Interscience publishers, inc., New York, 1960. - 1339 p. In this third volume, we have collected all the numerical data puhlished about concentrated
hinary mixtures with at least one metallic compound. As in the former volumes, there are in some cases considerable discrepancies between
the data given by different authors; this is due not only to some impurity of the chemical samples hut also to the use of experimental methods or apparatus which are not aluays adequate, For example, even if the thermome ters are accurately calibrated, some authors made the mistake of not taking into account the correction for the emergent column, which is positive or, for low temperatures, negative; with electrical thermometers, where the accuracy can be much greater, there are errors connected with the temperature coefficient of the resistant coils, if the thermostat is not very good: in viscosity measurements, the method of Poiseuille is accurate only if the diameter of the capillary, relativet o its length, is adequate; and so on. Today, all physical chemists who want to make accurate measurements should test their physical methods with great care.