Convection and crystal settling in sills

Abstract

It has been advocated that convective and crystal settling processes play significant, and perhaps crucial, roles in magmatic differentiation. The fluid dynamics of magma chambers have been extensively studied in recent years, both theoretically and experimentally, but there is disagreement over the nature and scale of the convection, over its bearing on fractionation and possibly over whether it occurs at all. The differential distribution of modal olivine with height in differentiated alkaline basic sills provides critical evidence to resolve this controversy, at least for small to medium-large magma chambers. Our own and others' published data for such sills show that, irrespective of overall olivine content, modal olivine contents tend to increase in a roughly symmetrical manner inwards from the upper and lower margins of the sill, i.e. the distribution patterns are more often approximately D-shaped rather than the classic S-shape generally ascribed to gravity settling. We concur with the majority of other authors that this is an original feature of the filling process which has survived more or less unchanged since emplacement. We therefore conclude that the magmas have not undergone turbulent convection and that gravity settling has usually played only a minor modifying role since the intrusion of these sills. We offer a possible explanation for the apparent contradiction between fluid dynamical theory and the petrological evidence by suggesting that such sills rarely fill by the rapid injection of a single pulse of magma. Rather, they form from a series of pulses or a continuous pulsed influx over a protracted interval during which marginal cooling severely limits the potential for thermal convection.