46 ORGANIC EVOLUTION 



We do not care here to discuss in detail the Mendelian 

 laws, their cytological explanation, and the exceptions to 

 them, though these subjects are most interesting and im- 

 portant. We are chiefly interested, in the present connec- 

 tion, in the fact that if the individuals crossed be sufficiently 

 divergent the result is not a mere admixture of the qualities 

 of the two parents in the young, but that individuals of 

 pure strain, showing no admixture, appear in the third 

 generation and in succeeding generations. Very divergent 

 individuals which arise by variation are commonly called 

 " sports." It is easy to see that if a single brood of sports 

 arose which were especially well adapted to their environ- 

 ment, although they might breed with non-divergent indi- 

 viduals of the species, yet among the offspring of the third 

 generation there would be individuals like the original 

 sports. It might, therefore, be possible for natural selection 

 to change the character of the species from the old type 

 to that of the sport, by preserving the sports and allowing 

 them by competition to destroy the individuals of the old 

 type. Should the sports prove to be more fertile when 

 crossed with one another than when crossed with individ- 

 uals of the old type this would increase the probability of 

 the new type becoming predominant. 



It may be that less divergent characters also may be 

 preserved from immediate swamping by intercrossing, but 

 it is too early in our study of the Mendelian phenomena 

 for us to be able to say. We do not know whether the 

 Mendelian laws apply at all to ordinary varieties or only to 

 sports. If they apply to ordinary varieties of course the 

 possible effect upon evolution would be greater. 



We should also note that in the experiments of Mendel, 



