42 



Scientific Proceedings, Royal Dublin Society, 



For this purpose a nickel wire 226 ems. long, 0'123 cm. diameter, and of 

 simple rigidity about 708 x 10^ grammes per square centimetre was prepared 

 for experiment in the manner already stated. 



The wire was suspended vertically in the centre of the long solenoid, with 

 a load on its lower end equivalent to 3 x 10' grammes per square centimetre ; 

 on the rim of the brass vibrator there was fixed a finely divided paper scale, 

 by means of which, with the aid of a reading-telescope with a fine vertical 

 hair in the eye-piece, the amplitude of oscillation could easily be read off to 

 one-fifth of a scale-division. 



The scale was divided so tliat 90° was equivalent to 127 scale-divisions ; 

 and since 0'2 of a division could be read with ease, the amplitude recorded 

 may be taken as correct to within 8*5 minutes of arc ; and since the time of a 

 complete vibration with no magnetic field round the wire was 5.23 sees., the 

 period was slow enough for the observer to note the turning-point with 

 accuracy. 



Several sets of observations were made on the subsidence of torsional 

 oscillations, starting from initial amplitudes of 90°, 70°, 50°, and 30° 

 respectively, when the wire in each case was subjected to different longitudinal 

 magnetic fields. 



Table V. 



