256 ON THE INHERITANCE OF THE FLO\\ERlNG TIME IN PEAS AND RICE 
U- The parent family of this raising had a variation type somewhat 
similar to that of the parent family of M in Table 16, but the variation types 
in the progenies are quite different in these two raisings. The variation 
types in XJ are quite similar to those in S and T> except in the non-pi-esence 
of the early constant. So we may assume that the F2 grand-parent of this 
raising was similar to the Fj hybrid plant in zygotic character, but in its F3 
offsprings did not produce any early constant individual. 
Interpretation of the Experimental Results. 
In the foregoing pages, we have recorded the results of our experimental 
work extending over eight years, and we now think that there has been ac- 
cumulated sufficient material to enable us to propose an interpretation of the 
inheritance of the flowering time in peas. 
In the first place, let us propose a hypothesis, and then let us apply it for 
the explanation of the results of our experiments, and see how far this hypo- 
thesis may work true. The hypothesis consists of two clauses. 
1. The inheritance of flowering time in the varieties of peas which were 
used in the present experiments, is governed by two Mendel ian allero- 
morphic pairs, each factor of which has a specific hereditary effect as 
follows : — 
A determines the proper flowering time of the late parent. 
B determines flowering a few days earlier than the proper time of the 
late parent and is hypostatic to A. 
a alleromorphic to A, determines the proper flowering time of the early 
parent and is epistatic to b- 
b alleromorphic to Bj determines the flowering a few days later than 
the proper flowering time of the early parent. 
2. Gametic contamination is caused by hybridization. 
Two-factor Hypothesis and Experimental Results. 
According to the proposed two-factor hypothesis, F^ families must con- 
