Search

Minute bubbles of fluid sealed inside crystals have long been studied for what they reveal about mineralizing, magmatic, and metamorphic fluids and the pressure-temperature ( P - T ) paths followed by the host rocks. The famous science-fiction author H.G. Wells, as a geology student, found these far

[Huang et al. (2011)][1] describe a granite that they interpret as crystallized from an A-type magma, derived from a metasedimentary protolith. [Sun et al. (2011)][2] had already described a similar example, with high initial 87Sr/86Sr and low ɛNd. The important question is whether these A-types

Huge volumes of S-type granitic and volcanic rocks, as well as many migmatites, are the products of fluid-absent partial melting in metasedimentary protoliths at granulite-facies conditions. Thus, it is essential to know the controls on the fertility of metasedimentary rocks as magma sources. In the present experiments on fluid-absent partial melting of biotite in metasedimentary rocks, cordierite proportion is inversely correlated with melt fraction. At constant pressure ( P ) and temperature ( T ), this variation in melt proportion is a function of bulk-rock Mg number [100 Mg/(Mg + Fe)] and activity of alumina ((alpha)Al(subscript)2O(subscript)3). However, melt compositions are essentially unaffected. At 0.5 GPa, the calculated H2O contents of cordierite decrease as T increases from 850 to 1000 ° C. In aluminous metasedimentary rocks, the cordierite stability field considerably overlaps with the conditions of fluid-absent melting in the middle to lower crust, severely restricting melt productivity. Consequently, metapelites may be much less fertile sources of granitic magma than metagraywackes.

Many I-type granitoid magmas are generated through partial melting of older metaigneous rocks, and the compositions of such melts are broadly calc-alkaline and metaluminous. These melts are granitic to tonalitic, and result from thermal extremes in their lower crustal source regions. Data on the experimental partial melting of common crustal rocks suggest that high-K, I-type granitoid magmas can be derived only from the partial melting of hydrous, calc-alkaline to high-K calc-alkaline, mafic to intermediate metamorphic rocks in the crust. Because of their low K2O contents, metabasaltic rocks of all kinds are unsuitable as sources, and models that propose mixing of mantle-derived basaltic magmas and crustal melts are also inadequate. There is no requirement that I- type calc-alkaline magmatism be related in any way to subduction processes.

Criteria that have been (and still are) used to characterize S-type granites of the Lachlan Fold Belt (LFB) of southeastern Australia are reviewed, and comparisons are made with various peraluminous granites of southwestern North America, some of which have been classified as S-types on the basis of insufficient data. Virtually all of the vast volume of S-type granites in the LFB are near-surface, batholithic granites that are commonly associated with S-type volcanics and are not associated with regional metamorphic rocks and migmatites. They are strongly peraluminous, as shown by the presence of cordierite. Granites with primary muscovite are rare. All are low in Na, Ca, and Sr as a result of chemical weathering during formation of the sedimentary sources. Peraluminous granites of various ages in southwestern North America are distinctly different. They rarely contain cordierite (a mineral characteristic of LFB S-types), but some are highly evolved such that Fe-Mn-rich garnet has crystallized. They are dominantly two-mica granites, indicating crystallization at higher water fugacities and greater depths than most peraluminous granites of the LFB. Cordierite-bearing volcanics (S-types) have not been reported. Sodium is generally high in the peraluminous granites of southwestern North America. Some of these rocks have trondhjemitic affinities; the parent magmas seem more likely to have been produced by partial melting of altered basaltic rocks. Locally, some peraluminous rocks (marginal to metaluminous types) may owe their compositions to high-level contamination of I-types; these are not S-type rocks. No compelling evidence has been presented that any of the peraluminous granites of southwestern North America are S-types.

The problem of muscovite in granites is also raised by Dickson. Muscovite is common in felsic S-type granites, as stated by Chappell and White (1974), but how much is primary and how much is secondary are debatable. We currently think that very little is primary (above the solidus). We are not arguing that muscovite (whether primary or secondary), garnet, and andalusite are not found in S-types; we are arguing that these minerals may also be found in some I-types. Dickson is quite correct in saying that replacement of cordierite by micas results from hydration, which may also affect the alkali/aluminum ratio of the rock. However, he is not necessarily correct in his assertion that water must be introduced into the system; it may be magmatic water released from the magma near the solidus during cooling and crystallization. More systematic goechemical and isotopic work is required on petrographically well-documented granite suites before this question is resolved.