Fig. 2.

The Ten-m oscillator. In the mathematical model, Eqs. (3,4), k ₂ is varied from 0.75 to l/100th of this, simulating control from FGF8. The Hill coefficient $\gamma$ is 2, and the other Hill coefficient m is 1.25, initially, approaching $\gamma$ when k ₂ goes down. k ₃ is taken initially to 0.49, close to k _3c (Eq. 15), and when k ₂ varies by two orders of magnitude, k ₃ varies with a modest factor 3.5. The result is a relatively rapid oscillation at first, with T _p = 10.5, but gradually, the period grows and reaches values of ~ 10 times larger than in the beginning. In experimentally observed gene expression waves, the period grows with a factor of about 6, after which the oscillation is trapped by a bistable system