ESTIMATION OF COLLOIDAL MATERIAL IX SOILS. 39 
The errors of sampling and of probable alteration of the colloidal 
material vary in magnitude in different soils. They are additive 
as a rule in the case of dye adsorption and tend to balance each other 
in the case of water or ammonia adsorption. 
The accuracy of the adsorption method is checked by comparing 
in the case of six soils the quantities of colloid shown by this method 
with the quantities shown by a gravimetric and microscopical method. 
When the adsorption ratios are corrected for possible alteration in 
adsorptive capacity of the colloid produced by extraction, the per- 
centages of colloidal material indicated by the adsorptions of mala- 
chite green, water, and ammonia agree well in nearly ad cases with 
the percentages of colloid determined gravimetrically and microsco- 
pically. When no correction is applied for alteration in adsorptive 
capacity of the colloid, the ratio obtained by the adsorption of watei 
in most cases is more accurate than rhe dye or ammonia ratio, and 
approaches the true value closely. 
A general procedure is suggested for estimating the quantit 
colloidal material present in a soil from the relative adsorptive 
capacities of the soil and of colloidal material extracted from the soil. 
It is pointed out that the wide discrepancy between the estimates 
of the colloid contents of soils given out by various investigators is 
due in part to differences in definition of "soil colloids*' and in part 
to faulty methods. The 32 soils selected for this investigation con- 
tained from 6 to 70 per cent of colloidal material as defined in this 
paper. Placing the maximum quantity of colloidal material in soils 
at 2 per cent is., obviously wrong, even if soil colloids are denned to 
include only that material which is dispersed into particles of sub- 
microscopic size. As much a> 38 per cent of such material has 
actually been isolated from a heavy soil. The probable reason is 
given for the low results obtained by certain investigators. 
LITERATURE CITED. 
1 Anderson, M. >.. Fry, W. H., Gile. P. L., Middleton, H. E.. and Robinson. 
W. 0. 
1922. Absorption by colloidal and noncolloidal soil constituents. U. >. Dept. 
Agr., Professional Paper, Bui. 1122, 20 p. 
2 Ashley. H. E. 
1909. The colloidal matter of clav and its measurement. U. S. t reological Sur- 
vey Bui. 388, 65 p., 1 pi. 
'3 Atterberg, A. 
1914. Die Eigenschaften der Bodenkomer und die Plastizitat der Boden. In 
Kolloidchemische Beihefte, Bd. 6, Heft 2-3. p. 55-89. 3 tig. 
4 Beaumont, A. B. 
1919 Studies in the reversibility of the colloidal condition of soils. Cornell 
Univ. Agr. Expt. Sta., Memoir 21, 479-524 p. 
5 Bemmelen, J. M. vax 
- 1910. Die Absorption. XI, 545 p., 1 pi. icol.), Dresden. 
6 Davi>. X. B. 
11916. The plasticity of clav and its relation to mode of origin. In Trans. Am. 
Inst. Min. Enein.. v. 51, p. 451-480, 4 figs. 
- a -. -is. R. O. E. 
1922. The interpretation of mechanical analysis of soils 
colloids. In Jour. Am. Soc. Agron., v. 14. no. 8, p. 2 -2 - 
(8 Ehrenberg, P. 
1918. Die Bodenkolloide. Aufl. 2. VIII. 717 p., illus., Dresden und Leipzig. 
I letcher, C. C, and Bryan, H. 
1912. Modification of the method of mechanical soil analysis. !". S. D. A. 
Bureau of Soils Bui. 84, 16 p., 7 pi. 
