30 HEREDITY IN EABBITS, RATS, AND MICE. 



could therefore arise only by crossing-over. A test mating of an Fi 

 individual with a double recessive would give zygotes as follows, it 

 being understood that the double recessive individual produces only 

 one type of gamete, viz, pr. Non-cross-over gametes, Pr and pR, 

 would give zygotes Pprr and ppRr, visibly red-eyed yellows and 

 pink-eyed yellows respectively. Cross-over gametes, PR and pr, 

 would give zygotes PpRr and pprr, visibly dark-eyed (black or gray) 

 and pink-eyed yellow respectively. The pink-eyed yellows could not 

 be distinguished readily from yellows arising from non-cross-over 

 gametes, but the dark-eyed young could be distinguished immediately 

 at birth from all other classes. Since theoretically thej'^ would con- 

 stitute half the total cross-overs, it is evident that the simplest way of 

 estimating with accuracy the proportion of cross-over gametes is to 

 double the number of dark-eyed young observed in new-born litters. 

 This number diiaded by the total number of young would give the 

 percentage of cross-over gametes. 



Following this procedure, we have reared from matin gs of Fi indi- 

 viduals with double recessives a total of 1,714 young, of which 174 

 were dark-eyed. Doubling the number 174, we have 348 as the prob- 

 able number of cross-over gametes among the 1,714 Fi gametes which 

 entered into the production of these young. This is a percentage of 

 20.3, a little higher than the calculation 18.5 of publication 241, based 

 on a study of a much smaller F2 population. The difference between 

 this figure, 20.3 and 50, the percentage expected where no linkage 

 occurs, would be a measure of the strength of the repulsion shown 

 between the genes for red-eyed yellow and for pink-eyed yellow 

 respectively, when they enter a cross in different gametes — that is, 

 each through a different parent, the condition realized in this cross. 



But, on the cliromosome theory, an attraction or "couphng" equal 

 in strength to this repulsion should occur between the same two 

 genes when they enter a cross together. Entering together, they 

 should tend to hold together, because they would lie in the same mem- 

 ber of a pair of chromosomes and so could pass out separately only 

 in consequence of a cross-over. This point, repeatedly verified in the 

 case of other organisms, was tested for rats by producing Fi individuals 

 through a cross of double recessive yellow (pprr) with a pure non- 

 yellow individual (PPRR). In reahty Fi zygotes of this same sort 

 were being produced in considerable numbers in the matings already 

 described to test the strength of repulsion. Such were the 174 dark- 

 eyed young already mentioned. Each resulted from the union of a 

 pr gamete with a PR gamete, the relationship which would give the 

 expected coupling. Accordingly many of these dark-eyed young were 

 used instead of Fi parents in the experiments to test the strength of 

 "couphng" between p and r. In these experiments, as in those to 

 test the strength of repulsion, Fi individuals were mated with double 



