with Crystalline Selenium. 411. 



(which were nearly filled with distilled Fig 4. 



water) two smaller horizontal glass 

 tubes, c and c f , were fixed, over each of 

 which a small saddle of white blotting- 

 paper, intended to act as a conducting 

 pad, was laid. The ends of the blotting- 

 paper dipped into the water, and formed 

 the connexion with a plate of crystalline 

 selenium, s s', which rested on the top. 

 The circuit was completed by inserting 

 a galvanometer (G) between the strips of 

 platinum foil. The whole was mounted 

 in a light-tight case with a diaphragm, 

 d d', so that light could be excluded 

 from or admitted to either side at plea- \ ^. } 



sure. When light was admitted to one ~~\Dg-~ 



side, it had of course to reach the contact 



surface of the selenium, after passing through the glass, the 

 water, and the blotting-paper pad. It was therefore diffused 

 and much weaker than had it fallen directly upon the face of 

 the selenium as in the previous experiments. The indica- 

 tions were therefore less in amount, but nevertheless perfectly 

 distinct. 



Next, instead of admitting light to the selenium plate, both 

 of the tubes were kept dark, and warm water was passed 

 through either c or c' ', so that the wet blotting-paper pad upon 

 it, and therefore the contact face of the selenium and water, 

 was slightly warmed. A current was immediately observed. 

 When this current had subsided, warm water was passed 

 through the other tube, and the new deflection noted. 



Now in both these tests it is evident that we are dealing, in 

 respect of light and heat, with the surface of the selenium only, 

 and have therefore eliminated any effects due to the molecular 

 condition of the interior. 



Lastly, the plate of selenium was removed, and a small bar 

 of copper and zinc, soldered together in the middle, was laid 

 across the pads, for the purpose of determining the direction 

 in which heat and light had acted in modifying the electro- 

 positiveness of the selenium. 



End c. End c'. 



(1) illuminated dark 



(2) warmed cold 



(3) copper zinc 



(4) dark illuminated 



(5) cold warmed 



(6) zinc copper 



The following are the results :- 

 Deflection. 

 -200 div. 

 -160 „ 

 — off scale. 

 + 50 div. 

 + 80 „ 

 + off scale. 



