Roman Jewelry

Roman jewelry reached every level of society: rings, earrings, necklaces, bracelets, brooches and hairpins, in bronze, silver and gold, set with engraved gems, moulded glass, pearls and emeralds from the eastern trade routes. The Metropolitan Museum's gold and emerald necklace, accession 21.29.2, 1st to 2nd century CE, marks the top of that range.

What reaches a collector is a narrow slice of it. Gold was routinely melted and reworked. Iron corrodes to oxide and rarely survives whole: one finds record, FindID 468587, a cast iron ring of the 1st to 2nd centuries, is typical, broken and corroded. Bronze survives well under a stable patina. Survival bias and Roman rank rules together, not modern taste, are why a bronze ring is the ordinary object in any Roman collection today, mine included.

Almost always a modern maker. A Roman-period ring carries no maker's mark and no fineness stamp, because neither was used on Roman finger rings, so a word or a number stamped into a hoop places the piece in the modern era on its own. ROMAN in Latin capitals is a trade name. The numbers are fineness marks: 925 is sterling silver, 585 is the gold alloy sold as 14 karat. If your piece reads ROMAN, 925 or 585, it is modern.

Four metals: bronze, silver, gold and iron, with bone, glass and rock crystal turning up more rarely. Pliny the Elder, writing in the 70s AD, treats the metal of a ring as a matter of rank rather than fashion, and records the iron ring as the older Roman habit, worn as an emblem of warlike valour.

Gold was regulated. Pliny reports a rule under Tiberius restricting the gold ring to a man free-born whose father and grandfather were free-born too, and rated at 400,000 sesterces, the equestrian census. The entitlement widened over the following centuries, reaching every citizen by Justinian's codification in the traditional account compiled in Smith's Dictionary of Greek and Roman Antiquities. A bronze Roman ring is therefore not a lesser object. It was the normal one.

What the trade calls bronze is a copper alloy, and the mix varies. Facsady and Verebes, analysing Roman finger rings from Aquincum in 2009, report zinc in every ring they measured, brass as the commonest alloy, and leaded gunmetal pieces running 8.3 to 23.1 percent lead. A Portable Antiquities Scheme record puts most gold Roman rings of the 2nd and 3rd centuries AD above 92 percent gold, a curatorial statement about British finds rather than a universal rule.

My stock reflects the same pattern. Bronze dominates my Ancient Roman Rings collection, silver is much less common and gold rarer still, with a bone ring among the remainder. A handful of late and Byzantine-period rings sit in that collection alongside the Roman ones. The pieces available follow the same order, bronze first, then silver, with gold rare.

A signet ring carries its device cut in reverse, sunk below the surface, so the impression it leaves in wax stands up in relief and reads the right way round. The reverse cutting is checkable: a Roman silver ring from Roundway Down in Wiltshire, published as Brit. 49.55 in Roman Inscriptions of Britain, carries the Greek word for victory incised backwards on its bezel.

The seal was then protected. A cord binding a writing tablet or a package passed through the pierced walls of a small bronze seal box and was knotted inside; wax was run over the knot and stamped with the ring, and the lid closed over it. Colin Andrews catalogued the British seal boxes in 2012, and their exact use is still argued over, so the mechanism is worth describing and a single purpose is not. None of this is the red shellac wax of later centuries.

Then the honest limit. Signet describes a form, not a proven use. Whether a ring was ever pressed into wax is rarely provable from the object. I say what a ring is, not what it did.

Carnelian is the commonest Roman choice, with nicolo, a banded onyx showing a pale blue upper layer over a black ground, and red jasper the other regulars. A checkable example: the Metropolitan Museum's gold ring with a nicolo intaglio of an ibis, accession 44.2.4, Roman-period Egypt, 2nd century CE.

The stone was not carved with a blade. The engraver held it against a rotating wheel or drill point fed with an abrasive slurry, and the grit, harder than the stone, did the cutting; a 2003 study in the Journal of Archaeological Science identified abrasive and lubricant residues in the engraved grooves. Glass was the cheaper substitute, moulded rather than engraved, and Pliny called glass ring-stones the ones ordinary people used.

Good fortune, mostly: Fortuna with a rudder and a cornucopia is among the commonest figures, with Bonus Eventus and Ceres also recurring. A jeweller's stock buried at Snettisham in Norfolk around AD 155 held 117 engraved carnelians of which only seven were mounted in rings, published by Catherine Johns in 1997, so gems were cut and sold loose, then mounted to order. On a worn gem the subject is often only attributed: a curator recording FindID 402164 read its figure as possibly Sol or Mercury and left it there. I write mine the same way.

Not in the modern sense. Rome had a betrothal ring given at the engagement, not a pair exchanged at a ceremony. Pliny records that an iron ring, and one without any stone in it, was sent as a gift to a woman when she was betrothed. The same passage shows how situational the custom was: men given gold rings for an embassy wore them in public and put the iron ring back on at home. Clasped right hands, the dextrarum iunctio, is the motif that carries marriage meaning, though the gesture runs across Roman art generally.

The only ancient authority for a reason is Aulus Gellius, writing in the 2nd century AD. In Attic Nights 10.10 he puts the ring on the finger next to the little finger of the left hand, citing the Greek writer Apion: Egyptian dissection was said to have found a very fine nerve running from that finger to the heart. Note the word. A nerve, not a vein. And the anatomy is Apion's, reported at second hand by Gellius, with no Egyptian anatomical source for it surviving.

Practice was looser than the story. Pliny describes the custom moving from the finger next to the little finger to the one next to the thumb, and then to the little finger as well; he notes that the Gauls and Britons used the middle finger, which in his own day was the finger Roman hands left bare. The phrase vena amoris is not classical at all. Its earliest known use is in a treatise on matrimonial contracts by the English lawyer Henry Swinburne, published posthumously in 1686.

By form, and the form gives a bracket rather than a year. Two typologies are standard: Martin Henig's, built on British material and published as BAR British Series 8 in 1978, and Helene Guiraud's, built on Gaul and published in Gallia 46 in 1989. Finds records quote both together, so a single ring can carry a Guiraud number and a Henig number at once.

The variables are the cross-section of the hoop and the way the hoop meets the bezel. Henig Type II, a hoop of circular section curving smoothly into an oval bezel, is the common Imperial form, generally given the 1st to 3rd centuries. Type III, flatter in section with a slight angle at the bezel, is dated in finds records to the second half of the 1st century and the 2nd. The heavy-shouldered Types X and XI are a later 2nd to 3rd century form.

The frameworks also disagree, openly. A bronze ring with eight external facets, FindID 591202, is recorded citing Henig for a 3rd century reading of polygonal hoops and Crummy's Colchester sequence for a 4th century tendency, carrying both rather than choosing.

A word about my own material. I catalogue every piece myself, and every piece is a single object from an old collection with no excavation find context, so attribution rests on typology alone. I grade it: determined, where an unambiguous typology settles it; attributed, where the typology is plausible but alternatives stay open; or date secure with the cultural region open. Almost every available ring in the collection carries a range spanning more than one century, the rare exception being a bronze ring given to the 3rd century. A single-year date on a Roman ring deserves a question.

Metal and condition set the price far more than age does. Bronze sits lowest, silver well above it and gold higher again. The date brackets overlap heavily across those three groups, so age is not what separates them. The metal moves everything.

A cut stone raises the floor: intaglio rings start well above the plain bronze bands, while the Roman signet rings in stock are all bronze. Prices here are my asking prices, not a survey of the market.

What a listing tells you matters as much as the number on it. I photograph every piece on its own, give both EU and US size, state a date range rather than a year, and write the condition, including old breaks and losses.

Many are wearable, and size is the constraint rather than the metal. Roman rings survive across most of the modern range, from roughly US 3 to US 12, so the assumption that ancient rings are all tiny does not hold. What is on offer at any moment is a narrower slice of that, and every listing gives its own EU and US size.

An ancient hoop should not be resized. The metal has been in the ground, it is work-hardened and often thinned by corrosion, and bending it is how a two-thousand-year-old object becomes two pieces. I do not resize them and I advise against it.

Bronze can leave a green mark where copper salts form with sweat and skin acids. In my experience it washes off skin and off most fabrics. A red, itchy or blistering reaction is a different matter and not something I can rule in or out for anyone; if that happens, stop wearing the piece and ask a doctor. A hoop with an old break or a paper-thin section is a display object, and the listing says so.

Coins behave differently from rings. A ring is placed by form against a published type series and comes out with a bracket. A coin is placed by legend, portrait and reverse type against Roman Imperial Coinage, the standard reference, and can come out with a tight date. Most Roman coins here carry an RIC citation in the listing itself, including every one available, for example RIC IV Severus Alexander 92.

The denarius was the everyday silver coin, about 3.9 grams under Augustus; four sestertii in brass, roughly 25 to 28 grams each, made a denarius, and the gold aureus was worth 25 denarii. Caracalla added the antoninianus in early AD 215, marked by the emperor's radiate crown to signal a double denomination, but it carried only about one and a half denarii worth of silver. That gap is the point of the coin.

The silver thinned steadily: above 90 percent fine in the 1st century AD, below 60 percent by AD 200, about 5 percent by AD 300. Butcher and Ponting's 2014 study put the early Imperial end of that curve on a scientific footing, from Nero's reform to Trajan's; the later figures are the standard summary numbers and are not theirs. Severan silver is the least expensive way into Roman coinage here, and each piece is determined by legend and portrait against RIC, naming Severus Alexander, Julia Mamaea or Julia Maesa.

I sell only pieces I can document as lawfully exported and not stolen, and I ship from Germany, which has required evidence of lawful import since 2016. Britain's Dealing in Cultural Objects (Offences) Act 2003 targets dishonest dealing in objects removed after it came into force. What usually decides a case is the source country rather than the buyer's: Italy and Greece treat undiscovered antiquities as state property, and the United States restricts imports of specified archaeological categories under bilateral agreement. The rules bite on how an object moved, not on how old it is, and I am not in a position to give legal advice for your country.

Intaglio: an image cut into a stone, sunk below the surface. Cameo: an image raised above it. Signet: a ring made to seal. An intaglio ring is a ring set with an engraved stone, which may or may not ever have sealed anything. Hoop and bezel: the band and the setting, the two features by which a Roman ring is dated. Nicolo: a banded onyx, pale blue over black.

Determined, attributed, and date secure with the cultural region open are the three grades I use, in that order of confidence. If a description does not say which of the three it stands on, ask. Roman is one strand of a longer story.

Every piece is a single object, so the collections change as things sell. These are the doors into the catalogue.