Electrical Properties of Selenium. 251 



traced and found to terminate very near the Pt electrode, 

 where no doubt platinous selenide existed. 



This experiment, though not conclusive, tends to show that 

 Se unmixed with selenide is not acted upon by the current. 

 It is possible that moist crystalline Se may be converted into 

 the red amorphous form by electrical action, but the question 

 cannot be definitely settled until perfectly pure Se is obtain- 

 able. 



§ 14. Photoelectric Currents. 



Another of the interesting discoveries of Adams and Day 

 was the fact that an electromotive force could be set np, and 

 a current started in annealed Se merely by the action of 

 light*. This observation has been fullv confirmed bv sub- 

 sequent experimenters, and the belief has been expressed that 

 the phenomenon affords an example of the direct conversion 

 of the energy of light into that of electricity. Others have 

 held the opinion that the effect is of a purely thermoelectric 

 nature. My own investigations have led me to the following 

 conclusions : — 



(1) The effect is undoubtedly due to light and not to heat ; 

 or, more accurately, it is brought about by the action of radia- 

 tion, and not by change of temperature. 



(2) The energy of the photo-electric current is primarily of 

 chemical or voltaic origin, and is not derived from radiation. 

 Light merely has the effect of facilitating the chemical pro- 

 cess, and can have no further influence when the supply of 

 unused material is exhausted. 



Exp. 27. — A rectangular plate of Se measuring 14 X 11 x 2 

 millim. had Pt wire electrodes fused into its two ends and 

 was " annealed " for several hours. The electrodes being 

 connected with a galvanometer, a piece of magnesium ribbon 

 was burnt near the Se : a deflexion immediately occurred 

 showing a current which ceased as soon as the light was 

 extinguished. By screening the two ends alternately, it was 

 ascertained that the effect was produced only at one junction, 

 the other being quite insensitive. (It may happen in differ- 

 ent cases that both junctions are sensitive or neither. As 

 will be shown later, sensitiveness points to imperfect electrical 

 contact between the Se and Pt due perhaps to a film of 

 red Se.) The direction of the current through the sensitive 

 end was from Se to Pt. 



The Se plate was afterwards placed under a glass receiver 

 with a basin of water, and connected with the galvanometer 

 by wires passing through a caoutchouc stopper. In a minute 



* L. c. p. 333. 



S 2 



