ABSORPTION BY SOIL CONSTITUENTS. 15 
ble hydrolytic products of colloidal nature. This action of course 
becomes more pronounced the finer the particles. Cushman (6) 
shows that this action takes place to an appreciable extent in a 
few hours and suggests that a colloidal fiJm is thereby formed on 
the surface of the particles. Michaelis (15) states that some finely 
powdered quartz which he prepared absorbed water and sw^elled; 
also that a silicic acid gel may be made by boiling a quartz suspension 
for several days. We found that treating powdered orthoclase with 
hot water for two days, then leaching free from soluble salts, increased 
its absorptive capacity for water vapor from 0.0016 gram to 0.0046 
gram. 
The fact that the most finely powdered minerals, even though 
they must have suffered considerable alteration, had absorptive 
capacities less than the ultra clays, affords some evidence that the 
colloidal matter in soils is not made up of merely finely divided 
minerals. More conclusive evidence of this, however, is afforded by 
analytical data which are being prepared for publication. These data 
show that no probable combination of soil minerals has th§ same 
chemical composition as the typical colloidal matter in soils. 
ABSORPTIVE CAPACITIES OF SYNTHETIC GELS. 
It was thought that further evidence on the nature of the col- 
loidal matter in soils mi^ht be obtained from the absorptive capaci- 
ties of inorganic gels; since, as previously stated, other investiga- 
tions in this bm-eau indicate that the colloidal material in soils is 
largely of a gel-like nature. An examination of the ultra clays 
extracted from many diff'erent soils has indicated that they consist 
chiefly of aluminum silicate with some ferric silicate and one or more 
of the following substances in the gel form; ferric hydroxide, alumi- 
num hydroxide, organic matter, and possibly silicic acid (20). If 
this is true, it would seem that synthetic preparations of these indi- 
vidual gels should have absorptive capacities of the same order as 
the ultra clays. Four of these gels were therefore prepared and their 
absorptions tested. 
An aluminum silicate gel was made from sodium silicate and 
aluminum chloride mixed in the proper proportions for forming 
kaolinite (AljOo. 2Si02. 2H2O) ; the ferric hydroxide was prepared 
from feri'ic chloride and sodium hydroxide; the aluminum hydroxide 
from aluminum chloride and sodium hydroxide; and the silicic acid 
from sodium silicate and sulphuric acid. The gels thus formed were 
repeatedly washed with large amounts of distilled water until prac- 
tically free from soluble salts. The wash waters were removed 
through Pasteur-Chamberland filters, using suction, and the gels, 
after being concentrated by air-drying, were ground to pass a 200- 
mesh sieve. 
For comparison with these freshly prepared ^els, a sample of 
silica gel was procured from a commercial chemical company and 
samples of alumina and iron gels from the Chemical Warfare Service. 
The absorptive capacities of these products all ground to pass a 
200-mesh sieve are shown in Table 5. Included in this table also 
are the absorptions of a peat soil and of an extract from it con- 
taining no particles above 1 micron. The peat soil contained very 
little mineral matter, so these samples are probably fairly repre- 
sentative of the organic colloids in soils. 
