May 31, 1924 
The Heat of Wetting of Soil Colloids 
929 
The heats of wetting of soil colloids are shown to be of an order comparable 
with those of organic and inorganic gels studied by various investigators. 
Rodewald (14) found 19.4 to 24 calories evolved per gram of starch. Bouyou- 
cos 4 (2) found 9.6 calories per gram for “ferric hydroxide” and 21 calories for 
“ aluminum hydroxide.” Various investigators working with different prepara¬ 
tions of silica gel have found from 6.9 to 24.2 calories per gram (5, 10). 
Table I shows that the heat liberated from the different soil colloids varies 
widely. For example, the Wabash colloid, which gives over 17 calories per gram, 
has nearly four times the heat value of the Cecil colloid with 4.5 calories. This 
difference is of about the same order as the difference in the extreme values for 
dye and ammonia adsorption. 5 The adsorption of water vapor, however, was 
much less variable. 6 
It is to be noted also in Table I that there is not much correspondence between 
the heat of wetting of the colloidal material and of the whole soil from which it 
was isolated. The heat liberated by the colloidal material was in some cases 
about twice that of the whole soil and in other cases more than six times the soil 
valhe. Certain of the soils gave higher heat values than the isolated colloids 
from other soils. The Miami, Sharkey, and Wabash soils, for instance, each 
gave a higher heat of wetting than the colloidal material from the Cecil. The 
heats of wetting of the different soils would not, therefore, be expected as a rule 
to show the relative quantities of colloidal material present. However, soils 
known to contain colloids of similar properties might be compared on the basis 
of heat of wetting of the whole soil and some idea of the relative colloid content 
be obtained. 
HEAT OF WETTING OF NON COLLOIDAL SOIL PARTICLES 
It was shown in a former publication of this bureau ( 1) that the adsorptive capac¬ 
ity of the soil was associated almost entirely with the colloidal matter, and that 
the minerals making up the noncolloidal portion possessed but little adsorbing 
power. Of the common soil fninerals, the micas were the only ones to show any 
significant adsorption. It was thought desirable to ascertain whether or not the 
heat of wetting of a soil was also practically all due to the colloidal material. 
The heat of wetting of mineral powders ground to a degree of fineness compar¬ 
able with that of the minerals in the soil should show whether or not the non¬ 
colloidal part of the soil contributes appreciably to the heat of wetting of the 
whole soil. 
For the study of mineral powders, commercial samples of quartz flour and or- 
thoclase were used, all of which passed a 130-mesh sieve and contained much 
very fine material. Samples of hornblende and muscovite were prepared by 
grinding the minerals several days in a steel ball mill. These samples also passed 
a 130-mesh sieve and contained much that was very fine. The quartz, horn¬ 
blende, and orthoclase each gave a temperature increase on wetting of less than 
0.05° C., which was not greater than the probable error. Thus these common 
soil minerals above colloidal size would not contribute appreciably to the heat of 
wetting of a soil. Muscovite gave a heat of wetting of 1.3 calories per gram. 
This quantity is small, but it might constitute a significant part of the heat value 
of a soil containing a small quantity of colloid and a high percentage of mica. The 
4 These calculations are made on the basis of air dry weight. 
* Additional data, as yet unpublished, indicate a close relationship between the heat of wetting power of 
a soil and its capacity to adsorb ammonia. 
6 It was indicated that the approximate constancy of the different colloids for water vapor adsorption 
probably applied only to an atmosphere whose humidity was similar to that produced by 3 peF cent sul¬ 
phuric acid (by weight). If stronger acid were used the results would be more variable U, p. 12). 
