Two phases, two different dangers
Mount Vesuvius erupted over roughly two days in 79 CE, in two distinct phases with very different consequences for Pompeii’s population. The first phase produced around eighteen hours of falling pumice, a slower and more survivable danger that gave many residents enough time to gather belongings and flee the city. The second phase brought pyroclastic flows, fast-moving, dense clouds of scorching ash and gas, which overwhelmed the town and killed those who had remained, whether by choice, circumstance, or the belief that the worst had already passed during the initial pumice fall.
Heat, not suffocation, as the likely killer
A 2010 multidisciplinary study examining the remains and site conditions concluded that extreme heat, rather than suffocation from ash inhalation as long assumed, was the primary cause of death for most victims, finding that pyroclastic flow temperatures of at least 250 degrees Celsius proved lethal even for people sheltering indoors as far as ten kilometres from the volcano. Archaeologists have recovered approximately 1,150 bodies from the site, a number that, set against Pompeii’s known population, suggests a substantial proportion of residents successfully evacuated during the earlier pumice phase rather than remaining to be caught by the more lethal flows that followed.
About 1,150 bodies, not a whole city
The eruption’s dating rested for centuries on the eyewitness letters of Pliny the Younger, written years after the event and describing what he had witnessed as a young man, which suggested an August date and was treated as authoritative for most of the intervening two millennia. This dating went essentially unchallenged for so long partly because Pliny’s account was the only detailed written testimony to survive, giving historians little independent basis on which to question it despite growing physical evidence from the site itself that seemed inconsistent with a summer eruption.
A letter that got the month wrong
That long-standing chronology was overturned by a collaborative 2022 study that used physical evidence recovered from the site itself, including the types of produce present, the thickness of clothing found on victims, and inscriptions on coins discovered in the ruins, to determine a more precise date of 24 to 25 October, several months later than Pliny’s letters had implied. This case stands out precisely because it shows a documented eyewitness source, generally treated as strong historical evidence, being corrected not by a rival written account but by physical archaeological evidence accumulated and analysed centuries after the fact.
Produce, clothing and coins correct the record
Pompeii remained largely forgotten for centuries before systematic excavation began in earnest in the mid-eighteenth century, and the site’s modern significance owes much to Giuseppe Fiorelli, whose nineteenth-century development of a plaster-casting technique allowed archaeologists to preserve the exact bodily impressions victims had left in hardened volcanic ash, giving later visitors and researchers direct physical access to the eruption’s final moments in a way few other ancient disasters can offer. Contemporary archaeological work at the site has shifted toward conservation and targeted investigation rather than further large-scale excavation, reflecting a broader shift in the discipline away from wholesale digging and toward preserving what has already been uncovered.
A technique that let plaster speak for the dead
This is a genuinely rewarding case for showing how even a well-known, long-settled historical date can be revised once the right combination of physical evidence is examined properly, correcting a chronology that had rested on a single eyewitness account for nearly two thousand years. It is particularly strong on distinguishing the eruption’s two separate killing mechanisms, an important corrective to any flattened popular image of the disaster. It is less useful if you want a confident total death toll, since the recovered body count establishes only a partial figure rather than the full scale of loss.