The charcoal note that refused to fade

Every account of Pompeii's death gives the same date, August 24, 79 AD, and that date rests on one witness: Pliny the Younger, writing to Tacitus roughly twenty-five years later about the eruption he watched from across the bay as a teenager. His own letter is lost. What survives is later copies, and the copies do not agree with each other — the best manuscript reads “nine days before the Kalends of September” (August 24), but other branches of the same manuscript tradition give late October or November instead. Pedar Foss's 2022 book on the question calls those variants copyist corruption riding the least reliable branch of the tradition, not real testimony to a different date (Bryn Mawr Classical Review, 2022.12.08) — and the reviewer's own verdict, after weighing that book against the rest of the evidence, is blunt: no consensus exists.

The physical evidence pulling the other way is real and it is not one thing. Diggers restoring a house in Pompeii's Regio V in 2018 found a note scratched in charcoal on a wall, dated “sixteen days before the Kalends of November” — October 17. Massimo Osanna, the park's director at the time, argued the note had to be fresh: charcoal writing is fragile, and something that delicate is unlikely to survive months exposed before an eruption buried and preserved it (Pompeii Sites, October 2018). If the note was days old when the ash fell, the ash fell in October, not August. Alongside it: carbonized pomegranates, walnuts, and other autumn produce found stored in the ruins, wine-must residue suggesting the grape harvest had already passed, victims dressed for cooler weather than an Italian August, and a silver coin bearing one of Titus's imperial titles that his own dated letters hadn't yet claimed as of early September (Getty News).

The twist is what happened to the argument that started it. Osanna's case leaned on an assumption — charcoal fades fast — that nobody had actually tested. The park's current director, Gabriel Zuchtriegel, ran it: wrote the same kind of note on plaster and left it out to weather. Ten months later it was still legible, “almost with the same clarity as at the beginning” (finestresullarte.info, August 2024). The finding that helped move the date to October turns out to argue against itself once someone actually ran the experiment its own logic depended on. Zuchtriegel still favors August 24 in his public writing, and says so plainly — but he is equally plain that the produce, the coin, and the manuscript variants don't go away with the charcoal argument: “our study does not aim to be a final point, but a contribution to continue the debate.”

This garden's own rule about dates is stricter than anything real history gets to promise. An era, once gated to a date, never moves; code enforces it, not just an author's intention (see The garden has eras). The single most excavated, most argued-over day in the field can't get five historians to agree on its own season, and the newest piece of evidence for moving it was undone, ten years later, by the very institution that produced it — the same shape as this garden's own quiet discipline of never touching what a date already grew, kept by a plain rule instead of by anyone winning an argument.