MENDELIAX P1IEXO.MEXA WITHOUT 

 DE VRIESIAX THEORY 1 



DR. W. J. SPILLMAN 

 U. S. Dept. of Agriculture 



That the phenomena of Mendel ian segregation of char- 

 acter pairs is inextricably linked with the de Vriesian 

 notion of pangenes is an opinion widely held, both by the 

 advocates of the theory of discontinuous variation and by 

 the opponents of that theory. Professor S. J. Holmes 

 was one of the first American biologists to point out the 

 fallacy of this idea. 2 He called attention to the fact 

 that due consideration of the phenomena reported by 

 Mendelian investigators did not lead to the theory of dis- 

 c -< »iit iiiuous variation as a necessary consequence. 



While in the present paper it is my purpose to present 

 an explanation of Mendelian phenomena without resort- 

 ing to the idea of unit characters, I do not wish to be un- 

 derstood as belittling the important work done by de Vries 

 and his followers. While contending that the de Vriesian 

 doctrine that organisms are aggregates of separately her- 

 itable characters is unsound, I believe this investigator 

 has uncovered a new type of variation which must be 

 reckoned with as a means of evolution, not by any means 

 the sole, or even the most important, means. For want 

 of space, 1 am compelled, in what follows, to present some 

 opinions without submitting all the evidence in favor of 

 them. However, an attempt will be made later to present 

 this evidence more fully. It seems to the writer that 

 what Darwin considered to be fluctuating variations, 

 amenable to the action of natural selection, may now be 

 regarded as consisting of four distinct types of variation. 

 First, we have those variations in the progeny of a given 

 individual, or pair of individuals, which are due to what 

 we may call "Mendelian recombination of characters.' ' 



