One ancient process for extracting the silver from lead was cupellation. Lead was melted in a bone ash 'test' or 'cupel' and air blown across the surface. This oxidised the lead to litharge, leaving a button of silver. Anciently, the litharge was discarded, but more usually it was re-smelted to lead. "Pigs" of Roman lead have been found marked EX ARG (argentum is Latin for Silver). This presumably indicated that the lead had already been de-silvered. This process was viable economically if the lead contained 8 troy ounces of silver per ton of lead (178 ppm).
In the 18th century, the process was carried on using a kind of reverberatory furnace, but differing for the usual kind in that air was blown over the surface of the molten lead from bellows or (in the 19th century) blowing cylinders.
The equipment consisted of a row of about 8-9 iron pots, which could be heated from below.
Lead was charged to the central pot and melted. This was then allowed to cool, as the lead solidified, it was skimmed off and moved to the next pot in one direction, and the remaining metal was then transferred to the next pot in the opposite direction. The process was repeated in the pots successively, and resulted in lead accumulating in the pot at one end and silver in that at the other. The process was viable down to 2-3 troy ounces per ton (45-67 ppm.
The Parkes process, patented in 1850 uses zinc to form a material which the silver enters. This floats on the lead and can be skimmed off, enabling the silver to be recovered.
The product of the blast furnace is pig iron, which contains 4-5% carbon and usually some silicon. To produce a forgeable product a further process was needed, usually described as 'fining, rather than refining. From the 16th century, this was undertaken in a finery forge. At the end of the 18th century, this began to be replaced by puddling (in a puddling furnace), which was in turn gradually superseded by the production of mild steel by the Bessemer process.
The term refining is used in a narrower context. Henry Cort's original puddling process only worked where the raw material was white cast iron, rather than the grey pig iron that was the usual raw material for finery forges. To use grey pig iron, a preliminary refining process was necessary to remove silicon. The pig iron was melted in a running out furnace and then run out into a trough. This process oxidised the silicon to form a slag, which floated on the iron and was removed by lowering a dam at the end of the trough. The product of this process was a white metal, known as finers metal and refined iron.