258 <-^*^ I'Tl"- INlIKRri'ANCK OK TIFK FLOWKRINi; TIMK IN TEAS AND RICE 
must be 6 : 10. The ratio of tlie actual numbers in tlie early flowering group 
in the 
late Hou'cring 
group is a 
s follows : — 
Eai 
ly group 
Late group 
Table 
1 1 2 
: 259 
Table 
8 A, 
: 72 
Table 
8 A.^ 
26 
: 43 
Treble 
8 
13 
: 29 
Table 
8 B_; 
32 
64 
Table 
8 C 
50 
: 90 
Sum 
258 
: 557 
Ratio 
6 
: 1 2.9 
The actual ratio exceeded the expected one, but this discrepancy is not 
so great as to be inadmissible. It is quite probable that such a discrepancy 
should occur, because, in the actual condition of plant raising, the hastening 
in flowering is rather rare, but its retardation is of very common occurrence, 
caused by environmental influences, both physiological and mechanical. 
According to the present hypothesis, there must be, in the F3 raisings, 
four kinds of constant families, aabb, aaBB, AAbb; and AABB- In Tables 
9, 10 and II (Pis. XII — XVI), we were able to group constant families into 
four types, — the early constant, the early intermediate constant, the late inter- 
mediate constant, and the late constant (see page 247), cxcej^t a comparatively 
few families, which we have designated as the pseudo-early or jxseudo-late 
constant, and also two abnormal families, No. 8 and No. 9 in Table i i. The 
early constants might correspond to aabb, the early intermediate constants to 
aaBB, the late intermediate constant to AAbb, and the late constant to AA- 
BB: and the pseudo-early and the pseudo-late constants might be considered as 
contaminated early and late constants respectively. We shall discuss the two 
abnormal families, No. 8 and No. 9 in Table i i later on. 
In the Fg variable families descended from the heterozygous plants, 
which belong to the early flowering group, we should expect two different 
variatic>n types, the tj^pes of the progen}^ of aabB and of that of aAbb- The 
