542 



PROF. C. G. KNOTT ON 



wire cooled more quickly, and again a gradually changing deflection was obtained. In 

 like manner when the current was so applied as to magnetise the one wire, the sudden 

 deflection due to the effect of the magnetisation on the resistance was followed by a 

 gradually growing deflection due to the different rates of heating in the coils. The 

 reading of the deflection could not be taken at once, but a few seconds had to elapse 

 before the spot of light ceased oscillating to and fro about its mean position. The 

 method of observation was to take the mean of the last distinct oscillation after the 

 magnetisation had been produced in the one wire, then break the current, take the 

 mean of the last distinct oscillation, and again make the current, and once more take 

 the mean of the last distinct oscillation, and so on until the operation was complete. 



RESISTANCE OF NICKEL IN VARIOUS FIELDS 

 AT DIFFERENT TEMPERATURES 



8 





















/ 













o 







<4 





6 























X. 

 



o 

 o 

 o 





M 







N / 











4 



K 





















a 













































«j 





















x 



T) 



/ / 





















<U / 





















i 





2 





4 







6 



o 













c 



> Field 



X 







4 







6 









Now the rates of heating when the current was passing were not related in the same way 

 as the rates of cooling when the current ceased to act. A little consideration will show, 

 what was clearly demonstrated by the experiment, that the effect will be to increase the 

 real deflection in the case of the one wire, and to decrease it in the case of the other. 

 In other words, the measured changes of resistance of the wire M in the higher fields 

 are less than the real values because of an uncompensated temperature effect, in virtue 

 of which the measured changes of resistance of the wire N are greater than the real 



