THE GROWTH OF GROUPS 5 



animals we may say that they are alike as regards certain 

 characteristics which may be enumerated, but we can 

 never be sure that the enumeration is final as regards 

 each individual. In defining any such collection as a 

 specific group, we close our description at a point which 

 is chosen in an arbitrary manner. Species defined in 

 this manner are merely conventional. On the other hand, 

 those botanists who speak of elementary species, must 

 believe, in regard to any particular species, that they 

 have explored the whole field of observation afforded by 

 it, that no one could see in it more than they have seen. 

 The botanist is more likely to arrive at this opinion than 

 the zoologist because the field for observation is simpler 

 in plants than in animals, there is less scope for disagree- 

 ment in observation ; he feels that he can see and 

 describe all that there is to see, that no man could see 

 more, hence to him species are indivisible, or real and not 

 merely conventional. 



If the reader is of the opinion that species of mammals 

 are also indivisible let him take another step higher and 

 regard mankind, the most complex of living things and 

 the one with which we are best acquainted. Among 

 mankind we see such a variety of inheritable charac- 

 teristics, present in some individuals but not in others, 

 that we cannot possibly regard any group as indivisible. 

 The individual alone is indivisible. 



When two persons discuss the origin of species they 

 will misunderstand one another if the one holds species 

 to be real and the other regards them as conventional. 

 If they differ on this preliminary point it is idle to 

 proceed further. 



In this book, species are assumed to be conventional. 



