Progressive alteration by seawater of an andesite in the Aegean Island of Milos and an ignimbrite in the Aegean Island of Kimolos, Greece, formed bentonites with or without zeolites. Both profiles are dominated by migration of alkalis and uptake of Mg, Fe and H2O, while Al and Ti are immobile. The relative removal of alkalis controls the formation of either smectite or zeolites. The behavior of Ca and Si depends on the chemistry of the parent rock. In the rhyolitic profile, alteration is controlled by gain of Mg, Fe2+ and Ca and loss of Na, K and Si, while in the andesitic profile by gain of Mg and Fe2+ and loss of Na, K and Ca. In both profiles, significant uptake of SO4- was not observed. Moreover Zr, Nb, V and Ni are immobile and have been enriched residually, while Sr, Rb and Y are lost in both profiles. Thorium is immobile in the rhyolitic profile but is leached in the andesitic profile. Also, the rare earth elements (REE) display fractionation in both profiles; the degree of fractionation increases with the degree of alteration to bentonite. Fractionation of the REE in both profiles and mobility of Th in the andesitic profile are related to the existence of monazite (rhyolitic profile) and apatite (andesite profile). The REE and Th appear to partition into phosphates rather than smectite.

The mobility of Y coupled with the immobility of Nb increases the Nb: Y ratio with advancing alteration, rendering discrimination diagrams that use this ratio to determine the nature of the protoliths misleading. Mass balance calculations showed that in the smectite-rich zones the water: rock (WR) ratio might be as high as 13:1 in both profiles, while in the zeolite-bearing zones it is about 5.5:1. Such WR ratios explain the observed extensive mass transfer and suggest that the pore fluid chemistry might overprint the chemical characteristics of the parent rocks controlling smectite and bentonite chemistry.

The vitric component of the silicic pyroclastic flows and surge deposits (Prassa ignimbrite unit), from the northwestern exposures of the pyroclastics formation of Polyegos Island, South Aegean Sea Volcanic Arc in Greece, is replaced by authigenic zeolite and clay minerals. Mordenite dominates, and clinoptilolite (heulandite type 3), illite and illite-smectite (I-S) are subordinate. Opal-CT, quartz, feldspar, biotite and halite complete the mineralogical suite. In the southwestern part of the pyroclastics formation (Myrsini pyroclastics unit), kaolinite, halloysite, alunite and amorphous silica are the major mineralogical constituents as a result of a strong hydrothermal alteration by solutions enriched in SO42-. Scanning electron microscope (SEM) examination proved that the 1st type of zeolites formed within the area were heulandite minerals, following the formation of smectite, as a result of the activity of pore fluids within the volcaniclastic pile. There was no substantial evidence to support a hypothesis that mordenite was formed with the heulandite minerals after the initial stages of glass dissolution, other than some very minor mordenite that was formed from a gel-like type of material. The majority of mordenite present is most often draped over and formed from the crystals of heulandite minerals. Heulandite minerals often show advanced dissolution effects. The mineral dissolution was due to the emplacement of rhyolite lava domes and flows and the associated temperature rise and circulation of hydrothermal fluids, which had also mixed with seawater. The volcanic activity raised the temperature and changed the pore fluid chemistry, and the more unstable members of the heulandite minerals group were transformed to mordenite. Clinoptilolite (heulandite type 3), which was found within a few samples, was thermally more stable than any heulandite type 1 or 2 phases initially present. Therefore, clinoptilolite was either transformed more slowly or within a very few cases, it has not been affected at all. A heulandite-minerals-derived material acted as the major precursor for the formation of mordenite. The temperature increase within the area and the later hydrothermal alteration effect are also indicated by the illitization of the smectite. Mordenite is also found to have formed from I-S clays. Overall, mordenite formed as a result of elevated temperature and high Na+ concentration.

Polytypism is common in micas, and the frequency of polytype occurrence is believed to be related closely to the crystallization conditions and chemical compositions of the corresponding fluids and melts. Coexisting multiple standard and complex/disordered polytypes in igneous rocks generally reflect a complicated magma evolution history. The purpose of the current study was to clarify the origin of coexisting biotite polytypes and their growth mechanism. Micro-X-ray diffraction (μXRD) and transmission electron microscopy (TEM) were used to investigate Fe-rich biotite phenocrysts in rhyolite from the Long Valley Caldera, California, USA. The μXRD analyses characterized various polytypes, and TEM observations revealed that common polytypes (e.g. 1M, 2M1, and 3T) and rare polytypes (e.g. 4M2 and 4A8) coexist within biotite monocrystals. The two 4-layer polytypes of Fe-rich biotite, 4M2 and 4A8, were identified via selected-area electron diffraction (SAED) and high-resolution scanning transmission electron microscopy (HRSTEM) at the atomic resolution, with unique stacking sequences ([0222] for 4M2 and [002$\overline{2}$] for 4A8). Energy-dispersive X-ray spectroscopy (EDS) results showed differences in their chemical compositions, especially Fe and K. The 4A8 polytype is reported for the first time. The present study suggested that environmental changes, such as rapid cooling and inhomogeneous compositional distribution, led to chemical and structural oscillations and complex nucleation of the two 4-layer polytypes. Screw dislocations producing spiral growth enhance polytype stability and form ordered long-period/complex polytypes. These results are useful to understand the origin of long-period/complex polytypes and the intergrowths of diverse polytypes formed in non-equilibrium crystallization environments.

Silicic magmatism, minor overall in the ∼65.5 Ma Deccan Traps continental flood basalt (CFB) province of India, was widespread in the Saurashtra region. We describe the physical volcanology of silicic volcanics and dykes exposed around Rajula–Savarkundla–Gariyadhar–Talaja towns in southeastern Saurashtra. The silicic volcanics conformably overlie basaltic lavas, suggesting rapid subaerial volcanism, and the sequence shows gentle tectonic dips (∼15°) towards the Arabian Sea. Rhyolites and dacites with preserved thicknesses of tens of metres show intense internal rheomorphic deformation, and a dacite shows a well-formed basal autobreccia. The rheomorphic rhyolites, and vitrophyres which often underlie them, lack vitroclasts (glass shards and pumice clasts). They have very similar mineral assemblages (quartz and alkali feldspar phenocrysts, and crystal cargoes dominated by calcic plagioclase and clinopyroxene or orthopyroxene, sometimes with olivine). The rheomorphic units are thus flood rhyolite and dacite lavas, apparently common in the northern-northwestern Deccan, and the vitrophyres their basal chilled parts. Tuffs (including crystal-vitric Plinian fallout ash) and eutaxitic ignimbrites formed from pyroclastic density currents; one tuff contains extraordinary numbers of lithophysae. Ridges of rhyolitic tuff breccias with pervasive secondary silicification and ferruginization represent pyroclastic eruptive fissures. The area thus records large-scale effusive and explosive silicic eruptions. Mafic and silicic dykes intrude the basaltic lavas and rarely the silicic volcanics. Mafic enclaves in several silicic dykes and some volcanics indicate magma mingling as a common phenomenon. The seaward-dipping volcanic units define a regional-scale flexure comparable to coastal flexures in CFB provinces worldwide, suggesting extensive block-faulting of this classical volcanic rifted margin.

Most continental flood basalt (CFB) provinces of the world contain silicic (granitic and rhyolitic) rocks, which are of significant petrogenetic interest. These rocks can form by advanced fractional crystallization of basaltic magmas, crustal assimilation with fractional crystallization, partial melting of hydrothermally altered basaltic lava flows or intrusions, anatexis of old basement crust, or hybridization between basaltic and crustal melts. In the Deccan Traps CFB province of India, the Barda and Alech Hills, dominated by granophyre and rhyolite, respectively, form the largest silicic complexes. We present petrographic, mineral chemical, and whole-rock geochemical (major and trace element and Sr–Nd isotopic) data on rocks of both complexes, along with 40Ar–39Ar ages of 69.5–68.5 Ma on three Barda granophyres. Whereas silicic magmatism in the Deccan Traps typically postdates flood basalt eruptions, the Barda granophyre intrusions (and the Deccan basalt flows they intrude) significantly pre-date (by 3–4 My) the intense 66–65 Ma flood basalt phase forming the bulk of the province. A tholeiitic dyke cutting the Barda granophyres contains quartzite xenoliths, the first being reported from Saurashtra and probably representing Precambrian basement crust. However, geochemical–isotopic data show little involvement of ancient basement crust in the genesis of the Barda–Alech silicic rocks. We conclude that these rocks formed by advanced (70–75 %), nearly-closed system fractional crystallization of basaltic magmas in crustal magma chambers. The sheer size of each complex (tens of kilometres in diameter) indicates a very large mafic magma chamber, and a wide, pronounced, circular-shaped gravity high and magnetic anomaly mapped over these complexes is arguably the geophysical signature of this solidified magma chamber. The Barda and Alech complexes are important for understanding CFB-associated silicic magmatism, and anorogenic, intraplate silicic magmatism in general.

A petrological and geochemical study has been made of the late Precambrian St. David's granophyre (SDG) and the Cwm Bach rhyolites (CBR). The SDG is trondhjemitic, peraluminous, and chemically zoned with lower Na/K ratios in its interior portions. Although the SDG contains only minor amounts of K, the Rb content is relatively high and ratios of immobile versus mobile elements indicate that K has been removed. The CBR are flow-banded and autobrecciated subaerial rhyolite flows which are more K-rich and less altered than the SDG, lying close to the ternary minimum composition of the system Q-Or-Ab-(H2O). K mobility is considered to result from the reaction K-feldspar→ K-mica which released excess K ions, imparting peraluminous compositions to the rocks. On the basis of this reaction the original K content of the rocks has been estimated, the values for which are more compatible with immobile element abundances and with normal igneous trends. Similar abundances of the relatively immobile elements Ti, Zr, Sr, Y and Nb, together with K-corrected K/Rb ratios, suggest that the SDG and CBR may be comagmatic volcanic-arc-collision-type granite and rhyolite.

Primary almandine and spessartine-rich garnet and zinnwaldite phenocrysts occur along with feldspar (plagioclase and sanidine) phenocrysts, in the rhyolite of Profitis Ilias, which is located on the SE coast of the island of Chios, Greece. The distinctive mineralogical composition of this rhyolite is described. Although formed in the back-arc tectonic environment of the Aegean volcanic arc, the Profitis llias rhyolite shows significant trace element compositional differences when compared with typical arc or back-arc volcanic rocks of the area. It shows extreme depletion in Sr and Ba and enrichment in Nb and Mn, and has much more affinity with A-type granites and particularly Li-mica granites.

Apparently, both zinnwaldite and spessartine-rich garnet can be generated as primary phases from a granite melt enriched in volatile constituents at low P–T. This granite melt could be the residual product of an un-exposed, earlier formed, typical back-arc granite of the area, enriched in volatile constituents from a subcrustal source above the active mantle of the eastern Aegean area.

The extensive and deep faulting in the broad eastern Aegean lithosphere section would have facilitated the rapid ascent of that volatile-enriched granite melt, the parent of the Profitis Ilias rhyolite.

Cordierite is a characteristic mineral of many peraluminous felsic igneous rocks. A combination of T-P-X parameters, which overlap the stability conditions for felsic magmas, control its formation. Critical among these parameters are relatively low T, low P, and typically high (Mg+Fe2+), Mg/Fe2+, A/CNK, aAl2O3, and fO2. Spatial and textural information indicate that cordierite may originate in one of three principal ways in felsic igneous rocks: Type 1 Metamorphic: (a) xenocrystic (generally anhedral, many inclusions, spatial proximity to country rocks and pelitic xenoliths); (b) restitic (generally anhedral, high-grade metamorphic inclusions); Type 2 Magmatic: (a,b) peritectic (subhedral to euhedral, associated with leucosomes in migmatites or as reaction rims on garnet); (c) cotectic (euhedral, grain size compatibility with host rock, few inclusions); (d) pegmatitic (large subhedral to euhedral grains, associated with aplite-pegmatite contacts or pegmatitic portion alone); and Type 3 Metasomatic (spatially related to structural discontinuities in host, replacement of feldspar and/or biotite, intergrowths with quartz). Of these, Type 2a (peritectic) and Type 2c (cotectic) predominate in granitic and rhyolitic rocks derived from fluid-undersaturated peraluminous magmas, and Type 2d (pegmatitic) may be the most common type in fluid-saturated systems.

Dissolution rates of small forsterite spheres in superheated melts of basalt, andesite and rhyolite composition have been measured at 1300°C, atmospheric pressure. The rate is constant (83 µm hr−1) in the basalt, regardless of run duration. In the andesite the initial dissolution rate is 200µm hr−1, followed by a decrease to a constant value of 16µmhr−1 in 2–3 hours. Dissolution rate in the rhyolite decreases from an initial value of 1.7 to <0.1 µmhr−1 over 280 hours and never reaches a constant rate. Once the rate of dissolution has become constant, the film of contaminated melt that forms in melt about a crystal does not thicken with time, indicating attainment of a steady-state condition. Steady state is attributed to natural convection arising from the difference in density between the film of contaminated melt surrounding a crystal and that beyond. The density difference is approximately 2% of the density of the rock melt.

Calc-alkaline dacites from the Setouchi volcanic belt contain garnet crystals that are classified petrographically and chemically into two types: type I and type M. Type-M garnets are characterized by acicular sillimanite inclusions or dissolution textures, and may be accompanied by xenolith fragments. They exhibit extensive compositional zoning with an increase in MgO/FeO and decrease in MnO content towards the margin. These petrographical and compositional features are identical to those of garnets from metamorphic xenoliths entrained in the Setouchi volcanic rocks, suggesting a xenocrystic origin for the type-M garnets. In contrast, type-I garnets lack sillimanite inclusions and have different rim compositions from the type-M garnets. Transmission electron microscope analysis has revealed the presence of minute glass inclusions in the type-I garnets, which indicate conclusively that these garnets grew in the presence of a melt. Type-I garnets have oscillatory zoning characterized by an antipathetic variation between FeO and MgO. This zoning was probably caused by magma heterogeneity within magma batch. Differences in rim compositions between the two types of garnets, and the presence of reaction rims indicate that the xenocrystic type-M garnets were incorporated into the magma after phenocrystic type-I garnet became unstable due to decompression during magma ascent.

Uranoan thorite crystals have been found occurring as a sparse constituent in lithophysae in 6.1 to 6.8 Ma alkali rhyolite flows at Thomas Mountain, Utah, USA. The crystals are associated with sandidine, quartz, topaz, hematite, magnetite, and calcite; they are leek to dark grass green, transparent, well-formed, euhedral prisms, showing development of forms {100}, {101}, and {111}. The mineral is both optically isotropic, with a refractive index of 1.86(1), and X-ray amorphous, indicating its metamict state. Electron microprobe analysis yields: SiO2 17.3, ThO2 56.8, UO2 25.4, total 99.5%, and a structural formula of (Th0.72U0.31)Σ1.03Si0.97O4.

This appears to be the first reported occurrence of thorite in volcanic rocks of rhyolitic composition, and may account for part or all of the Th reported in alkali rhyolites and so-called ‘topaz-rhyolites’ of the western USA.

The upper 35 m of a thick (≤175 m) Early Jurassic North Mountain Basalt flow at KcKay Head contains 25 cm thick differentiated layers that are separated by 130 cm sections of basalt. The lower layers are mafic, pegmatitic, and contain thin (2 cm), fine-grained 'rhyolite' bands. Evidence that the rhyolite represents a Si-rich immiscible liquid includes: (1) textures such as fiine-grained globules of Ferich pyroxene (once Fe-rich liquid) bordering pegmatite feldspar grains; (2) structureless, microcrystalline, interstitial, polygonal patches of Si-rich minerals and similar areas of Fe-rich stilpnomelane surrounding skeletal Fe-Ti oxide grains, with bulk chemical compositions (to a first approximation), relative proportions and total modat percentages suggesting they were once Si-rich and Fe-rich glasses respectively; (3) basalt and pegmatite compositions (particularly their Fe, and Ti contents) similar to rocks known to contain immiscible liquids; (4) rhyolite major element compositions generally consistent with formation from an immiscible Si-rich liquid; (5) mineral compositions and temperature of pegmatite formation compatible with immiscibility; (6) the inability of mass balance calculations (crystal fractionation) to explain rhyolite formation unless mesostasis stilpnomelane (representing the Fe-rich liquid) is included in the caculations. If, as we suggest, these rocks are the result of immiscibility, they shed light on the incipient formation of granophyres in mafic intrusions and support liquid immiscibility as an important rock-forming process.

The long-distance transport of the bluestones from south Wales to Stonehenge is one of the most remarkable achievements of Neolithic societies in north-west Europe. Where precisely these stones were quarried, when they were extracted and how they were transported has long been a subject of speculation, experiment and controversy. The discovery of a megalithic bluestone quarry at Craig Rhos-y-felin in 2011 marked a turning point in this research. Subsequent excavations have provided details of the quarrying process along with direct dating evidence for the extraction of bluestone monoliths at this location, demonstrating both Neolithic and Early Bronze Age activity.

Silicic volcanism at c. 168 Ma has been identified previously on the Antarctic Peninsula, and the Mapple Formation, which includes those volcanic rocks, has been defined and documented from one area of the east coast of Graham Land. Based on age and geochemical criteria, correlations have been made to the extensive Chon Aike Province of South America, which has been demonstrated to be one of the largest silicic volcanic provinces in the world. Rhyolitic and intermediate composition volcanic successions from six separate localities on the east coast of the Antarctic Peninsula are described here and are confirmed as correlatives of the Mapple Formation, based on newly acquired geochronology and field observations. They are dominantly rhyolitic crystal tuffs and/or ignimbrites with ages in the interval 162–168 Ma, overlapping with the age of the Mapple Formation (167–171 Ma) at the type locality. Andesitic agglomerates are also described, which are included in the same event and demonstrate the occurrence of rare intermediate volcanism, which is also seen in the Chon Aike Province. A new group, the Graham Land Volcanic Group, is defined here, and criteria are established which allow the separation of some volcanic successions out of the previously defined Antarctic Peninsula Volcanic Group, which takes no account of tectonic setting, eruption age or geochemistry.

Many continental flood basalt provinces contain rhyolites with ‘A-type’ compositions and many studies have concluded that these higher silica rocks are crustal melts from metapelitic or tonalitic country rock. However, although many of the low-Ti continental flood basalt sequences exhibit a marked a silica gap from ∼55–65 wt. SiO2, many incompatible element ratios, and the calculated eruption temperatures (950–1100°C) are strikingly similar between the rhyolites and associated basalts. Using experimental evidence, derivation of the low-Ti rhyolites from a basaltic parent is shown to be a viable alternative to local crustal melting. Comparison of liquid compositions from experimental melting of both crustal and mantle-derived (basaltic) source materials allows the two to be distinguished on the basis of Al2O3 and FeO content. The basalt experiments are reversible, such that the same melts can be produced by melting or crystallisation. The effect of increased water content in the source is also detectable in the liquid composition. The majority of rhyolites from continental flood basalt provinces fall along the experimental trend for basalt melting/ crystallisation at relatively low water content. The onset of the silica gap in the rhyolites is accompanied by an abrupt decrease in TiO2 and FeO*, marking the start of Fe–Ti oxide crystallisation. Differentiation from 55–65 wt. SiO2 requires ∼30 fractional crystallisation in which magnetite is an important phase, sometimes accompanied by limited crustal contamination. The rapid increase in silica occurs over a small temperature interval and for relatively small changes in the amount of fractional crystallisation, thus intermediate compositions are less likely to be sampled. It is argued that the presence of a silica gap is not diagnostic of a crustal melting origin for either A-type granites or rhyolites in continental flood basalt provinces. The volume of these rhyolites erupted over the Phanerozoic is significant and models for crustal growth should take this substantial contribution from the mantle into account.

The Jurassic Mount Poster Formation of eastern Ellsworth Land, southern Antarctic Peninsula, comprises silicic ignimbrites related to intracontinental rifting of Gondwana. The identification of less voluminous basaltic and sedimentary facies marginal to the silicic deposits has led to a reclassification of the volcanic units into the Ellsworth Land Volcanic Group. This is formally subdivided into two formations: the Mount Poster Formation (silicic ignimbrites), and the Sweeney Formation (basaltic and sedimentary facies). The Mount Poster Formation rhyolites are an intracaldera sequence greater than 1 km in thickness. The basaltic and sedimentary facies of the Sweeney Formation are consistent with deposition in a terrestrial setting into, or close to, water. The geochemistry of the Mount Poster Formation is consistent with derivation of the intracaldera rhyolites from a long-lived, upper crustal magma chamber. The basalts of the Sweeney Formation are intermediate between asthenosphere- and lithosphere-derived magmas, with little or no subduction-modified component. The basalt could represent a rare erupted part of the basaltic underplate that acted as the heat source for local generation of the rhyolites. U–Pb ion microprobe zircon geochronology of samples from the Mount Poster Formation yield an average eruption age of 183.4±1.4 Ma. Analysis of detrital zircons from a Sweeney Formation sandstone suggest a maximum age of deposition of 183±4 Ma and the two formations are considered coeval. In addition, these ages are coincident with eruption of the Karoo-Ferrar Igneous Province in southern Africa and East Antarctica. Our interpretation of the Ellsworth Land Volcanic Group is consistent with the model that the Jurassic volcanism of Patagonia and the Antarctic Peninsula took place in response to intracontinental extension driven by arrival of a plume in that area.

Very large volume (>1000 km3 of magma) crystal-rich dacitic ignimbrites that lack pronounced evidence of fractional crystallization or vertical zonation erupt in some continental magmatic arcs (e.g. the Lund Tuff of the Great Basin and the Fish Canyon Tuff of Colorado in western USA). Apparently, their magma chambers were only modestly heterogeneous and not systematically zoned from top to bottom. These ignimbrites have 40 to 50% phenocrysts set in a high-silica rhyolite glass. Mineral assemblages and mineral compositions suggest pre-eruption temperatures were 730 to 820°C and water and oxygen fugacities were relatively high. We have speculated that these very large volume ignimbrites are unzoned because crystallization and convection in slab-shaped magma chambers inhibited separation of crystals from liquids and resulted in a chamber filled with compositionally heterogeneous magma that lacked systematic chemical zonation or strong fractionation. However, many other very large volume silicic ignimbrites are strongly fractionated and may be vertically zoned (e.g. tuffs related to the Yellowstone hotspot). These rhyolitic tuffs typically have few phenocrysts, anhydrous mineral assemblages, low oxygen fugacities, crystallization temperatures of 830 to 1050°C, and a strong imprint of fractional crystallization. Yet these Yellowstone-type rhyolites are derived from chambers 40 to 70 km across which have sill-like shapes (depth/diameter ratios much less than 1). Thus, factors other than chamber shape must be important for establishing the degree of evolution and nature of zonation in silicic magma chambers. Here, the role of crystallinity-dependent viscosity on the evolution of these two types of contrasting magmas is explored. Calculated magma viscosities for the hot, dry, crystal-poor rhyolites are significantly lower than for the cooler, wetter, crystal-rich dacites. Perhaps these hot rhyolites had low enough crystal contents and viscosities to allow efficient crystal–liquid separation, probably by a combination of unhindered crystal-settling, floor crystallization (including compaction), and crystallization on the walls of large chambers. Clean separation of melt from residual solids at their sources may have been promoted by their high temperatures and low viscosities (<104.5 Pa s). In contrast, monotonous dacitic magmas may never have been crystal-free near-liquidus magmas. Their large magma chambers may have developed by progressive growth at a shallow level with repeated input of intermediate to silicic magma. Crystallization of the water-enriched dacitic magmas occurred at lower temperatures (<800 °C) where crystallinity and hence magma viscosity (>106.5 Pa s) were significantly higher. These characteristics inhibited all forms of crystal–liquid separation, hindered development of systematic vertical zonation, and promoted quasi-equilibrium crystallization in small domains within large heterogeneous magma chambers. Eruptions of these crystal-rich dacites may only occur if the roof fails over a growing magma chamber that is becoming increasingly molten.

Rhyolitic volcanism along the proto-Pacific margin of Gondwana occurred at intervals throughout the Jurassic. Silicic melt generation has been interpreted as a result of interaction between mantle plumes and subduction modified lower crust. The rhyolitic Mount Poster Formation of the southern Antarctic Peninsula is c. 184 Ma in age (V1), whereas silicic volcanism of the northern Antarctic Peninsula is c. 168 Ma (V2). A thin, (13.5 cm) reworked air fall tuff, interbedded with sandstone and mudstone of the Latady Formation in the southern Antarctic Peninsula has a REE pattern similar to V2 volcanic material but is isotopically similar to the extracaldera, low-Ti rhyolites of the V1 Mount Poster Formation. The tuff is interbedded with lithofacies that have been assigned a Callovian age (164–159 Ma) in the west of the area. Simple mixing between a MASH source and reworked Early Jurassic (184 Ma) V1 volcanic material during V2 volcanism in the area explains the apparent discrepancy between the faunal age and the isotopic characteristics of the ash fall. This supports a Middle Jurassic (168 Ma) age that also corresponds to a 167 ± 3 Ma age from Mt Rex on the periphery of the Mount Poster Formation, which was previously thought to be anomalous.

Viscosities of liquid albite (NaAlSi3O8) and a Himalayan leucogranite were measured near the glass transition at a pressure of one atmosphere for water contents of 0, 2·8 and 3·4 wt.%. Measured viscosities range from 1013·8 Pa. s at 935 K to 109·0 Pa. s at 1119 K for anhydrous granite, and from 1010·2 Pa. s at 760 K to 1012·9 Pa. s at 658 K for granite containing 3·4 wt.% H2O. The leucogranite is the first naturally occurring liquid composition to be investigated over the wide range of T-X(H2O) conditions which may be encountered in both plutonic and volcanic settings. At typical magmatic temperatures of 750°C, the viscosity of the leucogranite is 1011·0 Pa. s for the anhydrous liquid, dropping to 106·5 Pa. s for a water content of 3 wt.% H2O. For the same temperature, the viscosity of liquid NaAlSi3O8 is reduced from 1012·2 to 106·3 Pa. s by the addition of 1·9 wt.% H2O. Combined with published high-temperature viscosity data, these results confirm that water reduces the viscosity of NaAlSi3O8 liquids to a much greater degree than that of natural leucogranitic liquids. Furthermore, the viscosity of NaAlSi3O8 liquid becomes substantially nonArrhenian at water contents as low as 1 wt.% H2O, while that of the leucogranite appears to remain close to Arrhenian to at least 3 wt.% H2O, and viscosity–temperature relationships for hydrous leucogranites must be nearly Arrhenian over a wide range of temperature and viscosity. Therefore, the viscosity of hydrous NaAlSi3O8 liquid does not provide a good model for natural granitic or rhyolitic liquids, especially at lower temperatures and water contents.

Qualitatively, the differences can be explained in terms of configurational entropy theory because the addition of water should lead to higher entropies of mixing in simple model compositions than in complex natural compositions. This hypothesis also explains why the water reduces magma viscosity to a larger degree at low temperatures, and is consistent with published viscosity data for hydrous liquid compositions ranging from NaAlSi3O8 and synthetic haplogranites to natural samples. Therefore, predictive models of magma viscosity need to account for compositional variations in more detail than via simple approximations of the degree of polymerisation of the melt structure.

Isobaric crystallisation paths obtained from phase equilibrium experiments show that, whereas in rhyolitic compositions melt fraction trends are distinctly eutectic, dacitic and more mafic compositions have their crystallinities linearly correlated with temperatures. As a consequence, the viscosities of the latter continuously increase on cooling, whereas for the former they remain constant or even decrease during 80% of the crystallisation interval, which opens new perspectives for the fluid dynamical modelling of felsic magma chambers. Given the typical dyke widths observed for basaltic magmas, results of analogue modelling predict that injection of mafic magmas into crystallising intermediate to silicic plutons under pre-eruption conditions cannot yield homogeneous composition. Homogenisation can occur, however, if injection takes place in the early stages of magmatic evolution (i.e. at near liquidus conditions) but only in magmas of dacitic or more mafic composition. More generally, the potential for efficient mixing between silicic and mafic magmas sharing large interfaces at upper crustal levels is greater for dry basalts than for wet ones. At the other extreme, small mafic enclaves found in many granitoids behave essentially as rigid objects during a substantial part of the crystallisation interval of the host magmas, which implies that finite strain analyses carried out on such markers can give only a minimum estimate of the total amount of strain experienced by the host pluton. Mafic enclaves carried by granitic magmas behave as passive markers only at near solidus conditions, typically when the host granitic magma shows near-solid behaviour. Thus they cannot be used as fossil indicators of direction of magmatic flow.