of Cyclically Varying Temperature. 



291 



followed by about 20 blank strokes. The effect of the 

 thorough " washing-out " of the products of combustion from 

 the cylinder and of the hot cylinder jacket was to produce a 

 temperature in the cylinder as high as, if not higher than, if the 

 engine had been working at full load. The small number of 

 working strokes per minute enabled observations to be made 

 with the galvanometer which would not have been possible at 

 much closer intervals of time. The wear and tear on the 

 measuring-wire was reduced, it being a matter of consider- 

 able importance to expose the wire to as few explosions as 

 possible. 



In all the following experiments the thermometer measuring- 

 wire was placed on the axis of the cylinder and in the centre 

 of the compression space. 



Two thermometers were used, both practically of identically 

 the same construction as described above. The results of a 

 number of experiments are given below ; I have selected 

 those which I have reason to believe are the most complete 

 and accurate of the number which I have made. Wherever 

 I have been able, I have given the constants of the thermo- 

 meters which were determined before and after the experi- 

 ments ; where it is not given it is because the wire fused 

 before a re -calibration could be made. 



January 1st, 1895. 

 Thermometer No. 5. Wire 0003 inch diam. 



Air temperature 9° 0. 

 E before, 0'324 ohm. 



Circulating water, 46 c 

 Ei-Ko =0-1206. 3=1-52. 



C. 



E after, 



0-340 ohm. 







First Set. 



Percentage of 



Observed Eesist- 



pt. 



Temperature on 



stroke. 



ance, ohms. 



air-scale. 



10 



1-476 



948 



1120 



20 



1-443 



921 



1085 



30 



1-433 



913 



1070 



40 



1-362 



854 



990 



50 



1-351 



845 



975 



60 



1-299 



802 



925 



70 



1-286 



791 



900 



80 



1-238 



751 



850 



Second Set. 



10 



1-547 



1007 



1210 



20 



1-477 



949 



1120 



30 



1-466 



940 



1110 



40 



1-433 



921 



1085 



50 



1-382 



871 



1010 



