Brown and Smith — Subsidence of Torsional Oscillations. 217 



solenoid, and an equal alternating current immediately switched on to the 

 solenoid only. The amplitudes of the oscillations were then read off on the 

 scale until seventy vibrations had been observed. 



It has been shown' that, in nickel wires, the greater the longitudinal load 

 on the wire — up to a certain limit — the greater the twist of tlie free end, and 

 that the magnetic field round the wire which gives this maximum twisi 

 increases witli the load, the twist being due to the combined circular and 

 longitudinal magnetization. It was also shown that the greatest damping of 

 the torsional oscillations of the wire takes place under these conditions of load 

 and longitudinal magnetic field. It seemed likely that the same conditions 

 would hold with alternating magnetic fields ; and at the beginning of the 

 present investigation we proved experimentally that they do hold. Therefore 

 the magnetic fields employed in the present research, whether continuous or 

 alternating, were those corresponding to the given loads on the wire when the 

 maximum twist was obtained. Thus, when the load on the wire was 0'5 x 10^ 

 grammes per sq. cm., the magnetic field round the wire was 13 c. g. s. units, 

 and 17 e.g. s. units when the load was 10* grammes per sq. cm,, and 20 c. g. s. 

 units when the load was 1'5 x 10* grammes per sq. cm. 



The nickel wire first tested was a No. 16 s. w. g. of diameter 0'1675 cm. in 

 the physical state in which it was received from the manufacturer — that is, it 

 was hard or of simple rigidity about 770 x 10' gi-ammes per sq. em. There 

 were few tests made with the light load 0'5xl0* grammes per sq. cm., 

 because the torsional oscillations were so rapid that reliable readings of the 

 amplitudes of the light-spot on the scale could not be made. 



The load of 10* grammes per sq. cm. was therefore put on the wire, and 

 observations made on the damping of the torsional oscillations, when magnetic 

 fields of 17 c. g. s. units were round the wire — that is, a damping curve was 

 first observed when this magnetic field was produced by a continuous, steady 

 direct ourreut round the solenoid, and observations again made when the 

 magnetic field was produced by means of alternating currents of the same 

 R. M. S. value as the direct current, and of frequencies 20, 30, 50, 100, and 

 140 per second.- The results are shown in Table I. 



' Sclent. Proc. Roy. Dub. Soc, vol. xiii (N.S.), No. 3, pp. 31-37. 



'' Graphs of the alternating currents at the varioua frequencies were taken by means of the 

 oscillograph. Those for frequencies 20 to 50 were very nearly sine curves ; and the currents were 

 obtained from a single-phase alternator direct-coupled to a direct current shunt motor. The 

 currents at frequencies 100 and 140 were obtained from a small belt-driven inductor alternator ; and 

 their graphs, on analysis, showed the presence of the third harmonic of amplitude 2' 3 as compared 

 with 10'5, the amplitude of the fundamental on the same scale, the harmonic starting from zero in 

 nearly phase-opposition to the fundamental. 



2k2 



