398 Messrs. K. Honda and S. Shimi/Ai on the Chancfe of 



<lecidedly large in strong fields, producing an increase of 

 contraction. 



The relation between the change of length and the teni- 

 perature is given in fig. 7. It is remarkable to observe 

 that the maximum elongation in weak fields, which is 

 characteristic for iron, remains almost constant for tempe- 

 ratures ranging from —18(5'^ C. to 200° C. Above this 

 temperature the elongation increases till it reaches a maximum, 

 and then rapidly decreases. 



Tunqsten steel, — The results of experiments in tungsten steel 

 are given in figs. 8, 9, & 10. The course of the curves 

 and its variation with temperature are similar to those o£ soft 

 iron at temperatures higher than 500° C. The change of 

 length seems to disappear nearly at its critical temperature, 

 namely 900° C, a value obtained by Prof. H. Xagaoka and 

 Mr. S. Kusakabe. The former result obtained by one of us 

 approximately agrees with the corresponding result in the 

 present experiment. 



With tungsten steel we first studied the effect of tempera- 

 ture ; when the specimen was cooled down to its initial 

 temperature it underwent a considerable permanent change 

 with regard to the change of length. So the experiment in 

 liquid air was performed with another rod of square section 

 cut from the same specimen as the cylinder. The curves for • 

 10° 0. and —186° 0. are given in figure 10, which shows a 

 slio-ht effect of cooling on the change of length. Coolino- 

 decreases the elongation of the alloy in weak fields, but 

 increases it in strong fields 



Cast cobalt. — The results of observations on cast cobalt are 

 shown in figs. 11 & 12. As the temperature is raised the 

 magnetic contraction in weak fields gradually lessens, and the 

 elongation in strong fields increases till it reaches a maximum. 

 At temperatures higher than 800° (J. the initial contraction 

 altogether disappears, and the course of the curves resembles 

 that of iron and steel at high temperatures. If the tempe- 

 rature be further increased the elongation diminishes steadily 

 but at a diminishing rate, and even at such a high tempera- 

 ture as 1020° C. we still observe a considerable elongation of 

 the metal. From the course of the curves in fig. 12 it is easy 

 to see that for H = 800 the elongation does not vanish up 

 to a temperature of 1200° C, which is higher than its critical 

 temperature by 100° C. 



With our arrangement it w^as not possible to push the 

 experiments still further, as the melting-point of copper was 

 not far from that temperature. It is also to be observed that 



