Striking Errors

By Tom Buijtendorp

Ancient coins were produced by hand and as a result offer some striking examples of production errors. Although coin collectors tend to regard those coins as less perfect and less valuable, they tell an interesting story about the minting process, as some examples show. In addition, it explains some peculiarities collectors may notice and easily interpret wrongly, for example as simple coin wear.

Flan Production


Fig 1 Lamination defect on a half shekel minted in Tyre 18 BCE – 69 CE. Obverse laureate head of Melqart right, lion's skin knotted around neck; reverse ΥΡΟΥ ΙΕΡΑΣ ΚΑΙ ΑΣΥΛΟΥ, eagle standing left, right foot on ship's ram, palm frond behind, uncertain date over club left, KP over monogram right, Aramaic letter between legs. Weight 6.535 gram, maximum diameter 18.7 mm. Sold by Forum Ancient Coins nr. JD40460.


Fig. 2. Unstruck blank flan of a bronze diobol found on Cyprus. Paphos mint, active during the time of Cleopatra VII Thea Philopator. Weight 13.907 gram, maximum diameter 24.4 mm. Sold by Forum Ancient Coins nr. GB 17256.


Fig 3. Another unstruck blank flan (see red arrow) among a part of the coin find of Cyprus. Sold by Forum Ancient Coins nr. GB33634.

The minting process started with the raw material. In many cases this was an alloy. For silver, using the wrong alloy could generate quality problems. An example is the lamination effect: a part of the surface is lost due to impurities (fig. 2). After producing the right raw material, producing the flan is the next step. Some flans were cast as excavated examples of stone molds illustrate. In some cases, the circular forms were drilled, leaving a round hole with the trace of the sharp drill in the middle. Rare examples of unstruck flans show the resulting dimple (fig 1). It has been suggested this dimple served a goal as it could improve the flow of metal into the higher part of the obverse portrait. The bronze diobol shown of 13.907 gram and 24.4 mm maximum diameter was found on Cyprus and dates between 50 and 30 BCE. The coin was part of a find of hundreds of normal minted bronze coins from the same mint and period, with a few blank flans (fig 3).


Fig. 4. Two unstruck blank flans of two dichalkon found on Cyprus. Paphos mint, active during the time of Cleopatra VII Thea Philopator. Weight 2.420 gram, maximum diameter 13.9 mm (left) and 1.266 gram, maximum diameter 11.4 mm (right). The line shows how such flans once were connected as part of a casted string of flans before being cut into single flans. Sold by Forum Ancient Coins nr. GB 17258 (right) and GB 17259 (left).


Fig. 5 Group of 5 so called widow mites, lepta minted between 95 and 76 BCE in Jerusalem by Alexander Jannaeus, king of Judea, with the typical flan sprues. Obverse ΒΑΣΙΛΕΩΣ ΑΛΕΞΑΝ∆ΡΟΥ (barbaric and blundered), anchor upside-down in circle; reverse star of eight rays surrounded by diadem, crude barbaric style. Sold by Forum Ancient Coins nr. LT35676.

Forum Ancient Coins sold two smaller blank flanks of bronze diobols from the same find, the one 1.266 gram and 11.4 mm maximum diameter, the other 2.420 gram and 13.9 mm maximum diameter. These to flans illustrate how they were once part of a cast string of circular flans connected by small ‘bridges’ needed to fill the mold at once with the melted bronze. After, the blanks were cut into separate pieces, leaving straight sides at the ‘sprues’ (fig 4). These remains are most clearly visible at the small widows mites minted between 95 and 76 BCE in Jerusalem (fig 5).


Fig 6. Silver antoninianus of Gordian (238—244) minted in Rome with rare remains of the sprues created during the flan casting process. Minted between 1 January 241 and July 243 CE. Obverse IMP GORDIANVS PIVS FEL AVG, radiate, draped and cuirassed bust right, from behind; reverse LAETITIA AVG N, Laetitia standing left, wreath in right, anchor in left. Weight 5.094 gram, maximum diameter 22.9 mm. Sold by Forum Ancient Coins nr. RS47915.


Fig. 7. Sestertius of Gordian III (238-244 CE) with a flan with straight sides, minted in Rome in 240-243 CE. Obverse IMP GORDIANVS PIVS FEL AVG, laureate, draped and cuirassed bust right, from behind; reverse LAETITIA AVG N S C, Laetitia standing facing, head left, wreath in right and anchor in left. Weight 22.614 gram, maximum diameter 29.6 mm. Sold by Forum Ancient Coins nr. RB53605.

 
Fig 8. Antoninianus of Gallienus (260-268) with square flan, probably overstruck of old as. Obverse...GALLIENVS AVG, radiate head right; reverse uncertain legend, female figure standing left, holding cornucopia in left, uncertain object in right. Weight 4.075 gram, maximum diameter 18.6 mm. Sold by Forum Ancient Coins nr BB52458.

A very rare later example is an antoninianus of Gordian minted in Rome in 214-243 CE with the unusual large remains of the sprues created during the casting process (fig 6). The flans for the large bronze sestertii of Gordian were cast in the same way, however with much broader ‘bridges’. As a result, these coins in many cases have very straight sides what gives them a more rectangular then circular appearance (fig 7). A very rare example is an antoninianus of Gallienus (260-268 CE) with square flan, probably an overstruck of a cut fraction of an old coin, probably an as (fig 8).


Fig. 9 Drachm of Antoninus Pius and Marcus Aurelius minted in Cyprus, overstruck on an older coin. Undertype probably like other examples from this mint, a large three century old Ptolemaic coin. Obverse ΑΥΤΚΤΑΙΛΑΡ ΑΝΤΩΝΙΝΟΧ ΧΕΒΕς (or similar), laureate head of Antoninus Pius right; reverse M AV ΡΗΛΙΟ C KAICAP VIOCCEBAC (or similar), draped, bare-head of Marcus Aurelius Caesar right. Weight 21.318 gram, maximum diameter 33.5 mm. Sold by Forum Ancient Coins nr RP17550.

 
Fig 10 Silver overstrike of the Bar Kochba revolt (132-135 CE) with a denarius of Nerva (96-98) as undertype. Obverse Shim'on (in Hebrew), bunch of grapes; remains of legend denarius Nerva at the right side: TR P II COS III; reverse For the freedom of Jerusalem (in Hebrew), Lulav (palm-frond). Weight 3.326 gram, maximum diameter 18.3 mm. Sold by Forum Ancient Coins nr. SH28912.


Fig. 11 Antoninianus of Carausius (286/87 – 296 CE), overstruck on old denarius, probably of Severus Alexander (218-222 CE). Obverse IMP CARAVSIVS P F AVG, radiate, draped and cuirassed bust right; reverse MONETA AVG, Moneta standing left, scales in right, cornucopia in left, C in ex (obscured by overstrike effects). Weight 2.664 gram, maximum diameter 22.8 mm. Sold by Forum Ancient Coins nr. RS38378.


Fig. 12. Denarius of Septimius Severus (193-211 CE) with flan defect (hole), maybe as the result of being overstruck on an old holed coin. Minted in Rome in 198-200 CE. Obverse L SEPT SEV PERT AVG IMP X, laureate head right; reverse MARTI PACIFERO, Mars standing half left, helmeted, nude, foot on cuirass, branch in right, inverted spear in left. Weight 2.837 gram, maximum diameter 18.6 mm. Sold by Forum Ancient Coins nr. RS41882.

Old coins were often used as flan for new coinage. In some cases very old coins were used. An interesting example is a drachm of Antoninus Pius and Marcus Aurelius minted in Cyprus, overstruck on a probably about three century old Ptolemaic coin (fig 9). During the special minting circumstances of revolts, many old coins have been reused. An example is the silver Hebrew coins minted in 132-135 CE during the Bar Kochba revolt against the Romans. On the example shown, the obverse at the right bottom clearly shows the remains of the titles of a Roman denarius: TR P II COS III, in this case pointing to a denarius of Nerva (96-98 CE) as undertype (fig 10). Such overstrikes can offer interesting insights in the chronology of coins. An example is an overstrike of an antoninianus of Carausius (286/87 – 296 CE) on an old denarius, probably of Severus Alexander (218 – 222 CE). This indicates the value of the antoninianus (‘double denarius’) was worth at least one (possibly worn) old denarius. The reuse of old coins could generate flan defects like holes (fig 12).

Striking


Fig 13. Denarius minted in 46 BCE in Rome showing the implements of the Roman mint. Obverse MONETA, head of Juno Moneta right, hair in knot, earring and necklace, one lock of hair falls down on her neck; reverse T•CARISIVS above minting implements, die as wreathed cap of Vulcan above anvil, between tongue and hammer; all in wreath; Weight 3.56 gram, maximum diameter 20.6 mm. Sold by Forum Ancient Coins nr. SH09049.


Fig 14. Aureus of Severus Alexander (222-235 CE) in mint fresh state, minted in Rome from fresh dies in 230 CE. Obverse IMP SEV ALEXAND AVG, laureate bust right with drapery on left shoulder; reverse P M TR P VIII COS III P P, emperor advancing right, spear in right, trophy over shoulder in left. Weight 5.922 gram, maximum diameter 20.6 mm. Sold by Forum Ancient Coins nr. SH21939.

Assuming the right alloy was used and a good solid flan produced, the next moment in the minting process to generate potential minting errors was striking the coins. A denarius minted in 46 BCE in Rome on the reverse very nicely shows the implements used in the mint of Rome (fig 13). In the middle there is the anvil. One die would be mounted on the anvil. The other (shown above the anvil) would be mounted on a hand-held punch. The heated flan would be transferred with the tongue shown at the left side of the anvil towards the die mounted in the anvil. Then the hammer shown at the right side would be used to hit the hand-held punch to strike the flan. The minted coin would be removed to continue with the next flan. When successful, the end product would be a well centered beautiful freshly minted coin (fig 14).


Fig 15. Denarius of Julius Caesar minted Jan-February 44 CE in Rome by moneyer P. Sepullius Macer, with an unstruck area with raised pellets. Obverse CAESAR DICT PERPETVO, wreathed head of Caesar right; reverse P SEPVLLIVS MACER, Venus standing left, Victory in right, long scepter vertical in left, shield at feet right. Weight 3.667 gram, maximum diameter 19.4 mm. Sold by Forum Ancient Coins nr. SH34871.


Fig 16. Denarius of Julius Caesar minted Jan-February 44 CE in Rome by moneyer P. Sepullius Macer, with the reverse off center. Obverse CAESAR DICT PERPETVO, veiled and wreathed head of Caesar right; reverse P SEPVLLIVS MACER, Venus standing left, Victory in extended right, long scepter vertical in left, shield at feet right. Weight 3.25 gram, maximum diameter 19.6 mm. Sold by Forum Ancient Coins nr. SH06891. At the bottom a reconstruction of the size of the reverse die.

However, many errors occurred while striking the coins. A weak strike, for example, could result in an unstruck area. A nice example is a rare denarius of Julius Caesar, the first issue with his portrait and minted a few weeks before he got murdered, according to some the portrait on his coins being one of the reasons (‘the coin that killed Caesar’). They were minted in great hurry for his planned Parthian war, resulting in an above average number of minting errors. An example is this one minted by moneyer P. Sepullius Macer with an unstruck area with raised pellets, an error typical for Republican flans. The die axis of 255o explains the different position of the unstruck part on the obverse and reverse (fig 15). Another denarius of Caesar minted in Rome by the same moneyer and in the same months is off-center on the reverse, meaning the hand-held punch did not hit the flan in the middle. This helps to reconstruct the original size of the reverse die (fig 16). The small sharp marks on both coins, by the way, are banking marks, no striking errors.


Fig 17. Denarius of Julius Caesar minted in 49 BCE, and double struck with part of the name Caesar repeated at the left side of the obverse, along the back of the animal. Obverse CAESAR below elephant right trampling on snake; reverse emblems of the pontificate – culullus (cup) or simpulum (ladle), sprinkler, axe and apex (priest's hat); Weight 3.72 gram, maximum diameter 19.4 mm. Sold by Forum Ancient Coins nr. SH28927.

 
Fig 18. Bronze AE3 of Valentinian I (364-375 CE) with extreme double strike, minted in Rome in 364 – 367 CE. Obverse D N VALENTINI-ANVS P F AVG, pearl-diademed, draped and cuirassed bust right; reverse RESTITV-TOR REIP, emperor standing facing, head right, labarum (chi-rho Christogram standard) in right, Victory on globe presenting wreath in left, uncertain mintmark. Weight 2.497 gram, maximum diameter 21.2 mm. Sold by Forum Ancient Coins nr. BB54160.

Another series of errors has to do with the flan being hit more than once, with several different possible reasons. First there is the ‘double strike’, hitting the flan a second time, for example because the first hit was weak and the minter would like to avoid weak or unstruck parts. When the flan moved a little after the first strike, the second strike generated a shifted pattern. An example is a denarius of Caesar with an elephant on the obverse, the first type minted in his name, produced in a mobile workshop travelling with Caesar. As a result of the double strike, parts of the name Caesar appear twice on the obverse (fig 17). In this case, the flan only turned a little. In other cases the flan also shifted, resulting in a very strange pattern (fig 18).


Fig 19. Flip-strike error of a bronze AE3 of Licinius (308 – 324 CE), minted in the second officina of Nicomedia (Turkey) in 321-324 CE. Obverse IMP C VAL LICIN LIC[INIVS P F AVG], overstrike with reverse die (normal obverse - radiate, draped and cuirassed bust right); reverse IOVI CONSERVATORI, Jupiter standing left, nude, Victory on globe in right, long scepter vertical in left, flanked by eagle with wreath and captive at feet, X/II Γ right, SMNB in ex. Weight 3.580 gram, maximum diameter 21.0 mm. Sold by Forum Ancient Coins nr. ER60013.

 
Fig 20. A brockage of Domitian, denarius minted 81-82 CE in Rome. Obverse IMP CAES DOMITIANVS AVG P M, laureate head right; reverse incuse obverse type. Weight 3.046g, maximum diameter 17.7 mm. Ex John Quincy Adams Collection, 6th President of the United States, and His Descendants, ex Massachusetts Historical Society Collection, ex Stack’s Sale, 5-6 March 1971, lot 791. This coin later sold by Forum Ancient Coins nr. JA47618.


Fig 21. Special brockage of Tetricus I (270-273 CE), antoninianus minted in 271-274 CE in Gaul. Obverse IMP C TETRICVS P F AVG, radiate, draped and cuirassed bust right; reverse incuse of obverse. Weight 1.772 gram, maximum diameter 17.9 mm. Sold by Forum Ancient Coins nr. BB22036.


Fig 22. Reverse brockage, bronze AE3 minted in 330-333 CE in Thessalonica (Greece). Obverse Victory standing left on prow holding spear and shield, SMTSE in exergue, all incuse; reverse Victory standing left, right foot on prow, scepter in right, resting left on grounded shield, SMTSE in exergue. Weight 1.05 gram, maximum diameter 17.0 mm. Sold by Forum Ancient Coins nr. RL08146.


Fig 23 Solidus of Constans (337 – 350 CE) mule with the reverse of a solidus of Constantius II (337-360 CE), minted in 340-350 CE in Siscia (Croatia). Obverse CONSTANS AVGVSTVS, pearl-diademed, draped and cuirassed bust right, all within wreath; reverse VICTORIAE DD NN AVGG, two victories facing one another, holding wreath inscribed VOT XX MVLT XXX, TES in ex, all in wreath. Weight 4.238 gram, maximum diameter 22.0 mm. Sold by Forum Ancient Coins nr. SH24842.

In very rare occasions, during the second strike the flan not only shifted, but also flipped over. As a result, now both the obverse and reverse are have been hit by both the obverse and reverse die. This may produce a double image on both sides, a flip-strike error. In this example, part of the legend of the obverse is combined with the image of the reverse (fig 19). Another special case, less rare, is the brockage. Now the minted coin sticks to the hand-held punch. As a result the punch no longer provides the negative reverse die imprint, as normal, but hits the new flan with the positive relief of the obverse of the coin attached to the punch. Instead of hammering the reverse of the flan with the reverse die, it now leaves a negative impression of the obverse, as an interesting example of a denarius of Domitian shows (fig 20). The die-axis is in this case 0o. This means the next stroke was at the same angle as the first one. This interesting coin was part of the collection of the American president Adams (1797-1801) and his heirs.

In some cases, the old coin remained stuck to the hand-held anvil during a row of strokes. As a result, the obverse of the still hot coin flattened and left less deep incursions. An interesting example is a brockage of Tetricus I minted in 271-274 CE (fig 21). It appears the reverse die was much smaller than the one reconstructed in fig 16 as suggested by the imprint of what seems to be the border of the reverse die. While normally brockages are produced by a earlier coin stuck to the hand-held punch, some brockages are the result of a coin stuck to the anvil. These ‘reverse brockage’ coins are much rarer as this was much easier to detect. The coin shown in fig. 22 normally would depict a helmeted bust on the obverse were now the negative incuse of the winged Victoria of the reverse is clearly visible, with below the abbreviation of the minting place (SMTSE) negative as well (fig 22). A different error occurs when the wrong dies are used for striking, a mule. An example is a solidus of Constans (337 – 350 CE). While the obverse shows the portrait of Constans and his titles, the reverse is from a coin of Constantius II (337-360 CE). The coin was minted in 340-350 CE in Siscia. The error was understandable as the reverse of Constant looked quite the same, however with in the wreath the text VOT X MVLT XX and not the higher number of his co-regent Constantius II with VOT XX MVLT XXX as shown on this coin (fig 23).

Dies


Fig 24. Denarius of Augustus (27 BCE – 14 CE) minted in Lyon from 2 BCE onwards. Obverse CAESAR AVGVSTVS DIVI F PATER PATRIAE, laureate head right, die crack on neck; reverse C L CAESARES AVGVSTI F COS DESIG PRINC IVVENT, Caius and Lucius Caesars togate stand facing, each resting hand on a round shield with spear behind, X in center, above center on left a lituus r. and on right a simpulum l.; Weight 3.636 gram, maximum diameter 18.5 mm. Sold by Forum Ancient Coins nr. SH16756.


Fig 25. Example of a larger die damage - edge of the reverse die broken, a bronze AE4 minted for Jovian (363 – 364 CE) in Siscia. Obverse D N IOVIA-NVS P F AVG, pearl-diademed, draped and cuirassed bust right; reverse VOT V MVLT X within wreath, BSISC in ex. Weight 3.216 gram, maximum diameter 20.7 mm. Sold by Forum Ancient Coins nr. RL25293.


Fig 26 Example of the effect of a partly filled die, leaving some letters blank, bronze AE3 minted for Constantius II (337 – 361 CE) in328-329 in Siscia. Obverse FL IVL CONSTANTIVS NOB C, laureate, draped and cuirassed bust left; reverse PROVIDENTIAE CAESS, campgate with two turrets, decorated top and bottom rows, star above, SIS and double crescent in ex. Weight 2.776 gram, maximum diameter 18.6 mm. Sold by Forum Ancient Coins nr. RL38231.

A next group of minting errors has to do with the quality of the dies used. Based amongst others on some medieval statistics and experiments, it has been estimated that the die mounted in the anvil could at least produce 10.000 coins. The reverse die, mounted on the hand-held punch, received more stress and as a result lasted on average a shorter period. Analyses of die-links indicate a ratio of about 3 reverse dies used per single obverse die. During their use, dies got worn and left less sharp reliefs. This can be best noticed on coins with no or little circulation wear. Fresh dies can leaf very sharp reliefs (fig 14). Worn dies leaf a much less sharp imprint what wrongly may suggest the coin is worn by long circulation (fig 24). In the example of Augustus shown, the obverse shows a very sharp portrait and lettering, indicating the coin has hardly circulated. Still, the reverse looks quite worn. In this case, the old reverse die also was a little rusted what left the imprints of small rusted particles on a part of the reverse. In this case, the obverse die was cracked what left in high relief a line on the bust of Augustus. In other cases the damage of the die was not just a crack but a large hole, leaving a strange higher surface on the minted coin (fig 25). Another quality problem was the old die getting filled at some places. When for examples letters got filled, they could generate blank spaces in the legend. Although the example shown was a bold strike with strong relief, some letters are missing in the obverse legend: FL IVL CONSTANT[IV]S [N]OB C (fig 26). Due to the deterioration of the obverse dies and the more frequent use of reverse dies, detailed analyses of dies used for coins of a certain series, can reveal the chronology of the dies.


Fig 27. Denarius of Vitellius (69 CE) with an unrecorded engraving legend error at the end of the reverse legend, the last two letters being interchanged, minted in Rome. Obverse A VITELLIVS GERMANICVS IMP, bare head right; reverse XV VIR SACR FCA (correct: FAC), tripob-lebes with dolphin lying right on top and raven below. Weight 2.916 gram, maximum diameter 18.8 mm. Offered by Forum Ancient Coins nr. SH53562.

Finally, some errors already started with the engraving of the dies. Contrary to manual errors during the minting process, these errors are present on all coins minted by a certain die. Like the striking of the coins, the engraving was manual work. And the die engraver needed to work in negative, making it more difficult to notice errors. A quite common error occurs in the legend were letters are omitted or exchanged. An unrecorded example is the reverse of a denarius of Vitellius with in the reverse legend the past two letters being interchanged, ending with FCA instead of the normal FAC (fig 27).

As these examples show, the wrongly minted coins tell an interesting story about the way ancient coins were minted. As a quality mint would prefer to melt such coins instead of circulating them, the real number of minting errors must have been much larger than the survivors suggest. So what does remain requires full scientific attention and publication, for example in NumisWiki.