26 
BULLETIN 1004, U. S. DEPARTME]S"T OF AGEICULTURE. 
and that yields increase in proportion to the quantity of water con- 
sumed above this. For those yields under 2,000 pounds per acre 
there is a wide variation in the ratio of water used to yield. This 
is not unexpected, as such yields do not represent normal growth 
throughout the season or in many cases for any part of the season. 
A further cause of divergence is the wide range of conditions from 
the northern to the southern Plains covered by the data. There is a 
much closer grouping for the yields above 2,000 pounds per acre, as 
these are more nearly expressive of the requirements of normal 
growth. 
The yields have been averaged in groups of 1,000 pounds and the 
corresponding averages of the respective ratios determined. These 
averages are given in Table 4 and shown in Figure 8 by crosses 
Avithin circles. No attempt has been made to determine a mathe- 
matical expression of the relation evidenced in the data comprising 
Figure 8. 
Table 4. — Ratios and total yields of iclieat shown in TaMe 3 arranged in yield 
groups of 1,000 pounds and expressed a^ averages of such groups. 
Yield group. 
Number 
in group. 
Yield 
average. 
Ratio 
average. 
Yield group. 
Number 
in group. 
Yield 
average. 
Ratio 
average. 
1 to 1,000 pounds 
1,001 to 2,000pounds. 
2,001 to3,000pounds. 
3,001 to 4,000 pounds. 
4,001 to 5,000 poiinds. 
15 
26 
14 
18 
6 
Pounds. 
793 
1,413 
2,444 
3,492 
4,323 
2,165 
1,588 
951 
814 
754 
5,001 to 6,000 pounds 
6,001 to 7,000 pounds 
7,001 to 8,000 pounds 
8,001 to 9,000 pounds 
1 
2 

1 
Pounds. 
5,200 
6,225 
686 
615 
8,650 
382 
While the ratios of water to yield obtained in this work range 
higher than those obtained in controlled wat^r-requirement studies, 
such as those reported by Briggs and Shantz,^ there are sufficient 
reasons why this should be the case. These field studies include all 
loss from the soil itself which was excluded in physiological studies. 
Our yields are relatively lighter, because they do not include an un- 
determined residue that remains as stubble in the field plats from 
which the crop is harvested with the self-binder but which was 
included with the crop by clipping close to the ground in pot cul- 
tures. This difference, however, is partly offset by the difference 
between dry and air-dry material. Maximum efficiency was obtained 
in pot cultures by constantly maintaining an optimum water content, 
a condition which is the exception rather than the rule in the field. 
The results all show that the efficiency of water is increased (the 
water requirement reduced) by the maintenance of conditions that 
favor maximum production. 
CORRELATION BETWEEN THE USE OF WATER AND THE YIELD 
AT INDIVIDUAL STATIONS. 
A study was made of the correlation between water use and yield 
at two stations — Edgeley, N. Dak., and North Platte, Nebr. These 
stations were not selected because of their location but because re- 
sults for more years were available at these stations than at any of 
2 Briggs, D. J., and Shantz, H. L. The water requirement of plants. U. S. Dept. Agr., 
Bureau of Plant Industry Bui. 284, 49 pp., 2 fig., 11 pi. X913. 
