ON GASEOUS EXPLOSIONS. 259 
This increase in efficiency is to be ascribed mainly to the reduction 
of flame temperature—much as in Dugald Clerk’s super-compression 
experiments. A small increase is also to be expected from the increase 
of specific volume which occurs on explosion, and which may amount 
to as much as 10 per cent. volume in extreme cases. Although this 
volume change is insignificant compared with that occurring in ex- 
plosives of solids like gunpowder and cordite, yet it should strictly be 
included in calculations of thermal efficiency. It tends to make the 
efficiency of a petrol engine rather greater than that of a gas-engine, in 
which the compression and heat supply per cubic foot are the same. 
Radiation in Gaseous Explosions. 
The importance of radiation in its bearing upon the calculation of 
volumetric heats from explosion pressures was pointed out in the first 
Report. It is probable that the loss of heat from this cause, or at any 
rate that part of it which occurs during the progress of the flame, is 
independent of the size or surface area of the vessel, arid cannot therefore 
be allowed for by a comparison of vessels of different sizes. Any con- 
siderable amount of radiation of this character will seriously affect the 
values of the volumetric heat obtained by explosion experiments. 
Hopkinson has been investigating this question, and has made 
some progress during the year. He gives some of his results in Note 
No. 10. It will be remembered that at the meeting of the British Asso- 
ciation at Dublin he described the results of some experiments which he 
had made dealing with the effect upon explosion pressures of the nature 
of the surface of the vessel. He coated the inside of the explosion vessel 
with tinfoil, and compared the results of exploding identical mixtures, 
first with the tinfoil brightly polished, and, secondly, when it was covered 
with lamp-black. He found that the difference in maximum pressure 
was inappreciable, but that the rate of fall of pressure during cooling 
was considerably less with the bright lining than with the dark lining. 
This is in accordance with observations which have been made upon the 
effect of polishing the interior of the combustion space of a gas-engine, 
which has been found to result in a perceptible increase in mean pressure. 
During the year Hopkinson and his pupils have been carrying out 
further investigations on this subject. The results described in the last 
paragraph have been fully confirmed, and direct bolometric measurements 
have also been made of the radiation in an explosion. For this purpose 
a small portion of the surface of the explosion vessel was covered with 
thin copper strip, and the rise of resistance of this strip during the 
explosion was recorded by means of a quick period reflecting galvano- 
meter, a record of the pressure being taken at the same time. Com- 
parative experiments were made first with the strip polished as highly 
CH, were not directly measured, but were calculated by means of an empirical 
formula from the measured quantity of CO. Hopkinson and Morse worked out 
. the heat developed by burning the oxygen present to CO, CO,, and H,O in the 
proportions found in the exhaust. The result obtained is rather different from 
that calculated by Watson’s method, and this may account for the fact that Hop- 
kinson found a smaller increase in efficiency with richness of charge than did 
Watson, though the observations on which his calculations were based were in 
good agreement with Watson’s. : 
; : s 
