g6 Animal Life and Intelligence. 



of taste among the hens of each particular species, since 

 there is a uniformity of coloration among the cock-birds. 

 It may be said that in all their mental endowments there 

 is greater uniformity among animals than among men ; and 

 it is true that individuation has not been carried so far in 

 them as in human-folk. Still, careful observers of animals 

 see in them many signs of individual character ; and this 

 uniformity in the standard of taste in each species of 

 birds seems to many naturalists a real difficulty in the 

 way of the acceptance of sexual selection. We shall, 

 however, return to this point. For the present it is 

 clear that selection chooses out advantageous variations, 

 that the advantage is determined by the taste of the 

 selector, and that uniform selection implies uniformity of 

 taste. 



Turning to elimination, it is clear that it begins by 

 weeding out, first the more disadvantageous, then the less 

 disadvantageous variations. It leaves both the advan- 

 tageous and the neutral in possession of the field. I 

 imagine that many, perhaps most, of the variations 

 tabulated by Mr. Wallace and other observers belong to 

 the neutral category. Their fluctuating character seems 

 to indicate that this is so. In any case, they are 

 variations which have so far escaped elimination. And 

 I think they are of great and insufficiently recognized 

 importance. They permit, through interbreeding, of end- 

 less experiments in the combination of variations, some of 

 which cannot fail to give favourable results. 



It is just possible that it may be asked — If in natural 

 elimination there is nothing inOre than the weeding out 

 of the unfit and the suppression of disadvantageous varia- 

 tions, where is the possibility of advance ? The standard 

 may thus be maintained, but where is the possibility of 

 progress ? Such an objection would, however, imply forget- 

 f ulness of the fact that all the favourable variations remain 

 to leaven the residual lump. Given a mean, with plus and 

 minus variations : if in any generations the minus varia- 

 tions are got rid of, the mixture of the mean with the plus 



