After many years of systematic experimental investigations, it is now possible to quantify the conditions for optimum fertility to melt production of most common crustal rock types as functions of temperature and to about 30 kbar pressure. Quartzo-feldspathic melting produces steady increases in melt proportion with increasing temperature. The exact melt fraction depends on the mineral mode relative to quartz-feldspar eutectics and the temperatures of mica dehydration melting reactions. Mica melting consumes SiO2 from residual quartz during the formation of refractory Al2SiO5, orthopyroxene, garnet or cordierite.

A simple graphical interpretation of experimental results allows a deduction of the proportions of mica and feldspar leading to optimum fertility. In effect, the mica dehydration melting reactions, at specific pressure and are superimposed on quartz-feldspar melting relations projected onto Ab-An-Or. Fertility to melt production varies with the mica to feldspar ratio and pressure. Pelites are more fertile than psammites at low pressures (e.g. 5 kbar), especially if they contain An40 to An50 plagioclase. At higher pressure (e.g. 10-20 kbar) and for rocks containing albitic plagioclase, psammites are more fertile than pelites. For a typical pelite (e.g. with An25 at 20 kbar), the cotectic with muscovite lies at higher (≍·) and XAb (≍0·42) than with biotite :≍0·35; XAb(≍·), thus dehydration melting of muscovite requires 10% more plagioclase for fertility than does biotite.

The first melts from dehydration melting of muscovite (with Plg + Qtz) are more sodic and form at lower temperatures than the first melts from Bio + Plg + Qtz. With increasing pressure, to at least 30 kbar, granite minimum and mica dehydration melts become more sodic. This indicates that of such melts is greater than 0·3.

Melting experiments with and without added H2O on a model metagreywacke and a natural metapelite demonstrate how pressure and H2O content control the compositions of melts and residual assemblages. Several effects are observed under isothermal conditions. Firstly, the stability field of biotite shrinks with decreasing pressure and with increasing H2O content, whereas that of plagioclase shrinks with increasing pressure and H2O content. Secondly, the ferromagnesian content of melts at the source (i.e. coexisting with their residual assemblages) decreases with decreasing H2O activity. Thirdly, with increasing pressure the Ca/Mg and Ca/Fe ratios of melts decrease relative to those of coexisting garnet. As a consequence, a wide spectrum of melts and crystalline residues can be generated from the same source material. For example, H2O-starved dehydration melting of metagreywacke at low pressure (≤10 kbar) generates K-rich (granitic) melts that coexist with pyroxene- and plagioclase-rich residues, whereas melting of the same material at high pressure (≍15 kbar) and with minor H2O infiltration can generate leucocratic Na-rich and Ca-poor (trondhjemitic) melts that coexist with biotite- and garnet-rich residues. An increased H2O content stabilises orthopyroxene at the expense of garnet + biotite + plagioclase, causing melts to shift towards granodioritic or perhaps tonalitic compositions.

Dehydration melting of crustal rocks may commonly occur in response to the intrusion of mafic magma in the mid- or lower crust. However, the relative importance of melt buoyancy, shear or dyking in melt generation and extraction under geologically relevant conditions is not well understood. A numerical model of the partial melting of a metapelite is presented and the model results are compared with the Ivrea-Verbano Zone in northern Italy. The numerical model uses the mixture theory approach to modelling simultaneous convection and phase change and includes special ramping and switching functions to accommodate the rheology of crystal-melt mixtures in accordance with the results of deformation experiments. The model explicitly includes both porous media flow and thermally and compositionally driven bulk convection of a restitecharged melt mass. A range of melt viscosity and critical melt fraction models is considered. General agreement was found between predicted positions of isopleths and those from the Ivrea-Verbano Zone. Maximum melt velocities in the region of porous flow are found to be 1 × 10−7 and 1 × 10−1m per year in the region of viscous flow. The results indicate that melt buoyancy alone may not be a sufficient agent for melt extraction and that extensive, vigorous convection of partially molten rocks above mafic bodies is unlikely, in accord with direct geological examples.

The behaviour of trace elements during partial melting depends primarily on their mode of occurrence. For elements occurring as trace constituents of major phases (e.g. Li, Rb, Cs, Eu, Sr, Ba, Ga, etc.), slow intracrystalline diffusion (D ≍ 10−16 cm2 s−1) at the temperature range of crustal anatexis causes all effective crystal-melt partition coefficients to have a value close to unity and impedes further melt-restite re-equilibration. Usually, therefore, the trace element composition of crustal melts simply depends on the mass balance between the proportion and composition of phases that melt and the proportion and composition of newly formed phases. The behaviour of trace elements occurring as essential structural components in accessory phases (e.g. P, La-Sm, Gd-Lu, Y, Th, U, Zr, Hf, etc.) depends on the solubility, solution kinetics, grain size and the textural position of accessory phases. In common crustal protoliths a significant mass fraction of monazite, zircon, xenotime, Th-orthosilicates, uraninite; etc.—but not apatite—is included within other major and accessory phases. During low melt fraction anatexis the amount of accessory phases available for the melt is not sufficient for saturation, thus producing leucosomes with concentrations of La-Sm, Gd-Lu, Y, Th, U and Zr lower than expected from solubility equations. Low concentrations of these elements may also occur if the melt is prevented from reaching equilibrium with the accessories due to fast segregation. However, the first mechanism seems more feasible as leucosomes that are undersaturated with respect to monazite and zircon are frequently saturated, even oversaturated, with respect to apatite.

Finite difference numerical simulations were used to characterise the rates of diffusion-controlled dissolution and growth of zircon in melts of granitic composition under geologically realistic conditions. The simulations incorporated known solubility and Zr diffusivity relationships for melts containing 3 wt% dissolved H2O and were carried out in both one and thre dimensions under conditions of constant temperature, linearly time-dependent temperature and for a variety of host system thermal histories. The rate of zircon dissolution at constant temperature depends systematically on time (t½−12;), temperature (exp T−1) and degree of undersaturation of the melt with respect to zircon (in ppm Zr). Linear dissolution and growth rates fall in the range 10−19 10−15 cm s−1 at temperatures of 650-850°C. Radial rates are strongly dependent on crystal size (varying in inverse proportion to the radius, r): for r>30 μm, dissolution and growth rates fall between 10−17 and 10−13 cm s−1. During crustal magmatism, the chances of survival for relict cores of protolith zircons depend on several factors, the most important of which are: the initial radius of the zircon; the intensity and duration of the magmatic event; and the volume of the local melt reservoir with which the zircon interacts. In general, only the largest protolith zircons (>120 μm radius) are likely to survive magmatic events exceeding 850°C. Conversely, only the smallest zircons (<50 μm radius) are likely to be completely consumed during low-temperature anatexis (i.e. not exceeding ≍700°C).

The effects of stirring the zircon-melt system are unimportant to dissolution and growth behaviour; except under circumstances of extreme shearing (e.g. filter pressing?), zircon dissolution is controlled by diffusion of Zr in the melt.

New experimental determinations of water solubility in haplogranitic melts (anhydrous compositions in the system Qz-Ab-Or and binary joins) and of the viscosity of hydrous Qz28Ab38Or34 melts (normative proportions) and natural peraluminous leucogranitic melt (Gangotri, High Himalaya) are used to constrain the evolution of viscosity of ascending magmas, depending on their P-T paths.

At constant pressure, in the case of fluid-absent melting conditions, with water as the main volatile dissolved in the melts, the viscosity of melts generated from quartzo-feldspathic protoliths is lower at low temperature than at, high temperature (difference of 1-2 log units between 700 and 900°C). This is due to the higher water contents of the melts at low temperature than at high temperature and to the fact that decreasing temperature does not counterbalance the effect of increasing melt water content. In ascending magmas generated from crustal material the magma viscosity does not change significantly whatever the P-T path followed (i.e. path with cooling and crystallisation; adiabatic path with decompression melting) as long as the crystal fraction is low enough to assume a Newtonian behaviour (30-50% crystals, depending on size and shape). Comparison of the properties of natural and synthetic systems suggests that both water solubility and the viscosity of multicomponent natural felsic melts (with less than 30-35% normative Qz) can be extrapolated from those of the equivalent synthetic feldspar melts.

Progress in the understanding of the volumes and viscosities of granitic and related pegmatitic melts generated by experimental studies are reviewed. The results of a series of investigations of the volumes and viscosities of melts derived from a haplogranitic base composition, HPG8, located near the 2 kbar water-saturated minimum melt composition in the albite—orthoclase—silica system are discussed. Melt volumes, obtained using a combination of dilatometric and calorimetric methods at 1 atm and relatively low temperatures yield an internally consistent set of partial molar volumes for 18 components in granitic melts. These partial molar volumes, combined with an estimate for water, allow the estimation of melt densities for granitic and related pegmatitic magmas.

Melt viscosities, obtained using a combination of high and low range viscometry techniques, provide a template for the estimation of melt viscosities in more complex natural systems. The parameterisation of the non-Arrhenian temperature-dependence of the viscosity of such melts is presented, together with some structural implications of the variation of melt viscosity with temperature and composition. Outstanding questions related to the PVT equation of state of granitic melts and to the mechanical response to shear stresses are discussed, with an outlook for the experimental solutions to those questions in the next few years.

The rheological and chemical behaviour of the lower crust during anatexis has been a major focus of geological investigations for many years. Modern studies of crustal evolution require significant knowledge, not only of the potential source regions for granites, but also of the transport paths and emplacement mechanisms operating during granite genesis. We have gained significant insights into the segregation and transport of granitoid melts from the results of experimental studies on rock behaviour during partial melting. Experiments performed on crustal rock cores under both hydrostatic conditions and during deformation have led, in part, to two conclusions. (1) The interfacial energy controlling melt distribution is anisotropic and, as a result, the textures deviate significantly from those predicted for ideal systems—planar solid-melt interfaces are developed in addition to triple junction melt pockets. The ideal dihedral angle model for melt distribution cannot be used as a constraint to predict melt migration in the lower crust. (2) The ‘critical melt fraction’ model, which requires viscous, granitic melt to remain in the source until melt fractions reach >25 vol%, is not a reliable model for melt segregation. The most recent experimental results on crustal rock cores which have helped advance our understanding of melt segregation processes have shown that melt segregation is controlled by several variables, including the depth of melting, the type of reaction and the volume change associated with that reaction. Larger scale processes such as tectonic environment determine the rate at which the lower crust heats and deforms, thus the tectonic setting controls the melt fraction at which segregation takes place, in addition to the pressure and temperature of the potential melting reactions. Melt migration therefore can occur at a variety of different melt fractions depending on the tectonic environment; these results have significant implications for the predicted geochemistry of the magmas themselves.

To form a granite pluton, the felsic melt produced by partial melting of the middle and lower continental crust must separate from its source and residuum. This can happen in three ways: (1) simple melt segregation, where only the melt fraction moves; (2) magma mobility, in which all the melt and residuum move together; and (3) magma mobility with melt segregation, in which the melt and residuum move together as a magma, but become separated during flow. The first mechanism applies to metatexite migmatites and the other two to diatexite migmatites, but the primary driving forces for each are deviatoric stresses related to regional-scale deformation. Neither of the first two mechanisms generates parental granite magmas. In the first mechanism segregation is so effective that the resulting magmas are too depleted in FeOT, MgO, Rb, Zr, Th and the REEs, and in the second no segregation occurs. Only the third mechanism produces magmas with compositions comparable with parental granites, and occurs at a large enough scale in the highest grade parts of migmatite terranes, to be considered representative of the segregation processes occurring in the source regions of granites.

Diapirism has been discredited as a transport mechanism for magmas partly because diapirs seem to be unable to bring magmas to shallow crustal levels (<10km) and partly because recent developments in the theory of dyke propagation have shown that sufficiently wide dykes are able to efficiently transport felsic magmas through the crust. However, it is still unclear how felsic dykes grow to widths that allow them to propagate faster than they close by magma freezing. Ultimately, it may be the ability of felsic dykes to grow within the source that controls which mechanism dominates ascent.

The ability of dykes to propagate, from the top of rising diapirs depends among other factors on the changing temperature gradient of the wall rocks. The steep gradient around rapidly rising diapirs in the low viscosity lower crust will cause dykes to freeze. As diapirs rise to colder stiffer crust and decelerate, heat diffuses further from the diapir, resulting in shallower temperature gradients that favour dyke propagation. The mechanism may thus swap, during ascent, from diapirism to dyking. Calculations of the thermal evolution of diapirs and their surroundings show that basaltic diapirs may never form because they would be drained by dykes at a very early stage; felsic diapirs may be unable to give rise to successful dykes, whereas diapirs of intermediate magmas may propagate dykes during ascent.

Until the last few years, diapirism reigned supreme among granitoid ascent mechanisms. Granitoid masses in a variety of material states, from pure melt through semi-molten crystal mushes to solid rock, were believed to have risen forcefully through the continental crust to their final emplacement levels in a way analogous to salt domes. The structural analogy between granite plutons and salt diapirs, which gained acceptance in the 1930s, has clearly been attractive despite the pessimistic outcomes of thermal models and, at best, ambiguous field evidence.

In contrast with traditional diapiric ascent, dyke transport of granitoid magmas has a number of important implications for the emplacement and geochemistry of granites that have yet to be fully explored. Rapid ascent rates of ≍ 10 2m/s predicted for granite melts in dykes (cf. m/a for diapirs) mean that felsic magmas can be transported through the continental crust in months rather than thousands (or even millions) of years, and that large plutons can in principle be filled in <104 a. Granitic melts are likely to rise adiabatically from their source regions, leading to the resorption of any entrained restitic material. Ascending melts in dykes close to their critical minimum widths may have little opportunity to assimilate significant amounts of country rock, and if source extraction is sufficiently rapid, most crustal contamination will be restricted to the site of emplacement. Rates of pluton and batholith inflation will be determined by the amount and rate of melt extraction at source.

The construction of large plutons and batholiths piecemeal from a number of magma pulses separated by periods of relative quiescence provides a means of reconciling rapid ascent rates with times for batholith construction based on average rates. Field and seismic evidence that shows batholiths as large, sheet-like structures with flat roofs and floors is consistent with a general model for plutons and batholiths as laccolith-type structures, fed from depth by dykes. The overall geometry of this type of structure helps ameliorate the space problem, which developed as a consequence of the unrealistic volumes of upwelling granite associated with the classical diapir model.

Buddington (1959) pointed out that the construction of large crustal magma chambers involves complex internal processes as well as multiple country rock material transfer processes (MTPs), which reflect large horizontal, vertical and temporal gradients in physical conditions. Thus, we have attempted to determine the relative importance of different magmatic and country rock MTPs at various crustal depths, and whether country rock MTPs largely transport material vertically or horizontally, rather than seeking a single model of magma ascent and emplacement.

Partially preserved roofs of nine plutons and in some cases roof–wall transitions with roof emplacement depths of 1·5–11 km were mapped. During emplacement, these roofs were not deformed in a ductile manner, detached or extended by faults, or significantly uplifted. Instead, sharp, irregular, discordant contacts are the rule with stoped blocks often preserved immediately below the roof, even at depths of 10 km. The upper portions of these magma chambers are varied, sometimes preserving the crests of more evolved magmas or local zones of volatile-rich phases and complex zones of dyking and magma mingling. Magmatic structures near roofs display a wide variety of patterns and generally formed after emplacement. Transitions from gently dipping roofs to steep walls are abrupt. At shallow crustal levels, steep wall contacts have sharp, discordant, stepped patterns with locally preserved stoped blocks indicating that the chamber grew sideways in part by stoping. Around deeper plutons, an abrupt transition (sometimes within hundreds of metres) occurs in the country rock from discordant, brittle roofs to moderately concordant, walls deformed in a ductile manner defining narrow structural aureoles. Brittle or ductile faults are not present at roof–wall joins.

Near steep wall contacts at shallow to mid-crustal depths (5–15 km), vertical and horizontal deflections of pre-emplacement markers (e.g. bedding, faults, dykes), and ductile strains in narrow aureoles (0·1–0·3 body radii) give a complete range of bulk strain values that account for 0–100% of the needed space, but average around 30%, or less, particularly for larger batholiths. A lack of far-field deflection of these same markers rules out significant horizontal displacement outside the aureoles and requires that any near-field lateral shortening is accommodated by vertical flow. Lateral variations from ductile (inner aureole) to brittle (outer aureole) MTPs are typically observed. Compositional zoning is widespread within these magma bodies and is thought to represent separately evolved pulses that travelled up the same magma plumbing system. Magmatic foliations and lineations commonly cross-cut contacts between pulses and reflect the strain caused either by the late flow of melt or regional deformation.

Country rocks near the few examined mid- to deep crustal walls (10–30 km) are extensively deformed, with both discordant and concordant contacts present; however, the distinction between regional and emplacement-related deformation is less clear than for shallower plutons. Internal sheeting is more common, although elliptical masses are present. Lateral compositional variations are as large as vertical variations at shallower depths and occur over shorter distances. Magmatic foliations and lineations often reflect regional deformation rather than emplacement processes.

The lack of evidence for horizontal displacement outside the narrow, shallow to mid-crustal aureoles and the lack of lateral or upwards displacement of pluton roofs indicate that during emplacement most country rock is transported downwards in the region now occupied by the magma body and its aureole. The internal sheeting and zoning indicate that during the downwards flow of country rock, multiple pulses of magma travelled up the same magma system. If these relationships are widespread in arcs, magma emplacement is the driving mechanism for a huge crustal-scale exchange process.

Most granitic batholiths contain plutons which are composed of low-variance mineral assemblages amenable to quantification of the P– conditions that characterise emplacement. Some mineral thermometers, such as those based on two feldspars or two Fe–Ti oxides, commonly undergo subsolidus re-equilibration. Others are more robust, including hornblende–plagioclase, hornblende–clinopyroxene, pyroxene–ilmenite, pyroxene–biotite, garnet–hornblende, muscovite-biotite and garnet–biotite. The quality of their calibration is variable and a major challenge resides in the large range of liquidus to solidus crystallisation temperatures that are incompletely preserved in mineral profiles. Further, the addition of components that affect Kd relations between non-ideal solutions remains inadequately understood. Estimation of solidus and near-solidus conditions derived from exchange thermometry often yield results >700°C and above that expected for crystallisation in the presence of an H2O-rich volatile phase. These results suggest that the assumption of crystallisation on an H2O-saturated solidus may not be an accurate characterisation of some granitic rocks.

Vapour undersaturation and volatile phase composition dramatically affect solidus temperatures. Equilibria including hypersthene–biotite–sanidine–quartz, fayalite–sanidine–biotite, and annite–sanidine–magnetite (ASM) allow estimation of Estimates by the latter assemblage, however, are highly dependent on . Oxygen fugacity varies widely (from two or more log units below the QFM buffer to a few log units below the HM buffer) and can have a strong affect on mafic phase composition. Ilmenite–magnetite, quartz–ulvospinel–ilmenite–fayalite (QUILF), annite–sanidine–magnetite, biotite–almandine–muscovite–magnetite (BAMM), and titanite–magnetite–quartz (TMQ) are equilibria providing a basis for the calculation of .

Granite barometry plays a critical part in constraining tectonic history. Metaluminous granites offer a range of barometers including ferrosilite–fayalite–quartz, garnet–plagioclase–hornblende–quartz and Al-in-hornblende. The latter barometer remains at the developmental stage, but has potential when the effects of temperature are considered. Likewise, peraluminous granites often contain mineral assemblages that enable pressure determinations, including garnet–biotite–muscovite–plagioclase and muscovite–biotite–alkali–feldspar–quartz. Limiting pressures can be obtained from the presence of magmatic epidote and, for low-Ca pegmatites or aplites, the presence of subsolvus versus hypersolvus alkali feldspars.

As with all barometers, the influence of temperature, , and choice of activity model are critical factors. Foremost is the fact that batholiths are not static features. Mineral compositions imperfectly record conditions acquired during ascent and over a range of temperature and pressure and great care must be taken in properly quantifying intensive parameters.

Recent water-undersaturated phase equilibrium data on the subsystems of the granite-H2O system have provided important new constraints on the topology of the cotectic surfaces and hence on the compositional evolution of felsic magmas. The effect of water on phase relations can be deduced from a comparison of anhydrous and H2O-saturated data or from data obtained in the presence of a CO2-bearing fluid. However, although new experimental evidence indicates that the silica enrichment of evolving H2O-undersaturated, H2O-unbuffered melts during the co-precipitation of quartz and feldspar is as previously thought for orthoclase-rich compositions, it suggests that such a trend is considerably less for Ab-rich compositions. For water-poor trachytic melts, the newly recognised strong destabilisation of the sanidine melt component relative to the anorthite melt component with increasing water content indicates that the co-precipitation of two feldspars will result in saturation of the melt with ternary alkali feldspar at an earlier stage (i.e. higher melt anorthite content) than previously thought. This, in turn, implies that the melt differentiation path will have a greater component of anorthite depletion during the equilibrium co-precipitation of ternary feldspars and that the melt will remain in the peritectic region of the two feldspar plus liquid surface over a greater interval of crystallisation, thereby enhancing the possibility that the resoption of plagioclase during the early stages of equilibrium with alkali feldspar may go to completion. Comparison of CO2-free and CO2-bearing haplogranitic phase equilibrium data suggests that CO2 may be playing an independent part in the modification of phase equilibria and may induce a significant destabilisation of the orthoclase melt component.

Petrological investigations of granite commonly reveal multiple periods of growth punctuated by resorption for many of the constituent minerals. Complementary to such textures are mineral compositional heterogeneity manifested by zoning or grain to grain variability. These features ultimately reflect changes in the intensive parameters or activities of components during melt solidification. Such complexities of granite crystallisation can be simultaneously modelled in a reaction space constructed from the set of linearly independent reactions describing the equilibria among all phases and components in the system of interest.

The topology of the linearly independent reactions that define the reaction space for garnetmuscovite-biotite granites yields the following insights: (1) there is no one unique reaction that produces or consumes aluminous minerals (e.g. garnet); (2) minerals can alternate as reactants or products in different reactions accounting for textures indicating multiple periods of crystallisation separated by resorption; (3) mineral compositions are regulated by the reaction(s) producing them and vary as the stoichiometry of the reaction(s) producing them varies; (4) resorption of early crystallising garnet is likely to reflect decreasing pressure, presumably during magma ascent; (5) late crystallisation of garnet, at the expense of biotite, reflects an increase in melt aluminosity and does not necessarily require high Mn activities for the melt and (6) increasing melt H2O, at H2Oundersaturated conditions, favours the formation of biotite–muscovite granite.

Application of the reaction space method to other granite types holds considerable promise for elucidating reactions that regulate mineral assemblages and compositions during crystallisation.

Granites within suites share compositional properties that reflect features of their source rocks. Variation within suites results dominantly from crystal fractionation, either of restite crystals entrained from the source, or by the fractional crystallisation of precipitated crystals. At least in the Lachlan Fold Belt, the processes of magma mixing, assimilation or hydrothermal alteration were insignificant in producing the major compositional variations within suites. Fractional crystallisation produced the complete variation in only one significant group of rocks of that area, the relatively high temperature Boggy Plain Supersuite. Modelling of Sr, Ba and Rb variations in the I-type Glenbog and Moruya suites and the S-type Bullenbalong Suite shows that variation within those suites cannot be the result of fractional crystallisation, but can be readily accounted for by restite fractionation. Direct evidence for the dominance of restite fractionation includes the close chemical equivalence of some plutonic and volcanic rocks, the presence of plagioclase cores that were not derived from a mingled mafic component, and the occurrence of older cores in many zircon crystals. In the Lachlan Fold Belt, granite suites typically evolved through a protracted phase of restite fractionation, with a brief episode of fractional crystallisation sometimes evident in the most felsic rocks. Evolution of the S-type Koetong Suite passed at about 69% SiO2 from a stage dominated by restite separation to one of fractional crystallisation. Other suites exist where felsic rocks evolved in the same way, but the more mafic rocks are absent. In terranes in which tonalitic rocks formed at high temperatures are more common, fractional crystallisation would be a more important process than was the case for the Lachlan Fold Belt.

The paradox of Lachlan Fold Belt (LFB) granitoids is that although contrasted chemical types (S- and I-types) imply melting of distinct crustal sources, the simple Nd–Sr–Pb–O isotopic arrays indicate a continuum, suggesting mixing of magmatic components. The paradox is resolved by the recognition that the previously inferred, isotopically primitive end-member is itself a crust-mantle mix, so that three general source components, mantle, lower crust and middle crust, comprise the granitoids. Based on Nd isotopic evidence, mantle-derived basaltic magmas melted and mixed with Neoproterozoic-Cambrian, arc-backarc-type material to produce primitive I-type, parental granitoid magmas in the lower–middle crust. Ordovician metasediment, locally underthrust to mid-crustal levels, was remobilised under the elevated geotherms and is most clearly recognised as diatexite in the Cooma complex, but it also exists as gneissic enclaves in S-type granites. The diatexite mixed with the hybrid I-type magmas to produce the parental S-type magmas. Unique parent magma compositions of individual granite suites reflect variations within any or all of the three major source components, or between the mixing proportions. For example, chemical tie-lines between Cooma diatexite and mafic I-type Jindabyne suite magma encompass almost all mafic S-type granites of the vast Bullenbalong supersuite, consistently in the proportion Jindabyne: Cooma, 30:70. The modelling shows that LFB S-type magmas are heavily contaminated I-type magmas, produced by large-scale mixing of hot I-type material with lower temperature diatexite in the middle crust. The model implies a genetic link between migmatite and pluton-scale, crustally derived (S-type) granites.

Given the chemical and isotopic contrasts of the crustally derived source components, and their typically unequal proportions in the magmas, it is not surprising that the LFB granitoids are so distinctive and have been categorised as S- and I-type. The sublinear chemical trends of the granitoid suites are considered to be secondary effects associated with crystal fractionation of unique parental magmas that were formed by three-component mixing. The model obviates the necessity for multiple underplating events and Proterozoic continental basement, in accordance with the observed tectonostratigraphy of the Lachlan Fold Belt.

The Cretaceous granitic rocks and associated regional metamorphic rocks in SW Japan were formed by a Cordilleran-type orogeny. Southwest Japan is regarded as a hypothetical cross-section of the upper to middle crust of the Eurasian continental margin in the Cretaceous, comprising (1) high-level granitoids (called San-yo type) and weakly to unmetamorphosed accretionary complexes that are exposed on the back-arc side and (2) low-level (Ryoke type) granitoids with high-grade metamorphites up to migmatitic gneisses on the forearc side. All these granitoids are of the ilmenite series, and predominantly I-type, with a subordinate amount of garnet- or muscovite-bearing varieties in the Ryoke zone, but none of these contains cordierite. These mineralogical variations are likely to depend more on their slightly peraluminous chemistry rather than the pressure differences during crystallisation.

In the eastern part of SW Japan, the granitoids of both levels give K–Ar biotite ages of approximately 65 Ma, whereas the magmatic age of high-level granitoids is approximately 70 Ma, 15 Ma younger than the nearly 85 Ma old lower level granitoids. This implies that the formation of the middle crust started approximately 15 Ma before that of the upper crust. The middle crust material was kept over 500°C for 15–20 Ma after solidification, then it cooled together with the upper crust to 300°C, 6–7 Ma after the formation of the upper crust. The coincidence of cooling history below 500°C of the upper and middle crust may reflect the regional uplift of the crust.

The low-level granitoids have higher 87Sr/86Sr initial ratios than those of high-level granitoids in the middle-western part (Chugoku district), but the relationship appears to be opposite in the eastern part. This may imply that the two plutonic series formed by separate magmatic pulses at an interval of c. 15 Ma, even though they are not independent, but rather part of a larger episode of crustal growth.

In Patagonia a Triassic-Early Jurassic Cordilleran interior magmatic belt preceded the widespread eruption of Middle Jurassic syn-extensional rhyolites. Two plutons (La Calandria and La Leona) represent the easternmost plutonic rocks of this belt, > 750 km east of the present oceanic trench. They define a high-K calc-alkaline monzonite series in contrast with the main Andinotype arc magmatism of the Pacific margin: they are enriched in large ion lithophile elements (K, Rb, Ba, Sr and Th), LREE and P2O5 and depleted in HREE and Y, with low FeO*/MgO ratio. The range of observed compositions (56-76% SiO2) resulted from high-level fractionation of plagioclase, hornbleńde, biotite, K-feldspar and accessories (sphene, apatite and zircon).

Initial 87Sr/86Sr ratios, average εNdt and mean depleted-mantle Nd model ages of the two plutons are 0·70487, -0·5 and 1050 Ma for La Calandria and 0·70509, -1·4 and 1125 Ma for La Leona, respectively. They are thus isotopically more primitive than the Middle Jurassic rhyolites, previously attributed to partial melting of Mesoproterozoic mafic lower crust. The preferred model for the origin of the monzonites is remelting of an amphibole- + garnet-bearing, plagioclase-poor, high-K mafic source (?underplating). This occurred in a distal sector of a dying oblique subduction regime, immediately preceding the extensional silicic volcanism.