THE president's ADDRESS. 31 



If you ask wliat justification we have for arranging the series 

 as we do and for assuming that they represent actual stages 

 in evolution, I can only reply that we have exactly the same 

 justification as in the case of any other group of organisms. 

 Where we find a graduated series leading from simple to com- 

 plex it is difficult to account for it on any other hypothesis, especi- 

 ally when, as is actually the case in sponges, the various stages 

 exhibited by one system of organs are correlated with correspond- 

 ing stages exhibited by other systems. Thus in the tetraxonid 

 sponges the evolution of spicular form is accompanied by evolu- 

 tion in the arrangement of the skeleton as a whole and also in the 

 arrangement of the canal-system, and again in the Euceratosa 

 the correlation between the degree of complexity attained by 

 the skeleton and that attained by the canal-system is very striking. 

 Moreover we can, in some cases, appeal to the facts of ontogeny 

 or individual development. One of the most highly modified 

 forms of spicule found in tetraxonid sponges is the discotriaene 

 of the genus Discodermia. This spicule resembles nothing so 

 much as a carpet nail, the disc being formed by three of the rays 

 of the primitive tetract form expanded and fused together, 

 as is shown conclusively by the fact that the young spicule 

 passes through a stage in its development in which all three 

 rays are separate and it is, in short, an ordinary tetract. Plenty 

 of awialogous cases could be mentioned, and many more will 

 doubtless be brought to light as our knowledge of the subject 

 increases. 



Palaeontological evidence, unfortunately, teaches us but little 

 with regard to the evolution of sponges, for most of the fossils 

 known are highly specialised forms which owe their preservation 

 to the peculiar manner in which the spicules have become 

 united together in a coherent framework. For some unknown 

 reason, however, it appears that most of the evolutionary stages 

 through which recent sponges have passed are still represented 

 by existing forms, and that is why the group is so exceptionally 

 suitable for the study of evolutionary processes. 



When we come to consider the factors involved in the pro- 

 duction of these evolutionary series we have to note in the first 

 place that we can distinguish between two types of modifications — 



