Experimental Data Bearing on the Petrogenetic Links between Kimberlites and Carbonatites

The genetic relationships among kimberlites, carbonatites, and associated alkaline igneous rocks are being investigated by phase equilibrium studies in a number of silicate-carbonate systems, with special emphasis on liquidus studies. The precipitation of calcite from melts in the system CaO-CO_2- H_2O at temperatures near 650°, through a wide pressure range, was regarded by Wyllie and Tuttle as experimental verification for the magmatic origin of carbonatites. Addition of silicates has yielded significant results in the systems: CaO-SiO_2-CO_2-H_2O (J . L. Haas), CaO-MgO-SiO_2-CO_2-H_2O (G. W. Franz), albite-nepheline-CaCO_3-Ca(OH_2-H_2O (D . H. Watkinson), and a northitc-albite-Na_2-CO_3-H_2O (A. F. Koster van Groos). The heterogeneous phase relationships in these systems indicate that a variety of petrogenetic processes may be effective. There is a persistent thermal barrier on the liquidus of several systems which suggests that normal peridotite magmas are incapable of yielding a residual lime-rich carbonatite magma by fractional crystallization. However, there is evidence in other systems that crystallization of an alkaline magma may yield a residual carbonatite magma. Liquid immiscibilily has been discovered between some silicate melts and sodium carbonate melts, adding support to the suggestion that carbonatite magmas may be derived as immiscible fractions from parent alkali peridotite magmas. This fact also provides support for the hypothesis that a primary alkali carbonatite magma could form and persist without significant contamination by silicates until conditions (pressure, temperature, or composition) were reached where the immiscibility relationship ceased to exist. Mineral assemblages in the systems studied are closely comparable with those in many carbonatitc complexes. The wide distribution of melilite is notable.