The Extraction of Gold. 21 
gold attending the different re-treatments would be greater 
than that incurred in effecting decomposition by means of 
roasting. 
On the other hand, the cost of roasting in bulk, in addition 
to that of grinding, so increased the outlay that it was only 
in a few cases of exceptional richness where the waste could 
have been so operated on profitably, and therefore little ad- 
vantage would have been gained by following up this system. 
But as more than seventy-five per cent. of the gold lost was 
in the sulphides, and the greater part of the remaining gold 
in a form likely to be retained with them, another course 
was open, viz., to separate the sulphides from the compara- 
tively worthless sand, thus reducing the bulk of the material 
to be acted on and the consequent expense of extracting the 
gold. This is the system we have been endeavouring to 
bring to perfection for many years, and it appears to be the 
only course at present known by means of which we can 
hope to reduce the loss of gold within reasonable bounds, 
As regards the larger grains of pyrites, this concentration is 
partially effected on the blanket strakes, and it is the sul- 
phides obtained from them, together with a portion sepa- 
rated from the waste tailings, which have been operated on 
at the Clunes works for several years past at a cost of about 
£1 per ounce of gold extracted, leaving £3 per ounce for 
profit. 
The common reverberatory furnace was first tried for 
roasting, but it was found to require such a large expendi- 
ture of labour and time in turning over the sand, so aseto 
allow of every portion being exposed to the action of the 
heated air for a sufficient length of time to imsure perfect 
oxidation of the sulphides as to render it a very costly pro- 
eess. To remedy this defect a new oxidating furnace was 
designed by Mr. Latta, which has been in use at the Clunes 
works for three years. It is a reverberatory furnace, with an 
inclined bed from thirty to fifty feet long, and from five to 
six feet wide. The bed is set at an angle that will 
allow the undisturbed sand to remain at rest on it, but still 
make it easy to rake down through doorways at the side. 
The sand to be roasted is fed in at the upper end of the bed, 
and is gradually raked down, its place being supplied by 
fresh charges, until it reaches the lower end of the bed com- 
pletely. desulphurised, and is then discharged through a 
narrow opening between the bed and the firebridge. This 
furnace may be supplied with heated air by tubes over the 
e 
a 
