REVERSIBILITY OF THE COLLOIDAL CONDITION OF SOILS 483 
they also are mixtures of colloidal gels. Gans thought that the sub- 
stance was a chemical compound, but admitted the possibility of its 
existing in the colloidal state. 
Ehrenberg (1915) classifies soil colloids broadly into systems based on 
the physical condition of their phases. Colloidal silica and humus are 
placed in the liquid-liquid system, and the colloidal ferric oxide and alumina 
in the solid-liquid system. 
Colloidal silica originates from the remains of plants and from the 
weathering of rocks. The final state may be colloidal or erystalloidal, 
depending on the speed of the reaction. If the change is rapid, a colloid 
such as opal may be formed; while if it is slow, quartz crystals are formed. 
Iron compounds occur in soils in relatively large quantities and widely 
distributed. If these become dissolved in water, colloidal ferric oxide 
easily forms by hydrolysis under ordinary conditions. With difficultly 
soluble iron compounds such as silicates, weathering and consequent 
solution are necessary. 
Microscopic plant and animal life are colloidal. Altho bacteria may 
be present in large numbers, because of their comparatively small mass 
they are not very important as compared with other soil colloids. 
Colloidal humus originates from all sorts of plant and animal remains, 
is not chemically defined, and varies in all degrees between solid and 
liquid. It stands on the boundary line between a colloidal and a crys- 
talloidal condition, as shown by a slight dialyzability and conductivity. 
Spear (Zsigmondy and Spear, 1917) states that colloidal silica, silicates, 
hydroxides of iron, aluminum, and manganese, and colloidal matter such 
as humus, exist in the soil. 
From what has been said so far it is evident that a great many materials 
existing in the soil in the colloidal condition have been identified and 
studied. There is little doubt (Ehrenberg, 1915) that many other sub- 
stances exist in the soil in this condition, and that, as methods of colloid 
chemistry are further developed and the range of soils examined is extended, 
additions will be made to the list. It is possible that every soil constituent 
will be found to exist in the colloidal condition under some circumstances. 
Methods for determining the amounts of colloids in soils have not been 
sufficiently developed to enable one to determine these amounts with 
any great degree of accuracy. At best, the measurement is only an esti- 
mation. Evidence of the unsatisfactory status of methods exists in the 
