A review and assessment of experiments on Kimberlites, Lamproites and Lamprophyres as a guide to their Origin

Liquidus experimental studies on kimberlite, lamproite and lamprophyre compositions are reviewed with respect to the information they carry on the mantle origin of these rock types. This information is coupled with melting experiments on peridotite in the presence of H₂O and mixed H₂O+CO₂ volatile species. The origin of most lamproites is explained by the melting of mica-harzburgite assemblages at depths ranging from 40km for leucite lamproites to more than 150km for olivine lamproites. Clinopyroxene-rich, silica-poor lamproites remain enigmatic, but are possibly derived by the melting of a mica-bearing ultramafic source richer in clinopyroxene and under more oxidized, CO₂-bearing conditions. There are insufficient experimental studies on kimberlite to reasonably constrain their origin, and what remain are only general indications of the compositions of partial melts of mantle under volatile-bearing conditions. Melt compositions are not sufficiently well known to prevent very conceptual use of melt 'names' such as 'kimberlitic' or 'carbonatitic', and melts similar to alkaline and ultramafic lamprophyre may be hidden under this shroud. Clearer definition of the origins of alkaline melt compositions such as kimberlites and various lamprophyre types badly needs more exact bracketing of melt compositions of a variety of possible mantle mineral assemblages. The recently-developed sandwich reversal technique is ideally suited to study small degrees of partial melting, and could usefully be applied to lherzolitic and non-lherzolitic materials with hydrous and/or carbonate minerals.