The angle that never quite repeats

A sunflower head packs its florets by turning through the same angle for every new one, over and over: about 137.5°. Helmut Vogel wrote down the exact formula for it in 1979 — one multiply per floret, no growth process needed to draw the finished pattern. I built /spiral around that formula, with the angle itself as a control, because the interesting part isn't that 137.5° works. It's how narrow the band is where it works at all.

Nudge the angle to 138° — visibly the same number, off by half a degree — and the interlocking double spiral a sunflower is famous for collapses into obvious straight arms with gaps between them. 138° is close to 23/60 of a full turn, so roughly every 60th floret lands almost exactly on top of an earlier one, and the pattern reveals the fraction hiding underneath it. Try 90° and you get four bare spokes radiating outward — the fraction there is just 1/4, as plain as it gets. The golden angle, 137.5077640500…°, is 360°×(1 − 1/φ), where φ is the golden ratio — and φ has the odd distinction of being the real number that every rational fraction approximates worst. Its continued fraction is an infinite run of 1s, the slowest-converging expansion there is. Practically, that means no run of florets spaced by the golden angle ever nearly closes a loop back to an earlier direction, at any scale you check. New florets keep landing in whatever gap is currently widest instead of stacking into a row. That's the packing efficiency real botanists actually measure, and it's a direct consequence of irrationality, not a coincidence discovered by counting spirals on enough sunflowers.

None of that says a meristem — the growing tip where new leaves or florets actually form — is doing trigonometry. It isn't. A 2003 Nature paper (Reinhardt, Pesce, Stieger et al.) confirmed the mechanism directly, by blocking it: each existing primordium depletes the plant hormone auxin in its immediate neighborhood, and the next primordium forms wherever auxin is currently most concentrated — which, given the ones already there, ends up being the widest remaining gap. Block auxin transport and phyllotaxis breaks down entirely; the paper's own proof that this is the mechanism, not just a correlate of it. Run that purely local rule — form where auxin is highest, deplete it there, repeat — for long enough, and the golden angle falls out on its own, an emergent consequence of a rule that never mentions 137.5° anywhere. Vogel's formula and a real meristem's auxin rule arrive at the same destination by completely different means: one places every floret from a global angle chosen in advance, the other never sees more than its own immediate neighborhood. The room draws the first one, because it's the one honest to build without a growth simulation behind it, and says so.

Sources: Vogel, A Better Way to Construct the Sunflower Head, Mathematical Biosciences 44 (1979); Reinhardt et al., Regulation of Phyllotaxis by Polar Auxin Transport, Nature 426 (2003); Bravais & Bravais, 1837, the first to connect phyllotaxis divergence angles to Fibonacci numbers.