402 



Scientific Proceedings^ Royal Dublin Society. 



The percentage chemical composition of four of the specimens, omitting 

 the iron, is as follows : — 



These four wires were tested, for the damping of torsional oscillations 

 in the physical state in which they were received from the manufacturer ; 

 they were all of the same length and diameter and the longitudinal load on 

 each was 1*5 x 10* grammes per sq. centimetre. This was the load found 

 most suitable and was used in the comparison of these four wires. 



They were too hard magnetically to obtain witli them a " Wiedemann 

 curve," or curve relating twist and longitudinal magnetic field, from which 

 can be found the magnetic field corresponding to the peak or turning point 

 on the curve, and therefore tlie magnetic field which gives the greatest 

 damping to torsional oscillations. ' In order, therefore, to obtain this particular 

 magnetic fieli], damping curves were taken for eight or ten different values 

 of longitudinal magnetic fields, also for the same number of corresponding 

 alternating magnetic fields, and the results plotted as follows'' : — the values 

 of the magnetic fields employed for both the sets of experiments D.C. 

 and A.C. were taken as abscissm, and for ovdinates the corresponding values 

 of the amplitude of oscillation after 70 complete vibrations had taken 

 place, starting in each case from the division 300 on the scale. The 

 magnetic field corresponding to the lowest points on these two curves was 

 taken and employed in the final experiments on the damping of torsional 

 oscillations in the wires. For the three wires containing chromium this 

 magnetic field both direct and alternating was about 11 c.g.s. units, and for 

 the wire containing nickel, the field was about 16 c.g.s. units. The principal 

 results are shown in the following table, whioli gives only the amplitude of 

 oscillation — read by means of the light-spot on the scale — after 70 complete 

 vibrations had taken place starting from the division marked 300 on the 



' Scient. Proc. Eoy. Dub. Soc, vol. xii, 1910, jS"o. 36, p. 484. 

 * Scient. Proc. Roy. Dub. Soc, vol. xiii, 1911, No. 3, p. 37. 



