INFLAMMABLE GAS. 253 



Lei us endeavour, then, from the preceiling experinuents, Experiments 



to ascertain the constituents of this new gas. The reasoning *'^."'^ observa- 



ft,.,,, . ii-z-ii tions on the in- 



om which these constituents are redaced, is founded on a flammable gas 



hypothesis not yet strictly demonstrated, though sufficient!)' i^om peat. 

 probable to be admitted by chemists : the hypothesis is, that 

 when a mixture of the inflammable gas and oxygen are burnt, 

 all that portion of both which disappears is converted into 

 water and carbonic acid. The proportion of carbonic acid 

 formed is known from the experiments, while the proportion 

 of water is deduced from it in the following manner : When 

 oxygen gas isconverted into carbonic add, its bulk is not sen- 

 sibly altered; therefore, the quantity of carbonic acid formed 

 being subtracted from the quantity of oxygen consumed, leaves 

 a remainder of oxygen gas which entered into combustion, 

 but did not form carbonic acid. It is presumed that the re- 

 mainder went to the formation of water. It must, therefore, 

 have combined with the hydrogen contained in the inflam- 

 mable gas. Now, to obtain the weight of this hydrogen, it is 

 only necessary 'to know, that when oxygen is burnt with 

 hydrogen, it combines with very nearly twice its bulk of that 

 inflammable gas. 



Having thus obtained the quantity of carbonic acid and of 

 water, formed by the combustion of the gas, as the carbon in 

 the one and the hydrogen in the other were furnished by the 

 inflammable gas, while the oxygen was furnished by the oxygen 

 gas present, we add the weight of that carbon and hydrogen 

 together, and compare it with the weight of the inflammable 

 gas consumed. If the two weights are equal, we conclude 

 that the inflammable gas was composed of the proportion of 

 carbon and hydrogen obtained by the experiments. But if the 

 weight of the gas be greater than that of the carbon and hy- 

 drogen, we are obliged to have recourse to a new hypothesis, 

 and to suppose that the difference of weight is owing to a 

 portion of oxygen and hydrogen present in the gas, which 

 combined during the combustion, and formed water. The 

 proportion of these two substances deduced from the hypothe-» 

 sis, is added to the hydrogen and carbon previously obtained : 

 thus making up the whole weight of gas, and giving us the 

 constituents. 



From this account of the mode of analysing these gases, it 



is obvious, that it is liable to some degree of uncertainty. But 



Y i the 



