on the Nature of certain Bodies. 34 1 



of nitrogen to the hydrogen diminished as the pvoeess ad- 

 vanced. 



The first portions contained con'^iderably more^ mtrosjcn 

 in proportion, than the gases ovolved during ihc eleclriza- 

 lion of anjmonia, and the kist portions less. 



I shall give the results ol' an experinieDt, in which the 

 gases produced in distillatl'^n were collected in four diifcrent 

 vessels, and in which every precaution was taken to avoid 

 source? of inaccuracy. 



The barometer was at 29*8, thermometer 65° Fahrenhcif. 



Six grains of potassium absorbed 12 cubic inches of well 

 dried ammonia. The metal was heated in a trav oi platina, 

 and the gas contained in a retort of plate glass. 



5*8 cubical inches of hydrogen were produced. 



The fusible substance'-was distilled in an. iron tube of the 

 capacity of 3 cubical inches and half filled with hydroocn, 

 the adaptor? connected with the mercurial apparatus con- 

 tained -S of common air. 



The first portion of gas collected (the heat being very 

 slowly raised, and long before it had rendered the vessel 

 red,) equalled 7-5 cubical inches. It contained -6 of ami 

 nionia. 7 of the residuum, detonated with 4i of oxygen gas, 

 left a residuum of 4. 



The second portion, equal to 3 cubical inches, contained 

 no ammonia. 7'2 measures of it, detonated \\ilh 3*8 of 

 oxygen, left a residuum of 3*5. 



ihe third portion was equal to 5 cubical inches ; at this 

 time the tube was white hot; it contained no annnonia ; 

 fr5 of it detonated, with 4-'5 of oxvgen, diminished to 2- 5. 



The last portion was a cubical uich and half|, collected 

 when the heat was n)ost intense, -i'j nitasures, with 3'7J 

 of oxygen, left a residuum of 2'8. 



The iron tube contained, after tb.e experiment, (as was 

 ascertained by adniiiting hydrogen whc" :'c was cool,) 2*7 

 of gas; which steme<l of the same composition as the last 

 portion. The adaptors nmst have contained -S of a similar- 

 gas. 



The tube containcf! potash in its lowest part, and in its 

 Upper part potassium, whicli gave by its action uj<on water 

 1^ cubical inch of bvdrocen. 



As the largest quaniity of hydiogc-n is always produced 

 at that period of the process in which the potassium mujt 

 be conceived to be regenerated, and in which the gases 

 l)eing in the nascent slate, its |V)\vtr of neiion upon them 

 fWHild bo greatest, it occurred to me, that if U!troi»f.n was 

 lleeomposvd in ihc operation, there wouUi probably be a 

 V :\ largvr 



