ON GASEOUS EXPLOSIONS. 193 
the ‘ Philosophical Transactions of the Royal Society ’ dealing with the 
first of these matters. The rest of this work has not yet been pub- 
lished, but an abstract of the results obtained is given in the body of 
this Report. Professor Dalby has continued his measurements of the 
suction temperature in gas-engines and of the gas temperature reached 
in compressing and expanding air. Professor Watson has read a paper 
on the measurement of air-flow by means of an orifice, the results of 
which make available for use an accurate and simple method of 
measuring the supply of air to a gas-engine, and will therefore be of 
great value to those engaged on experimental work on such engines. 
In this Report the Committee propose to give a short review of the 
present state of knowledge with regard to the heat-flow from the work- 
ing substance of a gas-engine into the cylinder walls. It is unnecessary 
to insist on the importance to practical designers of this side of the 
theory of internal combustion engines. It is now fully recognised that 
a great part of the difficulties experienced in the construction and 
working of these engines is ultimately due to heat-flow, and the sub- 
ject has been brought into special prominence in recent years by the 
introduction of Jarge cylinders in which these difficulties have only 
partially been overcome. 
The rate of flow of heat from the gas to any part of the walls at 
each instant of time depends upon the then state of the gas as regards 
temperature, density, and motion, and also on the temperature and 
condition of the wall surface. It differs widely at different points of 
the expansion stroke, being far greater just after firing, when the gas 
is at a high temperature and highly compressed, than towards the end 
of expansion. There will, however, be a certain mean rate of heat-flow 
into any patch of the cylinder walls, and heat must be conducted from 
that patch on the whole as fast as it goes in. In order that the heat 
may be conducted away at the required rate there must be a certain 
temperature gradient in the metal, and there will be a corresponding 
mean surface temperature. Superposed on the mean surface tempera- 
ture are variations due to the varying rate of heat-flow at different parts 
of the cycle. The thermal conductivity and capacity for heat of cast 
iron are, however, so large that these variations on a clean metal sur- 
face must be small—a conclusion which has been verified by Coker, 
who found a maximum cyclical change of but 7° C. at a depth of 
0°015 inch in the wall of the combustion chamber of an engine 
running at 240 revolutions per minute. If the metal surface is not 
clean the variation at the surface of the carbon or other deposit may 
be much greater. 
The important practical question is the mean rate at which heat 
goes into each part of the surface, and the resulting mean distribution 
of temperature. The chief problem in designing large gas-engines is 
to control the mean temperature distribution by water-jacketing or 
otherwise in such a way that the metal does not get overstrained by 
unequal expansion nor reach a temperature sufficient to ignite the gas. 
The temperature gradient necessary to sustain the flow of heat from 
the inside of a combustion chamber to the external water is not likely 
to exceed 50° C. per inch. At places where the metal is not thick and 
1912. Oo 
