Though typically exhibiting considerable scatter, geochemical variations in granitic plutons and silicic volcanic deposits are commonly modelled as products of differentiation of originally homogeneous magmas. However, many silicic igneous bodies, particularly those classified as S-types, are internally heterogeneous in their mineralogy, geochemistry and isotope ratios, on scales from hundreds of metres down to one metre or less. The preservation of these heterogeneities supports recent models for the construction of granitic magma bodies through incremental additions of numerous batches (pulses) of magma derived from contrasting sources. Such pulses result from the sequential nature of the melting reactions and the commonly layered structure of crustal magma sources. Internal differentiation of these batches occurs, but not generally on the scales of whole magma chambers. Rather than being created through differentiation or hybridisation processes, at or near emplacement levels, much of the variation within such bodies (e.g. trace-element or Mg# variation with SiO 2 or isotope ratios) is a primary or near-source feature. At emplacement levels, the relatively high magma viscosities and slow diffusion rates of many chemical components in silicic melts probably inhibit processes that would lead to homogenisation. This permits at least partial preservation of the primary heterogeneities.

Abstract A new fracture-mediated intrusion model resolves the sequence of magma and rock displacements generating a felsic magma system with a lower crustal source, central conduit and shallow sill pluton. Idealized intraplate conditions are assumed, to neglect regional tectonism and to focus on juvenile cracking by magma-intrinsic hydraulic and buoyant loads. The magma source is conductively heated and develops by endothermic fluid-absent melting in approximately 10 6 years. The idealized domical thermal anomaly and endothermic heat focusing yield a low aspect ratio source, with outer-porous and inner-permeable partial melt zones. An anatectic core region is unrealized owing to magma segregation. Thermal stresses are readily relaxed and unimportant to source loading while crustal uplift generates tensile stress and, upon relaxation, lateral space for tensile fractures. Dilative melting generates buoyancy overpressure (Δ P B ) and a hydraulic contribution (Δ P V ) to the magma pressure ( P M ). Δ P V develops by elastic wall-rock compression as the ‘excess magma volume’, EMV , arises too abruptly for full relaxation by inelastic deformation, inducing a brittle response. Tensile rupture criteria are met in an effective tensile stress field with low differential stress induced by magma pore pressure and wedging by pressurized cracks, which initiate by source inflation and uplift. Preferred vein geometry reflects the starting stress field. For symmetric doming, radial vertical cracks with a central nexus form a natural conduit. A vertically extensive crack system, however, requires special explanation because wedging by Δ P V reorients dykes to sills just above source. The solution is that volumetric crack growth accommodating non-relaxed EMV ( EMV *) causes Δ P V →0. Magma transport becomes buoyancy-driven and the Δ P V problem does not arise. The critical sill intrusion depth, I , is where Δ P B exceeds the regional vertical stress curve, where columns must intrude owing to Δ P B alone. Sill growth is mainly by floor depression, involving ductile shear of lower crust, creating sill volume, suppressing roof uplift, expelling source contents, processing protolith through the melting zone, reducing stress, σ H , and widening conduits for sustained flow. Two intrusive regimes are identified; Δ P V >0 (hydraulic inflation) and Δ P V =0 (buoyancy pumping). Partitioning between three sinks for EMV – inelastic uplift (φ), crack growth (η) and a non-relaxed portion ( EMV *) generating Δ P V – defines four hydraulic subregimes. Disequilibrium dilation occurs during crustal relaxation prior to rupture, when η=0 and EMV partitions between φ and EMV *. Uplift occurs readily owing to the crust's weakness in flexure, so φ abruptly increases while EMV * decreases, causing abrupt variations in source failure mode, geometry and rate that smooth initial Δ P V variations. During equilibrium dilation source swelling continues with φ dominant over EMV *. Dilatant loading rates mean that positive Δ P V is always maintained however, keeping the source near-isotropically inflated and prepared for rupture. Disequilibrium cracking begins when uplift-driven horizontal stretching initiates rupture and crack growth (η). Crack volume is initially small, but readily enlarges as dykes propagate by conversion of stored Δ P V . EMV is minimized better and faster by η than by φ, owing to crack-tip stress concentration, giving abrupt augmentation of η and decreases in φ and EMV * in a crack growth-surge until the uplift-modified stress field is balanced. In equilibrium cracking , once EMV * (and Δ P V ) decrease to incipient levels, each new increment of EMV * partitions directly into crack growth, while continued uplift maintains vertical rupture, generating a vertically extensive fracture system. The absolute volumetric equivalence of EMV *, at most a few tens of km 3 , will be exhausted during dyking, sill intrusion or surface eruption. The system then becomes buoyancy-driven and, if depth I is reached, must intrude a sill. Relaxation of sill underburden initiates crustal decoupling and buoyancy pumping, where the downward underburden-flux drives and is balanced by upward magma flow.

Abstract It is probable that granitic magma ascent does not result from the intrinsic properties of the magmas. Within the uppermost crust, neither the reduced viscosity nor the density contrast between magma and surroundings are themselves sufficient to induce either low-inertia flow (diapirism) or fracture-induced magma propagation (dyking). Igneous diapirism is intrinsically restricted to the lower, ductile crust. Dyking is therefore the most probable ascent mechanism for granitic magmas that reach shallow crustal levels. A neutral buoyancy level in the crust, at which magma ascent should stall, is never observed. This is demonstrated by coeval emplacement of magmas with different compositions and densities, and the negative gravity anomalies measured over many granitic plutons. We suggest that deformation, through strain partitioning, is necessary to magma ascent. Pluton formation is controlled by local structures and rock types rather than by intrinsic magma properties. As a result of its intermittent character, deformation (both local and regional) induces magma pulses, and this may have important consequences for the chemical homogeneity of intruded magmas.