on the Xchirc of certain Bodies. 9 



\vhole qmn'iitv gf^nenUed, making the usual corrections for 

 teinpciature and pressure, and the poriion i.t hydrogen ori- 

 gin-illy in the tube, and the res.durm, wiirld h.ive been 

 fourteen cubical inches and a hall' at the mean degree of the 

 barometer and thermoincter. Of these, utarly a cubical 

 inch was ammonia and the remainder a gas, of which the 

 portion destruciible by detonation with oxygen, was to the 

 indfestructibie portion, as 2-7 to 1. 



The lo'.vcr part of the tube, where the heat had been in- 

 tense, was found surrounded with potash in a vitreous forjii ; 

 the upper part contained a ccnsiderable quantity of potas- 

 sium. 



In another similar experiment, mr>de expressly for the 

 purposes of ascertainii);:!; thequantitv of potassiua\ recovered, 

 the same elastic products were evolved. The tube was suf- 

 fered to cool, the stop cock being open in contact with 

 mercury, it was filled with mcrcurv, and the iiiLtcuiy dis- 

 placed by water ; when two cubical inches and ihrce quar- 

 ters of hydrogen gas v. ere generated, which proved that at 

 least two grains and "a. half of potassium had been revived. 



Now, it a calculation be made upon the produ* ts in these 

 operations, considering them as nitro>zen and hydrogen, and 

 taking th.e conunon standard of temperature and pre.-<sure, it 

 ■will be found, that by the decompojition «i II cubical in- 

 ches of ammonia equal to 2'05 grains, there is generated a(J 

 cubical inches of nitrogen, ecpial to 1'06 grains, and 9-9 

 cubical inches of hydrogen, which added to th:U disengaged 

 in the first operation equal to about 6- 1 cubical inches, are 

 together equal to '382 grains; and the oxygen added to 35 

 Crains of pot.i^-iiam would be "G grains, and the whole 

 amount is 2*04 grains; and 2-05 — 204 = •()] . But the 

 same quantity of ammonia, decomposed by elOclricity, 

 would have uivcn b'b cubical inches of nitrogen equal to rd 

 grains, and only 1-4 cubical inches of hydngcn * equal to 

 •33, and allowing thfj separation of oxygen in this |)roccs3 

 ill water, it cannot be estimated at more than •! 1 or -12. 



So that if the an.dvsls of an)mt)nia bv electricity at all ap- 

 proaches towards accuracy; in the process ju«i described, 



* See Philuspi^fiic.il Trj;ii.ic»!'.)iii for U03, p. -iO. • ■ • 



thcr« 



