100 THE PRINCIPLES OF HEREDITY 



a more or less early stage. 1 But now arises the question, 

 What originally caused two branches of the same race to 

 separate and so cease to interbreed ? Probably the initial 

 cause was always geographical : one branch took to the 

 mountain, the other to the valley; one to the forest, the 

 other to the swamp ; one to the island, the other to the 

 continent ; one to one side of a mountain-chain, the other to 

 the opposite ; one to one end of a continent, the other to the 

 other end ; and so forth. Doubtless in many cases much 

 intercrossing occurred between the two diverging branches at 

 first ; but, as the divergence grew greater, the intercrossing 

 became less. Suppose for instance a species inhabited the 

 whole of a large tract of land which interposed no obstacle to 

 free intercourse except distance. Then though there were 

 free intercrossing between contiguous sections of the species, 

 there would be none between distant sections. Two or more 

 increasingly divergent types might arise, as we know they 

 have arisen between continental races of mankind in hun- 

 dreds of instances. In the course of geological time mutual 

 sterility would supervene, though intermediate types, did 

 they survive, might long furnish a chain of fertility uniting 

 the most separated groups. If subsequent migrations 

 mingled the two newly-established varieties, mutual dis- 

 inclination, combined with comparative or complete infertility, 

 would keep them from interbreeding. 2 Probably, therefore, 

 mere lapse of time is sufficient in all cases to cause mutual 

 infertility. Professor Romanes sought to attribute specific 

 differentiation in some cases to accidental infertility; he 

 supposed that certain individuals in a race might vary so as 

 to be fertile amongst themselves, but, from physiological 

 causes, infertile with the rest of the race. The sections, 

 separated by function but not by space, would then evolve on 

 diverging lines. 3 But "this seems putting the cart before 

 the horse ; we require to seek why infertility accompanies 

 differentiation of type, and we are told that infertility may be 

 the source of differentiation of type ; but why then should it 

 be correlated with advantageous rather than disadvantageous 

 differences ? Why are not the chance reproductive variations 

 distributed between the two types ? " 4 



1 Origin of Species, pp. 387-8. Sterility is sometimes caused other- 

 wise than by incompatibility in the germs and embryos. Thus a big 

 dog could not easily mate with a little dog. But these causes of sterility 

 do not concern us here. 



2 Animals and Plants, vol. ii., p. 80 et seq. 



3 Darwin and After Darwin, vol. iii. 



4 Pearson, The Grammar of Science, p. 417. 



