Magnetic Field on Metallic Resistance. 733 



on a nickel resistance. There is a slight preliminary increase, 

 after which the resistance decreases uniformly until 

 H = 1800, when the rate of decrease diminishes but does 

 not reach a maximum. The form of the curve suggests that 

 the increase is due to at least two distinct factors. One of 

 these gives a decrease of resistance directly proportional to 

 H until H = 1800, while the other gives an increase of 

 resistance and soon reaches a maximum. On the assumption 

 that such is the case the data for figure 5 have been calcu- 

 lated. 



Fig, 5, PL X. shows the increase of resistance in a trans- 

 verse magnetic field, which would result if the second factor, 

 which gives a decrease of resistance, were absent. The shape of 

 the curve is similar to that of figure 1 and the increase 

 reaches a maximum in both curves at the same field strength. 

 As the increase in resistance in figure 4 is approximately 

 one-half of that of figure 1, it is probable that the cause 

 of the increase of resistance in a longitudinal field is the same 

 as is operative in the case of the preliminary resistance 

 increase in a transverse field. This view is strengthened by 

 consideration of figures 6 and 7. 



Fig. 6, PL X. shows the transverse effect at 93 0, 5 C. 

 Fig. 1, PL X. shows the increase in resistance due to a trans- 

 verse field at temperature 93°* 5 0. Comparing figures 4 and 

 7 we notice that the maximum increase in a transverse field at 

 93°*5 C, is about one-half of what it is in a longitudinal field 

 at 93°*5 C, and that the maximum increase occurs at the 

 same field strength in both cases. 



Figs. 8 and 9, PL X. show the hysteresis curves for the 

 resistance effects in a transverse field at temperatures 18° C. 

 and 93°*5 C. respectively. The difference in the curves is 

 interesting. 



It will be advantageous at this point to describe some 

 experiments carried out by the author a short time ago on 

 the elongation effect of a longitudinal magnetic field on a 

 nickel rod. The method used — suggested by Dr. Swann — 

 was that of placing a pure nickel rod inside a solenoid and 

 observing the length alteration by means of interference 

 fringes. The solenoid was so constructed as to allow of a 

 flow of water between the wire and the rod, while the rod 

 was firmly fastened at one end. 



Fig. 10, PL X. shows the hysteresis curve so obtained. Un- 

 fortunately the magnitude of the field used was not large, 

 but Bidwell * has obtained results for large field values, and 



* Phil. Trans. 1888. 

 JPM. Mag. S. 6. Vol. 27. No. 160. April 1914. 3 C 



