P hlogifiication of Spirit of Kitte . 14a 
it is o no confequence whatever to fay 5 that any particular 
fubftance, imagined to be decompofed, is prefent in a procefs, 
unlefs it can be {hewn that, in that procefs, there are agents 
capable of decompofing it. If mere heat (which is a!l°that 
my procefs requires) would decompofe phlogiflicated air, and 
reduce it to nitrous acid, the tranfmiffion of common air 
(which confifts of dephlogifticated and phlogiflicated air) 
through a red hot tube would have this effeft, which it is well 
known not to have. 
But what I have afferted above is a conclufion which I have 
drawn from comparing the decompofition of dephlogifticated 
air by the two proceffes with nitrous and inflammable air. 
That nitrous air, when mixed with dephlogifticated air, has 
no tendency to produce phlogiflicated air, is evident from the 
almoft total evanefcence of both of them, when they are 
very pure, and mixed in due proportions ; and that nitrous air 
has no effedt on phlogiflicated air is well known. If then the 
firing of dephlogifticated and inflammable air had a tendency 
to decompofe any portion of phlogiflicated air, which fliould 
happen to be mixed with them, lefs would remain after the 
firing of inflammable and impure dephlogifticated air than after 
mixing it with nitrous airj for as the impurities of dephlo- 
gifticated air confift of phlogiflicated air, thofe would difap- 
pear in a greater proportion in the former procefs than in the 
latter.. But by many careful trials I find, that I can reduce 
any kind of dephlogifticated air no farther by a mixture of in- 
flammable air than I can by nitrous air. When the propor- 
tions are well managed, the diminution is as nearly as poffible 
the lame in both the cafes. 
I muft obferve, however, that it requires more nitrous air 
than inflammable air (from iron by fleam) to produce this effect 
V 2 in 
