Evolution of Animal Life. 153 



The analogy of artificial breeding is deceptive in this : 

 that we select plants and animals for their desired peculiar- 

 ities, and prevent cross-breeding. We do not select out of 

 the aggregate of forms we desire to perpetuate those which, 

 besides having that peculiarity, are fertile with one another, 

 but sterile towards the rest. Hence our varieties are sub- 

 ject always to cross-breeding and reversion. In other 

 words, we do not get specific sterility, because we do not 

 breed for it. But Nature starts with that, and performs by 

 her selection the close-breeding which we secure by artificial 

 devices. 



Finally, this theory, which makes relative sterility with 

 special interfertility one of the protective modifications 

 upon which natural selection proceeds, is after all only a re- 

 statement of the Darwinian formula itself. For the surviv- 

 ors in each generation, retaining in most effective degree 

 the advantageous peculiarities which distinguish them, are 

 most likely to be the offspring of the protected parents on 

 both sides. Cross-breeding will be punished by reversion 

 and loss of advantageous peculiarities. 



This may be expressed in our fanciful symbolism by sub- 

 stituting for Darwinian equations 4 and 5, the following : 



( M ) IL 



(4) - — [ 1= 



w I K > K 2 



(5) \J J ^] n AB=A\^-\=R 

 I K 2 > I R > 



That is to say, natural selection acts twice on each genera- 

 tion, selecting from the fittest to survive (M ) the fittest to 

 breed; and this process, repeated through numerous (w) 

 generations produces physiologically permanent species, as 

 -artificial selection ($) would do, if it were directed towards 

 intersterility (B) as one of its objects, and continued 

 through a sufficient period (A). 



What Mr. Darwin apparently overlooked was the proba- 

 ble decrease in numbers of the pure-blooded variety, ac- 

 companied by a complete isolation from related forms, un- 

 til the new species takes its start, perhaps from a single 

 pair, which, in its swift multiplication thereafter, sweeps 

 away all the feeble varieties of the old stock which may 

 have accompanied its history. 



