620 
Journal of Agricultural Research 
Vol. XXVII, No. 9 
wet and to shrink on drying that constitutes one of the chief difficulties 
in the way of a clear understanding of its water-holding capacity. Soils 
differ greatly with respect to changes in volume that take place when 
they are alternately wetted and dried. Pure sand does not change its 
volume materially whether it is saturated with water or dry. Soil con¬ 
taining clay, on the other hand, swells extensively when wet and shrinks 
again when dried. 
It is possible to demonstrate that a certain volume of pure sand retains 
its volume w r hether it is saturated with water or dry. If its volume mass 
is determined when it is wet and again when it is dry it wall be found that 
the difference in weight between the dry condition and the saturated 
condition is substantially the same as the weight of the water required 
to fill the computed pore space. The computed pore space may range 
from 34 to 28 per cent, depending upon the assortment of sizes of the 
particles and the pressure to which the wet material has been subjected. 
The percentage of water occupying 34 per cent of pore space is approxi¬ 
mately 19.5, while that occupying 28 per cent of pore space is about 14.6. 
The water which occupies or fills the pore spaces between the soil parti¬ 
cles may be referred to as interstitial water. 
When soil contains clay as well as particles of sand of assorted sizes 
its reaction with water becomes strikingly different from the reaction 
of pure sand. The addition of water to soil that contains clay tends to 
increase the volume of the soil mass. When the soil is not subject to 
pressure the increase in volume when wet may be very great. In the 
case of subsoil it is probable that the tendency to increase in volume on 
wetting is met by the pressures that develop within the soil mass. The 
quantity of water that may be absorbed by the soil some distance below 
the surface is very much less than the same soil will hold if it is at the 
surface. 
These differences of water-holding capacity have been demonstrated 
by a series of tests made at the Newlands Experiment Farm by 
F. B. Headley. 2 Samples of subsoil were taken with a soil tube in places 
where free underground water was found within 2 to 3 feet of the surface. 
The soil samples were taken from below the level of the ground water. 
The moisture content was determined on the samples obtained with the 
soil tube, and the dried samples were then pulverized, sifted, and placed 
in metal cups of 20 cc. capacity having perforated bottoms. The soil 
in^ these cups was then saturated with water and weighed to determine 
its water-holding capacity in the absence of pressure. In texture the 
soil samples ranged from coarse sand to fine clay. For convenience in 
comparison they are arranged in Table II in groups with respect to 
texture. The results of these tests show that there are not only marked 
differences in the water-holding capacity of soils of different textures as 
they exist in the lower part of the root zone but that the water-holding 
capacity of the same soil is much greater when it is on the surface and 
freed from pressure than when it lies some distance below the surface 
and is subject to pressure. 
* These experiments were made at the writer's request during August, 1933. 
