90 CROSSING. 



delian segregation of size-influencing genes even in material 

 where their presence cannot be demonstrated at once. 



With this object in view we started an extensive series of 

 breeding-experiments on size in mice. We obtained very good 

 material in the Orient, very minute domestic mice which 

 proved extremely pure in respect to shape and weight. For the 

 other species we were fortunate to obtain a strain of large white 

 mice which had been purely bred by Dr. T. B. Robertson of the 

 University of California for his experiments on growth. These 

 mice were very pure in respect to weight, and their standard 

 weight had been tabulated for both sexes and all ages from one 

 week up to time of natural death. 



This made it possible to use these figures as a standard. In- 

 stead of expressing the weight of any of our animals in grams, 

 which would have restricted the use of this weight to compar- 

 ing it to that of animals of its own age and sex, we expressed 

 the weight of any given animal in procentages of the weight of 

 the standard animals of that sex and age. This made it possible 

 to compare the weights of brothers and sisters directly, and 

 also that of animals of approximately the same age. 



This is not the place to give more than brief outlines of our 

 problems and our results. Let it suffice to say that our problem 

 was "threefold. In order to show the existence of one or more 

 weight-influencing genes, we had to control as far as possible 

 the influences of the environmental factors, of age and sex, and 

 of heterozygosis in respect to other weight influencing genes. 



By keeping all the animals in the same kinds of cages, in uni- 

 form temperature and on the same diet, all through the experi- 

 ment we tried to eliminate as far as possible the effects of a dif- 

 ferent environment on weight. Differences due to sex and age 

 were equalized in the way indicated above. Rests heterozy- 

 gosis. 



The F! animals are necessarily heterozygous for all the genes 

 which are not common property of both parent species. Segre- 

 gation of a weight-influencing gene over the gametes of these 

 animals can be expected only if the influence of this gene is rel- 



