Classification of Animals. 19 



species as well as higher groups being marked out, not 

 by any distributive law, but by the purely non-significant 

 operation of human experience, which cannot transcend 

 the results of death and decay." 



To the very last Louis Agassiz maintained his convic- 

 tion that "there is no evidence of a direct descent of 

 later from earlier species in the geological succession of 

 animals"; and the famous Essay on Classification appears 

 throughout to involve a misunderstanding of what clas- 

 sification really is. At the same time, it must be re- 

 membered that this great naturalist saw clearly that the 

 various forms of life are not chaotic, that they can be 

 put in order, that there is a Systema Naturce, and a 

 progressive development which he chose to express only 

 in transcendental terms. 



The modern conception of species may be expressed 

 as follows: When we see individual organisms very 

 like one another, and so well marked off from their 

 nearest neighbours that it is possible to distinguish 

 them, we find it convenient to give them a specific name. 

 Before doing so, if there is opportunity, we take certain 

 common-sense precautions. We inquire whether the 

 distinguishing marks which have arrested our attention 

 have any real constancy, whether they persist through 

 successive generations. What is more difficult is, to 

 distinguish acquired characters or modifications, which 

 are assumed by each individual in its lifetime as the 

 result of external conditions, from inborn characters 

 which form the real basis of the specific inheritance. 

 We also inquire whether the distinctive characteristics 

 in question are greater than those variations which are 

 so often exhibited among the progeny of a single pair. 

 Thus, no one would propose to divide men into species 

 according to the colour of their hair or eyes, since that 

 would land one in the absurdity of placing two brothers 

 in different species. We also find out whether the 

 members of the proposed species are fertile inter se, and 

 tend to be sterile when crossed with the members of a 

 related species. 



To sum up, a species is a relative conception, con- 

 venient when we wish to include under one title all the 



