30 
Research Bulletin No 3 
dried at the end of 12 hours, thus permitting two charges, in- 
stead of a single charge, to be dried every 24 hours. In the 
tables in the following pages the amounts of water are expressed 
as percentages of the dry soil. 
FREE WATER. 
The free water represents the difference between the total 
water and the maximum hygroscopic water (hygroscopic coeffi- 
cient). The soil moisture in the following tables is expressed 
usually as the free water only but in a few both the total water 
and the free water are given. In all cases the total water may 
be found by simply adding the free water to the hygroscopic co- 
efficient. 
An expression of free water alone is almost as meaningless as 
that of total water alone. Thus corn plants may be dying in a 
soil with a hygroscopic coefficient of 15 and containing 3 per cent 
free water while plants of the same kind in a soil with a hygro- 
scopic coefficient of 0.8 may be very vigorous with only 1.0 per 
cent free water. For this reason in order to give tables of free 
water their full significance there should at the same time be 
given also corresponding tables of hygroscopic coefficients or of 
total water. Any two of the three tables are sufficient but some 
will find a certain combination the most satisfactory and others 
another. The writer prefers the statement of free water and 
hygroscopic coefficient, but those who have made many determi- 
nations of total water and none of free water will, at least at first, 
find the statement of total water along with the hygroscopic co- 
efficient to have more significance for them. 
Total tvater — hygroscopic coefficient = free water. 
The hygroscopic coefficients of the soil samples were de- 
termined by the Hilgard method with a slight modification. In- 
stead of using glazed paper to hold the soil, flat aluminum trays 
were employed. An extensive study of the determination of hy- 
groscopic moisture had shown that this modification of the Hil- 
gard method gives concordant and reliable results and also that 
the modification used in determining the hygroscopic coefficients 
as reported in previous publications 1 gave results much too low. 
The discussion of methods for the determination of the hygro- 
scopic coefficient is reserved for a future publication. 
For the determination of the hygroscopic coefficients the in- 
^lway, F. J., and McDole, G. R. Studies on the Soils of the Northern 
Portion of the Great Plains Region: the Distribution of Carbonates on the 
Second Steppe. American Chemical Journal, vol. 32, 1907, p. 275. — Alway, 
F. J. Some Soil Studies in Dry-Land Regions. Bui. 130, Bureau of Plant 
Industry, U. S. D. A., 1908, p. 360. Studies of Soil Moisture in the "Great 
Plains" Region. Journal of Agricultural Science, vol. 2, 1908, p. 333. 
