164 PROFESSOR H. B. DIXOX OX THE RATE OF EXPLOSION IX GASES, 
nature of the gas, and also probably on the intensity of the initial shock and the 
cooling power of the walls.’'* 
On the other hand, there is some evidence of the breaking up of ethylene in the 
explosion-wave. When ethylene is exploded with an equal volume of oxygen some 
carbon is deposited, and an appreciable quantity of marsh gas is found in the products 
of combustion. It is difficult to account for this formation of marsh gas except by 
the decomposition of the ethylene, either directly or by partial oxidation :— 
(i.) an,. . = c + cn,. 
(ii.) -f O = CO + CH,. 
When ethylene is exploded with less than its own volume of oxygen, the propor¬ 
tion of marsh gas found in the residue increases as the volume of oxygen is diminished.t 
This points to the formation of marsh gas as an intermediate product in the explosion 
of ethylene. 
Again, it is a matter of frequent observation in the ajralysis of coal gas, when the 
residue of marsh gas, hydrogen, and carbonic oxide is exploded in an eudiometer, 
with a large excess of oxygen, that carbon is deposited towards the bottom of the 
tube. When, however, the mixture is sufficiently diluted with air, the marsh gas is 
completely burnt. This fact is easily explained on the assumption that in the 
explosion-wave (which is set up in marsh gas with oxygen but not with air) the 
marsh gas is more or less decomposed in the wave-front, and some of the liberated 
carbon escapes combustion by cooling.;}; 
T do not forget that the question raised concerning the breaking up of yet unburnt 
compounds in the explosion-wave, may also be asked concerning the elementary mole¬ 
cules. What happens, for instance, to the molecules of oxygen in the wave ? AYhen 
a molecule of oxygen reacts with a molecule of hydrogen, is an oxygen atom liberated, 
and, if so, how far does its existence inlluence the rate ? It was with the object of 
obtaining some answer to these questions that experiments were made on the rate of 
explosion of hydrogen with nitrous oxide, in which reaction it is possible that no free 
atoms are liberated :— 
Ho + N2O = HoO + No. 
So far as the experiments go, I cannot trace the influence of liberated atoms of 
oxvo’en. 
^ O 
* Dixon and Haeker, “ On tLe Decomposition by Shock of Endothermic Compounds.” ‘ Manchester 
Memoirs’ [IV.], vol. .5, 1892. 
t Lean and Bone, “ The Behaviour of Ethylene on Explosion with less than its own Volume of 
Oxygen.” ‘ Chem. Soc. Jonrn.,’ 1892. 
X Results which might be explained on the same hypothesis—viz., the decomposition of unburut 
compounds in the wave-front, are afforded by the explosion of ammonia with nitrous oxide, and of 
chlorine with hydrogen. In both eases the I'ate is somewhat higher than might be expected. 
