In ancient times artisans made everything, including coins, using simple tools, and results depended very much on skill. Ancient coins come in many qualities, ranging from the hastily struck "widow's mites" of Palestine to the superbly artistic silver coins of Greek Sicily.

Dies were made of hard bronze or iron. Bronze dies were easier to engrave and did not rust, but wore out faster. The Greeks used iron dies for their largest coins, many of which show traces of rust. The obverse die was mounted on the anvil and the reverse die, or punch, was struck to make the impression. On the left is a reverse die for a tetradrachm of Athens. In the center are Roman die design improvements: first the socketed reverse die, which aligned with the obverse die, next the hinged die set which eliminated striking irregularities. On the right is a Roman denarius die.

How dies were sunk is still being debated. There is not much evidence, but we have the coins, and some idea of how many were struck. A coining team could produce up to 20,000 strikes, wearing out a set of dies, in one day. During the 2nd century about 17 million Roman denarii were issued each year, so a year's issue required up to 1000  dies; the bronze issues of the Constantinian period must have required many thousands of dies.

Roman coins all bear standardized portraits, and it is difficult to visualize how portraits could be so uniform if  dies were individually engraved by many artisans. Two replication processes were feasible, though we have no proof that the Romans used them: casting dies and hubbing (impressing a design into soft metal). The Greeks used bronze or lead die patterns to cast bronze dies, and at least two pattern specimens have survived.

If  a replication process were used for the central design, the process would have been similar to that described by the medallist Cellini in the 1500s:

1) The central design is sunk in the die blank, either by casting or hubbing.
2) The legend is added by letter punches which are individually struck, perhaps using a fixture to align them.
3) The border is similarly added by punching dots around the legend.
4) The completed design is hand engraved for touchup where needed, then polished.

Alternatively, the die-making workshop may have employed apprentices working under master engravers who cut the portrait. This would explain why engraving of  the portrait is frequently much better than the surrounding work. Standardization of portraits may be explained by the master engravers copying official portrait medals.

Blanks or flans were prepared by cutting from bars or strip and hammering into shape, or by casting. Bronze flans could not be hammered, and were either cast or struck as cut. On the left is a piece of a mold for casting bronze flans. The flans were connected edge to edge, or by runners. We find evidence of this on many coins from which the runners were not completely removed. Coins were often struck by moving strips of cast flans through the dies and later cutting them apart. Before striking, the flans were placed in an oven and heated.

With flans hot and dies ready, the striker (usually with an assistant handling the hot flans) placed the flan between the dies with the tongs, set the reverse die or punch on the flan, and struck the punch with a coining hammer. This sounds like a slow process, but strikers developed a rhythm and a two-man team could strike thirty coins per minute or more. From mint marks on Roman coins we know that a mint had four to twelve workshops, each with least one team of coiners making up to 20,000 coins per day. No doubt it was hot work!

All images © Classical Coins

Privacy and Security

About the Images

Contacting Classical Coins