On the 10th of November 1975, the Edund Fitzgerald, a freighter that had once been the largest on the Great Lakes, went missing in a Lake Superior storm.
All 29 men on board died.
No Mayday, no survivors, and no bodies ever recovered.
For almost 50 years, the official story has blamed monster waves and leaking hatch covers.
But when underwater drones were finally sent down to the wreck, they filmed things the original investigators had missed or had no way to see in 1975.
What those cameras recorded has kept one of the darkest mysteries on the Great Lakes alive.
And it challenges a lot of what we believed about that November night.
I’m Ryan Laauo, and of everything those cameras picked up, the detail I keep coming back to is a pair of steel hatch covers.
They belong to the front half of the ship, and they were found lying next to the back half.
Hold on to that picture because the official story has a hard time explaining how they got there.
Here’s the irony at the start of all this.
The thing that made the Fitzgerald such a good earner would later sit right in the middle of the argument over why she sank.
In 1958, though, she was simply the new giant of the lakes.
She was launched at River Rouge, Michigan that June, 729 ft long, and for the next 13 years, nothing on the Great Lakes was bigger.
People in the shipping business called her unsinkable.
In 1969, she hauled 27,42 long tons of cargo in a single trip, which was a record.
And over her career, she set seasonal hauling records six times.
Then the rules around her changed.
In 1969, 1971, and 1,973, the Coast Guard raised her load line, the mark that says how deep a ship is allowed to sit in the water when she’s full.
Each change looks small on paper.
Added together, they let her sail with a little more than 3 ft less freeboard, which is the height of the deck above the waterline than she was originally designed for.
Captain Pquette of the Wilfried Sykes, another freighter, said that change was worth about 4,000 more tons of cargo.
Keep that in mind because on her last night, it meant her deck was riding about 11 12 ft above the water in a storm throwing up waves of 35 ft.
Her captain on that final trip was Ernest Msurley.
He was 63.
He had 44 years of sailing these lakes behind him.
And the story goes that he planned to retire once the season was over.
He’d been in command of the Fitzgerald since 1972, and she wasn’t a ship without scrapes.
You’ll sometimes hear that she’d been lengthened over the years.
She wasn’t.
What did change was her machinery.
She got a bow thruster in 1969 and over the winter of 1,971 into 1,972 her boilers were converted from coal to oil.
She also had a rough stretch around the locks at so Staint Marie.
She ran a ground near the Soolocks in September 1969, collided with another freighter, the Hochalaga, in April 1970, and struck lock walls in 1970, 1973, and 1,974.
At 2:15 on the afternoon of the 9th of November 1975, she pulled away from Superior, Wisconsin, carrying 26,116 long tons of tachinite pellets bound for Detroit.
Less than half an hour later, the National Weather Service issued gale warnings.
A second freighter, the Arthur M.
Anderson, under Captain Jesse Cooper, was heading the same way, and the two ships ended up crossing the lake together with the Anderson trailing behind.
By the next afternoon, the storm had blown far past the forecasts.
Around 3:15, Cooper watched the Fitzgerald come around Caribou Island, and he later said she passed closer to a shallow patch called Six Fathom Shaw than he would have liked.
About 15 minutes after that, Msory came on the radio.
He told the Anderson he had a fence rail down.
He’d lost two vent covers, and the ship had taken a list to starboard.
Two of her six billagege pumps were running non-stop.
He asked Cooper to stay with him.
The wind was brutal.
The Anderson logged northwest winds of 50 knots, about 57 mph.
And a 2005 computer simulation by Noah and the National Weather Service put the sustained winds near where the Fitzgerald went down at close to 70 m an hour with gusts up to 86.
Sailors on the lakes have an old name for the biggest waves in a storm like that, the three sisters, three huge seas rolling in one behind another.
Around 6:30 that evening, Cooper reported two seas of 30 to 35 ft hitting the Anderson.
Now, that radio call is the hinge this whole story turns on.
Something had already hurt that ship by 3:30 in the afternoon, almost 4 hours before she disappeared.
Was it the waves breaking over her?
Or was it something under the water?
Every theory you’re going to hear in this video starts right there with that one call and the drones would eventually be sent down to settle it.
From that moment, the clock was running and it got worse quickly.
Just after 4:10, Msory reported that his radar had failed and the light he needed most wasn’t there for him either.
At about 4:39, the Fitzgerald couldn’t pick up the Whitefish Point radio beacon at all, and she called the Coast Guard station at Grand Marray to ask about it.
Around that time, another vessel reported that the light and the beacon at Whitefish Point weren’t working.
Sometime between 5 and 5:30, Msory told the captain of another ship, the Avaors, that both radars were gone and heavy seas were rolling across his deck.
He called it one of the worst seas he’d ever been in.
So the Fitzgerald was running blind, leaning to one side, taking water and steering for a harbor entrance whose light had gone dark.
At 7:10 that evening, the Anderson radioed over to say it was gaining on her.
The answer from the Fitzgerald was short and calm.
We are holding our own.
Those were the last words anyone ever heard from her.
Around 7:20, she dropped off the Anderson’s radar, and Cooper later said her lights didn’t fade out slowly.
They were there and then they were gone.
The Coast Guard’s own report would later point out that she was about an hour from a safe harbor when she went down.
The search that followed was miserable.
The Coast Guard had no rescue ships close enough, so it asked the Anderson to turn around and head back out into the storm to look.
The first search plane didn’t arrive from Traverse City, Michigan until 10:53 that night.
By the next day, searchers were pulling up life rafts and life jackets.
They never found a single man.
4 days after she sank, on November 14, a Navy patrol plane flown by Lieutenant George Connor picked up a magnetic signal from something huge on the lake bed about 17 miles from the entrance to Whitefish Bay on the Canadian side of the border.
Over the next two weeks, a Coast Guard sidescan sonar survey and then a second survey by a company called Seawward found two large objects lying on the bottom.
And in May of 1976, the first underwater drone finally reached her.
It was an unmanned robot called Curve 3.
And from May 20 to May 28, it worked over the wreck 530 ft down, sending pictures back to the surface.
Those pictures showed a ship in two big pieces.
The front section, roughly 276 ft of it, was sitting upright in the mud, while the stern, about 253 feet, had flipped over, and the two halves were lying about 170 ft apart.
The Coast Guard’s Marine Board published its report on the 26th of July, 1977.
It said the most probable cause was massive flooding of the cargo hold, most likely through hatch closures that didn’t do their job.
And it concluded the ship split in two when she hit the bottom.
Then on the 4th of May 1978, the National Transportation Safety Board put out its own version.
In the Safety Board’s version, one or more hatch covers caved in under giant waves sweeping the deck and the hold flooded suddenly.
But even inside the safety board, not everybody agreed.
One dissenting opinion backed a completely different idea that the Fitzgerald had dragged herself across that shaw near Caribou Island.
Cooper was one of the people who saw it that way from the start.
He believed the Fitzgerald bottomed out on the Shaw and the Lake Carriers Association, the group representing the shipping companies, made the same argument.
The families pushed back on the hatch theory, too, pointing to recent inspections that hadn’t found any serious problems with the hatches, and one detail bothered a lot of people.
The crew had been talking with the Anderson just minutes before she vanished.
Yet, there was no mayday, and nobody launched a lifeboat.
Whatever happened at the end, it happened faster than any man aboard could reach a radio.
The Shaw idea had real evidence behind it.
In 1976, the Canadian Hydrographic Service surveyed the area and found that six Fathom Shaw ran about a mile farther east than the Canadian chart showed.
Those charts were old, too.
The 1,973 chart was based on surveys done in 1916 and 1919.
That’s not a typo.
And it gets worse because the Fitzgerald didn’t carry a fthometer, a depth finder at all.
She wasn’t required to.
So if the water beneath her was getting shallow, there was no instrument in the pilot house that would have warned anyone.
Then there was the ship herself.
In the years after, men who’d sailed on her described a hull that flexed in heavy weather.
Richard Orle, a former second mate, said she bent like a diving board.
A former cook named George Bergner, known as Red, who’d stayed aboard over the winters as her keeper, said in a 1978 deposition that the hatch clamps were loose and the keel was only tack welded.
And according to his account, Msurley waved off his worries about the keel.
A marine engineer, Joseph Fiser, recalled a 1,969 investigation that found cracks in the keel welds.
Cracks that were patched up every year.
And in 2012, two engineers, Sha Kerry and Ben Fischer, put forward a third explanation that starts with those vents Msurley lost at 330.
In their reading, the broken vents let water into the ballast tanks.
The ship took on her list.
The forward hatch covers were eventually blown in by the seas and finally the hull itself failed.
In that version, there’s no rock at all, just a chain of small failures piling up in a ship that was already riding low.
So, you’ve got three theories, and each one points at a different spot on the wreck.
The Shaw theory needs scrapes on the bottom of the hull.
The hatch theory needs to know what happened to those covers.
The structural theory needs to know where she broke apart and when.
All of it was sitting 530 ft down in the dark in water that stays barely above freezing.
That’s why the drones mattered so much.
Here’s what that first robot actually filmed.
Hatch cover number one, the one closest to the bow, was down inside its own hatch, buckled.
Covers two and five were missing.
Cover six, was standing on its end.
Seven and eight were gone, too.
Then the investigators studied the clamps, the heavy fasteners that hold each cover down.
Along the edge of hatch number five, there were 16 clamps in a row, and only two of them, the first and the eighth, were bent or broken.
A cover that’s gone, while most of its clamps look fine, suggests it may not have been clamped down tight, and the Coast Guard’s final report said the hatch covers may not have been properly secured.
If you were one of those families, that was a hard thing to read because it pointed straight back at the crew.
It would be 13 more years before another robot went back with better eyes.
In between, in 1980, Jean Michelle Kustoau’s team made a short submarine visit and claimed the ship broke in two on the surface.
A claim nobody could back up.
What Kustau did come away with was a sense of how well the lake keeps things.
The ship was amazingly beautiful, he said, because it’s deep and cold and almost nothing grows on it.
Then in late August 1989, Michigan Sea Grant put together a team with Noah, the National Geographic Society, the Army Corps of Engineers, and the Great Lakes Shipwreck Historical Society.
Working from a research ship called the Grailing, they sent down a remotely operated vehicle and came home with about 5 hours of video.
The glass in the pilot house windows was still intact.
One pilot house door was standing open and it wasn’t dogged, which means it wasn’t latched shut.
And the team said the damage to the bow was too extensive to have been caused by the storm.
If the waves didn’t do that, the lake bottom did and she hit it with enormous force.
What came next is the part that still doesn’t fit neatly into anybody’s theory.
In July 1994, an independently financed expedition led by Frederick Shannon made seven dives over three days and brought back more than 42 hours of video.
Great Lakes historian Rick Mixster was down there in the submersible and his description is the closest thing we have to a walk across the wreck.
Mixster remembers thinking that if he dropped his camera it would implode at 300 ft and he said that all of them were lost down there.
At first, he says they didn’t even realize they were on the deck because the bow was buried into a slope of mud on the port side.
Then they found a massive crack running across the deck.
And next to the upside down stern, they found two twisted hatch covers numbers two and three.
There they are, the covers from the front of the ship, lying beside the back of it.
Remember, the official finding was that she broke in two when she hit the bottom.
So, how did two hatch covers from near the bow end up next to the stern, which by 1994 was lying about 255 ft from the bow?
To Mixster, those covers broke the belief that she went down in one piece.
He also noticed that most of the Takenite cargo was lying behind the stern with none around the bow and that convinced him she didn’t just nose dive straight down.
Then there’s the stern itself.
It’s lying upside down which means its bottom is facing up and Mixster says it had no visible scratches.
Think about why that matters.
If the Fitzgerald had scraped across a rock shelf, the bottom of the hull is exactly where you’d expect to see the scars and the one stretch of bottom anyone could actually look at was clean.
His team saw no sign of a Shaw strike.
Divers who checked the Shaw itself found no evidence of a recent grounding there either.
Mixster has even suggested the Shaw may not exist at all.
Shannon came away blaming what he called massive and advancing structural failure.
Another team that same summer led by Joseph McKinnis found more damage than anyone had reported before with tachinite pellets scattered across the site and one of its members Tom Farnquist concluded she probably did not break up on the surface.
So, two teams looked at the same wreck in the same summer and came home with opposite answers.
And on one of Shannon’s dives, the cameras found a crewman’s body beside the bow wearing a life jacket, the only one of the 29 anyone has ever seen.
So, what does all that footage really add up to?
Put it next to that 3:30 radio call, and a pattern starts to show.
If she had struck the shaw around 3:15, the damage Msory reported 15 minutes later would fit.
The list.
The pumps running, water getting in somewhere nobody could see.
That’s the case Cooper made.
But the clean stern and the empty shaw pull hard the other way.
And the Carrie and Fisher version explains that exact same radio call without any rock at all.
If those vents were gone, water could find its way into the ballast tanks, and that alone could have given her the list.
Mixurly reported.
Here’s where every version lands in the same place, though.
The Fitzgerald went into that storm, loaded under rules that let her ride more than 3 ft lower than she was designed to.
Her cargo hold wasn’t divided by watertight walls, so water that got in anywhere could spread.
She had no depth finder to warn her about the bottom, and no way to read her own draft, which meant her crew couldn’t measure how much freeboard they’d already lost.
So, when Msurley said they were holding their own, he may well have believed it.
Nothing on board could have told him he was wrong.
And if the Schaw theory is right, the failure wasn’t only on that ship.
It was in charts built on surveys nearly 60 years old.
That’s why the changes that came afterward matter so much.
In 1977, depth finders became required on vessels of 1,600 gross tons and up.
In 1980, survival suits were required and a navigation system called Lauran Sea came in, later replaced by GPS.
The Coast Guard began inspecting ships every year before the November storms, and it took back the 1,973 rule change that had let ships load deeper.
Every one of those fixes goes after something the Fitzgerald’s crew didn’t have that night, even while the argument over the wreck itself carried on.
And one thing plenty of retellings get wrong is who protects the wreck now.
The Fitzgerald lies in Canadian waters, so it’s Ontario’s law that guards her.
Since an April 2005 change to the Ontario Heritage Act, a license can be required to dive, run a submersible, or aim a camera or sonar at a protected site with fines of up to a million Canadian dollars.
In January 2006, the province drew a protected circle 500 meters wide around the wreck.
After the 1994 dives found that crewmen, many of the families, pushed back against any more expeditions.
The one exception they wanted was the bell.
On the 4th of July 1995, it was raised and replaced with a replica carrying all 29 names, and the families lobbyed for the recovery on the condition that no video of any bodies would be recorded.
Every November 10 at the Great Lakes Shipwreck Museum on Whitefish Point, they ring the original bell 29 times.
One ring for every man aboard, and then a 30th time for every sailor the Great Lakes have taken.
Only two people are known to have actually touched her.
On the 1st of September 1995, Terrence Tyel and a dive partner went down on scuba breathing a mix of helium, nitrogen, and oxygen to roughly 530 ft in water around 34° F.
They spent about 15 minutes going down and looking around and about 4 hours coming back up to decompress.
Tyel gripped a rail near the bridge and called it a very special moment.
They didn’t hook onto the wreck out of respect for the dead and he says they weren’t there to look for bodies.
He later deleted his footage.
Now, you may have seen videos claiming a high-tech drone dove to the Fitzgerald recently and finally cracked the case.
I went looking for that expedition.
At the 50th anniversary last November, the last approved expedition to the wreck was still the 1,995 bell recovery, and the Canadian government hadn’t approved another one since.
Larry Elliot, who worked on that bell recovery, doesn’t think anyone will go back down in his lifetime.
Mixster says he’d go tomorrow.
The clues are all down there, he says.
So, for now, the Fitzgerald is still lying where that first robot found her.
The bow upright in the mud and the stern on its back with that unscratched hull facing the surface and two hatch covers from the bow still sitting beside it.
Any full explanation has to account for those covers, and so far none of them has done it to everyone’s satisfaction.
The last thing anyone heard from the Fitzgerald was that she was holding her own.
Almost 51 years later, the wreck is still holding on to the answer.
I’m Ryan Laauo and if you want to see what the cameras really found behind the shipwreck stories everybody thinks they know, there’s more of that on the way here.
