What happened to Flight 19 is not a Bermuda Triangle mystery. It is a US Navy investigation that was digitised in full, marked Unrestricted, and has been public the whole time. This film reads it - NARA microfilm publication M1657, page by page - and then reads the 1964 magazine article that turned those pages into the Bermuda Triangle.
On the fifth of December, 1945, five aircraft took off from a naval air station in Florida on a training exercise. Fourteen men were aboard. A rescue aircraft went out that evening with thirteen more. Neither the five nor the rescue plane came back.
Twenty-seven men were lost in a single afternoon, and no wreckage from any of the six aircraft has ever been identified. That afternoon is the reason the Bermuda Triangle exists. Every book and every documentary about that stretch of ocean begins there. The United States Navy investigated.
A board of officers took testimony, gathered the radio logs, and set out its findings in fifty-six numbered paragraphs. That record was declassified and has been digitised by the National Archives, page by page, five hundred and forty-nine of them. It is free, it is marked unrestricted, and anyone can read it this afternoon. It names the men, the aircraft, the compass headings they were given, the time each transmission was received, and the position where a tanker watched something burn on the water.
Naval Air Station Fort Lauderdale trained torpedo bomber crews. The aircraft was the Grumman Avenger, the largest single-engine aircraft of the war, built to carry a torpedo in an internal bay. It carried three men: a pilot, a turret gunner behind him,
and a radio operator and bombardier below that. Trainees flew navigation exercises of increasing difficulty. The one scheduled that afternoon was the third and last of the basic series, and the board's findings record it by name: Navigation Problem Number One. The route is written down.
Depart Fort Lauderdale. Fly zero-nine-one degrees — one degree south of due east — for fifty-six miles, to a shoal called Hen and Chickens, and there practise low-level bombing. Continue on the same heading for another sixty-seven miles. Turn to three-four-six degrees, slightly west of due north, for seventy-three miles.
Then two-four-one degrees, southwest, for a hundred and twenty miles, which brings you home. A triangle, flown out over the water and back. They took off at about ten past two and were expected back at twenty-three minutes past five. Fourteen men: six Navy, eight Marines.
The board lists all of them, aircraft by aircraft, with the bureau number of every machine — the serial number stamped on a naval aircraft. One of the five, FT-81, went out with a pilot and one crewman instead of two. The board records why. Corporal Allan Kosnar of the Marine Corps had asked to be excused from the mission.
He lived. The board's findings name the third man on the flight leader's aircraft as Robert Francis Harmon, and separately record Harmon as being in a flight duty status with the Gunnery Department. The Navy's accident card, filled in on a different form, names somebody else. Two official records of the same aircraft, on the same afternoon, disagree about who was in it.
This film will show you both. There are no photographs of these men — not in the National Archives, not in the Navy's collections, not in any free repository. For one of the twenty-seven there is a single small portrait whose licence could not be established; for the other twenty-six, nothing. There is no film of Naval Air Station Fort Lauderdale in the 1940s either.
Where the men appear in this film, they appear as the Navy wrote them down: names, ranks, service numbers, in naval typescript. The four student pilots had roughly three hundred flying hours each, about sixty in the Avenger. Two were Marine Corps captains, one a second lieutenant; the fifth aircraft was flown by a Navy ensign. On a training navigation exercise the students navigate.
The instructor lets them make their own errors, because making them is the point. The instructor is the reserve — the man who knows where everybody is when nobody else does. The instructor was Lieutenant Charles Carroll Taylor,
of the Naval Reserve. Two thousand five hundred and nine flying hours, six hundred and six of them in the Avenger, a year flying out of Key West, and a combat tour as a carrier torpedo pilot aboard the USS Hancock. He had come to Fort Lauderdale from Naval Air Station Miami, and this was the first time he had led this particular problem. He knew the water south of Florida — the Keys, the chain of low islands trailing southwest from Miami.
He did not know the Bahamas, which lie to the east. Taylor arrived late for the briefing. He asked the duty officer to find another instructor, because he did not want to take this flight. He gave no reason.
The record does not say he was ill, or tired, or worried. It says he asked to be replaced, that he offered no explanation, that no other instructor was available, and that the request was refused. Anything else you have heard about that moment was supplied by somebody who was not there. At about ten past two they flew east on the written route, did the bombing exercise at Hen and Chickens Shoals, and turned onto the next leg.
An aircraft in 1945 over open water navigated by dead reckoning. You know where you started, your compass heading, your airspeed and how long you have been flying, and from those you calculate where you must now be. It is arithmetic, and it works as long as your starting assumption is right. If you are wrong about where you are when you begin, every correct calculation afterwards takes you further from the truth.
At about ten past four, Taylor transmitted this: Both my compasses are out and I am trying to find Fort Lauderdale, Florida. I am over land but it is broken. I am sure I'm in the Keys but
I don't know how far down and I don't know how to get to Fort Lauderdale. Broken land below him — small islands with water between them. He was over the Bahamas, northeast of Fort Lauderdale. He read them as the Keys, southwest of Fort Lauderdale.
If you believe you are over the Keys, Florida is to the north and east, and you fly northeast to reach it. Over the Bahamas, Florida is to the west, and flying northeast takes you into the open Atlantic. Taylor was not drifting. He was navigating carefully in a direction that was exactly wrong, and every correction he made was a correction in the wrong direction.
Both my compasses are out. An Avenger had two. A magnetic compass, which points at magnetic north, swings in a turn and takes time to settle. And a gyro compass, steadier but drifting over time, to be reset against the magnetic one periodically.
Lose confidence in both and you have no absolute reference — only the sun, which was going down, and what you believe about where you started. Three further details are in the record. He may have had no watch, and the aircraft had no clock. He was heard asking the others how long they had been on a heading, which is the one number dead reckoning cannot do without: without elapsed time a heading cannot be converted into a distance.
He was asked to switch to the search and rescue frequency, which would have given him a cleaner channel to the shore stations. He declined — changing frequency risked scattering the formation, and he had four inexperienced pilots behind him in deteriorating weather. His emergency identification transmitter, which makes an aircraft show up distinctly on radar, was not switched on. Each was a reasonable decision or an ordinary oversight in isolation.
Together they removed, one by one, the ways the shore could have found him. At eleven minutes past four, Fort Lauderdale understood that the flight leader was uncertain of his position — over two hours after take-off, and roughly when they should have been turning for home. Another aircraft, flown by a Lieutenant Cox, made contact; the board records the exchange as a formal finding. Cox told him to put the sun on his port wing and fly north.
Over the Bahamas in the late afternoon that is correct, and north eventually gives you Florida. From the Keys, the same instruction takes you into the Gulf. At forty-seven minutes past four, the shore stations were asked to start taking bearings. The weather was worsening and the light was going.
High-frequency direction finding works like this. When an aircraft transmits, a ground station with the right aerial can measure the direction the signal arrived from — not how far away, only which way. One station gives you a line, and the aircraft is somewhere on it. Two stations give you two lines, and the crossing point is the aircraft.
Six stations enclose a small patch of ocean. By six o'clock that evening, six of them had bearings on Flight 19. The fix put the flight north of the Bahamas and east of Florida, inside a circle of about a hundred miles' radius. The system had been listening to a man who kept transmitting because he was lost, and his transmissions were what allowed it to find him.
The board recorded what happened next. No one thought to transmit the fix in the blind for Flight 19 to hear. In the blind means broadcasting without knowing whether anyone
is receiving — you send it out and hope somebody is listening. It required no equipment they did not have and no permission they could not get. Somebody would have had to decide to do it, and nobody did. On the radio, two of the students were heard saying this to each other: Dammit, if we could just fly west we could get home; head west dammit.
West was correct, and had been correct from the moment they crossed the Bahamas. They said it on an open frequency that shore stations were listening to. The formation stayed with its leader. At twenty past six a Coast Guard aircraft went up from Dinner Key to regain contact.
It did not find them. The last transmission the Navy logged came at about four minutes past seven. The accident cards timestamp the loss at approximately nineteen-oh-four. They had taken off with fuel for about four hours in the middle of the afternoon.
By then they had been in the air for five. The Coast Guard's Seventh Naval District opened a file — number 2273, headed "Five TBM's, Naval Air Station Fort Lauderdale, Lost." Forty-five aircraft flew the search on the first morning. Fifty-one flew it that afternoon. Ships and aircraft worked the area from the sixth of December to the tenth.
One entry in that log: a raft of survivors was reported sighted. It was investigated. It was packing boxes. Thirteen of the twenty-seven men lost that day were not on Flight 19, and what happened to them is one of the better-documented aircraft losses of the war.
At about half past seven that evening a PBM Mariner flying boat left Naval Air Station Banana River to search. Bureau number 59225, callsign Training 49, thirteen men aboard, fuel for twelve hours. The pre-flight inspection is in the record, including the line that it showed no indication of petrol fumes. The PBM's service nickname was the flying gas tank.
Its last contact was a routine departure report at about half past seven. The position report due an hour later never came. Two witnesses recorded what happened, independently, and the board printed both. The tanker SS Gaines Mills: At 0050 Greenwich, observed burst of flames, apparently explosion, leaping flames one hundred feet high, burning ten minutes.
Position twenty-eight degrees fifty-nine minutes north,
eighty degrees twenty-five minutes west. At present passing through big pool of oil. Stopped, circled area using search lights, looking for survivors. None found.
The escort carrier USS Solomons, watching on radar: Our air search radar showed plane after takeoff from Banana River last night joining with another plane, then separation and proceeding on course zero-four-five degrees, at exact time SS Gaines Mills sighted flames in exact spot the above plane disappeared from the radar screen and never reappeared. The tanker did not know it was watching an aircraft; it reported a burst of flames at a position, stopped, and looked for survivors. The carrier was reading a radar screen and did not know what the tanker had seen. The two accounts agree on the minute and on the spot.
The board reached no formal conclusion on the cause, but its questioning went repeatedly to petrol fumes in the hull and smoking regulations aboard PBMs. The Navy's historians say the aircraft probably caught fire in flight. Searches from the sixth to the tenth of December found no debris and no crew, for either the Avengers or the Mariner. Heavy seas prevented the site being marked.
The board's conclusion about the flight leader, in its own words: he had become so hopelessly confused as to have suffered something akin to mental aberration. On the Navy's accident cards, the cause of the loss of five aircraft and fourteen men was entered in three words. Cause: poor navigation. It fits the radio logs, the geography and the direction-finding fix.
Charles Taylor's mother did not accept it. Katherine Taylor's objection was evidential: there were no bodies, no aircraft and no wreckage — nothing had been recovered from any of the five machines — and on that evidence the Navy could not fairly attach blame to her son's name. She engaged an attorney, conducted her own investigation, and petitioned the Board for Correction of Naval Records. That board is not an appeal court and does not re-investigate accidents.
It exists to correct injustices in a serviceman's official file — a wrong date, a missing decoration, a finding the evidence does not support. Its question is never what happened. Its question is whether the record treats the man fairly on the evidence. In the years that followed it found —
in its own phrasing — that an injustice is found in subject officer's record under applicable standards of Naval Law. Taylor was cleared of blame, and the finding in the file was changed. Where it had read poor navigation, it now read: The cause of the accident remains unknown. Nothing new had been found.
No wreckage had surfaced, no witness had come forward, no fresh transmission had been recovered. The change was a ruling about fairness to the reputation of one dead officer. Twenty years later, writers looking for material found a United States Navy file on the loss of five aircraft and fourteen men that concluded in three words: cause unknown. They did not mention the mother, the attorney, the Board for Correction of Naval Records, or that the Navy's investigators had said plainly what they thought had happened.
The words Bermuda Triangle did not exist until February 1964. They were written by Vincent Gaddis, in Argosy — an American pulp magazine — in an article titled "The Deadly Bermuda Triangle." Draw a line from Florida to Bermuda, another from Bermuda to Puerto Rico, and a third line back to Florida through the Bahamas. Within this area, known as the Bermuda Triangle, most of the total vanishments have occurred. At the moment those words were printed it was known as that by nobody, because he had named it in the same sentence.
Naming a region is not neutral. Once the shape exists, every loss inside it is evidence for the shape and every loss outside it is somebody else's problem. The boundary does the argument's work before the argument begins. Gaddis claimed hundreds of ships and aircraft over two decades and roughly a thousand lives.
His proposed mechanism was an atmospheric aberration — in his phrase, a hole in the sky. His published list: Flight 19 and the Mariner. The Marine Sulphur Queen. The Sno' Boy.
The Sandra. The Rubicon. The yacht Connemara the Fourth. The USS Cyclops.
A DC-3 airliner in 1948. A Super-Constellation in 1954. A KB-50 tanker in 1962, two KC-135s in 1963, a Navy patrol bomber in 1956.
And the Star Dust — a British airliner that disappeared in 1947 on a flight to Santiago, found in 1998 in a melting glacier on Mount Tupungato, in the Andes, in Argentina, some four and a half thousand miles from the triangle Gaddis had just drawn. One claim has been cut from this film. You will often hear that Lloyd's of London charges no extra premium for ships crossing this water. No citable statement from Lloyd's saying it could be found.
It may be true. Using an unverifiable claim to attack unverifiable claims is the error this film is describing. Three organisations have a documented position, and all three are readable today. The National Oceanic and Atmospheric Administration states that there is no evidence that mysterious disappearances occur with any greater frequency in the Bermuda Triangle than in any other large, well-travelled area of the ocean.
The Naval History and Heritage Command — the Navy's own historians, writing about their own missing aircraft — call the Bermuda Triangle a myth. And the United States Coast Guard, the service that would be sent if a ship went missing there, has a page of frequently asked questions. One question is whether the Bermuda Triangle really exists. This is the entire answer: The Bermuda Triangle, or Devil's Triangle, is a mythical geographic area located off the southeastern coast of the United States.
It is noted for an apparent high incidence of unexplained losses of ships, small boats, and aircraft. The Coast Guard does not recognize the existence of the so-called Bermuda Triangle as a geographic area of specific hazard to ships or planes. In a review of many aircraft and vessel losses in the area over the years, there has been nothing discovered that would indicate that casualties were the result of anything other than physical causes. No extraordinary factors have ever been identified.
The first adjective is mythical. By the mid-1970s the Triangle was an industry. The cases had been copied from one author to the next for a decade, each retelling taking the previous one on trust. In 1975 a research librarian at Arizona State University named Lawrence David Kusche went back to the sources — contemporary newspapers, weather records, official casualty reports — and checked the cases one at a time.
His book is under restricted access and could not be read for this film, so what follows is what the contemporary record shows, which anyone can check. Kusche is credited with the method: going back and looking. The errors sort into kinds. The wrong ocean.
A vessel called the Freya appears in Triangle books as having sailed from Manzanillo, Cuba. She sailed from Manzanillo, Mexico — a port of the same name, on the Pacific coast. The incident that never happened. An aircraft is described as lost off Daytona Beach in 1937.
The newspapers of 1937 record no crash, no search and no report. The storm retold as a calm-sea vanishing. The Japanese freighter Raifuku Maru is cited for decades as a Triangle disappearance, always with the same detail — her final radio message, a sentence quoted in dozens of books: Danger like a dagger now. It appears in the radio logs nowhere.
What she sent was: Now very danger. Come quick.
And she did not vanish. She sank off Nova Scotia — roughly a thousand miles north of the triangle — in heavy weather, with the liner RMS Homeric standing by, trying and failing to launch boats in the sea state. Her sinking was watched from close range. Then there is the case where the official report is quoted accurately, and then stops being quoted.
On the thirtieth of January, 1948, a British airliner called the Star Tiger disappeared between the Azores and Bermuda with thirty-one people aboard. She is in every Triangle book, introduced the same way: an aircraft that vanished from a calm sky with nothing wrong with her. The British government held a court of inquiry under a judge, Lord Macmillan, sitting in public at Church House in Westminster for eleven days. It published its report as a command paper, Cmd.
7517, in September 1948. Sixty pages. It is online and free. The books quote its last sentence.
What happened in this case will never be known, and the fate of Star Tiger must remain an unsolved mystery. That is what the court wrote. Here is what is in the other fifty-nine pages. She took off about nine hundred pounds over her maximum permitted weight, because her commander had insisted on full tanks.
She crossed nineteen hundred and sixty nautical miles of open Atlantic at two thousand feet, the whole way, because the headwinds were weaker down there. The court's comment on that choice: the lower the altitude, the less opportunity there is for dealing with any emergency. The route routinely came close to running aircraft out of fuel. In the court's words, the winds not infrequently contain such a headwind component as to prevent a Tudor aircraft with full tanks from reaching Bermuda either with the customary two-hour reserve — or at all.
The Star Tiger had twice before diverted to Newfoundland. Two months earlier another aircraft of the same type landed at Bermuda with its fuel gauges reading nil, having alerted air-sea rescue and prepared to ditch. The wind forecasts for the critical stretch were wrong by a factor of two: Bermuda gave twenty to twenty-five knots; the actual wind, calculated by another aircraft on the same track, was forty-eight, from a different direction. That other aircraft matters.
A second British airliner flew the identical route one hour ahead of her and landed safely. Its navigator found he had been carried sixty-eight miles north of his intended track in a single hour. The court found — its phrase — that in this flight there was want of care and attention to detail. Star Tiger's hourly position reports gave her altitude as twenty thousand feet for the entire crossing.
She was at two thousand. None of that solves it, and the court said so. An overloaded aircraft at low altitude on a marginal route in a wind nobody had predicted correctly is a list of ways to die, not a demonstration of which one happened. The mystery and the mundane explanations are in the same document; six words of it are famous and the other sixty pages are not.
The largest case on Gaddis's list is the oldest. On the fourth of March, 1918, the USS Cyclops left Barbados for Baltimore carrying manganese ore and three hundred and nine men. She never arrived. It is the greatest loss of life in the history of the United States Navy that did not involve combat.
The Secretary of the Navy wrote about her in his annual report that year: There has been no more baffling mystery in the annals of the Navy than the disappearance last March of the USS Cyclops. A government document, public domain, and where the Cyclops enters the literature — a Secretary of the Navy, in print, using the word mystery. Germany's naval records were examined after the war. There was no submarine in that part of the Atlantic, which removed the explanation everyone had assumed.
The Navy's historians say today that she probably sank in an unexpected storm. The National Archives holds, for that ship, her stability calculations — the naval architect's booklet showing how much the hull could be loaded and how far it could heel before it would not come back up. Cross curves of stability. Righting arms.
Body plan. Seven numbered sheets. It also holds a United States Weather Bureau chart of the weather she sailed into. She was carrying manganese ore, which is dense and which shifts.
She had a cracked cylinder on one engine, and she was overloaded. The calculations that describe how a ship like that behaves in weather like that have been in the public archive for a century, in the same building as the legend that says her loss cannot be explained. The explanations were sitting in public archives the entire time. Sitting in an archive is not the same as being read.
This photograph — five Avengers in formation over open water, in wartime paint — is the picture of Flight 19. It is on the book covers, it is the thumbnail on the videos, it runs behind the narration in the documentaries. It is not Flight 19. It is squadron VGS-29, over Norfolk, Virginia, on the first of September, 1942 — three years and three months before Flight 19 took off, seven hundred miles up the coast, a different unit, whose crews landed safely.
The Navy catalogued it properly. It carries a photo number, eighty-dash-G-four-two-seven-four-seven- five, and a caption naming the squadron, the place and the date. The correct information has been attached to the image the entire time. A second version circulates, colourised.
A third is not a photograph at all — a digital illustration made in 2008, whose artist describes it as a homage to the wartime image, redrawn so that squadron numbers could be added. That reconstruction now appears in places as though it were the photograph: a picture of Flight 19 manufactured out of a picture that was already not Flight 19. Nobody faked the original. Somebody needed a picture of five Avengers in formation, used the one that existed, and everybody afterwards took it without asking where it came from.
It is the same mechanism as the dagger sentence, the boundary on the map, and the ship from the wrong ocean: a convenience, repeated until it hardened.
In 2026 a story went round the internet — the same paragraph, in slightly different words, across dozens of sites. A NOAA oceanographic study had mapped enormous fields of methane seeps on the sea floor beneath the Bermuda Triangle, and rising gas had been sinking ships all along. Methane hydrate is real: natural gas locked into an ice-like solid in cold sea-floor sediment. If a large volume comes out at once the water above becomes frothy — less dense — and a ship floating in it can lose the buoyancy holding it up.
One peer-reviewed paper stands behind the idea, a 2003 study in the American Journal of Physics, done in a glass tank and by computer model, which asked whether a single bubble could sink a ship. It found that one can, if the hull sits in a narrow band between two points in the rising plume. The work was motivated by the North Sea. There is no such NOAA study.
Not disputed — absent. No digital object identifier, no NOAA product number, no press release, no expedition report, no trace of it in NOAA's or the Geological Survey's publication records. The articles credit a research team led by a named scientist at two well-known institutions. That scientist cannot be found to exist.
What does exist is a real piece of work, published in Marine Geology in May of 2024, which mapped about eleven hundred methane seep sites along the United States Atlantic margin. It says nothing about the Bermuda Triangle, or about any ship or aircraft. Ninety-eight per cent of the seeps lie between Cape Hatteras and southern New England — several hundred miles north of the triangle as Gaddis drew it in 1964. And the water where Flight 19 actually was, over the Bahama Banks, is far too shallow for methane hydrate to be stable.
A real study, from the wrong year, about the wrong stretch of coast, attached to a mystery it never mentions, by writers who did not read it. Not a place. A sixty-two-year-old habit of not checking, which has never stopped, and which produced a new artefact this year. The men are the part that was always real.
Twenty-seven of them, on the fifth of December, 1945, and a file in the National Archives that says, in fifty-six numbered paragraphs, what happened to them. NOAA's own page, listing the explanations people reach for, files the methane idea under a phrase worth ending on. They call it more grounded in science — if not in evidence.
Wonder Minute