OF THE FLAME IN THE EXPLOSION OF GASES. 
327 
owing to the last uncombined molecules being surrounded by the products ol 
combustion. In either case it is not probable that the passage of a compression-wave 
would greatly influence the chemical combustion. 
{a) If combination were limited by dissociation, e.y., in the explosion of cyanogen 
with twice its volume of oxygen, there miglit be in the vessel some little time after 
the passage of the detonation-wave a mixture of COo, CO, 0^, and No at a tem¬ 
perature (let us say) of 3000° C. and a pressure of 15 atmospheres. Let us take as a 
probable number the figure calculated by Le Chatelier^ for the dissociation of COo 
under the partial pressure of 10 atmospheres and 3000° C. About one-fifth of the 
carbonic acid would be dissociated. Now if gases were comjDressed adiabatically 
(even if we take the molecular specific heat of COo as 20 and the molecular specific 
heat of the diatomic gases as 7 at constant volume and 3000° C.), the rise of tem¬ 
perature due to compression would more than counteract the increase in pressure, and 
the dissociation would tend to increase. 
(h) AVe have found that the addition ot an inert gas to the mixture of ecpial 
volumes of cyanogen and oxygen prolongs the luminosity of the ex})losion.t As the 
inert gas must lower the temperature, and therefore reduce the possible dissociation, 
this prolonged luminosity must be due to a combustion of residual gases not caused 
Ijy dissociation. But in undiluted mixtures this residue is small, as is shown l^y the 
very rapid fall in luminosity, and therefore the action of a pressure-wave on this 
residue can have little chemical efiect. 
We are thus brought to the conclusion that the reflected wave produced by the 
collision of a detonation-wave with the end of the tube is mainly a compression-wave. 
The velocity of the reflexion-wave may be readily compared with that of the 
detonation-wave, their relative velocities being as the sines of the angles made by 
the two waves with the horizontal. These angles have been measured by fixing the 
films on a glass plate (illuminated from behind), and covering the films with tracing- 
paper, on which the lines could then be traced and extended Ijy the ruler. When 
the rehexion-wave was curved, a line was drawn touching the curve as nearly as 
possible at the point where the movement of the gas itself was nil. The velocity of 
the detonation-wave being known, the velocity of the reflexion-wave could readily be 
calculated. By this means the velocities given in the following table were measui-ed, 
each velocity being the mean of those oljtained from four to eight photographs :— 
* ‘ Zeit. Phys. Cliem.,’ vol. 2, p. 782 (1888). 
t ‘ Joum. Chein. Sue.,’ vol. 79, p. 772. 
