Fig. 6

a Variation with temperature of the slope α of δCt(fgf8(t)) vs time (Fig. 5), which is linearly related to the exponential decay rate of <[fgf8m]>: 1/τ = ln(2) α. The red point at 27 °C is from PSM only data. The data can be fit (blue line) by α = a (T−Tc) with: a = 0.022 ± 0.0038; T_c = 14.4 °C (Χ² = 6.57; DF = 6). Inset: variation of the clock frequency f_s = 1/τ_s vs temperature² (data from Schröter et al.²) which can be fit by f_s = a_s (T − Tc) with a_s = 0.188. b Another way to represent the data shown in a: since δCt(fgf8 (t)) varies with temperature in a similar way as the segmentation period, τ_s, the δCt(fgf8(t)) data taken at different temperatures collapse on the same curve as a function of somite stage: s = t/τ_s (=number of segmentation periods). The data at all temperatures can be fit (blue line) by δCt = a′ s + b, with a′ = 0.122 ± 0.0098 and b = 3.955 (χ² = 26.5; DF = 55). Notice that as expected: a = a′∙ a_s. c variation with temperature of the growth rate (blue points) and the PSM shrinkage rate (red points) and linear best fits (blue lines): y = β(T − Tc) with β = 4.79 ± 0.46 (χ² = 6.35; DF = 2; blue points) and β = −5.08 ± 0.20 (χ² = 0.51; DF = 2; red points).