With 23% of metakaolinite the mixture of expansive cement really contained a substantially smaller quantity of gypsum rock than the theoretically possible formation of ettringite should anticipate. The charged proportion of gypsum rock may rather cause the formation of monosulDhate, with some reserve being present. When more gypsum rock and other sulphates are added, there takes place, as it follows from the Fig. 2 of our paper, an uncontrolled expansion which may lead even to a total destruction of samples after a longer period. Evidently the authors had to avoid this phenomenon.
The fact that the superfluous AI203 in the mixture will cause the later transformation of ettringite to monosulphate was also originally expected by the authors. This was also supported by the appearance of cement paste microstructure after a longer period of hydration, when by SEM findings the needle-like crystals of ettringite totally disappeared and the systems resembling monosulphate in their morphology were formed. Following the x-ray analysis no monosulphate was identified and there was no diffraction shift from 9,1 ° to 9,3 ° . The determined diffraction values correspond to the table values of ettringite (i.e. also in the samples of long hydration period).
The calculation of possible link to monosulphate carried out by Mehta is quite correct. Only the x-ray findings speak against its validity in practical samples.
As far as the method applied for the determination of expansion is concerned, Mehta rightly guessed that the free expansion of hardened cement pastes deposited * CCR 4, 113 (1974).