The upper part of the Great Estuarine Group (Middle Jurassic) comprises three lithostratigraphic formations, the Duntulm, Kilmaluag and Skudiburgh Formations. They record a series of paralic palaeoenvironments, marine-brackish lagoons. low-salinity, ephemeral lagoons and alluvial plains respectively. This general transition of environments was produced by late Bathonian regression. Algal limestones from the Duntulm Fm. are interpreted as lagoon-marginal, algal marshes which underwent synsedimentary calcification by magnesium-calcite. Stable-isotope ratios of oxygen are light, delta18O = -4%, which suggests this calcification occurred in an environment with significant freshwater input. Stable-isotope ratios of carbon are also light and imply a pore-water system with abundant light organic carbon. Early diagenetic carbonate microfabrics were probably influenced by microbial systems. The algal marsh environment was hostile to the usual lagoonal fauna, mainly through exposure. The Kilmaluag Fm. is dominated by a low-salinity fauna. Accumulation of muddy detrital sediment was common, while fine-grained carbonates were probably primary precipitates of calcite and magnesium-calcite. During evaporative periods concentration of Mg allowed synsedimentary dolomitization of these primary carbonates. Stable-isotope ratios of oxygen for dolomite are the heaviest reported from the Great Estuarine Group and corroborate the theory formation of dolomite from evaporated waters of low salinity. Depositional clay mineral suites from mudrocks are largely unaffected by burial diagenesis. The assemblages contain significant amounts of smectite, and one bentonite horizon in the Kilmaluag Fm. suggests that smectite was mainly of volcanic origin. Immediately adjacent to Tertiary igneous centres, mild metamorphism caused destruction of the smectite phase. The formation of calcified, freshwater-influenced algal marshes, probable seasonal evaporites, early diagenetic dolomites interspersed with calcitic limestones in ephemeral lagoons and nodular calcretes in alluvial sediments, all suggest a warm palaeoclimate with marked rainy and dry seasons.

The problem of strategic decision making in the metalliferous minerals industry has, to date, tended to have been solved by a stochastic process. This thesis describes a new approach to this problem involving rational decision making for the orientation of mineral exploration efforts. The thesis is composed of two basic parts, the first being the specific statement of the problem, underlaying assumptions and constraints, and its theoretical solution. The second part being an example of the use of the theory by a hypothetical mining company to determine the best exploration strategy, and a review of the status of known deposits in the light of the results of the strategy developed. Success is defined, in general, as the excess of reality over desire. Using this concept in exploration, reality is expressed as a series of grade-tonnage curves representing the sources of the commodity. Financial desire is initially defined as an internal rate of return, but this is then translated to equivalent grade-tonnage combinations and is then also depicted as a series of grade-tonnage curves. The chances of exploration success are then determined by overlaying the grade-tonnage curve of reality on that of desire. On the basis of this overlaying specific deductions are made regarding the relative amount of effort that can be rationally justified for each commodity. In addition, specific, attractive deposit types are identified and minimum grade and tonnage criteria are calculated for each deposit type within each commodity. Finally, these specific conclusions are combined to form the best overall strategy for investment in mineral exploration by a hypothetical company.

Late Precambrian meta-sediments and Caledonian granitic intrusions are exposed in Leicestershire. Similar rocks are found at shallow depth in boreholes from Leicester to East Anglia. Two seismic refraction experiments, each comprising two 10-30 km. profiles, have been carried out over potential field anomalies in the East Midlands. The Melton Mowbray experiment, located to the east of the Charnian Inlier, defined the top surface of the Melton Mowbray granitic intrusion at less than 0.5 km. depth, and the southern margin of the Carboniferous Widmerpool Gulf. The Melton intrusion was found to have a similar p-wave velocity (c. 5.7 km.s-1) to the surrounding Precambrian basement. The refraction interpretations, together with some seismic reflection data, are used to constrain 3-dimensional modelling of the aeromagnetic anomalies in this area. The Peterborough experiment was located over a negative Bouguer gravity anomaly to the north of Peterborough. Refractor velocities typical of late Precambrian basement were detected at shallow depth (0.5 km.) across the gravity anomaly. The gravity data is modelled as a intra-basement intrusion. Measurements have been made of physical properties of samples of basement rocks from outcrop and from boreholes within the area. These are used in the interpretation of the potential field data. A pre-Carboniferous palaeogeological map has been compiled from borehole and geophysical data. A compilation of non-confidential borehole data is included as an appendix. The potential field and seismic studies show that two groups of granitic intrusions can be recognised, one group is more basic than the other.

Carbonatite magmas are alkali-bearing, ionic melts, analogous to synthetic carbonate melts. REE complexing suggests that carbonato-complexes dominate the melt, and these are more stable for the LREE than the HREE, leading to the extreme LREE/HREE enrichments characteristic of carbonatites. Crystal settling is viable in static magma chambers. however in thin carbonatite dykes as seen on San Vicente, turbulent flow velocities preclude vertical particle transfer against the flow of liquid, and favour fractionation by rapid crystal growth in a boundary layer close to the dyke margins, and particle accretion to the dyke walls. The Vale de Cavaleiros sovite on Fogo, displays mineralogical variation as a result of adcumulation, gravitationally controlled sedimentalogical sorting and post-lithification deformation. Minor and trace element variations are explained partly as a function of the heterogeneity of mineralogy, but require chemical modification of the magma by fractional crystallization and auto-metasomatizm. O18o and O13C of the carbonatites on San Vicente are decoupled, however calcite and dolomite may have equilibrated with seawater at temperatures close to 100°C. Fractionation of calcite and apatite from a carbonatite magma, produces cumulate microsovite and conjugate evolved carbonatite liquids which form the ferrocarbonatite suite. Trace element variations within the ferrocarbonatite suite indicate that the unstained ferrocarbonatites (unaltered 'quench' carbonatite) are heterogeneously metasomatized and mineralized by evolved (contaminated) orange ferrocarbonatites to produce brown-black ferrocarbonatites. The degree of dolomitization of the Camile dyke on San Vicente, relates to the activity of Ba2+ released during recrystallization of the original carbonate minerals. The arrays on plots of 87Sr/86Sr vs 143Nd/144Nd vs. 206Pb/204Pb for rocks from the Cape Verde Islands are interpreted as a two component mixing lines between recycled pelagic sediment and recycled altered MORB. The melilitites and nephelinites of the Malhada Pedra formation on Maio defines an array from CSn = 157. to CSN = 07 (in CSn-(Or+Ab) n-Din space), which may be interpreted as describing the loci of migration of the peritectic melt during the melting episode. There is little evidence for the genesis of carbonatite by ultra - fractionation of carbonated, alkalic, silicate magma. Neither major or trace element evidence can discriminate between liquid immiscibility, or direct partial melting of carbonated peridotite at pressures greater than 20kbar as mechanisms for generating carbonatite magmas.

About 100 km of seismic reflection sections down to 1 sec commercially produced for the NCB for prospects at Loughborough, East Leicestershire and Amington were interpreted. The Loughborough data show a NW-SE syncline, six principal faults, and thin Coal Measures. The East Leicestershire data show considerable undulations up to 200 m and an E-W fault. Variable Coal Measures 75--175 m thick existed throughout the area. Eleven coal seams up to 3.5m thick reflects the economic potential of the area. Three diorite horizons and Mountsorrel granite work type are encountered. The Amington data show undulations which increase in complexity downwards, five faults with throws up to 60 m, and relatively thin coal seams. Raw gravity data for the 3000 km2 between 41--46E and 29--35N were checked, contoured and then gridded at 0.5 km to produce 101 x 121 array which was used for processing. Ten profiles were selected for spectrum analyses and showed that the main gravity data features could be explained in terms of structures at depths of 1.3, 3.5, 6.6, 9.5 and 17.3km. Wavelength filtered, upward and residual maps indicate a NW-SE regional trend and suggest various densities for the Charnian rocks, and northern extension of the Warwickshire Coalfield. The downward continuation estimated the depth to the different gravitating sources. The second derivative estimated the width of the different structures. The lateral density variations using polynomials produced a new Bouguer map on which most of the anomalies were shifted and matched the known geology. A 2D-model along a 50 km profile shows a faulted layered basement which includes intrusives overlain by folded and faulted sediments. The aeromagnetic map of the area was digitised at 1 km intervals and used to produce a reduced to the pole map which was processed. The main feature of the magnetic basement shown by the maps is a NW-SE trough with its deepest part around Ashby. The spectral analysis, filtering and continuation techniques confirmed the gravity results and suggested that the diorites in Charnwood and South Leicestershire are more extensive than the outcrops indicate. The second derivative was useful for outlining the shapes of the magnetic sources. The Mountsorrel granite appears to extend considerably to the east, north and west by about 15 km at an estimated 2D-model depth of 1.3 km in the NW.

Five phases of deformation within six fault-bounded structural blocks are recognised in the Palaeozoic rocks of central Notre Dame Bay. The Lukes Arm-Sops Head (LASH) Fault, or Red Indian Line is a major sigmoidally shaped terrane-bounding fault in the Dunnage Terrane. The LASH Fault controlled deformation, which was protracted, progressive and heterogeneous. D0 involves wet-sediment deformation, is non-time specific and occurred during D1. D1 pre-Acadian structures include NW-SE to NNW-SSE upright macroscopic F1 folds, and local recumbent F1 folds related to SSV directed D1 thrusting. D1 structures were controlled by the restraining bend shape of the LASH Fault from mid-Ordovician time onwards. F2 Acadian folds trend NE-SW are transected by the anastomosing S2 cleavage, are upright, tight to isoclinal and trend NE-SW. F2 fold transection is related to the position of the fold with respect to the LASH Fault. S2 changes from a predominantly clockwise transection of F2 (<10°) to the south, through axial planar, to anticlockwise transected northwards towards the LASH Fault. D2 is interpreted to have formed under a N-S sinistral shear couple from early-Silurian time, with local dextral shear along E-W to ESE-WNW trending areas of the LASH Fault. F3 late-Acadian folds were associated with strike-slip faults, have a local axial planar S3 crenulation cleavage, and deform D1 and D2 structures about upright NE-SW axes. D2 to D3 transition was progressive in the late-Silurian, under widespread E-W dextral shear. D3 and older structures occur in rocks as old as Wenlock and are crosscut by Upper Ludlow dykes. Post-Acadian F4 folds are Devonian and are associated with kink bands that refold earlier structures about NW-SE axes. D5 involves complex crosscutting joints, faults, veins and shear arrays and ranges from Devonian to Jurassic.

(i) This study uses mineral magnetic parameters (e.g. magnetic susceptibility, saturation isothermal remanence and anhysteretic remanence) and lithostratigraphical description to correlate a suite of sed~ent cores sampled from a closely-meshed grid network covering the whole of a small upland lake basin in Gwynedd, North Waleso (ii) The core correlations are used to define three horizons within the lake sediment deposits: (1) the Loch Lomond Stadia I Clay - early Flandrian clay-mud contact, at the base of the Post-glacial sediment sequence. (2) A peak in magnetic susceptibility which is identified as a concentration of paramagnetic iron and manganese which are tentatively assumed to have entered the lake basin in solution. (3) A peak in magnetic susceptibility and magnetic remanence in the uppermost sections of the cores which is identified as a concentration of ferrimagnetic material. The correlated horizons define three contiguous layers of sediment within the lake basin. Contour maps of the thicknesses of the sediment layers are constructed and USed to estimate the volume of each layer. (iii) Radiometric dating (C-14 and Cs-137) is used to estimate the ages of the boundaries of the sediment layers, thereby enabling rates of total sediment influx and sediment yield per unit area of the terrestrial part of the watershed to be obtained. Sediment yield estimates of 36 kg/ha/yr, 1~6 kg/ha/yr and 263-326 kg/ha/yr were obtained for the periods (represent1ng the duration of the deposition times of the sediment layers) 10 400- 800 bp; 800 bp-AD 1951 and AD 1951-1977, respectively. (iv) Pollen analyses are used together with documented regional and local ?nvironmental history to provide a ccmtext for the eva'Lua.tdon of the est- ~mated rates of sediment yield. During the majority of the 10 400-800 bp period the watershed is likely to have been occupied by relatively undisturbed deciduous woodland, and during the 800 bp-AD 1951 period by grazed and possibly burnt-over moorland. The sediment yield for the AD 1951-1977 period is ascribed predominantly to soil erosion resulting from a forest :ire within the watershed in AD 1951. Comparison of the estimates of sed- ~~ent yield with those obtained using conventional geomorphological techn1ques in watersheds under comparable forms of land-use suggests some accordance. (v) The core correlations and sediment thickness contour maps depict the history of the spatial pattern of sediment accumulation within the lake baSin. The lake basin appears to have infilled asymmetrj.cal1y, with the f?CUS of sediment accumulation having moved progressively northwards with t1me. Marked variations in rates of sediment accumulation over short dist~ ces (tens of metres) are also evident. In contrast to the results ob- ~a~ed from other lake basins, the currently deepest part of the lake bas- 1n 1S virtually devoid of sediment. The observed patterns of sediment accumulation are tentatively assumed to be a response to wind stress.

Magnetic and geochemical properties of five late Holocene peat and loessial soil sections in SE Iceland have been studied. The aims were (i ) to establish a reliable tephrochronology for the sections, using geochemical analysis of tephra shards; (ii) to establish high resolution bulk sample records of mineral magnetic and geochemical properties; (iii) to test the ability of magnetic properties to correlate distal tephra layers; (iv) to assess the possible environmental significance of mineral magnetic time series and (v) to use records of loessial accumulation rates to assess human impact on the environment in SE Iceland. The work has been successful in establishing a reliable tephrostratigraphy for the studied sections. This was achieved using electron microprobe analyses (EMPA) of the geochemistry of tephra shards concentrated using a density separation technique. The results of detailed mineral magnetic measurements suggest that the main magnetic minerals in the tephras are probably ferrimagnetic minerals (e.g. magnetite) and canted anti ferromagnetic minerals (e.g. haematite), with abundant paramagnetic material also present. Cross plots of MrJMs vs. (Bo)c.l(Bo)c and Xfd% vs. XARM/SIRM indicate the main magnetic grain size in the tephras are pseudo-single domain (PSD) and multi domain (MD). The results indicate that the 'fingerprint' of both hysteresis and Curie temperature properties are insufficiently consistent within the tephra layers to provide an obviously sound basis for correlating tephra layers from one profile to the next. However, initial correlation of tephra layers was achieved, using all the measured magnetic parameters, by the use of the multivariate statistical measures of Similarity Coefficient (SC) and Euclidean Distance (ED). Tephra layer correlation was performed first time here using magnetic measurements. This demonstrates that magnetic techniques can potentially assist in the identification and correlation of distal tephra. Iceland offers a unique opportunity to study the role of humans in land degradation processes because of the late and well documented literature of settlements 1100 years ago. In this context, higher sedimentation rates in the upper part of two soils and one peat profile are interpreted as reflecting the effect of Landnam on deforestation and soil degradation.

This study is concerned with the distribution, age, chemistry and geotectonic setting of the 21 granites in Shetland. The granites range in size from a few metres to 8 Km across and in age from Pre-Moine to early Carboniferous. The granites occupy a unique position in being placed approximately half way between Scotland and Norway Caledonides and about 320 Km east of southeast coast of Greenland, if Greenland is restored to its pre-continental Drift position as predicted by the best fit of Greenland and Europe. The Shetland granites have been subdivided into 5 groups according to their location to the east or west of Walls Boundary Fault ( a continuation of Great Glen Fault), mineralogical content and their occurrence in situ as granitic pebbles, as follows; Granites to the east of WBF include (1) hornblende-bearing granites (2) hornblende-free granites and (3) Granitic pebbles a) the Rova Head conglomerate & b) the Funzie conglomerate. Granites to the west of WBF include (4) Ronas Hill granite and its satellites and (5) Sandsting & Bixter granites which are grouped together because of their proximity and because the Bixter granite seems to be the acidic end product of the Sandsting granitoid. A detailed petrological and geochemical study has been made of each group. Samples from each occurrence have been examined for mineralogical content and analysed by a variety of methods for major and trace elements (XRF, INAA, RNAA). This infonnation has been used to derive the mechanisms which may have been responsible for the observed differentiation trends. To the east of the Walls Boundary Fault (WBF) are hornblende-bearing granites, hornblende-free granites and Funzie and Rova Head granitic pebbles. The hornblende-bearing granites are truncated by the WBF and characterized by high Sr and Ba values. The Spiggie granite within this hornblende-bearing group also contains considerable amounts of primary epidote. The hornblende-free granites are a miscellaneous group of granitic rocks ranging between two-mica granites and garnet bearing granites, albite keratophyre and trondhjemite dykes. To the west of WBF are Ronas Hill granophyre and its satellites (with drusy cavities containing crystals of stilpnomelane, quartz and epidote), Sandsting complex and Bixter granites. Major element modelling suggests that a plagioclase, biotite and hornblende fractionation process is appropriate for hornblende-bearing granites and Sandsting granite. Graphical and trace element modelling do not conflict with this too, but rare earth element modelling requires extraction of a quartz-feldspar phase. The classification of Shetland granitoids on the basis of ages and petrology has not been previously attempted. According to the Read classification, the Caledonian hornblende-free granites could be classified as Pretectonic and Syntectonic intrusions equivalent to his older granites. In contrast to the hornblende-bearing granites to the east ofWBF (Graven, Brae complexes and the Spiggie granite), the Graven and Brae appear to belong to the Appinite Suite while the Spiggie granite seems to be a forceful Newer Granite. According to Read the forceful Newer Granites were emplaced just after the Appinite Suite but in Shetland the Graven and Spiggie granites appear to be the same age (400 Ma). The Ronas Hill granite and its satellites, Sandsting and Bixter granites to the west of WBF are Upper Devonian granites and equivalent to the permitted last intrusions of Read, but do not appear to be of the caulderon or ring complex type typical of those in Scotland In terms of ages the hornblende-bearing granites have given K-Ar ages of about 400-430 Ma. The Ronas Hill granite and its satellites, Sandsting complex and Bixter granite give K-Ar ages about 360 Ma. Consideration of typology indicates that the hornblende-bearing granites, Ronas Hill & its satellites, the Sandsting complex and Bixter granite are I-Caledonian type whereas some of the hornblende-free granites are close to S-type. On the De La Roche classification system the Shetland granites are high-K calc-alkaline (except trondhjemite dykes and keratophyre in the hornblende-free granites group) and mostly plotted in the shoshonitic trend The geochemical comparison of the the post Devonian Ronas Hill granite and its satellites on the one hand and the closely associated late Devonian Shetland volcanics on the other, reveals that there is no link between them. A comparison of the Shetland and Scottish granitoids in terms of major, trace and rare earth elements reveals very close similarities between these two Caledonian regions. Both show the same high K-calc-alkaline (peccerillo & Taylor, 1976), the high alumina and alkali (Kuno, 1966) character, also show a general alkali-calcic character (Peacock, 1930) and they also have very good correlation in terms of some trace element variation diagrams such as Sr and Ba

Sediments of Cretaceous age in N.W. Scotland outcrop in small, often isolated exposures throughout the Inner Hebrides and Morvern, and have been dealt with cursorily in most previous work on the Cretaceous rocks of Britain. The aims of this study were (i)to propose a formal integrated stratigraphic scheme for the Cretaceous strata of N.W. Scotland and (ii)to model the development of the Inner Hebrides Basin (where these strata outcrop) during the Cretaceous Period. Detailed field observations, macro- and micropalaeontology (including palynology), sedimentology and structural data were integrated in an attempt to achieve these aims. Fieldwork was conducted in Morvern (Argy 11) and the Inner Hebridean islands' of lYlul], Eigg and Skye. In the proposed lithostratigraphy the Mo~ern Greensand becomes the M.orvern Greensand Formation, of which the former "Lochaline Glass Sand" or "Loch Aline White Sandstone" becomes the Lochaline White Sandstone Member. The overlying silicified chalk, outcropping in Morvern and Mull, becomes the Gribun Chalk Formation. Dark grey micritic limestone, previously undifferentiated from the silicified chalk, becomes the Strathaird Limestone Formation of which there are two clastic members: the basal Laig Gorge Sandstone IVIember and the Clach Alasdair Conglomerate lVlember. The "Upper Estua.rine Series" of Judd (1878), becomes the Beinn Iadain Mudstone Member of which there is a coarser clastic member: the Feorlin Sandstone ~ember. r]:lhese Formations comprise the Inner Hebrides Group. Biostratigraphic evidence (based primarily on dinoflagellate cysts and foraminifera) indicates a latest-Albian to IVlid-Cenomanian age for the Morvern Greensand Formation; a Late Cenomanian age for the Gribun Chalk Formation, and an Early - IVliddle Turonian age for the Str2thaird Limestone Formation. The most refined biostratigraphical range for the Beinn Iadain lVIudstone F'ormation was Albian to Palaeocene. The IVlid-Late Cretaceous development of the Inner Hebrides Basin includes two periods of major transgression, the first of which began in the latest Albian and continued through the Early Cenomanian with the deposition of the marginal clastic facies of the Morvern Greensand Formation. A minor period of regression preceded the onset of carbonate deposition in the Late Cenomanian, recorded in the Gribun Chalk Formation. A second major transgressive episode followed the silicification, uplift and erosion of the Gribun Chalk, and reflects rapid deepening of the basin during the Early to Middle Turonian, poorly sorted clastic sediments (the Laig Gorge Sandstone Member) being overlain by biomicritic limestones intercalated with debris flows (the Strathaird Limestone Formation). These two major transgressive episodes are marked by the deposition of similar lithofacies throughout N.W. Europe (although no in situ deposits of Turonian age are found in Northern Ireland). Some feature; of the Cretaceous sediments of N.W. Scotland are found elsewhere, for instance the high abundance of calcispheres and organic-rich nature of the Strathaird Limestone are typical features of the Early Turonian transgressive episode in the AngloParis Basin. However, in the Inner Hebrides Basin, these eustatic changes in sea level were imposed upon local tectonic movements as indicated Qy the weathering and erosion of the silicified Gribun Chalk prior to the deposition of the Strathaird Limestone Formation, and the debris flow events, apparently related to synsedimentary movements along the Camasunary Fault, recorded within that Formation.