MAKING jyHW KINDS OF PLANTS 451 



that this aloue can give rise to a new species. Further- 

 more, he includes in mutation the effects of crossing: 

 hf' believes that new species can arise from hybridiza- 

 tion and that there are well-estalilished instances of 

 this. 



Professor de Vries does not deny the great value of 

 selection for cultivated plants: he merely insists that 

 the eifect of selection is only temporary and soon 

 ceases when selection stops. Each year the seedsman 

 must carefully go through his beds and "rogue," i. e., 

 remove the rogues or plants which do not come true 

 to seed or which in any w&y fall short of the standard: 

 only so can he keep the seed at all pure and true 

 to type. And this is the best proof that he has not 

 succeeded by selection in making a true species. 

 Where he does get a constant form. Professor de Vries 

 believes it is due to the selection of mutations which 

 the seedsman does not distinguish from fluctuating 

 variations. He believes that our garden varieties have 

 mostly originated in this way, since he finds that most 

 of those with which he has experimented are constant. 

 The reason why they are not regarded as such is that 

 they are usually grown side by side with the parent 

 forms (or other nearly allied forms) , so that crossing 

 takes place and hybrid seeds are produced: these, when 

 sown, give variable forms after the manner of hybrids, 

 and much confusion results. His experience is that 

 if a pure (i. e., not crossed) white ^'ariety of a 



