrange- 

 ment of the supporting skeleton, which could not be obtained 

 w T ith spicules of the primitive tetraxon form. That the evolu- 

 tion of the necessary linear spicules has taken place along 

 different paths in different cases is, however, nothing to be 

 surprised at ; it is merely one of those instances of convergence 

 which are quite as common amongst sponges as amongst other 

 groups of the animal kingdom. 



In the most primitive tetraxonid sponges, which represent 

 more or less closely the ancestral forms from which both 



