430 
Journal of Agricultural Research 
Vd. XXIV, No. 5 
ice capillary columns, or long needle-like crystals. The force of crystalli¬ 
zation seems to pull the water from below and bring it to the sur¬ 
face, where it freezes into these massive ice capillary columns or com¬ 
pact needle-like crystals. In Plate I a typical example of this phenome¬ 
non is shown. This picture was taken on a muck soil during the latter 
part of November, when the soil temperature below the surface was still 
considerably above the freezing point. The ice capillary columns would 
be formed at the surface of the soil without penetrating the lower depths, 
growing upward as straight needles or thin capillary tubes massed to¬ 
gether. The growth seems to take place at the lower end and push 
the entire column upward, as the capillary tubes are elongated from 
below. The ice column shown in Plate I is about 4 inches thick, and 
was formed during three nights. The formation for each night is indi¬ 
cated by the lines or layers seen in the column. 
As previously stated, the water which went to make this 4-inch column 
of ice came from the capillary water of the soil at a lower depth, and 
was brought to the surface by the pull or force of c^stallization. From 
these results it is easily understood that it is possible for the moisture 
to move from the finer capillaries, and from around the particles 
as films, to the larger capillaries of a soil short of saturation. This 
phenomenon of the transference of moisture from the smaller to the 
larger capillaries upon freezing is somewhat analogous to another phe¬ 
nomenon—^the tendency of small drops of liquid to unite into a single 
drop, which is accomplished'either by actual contact or by the trans¬ 
ference of vapor from the smaller to the larger drops. 
In referring again to the effect on the freezing-point depression, another 
question arises: Why should the water in the large capillaries affect the 
freezing-point depression differently from that in the finer capillari^? 
These differences can be easily explained if the hypothesis® previously 
advanced is correct. This hypothesis assumes that the solution immedi¬ 
ately around the soil particles and in the very fine capillary spaces is 
less concentrated than the mass of the solution. This assumption which 
accords with the results presented in this paper, holds that the force of 
crystallization tends to draw the moisture from the finer capillaries and 
from around the particles as films into the larger capillaries. It is readily 
seen that during freezing and thawing the dilute solution from the finer 
capillaries and the films from around the particles go to dilute the 
solution in the larger capillaries or the mass of the solution. The con¬ 
sequence is that the original freezing-point depression is diminished. 
When the soil mass is stirred the moisture is again redistributed and 
readjusted and the freezing-point depression becomes as before. 
If the h3q)othesis of the difference in concentration between solution 
in mass and that in the finer capillaries and around the soil particles is 
true, the above explanation is probably the correct one. But Parker • 
has published results to show that when the water is reduced to film or 
capillary form it has a decided physical effect upon the freezing-point 
depression. If his claim is true the explanation that immediately 
suggests itself is that the water in the larger capillaries has less physical 
® Bouyoucos, George J. op Cit. _ 
• PaRKBR, F. W. MBTHODS op studying THB CONCBNTRATION and composition op THB son, SOLUTION. 
In Soil Sci., v. 12, p. 209-233. 1931. References, p. 231-333. 
