Tag: antique identifier by picture

A mark narrows an antique’s maker, country and date, but on its own it is not proof of authenticity: marks are the easiest part of any piece to fake.

This guide is published by Antique Identifier – Antiqly, which makes a photo identification app that is free to download on iPhone and free on Android with no sign-up, and it was last updated on 17 September 2026. It is a review of the published sources linked in the text — the Goldsmiths’ Company Assay Office in London, the Sheffield Assay Office, The National Archives, UK hallmarking law and the US import marking rules — and not a hands-on test: no object was examined for it, and it puts no value on any individual mark. Where a figure would need a dated, comparable sale behind it, this guide sends you to the records instead of inventing a range.

What a mark can and cannot tell you

A mark is evidence, not a verdict. It answers some questions well, answers others only partly, and answers the two questions collectors most want settled — is it real, and what is it worth — not at all. Knowing where that line falls is the whole skill, so it is worth fixing before the families of marks are worked through one by one.

The first row is the one most identification guides get wrong, and the offices that strike the marks are blunt about it. The Sheffield Assay Office says of the sponsor’s mark — what older books call the maker’s mark — that it “will only tell you who registered the mark and submitted the item to be tested and hallmarked – which is not always the same thing as the manufacturer of the item”. British design registration has the same gap: The National Archives records that the proprietor named in a registered design is “usually the manufacturer or retailer rather than the designer”. In both systems the mark names a business that took responsibility for the object, which is not the same as the hand that made it.

The four families of marks and where to look

Each material marks in its own way, in its own place, and gives up a different kind of date. Turn the piece over before anything else; on furniture, open it.

For ceramics the mark is usually the most informative thing on the piece, which is why our guide to decoding a ceramic mark is organised by mark type and the earthenware and porcelain reference covers the body and glaze that have to agree with it. Silver is the opposite case: the metal itself is anonymous and the punches carry the record, so the British silver hallmarks chart and the American silver hallmarks guide are arranged by punch rather than by maker. Glass signatures are faint and easy to add after the fact, and the visual dictionary of glass marks and the French glass makers’ marks guide show what genuine ones look like at scale. Furniture is the hardest of the four, because most period furniture was never marked at all: furniture stamp identification and the alphabetical list of furniture makers’ marks cover what does turn up, and dating by hardware is usually the stronger evidence. For symbols that belong to none of the four, our marks and symbols guide is the wider index.

British hallmarks: the one system built to be read

British hallmarking is the most systematic marking scheme an ordinary collector will meet, and the place to learn what a mark can carry. Assay Office London, the Goldsmiths’ Company office, defines a hallmark as “a set of component marks applied to articles of the precious metals gold, silver, platinum or palladium” and says that it “Guarantees that it conforms to all legal standards of purity (fineness)”. Read that guarantee narrowly. It is about metal content, tested by an office; it says nothing about who made the object, how old the object is, or whether the piece around the punches is what it appears to be.

Three marks are compulsory on a modern British hallmark: the sponsor’s mark, the assay office mark and the fineness mark. The date letter is now optional and “changes annually on January 1st”. On antique silver the group is usually larger, and each punch answers exactly one question.

Two of those rows carry traps. The English hallmarking system began with a statute of Edward I, so the English sterling statute dates from 1300 and anything claiming to be marked English sterling before it is claiming too much; and from 1697 there was also a higher standard of silver, known as the Britannia standard, with its own marks, so a piece that fails a sterling reading is not automatically wrong. The date letter is the other trap. It started as the mark of the Assay Master, “who was sworn in every May”, and it advanced when a new master was sworn rather than at New Year — which is why an antique date letter marks an assay year that straddles two calendar years, and why “1889” on a dealer’s ticket is shorthand rather than a date of manufacture.

Which office, and why a rose is not always Sheffield

There are four assay offices left in the UK — London, Birmingham, Sheffield and Edinburgh — and their town marks are the part of the hallmark that identification guides most often get wrong, this one included until now.

Assay Office London states that the leopard’s head is the town mark for London. The Sheffield row is the correction this refresh owes its readers: earlier versions of this guide repeated the common shorthand that a rose means Sheffield. The office’s own record says otherwise for anything antique — “The Crown was used by Sheffield Assay Office to mark silver from 1773 until 1973” — and the rose took over on silver only when the crown was retired, so a rose struck on silver is a modern Sheffield mark, not a Georgian or Victorian one. Birmingham has drifted the other way: since the office began striking through an overseas subsidiary, its own history records that “The anchor mark no longer indicates that an item was assayed in Birmingham.” Marks do not stand still, which is why the register beats the symbol chart: the Sheffield Assay Office holds registers of sponsors’ marks going back to 1773 and will identify a Sheffield mark from a photograph.

Dating clues that actually hold

A few marks fix a date because the law, the registry or the technology behind them did not exist earlier. Those are worth memorising, because they set a floor that cannot be argued away. Everything else gives a range.

Three cautions go with the table. A registry mark dates the design, not the object: a pattern registered in 1876 could be pressed for decades afterwards, and The National Archives adds that “As registration was not compulsory, there are many commercially produced designs which you will not find in our records”, so the absence of a registry mark proves nothing at all. The diamond mark itself was “printed, engraved or otherwise marked on an object”, most often on the underside of china, so it is easy to miss under a foot rim. And the two modern floors date the printing rather than the object: the five-digit ZIP Code arrived in 1963 and the first UPC barcode was scanned in 1974, so each fixes the age of a label, a box or a reprinted backstamp that may be decades younger than the piece it is attached to.

“England” and “Made in England”: a rule of thumb, not a law

The version of this rule that circulates among collectors — “England” alone means 1891 to 1920, “Made in England” means after 1920 — is tidier than anything this guide can trace to a statute, and it is treated here as trade shorthand rather than a dated legal boundary. What US law actually requires, under section 304 of the Tariff Act of 1930, is that an imported article be marked so as to indicate to an ultimate purchaser “the English name of the country of origin of the article, at the time of importation”. That is a rule about naming the country in English. It sets no day on which factories had to add the word “Made”, and in practice a pottery changed its backstamp when it cut a new printing plate, not when a tariff clerk asked. Use the wording as a soft indicator, then let the body, glaze and pattern carry the date.

The same discipline applies to Chinese reign marks, where a mark is very often a mark of respect for an earlier emperor rather than a date of manufacture — our guides to the Kangxi reign mark and the Qianlong mark set out how apocryphal marks are read.

Why the mark is the first thing a forger copies

A mark is the smallest and softest part of an object to alter, and the law has treated it that way for a long time. The Hallmarking Act 1973 makes it an offence to make, with intent to defraud or deceive, “a counterfeit of any die or mark” — and, in the same section, to remove “any mark from an article of precious metal with intent to transpose it to any other article”. That second offence is the transposed hallmark, and it is the trap that catches careful buyers of antique silver: a genuine Georgian punch group cut out of a small, low-value piece and let into a larger, later one. The hallmark is real. The object carrying it is not.

Porcelain has its own version, and it starts at the top of the market. Meissen’s crossed swords were introduced in 1720 and “used consistently after 1731 by official decree”, and “Variations in the logo allow approximate dating of the wares” — but the same account of the factory notes that in the eighteenth century “the mark was not considered important, and it was commonly painted on in a crude manner”. A crude eighteenth-century Meissen mark can therefore be right, and a crisp, carefully drawn one can be wrong, which is the reverse of what most checklists say. Our real versus fake crossed swords guide works through the variants and the imitators.

Check a suspect mark in this order:

1. Placement. Is the mark where that maker and that period actually put it? A mark in an unusual spot, or one that would have been awkward to apply during manufacture, is worth more suspicion than a mark that reads badly.
2. Relationship to wear. A mark that sits on top of surface wear, or that is crisper than the wear around it, was applied after the object had a life.
3. Relationship to the surface. On ceramics, whether the mark sits under or over the glaze narrows the period and the technique. On metal, a struck punch displaces material; a mark with no corresponding stress or distortion around it is not a struck mark.
4. The rest of the object. Form, body, glaze, construction and hardware all have to agree with the mark’s date. When they disagree, the mark is the part that changed. Our guide to telling a real antique from a reproduction covers the rest of that reading.
5. An independent register. Take the mark to the body that keeps the records rather than to a symbol chart, and compare against a photographed example of the same variant. Kovels is the standard printed and online marks dictionary for American and European wares; the assay offices identify their own punches.

Retailer, decorator, import and registration marks

Second marks are common and they are the ones most often misread as maker’s marks. A retailer’s stamp — a department store, a jeweller, a china dealer — records who sold the piece, and on silver it often sits beside the assay group without belonging to it. A decorating studio’s mark records who painted a blank that another factory made, which is why the mark on the base and the style of the painting can honestly point to two different firms. Import marks were added at the border and date the arrival of the object, not its manufacture. Registry marks, as above, date the design.

None of these is a flaw. A full set of secondary marks is a provenance trail, and a piece that carries a maker, a retailer and a registry mark together is easier to place than one with a single stamp. The error is only to read a retailer or an importer as the maker. For signatures rather than punches — artists, silversmiths and studio potters — see our antique signature identification guide.

What a marked piece is worth

This is where mark guides usually stop being useful, because a premium gets quoted as a percentage with no sale behind it. Earlier versions of this page carried several such figures and a column of price ranges for famous makers. None of them had a source, and they are gone rather than replaced with different invented numbers.

What can be said honestly is the mechanism. A mark makes a piece attributable, datable and easier to describe, and buyers pay for that certainty rather than for the stamp itself. How much, for a particular piece, depends on the maker, on which variant of the mark it is, on condition, and on how recent and how comparable the sale you are measuring against happens to be. That number has to come from records.

Two routes get you a real one. Search completed sales for the same maker, form and period — WorthPoint and auction-house archives both index them — and check that the comparables carry the same mark variant and not merely the same name. Read those figures the way the sibling guides on this site do: a hammer price is a gross figure, and what reaches a seller is smaller once the buyer’s premium and the seller’s commission are accounted for, so a marked piece is not worth the headline number. Then, if the piece matters, commission a written appraisal from a credentialed appraiser through the International Society of Appraisers. Our guide to looking up antique values and the professional appraiser’s method walk through both routes in detail.

A checklist for the mark in front of you

1. Photograph the mark square-on, in raking light, with a ruler or a coin in frame for scale.
2. Note exactly where it is, and what else is near it.
3. Decide what kind of mark it is: impressed, printed, painted, incised, struck, etched, moulded or a label.
4. Look for the other marks. Hallmarks travel in groups, and a single punch is rarely the whole record.
5. Read the mark against the object: does the form, body, glaze, joinery or hardware agree with the date the mark suggests?
6. Compare against a photographed example of the same variant, not a line drawing.
7. Take it to the register that covers it — an assay office for British precious metal, a marks dictionary for ceramics and glass, a museum collection for a documented maker.
8. Write down what you could not resolve. The unresolved part is what a specialist is for.

Where a photo app fits

An identification app is a fast first pass, not a substitute for the registers. Antiqly reads a photograph, suggests which family a mark probably belongs to and gives you a starting point for the reference you should check next; it is free to download on iPhone and free on Android, with no sign-up. It works from an image, so it cannot confirm that a mark was struck rather than added later, and it does not replace an assay office register or a written appraisal. For the longer comparison, see an app against a professional appraiser and the Antique Identifier app page.

Frequently Asked Questions

Can an app identify a maker’s mark from a photo?

A photo app can narrow the field quickly. It compares the shape, lettering and layout of the mark against known patterns and tells you which family you are probably in — a British assay group, a printed factory backstamp, a registry mark, a retailer’s stamp — which is usually the step that takes a beginner longest. What it cannot do is confirm that the mark belongs to the object it is on, because that judgement needs the mark’s relationship to wear, glaze and construction, not its outline. Treat the result as a lead to check against a register.

How do I identify a pottery or porcelain mark I don’t recognise?

Photograph it square-on and decide first how it was applied: impressed into the clay, incised, painted, or printed. That single fact narrows the period before any name is read. Then compare against a specialist reference such as Kovels’ marks dictionary or Geoffrey Godden’s Encyclopaedia of British Pottery and Porcelain Marks, and check the body and glaze against what the mark claims. If the mark is a diamond or a registration number rather than a name, the date comes from the registry rather than the factory.

What do British silver hallmarks tell you?

A full group carries the sponsor who registered the mark, the standard the metal met, the office that tested it, the assay year and, between 1784 and 1890, that duty had been paid. Together those let you place a piece in a narrow window and trace it to a registered business. They do not tell you that the piece is entirely of that date: because punches can be transposed, a genuine group can sit on a later object, so the hallmark is read as one piece of evidence alongside the form, the construction and the wear.

Can a fake mark fool a professional appraiser?

Sophisticated work occasionally passes a first look. What catches most of it is not a better eye for the mark but the refusal to read the mark alone: magnification, ultraviolet light, X-ray fluorescence for metal content and provenance research are used together, and credentialed appraisers such as those listed by the International Society of Appraisers are trained to weigh them against each other. The common tells are a mark sitting above existing wear, letterforms that belong to the wrong period, and metal around a punch that shows no displacement.

Does a mark always increase an antique’s value?

No. A mark helps when it can be tied to a known maker and a known period, because it turns a guess into an attribution — and buyers pay for the certainty, not for the stamp. It does nothing when the mark is a retailer’s, an importer’s or an unrecorded workshop’s, and it actively hurts when it is wrong for the object, because a mark that does not fit makes the whole piece suspect. The size of any premium is a matter for comparable sales records rather than a rule.

What does “Made in England” versus just “England” tell me about dating?

Less than the shorthand suggests. US law required imported goods to name their country of origin in English, but no statute this guide can point to switched the required wording from “England” to “Made in England” on a set date, and factories changed backstamps when they replaced a plate. Treat the wording as a weak indicator that sits alongside the pattern, the body and any registry mark, rather than as a boundary between two decades.

What to take away

A mark is the fastest route into an object’s history and the least reliable single thing on it. It names a business that took responsibility for the piece, it often gives a usable date range, and on British precious metal it gives a legally tested statement about the metal. It does not, on its own, tell you that the object under it is genuine — and because it is the smallest thing to copy, it is the first thing that gets copied.

So read the mark, then read against it. Ask whether the form, the body, the construction and the wear agree with what the mark claims, take the mark to the register rather than the symbol chart, and be willing to end with a range and an open question rather than a name. That is what the people who keep these records do, and it is the habit that keeps a mistake from becoming a purchase.