OF PHLOGISTON. 33 



that of the atmofphere, the phlogiftlcated air that is 

 found by heating fteain in a copper veflel muft have been 

 formed from phlogiilon in the copper, and the pure part 

 of the air contained in the water ; and whenever I have 

 heated water in this manner and have kept it a confider- 

 able time in the form of fteam, I have found a quantity 

 of air completely phlogiftlcated, and the longer 1 kept 

 it in this ftate the more of this air I found. 



I have obferved that when metals are calcined in com- 

 mon air over water, the air is always diminifhed, and if 

 it be done over lime water, fixed air is produced. If the 

 calcination be continued after the greateft diminution of 

 the air, it will be increafed by an addition of inflamma- 

 ble air. If this inflammable air came from the decom- 

 pofition of the water, the water over which the procefs 

 was made would either be acid, or contain pure air, but 

 this is never the cafe. This water is both free from all 

 acidity, and gives out air lefs pure than that of the at- 

 mofphere. Alfo the air confined in the fame phial with 

 it is lefs pure than that of the atmofphere. If the oxy- 

 gen of the water entered into the calx that is formed, 

 hydrogen, or inflammable air, ought, according to the 

 new theory, to be formed, which it is not. 



Alfo air from water in which mercury has been agitat- 

 ed is confiderably worfe than common air. A candle 

 went out in it. Had the black powder which is formed 

 in this procefs been owing to the decompoiition of the 

 water, fince this powder is mercury fuper-phlogifticated, 

 the remaining water would have been in a ftate of oxy- 

 genation ; and therefore the air expofed to it would have 

 been purer than that of the atmofphere. 



It is faid that metals become calces by imbibing oxy- 

 gen ; but no oxygen has yet been difcovered in finery cin- 

 der, and very little, if any, in flowers of zinc. If minium 

 or red precipitate, be diflTolved in marine acid, none of the 



Vol. V. E dephlo- 



