Whirl

A maple seed doesn't fall so much as fly. Cut loose from the branch, a samara — the papery, one-winged fruit hanging in clusters in a maple canopy — doesn't tumble and doesn't glide either. It spins, nose down, around a nearly vertical axis, at a steady rate, sinking at a nearly constant, slow speed the whole way down: a tiny rotor that gave up on going anywhere and just concentrated on staying up. Norberg named the behavior in 1973 — autorotation, a passive, self-stabilizing spin with no muscle behind it, that a samara falls into on its own the instant it's released, and never falls out of.

The point of that slow, steady spin isn't the spin. It's the time. A samara held up by autorotation stays airborne many seconds longer than the same mass falling any other way, and every extra second aloft is another chance for whatever crosswind happens to be blowing to carry the seed a little further from the tree that's shading it. Schaeffer, Truman, Truscott & Dickerson trimmed the wings of 160 real samaras from eight maple species by measured amounts and clocked what happened next: descent speed climbs as a clean power law of the wing area left behind, and once a wing has lost somewhere between a fifth and two-fifths of its area — the exact edge varies from seed to seed — autorotation collapses outright and the seed can no longer hold a stable spin at all. Try it below.

The formula behind the slider is the study's own: V_d/V_d0 = (A/A₀)^−0.79 (R² = 0.91), where V_d0 is the baseline descent speed of 0.83 m/s the study measured across all 160 intact seeds and A/A₀ is the fraction of wing area left. Trim the wing here and that's the exact curve deciding how much faster it falls — not a straight line, a real measured exponent, so the first 10% you trim costs less speed than the last 10% before it fails. Every release also rolls its own failure edge, once, from the paper's own stated range (60– 80% of the original area, since individual seeds gave out at different points): trim past that edge and the spin doesn't survive the drop. Baseline spin, while it holds, is the study's other measured number: 18.8 rad/s, about 3 full turns a second.

Honest gaps. The V_d0 and the −0.79 exponent are exactly the paper's own numbers; the failure edge is drawn once per release from its own stated range, not invented. Three things here are not from the cited sources, and are flagged as such: the branch height (4m, a plausible low limb, not measured), the crosswind (a steady, gentle 0.6 m/s, so the room can show wind actually doing something rather than nothing), and the fall speed once a seed tumbles (∼3.2 m/s, a rough stand-in for “about as fast as a compact wingless seed of similar weight,” not a number either study reports — they measured that autorotation fails past a point, not how fast the resulting tumble falls). The wing's own drawn length shortens in step with the trim so it reads as cut, which is a simplification of area lost, not length; the physics only ever uses the area fraction. And real autorotation spins around a near-vertical axis, traced from above as a cone — drawn here, like every side-on scene on this site, as a spin in the picture plane, the same liberty a falling leaf or a wind-blown flag already takes anywhere else here.

Sources: Norberg, “Autorotation, Self-Stability, and Structure of Single-Winged Fruits and Seeds (Samaras) with Comparative Remarks on Animal Flight,” Biological Reviews 48(4):561–596, 1973, for naming and first describing autorotation. Lentink, Dickson, van Leeuwen & Dickinson, “Leading-Edge Vortices Elevate Lift of Autorotating Plant Seeds,” Science 324(5933):1438–1440, 2009, for the aerodynamic mechanism — more in the field note. Schaeffer, Truman, Truscott & Dickerson, “Maple Samara Flight is Robust to Morphological Perturbation and United by a Classic Drag Model,” Communications Biology 7:248, 2024, for the descent-velocity formula, the baseline numbers, and the wing-ablation failure range this room's trim control is built from.