Page:Philosophical Transactions of the Royal Society A - Volume 184.djvu/109

110 The small differences in the rate of explosion of hydrogen and oxygen, and of ethylene and oxygen at 10° and at 100º show that the effect of ordinary atmospheric changes of temperature may be neglected.

The results obtained, under ordinary conditions of pressure and temperature, with hydrogen and oxygen, with hydrogen and nitrous oxide, and with marsh gas and oxygen, in exact proportions for complete combustion, were in close accordance with the mean results of BERTHELOT ; for ethylene, acetylene, and cyanogen my numbers differed appreciably, but in no case differed by more than 7 per cent. from the rates observed by BERTHELOT. In the following table our measurements for the same mixtures are compared :

TABLE X.--- Mean Velocity of Explosion in Metres per Second. BERTHELOT. Dixon. Hydrogen and oxygen H+0. 2810 2821 Hydrogen and nitrous oxide H+N,0. 2284 2305 Marsh gas and oxygen. 2287 2322 CH +0,- C..+0 Ethylene and oxygen 2210 2364 Acetylene and oxygen. C,H,+ 0. 2482 2391 Cyanogen and oxygen. C,N,+0, 2195 2321

The general agreement between these measurements leaves, I think, no room for doubt about the substantial accuracy of the results.

The results already described show that the formula proposed by BERTHELOT expresses with a close degree of approximation the rates of explosion of several gaseous mixtures. The formula fails for the explosion of carbonic oxide with oxygen or with nitrous oxide. This was to be expected if, in the detonation of carbonic oxide in a long tube, the oxidation is effected indirectly by means of steam--as it is in the ordi- nary combustion of the gas. Measurements of the rate of explosion of carbonic oxide and oxygen in a long tube showed that the rate increased as steam was added to the