Striking of the coins 

The largest silver and bronze coins were normally struck using iron dies, because bronze dies did not work well for large coins. This was due not only to the greater striking pressures involved but also to the much greater heat transfer to the dies from the large hot flans during striking. To soften the material, flans were heated red hot in a furnace, to a temperature of 400º to 500ºC before striking. Water may have been poured on the obverse dies between strikes to cool them, because the obverses of some large silver coins (such as Syracusan dekadrachms) often show clear evidence of die rust. It has been argued that this rust could also be related to long term die storage, but that does not easily explain why it is so much less commonly seen on reverse dies.

During striking, several reverse dies or punches could be used in succession before the first struck another coin, giving them a chance to cool between strikes. This rotation could not be done with obverse dies, which were fixed in an anvil.

The fact that iron dies were used strongly suggests that dies for large bronze coins tended to suffer from wear and breakdown to a greater degree than dies used on smaller coins. Engraving these large dies required greater effort, and also often involved more artistic work, so mints would naturally have an incentive to do everything possible to extend the life of the dies. 

Flan Preparation 

One way to maximize the life of a die set would be to prepare the flans before striking so that the flan surface presented to the die was as uniform as possible, and free from any imperfections and inclusions. The hypothesis examined here is that flan preparation frequently included machining the surfaces to be struck, both to smooth them and to remove the hard and somewhat irregular “skin” formed during solidification of the flan. Experience with modern castings suggests that the amount of material that would have to be removed to achieve this would be 0.1 to 0.2 mm. For aesthetic reasons it was also desirable to remove remnants of the sprues from the periphery of the flan. Both of these objectives could be achieved by machining the flan on a lathe, and there is significant evidence that this is the process that was actually employed. 

Ptolemaic Bronzes 

The flans of large Egyptian bronze coins, both those struck under the Ptolemies and those later struck under the Romans, have markedly conical edges [p.1, Fig. 1] which cause the diameter of the reverse side to be considerably greater than that of the obverse side.  The shape of these flans suggests that the molds were prepared by making repeated impressions of lathe turned patterns. Bouyon et al. illustrates [pp. 8-9] how such patterns could easily have included a central cavity that would form a raised feature in the mold and a centered dimple in the cast flan. Examination of the struck Ptolemaic coins has suggested that at least in some cases the dimples were cast into the flans. As will be seen, in a lathe turning process, the flan could be efficiently faced (its flat surfaces machined) without a preexisting centered dimple, however smoothing the conical edge would be very much simplified if a well centered dimple existed as a locating feature. After the flans were removed from the molds and (if not individually cast) detached from the sprues, they were prepared for striking by a machining process. The central dimples of large early Ptolemaic bronzes tend to be quite well centered on the flans and the flans tend to be relatively symmetrical. This suggests that they may all have been machined to a uniform size and shape, variations in the struck coins being explained by flan distortion during striking. Some existing specimens also exhibit clear evidence of tool marks on both the coin faces and edges. Thus, there is reason to believe that the edges as well as the flat surfaces of the flans for large early Ptolemaic bronzes were machined prior to striking.

Fig. 4a
Bronze Coin of Ptolemy VI

Later Ptolemaic bronzes are not as well struck as the early issues and there is also evidence that the flan preparation was not as careful. The central dimples are not as uniform or as concentric, making it likely that if the edges had been machined on the early coins, this practice was later abandoned. Fig. 4a shows an example of a bronze coin of Ptolemy VI in which there is a  prominent remnant of the casting runner at the 7:00 position.

Roman Provincial Bronzes 

More than 300 years after issues of large Ptolemaic bronzes ended, civic bronze coins issued under the Roman Empire by many cities in Moesia, Thrace and Asia Minor exhibit central dimples that appear to have been produced by a different process. The dimples on denominations of 4, 4 1/2  and 5 assaria are only approximately centered on many examples, and an adjacent raised bump is often present which appears to be related in some way to the dimple.  Close examination of these coins never reveals any evidence of the kind of metal deformation that would result from impressing the dimples after striking, so the dimple must have been a feature of the flan. This suggests that on these coins, the dimples were produced after casting, during the machining process. These central features are both well illustrated in Figure 5, which shows a 4 1/2 assaria bronze coin of Gordian III struck in Tomis, Moesia Inferior. 

Fig. 5
Roman Provincial Bronze of Tomis 

Here the dimple is clearly off center and is also next to a raised bump, which is much nearer to the actual center of the flan.  

This central raised bump may possibly have been a feature cast into the flan, or it may alternatively have been a feature cut into the die, possibly as a centering device used for turning the die blank and/or to align certain features that were cut into the face of the die. The argument that the bump is a die feature does not easily explain the extreme variability observed in the appearance of these bumps, which are often not visible at all even though the dimple is so far off center that it could hardly have eradicated the bump. Figure 6 illustrates an example of this on a bronze of Philip II struck in Tomis.  

Fig. 6
Off center dimple without adjacent bump 

How could a raised bump cast into the flan surface survive the striking process? Perhaps in the same way that the dimples themselves survived. The explanation may lie in the way that metal in the surface of the flan flowed during and after striking. There is always a certain amount of “elastic rebound” in striking, and the surface of a coin does not necessarily become an exact replica of the surface of the die. Small irregularities  in the surface of a flan tend to persist in rapidly struck coins, more so than in the case of medals which are much more slowly impressed under greater pressure.  The surface instantly deforms under the striking impact but there is not enough time for the metal to flow enough to become completely flat in that area, and it springs back. The greater hardness of the metal at the surface is also a factor. In the case of many coins with such central bumps, which show little wear and yet have flat areas on the busts, a flat or incomplete strike may also have contributed to the survival of the bump. Such bumps, incidentally, always appear to be much higher in the images than their actual elevation above the surface.

Central bumps were not in any way necessary for the machining of the flans, for they are also found on late Roman bronze coins of certain mints, whose flans could not possibly have been machined, given the enormous issues of these coins and the many  irregularities noted in the struck coins.

NEXT PAGE

Website Design By