/ 



THE EVOLUTION OF ANIMAL FUNCTION 477 



corresponding elementary parts of the early embryo. 1 So the 

 claim of structural characters to form the sole criterion of 

 homology, and hence of classification, appeared to be established. 



For the rest the story is a simple one. When the idea 

 of evolution was formally enunciated by Darwin, and so became 

 a working hypothesis, the old methods of classification were 

 accepted without inquiry as the methods by which the course 

 of evolution must be traced. The relationships of animals 

 were determined solely on the basis of structural homology. 

 No more striking evidence could be found for the completeness 

 with which questions of function had been excluded from all 

 inquiry into the relationship of animals. 



So it has come about, mainly through the work of Geoffroy 

 St. Hilaire, that the study of animal function is almost wholly 

 -separated from the study of evolution. The separation is 

 greatly to be deplored. For in the first place it is not im- 

 probable that the relationships of animals might have been 

 better determined if morphology had co-operated with physio- 

 logy in the work. And beside this, a point which has been 

 of far greater importance to biology, the study of the evolution 

 of function, might have developed, given only the requisite 

 stimulus which the tracing of genealogy would have supplied, 

 into a body of knowledge of far greater moment for the under- 

 standing of evolution than any purely structural study can ever 

 be. If physiologists had felt that the comparative study of 

 function could form a science really essential to the under- 

 standing of evolution, it is hard to believe that they would 

 not have hastened to remove the reproach so commonly made 

 against them, that their animal kingdom comprises only the 

 frog, the rabbit, the cat and the dog. It is easy to urge 

 by way of excuse and explanation that a study of the 

 functions of invertebrate animals demands a technique which 

 is difficult and little understood. But this is merely to state 

 the same fact from the opposite side. The technique of 

 experiment on invertebrate animals is not in itself essentially 

 difficult ; it has merely failed to develop because the doctrine 



1 The whole question of the validity of "homologies" determined on the 

 ground of embryonic development has been raised in recent years by such 

 workers as Wilson, Braem, and Gaskell. The question is one of particular 

 interest because it touches the fundamental distinction of structural and 

 functional characters. But space forbids any discussion in this place. 



\ 



