ON GASEOUS EXPLOSIONS, 201 
working of a gas-engine seems, however, to require more direct 
methods of investigation than the ordinary tests. During the past 
year Dugald Clerk has applied his method of indicating the engine 
with tripped valves, so as to obtain a ‘ zigzag’ diagram, to the investi- 
gation of this point. During the first expansion line in such a diagram 
there is present the normal amount of turbulence which obtains in 
the ordinary working of the engine; during the second and later 
expansions of the ‘ zigzag’ the turbulence has practically died out. 
We have here obviously a method of considerable delicacy for detecting 
and measuring the effect of turbulence in causing heat-loss on the 
expansion line. Clerk has found that in the compression and expansion 
of air or carbon dioxide without firing, the engine being simply motored 
im?) 
Fie, 1.—Ordinary ignition, a to b, takes 0:037 second ; trapped ignition on third 
compression ; line a’ to b’ takes 0°092 second ; mixture in both cases, 1 vol. 
gas, 9°3 vols. air and other gases. 
round, the rate of heat-loss at a given temperature is greater in the 
first compression after drawing in the charge than in the subsequent 
compressions. 
For the purpose of studying by this method the effect of turbulence 
on heat-loss in the ordinary working stroke of a gas-engine, Clerk 
tried the experiment of drawing in a combustible charge into the 
engine in the ordinary way and then tripping the valves and com- 
pressing and expanding this charge for one or two revolutions before 
firing. By this means the turbulence which, in the ordinary method 
of working, persists till the moment of firing was given time to die 
away. It was expected that a comparison of an expansion line 
