162 The Earl of Rosse on the Electric 



sensibility of selenium to the various rays of the solar spec- 

 trum. 



As it was the opinion of one or two friends of much experi- 

 ence in kindred questions in physical science that heat, not 

 light, is the active agent, and that selenium might with advan- 

 tage replace the thermopile in the measurement of radiant heat, 

 I procured the necessary apparatus for a preliminary trial. 



Although my experiments are hardly sufficient to justify a de- 

 cided opinion in favour of or against its applicability to photo- 

 metric purposes, I think that they remove all hope of selenium 

 replacing the thermopile in the measurement of heat of low 

 refrangibility. 



I exposed a rod of selenium during a sufficient interval to 

 obtain the maximum effect from the light of a candle 3^ inches 

 distant. The diminution of resistance produced amounted to 

 24*3 per cent. A second trial gave a diminution of 24'2 per 

 cent., the candle being at 4 inches distance. A vessel of hot 

 water of about 9 or 10 inches diameter and 9 inches height pro- 

 duced no perceptible change when it stood for several minutes 

 with its centre at 13 inches distance, nor even when it was 

 brought 6 inches nearer. On the other hand, by means of the 

 thermopile a deviation to the extent of 136 divisions of the scale 

 of a Thomson's galvanometer was produced by the radiation from 

 the vessel of hot water whose centre was at 13 inches distance; 

 and when it was replaced by the candle at 4 inches distance, 

 the deviation was increased to only 315 divisions. Thus it was 

 shown that the two instruments were not comparable as measu- 

 rers of the radiation from the two sources of heat. 



The change of resistance produced by exposure to the radia- 

 tion from a candle when a sheet of glass was alternately inter- 

 posed and removed was next measured. The effect due to ab- 

 sorption by the glass appeared to be small, certainly not much 

 more, possibly even less than the average absorption of light by 

 glass. Perhaps 90 per cent, was transmitted. Owing to the 

 inconstancy of the light, much accuracy was not readily obtain- 

 able in this experiment. The piece of glass had been shown by 

 the thermopile to transmit 80 per cent, of the solar rays, and 

 under 1 per cent, of the radiant heat from a blackened tin ves- 

 sel of hot water. It was now by the same means ascertained 

 that 48 per cent, of the radiant heat from a candle was trans- 

 mitted by it. 



A glass cell filled with a solution of alum was now taken; and 

 it was found that, while the exposure of the selenium bar to the 

 radiation from a naked candle at 3| inches distance produced a 

 diminution of resistance amounting to 9*4 per cent., when the 

 alum solution was interposed the decrease was still 8*95 per 



