292 Mr. Dugald Clerk [Jan. 21L 



the second ; while at CO revolutions per minute the difference was- 

 only 20 per cent. 



In these comparative experiments the temperature and density 

 of the gas were the same, and the difference could only be due to 

 the motion. From measurements of heat loss from a similar wire 

 in a closed vessel, Hopkinson infers that the motion of the gas with 

 a fan speed of 2500 revolutions per minute is probably considerably 

 greater than that obtaining in the gas engine. xA.t this speed the 

 heat flow at a temperature of IGOO^ C. or over is increased by an 

 amount of the order of 25 per cent. 



These experiments of Clerk and Hopkinson clearly prove the 

 importance of what may be called Suction Turbulence in determining 

 the time of explosion and also the rate of cooling in all internal 

 combustion engines. 



Determination of Temperatures. — So far the question of the 

 rising hne and the phenomena of ignition only have been dealt with. 

 Before proceeding further it is necessary to say something on the 

 question of temperature determinations. Upon an accurate know- 

 ledge of temperature and temperature distribution in closed vessel 

 experiments and in gas engine cylinders depend all conclusions as to 

 properties of the gases, heat flow to the walls, and efficiencies possible 

 under the varying conditions. Accordingly, much consideration has 

 been given by the Members of the Committee to all questions re- 

 lating to the determination of temperatures under varying conditions. 

 In closed vessel experiments the initial pressure and temperature of 

 the gas mixture is known with great accuracy. Accordingly, if it be 

 assumed that no volume change occurs due to chemical action, the 

 temperature determination is easy, because absolute pressures are 

 then proportional to absolute temperatures. It is only necessary 

 then to determine the ratio between absolute pressure before explosion 

 and absolute pressure at explosion to get the mean temperature of 

 the mass of gas at the highest pressure point. Unfortunately, 

 chemical change introduces complications, and in ordinary coal gas 

 and air experiments which have been dealt with here the result of 

 combustion is to produce a contraction of volume. In ordinary gas- 

 engine mixtures the contraction is not great — it amounts to about 

 4 per cent — but it must be allowed for, and it is difficult to determine 

 with accuracy, because so far no absolute knowledge exists as to the 

 precise stage which the combustion has reached at the maximum 

 temperature of an ordinary gaseous explosion. Allowing for all 

 uncertainties, however, it may usually be taken in closed vessel 

 experiments that combustion has certainly consumed about 80 per 

 cent of the gases present at the point of maximum pressure. Applying 

 this proportion to the total contraction possible, a close approximation 

 to the maximum temperature can be obtained. In this way the 

 combining gases are themselves used to form a gas thermometer. In 

 closed vessel experiments many conclusions can be deduced from 



