RESEARCH METHODS IX STUDY OF FOREST ENVIRONMENT. 97 
be remembered that, while the table indicates three times as much 
water-holding capacity in the granitic loam of high humus content, 
the actual increase over the same type of soil with little humus is only 
about 65 per cent on a volume basis. 
The sum of these various effects of different soil properties on the 
moisture-holding properties is shown in the final line of Table 4, 
where it is clearly indicated that there is a closer parallelism between 
wilting coefficients and moisture equivalents than between wilting 
coefficients and capillarity. Eliminating the granitic gravels, the 
average variation of the group ratios is only 7.5 per cent from 
a mean value W. C./M. E. of 0.318. The explanations given, more- 
over, all tend to confirm the belief that the moisture equivalent ob- 
tained with a much greater centrifugal force would give a still 
closer index to the wilting coefficient of any of these types of soil. 
The hygroscopic coefficient is an expression of the amount of water 
held by a soil after a limited exposure to saturated water vapor under 
certain conditions. As in the case of the capillary moisture measure, 
it appears that Hilgard was the first to make practical use of the 
absorption powers of soils, to compare them generally as to physical 
properties, and to obtain an approximate measure of their wilting 
coefficients. More recently Alway (102, 103) has done a large amount 
of work on this subject, using Hilgard's methods very largely, but 
also investigating many possible sources of error in the routine 
treatment of samples. 
It is a very well-known fact that a soil is never entirely devoid of 
moisture if dried in the air for an indefinite period. On the con- 
trary, if atmospheric conditions did not fluctuate so rapidly there 
would be at all times an amount of moisture in the soil somewhat 
proportionate to the amount of vapor in the atmosphere. The amount 
so held is a measure of the soil's hygroscopicity, but not a useful 
measure because of the changing conditions of the atmosphere. 
Similarly a soil undoubtedly still possesses some hygroscopic moist- 
ure when dried in an oven at, say, 100° or 110° C. The only way in 
which the soil can eventually be robbed of all its moisture is by 
drying in a vacuum, by means of which the constant withdrawal of 
the atmospheric vapor is assured. For practical purposes, how- 
ever, drying in an ordinary atmosphere at 110° C. gives a good basis 
for moisture calculations, since at that temperature the vapor in the 
atmosphere will be very much rarefied in comparison with its satura- 
tion capacity. This point is mentioned because it is not infrequently 
noted, in drying large samples, that they may gain moisture in the 
hot-air oven if there is a decided increase in atmospheric moisture. 
To avoid appreciable errors it has been found necessary to avoid 
final weighings of oven-dried samples on excessively moist days. 
8276^—22 7 
