When a lawn sprinkler runs backwards, sucking water in rather than spraying it out, which way does it turn? The question has quietly nagged physicists since the 1880s. Richard Feynman popularised it in the 1980s, even rigging his own ill-fated experiment that ended in an explosion. Now a team at New York University has supplied the clearest answer yet, not through elegant equations alone but by watching custom-made silly sprinklers spin in a tank of water.
The researchers designed sprinklers with loops, spirals and other eccentric geometries. They ran them in both forward and reverse modes, tracking every rotation, the internal currents, the external flows and the tiny torques produced. What they found is striking in its simplicity. The reverse sprinkler does rotate, but in the opposite direction to the forward mode and roughly 50 times more slowly. The driver is the angular momentum carried by water as it flows inside the arms.
The geometry of the sprinkler arms governs the conversion of mass flux to momentum flux, controlling the direction and magnitude of rotation.
That sentence captures the heart of the work. The shape of those arms decides exactly how incoming water changes direction and hands over its momentum. Measure it carefully and the puzzle falls into place. The same principle holds for ordinary S-shaped sprinklers and for the whimsical variants the team tested. Previous ideas, including one from Ernst Mach in 1883 that invoked opposing fluid swirls and another linked to Feynman that emphasised external flows around the arms, do not survive the data.
Leif Ristroph, associate professor at New York University's Courant Institute School of Mathematics, Computing, and Data Science and senior author of the study, put it plainly. "This work provides the experimental answer for Feynman’s Sprinkler Problem by showing, across several sprinkler types, how the angular momentum of water flows drives sprinklers’ rotation."
The paper appeared in the Proceedings of the National Academy of Sciences on 13 July 2026. It builds directly on the team’s 2024 experiments with a conventional sprinkler that first pointed toward the momentum flux explanation. This latest round extends the test to deliberately odd geometries, closing off loopholes and showing the mechanism is robust.
Ristroph again: "By showing that momentum flux is the answer to Feynman’s Sprinkler Problem, our findings address a long-standing open problem in flow physics and provide useful knowledge about how these devices work and their effectiveness." Brennan Sprinkle, assistant professor at Colorado School of Mines, added a forward-looking note. "Our findings provide a firmer understanding of how components respond to fluid flows—knowledge that can guide future engineering and technological advances for devices, such as turbines, that convert these flows into energy."