with Crystalline Selenium. 409 



cipation of a changed (increased or decreased) resistance, 

 before inverting the selenium, the galvanometer (introduced 

 immediately after the inversion) showed a deflection always 

 indicating that the resistance had not entirely changed but 

 had still a value in the direction of that last measured ; and it 

 required two or three minutes to arrive at the settled value 

 corresponding to the new conditions. This difference was 

 small but distinct, and would be such as would correspond with 

 change from heat to cold or vice versa. 



On removing the battery and inserting a galvanometer, the 

 discharge current which issues from the selenium is in the 

 opposite direction to the battery current, and agrees with the 

 thermoelectric current which would result from the Peltier 

 effect. 



On the other hand, the supposition that the behaviour is 

 due to polarization is the more probable. The increase of the 

 resistance with increasing current (when weak) in one direc- 

 tion, indicates the existence in the selenium of a small inde- 

 pendent electromotive force, and leads to the suspicion that a 

 portion of the material in contact with one of the platinum 

 electrodes is in an electrolytic condition, or both perhaps, one 

 being more so than the other. As the measuring current in- 

 creases in strength, it appears to polarize the electrodes in the 

 selenium, as in an ordinary electrolytic conductor, and the 

 small independent current is overpowered and lost sight of. 

 The apparent decrement of resistance by increasing the battery, 

 is probably due to the fact that the polarization increases in a 

 less ratio than the measuring current, so that when this current 

 is weak the polarization is proportionally stronger, and the 

 apparent resistance higher, than when the measuring current 

 is strong. The discharge after removing the battery is such 

 as would answer to the depolarization of an electrolyte. 



Electromotive force of crystalline Selenium. — The action of 

 light in modifying the conductivity of selenium is evidently 

 a surface action, the effect of which penetrates very little, if 

 at all, into the mass. It therefore occurred to me that the 

 phenomenon could best be studied, particularly in relation to 

 heat, by making up the selenium plate in the form of a gal- 

 vanic element. By this means we can deal with the surface 

 without reference to the interior, both as regards light and 

 heat. 



A plate of crystalline selenium was prepared at 200° C. with 

 a platinum wire fused into it, by which it was suspended in a 

 test-tube. The back of the plate and the platinum wire near 

 it were covered with a black insulating varnish. The tube 

 was then placed in a light-tight box, in which a shutter could 



