ON THE METALLURGY OF NICKEL AND COBALT. 129 
These results were obtained with a somewhat friable ore con- 
taining much ferrie oxide, but it was observed that some dense 
varieties yielded scarcely any chloride of nickel by this method. 
When a charcoal chaufer was used for heating the tube, on 
t 1 
oxide, and probably contained ferroso-ferric oxide with the 
whole or greater part of the ferrous oxide replaced by nickel 
oxide, but sufficient for an analysis was not obtained. They’ 
ated, 
formation, as repeated trials to obtain them in larger quantities 
failed. 
These experiments, although failing to separate the nickel 
from the silicate, showed that oxide of nickel could be separated 
from ferric oxide by treatment with gascous a ag acid, 
so the behaviour of the mixed oxides was next ex 
The New Caledonian ore is very infusible, Fon the oe e pro- 
portion of magnesia it contains, and some cheap flux ontaining 
other silicates was necessary to obtain the nickel in a metallic 
or reguline state. This was found in copper ore furnace slag, 
which with its weight of ore yielded a sufficiently fusible 
charge. The alloy of iron, nickel, and copper obtained by fusing 
with carbon ee as a reducing agent was too tough to grind d and 
was therefore era gee of conversion into oxides, but the im gees 
the case with iron pyrites. Thus a aiceuk of 4 00 grains a 
100 grains apa and 800 grains wd ane gave a 
button Sega W756 grains, a and mS ins 
was obtained co! sini, neat all A ~~ but a tthe same 
time a peerente uantity of ferrous ainda, whilst y sat oxide 
with traces of ckel remained, ; 
