38 Scientific Proceedingfi, Royal Dublin Society, 



hard by being allowed to cool under the tension due to the load. The wire 

 was then placed in the solenoid, with a given longitudinal load on its lower 

 end, and the rate of subsidence of the torsional oscillations observed when it 

 was placed successively in longitudinal magnetic fields up to a maximum of 

 80 c.g.s. units. 



Three different longitudinal loads were used in this set of experiments,- 

 and, when tlie observations were plotted in curves, results similar to those 

 given in fig. 2 were obtained — wlieu the same load was on the wire — but 

 with somewhat sharper slopes, as would be expected fi'om a liarder wire. 



Curves were also obtained similar to those given in fig. 3, but with 

 slightly lower values throughout, and with the minimum points occurring in 

 the same magnetic fields as those obtained with the softer wire. 



This latter circumstance also was to be expected, as it has been shown 

 that when a nickel wire is subjected to the simultaneous actions of circular 

 and longitudinal magnetism, tlie longitudinal magnetic field in which the 

 maximum twist of its free end takes place is independent of the hardness of 

 the wire.i 



Section 2. — Iron Wire. 



It is known (1)^ that, when an iron wire is tested for magnetisation and 

 torsion, tlie longitudinal magnetic field in which the maximum twist occurs 

 is independent of the longitudinal load on the end of the wire between 

 certain limits; (2)^ that tlie magnetic field in which the maximum twist takes 

 place is greater the larger the diameter of the wire. 



Experiments on the subsidence of torsional oscillations were therefore 

 carried out with iron wires similar to those already done with the nickel 

 wires. Tiiis was in order to find out if, like the nickel, the iron wires 

 attained their greatest internal friction when they were placed in that 

 longitudinal magnetic field which gave the largest twist when they were 

 tested for magnetism and torsion. For this purpose two wires were tested — 

 namely, one No. 16 S.W. G. when it was subjected to two different longi- 

 tudinal loads, and one No. 14 S.W. G. when subjected to one load. 



Both wires were used in the physical state in which they were received 

 from the manufacturer, the simple rigidity of each being about 815 x 10° 

 grammes per square centimetre, 226 cms. long, the cross-sectional areas 

 being 20*2 x 10"^ sq. cm., and 33'66 x 10"^ sq. cm. respectively. The 



' Soient. Proc. Roy. Dub. Soc, vol. xii, JSTo. 37, p. 505. 

 - Scient. Free. Eoj'. Dub. Soc, vol. xii, No. 36, p. 484. 

 ^ Loc. cit., p. 495. 



