USE OF WATER BY SPPJXG WHEAT OX GREAT PLAIXS. 7 
introduction of irrelevant data in a coiTelation of the rate of the 
use of water with the yield. 
There remained 53 years in which the daily rate of the use of 
water could be deteiTained under the conditions imposed. These are 
distributed among 14 stations located in the Great Plains from 
Huntley. Mont., and TTilliston. X. Dak., on the north, to Amarillo, 
Tex., on the south. Two of these stations — Ardmore. S. Dak., and 
Garden City. Kans. — furnished data for only a single year each, 
while at Edgeley, X. Dak., a measure of the rate was determined for 
10 years. 
The rate was determined separately for two cultural methods at 
each station. The one designated plat A is continuously spring 
plowed and cropped to wheat. The one designated plat C or D is 
alternately summer fallow and wheat. This method occupies two 
plats, so there is available for study each year a crop of wheat grow- 
ing on land bare summer tilled the previous year. 
In the average of conditions this method contains more water in 
the soil at the time the crop commences rapid growth, supports a 
heavier growth of vegetation, and produces a larger crop than the 
plat continuously cropped, but exceptions are to be noted when one 
or more of these conditions are reversed. 
The results of this study are presented in Table 1. This gives the 
station, the year, the dates of the soil-moisture determinations mark- 
ing the beginning and end of the period, the average daily precipita- 
tion during the period, the average daily evaporation from a free 
water surface during the period, the daily rate of water use and 
yield of plat A, and the dailv rate of water use and vield of plat 
C or D. ' . 
The daily rate of water use has been charted with calendar dates 
as abscissa and total use as ordinates. The slope of the curve then 
indicates the rate of use. This chart is not presented, because a con- 
siderable part of its value for study lies in the identification of each 
line with the year and station it represents. It is impracticable to 
attempt this on the scale to which it necessarily would be reduced in 
publication. Such charting shows that the rate of use is not deter- 
mined by the dates within which or the part of the period of rapid 
growth during which the rate was determined. This phase of the 
subject is more fully developed in the study of the rate of the use 
of water during the season. In this it is shown that after beginning 
rapid growth, about the time tillering is completed, the rate of use 
remains fairly constant until harvest if not interrupted by failure of 
the water supply or by some destructive or inhibiting agency. 
These determinations afford opportunity for a study of the rela- 
tive importance of water from the soil and of precipitation in deter- 
mining the rate of use and for the examination of the reasons for 
combining the two to obtain this rate. 
The material has been studied in various ways. Charting the two 
quantities and the total which they make up shows very clearly that 
they are complementary. Under given conditions a given crop re- 
quires a certain quantity of water. If this is supplied by precipita- 
tion the quantity of water in the soil will not be reduced, and if 
more than the required quantity is supplied it will be increased: 
but if the rainfall is not sufficient to meet the demand the available 
