227 



enemies, the weasels, greyhounds, &c., would have improved 

 likewise, so that perhaps the hare would not be really better 

 off; but at any rate the direction of the change would be from 

 a war of pigmies to a war of Titans. Opinions may differ as 

 to the evidence of this gradual perfectibility of all things, but 

 it is beside the question to argue this point, as the origin of 

 species requires not the gradual improvement of animals retain- 

 ing the same habits and structure, but such modification of 

 those habits and structure as will actually lead to the appearance 

 of new organs. We freely admit, that if an accumulation of 

 slight improvements be possible, natural selection might im- 

 prove hares as hares, and weasels as weasels, that is to say, it 

 might produce animals having every useful faculty and every 

 useful organ of their ancestors developed to a higher degree ; 

 more than this, it may obliterate some once useful organs when 

 circumstances have so changed that they are no longer useful, for 

 since that organ will weigh for nothing in the struggle of life, 

 the average animal must be calculated as though it did not exist. 

 We will even go further : if, owing to a change of circum- 

 stances, some organ becomes pre-eminently useful, natural selec- 

 tion will undoubtedly produce a gradual improvement in that 

 organ, precisely as man's selection can improve a special organ. 

 In all cases the animals above the average live longer, those 

 below the average die sooner, but in estimating the chance of 

 life of a particular animal, one special organ may count much 

 higher or lower according to circumstances, and will accordingly 

 be improved or degraded. Thus, it must apparently be conceded 

 that natural selection is a true cause or agency whereby in some 

 cases variations of special organs may be perpetuated and accu- 

 mulated, but the importance of this admission is much limited 

 by a consideration of the cases to which it applies : first of all 

 we have required that it should apply to variations which must 

 occur in every individual, so that enormous numbers of indivi- 

 duals will exist, all having a little improvement in the same 

 direction ; as, for instance, each generation of hares will include 

 an enormous number which have longer legs than the average 

 of their parents, although there may be an equally enormous 

 number who have shorter legs ; secondly, we require that the 

 .variation shall occur in an organ already useful owing to the 



02 



