Volcanoes provide a wonderful laboratory for the study of crystallization. Crystallization, in turn, changes magma and lava rheologies, and may thus affect the course of volcanic eruptions. Magmas crystallize when timescales of decompression and volatile loss (in conduits) or of cooling (during surface flow) exceed kinetic limits to crystal nucleation and growth. Rates of crystallization are dictated by magma viscosity (diffusion rates) and the driving force provided by supersaturation. Crystal textures (total crystallinity, crystal number density, size, shape and spatial distributions) preserve information on the conditions under which crystallization occurred and, when calibrated, may provide quantitative constraints on timescales of magma ascent and lava flow cooling.
The addition of crystals to a silicate melt will increase viscosity, thus altering the ability of the suspension to flow. If crystallization is extensive, adjacent crystals will interact sufficiently to create a yield strength, and flow will occur only by increasing applied stress. While the effects of increasing viscosity and yield strength on rates and styles of lava effusion have long been recognized, the relationship between crystal microstructure and yield strength development is poorly understood, and represents an important area of future research. What of the consequences of crystallization-related viscosity increases and yield strength development in volcanic conduits? Degassing-induced crystallization accompanies the slow emplacement of lava domes of all compositions. However, higher syn-eruptive crystallinities are achieved in lower viscosity (more mafic) melts. Thus ascent of lava during obsidian dome formation typically results in < 10% very small (<10 µm) microlites, while slow ascent of basaltic andesite (e.g., Merapi, Indonesia) may produce up to 50% microlites and microphenocrysts (< 100 µm) and substantially affect magma rheology. Syn-ascent crystallization also occurs during explosive sub-plinian eruptions, which are typically pulsatory and produce juvenile pyroclasts with a wide range of densities and microlite crystallinities. Again, the extent of crystallization appears to be dictated by melt viscosity, thus the predominance of microlite-free pyroclasts in rhyolite and dacite eruptions contrasts with the highly crystalline pyroclasts produced during explosive eruptions of basaltic andesite (e.g., 50-80% at Mt. Spurr, Alaska, 1992). The degree to which extensive syn-eruptive crystallization modulates the eruptive process itself remains a key question in the study of intermediate-intensity volcanic eruptions, and the answer to this question is likely to have important consequences for improved monitoring of active volcanic systems.
We used experimental and numerical simulations to study the growth of water bubbles in silicic melts under variable pressure. We hydrated samples of rhyolitic obsidian at 150 MPa (850°C) and decompressed them to 110 MPa (700° C), where bubbles were allowed to grow to their final equilibrium radius. The bubble-bearing samples were quenched, photographed, and then re-loaded to 110 MPa, 700°C. After reaching thermal and mechanical equilibrium we released pressure gradually, from 110 MPa to 70 MPa during 32 seconds, and quenched again. We re-examined the samples under the microscope, compared the initial and final radii of individual bubbles, and documented the number and distance of neighboring bubbles. We found no new bubbles. To simulate our experiments, we used the model of Lyakhovsky et al. (1995), that follows the growth of water bubbles from finite shell of melt and modified it to include the effects of variable pressure and the dependence of viscosity and diffusivity on both temperature and concentration of water. We simulated individual bubbles and used the measured initial radius, calculated initial separation between bubbles (which agrees closely with average measured separation), and the path of decompression vs. time. Calculated radii are in excellent agreement with measured radii for a wide range of bubble sizes. Small bubbles closely follow the equilibrium growth-path (e.g., initial radius 5.5 µm; measured 8.0 µm; calculated 8.6 µm; and final equilibrium radius 9.0 µm). On the other hand, large bubbles are far from equilibrium radius but closely agree with the radius predicted by the model (e.g., initial 35 µm; measured 42 µm; calculated 43 µm; and final equilibrium radius 57 µm). This deviation reflects the less efficient diffusive transfer of water from the larger melt shells around the large bubbles. These results, together with previous success of the model in simulating growth under constant pressure (Lyakhovsky et al., 1995; Navon et al., 1998) assure the accuracy of our mathematical model. The model is now applied to examine the effect of bubble growth on bulk viscosity and the equation of state of vesiculating magma.
Lyakhovsky V, Hurwitz S & Navon O, Bull. Volcanol., 58, 19-32, (1995).
Navon O, Lyakhovsky V & Chekhmir A, Earth Planet Sci. Lett., 160, 763-776, (1998).
Crystals showing textures of rapid grow rates are common in many rock-types: pillow-lavas, chondrules, tektites, comb-layered rocks, xenoliths, komatiites... Olivine is the most studied quench mineral but at the moment there is not a coherent model of growth process. We are studying quench textures of forsterite in the CMAS system. The experimental techniques used in this study are the same as those used in dynamic crystallization. The charges (glass) are first heated above the liquidus and then cooled at a constant rate varying from 10°C/h to 1590°C/h. At the end of the experiment, the samples are quenched by dropping them into water. The charges are mounted in epoxy and doubly polished thin sections are prepared for optical, SEM and TEM study. The shape of crystals is not controlled by the cooling rate but by the degree of undercooling (-T). There are only two main shapes: hopper and swallow-tail. Other morphologies observed by previous authors (chain, lattice, plate, branching and feather olivine) are sections cut from those two shapes. Hopper crystals are elongated parallel to the a direction with an hollow core filled by glass and formed by (100), (021) and (010) faces. When (-T>100°C, the hopper shape persist but the corners, formed by intersection of {021} faces develop dendrites. Dendrites apparently grow along two directions: [101] and [-101]. They are made of numerous parallel units in the (010) plane. These units are similar to hopper crystal except that the (001) face is present. These units are connected along the (001) face and then actual growth proceeds in 3 directions. Ostwald ripening alter quickly units to form H-shape and hollow tube at the end.
Quantification of volcanic textures has different goals because it not only provides insights into the dynamics of explosive and effusive eruptions but is also a powerful tool to constrain the physical parameters necessary to develop reliable eruptive models. We have investigated the textural characteristics of the white and gray pumice from the tephra deposit of the 1991 Pinatubo eruption using image analyses techniques. Textural parameters which have been measured are 2D clast vesicularities, crystallinities, vesicle size, shape and orientation distributions, degree of vesicle coalescence and vesicle wall thickness. The main differences between the two pumice types can be summarized as in the following: a) white pumice has higher vesicularity and vesicle number density, elongated and highly interconnected vesicles, thin vesicle walls, presence of euhedral phenocrysts and microphenocrysts and no microlites in the groundmass; b) gray pumice has lower vesicularity and vesicle number density, thick vesicle walls, less elongated and coalesced vesicles, abundant broken phenocrysts and microphenocrysts with solution pitting surfaces, ubiquitous microlites (<10 µ) and crystal fragments in the groundmass. The results of this study have been integrated and compared with both field observations and conduit flow modeling. The detailed analysis of pumice microstructure lead us to highlight the likely occurrence of magma ascentmechanisms which have been poorly addressed so far, such as viscous dissipation, magma zonation and mechanical breakage of crystals at the conduit walls. These processes may be responsible for the contemporary occurrence and discharge of a white and gray magma along the conduit of this explosive eruption.
Muscovite occurs as large (average 3 cm) phenocrysts and defines the so-called Type III granite of the Northern Unit of the Leinster Batholith, S.E. Ireland. The Type III occupies a central elavated position with shallow or flat gradational contacts to other two-mica granite varieties. The muscovite phenocrysts themselves display optically complex growth, dissolution and inclusion textures. These must necessarily be interpreted in the light of all other avialable facts. Perhaps of most significance is isotopic work on this granite that demonstrates a cryptically preserved, systematic, inherited inhomogeneity that seems to rule out large-scale convection or fractionation as having operated; the pattern seeming to be an image of its source. The muscovite phenocrysts show individual stratigraphies but a generality emerges of a complex early growth, a mid-way major corrosion surface and quiescent late growth. The major corrosion event is almost universally seen and marks a boundary between distict magmatic events for this granite. Examination of this event in muscovite, comparison with associated plagioclase growth stratigraphies and taking into account the isotopic and field evidence indicates an influx of hot water (volatiles) may be responsible and could mark the 'moment' of intrusion. If so then this has knock-on implications for the pre- and post-intrusion history of the granite as recorded in the muscovite and other magmatic minerals.
High-resolution X-ray computed tomography (CT) of igneous rocks provides a rapid, non-destructive means to acquire textural data and imagery in three dimensions. The digital nature of CT data facilitates computer-aided visualization, opening new possibilities for exploration and interpretation of 3-D textural relationships. The numerical character of CT data makes it inherently rich in the metrical information required for quantitative characterization and statistical analysis of the sizes, shapes, locations, and interrelations of textural components in igneous rocks.
CT measurements produce 3-D maps of spatial variations in the average linear attenuation coefficient for the input X-ray spectrum, a quantity that depends strongly on mass density but also on mean atomic number. Differences of a few percent in these properties yield visible contrast in a CT image, allowing discrimination between minerals as similar as quartz and feldspar. Instrumentation at the University of Texas High-Resolution CT Facility combines both a high-energy source and detectors for analysis of cm- to dm-scale specimens with spatial resolution at sub-mm scale, and a microfocal source and low-energy detector for analysis of mm- to cm-scale specimens at spatial resolutions of tens of micrometers. Visualization of the data profits from the ability to view arbitrarily oriented sections through the 3-D volume, and from the ability to selectively extract features of interest and display perspective views of them using methods of isocontouring or volume rendering, which can then be animated to further enhance their impact.
Examples of applications to textural analysis of igneous rocks are diverse. In slowly cooled basalt, CT images reveal plagioclase crystals linked in monomineralic chains that form continuous 3-D networks. CT-based measurements of size and shape distributions of vesicles in basalts yield information on the principal driving forces for volcanic eruptions by revealing details of bubble growth, nucleation, and coalescence. Diamonds in Siberian eclogites are seen in CT imagery to have a close spatial association with a 3-D network of subplanar alteration zones, implying that the diamond does not have an igneous origin, but is instead a product of later metasomatism. Connectivity and geometry of leucosome distributions in migmatites constrain melt flow paths during anatexis. Comingling of compositionally dissimilar magmas occurs across a 3-D interface that exhibits delicate interfingering of the contrasting magmas at all observed length scales. Determinations by CT of relative masses and volumes of both chondrules and metal-troilite particles in mesosiderite meteorites constrain processes of sorting in the solar nebula. CT images of the 3-D distribution of metal veins and particles in lodranite meteorites, made without destructive sectioning, define scales and mechanisms of transport during silicate-metal segregation, a process likely to operate during core-mantle differentiation.
Quantitative information on processes below the brittle-plastic transition that lead to heat and mass transport is limited, because only minimal data are available concerning transport properties of crustal melts and movement of magma under conditions appropriate to lower crust. Permeability of partially molten crustal materials is one factor that controls rate of melt segregation and magma migration, but this parameter is not well constrained for the dynamic conditions that characterize natural melting during orogenesis. To address this issue we have determined the geometry of leucosome in two migmatites formed by syntectonic anatexis: a stromatic migmatite derived from pelite, comprising sheets of leucosome (quartzo-feldspathic layers with garnet) with walls of melanosome (biotite-rich selvedges) in schistose mesosome; and, a migmatitic garnet-amphibolite derived from basalt, comprising spindle-shaped quartzo-feldspathic leucosomes, associated with garnet, in melanosome (hornblende and quartz, ±clinopyroxene). Three-dimensional images were generated from two-dimensional representations of spatial data obtained by serial grinding and HR X-ray CT.
Projections of three-dimensional images of the stromatic migmatite show the nature of the planar leucosomes throughout the sample. In the image derived from HR X-ray CT data, garnet in the leucosome is only rarely in contact with the melanosome, which suggests leucosome garnet was suspended in melt during flow. If the walls of these planar conduits had impinged on the garnet, in-plane tortuosity would have increased as melt would have had to flow around the pinned crystals.
Projections of three-dimensional images of the migmatitic garnet-amphibolite do not reveal the full extent of leucosome connectivity. Connectivity in this sample can be shown, however, by virtual reslicing of three-dimensional images perpendicular to the plane of the two-dimensional representations (approximately parallel to the lineation defined by leucosome), and by using three-dimensional projections of a single leucosome connectivity 'tree' constructed by projecting leucosome patches from slice to slice and noting overlap. Based on leucosome geometry and volume, we estimate effective porosity in this sample to have been 20 vol.% at stagnation, which may represent the minimum for significant melt flow through this network. Leucosome in the migmatitic garnet-amphibolite occurs in strain shadows around garnet, which were obstacles to melt flow along a linear path. Blocking of inferred flow channels by garnet contributes to the high tortuosity of this sample (values of 2-6 derived from leucosome geometry). Individual inferred flow paths have highly variable cross sectional areas (by 2-3 orders of magnitude), which imply strong local flow divergences, and minimum channel apertures (by a factor of up to 50), which means that lineation-parallel pore non-uniformities were significant. Based on these data, unusually straight channels would have dominated the mesoscopic melt flux through this rock when partially molten, because the velocity of melt in these highly-conductive flow paths will be larger.
Ideal populations of crystals in viscous magmas display cyclic shape preferred orientations (SPO). In theory, this should allow the determination of paleo-flow directions by the detection of obliquities between grain populations of different shape. However, analogue experiments and simple calculations indicate that the cyclicity of different shape populations will be out of phase, resulting in "steady state" fabrics with uniform average preferred orientation. This effect is further amplified by the continuous appearance of new grain populations in the magma due to crystallization, and/or by interaction between grains. The low intensity of the resulting SPO therefore necessitates physical quantification using methods such as the anisotropy of magnetic susceptibility, and/or image analysis of mineral sub-fabrics on mutually perpendicular faces and subsequent 3D shape tensor determinations.
These approaches were used to study SPO sub-populations of crystals of different minerals in suites of plutonic rocks from the Lebel syenite, Ontario, Canada, the Dinkey Creek granodiorite, Sierra Nevada batholith, California, and gabbros of the Bushveld Complex, South Africa. In each case mineral SPO's always display low intensities but very stable orientations from site to site. However, the significance of such observations becomes uncertain when the SPO is modified or even destroyed by recrystallisation. In such cases it is necessary to constrain the crystallisation history of the rock in order to verify that its texture still retains information from the magmatic flow stage. This is evaluated by measurements of crystal size distributions (CSD) for each mineral population in the samples. We will pay a particular attention to CSD patterns that can indicate physical processes such as crystal sorting during flow, which can induce various obliquities between SPO's and compositional layering.
The shape of isolated crystals may evolve towards textural equilibrium by way of three independent processes: solid-state creep, surface diffusion or dissolution/precipitation. The driving force of creep is surface stress, and it is resisted by internal friction. The driving force of surface diffusion is chemical-potential gradient across the crystal surface, and it is resisted by two dimensional Fick's law. The driving force of dissolution/precipitation is chemical-potential gradient through the crystal surface, and it is resisted by interface kinetics and/or by three dimensional Fick's law in the embedding medium.
Textural equilibration processes are modelled by partial differential equations. The differential equations are reduced to non-dimensional form then solved numerically. Any initial crystal shape may be considered but simple shapes (corner, ellipse, etc.) have more predictive impact. Anisotropy of interface energy is taken into account with no additional difficulty.
Equilibration kinetics is influenced by the initial shape and size of crystals, and by temperature. Creep is probably never important for the textural equilibration of silicate crystals. Surface diffusion may be the main process for small grain (< 1 mm) at high temperatures (> 500°C). Dissolution/precipitation is the most efficient process for larger grains or lower temperatures, if there is sufficient solubility of the crystal constituents in the embedding medium.
A series of experiments on quartz in hydrous granitic liquid at 900°C - 1 GPa is being conducted in order to check the numerical simulation and to quantify the kinetics of Ostwald ripening.
The 930-920 Ma old Rogaland anorthositic province of southwest Norway is part of the Proterozoic Sveconorwegian/Grenvillian orogenic belt. Most of the Rogaland complex is occupied by massif-type anorthosite and by a large anorthosito-charnockitic layered intrusion, the Bjerkreim-Sokndal massif (BKSK). Diapirism is thought to be the main emplacement mechanism of the Rogaland anorthosite massifs and gravity-induced subsidence has been described in the BKSK lower part made up of anorthosito-noritic cumulates.
A structural study using the anisotropy of magnetic susceptibility method (AMS) has been conducted on BKSK. This study focused on massive charnockitic acidic rocks caping the BKSK layered rocks and on a contemporaneously intruded and (highly) solid-state deformed BKSK apophysis (the so-called Apophysis). The magnetic mineralogy survey shows that titanomagnetite is governing the magnetic fabrics of the BKSK various petrographic rock types. The mutual variations of the AMS parameters (mean magnetic susceptibility, anisotropy degree, shape parameter) have clear analogies with these observed in magnetite-bearing, non charnockitic granitoids. Combination of petro- and magnetofabrics indicates a converging flow of material across all BKSK lithological boundaries towards a central funnel-shaped trough formed by the interference of the metamorphic envelope and three anorthosite diapirs. This pattern together with field and microstructural data demonstrate that the influx and the crystallization of the several magma batches giving the different stratigraphical BKSK units took place in a magma chamber deforming by gravity-induced subsidence of its floor (sagduction). This BKSK sagging was concomitant with the development of a (gravity-induced) instability zone at the contact between an anorthosite diapir and the granulitic envelope, favouring emplacement of the Apophysis material.
In the general framework of gravity-induced tectonic, BKSK thus represents a nice example of rim-syncline, along diapiric anorthosite massifs, produced by a return sagduction flow of material.
Evidence of magmatic flow (Paterson et al. 1989) includes: (a) parallel to sub-parallel alignment of elongate euhedral crystals (e.g., of feldspar and hornblende) that are not internally deformed, (b) imbrication ("tiling") of elongate euhedral crystals that are not internally deformed, (c) insufficient solid-state strain between aligned or imbricated crystals, (d) elongation of microgranitoid enclaves without plastic deformation, (e) magmatic flow foliations and elongate microgranitoid enclaves deflected around xenoliths, and (f) schlieren layering (if due to flow sorting) in the absence of plastic deformation of the minerals involved. These features are consistent with rotation of crystals in a much weaker medium, namely a melt, at a stage when the magma had become viscous enough to preserve the alignment.
Evidence of solid-state flow includes: (a) internal deformation and recrystallization of grains, (b) recrystallized "tails," (c) elongation of recrystallized aggregates (e.g., of quartz and mica), (d) grainsize reduction, (e) fine-grained folia anastomosing around less deformed relics, (f) microcline twinning, (g) myrmekite, (h) flame perthite, (i) boudinage of strong minerals, typically with recrystallized aggregates of weaker minerals (e.g., quartz and mica) between the boudins, (j) foliation passing through, rather than around enclaves, and (k) heterogenous strain with local mylonitic zones.
Recent experimental studies (reviewed by Paterson et al., 1998) indicate that a change from grain-supported flow to suspension flow occurs in deforming magmas between 20% to 40% melt, and that large amounts of strain may accumulate in magmas without being recorded by the final fabric. At lower melt percentages (perhaps as low as a few percent depending on the minerals and their shapes), strain may be accommodated by: (a) melt-assisted grain-boundary sliding, (b) contact melting-assisted grain boundary migration, (c) strain partitioning into melt-rich zones, (d) intracrystalline plastic deformation (c-slip in quartz indicating plastic deformation at near granite solidus temperatures), and (f) transfer of melt to sites of low mean stress. The only indication of strain in the absence of crystal plasticity may be a "magmatic-looking" fabric (e.g., Nicolas, 1992). Moreover, Park and Means (1996) emphasized that some deformation mechanisms are difficult to infer from preserved microstructures. In addition, deformation mechanisms previously inferred to reflect subsolidus deformation are active during near-solidus, grain-supported flow, so that magmatic flow microstructures may be destroyed by fracturing, crystal plasticity and recrystallization, even before the magma reaches its solidus.
Evidence of magmatic flow may be preserved in deformed metamorphic rocks (e.g., alignment of K-feldspar megacrysts). However, absence of alignment does not necessarily preclude a magmatic origin for K-feldspar megacrysts in felsic gneisses, because magmatic flow could have ceased before the magma became viscous enough to preserve an alignment.
Nicolas A, Journal of Petrology, 33, 891-915, (1992).
Park Y & Means WD, Journal of Structural Geology, 18, 847-858, (1996).
Paterson SR, Fowler TK, Schmidt KL, Yoshinobu AS, Yuan ES & Miller RBES, and Miller, RB, Lithos, 44, 53-82, (1998).
Paterson SR, Vernon RH & Tobisch, OT, Journal of Structural Geology, 11, 349-363, (1989).
Granitoïdes des Planètes Telluriques, Département des Sciences de la Terre, Université de Paris-Sud, F-91405 Orsay Cedex France
Textures of hybrid rocks produced by magma mixing-mingling have been examined in the gabbro-granite association of Porto (Corsica), taken as an example of incomplete processes in which small volumes of magmas were involved and crystallised rapidly under the influence of high chemical, thermal and viscosity contrasts. Homogeneous hybrid rocks and heterogeneous magmatic breccias were generated by hybridisation processes. The abundance of hybrid rocks is rather low (2%) and the mean ratio of the mafic component incorporated into the hybrid rocks (mixing ratio) is 26%, with a large standard deviation of 20%. No hybrid rocks with mixing ratios higher than 49% have been observed, showing mixing did not produce all the compositions possible on the mixing line. According to Sparks and Marshall (1986), complete chemical mixing can happen only if the mafic component exceeds 50%. This was not the case in the hybrid rocks. The low mixing ratios calculated for the hybrid matrix of the magmatic breccias and their highly heterogenous texture afford further evidence against thorough mixing. Both gabbro and subsolvus granite end-members are porphyritic, indicating that the two liquids were below their liquidus temperatures when they came into contact. Hybrid rocks and magmatic breccias yield rounded quartz and plagioclase, issued from the granitic magma, and fine-grained to aplitic to subvariolitic textures, with crystals evolving from acicular to skeletal in the groundmass. Partially remelted phenocrysts and quench textures suggest that hybridisation was basically a disequilibrium process. Textural relationships are interpreted as the result, first, of superheating of the granitic magma and, then, of supercooling of the hybrid liquids increasing as a function of mixing ratio. The textural evolution can be portrayed in a mixing binary system where: (i) sudden reheating induced partial resorption of the granitic metastable phenocrysts within the hybrid liquids, (ii) when thermal equilibration was reached, supercooling increased as a function of mixing ratio and prevented important chemical mixing and favoured the development of quench textures, (iii) because complete mixing could not operate, the mixing ratios (26 ± 20%) yielded by the hybrid rocks can be at least three times lower than the actual ratio between the mafic and granitic liquids involved in the mingling process.
In order to understand microtextures in magmatic and metamorphic rocks, the knowledge of the nucleation and growth histories of the rock forming minerals are necessary. This is most important in partially molten migmatites, which have a solid (metamorphic) and a liquid (magmatic) component. The understanding of crystallisation processes in those rocks will also contribute to distinguish between solid state and liquid processes. Therefore, the microstructures of cordierite and feldspar inside migmatites have been quantified using crystal size distributions (CSD) and by using quantitative characterisation of grain boundary shapes.
The investigated cordierites are a product of a melt producing reaction and crystallise in the presence of melt. Those crystals show a straight-line CSD in a ln h(L) versus L diagram, indicating linear nucleation- and growth rates (L= 3D length of crystals; calculating with the program "StripStar" by R. Heilbronner; h(L) = frequency of L). Those shaped CSD have been mainly described in magmatic rocks, in contrast to log-normal distributions, which are commonly reported for metamorphic porphyroblasts. This indicates, that crystallisation in those migmatites are more similar to magmatic processes. However, the cordierites can also seen as the product of a metamorphic reaction. For the reaction kinetics the linear shaped CSD indicate that nucleation and growth are a rate controlling step during melt producing reactions. This is contrast to most metamorphic porphyroblasts, where crystal growth is mainly diffusion controlled.
Additionally, the crystallisation of segregated melt inside of leucosomes have been investigated. Alkalifeldspar-quartz leucosomes with a cotectic composition represent most likely segregated and crystallised melt. Inside segregated melt homogeneous nucleation and magmatic crystal growth is most likely. Those leucosomes show a "overproduction" of large grains in a ln h(L) versus L diagram. The numerous large grains indicate a change in nucleation rate in the early part of crystallisation history. In cases of decreasing undercooling in early state of the crystallisation history, nucleation rate will decrease (or stopped), but growth of the crystals can be continue. This would produce the observed CSD. The change in undercooling may caused by the heat input form the latent heat of fusion. The latent heat of fusion may influence undercooling in cases without a significant temperature hiatus between melt layer and country rock (present in migmatites). After a certain time interval heat production and heat flow is equilibrated and constant nucleation- and growth rates occur. They are responsible for the straight-line CSD in the small and middle size classes.
Experimental studies predict that partial melting of metapelitic lithologies will occur under the pressures, temperatures and low water-activity conditions characteristic of granulite-facies metamorphism. Major melt producing reactions are volatile-phase absent and produce water-undersaturated granitic melts across reaction slopes characterised by positive volume changes. Volume changes are small - fluid absent melting of muscovite produces a total positive volume change of only 2.7% (Rubie & Brearley, 1990) - but significant enough to account for the melt filled microfractures in quartz grains recorded in several experimental studies. Microfracturing may therefore be an important mechanism for extracting melts from migmatites by the generation of short-lived permeability. Laboratory experiments of reaction-induced microfracturing have used backscattered electron imaging (BEI) to detect trapped melt in fractures in quartz (Connolly et al., 1997). During anatexis of natural rocks, however, melt will be expelled from microfractures by deformation. Furthermore, when quartz anneals it is unlikely to retain melt because it has a very small void space. Conventional techniques such as BEI which rely on compositional contrasts cannot therefore be used to look for evidence of microcracking in quartz.
In order to ascertain whether reaction-induced microfracturing occurs in natural migmatites I have used an alternative technique, using recently developed high efficiency cathodoluminescence (CL) detectors to image structural defects introduced into the crystal lattice during microfracturing. This technique has been used successfully to demonstrate internal zoning of quartz in granites (D'Lemos et al., 1996) and volcanic rocks (Watt et al., 1997) Cathodoluminescence occurs when electron hole pairs (formed by interaction between the sample and a beam of electrons) recombine and emit radiation over the wavelength range from ultra-violet to near infra-red (200 - 2000 nm). Luminescence centres (sites of photon generation) occur at inhomogeneities and dislocations in the crystal lattice, and may also occur at sites of substitution or insertion of foreign elements (activators) and at lattice vacancies. Microcracks in quartz anneal extremely rapidly, especially when an aqueous fluid is present, and melt trapped in microfractures is thought to have a similar effect on annealing rates (Connolly et al, 1997). Because rapid annealing along microfractures will lead to higher defect densities, luminescence along healed microfractures will be enhanced and CL can therefore be used to image annealed reaction-induced microfractures in quartz. Preliminary studies of diatexites formed by water-undersaturated melting show that cracked metamorphic quartz grains are entrained in the leucosome and mantled by anatectic, oscillatory zoned quartz. Further application of this technique to a variety of migmatite and granite types will hopefully allow a fuller understanding of melt extraction mechanisms and the behaviour of magma-solid mixtures during granite ascent and emplacement.
Connolly JAD, Holness MB, Rushmer T & Rubie DC, Geology, 25, 591-594, (1997).
D'Lemos RS, Kearsley AT, Pembroke JW, Watt GR & Wright P, Geol. Mag, 134, 549-552, (1996).
Rubie DC & Brearley AJ, High-Temperature Metamorphism and Crustal Anatexis, Allen & Unwin, 407-435, (1990).
Watt GR, Wright P, Galloway S & McLean C, Geochim. Cosmochim. Acta, 61, 4337-4348, (1997).
The Scottish Ballachulish Igneous Complex (412 ± 28 Ma) comprises a central granite surrounded by quartz diorite, emplaced at about 10 km, initially at 1000 to 1050°C. The contact aureole occurs in a variety of metasedimentary rocks. We focused on a textural examination of the arkosic Appin Quartzite, on the eastern margin of the intrusion, where the contact is vertical. Prior to the present study, these aureole rocks were believed to have melted only within a few metres of the contact.
The Appin Quartzite contains clastic quartz and feldspar grains (perthite Or90 with rare albite) and a few rock fragments. Temperatures in the aureole exceeded the solidus in Qtz-Ab-Or-H2O to a distance of nearly 500 m from the contact. Experimentally determined, equilibrium, quartz-H2O dihedral angles are consistent with no grain-boundary permeability in the Quartzites - in agreement with previous statements concerning a very limited melting zone. Despite this, textures we observed in the Quartzite imply partial melting at distances up to 500 m from the contact. Given the peak temperature of ~ 775°C, this would only have been possible in the presence of aqueous fluid. This points to fluid infiltration, and allows us to infer both the fluid flux and the infiltration pathways, using the distribution and abundance of the crystallised former partial melts.
About 500 m from the contact, most feldspars have tiny cuspate extensions, and these increase in abundance toward the intrusion. These features generally form along quartz-quartz grain boundaries, but also along feldspar-quartz boundaries, close to planar arrays of fluid inclusions. Quartz-quartz-feldspar dihedral angles show a bimodal distribution with a peak at ~ 105°, corresponding to the equilibrium solid-state value, and a broader peak between 40 and 60°, that represents the extensions.
Within about 150 m of the contact, the extensions become increasingly elongate. Some are disjointed, like strings of beads in optical continuity. Small grains of feldspar commonly occur at three-grain junctions in the quartz. The quartz-quartz-feldspar dihedral angle of these highly elongate apophyses is, again, 40 to 60°.
Within about 60 m of the contact equigranular feldspar-rich microgranitic patches and stringers occur. Large clastic feldspars are commonly surrounded by microgranite. In some rocks, very close to the contact, there is evidence of melt-enhanced cataclasis.
Melting efficiency decreased with distance from the contact and evidence points to the intrusion as the fluid source. Calculated fluid flux was ~ 7000 kg/m2, and we conclude that the aqueous fluid passed through the aureole by a fracture network. Fluid transfer was highly efficient and only a small fraction of the true flux is texturally recorded in the rocks.
Quartz bipyramids and Ba-zoned orthoclase megacrysts grew in response to stress in quartz monzonite and adjacent country rocks at Papoose Flat, Inyo Mountains, California, during post-consolidation deformation. Studied were thin-sections and stained polished surfaces of orthoclase crystals from more than 100 localities, cut through centers parallel to (010). Orthoclase megacrystals contain spatially and crystallographically oriented epitaxial mineral inclusions, selectively concentrated in face-sector volumes; and 20 to 50 Ba-zoned internal layers that do not correlate from crystal to crystal. The monoclinic crystals show slightly triclinic outer shapes throughout most of the pluton where folia are weak, and rounded to subhedral shapes where folia are strong. Shape-controlling factors prevailed during growth, shown by parallelism of oriented internal features to outer boundaries, regardless of shape, from crystal centers to margins. Ductile flow by physical movement was not involved. Host minerals were replaced by passive coupled dissolution-crystallization reactions. Self-organization took place in response to stress, which supplied energy and disturbed equilibrium at crystal faces. Sparse fluid coatings promoted diffusion and kinetically dominated surface reactions. Mild stress at Papoose Flat, sustained for short geologic times, caused major solid-state rearrangement to new minerals with disequilibrium textures.
Chemical forces causing replacement are driven by gradients in temperature, pressure, and composition. Coupled reactions cycle energy. Exothermic energies of crystallization cause simultaneous endothermic dissolution. Elements released by dissolution may be utilized by growing crystals. Destabilized host minerals vanish, regardless of identity, giving way to equal volumes of new phases. Replacement requires: open systems; energies in excess of equuilibriium requirements; inward paths to reaction sites; supply of components, delivered by fluid flow or diffusion; nucleating mechanisms; instability of replaced phases; and outward paths for unconsumed components. To write balanced soichiometric reactions require knowledge commonly not available in open systems. Mechanical displacement of surroundings contrasts to passive retreat of host by coupled dissolution-growth reactions.
Quartz grain boundaries can be treated as fractal curves. Their fractal dimensions can be measured by the divider method, in which an inverse power law relation between perimeter and ruler length gives the fractal (coastline) dimension, or by the relation between perimeter and area. The fractal dimensions of both relationships have been calibrated against temperature of deformation (Kruhl and Nega 1996, Takahashi et al. 1998), and can be used in principle to identify intrusion-related deformation. The perimeter-area method is impractical for use in most granitic rocks because it is restricted to the few quartz grains with no other phase boundaries, and because it requires measurements of many grains/sample, which is impractical from thin sections due to the relatively large sizes of quartz grains. Important limitations of the coastline method are systematic and non-systematic breakdowns of the perimeter-ruler relationship at large ruler lengths. The relationship breaks down systematically by perimeter increasing with length in a series of discrete steps at larger ruler lengths, which can be predicted as a consequence of the method. Non-systematic breakdown of the relationship occurs when ruler lengths approach the chord length of the grain boundary. A more fundamental problem may be the dependance of fractal dimension on strain rate as well as temperature. However, the calibration of Kruhl and Nega suggests that the strain rate effect is negligible over a range of temperatures in naturally deformed samples corresponding to greenschist to granulite facies. Fractal dimensions of quartz grain boundaries in the late Archean Murehwa batholith range from 1.10 to 1.24, with an average of 1.17, giving a deformation temperature of 550 ± 123°C (1 S.D.) that is consistent with magma emplacement and cooling temperatures. As independant field evidence demonstrates that fabrics in the batholith are related to granite intrusion, this constitutes a successful field-based test for the method. Low fractal dimensions of quartz grain boundaries, indicating high temperatures of deformation, can be therefore be used as an indication of deformation associated with intrusion, provided that there has been no overprinting high-temperature deformation.
Kruhl JH, & Nega M, Geol. Rundsch, 83, 38-43, (1996).
Takahashi M, Nagahama H, Masuda T & Fujimura A, J. Struct. Geol, 20, 269-276, (1998).
Description and interpretation of the microstructural state (fully magmatic, subsolidus, HT solid-state and LT solid-state) at every possible location in a pluton, and comparison with the fabric patterns, produce meaningful maps allowing construction of a detailed emplacement model. If the continuity of a fabric pattern across different microstructural domains can be established, non-instantaneous models can be put forward that constitute a powerfull tool for solving the role of tectonics in the emplacement process. This technique has been aplied on late-Cretaceous granite plutons from the Sierra Nevada (the Mono Creek Pluton, MCP) and White-Inyo Range (the Papoose Flat Pluton, PFP) (California), leading to a better understanding of the respective roles of tectonics and magmatisms in their emplacement histories.
Microstructures in the MCP and PFP show a continuous fabric transition between magmatic and orthogneissic. In the MCP, presence of low-temperature subgrain boundaries attest that deformation continued after full cristallization of the magma. In the PFP, orthogneissic fabrics and perfectly polygonized quartz grains showing no sub-structures, attest for 1/ a HT static restoration of quartz after a large deformation ascribed to emplacement, and 2/ no deformation after emplacement.
The AMS fabric pattern within the MCP displays a dextral sigmoid of foliations and lineations. Continuity of fabric orientation through all the microstructural domains points to a continuous dextral shearing during emplacement. Within the PFP, the foliation pattern is concentric. In the western part of the pluton the NNW-SSE trending magnetic lineation pattern recorded in all microstructural domains suggest that the magmatic and solid state domains synchronously underwent the same kinematics, related to pluton construction by magma inflation.
These results demonstrate that fabric patterns cannot be analyzed independantly of microstructural record. Misinterpretation, such as post-emplacement margin deformation for the PFP, and consequent erroneous tectonic reconstructions, can be avoided. These two plutons being contemporaneous and adjacent, interesting consequences on magma transport and emplacement mechanisms can be derived, in particular on the respective roles of regional (tectonic) stress versus magmatic pressure. Finaly, our approach helps in evaluating the memory of fabrics...
Methane and higher hydrocarbons are commonly present in alkaline igneous rocks, sometimes in anomalously high concentrations of, for instance, up to 108 cm3/kg CH4 in the Kola igneous province. Our recent data show that much of the methane and associated hydrocarbons measured from boreholes and decrepitated samples originate from fluid inclusions. However, whether these hydrocarbons were derived from the mantle, represent sub-solidus magmatic reactions or contamination by biogenic basinal fluids is a matter of some controversy. Although the apparently primary nature of some of the inclusions might suggest a magmatic origin and be used in support of a mantle source, chemical, textural and mineralogical evidence point to a sub-solidus magmatic origin.
Microthermometric, laser-Raman, and C isotopic data from CH4 -bearing inclusions in minerals from the Khibina and Lovozero complexes, in the Kola Peninsula, and PVTX modelling of these data demonstrate the inclusions to have a low P-T origin at 0.8 - 1.5 kbar and ~340oC.
Petrographic and microtextural studies show that the CH4 -bearing inclusions are structurally controlled occuring along curvi-planar fractures and cleavage planes. Commonly they are located in zones of incipient hydrothermal alteration involving growth of phases such as aegerine, biotite, cancrinite, analcime, pectolite and magnetite at the expense of arfvedsonite, nepheline and Ti-magnetite. The inclusions are always found in association with small Ti-magnetite crystals which do not occur elsewhere in the rocks. These crystals show a reaction rim of pure magnetite, biotite and aegerine.
The fluid inclusion data and textural relationships, including the association of CH4 inclusions with magnetite and late-stage hydrated phases, suggest an abiogenic origin for CH4 during hydration within a reducing environment . This involved subsolidus Fischer-Tropsch reactions of the type CO2 + 4H2 --> CH4 + 2H2O coupled, through an autocatalytic positive feedback mechanism, with hydration reactions of the type assA + H2O = assB + mt + H2. In this case, hydration reactions, accompanying hydrothermal alteration would produce abundant H2 that would react with magmatic CO2 -rich fluids to produce CH4 and H2O. The H2O produced would drive further hydration reactions thus creating a self-perpetuating cycle enabling large-scale CH4 production in a relatively closed system. Access of the small amounts of magmatic H2O needed to initiate the cycle was by small fractures distributed through the rock volume that are now populated by CH4 inclusions. Such inclusions are limited to those areas not flushed by large volumes of H2O, as at the Sokli complex, where open system behaviour suppressed the efficiency of the Fischer-Tropsch reaction.
A number of properties of rocks are sensitive to grain size, for example the transition from diffusion to dislocation creep or seismic velocities. Existing grain growth data for olivine is restricted to experiments with monomineralic aggregates either at 1 atm or relatively short experimental durations at 0.3 GPa. The aim of the present study is to determine grain growth rates of olivine in melt-free and melt added aggregates at durations up to 4 weeks at high pressure (1 to 3 GPa) and temperatures in the range from 1200° to 1450°C. The starting material is produced by a solution-gelation process which ensures high purity and uniformly fine grain size of about 2 µm of composition Fo90. Without added melt, this material undergoes a transition from normal grain growth at 1250°C (mean grain size 11 µm after 24 h) to abnormal grain growth at 1350°C (100 µm). No abnormal grain growth occurs in partially molten aggregates; at 1350°C the mean grain size is 18 µm after 24 h and 40 µm after 430 h with about 1.5% basaltic melt present.
As in previous experiments with ground natural olivine as starting material, the melt distribution in long duration experiments is characterised by small triple junction tubules, some wetted two-grain boundaries and larger, irregular melt inclusions. In short duration, fine grained samples the melt distribution is more homogeneous, wetted grain boundaries are largely absent. Normalised grain size distributions indicate that both melt free and melt present grain growth is not diffusion but interface reaction controlled. For melt free aggregates the distribution is similar to first order surface reaction, for melt present aggregates to second order surface reaction controlled growth, consistent with the presence of faceted crystal-melt interfaces, which also indicate growth by a layer spreading mechanism of an atomically smooth surface. Melt infiltration of a dunite or harzburgite without a large temperature increase will therefore not lead to rapid grain growth as is inferred for some ophiolites and xenolith.
The origin of poikilitic K-feldspar phenocrysts was the subject of intense controversy in the literature until the mid-1980s. The subject has been more or less dormant since the authoritative paper by Vernon (1986) where he convincingly argued that most described Kf poikiloblasts were not a result of the reaction of pre-existing solids with volatile-rich fluids (metasomatism), but were in fact crystallized directly from a magma. In this talk, I wish to revive the subject by demonstrating that metasomatism was responsible for the growth of Kf poikiloblasts and a range of other textures commonly found in the water-rich leucogranites of the Imja Khola, Khumbu Himalayas, Nepal. Metasomatism, driven by volatile-rich fluids exsolved from granitic magma, was responsible for (1) growth of poikiloblastic K-feldspar megacrysts, (2) mm-wide microveins of albite which replaced magmatic quartz, plagioclase and K-feldspar, and (3) growth of tourmaline in schist. The metasomatic origin of K-feldspar poikiloblasts is indicated by textures such as partly resorbed inclusions of muscovite, plagioclase and quartz coexisting with well-preserved biotite inclusions. K/Na and Ba profiles across grains interpreted on textural grounds to be either metasomatic or magmatic, differ strongly, and support textural interpretation: metasomatic grains yield flat profiles, whereas magmatic grains yielded characteristic igneous Ba and K/Na zoning. High Ba-content in metasomatic Kf compared to low Ba-content in magmatic grains, suggest that volatile exsolution occurred early in the history of magma crystallization, mostly before crystallization of igneous Kf. Otherwise, Ba would have been enriched in the magmatic Kf, leaving the melt and late volatiles impoverished in Ba. This conclusion suggests that exsolution of volatiles may have played an important role in the origin of the granite injection complex outcropping in the upper Khumbu area.
Vernon, RH, Earth-Sci. Rev, 23, 1-63, (1986).
Among the various AMS parameters, P has been successfully employed in granitic rocks as an index of subsolidus deformation. Generally, P values lower than 4-6 have been correlated with magmatic/submagmatic microstructures whereas those values higher than 6 reflect different intensities of subsolidus imprint (Bouchez and Gleizes, 1995, Bouchez et al., 1990).
Paleozoic granitoids (tonalite to monzogranite) from the SW border of NPM display a magmatic fabric defined by a parallel alignment of subhedral zonal plagioclase, sometimes combined with tiling, along with an ill defined biotite orientation, in tonalites and granodiorites; an uncommon parallel alignment of microcline megacrysts, sometimes enclosing small oriented plagioclase and biotite inclusions, is considered as a primary magmatic feature in monzogranites. Magmatic microstructures are overprinted by an upper-greenschist facies ductile deformation. This subsolidus imprint is characterized by an unevenly distributed protomylonitic texture with plagioclase and K-feldspar porphyroclasts; muscovite fishes are developed in most strained domains (Cerredo and López de Luchi, in press).
A pilot AMS survey carried out in MCG revealed the existence of both ferromagnetic (107-1105 x10-5 SI) and paramagnetic (2-44 x10-5 SI ) K values. P% in the paramagnetic population varies from 4.1 to 21.5, being biotite the only iron-bearing phase. Planar magnetic fabrics dominate over linear (Flinn parameter < 1) except for a 20% population which display prolate magnetic ellipsoids.
Relationship between P values and deformational microstructures is not straightforward. Even the samples with the lowest P values are highly deformed. Since K in paramagnetic rocks depends on Fe-bearing silicate modal amounts, the P values reflect the biotite orientation degree which in these granitoids is the result of magmatic plus subsolidus events. When the magmatic fabric and mylonitic foliation planes lie parallel (or nearly so) to each other (as indicated by parallelism between inclusion trails within K-feldspar megacrysts and external mylonitic foliation) the highest P values are recorded, on the contrary when both planes lie at high angles (as evidenced by robust magmatic biotite plates that are orientated at high angle to the mylonitic foliation), lowest P values are recorded, regardless of the intensity of the mylonitic imprint.
P parameter may be a powerful tool to rapidly distinguish granitoids bearing only magmatic microstructures from those which have undergone a subsolidus overprint only if the former were not strongly controlled by the orientation of Fe-bearing silicates. P values in our example are strongly dependent not only on the subsolidus imprint intensity but also in the spatial orientation of the magmatic fabric and the mylonitic foliation.
Bouchez JL & Gleizes G, J. Geol. Soc. London, 152, 669-679, (1995).
Bouchez JL, Gleizes G, Djouadi T & Rochette P, Tectonophysics, 184, 157-171, (1990).
Cerredo ME & López de Luchi MG, J. South Am. Earth Sc, (in press).
Migmatites are often the site of intense fluid circulations, the effect of which is to reset isotope ratios (O, Sr) and to obscure melting processes. Since these fluids flow through heterogeneous materials, they are successively faced with different mineral assemblages along their pathways. This is a strong driving force for fluid-rock exchange, particularly when temperatures are high enough for local chemical equilibrium to be approached. Migmatitic metapelites from the Velay central (France) offers one example where such subsolidus reactions are particularly extensive.Subsolidus replacement textures involving aluminosilicates and feldspars are imaged and interpreted under optical cathodoluminescence. In the neosomes, Al silicates (sillimanite, cordierite and garnet) are partly converted to plagioclase while, in the leucosomes, K-feldspar develops at the expense of quartz and plagioclase, locally producing monomineralic veins. Depending on the nature of the substrate (Al-rich versus Si-rich), a different secondary feldspar is formed (plagioclase versus K-feldspar), but the bulk transformation is clearly a feldspathisation of the melted metapelite. This transformation converts a strongly peraluminous mineralogy into a less peraluminous one, and it involves a substantial addition of Ca and the probable removal of Fe from the metapelites. Moreover, it seems to be linked with the presence of metabasalt layers, interbedded with the pelitic material. We thus interpret this feldspathisation as a metasomatic effect due the strong change of equilibrium fluid composition when it moves from the (calcic) metabasalt towards the (aluminous) metapelite. Incidentally, this transformation and the related replacement textures are shown to have strong similarities with those observed in the vicinity of many skarns (in the so-called endoskarns).
Late-stage dikes in high-level granitic systems provide crucial information on the final stages of crystallization. We mapped dike complexes in granodiorites in the equigranular Half Dome near Tenaya Lake and megacrystic Cathedral Peak (CP) on Lembert Dome of the Tuolumne Intrusive Suite (CA). Dikes exhibit internal textural variability; many are aplites, pegmatites, or related textural variants, and are mineralogically simple. Field relations suggest that some dikes are locally derived, with diking occurring within a partly molten crystallization interval. We suggest that changes in dike texture along strike are a function of changes in host crystallinity and temperature during diking. In some cases, individual dikes can be followed along strike (for 10's of meters) from areas where the dike is wider, has faint or undulose borders, and has a texture similar to the groundmass of the host granodiorite, to areas farther along strike where the dike is thinner, contacts are sharp, and the dike texture does not resemble the host texture. In the direction of decreasing dike width, decreasing plasticity, and increasing undercooling, the dikes exhibit: 1) Fe-poor CP groundmass-like granite (internally nucleated), 2) a texture with euhedral to anhedral quartz "eyes", 3) pegmatitic texture (externally nucleated), 4) poorly-developed graphic texture, 5) well-developed graphic texture, and 6) in places, aplitic (i.e. fine-grained) borders. The textures (away from the site of dike melt extraction) suggest that cooling of melt below the liquidus in response to crystallization and the magnitude of undercooling may not approach crystal-melt equilibrium prior to solidification of the dikes. Our observations of changing physical and textural character within dikes are consistent with intrusion into a partially molten crystallization interval with varying degree of undercooling along dike length, indicating the importance of undercooling in the development of aplitic, pegmatic, graphic and quartz "eye" textures.
Igneous cumulate generation and the formation of igneous layering has been the focus of much debate, with many conflicting models e.g. crystal settling, in-situ growth, gravity currents etc.. Paramount for attempting to answer how igneous cumulates form is a detailed understanding of the packing structure of crystals, whether cumulate phases produce touching frameworks in 3-D and the nature of the building blocks of the cumulate pile. With these aims in mind the present study reports on the detailed textural analysis of igneous cumulates using komatiite cumulates as a case study. The rationale for looking at komatiite cumulates is that they have a relatively simple cooling history compared to cumulates in layered igneous intrusions and may, therefore, provide primary cumulate textures.
Textures from a detailed profile through the cumulate B2 zone of Joe's flow komatiite, Belingwe greenstone belt, Zimbabwe, were analysed. The grain-size distribution of the flow is very constant throughout the flow, and is moderately sorted. The Spatial Distribution Pattern of Joe's flow is characterised by variations in R value (a measure of how ordered, random or clustered the texture is) between 1.12-1.27 which indicates a clustered distribution, and shows a packing variation trend on a porosity vs R plot. Cluster analysis of the flow reveals a cluster population size range of 0.3-2.5 mm in 2-D section, with clusters occurring as glomerocrysts of olivine. The origin of the olivine glomerocrysts is constrained using geochemistry, and dihedral angles, and is likely to represent a mixed population of entrained olivine clumps and olivines which grew during eruption and flow of the komatiite lava. The analysis of other komatiites cumulates indicate that komatiites display a variety of packing arrangements of their constituent olivine grains. Such variations can be explained by differences in the original morphology of the packing units, whether they are individual olivine crystals or irregular chains and clusters of olivines before accumulation.
Quantification of rock textures requires precise recognition of geometrical and spatial relationships between grains at all scales. For this aim, we propose a method developed on the basis of the 2D Anisotropic Wavelet formalism. The Normalised Optimized Anisotropic Wavelet Coefficient (NOAWC) method enables to detect and quantify the size, shape, orientation and location of the different levels of grain organization of a rock section (Gaillot et al.,1997).
A first example allow to identify and evaluate the effect of possible magnetic interactions of magnetite grains on the rock AMS. For this use, we compared the 3D Shape Preferred Orientation ellipsoid (SPO) of natural magnetite grains obtained analyzing a set of perpendicular thin sections and the AMS ellipsoid of this granitic sample.
A second natural example is a multi-scale and multi-approach study of the mineral organization performed in the Sidobre granite (Tarn, France). At whole massif scale, AMS data gives a well defined NNE-SSW sub-horizontal trend interpreted as the direction of the finite extension. At medium scale (100 x 80 cm), alkali feldspar megacrysts seems to be broadly parallel to the AMS directions. However, a detailed exploration using the NOAWC analysis exhibit peculiar structures interpreted as "magmatic normal faults". At small scale (20 x 40 cm), Shape Preferred Orientations of the main phases defined by the intercept method (Launeau and Robin, 1996), are parallel and underline the magmatic finite extension except the interstitial K-feldpars which are sub-perpendicular to the elongation and form discontinuous and irregular millimetric-scale veinlets. It demonstrates that the residual melt takes place in small tensional domain sub-perpendicular to the magmatic lineation defined by the other main phases (boitite, quartz, plagioclase, K-feldspar megacrist).
Hence, the use of a multi-scale and multi-approaches analysis provides a satisfying quantification of the rock texture. It provides important insights about the physical processes involved in the rock formation. The rock history can be then reconstructed from magmatic to solid state.
P Launeau, PY Robin, Tectonophysics, 267, 91-119, (1996).
P Gaillot, J Darrozes and Mde Saint Blanquat, Geophysical Research Letters, vol. 24, n° 14, 1819-1822, (1997).
The Lachlan Fold Belt (LFB) is composed of N-S folded, early Palaeozoic deep marine sedimentary and volcanic rocks and a series of N-S elongated batholiths.
The I-type Carcoar (CG) and Barry-Granit (BG) square to tabular wedge-shaped bodies, show a weak deformation, whereas the Sunset Hills Granit (SHG) exhibits a moderate deformed S-type granit. The granites intruded into deformed and metamorphosed early to late (?) Ordovician Adaminaby Group in the south and Middle Ordovician Coombing Formation and late Middle to Late Ordovician Blayney Volcanics in the north (Wyborn et al. 1996).
Paramagnetic minerals like hornblende and biotite control the magnetic susceptibility and the linear as well as planar flow pattern of the granitoid bodies. Microstructural observation support the magnetic measurements. The dominant foliation is variably developed and increases from the CG (north) to the BG (center) to the SHG (south). Macroscopic foliation is visible as a planar alignement of biotite and sometimes of hornblende. The foliation of the BG, the SHG and their country rocks is subparallel to the magnetic foliation and also to the major fault zones. The macroscopic and magnetic fabrics of the CG are almost subparallel to the major shear zones. The AMS-ellipsoid also shows a general north-south trend, i.e. a more oblat shape in the CG and BG to a prolat shape in the SHG.
The magma emplacement was controlled by transcurrent movements along reactivated shear zones related to the late Early Silurian oblique opening of the adjacent Hill End Through.
Integrative gravity and structural modelling of the granitoids in the Eastern Lachlan Fold Belt revealed contrasts in emplacement mode and deformation style between I-type and S-type granites. The Carcoar Granodiorite was emplaced into a transtensional pull-apart structure and the Barry Granodiorite and the Sunset Hills Granit intruded into transpressional shear zones.
Henderson GTM, Morgan EJ, Raymond OL, Scott MM, Warren AYE & Wyborn D, Geological Society of Australia, 41, 189, (1996).
The Rieserferner Pluton is a tonalitic to granodioritic intrusion within the austroalpin nappe complex of the eastern Alps. It is related to the series of tertiery granitoids which intruded along the Periadriatic lineament (PL). In case of the Rieserferner Pluton this fault is the Defereggen-Antholz-Valles - Line (DAV). The east-west elongated pluton is located to the north of the DAV.
The deformation history of the tonalite was studied by magnetic fabrics, rock fabrics and microstructures. Different shapes of the AMS-ellipsoid could be observed in the northern and central part where the roof of the pluton is exposed and in the southern part next to the DAV. The deformation in the northern and central part of the tonalite shows oblate shapes. The magnetic foliation and the macroscopic foliation which is visible by the shape fabric of the biotite are subparallel to each other. Low KVol (<400x10-6 SI) also point to a control of the magnetic properties by paramagnetic minerals. The influence of ore minerals seems to be less important. A flat lying foliation could by observed next to the roof of the pluton, whereas to the north the foliation is from intermediate to steep north dipping and almost parallel to the foliation in the country rocks. To the south the shape of the magnetic ellipsoid is prolat with a steep west dipping Kmax and a foliation plane parallel to the contact. Almost isotropic fabrics were detected in the deeper parts of the pluton along the lower Geltal. Within the isotropic samples some of the quartzes shows a typical chessboard pattern and a lack of recrystallization features pointing to a high-temperature solid-state deformation. Evidence for a submagmatic deformation is preserved by brittle deformation of plagioclase where quartz and sometimes also K-feldspar crystallises within the fractures. In samples which were taken next to the contact quartz shows solid-state deformation features. In sections perpendicular to the foliation a layering of recrystallized quartz bands together with biotite is obvious.
Recently, the AMS is frequently used for the structural mapping of granites. To test the application of the AMS for a quantitative description of granite emplacement processes the Ardara pluton (Donegal, Ireland) was selected. The emplacement of the pluton has been attributed to various mechanisms, including diapirism, balooning, a set of nested diapirs with a high amount of stoping. The Ardara pluton is roughly circular, about 8 kilometers in diameter with a "tail"-like apophysis in the SW. Three distinct compositional units make up the pluton. The outermost quartz monzodiorite (GI) has a sharp and smooth contact to its wall-rock. The medium- to coarse-grained GI show a sharp contact to the inner part, i.e. a tonalite (GII) and the central granodiorite (GIII). The outer rim (GI) shows a strong foliation parallel to the contact wich is dying out towards the inner core (GIII). Most of the foliation is of magmatic origin. The values of the bulk susceptibility (K) from the investigated core samples differ from 37x10-6 to 13200x10-6 SI. The zoning in the bulk susceptibility of the Ardara pluton corresponds well to its petrographic zoning. For the pluton as a whole the magnetic fabric is very well developed. Like the macroscopic foliation the magnetic foliation is mostly steep or even vertical. The strike of the magnetic foliation is subparallel to the respective zone margin. The magnetic foliation is dying out progressively into the periphery of the central granodiorite, which is well documented in the planar anisotropy degree. The T values also shift from more distinct oblate (GI) to weak oblate or even prolate forms (GIII) indicating that the magnetic foliation is much better developed than the magnetic lineation in nearly all localities. The magnetic lineation shows variable plunge, mostly subhorizontal. The Kmax-axes are arranged along a broad E-W striking girdle. The magnetic fabric data, namely the determination of the geometry and intensity of the magnetic foliation and lineation obtained through the AMS was often used as a quantitative measure for granite emplacement processes. In order to constrain the magnetic fabric of the Ardara plutonic rocks numerical modelling has been performed to obtain quantitative information about the magnetic carrier. Additionally, HFA analyses support the modelling results. Neither the different magnetic measurements nor the numerical modelling support any kind of diapiric ascent of the Ardara pluton
The ellispoidal shaped Chinamora batholith is surrounded by greenstonebelts belonging to the western succession of the late Archean (2.7 - 2.6 Ga) upper Greenstones. The granitoid rocks can be divided into at least 26 different lithological units. They can be summarized on the basis of the age coherence and their macroscopical fabrics into 4 major units: (i) old gneisses; (ii) gneissic granites; (iii) porphyritic granites and (iv) late granites (where (i) is the oldest and (iv) the youngest unit respectivly, see Siegesmund et al. 1998 for details). The emplacement mechanism of this batholith is the subject of a controversial discussion. While Snowden & Bickle (1976) and Snowden (1984) proposed interference folding during and after emplacement of the late granites to explain the dome- and keel configuration of the batholith and envelopping greenstonebelts, Ramsay (1989) interpreted the batholith as a spherically inflated (ballooning) pluton. Jelsma et al. (1993) explain the structural- and strain- data with a diapiric intrusion. A study of the magnetic properties has been performed to get a detailed view of the orientaion and distribution of the magnetic properties in the granitoid rocks and hence their fabrics. The AMS (Anisotropy of the Magnetic Susceptibility) measurements revealed a very heterogenous distribution of the volumesusceptibility (kVol ranging from 100x10-6 SI up to 10000x10-6 SI) throughout the batholith. Neither the different plutons nor the 4 major units can be correlated with a certain volumesusceptibility, only a very weak tendency of samples with low kVol-values lying towards the margin of the batholith is evident. The degree of anisotropy ranges from near to 1 (late granites), 1-1.3 (porphyritic granite) up to 1.4 (gneissic granites and old gneisses), while the majority of the measured samples plot in the field of only low anisotropies (approx. 1.1). The form of the magnetic ellipsoids of the samples range from nearly perfect oblat to nearly perfect prolat without any lithological significance. The magnetic foliation of the granitoid rocks has an EW strike and seems to preferrably dip with high angles. The magnetic lineation in the gneissic granites and the porphyritic granite trends EW with a mainly shallow dip. The late granites show a preferred moderate dip to the NW. From the old gneisses there is no clear orientation pattern of the magnetic lineation. A High Field Analyses (HFA) of some selected samples has been made in advance to seperate the magnetic properties from paramagnetic and ferrimagnetic minerals. The magnetic properties described above suggest a syn- to post-emplacement deformation of the rocks on a regional scale which at least in part overprints the emplacement related magnetic fabric.
Jeslma HA, van der Beek PR & Vinuy ML, J. Struct. Geol, 15, 163-176, (1993).
Ramsay JG, J. Struct. Geol, 11, 191-209, (1989).
Siegesmund S, Becker JK, Becker T & Jeslma HA, Freiberger Forschungshefte, C471, 212-214, (1998).
Snowden PA & Bickle MJ, J. Geol. Soc. London, 132, 131-137, (1976).
Snowden PA, In: Precambiran textonics illustrated. Eds: Kroener A. & Greiling R, 135-145, (1984).
Methods currently available for the description and quantification of rock fabrics fall into two categories. 1) object methods based on the shape of individual grains or 2) bulk image analysis techniques such as 2D fourier transforms. In general, most of these methods suffer from being time consuming to apply or the results are difficult to interpret. In this study, SURFOR methods of strain analysis in deformed rocks have been combined with a simple method of visualisation to investigate the textures of igneous rocks.
A 2D image of a rock slice or thin section is thresholded to separate the different phases. Cumulative frequency histograms of grain dimension are calculated for a series of traverses through the image. These traverses are carried out over different orientations to give a radial picture of the cumulative grain spacing. The radial data are then plotted on a single diagram and contoured for visualisation purposes. This technique has been applied to synthetic images representing different fabric distributions and examples of natural igneous fabrics. The technique can be easily automated, is scale independent, and produces a resulting diagram that may be compared with diagnostic patterns.
Durbachites belong to the oldest (340 - 330 Ma) Variscan magmatites. The members of this suite grade from K-rich diorites to syenites and granites. The characteristic features for the whole suite are high concentrations of LILE (K, Rb) and radioactive elements (U, Th) which are accompanied by high concentrations of Mg, Cr, and Ni. The rocks with the highest mg number display the highest K/Na ratio. Holub (1997) and Gedres et al. (1998) interpreted these rocks as being a product of mixing between an enriched mantle magma and a crustal melt. The Trebic batholith is the largest (500 km2) body of durbachites in the Bohemian massif. The mineral assemblage - perthitic K-feldspar, plagioclase (oligoclase-andesine), biotite, actinolite pseudomorphs after clinopyroxen, apatite, and zircon - indicates some late stage overprinting.
Cathodoluminiscence+optical microscopy, combined with electronmicroprobe analyses, was used to distinguish and characterize magmatic and subsolidus structures.
The plagioclase is dominated by primary magmatic structure - distinctive oscillatory zoning, with some later alteration - sericitization, carbonatization and, rarely, scapolitization. The up to 2 cm large euhedral to subhedral phenocrysts of K-feldspar exhibit a more complex internal structure. The grains contain some small, altered plagioclase inclusions. The shape of the individual inclusions is irregular, indistinct. These inclusions are often arranged in several zones, which are parallel with the external shape of the crystal. Some inclusions are very small and are observable only as very thin zones, which could be confused with real magmatic oscillatory zoning. Such a style of zoning indicates an important role of the replacement processes in the origin of the K-feldspar phenocrysts. Subsolidus structures were identified in the amphibole and plagioclase. These Ca-rich minerals are sometimes replaced by carbonate. The carbonate builds irregular "clouds" in the amphibole or replaces, together with scapolite and sericite, the An-rich zones in plagioclase. The presence of carbonate with scapolite indicates subsolidus reactions of the rock with CO2- and Cl-rich fluids. Other subsolidus structure was observed in the veins of the extremely radioactive alkali-feldspar-syenites which accompanied the durbachites. The radioactive elements (U,Th) are bound in zircon, which is therefore strongly metamict. The change in volume during the metamictization of zircon is responsible for the origin of very thin cracks which penetrate the whole rock. Some K-feldspars exhibit a change in luminescence along such cracks (from blue to violet). This luminiscence change is probably due to radioactive damage in the feldspar structure, which was induced by circulation of radioactive fluids in the cracks.
The observed structure indicate a probable substantial role of late magmatic to subsolidus process in the origin of the durbachites and the high activity of fluids rich in K, CO2 and Cl in this process.
The magmatic and tectonic history of intrusive suites in high-grade terrains is paramount in understanding magmatic processes at depth, unravelling polyphase orogenic events and documenting crustal rheology (Corriveau et al., in press). However, these intrusive rocks can easily be overlooked in regional studies. Careful field work on so-called "metamorphic rocks" in the deeply-exhumed Grenville province reveal that many are plutons and dykes with misinterpreted magmatic fabrics. Examples include 1) compositional magmatic layering induced by crystal sorting and repeated magma injections, confused for gneissosity, 2) sub-vertical sheet-like intrusions and modal layering interpreted as tilted or transposed from horizontalness by tectonics, 3) sub-ophitic gabbroids getting unnoticed because of prevailling, "pseudo-metamorphic", granoblastic texture, and 4) episodes of mafic dyke intrusion overlooked, because of a wrong perception of timing relative to regional metamorphism.
Criteria used in textural analysis of deep orogens need to be reassessed. For instance, many igneous textures such as flow foliation and ophitic to intergranular textures are built around a framework of feldspar laths. Lath-shaped crystals are shown to be unstable, being either sutured or partly to completely recrystallized into a granoblastic texture. As the distribution of mafic versus felsic minerals commonly mimics the original crystal shape, macroscopic-scale textures can be better indicators of magmatic fabrics and permits to see beyond recrystallization. Microscope-based textural analysis may hinder magmatic process studies. With that approach, we re-evaluate the origin of Grenvillian mafic and felsic gneiss as primary sheet-like mafic and felsic intrusions.
Recent mapping also questions current dogma about the origin of vertically-layered mafic intrusions and the competency of anorthosite. Transposition and tilting of horizontal beds and original bottom-up accumulation are commonly invoked to account for vertical modal layering and guide mineral exploration for Ni-Cu. Field research reveals that verticality of layering is primary and most plausibly results from side-wall crystallization in a sub-vertical magma conduit. These gabbros were emplaced in a compressive regime (Corriveau and Rivard, 1997). With their competent behavior once crystallized, they remained undeformed. In contrast, coeval anorthosite massifs appear strongly metamorphosed. Sheath folds involving delaminated anorthosite septa in mafic dykes reveal that anorthosite can be extremely ductile. Such a rheology provides insights on the systematic recrystallization of post-metamorphic Grenvillian anorthosite massifs.
Another application to look at is dating peak metamorphism with metamorphic zircons from "migmatitic" mafic dykes. In western Grenville, strongly deformed mafic-felsic dykes with commingling textures postdate peak metamorphism. Their transposed commingled felsic component can easily be misinterpreted for deformed leucosomes. Zircons from the mafic component would record timing of reactivation not peak metamorphism. These very simple observations illustrate the impact of field analysis of intrusive rocks for the understanding of deep crustal processes and provide a sounder scientific database for mineral exploration in high-grade terrains.
Corriveau L, Rivard B & van Breemen O, Journal of Structural Geology, (In press).
Corriveau L & Rivard B, From source to surface: the extraction, transport and emplacement of magma in a Grenvillian perspective, Geological Association of Canada/Mineralogical Association of Canada, B4, 82p, (1997).
The Sotillo granitoid (Sanabria, NW Spain) is a porphyritic two-mica granite, rich in mafic schlierens and raft trains of country rock xenoliths. It is located in the southern limb of the Ollo de Sapo antiform, a N120°E-trending domain belonging to the Central Iberian Zone of the Variscan Iberian belt. The granite intruded into migmatitic glandular gneisses of supposed late Precambrian or Cambrian age and into sillimanite-bearing schists of Ordovician age.
In plan view, the Sotillo pluton (approximately 60 square kilometers) displays an asymmetric blister shape. The combination of detailed structural mapping, magnetic susceptibility and anisotropy of magnetic susceptibility measurements have been used to constraint the dominant structural control during granite emplacement. The western tail is pervasively affected by a synmagmatic NW-SE dextral shear zone. The intensity of deformation recorded by S-C structures, X/Z ratio in enclaves and grain sizes increases from north to south towards the contact with the Ordovician schists. S-planes and stretching lineations strike parallel to the N120°E-trending southern contact and dip more than 70° SW and less than 20° ESE, respectively. In contrast, the eastern portion of the pluton, which outlines an EW-elongate ellipse in map view, shows coarse-grained porphyritic granites where the imprint of solid-state deformation is very weak. These features can be taken as evidence for synkinematic granite emplacement related to a dextral strike-slip shear zone. The abundace of trains of country rock xenoliths parallels to field and magnetic structures suggests that the ascent of magma was mainly achieved by dyke-propagation. Structural observations from the granite tail lead us to recognize the continuous interplay between magma extraction and shearing, since different sets of shear zones overprint successively younger dykes.
In order to establish the kinematics and the temperature of deformation affecting e.g. granitic rocks, it is common to carry out a crystallographic study of the XZ plane which can normally be determined from the observed foliation. In the Sequeros Pluton (SP) (Central Iberian Zone, Iberian Massif), however, the solid state deformation is weak, and it has not developed a visible foliation. In this case, the opposite reasoning has been applied as study method, i.e. the stress system has been deduced from the crystallographic study of randomly oriented thin sections.
The [c] axes orientation of the undeformed or weakly deformed quartz of the SP has been measured. A N-S subhorizontal trend can be distinguished in their apparently random distribution. This trend correlates with the average direction of magmatic flow of the pluton, which had been determined by Anisotropy of Magnetic Susceptibility (ASM) methods. When the quartz grains develop basal subgrain boundaries, the [c] axes trend preferentially between N-S/N30E in subhorizontal planes. This indicates that during magma emplacement and crystallization, some quartz grains position their [c] axes parallel to the direction of magmatic flow. These grains are favorably oriented to be deformed in subsolidus conditions by the same stress system that triggers magma emplacement.
Sometimes, quartz [c] axes have developed prismatic subgrain boundaries oriented in two main directions: one between N-S/N30E and another between N120E/N150E, both subhorizontal. The weak deformation affecting the rocks is not enough to produce a rotation of the subgrains which would explain the observed values. Thus, a favorable preorientation of the quartz grains in relation to the active stresses can be suggested. The quartz grains with the [c] axes oriented between N120E and N150E are favorably positioned to be deformed in solid state at high temperatures by the same stress that determine the magma emplacement. The case is different for the quartz [c] axes oriented between N-S and N30E. In order to develop prismatic subgrain boundaries, this quartz population needs a stress system that provokes maximun subhorizontal extension in a direction between E-W and N120E. This stress system is compatible with the third regional deformation phase. Therefore, we assume that part of the solid state deformation of these rocks is due to this phase.
In summary, the used method has allowed us to suggest that deformation in the quartz grains is due to two different stress systems, the first of which is related to the magma emplacement. A second system, with different orientation, is related to a regional phase of deformation.
Acknowledgements: Study financed by the DGICYT, Proyect PB-96-1452-C03-02.
It is one of the most important subjects in earth and planetary sciences to analyze the textures of rocks quantitatively. For the analysis of rock textures in 2D, some digital imaging techniques have been applied, such as back-scattered electron imaging and characteristic X-ray imaging with a scanning electron microscope (SEM) or an electron probe micro-analyzer (EPMA) (e.g. Dilks et al., 1985 and Launeau et al., 1994), and optical imaging with a CCD digital camera or an image-scanner etc. Here we have applied a scanning X-ray analytical microscope (SXAM) for the first time in the field of earth and planetary sciences to obtain X-ray fluorescence (XRF) images of rock sections (hereafter called X-ray maps). In this paper, we report the new method of image processing in which the X-ray maps of a rock are transformed to the maps that show distribution of mineral composition. As a test case, the X-ray maps of the Ryoke granite from Teshima, SW Japan, were processed to make distribution maps of major minerals.
The SXAM has been recently commercialized (Hosokawa, 1997). Continuous X-ray is focused by the X-ray guide tube (XGT) to 100 µm in beam diameter, and irradiated onto the surface of the sample plate which is set on a scanning table in the ambient atmosphere. Fluorescent X-ray photons are detected by the high purity Si detector, with which photon energy is measured. The sample is set on a scanning table, then XRF imaging of the sample in 2D is possible. The scanning area ranges from 2.56 by 2.56 mm2 to 200 by 200 mm2, and is divided into 256 by 256 pixels. The instrument can obtain XRF images of 30 elements selected from 13Al to 92U simultaneously.
In the image processing, XRF intensities of major elements (Al, Si, K, Ca and Fe) were assumed to have linear relationship with the composition of major rock-forming minerals (quartz, plagioclase, K-feldspar and biotite) in each pixel. The coefficients between XRF intensities and mineral compositions were determined by picking up the some pixels in which pure minerals exist, respectively. The composition of minerals was subsequently calculated by the maximum likelihood method for Gaussian distribution i.e. a least-square method. A numerical experiment showed that it is plausible to apply the least-square method if the operation time of the SXAM i.e. accumulation time of XRF is sufficiently long. The sources of errors in the calculated mineral compositions are derived from statistical errors of X-ray counts, variation of chemical composition in each mineral, and a condition of a sample's surface etc. We propose a method to estimate these errors. As an example, such mineral maps were applied to study a modal composition analysis of minerals.
Dilks A & Graham SC, J. Sed. Petrol, 55, 347-355, (1985).
Launeau P, Cruden AR & Bouchez J-L, Can. Mineral, 32, 919-933, (1994).
Hosokawa Y, Ozawa S, Nakazawa H & Nakayama Y, X-ray Spectrometry, 26, 380-387, (1997).
Lithified volcaniclastic rocks form significant strata in the sedimentological record, but alteration by diagenetic/metamorphic processes in many cases prevent characterization of eruption and emplacement histories. Indeed, lithification prevents adequate disaggregation and separation of pyroclasts from these tuffs. Volcanic ash shards, pumice fragments and magmatic minerals have been partially or totally replaced by secondary minerals, and have been cemented through diagenesis or metamorphism. The alteration/lithification processes can also provide the environment for rare mineral formation. Especially secondary REE-, Zr- and Y-bearing minerals are far more widespread than assumed. In developing the SEM as a petrographic tool to unravel the multitude of textural features of altered and lithified volcanic ash, we discribe a technique of using digitized backscattered electron (BSE) images that enhance the pyroclast shapes or their relict textures (Heiken & Wohletz, 1984) in polished thin sections. Being able to observe the shard shapes and pumice fragments allow conclusions to be made about eruption mechanisms responsible for the tuff deposit. The reconstruction of chemical and physical properties of ancient tuffs eases comparison with more recent ashes (Obenholzner & Heiken, 1999).
Heiken G & Wohletz K, VOLCANIC ASH, California Press, (1985).
Obenholzner JH & Heiken G, Ann. Naturhist. Mus. Wien (in print), 100A, (1999).
Solidification of igneous rocks must start by the nucleation of crystals (if none are already present in the liquid), but completion can follow several different paths: 1) More crystals can nucleate and grow. 2) Existing crystals can grow without nucleation of new crystals. 3) Textural coarsening (Ostwald Ripening, textural maturation) combined with overall growth can make large crystals grow at the expense of smaller ones, which dissolve. 4) Compaction can remove liquid. The crystal size distributions (CSD) of a series of samples from an intrusion can distinguish between these models. The CSDs of plagioclase, olivine and pyroxene (where present) were measured for a series of 11 samples taken from throughout the Kiglapait intrusion, Labrador and the data reduced using the CSD Corrections program (Higgins, submitted). Each CSD had a sharp cut-off at small and large sizes and many were slightly concave upwards between these limits. The right part of the CSD was regressed to examine variations between samples. The 'Characteristic Length' (CL) of the slope was calculated from CL = -1 / Slope. There is a broad negative correlation between CL and Intercept for all three minerals. The quality of the correlation is improved considerably if the intercepts are corrected for differences in the abundance of the minerals. Such a correlation is expected if textural coarsening and overall growth control the textures of these rocks (Higgins, 1998, submitted). Compaction either alone or combined with continuous nucleation and growth, or just growth, cannot produce these CSD variations. Hence the rocks of the Kiglapait intrusion solidified by textural coarsening and overall growth. Variations in initial nucleation densities are not now observable. Limited evidence from other intrusions indicates that this is true for all plutonic rocks.
Higgins, MD, American Mineralogist, (Submitted).
Higgins, MD, Understanding Granites: Integrating Modern and Classical Techniques, Geological Society of London Special Paper, (Submitted).
Higgins, MD, Journal of Petrology, 39, 1307-1325, (1998).
Index of EUG 10 Volume
Further EUG 10 Information
Index of the Journal of Conference Abstracts
Cambridge Publications Home Page
Last Updated on Wednesday, March 17, 1999.
© 1997 Cambridge Publications