316 THE VARIATION OF ANIMALS IN NATURE 



there is little evidence at present that such specialisation is 

 initiated through gene-mutations susceptible to selection. 



Finally, we have considered the special difficulty of the 

 evolution of complex organs and of co-adaptations, of which 

 the interrelations of the male and female genitalia are one 

 example. The argument employed by Fisher and Haldane 

 to show that Natural Selection might account for the evolution 

 of such structures, depends on the assumption that very minute 

 changes in a complex situation will, as likely as not, lead to an 

 improvement. As we have previously stated (p. 224), we are 

 very doubtful whether the enhanced survival value conferred 

 by such minimal variants would give a sufficiently steady 

 selection-rate to ensure the establishment of the variant. We 

 prefer, rather, to regard such complex structures as a special 

 case of the elaborate internal organisation characteristic of all 

 living organisms. 



In short, we do not believe that Natural Selection can be 

 disregarded as a possible factor in evolution. Nevertheless, 

 there is so little positive evidence in its favour, so much that 

 appears to tell against it, and so much that is as yet inconclusive, 

 that we have no right to assign to it the main causative role in 

 evolution. 



