On the Basic Principles of Igneous Petrology

Thursday, 18 December 2014: 1:40 PM
Bruce D Marsh, Johns Hopkins University, Baltimore, MD, United States
How and why Differentiation occurs has dominated Igneous Petrology since its beginning (~1880) even though many of the problems associated with it have been thoroughly solved. Rediscovery of the proverbial wheel with new techniques impedes progress. As soon as thin section petrography was combined with rock and mineral chemistry, rock diversity, compositional suites, and petrographic provinces all became obvious. The masterful 1902 CIPW norm in a real sense solved the chemical mystery of differentiation: rocks are related by the addition and subtraction of minerals in the anciently appreciated process of fractional crystallization. Yet few believed this, even after phase equilibria arrived. Assimilation, gas transfer, magma mixing, Soret diffusion, immiscibility, and other processes had strong adherents, even though by 1897 Becker conclusively showed the ineffectiveness of molecular diffusion in large-scale processes. The enormity of heat to molecular diffusion (today’s Lewis no.) should have been convincing; but few paid attention. Bowen did, and he refined and restated the result; few still paid attention. And in spite of his truly masterful command of experiment and field relations in promoting fractional crystallization, Fenner and others fought him with odd arguments. The beauty of phase equilibria eventually dominated at the expense of knowing the physical side of differentiation. Bowen himself saw and struggled with the connection between physical and chemical processes. Progress has come from new concepts in heat transfer, kinetics, and slurry dynamics. The key approach is understanding the dynamic competition between spatial rates of solidification and all other processes. The lesson is clear: Scholarship and combined field, laboratory and technical expertise are critical to understanding magmatic processes. Magma is a limitlessly enchanting and challenging material wherein physical processes buttressed by chemistry govern.