34,7 



does not think the experiment conclusive, with regard to the exist- 

 ence of oxygen as an ingredient essential to the constitution of am- 



New analytical Researches on the Nature of certain Bodies, being an 

 Appendix to the Bakerian Lecture for 1808. By Humphry Davy, 

 Esq. Sec. R.S. Prof. Chem. R.I. [Phil. Trans. 1809, p. 450.] 



Mr. Davy, having in the experiments described in the late Bakerian 

 lecture, found that a quantity of nitrogen disappeared during the ac- 

 tion of potassium on ammonia, and that it could not be made to re- 

 sume its gaseous state but by the agency of oxygen in water, has 

 been from that time much occupied in determining, with certainty, 

 all the circumstances of the process ; and though he cannot yet speak 

 with precision as to the quantities, he thinks the general results de- 

 cisive with respect to a decomposition of nitrogen having been ef- 

 fected. 



When potassium is heated in ammoniacal gas, it becomes an olive- 

 coloured fusible substance, losing all its metallic properties, a quan- 

 tity of hydrogen is evolved, and ammonia disappears. 



In the Bakerian lecture it was stated, that upon heating the olive- 

 coloured substance a part of the ammonia is recovered ; but when all 

 moisture is carefully excluded, this quantity of ammonia does not 

 amount to one tenth part of the quantity absorbed ; and even this 

 quantity appears to be owing to the almost unavoidable presence of 

 moisture or oxygen. 



In the present experiments, after taking all possible care to ex- 

 clude moisture, since the glass of the vessels might possibly yield 

 oxygen when in contact with potassium, a small tray of platina, con- 

 taining the potassium intended to act upon the ammonia, was intro- 

 duced into a retort containing the gas, and afterwards transferred ex- 

 peditiously into a clean iron or platina tube made air tight, and fur- 

 nished with a stop-cock. 



In one experiment it was ascertained that 12| cubic inches of am- 

 monia are decomposed by nine grains of potassium, and evolve 8J of 

 hydrogen. 



In a second experiment, instead of reserving the gas for experi- 

 ment, an equal quantity of potassium was used, and immediately 

 transferred into an iron tube. The tube being then filled with hy- 

 drogen, and connected with a mercurial apparatus, was heated gra- 

 dually. The quantity of gas collected previously to its acquiring a 

 red heat, amounted to 9| cubic inches, and four inches more were 

 collected by raising the heat to whiteness. Of the former portion, 

 about three fourths of an inch were ammonia ; the remainder of the 

 gas consisted of hydrogen and nitrogen, in the proportion of 8 to 3 ; 

 so that the total quantity of hydrogen collected in this analysis, 

 amounted to 16-^ inches, and the nitrogen to three cubic inches. 

 If the same quantity of ammonia had been decomposed by electricity, 

 it would have yielded 15 of hydrogen, and 6 or 7 of nitrogen; so 



