18 HEREDITY IN RABBITS, RATS, AND MICE. 



hand, the back-cross with tan (Fi X T) shows bimodal variation with 

 the modes at and 5 respectively. This indicates that the gametes 

 formed by Fi are really of two types, extracted dark and extracted 

 tan. The former uniting with pure tan produces a group like Fi but 

 apparently of even slightly lower grade, since the modal condition is 

 now 0, not 1. The extracted tan gametes uniting with pure tan pro- 

 duce a group like pure tan, but of apparently higher grade, since the 

 mode is now on 5, not 3 as in pure tan. Hence extracted tan is poten- 

 tially of higher grade than pure tan, a conclusion supported by Fj 

 from dark crossed with tan, for here we observe that extracted tan 

 meeting extracted tan produces zygotes of grade 5 to 8 or even higher, 

 whereas pure tan does not exceed grade 5. Yet, to return to the imper- 

 fection of our grading scale, these higher grades consist merely in 

 combining a wider collar with the same form of head markings as 

 are found in the tan series of grade 4 or 5. Hence it appears that tan 

 is regularly modified through its contact with dark in an Fi zygote in 

 the way of acquiring a under collar, whereas dark is modified bj' the 

 same agency in the way of acquiring a narrower collar. Yet there is 

 no indication that head marking and collar marking are due to dis- 

 tinct single genetic factors, but merely that they are qualitatively 

 different. This difference tends to disappear tlu-ough mutual influ- 

 ence in the heterozygous condition, but the difference disappears more 

 rapidly in collar markings than in head markings; hence the extended 

 range of grades in F2. Because their collars become more alike the 

 extracted darks rank lower in grade and the extracted tans rank 

 higher; for it will be recalled that the uncrossed darks, though having 

 wider collars than the tans, were graded lower on account of their 

 dark heads. The collar changes, then, are actuallj^ blending in this 

 cross, as in all the others studied, but give the appearance of segrega- 

 tion with differences emphasized, merely because of the inadequacy 

 of our linear grading scale to record simultaneously changes in head 

 and in collar markings when these occur with unequal rapiditJ^ 



It has already been shown that we ha^'e concIusi^'e e^'idence that 

 Fi, from the cross of tan with dark (text-figure 5), produces two types of 

 gametes, not four types, this evidence being (a) the bimodal variation 

 seen in the back-cross of Fi with tan and (6) the bimodal variation 

 seen in the cross of Fi with white. These results indicate that tan 

 and dark are to be regarded as allelomorphic but mutually modifying 

 conditions, as had already been found to be true for tan and white, 

 for dark and wliite, and for self and white. AYe have, then, a condition 

 of multiple allelomorphs in white-spotting patterns of Dutch-marked 

 rabbits, which includes the forms self, dark Dutch, tan Dutch, white 

 Dutch, and possibly many other types or conditions of white-spotting 

 wliich with sufficiently accurate observation might be distinguished 

 from each other. 



