LEE BARKER WALTON 83 



than a tentative opinion, there are several conclusions 

 concerning mutation which appear justified. The nature 

 of the results obtained through the various agencies make 

 it quite evident that they are not all due to a single under- 

 lying principle. There are many "mutants" the origin of 

 which is most certainly to be explained on the basis of a 

 heterozygous condition of the gametes, and much evi- 

 dence has accumulated that O. lamarckiaua of De Vries 

 on which the mutation theory was founded belongs to 

 this class. Furthermore there are mutants developing in 

 connection with the action of abnormal stimuli although 

 it is not at all improbable that some of these result from 

 heterozygotes. It may be mentioned that Humbert (1911) 

 in experiments with 7,500 pure line plants of Silene 

 noctiflora, one of the "pinks" utilizing methods similar to 

 those of MacDougal, failed to obtain any "mutants." 

 Another explanation of the results in connection with the 

 influence of abnormal stimuli is that the modification 

 takes place through the destruction of a factor and thus 

 the process is one of subtraction instead of addition. 

 There are also investigations, notably those of Gates, in 

 which the aberrant organism apparently results from the 

 abnormal behavior of the chromosomes at some stage dur- 

 ing the life cycle. CEnothera gigas with its tetraploid 

 chromosomes is here of much interest. 



Notwithstanding these diverse results, there is little 

 indication that anything actually new has been added to 

 the organism which would not have occurred within a 

 pure line. If this is true the heterogeneous school of 

 mutationists can be of little assistance beyond suggesting 

 the way in which evolution did not take place. 



The experiments on the basis of pure line breeding be- 

 long to a comparatively recent period and are of the 

 utmost importance. Johannsen in 1903 published results 

 based on a pure line of beans self-fertilized for successive 

 generations and evidently homozygous. From a bean 

 weighing 95 centigrams and far above the average in size 

 he obtained plants producing beans varying in weight 

 from approximately 3 5 to 70 centigrams, but all far be- 

 low the weight of the parent. Utilizing these in turn as 



