142 Controversial Questions. 



have 2 or 3 leaves in addition to the cotyledons. If, 

 after that, I make the conditions as favorable as possible, 

 I get large and vigorous plants with but few stamens 

 transformed into carpels. 



By this device therefore it is possible to separate the 

 two acquired characters, previously associated. The ac- 

 quisition of the visible monstrosity is prevented ; but the 

 seeds attain their full development. 



Parallel experiments with other species and other 

 characters have convinced me that we are dealing in this 

 case with a universal and very important principle. I 

 mean the simultaneous influence of the conditions of life 

 on the visible character of an organism and on its germ 

 cells. In other words (using the word nutrition in its 

 old broad sense) we may say that selection is the choice 

 of the best nourished individuals. 



This statement can only be taken in a very general 

 sense; for there are individual cases to which it does not 

 seem to apply, as for instance the case of the selection 

 of ^/znrz/.S'-variations. Moreover in other cases there are 

 special circumstances which prevent its exact application, 

 for example in agricultural selection where the plants 

 have to be adapted to a supply of manure which can never 

 be copious.^ 



We are thus led to see that a proper understanding 

 of the difference between the theories of Selection and 

 Mutation opens up the possibility of a solution of the 

 question of the inheritance of acquired characters. Spe- 

 cific characters are excluded once and for all from this 

 discussion ; they arise suddenly by mutation and are not 

 acquired. Individual deviations from the mean of the 

 specific character are to be regarded as acquired char- 



' On this point see the previous chapter especially § I2. 



