Somites, the precursor of vertebrae, are formed in the presomitic mesoderm (PSM), which consists of cells showing oscillatory signalling activity. The PSM is elongated in space, with somites forming in one end, the anterior PSM, which is opposite to the posterior PSM. Recent experiments in mice (in particular experiments conducted by the Aulehla group in Heidelberg, Germany) have shown that the oscillatory signalling activity organise in such a way, that waves of signal seem to be travelling from the posterior PSM to the anterior PSM. Experiments have also indicated that exactly one full wave can be seen in the PSM at any time, and that a period gradient is present in the PSM – the shortest period being in the posterior PSM. In this talk, we present a minimal model of the coupled cells in the PSM, that reproduce central results from the experiments. We also investigate several of the experimental findings analytically, by approximating the cells as weakly coupled oscillators. Do we know the shape of the individual oscillators? What are the consequences of a period gradient in the PSM? Why does it take 20h of “growth phase” before somites are formed in experiments? Can we alter the width of somites? Exploiting our knowledge of weakly coupled oscillators, we answer these questions and link the period gradient to several of the experimental findings, indicating that the PSM cells may be considered weakly coupled oscillators.