42 



K. TSURUTA. 



facility that which, whatever it is, mainly contributes to the produc- 

 tion of the hysteresis, comes and goes with agitation. We cannot 

 indeed help thinking of the thermo-electric hysteresis as due to some- 

 thing like molecular viscosity or constraints in iron. 



§ 17. The following experiment on the same wire as in § 12 

 was made to complete the experiment of § 14. 



After a cycle of loading and unloading had been gone through 

 and the familiar curve obtained (not here given), the wire was tapped 

 at a certain stage (6 kilos.) in the next process of loading, the repre- 

 sentative point being thus shifted below (See Fig. 16., PJ. III). The 

 wire was then unloaded down to 1 kilo, and successively loaded 

 on again. The course of the curve is exactly the same as that we had 

 before in the minor cyclic curves. Arriving at the uppermost limit 



of loading (11 kilos.), I began to unload and tapped the wire again 

 at a certain stage (2 kilos.). The effect of hysteresis being thus 



