“…Having gained con®dence in our methodology, we have decided to apply it to a much more demanding subject: the case of a ternary alloy. In a previous study (paper I), we had shown that the Cu±Ni±Fe system is an interesting case: the decomposition follows the tie-line of the calculated phase diagram (Chart et al, 1982;Servant et al, 2000), in agreement with atomic probe measurements of the solute concentrations (Wagner et al, 1984;Lopez et al, 1993). We have also shown, however, that this description is a rough ®rst approximation: this system does not simply decompose into well de®ned precipitates (enriched in Ni and Fe), embedded in a Cu matrix, itself depleted in solute elements.…”