248 A. Durward — Temperature Coefficients of 



of 4 per cent or 5 per cent, but this loss seemed to disappear 

 entirely after the magnet had been allowed to stand for about 

 24 hours. The numbers in the second column of Table I 

 show the difference between the moments of a certain cylin- 

 drical magnet, 05 cm in diameter and 15 c,n long, at 17° and at 

 the temperature indicated by the corresponding numbers in 

 the first column. These differences are expressed in terms of 

 the original moment at the room temperature of 17°. 





T 



AJBLE I. 









Loss of moment in terms of the 



Temperature. 







original moment at 17°. 



17° 







o-oooo 



48 







0-0043 



64 







0-009 1 



81 







00131 



100 







0-0180 



81 







0-0134 



60 







00094 



19-5 







0-0012 



Effects of a similar nature might have been looked for if obser- 

 vations had been taken without allowing ample time for the mag- 

 net to attain a uniform temperature at each stage, but it is 

 certain that no sensible error from this source enters into the 

 results here given. 



Many seasoned magnets show no measurable time lag in 

 attaining their original moments after they have been heated 

 to 100° C, and none show any such lag if they have been 

 heated to only about 20° above the room temperature and 

 then cooled, or if they have been heated to 100° shortly before 

 the experiment has begun. In other words, a second heating 

 immediately after the first does not add to the temporary loss 

 of moment brought about by the first heating. When a sea- 

 soned magnet is exposed to constantly higher and higher tem- 

 peratures it attains at each stage, in a fraction of a minute, a 

 state which it would keep sensibly unchanged if exposed for 

 hours to that temperature. 



With very few exceptions all the magnets tested had dis- 

 tinctly greater temperature coefficients at higher temperatures 

 than at lower temperatures, so that if the losses of moment in 

 terms of the moment of the original low temperature be used 

 as ordinates and the temperatures as abscissas, the curve is con- 

 cave upwards. In the case of magnets made of a single 

 specimen of tool steel 1*1 l cm in diameter, the concavity was 

 extremely small and in a solitary instance there was a distinct 

 convexity, but other magnets made of the same bar as this 



