268 Darwin, and after Darwin. 



amount of attachment is already so great that teeth 

 will break before they can be drawn by anything 

 short of a dentist's forceps. Therefore I conclude 

 that this peculiarity in the dentition of the genus 

 must have arisen in its parent species by way 

 of what Darwin calls a " fluctuating variation," with- 

 out utilitarian significance. And I adduce it in 

 the present connexion because the peculiarity is one 

 which is equally unamenable to a utilitarian ex- 

 planation, whether it happens to occur as a generic 

 or a specific character. , 



Numberless similar cases might be quoted ; but 

 probably enough has now been said to prove the 

 inconsistency of the distinction which our opponents 

 draw between specific and all higher characters 

 in respect of utility. In point of fact, a very 

 little thought is enough to show that no such 

 distinction admits of being drawn ; and, therefore, 

 that any one who maintains the doctrine of utility 

 as universal in the case of specific characters, must 

 in consistency hold to the same doctrine in the case 

 of generic and all higher characters. And the fact 

 that our opponents are unable to do this becomes 

 a virtual confession on their part of the futility of 

 the generalization which they have propounded ^. 



' It may be observed that this distinction was not propounded by 

 Mr. Wallace — nor, so far as I am aware, by anybody else — until he 

 joined issue with me on the subject of specific characters. Whether he 

 has always held this important distinction between specific and generic 

 characters, I know not ; but, as originally enunciated, his doctrine of 

 utility as universal was subject to no such limitation : it was stated 

 unconditionally, as applying to all taxonomic divisions indifferently. 

 The words have already been quoted on page i8o; and, if the reader 

 will turn to them, he may further observe that, prior to our discussion, 

 Mr. Wallace made no allowance for the principle of correlation, which, 



