2 BULLETIN 1311, U. S. DEPARTMENT OF AGRICULTURE 
viously desirable. The use of a high-power centrifuge has enabled 
us to isolate from a large number of different soils material exhibit- 
ing colloidal properties to a marked degree. It has thus been pos- 
sible to make chemical analyses of purely colloidal material which 
has been separated from the noncolloidal part of the soil. These 
chemical analyses and data based on them are discussed in the fol- 
lowing pages. 
Under normal conditions very little of the soil colloidal material 
is present as a sol; practically all of it is present as a gel or as what 
Freundlich terms "coagel" (15, p. 905). 1 Before the colloidal ma- 
terial can be separated from the larger, noncolloidal particles of the 
soil and analyzed it must be brought into the suspended or sol con- 
dition. This bulletin therefore deals almost entirely with the com- 
position of the colloidal material which can be brought into a finely 
dispersed condition in water without alteration of the discrete non- 
colloidal particles. It is recognized that there may be present in 
some soils, besides colloidal material capable of forming a suspen- 
sion, coarse particles which are colloidal in the sense that opal and 
coal are colloidal. These materials are not considered here. The 
soil colloidal material may be considered as essentially a mechanical 
separate of the soil, similar to the clay fraction, except that the 
upper size limit of the colloid fraction is smaller than that of the 
clay fraction. The largest particles in the colloidal material 
analyzed were apparently less than 0.3 micron in size. 
REVIEW OF EARLIER WORK 
ULTIMATE CHEMICAL COMPOSITION 
Previous knowledge of the chemical composition of the earthy col- 
loidal matter has been contributed by the soil investigator, the geol- 
ogist, and the workers in the ceramic industries. 
Considerable knowledge of the composition of soil colloids can be 
gotten from the composition of the finest fractions that are obtained 
in the mechanical analysis of soils. The clay fraction of the Ameri- 
can system of mechanical analysis (14) may not be entirely colloidal, 
but it is nearly so, and the " Schlamm " fraction of the European 
workers is probably nearly all colloidal. Previous work on the com- 
position of these finest mechanical fractions will be considered in 
detail. 
Failyer, Smith, and Wade (13) studied the partial composition of 
the sand, silt, and clay separates and found that the finest soil sepa- 
rates—that is, the more nearly colloidal — were richer in potash, lime, 
and phosphoric acid than the coarser separates. Loughridge, cited by 
Hilgard (22, p. 385) and Tolman (44) have reported some work on 
the composition of clay separates as determined by digestion in 
strong acids. In every case they found the clay separates were higher 
in soluble iron and aluminum than the whole soil. 
A few chemical analyses for the total constituents of the finest soil 
separates obtained by several other investigators are available in the 
literature. 2 These analyses are given in Table I. 
1 Reference is made by number {italic) to " Literature cited," p. 39. 
3 A number of analyses by acid digestion of fine soil separates of colloidal dimensions 
are to be found in the literature, but on account of the difficulty of comparing such 
analyses for the total constituents and also on account of the difficulty of interpreting 
the figures for the " insoluble residue," the acid digestion analyses are not giveih in the 
table. 
