512 
Journal of Agricultural Research 
Vol. XXVIII, No. & 
GENERAL DISCUSSION 
Some of the factors influencing the binding power of soil colloids have been 
outlined above. It was first necessary to establish a suitable method for deter¬ 
mining the binding power and to ascertain the conditions under which the results 
from this method could be relied upon. The method of determining the breaking 
strength of briquettes was adopted as most applicable, and when proper care was 
taken in making them, the results could be duplicated with sufficient accuracy. 
In order to obtain duplicable and therefore comparable results the briquettes 
must be made up with (a) the critical amount of moisture present, that is, the 
moisture content when the briquettes with 2,000 pounds pressure applied assumed 
the smallest volume, (b) the same degree of dispersion of the colloid and a uniform 
distribution, (c) the same pressure per square inch (2,000 pounds) for the same 
length of time and (d) drying in such a manner as to prevent cracking of the 
briquettes. A variation of any of these factors might so affect the breaking- 
strength of the briquettes that the values obtained would be very much in error. 
These four factors, which are involved in the formation of the briquettes ancl 
influence the values obtained, were standardized as nearly as possible. Working 
under the standardized conditions, the effect of the following factors on the 
binding power of the colloid was studied: (a) The amount of colloid present, (b) the 
kind of colloid, (c) the size of the mineral grains, (d) the grading of the sizes of the 
mineral grains, and (e) the surface area of the mineral grains. These factors are 
variable and interdependent so that great accuracy is not obtainable in deter¬ 
mining them. 
Tests on soils show T ed a general relation expressed in the formula L= .42C 1 * 24 . 
An additional factor is involved, in the case of soil, in the determination of the 
amount of colloid in the soil. The variation in load is in direct proportion to the 
amount of colloid present, the constants in the equation changing with the char¬ 
acter of the colloid and the mineral portion of the soil. 
Tests on mixtures of colloid and sand show that with decreasing size of sand 
grains there is an increasing value for the binding power of the colloid. The 
increase in binding power is not directly proportional to the decrease in the average 
diameter of the sand grains. For example, in Table V the medium sand with 25 
per cent colloid gives a load per gram of colloid of 17 kgm., but when very fine 
sand is used with a mean diameter one-fifth that of medium sand, the binding 
power is approximately twice as great. However, with the silt and clay sizes 
the increase in binding power is larger in proportion than the decrease in diameters. 
This indicates that the amount of surface has a very great influence. The 
calculation of surfaces as shown in Table VII shows that it is probable that a rela¬ 
tion does exist, though the data for a complete verification of the relation are not 
available. 
The kind of colloid is of great importance in determining the binding power of 
mixtures. Of those colloids tested, that from the Stockton clay adobe soil had 
the greatest binding power and that from the Cecil clay loam subsoil the least. 
The reason for this difference has not been determined, but it may be due to a 
difference in the chemical composition, in the size or aggregation of the colloidal 
particles, or to other factors. * 
This discussion gives some idea of the nature of the problem of determining 
the binding power of colloidal material in the soil. The complexity of the problem 
is pointed out and the influence of the various factors is recognized. While the 
factors and their influences have been partially determined, it yet remains to 
coordinate and fully evaluate them, so that more exact information of the character 
of a soil may be derived from the determination of the colloid content and one or 
more of these factors. 
