

ON GASEOUS COMBUSTION. 471 



- . -'• i*l : - 



notion of ' discontinuity ' or variation per saltum in regard to gaseous 

 combustion is still occasionally met with in technical literature. 



A similar error crept into Bunsen's interpretation of the results of his 

 measurements of the pressures produced when either hydrogen or carbon 

 monoxide is exploded with half its own volume of oxygen at atmospheric 

 pressure in a closed vessel. He contended that, in either case, one-third 

 only of the gases combine in the first instance, whereby the temperature 

 of the system is raised to some point between 2844° and 3033°; that it 

 then falls by radiation to 2558°, between which point and 1140° a 

 further one-sixth of the mixture combines, leaving the remaining half 

 to burn as the system cools down to the ordinary temperature. This idea 

 of combustion per saltum was revived again by von Oettingen and von 

 Gernet in 1888, in connection with their photographic researches on the 

 explosion of electrolytic gas, 1 but it has been clearly proved by H. B. 

 Dixon that their observations can be explained on other grounds. 



Bunsen's first measurements of the rates at which flames are propa- 

 gated in gaseous mixtures (namely, 34 metres per second for a mixture 

 2H 3 + 2 , and 1 metre per second for a mixture 2CO + 2 ) have since been 

 shown to apply only to the initial stages of an explosion, where the gases 

 combine with relatively very slow velocities compared with those charac- 

 teristic of ' detonation. ' It was in the year 1881 that Berthelot, and 

 independently Mallard and Le Ohatelier, announced the discovery of the 

 rapid acceleration of the initial velocity of inflammation and the final 

 attainment of the enormously higher constant velocity of the 'explosion 

 wave.' 



Section I. — Ignition Temperatures and the Initial Phases of 

 Gaseous Explosions . 



Chemical change may be determined in a gaseous explosive mixture 

 at a much lower temperature than its ignition point. Thus, if electro- 

 lytic gas be heated in a sealed bulb to a temperature somewhat higher 

 than 400°, the formation of steam can usually be detected after a lapse 

 of a few days. Between 450° and 500° the rate of combination, although 

 considerably greater, would still be insufficient to cause any self-heating 

 of the mixture. If, however, the temperature of the enclosure be 

 further slowly raised, a point (probably about 550°) would soon be 

 reached at which self-heating of the mixture would begin; its tempera- 

 ture would thus be raised above that of the enclosure, and the rate of 

 combination rapidly accelerated until explosive combustion would be set 

 up. The precise temperature at which this would occur would obviously 

 depend upon the amount of slow combustion which had taken place 

 during the preliminary heating-up of the mixture. Thus it follows that 

 the only way of determining the true ignition temperature of such a 

 mixture, undiluted by the products of its own slow combustion, would 

 be either to make the preliminary heating-up period negligibly short, 

 or, better still, to heat separately the combustible gas and the air or 

 oxygen to the ignition temperature before allowing them to mix. 



The work of Victor Meyer and his pupils, 2 as also that of Helier 3 



1 Annakn der Phys. und Ghemie, 33, 58C>. 



2 Ber., 1892, 25, (522 ; 1893, 26, 2421 ; Zeit. Phys. CJtem., 1893, 2, 28. 

 8 Ann. Chim. Phys., 1897 Tvii.], 10, 521. 



