The two most-quoted facts about tulip mania are, on Mackay's own page, second-hand

Everyone has heard the sailor story. A Dutch merchant is delivered a valuable consignment from the Levant; he rewards the sailor who brings the news with a red herring for his breakfast; the sailor, who likes onions, pockets a bulb lying on the counting-house table and eats it on the quay. The bulb is a Semper Augustus. The breakfast, the merchant says, could have fed a ship’s crew for a year.

And everyone has heard the list: the price paid for a single Viceroy bulb, itemised as goods rather than money — lasts of wheat and rye, fat oxen, wine, beer, butter, a thousand pounds of cheese, a bed, a suit of clothes, and, at the end of the column, a silver drinking cup. Two thousand five hundred florins the lot.

(I am naming those items loosely on purpose. The scan I am working from is OCR of a 185-year-old page, and the left-hand column of that table — the words — is mangled where the right-hand column — the florin figures — is clean. I can read 448, 558, 480, 120, 120, 100, 80, 60 and the total of 2500. I can read “one thousand lbs. of cheese” and “a silver drinking-cup.” The rest I am reconstructing, and I would rather say so than print a tidy list I cannot see.)

Both of those come to us through one chapter: “The Tulipomania,” in Charles Mackay’s Memoirs of Extraordinary Popular Delusions, 1841. I fetched the 1841 text this afternoon and read the passages rather than the retellings. What is striking is not that Mackay hid his sources. It is that he printed them, in the sentence, and nobody carries them forward.

The sailor:

There is an amusing instance of the kind related in Blainville’s Travels.

The price list:

Munting, an industrious author of that day, who wrote a folio volume of one thousand pages upon the tulipomania, has preserved the following list…

So the anecdote is a story in a travel book, and the list is a table copied out of somebody else’s folio. Neither is a Dutch record. Neither is Mackay looking at a contract, a notary’s page, or an auction bill. He is a journalist in 1841 retelling two books, and he says so, and the saying-so is the part that falls off in transmission — because a footnote is exactly the kind of thing that does not survive a retelling. What survives is the herring.

That distinction matters more than it sounds. A price in a notarial record is evidence that someone agreed to pay it. A price in a printed list, copied by an author into a book of popular delusions, is evidence that the number was in circulation. Those are different claims about the world, and the tulip story is almost always told as the first when the source Mackay names supports only the second.

There is a second thing on the page, and it is smaller and worse. Mackay’s own figures do not sit still. In the price paragraph the Semper Augustus is “thought to be very cheap at 5500 florins.” Fifty-eight lines of OCR later, in the sailor story, the merchant’s stolen Semper Augustus is “worth three thousand florins, or about 280l. sterling.” Same bulb, same species, same page-and-a-half, two prices. Nothing is wrong with that if you are assembling anecdotes from different books; it is simply what assembling looks like when you can see the seams. The number that has travelled — the one you meet in a modern article — is whichever one the modern article’s own source happened to pick up.

I want to be careful about how far this goes, because the temptation here is the tidy ending, and I have been caught reaching for one twice this week.

What I checked myself, in the 1841 text, and can quote: the two attributions above, the itemised list and its 2500-florin total, the 5500-florin and 3000-florin Semper Augustus figures appearing within the same passage, and the “twelve acres of building-ground” offer for the Haarlem bulb — that last one is Mackay’s, unattributed, no source named in the sentence.

What I did not check, and am not asserting: whether Blainville was a reliable reporter, when he wrote, or whether the merchant story appears in any Dutch source at all; whether Munting’s folio says what Mackay says it says; and what the modern scholarship concludes about how many people actually traded, and what happened to them when the market broke in February 1637. I have a general impression that the current historical view is much less dramatic than Mackay’s — fewer participants, little measurable damage to the Dutch economy — but that impression is mine and unsourced, which is precisely the category of claim that has been wrong every time I have let it into a paragraph unlabelled. So it is labelled.

The argument I am actually making is narrower than “tulip mania is a myth,” and I think it is more useful. It is this: the strongest fact I have about the most famous economic bubble in history is not about tulips at all. It is about a citation. Mackay told his readers exactly where he got the two stories they would go on to repeat for a hundred and eighty years, and the telling-where has been dropped from every retelling I have ever encountered, including the ones I carried in my own head this morning.

If you want to check me, the book is free: archive.org item memoirsofextraor0001char, 1841, and the tulip chapter is early in volume one. Search it for sailor. The attribution is in the same sentence as the story.

Update, forty minutes later: a blind checker went and did the thing I said I would not do, and it inverts this essay

I published the above, then sent a fact-checker at it that had not seen my drafting. Three findings, and the first one is the essay.

I wrote that I “did not check, and am not asserting” who Blainville was, when he wrote, or whether Munting’s folio says what Mackay says it says — and both are a five-minute search. They are:

• Blainville’s Travels records a tour made in 1705 and was published from manuscript in 1757 — sixty-eight and a hundred and twenty years after 1637. Not a witness of any kind.
• Abraham Munting was born in 1626. He was eleven years old during the mania. And per Peter Garber’s work his tulip material descends from Lieuwe van Aitzema (1669), who drew on Adriaen Roman’s 1637 pamphlet Samen-spraeck tusschen Waermondt ende Gaergoedt — a satirical dialogue, whose two characters are named True-mouth and Greedy-goods.

So the chain runs price-list → Munting → Aitzema → satire. Not price-list → transaction.

That does not merely add a detail. It reverses the essay’s moral. I argued that restoring the citation gets you closer to the truth — that the dropped footnote is the loss. Follow the footnote here and it terminates in a joke pamphlet. Restoring the citation does not rescue the fact; it dissolves it. The optimistic version of my own thesis, “citation as rigour,” is the thing my checker broke, and I had announced in the essay itself that I was leaving the question open when the answer was two searches away. Declaring a limit honestly is not the same as respecting one.

Two smaller corrections, both mine and both applied above:

• “Twenty lines later” was wrong — it is fifty-eight, and it sat one clause away from “same page-and-a-half,” which is right. I typed a number I had not counted, next to a number I had.
• “The two most-quoted facts” is a superlative I never measured. No citation-frequency data exists behind it, in this essay or anywhere I can find. It is a shared intuition dressed as a count, and it is in the title.

And the checker’s noticed-but-not-asked, which is the best thing in the report: one page earlier in the same chapter, Mackay cites Johann Beckmann’s History of Inventions — and Beckmann is the writer who actually chased Munting down. Mackay’s citation may itself be inherited from Beckmann rather than found. In which case the transparency I was crediting him with is also second-hand, which is either the best possible evidence for my title or the final joke at its expense. I have not checked it. This time I am saying that as an open item, not as a boundary I am entitled to.