12 BULLETIN 1193, U. S. DEPARTMENT OF AGRICULTURE. 
capacity of a soil by 0.298 which was the average water-adsorptive 
capacity of the different colloids tested (Robinson, 26). It should 
be noted, however, that this relative constancy in the water-adsorp- 
tion values of the different colloids may well hold only for the par- 
ticular conditions of the method used. ~ Preliminary work mdicates 
that greater variations may be expected to occur in the water ad- 
sorption values of the different colloids when the materials are in 
equilibrium with the vapor pressure of water afforded bv 10 per 
cent sulphuric acid. 
There is only a general parallelism between the quantities of mala- 
chite green, water vapor, and ammonia adsorbed by the different 
colloidal materials. The colloids of the Sharkey soil and of the 
Stockton, Wabash. Carrington, and Marshall soils and subsoils fall into 
one group, the least adsorptive member of which has an appreciably 
higher adsorptive capacity for dye or for ammonia than any of the 
other colloids. However, many of the other samples of colloidal 
material differ widely in their adsorptive capacities for these two sub- 
stances. The much smaller variations in the adsorptive capacities 
for water vapor show practically no correspondence to the varia- 
tions in the adsorptive capacities for dye or for ammonia. 
This lack of correspondence in the adsorptive capacities of the 
colloids for malachite green, water, and ammonia is similar to that- 
observed in the adsorptions by mineral powders and by synthetic 
inorganic gels, reported in a previous publication (Anderson, 1). 
The silica gel, for example, which was highly adsorptive of ammonia, 
adsorbed practically no dye. Thus far we have obtained very little 
indication of any correspondence between the empirical chemical 
compositions of the colloidal materials and their adsorptive capacities 
for the three substances. 
It is an interesting fact that the colloid of nearly every subsoil 
has an adsorptive capacity for malachite green, water or' ammonia 
which is almost the same as that of the corresponding surface soil. 
The only marked exceptions to this generalization occur in the 
adsorption of dye by the Clarksville, Orangeburg, and Wabash 
colloids and in the adsorption of water by the Ontario colloid. This 
similarity in adsorptive capacities would indicate that the colloidal 
material in the subsoil is in most cases very nearly the same as that 
occurring in the surface soil. The adsorption evidence is particu- 
larly strong on this point, inasmuch as the adsorptive capacity of 
any one colloid for malachite green shows little correspondence to its 
adsorptive capacity for water or ammonia. The surface and sub- 
soil colloids also seem to be very nearly alike in their ultimate chemical 
composition. (Robinson and Holmes, 27.) 
PERCENTAGES OP COLLOIDAL MATERIAL INDICATED BY THE ADSORPTION METHOD. 
As already pointed out, the relative adsorptive capacities of the soil 
and of the corresponding colloid should indicate the quantity of colloi- 
dal material present in the soil, inasmuch as practically all the adsorp- 
tive capacity of a soil seems to be due to the colloidal material. From 
the data given in Table 1 the adsorption ratios 
adsorption per gram soil 
adsorption per gram colloid 
