8 ART. 10.— K. HOXDA &. S. SHIMIZü. 



In experiments at high temperatures, the internal field for a 

 given magnetizing current varies with the temperature, as the inten- 

 sity of magnetization changes with it. A full knowledge of the field 

 in which experiments were carried on requires the determination of 

 the intensity of magnetization at each temperature and field. Since 

 our experiments did not extend so far, the curves of the length 

 change at different temperatures were drawn for the external field. 

 But if we refer to Fig. 1, it is easy to see how the forms of these 

 curves are to he changed, if the effective field be used instead of the 

 external one. 



Nickel. The change of length in nickel under high and low 

 temperatures is graphically shown in Fig. 2. The rise of temperature 

 markedly reduces the magnetic contraction of the metal. At a 

 temperature of 240°C, the contraction in H = 800 is already reduced 

 to half its ordinary value, and at 400°C, it almost vanishes. With 

 the ovoid of the same specimen, Professor H. Nagaoka and Mr. S. 

 Kusakabe found the critical temperature to be 400°C. In liquid air, 

 the contraction is reduced in weak fields, but is increased in strong 

 fields. The relation between the change of length and the tem- 

 perature for given external fields is given in Fig. 3. Each curve has 

 a minimum point, the temperature of which decreases as the field is 

 increased. 



We also notice that the contraction vanishes asymptotically, as 

 the temperature approaches to 400°C. It is to be remembered that on 

 account of the demagnetizing force, each curve does not represent the 

 contraction in a constant effective field, but shows the general feature 

 of contraction with regard to temperature. The former results are 

 consistent with the corresponding results'in the present experiment. 



Soft iron. The change of length in soft iron is ffiven in 

 Fig. 4. As the temperature is raised," the contraction in high fields 



