* 1 2 X On definite Proportions. 



large scale, and found, that when pure nretals are used, the! 

 potass contains about 15 per cent, of oxygen, the soda 2i 

 to 27." Consequently his determination agrees, especially 

 with respect to soda, prettv well with mine. [And in the 

 last Bakerian Lecture we find tliat lOO parts of potassium 

 ahsorb IS of oxygen, and 100 of sodium 34; affording pure 

 alkalies in a state of extreme moisture. Gilbert.'] 



XVI. Ammonia. 



It would be useless to relate here all the fruitless experi- 

 ments which I performed, nearly in Davy's manner, in 

 order to obtain the base of ammonia in a separate stale. It 

 is utterly impossible to dry an amalj.';am of this basis, which 

 Is formed in a fluid. I therefore endeavoured to obtain it 

 by the operation of dry bodies. For this purpose 1 mixed 

 dry amalgam of potass with dry sal ammoniac, finely pow- 

 dered, in a tubulated retort, provided with a receiver. Both 

 vessels were previously filled with hydrogen, which I had 

 caused to pass ihrouoh a long tube filled with fused mu- 

 riate of lime. The sal ammoniac began, after some time, 

 to bt- d(.compf)sed, and the retort, in an hour and a half, 

 was filled with an amalgam of the consistence of butter* 

 When I wished todisiii the mass, the amalgam subsided 

 into the original vohiuie of the mercury; and when the 

 apparatus was opened, annuoniacal gas and hydrogen gas 

 escap'.'d, with a slight explosion. The neck of the retort 

 was full of drops of watei. This result is easily explained 

 when we consider that sal ammoniac contains water of cry- 

 stallization, amounting, according to the analysis hereafter 

 to be relaied, to 19 per cent. The potassium is oxidated 

 at the expense both of the water and of the ammonia 

 in the salt, and this latter is reduced to a metallic form : 

 but it is converted again at the expense of the water of a 

 neighbouring portion of the salt, into ammonia; so that 

 after the completion of the whole operation, only the oxy- 

 gen of the water has vanished, having been employed in 

 the formation oF the potass, by which the sal ammoniac 

 has been decomposed as a salt. 



In order to separate the newly formed amalgam from the 

 powder of sal ainii:oniac adhering to it, I made an instru- 

 ment of a glass tube, at the ends of which I blew two 

 bulbs, one of them running into a long and thin point : 

 it was filled at the temperature of the freezing point, under 

 boiled quicksilver, with dried hydrogen gas, passed through 

 the point, which was then sealed, and stuck through a thick 

 cork, which had been previously fitted to a bottle, in which 



wai 



