Apr. i, 1921 
Effect of Time of Irrigation on Kernel Development 
35 
be considerably lower than those of kernels from plot 3 or 4. A part of 
this may be due to soil conditions, but a part is due also to the greater 
culm formation. The irrigation of July 14 on plot 2 stimulated stooling 
to such an extent that more spikes were produced in this plot than in 
any of those irrigated later. As it received no further irrigation, the 
competition was more intense and the plants suffered from drouth toward 
maturity. As maturity approached, the diameters of kernels from plots 
6 and 7 were greater than those of kernels from plot 8, showing benefit 
from late irrigation. 
The dorsoventral diameter (fig. 3) does not reach its maximum until 
just before maturity. There was little difference between these measure¬ 
ments of kernels from the different plots until July 27. After this date 
there is an evident relation between the dorsoventral diameter and the 
date of the application of water, excepting in plot 2, which was some¬ 
what abnormal. The kernels from plots 3 and 4 are almost equal to 
those of plot 1. These kernels, it will be remembered, are from selected 
I 
Fig. 3.—Graph showing dorsoventral diameters of barley kernels during growth, from plots variously 
irrigated. 
spikes, the total number of spikes being much less in those plots. The 
plants in plots 6 and 7 were able to use the water applied as late as July 
26 and 29. They showed an increase in the dorsoventral diameter of 
the kernels as compared with those from plot 8, and the maturation 
was delayed by the application. 
The average weights per kernel of all intermediate plots fell between 
those of plot 1 and plot 8. The data on weights present one surprising 
feature. The wet weight per kernel is an earlier index of fingl size of 
kernel thanisthedry weight. In figures 4 to 10 it appears that the moisture 
of the irrigation of June 23 was sufficient to maintain a mavirnnm growth 
on plot 8 until July 27. After that date, the wet weight of kernels from 
plot 8 was less than that of those from plots 1 to 4. In figures 11 to 17 it 
will be seen that this is not the case with the dry matter. The weight of 
dry matter in kernels from plot 8 is equal to that of those from plots 
2, 3, and 4 until July 30. The cause of this is not evident. It would 
appear that the size of kernel is determined before the rate of deposit 
of material is checked. In figure 18, the water content of the kernels 
from plot 8 is seen to have reached the maximum by July 27. This is 
not true in any other plot. This means that whatever change takes 
