86 PROCEEDINGS OF THE AMERICAN ACADEMY. 



following table shows by litters the numbers and sorts of young obtained 

 from all the crosses in which the two parents were hybrids between the 

 black-white stock and albinos. 



This table shows that out of a total of 214 mice obtained there are 

 actually 55 albinos, where 53.5 are expected. The experimental result 

 is thus closely in accord with the theoretical. 



As in the case of the heterozjgotes of house-mouse and albino 

 parentage, the relative numbers of pigmented and albino young in the 

 individual litters are rarely those demanded by the Mendelian formula. 

 In fact, the precise theoretical ratio obtains in but 7 out of the 56 

 litters (Nos. 2, 12, 20, 23, 25, 26, 40), where in each case there 

 were three pigmented and one albino young. In two litters (Nos. 25, 

 26) from (^ gr. 4 X 9 gi'. 64 this precise proportion was maintained, 

 but a third lot of six 3'oung (No. 27) by the same parents consisted 

 wholly of pigmented mice. Cases in which all the young at a birth 

 are albinos are practically wanting in these experiments, being limited 

 to two litters (Nos. 3, 5), where but a single albino represented the 

 entire progeny. Among the 56 litters, however, there is occasionally 

 an excess of albinos over pigmented young, the most notable case, 

 perhaps, being one (No. 6) in which the expected ratio was precisely 

 reversed. There were in this litter 3 albinos to 1 pigmented indi- 

 vidual. In all but 9 of the 56 lots of young, there was a greater or less 

 preponderance of pigmented mice, which is a result of the nature 

 expected. Two of these 9 litters (Nos. 11, 30) showed an equality of the 

 two classes, the total number of young in each of the two cases being four. 

 In one large litter (No. 28) of 9 young, four were albinos ; in another 

 (No. 49) of 8 young, but one was an albino. In the latter case, however, 

 the excess of pigmented mice is but one. It is obvious that in cases where 

 the litter consists of an odd number of individuals, it is impossible in that 

 particular case to obtain a 3 : 1 ratio of the two classes. Hence in a lot 

 of 5 young the nearest approach to the Mendelian ratio is 4 pigmented 

 to 1 albino. In this case there are but .25 too few albino individuals. 

 If two of the five young are albinos, the excess is still represented by 

 less than an integer. Hence the irregularity in distribution of the two 

 characters is a thing to be expected in small litters. In a long series of 

 litters, however, the discrepancies noticed in individual lots tend to coun- 

 terbalance one another, so that the final result shows a close approxima- 

 tion to the 3 : 1 ratio. This is seen also in cases where a number of young 

 are obtained from the same fair. For example, $ bl. 342 X $ bl. 344 

 produced two litters, the one (No. 38) of 5 pigmented to 1 albiuOj 



