ORIGINAL PAPER T. W. Sisson Æ K. Ratajeski Æ W. B. Hankins A. F. Glazner Voluminous granitic magmas from common basaltic sources Received: 22 August 2003 / Accepted: 22 October 2004 / Published online: 17 December 2004 Ó Springer-Verlag 2004 Abstract Granitic—rhyolitic liquids were produced experimentally from moderately hydrous (1.7–2.3 wt% H 2 O) medium-to-high K basaltic compositions at 700 MPa and fO 2 controlled from Ni-NiO 1.3 to +4. Amount and composition of evolved liquids and coex- isting mineral assemblages vary with fO 2 and tempera- ture, with melt being more evolved at higher fO 2 s, where coexisting mineral assemblages are more plagioclase- and Fe–Ti oxide-rich and amphibole-poor. At fO 2 of Ni–NiO +1, typical for many silicic magmas, the sam- ples produce 12–25 wt% granitic–rhyolitic liquid, amounts varying with bulk composition. Medium-to- high K basalts are common in subduction-related mag- matic arcs, and near-solidus true granite or rhyolite liquids can form widely, and in geologically significant quantities, by advanced crystallization–differentiation or by low-degree partial remelting of mantle-derived basaltic sources. Previously differentiated or weathered materials may be involved in generating specific felsic magmas, but are not required for such magmas to be voluminous or to have the K-rich granitic compositions typical of the upper continental crust. Introduction Nearly 50 years after Tuttle and Bowen (1958), the origins of rhyolitic and granitic magmas remain subjects of debate, disagreement, and study. While it is generally accepted that strongly peraluminous magmas with crust- like radiogenic and stable isotope values originate by partial melting of micaceous metasedimentary rocks (White and Chappell 1983; Vielzeuf and Holloway 1988), diverse origins are proposed for the more abun- dant metaluminous-to-weakly peraluminous silicic magmas (Fig. 1). Explanations include crystallization– differentiation of basaltic parents (Bowen 1928), partial remelting of igneous rocks (White and Chappell 1983), assimilation of sialic rocks into differentiating basaltic magmas (DePaolo 1981), and partial melting of imma- ture volcanogenic sediments with igneous-like composi- tions (some metagreywackes (Ewart and Stipp 1968)). Advantages and disadvantages accrue to each of these interpretations. Intermediate-to-silicic igneous rocks are close to the desired melt compositions, but contain little water and therefore cannot partially melt to yield large amounts of silicic magma except at high temperatures (Beard et al. 1993). Metagreywackes are hydrous, at least at low metamorphic grades, but it is not apparent that metagreywackes are sufficiently widespread to be a common source for granites and rhyolites. Assimilation of sialic rocks into basalt, partial melting of intermedi- ate-to-silicic igneous rocks, and partial melting of metagreywackes certainly take place, but each of these processes requires the earlier formation of silicic mate- rial to produce later silicic magmas. The questions arise, ‘‘What created the earlier silicic material?’’ and ‘‘Is there a simpler way?’’ Derivation of granitic and rhyolitic magmas directly from basaltic sources, either by advanced crystalliza- tion–differentiation or by partial melting, has some distinct advantages over the aforementioned models. For differentiation, basaltic parent magmas carry their own heat, obviating the need to obtain heat from Editorial Responsibility: T. L. Grove T. W. Sisson (&) Æ W. B. Hankins Volcano Hazards Program, U.S. Geological Survey, 345 Middlefield Rd., Menlo Park, CA 94025, USA E-mail: tsisson@usgs.gov Tel.: +1-650-3295247 Fax: +1-650-3295203 E-mail: bhankins@usgs.gov K. Ratajeski Æ A. F. Glazner Department of Geological Sciences, University of North Carolina, Chapel Hill, NC, 27599-3315 USA E-mail: afglazne@unc.edu Present address: K. Ratajeski Department of Earth Sciences, Montana State University, 173480, Bozeman, USA E-mail: kratajeski@montana.edu Contrib Mineral Petrol (2005) 148: 635–661 DOI 10.1007/s00410-004-0632-9