38 BULLETIN 1311, U. S. DEPARTMENT OF AGRICULTURE 
(6) The stoicliiometrical calculations showed that all the silica, 
alumina, and iron oxide in soil colloids could not be present as 
hydrated silicates of the composition of kaolinite and nontronite. 
Some suggestive relations between the color and the calculated ex- 
cesses of free iron oxide in the various colloids were brought out. 
but on the whole the stoicliiometrical calculations were not convinc- 
ing enough to disprove or to establish the presence of the compounds 
assumed in the calculations. 
The evidence presented shows that the soil colloids studied behave 
as very intimate mixtures and that there is little tendency to break 
up into compounds of definite composition. There is, however, a 
growing belief among colloid chemists that many colloidal substances 
are ultimately composed of small crystals, which are probably defi- 
nite compounds. This belief, founded on theoretical considerations, 
has been strengthened by the identification by X-ray analysis of 
crystals in such pure colloids as colloidal gold, colloidal silver, 
silicic acid, ramie (36), chalcedony flint (45), and limonite (25). 
With a further development of X-ray methods it may be possible 
to decide definitely if such complex mixtures as soil colloids are 
composed ultimately of crystals, and what compounds, if any, are 
present. It seems probable that the soil colloidal matter contains 
some minerals like those constituting larger soil particles, and also 
ferric oxide, probably hydrous. No strong evidence, however, has 
been obtained for the presence of other compounds. 
SUMMARY 
This bulletin deals with the composition of soil colloidal matter 
that can be brought into a finely dispersed condition in water. 
Analyses are given of the colloidal matter isolated from 45 soils, 
which represent important agricultural types in the United States 
and cover wide ranges of composition, but do not include such 
extremes as peats and laterites. 
The colloidal matter is composed mainly of silica, alumina, iron 
oxide, and water, with smaller amounts of lime, magnesia, potash, 
soda, phosphorus, manganese, sulphur, chlorine, and organic matter. 
There is a rather wide variation in the proportions of these con- 
stituents present in different colloids, though a considerable number 
of the colloids show almost a constant composition. In general, the 
sum of the lime, magnesia, potash, and soda is low when the silica 
is low and high when the silica is high. Silica and alumina usually 
vary inversely. 
The colloidal matter from the soil and corresponding subsoil are 
much alike in composition. 
The composition of the colloidal matter differs from that of the 
whole soil in being high in alumina, iron, water of combination, 
organic matter, magnesia, phosphorus, and sulphur and lower in 
silica. There is the same general difference in composition between 
the colloidal matter and the coarser mineral particles that there is 
between the colloidal matter and the soil, but the difference is more 
pronounced. 
The part of collodial matter which is most readily dispersed is 
fairly representative in composition of all that it is possible to 
isolate by the means employed. In some cases, however, the col- 
