70.] PASSAGE OF GASES THROUGH CAOUTCHOUC. 119 He says gases, under these circumstances, as due to a species of solution. (p. 403), referring to the action of rubber on carbonic anhydride, The first absorption of the gas by rubber must depend upon a kind of chemical affinity subsisting between the material of the gas and substance of rubber, analogous to that attraction which is admitted to exist between a soluble body and its solvent, conducing to solution. Carbonic acid being soluble in ether and volatile oils, it is not wonderful that it is also dissolved by the hydrocarbons of rubber. The rubber being wetted through by the liquefied gas, the latter comes to evaporate into the vacuum, and reappears on the other side of the membrane.' In a series of experiments made upon the absorptive power of rubber upon different gases, Graham employed a diffusion tube about 22mm. in diameter, and a metre in length, closed at the upper end by a thin plate of stucco, and open below. A thin film of rubber from a small caoutchouc balloon was stretched over the plate of stucco, secured with copper wire, and cemented to the glass with gutta percha softened by heat. The tube was filled with mercury, and inverted in a vessel of mercury so as to obtain a Torricellian vacuum, into which different gases were allowed to penetrate, by enclosing the upper end of the tube in a hood of thick vulcanized rubber, provided with an entrance and exit tube for gas. The gas to be operated on was then transmitted steadily through the chamber formed above the film of rubber by the hood; the excess of gas being allowed to escape into the air. The mercurial column gradually fell in the diffusion tube, and the time occupied during a fall of 251 was observed. mm. Taking the time occupied for the passage through the film of a constant volume of carbonic anhydride as 1, the time occupied by the entrance of equal volumes of the following gases was found by Graham to be represented by the numbers subjoined :— No relation between either the diffusibility of a gas or its solubility in water is exhibited by these numbers. The penetration of rubber by gases is much affected by temperature, the softening of the film by heat increasing the rate at which permeation takes place, notwithstanding the increase of volume conferred upon the various gases by rise of temperature. Air, for instance, passed through a septum of rubber at 4 C. with only of the velocity with which it passed through the same septum at 60° C. The difference in the facility with which oxygen and nitrogen pass through caoutchouc suggested the possibility of mechanically separating the constituents of atmospheric air to a considerable extent; and the method by which Graham was enabled to accomplish this object most readily and completely was the following:-A Sprengel's air-pump (Chem. Soc. Journ. 1865 [2], iii. 9) is attached by means of the branch tube, t, to an air-tight gas-bag, B. This bag is best made of a thin but close silk fabric, coated with black rubber upon one side only. The varnished side is turned inwards, and between the two surfaces a double thickness of common felt carpet is placed. A glass quill tube projects inwards for a few inches, and is connected by means of black vulcanized tubing, secured with coils of thin copper wire, and varnished with fused rubber. An exhausting syringe connected with the tube, c, is used to commence the removal of the air, 120 PASSAGE OF GASES THROUGH METALLIC SEPTA. [70. the tube being closed above by the clamp at c; the syringe is removed, and the funnel, F, is filled with mercury, and on relaxing the clamp, the residual air in B gains access to the descending tube, and is swept on by the falling mercury. It may be collected when desired by placing a small receiver, E, over the recurved extremity of the descending tube, which should be of 2.5mm. internal diameter (b inch). After all the contents of the bag had been thus extracted, and the collapse was complete, the Sprengel tube began to throw out air in a slow but very regular manner. The air thus collected was found to have experienced what is regarded by Graham as a process of dialysis. A square metre of such silk (at 20° C.) allowed the passage of about 2:25 cub. centim. air per minute, and the proportion of oxygen was usually from 41 to 42 per cent. If a glowing chip be introduced into such air, it is rekindled. In one observation, at a temperature of 4° C., the proportion of oxygen in the dialysed air was found to rise as high as 47'4 per cent. No doubt, if on a large scale it were possible by such mechanical means to reduce the proportion of nitrogen to one-half of that which exists in air, the concentrated oxy. gen mixture might be applied to many useful purposes. (70 a) Passage of Gases through Metallic Septa.-Something analogous to this action of caoutchouc has been proved by the experiments of Deville and Troost to occur in the case of certain metals-viz., platinum and iron, which they found to become permeable to hydrogen at elevated temperatures in a very remarkable and unexpected manner :-. -A tube of hammered platinum was fitted by means of corks into the axis of a shorter and wider porcelain tube; a slow current of pure and dry hydrogen was directed through the outer tube, whilst a current of dry air was transmitted through the platinum tube. At ordinary temperatures no moisture was observed in the air which passed through the platinum tube; the porcelain tube and the platinum within were then gradually raised to a red heat; at a temperature of 1100° C. (2012° F.) the oxygen contained in the air had entirely PASSAGE OF GASES THROUGH METALLIC SEPTA. 121 70 a.] disappeared, nothing but nitrogen and steam passing out of the platinum tube; whilst at still higher temperatures the moist nitrogen was mixed with hydrogen. As the tube cooled, the same phenomena were observed, but in the inverse order, till at ordinary temperatures no diffusion of hydrogen was perceptible. A cast tube of platinum 2mm. thick gave similar 2" results. (Comptes Rendus, 1863, lvi. 977, and Phil. Mag. 1863 [4], xxvi. 336.) Carbonic anhydride does not experience diffusion in the platinum tube. Analogous experiments upon a tube of soft cast steel drawn down to a tube the sides of which were from 3 to 4mm thick, showed that this material also is porous at high temperatures. In this experiment the steel tube was soldered at each end by means of silver solder to narrow tubes of copper; it was then enclosed in a porcelain tube, to protect it from the fire, and raised gradually to a strong red heat: one end of the copper tubing was connected with an apparatus which disengaged hydrogen; the other terminated in a long glass tube, which was plunged into mercury. After the hydrogen had been transmitted for several hours through the ignited tube, the tube which conveyed the gas was sealed by fusion; and immediately the mercury rose in the glass tube to a height of 740mm. (29'13 inches), so that an almost perfect vacuum was produced by the escape of the hydrogen through the substance of the hot steel tube (Comptes Rendus, 1863, lvii. 965). I sq. This subject has been pursued by Graham in the paper already quoted (Phil. Trans. 1866):-By means of the Sprengel pump attached to one of Deville's tubes, the platinum tube was found to be tight to either atmospheric air or hydrogen at ordinary temperatures, and at all temperatures below a dull red heat; but as soon as the outer porcelain tube became visibly red, the hydrogen passed through the pores of the platinum, and in 7 minutes, 15'47 cub. centim. of gas were collected, of which 15 27.. were hydrogen. The platinum tube which was used in this experiment was joined without solder; it was drawn from a mass of the fused metal, and was I'I millimetre in thickness. metre of such a tube delivered hydrogen, at a temperature approaching a white heat, in quantity of 489 2 cub. centim. per minute: this was nearly 4 times as fast as the passage of hydrogen at 20° C. through a septum of rubber 0.014' thick. Oxygen and nitrogen were tried under similar circumstances, but neither of these gases permeated the platinum. Carbonic anhydride, chlorine, hydrochloric acid, and aqueous vapour were equally unable to penetrate the ignited metal. No hydrogen was found when either of the gases last named was used; therefore it may be inferred that no dissociation of their constituents occurred at this temperature. When either ammonia, coal gas, or hydrosulphuric acid was tried, pure hydrogen was found to have penetrated the tube, but none of the undecomposed gas itself passed through. The experiments were then varied in the following manner :-/ -A porcelain tube was fitted with quill glass tubes and sound corks, and either closed hermeti 122 PASSAGE OF GASES THROUGH METALLIC SEPTA. [70 a. cally, or, when necessary, connected with a gas delivering apparatus at one extremity, and at the other with a Sprengel tube. A quantity of clean platinum wire drawn from fused metal was introduced into the porcelain tube, exhausted, and heated for an hour. Then pure and dry hydrogen was admitted into the porcelain tube at a cherry-red heat, and the temperature allowed to fall gradually, so as to expose the metal for about twenty minutes at a heat a little below redness to an atmosphere of hydrogen. It was then allowed to cool completely, and the excess of hydrogen was swept out by a current of air or of nitrogen. The closed tube was now exhausted in the cold, but no hydrogen came off. The platinum being still retained in a good vacuum, heat was again very gradually applied, and the action of the Sprengel pump maintained. Gas began to come off as soon as visible redness was attained; and in the course of one hour 201 grms. of platinum gave off 2*12.c. of gas, 1'93 of which consisted of hydrogen, and o'19 of nitrogen. The platinum had consequently absorbed one-fifth of its bulk of hydrogen (or nearly half its bulk, if calculated at a red heat) without any sensible change in lustre, or otherwise. Repeated heatings somewhat diminished the volume of hydrogen absorbed by this mass of platinum. Spongy platinum, when similarly treated, absorbed 148 its volume of hydrogen, with a sensible amount also of nitrogen. Sheet platinum from an old crucible absorbed a still larger quantity of hydrogen, the amount exceeding five volumes. The gas thus absorbed appears to be permanently retained. In one experiment a piece of platinum which had been charged with hydrogen was hermetically sealed in a tube with air, and left for two months. No hydrogen escaped into the tube; and on heating the platinum in the manner already described, it gave off, in different experiments, from 2:28 to 3.79 times its bulk of hydrogen. Platinum foil was observed, when heated in hydrogen only to 230° C. for three hours, to absorb 145 its volume of hydrogen, which could not be extracted in vacuo until the temperature was raised to redness. Copper, by similar modes of experiment, was found to absorb from 0.3 to 06 of its bulk of hydrogen. Gold absorbed o 48 its volume of hydrogen, and 03 of carbonic oxide. It also absorbed nitrogen from the air to the extent of about 0'4 of its volume; 86.3 per cent. of the absorbed gas consisting of nitrogen, 53 of oxygen, and the remainder 84, of carbonic anhydride. Silver wire absorbed o'21 its volume of hydrogen, and 0.745 of oxygen. In three different experiments silver sponge (reduced from the oxide) retained 6.15, 8:05, and 747 volumes of oxygen, at all temperatures below incipient redness, but showed no visible tarnish of its surface. Iron wire retained from 0.46 to 0'42 its volume of hydrogen. The same iron, after the expulsion of the hydrogen, was exposed to carbonic oxide: the pure iron was thus found to be capable of taking up at a low red heat, and holding when cold, 415 volumes of carbonic oxide. The wire remained soft, and did not become hard when heated red hot and suddenly cooled, neither was it altered in aspect or in solubility in acids. Indeed, as Graham remarks, 'in the course of its preparation wrought iron may be supposed to occlude six or eight times its volume of carbonic oxide gas, which is carried about ever after.' This diffusion of carbonic oxide through the mass of the iron is regarded by Graham as the ordinary means of distributing carbon through iron in the process of converting it into steel. The Lenarto meteorite was found to contain 2.85 times its volume of gas, of which 85.68 per cent. was hydrogen, 9.86 nitrogen, and 4'46 carbonic oxide (Proc. Roy. Soc. 1867, xv. 502). The most remarkable results upon the absorption of gases by the metals were obtained in the case of palladium. A particular piece of palladium foil was first heated in vacuo to redness, and then exposed, at a heat not exceeding 245° C., to hydrogen, after which it was allowed to cool very slowly. When exposed at ordinary temperatures in vacuo, it gave off no hydrogen; but as soon as heat 70 a.] GENERAL REMARKS ON DIFFUSION. 123 was applied, the gas was evolved freely, and amounted to 5:26 volumes for I volume of the metal. The same piece of foil, when heated for three hours to between 90° and 97° C., absorbed 643'3 volumes of hydrogen; and if the metal, after heating to redness, was allowed to cool in vacuo, it absorbed, even at common temperatures, fully 376 volumes of the gas. No alteration was sensible in the metallic appearance of the foil. Spongy palladium took up 686 volumes of hydrogen, but exhibited no tendency to absorb oxygen or nitrogen. Palladium foil prepared from a piece of the metal which had been fused, absorbed hydrogen much less abundantly; the maximum amount not exceeding 68 volumes. A still larger quantity of hydrogen can be occluded by palladium when a wire of this metal is made the negative pole of a voltaic cell, decomposing water acidulated with sulphuric acid. As much as 936 volumes of hydrogen may thus be absorbed by 1 volume of palladium. During this occlusion, the palladium increases in bulk in the proportion of 100 to 104'908, or sixteen times the expansion it would undergo in being heated from o° to 100° C. On heating the metal charged with hydrogen in vacuo the gas escapes, and the metal contracts to an extent nearly double of that which it had expanded in absorbing the hydrogen thus a wire of palladium 609°144mm long expanded, on absorbing 936 volumes of hydrogen, to a length of 618.923, showing an increase of 9'77 ; mm. or a contraction after heating in vacuo the length was found to be 599'444* from the original length of 97. This contraction is in length only, for the density of the palladium at the conclusion of the experiment is almost exactly : what it was at the commencement. On making a piece of palladium charged with hydrogen the positive pole of the battery, the hydrogen is rapidly converted into water by the nascent oxygen; and by proper arrangements the expansion and contraction of the metal can easily be observed on reversing the current in the cell. Palladium charged with hydrogen is liable to become suddenly hot when exposed to the atmosphere, and the hydrogen removed by oxidation; on heating the end of a piece of charged wire by a spirit-lamp, the hydrogen burns, and the flame travels along the wire. From the measurements of the expansion and contraction of palladium on absorbing and losing hydrogen, Graham has calculated that the density of hydrogen in this condition (or hydrogenium, as he has named it) is between 0.854 and 0-872. The metal when charged with hydrogen, perfectly retains its metallic lustre, and will even receive an impression in a coining press without alteration.. A solid palladium tube 1mm. thick was found to be tight to atmospheric air when heated to redness; but at a temperature of 240° C., hydrogen began to pass through it; and when it was heated to a point just short of redness, the gas passed at the rate of 423 cub. centims. per minute for a square metre of surface. When coal gas was employed instead of pure hydrogen, and the heat was raised to 270°, pure hydrogen, completely free from any smell of coal gas, passed into the interior of the metallic tube. Ether vapour was found to pass through palladium foil at common temperatures, although hydrogen did not pass. From these remarkable investigations Graham concluded that there is a progression in degrees of porosity. He says, there appear to be (1) pores through which gases pass under pressure, or by capillary transpiration, as in dry wood and many minerals; (2) pores through which the gases do not pass under pressure, but pass by their proper molecular movement of diffusion, as in artificial graphite; and (3) pores through which gases pass neither |