Resistance of Selenium ivith Temperature. 



465 



that the experiment should be repeated in all its details in 

 exactly the same way, with the same rate of variation of 

 temperature. It is clear that in order to be sure that the 

 thermo-couple records truly the temperature of the whole of 

 the selenium, changes of temperature, whether heating or 

 cooling, must be arranged to take place sufficiently slowly. 

 This precaution has been taken, with the result that the 

 cooling curves show satisfactory agreement with the heating 



o 



curves 



3. Results. 

 In each case the resistance observed was plotted against 

 the temperature; and for the several cells of the same dimen- 

 sions examined, although they differed considerably in their 

 initial resistance, the form of the curve was the same, and at 



Fi- l. 



70x10 



60x10 



50*10 



40x10 



30x10 



cc 20x10 



10x10 



75 100 125 150 



Temperature in degree centigrade-* 



temperatures higher than about 130° C. the different curves 

 very nearly coincided with one another. Two typical cases 

 are shown in fig. 1. Curve I. shows the effect of time in 

 Phil. Mag. S. 6. Vol. 42. No. 249. Sept. 1921. 2 I 



