a 86 Mr. Hatchett’s Experiments on the various Alloys , 
just capable of being rolled and stamped, seems to be equally 
improper for the purpose of coin. For, even supposing that hard 
gold suffered, in every case, less by friction than that which is mo- 
derately ductile, (which is not however the fact,) and allowing 
that standard gold may, by a mixed alloy, be rendered as hard 
as gold reduced by copper to 18 carats, without changing the 
standard proportion of gold, yet it w r ould be very difficult always 
• to make such standard gold of an uniform degree of hardness. 
Moreover, by some experiments which I purposely made at the 
Mint, upon the rolling and stamping of gold of different qualities, 
it evidently appeared, that gold equal in hardness to that of 18 
carats, could not be employed with advantage ; for, the addi- 
tional labour which was required for the rolling and stamping of 
this hard gold, the frequent failure in making the impression, 
and the battering and breaking of the dies, fully proved, that 
the expense and difficulty attending the working of such gold, 
would by no means be compensated by any small degree of 
durability which it might possess over any other. 
The extremes of ductility and of hardness being therefore 
equally objectionable, it follows of course, that gold of moderate 
ductility must be that which is the best adapted for coin ; and, as 
nothing but silver or copper can be employed to alloy gold which 
is intended to be coingd, it may be here observed, that whatever 
might have been the original motive for introducing the present 
standard of 22 carats, yet it appears, from the experiments 
lately described, that this proportion of of the abovemen- 
tioned metals, is (every circumstance being considered) the best, 
or at least as good as any, which could have been chosen. 
There is, however, some difference in the quality of gold, 
when alloyed with the standard proportion of silver, of silver 
