240 Mr. Shelford Bid well on the 



5 per cent, of selenide was added to the insensitive Se. Its 

 resistance in the dark was 105,000 ohms, and when illumi- 

 nated by the lamp 70,000, a fall of 25 per cent. 



Exp. 8. — Another similarly sized cell had 3 per cent, of 

 lead selenide mixed with the Se. Its resistance dark was 

 363,000 ohms, and illuminated 10^ per cent. less. 



It appears, therefore, that the addition of a small quantity 

 of cuprous selenide imparts a fair degree of sensitiveness to 

 the anomalous Se, at the same time bringing down its resist- 

 ance to nearly the normal value. An equal proportion of 

 lead selenide produces a similar, though less marked effect. It 

 would be interesting to try selenides of arsenic and other 

 metals, but I have not yet had time to do so. 



It is certain that an ordinary well " annealed " Se cell 

 with copper electrodes contains a relatively large quantity of 

 copper selenide. Mr. W. Hibbert was kind enough to analyse 

 the Se taken from one of my old cells to which no foreign 

 substance had been intentionally added, and found that it 

 contained 2*28 per cent, of combined copper, equivalent to 

 3*71 per cent, of cuprous selenide. Of course this must have 

 been derived mainly from the copper electrodes. 



§ 7. Effect of Time. 



The fact that the resistance of crystalline Se with metallic 

 electrodes fused into it decreases very greatly in course of 

 time was first noticed by Prof. W. G. Adams and Mr. R. E. 

 Day*, who thought that it was due to the Se having become 

 more completely annealed. In a former paper f I have sug- 

 gested that the decrease is more probably to be attributed to 

 the short-circuiting of the cell by an excessive amount of 

 conducting selenide, formed by the gradual union of the Se 

 with the metal of the electrodes. This view receives strong 

 support from the following observations. 



Exp. 9. — In the year 1891 I constructed a copper- wire 

 cell but did not crystallize nor anneal the Se, which remained 

 therefore in the vitreous state : its resistance was sensibly 

 infinite. The cell was laid aside in a box until March 1895, 

 when it was tested and its resistance found to have fallen to 

 about 8 ohms. The cell was quite unchanged in appearance, 

 the selenium surface being perfectly black and glossy. The 

 decrease of resistance could not possibly be accounted for by 

 spontaneous crystallization, and could hardly have been due 

 to any other cause than the formation of selenide in the 



* Phil. Trans, vol. clxvii. pt. i. p. 348. 

 f Phil. Mag. March 1891, p. 250. 



