monly employed than that just described, consists in the dialysis of a salt solution in which a colloidal base or acid is present, either owing to natural hydrolysis or to the previous addition of an alkali or acid. Thus colloidal silicic acid may be prepared by dialyzing either a solution of sodium silicate alone, or one to which hydrochloric acid has been previously added. A dark red but perfectly clear colloidal suspension of ferric hydroxide is obtained by the dialysis of a ferric chloride solution which has been treated with ammonium carbonate until a permanent precipitate begins to form. This process of dialysis is commonly resorted to also for freeing colloidal solutions or suspensions prepared in other ways from mineral impurities. It is most conveniently carried out in parchment tubes, which are now an article of commerce. As the surface exposed by these is large, the process is a comparatively rapid one. The solution to be dialyzed is placed within such tubes, and these are immersed first in running tap water and afterwards in distilled water which is frequently renewed.
There is one other method of sufficient importance to deserve mention, and this is the process recently described of preparing colloidal suspensions of metals by producing an electric arc under water between electrodes of the metal in question. This is most readily carried out with the non-oxidizable metals, such as gold or platinum. When gold is used, red clouds of colloidal gold are formed near the arc, and in half a minute the whole liquid assumes a red color. The method depends on the fact that the metal is volatilized into the arc or spattered into it in an extremely finely divided form, and is then condensed or absorbed by the water, which, owing to the absence of electrolytes, has little tendency to cause aggregation of the particles.
Besides these colloidal suspensions artificially prepared from mineral substances, others can be obtained by dialysis and other treatments from animal and vegetable sources. Among the most fully investigated of these are heated albumen and gum mastic.
Properties indicating Heterogeneity.—Turning now to the properties of such colloidal suspensions, it seems appropriate first to refer to those which indicate that these mixtures really are suspensions of minute particles and not true solutions. The fact that the components of the mixture are separated by filtration through animal membranes or close-grained porcelain filters is not of itself an evidence of physical heterogeneity; for by copper ferrocyanide membranes, prepared by depositing a precipitate of this substance in an unglazed porcelain cylinder, sugar and even salts can be separated from true solutions. In some cases, the presence of particles in suspension in so-called colloidal mixtures has been proved directly by microscopic observation; thus this is the case with the colloidal mercuric sulphide and with colloidal arsenious sulphide when prepared under certain conditions, but not
