YUZO HOSHINO 
I. EXPERIMENTS WITH PEAS. 
Experimental Methods. 
The Experiments were carried on in the College Vegetable Garden from 
1907 to 191 1, and in the nursery ground of the College Orchard from 19 12 
to 1914. As peas do not flourish under continuous cropping, care was taken 
to change the raising place every year, selecting even and homogeneous plots. 
The method of raising the offsprings was as follows. At first, we select- 
ed a certain number of F-^ plants, and took 30 seeds from each of these 
selected plants and sowed them according to their origin on different rows. 
Thus we got a certain number of F.^ families, each of which consisted of 30 
individuals. In raising the F3 progenies, we selected certain F^ families 
(rows), and took 30 seeds from every plant of these selected families, sow- 
ing them on different rows according to their different origins, just as in 
the case of the F.^ raising. F^^ raisings were also made according to the same 
principles. 
Thus our method was to raise 30 individuals in one family and to record 
their flowering time, but, in reality, there were some seeds which failed to 
sprout, and some individuals which, on account of retarding of flowering 
time caused by some injuries, such as breaking of stem, insect injury etc., 
had to be discarded later from records. And moreover, there occurred some 
few individuals which flowered normally but did not produce a sufficient num- 
ber of seeds for raising 30 offsprings. Such were the causes of the insufficient 
number of the individuals in one family that we noticed everywhere through- 
out the experiments. We did not raise the progenies of those plants which 
produced too small a number of seeds. 
The rows, on which the individual plants were grown 6 inches apart, 
were 15 feet long and 2 feet apart. When the plants had grown about 10 
inches, we gave to every plant a bamboo cane for support, binding the stem 
to it lightly. Afterwards we repeated this binding two or three times to 
