Characters as Adaptive and Specific. 195 



ary refuge in the argument from ignorance is most 

 effectually closed. For — 



"When the pasture is tolerably long, these cattle feed as well 

 as common cattle with their tongue and palate ; but during the 

 great droughts, when so many animals perish on the Pampas, 

 the niata breed lies under a great disadvantage, and would, 

 if not attended to, become extinct ; for the common cattle, like 

 horses, are able to keep alive by browsing with their lips on the 

 twigs of trees and on reeds ; this the niatas cannot so well do, 

 as their lips do not join, and hence they are found to perish 

 before the common cattle. This strikes me as a good illus- 

 tration of how little we are able to judge from the ordinary 

 habits of an animal, on what circumstances, occurring only at 

 long intervals of time, its rarity or extinction may depend. 

 It shows us, also, how natural selection would have determined 

 the rejection of the niata modification, had it arisen in a state 

 of nature '." 



Hence, it is plainly impossible to attribute this 

 modification to natural selection, either as acting 

 directly on the modified parts themselves, or indi- 

 rectly through correlation of growth. And as the 

 modification is of specific magnitude on the one 

 hand, while it presents all " the most essential fea- 

 tures of specific characters " on the other, I do not 

 see any means whereby Mr. Wallace can meet it 

 on his a priori principles. It would be useless to 

 answer that these characters, although conforming to 

 all his tests of specific characters, differ in respect 

 of being deleterious, and would therefore lead to ex- 

 termination were the animals in a wholly wild state ; 

 because, considered as an argument, this would involve 

 the assumption that, apart from natural selection, 

 only deleterious characters can arise under nature 



' Darwin, Variation, Sec. vol. i. p. 94. 

 O 2 



