130 
Journal of Agricultural Research 
Vol. XI, No. 4 
With respect to the number of rows of kernels on ears, no significant 
difference is shown to exist between the upper and the lower ears. 
While the upper-ear plots were continued for only four years, it seems 
that it may be of value to compare for these four years the upper ears 
of the upper-ear plots with the upper ears of the lower-ear plots and the 
lower ears of the upper-ear plots with the lower ears of the lower-ear 
plots. From such a comparison (see fig. 9 to 12) we may assert that in 
mean length of ears, circumference of ears, and weight of ears, the upper 
ears of the upper-ear plot are larger in each case than the corresponding 
upper ears of the lower-ear plot. The lower ears of the upper-ear plots 
exceed in mean length and weight the lower ears of the lower-ear plots, 
but in circumference no significant difference exists. 
DIFFERENCES IN VARIABILITY 
[Table III and fig. 13 to 16] 
The standard deviations of length of ears have in general been only 
slightly different for the upper and lower ears of the lower-ear plots. 
In circumference of ears for the lower-ear plots it seems that the dif¬ 
ferences in standard deviation are insufficient to persist from season to 
season. In the weight of ears the standard deviations of the upper ears 
of the lower-ear plots have generally been significantly greater than those 
of the lower ears of the lower-ear plots. The year 1915 gives an excep¬ 
tion to this rule. The standard deviations of the upper ears on the up¬ 
per-ear plots for length of ears, weight of ears, and circumference of ears 
are in general smaller than the corresponding standard deviations of 
the lower ears on these plots. 
The coefficients of variability for length and weight of the lower ears 
of the lower-ear plots are decidedly larger than the coefficients of varia¬ 
bility for these characters in the upper ears of the lower-ear plots. Fur¬ 
thermore the coefficients of variability of length, circumference, and 
weight of the upper ears are larger for the lower-ear plots than for the 
corresponding upper-ear plots. 
COMPARISON OP SINGLE EARS ON UPPER- AND LOWER-EAR PLOTS WITH 
THE UPPER AND LOWER EARS 
[Table III and fig. 9 to 16] 
DIFFERENCES OF MEANS 
[Fig. 9 to 12] 
The means of single ears for weight, length, and circumference are 
larger in each case than the corresponding means of the lower ears. 
In length of ears the mean values of the upper ears are larger than those 
of the corresponding single ears in io out of n cases. To judge by 
probable errors, no more than half of these differences are significant. 
