How Much Silver Was in a Roman Denarius? Debasement Tables From Augustus to Diocletian
An Augustan denarius held about 3.8 grams of silver. An antoninianus of Claudius II, struck about 290 years later and tariffed at two denarii, held about six hundredths of a gram. The face value stayed put. Almost everything else went.
Nobody sat down and decided to wreck the Roman currency. It happened in small steps, most of them defensible in their year, and then all at once in the 260s. The numbers below track how.
A note on the figures. Fineness estimates come from metallurgical analysis of surviving coins, and different studies give different results, sometimes by ten percentage points for the same reign. Weights also varied coin to coin. Treat every figure here as a typical value with a margin around it. The shape of the decline is not in dispute.
Silver Content, Reign by Reign
| Date | Emperor | Coin | Typical weight (g) | Silver fineness | Silver per coin (g) | vs Augustan denarius |
|---|---|---|---|---|---|---|
| c. 20 BC | Augustus | Denarius | 3.9 | 95-98% | 3.8 | 100% |
| 64 AD | Nero | Denarius | 3.4 | c. 93% | 3.2 | 84% |
| 107 | Trajan | Denarius | 3.2 | c. 85-90% | 2.8 | 74% |
| 148 | Antoninus Pius | Denarius | 3.2 | c. 83% | 2.7 | 70% |
| c. 190 | Commodus | Denarius | 3.2 | c. 71% | 2.3 | 60% |
| c. 200 | Septimius Severus | Denarius | 3.2 | c. 50-57% | 1.7 | 45% |
| 215 | Caracalla | Antoninianus | 5.1 | c. 50% | 2.6 | 67% |
| c. 250 | Trajan Decius | Antoninianus | c. 4.0 | c. 41% | 1.6 | 43% |
| 260 | Valerian and Gallienus | Antoninianus | c. 3.0 | c. 15% | 0.45 | 12% |
| c. 268 | Gallienus | Antoninianus | c. 2.8 | c. 5% | 0.14 | 4% |
| c. 270 | Claudius II | Antoninianus | c. 2.8 | c. 2% | 0.06 | 1.5% |
| 274 | Aurelian | Reformed radiate | c. 3.9 | c. 5% | 0.2 | 5% |
| 294 | Diocletian | Argenteus | c. 3.4 | c. 90-95% | 3.2 | 84% |
The first two centuries look almost gentle. Nero’s reform in 64 cut the silver by about a sixth. It then took until Commodus, some 125 years later, to lose roughly another quarter of the original. Then Severus took the coin below half.
The Antoninianus Trick
The raw silver column flatters Caracalla. His new coin, the antoninianus, looks like a recovery: 2.6 grams against the 1.7 of his father’s denarius. The catch is its tariff. It’s generally thought to have circulated at two denarii while containing only about one and a half denarii’s worth of silver.
So the useful measure is silver per denarius of face value.
| Date | Coin | Silver per coin (g) | Face value (denarii) | Silver per denarius of face value (g) | vs Augustus |
|---|---|---|---|---|---|
| c. 20 BC | Denarius | 3.8 | 1 | 3.8 | 100% |
| c. 200 | Denarius | 1.7 | 1 | 1.7 | 45% |
| 215 | Antoninianus | 2.6 | 2 | 1.3 | 34% |
| c. 250 | Antoninianus | 1.6 | 2 | 0.82 | 22% |
| 260 | Antoninianus | 0.45 | 2 | 0.22 | 6% |
| c. 268 | Antoninianus | 0.14 | 2 | 0.07 | 1.8% |
| c. 270 | Antoninianus | 0.06 | 2 | 0.03 | 0.7% |
The two-denarii tariff is a modern reconstruction (nothing on the coin says so) and some numismatists question it. If it holds, Caracalla’s introduction was a quarter-cut in silver disguised as a bigger coin.
Each Halving Came Faster
Measured by silver per denarius of face value, the pace of decline is the real story.
| Silver remaining | Reached by | Time taken from previous step |
|---|---|---|
| 100% | Augustus, c. 20 BC | Starting point |
| Below 50% | Septimius Severus, c. 200 | About 220 years |
| Below 25% | Trajan Decius, c. 250 | About 50 years |
| Below 6% (two more halvings) | Valerian, 260 | About 10 years |
| Below 1% (three more halvings) | Claudius II, c. 270 | About 10 years |
Two centuries to lose half. Two decades to lose nearly all the rest. By 270 the “silver” coin was a bronze blank with a thin silver wash that rubbed off in circulation, and the mints were striking them in enormous numbers.
Who Was Paying for It
The army was the biggest line in the imperial budget, and pay rises track debasement closely. Legionary base pay is itself reconstructed from scattered evidence, so the later figures come as ranges.
| Pay rise | Annual base pay (denarii) | Silver per denarius at the time (g) | Pay in silver (g per year) |
|---|---|---|---|
| Augustus | 225 | 3.8 | 855 |
| Domitian, c. 84 | 300 | 3.1 | 930 |
| Septimius Severus, 197 | 450-600 | 1.7 | 765-1,020 |
| Caracalla, 212 | 675-900 | 1.5 | 1,013-1,350 |
| Maximinus Thrax, 235 | 1,350-1,800 | c. 1.4 | 1,890-2,520 |
Look at the right-hand column. Measured in silver, the legionary’s pay held level or rose through the whole period. The soldiers kept their silver. Debasement let the state keep raising their pay without mining any more of it.
The people who lost were everyone else holding coin: civilians with savings, cities with endowments, the state itself when taxes came back in its own debased money. Soldiers also saw a large share of pay deducted for food and kit, and in the third century the real value increasingly came through donatives and payments in kind. The silver column is gross pay, not take-home.
The Reforms
Aurelian’s reformed radiate of 274 fixed the silver at around 5% and apparently marked it (the XXI on many issues is usually read as a 20:1 ratio of bronze to silver). It stabilised the coin at a low level. It didn’t restore anything.
Diocletian’s argenteus of 294 did restore something. At roughly 3.4 grams and over 90% fine, it put about as much silver into one coin as Nero’s reform denarius had. It was struck in small numbers, and within a few years his Price Edict of 301 was trying to hold down the prices that three decades of bad money had pushed up.
The good coin came back. The trust didn’t.