ESTIMATION OF COLLOIDAL MATERIAL IX SOILS. t> 
"berg evidently labors under the same misconception as Schloesing in 
believing that the material which remains longest in suspension 
differs essentially from material which does not remain so long in 
suspension. 
The chief object of much of the earlier work, in which colloidal 
material was isolated from the other soil constituents, was not to 
make a determination of all the colloidal material present in the soil, 
but merely to separate the soil into different-sized particles. It was, 
however, clearly recognized by Hilgard and Williams that one of the 
chief problems in such a separation was the disintegration of lumps 
of material into the individual particles of which they were formed. 
The basis of the separation was entirely one of size. 
As a matter of fact, under normal moisture conditions probably no 
soil contains more than a trace of colloidal material in the sol con- 
dition. The colloidal .material must be virtually all in a coagulated 
or gel form, since there is not sufficient water present to form a 
disperse medium. The problem of separating the colloidal material 
by a method of mechanical analysis, therefore, becomes largely one 
of converting the colloidal material from the gel to the sol form. 
Since a considerable part of the gels in most soils is in a more or less 
indurated condition that is difficultly dispersable, it might be expected 
that the exact character of the suspension obtained, its stability, and 
the size of the particles would be a more or less fortuitous matter, 
depending on the means of dispersion employed. This seems especially 
probable from a consideration of the well-known coagulating, peptiz- 
ing, and stabilizing effects of various concentrations of different 
ions. 
From this point of view, the ideas of Schloesing and Ehrenberg, that 
the particular preparations they obtained represented the only 
truly colloidal matter in the soil, appear illogical. It is of course 
true that size is one of the most dependable criteria for colloids, but 
in the case of many preparations the question arises, size of what? 
In the case of a coagulated suspensoid colloid, such as colloidal gold, 
the size of the floccules is not the criterion, but the size of the indi- 
vidual particles that are coagulated. But in the case of coagulated 
emulsoid colloids or of partially dehydrated emulsoid gels, structure 
or size of the molecular aggregates is evidently the criterion, and 
size in this case can not be determined by direct observation. 
In recent years a number of rapid methods have been proposed for 
estimating the comparative amounts of colloids in soils. Many of 
these methods are based on the fundamental misconception that by 
some simple treatment of the soil, such as by mere agitation with 
water, all the colloidal matter, or some definite part of it, is at once 
brought into suspension. Among such methods may be mentioned 
the method of Sokol 1 35 ) based on the relative volumes occupied by 
the colloid when first it is flocculated and when it has settled for 24 
hours: and the method of Scales and Marsh (30) based on the turbidity 
of the suspension obtained on shaking the soil for five minutes with 
distilled water. 
These methods fail to take into account that the quantity of 
colloidal matter brought into suspension by mere agitation of a" soil 
with water may vary greatly according to a transitory condition of 
the soil. Our experience shows that some soils yield practically no 
colloidal matter on the first agitation with water^ but after decanta- 
