274 INHERITANCE OK THE FLOWERING TIME IN PEAS AND RICE 
a parallelism botwccii them to some extent. The nowering time (shooting time 
in rice) of the first generation of a hybrid was not just intermediate between 
parent varieties. It inclined towards one of the parents, in peas towards the 
late parent and in rice towards the early parent. In F.^, the variation range 
extended within the combined ranges of both parents, having the minimum 
frequency class in the middle. When all individuals were arranged into 
two groups, the early flowering and the late flowering, using the minimum 
frequency class as a demarkation line, the group towards which F-^ inclined 
contained a greater number of individuals than the other group, — in the rice 
the early group and in peas the late group. In Fg raisings of rice hybrids, 
there appeared few families which seemed to be constant, though the distribu- 
tion of these apparently constant families was not similar to that in the peas, 
and in F^ raisings we had one case of variation types closely suggesting those 
of monohybri dsegregation (B Table 21). 
From the above cited facts, we may safely assume that the inheritance of 
the shooting character in rice is not different from that of the flowering charac- 
ter in peas in its underlying principle and is interpretable by the multiple- 
factor hypothesis. But the two-factor hypothesis which was proposed for 
the interpretation of the inheritance of the flowering time in peas is not applic- 
able in the case of rice, because of the following two facts. 
1. The actual number of F^ individuals which seem to be in the same 
zygotic constitution as the original parents is much smaller than would be 
expected on the two-factor hypothesis. From actual results in Fg raising, 
we might assume those F^ individuals which shot forth within the variation 
range of the late parent as the late constant, and those which shot forth during 
the two days, 88 and 89, as the early constant. The actual number of the 
former is 4 and that of the latter 5, while the expected number of each of the 
early and late constant on the two-factor hy[)othesis is 340/16 = 21.2. 
2. According to the two-factor hypothesis, all variable families in F^ 
have to be divided into two categories, monohybrid segregates (descendants 
of aabB, aAbb, aABB or AAbB) and dihybrid segregates (descendants of 
