32 THE MENDEL JOURNAL 



Selection undoubtedly decided in the affirmative 

 the general question of organic evolution, for there 

 can be no reasonable doidot either as to the fact of 

 selection taking place in nature, or as to the occur- 

 rence of favourable variations to be selected. But it 

 was more or less tacitly assumed that all kinds of 

 variations could be inherited and thus serve as 

 material for the production of new forms by selection. 

 Variation phenomena, however, as is now well known, 

 are of many different kinds, and, apart from the 

 variations due to the recombination and interaction 

 of characters alluded to above, there has been much 

 uncertainty as to the role, in the process of evolu- 

 tion, of continuous variations (i.e., variations which 

 are connected with one another by a continuous 

 series of intermediate forms) on the one hand, and 

 discontinuous variations on the other. It is here 

 that the evidence from the Mendelian researches 

 seems to come to the support of the view that it is 

 the discontinuous variations which really count in 

 evolution. For if the factors for the differentiating 

 characters between races and species are treated as 

 separable entities by the cell-divisions concerned 

 in the production of the germ-cells, it is extremely 

 improbable that these characters could have been 

 evolved from one another by the selection of im- 

 perceptibly small variations. In other words, 

 segregation of the character-factors involves an 

 antithesis between the factors which is inconsistent 

 with the idea of continuous variation. So far, 

 therefore, as Mendel's Law holds good, and it is 



