GAS BATTERY. 275 



exactly 0*5 cubic inch. The apparatus was left in this state for 

 several days, but without any farther effect ; the voltaic action 

 had thus perfectly exhausted the hydrogen and there stopped. 



These experiments are sufficient to prove the accurate 

 eudiometric action of the gas battery. Performed on a large 

 scale this method of eudiometry appears to me likely to possess 

 some advantages. In the eudiometer of Volta, when gases 

 containing oxygen are to be analysed, if the hydrogen added 

 for detonation be impure the result is of course erroneous. 

 The same may be said of the detonation by spongy platinum, 

 or by a wire heated by a voltaic current, which I formerly 

 proposed.* 



If, on the other hand, gases containing hydrogen are to be 

 analysed, errors may result from any impurities in the oxygen 

 which is added, or from inaccuracy in the measurement of 

 either gas ; in the electrolytic method of eudiometry the quan- 

 tity or purity of the hydrogen, in the one case, is of no im- 

 portance ; and in the other the quantity or purity of the 

 oxygen ; that is, provided there be sufficient to exhaust the 

 equivalent to be abstracted from the mixed gas subjected to 

 examination. 



It should be observed, that in these experiments only a 

 single pair of the gas battery can be used, as, if more be 

 employed, the electrolyte is likely to be decomposed, and gas 

 added to the compound.^ The process is rather slow, but I 

 think very sure. Another valuable application of this process 

 is, that it affords (in Experiment 24) a simple method of 

 obtaining nitrogen of unquestionable purity. I know of no 

 method which effects this object so perfectly. All the oxygen 

 of the air is abstracted, as well as that free oxygen which may 

 be contained in the liquid ; and by subsequently introducing 

 a little lime-water into the tube a y the trifling quantity of car- 

 bonic acid may be removed, or the same thing may be at 

 once effected by using caustic potash as the electrolyte in the 

 apparatus, fig. 12. 



Probably many other applications of the gas battery may 

 suggest themselves to other experimentalists, and obviously 



* Phil. Mag., August 1841, p. 99. 

 f See Postscript. 



T 2 



