USE OF WATER BY SPPJXG WHEAT OX GREAT PLAINS. 5 
duction of the soil-moisture chart for wheat plat C in the continuous- 
cropping series at Belle Fourche in 1916. Determinations on that 
plat were made to the depth of 4 feet that year. Its water content 
at the several dates of sampling is consequently shown by the points 
of four separate curves. Immediate reduction of these to a single 
curve by averaging the four would result in an error of a magnitude 
dependent upon the differences in the volume weight of the several 
imits of soil. Soils of the several stations also differ in volume 
weight, and consequently any given percentage of water content may 
not indicate the same actual quantities of water in different soils or 
soil units. Further, it is not possible directly to combine percentages 
of water in the soil with inches of water added to it by rainfall 
during any period. Consequently it has been necessary to convert 
the percentage of water in each foot section of soil to terms of inches 
of water. 
Prerequisite to such conversion is the determination of the volume 
weight of each soil unit under study. This has been determined by 
calculation from the weight of the cores taken by the soil tube. For 
this determination periods were selected when the soil tubes were 
known to be new and in good condition with a sharp cutting edge 
of a circle of the known diameter of 20 millimeters. As large a 
number of weights as possible, not less than 50 and i)referably more 
than 100, were averaged for each unit, and the weight per cubic foot 
was calculated from the weight of a known volume. "Weights of 
obviously defective cores were arbitrarily rejected. It is believed 
that the volume weights determined by this method are as accurate 
as those that might be determined b}' any other method based upon 
a smaller number of cores of larger volume. Tlie volume weights 
so determined and used as bases of conversion have been carefully 
considered in the light of all knowledge and information on the 
several soils and accepted as fair bases of comparison. 
This study is concerned with the amount of change in the water 
content of the soil during a period rather than with the actual 
water content. The process, therefore, has been to determine the 
percentage of change and then to convert this into inches of water, 
instead of reducing the entire water content to inches and then 
determining the change by subtraction. 
In determining the water loss from the soil for any period the 
first step is to subtract the water content of each foot section at the 
end of the period from the water content of the same unit at the 
beginning of the period. At this time the water content is expressed 
as a percentage of the dry weight of the soil. The next step is to 
convert these expressions to terms of inches of water by calculation 
from the determined weights of the soil of each unit. An increase 
in water content is recorded as a minus loss. The algebraic sum of 
the losses from all the units (foot sections) involved is the loss from 
the soil of the plat under determination. 
The total water loss or the total water used by the crop in any 
period is determined by adding the precipitation for the period to 
the water lost from the soil. It is the algebraic sum of the two 
quantities. The relative importance of the two quantities will be 
considered later, and the necessity of combining the two into a 
single quantity will be shown. 
