P. Crepusius, Not Otherwise Known
Publius Crepusius must have the rare distinction of being essentially unknown outside the coins he issued as moneyer, yet having a significant amount of research dedicated to these coins. What we know of him and his family is practically limited to a single short statement: he was one third of the triumvir monetalis in 82-81 BC. So little is known about P. Crepusius that Michael Crawford, author of perhaps the most important reference book on Roman Republican coinage, had only this to say of him:
“The moneyer is a P. Crepusius, not otherwise known”
How can so much research be dedicated to an otherwise inconsequential figure of Roman history? As a moneyer, he and two others were responsible for issuing coins for the Republic during this time. A key feature of his coinage is the way he documented the dies used to mint the coins using control marks and, according to Crawford, Crepusius’ coins have the most carefully designed control mark systems of all Republican coinage. Die studies are important to the field of numismatics because they can shed light on the manufacturer of coinage, such as when a coin was minted, how long the issue lasted, how many dies were created, how many coins were struck from each die, and much more. Crepusius’ coinage alone can help us answer many of these questions.
It’s not uncommon for a coin issue to include a symbol or other mark to signify which mint the coin came from, or perhaps which magistrate or other authority was in power at the time. What is relatively unique about Crepusius’ coins is that his coins featured multiple markers covering both the obverse and reverse. The coins could feature as many three distinct markings: one of 23 symbols (e.g. a feather, wheat ear, grapes, etc) and a letter from the Latin alphabet on the obverse, and a sequence of numbers up to 519 on the reverse. No two obverse dies could share the same letter-symbol combination and no two reverse dies could share the same sequence number.
If we can find at least one coin for every symbol-letter combination and one for every sequence number, we can estimate the total number of obverse and reverse dies used for the issue to a very high degree. This is exactly what T. Buttrey did in his 1976 publication on the Crepusius coinage. Previous research had found that each obverse symbol would likely have a die for every letter of the alphabet, so a naïve approximation would expect 23 symbols * 21 letters for a total of 483 obverse dies. However, there are some exceptions to this classification. The first obverse die had no letter or number and is extremely rare. A second group of obverse dies has a letter but no symbol, and numbers approximately 21 dies. The remaining groups have both letters and numbers and Buttrey counts 457 dies across these groups. Thus, a combined total of 479 obverse dies.
At the time of Buttrey’s study, only 321 obverse dies had been accounted for but he argues we can expect 457 dies were created in reality but not all have yet come to light. For the reverses, most of the dies with numbers from 1 to 519 have been found. The combination of the reverse die with the obverse can help us understand in which order the obverse were carved. If we assume the reverses were carved in the order of their number, we can infer which obverses were carved first based on the earliest reverse die they are paired with. If the obverse die with a wheat ear symbol has the lowest reverse die pairing of 32, we know it was likely carved after an obverse that is paired with a lower reverse die
ROMAN REPUBLIC. Pub. Crepusius
Denarius (82 BC)
Dates: 82 BC
Mint: Rome
Obv: Head of Apollo right, scepter over shoulder, E behind, symbol below chin flatly struck
Rev: P CREPVSI in exergue. Horseman right, brandishing spear; control number behind horseman
Dimensions: 18.03mm, 3.87g, 9h
Ref: Crawford 361/1c
This sequence of obverse-reverse paired dies gets us a step closer to understanding the manufacture of coins. It is unlikely all obverse dies were used at once, as this would require as many as 21 anvils to be in use simultaneously, so we can assume obverse dies with a lower reverse number pairing were also struck at an earlier time. As dies wore or broke, they would be swapped out for newer dies, likely with a later reverse number. Statistically examining this relationship led Buttrey to deduce only two anvils were in operation at the same time. Each anvil worked from a group of different obverse symbols, i.e. one anvil went through all 21 obverse dies for the wheat ear symbol while the second anvil was using all 21 obverse dies for the feather symbol.
There is even more that one can glean from this coinage. Buttrey went on to hypothesise about the die engravers of this coinage: if there were two anvils being used, perhaps there was a die engraver for each anvil. He was able to categorise the obverse dies into a “fine” style and “gross” style, purportedly representing two different engravers. Reinforcing the hypothesis around two anvils with one engraver each, he noted that the obverse dies grouped by their symbol alternated chronologically in fine-gross-fine fashion.
Buttrey further estimated how many dies would have had to have been cut each day and thus how quickly they were consumed. His estimates landed at approximately 4 obverse dies and 4 reverse dies cut and consumed every day, meaning the entire issue of 519+479 dies were used over a period of only 125 days.
Research on Crepusius’ coins did not stop at Buttrey’s 1976 article. Others have picked up from where he left off on the question of die longevity, another academic suggests dies missing from the expected series may be explained by shipwrecks, and more research has studied this coinage using computer modelling.
Why Crepusius and other moneyers used control marks on their coins while others did not is not entirely clear and has been debated for decades. Some theories as to why control marks were used, include: detecting fraud, tracking the performance of a mint or its workers, or just for the interest of that particular moneyer. One recent paper discusses these possibilities and more. The authors conclude that control marks were most likely used to prevent fraud and theft by allowing the tracing of coins down to the anvil it was struck at, who was working the anvil, who was weighing the unstruck flans and resulting coins, and who was responsible for the distribution of that allotment of coins on that day to the treasury. Thus a coin found to be under or over weight, or struck using official dies on a counterfeit plated flan, could be isolated from circulation and traced back to those responsible for its manufacture.
Perhaps reinforcing this hypothesis is an interesting observation that can be made about the very first Crepusius die of this type: it had no obverse control marks — neither letter nor symbol — and is known for the incredible style of its portrait. It would make sense that this die had no control marks; if it served as an archetype from which all later dies would be modelled there is no need for tracing its production and circulation. Though we may never know the real reason for control marks on Republican coinage but as long as this question remains unanswered, Crepusius’ coins are likely to be studied for years to come.