70 Profs. J. Dewar and J. A. Fleming. Changes produced in 



magnetism. On cooling it to 185 C., it slightly increased in mag- 

 netic moment, and on heating up again to 5 C., its magnetic moment 

 decreased to the extent of about 3 per cent, of its moment at 185 C. 

 There was no initial decrease of moment in this case. In this 

 respect, therefore, it resembled the chromium steel magrets. 

 Broadly speaking, the results so far obtained are : 



(1) That the sudden cooling to the temperature of liquid air 

 usually permanently decreases the magnetic moment of short mag- 

 nets made of many varieties of steel, assuming them to have been 

 initially magnetised in a strong field. 



(2) This initial decrease is found both in hardened steels having 

 great coercive force, and also in the same steels in a soft or annealed 

 condition, and is especially conspicuous in the case of the 19 percent, 

 nickel steel. 



(3) In the case of most steels so far examined, the effect of cooling 

 magnets made of them to 185 C. is to temporarily increase the 

 magnetic moment after the permanent magnetic condition has been 

 reached. 



(4) The exceptions to the above rule so far noted are the nickel 

 steels with percentages of nickel from 39 to 29 per cent., in which 

 case the magnetic moment is always decreased temporarily by cooling 

 to 185 C., after the permanent magnetic condition has been 

 reached. 



(5) It appears from these experiments that one of the best ways of 

 ageing a permanent magnet is to dip it several times into liquid air. 

 It then arrives at a constant condition in which subsequent tempera- 

 ture changes have a definite effect, and in which the subpermanent 

 magnetism is removed. 



Note added May 4. 



Since the 19 per cent, nickel steel magnet increases 'in magnetic 

 moment when heated from 185 C. to +oC., and since it is well 

 known that at some higher temperature it would lose magnetic 

 moment altogether, it was considered very desirable to ascertain the 

 temperature at which it would have its maximum magnetic moment. 

 The magnet was accordingly heated (on April 2) in an oil bath 

 gradually up to a temperature of about 300 C., and the deflections of 

 the magnetometer observed at intervals, both as the temperature rose 

 and as it fell. The result showed that this nickel steel magnet con- 

 tinued to increase in magnetic moment, until a temperature of about 

 30 C. was reached, and the magnetic moment then began to de- 

 crease. 



At a temperature of +300 C., the moment of the magnet was not 

 much greater than it was at 185 C. On cooling down again from 



