312 HABIT AND INTELLIGENCE, [chap. 



if one dog in every hundred excels in one of those qualities, 

 and the same proportion in the other, to a sufficient degree 

 to attract the attention of the selector ; the chance of any 

 dog excelling in both will be only the second power of one 

 in a hundred, or one in ten thousand, which chance is so 

 small that it may be left out of account, 

 and ill It is obvious that selection by natural agency may act 



wild races. . j_n ^, i . , , 



m exactly the same way as selection by human agency. 

 WUd beasts of prey that excel either in fleetness or in scent 

 will have the best chance of surviving and leaving off- 

 spring ; and these favourable variations will be added Tip 

 and accumidated through an indefinite number of genera- 

 tions. As I have just shown, it is not at all likely that two 

 favourable variations will occur in the same individual ; 

 and an unusual degree of fleetness in one of a "race that 

 hunts by scent, or an unusual power of scent in one of a 

 swiftly running race that hunts by sight, would probably 

 be comparatively useless, and consequently would not tend 

 to the preservation of that individual. What will be most 

 valuable, and consequently most conducive to the preser- 

 vation of the individual, is probably a slightly increased 

 degree of some favourable peculiarity that the race already 

 has in a tolerably high degree ; especially as, by the laws 

 of self-adaptation and hereditary habit, the race will have 

 begun to be adapted in its whole structure to the mode of 

 life to which its favourable peculiarity is suited. Thus we 

 may expect sagacity and wariness to become characteristic 

 of those animals which hunt by scent, especially if they 

 are gregarious. Besides, a race produced by the selection 

 of individuals for any one character will necessarily have 

 that character, on the whole, more variable than the rest of 

 the organism ; and, as we have seen in speaking of the 

 laws of variation, a character, or an organ, that begins to 

 vary is apt to continue variable ; so that favourable varia- 

 tions will be more likely to occur in the character that 

 constitutes the principal differentia of a breed, than in any 

 other of its characters. It is, as already stated, an observ^ed 

 fact that this is so. For these reasons, different breeds will 

 continue distinct and divergent. 



