74 PROCEEDINGS OF THE AMERICAN ACADEMY. 



V. The Hkrkditv of Albinism. 

 A. Complete Albinism. 



1. House mice and albinos. In a preliminary paper (Castle and Allen, 

 : 03) Professor Castle and the writer have shown that complete albinism 

 in mice and rats is a recessive character in a Mendelian sense, and onr 

 observations corroborate Cuenot's (:02) results along similar lines. The 

 writer bred wild house mice, captured in stables and houses in Cambridge, 

 to a stock of albino mice obtained from a Boston dealer. Nothing could be 

 learned as to the ancestry of these albinos, but when bred inter se they 

 had produced only albinos. As a result of this cross of wild gray with 

 albino mice, only pigmented individuals were obtained. By breeding 

 albino males to female house mice, eight litters were secured, there being 

 a total of 28 young. The average number of young proiluced at a birth 

 was thus 3i, and the commonest number was 3 or 4. There was one 

 litter of 7 and one of G, while a third consisted of but 2 individuals. By 

 the reciprocal cross, in which the gray house mouse was the male, and the 

 albino the female parent, 38 young were produced in nine litters, an 

 average of 4;| to a litter. The sex was ascertained of 46 of these GG 

 mice, and there were fuu id to be 21 females and 2.5 males, an excess of 

 about 8 per cent of males. This is in accord with the work of Schultze 

 (:03), who usually obtained a slight excess of males among the offspring 

 of his white (albino) mice. 



Several of the cross-breds produced by the matings described were 

 paired. The Mendelian expectation in this case is, by formula (2), that 

 a quarter of the progeny will be albinos, a quarter will be pure dominants, 

 and the remaining half will be, like their parents, heterozygous, pig- 

 mented but containing the albino character in a recessive condition. 

 Quantitatively, this expectation was not precisely realized. For of the 

 09 mice produced, theoretically 17 should have been albinos, while 52 

 should have been pigmented. There were actually 25 albinos and 44 

 pigmented individuals, or 3G per cent albinos and 64 per cent pigmented. 

 The excess of albinos is thus 11 per cent, or there are about half again 

 as many albinos as expected. 



The distribution of the young by litters in this generation is shown in 

 Table A. 



Disregarding the first two litters in which the precise Mendelian ratio 

 of pigmented and albino young is realized, we find that six of the eight 

 remaining lots show an excess of albinos ranging from 0.25 to 4 per litter, 

 or an average of less than two (actually 1.5). Only two litters con- 



