1915~\ Outline on the Theory of Descent 15 



Johanssen, to cite the work of but one of them, experimenting 

 with beans, isohited nineteen such groups of pure lines, 

 " . . . . the progeny of each of these pure lines of beans 

 varied around its own mean, which was different in each of 

 the nineteen instances." It matters not whether the smallest 

 or the largest individuals are chosen to breed from, the 

 progeny in either case is the same, which is the average for 

 the pure line in question. The conclusion then is that ^'Se- 

 lection within a pure line is absolutely without effect in 

 modifying a particular character in the offspring of the line 

 in question." I^or can selection from the mixture of pure 

 lines which constitute the species accomplish more. The 

 utmost that it can do is to isolate that pure line which ex- 

 hibits the character in question developed to the highest de- 

 gree. Manifestly this is a discovery of far-reaching signifi- 

 cance, for if these things be so, how can the transition from 

 one pure line to another have been made ? That such tran- 

 sitions have repeatedly occurred is, it would seem, beyond 

 intelligent question, but how, looms larger today than ever 

 before. There seem to be but two answers now made to this 

 question : one of these is mutation, which makes the passage 

 from one pure line to another at a single bound, but which 

 seems to me much too questionable a support for any great 

 super-structure ; the other is the inheritance of acquired 

 characters, an ancient dogma but not now in good standing in 

 most approved scientific circles. Later on I shall discuss each 

 of these theories at some length, but now before leaving this 

 topic it should be added that Johanssen has recently reported 

 mutations within his pure lines. This is, however, as Walter 

 has pointed out, a clear case of the logician's "vicious circle." 

 For so long as a variation does not reappear in the progeny 

 it is taken to prove that such variations are individual, due 

 to effects of the environment, and not heritable, but, when- 

 ever such variations do reappear, they are at once styled 

 mutations. 



DARWINISlSr 



With your indulgence, I will now sketch in as briefly as 



