86 BULLETIN 1059, U. S. DEPARTMENT OF AGRICULTURE. 
that may be held by the soil against the force of gravity. This 
amount decreases as the height of the column of soil increases, and 
may also be considerably influenced by the packing of the soil. 
Since it is almost impossible to treat any soil in the same state of 
compactness in which it is found in the field, or to establish a stand- 
ard condition for soils in vessels, this measure of the water-holding 
power of a soil is not likely to have precise value. The greatest 
theoretical objection to it is. that the force tending to remove the 
water from the soil is of an entirely different magnitude from that 
at work as the plant makes its final struggle for water, and thai 
the effect produced by the one force can not serve as a measure of 
the effect which might be produced by the other. There seems also 
to be an impression that salts in the soil water operate to raise the 
wilting coefficient, while decreasing the capillary moisture by lower- 
ing the surface tension of the liquid. Such an impression arises 
from the well-known effect of foreign substances on the surface of a 
liquid. It has been pointed out by Free (121) that salts in solution 
actually increase the surface tension of the liquid, and this is entirely 
in keeping with the known properties of solutions. While the pres- 
ence of solutes may have the effect of weakening the affinity of one 
water molecule for another, this is fully counterbalanced, in its rela- 
tion to capillarity, by the greater density of each group of molecules 
of which the solute forms a nucleus, and the consequent greater 
affinity between such groups and the solid surface. This affinity is 
known by the name of " capillary attraction." Furthermore, even 
while admitting that in either the capillary moisture test or the 
moisture equivalent test some of the solutes may be lost with the 
water which is drained out of the soil, considerable satisfaction is 
gained from the idea previously set forth that, at the wilting point 
of soils, these solutes may be absorbed by the colloids. 
It is believed that Hilgard (1'25) was the first to employ the 
principle of capillarity for comparing soils. He used a sieve cylin- 
der only 1 centimeter high, which, after a layer of filter paper was 
placed in the bottom, was filled level with the soil. This was im- 
mersed to a depth of 1 millimeter in distilled water, allowed to stand 
for an hour, and then weighed. The amount of water absorbed, of 
course, was dependent on the ability of the soil to lift it. a maximum 
distance of 9 millimeters. 
Briggs and Shantz (114) compared this measure of absorbing 
capacity with the directly determined wilting coefficients of 15 soils 
whose wilting coefficients ranged from 0.9 to 1G.7 per cent. From 
these comparisons it is evident that a soil which is able to withhold 
almost no moisture from plants has a fairly high capillarity, but 
that the latter does not increase in so great a proportion as the 
wilting coefficient with more retentive soils. Thus it was neces- 
