100 Hagedoorn. 



work witli mice of one of us, a gene differentiating agouti from non- 

 agouti was found to be transmitted independently from a gene diffe- 

 I'entiating coloured from albino. But in one series there was found ;i 

 complete "repulsion" between two such genes. Recently another author, 

 finding that in his material two such genes were wholly independent, 

 actually called in doubt our facts. He would not have done so, if he 

 had realized, that the fact, that a gene differentiates between albino 

 and coloured, does not prove it to be identical with other genes having 

 this influence in other series. In fact, we now see that probably in 

 our own material we had either two different "agouti" genes or two 

 different "pigment" genes. 



There is only one safe way of denoting genes, and this is to make 

 the naming as provisional and temporary as we know how. If we use 

 a set of symbols. A, B, C, 1), E, etc. for the genes treated in a certain 

 paper, and carefully explain what these symbols stand for in this parti- 

 cular paper, we have done all we can to avoid confusion. We have 

 found, that the only reason for affixing a definite symbol to a gene, is 

 a feeling that this gene determines a certain definite quality, whereever 

 it is present, and we know that this idea is erroneous. 



In composing this paper, we have used the first letters of the 

 alphabet to denote the genes discovered. These letters do not corres- 

 pond to those used in our work with mice, not even with those we 

 employed in a former papei' treating of these same experiments with rats. 

 Whereas the Horticulturists at least concede, that al)Solutely new 

 characters in plants often result from ci-osses, and tend to ascribe the 

 origin of all their novel gnrden forms to such crosses, the animal breeders 

 and especially the geneticians who have worked with animals, are more 

 apt to ascribe the origin of new characters, shapes, colours, to spon- 

 taneous variation, unconnect(!d with crossing. Let us briefly examine 

 the evidence for our contention, that in animals as well as in plants, 

 crossing causes novel characters to appear. 



First we have the strongest of circumstantial evidence. In some 

 domestic animals only one species seems to have been taken into cul- 

 tivation, and no species are closely enough related to this one to pro- 

 duce fertile hybrids. In other domestic animals however, two or more 

 species have been domesticated, or evidence is at hand, showing that 

 fertile hybrids with related species are sometimes produced. Examples 

 of the first kind are the peacock and the guineafowl. Although both 

 birds aie among the very oldest of domestic animals, the peacock so 



