chemical Properties attributed to Light. 451 
through the riband, and I found the stain to be perfectly in- 
delible. I endeavoured to wash it out ; but nothing I applied 
to it, and among other things I tried super-oxygenated marine 
acid, appeared in the smallest degree to diminish its lustre. 
The hue was not uniform, but varied from a light crimson to 
a very deep purple, approaching to a reddish browm. 
I searched, but in vain, for traces of revived gold, in its re- 
guline form and colour ; but, though I could not perceive that 
the riband was gilded, it had all the appearance of being co- 
vered with a thin coating of the most beautiful purple enamel, 
which, in the sun, had a degree of brilliancy that was some- 
times quite dazzling. 
Experiment No. 2. A piece of the riband which had been 
wetted with the aqueous solution of the oxide, was carefully 
dried in a dark closet, and was then exposed, dry, over the 
flame of a burning wax candle. The part of the riband which 
had been wetted with the solution, (and which on drying had 
acquired a faint yellow colour,) was tinged of the same bright 
purple colour as was produced in the last-mentioned experi- 
ment, when the riband was exposed wet to the action of the 
heat. * 
Experiment No. 3. A piece of the riband which had been 
wetted with the solution, and dried in the dark, was now wetted 
with distilled water, and exposed wet to the action of the ascend- 
ing current of hot vapour which arose from the burning candle : 
the purple stain was produced as before, which extended as far 
* We shall hereafter find reason to conclude, that the success of this experiment, or 
the appearance of the purple tinge, was owing to the watery vapour which existed in 
the hot current that ascended from the flame of the candle. 
3 M 2 
