Ancient Mysteries - 4+ HOURS of Deep Sea Mysteries We Still Can’t Explain

Episode Date: October 2, 2026

What mysteries are still hiding beneath the surface of our oceans?For thousands of years, rising seas, earthquakes, storms and geological change have swallowed entire landscapes, ancient settlements and traces of civilizations that once existed above the water. Today, some of their remains can still be found on the seafloor — while other discoveries continue to raise questions.In this 4+ hour documentary, we explore some of the most fascinating mysteries ever discovered beneath the sea: submerged cities, forgotten coastlines, unusual underwater formations, ancient shipwrecks, mysterious artifacts and places that disappeared beneath the waves.We travel from Doggerland, the prehistoric landscape that once connected Britain to continental Europe, to Pavlopetri, one of the world’s best-known submerged ancient settlements. We investigate Yonaguni, the controversial underwater formation off Japan, and the famous Bimini Road, asking where natural geology ends and possible human activity begins.The journey also takes us to the remains of Heracleion, an ancient Egyptian port city rediscovered beneath the Mediterranean, and to Rungholt, the legendary medieval settlement whose story became intertwined with the devastating storms of the North Sea.And then there are discoveries that changed what we thought we knew about ancient technology.The Antikythera Mechanism, recovered from an ancient shipwreck, revealed an astonishing system of precisely engineered gears capable of modeling astronomical cycles — evidence of technological knowledge that survived only because fragments of the device were preserved beneath the sea.But not every underwater mystery requires a lost civilization or extraordinary explanation.Throughout the documentary, we separate archaeological evidence, geological explanations and historical records from speculation, asking what these discoveries actually tell us — and what remains uncertain.How many chapters of human history disappeared when the oceans rose?And if entire settlements can vanish beneath the waves for centuries or even millennia…what else could still be waiting on the ocean floor?Watch to the end for more than four hours of submerged ruins, lost landscapes, ancient shipwrecks, archaeological discoveries and deep-sea mysteries.💬 Which underwater mystery do you think deserves further investigation?🔔 Subscribe for more documentaries exploring ancient civilizations, archaeology, lost history, underwater discoveries and the unexplained mysteries of our world.

Visit BetMGM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and Game Sense remind you to play responsibly. 19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact connects Ontario at 1-866-531-2,600 to speak to an advisor,

free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. This family is on the brink of Civil War. The Mobland, the hit original series is now streaming on Paramount Plus. We are the Harrogans. Don't know than yet? And Google us. From the underworld of Guy Ritchie. Do you want to step up the ladder?

I won't comment dead. Starring Tom Hardy, Pierce Brosnan, and Helen Mirren. Do I have to do everything myself? You want to want! I'll give you a... Moblad, new season is now streaming on Paramount Plus. Hey there, Deep Sea Detectives. Let me show you three things the ocean handed back to us.

a lump of corroded bronze 2,000 years old, packed with gears nobody could match for the next 15 centuries. A cup of coffee, still warm, on the deck of a yacht with the sails up and nobody on board. And a perfect rectangle, 60 metres wide, lying in the dark at 87 metres down in the Baltic Sea. Three objects. Three very different questions, because the ocean stays quiet in three separate ways. It preserves what we thought was lost. It takes ships and crews without leaving so much as a floating boot. and it fakes, handing us rocks that look like staircases and iceberg noises that sound like monsters. So here's my rule for today. For every story, I'll tell you whether there's a real expedition and a published paper behind it,

or whether it's fan fiction with a very good publicist, one number before we go under. Over 80% of the seafloor has never been mapped in high resolution. We have better images of Mars than of the ground beneath our own shipping lanes. So most of the answers are not missing. they're just still down there in the dark. Smash that like button and drop a comment telling me what city you're watching from. Let's go down, and we are not starting with a shipwreck. We are starting with a country.

Because before the ocean took its first ship, it took an entire inhabited landscape, and it did it slowly enough that the people living there had time to notice, complain, pack up and move, generation after generation, for thousands of years. Today that landscape is called Doggerland, and it sits under the North Sea. in the gap between the East Coast of Britain and the Coast of Denmark and the Netherlands.

Right now that gap is a few hundred kilometres of grey, choppy, thoroughly miserable water full of container ships and wind turbines. Twelve thousand years ago it was solid ground. Not a swamp, not a sandbar, not a technicality. Ground. With rivers on it, with hills, with forests, marshes, deer, wild horses, and people who lived there, cooked there, buried their dead there and had absolutely no idea they were standing on the future floor of a shipping lane.

Here is the physics behind that, and it is refreshingly simple compared to everything else in this video. During the last glacial maximum, roughly 20,000 years ago, a genuinely obscene amount of the planet's water was locked up in ice sheets, kilometres thick sitting on top of Scandinavia and Canada. Water that goes into an ice sheet has to come out of somewhere, and it came out of the ocean. Global sea level was around 120 metres lower than it is today. 120 metres is not a subtle adjustment. That is a 40-story building of missing water.

Every shallow shelf sea on earth simply switched off and became land, and the North Sea, which is embarrassingly shallow for something with such a dramatic reputation, was one of the biggest winners. Britain was not an island. It was a soggy north-western corner of the European mainland, and you could have walked from the future site of Norwich to the future-site. of Amsterdam, without getting your feet wet, assuming you had the free time and no strong feelings about mud. Now, the fun part. Everybody imagines that a land bridge is a bridge, a narrow strip,

a prehistoric causeway with a nice sign saying you're now leaving Europe. Thanks for visiting. That is completely wrong, and the truth is much better. Doggerland was not a bridge between two places. Doggerland was the place. At its maximum it covered something on the order of tens of thousands of square kilometres of open plain, and it was not the leftover scrap of the continent, it was arguably the best real estate on it. Because what did it have? It had the outflow of enormous rivers, including systems that are now the Thames and the Rhine, meeting up and wandering across a nearly flat plain, which is exactly the recipe for the wetlands, estuaries and shallow lakes that a hunter-gatherer economy dreams about. If you live by hunting, fishing and gathering,

you do not want a mountain range. You want a rich, flat, boring, waterlog, floodplain absolutely stuffed with fish, waterfowl, reeds, shellfish and animals coming down to drink. Doggerland was that on an industrial scale. So while we picture our ancestors clinging to the edge of the known world, they were more likely sitting in the middle of the good stuff, and modern Britain and Denmark were the awkward hilly outskirts, the suburbs. We are living in what used to be the boring bit. So how do we know any of this, given that the entire archive is currently at the bottom of the sea and under several metres of sediment. This is where the story gets genuinely funny, because the people who discovered a lost country were not archaeologists. They were fishermen,

they were not looking for it, and they were mostly annoyed about it. For well over a century, trawlers dragging heavy nets across the bottom of the North Sea have been hauling up things that have no business being in a fishing net. Mammoth teeth, mammoth tusks, bones of woolly rhino, giant deer, aurochs, bison, horse and reindeer. Chunks of what the fishermen called moorlog, which is a lovely blunt word for a lump of ancient peat, meaning compressed dead swamp, meaning the ground of a landscape that no longer exists

coming up in a net alongside the soul and the place. Imagine going to work, catching your fish, and pulling up a two-meter tusk from an animal that has been extinct for 10,000 years, and imagine the standard reaction on deck for decades being irritation, because that thing is heavy, it is ruining the catch and it is tearing the net. There is a real argument that the North Sea fishing fleet was, accidentally, the largest and worst-funded archaeological survey in European history.

They were excavating a lost world by accident, and throwing most of it back overboard because it was not worth money at market. There were scientists paying attention to be fair. A geologist named Clement Reed had already published work on submerged forest before the First World War, tramping around British coasts looking at tree stumps that stick out of the sand at low tide, standing exactly where they grew, roots down in soil that is now beach. Tree stumps do not swim. If a rooted stump is under the tide line, the land went down or the sea came up,

and either way the coastline moved. So the concept of drowned landscape was already sitting there in the literature, respectable but abstract, the sort of thing that gets a chapter and a polite nod. Then came 1931 and a trawler out of Loewstoft called the Kalinda, working the shallows out near the Lehman and Oa Banks. The crew hauled up a heavy lump of that moorlog peat, and rather than dumping it, someone broke it open. Inside was a barbed point carved out of Antler,

about 20 centimetres long, with a neat row of barbs cut down one side. That is not a rock that looks interesting. That is not a maybe. That is a tool made by a person with a specific job, which is spearing fish or eels and not letting them wriggle off. It had been sealed inside a block of ancient peat, which means it had been lying in a freshwater swamp when that swamp was still a swamp,

and it had gone into the water long before there was any water there in the modern sense. Later dating put the thing somewhere in the region of 11,000 years old. That single object changed the entire question. Up to that point, drowned North Sealand was a geology story, interesting mainly to people who enjoy sediment. The moment a human tool comes out of the same peat as the mammoth bone, it becomes an archaeology story, and archaeology stories have people in them. Somebody made that point. Somebody sat down, chose the antler, worked the barbs with a flint blade, probably swore when one snapped off, and then lost the finished thing in the reeds, which is the single most relatable human experience in this entire video. Every one of us has dropped something in water and watched it go.

That man just happened to drop his inner place that would become the North Sea, which is a hell of a way to become archaeologically significant. It took a surprisingly long time for that world to get name. The term Doggerland was coined in 1998 by the archaeologist Brian E. Coles, who took it from the doggar bank, the big shallow rise that is now the most famous shoal in the North Sea. And the name of the bank itself comes from an old Dutch word for a type of fishing boat, a doger. So we have a lost prehistoric civilization named, indirectly, after medieval Dutch

fishing vessels. No lost golden city naming ceremony, no dramatic Greek etymology, somebody's boat. Honestly, it fits. Everything about the discovery of this place is aggressively unglomerous. Naming it was one thing, mapping it was the problem, because you cannot survey a landscape that is under 50 metres of water, and then another layer of mud, at least not with a trowel and a attitude. This is where somebody had one of the great lateral thinking moves in modern archaeology. The oil and gas industry had spent decades shooting three-dimensional seismic surveys across the North Sea, because that is what you do when you're looking for hydrocarbons. You send sound down, you record what bounces back, you build a three-dimensional model of the layers under the seabed,

and you look for the shapes that mean oil. And in doing so, purely as a side effect, they had recorded in extremely high resolution the shape of every ancient buried surface down there, including, sitting near the top of the stack completely ignored the actual ground surface of doggarland. The team led by Vince Gaffney went and got that data and asked it a question nobody in the petroleum sector had bothered to ask, which is what did this place look like when you could stand on it, and out came a map, not a vague blob, a map with river channels meandering across it, with the oxbow bends you get from a sloth, you get from a slug, river on flat ground, with lake basins, with coastlines, with salt marsh, with valleys and low ridges,

tens of thousands of square kilometres of buried topography, reconstructed from data collected by people hunting for oil. There is a major river system down there that got named after a British geologist Frederick Schotton, because it had never had a name before, having last been seen by people who did not leave us their vocabulary. There is a large ancient lake basin out in the area now called the outer silver pit. You can look at the reconstruction and pick out where you would have put a camp. Right there on that bend, above the floodline near the fish trap, downwind of the smoke, it is a landscape you can read, and then it got even more granular because the next generation of work went after the sediment itself. Push a core tube down into the old land surface,

bring up a column of mud, and inside that mud is pollen, plant fragments, insect parts, and, in the best cases, environmental DNA, the genetic traces shed by everything that lived and died in that spot. You can pull that apart and get a species list for a place that has been underwater for 8,000 years, oak, hazel, alder, grasses, the plants of an open woodland and a wetland edge. This is a bit like finding out you can reconstruct a whole neighbourhood from the contents of one very old vacuum cleaner bag, and it works, and it is one of the most quietly astonishing things happening in archaeology, right now. Meanwhile, the human evidence kept turning up, mostly through a channel nobody planned. The Netherlands, being the Netherlands, has an ongoing national relationship with sand. Large volumes

get dredged from the seabed and pumped onto beaches to keep the coast where the Dutch wanted, which is a very Dutch solution to a very Dutch problem, and that dredged sand comes from exactly the drowned landscape we have been talking about, which means that for years amateur fossil hunters have been walking Dutch beaches finding bones and teeth of ice age animals, lying in the fresh sand, essentially a lost continent being delivered to the shoreline by pipe. In one batch of that dredged material came a fragment of a human frontal bone from above the eye. It was studied and identified as Neanderthal, the first Neanderthal fossil found in the North Sea. He got the nickname Creene, and he is far older than the Mesolithic fisherman with the Antler Point, tens of thousands

of years older, which tells you that this drowned plane was not occupied once. It was occupied over and over by more than one kind of human, whenever the ice let go and the land opened up. It was not a fluke, it was a habitat, and it was busy. So we have a place, we have a map, and we have residence. Now the eviction. The end of Doggland was not one event, and this is the part that usually gets flattened in the retelling, because a slow flood is a terrible story and a tsunami me is a great one. The truth is mostly the slow flood, and the slow flood is more unsettling. As the ice sheets melted, sea level came up steadily over thousands of years, not a wall of water, not a punishment, just the tide line arriving a bit further inland than your grandfather remembered,

decade after decade, century after century. In a landscape as flat as doggarland, a small vertical rise eats a huge horizontal distance. Raise the water a meter on a plane with almost no slope and you can lose kilometres of ground. So the experience for the people living there was not disaster. It was a background fact. The good marsh gets salty. The eel run moves. The place your family used to camp is now a lagoon, then a bay, then just water. The river's back up. The fresh water goes brackish. And you shift the camp a little further inland each generation, and you do not think of it as the end of the world. You think of it as this year, because nobody experiences a 2000-year process as an event. That is the genuinely eerie thing. Nobody there ever knew

their country was ending. They just knew the fishing had changed. Piece by piece, the plane shrank. The high ground held out longest, and for a while the doggar bank itself was an island. A last big chunk of the old world standing alone in the middle of the new sea, which is either a nice thought or an extremely bleak one depending on how your week is going. And then, around 8,200 years ago, something did happen quickly. Off the coast of Norway, an enormous section of the continental shelf let go. This is the Sterega slide and the numbers on it are ridiculous. We are talking about thousands of cubic kilometres of sediment breaking loose and sliding downslope in a submarine landslide, a scar hundreds of kilometres long. When you displace that much material under water that fast,

the water has to go somewhere and it goes outward as a tsunami. Geologists have tracked the deposit it left behind, a distinctive layer of marine sand shoved inland and dropped where marine sand has no business being, and they find it in eastern Scotland, in the Shetlands, along the Norwegian coast. In some places the run-up estimates are genuinely alarming, especially in the north where the wave arrived with the least warning and the most energy. Now aim that at a low, flat, water-logged remnant plain with essentially no high ground. Doggerland had no hills to run to, it had already been cut down to marsh, shoals, tidal flats and the last islands. A wave like that would have gone straight across it and kept going. But here is the correction that recent work has made, and it matters

because it is a nice example of science declining to take the dramatic option. For a while, the popular version was that Sterega was the killing blow, the moment Doggerland ended, curtains, water, done. Then researchers went and pulled sediment cores from the southern North sea from the drowned landscape itself, looking for the tsunami layer in place. They found evidence of it, and they also found evidence of what came after, which is that in places the landscape recovered. The wave came through, dumped its sand, wrecked whatever it wrecked, and then vegetation grew back on top. Which means Stereger did not delete Doggerland. It hit an already shrinking, already dying, already half-drowned country, and made a terrible day out of it, and then the

slow, patient, undramatic rise of the sea, finish the job over the following centuries. The landslide gets top billing because it is loud. The actual cause of death is a very slow, very quiet number going up, which is a lesson that will keep showing up for the rest of this video, and possibly outside it. It is worth sitting with the human cost for a second, without getting mordland about it, because this is where the scale lands. That coastline was some of the richest ground in the region, so the areas hit hardest with the areas most, likely to be occupied. Not thousands of people at once, this was a thinly populated world, but real people, in real camps, at the water's edge, who had built their entire economy on exactly

the environment that was about to be reorganised by physics. And the survivors did what survivors do, which is walk, west into what became Britain, east into what became the continent. Same people, same tools, same traditions, split by a rising sea into two populations that would spend the next 8,000 years developing slightly different accents and then arguing about fishing rights in the exact same water that separated them. History has a sense of humour and it is not a subtle one. So that is the coordinate system for everything that follows and it is why we started here instead of with a nice photogenic shipwreck. Because the standard instinct, when someone shows you a picture of a wall or a street or a foundation underwater, is to say that is impossible. The sea has

always been there, human beings do not build on the seafloor. And the Doggland file demolishes that instant completely. The sea has not always been there. On a geological time scale, it is a recent and pushy arrival. There are entire drowned landscapes on every continental shelf on this planet, and the shelves are exactly where people always preferred to live, because coasts are where the food is. If you want to find early coastal settlement, you have a problem, which is that the coast has moved and taken everything with it. So a drowned settlement is not extraordinary. It is expected. It is close to inevitable. Which brings us straight to the trap and to the second file, because once you accept that the sea can swallow inhabited land, you become extremely easy to sell

things to. If a country can go under, why not a city? If a city, why not a temple? If a temple, why not a whole lost civilization with technology we cannot explain? The line between reasonable and ridiculous stops being obvious and you need a method instead of a feeling. So let us build one. And the way to build it is not to argue about the disputed sites first. It is to go find a site nobody disputes, look very carefully at what it actually looks like on the bottom and use it as a control sample, get the reference specimen, then hold everything else up against it. Our control sample is in southern Greece, off the coast of Laconia, in the shallow water between the mainland and the little island of Elephonisos.

It is called Pavlopetri, and it has one enormous advantage over everything else we're going to look at today, which is that it is in three to four metres of water. That is snorkel depth. That is stand-up and wave depth. You do not need a submersible, a support ship or a lawyer.

You need a mask, and on a calm day with the sun overhead, you can float on the surface and look down at a bronze-aged town, and it is a town. Not a formation, not an anomaly, not something that looks a bit like a town if you squint and are already convinced. Pavlapetri has streets, actual streets, running in more or less straight lines, and crucially they intersect. They meet each other at right angles and continue on the other side forming blocks.

Along those streets there are building foundations, and the buildings have internal divisions, meaning rooms, meaning walls that meet other walls in a way that encloses space and creates doorways. There are courtyards. There are what appear to be gardens or yards. attached to specific houses, and around the edges there are graves, both chamber tombs cut for multiple burials and stone-lined cyst graves for individuals. The site was noticed as early as the beginning of the 20th century by a Greek geologist, then properly rediscovered in 1967 by Nicholas Fleming, and mapped the following year by a Cambridge team who did it the hard way, with tape measures and drawing boards and an enormous amount of patients underwater.

Decades later, a new project went back in with Modern Kit, led from the University of Nottingham working with the Greek Underwater Antiquities Authorities, and produced a full digital three-dimensional survey of the whole place, using high-resolution sonar and robotic mapping systems developed for entirely different purposes. That survey pushed the occupation back further than anyone expected. This is not just a Mycenaan Bronze Age town, although it was that too. There is material going back to around 5,000 years ago, which makes it one of the oldest submerged planned settlements we have found anywhere. Now, the important bit, the bit that makes it a control sample. Forget the layout for a moment. What makes Pavlopetri unarguable is the junk, because the site is covered in the

debris of daily life, broken pottery and quantity. Not one beautiful vase, thousands of fragments of ordinary storage jars and cooking pots, the prehistoric equivalent of a landfill full of takeaway containers, loom weights, the little perforated clay lumps that hang on the bottom of a vertical loom, which tell you somebody was making cloth here, at home, routinely. Obsidian blades, which are extra interesting because obsidian does not occur locally, so it came in by boat from the islands, which means trade, which means this settlement was plugged into a network, large storage jars for oil or grain, grinding stones, and the graves, which are the single most human piece of evidence available,

because burying your dead in a specific formal way, in a specific place, is something no natural process has ever once done by accident. That is the checklist. Write it down, because we are going to use it twice in the next ten minutes. A real human site has geometry. Yes, but geometry is the least of it. A real human site has garbage. It has wear, meaning worn thresholds, worn steps, tool marks, repairs.

It has imported material that could not have arrived by itself. It has standardisation, the same brick size, the same wall thickness, the same room proportions repeating because somebody was following a plan. It has the boring infrastructure of survival, drains, hearths, storage, and it has burials. People generate mess at an incredible rate. Even a small settlement occupied for a few centuries leaves behind an absolute mountain of broken, discarded, half-finished, lost and thrown-away material. And water does not clean it up. Water preserves it beautifully. Oh, and one more thing about Pavlopitri,

which is the least mysterious part of the whole story. It did not sink because of a curse, a war, an angry seagod, or a lost technology. That stretch of coast sits in a seismically busy part of the Mediterranean. The ground there has been dropping over the long term, and a few thousand years of subsidence plus earthquakes will comfortably put a coastal town in four metres of water. The mechanism is dull, the mechanism is always dull, remember that. control sample established. Now let us go to the Pacific and see how the method holds up

because our second exhibit is the most photogenic underwater structure on the internet and it is genuinely honestly strange-looking. Off the southern coast of Yonaguni, the westernmost inhabited island of Japan, closer to Taiwan than to most of Japan, there is a massive rock in shallow water that does not look like it should be there. It was found in 1986 by a local dive operator named Kihachiro Aratake

who was not looking for archaeology at all. He was scouting the area because Yonaguni is one of the great places on earth to dive with hammerhead sharks in winter, and he was working out where to take customers. Instead of sharks, he found a wall of stone with what appeared to be enormous flat terraces, sharp corners and a series of steps, and it is big. The main formation runs roughly 200 metres or so along,

and rises a good 25 metres, and the top surfaces are startlingly flat. There are vertical faces that meet horizontal faces at clean 90-degree angles. There is a feature that gets described as a staircase, another described as a road running along one side, another as a terrace or a plaza. In the photographs it looks like the corner of a ruined temple complex that somebody dropped into the ocean.

If you have seen one image of a lost underwater city in your life, statistically there is a good chance it was this. The case for it being artificial has been argued for decades, Most prominently by Masaki Kimura, a marine geologist at the University of the Rukius, who has dived the site many, many times. His interpretation has grown over the years into a full site plan, complete with identified features. He argues for quarried terraces, for a processional road, for drainage channels, for post holes, and for carved figures, including a formation he reads as a face and another he reads as a turtle. He places its construction well before the end of the last ice age on the reasonable grounds that the site is now in water, and, as we established at the top of this video, sea level was much lower back then.

If it is artificial, it has to be very old, because the last time you could have walked around on it, mammoths were still a going concern. The case against comes most prominently from Robert Schoch, a geologist from Boston University, who dived the site and came back with a much less exciting answer, which is that it is bedrock. Specifically, it is a fine-grained sandstone and mudstone formation, the same rock that makes up the local geology and that rock does something very particular. It is deposited in horizontal layers, called bedding planes, and it fractures along vertical joints. Layered rock with a strong joint set does not crumble into random rubble. It breaks into rectangular blocks with flat tops and vertical sides, because those are the

two directions of weakness. Tectonics to tilt everything slightly, add a few thousand years of extremely energetic surf, throw in the occasional typhoon and the occasional earthquake, and you get flat terraces separated by clean vertical steps, which is exactly the objection you make when someone says nature does not make right angles. Nature makes right angles constantly and it does not even have to try. Go look at columnar basalt sometime, at the giant's causeway or any of a hundred other places, where cooling lava splits into neat hexagonal pillars so regular that people spent centuries insisting a giant must have stacked them. Nature is an extremely competent geometre. It just does not

have a permit office. The strongest single argument here is not about any one feature. It is about the neighborhood. The same rock formation with the same bedding and the same joint patterns is exposed above water on the nearby coastline, where you can walk over to it in shoes. And it does the same thing, flat surfaces, sharp steps, rectangular blocks, right angles, all above the water line where no lost civilization has been proposed, because when a structure like that is sitting in a field, nobody feels the urge to build a mythology around it. The mystery is significantly enhanced by being underwater, poorly lit, hard to reach, and only viewable in short bursts by people with a limited air supply and a strong prior belief. But we are not going to settle this on geology

alone, because I am not a geologist and neither statistically are you. We are going to run the checklist. Pavlopetri gave us the list. Apply it. Garbage. Where is it? A settlement or a monument complex of the scale being proposed would have been in use for a long time by a lot of people, and those people would have left behind broken pottery, food waste, charcoal, discarded tools, failed blanks and general debris. At Yonaguni, that assemblage has not been found. There is no midden. Imports. Nothing has come out of the site that had to be carried in from somewhere else. No obsidian equivalent, no exotic stone, no material that fails to match the local bedrock, which is a problem because everything there matches the local bedrock a bit too well.

Standardisation. Human building repeats itself. Same block dimensions, same tread height on steps, same wall thickness, because a crew building something works to a measure. At Yonaguni, the so-called steps are famously irregular, with treads and rises of wildly varying height, which is fine for a rock and terrible for a staircase. Try walking up steps where each one is a different height, and you will discover very quickly that this is not how humans build things they intend to walk on twice, wear and tool marks, cutting stone leaves evidence, chisel scars, wedge slots, drill holes, the little repetitive marks of somebody working a face, quarrying leaves waste, spalls, half-detached blocks, abandoned work. On the disputed features, these have not been convincingly

documented, and the marks that have been proposed are also consistent with erosion and marine organisms. Burials. None. So the score on the control checklist is close to zero, and that is the actual argument. Not that the formation looks natural, because to a lot of eyes it does not. The argument is that a human construction site is not just a shape. It is an entire ecosystem of leftovers, and none of the leftovers are there. If a civilisation built that thing, they built it, used it, and then took every single object they had ever owned home with them and swept up afterward, which would make them the tidiest people in the history of the species and a completely unprecedented archaeological phenomenon in their own right. Could that change tomorrow? Yes,

absolutely. That is the honest position. Find one hearth with charcoal in a sealed context, one worked bone, one pot shirt of a known type, one securely dated toolmark, and the whole file reopens instantly. Nobody serious is claiming it is impossible, only that the evidence is currently missing. And missing evidence is not a conspiracy, it is a to-do list. Now the third exhibit, and this one is my favourite, because the timing of it is so absurd that if a screenwriter pitched it, they would be told to tone it down. In the Bahamas, off the coast of North Bimony, in about five or six metres of clear turquoise water, there is a line of large stone blocks running along the seabed for the better part of a kilometre, curving at one end into a hook shape. The blocks are roughly

rectangular, some of them three or four metres across, and they are laid out in a way that reads to any human eye as a paved road, a causeway or the top of a harbour wall. It was reported in 1968 by Joseph Manson Valentine, who was diving with colleagues in the area, and it went into the world as the Bimini Road. Here is where the comedy. comes in. The psychic Edgar Casey, decades earlier, had made a very specific prediction. He said that a portion of Atlantis would rise again, and that it would happen near Bimony, and he named the years 1968 and 1969. So a formation gets reported off Bimony, in 1968, that looks like ancient paving. You could not construct a better trap for human pattern-seeking if you spent a year designing one.

The prophecy did not just predict the discovery. It created the audience, the expectation and the interpretation before anybody got in the water. And to be completely fair to everyone involved, if you had been told for years that something was going to appear right there, right then, and then you swam over a line of enormous flat blocks lying end to end in a curve, you would have surfaced yelling too. So we run the test. And in this case, the geologist had a beautifully direct method available, which I love because it does not require anyone to argue about vibes. the blocks are made of what is called beach rock. Beachrock forms when sand and shell fragments on a beach

gets cemented together by carbonate minerals precipitating out of the groundwater, and it happens along the shoreline in a slab that follows the beach. Then, as conditions change, that slab cracks. It cracks into blocks, along fairly regular joints, and because it formed on a gentle slope, the blocks tilt slightly seawward, and the whole thing looks alarmingly like paving. Now, how do you tell a cracked natural slab from blocks that were quarried and laid by people? You drill it.

Take cores through multiple blocks and compare the internal structure. If people cut these blocks somewhere and haul them here and set them down, the internal layers inside each block will be oriented randomly because nobody putting down paving stones cares which way the grain runs. But if it is a single slab that cracked in place, then the internal bedding will line up continuously from block to block to block, all dipping the same way at the same angle, because it is all one rock that simply broke. Eugene Shin and colleagues from the United States Geological Survey did precisely that.

They cored the blocks, and the internal layering matched across them, consistently, dipping the same direction, exactly as it would in a continuous beach rock slab that fractured in place. Dating put the rock in the range of a few thousand years old, not the 10,000 plus required by the Atlantis version. same story as Yonaguni, different rock type, same verdict, and this time with a physical test that anyone can understand. There have been counter-arguments over the years, including claims of toolmarks, of a second layer of blocks underneath, and of prop stones wedged beneath the edges. The most entertaining alternative theory is not Atlantis at all, but ballast, the idea that the blocks are stone dumped from ships, which is a genuinely reasonable thing to check because harbors do accumulate weird stone. That has its own problems and has not held up well either. But notice what none of these arguments have produced.

Run the checklist. No dwellings alongside the road. No pottery scatter. No refuse. No burials. No wear surface, which is a big one. Because a road that people actually used for centuries would be polished, rutted and repaired. A road going nowhere, connecting nothing, next to nothing.

With nothing on it is not really a road. It is a line. and that is the tool I want you to walk away with, because we are going to need it repeatedly for the rest of this. Call it the right angle test, and the punchline of the right angle test is that the right angle is not the test. Straight lines and clean corners are cheap. Rock produces them by fracturing,

ice produces them by scouring, currents produce them by carving. Your brain, meanwhile, is a machine built specifically to find edges, faces and patterns in bad lighting, because for most of our history, the alternative to spotting a shape in the murk was being eaten by it. Point that machine at a low-resolution image of an unfamiliar rock formation and it will hand you a staircase, a wall and a face, free of charge, with total confidence.

What is not cheap is everything else. Debris is not cheap. Standardisation is not cheap. Imported material is not cheap, and the dead are not cheap, because burial is the one behaviour that geology has never once faked. So the difference between Pavlopetri and the other two is not that one is prettier or more convincing in a photograph. Honestly, in a photograph, Yonaguni wins. Pavlopetri looks like a bunch of rubble outlines in shallow sand, and if you swam over it without a guide, you might not notice you are floating above a 5,000-year-old town.

Yonaguni looks like the set of a film about a lost empire. That is precisely the point, and it is the reason this chapter exists before all the others. Looking like something is not evidence. It is a hypothesis that has not been checked yet. And it works both ways, which is the part people miss when they think scepticism means saying no. The same method that knock down two famous sites is the method that finds the real ones, because there are drowned cities out there that are completely genuine,

that were dismissed as legends for centuries, that got laughed at in exactly the tone we have just been using, and that turned out to be sitting there the whole time waiting for somebody to bring the right instrument. Two of them are coming up next, and in both cases the people who found them were not holding a shovel. They were holding a magnetometer, which is a much less romantic image but a significantly better success rate. Let us start in Egypt, in Abu-Kir Bay, a stretch of shallow, murky green water about 30 kilometres east of Alexandria. It is not a beautiful dive site.

Visibility is often measured in centimetres, rather than metres. The bottom is soft grey sediment, and for most of the 20th century the main reason anyone paid attention to that bay was a naval battle fought there in 1798, when the British Navy caught the French fleet at anchor and comprehensively ruined Napoleon's afternoon. Nobody thought there was a city under it. There were only rumours written down by dead Greeks, because the city was in the sources.

That is the maddening part. It had been sitting in the literature the entire time described by people who had actually been there. Herodotus, writing in the 5th century before the common era, mentions a place at the westernmost mouth of the Nile, with a great temple and a customs post, and tells a version of the Trojan war story in which Helen and Paris get blown off course

and end up stuck there. Later writers mention it too. It appears in Greek as Heraclean, and separately in Egyptian records there is a port called Thonis, and for a long time scholars assumed those were two different towns in roughly the same area, because why would one city have two completely unrelated names? That would be silly.

The problem was that none of these authors included coordinates, and the Nile Delta is the single worst filing cabinet on earth. It is a giant pile of mud that the river has been rearranging for thousands of years. Channels move, new mouths open, old ones silt up and vanish. Whole branches of the Nile that were major shipping routes in antiquity are now dry farmland or a vague depression in a field, and the ancient coastline is nowhere near the modern coastline. So you have a city that is definitely real in the texts,

in a landscape that has completely reshuffled itself, in a bay where you cannot see your own hands. Good luck. Which is exactly why the breakthrough came from a piece of physics gear rather than from anyone with a trowel. The French underwater archaeologist Frank Godio and his institute spent years working Abu Kier Bay with an approach that is basically medical imaging for the seabed. The key instrument was a highly sensitive magnetometer,

developed in collaboration with French nuclear physicists, which measures tiny variations in the local magnetic field. Here is why that matters. You do not need to see an object to detect it. You need it to disturb the magnetic field. Stone, brick, fired pottery and especially metal all sit in the sediment quietly warping the field around them. And if you tow a sensitive enough sensor back and forth across the water in a tight grid,

you can build a map of what is buried under the mud without ever getting in the water, so they mowed the lawn. Kilometer after kilometre, line after line for years, in a bay that looks like nothing, building a magnetic map of an area of many square kilometres, add side-scan sonar to see the shape of the bottom, add a sub-bottom profiler to look at the layers underneath it, and slowly a pattern comes out of the noise, not one blob, a structured arrangement of anomalies, straight edges, concentrations,

channels, a layout, and then, in the year 2000 they went down and confirmed it, and the bay coughed up a city. The single most important object came up early, and it is the sort of find archaeologists dream about because it does not require interpretation. It is a stealer of black granadaerite, close to two metres tall, and it is in almost perfect condition, sharp and crisp as if it were carved last year rather than in the fourth century before the common era.

It was put up by Necdinebo I, one of the last native Egyptian pharaohs, and its text is not poetry or myth. It is a tax decree. it orders that a 10% levy on certain goods arriving from the Greek world be paid to the temple of the goddess Neith at Saes. It is, in essence, a large stone invoice, which is a wonderfully deflating thing to find at the bottom of the sea when you were hoping for a prophecy.

But the reason it settles everything is at the bottom, where the text specifies where the stone is to be erected, and it names the place, the city known in Egyptian as Thonis, which the Greeks called Heraclean, two names, one town, The entire scholarly debate is closed by a line of hieroglyphs on a tax notice, discovered by people who found it with a magnetic field detector.

The Steeler even has a near twin, an almost identical decree found on land a century earlier at another delta site, which means we can compare the two and confirm this was standard administrative practice and not a one-off. Bureaucracy, once again, quietly outlasting everything built to impress the gods. Then came the rest of it, and the rest of it is a one of it is. enormous in the most literal sense. Red granite colossi, including a figure of Happy, the god of the Nile flood, standing well over five metres tall, which was raised out of the water in one piece, and now stands upright in daylight for the first time in more than a thousand years. A colossal pharaoh, a colossal queen. The remains of the great temple of Amun Gerib, which was not just any temple,

but the place where a pharaoh was formally granted the title of Universal Ruler, meaning this was not a provincial trading post. This was a site of state ceremony. Below the temple, ritual deposits. In the Sacred Canal, dozens of small lead models of ceremonial boats, votive offerings tied to the festival of Osiris, which tells us not just what the city did but what it believed and when it parted. And the small stuff is arguably better, because the small stuff describes an economy rather than a monument. Gold, silver and bronze coins, jewelry, weights used by merchants, hundreds of little statuettes. And then the numbers that really tell you what this place was, more than 70 shipwrecks and several hundred stone anchors spread across the bay,

accumulated over centuries. That is not a disaster. That is a working harbour doing what harbors do, which is losing ships steadily as a cost of business. Heraclean was the mandatory customs gateway for anything Greek entering Egypt by the Western Nile, the checkpoint where everything got taxed before it went up river. It was a port town whose whole personality was paperwork and cargo handling and it got rich doing it. One of those wrecks deserves its own moment, because it closes a very old argument in the funniest way possible. Herodotus described a distinctive type of Nile cargo boat, and his description was strange. He said it was built from short planks of acacia, assembled almost like bricks, arranged in a pattern and pinned together with an internal

structure unlike normal Mediterranean shipbuilding. For a long time, this was filed under Herodotus being Herodotus, meaning charming, curious, sometimes brilliant and occasionally repeating something a guy told him in a bar. Nobody had ever found a boat built that way. Then the team excavated one of the wrecks in the bay, and it was exactly the vessel he described, built exactly the way he said, right down to the planking method. The man was taking accurate technical notes on shipbuilding two and a half thousand years ago, and we spent most of that time patronising him about it. Somewhere out there his ghost is holding up a receipt, so what killed the place? And here you can pick your favourite dramatic option, because the popular imagination has plenty.

Divine wrath, a single catastrophic wave, a curse for the usual assortment of sins. The actual answer is a geotechnical process with a name so boring it's almost aggressive, liquefaction. Here is how it works, and it is genuinely working. understanding because it explains a lot of drowned coastal sites. The Nile Delta is not rock. It is a thick pile of water-saturated clay and sand that the river dumped there. Sediment like that gets its strength from grain-to-grain contact, with the water sitting in the spaces between the grains. Now shake it, or load it, or both. The pressure of the water trapped between the grains rises, the grains lose contact with each other, and for a moment the whole mass stops behaving like a

solid and behaves like a thick liquid. Anything heavy sitting on top does the obvious thing, it sinks into it. Then the water pressure drops, the grains resettle, and the ground goes back to being ground, now with a temple three metres lower than it was that morning, and Heracleion was doing everything wrong for that particular failure mode. It was a heavy stone city, with colossal statues and massive temple blocks, built on saturated delta clay, on a coast that was already slowly subsiding, region that gets earthquakes while the sea level was creeping up. That is not one catastrophe, it is a subscription service. The archaeological evidence points to repeated events over centuries rather than one bad night, with major collapses in the second century before the common era

and continuing losses afterward, until the last stubborn islands of the city finally went under around the 8th century of the common era. There is also a purely economic murder happening in parallel, and it is my favourite part of this story because it is so unromantic. Alexandria was founded in 331 before the Common Era, just down the coast, with a superb deep harbour and enormous imperial investment behind it. Once Alexandria was up and running, why on earth would you route your cargo through the shallow, silting, sinking customs post next door?

Trade moved. Heraclean's entire reason to exist evaporated. So by the time the ground finished swallowing it, the city had already been economically dead for a long time. It was not struck down at the height of its glory. It went bankrupt first and drowned afterward, which is the least cinematic possible ending, and almost certainly the true one.

Now let us jump 5,000 kilometres north and about 1,500 years forward to a place where the water does something even weirder, which is leave. Visit BetMDM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMGM and GameSense remind you to play responsibly. 19 plus to wager.

Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact connects Ontario at 1-866-531-2600 to speak to an advisor. Free of charge. BetMGEM operates pursuant to an operating agreement with Eye Gaming Ontario. The Wadden Sea runs along the coast of the Netherlands, Germany and Denmark, and it is one of the strangest environments in Europe.

Twice a day the tide goes out and exposes an enormous plain of mud and sand kilometres wide. You can walk on it. People do as a hobby with guides, wearing shorts and an expression of mild concern, and then twice a day the water comes back quickly, and if you have misjudged the timing, the flat, harmless mud you are strolling across becomes a tidal channel with a current in it. In the North Frisian part of that coast, in Germany, there is a legend about a town called Rungholt.

The story, as it has been told for centuries, is a proper morality tale. Rungholt was rich and got arrogant. In the standard version, a group of drunk men decided it would be hilarious to force a priest to administer the last rights to a pig, and when the priest refused and prayed for justice, the punishment arrived on schedule in the form of a storm that wiped the town off the map in one night. Locals will tell you that on quiet days you can still hear the church bells ringing under the water, which is a detail that appears in drowned town legends across Europe. with such reliability that it is practically a genre requirement.

The legend got a serious upgrade in 1882 when the German poet Detlev von Lillienkron published a dramatic ballad about the flood, written in dialect, with the sea itself boasting about what it had done. It is genuinely good, it is memorable, and it did to Rungholt exactly what a hit song does to a real event, which is to permanently glue the drama to the facts. Generations of German schoolchildren learned about Rungholt as poetry,

And once a place is famous as poetry, it is very hard for it to be taken seriously as geography. It becomes the local Atlantis, which is a compliment socially and a disaster academically. The storm at least is unambiguously real, and it is one of the worst natural disasters in the history of Northern Europe. In January 1362, a monstrous storm surge hit the North Sea coast, remembered as the Grota Mandrenke, a name that translates roughly to the great drowning of men and does not require any. embellishment. It tore into the low-lying reclaimed marshland along the Frisian and Danish coasts, breached dikes, drowned tens of thousands of people, and physically redrew the coastline. Entire parishes ceased to exist. Islands were split in half, in places the sea permanently

claimed land that had been farmed for generations, and the modern map of that coast is essentially the after-photo. So the flood is history. The question was always whether Rungholt itself was history or folklore, and the mainstream answer for a long time leaned toward folklore, or at least toward the assumption that the real place was a modest scatter of farms that the legend had inflated into a wealthy metropolis. That scepticism was not unreasonable, documentary evidence is thin, and a settlement of timber, thatch and earth, hit by a surge and then worked over by six and a half centuries of tides, is not going to leave you a nice ruin to photograph. But it was leaving something, and the people who noticed were, once again, not professional archaeologists,

they were mud-flat walkers. For about a century, locals and amateurs have been going out onto the flats at low tide near the island of Sudfall and finding things. Pottery sherds, the remains of wells, which are extremely useful because a well as a whole somebody dug on purpose in a specific spot to reach fresh water, which means somebody lived right there, traces of ditches, the occasional exposed line of an old dike. The evidence kept coming slowly, piece by piece, in the two-hour windows the tide allows, and it kept saying the same thing, which is that there was a real, organised, engineered landscape under the mud. Then, in 2023, the instruments arrived, and the whole picture flipped in a single field season.

A team from several German universities went out onto the flats with the full geophysics kit, and the way they had to work is my favourite detail in this entire chapter. They used magnetic gradiometry, which measures tiny differences in the magnetic field between two sensors and is superb at picking up things like fired brick, hearths and disturbed soil. They used electromagnetic induction, which pushes a field into the ground and reads how the material responds, letting you distinguish a ditch full of soft fill from undisturbed clay. And they had to drag this equipment across a tidal flat, on foot and on sleds, in a race against the incoming sea, mapping about 10 square kilometers of seabed a few hours at a time. Not a research vessel, not a submersible.

Scientists jogging across mud with expensive electronics, checking their watches, which is a much more accurate picture of modern archaeology than anything with a fedora in it. What they mapped was not a village. It was a system. They traced dozens of dwelling mounds, the artificial earth platforms that people in this region built to raise their houses above the floodline, arranged across the landscape in a coherent pattern. They found kilometres of drainage and dike structures, the engineering that made this marsh farmable in the first place. They found evidence of a tidal harbour,

and in the middle of it, the outline that made the news a large building foundation on its own mound, roughly 40 metres by 15. That size is the punchline. A building 40 metres long is not a farmhouse and it is not a barn. In a medieval North Frisian marsh community, a stone-founded structure of that scale is a church, and specifically it is a church big enough to serve a real parish with real money,

which means the legend, ridiculous as it sounds, when it involves an intoxicated pig baptism attempt, was not wrong about the fundamentals. There was a settlement, it was substantial, it was wealthy enough to build big, and it was exactly where the mudflat walkers had been, quietly saying it was for the last hundred years while everyone else was quoting the poem.

Put these two together, and you get the, actual moral, which is not that legends are true because most of the time they are not. It is subtler and more useful than that. Legends are terrible witnesses about causes and surprisingly decent witnesses about locations. Nobody in either story got the mechanism right. Heracleon did not sink because a god was annoyed, it sank because saturated clay does not enjoy being asked to hold up a granite temple. Rungholt was not deleted for blasphemy.

It was deleted because people had built productive farmland on low-reclamed marsh behind dikes that could not survive a once-in-several century's storm surge, which is a risk calculation, not a sin. But both traditions were pointing at a real place. They just described it in the only vocabulary available at the time, which was moral rather than geological. And in both cases, the thing that converted the story back into a site was not new belief, and it was not better diving. It was an instrument that could see through mud, the magnetometer does not care about Herodotus, and the gradiometer has never heard of a poem. They both just report an anomaly at a location, and then somebody goes and checks. The gap between myth and archaeology most of the time

turns out to be a hardware problem, which sets up the next question rather neatly, because so far everything we have found has been in the ground. Now we go after the things that were moving. When a ship goes down, we tend to obsess over the vessel, but the vessel is usually the least interesting thing about the event. A wooden hull is a container. What was a inside it is the actual historical document because cargo tells you who was trading with whom, what was worth moving, who was rich and what they thought was valuable. And there are two shipwrecks that prove this point from opposite directions in the most extreme way possible. In one case, the cargo was found and it rewrote the history of technology. In the other, the cargo has never

been found and it has been generating international arguments for five centuries. The first one starts, as the best archaeology often does, with people who are doing a completely different and much more dangerous job. In the year 1900, a crew of Greek sponge divers from the island of Simi was working its way home from a season in North African waters under Captain Demetrios Contos. Bad weather forced them to shelter near Antikythra, a small rocky island in the channel between Crete and the Peloponnese, with a coastline that is essentially a series of hints that you should be somewhere else. While they were waiting out the wind, they did what sponge divers do, which is go down and look for sponges. Understand what diving meant for these men, because it puts the

rest of the story in perspective. They were working in canvas suits with copper helmets, air pumped down by hand from the boat above, at depths of around 45 metres and deeper. Decompression theory at the time was somewhere between primitive and folklore. These men accepted a level of occupational risk that would give a modern safety officer a spontaneous nosebleed, and they did it for sponges, which is a commodity you now buy in a plastic bag for the price of a coffee. One of the divers, Aleus Stadiartis went down and came back up faster than anybody should. He was by every account badly shaken and described a heap of dead bodies on the bottom. Naked people, horses, piles of them, lying in the sand. The captain, reasonably suspecting that his diver had either found something horrific

or had gotten a bit of a nitrogen buzz went down himself, and he came back up with an arm, a bronze arm from a statue. That is the moment. Not a treasure chest, not a glint of gold. A single, detached bronze limb, brought up by a working man who immediately understood he was looking at something the Greek state would want to know about. And here is the part that deserves genuine respect.

They reported it. They could have quietly stripped the site for years. Instead they took it to Athens, and the result was the first large scale, underwater archaeological operation in history, conducted in 1901 with the sponge divers doing the actual work. Supported by the Greek Navy, under the direction of the Education Ministry at cost them, the divers worked at the limits of what was survivable, hauling marble and bronze up a steep underwater slope, and the physiological price was exactly what you would expect from repeated deep

dives with no real understanding of decompression. One of them died, others were left paralysed. The founding operation of an entire scientific discipline was carried out by sponge divers who were injured doing it, and if you want a single image for how underwater archaeology actually got started, that is it. What came up was spectacular. Bronze statues, including a magnificent youth with one arm raised, and a haunting bearded head usually called the philosopher. Marble figures, many of them grotesquely lopsided, eaten away on one side and pristine on the other, because whichever half stuck out of the protective sediment spent 2,000 years being consumed by marine organisms with no appreciation for classical art. Glassware, jewelry, coins, furniture fittings, fragments of a

bronze lyre. It was, by any measure, one of the richest ancient cargoes ever recovered, and the museum in Athens filled up with it and everyone was thrilled. And in the middle of all that, ignored, sat several lumps of corroded green rubbish. They were unremarkable, encrusted, shapeless, the sort of thing that gets put on a shelf while the important material is dealt with. It took until 1902, more than a year later, before an archaeologist named Valerius Stais looked properly at one of those lumps and noticed something impossible sticking out of the corrosion, a gear wheel, a precision-toothed gear, cut in bronze, sitting inside a 2,000-year-old shipwreck deposit. The reaction, for decades afterward, was mostly denial, and honestly that was a defensible

position. Everything known about ancient technology said this could not exist. Toothed gearing of that fineness in that arrangement belongs to the world of mechanical clocks, and mechanical clocks show up in Europe in the 14th century of the common era. Finding one in a wreck from the first century before the common era is like opening an Egyptian tomb and finding a laptop, except worse, because at least a laptop would be obviously modern and therefore obviously a hoax. This thing was unambiguously ancient and unambiguously advanced, which is the most annoying combination available. Cracking it took most of the 20th century and three generations of imaging technology, and it is a great example of a mystery being solved not by a genius insight, but by better cameras. The first serious assault came from

Derek de Sola Price, a historian of science, who worked with a Greek nuclear physicist to get gamma-ray and x-ray images of the fragments in the early 1970s. Radiography is the key move here, because the object is a solid lump of corrosion product with the mechanism entombed inside it. You cannot open it, take it apart and you destroy it, but you can shoot radiation through it and read the shadows of the metal inside. Price published a reconstruction showing more than 30 gears and crucially established that this was a real geared calculating device, and not a fluke. Then came linear tomography work in the 1990s, and then the big one. In the mid-2000s, an international research project brought in equipment that had no business being pointed at an archaeological artifact.

A custom-built micro-focus computed tomography scanner weighing several tons designed for industrial inspection was transported to Athens specifically to scan the fragments in three dimensions at very high resolution. Alongside it, they used a photographic dome that lights an object from dozens of different angles and combines the images to make surface detail pop out of a flat, worn, illegible face. The results were absurd. The tomography. Let researchers count teeth on gears that nobody had ever seen, buried inside solid corrosion, and trace how they meshed. The surface imaging let them read inscriptions that had been invisible for two millennia, thousands of characters of text, including what amounts to a user manual engraved on the plates.

The device started explaining itself. Here is what it does, and this is the part where you should probably sit down. On the front, a dial with the Zodiac and an Egyptian calendar ring, with pointers to the zodiac and an Egyptian calendar ring, with pointers showing the position of the sun and the moon against the stars, and very probably the five planets known to antiquity. There is a little ball, half-silvered and half-dark, that rotates to display the phase of the moon on any given date, which is a design touch so elegant it would look good on a wristwatch today. But the real showpiece is hidden in the gearing, and it is the thing that makes engineers go quiet. The moon

does not move at a constant speed across the sky. Its orbit is elliptical, so it speeds up and slows down over the course of a month, and that variation is a genuine pain to model. The builder of this device handled it with a mechanism in which one gear carries a pin that sits in a slot on another gear mounted slightly off centre, so that as the pair rotates, the pin slides in and out of the slot, and the output speed smoothly varies, fast, then slow, exactly matching the lunar behaviour. That is a mechanical solution to a problem in celestial mechanics, built by hand, before the invention of the screw press. It is not just clever.

It is the kind of clever that suggests a long tradition of similar work behind it, because nobody produces that on their first attempt. On the back spiral dials. One tracks the 19-year cycle over which the lunar months and the solar year come back into alignment, laid out as a spiral of 235 months so you can follow the calendar around and around. Another handles a longer refinement of that same cycle. Another is an eclipse predictor running on the roughly 18-year cycle after which eclipse is repeat in a similar pattern, and it is not a vague indicator.

It carries little symbols specifying which kind of eclipse to expect and even what time of day. Visit BetMGM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and Game Sense remind you to play responsibly. 19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone.

close to you. Peace contact connects Ontario at 1-866-531-2,600 to speak to an advisor, free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. My name is Siaia and I live in the most beautiful place on earth. Seven months ago, I blew up my life and a lot has shifted since then. Everyone is moving forward. I'm still working on that. And that's me.

Welcome to Ice Cove! Where are the Penguins? That's Antarctica. We're the other one. North of North. Stream all episodes, available on CBC Gem. There is a further dial correcting for the fact that the eclipse cycle does not come out to a whole number of days, so the times shift by a set number of hours each time around. And there is a dial listing athletic games on a four-year cycle, including the Olympics,

because apparently the same instrument that predicts eclipses will, also remind you when the sports are on. Priorities. Nothing of comparable mechanical complexity is known anywhere for the next 14 centuries. The mechanism is generally dated to the second century before the common era, and the wreck itself a few decades later, probably a heavily loaded ship carrying Greek luxury goods and art westward toward wealthy Roman buyers, which is a polite way of saying that a lot of this material was leaving Greece under circumstances the original owners would not have described as a sale. Work at the site. never really stopped. Jacques Cousteau dived it in the 1970s. A modern long-term project has been

returning to the wreck since 2012 with rebreathers, robots, and even an armoured atmospheric suit, and it keeps producing, including a bronze arm, human skeletal remains from someone who went down with the ship, and in 2022 a marble head that appears to belong to a statue recovered over a century earlier. The site is not done. It may never be done. Now hold that outcome in your head, and look at the exact opposite case. In 1511, a Portuguese carac called the Flood de la Maire was sailing from the newly conquered city of Malacca toward Goa, commanded in practice by the fleet of Afonso di Albuquerque,

the man who had just taken Malacca by force. Malacca sat on the strait that controlled the trade route between the Indian Ocean and the South China Sea, and taking it was a strategic jackpot. Albuquerque spent months there, and when he sailed he took the accumulated wealth of the Sultan's. with him. The manifest, as it has come down to us through a long chain of retelling, is the kind of list that makes treasure hunters go weak. Gold, jewels, diamonds, porcelain, furnishings from the palace,

including the Sultan's throne, diplomatic gifts from neighbouring kingdoms. Popular accounts throw around numbers like 60 tonnes of gold, and describe it as the richest cargo ever lost at sea, and I want to flag clearly that those figures come from centuries of enthusiastic repetition, rather than from a surviving customs document, which is a distinction that gets flattened every single time this story is told. What is not in doubt is that the ship was carrying the concentrated portable wealth

of a major trading capital. There was one obvious problem, and everyone involved knew about it in advance. The floor de la Mar was old. She had been built around 1502. She had done years of hard service in tropical waters, she leaked, and she had a reputation as a vessel that needed constant repair.

and probably should have been retired. The decision to load the greatest treasure hall of the campaign onto the most tired ship in the fleet is a choice that historians have been raising an eyebrow at for 500 years. It is the naval equivalent of moving house and putting all your valuables in the car with the check engine light on. In November 1511 the fleet hit a storm in the Strait of Malacca. The old Carrick was driven onto shoals off the coast of Sumatra and broke apart.

Albuquerque survived, reportedly getting off on an improvised raft, which is the kind of luck that follows certain historical figures around like a loyal dog. Much of the crew did not, and the cargo went into shallow water off a coast, that is, from a search perspective, close to the worst possible place on earth to lose something. Here is why.

The strait of Malacca is shallow, which sounds helpful and is not. It is fed by rivers pouring off Sumatra, carrying enormous volumes of fine sediment. It has strong tidal currents that constantly move sense, sand. Visibility in the water is frequently close to nothing, and the seabed is not a stable surface with a wreck lying on it. It is a moving system of banks and channels covered in metres of soft mud. Anything that went down there five centuries ago is not sitting on the bottom waiting to be photographed. It is buried, deep, somewhere under a landscape that has been reshuffling

itself continuously since the day it sank. The contemporary accounts give an approximate area. Approximate in that environment is not a location, it is a region. So the searching began and it has never stopped. Treasure hunters, salvage companies and self-funded expeditions have gone after this wreck for decades. Claims of discovery have been made repeatedly. Permits have been sought, granted, revoked and fought over. Estimates of the cargo's value have been announced to the press with a confidence that is inversely proportional to the evidence,

because in the salvage business a big number is a fundraising tool. And to this day, nothing verified has come up. No confirmed hull, no confirmed cargo, five centuries, dozens of attempts, zero authenticated results. What has been produced instead is a legal problem that is honestly more interesting than the gold. Because who owns it? Portugal has a claim because it was a Portuguese state vessel and states generally retain title to their sunken warships, regardless of war. where they sink or how long ago. Malaysia has a claim because the treasure was taken from Malacca, which is Malaysian territory, and the wealth was seized in a conquest, which makes the whole thing look less like a cargo

and more like stolen property in transit. Indonesia has a claim because the wreck is almost certainly in Indonesian waters, and a country generally has jurisdiction over what sits on its own seabed. Three completely reasonable arguments, pointing three different directions over an object that nobody has found. This is a custody battle over a hypothetical, which brings us to the comparison, and it is a genuinely sharp one,

two wrecks, two cargoes, opposite outcomes. At Antiquithra, the item that changed the world was worth nothing at market. If you had handed that lump of corroded bronze to a scrap dealer in 1901, he would have laughed at you. It looked like garbage. It was treated as garbage for over a year,

and it turned out to contain a mechanical model of the heavens that forced a complete rewrite of what we thought ancient engineers could do. Its value was entirely informational, and informational value is invisible until somebody looks properly. The cargo of the floor de la mar is the reverse. Its value is entirely monetary, and monetary value is extremely visible, which is precisely why it has attracted 500 years of attention and produced almost no knowledge. Even in the best case, if it were found tomorrow, intact, what would we learn? That Albuquerque took Malacca, which is documented.

that the Sultanate was wealthy, which is documented. That 16th century Portuguese Carrick sank in storms, which is spectacularly documented. It would be a magnificent hall of beautiful objects and a fairly thin historical contribution, and it would immediately be swallowed by litigation. There is a rule hiding in that pairing, and it is worth stating plainly.

The most valuable thing on a wreck is almost never the most expensive thing on it. Gold is inert. Gold does not tell you anything. because gold from 1511 looks exactly like gold from 1811. Whereas a broken cooking pot tells you what people ate, a custom seal tells you what the tax regime was, a personal seal tells you who was a board, and a lump of green corrosion can tell you that somebody in the Hellenistic world

could model the wobble of the moon in bronze, which is also why the search patterns look so different. One site was found by accident by working men, and then investigated for over a century by scientists with increasingly ridiculous imaging equipment, and it is still producing new material. The other has been hunted by people motivated by an estimated dollar figure, in conditions where the sediment guarantees failure, and it has produced press releases and lawsuits, and notice what the successful cases in this half of the video have in common.

Heraclean, Rungholt and the mechanism inside that lump all came down to seeing through something opaque. Mud, in the first two, corrosion, in the third. In every case, the only case, the object had been sitting there the whole time, complete, unhidden in any meaningful sense, and utterly inaccessible until somebody had a machine that could look through the thing in the way. Which naturally raises an uncomfortable question. If the answer is usually sitting right there waiting for the right instrument, what happens when there is nothing left to look at? No mud to scan, no corrosion to penetrate, no wreck on the bottom at all, just an intact vessel, floating perfectly well in good condition with everything in its place and nobody aboard.

Because that category exists, there are more cases of it than you would like, and the explanations are much stranger than the ones people usually reach for. The traditional way to handle this material is to give each abandoned ship its own dramatic segment, complete with spooky music and a shrug at the end. We are not doing that, because if you line the cases up side by side, the same handful of variables keep deciding the outcome, and once you can see the variables the fog burns off remarkably fast. So here is the checklist we're going to run on all five,

and I want you to keep score as we go. First, does the ship exist in the paperwork, meaning a registry entry, a home port, an owner, a crew list? Second, is there a physical hull that qualified people actually walked around on and wrote down what they saw? Third, what is missing compared to what stayed, because a crew that leaves in a hurry takes a very specific set of things

and leaves an equally specific set behind. Fourth, was the vessel genuinely sinking or did it merely look like it was sinking to the people standing on it? Fifth, how much time passed between the last known moment aboard and the moment somebody else climbed the rail

because that gap is where all the good false evidence gets manufactured by wind and water and sixth, the one nobody applies often enough, who profits from the story being unsolved. Start with the one everybody has heard of and prepare to lose some cherished details because most of what you know about it was written by a novelist who was making things up on purpose.

In November of 1872 a merchant brigantine sailed out of New York Harbour bound for Genoa, carrying 1,1 barrels of industrial alcohol destined for the wine trade where it would be used to fortify cheap Italian product. A board were the captain, Benjamin's, Spooner Briggs, his wife Sarah, their two-year-old daughter Sophia, and seven crew. Ten people. Briggs was, by every surviving account, a careful, sober, religious professional

who did not take unnecessary risks, which is worth stating up front because a lot of the wilder theories require him to behave like a lunatic. A month later, roughly midway between the Azores and Portugal, another ship spotted her moving oddly. That ship was the day Gratia, and by a coincidence that has fueled two centuries of accusations, her captain David Morehouse knew Briggs personally, and their vessels had loaded in New York within days of each other. Morehouse sent his mate, Oliver DeVoe, across in a boat to see what was wrong. What DeVos found is the entire case, so let us be precise about it rather than atmospheric. The ship was afloat and in fundamentally sound condition. Some sails were set, others were furled, and a few were

damaged or missing, consistent with days of unattended sailing in variable wind. There was about a metre of water sloshing in the hold, which sounds alarming and is not. On a wooden ship of that period, that quantity is a job for the pumps, not an emergency. The pumps themselves were a story, though, because one had been taken apart, apparently while somebody was trying to sound the depth of water in the hold. The cargo was largely intact. There was food and water aboard for months. The crew's personal belongings were still there, along with items of value, which immediately makes piracy an awkward theory, unless you want to propose pirates who came all that way and then took nothing. And the yawl was gone. The ship's small boat had been launched, and there is

evidence it was taken from a specific spot with the hatch nearby opened, and there are indications a line had been rigged. That is not chaos. That is a procedure. Somebody made a decision and executed it in an orderly fashion. The last log entry was made on the 25th of November, placing the ship near the island of Santa Maria in the Azores, and it recorded nothing worrying whatsoever. Nine days of silence follow. Now, the salvage hearing in Gibraltar,

which is where the mythology gets its foundation. The presiding attorney general was a man named Frederick Solly Flood, who approached the case with the calm neutrality of a bloodhound that has just been insulted. He was convinced there had been. foul play, possibly a mutiny fuelled by the alcohol cargo, possibly a conspiracy between the two captains to split a salvage payout. Stains were examined and hyped as blood. A sword was found under a bunk and hyped as the murder weapon. Marks on the bow were described as deliberate cuts.

Analysis of the stains did not support blood. The sword was rusty, not bloody. The marks were consistent with ordinary wear. Nothing was proven, no charges were possible, and the salvage award came through at a notably reduced level, which was the legal system's way of saying we cannot prove you did anything but we would like you to know we have our eye on you. Then comes the detail that quietly solves most of the puzzle, and it turned up in Genoa when the cargo was finally unloaded.

Of the 17001 barrels of alcohol, nine were empty. Not smashed, not stolen, empty. And the reason is beautifully mundane. Those nine barrels were made of red oak, while the rest were white oak. Red oak is more porous. It leaks, which means that during the voyage, nine barrels had been quietly venting alcohol vapour into a sealed hold in a warming climate as the ship moved south. Put yourself on that deck. You're carrying an enormous quantity of flammable liquid.

Below you, the hold is building pressure and fumes. You open a hatch to check and you get a rush of vapour, possibly a noise, possibly a bang as pressure equalises. Your pump is in pieces and you cannot get a reliable reading on how much water is below. You have your wife and your two-year-old aboard. What is the safe move in your judgment in that moment? You get everyone off the ship, into the boat, on a line, and you sit fifty meters astern for an hour until you're confident it is not going to detonate. Then you pull yourself back and get on with your day. That is almost certainly what happened. And then the wind freshened, or a wave snapped the toe line, and a fully rigged ship with sail set began doing what fully rigged ships do, which is sail. Faster than ten people can row,

downwind, over the horizon, in minutes. Ten people in an open boat in the mid-Atlantic in December with no ship is not a mystery. It is an outcome. There are supporting variants worth mentioning. A water spout passing over the vessel would account for the water in the hold, the disturbed rigging and a sudden panic all at once. There is also a navigational argument that Briggs may have believed himself closer to land than he was. But the underlying shape is the same in all of them, and it is the shape you should now recognise. The ship was never in serious danger, and the people believed it was, and the people were the fragile part. So why does everyone think there was a half-eaten breakfast still warm on the table, and a cat asleep on a locker and cups of tea steaming away

in the galley? Because in 1884, a young and not-yet-famous doctor named Arthur Conan Doyle published a short story about the case, anonymously, in a magazine. It was fiction. He knew it was fiction. He invented the still-worn meals, the eerie, perfectly preserved domesticity, and a lurid plot involving murder and revenge on the high seas. He also slightly misspelled the ship's name, and that misspelling is the version most people still use today. It was so effective that officials involved in the real case reportedly wrote in to complain about the invented details being taken as fact. A century and a half later, the fiction is still winning, which is either a testament to Conan Doyle's talent or an indictment of everyone's fact-checking, and I genuinely cannot decide which.

Score the checklist. Documented ship. Yes, thoroughly. Physical hull inspected by professionals within hours, yes. Missing versus present, extremely informative. Actually sinking, no. Time gap. nine days, long enough for weather to rearrange the evidence but not long enough to erase it, who profits from the mystery, novelists and later an entire publishing industry. Now the second case, which is far less famous, considerably more disturbing and much better documented because it happened in the age of a radio and government inquiries. In October of 1955, a motor vessel called the Joyita left the port of Apia in Samoa, heading for the Tokulow Islands with 25 people aboard, 16 crew and nine passengers,

plus cargo including medical supplies, timber and food. The captain was an experienced local skipper. The trip should have taken about two days. The ship itself had an entertaining backstory. She had been built in the early 1930s in California as a luxury yacht for a Hollywood film director, then requisitioned by the United States Navy during the war for patrol and supply duties, then sold into the far less glamorous world of Pacific Island trading and fishing, and during her working life the holds had been lined with cork for insulation.

Remember that, because it is the single most important physical fact in the entire case. She never arrived. A search was mounted, aircraft covered a huge area of ocean, and nothing was found. Then, five weeks later, a merchant ship spotted her about a thousand kilometres from where she should have been, drifting, listing heavily to one side, partially swamped with her decks awash and empty. Everyone was gone, all 25. Also gone were the three liferaths, the logbook, the sextant and chronometer, the firearms and a portion of the cargo.

The ship was in an appalling state, wallowing, corroded and half full of water, but she was still floating, and here is where the cork comes in. A wooden hull with cork-lined holds is close to unsinkable in any practical sense. Fill her with water, and she becomes a miserable, waterlogged, unmanageable barge that continues to float indefinitely because the cork will not let her go down. Visit BetMGM Casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and Game Sense remind you to play responsibly.

19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, Peace contact connects Ontario at 1-866-531-2,600 to speak to an advisor. Free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. The captain reportedly knew this and had said so,

which sets up the cruelest possible version of the mismatch we saw before. The vessel could not kill them. The vessel was, by construction, a lifeboat with an engine, and 25 people got off it anyway. What was found aboard is where it gets grim. There was a doctor's bag on deck containing surgical instruments, and there were bandages, four of them stained with blood. That single detail has powered decades of speculation about violence and murder.

It is also entirely consistent with something much simpler, because there was a doctor among the passengers, and in an emergency involving a failing engine room, a heavy list and people moving around a pitching deck, injuries are not just possible, they are expected. The radio told a similar story of frustration rather than horror. The set was tuned to the international distress frequency, which is exactly what you would do if you were in trouble. But the antenna lead was broken, so they could listen and they could transmit a signal

that would carry a couple of miles into an empty ocean from a ship that was already far off any shipping lane. They were calling for help on a phone with no service, and they had no way of knowing whether anyone could hear them. The engineering findings finished the picture. There had been a failure in the cooling system involving a corroded pipe and someone had jammed a mattress against the leak in a desperate improvised repair, which tells you everything about how the last hours went.

The bilge pumps were blocked and not doing their job. One engine was out of action. Water was coming in below the water line, the ship was taking a list, and to anyone standing on that sloping deck at night, watching water rise while the pumps did nothing, and the radio silently failed. The conclusion would have been obvious and reasonable and completely wrong. She looked like she was going down.

She was not. The official inquiry concluded that the flooding was explicable and that the abandonment was not, and it was fairly blunt about the ship's condition being unfit for the voyage in the first place. The most plausible reconstruction is that the captain, the one man aboard who understood the cork-lining and would have told everyone to stay put, was either injured, incapacitated or dead by the time the decision was made. Take away the one person with the correct mental model of the ship, add darkness, a growing list, a dead radio and 24 frightened people, and the group makes the intuitive choice

which is to get into the rafts. Three small rafts, in open Pacific water, hundreds of kilometres from anywhere, at the start of a five-week disappearance. There are alternative theories, of course, piracy, insurance fraud, a wartime holdover conspiracy. They exist mainly because the honest answer is unbearable, and people would genuinely rather believe in villains than in a corroded pipe and a broken antenna wire. Checklist score. Documented, inspected, richly informative, not actually sinking, five-week gap, and the beneficiaries of the mystery

are anyone selling a more exciting explanation than plumbing. The third case takes place in the modern era, with a coroner, forensic examination and video footage, and it is important precisely because it is so recent that the usual excuses do not apply. In April of 2007, a catamaran of a catamaran of, about 10 metres set out from the Queensland coast in Australia with three men aboard, an experienced owner and two brothers who were less experienced but fit and capable. Three days later the boat was

found drifting off the coast with nobody on it, and the tableau aboard was, on first reading deeply strange, the engine was running, a laptop was switched on and operating, the GPS and the radio were both functional, there was food set out, the life jackets were still aboard, all of them, Fenders, the padded bumpers you hang over the side when you're coming alongside a dock, were deployed over the rail out in open water where they are completely pointless. One sail was up and shredded to ribbons. It was, as a scene, so odd that the story went around the world instantly and got exactly the treatment you would expect,

with the strong implication that something had come aboard and taken them. Then the investigation actually happened, and it is a masterclass in how weird a normal accident can look after three days of no. nobody being there to tidy up. The boat carried a camera, and footage recovered from it showed the men earlier in the voyage, relaxed, fishing, entirely ordinary, with no sign of conflict, fear or a fourth party. The fenders had almost certainly never been stowed after leaving the marina, which is the single most human detail in this chapter, because absolutely everybody who has ever

owned a boat has motored out with the fenders still hanging over the side while somebody's spouse points it out. The shredded sail is what any sail does when a boat sails itself unattended for days, luffing and flogging in changing wind until the cloth destroys itself. The running engine and the open laptop are not signs of a sudden interruption. They are signs that nobody was there to switch them off, which is what happens when the crew is no longer aboard. The coroner's finding was that this was an accident, and the reconstruction is depressingly ordinary. One man went into the water, most likely while dealing with a snagged fishing line, or when the boat moved unexpectedly. The others reacted the way human beings always react, which is by going in after him, without

thinking about the one detail that decides everything. A vessel under sail, even sailing badly and unattended, moves faster than a person can swim, not dramatically faster, just faster. Enough that within a minute the gap is unbridgeable, and within ten minutes the boat is a shape on the horizon, and you are three people in open water off the Australian coast with no flotation. That is the mechanism. Not a monster, not a wave, not a conspiracy. Just the fact that a boat does not wait for you, and that rescuing someone from the water is one of the most lethal things a person can attempt. Because the instinct to jump in is overwhelming and almost always wrong. The single most useful piece of information in this entire video is arguably

that you throw something that floats, you do not jump. Checklist, documented to the point of having its own coroner's report, Hull examined, a three-day gap, not sinking, and no beneficiary of mystery except the newspapers who ran the first version. The fourth case is barely a mystery at all, and I am including it because it shows the myth-making machinery operating in real-time, in our lifetimes, with satellites watching. There was a cruise ship built in the 1970s, ice-strengthened, named after a famous Soviet film actress, which spent years running expedition voyages to polar waters. Eventually she ended up impounded in Newfoundland over unpaid bills, with her crew stuck aboard and unpaid for months, which is an ugly and very common way for an old ship to die. She was sold for scrap, and in January of 2013 she was taken under tow toward the Caribbean to be broken up. The tow line parted. The ship drifted off into the North Atlantic.

Canadian authorities dealt with the immediate hazard by towing her clear of the offshore oil installations, and then, having established that she was no longer in Canadian waters and no longer their problem, they let her go. Officially, she was released as no longer a threat. Un Officially, one of the world's largest countries dealt with a derelict cruise liner by nudging it firmly toward the middle of the ocean and looking the other way, which is a solution I respect for its honesty. Off she went, unlit, unmanned, uninsured, drifting east. In March, two of her emergency beacons activated, which happens when a beacon gets immersed in water, and that is generally interpreted as the vessel taking water or breaking up. After that, silence, with a few unconfirmed

reports over the following years, and no verified sighting. She's presumed to be on the bottom, and here is why she belongs in this chapter. Within weeks, the story mutated. The British tabloids discovered that an abandoned ship might plausibly contain rats, and that rats on a ship with nothing to eat might plausibly eat each other. And so the country was treated to a sustained panic about a derelict liner full of cannibal rats bearing down on the coast. It was covered seriously, it was covered repeatedly, people wrote follow-ups, there were maps with arrows on them. There is no meaningful evidence that this ship was full of rats, cannibal or otherwise, and there is a fairly strong argument that a vessel that had been laid up and repeatedly inspected in port

was not exactly a rodent paradise. But that is not the point. The point is that a completely explicable event involving a broken tow cable and a scrapping contract, both of which are boring commercial facts, produced a full folklore layer in under a month in the age of satellite tracking. Ghost ship generation is not a Victorian craft that died out. It is a live industry with good distribution. Checklist. Documented. Never boarded after the tow parted. No crew to lose because there was no crew. Nothing mysterious about the mechanism and the beneficiary of the mystery is whoever wanted the clicks. Which brings us to the fifth case, the most famous one in the genre, the one with the most horrifying details and the most enthusiastic retelling, and the one where the

checklist collapses on the very first question. The story goes like this. In the late 1940s, in the waters near the Strait of Malacca, distress signals were picked up from a Dutch freighter. The messages reported that the officers and crew were dead, followed by a final incoherent transmission, stating that the sender was dying and then nothing. A rescue ship located the freighter. A boarding party went aboard and found the entire crew dead, scattered across the decks and in the wheelhouse, faces frozen in terror,

eyes open and staring upward, arms outstretched toward nothing, with no wounds on any of them. Even the ship's dog was dead, mid-snarl. The boarding party attempted to take the vessel under tow, then noticed smoke coming from the hold, evacuated, and the ship exploded and sank behind them. It is a magnificent piece of horror. It has everything. It has the dog. And it has one small structural problem, which is that there is

no evidence the ship ever existed. Run the first item on the checklist. Registry. Nothing. The name does not appear in the shipping registers where every commercial vessel of that period is recorded, and shipping registration in the mid-20th century was extremely thorough, because in Insurance existed, and insurers are not romantics. Home port, unspecified in every version. Owner, never named. Crew list never produced. So we have a ship on which everybody supposedly died in a manner unprecedented in maritime medicine,

and not one of the dead has a name. Rescue vessel. The identity varies between retellings, which is not what happens with real incidents, because the ship that finds a disaster files a report and its own crew tell their families. wreck none conveniently destroyed by explosion which is the narrative

equivalent of the homework being eaten by the dog that also died where the story actually comes from is far more interesting than the story it surfaced in the late

1940s in newspapers in the Dutch East Indies as a serialised piece attributed to a source whose account nobody could verify and it was told in the register of adventure writing rather than

repertage from there it was picked up and retold and at one point a version appeared in a publication associated with the United States Coast Guard in a section that was essentially maritime storytelling rather than investigation. That single appearance is the load-bearing beam of the entire legend, because ever since, people have been able to say the story appeared in an official government publication, which is technically true and completely meaningless.

A magazine printing a spooky sea yarn is not an inquiry, but the moment a legend can point at a government letterhead, it graduates. The explanations offered for the deaths are just as instructive, because people have gone to real trouble to solve a case that does not exist. There is a chemical cargo theory, in which the ship was secretly carrying nitroglycerin and cyanide compounds, and seawater in the hold released a lethal gas. There is a carbon monoxide theory involving a smouldering fire in the boiler room.

There is a clandestine wartime leftovers theory. Every one of these is a serious attempt to explain a set of symptoms that were described by an anonymous writer in a newspaper feature for entertainment. That is the intellectual equivalent of performing a detailed autopsy on a character from a novel and publishing the results. And now count the popularity. Of the five cases we have gone through, four have hulls, inquiries, coroners, engineering reports, cargo manifests and in one instance video footage. The fifth has none of those things, and the fifth is by a wide margin the most retold, most illustrated and most confidently described. There is a rule in there, and it is a

bleak one. The amount of vivid detail in a story is inversely related to how much of it can be checked. Real evidence is boring and specific. Fiction is free to be spectacular, because nothing constrains it. So what does the completed table actually tell us? Four things, and they generalise far beyond ships. The first is the one we have watched play out repeatedly today. Crews do not abandon good vessels. They abandon vessels they believe are about to kill them, and belief is the operative word. In every genuine case here, the gap between the real condition of the ship and the perceived condition of the ship is the whole story. Alcohol vapor is not an explosion, but you cannot tell from the deck. A list plus flooding is not sinking when the hull is stuffed with cork,

but you cannot tell in the dark with the pumps clogged. The sea does not need to kill anybody. It just needs to make a reasonable person make a reasonable decision. The second is a piece of pure physics that explains why we keep finding empty ships instead of empty ocean. An abandoned vessel is remarkably good at surviving. It has volume, buoyancy, a self-writing tendency, freeboard, a rig that will let it sail after a fashion for weeks. It does not get thirsty, it does not get hypothermic, and it does not get tired. A human in a raft has none of that, and a human in the water has led. than none. So the ship persists, sometimes for months, and gets found, while the people who

left it are gone within days. The empty deck is not evidence of something strange happening. The empty deck is the predictable end state of a very ordinary tragedy. We are essentially finding the packaging and being surprised that it survived better than the contents. The third is about time, the interval between the last human moment and the first inspection is a period during which wind, sea and gravity actively fabricate evidence. Sails shred, objects slide and fall, doors bang open, water enters. Rust runs and looks like something worse. Everything that looks staged in these cases was staged, just not by a person.

Three days will do it. Five weeks will do it comprehensively. If you want to know why the tableau on an abandoned ship always looks so wrong, it is because you are not looking at the scene of the event. you're looking at a scene that has been alone in the weather ever since, and the fourth is the one to actually take away. The mysteries that get solved are the ones that had a physical object at the centre of them, a hull to board, a pipe to inspect, a barrel to count, a camera to recover. When there is a thing, you eventually get an answer,

even if it takes 80 years and a court hearing conducted by a man with a grudge. When there is no thing, the story can never be solved, and it therefore never dies. which is exactly why it is the one you already knew. Which raises the obvious next question, what happens when the ship does not stay afloat for us to examine, when it does not turn up empty off the Azores or half swamped in the Pacific, but simply stops existing, mid-voyage,

with everybody still aboard, leaving nothing at all, no hull, no boat, no bodies, no wreckage, no distress call? Because that is a different category of disappearance entirely, and the way we eventually cracked it did not involve looking at all. It involved listening, and we will get to that. But first we need to understand what we are listening for, and that means understanding how a large ship can go from completely normal to completely gone in less time than it takes to make a phone call.

Because that is the actual problem with this category. We instinctively assume that a big vessel dies slowly. That is what every film has taught us. The hull is breached, the alarm sounds, there is a long, fraught sequence of men shouting in engine rooms, someone sends a message, boats are lowered, the band plays. That version of sinking exists and it does happen, and it is precisely why we find so few of these ships, because when it happens that way you get a position, a distress call,

survivors, floating debris and a search area the size of a postage stamp. The ships we are talking about now did none of that. They produced nothing, not a message, not a life belt, not a plank, and there are only a handful of physical mechanisms that can do that to a large steel vessel. All of them fast, all of them silent from the outside, and all of them far more common than the mythology surrounding them would suggest. Our first case is the most famous unexplained loss in the history of the United States Navy, and it is famous for entirely the wrong reason. In early 1918, a naval collier called the Cyclops was working the South Atlantic supply run.

She was a big ship for her era, over 160 metres long, built to carry coal to fuel the fleet, with a distinctive silhouette. Rather than a conventional superstructure, she had a row of enormous coaling derricks along her length and the crew accommodation and bridge perched up on towers above the hull, which made her look less like a ship and more like a floating industrial estate that had wandered out to sea by mistake. On this voyage she was not carrying coal. She had loaded roughly 11,000 tonnes of mangan. or in Brazil, bound for Baltimore, and that substitution is the first thread we're going to pull. She stopped unexpectedly at Barbados in early March, took on extra provisions, and sailed on the fourth. Abord were just over 300 people. She never arrived. No distress call was received, no wreckage was ever identified, no body was ever recovered. It remains the largest loss of life

in United States naval history, not connected to combat, and a century we still do not have the hull. Now, the reason you have probably heard of this ship is that she disappeared in a part of the Western Atlantic that, several decades later, an American writer would christen the Bermuda triangle in a magazine article because the phrase sounded good. The concept caught on with such enthusiasm that it spawned an entire publishing genre, and the Cyclops became the star exhibit, the case that supposedly could not be explained by ordinary means. It is worth pausing on how that industry worked, because it is beautifully lazy. The triangle cannon was assembled largely by copying earlier lists, and when a researcher named Larry Kusher went back to the primary

sources in the 1970s and checked the cases one by one, the results were devastating. Ships listed as vanishing in calm weather turned out to have gone down in documented storms. Vessels described as disappearing inside the triangle turned out to have been lost hundreds or thousands of kilometres away, in some cases in a different ocean. Incidents were listed with no mention that wreckage had in fact been found, or that the cause had been formally determined. The mystery was not discovered in the ocean. It was manufactured in the retelling,

mostly by people who could not be bothered to look things up, which in the pre-internet era was admittedly more work than it is now, but was also literally their job. So set the triangle aside and look at the ship instead, because the ship is genuinely interesting and the answer is almost certainly sitting in her cargo hold. Mangonese ore is dense. extremely dense. Coal, which is what this vessel was designed around, is bulky and light,

so a holdful of coal fills the entire space, and the weight is distributed through the volume of the ship, or is the opposite. 11,000 tonnes of it occupies a fraction of the space, which means it all sits in a compact heavy mass low down in the bottom of the holds, with a large empty void above it. That does two nasty things. The first is a stability problem that sounds like the opposite of a problem. Putting all that weight low gives the ship a very low centre of gravity, which makes her extremely stiff. Stiff sounds safe. Stiff is not safe. A stiff ship does not roll gently, she snaps. She rolls hard, stops abruptly and whips back over and over, and every one of those violent recoveries slams stress through the structure and through

the cargo. Sailors on ore carriers describe the motion as vicious rather than uncomfortable. The second thing is worse, and it is still killing ships today, which is, is the detail that makes this case a technical story rather than a spooky one. A dense granular cargo with a void above it can move. If the load shifts to one side during heavy rolling, the ship lists, the list makes the next shift easier, and the whole thing goes over in a couple of minutes with no warning at all, even more insidiously certain ore cargoes can liquefy. If the material contains enough moisture, sustained vibration from the engines and repeated wave impacts can raise the pressure in the water between the grains, until the pile

stops behaving like a solid heap and starts behaving like a heavy slurry, at which point it sloshes freely from side to side. This is the same physical process we met under the temple in Abukir Bay, just relocated into a cargo hold with a schedule to keep. That is not a historical curiosity. Bulk carriers loaded with nickel ore and similar cargoes have been lost this way in the 21st century, with cruiserboard, repeatedly in the era of satellite communication and modern hull design. There are international regulations specifically about moisture content in all cargoes, because ships keep capsizing without warning. We solved the mystery of what happened to these vessels, and the answer was so undramatic that it became a compliance matter, and ships still go

down doing it. Add to that the specific condition of this particular ship. One of her engines had failed, with a cracked cylinder, so she was steaming on a single engine at reduced speed and reduced manoeuvrability, which is a bad place to start from if the weather. turns. Her class had a hull structure that has been criticised ever since, with reports of buckling and corrosion in the longitudinal support members in ships of that type. She was heavily loaded, possibly overloaded, with cargo she was never designed to carry, running on half her power plant in the Western Atlantic in March, which is not a season known for its gentleness. And then there is the human garnish, which is where the story gets its flavour and none of its evidence.

Her captain was a German-born American with an eccentric reputation, described by some crew as tyrannical, and known for appearing on deck in an outfit best described as informal, which in wartime, with anti-German feeling running at maximum, was more than enough to generate a rumour. So for years the leading exotic theory was that he had defected and handed the ship over to Germany. After the war, German records were checked. No German ship, submarine or raider ever claimed the attack, and no German operation of any kind accounts for the vessel, which leaves us with a man who was mainly guilty of having an accent and questionable dress sense during a moral panic. Now here is the piece of evidence

that should have ended the argument decades ago, and it is the reason this case belongs in a chapter about mechanics rather than mysteries. The Cyclops had sister ships, identical vessels built to the same design, and two of them, named Proteus and Nerius, were later sold into commercial service. In late 1941, both of them were carrying dense or cargo in the North Atlantic. Both of them disappeared without a trace, within weeks of each other. Three ships, same class, same design, same category of heavy or cargo, same result. No wreckage, no signal, no explanation, spread across 23 years and two different wars, which rules out any single external cause and points directly at something intrinsic to that design carrying that load.

If three cars from the same production run, all leave the road on the same kind of corner, you do not investigate the corner for supernatural activity, you investigate the suspension, and notably the geographic argument collapses immediately, because the two later losses happened well outside the fashionable triangle, which means the one factor the popular story insists on is the one factor the ships do not share.

Our second case moves to the upper. other side of Africa and swaps a design floor for a piece of ocean that is genuinely measurably hostile. In 1909 a large passenger and cargo liner called the Waratah was on her second voyage running the emigrant route between Australia and Britain. She was a substantial modern ship of over 9,000 tonnes, less than a year and a half old, with accommodation for hundreds of passengers and enormous cargo capacity, and she was very much a symbol of a shipping line doing well. She left Durban on the 26th of July heading for Cape Town with 211 people aboard. Early the following morning she overtook another steamer, the Clan McIntyre, and the two exchanged

routine signals in the ordinary way, one of those completely forgettable maritime interactions that becomes historically enormous a few hours later. Then the weather turned into a serious southwesterly gale. Nothing was ever found. Not a boat, not a life belt, not a body, not a deck chair. And that total absence is unusual even for a sinking, because ships shed floating material generously. Wooden fittings, cork life belts, provisions, furniture. A ship that goes down over any length of time leaves a slick of evidence behind. This one left the ocean exactly as it found it. There had been unease about her before she vanished. Passengers on earlier legs had commented that she felt strange in a seaway, that she rolled in an unusual manner, that she was

slow to come back upright. One passenger disembarked at Durban and did not continue the voyage, having become convinced there was something badly wrong with the way she handled, a decision which at the time made him look like a nervous crank, and which in retrospect makes him the luckiest man in the Southern Hemisphere. The subsequent formal inquiry could not establish a cause, recorded that she was presumed lost in the gale, and was notably unimpressed by the state of the stability documentation, because full stability testing was not the rigorous standard requirement then that it is now, which is to say nobody could actually prove how she would behave when heavily loaded, and the only body of evidence on that question went down with her. But the environment

is where this case really lives, and the environment there is a genuine monster with a scientific explanation. Running down the southeast coast of Africa is the Agulhas current, one of the strongest western boundary currents on the planet. It carries an immense volume of warm water. southwest along the coast, at speeds that can reach several knots, hugging the edge of the continental shelf in a relatively narrow band. Now bring in a southwesterly gale, which blows in the opposite direction, and you have waves running head on into a fast-moving river of water. When a wave train meets an opposing current, the waves get compressed, wavelength shortens, height increases, faces steepen. Worse, the current is not uniform, and the variation in its speed

across the band acts like a lens, refracting wave energy and focusing it into narrow zones. Add the ordinary statistical behaviour of a random sea in which wave trains of different lengths occasionally line up crest to crest, and you can get an event that mariners reported for centuries and that oceanographers spent most of the 20th century politely disbelieving. A single wave two or three times the size of everything around it, often preceded by an unusually deep troth, described by the people who survived one as a hole in the sea with a wall behind it. The scientific establishment treated these accounts as exaggeration. On the reasonable grounds that standard wave models said such things should be vanishingly rare.

Then, on the first day of 1995, an oil platform in the North Sea recorded one directly with a downward-facing laser instrument. The wave measured over 25 metres in a sea state, where the significant wave height was around half that. It was not a story from a frightened bridge officer. It was a data file. After that, the argument was over, and rogue waves went from folklore to a field of study with conferences and funding, which is a transition we have now watched happen several times in this video.

The coast we are talking about is one of the world's hotspots for exactly this phenomenon, and it has a track record that is not remotely ambiguous. A large tanker broke completely in two off that coast in the late 1960s. In the 1970s, Other vessels in the same waters took catastrophic bow damage from single waves, arriving in port with their forward sections crushed downward like a soft drink can. These are documented incidents with surviving crews, damage surveys and insurance claims,

in the exact stretch of water where our missing liner was last seen steaming into a gale. Here is what such a wave does to a long ship, and it explains the total absence of debris. A hull is a beam. It is designed on the assumption that it will be supported along its length by water. water. Drop the bow into an abnormally deep trough while the stern is still lifted, or perch the vessel with its midsection on a single crest, and both ends hanging in air, and you have loaded that beam far outside its design case. Steel is strong, but it is not infinitely strong, and if the hull fails amid ships, the ship does not sink, it comes apart. Two halves, each one open to the sea

at the break, each one full of dense cargo, each one going down almost immediately. Everyone aboard is inside a steel structure that is filling and descending within seconds. Nobody makes it to a boat. Nobody gets a message out, and the sea surface closes over the whole thing having barely noticed. The search for that wreck became one man's decades-long obsession. From the early 1980s onward, Emlyn Brown ran expedition after expedition along that coast, using side-scan, sonar and magnetometers, chasing targets across a seabed that is deep, current swept and full of other wrecks, because that stretch of coastline has been eating ships for as long as ships have gone around it. In 1999 he announced he had found her. The target was the right size in a

plausible place sitting on the bottom. Then, in 2001, they got cameras onto it, and it was the wrong ship. It was a British cargo vessel torpedoed during the Second World War. Same rough dimensions, same rough spot, entirely different decade. That is the single most demoralizing sentence in the entire history of shipwreck hunting, and Brown eventually announced that he had exhausted his options and was done, which after 20 years of looking is not surrender, it is arithmetic. The false target is the important lesson, not the failure. A sonar contact is a shape, and shapes lie, exactly as they did on the rock formations earlier in this video. A hundred-meter steel object lying on its side on a sandy bottom looks like a hundred-meter steel object lying on its side

on a sandy bottom, regardless of what year it got there. The only thing that resolves it is going down and reading the identification, which is expensive, which is why so many rec searches consist of years of raising money to confirm that a promising blob is a different promising blob. So two cases, two total disappearances, and in both of them the underlying reason we know so little is the same. No signal came out. Nobody at any monitoring station recorded anything at the moment it happened, because there were no monitoring stations, and the fastest information channel available was a radio operator who needed to be conscious, at his post, with power, and with more than a few seconds of warning. Under the mechanisms we have just described, none of those conditions are met.

A capsize from a shifting or cargo can be over in under two minutes. A hull failure from a single wave is over in less than one, which is exactly why the third case is here, because it is the same category of event with one crucial difference, and that difference is the entire point of this chapter. In November of 2017, an Argentine Navy submarine, the San Juan, was returning from a patrol in the South Atlantic to her base at Mardel Plaza, with 44 crew aboard, including an officer who had become nationally known as the first woman to serve as a submarine officer in South America. On the 15th, the submarine reported a problem. Seawater had entered through the snorkel mast into the battery compartment, causing a short circuit and the beginning of a fire in the forward

battery bank. Water plus batteries is a genuinely serious situation on a submarine, both because of fire and because certain battery chemistries produce toxic gas when flooded. The crew reported that they had isolated the affected section and dealt with the immediate problem and they were ordered to head for base, routine in the sense that this is exactly the kind of casualty submarine crew's train for and survive. That was the last communication. They're nothing, and the largest international search operation in the region's history spun up, with ships and aircraft from more than a dozen countries covering an area of ocean so vast that finding a 70-meter object in it by looking was statistically close to hopeless, except that the ocean had already recorded what happened

and nobody had checked yet. There is a global network of hydroacoustic monitoring stations run under the treaty organisation that verifies compliance with the ban on nuclear testing. Its purpose is to catch anyone detonating a device underwater anywhere on Earth, and to do that it uses hydrophones suspended at a specific depth in the ocean, in a layer where sound gets trapped and can travel across entire ocean basins with very little loss. We will come back to that layer, because it turns out to be one of the strangest features of the planet. The point for now is that these instruments are extremely sensitive. They're always running, and they're always running, and they're they are sparse. There are only a handful of them scattered across the world's oceans, and they

are enormously far apart. When the recordings from stations located thousands of kilometres away were analysed, they contained a short, sharp, violent event registered about three hours after the submarine's last transmission in the vicinity of her last known position. It was not the signature of an explosion of ordnance, it was consistent with an implosion, the sound of a pressure hole collapsing. Visit BetMGM casino and check out the newest exclusive. The Price is Right Fortune Pick. BetMDM and GameSense remind you to play responsibly.

19 plus to wager. Ontario only. Please play responsibly. If you have questions or concerns about your gambling or someone close to you, please contact Connects Ontario at 1-866-531-2,600 to speak to an advisor. Free of charge. BetMGM operates pursuant to an operating agreement with Eye Gaming Ontario. That single acoustic event did two things at once.

It told the searchers when, and it told them, by comparing arrival times at widely separated stations, roughly where. It converted an open-ended disappearance into a bounded search problem, and it also, brutally answered the question everyone was avoiding, because a hull collapse at depth is not a survivable event. It is not even a perceivable event. The structure fails faster than a nerve impulse can travel, which is the only genuinely merciful sentence of a event.

available in this story. Even with that, it took a year. A commercial survey company took on the search with autonomous underwater vehicles under an arrangement in which they would only be paid if they actually found it, which is a business model that concentrates the mind wonderfully. In November of 2018, almost exactly a year after the loss, they located the wreck at a depth of around 900 metres, several hundred kilometers offshore, crushed and broken across a debris field. The physical evidence matched the acoustic evidence exactly. So compare the three outcomes, and the pattern is stark. Two ships lost in the first half of the 20th century,

one of them with a design and cargo combination that we can now identify as lethal, and that we watched kill two of her sister ships, one of them steaming into the most notorious rogue wave zone on the planet, both of them producing zero information at the moment of loss, both still missing. Both still, in popular culture, filed as mysteries, when what they actually are is well-understood physics that nobody was standing there to record.

And one submarine, lost in an event just as sudden and just as total, found and photographed, because a set of microphones several thousand kilometres away happened to be listening at the moment it happened. That is the whole difference, not better ships, not better luck, not a different ocean. The difference between missing and found is whether there was a sound in the archive. And once you see it that way, the entire history of ocean disappearance, splits neatly into two eras, with the dividing line falling somewhere in the middle of the 20th century. Before the ocean was wired for sound, a large vessel could simply cease to exist, and the only evidence generated was the absence of the ship at the far end of the voyage.

Afterward, the ocean began keeping a continuous record, and events that would previously have been permanent enigmas started leaving fingerprints in a data file on land. It is worth pointing out the uncomfortable corollary. Everything we have covered so far in this chapter concerns objects that failed to arrive somewhere and were therefore missed. Loss is defined by expectation. A ship disappears because somebody was waiting for it with a manifest and a schedule, which means the record is fundamentally biased toward things that were being tracked and says nothing whatsoever about anything that went down while nobody was keeping the books, which for most of maritime history is most of everything.

But there is a much stranger consequence of wiring up the ocean, and it is where we are going next. Because the network built to catch enemy submarines and nuclear tests was not designed to be selective. It cannot choose to hear only the things we're interested in. It hears everything the ocean does, and the ocean, it turns out, is extremely noisy, in ways that nobody expected, at volumes that nobody thought were possible, producing sounds that took decades to attribute to anything at all. Some of those sounds were sold to the public as evidence of something enormous and unknown living down there. One of them was so loud it was picked up on instruments across an entire ocean, and the truth behind them is, in the most literal sense,

colder and larger than a monster. To understand why the ocean can be wired for sound at all, you have to understand a piece of physics that sounds made up, and is arguably the single most useful accident in the natural world. Sound travels through seawater at a speed that depends on two things, pulling in opposite directions. Colder water slows sound down. Higher pressure speeds it up. As you descend from the surface, the temperature drops fast, so the speed of sound drops with it.

Keep going, and eventually you reach a depth where the water is about as cold as it is going to get, but the pressure keeps climbing, so from there down the speed of sound starts rising again, which means there is a depth, typically around a kilometre down in temperate oceans, where the speed of sound hits a minimum, and a minimum is a trap. Sound waves bend towards slower media,

so any sound that tries to leave that layer by heading up gets refracted back down, and any sound that heads down gets refracted back up. It is a channel, a horizontal wave guide running through the middle of the ocean, in which sound is prevented from escaping to the surface or the bottom, where it would be absorbed or scattered. Instead, it just keeps going sideways, and it goes an absurd distance. Researchers working on this in the 1940s established that a small explosive charge detonated in that layer could be picked up thousands of kilometres away. Not a nuclear device, a modest charge,

and instruments on the far side of an ocean basin hearing it clearly. The military applications were obvious immediately, and one of the first was a survival system in which downed air crew could drop a small charge set to detonate in the channel, allowing shore stations to fix their position by comparing arrival times, which is a rescue beacon powered entirely by the structure of the sea itself. Then the Cold War arrived and somebody realized that if the channel carries your sounds across an ocean, it also carries theirs. So the United States Navy built a system of underwater listening arrays, laid on continental slopes and sea mounts, and connected back to shore stations by cable, running continuously all day and all night for decades. The goal was

submarine detection, and the method was frequency analysis. A submarine, however quiet, is a machine, and machines produce tonal lines, a shaft turning at a particular rate, a pump running at a particular frequency, a propeller with a specific number of blades all put narrow spikes into the noise, feed that into an analyser that displays frequency against time, and a submarine stops being a stealthy shadow and becomes a set of horizontal lines slowly drifting across a chart. Trained operators could not only detect a boat across an ocean basin, they could often tell you what class it was. The whole thing was classified to a degree that is genuinely impressive given its scale, and it stayed that way until the early 1990s, when the Cold War ended,

and somebody in Washington was left holding an enormous, expensive, superbly maintained global listening network with a suddenly reduced number of Soviet submarines to point it at. And here is where the story becomes wonderful, because rather than switching it off, they let scientists in. Oceanographers were given access to portions of the system, and a research program was set up to use military hydrophone data for civilian science. Imagine being handed the keys to the most sensitive listening apparatus ever built, and being told to go and find out what the ocean actually sounds like when nobody's looking for a submarine. The results were immediate and slightly embarrassing for everyone who had assumed the deep sea was quiet. They could hear underwater volcanoes erupting in real time. In the early 1990s, the network picked up a swarm of seismic events on a mid-ocean ridge.

Researchers correctly interpreted it as an eruption in progress, and a ship was sent out to arrive while it was still happening, which had never been possible before. Previously discovering a deep sea eruption meant finding fresh lava on a survey years later and working backward. Now it was a live feed. They could track whale calls across entire ocean basins and follow individual animals for weeks. They could catch earthquakes, landslides and the low grinding of ice. They could also hear a great many things that nobody could identify at all. And being scientists,

they gave them names. And being scientists rather than marketing professionals, the names were terrible in a way that is now permanently baked into internet culture. The most famous of them arrived in 1997. It was a rising, warbling, low frequency sound lasting about a minute, and the striking thing about it was the power. It was picked up on sensors more than 5,000 kilometres apart, which means the source pumped an enormous amount of energy into the water. When researchers compared the profile with their catalogue of known sources, it did not match a ship, a volcano, an earthquake or any submarine that had ever been detected.

What it did loosely resemble in the shape of the rising frequency sweep was a biological call. It got the informal name Bloop, and the internet did the rest. Because the way the story got told to the public was that the loudest sound ever recorded in the ocean came from something alive and unidentified, which is a sentence engineered in a laboratory to produce nightmares. The estimates going around suggested a source far louder than a blue whale, and blue whales are the largest animals that have ever existed. So the implied logic was that something bigger than the biggest animal in history was out there, calling, and we did not know what it was.

Then somebody noticed that the rough position was in the southern Pacific, in the same general region of ocean where a certain horror writer had placed a sunken city inhabited by an enormous sleeping entity, and the case was closed as far as the internet was concerned. HP Lovecraft was vindicated, the old ones were real, and thousands of forum posts wrote themselves. I should note that the researchers running the program never claimed it was an animal. The animal framing came from the way the finding was communicated,

and then amplified. But it took years for the actual answer to arrive, and the actual answer is better, in the specific sense that it is larger and colder and involves no tentacles whatsoever. It was ice. Specifically, it was an ice quake. When a very large iceberg cracks, fractures, or breaks up, or when a mass of ice grinds against the seabed or against another mass of ice, it releases a colossal amount of energy into the water, and the resulting sound has a frequency structure that sweeps as the fracture propagates. The confirmation came from comparing the mystery recording with acoustic monitoring done in Antarctic waters, where researchers deployed instruments near ice shelves and recorded

ice doing exactly this, producing signals with the same character. The profile matched. Ice, on the scale of a moving mountain, sounds a lot like something enormous groaning. And that is not a disappointment. It is a recalibration. Consider what is actually being described, a slab of ice the size of a small country, weighing billions of tonnes, breaking, and the resulting crack being audible from 5,000 kilometres away through the water. That is a monster. It just has a different job description than the one we hoped for. The same explanation cleaned up most of the rest of the catalogue, because once you know what ice sounds like, the mystery list starts looking suspiciously polar. There is a sound from 1999 given the name Julia,

which lasts about 15 seconds and was heard across the equatorial Pacific array, and the leading interpretation is a very large iceberg running a ground, which is to say the sound of an ice mass the size of a city arriving somewhere it did not intend to be, and stopping in a hurry. There is a sound recorded in 1997 that descends slowly in frequency over about seven minutes, which earned it the name Slowdown. Seven minutes is a very long single event, and the interpretation again involves ice, this time.

sliding, with friction against rock producing a long, gradually changing tone. It is essentially the acoustic version of a very heavy object dragging, and the deceleration you can hear in it is the source physically slowing to a halt. There is upsweep, which is different and more interesting because it is still going. It was identified in the early 1990s. It consists of a long train of upward sweeping tones. It is seasonal, with clear peaks at certain times of year, and it has been traced to a region of the South Pacific associated with volcanic seamount activity. The current best explanation involves hot magma or hydrothermal activity interacting with seawater, and the seasonality is a genuinely open question.

It has been playing for over three decades and has never stopped, which makes it less a mystery sound and more a permanent installation. And there is whistle, which is a nice cautionary case because it was captured on a single hydrophone. One sensor gives you a signal and nothing else. You cannot triangulate a position with one microphone any more than you can find someone in a dark warehouse with one ear. Without a location, you cannot compare the event to known ice, known volcanoes or known shipping, so it stays unattributed, not because it is inexplicable, but because it was insufficiently witnessed. The unsolved status is a limitation of the observation, not a property of the thing.

That pattern is the real point of this whole catalogue. Nearly every famous mystery sound of the deep turned out to be geology or ice. Almost none of it turned out to be biology, which is precisely the opposite of what the popular versions assumed, and the reason for the mistake is worth naming. A frequency sweep, a rising or falling tone, a pulse with structure, all of these feel like communication to us, because our brains are optimized for voices.

We are the species that hears words in a washing machine, show a human arising warble, and they will hear a call, and no amount of professional training completely removes that reflex. It just teaches you to check. There is also a technical trap in the loudness comparisons that got everybody excited, and it is worth 30 seconds because it explains why underwater numbers always sound apocalyptic. Sound levels in water are quoted against a different reference pressure than sound levels in air, and the convention involves stating the level as it would be at one metre from the source,

even when the source is enormous and distant. The result is that perfectly ordinary underwater events get quoted at decibel figures that, if you naively compared them to air values, would suggest something louder than a rocket launch. Every time you see an underwater sound described with a terrifying number, some of that terror is a unit conversion. Not all of it, because the ocean genuinely is loud, but enough that the number alone should never be the argument. Now, having spent all that time establishing that the mystery sounds are mostly ice, let me hand you the counter-example, because there is one case where the answer really was an animal,

and getting to that answer took a decade and a machine learning model. In the mid-2010s, autonomous underwater gliders were deployed in the Western Pacific near the Mariana Trench, drifting and diving on programmed patterns and recording as they went, which is a much cheaper way to monitor an ocean than parking a ship in it. In among the data was a call that nobody could place, and it was strange in a very specific way. It had five parts. It swept across an enormous frequency range, from a deep, low rumble all the way up into the high thousands of hertz, and it lasted only a couple of seconds. And the ending was the problem, because the final component was metallic. It had a tinny, ringing artificial quality, the sort of sound you would expect from struck sheet metal or a synthesizer, not from a large mammal. It got named the Western Pacific Bio-Twang, which is another entry in the ongoing evidence that, O'SRexeckel.

oceanographers should not be allowed to name things. The immediate suspicion was that it was some kind of baleen whale because the low-frequency component fits, and because whales in that part of the world are known for producing calls that sound like sound effects from a science fiction film. There is a well-known minky whale called from Australian waters that has a mechanical pulsing quality so odd that it was originally attributed to boats or even to the seafloor.

So a strange metallic whale was not unprecedented, But suspicion is not attribution, and the problem was that nobody had ever been sitting on the surface, watching an identified animal. At the moment one of these calls was recorded, which is the recurring difficulty of acoustic biology. You have a voice and no face. The way it was eventually solved is thoroughly modern and slightly unromantic. Rather than trying to catch an animal in the act, researchers took an enormous archive of recordings on the order of hundreds of thousands of hours and trained a neural network

to identify every occurrence of the call. Then they took the resulting map of when and where the call happened, across years and across a huge stretch of the Pacific, and compared it against records of where particular whale species had actually been observed from ships and surveys. If the acoustic detections and the visual sightings

of one species line up in space and time, over and over, across seasons and across years, you have a statistical identification. The answer that came out was Bride's Whales, a Baleen species that lives in warmer waters and is far less studied than its more charismatic relatives. The pattern held up, including seasonal movements, and the work also picked up variations that appear to track large-scale climate shifts in the Pacific, which is a bonus result that nobody was even asking for. Notice how different that solution looks from everything else in this video. There was no expedition, no submersible, no dramatic encounter.

somebody solved a decade-old mystery by pointing a classifier at a hard drive. The heroic version of ocean science involves a ship and a storm. The version that actually closes cases increasingly involves a data centre and a very patient graduate student, and I say that with total admiration, because the whale did not care how it was identified and neither should we. So that is the natural half of the archive. Now for the human half, which is funnier and considerably more expensive. A network that hears everything has a specific, occupational hazard, which is that it will hear things and you will have to decide what they are,

and during a period of intense military tension, the default interpretation of an unexplained underwater contact is not going to be marine biology. It is going to be the enemy. This started well before the deep sound channel was wired up. During the Second World War, escort vessels hunting submarines with active sonar spent an extraordinary amount of effort attacking things that were not submarines. Sonar returns came back from whales, which are large, mobile, air-filled, and give a beautiful echo.

They came back from wrecks on the bottom, which do not move, but do look promising until you have wasted an hour. They came back from thermal layers in the water, boundaries between water masses of different temperature, which reflect and bend sound, and can produce a strong false return or more dangerously hide a real target completely,

and they came back from the single, strangest false contact in naval history. Sonar operators began reporting a solid bottom at a few hundred meters in places where the actual seafloor was kilometres down. Worse, this false bottom moved. It rose toward the surface at night and sank back down during the day, reliably every day, which is not behaviour anyone expects from a seabed. For a while this was a genuine tactical concern, because a large moving reflective layer in the water is exactly what you would used to hide under if you were a submarine commander with imagination. What they had found was the deep scattering layer, and it is one of the great discoveries in biology, accidentally made by people

trying to kill each other. It is not geology at all, it is animals. A dense layer of small fish, squid, cyphernophores, crustaceans and plankton spread across the world's oceans, sitting in the dark during the day, and rising toward the surface at night to feed, then descending again before dawn to avoid predators. Many of the fish involved have gas-filled swim bladders, which is why they reflect sonar so well, because a bubble in water is an excellent acoustic mirror. This is the largest daily movement of biomass on the planet,

happening every single day, in every ocean, and we discovered it because it kept interfering with anti-submarine warfare, which sets the tone perfectly for the strangest peacetime chapter in this story, which took place in Swedish waters. Sweden spent much of the Cold War in the geographically awkward position of being neutral, heavily armed, and sitting directly on top of the Baltic, with a Soviet naval presence on the other side of it. Swedish waters are a maze of shallow archipelago,

thousands of islands, narrow channels and complicated bottom topography, which is a nightmare for acoustics because sound bounces off everything. In October of 1981, the awkwardness became international news when a Soviet submarine ran aground on rocks near a Swedish naval base, not detected, not chased, not tracked, stuck, in plain view, in a restricted military area. The Swedish Navy did not need a hydrophone array for that one, they needed a camera and a tow rope. The submarine sat there for a week and a half while diplomats argued, and it remains one of the most humiliating navigational errors ever committed by a major power,

and also, crucially, proof that foreign submarines really were operating where they should not have been. Keep that in mind because it is the reason everything that followed was taken so seriously. The following year came a much bigger event. Swedish forces detected what they believed were foreign submarines inside a sensitive area of the Stockholm Archipelago and mounted a full-scale hunt lasting weeks. Depth charges were used. Mines were detonated. Helicopters and surface ships combed the water.

Investigators found what appeared to be track marks on the seabed, and nothing surfaced. no submarine was brought up, no vessel was identified, no crew was captured. Over the following years, reports of underwater intrusions in Swedish waters piled up into the hundreds. It became a permanent feature of national politics, with enormous defence spending attached, and a general public understanding that the Baltic seabed was essentially a highway for foreign submarines with a taste for provocation. Then, in the 1990s, Swedish researchers went back to the acoustic evidence and asked what turned out to be an extremely unpopular question, which was whether the characteristic sound the Navy had been treating as a signature

of foreign propulsion actually came from a submarine at all. The answer is my favourite fact in this entire chapter, so brace yourself. It was fish. Specifically it was herring. Sholes of herring release gas from their swim bladders, venting bubbles, and a large shoal doing this produces a distinctive crackling, hissing acoustic signature. Researchers recorded it, characterized it, compared it with the archived recordings the Navy had been using as evidence, and demonstrated that the typical signal matched a shoal of fish rather than a machine. A substantial part of one of the most expensive security panics in modern Scandinavian history came down to the digestive habits of a small silvery North Atlantic food fish. I want to be fair here, because it would be easy to turn this

into pure comedy, and the situation was not that simple. Real intrusions did happen, as the grounded submarine proves beyond argument. Not every contact was fish, and the Swedish Navy was not staffed by idiots. What went wrong was interpretation under pressure. When you have a national security expectation of finding submarines, an ambiguous acoustic signature does not stay ambiguous for long. It gets resolved in the direction everybody's already looking. That is not a failure of hardware, it is a failure of the process sitting on top of the hardware, and it is the same failure we saw with a lump of sandstone becoming a staircase. The coda is almost too neat. In more recent years, another major submarine hunt in the archipelago produced a genuine confirmed contact, and separately the discovery of an actual submarine wreck on the bottom. When it was examined, it turned out to be a submarine from the era of the First World War, sunk around a century earlier, sitting quietly in the mud the entire time and finally getting the attention it deserved several generations late. It was not the intruder anyone was hunting. It was, however, a submarine, which, after all, that searching feels like the ocean offering a consolation prize with a straight

face. So pull the two halves of this together, because they say the same thing from opposite directions. On the natural side, a listening network built for war became the single most productive instrument in the study of the deep ocean, and the mysteries it generated were resolved slowly by comparison. Every case was cracked the same way. Get more recordings. get better coverage, build a reference library of what known things sound like, then match. Ice was identified by going to the ice and recording it. A whale was identified by cross-referencing millions of detections against sightings. Nobody solved anything by staring at a single spectrogram and thinking hard about it.

On the human side, the same network demonstrated that detection and attribution are completely different problems and that the second one is much harder than the first. hearing something is easy, and modern sensors will hear far more than you want them to. Knowing what made the noise requires context, comparison, and, above all, a willingness to accept an answer that makes your budget request look silly. And this is the honest state of play. There are still unattributed sounds in the archive. New ones get recorded regularly, because coverage keeps expanding and because most of the ocean has never been listened to at all, which means the correct way to read a list of unidentified ocean sounds.

is not as a catalogue of monsters. It is as a to-do list, written by instruments that are much better at detecting than we are at explaining, and work through at a rate of roughly one entry per decade. Sound, though, is a relatively honest medium. A recording is a measurement. It has a timestamp. It can be compared with other recordings and analysed by people who are not there. It resists interpretation for a while. Images are much weaker. An image arrives already looking like something, and it arrives inside a human visual system that has spent millions of years learning to find familiar shapes in poor conditions, which is why the next set of cases involves objects that are considerably less mysterious than the sounds we just discussed,

and yet have proven far more durable in the public imagination, purely because somebody took a bad photograph. In June of 2011, a small Swedish salvage outfit was working the Northern Baltic, in the gulf between Sweden and Finland, doing what small salvage outfits do, which is look for old wrecks with valuable cargo. The Baltic is genuinely superb for this, because it is cold, dark, low in salt and low in oxygen in places, which means wooden ships and their contents survive there in a state that would be impossible anywhere warmer. People have pulled drinkable champagne out of that seabed. It is a serious business with serious money in it. What they got instead was a sonar image of a large circular object sitting on the

bottom, at around 87 metres, roughly 60 metres across, with what looked like a track or a groove running several hundred meters behind it, as though the thing had slid or been dragged into position. A second, smaller object sat some distance away, and the image went around the world in about a day and a half for one reason that has nothing to do with archaeology. It looked like a spaceship. Specifically, it looked like a particular famous fictional spaceship from a famous film franchise. And once that comparison was made, the story stopped being a sonar contact

and became a cultural event. Newspapers ran it. Forums exploded. The number of people who could tell you the depth of that object rose by several orders of magnitude within a week, and approximately none of them

had ever looked at a side-scaned sonar image before in their lives. Which is the actual subject of this chapter, so let us slow down and explain what people were actually looking at, because almost nobody does, and it matters in no. enormously. Side-scan sonar is not a camera. It does not produce a photograph and it does not

see anything. A toad instrument dragged behind the boat, fires fan-shaped pulses of sound out to either side and then records how much energy comes back and how long it took. Hard, rough, upward-facing surfaces return a lot of energy and appear bright. Soft mud returns little and appears dark. Anything standing up off the bottom blocks the sound behind it and cast an acoustic shadow, which is why sonar images have shadows that look like sunlight from a low angle, even though there is no light down there at all. Then software takes those returns, which arrive as a series of individual scan lines and stitches them into a picture. And that stitching involves assumptions. It assumes the tofish was moving at a steady speed, on a steady heading at a steady height above the bottom. In reality,

it was swinging on a cable behind a boat in a current. So the software stretches and squeezes the image to make it look right, and every one of those corrections is a chance to introduce a smooth curve, a straight edge or a symmetry that does not exist in the water. Sonar artefacts are so common that reading them is a professional skill, and the standard experience of a working hydrographer looking at an amateur sonar mystery is a very long sigh. None of which proves that particular object was an artifact of processing. It probably is a real object because something is genuinely there. The question was always what it is made of. The team went back the first. following year and dived it, which deserves credit because that is the correct move and a lot of

people would have just kept selling the image. Visibility was poor, the depth is well beyond recreational limits, and what they described afterward was a rounded mushroom-like mass with a hole in it, rounded edges and stone-like material. They also reported that electrical equipment stopped working within a couple of hundred metres of the object, which is the sort of claim that is impossible to verify, impossible to disprove, and enormously helpful for a doctor. I will simply note that the number of independent parties who have ever measured an equipment failure at that site is zero. They did, however, bring up samples, and this is where the case gets settled to the degree it can be. The rock was analysed at a Swedish university and identified

as a mixture of the sort of material that turns up in glacial deposits, including granites, gneiss and sandstone. The kind of assortment you get when an ice sheet grinds across a landscape, picks up everything it passes, carries it for a few hundred kilometres, and then drops it in a heap when it melts. And that is exactly what the northern Baltic is. It is a basin that spent the last glacial period underneath a continental ice sheet several kilometres thick. That ice sheet did two things relevant here. It ground the underlying bedrock and deposited enormous quantities of till, which is the technical word for an unsorted pile of rock debris all over the region,

and it left behind, in the water column and around the margins. floating and grounded ice that dragged along the bottom. Which also explains the runway. A large boulder or an ice keel dragging across a seabed leaves a groove, sometimes hundreds of metres long, with the objects sitting at the end of it. Plow marks like that are so well known in formerly glaciated seas that they appear in geology textbooks, and they look exactly like something heavy skidded to a stop because something heavy did. It just did it during a deglaziation rather than during a landing sequence. So the honest state of the Baltic case is, is this. There is a real feature on the seabed. Its shape is unusual enough to be worth looking at.

The physical samples that have been taken point to glacial deposit. And crucially, nobody has ever gone out with a modern high-resolution multi-beam system mapped it properly at centimetre scale and published the raw data for anyone to check. That last point is why the story is immortal, and it is not because of any conspiracy. It is because of cost. A proper survey requires a suitable vessel, a mapping system, qualified operators and several days of ship time in the Baltic, which runs into serious money and produces, at best, a paper confirming that a rock is a rock. Nobody is funding that. Meanwhile, keeping the mystery alive requires a laptop and an existing sonar image, and it produces views, documentaries and crowdfunding. The economics point in exactly

one direction, and it is not toward resolution. That is symmetry between the cost of making a claim and the cost of checking one is the engine behind most of the cases in this chapter. Our second case is older, prettier, and, I would argue, the most perfect story in this entire video, because the answer to the mystery was published in a scientific book 76 years before the mystery was created. In August of 1964, a research vessel called the Eltonin was working the far southern ocean, west of Cape Horn, doing deep sea photography. This was serious oceanographic work, of the sort that involves lowering a camera on kilometres of cable into pitch darkness, firing a strobe and hauling it back up hours later with no idea whether you got anything.

Deep-sea photography in that era was closer to fishing than to photography. One frame, taken at nearly 4,000 metres, showed something with a very particular quality. Sitting on the sediment was a slender vertical stalk, apparently anchored on a splayed base like a tripod, and near the top was several perfectly regular horizontal crossbars, each ending in small, evenly spaced knobs, radiating outward in a symmetrical arrangement. It looked like an antenna, not vaguely, not if you squint. It looked like a piece of manufactured hardware, the sort of thing you would bolt to the top of a mast, sitting alone on an abyssal plane, 1600 kilometres from the nearest land, at a depth where the pressure is around 400 atmospheres. The photograph existed quietly in the scientific record for a while,

and then got a second life decades later, when writers of a certain persuasion rediscovered it and correctly identified it as the best-looking piece of evidence they had ever seen. And it is. Compared to a blurry sonar smear, this is a sharp, well-lit, well-composed photograph of an object with real structural regularity. If I were building a case for engineered hardware on the deep-sea floor, this is the image I would lead with. Here is what it is.

It is an animal. Specifically, it is a sponge of a genus that lives on storks in deep water with radiating arms bearing regularly spaced projections. The identification is not a guess, and it is not recent. A naturalist named Alexander Agassi described this form in the 1880s from specimens collected by deep sea dredging expeditions, and published an illustration of it. If you place that 19th century engraving next to the 1964 photograph, they match.

Same stalk, same base, same arranging. of arms, same little regular knobs. The organism had been formally described, drawn and printed in a monograph, generations before anybody pointed a camera at one in situ. So the sequence of events runs as follows. Science identifies an animal in the 1880s. A research vessel photographs the same animal in 1964. The photograph is later presented to the public as an unexplained structure of unknown origin, and the refutation required no new expedition, no new technology and no new funding. It required a library card. There are two details that make this even better. The first is scale, and it is the most underrated trap in deep-sea imagery. That object is roughly the size of a

dinner plate in height. It is small, and you cannot tell that from the photograph, because a deep-sea image has almost no scale cues. There is no tree, no person, no coin, no horizon, no atmospheric haze, no familiar object of any kind. You're looking at a strobe-lit patch of sediment with something on it, and your brain assigns a size based on nothing whatsoever. Show people an unlabeled photograph of a seafloor object, and their size estimates will span two orders of magnitude, and every one of those estimates feel certain to the person making it. Half the drama in deep-sea photography is generated purely by the absence of a ruler. The second detail is that the actual biology is better than the alien version, and this is a hill I will die on. Spongers of this group

carnivorous. In the deep ocean, where the flow of food particles through the water is thin, some sponges abandoned the standard sponge business model of filtering water and switched to catching prey. They snag small crustaceans on hooked microscopic spines, and then engulf and digest them slowly. That symmetrical radiating structure is a trap. Carnivorous sponges were not fully confirmed until the 1990s, when researchers studying an underwater cave found one, and demonstrated the mechanism, and there are now dozens of species known. So the true story is that a photograph taken in 1964 shows a predatory animal, sitting alone in permanent darkness under 400 atmospheres of pressure,

having given up filter feeding in order to catch and eat live prey with hooks, in a body plan so geometrically regular that it fooled the internet into thinking it was manufactured. And people found that disappointing. Honestly, the human capacity to be let down by reality is one of the great renewing, resources. The third case is the most ambitious of the three, and it is the one where the failure is not about interpretation of a single image, but about a claim that was never followed up by anybody at all, including the people who made it. In 2001, a survey company working with the Cuban government in the waters off the western tip of Cuba, announced something extraordinary. They had been

conducting seabed survey work in the area, primarily looking for historic shipwrecks, which is a reasonable thing to look for in the Caribbean, given the several centuries of Spanish treasure shipping that passed through it. Their sonar data, they reported, showed a field of symmetrical structures spread over a couple of square kilometres, described as blocks, pyramidal shapes and features suggesting roads and plazas, a stone complex, a city, at a depth of somewhere between 650 and 700 metres, and that depth figure, which sounds like a minor technical footnote, is the entire refutation, delivered before anyone has to argue about a single pixel. Here is why. We established at the very beginning that global sea level during the last glacial maximum

was around 120 metres lower than today, because that is how much water was locked into ice sheets. That figure is not a rough guess. It is one of the better constrained numbers in Earth science, cross-checked by drowned coral reefs, by sediment cores, by isotope records, and by drowned coastal landscapes on multiple continents. The ocean has not been 600 metres lower at any point during the existence of anatomically modern humans. There is not enough water in the world's ice to do it. Even if you melted nothing and froze everything available, you do not get anywhere close. So for a human-built structure to be sitting at 700 meters, the sea level explanation is unavailable, and you need the other option, which is that the ground itself went down. Tectonic subsidence. The block of

crust it was built on sank, and that is where it dies, because subsidence has a rate. Crustal blocks move vertically at speeds typically measured in fractions of a millimeter to a few millimeters per year. To drop a piece of seafloor six or 700 meters, you need a very long time, and the honest estimates run into the tens or hundreds of thousands of years at minimum, and that is being generous. Cuban geologists made precisely this point at the time, and it has never been answered. The structures would have to predate not just civilization, but, on most estimates, our species arrival in the Americas by an enormous margin. You cannot rescue the claim by making the builders older, because there is no plausible builder that old. There is an alternative, which is a

catastrophic single-event drop, and it is worth taking seriously for exactly as long as it takes to ask what evidence would exist. A block-dropping hundreds of metres in one go is an event that leaves a fault-scarp, a slump structure, a tsunami deposit ring around the entire Gulf in the Caribbean, and a seismic signature in the sediment record. None of that has been produced. Now, what happened next is the actual scandal, and it is a scandal of absence. Almost nothing. There was talk of further expeditions, there was footage from a remote vehicle, and it was murky, low-contrast video of a dark seabed which shows shapes if you want to see shapes. There was no peer-reviewed publication, there was no recovery of a single worked stone,

there was no independent survey by another institution with its own equipment. Two decades on, one of the most spectacular archaeological claims of the century sits with no follow-up data whatsoever, in an era when marine survey capability has improved dramatically and when a genuine find of that magnitude would be the easiest funding application in the history of science, that silence is informative. Not proof of dishonesty, and I want to be careful here, because there are perfectly boring reasons a survey does not get repeated, including money, permits, politics, and the fact that the company involved was in business to find shipwrecks and not to run a 20-year archaeology programme. But an extraordinary claim that generates

no subsequent work is a claim that nobody with the relevant expertise thought was worth the fuel. So we have three cases and one shared failure mode. Let us name the mechanisms precisely, because these apply far beyond the ocean. The first is that low resolution does not mean less information, it means more room. A blurry image is not a partial picture of the truth. It is a partial picture plus a large empty space that your visual system will fill in automatically

and without asking permission. And it fills that space with the most probable familiar object because that is what perception is for. The classic demonstration is not even underwater. In the 1970s a spacecraft photographed a hill on Mars under low sun angle at low resolution, and the resulting image looked like a human face staring up at the sky. It launched a genuine cultural phenomenon. Then better cameras arrived in later decades, and at high resolution it is a hill. Not a damaged monument, not a weathered sculpture.

A perfectly ordinary lump of Martian geology whose shadows happen to fall in a suggestive pattern on one particular afternoon. Resolution did not reveal what the face really was. Resolution removed the space in which the face was being generated, and the face went away, because it was never in the rock, it was in the gap. The second mechanism is the missing ruler, which we covered with the sponge, and which is worse underwater than anywhere else, because there is no natural lighting, no distance queue, no atmosphere and no context. The third is that sonar imagery arrives pre-formatted to look like a photograph of a landscape, and it is not one. It is a map of acoustic reflectivity that has been processed into something the human eye can pass,

and every processing decision, from contrast stretch to interpolation to colour palette, changes what a viewer perceives. A hydrographer looks at that image knowing what the software did, everybody else looks at it as if it came out of a camera. The fourth is the raw data rule, and it is the single most useful filter you can apply to any claim of this type. Ask whether the original unprocessed data has been made available for independent analysis.

not the documentary, not the rendered image, not the enhanced version, the raw sonar file, the uncorrected photograph, the sample and the location. When the answer is yes, real work follows, sometimes confirming and sometimes demolishing. When the answer is that the data is proprietary, or lost, or being saved for the next production, that is not evidence of fraud, but it is a guarantee that nothing will ever be resolved, and often that is the point. And the fifth is the economics I mentioned, which is the ugliest one. Producing a mystery is cheap and profitable.

Resolving a mystery is expensive and thankless. Ship time in deep water cost enormous sums per day, and funding bodies exist to support research programmes, not to send survey vessels out to disprove things on the internet. So false positives accumulate permanently, while corrections are rare, underfunded and much less fun to watch. Every one of these cases is still alive today, not because the evidence is balanced, but because nobody has the budget to bury them properly,

and the people who could have don't consider it their job. There is a fair objection to all of this, and I want to take it seriously rather than pretending it does not exist. Sonar and low-resolution imagery have found real things. The drowned settlements we looked at earlier in this video were found by remote sensing, in low-resolution, by people looking at squiggles on a screen and going, hang on. The technique is not the problem. Blurry anomalies are how discovery starts. The difference is entirely in what happens next. In the successful cases, the anomaly triggered a targeted follow-up. Samples came up. The material was analysed by people who had no stake in the outcome,

and the results were published in a form that let other researchers argue about them. The anomaly was a beginning. In the cases in this chapter, the anomaly was the product. It got announced, illustrated, monetised and defended, and the follow-up either did not happen or did not confirm. So the working rule for this whole category is a single sentence, and it is not sceptical so much as procedural. An unexplained image is a request for a second visit, and any claim that never received one has not survived scrutiny,

it has merely avoided it. Which puts us in an interesting position for the last part of this video, because so far every doubtful case we have examined has involved something that does not move. rocks, deposits, structures, seabed features, objects that sit there being misread. Living things are much harder and much more interesting, because they genuinely can be enormous, they genuinely can be unknown to science, and unlike a sonar smear, they have every reason to be somewhere else by the time anybody arrives with a camera, which means the same evidentiary rules have to be applied to a category

where, occasionally, the ridiculous answer turns out to be correct. Before we get to bodies, though, we need to deal with a variable that has been quietly deciding every single case in this video and that almost nobody accounts for, not the object, the water it happened to end up in. Because the ocean is not one environment. Treating it as a single place is like treating land as a single place, and then being surprised that Egypt and Siberia have different rules about what survives in the ground. Some patches of ocean are ruthless archivists that hold on to everything. Some patches are industrial shredders that will remove all trace of an event within a few weeks, and which one you fell into is mostly a matter of chemistry.

So let us look at the extremes, starting with the single most efficient preservation machine on the planet, which also happens to look like a hole punched through a coral reef by something in a bad mood. In the South China Sea, in the Paracel Island Group, there is a near-perfectly circular shaft in the reef. From above it appears as a dark blue disc surrounded by pale, turquoise shallows, which is why features like this are called blue holes, and the colour is pure optics. The surrounding reef is a couple of metres deep and bounces light back up. The hole does not bounce anything back at all. This particular one was properly surveyed in the middle of the 2010s with a remotely operated vehicle lowered on a tether, and the bottom came in

at just over 300 metres. For a while it held the record for the deepest blue hole known anywhere, and I should mention that the record has since been challenged by a hole off the coast of Mexico, which has been measured deeper, and whose bottom, at last report, had not actually been reached, not measured and found to be very deep, not reached. The instrument ran out before the hole did, which is a sentence that should make you slightly uncomfortable in the best possible way. The formation story is simple once you know that limestone dissolves in mildly acidic rainwater. During glacial periods, when the sea was far lower,

that reef platform stood in open air and rain worked its way down into the rock, opening voids and cave passages. Eventually the roof of one collapsed, leaving a vertical shaft. Then the ice melted, the sea returned and the shaft filled up. It is a drowned cave and its shape was inherited from a time when nobody could have swum into it because it was a hole in a field. Now the chemistry, which is where it becomes remarkable. The shaft is narrow and walled in by reef. so wind and waves cannot stir the water inside it.

Nothing mixes. An oxygen only enters seawater at the surface, from the air and from photosynthesis, after which it has to be physically carried down by circulation. Cut off the circulation and the oxygen simply stops arriving. Whatever lives down there uses up the small amount present, and then there is none, permanently. Measurements in that shaft show oxygen collapsing sharply below about 100 metres,

and beneath that the water is chemically dead as far as animals. animals are concerned. It does not look dangerous. It is clear and calm and utterly empty. Roughly the top third of the hole holds fish and normal reef life. The bottom two-thirds hold nothing that breathes. That combination is a preservation machine because decay is not primarily a chemical process. It is a biological one. Things rot because organisms eat them and the overwhelming majority of those organisms need oxygen. Take away the oxygen and you have removed the scavengers, burrowers, the wood eaters, and most of the bacteria that break tissue apart, anything that falls into the bottom of that shaft is not consumed. It settles and it waits. This is not a theoretical

property. Similar drowned caste systems elsewhere have produced spectacular results for exactly this reason. Sites in the Caribbean have yielded the remains of animals that tumbled into holes thousands of years ago and were preserved in the oxygen-free layer in conditions that would be impossible in ordinary sediment, including fragile material that normally does not survive at all. And in a flooded cave system in Mexico, divers found a nearly complete skeleton of a young woman who fell into a chamber more than 12,000 years ago, back when it was still dry, and who has since become one of the most important individuals in the story of how people first spread through the Americas. She is still with us because the water that eventually filled that cave

contained nothing hungry. So that is the archive end of the scale. Now the version that is actively hostile, which is the same chemistry, with the volume turned all the way up until it becomes lethal. On the floor of the Gulf of Mexico, there are lakes, genuine lakes, at the bottom of the sea, with a defined shoreline, a visible surface, and a boundary sharp enough that a submersible can hover just above it. Disturb one and ripples travel across the surface exactly as they would on a pond, which is an image that quietly rearranges your understanding of what a planet is permitted to do. They exist because of a leftover from the Jurassic. That whole basin was once a restricted sea that repeatedly evaporated,

leaving behind gigantic deposits of salt, which then got buried under kilometres of river sediment. Salt has an inconvenient property for geologists and a wonderful one for storytellers. It is less dense than the rock piling on top of it, and under pressure it flows slowly, like extremely unmotivated toothpaste. Over millions of years it rises in columns and sheets, pushing up through the overlying rock, and wherever it reaches the seafloor and meets seawater, it dissolves. The result is brine, water so loaded with dissolved salt

that it can be several times saltier than the ocean surrounding it. Dissolved salt makes water heavier, so this brine does not mix upward and disperse. It sinks and pools in whatever depression it can find, and the density contrast with the normal seawater above it is sharp enough that the boundary behaves like an actual surface. Then it gets worse, because these systems usually come with methane seeping up alongside them,

and hydrogen sulfide produced by bacteria in the sediment. So the pool is not simply very salty. It is oxygen-free, chemically aggressive, and carrying a gas that is poisonous to nearly everything that swims. Anything that blunders in dies quickly and then fails to decompose. Surveys of these features have documented dead crabs, fish and other animals lying around the margins and in the shallows of the pools, effectively pickled.

some of them still holding their final posture. One well-publicised example ended up with a nickname referencing a hot tub, which, combined with the footage, gave the internet precisely what it was looking for. But here is the part that redeems the whole arrangement, and it is genuinely magnificent. The rims of these pools are often crowded with life. Dense beds of muscles grow around the edges, and they are not there by accident, and they are not eating anything that drifted down from above. They carry symbiotic bacteria inside their own tissues which metabolise the methane, and the muscle gets fed in return.

Tube worms and other chemical-powered communities cluster nearby, running an entire functioning ecosystem on chemistry instead of sunlight. So the same seep that produces a lethal pool also produces an oasis around its edge, and you get a thriving, densely packed neighbourhood built directly on the shoreline of a lake that kills everything that enters it. Prime waterfront property, terrible swimming, and unsurprisingly no shortage of tenants. Now scale that principle up from a pool to an entire sea, because there is one place where oxygen-free water covers a basin large enough to see on a world map, and it has produced the best preserved shipwrecks ever found. The Black Sea is layered, and the layering is permanent. Enormous rivers pour fresh water in from the north and west, and fresh water is lighter, so it floats.

dense, salty water enters along the bottom from the Mediterranean through a narrow strait. The result is light water sitting on heavy water with a sharp boundary between them, and crucially, no seasonal turnover. Most seas mix vertically, at least occasionally, because surface water cools in winter, gets denser, sinks, and drags oxygen down with it. In the Black Sea, that process cannot punch through the density barrier, so below roughly 150 to 200 metres, the water contains no oxygen, whatsoever and does contain dissolved hydrogen sulfide. That oxygen-free zone makes up the overwhelming majority of the sea's volume, which makes it the largest body of anoxic water on earth, and it has been in that state for thousands of years. For maritime archaeology, this is

close to a miracle, and the reason comes down to one specific group of animals. In ordinary seawater, the great destroyer of a wooden wreck is not the current, and it is not the sediment. It is wood-boring mollusks, commonly called shipworms, which are not worms at all but by-valves that went into the demolition industry. They bore into submerged timber and eat it, and they are ferociously efficient. A wooden hull in warm oxygenated water can be reduced to a scatter of nails, pottery and ballast stones within years to decades. This is precisely why the standard mental image of an ancient wreck is a neat pile of amphorae with no ship attached. The cargo survives because it is fired clay. The vessel gets eaten. Shipworms need oxygen. Below the boundary layer in the

Black Sea they cannot survive, and neither can the burrowing animals that normally churn the seabed, nor the scavengers, nor most of the microbial pathways that take timber apart. What that means in practice was demonstrated by a large mapping project that ran in the middle of the 2010s. Its original purpose was not treasure at all. It was research into ancient sea level change and the drowning of prehistoric coastlines, which is a serious geological question in that basin. The team was running deep survey systems and remotely operated vehicles across huge areas of seabed, and along the way they kept tripping over wrecks, dozens of them. Ottoman vessels, Byzantine vessels, Roman period ships spread across two and a half thousand years of maritime activity,

in states of preservation with no precedent anywhere, and then, at around two kilometres of depth, they found the one that stopped everyone. It is a Greek merchant ship, a little over 20 metres long, and it is intact. Not a hull outline in the mud, not a debris scatter, the mast is standing, the rudders are in position, the rowing benches are still in place, the timbers hold their shape, it sits there looking less like an archaeological site and more like a vessel, that was carefully set down and then left in a hurry, which at some point it presumably was. Radiocarbon dating put it at roughly 2,400 years old, making it the oldest intact shipwreck ever documented.

And here is the detail that gives it real weight. We already knew that style of vessel, but only from art. There is a celebrated painted vase in a museum collection showing a ship of essentially this design, with the same rig and the same lines, an image reproduced in textbooks for a century and argued over by specialists precisely because no physical example existed to check it again.

Now one does. A painting has been confirmed by the object it depicted two and a half millennia later by a survey team that had gone out there to study coastlines. Which brings us to the flip side because if some water preserves everything, other water destroys everything, and there is a region whose reputation for eating ships turns out to be entirely deserved and entirely explicable. South of Japan, running down through the Isu and Bonin Island chains, there is a stretch of the Pacific that Japanese mariners have long treated with caution. and that Western writers in the 1970s enthusiastically rebranded as another mysterious triangle, complete with the claim that the Japanese government

had formally declared it a danger zone after losing a series of vessels and hundreds of lives. That claim spread widely and is still repeated with confidence. When researchers went back and checked it against the primary sources, the pattern was extremely familiar.

The lost vessels were mostly small fishing boats rather than military ships. several went missing outside the boundaries of the supposed triangle, which had been drawn generously enough to scoop them in. The casualty figures had inflated in transit. The whole structure was assembled the same way its Atlantic cousin was by copying earlier lists instead of verifying them. But underneath the invented version sits a real event,

and it is genuinely appalling, and it explains the reputation without needing a single supernatural ingredient. In September of 1952, a Japanese hydrographic survey vessel was sent to investigate a submarine volcano in that island arc. This was not sightseeing. The volcano was erupting, and it was doing something extraordinary, repeatedly constructing a new island above the waterline, and then having it blown apart or washed away over and over across a period of months. For a hydrographic office, whose entire professional purpose is telling mariners where the seabed is, A feature that keeps changing its mind about whether it is above or below sea level is an urgent problem.

Charts needed updating. Shipping needed warning. So a survey ship went out with 31 people aboard, including scientists, to observe an actively erupting volcano at close range. It was destroyed. Debris was recovered later, and the conclusion was that the vessel had been directly over the vent during an explosive eruption. Everyone aboard was lost. The physics is worse than an explosion in air. and it is worth understanding. A submarine eruption at shallow depth involves magma-meeting seawater, which flashes into steam instantly and violently, capable of hurling blocks of rock upward through the water column at enormous speed.

It can also do something even nastier, which is fill the surrounding water with gas. Waterful of bubbles is dramatically less dense than water, and a ship floats purely because of the density of the fluid it displaces. reduce that density enough, and the vessel simply stops floating, no matter how well built it is. Whether the ship was struck by ejected material, engulfed by the blast, or dropped straight through aerated water, no mystery is required at any point, and the region's reputation needs no help beyond that, because every ingredient is documented. It sits on one of the most volcanically active

arcs on the planet, with dozens of submarine volcanoes and sea mounts, some of which erupt with effectively no warning. A powerful warm current runs through it, it gets typhoons, and for centuries it was worked by small wooden fishing boats operating far offshore with no radio and no way to call for help, which is exactly the combination that produced grim disappearance statistics in every fishing region on Earth before the marine radio existed. So that stretch of ocean genuinely does take ships. It just does it with volcanoes, weather, and small boats, which is the least marketable explanation available and naturally the correct one. Now assemble the chapter, because there is one principle underneath all of it, and it is among the more important ideas

in this entire video. Everything we believe about the ocean's past is filtered through the question of where evidence was chemically capable of surviving. In normal oxygenated deep water, an event leaves almost nothing. A wooden vessel is consumed. A body is dealt with by the scavenging community with startling efficiency because the deep sea is chronically short of food and the arrival of a large carcass

is a windfall that gets processed at speed by hagfish, amphipods, crabs and sharks. Even the skeleton does not last indefinitely because there are specialised worms that colonise bone and break it down over the following years. Nothing is wasted down there, which is admirable, ecologically and catastrophic evidentially.

That is why the ships in the earlier chapters left nothing floating and nothing findable on the bottom. It was not strange. It was the default setting. So the map of what we have recovered from the sea is not a map of what happened at sea. It is a map of the handful of places where the water was chemically incapable of destroying the evidence. Oxygen-free basins, brine pools, sealed cast shafts and cold deep water are in effect

the ocean's hard drives. Everywhere else is a shredder that has been running continually. for as long as there has been anything to shred, which means the honest way to read the historical record of the sea is as a wildly biased sample. We have a beautiful, complete Greek merchant ship because it happened to sink into one of the very few environments on Earth where such a thing can persist.

There were thousands of others. They are gone, not because they mattered less, but because they sank into water with oxygen in it. And that same principle is about to make the final part of this video considerably more complicated, because if a large object in the deep ocean can be erased entirely within months, then the absence of a specimen tells you much less than you would like it to. It does not prove nothing was there.

It proves you are not standing in the right chemistry at the right moment holding a camera, which happens to be exactly the argument people make when they insist there are enormous animals down there that we have never documented. And the awkward thing, the thing that makes this last section genuinely difficult rather than a straightforward debunking, is that they are not entirely wrong. So we have to do this carefully, because this is the one category where the sensible position and the ridiculous position occasionally swap places without warning. Start with the ridiculous one and start with the most recent legend in this entire video, which is useful because we can watch

it being born rather than guessing at its origins. Somewhere around 2007, stories began circulating on Japanese internet forums about something seen in Antarctic waters. The descriptions were consistent in a way that is either impressive or suspicious depending on how you feel about internet forums. A vast pale figure in the water, tens of metres long, smooth and white, without scales or fur, with a body that appeared to have a head,

a torso, and structures like arms. Cruise on Japanese vessels working in the southern ocean were said to have seen it, sometimes at a distance, sometimes alarmingly close. It was given the name Ningen, from the Japanese word for a human being, and the whole thing came packaged with the usual accessory,

which is the claim that the government knew and had classified the material. Now run it through the checks we have been building all video, named witnesses, none. Not a single crew member has been identified, come forward or been interviewed by anyone. For an event supposedly witnessed by professional mariners on multiple vessels, that is an extraordinary silence because sailors are not famously reticent people, and a sighting like that ends careers of anonymity in about a week.

Vessel records. None. No ships log, no expedition report, no position, no date, no time. Every real maritime event in this video generated paperwork, because ships run on paperwork even when they are lying. Photographs. This is the fun part. The images circulating a uniformly low resolution poorly lit, and when traced back are variously computer-rendered, unrelated pictures of ordinary things, or so ambiguous that they could be a whale, a wave, or a stain on a lens. Not one has a verifiable provenance chain, and by now you know exactly why blurry is the only format this creature can exist in. Physical material, none. Not a tissue sample, not a stranding, not a bone. Institutional trace, none. No museum, no fisheries agency, no research institute,

no marine biologist anywhere has produced a single line about it in a professional context, so it fails everything comprehensively. But rather than stopping there, let us do the more interesting thing and ask why that particular ocean is so good at producing a story like this, because the conditions down there are close to an optimal laboratory for visual error. First, the ice. Antarctic waters are full of it in every size class, from tabula bergs the size of counties down to growlers, which are chunks of ice barely breaking the surface, and which are genuinely

dangerous to shipping precisely because they are hard to see. And ice does not weather into geometric blocks in the water. It melts and erodes into smooth, rounded, organic shapes, with lobes and bulges and glossy blue-white surfaces, and it rolls unpredictably as it loses mass, so a burgh can suddenly present a face that has been underwater for weeks, wet and gleaming and smooth as skin. Ice in the water looks alive, Ask anyone who has stood a watch in it. Second, the light. High latitudes give you extended twilight, low sun angles, flat grey overcast,

and a phenomenon where sea and sky lose their boundary entirely. So the horizon disappears and everything sits in a featureless white bowl with no depth cues and no scale reference. Add fog, add sea smoke rising off relatively warm water into freezing air, and you have an environment specifically engineered to defeat human visual judgment. Polar aviation has an entire vocabulary for the ways this environment lies to you. Third, the animals. Southern Ocean whales are enormous.

They surface unpredictably, and they show you a fraction of themselves for a couple of seconds at a time. A pectoral fin, rising and rolling looks disturbingly like an arm. The blunt rounded head of a sperm whale breaking the surface at distance in poor light is a large pale mass with no obvious animal features on it. And pale individuals genuinely exist, because lucidious. and albino whales turn up across multiple species, including a white humpback in Australian waters famous enough to have a name and a fan following. A partly white whale, at the edge of visibility, in fog, in flat light, glimpsed for three seconds by a tired person on a cold deck,

is a perfect generator of an unexplained sighting. Put those together and you do not need a hoax, a conspiracy or a monster. You need one crew member with a story, an internet forum, and a species that is compulsively good at seeing bodies in ambiguous shapes. Notice the specific shape of the claim too. Of all the forms life could take in the Southern Ocean, the legend settled on a giant version of us. That is not a fact about Antarctica. That is a fact about human beings,

who have been putting our own outline onto unfamiliar things for as long as we have had stories. Now the second case, which is the opposite in every respect, because here there was a real object, real photographs, real measurements and real measurements and real physical samples, and the answer was still wildly misread by an entire country for a while. In April of 1977, a Japanese trawler was fishing off the coast of New Zealand, east of Christchurch, working nets at a few hundred metres. When one neck came up, it contained something rotting and

enormous. The crew estimated it at around 10 metres long and close to two tonnes, and the shape of it was the problem. It had a long slender neck well over a metre, ending in a small head with no obvious features left. It had four large flipper-like appendages, it had a tail, and it had almost no smell of fish, which the crew noted at the time. If you have ever seen a drawing of a plesiosaur, you already understand what happened next. This looked exactly like a pleasiosaur, a marine reptile that has been extinct for around 66 million years, hanging in a net off New Zealand, dripping. An official from the fishing company aboard the vessel did what a sensible person would do. He took photographs, five of them. He measured the carcass carefully and made a detailed sketch with dimensions,

and he cut samples of tough fibrous material from one of the front flippers. Then the captain ordered the thing thrown back overboard. I want to be fair to that captain, because he gets abused for this decision, and he was completely right. He had a hold full of mackerel and a rotting two-ton carcass leaking fluid all over his deck and his catch. The commercial value of that catch was real, immediate and belonged to his employer. The scientific value of the carcass was speculative and belonged to nobody in particular. He made the correct operational call

and in doing so he became the most cursed man in the history of cryptozoology, which seems like a rough trade for a clean deck. What followed in Japan was a national event. The story broke and the country went briefly and enthusiastically pleasaur-shaped. It was front-page material. There were toys.

There was a commemorative postage stamp. The fishing company's share price reportedly benefited, which is my favourite detail because it means somebody, somewhere, made money on the stock market because of a dead shark, and we will get to why it was a dead shark in a moment. The samples were the entire case, and this is why the story belongs in a video about method rather than monsters. Those tough fibres from the flipper were analysed, and their protein composition was compared against known reference material. The amino acid profile matched a structural protein found in the fin rays of shark. Not reptile, not mammal, shark. And once you have shark, the whole picture snaps into focus, because there is a specific, well-documented, entirely predictable way that a large shark decomposes, and it produces this exact

result almost every time. The species involved is the basking shark, the second largest fish alive, a filter feeder that reaches around 10 metres and swims around the temperate oceans with its mouth open. Its skeleton is cartilage, not bone, and the way that skeleton is put together is the key. The enormous jaw and the huge gill arch structure, which together make up the vast bulk of that famous open-mouthed head, are attached relatively loosely. When the animal dies and decay sets in, those parts drop away first, all of it, the jaws, the gill baskets, the whole apparatus. What is left is a small brain case perched on the front of an exposed spinal column,

sticking out ahead of the body. In other words, a long, thin neck ending in a small head. Meanwhile, the dorsal fin, being soft tissue on the top, rots and disappears, and the upper lobe of the tail can go too, leaving a stumpy tail shape.

But the paired pectoral and pelvic fins, supported by those tough fibres, survive much longer. Four appendages arranged front and back, so a decomposing basking shark does not merely resemble a plesiosaur. It converts itself into one, by a purely mechanical process with no imagination required at all.

This is so consistent that the same misidentification has been happening for centuries. A famous carcass washed up in the Orkney Islands in the early 19th century and was described to a scientific society as an unknown serpentine animal and it was the same thing, doing the same trick 200 years earlier. There is a broader lesson here about mystery carcasses in general and it applies to the whole category of shapeless masses that wash up on beaches and briefly convince everyone that something impossible has died. Several of the most famous ones, huge pale, rubbery blobs with no discernible features

have been analysed with modern techniques, and the answer keeps coming back the same. They are collagen. Specifically, they are the tough fibrous matrix from the blubber layer of large whales which holds together as a mass long after everything else has been stripped away and can drift for a very long time. A featureless blob on a beach is not a new species. It is a whale's structural underwear. The unifying principle is that decomposition is not random. It follows a sequence dictated by anatomy, and if you know the sequence you can read backward from the ruin to the original animal.

A monster carcass is not a zoology problem. It is a decay problem, and it has been solved. Which leaves the last section of this video, and I have been looking forward to it, Because after nine chapters of taking things apart, it is time to hand over the genuine article. Here is the honest situation. Everything we just dismantled was fake or misread, and yet the underlying intuition, that there are large animals in the deep ocean that we barely know, is completely correct, and the real examples are stranger than the invented ones by a wide margin. Start with a fact that sounds like a joke.

There is an animal in the deep sea that is structurally a woodlouse, the same basic body plan as the little grey creature that scuttles out when you turn over a plant pot in the garden. Segmented armour, seven pairs of legs, the ability to curl up. Its shallow water relatives are a couple of centimetres long and are famous for being harmless and slightly boring. The deep sea version reaches something like 40 centimetres, and there are records of individuals approaching half a metre. It is a woodloose the size of a small dog, it walks on the abyssal floor and it turns up in numbers around a bait trap looking like a scene from a film nobody wants to watch during dinner. Photographs of them being held by researchers are widely disbelieved on first viewing,

and I encourage you to look one up and then come back, ideally not immediately after eating. That size difference is a real, systematic phenomenon called deep sea gigantism, and it has several overlapping explanations, none of which involve anything exotic. The first is temperature, the deep ocean is cold, uniformly everywhere, a couple of degrees above freezing. Cold slows metabolism dramatically. A slow metabolism means a slow burn rate, which means an animal can afford to be large in a place where food is scarce, and it also tends to mean slower growth, later maturity, and a much longer life. Many deep-sea animals simply keep growing for decades, and if you keep growing for long enough, you get big by

default. There are deep water species with lifespans measured in centuries. The second is oxygen. Cold water holds more dissolved oxygen than warm water, which sounds like a footnote and is actually a major constraint on body size for animals that do not have efficient circulatory systems for their size. Give an invertebrate colder, more oxygenated water, and you have relaxed one of the main breaks on how large it can grow. The third is food supply, and it works in the least obvious direction. The deep sea is not merely short of food, it is unpredictably short of food, with long stretches of nothing. punctuated by rare windfalls when something large dies and sinks. That environment does not reward being nimble and small. It rewards being able to gorge enormously when the opportunity arrives and then survive on the reserves for an extremely long time. A larger body is a bigger fuel tank.

There is a well-known case of one of these giant isopods in an aquarium refusing food for years, going about half a decade without eating anything at all and remaining by all appearances entirely fine, which is a survival strategy I find personally aspirational, and the fourth is that a great deal of the pressure that keeps animals small in shallow water in the form of fast visual predators is simply weaker down there. So gigantism is real, it is explicable, and it is not remotely mysterious. Which brings us to the animal that closes this entire video, because it demonstrates every single theme we have covered in one specimen. There is a squid in the southern ocean called the colossal squid. It is not the same animal as the giant squid, which is a different genus with a different body plan,

and the colossal one is the heavier of the two, with a shorter but far bulkier body, potentially reaching several hundred kilograms. Its eyes are around 30 centimetres across, which are the largest eyes of any animal known to have existed, and its arms carry rotating hooks rather than just suckers, which is not a design feature that inspires confidence. It was formally described in the 1920s, and here is the crucial part. the part that ties this to everything else in this video. It was described from two arm crowns found in the stomach of a sperm whale.

That is how we know most of what we know about deep sea cephalopods. Sperm whales dive to enormous depths and eats squid in industrial quantities, and squid beaks are made of a tough material that resist digestion, so they accumulate in the whale's stomach. Which means that for the better part of a century, the most productive deep sea sampling program on the planet has been sperm whales, working for free, on their own schedule, with no funding applications and no ship time to book. We have been reading the ocean's contents by going through the pockets of its largest predator.

For a hundred years after that description, nobody photographed one alive in its own habitat. Specimens turned up occasionally, hauled up dying on fishing lines or found in whale stomachs, and everyone was a new story. But the living animal, in the deep water where it belongs, doing whatever it does was simply absent from the record. A 500-kilogram predator with the largest eyes in the history of life, and we had never once seen it at home. That changed only very recently, when an expedition operating a remotely operated vehicle in the Southern Atlantic

filmed a young one at several hundred metres, a small transparent juvenile drifting in the dark, roughly a century after science first gave the species a name. The individual was a fraction of the adult size, which somehow makes it better because the first footage of one of the largest invertebrates ever to exist is of a baby minding its own business. And the reason it took 100 years is the most instructive thing in this chapter, because it was not depth and it was not luck. It was our own equipment. The standard way to look at the deep ocean is to send down a machine with powerful

floodlights and multiple thrusters. That machine is loud, it pushes water around, and it blasts an environment that has been in permanent darkness for its entire higher evolutionary history with the equivalent of stadium lighting. Every animal down there with any sense leaves immediately, and the animals with the largest eyes and the best sensitivity to light and vibration leave first and fastest. We built a survey tool that specifically repels the things we most wanted to survey, then spent decades concluding they were rare. The fix, when it came, was to stop doing that. Researchers built quiet camera platforms that drift rather than thrust, using far-red light which most deep-sea animals cannot see, and in some cases

a lure that imitates the bioluminescent distress display of a jellyfish, which in the deep sea functions as an advertisement that something is being eaten nearby and is therefore worth investigating. Using that approach, teams have filmed the giant squid in open water, including in the Gulf of Mexico, after more than a century of failing to do so with brighter and louder machines. So the answer to where the monsters were, is not that they were hiding in some unreachable trench. They were right there. In water we had been visiting for decades, avoiding us competently, because we kept arriving with the underwater equivalent of a helicopter and a searchlight, and wondering why nothing came out, which is, I think, the right place to end because it reframes the whole video. Look back at what

the ocean has actually been doing across these ten chapters. It preserved a drowned country under the North Sea and a Bronze Age street plan in four metres of water, and a Greek merchant ship with its mass still standing, and it handed those back to us once we built instruments capable of asking. It took ships and crews with total efficiency, not through malice, but because a shifting cargo, a rogue wave or a collapsing hull operates faster than any human response, and because the deep sea removed evidence as a matter of routine metabolism, and it faked constantly, generating right angles in bedrock, staircases in sandstone, antennas out of sponges, spacecraft out of glacial debris and monster cools out of breaking ice,

not because it was trying to fool anyone, but because we are a species that cannot look at an ambiguous shape without finishing the sentence. Every case we solved was solved the same way. Somebody went back with a better instrument, or checked the actual sediment, or read a book from the 1880s, or counted the barrels,

or pulled the samples, or built a reference library of what known things sound like and did the comparison. Nobody solved anything by staring harder at a bad, photograph. And every case that remains open remains open for one of two reasons, both boring. Either the evidence was chemically destroyed before anyone could reach it, or nobody has yet paid for the ship time to go and look properly. Which loops us back to the number this whole thing started with. The overwhelming majority of the seafloor still has not been mapped at high resolution,

which means the honest summary of the state of ocean mystery is not that the sea is hiding things from us. It is that we have surveyed a small fraction of it, mostly the part of the part of the near our own coastlines, mostly recently, mostly on a budget. The unexamined remainder is not full of ruins and monsters. It is full of ordinary seafloor and somewhere in it, statistically, are a small number of genuinely extraordinary things that will look completely obvious the moment somebody finally points the right machine at them. The century of ocean discovery is not behind us.

It has barely been funded yet. If one of these files could be reopened tomorrow with unlimited ship time ship time and the best equipment we can build, which one would you send it after? The drowned plain under the North Sea, the wreck nobody has ever found in the mud of the strait, or a quiet camera drifting in the dark waiting for something with very large eyes to come and see what the light is. Tell me which, and tell me why, and I will see you in the For 50 years, Canadians have wondered what the names of IKEA furniture mean, like Natopa, the guardrail for children's beds. Natopa means giving your favourite person.

room to grow. It means their safety and your piece of... No. In Swedish, Natapa means night monkey. Mark!

IKEA, 50 years of Swedishness at home. Look, he's wearing little pajamas. So cute. Next one.