THE MICROSCOPIC ESTIMATION OE COLLOIDS IN 
SOIL SEPARATES* 
By William H. Fry 
Scientist in Soil Laboratory Investigations, Bureau of Soils, U S. 'Department of 
Agriculture, 
In a recent work upon soil colloids,^ in which a description of the soil 
materials classified as cdlloids was given, it became desirable to compare 
the values obtained by an adsorption method for determining the col¬ 
loidal content of soils with values obtained by some method independent 
of adsorption phenomena. The most satisfactory method for this pur¬ 
pose would be a mechanical separation of colloidal from noncolloidal 
soil materials. But it has been shown in previous publications that 
a mechanical analysis probably does not effect a complete separation. 
However, it seemed possible that colloidal matter not separated from the 
soil by mechanical analysis might be estimated by microscopical means- 
Then the sum of the colloid extracted and that estimated microscopi¬ 
cally would give a value for the total colloid in the soil by methods quite 
independent of adsorption. 
For the purpose of separating the colloids from the coarser particles 
of the soils the samples were repeatedly treated by agitation with dis¬ 
tilled water containing a trace of ammonia, rubbed witii a rubber pestle, 
and the supernatant liquid decanted and centrifuged. From 40 to 60 
such treatments yielded practically all of the colloid extractable by this 
method. For convenience of manipulation the soils were divided, during 
the washing and rubbing process, into the colloidal fraction made up of 
particles less than 0.001 mm. in diameter, a fine fraction made up of 
particles ranging approximately from 0.001 to 0.050 mm. in diameter, 
and a coarse fraction consisting of particles larger than about 0.050 mm.^ 
The adsorptive capacities of these fine and coarse residues indicated 
that they contained colloidal material.^ Observations of the residues 
with the ultramicroscope showed that the mineral particles had been 
fairly well cleaned of adhering colloidal matter by the washing and 
rubbing, and that the colloidal material remaining in these fractions 
was in the form of lumps or aggregates made up, at least superficially, 
of very large numbers of particles less than 0.001 mm. diameter. But 
since the light coming to the eye from the ultramicroscope is reflected 
from the stmaces of the particles under examination it was possible that 
the colloidal aggregates thus observed were simply mineral grains com¬ 
pletely coated with colloids or that, at least, the colloidal aggregates 
^ Accepted for publication Apr. 7,19*3. 
* Andbrsch^, M. S., Fry, W. H., Gilb, P. L.. Middlbton, H. E., and Robsstson. W. O. absorption 
BY COLLOIDAL AND NONCOLLOIDAL SOIL coNSTiTUBNTs. U. S. Dept. Agr. Bul. iiaa, ao p. X933. Litera¬ 
ture cited, p. igr-ao. 
» Davis, R. O. E- thb interpretation op mechanical an^ysis op soils as appected by soil col¬ 
loids. In Jour. Amer. Soc. Agron., v. 14, p. a93“»98. 1933- Literature cited, p. a98. 
Journal of Agricultural Research. Vol. XXIV, No.xo 
Washington. D. C. June 9, 1933 
acv Key No. H-‘>' 
(879) 
