492 On the Latent Capacity for Mutation. 



a way as to make it highly variable or mutable. Louis 

 ViLMORiN, who used the French word affoler for this 

 operation, proposed the following method :^ Look through 

 a crop for individuals which differ most from the normal 

 in any direction, not necessarily the one leading to the 

 ideal aimed at. Sow its seeds; and in the crop raised, 

 do not, as in ordinary selection, choose the individuals 

 which deviate most in the same direction, but such as are 

 abnormal in the opposite one. Repeat this mode of se- 

 lection through a series of generations and it is antici- 

 pated, says ViLMORiN, that the variability will gradually 

 increase, until finally it becomes so great that it will 

 give rise to any new character desired. 



It does not appear that Vilmorin carried out such an 

 experiment and still less that he ever saw the result he 

 anticipated. 



Nevertheless his suggestion deserves attention ; it may 

 contain the germ of truth. And anyhow the method 

 would be likely to lead to the discovery of latent muta- 

 tions. 



One of the most important conclusions which may 

 be drawn from these general considerations is that a 

 premutation must be due partly to internal and partly 

 to external causes. The former determine what shall 

 arise ; the latter when it arises. 



The external causes must be other than the ordinary 

 conditions of life under which the species remain con- 

 stant. On the other hand they must be such as appear 

 not too rarely in the natural state. I assume these causes 

 to be perhaps a combination of extremely favorable with 

 extremely unfavorable influences : this view would at any 



_^ Louis Vilmorin, Notice sur ramcUoration des plantes par le 

 semis. Nouv. Edition, 1886, p. 36. 



