(the experiment lasted several months), which would set up currents in the liquids and intermix their strata. Notwithstanding the uniformity of the temperature, the substance in solution in time ascended into the water and distributed itself uniformly through it, proving that there exists between water and a substance dissolved in it a particular kind of attraction or striving for mutual interpenetration in opposition to the force of gravity. Further, this effort, or rate of diffusion, is different for salt or sugar or for various other substances.[16 bis] It follows therefore that a peculiar force acts in solution, as in actual chemical combinations, and solution is determined by a particular kind of motion (by the chemical energy of a substance) which is proper to the substance dissolved and to the solvent. [16 bis] Graham, in the jelly formed by gelatine, and De Vries in gelatinous silica (Chapter XVIII.) most frequently employed coloured (tinted) substances, for instance, K_{2}Cr_{2}O_{7}, which showed the rate of diffusion with very great clearness. Prof. Oumoff employed the method described in Chapter X., Note 17, for this purpose. Graham made a series of experiments similar to those above described, and showed that the _rate of diffusion_[17] in water is very variable--that is, a uniform distribution of a substance in the water dissolving it is attained in different periods of time with different solutions. Graham compared diffusive capacity with volatility. There are substances which diffuse easily, and there are others which diffuse with difficulty, just as there are more or less volatile substances. Seven hundred cubic centimetres of water were poured into a jar, and by means of a syphon (or a pipette) 100 cub. centimetres of a solution containing 10 grams of a substance were cautiously poured in so as to occupy the lower portion of the jar. After a lapse of several days, successive layers of 50 cubic centimetres were taken from the top downwards, and the quantity of substance dissolved in the different layers determined. Thus, common table salt, after fourteen days, gave the following amounts (in milligrams) in the respective layers, beginning from the top: 104, 120, 126, 198, 267, 340, 429, 535, 654, 766, 881, 991, 1,090, 1,187, and 2,266 in the remainder; whilst albumin in the same time gave, in the first seven layers, a very small amount, and beginning from the eighth layer, 10, 15, 47, 113, 343, 855, 1,892, and in the remainder 6,725 milligrams. Thus, the diffusive power of a solution depends on time and on the nature of the substance dissolved, which fact may serve, not only for explaining the process of solution, but also for distinguishing one substance from another. Graham showed that substances which rapidly diffuse through liquids are able to rapidly pass through membranes and crystallise, whilst substances which diffuse slowly and do