ORIGIN 01^ SPEiCIliS. 313 



points out that "mutation" is complementary and not antagonistic to natural 

 selection, and that the two are necessary in evolution — that is, "mutation" 

 explains the origin of variations in evolution, and natural selection their 

 preservation. 



As I understand natural selection, Darwin never claimed that it originated 

 anything, but only worked upon that which was already in existence ; but we 

 must admit that since any variation depends upon the antecedent series of 

 stages in the organism, that any stages in the organism which survive through 

 natural selection or are not eliminated thereby, as in characters of non- 

 selective value, exert a powerful influence in evolution in that they limit 

 variation to certain possibilities. Morgan calls this "very plausible," but 

 careful study of any set of organisms and their variations shows that it is 

 actually true. This fact we saw clearly in the second and third chapters of 

 this paper. In this way selection, by determining the survival, acts also to 

 eliminate certain possibilities from future evolution. 



Of recent years there has been a deal of writing in which natural selection 

 is denounced as incompetent and trivial. However, few, if any, of these 

 "incontrovertible proofs" rest upon anything more than poor logic. Natural 

 selection is a subject for investigation, not for argumentative denunciation. 



In relatively few cases do we know just how selection works — just what 

 selection does — and not until we know accurately the action of selection can 

 we begin to estimate its effect in evolution. In various ways selection has 

 been shown to he actively at work in Leptinotarsa, hut as far as discovered 

 always in conservative ways, eliminating extremes and limiting the reproduc- 

 tive population to the individuals nearest to the racial mean. Thus in hiber- 

 nation the extreme variations are all eliminated as far as the evidence goes, 

 and a "mutant" like pallida or melanicum has only the remotest chances of 

 becoming established in nature. It is not possible to say at present how poor 

 their chances are, but their occurrence once in 6,000 cases is not favorable to 

 their becoming established. If enough mutants arose at one time we could 

 easily imagine how they might become established, but this does not seem to 

 happen. In the tropics L. melanothorax has been known for nearly fifty 

 years, yet it has not been able to gain a foothold, although it is far more 

 numerous than pallida. L. ruhicunda seems more fortunate, but its distribu- 

 tion is as yet very limited, and seems to be decreasing, being one-third less in 

 1905 than in 1904. Its future history will depend on whether it will or will 

 not be able to meet the conditions of its existence. 



The real question in the method of evolution at present is whether species 

 arise by the preservation of large variations or "mutants" or by small 

 accumulated variations. It appears that in both plants and animals large 

 variations occur in nature, but these, as far as all evidence goes to show, are 

 most rigorously exterminated by natural selection, and only the mean and 



