432 Effects of Temperature on Magnetized Iron Wire. 



Verification and Discussion of Results. 



It is interesting to notice in fig. 3 the limiting effect of the 

 circular magnetization in high fields. The curves up to 

 1 2 amps, in the wire almost coincide at a. longitudinal field 

 of 150 C.G.S. 



To test how far the results in fig. 3 were reliable, the wire 

 was mounted in a glass tube and a water circulation allowed 

 to flow through the tube in contact with the wire. Thus the 

 temperature of the wire was kept practically constant for all 

 values of the current. The results of these tests agreed very 

 closely indeed with those shown in fig. 3, even down to a field 

 of 10 C.G.S. 



Klemencic (Wied Ann. vol. lvi. p. 574) investigated the 

 circular magnetization of iron wires together with the axial 

 magnetization by a different method. By including the wire 

 as an arm of a Wheatstone's bridge and using a ballistic 

 galvanometer, the circular magnetization was deduced from 

 the observed value of the self-induction for different currents 

 in the wire. Here the change of temperature of wire intro- 

 duces difficulties. It seems that the magnetometer method, 

 when the wire is kept at a constant temperature, is much 

 simpler and less troublesome. 



The results shown in fig. 4 were found to agree closely with 

 the tests obtained for soft iron by the platinum-tube method. 

 The point of demagnetization was obtained a trifle lower by 

 the latter method, and the initial slope of the curve at a field 

 of 290 c.G s. was a little less. It will be noted that the point 

 of demagnetization is not absolutely sharp ; the curves suddenly 

 change their direction and I decreases more slowly. This 

 was also investigated in special tests made by the platinum- 

 tube method. The value of I at 750° C. was observed to be 

 about 7 C.G.S. 



These results are found to agree very well with those for 

 soft iron in high fields obtained by Curie (Comptes Rendus, 

 vol. cxviii. p. 859). He heated his specimen in a platinum 

 heating-coil and measured the temperature inside the coil by 

 a thermo-couple. The point of demagnetization which he 

 obtained is rather higher than that obtained by the platinum- 

 tube or current-in-the-wire methods, but not so high as that 

 given by Hopkinson (loc. cit.) who used a copper wire for 

 his temperature measurements. More recently a paper has 

 appeared by Morris (Phil. Mag. vol. xliv. Sept. 1897), who 

 employed the same method of heating as Curie, but measured 

 the temperature with a platinum wire. It will be seen that 

 the method of heating with a coil is less perfect than with 



