80 EVOLUTION OF LIVING OEGANISMS. 



thought that a variation which appeared in one or even 

 in several individuals of a species would have little chance 

 of establishing itself, since it might be reduced and 

 finally swamped by constant interbreeding with the more 

 numerous individuals of the species not possessing it. 

 Such a swamping by intercrossing, however, does not 

 occur. We have already seen (p. 52) that factors of in- 

 heritance are as a rule transmitted complete ; even when 

 the appearance of the character to which it gives rise is 

 subject to various inhibiting influences, it is liable at any 

 time to reappear in full force, as is seen in reversion. 

 Characters due to factors of inheritance will persist un- 

 less eliminated by variation and selection, as can be 

 shown by experiments and by mathematical reasoning. 

 The relative scarcity of the mutation at the start does 

 not prevent that a number of individuals interbreeding 

 at random, some with and others without a certain 

 factor, will give rise to a population of impure hetero- 

 zygotes and pure homozygotes in which the proportion 

 of the three classes will be in equilibrium so soon as the 

 square of the number of heterozygotes equals the number 

 of pure "dominants" multiplied by the number of pure 

 " recessives." If this proportion is not already present at 

 the beginning it will soon become established, and will 

 continue, provided there is no selection to disturb the 

 equilibrium. In fact a species of interbreeding indi- 

 viduals of unequal hereditary constitution soon reaches 

 a state of stability. 



CHAPTER VI 



ISOLATION AND SEXUAL SELECTION 



FREE intercrossing can only mix hereditary strains, hence 

 the necessity for isolation if divergence along various 



