CROSSING. 73 



ity of a hitch in the process, resulting in the absence of some 

 gene in a few cells. 



It must however ; be well-nigh impossible to distinguish be- 

 tween a real loss-mutation in animals and the production of a 

 recessive novelty through re-arrangement of genes. And it 

 is possible that, notwithstanding the breeding-tests performed 

 upon the "mutating" individual mice in our colour-inheritance 

 series, the production of the recessive novelties may have been 

 caused by simultaneous absence of two genes in respect to 

 which each grand-parent was pure for one. 



In the light of this difficulty, it becomes significant, that 

 hardly any cases of "dominant mutations" are recorded, and 

 none at all in "pure lines. "We think, we can, practically speak- 

 ing, exclude mutation as a factor in that variation, which can 

 furnish the material for evolution. 



A number of very good instances of the origin of recessive 

 novelties, which simulate loss-mutations have been studied by 

 Brainerd. He found numerous cases, in which apparently a 

 good species had spontaneously produced one or more varie- 

 ties. He found a few Viola affinis plants with black seeds in- 

 stead of seeds of the normal yellow colour. These plants bred 

 quite true. In every character, other than seed-colour they were 

 typically normal Viola affinis. Another plant found was a nor- 

 mal Viola cucullata with the exception of having dark purple 

 capsules instead of clear green ones. A plant of Viola nephro- 

 phylla was found, which carried buff seeds instead of black ones 

 and which bred quite true to the new character. Later Brainerd 

 could show by cross-breeding experiments that these apparent 

 mutations v/ere the result of crossing between species. He was 

 able to reproduce them and produce new similar instances. 



Very striking examples of the origin of new characters by 

 cross-breeding are found in a series of experiments with rats of 

 the Mus rattus group, which we have been continuing for a 

 number of years. Some of this work was started by Bonhote, 

 and continued by us, when he left Europe. Bonhote mated 

 white-bellied tree-rats from Egypt with the gray-bellied 



