EVOLUTION IN NATURE AND UNDER DOMESTICATION. 245 



the statement by adding — or fitting different ecological en- 

 vironments, we would not state the whole truth. We would 

 certainly have enumerated the main causes tor specific dis- 

 tinctness. To go to the root of the matter we have to state, 

 that species remain distinct, when the automatic reduction of 

 the potential variability of the groups through any or all 

 reasons, outweighs the heightening of the potential variability 

 by crosses or otherwise. 



Evolution, the establishing of groups with new genotype is 

 impossible within freely crossing populations. Whenever 

 there exist causes which limit free crossing to a certain extent, 

 and which depend upon nature of the factors reducing the 

 potential variability, such as colonizing habits, rapid repro- 

 duction, evolution can proceed. Geographic barriers, slow dis- 

 persal, autogamy, different size etc. are all at certain times 

 causes for the differentiation of species. And we have seen that 

 neither of these causes is in itself indispensible, one can replace 

 the other in the establishment of the final balance between the 

 factors heightening and those reducing the potential variabil- 

 ity of groups. 



Is it possible from an inspection of groups of organisms as 

 they occur in nature, to estimate what factors from among 

 the list of those possible have cooperated to establish them ? 

 To a certain extent it is. 



If we find sharply demarcated groups of closely-related organ- 

 isms differing in trivial characters, such as the local species of 

 deer-mice or ground-squirrels, we may feel safe in assuming 

 that just chance determined for what type, possible of reali- 

 zation in the potential variability of isolated colonies the group 

 has become pure, and if intermediate forms are absent we 

 can feel reasonably sure of the existence of real barriers. 



Should we on the other hand, meet with forms which grade 

 more or less gradually from one local species to the other, the 

 existence of barriers becomes doubtful, and we have reason to 

 suspect that the animals in question simply keep from mixing 

 because of a slow rate of dispersal. In such cases where bar- 



