286 iMAGNETISM 



A precaution is necessary in this method. Foueault currents 

 are induced in the bismuth at both make and break in the 

 primary. If each Foucault current has time to he formed and 

 die away before the next change in the primary, it will induce 

 equal and opposite currents in the secondary and may be left 

 out of account. But if the interval between make and break 

 is too short, the Foucault current formed in one interval extends 

 partly over the next interval, where its effect on the secondary 

 is commuted, and so a balance is left over to affect the gaKano- 

 meter. Experiments showed that the frequency must not exceed 

 more than 8 or 9 per second. 



In a second method a horizontal bismuth cylinder was suspended 

 from one end of a torsion rod so as to be partlv within, partly 

 without a horizontal solenoid through which a known cunvnt 

 passed and the force on the bismuth was meaMired. That force 



is, as we have seen, I g - II could be calculated for each 



element of the cylinder and hence K was determinate. 



A third method resembled that of Rowland and Jacques, and 

 need not be described. 



In a fourth method a bismuth cylinder was brought bet 

 the poles of a strong magnet. At a distance from tin poles cither 

 9 or 17cm. was a magnetometer nce<U< :<>re the bismuth 



was inserted the effect of the magnet on the needle was compen- 

 sated by a second magnet. When the bismuth was inserted it 

 behaved as a feeble reversed magnet and deflected tin i 

 meter needle through 



where M is the magnetic moment of the bismuth and II is Oil- 

 field at the magnetometer. The deflection was too minute for direct 

 measurement. The bismuth was inserted and withdrawn with a 

 periodicity the same as that of the needle, producing ultim itelv a 

 swing g. If K is the dampii.. .cut of the suin^s the d. 



deflection A is then given oy 



The values obtained by the successive methods for 10*/c were : 

 I, 13-57; II, 14-11 ; III, 15%S; IV, l:i-<>. Ti,e noond method is 

 probably the most exact. 



Curie's experiments. P. Curie * determined the value of 

 K for a number of substances and investigated the effect of change 

 of temperature, u>ing a method like that of Bt|uerel and 

 Faraday. The body to be tested was on a torsion arm, and the 

 torsion was measured which was needed to keep it at o in the field 

 * inn. de Chim. et n hy. , 7 ser. V. (1895), p, 289 ; or (Euvre* df P. Curie, p. 232- 



