Magnetic Influence of Solar Light 379 



oxides, as of iron and cobalt, belong themselves to the class 

 of independent magnetic bodies, It is more than probable 

 that the red colour of the glass employed in these experi- 

 ments was imparted by means of oxide of iron in a very 

 large proportion, mixed with a minute quantity of copper, 

 so that had there been evidence of increase of magnetic 

 intensity in the cylinder, it would have been necessary to 

 take the presence of the former substance into considera- 

 tion, and to determine clearly the influence it had produced, 

 ere any decisive inference as to the real source of the 

 magnetism, could have been ventured on. In this point of 

 view, one of Mrs. Somerville's experiments is peculiarly 

 objectionable, and cannot possibly, I think, be received as 

 any proof of the magnetic action of light. The instance 

 alluded to, is that wherein it is stated that the light was 

 transmitted through a glass coloured blue by cobalt, a body 

 which we well know, possesses the most distinct and de- 

 cisive independent magnetic properties, and which therefore 

 ought never to have been employed in such delicate ex- 

 periments as those on the present subject, in consequence 

 of the certainty of its interfering action, vitiating the re- 

 sults. I regret I had not the means of making some 

 direct experiments with cobalt to shew its interfering 

 effects, but of their existence there can be no reasonable 

 question. 



2. Experiments on the magnetic influence of Solar Light, 

 transmitted through Coloured Silk Media. 



Having by the active and valued kindness of a friend, 

 been supplied with an abundant assortment of coloured 

 silks and cloth, I selected such colours as appeared to me 

 best calculated to suit my purposes, and with these covered 

 half of each experimental cylinder, exposing it subse- 

 quently, with due precaution, to the direct light of the 

 sun. 



The subsequent tables exhibit the results obtained. 



