June 9, 1923 
Microscopic Estimation of Soil Colloids 
881 
to cleave into oblong particles v/ith the greater dimensions several times 
the magnitude of the smaller, thus presenting a widely different shape 
from the colloidal aggregates with which it was to be compared. Two 
samples were prepared consisting of 0.3 gm. colloid to 1.7 gm. hornblende 
and 0.4 gm. colloid to i.6 gm. hornblende. The first of these was mixed 
dry and the second wet, as with the Marshall colloid-quartz samples. 
The results of the microscopic examination of these samples are given 
in Table I. 
Table I .—Microscopic estimation of colloids in synthetic samples 
Sample. 
Colloid 
present. 
Colloid 
observed. 
Marshall colloid and quartz, mixed dry. 
Per cent. 
60 
40 
15 
20 
Per cent. 
57 
45 
12 
23 
Marshall colloid and quartz, mixed wet. 
Orangeburg colloid and hornblende, mixed dry. 
Orangeburg colloid and hornblende, mixed wet. 
These results showed that the colloid in the fine and coarse residues of 
the soils could be estimated with fair accuracy by this microscopic method. 
Microscopic examinations were therefore made of the fine and coarse 
residues of eight soils which had been subjected to the repeated washing 
and rubbing process. The results are given in Table II. The quantity 
of the unextracted colloid in the fine and coarse residues, as determined 
by this microscopic method, are given in column 3. The additional 
data in Table II are given for the purpose of showing the relative amounts 
of “extractable’' and “unextractable” colloids in the soils. 
From Table II it will be seen that from 9.4 to 42.4 per cent of the soil 
was made up of colloids extractable by the methods employed. The 
microscopic examination of the residues shows that from 25 to 97 per 
cent of the fine residues and 2 to 25 per cent of the coarse residues 
were colloidal aggregates not extractable by the methods employed. 
These results throw considerable doubt upon the results of several 
investigators who have determined the colloidal contents of soils and 
clays by purely mechanical methods. Hissink^ and Sven Oden,® for 
example, have dispersed colloids in soils by very elaborate washing 
and rubbing processes. It is of course possible that they may have 
effected complete dispersion by their methods, but apparently no inves¬ 
tigations were undertaken to ascertain whether, or to what extent, 
aggregates of undispersed colloids remained in their soil suspensions. 
Williams,® in the course of a very elaborate and painstaking separation 
of colloids from soils by washing, rubbing and boiling, did examine his 
residues microscopically. Details of his microscopic method are not 
given, but it is apparent that he used an ordinary chemical microscope 
and mounted his residues either in air or water. Owing to the wide 
differences in refractive indices between these mounting media and the 
ordinary soil minerals, scarcely more than surface phenomena could be 
* Hissink, D. J. di® mbthode der mechanischen bodenanalyse. In Interaat. Mitt. Bodenkunde, 
Bd II, p. i-ii. 1921. 
5 Od6n, Sven, uber die vorbehandltjng der bodenproben zur mechanischen analyse. In Bui. 
Geological Inst., Univ. Upsala, v. i6, p. 125-134. 1918-19. Bibliographical footnotes. 
8 Williams, W. R. untersuchung uber die mechanische bodenanalyse. In Forsch. Geb. Agr.- 
Phys. (Wollny), Bd. 18, p. 225-350. 1895. Bibliographical footnotes. 
