18 BULLETIN 1452, IT. S. DEPARTMENT OF AGRICULTURE 
Each of these determinations — quantity of liquid water held, 
water vapor adsorbed over strong sulphuric acid, and heat of wet- 
ting — may be regarded as a measure of some phase of the affinity of 
the material for water. The correspondence between the different 
determinations, shown by the constancy of the ratios in Table 8, 
indicates that in the case of most soil colloids any one of these de- 
terminations is sufficient for characterizing the affinity of the ma- 
terial for water. 
Some light is thrown on the nature of the soil colloids by com- 
paring the quantities of water held (Table 7) with the volume 
changes (Table 6) of the different materials. Gelatin and ben- 
tonite hold large quantities of water and increase in volume in pro- 
portion to the water held. The inelastic silica gel holds much less 
water than these materials and shows practically no volume change. 
Nearly all the soil colloids hold somewhat more water than silica 
gel. The Fallon, Sharkey, and Marshall colloids hold approxi- 
mately twice as much, and they increase in volume markedly. The 
volume increase in the case of the Fallon and Susquehanna colloids 
is approximately equal to the weight of water held; in the case of 
the other soil colloids it is 0.6 to 0.8 of the weight of water held. 
On the basis of these results the soil colloids might be regarded as 
elastic gels. But in view of the fact that all the soil colloids showed 
more or less fracturing under the microscope, probably none of them 
should be considered typically elastic gels. 
VISCOSITY 
The viscous behavior of a colloid has long been considered espe- 
cially important for characterizing the general nature of the mate- 
rial. At low concentrations the lyophile colloids are much more 
viscous than the lyophobe, and the critical concentration at which a 
sharp increase in viscosity occurs is usually much lower for the 
typical lyophile than for the typical lyophobe colloids. As a rule 
the viscosity of the lyophiles is more affected than that of the lyo- 
phobes by changes in temperature, manipulation, time of standing, 
and the like. 
Although colloids are characterized by their viscous behavior, vis- 
cosity apparently can not be considered the expression of a single, 
definite property of a colloid, such as size of particle. Although vis- 
cosity seems to be largely determined by the volume of the dispersed 
phase in proportion to the volume of the dispersion medium, few 
sols follow exactly Einstein's formula (12) for increase in viscosity 
with increasing volume concentration. It has usually been assumed 
that deviations from this formula are due to liquid which is asso- 
ciated with the dispersed phase so closely that it acts as a part of the 
solid volume of the dispersed phase in so far as viscosity is concerned. 
Aggregation and variation in electrical charge seem to affect the vis- 
cosity, presumably by affecting the volume of the associated liquid. 
According to Oclen's (42) investigations of sulphur sols, size of 
particles also influences viscosity. 
The viscosity of the soil colloids was determined with an Ostwald 
pipette of such dimensions that 25 cubic centimeters of water was 
delivered in about 60 seconds. Determinations were made at 25° C. 
and could be duplicated within 0.4 second if the sol did not stand 
