90 
Journal of Agricultural Research 
Vol. XXXI, No. 1 
fication for angularity also was made within groups of ears of the 
Garrick variety having the same number of rows of kernels. Data 
from these experiments are given in Tables IV and VII. 
As the data in Table IV are for different lots of ears and those in 
Table VII are for the same lots of ears grown in different places, four 
distinct comparisons are provided in the two tables. The ears with 
kernels of intermediate angularity produced more than either of the 
extreme classes in all four comparisons. In two of the comparisons 
the ears with rounded kernels produced significantly more than those 
with angular kernels and in the other two comparisons the differ¬ 
ences were negligible. 
SIZE OF KERNEL AND YIELD 
In connection with the experiments at Lykesland in 1916, the ears 
which had been classified for number of rows of kernels were grouped 
further into those with large, midsized, and small kernels. The data 
from this experiment are shown in Table IV. 
Table VII. — Data on the relation of weight and angularity of kernels on the 
parent ears to the yield of 12-rowed ears of the Garrick variety grown at Darlington 
and Lykesland } S. C., in 1915 
Shape of kernels 
Weight 
per 1,000 
kernels 
Grown at Darling¬ 
ton, S. C. 
Grown at Lykes* 
land, S. C. 
Perfect 
hills 
Relative 
yields 
Perfect 
hills 
Relative 
yields 
Angular____ _ 
Qm. 
382 
399 
397 
379 
968 
906 
3,852 
3,778 
Per cent 
88 
99 
102 
96 
765 
- 798 
3,330 
3,332 
Per cent 
88 
102 
102 
97 
Midangular_ 
Midrounded__ 
Rounded_ 
The yields tended to be directly proportional to the size of the 
kernels, and the yield of the ears having both the fewest rows and the 
largest kernels was very distinctly the highest. It will be noted in 
Table VII that the ears with midangular and midrounded kernels had 
heavier kernels than those of the extreme classes. The larger yields 
of these intermediate groups, therefore, may have been a result of 
their having heavier kernels, although this is not certain. 
INTERRELATION OF NUMBER OF KERNEL ROWS, ANGULARITY OF KERNEL, AND 
SIZE OF KERNEL 
It has been shown in connection with some of the preceding experi¬ 
ments that angularity of kernel is related to weight of kernel and that 
both of these characteristics are related to the number of rows of 
kernels on an ear. These interrelations probably are universal 
within a variety, and it is not possible to separate the effect of each 
individual factor in these experiments. Fortunately, however, this 
is not necessary from the present point of view, as it is necessary to 
know only that ears chosen within a variety to represent similar 
combinations of these characters react in a similar manner. Both 
in the varietal experiments and in the comparisons within varieties, 
the groups of ears with fewer rows of kernels and with less angular 
kernels have produced higher yields than those with more rows and 
more angular kernels 
