Gastropods, like many other marine invertebrates undergo a two-stage life cycle. As the adult body plan results in narrow environmental tolerances and restricted mobility, the optimum opportunity for dispersal occurs during the initial larval phase. Dispersal is considered to be a major influence on the evolutionary trends of different larval strategies. Three larval strategies are recognised in this research: planktotrophy, lecithotrophy and direct development. Planktotrophic larvae are able to feed and swim in the plankton resulting in the greatest dispersal potential. Lecithotrophic larvae have a reduced planktic period and are considered to have more restricted dispersal. The planktic period is absent in direct developing larvae and therefore dispersal potential in these taxa is extremely limited. Each of these larval strategies can be confidently inferred from the shells of fossil gastropods and the evolutionary trends associated with modes of development can be examined using both phylogenetic and non-phylogenetic techniques. This research uses Cenozoic gastropods from southeastern Australia to examine evolutionary trends associated with larval mode. To ensure the species used in analyses are distinct and correctly assigned, a taxonomic review of the six families included in this study was undertaken. The families included in this study were the Volutidae, Nassariidae, Raphitomidae, Borsoniidae, Mangeliidae and Turridae. Phylogenetic analyses were used to examine the relationships between taxa and to determine the order and timing of changes in larval mode throughout the Cenozoic. Traditionally, planktotrophy has been considered the ancestral mode of development. However, using maximum parsimony and maximum-likelihood analysis, this research suggests that the ancestral developmental mode cannot be confidently determined in gastropods from southeastern Australia. Similarly, evidence that transitions between larval strategies might be reversible contradicts the general view that regaining the specialised structures associated with planktotrophy is so difficult that it is considered extremely unlikely to occur. When the timing of switches in larval mode was examined they were found to be scattered at different points in time rather than clustered to specific periods and therefore no inference can be made as to the likely factors driving transitions between larval modes. The correlation between mode of development and macroevolutionary trends was examined using non-phylogenetic techniques. The results do not concur with the hypothesis that species with planktotrophic larvae will exhibit wider geographic ranges, longer species durations and lower speciation rates then lecithotrophic or direct developing taxa. The analyses are thought to be hindered by a strong preservation bias and gaps within the fossil record. The quality of the fossil record and the congruence between phylogenies and stratigraphy is examined using the Stratigraphic Consistency Index, the Relative Completeness Index and the Gap Excess Ratio.

The principles of archaeomagnetism can be applied to answer archaeological and geomagnetic questions and examples of both are presented in this thesis. Firstly, this thesis demonstrates the use of archaeomagnetism to establish the maximum palaeotemperature reached in a kiln at the Oylum Höyük archaeological site in Turkey. A maximum temperature of between 600 and 700 °C was determined confirming that the feature was more likely to have been a lime kiln than a bread oven. Archaeomagnetism was next used to determine the relative ages of different construction events on the St Jean Poutge archaeological site in Southern France. The results of archaeointensity experiments on 137 core samples taken from bricks and tiles confirmed the different relative ages of two construction events. Average intensity values determined using Thermal Thellier-Thellier methods were 56±7 and 58±8 for the 2nd Century AD and 68±6 and 68±7 for the 3rd Century AD. There is an increasing body of evidence that the geomagnetic field in the Middle East during the Bronze Age reached exceptionally high field values extremely quickly. Both archaeomagnetic jerks (marked by sharp cusps in geomagnetic field direction coinciding with intensity maxima) and geomagnetic spikes (where the field rises and falls over a period of less than 30 years with associated virtual axial dipole moment fluctuations of at least 70 ZAm2 /~38 µT) have been proposed to have occurred in the Middle East between 3000 BC and 0 BC. Here, Coe and IZZI method archaeointensity experiments were carried out on 154 Bronze Age pot sherds from two archaeological sites in Turkey, Tell Atchana and Kilise Tepe, and 2 archaeological sites in Cyprus, Marki Alonia and Bellapais Vounous. In addition, thermal Thellier experiments were conducted on 18 mud brick cores from Tell Atchana. The results of these experiments were corrected for cooling rate whilst experimental design mitigated the effects of anisotropy. A success rate of 56% was recorded overall. The effects of applying cooling rate corrections, anisotropy corrections and the impact of varying archaeointensity selection criteria cut-off values, on the results, are discussed in this thesis. An average field value of 47µT was determined for Turkey over the time period ~2200-~700 BC which is indistinguishable from the current average field value. An archaeointensity value of 84 µT (153 ZAm2) was measured for the time period 800-600 BC. This is consistent with data from other authors who found evidence of high geomagnetic field intensity in the Middle East around 1000 BC. It is proposed here that this geomagnetic intensity high was of a longer duration and felt over a wider geographic area than has previously been suggested. Contrary to previously published studies based in Syria, evidence is presented here of decreasing geomagnetic field intensity in Cyprus between 2400 BC and 1900 BC. The proposal that an archaeomagnetic jerk was experienced in Cyprus over this time period is therefore rejected.

Geodetic observations of crustal deformation through the earthquake cycle provide unique opportunities to gain essential knowledge of faulting mechanisms, lithospheric rheology, and fault interaction. Normal faults, an integral geologic unit responsible for crustal deformation, are specifically investigated in this thesis, via three case studies in two significantly different tectonic environments: the 2008 Mw 6.3 Damxung and Mw 7.1 Yutian earthquakes on the Tibetan Plateau, and the 2005 Mw 7.8 Tarapaca earthquake in the northern Chile subduction zone. To move toward realistic slip models, I consider crustal layering for the Damxung earthquake, and non-planar rupture for the Yutian earthquake. The Damxung study shows that assuming a homogeneous crust underestimates the depth of slip and overestimates the magnitude, in comparison to a crustal model with a weak sedimentary lid. A curved fault model composed of triangular dislocation elements (TDEs) for the Yutian earthquake recovers the geodetic observation better than rectangular fault segments. Normal faulting earthquakes are characterized by shallow slip deficit, which is partially compensated by patchy afterslip around, but no deeper than, the coseismic rupture. The complementary and partially-overlapping relationship between coseismic slip and afterslip implies complexity of frictional properties in both down-dip and along-strike directions. Postseismic deformation induced by viscoelastic relaxation (VER) following normal faulting earthquakes is fundamentally different in pattern from that produced by afterslip. This difference enables identification of afterslip as the major postseismic mechanism for the Damxung and Yutian earthquakes, and VER for the Tarapaca earthquake. In addition to understanding the faulting mechanism, I also place constraints on local rheological structure. In central Tibet, where the Damxung earthquake occurred, lack of noticeable VER-related signal allows a lower bound of 1 × 1018 Pa s for the viscosity of the lower crust/upper mantle. In northern Chile, geodetic observations following the Tarapaca earthquake indicate a weak layer with viscosity of 4 – 8 × 10^18 Pa s beneath a higher-viscosity lower crust and mantle lithosphere, and a strong continental forearc. Based on the co- and post-seismic models, I investigate fault interaction from a perspective of static stress change. Stress computation suggests that the 2014 Mw 6.9 strike-slip event close to the Altyn Tagh fault occurred on a fault that was positively stressed by the Yutian earthquake, and the Altyn Tagh fault to the east of the 2014 rupture is a potential locus for future failure. Although the Coulomb stress change on the 2014 Iquique earthquake rupture is negative due to the Tarapaca earthquake and its postseismic VER process, positive loading on the shallow-dipping nodal plane of its M 6.7 preshock suggests that the Tarapaca earthquake may have acted as an indirect trigger of the 2014 Iquique earthquake. Both studies reveal the role played by normal faults in interacting with other types of faults and have implications for seismic hazard assessment.

Low velocity hydrous minerals in the subducting plate deliver water to the mantle and are thought to cause intermediate depth Wadati-Benioff zone (WBZ) seismicity through dehydration embrittlement. High frequency seismic energy (> 2 Hz) from intermediate depth earthquakes that occur within this low velocity oceanic crust is retained and delayed by the crustal waveguide while lower frequency (< 0.5 Hz) energy travels at the faster velocities of the surrounding mantle. These guided waves therefore spend longer interacting with the low velocity oceanic crust than any other seismic phase, and have the potential to reveal a large amount about the velocity structure of the WBZ. Dispersive arrivals recorded in the forearc of Northern Japan are directly compared to synthetic waveforms produced from full 2D and 3D waveform simulations. Comparing the relative amplitude and arrival time of a given frequency using the velocity spectra and spectrogram respectively, allows the full dispersive P-waveform to be constrained. Analysis of dispersive arrivals from upper plane WBZ events at 150 – 220 km depth place the first observational constraints on the metamorphic reactions occurring before full eclogitisation of the subducting oceanic crust. I show that blueschist and lawsonite bearing rocks may persist well beyond the depths inferred from established thermo-petrological subduction zone models, and that full eclogitization may occur at much greater depths than is inferred by receiver function studies. The persistence of meta-stable hydrous minerals explains the occurrence of WBZ seismicity at 200 - 250 km depth, and may be due to the partially hydrated oceanic crust. Dispersion from events that occur well below the upper plane of WBZ seismicity can be explained by the occurrence of low velocity hydrated outer rise normal faults at intermediate depths. At depth, these faults are inferred to be 2 - 3 km thick and 12 - 15 % slower that the surrounding mantle, suggesting they are 50 - 71 % serpentinised. We suggest that the extended P-wave coda observed at stations close to the trench in Northern Japan are explained by low velocity dipping faults of a range of scale lengths forming a scattering medium. This scattering medium is simulated using a von Kármán function, and the synthetic waveforms produced are compared to the observed P-wave coda, that decays in amplitude with distance from the trench. The magnitude of this spatial coda decay is sensitive to the average bulk velocity of the scattering medium and provides a constraint on the hydration of the lithospheric mantle subducted beneath Japan. This first in-situ constraint on the degree of slab mantle hydration at intermediate depth suggests that 170.4 - 318.7 Tg/Myr/m of water is subducted beneath Northern Japan by the slab mantle. In summary we have shown that up to 94 % of the water subducted beneath Northern Japan is transported by the lithospheric mantle, and that upper and lower planes of WBZ zone seismicity are directly related to hydrous mineral assemblages, and so may occur through dehydration embrittlement. This work shows that guided waves have the potential to resolve new details of the WBZ velocity structure and the techniques developed here can be applied to other subduction zone settings.

A palaeomagnetic study of some of the best preserved Palaeoarchaean (3.2 -3.6 Ga) rock successions in the world – the Barberton Greenstone Belt, South Africa – has been carried out in order to shed light on Early Earth processes. The aim is to improve the understanding of the long term evolution of the Earth and the surface conditions under which the first forms of life originated through using palaeomagnetic records. The study follows on from work by Layer et al., (1998), Yoshihara & Hamano (2004), Usui et al., (2009), Tarduno et al., (2010) and Biggin et al., (2011) that indicate that rocks from the Barberton Greenstone Belt have the potential to record a near-primary direction of remanence and a reversing geomagnetic field at ca. 3.5 Ga. The rocks of the Barberton Greenstone Belt are excellently preserved and have only been subjected to low grade metamorphism (greenschist facies), making them good candidates for palaeomagnetic studies. Here, new data obtained from three Onverwacht Group Formations and from the Nelshoogte Pluton are presented. The reliability of the new palaeomagnetic data is affected by the complex history and the age of the rocks but no more so than any other published study of rocks of this age. The Noisy Complex and the Nelshoogte Pluton are affected by lightning induced isothermal remanent magnetisations (IRMs) which result in poor quality results. Nevertheless, two new poles, produced by combining the new findings with previously published data, are calculated for the Komati and Hooggenoeg Formations. The new Komati pole shows improved clustering when compared to previous studies. Whilst the data of the Hooggenoeg Formation are encouraging, the results are ambiguous and open to interpretation. When combined with the results of Biggin et al., (2011) they exhibit considerably improved clustering when the directions are corrected for the tectonic deformation resulting in the formation of the Onverwacht Fold, dated at 3.23 Ga. The new results presented here support the findings of previous palaeomagnetic studies of the Barberton Greenstone Belt and support the existence of a stable geomagnetic field at ca. 3.5 Ga. The results presented here support the findings of Biggin et al., (2011) of moderate latitudinal plate motion during this time and do not rule out the hypothesis that the Pilbara Craton (Western Australia) and the Kaapval Craton (South Africa) were conjoined in the Palaeoarchaean.

The structures and dynamics of intra-salt layers have received limited study in comparison with the external shape of salt structures. Our limited understanding of the behaviour of intra-salt layers generally comes from salt mines, outcrops, analogue, and numerical modelling where the full three-dimensionality of intra-salt layers is barely observed. To understand the internal dynamics of giant salt structures and the response of their intra-salt layers during regional tectonics, this thesis provides detailed interpretation and analysis of intra-salt layers from the Silverpit Basin, in the Southern North Sea Basin, and the Birba Area, in the South Oman Salt Basin. These two locations provide unique natural laboratories where driver mechanisms for salt tectonics are investigated using high-resolution, high-quality three-dimensional (3D) seismic reflection data. The Silverpit Basin is a buckled basin formed during the Mid Eocene to Late Oligocene, while the Birba Basin was affected by massive sediment loading, which generated differential loading from the Early Cambrian to the Late Permian. Differential loading of the basin caused down-building and influenced the growth of diapirs and minibasins, which later led to intense deformation and fragmentation of the intra-salt carbonate stringers. In the Silverpit Basin, regional salt anticlines encapsulated a 23–63 m-thick intra-salt layer known as the Z3 Stringer. Lithologically, the Z3 Stringer is composed of anhydrite, and it represents a strong seismic marker across the Southern North Sea Basin. Relative to regional anticlines and synclines at the Top Salt level, the Z3 Stringer deformed in a ductile manner comparable in geometry and attitude to the regional salt structure. Non-cylindrical stringer folds, which vary from gentle to isoclinal, are related to the intensity of the regional-scale structure, whereby tighter vii stringer folds are observed under well-developed Top Salt anticlines and synclines. Synclines at the Top Salt level include long-wavelength gentle folds. Extreme thinning of the Zechstein by the downward displacement of the Top Salt causes the stringers to extend and finally break laterally in a mode-1 tensile fracture mechanism. This thesis highlights the complexity of intra-salt deformation and forms a good large-scale case study for the analysis of the kinematics and rheology of competent material enclosed within an incompetent medium. Understanding the complexities and attitudes of intra-salt layers and their encasing salt structures has broader implications for regional tectonic history, hydrocarbon prospectivity, and industrial applications.

Mafic sill complexes are increasingly being shown to play a major role in the movement of magma around the upper crust in volcanic terranes and to have a role in mass extinction events in Earth history. Most of the current models of sill emplacement assume that brittle fracture operates at all points of sill emplacement. Within this thesis, a series of observations are presented from sheet intrusions in South Africa, USA and the UK showing that in certain situations, dependent on host rock lithology, the propagation of magma through normal brittle fracture can cease. In this circumstance a prevalent fluid/fluid or fluid/ductile relationship between host rock and intruding magma is often developed. Once this occurs, the evolution of a given sheet intrusion becomes distinctly different from that produced by normal brittle fracture alone. The break down in brittle fracture often leads to the development of magma fingers, which accelerate ahead of the main sheet of magma. It is important to note that it is ultimately the host rock lithology and its coupled response to intrusion of magma that dictates the ongoing evolution of the morphology of sheet intrusions in high-level magmatic systems.

The Much Wenlock Limestone Formation (Silurian: Wenlock, Homerian) of England and Wales contains a diverse invertebrate fauna including many rare and problematical taxa. This study investigates the palaeobiology and palaeoecology of five such groups: asteroids (starfish), the crinoid Calyptocymba mariae gen. et sp. nov., rostroconch molluscs, machaeridians and cornulitids. Six species of asteroids are recognized, including three new forms (Palasterina orchilocalia sp. nov., Hudsonaster? carectum sp. nov., and Doliaster brachyactis gen. et sp. nov.), that show a diverse range of morphologies. Most distinctive is the 13-rayed Lepidaster grayi Forbes, 1850, the oldest known multiradiate starfish. The variety of asteroid body shapes indicates a diversification of behaviour, particularly feeding strategies. By functional convergence, L. grayi is interpreted as an active predator. C. mariae is an unusual species of camerate crinoid, with a small, thinly plated calyx. This, combined with its slender body morphology, is interpreted as an adaptation to life in a low energy environment. Four rostroconch taxa [Mulceodens aedicula sp. nov., M? aequicostatus (Phillips, 1848), M. latus sp. nov., and Redstonia sima sp. nov.] occur, and are interpreted as mobile, semi-infaunal, deposit feeders. Differences between Mulceodens and Redstonia probably reflect adaptations by closely related taxa to specific environments, with Mulceodens living in higher energy conditions, rather than indicating that they belong to separate families as has been suggested previously. Machaeridians are problematical taxa having elongate bodies covered with serially repeated calcitic sclerites, and three species – Turrilepas wrightiana (de Koninck, 1857), Lepidocoleus ketleyanus (Reed, 1901) and L? extraplax sp. nov. – are recognized in the Much Wenlock Limestone Formation. By analysis of their body morphology, skeletal structure and growth pattern, machaeridians are interpreted here as molluscs, probably the sister group of Polyplacophora + Conchifera. Using a similar approach, Cornulites is interpreted as a stem group anthozoan. Cornulites species show various life positions: C. scalariformis Vine, 1882, attached itself to live brachiopods, with its aperture positioned close to the host’s feeding currents, C. gremialis formed clusters growing upon one another, and C. cellulosus was solitary and unattached.

Sub-volcanic intrusive networks, of which cone sheets are recognised as a major constituent, control volcano growth and eruption style. The accepted cone sheet model is that these confocally dipping intrusions originate from an unexposed central magma chamber through dip-parallel magma flow. However, the emplacement mechanism of cone sheets has remained poorly understood. The classic ~58 Ma cone sheet swarm of Ardnamurchan, NW Scotland, offers an excellent opportunity to further resolve the emplacement dynamics of cone sheets through studying their magma flow. Structural measurements and anisotropy of magnetic susceptibility (AMS) analyses have constrained a lateral magma flow regime, consistently oriented NW-SE, in the majority of the Ardnamurchan cone sheets. This is not consistent with previous emplacement models. In this thesis, it is suggested that emplacement of the Ardnamurchan cone sheets occurred through the deflection of laterally propagating, NW-SE trending regional dykes, sourced from laterally adjacent magmatic systems (likely the Palaeogene Mull central complex). Field observations highlight the importance of host rock structure and interference between locally compressional and regional extensional stress fields in controlling intrusion geometry. Implicitly, edifice construction and potential eruption precursors observed at a volcano may instigate, or result from, magmatic activity in laterally adjacent volcanic systems.

There has recently been a steady increase in the number of licenses granted for the abstraction of water from the Chalk aquifer beneath London to supply "open loop" geothermal systems (Environment Agency, 2007). However, there has been little research conducted on how the water re-injected by these systems, which often differs in temperature by as much as 10°C, will interact with the fractured Chalk aquifer in both the short and long term. An analytical solution developed by Bodvarsson (1989) was used to show that, for most configurations of a geothermal system, thermal transport would be governed by fractures. It was then proved that the United States Geological Survey SUTRA code could be used to construct a more detailed model of the aquifer. A thermal test was devised to collect hydrogeological and thermal data. This test, along with conventional site investigation techniques, was used at a site in central London. A detailed numerical model of the geothermal system and the aquifer was then constructed in SUTRA. The results showed that the fracture zones found during testing would affect the system performance. Building on these results a procedure was developed for designers, to ensure such systems function in an appropriate way.