We no longer have the original letters, only translations and transcriptions of copies, the first of which was made in the fifth century. “Many manuscripts of Pliny’s letters came down to us with differing dates,” said classicist Daisy Dunn. Her 2019 biography of Pliny, The Shadow of Vesuvius, is the definitive guide to him and his uncle, naturalist Pliny the Elder, who died during the eruption. “August 24 was chosen as the most secure on textual grounds,” Dunn said.

In sticking by Pliny, the park walked back some of the recent enthusiasm for October 24 as a possible start date for the eruption, a theory that had been fuelled by the 2018 discovery of a scrap of graffiti on a wall of the site’s freshly excavated House of the Garden. The charcoal scrawl records a date that translates to October 17 in the modern calendar, suggesting that the eruption might have occurred after this time. The find, which did not specify a year, seemed to corroborate other unearthed clues that pointed to cooler weather than is typical in August: remnants of unripe autumnal fruits such as chestnuts and pomegranates; heavy wool clothing found on bodies; wine in sealed jars, indicating that the grape harvest was over; and wood-burning braziers in homes.

Massimo Osanna, general director of the park at the time of the discovery, was convinced that the graffiti was idly doodled a week before the explosion. “This spectacular find finally allows us to date, with confidence, the disaster,” he said. Dunn found it improbable that Pliny would have forgotten such a momentous date; still, she said, “in my view, the traditional date of August 24 is just too early in the year to be accurate”.

The dating game

The park’s recent about-face from October to August relied in part on a forensic analysis of Pliny’s letters by Pedar Foss, a classicist at DePauw University in Indiana. For his 2022 book, Pliny and the Eruption of Vesuvius, Foss examined 79 early hand-copied manuscripts of the letters and mapped out how textual errors had been compounded. He concluded that a simple scribal mistake, made in the 1420s, of switching a “u” for an “n” had resulted in an incorrect eruption date of November 1. The error appeared in the second print edition of Pliny’s letters, in 1474, and gave rise to further misreadings, misunderstandings and misuses.

By the 20th century, seven different possibilities were in circulation – eight, counting November 9, which Mark Twain casually proposed in Innocents Abroad, his 1869 travel narrative. “Those many options gave the appearance of doubt concerning what Pliny actually wrote, but upon examination, I was able to explain away each of the mistaken alternatives,” Foss said.

He also explained away each of the archaeological alternatives to August 24, some of which he believes fail based on the evidence and some on the basis of faulty reasoning. He argued that the pomegranate rinds were used for dyeing, not eating; that the Romans commonly used braziers for cooking, not just heating; that wool clothing was standard gear for Roman firefighters; and that Roman agricultural and storage practices allowed for the preservation of fruits beyond their natural harvest seasons.

As for the House of the Garden doodle, on October 12, 2023, researchers commissioned by Osanna’s successor, Gabriel Zuchtriegel, left their own charcoal message on the same wall on which the graffito appeared. Ten months later – on August 24 – the text was still perfectly legible. “The inscription could have been put on the wall during October of any number of previous years,” Foss said.

So much for dating the disaster with confidence.

Letting off steam

Claudio Scarpati, a volcanologist at the University of Naples Federico II, favours the traditional date. “In my mind, the eruption occurred in August, on a sunny day,” he said. Scarpati is the lead author of two recent studies of the catastrophe published in the Journal of the Geological Society. One offered an hour-by-hour reconstruction, extending the chronology to 32 hours from the previously estimated 19. The other revealed a dynamic sequence with 17 distinct “pyroclastic density currents,” many of them previously undocumented.

Pyroclastic currents are hot, swift-moving mixtures of volcanic particles – ash, pumice lava fragments and gas – that flow according to their density in relation to the surroundings. Scarpati said that contrary to popular belief, Pompeians were neither entombed by molten lava nor poisoned by gas. “No lava reached Pompeii, and the gas was predominantly vaporised water and, to a lesser extent, carbon dioxide,” he said. “According to our studies, the victims died primarily from asphyxiation caused by inhaling ash.”

To measure the distribution and the volume of the ash and the pumice layers, the team measured the thickness of the single layers over a 775-square-mile (2007-square-km) area around Mt Vesuvius. The deposits recorded dramatic, increasingly violent pulses from the volcano.

At noon on Day 1, Vesuvius began to eject a plume of rocky volcanic fragments and gas into the air, known as an eruption column. The mushroom cloud that Pliny observed at 1pm was typical of what is now known as a Plinian eruption, in tribute to his richly detailed testament.

Scarpati said the first currents had flowed to the town of Herculaneum, west of Vesuvius, bringing intense heat that essentially roasted inhabitants and, in one documented case, turned human tissue into glass, a process known as vitrification. At Pompeii, south of the volcano, the currents were cooler, and only the final eight penetrated the town.

During the first 17 hours, Scarpati said, Pompeii was blanketed with pumice lapilli from the column, which fluctuated like a giant fountain through 12 distinct pulses. At 2pm, the volcano began to spew pumice mixed with gas. Over the next four hours, roofs began to cave in under the weight of the pumice lapilli, causing some supporting walls to crumble as well. After 17 hours, the debris in Pompeii was up to 2.7 metres thick. Enough was ejected to bury the New York City borough of Manhattan roughly 137 metres, or 45 building storeys, deep.

The eruption peaked when the column reached its maximum height of 34km, at about 1am on day two. “The column rose as long as its density was lower than that of the air, like a balloon,” Scarpati said. At daybreak, enormous amounts of fine ash and pumice collapsed the eruptive column, forming pyroclastic currents.

During a brief lull, Pompeians presumably tried to flee the town. Then, just after 7am, the 13th and most lethal current struck – a thick concoction of ash was disgorged for nine hours, spreading detritus 16 miles across the plain and into the Lattari Mountains. In Pompeii, many victims of the volcano were found in the streets encased in this layer.

Around 4pm, the magma in the volcano’s conduit interacted with groundwater, causing the magma to break up into fine ash. No human remains were found in any of the layers after the 13th, suggesting to Scarpati that the morning’s devastation left no survivors. The eruption ceased at 8pm.

Paul Cole, a volcanologist at the University of Plymouth in England who was not involved with the project, said, “The work places a finer timeline on the events of 2000 years ago, and also provides fresh evidence for how the hazard from such large, explosive eruptions can change even during the event.”

Rumpus involving Vesuvius may go on endlessly, but unlike Pliny’s letters, the geological history of the eruption seems to have been written in stone.

This article originally appeared in The New York Times.

Written by: Franz Lidz

Photographs by: Gianni Cipriano

©2025 THE NEW YORK TIMES