2 BULLETIN 1004, U. S. DEPARTMENT OF AGRICULTURE. 
the work was started. The wide range of the curve of every rela- 
tion that is plotted may reveal a fundamental principle that could 
not have been suspected from a single group of the data and makes 
it possible to test severely hypotheses to which its study leads. 
The present study is the first that has been presented in which the 
data are used to develop a general problem other than one dealing 
primarily with methods of culture and the results as measured in 
terms of resulting yield. T\^iile it develops effectively the problem it 
undertakes, it is by no means exhaustive and suggests and invites 
study of more questions than it endeavors to dispose of. It is only 
suggestive of the possibilities of such data. 
While the importance of an adequate supply of water has long been 
recognized, the dependence of yield upon an uninterrupted supply 
has perhaps never before been so well established. Determination of 
the daily rate of the use of water and the dependence of yield upon 
the maintenance of this rate lays a foundation for more exact predic- 
tion of yields than are possible without such knowledge. In the daily 
rate Qf the use of water is to be found a reason for the difference in 
effectiveness of rainfall in different sections of the Great Plains. 
The writers have done enough checking of results outside the Great 
Plains to be satisfied that in the same factor is to be found a basis 
of comparison and explanation of the results in other sections and 
under other types of rainfall. 
E. C. Chilcott, 
Agriculturist in Charge. 
STATEMENT OF THE PROBLEM. 
Dry farming is practiced only in regions where the water avail- 
able to crops is a factor of prime importance. As the initial water 
supply stored in the soil and the quantity that may be supplied by 
rains during the growth period are limited and susceptible of meas- 
urement, it is of fundamental importance in any study of the sub- 
ject to know the daily rate of the use of water by the growing crop 
and the total consumption by the crop during its life period. In 
investigational work it has appeared especially important to deter- 
mine the actual rate of use in the field in order to properly evaluate 
quantities of water in terms of the length of time they would meet 
the requirements of the crop. 
Other investigators have made exhaustive physiological investiga- 
tions of the water requirements of crops directed chiefly to the 
measurement of the water actually transpired by the crop plants 
alone. 
In the field the crop has an initial supply measured by the avail- 
able water stored in the soil within the zone of depletion by the crop. 
This is replenished or added to from time to time by the precipitation. 
The water is used by the crop and by the weeds that accompany it 
and is lost by direct evaporation and very seldom under dry-farm- 
ing conditions by leaching. A portion of the precipitation may be 
lost by run- off. 
In the field it is not possible to separate the water actually used 
by the crop in the sense of passing through the tissues of the plants 
comprising it from that lost by other means. If it were possible, 
it does not necessarily follow that it would be desirable. The water 
