226 UNIVERSITY OF MISSOURI BULLETIN 



of evolution it has nothing to do. Only after the step has 

 been taken does the sieve act, eliminating the unfit." (de 

 Vries.) The single steps, recent investigations have shown, 

 are taken by mutation. 



Before leaving this subject, I should not fail to mention 

 the fact that this fundamental distinction between fluctuations 

 and discontinuous variations has received strong support from 

 the brilliant experiments of Professor Johannsen of Copen- 

 hagen, who has apparently established the fact that a gradual 

 accumulation of minute fluctuating variations through a con- 

 tinuous selective process is impossible. That such variations 

 could be so accumulated is the basic assumption of Darwin's 

 theory. 



One of Johannsen's investigations, which deals with the 

 inheritance of the weight of beans, will make the matter clear, 

 although it will be impossible to do more than indicate the 

 general result. If seeds are gathered at random from a patch 

 of beans, it will be found that they vary considerably in weight, 

 but that the large majority will not differ greatly from the 

 average weight of the total number. As we approach the 

 upper and lower extremes, however, the number of beans 

 possessing these less frequent weights will constantly and sym- 

 metrically decrease. It has been hitherto customary to re- 

 gard such variations as of a continuous nature, that is, as all 

 chance variations in a homogeneous mass or population, and 

 it has been supposed by many that these slight differences 

 could be accumulated and increased by selective breeding. 

 Such, however, Johannsen has shown not to be the case. He 

 has proved clearly that the material, instead of constituting a 

 homogeneous population is in reality composed of a mixture 

 of several different strains or so-called pure-lines, each one 

 of which has its own average weight, and that around this 

 average the individuals composing a given pure-line fluctuate 

 in the usual way. Although the different pure-lines may over- 

 lap at their extremes, they are in reality discontinuous, as 



