Experiments ivith Selenium Cells. 251 



known as the black amorphous or vitreous, the red amorphous, 

 and the crystalline forms. Selenium is, I believe, always 

 supplied commercially in the vitreous condition. Vitreous 

 selenium is as structureless as glass, and, when its surface is 

 smooth and clean, it is very like black sealing-wax in appear- 

 ance. In thin films, however, it is seen to be transparent and 

 of a ruby-red colour. It melts, after first softening, at a tem- 

 perature of about 100° C. Ked amorphous selenium is a finely 

 divided brick-red powder, which, at a temperature of from 

 80° to 100°, becomes transformed into the ordinary dark- 

 coloured vitreous form. Crystalline selenium is obtained by 

 keeping melted vitreous selenium at a temperature of from 

 100° to 200° for some time, the black liquid mass being 

 gradually converted into a hard slate-coloured solid. This 

 form of selenium melts at 217° : it resembles a metal in 

 appearance, and even in the thinnest films it is quite opaque 

 to light. 



In the vitreous condition selenium is a practically perfect 

 non-conductor of electricity : so it is also in the powdery 

 condition if perfectly dry. Crystalline selenium, on the 

 other hand, is, according to the books, a good conductor. 

 In comparison with true metals, however, it conducts rather 

 badly. I found some time ago that the specific resistance 

 of a plate of crystalline selenium, which had been annealed 

 for several hours in a glass mould, out of contact with any 

 metal, was 2500 megohms, while that of another specimen 

 annealed in contact with copper electrodes was 0*9 megohm, 

 the lower resistance in the latter case being due, as I then 

 suggested and still think probable, to the formation of copper 

 selenide during the heating. 



A very remarkable fact, first published by Mr. Willoughby 

 Smith in 1873, is that the resistance of crystalline selenium 

 appears to be temporarily diminished by the action of light. 

 This effect is easily shown by means of a small piece of 

 crystalline selenium fitted with wire electrodes and joined up 

 in circuit with a battery and a reflecting-galvanometer. The 

 galvanometer-deflexion is increased when the selenium is 

 illuminated. 



In order that the resistance of a so-called " selenium cell M 

 may be kept as low as possible, and the light made to produce 

 a maximum effect, three conditions have to be aimed at. 

 The electrodes should be large ; they should be near 

 together ; and the selenium, in order to expose a large 

 surface relatively to its volume, should consist of a thin film. 

 Several contrivances have been proposed with the view of 

 fulfilling these requirements. A simple and, I believe, 



