194 



THE MAGNETIC EFFECTS OF LIGHT. [MEMOIR XII. 



Fig. 24. 



compass-boxes are generally made of sheet-brass with a 

 soldered seam in the side, it was possible that the fine 

 line of solder acted with the brass as a thermo-electric 

 couple, capable of excitation by the warmth of the sun- 

 beam. I therefore made a compound cylinder, half of 



copper, the other half of zinc, c z, 

 Fig. 24, the edges of which, at a 

 and b respectively, are soldered to- 

 gether, one junction, , being pol- 

 ished, the other, #, blackened. The 

 needle was suspended in an ex- 

 hausted receiver by a silk fibre, </, 

 ^ and a ray of light, rf, coming from 

 an aperture half an inch wide in 

 the shutter fell upon the junction. 

 The needle used in this experiment 

 was of watch-spring. Its first vi- 

 bration was performed in an arc of 40, and when the 

 compound cylinder was taken away it made thirty-two 

 vibrations in sixty seconds in vacuo. On placing it con- 

 centrically with the compound cylinder, and suffering 

 the ray to impinge on the polished junction, the moment 

 that the arc of vibration had become 40 the number of 

 oscillations in one minute was observed ; six experiments 

 gave severally thirty -two. On turning the blackened 

 junction to the light the result was still thirty-two, and 

 on substituting the solid brass cylinder three consecutive 

 trials gave thirty -two. The thermometer stood in the 

 sunshine at 103. 



By some this magnetic action of light has been at- 

 tributed to the violet or more refrangible rays only. A 

 needle of watch-spring about four inches long, which in 

 an exhausted receiver suspended by a filament of silk 

 exhibited no polarity, had one half exposed to the violet 

 ray dispersed by a flint-glass prism. This ray was sepa- 



