Harold's Arrow and the Coal-Tar Dyes That Dated the Bayeux Tapestry's Repairs
Table of Contents
📋 What You'll Learn
This guide walks you through harold's arrow and the coal-tar dyes that dated the bayeux tapestry's repairs with detailed instructions.
The most famous injury in English history may have been stitched in about 800 years after the battle. The best evidence for when the repairs were made doesn’t come from a historian. It comes from two dyes that didn’t exist in 1066.
On 10 September 2026 the Bayeux Tapestry opened at the British Museum, on loan from Normandy until 11 July 2027. It is the first time the embroidery has been shown in Britain since it was made. Public tickets went on sale on 1 July, it became the fastest-selling exhibition in the museum’s history, and every date to the end of the year is sold out. Most of the coverage has been about the history. This article is about the chemistry: how analysis of the dyes found a Victorian-era repair crew in the threads, why that matters for the arrow, and why two of the three original colours only stay put because of a metal salt.
We sell that metal salt, aluminum sulfate, and say so at the bottom. Everything above that point is sourced to the researchers, the museum and the published literature, and every source is linked in the reference list.
What the chemists found in the threads
Short answer: the 19th-century repairers used brand-new synthetic dyes, and those dyes can be dated. Clarisse Chavanne did the work for her PhD, supervised by Philippe Walter at the Laboratory of Molecular and Structural Archaeology at Sorbonne University. She analysed yarns from the tapestry spectroscopically, and Chemistry World set out the results in a feature on 7 September 2026. The original threads carry the three dye families you would expect from 11th-century England: madders, “a group of red dyes from anthraquinone compounds”; welds, “giving yellows from flavonoid compounds”; and indigos or woads for the blues.
The repair threads were different. Two synthetic dyes turned up that no medieval dyer could have had:
| Dye found in repair threads | Origin | Earliest possible use | What it tells you |
|---|---|---|---|
| Aurin | Found by Friedrich Ferdinand Runge while distilling coal tar, 1834 | After 1834 | The repair is 19th century or later |
| Methyl violet (methylaniline violet, “Violet de Paris”) | First made by the French chemist Charles Lauth; sold commercially from 1866 | 1866 | Sets the earliest date for the repair |
| Photographs of the tapestry | The repaired areas already appear in photographs taken in 1872 | — | Sets the latest date: 1872 |
Put the two bounds together and you get a restoration between 1866 and 1872. This is standard dating by first appearance, the same logic that dates a painting by its titanium white. A material can’t turn up in a repair before the year it was first sold, so a dye whose first sale date is known works like a timestamp. The same analysis shows how the repairers worked: Chavanne found both natural indigo and a synthetic rosaniline blue, suggesting they simply took whichever dyed yarn matched the faded original best. They were matching colours, not following the recipe.
The one-sentence version. The repair threads contain a dye that didn’t go on sale until 1866, and the repairs appear in photographs by 1872, so the repair crew was at work in those six years.
The arrow: what the chemistry does and doesn’t show
The scene labelled Harold Rex interfectus est, “King Harold is killed”, shows a figure gripping a shaft near his eye. It is the source of the schoolbook story that Harold II died with an arrow in the eye at Hastings in 1066. Doubts about it are old. The earliest drawings of the tapestry, from the early 18th century, show the figure holding something that reads more like a spear or lance. There are also needle holes in the linen where stitching was lost and later replaced.
Engraved for Montfaucon, 1729–30
The tapestry today
The new dye work adds weight to the theory that the arrow was not in the original tapestry. Here is exactly how far the evidence goes. Chemistry World reports that the yarn in the arrow “seems to be synthetic” but is so faded that it is unlikely to be identified by spectroscopy. So nobody has chemically fingerprinted the arrow itself. What the chemistry did was date the restoration campaign that the arrow’s critics have long blamed. Together with the marks on the reverse and the holes visible in the earliest drawings, the case is now much harder to wave away.
Not everyone goes as far. The Art Newspaper’s summary for the opening says the pointed object “was probably shifted slightly” during the 19th-century restoration, and notes that specialists still disagree about whether the figure is Harold at all. Some read the scene as two events, one man hit by an arrow and another cut down by a horseman. So the honest summary is this: the curator of the exhibition you can now visit says the arrow was added, the chemistry puts the restoration in 1866–1872, and the arrow’s own thread is too faded to test directly.
There is a chemistry joke inside the history. The arrow story may have been cemented by aniline chemistry: the same coal-tar dye industry that, in exactly those years, was making the old plant dyes obsolete. In 1868 the German chemists Carl Graebe and Carl Liebermann worked out how to make alizarin, the main red compound in madder root, from anthracene, a coal-tar hydrocarbon. It was the first natural dye ever made synthetically, and the madder-growing trade collapsed within a few years. The tapestry was being repaired in the middle of that switch.
Why madder and weld needed a metal to stay on the wool
Short answer: madder and weld are mordant dyes, which means they hold on wool through a metal ion, historically aluminium, that links the dye molecule to the fibre. Without the metal, most of the colour rinses out.
A mordant, from the French mordre, “to bite”, is a metal salt the fibre is treated with before or during dyeing. In water, aluminium sulfate or potash alum releases aluminium ions, Al3+. Wool is a protein, and its carboxylate and amine groups bind those ions. The dye molecules then coordinate to the same aluminium centres. The result is an insoluble metal–dye complex called a lake, sitting in and on the fibre. It resists washing far better than the dye alone.
The two medieval dyes that behave this way share a structural feature: hydroxyl and carbonyl groups sitting next to each other, which is the shape a metal ion likes to grab.
| Original dye | Main colour compound | Chemical class | How it holds on wool | Colour |
|---|---|---|---|---|
| Madder (Rubia tinctorum root) | Alizarin, C14H8O4 | Dihydroxyanthraquinone | Mordant dye: forms an aluminium lake | Reds; the mordant shifts it anywhere from pink through purple to brown |
| Weld (Reseda luteola) | Luteolin | Flavonoid | Mordant dye: forms an aluminium complex | Bright yellow; dyed over blue it makes green |
| Indigo or woad | Indigotin | Indigoid | Vat dye: reduced to a soluble form, soaked in, then re-oxidised in air inside the fibre. No mordant. | Blues; over weld, greens |
That split explains the tapestry’s whole palette. Ten colours from three plants is possible because a dyer can vary the mordant strength, the dye strength and the number of dips, and can over-dye one colour on another. Weld on top of woad gives greens. A strong madder bath on well-mordanted wool gives a deep red; a weak one gives pink. The colours differ in chemistry as well as shade. It is also why the hyperspectral reconstruction in the Sorbonne work could rebuild the original colours: the fading of anthraquinone reds and flavonoid yellows follows known chemistry.
Why the yellows faded worst. Flavonoid yellows like luteolin are the least lightfast of the three families. The Chemistry World feature notes that the reconstructed original colours closely match a hand-coloured engraving from 1820, which suggests most of the fading happened after that date, not in the Middle Ages.
The English copy used the old dyes on purpose
There is a second 19th-century tapestry in this story, and it makes the opposite choice. In 1885 Elizabeth Wardle and a team of 35 women from the Leek Embroidery Society in Staffordshire embroidered a full-size copy, now at Reading Museum. Her husband, the silk dyer Thomas Wardle, supplied the worsted wool, and according to Reading Museum he dyed it with vegetable dyes rather than chemical ones. Within about 15 years of each other, the French repairers were putting new coal-tar synthetics into the original while the English copyists were deliberately using the medieval kind.
The copy has its own famous alteration. A naked man in the border of the original wears a pair of shorts in the Reading version. The museum points out that the Leek embroiderers didn’t add them out of modesty. They traced from a set of photographs that staff at the South Kensington Museum had already “cleaned up”. So both 19th-century interventions came from other people, one working in dye and the other in photographs.
What this means if you dye wool now
The chemistry hasn’t changed in 950 years. Natural dyers today still pre-treat protein fibres (wool, silk, alpaca) with an aluminium salt before a madder or weld bath. Two aluminium salts are in common use:
| Salt | Formula | What it contributes | Notes |
|---|---|---|---|
| Potassium aluminium sulfate (potash alum) | KAl(SO4)2·12H2O | Al3+; the potassium and sulfate are spectators | The traditional “dyer’s alum” of the medieval trade |
| Aluminium sulfate | Al2(SO4)3·xH2O | Al3+, with more aluminium per gram than potash alum | Widely recommended by natural-dye suppliers for protein fibres |
The working quantity is set as a percentage of the dry weight of fibre (WOF). Botanical Colors, a natural-dye supplier, recommends aluminium sulfate at 12% WOF for animal fibres, up to 20% for deeper shades. That is about a scant tablespoon per 100 g of dry wool. More mordant generally gives deeper colour, especially with reds such as madder. Follow your dye supplier’s recipe for times and temperatures; this article is about why it works, not a step-by-step method.
Aluminum sulfate solutions are acidic and irritating to the eyes. Read the safety data sheet, which is linked from the product page, before you open the bag. Keep dye pots and utensils for dyeing only, never for food.
Common questions
Was the arrow in Harold’s eye added to the Bayeux Tapestry later?
The evidence increasingly points that way. The British Museum’s curator, Prof Michael Lewis, told The Telegraph in September 2026 that the arrow “has definitely been added” during 19th-century restoration. Sorbonne dye analysis dates the main restoration to 1866–1872. The arrow’s own yarn is too faded to identify by spectroscopy, and some specialists still describe it as shifted rather than added.
How did scientists date the Bayeux Tapestry restoration?
By finding synthetic dyes in the repair threads. Methyl violet, one of them, was only sold commercially from 1866, and photographs taken in 1872 already show the repairs. A dye can’t appear in a repair before it existed, so the restoration falls between 1866 and 1872.
What dyes were used on the original Bayeux Tapestry?
Three plant dye families: madder for reds, weld for yellows, and indigo or woad for blues. Over-dyeing and different mordant strengths turned those three into about ten wool colours, stitched onto a bleached linen ground.
What is a mordant in dyeing?
A metal salt, most often an aluminium salt, that binds a dye to a fibre. The metal ion attaches to the fibre and to the dye molecule, forming an insoluble complex called a lake that resists washing. Madder and weld need a mordant; indigo does not, because it is a vat dye.
Can you use aluminum sulfate as a mordant for wool?
Yes. Aluminum sulfate supplies the same aluminium ion as traditional potash alum and is widely recommended by natural-dye suppliers for protein fibres such as wool and silk, typically at about 12% of the dry fibre weight and up to 20% for deeper shades.
When can you see the Bayeux Tapestry at the British Museum?
From 10 September 2026 to 11 July 2027. It is the fastest-selling exhibition in the museum’s history, and tickets for dates to 31 December 2026 sold out ahead of the opening.
References
- The chemistry helping to unravel the Bayeux Tapestry’s secrets — Rachel Brazil, Chemistry World, 7 September 2026. Dye identification, the 1866–1872 dating, the arrow yarn, hyperspectral reconstruction.
- Bayeux Tapestry: First use of early synthetic dyes for the restoration of a masterpiece — Chavanne et al., Dyes and Pigments, 2022. The primary dye study.
- Bayeux Tapestry curator on the Harold arrow — The Telegraph via AOL, 3 September 2026. Prof Michael Lewis, British Museum.
- Ten surprising facts about the Bayeux Tapestry — Martin Bailey, The Art Newspaper, 4 September 2026.
- What to expect at the British Museum’s Bayeux Tapestry exhibition — Medievalists.net, September 2026. Dates, ticket demand.
- The history of Britain’s Bayeux Tapestry — Reading Museum. The 1885 Leek copy, Wardle’s vegetable dyes, the photographs.
- Alizarin — structure, mordant colour range, the 1868 Graebe–Liebermann synthesis.
- Weld (Reseda luteola) — luteolin, use in the Bayeux Tapestry, greens with woad.
- How to mordant with aluminum sulfate — Botanical Colors. 12–20% WOF for protein fibres.
- Harold scene after Montfaucon — Wikimedia Commons, public domain. Engraving from Les Monumens de la Monarchie Françoise, reprinted in Ducarel, Anglo-Norman Antiquities, 1767.
- Bayeux Tapestry, scene 57 — Wikimedia Commons, public domain. The image used in this article.
- Aluminum Sulfate: The Ultimate Guide — Alliance Chemical. The evergreen aluminum sulfate guide this article hangs from.
The two we stock
Aluminum Sulfate Hydrate ACS Grade
The dry crystalline salt you weigh against fibre weight. 2 lb jars for a few skeins, 50 lb bags for a studio.
Aluminum Sulfate 50% (liquid alum)
Pre-dissolved for metering pumps in dyehouses, water treatment and paper sizing. Quart to tote.
Dyeing at studio or dyehouse scale?
Tell us the fibre, the batch weight and whether you dose dry or by pump, and we will point you to the form and pack size that fits.
See aluminum sulfate sizesKey numbers and sources
| Number | What it is | Source |
|---|---|---|
| 1866–1872 | Date window of the major restoration, from methyl violet’s commercial launch and the 1872 photographs | Chemistry World; Chavanne et al. |
| 1834 / 1866 | Aurin discovered in coal tar by Runge; methyl violet sold commercially | Chemistry World |
| 10 colours, 3 dye families | Madder, weld, indigo or woad | Chemistry World |
| 1868 | Graebe and Liebermann synthesise alizarin, the red of madder | Alizarin (reference) |
| 1885, 35 women | The Leek full-size copy, worked in vegetable-dyed wool | Reading Museum |
| 12–20% WOF | Aluminum sulfate mordant for protein fibres, by dry weight of fibre | Botanical Colors |
| 10 Sep 2026 – 11 Jul 2027 | British Museum display dates | Medievalists.net |
Frequently Asked Questions
Was the arrow in Harold's eye added to the Bayeux Tapestry later?
The evidence increasingly points that way. The British Museum's curator, Prof Michael Lewis, told The Telegraph in September 2026 that the arrow "has definitely been added" during 19th-century restoration. Sorbonne dye analysis dates the main restoration to 1866–1872. The arrow's own yarn is too faded to identify by spectroscopy, and some specialists still describe it as shifted rather than added.
How did scientists date the Bayeux Tapestry restoration?
By finding synthetic dyes in the repair threads. Methyl violet, one of them, was only sold commercially from 1866, and photographs taken in 1872 already show the repairs. A dye can't appear in a repair before it existed, so the restoration falls between 1866 and 1872.
What dyes were used on the original Bayeux Tapestry?
Three plant dye families: madder for reds, weld for yellows, and indigo or woad for blues. Over-dyeing and different mordant strengths turned those three into about ten wool colours, stitched onto a bleached linen ground.
What is a mordant in dyeing?
A metal salt, most often an aluminium salt, that binds a dye to a fibre. The metal ion attaches to the fibre and to the dye molecule, forming an insoluble complex called a lake that resists washing. Madder and weld need a mordant; indigo does not, because it is a vat dye.
Can you use aluminum sulfate as a mordant for wool?
Yes. Aluminum sulfate supplies the same aluminium ion as traditional potash alum and is widely recommended by natural-dye suppliers for protein fibres such as wool and silk, typically at about 12% of the dry fibre weight and up to 20% for deeper shades.
When can you see the Bayeux Tapestry at the British Museum?
From 10 September 2026 to 11 July 2027. It is the fastest-selling exhibition in the museum's history, and tickets for dates to 31 December 2026 sold out ahead of the opening.
About the Author
