The Eruption of AD 79, Hour by Hour

The eruption of Vesuvius in AD 79 lasted less than two days, according to Italy's Civil Protection Department, and it destroyed Pompeii in two different ways. First came hours of falling pumice and small rocks, which piled up about three meters deep and pulled roofs down. Then fast, hot clouds of gas and ash, called pyroclastic surges, ran through the streets. There was no lava flow. Herculaneum, on the other side of the volcano, got the surges first and almost none of the pumice. The only clocks that hold up are Pliny the Younger's, so this page gives hours where he does, and calls everything else an estimate.

Two letters from Pliny the Younger to the historian Tacitus (Epistles 6.16 and 6.20) are the only eyewitness account. Pliny was seventeen, staying at Misenum across the bay, and he wrote a generation later. His uncle, Pliny the Elder, commanded the fleet there and died on the other side of the gulf, at Stabiae. The letters never name Pompeii or Herculaneum. They name Misenum, Stabiae, Capri, a woman called Rectina, and a mountain that, Pliny says, was "found afterwards to come from Mount Vesuvius."

So when a page tells you what time the surge reached the Forum, it is quoting a volcanology paper, not an ancient witness. Pliny's clocks are few. A cloud is seen at Misenum around the seventh hour of the first day. Through the night the earth shakes and stones fall on his uncle's host at Stabiae. At the first hour of the second day the light is still weak and doubtful, and then the cloud comes down over the sea. His uncle's body is found when daylight returned, the third day counting from the last he had seen (Epistles 6.16.20).

The hour needs care too. Pliny's Latin reads hora fere septima, about the seventh hour, and Romans counted hours from sunrise, so the seventh falls in the early afternoon. Printed translations that say "one in the afternoon" are making a conversion, not quoting him. The same goes for the date, which gets its own section below.

Volcanologists number the deposits differently: S1 to S6 in Sigurdsson and Mastrolorenzo, EU units in Cioni, letters in Luongo. The table uses no numbers of its own. Where a figure comes from one study, the last column names it, and the estimates are estimates.

The pumice phase is the part most people skip, and it is where more than a third of the recovered victims died. For at least 12 hours, and by some estimates 20 or 22, small rocks and pumice fell on a town with no way to shake it off. The Park of Pompeii says the victims of this first phase are known only from bones, because pumice does not keep the shape of a body the way fine ash does. Roofs gave way under the load. Giacomelli and colleagues found that most people who died in this phase, between 80 and 88 percent, were inside buildings.

Then the surges arrived. One of them left a deposit about 3 cm thick, spread over some 80 square kilometers, yet Mastrolorenzo and colleagues attribute most of the deaths by flow in Pompeii to it. A thin layer does not mean a mild event: it was fast, hot and left little behind. At Pompeii, about 10 km from the vent, the temperatures measured on rock fragments in the deposits sit between 180 and 380 degrees C, mostly between 240 and 340.

The two phases killed in different ways. The first phase killed with weight and collapse over many hours. The second killed in minutes. A single sentence like "everyone died instantly" fits only the second.

Pliny's Latin reads nonum Kal. Septembres hora fere septima: the ninth day before the Kalends of September, about the seventh hour. Counted the Roman way, that gives 24 August. The date on most pages is that conversion, and it is the traditional one. The manuscripts do not fully agree: the ninth-century Mediceus reads "the ninth day before the Kalends of September" (24 August); a fifteenth-century copy in Paris gives no month; a printed edition of 1474 reads November; Rosini in 1797 proposed December (Getty Museum summary of the textual record).

In October 2018, archaeologists in Regio V found a charcoal inscription in a room being renovated in the House of the Garden: XVI K Nov, the sixteenth day before the Kalends of November, which is 17 October. It has no year. The Park first presented it as a sign the eruption came later than August. The reply is that charcoal can survive on a wall for more than a year, so the line may be a year old. A 2022 review by Doronzo and colleagues in Earth-Science Reviews argued for late October or early November, and a critique by Altamura appeared in 2023. P. W. Foss, a philologist, defends August in his 2022 book on Pliny and Vesuvius.

In December 2024 the Park's director, Gabriel Zuchtriegel, told the Guardian that the manuscripts leave no alternative to August: "either you confirm Pliny or you have to simply assume that he made a mistake." Plant remains such as chestnuts prove nothing, he argued, because August varieties exist.

We write "the traditional date, 24 August" and mean exactly that: the date in the textbooks, not a settled fact. A page that states the date without mentioning the argument has picked a side without telling you.

Researchers do not agree, and a fair summary gives the positions rather than a verdict.

Heat. Mastrolorenzo and colleagues (2010) studied 93 casts at Pompeii, 37 bodies at Oplontis and 78 skeletons at Herculaneum, and concluded that heat was the main cause of death, not suffocation by ash. Their threshold: at least 250 degrees C at 10 km from the vent could kill instantly, even indoors. The Park's own page on the casts says victims of the surge died of thermal shock.

Suffocation. Giacomelli and colleagues (2003) saw no burns on bodies, clothes, objects or frescoes, and noted that the victims lie above the base of the flow deposit, so the first currents were not what killed them. For them the later flows asphyxiated people.

Collapse. Recent work adds a third cause. Sparice and colleagues (2024) tie wall collapses in the Insula of the Chaste Lovers to earthquakes during the eruption, and report crushing injuries on two skeletons. Giacomelli and Scandone (2026) suggest many people died as they fled the tremors, after roofs had held for hours.

These can all be true of different people on different streets. The shapes of the casts do not settle it: Mastrolorenzo reads them as cadaveric spasm from heat, others as a slower agony.

Pliny the Elder's own death is an opinion. His nephew writes that he fell and died, "suffocated, as I conjecture, by some gross and noxious vapour." The Latin is ut ego colligo, "as I gather." His body, found when daylight returned, the third day counting from the last he had seen (Epistles 6.16.20), looked to Pliny more like a sleeper than a dead man.

Pompeii has a precise number of bodies found and no precise number of dead. Herculaneum has a range. The totals you see online are sums of unlike things.

How many people escaped cannot be known. A figure of about 2,000 deaths circulates; it comes from one volcanology glossary entry and no archaeological count backs it, so we leave it out. For how the casts relate to these numbers, see the page on the body casts.

The empire answered fast. According to Suetonius (Titus 8), the emperor Titus chose commissioners by lot from among the former consuls to relieve Campania and applied the property of victims who left no heirs to rebuilding the damaged towns. Cassius Dio (66.24) tells a similar story, with the property of the heirless dead going to the inhabitants. The Park says the cities were abandoned, survivors came back to recover what they could, and centuries of looting followed. The story of what happened next is told in our history of the excavations. For the volcano itself, see Vesuvius and Pompeii, including what the Civil Protection Department says about monitoring.