418 CONCLUDING REMAKES. Chap. XXVIII. 



country, the two lots being supplied with different kinds of 

 food, would they after many generations differ ? AYhen we 

 reflect on the cases given in the twenty -third chapter, and 

 on such facts as the difference in former times between the 

 breeds of cattle, sheep, &c, in almost every district of Europe, 

 we are strongly inclined to admit that the two lots would be 

 differently modified through the influence of climate and food. 

 But the evidence on the definite action of changed conditions 

 is in most cases insufficient ; and, with respect to pigeons, I 

 have had the opportunity of examining a large collection of 

 domesticated kinds, sent to me by Sir W. Elliot from India, 

 and they varied in a remarkably similar manner with our 

 European birds. 



If two distinct breeds were mingled together in equal 

 numbers, there is reason to suspect that they would to a 

 certain extent prefer pairing with their own kind ; but they 

 would often intercross. From the greater vigour and fertility 

 of the crossed offspring, the whole body would by this means 

 become jnterblended sooner than would otherwise have oc- 

 curred. From certain breeds being prepotent over others, it 

 does not follow that the interblended progeny would be strictly 

 intermediate in character. I have, also, proved that the act 

 of crossing in itself gives a strong tendency to reversion, so 

 that the crossed offspring would tend to revert to the state 

 of the aboriginal rock-pigeon ; and in the course of time they 

 would probably be not much more heterogeneous in character 

 than in our first case, when birds of the same breed were 

 confined together. 



I have just said that the crossed offspring would gain in 

 vigour and fertility. From the facts given in the seventeenth 

 chapter there can be no doubt of this fact ; and there can be 

 little doubt, though the evidence on this head is not so easily 

 acquired, that long-continued close interbreeding leads to evil 

 results. With hermaphrodites of all kinds, if the sexual ele- 

 ments of the same individual habitually acted on each other, 

 the closest possible interbreeding would be perpetual. But 

 we should bear in mind that the structure of all herma- 

 phrodite animals, as far as I can learn, permits and fre- 

 quently necessitates a cross with a distinct individual. AVith 



