68 Profk J. Dewar and J. A. Fleming. Changes produced in 

 too- 

 5o- 



FIG. 16. Silicon steel. 



Si = 2-67 

 C =0-20 

 Mn = 0-25 

 Fe = 96 -88. 



into a U shape, with ends about 10 mm. apart, and in this case the 

 changes of magnetic moment, as shown in fig. 17, were similar to 



2oo-- 



loo- 



o- 



FIG. 17. Soft iron. 



those in the case of the carbon steels, only with very much narrower 

 limits of variation. The first cooling hardly produced any change 

 whatever in the magnetic moment of the magnet. On allowing it to 

 heat up again, the magnetic moment was very slightly diminished, 

 and thenceforth the changes of moment were such that the magnetic 

 moment was greater when the magnet was cold than when it was 

 warm, by about 2 or 3 per cent, of the latter value. In the case of a 

 straight, soft iron magnet, formed of annealed transformer iron, the 

 curious fact was noticed that whereas a rapid cooling of the magnet 

 by plunging into liquid air hardly produced any effect on the magnetic 

 moment after the first initial loss of magnetism had taken place on 

 cooling, the effect of a slow cooling down to the temperature of 

 185 C. was always to produce a permanent diminution of magnetic 

 moment. Hence the magnetism of this soft iron sample could be 

 frittered away by a process of slow cooling to 185 C., and inter- 

 mediate heating up to 5C. These changes of moment are repre- 

 sented in the diagram of fig. 18. 



Hard Iron. A sample of the same iron, hardened by hammering, 

 was tested, and was found to behave in a very similar manner to the 



