Home Blog Which Way Does a Reverse Water Sprinkler Spin? Feynman's Unsolved Mystery Solved

Which Way Does a Reverse Water Sprinkler Spin? Feynman's Unsolved Mystery Solved

TahaSite Editor
Nov 04, 2026
230 Views
Which Way Does a Reverse Water Sprinkler Spin? Feynman's Unsolved Mystery Solved

Richard Feynman blew up a laboratory to discover the direction of rotation of a water sprinkler that sucks in water; however, the secret of this century-old mystery has just been revealed using a laser.

Close your eyes for a moment and imagine a lawn sprinkler in a park; that S-shaped pipe where water squirts out from both ends and spins around. Now, imagine picking it up, plunging it into a pool of water, and reversing the flow so that instead of spraying water, it sucks water into the tips of its arms. Which way does the sprinkler spin?

Before you answer, hold on a second. An answer probably immediately came to mind, and you are likely completely sure about it. That is where the problem begins. Richard Feynman, one of the most brilliant minds of the 20th century, wrote about this exact question, noting that the answer seems completely obvious at first glance. He then provided a strong argument for rotation in one direction, and an equally strong argument for rotation in the opposite direction. And about what the sprinkler actually did, he didn't write a single word.

A question that looks like a fifth-grade puzzle blew up a glass bottle, threw a young physicist out of the laboratory, and sparked arguments among physicists for nearly 140 years. And its answer was ultimately found not in equations, but in a place no one expected.

The Explosion in Princeton; The Day Feynman Was Expelled from the Laboratory

Around 1940, years before his name became intertwined with quantum physics and the Nobel Prize, Richard Feynman was a young graduate student at Princeton who enjoyed tinkering with physical objects much more than equations. The riddle of the garden sprinkler was being passed around the corridors of the physics department, and professors and students were arguing over its answer. Everyone had a solution, and every solution seemed obvious. Feynman preferred to take action rather than argue.

The experiment was conducted in the cyclotron laboratory, one of the university's most advanced scientific facilities. Feynman built a small, S-shaped sprinkler and first made sure it worked like a normal sprinkler. Then, he painstakingly lowered the entire suspended apparatus through the mouth of a large glass bottle filled with water.

Feynman's sprinkler gave just a short twitch and stopped moving as the flow continued.

His plan was for compressed water inside the bottle to push water into the sprinkler nozzles and force it out through a pipe. In other words, the submerged sprinkler would suck in water—the exact scenario everyone was arguing about.

Edward Creutz, the man in charge of the cyclotron, stood beside him and helped set up the experiment. They increased the pressure. The sprinkler rotated about five degrees, and in the direction of a normal sprinkler. Then it gave a short shudder and returned to its original position.

Creutz wrote decades later that from that moment on, they increased the pressure about five times; water passed through the device in a free and continuous stream, but the sprinkler moved no more.

Wheeler, Feynman's thesis advisor, described the incident more concisely. At the very first moment of applying pressure, a small tremor was seen, but as the flow continued, there was no reaction. So Feynman increased the pressure. Then more, and yet more.

The glass bottle couldn't withstand it and exploded. Water and pieces of glass scattered across the cyclotron room floor. The janitor came and helped collect the broken glass and mop up the water. The verdict was also delivered; from that day on, Feynman was not allowed to enter the cyclotron laboratory.

Krotch later wrote that he didn't know what Feynman was expecting, but his own vague notions about time reversibility were shattered, just like the bottle. Feynman didn't say anything clearer either, and years later, when he recounted the story in his memoir, he said nothing about the outcome.

The real issue begins right here: if the sprinkler didn't move except for that first twitch, why did Feynman keep increasing the pressure to the point of explosion? Perhaps he was waiting for a rotation that never came, and perhaps the answer was simpler than anyone, even Feynman, could have imagined.

The Shadow of Mach; A Mystery Older Than Feynman

چهار مدل مختلف از مسیر حرکت آب در دستگاه‌های آب‌پاش
چهار هندسه‌ی آزمایش‌شده. پژوهش ۲۰۲۶ آب‌پاش‌هایی با هندسه‌های S، مارپیچ، خم بازگشتی و حلقه‌ای را آزمود تا نقش شکل بازوها در ایجاد گشتاور روشن شود

There was an interesting point about the Princeton incident; none of those who were debating the sprinkler in the faculty hallways apparently knew that the answer to this question had been written about half a century earlier in a famous book.

In 1883, Ernst Mach, an Austrian physicist and philosopher, published a book titled "The Science of Mechanics," which later deeply influenced the young Einstein. Mach's name is mostly familiar to us through the Mach number and the speed of supersonic aircraft, but in the third chapter of this same book, he examined a device that bore a strange resemblance to Feynman's sprinkler.

Half a century before Feynman, Ernst Mach had performed the exact same experiment with air

Instead of water, Mach used air. He connected a small wheel with curved arms to a hollow rubber ball. By squeezing the ball, air rushed out from the tips of the arms, and the wheel spun, just like a garden sprinkler. The real excitement came in the next stage, because when he released the ball, air was sucked into the arms from the outside. Mach looked closely and wrote that the device showed "no discernible rotation."

Mach's answer was simple and calm, and no one made a fuss or kicked up a storm over it. His explanation also seemed intuitive; the air blowing outward creates a structured, directional jet, but the air being pulled inward comes from all directions, and its pushes cancel each other out. The story was over; at least for a few decades.

Feynman himself later realized the puzzle wasn't his. In 1966, the editor of the Physics Teacher magazine asked him to write about the sprinkler. Feynman declined the offer, protested it as the "Feynman problem," and referred the editor to Mach's book.

However, names don't always stick to their rightful owners. Feynman's memoir, with its charming account of the exploding bottle, made it to the global bestseller lists, and the puzzle forever bore his name. The more interesting point is that both Mach and Feynman missed something. Perhaps there was nothing to see, or perhaps there was something that the eyes and equipment of those days could not reach.

Half a century of dispute over a sprinkler

جریان آب در یک دستگاه آب‌پاش معکوس با دو جت ورودی و گردابه‌های بازچرخشی
نمای شماتیک جریان درون آب‌پاش معکوس؛ دو جت نامتقارن در محفظه‌ی مرکزی چهار گردابه می‌سازند و گشتاوری ضعیف ایجاد می‌کنند.

Following Feynman’s death in 1988, the mystery of the sprinkler did not settle; instead, it gained a new lease on life. Everyone had a narrative, and every narrative refuted another's answer. James Gleick, the science journalist, brought the story to a strange climax in his famous 1992 biography, Genius.

He wrote that Feynman and Wheeler obsessively kept the answer to the question hidden from everyone, and a few lines later claimed that Feynman's answer had been clear from the beginning: the sprinkler does not spin. The explanation Gleick offered to prove his point was nothing more than a repetition of Mach's old argument—the same incoming water coming from all directions.

Experiments, ranging from stillness to rotation in both directions, showed contradictory results.

A few years later, an unexpected figure entered the debate. Marilyn vos Savant, a writer who claimed to possess the highest recorded IQ, wrote an account of Feynman's experiment in a 1996 column for Parade magazine, declaring the case closed because the reverse sprinkler does not move.

Laboratories did not have a unified voice either; for example, at the MIT Edgerton Center, a suction-driven sprinkler was exhibited that showed no movement at all. In contrast, Berg and Collier in 1989, and two years later along with Ferrell, claimed that the reverse sprinkler rotates toward the incoming water—meaning precisely the opposite direction of a conventional sprinkler. Other reports emerged stating that by manipulating the internal shape of the device, rotation in both directions could be observed.

The theorists didn't sit idly by either and entered the fray. Most calculations concluded that an ideal reverse sprinkler, once the water flow stabilizes, experiences no torque. It might twitch a little at the moment of start, but then it stops. Wheeler and Krotkov's account of the initial twitch of Feynman's sprinkler also matched this picture.

So why did the answers seem so scattered? The main culprit wasn't physics, but an old and silent enemy: the friction factor, which alters many equations. Any effect that caused the reverse sprinkler to move was so weak that the friction of the device's axis easily neutralized it.

Axis friction neutralized the weak reverse rotation effect before it could be seen

Simply put, the devices didn't see anything because the signal of the effect was weaker than the noise of the error. Only in a few experiments where friction was minimized and the water flow was strong, a feeble, reverse rotation was observed.

A century later, the problem statement changed slightly. The question was no longer which way the sprinkler rotates; what mattered was how such a delicate thing could be observed. Before getting to the final answer, we need to take a step back and see why a regular sprinkler rotates in the first place.

Why a regular sprinkler rotates: A rocket on the grass

نمودار علمی یاتاقان مبتنی بر کشش سطحی با اجزای داخلی
پژوهشگران با شناورکردن دستگاه روی لایه‌ای از آب، تماس با سطوح جامد و اصطکاک محور را به حداقل رساندند.

If you've ever carelessly let go of a garden hose, you've seen how it writhes in every direction like a wild snake. Or think of a balloon you've blown up and released, darting around in the air. Both cases noisily repeat the lesson of Newton's third law: anything that is forcefully ejected pushes its launcher in the opposite direction.

The garden sprinkler follows the exact same rule, with just a twist. Water enters from the center, moves forward along the arms, and shoots out at high speed from the end of each arm where the pipe bends. Every burst of water pushes the tip of the arm backward, and because the arms are mounted on a free axis, the result of these pushes is the rotation of the entire device.

The gush of water pushes the arms of a conventional sprinkler backward like a rocket thruster.

Ristroph, the physicist who later solved the puzzle, considers the ordinary sprinkler to be a rotating version of a rocket—a rocket that spins around its own axis instead of heading toward the sky.

Physicists use a concept called angular momentum to describe this rotational motion. The curved arms of the sprinkler direct the water outward in a way that its motion creates angular momentum relative to the central axis of the sprinkler. According to the law of conservation of angular momentum, the sprinkler rotates in the opposite direction, simple and neat.

Now, reverse the film. The submerged sprinkler draws water inward through the tips of the arms. What happens then? Two arguments, both logical, lock horns.

The first argument states that the tip of the arm acts like a vacuum cleaner nozzle. The water pressure at the suction opening is lower than the water pressure behind the arm, so it pushes the surrounding water toward the opening. As a result, the sprinkler should rotate toward the incoming water, which is the reverse of the normal state.

In the reverse state, the forces at the nozzle and the pipe bend ideally cancel each other out.

The second argument states that the water entering the arm strikes the inner wall at the pipe bend and pushes it in the opposite direction; therefore, the sprinkler should rotate just like in the normal state.

In an ideal world, two forces cancel each other out and the sprinkler does not move. Mach's explanation is also consistent with these two arguments; the incoming water, unlike the outgoing jet, does not form a uniform stream and comes from all sides, so it does not bring any rotation with it either. In Ristroph's words, the main mystery lies right here.

The water that is sucked in looks nothing like a jet. Therefore, if the reverse sprinkler is supposed to rotate, its engine must be sought elsewhere.

Hunting the Invisible Vortex with a Laser; The Hidden Motor of the Sprinkler Found

جریان آب در دو حالت عادی و معکوس درون یک آب‌پاش
در حالت عادی، آب‌پاش خلاف جهت جت‌های خروجی می‌چرخد؛ مکش آب در حالت معکوس، چرخشی آهسته در جهت مخالف می‌سازد.

Leif Ristroph, an applied mathematician at New York University, has a peculiar interest in tangible real-world puzzles; for instance, his lab had previously tackled questions like the formation of the stone forests of China and Madagascar. Feynman's sprinkler was of this exact same nature; a simple question that had no simple answer.

There were four people on Ristroph's team: Kaijie Wang, then a PhD student at NYU; Mingxuan Zhu; Ristroph himself; and a mathematician named Brennan Sprinkle from the Colorado School of Mines, whose name curiously matches the subject of the research. Sprinkle in English relates to scattering water; the root of the word from which "sprinkler" originates.

A century of failures showed that until friction was removed from the path, nothing could be seen. Instead of ready-made sprinklers, the researchers built their own devices and mounted them on a bearing that operated using the surface tension of the water itself.

The sprinkler had no contact with any solid surface while rotating, and friction was reduced to a negligible level. The water that was supposed to create the mystery had now become the bearing itself. Ristroph recounts that at every stage, they were forced to invent a new method.

For their next big step, the research team needed to see inside the device; therefore, they designed the sprinkler in such a way that the flow of water outside, inside, and between the arms was visible—a feat that, according to Ristroph, had never been done before and became the key to solving the mystery. The researchers added dye and tiny particles to the water, illuminated it with a laser, and captured the water's movement using high-speed cameras.

Two angled jets in the central chamber slowly rotated the sprinkler in the reverse direction of the usual.

The result finally put an end to the century-old debate. The reverse sprinkler does rotate, and in the opposite direction of a regular sprinkler. A stable and repeatable rotation, yet very slow; about 50 times slower than normal. Importantly, both modes essentially share the same mechanism. The reverse sprinkler was also a rocket; just an inverted one.

The story came down to that small bend at the end of the arms. The water sucked from the tip of the arm passes through the pipe's curve on its way toward the center. At this bend, centrifugal force throws the water toward the outer wall of the bend; much like a passenger pressed against the car door on a sharp road curve. The result is an angled and asymmetric flow.

After passing through the arms, the water enters the central chamber in the form of two small jets; right where the two arms meet. Ristroph says the secret of the story was hidden right here, inside the sprinkler. If the two jets had collided directly head-on, they would have canceled each other out. But the deflected flow caused the two jets to collide slightly off-center; much like two people pushing a revolving door from opposite sides, not right from the middle. The result is a small vortex at the heart of the device and a torque that slowly rotates the entire sprinkler.

Physicists named this explanation the flux of angular momentum, a stream of rotation that moves along with the water. The motor that Mach, Feynman, and a generation of physicists wanted to discover was inside the device the whole time.

Is the problem really solved? When geometry rules the flow

نمودار جهت چرخش آب‌پاش در دو حالت مختلف جریان آب
Wikimedia Commons

In January 2024, scientific journals reported on the solution to the century-old mystery of Feynman's sprinkler. The headlines were definitive, and there was no reason for doubt. But the researchers themselves knew their work still had a very important missing piece.

The 2026 study tested various geometries to rule out competing theories

The 2024 experiments were performed only on classical sprinklers with S-shaped arms. No one knew whether this same explanation would hold if the shape of the arms became more complex. On the other hand, the 2024 data served as good support for the team's theory, but they could not eliminate competing theories. In science, proving the validity of one explanation is not enough; one must show that other explanations do not work.

The answer this time came from backyards. Every summer, American lawns are filled with fancy and whimsical sprinklers—sprinklers with coiled and spiral pipes that spray water in bizarre shapes. Ristroph and Sprinkle's team, joined this time by two new students, designed a collection of these very sprinklers with various curvatures and tested all of them in both spraying and suction modes.

This time, the researchers did not just look at rotation. They recorded the flow of water in and out of the device and, in a separate experiment, held the sprinkler firmly to directly measure the torsional force exerted on it. Then, three theories were put under the microscope.

The first theory went back to Mach. Mach, of course, had not only written about the immobility of the sprinkler; in the same book, he had also described another device that rotated in the opposite direction during suction. The explanation attributed to him based on the fluid's angular momentum states that the water spins in one direction and the sprinkler in the opposite.

The data did not match the first argument; the rotation and reverse torque recorded by the researchers had no explanation within Mach's framework.

Only angular momentum flux could explain the movement and torque of all geometries.

The second theory was the explanation attributed to Feynman and his followers, which focused on the flow of water in the outermost part of the arms. Fresh experiments showed that the outer part of the arms and the surrounding flow play no role in the sprinkler's movement.

The third theory, namely angular momentum flux, was the only one that emerged victorious from the test. The team expanded the theory and saw that it works for both reciprocal states and all strange and unusual shapes. Newer findings also showed that the shape of the arms can control internal jets. The title of the paper published in the PNAS journal in 2026 says just that: "Geometry, Momentum Flux, and the Sprinkler Problem."

The story of Feynman's sprinkler is a reminder that the phrase "solved" in science is often a station, not a destination. It took two years for an answer that seemed correct to turn into an answer with no rivals left.

From a Backyard Sprinkler to Future Turbines

پرتره جیمز گلیک روزنامه‌نگار علمی- مردی میانسال با لبخند
جیمز گلیکWikimedia Commons

Let's return to the first thought experiment of the article. An underwater sprinkler sucks in water. What was your guess? If you said it doesn't spin, you were in agreement with Mach, Gleick, and Vos Savant. If you said it spins in reverse, you guessed correctly; it just spins so slowly that it took a century for someone to notice.

Now, the Princeton explosion takes on a new meaning. If Feynman's sprinkler had also spun in reverse, the friction of its axis would have choked off the rotation—which was 50 times slower than normal—before it even started. Mach also saw correctly that he saw nothing; he just got the reason wrong.

Ristroph himself admits that the reverse sprinkler has little practical use; after all, who wants to collect water from their lawn instead of watering it? However, the story doesn't end with just a sprinkler. Oceans and rivers are full of currents that carry massive amounts of energy, and harnessing this energy is impossible without a precise understanding of fluid dynamics. Sprinkle also states that this exact understanding of how components react to fluid flow can pave the way for designing devices like turbines.

Perhaps the most beautiful part of the story is what Feynman didn't do. The man known for his boldness in proposing strange ideas remained silent about the sprinkler for years; he neither sold his guess as the answer nor accepted the label of a riddle. Feynman preferred to honestly accept not knowing.

And perhaps next summer, when you see a sprinkler in a yard quietly spinning around itself, remember that sometimes the simplest devices around us hide the most difficult questions within themselves.

From Other Members of the Pen Family

آب‌پاش چرخنده با پره‌های منحنی که جهت چرخش هوا را نشان می‌دهد
تصویر از کتاب مکانیک ارنست ماخ. او نوشته بود وقتی توپ لاستیکی توخالی فشرده می‌شود، هوا در جهت پیکان‌های کوتاه جریان می‌یابد و چرخ در جهت پیکان بلند می‌چرخد. وقتی توپ لاستیکی رها می‌شود، جهت جریان هوا معکوس می‌شود، اما ماخ «هیچ چرخش مشخصی» در دستگاه مشاهده نکرد.wikimedia commons

Destruction of one of Tehran's modern architectural jewels in Amjadieh neighborhood

ارنست ماخ با عینک و ریش بلند در یک پرتره سیاه و سفید
ارنست ماخWikimedia Commons

Trump: We have no immediate plans to attack Iran

ادوارد کروتز فیزیکدان برجسته با چهره‌ای خندان
ادوارد کروتزWikimedia Commons

A motorless and gearless machine; a masterpiece of traditional Iranian engineering

ساختار آزمایش آب‌پاش پرینستون (Princeton Water Sprinkler) با لوله‌های مسی S شکل
شماتیک آزمایش پرینستون؛ افزایش فشار در مخزن شیشه‌ای سرانجام ظرف را شکست، بی‌آنکه پاسخ روشنی به معما بدهد

Crucial hours for Trump; will the balance of power in Washington shift?

چهره جوان ریچارد فاینمن (Richard Feynman) با کت و پیراهن
Wikimedia Commons

Is Call of Duty: Modern Warfare 4 worth buying? Comprehensive game review

آب‌پاش‌های کشاورزی در حال آبیاری زمین‌های زراعی در یک دشت
pexels

50 horror films chosen by great cinema directors | From Tarantino to Scorsese - Filmzi

Why has the Tomb of Cyrus the Great survived without mortar?

Kimi Antonelli became the youngest Formula 1 world champion

A 2,500-year-old engineering masterpiece; how Xerxes built a bridge over the sea

15.5 million households did not even take a single trip

Specifications and price of Jaecoo J7 (Lucano L7) Plug-in Hybrid in Iran

For the day sanctions are lifted

Renewed division among Trump's inner circle over Iran

Fear of the cosmic unknown; an introduction to Lovecraftian horror

The secret of producing ice in the heart of the desert; the miracle of ancient Iranian architecture

Tags:
reverse water sprinkler Richard Feynman physics mystery fluid mechanics laser experiment
TahaSite

TahaSite

Since childhood, I have always been fascinated by computers and technology. My passion for programming and web development has driven me to constantly learn and create. At TahaSite, my goal is to provide the latest tech insights and build cutting-edge software infrastructure.

Comments

No comments yet. Be the first!