GAS BATTERY. 281 



command, and here for the present I leave the gas battery 

 and its theory. 



Postscript, July 7. The length of time which has elapsed 

 between the communication and printing of this paper, as it 

 has enabled me to procure the apparatus fig. 8, will, I trust, 

 be deemed a sufficient reason for my adding a postscript con- 

 taining a few experiments with this form of battery, some of 

 which I cannot but consider important. 



Experiment 28. In order farther to test the opinion ex- 

 pressed, p. 274, six cells of this battery were charged with 

 pure hydrogen and dilute acid in the alternate tubes. When 

 first charged they decomposed water freely, but after the cir- 

 cuit had been closed for a short time, to, exhaust the oxygen 

 of the atmospheric air in solution, they produced no voltaic 

 effect ; the whole series of six would not decompose iodide of 

 potassium ; when, however, a little air was allowed to enter 

 any one of the tubes containing liquid that single cell instantly 

 decomposed the iodide ; three cells were put aside, each in 

 closed circuit ; at the expiration of a week these produced no 

 effect upon a galvanometer, nor was there any gas evolved in 

 the tubes containing liquid ; the stoppers were now taken out, 

 and the liquid in the hydrogen tubes rose to an average of 

 0*3 cubic inch ; each cell contained a pint, and we may there- 

 fore regard 0*15 cubic inch as the amount of oxygen held in 

 solution by this quantity of acidulated water. Were it not 

 for the extreme practical difficulty of perfectly excluding 

 atmospheric air for a long period, the above would furnish an 

 excellent method of examining the quantity of oxygen held 

 in solution by water, and by applying the proper calculus we 

 might read off on our galvanometer scale the infinitesimal 

 bubbles of gas contained in a given bulk of liquid ; if, how- 

 ever, the acid water or the hydrogen contain foreign ingre- 

 dients, a very different result follows, and the liquid, for reasons 

 which will now be obvious, frequently rises considerably in" 

 the hydrogen tubes. 



Experiment 29. I repeated experiment 24 with the battery 

 fig. 8, expecting that as the external air was shut out I should 

 obtain the result more speedily ; I was indeed not without a 

 vague hope of producing some effect upon the nitrogen. The 



