How two diffusing chemicals create the patterns of life
In 1952, two years before his death, Alan Turing published his final paper: The Chemical Basis of Morphogenesis. He asked a question that had puzzled biologists for centuries: how does a uniform ball of cells develop into an asymmetric organism? How does an embryo know where to put the head?
His answer was startlingly simple. Two chemicals, diffusing at different rates and reacting with each other, can spontaneously create stable patterns from a uniform mixture. No blueprint required. Just chemistry and diffusion, playing out according to simple rules. He called these reaction-diffusion systems, and the patterns they produce — spots, stripes, spirals, labyrinths — are now known as Turing patterns.
The paper was largely ignored. Biologists found it too mathematical; mathematicians found it too biological. Decades later, Turing was vindicated. His patterns were found on the skins of angelfish, in the arrangement of hair follicles, in the ridges of your fingertips, and in chemical reactions in petri dishes. The mathematics of two diffusing chemicals turns out to be a fundamental language of pattern formation in nature.
The system has two chemicals: an activator (V) that promotes its own production, and a substrate (U) that the activator consumes. The substrate diffuses faster than the activator — and this difference in speed is the key to everything.
Imagine dropping a spot of activator into a uniform field of substrate. The activator consumes nearby substrate and produces more of itself. But the substrate, diffusing faster, rushes in from the surrounding area to fill the gap — except it can't quite keep up at the centre. The result: a peak of activator surrounded by a depletion zone, surrounded by fresh substrate. This is diffusion-driven instability. Diffusion, which normally smooths things out, here creates structure.
The simulation below uses the Gray-Scott model, which captures this with two parameters: the feed rate F (how fast fresh substrate is added) and the kill rate k (how fast the activator decays). Small changes produce dramatically different patterns.
The terms: D∇² is diffusion (chemicals spread out). UV² is the autocatalytic reaction (where U and V meet, U is consumed and V is produced). F(1−U) feeds fresh substrate. (F+k)V removes activator.
Click or drag on the canvas to add activator chemical. Use the presets to explore different pattern types, or adjust F and k manually to discover your own.
The Gray-Scott parameter space is a landscape of pattern types. Each region produces distinct forms, many of which have striking parallels in biology. Click any card to load its parameters into the simulation above.
Turing never saw his theory confirmed. He died in 1954, and the first compelling experimental evidence didn't arrive until the 1990s, when Shoji Kondo and Rihito Asai showed that the stripe patterns on marine angelfish actively maintain themselves through exactly the mechanism Turing described. When the pattern was disrupted, it regrew — not from a template, but from the local chemistry.
Since then, Turing patterns have been found in the spacing of hair follicles, the arrangement of feather buds, the ridges of the hard palate, the tooth-like scales of sharks, the distribution of vegetation in arid ecosystems, and the branching patterns of lung airways. The Belousov-Zhabotinsky chemical reaction, discovered independently in the 1950s, produces the travelling waves and spirals that Turing's equations predict — visible proof that chemistry alone can create spatial order.
The deep lesson is about emergence. Complex, ordered patterns don't require complex instructions. Two simple rules — react and diffuse, at different rates — are enough. The pattern is not encoded anywhere; it emerges from the interaction. This is perhaps the most beautiful idea in mathematical biology: that form is not imposed from above, but grows from below, out of the blind mathematics of molecules bumping into each other.