Mitchell on the Penetrativeness of Fluids. 103 



As in all the previous experiments, different gases were placed in 

 comparison, I placed the same gas on both sides, and expected, for 

 the * sufficient reason,' no change. The experiment accorded with 

 expectation. The membrane remained stationary. 



The circumstances essential to the transmission of gases through 

 the membrane formed an interesting subject of inquiry. 



My first attempt was to produce a vacuum, by placing the gas in 

 a bottle, and exhausting, by means of the air-pump, the bell-glass 

 which covered it. The gases effected their escape from the bottles 

 thus treated, with a velocity proportional to the rate of permeation 

 already ascertained ; sulphuretted hydrogen passed out more rapidly 

 than carbonic acid, and that than hydrogen. Still as some air is 

 always found in an exhausted receiver on the finest air-pump, I 

 passed a tube containing carbonic, acid into a Torricellian vacuum, 

 where it very speedily escaped and caused the descent of the mer- 

 cury. Even this experiment could not prove perfectly satisfactory, 

 as mercurial vapour occupies the barometric vacuum. A perfectly 

 empty bag carefully closed was placed in carbonic acid and nitrous 

 oxide successively, without undergoing the slightest inflation. If a 

 very small portion of any kind of air remained in the bag, inflation 

 followed, provided the bag were exposed to a different gas. 



By another arrangement I obtained my object more unexception- 

 ably. Having found, by inverting a bottle holding confined gas, 

 and thus plunging it into mercury, that no gas escaped, and that 

 consequently mercury could not promote or sustain the permeation of 

 the gas, I reached my object by the following means. Closing a 

 tall cylindrical lamp-glass at one end with gum elastic, and filling it 

 with mercury, it was placed, so filled, on the shelf of the mercurial 

 trough, having the end closed by the membrane uppermost. 

 Through this fine film the mercury could be plainly seen in close 

 contact with its under surface, while the deep depression of the 

 membrane showed the power of the column of mercury by which 

 it was drawn down. By leaving it in the air, or by placing over it. 

 a bell-glass of any gas, more slowly, but at their settled rates, the 

 gases penetrated the membrane and accumulated in the cylinder, 

 thus permitting the descent of the mercury. The process continued 

 long after the mercury had abandoned the surface of the membrane, 

 and the space was occupied by the gas, in, of course, a rarefied 

 state. 



It became then evident, that anything which could remove the 

 gas from the surface of the membrane at which it had arrived 

 by penetration, would continue its transmission. Of course then 



