Brown — Mechanical Stress and Magnetisation of Iron. 481 



centring clutch, and allowed to hang loosely under its own weight only. It 

 was then raised to a cherry-red heat by heating from the top downwards, so 

 that the hot air ascending tended to anneal the part just heated, tlie degree of 

 hardness thus obtained being here distinguished by the symbol Ho. The next 

 degree of hardness was obtained by the same process, but witli a weight on 

 the lower end of the wire amounting to 10" grammes per sq. cm. ; this 

 hardness we call Hi, and the other degrees were obtained by hanging on 

 weights of 2 X 10*, 2-5 x 10*, and 3 x 10* grammes per sq. em. respectively, 

 and their hardnesses are distinguished by the symbols Hj, H2.5 and H3. 



The first heating of the wire did not change its diameter, which was in 

 every case carefully measured when the oxide had been all cleared off by 

 means of emery paper; new wires of full No. 16 size were taken for each of 

 the other tempers, so as to get all the wires of different degrees of magnetic 

 softness, and very approximately of the same diameter. 



To get a measure of the hardness of the wire, two tests were applied, first 

 the simple rigidity of each was measured ; and, secondly, the electrical 

 conductivity of each was determined by comparison with a standard of 

 electrolytic copper wire whose conductivity was lOl'Ol, Matthiessen's 

 standard being 100. The simple rigidity was measured by a statical method 

 in which the horizontal forni of Searle's torsion apparatus was employed. 



This apparatus consists essentially of a spindle, mounted on ball- 

 bearings, with a self-centring three-jaw clutch at one end of the spindle, 

 and a wheel at the other end, by means of which the required couple can be 

 applied. One end of the wire under test is caught in the self-centring 

 clutch, and the other end fixed in a firm support at a suitable distance (in 

 this case 40'3 em.) from the clutch. 



In the apparatus used for these experiments, the end of the spindle at the 

 wheel was prolonged about 5 cm., and a plane mirror fixed on it with its 

 reflecting surface in the same plane as the axis of the spindle ; and the twist 

 of the wire under test was observed in the usual way by means of a vertical 

 scale and telescope with cross-hairs, the scale being 64'5 cm. from the mirror. 



In observing the deflection on the scale, or the twist of the end of the 

 wire, the maximum of which was about 6°, four or more different weights 

 were used, and a double reading taken in each case by hanging the weight 

 from either side of the rim of the wheel. The values of the scale-readings 

 were then plotted against the corresponding values of the weights, and a 

 straight line drawn from the origin through the points ; the mean values 

 of the weight and twist were picked off for calculating the rigidity. Tlie 

 scale used was divided into millimetres, and could be easily read to 0'2 mm., 

 so that an error in the scale-reading of 0'2 mm. made au error in the values 



3 y2 



