Stephenson, D., Loughlin, S.C., Millward, D., Waters, C.N. & Williamson, I.T. 2003. Carboniferous and Permian Igneous Rocks of Great Britain North of the Variscan Front. Geological Conservation Review Series, No. 27, JNCC, Peterborough, ISBN 1 86107 497 2.
The original source material for these web pages has been made available by the JNCC under the Open Government Licence 3.0. Full details in the JNCC Open Data Policy
D. Stephenson
Volcanism continued intermittently throughout Dinantian time in the Midland Valley of Scotland, but as the sedimentary basins developed and the palaeogeography changed from fluvial plains with lakes to deltas, estuaries and shallow seas, so the nature of the volcanism changed. The vast subaerial lava plateaux and linear eruptive centres that dominated earlier Dinantian times were replaced in later Dinantian and Namurian times by more localized, short-lived, central vent complexes, characterized by more explosive, phreato-magmatic eruptions. Pyroclastic deposits are a major product of these eruptions and the lavas are almost exclusively basaltic (of micro-porphyritic 'Dalmeny' and 'Hillhouse' types); many are silica-undersaturated alkali olivine basalts and basanites. Basaltic hawaiites are rare, and the more fractionated rocks that are a feature of many early Dinantian successions, are absent.
These volcanic conditions were already widely established by the beginning of Namurian time and continued intermittently until at least mid-Westphalian times in the eastern Midland Valley, though at an overall less productive level than in Dinantian times. In the western Midland Valley, volcanism ceased following a major outpouring of lavas in late Namurian to earliest Westphalian times. Since there are no strata of Stephanian age preserved in Scotland we cannot be certain if there was any volcanism at this time. Large volumes of tholeiitic magma were intruded as dykes and sill-complexes during early Stephanian times (see Chapter 6) but, according to the model of Francis (1982), the magma was under insufficient pressure to reach the surface and there is no evidence of it having done so. Volcanism did resume in latest Stephanian or very early Permian times, when conditions had once again become continental, with localized outpourings of alkali basalt and basanite lavas in western and south-western areas and possibly also in the eastern Midland Valley where some sub-volcanic plugs and necks have late Stephanian to Early Permian radiometric ages.
Most of the Silesian volcanic rocks are interbedded with well-established, commonly fossiliferous, shallow-water sedimentary successions, and it is relatively easy to trace their strati-graphical development
Lavas dominate volcanic successions in the early Namurian rocks of the Bathgate Hills, the late Namurian rocks of north Ayrshire and the Early Permian rocks of south-west Scotland
The most continuous volcanic activity from late Dinantian through to mid Namurian times occurred in the area around the Bathgate Hills (Peach et al., 1910; Cadell, 1925; Macgregor and Haldane, 1933; Smith et al., 1994; Stephenson in Cameron et al., 1998). Here, a 600 m-thick succession of basaltic pyroclastic rocks and lavas, the Bathgate Hills Volcanic Formation, extends from the upper part of the Hopetoun Member, West Lothian Oil-shale Formation, to just above the Castlecary Limestone at the base of the Passage Formation. Unfortunately, there are no reliable radiometric dates for this prolonged sequence, which spans the Dinantian–Silesian boundary; some precise Ar-Ar determinations would have clear international value. This important and extensive sequence is not represented at present by a GCR site.
Borehole and mining information to the west of the outcrop suggests that the volcanic deposits are restricted to a sub-circular area, 20–25 km in diameter, that coincides with spectacular positive gravity and magnetic anomalies. The source of these anomalies has been the subject of much debate, as the combined thickness of volcanic rocks (Bathgate Hills and earlier volcanic formations) is insufficient to generate the observed anomalies. Current interpretations favour up to 1 km of volcanic rocks, intruded in their lower part by a large basic mass extending to a depth of about 8 km, and all sited upon a WNW-trending structural high that may have acted as a focus for the igneous activity (Rollin in Cameron et al., 1998). Several necks and plugs that cut older strata to the east of the main outcrop of volcanic rocks give some idea of the former eastward extent of the volcanic field, prior to erosion.
The basal part of the volcanic sequence, up to the base of the Lower Limestone Formation, is predominantly of pyroclastic rocks, except for some more persistent lavas in the Riccarton–Longmuir area. Sporadic graded bedding and load casts suggest subaqueous deposition in places, but numerous seatclays indicate frequent emergence. Layers of chert and spherulitic carbonate within the freshwater East Kirkton Limestone, famous for its terrestrial vertebrate, arthropod and plant fossils (Rolfe et al., 1994; Dineley and Metcalfe, 1999) , have been interpreted as evidence of hot spring activity associated with the volcanism (McGill et al., 1990, 1994). From the top of the Lower Limestone Formation upwards, lavas become dominant at outcrop, although in boreholes to the west, thick, proximal pyroclastic deposits suggest the presence of a long-lived volcanic centre or centres in this area. The petrography of the lavas has been described by Falconer (1906) and Flett (in Peach et al., 1910). They are all alkali olivine basalts or basanites, and are microporphyritic with phenocrysts of olivine and variable amounts of augite ('Dalmeny' and 'Hillhouse' type). Kaolinized or reddened flow tops indicate subaerial eruption.
The overall picture of the Bathgate Hills area is of a low-lying, heavily vegetated coastal plain in late Dinantian times, giving way periodically to shallow marine conditions in Lower Limestone and Upper Limestone formation times. Volcanoes accumulated above sea level to form islands surrounded by coastal plains, restricted lagoons and a variety of carbonate reefs, all neatly modelled by Jameson (1987) (see Petershill Quarry GCR site report in the British Lower Carboniferous Stratigraphy GCR volume — Cossey et al., in prep). Carbonaceous mudstones and argillaceous limestones, interpreted as having formed in back-reef lagoons, contain synsedimentary Pb-Zn mineralization related to the volcanism (Stephenson, 1983).
The Saline Hills of western Fife are broadly along strike to the NNE of the Bathgate Hills along the Bo'ness High, a zone of relatively low subsidence that forms the eastern margin of the Silesian Kincardine Basin (Read, 1988). Here, thick tuffs, volcaniclastic sedimentary rocks and rare thin basalt flows occur in the Limestone Coal Formation and parts of the Upper Limestone Formation (Francis, 1961a). The succession is cut by several necks and plugs. Within the Kincardine Basin, boreholes and mine workings have revealed distal tuffs at various levels in the Limestone Coal Formation and Upper Limestone Formation that may correlate with the volcanism in the Bathgate Hills to the south (Francis et al., 1961). These are mostly on the eastern side of the basin, on the flank of the Bo'ness High, where several necks have also been recognized (Francis, 1957, 1959; Barnett, 1985). However, a borehole at Tillicoultry in the centre of the basin revealed proximal agglomerates and some lavas in the top part of the Upper Limestone Formation. Rippon et al. (1996) have suggested that these were related to the NNE-trending Coalsnaughton Fault which may have been an active extensional fault parallel to the Bo'ness Line.
In the west of the Midland Valley, basic pyroclastic rocks are interbedded with sedimentary strata throughout the Limestone Coal Formation and to a lesser extent the Upper Limestone Formation to the west of Dalry; they are known mainly from borehole records (Richey et al., 1930). All may have been derived from necks that cut older strata to the west.
A major episode of volcanism is represented in the upper part of the Passage Formation by the Troon Volcanic Member. This is recognizable over an area that extends from Ayrshire south to Stranraer and west to Arran, Kintyre and possibly to Northern Ireland (Richey et al., 1930, fig. 25), but the main development occurs beneath the Coal Measures of the Ayrshire Basin (Monro, 1999). Outcrops of lava occur on the northern and southern flanks of the basin and the thickest development of over 160 m is just to the north of the town of Troon
The Troon Volcanic Member is composed almost entirely of subaerial basaltic lavas (olivine-microphyric, 'Dalmeny' type) with some interbedded sedimentary rocks. Petrographical details were given by MacGregor (in Richey et al., 1930; in Eyles et al., 1949). Like the earlier, Dinantian lavas, these are transitional in nature and range from hypersthene- to nepheline-normative; a few are basanites (Macdonald et al., 1977; Wallis, 1989).
The lavas are generally decomposed, commonly with a characteristic red speckled appearance due to sideritic alteration. More advanced decomposition produces pseudostratified greenish-blue clays, thought to be the result of penecontemporaneous subaerial weathering. This weathering is particularly well developed at the top of the member, and on the northern side of the Ayrshire Basin it grades upwards into aluminous clayrocks interbedded with laminated mudstones, seatclays and coals that together comprise the Ayrshire Bauxitic Clay Member (Wilson, 1922; Eyles et al., 1949; Monro, 1999) (see High Smithstone Quarry GCR site report in the British Upper Carboniferous Stratigraphy GCR volume — Cleal and Thomas, 1996). This complex member is up to 20 m thick in places, but is more typically 2–4 m thick and has been extracted as a source of alum and specialist refractory clay. It contains some of the highest quality fireclay in Britain, with up to 42% Al2O3. Whilst the basal parts were undoubtedly formed by the weathering of basalt in situ, there has been much debate about additional sedimentary and diagenetic processes that may have operated (Monro et al., 1983). Bauxitic laterites only form in tropical climates and require a significant time to build up any thickness. A variety of clastic clayrocks demonstrate the local reworking of the weathered crust and, in the absence of any other sediment input, we can deduce that the topography of the underlying lava surface was low. Coals, seatrocks and plant remains, including tree trunks, are abundant and, together with the presence of at least one marine band, suggest that the area was reduced to a heavily vegetated, flat, low-lying coastal plain when the volcanism ended at the close of Namurian time.
To the west of the main outcrop, on the Isle of Arran, a thin succession of red tuff and basaltic lava overlain by bauxitic clays and Middle Coal Measures occurs in the Merkland Burn, near Brodick Castle. Tuffs with thin lavas occur in two outcrops to the west and WSW of Lamlash: in the Benlister Burn, bauxitic clays are found and in the larger outcrop at the head of the Sliddery Water, tuffs are overlain by red mudstones with mussels of Middle Coal Measures type (Leitch, 1942). Both occurrences have been assigned to the Passage Formation. Farther west, in Kintyre, about 100 m of thick basaltic lavas with thin tuffs and red lateritic mudstones are known only from boreholes at Machrihanish. The highest lavas have been weathered to bauxitic clay and there is a marked non-sequence beneath the overlying Lower Coal Measures. In Galloway, a single thin flow of basaltic lava near Kirkcolm, on the west side of Loch Ryan, occurs within a sequence of mottled sandstones, mudstones and seatclays that is overlain unconformably by Permian breccias.
Evidence of volcanicity in the form of interbedded tuffs and tuffaceous sedimentary rocks is widespread in the Silesian successions of Fife, south of the East Ochil Fault. Basaltic lavas are rare. Much of the information comes from underground coal mining and boreholes, both onshore and offshore, though there are some good exposures, particularly in coast sections. The area is renowned for its numerous, mainly small necks, many with plugs of alkali olivine basalt or basanite. These are particularly abundant in East Fife where over 100 are recorded, and are represented by the East Fife Coast GCR site. Here, the continuing influence of deep-seated Caledonian structures is well illustrated by the NE-trending Ardross Fault, along which ten or more necks are sited over a distance of 4 km (Francis and Hopgood, 1970). On the opposite side of the Firth of Forth, Howells (1969) recorded high-level vent structures near Longniddry that cut Namurian strata and lie directly on the projected north-eastwards continuation of the Southern Upland Fault, which may have controlled the rise of magma during Visean time (see Chapter 2). The East Fife necks are also renowned as some of the most productive sources in Scotland of megacrysts and/or rock clasts of deep-seated igneous material or metamorphic basement (e.g. Colvine, 1968; Chapman, 1974, 1976; Macintyre et al., 1981; Donaldson, 1984) (see Chapter 1).
Onshore outcrops of interbedded volcanic rocks are concentrated mainly in two areas. Around the former Westfield opencast coal site in central Fife, there are five flows of basaltic pillow lava with associated tuffs and hyaloclastites in the top of the Upper Limestone Formation and the basal Passage Formation. The large complex necks of Largo Law and Rires in East Fife are surrounded by bedded tuffs and a few lavas that seem to be interbedded with the Upper Limestone Formation and Passage Formation, though in many places it is difficult to separate them from pyroclastic rocks within the necks (Forsyth and Chisholm, 1977). Offshore, boreholes in the Firth of Forth have proved pyroclastic rocks and rare basalt lavas, ranging from the topmost Passage Formation up into the Middle Coal Measures.
The necks cut almost the full local age range of Dinantian and Namurian strata and it is likely that some are the source of the Namurian bedded tuffs. A few necks cut Coal Measures strata. The Viewforth Neck contains foundered blocks of tuffaceous sedimentary rocks that have yielded Langsettian spores, and the Lundin Links Neck cuts (Duckmantian) Middle Coal Measures. Some necks are thus of Duckmantian age or younger and it has been suggested that blocks of quartz-bearing dolerite in several necks (e.g. Ardross, St Monance, Viewforth, Lundin Links) are derived from the early Stephanian tholeiitic intrusive suite, though this has not been proved. The most reliable K-Ar radiometric dates from four of the necks are within a range of 295–288 Ma, which dates their time of emplacement as close to the Stephanian–Permian boundary (Wallis, 1989). The highly silica-undersaturated nature of plugs and other minor intrusions (basanites, olivine nephelinites and olivine analcimites) within many of the vents is characteristic of the latest magmatism elsewhere in the Carboniferous–Permian Igneous Province of northern Britain, especially the undoubted Early Permian lavas and associated necks of south Ayrshire (see below), and hence is compatible with the age estimates.
The morphology and physical volcanology of the Fife volcanoes and their relationships to surrounding contemporaneous sedimentation have been studied more intensively than in any other part of the Carboniferous–Permian Igneous Province of northern Britain. Much of this work has been by E.H. Francis, building upon earlier Geological Survey work and making extensive use of boreholes and mine sections (Francis, 1960, 1961b,c, 1968a,b, 1970b; Francis et al., 1961; Francis and Ewing, 1961; Francis and Hopgood, 1970; Francis in Forsyth and Chisholm, 1977).
The bedded tuffs consist of a mixture of basaltic and comminuted sedimentary debris and are often well graded, indicative of ash fall into shallow water. Thin distal representatives have been traced for up to 30 km from their implied vents and have correlation value as effective time-planes (Francis, 1961c, 1968a). Many of these thin tuffs have been altered diagenetically to kaolin, especially in, or near to, coal seams, and have been likened to the 'tonsteins' of north-west European coalfields (Francis et al., 1961; Francis, 1969, 1985). Most of the necks appear to be funnel-shaped tuff-pipes filled by varying proportions of basaltic material and country rock. Initial updoming, with associated radial and concentric fracturing, was probably followed by gas-fluxioning and wall-rock stoping, prior to the rise of basaltic magma that interacted with groundwater and surface water in explosive, phreatomagmatic eruptions. Many of the larger necks contain inward-dipping bedded pyroclastic rocks and sediments, with fragments of fossil wood that show that they accumulated at the surface. The inward dips were generated by inward collapse of the areas surrounding the initial vent during the eruption, possibly along ring-fractures, and by immediate post eruptive subsidence (Francis, 1970b). Such features, together with the interpretation of cross-bedding and large-scale slumping as the results of base-surge eruptions, led Francis (in Forsyth and Chisholm, 1977) to compare the volcanoes with modern Surtseyan ash-rings of wide diameter and low height, which are typical of basaltic eruptions into shallow water. Magmas failing to reach the surface were emplaced as a variety of minor intrusions. Details, as illustrated by individual necks, are given in the East Fife Coast GCR site report.
The volcanic rocks interbedded with the sedimentary succession of East Lothian are undoubtedly of Dinantian age, as are most of the closely related vents (see North Berwick Coast GCR site report). However, some of the associated intrusions are basanitic or foiditic and it seems reasonable to suggest that these may represent a south-eastern continuation of the East Fife late Stephanian to Early Permian volcanic field, from which they are separated by only 15 km. Available K-Ar whole-rock dates are equivocal, but suggest minimum ages in the range 295–229 Ma (Snelling and Chan in McAdam and Tulloch, 1985, recalculated; Wallis, 1989). Notable examples of these intrusions occur at Oldhamstocks, Kidlaw, Limplum, Gin Head, Yellow Man, Yellow Craig Plantation and North Berwick Abbey, and it is probable that some of the breccia-filled necks in the area belong to this late phase of activity.
A large plug of nepheline basanite at Southdean, south of Jedburgh, is anomalous among the Dinantian plugs of south-east Scotland (see Chapter 3) and may be an isolated Early Permian occurrence.
Bedded volcanic rocks of Early Permian age crop out in south Ayrshire, in an elliptical outcrop around the overlying aeolian sandstones of the Mauchline Basin, and in the NNW- to NW-trending, fault-controlled Sanquhar and Thornhill basins, within the Southern Uplands. Together these mid-Carboniferous-Permian basins define a broad NW-trending lineament that continues to the south-east through the Permian sedimentary basins at Dumfries, Lochmaben and the Vale of Eden. Contemporaneous necks and sub-volcanic intrusions within and around the Mauchline lavas and in the Sanquhar Basin extend the known limits of the volcanic fields, but it is not known if they were ever interconnected.
The Mauchline Volcanic Formation, represented by the Howford Bridge GCR site, is up to 238 m thick and rests unconformably but with no marked discordance upon the 'Barren Red Beds' of the Upper Coal Measures. Plant debris found near the base of the volcanic sequence suggests an earliest Permian age (Wagner, 1983) and K-Ar whole-rock dates around 286 ± 7 Ma are Early Permian (De Souza, 1982). Palaeo-magnetic measurements have also indicated pole positions close to those of Carboniferous-Permian boundary time (Du Bois, 1957; Harcombe-Smee et al., 1996). The lavas are predominantly microporphyritic olivine basalts ('Dalmeny' type), but basanites are common and some strongly silica-undersaturated olivine nephelinites are present. However, analyses include some hypersthene-normative transitional basalts (Macdonald et al., 1977; Wallis, 1989). Pyroclastic rocks comprise a large part of the succession, becoming more abundant in the thicker, eastern parts. Sedimentary rocks within the volcanic sequence contain wind-rounded sand grains, indicating that the lavas were erupted in predominantly desert conditions, but fluvial sandstones and mudstones imply spasmodic sheet-floods and ephemeral lakes.
Over 60 necks are known, mostly within a 20–30 km radius of the centre of the Mauchline Basin, but also extending to West Kilbride in the north (Alexander et al., 1986), Muirkirk in the east and Dalmellington in the south
In the Thornhill Basin, the Carron Basalt Formation, represented by the Carron Water GCR site, is up to 50 m thick and rests unconformably on reddened Coal Measures and Lower Palaeozoic rocks. It comprises subaerial olivine-microphyric basalts and basanites similar to those of the Mauchline Basin and interbedded sedimentary units. Conglomeratic sandstones and breccias below the lavas, and beyond the present lava outcrop, contain angular fragments of basalt indicating earlier, possibly more extensive, flows. Although some basalts incorporate wind-blown sand in their matrix, the sedimentary rocks interbedded with and overlying the volcanic rocks are predominantly fluvial. Aeolian desert conditions did not become fully established here until well after the volcanic period. Small outliers of olivine basalt rest upon Middle Coal Measures at the southeast end of the Sanquhar Basin and five small necks pierce the Coal Measures nearby.
Across the North Channel, in a borehole at Larne in Northern Ireland, over 600 m of basic volcanic rocks were proved at a depth of over 2000 m, beneath an undoubted Permo-Triassic succession (Penn et al., 1983). A K-Ar whole- rock date of 245 ± 13 Ma is almost certainly too young and the authors suggested an Early Permian age. However, there are marked petrological differences between these lavas (which have possible tholeiitic affinities) and those of south-west Scotland. They occupy a separate sedimentary basin and are certainly a separate lava field.
In the centre of the East Irish Sea Basin, a borehole has penetrated 45.5 m of altered basalts overlain by 22 m of volcaniclastic rocks at the base of the Permian succession (Jackson et al., 1995, 1997). These Tormentil Volcanics occur on a high that trends north-east towards south Cumbria, where clasts of olivine-dolerite and vesicular basalt are abundant in basal Permian conglomerates in the Humphrey Head Borehole (Adams and Wadsworth, 1993). The source of these clasts is not known, but they are likely to be Carboniferous to Early Permian in age and relatively local. Thin subaerial basaltic flows are also recorded from the Lower Permian rocks of the North Sea Basin (Dixon et al., 1981).
On the Isle of Arran, the sequence at the head of the Sliddery Water that includes possible Passage Formation volcanic rocks (see above), passes up into gritty feldspathic sandstones and slaggy basaltic lavas with thin tuffs that have been assigned to the base of the Permian succession on the current British Geological Survey 1:50 000 map (1987).
A 120 m-thick Permian succession is exposed on the small island of Glas Eilean in the Sound of Islay, between the Isles of Islay and Jura (Pringle and Bailey, 1944; Upton et al., 1987)
Evidence of more widespread Early Permian volcanism in the Highlands and Islands comes from the distribution of small sub-volcanic necks