

Measurement of Gas-Engine Temperatures. So 



once reached them. For the complete study of gas-engine 

 temperature therefore some method of local thermometry is 

 necessary. 



The platinnm resistance-thermometer is the onlv means 

 available for the direct measurement of rapidly changing 

 gas-temperatures, since it alone can be constructed with the 

 requisite low thermal capacity. Moreover, it gives the local 

 temperature as distinct from the mean value obtained from a 

 study of the pressure. It has been employed by Callendar 

 and Nicholson * with great success for the measurement of 

 -team temperatures in the steam-engine cylinder ; and 

 Burstall has nsed it in the gas-engine, but experienced great 

 difficulties f. I have also used it in studying the explosion 

 of gaseous mixtures in a closed vessel ±. The difficulty of 

 the gas-engine experiments is that the temperature rises 

 locally much above the melting-point of platinnm, and the 

 wire fuses unless it is so thick that it fails to reach the 

 maximum temperature of the gas surrounding it. But if the 

 latter condition be fulfilled, it is obvious that the thermometer 

 does not give the temperature of the gas at other points of 

 the cycle, unless corrections are applied for the time-lag. 

 Professor Burstall found it necessary to use wires j -,,-„„ of an 

 inch diameter, in order to secure sufficient permanence in the 

 thermometer to take a complete series of measurements of its 

 resistance throughout the cycle by means of a rotating contact- 

 maker. In my own closed- vessel experiments I was able to use 

 the finest obtainable wire (foW m ' caame t er )? au d it nearly 

 always melted if placed near the point of ignition. But as a 

 continuous record of the resistance was taken in each explosion 

 prior to melting, this did not affect the measurements. Even 

 with such fine wire the correction for time-lag was by no 

 means neolio-ible; and I was able to show that a wire double 

 the diameter might be as much as two hundred degrees hotter 

 or cooler than the gas surrounding it when the temperature 

 changes so rapidly as in the working stroke of a gas-engine. 

 The determination of the time-lag of the wire, or to speak 

 more precisely of the relation between the temperature of the 

 wire and that of the gas, is therefore a matter of much im- 

 portance in these measurements. The difference of the two 

 temperatures is due to the constant exchange of heat, which 

 must go on between the wire and the gas to provide for the 

 warming or cooling of the wire and for the loss of heat from 

 the wire by radiation. If the temperature of the wire is known 



* Min. Proc. Inst. C. E.vol. cxxxi. 

 r Phil. Mag-, vol. xl. p. 282 (1895). 

 \ Proc. Roy. Soc. A. vol. Ixxvii. p. 387 '. 



