in Nickel and Cobalt. 163 



Mr. Hopkinson has suggested " rotational coefficient of resist- 

 ance ;" and possibly some quantity might be found which 



W 

 would demand that title. At first sight ^, which is an elec- 

 tromotive force divided by a quantity proportional to a cur- 

 rent, would seem to be of the nature of a resistance ; but it is 

 to be noticed that the electromotive force E' is not the cause, 

 but the effect, of the current implied in V. 



In the experiments which I have described in previous 

 papers, no account was taken of the temperature of the con- 

 ductor experimented upon. When these experiments upon 

 nickel, however, had been going on for a long time, it began 

 to be suspected that the temperature of the room, and so of 

 the nickel plate, did exercise a very considerable influence 

 upon the magnitude of the transverse effect as expressed by the 



ratio ^r- 



A few hasty experiments with considerable ranges of tem- 

 perature in the room indicated very decidedly that the tempe- 

 rature was a factor to be considered, and that the higher the 

 temperature the greater the value of W, other things being 

 equal. The magnitude of this influence can hardly be deter- 

 mined from results thus far reached. It may prove that the 

 transverse electromotive force E / is no more increased by a 

 rise of temperature than the direct electromotive force E* is ; 

 and in this case it would appear, as intimated above, that the 



ratio -^r is the one to be investigated rather than ~ • 



Future investigation must determine this matter; and mean- 

 while it has been sought to avoid evil consequences by regu- 

 lating, as well as practicable, the temperature of the nickel 

 plate. Sometimes an experiment had to be made at a rather 

 high temperature for instance; and an attempt would then be 

 made to balance this by making another with about the same 

 strength of magnetic field but at a low temperature, or vice 

 versa. There was, however, even now no attempt to deter- 

 mine the actual temperature of the nickel; but a thermometer 

 was hung up with its bulb close to the plate, and as nearly as 

 practicable always in the same position with respect to the 

 latter, and both plate and thermometer were protected from 

 sudden changes of temperature. As the nickel was of course 

 heated by the current, its temperature must have been always 

 considerably higher than that indicated by the thermometer. 



* Apparently E' is in nickel affected by temperature more than E in 

 most metals would be ; but the rate of increase of the resistance of nickel 

 with rise of temperature seems not to be known. 



