chemical Properties attributed to Light. 45,5 
the riband had always acquired, when exposed to the light in 
the preceding experiments. 
On taking the riband out of the phial, and wetting it tho- 
roughly with distilled water, and exposing it again, while thus 
wetted, to the sun’s rays, it almost instantly began to change 
colour, and soon became of a deep purple tint ; but, though I 
examined the surface of the riband with the utmost care, and 
with a good lens, both during the experiment and after it, I 
could not perceive the smallest particle of revived gold , nor did 
I see any vestige remaining that appeared to indicate that any 
had in fact been revived. 
This experiment was repeated several times, and always with 
results which led me to conclude, (what indeed was reasonable 
to expect,) that light has little effect in changing the colour of 
metallic oxides, as long as they are in a state of crystallization. 
The heat which is generated by the absorption of the rays of 
light must necessarily, at the moment of its generation at least, 
exist in almost infinitely small spaces ; and consequently, it is 
only in bodies that are inconceivably small that it can produce 
durable effects, in any degree indicative of its extreme intensity. 
Perhaps the particles of the oxide of gold dissolved in water, 
are of such dimensions ; and it is very remarkable, that the co- 
lours produced, in some of my experiments on white ribands, 
Jby means of an aqueous solution of the oxide of gold, are pre- 
cisely the same as are produced from the oxide of that metal, by 
enamellers, in the intense heat of their furnaces. 
As the colouring substance is the same, and as the colours 
produced are the same, why should we not conclude that the 
effects are produced in both these cases by the same means, 
that is to say, by the agency of heat ? or, in other words, and 
