Electron-optical analyses of the phases in a portland cement clinker, with some observations on the calculation of quantitative phase composition