90 



The Ohio Journal of Science [Vol. XVI, No. 3, 



The first reading shown was taken one minute after the 

 potential was applied. This was permitted to cool to room 

 temperature and remain in that condition without the applica- 

 tion of a potential for two days. The specimen was reheated to 

 ■approximately the same temperature as before and the same 

 potential applied. Figure 5 shows the result on the same scale. 

 The first reading shown is smaller than the first one in Figure 4 

 and was taken 15 seconds after the potential was applied. The 

 potential indicated was left on for four hours and the specimen 

 then allowed to cool. It was reheated after 20 hours and the 

 curve in Figure 6 obtained. 



Time m ITI mutes 



This and many similar experiments show that the resistance 

 of bodies of this character cannot be determined in a satis- 

 factory way by direct current measurements. The behavior 

 of the body is determined in part by its previous history. In 

 this respect it is analogous to the behavior of magnetic bodies. 

 It would be of interest to put these bodies through a cycle such 

 as is done in the study of magnetic hysteresis. In fact a possible 

 way of throwing more light on the interesting and difficult 

 subject of dielectric hysteresis might be carried out in this 

 manner. The author had in mind attempting such a series 

 of experiments, but the close temperature regulation required 

 .and the length of consecutive time required to take a series of 

 observations did not appear feasible for one carrying a teaching 

 schedule. 



Physical Laboratory, Ohio State University, December, 1915. 



