Library of Alexandria: No Source on File Names Its Last Day

Who burned the Library of Alexandria?

In 48 BC, fire spread from the harbor of Alexandria. Seneca, who names no year, later put the books burned there at 40,000, and writers who lived centuries after the fire gave figures of 400,000 and 700,000. The commander on the scene, Julius Caesar, names something else. “He burnt all those ships and the rest that were in the docks,” his Civil War records, before moving straight on to the landing on Pharos.

The Alexandrian War, a continuation of Caesar's commentaries, calls Alexandria “well-nigh fire-proof,” because its buildings had no wooden joinery. That does not rule out a harbor fire. But it sits uneasily beside later claims that the whole collection burned.

The later writers do not agree on what burned. According to Plutarch's Life of Caesar, the fire “spread from the dockyards and destroyed the great library.” In Herbert Baldwin Foster's 1905 translation of Cassius Dio, “the docks and the storehouses both of grain and of books were burned.” Orosius says the 400,000 books were stored by chance in nearby buildings, proximis forte aedibus condita.

Fix: books burned near the harbor during Caesar's fighting in Alexandria, a point on which Plutarch, Dio and Orosius agree; Seneca gives only a count, with no date. On what burned, a library or books stored by the docks, and on how many, the sources disagree.

The year 48 BC comes from the accounts themselves. Caesar tells of burning the ships in the book of his Civil War that opens with his consulship alongside Publius Servilius, and the 1905 English edition of Dio marks the year of the fire as 48 BC. The Catholic Encyclopedia's 1907 article on the library gives 47 BC instead.

Open the full-size diagram

What do the ancient accounts actually say?

Table: compiled from the passages listed under Sources; birth years approximate, with the source for each in the CSV.

Download the table (CSV, library-of-alexandria-accounts.csv) (file notes: sources and license) Data: Julius Caesar, Civil War 3.111, trans. A. G. Peskett (Loeb, 1914), retrieved 2026-10-07, public domain in the US, short quotation; Strabo, Geography 17.1.8, trans. H. L. Jones (Loeb, 1932), retrieved 2026-10-07, public domain; Seneca, De Tranquillitate Animi 9.5, Latin text, retrieved 2026-11-01, public domain; Plutarch, Life of Caesar 49, trans. B. Perrin (Loeb, 1919), retrieved 2026-10-07, public domain; Aulus Gellius, Attic Nights 7.17, trans. J. C. Rolfe (Loeb, 1927), retrieved 2026-10-07, public domain; Cassius Dio, Roman History 42.38, trans. H. B. Foster, Dio's Rome, vol. 2 (1905), retrieved 2026-11-01, public domain; Ammianus Marcellinus, Res Gestae 22.16, trans. J. C. Rolfe (Loeb, 1940), and Latin text, retrieved 2026-11-01, public domain; Orosius, Historiae adversus paganos 6.15, Latin text, retrieved 2026-11-01, public domain; Edward Gibbon, The History of the Decline and Fall of the Roman Empire, ch. 28, with notes by H. H. Milman (1845), retrieved 2026-11-01, public domain; Edward Gibbon, The History of the Decline and Fall of the Roman Empire, ch. 51, with notes by H. H. Milman (1845), retrieved 2026-11-01, public domain; George Bolling, "The Alexandrian Library," Catholic Encyclopedia, vol. 1 (1907), retrieved 2026-11-01, facts only, paraphrased; Project Gutenberg catalogue author records (birth and death years), retrieved 2026-11-01, factual data; Johann Peter Kirsch, "Paulus Orosius," Catholic Encyclopedia, vol. 11 (1911), retrieved 2026-11-01, public domain

Every writer in the table who reports burned books and has a birth year on file was born after the fire, Seneca about 43 years after and Ammianus about 377 years after. Orosius, whose birth year is not on file, wrote more than four centuries after it.

Seneca brings up the fire in a passage on excess in On Tranquillity of Mind, where he lets “someone else” praise the collection, as Livy did, and calls it studiosa luxuria, learned luxury, gathered for show rather than study. In Latin, his number, quadraginta milia (40,000), and Orosius's quadringenta milia (400,000) look almost alike. By dead reckoning, one figure could be a copyist's slip for the other, but the texts on file cannot say which. The navigator's method behind that label is explained in Dead Reckoning: Why Bowditch Advised Shorter Log-Line Knots.

Open the full-size diagram

Was the library still there after Caesar?

Strabo describes the Museum in his Geography as part of the royal palaces. It had a public walk, an exedra with seats, and a hall where its scholars ate together. Its priest “formerly was appointed by the kings, but is now appointed by Caesar.” The passage mentions no fire and no library.

Open the full-size diagram

The Museum also appears in imperial records. According to Suetonius, Claudius added a new Museum named after himself and had his two Greek histories read aloud yearly, one in each. Suetonius adds that Domitian renewed libraries destroyed by fire, “sending scribes to Alexandria to transcribe and correct” copies of the lost works. The Catholic Encyclopedia's 1909 article on Domitian dates his reign from 81 to 96.

Illustration

By dead reckoning, a city where an emperor sent scribes for lost works 128 to 143 years after Caesar's fire still held copies worth the voyage. Suetonius does not say which libraries had burned.

Why did the Library of Alexandria decline?

Asking which fire ended the library assumes the answer is a fire. A more useful first question is who stopped copying. By dead reckoning, the record points to people before it points to flames.

Athenaeus, citing older chroniclers, says a Ptolemy the Alexandrians nicknamed Kakergetes, “Malefactor,” murdered many citizens and drove scholars into exile. Philologists, philosophers, mathematicians and physicians scattered to the islands and towns and taught there.

Under Emperor Aurelian, Alexandria lost “the greater part of the district called Bruchion, which had long been the abode of distinguished men,” according to Ammianus Marcellinus. The Catholic Encyclopedia's 1907 article on the library dates that loss to AD 272.

Open the full-size diagram

By dead reckoning, the record fits a slow loss: scholars driven out under the king the city called Malefactor, the Ptolemies' patronage gone once Rome appointed the Museum's priest, and the Brucheion wrecked under Aurelian. None of the sources on file names a last day.

What happened at the Serapeum around 390?

The Serapeum, the temple of Serapis, held a smaller collection of its own, according to the Catholic Encyclopedia's 1907 article on the library. Ammianus praises the temple in the present tense, then writes in quo bybliothecae fuerunt inaestimabiles: in it there were priceless libraries. The verb for the libraries, fuerunt, is past tense. The temple's, eminet (“stands out”), is present.

Open the full-size diagram

Around 390, under an imperial order, Bishop Theophilus had the Serapeum torn down; Gibbon allows either 389 or 391, and the Catholic Encyclopedia's 1910 article on libraries gives 390. The church historian Socrates of Constantinople describes the demolition and the street fighting that followed and says nothing about books. Orosius says he had seen empty book chests in temples, and he calls true the report that they were emptied “by our own men in our own times.” Gibbon puts that sight near twenty years after the Serapeum fell; the Catholic Encyclopedia's 1911 article on Orosius says his history was completed in 418.

Edward Gibbon later cited those empty chests to argue that “the valuable library of Alexandria was pillaged or destroyed” when the Serapeum fell. Whether any library was left to lose by then is not recorded.

Did the Arabs burn the Library of Alexandria?

In the story, after the Arab capture of Alexandria, Caliph Umar ruled that Greek books agreeing with the Quran were useless and those disagreeing were harmful. The books then fed the city's 4,000 bathhouses for some six months. The sources on file date the capture itself differently: Gibbon gives 22 December 640, and the Catholic Encyclopedia gives 641 in one article and 642 in another.

Gibbon traced the story to the Dynasties of Gregory Abulpharagius, which Edward Pocock published in a Latin version in 1663, and wrote, “I am strongly tempted to deny both the fact and the consequences.” He called it “the solitary report of a stranger who wrote at the end of six hundred years,” set against the silence of two earlier Christian annalists from Egypt. In a note, he credited Renaudot with doubting it.

A note in the 1845 edition of Gibbon, with notes by Henry Hart Milman, lists Muslim writers brought forward since then in support of Abulpharagius, beginning with one it calls Abdollatiph. The note gives no dates for them.

Open the full-size diagram