Trap

Dionaea muscipula has no brain, no nerves, nothing an animal would call a nervous system — and it still won't close on one touch. Brush a single trigger hair inside the trap once and nothing happens: a real electrical spike fires, and the trap stays open. Brush a hair a second time, within roughly thirty seconds of the first, and it snaps shut in about a tenth of a second, one of the fastest movements documented in any plant (Forterre, Skotheim, Dumais & Mahadevan, Nature, 2005). One touch isn't a signal worth acting on — it could be rain, a falling leaf, wind. Two touches close together are what a moving, worth-catching animal actually produces. The trap isn't just reacting. It's counting.

Suda et al. (Nature Plants, 2020) filmed the mechanism with a genetically encoded calcium sensor: each touch sends a calcium wave through the leaf, and closure only fires once that concentration crosses a threshold. A lone touch's calcium rises and then decays back toward rest. If a second touch lands while the first signal is still elevated, the two waves overlap and cross the threshold together. Wait too long, and the first wave has already faded below it by the time the second one arrives — no overlap, no closure, nothing carried over. That decay is the whole memory: not stored anywhere, just an ion concentration on its way back down, with roughly thirty real seconds before it's gone. Touch the hair below.

Closing isn't the end of the counting, either. A real catch keeps struggling once it's shut in, firing the same hairs again and again. Böhm et al. (Current Biology, 2016) traced what those later touches do: the third stimulus starts ramping jasmonate, the plant's wound hormone, and by the fifth the trap switches on sodium uptake in its glands and begins secreting the enzymes that actually digest the catch. A struggle that stops after two touches gets a trap that reopens, empty-handed, in about a day. A struggle that keeps going gets digested. The trap is reading effort, not just contact — still counting after it has already decided to close.

Honest gap: this room draws one trigger hair; a real trap carries about six, three to a lobe, and any two hits in any combination count toward the same threshold. The calcium bar's decay is drawn as a straight line for legibility — the cited papers establish the roughly-thirty-second cutoff and the overlap-a-threshold mechanism, not a published curve for the concentration in between, so a straight line is this room's own invented shape, not a measured one. The closing animation itself runs a few hundred milliseconds slower than the real ~0.1s snap, the same visibility trade /pod and /touch already make with their own plants' real speed. And the third, fourth, and fifth touches here are just clicks in a row; a real trap gets them from a live catch struggling for hours, not from a visitor's own hand a few seconds apart. No date, no rng() plant.js could ever touch — only a visitor's own touch and the clock, same discipline every by-hand room here keeps.

Sources: Suda, Mano, Toyota, Fukushima, Mimura, Tsutsui, Hedrich, Tamada & Hasebe, Calcium Dynamics During Trap Closure Visualized in Transgenic Venus Flytrap, Nature Plants 6, 1219–1224 (2020); Böhm, Scherzer, Krol et al., The Venus Flytrap Dionaea muscipula Counts Prey-Induced Action Potentials to Induce Sodium Uptake, Current Biology 26(3), 286–295 (2016); Forterre, Skotheim, Dumais & Mahadevan, How the Venus Flytrap Snaps, Nature 433, 421–425 (2005).