ABSORPTION BY SOIL CONSTITUENTS. 17 
determining the colloidal content of a soil obviously would not be 
affected appreciably by the noncolloidal constituents. One of the 
fundamental objections against absorptive methods of estimating 
the colloids is thus removed. It is not intended to imply, however, 
that even relative colloidal contents of different soils are indicated by 
the mere absorptive capacities of the soils for such substances as 
malachite green or ammonia; since the colloids from different soils 
absorb widely varying amounts of these substances. 
The almost complete localization of absorption in the colloidal 
matter probably is due primarily to the colloidal material being of a 
quite different nature from the larger soil particles. While the non- 
colloidal part of the soil is made up of unaltered soil minerals, the 
colloidal material is doubtless made up chiefly of inorganic gels, the 
insoluble decomposition product of soil minerals, together with 
amorphous organic matter. 
The results on the noncolloidal absorption of soils are evidently at 
variance with some of the older, but still prevalent, concepts regard- 
ing soil absorption. It has been generall}' recognized that the organic 
matter constitutes an especially absorptive part of the soil, and the old 
view was that the rest of the absorption was due to mineral particles 
of all degrees of fineness, the finer the particles the more absorptive 
the soil. Absorption was thus regarded as proportional to the sur- 
face exposed, and the sm-face. except for the organic matter, was con- 
sidered as beino^ almost entirely that of mineral particles. The more 
highly absorptive soils were supposed to^contain some especially ab- 
sorptive minerals called zeolites. It is now well recognized that 
zeolites, if present at all in soils, can be there only in very small 
amounts. 
It is now coming to be realized that the highly absorptive inorganic 
material in soils is not the definite crystalline zeolites, but amor- 
phous material such as ferric hydroxide, aluminum hydroxide, silicic 
acid, and aluminum silicate gels. The data here presented tend to 
confirm the more modern view of soil absorption, and gives for the 
first time a quantitative estimate of the amount of absorption due 
to the coUoidal material and the amount due to the noncolloidal 
particles. 
SUMMARY. 
It was impossible to determine directly the amount of absorption 
in soils due to colloids and noncoUoids from a fractionation of the 
soil into these two classes of material, since we could not separate 
all the colloidal matter from the finer mineral particles. 
Since the noncolloidal part of soils is made up of minerals of 
different degrees of fineness, it should be possible to calculate the 
magnitude of the noncolloidal absorption in the soil from the ab- 
sorptive capacities of minerals powdered to definite sizes. 
Each mineral has a characteristic absorptive capacit}^ for different 
substances. Several of the most abundant minerals in soils, such 
as quartz and the feldspars, absorb practically no malachite green, 
water or ammonia. 
It was calculated from the absorption of mineral powders and the 
mineralogical compositions of soils that as a rule less than 5 per 
cent of the total absorption of the soil is due to the noncolloidal 
part. In the case of the most highly micaceous soils, however, the 
