250 ON THE IMlIERtTANCE OF THE KI,( )\V ERIXC; TIME IN PEAS AND RICE 
The results are slunvn in Tables 12 and 13. 
Table 12, A. The parent family of this raisin;^ was one of the families 
designated as the early intermediate constant. Now it is proved actually 
that this latter designation was well chosen. The variation types of all fami- 
lies are almost equal, in correspondence with the assumjjtion that the I*",, 
grand-parent might be constant, and their variation means situated nearer to 
that of M. p. rather than in the intermediate position between M- P. and G- P- 
Table 12, B. The parent fatnil)- of this raising varied within the com- 
bined range of M- P- and the early intermediate constants ; and within the 
range, there was a minimum frequency class, which makes it possible to 
divide the individuals into two groups, the early and late flowering. Now, 
in the table, it will be seen that all families descended from the early flower- 
ed parents (No. i to No. 6) show the variation type of the constant families 
which are very near to that of M- P. (pseudo-early) '\ and among the fami- 
lies descended from the late flowered parents, there are two types of variation, 
one varying just like the early intermediate constant (from No. 2i to No. 25), 
the other varying within the combined range of the pseudo-early and early 
intermediate constants, exactly like the parent family. These facts suggest 
strongly that the F.^ grand-parent was monohybrid, being heterozygous in 
the determiners of the early (pseudo) flowering time and of the early inter- 
mediate flowering time. The ratios of the earl}- constant, the variable, and 
the early intermediate constant families are 6:14:5, quite close to the ex- 
pected ratios i : 2 : i. 
Table 13, C, Di E, F. All four parent families of the raisings shown 
here varied from the range of the early intermediate constant to that of the 
pseudo-late constant or of the late constant, and their variation types were 
all similar, having vacant classes near the middle of the range, which makes 
it possible to divide all individuals into two groups, the earlier and the later 
flowering. Now, in the actual raisings of the F ^^ families, we see that all 
l) One F,3 individual which flowered in 47 days produced too small a number of seeds and was 
discarded for the F4 raising. 
