CORRELATION 177 



might have allowed us to infer that the history of species in 

 time would be exceedingly complex. Groups of living animals 

 are broken up into hierarchies of divergence isolated from one 

 another to a varying extent. We can, therefore, if we wish, 

 separate any two groups by a single differentiating character, 

 but only at the expense of ignoring all the other features in 

 which they may happen to agree or differ. The palaconto- 

 logical evidence that single characters evolve more or less inde- 

 pendently of one another is only a corollary of their failure 

 as group-indicators in living forms. From this point of view 

 evolution is a relatively simple process with two main aspects — 

 (1) the origin, in a relatively small number of individuals, of 

 new characters, some of which spread throughout large popu- 

 lations, and (2) the ' trying-out ' of such material in all sorts 

 of combinations. 



It does not require a highly developed critical faculty to 

 see that this is a very simplified and abstract account of the 

 living organism. The picture of species as being built up like 

 houses from bricks is very hard to reconcile with any theory of 

 development. The phenomenon of regulation in the individual 

 is so like that of correlation in the species, that it is difficult to 

 believe that the modern genetic concept of species as mosaics 

 of gene interaction illuminates more than one aspect of our 

 problem. If the regulatory activity of organisms can deter- 

 mine the development of a single blastomere into a whole 

 rather than a fractional organism, it would be strange if the 

 relations between specific characters were not also in some way 

 controlled. 



We may consider first how far it is possible to sum up an 

 organism as a mosaic of unit characters. Such a question 

 might be asked not only with regard to the relatively crude, 

 unanalysed specific characters, but also with regard to the 

 supposedly more fundamental genes of the geneticist. In 

 discussing specific characters as they appear in taxonomy we 

 have not indicated how far the taxonomic definitions are in- 

 complete. Actually everyone knows that specificity is not 

 something superficial and external, like the last coat of paint 

 on a new car, but something which permeates the organism 

 through and through. It may show itself in any part of the 

 organism, whether structural, physiological or psychical. It 

 is seen perhaps most characteristically in the apparently 



N 



