gelatinization, and their property of permitting the free passage through them of non-colloidal substances, but preventing entirely that of other colloids of either of the two classes. This last property is the same one that is involved in the permeability of animal membranes for crystalloids and their impermeability for colloids, since such membranes are themselves nothing more than gelatinized colloids. Yet it deserves, on account of its great importance, a somewhat fuller consideration. This difference in behavior towards crystalloids and colloids may be readily illustrated by immersing sticks of gelatine or agar jelly in one experiment in a colored salt solution and in another in a colored colloidal suspension, and allowing them to remain for a day or more. Such comparative experiments may be made in a striking way with a solution of copper sulphate and ammonia and with a colloidal suspension of ferric ferrocyanide or Prussian blue made by mixing equal volumes of dilute solutions of ferric chloride and potassium ferrocyanide. Upon removing the sticks after some hours and cutting them in two, it will be noticed that the ammoniated copper sulphate has permeated the stick uniformly to its center, while the Prussian blue has not entered it at all.
Not only are gelatinized colloids permeable to salts, but, remarkably enough, they offer only a very slight, often scarcely appreciable, hindrance to the passage of these substances through them. Thus, accurate experiments have shown that the rate of diffusion of salts and mineral acids is the same, at any rate within one per cent., in a solid jelly containing 3 to 5 per cent, of agar-agar as it is in one containing only 1 per cent, of agar-agar, and it is, therefore, presumably the same as in water itself, though the accuracy with which this latter conclusion has yet been directly tested is much less. It has also been shown that the electrical conductivity of salts in a gelatine jelly is only a few per cent, different from that in pure water, and that there is no sudden change in its value when the jelly sets. This property is, however, dependent on the rate of motion of the ionized molecules of the salt through the medium between the electrodes, and the slight variation in it caused by the presence of colloids, even in quantity sufficient to produce gelatinization, proves that the flow of such molecules is but little impeded by the colloid.
Returning now for a moment to the other side of the phenomenon—the impermeability of one colloid by another—attention may be called to an apparently related fact of much importance, namely, to the fact that the presence of a gelatinizing colloid in a liquid in fairly small quantity prevents the coagulation of colloidal suspensions by salts, and, therefore, usually prevents the formation of a coagulated precipitate when the solutions of two chemical substances are mixed which, under ordinary conditions, give rise to such a precipitate. Thus when aque-
