How Did Alexander the Great Die? The Only Witness Wrote 'Clouds'

Ask me how Alexander the Great died and, off the top of my head, I will tell you this: he died in Babylon at thirty-two. Some say it was a great drink in honour of Heracles, some say poison, others fever and disease. That is not a muddled memory. It is a fair summary of the ancient evidence, which cannot even agree on his age: one source makes him thirty-two years and eight months old, another thirty-three and a month.¹ Ask the old writers four questions and you get four answers, because the answers were written by different people with different interests. Only one text about his death was written in the month it happened. It is a Babylonian astronomer's note, and in the standard edition it comes to a few words: the king died; clouds.² Every other account passed through the men who divided his empire, or was copied from them centuries later. Every modern diagnosis, from poison to typhoid to Guillain-Barré, is a diagnosis of those copies.

The king died; clouds

For centuries, Babylonian astronomers kept a running record of the sky, night by night. Beside the positions of the moon and planets they noted the weather, market prices and events in the city. That steady record-keeping is what made Babylonian prediction possible. One of these diaries, now in the British Museum as BM 45962, covers the weeks around Alexander's death. It was published by Abraham Sachs and Hermann Hunger in the first volume of their Astronomical Diaries (1988), filed under the year −322, which is 323 BC.²

The tablet is broken. Its opening section is lost, and several days have gaps. Under the 29th of the month Ajaru it records that the king died, and then, in the next breath, the weather: clouds. Leo Depuydt worked through the Babylonian calendar and the astronomical data and put the day at 11 June 323 BC, probably between about four and five in the afternoon.² The same week was heavily overcast, which is why the tablet's observations for those days are thin. In the week the king died, the sky the astronomers watched for omens could hardly be seen.

What the tablet does not say matters as much. It names no illness. It mentions no party, no cup, no poison, no doctors, no last words. It gives a date and a death, then goes back to the weather. That is the whole of the contemporary record. Everything else said about how Alexander died comes from somewhere else.

Five storytellers, none of them there

Five connected narratives of Alexander's death survive from antiquity. None was written by anyone who was in Babylon. The oldest is by Diodorus of Sicily, who wrote his universal history in about 60–30 BC, some 265 to 295 years after the event. Alexander goes to a revel given by Medius the Thessalian, drinks a great deal of unmixed wine, and at the end drains a great cup of Heracles. He cries out as if struck by a heavy blow and is led away. Before he dies he gives his ring to Perdiccas, and asked who should have the kingdom, he answers: to the strongest.³ This is where the great drink in honour of Heracles comes from, and Diodorus is the only one of the five to tell it.

Diodorus then reports a rival version. Antipater, Alexander's regent in Macedonia, had quarrelled with the king's mother Olympias and had been alarmed by the executions of the generals Parmenion and Philotas. His son, the royal cupbearer, gave Alexander poison. He adds that many writers did not dare to set the poison story down, because Antipater and then his son Cassander held power in Greece.³ On his own account, then, what historians wrote about the death depended on who was ruling when they wrote it.

Plutarch, writing his Life of Alexander about AD 100–120, over four centuries after the death, rejects the dramatic cup outright. Some writers, he says, claim that Alexander drained a bowl of Heracles and felt a sudden pain in the back as if stabbed; but that, in Bernadotte Perrin's 1919 translation, was invented "in tragic fashion" to give a great life a moving ending. On poison he is just as direct. Nobody suspected it at the time. The story surfaced in the sixth year after the death, when Olympias, acting on information, put people to death and scattered the ashes of Iollas, the cupbearer. Plutarch adds that most writers think the poison story a fabrication.⁴

Plutarch offers a fever instead. After drinking with Medius, Alexander bathes and falls ill; the fever grows day by day, and he loses his voice. The soldiers, suspecting he is already dead, demand to see him and file past his bed. He dies, by this reckoning, on the 28th of the Macedonian month Daesius. Plutarch says he took most of this word for word from the court "Journals".⁴

Arrian, a Greek official in the Roman empire who wrote in the second century AD, some 440 to 480 years after the event, tells the same fever story from the same kind of source, which he calls the Royal Diary. Alexander bathes, sleeps, sacrifices and gives orders to his officers each day while the fever grows, until he can no longer speak but still recognises his commanders; the soldiers file past as he greets them with his eyes. The famous deathbed line appears in Arrian only as something "some authors" say, set outside the Diary's account. Arrian then lists the poison stories, Antipater, Aristotle, Cassander carrying the drug, Iollas pouring it, and makes clear that he records them only so as not to seem unaware of them.⁵

Quintus Curtius Rufus, probably writing in the first century AD, some 360 to 380 years after the death, is the fourth. His account of the illness is missing, for reasons that come later. What survives includes the dying king greeting his troops, his voice failing, and a long passage on poison: most people, Curtius says, believed Iollas had given it on his father Antipater's orders, and that Cassander had brought it, water from the river Styx carried in a mule's hoof; the rumour was later suppressed by Cassander's power.⁶

The fifth is Justin, whose Latin epitome shortens the lost history of Pompeius Trogus, written under Augustus. Justin himself is placed anywhere from the second to the fourth century AD. He has no doubts. Antipater plotted; Cassander carried the poison; the brothers who tasted and mixed the king's drink added it. Alexander groans over a cup at Medius's house as if stabbed, asks for a sword to end the pain, and passes the ring to Perdiccas, naming as his heir "the most worthy".⁷

Side by side, the five sort into two families. One, shared by Diodorus, Curtius and Justin, has a sudden pain over a cup, a ring, a famous last line and poison. The other, in Arrian and Plutarch, has a slow fever, a daily routine, soldiers filing past and no last words. So the list at the top of this essay, a drink for Heracles, poison, fever, is not three memories jumbled together. It is the two ancient traditions and the argument between them.

On his own account, then, what historians wrote about the death depended on who was ruling when they wrote it.

The diary nobody has seen

The fever version rests on the document Arrian calls the Royal Diary and Plutarch the Journals, the Ephemerides. If it was what it claims to be, a daily record kept by the king's secretariat in Babylon, it is the nearest thing to an eyewitness account, and this essay's thesis must be stated with care. Athenaeus, a later collector of anecdotes, names its compilers as Eumenes of Cardia, Alexander's chief secretary, and Diodotus of Erythrae. The surviving quotations are gathered in Jacoby's collection of the fragmentary Greek historians as number 117. Apart from the death narrative, they are cited mainly for Alexander's heavy drinking and for the days he spent sleeping it off.⁸

The disagreement over the Diary is real. The older view, held by Wilcken, Berve and Jacoby himself, took it as a genuine court journal. In 1954 Lionel Pearson argued that it was a later literary forgery using Eumenes' name. In 1965 A. E. Samuel suggested that the account of the last days was based on Babylonian records. Brian Bosworth (1971, 1988) called it propaganda, issued by Eumenes soon after the death to answer the poison rumours: drink had worsened a natural illness. Ernst Badian treated it as a literary composition. Robin Lane Fox, according to a recent survey of the debate, connected it with propaganda from Cassander's circle. N. G. L. Hammond (1988) defended it as genuine, as have Edward Anson (1996), Andrew Chugg (2005) and Cinzia Bearzot (2017). P. Zaccaria (2022) argued that its portrait of the king was positive, and that the stress on drinking comes from the authors who quoted it.⁹

Both sides have reasonable arguments. Against: an official journal recording the king's hangovers is odd; the text suits an apologetic purpose, natural death and no poison, very neatly; and it sends officers to consult a god in a temple of Serapis, a cult usually tied to early Ptolemaic Egypt. For: the detail, with named officers and exact days; Arrian's statement that Ptolemy and Aristobulus broadly agreed with it; and the fact that a royal secretariat kept records of some kind.⁹ The fragments cannot settle it. On either reading the Diary survives only as paraphrase in authors writing four centuries later, and even on the most generous reading it records no cause. It lists baths, sacrifices, orders and fever, which is exactly why doctors have been able to read so much into it.

The officers who were present did write, and their books are lost. Ptolemy, a general and later king of Egypt, wrote a history that Arrian used as a main source; Aristobulus, who served on the campaign, wrote another, reportedly begun at eighty-four. Both survive only in later quotation. Medius of Larisa, the host of the party at which the fatal illness began, was himself a historian, and nothing he wrote about the death survives. Callisthenes, the court historian, cannot be a lost witness: he died in 327 BC.¹⁰ A few hostile voices from near the time survive in quotation: the pamphleteer Ephippus had Alexander drain a huge cup in a toast, fall sick and die, and blamed the anger of Dionysus.⁸ The fullest poison story, a political pamphlet on Alexander's last days and will, naming Antipater, Cassander and Iollas, survives embedded in later works and is usually dated within a decade or two of the death.¹¹

So the two traditions were both in circulation within a few years of 323, and both came from people with something at stake. Ptolemy took the body to Egypt and became a king. Eumenes, if he was behind the Diary, needed the death to be natural. The poison story blamed the family of Antipater at a time when his enemies were fighting for the empire. The eyewitnesses did not vanish. They became the interested parties.

The pages filled in 1639

Return to Curtius, whose illness narrative is missing. The manuscripts of his tenth book have a gap. The passage that should describe the last banquet, the illness and the days of decline is not there, and the text resumes with the dying king greeting his troops. Anyone who reads Curtius in a printed edition will nonetheless find a full account in the gap: Medius begging Alexander to join his party, a night of drinking, the illness taking hold. It even includes the death of Hephaestion, Alexander's closest friend, from an illness caught after a drinking bout, and the king's enormous grief.¹²

None of it is Curtius. It is the work of Johann Freinsheim, a German classical scholar, who wrote Latin supplements to fill the missing parts of Curtius in the style of the original. They first appeared at Strasbourg in 1639. The online Latin text at LacusCurtius marks the point in a note: what follows, up to the start of chapter 5, is Freinsheim's. Earlier in the same book the editor's note is blunter still: a large gap in the manuscripts, filled by Freinsheim's own invention.¹² The supplements travelled with the ancient text: a Latin edition of Curtius printed in 1685 lists them on its title page, and the online text still prints them in place, flagged only by the editor's notes.¹²

So, at the exact point in one of the five ancient narratives where the cause of death should be, the text printed in many editions was written in a scholar's study more than 1,900 years after the event. It was a learned reconstruction from the other surviving authors, which is why it fits so smoothly: a copy of the copies, printed under the ancient author's name. Nothing else in the tradition is quite so literal, but the method is the same all the way down. Somebody fills a gap from what the other sources say, and the fill becomes part of the source.

Diagnosing a document

Modern medicine has taken the case many times, always from the same pages. In 1978 the historian Donald Engels argued for malignant malaria, caught on a recent journey through marshland near Babylon.¹³ In 1998 a clinicopathological conference at the University of Maryland, published in the New England Journal of Medicine, proposed typhoid fever with perforation of the bowel and an ascending paralysis. One of its authors, the historian Eugene Borza, was quoted as saying that the earliest surviving accounts were "written three centuries after he died".¹⁴ In 2003 J. S. Marr and C. H. Calisher proposed West Nile encephalitis, resting much of the case on Plutarch's story that ravens fell dead at Alexander's feet as he entered Babylon; they themselves mentioned the problems of author bias and translation.¹⁵ John Maxwell O'Brien made the case for alcoholism in 1992, and in 1997 C. N. Sbarounis argued for acute pancreatitis after heavy drinking and a large meal.¹⁶ In 2014 Leo Schep and colleagues, writing in Clinical Toxicology, asked which poison fitted best, if he was poisoned, and chose white hellebore, Veratrum album, a plant the Greeks used as a purge; they conceded that it could not be proved.¹⁷

The most widely reported recent theory came from Katherine Hall of the University of Otago, published in the Ancient History Bulletin in 2018. She proposed Guillain-Barré syndrome, a paralysis that can leave the mind clear, and argued that Alexander may have been declared dead up to six days before he died. Headlines turned this into "buried alive". That is press shorthand for a mistaken declaration of death, and the 1998 typhoid paper had already suggested that ascending paralysis could make a living man look dead. Hall says openly that none of the main accounts is contemporary and she declines to judge between them.¹⁸

Part of her case, like part of the typhoid case, rests on the report that the body lay for days in the Babylonian heat without decaying. That report comes from Curtius and Plutarch. Plutarch uses it as evidence that there was no poison. Curtius, having told the story, adds that he reports it as it was handed down rather than as something he believes.⁶⁴ The ancient author doubted the detail that the modern diagnosis needs.

Each theory chooses a tradition. Typhoid, malaria and Guillain-Barré lean on the slow fever of the Diary. Pancreatitis and poison lean on the sudden pain over the cup, the very detail Plutarch called invented. West Nile leans on an omen. None has anything new to examine. There is no body.

The ancient author doubted the detail that the modern diagnosis needs.

The body itself became a prize. It was left untended for days while the generals quarrelled; Curtius says seven. Alexander had asked to be taken to the shrine of Ammon in Egypt, but when the funeral carriage finally set out, about two years later, Ptolemy seized it in Syria and took it to Egypt himself. It was buried first at Memphis and later moved to Alexandria; the sources disagree about who moved it and when.¹⁹ Augustus viewed the body in 30 BC and laid flowers on it; Cassius Dio says a piece of the nose broke off at his touch. Septimius Severus had the tomb locked so that no one could view the body again, and Caracalla left his robe and rings on it in AD 215.²⁰ By the end of the fourth century the preacher John Chrysostom could say that even Alexander's own people did not know where his tomb was. The same city that let its Library fade lost its founder's grave.²¹

The newest copy

There is one more answer to the question, and it is the most recent. I remembered a jealous lover as part of the story of Alexander's death, a lover jealous of Alexander's queen. When I checked, I found it in no ancient writer; it came from Oliver Stone's film Alexander (2004).

The film does have that triangle, though turned around. In Stone's version, as the historian A. E. Larsen describes it, it is Roxane, Alexander's Bactrian wife, who is jealous of Hephaestion; when Hephaestion dies, Alexander suspects she poisoned him and confronts her; he then drinks heavily at a party and falls ill with similar symptoms. At the end of the film the aged Ptolemy, dictating his memoirs, says Cassander faked the diaries to paint Alexander as a drunk, and then tells his scribe to write it differently.²² Stone put the modern scholarly quarrel over the Diary into the mouth of a man who was there. The film's historical adviser was Robin Lane Fox, the same scholar named above among those who connect the Diary with Cassander.²²⁹

The ancient sources do record jealousies around Alexander, but they point in other directions. Diodorus says Olympias, the king's mother, was jealous of Hephaestion and wrote him threatening letters. Plutarch says that after Alexander's death Roxane, jealous of his Persian wife Stateira, had her and her sister killed.³⁴ The only lover the ancient writers attach to the death itself is in one of the rumours Arrian lists and dismisses: that Medius, who hosted the party, was the lover of Iollas, the cupbearer, and took part in the plot.⁵ None of the five narratives, nor the other ancient writers searched for this essay, makes a jealous lover part of Alexander's death. In the form remembered here, it comes from a screenplay. A filmmaker retelling an ancient story is doing what Freinsheim did in 1639: filling in what the record does not say.

A Babylonian astronomer wrote down a date and a death, then went back to the weather. Within a few years two versions were circulating, each useful to someone among the men who divided the empire. Three and four centuries later, writers who could no longer check anything chose between those versions and copied them. A scholar in 1639 filled the gap in one of them. Doctors in the twentieth and twenty-first centuries diagnosed the copies. A film in 2004 added a lover. Alexander has never stopped being copied, and the copying has not stopped at his death. Each new layer is read as if it were part of the record. The only line nobody has improved on is the one from the month itself: the king died; clouds.