USE OF WATER BY SPRING WHEAT OX GREAT PLAINS. 6 
lost by weeds, evaporation, rim-ofi'. and other factors is as truly a 
part of the water used by a crop under a given method of cultivation 
in any region as is the water actually transpired by the vegetative 
surface of the crop. 
To know the requirements of the crop in the field it is conse- 
quently necessary to determine the quantity of water that is lost 
from the soil and the quantity that is supplied by precipitation. 
The sum of these quantities for any period constitutes the water used 
by the crop. When this is known for any period the daily rate of 
use can be calculated directly from the number of days in the period. 
Many problems arise in addition to or in connnection with the 
main one. Some of these are considered in the present study. 
SOURCE, CHARACTER, AND METHOD OF STUDY OF THE DATA. 
In the 14-year period from 1907 to 1920. inclusive, soil moisture 
determinations have been made at 23 field stations for a total of 
205 station years. ^Aliile data have been obtained from all the prin- 
cipal crops.' the present study is confined to the wheat crop. This 
has been durum wheat and mostly of the Kubanka variety or type. 
From the data available, that from two plats has been selected for 
the present study : Plat A. continuously spring plowed and cropped 
to wheat, and plats C and D. which are alternately fallowed and 
cropped to wheat, so that wheat is grown on one of them on fallow- 
each year. 
The data for individual years are not all that might be desired, 
as they were not taken especially for the present study. Greater 
intensiveness through more frequent sampling would have added 
to their value and made possible the use of the data of many years 
that must be rejected because determinations were not made at the 
necessary- stage of growth or condition of soil moisture, but the cases 
that can be used make up a volume of observations that probably 
more than offset the intensiveness that might have been obtained 
had the work been limited to a smaller number. 
All determinations here studied have been made on 1-foot sections 
of soil. 
Soil samples are taken with a tube having a diameter at the cut- 
ting edge of 20 millimeters. To determine the water content of a 
plat four cores are taken from it at locations representing its four 
quarters. The four cores are handled as two samples, the two cores 
from opposite corners being placed together. Determinations are 
thus made in duplicate, each one of the samples being made up of 
two cores and having a weight of dry soil var^-ing from 200 to 800 
grams. 
Each core of soil when drawn from the ground is discharged from 
the tube into a can of sufficient size to hold two cores. The cans are 
pro^^ded with tight covers, which prevent any appreciable errors 
from loss of water before weighing, even though weighing should be 
considerably delayed. To facilitate weighing, each can is made to 
balance a weight of 110 grams. The samples are immediately taken 
to the laborator}^ and the net weight determined to the nearest one- 
tenth of a gram. 
After the wet weight is determined the cans are opened and placed 
in an oven, where they are subjected to a temperature ranging from 
