The mechanism for the slow rate of reaction between portland cement and water during the early stage is not well understood, but it probably is controlled by either the rate that the reactants diffuse through a barrier that surrounds the unreacted cement grains or by the rate that nuclei of the stable product form and grow or by both rates. New evidence using environmental scanning electron microscopy is presented about the structure of a layer that forms around the particles of cement. Preliminary observations that relate mixing to the structure of the layer are also presented.