68 DISCONTINUOUS VARIATION. 



suggests, slowly evolved from an originally almost 

 imperceptible degree of infertility, but that this takes 

 place only simultaneously with the evolution of mor- 

 phological character, in consequence of some form of 

 isolation. Thus, suppose a number of individuals 

 of a species become for a time separated from the 

 remainder of the species by a geographical barrier, 

 by migration, or some other cause of isolation, whereby 

 they are enabled to vary independently of the gen- 

 eral stock in response to changed conditions of life. 

 Then as they gradually become more and more diver- 

 gent from the parent stock in respect of morphological 

 characters, it is highly probable that they may concur- 

 rently perhaps from the direct action of the body 

 tissues on the reproductive system diverge also in re- 

 spect of physiological characters. Should any of them 

 now happen to meet and intercross with individuals of 

 the parent stock, or even if they should occupy the 

 same breeding area again, their newly acquired mor- 

 phological characters would no longer be in danger of 

 being swamped, for the simple reason that few or no 

 hybrid offspring would result from such crossing. In 

 the case of the higher animals, also, it is probable that 

 individuals of different varieties or sub-species, once 

 these are formed, instinctively tend to breed amongst 

 themselves, and hence the chance of production of hy- 

 brid offspring is still further diminished. Thus Dar- 

 win records * that in Paraguay it is believed " that the 

 native horses of the same colour and size prefer asso- 

 ciating with each other, and that the horses which have 

 been imported from Entre Rios and Banda Oriental 



* Ibid., vol. ii. p. 80. 



