MUTATION. 141 



translate itself in a discontinuous variation, are essential to 

 the theory, that such abrupt changes may occur and may play 

 a role in evolution. We know, that the effect of a great number 

 of genes studied in the most diverse organisms, is very slight. 

 We do not think any author on the subject has accepted the 

 idea of a periodicity in mutation. 



If genes are lost or acquired spontaneously at times, such an 

 occurrence may very well pass unnoticed, if the gene in question 

 does not happen to be an obvious factor in the develop- 

 ment of organisms with this genotype. 



As, theoretically, a mutation can consist of either the spon- 

 taneous acquisition of a gene, or of the spontaneous loss, we 

 are confronted with a difficulty. Is it possible in a given in- 

 stance to know whether a gene has been added or whether a 

 gene has been lost? The point is rather important, because of 

 the obviously unequal interest attached to these two theoret- 

 ical processes. If we see two organisms who obviously differ 

 in one gene, how can we judge which one has the gene in ques- 

 tion and which one lacks it ? We think that dominance of any 

 quality of an individual over a corresponding character of an- 

 other, proves the presence in the first one of something which 

 is absent from the other. If two animals differ in colour be- 

 cause from the germ of one of them there is lacking an indis- 

 pensable link in the chain of factors necessary for pigmenta- 

 tion, the hybrid will be coloured because it inherited that nec- 

 essary factor, if only from one parent. In such a case we as- 

 sume that an individual which is impure, heterozygous for a gene 

 will show in its development the action of this gene as strongly 

 or nearly as strongly as an individual which is pure for it, in 

 other words that a single dose of the gene ultimately has ap- 

 proximately the same effect on the development as a double 

 dose. This is the "presence and absence" theory as I proposed 

 it in 1908. 



We have to go carefully here, so as not to be deceived by the 

 apparent fitness of the hypothesis, for, as Shull has pointed 

 out, a hybrid from a cross between individuals of which one 



