94

CONDITIONS OF OSMOSE.

[61.

ox bladder with the muscular coat removed, or else an artificial membrane formed by calico soaked in white of egg and dipped into boiling water to coagulate it; to

FIG. 36.

the upper aperture of the bell-jar, a tube, one-tenth of the diameter of the lower opening of the jar, is fitted. This tube is open at both ends, and is graduated into millimetres. A rise or fall of liquid in the narrow tube amounting to 100 millimetres therefore represents the entrance or removal of a stratum of liquid of 1 millimetre in thickness over the whole surface of the membrane. In using the instrument, the membrane is well macerated in pure water, and the saline solution introduced into the jar, A, until it stands at a fixed mark in the narrow tube. The apparatus is then placed on a tripod stand in a tall cylindrical jar, B, and distilled water poured in until it stands exactly at the level of the liquid in the tube. During the whole experiment this level is carefully maintained, by the addition or removal of water in the outer jar, as circumstances require.

The principal points which were ascertained by experiments conducted in this way were the following:

i. Neutral organic substances, such as urea, gum arabic, sugar of milk, gelatin, and salicin, exercise little or no osmotic action. ii. Strictly neutral salts, such as magnesic sulphate, sodic chloride, and baric chloride, exercise no peculiar osmotic power, but appear to follow nearly the same rate of diffusion as that which is observed when no porous partition is used.

iii. Alkaline solutions, and especially the solutions of the potassic and sodic carbonates, on the contrary, produce endosmosis to a most remarkable extent. This effect is observed even in solutions which contain not more than 1 part of the salt to 1000 of water. Indeed, it was found generally that these osmotic phenomena were most strongly developed in dilute solutions, such, for instance, as did not contain more than 2 per cent. of the salt. In these experiments a large bulk of water entered the osmometer whilst only a very small portion of the alkaline salt escaped into the water of the outer jar. For example, in 5 hours, when a solution of potassic carbonate containing part of the salt in 1000 of water was placed in the osmometer, the liquid in the stem of the instrument rose through 192 divisions; and for each milligramme of potassic carbonate that became diffused into the outer cylinder, upwards of 550 mgrms. of water entered the osmometer; but when a solution which contained 1 per cent. of potassic carbonate was used, not much more than 63 mgrms. of water entered the instrument for each milligramme of carbonate that became diffused into the outer cylinder. When the liquid rises in the osmometer, Graham distinguishes it as positive osmose.

61.]

OSMOSIS.

95

iv. On the other hand, dilute acids, and solutions of acid salts generally, produce a current in the opposite direction; consequently the column falls in the stem of the osmometer. effect is distinguished as negative osmose.

This

Salts which admit of division into a basic salt and free acid exhibit osmotic properties in a high degree. This is well seen in the case of aluminic acetate, plumbic nitrate, zincic chloride, and the chromic and ferric salts. The acid travels outwards by diffusion, and the inner surface of the membrane is left in a basic condition, whilst the outer surface is acidconditions highly favourable to rapid positive osmose.

v. In every instance in which osmotic action is observed (except in the cases of alcohol and cane sugar), a chemical action on the material of the septum, whether it consists of bladder or of earthenware, invariably occurs; and it is remarkable, that if porous materials, not susceptible of decomposition by the liquids, be made use of as a partition, the osmotic phenomena become insignificant :—for instance, a plug of gypsum, of washed unbaked clay, of tanned leather, or of compressed charcoal, although sufficiently porous, gives rise to little or no osmotic action. To induce osmose under the most favourable circumstances, the chemical action on the septum must be different on the two sides, not only in degree, but also in kind; such as is produced by the presence of acid upon one surface, and of alkali on the other, and the septum itself must belong to the class of colloids (62). These circumstances are specially interesting from their chemical bearings, as is also the next point, which is no doubt connected with the influence of the colloid employed as the septum.

vi. Two salts, when mixed, often have an osmotic action very different from that which they exercise separately. For example, perfectly neutral potassic sulphate has a feeble positive osmose, represented by a rise of 20mm. in 5 hours. The addition of I part of potassic carbonate to 10,000 of the solution raised it to nearly 100mm. in 5 hours, whilst an equally minute trace of hydrochloric acid stopped the osmose almost entirely. Similar results were obtained with sodic sulphate. Sodic chloride, on the other hand, exhibits a remarkable power of reducing osmotic action in other salts. The osmose of a solution of sodic carbonate, containing of the carbonate, was reduced from 179' 32,, by the addition of 1 per cent. of sodic chloride. From other experiments, it appears further that two different saline solutions, one placed in the osmometer, the other in the outer

mm.

to

96

OSMOSIS-DIALYSIS.

[61.

jar, each solution holding equal weights of the different salts dissolved in the same bulk of water, may also give rise to osmotic action, when separated by a suitable porous partition.

Liebig has shown that the mechanical force of the osmotic current may be measured by the following simple means :-Let the open extremity of the shorter limb of a glass tube bent into the form of a syphon be closed by a piece of bladder, pour a little mercury into the bend of the tube, and fill the shorter limb with the saline liquid under experiment; immerse the bend of the tube and the membrane in water, leaving the extremity of the longer limb open: as the water enters the tube, the mercury will be raised in the longer limb, and when the column reaches a certain height, the two liquids will intermix without change of volume. The length of the column which has been raised above the level of the surface of the mercury in the shorter limb must be measured, and when compared with the length of the column obtained with other liquids under similar circumstances, it offers a comparative measure of the osmotic force for each. Osmotic action thus offers an interesting case of the direct conversion of chemical attraction (on the septum) into motive power, the extent of which admits of ready numerical expression.

Osmotic phenomena are constantly going on both in plants and in animals; for in their tissues, liquids of very different natures, sometimes acid, still more often alkaline, are circulating through vessels necessarily constructed of flexible and porous materials; and in the economy both of the vegetable and of the animal creation such actions are of the highest importance to the due performance of the vital functions. In fact, we as yet know not how intimately the entire processes of absorption, nutrition, and secretion are connected with the operations of liquid diffusion and of endosmosis.

(62) Dialysis-Crystalloids and Colloids. In a subsequent memoir (Phil. Trans. 1861, 183), Graham has pursued the subject of liquid diffusion, and applied the process to the purposes of chemical analysis. The most remarkable conclusion at which he arrives in this memoir is that all bodies are chemically referable to one or other of two great classes, which he distingushes as crystalloids and colloids. Bodies susceptible of crystallization, or crystalloids, form a solution generally free from viscosity, and they are always sapid; they are especially endowed with the tendency to diffusion through a porous septum: whilst the colloids, or jelly-like substances (from xóλλŋ, glue), such as gum, starch, dextrin, tannin, gelatin, albumin, and caramel, are charac

62.]

DIALYSIS.

97

terized by a remarkable sluggishness and indisposition to diffusion or to crystallization; when pure, they are also tasteless or nearly so.

Magnesic sulphate, for instance, one of the least diffusible crystalline bodies, has a diffusibility 7 times as great as that of albumin, and 14 times as great as that of caramel, if compared by determining the relative weights which are diffused in equal times under similar circumstances.

If we compare together the times required for the diffusion of equal weights of different substances, calling the time of hydrochloric acid, the most diffusible of known bodies, unity, the following table may be formed of

In making these experiments, some insoluble colloid, such as a sheet of the paper modified by sulphuric acid, which is well known under the name of 'parchment paper,' is employed as a septum through which the diffusion may take place. A ready dialysis (from dia, asunder, Avoiç, separation), or separation of crystalloid and colloid bodies, may be effected in the following manner:- Prepare a shallow tray by stretching a sheet of parchment paper over one side of a hoop of gutta percha; place the mixture for experiment in the tray, and then float it in a shallow dish of pure water, the volume of the water being from 4 to 10 times that of the mixture. In the course of 24 or 48 hours, the separation will have taken place more or less completely. In this way a solution of arsenious acid, for instance, after admixture with various articles of food, readily diffuses out. If the diffusate be evaporated down to a small volume, the arsenious acid may be obtained nearly free from organic matter, and sufficiently pure to yield a yellow precipitate with sulphuretted hydrogen.

The process of dialysis has already received important applications both in pharmacy and in the laboratory. In the examination of organic mixtures for poisons, it affords a simple method of separating almost all crystalline bodies, such as the mineral poisons and the vegetable alkaloids, from the mass of organic

98

DIALYSIS CRYSTALLOIDS, COLLOIDS.

[62.

fluids without introducing any extraneous substance, thus leaving the mixture perfectly fit for other modes of examination.

By dialysing solutions of salts of aluminium, chromium, aud iron, the hydrates of the metals remain in the soluble condition, the crystallized acids having been separated. When a solution of an alkaline silicate to which an excess of dilute hydrochloric acid has been added is treated in a similar manner, the sodic chloride and free hydrochloric acid are removed and the silicic acid remains in the soluble colloid condition. The solutions of aluminic, ferric, and chromic hydrates and silicic acid thus obtained in the colloid condition are very unstable, showing a great tendency to gelatinize, a change which is hastened by heat, and by the addition of certain salts. A solution containing Tooth part of any alkaline or earthy carbonate effects the coagulation of the colloid silicic acid in a few minutes, and a few drops of well water, caused the gelatinization of a solution containing 2 or 3 per cent. of alumina.

Many colloidal bodies of organic origin, such as gum, albumin, or caramel, may in a similar way be freed from saline impurities, which it is very difficult, if not impossible, to remove by other means. Diffusion, indeed, takes place very perfectly from solid hydrated colloids; this may be shown in a striking manner by the following experiment:

Let 10 parts of common salt and 2 of gelose, or Japanese gelatine, be dissolved in hot water, which must be added till it forms 100 parts of solution. If this be poured into a glass jar, it will set, on cooling, into a firm jelly; now pour upon this 700 parts of a similar solution of gelose, but containing no sodic chloride; this also will set into solid jelly. In an experiment made in this way, the whole was left undisturbed for eight days, and the result was compared with a similar experiment, in which diffusion of the salt was allowed to take place into an upper stratum of water instead of one of gelose. The rate of diffusion was found to be nearly the same in the two cases, but rather the more rapid in the case of the solid jelly. This process may be watched very readily by substituting a coloured salt, such as cupric sulphate, or potassic dichromate, for sodic chloride.

Graham has suggested the following explanation of the process of dialysis:-The water in the colloidal septum is not directly available as a medium for diffusion, being in a state of true chemical combination, feeble though it be. Soluble crystalloids, however, can separate water, molecule after molecule, from the hydrated colloid constituting the septum; the crystalloid in this manner obtains the liquid medium required for its diffusion, and thus makes its way through the gelatinous septum.

Graham indeed supposes that the coats of the stomach dialyse the food during digestion, absorbing the crystalloids and rejecting