ABSOBPTION BY SOIL CONSTITUENTS. 13 
Although the mineral content assumed in this case is about the 
average of a large number of soils, most soils do not contain this 
amount of muscovite. The mica average of 5 per cent is due to the 
very high mica content of a few important soil types. Usually the 
muscovite content is below 2 per cent. If the hypothetical soil just 
considered had contained 2 per cent of muscovite, only approximately 
1 per cent of the total absorption would have been due to the crystal- 
line minerals. If, however, the muscovite content had been as high 
as 37 per cent, which has been reported for one of our most highly mi- 
caceous soils, the mineral absorption values might have been as high 
as 7 to 20 per cent of the total soil absorption, the dye absorption 
being most affected by the mica. 
In the case of a peat soil, the total mineral matter is small in 
amount, so the absorption of the minerals would be negligible. The 
organic matter, as indicated before, may be considered essentially 
colloidal. 
EVIDENCE CONCERNING THE NATURE OF SOIL COLLOIDS AFFORDED 
BY ABSORPTION DATA. 
It is difficult to understand why so small a part of the absorptive 
capacity of the whole soil should reside in the noncolloidal constitu- 
ents unless the colloidal constituents differ from the noncolloidal in 
other respects than mere degree of subdivision. The organic matter 
of soils, which is classified as colloidal, is obviously of a different 
nature from the mineral particles; but in many soils organic matter 
forms only a small part of the total colloidal content. Other investi- 
gations in this bureau indicate that the larger part of the inorganic 
colloidal matter in soils is different from the noncolloidal mineral par- 
ticles in being similar to gels. This view that the soil colloids are 
mostly of a gel-like nature is also held by Van Bemmelen (5) , Rohland 
(22), Konig (13), Ehrenberg (9), Wiegner (25), and others. 
The fact that the mineral powders just above the colloidal size 
had such low absorptions as compared with the ultra clays is in 
harmony with the view that the two classes of material are of a 
different nature. However, since most of the colloidal particles 
are so small that their size can not be accurately determined micro- 
scopically, the possibilities would still remain, so far as microscopical 
observation is concerned, that the colloidal and noncolloidal parts 
of the soil are of the same nature and that the differences in absorp- 
tian are merely a function of the size of the particles, the colloidal 
particles being exceedingly fine as compared with the noncolloidal 
particles. 
It therefore became desirable to determine the absorptions of 
mineral powders ground as finely as possible and to compare the 
absorptions of these powders with those of the coarser powders and 
ultra clays. 
ABSORPTIVE CAPACITIES OF MINERAL PARTICLES BELOW 1 MICRON IN DIAMETER. 
Six minerals which showed but little tendency to flocculate when 
suspended in water were selected for grinding to colloidal dimensions. 
The minerals were ground dry in a steel ball mill — muscovite for 10 
days and the other minerals for about 3 days. The powders from the 
ball mill were then shaken with distilled water and the greater part 
