The Last Known Position
Where The Trail Ends
From a deserted lighthouse to a missing airliner, six famous cases reveal how much can be known without knowing the ending.
The ship was still there. Its people were not.
When men from the Dei Gratia boarded the Mary Celeste in December 1872, they found a vessel that could be brought to port, but nobody to explain why it had been abandoned. That surviving ship made the disappearance harder to understand. If the vessel had sunk, the broad outline of the disaster would have seemed more obvious.
The world’s most mysterious disappearances often contain this uncomfortable mismatch: enough evidence to establish that something happened, but too little to reconstruct the decisive moment. They are rarely cases in which nothing is known. More often, the missing piece is unusually important.
Six cases show the difference: the Mary Celeste, the Flannan Isles keepers, Roanoke’s colonists, Amelia Earhart and Fred Noonan, D. B. Cooper, and Malaysia Airlines Flight MH370. These are not an objective ranking of human suffering or strangeness. They are enduring mysteries with substantial records, consequential unanswered questions and sharply different reasons why an answer remains difficult.
The Mary Celeste: Why Leave A Ship That Stayed Afloat?
On 4 December 1872, the Dei Gratia encountered the Mary Celeste near the Azores. Captain Benjamin Briggs, his wife Sarah, their young daughter Sophia and seven crew members were missing. The vessel had sailed from New York towards Genoa carrying industrial alcohol.
The smaller boat was gone, as were the ship’s papers and the captain’s navigational instruments. Personal belongings remained. There was water inside the vessel, damaged rigging and a partly dismantled pump. Those details support abandonment, but they do not establish what made it necessary or where the people subsequently went.
The distinction is crucial. A ship found afloat is evidence that it survived until discovery. It does not establish that somebody standing aboard days earlier could confidently predict that outcome.
Imagine the decision in its proper order, without pretending to know Briggs’s thoughts. A captain must judge the condition of a vessel using the information available at the time. If that information suggests an immediate threat, taking to a boat might appear safer. The later survival of the ship cannot tell us exactly what warning prompted the decision or whether it was correctly understood.
This is a plausible framework for interpreting abandonment, not a reconstruction of what Briggs actually did. It explains why the apparently decisive question — why abandon a seaworthy ship? — contains an assumption. We know the vessel’s eventual fate more clearly than we know the crew’s assessment of it.
Investigators considered foul play, but the salvage proceedings did not establish a conclusive explanation. The absence of an agreed answer leaves room for competing mechanisms, not an unlimited licence to invent events.
Any convincing solution must explain two stages. First, why did everyone leave? Second, why did they fail to return or reach safety? A theory that addresses only a frightening condition aboard the ship is incomplete. So is one that explains how a small boat might be lost without accounting for why it was launched.
The surviving vessel is therefore both the case’s greatest clue and its most misleading invitation. It lets us inspect the aftermath while withholding the emergency, real or perceived, that may have preceded it.
The Flannan Isles: An Empty Lighthouse And A Violent Shore
On 26 December 1900, a relief visit to the Flannan Isles lighthouse, off Scotland’s west coast, found three keepers missing: James Ducat, Thomas Marshall and Donald McArthur. Joseph Moore went ashore and discovered the deserted station.
The Northern Lighthouse Board preserves accounts from Moore and Superintendent Robert Muirhead. These describe a properly prepared light and serious damage around the western landing. Railings had been broken or displaced, and equipment had been washed from its usual position.
The official explanation was that the men had gone outside, possibly to deal with equipment or damage, and were swept away by a large wave. Their bodies were not recovered. The precise sequence remained unknown.
That is a substantial explanation, even though it is not a witnessed account. It connects the disappearance to a dangerous location and physical evidence of powerful seas. A supernatural theory does not become equally credible merely because the men’s final movements cannot be established.
Yet it would also be an overstatement to say that every important question has been solved. Why were all three exposed? Did they go out together, or did one follow the others? Was someone attempting to help? These are possibilities to investigate, not scenes that can responsibly be narrated as fact.
The case makes an important distinction between knowing the likely cause and knowing the full event. An investigation can reasonably favour accidental loss at sea while remaining unable to explain the order in which three people left a building.
It also shows why a peaceful interior need not contradict a fatal accident outside. An orderly room records the condition in which it was left. It offers no guarantee about what happened beyond its walls. The contrast is emotionally powerful, but the contrast itself is not evidence of an extraordinary cause.
A good account of Flannan should leave the reader with a smaller, sharper mystery than the one with which they arrived. The unanswered question concerns the men’s movements and the immediate circumstances of their deaths. It does not require us to discard the documented power of the sea around them.
Roanoke: A Disappearance That Included A Destination
Roanoke is often imagined as an English settlement erased from the world. The evidence is more awkward for that version: the people who left appear to have supplied a clue about where they were going.
John White returned to Roanoke Island in August 1590 after leaving for England in 1587. His efforts to bring assistance had been delayed by the conflict with Spain and the difficulty of securing passage. The absence of regular contact created a long interval in which the settlement’s circumstances could have changed.
White found the site abandoned. The letters “CRO” were carved into a tree, and “CROATOAN” into a palisade. Croatoan referred to both an Indigenous people and their island, now Hatteras Island. White understood the message as an indication of relocation, but weather frustrated his attempt to follow the clue.
This was not an incomprehensible word left for future generations to decode. Within the circumstances of the colony, it had a recognisable geographical meaning.
That changes the central question. Instead of asking how an entire community could disappear instantaneously, we should ask where its members moved, whether they remained together and what happened after their departure. Those questions allow for a process unfolding over months or years.
The National Park Service considers several explanations, including movement inland and relocation to Croatoan. European objects discovered in relevant areas are suggestive, but trade offers another route by which objects could travel. An artefact associated with English settlement does not automatically establish that a particular missing colonist lived beside it.
This is the central archaeological difficulty. A find must do more than fit the right broad period. To settle the mystery, it would need a sufficiently secure context and connection to distinguish these settlers from other visitors, exchanges or occupations.
Nor should a proposed answer assume that every colonist shared the same fate. A community under pressure could separate. Some people could survive while others died. That is a possible model, not an established history, but it exposes the weakness of demanding one dramatic event to explain every missing individual.
There is also a question of perspective. People can disappear from one record while remaining present in another community’s life. If some colonists joined Indigenous households, their absence from subsequent English accounts would not mean that they vanished from human society.
The evidence does not let us turn that possibility into a complete survival story. It does, however, make relocation a serious explanation. Roanoke remains unresolved because a meaningful departure clue has not been followed by an equally secure account of arrival and later lives.
Amelia Earhart And Fred Noonan: When A Promising Shape Is Not A Plane
On 2 July 1937, Amelia Earhart and navigator Fred Noonan left Lae for Howland Island during their attempt to fly around the world. They never reached their intended destination. The subsequent search failed to locate them or establish the aircraft’s final resting place.
The broad alternatives include a crash at sea and the possibility that they reached land elsewhere. The Nikumaroro castaway hypothesis proposes that they landed on the island’s reef and survived there for a time. It remains a hypothesis, rather than a confirmed account of their last days.
The search produced a particularly useful development in October 2026. An expedition involving the Archaeological Legacy Institute, Purdue Research Foundation and Purdue University examined a feature known as the Taraia Object, previously identified in imagery of Nikumaroro.
On 19 October local time, the team announced that it had found no conclusive evidence of Earhart’s aircraft in the lagoon. Its assessment was that the apparent object was likely accumulated trees, with a reflective coating that had helped create the impression of metallic material. The team continued to support the wider castaway hypothesis, but this particular target had not delivered the hoped-for wreckage.
That outcome deserves attention because it is genuine progress of a less marketable kind. A specific candidate was investigated and weakened. An unresolved case had acquired a more definite answer to one of its subsidiary questions.
It does not follow that the entire island hypothesis has been disproved. Equally, the survival of that broader hypothesis cannot rescue the rejected identification. Each claim has to bear its own evidential weight.
The same discipline applies to historical records. The US National Archives’ guide to Earhart material explicitly includes unverified reports among its holdings. A document being preserved in an official archive establishes that the document exists. It does not make every allegation inside it true.
For a convincing identification of the aircraft, resemblance alone would be inadequate. Investigators would need features that establish what the object is and connect it securely to Earhart’s plane. The more ambiguous the image, the more work must be done by physical inspection and context.
This is why the case can remain open without every new claim deserving equal enthusiasm. Earhart and Noonan’s disappearance presents a geographical problem, but also a problem of identification: how to distinguish evidence of their journey from the countless other things that can be found in an ocean or on an island.
D. B. Cooper: Escaping The Aircraft Was Only The First Stage
On 24 November 1971, a man using the name Dan Cooper boarded a flight from Portland to Seattle. He threatened the crew with what he said was a bomb and demanded $200,000 and parachutes.
After the passengers were released in Seattle, the aircraft took off again with crew members still aboard. Cooper subsequently jumped from the aircraft with the ransom. His identity and ultimate fate remain unresolved in the FBI’s public account.
The FBI has highlighted the hazards of the jump, including darkness, unsuitable clothing and a parachute that could not be steered. In 1980, a child discovered $5,800 in deteriorated notes whose serial numbers matched the ransom. The money established a physical connection to the crime without settling what had happened to the man.
The familiar description of a successful escape therefore goes further than the evidence. Cooper succeeded in leaving the aircraft. Survival, recovery of the money, movement away from the landing area and a subsequent life under another identity would each require additional steps.
It is tempting to make the leap itself proof of expertise. But willingness to attempt a dangerous act does not establish the ability to complete it. Conversely, the danger does not prove that Cooper died. Both judgements need evidence beyond the boldness of the plan.
The recovered banknotes create a similar trap. Their authenticity answers which crime the money came from. It does not, by itself, answer how the notes reached the discovery site or whether Cooper was alive when they did. An object can preserve a connection while losing the history of its movements.
A proposed suspect therefore needs more than a resemblance to a composite image or a biography containing aviation experience. The case calls for a reliable connection to this flight and its evidence, assessed alongside details that might exclude the person.
The difference between a striking candidate and a theory supported by evidence becomes especially important when real people and their families are named. A compelling life story can supply a character for the mystery without supplying its solution.
Cooper’s disappearance also resists the reassuring assumption that knowing someone’s name would explain everything. Identification could settle who boarded the aircraft while leaving the landing, the money and the final outcome uncertain.
MH370: A Missing Aircraft With A Surviving Evidence Trail
Malaysia Airlines Flight MH370 disappeared on 8 March 2014 while travelling from Kuala Lumpur to Beijing with 239 people aboard. It remains a profound challenge to the expectation that a modern airliner’s journey should be recoverable in full.
But “vanished without a trace” is an inaccurate description. Satellite communications data helped investigators assess the aircraft’s later movements towards the southern Indian Ocean. Aircraft debris found on Indian Ocean shores also informed the search. The missing main wreckage and flight recorders leave the crucial gaps.
The Australian Transport Safety Bureau explains that its search relied on modelling derived from the satellite communications and other evidence. Its underwater operation covered more than 120,000 square kilometres. The aircraft was not located, and the Malaysian safety investigation could not reach a definitive conclusion about what happened without the missing evidence.
The scale of that effort creates an understandable question: how can such a large search fail? The answer begins with the difference between a large area and the correct area. Search coverage matters in relation to where the target could plausibly lie, not simply as an impressive total.
Consider a simplified comparison. Searching every room in the wrong building thoroughly does not compensate for uncertainty about the address. Ocean searches are more complex, but the underlying distinction remains useful: the quality of the location estimate and the quality of the inspection are separate requirements.
There is also no automatic route from reconstructing movement to establishing intention. Data about where an aircraft travelled cannot alone explain who controlled it, what conditions existed aboard or why particular actions occurred. Accusations against named individuals require evidence beyond an interpretation of the route.
The search has continued to develop. On 29 June 2026, Malaysia’s transport ministry announced an extension of its agreement with Ocean Infinity, running from 1 July 2026 to 30 June 2027. The statement identified a remaining search area of approximately 7,429 square kilometres and a planned operational window between November 2026 and April 2027. It retained a $70 million payment conditional on finding the wreckage.
As of 27 October 2026, that announced window was still ahead. An agreement to search is not a discovery, and a scheduled deployment should not be described as completed work.
Finding the main wreckage would be enormously important. It could establish the location, enable examinations and potentially recover evidence about the aircraft’s final condition. It would not guarantee that every question about the preceding hours could be answered.
That is not a reason to diminish the search. It is a reason to be precise about its value. Locating the aircraft would replace a major uncertainty with physical evidence, even if explaining the entire sequence required further work or remained partly impossible.
Why More Detail Does Not Always Mean More Knowledge
These cases expose a weakness in how disappearance stories are often told. A narrative can become increasingly detailed while the evidence stays exactly the same.
A reconstruction might specify who left first, what somebody feared, why a message was missed and how a final decision was made. If those details are inferred rather than documented, the story has become easier to imagine without necessarily becoming more accurate.
The problem is especially acute when a reasonable general explanation is expanded into an exact sequence. The Flannan keepers may very plausibly have been swept away. That does not authorise a scene in which one man cries for help and the other two rush to rescue him. Such a sequence might fit the outcome, but fitting is not confirmation.
The strongest alternative is not always a rival grand theory. Sometimes it is a modest adjustment: the event took longer, people separated, an object moved independently, or a decision responded to information we no longer possess. These possibilities matter because they test whether the dramatic account is necessary.
The same principle applies to apparently objective images. A camera can capture a departure while leaving the next street unseen. Understanding what footage can and cannot establish helps prevent a recorded moment from being mistaken for a complete explanation. A last known position is a limit of observation, not necessarily the place where the decisive event occurred.
For readers, the useful question is therefore unusually concrete: which part of this account is actually new evidence? A newly published retelling, a more vivid map or a sharper-looking reconstruction may change the presentation while adding nothing to the underlying record.
What Would Count As Solving A Disappearance?
An answer should be judged against the question it resolves. Locating remains, identifying a person, explaining a departure and establishing responsibility are different achievements. A discovery can be decisive for one and inadequate for another.
For Roanoke, an archaeological connection to particular settlers could establish survival at a particular place without accounting for the whole community. For the Mary Celeste, evidence of why the boat was launched would still leave the passengers’ subsequent fate to explain. These are substantial advances, not failures to provide an all-or-nothing ending.
The practical lesson from how cold cases are investigated is that a useful development must be connected to the existing record. A discovery matters because it answers a defined question or rules out a serious possibility. Novelty alone is not enough.
That gives future claims a demanding but fair standard. What exactly was found? How was it identified? Which explanation does it strengthen? What remains unresolved afterwards? A search team should be able to describe an unsuccessful target examination as worthwhile without having to imply that the wider mystery is nearly solved.
There is a human reason to preserve those distinctions. An audience can leave a mystery unfinished and move to the next story. Relatives must live with decisions, anniversaries and public claims attached to the missing person’s name. Uncertainty should not become an excuse to assign them an invented ending.
The next important breakthrough may therefore be less spectacular than the legend demands: a securely identified component, a well-documented object or an overlooked record that narrows a journey. Its value will come from the specific uncertainty it removes. That is the point at which a disappearance begins to yield an answer.
Sources
Royal Museums Greenwich — The Mystery Of The Mary Celeste — People aboard, discovery, condition of the vessel, missing items and the unresolved salvage investigation.
Northern Lighthouse Board — Flannan Isles — Contemporary accounts, named keepers, damage at the western landing and the official explanation.
National Park Service — 1590 Rescue Voyage — White’s delayed return, the carved messages and the interrupted attempt to reach Croatoan.
National Park Service — Major Theories Of The Lost Colony — Relocation hypotheses and the limitations of archaeological finds.
Purdue University — Search For Amelia Earhart’s Plane Continues After Taraia Object Expedition On Nikumaroro Rules Out Latest Hypothesis — Earhart and Noonan’s final journey, October 2026 expedition findings and the team’s stated position on the castaway hypothesis.
US National Archives — Records Relating To Amelia Earhart — The character of the archived material, including search records and explicitly unverified reports.
FBI — D.B. Cooper Hijacking — Hijacking chronology, ransom, parachute jump, recovered money and unresolved identity and fate.
Australian Transport Safety Bureau — MH370 Search Overview — Flight details, satellite and debris evidence, search scale and investigative limitations.
Ministry Of Transport Malaysia — Extension Of The Agreement Between The Government Of Malaysia & Ocean Infinity For The Search Of MH370 Wreckage — Agreement extension, remaining search area, planned operational window and conditional payment, announced on 29 June 2026.
Next Reads
How To Tell Whether A True Crime Theory Is Supported By Evidence — A practical way to assess claims about unresolved cases.
How Cold Cases Are Solved: DNA, Witnesses And The Evidence That Changes Everything — How preserved material and new leads can move an old investigation forward.
How Reliable Is CCTV In A Murder Investigation? What Footage Can—and Cannot—Prove — Why an image can establish a moment without explaining the whole event.
