10 



AKT. 8.-H. XAGAOKA AND K. HONDA 



then (jb.served by nicaii.s of :i inicroscope provided with uiicrometer 

 ocular. l>y using filar micrometer, the accuracy of tlie measurement 

 may be increased several times, but it was found (|uite superfluous. 

 One division of the micrometer, the tentli of whicli can be easily esti- 

 mated, was equivalent to an elongation of 4.35 x 10~^ 



Fig. c shows the horiz(3ntal plan of the arrangement while the 

 perspective view is given in Fig. d. 



Figr. d. 



i 4. Change of Length in Differenl Ferromagnelic Bodies. 



AVe consider the different sorts of iron, w^hich were examined in 

 our former experiments as homogeneous, as the general character of 

 the change is both qualitatively and quantitatively nearly alike, and 

 they are ahvays subject to very careful annealing. Any new 

 experiment would therefore have been superfluous. 



(rt) Cast Steel (Fig. 2, PI. I.). 



Dr. H. du Bois was kind enough to give us a piece of cast steel, 

 which was regarded as the most homogeneous specimen at present 

 obtainable. The rod was turned into an ovoid and its chang-e of lensfth 

 measured in the usual manner with the followinjï result. 



