Morphogenesis
In 1952, Alan Turing proposed that biological patterns — leopard spots, zebrafish stripes, fingerprints — arise from simple chemistry: an activator that promotes its own production and a faster-diffusing inhibitor that suppresses it. This instability spontaneously breaks spatial symmetry and creates structure from noise.
∂A/∂t = DA∇²A + f(A,I) ∂I/∂t = DI∇²I + g(A,I)
A uniform steady state can be destabilized by diffusion — counterintuitively, diffusion creates patterns rather than smoothing them. The key requirement: the inhibitor must diffuse faster than the activator. Small perturbations at certain spatial wavelengths grow exponentially, selecting a preferred pattern scale.
This simulation uses the Gray-Scott reaction: U + 2V → 3V (autocatalytic), V → P (kill). The feed rate f replenishes U; kill rate k removes V. The ratio DA/DI controls pattern wavelength. This deceptively simple system generates spots, stripes, spirals, worms, and self-replicating spots.
Turing's mechanism has been confirmed experimentally in fish skin (zebrafish stripes require specific diffusing signals), mammalian digit formation, sea shell patterns, and hair follicle spacing. The 2012 discovery of the Nodal-Lefty signaling system — a molecular activator-inhibitor pair — provided direct biochemical validation of Turing's 1952 prediction.
Different values of feed rate f and kill rate k produce dramatically different patterns. Near the lower-left: stable spots. Toward the diagonal: worms and labyrinths. Upper regions: holes and inverted spots. The boundary between regions shows complex transients. Biological systems are thought to sit near phase boundaries for robustness and evolvability.