508 
Journal of Agricultural Research 
Vol. XXVIII, No. 6 
wide variation in the types and classes of soil, and the numerous factors affect¬ 
ing the binding power of the colloid, this relation seems to be very satisfactory. 
It is evident from the relation between the load and the amount of colloid 
that the amount of colloid in the soil could be calculated provided the load were 
known. This is somewhat analogous to the water adsorption factor as de¬ 
veloped by Robinson ( 9 ). While this method is not adaptable to as great re¬ 
finement and accuracy as the adsorption methods, it would seem to indicate the 
colloidal properties of soils for comparative purposes, particularly from the 
physical standpoint, more readily than the mere knowledge of the amount of 
colloid in each of the different soils under comparison. By combining the two 
methods, as was done in working out the above relation, one may be used as a 
control on the other. However, for the purpose of this discussion, it was as¬ 
sumed that the amount of colloid in the soil was that shown by the water vapor 
adsorption. 
THE BINDING POWER OF SOIL COLLOIDS 
In order to determine the binding power of soil colloids it was thought best 
to determine their breaking strength in controlled mixtures of sand. However, 
the binding power of the colloids was found to be influenced by a number of 
factors, as, for example, the size and grading of the sand particles wdth which 
they were mixed, the kind of colloid, that is, the soil from which the colloid was 
extracted, and to some extent by the amount of colloid present. 
The particular colloid used in these determinations was extracted from the 
Sharkey clay soil by means of centrifuging and filtering as described by Moore 
et al. (7). The material was reduced to as low a moisture content as possible by 
means of the filters. It was then worked through a 1-millimeter sieve to make it 
uniform in consistency and placed in an air-tight jar to prevent evaporation. 
This material contained 30 per cent colloid (oven dry) and 70 per cent water. 
In mixing up a mixture of colloid and sand containing 25 per cent colloid, for 
example, it was necessary to mix 83.3 gm. of the above material with 75 gm. of 
sand. This made a total of 100 gm. of sand and colloid and 58.3 gm. of water. 
The materials were thoroughly mixed and then allowed to dry down to the criti¬ 
cal moisture content, which had to be determined for each mixture. The bri¬ 
quettes were made up and treated in the same way as those of soil material. 
The sand used in these experiments was separated from a clean, white river 
sand and from quartz flour. The silt and clay fractions were separated from 
the quartz flour by subsidence. The quartz flour contained a trace of colloidal 
material and the clay fraction was freed from this colloidal material by repeated 
centrifuging. Particular pains were taken to have all of the separations as exact 
as possible. 
THE EFFECT OF SIZE AND GRADING OF THE MATERIAL 
The first series of briquettes was made up of a mixture of 25 per cent colloid 
with each of the six grades of sand from coarse sand to clay. The results are 
shown in Table V and indicate that the binding power of the colloid increases as 
the size of the material decreases. 
Table V. —Effect of the size of the material on the breaking strength of the briquettes , 
25 per cent Sharkey colloid 
Material 
Size of 
particles 
Weight 
of dry 
briquette 
Average 
load per 
briquette 
Load per 
gram of 
briquette 
Load per 
gram of 
colloid 
Coarse sand_ 
Millimeter 
1.0 -0.5 
Grams 
21.7 
Kilograms 
82 
Kilograms 
3.8 
Kilograms 
15.1 
Medium sand_ _ 
.5 - . 25 
21.4 
91 
4.3 
17.0 
Fine sand__ 
. 25 - . 1 
21.5 
134 
6.2 
24.9 
Very fine sand__ 
.1 - . 05 
21.6 
211 
9.8 
39.1 
Silt___ 
. 05 - . 005 
21.4 
526 
24.6 
98.4 
Clay____ 
. 005- . 001 
21.5 
1,325 
61.6 
246.6 
