specific Gravity, and comparative Wear of Gold. 87 
It then became ductile, but not quite in so great a degree as 
in Exper. xm. 
Experiment xvi. 
Six ounces of fine gold, and six ounces of the finest Swedish 
copper, were melted, and mixed. This metal, being cast in the 
mould of iron, was ductile. 
Experiment xvii. 
The bar was melted again, and, being cast in sand, was then 
found to be very brittle. 
From these experiments it appears, that the varieties of copper 
in commerce, although similar in aspect, and other obvious 
properties, are far from being uniform in quality ; so that many 
of them are by no means sufficiently pure to be employed as an 
alloy for gold. 
Moreover, the different effects produced by the moulds of 
iron and those of sand, are such as fully prove, that copper 
which is not perfectly pure, and which has a tendency to render 
gold brittle, acts more powerfully, in this respect, when the al- 
loyed mass is cast in sand than when it is cast in iron ; and, all 
things being considered, we have reason to conclude, that moulds 
of iron are much to be preferred to those of sand.* 
The ores of antimony and of lead frequently accompany those 
of copper ; and it has already been proved, that of either 
* Bars of alloyed gold (particularly those which are alloyed with copper) are gene- 
rally discoloured on the surface, when cast in moulds of sand ; but not so when cast 
in iron. It may be suspected, that the alloy is superficially oxidized when sand is 
employed, in consequence of the air which is lodged in the interstices, together, per- 
haps, with some degree of moisture. 
