264 



exceptions possibly attributable to some predisposition in the needle 

 itself to magnetism too slight to be observed. The distance of the 

 needle from the prism was varied without materially varying the 

 effect. It was found unnecessary to darken the room, provided the 

 spectrum was thrown out of the direct solar rays. 



The next object was to ascertain whether the other prismatic 

 colours had the same property as the violet. Needles (previously 

 ascertained to be unmagnetic) exposed to the blue and green rays 

 sometimes acquired magnetism, though less frequently, and requiring 

 longer exposure ; but the magnetism when communicated seemed 

 equally strong as in the case of the violet rays. The indigo suc- 

 ceeded nearly as well as the violet. In all cases, the exposed end 

 usually became a north pole. But in no one instance was magnetism 

 produced by the yellow, orange, or red rays, though in some cases the 

 same needles were exposed to their influence for three successive 

 days, neither did the calorific rays produce any effect. 



Pieces of clock- and watch-spring were next tried with similar 

 success, and they were even found more susceptible of the peculiar 

 magnetic influence than needles ; possibly from their greater pro- 

 portional surfaces, or their blue colour. The violet rays concen- 

 trated by a lens produced magnetism in a shorter time than the 

 prism alone. 



Experiments were next instituted by transmission of the solar rays 

 through coloured media. Three needles, half covered with paper as 

 before, were exposed on a stone outside a window, under a blue glass 

 coloured by cobalt, to a hot sun for three or four hours. They were 

 rendered feebly magnetic, and when examined next day were found 

 to have lost their magnetism ; a circumstance found occasionally to 

 take place afterwards, as the season advanced and the force of the 

 sun diminished. No iron was near, nor did the stone show any signs 

 of influence on a suspended magnetic needle. The experiment was 

 repeated next day, leaving two needles six hours exposed, when the 

 magnetism communicated was found to be very sensible, and is still 

 retained by them at the distance of six months. 



Pieces of clock-spring which had been heated to deprive them of 

 magnetism also became magnetic when so exposed. The rays trans- 

 mitted through the glass employed in this experiment, blackened mu- 

 riate of silver as powerfully as those transmitted through uncoloured 

 glass, proving its free permeability to the chemical rays. 



Green glass 'was next tried, and the rays which penetrated it were 

 found to communicate magnetism. 



Neutral pieces, of watch-spring were then half exposed (when 

 wrapped in violet and green ribands, and fastened inside of the 

 frame of a window,) to the .sun's rays all day. In the evening both 

 had become magnetic, and the parts which received the rays trans- 

 mitted through the ribands became north poles. 



Similar pieces were then freely exposed (half covered as all along 

 with paper) to the white light of the sun, with neither prism glass 

 nor riband ; and though the heat was greater than before, no mag- 



