Coin Relief – Issue Nine

Here is the next edition in a series of blog posts written by Dr. Sam Moorhead and Dr. Andrew Brown, the PAS Finds Advisers for Roman and Iron Age coins.

Other 4th century silver denominations

As we saw in the last edition of Coin Relief, during the second half of the 4th century the silver siliqua became the standard silver denomination in circulation and recorded through the PAS. However, this was by no means the only silver coin struck during the 4th century. Indeed, even the nummus contained a percentage of silver (perhaps initially a notional c.5%) within its alloy. In this edition we will look at three other coin types struck in good silver during this period that although appearing in much smaller quantity on the PAS also constitute important components of 4th century Roman coinage.


Argenteus of Diocletian c.AD 300. British Museum coin B.1474 (Copyright: Trustees of the British Museum).

Diocletian’s currency reforms in c.AD 294 introduced a new silver denomination, the argenteus (literally ‘silvery’ or ‘of silver’). On the incomplete Aphrodisias currency inscription of c.AD 301 the value of the denarius argenteus (‘silver denarius’) is fixed at 100
denarii communes (‘common denarii’; now more a unit of account that is perhaps represented by small billon coins with laureate busts) – four times the value of the newly introduced
nummus. This high value silver denomination contained a substantial percentage of silver and was struck at 96 to the Roman pound until c.AD 313. On some argentei (such as the one above) this value of 1/96 of the pound is explicitly represented by the reverse legend XCVI. The argenteus was a relatively short-lived phenomenon and it is no-longer struck after c.AD 310-313 once inflation increased the bullion value of silver. The argenteus is a rare coin generally, but especially so in Britain and within the material recorded through the PAS (see this search). Indeed, from the early period of the Tetrarchy, there are just two examples on the PAS (ESS-C83012 and WILT-CA45F1) but both are contemporary copies of the same type for Diocletian. The PAS data also includes several examples from a small group of base silver coins struck at Trier at the end of the period argentei were issued, c.AD 310-313. These contain perhaps as much as c.25% silver and are regarded as either highly silvered (or silver washed) nummi or base/pseudo argentei. They are struck for Constantine I, Licinius I, and Maximinus Daia, each with a distinctive reverse type.


Alongside the appearance of the siliqua in the AD 320s was a larger silver denomination struck at about 4.5g or 72 to the Roman pound and called a miliarensis. This was in circulation contemporary with the siliqua although is much less common as a single find or indeed in hoards – Hoxne had just 60 examples. The term scrinium a miliarensibus appears in later Roman documents and has plausibly been linked to these larger silver coins we call miliarenses, so-called due to their value at 1/1000 of a pound of gold. A heavier silver coin struck at about 60 to the Roman pound or c.5.4g is also known and termed a ‘heavy
miliarensis’. This is a much rarer coin – there are none in Hoxne but there is one PAS example (HAMP-2197A7). The majority seen as single finds, though, are of the lighter weight variety. In total, the PAS records 20 single examples of miliarenses, covering the entire period of their production and usage with examples in Reece Periods 16-21 (see this search). The majority cluster around Reece Periods 18 (4 coins) and 19 (10 coins) and it is interesting to note that of those, one is of uncertain Reece period, two lack images with mints identified, coins from the eastern mint of Thessalonica are the most frequently found (albeit only with 5 examples!).

Half siliqua

Half siliqua of Arcadius dating to AD 395-402. Record ID is HAMP-03C5B0 (Copyright: WInchester Museums Service, License: CC-BY).

While the majority of siliquae recorded through the PAS are full weight coins, fractional half siliquae were also struck in the late-4th century at the western mints of Trier, Milan, Rome, and Aquileia. These circulated alongside the siliqua and miliarensis but are rare in both hoards (Hoxne has just 5 examples) and as single finds. They can be separated from the siliquae due to their size and weight (up to c.1.5g), but also the types represented and the mintmarks on the coins themselves. Whether they functioned as currency in the same way as the other silver denominations of this period is unclear and it is possible they had a more donative or ceremonial function (see R. Bland, 2010: p. 206). The first example of a very rare anonymous issue from the House of Constantine (BERK-4AF264) recorded in Britain appeared on the PAS database in 2014. To date, this is exceptional within the PAS data and the remaining examples all belong in the period from c.AD 367-402. A good starting point for these fractional siliquae is the articles by S. Bendall and R. Bland but there is also brief discussion of the type in P. Guest’s Hoxne volume (Guest, 2005: pp.

Only one coin has been identified on the PAS as a half siliqua with full legends on obverse and reverse attributed to a specific ruler (HAMP-03C5B0), although there are examples from hoards. It carries a Victory reverse type with the legend VICTORIA AVGG – the Victory type is the most typical for the smaller denominations, although there are other known examples too. Note that on this coin the mintmark is shortened to just MD for Milan rather than MDPS. In his analysis of Hoxne, Guest (2005: pp. 44-45) noted how it was originally assumed that the lack of PS in the mintmark would mean these types were not struck from the refined silver of the siliquae. However, examination of the metallurgy suggested this was not the case and they are struck from good silver. The remaining three coins belong to a group of anonymous half siliquae that distinctively do not display the name of the ruler, simply a vota reverse legend and mintmark. Roger Bland’s
(2010) analysis of anonymous half siliquae notes that the vota reverse may be significant in terms of dating these issues, potentially linking it to the 10th anniversary of Theodosius I (c.
AD 388-389) or the 10th anniversary of Arcadius (and 15th of Theodosius, c. AD 392-393). He suggests the second of these is perhaps most likely, but their precise date of issue or reason for issue is still not entirely certain. This type is issued from the two mints of Trier and Aquileia. At Trier the obverse has a helmeted bust of Roma left, while the reverse contains either an X or XV within a wreath. In contrast, coins from Aquileia have the bust facing right and with the vota XV. One coin of this type, from Trier, was identified in Hoxne (Hoxne no. 759) and a nice example appeared in a hoard from Somerset in 2010 (SOM-6E89B0), but these are still not common coins (see this search for all examples on the PAS database).