10 
average rate of use at Archer agrees with the rate in North Dakota 
and South Dakota. This station is located at an altitude of more 
than 6,000 feet and has a summer temperature, evaporation, and other 
climatic factors more nearly like the Northern States than stations 
in the same latitude. 
Table 2. 
-Rate of the use of toater per day during a period of rapid groioth on . 
both plats A and C or D averaged for each station. 
Use of water. 
Station. 
Use of water. 
Station. 
Plat 
A. 
Plat 
CorD. 
Plat 
A. 
Plat 
CorD. 
Huntley 
Inch. 
0.14 
.15 
.15 
.15 
.17 
.15 
.15 
Inch. 
0.20 
.21 
. lo 
.16 
.19 
.18 
Inch. 
0.21 
.18 
.20 
.15 
.20 
.17 
.24 
Inch. 
0.20 
Williston 
Scottsbluff 
.21 
■Oinldnson 
North Platte 
.23 
Archer 
.15 
Edgeley 
Akron 
.20 
Hettinger 
Garden City 
.22 
Belle Fourche 
.30 
With plat C or D the regional effect is more obscured by the in- 
fluence of the size of the crop on the rate of use. It is most fully 
exemplified in the high rate at Amarillo, which averages 0.3 inch 
per day. This is the average of the rates of 0.33 inch and 0.26 inch, 
respectively. One of these is the highest yet determined and tlie 
other has been equaled only once, which was at North Platte in 
1909. The rates of use at North Platte and at Garden City also 
average high in comparison with those at more northern' stations. 
Archer has a comparatively low rate on this plat, the same as on 
plat A. 
In general terms, the average rate of use on plat A ranges from 
an average of 0.15 inch per day at the northern stations to 0.24 inch 
at the most southern station and on plat C or D from nearly 0.2 
inch per day at the northern stations to 0.3 inch at the station farthest 
south. The most southern station therefore shows a daily rate of 
use from 50 to 60 per cent greater than the northern stations. 
The aA^erages of plats A and C or D show a relation between the 
rate of use and yield. A close correlation between the two should 
not be expected, because the rate of use of water was determined 
for a period that covered only part of the life of the crop. 
In many cases the crop suffered severely from drought at some 
time after this rate was determined. Other factors, such as ex- 
tremely hot weather, may have influenced it after the rate had been 
determined. It was thought, however, that if any relation between 
the rate of use of water and yield existed, it would be shown through 
a positive correlation, even though the correlation might not be 
strong. 
The first correlation was made by using the average yield and the 
average rate of all stations as bases from which to compute de- 
partures. The resulting correlation for plat A was 0.24, with a 
probable error of ±0.09. This is a very weak correlation and only 
indicates that a relation may exist. It was not thought possible 
that a close correlation can be arrived at in this way. The facts 
that the northern stations yield heavier than the southern ones and 
