18 
BULLETIN 1311,, U. S. DEPARTMENT OF AGRICULTURE 
Table 4. — Average compositions of soils, colloids, and coarser mineral 
particles compared 
Si0 2 
AI2O3 
Fe 2 3 
CaO 
MgO 
KiO 
Na 2 
Com- 
bined 
H2O 
Colloidal matter ... . 
P.d. 
43.34 
87.2 
83.3 
P.d. 
26.84 
6.0 
7.3 
P.d. 
10.41 
1.9 
1.5 
P.d. 
1.04 
.5 
2.0 
P.d. 
1.72 
. 5 
1.7 
P.d. 
1.43 
1.5 
2.9 
P. d. 
0.38 
.9 
.9 
P.d. 
9.93 
Coarser mineral particles 
Mineral particles in fine sands and 
silt of a different series of 26 soils. . . 
.6 
.4 
The average composition of the coarser mineral particles are 
approximately alike, although they were calculated for two different 
series of soils by two different methods. 
It may be seen from the table that the colloidal matter averages 
half as high in silica as the mixture of coarser mineral particles. 
4 times as high in alumina, 5 times as high in iron, and about 20 times 
as high in combined water. 
In tracing the relation between the composition of the colloid and 
of the corresponding soil, it should be borne in mind that the compo- 
sition of the whole soil is a composite of the colloidal matter and the 
larger mineral particles. The extent of the difference in composition 
between the whole soil and the colloidal matter is, therefore, partly 
due to the degree of difference in composition of the colloids and 
minerals and partly due to the relative quantities in which these con- 
stituents are present. 
By comparing Tables 2 and 3 it will be seen that the colloids are 
lower in silica and higher in all other constituents than the corre- 
sponding soil. The concentration of organic matter, magnesia, phos- 
phoric acid, and sulphur in the extracted colloid is quite marked. In 
many cases the colloid contains over five times as high a percentage 
of these constituents as the soil. 
It is known that surface soils are usually higher in silica (34, S5) 
and lower in alumina and iron oxide than subsoils. This same rela- 
tion holds true for the soils and subsoils of Table 2. 
The difference in composition of soil and subsoil is of the same 
nature as the difference between coarse particles and colloidal mat- 
ter. Since the colloidal matter in soil and subsoil is very similar, the 
general difference between the analyses of soil and subsoil can usually 
be accounted for merely by a difference in the quantity of colloidal 
matter present. As a rule, then, one would expect a marked differ- 
ence in texture when there is much difference in composition between 
the surface soils and subsoils, and, conversely, there would be little 
difference in texture between the two layers when there is little differ- 
ence in chemical composition. Experience bears this out. Of course 
there are many exceptions to this rule, such as when calcium car- 
bonate, iron concretions, etc., are unevenly deposited in the various 
soil horizons. 
RELATION OF SAMPLE ANALYZED TO TOTAL COLLOIDAL MATTER IN THE SOIL 
The samples analyzed were only a small part of the total quantity 
of colloidal matter estimated to be present by adsorption methods. 
The question naturally arises: Is this small part representative of 
the whole? An attempt was made to answer this question by analyz- 
