Reply to Comments by Fergus G. F. Gibb and C. Michael B. Henderson on 'Mafic-Ultramafic Sills: New Insights from M- and S-shaped Mineral and Whole-rock Compositional Profiles'

Abstract

The comments by Fergus Gibb and Michael Henderson on a recent and some previous publications are most welcome and I am glad of the opportunity to clarify some of the issues they raise. Recent studies of chemical zonation in mafic dykes (Chistyakova & Latypov, 2009a, 2009b, 2010, 2011, 2012), dolerite sills (Latypov & Egorova, 2012; Egorova & Latypov, 2013) and layered intrusions (Latypov et al., 2011; Egorova & Latypov, 2012a, 2012b) have provided several new and important constraints on the initial stages in the development of intrusive bodies. In particular, we have found that the An content of plagioclase cores in the marginal zones of mafic sills and layered intrusions systematically increases inwards, indicating that these zones were not formed from a single pulse of parental magma of uniform composition. Rather, they resulted from an early stage of filling when the inflowing magma became increasingly more primitive. This and other findings resulted in an abandonment of Soret differentiation, a hypothesis that I had supported earlier for the origin of marginal reversals (Latypov, 2003a, 2003b), in favour of the ‘leading-edge fractionated magma’ hypothesis (see Latypov, 2014). This is quite an old concept in petrology that implies the initial emplacement of batches of evolved magma produced by three principal mechanisms: (1) the early stages of fractional melting without complete mixing in the mantle source; (2) magma stratification in a subjacent chamber; (3) fractional crystallization of magma while flowing through conduits. In accepting this hypothesis as an explanation for marginal reversals I referred to Morse (1979, 1981, 2008), who has been supporting this idea for many years for layered intrusions. I fully agree, however, that it was a grave omission not to mention the work of Gibb and Henderson, who proposed essentially the same process for mafic^ultramafic sills (Gibb & Henderson, 1978, 2006; Henderson & Gibb, 1987; Henderson et al., 2000). References to their publications would certainly have made our case even stronger. It should be noted that the conclusion about the formation of basal reversals by inflowing magma that gradually becomes more primitive is not trivial. No absolutely unequivocal evidence for this process has been forthcoming from petrological studies (including those by Gibb and Henderson) of mafic^ultramafic sills owing to their enrichment in olivine. This is because the primary composition of olivine is commonly altered to varying degree by reaction with interstitial melt, making this mineral unreliable for deciphering igneous processes. It is therefore understandable that most workers considered such sills as having formed from a single pulse of magma followed by gravitational accumulation of olivine (and/or pyroxene), either already present in the magma at the time of emplacement (Marsh, 1996; Ubide et al., 2012) or that grew in the intrusion itself (Frenkel’ et al., 1989; Worster et al., 1993; Ariskin & Yaroshevsky, 2006; Lo¤ pez-Moro et al., 2007). This crystallization model suggests that an upward increase in whole-rock MgO in basal reversals is a consequence of a decreasing solidification rate that ensures an increase in the amount of olivine able to settle to the bottom of the magma. In other words, basal reversals are considered to be mixtures of melt and cumulus olivine