A digital edition · public domain, 1822

The Book of Fate

"Formerly in the possession of, and used by, Napoleon" — an English oracle chapbook that sold itself as a translated Egyptian papyrus, printed the same decades tarot was acquiring its own invented Egyptian pedigree. It's not a tarot deck. It belongs on this site anyway, as the clearest sister case we have.

⚑ Presented as fact by the book; it was a hoax ✎ Draft — 10 of 32 questions transcribed so far

Read the original 1822/1923 scan on the Internet Archive — everything on this page is transcribed from it.

The story the book tells about itself

In 1801, the book claims, the naturalist M. Sonnini — travelling with Napoleon's Egyptian expedition — broke into a royal tomb near Mount Libycus and found a mummy with a roll of papyrus bound to its chest. A Coptic scholar decoded the hieroglyphics for the First Consul, who had them translated into German for secrecy and kept the manuscript in his private cabinet ever after, consulting it, the book says, before every campaign and on the eve of every battle. It survived him only by accident: lost at Leipzig in 1813, sold for a few coins by a Prussian officer to a French general who died before he could return it, passed eventually to the Empress Marie Louise, who granted an English translator — one "H. Kirchenhoffer, Fellow of the University of Pavia" — permission to publish it. London, 1822. Five shillings, boards.

"He accordingly secured the original and translated Manuscripts, in his private cabinet, which ever after accompanied him... They were held by him as a sacred treasure, and are said to have been a stimulus to many of his grandest speculations." — Translator's Preface, The Book of Fate (1822)

None of it happened. There was no Kirchenhoffer, no Sardinian knighthood, no papyrus, no Coptic scholar — the whole apparatus, dedication to a real empress included, was invented for the London chapbook market. It's the same move Antoine Court de Gébelin made for tarot in 1781: an invented Egyptian pedigree, dressed in real names and real dates, sold as the frame that would make an oracle worth buying. Where A History of Tarot corrects that record for the cards, this page just leaves the correction sitting in plain sight, up top, rather than at the end.

How the oracle actually worked

The consultation method is a real, functioning piece of design — structured randomness with a lookup table, no cards, no dice, just a pen and a count. Here it is, condensed from the book's own instructions:

  1. Trace five rows of upright lines with a pen, quickly and without counting — each row needs to come out to more than twelve marks, but exactly how many is left to chance.
  2. Count off each row in twelves. What's left over — the remainder — is either odd or even.
  3. Mark one star beside a row if its remainder is odd, two stars if it's even. Five rows, five marks: this double column of stars is your Sign.
  4. Find your Sign among the printed columns in the book's table. Each Sign points to one Hieroglyphic — a named image (a Pyramid, a Plough, a pair of Cross Bones).
  5. Ask your question aloud, by its number in the list of thirty-two. Look up that Hieroglyphic under your question, and the words beside it are your answer.

The translator adds, dryly, that the blood-dipped reed and the temple sacrifices can be safely skipped — "he and his friends have, invariably and without the least detriment, used a pen dipped in common ink."

The thirty-two questions

As listed on the book's frontispiece. The verified ones are checked against clean, quoted text in the translator's preface — wording and numbering confirmed. The rest are a best-effort reading of a badly damaged scan (the frontispiece was set in italic type the OCR mangled badly) and may not match the original phrasing or order exactly; treat them as a sketch, not a transcription.

Cast the oracle

Ten questions, each with a set of the book's own answers behind it — fifty-eight in all, transcribed from the scan. Pick a question and cast. Your five rows are read as five odd/even marks, which makes one of thirty-two possible Signs, exactly as the book has it, and the Sign is what selects your answer: different marks, different answer. Each result tells you where it comes from — a pairing the book documents outright, or an answer from its bank matched to your question by subject.

Ask:

The hieroglyphics, as cards

The documented pairings only — the twelve cases where the book itself names the Hieroglyphic alongside the Question and the Answer, several of them recorded as Napoleon's own consultations. Click one for the full answer and its source.

Sources & honesty. Text drawn from the 1923 H. S. Nichols facsimile of the 1822 London original, digitized by the Internet Archive (full scan) and the Library of Congress. The hoax attribution — no historical Kirchenhoffer, the Sonnini/Napoleon story invented for the market — follows the Public Domain Review's account, "Hieroglyphics of a Rope-Dancer".

What is verified, and what isn't. Twelve answers are documented pairings: the translator's preface prints the Question, the Hieroglyphic and the Answer together, several of them as Napoleon's own recorded consultations. The remaining forty-six answers are the book's own words, transcribed from its answer bank in the body of the volume, and matched to a question by their plain subject matter — "the patient will recover" belongs to the question about the patient. That matching is this edition's, not the original's, and each cast result says so. The reason for it: the book routes a Question and a Sign through an engraved fold-out plate to a Hieroglyphic, and the Hieroglyphic's own page then gives the answer — but the plate and the engraved Hieroglyphic headers are images, and they did not survive the OCR this edition was built from, so the exact routing cannot be reconstructed here without going back to the page scans. Obvious scanning damage in the answers has been repaired (a dropped letter, a broken word); no wording has been modernised or rephrased. Twenty-two of the thirty-two questions are listed above but not yet castable, and the full body of answers is much larger than the fifty-eight here.

Want to take it further? The wishlist has this written up as an open job, with the data shape spelled out — even a handful of newly-verified questions is a welcome pull request. This page is a companion piece to Tarot as a Source of Fiction.