Adsorption isotherms have been determined for mixed Ca-Na montmorillonites, with various values of f, the proportion of (½ Ca) to Na (Fig. 1). These are coincident from f=0 to f=0·3, and then diverge. This is interpreted as meaning that “demixing” occurs at f=0·3, a system being formed in which some interlayer spaces are occupied mainly by Ca, others mainly by Na.

X-ray analysis confirms this view. Plots of the water uptake against f, for various relative humidities (Figs. 2–5) show that, for R. H.'s lower than 75%, the demixing is incomplete.

The observations can be explained by the occurrence of a limited number of doubly-charged sites (by Al-for-Si exchange spots being accidentally contiguous). For the montmorillonite used, the statistical probability of occurrence of such sites is found to be 0·2 if the layers have symmetry C-2/m and 0·33 for C-2, so that the observations are in favour of the latter symmetry.

(1) Long spacings (29·6 and 26·7Å) are confirmed from certain clays from Japan. (2) These clays have very complicated clay mineral compositions including kaolin minerals, chlorite, sepiolite and montmorillonite. (3) Each long spacing can be explained by a regular interstratification between two kinds of lattices selected from the following ones: montmorillonite (cell ht. 15·4Å), a magnesian chlorite (14·5Å), sepiolite (12·7Å), and a new lattice (14·0Å); the last is a regular interstratitication of gibbsite and silicate (pyrophyllite unit) layers. The new lattice is proposed not only to explain the spacing but all other mineralogical properties of the clay. Thus the 29·6Å spacing agrees well with that of a 15·4/14·0Å super-lattice, and the 26·7 Å spacing with a 12·7/14·5Å super-lattice. (4) These long spacings are liable to be found in a clay in which several kinds of clay minerals co-exist, and such a complex clay tends to be found in special conditions, such as a boundary area between two different kinds of alteration zones, and/or an area affected by successive hydrothermal alteration.