Anisotropy in the physical and transport properties of crustal rocks is a key influence on crustal evolution and energy resource management. Data from deep seismic soundings, borehole logging and laboratory measurement all show that the physical properties of the earth are anisotropic. Such anisotropy generally results from the superposition of fabric development during diagenesis and/or petrogenesis, and the application of anisotropic tectonic stresses. This leads to an aligned crack and pore fabric in crustal rocks that, in turn, leads to seismic velocity anisotropy and permeability anisotropy. This thesis describes an experimental study which aims to investigate the relationships between pressure, pore fabric geometry and seismic and permeability anisotropy under hydrostatic pressures from room pressure to ~4km depth equivalence within the Earth's crust. Firstly, pore fabric analyses of three representative crustal rock types is presented. These rock types represent a range of crack and pore fabrics. The average void space shape and orientation is determined 3-D using the methods of anisotropy of magnetic susceptibility and velocity anisotropy. Scanning electron microscopy and fluorescent-dye crack imaging techniques further aid in the void space characterisation. Secondly, the development and application of an apparatus capable of contemporaneously measuring elastic wave velocity, porosity and permeability at effective pressures of up to 100 MPa is described. Results are analysed in terms of applied effective pressure and the rock pore fabric type and orientation. Finally, the laboratory data are used to test models that attempt to predict geophysical parameters such as permeability and elastic wave velocity from microstructural attributes. This multi-facetted analysis allows a number of conclusions to be drawn, expanding the state-of-the-art in how the pore fabric microstructure of crustal rock is represented by the methods of elastic wave velocity and porosity, with reference to the hydrostatic confining pressure and hence the burial conditions of the rock.

The emplacement processes of the Whin Sill complex and its associated dykes have been studied using a combined approach of detailed field and magnetic investigations. Regional palaeodirectionai results show two different but consistent palaeomagnetic signatures and allow the sill complex to be subdivided into three geographically separate intrusions: the 'Holy Island Sill', the 'Alnwick Sill' and the ' Hadrian's Wall-Pennine Sill', The Little Whin Sill has been identified previously as a fourth, separate sill on petrological evidence. On the basis of the palaeodirectional results it is also possible to relate the exposed contemporaneous dykes to the individual intrusions. The magma flow geometries within the individual Whin Sill intrusions were detennined by AMS analyses, and both traditional and newly identified magma flow indicators. For the Holy Island Sill the results suggest that the magma flow was homogeneously southwards directed and that the Holy Island Dyke acted as the feeder to this intrusion. In the Alnwick Sill the magma flow was homogeneously westwards directed and it is proposeU that an offshore, en echelon segment of the High Green Dyke fed the Alnwick Sill. The magma flow geometry for the Hadrian's Wall-Pennine Sill is complex. The results suggest that this intrusion was fed by the Hett Dyke and that the magma flow within the sill was generally north and north westwards directed. It is proposed that the intrusion was emplaced during a phase of late Variscan, E-W compression and that pre-existing faults at a high angle to this compression direction acted as a magma flow barrier along which the magma became deflected. The geometry of the sills is approximately that of a quarter- to half-saucer-shape, with the dykes being situated at the saucer truncation. From the feeder dykes magma was injected into the individual sill bodies and the magma flow was generally up dip or parallel to the strike of the host rock bedding, thus out of the basin centres and into levels of lower lithospheric pressure. It is proposed that a compressional stress field and locally overpressured horizons had a significant impact on the initiation and emplacement of the Whin Sill complex.

No description available.

No description available.

The aim of this thesis is to get a better understanding of the composition of a primary melt. Kimberlites are the deepest magmas known to erupt on the Earth's surface, and thus provide us with discrete, direct sampling of the upper mantle. Kimberlites are thought to reach the surface of the earth directly from the source region (150 - 200 km), but in spite of this, they do not correctly represent the source region due to the alteration experienced upon ascent and after emplacement (eg. serpentinisation). Determining the composition of a primary melt that gives rise to kimberlites, can give us important clues on melt generation in the mantle and the composition of the upper mantle. Constraining the true volatile content of these magmas has important implications regarding the eruption dynamics and the ascent ofkimberlite magma from the mantle. We use the petrology and geochemistry of the Igwisi Hills Volcano (IHV), Tanzania, and experiments on a simple peridotitic model system CaO-MgO-Ah03 -Si02 -H20-C02 -K20 (KCMASH-C02) to gain insight into the composition of a primary melt. From the ground mass mineralogy of IHV, there is evidence for magma mixing and stalling which is disrupted by a second pulse of kimberlite magma. This is reflected in the presence of forsterite inclusions in a titanomagnetite host, two different perovskite species with each ex- periencing a different range of oxygen fugacity, the presence of annite (Fe-rich) and phlogopite (Fe-poor), and the two types of NiO diffusion profiles seen in olivines that are next together in the same sample. Phosphorous zoning provides evidence for rapid crystallisation which is likely due to the onset of an eruption. Igwisi olivine is exceptionally wet, with up to 150 ppm (or 245 ppm according to FTIR) H20 found in the cores of several grains. This would imply over 5 wt% H20 in the source region. A change in H speciation is accompanied with variable H20 concen- tration, reflecting the changing conditions that the melt sees during ascent. The homogenous H20 content in olivine from pyroclastic deposits suggests that equilibrium was reached between the water content in the melt and olivine. Experiments in the KCMASH-C02 system illustrate that at isobaric univarience, phlogopite does not take part in the melting reaction. Melting starts between 1000 and 1050 QC, 100 QC lower than the solidus in CMAS-C02 . At 30 kbar and 1400 QC, a melt with a kimberlitic composition is in equilibrium with a carbonatitic melt and demonstrates a kimberlitic melt forming at relatively low pressures in the melting of a K-bearing peridotite.

Constraining the structure of arc crust is critical to our understanding of arc magmatic systems and to test the hypothesis that continental crust is generated at arcs (e.g. Tay- lor, 1967). In this study, petrological information has been combined with information on rock physical properties from laboratory measurements and theoretical calculations to produce the most thoroughly constrained model to date of the structure of the crust beneath the active volcanic island of Montserrat. Ultrasonic velocity measurements together with the seismic tomography model of Paulatto et al. (2012), support geological evidence that andesite dominantly comprises the shallow crust. Igneous crustal inclusions found in host andesite and high velocity regions identified in the seismic tomography model are consistent with shallow intrusive complexes beneath the three volcanic centres. A low velocity region is distinguished in the seismic tomography model beneath the currently active Soufriere Hills centre, most likely due to the presence of melt. The volume of this magma storage region is highly dependent on the geometry of melt pockets. Using constraints from this study, the best model to explain this low velocity zone is a magma storage region between 5.5-7.5 km depth with a volume of ~13 km ' and melt fraction of ~0.30-0.35 (Paulatto et al., 2012). Cumulate-textured inclusions provide direct evidence for fractional crystallisation at Montserrat. Thermobarometry calculations indicate that most of these inclusions crystallised between 3-6 km depth, providing further evidence in support of a shallow magma chamber. Some inclusions however, including one erupted in the current eruption, are calculated to have formed at depths of up to 22 km from a highly water-rich magma with up to 8.9 ±1.3 wt% H20. Magma may therefore be ascending from the lower crust to feed the current eruption.

This is a study of kimberlite and lamproite intrusions located in southern India within the Eastern Dharwar craton (EDC). Bulk-rock major-element, trace-element, REE- and Nd- isotopic data support previous studies that have classified the ultramafic magrnatism located in the EDC as either Group I kimberlite or lamproite. This study reports new high-precision Ar-Ar age data for southern Indian ultrapotassic 'rocks (kimberlites and lamproites). Previously, the Chelima lamproite (ca. 1400 Ma) was considered to be one of the oldest lamproites in the world. However, our age data suggest that at least one lamproite (Pochampalle from Krishna lamproite field) was generated in the same region 100 Ma before the Chelima event. The Pochampalle lamproite was emplaced around ~ 1500 Ma as shown by the Ar-Ar data in this study, roughly 250 Ma before the other Krishna lamproites. It would seem that the Pochampalle lamproite was also derived from an isotopically distinct source region with a lower 143Ndl144Nd ratio than other Krishna lamproites. These findings not only have implications for regional ultramafic/ultrapotassic magrnatism, but also demonstrate that the mantle processes for producing lamproitic melts existed earlier than previously thought. This study also presents the first Hf isotopic analyses of perovskites from Indian kimberlites. Perovskites fall into two distinct groups in Hf isotope space, suggesting a degree of heterogeneity in the source of the kimberlites in the EDC. The kimberlites are all found in such close proximity that lateral variations in the source are unlikely, but this does not preclude vertical heterogeneity in the source region(s). I propose a model of kimberlite generation with a vertically heterogeneous source that has undergone separate periods of enrichment and depletion before kimberlite melt generation and emplacement at ~ 11 00 Ma.

Zircon (ZrSiO4) is a durable mineral found in most igneous rocks; it is highly retentive of the trace element concentrations it acquires at crystallisation, and its high initial concentrations of U and Th relative to Pb make it the most important mineral for radiometric dating. Therefore, it is a valuable archive of magmatic processes, particularly with regard to those occurring on the early Earth. Based on crystal chemistry, anomalous Ce and Eu concentrations in zircon relative to other rare earth elements (REE) seem likely to reflect the oxidation state of the magma. Zircons were grown experimentally under controlled conditions of oxygen fugacity (fO2) and the crystals and coexisting glass were analysed by SIMS and LA-ICP-MS to examine the covariation of the partition coefficients of Ce and Eu, as well as those of other trace elements. This revealed that with increasing fO2, Ce becomes more compatible and Eu and U become less compatible. There is a narrow window of fO2s in which a Ce and a Eu anomaly coexist. Literature data allow the partitioning data obtained for the heavy REE in this study to be extrapolated to other temperatures. To allow extrapolation of the partitioning experiments, Ce- and Eu-doped glasses of various melt compositions were prepared at a range of fO2s and temperatures. X-ray absorption near edge structure (XANES) spectroscopy of these glasses was carried out at the LIII-edge of these elements to determine their oxidation state ratios. Because of beam damage effects for the Eu-bearing glasses, a limited number of XANES spectra were recorded in situ at 1400 °C, and some samples were analysed by electron paramagnetic resonance spectroscopy. The results obtained were compared to trace element concentrations in zircons from some natural samples, and suggestions for future work made.

Afin de mieux comprendre la serpentinisation des roches mantelliques au niveau des dorsales océaniques lentes, diverses séries d’expériences ont été réalisées à 50 MPa et 250, 300 et 350 ℃ pour quantifier l’H₂ produit et clarifier les liens entre la production d’ H₂, les phases minéralogiques produites et les propriétés rédox dues à la présence d’ H₂. Les résultats, qui constituent un effort cohérent de prise en compte des paramètres et conditions rédox lors de la serpentinisation, sont les suivants : – Une nouvelle méthode expérimentale de mesure in situ de la \fH a été mise au point à 250 et 300 ℃, 50 MPa. Les deux variables exprimant la concentration d’H2 dans le fluide, \molalHydAq et \fH, ont été reliées quantitativement. Ces résultats permettent le calcul de la \fO dans les systèmes hydrothermaux océaniques à partir de la concentration en hydrogène dissout. La production d’ H₂ commence précocement, augmente rapidement et est continue au cours de la serpentinisation. – Les expériences d’interaction harzburgite-eau de mer montrent que la serpentinisation est une dissolution irréversible de l’olivine et de l’orthopyroxene et qu’elle se décompose en une succession d’étapes impliquant des assemblages de phases hors d’équilibre et métastables. – Les mécanismes de la serpentinisation et les assemblages de phases produits (en particulier les oxydes de fer) dépendent de la \fH laquelle est étroitement contrôlée par les régimes de circulation des fluides dans la croûte océanique. – La modélisation thermodynamique des fluides hydrothermaux riches en H₂ souligne le besoin de connaître précisément \yHydAq , le coefficient d’activité de \hydAq. / To better understand serpentinization of mantle rocks at slow-spreading ridges, several series of experiments have been conducted at 50 MPa and 250, 300 et 350 ℃ to quantify the H₂ production associated with the serpentinization process, and to clarify the relations between the H₂ generation, the nature of product mineral phases and the redox properties of H₂ bearing hydrothermal systems. The main results of this work, which represents a significant effort toward the consideration of redox parameters and processes during serpentinisation, are the following: – A new experimental method of in situ monitoring of the \fH has been set up at 250 and 300 ℃, 50 MPa. The two variables expressing the H₂ production, \molalHydAq; aq and \fH , have been related quantitatively. Results enable the \fO of hydrothermal oceanic systems to be computed from the dissolved hydrogen concentration. H₂ production starts early, increases rapidly and is continuous in our serpentinization experiments. – Harzburgite-seawater interaction experiments show that serpentinisation is an irreversible dissolution reaction of olivine and orthopyroxene and that it consists of a sequence of discrete steps involving metastable and disequilibrium phase assemblages. – Serpentinisation mechanisms and phase assemblages (especially Fe oxides) depend on \fH, the latter being closely controlled by processes of fluid circulation in the oceanic crust. – Thermodynamic modelling of H₂ rich hydrothermal fluids stresses the need to know precisely \yHydAq, the activity coefficient of \hydAq.

Η Φυλλιτική-Χαλαζιτική ενότητα αποτελεί τμήμα των Εξωτερικών Ελληνίδων και έχει υποστεί μεταμόρφωση σε συνθήκες υψηλής πίεσης. Στην Κεντρική Κρήτη εμφανίζεται κυρίως στο βόρειο τμήμα του παραθύρου των Ταλαίων ορέων. Σκοπός της διπλωματικής εργασίας είναι να προσδιοριστεί η γεωμετρία του τριαξονικού ελλειψοειδούς της παραμόρφωσης και το ποσό της παραμόρφωσης στη Φυλλιτική-Χαλαζιτική ενότητα. Για το σκοπό αυτό συλλέχθηκαν 18 προσανατολισμένα δείγματα χαλαζιακής σύστασης και μετρήθηκαν ελλειπτικοί δείκτες της παραμόρφωσης σε δύο κάθετες μεταξύ τους τομές από κάθε δείγμα. Τα στοιχεία φανερώνουν ότι κατά την πλαστική φάση παραμόρφωσης D1 τα πετρώματα της Φυλλιτικής-Χαλαζιτικής ενότητας παραμορφώθηκαν κυρίως σε συνθήκες πλάτυνσης και σε μικρότερο βαθμό σε συνθήκες επίπεδης παραμόρφωσης. Επιπρόσθετα το ποσό της παραμόρφωσης έχει μια συστηματική και μη γραμμική αύξηση σε σχέση με την απόσταση από την επώθηση βάσης, ενώ ο τύπος του τριαξονικού ελλειψοειδούς δε συνδέεται με τη δομική θέση των δειγμάτων. / The Phyllite-Quartzite (PQ) unit is a part of the External Hellenides and has been subjected to a high pressure metamorphism. In Central Crete PQ unit mainly appears in the northern part of the Talaia window. The aim of this study is to determine the strain ellipsoid geometry and the intensity of deformation in the PQ unit. To do so 18 quartz-rich samples were collected, in which elliptical strain indicators were measured in two mutually perpendicular cross sections. Measurements show that during the D1 phase of the ductile deformation, rocks of the PQ unit were deformed mainly by flattening and secondary by plane strain conditions. Additionally, the amount of deformation shows systematic and non-linear increase relative to the distance of the Basal thrust while the type of strain ellipsoid is not related to the structural position of the samples in the PQ.