410 Mr. T. Graham on the Absorption and 



To produce this concentration of oxygen, it is quite necessary 

 that the operation be interrupted at an early stage, as was done 

 in the last experiment ; otherwise the oxygen diminishes again in 

 proportion to the nitrogen, falling at last to the normal propor- 

 tion of 21 per cent, as in the external air. Thus a balloon in- 

 flated by carbonic acid to 150 millims. in diameter, was found 

 to lose nearly all its carbonic acid in the course of twenty- four 

 hours. It gave 150 cub. centims. of gas after treatment with 

 caustic potash. This was air of the composition, 



Oxygen 22*6 



Nitrogen 77*4 



KXK) 



and exhibited therefore no material augmentation in the propor- 

 tion of oxygen. 



It may be inferred from the familiar fact that air dissolved in 

 water contains so high a proportion as 30 per cent, of oxygen, 

 that if carbonic acid gas were divided from atmospheric air by a 

 film of water, the former gas would come to be charged through 

 the film with air bearing the same high proportion of 30 per 

 cent, of oxygen. But it is not easy to imitate this experiment 

 unless the dividing film is supported by a membrane of some 

 sort. The air from the atmosphere, which entered a fresh ox- 

 bladder preserved humid and inflated with carbonic acid, was 

 found to possess 24*65 per cent, of oxygen to 75*35 of nitrogen, 

 which is but a small increase in the proportion of oxygen. But 

 the thickness of the membrane here was too great, and other 

 circumstances of the experiment were unfavourable. 



A balloon of rubber inflated to 150 millims. in diameter with 

 carbonic acid was submerged in water, at 22° C, for forty-eight 

 hours. Only a small portion of carbonic acid remained in the 

 residual gas, which, after being washed with potash, consisted of 



Oxygen . . . . 25*77 

 Nitrogen .... 74*23 



100*00 



2. With the colloid septum properly supported, as by a stucco 

 plate in the diffusiometer covered by a film of rubber (p. 403), a 

 considerable separation of mixed gases can be effected. The 

 constituents of atmospheric air appear to be carried through a 

 film of rubber into a vacuum, nearly in the same relative propor- 

 tion as the same gases penetrate singly (p. 405). The veloci- 

 ties of nitrogen and oxygen passing separately were observed to 

 be as 1 to 2*556, and hence by calculation, 



