40 Scientific Proceedings, B.oi/al Dublin Society. 



was 1-5 X 10^ grammes per sq. cm., and that on the iron 10= grammes per 

 sq. em. These values were obtained when the wires were placed successively 

 in zero field and in a magnetic field of H = 20 c.g.s. units, in each case 

 starting from an initial amplitude of 300 divisions on the scale or from an 

 angle of 7-f'^'. 



When the results obtained with the iron wires, witli magnetic fields 

 lying between these limits of and 20 units, were plotted in curves, they 

 showed that the internal friction of the wire was first increased and then 

 decreased, as in the case of the nickel wires, but to a more limited extent. 



Gray and Wood,' in the papers already referred to, state that no cliange 

 of this nature in the internal friction of iron took place in their experiments ; 

 they did not, however, employ any magnetic field round their wire between 

 and 22 units, whereas the present experiments show that the phenomenon 

 occurs in low fields, that is, in a magnetic field of about 2'5 units with a No. 16 

 size wire, and about 3'5 units with a No. 14 wire. 



H. Tomlinson^ has also observed that the internal viscosity in an iron wire 

 is perceptibly greater when it is surrounded by a magnetic field than when it 

 is not ; he does not indicate, liowever, that any definite magnetic field was 

 found to give a maximum effect. 



Tlie points where the greatest internal friction in iron wires oeeiu's in the 

 present experiments, though much less pronounced than in the nickel wires, 

 are shown in the accompanying fig. 4, where as abscissse are plotted the 

 values of the longitudinal magnetic field round the wire, and as ordinatesthe 

 corresponding values of the amplitudes of oscillation when seventy vibrations 

 had been made, starting in every case from an initial amplitude of 300 divisions 

 on the scale (mm.) or from an angular twist on the lower end of the wire 



ofTF- 



The two lower curves are those obtained with the No. 16 iron wire ; the 



number at tlie end of each curve indicates the longitudinal load in grammes 



per square centimetre that was on the wire when being tested. 



In each of these two curves the lowest point is reached wlien the wire is 



in a longitudinal magnetic field of about 2'5 units, which is the field in which 



the maximum twist occurred in the tests for magnetism and torsion already 



mentioned. The top curve in fig. 4 is that obtained with the No. 14 iron wire 



when loaded at the rate of 2 x 10' grammes per sq. em. ; and its lowest point 



occurs in a magnetic field of about 3"5 c.g.s. units, which is also the magnetic 



field in which the greatest twist took place in the magnetism and torsion tests. 



' Loc. cit. 



-Phil. Trans. 1888, vol. clixix., A, p. 13. 



