Common salt diffuses itself in a solid mass of jelly almost as easily and extensively as in the same bulk of free water. Thus colloid bodies do not interfere with the diffusion of crystalloids such as salts, but they almost entirely arrest the diffusion of one another. Solutions of salts, sugars, and other crystalloids pass freely through colloid substances, such as parchment-paper, vellum, and membrane into water, although they have no pores, because the particles of the crystals unite diffusively with the water combined in these substances, which solutions of gum, caramel, and other colloids cannot do. These colloid substances are permeable to solutions of crystalloids, impermeable to solutions of colloids. This constitutes Dialysis. The instrument used by Mr. Graham was a little tray formed of vellum or membrane stretched tightly over a hoop of gutta-percha and capable of holding a liquid and floating on water. When a mixed solution of equal parts of salt and gum is put into the tray, after a time all the salt will have passed into the water below, leaving nothing in the tray but an aqueous solution of gum. The following is one of the most extraordinary results of dialysis. Mr. Graham took a silicate of soda, a soluble crystalline salt formed by fusing quartz with carbonate of soda at a red heat, which diffuses readily. He acidulated the aqueous solution of the salt with hydrochloric acid, which changes the constituent silica from being a crystalloid substance into a colloid form. When the liquid was poured into the tray floating on water, after four days, the whole of the acid and the chloride of sodium had been diffused in the water and nothing remained in the tray but an aqueous solution of quartz. There remained in fact, a solution of sand in water, a substance so hard that no pure aqueous solution of it had ever been obtained. Many other crystalline substances besides quartz can exist both in the colloid and crystalloid states. All colloid substances are characterized by non-crystalline habits, low diffusibility, chemical inertness, high atomic weight, and above all by their mutability. The aqueous solution of quartz is limpid and liquid, even if it contains 14 per cent. of silica, but after a time it becomes opalescent, viscous, and ultimately sets into a firm insoluble jelly, capable however of solution by chemical means. This jelly gradually shrinks, exudes pure water, and when perfectly dry it forms a glassy, transparent, but not anhydrous substance, and the residue left by ignition has a specific gravity of 2·2, that of crystallized silica being 2·6.