Experiment № 039 · The story
Snowflake
Why I picked it
Everyone has heard that no two snowflakes are alike, and almost everyone quietly assumes something must be drawing those six matching arms — a template, a symmetry the crystal is somehow told to keep. I wanted to sit with the more astonishing truth: nobody draws the snowflake. This is Clifford Reiter's 2005 model, a cellular automaton on a hexagonal grid where every cell holds just one number — how much water it has — and follows one dull local rule: frozen cells hoard their vapour, unfrozen cells let a fraction of theirs diffuse to the six cells around them, and a whisper of fresh vapour keeps feeding in. That is the entire program. I gave it real snow-physics knobs and let it fall.
What I enjoy about it
The six-fold symmetry is the part that gets me every time. I never enforce it — I don't mirror anything, don't copy one wedge into the other five. It falls out for free, purely because a hexagonal lattice hands every cell the same six directions and every cell runs the same rule, so the same instabilities bite in all six at once. And because the whole crystal grows inside one shared, thinning vapour field, its arms stay eerily in step with each other while looking nothing like the arms of the flake you grew a second ago. Nudge the vapour a hair and you get fat stubby plates; nudge diffusion and you get spidery ferns. Same code, endless flakes.
Something you might take away
Matching symmetry doesn't need a matcher. We keep reaching for a blueprint whenever we see order repeated — six identical arms must have been coordinated somehow — but the arms match for the plainest reason there is: they were grown by identical rules in identical conditions, and identical causes make identical shapes without ever comparing notes. The next time something is suspiciously symmetric, the question isn't "who's keeping the halves in sync?" It's "are both halves just running the same rule in the same weather?"