88 Journal of Agricultural Research v 0 i. xxxi, No. 1 
the ears with 14 rows, the modal number in the Koger variety. 
Each of the three few-rowed groups (10, 12, and 14 rows), however, 
yielded more than any one of the three many-rowed groups. 
The data show a consistent relation between the number of rows 
on the seed ears and the yield produced from them, the ears with 
the lower number of rows yielding more as a class than those with 
the higher number of rows. 
The Garrick variety was included in both the South Carolina and 
Georgia varietal experiments and was highest yielding in each. “ From 
the standpoint of selection, this variety is the most perfect product 
of the prolific ideal that was studied. It was learned by interview 
that Mr. Garrick had grown this corn at Weston, S. C., for about 
30 years. His ideal had been the production of two or more ears 
per plant, and he had given no special attention to the character 
of the ears and kernels. Several thousand ears of this variety were 
studied on the home farm. The ears were large for the prolific class 
of corn, the kernels were intermediate between angular and rounded, 
and the characteristic numbers of rows of kernels on an ear were 
10 and 12. The indentation was distinct and smooth to slightly 
rough. 
Data on the relation of number of rows of kernels to yield in the 
Garrick variety are shown in Table IV. As different lots of ears 
were used in each experiment, eight distinct comparisons are shown. 
Ears with the fewest rows of kernels gave the highest yields in six 
comparisons, and ears with the most rows gave the lowest yields in 
seven comparisons. Ears with 12 rows of kernels yielded more than 
those with more than 12 rows, in all of the eight comparisons. 
Table IV .—Data on the relation of number of kernel rows , angularity of kernels , 
and weight of kernels in the parent ears to yield in the Garrick variety gro^n 
at different places and in different seasons 
Place and date of experi¬ 
ment, and character of 
kernel classes compared 
in 1914 and 1916 
Relative yields from ears 
with stated number of 
kernel rows (per cent) 
Perfect hills grown from ears 
with stated number of ker¬ 
nel rows 
Weight of ker¬ 
nels (grams) per 
1,000 from ea”s 
with the stated 
number of ker¬ 
nel rows 
8 
10 
12 
14 
16 
Av¬ 
er¬ 
age® 
8 
10 
12 
14 
16 
To¬ 
tal 
10 
12 
14 
Av. 
Lykesland, S. C., 1914: 
TTprnpls angular 
91 
84 
83 
87 
317 
291 
295 
903 
TTprnp.ls int.prmp.Hiatp. 
99 
101 
96 
99 
319 
295 
292 
906 
TTfirnfils rnilndp/l 
89 
93 
84 
88 
264 
300 
315 
879 
Total or average « 
94 
93 
87 
900 
886 
902 
Lykesland, S. C., 1916: 
Kernels large_ 
119 
100 
89 
102 
606 
619 
657 
1,882 
1,924 
1,988 
466 
418 
374 
419 
Kernels midsized 
101 
92 
86 
93 
644 
685 
595 
387 
378 
346 
370 
Kernels small_ 
94 
93 
86 
91 
715 
602 
671 
355 
334 
324 
338 
Total or average « 
104 
95 
87 
1,965 
1,906 
1,923 
403 
377 
348 
Lykesland, S. O., 1917 __ 
112 
109 
100 
1,085 
-.650 
579 
572 
"Rrooksville, Pla , 1917 
104 
”161 
100 
"94 
95 
630 
1,220 
1,144 
599 
1 
a Computed directly from the basic data. 
Both in the prolific and the nonprolific varieties that were studied 
there was a strong general tendency for the ears with fewer rows 
of kernels to yield more than those with larger numbers of rows. 
