Along the forest streams of Central America, a lizard escapes a predator by doing something that should be mechanically impossible. It runs across the water's surface. Not floats — runs, bipedally, at roughly two meters per second, for distances of several meters before sinking and switching to swimming. Basiliscus basiliscus, the common basilisk, earns its local name: the Jesus Christ lizard.
The observation is ancient. The mechanism took until the 1990s to characterize precisely.
Three Phases, One Stride
The work of Glasheen and McMahon at Harvard in 1996 broke down what happens during each stride. Each foot goes through three distinct phases:
The slap phase: the foot strikes the water surface moving downward and slightly backward. The impact creates an air pocket — a cavity in the water — before the surface tension and hydrodynamic resistance close it. This phase generates an upward force from the momentum of the water displaced downward.
The stroke phase: the foot pushes through the water while the air cavity is still open. Because the cavity exists, the foot is pushing against water on one side and air on the other, which allows a large stroke surface to generate thrust without the foot needing to be entirely submerged. This is the thrust-generating phase.
The withdrawal phase: the foot exits the water before the cavity collapses. Pulling a foot out of a closed cavity would require energy to break the surface tension. The timing — withdrawing before collapse — avoids that energy penalty and allows a fast stride frequency.
The result: each stride generates a vertical force roughly 1.5 times the lizard's body weight during the slap phase, enough to keep the animal above the surface for the stride. The timing requirements are precise. Get the withdrawal wrong and the foot is fighting water on exit; get the slap wrong and the cavity doesn't form properly.
Why Speed Is Non-Negotiable
The physics are speed-dependent in a way that creates an engineering constraint. The upward impulse from slapping the water depends on the foot velocity at impact. Slow down and the impulse drops below what's needed to support body weight for the stride duration. This sets a minimum speed — roughly 1.5 meters per second — below which water-running becomes impossible and the lizard sinks.
Glasheen and McMahon estimated the power output required for water-running at roughly 10 times what's needed for land running at the same speed. Basilisks achieve this through burst fast-twitch muscle activity — the kind of anaerobic sprint that's unsustainable for more than several seconds. Longer flight distances exceed their power budget, and they transition to swimming.
The Fringe and the Flow
Look at a basilisk's hind foot and you'll see something distinctive: fringes of skin extending laterally from the toes. These aren't decorative. Hsieh and Lauder at Harvard used particle image velocimetry (PIV) in 2004 to track the actual water flow during basilisk strides. The toe fringes increase the effective foot area during the slap and stroke phases, enlarging the air cavity and improving force production without adding much mass or drag during the withdrawal.
The fringes are deployable — they're folded during the withdrawal phase and extend during the slap and stroke. This deployment is passive, driven by the forces of water contact rather than active muscle control. An elegant solution that costs nothing in additional neural complexity.
Why Humans Can't Do This
A question that comes up with some regularity: could a human water-run if they had large enough feet? The answer is no, and the reason is informative about the geometry of the problem.
Scaling matters enormously here. A small lizard's body weight relative to the upward impulse available from foot slaps falls within a range where water-running is possible. As body mass scales with volume (the cube of linear size) and force scales with area (the square of linear size), a larger animal requires disproportionately more force per unit foot area to stay up. Glasheen and McMahon estimated that a human-sized water-runner would need roughly 15 times the muscle power available in human legs — and feet large enough to generate sufficient slap force that sprinting would itself be impractical.
Some engineers have built artificial water-walking devices for humans — essentially boats with leg-driven paddles. They work but they're not biomimetic in the basilisk sense; they use a fundamentally different mechanism (sustained paddle thrust rather than slap-stroke-withdrawal strides).
Who Else Does This
Basilisks aren't alone, though they're the most studied example. The western grebe (Aechmophorus occidentalis) runs on water during its courtship rushing display — a brief sprint across the surface at high speed that can last several seconds. The mechanism appears similar in broad outline: rapid foot strokes that generate upward force from water displacement.
The convergent evolution here is striking. A lizard and a bird, separated by deep evolutionary history, both solve the same problem with a similar three-phase stride. What this suggests about the physics isn't surprising — there may be a narrow solution space for water-running in air — but it's satisfying as a biological observation. When different lineages independently discover the same answer, it usually means the problem has few answers.
There's something instructive in the withdrawal phase specifically. The energy-expensive step in any reciprocating mechanism is often the return stroke — the phase that repositions for the next power stroke. Basilisks solve this by timing the return to avoid the cost of closing cavities. It's the kind of optimization that looks obvious in retrospect and requires careful measurement to identify in the first place. Two meters per second across water, and the critical trick is knowing when to pull your foot out.
Maren covers biology and biomechanics at Anethoth. Follow what we're building at builds.anethoth.com.