GASEOUS COMBUSTION AT HIGH PRESSURES. 297 



propagation of flame by conduction) of mixtures of hydrogen and air are much 

 higher than those of methane-air mixtures. Thus, according to LE CHATELIER.* 



Hydrogen- Air Mixtures. 



Hydrogen per cent. . . 10 20 30 40 50 60 70 

 Metres per second . . . 0'60 1'95 3'30 4'37 3'45 2'30 I'lO 



Methane-Air Mixtures. 



Methane per cent 6 8 10 12 14 16 



Metres per second .... 0'03 0'23 0'42 0'61 0'36 O'lO 



Moreover, also, H. B. DIXON has shown that the rate of detonation of electrolytic 

 gas, 2817 metres per second, is greater than the fastest rate for any mixture of 

 methane and oxygen, namely 2528 metres per second for the equimolecular mixture 

 CH 4 +0 2 .t 



But inasmuch as ignition temperatures and rates of explosion are known to be 

 governed chiefly by physical factors, such as heats of combustion, specific heats and 

 molecular weights of products, and (possibly also) ionisation effects, nothing can be 

 deduced from such data as to the relative affinities of the gases for oxygen. On the 

 chemical side, however, certain experiments carried out, in the year 1856, by 

 LANDOLT} in BUNSEN'S laboratory, in which the partly burnt products of a coal gas 

 flame were sucked off for analysis through a fine platinum tube at different vertical 

 heights (0, 10, 20, 30, 40, 50 mm.) along the vertical axis of the flame (100 mm. 

 in height) have been cited as proving the vast superiority of the. affinity of hydrogen 

 over that of methane for oxygen in flames. LANDOLT, it is true, concluded that with 

 regard to the different constituents of coal gas : " Der Wasserstoff ist unter alien 

 Gasarten diejenige, welche am leichtesten verbrennt, es nimmt daher derselbe auch in 

 der Flamme am schnellsten ab ; etwas langsamer verschwindet das Grubengas, und 

 zuletzt kommen die schweren Kohlemvasserstoffe, deren Verbrennung hauptsachlich 

 erst in der oberen Halfte der Flamme vor sich geht." 



The circumstance that LANDOLT employed a platinum tube of narrow bore, fixed 

 along the vertical axis of the flame, for the withdrawal of the partly burnt products, 

 is in itself sufficient to vitiate his conclusion, inasmuch as recent researches in my 

 laboratory have proved that in " surface combustion " (e.g., in contact with firebrick 

 at 500 C.) the usual order of the affinities of various combustible gases for oxygen 

 in flames are entirely reversed. 



* ' Le9ons sur le Carbone,' p. 279. 



t Bakerian Lecture, 'Phil. Trans.,' 1893, vol. 184, pp. 177, 181. 



t ' Habilitationsschrift, Breslau,' 1856 ; 'Pogg. Ann.,' vol. 99, 1856, pp. 389 to 417. 



