Ancient Jewelry

Ancient means made in the period it is named for. Antique, in the jewelry trade, means something far younger.

The word does two jobs on the open web, where dealers in excavated material sit beside brands selling new pieces in an ancient manner. The line worth holding is simple: an ancient piece was made in antiquity or in the medieval centuries and has survived. A piece made last year in a Roman manner is modern, however well made.

Antique is a different word with a different threshold. United States customs uses it numerically: tariff heading 9706 covers antiques of an age exceeding one hundred years and admits them duty free, with the age claim tested at entry. In the jewelry trade the same word points at Georgian, Victorian, Edwardian, Art Nouveau and early Art Deco work. So antique rings, as the trade uses the phrase, almost never means a Roman ring.

One trap sits inside that hundred-year line. Being over a hundred years old does not make an object freely tradeable. Under Council Regulation (EC) No 116/2009 an archaeological object more than a hundred years old needs an export licence to leave the European Union, with no minimum value, so a bronze ring at two hundred euro carries the same paperwork as a far more valuable piece. Tariff category and permission to ship are separate questions.

No one can answer that with a single object, and ornament in general is far older than metal.

The oldest securely dated ornaments include forty-one perforated Nassarius shell beads from Blombos Cave in South Africa, from sand dated by luminescence to about 75,000 years ago, published in the Journal of Human Evolution in 2005. A 2021 report on Bizmoune Cave in Morocco puts thirty-three perforated shell beads at roughly 142,000 years, by uranium series on the surrounding strata. The figure moves as excavation moves, so the honest sentence names the site and the method.

Worked gold begins much later: the copper age cemetery at Varna in Bulgaria, roughly 4600 to 4200 BC, produced around three thousand gold objects weighing some six kilograms. Metal finger rings are documented from the third millennium BC, among them rings from the tombs at Ur in southern Iraq around 2500 BC.

For the ring itself there is no first place. Candidates come from several regions and several millennia, and none could be traced to a museum record in the research behind this page.

My own catalogue covers a much narrower window. Most of it runs from the 5th century BC to the 15th century, with a small later tail dated from the 16th century onward, sold as later work rather than as ancient. Counting by the leading century in each title, the material clusters in the 1st to 3rd centuries and again in the 11th to 13th, and the earliest pieces carry BC dates.

Four groups carry the market: Roman, Byzantine, medieval European and, in far smaller numbers, Islamic. Each is read from what is visible on the object, never from a label alone. These are typological attributions, not excavation records, because pieces from old collections arrive without find context. Each listing carries its grade: determined, attributed, or date secure with cultural region open.

Roman finger rings are classified by the shape of hoop and bezel rather than by metal. Two frameworks carry most of the weight: Martin Henig's Corpus of Roman Engraved Gemstones from British Sites (BAR British Series 8, 1978) and Helene Guiraud's typology of rings in Roman Gaul (Gallia 46, 1989). A raised oval bezel on a plain D-sectioned hoop, Henig type III, is conventionally placed in the 2nd and 3rd centuries AD. Read that as a convention, since authors shade the boundaries differently, and treat a bare type number with no feature named on the object as worth nothing. For the object-level checks see How can you identify a Roman ring? The full typology is on Roman jewelry.

Byzantine rings turn to crosses, monograms, personal names and Christian images, which What does Byzantine jewelry symbolize? reads one by one. One museum gives a chronology you can check: the Metropolitan Museum of Art holds finger ring 52.7, gold with niello and a sapphire intaglio, 5th to 6th century; the Ring of Leontios, 1982.282, gold and niello, about AD 1000; and the gold signet ring of Michael Zorianos, 18.145.42, about 1300. Niello, a black sulphide inlay burnished into engraved lines, appears on two of those three.

The reference shelf is Dalton's 1912 British Museum catalogue of finger rings, free on the Internet Archive; Ward-Perkins's London Museum Medieval Catalogue of 1940; and Egan and Pritchard's Dress Accessories c.1150 to c.1450, publishing over two thousand excavated London dress items. Stratified material is what makes those volumes usable for dating. In my own stock the signet is the dominant form: a large share of the medieval and Crusader rings also sit in a signet, seal or devotional collection, among them the medieval signet rings. Non-ring forms are thin: medieval pendants are far outnumbered by rings. The period in full is on medieval jewelry.

The pieces catalogued here as Islamic are read from Arabic script on the bezel where it survives; where it does not, the attribution is graded as attributed rather than determined. They are dated between the 10th and the 18th century, so some run past the medieval centuries. Greek, Celtic and Nordic material appears occasionally.

Coins are the exception to everything in the next section. Roman Imperial Coinage, thirteen volumes covering 31 BC to AD 491, ties a coin to a named ruler, which is why coin listings carry RIC references. What to check on the coin itself is in How to authenticate an ancient Roman coin.

By comparison, not by machine. There is no laboratory method that returns an absolute date for a metal object.

Radiocarbon reads organic carbon, and a cast bronze or gold ring contains none. Metal is dated by find context, by datable material lying with it, and by typology measured against published type series, which at the level of one object looks like How to figure out how old a ring is. Radiocarbon can reach a metal object only through adherent organics, never through the metal itself. Anyone who tells you a bronze ring was carbon dated is describing something that did not happen.

Composition analysis is a real check, but it is not a date. Roman workshops made brass by cementation, heating copper together with zinc ore, from the 1st century BC onward, and the zinc measured in Roman brass runs at most to about 28 percent. A piece carrying appreciably more zinc was alloyed with metallic zinc, a later technique no Roman workshop had. That excludes rather than confirms. Portable XRF reads only the surface, which corrosion can bias.

Museums bracket too, and the width of the bracket is a statement about evidence. The Metropolitan Museum dates a Greek gold ring, 06.1123, to the third quarter of the 5th century BC, and leaves a loose Roman pink quartz intaglio, 41.160.1047, open from the 1st century BC to the 3rd century AD. Evidence sets the width, not confidence.

I sell single pieces from old collections. They carry no excavation find context, so the date rests on typology alone, and I say so on the listing rather than imply a record that does not exist. Determined means find context or unambiguous typology fixes the piece; attributed means the typology is plausible but alternatives stay open; date secure, cultural region open means the century holds but the workshop does not. I claim no laboratory dating, and I name no style unless I can point at the feature that carries it.

Mostly bronze, because bronze is what survives in numbers.

Bronze dominates the catalogue, silver is much less common and gold is rare, among the rings as much as across everything else. Gold does not corrode, so an ancient gold ring can leave the ground close to how it was made, and the absence of patina is no argument against age. Gold was also worth melting in every period. A bronze ring was not, which is why base metal reaches us in quantity and gold does not.

Workshops also worked iron, glass and engraved stone. The Roman intaglio repertoire runs on a few hardstones: carnelian above all, with jasper, nicolo, a banded stone with a pale blue-grey layer over a dark base, and moulded glass. An intaglio is cut into the stone so its impression stands up in relief. A cameo is the inverse.

Gem cutting was a separate trade from ring making. The Snettisham jeweller's hoard, buried in Norfolk around AD 155, held 117 engraved carnelians of which only seven were mounted in rings. A customer could choose a stone and have it set, which is one reason a stone and its mount can carry different dates.

Yes. Bronze, silver and gold are worn comfortably every day; iron is for display and should never be wetted. Please wear them with care: it always depends on the individual ring, and an ancient ring can still be damaged.

The patina is structural, not decoration. On buried bronze the stable corrosion products are cuprite, a red-brown oxide, and malachite, a pale green carbonate. Those layers hold thinned metal together and carry what a conservator reads, so cleaning them away weakens the ring. Take a ring off for swimming, chlorinated water, perfume and cleaning products.

One condition problem is worth naming, because it is a fair question to put to any seller. A bright, powdery, pale green that wipes off and keeps coming back is not patina but chloride corrosion, known as bronze disease. Chlorides form nantokite beneath the oxide and carbonate layers, and water turns it into hydrated copper chlorides of much greater volume that erupt through the surface. It is progressive and needs a conservator. Buried silver has its own version, horn silver, a brittle chloride crust.

Green marks on the finger are ordinary chemistry: copper reacts with the moisture, salt and mild acids in sweat to form salts that rub off and wash away with soap and water. Itching or a rash is a different matter. It can point to a metal sensitivity, which a clinician should assess by patch testing, and it can also come from powdery chloride corrosion transferring off an unstable surface, so it belongs with a clinician and with the seller. The fuller treatment, including the lead question, is in Is it okay to wear bronze jewelry?

An ancient ring is a fixed size. A hoop thinned by corrosion cannot be resized or re-shanked without destroying original surface, so the size has to be right before the piece is bought. Every ring I list carries its EU and US size in the title for that reason.

Less than most people expect, and from wherever the paperwork actually exists.

Prices vary by collection, and the collections overlap, so no single range covers them all: Roman rings, medieval and Crusader rings, Byzantine rings, signet rings, intaglio rings and Roman coins each sit in their own band, the coins almost all silver with billon the exception. Stock turns over, so read any figure you see as the shape of the market on the day, not a standing price list.

Ask for the collection history, ideally reaching back before 1970. The UNESCO convention on the illicit import, export and transfer of ownership of cultural property was adopted on 14 November 1970 and entered into force on 24 April 1972, and its date is the line the market uses. Two trade bodies publish codes you can hold a dealer to: the International Association of Dealers in Ancient Art and the Antiquities Dealers' Association.

Buying and owning ancient jewelry is legal in most of Europe and in the United States, provided the piece left its country of origin lawfully and has been traded lawfully since. Two articles cover that ground country by country: Are ancient Roman coins legal to own? and Can you legally buy Egyptian artifacts? Since 28 June 2025 the European Union has required an import licence for archaeological objects at least 250 years old under Regulation (EU) 2019/880, handled through a centralised electronic system. That is a summary as it stands in September 2026, not legal advice.

I catalogue and examine every piece myself. My pieces come from a private family collection (initials K.V.D., Belgium, on the listing), from private collections in the UK and elsewhere in Europe, and from European dealers and antique markets. Each listing gives the collection history as far as it is recorded and the date range; coin listings add the RIC reference, ring listings the EU and US size. Returns run thirty days. If a piece bought from Aurora is ever shown not to be authentic, or not of the period stated at purchase, we refund the full purchase price. There is no time limit. The guarantee applies to the original buyer and to anyone who inherits the piece. Outside the European Union I prepare and file the export licence application myself. The turnaround is not mine to promise: the licence is issued by the heritage authority of the country the piece ships from, for me the Belgian one, and that review typically takes about three weeks.