HEREDITY, AND THE OEIGIN OF SPECIES MACDOUGAL. 519 



the primary idea, although the action might become intensified in 

 certain periods of more or less definite limits. 



Finally we may consider the causes inducing or affecting muta- 

 tions. Mutants are found to be the most numerous under conditions 

 most favorable to the growth and reproduction of the parental type, 

 and this is also true of all anomalous structures. It is thus to be seen 

 that mutations probably do not give rise to species most readily 

 under the stress of unfavorable environment, or under anv conditions 



Fig. 6. — Correlation table of length and width of leaf in 0. Lamarckiaiia. (p=0.701(5± 



0.0090.) (After Sbull.) 



which weaken the parental form, but when it is at a maximum of 

 activity. 



In the attempts to localize the changes in the cells, or in the chro- 

 mosomes, which result in the formation of atypic individuals from 

 seeds, we are confident for theoretical reasons that these ensue pre- 

 vious to the reducing or qualitative divisions in the' formation of the 

 egg, or in the pollen mother cells. Just what this change may be we 

 are unable to say, and shall probably know no more about it until 

 88292— SM 1908 34 



