38 THE EVOLUTION OF LIVING BEINGS. 



be explained by the acceptance of numerous so-called 

 factors, because aU this indicates complication and 

 therefore the possibility of errors of interpretation. 



We are therefore, for the present, forced to require 

 from him who wants to furnish proof for such an unex- 

 pected novel fact, as mutation would be, at the very 

 least: 



istly . that he has previously investigated, to the best of 

 his ability, by hybrid analysis, the purity of the 

 form from which the supposed mutant arose. 

 2dly . that the supposed mutant, crossed with the spe- 

 cies from which it arose, does not behave as a 

 recessive, but either dominates in the homo- 

 geneous and identical, reciprocal Fi generati- 

 ons, or shows in such FiS, characters interme- 

 diate between those of the original species and 

 the mutant; furthermore that the mutant and 

 the form from which it arose reappear in the 

 uncontaminated F,. generation, either in the 

 proportion 3 : i or in the proportion i mutant : 

 2 intermediates : i original form. 

 This must be at present the test required because, so 

 only, it can be proved that a homozygous individual can 

 become heterozygous by itself, in other words can 

 mutate. 



As far I as am aware, no pretended case of mutu- 

 tion can stand this test. 



There is consequently not the slightest proof for the 

 existence of mutation and Davis with whom I agree in 

 the main, goes in my opinion decidedly too far when 

 he says (Science XLII Nov. 5. 1915) : 



