34 BULLETIX 1193, U. S. DEPARTMENT OF AGRICULTURE. 
acters. Mellor (W) gives the content of colloidal clay, "l'argiie col- 
loidale." in Cornish china clay and Devonshire ball clays as being 
probably something more than 0.005 and 0.05 per cent, respectively. 16 
The data given in the preceding pages show that the total quantity 
of colloidal material in soils is far greater than many investigators 
had supposed was present. It is more in accord with the quantities 
of the Schlamm fraction obtained by Williams (39) and the "clay" 
fraction obtained by Hilgard (15). 
Naturally views concerning the quantity of colloidal material in 
soils will vary according to differences of opinion regarding just 
what soil materials should be classed as colloidal. The older view 
that the inorganic material in soils consisted of definite mineral 
particles of all degrees of fineness has lately been modified somewhat. 
The more general view at the present time is that the mineral par- 
ticles are coated with a film of colloidal, gelatinous material. Even 
at the present time, however, the prevalent idea seems to be that 
in the finer soil fractions mineral fragments of quartz, feldspar, etc., 
predominate. For instance, Sven Oden (#4> p. 328) states that: 
"The true soil colloids are the innumerable fragments of both weath- 
ered and unweathered minerals, crystal chips and amorphous sub- 
stances, which, in a fine state of division, constitute the clay." If 
the colloidal material in soils were made up entirely of small crys- 
talline particles of such minerals as form the coarser soil particles, 
there would probably be little difference of opinion regarding the 
quantity of such material in soils, once the upper limit of colloidal 
size was agreed upon. The colloid could probably be determined 
by relatively simple methods of mechanical analysis, since the ulti- 
mate particles would be definite and would doubtless readily de-floc- 
culate. 
Three different lines of evidence, however, indicate that the 
colloidal material in soils is not composed chiefly of fragments of min- 
erals, at least of such minerals as constitute the larger soil particles. 
It is apparent from the results given in this paper that merely shaking 
with water deflocculates only a small amount of colloidal material 
and that about 50 per cent of the colloidal material in some soils 
is very resistant to deflocculation or peptization. Fragments of com- 
mon soil minerals probably would deflocculate readily. Also, adsorp- 
tion data given in a previous bulletin indicate that the larger part 
of the colloidal material in soils is not made up of major soil minerals 
but rather of substances in a gel-like condition. Finally, chemical 
analyses of colloidal material extracted from some 40 different soils 
show that chemically the colloidal material is quite different from 
the larger soil particles (Robinson and Holmes, 27). 
There is every reason to suppose, as Van Bemmelen (5), Ehren- 
berg (8), Wiegner (38), Rohland (.28), and others have previously 
pointed out, that the inorganic colloidal material in soils is similar in 
nature to artificially prepared gels of silicic acid, aluminum hydroxide, 
or iron hydroxide. Such materials are, of course, colloidally dis- 
persed when sufficient water is present, but when they are in a par- 
tially dry condition the silica, alumina, and iron, may be regarded 
as constituting the dispersion medium and water or air the dispersed 
phase; in other words, the particles are colloidal because of the large 
16 However, Mellor points out as a significant fact that when these quantities of colloidal material arc 
extracted from the clays no detectable difference in plasticity is noted. 
