There is a specific kind of historical case file that the archival profession quietly enjoys, which is the one we have not closed and are not going to. The file is not empty. It contains evidence. It contains witness testimony, technical analyses, provisional theories, and, in several of the cases that follow, a hundred years of subsequent scholarship. What it does not contain is an answer.
I have spent time with each of the six files below, in one library or another. I am not going to pretend that any of them are solved. What I can do is describe what the evidence actually looks like, and what the competing theories actually claim. The confidence with which popular accounts often resolve these cases does not, as far as I can tell, reflect the state of the underlying documents.
1. The Voynich Manuscript
The Voynich Manuscript is a 240-page illustrated codex, written on vellum, that has been carbon-dated to between 1404 and 1438. It is written in a script that does not match any known language or cipher system. It contains detailed botanical illustrations of plants that do not obviously correspond to known species, astronomical or astrological diagrams, and a large number of figures of unclothed women bathing in what appear to be interconnected tubes.
The manuscript surfaces in the documentary record in 1639, in a letter from a Prague-based alchemist named Georg Baresch, who describes having owned it for several decades. It has been subjected, since, to analysis by some of the best cryptographers of the twentieth century, including William Friedman (who broke the Japanese PURPLE cipher in World War II) and John Tiltman (who led Bletchley Park's work on the Fish cipher). Neither made substantive progress.
Three major theories remain on the table. One: the manuscript is a genuine but as-yet-undeciphered text, in an unknown constructed language or an unattested natural one. Statistical analyses of its letter frequencies, word lengths, and morphological patterns are consistent with a real language rather than random glyphs. Two: the manuscript is an elaborate but meaningless hoax, produced to sell to credulous aristocrats who had money for exotic books. The statistical patterns can be produced by certain generative techniques (Gordon Rugg has demonstrated a method using a grille and tables of syllables). Three: the manuscript is a cipher of an underlying text that has been obscured beyond practical reconstruction. I do not know which of these is correct. No one, currently, does.
2. The Amber Room
The Amber Room was a chamber of the Catherine Palace in Tsarskoye Selo, near St. Petersburg, built between 1701 and 1709 as a gift from King Frederick I of Prussia to Tsar Peter the Great. It was paneled, floor to ceiling, in approximately six tons of amber, gold leaf, and mirrors, and was considered one of the great decorative treasures of Europe. It was, by the standards of the time, roughly comparable in value to the Sistine Chapel.
In October 1941, advancing Wehrmacht forces dismantled the Amber Room and shipped it west, in crates, to the Königsberg Castle museum in what is now Kaliningrad. It was displayed there through 1944. In April 1945, as the Red Army approached, it was crated again and, according to the last confirmed documents, stored in the castle cellars. The castle was destroyed in Allied bombing and subsequent Soviet shelling. What happened to the crates is unknown.
Theories include: destruction in the Königsberg fires of April 1945 (most widely accepted by professional historians); removal to an unknown location elsewhere in East Prussia, possibly a salt mine or bunker system (favored by treasure-hunters, supported by occasional but inconclusive discoveries of crates and amber fragments); removal to the United States, the Soviet interior, or Switzerland. The Russian government funded a ten-year reconstruction, completed in 2003, at a cost of approximately twelve million dollars. The original is still, formally, missing.
3. The Antikythera Mechanism
In 1901, Greek sponge divers working a Roman shipwreck off the island of Antikythera recovered, among the cargo, a corroded lump of bronze approximately the size of a shoebox. It sat in the National Archaeological Museum in Athens for fifty years before anyone took a close look. When they did, what they found was a gear mechanism of extraordinary complexity, containing at least thirty interlocking bronze gears, graduated scales, and inscriptions in ancient Greek.
The mechanism has been dated to approximately 100 BC. Subsequent analysis, using computed tomography and reflectance transformation imaging, has established that it was an analog computer capable of predicting lunar and solar eclipses, tracking the 19-year Metonic cycle, and displaying the positions of the five known planets. It was, in other words, a working astronomical predictor approximately fifteen hundred years in advance of anything comparable in the European record.
The mystery is not how it worked (which we now largely understand) but what produced it. No other mechanism of comparable complexity has ever been recovered from the ancient Mediterranean. There are, in the textual record, a few references that might plausibly describe similar devices (Cicero mentions one made by Archimedes and one owned by Posidonius), but we have no other physical example. The Antikythera Mechanism is, as far as the physical record goes, the only one. Which means either there was an entire tradition of Greek astronomical clockwork that has been lost wholesale, leaving behind one accidental survivor, or the mechanism represents a single individual inventor's achievement that did not propagate. Neither possibility is, on its face, satisfying.
We know what the Antikythera Mechanism did. We do not know what it was part of.
4. The Lost Colony of Roanoke
In August 1587, a group of 115 English settlers were left by their supply ship on Roanoke Island, off what is now the North Carolina coast. Their governor, John White, returned to England to arrange additional supplies. The return voyage was delayed for three years by the outbreak of the Anglo-Spanish War. When White finally reached Roanoke in August 1590, the settlement was abandoned. There was no sign of struggle. The word CROATOAN was carved on a palisade post. The letters CRO were carved on a tree. A storm prevented White from reaching Croatoan Island (modern Hatteras), and he never returned.
Four centuries of archaeological and historical investigation have produced several candidate explanations but no definitive answer. The most widely accepted current theory, supported by recent excavations at Hatteras Island, is that at least some of the colonists did, in fact, integrate with the Croatoan people. Pottery and metal artifacts consistent with English settler origin have been recovered from Hatteras contexts dated to the early seventeenth century. Another theory has some colonists moving inland to the Chesapeake and being killed, along with indigenous peoples sheltering them, by the forces of the Powhatan chiefdom in the first decade of the 1600s. The colonists may have split up. Some probably died; some probably assimilated. A definitive answer is unlikely to appear at this point.
5. The Tamam Shud Case
On December 1, 1948, the body of an unidentified man was found on Somerton Beach in Adelaide, South Australia. He was approximately forty-five years old, well-dressed, had recently shaved, and carried no identification. A pathologist ruled the death a probable poisoning, but no poison could be identified by the toxicology of the period. A small tightly-rolled piece of paper was later found sewn into a hidden pocket of his trousers. The paper bore two words in Persian: تمام شد, tamám shud, which mean "it is finished" or "ended."
The phrase was identified as the final line of Edward FitzGerald's English translation of the Rubáiyát of Omar Khayyám. A copy of the Rubáiyát from which the phrase had been torn was located, some months later, in an unlocked car parked near the beach. The book contained a previously-unnoticed set of pencilled letters on the back cover, arranged in five lines, which appear to be a code or cipher. No one has decoded them. It may not, in fact, be a cipher; it may be shorthand, or a memory-aid, or notes to a woman's phone number, or something else. The body was exhumed for DNA analysis in 2021, and provisional identification as a man named Carl Webb was announced in 2022, but this identification is not universally accepted and explains neither the poisoning nor the cipher.
6. D. B. Cooper
On November 24, 1971, a man traveling under the name Dan Cooper (later misreported as D. B. Cooper) boarded Northwest Orient Flight 305 from Portland to Seattle. Shortly after takeoff, he showed a flight attendant what appeared to be a bomb in his briefcase and demanded $200,000 in cash and four parachutes. In Seattle, the passengers were released, the cash and parachutes were delivered, and the plane took off again, as instructed, southbound, at low altitude, with Cooper and crew aboard. Somewhere between Seattle and Reno, Cooper opened the rear airstair, jumped, and was never seen again.
The FBI's investigation, which ran for forty-five years and was formally suspended in 2016, produced no reliable identification of the jumper. Approximately $5,800 of the ransom money was recovered in 1980 on a riverbank along the Columbia, partially decomposed. The remaining money has never surfaced. The case is the only unsolved act of air piracy in U.S. history.
The theories divide into: he survived (and is therefore someone who lived a normal life afterward without ever spending the serial-numbered ransom bills), he died in the jump (the terrain below his likely exit point is heavily forested; a body could easily have gone unrecovered), or he was someone specific whose identity has been proposed and debated (dozens of candidates, none confirmed). There is no physical evidence that settles which of these is correct.
On leaving things open
I do not have a conclusion for this piece, which is appropriate. What I will say is that each of these files has attracted a subculture of amateur investigators whose enthusiasm is a small good in the world and whose theories are often, though not always, wrong. Professional archivists tend to be quieter about these cases, because we are the ones who have to remember that the actual evidence is thinner than it looks, that most of the good leads have already been followed to their ends, and that the remaining probability of a neat solution is, for each file, quite low.
That does not mean nothing. Cases do occasionally break. The Dyatlov Pass incident was probably resolved (by a 2021 glaciological analysis) after sixty-two years. The Somerton Man may eventually be identified. New techniques will uncover evidence we do not currently have. The files stay open because there is a nonzero chance, in each case, that they will not stay open forever.
What I think the archivist's contribution to this conversation is, ultimately, is a reminder that "unsolved" does not mean "empty." Each of these cases is a substantial intellectual object. They are, in their specific incompleteness, some of the most interesting documents we have.