88 CROSSING. 



The only really satisfactory proof of the existence of a gene 

 consists of showing its distribution over one half of the gametes 

 produced by a heterozygote, and its distinctness from others. 

 In the most favourable cases, this distribution is seen directly, 

 as a Mendelian segregation of the off-spring of a hybrid into two 

 classes. In other cases, where the influence of the gene is not as 

 striking, test-matings will have to be made to determine the 

 genotype of the off-spring of heterozygotes. 



Another difficulty in the path of germinal analysis is given in 

 the fact, that only very rarely we can so choose our material as 

 to reduce heterozygosis to the gene we want to study. Ordinar- 

 ily we can never know, how large the number of genes may be 

 for which two individuals we are crossing are not identical. 

 Even in those cases where the heterozygosis of the hybrids is 

 great, the distribution of such genes as have a very striking in- 

 fluence on the development is easily studied. But it is evident 

 that in very many cases several genes for which the parents 

 differ may influence the same character. 



When we are dealing with two distinct genes which influence 

 the colour of our rodents, it is often relatively easy to study the 

 distributions and mutual independence of both. But in other 

 cases the effect of the same two genes may become so much 

 alike that the analysis becomes hopelessly difficult. The only 

 way out in such cases is to reduce heterozygosis, to study A in 

 families pure in respect to B and vice versa. 



When we are studying the distribution, and the effect of a 

 gene, which in our material has a very strikking influence upon 

 the development, we can afford to neglect for the time-being 

 the influence of other genes, and of environmental factors, so 

 long as these do not influece the character we are interested in, 

 in such a way as to interfere with our analysis. But it is going 

 altogether too far to expect this great simplicity wherever we 

 want to study a gene, or to conclude, from the fact that such a 

 simple direct relation does not exist in every case between a 

 gene and a pair of contrasted characters, that there is a funda- 

 mental difference between the genes themselves. 



