28o ON THE INHERITANCE OF THE EIXAVERING TIME IN PEAS AND RICE 
1 1 are shown, we see again uniform distribution of whites in variable families 
(page 252). Family No. 21 in Table 1 1 appears to conflict with the present 
hypothesis, because the distribution of whites in it is quite uneven, while from 
its parent class and variation type we have to assume it to be progeny of aabB. 
On the other hand, we may also assume it to be the progeny of aAbb, deduc- 
ing from the example of the similar typed family No. 22 which are shown to 
be the segregates of aAbb by raising progenies (Table 14). 
Among those F3 families which are to be taken as the segregates of 
AAbB, the majority are homozygous red. Only one family, No. 81 of Table 
I I, which is variable in colour, has uniform distribution of whites. No. 79 
of Table 1 1 seems to be the segregates AAbBj but by F^ raising it was 
proved to be those of aABB (Table 16 0). 
(2) According to the proposed hypothesis, F^ plants which were homo- 
zygous in the factor 3 or b but heterozygous in the factors A £i"d a and also 
in colour, must produce such F3 families which have the following distribu- 
tion of different coloured individuals. 
Early group (earlier constants) Late group (variables & later constants) 
white red white red 
49 15 '5 177 
And when we raise progenies of these Fg families, there must appear the 
same distribution of different coloured individuals in variable families. 
As the variation types of the supposed aAbb segregates in F3 raising in 
1913 (Table 1 1) are irregular and do not allow us to distinguish two groups, 
the early flowering and the late flowering, by taking a ciefinite minimum fre- 
quency class, we shall take only Tables 9 and lO, and examine the distribution 
of different coloured individuals in those families which suggest the segre- 
gates of aAbb- 
