414 EVOLUTION, GENETICS, AND EUGENICS 



Kammerer found that salamanders with black and yellow spots 

 when reared on yellow soil gradually lose their black color, becoming 

 more yellow, and their young continue to grow more yellow until 

 finally almost all black may disappear. The offspring of such sala- 

 manders are said to be more yellow than normal; but this work has 

 been called in question and needs confirmation. Even if confirmed 

 the result may be an after effect or "induction" which would soon 

 disappear under usual conditions, and there is no evidence that it is 

 really inherited. 



Such cases are not instances of true inheritance; they do not 

 signify a change in the hereditary constitution but an influence on the 

 germ cells of a nutritive or chemical sort comparable with what takes 

 place when fat stains are fed to animals; the eggs of such animals are 

 stained, and the young which develop from such eggs are also stained , 

 though the germinal constitution remains unchanged. The very fact 

 that the changed condition is reversible and that it disappears within 

 a short time is evidence that it is not really inherited. 



In conclusion: (i) Developed characters, whether "acquired" 

 or not, are never transmitted by heredity, and the hereditary constitu- 

 tion of the germ is not changed by changes in such characters. (2) 

 Possibly environmental stimuli acting upon germ cells at an early 

 stage in their development may rarely cause changes in hereditary 

 constitution, but changes produced in somatic cells do not cause 

 corresponding changes in the hereditary constitution of the germ cells. 

 (3) Germ cells like somatic cells may undergo modifications which are 

 not hereditary; if starved they may produce stunted individuals and 

 this effect may last for two or three generations; they may be stained 

 with fat stains and the generation to which they give rise be similarly 

 stained; they may be poisoned with alcohol or modified by tempera- 

 ture and such influence be carried over to the next generation without 

 becoming hereditary. All such cases are known as "induction" and 

 many instances of the supposed inheritance of acquired characters 

 come under this category. (4) Environment may profoundly modify 

 individual development but it does not generally modify heredity. 



THE OTHER SIDE OF THE QUESTION 



It will have been noted that the chief objection to the idea of the 

 possibilit)' of acquired characters being inherited comes to us as a 

 heritage of the rather extreme Weismannian concept of the "germ 



