12 BULLETIN 1122, U. S. DEPAKTMENT OF AGEICULTUEE. 
cent, potash feldspars 7 per cent, muscovite 7 per cent. Miscella- 
neous minerals, no one of which is present in important amount, 
together with colloidal aggregates, make up the remaining 35 per 
cent of the group. ^*^ The quartz and feldspar mineral powders 
reported in Table 2 had no appreciable absorption for malachite 
green, water, or ammonia. It is therefore evident that on an aver- 
age at least 68 per cent of the mineral particles which are from 1 to 
50 microns in size would have practically no absorption. (This is 
assuming that the minerals in the clay group are essentially the same 
as those in the silt group.) If we take the average absorption of the 
other mineral powders as representing the absorption of the remain- 
ing 32 per cent of silt and clay particles, the average absorptions per 
gram of the unaltered minerals in the soil just above the colloidal 
size would be only 1.7 per cent, 1.1 per cent, and 0.8 per cent that of 
the average absorption of the ultra clays for malachite green, water, 
and ammonia, respectively. 
In the mechanical groups larger than silt, somewhat over 68 per 
cent of the mineral particles would probably be quartz and feldspars, 
since quartz is more abundant in the larger soil separates. Hence in 
calculating the absorption of mineral particles in the fractions larger 
than silt, it will be safer to assume that so far as their mineralogical 
composition is concerned their absorption will be less than that of 
particles of silt size. 
Particles of different sizes of any one crystalline mineral must have 
absorptive capacities that are to some degree inversely proportional 
to the size of the particles. On this basis the absorptive capacities 
of particles from 2 to .25 millimeters in diameter, included in the 
gravel, coarse sand, and medium sand groups, would be only about 
one-fortieth as great as that of particles of silt and clay size, and may, 
therefore, be disregarded. On the same basis, the finer mineral 
sands, with diameters from 0.05 to 0.25 millimeters would have ab- 
sorptive capacities one-sixth that of particles of the silt and clay size. 
The application of the above facts to a few hypothetical soils will 
make evident what part the noncolloidal soil particles may be ex- 
pected to play in the absorption of different types of soils. 
Let us assume a loam soil of about average mineral content made 
up of 45 per cent quartz, 5 per cent muscovite, 5 per cent feldspars, 15 
per cent other crystalUne minerals, and 30 per cent of colloid. The 
maximum amount of particles under 0.05 millimeter for a loam 
soil, according to the classification used in the Bureau of Soils, is 60 
per cent. We should then have 60 per cent of material smaller than 
0.05 millimeter and 40 per cent larger than 0.05 millimeter. The 
average size of the latter may well be taken at 0.5 millimeter. If we 
assume average absorptions for the colloidal matter and calculate 
the mineral values from Table 2, we should have a gram of the soil 
absorbing 0.0618 gram dye, 0.0914 gram water, and 0.0121 gram 
ammonia. The percentages of these total absorptions which were 
due to the minerals would be respectively 2.9 per cent, 2.2 per cent 
and 1 per cent. Most of the mineral absorption in this case would 
be due to mica. 
10 Recent work has shown that 15 per cent or more of the silt group may consist of colloidal aggregates 
when this group is separated by the method of mechanical analysis practiced in the Bureau of Soils. 
Assuming 15 per cent of colloidal aggregates in this group, the average composition of the crystaUine min- 
erals would be: Quartz 60 per cent, potash feldspars 8 per cent, muscovite 8 per cent, and miscellaneous 
minerals 24 per cent. 
