Three Coins of Gordian III & Tranquillina

Centration Dimples & other features

The three subjects of this page are all 5 assaria coins of Gordian III struck at the Roman Provincial (Greek Imperial) mint at Anchialus, Thrace. Made Emperor in 238 AD at age 13, Gordian was a puppet under the control of his adult 'advisors'. In 241, he married Tranquillina, daughter of his Praetorian Prefect Timisitheus. Long time visitors of this site will recall we previously showed a twin portrait coin of the teen couple. The purpose of this page is to examine these three coins and see what we can learn about a technical feature of their production.

Anchialus (Pomorie, Bulgaria) was one of several cities in the Black Sea region that issued these twin portrait coins. Since the wedding (241) and death (244) of Gordian were separated by a very short time there should be little difference in the dates of the three. Activity in the area was at a peak near the date of the wedding when Roman legions were quelling barbarian raids from the North. The last year of the reign saw attention (and soldiers with their spending power) redirected to troubles with the Sassanians in the East so the greatest demand for local coinage would have been 241-242 AD.

All three coins are very similar in style. In fact, close inspection will reveal that the same obverse die produced both the center and right coins. The reverses are all different in subject but all could certainly have been cut by the same hand. Coins of the region regularly show a pit (the centration dimple) in the center of each side. The purpose or cause of these marks have been a matter of controversy which this page is not going to put to rest. We will, however, examine the marks on these three coins and see some conflicting situations that make it difficult to stick to one theory.

AVT K M ANT GOPDIANOC AVG CAB / TPANKVLLI / NA
OVLPIANWN AGXIALEWN
Tyche with rudder & cornucopia
14.4g
Demeter (?) with grain & staff
10.0g
Victory with wreath & palm
11.4g

Of the three coins, the most telling is the Tyche (left) struck considerably off center on both sides. This is one case where having a fault is an advantage. Beyond the border of dots and the edge of the die is an area where the flan was not altered by the striking. Here we see several rows of concentric ridges which seem to center on the dimple (here falling just below Gordian's eye). This suggests that a cutting or grinding tool was used to flatten the flan. The two small round marks on the two sides of the flan do not line up with one another suggesting they were not produced at once by a lathe turning the edges. Perhaps they were made by a tool like a forstner bit with a small center spur. This would have been used to smooth the flan surfaces one side at a time. The fine ridges suggest the possibility that the actual cutting was done by a course grinding stone or a finely cut rotary file.

Two Very Different Centration Dimples
Unrotated dimple
Arrow indicates raised layout dot - see below.
Rotated dimple
Arrow indicates concentric ridges.
The Demeter coin (center) is rather well centered on an oval flan and has very prominent centration marks. The one on the obverse removes Tranquillina's nose and mouth (see enlargement). Again, the two sides do not line up. The information added by this coin is that the marks here are not rounded but slightly oval in shape. This suggests that the tool did not rotate inside the dimple. This observation is exactly the opposite of that made when we examined the Marcianopolis coin of Philip II as Augustus not long ago (see enlargement). Note a few of the concentric ridges survived striking and show in the field of the coin (right arrow). Our theory needs to allow for some coins to rotate on the point while others remain stationary.

The Victory coin (right), as mentioned above, uses the same obverse die as the Demeter. It, however, shows only a faint trace of an obverse pit located on the bridge of Gordian's nose. This, at least, proves that the marks were not features cut into the dies! We always must allow for the possibility that a lack of the obverse pit is the result of its being erased by the striking process. The obverse also shows a small raised dot between the portraits. This is a center point for a compass used to lay out the beaded border on the die. These were present on most coin dies but would be erased by the cutting of a central portrait or figure. This mark is in the same position on our die linked pair (Demeter and Victory) as we would expect (see left arrow in above enlargement). The equivalent mark on the Tyche obverse is slightly lower and closer to Tranquillina. The presence of these die marks is in no way related to the centration dimples. The reverse dimple is quite large and rounded suggesting a tool rotated inside it. This pit is similar to the more clear example in our enlargement but is slightly filled by patina. Differences between the dimples on the two sides of many coins suggest they were produced independently.

Some students suggest that the dimples are actually marks of tongs used to handle the heated blanks and the marks were made intentionally as a test of the flan being soft enough for striking. I see three problems with this theory. First it would seem that a flan made soft enough to be marked by the tongs would have been soft enough to have erased the marks when subjected to the much greater force of striking. Second, it seems odd so many coins are seen with a very distinct mark on one side and little or none on the other. If the purpose was to determine softness, it would seem the flan would have been equally soft on both sides. Finally, this theory fails to address either the concentric ridges or the dimples that show rotation. Misaligned tongs would hardly rotate so smoothly. Of our coins, only the Demeter would seem to support the tongs theory.

What is the answer? I simply do not know. If I had to select a theory based on what I have observed on the coins I have handled, I would tend toward the marks being made by a tiny spike driven into one face of a coin allowing the surface to be smoothed by some sort of cutter. If the spike were driven in securely and held firmly, the dimple would retain the shape of the tool. If the grip of the spike was weaker than the force of the grinding tool, the coin could rotate on the point producing a perfectly round dimple. I do not claim this answer is without problems but, for the present, it seems the best of the choices.

Why were these marks made? This question bothers me greatly. It would seem that the same minds that thought it desirable to smooth the flan surfaces at all would have been terribly disturbed by the pits that remained. Weight control seems an unlikely reason since our three coins vary greatly from each other. This is simply a subject that requires more study.

There is more to these coins than the pits!

The long section above completely ignores that fact that these coins have much more to offer than their pits! In fact there are several other points of interest that will speak to different groups of people. Those interested in epigraphy may note the ligate NK, NT, VG and two different ligatures for Omega-Nu. On the reverse OULPIANWN (referring to the city charter granted under Trajan whose family name was Ulpius) ends WN while the city name ends WN. Unfortunately the off center strikes of the other coins makes this only seen on the Demeter. Persons trying to assemble a set of portraits of Roman rulers will find Tranquillina rare and costly in the regular Rome mint series but her coins are common among the Greek Imperials. Those attempting to assemble a series of reverse deities and personifications will find the range available quite impressive and well executed. Some are not easily identified. (Am I correct on this Demeter??) Others will try to assemble a series of the various vis a vis twin portrait combinations. These coins have much to offer the collector including, certainly, the mystery of the pits.


Back to Main page


(c) 1999 Doug Smith