STUDIES IN ANIMAL LIFE. 97 



species. If these varieties inhabited different coun- 

 tries if the pug were peculiar to Australia and the 

 mastiff to Spain there is not a naturalist but would 

 class them as of different species. The same re-, 

 mark applies to pigeons and ducks, oxen and sheep. 

 The reason of this uncertainty is that the thing 

 species does not exist : the term expresses an ab- 

 straction, like virtue, or whiteness ; not a definite 

 concrete reality, which can be separated from other 

 things, and always be found the same. Nature pro- 

 duces individuals ; these individuals resemble each 

 other in varying degrees ; according to their resem- 

 blances we group them together as classes, orders, 

 genera, and species; but these terms only express 

 the relations of resemblance, they do not indicate the 

 existence of such things as classes, orders, genera, or 

 species.* There is a reality indicated by each term 

 that is to say, a real relation ; but there is no ob- 

 jective existence of which we could say, This is 

 variable, This is immutable. Precisely as there is 

 a real relation indicated by the term goodness, but 

 there is no goodness apart from the virtuous actions 

 and feelings which we group together under this 

 term. It is true that metaphysicians in past ages 

 angrily debated respecting the immutability of vir- 

 tue, and had no more suspicion of their absurdity 

 than moderns have who debate respecting the fixity 



* CUVIER says, in so many words, that classes, orders, and gen- 

 era are abstractions, et rien de pareil n'existe dans la nature ; but 

 species is not an abstraction ! See Lettres a Pfaff, p. 1 79. 



E 



