2 BULLETIN 1193, U. S. DEPARTMENT OF AGRICULTURE. 
and by the addition to the soil of various substances, such as lime r 
fertilizers, and manures, are to be attributed to the influence of these 
treatments on the condition of the colloidal matter present (Ehren- 
berg, 8). In order properly to evaluate the part played by colloids 
in various soil phenomena, a method for determining the total quan- 
tity of colloidal material present is greatly needed. 
COLLOIDAL AND NONCOLLOIDAL SOIL PARTICLES. 
A fairly sharp distinction can be made between material in true 
solution and that in colloidal suspension, but a sharp demarcation 
between colloidal material and that in a coarser state of subdivision 
is bound to be more or less arbitrary. The negative characteristics 
of colloidal material, slight influence on the boiling or freezing point 
of the dispersing medium, and hardly sensible osmotic pressure 
serve to differentiate colloidal suspensions from true solutions. The 
more positive characteristics of colloidal material, those associated 
with small size, such as the property of remaining in suspension in the 
dispersing medium, Brownian movement, and high adsorptive 
capacity are rather unsatisfactory criteria for a rigid separation of 
colloidal and larger particles. 
The criterion of colloidal condition most generally adopted is that 
of size of the dispersed particles, the upper limit of colloidal particles 
usually being set at 0.1 micron. 2 The selection of exactly 0.1 micron 
as the upper limit of colloidal size is of course purely arbitrary. 
There are no grounds for assuming that the adsorptive power, Brown- 
ian movement, or reactivity of dispersed particles shows a sharp 
break at this particular point. Moreover, in the case of emulsoid 
colloids in the gel condition, the classification on the basis of size 
is not determinate. The division at 0.1 micron does, however, throw 
the colloidal particles into the submicroscopic region. 
In the case of soil constituents, also, a rigid distinction can not be 
made between colloidal and noncolloidal material. Nevertheless, 
there will probably be little disagreement concerning the colloidality 
of most of the soil material which is classified as colloidal in this paper. 
In the investigation described here all the organic matter of the 
soil is regarded as colloidal. A considerable part of the organic 
matter in normal soils can be readily peptized or dispersed into par- 
ticles 0.3 micron or less in diameter, and the part of the organic 
matter which is not so readily dispersable — the less humified material, 
such as partially decomposed roots and plant residues — is doubtless 
just as strictly colloidal as the material in a more advanced stage of 
decomposition. 3 The celluloses and structural plant substances, as 
well as most of the intracellular material, is generally conceded to be 
of a colloidal nature, although in the dried state these substances 
can not readily be converted into the sol condition. 
In this investigation all inorganic material was classified as colloid 
which could be dispersed into particles less than 1 micron in diameter 
without subjecting the soil to a drastic chemical or physical treat- 
ment that would disintegrate the mineral particles. One micron 
2 Freundlich (10, p. 2) selects 0.5 micron as the upper limit. On the other hand, many soil investigator! 
have used 2 microns as the dividing line between colloid and noncolloid (Atterberg, S). 
» An unextracted peat soil had just as high an adsorptive capacity for water as the material extracted 
from it, which was composed of particies less than 0.3 micron in diameter. The extracted colloid was 
slightly less adsorptive of ammonia and more adsorptive of malachite green than the unextracted peat. 
