42 
ECONOMICAL MINERALOGY. 
subjected to high heat in contact with oxide of iron, unites with it, and forms a silicate which 
cannot be reduced by carbon. It is, therefore, of great importance to prevent the formation 
of this compound. One of the means by which this end may be attained, is the roasting of 
the ore, an operation which is too often neglected by our manufacturers. By being previously 
roasted, the ore loses its water, if it contains any; the carbonic acid and volatile matters are 
also driven off, and it is brought to that porous state which permits the flame to come into 
contact with each of its particles, and thus the metal is reduced before the silicates can be 
formed. Another advantage gained by the roasting of the ore is, that we avoid in a great 
measure the evolution of the gases in the furnace, which not only obstruct the free passage 
of flame, but have a tendency to cool it. It is, therefore, of the greatest importance that 
particular attention should be paid in all cases to the process of roasting; but especially so, 
when the more compact kinds of magnetic and specular ores are employed. As the operation 
is analogous to the burning of limestone, it may be performed either in the open air, or in 
furnaces constructed for the purpose, as may be found most convenient and economical.* 
In most cases, however, the roasting is not sufficient to effect the complete separation of 
the matters which prevent or retard the reduction of the metal. In order to facilitate the 
operation, substances are added, which are denominated fluxes, the use of which is to bring 
the earthy matters, mixed or combined with the oxide of iron, to a state of thin fusion or 
liquidity, so that the particles of metallic iron, having been formed by the action of the car¬ 
bon, may run together and fall down to the bottom of the furnace. It is evident, therefore, 
that the nature of the flux must differ with that of the ore. When siliceous matter predomi¬ 
nates, limestone is used, so that the silica may form with it a fusible compound, and thus 
liberate the metal. On the contrary, when an ore contains carbonate of lime, this substance 
being also infusible, a siliceous flux is then to be employed. 
From this view of the nature and use of fluxes, it is apparent that they cannot be employed 
in an arbitrary or empirical manner. The particular kind, as well as the proportion, must be 
determined by the chemical composition of the ore; and it is very seldom that two ores so 
closely resemble each other as to allow the use of exactly the same flux. 
With the exception of the lenticular argillaceous oxide of the western part of the State, 
and the magnetic iron from Warrensburgh, all the ores which I have examined are siliceous 
or aluminous, and require for their reduction a flux composed chiefly of carbonate of lime. 
Limestones well adapted to this use, abound in various parts of the State. 
That the proper adjustment of the flux to the nature of the ore is a subject not yet suffi¬ 
ciently attended to by our iron smelters, I think is evident, from an examination of the scoriae 
of many of our furnaces. Small globules or buttons of nearly pure iron, or of the metal 
slightly oxidated, are sometimes found in the hard and glassy compound, and by treating this 
refuse with additional portions of flux, a large proportion of metallic iron is obtained. I was 
informed by the conductor of a furnace in Dutchess county, that a heap of this slag, which 
* For descriptions of processes and apparatus, see Dumas, Traite de Chimie appliques mix Arts, IV. 619. 
