ON GASEOUS EXPLOSIONS. 953 
large variations of temperature in the gas, and referred the change of 
heat-flow not to the metal surface, whose temperature (he thought) might 
for this purpose be supposed to be constant, but to changes in the tempera- 
ture gradient in the layer of gas close to the walls. The controlling in- 
fluence of the condition of the surface layer on the rate of heat-flow from 
the gas appears in many experiments. In a gaseous explosion in a closed 
vessel, for example, the flow of heat from the gas to the walls is at first 
intensely rapid, for the hot flame is brought into immediate contact with 
the cold wall and the heat is drawn from the still hot surface layer, and 
has not very far to travel. As time goes on, however, this layer becomes 
cooled down, and serves as a jacket resisting flow of heat from the hotter 
core within. Study of the curve of cooling after explosion in a closed 
vessel shows that the rate of fall of temperature diminishes in much greater 
proportion than the temperature itself, and the same thing was shown by 
Hopkinson’s experiments with a recording calorimeter, in which the rate 
of heat-flow was directly measured and found to vary nearly inversely as 
the square root of the time elapsed.’ 
The basis of Clerk’s method is a comparison of volume and pressure 
changes in the rapid compression of a gas in a closed cylinder. In adia- 
batic compression the relation between these two quantities is— 
PV” = a constant, 
and the mean value of the volumetric heat over the range of compression 
oe, Ee : , 
Bea where R is the gas constant and equal to 7°75 foot lb. per cubic 
foot or 1°96 calories per gm. molecule. The value of y for true adiabatic 
compression is deduced from the actual indicator diagram by making 
corrections for heat loss in the manner to which reference has already 
been made. Another method of obtaining y-is.to find the relation 
between corresponding pressure and temperature changes in rapid com- 
pression. For adiabatic compression this relation is of the form 
where @ is the absolute temperature. It has been applied for small ranges 
of temperature by Lummer and Pringsheim, and also by Makower, who 
suddenly opened to the atmosphere a large glass globe, containing air at 
a pressure of a few centimetres of mercury above atmosphere, and 
measured the resulting fall of temperature at the centre of the globe by 
means of a platinum thermometer. He obtained in this way the value 
19°3 foot lb. per cubic foot or 4°9 calories per gm. molecule for the 
volumetric heat of air at 20° C., which is certainly within 2 per cent. of 
the truth, without making any correction for heat loss from the air at the 
centre of the globe where the temperature was measured. 
Hopkinson has commenced experiments, described in Note No. 13, 
with the object of applying a similar method to the compression of air in 
a gas-engine cylinder. The engine is motored round, taking in a charge 
of air at every other revolution, compressing it, expanding it, and exhaust- 
ing it in the usual way. A fine platinum wire at the centre of the 
combustion space of the engine measures the temperature of the air at 
tv 
? Proc. Roy, Soc., Series A, vol. Ixxix. p. 138. 
