Comments on the paper "Megacryst suites from the Lekkerfontein and Uintjiesberg kimberlites, southern Africa: evidence for a non-cognate origin"

Abstract

Zoned rims to garnets in sheared peridotites have been ascribed to reaction with the liquid that crystallized the megacryst suite (Griffin et al. , 1989; Moore and Lock, 2001). To preserve such zoned margins, this magma must have been present immediately prior to, or at the time of kimberlite eruption (Griffin et al. , 1989). This conclusion is supported by the isobaric range in megacryst equilibration temperatures in individual pipes (Gurney et al. , 1979). It is also consistent with U/Pb ages for megacryst zircons, which typically span an interval from the time of emplacement of the host kimberlite to some 5–10Ma earlier (Moore et al. , 2008). Megacrysts thus offer potential keys to understanding the significance of the sheared peridotites in terms of mantle composition and structure, as well as the processes involved in the generation of kimberlite magmas. Accordingly, the detailed characterization of the chemical signatures of populations from the Lekkerfontein and Uintjiesberg kimberlites (Merry and Le Roex, 2007) is a very welcome addition to the database of this important suite. Nevertheless, it highlights a persistent conundrum that must be addressed. REE modelling, using published partition coefficients for clinopyroxene and basaltic magma compositions argues against crystallization from the host kimberlite magma. However, this conflicts with field, petrographic, mineral chemistry and experimental data, summarized below, which appears to demand a cognate origin ( e.g. Jakob, 1977; Eggler and Wendlandt, 1979; Kotrovitsky et al. , 2004, Moore and Belousova, 2005 and Rankenberg et al. 2005. ### Field evidence Magmas such as meimechites (Moore et al. , 1992) and basanites and …