Content and licensingview original scan buy a printed copy

Geology of the Whithorn, Kirkcowan and Wigtown district. Memoir for 1:50 000 geological sheets 2, 4W and 4E (Scotland)

Bibliographical reference: Barnes, R P. 2008. Geology of the Whithorn, Kirkcowan and Wigtown district. Memoir of the British Geological Survey, sheets 2, 4W and 4E (Scotland).

British Geological Survey. Keyworth, Nottingham: British Geological Survey 2008. First published 2008. ISBN 978 0 85272 583 2. © NERC copyright 2008.

The grid used on the figures is the National Grid taken from the Ordnance Survey map. Topography is based on material from Ordnance Survey 1:50 000 scale maps. © Crown copyright reserved. Ordnance Survey licence number 100017897/2008

(Front cover) Cover photograph: Pibble Mine engine house situated on the north-west slope of Pibble Hill, 5 km east-north-east of Creetown [NX 5255 6068] (Photographer T S Bain; P001527).

(Rear cover)

Acknowledgements

The author and contributors extend their thanks and appreciation to numerous colleagues both in BGS and academia for assistance in the field and for access to unpublished data. Structural and stratigraphical interpretations benefited from the enthusiastic support of Dr P Stone, the late Dr B C Lintern, and Dr T B Anderson (Queen's University, Belfast) along with numerous participants on Southern Uplands field workshops over many years. Work on the intrusive rocks was inspired by the late Dr N M S Rock. Dr D Loydell (University of Portsmouth) provided additional biostratigraphical information.

Work was initially carried out under the programme management of Mr D H Land and Dr D J Fettes, and was completed under Dr M Smith.

Text was edited by Drs P Stone and E A Pickett. Series editor is A A Jackson: pagesetting by A Hill: cartography by J Bain.

Notes

Foreword

Geology underpins a wide range of activities vital to the creation of wealth, particularly in relation to the exploration for and exploitation of resources. It is also important that we have the best possible understanding of the geology of the United Kingdom if we are to maintain our quality of life, whether through the identification of potential hazards prior to development, or by helping to ameliorate the problems caused by earlier developments. The strategic core programme of the British Geological Survey is funded by the Natural Environment Research Council (NERC) to improve our understanding of the geology of the UK through a national programme of geoscience surveying, data collection, interpretation, publication and archiving.

The Whithorn, Kirkcowan and Wigtown district lies in the Dumfries and Galloway Region of south-west Scotland. The highest ground is around Cairnsmore of Fleet (711 m) in the north-east with lower ground to the west and on The Machars. The small market town of Newton Stewart is the main centre of population, with smaller centres such as Kirkcowan, Wigtown, Whithorn and Creetown serving a predominantly rural hinterland and also supporting an expanding tourism industry. Agriculture and forestry are the principal land uses, although in historical times the exploitation of the Galloway base-metal mining field, most of which lies within the district described, gave a local industrial base.

The extensive coastal sections provide excellent exposure through the late Ordovician to middle Silurian, southern sector of the Southern Uplands accretionary thrust belt. New data gathered there have proved crucial to the regional understanding of this important part of Britain's geological framework. A combination of biostratigraphy, petrography, geochemistry, sedimentology and structural geology has been applied in detail, complemented at a more regional scale by assessment of geophysical datasets. The results are described in this memoir, which also deals with the major, late Caledonian granitoid pluton of Cairnsmore of Fleet. Together with several smaller igneous bodies and an associated dyke swarm, this pluton provides critical geochemical and geochronological evidence for the igneous processes operating in Siluro-Devonian times across Scotland and northern England. More recent geological phenomena are also addressed, with extensive till, glaciofluvial sand and gravel and raised beach and estuarine deposits all contributing evidence for Quaternary glacial and interglacial events.

This memoir is the first Geological Survey publication specifically describing the Kirkcowan and Wigtown districts since the original explanation was published in 1878, followed by a regional Southern Uplands summary in 1899; the opportunity has also been taken to integrate a description of the adjacent Whithorn district, last published as a separate memoir in 1989. The 1:50 000 Series maps for the three districts have been published as separate sheets of the geological map of Scotland: Sheet 2 Whithorn, Solid and Drift 1987; Sheet 4W Kirkcowan, Solid 1992, Drift 1982; Sheet 4E Wigtown, Solid 1992, Drift 1981.

John N Ludden, PhD, Executive Director, British Geological Survey, Kingsley Dunham Centre, Keyworth, Nottingham. NG12 5GG

Geology of the Whithorn, Kirkcowan and Wigtown district

The district described in this memoir is covered by Sheet 2 Whithorn, Sheet 4W Kirkcowan and Sheet 4E Wigtown of the 1:50 000 geological map of Scotland. The extensive coastline includes the large embayments of Luce Bay and Wigtown Bay. Between these, the generally low-lying Whithorn–Wigtown peninsula, known as The Machars, rises towards the north-east into the upland area around Cairnsmore of Fleet. Much of the district is devoted to sheep and cattle farming with large coniferous forestry plantations on the higher ground.

Geologically the district forms part of the extensive outcrop of Lower Palaeozoic strata comprising the Southern Uplands terrane. During the Ordovician and Silurian periods, sediment transported by turbidity currents was deposited adjacent to the margin of Laurentia, a continental landmass on the northern side of the contemporary Iapetus Ocean. As the oceanic crust was subducted northwards beneath the Laurentian margin the accumulated sedimentary strata were progressively deformed until, eventually, Laurentia collided with the continental mass to the south of the ocean, Avalonia, in the late Silurian and early Devonian. A suite of igneous rocks was emplaced as dykes and small plutons during later stages of this process, culminating in emplacement of the large Cairnsmore of Fleet granite pluton, part of which crops out in the north-east of the district. Large-scale strike-slip tectonism was important at this time. Permian and Triassic rocks, mainly breccia and fluvial and aeolian red sandstone, are preserved offshore in Luce and Wigtown bays within extensional half grabens.

The district has a variable, patchy cover of till laid down during the Devensian glaciation. Over much of the low-lying areas, the till forms distinctive drumlins that rest on the glacially scoured surface of the Lower Palaeozoic rocks. Thin lacustrine deposits and peat occur locally in the hollows between the drumlins. More extensive spreads of peat blanket parts of the higher ground. Raised beaches are present along some parts of the coast.

Lead, zinc, copper, nickel and arsenic mineralisation occurs in the district and was worked at a number of sites in the 19th and early 20th centuries. Several mineralisation styles are present but the ages of the metallogenetic events are uncertain.

(Frontispiece) Refolded folds (D1 and D2) in the Kirkmaiden Formation, Back Bay (Photograph F I MacTaggart; P001123).

(Geological succession)

Chapter 1 Introduction

The District described in this memoir (Figure 1); (Figure 2) covers the area of 1:50 000 sheets 2 Isle of Whithorn, 4W Kirkcowan and 4E Wigtown. Much of the district consists of the Wigtown peninsula, known as The Machars, and comprises low-lying countryside in which extensive areas of peat 'moss' with shallow lochs surround rocky hills and ridges of subdued relief and rounded till drumlins. Here, the topography is generally less than 100 m above OD with the highest hills at about 200 m. Most of The Machars drains eastwards into Wigtown Bay or the River Cree, which separate it from the eastern part of the district where the relief is much higher. In this area, the ground rises to a watershed that takes an irregular course northwards from Ben John and Cairnharrow (456 m), via Cambret Hill, Pibble Hill and Craig Hill at about 350 m, up onto Cairnsmore of Fleet which, at 711 m, is the highest point in the district. Land use is almost entirely agricultural with extensive forestry plantations in the north-east of the district. Base metal vein mineralisation, occurring principally in a belt from Newton Stewart to Pibble Hill, was worked intermittently for about 200 years from the middle of the 18th century, although all of the mines are long since abandoned.

Geological setting

Geologically the district (Figure 1); (Table 1) is part of the extensive outcrop of Lower Palaeozoic strata which forms the Southern Uplands terrane (Figure 3). Bounded to the north by the Southern Upland Fault and to the south by unconformably overlying Devonian and Carboniferous strata, the terrane is dominated by a sandstone-rich turbidite sequence overlying a condensed, fossiliferous black mudstone and chert (with some basaltic lava) sequence. The outcrop is divided by north-east-trending, strike-parallel, faults into 1 to 5 km wide 'tracts' (Figure 1) in which the stratigraphy is usually steep to vertical, younging northwards from the black mudstone, preserved in a narrow discontinuous zone at the southern tract margin, into the turbidite sequence. Biostratigraphical evidence shows that the age of the latter in any one tract is within one graptolite Biozone, usually corresponding with the youngest biozone represented in the underlying mudstone (see Chapter 4). Hence, despite their volumetric predominance, the turbidites in each tract occupy a relatively small time interval (<1.5 Ma) compared with the underlying Moffat Shale Group, which may represent up to 25 Ma in its southernmost outcrops. The biostratigraphical data also show that, in contrast to the typical northward younging within each tract, the age of onset of turbidite deposition becomes progressively younger in successive tracts southwards. Igneous rocks were emplaced as minor intrusions during the later phases of deformation of the sedimentary rocks and as large, post-tectonic granitoid plutons (see Chapter 7). Low-grade regional metamorphism also occured during the later phases of deformation, overprinted by thermal alteration adjacent to intrusions of intermediate to large size.

The application of plate tectonic theory to the British Caledonides (e.g. Dewey, 1971) led to interpretation of the Ordovician and Silurian sedimentary rocks preserved in the Southern Uplands as representing oceanic and continental margin deposits formed near the northern margin of the Iapetus Ocean. The tectonostratigraphical configuration of the Southern Uplands, strongly suggestive of deformation relatively soon after deposition, became the classical evidence for the fore-arc accretionary prism model (e.g. McKerrow et al., 1977; Leggett et al., 1979). Although still favoured by many authors (e.g. Needham, 1993), this model cannot easily account for all of the observed features in the Southern Uplands terrane, particularly some palaeocurrent and provenance data. A number of alternative tectonic models have therefore been proposed (e.g. Murphy and Hutton 1986; Stone et al., 1986, 1987; Morris, 1987). These authors concur with the previous model in terms of the finite structure but they offer different interpretations of the tectonic setting and, to a limited extent, timing of deformation. The continuing debate thus hinges on detailed sedimentology, stratigraphy, petrology and geochemistry. For example, basic volcanic rocks at the base of the sequence in northern parts of the Southern Uplands were originally interpreted as alkaline basalt and ocean floor tholeiite, consistent with oceanic basement (Lambert et al., 1981). More recent work (Barnes et al., 1995b; Phillips et al., 1995a) suggests a more complex pattern. Some components have affinities with island-arc volcanic suites such as those which crop out extensively in the more deeply eroded equivalent sequence in Newfoundland (Colman-Sadd et al., 1992), in what was the continuation of the Southern Uplands terrane prior to the opening of the Atlantic Ocean.

The post-Caledonian rock-record in the district is limited to Permian and Triassic rocks preserved offshore in fault-bounded basins, and Palaeogene dykes. Extensive early Permian breccia deposits are preserved elsewhere in south-west Scotland and were deposited in association with active faults. Fault movement of this age has not been proved in the district but is assumed to have occurred. Offshore boreholes establish that regional subsidence in the late Permian and Triassic led to the deposition of fluviatile and aeolian sandstones and mudflat or playa lake mudstone and evaporites.

Chapter 2 Applied geology

Planning and conservation

With a relatively small and stable population in the district, there is generally little development pressure for new building, with only minor housing, farm and light industrial units to be considered. The major construction work in the district in recent years has been associated with the various road re-alignment, bridge rebuilding and by-pass schemes forming part of the long-term plan to upgrade the A75 'Euroroute' from Dumfries to Stranraer. This work has greatly improved the quality of life in those towns, such as Newton Stewart, Gatehouse of Fleet and Glenluce, which previously straddled the trunk road. The track-bed of the former railway line between Dumfries and Stranraer traverses the north of the district and includes an attractive legacy of stone bridges and other engineering works.

Agriculture and forestry

Agriculture and forestry account for by far the largest proportion of land use in the district. Agriculture is extensive in the lower lying western and southern parts of the district. Where the soils are underlain by till or alluvium resting on Lower Palaeozoic rocks, the land is generally of good quality. It is used mostly for grazing cattle and sheep, with associated arable use. Thin peaty soil on rock, common between isolated drumlins in central parts of the district, gives rise to poor quality land only suitable for rough grazing. In the more elevated areas in the north and east of the district, soils are generally thin, acidic and poor in nutrients, forming moorland which mainly supports sheep farming. Forestry is long-established on the upland area formed by the thermally altered rocks west of the Cairnsmore of Fleet granite pluton and is expanding into other areas.

Ground stability and foundation conditions

Bedrock or till usually form good bearing ground in the Southern Uplands region. The main stability problems are likely to occur on steep slopes or within the alluvium or peat covered areas, where soft or otherwise unpredictable ground conditions may prevail. Site investigation by trenching will usually be sufficient to establish the actual conditions at any particular location. Underground mining (Chapter 10) is present around the western end of the Cairnsmore of Fleet granite, concentrated in a belt south-east of Newton Stewart, from Blackcraig [NX 447 645] to Pibble [NX 527 605] and at Penninghame [NX 387 697], with shafts and underground cavities which may cause difficulties locally.

Groundwater resources

Although groundwater is present in all rock types across the area, there are no major, regionally important, aquifers in the district and hence no large-scale exploitation of this resource. There is a large contrast between the mode of storage and transmission of groundwater in the metasedimentary rocks and intrusive igneous rocks that underlie most of the district and the granular superficial deposits found in the valleys.

Aquifer characteristics

The Ordovician and Silurian sedimentary rocks and the granitic intrusions are poorly permeable and rely entirely on fractures within the rock to store groundwater, so that many parts of the rock mass contain little water. Groundwater flow rates and paths also depend on discontinuities and, where flow does occur, it tends to be in restricted, heterogeneous zones such as fracture zones commonly associated with faults. The success of a water supply borehole therefore depends entirely on the number of voids intercepted below the water table. Where boreholes have been sited near fault zones, groundwater yields may reflect the enhanced permeability, but other boreholes located nearby may yield almost no groundwater.

Superficial deposits contrast with bedrock by containing groundwater within intergranular void space. The space available for water storage is called the effective porosity, with up to 20 per cent of the bulk volume of a sand body potentially comprising groundwater in storage. Movement of water within such intergranular space is more predictable than in fractured bedrock, but the heterogeneous nature of much of the alluvial and glacial deposits means that flow can be concentrated along the line of buried channels etc. Superficial deposits that contain amounts of groundwater usable for small communities are restricted to the main valleys, particularly the River Cree. Elsewhere, on higher ground, low-permeability till, moraine deposits and peat are important in helping to maintain stream flows in summer months by releasing small volumes of groundwater from storage. In the Cree estuary, shallow groundwater contributes to the existence of wetlands on the Moss of Cree.

Trial boreholes have been drilled near Newton Stewart, both to the north of the town and at Minnigaff, and pumping tests carried out. The Boreland Farm Borehole [NX 3912 6784] was drilled into mounded glacial meltwater deposits to a depth of 9.8 m. Below 3.6 m of sandy clay, coarse-grained sand and gravel is present to the base of the borehole. A very promising yield of 14 litres/second (1200 m3/day) was obtained for a drawdown in borehole water level of 1.2 m after 3 hours pumping. A similar result was found at Minnigaff [NX 4160 6545] where a borehole was drilled through late-glacial raised beach gravel and sand. These deposits are restricted to the Newton Stewart area, within the drainage system of the River Cree, but indicate the high values of hydraulic conductivity within small areas. Unfortunately, from a water resources viewpoint, much of the surface area of these deposits has been developed for housing in the Newton Stewart urban area, thereby reducing its potential for exploitation.

Large spreads of postglacial raised beach deposits are present in the lower reaches of the Cree and Palnure Burn valleys. The fine-grained nature of these deposits reduces the overall hydraulic conductivity and hence borehole yields. A significant proportion of silt within granular layers helps to reduce the effective porosity and many sandy layers become increasingly clayey with depth. A series of trial boreholes has been drilled into alluvium and underlying estuarine deposits in the valley of the Palnure Burn [NX 458 658]. Thin, shallow gravel beds overlie silty sand and laminated clay to 17 m below ground level (10 m below OD). At Palnure, borehole yields are in the range 1 to 3 l/s (80 to 430 m3/day). A narrow raised beach platform is present almost continuously along the coast between Creetown and Ravenshall Point and along much of the western side of Luce Bay. Its sediments contain small amounts of groundwater, but near Creetown any development potential has been reduced by the re-alignment of the main A75 trunk road.

Use of groundwater

The lack of any extensive, highly permeable aquifers in the district means that groundwater has not been exploited to any significant extent. The total number of water supply boreholes drilled into the bedrock is unknown, but several have been constructed in recent years, mainly for farm and domestic use. Little work has been done on pump testing or sample collection for chemical analysis from these sources, which generally yield less than 50 m3/day and are of potable quality.

Springs are common near exposures of bedrock, indicating active groundwater movement at shallow depths. Many of these sources have been captured for potable use, although a number of them are failing to meet increased consumption and some owners are turning to more reliable, deeper, borehole supplies.

Mains water supplies to small communities such as Creetown have, in recent years, proved vulnerable to prolonged periods of dry weather. This is because many sources of supply are simply surface water intakes close to the centres of population. No permanent public supply groundwater sources are yet in place, but there have been efforts in recent years to establish the feasibility of shallow groundwater as a source for public consumption.

Groundwater quality

Limited data are available for bedrock groundwaters, but quality is thought to be generally good. An exception may be the narrow bands of black shales near Newton Stewart where mineralised zones could introduce higher than normal concentrations of metals to groundwater. Bacteriological quality is excellent where local sources of pollution are avoided. Many shallow springs suffer from a certain amount of contamination from animal wastes, another reason for some users opting for a deeper supply. Apart from diffuse sources of pollution such as fertilisers, there are few other major threats to groundwater quality. The low population density of the district means that waste disposal sites are of small scale and generally located on either peat or till. There are no large industries that could pose a threat to local aquifers. One potential source of pollution is the mine waste tips in the Blackcraig [NX 447 645] and Pibble [NX 527 605] areas. Here, very high concentrations of copper, lead and other elements are present, which are leached into surface watercourses. Development of water supply boreholes sited in these areas should take account of the potential presence of these pollutants.

The glacial and Late-glacial gravels near Newton Stewart are vulnerable to pollution from surface sources and this is reflected in high concentrations of nitrate, derived from agricultural fertilisers and animal wastes. Analyses of groundwater from the alluvial and lacustrine deposits of the Palnure valley indicate that high concentrations of iron and manganese are present in groundwater from silty gravels. This is a common phenomenon in certain superficial deposits near the Galloway coast. Poor natural water quality, from limited knowledge, appears to be a feature of raised beach granular deposits in certain areas, but insufficient data are available to make general statements about the chemical characteristics of shallow groundwater in the area.

Groundwaters are, on the whole, weakly mineralised across the district within Silurian rocks and superficial deposits. They are, nevertheless, capable of resisting acidification from rainfall containing acidic pollution. Weakly buffered groundwaters in granitic rocks and some isolated superficial deposits may be prone to acidification.

As with groundwater, the chemistry of surface waters also shows a close relationship with the local bedrock/superficial geology. The pH has a negative correlation with the available Ca, Mg and total carbonate content of bedrock. It shows strongly acidic drainage (pH 5–6) over the granite areas that are poor in these factors, broadly neutral conditions (pH ~7) over the Gala Group rocks and mildly alkaline surface waters (pH ~8) over the carbonate-rich Hawick Group.

Flooding and land drainage

Local flooding may occur from time to time during periods of spate in low-lying areas adjacent to the main lochs and rivers. Most of the arable land is drained by field drains into local watercourses. The recent expansion in commercial forestry must have increased the run-off rate of rainwater, and therefore the risk of flooding, due to the extensive networks of drains and ditches dug before planting. Otherwise, the peaty soils in these upland areas form a natural reservoir that stores rainwater and releases it slowly, thus reducing the peak run-off during periods of high rainfall.

Mineral resources, quarrying and mining

Potentially there are very extensive hard rock resources within the district, with the granite pluton in the north-east, several smaller granodiorite to diorite intrusions and large areas underlain by variably metamorphosed Lower Palaeozoic sedimentary rocks. Three small- to medium-sized, hard rock aggregate quarries operate within the district. Two of these are in the northern part of the Gala Group: Barlockhart Quarry [NX 213 565], south of Glenluce, is in bedded sandstone of the Kilfillan Formation in the contact metamorphic aureole of the Glenluce diorite intrusion; Boreland Fell Quarry [NX 348 593], 2 km south-east of Kirkcowan, works massive sandstone of the Mindork Formation. Creetown Quarry [NX 480 565] (fomerly Glebe and Kirkmabreck Quarries), situated adjacent to the A75 on the coast 2 km south of Creetown, is in the Kirkmabreck granodiorite, a large dyke-like intrusion. Creetown is the lowest of several quarries along the length of the dyke; topographically higher workings include Silver Grey Quarry [NX 487 568] (Plate 1a) and Fell Quarry [NX 490 569]. Together with Bagbie Quarry [NX 489 549] in the Carsluith granodiorite dyke, these quarries took advantage of the well-developed jointing in the granodiorite to manufacture sets (Plate 1b), kerb stones and dimension stone for building and monumental purposes. All of these products are much in evidence in the surrounding towns and villages.

Various temporary and small-scale quarries, mostly in till or weathered bedrock, are operated from time to time by the forestry companies to obtain bulk material for unsurfaced roads. Deposits of sand and gravel are scattered throughout the district, some being worked on a small-scale for local usage, although those at Glenquicken [NX 510 596] and near Garrochar [NX 497 589] and [NX 500 558] have previously been worked on a larger scale.

Vein-hosted, base metal mineralisation (Chapter 10) has been worked in the Blackcraig and Pibble mines and at a number of small mines or trials in a zone around the western end of the Cairnsmore of Fleet granite. Fracture zones containing traces of iron (siderite) and copper (chalcopyrite and malachite) and numerous baryte veins occur in the coastal section from Burrow Head to Port Castle Bay [NX 426 358], in the south of the district. These have been explored locally but there is no record of exploitation.

Chapter 3 Geophysical interpretation

Regional gravity and aeromagnetic survey data have been used to investigate the concealed geology of the district. A pronounced gravity low occurs over the Cairnsmore of Fleet pluton in the north-eastern corner of the district, and can be used to interpret its sub-surface form. A second low over Luce Bay, in the south-west, is due to the Permo-Triassic rocks in the Stranraer basin and there is evidence of a further, smaller basin beneath Wigtown Bay. In addition to these features, there are gravity variations that appear to relate to changes in the depth and/or density of a concealed basement. The magnetic field in the region is dominated by the long wavelength anomaly due to a major, deep magnetic source. Near-surface magnetic disturbances are evident over the marginal phase of the Cairnsmore of Fleet pluton, particularly in the east, over minor, late Caledonian intrusions at Culvennan Fell and Glenluce, and over Palaeogene (Tertiary) dykes. Gravity and aeromagnetic data for the district are summarized in (Figure 4) and (Figure 5), respectively.

Rock physical properties

The results of density measurements on rocks from the western part of the Southern Uplands have been summarised by Kimbell (1991) and Floyd (1999). The estimated average saturated bulk density for the Lower Palaeozoic sedimentary rocks is 2.72 Mg/m3 . Local variations in the density of these rocks may occur because of varying relative proportions of quartz and the denser minerals (Floyd, 1999). Density determinations on specimens from the Cairnsmore of Fleet pluton (Parslow and Randall, 1973; Gardiner and Reynolds, 1937) indicate that there is a small but significant difference between the density of the inner biotite-muscovite granite (2.61 Mg/m3) and the outer biotite granite (2.63 Mg/m3). The diorite exposed at Culvennan Fell [NX 330 650] has an estimated density of 2.74 Mg/ m3 (Parslow and Randall, 1973). The density of near-surface Permian sandstone is approximately 2.3 Mg/ m3 (Bott and Masson Smith, 1960), but higher densities are likely at depth. For example, in the Silloth No. 1 borehole [NY 1231 5485] the Permian Penrith Sandstone lies at depths of 1 to 1.3 km and has an average density of 2.46 Mg/ m3 (Lee, 1989).

The available magnetic susceptibility measurements (Floyd and Trench, 1989) and observed magnetic field measurements indicate that the Lower Palaeozoic formations that crop out within the district are only weakly magnetic (susceptibilities typically less than 1 x 103 SI units). There is evidence of higher magnetisations over some of the intrusive rocks in the region. Modelling and analogy with similar rocks elsewhere (e.g. Robson and Green, 1980) suggest that the magnetisation of the late Caledonian intrusions is likely to be predominantly induced (i.e. in the direction of the Earth's present field), whereas Palaeogene intrusions typically have a strong remanent magnetisation which may be normal or reversed (more commonly the latter) depending on the polarity of the Earth's field at the time of intrusion.

Gravity surveys

Gravity data from the region are displayed as a contoured Bouguer gravity anomaly map in (Figure 4). This map is based on the results of gravity surveys conducted by BGS, Parslow and Randall (1973), Mansfield and Kennett (1963) and Bott (1964).

Local gravity features within the district are superimposed on the general increase in gravity values towards the Irish Sea that is observed across south-west Scotland and north-west England. The Caledonian Suture Seismic Profile (CSSP; Bott et al., 1985) provided seismic refraction data across this gravity feature (the profile is orientated in an east-north-east–west-south-west direction and passes between the Wigtown peninsula and the Isle of Man). Lewis's (1986) interpretation of the CSSP placed a 5 km thick, high velocity and high density zone (a 'transitional Moho') at the base of the crust beneath the Irish Sea and identified this as the cause of the regional gravity high; it was suggested that this layer may have evolved during the periods of extension which were responsible for basin formation in Carboniferous and Permo-Triassic times.

The north-eastern part of the Bouguer anomaly map of the district (Figure 4) is dominated by the low due to the Cairnsmore of Fleet pluton, which is approximately 0.1 Mg/ m3 less dense than the Lower Palaeozoic rocks it intrudes (see above). Parslow and Randall (1973) modelled the observed gravity anomalies as the signature of a large batholith extending to a depth of 11 km with contacts dipping outward at about 50°; they identified a region of relatively shallow granite extending for several kilometres to the west of the exposed western margin. Dawson et al. (1977) developed a three-dimensional model encompassing both the Cairnsmore of Fleet and Loch Doon plutons; the depths to the bases of these plutons were estimated at 12 km and 7 km, respectively, and their model indicated subsurface granite extensions both to the west of Fleet and east of Loch Doon.

A second major Bouguer anomaly low, which lies in the south-west corner of (Figure 4), is due to the Stranraer Permian Basin. The form of this basin has been modelled by Mansfield and Kennett (1963), Bott (1964) and Kimbell (in Stone, 1995). It is markedly asymmetrical, with a steeper eastern flank, probably indicating a half-graben structure. Kimbell (in Stone 1995) estimated that this basin extends to a depth of approximately 2 km, assuming basin-fill with an average density of 2.45 Mg/ m3 above a 2.72 Mg/ m3 Lower Palaeozoic basement. The Lower Palaeozoic outcrop on the Scares [NX 250 330] (Barnes et al., 1988) demonstrates the truncation of the basin in the south, and the gravity data suggest it terminates against a north-east-trending lineament. Kimbell (in Stone, 1995) traced this lineament farther to the south-west, and there is gravity evidence that it also extends north-eastwards across the Wigtown peninsula (feature G1 in (Figure 6); see below).

A local Bouguer anomaly low in Wigtown Bay appears to extend a short distance onshore around [NX 450 600], over an alluvium-covered area to the north of Wigtown Sands. This anomaly is not well resolved by the data shown in (Figure 4) as there are no gravity stations between the onshore data and some sea-bottom stations (Bott, 1964) close to the mouth of the bay. However, additional measurements by Parslow and Randall (1973) suggested the continuity along the bay of a short wavelength low with an amplitude of about 6 mGal. Parslow and Randall interpreted this anomaly in terms of a small basin of Permo-Triassic or Carboniferous age with a similar orientation to the neighbouring Stranraer Basin (Figure 6). The location of the basin may have been controlled by the north-west-trending fault which has been identified within the Lower Palaeozoic rocks to the north-west of Creetown [NX 475 590], approximately coinciding with geophysical lineament L1 ((Figure 6); cf. Floyd, 1999). The Wigtown Bay basin appears to extend across the extrapolation of lineament G1, although there is some indication of a change in strike or possibly a sinistral offset across this line. There is no direct evidence of the nature of the rocks contained in the Wigtown Bay Basin. If these are of Permian age, as seems most likely, they are likely to be up to about 0.4 Mg/ m3 less dense than the underlying Lower Palaeozoic rocks (Bott and Masson Smith, 1960). With such a density contrast, the observed anomaly corresponds to a minimum basin thickness of approximately 360 m. The presence of low density alluvium may reduce the thickness estimate somewhat, but it is unlikely that the entire anomaly can be explained in terms of recent deposits.

To the south of the Wigtown Bay basin, in the south-eastern corner of (Figure 4), Bouguer anomaly values decrease southwards. This is due to the southward thickening of the Carboniferous and younger sedimentary rock succession into the Solway Basin.

There is some evidence of gravity effects relating to variations in the density of the Lower Palaeozoic rocks immediately north of the district (Floyd, 1999). The most conspicuous of these are residual highs that lie approximately parallel to the regional strike in the vicinity of the northern margin of the Portpatrick Formation (G2 and G3) and within the Shinnel Formation (G4). These may indicate relatively dense units within the Lower Palaeozoic sequence. Within the district there is little evidence of density contrasts within the Lower Palaeozoic sequence, although the relatively sparse gravity station distribution makes such features difficult to identify.

Gravity values rise southwards along the Wigtown peninsula, with gravity lineament G1 (Figure 6) defining the line of maximum gravity gradient. Modelling suggests that only part of this trend can be explained by the influence of the Cairnsmore of Fleet pluton because of the distance from this body. There is no evidence in the exposed Lower Palaeozoic sequence of the significant, gradational change in composition that would be required to explain the observed feature. The most likely explanation involves variation in the depth to an underlying basement and/or lateral variations in basement density. Seismic refraction surveys have detected a relatively shallow (<4 km), high velocity (³6 km/s) refractor beneath the north-eastern part of the Southern Uplands, which has been interpreted as the top of a crystalline basement (Hall et al., 1983; Kamaliddin, 1991). There is only limited seismic refraction coverage in the vicinity of the district (Al Mansouri, 1986; Kamaliddin, 1991; results summarised by Floyd, 1999). High velocity basement was not detected in the north-western part of the district (down to the exploration depth of about 6 km) but the surveys do not extend to the Wigtown peninsula. It is therefore possible that the gravity gradient in this area is due to shallow, relatively dense basement to the south of G1, although further refraction experiments would be required to confirm this. The continuation of the gravity gradient immediately to the east of Wigtown Bay is not easily discerned because of the proximity of the Cairnsmore of Fleet pluton, but modelling suggests that the feature is still present (see below) and its probable extrapolation is clearly defined farther north-east, in the Thornhill district (McMillan, 2002).

Aeromagnetic data

An aeromagnetic survey of the region was flown in 1959. North–south flight lines were flown at 2 km spacing with orthogonal tie lines 10 km apart; the nominal sensor elevation was 305 m (1000 ft) above terrain. (Figure 5) shows the location of digitised data points and contoured, reduced-to-pole aeromagnetic anomalies. The digitisation procedure has been described by Smith and Royles (1989). Reduction to the pole centres magnetic anomalies over source bodies, provided magnetisation is in the direction of the Earth's present field.

Long wavelength effects

The aeromagnetic map is dominated by a long wavelength high with an axis orientated north-east–south-west along the geological strike (the 'Galloway high' of Powell, 1970). The horizontal gradient of the pseudogravity anomaly (calculated from the magnetic anomaly using Poisson's relation) can help to delineate deep-seated magnetic sources (Cordell and Grauch, 1985) and has been used by Kimbell and Stone (1995, fig. 2b) to investigate correlations between the source of the Galloway high and structures seen at surface. There is a very close correspondence between the course followed by the axis of the pseudogravity anomaly, which defines the 'centre of gravity' of the magnetic source, and that of the Orlock Bridge Fault, the latter lying about 2 km to the north of the former (Figure 6). This correspondence applies for more than 100 km along the arcuate trace of the fault and thus appears unlikely to be circumstantial. The anomaly cannot however be modelled as a 'dyke-like' source in the fault zone but requires a belt of magnetic rocks several tens of kilometres wide. Although it is theoretically possible for a long wavelength anomaly to arise because of gradual variations in the depth to a relatively shallow magnetic basement, such an explanation is unlikely in this case. This is because the magnetic basement would be shallowest where it should be punctured by the largely non-magnetic Cairnsmore of Fleet pluton, and there is not the pronounced magnetic effect that would be expected with such a configuration (Powell, 1970). There is a slight magnetic low superimposed on the regional high over the Cairnsmore of Fleet pluton which is reflected in the disruption of the axis of the pseudogravity anomaly in the vicinity of this body (Kimbell and Stone, 1995, fig 2b); this may be due to limited interaction between the base of the granite and the top of the magnetic zone at depths of around 10 km (Powell, 1970).

Kimbell and Stone (1995) proposed that the deep magnetic source beneath the Southern Uplands is a fragment of an Ordovician arc terrane or Precambrian crystalline basement. Its boundaries dip to the north, but the northern one is inferred to have a steeper dip than that to the south. The southern boundary may be the Iapetus suture (as identified by seismic reflection surveys, e.g. Soper et al., 1992). The basement structure to the north occurs in the vicinity of, and was correlated with, the Moniaive Shear Zone (see Chapter 7) in the central part of the Southern Uplands. It was considered to predate the shear zone and to have controlled its development. Support for a change in basement composition in this vicinity is found in the contrast in the geochemistry and isotopic characteristics of granite plutons, the group to the south of this line showing similarities to the granitoids of the Lake District (Thirlwall, 1988; Stone et al., 1997). This, together with evidence of magnetic (?Precambrian) basement beneath the Lake District, led Kimbell and Stone (1995) to suggest that the magnetic block may have rifted from the Avalonian continent during its northward drift and subsequently become trapped in the hanging wall of the Iapetus suture.

Local magnetic features

Two weak, north-west-trending, narrow magnetic highs cross the district (M1 and M2 in (Figure 6)). They can be traced most readily in residual maps although their continuity through the central part of the area is difficult to define with the existing data. These anomalies are probably due to normally magnetised Palaeogene dykes of the Arran swarm. Further subparallel magnetic anomalies have been identified from detailed aeromagnetic survey data to the north-east of the district (Floyd, 1999). The only indication of a reversely magnetised Palaeogene dyke in the area is negative anomaly M3 to the south-west. A local magnetic closure (M4) suggests the presence of a Palaeogene plug or sill where this dyke intersects the lineament forming the southern margin of the Stranraer Basin and there is also an offset in the course of the dyke at this feature (Stone, 1995).

Local magnetic anomalies at Glenluce and Culvennan Fell (M5 and M6 respectively in (Figure 6)) are due to late Caledonian intermediate intrusions, but there is no discernible aeromagnetic anomaly over intrusions of similar age near Creetown [NX 480 565]. Ground traverses at Culvennan Fell indicated that the south-east part of the intrusion is more magnetic than the remainder, and possibly represents a different intrusive phase (Parker et al., 1981). A small positive residual gravity feature indicates that this body is slightly more dense than the Lower Palaeozoic country rocks. A number of local magnetic disturbances are evident in ground survey data from the region around the intrusion; the source of these is not known, but Parker et al. (1981) suggested the possibilities of dykes or concentrations of magnetic minerals in the country rocks.

An arcuate magnetic high around the eastern margin of the Cairnsmore of Fleet pluton (M7 in (Figure 6)) correlates with the outcrop of the outer biotite granite, indicating that this is more magnetic than the central muscovite-biotite granite and the surrounding country rock. On the western side of the pluton there are two local magnetic highs (M8 and M9 in (Figure 6)), the sources of which are not exposed. These are not well displayed by the contour map (Figure 5), because they coincide with the high in the long wavelength anomaly, but they are resolved by residual images, which indicate that they are bounded to the west by lineament L1 and have a considerably shallower source than the anomaly upon which they are superimposed. The most likely explanation is that these are due to concealed late Caledonian intrusive rocks. Feature M8 overlies Parslow and Randall's (1973) shallow projection of the Cairnsmore of Fleet pluton, suggesting a possible common source (see below). Feature M9 lies on the flank of the Cairnsmore of Fleet intrusion and is somewhat elongated along the trend of the Orlock Bridge Fault. It is possible that the source is a concealed magnetic intrusion whose location is controlled by this fault, in common with the Culvennan Fell and Glenluce bodies, and that postulated to occur at Sandhead [NX 100 500] (M10 in (Figure 6); Kimbell and Stone, 1992). A further concealed magnetic intrusion may be responsible for weak magnetic high M11 to the north of the district (Floyd, 1999).

Using detailed airborne survey data, zones of magnetic disturbance (M12 and M13) have been identified along the Morrach Bay (Fardingmullach) and Leadhills faults respectively immediately to the south-west of the Loch Doon pluton (Floyd, 1999). These may be due to secondary magnetic minerals (possibly pyrrhotite) formed within the fault zones. There is no evidence that the anomalies extend into the district; this could be due to the lower resolution aeromagnetic coverage, although the fact that shallower structural levels are probably exposed west of L1 (Stone et al., 1995) may also be significant.

Geophysical modelling

Simple geophysical models (Figure 7); (Figure 8) are considered along two profiles (A and B on (Figure 6)) that cross the Cairnsmore of Fleet pluton. Two and a half dimensional modelling methods have been used, in which the model profiles are assumed to bisect bodies of finite strike length and constant cross section along strike.

Profile A (Figure 7) crosses both the Loch Doon and Cairnsmore of Fleet plutons although, since the former is cut obliquely to its elongated north–south axis, it is not modelled accurately by two and a half dimensional methods. Only the central granitic core of the Loch Doon body is modelled, as the outer tonalite has a similar density to the country rocks. The Cairnsmore of Fleet pluton is interpreted to be steep sided and extending to depths of about 12 km. The base of the Loch Doon intrusion is shown at a shallower depth (as suggested by Dawson et al., 1977), although the model is in fact insensitive to this depth. The biotite granite and muscovite-biotite granite are included as separate components in the model of the Cairnsmore of Fleet pluton. A small magnetic anomaly coincides with the southern outcrop of biotite granite; modelling suggests that a susceptibility contrast of about 1 3 103 SI units (induced magnetisation) is required to explain the observed anomaly amplitude. The magnetic effect of the northern biotite granite is barely discernible and implies an even smaller magnetisation contrast.

The rise in gravity values towards the southern end of profile A has been modelled by introducing a relatively dense concealed unit. An alternative explanation would involve topography on a continuous basement surface, stepping up to the south with a similar amplitude to the body thickness shown (the density chosen lies within the range likely for crystalline basement). The depth of this interface is not accurately constrained but it appears probable that it lies at a shallower crustal level than the underlying magnetic basement. If relatively dense basement is present throughout the area it may be intruded by the lower parts of the major plutons. This would have the effect of reducing their estimated depth extent (assuming that their density does not vary with depth).

A combined gravity and magnetic model for profile B is shown in (Figure 8). It should be noted that, because of the flight line orientation (Figure 5), the magnetic field is effectively only sampled at 2 km intervals, so only a very simple model is justified. The long wavelength magnetic variation along this line (shown as an estimated 'background' effect in (Figure 8)) probably includes a component due to interaction at depth between the granite and a deep magnetic source. The gravity model supports Parslow and Randall's (1973) inference of a south-westward subsurface extension of the pluton, and it has been assumed that this concealed component is of biotite granite composition. This is reasonable in view of the general presence of this phase towards the margins of the pluton and also in view of the magnetic anomaly (M8 in (Figure 6)), which can be explained in terms of such a configuration. There may therefore be a local cupola of relatively magnetic biotite granite reaching near to the ground surface, as indicated in (Figure 8), although an explanation for the magnetic anomaly in terms of an overlying body of different composition (e.g. magnetic mineralisation in the country rocks) is possible. The small magnetic anomaly on the eastern side of the pluton can be reproduced assuming a source coinciding with the mapped biotite granite. An increase in anomaly amplitude to the south of the profile suggests that this phase may become more magnetic in this direction. It is possible that higher magnetisations within metamorphosed country rocks adjacent to the pluton contribute to the magnetic anomaly, although this effect cannot be quantified using currently available data (Kafafy and Tarling, 1985; McMillan, 2002). The calculated gravity field is lower than that observed over the north-eastern part of the granite outcrop ((Figure 8)); this could be because of a north-eastward increase in the bulk density of the biotite muscovite granite (of about 0.01 Mg/m3) or a thinning of the pluton in this direction, although the modelling assumptions (in particular the shape of the background field and properties of the concealed basement) also influence the interpretation.

Local surveys

In addition to the magnetic surveys mentioned above, Parker et al. (1981) carried out an induced polarisation (IP) survey as part of exploration for porphyry copper mineralisation associated with the Culvennan Fell dioritic intrusion. IP anomalies (zones of higher chargeability) were detected over the sedimentary rocks to the west-south-west of the intrusion but these lay parallel to the local strike and were apparently unrelated to the intrusive body, which was characterised by low chargeabilities. Parker et al. (1981) concluded that the anomalies were more likely to be due to polarisable units within the Lower Palaeozoic sequence than porphyry-style mineralisation.

Magnetic and Very Low Frequency electromagnetic (VLF-EM) surveys were conducted around the Talnotry Mine [NX 479 704] in order to investigate possible extensions to the nickel-cobalt mineralisation which occurs there (Parker, 1977). Large amplitude magnetic anomalies, probably due to pyrrhotite associated with the mineralisation, were detected directly over the mine and an adjacent tip. However, no comparable anomalies were found elsewhere in the area and there was no significant VLF-EM response.

Some geophysical trials were undertaken by Rollin (1983) on the west side of the Wigtown peninsula near Culroy [NX 253 540], Garheugh [NX 274 504] and Port William [NX 340 435]. The aim was to assess the value of IP and VLF-EM surveys for mapping mudrocks of the Moffat Shale Group beneath drift cover with a view to applying these techniques in the poorly exposed ground around Port William. Orientation studies at Dob's Linn had suggested that the Moffat Shale Group is considerably more conductive and polarisable than neighbouring greywacke. Strong electrical and electromagnetic responses were obtained over exposed mudrocks at Culroy and Garheugh, although in some cases the geophysically defined boundaries did not coincide exactly with those identified at outcrop. In the Port William area, however, the geophysical methods were unsuccessful because of a combination of interference from man-made structures and the screening effect of thick, electrically conductive drift.

Chapter 4 Ordovician and Silurian rocks

The district is underlain predominantly by sandstone-dominated sedimentary sequences of Ordovician and Silurian age (Figure 9). Although generally steeply dipping and variably folded and cleaved (Chapter 5), over much of the district the rocks have been affected by only sub-greenschist facies regional metamorphism (Chapter 6) and details of the original sedimentary structures have largely been preserved. Thermal alteration of rocks adjacent to the late Silurian to Devonian granitoid intrusions (Chapter 7) and locally intense cleavage, particularly in the Moniaive Shear Zone in the north-east of the district, obscure original texture to a greater degree.

The sedimentological characteristics of the sandstone-dominated sequences in the district are consistent with transport and deposition by turbidity currents (e.g. Pickering et al., 1986, 1989). Well-bedded sequences were deposited from relatively low concentration turbidity currents; individual sandstone beds display grading and the typical range of internal sedimentary structures (Figure 10) first recognised by Bouma (1962). Sandstone beds commonly also exhibit external bedding plane or sole structures such as flames and flute/groove casts on the base of beds and ripples on bed tops which provide valuable evidence of palaeocurrent directions. Internal and external sedimentary structures allow the 'way-up' of the sequence to be determined locally. Massive sandstone bodies, typical of the Shinnel Formation and parts of the Gala Group, generally without obvious sedimentary structures and sometimes packed with intraclasts, were deposited from high concentration turbidity currents.

Throughout this account, sequences are related to the turbidite facies model of Pickering et al. (1986, 1989) and the structures within individual beds are described using the standard 'Bouma' terminology of Tab, Tabc, etc. The term greywacke (sensu Pettijohn, 1975) describes sandstone that contains more than 15 per cent of matrix material. In this account, evidence of deposition of a sandstone from a turbidity current is taken to indicate there is a likelihood of more than 15 per cent matrix and hence that greywacke is an approporiate term. Sandstone is used as a more general term where there is no field evidence for deposition from turbidity currents; these rocks might be either greywackes or quartz/lithic arenites sensu stricto. All palaeocurrent data cited have been corrected as far as possible for rotation of bedding during folding and tilting of the succession. This correction is clearly uncertain, particularly in areas with complicated patterns of fold plunge, but the general azimuth or direction inferred from the bulk of the data is likely to be a reliable indication of the average palaeocurrent orientation.

Evolution of stratigraphical subdivision

The first systematic accounts of the geology of the district by Craik (1873) and Irvine (1878) followed the primary geological survey of sheets 2 and 4 respectively. The lithostratigraphy, largely reflecting survey work in adjacent areas, was based on broad lithological character and the perceived order of superposition but lacked biostratigraphical or structural control. The 'Ardwell Group' (= Hawick Group), cropping out over a large area in the south of the district, was considered the oldest part of the succession, succeeded by the 'Lower or Moffat Black Shale Group' from its first occurrence northwards at Elrig Loch [NX 323 492], and then the 'Queensberry Grit Group' (= Gala Group). Inliers of the black shale within the latter were thought to occur in the cores of large anticlines. Four divisions of the strata in the north of the district were based mainly on variations in character of the turbidite sequence defined to the north-east, but included the 'Upper Black Shales' near the top of the succession. It was noted by Irvine (1878, p. 15), however, that the latter 'bears so strong a resemblance to the lower unit that, but for the evidence of superposition, it might readily be identified with that band'. The whole of this sequence was assigned to the Lower Silurian (= Ordovician), although Upper Silurian graptolite species listed from Burrow Head suggested the presence there of the 'Riccarton Beds' of Wenlock age considered to lie unconformably on, or be faulted against, the older rocks (Lapworth and Wilson, 1871).

More or less in parallel with the work of the primary survey in the Southern Uplands, Lapworth (1876, 1878) developed a new stratigraphical framework based on graptolite faunas and demonstrated that the black mudstones were in fact a single sequence. Consequently, the large-scale stratigraphy of the Southern Uplands was radically revised by Peach and Horne (1899), who, 'for the sake of convenience', described three strike-parallel belts reflecting the age of the sandstone succession:

This stratigraphy, and the recognition that the 'oceanic deposits' of the Moffat Shale had been deposited continuously as the overlying coarse terrigenous materials were carried progressively farther south during the Ordovician and Silurian, have remained essentially unchanged to the present day.

There was little subsequent published work until Walton (1955, 1961) and Craig and Walton (1959) demonstrated that the southward younging of the succession regionally was inconsistent with the dominantly northward younging observed locally at outcrop. This instigated a long debate about the structure of the Southern Uplands (Chapter 5).

In the Northern Belt, major advances in defining a stratigraphy in the Ordovician sandstone succession have come from the definition and mapping of lithostratigraphical units based largely on consistent petrographical differences (Kelling, 1961, 1962; Welsh, 1964; Floyd, 1976, 1982; Stone et al., 1987; Floyd and Rushton, 1993; Stone, 1995). Recently, the Barrhill and Scaur groups have been defined on the basis of the dominant compositional type of sandstone, quartzo-feldspathic or volcanic-rich, within a Leadhills Supergroup (Floyd, 1996).

In the absence of comparable compositional variation, subdivision of the Central Belt sandstone has been based mainly on variation in sedimentary facies (Gordon, 1962; Rust, 1965; Barnes et al., 1987; Kelling et al., 1987; Davies, 1990; Stone, 1995). The two principal divisions, as recognised previously, were termed the Gala and Hawick groups by Barnes et al. (1987) and White et al. (1991). The biostratigraphical age of the Gala Group 'formations' is generally relatively well-constrained from interbedded fossiliferous material or by the youngest fauna in the underlying Moffat Shale Group. In the absence of either control, the age of the Hawick Group was for some time a matter of conjecture based mainly on relationships with the fossiliferous Wenlock 'Riccarton Beds' (Craig and Walton, 1959; Warren, 1964; Clarkson et al., 1975; Kemp and White, 1985). However, in Wigtownshire, Rust (1965) found sparse graptolites considered by Strachan to be indicative of a late Llandovery age in one of two formations defined in the southern part of the Hawick Group outcrop. This age has been supported by new collections made during this resurvey and a review of existing material (Barnes, 1987; White et al., 1991; Lintern and Floyd, 1998). This work has led to a four-fold division of the Hawick Group in south-west Scotland as described in this account, including the early Wenlock Ross Formation, which crops out in Kirkcudbrightshire (Clarkson et al., 1975; Kemp and White, 1985) and also on the tip of Burrow Head and at Isle of Whithorn in Wigtownshire (Craik, 1873; Irvine, 1878; Rust, 1965). Younger Wenlock strata exposed in Kirkcudbrightshire and to the east form the Riccarton Group which can be subdivided both lithostratigraphically and biostratigraphically (Kemp and White, 1985; Lintern and Floyd, 1998).

The lithostratigraphy of the Moffat Shale Group, summarised by Floyd (1996), has been little altered since Peach and Horne (1899). However, volcanic rocks (described by Peach and Horne (1899), Lambert et al. (1981), Barnes et al. (1995b), and Phillips et al. (1995a)) and associated chert and cherty mudstone seen locally at the base of the exposed succession were separated by Floyd (1996) as the Crawford Group.

Biostratigraphy

The biostratigraphy of the district (and indeed of the entire Southern Uplands terrane) is largely defined by graptolites; details of faunas and fossiliferous localities are summarised in appendices 1–4. Fossiliferous horizons are well developed in the Moffat Shale Group, otherwise are generally sparsely distributed in the Southern Uplands and tend to have relatively poorly preserved graptolite faunas of low diversity. Hence the biostratigraphy of the Moffat Shale Group is critical to the interpretation of the geology of the Southern Uplands, dating not only the mudstone sequence itself, but also constraining the age of the transition into the overlying greywacke succession (Figure?9). The zonal scheme is based very largely on the distribution of graptolites in the Moffat Shale Group as worked out by Lapworth (1878) and subsequently elaborated by various other authors. The use of the Silurian zones essentially follows that of Rickards (1976), except that the persculptus Biozone is now treated as uppermost Ordovician rather than basal Silurian (Cocks and Rickards, 1988). The zones of the Ordovician are more problematical, as discussed by Strachan (1971) and Rushton (1990). Elles and Wood (1901–1918) and Elles (1925) listed the species according to their zonal occurrences, and for the lower part of the sequence (gracilis to 'peltifer' biozones) there is only partial revision. The use of these zones is discussed further below. Williams (1982a, 1982b, 1983, 1987, 1994; Zalasiewicz et al., 1995) recorded the stratigraphical distribution of higher Ordovician graptolites, from the wilsoni Biozone to the acuminatus Biozone, mainly in sections at Dob's Linn, Moffatdale. This new precision in the higher zones has been exploited in more recent work in the Southern Uplands, although some discrepancies between Williams' results and Elles and Wood's table have yet to be resolved.

The recognition of the gracilis Biozone was discussed by Finney and Bergstrom (1986), who took the base of the zone at the appearance of Nemagraptus gracilis itself. Hughes (1989, p.22) provisionally took the top of the zone in the Welsh Borderland at the disappearance of N.?gracilis. Besides N. gracilis, Didymograptus superstes appears to be the most diagnostic element of a large fauna, which otherwise has much in common with that of the overlying 'peltifer' Biozone. The base of the gracilis Biozone forms the base of the revised Caradoc Series of Fortey et al. (1995).

The 'peltifer' Biozone is not satisfactorily defined (Strachan, 1971, p.5), and furthermore the zonal species is merely an astogenetic variant of Climacograptus bicornis (Riva, 1976, p.595). A fauna from Camrie Burn [NX 2050 6055] that includes Climacograptus bicornis, C.?brevis? Corynoides calicularis Cryptograptus tricornis, Dicranograptus nicholsoni, D.?ziczac, Lasiograptus harknessi, Orthograptus calcaratus (group), O. cf. whitfieldi and Pseudoclimacograptus scharenbergi is referred to the 'peltifer' Biozone on account of the presence of C.?calicularis and L. harknessi, and the absence of N.?gracilis. Corynoides curtus seems to occur in the gracilis Biozone but C.?calicularis has not so far been found in association with N.?gracilis in south-west Scotland.

The wilsoni Biozone (revised by Williams, 1994) is characterised by a relatively low-diversity fauna mainly of diplograptids, but has not been recognised in the district. The clingani Biozone includes a diverse fauna characterised by Dicranograptus spp. and by several species of Dicellograptus that are not known at lower horizons (Elles, 1925). Zalasiewicz et al. (1995), on the basis of the distribution of species at Hartfell Spa, proposed a twofold subdivision of the clingani Biozone which appears to have value in southern Scotland and also in Wales: the caudatus Subzone below, characterised by Ensigraptus caudatus, Dicranograptus clingani and Climacograptus spiniferus and the morrisi Subzone above, with Climacograptus dorotheus and Dicellograptus morrisi.

Several species present in the clingani Biozone pass up into the linearis Biozone, so that it may not be possible to assign small collections to one or other zone with certainty. Williams (1982a), who recorded a succession of the first appearances of species at Dob's Linn, took an arbitrary base to the linearis Biozone at the level where Neurograptus margaritatus disappears. Pleurograptus linearis itself occurs at a higher horizon and is characteristic of the zone, but can be difficult to determine with certainty if only fragmentary material is collected. Therefore, Zalasiewicz et al. (1995) preferred to take the base of the linearis Biozone at the appearance ofClimacograptus styloideus and/or C.?tubuliferus. A typical fauna from Camrie Burn includes P.?linearis with Arachnograptus sp., Climacograptus cf. dorotheus, Leptograptus flaccidus subspp., Orthograptus amplexicaulis and O.?quadrimucronatus. Elsewhere, the presence of L.?flaccidus macer or Climacograptus tubuliferus, coupled with the absence of species typical of the clingani Biozone, is taken to represent the linearis Biozone. Although species of Corynoides and Dicranograptus have been recorded from the linearis Biozone elsewhere, neither genus is known to occur at this level either at Dob's Linn or in the district.

The complanatus and persculptus biozones have not been recognised in the district. The anceps Biozone is characterised especially by Orthograptus abbreviatus. Some of the collections include O.?fastigatus, suggesting the presence of the complexus Subzone, the lower of two subzones that Williams (1982b) recognised in the anceps Biozone at Dob's Linn; but at one outcrop O.?fastigatus is associated with Dicellograptus ornatus which Williams recorded only in the upper or pacificus Subzone, although other species indicative of the pacificus Subzone were not collected there. Further work is therefore needed to discover how far Williams' distributional information is applicable in south-west Scotland.

The zonal distribution of graptolites from the Llandovery and Wenlock was evaluated at the time of the mapping (between 1982 and 1989) following Rickards (1976). The later Llandovery and Wenlock graptolites were reported by White (in White et al., 1991). Dr?D?K?Loydell subsequently re-examined the late Llandovery collections, revised their identifications and, where necessary, reassigned their zonal positions according to the refined zonal scheme more recently developed in the Llandovery of mid-Wales (Loydell, 1992, 1993; Loydell & Cave, 1993, 1996). The differences between the zonal schemes can be seen by comparing the biostratigraphical summary diagrams on the maps and Figure?9.

Tectonostratigraphical subdivision of the sandstone succession

As described above, the sandstone succession in the Southern Uplands is divided into 'groups', which primarily reflect the differences in character of the deposits, although are based partly on age. All of these groups except the Riccarton Group lie stratigraphically above the Moffat Shale Group; hence there may well be lateral and/or vertical transitions between them. Such transitions are recognised locally by interbedding of Ordovician sandstones of different composition in the north-west of the district. Otherwise, however, boundaries tend to be drawn at strike-parallel faults. Whilst this may be the case locally, it is undoubtably in general an over-simplification.

Further subdivision, into 'formations', separates units in two ways: by contrasting petrography and age in the Ordovician Leadhills Supergroup in the north-west of the district, and also to some extent in the case of the Mindork Formation in the Gala Group; otherwise by more subtle variations in sedimentary facies and, in the Gala Group, by age. At this scale, the boundaries are almost always taken at strike-parallel faults as a matter of convenience and they do not represent the original sedimentary geometry of the natural subdivisions of the succession. This approximation is not unreasonable when constrained by variation in petrography. However, it is less satisfactory where based only on facies variations that may vary rapidly in any direction. Formations have been defined on the latter basis in the Gala Group in the west of the district and can sometimes be correlated for up to 20 km along strike to the west (e.g. Barnes et al., 1987). However, with a large area of poor exposure in the centre of the district and less distinctive facies in the east, it is the tract-bounding fault traces that tend to be correlated and not the rocks between. In recognition of this problem, a tectonostratigraphical nomenclature (Gala 1, Gala 2 etc.) is employed in parallel with the lithostratigraphical nomenclature in the Gala Group in the western parts of the district, but is continued eastwards in isolation as the lithostratigraphical distinction breaks down.

The problem is less obvious in the Hawick Group, which is generally of remarkably uniform character with relatively poor biostratigraphical resolution throughout the district. The lowest 'formation' is defined by distinctive turbidite facies locally but its boundary at the Innerwell Fault is again poorly constrained. The other divisions are distinguished by relatively minor interbedded components on which basis the boundaries can be distinguished within the limited number of tracts defined (see Chapter 5). It is noteworthy that some of these variations in the Hawick Group seem to be relatively local in extent, dying out north-eastwards in the Kirkcudbright district (Lintern and Floyd 2000).

Crawford Group

The Crawford Group (Floyd, 1996) underlies the Moffat Shale Group in the Southern Uplands and consists of a succession of grey and/or red, bedded chert (radiolarite) and cherty mudstone, often intimately associated with basaltic pillow lava. The Crawford Group is best seen at several localities in its type area near Abington (Peach and Horne, 1899, pp.286–289; Smith, 1904; Armstrong et al., 1996). Here, bedded chert and red and brown cherty mudstone of undoubted Arenig age are associated with basic volcanic and intrusive rocks.

In the north of the district, basic igneous rocks crop out at three localities in association with the Moffat Shale Group. Outcrops of vesicular pillow lava and chert west of Dirneark [NX 257 703] and on Glenhapple Moor [NX 360 703] could not be re-examined during the present resurvey but were considered by Peach and Horne (1899, pp.390 and 399) to underlie the mudstone of the Moffat Shale Group. These volcanic rocks may therefore represent the Crawford Group, although this division was formally defined (Floyd, 1996) subsequent to publication of the geological sheets for the district (BGS, 1992a & b) wherein the lava and chert were included within the Moffat Shale Group. At the third locality, in the upper Gabsnout Burn area [NX 216 617], the igneous rocks are more intimately associated with the Moffat Shale Group as described below.

Moffat Shale Group

The Moffat Shale Group is a condensed mudstone sequence that ranges in age from lower Caradoc at the base, up to upper Caradoc or Llandovery, depending on its location within the district. In the north-west, the Moffat Shale Group is overlain by greywacke of varying Caradoc ages, whilst in the southern tracts, the group extends well up into the Llandovery before being succeeded by greywacke (Figure 9). At its fullest development, the Moffat Shale Group may be about 80 m thick but, because of its relative incompetence and its exploitation as the basal décollement of the Southern Uplands thrust slices, it is commonly highly tectonised and continuous successions are rare. It is generally less well preserved beneath mid- to late-Llandovery sandstone sequences where its absence may be due to lack of preservation or may accurately reflect its original distribution.

Lithostratigraphy

Despite the structural complexity of the Moffat Shale Group, a distinctive lithostratigraphy has been known since Lapworth's (1878) biostratigraphical zonal scheme allowed the correlation of discontinuous exposures. The base of the sequence is taken at the first black shale overlying the bedded chert of the Crawford Group (Floyd, 1996). The oldest part of the group, the Glenkiln Shale (gracilis and peltifer biozones) is composed of hard black mudstone or siltstone with thin black cherty layers. This is succeeded by the Hartfell Shale, divided into a lower division (wilsoni to linearis biozones) of black mudstone with thin cherty layers and an upper division (complanatus and anceps biozones) composed of massive grey mudstone or siltstone with a few fossiliferous black mudstone bands. The upper, largely Silurian, part of the Moffat Shale Group, the Birkhill Shale, is composed of grey to black mudstone with frequent thin beds of pale metabentonite representing original tuffaceous horizons (Batchelor and Weir, 1989; Merriman and Roberts, 1990).

Detail of Moffat Shale Group outcrops

Other than the Moffat Shale Group outcrops associated with the Leadhills (Killantringan) Fault and the north-east part of the Morroch Bay Fault Zone, all of the outcrops in the district have been remapped in detail. Four of the five Moffat Shale Group inliers present in the west of the district have also been re-examined biostratigraphically from new collections made at selected localities; no identifiable fossils are preserved within the Cairnsmore of Fleet metamorphic aureole. No exposure of the southernmost outcrop of Moffat Shale Group in the district, previously recognised in the vicinity of Elrig Loch [NX 325 493] by Irvine (1878), was found during the resurvey. All of the available material from the graptolite collections made during the primary survey (Craik, 1873; Irvine, 1878; Peach and Horne, 1899) was also re-examined. A listing of species identified is given in Appendix 1; selected forms are illustrated in (Figure 11).

Because of their structural position as the décollement horizon at the base of each of the fault-bounded tracts, most outcrops of the Moffat Shale Group are highly disrupted. Outcrops along the fault traces are discontinuous and range in width from a few metres to 500?m due to tectonic excision and repetition. Commonly, imbrication can be demonstrated at two scales: the outcrops may consist of several strands due to repetition by imbrication of the Moffat Shale Group and the base of the overlying greywacke sequence, or, within an outcrop, imbrication of the mudstone is shown by detailed biostratigraphical data.

In the most extreme case of the latter so far identified, in a 12?m-long exposure in the roadside near Garheugh Farm [NX 276?503], there are four tectonic slices: three repetitions of Lower and Upper Hartfell Shale, including the clingani Biozone, and a 3?m thickness of faulted Birkhill Shale which yielded atavus/acinaces and triangulatus Biozone faunas. A well-exposed section through the Moffat Shale Group in Gillespie Burn [NX 257?539] also includes at least three repetitions of the Lower and Upper Hartfell Shale (Figure?12) overlying an unusual pale green and red mudstone sequence. At this locality the acuminatus to acinaces biozones are represented by 12?m of interbedded greywacke and mudstone rather than by the more characteristic lithologies of the Birkhill Shale. One of these greywacke beds yielded abundant Cystograptus vesiculosus and Normalograptus normalis, together with rare Parakidograptus acuminatus, indicating the upper part of the acuminatus Biozone. Overlying mudstone beds yielded sparse faunas suggestive of the atavus, acinaces and cyphus biozones, the last immediately overlain by turbidites of the Sinniness Formation from which cyphus Biozone faunas have been found at a number of localities.

Graptolite faunas from the particularly wide Moffat Shale Group outcrop in the upper Gabsnout Burn area are mainly from the clingani Biozone although faunas from exposures north of Camrie Loch indicate the presence of older material (gracilis/'peltifer' Biozone), suggesting tectonic imbrication (Figure?13). A sedimentary transition from the mudstone to greywacke of the overlying Portpatrick Formation is exposed in Gabsnout Burn near Mid Gleniron (Sheet?3 [NX 189?612]) and in Garwachie Quarry [NX 3468?6964] (Figure?14c). Alternating exposures of the Moffat Shale Group and the Portpatrick Formation in the vicinity of Camrie Fell [NX 200?610] are interpreted as tectonic repetitions of this transition (Figure?13) although the structure is likely to be more complicated in detail.

Basic igneous rocks in the upper Gabsnout Burn area

Basic igneous rocks appear within the outcrop of the Moffat Shale Group that is associated with the Morroch Bay Fault Zone between the farms of Mid Gleniron and Garvilland [NX 216 617] (Figure 13). These were mapped by Irvine, during the primary survey, as dolerite dykes. Peach and Horne (1899 p. 394) referred briefly to an exposure of 'lava' and to 'knobs' of dolerite at this locality. More recently Lambert et al. (1981) provided analyses of five specimens of 'basalt' from this locality as part of a regional study of basalt geochemistry in the Southern Uplands but provided no detail of the field relationships. A new interpretation of the outcrop pattern and geochemistry of these occurrences of basic rocks (Barnes et al., 1995b) is summarised below.

Basalt and dolerite crop out as lenses surrounded by black mudstone. Several of the smaller bodies are composed of dolerite alone, others include both basalt and dolerite and one is entirely formed of basalt. The dolerite is massive and usually of a uniform grain size although local variations suggest that it may include dolerite dykes. Most of the basalt exposed is also structureless, however, in a few places, curved chilled faces and small mudstone inclusions show that it is, at least in part, pillow lava. Two lenses include both basalt and dolerite; in the westernmost of these, basalt occurs to the north of dolerite; in the other, dolerite appears to have an intrusive relationship to the lava. The two largest lenses of basic rocks have an unusual mottled or banded, grey to black chert exposed at their northern margins; this lithology is also seen near Garvilland (Figure 13).

Only the western body of basic rock shows any significant effects of deformation. The dolerite is broken by randomly orientated, carbonate-filled fractures and includes a weak, strike-parallel fabric that becomes more intense towards the southern margin of the outcrop. Basalt in exposures near the northern and eastern margins of the body is also foliated. These features establish that the igneous rocks had an early, pre-tectonic genesis and are not post-tectonic dykes.

The lenticular units of basic igneous rocks and associated chert, all enveloped in mudstone of clingani Biozone age, may have been formed in one of three ways:

Although there is a suggestion of a relict volcanic stratigraphy, rising northwards from dolerite into basaltic pillow lava overlain by chert, other features suggest that these rocks were not formed in situ. Basalt is subordinate to dolerite which shows no indication of having been intruded into the host mudstone. These outcrops are very localised and in situ basic volcanic rocks are otherwise unknown from within the Moffat Shale Group. In view of the position of the basic rock bodies within the younger part of the Gabsnout Burn mudstone sequence, the shape of the outcrops and the minimal amount of deformation which they have suffered, it also seems unlikely that they were tectonically introduced. Hence an interpretation as large rafts of material, shed from a localised, elevated area of basic igneous rocks during mudstone sedimentation, is preferred. This material probably represents the Crawford Group, which forms the Arenig–Llanvirn chert and basaltic igneous substrate to the Moffat Shale Group.

Despite very low grade regional metamorphism, evident from the presence of actinolite and prehnite in some samples, the basic igneous rocks are chemically relatively unaltered (Barnes et al., 1995b). Contrary to what might be expected from their association with the basinal mudrock sequence and from some previous regional interpretations of the geology (cf. McKerrow et al., 1977; Leggett et al., 1979; Lambert et al., 1981) the rocks are geochemically distinct from typical mid-ocean ridge basalt. Instead, they have characteristics similar to that of basalt formed in an island-arc setting. Their occurrence as basic olistoliths within the Moffat Shale Group is, therefore, enigmatic and difficult to reconcile with either a back-arc or fore-arc setting for the Southern Uplands basin during the late Ordovician.

Leadhills Supergroup

The Leadhills Supergroup, broadly equivalent to the 'Northern Belt' of Peach and Horne (1899), includes the Ordovician turbidite sequences in the northern part of the Southern Uplands. At its northern margin in the Girvan district, the Leadhills Supergroup is separated from rocks of a similar age (but very different lithology) by the Stinchar Valley Fault (Figure 3). This is the local expression of the Southern Upland Fault, which elsewhere forms the northern boundary to both the Leadhills Supergroup and the Southern Uplands terrane. The southern margin of the supergroup, with the Gala Group, is formed by the Orlock Bridge Fault. However, its diachronous base, cropping out in several fault-bounded tracts, rests with stratigraphical conformity on the Moffat Shale Group (Figure 9). Five of its component formations crop out in the district ((Figure 10); inside front cover) and are distinguished by their contrasting quartzo-feldspathic and volcaniclastic sandstone compositions (Floyd, 1982; Stone et al., 1987; Floyd, 1996). Representative measured sections (Figure 14) illustrate typical sequences in the different sandstone formations. Detailed sedimentological interpretation of equivalent strata, along strike to the south-west in better-exposed sections on the Rhins of Galloway, is given by Kelling et al. (1987).

Barrhill Group (Kirkcolm and Galdenoch formations)

The north-western tract in the district, north of the Killantringan–Leadhills Fault, comprises interbedded turbidites of two compositional types. The Kirkcolm Formation (e.g. Figure?14a) typically consists of fine- to medium-grained, medium- to thick-bedded greywacke (Tabc, Tabcd and Tbcd, Facies C2.1 and C2.2) with thin mudstone interbeds, often associated with packets, up to 10 m-thick, of laminated siltstone with minor thin-bedded, fine-grained greywacke (Tc and Tcd, Facies D2). The Galdenoch Formation (Figure?14b) is broadly similar to the Kirkcolm Formation in appearance and facies, although the greywacke is medium-grained with lithic clasts commonly about 0.5?mm in diameter but ranging up to 5?mm. The Galdenoch Formation also tends to be more thickly bedded and thick siltstone packets are absent. The two formations are primarily distinguished on composition (see below), the Galdenoch Formation containing abundant andesitic volcanic debris that is absent in the Kirkcolm Formation. Although these petrographic differences cannot usually be determined in the field, measurement of magnetic susceptibility can provide a useful alternative guide since the Galdenoch greywacke is appreciably more magnetic than that of the Kirkcolm Formation (Floyd and Trench, 1989). Palaeocurrent indicators from the Kirkcolm and Galdenoch formations in the district and along strike in the Carrick-Loch Doon district (Floyd, 1999) show that the two formations generally have different transport directions, typically from the north-east and south-east respectively.

The north-western tract in the district, north of the Killantringan–Leadhills Fault, comprises interbedded turbidites of two compositional types. The Kirkcolm Formation (e.g. (Figure 14)a) typically consists of fine- to medium-grained, medium- to thick-bedded greywacke (Tabc, Tabcd and Tbcd, Facies C2.1 and C2.2) with thin mudstone interbeds, often associated with packets, up to 10 m-thick, of laminated siltstone with minor thin-bedded, fine-grained greywacke (Tc and Tcd, Facies D2). The Galdenoch Formation ((Figure 14)b) is broadly similar to the Kirkcolm Formation in appearance and facies, although the greywacke is medium-grained with lithic clasts commonly about 0.5 mm in diameter but ranging up to 5 mm. The Galdenoch Formation also tends to be more thickly bedded and thick siltstone packets are absent. The two formations are primarily distinguished on composition (see below), the Galdenoch Formation containing abundant andesitic volcanic debris that is absent in the Kirkcolm Formation. Although these petrographic differences cannot usually be determined in the field, measurement of magnetic susceptibility can provide a useful alternative guide since the Galdenoch greywacke is appreciably more magnetic than that of the Kirkcolm Formation (Floyd and Trench, 1989). Palaeocurrent indicators from the Kirkcolm and Galdenoch formations in the district and along strike in the Carrick-Loch Doon district (Floyd, 1999) show that the two formations generally have different transport directions, typically from the north-east and south-east respectively.

Two units of the Galdenoch Formation, each about 300?m-thick, are mapped as being interbedded with the Kirkcolm Formation in the district, although the degree of exposure does not preclude more intricate interbedding or tectonic contacts. No contacts are exposed in the district but along strike to the south-west on the Rhins of Galloway (Stone, 1995) and to the north in the Barrhill area (Floyd, 1999), similar units, ranging from 25?m to 300?m thick, do have sedimentary contacts. North of the Killantringan Fault, near Tannylaggie [around 287?717], the Kirkcolm Formation overlies mudstone of the Moffat Shale Group. The mudstone yields faunas of the gracilis Biozone, including Nemagraptus gracilis, Climacograptus bicornis (with subsp. 'peltifer'), Dicellograptus exilis, D.?salopiensis, Dicranograptus cf. furcatus minimus, Didymograptus superstes and Hallograptus (Peach and Horne, 1899) suggesting that the overlying sandstone is of gracilis to 'peltifer' Biozone age (Figure?9).

Peach and Horne (1899, p.397) recorded another fauna, from a bed of sandy mudstone in greywacke south of Tannylaggie farmhouse [NX 287?715], which appears to be more probably referable to the clingani Biozone, although the original specimens are missing. Along strike to the south-west, on the Rhins of Galloway (Stone, 1995), the Kirkcolm Formation is shown to crop out in two tracts, separated by the Glaik Fault. In the northern tract, dark mudstone interbeds within the Kirkcolm Formation include gracilis and 'peltifer' Biozone faunas, whereas in the southern tract, greywacke referred to the Kirkcolm Formation overlies the Moffat Shale Group with faunas as young as clingani Biozone. The Glaik Fault is mapped eastwards only as far as Loch Ryan with only the northern tract shown in the eastern part of the district. However, if the clingani fauna from near Tannylaggie farmhouse is correctly assigned, the Glaik Fault, though not recognised, must pass through the district. As on the Rhins of Galloway, it would separate an older, northern Kirkcolm/Galdenoch tract of gracilis-'peltifer' biozone age from a younger southern Kirkcolm tract of clingani Biozone age. An identical subdivision of the Kirkcolm Formation, complete with Glaik Fault equivalent, has been identified in the Peebles area (Rushton et al., 1996).

Scaur Group (Portpatrick and Glenwhargen formations)

The Portpatrick Formation forms a broad belt within the district, between the Morroch Bay and Killantringan faults. It is well exposed on several rocky hills and comprises turbidites poor in quartz and rich in andesitic debris. Contrasting highly quartzose sandstone beds with no andesitic debris is locally interbedded and distinguished as the Glenwhargen Formation. They crop out near the northern margin of the district but are better known from the Carrick–Loch Doon district to the north, where stratigraphical continuity with the Portpatrick Formation can be demonstrated (Floyd, 1999). As with the Kirkcolm–Galdenoch formations in the tract to the north, visual discrimination of the two sandstone types in weathered exposures can be difficult, though the two formations can again be distinguished by their magnetic susceptibility (Floyd and Trench, 1989). On fresh surfaces, the Portpatrick greywacke is also generally dull, dark bluish grey and massive whereas the Glenwhargen greywacke is vitreous, pale grey and commonly laminated. Abundant palaeocurrent data from the Portpatrick Formation shows dominantly east to north-east azimuths with flute casts demonstrating transport from the west and south-west. The Glenwhargen Formation sandstone is better exposed in the Carrick–Loch Doon district to the north where bed thickness is generally in the range 0.2 to 0.8 m, though greywacke units up to 2 m occur along with conglomerate units up to 5 m or more, and transportation is from the north-east.

The Portpatrick Formation is composed of medium- to very coarse-grained greywacke that is pebbly in places and occurs in beds ranging from 0.10 to 2?m thick, usually separated by thin mudstone beds (Facies?C2.1 and C2.2). Pebbles ranging up to 1?cm diameter, composed of quartz or fine-grained volcanic rock, are dispersed through the lower parts of beds or are concentrated in a pebbly sandstone base. They are included in the graded Ta division of the sandstone beds, which is generally prominent, although laminated and cross-laminated finer grained bed tops are commonly present (Tab, Tabc) and some beds are laminated throughout (Tbc, Tc). The mudstone beds are usually 5 to 15?cm thick but range up to 30?cm. Very thick (up to 6?m) massive sandstone beds (Facies?B1.1) also occur, dispersed throughout the sequence, and locally form sequences of massive sandstone in which bedding can be difficult to define.

The base of the Portpatrick Formation can be seen in a quarry at Garwachie [NX 3368 6964] where it overlies variably tectonised black shale of the Moffat Shale Group. Here, a single thin bed of quartzose sandstone, with no andesitic debris, occurs about 1 m above the first andesite-rich sandstone sequence at the base of the Portpatrick Formation ((Figure 14)c). Another sedimentary transition from Moffat Shale Group to the Portpatrick Formation is exposed in Gabsnout Burn [NX 1925 6090], just beyond the western margin of the district, where dark grey mudstone with a few thin greywacke beds passes northwards into greywacke, mudstone and black mudstone. Several graptolite collections from this vicinity have yielded faunas of the clingani Biozone or are consistent with that zone, though indefinite by themselves. Hence the age of most of the Portpatrick Formation is taken as clingani Biozone (or younger), although no fossils have been obtained from within the formation in the district.

The Morroch Bay Fault, at the southern margin of the outcrop of the Portpatrick Formation, is mapped as two parallel strands up to 1 km apart based on the disposition of Moffat Shale Group outcrops. These separate a narrow tract within which the Portpatrick Formation overlies mudstone of linearis Biozone and is therefore younger than that to the north of the northern strand of the fault. This relationship was first described from the Rhins of Galloway, to the west of the district (Stone, 1995) and has also been discussed by Rushton et al. (1996) and Floyd (1999). Further cryptic strike-parallel faults may be present within the northern parts of the outcrop but are not recognised since they have no associated Moffat Shale Group rocks.

Shinnel Formation

The Shinnel Formation is well exposed in the district around Glenluce, north of Kirkcowan and north and north-east of Newton Stewart and is characterised by quartzo-feldspathic sandstone without obvious intermediate volcanic debris. The turbidite facies (e.g. (Figure 14)d–f) show an unusual degree of vertical and lateral variability compared with other formations in the region, in some places being mud-dominated and in others predominantly coarse-grained pebbly sandstone.

A thick siltstone unit occurs at the base of the formation. In the west of the district this lithology is exposed south of Glenluce e.g. [NX 204 571] and at Wood of Dervaird [NX 225 577] where it is intensely deformed by the Orlock Bridge Fault (Barnes et al., 1996). It is seen again north of Newton Stewart in the River Cree [NX 402 670] and the Penkiln Burn, where it overlies Moffat Shale Group (e.g. north of Glenhoise [NX 429 684]), and is relatively undeformed.

Above the siltstone, the formation is dominated by turbidites of facies C2, comprising medium- to thick-bedded (Ta,Tab,Tabc), medium-grained greywacke, some beds with coarse-grained, rarely pebbly, bases. Sandstone beds are usually separated by thin mudstone partings although thicker (0.20?–1?m) mudstone beds, with very thin siltstone or cross-laminated (Tc) fine-grained greywacke beds, are common. Massive, structureless sandstone (Facies?B1.1) may form units tens of metres in thickness, with grading absent, poorly developed or sometimes repeated, suggesting amalgamated beds of Facies?C2.1.

The well-bedded turbidite facies are exposed north-east of Glenluce (e.g. south of Carscreugh [NX 225 600]), extending for over 600 m north of the Orlock Bridge Fault in a highly folded sequence. This passes upwards and laterally into a facies in which mudstone is much more abundant, occurring in units ranging from 1 m to several tens of metres thick and interdigitated with well-bedded or massive sandstone. The mudstone may be structureless (Facies E1.1) or laminated (Facies E2.1) and may include interbedded siltstone (Facies D2.1), very thin fine-grained cross-laminated greywacke beds (Facies C2.3) and some thin greywacke beds in sections transitional into the greywacke-dominated facies. A thin unit of dark grey to black mudstone can be traced for about 1 km along strike in a series of small outcrops near Glenhowl [NX 2045 5912] to [NX 2150 5981] (Peach and Horne, 1899, p.393).

The mudstone-dominated facies occurs again at the north-east end of the Shinnel Formation outcrop in the district. At Tors of Glenmalloch [NX 429 704] mudstone occurs in units up to 30 m-thick with a variable proportion of very thin- to thin-bedded, fine-grained greywacke (Tb, Tbc and Tc), commonly with well-developed convolute lamination. These units have sharp contacts with sequences, up to 50 m-thick, of thick- to very thick-bedded massive greywacke, locally with rounded quartz pebbles up to 10 mm diameter. A similar interdigitation of facies occurs south of Culvennan although here massive, normally parallel-laminated, very thick-bedded, coarse-grained, locally pebbly greywacke is dominant but includes units of mudstone and fine-grained, thin-bedded, laminated or cross-laminated greywacke.

The basal siltstone of the Shinnel Formation in the River Cree [NX 4016 6696] rests on mudstone of the Moffat Shale Group which has yielded fragmentary graptolites, probably representing the linearis Biozone. A fragmentary Dicellograptus collected from dark grey/black mudstone exposed within the Shinnel Formation near Glenhowl [NX 207 594] proves the Ordovician age of the formation and is consistent with a late Caradoc, linearis Biozone age inferred from the underlying mudstone. To the north-east, near Moniaive in the New Galloway–Thornhill district, the top of the Shinnel Formation ranges up into the anceps Biozone (Floyd and Rushton, 1993).

Sandstone petrography and composition

The various formations of the Leadhills Supergroup, described above, are recognised by means of sandstone petrography and tectonostratigraphical position. Petrographically, the sandstone comprises two types, distinguished by the proportion of volcanic lithic and mafic mineral debris.

Quartzo-feldspathic sandstone The background sandstone, forming the Kirkcolm, Glenwhargen and Shinnel formations, is predominantly composed of quartzo-feldspathic material. White quartz pebbles are common in the Shinnel Formation. Lithic detritus forms up to 20 per cent of the sand-grade material; it consists mainly of fine-grained volcanic rocks, most commonly variolitic, spilitic rock fragments and a variety of felsic rocks, but there is also plentiful sedimentary and metamorphic detritus. Mica is common in finer grained samples and is often conspicuous in hand specimen. Tourmaline and zircon are ubiquitous in minor amounts, along with rare garnet in the Shinnel Formation. Chlorite and carbonate, present in some samples in significant amounts, have replaced some grains.

Sandstone with volcanic debris The sandstone in the Galdenoch and Portpatrick formations is, in some respects, similar to that of the other formations of the Leadhills Group. It includes substantial amounts of quartzofeldspathic debris and a range of lithic fragments including a small amount of gabbro, diorite and granite where coarser grained sandstone provides an appropriate sample (notably in the Portpatrick Formation). However, it is distinguished by the occurrence of abundant andesite and dacite rock fragments (up to 25 per cent of the sand grade or coarser material). Andesite has phenocrysts of plagioclase, hornblende and augite, all of which are also common in the sandstone as single mineral clasts. The mafic minerals are usually very fresh, contrasting sharply with the altered nature of the feldspar. The pyroxene detritus in the Galdenoch and Portpatrick formations has transitional calk-alkaline to tholeiitic and more evolved calc-alkaline affinities respectively (Styles et al., 1989, 1995). The apparent freshness of the volcanic detritus has been taken to indicate penecontemporaneous volcanicity and sedimentation (e.g. Stone et al., 1987). However, poorly-constrained radiometric dates from detrital hornblende (Kelley and Bluck, 1989) suggest that the volcanicity was older, Cambrian or possibly even earlier.

Gala Group

The Gala Group of Llandovery turbidites is separated from the Ordovician Leadhills Supergroup to the north by the Orlock Bridge Fault and either passes stratigraphically into or is faulted against the Hawick Group to the south (Figure 1). The Gala Group strata occupy the northern half of the Central Belt of the Southern Uplands as defined by Peach and Horne (1899) and have been variously named Gala or Queensberry 'grits' by early workers (e.g. Lapworth and Wilson, 1871; Peach and Horne, 1899). As in the Leadhills Supergroup, turbidite sequences occur in a number of near-vertical, fault-bound tracts, usually stratigraphically overlying and younging northwards away from the Moffat Shale Group (Figure 9).

The Gala Group is characterised by a limited range of relatively 'proximal' turbidite facies (e.g. (Figure 15)). The sandstone is typically medium to coarse grained and uniformly quartzofeldspathic (e.g. (Plate 2)) apart from the Mindork Formation, which locally contains more basic volcanic material and mafic crystal debris. Most of the succession includes units, tens and possibly hundreds of metres thick, of very thick-bedded greywacke or massive sandstone in which it is difficult to identify any indication of bedding (Facies C2.1 and B1.1). Such strata are interbedded with or pass laterally into medium- to thick-bedded greywacke with thin mudstone beds (Facies C2.1 and C2.2) with packets of thin-bedded greywacke and mudstone (Facies C2.3) or rare thicker mudstone units (Facies E1.1 and/or G2.1) up to a few metres in thickness. Detailed sedimentological interpretation of exposures in the west of the district and in equivalent sections along strike on the Rhins of Galloway is given by Kelling et al. (1987).

Other than the Mindork Formation, units have been defined from facies variation in the best exposed parts of the outcrop, in sections along and slightly inland from the west coast of the district and on the Rhins of Galloway (Stone, 1995), where much of the revision mapping and biostratigraphical investigation has been concentrated. Exposure is relatively poor throughout most of the central part of the district and, although it improves east of the River Cree, most of the Gala Group in that area is within the thermal metamorphic aureole of the Cairnsmore of Fleet granite. These factors, together with facies variation along strike, mean that the facies-based subdivision is of limited use beyond the areas in which it was defined. Apart from the Mindork Formation, which can be reliably traced along strike based on sandstone petrography, mapping within the Gala Group generally relies on tracing the tract-bounding faults by means of Moffat Shale outcrops and major features. Recognising that there is no implicit association between these divisions and the natural lithostratigraphy of the original succession, a tectono- stratigraphical nomenclature is employed for the Gala Group throughout the district.

Kilfillan Formation: Gala 1

The 'Kilfillan Formation' was first defined by Gordon (1962) as including all of the strata between the Orlock Bridge Fault and the Moffat Shale Group outcrop on the Gillespie Burn Fault. New biostratigraphical and petrographical evidence suggests, however, that this area encompasses at least three stratigraphical divisions. The formation is thus now restricted to the northernmost tract, Gala 1, between the Orlock Bridge and Stair Haven faults. A coastal section close to the farm of Kilfillan [NX 205 546], and patchy exposure inland in the west of the district, are characterised by units of very thick-bedded to massive sandstone, from a few metres to several tens of metres thick (e.g. (Figure 15)a; (Plate 3)). These are separated by thinner units of medium- to thick-bedded greywacke (Ta, Tac) with mudstone beds that commonly include thick, fine-grained greywacke laminae (Tc) or isolated ripples (Facies D2.2). An 80 m-thick slumped unit, composed of lenses and larger bedded blocks of greywacke and mudstone, all set in a mudstone matrix, is exposed in the northern part of the coastal section [NX 199 153]. Palaeocurrent data are primarily from linear grooves (e.g. (Plate 4), (Figure 16)a), which commonly lie close to, and are accentuated by, the bedding-cleavage lineation; restoration thus places them parallel to the strike of bedding. Rare flute casts suggest derivation from the north-east.

Sparse graptolite faunas have been found in both the mudstone clasts and the matrix of the slumped unit and also in dark grey to black mudstone that, in a few places within the turbidite sequence in the coastal section, forms interbedded units from several metres to 10 m thick. These faunas are not generally diagnostic although the mélange matrix yielded a diplograptid fauna including Climacograptus spp., Diplograptus modestus and fragmentary Parakidograptus acuminatus, suggesting an undetermined level in the acuminatus Biozone.

Between Newton Stewart and the Cairnsmore of Fleet granite, a thin sliver of black mudstone is taken to be Moffat Shale Group at the southern margin of the Gala 1 tract, although no fossils have been recovered from it. Exposure in the turbidite sequence to the north is dominated by massive sandstone, although locally interrupted by relatively thinly bedded strata. Medium- to thick-bedded greywacke with thin mudstone interbeds generally becomes more important towards the northern margin of the tract where a 60 m-thick unit of siltstone and mélange is exposed in a southward younging sequence in Penkiln Burn at Queen Mary's Bridge [NX 413 669]. Large greywacke lenses are included in the mélange that forms part of the northern (?basal) 30 m, and about 4 m of dark grey mudstone, overlain by greywacke, occurs at the southern margin.

Mindork Formation: Gala 2

Between the Stair Haven and Sandhead faults, the Gala 2 tract comprises the Mindork Formation, the only compositionally distinctive sandstone in the Gala Group in the district. Massive, medium to coarse-grained sandstone including a variable proportion of medium to very thick-bedded greywacke is well exposed in several inland areas. The Mindork Formation is named from Mindork Fell [NX 320 582] where approximately 1000 m of dominantly massive, petrographically distinct sandstone is well exposed; the sandstone contains pyroxene and amphibole, both within basic volcanic lithic clasts and as single mineral debris (Plate 5). This passes northwards into medium to very thick-bedded greywacke and massive sandstone around Low Mindork [NX 317 590], at least in part of more typical Gala Group petrography, although with mafic clast-bearing sandstone re-appearing to the north. The location of the bounding faults in the vicinity of Mindork has been taken at the southern and northern extent of petrographically distinct sandstone.

A small quarry on Mindork Fell [NX 3207 5827] exposes an unusual unit, 12 m or more thick, of laminated siltstone and fossiliferous dark grey mudstone (Facies E1.1 or G2.1) within the massive sandstone. This yielded a graptolite fauna of the acuminatus Biozone, including Climacograptus trifilis (Figure 17) indicating the middle part of the zone (Rickards, 1976).

The visible mafic debris in the sandstone seems to die out rapidly west of Mindork Fell, giving way to sandstone of similar facies, but of normal Gala Group petrography. Consequently there is little control over the location of the tract-bounding faults as they are extrapolated westwards towards the coast, other than fossiliferous mudstone units attributed to the Kilfillan and Sinniness formations to the north and south of Stair Haven. The southern fault is taken to intersect the coast at the southern margin of a 120 m-thick, largely sedimentary mélange. This varies from greywacke blocks in a variable proportion of mudstone matrix, commonly finely laminated with thin siltstone laminae, to less disrupted strata in which thicker beds have been deformed by boudinage, forming pinch and swell structures, and thin beds have been extended by small normal faults. The mélange passes gradationally into well-bedded greywacke near Stair Haven where there is a significant break in exposure.

North-east from Mindork Fell, exposure is poor, although Boreland Fell Quarry [NX 347 594] is worked in massive sandstone with an unusually high proportion of mafic lithic and crystal detritus. West of the River Cree, extensive exposure of massive sandstone south of Low Knockbrex and patchy exposure around Moor Park of Barr and Upper Barr are all composed of sandstone with mafic crystal detritus. East of the River Cree, the outcrop of the Gala 2 tract lies within the metamorphic aureole of the Cairnsmore of Fleet granite except for a small area around the south-west side of Larg Hill [NX 435 653]. Here again, bedding is only locally visible in medium- to thick-bedded greywacke packets within dominant massive sandstone. The mafic crystal detritus is not apparent in the sandstone east of the River Cree, probably because of alteration in the metamorphic aureole. However, Cook (1976 and in Cook and Weir, 1980), classified whole rock analyses of samples from the Gala 2 tract in the aureole as 'basic', consistent with the Mindork Formation.

Metamorphosed black mudstone of the Moffat Shale Group, in outcrops up to 100 m wide, is exposed east of Bardrochwood in three tectonic repetitions of the basal part of the Mindork Formation. About 600 m north of these faults, an 80 m thick unit of black mudstone and siltstone is exposed in the forestry track [NX 4782 6657] and in Mill Burn [NX 4750 6637]. Here, interbedded greywacke and a sedimentary contact with greywacke to the north are clearly exposed. Two other laterally discontinuous black mudstone units, between 10 and 30 m thick, are exposed near the granite contact on Blarbuies Hill.

Although the term Mindork Formation should ideally be restricted to the part of the sequence which includes mafic crystal detritus, this is not possible in practice as the distinction can only be made in thin section. A suite of eight samples collected at 200 to 300 m intervals across the width of the tract in the Mindork Fell–Low Mindork area shows mafic crystal detritus in the southern 1 km and in an outcrop of massive sandstone 500 m north-west of Low Mindork. Mafic material is not, however, apparent in samples from the intervening area around Low Mindork. A study of magnetic susceptibility (cf. Floyd and Trench, 1989) could not distinguish the two sandstone types, giving variable but locally high values between and within beds, even around Low Mindork, suggestive of very variable volcanic lithic or mafic crystal component in the sandstone. Hence the term Mindork Formation has been applied to the tract as a whole and can be defined as acuminatus Biozone greywacke, which, at least locally, is known to include mafic crystal detritus. However it may in part be equivalent to the Kilfillan Formation.

Sinniness Formation: Gala 4

The Sinniness Formation is well exposed in a 2.5 km-long coastal section between the Sandhead Fault and a 100 m- wide zone of broken and disrupted strata that marks the Gillespie Burn Fault at the Mull of Sinniness [NX 2240 5170]. The turbidite sequence is characterised by well-bedded, fine- to coarse-grained greywacke, predominantly in Ta units, some with cross-laminated tops (Tac), but with lami- nation better developed in thinner beds (Tabc, Tbc, Tc). Thickness of the greywacke beds is very variable with sporadic very thick massive sandstone beds. Thickness variation is usually random, but locally successive beds of similar thickness occur and near Laigh Sinniness several upward-thinning sequences are apparent (e.g. (Figure 15)b). Thin beds of mudstone are widely present between the greywacke beds, except in thick-bedded greywacke sequences, and the mudstone also forms thicker (2–3 m) beds that may include very thin, cross-laminated siltstone or very fine-grained sandstone beds (Facies D2.2). Evidence for soft sediment deformation is seen in a number of places ranging from slump folding of thin greywacke interbeds in mudstone to complete disruption of bedding, locally forming mélange over sequences from a few metres to 70 m thick. Palaeocurrent data ((Figure 16)b) from sole marks (grooves at a range of scales and locally well-preserved flute casts) in the Luce Bay coastal section generally show south-directed transport, although restoration is uncertain because of variable fold plunge in this tract. Some bed bases show several sets of markings in different orientations, suggesting that some were formed by non-depositional currents.

Inland, bedding is less clearly displayed and much of the exposure appears to be massive sandstone, although with rare thick (10–30 m) silty mudstone units. East of Culroy [NX 253 540] predominantly thick-bedded greywacke overlies interbedded greywacke and mudstone at the top of the Moffat Shale Group exposed in Gillespie Burn (Figure 12). East of the River Cree, exposure of greywacke in the Gala 4 tract within the thermal aureole is commonly of massive sandstone, although near Cairnsmore Farm [NX 472 640] two mudstone units reach 4 to 6 m in thickness. Nearby, almost 2 km of discontinuous exposure oblique to strike in Graddock Burn [NX 485 637] includes massive sandstone but is more commonly of medium- to thick-bedded greywacke with thin mudstone partings in places.

The thicker mudstone beds in the Luce Bay coastal exposures of the Sinniness Formation are sparingly fossiliferous; one of these [NX 2154 5214] yielded Atavograptus atavus, Coronograptus gregarius?, C. hipposideros and Glyptograptus tamariscus linearis, indicating the cyphus Biozone. This is consistent with the cyphus Biozone fauna collected from the base of the formation near Culroy.

Garheugh Formation: Gala 5

The Garheugh Formation is well exposed in a section across strike between Garheugh Port [NX 270 500] and the Gillespie Burn Fault. The greywacke is medium to very coarse grained and predominantly very thickly bedded (e.g. (Figure 15)c; (Plate 6)) to massive (Facies C2.1–B1.1), although it includes packets of medium- to thin-bedded greywacke, some with very thin mudstone partings (Facies C2.2–C2.3). Massive units are usually homogeneous although they may include irregular coarse-grained patches; thicker beds are commonly well graded (Ta), locally with pebbly bases. Cross-lamination and convolute lamination are best developed in the top of thinner beds (Tac). They are locally well developed in the unusually thick (about 500 m) succession of medium- to thick-bedded greywacke ((Figure 15)d; (Plate 7)), which is well exposed in the short coastal section at Rocks of Garheugh [NX 268 501]. Casts of linear and fluted sole marks, well displayed on the base of many beds in this section, form a tight west-south-west-trending group ((Figure 16)c) when corrected for folding.

The Rocks of Garheugh include a distinctive 10 m-thick unit of laminated siltstone and mudstone with very thin red mudstone beds and two very thin bentonite beds (Facies D2.1–E1.2). In this unit the bedding is contorted locally by slump folding, and soft sediment deformation is also evident locally in the enclosing sandstone sequence. Further evidence of early disruption occurs near the northern edge of the Gala 5 tract at Barhaskine [NX 260 536], where a few metres of intraclast rudite (Facies F2.2, see Mull of Logan Formation) are exposed.

Also at Barhaskine, a single bed, about 30 m thick, of conglomerate, varying from matrix supported to clast supported, is well exposed for about 200 m along strike. Large clasts (up to 30 cm in diameter) are dominated by well-rounded cobbles of crystalline (possibly vein) quartz and quartz-arenite of varying grain size (up to 1 mm) and with common healed fractures, and rare granodiorite. Other clasts, including black siltstone and mudstone, fine-grained laminated sandstone and a highly ferruginous lithology are up to a few centimetres in length and vary from angular to well rounded. One of the mudstone clasts yielded Ordovician brachiopod fragments. Although apparently laterally discontinuous, similar deposits occur in an analogous situation at Challochglass [NX 299 550] and near Craigeach [NX 312 561] up to 6 km along strike to the north-east, suggesting an extensive system of coarse-grained channel-fill deposits at this stratigraphical level within the Garheugh Formation.

North-east from Garheugh, thick-bedded to massive greywacke at the base of the Garheugh Formation occurs to the north of an intensely deformed and imbricated outcrop of the Moffat Shale Group. The latter extends for about 8 km along strike to the north-east, and then reappears south of Cairnsmore of Fleet in the east of the district. Here the Gala 5 tract is again dominated by massive sandstone, although medium- to thick-bedded greywacke is developed locally, for example at Craig [NX 517 622] and south-east of the Clints of Drummore [NX 537 635].

The Garheugh Formation has not yielded any graptolites and no sedimentary contact is seen between the greywacke and the underlying fossiliferous Moffat Shale Group. The youngest zone represented in the Moffat Shale Group in the Garheugh area is the triangulatus Biozone and this is taken to be the maximum age of the turbidite sequence. The younger, magnus Biozone is present at the base of the sandstone sequence in the Gala 5 tract on the Rhins of Galloway (Stone, 1995).

Gala 6

No lithostratigraphical formation name has been applied to this tract in the district although it is likely to be equivalent to the Grennan Point Formation of the Rhins of Galloway (Stone, 1995). In the district this tract is again composed largely of medium-grained massive sandstone where it is best exposed at Crouse Moor [NX 354 546] and, in the thermal aureole of the Cairnsmore of Fleet granite, around Mark [NX 507 607] and north of Pibble Hill [NX 535 605]. Thin-, medium- and thick-bedded greywacke occur locally, some with very thin mudstone beds. A millimetre-scale parallel lamination is common in metasandstone in the aureole and may reflect accentuation of a well developed, sedimentary Tb lamination.

In both of the areas where the Gala 6 tract is well exposed, the greywacke occurs to the north of outcrops of Moffat Shale Group. No graptolites have been found in the turbidite sequence but Peach and Horne (1899, pp.177–179) record a fauna from mudstone of the Moffat Shale Group exposed in the Water of Malzie south and south-east of Crailloch [NX 325 525]. Their extant material comprises good collections representing the gracilis Biozone together with Climacograptus, Dimorphograptus, Parakidograptus acuminatus and monograptid fragments indicative of an early Llandovery fauna. At Grennan Point on the Rhins of Galloway (Sheet 3) the turbidite sequence in the Gala 6 tract has yielded a convolutus Biozone fauna at its base (Stone, 1995).

Mull of Logan Formation: Gala 7

The Mull of Logan Formation, well exposed in pastureland between Alticry [NX 280 499] and Elrig [NX 322 476], shows a decrease in the proportion of massive sandstone in the younger part of the Gala Group. Tectonostratigraphically it is divided into two tracts, Gala 7a and Gala 7b, by the Laurieston Fault, although farther east this marks the junction between the Gala and Hawick groups, the Gala Group being present only to the north of the fault.

Gala 7a

Most of the turbidite sequence in the northern tract is composed of Facies C2.1–C2.2, dominated by medium- to thick-bedded, fine- to medium-grained greywacke. Mudstone beds are generally very thin but locally range up to 40 cm with silty laminae or very thin beds of fine-grained greywacke in thicker units. Greywacke beds are commonly simple graded (Ta) units, although even grading may be imperceptible in fine-grained beds, and parallel and cross-lamination may be well developed near the tops of beds (Tabc, Tac). Linear sole markings have variable orientation but are predominantly south-west-trending; one set of well developed flute casts shows current flows in this direction ((Figure 16)d), after structural correction. Medium- to coarse-grained massive sandstone occurs in sporadic, very thick beds (up to several metres) that are distributed throughout the sequence, but constitute more than 150 m of the sequence on the south-east side of Bennan Hill [NX 299 484] and a thick sequence in the northern part of the tract. Common minor disruption of bedding by ductile boudinage and mudstone injection, or by brittle fracture of greywacke beds and rotation of fragments into mudstone, are probably due to limited soft sediment deformation.

Towards the presumed top of the sequence in the northern part of the tract, two units of intraclast breccia or pebbly sandstone (Facies F2.2; (Plate 8)) occur in association with massive sandstone and mark major soft sediment slump events. The breccia is composed of angular to sub-rounded siltstone and fine- to medium-grained greywacke clasts, commonly from a few centimetres to 20 cm diameter but up to 50 cm, in a variably medium- to very coarse-grained sandstone matrix. It first appears interbedded with thin- to medium-bedded greywacke near Corwall [NX 288 494] to the north-east and passes up and along strike to the south-west into about 150 m of massive intraclast breccia. Between Corwall and Alticry this is overlain by about 350 m of massive or locally thick- to very thick-bedded, medium- to very coarse-grained sandstone that passes up into a further 200 m of intraclast breccia. The intraclast breccia and massive sandstone sequence, about 700 m thick in the western part of the district, can be matched with similar deposits in a similar structural situation within the Gala 7 tract on the Rhins of Galloway (Kelling et al., 1987, p. 800) about 14 km to the south-west. However, in the district the intraclast breccia dies out within 2 km to the north-east of Alticry although the massive sandstone can be traced to High Glenling [NX 318 513]. Near Alticry the upper intraclast breccia has a planar top, overlain by about 130 m of dominantly thin-bedded greywacke and mudstone succeeded by medium- to thick-bedded greywacke before the sequence is truncated by the Garheugh Fault.

In the eastern part of the district the turbidite sequence within the Gala 7 tract is exposed south of Creetown and south of Pibble Hill. Greywacke commonly occurs in poorly defined beds, which range from thin to thick with Ta or parallel-laminated (Tb) units. Massive sandstone crops out locally but probably represents scattered very thick massive sandstone beds rather than the continuous sequences seen farther west.

A sparse and indefinite fauna was recorded by Peach and Horne (1899, p.180) from the narrow outcrop of Moffat Shale Group at the southern margin of Gala tract 7a near Elrig Loch [NX 323 491], although this has not been relocated. However, the thin-bedded sequence overlying the intraclast breccia near Alticry has yielded limited graptolite faunas at two localities [NX 2846 5024] and [NX 2860 5032]. Initial interpretation of these suggested the turriculatus Biozone or the base of the overlying crispus Biozone. However, Loydell (written communication, 1997) identified Torquigraptus cf. proteus and Torquigraptus cf. germanicus and concluded that, whilst the faunas are too limited to enable any biostratigraphical precision, they suggest a biozone higher than guerichi.

Gala 7b

Two units of very thick-bedded to massive, medium- to very coarse-grained sandstone, both about 400 m thick, can be traced from Milton Fell [NX 315 477] to Elrig Fell [NX 332 485], although the southern unit, situated immediately north of the Mochrum Fault, starts to break down into medium- to thick-bedded greywacke to the west. They are separated by 500 to 600 m of thin- to thick-bedded greywacke and thin silty mudstone beds with a few thicker beds of massive sandstone and silty mudstone. Linear sole markings have easterly trends assuming gentle north-east-plunging folds as in Gala tract 7a. No fossils have been recovered from the Mull of Logan Formation in Gala tract 7b.

The equivalent tract east of Wigtown Bay is composed of greywacke attributed to the Cairnharrow Formation of the Hawick Group, and the few exposures on the eastern side of the Wigtown peninsula, near Braehead [NX 420 522] also show affinities with the Cairnharrow Formation. A graptolite fauna, collected during the primary survey by Maconnochie at Baldoon Mains [NX 424 535], from the Cairnharrow Formation near the northern margin of Gala tract 7b, was considered by White et al. (1991) to be of the turriculatus Biozone, similar to that from the Mull of Logan Formation at Alticry in Gala tract 7a. Consequently, the Mull of Logan Formation outcrop in the western part of the Gala tract 7b in the district was interpreted to pass stratigraphically laterally into Cairnharrow Formation west of Braehead (BGS, 1992a, b). Recent re-examination of the Baldoon Mains fauna (Loydell, written communication, 1997), however, indicates that it is somewhat older (see below and (Figure 9)), suggesting that the Gala tract 7b pinches out within the Wigtown peninsula ((Figure 1); Chapter 5).

Port Logan Formation: Gala 8

The Port Logan Formation is best exposed on the Rhins of Galloway, where it has yielded graptolites of crispus Biozone age (Stone, 1995), but strata with similar characteristics are poorly exposed between Elrig and Mochrum on the margin of a large area of continuous drift cover around Port William. The sequence is relatively thinly bedded, dominated by Facies C2.2 and C2.3, with thin- to medium-bedded fine-grained greywacke typically poorly graded (Ta) or with laminated or cross-laminated tops (Tab, Tabc, Tac). Linear sole markings are well developed and show predominantly southerly orientated palaeocurrent azimuths assuming gentle fold plunge, although one occurrence of flute casts suggests transport from the east ((Figure 16)e). Interbedded mudstone ranges from thin partings to beds 40 cm thick. Scattered small exposures of massive sandstone suggest the presence of some very thick-bedded greywacke.

Sandstone petrography and composition

Compared with the Leadhills Supergroup, sandstone petrography in the Gala Group in south-west Scotland is generally much less variable, with most being similar to the quartzofeldspathic material of the Kirkcolm, Glenwhargen and Shinnel formations. However, the Mindork Formation and, on the Rhins of Galloway, the Money Head Formation (Stone, 1995) include andesitic volcanic lithic and mafic mineral debris indicating that the pattern apparent to the north continues at least into the older parts of the Gala Group.

Quartzo-feldspathic sandstone Typically, the Gala Group sandstone (Plate 2) is composed predominantly of monomineralic grains of quartz (including vein quartz) and feldspar (plagioclase and K-feldspar). Their occurrence together in some coarse-grained polycrystalline fragments suggests that they were derived mainly from quartzite or granitic to more basic igneous rocks. Other lithic detritus comprises a variable assemblage of fine-grained igneous rocks, including common altered basic and a variety of felsic types, chert, fine-grained sedimentary rocks (locally abundant as intraclasts) and sparse metamorphic (foliated) detritus. Minor constituents include mica, common in finer grained samples, and ubiquitous tourmaline and zircon. Secondary chlorite and carbonate are present in significant amounts in some samples, replacing the matrix and some clasts. Gordon (1962) noted slight variation between the Garheugh Formation and the Kilfillan and Sinniness formations. The Garheugh Formation locally includes rare clinopyroxene not present in the older rocks although coarse-grained basic igneous rock debris is not apparent. A tendency for the quartz content to increase in younger rocks at the expense of feldspar and lithic detritus is a general characteristic of the Gala Group sandstone, although rare mafic mineral clasts occur, even in the younger sandstone.

Sandstone with volcanic debris The sandstone in the Mindork Formation exposed around Mindork Fell and to the east is similar to that described above except that it includes a variable but usually small proportion of andesitic to basaltic lithic clasts and single mineral grains of clinopyroxene, amphibole and minor amounts of olivine and epidote (Plate 5). The monomineralic grains are usually angular to subrounded with some amphibole grains in particular preserving broken euhedral crystal shapes. The mafic mineral grains vary in proportion up to about 10 per cent with clinopyroxene usually dominant. Amphibole is usually absent or minor relative to clinopyroxene but in some samples the two are present in approximately equal proportions. Similarly, igneous lithic clasts are dominantly basaltic but in samples with a significant proportion of amphibole, intermediate to felsic clasts also occur. Lithic clasts increase in proportion in coarser grained sandstone where the volcanic clasts commonly include pyroxene or amphibole and feldspar phenocrysts.

Hawick Group

Since the 'Hawick Rocks' were first described, their age, structure and stratigraphical relationships have been the subject of considerable debate. Initially assigned to the 'Lower Silurian' (now equivalent to the Ordovician and Llandovery Series of the Silurian) by Lapworth and Wilson (1871), they have subsequently been variously described as Upper Llandovery (Lapworth, 1889; Peach and Horne, 1899; Kemp and White, 1985), Wenlock (Warren, 1964) and Ludlow (Craig and Walton, 1959; Clarkson et al., 1975). Rust (1965) defined two divisions of the Hawick Group outcrop in the district (the Kirkmaiden and Carghidown beds) and discovered graptolites indicative of a late Llandovery age at Kirkmaiden. Following the resurvey of south-west Scotland, this age has been supported by new collections from some of Rust's localities and a regional review of all biostratigraphical data from the Hawick Group by White et al. (1991). This work has also led to modification of the stratigraphy of the Hawick Group. Thus, the northern part of Rust's 'Kirkmaiden Beds' is separated as the Cairnharrow Formation and the Ross Formation, formerly regarded as part of the Riccarton Group (e.g. Kemp and White, 1985), is now incorporated into the Hawick Group as described below.

All formations of the Hawick Group are characterised by very uniform sequences of thick- to thin-bedded, generally fine- to medium-grained but locally coarse-grained, greenish grey greywacke with interbedded mudstone (e.g. (Figure 18)). The sandstone is quartzo-feldspathic but with significant amounts of basic volcanic material and carbonate debris (Plate 9). The latter, together with a calcareous cement, gives rise to a thick brown crust on weathered exposures. Variations in sedimentological character at a scale of a few metres to tens of metres thickness can be described as interfingering mid-fan turbidite facies with reference to the scheme of Pickering et al. (1986, 1989). The sequence is dominated by medium- to thick-bedded greywacke with thin mudstone partings (Facies C2.1–C2.2), generally in packages a few tens of metres thick separated by units a few metres thick composed of thinly interbedded greywacke and mudstone (Facies C2.3–D2.3). Very thick-bedded, massive sandstone, either single beds or packets of a few beds (Facies B1.1) are dispersed through the sequence. This basic character is represented by the Kirkmaiden Formation ((Figure 18); (Plate 10a), (Plate 10b) and variations from it are used to define the other formations: more thickly bedded sandstone facies characterise the Cairnharrow Formation whereas the addition of red mudstone and laminated carbonaceous siltstone hemipelagite beds characterise the Carghidown and Ross formations, respectively. The well-bedded nature of the sequence is disrupted in most of the Carghidown Formation and to a lesser extent in the Ross Formation by soft sediment deformation concentrated in mélange units but affecting much of the sequence to varying degrees.

Facies C (mainly C2.1) members are made up of greywacke beds of variable thickness, usually 20 to 60 cm, although ranging from a few centimetres to 1 m thick. Groups of beds of uniform thickness or forming upward thinning sequences are locally recognisable. Mudstone partings are thin (less than 10 cm) to very thin or absent; pelagic mudstone (Te) is only very rarely developed as dark laminae. Sandstone beds have parallel sharp tops and bases and, except in slumped units, even thin beds are laterally continuous within the limits of foreshore exposure (up to 100 m along strike). Ripples, usually symmetrical with amplitudes of 2 to 3 cm and wavelengths of about 20 cm, are common and small dewatering structures (Plate 11) are preserved locally on the tops of beds. Sole marks are also common, typically as casts of linear groove and flute marks (Plate 12). Current orientations determined from sole marks and ripples (Figure 19) are commonly mutually oblique or perpendicular, with the ripples probably the result of reworking by non-depositional currents. The sandstone is usually fine to medium grained although locally includes coarse-grained detritus and mudstone intraclasts. The Bouma Ta, Tb and Tc divisions are variably developed, usually alone or in Tab Tacand Tbc combinations, where the lowest division characteristically constitutes most of the bed thickness.

Facies D (principally Facies D2.3 but transitional to Facies C2.3) is dominated by mudstone, commonly finely laminated, with a variable proportion of very thinly bedded (1 cm) siltstone and beds, up to 5 cm thick, of very fine-grained sandstone. Individual greywacke beds are parallel-sided and laterally continuous Tc units. Facies D members range in thickness from 0.10 to 10 m (typically 2 to 3 m) and locally form the upper part of upward thinning and fining sequences. Trace fossils are abundant in places (e.g. Benton, 1982).

Facies B (transitional between Facies B2.1 and Facies C2.1) members are composed of very thick-bedded, massive or parallel-laminated, weakly graded sandstone beds that range up to several metres in thickness. The sandstone is typically medium grained although it commonly includes coarse-grained detritus. Some of the thick greywacke beds are erosive, cutting sharply down into the underlying beds and thinning rapidly. Others are laterally more continuous and can be traced for several hundreds of metres along strike through discontinuous exposure.

Cairnharrow Formation

The Cairnharrow Formation comprises the northern part of the 'Kirkmaiden Beds' as defined by Rust (1965). It crops out in three fault-bounded tracts and is best exposed in the hills to the west and north of Gatehouse of Fleet, particularly around Cairnharrow [NX 533 561], but is also exposed around Monreith [NX 359 411] in the west of the district. Thin slivers of Moffat Shale Group are poorly exposed beneath the formation in the Kirkcudbright district (Lintern and Floyd, 2000).

As in other parts of the Hawick Group, the Cairnharrow Formation is dominated by turbidites of Facies C ((Figure 18)a). It is distinguished by relatively thickly bedded sandstone with well developed parallel-lamination (Tb) throughout the bed thickness, grading to a thin cross-laminated layer near the top. Thin-bedded units (facies D2.2 or C2.3) range from tens of centimetres up to 40 m thick. Massive sandstone units occur in the better exposed sequence of the southernmost tract west of Gatehouse of Fleet where palaeocurrent data show dominant transport from the north-east.

Graptolites have been found at only three places within the Cairnharrow Formation in the district. Peach and Horne (1899, p.215) reported graptolites, collected during the primary survey by Maconnochie, at two localities in 'occasional dark seams in flags, shales and greywackes' — in railway cuttings near Whaup Hill [NX 405 502], and at Baldoon Mains [NX 424 535]. A search for these localities was unsuccessful, the railway cuttings being long since disused and overgrown. However, the small number of specimens preserved in the collections include one identified as Stimulograptus?halli from Whaup Hill, suggested as indicative of the turriculatus Biozone sensu lato (White et al., 1991); this biozone has since been divided into two, an older guerichi and a younger turriculatus biozone (Loydell et al., 1993). Having re-examined the faunas collected by Maconnochie, Loydell (written communication, 1997) confirmed the presence of Stimulograptus halli at Whaup Hill. However, with Streptograptus plumosus and fragments of Torquigraptus planus or T. cavei at both localities, the age is upper guerichi Biozone (with Baldoon Mains possibly being lowermost turriculatus Biozone).

Graptolite collections from three localities in the Cairnharrow Formation in the adjacent Kirkcudbright district were considered to be of either turriculatus or crispus Biozone age (White et al., 1991; Lintern and Floyd, 2000). However, Loydell (written communication, 1998) confirmed that these faunas represent the guerichi Biozone. Old collections from Trowdale Glen [NX 761?682] and Tarff Glen [NX 674?613] both include fragments of Stimulograptus halli but Torquigraptus?minutus with Petalolithus?hispanicus in the former and T.?cavei in the latter indicate the lower and upper parts of the biozone respectively. Poorly preserved graptolite fragments from Edgarton [NX 672?631] include Paradiversograptus runcinatus, indicating the guerichi Biozone.

The boundary between the Cairnharrow and Kirkmaiden formations in the district is taken for convenience at the Innerwell Fault because no other relationship is demonstrable. However, they are composed of similar turbidite facies and the Cairnharrow Formation probably passes gradationally up and/or laterally into the Kirkmaiden Formation (Figure 20).

Kirkmaiden Formation

The Kirkmaiden Formation comprises the southern part of the 'Kirkmaiden Beds' of Rust (1965). It is well exposed in coastal sections south of Back Bay [NX 365 394] on the east side of Luce Bay, in Wigtown Bay and in Fleet Bay (Weir, 1974). As described above, this formation represents the essential character of the Hawick Group without additional components. It is dominated by Facies C2.1–C2.2 (e.g. (Figure 18)b, c; (Plate 10a), (Plate 10b)); Facies C2.3–D2.3 units are common and usually less than 2 m thick, and Facies B units are rare — thickly bedded greywackes in packets up to about 10 m thick. Palaeocurrent data ((Figure 19)a) show sediment transport from the north-east.

Rust (1965) reported several graptolite localities in a short section of the coast near Monreith, between Kirkmaiden [NX 358 407] and Cairndoon [NX 375 390], the faunas being considered by Strachan (in Rust, 1965) to be indicative of a late Llandovery age. These localities lie close to the Innerwell Fault in this area (Figure 1), although the precise location and width of the fault zone are uncertain. One strand of the fault is likely to pass immediately north of the steeply plunging folds at Black Rocks (Chapter 5) and all of the graptolite localities are thus likely to be in the Kirkmaiden Formation. Identifiable graptolites were collected from Rust's southern localities, near Cairndoon [around 372 392], during the resurvey but the northern localities near Kirkmaiden could not be relocated. Graptolites are also available from one locality in the Kirkmaiden Formation in the adjacent Kirkcudbright district (Lintern and Floyd, 2000).

At Cairndoon, thin black mudstone laminae in the grey mudstone between greywacke beds contain sparse graptolite faunas, generally poorly preserved, and completely unidentifiable in some of the bands. However, slender monoclimacids of vomerina type were first considered to indicate a late Llandovery age by Strachan (in Rust, 1965). New collections made during the resurvey include a number of examples of Monoclimacis priodon and Mcl. griestoniensis,and wereconsidered by White (in Barnes, 1989 and White et al., 1991) to represent the upper part of the griestoniensis Biozone. However, Loydell (written communication, 1998) also recognised Mcl. woodae and one proximal end of Mcl. crenulata, suggesting the lower part of the crenulata Biozone.

In the Kirkcudbright district, dark mudstone laminae in the Kirkmaiden Formation exposed in a small quarry south-west of Waterside [NX 665?587] yielded graptolite fragments considered by White et al. (1991) to represent a single example of Monograptus cf. tullbergi, a species recorded elsewhere from the griestoniensis and crenulata biozones. Loydell (written communication, 1998) determined this specimen to be of Torquigraptus, possibly the species pragensis, sensu lato, or minutus, but it is too fragmentary to be identified with confidence and thus biostratigraphically it is of little value.

/p>

The boundary between the Kirkmaiden and Carghidown formations is drawn at the northern limit of the occurrence of red mudstone beds. In Wigtownshire there is no evidence of a major structural break between them and so the boundary there is interpreted as a conformable upward transition within a fault bounded tract, from Carghidown Formation into Kirkmaiden Formation, as the red mudstone beds die out up sequence. However, since the Carghidown Formation in tracts to the south is apparently of the same age or younger than the Kirkmaiden Formation (Figure 20) there is still uncertainty over the stratigraphical relationship.

Carghidown Formation

The Carghidown Formation is well exposed in almost continuous sections on both sides of the southern part of the Wigtown peninsula, being named from a locality [NX 435 350] on the west coast by Rust (1965). In the northern part of the outcrop, the Carghidown Formation is generally very similar in appearance to the Kirkmaiden Formation. It is dominated by Facies C with subordinate Facies D members generally less than 2 m thick; Facies B occurs sporadically but is less common than in the Kirkmaiden Formation. The Carghidown Formation is distinguished by red mudstone interbedded with the ubiquitous grey-green mudstone, particularly in Facies D (Plate 13). Distinctive red mica flakes, also present in the transitional part of the Kirkmaiden Formation, are commonplace in sandstone in the Carghidown Formation. South of the Glasserton Thrust, bedding has been variably disrupted by soft sediment deformation and intraformational mélange units are common. Palaeocurrent data ((Figure 19)b) show a wide range (probably due at least in part to complex folding making restoration difficult) although the predominant trend is from east to west.

The red mudstone beds are thin and rare in the northern part of the outcrop but their maximum thickness increases up to 6 m as they become more common southwards. They are laterally persistent within the extent of the exposures. The red mudstone appears structureless, although bed margins are diffuse (commonly mottled) or interlaminated with the grey-green mudstone. A primary origin for the red colouration is indicated by the occurrence of patches of red mudstone mixed with green in the matrix of slumped units (e.g. north of Chapel Port West, Isle of Whithorn [NX 478 361]).

Soft sediment deformation is widespread in the southern part of the Carghidown Formation in sequences up to several tens of metres thick, the transition from well-bedded strata coinciding with a marked change in structural character (see Chapter 5). Disruption varies from incipient pinch and swell of otherwise continuous bedding to disrupted zones up to 30 m thick in which bedding has completely disaggregated into lenses and irregular blocks of sandstone of varying size and density in a mudstone matrix. Small folds, many of which may have originated as slump structures, characteristically have a wide range of axial orientations on bedding-parallel axial surfaces. The deformation of many of the disrupted zones has clearly been accentuated by tectonic effects, but gradational boundaries with coherent strata, blocks lying across the matrix foliation, and variations in matrix lithology all suggest earlier, soft sediment disruption. Downward-facing tectonic folds in inverted coherent units between disrupted zones (Lintern and Floyd 2000) suggest that in some cases the mélange units mark the sheared hinge zones of large-scale slump folds.

Relationships with the Kirkmaiden and Ross formations suggest that the Carghidown Formation is of late Llandovery to earliest Wenlock age, although no macrofossils have been found in the Carghidown Formation in the district or adjacent areas. This age assignment is supported by microfossil occurrences outwith the district and two graptolite localities along strike to the north-east near Eskdalemuir which have yielded faunas assigned to the Monoclimacis crenulata Biozone (White et al., 1991).

The nature of the junction between the Carghidown and Ross formations has long been the subject of debate as to whether it is a gradational sedimentary change (e.g. Clarkson et al., 1975; Barnes, 1989; Lintern and Floyd, 2000) or a tectonic break (e.g. Rust, 1965; Kemp, 1986). In the district, the junction is exposed at Burrow Head in an unusually broad belt of southward-younging strata within which the Ross Formation occurs to the south of the Carghidown Formation. The southernmost red mudstone bed is separated from the northern hemipelagite bed of the Ross Formation by about 30 m that includes a 6 m zone deformed by folds and strike-parallel faults. Rust (1965) proposed a major fault at this locality but the deformed zone may equally well be a relatively minor structure in an otherwise conformable sequence. Red mica grains in a few sandstone beds south of the deformed zone, but otherwise not seen in the Ross Formation, suggest a gradational junction. A transitional facies, including both red mudstone and hemipelagite beds, has been described from the only other coastal exposure of this junction in Kirkcudbrightshire (Clarkson et al., 1975; Lintern and Floyd, 2000), supporting a stratigraphical transition.

Ross Formation

Formerly regarded as part of the Riccarton Group (Craig and Walton, 1959), the Ross Formation has been incorporated into the Hawick Group because of the close similarity between their dominant turbidite facies (White et al., 1991). A major structural break, coincident with a marked change in structural style and metamorphic grade occurs farther south at the boundary between the Ross and the Raeberry Castle formations (Lintern and Floyd, 2000).

Occurrences of fossiliferous dark grey, carbonaceous siltstone beds define outcrops of the Ross Formation in the district at Burrow Head and in a tectonically complex zone in the southern part of the Isle of Whithorn. The formation, well exposed in the coastal sections at Burrow Head, is dominated by Facies C2.1–C2.2, in units up to 50 m thick and commonly of uniform bed thickness, separated by more thinly bedded Facies C2.3–D2.3 units. The siltstone occurs in beds ranging from a few centimetres to 1 m in thickness, and is characterised by thin, alternately light and dark parallel laminae except where disrupted by bioturbation. The siltstone beds are sparsely distributed and usually interbedded, with sharp contacts, in mudstone in Facies C2.3–D2.3.

Graptolite faunas from the carbonaceous siltstone beds indicate an early Wenlock age (Whiteet al., 1991). Cyrtograptus centrifugus, Monoclimacis vomerina vomerina, Monoclimacis vomerina cf. basilica and Monograptus cf.danbyi, from a cliff-top exposure at Burrow Head [NX 4538?3411], demonstrate the presence of the earliest Wenlock centrifugus Biozone. This zone was also recognised on the Isle of Whithorn e.g. [NX 4800?3597]. At Burrow Head the murchisoni Biozone may be present at the base of the cliff [NX 4633?3437] where Monograptus priodon and Monoclimacis spp. are associated with a cyrtograptid with a high proximal thecal count which may be Cyrtograptus murchisoni. A nearby cliff-top exposure [NX 4633?3437] yielded Monograptus cf. riccartonensis Lapworth, indicating the riccartonensis Biozone.

Sandstone

Petrographically, the Hawick Group sandstone is remarkably uniform throughout the sequence. It is mostly fine- to medium-grained, poorly sorted lithic greywacke, consisting of angular to subrounded sand grains and up to 40 per cent silt-grade matrix, although some coarser grained sandstone tends to be better sorted. The sand fraction is dominated by quartz with significant amounts of feldspar, carbonate, lithic fragments and mica, relative abundances depending to some extent on grain size. Feldspar, generally comprising about 10 per cent of the sand fraction, is mainly plagioclase with some potassic feldspar. Carbonate forms up to 15 per cent of the rock, and probably originated largely as detrital grains, some of which are still recognisable although this mineral has been extensively recrystallised and may have partially replaced other grains. The proportion of carbonate is particularly variable in the Cairnharrow Formation and in northern parts of the Kirkmaiden Formation (around Fleet Bay), where it is low in some beds.

Lithic detritus, varying from about 15 per cent in medium-grained sandstone to the dominant part of the sand fraction in very coarse-grained sandstone, consists of volcanic and sedimentary rocks, polycrystalline quartz and quartzo-feldspathic material. Volcanic detritus is generally basic to intermediate in composition and highly altered. Acid rock types are only rarely recognisable, other than as relatively coarse-grained quartzo-feldspathic material. Sedimentary rock types include intraclasts but also siltstone that was cleaved prior to incorporation. Red- or pink-stained cherty material is common in some samples from the Carghidown Formation.

Mica varies in abundance inversely with grain size within beds, from 3 per cent or less in medium-grained sandstone at the base up to 15 per cent in fine-grained sandstone in the upper parts. 'Red mica', commonly visible in hand specimen, is a minor constituent of the sandstone in southern exposures of the Kirkmaiden Formation, throughout the Carghidown Formation and locally within the Ross Formation. However, even within individual beds, its occurrence may be laterally discontinuous on a scale of several metres. The colour is due to a hematite coating on the mica flakes and may be associated with patchy hematite staining of the rock.

Accessory minerals include green tourmaline, zircon and garnet. Ubiquitous scattered opaque minerals probably include detrital and diagenetic materials.

Sandstone whole rock and mineral composition — source evolution

Variations in the petrography of the sandstones that dominate the Lower Palaeozoic succession in the Southern Uplands, described above, have been known since the work of Walton (1955), Kelling (1961, 1962), Gordon (1962) and Floyd (1976, 1982). These and subsequent studies demonstrated that the petrographically distinctive Ordovician sandstones may be closely interbedded but also that they may have different palaeocurrent directions suggesting not only different source rocks but possibly different source terranes (e.g. Stone et al., 1987; Stone, 1995; Floyd, 1999 and references therein). All of this work underpins the understanding of the Ordovician and locally the Silurian sandstone stratigraphy as described above. Recent studies of detrital pyroxene chemistry (e.g. Styles et al., 1989, 1995) and an extensive suite of whole rock analyses (Duller and Floyd, 1995) are dominated by data from the Ordovician sandstones. However, varying amounts of data are available from most parts of the succession allowing a different perspective on the sandstone composition and its implications for correlation and source evolution.

Pyroxene composition

Clinopyroxene is found as single grains and within volcanic and plutonic andesitic lithoclasts in the Galdenoch, Portpatrick and Mindork formations and defines two compositional groups (Styles et al., 1989, 1995). Those from the oldest unit, the Galdenoch Formation, generally represent material derived from a tholeiitic volcanic arc source although one member includes material transitional towards a calc-alkaline composition. The material from the younger formations shows a more complex signature, possibly due to a mixture of tholeiitic and calc-alkaline components. Parts of the Ballantrae Complex contain andesitic rocks with appropriate pyroxene composition to have been the source for some of the material in the Galdenoch Formation. However, no suitable provenance for the more calc-alkaline material can be readily identified. Styles et al. (1995) suggest that it may originate from one or more volcanic arc sequences of the type preserved beneath the Caradoc black shale (equivalent to the lower part of the Moffat Shale Group) in central Newfoundland (e.g. Colman-Sadd et al., 1992). Material of this nature is also seen more locally in the Southern Uplands in the Crawford Group and may contain pyroxene of appropriate composition. For example, dolerite and basalt exposed north of Glenluce in the Gabsnout Burn area (see p. 25) contain pyroxene of similar composition to that which occurs in the immediately overlying Portpatrick Formation. However, the dominant detrital lithic component with recognisable phenocrysts is andesitic rather than basic in composition

Sandstone composition

The bulk chemical composition of a sandstone reflects, at least in part, the detrital material contained within it, not only as fragments visible petrographically but also as material too fine grained or too altered to be distinguished. It is thus a complex interplay of a range of variables relating to provenance (clast content), transport and deposition, diagenesis/metamorphism and weathering (e.g. Bhatia, 1983). Many of these variables may themselves be strongly influenced by the tectonic situation of the source or depositional environment. Consequently, sandstone composition has been utilised to identify tectonic setting (e.g. Pettijohn et al., 1972; Blatt et al., 1980) in a similar way to the more established geochemical investigation of igneous rocks (e.g. Rollinson, 1993). However, geochemical composition also provides a convenient means of characterising sandstones simply for the purposes of comparison and correlation, particularly as it may be less susceptible to distortion through variation in grain size and weathering than petrographical methods.

Characteristic geochemical data for the Lower Palaeozoic sandstones within the district are presented in Appendix 5. Graphical summary (e.g. (Figure 21)) of the whole Southern Uplands sandstone geochemistry dataset (Duller and Floyd, 1995) defines the abundance trends of each element in progressively younger formations across strike. These highlight both complementary and different trends to those derived from the traditional petrographical approach to sandstone composition (e.g. Kelling, 1961; Floyd, 1982). Elemental abundances from the northern part of the Southern Uplands have irregular or saw-toothed profiles consistent with interdigitating volcanic-rich and quartzo-feldspathic greywackes. In the younger Ordovician rocks (the Portpatrick and Shinnel formations) and the Silurian rocks, smoother profiles dominate and imply a more gradual evolution of the detrital input from volcanic-rich debris in the north to quartzo-feldspathic debris in the south. In south-west Scotland, mafic input (e.g. represented by MgO, (Figure 21)b) increases southwards to peak in the Shinnel Formation. A subsidiary peak appears in the middle of the Gala Group sequence, followed by a progressive decline in younger Gala tracts. A dramatic compositional change is apparent with the introduction of large amounts of carbonate into the Hawick Group (diluting the SiO2 content; (Figure 21)a), although the difference between this and the younger parts of the Gala Group may otherwise be less significant. The parallel increase in MgO suggests that some of the carbonate may be dolomitic but this is difficult to separate from the potential contribution from spilitic volcanic material, which is seen petrographically to increase significantly in these younger rocks. In general, however, the Hawick Group formations have remarkably uniform composition relative to the older rocks. It should be noted that rocks originally correlated with the Cairnharrow Formation on the Rhins of Galloway (Stone, 1995) are compositionally part of the Gala Group, consistent with their subsequent re-assignment as Gala tract 9 (McCurry and Stone, 1996).

Implications for provenance interpretation

Whole rock analyses for the Southern Uplands sandstones indicate a progressive evolution of detrital supply with the distinction between volcanic-rich and quartzo-feldspathic sequences being less apparent than from petrographical studies. This difference may be due to the fact that relatively unstable materials, notably the volcanic and mafic components, will have variable preservation potential and will be difficult to recognise petrographically in an altered or recrystallised state, or when finely comminuted in the matrix of a greywacke. Alternatively, minor source components may be masked and difficult to detect by whole rock analyses, which are also unable to differentiate between contrasting mineral assemblages that might provide similar element abundances from different proportions of host grains.

The sandstone geochemical data cast some doubt on the evidence for compositional distinctness in the Southern Uplands greywacke sequence: the apparent southern and south-western derivation of 'volcanic-rich' sandstone as compared with the northerly or north-easterly derivation of more 'quartzo-feldspathic' sequences, from volcanic arc and basement terranes respectively (e.g. Stone et al., 1987). It appears more likely that both components were continuously available in the Ordovician and early Llandovery, with input of the volcanic component then waning during the Llandovery. Most elements show less variation within the youngest parts of the Gala Group and in the Hawick Group than in older sequences ((Figure 21)). This is consistent with the lack of compositional variation between Hawick Group formations, suggesting that much of the debris may have been derived by the reworking of older sedimentary rocks. The Hawick Group is, however, distinguished from the Gala Group by the appearance in Hawick Group greywackes of bioclastic carbonate, and possibly by an increase in the detrital spilitic volcanic content.

Chapter 5 Structure of the Lower Palaeozoic rocks

The tectonostratigraphical configuration of the Southern Uplands terrane strongly suggests that the Ordovician and Silurian sedimentary rocks were deformed during or soon after their deposition at the northern margin of the Iapetus Ocean (e.g. Leggett et al., 1979; Stone et al., 1987). The tectonic regime was dominated by north-west–south-east shortening as the ocean was closed by subduction. Initially the shortening was orthogonal but there was an increasing component of sinistral shear (Barnes et al., 1989) as the terrane was progressively deformed between the leading edges of the northern, Laurentian, and southern, Avalonian, continents. This period of tectonic activity, part of a southward migrating sequence of events broadly termed 'Caledonian', produced the steep dips, cleavage and locally complex folding, faulting and regional metamorphism of the Southern Uplands terrane. Several intrusive swarms were emplaced contemporaneously with deformation during the late Silurian, culminating in the early Devonian emplacement of the larger, essentially post-tectonic granitic plutons and superimposition of a thermal metamorphic overprint on their host rocks (Figure 22). With younger deposits being preserved only offshore, evidence of the post-Caledonian tectonic history of the district (Chapter 8) stems largely from adjacent areas (e.g. Chadwick et al., 1994, 1995; Stone, 1995; Lintern and Floyd, 2000).

Evolution of the regional structural model

During the primary geological survey of the Southern Uplands in the 1870s, difficulties were experienced in interpreting the complex folding apparent in the coastal sections whilst the absence of a clear stratigraphical framework allowed little appreciation of the overall structure of the region. Regional biostratigraphical correlation based on graptolite faunas (Lapworth, 1889) led to better understanding of the stratigraphy and formed the basis for a large-scale anticlinorium/synclinorium structural model (Peach and Horne, 1899) within which isoclinal folding brought the Moffat Shale to outcrop as inliers in anticline cores and repeated the overlying sandstone sequence on either side of the hinges.

Applying a then-new technique utilizing sedimentary way-up evidence, Craig and Walton (1959) demonstrated that the Hawick Group strata in the Kirkcudbright area were predominantly northward-younging, inconsistent with the anticlinorium/synclinorium model. Furthermore, recognition that the detrital components of the sandstone may be markedly different on either side of the main shale outcrops emphasised the lack of stratigraphical continuity (e.g. Kelling, 1961; Floyd, 1982). However, the terrane-wide, south-east-younging trend was confirmed, with the oldest rocks cropping out in the north-west and progressively younger strata appearing southwards (e.g. Toghill, 1970). Consequently, a large-scale structural model comprising a series of strike-parallel, fault-bounded tracts within which strata were dominantly north-younging but which become progressively younger to the south-east was invoked to explain the relationships (e.g. McKerrow et al., 1977). Development of the detailed structural model by Rust (1965), Weir (1968) and Cook and Weir (1979) in different parts of south-west Scotland, led to interpretation of up to five phases of folding. However, Stringer and Treagus (1980, 1981) suggested that the structure was actually relatively simple, with only two main phases of deformation, a view supported by more recent work (e.g. Knipe and Needham, 1986; Barnes et al., 1987, 1989; Kemp, 1987; Needham, 1993; Lintern and Floyd, 2000).

Regional structural pattern

The Ordovician and Silurian turbidite sequences of the Southern Uplands are typically steeply dipping to vertical, north-east- to east-north-east-striking, and generally young northwards in a series of fault-bounded tracts (Figure 3). In northern and central parts of the Southern Uplands, including the northern part of this district, the bounding fault-traces are marked by discontinuous slivers of the thin, fossiliferous Moffat Shale Group (Chapter 4) which is commonly preserved in stratigraphical continuity beneath the turbidite sequence. In some tracts the minimum biostratigraphical age of the Moffat Shale Group is the only indication of the maximum age of the turbidite sequence. Where faunas are available from within a turbidite sequence, they usually represent a single biozone, either equivalent to the youngest fauna within the underlying Moffat Shale Group or one biozone younger. Such biostratigraphical evidence defines a sequential decrease in the age of the turbidite sequence from north to south across the Southern Uplands (Figure 9). The change at the base of the sequence from tract to tract may simply be a result of progradation and thus may have no particular tectonic significance. However, the parallel temporal curtailment of the sequence in each tract suggests the tectonic mechanism was likely to have been southwards propagation of the thrust front into progressively younger strata contemporaneously with continuing deposition to the south (e.g. Leggett et al., 1979; Stone et al., 1987). This progressive accretion of new material may have been responsible for much of the rotation of the system into its present, near-vertical orientation, subsequently completed by collisional processes as the Iapetus Ocean finally closed.

Regional deformation (D1)

Early movement on thrusts, some of which eventually became the tract-bounding faults now recognised, propagated at a low angle to stratigraphy and was probably associated with the only phase of ductile deformation (D1) to have affected many of the rocks in the Southern Uplands. From the biostratigraphical arguments above, it is likely that this was diachronous, becoming younger southwards (Figure 22). Folds, typically gently plunging and tight to isoclinal, were developed very variably. Across-strike, highly folded zones occur interspersed with long homoclinal sections, usually of steeply inclined north-younging strata. This variation in structural style is, at least in part, related to the nature of the strata, with thickly bedded, massive greywacke less likely to be intensely folded than more thinly bedded strata. Slickensides or slickenfibres in thin veins along bedding surfaces are perpendicular to the fold axial orientation, and so demonstrate early fold growth by flexural slip, although they were themselves folded in the later stages of fold development. Individual tectonostratigraphical tracts are often marked by subtle variations in the style, orientation and intensity of D1 folding (e.g. Barnes et al., 1989 and below). D1 deformation was particularly intense in the generally finer grained, more calcareous rocks of the Hawick Group in the southern part of the Southern Uplands, where it is associated with the highest grades of regional metamorphism apparent in the terrane.

D1 is apparent regionally as a penetrative slaty cleavage (S1). In northern and central parts of the Southern Uplands, S1 is best developed in the fine-grained, muddy lithologies, although even this can be quite weak. It is only locally apparent macroscopically in sandstone and is generally represented only as a rough anastomosing fabric apparent in thin section. In parts of the Hawick Group, however, the foliation is more pervasive and is well developed in sandstone, where it is strongly refracted through graded beds, and locally also in felsic and lamprophyre dykes. A pervasive foliation of similar orientation to S1 is also present in all of the sedimentary lithologies in the Moniaive Shear Zone, although its association with S1 is unclear (see p. 58).

Cleavage tends to be congruous with the D1 folds, but may vary either in dip or strike from truly axial planar. Other than in the immediate vicinity of fold hinge zones, significant variation in the dip of the cleavage from that of the fold axial surface may cause bedding to be downward facing. This is particularly apparent in overturned, south-dipping beds where cleavage commonly dips more steeply than bedding. As a consequence of this effect, the assessment of way-up or vergence from bedding-cleavage relationships is generally unreliable in the Southern Uplands. Some downward-facing folds are apparent locally in the southern parts of the Hawick Group (see p. 62). Typically, the S1 cleavage plane contains the fold axial orientation in the northern part of the Southern Uplands but locally transects the fold axis clockwise by up to 20° in central parts. This effect is commonplace throughout the Hawick Group due to systematic variation between the strike of the cleavage and that of the fold axial surface (cf. Anderson, 1987). The cleavage transection has been explained in various ways to result from the evolution of the D1 stress system (Stringer and Treagus, 1980; Gray, 1981; Sanderson et al., 1985).

Post- D1 deformation

The extent and character of post- D1 deformation varies widely across the Southern Uplands. In northern and central parts, post- D1 structures coaxial with D1 tend to occur only very locally and are difficult to correlate. To the south, in central and southern parts of the Hawick Group outcrop, coaxial post- D1 deformation is widespread. Minor to mesoscale, upright to inclined and recumbent D2 folds are commonly seen to refold D1 structures. The recumbent folds are associated with a widely developed, gently dipping, S2 crenulation cleavage. In contrast, steeply plunging sinistral folds are developed locally throughout the Southern Uplands, usually in narrow zones adjacent to tract-bounding faults and therefore probably associated with reactivation of these structures (e.g. Barnes et al., 1995a). The relationships of these folds to D2 are ambiguous (e.g. Barnes et al., 1989; Stone, 1995), suggesting episodes of sinistral shear interspersed with refolding co-axial with D1 at various times.

The Moniaive Shear Zone (Phillips, 1994; Barnes et al., 1995a; Phillips et al., 1995b), named after the area around Moniaive, north-east of the Cairnsmore of Fleet granite, is a zone of high strain kinematically similar to, but much wider than, the narrow shear zones associated with tract- bounding faults. It has been recognised over a strike length of about 100 km through the central part of the Southern Uplands. Other than in the immediate vicinity of the granite, the Moniaive Shear Zone is up to 5 km wide, truncated to the north by the Orlock Bridge Fault and dying out southwards within the northern tracts of the Gala Group. It is characterised by the intermittent development of a pervasive foliation near-parallel to bedding, locally with a strong linear component and commonly overprinting all original structure. Strain within the zone is very variable (e.g. Phillips, 1992, 1994) but a variety of kinematic indicators consistently show a sinistral sense of shear. Because the fabric in the high strain zone is subparallel to the regional, relatively weak S1 cleavage, the two can only rarely be differentiated and unequivocal relative age relationships are difficult to establish. Close constraint on the final part of the development of the Moniave Shear Zone is provided by relationships in the thermal metamorphic aureole of the early Devonian (about 395 Ma, see Appendix 7) Cairnsmore of Fleet pluton. Cordierite porphyroblasts, widely distributed throughout the aureole, are deformed by the shear zone foliation. However, the foliation is generally overprinted by the biotite hornfelsing and later stages of the thermal metamorphism. Relatively high grades of regional metamorphism in the zone (see below) indicate that it formed at depth. Barnes et al. (1995a) suggested that the Moniaive Shear Zone is a composite feature, representing progressive but intermittent sinistral deformation over a long time period from its initiation during D1, possibly in the early Silurian, until the early Devonian.

Large-scale tract structure of the district

The model of the large-scale structure of a map area, based on interpretation of data from the area itself or drawing on more regional information, is crucial to the way in which any geological map is drawn. The structural model of the district is based largely on regional considerations of the geology of the Southern Uplands as discussed above. Laterally continuous, steeply dipping, east-north-east-trending (strike-parallel) faults spaced at 1 to 5 km are interpreted to define a series of near-vertical tracts. Typically, strata in each tract young northwards from Moffat Shale Group, discontinuously preserved at the southern margin, through the stratigraphically overlying turbidite sequence. This is substantiated within the district by biostratigraphical evidence, the disposition of outcrop of the Moffat Shale Group and evidence of a sedimentary transition into the dominantly northward younging turbidite sequence. Faults are rarely exposed, but have been located by lithological, biostratigraphical and/or structural considerations. Where well constrained, individual tract bounding structures typically comprise zones of closely spaced, anastomosing strands; otherwise they tend to be represented as a single straight strand.

Outcrops of the Moffat Shale Group in the north and east of the district locally define all of the tract-bounding faults mapped in the Leadhills Supergroup and Gala Group outcrops. However, many of the mudstone outcrops lie entirely within the thermal aureole of the Cairnsmore of Fleet granite and, along with the sandstone sequences, yield no useful fossils, so making along-strike correlation difficult. Consequently, most of the biostratigraphical data stem from the western part of the district where only the Morroch Bay, Gillespie Burn, Garheugh and Laurieston faults are associated with outcrops of the Moffat Shale Group. Other tracts within the Gala Group are defined by changes in the biostratigraphical age, composition (Gala tract 2) or sedimentary facies of the sandstone sequence. The trace of the Morroch Bay Fault is well constrained by closely spaced Moffat Shale Group outcrops and its situation at the southern boundary of the petrographically distinct sandstone of the Portpatrick Formation. To the south, however, extrapolation and correlation of other tract-bounding faults across the district is less reliable because of the lack of biostratigraphical constraints in the east of the district and the more uniform sandstone composition and facies.

The Laurieston Fault is extrapolated from the Kirkcudbright district, where it is associated with Moffat Shale Group (Lintern and Floyd, 2000), via the Kirkmabreck dyke [NX 480 565] to the small outcrop of Moffat Shale Group in the vicinity of Elrig Loch [NX 325 493] (Irvine, 1878). West of Wigtown Bay, the Laurieston Fault is taken as the boundary between the Hawick and Gala group sandstone sequences, although a precise boundary is difficult to define. An overlap in the biostratigraphical age of the Gala and Hawick groups is now apparent from revised interpretation of the limited data available in this area (Chapter 4; (Figure 9)) and the lateral transition previously interpreted between the two groups (BGS, 1992a, b) is probably a tectonic boundary, effectively pinching out the Gala 7b tract (Figure 1).

The limited biostratigraphical control on the age of the Hawick Group in the district is sufficient only to confirm the Southern Uplands paradox of overall southward decreasing age in sequences which individually young northwards from sedimentary way-up and fold vergence. This requires that strike-parallel faults are present, but their identification is difficult without evidence from outcrops of Moffat Shale Group. The Innerwell Fault is relatively well constrained because it is associated with a distinctive zone of post- D1 deformation, although, as with other faults, the precise location of its trace is locally uncertain. To the north, the Cambret and Cairnholy faults are interpreted to follow major, strike-parallel topographical features, the former including two large granodiorite dykes. The Cairnholy Fault is correlated with a structure marked by Moffat Shale in the Kirkcudbright district (Lintern and Floyd 2000). The Garlieston Fault is defined from the detailed structure of the coastal section north of Eggerness Point.

Clearly, there is uncertainty in the location and extrapolation or correlation of tract-bounding faults across the district. Their resolution and detail depend on the nature of the succession on either side of the fault and the amount of available exposure. Locally, where associated with well-exposed outcrops of the Moffat Shale Group, the high level of biostratigraphical resolution of the condensed stratigraphy and repetition of the base of the overlying sandstone succession allow definition of very closely spaced, anastomosing fault systems. This degree of tectonic imbrication may be the result of the relative incompetence of the mudrocks and the structural situation, but higher stratigraphical levels within each tract may be significantly more imbricated than can be demonstrated. For example, biostratigraphical control in the Ross Formation turbidite sequence east of the district in Kirkcudbright Bay shows at least six thrust slices in a 2.5 km section, from as little as 200 m thick (Kemp and White, 1985; Lintern and Floyd, 2000, fig. 6.4d). In the west of the district, closely spaced (from a few hundred metres to 2 km) strike-parallel faults can be defined locally in the Gala Group outcrop and through much of the Hawick Group outcrop. Unfortunately, there is no other evidence by which to substantiate or extrapolate tectonic subdivision to this extent, even in these areas. However, it is likely that the large-scale strike-parallel fault structure mapped throughout the district is oversimplified. The lateral persistence of subparallel tract-bounding structures over long distances is particularly questionable. An anastomosing system, as seen locally at the smaller scale, is more probable, if unresolvable.

A major change in structural style in the southern part of the district, also recognised in the Kirkcudbright district (Lintern and Floyd, 2000), occurs at a boundary oblique to the regional strike. It is coincident with an irregular drift-filled hollow that crosses the Wigtown peninsula, but is not exposed at the coast due to late faults. The fault on the east coast [NX 477 397] appears to be relatively minor and numerous small, north-west-dipping, syn- D2 thrusts are exposed immediately to the north, suggesting that the boundary is an unusually large D2 thrust, which overprints the tract structure.

Structural dip variation

Throughout the district the strike of bedding is generally quite consistent (050°–070°), although locally it becomes more easterly (e.g. north of Cairn Head) or northerly (e.g. adjacent to the western end of the Cairnsmore of Fleet granite). In the north-west of the district, the dip of bedding in the Leadhills Supergroup and the northern Gala Group tracts is also constant, being near vertical. However, to the south there are substantial variations in the general attitude of bedding, both across strike from north to south and, in central parts, along strike (Figure 23a), (Figure 23b), (Figure 23c)

Traversing southwards in the west of the district, bedding passes from vertical to south-east-dipping in Gala tracts 4 and 5 and in the northern part of Gala tract 7 before swinging through vertical into progressively lower northward dips through the northern Hawick Group tracts. It then steepens southwards, passing through vertical again in the Carghidown Formation outcrop to become steeply south-east-dipping at Burrow Head.

Eastwards, the south-east-dipping Gala Group steepens through vertical south of Kirkcowan to dip northwards near the River Cree. East from the River Cree to the Cairnsmore of Fleet granite, the entire Gala Group section dips relatively gently (30°–60°) towards the north-north-west. In the tracts which pass south of the granite, this dip persists for about 5 km along strike where bedding steepens to vertical or steeply south-south-east-dipping in a 1 km-wide, north-trending transition zone around Pibble Hill. Within this zone, the dip transition seems to occur along a zigzag trace (Figure 23a), (Figure 23b), (Figure 23c) suggestive of a series of open, south-south-east-inclined synformal flexures.

The dip variation appears to be superimposed on the tract structure because changes in dip are gradual, generally not associated with bounding faults and correspond to similar variations of the S1 cleavage and the Moniaive Shear Zone foliation. Thus the dip variation probably also affects the tract-bounding structures themselves. However, a gently dipping crenulation cleavage (S2) developed locally in association with recumbent 'D2' folding is not affected. As this S2 cleavage is overprinted by the thermal alteration associated with the Cairnsmore of Fleet granite, establishment of the dip variation must predate emplacement of the granite. Where strike-parallel, the granite contact follows the dip of the host rocks. Locally, adjacent to the western margin of the granite, the strike of the moderately dipping fabric has been modified by the granite.

Detailed (within-tract) structure of the district

Resurvey of the district was based on the extensive coastal exposures in the Hawick Group in Wigtown Bay and the west of Luce Bay and limited coastal sections in the Gala Group in Luce Bay. Additional detailed information was gathered along two transects across strike:

The description of the detailed structure is based largely on these areas. Correlation of structural style across the district and with areas along strike to the west (e.g. Barnes et al., 1987) and east (e.g. Lintern and Floyd, 2000) suggests that the structure displayed in these sections is generally representative of internal tract structure over long distances. However, this is not true in the north-east of the district where the rocks in the thermal aureole of the Cairnsmore of Fleet granite pluton appear to have a more complex structural history. Kink bands, locally well developed in a range of orientations in strongly foliated rocks, are not considered further.

Structure of the Leadhills Supergroup and Gala Group

As described above, the regional structural configuration of the district, comprising a series of fault-bounded, steeply dipping to vertical strike-parallel tracts, is thought to have been largely a consequence of the D1 deformation. However, in the north of the district, little of this is apparent at outcrop. The strata are generally relatively weakly deformed, although cleavage is generally present in finer grained lithologies and tight to isoclinal folding is apparent locally. The Moniaive Shear Zone affects a wide area east of the River Cree, mainly in the thermal aureole of the Cairnsmore of Fleet granite. Otherwise, in the west of the district, post- D1 structures occur only locally, with upright 'D2' folds present in Gala tract 5 and steeply plunging 'D3' folds in a narrow zone adjacent to the Orlock Bridge Fault.

D1 structure

Strata are commonly consistently northward younging, suggesting that wide zones across strike, although steeply dipping, are not significantly folded. This seems partly to be a result of sedimentary facies, with sequences dominated by very thickly bedded or massive sandstone generally not folded (e.g. in Gala tract 2 and parts of Gala tracts 5 and 7). However, some well-bedded sequences (e.g. a wide area of the Portpatrick Formation around Gleniron Fell [NX 201 621] and in parts of Gala tract 1) also show little or no evidence of folding. Folds, of generally similar form where visible, are developed locally at a range of scales in zones recognisable up to a kilometre across strike in areas where exposure permits. Intensely folded zones, with alternating belts of southward and northward younging up to 200 m thick, occur in medium-bedded sandstone with interbedded mudstone of the Shinnel Formation in well-exposed ground south of Carscreugh [NX 225 595] and in similar facies in Gala tract 7 exposed around Chippermore [NX 297 490]. Less intensely folded zones are well exposed on the coast in Luce Bay in Gala tracts 4 and 5 in facies with more variable bed thickness, including some very thickly bedded units.

In the west of the district, the attitude and interlimb angle of the D1 folds vary across strike from tract to tract, influencing the variation in bedding dip (Figure 23a), (Figure 23b), (Figure 23c). Folds, locally well developed at a range of scales in the Portpatrick and Shinnel formations, are gently plunging and dominantly isoclinal with near-vertical axial surfaces. Folds seen in the coastal section in the northern part of Luce Bay have axial surfaces dipping south-east with southwards bedding dip in both limbs. A fold pair exposed in Gala tract 1 (Plate 14) and several larger scale, close to tight folds exposed in the northern part of tract 4 are south-west-plunging, but in the southern part of the section they switch to easterly plunging, one syncline being almost reclined. In the southern part of Gala tract 5, folds (Plate 15) are upright to south-east-inclined, consistently gently north-east-plunging and typically tight to isoclinal, although a more open structure is exposed on the foreshore at Rocks of Garheugh [NX 268 501]. Large, tight to isoclinal folds are also well developed in well-bedded sequences in Gala tracts 7a and 7b, again gently plunging but with vertical axial surfaces. Gala tract 8 is relatively poorly exposed, but a change from vertical to steep north-west-dipping is apparent southwards across the tract, transitional to the dip of the Hawick Group to the south. The only evidence for folding is a broad zone of southward younging, overturned beds in the middle of the tract, the full extent of which is not known. However, equivalent zones along strike to the south-west are characterised by wide belts of southward younging strata (Barnes et al., 1987; McCurry and Anderson, 1989; Stone, 1995).

D1 fold axial surfaces must vary with the structural dip in central and north-eastern parts of the district in line with changes in the dip of bedding; no fold plunge information is available from these areas.

The S1 cleavage is usually apparent macroscopically only in finer grained strata (mudstone to very fine-grained sandstone), although it may also be visible in thin section in coarser grained sandstone as a weak, anastomosing foliation. In the east of the district, cleavage in the Gala Group is generally more pervasive than elsewhere. Other than in fold hinge zones, the S1 cleavage is generally near-parallel to bedding. It commonly strikes a few degrees clockwise of bedding in the dominantly northward younging strata, although in the west of the district, in the southern part of the Portpatrick Formation, cleavage consistently strikes anticlockwise of bedding. The dip of cleavage may vary from bedding by up to 30° in mudstone, is downward facing in places, and it refracts through sandstone beds.

Moniaive Shear Zone

The Moniaive Shear Zone in its type area, north-east of the Cairnsmore of Fleet granite, is up to 5 km wide in the northern tract of the Gala Group (Barnes et al., 1995a; Phillips et al., 1995b). A zone of high strain in the north-east of the district lies along strike from, and is equated with, the Moniaive Shear Zone, although it is wider (up to 8 km) and affects several of the Gala Group tracts. To the west, the zone is truncated by the Cree Fault; sandstone exposed south-west of Newton Stewart is little-deformed compared with that to the east, consistent with a marked change in metamorphic grade across the fault (Chapter 8).

The high strain zone is characterised by an intense, pervasive cleavage near-parallel to bedding in the sandstone-dominated succession, locally with a strong, gently plunging linear component (Plate 16). The fabric is variably developed, consistent with variations in strain quantified elsewhere in the Moniaive Shear Zone (Phillips et al., 1995a), but it commonly transposes all original structure. It is particularly intensely developed in the northern part of the shear zone where abundant millimetre-scale quartz segregations occur parallel and at a low angle to the foliation. Within this part of the zone, a 70 m-wide quartz mylonite, with common syntectonic quartz veins, is intermittently exposed parallel to, and about 400 m south of, the main strand of the Orlock Bridge Fault. The mylonite is brecciated with abundant internal faults indicating brittle, probably polyphase, reactivation. The protolith is unclear but it seems to have developed along a fine-grained horizon within the Gala Group, which was particularly susceptible to ductile deformation during development of the shear zone.

The timing of the development of the high strain zone is unclear. The foliation is parallel to cleavage (S1) outside the zone and the transition is gradational such that generally one cannot be seen to overprint the other. North of the district, the foliation is locally deformed by folds that are similar in style to D2 and D3 developed elsewhere (Barnes et al., 1996; Floyd, 1999), suggesting that it may be essentially a more pervasive form of S1. However, within the district, relationships with intrusions suggest that the foliation must be at least in part younger than S1. For example, it cuts dioritic intrusions e.g. [NX 466 671] which are characteristically post- S1 in age and it deforms porphyroblasts in the thermal aureole of the Cairnsmore of Fleet granite. Together, these features suggest that the high strain zone developed over a protracted period of time. A minimum limit is provided by the later stages of the thermal metamorphism around the Cairnsmore of Fleet granite (cooling age about 395 Ma, see Appendix 7), with biotite hornfelsing and skarn alteration overprinting the foliation.

In thin section it can be seen that deformation of the sandstone was largely taken up within the matrix, resulting in the development of a locally intense ductile fabric defined by very fine-grained to cryptocrystalline granular quartz and chlorite (± feldspar and white mica) some with wispy domains of aligned, very fine-grained chlorite. Shearing also resulted in the stretching and attenuation of the more unstable lithic clasts. Quartz and feldspar porphyroclasts, which represent relict detrital grains, show evidence of cataclastic deformation (paracrystalline microboudinage) and exhibit a preferred dimensional orientation parallel to the fabric developed within the matrix. Quartz grains show evidence of a weak to well-developed undulose extinction and localised development of new grains. The preserved detrital clasts are wrapped by the fabric present within the matrix and are enclosed within locally well developed, symmetrical to asymmetrical pressure shadows composed of very fine-grained quartz and chlorite. Minor etching of quartz grain boundaries occurs adjacent to these pressure shadows. Within the aureole of the granite (Chapter 7) the fabric has been modified during subsequent thermal metamorphism by variable recrystallisation and increase in the grain size of the matrix and the development of biotite, although deformed primary clastic textures are preserved.

Gently plunging post-S1 folds, 'D2'

Gently plunging, post- S1 folds are generally rare in the Leadhills Supergroup and Gala Group across the Southern Uplands. However, rather unusually, they occur at two localities in the west of the district, and over a wider area in the east of the district.

Small recumbent folds, with a gently dipping crenulation cleavage in axial regions, are sporadically developed in mudstone with very thin sandstone beds of the Shinnel tract near Glenluce e.g. [NX 210 581] and in northern Gala Group tracts in the east of the district e.g. [NX 466 671]. Minor recumbent folds and associated cleavage are also locally well developed in mudstone and thin- to medium-bedded sandstone in the southern part of the Gala Group around Pibble [NX 517 604]. From there, the gently dipping cleavage extends across the Garheugh Fault north of Pibble Hill, and thence through Gala tract 7 to the south-west, manifested as a gently plunging crenulation lineation sporadically apparent on S1. It predates growth of biotite in the aureole of the Cairnsmore of Fleet granite, but overprints the zone of bedding dip transition south of the granite. The folds are mainly concentrated in Gala tract 6 but are also seen in the northern margin of Gala tract 7 on Pibble Hill. Their asymmetry varies with the dip of bedding from south-verging in north-west-dipping beds through neutral vergence in vertical strata to north-verging in south-east-dipping beds. These recumbent folds and cleavage are closely comparable to D2 features in the Hawick Group to the south (see below).

A few larger scale, upright, north-verging folds of bedding and S1, with a weak crenulation cleavage in mudstone in the faulted hinge zones, are exposed in thick-bedded sandstone in the south of Gala tract 5 at Rocks of Garheugh [NX 264 504]. They may, however, be more widely developed and be responsible for dip variation through much of the tract in the west of the district. The only analogous structures in the district occur in the Carghidown Formation tracts (see p. 62).

Steeply plunging post- S1 folds, 'D3'

Narrow, discontinuous zones of post- S1 deformation are associated with several tract-bounding faults throughout the Southern Uplands. Two such zones are exposed in the district, one associated with the Orlock Bridge Fault at the boundary between the Leadhills Supergroup and the Gala Group and the other in the outcrop of the Hawick Group (see below). The former is exposed at Wood of Dervaird [NX 216 577] (Barnes et al., 1995a, 1996). Deformation is largely concentrated in a thick unit of mudstone to the north of the fault, but a strong cleavage is present in sandstone immediately to the south. Close to the fault, the mudstone contains an intense, steeply dipping, bedding-parallel cleavage with common, foliation-parallel quartz veins up to 2 cm thick. The veins are boudinaged and folded by small, steeply plunging, sinistral folds (Plate 17); these are widely developed and deform cleavage through the mudstone unit. The style of deformation close to the fault is similar to the most intensely deformed part of a broader shear zone that locally marks the Orlock Bridge Fault in Ireland (Anderson and Oliver, 1986), pointing to important sinistral reactivation of this tract-bounding fault.

Structure of the Hawick Group

D1 deformation of the Hawick Group is generally more intense than is apparent in other parts of the Southern Uplands, consistent with the relatively high metamorphic grade, although D1 structures vary in intensity and style. Folds are abundant except in the northern part of the outcrop. They show an overall similar style, but commonly form more complex disharmonic structures in strata of widely varying bed thickness. Fold hinge zones may be faulted to varying degrees. Slickensides on bedding planes and mineral lineations in fine, bed-parallel veins perpendicular to fold axes and crenulated by cleavage, testify to bedding plane slip prior to tightening of the folds during cleavage development. The S1 cleavage is well developed throughout the Hawick Group and is commonly pervasive in sandstone, associated with flattened concretions and sand volcanoes. It tends to lie at a low angle to bedding in fine-grained lithologies, except in the hinge zones of tight folds, but may be refracted to a high angle to bedding in sandstone. Cleavage is congruous with the folds in profile although variably transects up to 20° clockwise in strike.

Post- S1 deformation comprises a relatively widely developed second phase of folding (D2) associated with a gently dipping crenulation cleavage, locally deformed by a zone of steeply plunging (D3) folds associated with two of the tract-bounding faults. Numerous lamprophyre and felsite dykes (Chapter 7) emplaced into the Hawick Group at various times during its deformation history are also important to its structural characterisation. A major swarm of early, strike-parallel felsite dykes was emplaced into the Cairnharrow Formation tracts during D1 and are variably deformed by the S1 cleavage, along with some lamprophyre intrusions. Later dykes, mainly lamprophyric, were emplaced prior to or synchronous with D2 deformation.

D1 — North of the Innerwell Fault

The structural style of the Cairnharrow Formation outcrop in the three tracts mapped north of the Innerwell Fault is transitional from the Gala Group to the north into the Hawick Group. Unlike the tracts to the south, there is no coastal exposure through much of this zone other than the southern part north of the Innerwell Fault. However, inland areas east of Wigtown Bay and in a section from east of Monreith northwards in the west of the district are locally well exposed.

Bedding varies from steep south-east-dipping or vertical in the north-east of the zone to moderate north-west-dipping in the south-west; strike is variable in the tract north of the Innerwell Fault due to post- D1 folding. Younging is dominantly northwards, particularly north of the Cambret Fault. To the south, southward younging strata occur locally in zones up to 100 m wide with dip similar to that in northward younging limbs indicating tight to isoclinal, gently plunging D1 folds, consistent with the plunge of most of the fold hinges exposed. A few moderate to steeply north-east-plunging minor folds occur near Orchardton, apparently related to S1. A zone about 350 m wide of south-younging strata on the west flank of Cairnharrow occurs north of a marked feature suggesting a strike-parallel fault and is also associated with D1 folding plunging more steeply north-eastward. A large structure, comprising a disharmonic set of complex folds is well exposed on the coast below Kirkclaugh [NX 530 522]. Otherwise, small-scale south-east-verging structures, varying from close to tight, appear more common in the coastal sections; this may reflect a real increase in the intensity of small folds towards the Innerwell Fault or may simply be due to the better exposure.

D1 — Innerwell Fault to the Glasserton Thrust

The tracts containing the outcrop of the Kirkmaiden Formation and the transition from the northern part of the Carghidown Formation outcrop (Figure 23a), (Figure 23b), (Figure 23c) are very well exposed in extensive coastal sections (e.g. Back Bay, Monreith; (Plate 18); Barnes, 1996), and inland in a strip parallel to the coast in the west of the district. The dip of bedding through most of this zone is variable due to changes in the interlimb angle and attitude of D1 folds (Figure 23a), (Figure 23b), (Figure 23c), modified locally by coaxial D2 refolding. D1 folds occur at a wide range of scales, with the maximum size of folds and the overall fold vergence varying across the district.

In northern parts of the zone in the Fleet and Wigtown Bay sections, the D1 folds tend to be relatively open (interlimb angles typically 70°–90°) with axial surfaces upright to steeply north-east-dipping; the north-younging limbs dip moderately northwards and the south-younging limbs are vertical or dip steeply southwards. South of Eggerness Point the folds are tighter with axial surfaces still inclined north-eastwards, such that south-younging limbs are near vertical to steeply north-dipping. Fold axes are generally linear and typically subhorizontal to gently north-east- or south-west-plunging throughout Fleet Bay and north of Rigg Bay [NX 480 445]. South of Rigg Bay, however, they plunge consistently north-eastwards, locally becoming almost reclined with relatively steep north-north-east plunge. The overall structure in these sections is dominated by north-younging beds with south-east-verging folds of relatively small size (short limbs are generally less than 20 m wide). In areas of wide foreshore exposure such as Eggerness Point [NX 493 463], these folds are seen to be laterally impersistent, dying out over a few hundreds of metres. Larger folds occur in a zone near Mossyard [NX 550 520] and also near Garlieston [NX 490 470] where they are associated with strike-parallel faults interpreted to represent a tract-bounding fault zone.

In the west of the district, the folds are close to tight and gently north-east-plunging on north-west-inclined axial surfaces, with north- and south-younging beds dipping northwards through most of the zone. Towards the south of the zone, the folds become tight to isoclinal and upright, with bedding close to vertical. The northern part of the zone is again dominated by northward younging beds with south-east-verging folds of relatively small size. However, southwards from Cairndoon [NX 380 390] the zone is characterised by regular alternations in younging on a scale of 200 m or less suggesting folds with essentially neutral vergence. This gives the appearance of a gentle sheet dip, but a number of faults parallel or slightly oblique to strike can be inferred and are undoubtedly important to the large-scale structure. Unfortunately, the significance of these faults is difficult to assess in the absence of detailed biostratigraphical control.

D1 — South of the Glasserton Thrust

The southernmost parts of the district, comprising the southern, disrupted portion of the Carghidown Formation and the lower Wenlock Ross Formation, are well exposed in continuous coastal sections around Burrow Head (Barnes, 1996). The dip of bedding varies systematically from steeply northward in the north-west, through vertical to steeply southward in the south ((Figure 23a), (Figure 23b), (Figure 23c)), continuing the trend from the zone to the north. Deformation is distinguished from that to the north by the widely variable plunge of D1 folds, including steeply plunging, sinistral structures. Cleavage is variably developed but may be intense in finer grained lithologies. It is congruous with the folds but is typically non-axial planar in gentle to moderately plunging structures where it strikes up to 20° clockwise of the axial surfaces. Generally, the cleavage becomes more closely axial planar as fold plunge increases.

Zones of disrupted bedding up to 30 m wide, separating packets of more coherently bedded strata, are common in the Carghidown Formation in the district and occur locally in the Ross Formation. They probably originated through soft sediment deformation processes (Chapter 4), possibly as zones of décollement (Lintern and Floyd, 2000) but in many cases appear to have focused subsequent deformation (cf. Kemp, 1987). Boundaries have been replaced by anastomosing sheared zones and internal disruption has commonly been accentuated by boudinage and intense veining. The veins themselves have frequently been folded and sheared, indicating that these zones have acted as a locus for deformation over a long period of time.

In coherently bedded strata, D1 folding is also intense, generally comprising relatively small south-east-verging folds in dominantly northward younging strata. Around Burrow Head, however, a number of larger scale structures produce zones of southward younging strata, with small north-verging D1 folds, up to 300 m wide. One such zone contains the transition from Carghidown Formation to Ross Formation (Chapter 4). Folds are tight to isoclinal and hinge zones are typically variably curved, some strongly so. This commonly causes fold plunge to change within metres, varying between moderate north-east and south-west (Plate 19). Locally the plunge varies to steeper angles and may pass through vertical, when the folds become downward facing. This may be due to local variations in strain (e.g. Knipe and Needham, 1986). However, broad zones of consistently downward facing, gently plunging folds in packages of strata between disrupted zones in the Kirkcudbright district (Lintern and Floyd, 2000, fig. 22c) cannot be explained in this way, suggesting instead pre- D1 inversion, probably during slump folding.

Discrete sinistral folds with constant steep to vertical plunge and short limbs ranging up to several tens of metres (Plate 20), occur locally throughout the southern part of the Whithorn peninsula. One such fold, plunging 66° to 87° west, is well exposed south of Port Castle Bay [NX 427 356] (Barnes, 1996, locality 3). The S1 cleavage, developed in muddy interbeds, is axial planar to the steeply plunging folds as well as the more gently plunging D1 folds. This emphasises that, despite their widely variable attitude, all of these folds are the products of one phase of deformation. However, the steeply plunging folds may have been initiated at a relatively late stage as they are rarely seen to refold earlier structures (e.g. Kemp, 1987, fig. 9). At Cairnhead Mote [NX 486 383] (Barnes, 1996, locality 6), minor folds in the short limb of a steeply west-plunging sinistral fold refold tight folds in which a tectonic fabric is apparent in the hinge zones, suggesting that they are not simply slump folds. This locality was described by Rust (1965) as indicative of two phases of deformation. However, it seems most likely that all of these folds are part of a 'D1' event in which the stress regime evolved from dominantly orthogonal to one involving a major component of sinistral shear.

Gently plunging post- S1 folds, 'D2'

Two sets of gently plunging structures, co-axial with D1 folds and generally of small to intermediate size, are grouped as D2 because of their conjugate geometry; they form box folds where they occur together. They are best developed within 2 km either side of the Innerwell Fault, being well displayed in the section south from Back Bay [NX 368 394] (Plate 18), where they fold bed-parallel lamprophyre dykes, and in sections north of Ringdoo Point.

South-verging open folds with gently dipping short limbs (generally less than 10 m long, although locally larger) and steep south-dipping axial surfaces (Plate 21), have only a weak crenulation cleavage developed in axial regions. These tend to be sporadically developed, although they are common locally and exaggerate the D1 variation in the dip of bedding and modify the dip of S1 cleavage.

North-verging or neutral, open to close folds with recumbent axial surfaces are generally of small size (short limbs less than 5 m and commonly less than 20 cm) and locally associated with minor thrust faults gently dipping north-west or south-east. The folds are associated with a gently dipping crenulation cleavage (S2), which is much more widely developed than the folds. This cleavage is also well developed around Cairnharrow [NX 533 561], where gentle flexures of bedding may be a weak form of the recumbent folding. Otherwise, evidence of S2 is seen sporadically over a wide area of the Hawick Group north of the Glasserton Thrust as a gently plunging crenulation of S1.

The orientation and geometry of the D2 structures suggests that they formed by shortening in the plane of bedding. This may have occurred as a continuation of D1 after locking of the D1 folds, or by subsequent renewal of shortening on the tract-bounding faults. However, if the recumbent form of D2 and the associated gently dipping S2 cleavage in the Hawick Group can be equated with similar structures around Pibble Hill in the Gala Group and near Glenluce in the Shinnel Formation, then its relatively widespread development suggests that some other mechanism is required. Recumbent D2 folds in the Innerwell tract occur in a zone in which bedding is dominantly north-west-dipping and the folds are north-west-verging, but in the Pibble area, where the dip of bedding passes through vertical, the D2 fold vergence varies with the dip of bedding. If the latter is representative of the situation regionally, then the recumbent folds appear to have been formed by the vertical shortening of bedding in more-or-less its present attitude, rather than by a consistent sense of shear on tract-bounding or other faults.

Steeply plunging post- S1 folds, 'D3'

Steeply plunging, similar and concentric folds are present mainly in the tract north of the Innerwell Fault; they are open to close and sinistral with the short limbs striking east to south. They are best developed near the fault at Black Rocks [NX 358 408] (Plate 22) and Skyreburn Bay [NX 576 545], with short limbs up to 16 m long cut by an axial planar, pervasive, east-striking cleavage. Thin mica lamprophyre dykes occur parallel to, and are deformed by, this cleavage at Black Rocks. Larger, gentle to open folds of this style are developed more widely through the tract north-east of Monreith in the west and east of the district and over a wide area to the east in the Kirkcudbright district (Lintern and Floyd, 2000).

The 'D3' folds indicate sinistral shear parallel to the bounding faults, particularly along the Innerwell Fault. Their relative timing with respect to recumbent D2 folds can be demonstrated at Black Rocks, although with some difficulty because both sets of structures are open and their axial surfaces lie at a high angle to one another. However, the recumbent folds and cleavage appear to be deformed by the steeply plunging folds, whilst the cleavage associated with the steeply plunging folds cuts across the recumbent folds. Consequently, the steeply plunging folds are assigned to a D3 event. They are geometrically similar to structures that occur adjacent to other tract-bounding faults, including the Orlock Bridge Fault farther north in the district as described above, and hence they are assumed to be equivalent. Bearing in mind the diachronous nature of D1, the D3 structures may have formed contemporaneously with the development of steeply plunging D1 folds in the younger part of the Hawick Group in a single regional sinistral shear event (e.g. Barnes et al., 1989).

Small, angular, close, steeply plunging folds (short limbs up to 15 cm long), with an east-trending axial planar crenulation cleavage in axial regions, also occur locally in exposures in Kirkbride Burn, near Bagbie [NX 488 553]. Their occurrence caused the trace of the Cambret Fault to be drawn north of the Carsluith intrusion, although the latter may have been intruded into a second strand of the tract-bounding fault.

Miscellaneous post- S1 folds in the southern part of the district

Post- S1 structures are rare south of the Glasserton Thrust, although three types of structure are present locally. Southward verging, gently plunging, open to close folds with steep south-east-dipping axial planes are associated with a weak crenulation cleavage in hinge zones. These are generally seen as minor folds, but locally with short limbs up to 20 m long, and are comparable to the steeply inclined set of D2 folds seen in the Hawick Group to the north. Rare, gently plunging, minor north-verging folds (short limbs are less than 2 m long) with steeply inclined axial planes are apparently unrelated to the 'D2' folds, but their relative age is unknown. Steeply plunging, dextral, open minor folds, with vertical axial planes striking north to north-north-east, some with angular and fractured hinge zones, are widely developed, particularly so in the section from Burrow Head to Port of Counan. Short limbs are typically 10 to 20 cm long but range up to 1 m. Rare minor folds of the same style but with sinistral sense and axial planes that strike north-east represent a conjugate set of structures, which may be comparable to D3 to the north.

Faults

Faults in a range of orientations (Figure 24); (Figure 25) are very common in the coastal sections. Their movement direction may be determined from fault-rock fabrics or structures in the wallrocks, although the throw can be determined only if there is evidence within the extent of the exposure. Faults are presumably equally abundant throughout the district, but only the larger faults that form distinct features (e.g. north-north-west-trending faults over Cairnharrow) or offset a marker horizon (usually dykes or the Moffat Shale Group outcrops) can be recognised inland and even then only in reasonably well-exposed ground. Four principal groups of faults can be recognised in the district:

The early strike-parallel faults include the tract-bounding fault structures discussed above. However, many other faults are apparently related to D1 folding because they replace hinge zones or cut out fold limbs and, in the south of the district, form sheared zones related to both steeply plunging D1 folds and the margins of disrupted zones.

D1 folds are deformed at a number of localities in the Hawick Group tracts by reverse faults with small displacement, dipping gently to moderately northwards or less commonly southwards. These faults are locally associated with recumbent D2 folds and may generally be associated with this stage of the deformation sequence. The inferred Garlieston Fault is the only large structure of this group.

Steeply dipping faults oblique to strike (Plate 7) are abundant in the coastal sections throughout the district, particularly in a fracture zone which dominates the section for 2 km north-west from Burrow Head. Fault breccia, up to 3 m thick, and locally intense quartz and/or carbonate veining are characteristic of these structures, even where the net displacement is demonstrably small. Some fault surfaces display several generations of slickensides, ranging from steep to gentle pitch, showing multiple phases of movement. Vertical displacement is generally difficult to quantify but offset of bedding, dykes and other indicators of the sense of movement commonly show a component of lateral displacement. This broadly defines two groups, with north- to north-east-trending faults generally sinistral whereas north-west-trending faults are usually dextral, a pattern common throughout the Southern Uplands (cf. Stone, 1995, fig. 23). Within these groups, however, several sub-sets distinguished by strike clusters are apparently variably developed in different parts of the district (Figure 24), although the data are to some extent influenced by the nature of the exposure and orientation of the coastal sections.

Most orientations are represented in the coastal sections in the south of the district where, from complex cross-cutting and termination relationships and relationships with minor intrusions (Chapter 7), it is apparent that they formed contemporaneously. This is understandable within the regime of sinistral shear evident from D1 to D3 in the south of the district. In a brittle shear zone orientated parallel to the tract-bounding faults (Figure 25), the sinistral north- to north-north-east-trending and dextral north-north-west-trending faults correspond with synthetic and antithetic Reidel shear directions (R1 and R2). The east-north-east- and east-south-east-trending faults may represent the P and X shear directions within the same model or Reidel shears in a rotated stress system. Large-scale movement was concentrated on the sinistral faults in the synthetic Reidel shear directions. In the southernmost parts of the district this was close to the orientation of bedding and some faults (e.g. those through Port Castle Bay [NX 426 358] and Isle of Whithorn) are associated with broad zones of distributed bedding-plane slip. Subhorizontal shearing along bedding formed slickensides and slickencrysts in thin veins along bedding surfaces. These are sinistral where the sense of movement can be demonstrated. More discrete fractures, commonly preserving several generations of slickensides, show dominantly horizontal displacement but also oblique and vertical components. Small sinistral displacements ranging from a few centimetres to 3 m can be demonstrated where bedding-parallel faults cut oblique east-south-east-trending lamprophyre dykes.

Some of the faults oblique to bedding were reactivated during episodes of Upper Palaeozoic and Mesozoic extension. Mineralisation of mainly north-west- to west-north-west-trending structures in a zone south-east of Newton Stewart (Chapter 10) implies reactivation during the Carboniferous. North-west- to north-north-west-trending faults, spaced 20 to 30 km, across the Southern Uplands form a horst/graben system (e.g. Stone, 1995, fig. 4) bounding Permo-Triassic half-graben basins. In the district, one such fault passes along the western side of Luce Bay (Chapter 8) and another may be present parallel to the western shore of Wigtown Bay. Reactivation of segments of some of the north-east-trending, tract-bounding faults in conjunction with movement on the cross-strike faults allowed net differential movement of blocks, as shown by the juxtaposition of rocks of quite different metamorphic grades (Chapter 6).

Chapter 6 Regional metamorphism of Lower Palaeozoic rocks

Early studies of metamorphism in the Southern Uplands established that the regional grade is generally very low, with mineralogical evidence of prehnite-pumpellyite facies widely developed in volcaniclastic sandstone and basic volcanic rocks (e.g. Oliver and Leggett, 1980). Reconnaissance studies of metapelite grade showed that mudrocks are most commonly in the anchizone, and that K-white mica b lattice dimensions indicate medium to high pressure facies series conditions (Oliver et al., 1984; Kemp et al., 1985).

A more detailed survey of metapelite grade, based on one mudrock sample per 2–3 km2 has been carried out as part of the geological resurvey of the Southern Uplands. Samples were collected from typical pelitic lithologies interbedded within the greywacke-dominated sequences and the Moffat Shale Group (246 samples in total from the Wigtown district). Variations in grade, interpreted in the context of the structural configuration, are presented as a contoured metamorphic map (Figure 26). Contours of equal crystallinity (isocrysts), were defined by manual contouring of the white mica crystallinity indices. The errors and precision involved in contouring the crystallinity data (Robinson et al., 1990), were determined by multiple samples at several sites in an adjacent district (Merriman and Roberts, 1992, 1996), and elsewhere (Roberts et al., 1990). Results from multi-sampled sites indicate that 95 per cent of the samples have indices within the range of values delineated by zonal isocrysts.

Metapelite mineralogy and texture

The mineralogy of the fractions that are less than 2 µm and white mica (illite) crystallinity indices of the pelite samples were determined by X-ray diffraction (XRD) analysis using the preparation methods and machine conditions detailed by Roberts et al. (1991), and the measuring conditions recommended by Kisch (1991). In response to increasing metamorphic grade, authigenic clay micas recrystallise to form thicker crystallites. The thickening of a population of white mica crystallites can be measured by two methods. Using XRD techniques, the Kubler index (KI in D°2q) measures the small reduction in the half-height width of the 10Å white mica peak which results from thicker crystallites. Another method uses transmission electron microscopy (TEM) to directly measure the thickness of a population of white mica crystallites, and results show good correlation with the Kubler index. Both methods represent a statistical measurement of the progressive thickening of white mica crystallites with increasing grade (Merriman et al., 1990, 1995).

The Kubler index is used to define a series of metapelite zones of low- and very low-grade (sub-greenschist) metamorphism, rising from late diagenetic (KI > 0.42) through the anchizone (low anchizone KI 0.30–0.42; high anchizone KI 0.24–0.30), to the epizone (low epizone KI 0.20–0.24; high epizone KI 0.16–0.20). Rocks at the lowest grade have not been identified in the district, but are present in adjacent districts (Stone, 1995; Lintern and Floyd, 2000). Typically, they are composed of chlorite and illite, with subordinate corrensite and minor amounts of kaolinite, mixed-layer illite/smectite, albite, quartz and hematite. At anchizonal grade, the mudrocks are composed typically of chlorite and phengitic 2M1 K-mica, with minor albite and quartz; neither corrensite nor paragonite are detectable in the < 2 µm fractions. In the epizone, they contain abundant 2M1 K-mica and chlorite, with minor albite, quartz and rutile.

This increase in grade is generally apparent macroscopically by the increasing development of foliation in mudrock, with the development of a bedding parallel fissility (shale) passing into cleaved slate through the anchizone. Petrographically, weakly to moderately cleaved shale and slate, typical of the low anchizone (KI 0.30–0.42), show a bedding lamination of abundant detrital muscovite, albite and quartz, anatase and apatite, and a bedding-parallel microfabric of anastomosing white mica and chlorite intergrowths. The cleavage microfabric is better developed in slates of the high anchizone (KI 0.30–0.25) and epizone (KI < 0.25), where it forms well-orientated intergrowths of white micas and chlorite flakes, typically 1–4 µm thick, enclosing grains of quartz, albite and relicts of large detrital phyllosilicates.

Variation in metamorphic style

The isocrysts (Figure 26) show that the district includes two areas of contrasting metamorphic style. Across most of the district, variations in regional grade between low anchizone and high epizone are apparent, offset by faulting. However, in the north-east of the district, east of the Cree Fault, a wide area of high epizone marks thermal alteration associated with the Cairnsmore of Fleet granite.

Regional metamorphism west of the Cree Fault

Mean KI values west of the Cree Fault in each tract in the district (Figure 27) suggest that the metapelite grade generally lies close to the anchizone–epizone transition, with no large changes across tract-bounding faults. The mean grade is lowest in the northern part of the Gala Group, increasing gradually southwards into Gala tract 5. Too few samples were analysed to provide meaningful mean KI values for Gala tracts 6 and 8, but a more pronounced trend of increasing grade is apparent south of the Garheugh Fault. This culminates in a belt of relatively high grade epizonal rocks in the northern part of the Hawick Group. This area is characterised by a relatively well-developed slaty cleavage in mudrocks, pervasive in sandstone and locally crenulated by a second cleavage, which may be partly responsible for the relatively high grade. However, the Hawick Group is relatively intensely deformed throughout its outcrop, although the grade decreases southwards to low anchizone in the south and east of the Wigtown peninsula.

A more detailed picture of the spatial variation of metamorphic grade across the district is provided by the outcrop of the metapelite zones interpreted from the sample KI values (Figure 26). These do not show systematic variation in grade within tracts in relation to the younging direction of strata, from relatively high at the base becoming lower upwards, as might be expected from burial metamorphism prior to thrust imbrication. Instead, the isocryst traces are generally north-east-trending, oblique to the tectonostratigraphy, although locally, notably south of Wigtown, they become strike-parallel. The direction of decreasing grade locally suggests that the metamorphic zones are south-west-dipping in the north-west and south of the district, but north-east-dipping over central parts. The amount of dip is difficult to define but is related to the outcrop width of the zones. The broad areas of high anchizone in the north-west suggest gently dipping zones, whereas in the centre and south of the district the high anchizone has a much narrower outcrop width (1–2 km), implying steeper dips. In a burial situation, the anchizone represents a temperature range of about 100°C, this is about 4 km thickness at a relatively low geothermal gradient of 25°C/km. Taking the high anchizone as about half of this thickness, the metamorphic zones in the centre of the district are near vertical as drawn in (Figure 26). The alternative is that there are more north-north-east-trending faults north of Port William than have been recognised, with one or both margins of the narrow zone of high anchizone grade being faulted.

Fault displacement of the zones is down to the north on most of the strike-parallel faults. Opposed, down to the south, displacements on the Killantringan Fault and parts of the Laurieston and Cambret faults introduce tract segments with low anchizone grade in the north of the district and between Port William and Wigtown. Faults trending north-east generally displace the metamorphic zones down to the north-west, although downthrow switches to the south-east in some segments between tract-bounding faults.

East of the Cree Fault — thermal alteration

East of the Cree Fault, the regional pattern has been extensively overprinted by the thermal aureole surrounding the Cairnsmore of Fleet intrusion. The mapped limit of thermal alteration (Figure 26) is defined by the extent of macroscopically recognisable biotite hornfels (Chapter 7), but the metapelite grade shows a more extensive cryptic aureole. Both the biotite hornfels and the metapelitic aureole extend across the Orlock Bridge Fault zone without apparent offset, suggesting that they overprint the latest movement. At its western margin the aureole, like the Moniaive Shear Zone which it overprints, terminates against the Cree Fault, suggesting that the Whithorn peninsula block is downthrown to the west. South of the granite, the cryptic aureole appears particularly extensive, but it probably merges with high epizone regional grade rocks around the Cairnholy Fault as to the west.

Grade in the thermal aureole ranges up to epizonal and hornblende-hornfels facies close to the intrusive contact. However, high KI values more usually associated with non-metamorphic (late diagenetic) pelites are anomalously developed in some spotted slates within the aureole. Preliminary results of TEM studies of similar anomalies in the aureole of the Criffel–Dalbeattie granite (Lintern and Floyd, 2000) suggest that late growth of very thin white mica crystallites has replaced cordierite spots as a result of local retrogression of the pelitic hornfels.

Metamorphic history

The regional pattern of metamorphism found in the Southern Uplands is not a normal basinal burial pattern that results from grade increasing into older strata with greater thickness of overburden, the zones forming parallel to the stratigraphical surfaces (cf. Roberts et al., 1991, 1996). In the Rhins of Galloway (Stone, 1995; Merriman and Roberts, 1996) a pattern of grade increasing southwards into sequentially younger tracts of strata is more consistent with accretionary burial and underplating into a thrust stack. In this situation, younger tracts of strata are progressively buried and metamorphosed beneath the tectonic overburden of older strata, as seen in the underplated slate belt of the Kodiak accretionary complex (Sample and Moore, 1987). TEM studies of the fabric-forming phyllosilicates developed in anchizonal to epizonal metapelites on the Rhins of Galloway indicate accretionary burial depths of 8 to 13 km (Merriman et al., 1995).

In the district, the regional pattern of metamorphism could have evolved as the thrust stack was assembled and rotated. The metapelite zones formed by burial within the developing thrust stack (e.g. Merriman and Roberts, 1992) would have been approximately ground-parallel and at a high angle to the stratigraphy, as may still be seen in the north-west of the district. However, as the stack continued to thicken, further rotation and steepening of the imbricated strata in some parts resulted in previously established metamorphic zones developing a tectonic dip and becoming folded. In the district, the zones appear to dip towards a large area of low anchizone grade around Kirkcowan, suggesting an open synformal fold of the zones. To the south, the zones appear to dip away from the belt of highest grade rocks, the high epizone, suggesting an antiformal axis close to the trace of the Cairnholy Fault. Folding of the metamorphic zones must have been accompanied by uplift to shallower structural levels, so that lower grade rocks were not reset, and occurred prior to offset on the strike-parallel and oblique faults.

The apparent 'folding' of the zones may to some extent simply reflect non-uniform depth ranges within which they formed in response to local variations in heat flow across the district. For example, the belt of high epizone rocks associated with the Cairnholy Fault in the south of the district is coincident with a zone of abundant early felsic dykes, the emplacement of which may have been associated with a pulse of hot fluid causing elevated grades locally. However, the steeply dipping zones in central and southern parts of the district require rapid lateral changes in grade which seem unrealistic in the absence of major intrusions, and thus they appear likely to have been tilted. One possibility is that individual fault blocks were tilted during late normal faulting (cf. Stone, 1995; Lintern and Floyd, 2000). However, in the district this seems unlikely as the belt of steeply dipping zones is continuous across several faults, one with an opposed sense of displacement, and the dip of the zones is oblique to the faults. Folding of the zones about north-north-east-trending axes to the extent required to generate vertically dipping zones would have a perceptible effect on bedding and the tract-bounding faults, which is not apparent. As noted above, the apparently steeply dipping zones may be explained by the presence of more cross-strike faults than have been mapped, separating narrow blocks of different grade. Alternatively, the late sinistral bedding-parallel shear, for which there is much evidence in the south of the district, could have caused rotation of metamorphic zones into their present attitude from an original orientation much closer to the structural dip.

Post-metamorphic faulting took place prior to the intrusion of the Cairnsmore of Fleet granite, with neither the hornfels aureole nor the cryptic metapelitic aureole showing any offset on faults other than the Cree Fault which seems to truncate its western end.

Chapter 7 Intrusive rocks and thermal metamorphism

Several suites of igneous rocks are present in the Lower Palaeozoic strata of the Southern Uplands. They were first described systematically and characterised petrographically during the primary survey (Craik, 1873; Irvine, 1878 in the district; Peach and Horne, 1899). The three large granite plutons that crop out in south-west Scotland have dominated subsequent study (Floyd, 1999 and Lintern and Floyd, 2000 and references therein). The Cairnsmore of Fleet pluton, in and to the north-east of the district, was described by Gardiner and Reynolds (1937) and has since been considered in more detail by Parslow (1964, 1968, 1971), Parslow and Randall (1973) and Cook (1976). Smaller granitic to dioritic intrusions scattered throughout the Southern Uplands were reviewed for their mineralisation potential by Cooper et al. (1982) and Parker et al. (1981). Intrusive rocks in the areas adjacent to the district have been recently described in Stone (1995), Floyd (1999) and Lintern and Floyd (2000).

Dykes showing a range in composition are common throughout the Southern Uplands but are best developed and well exposed in southern parts in south-west Scotland and on the Ards Peninsula in Northern Ireland. Early workers (Read, 1926; Reynolds, 1931; Blyth, 1949) described the main rock types and recognised that certain suites occurred in well-defined zones. These, and later, predominantly structural studies (Anderson, 1962; Rust, 1965; Weir, 1968; Anderson and Cameron, 1979) also demonstrated that the dyke suites show a range of relationships to cleavage and fold and fault structures, suggesting that they 'punctuated' the tectonic events.

Recent work has focused on geochemical and isotopic characterisation of the full range of intrusive igneous rocks (Barnes et al., 1986; Rock et al., 1986a, b, 1988; Shand, 1989; Henney, 1991; Swarbrick, 1992; Shand et al., 1994). This has shown that the lamprophyric rocks represent magmas which originated deep in the mantle, the different groups probably resulting from slight variations in melting conditions and differing degrees of modification in the upper mantle and crust. The main regional zone of dykes across the southern part of the Southern Uplands is dominated by mica-bearing lamprophyres that are chemically similar to those found in the Lake District. This has been taken as evidence for the presence of Lake District mantle underthrust below the Southern Uplands (Shand et al., 1994). Otherwise, hornblende-bearing lamprophyre dyke swarms are locally spatially associated with the larger dioritic to granitic intrusions, particularly in south-west Scotland. They have been interpreted as having a genetic (parental) link with the plutons through differentiation and mixing (Barnes et al., 1986; Henney, 1991). A few of the felsic dykes may represent differentiates of the lamprophyric magmas. However, the majority of the intermediate to acid lithologies are petrogenetically unrelated to the lamprophyre suites and were probably produced during melting events in the lower crust or upper mantle. Some suites are spatially associated with, and have similar chemical compositions to the plutons, implying that they were derived from the same magma chambers that ultimately produced the plutonic rocks.

Isotopic dating of the intrusive rocks has provided a geochronological framework for the tectono-magmatic evolution of the Southern Uplands (Figure 22). The oldest ages, from 400 ± 9 Ma up to 418 ± 10 Ma obtained from the later, post-tectonic dyke rocks (Rock et al., 1986b; Rock et al., 1988), approach the biostratigraphically constrained age of the younger Llandovery and Wenlock host rocks (Barnes et al., 1989, fig. 5.1). Attempts to date the older, deformed intrusions have as yet been unsuccessful. The larger, late Caledonian intrusions show a contrast in ages between the northern and southern parts of the Southern Uplands (Halliday et al., 1980; Thirlwall, 1988). The northern plutons (e.g. Priestlaw, Carsphairn, Loch Doon) are dated at around 410 Ma whereas the southern plutons nearer the Iapetus Suture (e.g. Criffel, Cheviot, Cairnsmore of Fleet) are dated at about 395 Ma. A recent U-Pb age of 395 ± 3 Ma determined for three monazite and zircon fractions from the Cairnsmore of Fleet pluton (Appendix 7) confirms the Rb-Sr age (397 ± 2 Ma, Halliday et al., 1980) obtained for this intrusion. All of these intrusions are early Devonian in terms of recently published time-scales (e.g. Gradstein and Ogg, 1996) but the older ones have been previously described as late Silurian in some earlier accounts that utilized different time-scales.

Studies of the isotope ratios of the igneous rocks have provided further insight into the origin of the magmas and their source rocks. The granitoid plutons in the north of the Southern Uplands show similarities to the igneous rocks in the Midland Valley whilst those in the south have characteristics more similar to igneous rocks in the Lake District, suggesting different basement terranes (Shand, 1989; Shand et al., 1994). However, there is considerable debate as to the origin of the parental magmas to the plutons. The Criffel pluton was considered to be infracrustal by Chappel and Stephens (1988), whereas Shand (1989) argued from isotopic evidence that a dominant mantle component was present. Others, such as Cairnsmore of Fleet and the centre of the Criffel intrusion, do show characteristics of crustal-derived granites and were probably formed from partial melts of lower crustal rocks related to underplating and intrusion of basic mantle-derived melts. However, the isotopic evidence shows that the plutonic rocks were not derived from a single source and that crustal melts formed at least a component of the plutons.

The younger dolerite dykes that occur in the Southern Uplands and adjacent areas (Floyd, 1999; Lintern and Floyd, 2000) are thought to be mainly members of the Mull or Arran Palaeogene (Tertiary) dyke swarm, although it is possible that some may be of Carboniferous age.

Minor intrusions

Most of the numerous minor intrusions exposed in the district form part of the Southern Uplands regional 'Caledonian' swarm. These intrusions are generally termed 'dykes', because of their predominantly steep attitude, although many lie parallel to the steeply dipping bedding. Two broad suites of dyke rocks are present (Barnes et al., 1986; Rock et al., 1986b; Rock et al., 1988):

No absolute age dates are available from the dykes in the district, although a range of relationships with tectonic structures show that their emplacement was more-or-less coeval with deformation (Barnes et al., 1986 and discussion below). Radiometric age determinations obtained for the lamprophyre dykes in the Kirkcudbright district are in the range 395–418 Ma (Rb-Sr whole-rock ages and K-Ar biotite and hornblende mineral ages, Rock et al., 1986b; Rock et al., 1988).

Field relationships

The district may be divided into three parts on the basis of dyke composition and orientation (Figure 28).

In the north of the district, in the Leadhills Supergroup and Gala Group tracts, hornblende-lamprophyre (commonly appinitic) and felsic dykes are typically several metres to 100 m thick and north- to north-north-east- and north-west-trending. Their distribution, although to some extent obscured by available exposure, is not uniform. They are most common in the coastal section in the north-east of Luce Bay and in a broad north-north-east-trending belt to the east, encompassing the Culvennan diorite. Within this belt, there are notable clusters of dykes near Culvennan Fell [NX 310 645], north of Garheugh [NX 265 525] and around Elrig [NX 330 480]. All these occurrences include both lamprophyric and felsic types suggesting that they are petrogenetically related and/or that they utilised the same conduit system. In contrast, dykes are rare over a wide area around the Cairnsmore of Fleet granite. Mica-lamprophyre dykes are, in general, absent from northern parts of the district, although one occurs in an 8 m-wide zone, comprising several dykes emplaced along a north-north-east-trending fault, 500 m south of Stair Haven [NX 210 531]. In general the dykes in northern parts of the district were emplaced along, and parallel to, faults. However, a few relatively thin, bedding-parallel intrusions do occur in the coastal section north of Mull of Sinniness [NX 213 523]. Some of the felsic dykes, including many of those in the cluster around Elrig, contain a strong fabric parallel or oblique to the dyke margins. This fabric is locally folded on steeply plunging axes consistent with continuing movement along the controlling faults during or after emplacement.

In the southern part of the district, in the Hawick Group tracts, two distinct zones of minor intrusions are apparent (Figure 28). The Cairnharrow Formation tracts host predominantly felsic dykes, typically up to a few metres in thickness, although ranging up to 20 m, the great majority of which were emplaced parallel to bedding. The distribution is again not uniform, with many dykes occurring in a 2 to 3 km-wide belt along the Cairnholy Fault. Within this, marked clusters of dykes occur in the area east of Carsluith around [NX 535 550] and [NX 570 570] and in the central part of the Wigtown peninsula [around 380 450]. Although postdating the main D1 folding (dykes locally having been emplaced along fold axial surfaces), many of the dykes in this zone contain a weak to intense fabric (Plate 23); (Plate 24), which is evident in the field (e.g. Barnes and Fettes, 1996). The fabric is commonly oblique to the dyke margins and as a consequence was interpreted by Blyth (1949) as resulting from dextral shear along the zone during or after emplacement of the dykes. However, it is continuous with the S1 cleavage in the host sedimentary rocks, but varies in orientation in the dykes due to refraction (as it does in sandstone beds in the host rocks). At one locality, near High Auchenlarie [NX 539 535], a 2 m felsite dyke and the cleavage within it are folded around an upright, south-verging D2 fold. In the southernmost part of this zone, adjacent to the Innerwell Fault, several thick, bedding-parallel lamprophyre dykes exposed in Skyreburn Bay [NX 576 544] are cut by a irregularly developed, oblique fabric, which is parallel to the S1 cleavage in adjacent sandstone.

A marked change in the nature of the minor intrusions occurs in the vicinity of the Innerwell Fault, to the south of which dykes of all types, including biotite lamprophyre, are abundant and occur in a wide range of orientations (Figure 28). Dyke thickness is generally less than to the north, typically from a few centimetres to 2 m, although ranging up to 20 m. Many of the dykes occur parallel to bedding (Plate 25) and are not associated with other features which allow their relative age of emplacement to be determined beyond postdating the S1 cleavage. However, some dykes show critical relationships which suggest that, as in the zone to the north, many were probably emplaced relatively early in the tectonic history. Strike-parallel dykes are common in the axial zones of D1 folds and cross-cut these folds (Plate 26) suggesting that they generally postdate D1 folding. However, two examples on the foreshore in Wigtown Bay, south-east of Creetown, show that, at least locally, the earliest dykes were emplaced prior to or during D1 deformation. One 33 cm-thick bed-parallel biotite-lamprophyre dyke [NX 540 518] is folded by a small, upright fold with transecting cleavage visible in the dyke (Plate 27). To the west [NX 520 524], two lamprophyre dykes each about 1 m thick are repeated across a larger scale, tight syncline and may also be folded; a third dyke occupies the hinge zone of this syncline. In the section south of Back Bay [NX 368 394], bedding-parallel lamprophyre dykes are folded by both the upright and recumbent D2 folds ((Plate 28); Barnes, 1996). Here, however, one dyke also shows evidence of emplacement during D2, utilising recumbent D2 folds to transgress laterally (Plate 29). Relationships between dykes and steeply plunging folds in more southerly tracts cannot be demonstrated, although at Black Rocks [NX 358 408], thin mica-lamprophyre dykes were emplaced along, and are deformed by, the cleavage associated with D3 folds.

These key relationships show that, in the Hawick Group tracts, lamprophyric and felsic dykes were emplaced during D1, D2 and D3 deformation. During thrust-related D1 shearing and subsequent formation of conjugate D2 folds, the principal direction of shortening (s1) lay close to the dip direction of bedding. At an early stage, when the thrust slices were gently dipping, the overburden pressure would have been substantial and s2 therefore perpendicular to bedding. As the imbricate stack steepened, however, it appears that s2 and s3 switched, allowing bedding to be utilised for emplacement of the dykes. Subsequently, a change in the D1 stress regime led to a larger component of sinistral shear in the more southerly Hawick Group tracts. This accompanied continued dyke emplacement to the north in areas actively undergoing D3 deformation.

Succeeding stages in the intrusive history were dominated by the brittle deformation related to continuing sinistral shear (Chapter 5). In northern parts of the district, thick dykes may have been emplaced at this time into active fault zones developed parallel to the Reidel shear directions (Figure 25). Dykes oblique to bedding in the Hawick Group tracts have a wide range of orientations comparable to that of the faults in this part of the district (Figure 24). East-south-east- to south-east-trending intrusions (the X shear direction of the model in (Figure 25)) are common in a fault zone along the coast north-west of Burrow Head [NX 444 345] where variable deformation of the dykes indicates that emplacement occurred whilst the faults were active. Variable cross-cutting and offset relationships indicate that most, if not all, of the fault orientations were active contemporaneously during the intrusion history.

Relationships between the dykes and the larger intrusions (described below) in the district suggest that, with the exception of the Cairnsmore of Fleet granite, they were broadly coeval. Spessartite and felsic dykes within the thermal aureoles of the Glenluce and Culvennan diorites are altered and hornfelsed and therefore probably predate emplacement of the diorite bodies. One dyke [NX 325 655] is demonstrably cut by the Culvennan diorite, although dykes cut a small outcrop of diorite to the south of the main intrusion [NX 329 645] and the adjacent, broadly contemporaneous (Cooper et al., 1982) vent breccia. The precursor microgranodiorite dyke swarm to the Kirkmabreck granodiorite intrusion [NX 480 565] is compositionally similar to the early (cleaved) felsic suite, suggesting that this too may have been emplaced relatively early (see p. 78).

Elsewhere in the Southern Uplands, relationships show that intrusion of the lamprophyre and felsic dyke suites continued during emplacement of the granite plutons. Felsic and, less commonly, lamprophyre dykes appear to be spatially related to the Criffel–Dalbeattie pluton (dated at 397 Ma by the Rb-Sr method (Halliday et al. 1980) and 406 Ma by the zircon (U-Pb) method (Pidgeon and Aftalion, 1978). The felsic dykes in particular show a close compositional relationship to, and some cut, the plutonic rocks, leading to the suggestion that these different intrusive suites are petrogenetically related (Henney, 1991; Lintern and Floyd, 2000). The youngest lamprophyre dykes locally cut the Lower Old Red Sandstone conglomerate (Rock and Rundle, 1986) and may have been emplaced contemporaneously with Lower Devonian volcanism. Radiometric ages from the younger members of the dyke swarm (Rock et al., 1988) and the syntectonic relationships of the older parts of the swarm suggest that emplacement occurred over a period of some 40 Ma (Figure 22).

Lamprophyre petrography

The lamprophyres are typically highly porphyritic, characterised by the presence of large euhedral crystals, mostly of hornblende or mica, within a finer grained groundmass (Plate 30a), (Plate 30b). Olivine, more abundant in mica-lamprophyre, is restricted to phenocrysts and is pseudomorphed by chlorite and carbonate. Pyroxene, generally a pale green diopsidic clinopyroxene, is relatively rare but where present forms euhedral phenocrysts, megacrysts or cumulophyric clusters. Felsic segregations (up to 10 per cent of the rock and in some cases substantially more) form globular structures (Rock, 1984), veinlets and irregular patches, which are either sharply defined or gradational into the host rock (Barnes et al., 1986). Coarse-grained 'appinitic' segregations may occur in hornblende-lamprophyre. They are variably altered, with the development of chlorite, carbonate, epidote, sericite and minor opaque minerals, sometimes precluding precise classification. These altered lamprophyres locally possess a weakly to moderately well-developed cleavage defined by thin foliae of chlorite and opaque minerals (± sericite). Pseudomorphs after possible biotite phenocrysts within these cleaved rocks have also undergone intracrystalline defor- mation suggesting that alteration may have occurred during cleavage development.

Xenoliths and inclusions are in general uncommon, although some lamprophyre dykes contain xenoliths of quartz. These are well displayed in a thick spessartite dyke south of East Culvennan [NX 307 652] in which angular to very irregular xenoliths up to 20 cm, apparently of vein quartz, are locally abundant (Plate 31). One dyke exposed near Garheugh [NX 268 501] has been found to contain an unusual assemblage of rounded to irregular xenoliths of a variably altered or metamorphosed, foliated to non-foliated dioritic/tonalitic rock (Phillips, 1996; Floyd and Phillips, 1999).

Mica-lamprophyre Mica-lamprophyre (Plate 30a) is typically fine-grained, although biotite phenocrysts may be large (5–10 mm) and are usually recognisable in hand specimen. A well-developed pilotaxitic fabric is defined in some examples by shape aligned phenocrysts. Because of the fine grain size of the matrix feldspar and alteration, the feldspar composition cannot be determined in the majority of mica-lamprophyre samples. However, where it is recognisable it is plagioclase, and the lamprophyre is thus kersantite. Felsic segregations and globular structures (defined by Rock, 1984), composed of quartz, feldspar, carbonate, epidote, pyrite and chlorite, are abundant.

Hornblende-lamprophyre Hornblende-lamprophyre (spessartite) varies from fine to coarse grained, and is composed of randomly orientated hornblende phenocrysts (> 20% of the rock) in a fine-grained groundmass of plagioclase (oligoclase to andesine), minor interstitial K-feldspar and quartz (Plate 30b). Hornblende forms subhedral to euhedral prisms and laths in the coarser grained rocks, but tends to be more acicular in finer grained lithologies. Felsic segregations are common, grading from the groundmass of the host lamprophyre and composed of the same minerals or including quartz, carbonate, epidote, chlorite and pyrite.

Appinite Appinite is the coarse-grained equivalent of the hornblende-lamprophyre, and generally forms the thicker dykes and segregations within spessartite. It is petrographically similar to the hornblende-lamprophyre, although non-porphyritic and with mafic and felsic segregations commonly well developed. Additional textural features include myrmekite, complex reaction relationships between green and brown hornblende, pyroxene and biotite, quartzose segregations and resorbed hornblende phenocrysts.

Felsic lithologies

The felsic dykes, including microdiorite, microgranodiorite and microgranite, consist of plagioclase, amphibole, biotite and quartz phenocrysts in varying modal proportions in a fine- to very fine-grained quartzo-feldspathic groundmass. They commonly contain abundant pale phenocrysts of feldspar that are typically shape-aligned parallel to the dyke margins. The 'porphyrite-porphyry' classification of these rocks (described by Barnes et al., 1986) as used on the geol-ogical maps of the district is here replaced with the more modern terminology of Le Maitre (1989).

Phenocrysts within the more mafic microdiorite ('porphyrite' of Barnes et al., 1986) are predominantly of plagioclase with variable amounts of amphibole, biotite and, in some cases, clinopyroxene. Subhedral to euhedral hornblende phenocrysts are commonly altered to actinolite and chlorite (± Fe-Ti oxides). Clinopyroxene, where present, occurs as corroded, anhedral crystals. The more evolved, more leucocratic microdiorite ('acid porphyrite' of Barnes et al., 1986) contains little or no amphibole, with more K-feldspar and quartz in the groundmass. Biotite is more abundant as sparse phenocrysts and, more commonly, as small flakes within the groundmass. The microgranodiorite ('porphyry' of Barnes et al., 1986) is characterised by phenocrysts of quartz and K-feldspar with biotite as the principal mafic phase (Plate 32). K-feldspar is common in the groundmass where it occurs as isolated crystals and in micrographic intergrowth with quartz. Hydrothermal alteration during and subsequent to emplacement has resulted in the sericitisation of feldspar, chloritisation of the mafic phases (both phenocrysts and groundmass), and silicification of the groundmass to varying degrees.

The early, cleaved, felsic dykes have been deformed by a homogeneous to domainal fabric. Locally well developed within the groundmass of the dykes, it is defined by aligned white mica and thin foliae of chlorite (± opaque minerals) and wraps around plagioclase phenocrysts (Plate 24). It may itself be deformed by narrow shear bands, which define a variably developed extensional crenulation cleavage or S-C fabric. Asymmetrical pressure shadows developed upon the feldspar phenocrysts are composed of cryptocrystalline quartz/feldspar and are locally replaced by apparently later carbonate. In some samples, plagioclase and, where present, quartz phenocrysts are relatively undeformed and may locally possess well-developed crystal faces indicating that deformation was preferentially partitioned into the groundmass. However, plagioclase phenocrysts, in particular, may exhibit a preferred shape alignment parallel to the shear fabric developed within the groundmass. In contrast to plagioclase, chlorite pseudomorphs after biotite phenocrysts are highly deformed and sheared out into this fabric. These microtextural relationships support field evidence indicating that emplace- ment of the dykes occurred during regional deformation.

Younger basic dykes

The younger basic dykes are fine-grained, essentially aphyric rocks consisting of small randomly orientated plagioclase laths with interstitial, granular clinopyroxene and opaque oxide. Pyroxene and possible interstitial glass are variably replaced by dark green chlorite. Rounded to irregular amygdales present within these dykes are composed of very fine-grained to cryptocrystalline chlorite and carbonate.

Geochemistry

Samples were selected for analysis from the least altered minor intrusions so that the full petrographical range of compositional variation was covered; representative analyses and summary data are given in Appendix 6. A wide range of silica content is present (Figure 29), from 52 to 61 per cent in the lamprophyres and up to 76 per cent in the felsic rocks. The lamprophyres are generally intermediate in composition but with high alkali (K2O, Na2O) for their silica content ((Figure 30)a) indicating that they are high-K calc-alkaline to shoshonitic in character ((Figure 30)b). The high alkali content causes the lamprophyres to be misclassified as trachyandesite ((Figure 30)a); they are quartz-bearing calc-alkaline rocks, and are consequently basaltic andesite or andesite. Some of the 'felsic' rocks also possess a high alkali content and exhibit some compositional overlap with the lamprophyres, but can be classified as dacite to rhyolite.

Samples classified petrographically as appinite and spessartite (hornblende-lamprophyre) are compositionally very similar (Figure 31). Mica-lamprophyre samples are generally compositionally similar to the spessartite-appinite suite but are discriminated by their lower SiO2, Al2O3 and Na2O content, and higher CaO and, in particular, P2O5 content. These differences are reflected in MORB (mid-ocean ridge basalt) normalised multi-element diagrams (Figure 32) and suggest that the mica- and hornblende-lamprophyres were derived from two separate, if closely related, magma types (Barnes et al., 1986; Rock et al., 1988). Both are significantly enriched in low field strength elements (Sr, K, Rb, Ba, Th) and, to a lesser extent, some high field strength elements (Nb, P2O5) and show minor depletion of Ti and Y. This pattern is typical of K-rich rocks in both island-arc and continental within-plate settings (Pearce, 1982, 1983; Rock et al., 1986b).

Lamprophyre composition may vary significantly with grain size and texture in a single intrusion, particularly within larger dykes. A suite of analyses from a broad dyke exposed at Rocks of Garheugh (Barnes et al., 1986) suggests that this variation generally lies well within the compositional range of the lamprophyre suite as a whole (Figure 31). Overall, the two lamprophyre suites define a single steep trend for most elements plotted against SiO2 (Figure 31). Two relatively felsic segregations from the Garheugh dyke have compositions at or just beyond the silica-rich end of the lamprophyre trend, but are compositionally distinct from the felsic dykes (Figure 31).

The various felsic dykes show a wide variation from microdiorite with around 61% SiO2 through microgranodiorite to microgranite with values up to 76% (Figure 29). An early study of samples from the west of the district (Barnes et al., 1986) suggested a silica gap at 58 to 63% SiO2 between two separate parental magmas. Subsequent data for the Culvennan Fell dykes and intermediate to felsic rocks exposed in the east of the district, have partly filled this gap. However, there is still a relative paucity of dyke rocks with SiO2 in the range 61 to 64%, and little compositional overlap, with respect to SiO2, between the lamprophyric and felsic suites.

Selected major oxides (e.g. Fe2O3T {where T indicates total Fe}, MgO, CaO, TiO2) and trace elements (V, Y) in the felsic suites form an apparent continuation of the lamprophyre trend with respect to SiO2 (Figure 31), with their distribution being controlled by the fractional crystallisation of mafic minerals. Other oxides (Al2O3, Na2O, K2O, P2O5) and trace elements (Rb, Zr, Ni, Cr), however, clearly discriminate between the felsic and lamprophyre suites which define two discrete trends indicating that they formed from different parental magmas (Figure 31). It is noteworthy that many of the S1 foliated felsite dykes are compositionally very similar to the granodiorite and associated dyke swarm as exposed at Kirkmabreck, south of Creetown [NX 477 564].

Diorite dykes, locally abundant at Culvennan and occurring sporadically in the north-east of the district, as well as a single example of an 'andesite' at Burrow Head, are petrographically enigmatic in relation to the other dyke suites. However, geochemically they appear to form a more basic component of the felsic dyke suite (Figure 31). This is supported by the MORB-normalised multi-element diagrams which show that the dioritic and felsic rocks possess a very similar pattern of enrichment that is clearly distinct from that of the lamprophyre suites.

Larger Diorite to granite intrusions

A number of relatively large dioritic to granodioritic intrusions, some composite, occur within the district together with part of the larger Cairnsmore of Fleet pluton (Figure 1). These range from a few tens of metres to 2.5 km in longest dimension and from equant to dyke-like in outcrop. The larger intrusions are described individually below. No radiometric ages are available for most of these intrusions. Field relationships suggest that they are broadly coeval with the post-tectonic dykes and predate emplacement of the Cairnsmore of Fleet granite. They are, therefore, likely to be late Silurian or early Devonian in age.

Four small granodioritic intrusions occur close to the western end of the Cairnsmore of Fleet granite within the Moniaive Shear Zone. Two intrusions, an irregular body about 150 m wide and a 25 m-thick 'dyke', occur close together in the vicinity of the granite near Talnotry [NX 478 704]; to the south near Bargaly, two ovoid intrusions (both less than 70 m in diameter) are situated about 1.5 km from the granite [NX 466 670]; [NX 472 665]. All these intrusions were intensely deformed within the shear zone and subsequently metamorphosed by the Cairnsmore of Fleet granite, with recrystallisation of quartz and development of biotite overprinted by fine acicular metasomatic amphibole.

Two small granitic intrusions are associated with the Cambret Fault south of the Cairnsmore of Fleet granite, at Cambret [NX 519 567] and near Arkland [NX 565 593] (Figure 28). The former is a poorly exposed biotite-granite up to 100 m thick between strike-parallel margins but the length along strike is unknown. A thermal aureole extends about 130 m north of the intrusion. The Arkland dyke is composite, comprising biotite-granite in the west and feldspar porphyritic microgranodiorite with micrographic groundmass in the east, the two phases interfingering in the middle. Along strike from the Kirkmabreck intrusion, a small area of hornfels exposed on Glenquicken Moor [NX 517 584] indicates a further concealed intrusion within or close to the Laurieston Fault zone.

Glenluce diorite

The Glenluce diorite forms two main outcrops, at Corsehead [NX 202 568] and Balcarry [NX 203 560] with a small outcrop near Barlockhart [NX 210 563]. The Balcarry outcrop trends along the strike of bedding in the host rocks but the others are perpendicular or oblique to strike. With the exception of a single isolated exposure [NX 203 568] of coarse-grained quartz-diorite, they are all composed of pink or grey feldspar porphyritic pyroxene-diorite. In the central part of the Corsehead body and at the western end of the Balcarry body, the proportion of pyroxene (forming crystals up to 8 mm at Balcarry) increases into a core of dark grey, relatively mafic diorite or andesite.

The biotite hornfels aureole follows the form of the three intrusions. It ranges in width from about 50 m either side of the western end of the Balcarry body to more than 300 m east and north of the small outcrop of the Barlockhart intrusion. Outwith the aureole, a single exposure of a breccia 500 m east of the Corsehead intrusion [NX 208 568] contains clasts of a highly altered basic rock within a microgranite matrix. There is a strong similarity to the Culvennan intrusion breccia, described below.

Petrography

The porphyritic andesite within the central part of the Coreshead body is a fine-grained rock comprising randomly orientated plagioclase, clinopyroxene, orthopyroxene and minor biotite phenocrysts in a very fine-grained to cryptocrystalline groundmass of plagioclase, quartz and possible K-feldspar. Anhedral to weakly subhedral plagioclase phenocrysts (0.4–1.7 mm in length) are zoned and twinned with minor patchy sericitic alteration. Pale green to colourless clinopyroxene forms anhedral to rarely subhedral microphenocrysts (0.2–0.4 mm), which occur either in isolation or as glomerophyric aggregates of several small crystals. Clinopyroxene is only locally rimmed by hornblende and/or biotite, and may partially include earlier pale brown orthopyroxene. Both pyroxene and hornblende are locally altered to, or pseudomorphed by chlorite and actinolite.

The pyroxene-rich diorite from the western end of the Balcarry body is a fine- to medium-grained rock that is largely composed of anhedral plagioclase laths that are variably aligned, defining a moderately developed igneous foliation. Large (1.5–2.0 mm long), intergranular brown to brown-green hornblende crystals contain rounded to irregular relict clinopyroxene, orthopyroxene and opaque minerals. Pale green clinopyroxene cores to smaller hornblende crystals are common. Primary hornblende is variably replaced by, or recrystallised to green hornblende or actinolitic hornblende.

The pyroxene-diorite that dominates the Glenluce intrusion comprises large (< 4 mm) anhedral hypersthene and smaller clinopyroxene (< 2 mm) phenocrysts in a groundmass of plagioclase (< 0.5 mm, 60% of the rock), biotite (< 0.1 mm, 10%), pyroxene and interstitial quartz (5%) (Plate 33). Both clinopyroxene and to a lesser extent hypersthene, show varying degrees of alteration to pale green amphibole. Locally, plagioclase forms subhedral, unzoned phenocrysts (< 6 mm in length), which form up to 30 per cent of the rock. Finer grained examples of the plagioclase porphyritic diorite also contain small, granular phenocrysts of hypersthene and clinopyroxene. Accessory minerals include granular opaque oxides (typically associated with biotite) and apatite.

The quartz-diorite is characterised by large (< 8 mm) crystals of subhedral plagioclase (about 50% of the rock) intergrown with subhedral, brown amphibole (< 4 mm, about 35%). The plagioclase is zoned from labradorite (An54) to andesine (An35) in composition, and exhibits patchy alteration to sericite. Minor pyroxene is present, comprising large (< 2 mm), anhedral hypersthene and smaller clinopyroxene crystals; the latter are variably replaced by pale green amphibole. Interstitial quartz, varying from 2 to 15 per cent of the rock, forms anhedral crystals and a micrographic intergrowth with small amounts of K-feldspar. Accessory minerals include apatite, opaque oxides, epidote and secondary carbonate.

Culvennan diorite

The Culvennan intrusion was investigated by Parker et al. (1981), as a potential site of copper mineralisation, and by Stephens and Halliday (1984). The main igneous body covers an area of approximately 1.5 km2 but, from the irregular shape of the enclosing biotite hornfels aureole and several outlying pods of diorite, its form in the near subsurface may be significantly different. Most of the exposed intrusion comprises a uniform dark grey pyroxene-biotite-diorite, although the western end of the main body was described as tonalite by Parker et al. (1981). Near the southern edge of the intrusion, on Bennan Hill [NX 327 649], a small exposure of a coarse-grained granitic lithology is surrounded by the diorite into which it was probably intruded.

An intrusion-breccia (Plate 34) exposed on the summit of Culvennan Fell [NX 311 650] was described in detail by Parker et al. (1981). It forms an oval outcrop about 100 m long within the thermal aureole, about 400 m west of the main diorite body. Although it is hosted by greywacke and interbedded siltstone, clasts in the breccia are dominated by silty mudstone and siltstone. Clasts of quartzose sandstone are also present but were not derived from a local source. The majority of clasts are subangular to rounded and less than 15 cm in length, but some angular blocks are up to 1 m long. The highly feldspar-porphyritic, biotite-microgranite matrix is vuggy with carbonate, hematite and crystalline quartz in cavities. The breccia occurs within the aureole of the diorite and both the clasts and probably also the matrix are metamorphosed, suggesting that the breccia predated this intrusion. The breccia is cut by porphyritic microgranite dykes which do not appear metamorphosed; their relationship to the diorite is unknown.

Petrography

The main intrusion is composed of a biotite-bearing pyroxene-diorite (Plate 35) in which anhedral to granular clino- and orthopyroxene are variably replaced by pale green amphibole. Pyroxene forms up to 30 per cent of the rock, clinopyroxene as large (1.5–3 mm) skeletal to sieve-textured crystals and subordinate orthopyroxene typically finer grained (0.25–0.5 mm). Rare olivine occurs in some rocks where it is mantled by clinopyroxene, both showing replacement by amphibole. Plagioclase (andesine to labradorite) laths (about 2 mm in length; 45–55%) exhibit a preferred shape alignment and define a locally well-developed pilotaxitic fabric (Plate 35). Biotite (< 10%) forms ragged crystals up to 2 mm. Minor to accessory minerals include interstitial quartz, opaque oxides and apatite.

The tonalite described by Parker et al. (1981) is a medium-grained, non-porphyritic, homogeneous, non-foliated granitic rock, consisting of zoned plagioclase (andesine), quartz, biotite and amphibole. The granite exposed at Bennan Hill is composed of large (up to 12 mm) microcline crystals (40% of the rock) which are intergrown with finer grained, anhedral quartz (15%). Plagioclase (oligoclase, 35%) forms large crystals up to 6 mm across. Anhedral, brown biotite (5%) is the main mafic mineral and may exhibit minor alteration to chlorite.

Two smaller bodies exposed to the south of the main diorite intrusion [NX 337 638] and [NX 340 642] (Figure 1) are composed of relatively fresh, two pyroxene-diorite similar to the main intrusion. In contrast, the small bodies exposed to the west of the main intrusion [NX 308 651] and [NX 309 653] are more highly altered.

Carsluith intrusion

The Carsluith intrusion is bedding parallel, near vertical and composite. It is up to 200 m thick at its western end as exposed, but wedges out over a strike length of 700 m. It comprises an outer carapace (up to 70 m thick) of grey plagioclase-porphyritic microdiorite enclosing a white microgranodiorite body up to 100 m thick. Both phases are well exposed in Bagbie Quarry in Carsluith [NX 489 549] (Barnes and Fettes, 1996) and are cut by an irregular, 30 to 40 cm-thick, fine-grained dolerite dyke thought to be Palaeogene (Tertiary) in age. The host Cairnharrow Formation is hornfelsed over a relatively wide zone (up to 250 m) north of the intrusion, which appears to be much narrower (< 100 m) to the south, although exposure is limited.

The contact between the two phases of the main intrusion, exposed on the north side of Bagbie Quarry, is sharp and steeply dipping, though irregular in detail, with the microgranodiorite becoming finer grained close to it. In the south side of the quarry, two vertical slabs of microdiorite (2–4 m thick) are aligned near-parallel to the length of the intrusion and are interpreted as screens between microgranodiorite sheets. In the south-west of the quarry, the microgranodiorite contains platy xenoliths of mudstone, up to several metres thick, which are also aligned parallel to bedding outwith the intrusion, again suggesting emplacement of the intrusion as a series of dykes. The contact between the microdiorite and the host rock is well exposed in Carsluith Burn, which flows along the southern edge of the intrusion. It is generally parallel to the vertical bedding but sporadic northward steps across bedding are associated with thin, bed-parallel dykes.

Steeply dipping joints are well developed in both phases of the intrusion. Joints in the microgranodiorite are predominantly north-east-trending, commonly with a quartz veneer containing slickensides pitching 20 to 60° north-east. North-west-trending joints are also developed in the microgranodiorite but are dominant in the microdiorite in which they are locally closely spaced.

Petrography

The outer feldspar-porphyritic microdiorite is characterised by subhedral, zoned plagioclase microphe- nocrysts (< 2 mm long, forming about 40% of the rock) in a fine-grained (about 0.1 mm) groundmass of interlocking quartz, plagioclase, K-feldspar and biotite. Plagioclase has andesine composition and exhibits minor alteration to sericite. Biotite (> 5%) occurs as scattered, randomly orientated tiny (about 0.1 mm) flakes and irregular aggregates (< 0.5 mm in length). Minor secondary carbonate and epidote occur. In the inner, coarser grained part of the dykes the plagioclase phenocrysts are enclosed by micrographic quartz-feldspar intergrowths.

The main part of the intrusion (Plate 36) is composed of a coarse-grained microgranodiorite consisting of large, zoned anhedral to subhedral plagioclase (oligoclase, 55–65% of the rock) within a finer grained (0.25–0.5 mm) groundmass of anhedral interlocking quartz, plagioclase, K-feldspar and biotite. Plagioclase exhibits preferential sericitisation of the core of crystals and later replacement by small muscovite flakes. Biotite (1%) occurs as ragged flakes and aggregates (< 2 mm long), and is variably replaced by chlorite. Microcline (> 5%) is common within the groundmass, occurring as anhedral crystals (< 1 mm across). Quartz (> 20%) occurs as anhedral crystals which exhibit an undulose extinction. Accessory minerals include granular to subhedral sphene, apatite and epidote/clinozoisite.

Kirkmabreck microgranodiorite

The Kirkmabreck (Creetown) microgranodiorite is a dyke, about 150 m thick, dipping approximately 50° north-north-west and extending for at least 2 km parallel to bedding in the host rocks. It is interpreted as having been emplaced into the Laurieston Fault at 390 ± 12 Ma (Rb-Sr whole-rock method; Greig, 1971). The intrusion has been extensively worked (Plate 1a) in several quarries (Watson, 1911), a complete section being exposed in the only quarry still active, adjacent to the A75 near Kirkmabreck [NX 480 565] (Barnes and Fettes, 1996). Here, evidence for a complex emplacement history is preserved within the wall rocks to the north of the dyke (Barnes and Fortey, 1997). A biotite hornfels aureole is apparent macroscopically, over a width of about 100 m around the intrusion, although it may be more extensive to the north, joining with that of the Cairnsmore of Fleet granite. Bedding and cleavage in the host rocks have been progressively overprinted within the aureole, being barely perceptible close to the intrusion. Early development of the hornfels is demonstrated by its relationship to a demonstrably later phase of skarn alteration and quartz carbonate veining (described below), which itself predates the intrusive rocks.

The first phase of intrusion formed a swarm of closely spaced, anastomosing dykes (Plate 37), ranging from 0.30 to 4 m thick. These are composed of fine-grained, grey porphyritic microdiorite with a strong pilotaxitic fabric in all but thin chilled zones against the sharp contacts with the host rock. The dykes are geochemically similar to the microgranodiorite of the main intrusion (see p. 92), but represent a slightly less evolved magma. However, they were cut and metamorphosed by the microgranodiorite, which locally contains xenoliths of the dyke rocks.

The microgranodiorite is light coloured and generally uniform, although locally it contains a perceptible igneous fabric. It is cut by a few irregular aplite dykes in the higher quarries. It is well jointed, with the dominant sets perpendicular and parallel to the length of the body. Fluid migration through these joints has locally caused reddening of the wall rocks over a narrow zone, although more extensive reddening and kaolinitic alteration are associated with faulting.

Petrography

The porphyritic microdiorite dykes contain abundant subhedral, oscillatory zoned andesine microphenocrysts (< 0.5 mm) and rare hornblende phenocrysts within a very fine-grained, devitrified groundmass. Hornblende has typically been replaced by aggregates of fine-grained biotite. Thermal metamorphism associated with the emplacement of the later microgranodiorite resulted in the widespread development of biotite within the groundmass, where it has mimetically grown along the pre-existing igneous fabric. This fabric is now defined by fine-grained biotite and optically aligned quartz and feldspar within the groundmass. The fabric wraps around subhedral to euhedral, zoned plagioclase phenocrysts which are themselves variably aligned parallel to this foliation, but lack any evidence of intracrystalline deformation. Although recrystallised, patches of micrographic intergrowth and rare relict plagioclase feldspar crystallites are preserved within the groundmass. Localised metasomatic alteration resulted in the development of secondary muscovite, actinolite, epidote and calcite.

The main intrusion is composed of medium- to coarse-grained, aphyric biotite-microgranodiorite, comprising an inequigranular assemblage of plagioclase, quartz, biotite, K-feldspar and rare hornblende (Plate 38). Accessory minerals include apatite, allanite and zircon. Equant to prismatic, randomly orientated plagioclase (50–60% of the rock) and subordinate biotite crystals (5–10%) form an open primary crystal framework. Interstices are filled by anhedral to equant quartz with minor amounts of interstitial K-feldspar. Plagioclase has oligoclase composition and forms subhedral to anhedral, weakly zoned crystals that exhibit preferential sericitisation of their cores. Rare, subpoikilitic, intergranular K-feldspar crystals are also present. Biotite is brown to brown-green in colour and forms anhedral to weakly subhedral flakes that are variably altered to muscovite and chlorite. Localised hydrothermal alteration of the microgranodiorite resulted in the development of a secondary assemblage comprising white mica, carbonate, hematite, chlorite and clinozoisite.

Cairnsmore of Fleet pluton

The Cairnsmore of Fleet pluton is the most evolved of the major granitic plutons in the Southern Uplands and is the youngest. Dates recorded include: 392 ± 2 Ma (Rb-Sr whole-rock, Halliday et al., 1980), 390 ± 6 Ma (U-Pb zircon, Pidgeon and Aftalion, 1978), 390 ± 6 Ma (K/Ar biotite, Brown et al., 1968); 397 ± 2 Ma (U-Pb zircon and monazite, Appendix 7). The outcrop of the pluton is ovoid in form, 18 3 11 km, with its long axis parallel to the regional strike of the country rocks. North of the district, its northern boundary lies close to the trace of the Orlock Bridge Fault (Floyd, 1999) and to the south it coincides with the Gillespie Burn Fault, but between these it crosscuts several major tract-bounding faults (Figure 1). Geophysical studies (Parslow, 1968; Parslow and Randall, 1973; Chapter 3) suggest that the pluton is a steep-sided intrusion down to 10 to 12 km and lies above a magnetic basement (Kimbell and Stone, 1995).

The outcrop of the pluton comprises two phases, dominated by a coarse-grained, feldspar-phyric granite, which encloses a core of fine- to medium-grained aphyric granite. The junction between these is sharp but very irregular, locally producing a complex pattern of coarse- and fine-grained varieties, with xenoliths of one within the other showing that the fine-grained variety is later (Parslow, 1968, 1971). There are very few country rock xenoliths present within the granite, but it is locally cut by numerous aplite and quartz veins.

The detail of the margin of the granite is exposed at several localities around the western end of the pluton in the district (Barnes and Fettes, 1996). The contact is seen cutting across bedding and the Moniaive Shear Zone fabric in Graddoch Burn [NX 490 646] and at McClaves Pantry [NX 496 660]. The granite in the main mass and veinlets and apophyses adjacent to the contact cut the shear zone foliation and are not deformed. To the south-east, the cliff at Door of Cairnsmore shows a large screen of metasedimentary rock a few tens of centimetres thick and parallel to bedding to the south, but about 200 m into the granite. This, together with the mapped form of the contact to the east where it interleaves with Moffat Shale Group along the Gillespie Burn Fault, suggests that at least this part of the granite was emplaced as a series of sheets. Exposures around Talnotry show that the north-western margin of the pluton forms a series of large steps that 'plunge' down the dip of bedding. In general, the contact dips north-north-west, parallel to the bedding and shear zone foliation, over sections of tens to hundreds of metres long; between these sections it cuts up at a high angle to bedding.

Petrography

The Cairnsmore of Fleet pluton is compositionally zoned, comprising a marginal coarse-grained, feldspar-phyric facies enclosing a central fine- to medium-grained aphyric facies, both of which can be classified as true granites (Le Maitre, 1989). The boundary between the two phases is irregular (Cook, 1976), but generally sharp. Blocks of the coarse-grained granite within the medium-grained facies indicate that the latter was a later intrusive phase. Parslow (1968) divided the outer coarse-grained facies into a marginal biotite-granite and an inner biotite-muscovite-granite. The boundary between the two types is gradational and is defined by an overall decrease in modal biotite and antithetic increase in primary muscovite away from the margin of the pluton. Large microcline phenocrysts ( 30 mm long), common in the marginal facies, decrease in size and number towards the centre of the pluton, with oligoclase becoming slightly more albitic in composition. Similarly, microperthite and myrmekite, which are common in the coarse-grained granite, are largely absent from the finer grained lithologies, whereas micrographic intergrowths increase towards the centre of the pluton. The intensity of later alteration, interpreted as resulting from a late stage metasomatic/hydrothermal event (Cook, 1976), increases towards the centre of the pluton leading to the development of chlorite from biotite, the sericitisation of feldspar and the growth of secondary muscovite, zoisite and epidote.

Locally, the granite exhibits a weak foliation defined by the aligned feldspar and mica crystals. Parslow (1968, fig. 1) demonstrated that the foliation occurs subparallel to the granite margin and increases in dip from subhorizontal near the centre to over 60° at the margin, where it is most intense. Varying degrees of deformation are associated with fabric development. Quartz, where deformed, possesses an undulose extinction and sutured margins. Feldspar crystals are locally fractured, but were apparently more resistant to deformation. Secondary mica has clearly overgrown these fractures indicating that the main phase of hydrothermal alteration post-dated fabric development.

Emplacement of the Cairnsmore of Fleet Granite Pluton

Field and petrographical relationships suggest that the coarse-grained granite was intruded first and began to fractionate before emplacement of the finer grained granite. The overall increase in the intensity and dip of the foliation developed within the granite, towards the margins of the pluton, is consistent with fabric development occurring as a result of forceful intrusion. However, the mode of emplacement of the Cairnsmore of Fleet pluton is uncertain. The cross-cutting nature of the eastern and western contacts, and absence of country rock fabrics parallel to the margin of the granite indicate that diapirism was not the main mechanism of emplacement. Evidence in the south of the intrusion suggests that this part at least was emplaced as thick sheets. The irregular western margin and the stepped bedding/fabric-parallel nature of the northern contact is consistent with stoping, although few xenoliths have been preserved at the level of exposure. Local deflection of the bedding and the fabric near to the western end of the intrusion points to some expansion of the pluton by ballooning. The pluton occurs entirely within the Moniaive Shear Zone and porphyroblasts which appear to have developed at an early stage of thermal meta-morphism are deformed by the ductile shear fabric (Barnes et al., 1995a). This suggests that early emplacement of the pluton may have been contemporaneous with the late stages of movement along the shear zone and was perhaps facilitated by the adjacent tract-bounding faults (cf. Hutton et al., 1990).

Geochemistry of the larger dioritic to granite intrusions

The larger intrusions include various rock types ranging from basic two-pyroxene-diorite with 49 to 58% SiO2 to granite with up to 76% SiO2. These rocks are calc-alkaline in composition (Figure 33) and form part of a suite of late Caledonian plutonic rocks in the Southern Uplands (Brown et al., 1979; Tindle and Pearce, 1981; Halliday, 1983, 1984; Stephens et al., 1985).

The more basic rocks, mainly in the Glenluce and Culvennan intrusions, show some compositional overlap and generally define a common linear trend (Figure 33); (Figure 34). However, the Glenluce diorite can be distinguished by its higher Fe2O3T and MgO, and lower K2O, Y and Ni contents (Figure 34). The more acid rocks from the Carsluith and Kirkmabreck intrusions lie on an extension of the diorite trend characterised by decreasing Fe2O3T, MgO, CaO and TiO2, and increasing Rb and total alkalis with respect to SiO2 (Figure 34). It is therefore possible that all of these intrusions were derived from a single parent dioritic magma with the fractionation of pyroxene and plagioclase, and later hornblende and plagioclase. However other elements (e.g. K2O, P2O5, Y and Zr) clearly show that the Carsluith and Kirkmabreck intrusions are geochemically distinct and lie on a separate trend at a high angle to that of the diorites, but parallel to that of the Cairnsmore of Fleet pluton.

Compared with the diorite and granodiorite intrusions, the Cairnsmore of Fleet granite is considerably richer in K2O, Rb, and Nb, and has lower MgO, CaO, TiO2, Sr and V contents (Figure 34). However, the granite from the Culvennan intrusion plots with the Cairnsmore of Fleet data on all of the variation diagrams. Field evidence suggests that this granite is an integral part of the Culvennan intrusion and was, therefore, derived by fractionation of the dioritic magma. If correct, this may be used to tentatively suggest a possible petrogenetic link between the Cairnsmore of Fleet granite pluton and the Culvennan and related dioritic intrusions. Volumetrically, however, this seems highly unlikely.

Petrochemical studies of the Cairnsmore of Fleet pluton (e.g. Cook, 1976; Fettes and Timmerman, 1992) have shown that it is a highly evolved, Ca-poor granite which exhibits albite-anorthite-orthoclase-SiO2-H2O systematics typical of subsolvus granites. The marginal coarse-grained granite defines an increasing fractionation trend inwards from the margin of the pluton, which is continued, with a slight overlap, by the finer grained granite (Figure 33); (Figure 34). High P(H2O) conditions during crystallisation resulted in the high mica content of the granite as well as the development of two primary feldspars and subsequent exsolution of albite (microperthite). Both the coarse- and fine-grained granites possess geochemical characteristics similar to S-type granites indicating that the parental magma was probably derived from the melting of Lower Palaeozoic or earlier sedimentary material. Isotopic systematics support an S-type model for the Cairnsmore of Fleet pluton with 87Rb/86Sr ratios of 0.706 to 0.707 and 18O values of 11.17 to 11.33 per cent (Halliday et al., 1980). The ∈Nd values of 22.4 to 23.0 are slightly higher than, but still comparable to, the values of 24.2 to 26.7 reported for the host Gala Group greywacks (Stone and Evans, 1995). The 87Rb/86Sr and 87Rb/86Sr isotopic ratios indicate that the Cairnsmore of Fleet pluton represents the most evolved member of the group of three major Southern Uplands plutons (Halliday et al., 1980).

Evolution of the intrusive suite

Comparison of the intrusions in the district with other parts of the regional swarm described from elsewhere in the Southern Uplands (Barnes et al., 1986; Rock et al., 1986b; Rock et al., 1988) shows that they are compositionally very similar. For example, the Kirkmabreck, Carsluith and Arkland intrusions are comparable to the sheared/cleaved felsite dykes and granodiorite dykes and appear to lie on a single trend on the majority of the variation diagrams (Al2O3, Fe2O3T, CaO, P2O5, TiO2, Rb, Sr, Zr, V, Y) (Figure 34). These intrusive rocks also display the same pattern of enrichment on MORB-normalised multi-element diagrams (Figure 32). The only minor exception is the Carsluith intrusion which contains higher K2O, but lower Na2O than the other granitic intrusions and dyke rocks (Figure 34). This may be due to the alteration of feldspar and/or the variable growth of secondary white mica. The granodiorite and microdiorite dykes may also provide a petrogenetic link between the more basic Culvennan and Glenluce diorites and the more evolved Carsluith and Kirkmabreck bodies. A similar petrogenetic link between the felsic dykes and the larger Criffel– Dalbeattie pluton was suggested by Henney (1991) and Lintern and Floyd (2000). The Culvennan and Glenluce diorites clearly plot at the more basic end of the geochemical trend defined by the dyke rocks (Figure 34). Although the dioritic rocks show some compositional overlap with the mica- and hornblende-lamprophyres, the lamprophyric rocks are compositionally distinct and define steeply dipping trends on a number of variation diagrams (Figure 34). This supports the conclusion of Barnes et al. (1986) that the microdioritic to felsic dykes of the district do not represent simple lamprophyre fractionates. Furthermore, this indicates that two separate magmas were available, facilitating the simultaneous intrusion of both the lamprophyric and felsic dyke suites.

The possible petrogenetic link between the Kirkmabreck, Carsluith, Glenluce and Culvennan intrusions and the various microdioritic to microgranitic dykes poses a number of questions concerning the age of emplacement of these larger bodies. These intrusions are thought to largely postdate regional deformation within the Southern Uplands, with the Kirkmabreck micrograno- diorite having been dated at about 390 Ma (Rb-Sr whole-rock; Greig 1971); this is comparable with age data (c. 395 Ma; Halliday et al., 1980; Appendix 7) obtained for the Cairnsmore of Fleet pluton. They are, in general, intruded into, or associated with, the main tract-bounding faults. The apparently primary igneous fabric developed within the microdiorite dykes associated with, and chemically related to, the Kirkmabreck intrusion may be interpreted as having developed in response to syn-emplacement deformation (compare with pre-full crystallisation fabrics of Hutton 1988), with similar fabrics also being developed within the main microgranodiorite body. These fabrics could be interpreted as recording the reactivation of the tract-bounding Laurieston Fault during emplacement. However, the marked geochemical similarity between the Kirkmabreck suite and parts of the S1-foliated felsic dyke swarm suggests that emplacement of at least the Kirkmabreck dykes and possibly also the main intrusion may be much earlier than had previously been considered, possibly during the later stages of regional deformation.

On a plot of Rb against Y+Nb, data from the Kirkmabreck, Carsluith, Glenluce and Culvennan intrusions fall within the field of volcanic-arc granites, with the more evolved Cairnsmore of Fleet granite plotting on the boundary between this and the field of syn-collisional granites (Figure 35). By the time of their intrusion the Iapetus Ocean had effectively closed, whichever geotectonic model is applied to the Southern Uplands. Consequently, an island-arc setting for the Kirkmabreck, Carsluith, Glenluce and Culvennan intrusions is unlikely, with these intrusions possibly having been partially derived from a subduction-enriched mantle source. Although there is considerable evidence of sedimentary parentage for the magma, it is possible that the initial melts were derived from a deep-seated low 87Sr/86Sr source. The magma may have differentiated and been progressively contaminated with sedimentary material during its ascent (Halliday et al., 1980). Rock et al. (1986b) demonstrated that the Southern Uplands lamprophyres comprise three components:

Consequently, although the various minor Caledonian intrusions exposed in the district may not be directly related, they may share a similar petrogenesis.

The geochemistry of the lamprophyres requires a deep mantle origin located some several hundred kilometres behind any subduction zone (Barnes et al., 1986; Rock et al., 1986b). This suggests that the 'Iapetus Suture', at the time of dyke emplacement, was much farther south than its currently inferred position, and places the Southern Uplands in at least a back-arc situation. Rock et al. (1986b) concluded that the lamprophyric dykes within the Southern Uplands were derived from one of two sources:

Consequently, it is possible that the source area was already contaminated by an earlier subduction component, hence, effectively divorcing the lamprophyric and felsic dykes from an active island-arc-subduction system. Recently, Vaughan (1996), working on comparable lamprophyric to shosonitic dykes in eastern Ireland (cf. Barnes et al., 1986), suggested that dyke emplacement was coeval with sinistral transtension. The lamprophyric dykes were emplaced along the Reidel shears, with the shoshonitic dykes being intruded along the primary shears developed within this overall sinistral strike-slip regime. Vaughan (1996) argued that transient decompression of a metasomatised mantle source during simple shear facilitates the release of volatile-rich potassic magma, which ascends to crustal levels along steep strike-slip faults. A similar model can be applied to the Southern Uplands lamprophyric and microdioritic to microgranitic dykes. Field relationships clearly indicate that these dykes were emplaced into active fault zones, which developed parallel to the Reidel shear directions within a sinistral strike-slip regime active during regional D3 deformation.

Contact metamorphism and skarn alteration

Thermal aureoles occur around all of the larger intrusions in the district and generally represent simple recrystallisation of the host rocks. Such aureoles are typically regarded as being due to the heating of the host rocks by conductive transfer of heat from the cooling body of magma (e.g. Kerrick, 1991). However, relationships between different phases of alteration and intrusion in the aureoles around the Cairnsmore of Fleet pluton and the Kirkmabreck granodiorite suggest a more complicated process.

The absolute width of an aureole at any point may vary substantially depending on the method used to define it. The mapped aureole is based on visual detection in the field of the onset of a purple tint to sandstone caused by the development of microscopic biotite in the matrix. Thin sections may, however, show some biotite development beyond this limit. A hardening and annealing of mudstone may also be apparent and more extensive than the biotite hornfels. More sensitive techniques, such as illite crystallinity (Chapter 6) show that relatively weak thermal effects may extend for consider- able distances from the outcropping intrusive rocks (Figure 26).

Biotite hornfels

The width of a biotite hornfels aureole, as defined in the field, generally varies from a few tens of metres to 1 km around the small to intermediate size bodies, and up to several kilometres around the Cairnsmore of Fleet pluton. Aureole width is not always in proportion to the size of intrusions and the exposed intrusive bodies may not lie symmetrically within their aureole. Particularly irregular aureoles occur around the Culvennan diorite and the Cairnsmore of Fleet pluton, suggesting that the near subsurface form of the intrusions may be quite different from that of their outcrop. Geophysical data shows this to be the case for the very extensive aureole south-west of the Cairnsmore of Fleet pluton (Chapter 3). Parts of the Cairnsmore of Fleet aureole, particularly north of the granite outcrop, are less than 800 m wide. This is similar to the narrower parts of the aureole around the much smaller Culvennan diorite, although the latter may have been emplaced at higher temperature. The aureoles around the narrow Carsluith and Kirkmabreck intrusions are, on the other hand, relatively wide.

Petrographically, it can be seen that the sedimentary host rocks have undergone pervasive recrystallisation during thermal metamorphism. This generally resulted in the development of fine-grained biotite, feldspar (oligoclase and andesine) and minute granules of rutile in the matrix of sandstone and more pervasively in finer grained, more pelitic rocks. Biotite generally occurs as very fine-grained, anhedral to rounded flakes, which, in cleaved rocks, mimetically overgrow the pre-existing fabric. With an increase in metamorphic grade, biotite becomes coarser grained, forming larger subhedral porphyroblasts, which may overprint original clastic texture in the inner part of an aureole. In general, although recrystallised, the original sedimentary texture and/or cleavage can still be recognised even within the highest grade parts of the aureole. Fine-grained actinolitic amphibole may also be present, particularly in the inner parts of an aureole, with sieve-textured poikiloblasts of hornblende and, in some rocks, clinozoisite immediately adjacent to intrusive rocks.

The aureoles around the Cairnsmore of Fleet pluton and the Kirkmabreck microgranodiorite are more complex. Throughout the Cairnsmore of Fleet aureole, biotite forms abundant small (< 5 mm) clots. These may represent pseudomorphs after cordierite porphyroblasts, chlorite foliae where they occur within the Moniaive Shear Zone or, in a few cases, mudstone lithic clasts. Cordierite is well developed in the hornfels at several localities associated with Moffat Shale Group inliers. It is best seen in Culchronchie Burn [NX 510 638] and on the adjacent Culchronchie Hill [NX 515 637]. Lenticular cordierite psuedomorphs, up to 2 cm in length, are flattened and show a weak, gently plunging lineation in the pervasive foliation (Plate 39). In thin section it is clear that the cordierite, pseudomorphed by very fine-grained white mica, and small garnet porphyroblasts (Plate 40) predates at least part of the deformation. Both the cordierite and the fabric are overprinted by the biotite hornfels. In northern parts of the aureole, fine (< 2 mm) quartz veins are abundantly developed parallel to the pre-existing Moniaive Shear Zone fabric but are not strained and, therefore, also appear to be associated with the metamorphism. In the southern part of the aureole, well displayed around Mark [NX 507 607] and Pibble [NX 516 606], the hornfels shows well-developed domainal schistosity with 1 to 5 mm mica-rich layers between quartzose bands up to 2 cm thick, giving the rock a strongly banded appearance (Plate 41). Retrogression of the biotite hornfels has occurred in western parts of the aureole with biotite having been altered to chlorite.

Skarn alteration

In the Cairnsmore of Fleet and Kirkmabreck aureoles, the dark coloured biotite hornfels locally shows evidence of metasomatic (skarn) alteration to a distinctive pale green calc-silicate assemblage in lenses/bands with gradational contacts. In the Cairnsmore of Fleet aureole, this occurs between Cairnsmore Farm [NX 479 639] and the granite contact 2.5 km to the east. Here, oblate lenses, 1 to 5 cm thick and 20 cm to several metres in length, of calc-silicate generally occur parallel to both bedding and foliation in the metasandstone. These calc-silicate lenses are also well exposed at a number of localities in Graddoch Burn, south-east- wards from the contact with the Cairnsmore of Fleet pluton [NX 496 645] to [NX 476 637] where they are cut by the granite. The lenses are composed of epidote and amphibole that clearly overgrew and replaced earlier-developed biotite. In a single exposure close to the granite at Talnotry [NX 488 716] in the Loch Doon district to the north, smaller lenses (up to 2 cm thick) of calc-silicate are composed of fine-grained granular diopsidic pyroxene.

The aureole of the Kirkmabreck granodiorite includes a 20 m-wide zone of metasomatic alteration in which the biotite-hornfels contains numerous planar to anasto- mosing, parallel-sided to lenticular, pale green skarns, which predate the microgranodiorite dyke swarm (Plate 42a), (Plate 42b). The skarns range from a few millimetres to 2 cm in width and locally enclose garnet-bearing quartz veinlets. The skarns are fine grained and comprise an inner zone rich in salitic clinopyroxene, which passes gradationally through an outer zone of clinozoisite, acti- nolite and minor quartz into the host biotite- and amphi- bole-bearing hornfels. Veinlets enclosed within these calc-silicate bands are composed of grossular garnet (Gr82–And18 to Gr90–And10), quartz, clinopyroxene, clinozoisite and minor titanite. Large (up to 5 mm), typically euhedral, garnet crystals are birefringent and exhibit sector twinning and oscillatory zonation (Plate 42c). Locally they extend into the wall rock and contain inclusions of fine-grained, granular clinopyroxene (± clinozoisite). Textural evidence indicates that there were several successive stages of metasomatic alteration, all of which postdated the formation of the host hornfels.

Formation of contact metamorphic aureoles

The sequence of events recognised in the Cairnsmore of Fleet and Kirkmabreck aureoles suggests that their formation involved a more complex sequence of events than the simple model of conductive heating from an emplaced body of magma (e.g. Kerrick, 1991) would allow. The Cairnsmore of Fleet aureole was superimposed on a broad zone of intense foliation which is continuous with the Moniaive Shear Zone outwith the aureole to the east. The foliation deforms early cordierite and garnet porphyroblasts but was overprinted by the biotite hornfels and locally altered metasomatically to form skarn phases prior to final emplacement of the outer parts of the granite.

A similar, but more detailed sequence of events can be established for the Kirkmabreck aureole. The hornfels, developed at an early stage, was altered locally to form the calc-silicate layers. These were then cut by several generations of fracture-hosted quartz-carbonate veins (Barnes and Fortey, 1997), probably associated with an influx of meteoric water. A swarm of porphyritic microdiorite dykes was emplaced subsequently; the chilled margins of these dykes suggest that the aureole had cooled by the time they were intruded. The dykes were followed by the main intrusion, causing thermal alteration of the dyke rocks and possibly modifying the extent of the thermal aureole.

In both intrusions it is apparent that at least part of the 'contact' metamorphism occurred before emplacement of the intrusion, which would normally be supposed to have caused it. Although early metamorphism may have been initiated by precursor intrusions for which no other evidence now remains, this seems an unsatisfactory explanation. Instead, the sequence of events apparent in these aureoles suggests that metamorphism occurred as a continuing process ahead of the granitic magma as it rose through the upper crust. At Kirkmabreck, the ascent of the magma body appears to have been arrested beneath the present exposure level, allowing alteration and veining of the hornfels prior to a further pulse of intrusion. Such a model may also account for other features apparent in the district, such as the variable width of individual aureoles, which in some cases are not concentric with their intrusion, and also the variable size of aureoles relative to the size of intrusions.

Palaeogene dykes

Palaeogene (Tertiary) dolerite dykes occur sporadically throughout the district. Generally they have northerly trends and are typically a few metres thick. The dyke margins are chilled with the central portion being composed of coarse-grained dolerite with ophitic to subophitic clinopyroxene (augite, 20–45% of the rock) enclosing plagioclase (labradorite to bytownite, 40–60%). Pseudomorphs after olivine (in varying modal proportions) are composed of bowlingite, chlorite and/or serpentine. Accessory minerals include opaque oxides, spinel and rare analcime. Amygdales, filled with carbonate, chlorite or zeolite, are common in the finer grained margins of the dykes. These dykes are similar to the 'Tertiary' dykes described by Harrison et al. 1987). This provides a likely maximum age for most of the dolerite dykes in the district, which are thought to be members of the Arran and/or Mull swarms.

Chapter 8 Upper Palaeozoic rocks and later geological history

Post-Silurian rocks are preserved only offshore in the district although Devonian to Triassic rocks also occur onshore in adjacent areas (Stone, 1995; Akhurst et al., 1997; McMillan in Lintern and Floyd, 2000) and early Jurassic rocks occur in the Solway and East Irish Sea basins to the south (Jackson et al., 1995). With no preserved rock record, the younger depositional and tectonic history can only be assessed from more regional considerations (e.g. Chadwick et al., 1994).

As continental collision between Laurentia and Avalonia removed the final vestiges of the Iapetus Ocean during the late Silurian, the ductile deformation front migrated southwards. There is no clear record in the Southern Uplands of the early Devonian (Acadian) deformation of the rocks deposited on the Avalonian margin (e.g Soper et al., 1987). The leading edge of the Avalonian plate was thrust beneath the Southern Uplands from the late Silurian onwards so that by the early Devonian there was considerable tectonic uplift and rapid erosion. Coarse-grained clastic deposits of the Lower Old Red Sandstone were deposited with marked unconformity on an irregular surface, but were themselves tilted and eroded and are only preserved locally. Regional extension from late Devonian times led to the development of major sedimentary basins to the north and south of the Southern Uplands in the Midland Valley (e.g. Leeder, 1982) and Northumberland–Solway trough (Chadwick et al., 1995; McMillan and McAdam, 1996) respectively. Associated reactivation of strike-parallel and cross-strike faults in the Southern Uplands (e.g. Anderson et al., 1995; McMillan and Brand, 1995; Stone, 1995) formed small basins, which preserve a fragmentary record of fluviatile, estuarine and marginal marine sedimentation from Devonian to late Carboniferous times. Localised volcanism occurred in the early Carboniferous in southern Scotland and in Cumbria (e.g. Leeder, 1982). These deposits were in turn gently folded and/or tilted and variably eroded during late-Carboniferous to early Permian times as the region experienced the peripheral effects of the Variscan Orogeny.

Renewed reactivation of dominantly north-trending faults occurred as a result of east–west extension during the Permian and Triassic, with more regional subsidence leading to the development of the major Solway and East Irish Sea sedimentary basins (e.g. Jackson et al., 1987, 1995; Jackson and Mulholland, 1993) to the south of the Southern Uplands. Extensive breccia fans, locally associated with volcanic rocks (e.g. in Southern Scotland, Brookfield, 1978; Stone, 1988; McMillan, 2002), were deposited during the Permian in close association with active faults. Late Permian breccias pass laterally into aeolian sandstone in the East Irish Sea basin. Breccia deposition continued on the fringe of the East Irish Sea Basin but a series of marine transgression-regression cycles led to deposition of evaporite within the basin, overlain by muddy siltstone deposited in ephemeral playa lakes. This sequence passes up, near the Permian–Triassic boundary, into the thick sequence of fluviatile and aeolian red sandstone that comprises the Sherwood Sandstone Group (e.g. Barnes et al., 1994; Jones and Ambrose, 1994). The succeeding mudrock and evaporite sequence of the Triassic Mercia Mudstone Group records a return to low-relief mudflat or playa deposition in a largely continental setting but with possible periods of marine influence (Arthurton et al., 1978; Talbot et al., 1994). At the top of the Triassic, a thin sequence of Rhaetian lagoonal and shallow marine deposits is known from the Carlisle area (Ivimey-Cook et al., 1995) and marks the transition to marine deposition. The Rhaetic rocks are conformably overlain by marine limestone and mudstone of the Lias Group at the base of the Jurassic. The latter, occurring in small outliers known in the Carlisle area onshore (Ivimey-Cook et al., 1995) and from seismic and borehole data offshore (e.g. Jackson et al., 1995) form the top of the preserved succession offshore in the vicinity of the district.

Regional considerations and depth of burial studies (e.g. Chadwick et al., 1994) suggest that the Triassic deposits may have extended over the present upland area of the Lake District in north-west England. These deposits are preserved in fault-bounded half-graben structures in the Southern Uplands (see below) where it is also likely they were originally much more extensive. Major post-Triassic faulting evident in the basins to the south of the Southern Uplands appears to have occurred during Jurassic to early Cretaceous times from stratigraphical considerations (e.g. Whitaker, 1985) and fault dating in the Sellafield area (Bailey, 1997). Extension at this time, leading to deposition of a thick sedimentary sequence in the East Irish Sea Basin, was followed by more regional subsidence and the probable widespread deposition of an Upper Cretaceous (Chalk) sequence (Chadwick et al., 1994). Regional uplift commenced in the latest Cretaceous or early Palaeogene (Tertiary) (e.g. Lewis et al., 1992) and has probably continued to the present, possibly accompanied in Miocene times by reversal of earlier normal faults. It has caused the removal of the younger cover from most of the Southern Uplands. Fault offsets in the district, shown for example by changes in metamorphic grade across some of the strike-parallel and cross-strike faults, are therefore the net effects of a long tectonic history.

Triassic rocks in the offshore area

Carboniferous and Permo-Triassic rocks have long been known from the shores of Loch Ryan to the north-west of the district and were thought to extend southwards beneath the Stranraer isthmus and Luce Bay. Geophysical surveys (Mansfield and Kennett, 1963; Bott, 1964) elucidated the half-graben structure of the basin, indicating up to about 2 km of Upper Palaeozoic strata adjacent to the Loch Ryan Fault along its eastern margin. Nothing is known of Permian strata offshore in the district, although it is likely that a variable thickness of breccia composed of locally derived basement material is present, comparable to the Loch Ryan Formation exposed on the shore in Loch Ryan (Stone, 1988, 1995) and the Brockram in west Cumbria (e.g. Akhurst et al., 1997). The presence of the Sherwood Sandstone Group in the continuation of the Loch Ryan Basin in Luce Bay in the district was, however, confirmed by red sandstone recovered from a shallow offshore BGS borehole (70/02, (Figure 1)). Red mudstone was penetrated in another borehole (73/47) south of the Scares (Jackson et al., 1995), suggesting that strata of the Mercia Mudstone Group, known in the core of the Solway Basin south-east of Burrow Head, may extend into the mouth of Luce Bay. A sliver of Mercia Mudstone Group may also be present along the eastern side of the bay, close to the Loch Ryan Fault (cf. Jackson et al., 1995, fig. 45).

The Scares, a group of small islands in the mouth of Luce Bay, are composed of Lower Palaeozoic rocks of the Hawick Group (Barnes et al., 1988), marking a relatively elevated block. The precise structural configuration of this block is unknown but it is likely to be due to post-Triassic reactivation of segments of one or more of the strike-parallel, tract-bounding faults in the Lower Palaeozoic basement to form a horst within the younger basin (cf. Stone, 1995).

Sherwood Sandstone Group has also been proved in Wigtown Bay in BGS borehole 71/63 (Jackson et al., 1995). The geometry of the Wigtown Bay basin is less well known, but from the straight eastern margin of the bay along a continuation of the Cree Fault (Chapters 5 and 6), parallel to the Loch Ryan Fault, it seems likely that it is a half-graben similar to the Loch Ryan Basin. Lower Palaeozoic rocks exposed on the western side of Wigtown Bay are pervasively reddened, suggesting that they lay at or just below the Permian erosion surface.

Chapter 9 Quaternary

Following the revelation in 1840 by Louis Agassiz that Scotland had been glaciated, convincing evidence for glaciation in Galloway was evocatively described by Jolly (1868). The primary geological survey in the 1870s confirmed many glacial phenomema and emphasised compelling evidence for marine inundation along the Wigtownshire and Kirkcudbrightshire coast (Geikie, 1878). During the first half of the 20th century the Quaternary geology of the district received little further attention apart from the extensive researches of Charlesworth (1926a, b) on the glacial deposits, landforms and source of erratics. During the last forty years the work of J B Sissons (reviewed by Ballantyne and Gray, 1984) and PhD research, in particular by Price (1961), Cutler (1979), Cornish (1980) and May (1981), has stimulated renewed interest in the late Devensian history of southern Scotland. Concurrently the research of Jardine (1967, 1975, 1980, 1981) has detailed the history of Flandrian sedimentation and relative sea level change along the north Solway coast. The implications of Jardine's interpretations of radiocarbon age determinations have been disputed, particularly in relation to sea level curves (e.g. Haggart, 1988, 1989; Sutherland, 1984). However, his work, together with studies of pollen (e.g. Moar, 1969; H H Birks, 1972; H J B Birks, 1989) and insect assemblages (Bishop and Coope, 1977), have thrown light upon the changing depositional environments and fluctuating climate during the early part of the Flandrian.

Summary of Quaternary history

Biostratigraphical and oxygen isotope evidence from cored ocean floor sediments indicates that there have been at least 16 major cold events during the Quaternary (the last 1.8 Ma) (Bowen, 1978; Shackleton and Opdyke, 1973, 1976; Boulton et al., 1991; Boulton, 1992). Some 14 climatostratigraphical stages of alternating glacial (cold) and interglacial (temperate) conditions are recognised in the British Isles (Boulton et al., 1992; Jones and Keen, 1993). It is unlikely that each of the cold events produced ice sheets and till deposits on the British landmass, where there is evidence for two major ice advances during the last 115 ka (thousand years) (McCabe, 1987; Bowen, 1989). The earlier glaciation developed at about 70 ka before present (BP) (within the Early Devensian substage) and the second at about 26 ka BP (at the beginning of the Late Devensian substage) (Rose, 1989; Boulton et al., 1991; Boulton, 1992; Gordon and Sutherland, 1993) (Table 2).

Ice sheet development during the Late Devensian substage was complex and has been modelled only in general terms (Boulton et al., 1977; Boulton, 1990; Boulton and Payne, 1994). Its effects were to remove or redistribute most deposits of previous Quaternary events in the ice-covered area. Because of this erosive action only the very latest events since about 26 ka BP are known with any degree of certainty for southern Scotland. Even during this period the record is incomplete, but from an assessment of geomorphological, lithological, biostratigraphical and sedimentological evidence the glacial and postglacial history may be reconstructed (Table 2) with reference to the following climatostratigraphical divisions:

Dimlington Stadial

The earlier of the two cold episodes during the Late Devensian substage, the Dimlington Stadial (Table 2) was associated with the build-up of glaciers in both the mountains of the Western Highlands of Scotland and in the western Southern Uplands in response to increasing precipitation and a cooling climate. During the first 3 ka of the Dimlington Stadial, these centres of ice accumulation nourished radiating patterns of expanding glaciers until Highland ice became confluent with Southern Uplands ice. Thereafter the whole of the Southern Uplands was overtopped by a single ice sheet. The ice sheet continued to expand and at its maximum extent some 22 ka ago covered most of the Scottish landmass and much of England. From theoretical reconstructions, Boulton et al. (1977) estimated that at the time of its maximum extent the relative elevation of the ice sheet surface over the Southern Uplands may have been in excess of 1600 m. More recent models indicate ice thicknesses in the order of 1000 m (Boulton et al., 1985) or 600 to 800 m (Lambeck, 1996).

The general pattern of ice movement over Wigtownshire and Kirkcudbrightshire during the Dimlington Stadial, as evidenced by glacial striae, roches moutonnées and drumlin forms, is shown in (Figure 36). Striations on well-preserved roches mountonneés, well displayed around Mochrum Loch [NX 300 535] and Castle Loch [NX 290 535], indicate an ice flow direction predominately towards the south to south-south-west (Geikie, 1878; Greig, 1971). These directions are generally confirmed by streamlined landforms throughout The Machars of Wigtownshire including drumlins, particularly on the low ground between Glen Luce [NX 199 574] and Kirkcowan [NX 329 610] (Charlesworth, 1926a; Cutler, 1979) (Figure 37). Well-formed drumlins composed predominantly of till are present around Kirkinner [NX 423 513], Sorbie [NX 436 469] and Wigtown [NX 434 554]. They generally form discrete, parallel ridges, but clusters of smaller interconnected ridges are also present. They range in height from 10 to 45 m and are commonly situated on bedrock and/or surrounded by peat bog. The distribution of erratic blocks and pebbles of granodiorite from the Cairnsmore of Fleet and Loch Doon plutons further confirms the south-directed flow of ice across the district (Charlesworth, 1926a; Sissons, 1967a; Greig, 1971).

Although the evidence outlined above (and summarised in (Figure 36)) confirms the general direction of ice flow during the Dimlington Stadial, it is likely on the basis of detailed analysis (particularly of drumlin orientations and till fabric and composition) that ice originated from two different sources, namely the Highlands and Southern Uplands. West of the district, on the Rhins of Galloway, Kerr (1982a) interpreted the presence of two sets of drumlins and two lithostratigraphically distinct tills (the lower of which is a shelly drift; Sissons, 1976) as the product of two distinct ice movements over the region. This interpretation was developed by Stone (1995), who confirmed that north- or north-west-trending drumlins and the lower shelly, lodgement till were associated with the Highland ice sheet whereas north-east-trending drumlins and the upper till (composed of sandy diamicton) were the product of ice flowing from the Galloway hills. The early research by Charlesworth (1926a) in the district indicated that drumlins orientated predominantly north-east to south-west and north to south lay within and outside kamiform sands and gravels. He considered that the kamiform deposits were deposited as moraine and concluded that they formed part of a discon- tinuous belt of similar deposits stretching from the Rhins of Galloway to the Nith valley, which he termed the Lammermuir–Stranraer Moraine. The moraine was considered the product of one or more readvances of ice following the main Late Devensian glaciation (Charlesworth, 1926b). This interpretation stood unchallenged for nearly fifty years and was used by Sissons, (1967a, b), together with evidence from other parts of Scotland (e.g. Simpson, 1933) and new radiocarbon dates, in the reconstruction of the Perth Readvance ice limit. Although Kerr (1982b) supported the concept that the drumlin field might be the product of a late readvance of ice from the Galloway ice cap following the deglaciation of the Irish Sea, many researchers (for a review see Sissons, 1974, 1983) have questioned and dismissed the readvance models on the grounds that there is no good morphological or stratigraphical evidence for a major readvance across the Scottish lowlands. For example, Cutler (1979), mapping in The Machars (Figure 2), concluded that Charlesworth's (1926a) morainic deposits were in fact the product of ice still-stand during final deglaciation, an interpretation supported by Sissons (1976). However, in view of the renewed interest in the concept of multiple readvances of 'Scottish' ice from the Southern Uplands impinging on the coast of west Cumbria (for a summary see McMillan in Akhurst et al., 1997), the possibility of small ice readvances from the Southern Uplands mountains affecting the Galloway lowlands following deglaciation of the Irish Sea cannot be entirely excluded.

Till, the product of glacial deposition, is the most widespread Quaternary deposit in the district. It is present only in drumlins and as thin, undulating sheets occupying generally low-lying areas of Wigtownshire in particular around Kirkcowan, Glenluce and Mochrum [NX 347 464], Newton Stewart [NX 410 655], Wigtown and Sorbie. East of Wigtown Bay, till is widely distributed only around Creetown [NX 475 585], on lower hill slopes of Cairnharrow [NX 534 561] and within the valley of the Water of Fleet (Figure 2). The deposits comprise sandy and clay-rich, stony and bouldery diamictons in which both the colour and composition of the matrix and the dominant clast lithologies reflect the substrate to the ice. Thus, till derived from the Lower Palaeozoic sedimentary rocks tends to be composed of a hard, tenacious clay matrix with subangular to rounded, hard sandstone and siltstone clasts, whereas in areas of granitic rocks, particularly around Cairnsmore of Fleet, a more gritty, sandy diamicton may predominate. The colour of the clay matrix may vary from grey to red depending on the content of red-weathering local bedrock. Ice-smoothed and striated pebbles and boulders are commonly found within the till. Locally, thin lenses of water-laid sand and gravel may be present within the till but otherwise the deposits are unstratified. Small areas mapped as moraine (IGS, 1981) are present on high ground north and north-east of Creetown around Graddoch Burn, Cullendoch Moor and Big Water of Fleet (Figure 2). The moraine comprises mounds and ridges of very poorly sorted gravelly sandy diamicton, dominated by granite clasts, with interbedded sand and gravel. These deposits show limited signs of water sorting (Goodlet, 1970) and are probably the debris product of stagnating ice at the margin of a retreating glacier around Cairnsmore of Fleet.

During deglaciation, prior to about 13 ka BP, glaciofluvial sand and gravel, deposited as outwash by meltwaters of the receding Dimlington Stadial glaciers, accumulated along the Solway coast and within the principal valleys of the district. Kames (moundy deposits) are present at a number of localities summarised below. Interpreted originally as representing ice readvance moraine (Charlesworth, 1926a), they are clearly the product of ice-contact and subglacial processes, and probably formed as a result of an ice still-stand during deglaciation (Cutler, 1979). The distribution of these deposits is shown on Geological Survey 1:50 000 Drift Edition maps (IGS 1981, 1982). Details of sections are described by Goodlet (1970).

In the Kirkcowan area [NX 330 610] kame deposits occupy ground north-east of Loch Ronald [NX 265 640]. Moundy terraces are also present around Shennanton [NX 343 633] and south of Kirkcowan along the sides of the River Bladnoch. Similar deposits, present between Challoch [NX 385 675] and Newton Stewart on the River Cree may have formed against ice occupying the centre of the valley. Typically, these deposits, which range up to 15 m thick, are water sorted and stratified sand and gravel with a range of grain sizes from fine-grained sand to coarse gravel with cobbles. Interbedded silt and thin diamictons are also characteristic of these ice contact deposits.

Windermere Interstadial

Before about 13 ka BP, the pollen record preserved in peat and lacustrine sediments indicates that a varied aquatic and semi-aquatic flora grew in and around pools left on the glacially moulded landscape. Tree cover, however, was very limited. By 13 ka BP, the record provided by assemblages of beetles (sensitive indicators of climatic change) preserved in organic sediments shows that a rapid amelioration of climate had taken place before the beginning of the Windermere Interstadial (Table 2). Beetle collections from organic lake silts and sands from sites in southern Scotland, dating from 13 ka to 12.5 ka BP, are indicative of temperatures similar to those of today (Bishop and Coope, 1977). Between 12.5 and 12 ka BP a period of rapid cooling is recorded in the coleopteran record (Bishop and Coope, 1977). The thermal maximum as shown by the pollen record occurred between 11.8 and 11 ka BP, an indication that vegetation recolonisation at the end of the glaciation lagged behind the climatic changes (Gordon, 1993).

The chronology of relative sea level changes in southern Scotland during the Windermere Interstadial is poorly known. Early rapid eustatic rise in sea level was compensated in due course by the slower process of isostatic recovery of the land. Lambeck (1996) computed the local sea level isobase at 13 ka BP, relative to present OD, as about 215 m. Thereafter, sea level rose intermittently to a Holocene maximum (Flandrian) at about 6 ka BP. Remnants of Late-glacial raised beach and estuarine deposits, comprising bedded silt, well sorted sand and fine gravel are preserved at elevations between 10 and 25 m above OD along the Wigtownshire and Kirkcudbrightshire coasts at several localities in the district (IGS, 1981; 1982). Well sorted, raised beach, shingly gravel deposits are present as pockets between Glenluce and Port William, between Newton Stewart and Bladnoch and south of Creetown. At Mains of Machermore [NX 423 645] a large kettlehole in the raised beach deposits may mark the position of a buried grounded iceberg.

Loch Lomond Stadial

Climatic deterioration culminated in the arctic conditions of the Loch Lomond Stadial (11–10 ka BP; (Table 2)) (Gray and Lowe, 1977) and evidence from studies of beetles in the eastern Solway Firth records this second sharp drop in temperature (see Windermere Interstadial above; Bishop and Coope, 1977; Gordon, 1993). In the Southern Uplands, corrie glaciers developed on a limited scale to the north-east of the district between Broad Law and Hart Fell (Price, 1963; May, 1981) and to the north of the district in the Galloway Hills (Cornish, 1981). Most of the south of Scotland, in common with other areas that remained unglaciated during the Loch Lomond Stadial, suffered arctic climatic conditions and the land was subjected to intense periglacial activity. Possible fossil frost wedges in glaciofluvial sands and gravels (Galloway, 1961; Watson, 1977) and solifluction deposits on hill slopes within the district testify to these effects.

Flandrian

The rapidity of climatic change at the beginning of the Flandrian (Holocene) (10 ka BP to Present, (Table 2)) is recorded in coastal deposits east of the district at Brighouse Bay [NX 635 455]. Here, present-day intertidal beach deposits overlie peat, yielding a radiocarbon age of 9640 ± 180 years BP, and containing a temperate insect assemblage indicative of a climate as warm or even warmer than that in south-west Scotland today (Bishop and Coope, 1977).

The colonisation of the Late Devensian, open and treeless land by birch (Betula) and hazel (Corylus) during the early Flandrian is recorded in pollen records from many sites in southern Scotland (Nichols, 1967; Moar, 1969; H H Birks, 1972; for summary see Price, 1983). By about 8.5 ka BP mixed deciduous woodlands including oak (Quercus), elm (Ulmus) and hazel had developed extensively. Pine (Pinus) began to grow in the Galloway hills by about 7 ka BP (H J B Birks et al., 1975; H J B Birks, 1989). As a wetter climate developed around 7.5 to 7 ka BP, peat started to accumulate on former lake floors and poorly drained flat land and continued to develop as ombotrophic Sphagnum mosses in raised bogs and in blanket peat on sloping ground. Between 7 and 6.5 ka BP, alder (Alnus) replaced willow (Salix) in wet habitats. From 5 ka BP to present, a slightly cooler, moister climate together with the impact of man resulted in forest decline (the so called 'elm' decline) and forests were replaced by blanket peat, heaths and grassland.

Several localities along the northern shores of the Solway Firth have provided evidence of relative sea level change during the Flandrian (Jardine, 1967, 1975, 1980, 1981; Jardine and Morrison, 1976, Haggart, 1989). During and immediately after the Loch Lomond Stadial relative sea level in the Solway Firth was lower than that of today (Jardine, 1980; Haggart, 1989). As relative sea level rose in response to eustatic rise exceeding isostatic rebound, early Flandrian peats that accumulated in coastal swamps were covered by carse (tidal flat and estuarine) deposits formed during the marine inundation of the Main Flandrian (Post-glacial) Marine Transgression (Jardine, 1980). Published provisional isobases (Sissons, 1976, 1983) for the Main Flandrian Shoreline describe an elliptic uplift dome centred on Rannoch Moor in the Western Highlands of Scotland and predict that isostatic uplift in the district was in the order of 8 m above OD.

The carse deposits comprise fine- to medium-grained sand and silt of former tidal flat, gulf and estuarine environments (Jardine, 1975, 1980). They form low-lying areas adjacent to the present high water mark of ordinary Spring tides. Included in 'Post-Glacial Raised Beach Deposits, undivided', on published Geological Survey 1:50 000 scale drift maps (IGS, 1981, 1982), they are overlain by recently deposited parallel-laminated, fine-grained sand, silt and clay of the present floodplain or saltmarsh. In the district the carse sediments occupy extensive tracts around Wigtown Bay. Raised beaches up to 10 m above OD commonly backed by steep wave cut sea cliffs are present at various localities. A narrow coastal strip of raised beach shingle gravels is present intermittently from Auchenmalg [NX 233 523] to Port William [NX 340 435]. The till or rock platform on which these deposits rest is locally exhumed. A particularly notable 2 km-stretch lies between Corwall Port [NX 276 490] and Chippermore Point [NX 292 478] in Luce Bay where the rocks have been smoothed and hollowed out into small caves by the action of waves (Plate 43). Between Creetown and Ravenshall Point [NX 520 525] in Wigtown Bay the raised beach deposits are backed by an old sea cliff.

At the head of Wigtown Bay, carse deposits with surface levels ranging from 7 to 10 m above OD occupy coastal ground between Newton Stewart and Wigtown and Creetown and the lower reaches of the Palnure Burn (Figure 2). Records of boreholes and foundations of a former railway viaduct put through the largely estuarine sequences indicate that they range from 7 to 25 m thick, and rest either directly on bedrock or on glaciofluvial sand and gravel. At the base of the estuarine deposits in the valley of the Palnure Burn, wood fragments have been age dated at 7960 ± 200 14C years BP (Jardine, 1975). This date, together with other dates from organic remains within the carse deposits, indicates that marine sedimentation in Wigtown Bay began between 7.5 and 7 ka14C years BP.

Sea level continued to rise until well after 6.6 ka BP. Although basal samples of peat and wood overlying estuarine clays at Palnure gave 14C ages ranging from 6540 and 6240 years BP, it is considered that peat formation began with the local development of a coastal marsh (Jardine, 1975). Elsewhere at the head of Wigtown Bay, estuarine sediments continued to be deposited until approximately 5 ka BP. These were overlain by brackish water sediments prior to later peat accumulation. The indications are that the end of the Flandrian marine transgression was later here than in the eastern Solway Firth (Jardine, 1975, 1980). Jardine's (1975) Flandrian sea level curve for the eastern Solway Firth (which he considered more reliable than that for Wigtown Bay) was reconsidered by Sutherland (1984) and Haggart (1988, 1989), who have shown that the available evidence does not necessarily conflict with a synchronous maximum for the Main Flandrian Transgression along the north coast of the Solway Firth. However it is accepted (Haggart, 1988) that small eustatically controlled fluctuations may have occurred during the period covered by the formation of the regional Main Flandrian Shoreline, and interacted with variable local uplift.

Peat mosses accumulated during the Late Flandrian over extensive tracts of lowland areas of Kirkcowan and higher ground to the north of Glenluce. Around Wigtown Bay, the Moss of Cree [NX 430 600] (for pollen profile see Moar, 1969), and Meikle Cullendoch Moss [NX 565 660] are the largest peat accumulations. Smaller bogs are present south of Whauphill [NX 404 498] and north-east of Creetown.

Alluvial tracts, comprising silt, sand and gravel, occupy the valley floors of the principal rivers and streams in the district, notably the Tarf Water, River Bladnoch, River Cree, Penkiln Burn, Palnure Burn, Big Water of Fleet and Little Water of Fleet (Figure 2). Small areas of lake alluvium are also present within enclosed hollows.'

Chapter 10 Metalliferous mineralization

Lead, zinc, copper, nickel and arsenic mineralisation is present in the district, and was worked at a number of sites in the 19th and early 20th centuries. Most of the mineralisation occurs within a 7 to 8-km wide, north-west-trending zone between Newton Stewart and Gatehouse of Fleet and within 6 km of the Cairnsmore of Fleet granite (Figure 38). The majority of the mineralised occurrences are epigenetic veins, emplaced in the Ordovician and Silurian metasedimentary rocks; most of the veins run parallel to the trend of the zone, although north- and north-east-trending examples also occur. In contrast, the Talnotry mineralisation occurs within diorite and is thought to have a magmatic origin. A more detailed account of the individual workings listed below is provided by Wilson and Flett (1921) and by Cook (1976).

Lead and zinc

East and West Blackcraig

The Blackcraig deposit, situated about 3 km south-east of Newton Stewart, was discovered in 1743 and worked intermittently until 1882 from mines at East and West Blackcraig [NX 447 645] and [NX 441 648]. These were the most important mines in the district. Output between 1853 and 1881 consisted of 3820 tons of lead and 1668 tons of zinc-blende, the bulk of which came from East Blackcraig. In addition, 28 tons of copper ore were raised from East Blackcraig in 1864–65. At the time of compiling their memoir (about 1919–20) Wilson and Flett (1921) recorded that the mines were under reinvestigation.

The mines were sited on a south-west-dipping structure; for most of its length the vein has a strike of 115° but towards its north-western end it swings round to 135°. The vein branches in the middle of the West Blackcraig workings [NX 441 648] with a northern arm following a more easterly course for about 200 m before turning parallel to the main vein. To the north-west, the main vein passes into a quartz vein that does not appear to have any metalliferous mineralisation. Three further quartz veins occur 100 to 500 m south-west of the mine workings.

The main vein occupies a shatter zone, which is up to 18 m wide. Within it the mineralisation extends over nearly the whole width of the structure and consists of strings and patches of galena, sphalerite and chalcopyrite in a gangue of broken country rock, calcite, dolomite, baryte and some quartz. A small pocket of chalcopyrite was recorded close to the East Blackcraig engine shaft. The mineralised structure extends along strike for about 1.2 km. However, Wilson and Flett (1921, fig. 7) indicated that the strike extent of ore was 490 m in West Blackcraig and 170 m in East Blackcraig, and that there was 170 m of unproven ground between the two mines. The same diagram shows that in both mines the ore extended to over 80 m in depth, and that along strike in both directions it passed from being galena- to sphalerite-dominated. There is evidence from the old workings that, for much of its length, the main vein is associated with an olivine dolerite, possibly Palaeogene (Tertiary) dyke, although this has not been recorded at surface.

Cairnsmore

The Cairnsmore mine [NX 466 635], located 2 km east-south-east of East Blackcraig, operated for at least 15 years in the mid 19th century when it produced a little over 3000 tons of lead ore. The deposit consists of two closely spaced veins dipping 60° towards 195°. They cut greywacke and shale, but are only productive within the latter. The main vein is between 0.3 and 1.5 m thick and consists principally of broken country rock with calcite, dolomite and baryte carrying galena and pyrite. The Cairnsmore veins lie on the easterly projection of the Blackcraig vein suggesting that they could be part of a major structure. Further evidence for the continuity of this structure is provided by records of trials for lead at Palnure Burn [NX 455 638], between East Blackcraig and Cairnsmore and 500 m east-south-east of the latter at Graddoch Burn [NX 472 633].

Pibble

Pibble Mine lies 7 km south-east of Cairnsmore and is one of four workings in the Pibble Hill area, which lie close to the south-easterly projection of the Blackcraig structure. The mine is situated within an outcrop of the Moffat Shale Group on Pibble Hill [NX 527 605], 5.5 km north-east of Creetown. No production figures have survived, but the workings comprised several shafts and levels and the ore was worked to a depth of at least 90 m. The vein, with a minimum length of 900 m, dips 75° towards 027°. The ore comprises galena, sphalerite and chalcopyrite together with linarite, cerussite, malachite, hemimorphite and pyromorphite. The vein consists largely of broken country rock, with quartz and baryte. Wilson and Flett (1921) specifically noted that graphitic shale on one of the dumps contained small crystals of sphalerite and galena.

Pibble Gulch

Considerably less is known about Pibble Gulch Mine [NX 522 615], situated 1 km north of Pibble. It worked a 1.8 m-wide vein (dip 80° towards 030°–040°), which can be followed for up to 800 m. The vein infilling is mostly broken country rock and quartz, but sphalerite in veins up to 10 cm thick is present on the dumps.

Rusco (Meikle Bennan)

The old workings are situated on the north-eastern slope of Meikle Bennan [NX 553 615], 3 km east of Pibble Gulch. The vein is around 0.3 m wide, dipping 70° towards 170°. It consists mainly of quartz, with a little pyrite, chalcopyrite, galena and sphalerite.

Coldstream Burn

This former mine, situated on the south side of the Coldstream Burn [NX 387 697], was worked extensively prior to 1919 by three shafts and three levels and was being reopened at the time of G V Wilson's visit in that year. The vein is 0.6 to 0.9 m wide, north-trending and near-vertical. It consists mainly of broken country rock with ribs of sphalerite and galena up to 10 cm thick.

Wood of Cree

The Wood of Cree Mine was opened around 1870, but little ore seems to have been produced at this time. At the time of G V Wilson's visit in 1919, it was in the process of being reopened and new machinery was being installed. The mine is situated on the east side of the River Cree [NX 387 695], some 200 m south-west of Coldstream Burn Mine. The ore occurs as strings and pockets within brecciated country rock in a north-north-west-trending shatter zone, which, where seen in the opencast section of the mine, is up to 4.5 m wide. The ore is zinc rich, comprising a complex intergrowth of sphalerite, galena, chalcopyrite and pyrite with 25.4 to 31.5% Zn, 5.9 to 12.9% Pb and 0.3% Cu. The distribution of ore within the vein is quite variable with enriched lenses that pinch out both vertically and horizontally.

Dromore

Dromore Mine is situated at the base of Pibble Hill [NX 538 622], 7 km north-east of Creetown. The site was opened as a copper mine just before the First World War, but seems to have contained more zinc than copper, with up to 10 cm of sphalerite at the footwall. The vein, dipping 80° towards 212°, is parallel to the structures at Pibble Gulch and Pibble. The structure is largely infilled with broken country rock, with sphalerite- and chalcopyrite-bearing quartz.

Minor occurrences

Lead veins at Bargaly [NX 466 682], Dallash [NX 471 693], Englishman's Burn [NX 480 587] and Chain Burn [NX 501 609] were also worked on a small scale prior to 1919. The latter vein occurs within a dextral wrench fault trending 110°, the mineralisation dominated by sphalerite and galena (Cook, 1976). Lead is also said to have been worked on Little Bennan Hill, and on the west side of the Big Water of Fleet near Upper Rusco [NX 566 612]. Quartz veins carrying traces of galena have been recorded at [NX 559 562] in the Skyre Burn, 4 km west of Gatehouse of Fleet.

Copper

Copper ore was formerly extracted in the Kirkcowan–Wigtown district at Drumruck, Lauchentyre, King's Laggan, East Blackcraig, Wauk Mill and Culchronchie. These workings appear to have been on much smaller scale than those for lead and zinc, and no production records are available. There is also a record of minor copper mineralisation to the south, in the Whithorn district, at Tonderghie.

Drumruck

The Drumruck Mine [NX 583 635] is situated on the east side of the Water of Fleet, 7.5 km north of Gatehouse of Fleet. The deposit was worked by three levels for a short period in the early part of the 20th century. It comprises a single vein, dipping 70° towards 195°, associated with a felsite dyke cutting Gala Group greywacke. The vein is 0.75 m wide and consists mainly of broken country rock, with thin ribs of quartz carrying a little chalcopyrite and malachite near the centre.

King's Laggan

The old mine is located on Doon Hill [NX 562 577], 90 m north-west of King's Laggan Farm and 6 km north-west of Gatehouse of Fleet. The vein, which is exposed just behind the farmhouse, dips 65° towards 235°. In common with most other veins in the district it consists predominantly of broken country rock with the ore present as disseminated strings and specks of chalcopyrite in quartz.

Lauchentyre

The former mine is located [NX 558 572] 750 m south-south-west of King's Laggan Mine and 6 km west of Gatehouse of Fleet. It appears to have been a relatively small-scale operation with one shaft and a short adit. The ore consisted of chalcopyrite, copper uranite, malachite and sphalerite, in a gangue of quartz, calcite and dolomite. As usual the structure is infilled mainly by broken country rock.

Culchronchie Burn

Cook (1976) described small workings for copper on three subparallel, east-trending veins in Culchronchie Burn [around 507 636], 5 km north-east of Creetown. The most extensive workings were on the middle vein and comprised a shaft, level and cross-cut adit. The mineralisation consists of chalcopyrite and pyrite in a quartz gangue.

Minor occurrences

A north-trending vein was formerly mined for copper at Wauk Mill [NX 333 603], 750 m south-east of Kirkcowan. Cook (1976) recorded minor chalcopyrite mineralisation in veins in a fault zone [NX 479 678] about 1 km south-east from the Bargaly lead mine. Farther south at Mary Mine, Tonderghie [NX 439 348], close to Carghidown on the coast between St Ninian's Cave and Burrow Head, working of minor chalcopyrite mineralisation was reported by Wilson and Flett (1921). The chalcopyrite mineralisation occurs in an approximately east–west-trending, brecciated zone, about 3 m across and with horizontal slickensides on its walls. The gangue minerals are baryte and quartz; some malachite is also present. Minor chalcopyrite is also seen in numerous small baryte veins within the Burrow Head to Port Castle Bay [NX 426 358] coastal section.

Following the recognition of porphyry-style disseminated copper mineralisation at Black Stockarton Moor [NX 723 551] to the east of the district, a number of small intrusions in southern Scotland, including those in the district at Glenluce, Culvennan, Mochrum and Kirkmabreck, were examined for indications of disseminated sulphide mineralisation (Cooper et al., 1982). Hydrothermal alteration was recorded at Mochrum and Glenluce; the latter was associated with common pyrite, and rare chalcopyrite and arsenopyrite. Faults in adjacent greywacke and mudstone have quartz and carbonate veins carrying disseminated pyrite and traces of chalcopyrite. A detailed geophysical and geochemical assessment of the area around the Culvennan diorite found no evidence of porphyry-style metal enrichment, but the geochemistry indicated the presence of weak local copper, arsenic, iron and lead mineralisation (Parker et al., 1981).

Nickel

The Talnotry nickel mine [NX 4785 7035] lies 8 km east-north-east of Newton Stewart, 230 m south-east of Talnotry. The mineralisation occurs at the southern edge of a sheet-like diorite body that intrudes Gala Group greywacke about 200 m west of the margin of the Cairnsmore of Fleet granite. The deposit was reputedly discovered around 1885, but although about 100 tons of ore were raised between then and 1900, little of this was actually removed from the site.

The unusual nature and possible origins of the deposit have attracted considerable interest, for example Russell (c. 1917), Wilson and Flett (1921), Jones (1924), Gregory (1928), Cook (1976), Parker et al. (1977) and Stanley et al. (1987). Russell visited the site shortly after the cessation of mining so his observations (quoted in Stanley et al., 1987) are probably the most comprehensive account of the deposit: 'He noted that the metamorphosed sediments formed the southern boundary of the deposit and that 3 m or so of massive pyrrhotine, hornblende, and minor chalcopyrite occurred immediately above the metasediments in the form of large irregular blocks. The massive pyrrhotite was separated from the diorite by an irregular but well defined line of demarcation, the diorite being traversed for a metre or so by stringers some of which formed considerable masses of intermixed nickel arsenides, chalcopyrite and pyrrhotine. The richest nickel ore formed against the diorite on the north side of the deposit'.

Wilson and Flett (1921) noted that the deposit was exposed over a 15 m vertical section and that it appeared to be lenticular in shape. Ground magnetic and VLF surveys carried out by IGS in 1977 confirmed that the extent of the mineralisation is largely as exposed, the body being 20 m long and 4 m thick, elongated in an east-south-east-direction (Parker, 1977).

Three types of ore mineral assemblage were recognised by Stanley et al. (1987), although there are only minor differences in the overall assemblages and in the sequence of deposition. They are nickelite-gersdorffite-rich, chalcopyrite-rich and pyrrhotite-rich. The pyrrhotite-rich assemblage occurs at the base of the diorite forming networks of blebs interstitial to the amphibole. Chalcopyrite-rich mineralisation forms stringers and lenses in non-mineralised diorite, and local enrichments in the pyrrhotite-rich ore. Nickelite-gersdorffite mineralisation occurs as blebs within plagioclase in amphibole-rich areas. The full list of ore minerals at Talnotry is: gersdorffite, nickelite, molybdenite, tellurobismutite, gold, pyrrhotite, pentlandite, sphalerite, argentopentlandite, chalcopyrite, pyrite, marcasite, violarite and goethite.

Arsenic

Talnotry Vein

The Talnotry Vein mine [NX 480 701] is located 200 m east of Talnotry Mine. The mine was opened at about the beginning of the 20th century and a few tons of ore were produced, although little of it appears to have been taken away. The vein has a north-east-strike and is north-west-dipping, appearing to follow the western margin of the Cairnsmore of Fleet granite. It is about 1 m wide and consists of scattered strings and patches of arsenopyrite in quartz gangue.

Glen of the Bar

Cook (1976) recorded a mine or trial for arsenic, 100 m west of Talnotry Mine, on a fault in Glen of the Bar [NX 477 703].

Molybdenum

Tandy (1974) recorded minor molybdenite mineralisation associated with quartz-muscovite veins in the marginal part of the Cairnsmore of Fleet Granite to the east of Talnotry, and in quartz and aplite veins within the southern contact zone of the granite between Culchronchie Burn [NX 510 639] and the Water of Fleet. Molybdenum drainage anomalies suggest that this type of mineralisation may be quite widespread within the intrusion (Leake and Brown, 1977).

Precious metals

Blackcraig Mines produced 4230 ounces of silver between 1854 and 1878, the bulk of which stemmed from East Blackcraig. There are no records of silver being recovered from any of the other mines, but assays of ore revealed the following concentrations of silver: Wood of Cree, 73 g/t; Cairnsmore, 107 g/t; King's Laggan, 3 g/t and Talnotry, 114–250 g/t. The Wood of Cree and Talnotry samples additionally contained 3 g/t gold. Stone et al. (1995) reported anomalously high gold values (up to 91.65 ppb) in stream sediment and pan concentrates from the vicinity of the Orlock Bridge Fault at Dervaird, 2 km east from Glenluce.

Metallogenesis

With the exception of the Talnotry nickel deposit, most of the significant base metal mineralisation in the Wigtown–Kirkcowan district is vein-hosted and demonstrably epigenetic. However, as in the catchment of the Penkiln Burn, to the north of the district (Stone et al., 1984), there is some evidence of disseminated sulphide mineralisation in country rocks (e.g. within the outcrop of the Moffat Shale Group at Pibble), which raises the possibility that this could, in part, be the source of the epigenetic mineralisation.

The epigenetic mineralisation forms part of the Fleet orefield (Cook, 1976), which covers approximately 300 km2 of the Fleet pluton and the country rocks around its western end, and contains over 25 veins hosting the most diverse suite of minerals in the Southern Uplands. There appears to be no direct evidence of a relationship between the vein mineralisation and Caledonian igneous rocks other than an association with a felsite dyke at Drumruck. However, Cook (1976) established that the distribution of the mineralisation has a tri-zonal arrangement clearly related to the exposed areas of the pluton. The arrangement comprises an inner zone, covering the granite and adjacent parts of the aureole, dominated by copper ores in a siliceous gangue and including high temperature hydrothermal minerals such as molybdenite, arsenopyrite and pentlandite. This assemblage grades outwards into a middle zone characterised by sphalerite, ankerite and calcite in a quartzose gangue. Deposits within this middle zone include Dromore, Rusco, Pibble Gulch, Wood of Cree, Coldstream Burn and Chain Burn. The outermost zone, as evidenced by the veins at Blackcraig, Pibble, Bargaly, Cairnsmore, Dallash and Englishman's Burn, is dominated by lead minerals with a carbonate gangue and minor baryte. Cook's zonal arrangement works well to the south of the pluton but is less well defined to the west. Nevertheless, the zonal distribution of base metals is mirrored in the drainage geochemistry, and the area of vein mineralisation south and south-west of the intrusion is roughly coincident with a zone of lithium enrichment in drainage (Leake and Brown, 1977), which overlies Parslow and Randall's (1973) gravity-inferred subsurface cupola of the Fleet Granite.

Gallagher (1958) concluded that the Blackcraig mineralisation was probably younger than the associated olivine-dolerite dyke, which he regarded as being Permo-Carboniferous in age. However, since no other dykes of that age have been recognised in this district, and the trend of the Blackcraig dyke (115°) lies close to the mean strike (122°) of six Palaeogene (Tertiary) dolerite dykes exposed on or close to the coast south-east of Creetown, it seems more likely that this dyke is part of the 'Tertiary suite'. If this is so, the Blackcraig mineralisation cannot be older than Palaeogene. However, a definitive age cannot be established since Gallagher's (1958) evidence did not unequivocally establish the relationship between the mineralisation and the dyke.

There is little doubt that the molybdenum mineralisation occurred during a late stage in the emplacement of the Fleet Granite. However, the relationship of the Talnotry nickel deposit to the host diorite is more conjectural. Russell (c. 1917) and Wilson and Flett (1921) considered the deposit to have a magmatic origin, whereas Gregory (1928) thought that it was hydrothermal, the fluids having been transported along a fault at the lower contact of the diorite. Cook (1976) agreed with Gregory's conclusion and postulated that the arsenic mineralisation at Glen of the Bar and at Talnotry Vein was continuous with the Talnotry Mine deposit and part of the same vein system. Stanley et al. (1987) supported Wilson and Flett's view that the most compelling evidence for an origin by magmatic segregation was that the deposit was at the base of the diorite sheet. They cited the homogeneity of the sulphide blebs and the low concentrations of characteristically hydrothermal elements (e.g. Pb, Zn and Ba) as further evidence of a magmatic origin, and pointed out that Ir and Pd concentrations in the pyrrhotite-rich and chalcopyrite-rich assemblages are typical of the range of values found in magmatic nickel deposits. Stanley et al. (1987) concluded that the Talnotry nickel deposit was unrelated to the arsenic mineralisation, but that the ultimate source of the As and S in both cases may have been the Lower Palaeozoic sedimentary rocks.

Information sources

Further geological information held by the British Geological Survey relevant to the district is listed below. It includes published material in the form of maps, memoirs and reports and unpublished maps and reports. Also included are other sources of data held by BGS in a number of collections, including borehole records, fossils, rock samples, thin sections, hydrogeological data and photographs.

Searches of indexes to some of the collections can be made on the Geoscience Index System in BGS libraries or the Data Index on the BGS web site. This is a developing computer-based system which carries out searches of indexes to collections and digital databases for specified geographical areas. It is based on a geographical information system linked to a relational database management system. Results of the searches are displayed on maps on the screen. The indexes which are available include:

Enquiries concerning geological data for the district should be addressed to the National Geoscience Information Service, BGS, Edinburgh.

Maps

Sheet Surveyor Date
NX 15 NE RPB 1984
NX 16 NE AAJ, PS 1987
NX 16 SE AAJ, RPB 1984
NX 17 SE PS 1987
NX 24 NE RPB 1984
NX 25 NW RPB 1984
NX 25 NE RPB 1986
NX 25 SW RPB 1984
NX 25 SE RPB 1984
NX 26 NW JDF 1986
NX 26 NE JDF 1986
NX 26 SW RPB 1986
NX 26 SE RPB 1986
NX 27 SW JDF 1987
NX 27 SE JDF 1987
NX 33 NE RPB 1984
NX 34 NW RPB 1984
NX 34 NE RPB 1984
NX 34 SW RPB 1983
NX 34 SE RPB 1984
NX 35 NW RPB 1986
NX 35 NE RPB 1986
NX 35 SW RPB 1986
NX 35 SE RPB 1986
NX 36 NW JDF 1986
NX 36 NE JDF 1986
NX 36 SW JDF 1986
NX 36 SE RPB 1988
NX 37 SW JDF 1987
NX 37 SE JDF 1987
NX 43 NW (including part of SW) RPB 1984
NX 43 NE (including part of SE) RPB 1984
NX 44 NW RPB 1987
NX 44 NE RPB 1985
NX 44 SW RPB 1985
NX 44 SE RPB 1985
NX 45 NW RPB 1988
NX 45 NE RPB 1988
NX 45 SW RPB 1987
NX 45 SE RPB 1988
NX 46 NW RPB 1989
NX 46 NE RPB 1989
NX 46 SW RPB 1988
NX 46 SE RPB 1989
NX 47 SW PS, RPB 1989
NX 47 SE RPB 1989
NX 54 NE BCL 1986
NX 55 NW RPB 1988
NX 55 NE RPB, BCL 1988
NX 55 SW RPB 1988
NX 55 SE RPB, BCL 1988
NX 56 NW DJF 1989
NX 56 NE BCL 1987
NX 56 SW RPB, DJF 1989
NX 56 SE RPB, BCL 1987
NX 57 SW RPB, DJF 1989
NX 57 SE DJF, BCL 1989

The current availability of these can be checked on the BGS web site at:

http://www.bgs.ac.uk/products/digitalmaps/digmapgb.html

Publications

Memoirs, books, reports and papers relevant to the district are arranged by topic. Some publications are out of print but may be consulted at BGS and other libraries.

Economic geology and mineralisation

Cooper, D C, Parker, M E, and Allen, P M. 1982. Investigations of small intrusions in southern Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No. 58.

Dines, H G. 1922. Barytes and witherite. Special Report on the Mineral Resources of Great Britain, Memoir of the Geological Survey, Vol. 2. Third edition.

Leake, R C, Brown, M J, Smith, T K, and Date, A R. 1978. A geochemical drainage survey of the Fleet granitic complex and its environs. Mineral Reconnaissance Programme Report,Institute of Geological Sciences, No. 21.

Leake, R C, Rollin, K E, and Shaw, M H. 1996. Assessment of the potential for gold mineralisation in the Southern Uplands of Scotland using multiple geological, geophysical and geochemical datasets. Mineral Reconnaissance Programme Report, British Geological Survey, No. 141.

Macgregor, M, Lee, G W, and Wilson, G V. 1920. The iron ores of Scotland. Special Report on the Mineral Resources of Great Britain, Memoir of the Geological Survey, Scotland, Vol. 11.

Parker, M E. 1977. Geophysical surveys around Talnotry Mine, Kirkcudbright, Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No. 10.

Parker, M E, Cooper, D C, Bide, P J, and Allen, P M. 1981. Mineral exploration in the area around Culvennan Fell, Kirkcowan, south-western Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences. No. 42.

Wilson, G V. 1921. The lead, zinc, copper and nickel ores of Scotland. Special Report on the Mineral Resources of Great Britain, Memoir of the Geological Survey, Scotland, Vol. 17.

Biostratigraphy

There is a collection of internal biostratigraphical reports. Those which include information on graptolites are listed in Appendix 1.

Bulk Minerals

Goodlet, G A. 1970. Sands and gravels of the southern counties of Scotland. Report of the Institute of Geological Sciences, No. 70/4.

Cameron, I B. 1977. Sand and gravel resources of the Dumfries and Galloway region of Scotland. Report of the Institute of Geological Sciences, No. 77/22.

Smith, C G, and Floyd, J D. 1989. Scottish Highlands and Southern Uplands mineral portfolio: hard-rock aggregate resources. British Geological Survey Technical Report, WF/89/4.

Environmental baseline geochemistry

British Geological Survey. 1993. Regional geochemistry of southern Scotland and part of northern England. (Keyworth, Nottingham: British Geological Survey.)

Hydrogeology/geochemistry

Gauss, G. 1969. Records of wells in the areas of Scottish one- inch Geological Sheets Kirkmaiden (1), Whithorn (2), Stranraer (3), Wigtown (4), Kirkcudbright (5) and Annan (6). Water Supply Papers of the Geological Survey of Great Britain, Well Catalogue Series.

Robins, N S. 1990. Hydrogeology of Scotland (London: HMSO for the British Geological Survey).

Documentary collections

Borehole record collection

BGS holds collections of records of boreholes which can be consulted at BGS, Edinburgh, where copies of most records may be purchased. For the district the collection consists of the sites and logs of about 620 boreholes.

Most were drilled in search of water supplies, with the balance divided between metalliferous mineral exploration and site investigation.

Index information, which includes site references, for these bores has been digitised. The logs are either hand-written or typed and many of the older records are drillers' logs.

Site exploration reports

This collection consists of site exploration reports carried out to investigate foundation conditions prior to construction. There is a digital index and the reports themselves are held on microfiche. For the district there are about 29 reports.

Mine plans

BGS maintains a collection of plans of underground mines for minerals other than coal and oil-shale. The mines known to have been worked or trialled within the district are discussed in Chapter 10 (see (Figure 38)) but only those noted below have abandonment plans deposited with BGS.

Mine Mineral worked Plan No.
Blackcraig lead, zinc, copper, silver AP 7906, AP 15151
Wood of Cree zinc, lead, copper AP 7907

Material collections

Geological Survey photographs

Ninety four photographs illustrating aspects of the geology of the district are deposited for reference in the libraries at BGS, Murchison House, West Mains Road, Edinburgh EH9 3LA and BGS, Keyworth, Nottingham NG12 5GG; and in the BGS Information Office, Natural History Museum Earth Galleries, Exhibition Road, London SW7 2DE. Many of the photographs can be viewed online at BGS web site.

Early photographs (black and white, dating from 1937–39) show the quarry and granite sett making operations at Kirkmabreck, near Carsluith. The others are modern colour photographs taken during the recent resurvey. A list of titles can be supplied on request. The photographs can be supplied as prints (black and white or colour), colour transparencies (35 mm) or digital files (tiff or jpeg at variable resolution), at a fixed tariff.

Petrological collections

The petrological collections for the district consist of about 1000 hand specimens and thin sections. Most of the older samples and thin sections are of the igneous rocks in the district whereas more recent collections concentrate on the Silurian sedimentary rocks. The collections are indexed on the basis of the 1:50 000 geological maps, but much of the older part of the collection cannot at present be searched by National Grid Reference.

Palaeontological collections

The collections of biostratigraphical specimens have been collected mostly from surface and temporary exposures throughout the district. They are essentially working collections and mainly used for reference. Some locality index data are held on a computer database.

References

Most of the references listed below are held in the Library of the British Geological Survey, Murchison House, Edinburgh, and at Keyworth, Nottingham. Copies of the references can be purchased subject to current copyright legislation. BGS Library catalogue can be searched online at: http://geolib.bgs.ac.uk

Akhurst, M C, and 24 others. 1997. Geology of the west Cumbria district. Memoir of the British Geological Survey, Sheets?28 (Whitehaven), 37 (Gosforth) and 47 (Bootle) (England and Wales). (Keyworth, Nottingham: British Geological Survey.)

Al Mansouri, D. 1986. Seismological studies of upper-crustal structure in the vicinity of the Girvan–Ballantrae area, SW Scotland. Unpublished PhD thesis. University of Glasgow.

Anderson, T B. 1962. The stratigraphy, sedimentology and structural geology of the Silurian rocks of the Ards Peninsula, County Down. Unpublished PhD thesis, University of Liverpool.

Anderson, T B. 1987. The onset and timing of Caledonian sinistral shear in County Down. Journal of the Geological Society of London, Vol.?144, 817–825.

Anderson, T B, and Cameron, T D J. 1979. A structural profile of Caledonian deformation in Down. 263–267 in The Caledonides of the British Isles?—?reviewed, Harris,?A?L, Holland,?C?H, and Leake,?B?E (editors). Special Publication of the Geological Society of London, No.?8.

Anderson, T B, and Oliver, G J H. 1986. The Orlock Bridge Fault: a major late Caledonian sinistral fault in the Southern Uplands terrane, British Isles. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?77, 203–222.

Anderson, T B, Parnell, J, and Ruffell, A H. 1995. Influence of basement on the geometry of Permo–Triassic basins in the northwest British Isles. 103–122 in Permian and Triassic rifting in northwest Europe. Boldy,?S?A?R (editor). Special Publication of the Geological Society of London, No.?91.

Arthurton, R S, Burgess, I C, and Holliday, D W. 1978. Permian and Triassic. 791–808 in The geology of the Lake District. Mosely, F (editor). Special Publication of the Yorkshire Geological Society, No. 3.

Armstrong, H A, Owen, A W, Scrutton, C T, Clarkson, E?N?K, and Taylor, C M. 1996. Evolution of the Northern Belt, Southern Uplands: implications for the Southern Uplands controversy. Journal of the Geological Society of London, Vol.?153, 197–206.

Bailey, D (compiler). 1997. Constraining the timing of movement on faults in the Sellafield area. British Geological Survey Technical Report, No.?WG/97/36.

Ballantyne, C K, and Gray, J M. 1984. The Quaternary geomorphology of Scotland: the research contribution of J?B?Sissons. Quaternary Science Reviews, Vol.?3, 259–289.

Barnes, R P. 1989. Geology of the Whithorn district. Memoir of the British Geological Survey, Sheet?2 (Scotland). (London: HMSO.)

Barnes, R P. 1996. Whithorn. 105–113 in Geology in south-west Scotland: an excursion guide. Stone, P (editor). (Keyworth, Nottingham: British Geological Survey.)

Barnes, R P, and Fettes D J. 1996. Creetown and Cairnsmore of Fleet. 140–150 in Geology in south-west Scotland: an excursion guide. Stone, P (editor). (Keyworth, Nottingham: British Geological Survey).

Barnes, R P, and Fortey, N J. 1997. Metamorphism and skarn metasomatism prior to emplacement of the Kirkmabreck microgranodiorite dyke, Creetown, SW Scotland. British Geological Survey, Technical Report, WA/97/94.

Barnes, R P, Rock, N M S, and Gaskarth, J W. 1986. Late Caledonian dyke-swarms in Southern Scotland: new field, petrological and geochemical data for the Wigtown Peninsula, Galloway. Geological Journal, Vol.?21, 101–125.

Barnes, R P, Anderson, T B, and McCurry, J A. 1987. Along-strike variation in the stratigraphical and structural profile of the Southern Uplands Central Belt in Galloway and Down. Journal of the Geological Society of London, Vol.?144, 807–816.

Barnes, R P, Floyd, J D, and Stone, P. 1988. Big Scare?—? resurveyed after 110 years. BGS short communications?8, Report of the British Geological Survey, Vol.?19, No.?2, 19–23.

Barnes, R P, Lintern, B C, and Stone, P. 1989. Timing and regional implications of deformation in the Southern Uplands of Scotland. Journal of the Geological Society of London, Vol.?146, 905–908.

Barnes, R P, Ambrose, K, Holliday, D W, and Jones, N S. 1994. Lithostratigraphical subdivision of the Triassic Sherwood Sandstone Group in west Cumbria. Proceedings of the Yorkshire Geological Society. Vol.?50, 51–60.

Barnes, R P, Phillips, E R, and Boland, M P. 1995a. The Orlock Bridge Fault in the Southern Uplands of SW Scotland, a terrane boundary? Geological Magazine, Vol.?132, 523–529.

Barnes, R P, Phillips, E R, and Merriman, R J. 1995b. Allochthonous Ordovician basaltic rocks of possible island arc affinity in the Southern Uplands of SW Scotland. 161–170 in Current perspectives in the Appalachian–Caledonian Orogen. Hibbard, J?P, van?Staal,?C?R, and Cawood,?P?A (editors). Geological Association of Canada Special Paper, No.?41.

Barnes, R P, Boland, M P, Phillips, E R, and Stone, P. 1996. The Orlock Bridge Fault and Moniaive Shear Zone. 151–164 in Geology in south-west Scotland: an excursion guide. Stone,?P (editor). (Keyworth, Nottingham: British Geological Survey.)

Batchelor, R, and Weir, J A. 1989. Metabentonite geochemistry: magmatic cycles and graptolite extinctions at Dob's Linn, southern Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?79, 19–41.

Benton, M J. 1982. Trace fossils from Lower Palaeozoic ocean–floor sediments of the Southern Uplands of Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?73, 67–87.

Bhatia, M R. 1983. Plate tectonics and the geochemical composition of sandstones. Journal of Geology, Vol.?91, 611–627.

Birks, H H. 1972. Studies in the vegetational history of Scotland II. Two pollen diagrams from the Galloway Hills, Kirkcudbrightshire. Journal of Ecology, Vol.?60, 183–217.

Birks, H J B. 1989. Holocene isochrone maps and patterns of tree-spreading in the British Isles. Journal of Biogeography, Vol.?16, 503–540.

Birks, H J B, Deacon J, and Peglar, S M. 1975. Pollen maps for the British Isles 5000 years ago. Proceedings of the Royal Society of London B, Vol.?189, 87–105.

Bishop, W W, and Coope, G R. 1977. Stratigraphical and faunal evidence for Lateglacial and Early Flandrian environments in south-west Scotland. 61–88 in Studies in the Scottish Lateglacial Environment. Gray, J M, and Lowe, J J (editors). (Oxford: Pergamon Press.)

Blatt, H, Middleton G V, and Murray R C. 1980. Origin of sedimentary rocks. (New Jersey: Prentice Hall.)

Blyth, F G H. 1949. The sheared porphyrite dykes of South Galloway. Quarterly Journal of the Geological Society of London, Vol.?105, 393–423.

Bott, M H P. 1964. Gravity measurements in the north-eastern part of the Irish Sea. Quarterly Journal of the Geological Society of London, Vol.?120, 369–396.

Bott, M H P, and Masson Smith, D. 1960. A gravity survey of the Criffel granodiorite and New Red Sandstone deposits near Dumfries. Proceedings of the Yorkshire Geological Society, Vol.?32, 317–332.

Bott, M H P, Long, R E, Green, A S P, Lewis, A H J, Sinha, M C, and Stevenson D L. 1985. Crustal structure south of the Iapetus Suture beneath northern England. Nature, Vol.?314, 724–727.

Boulton, G S. 1990. Sedimentation and sea level changes during glacial cycles and their control on glacimarine facies architecture. 15–52 in Glacimarine environments: processes and sediments. Dowdeswell, J A, and Scourse, J D (editors). Special Publication of the Geological Society of London, No.?53.

Boulton, G S. 1992. Quaternary. 413–444 in Geology of England and Wales. Duff, P?McL?D, and Smith, A?J (editors) (London: The Geological Society.)

Boulton, G S, and Payne, A. 1994. Northern hemisphere icesheets through the last glacial cycle: Glaciological and geological reconstructions. 177–212 in Long term climatic changes: data and modelling. Duplessy, J C (editor). NATO ASI Series, Vol.?122. (Stuttgart: Springer.)

Boulton, G S, Jones, A S, Clayton, K M, and Kenning, M J. 1977. A British ice-sheet model and patterns of glacial erosion and deposition in Britain. 231–246 in British Quaternary Studies. Shotton, F W (editor). (Oxford: Clarendon Press.)

Boulton, G S, Smith, G D, Jones, A S, and Newsome, J. 1985. Glacial geology and glaciology of the last mid–latitude ice sheets. Journal of the Geological Society of London, Vol.?142, 447–474.

Boulton, G S, Peacock, J D, and Sutherland, D G. 1991. Quaternary. 503–543 in Geology of Scotland. Third edition. Craig,?G?Y (editor). (London: The Geological Society.)

Bouma, A H. 1962. Sedimentology of some flysch deposits: a graphic approach to interpretation. (Amsterdam: Elsevier.)

Bowen, D Q. 1978. Quaternary geology: a stratigraphic framework for multidisciplinary work. (Oxford: Pergamon Press.)

Bowen, D Q. 1989. The last interglacial–glacial cycle in the British Isles. Quaternary International, Vol.?3/4, 41–47.

British Geological Survey. 1992a. Kirkcowan Scotland. Sheet?4W. Solid, 1:50?000 (Nottingham: British Geological Survey.)

British Geological Survey. 1992b. Wigtown Scotland. Sheet?4E. Solid, 1:50?000 (Nottingham: British Geological Survey.)

Brookfield, M E. 1978. Revision of the stratigraphy of Permian and supposed Permian rocks of southern Scotland. Geologische Rundschau, Vol.?67, 110–143.

Brown, G C, Cassidy, J, Tindle, A G, and Hughes, D J. 1979. The Loch Doon granite: an example of granite petrogenesis in the British Caledonides. Journal of the Geological Society of London, Vol.?136, 745–753.

Brown, P E, Miller, J A, and Grasty, R L. 1968. Isotopic age of late Caledonian granitic intrusions in the British Isles. Proceedings of the Yorkshire Geological Society, Vol.?36, 251–276.

Chadwick, R A, Kirby, G A, and Baily, H E. 1994. The post–Triassic structural evolution of north-west England and adjacent parts of the East Irish Sea. Proceedings of the Yorkshire Geological Society, Vol.?50, 91–102.

Chadwick, R A, Holliday, D W, Holloway, S, and Hulbert, A G. 1995. The structure and evolution of the Northumberland–Solway Basin and adjacent areas. Subsurface Memoir of the British Geological Survey. (London: HMSO.)

Chappel, B W, and Stephens, W E. 1988. Origin of infracrustal (I–type) granite magmas. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?79, 71–86.

Charlesworth, J K. 1926a. The glacial geology of the Southern Uplands, west of Annandale and upper Clydesdale. Transactions of the Royal Society of Edinburgh, Vol.?55, 1–23.

Charlesworth, J K. 1926b. The readvance, marginal kame–moraine of the south of Scotland, and some later stages of retreat. Transactions of the Royal Society of Edinburgh, Vol.?55, 25–50.

Clarkson, C M, Craig, G Y, and Walton, E K. 1975. The Silurian rocks bordering Kirkcudbright Bay, south Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?69, 313–325.

Cocks, L R M, and Rickards, R B (editors). 1988. A global analysis of the Ordovician–Silurian boundary. Bulletin of the British Museum (Natural History), Geology Series, No.?43.

Colman-Sadd, S P, Stone, P, Swinden, H S, and Barnes, R P. 1992. Parallel geological development in the Dunnage Zone of Newfoundland and the Lower Palaeozoic terranes of southern Scotland: an assessment. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?83, 571–594.

Cook, D R. 1976. The geology of the Cairnsmore of Fleet Granite and its environs, southwest Scotland. Unpublished PhD thesis, University of St Andrew's.

Cook, D R, and Weir, J A. 1979. Structure of the Lower Palaeozoic rocks around Cairnsmore of Fleet, Galloway. Scottish Journal of Geology, Vol.?15, 187–202.

Cook, D R, and Weir, J A. 1980. The stratigraphical setting of the Cairnsmore of Fleet Pluton, Galloway. Scottish Journal of Geology, Vol.?16, 125–141.

Cooper, D C, Parker, M E, and Allen, P M. 1982. Investigations of small intrusions in southern Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No.?58.

Cordell, L E, and Grauch, V J S. 1985. Mapping basement magnetisation zones from aeromagnetic data in the San Juan basin, New Mexico. 181–197 in The utility of regional gravity and magnetic anomaly maps. Hinze, W J (editor). (Tulsa: Society of Exploration Geophysicists.)

Corfu, F and Ayres, L D. 1984. U-Pb ages and genetic significance of heterogeneous zircon populations in rocks from the Favourable Lake area, northwestern Ontario. Contributions to Mineralogy and Petrology, Vol.?88, 86–101.

Cornish, R. 1980. Glacial geomorphology of the west-central Southern Uplands of Scotland, with particular reference to the 'Rogen moraines'. Unpublished PhD thesis. University of Edinburgh.

Cornish, R. 1981. Glaciers of the Loch Lomond Stadial in the western Southern Uplands of Scotland. Proceedings of the Geologists' Association, Vol.?92, 105–114.

Craig, G Y, and Walton, E K. 1959. Sequence and structure in the Silurian rocks of Kirkcudbrightshire. Geological Magazine, Vol.?96, 209–220.

Craik, J. 1873. Explanation of Sheet 2, Wigtownshire–Burrow Head. Memoir of the Geological Survey, Scotland. (London: HMSO.)

Cutler, H D. 1979. Glaciation and drumlins of the Moors and Machars of Galloway, southwest Scotland. Unpublished PhD thesis. University of Liverpool.

Davies, P. 1990. The geology of the Rhins of Galloway, SW Scotland: the west coast between Killantringan Bay and Clanyard Bay. British Geological Survey Technical Report, WA/90/67.

Dawson, J, Floyd, J D, Philip, P R, Burley, A J, Allsop, J M, and Bennett, J R P. 1977. A mineral reconnaissance survey of the Doon–Glenkens area, south-west Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No.?18.

Dewey, J F. 1971. A model for the Lower Palaeozoic evolution of the southern margin of the early Caledonides of Scotland and Ireland. Scottish Journal of Geology, Vol.?7, 219–240.

Duller, P R, and Floyd, J D. 1995. Turbidite geochemistry and provenance studies in the Southern Uplands of Scotland. Geological Magazine, Vol.?132, 557–569.

Elles, G L. 1925. Characteristic assemblages of the graptolite zones of the British Isles. Geological Magazine, Vol.?62. 337–347.

Elles, G L, and Wood, E M R. 1901–1918. A monograph of British graptolites. Monographs of the Palaeontographical Society (11 parts).

Fettes, D J, and Timmerman, H. 1992. The geology of the Cairnsmore of Fleet Granite. British Geological Survey, Technical Report,WA/92/89.

Finney, S C, and Bergström, S M. 1986. Biostratigraphy of the Ordovician Nemagraptus gracilis Zone. 47–59 in Palaeoecology and biostratigraphy of graptolites. Hughes, C P, and Rickards,?R?B (editors). Geological Society of London Special Publication, No.?20

Floyd, J D. 1976. The Ordovician rocks of west Nithsdale. Unpublished PhD thesis. University of St Andrews.

Floyd, J D. 1982. Stratigraphy of a flysch succession: the Ordovician of west Nithsdale, SW Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?73, 1–9.

Floyd, J D. 1996. Lithostratigraphy of the Ordovician rocks in the Southern Uplands: Crawford Group, Moffat Shale Group, Leadhills Supergroup. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?86, 153–165.

Floyd, J D. 1999. Geology of the Carrick-Loch Doon district. Memoir of the British Geological Survey, sheets 8W and 8E (Scotland). (London: The Stationery Office.)

Floyd, J D, and Phillips, E R. 1999. Xenoliths of Southern Uplands 'basement?' in a lamprophyre dyke, Central Belt, Glenluce, SW Scotland. Scottish Journal of Geology, Vol.?35, 57–62.

Floyd, J D, and Rushton, A W A. 1993. Ashgill greywackes in the Southern Uplands of Scotland: an extension of the Ordovician succession in the Northern Belt. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?84, 79–85.

Floyd, J D, and Trench, A. 1989. Magnetic susceptibility contrasts in Ordovician greywackes of the Southern Uplands of Scotland. Journal of the Geological Society of London, Vol.?146, 77–83.

Fortey, R A, Harper, D A T, Ingham, J K, Owen, A W, and Rushton, A W A. 1995. A revision of Ordovician series and stages from the historical type area. Geological Magazine, Vol.?132, 15–30.

Gardiner, C I, and Reynolds, S H. 1937. The Cairnsmore of Fleet Granite and its metamorphic aureole. Geological Magazine, Vol.?74, 289–300.

Gallagher, M J. 1958. Rock alteration in some mineralised basic dykes in Britain. Transactions of the Institution of Mining and Metallurgy, Vol.?73, 825–840.

Galloway, R W. 1961. Ice wedges and involutions in Scotland. Biuletyn Peryglacjalny, Vol.?10, 169–193.

Geikie, A. 1878. Explanation of Sheet 4, Wigtownshire (East Part) and Kirkcudbright (Portion of Western Division). Memoir of the Geological Survey, Scotland. (London: HMSO.)

Goodlet, G A. 1970. Sands and gravels of the southern counties of Scotland. Report of the Institute of Geological Sciences, No.?70/4.

Gordon, A J. 1962. The Lower Palaeozoic rocks around Glenluce, Wigtownshire. Unpublished PhD thesis. University of Edinburgh.

Gordon, J E. 1993. Bigholm Burn. 596–599 in The Geological Conservation Review Series 6: The Quaternary of Scotland. Gordon,?J?E, and Sutherland, D?G (editors). (London: Chapman and Hall.)

Gordon, J E, and Sutherland, D G (editors). 1993. The Quaternary of Scotland. Geological Conservation Review Series, No.?6. (London: Chapman and Hall.)

Gradstein, F M, and Ogg, J G. 1996. A Phanerozoic time scale. Episodes, Vol.?19 (1 & 2), 3–5.

Gray, D R. 1981. Cleavage-fold relationships and their implications for transected folds: an example from southwest Virginia, USA. Journal of Structural Geology, Vol.?3, 265–277.

Gray, J M, and Lowe, J J. 1977. Introduction. xi–xiii in Studies in the Scottish Lateglacial environment. Gray, J M, and Lowe,?J?J (editors). (Oxford: Pergamon Press.)

Greig, D C. 1971. British regional geology: the South of Scotland. Third edition. (Edinburgh: HMSO for British Geological Survey.)

Gregory, J W. 1928. The nickel–cobalt ore of Talnotry, Kirkcudbrightshire. Transactions of the Institution of Mining and Metallurgy, Vol.?37, 178–195.

Haggart, B A. 1988. A review of radiocarbon dates on peat and wood from Holocene coastal sedimentary sequences in Scotland. Scottish Journal of Geology, Vol.?24, 125–144.

Haggart, B A. 1989. Variations in the pattern and rate of isostatic uplift indicated by comparison of Holocene sea-level curves from Scotland. Journal of Quaternary Science, Vol.?4, 67–76.

Hall, J, Powell, D W, Warner, M R, El–Isa, Z H M, Adesanya,?O, and Bluck, B J. 1983. Seismological evidence for shallow crystalline basement in the Southern Uplands of Scotland. Nature, Vol.?305, 418–420.

Halliday, A N. 1983. Crustal melting and the genesis of isotopically and chemically zoned plutons in the Southern Uplands of Scotland. 54–61 in: Migmatites and crustal melting. Atherton, M P, and Gribble, C D (editors). (Orpington: Shiva.)

Halliday, A N. 1984. Coupled Sm-Nd and U-Pb systematics in late Caledonian granites and the basement under northern Britain. Nature, Vol.?307, 229–233.

Halliday, A N, Stephens, W E, and Harmon, R S. 1980. Rb-Sr and O isotopic relationships in three zoned Caledonian granitic plutons, Southern Uplands, Scotland: evidence for varied sources and hybridization of magmas. Journal of the Geological Society of London, Vol.?137, 329–348.

Harrison, R K, Stone, P, Cameron, I B, Elliot, R W, and Harding, R R. 1987. Geology, petrology and geochemistry of Ailsa Craig, Ayrshire. Report of the British Geological Survey, Vol.?16, No.?9.

Henney, P J. 1991. The geochemistry and petrogenesis of the minor intrusive suite associated with the Late Caledonian Criffel–Dalbeattie pluton, SW Scotland. Unpublished PhD thesis. University of Aston, Birmingham.

Hughes, R A. 1989. Llandeilo and Caradoc graptolites of the Builth and Shelve inliers. Monographs of the Palaeontographical Society, Vol.?141 (for 1987), No.?577.

Hutton, D H W. 1988. Granite emplacement mechanisms and tectonic controls: inferences from deformation studies. Transactions of the Royal Society of Edinburgh: Earth Sciences. Vol.?79, 245–255.

Hutton, D H W, Dempster, T J, Brown, P E, and Becker, S M. 1990. A new mechanism of granite emplacement: intrusion in active extensional shear zones. Nature, Vol.?343, 452–455.

Institute of Geological Sciences. 1981. Wigtown. Scotland Sheet?4E. Drift Geology 1:50?000. (Southampton: Ordnance Survey for Institute of Geological Sciences.)

Institute of Geological Sciences. 1982. Kirkcowan. Scotland Sheet?4W. Drift Geology 1:50?000. (Southampton: Ordnance Survey for Institute of Geological Sciences.)

Irvine, D R. 1878. Explanation of Sheet 4, Wigtownshire (east part). Kirkcudbright (portion of western division). Memoir of the Geological Survey, Scotland. (London: HMSO.)

Ivmey–Cook, H C, Warrington, G, Worley, N E, Holloway, S, and Young, B. 1995. Rocks of late Triassic and early Jurassic age in the Carlisle Basin, Cumbria (north-west England). Proceedings of the Yorkshire Geological Society, Vol.?50, 305–316.

Jackson, D I, and Mullholland, P. 1993. Tectonic and structural aspects of the East Irish Sea Basin and adjacent areas: contrasts in their post-Carboniferous structural styles. 791–808 in Petroleum geology of north-west Europe: proceedings of the fourth conference. Parker, J R (editor). (London: The Geological Society.)

Jackson, D I, Jones, S M, and Warrington, G. 1987. The geological framework of the East Irish Sea Basin. 191–203 in Petroleum geology of north west Europe. Brookes, J, and Glennie, K W (editors). (London: Graham and Trotman.)

Jackson, D I, Jackson, A A, Evans, D, Wingfield, R T R, Barnes,?R?P, and Arthur, M J. 1995. United Kingdom offshore regional report: the geology of the Irish Sea. (Nottingham: British Geological Survey.)

Jaffey, A H, Flynn, K F, Glendenin, L E, Bentley, W C and Essling, A M. 1971. Precision measurements of half-lives and specific activities of 235U and 238U. Physics Reviews, Vol.?C4, 1889–1906.

Jardine, W G. 1967. Sediments of the Flandrian transgression in south-west Scotland: terminology and criteria for facies distinction. Scottish Journal of Geology, Vol.?3, 221–226.

Jardine, W G. 1975. Chronology of Holocene marine transgression and regression in south-western Scotland. Boreas, Vol.?4, 173–196.

Jardine, W G. 1980. Holocene raised coastal sediments and former shorelines of Dumfriesshire and Eastern Galloway. Transactions of the Dumfries and Galloway Natural History and Antiquarian Society, Vol.?43, 1–59.

Jardine, W G. 1981. Holocene shorelines in Britain: recent studies. Geologie en Mijnbouw, Vol.?60, 297–304

Jardine, W G, and Morrison, A. 1976. The archaeological significance of Holocene coastal deposits in south–western Scotland. 175–195 in Geoarchaeology: Earth Science and the past. Davidson, D A, and Shackley, M L (editors). (London: Duckworth.)

Jolly, W. 1868. On the evidences of glacier action in Galloway. Transactions of the Edinburgh Geological Society, Vol.?1, 155–185.

Jones, N J, and Ambrose, K. 1994. Triassic sandy braidplain and aeolian sedimentation in the Sherwood Sandstone Group of the Sellafield area, west Cumbria. Proceedings of the Yorkshire Geological Society, Vol.?50, 61–76.

Jones, O T. 1924. Lead and zinc ores in the slaty rocks of Britain. Transactions of the Institution of Mining Engineers, Vol.?66, 219–242.

Jones, R L, and Keen, D H. 1993. Pleistocene environments in the British Isles. (London: Chapman and Hall.)

Kafafy, A M, and Tarling, D H. 1985. Magnetic fabric in some granitic aureoles, Southern Uplands, Scotland. Journal of the Geological Society of London, Vol.?142, 1007–1014.

Kamaliddin, Z A R. 1991. Seismic interpretation of the Southern Uplands. Unpublished PhD thesis. University of Glasgow.

Kelley, S, and Bluck, B J. 1989. Detrital mineral ages from the Southern Uplands using 40Ar–39Ar laser probe. Journal of the Geological Society of London, Vol. 146, 401–403.

Kelling, G. 1961. The stratigraphy and structure of the Ordovician rocks of the Rhinns of Galloway. Quarterly Journal of the Geological Society of London, Vol.?117, 37–75.

Kelling, G. 1962. The petrology and sedimentation of Upper Ordovician rocks in the Rhinns of Galloway, south-west Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?65, 107–137.

Kelling, G, Davies, P, and Holroyd, J. 1987. Style, scale and significance of sand bodies in the Northern and Central Belts, southwest Southern Uplands. Journal of the Geological Society of London, Vol.?144, 787–805.

Kemp, A E S. 1986. Tectonostratigraphy of the Southern Belt of the Southern Uplands. Scottish Journal of Geology, Vol.?22, 241–256.

Kemp, A E S. 1987. Tectonic development of the Southern Belt of the Southern Uplands accretionary complex. Journal of the Geological Society of London, Vol.?144, 827–838.

Kemp, A E S, and White, D E. 1985. Silurian trench sedimentation in the Southern Uplands, Scotland: implications of new age data. Geological Magazine, Vol.?122, 275–277.

Kemp, A E S, Oliver, G J H, and Baldwin, J R. 1985. Low-grade metamorphism and accretion tectonics: Southern Uplands terrain, Scotland. Mineralogical Magazine, Vol.?49, 335–344.

Kerr, W B. 1982a. Pleistocene ice movements in the Rhins of Galloway. Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society, Vol.?57, 1–10.

Kerr, W B. 1982b. How many ice advances in Galloway? Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society, Vol.?57, 11–15.

Kerrick, D M. 1991. Overview of contact metamorphism. 1–12 in Contact metamorphism. Kerrick, D M (editor). Mineralogical Society of America: Reviews in Mineralogy, Vol.?26.

Kimbell, G S. 1991. An assessment of geophysical data from the Rhins of Galloway district, south-west Scotland. British Geological Survey Technical Report, WK/91/10.

Kimbell, G S, and Stone, P. 1992. Geophysical evidence for a concealed Caledonian intrusive body at Sandhead, Wigtownshire. Scottish Journal of Geology, Vol.?28, 19–25.

Kimbell, G S, and Stone, P. 1995. Crustal magnetisation variations across the Iapetus Suture Zone. Geological Magazine, Vol.?132, 599–609.

Kisch, H J. 1991. Illite crystallinity: recommendations on sample preparation, X-ray diffraction settings and interlaboratory standards. Journal of Metamorphic Geology,Vol.?9, 665–670.

Knipe, R J, and Needham, D T. 1986. Deformation processes in accretionary wedges?—?examples from the SW margin of the Southern Uplands, Scotland. 51–65 in Collision tectonics. Coward, M P, and Ries, A C (editors). Special Publication of the Geological Society of London, No.?19.

Krogh, T E. 1973. A low contamination method for the hydrothermal decomposition of zircon and extraction of U and Pb for isotopic age determinations. Geochimica et Cosmochimica Acta, Vol.?37, 485–494.

Krogh, T E and Davis, G L. 1975. The production and prepa- ration of 205Pb for use as a tracer for isotope dilution analysis. Carnegie Institution of Washington Yearbook, Vol.?74, 416–417.

Lambeck, K. 1996. Glaciation and sea-level change for Ireland and the Irish Sea since Late Devensian/Midlandian time. Journal of the Geological Society of London, Vol.?153, 853–872.

Lambert, R, St J, Holland, J G, and Leggett, J K. 1981. Petrology and tectonic setting of some Ordovician volcanic rocks from the Southern Uplands of Scotland. Journal of the Geological Society of London, Vol.?138, 421–436.

Lapworth, C. 1876. On Scottish Monograptidæ. Geological Magazine, Vol.?3, 308–321, 350–360, 499–507, 544–552.

Lapworth, C. 1878. The Moffat Series. Quarterly Journal of the Geological Society of London, Vol.?34, 240–346.

Lapworth, C. 1889. On the Ballantrae rocks of the south of Scotland and their place in the upland sequence. Geological Magazine, Vol.?26, 20–24, 59–69.

Lapworth, C and Wilson, G V. 1871. On the Silurian rocks of the counties of Roxburgh and Selkirk. Transactions of the Edinburgh Geological Society, Vol.?2, 46.

Leake, R C and Brown, M J. 1977. A geochemical drainage survey of the Fleet granitic complex and its environs. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No.?21.

Lee, M K. 1989. Upper crustal structure of the Lake District from modelling and image processing of potential field data. British Geological Survey Technical Report, WK/89/1.

Leeder, M R. 1982. Upper Palaeozoic basins of the British Isles?—?Caledonide inheritance versus Hercynian plate margin processes. Journal of the Geological Society of London, Vol.?139, 479–491.

Leggett, J K, McKerrow, W S, and Eales, M H. 1979. The Southern Uplands of Scotland: a Lower Palaeozoic accretionary prism. Journal of the Geological Society of London, Vol.?136, 755–770.

Le Maitre, R W. 1989. A classification of igneous rocks and glossary of terms. Recommendations of the International Union of Geological Sciences Subcommission on the systematics of igneous rocks. (Oxford: Blackwell Scientific Publications.)

Lewis, A H J. 1986. The deep seismic structure of northern England and adjacent marine areas from the Caledonian Suture Seismic Project. Unpublished PhD thesis. University of Durham.

Lewis, C L E, Green, P F, Carter, A, and Hurford, A J. 1992. Elevated K/T palaeotemperatures throughout northwest England: three kilometres of Tertiary erosion? Earth and Planetary Science Letters, Vol.?112, 131–145.

Lintern, B C, and Floyd, J D. 2000. Geology of the Kirkcudbright–Dalbeattie district. Memoir of the British Geological Survey, Sheets 5W, 5E and part of 6W (Scotland). (London: The Stationery Office.)

Loydell, D K. 1992. Upper Aeronian and Lower Telychian (Llandovery) Graptolites from western Mid-Wales. Palaeontographical Society [Monograph], Vol.?146, Part?589.

Loydell, D K. 1993. Upper Aeronian and Lower Telychian (Llandovery) Graptolites from western Mid-Wales. Palaeontographical Society [Monograph], Vol.?147, Part?592.

Loydell, D K. 1994. Early Telychian changes in graptoloid diversity and sea level. Geological Journal, Vol.?29, 355–368.

Loydell, D K, and Cave, R. 1993. The Telychian (Upper Llandovery) stratigraphy of Buttington brick pit, Wales. Newsletters on Stratigraphy, Vol.?29, 91–103.

Loydell, D K, and Cave, R. 1996. The Llandovery–Wenlock boundary and related stratigraphy in eastern mid-Wales with special reference to the Banwy river section. Newsletters on Stratigraphy, Vol.?34, 39–64.

Loydell, D K, Štorch, P, and Melchin, M J. 1993. Taxonomy, evolution and biostratigraphical importance of the Llandovery graptolite Spirograptus. Palaeontology, Vol.?36, 909–926.

Ludwig, K?R. 1993. PBDAT. A computer program for processing Pb-U-Th isotope data, version 1.24. United States Geological Survey Open-file Report, No.?88–542.

Ludwig, K?R. 1994. ISOPLOT: A plotting and regression program for radiogenic-isotope data, version 2.75. United States Geological Survey Open-file Report, No.?91–445.

May, J. 1981. The glaciation and deglaciation of upper Nithsdale and Annandale. Unpublished PhD thesis. University of Glasgow.

McCabe, A M. 1987. Quaternary deposits and glacial stratigraphy in Ireland. Quaternary Science Reviews, Vol.?6, 259–299.

McClay, K R. 1987 (reprinted 1991). The mapping of geological structures. Geological Society of London Handbook. (Chichester: John Wiley & Sons.)

McCurry, J A, and Anderson, T B. 1989. Landward vergence in the Lower Palaeozoic Southern Uplands–Down–Longford terrane, British Isles. Geology, Vol.?17, 630–633.

McCurry, J A, and Stone, P. 1996. Rhins of Galloway. 121–139 in Geology in south-west Scotland: an excursion guide. Stone, P (editor). (Keyworth, Nottingham: British Geological Survey.)

McKerrow, W S, Leggett, J K, and Eales, M H. 1977. Imbricate thrust model of the Southern Uplands of Scotland. Nature, Vol.?267, 237–239.

McMillan, A A. 1997. Quaternary in The geology of west Cumbria. Akhurst, M C and 24 others Memoir of the British Geological Survey. Sheets 28, 37 and 47 (England and Wales). (Keyworth, Nottingham: British Geological Survey.)

McMillan, A A. 2002. Geology of the New Galloway and Thornhill district. Memoir of the British Geological Survey, Sheets 9W and 9E (Scotland). (London: The Stationery Office.)

McMillan, A A, and Brand, P J. 1995. Depositional setting of Permian and Upper Carboniferous strata of the Thornhill Basin, Dumfriesshire. Scottish Journal of Geology, Vol.?31, 43–52.

McMillan, A A, and McAdam, A D. 1996. Upper Palaeozoic to Quaternary regional geology. 15–20 inGeology in south-west Scotland: an excursion guide. Stone, P (editor). (Keyworth, Nottingham: British Geological Survey.)

Mansfield, J, and Kennett, P. 1963. A gravity survey of the Stranraer sedimentary basin. Proceedings of the Yorkshire Geological Society, Vol.?34, 139–151.

Merriman, R J, and Roberts, B. 1990. Metabentonites in the Moffat Shale Group, Southern Uplands of Scotland: geochemical evidence of ensialic marginal basin volcanism. Geological Magazine, Vol.?127, 259–271.

Merriman, R J, and Roberts, B. 1992. The low grade metamorphism of Lower Palaeozoic strata on the Rhins of Galloway, SW Scotland. British Geological Survey Technical Report, WG/92/40.

Merriman, R J, and Roberts, B. 1996. Metamorphism of the Lower Palaeozoic rocks. 13–15 in Geology in south-west Scotland: an excursion guide. Stone, P (editor) (Keyworth, Nottingham: British Geological Survey).

Merriman, R J, Roberts, B, and Peacor, D R. 1990. A transmission electron microscope study of white mica crystallite size distribution in a mudstone to slate transitional sequence, North Wales, UK. Contributions to Mineralogy and Petrology, Vol.?10, 27–40.

Merriman, R J, Roberts, B, Peacor, D R, and Hirons, S R. 1995. Strain-related differences in the crystal growth of white mica and chlorite: a TEM and XRD study of the development of metapelitic microfabrics in the Southern Uplands thrust terrane, Scotland. Journal of Metamorphic Geology, Vol.?13, 559–576.

Moar, N T. 1969. Late Weichselian and Flandrian pollen diagrams for southwest Scotland. New Phytologist, Vol. 68, 433–467.

Morris, J H. 1987. The northern belt of the Longford Down inlier, Ireland and Southern Uplands, Scotland: an Ordovician back arc basin. Journal of the Geological Society of London, Vol.?144, 773–786.

Murphy, F C, and Hutton, D H W. 1986. Is the Southern Uplands of Scotland really an accretionary prism? Geology, Vol.?14, 354–357.

Needham, D T. 1993. The structure of the western part of the Southern Uplands of Scotland. Journal of the Geological Society of London, Vol.?150, 341–354.

Nichols, H. 1967. Vegetational change, shoreline displacement and the human factor in the Late Quaternary of south-west Scotland. Transactions of the Royal Society of Edinburgh, Vol.?67, 145–187.

Noble, S?R, Tucker, R?D, and Pharaoh, T?C. 1993. Lower Palaeozoic and Precambrian igneous rocks from eastern England and their bearing on Ordovician closure of the Tornquist Sea: constraints from U-Pb and Nd isotopes. Geological Magazine, Vol.?130, 835–46.

Oliver, G J H, and Leggett, J K. 1980. Metamorphism in an accretionary prism: prehnite-pumpellyite facies metamorphism of the Southern Uplands of Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?71, 235–246.

Oliver, G J H, Smellie, J L, Thomas, L J, Casey, D M, Kemp, A E S, Evans. L J, Baldwin, J R, and Hepworth, B C. 1984. Early Palaeozoic metamorphic history of the Midland Valley, the Southern Uplands–Longford–Down massif and the Lake District, British Isles. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?75, 259–273.

Parker, M E. 1977. Geophysical surveys around Talnotry Mine, Kirkcudbright, Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No.?10.

Parker, M E, Cooper, D C, Bide, P J, and Allen, P M. 1981. Mineral exploration in the area around Culvennan Fell, Kirkcowan, south-western Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No.?42.

Parslow, G R. 1964. The Cairnsmore of Fleet granite and its aureole. Unpublished PhD thesis, University of Newcastle upon Tyne.

Parslow, G R. 1968. The physical and structural features of the Cairnsmore of Fleet granite and its aureole. Scottish Journal of Geology, Vol.?4, 91–108.

Parslow, G R. 1971. Variations in mineralogy and major elements in the Cairnsmore of Fleet granite, SW Scotland. Lithos, Vol.?4, 43–55.

Parslow, G R, and Randall, B A O. 1973. A gravity survey of Cairnsmore of Fleet Granite and its environs. Scottish Journal of Geology, Vol.?9, 219–231.

Peach, B N, and Horne, J. 1899. The Silurian rocks of Britain, Volume?1. Scotland. Memoir of the Geological Survey of the United Kingdom. (London: HMSO.)

Pearce, J A. 1982. Trace element characteristics of lavas from destructive plate boundaries. 525–548 in Andesites. Thorpe,?R?S (editor). (Chichester: Wiley.)

Pearce, J A. 1983. Role of sub-continental lithosphere in magma genesis at active continental margins. 230–249 in Continental basalts and mantle xenoliths. Hawkesworth, C J, and Norry, M J (editors). (Nantwich: Shiva.)

Pearce, J A, Harris, N B W, and Tindle, A G. 1984. Trace element discrimination diagrams for the tectonic interpretation of granitic rocks. Journal of Petrology, Vol.?25, 956–983.

Pettijohn, F J. 1975. Sedimentary rocks. Third edition. (New York: Harper and Row).

Pettijohn, F J, Potter, P E, and Siever, R. 1972. Sand and sandstones. (New York: Springer Verlag.)

Phillips, E R. 1992. Microfabric analysis of a series of sheared metasandstones exposed within the Moniaive Shear Zone, Southern Uplands, Scotland. British Geological Survey Technical Report,WH/92/45.

Phillips, E R. 1994. Microstructural study of the Moniaive Shear Zone, Southern Uplands, Scotland. British Geological Survey Technical Report,WG/94/2.

Phillips, E R. 1996. Mineralogy and petrology of a suite of xenoliths from a lamprophyric microdiorite dyke exposed in the Glenluce area, Southern Uplands, southwest Scotland. British Geological Survey, Short Report, MPSR/96/44.

Phillips, E R, Barnes, R P, Merriman, R J, and Floyd, J D. 1995a. The tectonic significance of Ordovician basic igneous rocks in the Southern Uplands, SW Scotland. Geological Magazine, Vol.?132, 549–556.

Phillips, E R, Barnes, R P, Boland, M P, Fortey, N J, and McMillan, A A. 1995b. The Moniaive Shear Zone: a major zone of sinistral strike-slip deformation in the Southern Uplands of Scotland. Scottish Journal of Geology, Vol.?31, 139–149.

Pickering, K T, Stow, D A V, Watson, M P, and Hiscott, R N. 1986. Deep water facies, processes and models: a review and classification scheme for modern and ancient sediments. Earth Science Reviews, Vol.?23, 75–174.

Pickering, K T, Hiscott, R N and Hein, F J. 1989. Deep marine environments: clastic sedimentation and tectonics. (London: Unwin Hyman Ltd.)

Pidgeon, R T, and Aftalion, M. 1978. Cogenetic and inherited zircon U–Pb systems in granites: Palaeozoic granites of Scotland and England. 183–220 in Crustal evolution in northwestern Britain and adjacent regions. Bowes,?D?R, and Leake,?B?E (editors). Geological Journal Special Issue, No.?10. (Liverpool: Seel House Press.)

Powell, D W. 1970. Magnetised rocks within the Lewisian of western Scotland and under the Southern Uplands. Scottish Journal of Geology, Vol.?6, 353–369.

Price, R J. 1961. The deglaciation of the Tweed drainage area west of Innerleithen. Unpublished PhD thesis. University of Edinburgh.

Price, R J. 1963. The glaciation of a part of Peebles-shire, Scotland. Transactions of the Edinburgh Geological Society, Vol.?19, 326–348.

Price, R J. 1983. Scotland's environment during the last 30?000 years. (Edinburgh: Scottish Academic Press.)

Read, H H. 1926. The mica-lamprophyres of Wigtownshire. Geological Magazine, Vol.?63, 422–429.

Reynolds, D L. 1931. Dykes of the Ards Peninsula, County Down. Geological Magazine, Vol.?68, 97–111 and 145–165.

Rickards, R B. 1976. The sequence of Silurian graptolite zones in the British Isles. Geological Journal, Vol.?11, 153–188.

Riva, J. 1976. Climacograptus bicornis bicornis (Hall), its ancestor and likely descendants. 589–619 in The Ordovician System. Bassett, M G (editor). (Cardiff: University of Wales Press and National Museum of Wales.)

Roberts, B, Morrison, C, and Hirons, S. 1990. Low grade metamorphism of the Manx Group, Isle of Man: a comparative study of white mica 'crystallinity' techniques. Journal of the Geological Society of London, Vol.?147, 271–277.

Roberts, B, Merriman, R J, and Pratt, W. 1991. The influence of strain, lithology and stratigraphical depth on white mica (illite) crystallinity in mudrocks from the vicinity of the Corris Slate Belt, Wales: implications for the timing of metamorphism in the Welsh Basin. Geological Magazine, Vol.?128, 633–645.

Roberts, B, Merriman, R J, Hirons, S R., Fletcher, C J N, and Wilson, D. 1996. Synchronous very low-grade metamorphism, contraction and inversion in the central part of the Welsh Lower Palaeozoic Basin. Journal of the Geological Society of London, Vol.?153, 277–285.

Robinson, D, Warr, L, and Bevins, R E. 1990. The illite 'crystallinity' technique: a critical appraisal of its precision. Journal of Metamorphic Geology,Vol.?8, 333–344.

Robson, D A, and Green, A G. 1980. A magnetic survey of the aureole around the Cheviot granite. Scottish Journal of Geology, Vol.?16, 11–27.

Rock, N M S. 1984. Nature and origin of calc-alkaline lamprophyres: minettes, vogesites, kersantites and spessartites. Transactions of the Royal Society of Edinburgh: Earth Sciences. Vol.?74, 193–227.

Rock, N M S, and Rundle, C C. 1986. Lower Devonian age for the 'Great (basal) Conglomerate', Scottish Borders. Scottish Journal of Geology, Vol. 22, 285–288.

Rock, N M S, Cooper, C, and Gaskarth, J W. 1986a. Late Caledonian subvolcanic vents and associated dykes in the Kirkcudbright area, Galloway, south-west Scotland. Proceedings of the Yorkshire Geological Society, Vol.?46, 29–37.

Rock, N M S, Gaskarth, J W, and Rundle, C C. 1986b. Late Caledonian dyke-swarms in southern Scotland: a regional zone of primitive K-rich lamprophyres and associated vents. Journal of Geology, Vol.?94, 505–522.

Rock, N M S, Gaskarth, J W, Henney, P J, and Shand, P. 1988. Late Caledonian dyke-swarms of northern Britain: some preliminary petrogenetic and tectonic implications of their province-wide distribution and chemical variation. Canadian Mineralogist, Vol.?26, 3–22.

Rollin, K E. 1983. Geophysical assistance to geological mapping in Wigtownshire, south-west Scotland. British Geological Survey, Applied Geophysics Unit Report, No.?156.

Rollinson, H R. 1993. Using geochemical data: evaluation, presentation, interpretation. (Harlow, UK: Longman Scientific and Technical.)

Rose J. 1989. Stadial type sections in the British Quaternary. 45–67 in Quaternary Type Sections: imagination or reality? Rose,?J, and Schlüchter (editors). (Rotterdam: Balkema.)

Rushton, A W A. 1990. Ordovician graptolite biostratigraphy in the Welsh Basin: a review. Journal of the Geological Society of London. Vol.?147, 611–614.

Rushton, A W A, Stone, P, and Hughes, R A. 1996. Biostratigraphical controls of thrust models for the Southern Uplands of Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?86, 137–152.

Russell, A. C.1917. Talnotry nickel mine, Newton Stewart, Kirkcudbright-shire. Unpublished notes, British Museum (Natural History).

Rust, B R. 1965. The stratigraphy and structure of the Whithorn area of Wigtownshire, Scotland. Scottish Journal of Geology, Vol.?1, 101–133.

Sample, J C, and Moore, J C. 1987. Structural style and kinematics of an underplated slate belt, Kodiak and adjacent islands, Alaska. Geological Society of America Bulletin, Vol.?99, 7–20.

Sanderson, D J, Anderson, T B, and Cameron, T D J. 1985. Strain history and the development of transecting cleavage, with examples from the Caledonides of the British Isles (Abstract). Journal of Structural Geology, Vol.?7, 498.

Shackleton, N J, and Opdyke, N D. 1973. Oxygen isotope and palaeomagnetic stratigraphy of Equatorial Pacific core V28–238: oxygen isotope temperatures and ice volumes on a 105 year and 106 year scale. Quaternary Research, Vol.?3, 39–55.

Shackleton, N J, and Opdyke, N D. 1976. Oxygen isotope and palaeomagnetic stratigraphy of Equatorial Pacific core V28–239, Late Pliocene to Latest Pleistocene. 449–464 in Investigation of late Quaternary paleoceanography and palaeoclimatology. Cline, R M, and Hays, J D (editors). Geological Society of America Memoir, No.?145. (Boulder: Geological Society of America.)

Shand, P. 1989. Late Caledonian magmagenesis in southern Scotland. Unpublished PhD thesis. Aston University.

Shand, P, Gaskarth, J W, Thirlwall, M F, and Rock,?N?M?S. 1994. Late Caledonian lamprophyre dyke swarms of southern Scotland. Mineralogy and Petrology, Vol.?51, 277–298.

Simpson, J B. 1933. The late-glacial readvance moraines of the Highland border west of the River Tay. Transactions of the Royal Society of Edinburgh, Vol.?57, 633–645.

Sissons, J B. 1967a. The evolution of Scotland's scenery. (Edinburgh: Oliver and Boyd.)

Sissons, J B. 1967b. Glacial stages and radiocarbon dates in Scotland. Scottish Journal of Geology, Vol.?3, 375–381.

Sissons, J B. 1976. The geomorphology of the British Isles: Scotland. (London: Methuen.)

Sissons, J B. 1983. Quaternary. 399–424 in The Geology of Scotland. Second edition. Craig, G Y. (editor) (Edinburgh: Scottish Academic Press.)

Smith, J. 1904. On the occurrence of conodonts in the Arenig–Llandeilo formations of the Southern Uplands of Scotland. Transactions of the Natural History Society of Glasgow, Vol.?7 (New Series), 235–252.

Smith, I F and Royles, C P. 1989. The digital aeromagnetic survey of the United Kingdom. British Geological Survey Technical Report, WK/89/5.

Soper, N J, England, R W, Snyder, D B, and Ryan, P D. 1992. The Iapetus suture zone in England, Scotland and eastern Ireland: a reconciliation of geological and deep seismic data. Journal of the Geological Society of London, Vol.?149, 697–700.

Soper, N J, Webb, B C, and Woodcock, N H. 1987. Late Caledonian (Acadian) transpression in north-west England: timing, geometry and geotectonic significance. Proceedings of the Yorkshire Geological Society, Vol.?46, 175–192.

Stacey, J S, and Kramers, J D. 1975. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letters, Vol.?26, 207–221.

Stanley, C J, Symes, R F, and Jones, G C. 1987. Copper-nickel mineralisation at Talnotry, Newton Stewart, Scotland. Mineralogy and Petrology, Vol.?37, 293–313.

Stephens, W E, and Halliday, A H. 1984. Geochemical contrasts between late Caledonian granitoid plutons of northern, central and southern Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?75, 259–273.

Stephens, W E, Whitley, J E, Thirlwall, M F, and Halliday, A H. 1985. The Criffell zoned pluton: correlated behaviour of rare earth element abundances with isotopic systems. Contributions to Mineralogy and Petrology, Vol.?89, 226–238.

Stone, P. 1988. The Permian successions at Ballantrae and Loch Ryan, south-west Scotland. Report of the British Geological Survey, Vol.?19, No.?2, 13–18.

Stone, P. 1995. Geology of the Rhins of Galloway district. Memoir of the British Geological Survey, sheets 1 and 3 (Scotland). (London: HMSO.)

Stone, P, and Evans, J A. 1995. Nd-isotope study of provenance patterns across the British sector of the Iapetus Suture Zone. Geological Magazine, Vol.?132, 571–580.

Stone, P, Leake, R C, and Jones, R C. 1984. Base metal mineralisation associated with Ordovician shales in south-west Scotland. Mineral Reconnaissance Programme Report, Institute of Geological Sciences, No.?69.

Stone, P, Floyd, J D, Barnes, R P, and Lintern, B C. 1986. A back-arc thrust-duplex model for the Southern Uplands of Scotland (abstract). Newsletter of the Geological Society of London, Vol.?15, No.?1,?5.

Stone, P, Floyd, J D, Barnes, R P, and Lintern, B C. 1987. A sequential back-arc and foreland basin thrust duplex model for the Southern Uplands of Scotland. Journal of the Geological Society of London, Vol.?144, 753–764.

Stone, P, Cook, J M, McDermott, C, Robinson J J, and Simpson,?P?R. 1995. Lithostratigraphic and structural controls on the distribution of As and Au in the southwest Southern Uplands, Scotland. Transactions of the Institution of Mining and Metallurgy Section B: Applied Earth Sciences, Vol.?104, B111–B119.

Stone, P, Kimbell, G S, and Henney, P J. 1997. Basement control on the location of strike-slip shear in the Southern Uplands of Scotland. Journal of the Geological Society of London, Vol.?154, 141–144.

Strachan, I. 1971. A synoptic supplement to 'A monograph of British Graptolites by Miss G L Elles and Miss E R M Wood'. Monographs of the Palaeontographical Society.

Stringer, P, and Treagus, J E. 1980. Non-axial planar S1 cleavage in the Hawick Rocks of the Galloway area, Southern Uplands, Scotland. Journal of Structural Geology, Vol.?2, 317–331.

Stringer, P, and Treagus, J E. 1981. Asymmetrical folding in the Hawick Rocks of the Galloway area, Southern Uplands. Scottish Journal of Geology, Vol.?17, 129–147.

Styles, M T, Stone, P, and Floyd, J D. 1989. Arc detritus in the Southern Uplands: mineralogical characterisation of a 'missing' terrane. Journal of the Geological Society of London, Vol.?146, 397–400.

Styles, M?T, Perez-Alvarez, M, and Floyd, J?D. 1995. Pyroxenous greywackes in the Southern Uplands and their petrotectonic implications. Geological Magazine, Vol. 132, 539–547.

Sutherland, D G. 1984. The Quaternary deposits and landforms of Scotland and the neighbouring shelves: a review. Quaternary Science Reviews, Vol.?3, 157–254.

Swarbrick, A. 1992. Caledonian minor intrusions from the Midland Valley and Southern Uplands of Scotland: their geochemistry, isotopic characteristics and petrogenesis. Unpublished PhD thesis. University of Birmingham.

Talbot, M R, Holm, K, and Williams, M A J. 1994. Sedimentation in low-gradient desert margin systems: a comparison of the late Triassic of northwest Somerset (England) and the late Quaternary of east-central Australia. 97–117 in Palaeoclimate and basin evolution of playa systems. Rosen, M R (editor). Geological Society of America Special Paper, No.?289.

Tandy, B C. 1974. Radiometric and geochemical reconnaissance of the Cairnsmore of Fleet granite, Kirkcudbrightshire. Radioactive and Metalliferous Minerals Unit Report, No.?324 (unpublished).

Thirwall, M F. 1988. Geochronology of late Caledonian magmatism in northern Britain. Journal of the Geological Society of London, Vol.?145, 951–967.

Tindle, A G, and Pearce, J A. 1981. Petrogenetic modelling of in situ fractional crystallization in the zoned Loch Doon pluton, Scotland. Contributions to Mineralogy and Petrology, Vol.?78, 196–207.

Toghill, P. 1970. The south-east limit of the Moffat Shales in the upper Ettrick Valley region, Selkirkshire. Scottish Journal of Geology, Vol.?6, 233–242.

Vaughan, A P M. 1996. A tectonomagmatic model for the genesis and emplacement of Caledonian calc-alkaline lamprophyres. Journal of the Geological Society of London, Vol.?153, 613–623.

Walton, E K. 1955. Silurian greywackes in Peeblesshire. Proceedings of the Royal Society of Edinburgh, Vol.?B 65, 327–357.

Walton, E K. 1961. Some aspects of the succession and structure in the Lower Palaeozoic rocks of the Southern Uplands of Scotland. Geologische Rundschau, Vol.?50, 63–77.

Warren, P T. 1964. The stratigraphy and structure of the Silurian rocks southeast of Hawick, Roxburghshire. Quarterly Journal of the Geological Society of London, Vol.?120, 193–218.

Watson, J. 1911. British and foreign building stones: a descriptive catalogue of the specimens in the Sedgwick Museum, Cambridge. (Cambridge: University Press.)

Watson, E. 1977. The periglacial environment of Great Britain during the Devensian. Philosophical Transactions of the Royal Society of London, Section B, Vol.?280, 183–198.

Welsh, W. 1964. The Ordovician rocks of north-west Wigtownshire. Unpublished PhD thesis, University of Edinburgh.

Weir, J A. 1968. Structural history of the Silurian rocks of the coast west of Gatehouse, Kirkcudbrightshire. Scottish Journal of Geology, Vol.?4, 31–52.

Weir, J A. 1974. The sedimentology and diagenesis of the Silurian rocks on the coast west of Gatehouse, Kirkcudbrightshire. Scottish Journal of Geology, Vol.?10, 165–186.

Whitaker, A (editor). 1985. Atlas of onshore sedimentary basins in England and Wales: post-Carboniferous tectonics and stratigraphy. (Glasgow and London: Blackie.)

White, D E, Barron, H F, Barnes, R P, and Lintern, B C. 1991. Biostratigraphy of late Llandovery (Telychian) and Wenlock turbiditic sequences in the SW Southern Uplands, Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?82, 297–322.

Williams, S H. 1982a. The Late Ordovician graptolitic fauna of the Anceps Bands at Dob's Linn, southern Scotland. Geologica et Palaeontologica, Vol.?16, 29–56.

Williams, S H. 1982b. Upper Ordovician graptolites from the top lower Hartfell Shale Formation (D. clingani and P. linearis zones) near Moffat, southern Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?72, 229–255.

Williams, S H. 1983. The Ordovician–Silurian boundary graptolite fauna of Dob's Linn, Southern Scotland. Palaeontology, Vol.?26, 605–639.

Williams, S H. 1987. Upper Ordovician graptolites from the D. complanatus Zone of the Moffat and Girvan districts and their significance for correlation. Scottish Journal of Geology, Vol.?23, 65–92.

Williams, S H. 1994. Revision and definition of the C. wilsoni Zone (middle Ordovician) of Southern Scotland. Transactions of the Royal Society of Edinburgh: Earth Sciences, Vol.?85, 143–157.

Wilson, G V, and Flett, J D. 1921. The lead, zinc, copper and nickel ores of Scotland. Memoirs of the Geological Survey of Scotland, special reports on the mineral resources of Great Britain, Vol.?17. (London: HMSO.)

Wilson. 1989. Igneous petrogenesis. (London: Unwin Hyman.)

Zalasiewicz, J A, Rushton, A W A, and Owen, A W. 1995. Late Caradoc graptolite faunal gradients across the Iapetus Ocean. Geological Magazine, Vol.?132, 611–618.

Appendix 1 Ordovician graptolite localities, sheet 4

'

No. Locality Grid ref NX BGS internal Palaeontological Dept report and item numbers Graptolite zone
1 Tannylaggie Flow [NX 287 715] & [NX 287 717] 87/99.14 gracilis
2 Camrie [NX 2018 6040] 82/41.10 gracilis or 'peltifer'
3 Camrie [NX 2050 6055] 82/41.5 'peltifer'
4 Camrie [NX 2050 6055] 82/41.2 clingani (caudatus Subzone)
5 Camrie [NX 2050 6055] 82/41.3 clingani (morrisi Subzone?)
6 Camrie [NX 2050 6055] 82/41.4 linearis
7 Camrie [NX 2050 6055] 82/41.9 poss. clingani
8 Camrie [NX 2085 6065] 82/41.1 prob. clingani
9 Balminnoch [NX 266 653] 87/99.11 prob. clingani to linearis
10 Drumabrennan [NX 296 671] 87/99.12 clingani or linearis
11 Carseriggan [NX 308 680] 87/99.23 'peltifer' or wilsoni
12 Garwachie [NX 344 693] 87/99.24 gracilis or 'peltifer'
13 Wood of Cree [NX 380 715] 87/99.25 clingani
14 River Cree, E of Knockstocks [NX 4016 6696] 88/234.4–6, 89/427.1–3 linearis?
15 Penkiln Burn [NX 4239 6806] 88/234.7 linearis to anceps
16 Risk v4494 6957] 89/378, 88/234.8 ?clingani or linearis
17 Gillespie Burn [NX 257 539] 83/29.4&5 clingani or linearis
18 Gillespie Burn [NX 257 539] 83/29.13 prob. linearis
19 Gillespie Burn [NX 257 539] 83/29.1,2, 12 anceps
20 Gillespie Burn [NX 257 539] 83/29.7, 8 lower anceps
21 Gillespie Burn [NX 257 539] 83/29.9 anceps
22 Gillespie Burn [NX 257 539] 83/29.15 clingani or linearis
23 Gillespie Burn [NX 2610 5400] 83/29.16&17 linearis
24 Gillespie Burn [NX 2610 5400] 83/29.18&19 linearis
25 Gillespie Burn [NX 2610 5400] 83/29.16 linearis
26 Barhaskine [NX 2688 5431] 87/99.4 wilsoni to anceps
27 Garheugh [NX 2752 5025] 84/23.8 clingani or linearis
28 Garheugh [NX 2756 5037] 84/23.7 anceps
29 Low Glenling [NX 325 519] to [NX 375 523] 87/99.18–20 gracilis

Appendix 2 Ordovician graptolites, sheet 4

Species Localities: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Arachnograptus sp. +
Climacograptus cf. antiquus +
Climacograptus bicornis + cf + +
Climacograptus [Normalograptus] brevis? +
Climacograptus dorotheus +
Climacograptus latus + ? + ?
Climacograptus longispinus group? + +
Climacograptus miserabilis + +
Climacograptus [Normalograptus] mohawkensis ? cf + cf
Climacograptus spiniferus cf ? +
Climacograptus styloideus + cf +
Climacograptus supernus + + + +
Climacograptus tubuliferus? ?
Climacograptus sp. + + + + + + + + + ? + +
Corynoides calicularis + + +
Corynoides curtus + + ?
Cryptograptus tricornis ss cf sl sp
Dicellograptus cf. angulatus + +
Dicellograptus complexus + +
Dicellograptus elegans + + cf
Dicellograptus flexuosus cf cf + ? cf +
Dicellograptus morrisi + cf +
Dicellograptus ornatus +
Dicellograptus pumilus cf +
Dicellograptus salopiensis +
Dicellograptus sextans +
Dicellograptus sextans exilis + cf
Dicellograptus sp. + + + + + + +
Dicranograptus cf. furcatus minimus +
Dicranograptus nicholsoni + cf + +
Dicranograptus ramosus aff +
Dicranograptus ramosus longicaulis +
Dicrograptus ziczac +
Didymograptus superstes + +
Glyptograptus sp. + ? ?
Hallograptus mucronatus nobilis + +
Hallograptus sp.
Lasiograptus harknessi ? cf +
Leptograptus flaccidus arcuatus +
Leptograptus flaccidus flaccidus sl + sl + sl
Leptograptus flaccidus macer +
Leptograptus flaccidus macilentus? +
Leptograptus flaccidus spinifer + + +
Leptograptus sp. + ? +
Nemagraptus gracilis + +
Species Localities: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Nemagraptus pertenuis +
Neuragraptus sp. +
Orthograptus abbreviates + ? + +
Orthograptus amplexicaulis group + + cf + +
Orthograptus basilicus + cf
Orthograptus calcaratus group + cf cf + + + + cf
Orthograptus fastigatus + + +
Orthograptus intermedius + cf
Orthograptus pageanus +
Orthograptus pageanus abnormispinosus +
Orthograptus pageanus micracanthus + cf
Orthograptus pauperatus? +
Orthograptus quadrimucronatus + + + +
Orthograptus whitfieldi? cf
Orthograptus sp. + + + + + + + + + +
Plegmatograptus nebula + + ?
Plegmatograptus sp. ?
Pleurograptus linearis + sp? + sp? sp
Pleurograptus lui +
Psendoclimocograptus scharenbergi +
Thamnograptus sp. +

Appendix 3 Silurian graptolite localities, sheets 2 and 4

'

No. Locality Grid ref BGS internal Palaeontological Dept report and item numbers Graptolite zone
30 Fishhouse [NX 1988 5526] 83/4.1 ?acuminatus
31 Fishhouse [NX 2007 5463] 83/4.2 acuminatus to acinaces*
32 Mindork [NX 3207 5827] 87/233.6 mid-acuminatus
33 Kilfillan [NX 2015 5427] 83/4.3, 87/233.1, 2 acinaces or cyphus
34 Low Sinniness [NX 215 521] 87/99.2 acuminatus to cyphus*
35 Fishhouse [NX 2095 5309] 83/4.4 cyphus (poss. upper)
36 Laigh Sinniness [NX 2154 5214] 83/4.6 cyphus
37 Gillespie Burn [NX 2565 5398] 83/29.10 acuminatus
38 Gillespie Burn [NX 2565 5398] 83/29.6 poss. atavus
39 Gillespie Burn [NX 2565 5398] 83/29.11 poss. acinaces
40 Culroy [NX 2555 5397] 83/29 prob. cyphus
41 Quarry near Barhaskine [NX 2806 5463] 87/99.5, 87/233.3 acuminatus & atavus
42 Garheugh [NX 2750 5020] 84/23.11, 12 acuminatus
43 Garheugh [NX 2752 5025] 84/23.10 atavus or acinaces
44 Garheugh [NX 2752 5025] 84/23.9 triangulatus
45 Garheugh [NX 2756 5037] 84/23.6 acuminatus
46 Garheugh [NX 2771 5057] 84/23.5 atavus or acinaces
47 Garheugh [NX 2783 5065] 84/23.1 persculptus to acinaces*
48 Garheugh [NX 2783 5065] 84/23.2, 3 atavus
49 Black Loch [NX 281 509] 87/99.6 acuminatus
50 Crailloch [NX 328 526] 87/99.17 atavus to low cyphus
51 Eldrig [NX 321 487] 87/99.16 ?cyphus
52 Alticry [NX 2846 5024] 84/52.A1, 2, 98/184.1 guerichi to crispus*
53 Whaup Hill [NX 405 502] 84/52.B, 98/184.3 upper guerichi
54 Baldoon Mains [NX 424 539] 84/52.B, 98/184.2 turriculatus or crispus upper guerichi or lowermost turriculatus
55 Kirkmaiden [NX 3632 4034] to [NX 3725 3913] 84/52C1-6, 98/183 crenulata
56 Burrow Head [NX 4538 3411] to [NX 4638 3430] 84/52D1-5 centrifugus–? '
57 Isle of Whithorn [NX 480 360 84/52E1-3 ?centrifugus

* indicates that the horizon falls within the zonal range and not that all of the listed zones are present.

Appendix 4 Silurian graptolites, sheets 2 and 4

Species Localities: 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57
Akidograptus ascensus +
Atavograptus atavus + + + sp?
Atavograptus gracilis? +
Atavograptus strachani +
Climacograptus innotatus ? + + +
Climacograptus medius + + + + + + ? + +
Climacograptus miserabilis + + + + + + + + +
Climacograptus normalis ? + + + + + + + + + + +
Climacograptus rectangularis + + ? ? +
Climacograptus trifilis +
Climacograptus tuberculatus +
Climacograptus sp. + + + + + + ?
Coronograptus cyphus? +
Coronograptus gregarius ? +
Cyrtograptus centrifugus + +
Cyrtograptus aff. insectus +
Coronograptus hipposideros +
Cystograptus vesiculosus + + ? ?
Dimorphograptus cf. longissimus + sp?
Diplograptus modestus + + ? +
Diplograptus modestus parvulus? +
Glyptograptus cf. enodis +
Glyptograptus cf. tamariscus ?
Glyptograptus sp. ? ? +
Glyptograptus tamariscus linearis +
Lagarograptus acinaces? +
Monoclimacis crenulata (sensu Elles & Wood) +
Monoclimacis griestoniensis +
Monoclimacis vomerina vomerina + +
Monoclimacis vomerina cf. basilica + +
Monoclimacis woodae +
Monograptus cf. danbyi +
Monograptus priodon + + +
Monograptus revolutus group +
Monograptus cf. riccartonensis +
Monograptus triangulatus triangulatus +
Orthograptus cyperoides sp? ? +
Orthograptus insectiformis +
Parakidograptus acuminatus ? + + + + +
Petalolithus sp. +
Petalograptus ovatoelongatus +
Pribylograptus sandersoni? cf + +
Pristiograptus sp. +
Pristiograptus concinnus? + c.f. +
Retiolites geinitzianus aungustidens (Elles & Wood 1908) +
Retiolites geinitzianus geintizianus (Barrande, 1850) +
Rhaphidograptus extenuatus +
Rhaphidograptus toernquisti +
Stimulograptus halli +
Streptograptus plumosus + +
Torquigraptus cf. germanicus +
Torquigraptus planus +
Torquigraptus tullbergi ?

Appendix 7 Data and analytical details for U/Pb dating of the Cairnsmore of Fleet granite

The Cairnsmore of Fleet granite was dated using the U-Pb zircon technique. A sample of between 30–50 kg was jaw crushed and disc milled and the <400 micron fraction was then sieved out. Heavy mineral concentrates were obtained using a Gemini shaking table followed by a superpanner. A >3.3gm/ml density separate was recovered using Di-iodomethane and the minerals separated magnetically using a Frantz LB-1 magnetic separator. The recovered zircons were hand picked under alcohol and abraded. U and Pb separations followed the procedures of Krogh (1973) with minor modifications of Corfu and Ayres (1984). Zircon fractions were spiked with a mixed 205Pb/235U isotopic tracer (Krogh and Davis, 1975) before digestion and chemical analysis. U and Pb were loaded onto outgassed single Re filaments with silica gel and were analyzed on a VG 354 mass spectrometer using a Daly detector as described by Noble et al. (1993). Chemistry blanks were c. 5 pg, and these were monitored in each batch of chemistry. Uranium blanks were <0.1 pg U. All results and errors were calculated following Ludwig (1993, 1994) and the Pb isotope ratios were corrected for initial common Pb in excess of laboratory blank using the Stacey and Kramers model (1975). Ages were calculated using the decay constants of Jaffey et al. (1971). The 2s error (95% confidence levels) for ages and isotope ratios, given in (Table 3), were obtained by propagating key sources of error through all calculations following the methods used by Ludwig (1993, 1994).

The Cairnsmore of Fleet granite samples yielded ample zircons for analysis, but they were generally highly fractured and of poor quality; it was difficult to select clear, unfractured grains. Nearly all the needle-like grains were fractured so, whilst they were by far the most abundant form, only two fractions of them were used. Two other fractions were selected from the rarer, but better quality, stubby crystals. The granite also contained monazite, which is a useful mineral in U/Pb geochronology because it does not usually show inheritance of older isotope signatures.

In (Figure 39), three fractions can be seen to lie on Concordia, CF1b, CF1c and the monazite fraction. These three determinations combine to give a concordant age of 395 ± 3 (2s) with a MSWD of concordance and equivalence of 2.6. The concordant age is in agreements with the published zircon age of 390 ± 6 Ma (Aftalion and Pidgeon, 1978) and a Rb-Sr age of 397 ± 2 Ma (2s) (Halliday et al., 1980). Fraction CF1d is slightly discordant, whilst fraction CF1a shows considerable inheritance and plots away from the main data grouping. Regression of the concordant data with CF1a gives an upper intercept age of 1793 ± 60 Ma (2s), suggesting that inherited grains were derived from an early Proterozoic source.

(Table 3) Analytical data for the Cairnsmore of Fleet zircon and monazite fractions.

Fractions Concentrations** Atomic ratios Age
Fraction* Number of grains size (µm) description Wgt** U ppm** Pb ppm** Common 206Pb/204Pb§ 208Pb/206Pb† 206Pb/238U†± %2σ 207Pb/235U†± %2σ 207Pb/206Pb†± %2σ 207Pb/206Pbage Rho¥
CF 1a M2° 20 100–250 eu <1:3 0.0188 34.2 453 23.5 1658 .1509 0.6149 ± .29 0.0714 ± 0.37 0.06256 ± .29 690 .79
CF 1b M2° 37 150–300 nee >1:3 0.0047 34.6 485 19.5 483 .1959 0.4761 ± .71 0.0635 ± .71 0.05438 ± .71 387 .73
CF 1c DM 45 c.100 st eu 0.0050 45.9 678 13.9 980 .1747 0.4728 ± .77 0.0629 ± .54 0.05454 ± .77 394 .61
CF 1d DM 22 50–100 st eu 0.0024 19.0 319 4.2 741 .200 0.4869 ± 1.1 0.0638 ± 1.0 0.05531 ± 1.1 425 .68
CF monazite 13 c.125 cl, eu 0.0018 946 719 58.5 107 15.96 0.4868 ± 1.8 0.0638 ± 1.1 0.05501 ± 1.4 413 .66
  • * Magnetic properties correspond to the tilt angle of a Frantz LB-1 Separator at full magnetic field. M = magnetic, NM = non magnetic, DM = diamagnetic, Eu = euhedral, pris; prismatic, st; stubby < 2:1 nee., needles > 3:1, cl = clear
  • ** Sample weight errors, and hence U and Pb concentrations, are approximate.
  • § Measured ratios are corrected for fractionation and common Pb in spike.
  • † Corrected for fractionation, spike, laboratory blank Pb and U, and initial common Pb at 400 Ma (Stacey and Kramers, 1975). Errors for the measured ratios propagated through the data reduction calculations were ± 2 standard errors of the mean. Errors are quoted for the last two decimal places.
  • ¥ Correlation coefficients of 207Pb/235U to 206Pb/238U are calculated using the procedures and algorithms of Ludwig (1993, 1994).

Fossil inventory

To satisfy the rules and recommendations of the international codes of botanical and zoological nomenclature, authors of cited species are listed below.

Some generic names revised [in square brackets].

Figures, plates and tables

Figures

(Figure 1) Solid geology of the district.

(Figure 2) Topography of the district.

(Figure 3) The district in its regional context within the Southern Uplands (1:50?000 Sheet?2 Whithorn, Sheet 4W Kirkcowan and Sheet?4E Wigtown).

(Figure 4) Bouguer gravity anomaly map of the district and the surrounding area. Crosses indicate gravity stations. Reduction density = 2.72?Mg/m3. Contour interval = 1mGal (10-5?m/s2).

(Figure 5) Reduced-to-the-pole aeromagnetic map of the district and the surrounding area. Crosses indicate digitised points along flight lines. Contour interval = 10?nT.

(Figure 6) Geophysical summary map showing local geophysical features discussed in the text.

(Figure 7) Gravity model for profile?A (Figure?6). Densities assumed in modelling (in Mg/m3) are indicated. Half strike length of bodies: 10?km (Cairnsmore of Fleet); 4?km (Loch Doon); 2D (dense body to south-east). Background field removed = 15?mGal.

(Figure 8) Gravity and magnetic model for profile B (Figure?6). Numbers indicate density (Mg/m3)/ susceptibility (10-3?SI units). Half strike length of bodies = 7?km. The assumed background gravity field decreases linearly from 19?mGal at 0?km to 9?mGal at 50?km.

(Figure 9) Biostratigraphical age profile of the Lower Palaeozoic rocks in the district. See (Figure 1) for abbreviations.

(Figure 10) Ideal sequence of sedimentary structures in a turbidite bed (from Pickering et al. 1989, after Bouma, 1962).

(Figure 11) Examples of Ordovician and Silurian graptolites from the Moffat Shale Group. All drawings are magnified by 3?4 except i which is 3?2.7. The parallel lines in figs.?f and g indicate the trace of the cleavage on the bedding-plane. The locality numbers refer to the distribution table in Appendix?1, where grid references and associated taxa are also given. All specimens are in the BGS palaeontological collections. a,?b Cyrtograptus centrifugus Boucek. a, 10E 886; b, proximal fragment restored from 10E 906 and 907. Ross Formation, centrifugus Biozone, Burrow Head (Loc.?56) [NX 4538?3411]. c Streptograptus plumosus (Baily), GSE 14866. Cairnharrow Formation, guerichi Biozone (see text), cutting N of Whauphill Station (Loc.?53) [NX 405?502]. d Monograptus triangulatus triangulatus (Harkness), 9E 6816. Birkhill Shale, triangulatus Biozone, Garheugh (Loc.?44) [NX 2752?5025]. e 'Orthograptus' cyperoides (Tornquist), 9E 6784. Birkhill Shale, triangulatus Biozone, Garheugh (Loc.?44) [NX 2752 5025]. f,?g Climacograptus supernus (Elles & Wood), both on 9E 6383. Upper Hartfell Shale, anceps Biozone, Gillespie Burn (Loc.?19) [NX 2570 5396]. h Orthograptus abbreviatus (Elles &?Wood), 9E 6202. Upper Hartfell Shale, anceps Biozone, Gillespie Burn (Loc.?21) [NX 2570 5396]. i Climacographtus trifilis Manck, GSE 14579. Mindork Formation, acumimatus Biozone, Mindork Fell (Loc.?32) [NX 3207?5827]. j Dimorphograptus longissimus (Kurch), GSE 14581. Birkhill Shale, acinaces Biozone, quarry 2?km ENE?of Barhaskine (Loc.?41) [NX 2806 5463]. k Atavograptus strachani (Hutt & Rickards), 9E 6155. Lower beds of Gala Formation, cyphus Biozone, Culroy (Loc.?40) [NX 2555 5397]. l Normalograptus normalis (Lapworth), part of 9E 6565, an external mould of a long specimen. Birkhill Shale, acuminatus to cyphus Biozone, 500?m NE of Garheugh (Loc.?47) [NX 2783 5065]. m Climacograptus tuberculatus Nicholson, distal fragment restored from 9E 6862 and 6863. Birkhill Shale, acuminatus Biozone, 250?m SE of Garheugh (Loc.?42–44) [c.[NX 2755 5048]. n Parakidograptus acuminatus (Nicholson), GSE 15036. Birkhill Shale, acuminatus Biozone, quarry of Barhaskine (Loc.?41) est. [NX 2815 5463]. o, p Cryptograptus tricornis (Carruthers); o, scalariform view, the apertures showing as holes, reconstructed from 8E 8383 and 8384; p, biprofile view, 8E 8394. Glenkiln Shale, peltifer Biozone, Camrie Burn (Loc.?3) [NX 2050 6058]. q Pleurograptus linearis (Carruthers), fragment of mature rhabdosome, 8E 8363. Lower Hartfell Shale, linearis Biozone, Camrie Burn (Loc.?6) [NX 2050 6057]. r Nemagraptus gracilis (Hall), GSE 15037. Glenkiln Shale, gracilis Biozone, Tannylaggie (Loc.?1) c.?[NX 287 717]. s Leptograptus flaccidus (Hall), proximal part of a long specimen, 8E 8347B. Lower Hartfell Shale, linearis Biozone, Camrie Burn (Loc.?6) [NX 2050 6057]. t Fragment of a broad plegmatograptid, possibly referable to Arachniograptus, 8E 8349. Lower Hartfell Shale, linearis Biozone, Camrie Burn (Loc.?6) [NX 2050 6057]. u Dicranograptus ziczac Lapworth, reconstructed from 8E 8383 with additional detail from counterpart 8384. Glenkiln Shale, peltifer Biozone, Camrie Burn (Loc.?3) [NX 2050 6058].

(Figure 12) Detail of the Moffat Shale Group outcrop in Gillespie Burn.

(Figure 13) Detail of the Moffat Shale Group outcrop around Camrie Fell and the Gabsnout Burn (after Barnes et al., 1995).

(Figure 14) Representative measured sections of formations in the Leadhills Supergroup.

(Figure 15) Representative measured sections of formations in the Gala Group.

(Figure 16) Palaeocurrent data from Gala Group formations. Data are from sole structures forming casts of linear grooves and directional flutes (arrows).

(Figure 17) Climacograptus trifilis Manck from a quarry on Mindork Fell [NX 3207?5827], specimen numbers GSE 14579 and 14580, magnification 3?5.7. Specimens showing the three basal spines are characteristically preserved in dorso-ventral orientation, as here. C.?trifilis occurs only in the basal Silurian acuminatus Biozone.

(Figure 18) Representative measured sections of formations in the Hawick Group.

(Figure 19) Palaeocurrent data from Hawick Group formations. Inner circle ripples; outer circle sole structures, directional flute casts shown by arrows.

(Figure 20) Postulated detail of the Hawick Group in Luce Bay in the district.

(Figure 21) Box plots showing the variation in SiO2 and MgO across strike in sandstones. Compilation of regional data from south-west Scotland including this district. See (Figure?1) for abbreviations.

(Figure 22) Summary of the regional evidence for the timing of deformation in the Southern Uplands terrane. Modified from Barnes et al. (1989), with time scale after Gradstein and Ogg, 1996. See Appendix?7 for details of more recent dating of the Cairnsmore of Fleet granite.

(Figure 23a) Summary of structural data from the district: Tract map contoured for dip of bedding.

(Figure 23b) Structural data summary: stereograms for the west of the district.

(Figure 23c) Structural data summary: stereograms for the east of the district.

(Figure 24) Summary of fault orientation data.

(Figure 25) Model for all of the 'cross-strike' fault sets in the district formed by sinistral shear on tract-bounding faults.

(Figure 26) Interpretational map showing the pattern of regional metamorphic grade.

(Figure 27) Range and mean values of white mica crystallinity (Kubler) indices for the main tectonostratigraphical units.

(Figure 28) Orientation and thickness of dykes in the district.

(Figure 29) Histogram of SiO2 content in analysed intrusive rock samples.

(Figure 30) Total alkalis (TAS) and K2O classification of minor intrusions using fields from Le Maitre et al. (1989). a Dyke rocks on a TAS v silica classification diagram (Q=normative quartz) b Lamprophyres on a plot to distinguish low, medium and high K calc-alkaline igneous rocks.

(Figure 31) Major oxide and trace element variation diagrams for minor intrusions; all plotted against SiO2.

(Figure 32) MORB normalised multi-element diagrams for minor intrusions; normalising values from Pearce (1983).

(Figure 33) Total alkali vs. silica classification diagram for the larger intrusions (fields from Wilson, 1989).

(Figure 34) Major oxide and trace element variation diagrams v SiO2 for larger intrusions; all plotted against SiO2. Key as Figure?33.

(Figure 35) Classification of the tectonic setting of the larger granitic intrusions; fields from Pearce et al. (1984). Symbols as in Figure?33.

(Figure 36) Ice flow direction from the orientation of drumlins and striations in the district (after Charlesworth, 1926a).

(Figure 37) Landsat image of drumlin fields in the area north-west of Wigtown.

(Figure 38) Location and mineralogy of mines and trials in the north-east of the district; mineralisation zones after Cook (1976).

(Figure 39) U/Pb concordia diagram for zircon and monazite fractions from the Cairnsmore of Fleet granite.

Plates

(Front cover) Cover photograph Pibble Mine engine house situated on the north-west slope of Pibble Hill, 5 km east-north-east of Creetown [NX 5255 6068] (Photographer T S Bain; P001527).

(Rear cover)

(Frontispiece) Refolded folds ( D1 and D2) in the Kirkmaiden Formation, Back Bay (Photograph F?I?MacTaggart; P001123).

(Geological succession) Geological succession and main events in the district.

(Plate 1a) Working the Kirkmabreck granodiorite c.?1939. a Silver Grey Quarry [NX 480?565]: general view showing the steam-powered crane and 'blondin' (a system of wire ropes and pulleys stretching across the quarry) used to move the granodiorite blocks (C3735).

(Plate 1b) Working the Kirkmabreck granodiorite c.?1939.b Sett-making yard where workers, sheltered in open huts, roughly squared granite blocks for paving, Glebe Quarry, Creetown [NX 479?564] (C3726).

(Plate 2) Photomicrograph of coarse-grained sandstone from the Gala Group (Garheugh Formation) showing typical clast assemblage. Sample from Rocks of Garheugh [NX 2590?5009] (P104254).

(Plate 3) Medium- to very thick-bedded sandstone and thin to medium mudstone beds, Kilfillan Formation, Kilfillan [NX 2023?5431] (D3638).

(Plate 4) Linear sole marks on the base of a thick sandstone bed, Kilfillan Formation, Kilfillan [NX 2023?5431] (D3639).

(Plate 5) Photomicrograph of fine- grained sandstone from the Mindork Formation showing amphibole-feldspar-phyric basaltic clast and amphibole mineral clasts. Sample from Mindork Fell [NX 3210?5838] (P104264).

(Plate 6) Very thickly bedded sandstone, Garheugh Formation, Rocks of Garheugh [NX 2636?5074] (D3625).

(Plate 7) Conjugate faults cutting across near-vertical, thick- to thin-bedded sandstone with thin mudstone partings, Rocks of Garheugh [NX 2665?5025] (D3626).

(Plate 8) Intraclast breccia (pebbly sandstone) from the Mull of Logan Formation, Alticry [NX 2717?4994] (P542162).

(Plate 9) Photomicrograph of coarse-grained sandstone from the Hawick Group (Carghidown Formation) showing typical clast assemblage including spilitic volcanic material and carbonate. Sample from Burrow Head [NX 4490?3415] (P104252).

(Plate 10a) Typical sequence of sandstone and silty mudstone beds in the Kirkmaiden Formation with 'packets' of varying bed thickness, younging from bottom left to top right.a Shore at Low Auchenlarie [NX 5396?5181] (D4004)

(Plate 10b) Typical sequence of sandstone and silty mudstone beds in the Kirkmaiden Formation with 'packets' of varying bed thickness, younging from bottom left to top right.b Shore south of Callie's Port [NX 3726?3912] (P104241)

(Plate 11) Dewatering structures in the Kirkmaiden Formation: small sand volcanoes on the top of a sandstone bed are distorted into elliptical shapes parallel to the cleavage intersection lineation. Shore at Low Auchenlarie [NX 5374?5189] (D3916).

(Plate 12) Flute casts (current from bottom to top) on the base of a steeply dipping sandstone bed, Carghidown Formation, Palmallet Point [NX 4827?4242] (P104242).

(Plate 13) Red mudstone beds in the Carghidown Formation, Burrow Head [NX 4429?3454] (D3631).

(Plate 14) Inclined, moderately plunging D1 fold pair in the Kilfillan Formation (Gala tract?1), shore at Kilfillan [NX 2011?5450] (D3621).

(Plate 15) Steeply inclined, gently plunging, tight D1 fold pair in the Gala Group (Gala tract?5), Rocks of Garheugh [NX 2614?5100] (D3624).

(Plate 16) Moniaive Shear Zone, strongly foliated metasandstone (Gala Group) in the aureole of the Cairnsmore of Fleet granite. Foliation and parallel quartz segregations are folded by kink bands, Bargaly Glen [NX 4680?6784] (D3995).

(Plate 17) Orlock Bridge Fault, quartz segregations parallel to cleavage in silty mudstone (Shinnel Formation) folded by small sinistral 'D3' folds, Wood of Dervaird [NX 2249?5770] (D3933).

(Plate 18) Gently plunging, steeply inclined D1 folds refolded by recumbent D2 folds in the Kirkmaiden Formation, Back Bay [NX 368?394] (P001123).

(Plate 19) Curved D1 fold hinges in the Carghidown Formation, Portyerrock [NX 478?389] (P104243).

(Plate 20) Steeply plunging, sinistral D1 fold pair in the Carghidown Formation, Carghidown [NX 4378?3502] (P104243).

(Plate 21) Upright, open, gently plunging, south-verging fold pair in the style of the larger D2 folds in the northern part of the Hawick Group, Carghidown Formation, Port of Counan [NX 4194?3613] (P104244 or MNS4682).

(Plate 22) Steeply plunging D3 fold associated with the Innerwell Fault, Cairnharrow Formation, Black Rocks [NX 3575?4081] (MNS?3895).

(Plate 23) S1 cleavage folded by kink bands, in a felsic dyke, Kirkbride Hill [NX 5630 5654] (D4009).

(Plate 24) Photomicrograph of S1 cleaved microgranodiorite dyke. Sample S77997 from Barholm Hill [NX 5333?5209] (P104274).

(Plate 25) Lamprophyre dyke parallel to bedding in the Kirkmaiden Formation, shore south of Mossyard [NX 5516?5157] (P104245).

(Plate 26) Lamprophyre dyke cutting D1 fold in the Kirkmaiden Formation, shore at Low Auchenlarie [NX 5407?5165] (P104246).

(Plate 27) Mica lamprophyre dyke, within the dashed lines, 33?cm thick, emplaced parallel to bedding and folded by D1 anticline, shore at Low Auchenlarie [NX 5402?5176] (D4600).

(Plate 28) Lamprophyre (kersantite) dyke (below hammer) folded by a recumbent D2 fold, Callies Port [NX 3706?3931] (D4604/MNS4680).

(Plate 29) Lamprophyre dyke, 1?m thick, stepping sideways in a D2 fold closure suggesting syn- D2 emplacement, south of Callies Port [NX 3723?3913] (D4606/MNS4681).

(Plate 30a) Photomicrographs. a Mica lamprophyre. Sample S72336 from Craignarget Hill [NX 2588?5216] (P104282).

(Plate 30b) Photomicrographs. b Spessartite/appinite. Sample S73286 from Garheugh Port [NX 2694?5010] (P104311).

(Plate 31) Quartz xenoliths in a spessartite dyke, Culvennan Fell [NX 3060?6502] (D4596).

(Plate 32) Photomicrograph of porphyritic microgranodiorite. Sample S78043 from Blackmyre [NX 4979?5724] (P104268).

(Plate 33) Photomicrograph of two-pyroxene microdiorite from the Glenluce intrusion. Sample S71494 from Barlochart Fell [NX 2008?5685] (P104297).

(Plate 34) Vent breccia containing angular fragments of the host sandstone and exotic clasts of a coarser-grained, white sandstone, Culvennan Fell [NX 3102?6498] (D4598).

(Plate 35) Photomicrograph of pyroxene-hornblende-biotite diorite with pilotaxitic fabric from the Culvennan intrusion. Sample S71494 from Bennan Hill [NX 3234?6474] (P104293).

(Plate 36) Photomicrograph of biotite granodiorite from the Carsluith intrusion. S78026 from Bagbie quarry, Carsluith [NX 4888?5492] (P104299).

(Plate 37) Upper (northern) contact of the Kirkmabreck granodiorite dyke and the adjacent swarm of porphyritic microdiorite dykes, Kirkmabreck (Glebe) Quarry [NX 4808?5661] (D3919).

(Plate 38) Photomicrograph of coarse-grained granodiorite, Kirkmabreck Quarry [NX 480?565] (P104303+05+07).

(Plate 39) Cordierite pseudomorphs flattened and aligned in the Moniaive Shear Zone foliation at outcrop, Graddoch Burn, west of Culcronchie Hill [NX 5108?6386] (P104247).

(Plate 40) Detail of cordierite pseudomorphs deformed in the Moniaive Shear Zone foliation in thin section. Sample S81725 from Culcronchie Hill [NX 5152?6378] (P576625).

(Plate 41) Well-developed banding in metasandstone caused by segregation of minerals in the aureole of the Cairnsmore of Fleet pluton, Chapelton [NX 4947?5976] (D4001).

(Plate 42a) Hornfels and skarn alteration.a Biotite hornfels (dark grey) with pale green skarn bands cut by white quartz veins forming host rock to and xenolith within a porphyritic microdiorite dyke, Kirkmabreck (Glebe) Quarry [NX 4802 5660] (D3921).

(Plate 42b) Hornfels and skarn alteration.b Detail of skarn banding in biotite hornfels. Quartz-garnet veins are surrounded by zones of alteration with pyroxene in inner part and epidote in outer part. Sample from Kirkmabreck (Glebe) Quarry.

(Plate 42c) Hornfels and skarn alteration. 42c Photomicrograph of grossular garnet in thin section from skarn vein. Sample from Kirkmabreck (Glebe) Quarry.

(Plate 43) Water-worn, wave cut platform, Chippermore [NX 285 482] (P104248).

Tables

(Table 1) Geological succession and main events in the district.

(Table 2) Summary of events, deposits and conditions in south-west Scotland during the latter part of the Quaternary.

(Table 3) Analytical data for the Cairnsmore of Fleet zircon and monazite fractions.

Apendices

(Appendix 2)

Geology of the Whithorn, Kirkcowan and Wigtown district

This memoir describes the geology of the BGS 1:50 000 map sheets of Whithorn (Sheet 2), Kirkcowan (Sheet 4W) and Wigtown (Sheet 4E), located in Dumfries and Galloway, south-west Scotland. This district includes the town of Newton Stewart and scattered smaller settlements including Glenluce, Kirkcowan, Wigtown, Port William, Whithorn and Isle of Whithorn. The land is mostly agricultural but includes part of the Galloway Forest Park.

Geologically, the district is part of the Southern Uplands terrane, an extensive outcrop of Lower Palaeozoic sedimentary strata formed during closure of the Iapetus Ocean. In the Ordovician and Silurian periods, sediment transported from the continental landmass of Laurentia was deposited in the northern part of the ocean as turbidites on top of muddy oceanic deposits. As the ocean crust was subducted northwards the accumulated strata were progressively deformed and accreted to the Laurentian margin. Igneous rocks, ranging from dyke swarms to the large Cairnsmore of Fleet pluton, were intruded in late Silurian and early Devonian times as collision with the continent to the south joined the landmasses that now form Scotland and England.

Little is preserved from the younger geological periods, although Carboniferous, Permian and Triassic rocks occur offshore in the extensional half-grabens of Luce Bay and Wigtown Bay. Quaternary deposits include till that was deposited during the Devensian glaciation and forms a distinctive pattern of drumlins in many parts of the district. Lacustrine deposits and peat occur locally and raised beaches are present along parts of the coast.

Geological issues relevant to land use, planning and development are described briefly in the Applied geology chapter, including groundwater resources, ground stability, mining and quarrying. Details of fossils and localities are given in the appendices and the book includes a full reference list.

The maps included in this district are available as separate Solid and Drift editions.

Index to the 1:50 000 Series maps of the British Geological Survey

Tables

(Table 2) Summary of events, deposits and conditions in south-west Scotland during the latter part of the Quaternary

Approximate age in years BP Stage Substage Climato-stratigraphical unit Deposit Climate Sea level relative to OD
Present–10 000 Flandrian† (Holocene) Alluvial silt, sand and gravel, peat; (marine and estuarine deposits) Rapid climatic amelioration Lomond Stadial. Climate possibly warmer than that of present day by 9.6 ka BP Withdrawal of sea to present coastline began about 5 ka BP.
10 000–11 000

Devensian

Late Devensian

 Loch Lomond Stadial* Solifluction deposits; peat; freshwaterpond deposits Arctic Relative sea level below OD
11 000–c.13 000 Windemere Interstadial* Peat and fluvial deposits; raised beach deposits Warm climate 13–12.5 ka BP. Climatic deterioration from 12.5–12 ka BP Early eustatic sea level

rise. Raised beach

deposits 10–25 m

above OD

Rising sea level
13 000–c.26 000 Dimlington Stadial Till; morainic deposits; glaciofluvial silt, sand and gravel Arctic Relative sea level below OD

* the informal term 'Late-glacial' covering the period of the Windermere Interstadial and Loch Lomond Stadial has been long established in Scottish Quaternary literature (see Gray and Lowe, 1977) and has also been applied to raised beach deposits on Geological Survey 1:50 000 Drift Edition maps of the district (IGS, 1981, 1982). † referred to on Geological Survey 1:50 000 Drift Edition maps of the district as 'Recent'; includes 'Post-glacial Raised Beach deposits' (IGS, 1981, 1982)