Abstract
Some numerical problems associated with the calculation of phase diagrams from thermochemical data are discussed. Two simple cases are considered. The miscibility gap in the Cr-Mo system is calculated using the free energies derived from calorimetric and vapour-pressure data. The liquidus and solidus curves for the Mg-Cd system were calculated in a similar way. These did not agree very well with the directly determined curves and a method is described for refining the estimated solidus curve using the measured liquidus curve. For both systems it is shown that unless care is taken in representing the thermochemical data spurious features appear in the phase diagram.