382 LOCK : STUDIES IN PLANT BREEDING 
It is perhaps noteworthy that a total so nearly approaching 
the expected should be derived from such small individual 
figures. 
144 of these seeds were sown on 3rd May. 1903. only those from 
the weakest of the above plants being excluded. 98 plants yielded 
fruit, but the majority gave only a poor crop. Most were very tall, 
but this effect was partly due to overcrowding. In general 
appearance all were much alike—the pale green foliage of 
Telephone being still visible—and the stipules were broad. The 
plants were badly mildewed and damaged by alternate rain and 
sunshine. 
The width of the pods appeared to be tolerably constant on each 
plant, but between different plants the variation was considerable. 
The width of a typical pod was measured in the case of each of 78 
plants—all that had pods typically developed—with the following 
result :— 
Width .. 19 mm. 18 17 16 15 14 13 
Plants .. 2 . . 9 . . 22 . . 17 . . 9 . . 14 . . 5 
If we draw an arbitrary line between the widths of 14 and 15 mm., 
as we are to some extent justified in doing, because (1) the width 
of the pods of No. 1 was from 13-14 mm., and (2) the figures 
show something of a natural demarcation at that point, we find 
the ratio of wide, intermediate, and narrow to be 59 : 19 or 
3.1 : 1. 
2. Telephone x Native pea No. 1.—The cotyledons fell into the 
same four groups as the above. In three cases it was necessary to 
remove the seed coat before the colour of the cotyledons could be 
ascertained with certainty : all three proved to be fully yellow, 
and this had been partly obscured by a greenish tinge in the testa. 
In these and in all other cases the two classes of colour were quite 
distinct, although some of the green seeds showed slight yellow 
patches. Smooth and wrinkled seeds were more difficult to distin¬ 
guish bv the eye, the former being all more or less pitted. The 
majority of the wrinkleds were thoroughly so. 
