254 INFLAMMABLE GAS. 



Experiments 155 measures of oxygen to 100 of the gas, while the average 

 andobserva- ^f j^j^g second gives us only 55 measures of the oxygen to 100 

 flammable gas of the gas. In the first, the oxygen consumed was one half 

 from peat. greater, and in the second, one half less, than the inflammable 

 gas. The first detonation was always louder than the second, 

 and accompanied by a white flame, while, in the second de- 

 tonatioii, the gas always burns with a blue flame. The dimi- 

 nutions of bulk are always greater after the second detonation 

 than after the first. 



If we examine the individual experiments, we shall find that 

 the proportion of oxygen consumed by the first detonation, is 

 a maximum, when the smallest quantity of oxygen present Is 

 the smallest possible, and that it gradually diminishes as we 

 increase the dose of oxygen. Thus, in the first experiment, 

 of all, the oxygen consumed by the first combustion was to the 

 gas consumed as 338 : 100; whereas, in the last experiment, 

 it was only as 117: 100. In the second combustion, on the 

 contrary, the proportion of oxygen consumed rather increases 

 with the ('ose. In the first experiment of all, it is not quite 

 equal to half the gas, while, in the last, it is rather more than 

 half the inflammable gas consumed. 



If we consider all these circumstances, it will appear ex- 

 tremely probable (hat the eflect of the first combustion is two- 

 fold: that one portion of the gas is burnt, while another com- 

 bines with oxygen without undergoing combustion, and forms 

 either carbonic acid, or some other inflammable gas still un- 

 known. The portion of this new gas formed, diminishes with 

 the doFes of oxygen, because the proportion of gas completely 

 burnt increases. It was doubtless the formation of this new 

 gas, in variable proportions, according to the dose of air em- 

 ployed, that occasioned the variations in the result when the 

 experiments were made with common air. 



As the whole quantity of inflammable gas was never con- 

 sumed in any one of the experiments in which the double de- 

 tonation was employed, and as the residual gas most probably 

 consists, at least in part, of the new inflammable gas formed 

 during the experimeuts, it is obvious that we cannot depend 

 upon these trials for determining correctly the proportion of 

 oxygen which the gas from peat consumes. The average of 

 the whole of them gives us 105.22 measures of oxygen as the 



proportion 



