A pound of purple silk was capped at six thousand days’ pay

Which colours were priced or reserved, and where?#

Eight cases. Purple was the colour most often priced out of reach or regulated: by the cost of the dye in Rome, by imperial law in Byzantium, by rank codes in Sui and Tang China and in Japan (where the chronicle records the grades but not their colours) and by statute in Elizabethan England. The last column says only what each source records.

What was a pound of purple worth in Rome?#

The best price list is the Edict on Maximum Prices, issued between 20 November and 10 December 301 by Diocletian and his co-rulers to check inflation. It capped more than 1,200 items, from fodder to lions, in denarii, which by then were a unit of account rather than a coin. These are ceilings, not market prices, and they include the wages of working people. Divide a price by a day’s pay and it becomes a number of days of work.

For a farm labourer at 25 denarii a day, a pound of purple silk equalled 6,000 days of pay, about 16.4 years of work without a day off. Both figures are maximums, so this is a ratio of two ceilings, not a record of what anyone paid or earned. Most of the price was the dye: undyed silk was capped at 12,000, so purple multiplied it by 12.5. And purple was not one price. Bright Tyrian wool, at 16,000 denarii, was capped at about 27 times as much as wool dyed with archil, a purple made from lichens, at 600. The lichen purple was capped at 3.75% as much.

Turning denarii into today’s money is guesswork, and the edict’s translator, Antony Kropff, says how he guesses: pegging the labourer’s 25 denarii to a modern day’s wage, he puts a denarius at around €1.5. On that peg the pound of purple silk comes to about €225,000. The days are the sounder measure, because they need no exchange rate at all.

Why so dear? The dye came from sea snails, a few drops from a gland in each. When the chemist Paul Friedländer set out in 1909 to identify the pigment, he processed the glands of 12,000 Murex brandaris snails and obtained 1.4 grams. That is about 0.12 milligrams a snail, from a modern laboratory, not an ancient dye-works.

The fashions had been moving for centuries. Pliny quotes the historian Cornelius Nepos, who died under Augustus: in his youth violet purple sold at one hundred denarii a pound, then Tarentine red came into favour, and then “the Tyrian dibapha, which could not be bought for even one thousand denarii per pound.” Dibapha was cloth dyed twice. Publius Lentulus Spinther, aedile in the year of Cicero’s consulship (63 BC), was the first to use it for the purple-bordered toga praetexta, “and he was greatly censured for it.” By the time Nepos wrote, he asked, who did not have purple hangings on his banqueting couches? In the translation: “who is there that does not have purple hangings to his banqueting-couches, even?”

Price was not the only guard. In his account of how Nero raised money, Suetonius writes that “having forbidden the use of amethystine or Tyrian purple dyes, he secretly sent a man to sell a few ounces on a market day and then closed the shops of all the dealers.” He adds that Nero is said to have had a matron stripped of her garment and her property for wearing the forbidden colour at one of his recitals.

Was purple worth its weight in gold?#

The denarius of Nepos’s day and the denarius of 301 cannot be set side by side: the first was a silver coin, the second a unit of account. Metal can. Each source in this chart states a price in coins whose weight in gold or silver is known, or, in 301, sets its own price for gold and silver. So each colour can be weighed against metal in its own time, in the metal its own source uses.

Nepos’s thousand denarii were silver coins struck 84 to the Roman pound, which is 11.9 pounds of coin. Allowing, as the dictionary cited for the 84 does, one thirtieth of a late-Republican denarius for alloy, that is about 11.5 pounds of fine silver, and a pound of Tyrian dibapha cost more than that. Three centuries later the edict priced a pound of gold at 72,000 denarii and a pound of refined silver at 6,000. Raw silk dyed purple, at 150,000, was capped at 2.1 times its weight in gold and 25 times its weight in silver. An ounce of Cennini’s best ultramarine cost eight ducats, coins holding about 28 grams of gold: roughly its weight in gold, or two-thirds of it if he meant Padua’s heavy ounce.

The comparison has limits. Gold and silver themselves changed in value, the edict gives ceilings, Nepos gives a memory of his youth, and Cennini prices a pigment, not a cloth. The 84 to the pound comes from Smith’s dictionary of 1875, which dates the first denarius to 269 BC; Crawford’s Roman Republican Coinage (1974), as summarised on Wikipedia, puts it about 211 BC, at first 72 to the pound, with 84 standard later, so 84 holds for Nepos’s youth. What the metal shows is the order of magnitude: the best purple and the best blue were priced like precious metal.

When did purple become the emperor’s alone?#

In the eastern empire, price gave way to monopoly. The Code of Justinian, a sixth-century collection of imperial laws, records how the restrictions grew. A law of Valentinian II, Theodosius I and Arcadius, issued between 383 and 392 (CJ 4.40.1), says that “no private person shall have power to dye or sell purple goods, silk or wool, which is called blatta, oxyblatta and hycinthina”, three named shades of purple, on pain of losing “his property and his head.”

In 424 Theodosius II went further (CJ 11.9.4, 16 January). “All persons of both sexes and of every rank, business, profession or birth must refrain from possessing vestments reserved for the emperor and his house alone,” and every tunic and cloak “dyed through and through with the blood of the purple fish” was to be produced and handed over. The people who gathered the snails were bound to the work as well. In October of the same year (CJ 11.8.11) gatherers who had abandoned their station and taken on the dress of rank were “recalled to the bonds of their own trade and birth,” and in 427 (CJ 11.8.15) their children were declared bound to it too.

The colour reached the emperor’s pen. In 470 the emperor Leo I ordered (CJ 1.23.6) that imperial orders appear only in “purple script”, which the law describes as marked by “the brightness of the cooked purple fish and crushed (triti) shellfish”, and that “the making of this sacred ink shall be permitted or granted to no one,” on pain of confiscation and death. The translator marks his wording as uncertain.

Which colour did a rank allow?#

Further east, colour was tied to rank in writing. Choose a code and a row: the page shows what the source records for it and, where the rows form a ladder, the rows above and below. The first five codes are the Japanese chronicle and the Chinese passages, in order of date; the last is England’s proclamation of 1574, whose rows are overlapping groups.

Choose a code or proclamation, then tap a row.

What happened to yellow under the Sui and early Tang?#

The Old Book of Tang’s treatise on carriages and dress follows yellow through four roles in about sixty years. Under the Sui dynasty, before 605, it says, emperors and noble ministers mostly wore robes of yellow figured twill, and the everyday dress of officials was “the same as that of the common run of people: all wore yellow robes.”

Then Emperor Yang changed it. In the first year of Daye (605) he had caps and dress redesigned, “from the Son of Heaven down to the clerks”, graded by rank, and allowed the fifth rank and above to wear vermilion and purple. In the sixth year (610), for officials accompanying him on distant journeys in military dress, the treatise records a code of colours: purple robes for the fifth rank and above; crimson and green for the sixth rank and below; blue for clerks, white for commoners, black for butchers and merchants and yellow for soldiers. By 610, then, purple stood at the top of the Sui scale and yellow was the soldiers’ colour.

The early Tang kept yellow for the emperor:

In the early Wude years, following the Sui system, the emperor’s informal dress, also called everyday dress, was only a yellow robe and tunic; later reddish-yellow was gradually used, and so officials and commoners were forbidden to use reddish-yellow for clothing or trim.

Old Book of Tang, chapter 45, Treatise on Carriages and Dress; translation by Marius Comper

The ban named a shade, not the whole colour, and the ranks already had colours of their own. An edict of the eighth month of the fourth Wude year (621) gave purple, with jade fittings, to the third rank and above; vermilion, with gold, to the fifth rank and above; and yellow to the sixth rank and, as the clauses on silver and brass fittings show, the ranks below it, with colours “generally yellow” for those outside the ranks and for commoners. So purple and vermilion stood above yellow from 621, not from 630. In the fourth year of Zhenguan (630) the ranks were set in purple, crimson, green and blue, and “although there is an ordinance, wearing yellow is still permitted.” In 662 a request to move the eighth and ninth ranks from blue to blue-green was approved, on the ground that deep blue could be mistaken for purple, with permission to wear yellow as well at court attendance. In 668, the treatise records, wearing yellow was no longer permitted at all; it does not say who is meant, and the sentences around it concern officials’ dress. In 674 the ranks received purple, dark and light red, green and blue, and commoners belts of copper or iron.

The emperor’s colour moved too. The emperor’s table cushions, bed cushions and bed curtains had all been decorated in purple; in 747 a rites official asked that they follow the colour of the imperial robe, reddish-yellow, and the request was granted. In 751 the guards’ banners changed from crimson to reddish-yellow “to match the virtue of earth.”

What the treatise shows is yellow in motion: the common colour of Sui everyday dress, then the soldiers’ colour in the code of 610, then the lowest band of the ranks in 621 and the colour of commoners, while a narrower reddish-yellow was claimed for the emperor and closed to officials and commoners, until in 668 yellow itself was no longer permitted. Purple and vermilion stay on top from 605. The treatise gives no reason for the first ban on reddish-yellow and says nothing about the cost of yellow dyes. Rules under the later Ming and Qing are outside the sources checked here.

What does the Nihon Shoki say about cap colours?#

The Nihon Shoki, in W. G. Aston’s translation, records under the twelfth month of the eleventh year of Empress Suiko (603 in Aston’s dating): “Cap-ranks were first instituted”: greater and lesser grades of virtue, benevolence, propriety, faith, justice and knowledge, “in all twelve grades. Each was made of sarcenet of a special colour.” In 604 the caps were first granted to the ministers, “there being a distinction for each.” The Japanese text says each cap was sewn of silk of its proper colour, 當色. In 611 “the colour of the clothing of the various Ministers agreed with the colour of their official caps,” and their hair ornaments were gold for the two highest grades, leopards’ tails for the next two and birds’ tails for the rest.

That is all the chronicle says about colour: each grade had one, and clothing matched the cap. It does not say which. The mapping comes from elsewhere. Aston’s notes of 1896 say that purple was for officials of the fifth rank and upwards “in imitation of the contemporary Sui dynasty of China,” that nin was green, rai red, shin yellow, gi white and chi black, and that princes and chief ministers wore the cap of the highest rank. Another of his notes says that these cap-ranks “did not remain long in use. They were subsequently modified, and ultimately abandoned altogether.” None of this is in the chronicle’s own text.

Japanese scholarship holds the colours unknown. The Japanese Wikipedia article on the system, which summarises work by Kishi Tetsuo (1982) and Takeda Sachiko (1984) among others, says that no source gives the colour for any grade and that every proposed mapping is unconfirmed. The commonest guess assigns the colours of the five Chinese elements: blue to benevolence, red to propriety, yellow to faith, white to justice and black to knowledge. Purple for the top grades is the long-standing guess, drawn from a later report that the minister Soga no Emishi privately gave his son a purple cap. The article adds that the Soga ministers and the imperial princes are now thought to have stood outside the twelve grades, which contradicts Aston’s note on princes and chief ministers, and that the system is no longer thought to have been modelled on Sui China. This summary is a tertiary source; the monographs behind it were not opened.

So the ladder of twelve colours often printed for this system is a reconstruction, and the picker above marks every colour in it as a reading. Nothing in these sources says the dyes were dear; the status came from the grade.

How much did the best ultramarine cost?#

The best grade was worth eight ducats an ounce, and a painters’ handbook ranked it above every other colour. Cennino Cennini’s handbook for painters, written around 1400, probably in Padua (a supposition in the Enciclopedia dell’Arte Medievale), opens its chapter on the pigment with praise: “Ultramarine blue is a colour noble, beautiful, and perfect beyond all other colours, and there is nothing that could be said of it but it will still exceed this (praise).” It tells the painter that with this colour, “together with gold (which adorns all the works of our art) let everything be resplendent.”

He also gives the price. Describing how the pigment is extracted from the crushed stone in successive washings, he says that “if you have good lapis lazuli, the azure from the first two extracts is worth eight ducats the ounce. The last two extracts are worse than ashes.” A ducat, like the Florentine florin, was a gold coin of about 3.5 grams, so eight of them held about 28 grams of gold. An ounce weighed about 28 grams in Florence and on the light pound of Padua, and about 25 on Venice’s light pound: by those ounces the best blue was worth about its weight in gold. Padua also had a heavy pound, whose ounce of about 40 grams would make it two-thirds of its weight in gold. The same stone yielded a blue worth that much and a blue worth less than ashes, and the painter had to tell them apart. The chapter heading calls it “azzurro oltre marino”, blue from beyond the sea. This source records value in the workshop; it does not record laws on who could wear or commission it.

Why was cochineal worth so much silver?#

The Science History Institute’s account of the cochineal trade (Samantha Nadel, 2025) points to three reasons. The dye, made from insects that live on prickly-pear cactus, was far more concentrated than other insect reds, so artisans got deeper colour from less material. The Spanish Crown held a monopoly through Mexico and Peru. And Europe and Asia would pay: “depending on the year, cochineal could be worth half its weight or more in silver,” and only silver itself consistently brought the Crown more wealth over three centuries of colonial rule.

The trade was guarded. In 1674 Spain required all cochineal shipments to travel on, or under the protection of, warships. In 1600 the chronicler Antonio de Herrera claimed that Tlaxcala grew more cochineal than all the other regions of New Spain combined, and an estimated 150,000 pounds were shipped to Spain that year. The dye was far older than the empire: the oldest Mexican textile known to contain cochineal is about 2,300 years old, and the Tlaxcalteca were using cochineal-based paints in their pictographic books by the 1400s. The colonial state tried to standardise production, appointing a cochineal judge in Puebla in 1572 and ordering sun-drying in 1573. In 1585 four Tlaxcalan noblemen asked Philip II for a stamp certifying their province’s cochineal, and did not get one.

This source documents the value of the dye and the control of its trade, not who was allowed to wear red. Rules on red dress are not covered here.

Who first sold purple made from coal tar?#

William Henry Perkin, who found the dye by accident at 18. In 1856, during the Easter vacation from the Royal College of Chemistry in London, he tried to make quinine from coal-tar chemicals and produced a purple instead. The Science Museum Group records that the Patent Office granted him a patent on 26 August 1856 for “a new colouring matter for dyeing with a lilac or purple colour stuffs of silk, cotton, wool, or other materials.” In 1857 he opened a factory at Greenford Green, near London.

The Science History Institute calls the result mauve, or aniline purple, “the first commercialized synthetic dyestuff.” Friedländer needed 12,000 snails for 1.4 grams of the ancient pigment; Perkin’s purple came out of a factory from coal-tar chemicals. What happened to purple’s standing afterwards is not covered by these sources.

What the cases have in common#

This is a reading across the sources above, not a finding any one of them states.

• Three ways of setting a colour apart, often combined. Price made purple dear in Rome, ultramarine in Cennini’s workshop and cochineal in the Spanish trade. Law backed price or replaced it: Nero’s ban, the Byzantine monopoly, England’s 1574 list for men. Rank codes gave colours to ranks in Sui and Tang China, and in Japan, where the chronicle records the grades and says each cap had a colour.
• Price alone did not decide. The 301 edict caps a lichen purple at under 4% of the Tyrian price. And a colour could move: yellow was the everyday robe of Sui officials, then the soldiers’ colour in 610, then the colour of the lowest ranks and of commoners in 621, while purple and vermilion stayed above it and a reddish-yellow was kept for the emperor.
• What these sources cannot say. They do not show whether a colour made the status or the status made the colour, and they stop at 1856 for purple. For ultramarine and cochineal they record value, not who was allowed to wear them. India, Africa, the Islamic world and pre-Columbian feather and shell colours are not covered.

Sources and method#

Quoted English is the translator’s, as linked; the Chinese passages are translated by Marius Comper. A second reader, OpenAI’s Codex, an AI model, checked the renderings against the Chinese and against parallel texts in the New Book of Tang and the Tang huiyao; where it disagreed, the wording follows its reading or says what is uncertain. Prices are the figures the sources give, in the units they give. The calculations are divisions: a price by a day’s wage; a price by the edict’s price for gold or silver; 1,000 denarii by 84 to the pound, less one thirtieth for alloy; and one multiplication: eight ducats times about 3.5 grams of gold each. The edict’s figure of 150,000 denarii for purple silk was read in Kropff’s translation and matches Crawford’s edition of the Aphrodisias text (purpurae metaxablattae, 150,000). A Roman pound is taken as 327.45 grams, the value Kropff follows from Lauffer. Colour swatches and cloths are drawn to suggest each colour; they are not reconstructions of ancient dyes. The images come from Wikimedia Commons, and each is credited under its figure.

1. Pliny the Elder, Natural History 9.63 (Cornelius Nepos on purple prices; Spinther and the toga praetexta), translated by Bostock and Riley
2. Suetonius, Nero 32 (Nero’s ban on amethystine and Tyrian purple), translated by J. C. Rolfe
3. Edict on Maximum Prices, 301 AD, English translation by Antony Kropff (2016): preface, weights, chapters VII, XXIII, XXIV and XXX (archived copy; the original server’s certificate is invalid)
4. Edict on Maximum Prices: Kropff’s translation as a PDF, University of Bergen (second copy)
5. Michael H. Crawford, Diocletian’s Edict of Maximum Prices at the Civil Basilica in Aphrodisias (Aphrodisias XIII, Reichert, 2023): the Latin text, chapter on purple (purple silk, 150,000)
6. Kenan T. Erim and Joyce Reynolds, “The Aphrodisias copy of Diocletian’s Edict on Maximum Prices”, Journal of Roman Studies 63 (1973), 99–110; the edition of the Latin text behind Kropff is Siegfried Lauffer, Diokletians Preisedikt (Berlin, 1971)
7. William Smith, A Dictionary of Greek and Roman Antiquities (1875), “Denarius” (84 denarii to the pound of silver; one thirtieth alloy)
8. Wikipedia, “Denarius” and “Roman Republican currency”, summarising Michael H. Crawford, Roman Republican Coinage (1974), pp. 594–595: first denarius about 211 BC, 72 and later 84 to the pound
9. Christopher J. Cooksey, “Tyrian Purple: 6,6′-Dibromoindigo and Related Compounds”, Molecules 6 (2001), on Friedländer’s 1909 extraction
10. Code of Justinian 4.40.1, 11.8.11, 11.8.15, 11.9.4 and 1.23.6, annotated translation by Fred H. Blume, University of Wyoming
11. Old Book of Tang (Jiu Tangshu), chapter 45, Treatise on Carriages and Dress, Chinese text
12. New Book of Tang (Xin Tangshu), chapter 24, Treatise on Carriages and Dress, Chinese text (parallel to the edict of 621)
13. Tang huiyao, chapter 31, on robes and dress, Chinese text (parallel to the regulation of 630)
14. Nihongi: Chronicles of Japan, book XXII, translated by W. G. Aston (1896), entries for 603, 604 and 611 and notes
15. Nihon Shoki, book 22, Chinese-character text, Wikisource (the entries for the eleventh and twelfth years of Suiko)
16. Japanese Wikipedia, 冠位十二階 (the twelve cap ranks): modern readings of the colours, citing Kishi Tetsuo (1982) and Takeda Sachiko (1984); tertiary source
17. National Astronomical Observatory of Japan, calendar database: Suiko 11, month 12, day 5 = 11 January 604 (Julian)
18. Cennino Cennini, The Book of the Art, English translation (first published 1899), chapter 62 on ultramarine
19. Cennino Cennini, Il libro dell’arte, chapter LXII, Italian text (1859 edition)
20. Deutsche Bundesbank, “Sign of the lily”: the Florentine gold florin
21. B. S. Tosatti, “Cennini, Cennino di Drea”, Enciclopedia dell’Arte Medievale (1993), Treccani: the book written around 1400, probably in Padua
22. Alan M. Stahl, “The Making of a Gold Standard: The Ducat and Its Offspring”, in J. H. Munro (ed.), Money in the Pre-Industrial World (the Venetian ducat, about 3.5 g)
23. Ministero di Agricoltura, Industria e Commercio, Tavole di ragguaglio (1877): the pounds of Padua, Venice and Florence in kilograms
24. Samantha Nadel, “Red the World Over”, Science History Institute, 18 December 2025
25. Elizabethan sumptuary statutes: proclamation of 15 June 1574 (16 Elizabeth I), transcription; printed in Paul L. Hughes and James F. Larkin (eds.), Tudor Royal Proclamations, II (Yale, 1969), no. 601, pp. 381–386 (not opened; the number is taken from later works citing it)
26. Statute of Apparel, 24 Henry VIII c. 13 (1533), transcription in Cox, “Apparel statutes 1509–1533” (the source of the royal list in the 1574 clause)
27. Science History Institute, biography of William Henry Perkin
28. Science Museum Group, William Henry Perkin (patent of 26 August 1856)
29. Wikimedia Commons, source of every image on the page (author and licence under each figure)