ON GASEOUS COMBUSTION. 485 



maximum pressure. Mr. Dugalcl Clerk put forward this suggestion as 

 long ago lis 1886 l in criticising the conclusions of Mallard and Le 

 Chatelier respecting the great increase in the specific heats of steam 

 and of carbon dioxide with temperature. He considered it highly prob- 

 able that combustion extends far into the actual ' cooling period ' in 

 gaseous explosions (and hence the long drawn out ' cooling curve '), so 

 that the system loses a certain part of the heat of combustion before 

 the chemical action is completed. This idea of a continued combustion 

 finds support in H. B. Dixon's photographic researches, and chemists 

 generally will concede its reality in any gaseous combination in which 

 detonation is not determined. But to what extent it may be held to 

 affect the pressures actually recorded by explosions is still a matter of 

 conjecture. 



3. To loss of energy by direct radiation. Thus in the explosion of a 

 mixture of hydrogen and oxygen it is conceivable that the initial action 

 results in the formation of an intensely vibrating molecular complex from 

 which steam issues as the first recognisable product. Some experiments 

 made in 1890 by Robert von Helmholtz - showed that non-luminous 

 hydrocarbon flames radiate about 5 per cent, of the heat of combustion 

 of the gas, and more recent experiments by Professor Callendar and 

 Mr. Nelson show that the heat radiated from an ordinary non-luminous 

 Bunsen flame may amount to between 15 and 20 per cent, of the total 

 heat of combustion, a figure which is in close agreement with the results 

 of experiments carried out by Mr. E. W. Smith under the auspices of 

 the Gas-Heating Research Committee appointed by the Institution of 

 Gas Engineers in conjunction with the University of Leeds. 3 There is, 

 therefore, little doubt but that this cause is truly operative in gaseous 

 explosions. 



4. To dissociation of products (steam and carbon dioxide). In the 

 case of two combining gases producing a dissociable product, it is clear 

 that if the average temperature in the system reaches that at which disso- 

 ciation begins the combustion must be delayed whilst heat escapes 

 from the system by radiation and conduction. Qualitatively the partial 

 dissociation of steam and carbon dioxide has been proved at temperatures 

 which are certainly exceeded by those of explosion flames, but it may 

 be urged that, inasmuch as all experiments upon dissociation have up to 

 the present involved contact with hot solid surfaces, there is no positive 

 evidence that the phenomenon would play any conspicuous part in an 

 unconfined gaseous system. On the other hand, there is direct experi- 

 mental evidence of the attainment of enormously high temperatures in 

 the explosion wave, temperatures which would generally be considered 

 as far beyond that of the initial, or perhaps even of the complete, dis- 

 sociation of steam or of carbon dioxide. Moreover, the fact that the rate 

 of explosion of electrolytic gas is retarded rather more by an excess of 

 oxygen than by a corresponding excess of nitrogen is inconsistent with 

 the supposition of any appreciable dissociation of steam in the explo- 

 sion wave, and photographic records give no evidence of continued 



1 Proc. Inst. Civil Engineers, 85. 



2 Beiblcitler, 14, 589. 



3 See pp. 10-13 of the Committee s Second Report, 1910. 



