> ee eS eee ee 
i 
1879.] on Coining Silver into Rupees. 59 
It will be seen that two meltings have been made for each proportion 
of copper. If C represent roughly one hundred tolas of copper, and 
we group these determinations in proportion to the quantity of copper, we 
shall have : 
GAIN IN 
Copper. Pots — Theory. Coins — Theory. Coins — Pots. 
oc — 0-075 + 0079 + 0°545 + 0°109 + 0°620 + 0-094 
2C + 0:050 + 0:073 + 0805 + 0108 + 0°755 + 0°091 
40 + 0°335 + 0:065 + 0°775 + 0°107 + 0440 + 0:088 
6C + 0°805 + 0-071 + 1180 + 0-090 + 0°375 + 0:065 
= P10 + 0°635 + 0:078 + 1:360 + 0:116 + 0°725 + 0:092 
It is evident that the refining of the Pots from the Theory is nearly 
proportional to C, and that the refining of the Coins above the Pots or the 
ingots is approximately constant, though irregular, as indeed might have 
been anticipated. 
If now we assume a + ma to be the refinage in melting, when m is the 
coefficient of C above, and y to be the refinage in passing from the Ingots 
to Coin; we shall have 
5 values of a + mw of nearly equal weight 
5 y of sufficiently equal weight 
and 5 values of a + mx + y, which being the sums of the others we may 
neglect. 
From these equations we get the following values : 
a = — 0:085 + 0:088 
2 = + 0109 + 0018 
y = + 0583 + 0-051 
The large probable error of « compared with its value renders it very 
doubtful if there is any real change in scissel melting. What there is 
seems to be towards loss of jineness and it is quite certain that silver evapo- 
rates ; for, in the Regenerators and flues of the Gas Furnaces (now disused) 
the soot was found to contain silver. 
The other quantities are clearly marked, and the small probable error 
of « shows that the hypothesis that free copper only burns is probably true. 
Had a@ been assumed = 0, the value of # would have been 0:095. 
a a oe 
