232 (). U. VONWILLEK. 



round it. After some time a steady temperature gradient 

 was established along the strip and the e.m.f., between 

 different pairs of copper molybdenite junctions could be 

 measured, readings at the same time being taken of the 

 e.m.f. obtained witli each copper-eureka junction, the other 

 junction in this case being at the room temperature outside 

 the enclosure in which the molybdenite was placed. After 

 a calibration was made later the latter readings gave the 

 temperatures of the four junctions. The copper wires of 

 the copper-eureka junctions were used to form the copper 

 molybdenite junctions. These tests were carried out with 

 a number of different heating currents so that the junctions 

 attained different temperatures. The calibration was 

 carried out by raising the temperature of the whole pre- 

 paration in the electric oven, the heating current at the 

 end of the strip being zero, and measuring the thermo-e.m.f. 

 for each copper-eureka junction, the temperature of the 

 whole being observed with the aid of a mercury in glass 

 thermometer. Such readings were taken at several tem- 

 peratures, and it was found that the four junctions gave 

 practically the same results. Curves were plotted giving 

 the potentiometer readings for various temperature differ- 

 ences and so the temperatures of the copper molybdenite 

 junction were readily determined with their aid. In all 

 these experiments the room temperature only varied over 

 a range of less than two degrees. 



On plotting the thermo-e.m.f. for each pair of copper 

 molybdenite junctions against differences of temperature 

 the curve obtained for any pair is practically a straight 

 line. It was found that the slope varied slightly with 

 different pairs of points on the same strip and with differ- 

 ent strips. The maximum difference of temperature in 



