i8 On the transposition of AIRS. 



tubes with water ; and expofing the top of it to the fire, 

 I found after fome time 2^ ounce meafures of air in it, 

 of the fame quaUty with the preceding. 



All the preceding experiments having been made with 

 the feveral kinds of air unmixed with any other, I was 

 willing to try the effect of a mixture of dephlogiflicated 

 and inflammable air, fuch as explodes with great violence 

 with the flame of a candle or an eledric fpark. In thefc 

 circumftances, however, this mixture did not explode 

 at all, the quantity was unchanged, and the quality was, 

 as before, nearly the fame as that of common air. 



To my great furprife, I found that this mixture of 

 dephlogiilicated and inflammable air did not explode in 

 a red hot gun barrel, a copper tube, or one of filver ; and 

 though the heat was applied ever fo fuddenly. When it 

 was put into a flint glafs tube, it was alfo heated without 

 explofion, but the tube became black, by the calx of 

 lead uniting with the inflammable air; but in a tube of 

 green glafs, in the compofition of which there is no calx 

 of lead, the mixture exploded. Why it fhould not ex- 

 plode in the earthen vefiel, the gun barrel, or the copper 

 and filver tube, I am unable to fay ; but it is probably 

 owing to the dephlogiificated air in the mixture uniting 

 with the metal, and forming a calx rather than with the 

 inflammable air, with which it was mixed. In an experi- 

 ment with the copper tube the quantity of the air came 

 out twice as much as it was when put in. Mixed with 

 an equal quantity of nitrous air, the ftandard v/as 1.4, 

 and it exploded like a mixture of common and inflam- 

 mable air. 



To diverfify this courfe of experiments, I put the dif- 

 ferent kinds of air into earthen retorts fuflSciently air-tight 

 for any common purpofes, and putting the open ends 

 into bafons of water, 1 placed the bulbs near to a fire, 

 where the heat was about that of boiling water, and noted 

 the following refults. 



Fourteen 



