408 EXPERIMENTAL INVESTIGATIONS. 



result obtained ; i.e. on removing the brass cylinder there was 

 a deflection of 12, the platinum exposed to light being posi- 

 tive to the sheltered one. 



This identity of electrical effect taking place with the re- 

 versed plates seemed so strongly in favour of the impact of the 

 solar rays having an initiatory effect in producing a voltaic 

 current, that the only remaining point seemed to be to ascer- 

 tain whether it was due to light or heat, to the chemical or 

 calorific rays of the sun ; yet the conclusion I then came to 

 was erroneous, as will presently be seen. 



In order to ascertain how far the effect was due to heat, I 

 arranged, in a room lighted by a small candle, the same appa- 

 ratus over a fire of asbetos heated by coal-gas, so that both 

 radiant heat and an ascending current of hot air impinged on 

 the side of the glass in which was the exposed platinum, while 

 the opposite side was entirely sheltered from the heat by a 

 metallic shelf, on which the cell rested ; this experiment was 

 continued until one side of the cell was uncomfortably hot to 

 the hand, while the other side was quite cool ; but not the 

 slightest deviation of the galvanometer took place. 



I now repeated the former experiment with sunlight, 

 changing the liquid each time. In three successive experi- 

 ments the deflections on the impact of light were in the same 

 direction, the exposed platinum being positive ; but in a fourth 

 the deflection was in the reverse direction, the exposed platinum 

 being negative. In several subsequent experiments there was 

 always a notable deflection which ensued on the impact of 

 light ; but it was sometimes in one direction and sometimes 

 in the other. I ultimately discovered that, in the deflection 

 produced by light, the needle of the galvanometer deviated 

 in the same direction which it took upon the first contact of 

 the wires connected with the platinum plates. The effect of 

 light was therefore to increase the deflection occasioned by 

 the polarisation of the platinum plates ; and this my subse- 

 quent experiments have, I think, fully established. Although 

 the experiment on the impact of heat seemed to show that the 

 heating effect of the solar rays was not the cause of the 

 phenomena, yet it might well be that the solar rays absorbed 

 by the platinum black, would produce a greater heating effect 



