72 CROSSING. 



which are shown to be present, without exerting any influ- 

 ence upon the development of the organism in which they 

 occur. 



The danger in the use of such special terms is that they make 

 appear exceptional what is usual. 



Crossing is practically the only cause of the production of 

 new dominant characters in a group of organisms. But it is 

 evident, that the new character as such need not be "intro- 

 duced" from a stock "carrying" it. Usually a new dominant 

 character, resulting from crossing must be looked upon as due 

 to a difference in the development from that in the common 

 type, caused by the presence in the zygote of one or more genes 

 not commonly present. Such "new" genes may so affect the 

 development in cooperation with all the other developmental 

 factors, that a new character results. The cross need not be a 

 wide one. Within one species with a sufficiently high potential 

 variability, combinations of individuals may be possible which 

 give off- spring with unexpected new characters. Dominant va- 

 rieties which can be made into domestic species, may originate 

 in this way, and such an instance may easily be mistaken for 

 positive mutation, for the spontaneous acquisition of a gene. 

 (Drosophila). 



New recessive characters may result from cross-breeding, and 

 it is evident, that this process looks so much like loss-mutation, 

 that it is only with the greatest difficulty, that it can be dis- 

 tinguished from it. Lotsy has accepted our view that almost 

 all the instances of loss-mutation recorded must have been 

 cases of the production of new forms, new characters, through 

 re-combination of genes. We would not subscribe to his state- 

 ment that mutation is a myth, because it appears difficult to 

 account for such mutations as have been seen to take place in 

 "pure lines" by both Johannsen and Nillsson Ehle, without 

 assuming that here a gene, present in the material was excluded 

 from some cells. If we remember, that every gene present 

 in a cell must be multiplied quantitatively and distributed over 

 millions of descendant cells, we can well imagine the possibil- 



