A fallen skin contains less water than when in the swollen condition, but the difference between the percentages of water in the skin in the two states is only relatively small, and obviously insufficient in itself to account for the great difference in the physical properties of the skin in the two conditions. It is evident, therefore, that the state in which the water is held in the two stages must be different. It is reasonable to suppose that, in the case of swollen skin substance, the whole of the capillary cavities of the organized structure of the skin are completely filled with water, which, owing to its incompressible nature, confers upon the skin elastic and unyielding properties which it always possesses in this condition. The water, entering the organized cell structure in the first place by osmotic pressure, is fixed there and confers upon the skin, in part, its own property of incompressibility. It is a well-recognized fact that, during the puering process, large quantities of the skin substance are dissolved by the action of the enzymes in the bate, and it is probable that the finer organized structure is first attacked. The walls of the cells which hold the water are partly destroyed, leaving the skin with, quite possibly, the same amount of water contained in it as a whole, but with the water dispersed at large throughout its interior structure instead of being held in the fibres. The result of this is, that when a small area of the puered skin is subjected to local pressure the water in the portion undergoing compression is free to move into the adjacent portions of the skin, and there is no force acting upon it, except that of capillarity, to cause it to return to that area when the pressure is removed from it. When a portion of a puered skin is compressed, the surrounding part is always swollen to an equivalent amount. That this view is, in the main, correct is supported by an observation of Abt (Paris), that, while cell nuclei may be demonstrated in the layers near the epidermis of an unpuered skin, they are entirely absent from a puered skin, showing that the puer has completely dissolved the nuclear structure of the cells. The swelling caused by alkalis is of a different nature to the swelling caused by acids. Procter has pointed out[59] that alkaline swelling is not repressed by sodium chloride, even when caused by sodium hydrate, but is repressed by sufficient concentration of the hydroxyl ion in the outer solution. [59] Colloidal Chemistry, p. 19.