A Diver Swam 140 Metres Into A Flooded Mine. The Charcoal Is 4,400 Years Old.
The gold in Ur's tablets came out of a hill in the Transvaal, and nobody could say when. A diver went 61 metres under its waterline for charcoal. Two laboratories put it at 2380 to 2335 BCE. The tablet is dated 2340.
MPUNZI RIDGE, South Africa — At 11:52 on Sunday morning a diver surfaced in the lower stopes of Mpunzi Ridge with forty grams of charcoal in a sealed tube.
Two laboratories, working blind on coded halves of it and told nothing about where it came from, put the fire that made it between 2380 and 2335 BCE.
Tablet PX-1207, which records eighteen tonnes of gold that did not arrive at Ur, was written in about 2340 BCE.
The locked door
In August this newspaper reported that forty-one of forty-seven Sumerian gold objects came out of this hill, and that it could not say when. The only charcoal anyone could reach had been burned over by a later fire and returned a bracket 3,110 years wide. The galleries that would fix the date were under standing water, and dewatering was a permit, a pump and about £400,000 that the operator had not applied for.
We did not wait for a pump. We sent a person.
The dive
Marit Onwuka-Lindgren has dived flooded slate, lead and gold workings for nineteen years and does not dive anything she has not been shown a plan of. Sableridge Minerals lent her its 2019 level plan, which shows three galleries off one junction below the waterline. Nobody alive had been in any of them.
She went down the old winze on 3 September, 61 metres of black water, and laid a line to the junction. Gallery 7 was behind a rockfall. So was Gallery 9. She looked at the fall for four minutes on her helmet camera and turned back.
On 7 September she went back down and along Gallery 8. It runs 140 metres from the junction on her surveyed line, fire-set on every face, with no drill mark, no shot hole and no blast round anywhere. At the foot of the last face is a hearth: a lens of charcoal under a skin of quenched spall, sealed by the rock the last fire broke and then by the water.
She spent eleven minutes at the face and two hours and forty minutes coming back up.
The laboratories
The forty grams were split and coded by a third party who held the key. Two laboratories on two continents received identical halves, no context, and no knowledge of each other.
One returned 2395 to 2335 BCE at 95 per cent confidence. The other returned 2380 to 2320. The overlap is 2380 to 2335 BCE, which is between 4,361 and 4,406 years ago, and is forty-five years wide.
The August bracket was 3,110 years wide. This one would fit inside it sixty-nine times.
Against the tablet
The scribe who wrote PX-1207 recorded a quota of gold for the road, a shortfall of seven hundred talents, and a line saying the fault was not with the workmen. Every reading of the fragment this newspaper has published dates it to about 2340 BCE.
The last fire lit in the gallery where the cutting stops was lit inside the bracket of the decades in which a scribe at Ur wrote down that eighteen tonnes had not arrived.
What the Assyriologists say
Dr. Halvard Sørensen-Ruiz, who has said in every story that cites the tablet that its fifth line is an ordinary calendrical formula, said it again on Tuesday and then said something else.
A date on a hearth, he said, is a date on a fire. It puts a fire in a hill in a particular half-century. It does not put a Sumerian in the hill, and two brackets that meet are the kind of thing that happens to any two brackets if you wait long enough.
Dr. Amaia Etxeberria, who agrees with him about line 5, answered by letter and asked that it not be cut. It is four sentences.
"I said in August that the tablet was ordinary. I still think the tablet is ordinary. I do not have an ordinary explanation for the hill. I would like one, and I would like it more than I would like to be right."
What the operator did on Monday
Sableridge's board met on Monday morning. By nine it had applied for the dewatering permit. By ten its chief executive had signed a heritage notice over the three lower galleries, which freezes nine kilometres of the company's best ground.
Her reply runs entire at the foot of this page. She accepts the date. She does not accept that it shortens the road, and she is right.
What this desk takes from it
That for three weeks the date of this working was a pump, a permit and £400,000 away, and it turned out to be a person with a line and a tube.
The bracket is forty-five years wide and it is the right forty-five years. What we do not have is anything between the escarpment and the Gulf, or Galleries 7 and 9, which are behind a fall that a diver with nineteen years' experience looked at for four minutes and would not pass.
When the water comes off, this newspaper intends to be standing at the junction.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
The August story ended on a locked door: the charcoal that would date the working was under standing water, dewatering would cost about £400,000 and a permit nobody had applied for, and the only sample anyone could reach was contaminated by a later fire on the same face. We did not dewater. We contracted a mine diver, obtained the operator's written consent and the regulator's conditions, and sent her down the old winze with a line and a sealed sample tube. The forty grams she brought out were split and coded by a third party who held the key, and the two laboratories received identical coded halves with no context and no knowledge of each other. The dive log, the line survey and the helmet footage are published in full, including the twenty-two minutes on the first dive that ended at a rockfall.
Who we spoke to
- Marit Onwuka-Lindgren, Mine and cave diver; nineteen years in flooded slate, lead and gold workings on three continents. Contracted at her standard day rate with a support team of three; dived twice, on 3 and 7 September; her dive log, line survey and helmet footage are published in full 1–7 September 2026 Turned back at the fall on the first dive and said she would not pass it on any dive. We did not ask her to.
- Two radiocarbon laboratories, working blind, Dr. Tobias Whitlock-Nkemdi and Dr. Sunniva Adebayo-Kraus, laboratory directors, on separate continents. Each received a coded split of the same forty grams with no information about origin, depth or context, and no knowledge that a second laboratory existed; both reports published entire 7–8 September 2026, on an expedited basis at the published rate One bracket is 2395 to 2335 BCE. The other is 2380 to 2320. The overlap is what we print.
- Ms. Anneke Ruthven-Mbatha, Chief executive, Sableridge Minerals Ltd, which holds the licence. Gave written consent to the dives under the regulator's conditions; sent the full draft and both laboratory reports on Monday morning; right of reply printed entire August–September 2026 Told us on Monday afternoon what her board had done that morning. It is in her reply and in the fourth section of this story.
- Dr. Halvard Sørensen-Ruiz, Assyriologist, reader in cuneiform palaeography. Sent both laboratory brackets and the draft; interviewed by telephone 8 September 2026 Reads line 5 of PX-1207 as calendrical, as he has said in every story that cites it. Says a date on a hearth is a date on a fire. His objection is printed at length.
- Dr. Amaia Etxeberria, Assyriologist. Sent the same material; answered by letter 8 September 2026 Also reads line 5 as calendrical. Her letter is quoted in full in the fifth section, because it is short and she asked that it not be cut.
Documents
- PX-1207 — Second Dynasty extraction schedule, tablet fragment (obverse and reverse) contested
What we could not confirm
- Who lit the fire. The charcoal dates a hearth to a bracket forty-five years wide. It does not say whose hand set it, and nothing in Gallery 8 is written on, marked or signed with anything but fire. A date that matches a tablet is a date that matches a tablet. It is not a name.
- Galleries 7 and 9. Both are behind a fall the diver would not pass, and the operator's 2019 plan is the only record that they exist at all. If the working ran longer in either of them than it did in Gallery 8, the bracket moves, and we cannot say which way.
- The route. It was nine thousand kilometres in August and it is nine thousand kilometres now. We have the metal at both ends and a date at one of them, and still nothing in between: no wreck, no way-station, no port, no intermediate assay. A date does not shorten a road.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Meridian TelegraphDATED: THE ANUNNAKI MINE WAS WORKING THE YEAR THE TABLET WAS WRITTEN
Ran the bracket as a signature and the hearth as a Sumerian hearth.
Had that we did not
Both laboratory brackets, correctly, credited to us.
Left out
That charcoal dates a fire and not a person; that both Assyriologists say so in our sources box; and that the two galleries which would confirm or wreck the date are behind a fall nobody has been through.
-
Signal & LedgerSableridge Applies To Dewater; Heritage Notice Freezes 9 km Of Strike
Read the whole thing as a licence story and asked what a heritage notice does to a junior miner's balance sheet.
Had that we did not
The permit application and the notice, both of which it had from the company before we did.
Left out
The date. The piece does not print the bracket, which is the only reason any of the rest happened.
-
They Buriedthis newspaperA Diver Swam 140 Metres Into A Flooded Mine. The Charcoal Is 4,400 Years Old.
Stopped waiting for a pump and a permit and sent a person under the water for the sample the August story said it could not reach.
Had that we did not
The dive log, the line survey, the helmet footage entire, both laboratory reports and the operator's reply unedited.
Left out
Our headline gives the diver's distance and the charcoal's age and not the word that matters, which is one. One hearth, in one gallery, dated by two laboratories from one lens of charcoal. We built the August story on two cores and were told so by the operator. We have built this one on one hearth and are telling ourselves. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Ms. Anneke Ruthven-Mbatha, chief executive, Sableridge Minerals Ltd on Monday 7 September at 08:00 local, in writing, with the full draft, both laboratory reports and the diver's line survey. Replied on Monday at 17:40 local, in writing, after a call of about half an hour. Printed entire and unedited, including the last paragraph, which she asked us to keep.
In August I told your correspondent that if the true figure was anything like eight thousand tonnes, my licence area was an archaeological site, a sovereign asset and a crime scene in whichever order the courts decided. I did not expect to be writing to you again inside three weeks, and I did not expect the thing that brought me back to be a woman with a rebreather and forty grams of soot.
The date is good. I have had both reports read by a geochronologist I pay and she cannot find a hole in either. A hearth at the foot of a fire-set face, under a skin of the rock that fire broke, under water that has stood since before my grandfather's time, is about as clean a context as this hill was ever going to give anybody. I accept the bracket.
Now I am going to say what I said in August, because it is still true. A date on a hearth is a date on a fire. You have dated the last fire somebody lit in one gallery of one level of a working that runs twenty-six kilometres that we know of, and you have set it against a tablet from a city nine thousand kilometres away, and the two brackets meet, and I understand why that reads as a match. I would ask your readers to notice that you still have nothing between the escarpment and the Gulf. Not a wreck. Not a port. Not a bead. The date does not shorten the road. It makes the road stranger.
Here is what my board did this morning, and you may print that it did it because of your diver and not because of anything I said to them in August, because that is what happened.
We have applied for the dewatering permit. The application went in at nine. It commits this company to roughly four hundred thousand pounds and a pump, and to a heritage notice over the three lower galleries, which I signed at ten, and which freezes about nine kilometres of strike that my shareholders were told in March was our best ground.
I have cost them a hill. I have done it on your newspaper's evidence, which is two cores, one hearth and a diver who would not pass a rockfall, and I want it on the record that I know exactly how thin that is and that I would do it again this afternoon.
Because the alternative is that somebody worked this reef by hand, in the same decades a scribe at Ur was writing down that eighteen tonnes had not arrived, and I go on selling it as a resource.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.