358 EDNA D. SAYCE. 



e.m.f. was measured independently by means of a Wolff 

 potentiometer, and the temperature calculated from tables 

 prepared from the calibration readings. The only other 

 alteration in the method was the substitution of a Weston 

 milliammeter and voltmeter for the wattmeter which Lees 

 used to measure the rate at which heat was supplied. 



The discs were first prepared in the vitreous 1 form by 

 pouring molten selenium into a hot iron mould. These were 

 transformed to the crystalline state by heating in an oil 

 oven to some chosen temperature. They were then ground 

 with fine emery and polished. 



On account of the contraction which takes place when 

 changing from the vitreous to the crystalline form, some 

 difficulty was experienced owing to discs cracking during 

 heating. This was overcome by performing all operations 

 very slowly. The vitreous discs were gradually cooled 

 from the molten state, and the oven was so regulated that 

 the required temperature was only reached after seven 

 hours continuous heating. This temperature was main- 

 tained for one hour, and the disc was allowed to cool slowly 

 in the oven. By this means crystalline discs were prepared 

 at temperatures which range from 160° 0. to 214° O. Some 

 specimens were also tested in the vitreous form. 



If the bath in which the discs were tested were heated 

 by a suitable current in the bath heating coil, the conduc- 

 tivity could be measured at any temperature higher than 

 that of the room. An upper limit is, however, placed on 

 the temperature of testing by the fact that selenium is so 

 readily changed by heating even to comparatively low 

 temperatures. Thus the available range was only about 

 25° 0. to 55° C. In order to obtain sufficiently steady tem- 

 perature conditions, it was found necessary to keep the 

 current in the heating coil constant for approximately six 



1 A. P. Saunders, Jour. Phys. Cheai., Vol. iv, p. 423, 1900. 



