131 
Copper . 
The crucible was then exposed to a heat sufficiently 
strong to melt the glass very fluid? Upon cooling and 
breaking the crucible, I found the silver at the bottom 
perfectly pure, as its oxyd alone could part from its oxy- 
gen without the access of gome inflammable substance. I 
find this process answers equally well for purifying gold,, 
and to me it seems to possess some advantages over all 
the former methods. The materials used are cheap, and 
a large quantity can be refined as soon and as easily as a 
small quantity, by merely altering the capacity of the era* 
©ible you use* 
I tried the same operation on gold and silver in round 
masses, but found it went on very slowly* and what I 
scarcely expected, in the first part of. the process of oxy- 
dating the metals, the remaining metal continued uniform- 
ly impure or nearly so, until the whole was oxy dated. 
I regret that I have been forced to make this matter 
public, before I; could do it in a manner satisfactory to 
myself. I wished to have given the exact proportions of 
alloy, manganese, and glass lo be generally used* and. to 
have ascertained if there is any truth in the old opinion, 
that saltpetre melted with gold destroys a part of it. I sup- 
pose that idea may have arisen from the oxygen given out 
by the nitre in a high heat, oxygenating the copper con- 
tained in the impure gold,, which has been the subject of 
the experiment. 
Since the above was written, I have been informed that 
this matter has actually been published, but know not in 
what work. I hope you will still have the gpodness to, in- 
sert this as an original communication^ as I do not think the 
person who has published it will have the impudence to 
call it his own, and as Mr. Kirwan, and other celebrated 
demists long ago advised me to publish it, I have already 
st&i©d:my reasons for not following-such good advice. . 
