The following is an excerpt from Caesar’s Last Breath: Decoding the Secrets of the Air Around Us by Sam Kean.

In addition to using computers, meteorologists last century took advantage of new balloon technology to explore the workings of the atmosphere, especially its upper reaches. And just as with computers, the balloon projects led to several important insights into how air works— as well as, in one memorable case, to no small amount of embarrassment for the scientists on the ground.

It all started one morning in June 1947 when a ranch foreman named Mac Brazel came across a trail of metal and plastic debris after a thunderstorm. Now, Brazel had no intention of igniting a half century of hysteria and conspiracy theories; he just wanted to clean up the damn ranch. So rather than leave the scraps there and risk his sheep chewing on them, he gathered them up, tossed them into a shed, and tried to forget about them.

Except, the more he thought about it, the more the scraps bothered him. He worked on a ranch near a few military bases in New Mexico, and scientists around there were always launching missiles and weather balloons that came crashing back down on people’s land. He’d in fact found downed weather balloons twice before. But this time was different. The crash landing had gouged out deep grooves in the earth, which seemed impossible for a soft balloon. And the shards of plastic and metal didn’t seem like balloon material. Most unsettling of all, the wreckage included a few short wooden beams with purple squiggles on them, like writing— but writing in no earthly language he knew of.

Caesar's Last Breath: Decoding the Secrets of the Air Around Us

Buy

A few days later Brazel showed the scraps to his neighbors. They in turn told him some rumors they’d heard lately about unidentified flying objects near their lands. This spooked Brazel good, so he visited the local sheriff in Roswell, seventy-five miles away, on July 7. The sheriff in turn called up officials at a nearby army air force base.

When they arrived at the shed, the air force officials examined the scraps and tried to reconstruct the thing; they soon gave up, baffled. They also tried to cut the metal bits with Brazel’s knife and burn the scraps with matches, and failed on both accounts. They finally examined the purple squiggles, which they began referring to as “hieroglyphics.” At this point they decided to confiscate the whole mess.

Thanks to Brazel the local gossip mill had been churning for days by this point. But rather than keep mum, the air force issued a spectacularly boneheaded press release claiming that “rumors regarding the ‘flying saucer’ became a reality yesterday.” A newspaper story made similar claims. To be sure, phrases like “flying saucer” and “unidentified flying object” had more neutral meanings back then, but it didn’t take long for people’s imaginations to endow them with very specific meanings indeed.

All the scuttlebutt still probably would have died down, except that senior air force officials swooped in and demanded the retraction of the press release; one actually drove around to local newspapers and radio stations and snatched up paper copies. This got even skeptics thinking hard about conspiracy. What was the air force scared of? What were they hiding? People grew more suspicious still when the air force insisted that all the scraps had come from a weather balloon— obvious bull sugar. And in fact, we can now say for certain that the air force lied about this: that was no weather balloon Mac Brazel found. Unfortunately, what the military was lying about probably isn’t what you’re hoping for— unless you’re a spy buff with some pretty esoteric knowledge about the atmosphere.

The whole Roswell fiasco started with an earthling named Maurice Ewing, a geophysicist at Columbia University who did contract work for the military. Like every other red-blooded American then, Ewing dreaded the prospect of the Soviet Union acquiring the Bomb. But in those days before satellites and fallout detectors, we had no idea what the Soviets were up to. So he started thinking about other ways to spy on the Reds. He finally hit upon a way to eavesdrop on atomic blasts from afar by suspending microphones in a region of our atmosphere called the sound channel, which is located roughly nine miles up in the sky.

To understand Ewing’s idea, you have to know three things about sound. First, sound moves faster in warm air than in cold air. That’s because sound depends on molecules knocking into one another. It’s pretty slapstick, actually. When someone speaks, the air molecules leaving her mouth crash into nearby molecules of air. These careen into a second layer of molecules, which blunder into a third, and so on, until the noise stumbles into your ear.* The key point here is that air molecules at high temperatures are moving faster than air molecules at low temperatures. And since sound is essentially a relay race of air molecules, the faster-moving molecules in warm air can transmit sounds more quickly: in air at 0°F sound travels at 718 miles per hour; at 72°F it jumps to 772.

The second thing to know is that sounds don’t always follow straight lines; they bend in certain circumstances. Specifically, if there are layers of warm and cold air around, sound waves always bend toward the slower layer— toward the colder air. This bending is known as refraction.

[Meet the teen behind the first ‘Nigerian prince’ scam.]

To see refraction in action, imagine a trumpet player standing in the end zone of a domed football stadium. Imagine too that the stadium’s air conditioners are struggling to keep the place cool: there’s a nice layer of cold air near the ceiling but the field below is bathed in warm air. Because of refractive bending, any tootles on the trumpet will curve upward toward the cooler air. This means that someone standing in the opposite end zone will have trouble hearing anything, since the sound will sail over her head. Conversely, imagine a game later in the season. Now the stadium’s heaters are struggling, leaving the dome with a layer of warm air up top and cold air below. In this case the trumpet notes might start to rise— but will soon get bent back toward the ground, making them easy to hear. Again, sound always bends toward colder air.

The third thing about sound involves the temperature profile of our atmosphere. We all know that air gets colder as you rise, which explains why mountaintops near the equator can be snowcapped. By around 45,000 feet, the air temp drops to –60°F, which slows the speed of sound to 672 miles an hour. And just as you’d expect, noises outdoors tend to curve upward toward this cooler air. This explains why early balloonists could hear dogs barking and roosters crowing with such clarity. The atmosphere was actually funneling noise up toward them.

But the air cools as you rise only up to a point— that point being around 60,000 feet, when ozone starts to appear. Ozone absorbs ultraviolet light that would otherwise scramble our DNA; life never could have oozed out of the ocean and onto land without it. And in absorbing ultraviolet light, ozone warms up. All the ozone in the atmosphere, if collected and pressed together, would form a shell just an eighth of an inch thick. But it absorbs ultraviolet light so well that even this trifling amount of gas can warm the air at 150,000 feet to a balmy 32°F. Overall, then, our air forms a sort of temperature sandwich: there are two layers of warm air (one near the ground, one at around 150,000 feet), with a slice of cold air in the middle.

And here’s the payoff: this temperature profile sends sounds on a wild ride. Consider a hunter on the ground who blasts a shotgun. Per the discussion above, that sound will rise, curving toward the cooler upper air. But the thing is, sounds don’t just stop when they reach this layer. They have momentum, they keep going. So after passing through the cold layer at 45,000 feet, the sound will inevitably run into the ozone-warmed air above it. And because sound always bends away from warm air and toward cool air, the shotgun noise will actually do a gentle U-turn at this point and begin falling like an arrow. In other words, ozone reverses the sound’s direction, as if it bounced off a wall.

What happens next is even stranger. After it starts falling, the sound still has a fair amount of momentum. So it plows right through that cold layer at 45,000 feet and heads toward the ground. But what happens as it approaches the ground? It encounters a layer of warm air. And because sound always (say it with me) bends away from warm air and toward cold air, the majority of the sound energy will turn tail again and start to rise. But this of course sends it on a recollision course with that upper layer of ozone-warmed air. Whereupon it pulls a third about-face, and starts to sink. And it keeps sinking— until it meets the warm air near the ground and bounces back toward the sky again. In other words, the sound gets stuck in a loop. It keeps rising and falling, rising and falling, oscillating around that cold layer of air. That’s why this cold layer is called the sound channel, because sounds get swept toward it and have trouble escaping.

There are a few caveats worth noting about the sound channel. First, only pretty intense noises have enough energy to rise that high and get sucked into it. It’s not like your sweet whispered nothings from last night are still bouncing around the stratosphere, thank God. What’s more, the most intense noises,* after their initial U-turn in the sky, do sometimes have enough energy and momentum to push through the layer of warm air near the ground and strike the ears of listeners below. We’ve encountered this already with Mount Saint Helens. Remember that people near the eruption heard nothing, while people far away got clobbered with noise. That’s because the boom initially curled upward toward the cooler air, sailing over the heads of those nearby and creating a sixty-mile-wide “sound shadow.” But the boom nose-dived when it hit those warmer pockets of air above, allowing people farther away to hear. Something similar happened with the nuclear bomb at Hiroshima. Survivors near the epicenter spoke of the pika, the flash, while those farther away recalled the pika-don, the flash-boom.

[Meet the women who brought us Apollo 11.]

Maurice Ewing first worked out the physics of the sound channel in 1944.* It seemed little more than a novelty, though, until he realized something else. He now understood what happened to noises that originated above or below the channel— they got funneled into it. But what about noises that originated within the channel? How would they behave?

Consider a shotgun blast again, but this time at 45,000 feet, at the temperature nadir. Like all sounds, no matter where they originate, the noise from this blast will initially start to spread in all directions. And in spreading out like this, sounds usually dissipate, weaken. But something unusual happens around this specific height. No matter what direction the sound waves go, up or down, they encounter warmer air and get nudged back toward the center. As a result, sounds that start within the sound channel don’t spread out much— meaning they don’t weaken. They’re therefore audible at much farther distances than normal. They’re effectively magnified.

In 1947 Ewing realized that this effective magnification of sounds offered a clever way to spy on the Soviets. Now, the Soviets weren’t going to explode nuclear weapons nine miles up in the sky— that’s awfully high. But Ewing knew that mushroom clouds often do rise that high. Mushroom clouds are pockets of hot gas that knock other air molecules around. Knocking air molecules around is basically the definition of sound, and Ewing hoped that Soviet mushroom clouds would raise enough of a ruckus at nine miles high for him to hear it halfway around the world. All the air force had to do was send balloons with microphones into the sound channel to eavesdrop. The air force called the scheme Project Mogul.

Ewing was pretty optimistic about Project Mogul at first, but when he began running tests at Alamogordo Army Air Field in New Mexico in early 1947, he ran into several problems. One involved keeping balloons at a constant altitude, since sunlight warmed the balloon envelope. This in turn warmed the gas inside and caused the balloon to rise out of the sound channel. Ewing’s team countered this tendency by using transparent balloons, which allowed sunlight to stream through. (Ewing ordered them from the same company that made the first balloon figures for the Macy’s Thanksgiving Day Parade. When his assistants saw the transparent balloons, they immediately thought of something else: titanic condoms.)

Another problem involved tracking the balloons, since they wandered aimlessly with the wind. Ewing proposed tracking them with radar, but the equipment at Alamogordo had trouble finding these tiny targets at high altitudes. So the scientists decided to send up not one but thirty balloons at once; they were yoked together in a column sixty-five stories tall, more than twice the height of the Statue of Liberty. They also added radar reflectors to the balloon column, metal surfaces that helped redirect the radar waves back toward the ground. Each reflector looked something like a metallic box kite, and Project Mogul in fact contracted with a toy company to make them. Because the scientists didn’t care about aesthetics, the toy company bound the reflectors together with Elmer’s glue and tape. And because tape was scarce due to lingering wartime shortages, the company dipped into a stock of wacky novelty tape it had on hand — tape covered in purple, squiggly hieroglyphics.

Because the scientists didn’t care about aesthetics, the toy company bound the reflectors together with Elmer’s glue and tape. And because tape was scarce due to lingering wartime shortages, the company dipped into a stock of wacky novelty tape it had on hand — tape covered in purple, squiggly hieroglyphics.

As you probably guessed, these ungainly columns of metal, plastic, and rubber accounted for many of the “unidentified flying objects” that invaded the skies of Roswell in 1947. When aloft, the columns moved in mysterious ways, with different parts snaking back and forth at different times, depending on the wind. The radar reflectors glinted eerily in the moonlight as well, and when the columns crashed down, the metal gouged the earth and produced far more debris than any weather balloon could have.

This tendency to strew debris about became a headache for Maurice Ewing. Crazy as it sounds, Project Mogul received the same ultra-double-secret classification that the Manhattan Project had. Not even the folks at Roswell Army Air Field ninety miles away knew about it, which meant that Ewing’s team had to scramble to retrieve every scrap from every one of the 110 flights they launched. Most of the time they found the downed balloons easily enough; when they lost one, they actually listened to radio reports of UFO sightings for leads. But some balloon columns escaped— including the one that crash-landed on Mac Brazel’s ranch.

Given all the hoopla that followed, in later years Brazel said he regretted not keeping his shed locked and his trap shut. But for whatever reason, people across the globe seized upon his story, and the debris he’d found in the dirt somehow acquired otherworldly powers. The military’s response only fueled people’s suspicions, and Roswell soon metastasized into the phenomenon we know today.

[Learn the Cold War origins of modern GPS technology.]

Meanwhile, Project Mogul continued in secret for another few years, and a few accounts claim that Mogul balloons did detect Joe-I, the first Soviet nuclear weapon test, in August 1949. But so did other, cheaper, more reliable methods, such as sending airplanes aloft to scour the skies for radioactive dust. After years of marginal results, the air force finally shuttered Mogul in 1950.

At this point, with Mogul in the dustbin of history, the military could have come clean. Paranoid to the end, however, officials continued to stonewall and insist on the silly weather balloon story. Apparently the threat of the Soviet Union loomed so large in their imaginations that they preferred to let rumors about an alien invasion fester rather than tip off the Soviets to even a failed attempt to spy on them. By the time the air force owned up to Project Mogul, in the 1990s, it was too late: the Roswell rumors had taken on a truth of their own.

In a topsy-turvy way, though, history has proved the conspiracymongers right. The air force was indeed lying all those years, and it was indeed desperately scanning the skies above Roswell in 1947— but for portentous rumbles of gas, not alien star cruisers. And to think the whole twisted tale started with an acoustic quirk of our atmosphere, which in turn depended on the energy-absorbing prowess of ozone. By protecting the DNA of landlubbing creatures, ozone arguably did as much as any other gas to accelerate the evolution of life on Earth. And by enabling Project Mogul, ozone also convinced more people than ever of the subject of our next chapter, the existence of life on other planets.

[It’s a bird! It’s a plane! It’s snarge!]

Excerpted from Caesar’s Last Breath by Sam Kean. Copyright © 2017 by Sam Kean. Reprinted with permission of Little, Brown and Company.