On the Molecular Mobility of Gases .
By THOMAS GRAHAM , F.R.S. , Master of the Mint .
Received May 7 , -Read June 18 , 1863 .
THE molecular mobility of gases will be considered at present chiefly in reference to the passage of gases , under pressure , through a thin porous plate or septum , and to the partial separation of mixed gases which can be effected , as will be shown , by such means .
The investigation arose out of a renewed and somewhat protracted inquiry regarding the diffusion of gases ( which depends upon the same molecular mobility ) , and has afforded certain new results which may prove to be of interest in a theoretical as well as in a practical point of view .
In the Diffusiometer , as first constructed , a plain cylindrical glass tube , about 10 inches in length and rather less than an inch in diameter , was simply closed at one end by a porous plate of plaster of Paris , about one-third of an inch in thickness , and was thus converted into a gas-receiver* .
A superior material for the Fig. 1 .
Fig. 2 .
porous plate has since been found in the artificially compressed graphite of Mr. BROCKEDON , of the quality used for making writingpencils .
This material is sold in London in small cubic masses about 2 inches square .
A cube may easily be cut into slices of a millimetre or two in thickness by means of a saw of steel spring .
By rubbing the surface of the slice without wetting it 4Jl upon a flat sand-stone , the thickness may be further reduced to iJI about one-half of a millimetre .
A circular disc of this graphite , 1 which is like a wafer in thickness but possesses considerable l1 tenacity , is attached by resinous cement to one end of the glass i tube above described , so as to close it and form a diffusiometer -I i ' ( fig. 1 ) .
The tube is filled with hydrogen gas over a mercurial | i| trough , the porosity of the graphite plate being counteracted for the time by covering it tightly with a thin sheet of gutta percha ( fig. 2 ) .
On afterwards removing the latter , gaseous diffusion immediately takes place through the pores of the graphite .
The 1i whole hydrogen will leave the tube in forty minutes or an hour , and is replaced by a much smaller proportion of atmospheric air ( about one-fourth ) , as is to be expected from the law of the diffusion of gases .
During the process , the mercury will rise in the tube , if allowed , forming a column of several inches in height-a fact which illustrates strikingly the intensity of the force with which the interpenetration of different gases is effected .
Native graphite is of a lamellar structure , and appears to have little or no porosity .
It cannot be substituted for the artificial graphite as a diffusion-septum .
Unglazed earthenware comes next in value to graphite for that purpose .
The pores of artificial graphite appear to be really so minute , that a gas in mass cannot penetrate the plate at all .
It seems that molecules only can pass ; and they may be supposed to pass wholly unimpeded by friction , for the smallest pores than can be imagined to exist in the graphite must be tunnels in magnitude to the ultimate atoms of a gaseous body .
The sole motive agency appears to be that intestine movement of molecules which is now generally recognized as an essential property of the gaseous condition of matter .
According to the physical hypothesis now generally received* , a gas is represented as consisting of solid and perfectly elastic spherical particles or atoms , which move in all directions , and are animated with different degrees of velocity in different gases .
Confined in a vessel the moving particles are constantly impinging against its sides and occasionally against each other , and this contact takes place without any loss of motion , owing to the perfect elasticity of the particles .
If the containing vessel be porous , like a diflusiometer , then gas is projected through the open channels , by the atomic motion described , and escapes .
Simultaneously the external air is carried inwards in the same manner , and takes the place of the gas which leaves the vessel .
To this atomic or molecular movement is due the elastic force , with the power to resist compression , possessed by gases .
The molecular movement is accelerated by heat and retarded by cold , the tension of the gas being increased in the first instance and diminished in the second .
Even when the same gas is present both within and without the vessel , or is in contact with both sides of our porous plate , the movement is sustained without abatement-molecules continuing to enter and to leave the vessel in equal number , although nothing of the kind is indicated by change of volume or otherwise .
If the gases in communication be different but possess sensibly the same specific gravity and molecular velocity , as nitrogen and carbonic oxide do , an interchange of molecules also takes place without any change in volume .
With gases opposed of unequal density and molecular velocity , the permeation ceases of course to be equal in both directions .
These observations are preliminary to the consideration of the passage through a graphite plate , in one direction only , of gas under pressure , or under the influence of its own elastic force .
We are to suppose a vacuum to be maintained on one side of the porous septum , and air or any other gas , under a constant pressure , to be in contact with the other side .
Now a gas may pass into a vacuum in three different modes , or in two other modes besides that immediately before us .
X D. BEENOULLI , J. HERAPATH , JOULE , KRONT , CLASCITJS , S CLEAR MAXWELL , and CAZIN .
The merit of reviving this hypothesis in recent times and first applying it to the facts of gaseous diffusion , is fairly due to Mr. HErAPATIH .
See ' Mathematical Physics , ' in two volumes , by JOHN HERAPATHI , Esq. ( 1847 ) .
386 The gas may enter the vacuum by passing through a minute aperture in a thin plate , such as a puncture in platinum foil made by a fine steel point .
The rate of passage of different gases is then regulated by their specific gravities , according to a pneumatic law which was deduced by Professor JOHN ROBISON from TORRICELLI'S wellknown theorem of the velocity of efflux of fluids .
A gas rushes into a vacuum with the velocity which a heavy body would acquire by falling from the height of an atmosphere composed of the gas in question , and supposed to be of uniform density throughout .
The height of the uniform atmosphere would be inversely as the density of the gas , the atmosphere of hydrogen , for instance , sixteen times higher than that of oxygen .
But as the velocity acquired by a heavy body in falling is not directly as the height , but as the square root of the height , the rate of flow of different gases into a vacuum will be inversely as the square root of their respective densities .
The velocity of oxygen being 1 , that of hydrogen will be 4 , the square root of 16 .
This law has been experimentally verified* .
The relative times of the effusion of gases , as I have spoken of it , are similar to those of molecular diffusion ; but it is important to observe that the phenomena of effusion and diffusion are distinct and essentially different in their nature .
The effusion movement affects masses of gas , the diffusion movement affects molecules ; and a gas is usually carried by the former kind of impulse with a velocity many thousand times as great as is demonstrable by the latter .
2 .
If the aperture of efflux be in a plate of increased thickness , and so becomes a tube , the effusion-rates are disturbed .
The rates of flow of different gases , however , assume again a constant ratio to each other when the capillary tube is considerably elongated , when the length exceeds the diameter by at least 4000 times .
These new proportions of efflux are the rates of the " Capillary Transpiration " of gasest .
The rates are found to be the same in a capillary tube composed of copper as they are in glass , and appear to be independent of the material of the capillary .
A film of gas no doubt adheres to the surface of the tube , and the friction is really that of gas upon gas , and is consequently unaffected by the tube-substance .
The rates of transpiration are not governed by specific gravity , and are indeed singularly unlike the rates of effusion .
The transpiration-velocity of oxygen being 1 , that of chlorine is 15 , that of hydrogen 2-26 , of ether vapour the same or nearly the same as that of hydrogen , of nitrogen and carbonic oxide half that of hydrogen , of olefiant gas , ammonia , and cyanogen 2 ( double or nearly double that of oxygen ) , of carbonic acid 1-376 , and of the gas of marshes 1-815 .
In the same gas the velocity of transpiration increases with increased density , whether occasioned by cold or pressure .
The transpiration-ratios of gases appear to be in direct relation with no other known property of the same gases , and they form a class of phenomena remarkably isolated from all else at present known of gases .
* " On the Motion of Gases , " Philosophical Transactions , 1846 , p. 573 .
t Ibidem , p. 591 ; and Philosophical Transactions , 1849 , p , 349 .
32 387 There is one property of transpiration immediately bearing upon permeation of the graphite plate by gases .
The capillary offers to the passage of gas a resistance analogous to that of friction , proportional to the surface , and consequently increasing as the tube or tubes are multiplied in number and diminished in diameter , with the area of discharge preserved constant .
The resistance to the passage of liquid through a capillary was observed by POISEUILLE to be nearly as the fourth power of the diameter of the tube .
In gases the resistance also rapidly increases ; but in what ratio , has not been observed .
The consequence , however , is certain , that as the diameter of the capillaries may be diminished beyond any assignable limit , so the flow may be retarded indefinitely , and caused at last to become too small to be sensible .
We may then have a mass of capillaries of which the passages form a large aggregate , but are individually too small to allow a sensible flow of gas under pressure .
A porous solid mass may possess the same reduced permeability as the congeries of capillary tubes .
Indeed the state of porosity described appears to be more or less closely approached by all loosely aggregated mineral masses , such as lime-plaster , stucco , chalk , baked clay , non-crystalline earthy powders like hydrate of lime or magnesia compacted by pressure , and in the highest degree perhaps by artificial graphite .
3 .
A plate of artificial graphite , although it appears to be practically impermeable to gas by either of the two modes of passage previously described , is readily penetrated by the agency of the molecular or diffusive movement of gases .
This appears on comparing the time required for the passage through the plate of equal volumes of different gases under a constant pressure .
Of the three gases , oxygen , hydrogen , and carbonic acid , the time required for the passage of an equal volume of each through a capillary glass tube , in similar circumstances as to pressure and temperature , was formerly observed to be as follows : Time of capillary transpiration of equal volumes .
Oxygen ... 1 Hydrogen ... .
044 Carbonic acid .
0-72 Now through a plate of graphite , half a millimetre in thickness , the same gases were observed to pass , under a constant pressure of a column of mercury of 100 millimetres in height , in times which are as follows : Time of molecular passage .
Square root of density ( oxygen 1 ) .
Oxygen ... .
11 Hydrogen ... 0-2472 0-2502 Carbonic acid. .
1-1886 1-1760 It appears that the times of passage through the graphite plate have no relation to the capillary transpiration-times of the same gases as first quoted .
The new times in question , however , show a close relation to the square roots of the densities of the 388 respective gases , as is seen in the last Table ; and they so far agree with theoretical times of-diffsion usually ascribed to the same gases .
These results were obtained by means of the graphite diffusiometer already referred to , which was a plain glass tube about 22 millimetres in diameter , closed at one end by the graphite plate .
In order to conduct gas to the upper surface of the graphite plate , a little chamber was formed above the plate , to which the gas was conveyed in a moderate stream by the entrance-tube e ( fig. 3 ) ; while the Fig. 3 .
gas brought in excess was constantly escaping e into the air by the open issue-tube i. The ?
chamber was formed of a short piece of glass tube , about 2 inches in length , cemented over the upper end of the diffusiometer .
The upper opening of this short tube was closed by a cork perforated for the entranceand exit-tubes .
i It will be observed that by this arrangement the upper surface of the graphite plate was constantly swept by a stream of gas , which was s under no additional pressure beyond that of the atmosphere , a free escape being allowed by the exit-tube .
The gas also was always dried before reaching the chamber .
The diffusiometer stood over mercury , and was raised or lowered by the lever movement introduced by Professor BUNSEN in his very exact experiments upon gaseous diffusioni .
To obtain the pressure of 100 millimetres of mercury , the diffusiometer was first entirely filled with mercury and then raised in the trough .
Gas gradually entered till the column of mercury in the tube Ii fell to 100 millimetres .
The mercury was then maintained at this height , by gradually raising the tube in proportion as gas continued to enter and the mercury to fall , so as to maintain a constant difference of level of 100 millimetres , as observed by the graduation inscribed upon the tube itself , between the level of the mercury in the tube and trough .
The experiment consisted in observing the time in seconds which the mercury took to fall 10 millimetre divisions with each gas .
The constant volume of gas which entered was 2-2 cubic centimetres ( 0 1342 cubic inch ) .
Tw6 experiments were made with each gas .
Oxygen entered in 898 and 894 seconds ; mean 896 seconds .
Hydrogen in 222 and 221 seconds ; mean 221-5 seconds .
Carbonic acid in 1070 and 1060 seconds ; mean 1065 seconds .
* BuSEi 's ' Gasometry ' by RoscoE .
389 In such experiments the same gas exists on both sides , and also occupies the pores of the diaphragm .
But the molecular movement within the pores in a downward direction is not fully balanced by the molecular movement in an upward direction , owing to the less tension , by 100 millimetres , of the gas below the diaphragm and within the tube than the gas above and without .
The influx of gas indicates the difference of molecular movement in opposite directions .
Taking the full tension of the gas above the diaphragm at 760 millimetres , that below would be 660 millimetres , and the movement downwards and that upwards are represented by these numbers respectively .
To increase the inequality of tension and favour the passage of gas through the graphite plate , a diffusion-tube was now used , 48 inches in length , or of the dimensions of a barometer-tube , by which a Torricellian vacuum could be commanded .
The pneumatic trough in which this gas-tube was suspended consisted of a pipe of gutta percha of equal length , closed at the bottom by a cork , and widening into a funnel-form at the top .
In one modification of the instrument it was found convenient to cement a capillary glass tube to the side of the glass diffusiometer , within about 15 millimetres of the upper end of the tube .
An opening into the upper part of the glass tube was thus obtained , by means of which the gas contained in the diffusiometer could escape when the latter was depressed in the mercurial trough .
A flexible tube with clip was attached to the capillary tube referred to , so that the latter could be closed .
From the same opening a specimen of the gas contained in the diffusiometer could be drawn when required for examination .
In another and more serviceable modification of this barometrical diffusiometer a large space was obtained above the mercurial column , by surmounting the long glass tube , unprovided with a graphite plate , by a glass jar about half a litre in capacity .
This jar was more correctly a small bell jar ( fig. 4 ) open at top .
It was fitted in an inverted position , as in fig. 5 , to the open end of the long glass tube d , by means of a cork and Fig. 4 .
Fig. 5 .
Fig. 6 .
i d , d cement .
The large upper opening was closed by a circular plate of gutta percha ( fig. 5 ) , about 10 millimetres , or nearly half an inch , in thickness .
This disc of gutta percha had two perforations atf and g ( fig. 6 ) , the former of which was fitted above with a wide glass 390 GASESi tube .
The tube f was closed below by the plate of graphite , and above with a perforated cork carrying a quill tube , e. This quill tube was the entrance-tube for gas , and was accompanied by the usual issue-tube , i. The other aperture in the gutta-percha cover was fitted with a plain quill tube h , which did not descend below the level of the gutta percha , and formed a tube of exit .
No difficulty was Fig. 7 .
found in making all these junctions air-tight , by applying the heated blade of a knife to fuse the gutta percha in contact with the glass .
Gutta percha is indeed of no ordinary value in the construction of pneumatic apparatus .
The graphite plate itself required to be not less than 1 millimetre in thickness , in order to support the pressure of a whole atmosphere , to which it is exposed in the present apparatus .
This barometrical diffusio| meter is supported from above by a cord passing over a pulley , and is duly counterpoised by a hanging weight .
In operating , the first point is to expel the air from the barometer-tube and upper chamber .
The instrument ( fig. 7 ) is sunk completely in the mercurial trough previously described , till the whole is filled , and mercury enters the quill tube of exit , h. The caoutchouc extension of this tube is then closed by a pinch .
The diffusiometer is now elevated 30 or 40 inches , when the mercury sinks in the glass tube till it comes to stand at the barometric height for the time , leaving the upper chamber | entirely vacuous .
The gas to be tried has in the mean time been made to stream over the upper surface of the 'i , graphite plate , exactly as in the experiment with the former diffusiometer .
The graphite is permeated by the gas , and the mercury in the diffusiometer-tube begins to fall , but it now falls slowly , owing to the considerable vacuous space to be filled .
It is allowed to fall about half an inch , and the exact time is then noted , by a watch , when the mercury passes a certain point in the graduation of the tube , and again when the mercury descends to another fixed point an inch or two below the former .
The time of permeation of a certain volume of gas is thus ascertained in seconds .
The experiment is immediately repeated with two or more gases in succession , in similar circumstances as to pressure , and with great care taken to ensure uniformity of temperature during the whole period .
In a series of four experiments made with hydrogen , the mercury fell from 758 to 391 millims. ( 29-9 inches to 27 inches ) in 252 , 256 , 254 , and 256 seconds ; mean 254-5 seconds .
In three experiments with oxygen the mercury fell through the same space in 1019 , 1025 , and 1024 seconds ; mean 1022*7 seconds : 10227 4018 .
254-5 The times of these gases appear therefore to be as I to 4-018 , while the times calculated as being inversely as the square root of the densities of the same gases are as 1 to 4 .
On another day , with a different height of the barometer , four gases were passed through the graphite plate in succession through a somewhat shorter range , namely , from 754 to 685 millims. ( 29-7 to 27 inches ) .
The time of permeation of air was 884 and 885 seconds ; mean 884-5 seconds .
The time of carbonic acid was 1100 and 1106 seconds ; mean 1103 seconds .
The time of oxygen was 936 , 924 , and 930 seconds ; mean 930 seconds .
The time of hydrogen was 229 , 235 , and 335 seconds ; mean 233 seconds .
These times of permeation are in the following proportion : Times of the permeation of equal volumes of gas through graphite .
Oxygen. .
1 Air .
0-9501 Carbonic acid. .
1-1860 Hydrogen ... 02505 These numbers approach so closely to the square roots of the density , or the theoretical diffusion-times of the same gases , namely , oxygen 1 , air 0-9507 , carbonic acid 1-176 , and hydrogen 0-2502 , that they may be held to indicate the prevalence of a common law .
They exclude the idea of capillary transpiration , which gives to the same gases entirely different numbers .
The movement of gases through the graphite plate appears to be solely due to their own proper molecular motion , quite unaided by transpiration .
It seems to be the simplest possible exhibition of the molecular or diffusive movement of gases .
This pure result is to be ascribed to the wonderfully fine ( minute ) porosity of the graphite .
The interstitial spaces appear to be sufficiently small to extinguish capillary transpiration entirely .
The graphite plate is a pneumatic sieve which stops all gaseous matter in mass , and permits molecules only to pass .
It is worth observing what result a plate of more open structure , such as stucco , will give in comparison with graphite .
For the graphite plate , a cylinder of stucco , 12 millims. in thickness , was accordingly substituted , and gas allowed to percolate at both low and high pressures , as in the former experiments with graphite .
1 .
Under a constant pressure of 100 millims. of mercury , gas was allowed to enter through 100 millim. divisions of the diffusiometer .
392 With air , the time in two experiments was 515 , and again 515 seconds .
With hydrogen 178 seconds , and again 178 seconds : 515 =2894 .
2 .
Under a pressure beginning with 710 millims. ( 28 inches ) and ending with 660 millims. ( 26 inches ) , the time with air , was 374 and 375 seconds ; mean 374-5 seconds .
The time with hydrogen was 129 and 130 seconds ; mean 129*5 seconds : ___ -=2 ' 891 .
The stucco cylinder of the preceding experiments had been dried over sulphuric acid , without the application of heat .
It was further desiccated at 600 C. for twenty-four hours , in order to find whether the porosity would be altered .
The ratio of the time of hydrogen to that of air now became I to 2'788 at the lower degree of pressure , and I to 2-744 at the higher degree of pressure .
It will be observed that the theoretical diffusion-ratio of hydrogen to air , which is 1 to 3*80 , is greatly departed from in these experiments with stucco .
The ratio appears ; to be tending to the proportion of the transpiration-times of the same gases , namely5 1 to 2 04 .
In an experiment recorded by BUNSEN , the ratio observed between the times of hydrogen and oxygen in passing , under a small degree of pressure , through stucco dried by heat was so low as 1 to 2'73 , the stucco being probably less dense than in the experiments before us .
With stucco the permeation of gases under pressure appears to be a mixed phenomenon-to some extent molecular diffusion into a vacuum , such as holds with the plate of graphite , but principally capillary transpiration of gas in mass .
The diffusiometer was now closed by a plate of white biscuitware , 2'2 millims. in thickness .
The time of fall at the constant pressure of 100 millims. , through a range of forty divisions of the diffusiometer , was , for air 1210 seconds , for hydrogen 321 seconds .
Air .
1210 ____= 3'769 .
Hydrogen ... 321 The time , again , from 736 to 685millims .
( 29 to 27 inches ) was , for air 685 and 684 seconds ; mean 684'5 seconds ; and for hydrogen 183 , 183 , and 184 seconds ; mean 183'3 seconds .
Air ... .
684S5 = 3.754 .
Hydrogen. .
183-5 The stoneware was evidently of a much closer texture than stucco , and the ratio appears again less influenced by capillary transpiration .
In fact the molecular ratio of I to 3-80 is approached within 1 per cent. Biscuitware therefore appears to be but little inferior to graphite for such experiments , a circumstance which is important , as the latter is not easily procured and cannot be converted into tubes and other convenient forms like plastic clay .
Further , the rate of passage of gas through the plate of graphite appears to be closely proportional to the pressure .
The resistance was increased by augmenting the thickness of the plate to 2 millims. ; and with air and hydrogen at a pressure maintained constant at 50 and 100 millims. , the time was observed that the gas took to enter 10 linear millimetre divisions of the tube .
Seconds .
Ratio .
Air under pressure of 100 millims ... 1925 1 Air under pressure of 50 millims ... 3880 2-015 Hydrogen under pressure of 100 millims. 497 1 Hydrogen under pressure of 50 millims. 1022 2*056 By halving the pressure the time of passage is doubled , or increased somewhat more .
Greater pressures might probably give a rate of passage corresponding more exactly with the pressure .
The ratio between the comparative times of the two gases in the last experiments may also be noticed , the observations having been made in similar circumstances as to pressure and temperature .
Barom. 760 millims. ; At pressure of 50 millims. Barom. 760 millims. ; At pressure of 100 millims. Therm. 12 ' ?
9 C. Therm. 120 ?
9 C. Air.3880 Air ... . .
19253 Hydrogen ... 1022 79Hydrogen .
497 The observation was repeated at the pressure of 100 millims. with barometer at 754 millims. and thermometer at 100 C. Air ... .
1920 Hydrogen ... 498 The velocity of hydrogen appears , as usual , to be nearly 3-8 times that of air ; Vo-6926 7994 An experiment was made at the same time as the -former series upon a mixture of 95 hydrogen and 5 air , which gave an unlooked for result that led to a great deal of inquiry .
It is.known , that such a mixture is effused through an aperture in a fine plate in a time which is as the square root of the density of the mixture , and therefore nearly the arithmetical mean of the two gases effused separately .
But in transpiration by a capillary , a mixture of 95 hydrogen and 5 air requires a considerably longer time than the gases transpired separately .
In fact 5 per cent. of air retards the transpiration of hydrogen nearly as much as 20 per cent. of air would retard the effusion of hydrogen* .
Now the mixture in question permeates the graphite plate in 527-5 seconds , while the calculated mean of the times of the two gases is 562'1 seconds .
The mixture has therefore passed neither in the effusion time , nor in a longer time as it would do by capillary transpiration , but , singular to say , in a time considerably shorter than either .
The gas that came through was found by analysis to be altered in composition .
It contained more hydrogen and less air than the original mixture .
Hence it passed through with increased rapidity .
On consideration it appeared that such a separation of the mixed gases must follow as a consequence of the movement being molecular .
Each gas is impelled by its own peculiar molecular force , which , as has been seen , is capable of causing hydrogen to permeate the graphite plate about 3'8 times as rapidly as air .
Each gas may permeate a graphite plate into a vacuum with the same relative velocity as it diffuses into another gaseous atmosphere , but it remains a question whether the velocities of permeation and diffusion are absolutely as well as relatively the same .
To illustrate this point , hydrogen and air were first allowed to permeate into a vacuum , and then to diffuse into each other , through the same graphite plate , which was 1 millim. in thickness .
The plate was a circular disc of 22 millims. in diameter .
The mercurial column in the barometrical diffusiometer fell from 762 to 685 millims. ( 30 inches to 27 ) with air in 878 seconds , and with hydrogen in 233 seconds .
Air .
8783 Hydrogen 233 72 The volume of gas which produced this effect was found by the calibration of the tube to be 8-85 cub. centims. Hence 1f-22 cub. centim. of the hydrogen entered the diffusiometer in 60 seconds , or one minute .
But the pressure under which the hydrogen gas entered was the mean of 762 to 685 millims. , or 723-5 millims. ; while a whole atmosphere ( the height of the barometer at the time ) was 765 millims. The volume of the gas has therefore to be increased as 723-5 to 765 to give the full action of a vacuum .
The volume becomes 1-289 cub. centim. in one minute .
When the diffusiometer was filled with hydrogen and the gas allowed to diffuse into air , the rise of the mercury was pretty uniform for the first five minutes , being 15-5 millim. divisions in the first two minutes , 7 in the third minute , 7-5 in the fourth minute , and 7 in the fifth minute , making 37 divisions in five minutes .
But as in diffusion 1 air may be supposed to enter the tube for 3-8 hydrogen which escape , the hydrogen which diffused was more than 37 divisions , by 1 , that is , by about 10 divisions .
Hence 3'8 47 divisions of hydrogen have diffused into air in five minutes .
These divisions measured , by the calibration of the tube , 6-215 cub. centims. One-fifth of this amount , that is , 1*243 cub. centim. , diffused in one minute .
The result of the whole is that in one minute there passed of hydrogen through the graphite plate , 1-289 cub. centim. by permeation into a vacuum , 1-243 cub. centim. by diffusion into air .
The numbers indicate a close approach to equality in the velocities of permeation into a vacuum and of diffusion into another gas , through the same porous diaphragm .
The diffusion appears the slower of the two by a small amount ; but this is as it should be , 3H2 895 our estimate of the diffusion-velocity being certainly underrated ; for the initial diffusion , or even the diffusion in the first minute , must obviously be somewhat greater than the average of the first five minutes , which we have taken to represent it the hydrogen necessarily diffusing out in a diminishing progression , or more slowly in proportion as air has entered the diffusiometer .
It is strictly the initial velocity of diffusion ( that of the first second if it could be obtained ) that ought to be compared with the percolation into a vacuum .
In fine , there can be little doubt left on the mind that the permeation through the graphite plate into a vacuum and the diffusion into a gaseous atmosphere , through the same plate , are due to the same inherent mobility of the gaseous molecule .
They are the exhibition of this movement in different circumstances .
In interdiffusion we have two gases moved simultaneously through the passages in opposite directions , each gas under the influence of its own inherent force ; while with gas on one side of the plate and a vacuum on the other side , we have a single gas moving in one direction only .
The latter case may be assimilated to the former if the vacuum be supposed to represent an infinitely light gas .
It will not involve any error , therefore , to speak of both movements as gaseous diffusion , the diffusion of gas into gas ( double diffusion ) in one case , and the diffusion of gas into a vacuum ( single diffusion ) in the other .
The inherent molecular mobility may also be justly spoken of as the diffusibility or diffusive force of gases .
The diffusive mobility of the gaseous molecule is a property of matter fundamental in its nature , and the source of many others .
The rate of diffusibility of any gas has been said to be regulated by its specific gravity , the velocity of diffusion having been observed to vary inversely as the square root of the density of the gas .
This is true , but not in the sense of the diffusibility being determined or caused by specific gravity .
The physical basis is the molecular mobility .
The degree of motion which the molecule possesses regulates the volume which the gas assumes , and is obviously one , if not the only , determining cause of the peculiar specific gravity which the gas enjoys .
If it were possible to increase in a permanent manner the molecular motion of a gas , its specific gravity would be altered , and it would become a lighter gas .
With the density is also associated the equivalent weight of a gaseous element , according to the doctrine of equal combining volumes .
Diffusion of mixed gases into a vacuum , with partial separation-Atmolysis .
Oxygen and hydrogen.-A diffusiometer of the same construction as that described ( fig. 3 , p. 389 ) , with a graphite plate of 1 millim. in thickness , was now employed .
The upper surface of the plate was swept by a current of the mixed gas proceeding from a gas-holder , the excess of gas being allowed to escape into the atmosphere , as usual , by an open exit-tube .
The gas was drawn through the graphite by elevating the diffusiometer containing a column of mercury , from its well , so as to command a partial vacuum in the upper part of the tube .
Care is taken that any gas , left in the upper part of the 396 diffusiometer-tube before the experiment begins , should be of the same composition as the gas to be allowed afterwards to enter , so that , on starting , the gas may be uniform in composition on both sides of the graphite plate .
The height of the mercurial column , which measures the aspirating force of the diffusiometer , is preserved uniform by gradually raising the tube in the mercurial trough in proportion as gas enters and the mercury falls .
The diffusiometer is suspended from the roof of the apartment by a cord passing over a pulley and properly weighted , as in former experiments .
The mixture to be diffused consisted of nearly equal volumes of oxygen and hydrogen .
The effect of different degrees of pressure on the amount of separation produced was first observed .
It will be seen that as the pressure or aspirating force is increased the amount of separation becomes greater .
Barom. 0 759 millim. ; therm. 18 ?
-3 C. Diffusion into a partial vacuum .
Oxygen .
Hydrogen .
Composition of original mixture in 100 parts.49-3 50 7 Diffused by pressure of 100 millims. 47 0 53 Diffused by pressure of 400 millims ... . .
375 62-5 Diffused by pressure of 673 millims. ( mean of 635-710 ) ... 264 73-6 Diffused by pressure of 747 millims. ( mean of 736-759 ) ... 22-8 77-2 In the last observation , or that with the greatest pressure ( 747 millims. ) , the oxygen is reduced to 22f8 per cent. and the hydrogen increased to 77f2 per cent. of the diffused mixture , showing a considerable separation .
The mixed gases appear to make their way through the graphite plate independently , each following its own peculiar rate of diffusion .
But it is only under the aspiration of a complete vacuum that the separation can attain its maximum , and reach the full difference that may exist between the special diffusibilities of the two gases .
The reason is that while we have the original mixture on both sides of the plate , and of equal tension , the gases are not at rest , but diffusion is proceeding as actively through the plate in opposite directions , as if the gases were different or the tension unequal on the two sides .
This is a condition of the molecular mobility of gases ( p. 386 ) .
The tension therefore being supposed to differ by 100 millims. only , as when the gas above the plate was of 759 millims. tension , and below of 659 millims. ( in the first experiment of the last series ) , then 100 volumes only out of 759 of the mixture are subject to separation .
But while these 100 volumes press through they are accompanied by 659 volumes of unchanged mixture .
The latter 659 volumes are replaced by an equal bulk of unchanged mixture diffused from below , so that the volumes are not disturbed by this portion of the molecular interchange .
The amount of separation , then , attainable by transmitting a mixed gas through a porous diaphragm by pressure will be in proportion to the pressure-that is , to the inequality of tension on different sides of the diaphragm .
Oxygen and nitrogen.-The separation of the gases of the atmosphere by transmission through the graphite plate has a peculiar interest .
In an experiment resembling those last described , atmospheric air was swept over the 397 upper surface of a graphite plate having a thickness of 2 millims. The gas that penetrated into the vacuum contained , as was to be expected , the lighter and more diffusible constituent in excess .
It gave by the pyrogallic acid and potash process of LIEBIG , Oxygen ... 20 Nitrogen ... 80 This was an increase in the nitrogen of quite 1 per cent. ; for air , analysed for comparison at the same time and in the same manner , gave oxygen 21'03 and nitrogen 78-97 .
It may be legitimately inferred from the last experiment , that if pure hydrogen in a diffusiometer were allowed to diffuse into the atmosphere through a porous plate , the portion of air which then enters the diffusiometer should also have its composition disturbed .
A diffusion of hydrogen through a graphite plate was interrupted before completion .
The air which had entered was found to consist of Oxygen ... .
1977 Nitrogen ... .
80-23 100-00 The increase of nitrogen is 1-23 per cent. While the nitrogen is increased and the oxygen diminished in the air which makes its way under pressure through the graphite , the converse effect must be produced on the air left behind .
But the latter result of atmolysis cannot be made apparent without a change in the mode of experimenting .
With the view of effecting an increase in the proportion of oxygen,.a volume of air , confined in a jar suspended over mercury , was allowed to communicate through a graphite plate of 2 millims. in thickness , with a vacuum sustained by means of an airpump , the gauge being about 1 inch only below the height of the barometer during the whole time of experimenting .
The jar containing the air to be atmolysed was formed of a plain glass cylinder , open at both ends , and about 400 millims. in height ( 15*75 inches ) .
The upper end was closed by a thick plate of gutta percha cemented on .
This plate was itself penetrated by a wide glass tube , descending about an inch into the jar .
The last tube carried the graphite disc , which was 27 millims. ( 1*04 inch ) in diameter , sufficient to close the lower end of the tube upon which it was cemented .
The other or upper end of the same tube was fitted with a cork and quill tube , and was put into communication with a large bell jar upon the plate of the air-pump .
The permeation was slow , owing to the unusual thickness of the graphite plate , occupying three hours to drain away one-half of the original volume of air in the jar .
The air remaining behind in the jar was examined in a series-of experiments , in which the original volume was reduced to one-half , one-fourth , one-eighth , and one-sixteenth .
The residual air , reduced to one-half , gave in two experiments 21-4 and 21-57 per cent. of oxygen , the air of the atmosphere being by the same analytical process 21 per cent. Reduced to one-fourth of its volume , the residual air gave , in two experiments , 21-95 and 22-01 per cent. , of oxygen .
398 Reduced to one-eighth of its volume , the air gave 22 54 per cent. of-oxygen .
Reduced to one-sixteenth of its volume , the air gave 23*02 per cent. of oxygen .
The proportion of oxygen had therefore increased about one-tenth in the last experiment , where the effect is greatest .
When the numbers are compared , it appears that by a reduction to half its volume the air gains about one-half per cent. of oxygen ; when this last air is reduced to onehalf again , another half per cent. of oxygen is gained , and so on the gain in the proportion of oxygen increasing in an arithmetical ratio , while the volume of air is diminished in a geometrical ratio , or as the powers of the number 2 .
Reduction of 1 volume of air Proportion of Increase of oxygen per cent. oxygen .
To 1 volume ... ... ... ... ... ... ... 21 0 To 05 volume ... ... ... ... ... ... ... 21-48 0*48 To 0-25 volume ... ... ... ... 21198 0.98 .
To 0*125 volume ... ... ... ... ... .
2254 1-54 To 0O0625 volume ... ... ... ... ... 2302 22 2 The densities of oxygen and nitrogen approach too nearly to admit of any considerable separation being effected by this method .
The density of oxygen being taken as 1 , that of nitrogen is 0-8785 .
The square roots of these numbers are 1 and 0 9373 , which are inversely as the diffusive velocity of the two gases .
'Diffusive velocity .
Oxygen .
1 Nitrogen e 1*0669 The velocity of nitrogen therefore exceeds that of oxygen by about 6*7 per cent. Hence by a simple diffusion of a whole volume of air the oxygen could only be increased 6'7 per cent. , according to theory .
In experiments such as the preceding only one-half of the volume of the air is diffused , and consequently only one-half of the stated amount of concentration of oxygen could possibly be produced at each step .
About three-fourths of the theoretical separation is actually obtained , although the apparatus works at an obvious disadvantage from the air within the jar being at rest .
This diffusive method of separation recalls the original observation of DOBEREINER on the escape of hydrogen gas from a fissured jar standing over water , which will always hold its place in scientific history as the starting-point of the experimental study of gaseous diffusion .
That observation proved to be an instance of double diffusion , air entering the jar by the fissure at the same time that hydrogen escaped by it-although , as DOBEREINER looked upon the phenomenon , it was more akin to single diffusion or the passage of gas in one direction only* .
The atmolytic power of other diffusing plates was tested , besides the artificial graphite .
The barometrical diffusiometer already described was closed by a plate of red unglazed Aunnales de Chimie , 1825 .
399 earthenware , 4 millims. in thickness , which was attached to the glass by resinous cement .
Dry air was swept over the upper surface , as in operating with the graphite plate .
With a mercurial column of 340 millims. falling to 200 millims. , the air which entered was found to contain 79'45 per cent. of nitrogen , instead of 79 .
With a column of mer. cury , maintained at 508 millims. in the tube , the air entering contained 79*72 nitrogen , and with a column beginning at 761 millims. , the full barometrical height , and falling to 679 millims. in seven minutes , the air entering contained 80-21 nitrogen .
This is a full degree of separation , exceeding 1 per cent. , while the time was greatly shorter than with graphite .
Thermometer 190 ?
5 C. With a diffusing plate of gypsum ( stucco ) 10 millims. in thickness , the proportion of nitrogen was also increased , although less considerably than with biscuitware .
The standard proportion of nitrogen observed in atmospheric air being 78*99 per cent. , the air drawn into the diffusiometer was as follows : Proportion of nitrogen per cent. In air entering over column of 330-200 millims. mercury .
79*26 In air entering over column 508 millims ... ... 79'32 In air entering over column 761-685 millims ... . .
79.53 In air entering over column 761-685 millims ... . .
79*69 The separation is sufficiently decided , and is certainly remarkable considering the comparatively loose texture of the stucco plate .
The gas entered in the two last experiments in about one minute , which appears too rapid a passage , and not to be attended with increased separation , compared with the immediately preceding experiment , in which the pressure was less and the passage of the gas proportionally slower .
In all such highly porous plates , we have always to apprehend the passage of a large proportion of the gas in the manner of capillary transpiration , where no separation takes place .
It may be concluded that all porous masses , however loose their texture , will have some effect in separating mixed gases moving through them under pressure .
The air entering a room by percolating through a wall of brick or a coat of plaster will thus become richer in nitrogen , in a certain small measure , than the external atmosphere .
The Tube Atmolyser .
In the application of diffusion through a porous septum to separate mixed gases , as a practical analytical method , it is desirable that the process should be more rapid than it can be made with the use of graphite and other diffusing-plates of small size , and also that the process should if possible be a continuous one .
Both objects are attained in a considerable degree by adapting a tube of porous earthenware to the purpose .
Nothing has been found to answer better than the long stalk of a Dutch tobacco-pipe used as the porous tube .
A tube of this description , about 2 feet long and having an internal diameter of 2-5 millims. , is fixed by means of perforated corks within a glass or metallic 400 TMiE MOLECULAR MOBILITY OF GASES .
tube , a few inches less in length and about 1inch in diameter ( e , i fig. 8 ) , as in the construction of a Liebig condenser .
A second quill tube ( v ) is inserted in one of the Fig. 8 .
end corks , and affords the means of communication between the annular space and the vacuum of an air-pump .
The external surface of the corks , and of those portions of the pipe-stalk which project beyond the enclosing tube , should be coated with a resinous varnish , to render them impermeable to air .
Now , a good vacuum being obtained within the outer tube , and sustained by the action of an air-pump , the mixed gas is made to enter and traverse the clay tube .
More or less of gas is drained off through the porous walls and pumped away , while a portion courses on and escapes by the other extremity of the clay tube , where it may be collected .
The stream of gas diminishes as it proceeds , like a river flowing over a pervious bed .
The lighter and more diffusive constituent of the mixed gases is drawn most largely into the vacuum , leaving the denser constituent , in a more concentrated condition , to escape by the exit end of the clay tube .
The more slowly the mixed gas is moved through that tube , the larger the proportion of light gas that is drained off into the vacuum , and the more concentrated does the heavy gas become .
The rate of flow of the mixed gas can be commanded by either discharging it from a gas-holder , or drawing it into a gas-receiver , in either case by a regulated pressure .
To observe the effect of a more or less rapid passage through the tube atmolyser , the impelling pressure was varied so as to allow a constant volume of half a litre of atmospheric air to pass through and be collected in different periods of time .
The clay tube used in these particular experiments was not a tobacco-pipe , but a wide unglazed tube , about 431 millims. ( 17 inches ) long and 19 millims. ( 0 75 inch ) in internal diameter .
It was required to place so wide a tube in a vertical position , and to admit the air by the upper and draw it off by the lower extremity of the tube .
The proportion of oxygen in the half-litre of air collected was as follows : Oxygen per cent. Experiment 1 .
Experiment 2 .
Mean .
When collected in 1 minute ... ... 21-00 ... ... .
When collected in 13 minutes ... ... 22*33 22*25 22*29 When collected in 75 minutes ... ... 2277 23-02 22*89 When collected in 120 minutes 23-25 23-22 23.23 When collected in 304 minutes ... ... 23-54 23-51 23-53 MDcccLxIH~ ... 3 ' - ; 401 3I The proportion of oxygen in the air circulated appears thus to increase with the slowness of its passage through the tube atmolyser .
The proportion of air drawn into the air-pump vacuum must be very large when the time is protracted ; but the additional concentration of oxygen appears small .
The preceding observations being made by means of a porous tube which may be considered wide and of considerable capacity with reference to its internal surface , the experiment was varied by substituting a porous tube about eight times as long , very narrow , and therefore of small internal capacity .
This second atmolyser was composed of twelve ordinary tobacco-pipe stems , each about 10 inches in length and of 1P9 millim. internal diameter , connected together by vulcanized caoutchouc adapters so as to form a single tube .
Having flexible joints , the tube was folded up and placed within a glass cylinder that could be exhausted .
Air was then circulated through this atmolyser by the pressure of several inches of water .
The instrument appeared to work with most advantage when the air delivered at the exit-tube amounted to about one-fourth of a litre per hour .
A volume of 268 cubic centimetres , which had circulated in one hour , was found to contain 24 37 per cent. of oxygen .
The current was then made slower , so that only 108 cub. centims. of gas passed and were collected in one hour , but with little further concentration of the oxygen .
The result , however , is interesting as being the highest concentration of oxygen yet obtained by an instrument of this kind .
The air collected was composed of Oxygen ... .
24-52 Nitrogen ... .
7548 100-00 The increase of oxygen is 3*5 per cent. ; that is , an increase of 16*7 upon 100 oxygen originally present in the air .
With the single pipe-stalk , 24 inches long , first described , the oxygen of atmospheric air was concentrated about 2 per cent. when one litre was transmitted in one hour .
Of 450 cub. centims. of air collected in that time , the composition proved to be Oxygen ... .
23-12 Nitrogen ... .
76-88 100-00 About 9 litres were drawn into the vacuum at the same time .
The separation of the gases of atmospheric air is a severe trial of the powers of the atmolyser , owing to the small difference in the specific gravities of these gases .
But where a great disparity in density exists , the extent of the separation may become very considerable .
Several experiments were made upon a mixture of equal volumes of oxygen and hydrogen carried through the single tube atmolyser , 24 inches in length .
1 .
Of the mixture described , 7*5 litres entered the tube and 0 45 litre was collected in one experiment .
The mixture was composed as follows : 402 Oxygen .
Hydrogen .
Before traversing the atmolyser. .
50 + 50 After traversing the atmolyser. .
92-78 + 7-22 2 .
In another similar experiment , 14 litres of the mixed gas entered the tube and 0 45 litre was delivered in a period of two hours .
The result wasOxygen .
Hydrogen .
Before traversing the atmolyser .
50 + 50 After traversing the atmolyser 95 +5 Here the proportion of hydrogen is reduced from 50 to 5 per cent. 3 .
Of the explosive mixture , consisting of 1 volume oxygen and 2 volumes hydrogen , 9 litres were transmitted and 0 45 litre collected in one hour .
The change effected was found to be as follows : Oxygen .
Hydrogen .
Before traversing the atmolyser .
33-33 + 66*66 After traversing the atmolyser. .
90-7 + 9-3 The result in such experiments is striking , as the gas ceases to be explosive after traversing the porous tube , and a lighted taper burns in it as in pure oxygen .
A mixture of oxygen and hydrogen is not explosive till the hydrogen rises to 11 per cent. To illustrate the analogy of diffusion into a vacuum with diffusion into air , the outer glass tube of the diffuser was now withdrawn , and the porous tube of the instrument was exposed directly to the air of the atmosphere .
A mixture of equal volumes of oxygen and hydrogen was again transmitted at the same rate of velocity as in experiment 1 .
The gas atmolysed and collected was found to consist of Oxygen ... . .
51-75 Hydrogen ... .
5*47 Nitrogen ... .
42-78 100-0.0 And may be represented as containing Oxygen. .
40-38 Hydrogen ... 5-47 Air .
54-15 100-00 A nearly similar concentration of the oxygen of the mixed gas is here observed as appeared in experiment 1 ; but the gas collected is now diluted with air which has entered by diffusion .
The external air manifestly discharges the same function in the latter experiment which the air-pump vacuum discharged in the former experiment .
Interdiffusion of Gases-double diffusion .
The diffusiometer was much improved in construction by Professor BUNSEN , from the application of a lever arrangement to raise and depress the tube in the mercurial 32 403 trough ; but the mass of stucco forming the porous plate in his instrument appears too voluminous , and , from being dried by heat , is liable to detach itself from the walls of the glass tube .
The result obtained of 3-4 for hydrogen , which diverges so far from the theoretical number , is , however , no longer insisted upon by that illustrious physicist .
It is indeed curious that my old experiments generally rather exceeded than fell short of the theoretical number for hydrogen ; V/ 0-0692637994 .
With stucco as the material , the cavities existing in the porous plate form about one-fourth of its whole bulk , and affect sensibly the ratio in question according as they are or are not included in the capacity of the instrument .
Beginning the diffusion always with these cavities , as well as the tube , filled with hydrogen , the numbers now obtained with a stucco plate of 12 millims. in thickness and dried without heat , were 3-783 , 3-8 , and 3-739 when the volume of the cavities of the stucco is added to both the air and hydrogen volumes diffused ; and 3-931 , 3-949 , and 3f883 when such addition is not made to these volumes .
The graphite plate , on the other hand , being very thin , and the volume of its pores too minute to require to be taken into account , its action is not attended with the same uncertainty .
With a graphite plate of 2 millims. in thickness , the number for hydrogen into air was 3-876 , instead of 3-8 ; and for hydrogen into oxygen 4-124 , instead of 4 .
With a graphite plate of 1 millim. in thickness , hydrogen gave 3-993 to air 1 .
With a plate of the same material 0-5 millim. in thickness , the proportional number for hydrogen to air rose to 3-984 , 4-068 , and 4-067 .
An equally considerable departure from the theoretical number was observed when hydrogen was diffused into oxygen or into carbonic acid , instead of air .
All these experiments were made with dry gases and over mercury .
It appears that the numbers are most in accordance with theory when the graphite plate is thick , and the diffusion slow in consequence .
If the diffusion be very rapid , as it is with the thin plates , something like a current is possibly formed within the channels of the graphite , taking the direction of the hydrogen and carrying back in masses a little air , or the slower gas , whatever it may be .
I cannot account otherwise for the slight predominance which the lighter and faster gas appears always to acquire in diffusing through the porous septum .
Interdiffusion of Gases without an intervening septum .
The relative velocity with which different gases diffuse is shown by the diffusiometer , but the absolute velocity of the molecular movement cannot be ascertained by the same instrument .
For that purpose it appears requisite that a gas should be allowed to diffuse into air through a wide opening .
In certain recent experiments , a heavy gas , such as carbonic acid , was allowed to rise by diffusion into a cylindrical column of air , pretty much as the saline solution is allowed to rise into a column of water in my late experiments upon the diffusion of liquids .
This method of gaseous diffusion appears to admit of considerable precision , and deserves to be pursued further .
A glass cylinder of 0-57 metre ( 22-44 inches ) in height had the lower tenth part of its volume occupied with carbonic acid , and the upper ninetenths with air , in a succession of experiments : thermometer 160 Cent. After the lapse of a certain number of minutes , the upper tenth part of the volume was drawn off from the top of the jar and examined for carbonic acid .
Before the carbonic acid appeared above , it had ascended , that is , it had diffused a distance of 0*513 metre , or rather more than half a metre .
After the lapse of 5 minutes , the carbonic acid so found in two experiments amounted to 0'4 and 0'32 per cent. respectively .
In 7 minutes , the carbonic acid observed was 1 02 and 0 90 per cent. ; mean 0 96 per cent. The effect of diffusion is now quite sensible , and it may be said that about 1 per cent. of carbonic acid has diffused to a distance of half a metre in seven minutes .
A portion of carbonic acid has therefore travelled by diffusion at an average rate of 73 millims. per minute .
It may be added that hydrogen was found to diffuse downwards , in air contained in the same cylindrical jar , at the rate of 350 millims. per minute , or about five times as rapidly as the carbonic acid ascended .
In these experiments the glass cylinder was loosely packed with cotton wool , to impede the action of currents in the column of air ; but this precaution was found to be unnecessary , as similar results were afterwards obtained in the absence of the cotton .
To illustrate the regularity of the results , I may complete this statement by exhibiting the proportion of carbonic acid found in the upper stratum already referred to , after the lapse of different periods of time .
Carbonic acid per cent. Experiment 1 .
Experiment 2 .
Mean .
After 5 minutes ... ... 04 032 0*36 After 7 minutes ... ... 12 090 o096 After 10 minutes ... . .
147 156 151 After 15 minutes ... ... 1*70 1P68 1.69 After 20 minutes ... 241 2*69 2 @-55 After 40 minutes 5 ... . .
5.60 5.15 5-37 After 80 minutes ... ... 8-68 8582 8575 In eighty minutes the proportion of carbonic acid had risen to 8*75 per cent. , 10 per cent. being the proportion which would indicate the completion of the process of diffusion .
The same intestine movement must always prevail in the air of the atmosphere , and with even greater velocity , in the proportion of 1 to 1'176 , the relative diffusion-ratios of carbonic acid and air .
It is certainly remarkable that in perfectly still air its molecules should spontaneously alter their position , and move to a distance of half a metre , in any direction , in the course of five or six minutes .
The molecules of hydrogen gas disperse themselves to the distance of a third of a metre in a single minute .
Such a molecular movement may become an agency of considerable power in distributing heat through a volume of gas .
It appears to account for the high convective power observed in hydrogen , the most diffusive of gases .
405

