4 Scientific Proceedings, Royal Dublin Society. 



In the case of iron, however, the value of the maximum fatigue is the 

 same for all the frequencies; whereas, for nickel, the value of the maxi- 

 mum fatigue increases with the frequency, up to a certain value of the 

 frequency. 1 



The wire was tested for subsidence of torsional oscillations in this hard 

 state : it was then taken down and heated, when hanging freely in a vertical 

 position, from the top downwards by means of a broad Bunsen flame. When 

 cold, it was cleaned up, the rigidity measured, and again put into the solenoid 

 and tested for torsional subsidence. The wire was then taken down and the 

 same process of heating gone through, the rigidity measured, and the torsional 

 subsidence again observed, and so on, so that the wire was tested when in five 

 different states of rigidity, as indicated below. 



The following five tables (II-VI) give only a few of the values observed, 

 which are perhaps sufficient to show the general trend that the curves would 

 take, with the wire in the various states of hardness, if the values were plotted 

 with the number of vibrations as abscissae, and the corresponding amplitudes 

 of oscillation as ordinates. 



In the tables, the mark D. C. means that the wire was subjected to the 

 influence of a direct longitudinal magnetic field, and A. C. that it was under 

 the influence of an alternating magnetic field at the different frequencies. 



Table II. 

 Eigidity == 815 x 10 6 grammes per sq. cm. 



1 Scient, Proc. Roy. Dub. Soc, 1915, vol. xiv, No. 39, p. 525. 



