Iron. 
453 
with carbonic acid. The salts of iron, are not used as ores. The 
combination of iron with phosphorus or its acid, I consider as too 
uncertain to be noticed. To reduce these ores to iron, the sul- 
phur or arsenic, the oxygen, the carbonic acid, must be driven off. 
This is done by means of heat and charcoal, whether of wood as 
in this country, or of fossil coals in England. 
When ores are roasted previous to smelting, with a heat not 
exceeding a red heat, they are 
1st. Rendered more friable: more easily broken. 
2d. Sulphur and arsenic if combined in them, are driven off, 
partly in the form of sulphur and white arsenic, and toward the 
close of the operation the sulphur is acidified by the air, and fiies 
off in the form of a suffocating sulphureous acid gas. 
3d. By heat also, the carbonic acid gas is partly driven off 
during roasting. Hence by roasting with coal dust, a part of the 
operations of the smelting furnace are forestalled by means of a 
cheap fuel : and where the ores contain sulphur this is more ef- 
fectually driven off by roasting in the open air, than it could be in 
the furnace, where its escape would be perpetually intercepted by 
the great body of ore, fuel, and flux. 
I believe itto be true, that the last portions of sulphur must be 
driven off by being acidified in contact with atmospheric air : but 
where the ore is, as it generally is, merely an oxyd, or an oxydule, 
I have no doubt but the oxygen is encreased instead of being di- 
minished, by roasting, unless plenty of charcoal dust be used ; 
which by yielding carbon, converts the oxygen of the ore into car- 
bonic acid. That iron in a red heat will greedily imbibe oxygen 
from the atmosphere, no one who has seen finery cinder, or the 
scales of a common blacksmith’s shop, can doubt (I think) for a 
moment. Nor can the iron be oxygenated to the utmost, without 
ignition, as any one may see by exposing yellow ochre to heat till 
it becomes red ochre, which is a peroxyd. 
Hence, where the iron is sulphurated, my correspondent’s re- 
marks apply fully? and they are important. Where the ore is a 
mere oxyd, not perfectly saturated with oxygen, as the ease gene- 
rally is, I think the roasting with plenty of charcoal dust, and stop- 
ping the operation rather before the charcoal dust is consumed, is 
the proper method to be adopted ; for if the oxygen of the ore be 
in great measure driven oft' by roasting, it will certainly be re- 
stored, if the ore be exposed to the action of the air in a red heat 
after the charcoal is consumed. 
