6 
BULLETIN 1139, U. S. DEPARTMENT OF AGRICULTURE. 
There is an experimental error in sampling, depending largely 
upon the lack of uniformity of the soil. Different soils vary greatly 
in this respect. Almost all soils in the Great Plains become less uni- 
form in character and structure as distance from the surface in- 
S 
2 
/ 
O 
33 
30 
23 
20 
X5" 
iS/?M 
fEM 
at/w. 
/?#?. 
M/?r 
<SW£ 
<J6/LY 
/&/& S£W 
OCT 
/yot? 
£>£C 
Site 
MINI 
Mill 
1 1 1 j ii 
Mill 
1 1 1 1 1 1 
Mill 
1 1 l l 1 1 
Mill! 
1 1 M 1 
1 M I I 
j 1 1 1 1 
1 1 M 1 I 
,1 
1 ! 
L. , 
.... . . i 
■ A. t 
/^r 
■»■— >, 
3 
O 
33 
30 
23 
20 
~ /S 
/o 
s 
o 
33 
30 
23 
20 
/3 
/O 
3~ 
O 
33 
30 
23 
20 
/3 
/O 
3 
O 
33 
SO 
23 
20 
Z3- 
/O 
O 
33 
SO 
23 
20 
/jS" 
/O 
.% 
"^\^, 
_-^^ 
\"2 ""— 
_ _ _. 
2™ 
. 
"^^C 
3^ r - 
^.rr 
SF*< 
— -^^ 
»**»CJ!^ 
6^ 
•inn 
- 1 
|i|||| 
i i nil i i I ii 
1 1 1 1 1 
i i i i i 
I I I I I 
i i i 1 1 
Mill 
1 1 1 1 1 
i i i ii 
LL! • 1 
Mill 
Fig. 1. — Diagram showing the percentage of water in the several foot sections of soil of 
plat B in wheat at Williston, N. Dak., at each date of sampling in 1910. 
creases. Thus, any single sampling can not be taken as giving an 
exact measure of the water content of a plat, but must be considered 
as giving a close approximation, the closeness depending upon the 
uniformity of the soil. In all of the years and stations under study 
