The sorption of boron by clay minerals from natural waters has been studied experimentally. The quantity of boron sorbed per unit weight of clay mineral is dependent on both the salinity and the boron content of the solution. Previous work has shown that illite is the best clay mineral sorbent, though kaolinite and montmorillonite do sorb some boron and this is confirmed by the present work. The new experimental results demonstrate that the amount of boron sorbed by illite is not affected by the original boron content of the mineral and suggest that the process of incorporation of boron into the lattice proceeds in two stages. The bearing of these experimental results on the use of boron as a palaeosalinity indicator in sedimentary rocks is discussed and this leads to the conclusion that the rate of sedimentation may also influence the boron content of such rocks.

The amounts of potassium retained, after exchange with normal ammonium acetate solution at pH 7, were measured with six montmorillonite-beidellite minerals, whose interlayer charge ranged 0·72-1·00 M+/Si8O22 and where tetrahedral substitution of Al for Si ranged 0·14-0·93 M+/Si8O22. The amount of potassium retained increased with increasing interlayer charge, but there was no correlation between the amount of potassium retained and the tetrahedral charge.

The mineralogy of the clay fractions of two soil profiles representing the end-members of a catena developed on a glacial till derived from basic lavas has been determined. Particular attention has been given to the assessment of the nature of the amorphous inorganic material in the clay fraction of these soils. Chemical dissolution techniques were used and their effects on the clay fraction were followed by X-ray diffraction, differential thermal, infrared absorption, electron-optical and surface area measurements. The principal conclusion is that the soil clays are a continuum from completely disordered, through poorly ordered to well crystallized material.

The removal of the alkali-soluble fraction from soil clays has been found to influence markedly the efficiency with which iron oxides can be removed from such clays. In clays pretreated with 5% Na2CO3, up to 40% more iron oxide was extracted by dithionite (Na2S2O4) than from soil clays treated with this reagent alone. Chemical analysis, electron microscopy, X-ray and differential thermal examination, and specific surface area and cation-exchange capacity measurements indicated an improved dispersion brought about by the removal of aluminosilicate binding material by Na2CO3. The electron-optical studies also showed that the dithionite-soluble iron was principally in the form of small granules. These were amorphous to X-rays and electrons and chemical analysis of the dithionite-soluble fraction suggested that they were ferruginous complexes containing considerable quantities of silica and alumina.

Mineralogical and chemical variation in a 24 ft wide argillized zone in basic hornfelses from Gwavas quarry, Cornwall, is described and compared with variation in unaltered wall rock hornfelses.

Hydrothermal minerals developed are well-ordered kaolinite, sericite, chlorite, quartz, and limonite. Biotite has replaced hornblende as the dominant mafic constituent in the argillized zone. With increasing argillization kaolinite progressively replaces plagioclase, biotite, chloritized biotite, quartz, and finally Fe ore. No montmorillonite is present.

From the argillized zone, Si, Na, Ca, Mg, and Fe were extensively leached, although an approximate equivalent of H was added to compensate for the loss. Al and K are relatively constant throughout the zone. F was expelled from the most heavily argillized areas and is now concentrated in the biotite of the zone's margin as a marked ‘fluorine front’.

Two tenuous kaolinite-sericite zones are observed which are thought to be due to local variation in a single hydrothermal pulse with time.

Hydrothermal alteration of the hornfelses was effected in two stages (1) K+ metasomatism (biotitization) and (2) H+ metasomatism (kaolinization and sericitization).

Varty & White's application of multivariate analysis to Grim & Kulbicki's measurements on montmorillonites has been re-examined and extended. Inconsistencies between their table of scored data and derived similarity table, and some unexplained errors in the similarity table, have only a small effect on their results. Similarities calculated from Grim & Kulbicki's data, using a different similarity coefficient, lead to a two dimensional representation which separates the groups of montmorillonites more clearly. The groups can be clearly separated even in a one dimensional representation, by changing the relationship between distance and similarity.