59 



The second parcel measured 34600 grains, & was reduced by the same means 

 to 30700 grains. 



The quantity of common air contained in the distilling apparatus, allowing 

 for the room occupied by the wood was about 1700 grains ; all of which must 

 have been forced into the inverted bottle along with the first distilled parcel of 

 air and would not be absorbed by the sope leys. 1700 is about -j-V of 12700 ; 

 so that the first distilled air when reduced by the sope leys, contains about 

 tV of its bulk of common air, or is a mixture of about 13 parts of pure facti- 

 tious air to 2 of common air. The last distilled parcel must have been in- 

 tirely free from common air. 



All that air which was absorbed by the sope leys may, I think, be fairly sup- 

 posed to be fixed air. The remaining air of each parcel was inflammable, but 

 required a much greater quantity of common air to make it explode than the 

 inflammable air from metals does : for a vial holding near 1200 grains mea- 

 sure being filled with 1 part of the first distilled air with 24^ of common air, 

 the moisture caught fire on applying a lighted candle to the mouth of the 

 vial & went off" with a small pufi^ ; but when the vial was filled with 1 part 

 of the same air to 2 of common air it would not catch fire. In like manner 

 a mixture of 1 part of the 2nd distilled air with 3 of common air went off 

 with a pufi^, but 1 part of the same air with 2i of common air would not. 

 So that the first distilled air required to be mixed with not less than between 

 2 and 2f times its bulk of common air, & the 2nd distilled air with be- 

 tween 2 1 and 3 times its bulk of common air, before it would explode ; whereas 

 the air from metals, when tried the same way, would explode though mixed 

 with only f its bulk of common air. 



I next tried which of these parcels of air would explode with most force when 

 mixed with considerably more common air than what was sufficient to en- 

 able them to catch fire. For this purpose I mixed some of each of these par- 

 cels of air & also some inflammable air from zinc with 4 times their bulk of 

 common air, & tried them in the same bottle. The first distilled air went 

 ofi^ with the least noise. As for the 2 others, I was uncertain which made 

 most ; but the air from zinc went off" with a sharper sound than the other, & 

 no light could be seen in the bottle ; whereas in the trial of each of the di- 

 stilled airs a small light was seen. 



The experiment was then repeated with mixtures of each of these airs with 

 5 times their bulk of common air. The first distilled air took fire, but scarce 

 any noise. The 2 others went off" as near as I could judge with the same de- 

 gree of noise, the distilled air with a small light visible in the bottle & a 

 duller sound, the air from zinc without any light & a sharper sound. 



It should seem therefore as if the 2nd distilled air contained about as 

 much phlogiston as the air from zinc, but that the first did not contain so 

 much : for when the quantity of common air is considerably more than suffi- 

 cient to consume the whole of the inflammable air, it seems likely that the 

 loudness of the explosion should be in proportion to the quantity of phlogiston 

 contained in the mixture. 



In all these experiments the air was measured in a cylindrical glass with 

 divisions on its sides, in such manner that I think I could not well err more 



