valent to that of 400 atmospheres would be necessary to
compress the water to an extent representing this shrinkage in volume.

On the other hand, gels when exposed to the air lose water by evaporation,
shrink in volume, and finally become hard inelastic solids, as in the case
of the familiar forms of glue, gelatin, agar-agar, gum arabic, etc.

The difference in the relation of gels and that of non-colloidal solids to
water may be illustrated by the different action of peas, beans, etc., and
of a common brick, when immersed in water. Each of these substances, under
these conditions, absorbs, or "imbibes," water; but the peas and beans
swell to more than twice their original size and become soft and elastic,
while the brick undergoes no change in size, elasticity, or ductility. In
all cases of colloidal swelling, the swollen body possesses much less
cohesion, and greater ductility, than it had before swelling. The essential
difference in the two types of imbibition is that in the case of the
non-swelling substances the cohesion, or internal attraction of the
molecules of the material, is too great to permit them to be forced apart
by the water; while in colloidal swelling, the particles are forced apart
to such an extent as to make the tissue soft and elastic. It is possible,
of course, to make this separation go still further, until there is an
actual segregation of the molecules, when a true solution is produced; for
example, gum arabic when first treated with water swells into a stiff gel,
then into a soft gel, and finally completely dissolves into a true
solution.

=Reversibility of Gel-formation.=--In some cases, the change of a sol to a
gel is an easily reversible one. Glue, gelatin, various fruit jellies,
etc., "melt" to a fluid sol at slightly increased temperatures and "set"
again to a gel on cooling, and the change can be repeated an indefinite
number of times. On the other hand, many gels cannot be reconverted into
sols; that is, the "gelation" process is irreversible. For example,
egg-albumin which has been coagulated by heat cannot be reconverted into a
sol; casein of milk when once "clotted" by acid cannot again be converted
into its former condition, etc. Irreversible gelation is usually spoken of
as "coagulation." Some coagulated gels, by proper treatment with various
electrolytes, etc., can be converted into sols, the process being known as
"peptization"; but in such "peptized" hydrosols, the material us