THE EVOLUTION OF ANIMAL FUNCTION 483 



the animal kingdom. And from this fact two consequences 

 would seem to follow : first, that there can be no sufficient 

 grounds whatever upon which any biologist of the past or 

 the present time could deny the legitimacy of the use of 

 functional characters as a test of animal relationship ; and, 

 second, that it is high time that the investigation of functional 

 capability should be begun in a conscious and systematic 

 manner. 



To this position we have been led from a study of the 

 historical connection between physiology and such attempts 

 as have been made to unravel the relationship of animals. 

 It is but natural, therefore, that the comparative study of 

 functional capability has been treated throughout as though 

 its one aim and justification must be sought in its contribu- 

 tion to animal genealogy. But this, I think, is far from Ithe 

 truth. Such a study, if ever it should advance to any sort 

 of coherence, must draw its chief interest from the direct 

 insight which it would afford into the actual course which 

 has been followed by the evolution of function. At present 

 a deep obscurity surrounds this problem. Physiologists accept 

 the fact that one cell spends its energy in contracting, another 

 in secreting useful ferments, a third in expelling waste products 

 from the organism. But there is hardly a suggestion as to 

 the past history of each type of cell. One thinks loosely of 

 the division of labour and of specialisation, and forgets the 

 ignorance which these words mask. And yet it is difficult 

 to escape the conviction - that here, certainly no less than 

 in the study of structure, lies the true history of animal 

 evolution. 



Broadly speaking, one may divide the study of functional 

 capability into two great parts — that which compares the cells 

 of different members of the animal kingdom, and that which 

 traces their gradual differentiation in the developing embryo. 

 These are the phylogeny and the ontogeny of function. And 

 it is hard to decide upon which part investigation may most 

 profitably begin. In a subsequent article I hope to discuss 

 this matter, and to attempt besides to indicate the general 

 lines upon which the study of functional capability may best 

 proceed. 



