Apr. 9 , i9if 
Water-Retaining Capacity of Soil 
67 
in coarse sands when the downward movement has become so slight as 
to be almost negligible, the upward movement in these is limited to very 
short distances. Prom these considerations it is evident that the amount 
of water which the deep subsoil can contribute to the growth of annual- 
crop plants will be of no practical importance. When a perennial crop 
with a root range of 20 to 30 feet follows an annual with one of only 4 to 6 
feet, the moisture of the deeper subsoil becomes of great importance; but 
here it is a case of the roots going to the moisture and not of this being 
elevated to them by capillarity. 
The experiments described, however, do not answer the question as 
to whether the moisture of the deeper subsoil may not, in the course of 
several years, or of a few decades, be elevated through a much greater 
distance. For instance, whether after the subsoil to a depth of 20 or 
30 feet has been exhausted of available moisture (brought to a ratio of 
1.0 to 1.1) by deep-rooted perennials, it may not eventually have the 
ratio restored to 1.7 to 2.4 by capillary movement from the deeper sub¬ 
soil instead of only by the portion of the precipitation reaching it from 
above. Even if a decade were required for such a transfer, the elevated 
moisture might still have some practical importance for the perennial 
crops. Field investigations and laboratory experiments which would 
definitely decide this point appear simple in principle. The character 
of suitable cylinder experiments will be evident from those discussed 
above. For these the calcareous loessial silt loams of the Great Plains, 
and the so-called “volcanic ash ” of eastern Washington and Oregon would 
be especially suitable. 
A field study, if sufficiently thorough, would give a more satisfactory 
answer, but it would be far more laborious, and in most places would 
appear to be quite impracticable. It should necessarily be conducted in 
a region of limited rainfall, and even there a wet year or two might 
cause the experiment to miscarry entirely. Desirable conditions would 
include a silt-loam subsoil comparatively uniform to a depth of 40 feet, 
free of any interrupting sand or gravel layers, and a water table at a 
depth of not less than 100 feet. The subsoil of the field practically 
exhausted of available moisture to a depth of 20 or 30 feet, as by a long 
continued stand of alfalfa, should be sampled at intervals of a foot or so 
from the surface through the dry zone and well into the moist deeper 
subsoil. The places of sampling should be sufficiently numerous to 
establish the uniformity of the distribution of moisture; and with all 
samples there should be a determination of the hygroscopic coefficient 
or moisture equivalent, as well as of the total moisture. In beginning 
the study the perennial crop on the experimental field should be at once 
killed to prevent further loss by transpiration, and only annuals should 
be allowed on it during the experiment. The thickness of stand and the 
period of growth of these annuals should be such as to intercept most 
