92 
Journal oj Agricultural Research 
Vol. XXXI, No. 1 
INDIRECT EFFECT OF SELECTION ON INDENTATION AND ANGULARITY OF KERNELS 
After the Garrick lines had been selfed and selected for different 
numbers of kernel rows for five years they were examined to deter¬ 
mine whether there had been any indirect effect upon the indenta¬ 
tion and shape of the kernels. The results of this examination are 
shown in Table IX, the classifications for indentation and angularity 
being based upon the characteristic tendency of the ears in the 
individual ear rows. These studies were made at the Arlington 
Experiment Farm, Rosslyn, Va., and in cooperation with the Pee 
Dee Substation of the South Carolina Agricultural Experiment Sta¬ 
tion, Florence, S. C. 
Table IX .—Data on the effect of selecting for different numbers of kernel rows 
upon the indentation and angularity of kernels of self-fertilized lines of Gar¬ 
rick grown at Florence, S. C., a and Rosslyn, Va., in 1921 
Kernel 
rows on 
parent 
ears 
Ear 
rows 
grown 
Percentages of ear rows producing ears with 
mode for kernel indentation stated 
Percentages of ear rows 
producing ears with 
mode for angularity of 
kernels stated 
None 
Trace 
on only 
a few 
kernels 
Shallow 
and 
smooth 
Mid¬ 
depth 
and 
wrinkled 
Deep 
and 
rough 
Round¬ 
ed 
Inter¬ 
mediate 
Angular 
Per cent 
Per cent 
Per cent 
Per cent 
Per cent 
Per cent 
Per cent 
Per cent 
8 
46 
26.1 
6.5 
45.7 
19.6 
2.2’ 
69.6 
28.3 
2.1 
10 
59 
16.9 
6.8 
52.5 
13.6 
'10.2 
59.3 
32.2 
8.5 
12 
59 
18.6 
8.5 
42.4 
16.9 
13.6 
55.9 
33.9 
10.2 
14 
65 
7.7 
1.5 
50.8 
20.0 
20.0 
46.2 
40.0 
13.8 
16 
52 
5.8 
3.8 
38.5 
26.9 
25.0 
42.3 
28.8 
28.8 
18 
34- 
8.8 
0.0 
35:3 
20.6 
35.3 
44.1 
20.6 
35.3 
• The data at Florence, S. C., were obtained in cooperation with the Pee Dee Substation of the South 
Carolina Agricultural Experiment Station. 
The classification for indentation was more or less difficult, but it 
is believed that it was fairly accurate in the extreme classes. The 
largest percentages of ear rows with no indentation were from the 
parents with 8, 10, and 12 kernel rows, and the largest percentages in 
the class with deep and rough indentation were from parents with 
14, 16, and 18 kernel rows. It is evident from this that selection for 
low numbers of rows on the ears tended to result in a different 
indentation from that obtained by selecting for higher numbers of 
kernel rows. Similarly, rounded kernels naturally followed selection 
for low numbers of kernel rows, and angular kernels followed selection 
for higher numbers of rows. 
RELATION OF SIZE OF EARS TO NUMBER OF ROWS OF KERNELS 
All of the ears produced in the Garrick selfed lines in 1921 were 
measured as to length and diameter. The total length of all the ears 
from each ear row was divided by the number of bearing plants in 
that row to obtain the total length of ear per plant. A yield index 
then was computed by multiplying this value by the square of the 
average diameter of the ears in the row. These data, and the average 
number of ears per plant produced by parent ears with different 
numbers of kernel rows, are shown in Table X. 
