100 BULLETIN 1059, U. S. DEPARTMENT OF AGRICULTURE. 
results reference is again made to the comparisons made by Briggs 
and Shantz between hygroscopic and wilting coefficients. In 17 tests, 
with soils varying from 0.9 to 16.5 per cent wilting coefficient, they 
found the ratio of hygroscopic to wilting coefficients to be on the 
average 0.680, with a probable error, or variation, in any single deter- 
mination of about 7.1 per cent of the wilting coefficient. It is to be 
noted that the hygroscopic is so much lower than the wilting co- 
efficient that serious error would result from considering them as inter- 
changeable, though this proposal has sometimes been made. 
Calculation of the Available Moisture. 
As has been stated, when the current moisture of the soil has been 
measured, and the nonavailable has been measured in the laboratory 
by the direct method of wilting tests, or indirectly through the 
capillaiw moisture, moisture equivalent, or hygroscopic coefficient, it 
is then only necessary to subtract the wilting coefficient from the 
whole moisture to have a measure of the amount of water which, 
under the most faA T orable circumstances, will be available for growth. 
For example, if in sand and clay, respectively, the whole moistures 
are 10 and 20 per cent, and the wilting coefficients of these soils 
are respectively 2 and 15 per cent, then it is evident that in the 
sand there is 8 per cent available moisture, and in the clay 5 per 
cent, or A—M— WO. The use of the last figures is certainly far mor^ 
expressive of the relative conditions in the two soils than would be 
the use of the whole moisture figures, although, on account of vary- 
ing concentrations of salts, even this figure for the available moisture 
does not give a direct means of comparing the moisture conditions 
of radically different soils. 
Of course, if the measure of available moisture is to be used most 
fully as an index to supply, the percentage should be transposed 
finally into cubic centimeters per cubic meter, or any other measure 
of soil volume. 
This is very readily done if the apparent density has been de- 
termined, as in the large capillary cans described, vdiere the apparenl 
density is obtained by dividing the dry-soil weight, in grams, by the 
volume in cubic centimeters, which is approximately 1,030 cubic centi- 
meters (usuall} T less after centrif uging) . 
Carrying the volume idea still farther, in studying any plant or 
group of plants it is obviously desirable to know how much soil sur- 
face can be drawn upon. Thus a yellow pine on a dry site may 
actually have a much greater supply of moisture than a crowded 
spruce on a moist site. Consideration of this point of view will lead 
to the conclusion that soil moisture figures, as ordinarily given in 
percentages of the dry-soil weight, have almost no significance 
