South west Scotland introduction to geology

By P. Stone

The rolling hills of south-west Scotland are underlain principally by Lower Palaeozoic clastic sedimentary rocks and Caledonian granitic plutons. Drainage in the region is controlled in the main by Permo-Carboniferous faults trending SSE, which define the margins of half-graben sedimentary basins. The Upper Palaeozoic basin infills have been preferentially eroded so that they now coincide with the main river valleys. Rivers such as the Cree, Ken and Nith flow generally southwards to the Solway Firth, the site of another major Upper Palaeozoic basin. The Lower Palaeozoic rocks are exposed on the west coast of the Rhins of Galloway and along much of the Solway coast westward from about Dalbeattie. The lower-lying part of the region, mainly in the south along the Solway coast, carries an extensive cover of glacial deposits, with drumlins particularly well developed between Wigtown and Glenluce.

Lower Palaeozoic regional geology

The geology of south-west Scotland testifies to processes active at the margins of a major ocean some 400 to 500 million years ago. This precursor to the Atlantic has become known as the Iapetus Ocean and its vestiges can be traced from the American Appalachians, through the maritime provinces of Canada, across Ireland and Scotland and into Scandinavia. The Iapetus Ocean probably spread to its maximum size during the Cambrian Period, and the Ordovician and Silurian rocks of the Scottish Southern Uplands record its later evolution and eventual destruction by oblique collision between the bordering continental plates. The one to the north is referred to as Laurentia, that to the south as Avalonia.

The oldest rocks seen form the early Arenig ophiolitic assemblage of the Ballantrae Complex, cropping out over about 75 km² on the Ayrshire coast south of Girvan (Figure 1). In broad terms ophiolites represent relics of the oceanic crust and mantle; their most complete development would show an upward sequence from ultramafic rock to gabbro, both layered and homogeneous, then through a dyke complex feeding spilitic pillow lavas with the whole pile topped by abyssal sedimentary rocks, typically black shale and chert. The Ballantrae example is fragmentary and only the ultra-mafic rocks and pillow lavas are well preserved. However, sufficient evidence can be gathered from these, particularly in respect of their geochemistry, to show that the Ballantrae ophiolite was not generated at a mid-ocean spreading centre, but instead is a tectonic mixture of supra-subduction zone island arc and oceanic island components (Thirlwall and Bluck, 1984; Smellie and Stone, 1992; Stone and Smellie, 1988 and references therein). The oldest volcanic components formed in an oceanic island arc above a subduction zone about 500 million years ago, but the oceanic crust forming the foundations of that arc might have been 70 or 80 million years older. Volcanic arc activity continued through the Arenig epoch, changing in character subtly as the arc evolved. Volcanic activity of different character occurred simultaneously, probably in the back arc region. This phase was brought to a close towards the end of the Arenig, about 478 million years ago, by collision of the arc with the Laurentian continental landmass. The volcanic arc, with its oceanic crust and mantle foundations, was structurally imbricated and emplaced on the continental margin, a process known as obduction. The subsequent foundering of the continental margin allowed a marine sedimentary sequence to be deposited above the deformed ophiolite. Conglomerates and shallow-marine limestone commonly form the base of the succession but sedimentary facies becomes more deep-water upwards as marine conditions transgressed northwards. The succession is almost complete from the Llanvirn to the lower Wenlock, with only a small break at the Ordovician–Silurian boundary.

The continuity of sedimentary sequence found above the obducted Ballantrae ophiolite is in stark contrast to the fragmented situation seen farther south in the Southern Uplands, where the Lower Palaeozoic rocks are divided by major strike faults into a series of blocks, each with only a limited strati-graphical range. These successions are dominated by turbidite deposits of greywacke, siltstone and shale. Graptolites, found most commonly in black shales, are important in establishing the ages of these, and prove that an upward lithological change from black shale to greywacke, present in each of the fault-bounded blocks, is diachronous, becoming sequentially younger southwards. Graptolite evidence for the age of the overlying greywacke formations is much more limited but generally proves the greywacke to be the same age or only slightly younger than the youngest shale. This emphasises the differences in sedimentation rates; within a single biozone (lasting very roughly a million years) a few metres of black shale may have been deposited but in the same time interval repeated turbidity current flows may have built up several thousand metres of greywacke. The mechanism of turbidity current action, and details of graptolites and their biostratigraphy, are discussed more fully below.

Throughout the Southern Uplands the greywacke turbidite beds are for the most part steeply inclined with a fairly uniform strike of about 060°. From the assemblage of sedimentary features it is clear that most of the succession youngs towards the NW, though many beds are slightly overturned to dip steeply SE. Folded zones consist of tight anticline—syncline pairs which do not interfere with the overall trend of younging. The dominance of NW-directed younging is at odds with the overall age relationships across the Southern Uplands; a glance at any geological map of the area shows that the oldest rocks crop out in the NW and the youngest in the SE. This paradox has been created by the major strike faults, all of which downthrow to the SE (Figure 1). In each fault-bounded block the oldest rocks lie at the SE side and youngest at the NW, but progressively older strata form the base of each block sequentially across the Southern Uplands from SE to NW (Figure 2) to produce the overall stratigraphical pattern. In the northernmost structural blocks spilitic lavas, chert and black shale form the base of the sequence, overlain by greywacke; in the more southerly blocks only shale is seen below the greywacke. From a comparison with the equivalent sequence now preserved in Newfoundland it seems likely that the volcanic component was originally much more extensive (Colman-Sadd et al., 1992) and represents arc and back arc material formed within the Iapetus Ocean.

The structural and stratigraphical relationships seen in the Southern Uplands have been explained by the 'accretionary prism' model formulated by Leggett et al. in 1979 and refined by Leggett (1987). This envisaged the greywackes being deposited in a deep ocean trench at an active continental margin of the Iapetus Ocean where the oceanic plate was subducting north-westwards. As the oceanic plate descended, its covering of sediment was scraped off and stacked at the continental margin in a series of underthrust slices. Since greywackes would only be deposited above the hemipelagic (black shale) sediments covering the oceanic plate as it approached the trench, the age of the greywacke in each thrust slice is younger than that in the pre-viously accreted slice. Final rotation of the thrust stack to the vertical was caused by continental collision as the ocean closed completely in the late Silurian or early Devonian.

The accretionary prism model elegantly explains the overall geological relationships within the Southern Uplands but many details are still difficult to reconcile. In particular, a close examination of the compositions of individual sand grains within the greywackes shows marked differences between adjacent beds in the same structural block and between the blocks themselves. These differences are particularly marked in the Ordovician sequence and form the basis of the lithostratigraphical divisions shown in (Figure 1) and (Figure 2). Two distinct provenances are indicated and palaeocurrent analyses of sole marks suggests a mature continental margin to the north and NE and an active volcanic island arc to the south and SW. The palaeogeographical indications have led ro an alternative interpretation, namely that the early history of the Southern Uplands was within a back arc basin which Stone et al. (1987) propose developed into a sequential back arc to foreland basin thrust system following collision between continent and arc in the early Silurian. Whichever model is preferred the thrust geometry is fundamentally the same and has one important rider; the thrust-related deformation was diachronous, with structures in the NW formed before those in the SE. This process has been partly quantified by Barnes et al. (1989) and its understanding remains dependent on graptolite biostratigraphy.

Within the Ordovician Leadhills Group (Figure 2) the compositional variation in the greywackes, when viewed in the light of the graptolite biostratigraphy, allows a stratigraphical division into defined formations. The same is possible in the mid-Silurian Hawick Group where the presence or absence of interbedded red mudstone provides a key lithostratigraphical indicator. However, the early Silurian Gala Group is more compositionally uniform and the units shown in (Figure 2) are arranged on the basis of biostratigraphy and deduced structural position. These units are regarded as tectonostratigraphical blocks or tracts and are numbered, from north to south, Gala 1 to Gala 9. Some formation names have been applied locally and are used informally.

Various fold geometries will be seen during the excursions, and most can be linked into a general picture of southward-propagating thrust development followed by superimposed sinistral shear. The thrust-related structures can be grouped into three styles which, at outcrop, form distinct structural domains:

Type 1 Uniform, usually steeply inclined or vertical bedding younging consistently NW. The regular pattern of strike and dip is only locally interrupted by sporadic fold pairs.

Type 2 Continuous sequences of small-to medium-scale folds separated by minor shears and/or narrow unfolded units of steeply inclined bedding younging NW. Most of the folds are tight or close withaxial planes steeply inclined and fold hinges which generally plunge gently; some folds are periclinal.

Type 3 Continuous sequences of close to open folds, ranging considerably in wavelength and amplitude. Axial planes are upright or steeply inclined and fold hinges plunge gently.

(Figure 3) shows how, in an idealised example, the three styles interrelate. Note that they may form at various depths within the thrust stack and therefore have no absolute implications in terms of structural depth. There are also local reversals of the regional trend, for example at the southern end of the Rhins of Galloway, where younging is dominantly southwards and the sense of thrust movement is towards the north (McCurry and Anderson, 1989).

Overall the regional (burial-related) metamorphic grade throughout the Lower Palaeozoic sequence of south-west Scotland is fairly low. The prehnite—pumpellyite facies is The maximum developed with many parts of the succession still at diagenetic grade. The clear implication is that thrusting occurred at relatively shallow depths as a thin-skinned tectonic phenomenon. Details of the techniques used to assess metamorphic grade and a discussion of its regional variation are given below.

The thrust deformation was a diachronous process, becoming progressively younger southwards. The NW part of the Lower Palaeozoic outcrop was affected in the late Ordovician and the SE part in the mid-Silurian. The first phase of deformation at the thrust front was accompanied by refolding within the thrust hinterland to accommodate the progressive steepening of the thrust sheets. However, a different factor came into play in the mid-Silurian with the increased importance of sinistral shear. Prior to that time deformation seems to have been more or less orthogonal and early-formed fold hinges generally have only a gentle plunge to NE or SW and an axial planar cleavage. However, from the mid-Silurian onwards, steeply plunging folds with an S-shaped down-plunge profile were superimposed on the thrust hinterland to refold earlier structures and cleavage. Simultaneously the deformation at the thrust front, by then coincident with the SE part of the Southern Uplands, changed in character so that the first folds formed there have variable, locally steeply plunging hinges; cleavage formed at this stage may transect the folds markedly. This diachronous tectonic history, spanning much of the Silurian or more, is the local manifestation of the Caledonian Orogeny, the mountain-building episode which accompanied the destruction of the Iapetus Ocean.

A numbering system has been widely applied to the Southern Uplands fold sequence. The early, thrust-related deformation is deemed D1, accommodation structures in the thrust hinterland are grouped together as D2, and the late sinistral shear is identified as D3. When specifically discussing fold hinges or cleavage planes D may be replaced by F or S respectively. Thus the D1 deformation produced F1 fold hinges and an S1 slaty cleavage. This shorthand will be used in many of the excursion accounts.

During the final stages of deformation a widespread dyke swarm was intruded into the region. Felsites, porphyritic microdiorites and lamprophyres are all present and some have been foliated by the later tectonic episodes. Other dykes are entirely post-tectonic and have been radiometrically dated at about 395–418 Ma (Rock et al., 1986). The climax of igneous activity was reached as deformation ended with the intrusion of the major granitic plutons. These have all been dated radiometrically to within a few million years of 400 Ma (Halliday et al., 1980). Evidence pertaining to the timing of deformation and intrusion is summarised in (Figure 4).

Graptolite biostratigraphy of south-west Scotland: R.A. Hughes

Graptolites are a long extinct group of tiny colonial animals which formed the greater part of the Ordovician and Silurian oceanic plankton. They are the most useful fossil group in Southern Uplands biostratigraphy. Each graptolite colony was composed of many (over 1000 in some cases) polyp-like animals (zooids) which lived in linked cuplike structures (thecae). These collectively comprised a single skeletal structure, called a rhabdosome, made of a protein-like material. Graptolite rhabdosomes are normally a few centimetres in length, but range from almost microscopic to a metre or so in extreme cases; all are slender and delicate. During Ordovician and Silurian times the graptolites evolved an extraordinary variety of shapes (Figure 5), which may have been adaptations to differing feeding strategies or to life in various parts of the water column. When the graptolites are well preserved, this variety enables today's palaeontologists to distinguish different groups and species, essential to the practice of biostratigraphy.

Graptolites inhabited the waters of the ancient oceans, and many different species co-existed at any one time. After death the graptolites sank to the sea bed. Graptolites today can be extremely abundant in 'hemipelagite' rock sequences, such as the Moffat Shale, formed mainly of very fine-grained sediment. Some Moffat Shale bedding planes are covered with graptolite remains. In marked contrast, graptolites are generally rare within the thick greywacke sequences, where hemipelagic sedimentation was mostly overwhelmed by coarse-grained sediment from continental sources, transported and deposited by frequent turbidity currents.

The different graptolite species on a Moffat Shale bedding plane constitute a sample of the graprolite population which existed at the time of deposition, and collectively are called an assemblage. Distinct assemblages of graptolite species are unique to individual graptolite zones, each of which represents a small slice of geological time. By calibrating the sequence of Silurian graptolite zones against the radiometric time-scale for the Silurian, we know for example that the duration of some of the Silurian graptolite zones was considerably less than one million years. The graptolites were a rapidly evolving group, especially in the Silurian, and different assemblages of species representing different zones are therefore indicative of different geological time periods.

In the latter half of the last century Charles Lapworth, working in the Moffat-dale area of the Southern Uplands (now the type area for the Moffat Shale Group) first recognised that graptolites occur in distinct assemblages, and established the sequence of assemblages and zones which record the evolution of the graptolites through geological time (Figure 5). Lapworth was the first person _to demonstrate the use of graptolites as a biostratigraphical tool, and his work has survived the test of time to the extent that his sequence of zones (Lapworth, 1878) remains almost entirely unchanged, and is today used throughout the world. His achievement was extraordinary, and without the tool of graptolite biostratigraphy the recognition of the gross structure of the Southern Uplands as an imbricate thrust stack would have been impossible. It is no exaggeration to state that graptolite biostratigraphy has been and will remain entirely fundamental to solving the great geological complexity of the Southern Uplands. Excursion 18 gives a selection of localities where they can be found. Further information on this fascinating group of animals is given by Palmer and Rickards (1991).

Turbidite sedimentology: M.C. Akhurst

Exposures of Lower Palaeozoic rocks in south-west Scotland are predominantly of interbedded greywacke, siltstone, mudstone, shale and, more rarely, conglomerate. The term greywacke refers to hard, poorly sorted sandstones containing grains of quartz, feldspar, dark ferromagnesian minerals and rock fragments set in a clay matrix which constitutes more than about 15 per cent of the rock. The sandstone beds may range from a few centimetres to more than 2 m thick. Bedding relationships are consistent:

the sandstone has a sharp base and grades up into siltstone and shale, reflecting their origin in a single depositional event. Sediment is transported downslope as a turbulent current. As the current slows, first coarse-grained then fine-grained sediment is deposited. The resulting graded bed is called a turbidite and shows a characteristic set of sedimentary structures forming a 'Bouma sequence' (Figure 6) (Bouma, 1962). Turbidity currents can flow very fast; one generated by an earthquake off eastern Canada in 1929 flowed at up to 20 metres/second (about 45 mph). Turbidity currents can erode the sea bed prior to deposition of the coarse-grained sediment. The erosive patterns, preserved as positive casts or 'sole structures' on the base of a turbidite bed, will then give a clear indication of way-up in the sedimentary sequence. Flute casts are distinctive sole structures, deepest and narrowest at the up-current end, which can be used to deduce the flow direction of the turbidity current. Linear groove casts are not quite so useful in this respect; they lie parallel to the current trend but do not show which way it flowed. Loading of sand into the underlying turbidite may also produce sole structures; sand at the base of the bed sinks into the underlying mud to form distinctive bulbous structures called load casts. Tapered 'flames' of mobile mud may penetrate up into the sand between the load casts.

Large volumes of sediment can be carried in suspension within the turbid flow. As current velocity wanes the coarsest sand and pebbles are deposited, followed by progressively finer-grained sand to form a graded bed (Figure 6). Deposition of fine sand and silt follows and is associated with the development firstly of planar lamination and then of ripple cross-lamination, reflecting changes in the flow regime of the current. Bedding in the fine sandstone and siltstone may also be contorted or convoluted where the rapidly deposited sediment was very wet and unstable. The ripple cross-laminated sediment (another indication of way up) grades into finely laminated or homogeneous siltstone and mudstone. These finest-grained sediments, which are darker in colour, accumulated much more slowly. The resulting mudstone or shale beds include both the finest sediment from the turbidity current and hemipelagic deep-sea sediment. The turbidites are predominantly unfossiliferous, although graptolites may be preserved in the finer-grained bed tops, perhaps only as broken pieces. However, if time is available between flows for a hemipelagic interval to develop, the resulting shale bed may carry a rich graptolite fauna.

Generation and flow of turbidity currents and subsequent deposition of turbidite beds, are sudden, catastrophic events. Turbidity currents can be initiated, for example, by collapse of unstable sediments on a submarine slope or by an earthquake shock. The resuspended sediment may be transferred for hundreds of kilometres on to the deep ocean floor. Currents are first funnelled down the continental slope via submarine canyons. Most sediment is then deposited as the current slows at the foot of the slope and spreads out from the confinement of the canyon. Accumulated deposits from many turbidity currents form a fan that has its apex at the canyon mouth (Figure 7). Submarine fans can be very large structures; modern examples are tens to hundreds of kilometres across.

The sediments deposited from a single turbidity current are not of uniform character along the length of the flow. Close to the source the turbidite bed will be dominated by coarse-grained graded sandstone, the lowest of the Bouma divisions (Figure 6), but far away from the source only fine-grained siltstone and mudstone of the higher divisions will occur. The complete Bouma sequence will only be deposited at intermediate positions. Studies of many ancient and modern submarine fan sequences have recognised inner, middle and outer fan divisions (Figure 7). The inner fan, nearest the sediment source, is crossed by a single channel within which turbidite deposition is mostly confined. Fine-grained sediment in suspension is deposited laterally to the channel. Inner fan deposits are, therefore, made up of coarse-grained sandstones and conglomerates forming lenses within a background of laminated mudstone. Down-current, in the middle fan setting, the channel divides and becomes much shallower. Turbidites dominated by the graded and planar-bedded divisions are deposited as more extensive, sheet-like beds. In the outer fan setting turbidity currents spread out over the fan surface to give laterally extensive beds with complete Bouma sequences. Finally, at locations farthest from the submarine canyon mouth, turbidite units may consist of only laminated siltstone and mudstone. Thus, by noting the style of bedding and sedimentary structures in a sequence of turbidites, it is possible to recognise where on a submarine fan they were deposited.

An extensive review of this remarkable group of sedimentary rocks is given by Pickering et al. (1989). Detailed sedimentological analyses are provided by Nilsen (1978) and Walker and Mutti (1973).

Metamorphism of the Lower Palaeozoic rocks: R.J. Merriman and B. Roberts

The development of the imbricate thrust system of the Southern Uplands was accompanied by regional low-grade metamorphism. Studies of this metamorphism have used the occurrence of hydrous Ca-Al silicate minerals in greywacke (Oliver and Leggett, 1980) and the white mica (illite) crystallinity of clay mineral assemblages in mudstone, shale and slate (Kemp et al., 1985) to assess the grade of metamorphism. These studies generally show that the prehnite—pumpellyite facies is widely developed in volcaniclastic greywacke and basic volcanic rocks; associated mudstones have white mica crystallinity indices typical of the anchizone. As deformation and metamorphic grade increase, mudstone and shale transform to slate, which commonly shows cleavage subparallel with bedding lamination. The grades detected suggest that across much of the Southern Uplands metamorphic temperatures did not exceed 300°C.

In the SW of the region, and in particular along the Rhins of Galloway, white mica crystallinity studies have been closely linked with the stratigraphy and structure (BGS, 1992a). X-ray diffraction (XRD) analysis was used to determine the Kubler index of white mica crystallinity and the mineralogy of the less than 2 pm fractions of mudstone samples (Merriman and Roberts, 1993). The Kubler index (KI in 0°20) measures very small changes in the profile of the 10A XRD peak which occur when authigenic clay micas recrystallise in response to advancing metamorphic grade. Metamorphism ranges from the zone of diagenesis (KI > 0.42) through the anchizone (KI 0.42–0.25) to the epizone (KI < 0.25). Metamorphic maps showing contours of equal crystallinity value (isocrysts) have been generated using Kubler indices (KI) from all mudstone samples and are shown as insets on the BGS solid geology maps of the region (e.g. BGS, 1992a, b, c). The metamorphic maps show that the isocrysts commonly trend subparallel to the traces of the strike-parallel faults. In many places the contours (which coincide with the diagenetic/ anchizone and anchizone/epizone boundaries) are discontinuities marking abrupt changes in grade across the tract-bounding faults. There is therefore a close relationship between the imbrication of the succession and the regional metamorphism. Isocrysts which cut across the tract-bounding faults mostly reflect contact overprinting of the regional metamorphism by igneous intrusions. Contact metamorphism typically reaches high-anchizone or epizonal grade and gives isocrysts broadly concentric with igneous outcrops, as seen around the Portencorkrie and Cairngarroch intrusions on the Rhins of Galloway. These concentric isocryst patterns are generally wider than aureoles delineated by recognisable hornfelsing. Such a pattern is also developed around the geophysically delineated but concealed intrusion at Sandhead (Kimbell and Stone, 1992). The concentric patterns developed around the Cairngarroch and Sandhead intrusions extend across the Orlock Bridge Fault, indicating that igneous emplacement occurred after the sinistral strike-slip fault movement proposed by Anderson and Oliver (1986). A similar relationship between metamorphism and fault movement is seen in the aureole of the Cairnsmore of Fleet granite (Merriman et al., 1991).

In relation to the tectonostratigraphy, the distribution of KI values shows distinctive trends of grade rising and falling from north to south through sequentially younger tracts of strata (Merriman and Roberts, 1993, (Figure 3); Stone, 1995 (Figure 32)). For example, on the Rhins of Galloway the oldest strata sampled from the Corsewall Formation (Figure 1) are mostly of late diagenetic grade. To the SE of the Glen App Fault the metamorphic grade increases abruptly into younger Leadhills Group strata so that most samples from the Kirkcolm, Portpatrick and Shinnel formations (Figure 1) are in the mid- or high-anchizone. South of the Orlock Bridge Fault in the Kirkcowan district (BGS, 1992b) the grade intially falls on crossing southwards into Gala 1, but then increases into the Gala 2 tract. Farther south, grade generally falls across sequentially younger tracts so that diagenetic mud-stones occur widely in the southern part of the group's outcrop around Port Logan. Grade generally increases abruptly in the Hawick Group (Figure 1) where both slates and interbedded sandstones have a penetrative cleavage.

The pattern of regional metamorphism found in south-west Scotland cannot easily be modelled in terms of a stratigraphical burial pattern whereby grade increases into older strata with increasing thickness of overburden (Roberts et al., 1991). Had the succession initially acquired such a pattern of burial metamorphism and subsequently been imbricated and rotated, older strata would still show higher grades than younger strata, whatever the final structure. The regional pattern in the Ordovician and parts of the Llandovery outcrop is the reverse of that generated by stratigraphical burial, in that grade may increase into younger tracts of strata. Such a pattern indicates that, from time to time, younger strata were buried beneath older strata, and in turn this suggests that the metamorphic pattern was generated by thrust-related tectonism. In the Ordovician outcrop the metamorphic pattern suggests that tracts comprising NW-younging Kirkcolm, Portpatrick and Shinnel formations were sequentially underthrust and buried beneath the older Corsewall Formation, which formed the upper unit of a thrust stack. It appears that a depth-related pattern of metamorphism was acquired after the strata were steepened, but imbrication continued to modify the pattern after burial. On the Rhins of Galloway, the widespread occurrence of diagenetic grade mudstones and shales in the SE-younging strata of the southernmost Gala Group outcrop is consistent with overthrusting to the upper part of the thrust stack (McCurry and Anderson, 1989). The outcrop of diagenetic zone strata has an abrupt southern termination at a significant metamorphic discontinuity across which grade and probable depth of burial increase sharply into the Hawick Group.

Upper Palaeozoic to Quaternary regional geology: A.A. McMillan and A.D. McAdam

Following the Caledonian Orogeny and intrusion of the Galloway granitic plutons, the Southern Upland region was subjected to uplift and erosion. Extension of the crust from late Devonian times led to the development of major sedimentary basins in the Midland Valley to the north and the Northumberland—Solway Trough to the south. Associated north- and NW-trending normal faults, which can be traced through the Southern Uplands, may have been active during early Carboniferous times, perhaps initiated by minor dextral move- ment on existing faults of Caledonoid (NE) trend (McMillan and Brand, 1995). Small half-graben basins defined by these faults as at Thornhill, Sanquhar and Loch Ryan (Figure 8), preserve a fragmentary record of fluviatile, estuarine and marginal marine sedimentation from Devonian to late Carboniferous times. A subsequent exten- sional event in early Permian times, gener- ated a short-lived volcanic episode as arid climatic conditions developed and exten- sive desert dune sandstones and associated breccia fans covered much of the region.

The north Solway coast, effectively the northern margin of the Solway Basin, provid es the best exposed reco rd of ea rly Carboniferous sedimentary rocks, together with remnants of late Devonian red beds (Table 1). The latter comprise conglomerates and arenites depos ited in fans and braided channels under hot, arid conditions. Calcareous palaeosols developed locally (Leeder, 1976). Following deposition of the Devonian strata, short-lived volcanic activity produced the lavas of the Birrenswark Volcanic Formation. The volcanism may be related to the ex tensional event which formed the Solway Basin (Leeder, 1982, 1988) .

The NE-trending Nonh Solway Fault which forms the northern boundary to the Solway Basin is probably a reactivated Caledonoid structure. Coarse clastic rocks of Dinantian age in the Rerrick (Wall Hill to Rascarrel Bay) and Colvend (Castlehill Point to Portling Bay) outliers provide evidence for periodic syn-sedimentary dip-slip movement on the fault (Deegan, 1973; Ord et al., 1988). Syn-sedimentary deformation of hanging-wall strata increases towards the fault. Cyclicity within alluvial fan deposits of the Rerrick Outlier may be attributed to the interplay between tectonic subsidence and changing sea level. Further away from the margin, Lower to Upper Border Group strata (Courceyan to Asbian, (Table 1)) of the Kirkbean-Southerness area (Craig, 1956) and at Langholm (Lumsden et al., 1967) also show evidence of changing sea level. Cyclically interbedded mudstones, siltstones and sandstones with thin cementstones and thin coals with seatearths reflect changing depositional environments in low-lying coastal areas. Locally, as at Powillimount on the Solway coast, there is convincing evidence in the Thirlstane Sandstone Member of synsedimentary seismic activity (Ord et al., 1988).

Marine inundations are also recorded at higher stratigraphical levels in the Lower and Upper Liddesdale groups (Asbian to Brigantian) at Kelhead near Annan, and at Langholm and Thornhill. At Thornhill, the Closeburn Limestone Formation records a marginal and probably quite short-lived marine incursion. These strata represent the earliest record of a breach through the Southern Uplands linking the Solway Basin to Sanquhar and the Midland Valley. Later Carboniferous sedimentation in the Southern Uplands was probably also restricted, particularly during Namurian times, but an attenuated Coal Measures sequence at Thornhill (Table 2) can be correlated with those at Canonbie and Sanquhar. At Thornhill, late-Carboniferous to Permo-Triassic oxidation has resulted in reddening of strata and replacement of coal. A fragmentary Upper Carboniferous sequence is also present at Loch Ryan. There has long been speculation on the extent of late Carboniferous to Permian sedimentation over the Southern Uplands. Pringle and Richey (1931) opined that it was once much more extensive and considered that the present distribution of strata could be attributed to post-Carboniferous downfaulting.

Crustal shortening and basin inversion took place in the Solway Basin during late Carboniferous times (Leeder and McMahon, 1988) resulting in the folding of Lower Carboniferous strata. The effects of deformation are particularly well seen on the Southerness coast where there are many examples of NE-orientated anticline—syncline fold pairs, structures probably controlled by inheritance of the Caledonian tectonic grain. Folding attributed to late Carboniferous inversion is also well displayed in the Liddel Water at Langholm.

Renewed extension activated N- and NW-trending normal faults during early Permian times to rejuvenate the Carboniferous basins of the Southern Uplands. At the same time the climate changed from tropical or semitropical to arid, and desert deposits of alluvial fans and aeolian dunes accumulated at Loch Ryan, Thornhill, Dumfries, Lochmaben and Moffat (Brookfield, 1978; 1980) (Table 2). In the Dumfries Basin, more than 1000 m of aeolian sandstones (Locharbriggs Sandstone Formation) and flash-flood breccias (Doweel Breccia Formation) conceal any remnant of Carboniferous strata. The extensive accumulation of breccia fan deposits on the western side of this basin may be attributed to synsedimentary activity on NW-trending faults, enhanced by relative uplift of the low-density crustal block of the Criffell granodiorite pluton (Bott and Masson-Smith, 1960). Evidence of a brief episode of volcanic activity associated with extension is seen at Loch Ryan, Thornhill and Lochmaben, as well as at Mauchline farther north in the Midland Valley. At Thornhill, this episode is represented by the Carron Basalt Formation which is succeeded by a series of fan breccias of the Durisdeer Formation and Locherben Formation, passing up into aeolian desert dune deposits of the Thornhill Sandstone Formation.

There are no strata preserved to record the geological history of the region between the Triassic and the Quaternary. The only rocks from this interval are Tertiary dykes, part of the swarms associated with igneous centres in Arran and Mull and thought to be about 55–60 million years old (Harrison et al., 1987; Hitchen and Ritchie, 1993).

During the last two million years Scotland has been repeatedly covered by ice sheets, which formed during glacial episodes and disappeared during periods of milder climate. Clear evidence is only preserved of the latest (Late Devensian) glaciation when ice accumulated on high ground in central Galloway and radiated out towards the lower ground. This merged with ice from the Highlands, which flowed down the Firth of Clyde and periodically encroached onto the Rhins of Galloway. The ice ground down the rocks, preferentially eroding softer lithologies, leaving striations on ice-smoothed rock surfaces, crag-and-tail features and oval drumlins as evidence of ice-flow direction. Particularly fine drumlin fields were developed in the south of Wigtownshire and Kirkcudbrightshire.

Glacial tills, the ground moraine of the ice sheet, blanket much of the lower ground. The colour and composition of the tills reflects the local rock type so that grey tills are common in the areas of greywacke outcrop whereas red sandy tills characterise the Permian basins. Marine shells are found in some of the tills deposited from Highland ice on the Rhins of Galloway. This suggests that the ice sheets scoured the sea floor of the Firth of Clyde.

Some 15 000 years ago the ice sheets began to melt. The higher hills became free of ice while glaciers still filled valleys. Meltwater from the ice cut channels, many now left as dry valleys, and deposited sand and gravel as eskers and kames and as raised beaches along the coast. Rivers have continued this process, depositing alluvial clay, silt, sand and gravel in their floodplains. However, the drainage pattern has been fundamentally influenced by the underlying geology. The main river valleys follow the trend of the Upper Palaeozoic basins, the softer rocks of which were preferentially eroded by the ice.

Carboniferous palaeontology: P.J. Brand

Carboniferous faunas (Figure 9) reflect the changing environments which affected south-west Scotland. Faunas in the Lower Border Group, for example, are of low diversity, the bivalve Modiolus latus being the dominant form. It appears to have flourished in restricted environments, possibly saline, such as lagoons which had only limited access to the sea and periodically dried up. The algal limestones which are a feature of the transition between the Lower and Middle Border groups are indicators of clear waters, no deeper than the photic (daylight) zone with limited access to marine conditions. Here the colonies could grow to large size. An exception to this restricted regime was the marine incursion represented by the Harden Beds of the Langholm area, with probable correlatives at Southerness and Orroland. At this horizon Syringothyris cuspidata, a distinctive spiriferoid brachiopod, and several species of productoid make their earliest appearance in the area. Current and wave action has disturbed and broken up the larger shells to some extent. Peculiar conditions prevailed during the deposition of the Glencartholm Volcanic Beds and these led to the preservation of a varied biota including numerous species of shrimps and fish. These comprised part of an offshore marine community periodically overwhelmed and buried by mud. Also present are rare pectinaceans and other marine bivalves a few of which appear to have been preserved in the bottom mud where they actually lived.

The faunas of the overlying Carlyle Beds at Langholm and in the corresponding beds at Southerness are much more varied, illustrating features associated with a habitat in soft bottom mud. Thus Megachonetes papilionaceus with its wide flat shell was able to lie on the mud without sinking, and the species of Prothyris were able to live partially buried in these conditions. Some horizons are corn-posed of materials which formed a harder substrate and colonies of Lithostrotion were able to establish themselves on these. However, current or wave action was sufficient to disturb the colonies from their growth positions, at least in the Southerness area.

The thick limestones of the late Dinantian Upper Liddesdale Group (as at Penton Linns) contain species of Gigantoproductus, which could have lived partially buried in the bottom sediment. Their thick shells, however, required an abundant carbonate supply, and this had a limiting effect on their distribution. Other brachiopods and the abundance of colonial corals, species of Siphonodendron, point to the availability of an abundant food supply, plenty of current movement and a low mud content in the water.

In the Namurian, conditions altered following the deposition of the Blae Pot Limestone and its correlatives resulting in the formation of a coal-bearing sandy sequence. In these beds rhynchonellids and Schellwienella, an orthotetoid, form the principal elements of the brachiopod fauna. They were presumably able to withstand increased current velocity and to utilise the different substrate. Marine incursions in the form of the Archerbeck Ochre bed and higher horizons contain examples of the brachiopod Latiproductus latissimus, the fine spines of which provided support on the calcareous muds which formed the sea bed where they lived.

In the Thornhill Basin, Coal Measures of Westphalian age occur. Mudstones and siltstones with varied nonmarine bivalve faunas form distinct horizons. Some of the valves of the shells lie oblique to the bedding of the strata, so may occupy positions of growth. The Westphalian 'mussels' may have lived in fresh water, though the actual salinity is unknown. What is clear, however, is that during the deposition of the beds containing these fossils there was no direct connection with the sea.

Few, if any, of the fossils that may be found are stratigraphically significant as individuals, but assemblages of forms can be used to provide a faunal framework for the subdivision of the Carboniferous rocks of the region.

References