ABSORBENT AND CONDENSING ACTION OF BODIES. 47 



a tissue. If, for example, we separated, by a medium of this kind, a certain volume of 

 ammonia from a like volume of nitrogen gas, though at the outset of the experiment 

 both the gases might be existing under the same pressure, yet this equality would very 

 rapidly be lost. The absorption of the ammonia taking place with much more rapidity 

 than the nitrogen, it would be presented to this latter gas, not under an equivalent pressure, 

 but in a state of great condensation. Under such circumstances, the transit of a gas 

 fe not, as will be shortly shown, analogous to the case where it flows under common 

 pressure into a vacuum, or into another gas, but the tissue, continually acting as a per- 

 petual condensing engine, brings the two media in contact with each other under ex- 

 tremely different conditions; the one in a compressed state, but ready to exert the 

 whole of its elastic force, the other in a state perhaps little varying from its normal 

 condition. 



152. If tissues really exert a power of this kind, some might inquire how it is that, 

 when a tube closed at one end with such a structure, and filled with mercury, is sunk 

 in the trough to its hydrostatic level, atmospheric air, or any gas to which it is ex- 

 posed, does not pass through and expel the mercury from the tube. If, it might be 

 said, the gas is existing in such a condensed state in the tissue, what is the reason it 

 does not expand, and drive the mercury down ? Experiment proves that this is not the 

 case, but no argument can be drawn from it at all affecting the position here taken ; 

 for, as soon as the gas has gained the under side of the tissue, there is no cause 

 soliciting it to escape any more this way than backward into its own atmosphere ; 

 the pressures each way are equal, and, therefore, counteract each other's effects ; or, 

 rather, the pressures are unequal, for that tending to expel the mercury is resisted by the 

 hydrostatic action of that fluid, and hence no gas can pass into the tube. 



153. We can now understand the rationale of action in Mr. Graham's experiment 

 with plugs of stucco. He found that this material exerted a very slight absorbent 

 power over the gases ; oxygen, hydrogen, nitrogen, &c., not being absorbed in any 

 sensible quantity. When, therefore, he diffused hydrogen into atmospheric air, the 

 stucco not acting mechanically on either of those substances, they were presented to 

 each other under equal and ordinary pressures, and they therefore began to flow into 

 each other just in the same way that they would have flowed into a vacuum ; but very 

 different is the result when we make use of sheets of India-rubber or moistened animal 

 membranes. The stucco plug serves only to make the experiment manageable by op- 

 posing a slight resistance to the escape of the gases, and acting, as I have said before 

 as a temporary valve ; so that, if a diffusion tube be fitted up in Mr. Graham's manner, 

 at the end of the arm of a balance, the gas does not escape so rapidly but that there is 

 time for a very accurate self-adjustment of the apparatus, and the volume of re-entered 

 air can be measured with precision. 



154. It might, perhaps, be objected to the view here taken, that the condensation 

 which some gases experience is more than sufficient to liquefy them; and that, there- 

 fore, they do not act simply as gaseous bodies would do towards each other. This 

 condition, however, when it does take place, appears not to change the resulting phe- 

 nomena, as the following experiment shows. The thermometer being at 38 K, and 



