Hallmarks & markings

Swiss assay marks have changed over time — the St Bernard dog head and the Helvetia head belong to different periods and metals. A mark inconsistent with the watch's claimed period is worth researching before it is worth reporting.

End links are reference-specific in fit, so a code inconsistent with the case is a real observation — but it is an observation about correctness of fitment, not about authenticity. Documentation for these codes is thin and largely community-compiled; the catalog records them at correspondingly limited confidence.

Treat exactly as the Tudor naval engravings are treated, and for the same reasons: an exterior engraving on an accessible surface that multiplies a watch's value, which no manufacturer can confirm because no manufacturer applied it. Three things should agree before an issue claim is repeated — the engraving's cutting method and letterform consistent with service practice, the watch's configuration consistent with the claimed issue period, and a documentary chain independent of the watch. Absent all three the correct output is unverified, not genuine and not fake. Note also that the stores reference denotes an equipment CLASS rather than a manufacturer: 6B/346 watches were supplied by several makers, so the marking does not by itself say the watch should be an IWC.

The mistake this record exists to prevent is the same one Cartier invites: a maker's mark that is not Patek's is normal and expected on a vintage watch, and reading it as evidence of a replaced or non-original case is exactly backwards. What the mark is genuinely useful for is consistency — a case whose maker's mark belongs to a workshop that had ceased trading before the movement was made is an inconsistency worth pursuing. Attributions rest on reading poinçons against hallmark registers rather than on any Patek documentation, so they are held at `probable`.

Three, and they matter more than the marking itself. It asserts a CONSTRUCTION, so the check is to look for the shock mount and the thick crystal inside rather than to compare the engraving against reference images — a turtle is trivial to reproduce and a shock-mounted movement is not. It postdates the DS concept: watches from 1959 and most of 1960 may be genuinely DS with no turtle, so its absence on an early example proves nothing. And it sits on a caseback, which can be swapped — where the turtle and the construction disagree, suspect the caseback before the watch.

It identifies the CASEBACK, and Rolex casebacks are separable, interchangeable between references of the same case family, and replaced at service. A period stamp inconsistent with the case serial is a caseback question first — most often a service replacement, occasionally a marriage, rarely anything worse. The quarter-and-year reading is collector-reconstructed and has never been published by Rolex, so it narrows a period and does not date a watch.

The stamp identifies the CASEBACK. A correct inside stamp on a caseback that was fitted at a service in 1985 is a correct stamp on a part with no original relationship to the watch, and the marking cannot distinguish the two cases. Treat it as one input beside the movement serial and the dial, never as the answer.

This is the mark that most often produces a false finding on Cartier, and the reason is the brand's history: Cartier was a jeweller that commissioned cases, so a maker's mark that is not Cartier's is normal and expected on a vintage watch. Reading it as evidence of a replaced or non-original case is exactly backwards. What the mark is genuinely useful for is consistency: the same watch should not carry marks from makers separated by decades.

It certifies nothing. It is reproduced on counterfeit casebacks and it appears on genuine casebacks fitted to the wrong watch. Its practical value is negative rather than positive: a dive reference whose caseback lacks it is worth asking about, but its presence establishes only that the caseback came from a Seiko diver.

The codes exist and are genuine; the mapping from code to year is collector-reconstructed and has never been published by Rolex. Use a clasp code to narrow a period and to check internal consistency with the case — never to assert a date. Bracelets are also the most commonly replaced part on a vintage watch, so a clasp code out of step with the case is a bracelet question before it is a watch question.

A claim to be tested, not evidence in itself — and the test is trivially available. A watch marked Accutron either hums and sweeps without increments or it does not contain a tuning-fork movement. Note that a dead watch cannot be tested this way, and a great many surviving Accutrons are dead: get it powered before concluding anything from the absence of a hum.

Widely presented as a test and it is not one. Counterfeiters have reproduced the signature for years, so its presence is consistent with a genuine dial and with a good fake; only its quality carries information, and only against the reference's expected execution. Two further limits: not every Cartier reference carries one, so absence must be checked before it is reported, and a Cartier service dial carries a correct signature while having no original relationship to the watch.

The single most important thing to know before applying any rule to a vintage Cartier: a rule derived from Paris practice will produce false findings on a London watch. Establish the house first. The signature also carries commercial weight — London-signed watches in particular command premiums — which makes it a target for redialling, so the signature must be assessed as printing rather than accepted as text.

The text is on the dial and the escapement is in the movement, so the two can disagree — most often on a watch whose movement was replaced, occasionally on a marriage of a Co-Axial dial to a non-Co-Axial watch. Where a display back allows it, read the escapement rather than the dial. Where it does not, treat the dial text as a claim to be checked against the reference's catalogued caliber.

This is the distinction between an SKX007 and a 5 Sports SRPD that a depth figure alone conceals: 200m with DIVER'S against 100m without it. Listings routinely describe both as 'divers'. On a condition or specification report the marking should be transcribed exactly as it appears on the dial rather than paraphrased.

The mark identifies the manufacturing entity, not a quality grade — Daini is not better or worse than Suwa, and listings that present one as superior are inventing a hierarchy. Seiko phased the marks out as the companies were renamed to Seiko Instruments and Seiko Epson, so a later watch carrying no factory mark is entirely normal. Because the mark appears on separable parts, a Daini dial in a Suwa-marked case is a parts question rather than a factory finding.

The most legible chronological boundary on any brand in this catalog — and the one most likely to be over-read, because it dates a DIAL rather than a watch. A TAG Heuer signature on a demonstrably 1960s watch is usually a service dial, which is a genuine part correctly fitted and an originality finding rather than an authenticity one. A plain Heuer signature on a modern reissue is correct. Establish the case and movement period first, then read the dial against it — never the reverse.

Widely misdescribed as a definitive test, and it is not one. Counterfeiters reproduce it, and the better reproductions are convincing under the magnification most examiners actually use. Three specific limits. It is in the CRYSTAL, so a crystal replaced at service carries whatever the replacement carries. It postdates the early 2000s, so its absence on an older watch is expected rather than suspicious. And a crude, oversized or obtrusive etching is a real finding against the crystal — which is a different and much weaker claim than a finding against the watch.

Each certified watch has a test result retrievable from METAS against its number, which makes this one of very few dial claims that can be checked against an outside register rather than against the manufacturer. Do that rather than accepting the dial. Note also that certification attaches to the watch as tested at the factory: a watch serviced with non-original parts still carries the printed text.

Frequently conflated with Master Chronometer, including by sellers. The two are different claims: Master Co-Axial is Omega's own specification, Master Chronometer is a METAS-certified standard tested by an outside body. A watch whose dial says one and whose paperwork says the other is inconsistent and the discrepancy should be resolved rather than smoothed over.

The Oystersteel naming is a marketing change and not a material one, so its absence from an older steel Rolex says nothing about the case. Where a case is represented as gold or platinum, the fineness mark is worth locating and the metal is worth testing non-destructively — plated counterfeits of precious-metal references are common, and the mark is easier to reproduce than the alloy.

Depth text is among the most frequently mis-set details on counterfeit dials — wrong font weight, wrong spacing, wrong unit conversion. Check the metric and imperial figures against each other: they must be a correct conversion, and a mismatched pair is conclusive. Never advise a client that a vintage rating still holds without a pressure test.

The text's most useful property is that it must agree with the movement. A dial claiming chronometer certification over a caliber that was never chronometer-grade — a 1520 in a 5513, most commonly — is an inconsistency that needs resolving, and the likeliest explanation is a service dial rather than a fake. Note also that the wording and line count changed over the decades, which collector research uses to place a dial in a period; that ordering is reconstructed rather than published, so use it to raise a question rather than to date a watch.

The legal test behind the words tightened in 2017 to require 60% of manufacturing cost in Switzerland. The designation is a statement about manufacture, not about authenticity, and it appears on counterfeits as readily as the watches they copy.

As with the Daini mark: it records who made the watch, not how good it is, and its absence on later production is expected rather than suspicious. On King Seiko the factory expectation is not uniform — 44KS and 45KS are Daini products while the 56KS is a Suwa one — so an examiner applying a single rule across the line will flag correct watches.

A powerful dating aid and a heavily reproduced one — counterfeit and redialled watches carry it routinely. Read it as evidence about the DIAL, and only then as evidence about the watch: a period-correct designation on a replacement dial says nothing about the case it sits in. The end year is approximate and varied by manufacturer.

The transition to 'Resist' is placed around 1970 but was neither instantaneous nor uniform across markets — North American 6139 examples are reported as 'Resist' from the start, so the transition cannot be applied as a single global date. Treat the wording as placing a watch broadly before or after the turn of the decade, not as dating it.

The single most useful check this mark supports is internal agreement: dial wording and caseback wording should tell the same story. A 'Proof' dial in a 'Resist' caseback means one of the two has been replaced — most often a service dial, occasionally a watch assembled from two donors. That is an originality finding to be recorded and pursued, not a verdict.

This marking records the origin declaration applied for a destination market's import requirements. Specialists disagree about where SKX watches and their movements were actually assembled, and the marking does not settle it — see the research gap on J and K designations. State what the marking is; do not state a country of manufacture on the strength of it.

14-carat cases were made chiefly for markets that preferred the harder alloy, so the mark carries a regional implication worth following up rather than treating as anomalous.

A fineness mark states what the metal is claimed to be. It is struck by the assay office or an authorised maker, and it is reproducible — a mark alone is not a metallurgical test. Where the metal itself is in question, the mark is a starting point for an assay, not a substitute for one.

Its evidential force runs one way. A correct eagle's head on a French-market gold Cartier is expected and is not evidence of authenticity — hallmarks are struck on cases, and a genuine case can be re-used. Its ABSENCE on a case that should carry it is the finding: either the watch was assayed elsewhere, the case is not what it claims to be, or the mark has been polished away, and those three have very different consequences. Note that a case assayed in another country carries that country's marks instead, so absence must always be read against where the watch was sold.

Identifies a SPECIFICATION and not a manufacturer. Hamilton, Benrus, Waltham and others built to it, so the marking does not establish who made the watch — the movement does. Reproduction casebacks carrying this marking are made; period stamping is shallow and slightly uneven where reproductions are uniform and often laser-etched.

The revision letter is not decoration and it is datable. Two changes make it a cross-check rather than a label: Rev B (1975) introduced the radiation trefoil and H3 on the dial and 'Dispose Rad. Waste' on the caseback, so those markings cannot appear on an earlier watch; and Rev E (1989) replaced tritium paint with tritium vials, which is physically obvious. Read the TYPE number alongside the revision — Types 1 and 2 are mechanical, Types 3 to 5 quartz — because the revision governs the era and markings while the Type governs the movement. As with GG-W-113 this is a multi-contractor specification and says nothing about the manufacturer: Westclox, Benrus, Hamilton, Stocker & Yale and Marathon/Gallet all appear. These watches were built not to be serviced, so a rebuilt example with mixed parts is an ordinary outcome rather than a finding.

The value of the decode is that it makes the movement marking CHECKABLE AGAINST THE WATCH rather than merely identifying. A caliber code containing C promises a sweep centre seconds, so a sub-seconds dial above it is an inconsistency; a code without D should not be driving a date aperture. That turns a movement marking into a cross-check on the dial and hands, which is exactly what an assembled or re-dialled watch fails. Two limits: the letter meanings apply only from the third character onward, and the jewel count is NOT encoded — the familiar '23 jewels' marking is separate from the caliber code and must be read on its own.

The declaration describes the movement as it left the factory. It survives servicing, re-regulation and even movement swaps, so it is a claim about that movement's original specification and nothing more.

Read the ENGRAVING first and the rest second. Chatons and blued screws can be reproduced by anyone willing to spend the money; freehand engraving cannot, because there is no fixed pattern to copy and the skill is scarce. Machine-cut or stamped engraving on a Lange is conclusive against the movement. Note also what the finishing does not establish: it is a property of the movement, and a movement can be moved.

The single most important marking on any Patek, because it is the one the archive can answer. Read it off the movement and check it against the extract rather than accepting an extract that names it — the direction matters, and a document produced first will encourage an examiner to see the number it expects. Note that the pairing is what detects a marriage: two genuine halves from two different watches will each check out individually and will not check out together.

This is the single most informative marking on the watch and the hardest for a counterfeiter to fake convincingly, because it requires a movement of the right architecture before any marking matters. Read the architecture first and the signature second: a correct-looking signature stamped on a movement of the wrong layout is conclusive, whereas an unfamiliar signature on a correct movement is more often a service part or a period variation. The serial found here dates the MOVEMENT, and on a serviced watch that is not the same thing as dating the watch.

It is the manufacturer certifying its own work, where the Geneva Seal was the Canton certifying it — a broader standard and a less independent one, and the distinction is worth being precise about with a client who asks what it guarantees. For an examiner its most useful property is chronological: Patek stopped applying the Geneva Seal when it introduced this, so a Patek Seal on a watch that predates 2009 is an anachronism, and a Geneva Seal on one made well after it is a question. Neither observation authenticates a watch on its own, and both point at the movement rather than at the case it sits in.

A finishing and origin certification, not a general quality ranking, and not held by many excellent movements made outside Geneva. Its absence proves nothing.

A claim about origin that the movement can confirm or refute, and the movement is visible: Villeret references carry display backs and Minerva calibers are unmistakable. Because the price difference between the two productions is an order of magnitude, this marking is worth more to a counterfeiter than any other on the brand — so treat it as the thing to be verified, never as the verification. The check is the arrow-shaped chronograph bridge and the hand finishing behind the glass, not the word on the dial. A useful preliminary before the watch is in hand: the caliber designation appears to encode the same distinction, with Villeret movements carrying an M — MB M16.29, MB M13.21 — where Le Locle calibers do not, as with the Sellita-based MB 24.17.

Certifies the individual watch that was tested, not the model and not the manufacturer — the same distinction that applies to every observatory certification in this catalog, and the one marketing most often blurs. Its presence is a strong positive where the strike is period-correct and consistent with the watch; its absence means only that this watch was not submitted. Note also that Besançon certification is available to any manufacturer, so the mark is not specific to Yema or to French watches.

Platinum and white gold are difficult to distinguish by eye and easy to confuse in a listing. The fineness mark is the first check; weight is the second, platinum being appreciably denser.

Relevant because platinum Cartier cases exist across the Tank line and are among the most valuable, which makes the metal claim worth verifying rather than accepting. A case represented as platinum without a platinum mark, in a market where the mark would be compulsory, is a metal question before it is an authenticity question — and metal is testable non-destructively.