40 & Researches respecting the 
place which you wish to stain yellow, and to heat the at 
tide gently to give it the colour. You must not employ 
too strong a heat : the degree will easily be found by 
practice. When the article has been sufficiently heated, 
you take it from the fire and separate the coating of oxyd, 
which will be found reduced to a regulus : you will then 
observe the place which it occupied tinged of a beauti- 
ful yellow colour, without thickness. It is chiefly on 
transparent glass that this process succeeds best. Very 
fine silver filings produce the same effect : but what seem - 
ed to me to succeed best in this case, was sulphat of 
silver, well ground up with a little water, that it may be 
extended very smooth. From what has been said, it 
may readily be seen that this yellow must not be em- 
ployed like other colours ; that it must not be applied 
till the rest have been fused ; for, as it is exceedingly fu- 
sible, and ready to change, it would be injured by the 
other colours; and as the coating of silver which is re- 
duced must be removed, the fluxes would fix it, and pre- 
vent the possibility of its being afterwards separated. 
Working on glass is not attended with this inconve- 
nience, because the silver-yellow is applied on the 
opposite side to that on which the other colours are 
laid. 
Green. 
Green is obtained directly from the oxyd of copper. 
All the oxyds of copper are good : they require little 
flux, which even must not be too fusible : one part or two 
of flux will be sufficient for one of oxyd. This colour 
agrees with all the fluxes, the saline as well as the me- 
tallic ; which tends to vary a little the shades. I have 
already pointed out the method of employing these 
oxyds on raw enamel : were not that method only fol- 
lowed in this case, the oxyd of copper would extend. 
