344 Mr. E. H. Hall on the "Rotational Coefficients " 



The mean of the numbers at low temperature is therefore 

 less than the mean for the higher temperature by rather less 

 than one per cent. The particularly small number 1703, 

 which causes this result, was obtained from a rather bad series 

 of observations, and is probably entitled to less weight than 

 the others. In fact, though the numbers as they stand seem 

 to indicate a decrease of about one per cent, in the value of 

 the transverse effect for a fall of about 30° C, I think it better 

 to say that we have here detected no certain effect of fall of 

 temperature. 



It is evident from these experiments that, if the value of the 

 transverse effect in gold varies at all with change of tempera- 

 ture, it varies far less than the electrical conductivity. We 

 must conclude therefore, that this effect depends rather upon 

 the magnitude of the current through the gold than upon the 

 fall of potential per unit of length. This conclusion was long 

 ago reached * ; but doubt had been cast upon its correctness by 

 the experiments upon nickel above mentioned. 



Turning again to the magnetic metals and taking a strip of 

 thin iron, experiments were made similar to those with gold 

 as above described. The results obtained, though by no 

 means so accurate as those obtained with gold, show that in 

 iron the rotational coefficient is very strongly affected by 

 change of temperature, the effect being an increase of perhaps 

 two thirds of one per cent, for a rise of 1° C. Possibly a 

 comparison of the effect of change of temperature upon the 

 magnetic permeability of iron, nickel, and cobalt with the 

 effect of the same change upon the rotational coefficient, will 

 be of value when both effects shall have been more fully 

 studied. 



Leaving the matter of effect of change of temperature and 

 referring again to the article on nickel and cobalt (Phil. Mag. 

 Sept. 1881), we see that the rotational coefficient in nickel 

 decreases as we increase the strength of the magnetic field ; 

 i. e. the rotational effect, other things being equal, increases 

 less rapidly than the intensity of the magnetic field. 



Experiments were made for the purpose of determining 

 whether a similar relation would hold in iron. The iron was 

 tested in magnetic fields varying in intensity from about 1000 

 to about 7500 in absolute C.G.8. units. Judging from the 

 behaviour of the strip of nickel previously examined in this 

 manner, the R. C. of that metal would be about 20 per cent, 

 greater in a field of intensity 1000 than in one of intensity 

 7500. For certain reasons I do not feel perfect confidence in 



* Silliman's American Journal, September 1880. 



