Verdicts from the sources · 14 September 2026

Historical Mysteries

Eighteen historical mysteries examined from the original documents. Each case ends in a verdict, a figure for how sure that verdict can be, what is new in it, and a comparison with what historians have said.

The eighteen cases below are questions historians have argued over for decades or centuries: who wrote a famous poem, where a battle was fought, what killed nine hikers on a mountainside in Russia's Ural Mountains in 1959, and whether a garden that ancient writers counted among the wonders of the world existed, and if so where. Each is examined here from the original documents: charters, court records, inscriptions, shipping lists, newspaper files and manuscripts, most of them now available online.

They stay unsettled because the record is thin, late or contradictory. The earliest surviving account of the Pied Piper was written about a century and a half after the event it describes. The Beale ciphers, three pages of numbers said to give the site of buried gold, exist only in a pamphlet printed more than sixty years after the gold was said to have been buried. The poem that celebrates the battle of Brunanburh names the place, and no source written before the twelfth century says where it was. Historians have taken sides on each question, and every page names the sides and who holds them.

Each case reads the surviving record in full where it exists online, sets out the strongest argument on each side, and ends in a verdict with a figure for how sure it can be, such as "about 85 in 100". The figures are judgements rather than measurements, and a label beside each verdict says how strong the case is; the notes under the table explain both. Every case links to its full report and to the code, data and notes behind it, so that any claim can be checked against the sources.

Of the eighteen cases, two rest on strong evidence, ten are probable, five have a leading candidate and one is unsettled. Seventeen agree with a position some historians already hold; one (The Bixby letter) provisionally shifts part of the accepted account; none overturns a prevailing view.

The mysteryVerdictHow sureEvidenceHistoriansRead
The Beale CiphersWhether the three Beale ciphers of 1885 are genuine or a hoaxA hoax, about 95 in 100: a treasure story written and partly encipheredabout 95 in 100, a composite judgement
Strong
Agrees withGillogly, King, Mateer, Nickell and Kruh; Hammer held the ciphers genuineCase pageFull reportFiles (640)
Where was Brunanburh?Where the battle of Brunanburh, won by Æthelstan in 937, was foughtBromborough, on the Wirral, at roughly 60 in 100, first among several possible sitesroughly 60 in 100 on the weights chosen
Leading candidate
Agrees withCavill, Harding, Jesch, Coates and Downham; Wood, Halloran and Breeze dissentCase pageFull reportFiles (40)
The Last Decade of the Ninth LegionWhere the Ninth Legion ended, after its last dated text of 108Probably ended in Britain, about 122 to 130; about 50 to 65 in 100about 50 to 65 in 100, central readings
Probable
Agrees withHodgson, Graafstal and Ritterling; other positions from Campbell, A.R. Birley and KeppieCase pageFull reportFiles (33)
The Princes in the TowerWhether Edward V and his brother died in the Tower in 1483Both died in Richard III's custody in the Tower, 1483: about 85 in 100about 85 in 100
Probable
Agrees withRoss, Horrox, Pollard, Hicks, Cunningham, Horspool and Skidmore; Langley opposedCase pageFull reportFiles (9)
The Hanging GardensWhere the hanging garden stood: Babylon, Nineveh, or no single structureBabylon, more likely than not, 55 in 100; no single structure 35; Nineveh 10about 55 in 100
Probable
Agrees withBichler and Rollinger, Bagg, Potts and Pedersén; Dalley opposedCase pageFull reportFiles (20)
The Ourang MedanWhether the Ourang Medan lost her whole crew in 1947 or 1948Fabricated: written by Silvio Scherli in Trieste in 1940; invented outright about 85%about 85 in 100 invented outright
Strong
Agrees withPaijmans's 2021 account, whose paper trail it completes; Hargraves and SkeptoidCase pageFull reportFiles (100)
Dyatlov PassWhy nine skiers left their tent in 1959, and what crushed threeCold killed six and snow crushed three; natural causes about 90 in 100about 90 in 100 natural causes
Probable
Agrees withIvanov's 1959 verdict and the prosecutors' 2020 ravine placement; Buyanov dissentsCase pageFull reportFiles (82)
The Pied Piper of HamelinWhat happened to Hamelin's 130 children on 26 June 1284Probably a real loss, 80 in 100; recruited emigration east ranks first at 4280 in 100 a real loss; emigration 42
Leading candidate
Agrees withDobbertin, Udolph, Humburg, Mieder and the Museum Hameln; Hucker opposedCase pageFull reportFiles (11)
Who wrote "A Visit from St. Nicholas"?Who wrote the poem that begins 'Twas the night before ChristmasClement Clarke Moore wrote it, about 85 in 100; Livingston 12, someone else 3about 85 in 100 for Moore
Probable
Agrees withthe traditional attribution, Nissenbaum, Kaller and Norsworthy; Foster and Jackson opposedCase pageFull reportFiles (266)
Nefertiti and SmenkhkareWho ruled between Akhenaten and Tutankhamun, and was King Neferneferuaten NefertitiNefertiti reigned as King Neferneferuaten, about 89 in 100; Smenkhkare a separate manabout 89 in 100
Probable
Agrees withDodson, Allen and Kawai; Reeves, Gabolde, Laboury and Krauss opposedCase pageFull reportFiles (39)
The dancing plague of 1518What caused the Strasbourg dancing of 1518, and whether anyone diedA behavioural epidemic among the poor, about 85 in 100; ergot rejectedabout 85 in 100
Probable
Agrees withClementz, Bischoff, Midelfort and Martin; shifts the popular account (Waller, Wikipedia)Case pageFull reportFiles (188)
The Man in the Iron MaskWho was the masked prisoner who died in the Bastille in 1703The valet delivered to Pignerol in 1669, entered as Eustache Dauger: 78 in 10078 in 100
Probable
Agrees withLair, Petitfils, Noone, Sonnino and Wilkinson, at lower confidence; Funck-Brentano opposedCase pageFull reportFiles (82)
The Roanoke ColonistsWhat became of the 1587 colonists left on Roanoke IslandMain body's seat on the Albemarle mainland, about 68 in 100; Croatoan 17about 68 in 100
Leading candidate
Agrees withHorn and Lawler's split-and-absorption reading, at lower confidence; Quinn and Dawson opposedCase pageFull reportFiles (104)
The Great Chicago FireHow the O'Leary barn in Chicago caught fire on 8 October 1871A flame carried into the barn, about 85 percent; the carrier cannot be identifiedabout 85 percent; no candidate above 30
Leading candidate
Agrees withthe Board of Police, Andreas and Smith; differs from Bales on SullivanCase pageFull reportFiles (1759)
The Bixby letterWhether Lincoln or his secretary John Hay composed the 1864 Bixby letterLincoln, narrowly, about 53 percent against Hay's 42about 53 percent for Lincoln
Unsettled
Shiftsthe post-2019 tilt to Hay (Burlingame, Grieve) back to near even, provisionallyCase pageFull reportIndependent check: the blind rerun's reportFiles (296)
Double Falsehood and CardenioWhether Theobald's 1727 Double Falsehood descends from Shakespeare and Fletcher's lost CardenioFletcher wrote the second half, Shakespeare probably part of the first; 64 in 100about 64 in 100
Probable
Agrees withHammond, Taylor and the New Oxford Shakespeare on Fletcher; Stern opposedCase pageFull reportFiles (532)
Hannibal's route over the AlpsWhich Alpine pass Hannibal and his elephants crossed in 218 BCThe Col du Clapier, first at about 40 in 100; Traversette second, Mont-Cenis thirdabout 40 in 100 for the Clapier
Leading candidate
Agrees withPerrin, Walbank, Lazenby, Lancel, Hunt and Hornblower on the Clapier; Mahaney opposedCase pageFull reportFiles (181)
The death of Meriwether LewisWhether Lewis shot himself at Grinder's Stand in 1809 or was murderedLewis shot himself, about 85 in 100; homicide about 15, accident under 1about 85 in 100 for self-infliction
Probable
Agrees withPhelps, Jackson, Cutright, Holmberg and Peck; Fisher, Guice, Starrs and Gale opposedCase pageFull reportFiles (101)

How to read the table

Strong

The evidence leaves little room for another answer. The verdict is still a reading of the documents, and a new document could change it.

Probable

The evidence clearly favours one answer over all the others, but does not prove it. The percentage moves when the same evidence is weighed differently.

Leading candidate

Several answers remain possible and are put in order. The first is ahead of each rival, and the order can change when one judgement changes.

Unsettled

No answer is clearly ahead of the others. The report may still say which one it favours.

Agrees with

The conclusion agrees with a position some historians already hold, not always the prevailing one. The case may raise or lower the confidence, correct a piece of supporting evidence, or explain how the event happened.

Shifts

The broad answer stands, but a central part of the accepted account changes: how the event happened, when it happened, or the weight given to a key piece of evidence.

Overturns

The verdict rejects the prevailing view outright.

How to read the percentages. "About 85 in 100" says how likely the verdict is judged to be on the evidence weighed and the assumptions stated in the report. Where a figure comes from a calculation, the calculation shows what follows from its inputs, and the inputs are themselves judgements. The range beside each figure shows how far it moves when the assumptions change.

What every case does

  1. States a verdict first, with a figure for how sure it is. "Cannot be settled on present evidence" counts as a verdict.
  2. Separates the primary sources from later scholarship and cites every load-bearing claim to something a reader can check: a URL, an edition and page, an archive reference, an inscription number. Where the wording matters, the source is quoted in its original language.
  3. Names the strongest opposing case and answers it, and says what evidence would overturn the verdict.
  4. Says what is new and what earlier scholars had already established.
  5. Keeps the code and data wherever a result was computed, so that it can be re-run.