ON GASEOUS EXPLOSIONS. 199 
cylindrical combustion space without changing its diameter, and to 
keep the mass of gas confined therein substantially constant so that 
the density goes up in inverse proportion to the length of the space. 
At the same time there will be a small rise in the temperature of the 
fired mixture consequent on the higher temperature before firing. 
This, however, would not be very much, amounting to about 100° C. 
for an increase in compression ratio from 4 to 6. 
The average heat-loss per square foot to the surface will increase, 
but not in proportion to the density. On the other hand, the area 
over which that loss is distributed is reduced, but again in a consider- 
ably less proportion than the density. For instance, with an engine 
of equal stroke bore ratio, having a cylindrical combustion chamber, 
the result of increasing the compression ratio from 4 to 6 will be to 
reduce the surface of the combustion chamber by nearly 16 per cent. 
The density is, of course, increased 50 per cent., and if the heat-loss 
increases in a greater ratio than the square root of the density, which 
is almost certainly the case, the effect of this increase of compress:on 
would be to increase the total heat-loss, and therefore to diminish the 
efficiency of the engine relative to the air standard. This in the case 
supposed would not, of course, lead to any reduction in actual efficiency, 
because the greater heat-loss would be more than counterbalanced by 
the increase in the efficiency due to increased expansion. But it is 
clear that if the process were carried sufficiently far the absolute 
efficiency might also be reduced. Some approach to this state of things 
was found by Burstall when the compression exceeded about 7.!° 
The conclusion gained from practical experience, that there is a 
point beyond which it will not pay to increase the compression in the 
gas-engine, is therefore in full accord with the results of laboratory 
experiments on the relation between density and heat-flow. Not only 
is there a point beyond which increasing compression is not followed 
by an increase in efficiency, but before that point is reached the flow 
of heat per unit area is increased to an amount at which trouble will 
begin to arise on account of the difficulty of cooling. It is sometimes 
supposed that the difficulties which arise from pre-ignition when the 
compression is increased too far are due in some way to the rise of 
temperature of the gas consequent on the high adiabatic compression. 
It is very improbable, however, that this has much to do with the 
matter. The real cause of pre-ignition is the overheating of some part 
of the interior surface of the metal or of a deposit thereon, due to 
excessive heat-flow following an increased density. If the metal could 
be kept clean and cool, compression could be carried to very much 
higher values than are now used in practice without any danger of 
pre-ignition. 
The effect of increasing density on heat-loss is, however, a matter 
on which further experimental evidence is needed. A comparison of 
the rates of loss after explosions in a series of cylinders of the same 
eo Proc. Inst. Mech. Eng., 1908, p. 5. See also Professor Callendar’s remarks 
in discussion on paper by Dr. Watson, Proc. Inst. Aut. Eng., vol. iii., p. 457, 
she oe limit of advantageous compression in the petrol motor is estimated 
as 0.5. 
