Chemical Properties attributed to Light. 95 



SUPPLEMENT. 



SINCE the foregoing paper was presented to the 

 Royal Society, I have had an opportunity of prosecut- 

 ing these inquiries a little farther; and the results of 

 two of my late experiments were so remarkable that I 

 have thought them deserving of being made known to 

 the public. 



Experiment No. 17. Into a thin globe of fine col- 

 ourless glass, about ij inches in diameter, with a short 

 cylindrical neck, I put equal parts of a weak solution of 

 gum arabic in water and of a diluted aqueous solution 

 of the oxide of gold ; and filling the globe about two 

 thirds full with these liquids, which being well mixed 

 together by shaking, the globe was suspended to a nail, 

 by its neck, near a window in an unfrequented room 

 fronting the north, where by accident it happened to 

 remain undisturbed and unobserved six weeks. 



When the globe was examined, it exhibited a very 

 curious appearance. The glass was beautifully tinged 

 in every part where it had been in contact with the 

 liquid, but the hues were very different in different 

 parts. The part of the globe in contact with the upper 

 surface of the liquid was of a very faint purple, but 

 this tinge gradually became of a deeper colour as it 

 descended by the sides of the globe, and ended below 

 in a rich gilding, which had all the metallic splendour 

 of pure gold. 



Experiment No. 18. Having provided a thin slip of 

 ivory, about half an inch wide and 3 inches long, I 



