﻿M and tester Memoirs, Vol. xlix. (1905), No. <>. 5 



we only calculate as far back as the great-grandparental 

 generation. 



We have seen, therefore, that hybrids were mated with 

 hybrids, and that they were also mated with albinos. In 

 this way two kinds of hybrids were produced which could 

 not be distinguished from one another by their outward 

 appearance but differed in the amount of their albino 

 ancestry ; for while the one kind, which we may call HH, 

 resulted from the union of two hybrids, the other, HA, 

 was the offspring of a hybrid and an albino. I took 

 those individuals to be hybrids which resembled the first 

 crosses (F a ) in coat and eye colour, t.e, } in the possession 

 of a coloured coat and pigmented eye. 



In mating hybrids of this generation (F 2 ) I did not 

 previously look up their ancestry in books containing 

 their genealogical record, so that mice of the categories HH 

 and HA were mated at random : in this way three kinds of 

 •crosses were made HH x HH, HH x HA and HA x HA. 

 In each type of union a hybrid was mated with a hybrid, 

 as I believed at the time : and as on Mendelian theory 

 there is no difference between the gametic constitution of 

 DR produced by DR x DR and DR with parentage 

 DR x RR, I argued* that, if that theory were true, each 

 type of union would produce a fraternity, half of which 

 would be composed of hybrids, a quarter of which would 

 b>e composed of pink-eyed mice with coloured coats, 

 while the remaining quarter would consist of albinos ; on 

 the other hand, it was evident that on Galton's theory of 

 heredity the proportion would not only not be the same 

 in each of the three cases but would differ in direct pro- 

 portion to the amount of albino blood in the parents of 

 the population. The subjoined table gives together with 

 the actual result, the proportions of albinos as predicted 



* Biom., Vol. III., pp. 23 — 25. 



