202 ON THE INIIERITAN'CE OF THE FLOWERING TIME IN PEAS AND RICE 
Tal)le 15 can be nothing but those of the segregation of aABB, according to 
the proposed hypothesis. Comparing the variation types of both earlier 
(woiild-be aaBB) and later (would-be AABB) constants of these raisings, 
however, we see some cHfference among them. From the variation types of 
J, K and L in Table 1 5, we can not take the F.^ grand-parents of these rais- 
ings as other than those with the zygotic formula aaBB? but within the 
variation types of the above three raisings we find complicated phenomena. 
The variation means of the majority of families in J are smaller than those of 
all families in Kj and the variation types in L, in general, remind us of those 
types which lluctuate between two different early intermediate constants in 
J and K- 
Now we are in a position to decide which is reasonable, to maintain the 
proposed two-factor hypothesis and to accept the occurrence of gametic con- 
tamination, or to assume gametic purity and propose the other hypothesis for 
the interpretation. Frecjuent occurrence of the distinct variation types of the 
monohybrid segregation in F_^ raisings disproved positively that the inher- 
itance of flowering time is non-Mendelian. If the inheritance be Mendelian, 
more than one factor must be concerned. To propose a new interpretation 
other than the two-factor hypothesis, it is necessary to suppose the presence 
of more than two-factors. To assume the presence of more than three factors, 
however, conflicts with the actual results. The actual numbers of the early 
constant families are too great, as one would expect in case more than three 
factors were concerned. In Table 3, two fau^iilies, Nos. 3 und 4, which had 
descended from the F^ plants, flowered in 45 days, had almost equal variation 
means and ranges as I. P. (early constant), and two out of the three families 
(i, 2 and II) which had descended from the F^ plants, flowered in 46 days 
varied like I. P. It might be concluded that all F.^ individuals which had 
flowered in 45 days, and some of those flowered in 46, were early constant. 
In Table 2, there are 8 individuals flowered in 45, and 7 flowered in 46, and 
it might be assumed that in the F2 raising there were present more than 10 
-early constant individuals among 372. Such a large number would not be 
