Electrical Properties of Selenium. 245 



In a former communication I stated my belief that these 

 polarization effects resulted from the electrolysis of the 

 metallic selenides contained as impurities in the Se. Whether 

 this be so or not, the following experiments plainly indicate 

 that the presence of moisture is necessary for the production 

 of the phenomenon. 



Exp. 17.— The electrodes of a Se cell were connected first 

 with the poles of a battery, and then by means of a " change- 

 over " switch with a galvanometer : this produced a deflexion 

 of 90 divisions, indicating an E.M.F. of rather more than y^ 

 volt. The cell was then placed in a glass receiver over 

 strong sulphuric acid, the connecting wires being passed 

 through an indiarubber stopper. From time to time it was 

 successively joined to the battery and the galvanometer, and 

 the polarization current was observed to be gradually di- 

 minishing. After 18 hours the deflexion was only 3 divisions. 

 The air was then exhausted from the receiver, and 24 hours 

 later no polarization current whatever could be detected, even 

 after prolonged connexion with the battery. A current 

 greater than ^q part of that which occurred in ordinary air 

 would certainly not have escaped observation. 



Exp. 18. — The cell was afterwards suspended for an hour 

 in a receiver over a basin of water and then tested for 

 polarization. The spot of light was immediately deflected far 

 beyond the limits of the scale. 



Probably therefore the polarization results merely from the 

 electrolysis of water ; in any case the presence of water seems 

 to be requisite for its production. 



It is perhaps worth noting that little if any polarization is 

 exhibited by crystalline Se of the coarsely granular form ob- 

 tained by gradually lowering the temperature of the molten 

 substance, instead of by heating solid vitreous Se. The 

 specific resistance of this form is much lower than that of the 

 other, and its sensitiveness is also less. 



§ 12. The Effect of Moisture upon Conductivity and 



Sensitiveness. 



The water ordinarily present in Se, even if it had no other 

 effect, would naturally influence its apparent conductivity by 

 acting as a shunt or alternative path for the passage of the 

 current. We should expect, therefore, that if a cell were 

 made as dry as possible, its resistance would be increased, and 

 this was found to be the case. 



Exp. 19.— A cell with a Se surface measuring 4 centim. by 

 1*5 centim., and containing 64 turns of 0*19 millim. copper 

 wire, was suspended in a receiver over sulphuric acid, and tests 



