LEE BARKER WALTON 91 



Hons. On the other hand, variations due to the structural 

 changes in the germ cells which were passed on from one 

 generation to another have been spoken of as inherited 

 variations. 



The evidence at present indicates that farther sub- 

 divisions must be made and that normal inherited varia- 

 tions consist of two quite distinct classes. The variations 

 where the results are due to the interaction of factors in 

 accordance with Mendelian principle, and which, adapt- 

 ing a term used by Plate (1913), may be called amphi- 

 ■mutations inasmuch as the condition is due to the ming- 

 ling of two lines of descent, the other variations, as a class, 

 in which the results — evolution in the abstract — are due 

 to a series of units added as increments, may well be called 

 cumulations. It is quite evident that the term "mutation" 

 can not continue to include both types. As a coordinate 

 term fluctuating variations may be spoken of as fluctua- 

 tions. 



Under abnormal variations must be classified forms 

 ranging from monstrosities to slight departures from the 

 ordinary condition, some of which are undoubtedly due 

 to the losses or modifications of unit characters through 

 the action of extraordinary stimuli, while others may be 

 due to abnormal and unequal distribution of chromo- 

 somes occurring at the time of their division. The idio- 

 mutations of Plate are here included. 



The answer to the question as to the progress made in 

 the application of evolution to the creation of new forms 

 rests in the statement that the attack on the problem is 

 becoming more concentrated. The selection of fluctua- 

 tions has been tried and has failed. Efforts by means of 

 amphimutations end in a maze of circles with no evident 

 progress. Idiomutations, so far as one may judge from 

 the evidence, present retrogression rather than advance- 

 ment. It is by means of pure lines under normal condi- 

 tions that one may search with advantage for cumulations, 

 the units by which to build the new. There the evidence 

 will be unobscured either by the pyrotechnics of Mendel- 

 ian formulse, or by the factitiousness of abnormal stimuli. 

 Fluctuations will be present, but statistical methods will 



