The Glass Beehive That Saved Ships: How Augustin Fresnel Reinvented the Lighthouse

On the evening of August 20, 1822, a strange machine was set up on top of the unfinished Arc de Triomphe in Paris. It was a drum of glass built from rings and wedges of polished crystal held in a brass frame, with an oil lamp at its heart. When the lamp was lit and the drum began to turn, it threw eight slow beams across the city and out into the countryside. Members of France’s lighthouse commission watched the test, and accounts of the night say King Louis XVIII and his entourage saw the flashes from about 20 miles away.

The man who designed it was Augustin-Jean Fresnel, a 34-year-old road engineer with weak lungs and a day job in the government’s bridges and roads service. In his spare time he had just helped overturn Isaac Newton’s theory of light. Now he had turned that theory into a tool. Within a year his apparatus was shining from a lighthouse at the mouth of the Gironde, visible to the horizon. Within a few decades, versions of it were turning in lighthouses from Scotland to Saigon, and it earned a nickname that still follows it around: “the invention that saved a million ships.”

The idea at its core is simple enough to fit in a wallet. Most of the glass in a lens does nothing but sit there. Cut it away, keep only the curved surfaces that bend the light, and you get a lens that is thin, light and far more efficient. That trick is why there is a Fresnel lens in theater spotlights, car tail lights, page magnifiers and some virtual reality headsets. It is also why a craftsman in Florida recently spent more than a year building a new one, prism by prism, for America’s tallest lighthouse.

Lighthouses that wasted their light

For most of history, a lighthouse was a fire on a tower. The Romans burned wood. Medieval keepers burned coal or tallow candles. Even the grandest light of the age, the Renaissance tower of Cordouan off the French coast, burned oak chips in an iron basket, and on a good night it could be seen perhaps five or six miles out. A ship that needed warning of a reef or shoal often saw the light too late, if it saw it at all. Sailors were only just getting a reliable way to fix their longitude; seeing the rocks at night was an older problem.

Things improved in the late 1700s. The Swiss inventor Aimé Argand’s oil lamp, with its hollow wick and glass chimney, burned far brighter and cleaner than a candle. Engineers began putting polished metal mirrors behind these lamps, shaped as parabolas to throw the light forward in a beam, and set the mirrors on rotating frames so the beams swept the sea. These “catoptric” lights were a real advance, but they were still wasteful. According to the National Park Service, a bare flame loses nearly 97 percent of its light for a sailor’s purposes, and a lamp with a reflector still loses 60 to 80 percent. Even the best mirrors reflected only about half the light that struck them.

The obvious alternative was a lens, which bends light by refraction instead of bouncing it. In 1789 a London glass cutter named Thomas Rogers installed lenses at Portland Bill on the English coast. But a lens big enough to gather a lighthouse lamp’s glow had to be enormously thick in the middle, and Rogers’s were 5.5 inches through. The thick glass absorbed much of the light it was supposed to concentrate. In 1748 the French naturalist the Comte de Buffon had suggested grinding a lens into concentric steps to make it thinner, but he wanted it for a burning glass, and nobody could make one large enough to matter.

A sickly engineer and a theory of light

Fresnel was born on May 10, 1788, in Broglie, Normandy, the second of four sons of an architect. He was a frail child and considered the slow one of the family, though he was famous among the local boys for making toy bows and guns that worked far too well. He entered the elite École Polytechnique at 16, trained as a civil engineer, and spent the rest of his working life in the government’s Corps of Bridges and Roads.

He came to optics almost by accident. In 1814, stuck on a road project in the Drôme, he wrote to his brother asking for papers on the new research into the polarization of light. “Though I break my head,” he admitted, “I cannot guess what that is.” The next year politics gave him time to find out. When Napoleon escaped from Elba in March 1815, Fresnel left his post to join the royalist resistance, fell ill, and was suspended from his job. He spent his enforced leisure at his mother’s house in the Norman village of Mathieu, and there he began his experiments on light.

The question he took up was the oldest argument in optics. Newton had held that light was a stream of tiny particles. A minority, most recently the English physician Thomas Young with his double-slit experiment of 1807, argued that light was a wave. Fresnel, working mostly alone, built a mathematical theory of light as waves that explained the colored fringes that appear at the edges of shadows. In 1818 the French Academy of Sciences announced a prize competition on diffraction, and Fresnel entered.

One of the judges, the mathematician Siméon Denis Poisson, was a committed believer in particles. He worked through Fresnel’s equations and found what he thought was a fatal flaw. If the theory was right, then the shadow of a small round disk should have a bright spot of light at its exact center, where common sense said it should be darkest. The head of the committee, François Arago, decided to try it. He fixed a metal disk about two millimeters across to a glass plate with wax, lit it, and found the spot. Fresnel won the prize in 1819, and the bright point in the middle of a round shadow is still called the Arago spot, or, with some irony, Poisson’s spot.

Lenses by steps

Arago did Fresnel another favor that year. France’s Commission of Lighthouses, created by Napoleon in 1811, had done little since, and Arago, one of its members, recommended Fresnel for a “temporary” assignment to look at better lighting. Fresnel joined on June 21, 1819. By the end of that summer he was proposing what he called lentilles à échelons, lenses by steps.

The physics was the same physics behind his prize. Light slows down when it enters glass, and the change in speed bends it at the surface. Only the surfaces do the bending; the glass between them just carries the light along while absorbing some of it. (That slowdown was confirmed in 1850, using water instead of glass, by Léon Foucault, the physicist later famous for his pendulum, in an experiment that showed light travels more slowly in water than in air, as the wave theory predicted.) So Fresnel kept the curved surface of a big lens and threw away the bulk. Imagine slicing a thick magnifying glass into concentric rings and then sliding each ring down until they all sit nearly flat. Each ring keeps the slope it had in the original lens, so it bends light toward the same focus, but the whole thing is a fraction as thick and as heavy.

To his embarrassment, one of the commissioners, the physicist and balloonist Jacques Charles, remembered that Buffon had proposed something similar. But Fresnel’s design had a practical edge. Buffon imagined a single slab of stepped glass, impossible to cast at that size. Fresnel’s lens was built from many separate pieces, each small enough to make well, mounted together in a metal frame. Large pieces of glass are fickle things to cast and cool, as anyone who has seen Prince Rupert’s drops explode can attest, and small prisms were far easier to get right. With a budget of 500 francs, he went to three manufacturers. The third, François Soleil, found a way to remove flaws by reheating and remolding the glass. Arago helped design a lamp with several concentric wicks to put more light at the focus, and along the way discovered by accident that fish glue could stand the heat, so it could hold the prisms in place.

The prototype, finished in March 1820, was a square panel about 22 inches on a side made of 97 small prisms. The commission was impressed enough to ask for a full eight-sided version. On the evening of April 13, 1821, that model was demonstrated side by side with the best reflectors France had, and it made them obsolete in a single night.

The light at Cordouan

The machine that went to the top of the Arc de Triomphe the next summer was a revolving light. Its eight main panels were “bull’s-eye” lenses made in rings by the Saint-Gobain glassworks, each focusing a wedge of lamplight into a beam that swept past a ship as a flash. Below them, 128 small mirrors stacked like the slats of a Venetian blind sent a fainter steady glow toward the horizon between flashes.

After the Paris test, the apparatus spent the winter in Bordeaux and was then carried out to Cordouan, the tower built between 1584 and 1611 on a rock at the mouth of the Gironde estuary to guard the Bordeaux wine trade. Fresnel supervised the reassembly and did some of it with his own hands. On July 25, 1823, the world’s first lighthouse Fresnel lens was lit, and as promised it could be seen to the horizon, more than 20 miles out. Cordouan, nicknamed the king of lighthouses, is still working today and was named a UNESCO World Heritage Site in 2021.

By then Fresnel had also finished the theoretical work that made his name in physics, showing that light waves vibrate sideways, across their direction of travel, which explained polarization and double refraction. Shortly after the Cordouan lens went into service, he began coughing up blood. He had tuberculosis. Warned that he needed to cut back if he wanted to live, he chose the lighthouses. He resigned his post as an examiner at the École Polytechnique, closed his research notebooks, and became the commission’s secretary and chief engineer.

The work he crammed into his last three years shaped lighthouses for the next century. He designed a fixed lens, shaped like a drum, that spread light evenly all around the horizon; the first went into service at Dunkirk in 1825. He drew up a national plan, the Carte des Phares, calling for 51 coastal lights in a hierarchy of sizes, with each giving a distinct signal so sailors could tell them apart: a steady light, one flash a minute or two flashes a minute. And in late 1825 he solved the last big inefficiency. The mirrors above and below the lens still lost light, so he replaced them with specially shaped prisms. Light enters the first face of such a prism, reflects off the second face by total internal reflection, the same effect that keeps light trapped inside an optical fiber, and leaves through the third, losing almost nothing along the way. A lens with refracting rings in the middle and reflecting prisms above and below is the “beehive” that people picture when they think of a lighthouse lens.

Fresnel never saw a full-size version. Arago brought him the Royal Society of London’s Rumford Medal on his deathbed, and Fresnel whispered that “the most beautiful crown means little, when it is laid on the grave of a friend.” He died eight days later, on July 14, 1827, at 39. His younger brother Léonor took over as secretary of the commission and spent the following decades carrying out his plans.

Orders, flashes and a bath of mercury

Fresnel sorted his lenses by size into what are still called “orders.” A first-order lens, the largest standard size, has a focal length of 920 millimeters, about 36 inches, from the lamp to the glass. The lens alone stands about eight and a half feet tall and six feet across, and on its pedestal it towers over the keeper. These went in major seacoast lights. Smaller orders went in bays, harbors and rivers, down to the sixth order, with a focal length of about six inches.

In Scotland, Alan and Thomas Stevenson, uncle and father of the novelist Robert Louis Stevenson, refined the design with new reflecting prisms, and the English firm Chance Brothers added “group flashing” lenses that gave two or three quick flashes so neighboring lights could be told apart. The biggest lenses ever made, called hyperradiant, had a focal length of more than 52 inches, and only about 30 went into service. One that had been shown at the 1893 World’s Columbian Exposition in Chicago ended up in the Makapuʻu Point Light on Oʻahu. Twelve feet tall, with more than a thousand prisms, it has been the largest lighthouse lens in the United States since 1909.

A turning lens needed a motor, and before electricity that meant clockwork. Heavy weights hung in the tower, and as they slowly dropped they drove the gears that rotated the lens. Keepers wound them by hand, sometimes as often as every two hours through the night. Later lenses floated on a bath of liquid mercury, which let a lens weighing several tons spin with almost no friction. A keeper’s other big job was cleaning. Soot from the lamp and salt from the air dulled the glass, so keepers polished hundreds of prisms by hand.

Sometimes keeping the light meant much more. During the hurricane that struck Galveston in August 1915, the storm that tested the city’s new seawall, about 60 people took shelter in the Bolivar Point lighthouse across the channel. The Smithsonian, which now owns that lens, says its turning mechanism broke around 9:15 that night, and assistant keeper J.B. Brooks turned the lens by hand. By 10 o’clock the tower was shaking so hard that he feared the glass would shatter, so he stopped it, kept the lamp trimmed and showed a steady light until morning. (We told the story of that storm and the city that raised itself after the 1900 hurricane.)

America’s stubborn auditor

The United States was slow to follow, and the man usually blamed is Stephen Pleasonton. He was the Treasury Department’s “Fifth Auditor,” a careful bookkeeper best remembered for carting the Declaration of Independence and other State Department papers out of Washington in linen sacks before the British burned the city in 1814. From 1820 he also ran the nation’s lighthouses, and he ran them cheaply. His favorite supplier was Winslow Lewis, a former sea captain whose patented system of Argand-style lamps and reflectors Congress had bought in 1812. Lewis also held contracts to supply whale oil and won bids to build many of the towers.

Mariners complained loudly. The publishers of the American Coast Pilot, Edmund and George Blunt, forwarded angry letters from ship captains and pressed for reform. Pleasonton replied that Lewis’s lamps and reflectors cost about one-sixth as much as French lenses and were good enough for the American coast. In 1838 Congress finally sent the naval officer Matthew Perry, later famous for forcing open Japan, to France to buy two lenses. A first-order fixed lens and a second-order revolving lens, made by the Paris firm of Henry-Lepaute, went into the Navesink Twin Lights on the New Jersey highlands in 1841, the first Fresnel lenses in an American lighthouse. Including installation, they cost $24,000.

Pleasonton traveled to see them that September and was dazzled. “The cost of these lenses, however, is nothing compared to the beauty and excellence of the light they afford,” he wrote, adding that pilots reported seeing the light 40 miles off in clear weather. In the same letter, though, he warned that unless a keeper was “an intelligent mechanic,” the delicate apparatus would be “unfit for use in the United States upon a large scale.” Congress lost patience. In 1851 it ordered a board of naval officers, Army engineers and scientists, including Joseph Henry of the Smithsonian and Alexander Dallas Bache of the Coast Survey, to investigate, and in 1852 that body became the United States Lighthouse Board. It began buying Fresnel lenses wholesale, and by the time of the Civil War every American lighthouse had one.

History has been hard on Pleasonton, and some historians now think too hard. In a 2018 paper in The Northern Mariner, James Risk argued that the American delay was about normal for the time, given how slowly the French makers could turn out lenses and how scarce the skills to set them up were. Pleasonton’s worry about keepers had some basis: a decade after the Navesink lenses were installed, the new board found the larger of the two lenses so badly out of alignment that the smaller one shone brighter.

A lens goes to war

No American lens had a stranger career than the one at Cape Hatteras, North Carolina, off the shoals known as the Graveyard of the Atlantic. It was one of the first two first-order lenses the government bought from Henry-Lepaute in 1852, and before it ever reached the coast it was the centerpiece of the Lighthouse Board’s exhibit at New York’s Crystal Palace world’s fair of 1853, where, as the historian Kevin Duffus has found, more than a million people filed past it. In 1854 it went to the top of the Hatteras tower, raised to 150 feet for the occasion.

Then came the war. In 1861 the Confederate Lighthouse Bureau ordered the lights along the Southern coast darkened and their lenses removed, so they could not guide Union warships enforcing the blockade. The National Park Service notes that most lenses were not smashed, as legend has it, but carefully taken down, packed and stored, often by their own keepers. At Hatteras the keeper, Benjamin Fulcher, helped dismantle the lens. Its prisms were packed into 44 cotton-lined pine crates and shipped to a warehouse in Washington, North Carolina. When Union troops came up the Pamlico River to seize it in March 1862, the crates were loaded onto a small steamboat the night before and carried up the Tar River, then sent by rail to the village of Townsville near the Virginia line and hidden on a nearby estate. In September 1865, months after the war ended, a Union patrol found the 44 crates neatly stacked on a railroad platform in Henderson, North Carolina. Nobody could say who had put them there. Many of the state’s other lenses turned up in Raleigh, where Sherman’s troops found them in 1865.

The Hatteras lens was shipped to Paris for repairs and came home in 1868, and when a new and taller Cape Hatteras tower was finished in 1870, the old lens went to the top of it. It worked for 66 more years. Then erosion drove the sea toward the tower’s base, and in the mid-1930s the light was moved to a steel skeleton tower and the brick lighthouse was abandoned. With the doors unsecured, souvenir hunters climbed the stairs and began prying out the glass. By 1950 all 24 of the central flash panels were gone, along with roughly two-thirds of the lens’s 1,008 prisms. What was left was taken down and replaced by an electric airport-style beacon. A National Park Service official called the loss “a disgrace.” The surviving pieces are now on display at the Graveyard of the Atlantic Museum in Hatteras village.

From beehives to headsets

By the middle of the 20th century, the great glass lenses were being retired. Electric lamps with compact plastic optics, rotating aerobeacons and eventually LEDs did the job with far less upkeep, and many old lenses went to museums. Some still turn in their towers. At Point Reyes in California, the first-order lens lit in 1870 still sits in its lantern above the Pacific, restored in 2019.

Fresnel’s idea, meanwhile, escaped the lighthouse. In 1852 a Brooklyn glassworks patented a way to press stepped lenses in a single piece, and small Fresnel lenses were soon common on railroad lanterns and steamboats. Once they could be molded from plastic in the 1950s, they turned up everywhere: under the glass of overhead projectors, in car tail lights and traffic signals, and in the domed window of a motion-sensor light, where plastic Fresnel facets gather infrared heat from across a room. Stage and film lights with Fresnel lenses are so common that crews just call them “Fresnels.” Aircraft carriers use them in the “meatball” landing lights that tell a pilot whether the approach is too high or too low.

The same geometry works in both directions. A large flat plastic Fresnel lens can gather sunlight onto a small spot, which is how solar cookers and concentrating solar cells work, the job Buffon had in mind for his stepped burning glass in the 1700s. Thin Fresnel lenses also helped make virtual reality headsets light enough to wear: the Meta Quest 2 and HTC Vive Pro used them, and so does the budget Meta Quest 3S, though pricier headsets such as the Quest 3 have moved to thinner, sharper “pancake” optics.

The beehive comes back

The tower that lost its lens is about to get one again. In 1999 the Cape Hatteras Lighthouse, all 4,800 tons of it, was rolled about half a mile inland to escape the eroding beach, and for 25 years it carried an ordinary rotating beacon. As part of a restoration project begun in 2024, the National Park Service hired Dan Spinella, an engineer at Walt Disney World who has spent decades restoring lighthouse lenses on the side, to build a replica of the original Henry-Lepaute lens. Restoring the original was ruled out: only 268 of its prisms survive, and its warped frame could no longer take new ones.

Spinella worked from measurements of the surviving pieces and from Thomas Stevenson’s 19th-century optical formulas. His 1,008 prisms are machined from optical acrylic, dyed to match the faint sea-foam green of the old crown glass, with each profile reworked for acrylic’s slightly lower refractive index. The replica weighs about 1,600 pounds, roughly a third of the original, and it will turn on a working clockwork pedestal, driven now by a small electric motor, to restore the old signal of one flash every 10 seconds, visible about 20 miles out. The Park Service says the lens was completed in October 2025 and will be installed near the end of the project, which has been slowed by corroded support brackets at the top of the tower and is expected to run through at least the end of 2026.

When it goes up, the light inside will be an LED, not an oil flame, and the prisms will be plastic, not glass. But the geometry will be exactly what Fresnel worked out with a budget of 500 francs, a few glassmakers and a pot of fish glue: throw away the glass that does no work, and bend every ray you can catch toward the horizon. He had only a few years to see it work. His idea has now been doing its job for more than two centuries.

Further reading