ON GASEOUS EXPLOSIONS. 199 
sents the heat produced by combustion during the change. The diff- 
culty is to separate these two during a change of state along the. 
expansion-line. It is probable that combustion is not quite com- 
plete at the point of maximum pressure; in fact some combustion may 
be going on right up to the point at which the exhaust-valve opens. 
If, therefore, two points are taken on the expansion-curve and this 
method of analysis is applied, neglecting O, the heat loss determined 
will obviously be too great. 
An analysis of the diagram by this method will be found in Dr. 
Clerk’s Gustave Canet lecture, and need not, therefore, be further 
pursued. 5 
Attention may be specially drawn to the curves in fig. 12, which 
show the results of trials made for the purpose of ascertaining the 
relationship between the suction temperature and the strength of the 
mixture used and on the speed. When the mixture is 9 parts of air and 
1 part of gas by volume the suction-temperature is about 70° C. at 
a speed of 100 revs. per minute. At 200 revs. per minute the suction 
temperature is increased to 783° C. At the constant speed of 200 revs. 
per minute the temperature gradually increases as the mixture becomes 
richer; with a 10 to 1 mixture the temperature is 75° C., and this 
increases to 964° C. with a 6 to 1 mixture. Af the lower speed the 
change in temperature is almost as great for a corresponding change 
in the mixture, namely from 673° C. to 82° C. With a modern engine 
using a higher compression it is probable that the temperatures would 
be generally higher. Fig. 13 shows the cyclical variation of tempera- 
ture aS determined by Dr. Coker on a more modern engine, and the 
suction temperatures given by him are of the order of 200° C. Dr. 
Coker explains this high suction temperature as being partly due to the 
retention of hot gas and partly due to the long exhaust-pipe which was 
used. 
Dalby and Callendar’s experiments have shown that when using 
rich mixtures the maximum temperature in the cylinder is probably 
about 2000° C., and these results have been confirmed by Coker and 
Scoble. For the mixtures used in ordinary working conditions the 
experiments of Dalby, Callendar, Coker, and Scoble show that the 
temperature is about 18009 C. It is hoped to continue the experiments 
on temperature measurements when engines of more modern construc- 
tion have been installed in the new engine laboratory of the City and 
Guilds (Engineering) College. 
The concentration of research on the accurate measurement of 
temperature is a necessary step towards a more certain knowledge of 
the specific heat of gases at high temperatures; and the vital import- 
ance of this subject is indicated by the brief explanation given above of 
the method by which the determination of heat exchange between the 
working charge and the walls of the cylinder can be made. So far the 
Committee have only been able to present the curves given in fig. 15 
as representing the most reliable data available. The practical use to 
which the curve can be put is illustrated by using the data given 
by it to find the efficiency of an engine working on the Otto 
cycle without loss of heat assuming that the mixture used is that 
