198 DR CARGILL G. KNOTT ON THE ELECTRICAL RESISTANCE OF NICKEL. 



palladium line,* continues parallel thereto till the temperature reaches 200° C, 

 after which it gradually bends up towards the palladium line. This it cuts 

 through at the neutral point (300° C), and almost immediately thereafter bends 

 round again into parallelism with the palladium line. Now these two rapid 

 bendings were found to occur just at the temperatures at which the peculiar 

 bendings occurred in the resistance curve. 



Similar experiments were tried on iron, with, however, doubtful results. 

 This was certainly in the main due to the non-efficiency of the method of 

 preparing and keeping a high temperature. 



The main results of these experiments may be thus described : — 



1. The rate of growth of the resistance of a given nickel wire with 

 temperature is greater, on the average, than the corresponding 

 quantity for platinum or palladium, and less than that for iron. 

 ' 2. The " logarithm rate "- — that is, the rate of change per unit rise of 

 temperature of unit resistance at any temperature — falls off more 

 slowly for nickel as the temperature rises to 200° C. than it does 

 for platinum or palladium. 



3. At about 200° C. the rate of resistance-growth for nickel increases 



markedly, and continues practically steady till about 320° C, when 

 a sudden decrease occurs, and thereafter the resistance steadily 

 increases at this diminished rate. In other words, between the 

 limits of temperature specified, the slope of the resistance curve 

 is much steeper than for any other temperature. The same pecu- 

 liarity is probably possessed by iron between the temperatures of 

 a dull red and a bright red heat. 



4. The peculiarity occurs (in each case) between the limits of tempera- 



ture within which the striking thermoelectric peculiarity discovered 

 by Tait also occurs — a peculiarity which is quite unknown in the 

 case of any other metal. 



5. There is thus a strong presumption that the Thomson effect in 



metals has a close connection with the mutual relations of resist- 

 ance and temperature ; at any rate in metals in which the 

 Thomson effect is proportional to the absolute temperature (accord- 

 ing to Tait's theory), the " logarithm rate " of change of resistance 

 seems to be very approximately inversely as the absolute tempera- 

 ture. In nickel and iron, in which the law of the Thomson effect 

 is peculiar, such a simple relation between resistance and tempera- 

 ture does not hold. 



* It is to be regretted that certain writers still persist in turning the diagram, as it were, upside 

 down, thus losing the advantage of Tait's improvement on Thomson's original form — an improvement 

 which fits in so admirably with the sign of the Thomson effect. 



