178 



THE AMERICAN NATURALIST [Vol. LII 



heredity and the definite formulation of a "mutation 

 theory" of evolution, some of Galton's more or less tenta- 

 tive views have crystallized into dogmas. Along with the 

 two just mentioned, there has been incorporated the prin- 

 ciple of the "continuity of the germ-plasm," a conception 

 which was likewise first clearly formulated by the great 

 English geneticist, though its modern expression we owe 

 to Weismann. 



These various hypotheses have been woven together 

 into a single fabric and made to reinforce one another. 

 It will hardly be denied that some rather flimsy reason- 

 ing has been employed at times by those here concerned. 

 Thus one familiar syllogism runs somewhat as follows: 

 Somatic modifications are not inherited; fluctuating varia- 

 tions are not inherited; therefore fluctuating variations 

 are somatic modifications, hideed, "somatic" and "non- 

 hereditary" have come to be used interchangeably by 

 many writers. Whether or not somatic modifications 

 ever become germinal is a matter to be settled by evi- 

 dence. But I must confess that I have never regarded 

 as self-evident the contention that because characters are 

 found to be "non-hereditary" they are, ipso facto, "so- 

 matic ' ' in origin. 



A certain sanctity and inviolability has come to be at- 

 tached to the units of heredity or "genes," according to 

 the neo-Mendelian creed. Not only do these units refrain 

 from any degree of blending, but— save for occasional 

 mysterious "mutations "—they are quantitatively and 

 qualitatively unchangeable. Thus, the only differences 

 upon which selection, natural or artificial, can act are dif- 

 ferences due to the presence or absence of different 

 genetic factors. "We know," say the Hagedoorns, in an 

 article (1917) which is typical of much of the recent lit- 

 erature of heredity, "that all the different genes, all the 

 different inherited factors . . . are each in themselves 

 invariable. . . . Liability to change by selection is synon- 

 ymous with genotypic variability, and this true variability 

 is synonymous with impurity." 



