THE CHEMICAL COMPOSITION OF SOIL COLLOIDS 5 
The fractions analyzed by the different investigators were without 
doubt all made up chiefly of colloidal matter, even if the upper limit 
of colloidal size is assumed to be somewhat smaller than that of the 
largest particle included in the fractions. 3 The fractions in question 
were therefore probably all similar, although the upper limit of size 
varies somewhat. The largest particles in the fractions obtained by 
Schloesing and by Novak and Smolik were probably considerably 
less than 1 micron and were in this respect more nearly like the 
material analyzed by the authors. 
Table 1 shows that the fine soil fractions are composed essentially 
of silica, alumina, " combined " water, or an unduly large proportion 
of organic matter, and generally iron oxide. The silica content 
does not vary greatly and is very much the same as that of the kaolin 
fractions analyzed by Schloesing. The alumina varies from 14.26 
to 39.30 per cent; the majority of the figures, however, being grouped 
around 30 per cent. Judging from these analyses alone, one would 
conclude that the colloidal matter of soils and certain ceramic clays 
(which are really soil material) is relatively constant in silica, but 
ratay vary widely in the amount of other elements present, although 
the alumina does not vary greatly in the majority of cases. 
It is significant that, in spite of the relatively wide variations in 
iron, alumina, etc., shown by certain samples, the colloidal materials 
from the three soils of widely different textures analyzed by Puchner 
show but little variation in any of the major elements. It is prob- 
able that these soils were from a restricted area, and this may have 
considerable to do with the constancy of composition of the colloid. 
The relatively small variations in the composition of the fine frac- 
tions within the series of 8 soils from the Rothamsted plots, the 
series of 9 soils from southeastern England, and the series of 10 
soils from northern Wales are further indications of the possible 
relation of the composition of soil colloids to conditions prevailing 
within restricted areas. The same point is further brought out by 
the small variation found by Blanck (5), in the fine fractions of 
seven soils taken near Breslau, Germany. These fine fractions con- 
tained from 60.75 to 68.31 per cent of silica and from 27.35 to 37.13 
per cent of the oxides of iron and aluminum. 
COMPOUNDS PRESENT 
Some of the earlier investigators give evidence concerning the 
way the constituents in the colloidal material are combined. 
Schloesing (37) examined three fine kaolins, analyzing the frac- 
tions which settled with different velocities in distilled water to 
which a few drops of ammonia had been added. Taking one clay 
as typical, the first three fractions which settled most rapidly and 
which comprised over 95 per cent of the whole mass had the com- 
position of kaolinite. AL,O s .2SiO.,.2H 2 0. The sixth fraction, about 
1.5 per cent of the whole, remained .suspended for 37 days. The 
analysis of this material is given in Table 1. It was slightly higher 
in silica and lower in alumina and water than kaolinite, and was 
regarded by Schloesing as the truly colloidal part of the kaolin. 
From these results one would conclude that the colloidal matter iso- 
3 In the usual system of mechauical analysis aggregates of particles less than 1 micron 
in diameter appear as larger particles. 
