The Theory of Evolution 215 



and suggestive their work may be, it needs to be carefully revised 

 and verified before it is safe to build on it a theory of evolution. 

 In the case of artificial selection it now seems probable that 

 selection of definite variations, the introduction of wild races, 

 and the results of hybridizing have all played a far more im- 

 portant r61e than the selection of -fluctuating variations. The 

 evidence is clearer in the case of plants than in that of animals. 

 In animals the domesticated races have been, in most cases,, 

 longer in confinement, and the history of the origin of the different 

 forms is lost in the remote past. Nevertheless it is not probable 

 that the method employed in producing domestic races of ani- 

 mals has been essentially different from that of plants ; and in a 

 few cases, at least, there is evidence that the same method has 

 been employed for both. 



The Influence of the Environment 



Zoologists and botanists alike have been impressed by the fact 

 that under different external conditions different species are found. 

 While this statement does not hold for all species, yet it holds 

 for many. Natural barriers also are likely to separate different 

 forms (races) of the same species, and isolation of any kind seems 

 to be connected in some way with the occurrence of different 

 forms. 1 It is not my purpose here to consider the different 

 hypotheses that have been built up on this evidence, but the 

 fact itself is so patent and so important for our present purpose 

 that it must receive here somewhat fuller treatment. 2 



A not unnatural inference from the facts of distribution is that 



1 Gulick has recently (1905) gone over this ground exhaustively, and the in- 

 fluence of isolation has been discussed by a number of specialists, Jordan, Allen, 

 Merriam, and others in recent numbers of Science (1905-1906). 



2 Isolation is often assumed to produce its results by preventing the new forms 

 (however they may arise) from crossing back with the parent species, yet it is 

 by no means true for all species and varieties that they are isolated from each 

 other except in so far as their special conditions of life are different. Internal dif- 

 ferences may produce physiological isolation which appears as potent a factor as 

 geographical isolation. It might be confusing cause and effect to assume in 

 the former case that the isolation produced the new forms. A new mutation 

 may itself cause its own isolation. 



