216 Scientific Pmceedinci.i, Roijal Dnhlin Society. 



003 cm. diameter, with a load of about 9 x 10* grammes per sq. em., which 

 is much larger than the loads employed in our experiments. 



The method of our experiment was as follows : — A solenoid, 236 centi- 

 metres long, consisting of 7,707 turns of No. 18 double cotton covered copper 

 wire arranged in four layers, and capable of producing an internal magnetic 

 field of 41 c.g.s. units per ampere, was fixed vertically on a wall with the 

 lower end about 30 centimetres from the surface of a slate slab. The nickel 

 wire under test was suspended in the centre of the solenoid ; the lengtli of tlie 

 wire was 226 centimetres, so that it was in a uniform magnetic field through- 

 out its whole length. The wire was firmly fixed at the top end by means of 

 a three-jaw centre clutch, which projected five centimetres into the solenoid. 

 On the lower end of the wire there was fixed by the same means a cylindrical 

 vibrator composed of lead and brass, with a small iron pin or spike projecting 

 from its lower surface for the purpose of dipping into mercury. 



A concave mirror was fixed on the stem of the vibrator in such a way that 

 its reflecting surface was in the same plane as the centre of the wire. The 

 torsional oscillations of the wire were observed by means of a spot of light on 

 a semi-transparent millimetre scale placed at a distance of 167 centimetres 

 from the mirror. 



The maximum deflection of the light-spot which was used was at the 

 division marked 300 on the scale, which corresponded to a torsion or twist of 

 the lower end of the wire equal to an angle of about 5° 10' on each side of 

 the zero. The wire under test was set to oscillate round its own centre line — 

 that is, without any pendulum-motion — until the light-spot was a little above 

 the three-hundredth division ; and, when it was just on this 300 mark, one 

 started to count the oscillations and to read off the amplitude at every fifth 

 vibration up till the twentieth, and then every tenth vibration until seventy 

 vibrations in all had taken place. 



For the purpose of setting the wire to oscillate properly, two independent 

 direct-current circuits were employed, {1) a circuit including the wire under 

 test, secondary battery, ammeter, reversing key,, and a rheostat ; (2) a circuit 

 including the solenoid, secondary battery, reversing key, rheostat, and 

 hot-wire milliammeter, which latter measured both the direct current and the 

 alternating current round the solenoid. 



By sending a given steady direct current round the solenoid, and a 

 separate steady direct current through the wire, then, by means of the 

 reversing key reversing the current through the wire, in unison with tlie 

 oscillations of the wire or light-spot, the required amplitude of oscillation 

 could be easily obtained ; and when this amplitude was a little above the 300 

 mark on the scale, the direct currents were switched off both the wire and 



