67.]

DESICCATION OF GASES.

109

(66) Desiccation of Gases.-It frequently happens that in the course of his operations, the chemist requires the gases which are the subjects of his experiments to be in a perfectly dry state. Gases are usually prepared in contact with water, and hence become charged with a variable quantity of aqueous vapour which adheres to them, and whether it be desired to ascertain their density, or to submit other bodies to their chemical influence, it becomes necessary to remove this moisture completely. The gas to be dried, which we will suppose to be in the act of formation in the glass bottle, A, fig. 38, is al

lowed to

pass

FIG. 38.

fused potash, or of calcic chloride, or of quicklime, or of phosphoric anhydride, or of pumice-stone moistened with oil of vitriol, according to the nature of the gas. The bulb, D, may contain the substance upon which the action of the gas is to be exerted, and the gas when it reaches it will be in a dry state; since all the bodies just mentioned possess the property of combining with water and aqueous vapour; and if allowed a sufficient length of time, will remove nearly every trace of moisture from the gases which are brought into contact with them. The different parts of the apparatus are connected by flexible tubes of caoutchouc, c, c.

Diffusion of Gases.

The process of intermixture in gases, and the motions of these bodies, have been even more completely investigated than the corresponding processes in liquids. The movements of gases may be considered under four heads; viz.,

1. Diffusion, or the intermixture of one gas with another. 2. Effusion, or the escape of a gas through a minute aperture in a thin plate into a vacuum.

3. Transpiration, or the passage of different gases through long capillary tubes into a rarefied atmosphere.

4. Osmosis, or the passage of gases through diaphragms. (67) Diffusion of Gases.-In consequence of the absence of cohesion among the particles of which gases and vapours consist,

110

the gases.

FIG. 39.

mixture takes place amongst these bodies very freely, and in all proportions. Very great differences in density occur amongst Chlorine is, for instance, nearly 36 times denser than hydrogen, the lightest of the gases, so that there is about three times as great a difference between the relative weights of these two gases, as between those of mercury and water. But the mingling together of gaseous bodies of different densities produces a result very different from that obtained by the mingling together of two liquids, such as mercury and water; for, if these liquids be mixed by agitation, they separate the instant that the agitation is discontinued. Chlorine and hydrogen, on the other hand, after they have once become mixed, never separate, however long they may remain at rest. Indeed, if the gases be placed in two distinct vessels, and be allowed to communicate only by means of a long tube, the hydrogen or lightest gas being placed uppermost, as represented at н, fig. 39, the heavier chlorine in A, will, in the course of a few hours, find its way into the upper jar, as may be seen by its green colour, whilst the hydrogen will pass downwards into the lower one, and ultimately the gases will be equally intermixed throughout. If a sufficient interval of time be allowed, this equal intermixture occurs with all gases and vapours which do not act chemically upon each other; and when once such a mixture has been effected it continues to be permanent and uniform. The rapidity with which this diffusion occurs varies with the density of the gases; and, contrary to what a superficial consideration might lead us to suppose, the more widely the two gases differ in density, the more rapid is the process of intermixture. If two tall narrow jars of equal diameter be about half filled, the one with hydrogen, the other with common air, which is more than fourteen times as heavy as the hydrogen, so that the water in both shall stand at the same level, and a small quantity of ether be thrown up into each jar, the ether will evaporate in both, and cause in each, ultimately, an equal depression; but the vapour of the ether will dilate the hydrogen at first much more rapidly than the air, for its vapour will become more quickly diffused through the lighter hydrogen. A

67.]

DIFFUSION OF GASES.

111

very simple and striking illustration of the rapidity with which a light gas becomes diffused into a heavier one, is shown as follows:

Take a tube 25 or 30 centimetres, or about 10 or 12 inches long, one end of which is closed with a porous plug of plaster of Paris, 5m. thick, that has been allowed to become dry, and fill it with hydrogen gas, without wetting the porous plug: this is readily effected by introducing the shorter limb of an inverted syphon, s, into the jar, b, fig, 40, till it reaches the top, and then lowering the jar in a deep vessel of water, A; when the air has escaped, the open limb of the syphon is closed with the finger, and the jar raised until the syphon can be conveniently withdrawn: the jar can now be filled with hydrogen prepared in a retort in the usual manner. If the jar after being filled with hydrogen be supported so that the water within and

FIG. 40.

without shall stand at the same level, the water in the jar will immediately begin to rise, and will continue to do so in opposition to gravity, until, in the course of three or four minutes, it will stand some inches higher than the surface of the water in the outer vessel, in consequence of the hydrogen passing through the pores of the stucco, and becoming diffused into the air much more rapidly than the air passes in and becomes diffused through the hydrogen.

Any dry porous substance may be substituted for the plaster; a film of collodion on paper gives excellent results, and unglazed biscuit-ware, or compressed plumbago, is still better.

By means of this simple diffusion tube, taking care to maintain the surface of the water within and without the jar on the same level, as shown at B, in order that the results may not be interfered with by the disturbing force of gravity, Graham has determined the law which regulates the rapidity of gaseous diffusion. Experiments so made show that the diffusiveness or diffusion volume of a gas is in the inverse proportion of the square root of its density; consequently the squares of the times of equal diffusion of the different gases are in the ratio of their specific gravities. For instance, the density of air being 1, the square root of that density is 1, and its diffusion volume is also 1; the density of hydrogen is o'06926, the square root of that density is o 26317, and its diffusion volume 7=3'7998; or as actual experiment shows, 3'83; that is to say, in an experiment conducted with due precautions, whilst I measure of air is passing into the diffusion tube, 3.83 measures of hydrogen are passing out of it.

112

DIFFUSION OF GASES.

[67.

In the case where different gases are mixed and then introduced into the diffusion tube, each preserves the rate of diffusion peculiar to itself. If, for instance, hydrogen and carbonic anhydride be mixed and placed in the diffusion tube, the hydrogen passes out with much the greater rapidity: a partial mechanical separation of two gases differing in density may thus be effected.*

2

*In particular cases advantage may be taken of this fact in the analysis of a mixture of different gases. Suppose it be desired to ascertain whether a certain gas be a mixture, or a single gas-to distinguish, for example, marsh gas (2 CH4) from a mixture of equal measures of hydrogen, and hydride of ethyl (H2+C,H)-the two would give exactly similar amounts of carbonic anhydride and water when detonated with oxygen. But suppose that, by eudiometric analysis of a portion of the mixture, the proportions of carbon and hydrogen have been determined, if another portion be submitted to diffusion, and the residue be again analysed, the proportions of carbon and hydrogen will remain unaltered if the gas consist of marsh gas only; whereas, if it be a mixture, the proportion of hydrogen will be diminished. Pebal has ingeniously applied this method to the examination of the question whether the vapours of certain compounds which, like hydrochlorate of ammonia, yield anomalous vapour volumes, are not really, as Cannizzaro supposes, mixtures at those high

FIG. 41.

temperatures, instead of chemical compounds, and he has succeeded in demonstrating the truth of this hypothesis (Liebig's Annal. 1862, cxxiii. 199). He places a plug of amianthus (c, fig. 41), on which some fragments of salammoniac d are supported, within a tube C, drawn out to a capillary end this tube is supported in a wider one D, and a current of hydrogen is transmitted through the tubes a A, b B, which are surrounded by a charcoal furnace, by which they may be sufficiently heated to volatilize the hydrochlorate of ammonia. The vapours of the hydrochlorate being formed in the inner tube, above the plug of asbestos, ammoniacal gas diffuses through the plug into the hydrogen in the tube C, and may be demonstrated by causing the issuing gas to come into contact with the reddened litmus paper in B; while the hydrogen which traverses the space between the two tubes contains the corresponding hydrochloric acid, and reddens blue litmus in 4. Hence it is evident that the sal-ammoniac when converted into vapour becomes, partially at least, dissociated, as Deville calls it, into its constituents, ammonia and hydrochloric acid, since the ammonia, which is the more diffusible, passes out through the plug the more rapidly of the two.

67.]

DIFFUSION OF GASES.

113

A still more remarkable result may be obtained by the use of an apparatus like the following:-A porons tube, such as the long stem of a tobacco-pipe, is fixed, like the inner tube of a Liebig's refrigerator, by means of perforated corks, within a glass or metallic tube a few centimetres less in length, and about 38mm. (1 inch) in diameter. A second quill tube is made to pass through one of the corks, and affords the means of communication between the annular space and the vacuum of an air-pump. If a vacuum be maintained in the outer tube, and a current of the mixed gases be transmitted through the tobacco-pipe, a portion of this gas will be drained off; whilst another portion will pass off at the open extremity of the porous tube, and may be collected. The stream of gas diminishes as it proceeds, the more diffusible gas being drained away the most rapidly; and the more slowly the gas is passed though this atmolyser, as the apparatus is termed by Graham, the more concentrated does the heavier gas become.

For example, a mixture of 1 volume of oxygen and 2 of hydrogen was transmitted at the rate of 9 litres per hour: o 45 litre of the mixed gas was collected. Instead of O 33 3, H 66'7, it now contained O 90'7, and H 9'3. It was no longer explosive, but rekindled a glowing match.

Since all gases expand equally (134) by the action of equal additions of heat, their relative densities are preserved, and the relative velocities of diffusion are therefore preserved also, whatever the temperature, provided that both gases be heated equally. The rate of diffusion of equal volumes of different gases becomes, as we might expect, accelerated by a rise of temperature; for by heat all gases are rendered specifically lighter; but the rate of diffusion does not increase so rapidly as the direct expansion of gases by heat. Consequently the same absolute weight of any gas will be diffused more rapidly at a low than at a high temperature.

The process of diffusion is one which is continually performing an important part in the atmosphere around us. Accumulations of gases, which are unfit for the support of animal and vegetable life, are by its means silently and speedily dispersed, so that this process contributes largely to maintain that uniformity in the composition of the aërial ocean which is so essential to the comfort and health of the animal creation. Respiration itself, but for the process of diffusion, would fail in its appointed end, of rapidly renewing to the lungs a fresh supply of air, in place of that which has been rendered unfit for the support of life by the chemical changes which it has undergone.

Graham applied to gases a method resembling that of jar diffusion employed for liquids (58). A glass cylinder of 0'57 metre high (22:44 inches) was filled for the lower tenth of its volume with carbonic anhydride, at 16° C., the remainder being filled with air. The air in the upper tenth of the jar was then examined after the lapse of a certain time for carbonic anhydride. After an interval of five minutes, o'4 per cent. in one experiment, and in a second 032 per cent. of this gas was found in this stratum, and in seven minutes the amount of CO, was nearly 1 per cent.; so that it appears that about 1 per cent. of carbonic anhydride will, under these circumstances, become diffused to a