The Glass Tear That Shrugs Off a Hammer: The 400-Year Puzzle of Prince Rupert’s Drops

Hold a Prince Rupert’s drop by its tail and you can lay the fat end on an anvil and hit it with a hammer. Nothing happens. People have squeezed the heads in vises hard enough to dent the steel jaws, and the glass survived. Researchers have fired bullets at them on camera. Then pinch the thin, curly tail between your fingers, snap it, and the whole drop bursts with a sharp pop into a spray of glittering powder, faster than any eye can follow.

The little glass tadpoles have been passed around since the 1600s, first as party tricks, then as a puzzle for the founders of modern science. “Rupert’s drops have been a curiosity for about 400 years,” the Purdue University engineer Srinivasan Chandrasekar said in 2017, after he and his colleagues finally measured what was going on inside one. The answer turned out to be a story about frozen-in stress, and the same idea sits in the side windows of your car and, in a different form, in the glass of your phone.

If you have read our stories on Euler’s Disk or the tippe top that flips itself over, you know the type. These are small objects that anyone can hold, that do something that seems to break the rules, and that took scientists far longer to explain than you would expect. Prince Rupert’s drop may be the oldest member of the club.

A toy from the glassworks

A Prince Rupert’s drop is made by letting a blob of red-hot molten glass fall into a bucket of cold water. If it survives the plunge, which many don’t, it freezes into a teardrop with a round head usually several millimeters across and a long, thin tail. Nobody knows who first made one. Glassmakers may have known the trick since Roman times, but the first reliable accounts come from Mecklenburg in northern Germany around 1625. From there the drops spread across Europe as toys and curiosities, under names like lacrymae Batavicae (Dutch tears) and lacrymae Borussicae (Prussian tears). The Dutch scientist Constantijn Huygens asked the writer and natural philosopher Margaret Cavendish, Duchess of Newcastle, to look into them. After experimenting, she concluded that a small amount of volatile liquid must be trapped inside. She was wrong, but no one had a better idea yet.

The English name comes from Prince Rupert of the Rhine, a nephew of King Charles I and one of the more colorful figures of his century. Born in Prague in 1619, he became the dashing Royalist cavalry commander of the English Civil War, later an admiral, and the first governor of the Hudson’s Bay Company, whose vast territory in Canada was called Rupert’s Land. He was also an enthusiastic experimenter who kept a laboratory with its own forges, and he became a noted artist in mezzotint, an engraving technique he helped bring to England. He did not invent the glass drops. But he is credited with bringing them to England around 1660, when the monarchy was restored, and showing them to his cousin, King Charles II.

The king’s homework for the Royal Society

The timing was perfect. A group of natural philosophers had held the first formal meeting of what became the Royal Society on November 28, 1660, at Gresham College in London. A few months later the king gave them a job. According to the society’s minutes for March 4, 1661, “The King sent by Sir Paul Neile five little glass bubbles, two with liquor in them, and the other three solid, in order to have the judgement of the society concerning them.” The society’s assistant made more drops just two days later, and they behaved the same way.

On August 14, 1661, the society’s president, Sir Robert Moray, gave a full report of the experiments. Because the society did not yet have its own journal, the account was recorded in its register book and first printed in 1662 as an appendix to Christopher Merret’s English translation of Antonio Neri’s The Art of Glass. It included a simple drawing, shown at the top of this story, that labeled the parts, from the thick body to “the point or end of the thread.”

The drops quickly became the talk of London. On January 13, 1662, the diarist Samuel Pepys wrote that a dinner guest “did show us the experiment (which I had heard talk of) of the chymicall glasses, which break all to dust by breaking off a little small end; which is a great mystery to me.” Samuel Butler worked them into his satirical poem Hudibras: “Honour is like that glassy bubble / That finds philosophers such trouble, / Whose least part crack’d, the whole does fly, / And wits are crack’d, to find out why.”

Robert Hooke gets cut

The person who did the most careful early work was Robert Hooke, the Royal Society’s curator of experiments, who described his tests in Micrographia, his famous 1665 book of observations through the microscope. Hooke broke drops in the open air and found that “the whole bulk of the drop flew violently, with a very brisk noise, into multitudes of small pieces, some of which were as small as dust.” The fragments scattered “so violently, that some of them pierced my skin.” He snapped them underwater, with the same result.

Then he got clever. He ground away nearly two-thirds of the round head on a stone, and the drop still did not explode, until he nipped off the very tip of its tail and it “flew immediately all into sand.” He coated drops in isinglass, a clear, tough glue made from fish, wrapped them in soft kid leather, and broke the tails. The drops cracked and kicked against his hand, but the skin of glue held every piece in place, so he could study the pattern of the fractures. And he heated drops red hot and let them cool slowly. Those lost their “fulminating or flying quality” completely and became ordinary, easily broken glass.

Hooke’s explanation was remarkably close. He reasoned that when the hot drop hits cold water, its outside is “presently cool’d and crusted,” while the inside is still hot and soft. As the inside cools, the hard crust will not let it shrink properly, so the glass is left “bent,” like a spring that cannot let go. He compared the drop to the stones in a vaulted arch, where “no one stone can be taken away but the whole Arch falls.” Modern researchers credit him with getting most of it right, including the role of quenching and the fact that slow cooling, or annealing, removes the effect. Where he slipped was in which way the forces pointed. He thought the outer surface was being pulled apart. In fact it is being squeezed. And no one in the 17th century had the theory of how cracks grow that would explain why that matters. That had to wait for the engineer A. A. Griffith’s work on brittle fracture in 1920.

Squeezed outside, stretched inside

Here is what happens in that bucket of water. The outer layer of the molten drop cools and turns solid almost instantly, forming a rigid shell around a core that is still hot. As the core cools, it tries to shrink, the way most materials do. But it is stuck to the shell, so it pulls inward on the shell from all sides. The result, once the whole drop is cold, is a thin outer layer that is squeezed very hard (compression) wrapped around an interior that is being stretched (tension). The two sets of forces balance each other, and the drop sits there, perfectly still, holding a lot of stored energy.

Glass almost always breaks because a tiny crack on its surface is pulled open. Every piece of glass has microscopic flaws, and under tension a crack at one of those flaws can run through the whole piece. Squeezing the surface works against that. A crack cannot easily open while the glass around it is being pressed shut. That is why the head of the drop can take a hammer blow. As Purdue’s Koushik Viswanathan put it, “if you could change the tensile stress to a compressive stress, then it becomes difficult for cracks to grow.”

Breaking the tail is different. The tail is thin enough that a finger can snap it, and the break gives a crack a path into the stretched core. Once a crack reaches that region, the stored energy drives it forward. In 1994, Chandrasekar and the Cambridge physicist Munawar Chaudhri filmed drops disintegrating at close to a million frames per second. They saw cracks racing from the tail toward the head at about 1,450 to 1,900 meters per second, more than 4,000 miles per hour, and splitting again and again as they went. The whole drop is gone in a few millionths of a second. The crack front slows dramatically when it reaches the squeezed outer layer, but by then the interior has already come apart.

How many pieces? In 2021, physicists at the University of Amsterdam broke a single small drop inside a latex glove, gathered the remains and scanned them with X-ray micro-computed tomography. They counted more than 20,000 fragments. Unusually for broken glass, the pieces tended toward one typical size rather than the usual huge spread from big shards to dust, a sign that the break is driven by the stress locked inside the drop rather than by the force of a blow.

Measuring the squeeze at last

One question was left: exactly how strong is that squeeze, and how deep does it go? Glass under stress bends polarized light, which is why a drop viewed between two polarizing filters glows with rainbow-colored bands. In work published in Applied Physics Letters and publicized in 2017, Chandrasekar and Chaudhri teamed up with Hillar Aben of Tallinn University of Technology in Estonia, a specialist in this kind of measurement. They suspended drops in a clear liquid, shone a red LED through them, and used a technique called integrated photoelasticity, with math similar to what reconstructs a CT scan, to map the stress throughout the drop.

The surface turned out to be squeezed much harder than earlier estimates suggested, with compressive stress reaching about 700 megapascals, nearly 7,000 times atmospheric pressure. Purdue compared the strength of the head to that of some grades of steel. Yet this armored layer is thin, only about 10 percent of the diameter of the head. That explains both halves of the mystery. Cracks that start on the surface of the head tend to run parallel to it, so they can’t dive into the stretched core. The tail gives them a way in. “The work has fully explained why the head of a drop is so strong,” Chaudhri told Phys.org. “I believe we have now solved most of the main aspects of this area. However, new questions may emerge unexpectedly.”

Bullets, and a flash of light

If a hammer can’t break the head, what about a bullet? Destin Sandlin, who makes the YouTube series Smarter Every Day, set out to find out with high-speed cameras filming at 150,000 frames per second and faster. A .22-caliber lead bullet splattered against a drop’s head. In his footage, the drop survived the hit itself, and when one did burst, it was because the impact sent a shock along the tail and set it whipping until it snapped. In a follow-up, a drop even took a hit from a .38 Special. Only faster rifle rounds finally destroyed one, and in that footage the exploding glass gave off a flash of light, which Sandlin connected to triboluminescence, the same light-from-breaking effect we explored in Spark in the Dark.

From party trick to safety glass

The drops were more than a curiosity. Their lesson, that glass cooled fast on the outside becomes far tougher, probably helped inspire the modern process for toughening glass. In 1874 the Frenchman François Barthélémy Alfred Royer de la Bastie patented a method of quenching nearly molten glass in a hot bath of oil or grease. In 1877 Friedrich Siemens in Germany developed another version that pressed the glass in cool molds. “The first patents to strengthen glass were in the 19th century, so there was a time lag of more than 200 years,” Chandrasekar noted.

Today’s tempered glass is made by heating a finished sheet to around 620 °C (about 1,150 °F) and then cooling its surfaces quickly with forced air, the same principle with better control. It is about four times as strong as ordinary annealed glass. It is used in car side and rear windows (windshields are a different, laminated glass), shower doors, glass tabletops and cookware. And like a Prince Rupert’s drop, it carries a squeezed skin around a stretched core, so when it does fail, it fails all at once. A sharp point that gets through the compressed surface, or a tiny nickel sulfide impurity inside, can make an entire pane crumble into small, blunt chunks instead of long, dangerous shards. That is the safety feature. It is also why car escape tools for breaking side windows are built around a small, hard point rather than brute force.

The glass on many phones uses a chemical cousin of the trick. Instead of heat, the glass is soaked in a bath of molten potassium salt, and larger potassium ions push into the surface in place of smaller sodium ions. Crowding the surface puts it in compression, up to about 690 megapascals, almost exactly what was measured in the head of a Prince Rupert’s drop.

Volcanoes make them too

Nature got there first. Volcanoes such as Kilauea in Hawaii fling out droplets of molten lava that freeze in the air into small black teardrops known as Pele’s tears, named for the Hawaiian volcano goddess, often trailing fine threads of volcanic glass called Pele’s hair. Volcanologists have long noticed the resemblance to Rupert’s drops, and some now use the drops as small laboratory models of how hot magma shatters when it meets water.

In a 2022 study in the Proceedings of the National Academy of Sciences, the volcanologist Katharine Cashman and colleagues broke drops in air, water and golden syrup, and found that the surrounding fluid changed the size of the fragments. They were trying to understand eruptions in which magma meets seawater, like the violent January 2022 eruption in Tonga. One of their more surprising findings came from a studio visit to Bristol Blue Glass in England. To make bubbly molten glass, the artists added a potato to the furnace, which released steam into the melt. No matter how many times they tried, the bubbly glass never formed an intact drop. It shattered on contact with the water instead. That is a clue, the researchers suggest, to why bubbly magma can fragment so explosively when it hits water.

A metaphor that stuck

For a while, Prince Rupert’s drops were common knowledge among educated people, and writers kept reaching for them. In 1867, the New York diarist George Templeton Strong described river ice that “flashed into fragments all at once like a Prince Rupert’s drop.” Sigmund Freud, writing in 1921 about how panic breaks up an army that has lost its leader, said, “The group vanishes in dust, like a Prince Rupert’s drop when its tail is broken off.” It is a good image for anything that looks solid but is held together by hidden tension.

It is also a reminder that the hidden stress inside a material can be as important as what the material is made of. Science is full of slow mysteries like this one. The tar in the University of Queensland’s famous pitch drop experiment takes about a decade to let go of a single drop. The Rupert’s drop goes the other way. It took centuries to explain and lets go in a few millionths of a second.

See it for yourself

You don’t need a furnace to see the stress. Tempered glass shows its locked-in strain pattern through a polarizer, such as a pair of polarizing sunglasses. Next time you are in a parked car on a bright day, look at the rear or side windows through polarized lenses and tilt your head. You may see faint dark patches or bands. That is the strain pattern frozen in by the rapid cooling, the industrial version of Rupert’s cold bucket. A pair of inexpensive polarizing film sheets works too, and is the same basic tool the Tallinn team used, minus the math.

Making your own drops is another matter. It takes a torch or furnace hot enough to melt glass, and as Hooke learned, the pieces fly hard enough to cut skin. If you ever get to break one, at a glass studio or a science demonstration, wear eye protection and do it inside a bag or a cupped cloth. The head will laugh at your hammer. The tail will not.

Further reading