CROSSING. 85 



A very unfortunate confusion of terminology has resulted 

 from this cooperation between de Vries and Mr. Burbank, for, 

 curiously enough, several very able Geneticians, including 

 Morgan and Davenport now use the term "mutation" in Bur- 

 bank's sense, instead of in the way the term was applied by de 

 Vries. In the chapter on Mutation we will try to show that, 

 whereas real mutation, in de Vries' sense of a spontaneous 

 change in the personnel of the genes is too rare and too insig- 

 nificant to be a real factor in evolution, it nevertheless doesoccur 

 occasionally. If we could be sure, that there did not exist real 

 cases of spontaneous loss-mutation, it would be perfectly allow- 

 able to use the term "mutation" for "individual with a wholly 

 new character" in the way it is used by so many American 

 authors after Burbank. For it is evident that almost all the 

 cases of the production of a new biotype with some wholly 

 novel characteristic, and which are claimed by de Vries as due 

 to spontaneous changes in the genes themselves, are due to 

 recombinations of genes. 



The very simplicity of the earliest cases of "germical ana- 

 lysis" has resulted in a temporary strengthening of the posit- 

 ion of those authors, who held fast to the Weismannian con- 

 ception of heredity as a transmission of determinants for de- 

 finite organs, and definite characters. When it became appar- 

 ent that in the majority of instances, there is no simple segre- 

 gation of the off-spring of hybrids into classes with clear-cut 

 characters corresponding to the parental ones, the result was, 

 that a distinction was made by several authors between two 

 kinds of inheritance, alternative inheritance and blending 

 inheritance. 



It is apparent, that, if we consider the characters of the or- 

 ganisms we cross, their phenotype, the inheritance of almost 

 every character envisaged is blending. Unless two individuals 

 are very closely related, but differ mainly through presence and 

 absence of one or two genes which have a very pronounced in- 

 fluence on some very marked character, the off-spring of two 

 diverse individuals is generally intermediate. This fact was 



