As the sun rose on July 2nd, 1937, Amelia Earhart knew she was in trouble. Over the radio, she called, "We must be on you but cannot see you. Gas is running low. Been unable to reach you by radio. We are flying at 1,000 feet." Beneath her was water in every direction as far as the eye could see.

She got herself into this predicament through a series of unfortunate events and bad decisions. Many of them could have been avoided with a better knowledge of physics. But even so, there was one thing she could have done in this moment, one switch she could have flipped that would likely have saved her life and changed history.

Amelia Earhart was vying to become the first female pilot to fly around the world. "I hope to accomplish something really scientifically worthwhile for aviation." And she wasn't taking any shortcuts. Other successful circumnavigations had followed a northern route, mostly staying close to land. But Earhart's route would be the longest by following a path close to the Equator.

This meant the last part of her journey was the hardest, crossing the full width of the Pacific Ocean. The starting point for this crossing was Lae, a city on the eastern side of New Guinea. At the time, it was one of the world's busiest airports, a hub of traffic from Asia and Australia. At 10:00 a.m. on a hot July day, Earhart piloted her Lockheed Electra down the runway and took off on what would be her final flight.

The Pacific Ocean is huge. It's way bigger than the Atlantic. I mean, if you look at the globe from that side, you see almost no land. The problem was, in 1937, most planes could only fly a maximum of a few thousand kilometers, so Earhart removed everything unnecessary from her plane. She ripped out the insulation to reduce weight, but that made the engine noise so overwhelming she had to communicate with her navigator sitting right beside her using written notes.

She packed almost nothing, telling her husband, "Extra clothes and extra food would have been extra weight and extra worry." She replaced the passenger seats with fuel tanks, effectively turning her plane into a flying gas can. But even so, the Electra's maximum range was between 6,600 and 7,200 kilometers in perfect weather. It could be just enough to reach Hawaii from Lae, or she might come up disastrously short. So Earhart needed a place to stop and refuel along the way.

Now, it might seem like there's no land here, but if you zoom in, there is this tiny island halfway between Australia and Hawaii. Howland Island is just over two kilometers long and less than one kilometer wide. The US had claimed it as part of the Guano Islands Act of 1856. But in 1937, it was barely inhabited with just a handful of colonists. It would be an ideal location to refuel, if only it had a runway.

Fortunately for Earhart, by the time of her around the world flight, she was already famous. In 1928, she became the first female passenger to cross the Atlantic by airplane. This made her an international celebrity. But she wanted to fly herself, saying, "Maybe someday I'll try it alone."

So, in 1932, she attempted to pilot a plane solo across the Atlantic heading for Paris. She brought with her only a toothbrush, one container of soup and three cans of tomato juice. But storms, ice, and dense fog battered her small plane. A seam in the exhaust manifold cracked and flames from the engine spewed out into the night. Gas leaked down her neck from a broken tank, and after 14 hours, she landed in a pasture in Northern Ireland. Her face was so covered in grease, a farmhand couldn't tell if she was a man or a woman. He asked if she had flown far. "From America," she replied.

These adventures brought her into the orbits of powerful people, like the First Lady, Eleanor Roosevelt. And using her new connections, she lobbied the president to hire her friend, Eugene Vidal, to head the Bureau of Commerce. Vidal had promised Earhart a runway on Howland Island, but red tape stalled progress only months before her planned takeoff.

So Earhart wrote directly to President Roosevelt. She explained that the airstrip funds required immediate approval, writing, "Please forgive troublesome female flyer for whom this Howland Island project is key to world flight attempt." The president approved the project four days later, and three runways were soon cleared.

So she had a place to land, but how would she find this tiny speck of an island in a vast ocean? Well, flying with her in the Electra was her navigator, Fred Noonan, and he would calculate the flight plan. They knew the direction of Howland, so they could use the onboard compass to set their bearing toward it. They knew their air speed and could figure out their ground speed by subtracting or adding the wind, and then they could calculate how long it should take to reach the island. This method is known as dead reckoning.

But they wouldn't aim directly at the island, because if they did that and they didn't see it at the prescribed time, they wouldn't know in which direction they were off. So instead, they intentionally picked a point either north or south of the island. Let's say they picked south. They estimated the trip would take 18 hours, so they would fly through day and night. And once they had traveled for the calculated length of time, they could confidently turn north and spot the island.

Before takeoff, the ground crew estimated they would encounter a headwind of 24 kilometers per hour. But just 20 minutes after takeoff, Lae radioed Earhart to warn that the headwinds would be stronger. She didn't acknowledge their message. Knowing the correct wind speed was critical because it would affect how long it would take to reach the island. If it took longer, Earhart would have to turn later. So she couldn't rely on dead reckoning alone to reach Howland.

As an independent check on their location, Noonan would take measurements of the sun, moon, and stars. This is known as celestial navigation. He had an almanac that listed 58 navigation stars and the point on Earth each one would be directly overhead for the day and time of his measurement. If they found themselves directly under a navigation star, well, then they would immediately know their position.

But generally they would not be that lucky, so Noonan would measure the angle above the horizon to a navigation star and use that to work out how far away they were from the point on the Earth where that star would be directly overhead. So he could trace out a circle on the globe of possible locations, and then he would measure the angle to another navigation star and draw out a second circle. And now they must be at one of these two circle intersections. Normally the circles were so large that only one of the intersections would be a plausible position. That way they could continually update their location and adjust bearings as needed.

But even with celestial navigation, errors could accumulate over long trips. Earlier in the journey, when Earhart crossed the Atlantic, they missed their intended airport in hazy conditions. Noonan's calculations were reasonable, but small errors put them off course. Luckily, in Africa, there were plenty of other places to land safely. The same could not be said for Howland.

So, for the flight across the Pacific, Earhart commandeered three US Navy and Coast Guard ships. The Itasca would be stationed at Howland Island, the Ontario would be halfway along the route, and the Swan was positioned midway between Howland and Hawaii. The Itasca would send out smoke signals as Earhart approached to help her spot the island. But even more importantly, all ships were equipped with radio.

Now, in 1937, radio was still fairly new tech. German physicist Heinrich Hertz discovered radio waves in the late 1880s. He excited electrons to oscillate back and forth in his transmitter, and a few meters away his receiver was a loop of wire with a small gap in it. When Hertz looked at it through a microscope in the dark, he saw faint sparks jumping across the gap. The sparks were strongest when the receiving loop was flat. If it was vertical, then no sparks were observed.

This demonstrated that radio waves are transverse waves with electric and magnetic fields oscillating perpendicular to each other and perpendicular to the direction of the wave motion. When the receiving loop was aligned with the direction the wave was traveling, the changing magnetic field through the loop induced an EMF that created the spark. But if the loop was facing the transmitter, then there was no change in magnetic flux through the loop, and so no spark was observed.

Now, Hertz couldn't see the future he had ushered in. He said, "I do not think that the wireless waves I have discovered will have any practical application." But within a few years, people started sending messages using radio. And by the 1920s, radio entertainment broadcast took off. Ships and planes routinely used radio to send Morse code, and some, including Earhart, could send and receive voice messages.

In fact, Earhart had five radio antennas around the plane, each for a specific purpose. The largest antenna could be reeled in and out like a fishing line behind the plane. It was 76 meters long, which was necessary to efficiently send and receive Morse code via the 4 or 500 kilohertz radio waves used by ships and remote stations. Ideally, an antenna should be at least 1/4 of the wavelength of the radio wave it's transmitting or receiving. This improves the efficiency of the conversion from electrical energy to radiated electromagnetic energy.

Earhart's trailing antenna was only around 1/8 of the wavelength. But it was connected to a high-power transmitter, so its signals could still be detected over 1,000 kilometers away. Next were two antennas for voice communications on higher frequencies. A transmitting V antenna on the roof of the plane and a receiving antenna along its belly.

Higher frequencies were useful for two reasons. First, they require smaller antennas, which save weight and can be better accommodated on small sparse planes. And second, high frequency radio waves can travel long distances by bouncing off a layer of the atmosphere called the ionosphere.

Starting about 50 kilometers above Earth's surface, radiation from the Sun splits electrons off molecules forming a layer of ions and free electrons. Radio waves with certain frequencies interact with these free electrons and are effectively reflected back to Earth. It's as if they've bounced off a big wobbly mirror in the sky. This effect is called skipping and it scatters radio waves all over the place. These radio waves can then reflect off the ocean and back off the ionosphere, making multiple hops to travel thousands of kilometers.

During the daytime, the intense radiation from the sun means the ionosphere starts lower in denser atmosphere. And because of this, lower frequency radio waves are more likely to be absorbed than reflected. So aviators would typically use the higher 6210 kilohertz to skip their signals during the day, and then the lower 3105 kilohertz at night once the bottom of the ionosphere had lifted into thinner air.

Four hours after takeoff, Earhart radioed an update to Lae on her daytime frequency of 6210. She reported her altitude at 7,000 feet and speed at 140 knots before concluding with her typical sign-off, "Everything okay." But she never acknowledged calls from Lae about the headwind. They radioed again at 11:20 and 12:20, but never got a response from Earhart.

In all likelihood, she never heard them. She did radio six hours into her flight to report stronger headwinds, but she makes no mention of Lae's earlier warnings. It's possible the receiving belly antenna was broken, fell off, or something in the receiving electronics wasn't working. But her ability to receive voice messages was clearly impaired.

Nine hours into the flight, Earhart expected to come upon the Ontario. She listened for Morse code Ns on 400 kilohertz, but she heard nothing. The original plan was that the Ontario would wait for her to radio them to request that they start transmitting. But the day before takeoff, Earhart realized she had made a mistake. The Ontario had told her they wouldn't be able to receive any high frequency signals, which meant no voice communication.

So she sent an urgent telegram asking the Ontario to transmit the Morse code Ns repeatedly 10 minutes after each hour. The purpose of the Morse code from the Ontario was actually to allow Amelia Earhart to make use of her two final antennas. So she had a loop antenna just like this one and a sense antenna. These were designed to allow her to locate the source of radio waves. This was the final and most critical way that Earhart planned to stay on course and locate Howland Island. She wrote, "I doubt if I'd try the flight to tiny Howland Island without it supplementing Fred Noonan's skill."

So the Ontario was sending out Morse code signals on their antenna, and here we have a transmitter tuned to about 3.6 megahertz. I'm gonna put on this blindfold and use the loop antenna to try to locate the transmitter. And because I already know where the transmitter is, we'll spin me around a few times to really disorient me. So the radio waves are gonna be emitted in all directions radially away from the antenna, the electric field will be oscillating up and down, and the magnetic field will be oscillating back and forth. So if I hold up this loop like this sort of parallel to the direction that the waves are traveling, then the magnetic field is gonna be changing through the loop. And because of that, it's going to create an EMF and current and I can pick that up because I'm tuned to the right frequency here.

So I got a fairly strong signal. Woo, it's very strong. But if I rotate the loop like that, well, now the magnetic field is oscillating back and forth but not changing through the loop itself because it's parallel to the loop. And so, in this orientation, I'm gonna get a null reading. If I turn it this way, there's a null. But if you turn it 90 degrees, now all the magnetic field is passing through this loop and so I can hear a maximum here.

So this is what Earhart wanted to measure using her loop antenna to detect the repeated N Morse code from the Ontario. She would turn it until she found the null and then she would know the direction to the ship. Something that's interesting is, if I turn it away, we get another null because again there's no magnetic flux passing through this loop.

Now, the first time she picked up the signal, she would probably be heading straight towards the ship or close enough, so she would know that it's roughly that way. But there's a chance that she's gone past it. And if you go past it, well, then you also get a null, but the ship is behind you, not in front of you. So that's where the sense antenna comes in. The sense antenna gives you a cardioid pickup pattern so it has a single null instead of two nulls, and so that allows you to determine whether it is in front or behind you.

With the sense antenna, the only null points directly away from the transmitter, so it's easy to use the sense antenna to check which null is correct but then only use the loop when navigating because it gives a sharper null. Now, aviators could have used where the signal is loudest and try to go in that direction, but it's actually easier to get a precise null, a point where the signal drops out. The loud section could range for quite a distance and so you wouldn't really know where it is, but the null is more precise so that's why they would look for the point where the signal drops out.

If Earhart could hone in on the Ontario using her radio direction finding loop, that would ensure she was on course and eliminate any navigation errors that may have occurred to that point. But her telegram asking the Ontario to transmit 10 minutes after each hour didn't make it to the ship in time. And since Earhart couldn't talk to the Ontario, they never sent out any signals. So they passed like ships in the night.

By this point, Earhart was around halfway to Howland. With no other landing strips within 1,000 kilometers, she would have to find the tiny island or return to Lae now. But multiple delays had already plagued her journey. In fact, this was not Earhart's first attempt to fly around the world.

Earlier that year, in March of 1937, she had taken off from California for Hawaii, heading west instead of east. On board were Fred Noonan and another crew member, Harry Manning. As a Merchant Marine captain, he was an expert in radio, Morse code, and traditional navigation. He was also a pilot. The flight to Hawaii was successful thanks in part to Manning using the loop antenna to hone in on a radio beacon on the destination island.

Three days later, the trio set off for Howland Island. But just as they were taking off, the plane drifted to the right. Earhart corrected by throttling back the left motor, but it was too much. The plane turned to the left and the right wing dipped down. Going up on one wheel, the right landing gear collapsed, then the left. The plane skidded out on its belly, spinning around to face the way it had come. Thankfully, no one was hurt, but the Electra took months to repair.

And during that time, the seasonal winds shifted. So, on her next attempt, Earhart would have to fly east instead of west. And most importantly, Captain Manning left the crew. Officially, the press reported that he needed to return to the Merchant Marines, but rumors spread that he had lost confidence in Earhart, or that Earhart believed Noonan was a better navigator than Manning and she could operate the radio on her own.

Whatever the case, when Earhart took off again three months later, she was accompanied only by Noonan. And now they had made it 80% of the way around the world. And in the dark of night, Earhart had to make the critical decision whether to keep going or turn back. The lack of signal from the Ontario must have been concerning, but maybe they never got her telegram. And she knew that at Howland, the Itasca would be transmitting the letter A over Morse code every half hour, even if they didn't hear from her. And they could send and receive voice signals. They promised to be ready on a range of different frequencies. So she flew on.

Around 6:15 a.m. local time, radiomen aboard the Itasca heard Earhart clearly. "Please take a bearing on 3105. We'll whistle into the mic. We are about 200 miles out." She then began to whistle. But the men were confused. They expected Earhart to take a bearing on them, not the other way around. And while they had told her that they had radio direction finding equipment, the signal needed to be lower frequency, between 270 and 550 kilohertz. Her voice frequency would skip off the ionosphere and reflect off the ocean, scattering in all directions. So there would be no way to find a null because the signal would be coming literally from everywhere.

In the Electra, Earhart heard only static. By now, she must have been worried that they hadn't heard anything from either ship. Almost blind from the rising sun and deaf from the roar of the engines, Earhart twisted the radio dial, listening for Itasca's response. Nothing. She may have expected Howland to have a high frequency radio direction finder called an Adcock antenna array. These systems solve the skipping problem with five vertical antennas at the corners and center of a square. The direction of the radio wave can be calculated from the slightly different arrival times and signal strengths at each antenna. But these antennas were massive, so they were really only installed at larger airports.

Now, as it happens, there was a portable high frequency radio direction finder on Howland Island, but the operator reported that Earhart's transmissions didn't last long enough for him to take a bearing. And trying to conserve his low battery, he missed parts of the later transmissions. Around 6:45, Earhart again asked them to take a bearing on 3105 kilohertz and report back in a half hour. But a bearing taken now and reported back in a half hour would be at best outdated and at worst misleading.

This confusion likely had to do with time zones. Earhart was using Greenwich Civil Time, but the Itasca set their clocks to their current position which was GCT -11.5 hours. And to make matters worse, Howland Island used Hawaii Time, which back in those days was GCT -10.5 hours. So the three parties attempting a rendezvous on a tiny island in the middle of the Pacific were on three different time zones.

And crucially, Earhart's hours didn't even line up with the others. Earhart told the Itasca she would be using GCT, but somehow it never made it to the radiomen. So, when the Itasca heard Earhart's request, it was 6:45 a.m. But in the cockpit, it was 6:15 p.m. So Earhart likely didn't say "in a half hour" but "on the half hour," which for her was only 15 minutes away. And also it was a prearranged time that Earhart would be listening for them.

Earhart was careful to set times she would transmit and times she would listen for the ship because she could only power one antenna at a time. And the ships used the same antenna for receiving and transmitting, so if they both broadcasted at the same time, they would miss each other's messages. If Earhart sent another message at a quarter after the hour, the Itasca blocked it with their own message. "Cannot take a bearing on 3105 very good. Please send on 500, or do you wish to take a bearing on us? Go ahead please." There was no response.

But she couldn't transmit on 500 kilohertz anyway because she had removed the long trailing antenna that could transmit lower frequencies. Since it could only be used for Morse code, something neither she nor Noonan were particularly well versed in, she saw it as dead weight after Manning left. So, after the Hawaii crash, it was removed during repairs. So she had no way of sending radio waves that would allow the Itasca to take a bearing on her. But she could take a bearing on the Itasca using her loop antenna, if they sent her the right frequency.

Before the trip, the Itasca had asked Earhart to specify the frequency they should broadcast. Earhart was unsure, so she consulted a radio expert in Lae and they recommended the Itasca send Morse code A, just repeated dot dashes, on the half hour at 750. But at that time it was typical to talk about radio waves using their wavelength, so the expert had meant 750 meters or 400 kilohertz. But Earhart made a terrible mistake relaying this plan to the Itasca. She requested the signal be sent on 7,500 kilohertz instead of 750 meters or 400 kilohertz. But she did explicitly say, "If frequencies mentioned unsuitable, inform me." But no one ever corrected her.

At 7:42 a.m., Earhart's voice came through so loud men even went above deck to see if they could hear a motor or spot the plane. She said, "We must be on you but cannot see you. But gas is running low. Been unable to reach you by radio. We are flying at 1,000 feet." On Howland, the high frequency radio direction finder was so low on battery the radiomen didn't even hear Earhart's message, much less take a bearing on it.

10 minutes later, Earhart said, "We are circling but cannot hear you. Go ahead on 7,500." The Itasca immediately sent As on 7,500 kilohertz. In the Electra, Earhart heard the stutter stop of As filling the cabin. The relief of finally hearing something must have been overwhelming. She quickly turned her radio direction finding loop to locate the null, but the signal never dropped out. The frequency was too high, so the radio waves from the Itasca were reflecting and arriving from different directions.

Joseph Gurr, a radio mechanic who worked on Earhart's plane, later said that they knew there were limitations to high frequencies which had a tendency to skip and bend, creating a false radio direction bearing. Without a minimum, she was still lost. Earhart frantically called Itasca. "We received your signals but unable to get a minimum. Please take a bearing on us and answer with voice." Itasca attempted to explain the problem. "Your signals received okay. It is impractical to take a bearing on your voice." No response. Without the belly antenna, she probably never heard any of their communications.

And it wouldn't have mattered if the Itasca had sent low frequency signals because Earhart's loop was tuned to pick up 7,500 kilohertz. So, why didn't the Itasca correct the frequency she suggested? Commander Thompson of the Itasca was aware of her radio direction finding limits. He had received messages both from Earhart's husband, George Putnam, and the Coast Guard's San Francisco division stating Earhart could only take bearings on frequencies between 200 and 1,500 kilohertz. But he either thought Earhart knew more about her radio equipment, or that it wasn't his place to make suggestions and take more responsibility for her flight.

When she asked the Itasca to tell her if these frequencies weren't suitable, she could have been referring to the ship's capabilities rather than her own. The Itasca said they'd be ready on the frequencies she wanted and more instead of giving specific suggestions. San Francisco's Coast Guard division tried to get Commander Thompson to take more responsibility for Earhart's radio communications by suggesting they directly tell Earhart which frequencies to use. But Thompson essentially told them to butt out. The Itasca communicated directly with Earhart from then on.

The radiomen continued to try to reach Earhart, and just before 9:00 a.m., Earhart's voice suddenly burst through again. "We are on the line 157-337. We will repeat this on 6,210 kilohertz. We are running on line north and south." Her voice was desperate. It sounded as if she was about to burst into tears or scream. This was the last message the Itasca heard.

There are a number of conspiracy theories about what happened to Earhart after that, but the evidence seems clear. She ran out of fuel somewhere over the Pacific and crashed into the sea. Two hours after her last message, the Itasca left Howland to search north and west for the Electra. Other Navy and Coast Guard ships and planes joined the search for over two weeks.

To that point in US history, it was the most intensive and expensive search and rescue operation, costing around $4 million, which is almost 100 million in today's money. No one has ever found a trace of Noonan, Earhart, or her Electra. All of these mistakes could have been resolved if Earhart had two-way communication, but her belly antenna somehow malfunctioned. Some theories suggest it fell off during takeoff in New Guinea, but without physical evidence, it's impossible to say.

But Earhart did confirm receiving signals on her loop antenna. Her loop could only direction find with lower frequencies, but it could receive signals on a wide range. If she had switched to using the loop for all communications, she could have received Itasca's voice messages and then the Itasca could have requested she take a bearing on a lower frequency, which would have guided her safely to Howland Island.

When I began researching this video, I expected to find that Amelia Earhart's demise was inevitable. That what she was trying to do was just so difficult that nothing could have saved her. But instead, I found the opposite. There were at least a half dozen things that if they went differently would have allowed her to land safely.

So to me, this story comes down to two things. Knowledge and responsibility. Earhart lacked knowledge of radio systems, which would've allowed her to specify the right direction finding frequency. But Commander Thompson of the Itasca had that knowledge. He knew her direction finding limits, but he didn't take on the responsibility to correct her. When attempting any challenging endeavor, you need someone with the right knowledge who will also take responsibility for getting things right. That's what you need to battle the inherent chaos and disorder of the universe. Otherwise, what you get is disaster.

This video was sponsored by KiwiCo, so I thought it would be appropriate to build their radio-controlled delivery bot. But my oldest kid was out for the afternoon, so I sat down to make it by myself, but very quickly I was joined by my 4-year-old and 6-year-old who were really excited to help. You know, we have been getting KiwiCo delivered to our door for years now and it's one of my kids' favorite things to do with me. They have a lot of fun while we're also spending quality time together. And somewhat secretly, they're also learning, not only technical knowledge, but also knowledge of how simple and understandable machines that at first appear complex can be. I think it's great that they're getting hands-on with wheels and motors and radio transmitters. I think it enables them to build up this resilience and an attitude that, well, I can understand that, I could probably build that. And KiwiCo takes responsibility for ensuring everything you need is right in the box, which means no extra trips to the store. When a crate turns up, you can sit down right away and get into it. They have something for kids of all ages and each crate is designed by experts and rigorously tested by kids to make sure it promotes learning and engagement and is just generally really fun to do. I noticed that their instructions keep getting better with helpful hints to ensure you've done everything correctly. And with KiwiCo Labs for ages six and up, kids use the problem-solving skills they've developed to engineer cool games, robots, or everyday products. So if you wanna try out KiwiCo, use my code, Veritasium, to get 50% off your first crate of a monthly club. And with the holidays around the corner, KiwiCo makes a great gift for families. Head to kiwico.com/veritasium to check out their great club options today. So I wanna thank KiwiCo for sponsoring this video and I wanna thank you for watching.