144 ANALYSIS OF AMMONtA. 



receiver in which it had not been sjenerated, and which h;i(i 

 confirmed the likewise been inverted over hoiliiii^ mercury. The Pas left 

 no perceptible resitbuuii, when absorbed by water deprived 

 of air boilincf. In this process, 15 measures of ammonia 

 expended, so as to fill 27 measures; and the hidrogen by- 

 detonation with oxiijen, over water freed as much as pos- 

 sible from air, proved to be to the nitrogen as 73 '8 to 26"2. 

 In the experiment three explosions were made, the oxigen 

 being deficient in the first two; so that no nitrogen could 

 have been condensed in the form of nitric acid. 



Except when precautions of this kind are employed, as I 

 have before noticed, no accurate data can be obtained re- 

 specting the proportions of permanent gasses obtained from 

 animoiiia by electricity. 

 Cautions. ■ When the gas is generated and decomposed over the 



same mercury, there is always a greater expansion than the 

 true one ; and when the mercury is not boiled in the tube, 

 and when common water is used, the nitrogen will be always 

 overrated, unless this errour is counteracted by an opposite 

 errour, that of detonaliug with an excess of oxigen*. 

 Water sup- Dr. Henry had tlie kindness to send me tbe apparatus, 



posed to be j^ which he conceived, at that time, that he had witnessed 

 formed m the ■ ■ n • . j •,- r • l 



decomposi- the tormation or water m tne oecomposition ot ammoaia by 



tioiiofammo- electricity, by his ingenious method of applying hygrome- 



trical tests. 



Appearance of 1 tried one experiment only with it, and in this there 



dedsive uroof seemed to me to be more moisture exhibited in the elastic 



ttf this. matter alter electrization than before, when it was cooled 



by the evaporation of ether: but on maturely considering 



this question, I do not think, that the appearance of mois- 



tute even offers k decided proof of the existence of loosely 



combined oxigen in ammonia. To common hygroinetrical 



tests, water must be less sensible in ammonia than in hidro- 



* It will be seen by Dr. Hrnry's letter, v/hich immediately precetTes 

 this commuiiicaiion, that in repeatinf:; his processes, since this paper was 

 written, he has gained results almost precisely the same as those indicated 

 in the te.xt ; and there is every reason to believe, that 100 of ammonia in 

 volume uniformly bfCoiMc 180;. when decomposed by electricity, and that 

 the gas produced consists ui 100 parts of 74 hidrogen and 26 nitrogen. 

 See Journal, vol. XXIV, p. 358. 



