There are several theoretical mechanisms that are able to generate spatial patterns autonomously without any pre-pattern.
1,2 Among them, the most plausible in the biological system is the reaction-diffusion (RD) mechanism, which was first presented
by A. Turing in 1952
3,4 and mathematically refined by mathematical biologists.
5,6 In the model, the spatial pattern is made as stationary waves generated by the interactive RD of putative chemical substances.
According to the mathematical study using computer simulation, an RD system is able to generate stable and evenly spaced patterns
when the whole network satisfies a condition: “local activation and long-range inhibition.”
6 The spatial patterns made by the system (e.g., “RD pattern” or “Turing pattern”) do not need any pre-pattern and autonomously
regenerates when artificially disturbed.
However, in spite of its theoretical importance, until very recently Turing's theory was not widely accepted by experimental
biologists, for two major reasons. The first is that the main concept of the theory, “the pattern is made by wave,” is quite
unfamiliar to many experimental biologists. The second is the difficulty of showing the existence of “wave.” To prove that
the RD mechanism functions in a morphogenetic event, we need to show that the pattern possesses the dynamic nature of the
RD wave.