244 ItJFLAMMAlBLE GAS. 



Experiments to become red during the whole process. These variations were 

 tK)ns°on7hcin- """^^"^^^ toasceitain how far the nature of the gas depended 

 fl.immable gas upon the temperature. But the results were not quite satis- 

 irom peat. factory. Sometimes the gas was the same, though the heat 

 differed; and sometimes ihegas varied, though all the circum- 

 stances of the process were as exactly as possible the same. 

 The differences I am disposed to ascribe to variations in the 

 properties of the peat employed. The gas began to come 

 over very speedily. At first it was mixed with much carbonic 

 acid; but the proportion of this gas diminished as the process 

 advanced, though in one instance only it disappeared com- 

 pletely. The quantity of gas obtained from a given bulk of 

 peat was much smaller than what is yielded by the same bulk of 

 woofl or pit-coal, owing probably to the great difference of 

 weight between them. 



I never succeeded in procuring the gas perfectly pure,, as, 

 besides the carbonic acid already mentioned, it always con- 

 tained a portion of common air, varying from l-8th to l-4th 

 of the mixture, according to the process. It was always 

 greatest w hen the cast-iron bottles were used, and least with 

 the stone ware retorts; owing partly to the smaller size of 

 the former, which did not allow me to throw away so great a 

 proportion of the gas which first came over. The presence of 

 common air cannot well be accounted for on any other suppo- 

 sition, *han that the vesse^s were not altogether air-light; for 

 the tubes which conveyed the gas to the water-trough were 

 very well filled. The stone-ware retorts are known already 

 "not to be imiervioiis to air. 



To remove the carbonic acid, 1 at first washed the gas in a 

 large quantify of water; but finding afterwards that a portion 

 of carbonic acid still remained, notwithstanding this process, I 

 removed it, by washing the gas in lime-waier. 



To ascertain the proportion of common atr contained In the 

 gas, 1 employed nitrous gas, according to the method of Mr. 

 Dal ton, af.er iiavir.g convinced myself of the accuracy of that 

 me'htKi by repeated experiments. Into a long narrow tube, 

 graduafec; to lOOthsof a cubic inch, a portion of the gas to 

 be examined, is inirod-iced, and its bulk being noted exactly, a 

 determinate qanti'.y of nitrous gas, previo«isly meaj,ured in a 

 siuiila*- tubr, ii> let up to it. If any common air be present, 

 the bulk of the two gases gradually diminishes. The diminu- 

 tion 



