INHERITANCE OF WHITE-SPOTTING IN RABBITS. 17 



Again, in the crosses of dark w-ith white (text-figure 1), we see the 

 mutual modification of the contrasted conditions taking place. Un- 

 crossed dark has its mode on grade 3, and uncrossed white on grade 17. 

 Extracted white as seen in F2 has its mode on grade 15, but in the back- 

 cross with pure white it has its mode on the intermediate grade, 16. 



As regards the modification of dark, it will be observed that the 

 original Fi individuals are of lower average grade than the lower group 

 of individuals produced by the back-cross with white. The mode of 

 the former also is on grade 7, that of the latter is on grade S. The one 

 consists of pure dark united with pure white, the other of extracted 

 dark united with pure white. The extracted dark has evidently been 

 whitened, exactly as extracted white has been darkened. 



Uncrossed tan and uncrossed white have their modes on 3 and 17, 

 respectively (text-figure 4). The mode of Fi is on grade 7, but the 

 mode of extracted tan united with white is on grade 10, as seen in the 

 back-cross of Fi with white. This shows that extracted tan has been 

 whitened as compared with uncrossed tan. That white has had its 

 grade lowered by the cross with tan is also shown in the back-cross 

 of Fi with white. Its mode lies at 15, not at 17 as in uncrossed white. 

 If, in the several cases considered, crossing tends to mutual modi- 

 fication and assimilation to each other of the contrasted conditions 

 brought together in the cross, why does crossing of dark with tan 

 extend rather than shorten the range of variation, producing in F2 

 individuals darker than either uncrossed race and others lighter than 

 either uncrossed race? The answer to this question is perhaps to be 

 found in the imperfection of our scale of grades. The scale is a linear 

 one, whereas the variation is not entirely linear; for dark Dutch and 

 tan Dutch have the same modal grade, 3, yet are different in somatic 

 character, as has already been stated. Tan Dutch has a white head and 

 narrow collar, dark Dutch has a dark head and wider collar. In Fi, if 

 pigment simply dominates whiteness, we may expect to get a dark head 

 and a narrow collar simultaneously, i. e., a condition with less white 

 than either parent possessed, which in general is the result observed. 

 Since uncrossed dark Dutch varies down to grade 1 and uncrossed 

 tan down to grade 2, an extension of tlie dark areas due to crossing 

 naturally carries the pigmentation in Fi down to grade (self) in a 

 certain percentage of cases (1 in 6 observed). See text-figure 5. 

 In Fj the percentage of selfs is still larger, being about 1 in 3. There 

 are also found in Fo whiter individuals than either uncrossed race 

 contained. How these have arisen is indicated by the back-crosses of 

 Fi with dark and with tan respectively. The back-cross with dark 

 (text-figure 5, Fi X D) produces a monomodal group closely resembling 

 the Fi group, but with slightly higher range. This shows that the 

 potent factor in lowering the Fi range was the dark gamete, since it 

 is the only common factor entering into both crosses. On the other 



