One consequence of this system is, that when a piece of gelatine swells, there is a considerable enlargement in the zones of compression; in other words, some, at least, of the imbibed water is compressed. Now the compression of water means that work is done, and when gelatine swells, therefore, we expect--and actually find--that heat is liberated (5.7 cal per g. gel). Hence also by the Le Chatelier theorem, we expect--and find--that gelatine swells best in _cold_ water. Further, the compression of water involves a decrease in volume, and we therefore expect--and actually find--that the volume of the swollen jelly is appreciably less than the volume of gelatine plus the volume of water imbibed. Another consequence of such a compressed system is that a gelatine jelly, even in water, will have a surface tension towards water just as the water itself has such a tension to the water vapour above the liquid. This interfacial tension of the jelly will of course have a contractile effect, and will tend to resist swelling and to limit it as far as it possibly can. This force, tending to contract the jelly and resist imbibition is therefore one of the main influences at work in the swelling of gelatine, and is one of the two principal factors which determine the extent of the maximum swelling when equilibrium is established. The force tending to resist swelling is, in the ultimate, just surface tension. Its actual magnitude depends, of course, mainly upon the extent of compression in the dispersion medium of the gel, and will be a resultant which is a function of this compression. The magnitude will thus vary with the average compression in the continuous network of compressed water. It will be obvious that as the jelly swells the power of resisting the swelling will decrease, and the interfacial tension with the external water will tend to disappear. If the force tending to swell were great enough the swelling would continue until the zones of compression were no longer in contact and the gel would become sol.