CLASSIFICATION 225 



According to this view a newly arisen species, at its first origin, 

 is already sharply distinguished from the parent species and the 

 extinction of intermediate forms is not required to separate the two. 



The elementary species, however, though apparently constant 

 in their characters, are often so similar to one another, and 

 they are moreover so numerous, that for practical purposes of 

 classification it would be undesirable to regard them as distinct 

 species in the ordinary sense of the word. It is far more con- 

 venient to maintain the older view that they are merely varieties 

 or subspecies, and their occurrence in no way invalidates our 

 conception of the slow, tree-like evolution of the organic world. 

 If de Vries be right as regards the importance he attributes to 

 them as the sole means by which species have originated, which 

 is extremely doubtful, it merely means that evolution has taken 

 place in a rather more jerky fashion than we previously supposed. 

 It will be necessary to return to this question when we come to 

 discuss the factors of organic evolution. 1 



It is not sufficient for the purposes of classification to arrange the 

 individual plants or animals with which we are acquainted in their 

 respective species. These species in turn must be arranged in 

 groups of higher order, which are termed genera. Each genus 

 stands to the species included within it in much the same relation 

 that the species themselves stand in with regard to individuals, 

 and related genera, though often sharply definable, are only so by 

 virtue of the disappearance of connecting links. The characters 

 by which genera can be distinguished from one another are of 

 a more deep-seated and fundamental nature than those which 

 distinguish species. Individual opinion, however, differs greatly 

 as to the exact limits which should be assigned to each, and it is a 

 common thing for an old established genus to be subdivided into 

 several as the result of the discovery of new species and the more 

 exhaustive study of old ones. 2 It is all a matter of convenience, 

 and one may justifiably make a separate genus for any group 

 of closely related species which can be distinguished by some 

 common character from all other co-existing groups of species. 

 Many naturalists, however, consider that the mutual relation- 

 ships of species are more accurately expressed by making use of 

 subgenera as intermediate groups between genera and species. 



1 Vide Chapter XXVII. 



2 A striking illustration of this is afforded by the recent history of the old genera 

 Peripatus and Amphioxus, each of which is now subdivided into several. 



B. Q 



