250 INFLAMMABLE GAS. 



Experiments From this table, it appears that the ditFerent experiments by 

 and observa- no means agree with each other. In the second, for example^ 

 flammable gas *^® oxygen which disappeared was more than double of the 

 fronapeat. gas consumed, while, in the other two it was considerably less 

 than the bulk of inflammable gas which disappeared. These 

 anomalies, which appear at first sight irreconcileable, are in 

 some measure accounted for by the subsequent experiments. 

 In the mean time they shew us that common air cannot be em- 

 ployed in the analysis of the inflammable gas from peat. 



8. If we substitute oxygen gas for common air, the combus- 

 tion is more violent, and the detonation louder. If the pro- 

 portion of oxygen be small compared with that of the inflam- 

 mable gas, scarcely any of it can be detected in the residuary 

 gas, after the detonation is over, though only a comparatively 

 small portion of the inflammable gas is consumed. Hence, if 

 we add a new portion of oxygen to the residuary gas, the mix- 

 ture will detonate a second time, as loudly as at first. This 

 double combustion continues till the oxygen amounts to about 

 two thirds of the inflammable gas. If it be increased beyond 

 that proportion, (he residuary gas becomes incapable of burn- 

 ing, with what portion soever of new oxygen it may be 

 mixed. 



9. The first set of experiments were made with oxygen ob- 

 tained from the black oxide of manganese. It was very im- 

 pure, containing very nearly half its weight of azote. Two 

 diflferent quantities of this oxygen were employed. The first 

 was composed of 



57.9 oxygen 

 42.1 azote 



100 

 The second portion was composed of 

 47.89 oxygen 

 52.11 azote 



100 

 To those experiments, in which the first portion of oxygen 

 was employed, the letter a is prefixed ; while those in which 

 the second portion was employed are distinguished by the 

 letter b, The inflammable gas was the same as that em- 

 ployed 



