14 HEREDITY IN RABBITS, RATS, AND MICE. 



grade to grade 11. (Text-fig. 5, Fs, DxT.) Indications of segregation 

 are plainly seen in Fo, but it is impossible to be sure of the number of 

 factors in^•olved because dark and tan are so similar in grade. But we 

 know that dark is an allelomorph of white. If it is also an allelomorph 

 of tan, then Fi individuals should produce dark gametes and tan game- 

 tes in equal numbers. These, in a cross of Fi with white, uniting with 

 white should produce two kinds of heterozygotes: (o) white-dark hete- 

 rozygotes like those of text-figure 1, Fi, and (6) white-tan heterozygotes 

 like those of text-figure 4. The former range from grade 5 to grade 11, 

 the latter from grade 6 to grade 9. The ranges are similar and the 

 mode of each group is on grade 7. Mutual modification of dark and 

 tan would tend to extend the range of segregates in both directions, 

 as observed in F2 (text-figure 5) . 



The observed back-cross generation (table 29 and text-figure 5, 

 Fi X W) is distinctly bimodal, as the hypothesis just formulated 

 (that dark and tan are allelomorphs) would demand. Indeed, the 

 evidence of 1 : 1 segregation are clearer than we should expect, but 

 the modes of the expected groups are farther apart than we should have 

 expected. (See text-figure 5.) The lower mode is at 4, not at 7 as in 

 the Fi produced by dark crossed with white; and the upper mode is 

 at 12 to 14, not at 7, as in the Fi of tan crossed with white. It appears, 

 therefore, that while dark segregates from tan in the gametes formed 

 by Fi individuals, each segregates in an altered form, the dark ha\'ing 

 become darker and the tan lighter as a result of their association in 

 the heterozygous Fi. This is indicated both by the bimodal condition 

 and increased range shown by the cross of Fi with white, as just described, 

 and also by the wide range of the F2 from the cross of dark with tan. 



.\n alternative hypothesis, which has been gi^•en careful considera- 

 tion and which in fact the cross of Fi with white was especially designed 

 to test, is this: that dark and tan are due to independent factors and 

 that the two together constitute white. With this hypothesis the 

 following facts harmonize : Dark crossed with tan produces in Fi and 

 Fo individuals which are self-colored (i.e., which have no white-spotting) 

 as well as others wliich are whiter than either the dark or the tan 

 race. The former may be interpreted as those which lack (or are 

 heterozygous for) both dark and tan; the latter as those which have 

 both dark and tan. But it should be observed that none of the 275 F2 

 young which have been produced extend into the range of the uncrossed 

 white race, where about 25 per cent of them should lie if the hypothesis 

 is correct. Compare text-figures 1 and 5. 



Another way of testing the hypothesis that dark and tan are due 

 to independent MendeUan factors is the cross of Fi with white. If the 

 hypothesis is correct, Fi individuals should produce four kinds of 

 gametes, viz, those which will transmit (a) both dark and tan, (b) 

 dark alone, (c) tan alone, and (d) neither dark nor tan. By the hypoth- 



