136 Controversial Questions. 



Mutations are inherited and, as a rule, constant from 

 the time when they appear. Reversions to the parent 

 form are not wanting; but they are very rare and take 

 place as sports and not by a series of transitional forms. 

 We call this atavism. 



The inheritance of variations or deviations from the 

 mean of the type is quite a different thing. The children 

 deviate less, on the average, from tlie mean than the 

 parents do; on the other hand some individuals of them 

 mav differ more, and these enable us to increase the 

 deviation by means of selection. 



The answer to the question whether acquired char- 

 acters are inlierited, is that they are not so in their en- 

 tirety, but with a reduction the amount of which is in- 

 dicated by Galton's law. On the other hand the gradual 

 change in the mean character of a race which can be 

 effected by selection is sufficient proof that these char- 

 acters really are inherited. The question whether such 

 variations are inherited becomes the question whether 

 they can be increased by selection. And as far as I am 

 aware no investigations have been made which prove 

 that this cannot be done. 



The so-called innate characters as opposed to acquired 

 ones are believed to be inherited while the latter are not ; 

 but it is obvious that they are merely inherited devia- 

 tions from tiie mean, and that the ancestors, which 

 showed these deviations must have acquired them them- 

 selves under the influence of external conditions. It 

 would lead us too far to follow up this line of thought 

 although such a discussion would undoubtedly contribute 

 to the ultimate solution of this question. 



If we regard individual variations as brought about 

 by environment and by nutrition in the widest sense, 



