REFINING THE METAL AND PRODUCING THE BLANKS
Once bullion had been obtained, the next step was to smelt and refine the metal to the desired purity or fineness, and then to make blanks of appropriate weight. The Greeks had been working metal long before the first coins were struck, and at least by the end of the sixth century they were able to extract metals from ores by smelting, to determine the purity of the refined product, and to make alloys or separate alloys into their constituent parts. Extensive remains of ore washeries and furnaces have been found in the Lavrion mining area, and ancient literary sources describe the processes. We know from the precious coins themselves that both Greeks and Romans were able to obtain a high degree of fineness of metal and to adjust the fineness to the desired level. When debasement began to occur, in some regions in the first century B.C., it was well controlled. Ancient craftsmen were also skilled in producing the various alloys used in the token coinages and knew the advantages of adding components such as lead.

To produce blanks of appropriate weight, molten metal was poured into open or two-part closed clay moulds. Experimentation has shown that, with practice, pouring into open moulds or even simply onto a flat surface can be controlled accurately enough to produce blanks within acceptable weight tolerances. Some scholars, however, disagree that such methods would have been used to produce flans for gold or silver coins, and it has been suggested that metal beads or granules of the correct weight were placed in a mould, which was then put into a furnace to melt the granules.

Flans of early coins tended to be spherical or lentoid in shape, and the coins made from them were often quite irregular, preserving the shape of the lentoid flan, flattened by the striking (see nos. 24 and 35). In all periods the weight of the coin mattered more than its shape or size (see nos. 2, 6, 23, 31, 33, and 39). In fifth-century B.C. Sicily, evidence for the use of two-part spherical moulds is preserved on the coins themselves. The flan made from this type of mould would have looked a bit like an unshelled walnut, with a ridge running around its middle produced by metal flowing into the joint where the two halves of the mould joined. When a coin was struck from such a flan, remains of the ridge appear as two small projections on opposite sides of the edge of the coin (see nos. 13 and 15). In Hellenistic times flatter, disk-shaped flans were cast, and this continued to be the normal shape, at least for gold and silver coins. The result was thinner coins of a slightly larger diameter (compare nos. 26 and 47).

It is most probable that silver and gold blanks were normally cast in individual moulds. Yet often the blanks for bronze coins, and sometimes those for silver, were cast en chapelet, that is, using open or closed moulds in which a number of mould hollows were connected by channels. Clear evidence for this method exists; the runners that connected the flans were not always completely removed.

Another method used to produce bronze blanks was to pour out a specific weight of bronze in a solid rod or sheet and then to divide it up into a certain number of blanks, either by sawing or chiselling a segment off the rod, or by cutting a section from the sheet. Unstruck blanks cut from a rod were found in the Athenian Agora, from a building probably used as the mint for early Roman bronze coins. And some later Roman sestertii and some Byzantine coins are slightly square in shape, as if they were squares cut from a sheet and then hammered into a more circular shape.

One further step often required before the flan could be struck would be to remove any impurities that may have risen to the top of the flan owing to oxidation during open casting. The flan could have been washed in some form of organic acid or pickle; in some cases the impurities seem to have been scraped off.

MAKING THE DIES (Continues...)


[LU Home] | [Bearers of Meaning] | [Contents] | [Essays] | [Catalogue] | [Glossary]

All contents copyright (c) 1996.
Lawrence University
All rights reserved.