172 



THE AMERICAN NATURALIST [Vol. XLIX 



ian manner, and, as they are capable of being separately 

 inherited and interchanged with others by hybridization, 

 we are justified in believing that they are separately 

 represented in the germ-cells by special factors. Im- 

 portant as this result is, I believe that at the present 

 time there exists a distinct danger of exaggerating its 

 significance. The fact that many new and apparently 

 permanent combinations of characters may arise through 

 hybridization, and that the organisms thus produced 

 have all the attributes of what we call distinct species, 

 does not justify us in accepting the grotesque view — as it 

 appears to me — that all species have arisen by crossing, 

 or even the view that the organism is entirely built up 

 of separately transmissible 1 'unit characters." 

 Bateson tells us that 



Baur has for example crossed species so unlike as Antirrhinum majus 

 and molle, forms differing from each other in almost every feature 

 of organization. 



Surely the latter part of this statement can not be 

 correct, for after all Antirrhinum majus and molle are 

 both snapdragons, and exhibit all the essential charac- 

 ters of snapdragons. 



I think it is a most significant fact that the only char- 

 acters which appear to be inherited in Mendelian fashion 

 are comparatively trivial features of the organism which 

 must have arisen during the last stages of phylogeny. 

 This is necessarily the case, for any two organisms suffi- 

 ciently nearly related to be capable of crossing are iden- 

 tical as regards the vast majority of their characters. 

 It is only those few points in which they differ that re- 

 main to be experimented on. Moreover, the characters 

 in question appear to be all non-adaptive, having no ob- 

 vious relation to the environment and no particular value 

 in the struggle for existence. They are clearly what 

 AVeismann calls blastogenic characters, originating in 

 the germ-plasm, and are probably identical with the mu- 

 tations of de Vries. These latter are apparently chro- 



