Dialysis (Gr. tavois, a separating), or Analysis by Diffusion, names given by Prof. Thomas Graham to a method proposed by him for effecting certain separations, usually of compound substances one from another, by means of the different rates at which substances diffuse through moist gelatine-like films or other septa, or upward through water or viscid masses. (See Exdosmose, and Gas, vol. vii., p. 634.) An example will illustrate the nature of the processes and results. A sheet of thin paper being thoroughly moistened, and depressed at the middle to form a sort of cup, a solution containing 5 per cent. each of cane sugar and gum arabic is poured into it; and the paper cup is then placed upon the surface of water in a deep basin narrow enough to keep its edges elevated, and left for 24 hours. The cup being removed at the end of the time, the quantity of liquid it contains is found increased by endosmose; but while a little of the liquid in the vessel below, tested with acetate of lead, shows a mere trace of gum, upon evaporation of the remainder the sugar crystallizes from it, in quantity equal to three fourths of that placed within the paper cup. The sugar, therefore, has rapidly made its way through the septum used, while the passage of the gum has been almost perfectly resisted.

The paper can be replaced by moist animal membrane, or by a thin layer or film of any substance having the character of a jelly, as hydrated gelatine, albumen, mucus, or gelatinous starch; but the most useful septum is parchment paper in thin sheets and without sensible flaw or porosity. It is found that, through moist films or partitions such as those here named, through masses of different substances in the gelatinous state, or through liquids, very many and perhaps all substances capable of crystallization in definite forms make their way by diffusion at rates which, though differing for the different substances, are rapid in comparison with those at which any substance having itself the gelatinous or jelly-like condition can traverse or diffuse in the same media. These facts led Prof. Graham to divide the great body of chemical substances (especially compounds) into two classes, which are readily characterized by the tendency of the former to crystallize definitely, either alone' or in combination with water, and to dissolve rapidly and generally in solutions free from viscosity; while the latter when dry incline to the vitreous structure, having little tendency to crystallize, dissolve slowly or only soften, and as a rule assume the viscid or gelatinous state.

To these classes, respectively, he gives the names of "crystalloids" and "colloids" (the latter from the Gr. kon, glue). In experiments such as those already referred to, the paper sized with starch, or other film o: membrane containing a jelly or viscid material of any sort, is a colloidal partition or body; and the colloid gum has very slight power to penetrate it. This is found true also of such substances as hydrated silicic acid, a number of hydrated metallic peroxides, starch, vegetable gums, dextrine, caramel, tannin, albumen, and vegetable and animal extractive matters, all of which are colloids. The crystalloids, however, as cane sugar, and a large number of chlorides, sulphates; etc, of metallic bases, readily penetrate the colloidal partitions or media; and the explanation given is, not that either class of substance is afforded or denied passage through any effect of capillary attraction as ordinarily understood, but that, the affinity of any colloid for water being of the feeblest character, one colloid cannot with any rapidity abstract molecule for molecule the water from another, by which process it could be conveyed through it; while the crystalloids brought in contact with a moist colloid, having a high affinity for water, can displace the colloid from solution particle by particle, and thus make their way through its mass.

These results are beautifully shown by placing at the bottom of two glass jars respectively, in a little starch jelly and then surmounted with several inches depth of the same jelly, a colored crystalloid, as bichromate of potash, and a colored colloid, as caramel; the gradual elevation of the former through the mass can be daily observed, while at the end of eight days the caramel has scarcely begun to discolor the jelly above its first position. The different rates of diffusion through such septa allow of the employment of the method thus discovered for separating, in degree or partially, one from among two or more crystalloids existing in mixture, but more readily and satisfactorily a crystalloid from a colloid. To this peculiar mode of separation Graham gives the name dialysis. It is conveniently effected with a "hoop dialyser," a sheet of parchment paper stretched beneath a hoop, and secured about it in the manner of a sieve. The sheet being moistened, and receiving in it a very thin layer of the solution from which some substance is to be separated (the separation being more rapid as the layer is thinner), is floated on a sufficient body of water in a larger vessel. To separate in degree two or more crystalloids, the simpler method of "jar diffusion" often suffices.

The mixed solution of crystalloids is conveyed by use of a pipette, so quietly as to leave the superincumbent liquid quite undisturbed, to the bottom of a jar of water or alcohol, and left at rest; the most diffusible substance rises most rapidly, and is more entirely separated from the others as the time is greater, and the height to which it ascends through the column increases. By carefully drawing off with a siphon, at the end of the experiment, successive strata of the liquid into separate vessels, and quantitatively analyzing their contents, the quantities of the "diffusates" in the strata from below upward, and so the diffusi-bilities of the substances, are determined. Thus, with 10 per cent. solutions in pure water, introduced to the bottom of separate vessels, beneath 4.38 in. of pure water, 1 per cent. of common salt in solution had at the end of 14 days reached the uppermost of 16 strata of equal depth in the column; while in the same time sugar had barely appeared (.005 gramme) in the uppermost stratum, gum had diffused itself to the tenth stratum, and tannin to the ninth, from the bottom. By such means, with proper care and noting of conditions, the absolute and comparative diffusibilities of substances can be determined.

Hydrochloric acid and the allied hydracids are found to be the most diffusive substances known; the solid chlorides are high in the scale, and of these apparently chloride of sodium highest. As an illustration of the results of series of experiments, the approximate times of equal diffusion of the following substances were found as here given: hydrochloric acid, 1; chloride of sodium, 2.33; sugar, and sulphate of magnesia, 7; albumen, 4!); caramel, 98. When two or more diffusible substances are mixed, the difference in their rates of diffusion is increased, and effective analysis by diffusion is thus favored. The rate of diffusion is much accelerated by elevation of temperature of the liquid or mass, so that separations may be effected in less time at high temperatures; but the degree of separation is less, since at the same time the less diffusible substances gain in the higher ratio. The rate of diffusion of all substances is less in alcohol, and probably in most other liquids, than in water, or in semi-fluid masses rendered such by water. The name "diftusate" has been given to any substance as diffused, or separated by dialysis. - The relations and applications of the new facts, and the principle which is their basis, are numerous, and some of them of high importance.

The dialyser affords an advantageous method for completely purifying soluble colloids without risk of decomposition, by the readiness with which all crystalloid substances pass from them into water; and Prof. Graham in his paper ("Philosophical Transactions," 1861, part i., p. 183) gives directions for the preparation and purification of many substances of this class. Besides the distinctions already referred to, it will be observed that crystalloid bodies tend to aggregate in plane films and with angular outlines, and are hard and solid; while the more usual condition of the colloid is that showing rounded outlines, a homogeneous mass, with more or less softness and toughness of texture. The water of crystallization in the former is represented by water of gelat-ination in the latter. The colloids are usually insipid; the crystalloids more commonly have a marked taste. Chemically, the former are the inert bodies; the latter, usually active, or energetic. But as observed in their most usual conditions, the rigid crystalloids are almost wholly unsusceptible to external impressions; while the soft colloids have a wide sensibility to external agencies, and thus great mutability of condition.

Even the simply mineral colloids cannot long be kept without change - pure hydrated silicic acid, or soluble silica, sealed up tightly, undergoing change within a few days or weeks; and the existence of many of them is only in and during a continued metamorphosis. This is especially true of albumen, gelatine, mucus, and related substances, as existing in the fluids and living tissues of the animal body. These colloids are plastic or nutritive, and apparently in good part because they are mutable or capable of those successive metamorphoses during which the conditions of vitality can be secured, and in turn vital force and action evolved and manifested. Thus, these elements stand physiologically in relations the reverse of those they show chemically; and Prof. Graham accordingly terms the crystalloid a statical, and the colloid a dynamical condition of matter. He suggests that the colloidal condition of matter may be looked upon as "the probable primary source of the force appearing in the phenomena of vitality;" while "to the gradual manner in which colloidal changes take place (for they always demand time as an element) may the characteristic protraction of chemico-organic changes also be referred;" in these intending to include, of course, the time required for application of the power of the will, for exertion of muscular force, and the physical changes that underlie the phenomena of sensation and thought.

The facts observed in connection with diffusion appear to lead to a new understanding of endosmose, as effected, in part at least, by the circumstances that a colloid cannot abstract water from (or dehydrate) another colloid or a crystalloid, while a crystalloid can readily dehydrate a colloid, and in so doing effect its own movement through the latter. - The method of dialysis can be employed for the extraction of arsenic, tartar emetic, corrosive sublimate, strychnine, morphine, and other crystalline poisons in the stomach, blood, milk, or any organic mixture. The crystalloid poisons will pass through the septum into the outer vessel, where their presence can be shown by the usual tests. By it soluble albumen may be obtained in a state of purity, by addition of acetic acid, and use of a colloidal septum. Nitrate of silver, from photographer's waste, when put into the dialyser readily separates from the albumen and other organic impurities, and can thus be saved. As early as 1864 Mr. Whitelaw took out a patent in England for the removal of chloride of sodium and nitre from the brine of corned and salted meats by means of dialysis.

Liebig has shown that the brine contains a large proportion of the nutritious constituents of the meat; and if we could remove the salts and evaporate the residue, we should have all of the properties of a good soup. This process is successfully accomplished by Mr. Whitelaw's apparatus, as the savory and valuable constituents of the meat are colloids, and will not therefore pass through a membrane. A further technical application of the doctrine of dialysis is in the extraction of sugar from the beet; and it has been proposed to apply the same method to the extraction of sugar from the cane. The contrivances employed by sugar refiners are called osmogenes, and they are now much used in Germany and France. Graham applied the principle of dialysis to the concentration of the oxygen of the air. When air is passed through shavings of India rubber, the rubber retains a portion of the nitrogen, and the proportion of oxygen can be increased to 41 per cent. The proposition has also been made to separate substances which fuse at different temperatures by passing them through porous walls made of refractory material.

This is called dialysis in the dry way.