THE CHEMICAL COMPOSITION OF SOIL COLLOIDS 27 
Although it seems improbable that the colloidal material contains 
any large amount of the common soil-forming minerals, it is like- 
wise improbable that the colloid should be entirely free of such 
material. In fact, in immature soils where there is an abundant 
supply of larger mineral particles other than quartz, as in certain 
glaciated and alluvial soils, there must be some small undecomposed 
particles of colloidal dimensions unless the decomposition process 
is instantaneous when the partiele is reduced to a certain size. 21 
Some of the' ordinary soil minerals, such as secondary quartz and 
magnetite, which may be forming in the soil, would be of colloidal 
size in some stages of development and still possess their character- 
istic mineralogical properties. 
It thus appears that the amount of undecomposed primary soil- 
forming minerals in soil colloids is generally small and probably 
variable, and that such particles are of a transitory character and 
are probably entirely lacking in very old or mature soils. 
ATTEMPTED FRACTIONATION OF COLLOIDAL MATERIAL 
By precijntation. — If the colloidal material consists of particles 
which differ from each other in physical properties or chemical 
composition, it would seem that some separation of the different 
kinds of particles could be made by fractional precipitation. For 
this purpose heavy dispersions of Cecil clay loam and Iredell clay 
colloids were evaporated to about one-fourth of the initial volume 
in a current of air at room temperature. By this process a large 
part of the colloidal matter was coagulated and fell to the bottom 
of the container, and a part of the colloid remained in suspension. 
The colloid which remained in suspension was poured off and col- 
lected on a baked-clay filter. Part of the collected suspension was 
analyzed and a part was again fractionated. The material remain- 
ing in suspension after the second evaporation was also analyzed. 
The compositions of the first and second suspended fractions agreed 
within the limits of analytical error with the composition of the 
original colloid. Therefore, no separation which affected the chemi- 
cal composition was obtained by this method, and no further at- 
tempts were made to fractionate the colloids from other soils. 
By freezing and thawing,- — Freezing and thawing was found by 
Czermak (9) to decrease the absorptive capacity of the soil. The 
process of freezing and thawing was tried on two of the colloids 
here reported with the idea that the} 7 might be separated into simpler 
constituents. The freezing and thawing of the colloidal suspension, 
however, resulted in an almost complete coagulation. The quantity 
of colloidal matter remaining in suspension was so small that little 
could be learned of its identity, and it seemed impracticable to make 
a separation in this way. 
Although the freezing and thawing process does not appear effec- 
tive in fractionating the colloid, it probably does affect the solubility 
of certain constituents or the tenacity with which they are held by 
the colloid. For example, a sample of Cecil clay loam soil, after 
freezing and thawing five times, yielded a water extract containing 
213 parts per million total solids compared with 180 parts per million 
yielded by the untreated sample; and a sample of Chester loam soil. 
-This possibility has been pointed out by Whitney (}6). 
