508 3Ir. DugaU Glerh [Feb. 22, 



igniter served admirably for the Otto cycle engine, igniting in usual 

 cases about 80 to 100 times per minute. Where, however, ignition 

 has to be more frequent, as in engines constructed in accordance 

 with the Clerk cycle, the Otto flame was too slow. It required too 

 nmch time for operation. Accordingly, I produced a modified flame 

 igniter, which I show in section. This igniter depended upon 

 supplying a port with inflammable mixture froui tlie cylinder itself ; 

 the gases, being already mixed, contained within themselves oxygen 

 sufficient to support combustion, and accordingly the flame filling 

 the port in this way can be very rapidly ignited, and is quite equal 

 to at least 300 ignitions per minute. This slide was much used in 

 Clerk cycle engines for many years. 



These flame igniters gave way to the hot tube igniter about 

 twenty years ago, and now electric igniters have been rendered rela- 

 tively so perfect that they are rapidly displacing all hot tube devices, 

 even in stationary engines. Petrol engines, as you know, very soon 

 passed from tube ignition to electric ignition. 



The regularity and certainty of the various operations : charging, 

 compressing, igniting, expanding and exhausting, in an ordinary 

 stationary gas engine, is marvellous, when the difliculties which have 

 been faced and overcome are considered. The ordinary gas engine of 

 medium size rotates at about 1<S0 revolutions per minute, that is, three 

 revolutions per second. The forward stroke is thus completed in a 

 sixth of a second, that is, the intensely hot gases are produced in the 

 cylinder, the pressure attains its maximum, and expansion occurs for 

 the whole power operation in one-sixth of a second. Naturally, 

 ignition must take less time, and in these engines we find the 

 ignition of the mixture from the beginning of the explosion to the 

 attainment of maximum pressure takes from one-twentieth to one- 

 thirtieth of a second. It is surprising enough to be able to accom- 

 plish these operations with such certainty at these speeds, but the 

 experience with the petrol engine is still more surprising. A petrol 

 engine, running at the very common speed of 1200 revolutions per 

 minute, takes but one-twentieth of a second per revolution, that is, 

 one-fortieth of a second for the power stroke. In these engines, 

 maxinuun ])ressure is attained very {|uickly, the ex])]()sion, or pres- 

 sure rise period taking from one-eightieth to one-luindredth of a 

 second. It is interesting to consider that in an ordinary four cylin- 

 der car, with the engine running at 1200 revolutions per minute, we 

 get 2400 explosions per minute, the whole of the operations of a 

 complete revolution being performed in the infinitesimal time of 

 one-twentieth of a second. 



Tliese considerations show wliat an advance has been made in 

 liandling flame as the working fluid during the last thirty years. It 

 has taken more than a Inmdred years to attain our present knowledge 

 of the means of handling and controlling flame as the working fluid 

 in these engines ; but the last thirty years have seen great develop- 



