Habitat loss is considered to be one of the greatest challenges currently facing society. An important part of the European Union's (EU) response to this problem is the Habitats Directive, the main aim of which is to protect biodiversity through the conservation and protection of natural habitats. Consequently, accurate mapping of specific habitats is of high importance in order to monitor ecosystem changes. Within this context, remote sensing has enormous potential as a source of land cover information which has been used widely for land cover mapping. In many instances this land cover mapping is only concerned with one particular habitat. However, in standard classification analysis, training data for all of the land cover classes contained in the area is typically required which means a wasteful use of resources. This thesis aims to address these issues by investigating advanced classification methods that focus on the accurate mapping of one specific protected habitat. The data used for this purpose is a Landsat ETM+ image of East Anglia, UK, acquired in June 2000 and ground truth data in the form of aerial photography. The habitat of interest chosen for this investigation is fens, a habitat protected by the EU Habitats Directive, whose diverse and dynamic nature is a particular challenge for its mapping and monitoring. A second protected habitat, saltmarshes, will be used for comparison purposes in order to determine any bias within the results. The methods considered to map the selected habitat consist of binary classifiers and one class classifiers. The binary classifiers chosen were Support Vector Machines (SVMs) and Decision Trees (DTs) which are two new methods very recently applied to land cover classification and remote sensing research. They are still very much the focus of current research for multiclass classification. In this thesis they are used in its binary form to classify the class of interest against. all the other classes. Both classifiers perform very well when compared with a classic parametric Maximum Likelihood Classification (MLC). Narrowing down the idea of classifying just one habitat of interest, one-class classifiers are put to the test. They have been explored in pattern recognition research but not yet within remote sensing image classification and land cover mapping. Specifically, the Support Vector Data Description (SYDD) classifier is considered particularly suitable for land cover classification as it is based upon the basis of SVMs which have already been applied in this area with success. When the results of the SYDD classification are compared against those obtained by the other classifiers these show an improvement in overall classification and a reduction in the errors of commission. Furthermore, another method is also put to the test to improve the accuracy of the classification of the class of interest. This method is the ensemble of classifiers, which in many research studies within pattern recognition has proven to improve accuracy of single classifiers. The results in this thesis also show an improvement in accuracy, although further investigation is needed. In conclusion, DT, SVM and SYDD classification methods offer clear advantages over standard classification analysis when concentrating on the classification and mapping of a particular habitat. All three classifiers obtained higher accuracies than the ML classifier with the use of significantly less training data. Furthermore, in the case of one-class classification only data from the class of interest was needed. Also in both binary and one-class classification approaches the attention was focused on separating the class of interest from all the other classes and therefore training efficiency was bigger than in a standard multiclass classification where efforts are directly to achieve a high overall accuracy. All three methods were found to be highly suitable for classifying and mapping a specific habitat and its application should be considered in future work involving the accurate mapping of protected habitats.

Late Quaternary (c. 11,400 [sup]14 C years BP) palaeoclimates were reconstructed from the sediments of closed lake basins in the western (Laguna de Cuitzéo) and eastern (Laguna de Zempöala) trans-Mexican volcanic belt (TMVB). Climate was separated out as a distinct agent of environmental change from the influence of anthropogenic, tectonic and volcanic activity by the use of a range of complementary techniques. These included mineral magnetic, geochemical, ostracod species assemblages, organic carbon, C/N, and authigenic [delta][sup]18 O and [delta][sup]13 C analyses, the establishment of independent chronologies and synthesis with previous work. At Laguna de Cuitzéo (core length, c. 11,400 BP) the Late Glacial was cold and wet with the Early Holocene (c. 1O,400-8,500 BP) arid. Wetter conditions established in the Mid Holocene (c. 7,500-1,300 BP) peaked with a period of methanogenesis (c. 6,600 BP). Drier conditions were recorded from c. 1,800 BP. An arid event c. 1,300 BP coincides with a peak in aridity identified in many records across the Sub-Tropics. At Laguna de Zempöala (core length c. 3,000 BP) the Late Holocene was wet before c. 2,600 BP with a drier phase between c. 2,600-2,000 BP before the return of wetter conditions until c. 600 BP. A 'Mega-Drought' recorded between c. 600-300 BP, coincides with a historically documented event. Wetter conditions existed from c. 300 BP until present. The high altitude closed lake basins of the TMVB provide high-resolution records of climate change. Mexico experienced a transition from a cold wet Late Glacial to an arid Early Holocene probably as a result of a northerly positioned Inter-Tropical Convergence Zone (ITCZ) displacing the north Atlantic subtropical high and weakening the easterly trade winds. Wetter conditions were established from c. 8,500 BP possibly because of orbital forcing extending the monsoon period. A transition occurred c. 1,800 BP into a period of apparent aridity (peaking c. 1,300-1,000 BP) probably resulting from solar forced variations, the periodicity of which modifies the intensity of the annual precipitation cycle leading to periods of drought and heavy rainfall on a bicentennial scale across the Sub-Tropics.

Key questions in the study of hydrothermal deposits related to magmatism are when, where, and how do magmas acquire their metals and fluids; and how do magmas contribute these components to the hydrothermal system for further ore concentration. This link between the magmatic ore source and hydrothermal deposits is unilaterally constrained due to emphasis on studies of ore bodies and diminished attention on the associated igneous evolution. The mid-Tertiary Sierra Madre Occidental (SMO) of Mexico is a large silicic igneous province in close spatio-temporal association to a premier silver epithermal region, and represents a suitable location to study the magmatic-hydrothermal link. To study the evolution of metal sources, rather than the pathway and final ore accumulation; this research has applied a magmatic point of view: employing geological fieldwork and zircon chronochemistry, together with rock, crystal and melt inclusion chemistry. This thesis presents the relation found between magmas and mineralization at a local scale in the Bolanos Graben, southern SMO. Pre-Extensional (Late Oligocene) and Syn-Extensional (Early Miocene) magmatic stages were recognized to record the volcanic history of the graben, with mineralization occurring during the second one. A novel igneous texture identified is interpreted to record the metal enrichment of magmas during the Pre-Extensional stage. In Early Miocene times, crustal recycling via the remelting of the enriched igneous crust formed in the Oligocene, aided by the onset of extension, may have enhanced magmatic enrichment and produced fertile magmas. These magmas may have contributed with metals, volatiles and other components to mineralization; or the formation of an enriched crust via a prolonged magmatic history provided material for hydrothermal fluids to scavenge. In conclusion, the petrogenesis of silicic igneous rocks and their evolution can give insights in when, where and how magmas acquire their metals for hydrothermal deposits to form.

During the last fifteen years the landscape of the Vale do Alto Tâmega, a region in Northern Portugal, has been subjected to environmental pressure leading to a change in land cover from predominantly forest to a mixture of shrub and rock outcrops. Both natural (e.g. wild fires) and anthropogenic (e.g. timber harvesting) factors have contributed to this change. Practical management techniques are necessary to preserve and manage this important environmental resource. New technologies, such as Geographical Information Systems (GIS), the Global Positioning System (GPS) and Remote Sensing (RS) provide tools that can be used for forest management and wild fire protection. This thesis explores the suitability of these technologies for managing forests located in the Vale do Alto Tamega region. It focuses on the development of data sets and models that can be used to assess the susceptibility of forests to fire hazards. The GIS analysis reveals that more than half of the study area presents a high fire hazard and nearly a third of the Pinus pinaster stands were burnt between 1986 and 1995. The GIS results also show that much forest land is occupied by unsuitable species (e.g. coniferous instead of deciduous trees), which contribute to restricted timber production and a high fire hazard. GPS was used to survey both the location of sampling plots and to map the boundaries of burnt areas or of post-fire recover areas, associated with traditional forest mapping and measurement techniques. These data were then used to update GIS and to support satellite image classification. Remote sensing derived data were used for mapping, providing up¬to-date land cover maps, and to derive predictive relationships which can be used to estimate forest biophysical variables (e.g. Leaf Area Index or Biomass) that can be used by forest managers. The results of the research demonstrate that GIS, GPS and RS could be used together in forest management and protection, such as in the classification of the land cover in order to calculate fire hazard indices or in the analysis of forests dynamics.

The Frontal Cordillera Composite Batholith (FCCB) is made up of a string of granitic plutons thought to be Permo-Triassic in age, that stretch with a N/S strike for the entire length of the Chilean and Argentine Frontal Cordillera (FC). Associated spatially and temporally with the batholith is a series of volcanic rocks called the Choiyoi Formation. Seven FC stocks were studied: the Boca del Rio, Cacheuta, Cerro Médanos, Cerro Arenales, Cerro Bayo, Punta Negra and Punta Blanca. They are high-K, I-type metaluminous, calc-alkaline granitic plutons, which range in composition from granodiorites to monzogranites to syenogranites. Contact relationships between the granitic stocks and the Choiyoi Formation, long thought to be their effusive equivalents, are intrusive, suggesting that the volcanic rocks are older. Chemically, the Choiyoi Formation rocks are also high K, alkalic and silica undersaturated, and similar to the granites. The stocks contain typical metaluminous modal mineralogies of plagioclase, quartz and alkali feldspar felsic phases, with amphibole and biotite as mafics, and magnetite, zircon, titanite, apatite and allanite as accessories. Field studies and P-T data from amphiboles suggest depths of emplacement of less than 10km. All the stocks have been affected by hydrothermal alteration, the main effects being sericitisation and/or argillisation of plagioclase cores, chloritisation of biotite and amphibole and oxidation of magnetite to haematite. The source of the hydrothermal fluids is thought to be the granites themselves. Four of the stocks were modelled with major, trace and rare earth elements. It was concluded that the Cacheuta stock evolved to aplitic compositions by ~35% fractionation of mainly plagioclase and alkali feldspar; the Cerro Arenales stock by ~30% fractionation of plagioclase, biotite and k-spar; and the Punta Negra stock by ~25% fractionation of plagioclase, alkali feldspar and quartz. The Punta Blanca stock was interpreted as consisting of more than one batch of magma. The granites from this segment of the FCCB differ from typical I-type granites in some crucial respects: 1) they are mainly high K, 2) alkali feldspar crystallised early and fractionated in the more evolved stocks, 3) the stocks contain high Ba and intermediate Sr, with U and Th enrichment in the leucocratic stocks, and 4) the stocks form part of a batholith in which granites predominate over granodiorites and tonalites. Typical I-type Cordilleran batholiths like the Coastal Batholith of Peru, are predominantly tonalitic to granodioritic. Three stocks from the study area were dated by the U-Pb zircon method. These are the first such dates for any plutons of the FCCB. The Punta Blanca stock was found to be 276 ± 1 Ma old, the Cerro Médanos stock 263 ± 1 Ma old, and the Cerro Bayo stock 262 ± 3 Ma old. All of these ages are in the Lower Permian. Sr-Nd isotopes suggest that the Choiyoi Formation and FCCB stocks were derived from different lower crustal sources which were variably enriched with [sup]87 Sr and unradiogenic Nd, extracted from the mantle at ~1.3 Ga and ~1.25 Ga, respectively. AFC processes are not thought to be important. The geodynamic circumstances which lead to the generation of the Choiyoi Formation and the FCCB are still not clear. The evidence collected during this study suggests that the Choiyoi phase of volcanism subsequent to the movements of the San Rafael Orogenic Phase at the end of the Carboniferous was possibly the result of extension, which caused extensive melting of a lower crustal source extracted from the mantle at ~1.3 Ga. This was followed by a switch in tectonic style in the Early Permian, possibly due to the intiation of oblique subduction, which lead to the sampling of a younger lower crustal source and the formation of the FCCB. This scenario is analogous to that recorded by Rapela et al., (1996) further to the south in the 'Gondwana' magmatism of Patagonia between the Late Triassic to Jurassic.

Increasingly farmers are being required to re-evaluate their role in relation to the conservation of wildlife habitats and landscape features on agricultural land. For many this challenges their construction of themselves as 'farmers', whose role is determined by their ability to harness the natural resources of the land and manage those resources to produce food. For many years this interpretation of their role was reinforced by government policies which encouraged the process of agricultural industrialisation. While many farmers still see themselves as 'stewards' of countryside, this stewardship has often been restricted to small areas of the farm, while the 'real' business of the farm continued to intensify. When the Countryside Stewardship Scheme was introduced in 1991, in response to EEC Regulation 2078/92/EEC, it represented a new type of funding scheme, one which required farmers to actively manage their land for conservation. Entry into CSS was not determined by dint of geographical location, and could only be attained by entering into a competitive 'bidding' process. For many farmers, the concept of being paid a grant to produce 'environmental goods' on agricultural land was completely alien. This thesis investigates a wide range of factors that can potentially influence a farmer's ability and willingness to adopt CSS. In addition to exploring the opportunities and constraints on adoption imposed by the farm business and farm household, the thesis explores the 'information environment' of the 112 farmer interviewees. The findings detailed in this thesis show that networks of institutional actors have developed around the promotion of CSS to farmers identified as having specific target habitats on their land. This thesis demonstrates that the availability of information and advice is a potentially powerful determinant of CSS adoption, whether the source of that information is institutional networks other influential local farmers who have adopted CSS.

It is standard practice for sewage treatment plants to discharge treated wastewaters to water-bodies and, in the UK, this has been the custom since the first sewerage system was completed in London around 1865. Before sewerage networks untreated sewage waste was discharged to rivers (wet carriage) or removed from domestic dwellings by "honey wagons" (dry carriage) and taken to rural areas to be used as fertiliser (Adams & Papa, 2000). Wastewater inputs to rivers are governed by discharge consents issued by the Environment Agency in accordance with existing legislation, taking into account the physical and ecological characteristics of the receiving river. However, it has become apparent in recent years that wastewaters may have a detrimental effect on receiving rivers, particularly in terms of nutrients as they frequently discharge phosphorus (P) in concentrations > 1.0 mg P 1[sup]-1. Metals are not specifically removed by the wastewater treatment process and, although their affinity for particulates results in some removal during processing, it is suspected that dissolved metals may be present in treated wastewaters. In 2000, the E.C. introduced the Water Framework Directive, which requires water bodies within member countries to attain "good ecological status" by 2015 in terms of biological and physico-chemical water quality. An umbrella Directive, it combines existing legislation on a range of contaminants including nutrients and metals with additional environmental standards to improve the quality of European waters. The purpose of this study was to consider whether the discharge of treated sewage wastewaters is likely to prevent rivers from achieving the standards required by this legislation. During an eighteen month period, water samples were collected from two second order rivers, the Bourne and Hogsmill, up and downstream of sewage treatment works. Samples were analysed for a range of P species and metals in filtered and unfiltered river water. Ultra-filtration was carried out on a selection of samples, as finer fractions are likely to be more bioavailable. River flow data was used to calculate potential contaminant loads downstream of the input source; this data was compared against published concentrations of agricultural derived P, historically considered the major contributor of P to UK riverine waters. Until the introduction of P removal processing at the Hogsmill, downstream concentrations of all P species in both rivers, were greater than upstream by an order of magnitude. P removal reduced P concentrations in the Hogsmill by more than 60% although they remain significantly greater than upstream. The majority of P occurred as soluble reactive P (SRP), the most bioavailable species, and the continuous nature of wastewaters discharge meant concentrations were high during the growth season. Current P-removal processes may not remove sufficient P to meet quality targets recommended by the U.K. Technical Advisory Group on the Water Framework Directive. Estimated P loads upstream of the wastewaters outflow are < 8 kg SRP day[sup]-1 for both rivers, the combination of increased P concentrations and greater river water volume downstream of the outflow result in estimated P loads of > 45 kg SRP day[sup]-1 in the Bourne and > 80 kg SRP day[sup]-l for the Hogsmill (after P-stripping). Comparison of P export figures from treated sewage wastewaters, calculated using daily load figures, with those from agriculture indicate that in all but the most rural catchments, sewage wastewaters are a greater source of P. Downstream, dissolved metals concentrations were not significantly different from upstream except for As in the Bourne and Pb in the Hogsmill; neither metal exceeded regulatory limits. River sediment from the Hogsmill and Bourne, and two additional rivers receiving sewage wastewaters, the Mole and Blackwater, were analysed for total metals and P. The ability of sediments to release (SRP) was investigated using kinetic release analysis. In the absence of U.K. standards, metals concentrations in sediments were assessed using the US EPA sediment quality guidelines (SQG). A range of metals in sediments from both the Hogsmill and Mole exceeded SQG to the extent that biota was likely to be affected. Although the greatest concentrations of total phosphorus in sediment were found in the Bourne, it is the Hogsmill which exhibits most potential for the release of SRP from sediment to the water column.

This project was initiated with the aim of determining the origin and evolution of the Belo Brdo Pb-Zn (Ag) carbonate-replacement and vein-type deposit in the Kopaonik metallogenic district of Serbia and Kosovo and to test its potential genetic association to nearby porphyry Cu-Au and skarn deposits. Hydrothermal-replacement, stockwork, skarn and vein-type base metal deposits in the region are related to a Tertiary (65.5 to 1.8 Ma) volcano-magmatic complex that intruded a tectonic nappe pile that includes a Triassic (251-199.6 Ma) Metamorphic Complex, a Jurassic (199.6 to 145.5 Ma) Ophiolite Unit and an Upper Cretaceous (99.6 to 65.5 Ma) Flysch Unit. Petographic and geochemical investigations of the Tertiary igneous rocks confirm a calc-alkaline chemistry for regionally distributed andesitic-dactic and quartz-latite rocks. These volcanic rocks show a clear subduction-related character. The Pb-ZN (Ag) minerlaisation at Belo Brdo evolved as a result of several successive stages of mineralisation includeing: (1) pre-ore stage, including early hydrothermal alteration and listwanisation followed by the deposition of Ni-Co-As sulphides; (2) main-ore stage, represented by massive sulphides including sphalerite, galena, pyrite, arsenopyrite, chalcopyrite, Ag-tetrahedrite-tennantite, bournonite; and (3) post-ore stage represented by quartz, carbonates + pyrite. A combination of fluid inclusion microthermometry, sphalerite and arsenopyrite geothermometry, and oxygen-isotype geothermometry suggests that the hydrothermal mineralisation at Belo Brdo was deposited from epithermal- to monderately high temperature (160-350[degrees]C), low salinity (6.5 eq. wt. % NaCl) fluids. Calculated [delta][sup]18O (6.5 to 10 ‰) and [delta]D (- 38 to - 62 ‱) values of water in equilibrium with sericite, tourmaline and quartz of the pre-ore stage are consistent with the ore fluids being derived from a magmatic source. Additionally, calculated oxygen isotopic composition of the main-ore stage fluid shows magmatic signatures (5.9 to 7.7 ‰). However, the water from fluid inclusions hosted by quartz from the late- ore stage breccia exhibits lower [delta]D (-93 to -96 ‰) and [delta][sup]18O (4.8 to 2.3 ‰) values, indicating an involvement of a more dilute, low-temperature meteoric hydrothermal fluids, Meteoric water played an increasing role toward the later stages of ore deposit evolution. The [delta][sup]34S values of the sulphides at Belo Brdo are homogeneous (average 4.4‰) and similar to those from nearby Cu-Au porphyry and skarn deposits in the study area, indicating a common source of sulphur for all deposits. The hydrothermal deposits in the study area show homogenous lead isotopic compositions ([sup]206P/[sup]204Pb= 18.68 to 18.76, [sup]207Pb/[sup]204PB= 15.67 to 15.68, and [sup]208Pb/[204]Pb = 38.87 to 38.91), similar to those of local calc-alkaline volcanic rocks ([sup]206Pb/[204]Pb - 18.76-18.82, [sup]207Pb/[204]Pb 15.66-15.68 and [sup]208Pb/[sup]204Pb 38.87-38.97 indicating a genetic link between minerlisation and magmatism. The data suggests that the Belo Brdo and the nearby Pb-Zn deposits are not Mississippi Valley type, but similar to a high temperature carbonate-replacement massive sulphide deposits associated with calc-alkaline magmatism in a supra-subduction zone setting.

The geology and commercial properties of china clay deposits in Cornwall and Devon have been investigated in the field and by laboratory methods. Two deposits were studied, the Lower Bostraze pit on Land's End and the larger Lee Moor pit on southwest Dartmoor. These china clay deposits are hosted in biotite monzogranites which have been subject to a history of hydrothermal alteration. SEM and XRD studies show that extreme argillic alteration is characterised by total destruction of the feldspars which are replaced by a fine-grained assemblage dominated by kaolinite and secondary mica. Partial alteration leads to assemblages of secondary mica, smectite and kaolinite. Pervasive alteration leads to a reduction of bulk rock density as a result of the leaching of alkali-metals and silica from the system. Kaolinite forms euhedral to anhedral platelets with larger booklet or vermiform stacks. TEM work has revealed that the kaolinite populations are composed of a mixture of euhedral and anhedral crystallites, a low disorder clay will be composed of euhedral crystallites with a subordinate number of anhedral kaolinites. The stack and booklet forms are common at higher structural levels within the deposits and are probably formed under supergene conditions. Fluid inclusion studies indicate a complex pattern of fluid/rock interaction. Inclusion abundances provide a measure of the amount of fluid that has passed through the granite, the most intensely altered granites having the highest values. The inclusion populations in kaolinized granites are dominated by regular to irregularly shaped liquid-vapour and vapour-only inclusions that sometimes contain a phase identified by SEM-EDS as kaolinite. The earliest episode of fluid activity was responsible for the formation of hydrothermal sheeted vein swarms, breccias, microfracturing and pervasive phyllie alteration. These formed from high temperature, dominantly magmatic fluids from between 250 to > 400°C with moderate to high salinities (10 to 40 eq.wt. % NaCl). These fluids were followed by cooler < 100 to 200°C, less saline (< 13 eq.wt. % NaCl) meteoric fluids that were responsible for the first stages of kaolinization. A small number of low/moderate temperature (180-220°C) high salinity (25-35 eq. wt. % NaCl) inclusions were found which are most likely to represent the influx of basinal brines during wrench fault re-activation and thus played a role in the kaolinization process. Homogenization temperature and salinity signatures for weakly and intensely kaolinized granites are similar and suggest that the degree of kaolinization is controlled by the volumes of fluid that pervade the granites. Stable oxygen and hydrogen isotope studies show that the isotopic compositions of kaolinite plot close to the kaolinite weathering line, indicating growth/re-crystallization in a supergene environment. Evidence for the epithermal regime under which kaolinization took place is found in both pits where late quartz-filled cross-course veins contain a significant amount of silica, iron oxides and re-mobilized/re-crystallized kaolinite. Stable oxygen isotope analyses of cross-course vein quartz from Lower Bostraze pit supports the fluid inclusion data which suggest a low temperature epithermal environment of deposition. The flow of fluids through the granite was controlled by the intensity of fracture systems present. Both pits host sheeted vein swarms and breccias which acted as regions of enhanced permeability and are coincident with zones of intense kaolinization. Pre-granite wrench-faults were re-activated following granite emplacement and played an important part in providing zones of high permeability during kaolinization. Laboratory studies have shown that the commercial properties of paper coating clays are controlled by their mineralogy, which is determined by the precursor mineralogy of the granite and its hydrothermal history. The yellowness of a clay is related to the amount of iron oxide present either adsorbed onto the kaolinite surfaces or as discrete particles. In general terms the higher the yellowness values the lower the brightness values. Abrasion is related to the levels of quartz and feldspar present. Anomalously high abrasion values at Lee Moor have been shown by SEM to be due to the presence of silica coatings on kaolinite and K-feldspar laths from antiperthites. Viscosity is related to the levels of mixed-layer smectite/mica phases and smectite in the ultrafine fraction. Viscosity is also related in a complex way to the particle size and shape of kaolinites. An intensely altered granite will contain well-formed blocky crystallites which will have good viscosity values, weakly altered granite will contain smaller thinner particles the viscosity of which is poor. The zones of intense kaolinization around vein swarms are dominated by blocky, well-formed low disorder kaolinite crystallites which have good viscosity and brightness values. Partially altered granite contains less kaolinite and more mica and smectite. The kaolinite will be dominated by less well-formed high disorder crystallites which may be stained by iron oxides thus increasing yellowness.

A method has been developed for the determination of gold and the platinum group elements at the low concentrations found in most geological materials, utilising the advantageous properties of Inductively Coupled Plasma-Mass Spectrometry (lCP¬MS) analysis. The method has been applied to assess the distribution of the precious metals in south-west England. Results from this study are combined with previously published data on the distribution of precious metal enrichments and new fluid inclusion data, to determine the origin of these enrichments. The methodology developed uses a sodium peroxide fusion of a small (1.0 g) sample of powdered geological material, with separation by tellurium co-precipitation and determination by ICP-MS. Isotope dilution calibration is used where possibe to overcome losses through the procedure, but data for monoisotopic elements (Au, Rh) are obtained by external calibration. Detection limits for the method are excellent, being < 0.5 ng g-l for Au, Pt, Pd and the minor PGEs: Ru, Rh and Ir. Good agreement with reference data for certified reference materials WGB-l, TDB-l and UMT-l was obtained. Pd data were reliable throughout the method testing. Data for Pt and Au improved through the method development phase. The minor PGEs gave reliable data for high concentration reference materials. Field sample data were only reliable for Pd, Pt and Au; minor PGE concentrations were too low to allow quantitation. Low concentration Ir results showed a systematic error.Small quantities of gold have previously been found in south-west England. New data confirm the presence of gold in mineral materials from Hope Nose and LoddisweIl. Most of the area shows low precious metal concentrations, probably due to the lack of potential source rocks. Anomalous precious metal concentrations are found associated with deep-seated cross-course structures. These may have allowed the ingress of precious metal-bearing fluids from deeper crustal levels. A spatial association with greenstone basic volcanics is also seen, but the greenstones analysed do not show anomalously high precious metal concentrations. Palladium shows a greater dispersion than either Au or Pt; small amounts being detected in most areas sampled.