YUZO HOSIIINO 245 
This fact coincides with what we have seen in Table 2. 
3. Keeble and Pellew(i9io) from their experiments, concluded that there 
might be a gametic coupling between stem characters and flowering 
character. In the present experiments, we could not find the existence 
of such a correlation. In general, the later the flowering, the thicker 
the stem and the longer the internode. For example, by exact measure- 
ment of the length of the internode of every individual of the raising 
in 1912, taking those which appeared to be representative in the length 
of the internodes in the individual plant, we got the following results : — 
Length of internode of M- P- plants 6.0 — 7.5 cm. 
„ „ „ ,, G- P. plants 9.5 — i[.ocm. 
of Fg plants in 
classes 46, 47 & 48. 
of F^ plants in j 
classes 49, 50 & 5 r. 
none shorter than 7.5 cm. 
jnone longer than 9.5 cm. 
majority 9.0 cm. 
[none shorter than 8.5 cm. 
(majority 9.0 cm. 
[none shorter than 9.0 cm. 
of F.J plants in the j 
late group 
majority 10.5 — 12. cm. 
Thus, the actual results in the present experiments suggested strongly 
the relation between the stem characters and flowering time to be only a 
physiological one. 
F3 Raisings, 
In 1912. 
The sowing was done on May 7, taking 30 seeds from each individual 
he following families : — 
I X 1-4 I 
[Crosses with the early flowering pure line of G- P- 
4 X XVI. 2 
