294 THE THEORIES OF EVOLUTION 



give rise to polymorphism, as observed by Kellogg 

 in butterflies, but it is hard to see how that divergence 

 could grow to the point of constituting a specific 

 difference. 



The last mentioned objection can be formulated 

 against anj^ isolation theory, whether the isolation be 

 geographical or physiological. Whenever a species is 

 split up into two groups by the arising of a natural 

 barrier, there are certain differences which existed 

 previously and which on account of this barrier are 

 not likely to be wiped out by interbreeding ; they will 

 therefore endure, but why should they become accen- 

 tuated in the following generation? 



Their ulterior fate may depend, according to their 

 environment, upon characters which have nothing in 

 common with them. This must be true, in particular 

 in all cases of apogamy. On the other hand, if the 

 division results from some character of specific im- 

 portance, we must account in some way for the origin 

 of the new character and whatever factor accounts for 

 it will also account for the formation of the new spe- 

 cies regardless of the intermediary fact of isolation. 



The same holds true when the accentuation of 

 differences between organic beings on both sides of 

 the barrier is due to a difference in conditions. Here 

 again the important factor is the conditions and not 

 the incident which submitted the individual to the in- 

 fluence of these conditions. 



