Abstract: The nature of the upper mantle below the ancient cratonic areas can be deduced by study of the xenolith suites in kimberlites. Studies on the proportions of xenoliths, together with their mineralogy and chemistry, suggest an upper mantle containing an upper harzburgite zone and a lower lherzolite zone, with both of these zones containing chemical and mineralogical variants, together with minor rock types such as glimmerites, MARID-suite rocks, pyroxenites, and eclogites. Isotopic studies of the phases in xenoliths have yielded restricted ranges of values for the isotopic composition of hydrogen, carbon, oxygen, and sulphur which are tentatively identified as the true isotopic values for these elements in the upper mantle. In addition, recent discoveries suggest that diamond may be a primary uppermantle phase.

The textures and fabrics of the xenoliths indicate that plastic deformation has taken place in connection with the intrusive kimberlite event, and also in earlier events unconnected with the kimberlite event. In addition, brittle fracture has been observed, this fracturing often being accompanied by the filling of the ensuing veins and joints by fluids that have crystallized potassium-, titanium-, and water-rich phases; limited metasomatism of peridotite wall rocks accompanies this vein filling and more widespread pervasive metasomatism may also be present.

Although most upper-mantle rocks are now metamorphic, in some rare instances there are relics of earlier rock types that have not been completely obliterated by subsequent metamorphic events; most of these could be attributed to an igneous origin and in most cases during subsequent metamorphism the original rock types have been subjected to increasing pressures and/or lower temperatures.

Although the source of most materials now present at the earth's surface can be directly attributed to some identifiable source within upper-mantle rocks, no source has yet been identified for CO₂, N, P, and the rare gases.