136 BULLETIN 1059, U. S. DEPAETVLEXT OE AGRICULTURE. 
Stjmmaby of Soils Discussion. 
The preceding discussion has attempted to bring out the 'theoreti- 
cal considerations which make the study of soils very important in 
forestry, from the standpoint of initiation of seedlings, later com- 
petition between individuals and species, and the rate and ultimate 
limit of height growth on any particular site. In the main all other 
soil conditions have been considered in their bearing on the supply 
of soil moisture. In view of the length of the discussions, it would 
appear desirable to repeat the salient points, as follows : 
1. It is believed that from every ecological aspect the important 
soil condition is the availability of the soil moisture. 
2. Plans for the study of this soil condition have been based on 
the assumption that the relation between the plant and the soil in 
which it grows can best be demonstrated if. at any time, the status 
of either may be expressed in terms of osmotic pressures. 
3. The generally coarse character of forest soils, and the presence 
of rocks which are as characteristic as any other part of the soil and 
can not properly be eliminated, give rise to the need for special 
methods of examining forest soils, and particularly for methods 
adapted to larger samples of the soil than have commonly been used 
in agricultural investigations. 
4. The total moisture of the soil, while not directly making possible 
the comparison of sites, if there be any variation in soil composition, 
must- be had for most of the indirect methods of comparison; and in 
forest studies it must be determined periodically through one or more 
seasons in order to discover the conditions that are critical. The 
quantity may sometimes be found through ordinary methods of sam- 
pling and drying the soil samples : but often, because of mechanical 
difficulties, and to insure greater physical uniformity in the samples 
from time to time, it is desirable to have " wells " of prepared soil 
from which successive samples will be taken. 
5. If it seems desirable to compute the moisture of the natural 
soil from that found in a soil well, this may be done, at least approxi- 
mately, by comparison of the capillarities, moisture equivalents, and 
wilting coefficients of well soil and natural soil, respectively. It seems 
probable, however, that up to a high moisture content osmotic equi- 
librium is more likely than capillary equilibrium between the well 
and the natural soil, so that if the moisture of the former may be 
expressed in terms of osmotic pressures, it is unnecessary to compute 
the moisture of the soil. 
6. For any study of the critical situations in soil moisture, either 
for seedlings or for older trees, it is necessary to know the wilting 
coefficient of each soil under consideration. The moisture content at 
which a plant may wilt, however, varies widely not only according to 
