English Silver Hallmarks: How to Identify and Date a Piece

A silver spoon, a christening mug, a teapot and a pocket watch case are all marked the same way, by the same offices, under the same Acts. If you can read one you can read all of them. The marks are not a maker’s decoration: they are the residue of a compulsory test, struck by a public office that had assayed the metal and found it up to standard.

Nearly every guide prints a single national chart of date letters. That chart is the reason so many pieces get dated wrongly, because there was never one national cycle. There were eleven separate offices running eleven separate alphabets, changing them on different days of the year, dropping different letters, and occasionally not marking anything at all. This page holds 1,822 lettered years across 86 cycles and 11 offices, 1559 to 1924, transcribed from Sir Charles James Jackson’s tables of 1921 — and a lookup that answers all 11 at once.

Start With the Town Mark, Never the Letter

The date letter is the last mark you should read, not the first. On its own it is close to worthless: the same letter in the same style was struck by different offices in wholly different centuries. What narrows a piece to one alphabet is the town mark — the symbol of the office that tested it. Find that, and the letter becomes a date.

The 11 offices covered here, with the span each one’s tables actually reach. Coverage ends where Jackson’s 1921 record ends, not where the office closed.

Office	Town mark	Covered here	Lettered years	Gaps in the record
London	A leopard’s head — crowned to 1821, uncrowned after	1678–1919	242	continuous
Birmingham	An anchor	1773–1924	152	continuous
Sheffield	A crown; the York rose on gold from 1904	1824–1920	97	continuous
Chester	Three wheatsheaves with a sword, from 1779	1695–1921	222	1698–1700
Exeter	A triple-towered castle; a crowned X before 1701	1701–1882	182	continuous
Newcastle	Three separate castles	1721–1883	163	continuous
York	A cross charged with five lions, from 1700	1559–1856	231	1714–1775
Norwich	A castle over a lion	1565–1697	50	1585–1623; 1644–1687
Edinburgh	A three-towered castle	1755–1921	167	continuous
Glasgow	A tree, a fish and a bell	1681–1921	128	1706–1818
Dublin	A crowned harp — the standard mark, not a town mark	1638–1920	188	1678–1772

Follow the office link and you get that office’s own two-way lookup, its plates photographed from Jackson, and the history that explains why its tables look the way they do. Use the tool below first if you are not yet sure which office you are holding.

Three Offices Used a Castle, and They Are Not Alike

The commonest misreading in provincial silver is a castle. Three offices used one, and a chart that prints them at thumbnail size will not separate them:

• Edinburgh struck one castle with three towers, and had done since 1485.
• Newcastle struck three separate castles in a shield, the arms of Newcastle upon Tyne.
• Exeter, from 1701, struck a triple-towered castle of its own — the arms of that city. Before 1701 it used a crowned letter X instead.

Look Up a Year Across All Eleven Offices at Once

Every other date-letter tool on the web asks you to pick an office first. This one does the opposite: give it a year and it shows you what every office was striking that year, so you can match the letter you are actually looking at. Give it a letter and it shows you how many different years, in how many different towns, could have produced it.

The lookup reads the same 11 data files the individual office pages read, so it cannot disagree with them. Where an office has no row for a year, it says so rather than guessing.

Why the Same Letter Means Eleven Different Things

Run a common letter through the tool above and the result is usually somewhere between forty and seventy candidate years. That is not a defect in the data; it is the actual state of the evidence, and it is what a single national chart conceals.

Three habits make it worse than simple repetition would:

• The offices disagreed about which letter to leave out. The rule everyone quotes is that cycles omit J. Birmingham ran two cycles that used the full twenty-six letters, and its 1824–1848 cycle did the opposite of the convention entirely: it dropped I and kept J. Chester’s 1726–1750 cycle does the same.
• Letter forms came round again. Chester’s cursive cycle of 1726–1750 was reused almost unchanged in 1901–1925. Only the other marks separate them — the later cycle carries no crowned leopard’s head.
• Some cycles printed both cases at once. Seven rows of one Chester cycle read B or b, D or d, I or i. In that cycle the case of the letter dates nothing at all.

The practical method is therefore the reverse of what the letter invites you to do. Read the town mark to fix the office; read the standard mark and the duty mark to bracket the century; then use the letter, and check its shape against the photographed plate on that office’s page rather than against a modern redrawing.

The Assay Year Is Not the Calendar Year

This is the error that quietly costs a year on a great many pieces. An office’s letter did not change on 1 January. It changed on the day that office elected its wardens or held its annual meeting, and every office picked a different day. A piece assayed in February wears the letter of the year that began the previous summer — which is why the tables print 1875–6 rather than 1875.

Changeover dates recorded in these tables. 6 of the 11 offices; for the remaining 5 the tables carry no changeover date, and none is asserted here.

Office	Letter changed	Detail
London	End of May	Britten gives 30 May; Jackson dates the older assay years from 29 May.
Birmingham	July	The Guardians met annually in July. The change moved to 1 January only in 1975.
Chester	July, then 5 August, then 1 July	July to 1839, 5 August to 1889, 1 July afterwards.
Exeter	After 7 August	The letter was appointed at the first Hall meeting after the wardens were elected on 7 August.
York	29 July, from 1585	Before 1585 the searchers were elected in January and the letter is taken to have changed with them.
Norwich	Michaelmas, 29 September	Each letter therefore served three months of one year and nine of the next.

The Standard Mark Is Not a Lion Everywhere

English sterling carries the lion passant, a lion walking with its right forepaw raised. Between 1697 and 1720 that lion was replaced by Britannia and a lion’s head erased, when the sterling standard was raised to stop silver coin being melted for plate. Both marks appear in the tables for London, York and Exeter.

North and west of that, the standard mark changes entirely:

• Edinburgh used no symbol at all before 1759 — the standard was guaranteed by a named person, the deacon, and from 1681 the assay master, whose own mark stood for it. The thistle arrives in 1759.
• Glasgow had no prescribed standard mark before 1819 and struck the letter S, sometimes O, instead. The lion rampant was prescribed from 1819 and a thistle added alongside it in 1914.
• Dublin has never had a separate town mark. The crowned harp is its standard mark, doing both jobs. Hibernia, added in 1730 and often mistaken for the city’s emblem, is a tax stamp for a duty on plate.
• Norwich used a crowned rose from 1581, the assay master’s oath binding him not to strike it on anything below standard.

The sovereign’s head duty mark is the most useful single mark on the piece for narrowing a date quickly, because it was struck only while the duty existed: from 1784 until its repeal in 1890. No duty mark on a piece that should carry one is as informative as the letter itself, and the reverse traps a good many reproductions.

Where a Hallmark Was Never Struck at All

The most useful thing these tables record is where the marking stops being a hallmark. Three cases matter, and no summary chart shows any of them:

• Norwich, for most of its history, had no working assay office. Jackson is explicit that the Norwich marks of those periods were in no sense hall marks but merely marks struck by the makers. A castle-over-lion is not evidence that anything was tested. That is why Norwich contributes only 50 lettered years to the table above out of more than a century and a half of marked silver.
• Exeter before 1701 had no statutory existence, and Jackson finds no evidence of a regular assay by cupel there before that year. A pre-1701 Exeter mark records who made a piece, not that anyone tested it. He believed Exeter used date letters earlier and argued the case — then deliberately refused to set them out in a column, so that his contention should not appear to gain support from any factitious aid.
• York’s town mark is frequently missing. Jackson’s own tables leave the York column blank row after row, so a genuine York piece may carry no York mark at all.

A chart that fills those spaces in has invented them. Where the record is silent, these pages say so — which is less satisfying than a complete grid, and considerably more useful when the question is whether a piece is what it claims to be.

Further Reading

• How to read English hallmarks — the step-by-step method with worked examples, mark by mark.
• Hallmarks on pocket watch cases — the same marks as they appear on cased watch work, with the gold and silver standards in full.
• London hallmarks — the largest set of tables here, 1678–1919, and the office every other one was measured against.
• Chester hallmarks — the office the Liverpool watch-case trade used, and the longest provincial record on the site.
• Makers’ marks — the fourth mark, and the one that names a person.
• Pocket watch cases — materials, construction and how to tell silver from plate.

All date-letter tables on this site are transcribed from Sir Charles James Jackson, English Goldsmiths and Their Marks, 2nd edn, Macmillan, 1921 — public domain. Each office page reproduces the plates the transcription was made from, so any answer given here can be checked against the printed punch rather than taken on trust.