272, 



Drs. J. T. Bottomley and J. C. Beattie. 



of the platinum wires was measured by means of a high resistance 

 reflecting galvanometer. 



This potential galvanometer had a resistance of about 5000 ohms, 

 and it was possible to insert in the galvanometer circuit an additional 

 resistance of 10,000 ohms. 



In order to ascertain the absolute value of the readings of the 

 potential galvanometer, a standard coil of platinoid wire, whose resist- 

 ance was known very accurately, was joined into the circuit, as shown 

 in fig. 2. This resistance was of considerable length, and it was 

 kept cool by being immersed in a bath of oil. 



The following was the order of experimenting. The pressure in the 

 tubes was first reduced as much as possible by means of the Sprengel 

 pump ; then a very small current, practically unable to heat any part of 

 the circuit, was sent through the two platinum wires and the standard 

 coil, and the potential difference between the two ends of each was 

 determined. This gave the ratio of the resistance of each of the 

 platinum wires to that of the standard coil, all being cold, and at the 

 same temperature. The current from the battery of storage cells was now 

 suitably increased, and readings were taken in the following order : — The 

 current passing was first read. Then the zero of the potential galvano- 

 meter was noted, and the deflection of the potential galvanometer 

 when connected to the two ends of the standard coil was observed. The 

 electrodes of the potential galvanometer were next applied to one of 

 the platinum wires, and the deflection noted ; then the deflection due 

 to the second wire was observed. A second reading was taken from 

 the first wire and also from the second wire. Usually these pairs of 

 readings were identical, or nearly so, as no reading was taken until 

 after the strong current had been passing through the circuit for 

 sufficient length of time to allow the temperature of the whole to 

 become perfectly steady. Generally speaking, five minutes or more 

 was allowed for this purpose. Lastly the current was again read, and 

 the zero of the potential galvanometer noted. 



The readings detailed above enabled us to calculate the current 

 passing through each wire and the resistance in that wire. The length 

 and cross section of each of the platinum wires (practically identical) 

 were also known. Thus the energy lost by radiation per square 

 centimetre per second, C^R/JS, could be calculated; C being the 

 current, E. the resistance and S the surface of the radiating wire, and 

 J being the dynamical equivalent of heat, all in absolute measure. 



The measurement of the electric resistance of the wires also enabled 

 us to calculate the temperatures of the wires by means of the results 

 of a separate determination of the electric resistances, at different 

 temperatures, of th^i wires themselves. ^ 



In a former paper,"*^ the precautions and difficulties connected with 

 * J. T. Bottomley, 'Phil. Trans.; A, 1887. 



