ON GASEOUS EXPLOSIONS. 195 
flow is not much different from that which would occur in a closed 
vessel of invariable volume having the form and size of the combustion 
chamber, the mixture fired having of course the same composition, 
density, movements, &c., as in the engine. Some allowance must be 
made for the fall of temperature and density which occurs in the initial 
stages of the expansion in the engine, but this will be of the nature 
of a correction, and will not affect the value of the general conclusions 
as to the effect of the various factors in heat-flow which may be drawn 
from closed-vessel experiments. 
The Factors in Heat-flow. 
1. The State of the Walls.—The loss of heat following a gaseous 
explosion in a confined space depends partly on the state of the gas 
and partly on the state of the walls of the enclosure. Dealing first 
with the walls, it is obvious that the higher the surface temperature the 
less rapid will be the flow of heat, which (generally speaking) depends 
on the difference of temperature between the gas and the surface. If 
the metal surface be clean the surface temperature cannot rise by more 
than an insignificant percentage of the temperature difference; but ‘if 
it be coated with a non-conducting layer the exposed surface may be 
heated by the first rush of heat after ignition to such an extent as 
materially to check the subsequent flow. For instance, Hopkinson 
found that a layer of brown paper ;;%,, inch thick pasted inside 
an explosion vessel of 1 cubic foot capacity would reduce the rate 
of heat-flow in the first tenth of a second following maximum pressure 
by more than 30 per cent.* The surface of the paper was not charred, 
but if must for an instant have reached a temperature of several 
hundred degrees Centigrade in order to produce such a result. This 
shows that a badly conducting deposit of carbon in a gas-engine may 
materially reduce heat-flow. Since the high surface temperature occurs 
just after explosion, it will not necessarily cause pre-ignition, though 
of course if the mean temperature be high, so that the surface remains 
red-hot throughout the cycle, it will have that effect. 
2. Radiation from the Gas.—Of more scientific, though perhaps of 
less practical, interest is the reduction in heat-loss which is found when 
the walls are highly polished. This is due to the fact that radiation is 
an important, if not the principal, agent in the transfer of heat from the 
gas to the metal. This matter was dealt with in the third Report of 
the Committee, and it is unnecessary to recapitulate the results there 
given. It has, however, been carried a good deal further by the 
researches of W. T. David, who has investigated the relation between 
the amount of the radiation and the mean temperature of the gas. He 
finds that the rate of loss from this cause varies roughly as the fourth 
power of the absolute temperature. Thus the products of exploding 
a 15 per cent. mixture of coal-gas and air in a cylindrical vessel 1 foot 
by 1 foot radiate about 5 gramme calories per square centimetre 
per second when the absolute temperature is 21009 C. (maximum 
? See Hngineering, September 11, 1908, p. 328. 
® Phil. Trans. Loy. Soc., A., vol. 211, p. 375. 
