It fits in two hands. Four flat sides of baked clay, with a hole bored straight through the middle so the whole thing can be turned on a rod while you read it.
What is written on it is a list of kings.
The first one ruled for 28,800 years. The second for 36,000. By the time the list reaches the flood, eight men have held the kingship of Sumer for 241,200 years between them.
Then the water comes. And the numbers start to fall.
A prism in Oxford
The object is catalogued as WB 444, and it sits in the Ashmolean Museum. It came into the collection through Herbert Weld-Blundell, an English traveler and patron who funded excavation in Mesopotamia in the years around the First World War. Its findspot is usually given as Larsa, in southern Iraq, though the chain of custody is not as clean as anyone would like.
Stephen Langdon, then the Assyriology professor at Oxford, published the text in 1923.
The prism stands about eight inches high and carries the fullest surviving version of a document scholars call the Sumerian King List. It is not the only copy. The list survives in more than a dozen manuscripts and fragments, some of them badly broken, disagreeing with each other in places. WB 444 is the one that is nearly whole.
It is generally dated to around 1800 BC. That makes the prism itself younger than most of what it describes by a margin that is hard to hold in the head.
What the clay actually says
It opens with a line that has never lost its force. In the standard rendering: after the kingship descended from heaven, the kingship was in Eridu.
Then it gets to work.
At Eridu, Alulim ruled 28,800 years and Alalngar ruled 36,000. The kingship moved to Bad-tibira, where En-men-lu-ana held it for 43,200 years, En-men-gal-ana for 28,800, and Dumuzid the shepherd for 36,000. Then Larak, one king, 28,800 years. Then Sippar, one king, 21,000. Then Shuruppak, and Ubara-Tutu, 18,600 years.
Eight kings. Five cities. 241,200 years.
And then five words that stop the column dead: then the flood swept over.
What follows the flood is the strange part. The kingship descends from heaven a second time, at Kish, and the list keeps going. But the reigns have shrunk. Hundreds of years now, and thousands, not tens of thousands. Etana of Kish, described as the shepherd who ascended to heaven, gets something over a thousand years depending on which manuscript you read. Gilgamesh of Uruk, lord of Kulaba, gets 126.
Keep reading and the numbers keep shrinking, until near the end they arrive at reigns of a few decades. Ordinary human reigns. Kings we can match to inscriptions and contracts and other people's records.
The document does not change its tone anywhere along that curve. The 28,800 and the 24 are written the same way, in the same hand, in the same format. Whoever compiled this was not switching from myth to history. He thought he was doing one thing throughout.
The number 3,600
Look at the reigns again and something jumps.
28,800. 36,000. 43,200. 21,000. Every one is divisible by 3,600, or by 600.
Those are not arbitrary figures. Sumerian counting ran on base 60. The unit of 3,600 was the sar, the unit of 600 the ner, the unit of 60 the soss. The antediluvian reigns are whole numbers of sars and ners, and nothing else. Alulim's 28,800 is exactly eight sars. Dumuzid's 36,000 is ten.
This is where the modern arguments enter, and they run in two directions at once.
Zecharia Sitchin, in The 12th Planet in 1976, took the sar as the key to everything. He argued that 3,600 was not a counting unit at all but a duration, the orbital period of the planet he called Nibiru, and that the reigns of the first kings were counted in orbits because the kings were not from here. On that reading the list is not exaggerated. It is measured in somebody else's calendar.
Assyriologists read the same pattern as arithmetic. If every figure in a table is a round multiple of your base unit, the likeliest explanation is that the figures were generated rather than recorded. Dwight Young and others have argued in the specialist journals that the antediluvian numbers are built from squares and multiples of 60, a scribal construction rather than a register of elapsed time.
The pattern is real either way. What it means is the whole fight.
The case against reading it as a record
The mainstream objection is not a quibble about numbers. It is about what kind of document this is.
Standard Assyriology holds that the King List is a political text. Its entire structure rests on one claim: that kingship over the land was single, indivisible, and passed from one city to the next in sequence. Eridu, then Bad-tibira, then Kish, then Uruk, then Ur, one at a time.
That is not what the archaeology shows. Several of those dynasties are understood to have ruled at the same time, in different cities, sometimes at war with each other. Lay them end to end and you inflate the total enormously. Lagash, a major Sumerian city with a well-documented line of rulers, does not appear on the list at all. On the usual reading, the omission is deliberate, because Lagash did not fit the story of a single crown moving through history toward whoever commissioned the copy.
There is a second problem for the flood section specifically. In 2003 the Assyriologist Piotr Steinkeller published an early manuscript of the list from the Ur III period, several centuries older than WB 444. It has no antediluvian kings. It begins after the flood. On the standard account, the 241,200 years were prefixed to an existing document later, as a way of running the crown back to the beginning of the world.
That is the case, and it is a serious one. It should be said plainly.
The part that will not go away
Here is what the argument does not reach.
The shape of the thing keeps turning up somewhere else.
Around 281 BC a Babylonian priest named Berossus wrote a history of his country in Greek, for a Hellenistic king. The work itself is lost. It survives only in quotations, passed through Alexander Polyhistor and preserved by later writers including Eusebius and Syncellus. In those fragments Berossus lists the kings who ruled before the flood, gives their total as 432,000 years, and ends with the flood hero.
432,000 is 120 sars. Written more than a thousand years after the prism, in a different language, for a foreign court, and still counting in the same unit.
Genesis 5 does the same job in a different register. Ten men from the first to the flood, with lifespans that run to 969 years and then stop. After the water, the numbers come down. The Hebrew figures are not sars and the two traditions are not copies of each other, but the architecture is identical: a long first age, a flood, and a diminished world afterward.
And then there is the curve itself. If the reigns were pure invention meant to impress, there is no obvious reason for them to decay so smoothly, from tens of thousands, to thousands, to hundreds, to the ordinary reigns of men we can verify from other documents. The list does not have a seam. It has a slope.
Whatever the compilers were doing, they were not throwing big numbers at a wall. They were working to a scheme. Nobody has recovered the scheme in a form that satisfies everyone who has looked at it.
What is left on the clay
241,200 years, in eight reigns, in five cities, on a prism you could carry under one arm.
Either the numbers are arithmetic, a base-60 exercise dressed as chronology by scribes who needed a crown to reach back to the moment kingship came down from the sky.
Or they are a count of something, kept in a unit we have not identified, by people who thought they were writing down what happened.
The clay does not say which. It only turns on its spindle, and lists them.