418 OUTLINES OF EVOLUTIONARY BIOLOGY 



If isolation is necessary for the establishment of new species 

 by divergent evolution, why, it may be asked, do we find closely 

 related species, which we must suppose to be descended from 

 common ancestors at no very distant period, actually inhabiting 

 the same areas at the present time ? The answer is that there 

 are other means by which groups of organisms can be prevented 

 from interbreeding besides geographical isolation. 



It used to be supposed that one of the best tests of the specific 

 distinctness of any two forms was their incapacity for breeding 

 together and producing fertile offspring. 1 The mule, it is true, 

 is the offspring of parents belonging to two distinct species, the 

 ass and the horse, but the mule is almost always sterile, and most 

 well characterized species 2 are incapable of breeding together at 

 all. This fact has been regarded by many people as a most serious 

 difficulty in the way of comparing the origin of species by natural 

 selection with the results produced amongst domesticated plants 

 and animals by artificial selection, for the products of artificial 

 selection, however much they may differ from one another, if 

 they have been derived from the same parent species will 

 remain capable of breeding together with perfect fertility. 



The solution of this difficulty is to be found in the theory of 

 " Physiological Selection," which we owe to Mr. Gulick, Dr. 

 Romanes and others. These writers point out that amongst 

 the endless variations to which plants and animals are subject 

 will be variations in the reproductive system, by which certain 

 individuals will be rendered infertile when crossed with others 

 of the same species, while remaining fertile with individuals 

 which have varied in the same manner as themselves. In this way 

 a species may be as effectively divided into two sections as by any 

 geographical barrier, and under these circumstances divergent 

 evolution may be expected to take place. According to this view 

 the mutual sterility which, to a greater or less extent, undoubtedly 

 does characterize distinct species in a state of nature, is the 

 cause and not the result of their distinctness, and cannot be 

 regarded as a reason for supposing that there is any essential 

 difference between the processes of artificial and natural 

 selection. In artificial selection it is merely another kind of 



1 Although Lamarck pointed out a century ago that this is really no criterion of 

 specific distinction. The idea that it is so is clearly expressed by Buffon (" Histoire 

 Naturelle," Tom. VI, 1756, p. Ifi). 



2 At any rate amongst the higher animals. 



