Magnetic Fields on Resistance. 815 



or moved slightly, the galvanometer flew off the scale. 

 Repeated settings of the contacts were made until this 

 difficulty disappeared, when the resistance always had an 

 approximately constant value. Most of the crystals were 

 free from troublesome cracks and fissures ; in cutting out the 

 bars the best part of the crystal was always chosen. 



Results are plotted with dR/R (the change of resistance 

 divided by the resistance in zero field) as ordinates, and with 

 field strength, H, as abscissse. 



Tellurium, cadmium *, zinc, gold. — Reference to the curves 

 for these metals shows that the transverse effect is always 

 greater than the longitudinal (B is for the transverse, A for 

 the longitudinal field), but it is noticeable that the difference 

 between the two effects approaches zero as they become 

 smaller for different metals. It is probable that for lead or 

 platinum, in which the effect is very small, the direction of 

 the field would have no influence on the amount of change of 

 resistance. In copper it would have a very slight influence. 



For small fields all the curves fit the equation 



dR/R=AH 2 , 



but for large magnetizing forces a law of direct proportion- 

 ality seems to express results better (see Grunmach's 

 experiments f). The values of A for the different metals are 

 as follows : — 



Transverse field. Longitudinal field. 



Tellurium 105'0 x 10" 12 27'7 x 10" 12 



Bismuth % 12000-0 x 10" 12 3200*0 x 10~ 12 (?) 



Lead sulphide ... 4*86 x 10" 12 2'36 x 10" 12 



Cadmium 2*88 x 10" ]2 1-77 X 10~ 12 



Zinc l-36xl0- 12 0-82xl0- 12 



Gold 0-38xl0- 12 0-32xl0~ 13 



The Hall effect in tellurium is abnormally large (sixty 

 times greater than in bismuth), so that if there is any close 

 connexion between the Hall effect and the change of re- 

 sistance in a magnetic field one would expect the curve for 

 the latter phenomenon to exhibit abnormal properties. The 

 relatively large values of dR/R in tellurium, however, are 

 small compared with dR/R in bismuth. 



* See Phil. Mag. xxii. p. 900 (1911). 



t Ann. d. Phys. xxii. p. 141 (1907). 



t From the results of P. Lenard, Ann. d. Phys. xxxix. p. 669 (1890) 



3H2 



