Prediction of a unique sequence of metasomatic zones that would develop by intergranular diffusion with local equilibrium is possible only for relatively simple systems, unless extensive thermochemical and kinetic information is available. The complexity of the problem for a given example will depend on what portion of the set of chemical components required to describe the example are ‘diffusing components’, that is, components that move relative to ‘inert markers’. Diffusing components are commonly K-components (Thompson, 1970) for the various local equilibria of a sequence of metasomatic zones, since diffusion tends to impose a monotonie variation of the chemical potentials of these components across the zones. The number of diffusing components may vary from zone to zone in a particular example, as may the number of diffusing components that are K-components. Calculation of the rate of growth of a specific sequence of zones is relatively straightforward only for cases where the zones are primarily due to the variation of the chemical potential of one independent diffusing component. Calculation of the material transfer involved in the growth of a sequence of zones, assuming a single sharp initial contact is meaningful only if ‘inert markers’ or a discontinuity in the otherwise-constant ratio of two components indicate the present location of the initial contact. Examination of some natural calc-silicate diffusion zones suggests that a diffusion-imposed gradient in the chemical potential of calcium is largely responsible for the observed zonation. Metasomatic zones developed at the boundaries of ultramafic bodies, however, are produced by diffusion-imposed chemical potential gradients of several components, notably silica and magnesia, the number varying from zone to zone.