Length of Ferrotnaffnetic Suhstances hy JLu/netizatio)!. '.\\)^ 



the field of maximiiin contraction grailually decreases as 

 the temperature is raised, and that the temperature of 

 maximum elonoation in a oiven tiehl diminishes as the field 

 is increased. 



With the same specimen the effect of high temperature 

 was also first studied, and when the specimen was co(d(Kl 

 tlown to its initial temperature, it totally changed its cha- 

 racter with regard to the magnetic change o£ length. It 

 was therefore not possible to examine the efPect of cooling in 

 the metal in its cast state. 



Amiealed cobalt. — The effect of high temperature> on an- 

 nealed cobalt presents an extraordinary feature, as may be 

 seen from tig. 13. The specimen was annealed in a charcoal 

 fire for about four hours after being carefully wrapped in 

 asbestos-paper. The change of length at ordinary tempera- 

 ture is abnormal. As the temperatAire is raised, beginning 

 with that of liquid air. the contraction increases at first >lowly 

 and then rapidly, till it reaches a maximum. It then de- 

 creases, and after passing the state of no contraction it is 

 changed to an elongation which again increases with the 

 temperature up to a maximum, and then gradually diminishes. 

 At such a high temperature as 1034^ C. we could still ob- 

 serve a considerable elongation of the metal. To judge from 

 the course of the curves, the temperature at which the elon- 

 gation at last vanishes is a little lower than in cast cobalt. 

 It is interestino- to observe that the curve of the length- 

 change at a temperature near 450° C. is similar to that of 

 iron at ordinary temperature. The cobalt slightly elongates 

 in weak fields, but it contracts in strong fields. At tempera- 

 tures higher than 500° C. the cast and annealed cobalts re- 

 semble each other in their behaviour in respect of the change 

 of length. 



The curves (fig. 14) showing the relation between the change 

 of length and the temperature present a peculiar feature, 

 having generally one maximum and one minimum. In low 

 fields, however, two small maxima and minima are observed. 

 Since each curve in the figure almost passes through a point 

 (464° C.) on the axis of temperature it follows that there is 

 a certain temperature at which the change of length in 

 annealed cobalt nearly disappears for all magnetizing fields, 

 and that the change occurs in an opposite sense in every field, 

 according as the specimen is heated above or l)elow that 

 temperature. It appears, then, that annealed cobalt under- 

 goes some molecular change at that temperature. 



(jreneral remarks, — On comparing the above results in soft 



