Before leaving this point, it is desirable to note the effect on the swelling of gelatine of the extremes of this lyotrope influence. Substances like iodides, thiocyanates and urea prevent a gelatine sol from setting to a gel at all, and a piece of gelatine in such solutions swells rapidly until it solates. On the other hand, sulphates, tartrates, etc., make a stiffer gel on account of the enhanced compression. Gelatine in such solutions may swell, but at a much slower rate than in water and with a decreased maximum extent. A gelatine gel may in such solutions not only fail to swell at all, but actually contract and in some cases, indeed, be practically dehydrated. If a gel be in a very concentrated solution of such a substance, it may be that the lyotrope compression in the external solution is greater than the compression in the dispersion medium of the gel; in which case the surface tension effect is reversed, and the external solution tends to increase in volume and the gel to contract. Hence we find that the saturated solutions of such substances as ammonium sulphate and potassium carbonate will dehydrate a gel almost completely, and will also, by a similar action on pelt, make a kind of white leather. It is important to remember this contractile effect of strong solutions of salts, because it is very easy to confuse this effect with a similar result produced in another manner, viz., by a reduction of the force tending to swell. 2. THE DISPERSE PHASE A very important feature of the colloid state is that the particles of the disperse phase appear to possess an electric charge, and if this charge be removed a colloid sol no longer remains such, but precipitates, flocculates, coagulates, etc. As to the origin of this charge several theories have been advanced, but the most generally accepted is that it is a result of the adsorption of electrically charged ions by the particles of the disperse phase. The enormous specific surface possessed by this phase renders it particularly liable to such adsorption. This view harmonizes well also with the general behaviour, of colloid sols and gels, in endosmosis, kataphoresis, precipitation, etc. According to this point of view the particles of the disperse phase are surrounded by a surface layer in which these ions are in much greater concentration than in the volume concentration of the dispersion medium. The hydrion and hydroxyl ion are particularly liable to such adsorption. In the case of a lyophile colloid, like gelatine, the charge may be either positive or negative, according to the nature of the predominant ions in the dispersion medium, and the amount of adsorption is determined by the concentration of these ions in accordance with the adsorption law.