NATURAL SELECTION 183 



Throughout the work such a process is suggested and assumed : 

 its actual occurrence is nowhere demonstrated. Stated briefly, 

 the argument is as follows : selection has plainly c worked ' 

 in domesticated races, analogous results and appropriate 

 processes and conditions are found in nature, therefore we 

 may assume that selection works in nature. In short, the 

 proof is based on circumstantial rather than direct evidence, 

 and the mainstay of the case is the analogy between Artificial 

 and Natural Selection. 



On the question of variation Darwin's mind evidently 

 hovered in some uncertainty. He clearly thought of it ' as 

 indefinite and almost illimitable ' (' Animals and Plants under 

 Domestication,' ii, 292). In the sixth edition of ' The Origin ' 

 (1884, p. 648) he was still under the impression that to some 

 extent ' physical, i.e. environmental conditions seem to have 

 produced some direct and definite effect . . . with both 

 varieties and species use and disuse seem to have produced 

 a considerable effect.' Nevertheless in ' Animals and Plants ' 

 (I.e.) he had doubted whether ' well-marked varieties have 

 often been produced by the direct action of changed condi- 

 tions without the aid of selection either by man or nature.' 

 Bateson (1909, p. 209) points out that Darwin originally held 

 that ' individual variation ' (i.e. mutation) was of high im- 

 portance, but subsequently abandoned the belief. With 

 these minor inconsistencies and changes of opinion we need 

 not occupy ourselves. 



It is far more relevant that, though the importance of 

 Natural Selection is always stressed, Darwin nowhere suggests 

 that it is the only modifying agency. He always laid stress 

 on isolation and correlation and, as we have seen, on the 

 effect of the environment. He even goes so far as to suggest 

 that the modification of a species may proceed without selec- 

 tion — that species may arise and be perpetuated ' for no ap- 

 parent reason.' He carefully disposes of a (for him) too rigid 

 and literal application of the theory — e.g. when he shows that 

 Bronn's objection to it, based on the occurrence of parent 

 species and their varieties living side by side, may be met by 

 assuming that, if both had become fitted for slightly different 

 habitats, they might subsequently extend their ranges and 

 overlap (1884, P- 2D 4). It is quite clear that he thought that 

 varieties might arise and species might exist without having 



