The Measurement of Magnetic Hysteresis. 



351 



In all these experiments, the curves showing W in terms of the 

 stress, bear a close resemblance to those showing B in terms of the 

 stress. To examine this point, curves were plotted showing how W 

 varies with B , when H is kept constant and B is varied by varying 

 the stress. 



For both tension and torsion each curve for a given value of H 

 takes the form of a straight line having a hook at one end. The 

 straight portions of the separate curves for different values of H all 

 pass, on prolongation, through a single point, generally on the line 

 B = 0. Thus the straight parts are represented by W = mB - b. 

 Plotting m against H it is found that m — aH h , and thus the formula 

 becomes W = «Ho ;i B - b, where a and b are constants. It is found 

 that this formula represents W closely when both H and B vary over 

 a considerable range in the neighbourhood of the maximum permea- 

 bility, the iron being now free from stress. 



§ 5. An electric current flowing along an iron wire magnetises it 

 circularly, and may be expected to diminish both B and W for the 

 given limits ± H . Experiment showed that the expected effect occurs, 

 a current of 1*123 ampere through an iron wire about 1 mm. in 

 diameter diminishing W by 22*7 per cent. 



§ 6. The numerical values of the quantity Q, which occurs, in § 2, 

 in the expression for the heat produced by the eddy currents in the 

 specimen, are calculated in Appendix I for rods of both circular and 

 rectangular sections. 



§ 7. In their experiments the authors have used straight iron wires 

 about 50 cm. in length. They discuss the effect of the de-magnetising 

 force due to the induced magnetism of the specimen, and show how 

 to apply corrections to the value of W calculated from the formula 

 1/47T . jH'c/B', where H' is the magnetic force due to the current, 

 and B' is the magnetic induction at the centre of the wire; they 

 also give numerical examples of these corrections. Appendix II 

 contains an account of experiments made to find the de-magnetising 

 force h under two sets of conditions. In the first case, h was determined 

 when H = H , after many magnetic cycles with the limits ± H . Using 

 a freshly annealed wire, and increasing H from to 124 G.G.S., h was 

 found to rise to a maximum, which occurred nearly when /x had its 

 maximum value ; the maximum was followed by a minimum of h, and 

 the value of h for the largest values of Ho was less than that which would 

 obtain if the induction through the centre of the wire flowed in and 

 out only by the ends of the wire. This small value of h implies the 

 existence, between the centre and either end of the wire, of a "pole " 

 of sign opposite to that of the pole at the end, a circumstance only 

 to be accounted for by the effects of hysteresis. In the second case 

 h was found for several points on the cyclic B-H curve, and curves 

 are given showing h in relation to both H and B. In both curves h 



2 B 2 



