RESEARCH METHODS IX STUDY OF FOREST ENVIRONMENT. 139 
capillary tensions are set up. and these do not react in the same way 
on relatively large and very small quantities of water in the soil. 
It appears, therefore, as suggested by Free (121), that the effective 
procedure is the employment of a variable force, sufficient in the 
case of any particular soil to extract all of the water which is extract- 
able. It seems probable that this quantity would correspond to all of 
the liquid water capable of moving from one part of the soil to 
another. The remaining water would probably correspond closely 
to the "unfree water" of Bouyoucos (106) and what in this bul- 
letin is termed " water vapor.*' or water whose molecules are too 
rigidly held by solid substances to have the motility of liquid mole- 
cules. While this method is as yet untried and would obviously be 
more laborious than the present standardized procedure, possibly pre- 
senting new mechanical difficulties, it promises so much as a direct 
and rapid means of determining a physical constant that it deserves 
serious investigation. In the meantime the 1,000-gravitv test should 
be employed as the basis for comparing the moisture conditions of 
various soils and in the detailed study of their wilting coefficients. 
10. Once the wilting coefficient of a soil has been determined, di- 
rectly or indirectly, the current moisture condition may be expressed 
in terms of the percentage of available moisture or the available 
moisture per unit of soil volume, by subtracting the nonavailable 
moisture from the whole. The amount of water per unit of volume 
recommends itself particularly in comparing the conditions of open 
and dense forest stands, provided that the root extent of the indi- 
vidual tree has been investigated. Without such information, very 
wrong conceptions of the moisture supplies available to the indi- 
vidual trees in different forest types are likely to be formed. 
11. For the purpose of expressing the condition with which any 
individual plant is coping, particularly the conditions against which 
a seedling must struggle in times of drought, it is very desirable to 
reduce the water content of the soil to terms of availability. If it 
is assumed that the wilting coefficient stands for a definite osmotic 
pressure in the soil, with which the osmotic pressure in the plant 
is in equilibrium (this, of course, being only approximately true, as 
pointed out in paragraph 6, and being further subject to the condi- 
tions of the plant, as indicated in the following paragraph), then, 
when the moisture content is equal to the wilting coefficient, the 
availability of the soil water is 0. 
When the moisture content of the soil is twice as great as the wilt- 
ing coefficient, about twice the osmotic pressure may be expected in 
the plant as in the soil, and this should make possible a fairly definite 
rate of absorption by the plant. The availability at this point may 
be called 0.50. Similarly, when the moisture content is three times 
the wilting coefficient, the availability may be expressed as 0.67. It 
