14 The Mechanism of Evolution in Leptinotaksa 



century, began a new attack upon the problem of species, which was based in 

 the main upon the conception that if species were discontinuous when fully 

 formed, a priori there was no reason why they may not have arisen discon- 

 tinuously at the start. In support of this proposition there was found in the 

 older writings of plant and animal breeders, and especially in Darwin's works, 

 many records of the rise of new domesticated races of organisms based upon the 

 appearance of an ancestral sport. The attack upon the problem was made from 

 quite diiferent points of view, but nevertheless all have arrived at a somewhat 

 similar result, namely, that certain discontinuous changes in the qualities of the 

 organisms were quite as likely to be productive of evolution as the methods of 

 slow accumulation so currently held by the neo-Darwinian school. Bateson's 

 compilation of instances of discontinuous variation and De Vries's experimental 

 investigation of the process which he calls " mutation " in Oenothera lamarck- 

 iana, the experimental production and observed rise of stable forms by Mac- 

 Dougal, Tower, Kammerer, Morgan, and others, in the main have provided the 

 basis for the hypothesis that effective changes in the qualities of organisms may 

 arise discontinuously in many directions, producing, as De Vries says, some- 

 thing quite new each time. 



MUTATION. 



The mutation theory of De Vries has raised a number of interesting and 

 important points for consideration, chief of which are his efforts to distinguish 

 between the conditions present in any population, dividing them into sharply 

 separated categories of " fluctuations," which are ineffective in transmutation, 

 and " mutations," which bear the burden of the evolution process. 



De Vries's argument is based largely upon the data of biometric study and 

 the more certain showing that in domesticated plants fluctuating variations are 

 not capable of accumulation beyond an amount which was present in the race at 

 the start as the limits of " variation," and further, upon the showing that all of 

 these variations are either plus or minus additions to the character and in no 

 other directions, while mutations are stable, effective, and in many directions. 

 There is no doubt that descriptively De Vries's account is correct, but I am 

 strongly of the opinion that both the account and the definitions are actually 

 incorrect. I have pointed out previously in this introduction that it is necessary 

 to recognize three general classes of characteristics in organisms, and that of 

 these characteristics one, especially, the attributes, which distinguish bodies 

 of the same kind, has been almost exclusively used in this selective accumula- 

 tion experimentation, and that there is not present in the process any mechan- 

 ism whereby the qualities of the race may be altered by accumulation of the 

 amount of the attribute present, and this in every instance is the effort which 

 the worker of cumulative selection seems to have made. With the sugar beet it 

 was increased in the amount of sugar. With other forms it is increase in stature 

 or increased or diminished amounts of one thing or the other within the race 

 which serves only to distinguish the individuals of the race, and not the races 

 from some other ; so that in all these operations there is present only the means 

 to attempt to accentuate the attributes present, and there is no possible mechan- 

 ism available for modifying the qualities of the race to which the attributes 

 belong. Whether these attributes vary only in one direction, plus and minus, is 

 largely a matter of definition and the method of investigation employed ; and in 



