CONTINUOUS SELECTION FOR EAR TYPE IN CORN 3 
^vas extremely favorable for corn production, and excellent yields 
were obtained, C. I. No. 119 averaging 87.5 bushels of air-dry shelled 
corn per acre. 
The method of comparison was an extension of the plan that has 
been described ^ for having the check plats constitute replications of 
the different sorts to be compared. The 21 kinds of seed, exclusive 
of C. I. No. 119, were divided into three groups, and C. I. No. 119 
was added to each group. The eight lots in a group were compared 
in a single section. The rows were 10 hills long. A single row of each 
of the eight kinds alternating with single-row check plats constituted 
a unit which was replicated eight times. Each of the eight kinds 
of seed was used as the check in one replication. Thus, in section 1, 
C. I. No. 119 was the check in the first replication, strains Nos. 1, 2, 
3, 4, 5, and 6, respectively, were the checks in the second to seventh 
replications, and the cross of strain No. 4 X strain No. 5 was the 
check in the eighth replication. This provided 17 plats of each 
-17 plats- 
-ITplats- 
\- 
t I 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is A. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is B. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is E. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is F. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is C. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is G. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is D. 
One single-row plat of each of eight strains or 
crosses, A to H, alternating with single-row 
check plats. 
Check is H. 
Fig. 1.— Diagram of the planting arrangement of a section in the comparisons at Arlington Experiment 
Farm in 1923. Eight lots were compared in each of three such sections. Extra rows of C. I. No. 119 
are indicated by x 
strain or cross. Eight extra plats of C. I. No. 119 were grown in 
each section for additional checks. As these were comparable with 
the other plats in every way, the data on C. I. No. 119 are based on 
25 replications in each section. A diagram of the arrangement in 
a section is shown in Figure 1. 
EXPERIMENTAL DATA 
No data were obtained on the length of ears or the indentation of 
kernels, although these characters were considered in selecting seed. 
In general it may be said that strains Nos. 1 and 2 produced larger 
percentages of very rough ears and that strains Nos. 3, 4, 5, and 6 
produced larger percentages of smoother ears in the later genera- 
tions. Moreover, it became increasingly difficult to obtain seed ears 
of strain No. 1 that were up to the standard length of 8 inches, 
whereas this difficulty did not occur in the other strains. A repre- 
2 Richey, F. D. Adjusting yields to their regression on a moving average as a means of 
correcting for soil heterogeneity. In Jour. Agr. Research, vol. 27, pp. 79-90, illus. 1924. 
