The Rate of Explosions in Gases. 149 



differ from the average velocity of the molecules in the 

 heated but yet unburnt layer, and this difference must be 

 taken into account. Lastly, there is a correction to be 

 applied for changes of density in the explosion which in 

 part counteracts the foregoing correction. In the oxidation 

 of hydrogen and of carbonic oxide there is a contraction 

 from three volumes to two ; in the burning of cyanogen to 

 carbonic oxide there is an expansion of two volumes to 

 three.* If each layer of the reacting gases is heated at 

 constant volume, a change in the number of molecules must 

 affect the temperature reached. 



§8. The criticisms, which I have ventured to make on 

 Berthelot's method of calculating the mean rate of trans- 

 lation of the gaseous molecules concerned in the propagation 

 of the explosion-wave, tend to show that the rate so calcu- 

 lated must be too small. But in attempting to apply 

 corrections suggested by these criticisms, one is beset with 

 difficulties. In discussing these difficulties in a lecture 

 given before the British Association, at Manchester, in 1887, 

 I remarked : — 



" I think I have made it plain that it is not so simple a matter to 

 calculate the rate of the explosion-wave ; but I also think that by 

 following up this path of experiment we may be led to discover the 

 true relation between the rate of explosion and the molecular rate. 

 Perhaps the chief obstacle to our seeing this relation is the fact that 

 few, if any, chemical changes are known which are sufficiently 

 simple to show it." 



I now advance as a working hypothesis the view that 

 the explosion travels with the velocity of sound in a gas of 

 the maximum temperature and of the density of the gases 

 in the wave-front. 



To obtain (approximately) the mean rate of translation 



• M. Berthelot rigluly takes this change into account in interpreting 

 Bunsen's experiments on the temperature produced in explosions — Ann. Chiin. 

 et Phys., v., xii. , 302. 



