Cone

Every other room in this garden moves because something in it is alive: turgor pressure shifting water between living cells on /touch, a circadian clock ticking in living tissue for a nyctinastic or heliotropic bloom, an insect's own nervous system timing a flash on /fireflies. A mature pine cone's scales move too, and by the time a cone is old enough to do it, every cell in those scales is already dead. The motion is passive — no nerve, no muscle, no metabolism, nothing that needs to still be alive to keep working. It runs on water content alone, the same physics that swells a wooden door shut in a humid summer and leaves it rattling loose come a dry winter.

Dawson, Vincent & Rocca worked out the mechanism in 1997: each scale is a bilayer, one side that swells more than the other when wet, so the whole thing bends the way a bimetallic strip bends to heat — except this one answers to humidity, not temperature, and it keeps answering for as long as the tissue holds together, long after the cell that built it died. Wet closes the scale; dry opens it, flaring the cone wide enough to let its seed catch the wind on the one kind of day that's actually good for dispersing it. Drag the slider below and watch it happen.

Notice the two directions don't move at the same speed. That's not an accident of the animation — it stands in for something real. Eger et al. measured the actual force a wetting scale generates as it closes against the force a drying scale generates reopening: about 1.3 newtons swelling shut versus about 0.9 newtons drying back open, in the same 2022 study. Closing genuinely pulls harder than opening pushes. That paper measured force, not speed, so turning a bigger force into a faster animation is this room's own liberty — but the direction of the asymmetry is real, not invented for effect.

Drag past about 15% RH and the cone marks its seed as released — a small, permanent thing, once. Real dispersal timing isn't incidental: a seed let go on a dry, breezy day can catch real distance on the wind; one released into rain just drops straight into the mud under its own parent tree, the worst possible place for it to end up. The bilayer mechanism is, among other things, a crude weather sensor with no nervous system attached to it at all.

Honest gap: real scale motion is nowhere near this fast. Eger et al. measured about 20 minutes for a scale to reach swelling equilibrium at a fixed humidity, and about 7 hours for full closure across a jump from 30% to 80% RH — this room compresses that into one to three seconds so a browser tab's patience is enough, exactly the liberty /touch already takes with a different plant's own timing. The slider also treats humidity as a single instant value a scale answers to the moment it lands; real hygroscopic tissue is a continuous cross-section whose response depends on more than the humidity right now, and the force asymmetry above is this room's only nod to that, not a full model of it. The scales themselves are drawn as one flat column of pairs, the same simplification /touch's own leaflet row makes for a different compound structure — a real cone's scales spiral around its own axis in the identical golden-angle packing /spiral already draws for a sunflower head, not two flat columns facing each other. And the seed-release moment is entirely stylized: a real seed needs an actual gust to shake it loose and carry it once the scale bares it, not just a crossed threshold.

Sources: Dawson, Vincent & Rocca, How Pine Cones Open, Nature 390, 668 (1997); Reyssat & Mahadevan, Hygromorphs: From Pine Cones to Biomimetic Bilayers, Journal of the Royal Society Interface 6(39), 951–957 (2009); Eger, Horstmann, Poppinga, Sachse, Thierer, Nestle, Bruchmann, Speck, Bischoff & Rühe, The Structural and Mechanical Basis for Passive-Hydraulic Pine Cone Actuation, Advanced Science 9(20), 2200458 (2022).