The term colloid is now generally applied to any system which contains a large area of phase boundary per unit volume. There can be no doubt that clay minerals rank as colloids and it is very desirable that clay mineralogists and colloid chemists should be brought together as in the present meeting.

The skill displayed by clay mineralogists in overcoming the technical difficulties with which they were confronted has been remarkable, and the benefits derived from their painstaking investigations are already being felt outside the immediate field of their studies and particularly in other branches of colloid science. To have precise information about the dimensions and crystal structure of the particles in a colloidal system can make all the difference between success and failure in distinguishing between alternative explanations of colloidal phenomena.

The mode of formation and the particle size distribution of potters' clays, particularly china clays and ball clays, were discussed in this paper, emphasis being laid on the proportion of colloidal material present. From this information, it was found possible to estimate the proportion of colloid in pottery bodies. After a brief indication of the structure of the clay minerals involved, the colloidal properties of clay were considered in some detail, particular attention being paid to the development of the Gouy-Freundlich diffuse double layer and the effect of exchangeable ions on this.

The theory of deflocculation of clay slips is based on the diffuse double layer and the development of an adequate potential. The anions are considered to be located on the surface of the particles which have associated lyospheres the size of which is dependent on the cation present.

From this investigation it was hoped to obtain information about the position of the exchangeable cations. That they take up a position between the layers of the mineral seems obvious from the varying c spacing of dried montmorillonite saturated with different ions, but in the complexes with water and organic liquids it was not known whether the ions remained close to the layers or were removed from them. It was also hoped that the Fourier syntheses would give some information about the actual layer structure of the mineral (montmorillonite) and thereby lend support to one of the two suggested structures, those of Hofmann, EndeU and Wilm on the one hand and of Edelman and Favejee on the other.

The montmorillonite used was a sample of Wyoming bentonite separated on the Sharples supercentrifuge. In order to get as large a contrast as possible between the two syntheses it was decided to use Li+-saturated montmorillonite and Ti+-saturated montmorillonite.

The hydroxides of the divalent metals have an octahedrally coordinated structure of hexagonal symmetry typified by the “cadmium hydroxide” structure, familiar to clay mineralogists as the “brucite” structure. The weakly electropositive metals form insoluble or slightly soluble basic salts in which the ratio increases as the initial concentration of the unhydrolysed salt is decreased. A pseudo-Cd(OH)2 structure is formed when this ratio is of the order of 20 to 25 in which the octahedral sheets are separated by a constant distance proportional to the radius of the anion of the unhydrolysed salt. The resultant characteristic interplanar spacing is stabilised by a very small number of anions, the remaining interlamellar space being filled with amorphous hydroxide. The latter is known as the “α -zinc hydroxide” structure since zinc shows its similarity to its fellow transitional group metals by forming basic salts with this structure (Feitknecht, 1938).

Two suggestions as to the structure of sorbed water on clays have received considerable attention. Macey (1942), arguing from the similarity between the structure of ice and of the oxygen atoms exposed at a sheet surface, has suggested that the first layers of such water might have a structure similar to that of ice. Hendricks and Jefferson (1938), on the other hand, suggested the well-known “hexagonal-net” structure, a concept which was later extended to montmorillonite (Hendricks, Nelson and Alexander, 1940), and which has recently been used increasingly in the interpretation of the properties of clays (Grim and Cuthbert, 1945; Grim, 1942; Mering, 1946).

Some difficulties in the application of these theories to montmorillonite have already been pointed out by Williamson (1947-48), but some other aspects of the problem appear to merit further consideration. For example, there is an obvious space problem which is influenced by the position of the exchangeable cations and their frequency of distribution in the space available for water.

While the writer was working on an unusual form of calcium carbonate usually found in caves, he collected some material closely resembling it from an adit of the disused Odin Lead Mine, Near Castleton, Derbyshire.

This material was pure white in colour, tending in places to translucency. It had a S.G. of 2.125 and a R.I. of about 1.5, though determination of the latter was rendered uncertain on account of included water. It took the from of a rippled flow covering the Carboniferous Limestone walls and roof of the passage. It was very soft, being obviously in the process of formation, though in places it had dried to form a slightly resistant encrustation.

As both attapulgite clays and some montmorillonite clays are termed “fuller's earths” in the literature, a comparison is made between the physical properties of the two groups of material, with special reference to those which are important commercially.