Electrical Properties of Selenium. 237 



in a mould formed by placing a glass ring, such as is used for 

 microscope slides, upon a glass plate. The thickness of a 

 disk was made approximately uniform, and its surfaces polished 

 by rubbing upon fine glass-paper laid upon a sheet of plate 

 glass. The disk to be tested was inserted between pads of 

 vulcanized rubber covered with tinfoil, which constituted the 

 electrodes, and the whole was placed between hinged boards 

 and loaded with a weight of 28 lbs. (12*7 kilos). It is un- 

 necessary to describe the tests in detail, inasmuch as the only 

 result arrived at was the negative one that crystalline Se has 

 no definite specific resistance. In different tests of disks pre- 

 pared from the same sample and as far as possible in the same 

 manner, it varied from 33 megohms to 1630. 



The specific resistance is certainly not always diminished, 

 as is generally believed, by prolonged heating at a high 

 temperature ; i. e. between 200° and 217°. It is indeed often 

 increased after such an operation, as in the following extreme 

 instance. 



Exp. 2. — A disk of Se was crystallized and heated for an 

 unrecorded number of hours in a copper air-bath at about 

 200°. Its specific resistance was afterwards found to be 38 '5 

 megohms. It was again heated at the same temperature 

 for three hours. When cold its specific resistance was 660 

 megohms. In both cases it was cooled slowly. 



§ 5. Action of Selenides. 



On the other hand, prolonged heating always diminishes 

 the resistance of a Se " cell v having electrodes fused into it. 

 This I have attributed to the action of the conducting metallic 

 selenide which would be formed during the heating at the 

 expense of the wires. 



Exp. 3. — For the purpose of testing this conjecture, a 

 number of cells were constructed of Se having incorporated 

 with it cuprous selenide in different proportions. The wire 

 used was platinum ; it was of the same gauge in all cases 

 (0*27 millim.), and the number of turns per unit of length 

 was also the same (10 per centim.). The surface area, 

 however, varied somewhat, and in order to facilitate com- 

 parison the figures, as given in the annexed table, express in 

 each case the resistance of a square centim. of surface. The 

 light employed was always, except where otherwise stated, 

 that derived from an 8 candle-power incandescent lamp, the 

 filament of which was 1 ft. (about 30 centim.) distant from 

 the face of the cell. None of the cells containing added 

 selenide were annealed, but were cooled as soon as the Se had 



