CROSSING. 87 



I 



by a mere inspection of the characters of the individual grow- 

 ing from it. Now such cases, in which a clear-cut segregation 

 of the off-spring of heterozygotes unsuitable test-matings goes 

 absolutely parallel with the segregation of a gene must be rela- 

 tivdy rare. For it is evident, that such a distinct segregation 

 into two classes of individuals can only be expected in those 

 cases where the influence of the gene is relatively very great. 

 Black and agouti in rodents are alternative, quaUtative char- 

 acters. This means, that in a mixed group of only blacks and 

 agoutis, the presence or absence of the gene which by its pres- 

 ence or absence produces the difference, can be immediately 

 detected by a mere inspection of the colour of the animals. In 

 this case, it is at once clear, that this fortunate circumstance, 

 which so simplifies germinal analysis, is not so much due to the 

 nature of this gene, as to the relation between the influence of ' 

 this gene and that of the whole further genotype of the mem- 

 bers of iiie group. This same gene, which in this case, in ani- 

 mals of this genotype, differentiates blacks from agoutis will 

 have very much less influence in other groups, and even none 

 at all in some. In light silvers, and mice of very hght shades 

 generally, the presence or absence of this same gene has so 

 little influence upon actual colour, that there is no possibility 

 of grading the off-spring of heterozygotes into two classes. 

 Only test-matings with suitable animals can show the presence 

 or absence of our gene in animals with doubtful coloiu". The 

 difference between animals with and without the gene is not as 

 great as differences in shade due to age, sex, moulting, general 

 health. Presence or absence of one and the same gene in the 

 same material, may result in some cases in qualitative, in others 

 in quantitative differences, according to the rest of the 

 genotype. 



A comparisori of the influence of one and the same gene in 

 groups of different constitution shows, that in one case ] res- 

 ence of one gene in some and absence from other members of 

 the group may result in a distinctly discontinuous variation, in 

 another case in a continuous variation. 



