i go THE VARIATION OF ANIMALS IN NATURE 



mated with like, which accelerates race-formation, while the 

 selection of parents on ' performance ' (i.e. by the quality of 

 their offspring) also increases the effectiveness of selection. 



We thus have two distinct and opposed views as to the 

 origin of domesticated races. According to the first they have 

 been produced mainly by the action of selection applied to a 

 plentiful stock of variations. According to the second they are 

 the result of appropriate crosses combined with pedigree breed- 

 ing and other devices. If the second view is correct, the success 

 of the breeder has been due to a procedure not fully repre- 

 sented in nature and the analogy between Artificial and Natural 

 Selection breaks down. If we disregard the question of muta- 

 tion-rate, as mutations are perhaps liable to turn up with equal 

 frequency in nature and under domestication, the issue can be 

 narrowed down to the question — is there as much opportunity 

 for crossing in nature as there is in the practice of stock-raising ? 

 If the numerous crosses made by man are the source of the 

 fresh steps in the development of domesticated breeds, and if 

 there is nothing comparable in nature, we think the analogy 

 must break down. The very great diversity of the means by 

 which isolation is established in nature between subspecies 

 and species inevitably suggests that the chances of factorial 

 recombination must be limited. It would seem a priori that 

 there could be no comparison between the amount of crossing 

 practised by man and that which occurs between natural 

 groups. Nevertheless some of the data in Chapter IV show 

 clearly that a large number of wild forms are highly polymor- 

 phic, and that the polymorphism is due to genetical causes. 

 We very frequently find subspecies and species that exhibit 

 various combinations of a common stock of characters, and 

 even among animals with a limited range, sedentary habits and 

 poor means of dispersal (such as land snails), there are numerous 

 instances of acute polymorphism. Nevertheless we do not 

 suggest that this polymorphism in any way approaches the 

 mixture of genotypes produced in domesticated forms. We 

 feel that some concrete measure of the difference is desirable 

 before this question is finally disposed of. However, the 

 critical point in this train of reasoning is that those who seek 

 to destroy the force of Darwin's analogy do not say that 

 selection is powerless. What they assert is that there is 

 more variation for it to work on among domesticated forms, 



