The Mystery of Moving Minima

Tuttle and Bowen's memoir clearly identified minimum melt compositions in the system Qz-Ab-Or, and cleared the way for recognition of a magmatic origin for granites such as those of SW England. However, F-rich granites and related volcanic rocks (e.g. ongonites) had bulk compositions displaced markedly towards the albite apex; Manning's determination of phase relationships in Qz-Ab-Or with added F demonstrated that these rocks too had bulk compositions consistent with their crystallization from F-enriched magmas. Thus it was possible to explain critical aspects of granite geochemistry in provinces such as SW England in terms of magmatic processes.

Enigmas remain in the form of granitic rocks whose compositions lie distant from the experimentally determined minima. For example, very fine grained rhyolitic porphyry (elvan) dykes that cross-cut the Cornubian batholith plot on the Qz-Or join. These have lost all vestiges of any original vitreous texture, leading to theories for their origin that involve K-Na exchange during or after complete devitrification associated with high temperature K enrichment via reaction with a supercritical fluid phase.

Young (<5Ma) perlitic rhyolites from Lopburi, Thailand have bulk compositions that plot close to low pressure minima in Qz-Ab-Or determined by Tuttle and Bowen. Glass compositions, however, lie on the Qz-Or join. The Thai rhyolites contain spherulites, indicating partial devitrification, whose bulk composition lies towards the Qz-Ab join. The development of glass compositions that exclude Na must involve processes of cation exchange requiring diffusion of alkalis within the melt, before extensive devitrification. This leads to the possibility that the SW England elvans also owe their composition to fluid-melt interactions prior to complete crystallization. Thus volatile-rich conditions late in the magmatic history of a granite can lead to gross perturbance of bulk rock compositions in rocks otherwise wholly crystallized from granitic magma.