Content and licensingview original scan buy a printed copy

Geology of the Eyemouth district: Memoir Geology of the Eyemouth district for 1:50 000 geological sheet 34 (Scotland)

By D. C. Greig

Greig, D. C. 1988. Geology of the Eyemouth district. Memoir British Geological Survey , Sheet 34 (Scotland), 78pp.

Geology of the Eyemouth district

Geology of the Eyemouth district for 1:50 000 geological sheet 34 (Scotland)

British Geological Survey

London: Her Majesty's Stationery Office 1988. © Crown copyright 1988. First published 1988. ISBN 0 11 884404 0. Printed in the United Kingdom for Her Majesty's Stationery Office Dd 0238934 4/88 C20 :398 12521

Other publications of the Survey dealing with this district and adjoining districts

Books

Maps

Preface

The Eyemouth district, covered by the Scottish one-inch to one-mile Geological Survey sheet 34, was first surveyed by A. Geikie, H. H. Howell and J. Geikie around 1860 and the one-inch map was published in 1863. Between 1962 and 1970 the district was resurveyed by D. C. Greig, I. B. Cameron, A. Davies, A. D. McAdam and T. E. Smith under R. A. Eden and G. I. Lumsden as district geologists. Lower Palaeozoic graptolites were identified by I. Strachan, and Carboniferous fossils by R. B. Wilson, while R. W. Elliot made a systematic petrographic study of the igneous rocks. J. D. Floyd assisted with sedimentary petrography and P. J. Brand with palaeontology. The section on groundwater was written by N. S. Robins. Some results have been included from off-shore geological work by Mr Eden and colleagues and members of the British Sub-Aqua Club. carried out at Burnmouth and St Abbs in 1967–68. Survey photographs (see Appendix 3) were taken by A. F. Christie and T. S. Bain. This memoir was written by Mr Greig, incorporating the contributions of his colleagues: and it was edited by D. H. Land.

The 1:50 000 geological map was published in 1982 (solid edition) and 1983 (drift edition). It is unfortunate that a variety of circumstances led to a long delay between completion of field work and publication of the maps and memoir. During this interval, plate tectonic theory with its implications has become accepted, while recent work by Casey and Oliver (1984) and Molyneux (in press) shows that the Coldingham and Linkim beds, previously regarded as older than the Llandoverian of the Lammermuir Hills, are of late Silurian age. This post-survey work has been taken into account in this memoir.

It is a pleasure to acknowledge our indebtedness to many professional colleagues for assistance and discussions at all stages of the project, and to landowners and occupiers for their helpful courtesy during the survey.

F G Larminie, OBE Director British Geological Survey Keyworth Nottingham NG12 5GG 15th January 1983

General notes

Chapter 1 Introduction

Geography

The district described in this memoir, covered by sheet 34 of the 1:50 000 geological map of Scotland, lies between Cockburnspath, Berwick-upon-Tweed and Duns (which is just outwith the south-west corner), and is within the county of Berwickshire (Figure 1).

It is a predominantly agricultural district, large areas of the lower ground being devoted to barley, potatoes and fodder crops, with cattle and sheep rearing important features of the economy. Eyemouth (population about 3500) is the only town and is traditionally a centre of the fishing industry. In recent years its economy has become more varied with the introduction of light industry. It is also a tourist centre, along with Coldingham, Burnmouth, St Abbs and Cockburnspath. Other large villages are Chirnside, Ayton and Reston. Coldingham was an important ecclesiastical centre from the foundation of its priory in the 11th century. Throughout history the district has been important as a corridor of communication between England and the Midland Valley of Scotland, traversed today by the main road and railway from London to Edinburgh, on the cross-border route which is least often blocked by winter weather.

The north-western half of the district is dominated by the eastern extremity of the Lammermuir Hills. Most of it is enclosed agricultural land grading up to marginal hill grazing and moorland in the higher parts, which nowhere exceed 270 m altitude. Moorland is most extensive between the railway and the coast. The terrain is plateau-like with rounded summits and grassy or heathery slopes, rock exposures being sparsely distributed. South-east of the Lammermuirs, the land is generally lower, more fertile and more densely populated. In the south-east of the district, Devonian lavas and Silurian greywackes form another thinly populated, largely pastoral, upland tract between Burnmouth and Mordington.

Mining and quarrying have never been large-scale operations. Sand and gravel, greywacke for road-building, and some sandstones for house-building have been widely quarried. Small abandoned workings are abundant within the outcrop of the greywackes and, less commonly, among the Lower Devonian lavas and tuffs. Larger quarries were operative until the early 1960's in greywacke at Old Cambus and Grantshouse, and sand was quarried at Kinnegar, south of Cockburnspath, but the only sand and gravel pit now working is at Causewaybank, 2 km north of Chirnside. Dolerite is quarried in the south-west corner of the district. Coal has been mined in a small way on the coast at Lamberton, but not since the early 19th century; and there was a trial for copper at Elba north of Preston.

Siccar Point [NT812 710], where the unconformable relations of Upper Old Red Sandstone on Silurian grevwackes are so clearly displayed, is of international significance for its importance in the early development of geology as a science.

James Hutton (1726–1797) who was one of the chief pioneers in this development, lived for 14 years Slighhouses [NT 822 594], and at Siccar Point (Frontispiece) found a splendidly clear section where he could convincingly demonstrate many of his ideas in the uniformitarian philosophy of geology (Playfair, 1805).

Physiography

Whereas the broad physiography of the district is determined by the solid rocks, details owe much to the effects of glaciation. Over most of the high ground solid rock is near the surface and the rounded topography is ascribed to glacial erosion of steeply dipping strata lacking groups of beds of different hardness. Drift deposits cover nearly all the lower ground. The granite of Stoneshiel Hill and the dolerite of Duns Law form upstanding features, and scarp and dip slopes are well developed among the lava-flows of St Abb's Head.

Striking features of the Lammermuir hills are the deeply cut narrow valleys of the Whiteadder and Eye waters. The latter, with its glacially modified former upper course north of Penmanshiel, traverses the uplands and affords an easy course for the main road and railway. Another gap to the south marks an abandoned link with the Whiteadder at Elba. To the north-east, for some 5 or 6 km inland from the coast, the upland terrain is dominated by parallel linear valleys and crag-and-tail features, produced by glacial erosion, later accentuated by meltwater flows. Channels formed by glacial meltwaters are readily recognisable over most of the district, varying from insignificant sinuous depressions to spectacular dry gorges such as those around Cockburnspath and Old Cambus.

Downstream from the uplands the valleys of the Whiteadder and the Eye widen significantly, and include extensive spreads of terraced gravels. In their lowest reaches the streams were all rejuvenated by the latest lowering of sea levels and have cut gorges in solid rock.

The soils of the district are predominantly stony sandy clays. Areas of heavy clay or of Stoneless clay or sand are very restricted. Where the glacial drifts are thin. Particularly over Lower Palaeozoic rocks, the soil may be packed with angular rock-debris. Soils are described by Ragg and Futty (1967).

Along most of its length the coast has high rocky cliffs overlooking a narrow foreshore (Plate 2), Between Siccar Point and St Abbs the steep cliffs are generally over 60 m high, reaching a maximum of 150 m above Mawcarr Stells. (Siccar Point may owe its name to a promise of safety which its sighting afforded to west-bound mariners: Scots sic-car = safe.) Between Eyemouth and Burnmouth the cliffs again exceed 100 m in height and here, as at St Abb's Head, they fall sheer into the sea in places. Sandy beaches occur only at Pease Bay, Coldingham Bay and Eyemouth. Landslips are widespread on the cliffs between Pease Bay and Meikle Poo Craig, and are hazards at Burnmouth and southwards from Hilton Bay.

Geological history

Throughout Lower Palaeozoic times the district lay beneath the Iapetus Ocean towards its north-western margin. The Ocean floor was being subducted beneath the north-western Laurentian continent, and the Southern Uplands Ordovician and Silurian strata are widely interpreted as part of a fore-arc accretionary prism developed above the subduction zone along the trench slope, and formed by turbidites of northern derivation with interbedded oceanic graptolitic shales (McKerrow and others, 1977; Leggett and others, 1979). However, the Geological Survey has recently proposed an alternative origin in a back-arc basin (Barnes and others, 1986; Stone and others, 1987). Whatever the origin, an imbricate structure, which developed during and after deposition, resulted in a series of fault-bounded slices of thick greywackes overlying thin oceanic shales, tightly folded and generally dipping steeply north-west and younging in that direction, with successive slices becoming younger to the south-east. The Coldingham and Linkim beds, depicted on the map as older than the main mass of Llandoverian strata, must now, as a result of recent work by Casey and Oliver (1981) and Molyneux (in press), be regarded as late-Silurian (see Chapter 2).

By the Lower Devonian or possibly late-Silurian, the ocean had closed in this region, and extensive calc-alkaline volcanism ensued. Large volumes of pyroclastic material were ejected in association with thick and extensive flows of lava. This activity was accompanied by intrusion of dykes of similar composition to the lavas. Larger bodies of more acid composition were also intruded, notably the granite of Stoneshiel Hill, which is part of a much larger generally concealed batholith. On the decline of volcanic activity and partly contemporaneous with it, the volcaniclastic Lower Old Red Sandstone was formed by subaerial erosion of the lava fields.

Further tectonic activity, much less intense than during the Lower Palaeozoic, caused the Lower Devonian rocks to be folded and faulted in broad structures which follow older alignments.

In Upper Devonian and Lower Carboniferous times the district at first formed part of a land-mass composed of hills of greywackes and volcanic rocks, fringed by extensive plains. Across these were spread the torrent and flood deposits of broad rivers, forming the Upper Old Red Sandstone.

By Lower Carboniferous times the lower areas had subsided essentially to sea level and the sea itself lay at no great distance to north, east, and south. In shallow lakes and coastal lagoons partial evaporation led to the formation of calcareous and dolomitic cementstones which occur as thin beds among the fluvial sandstones and mudstones of the lower beds. In the south-west of the district these sediments were preceded by basaltic lavas and tuffs erupted from terrestrial volcanoes, near the limit of a volcanic field which tends south-westwards to Annandale. In the south-east the later-formed Fell Sandstone Group and Scremerston Coal Group afford evidence of a large delta advancing from the north, the varied succession of lithologies of the latter group being due to the varied topography of the delta surface, where active river channels and quiet backwaters separated vegetated spreads of low-lying alluvium. Similar conditions prevailed in the Cockburnspath area. In both areas too the coastal plains were at times invaded by the sea. The Lower Limestone Group, in the south-east, shows more frequent and longer lasting marine incursions in its beds of limestone.

Of geographical conditions in later pre-Ouaternary times there is no direct evidence. Basic intrusions of late Carboniferous age cut the strata, and the Carboniferous rocks show Hercynian folding and faulting.

Pleistocene glaciations were initiated upon a topography different only in detail from that of today. The major pre-glacial rivers are thought to have flowed south-eastwards from a watershed along the north-western edge of the Lammermuir hills, the Eye, for example, from Edmond's Dean and the valley north-west of Penmanshiel. During the Devensian glaciation, which ended some 10 000 years ago, the whole district was over-ridden by ice moving east-north-east in northern and western areas, then towards the south-east across the lower Palaeozoic terrain, and eastwards towards the sea along the northern margin of the Tweed basin.

In late-glacial times the ice-cover was eventually reduced to remnants in some upland valleys, while North Sea ice of Scandinavian origin impinged on the coast. The local persistence of ice partly controlled the courses followed by the meltwater, and there are many instances where streams were diverted from older courses. Higher base-levels of drainage, due at least in part to ice obstruction, are attested by the wide occurrence of river gravels many metres above modern flood-levels. Evidence of higher sea level is uncommon but there are traces of a beach about 5m above the present beach, the local equivalent of the post-glacial shoreline widely developed elsewhere in Scotland. Earlier, probably during the Loch Lomond Re-advance, sea level was generally lower than today. At Eyemouth, a submarine channel about 18 in deep is thought to have been cut by the Eye Water during this period of low sea level.

In post-glacial time, as vegetation became generally established, deposits of peat built up in poorly drained areas. Many of these are on watershed cols, where some of the most persistent masses of ice had stood during deglaciation. The process of adjustment of levels of valleys to base-levels of erosion continued throughout post-glacial time and is by no means complete. Every stream is made up of a series of reaches separated by nick-points, at which the gradient increases sharply downstream and to which the reach above is graded.

Chapter 2 Silurian and Ordovician

The district includes the most north-easterly part of the Southern Uplands of Scotland, the Ordovician and Silurian strata of which have been considered to form part of a structurally compressed, fore-arc accretionary prism formed above a NW-directed subduction zone along the northwestern edge of the Iapetus Ocean (McKerrow and others, 1977; Leggett and others, 1979). However, recent work by the Geological Survey (Stone and others, 1987; Barnes and others, 1986) interprets the Southern Uplands as an imbricate thrust belt developed during crustal shortening of a back-arc basin. Whatever the origin, the eventual geometry is a series of steeply-dipping, fault-bounded slices which, while individually north-west younging, are sequentially younger in a south-east direction.

Silurian strata form three separate outcrops in the district. The largest comprises the north-east end of the Lammermuir Hills and is of Llandovery age with a small Ordovician inlier at Silverwells [NT 883 662]. This inlier probably marks the boundary between two major slices. The southern outcrop, south of Eyemouth, is unfossiliferous, but as the strata resemble the Hawick Rocks farther south-west, this outcrop is considered to be part of a third major slice of upper Llandovery age.

The third outcrop, of the Coldingham and Linkim beds (formations, properly speaking), is now thought to be of later Silurian age and perhaps not strictly part of the accretionary prism (Casey and Oliver, 1984). Geikie (1863) noted that these two formations are structurally the most complex in the district, and this was confirmed in more detailed studies by Shiells and Dearman (1963, 1966) and the present resurvey. From this it was deduced that they were the oldest in the district, and the Coldingham Beds are therefore shown on the map as Ordovician, while the Linkim Beds are shown as Silurian because they yielded poorly preserved Monograptus sp. Since the map went to press, Casey (1983; and Casey and Oliver, 1984) has demonstrated that soft-sediment deformation is an important element in the complexity of these formations and that their metamorphic facies is lower than that of the Lammermuir outcrop. Shales from the Linkim Beds have also yielded microfossils of probably Wenlock age (Molyneux, 1981 and in press). For these reasons, the Coldingham and Linkim beds are now regarded as the youngest Silurian strata in the district.

Lammermuir outcrop

Silurian Llandoverian strata are excellently exposed along the coast (Plate 2), where they' consist of a highly folded sequence of greywackes, siltstones and mudstones, hundreds of metres thick but lacking distinguishable horizons or groupings which might be used in subdivision or correlation. Craggy exposures of greywacke are abundant on high ground near the coast, and locally in streams, but inland exposures are too sparse to contribute to the correlation of structures seen on the coast. Poorly preserved graptolites from widespread localities show that the rocks are of upper Llandoverian age (Gala Group) with one or two inliers of middle Llandoverian and Ashgillian age (Birkhill and Upper Hartfell shales).

Lithology

The strata form an interbedded sequence of greywackes, siltstones and shales deposited from turbidity currents, and are now seen as a succession of repeated Bouma units (Bouma, 1962). Overall about half the strata are greywacke, the remaining half being divided equally between flaggy greywackes with siltstones, and shales.

While there are local variations in the proportions of the lithologies and in the thicknesses of units, there is a general appearance of lateral persistence and consistency of thickness. The measured sections fail to show any large-scale variation which might reflect the evolving geography of deposition. General conditions of deposition across the outcrop appear to be uniform.

Palaeontology

Graptolites have been collected from some 40 localities, but few of the faunas have enough different species to allow a precise determination of horizon (Strachan, 1982, p.158). All except one of the localities yielded Silurian graptolites, the exception being that at Silverwells Dean [NT 8831 6624] where a small fauna including Dicellograptus sp. was obtained indicating an Upper Ordovician (Ashgillian) age.

Most of the faunas collected can be assigned to the zones of Rastrites maximus, Monograptus turriculatus and M. crispus, but at a few localities earlier zones may be present. In Buskin Burn [NT 8893 6655], some 700 m ENE of the Ordovician exposure and nearly on strike with it, a Birkhill Shale (Middle Llandovery) fauna was obtained which probably belongs to the M. convolutus Zone. Two localities near Siccar Point [NT 8122 7085] and [NT 8126 7101] yielded poor faunas in which Monograptus sedgwickii is present, together with M. spiralis; these may belong to the M. sedgwickii Zone or the early part of the succeeding Rastrites maximus Zone. At the nearby Old Cambus Quarry [NT 8062 7050] Strachan (1982) recorded a possible M. turriculatus Zone fauna, in which a number of very slender forms of Monograptus are present.

Faunas in the majority of the localities are small, often consisting of comparatively long ranging forms. About five faunas can be recognised covering the zones of H. maximus, M. turriculatus and M. crispus. Strachan (1982, p.159) has interpreted the occurrence of beds containing abundant M. exiguus without M. crispus or alternatively abundant M. crispus without M. exiguus as the result of the segregation of local swarms within the depositional sequence. M. discus occur both in association with M. crispus and also in faunas not containing the latter species which may represent a slightly higher horizon. The presence of M. discus in these faunas contrasts with those to the west (Sheet 33) where it is very rare.

It is possible that the succeeding zone of Monoclimacis griestoniensis is present in the district. A poor fauna from the shore at Lumsdaine [NT 8744 7024] contains species suggesting the presence of this zone. Some 25 m to the NW [NT 8743 7026] another graptolite-bearing mudstone contains a fauna which may be referred to close to the boundary between the Monograptus turriculatus and M. crispus zones. This evidence points to a younging in a south-easterly direction in contradiction to the structural evidence (below). It is therefore possible that a large strike fault is present in the section, though this was not detected in the field survey.

Stratigraphical succession

An attempt has been made to work out the stratigraphic succession along the coast, but lack of distinctive lithological sequences, the sparsity of fossils and the impossibility of making allowances for the effects of faulting all render the exercise liable to error. The obliquity of the limbs of many folds and the consequent plunge of the fold-axes complicate the calculations and also govern the way in which the findings can be expressed. Calculations using observed fold-plunges lead to somewhat different results from those which allow indirectly for the plunge. Bed-by-bed measurements of strata in some cases give total thicknesses significantly less than those calculated geometrically.

(Figure 2) is a diagrammatic section along the coast, showing only major folds, and taking as datum the strata exposed [NT 906 690] at sea level in the axis of the Broadhaven Beach Syncline. With the provisos mentioned above, some 600 m of strata are estimated to crop out in the Redheugh to Pettico Wick coast section. This is considerably less than that estimated by Geikie (1863). The section shows NW younging strata between the axes of the Lumsdaine Shore anticline [NT 877 700] and the Brander Heugh syncline [NT 873 704], but the two fossil localities in this stretch show higher strata to south-east of lower, indicating a situation structurally more complex than shown.

Petrography

The greywackes consist predominantly of quartz grains with varying proportions of feldspars, micas, and small clasts of fine-grained sedimentary and igneous rocks. Quartz content varies from about 15 to 60 per cent, with feldspar in inverse proportion. Zircon, garnet, tourmaline, and apatite are seen in many thin sections and epidote and microperthite are not uncommon. Most greywackes show a bimodal grain-size distribution between clasts and matrix, but in many the distinction is blurred and there is an even gradation from smaller to larger constituents. The relative abundance of different sizes is very variable. Rock fragments are generally well rounded. Many are igneous with abundant feldspar laths, generally oligoclase-andesine and commonly exhibiting flow-structure. Sedimentary rock fragments are predominantly sericitic mudstones and siltstones, with scattered tiny grains of quartz. Metamorphic rocks appear as fragments of polycrystalline quartz, recorded in about one-third of the thin sections, and showing strain shadows, sutured grains and oriented mica flakes.

Quartz and feldspar, and indeterminate siliceous material, compose most of the matrix of the greywackes. Muscovite and biotite flakes, chlorite, lithic fragments, magnetite and leucoxene occur almost universally, and pyrite is recorded in several examples. Minor constituents include the accessory minerals mentioned above and, in a few rocks, grains of pyroxene and possibly amphibole. Calcite is not uncommon as a cement, and in some rocks seems to occur as detrital grains. Several specimens exhibit an interlocking texture, particularly developed in the quartz grains of the matrix, which is ascribed to pressure-solution during lithification. Sedimentary lamination is common, manifested by parallelism of the micas and of the larger dimensions of the lithic fragments.

Siltstones and mudstones consist largely of micas and clay minerals, with a high proportion of material not readily determinable in thin section. Quartz is always present, more abundantly in the coarser-grained rocks, where biotite, muscovite, iron ores, and feldspar are increasingly prominent. Parallelism of mineral grains is a common feature and, except for the most finely grained, the rocks show bands of different grain-size or mineral composition, both parallel and oblique to the principal bedding-planes.

In the field, the greywackes appear to be predominantly of medium to coarse sand grade. A few are finely pebbly in the basal few centimetres. Lamination is generally restricted to the mudstones, siltstones and finer-grained greywackes. Some greywackes are very rich in quartz or in muscovite, and a few include carbonate concretions. Variations from the common grey colour are due to abnormally abundant hematite or chlorite.

Quartz-rich greywackes (rock slices (S47987), (S48563) and (S49464) dominate the northern part of the area, but also occur to the south within generally quartz-poor greywackes.

Metamorphism

The strata have a low level regional metamorphism, to prehnite-pumpellyite grade. Biotite is developed in places, but there is no schistosity or slaty cleavage. Near the Stoneshiel Hill granite [NT 775 590] and immediately adjacent to minor intrusions the rocks are thermally metamorphosed, and are darkened and tougher and harder than normal. The shales have a pink tinge and a splintery non-fissile texture. Some rocks have abnormally abundant fine-grained magnetite or veins and segregations of silica and carbonate.

Sedimentary structures

Turbidity current deposits have a characteristic and widespread range of sedimentary structures, including graded bedding, internal laminations and assorted sole marks, from which younging directions of the strata and directions of flow of the depositing currents may be determined.

Greywackes locally show vertically restricted internal convolution developed during sedimentation before the succeeding sediment was deposited. Vertical grading is especially noticeable in the thicker greywackes which fine upwards from coarse, sometimes pebbly, sandstones to fine-grained sandstones and siltstones with

The upper surfaces of some greywackes carry ripple-marks, most commonly asymmetric wave-forms at right-angles to the direction of flow, with a wave-length of the order of 10 cm. Good examples are seen on the north-west slopes of Brockholes Hill [NT 811 653] and in a quarry [NT 836 694] beside the Dowlaw farm road. Larger ripple-marks are well seen on a crag [NT 8530 6956] 200 m SW of Lowries Knowes (Plate 3), and near the north end of the east side of Penmanshiel railway cutting [NT 7963 6720]. Ripple-marks of very short wave-length, between 1 and 2.5 cm, also occur, e.g. in fine-grained greywackes and siltstones in an old quarry [NT 9057 6850] 1180 m south-west of St Abb's Head Lighthouse, where their sharp parallel crests define many of the upper bedding-surfaces.

The commonest sole-markings are flute- and groove-casts, linear markings which may cover an entire exposed basal bedding-plane of a greywacke. Flutes were formed by the scouring action of the dense turbidity current, and grooves by solid particles suspended at its base being dragged along the bottom. Flute-moulds are smoothly rounded and tend to penetrate the sub-stratum more deeply and sharply than groove-moulds. (The term 'cast' is used for the material which tills the erosional depression, the 'mould'). Flute-casts have rounded or bulbous upstream ends, with downstream margins sinuous and cuspate in detail, and showing signs of internal turbulence in the current. There is every variation from markedly linear and roughly parallel to apparently random orientation, which may result from variation in the interaction between linear and turbulent flow. (Plate 4) illustrates a variety of small flute forms, including many sets of parallel casts which bear close comparison with the longitudinal ridges depicted by Dzulynski and Walton (1965, figs. 39, 48A, 49), some with crescentic ends suggestive of progressive expansion of a scour-mark both upstream and laterally. Most commonly flute-marks are of the order of 25 mm wide, but many are very much smaller and a few very much larger.

Groove-casts are common and conspicuous, and are markedly linear or only broadly curved, They may occur in relative isolation, as for example on the shore-stack [NT 8879 6947] 350 m SE of Mawcarr Stells, where the groove cuts across earlier formed flutes and narrower grooves at an angle of 15°. Another isolated curved cast at Pikie's Rock [NT 8902 6945], is 12 cm at its widest and appears to merge into flute-casts at each end (Plate 4). Structures in several directions on one bedding-plane are fairly common. Groove-casts more usually occur in groups, completely covering a bedding-plane in some exposures. At the foot of the cliffs [NT 8732 7029] 415 m S of Brander there are parallel linear grooves up to 180 mm wide (Plate 5), the margins of which carry fluting similar to the rill-marks of Dzulynski and Walton (1965, fig. 43).

In general groove-casts are themselves grooved, the outer skin being prominently fluted or ribbed in a direction closely parallel to the major feature. They are characteristically shallow, regularly convex in cross-section, and show little variation in depth along their length. It is most likely that grooves were eroded by rock fragments, or in the case of the wider grooves by armoured mud balls, carried across the sea bed by heavily-laden rapidly moving currents.

A few examples were seen of prod- or bounce-casts, trains of isolated marks caused by the recurring impact of a particle being carried along by the current. Frondescent marks also occur: they resemble converging flute-marks but appear to be caused by diverging currents, the deeply down-cut rounded leaf-like terminations marking their downstream limit. Dzulynski and Walton (1965, p.132) postulate that they are examples of load-cast.

Load-casts are immediately post-depositional gravity effects, irregularities in the sediments giving rise to differential compression and displacement of the deposits. Flute- and groove-casts may become over-deepened and develop a pear-shaped cross-section or ultimately become detached from the main body of sediment above. Load-casts are most conspicuously developed at the bases of thick greywackes and may show no clear orientation or pattern. A common feature is the development of flame-structures, which are tongues of mudstone penetrating up to several centimetres into the load-cast base of a greywacke. In some cases they bend over in the downstream direction.

Detached masses of sediment, generally rounded oblong in shape, are not uncommon in the upper parts of greywacke-siltstone units. A good example from the shore [NT 8904 6936] immediately south-east of Heathery Carr is shown in (Figure 3), where fragments of laminated siltstone, with thin concentric layers of greywacke in one are embedded in fine-grained greywacke, and are thought to be an extreme form of load-casting. Other detached masses have originated differently. The mass shown in (Figure 3)b was redeposited as a completely lithified rock fragment.

Where possible the orientation of current-generated structures was measured. (Figure 4) illustrates the azimuthal distribution of the inferred original orientations, after allowing for folding, and demonstrates a preferred orientation of N230°. Transverse ripple-marks reveal a markedly preferred current-direction towards N160°. The principal results of this analysis, the south-westward flow of the erosive currents responsible for the sole structures and the different direction followed by most of the depositional currents which rippled the upper surfaces, accord with the observations of other workers in the Silurian of the Southern Uplands (Kelling, 1962, 1964; Craig and Walton, 1962; Rust, 1965; Weir, 1974). Kelling (1964) explained the different direction of the currents rippling the upper surface of a greywacke as transverse displacement waves reflected by a surface such as the margin of a submarine channel. Most of the above analysis is from coastal exposures, and no geographical or stratigraphical variation was recognised.

Structure

Structurally the area is dominated by major open folds, more or less upright, with NE or ENE trend, strata dipping at high angles to NW or SE, and axial plunges generally low. This structural pattern is generally considered to have originated during sedimentation followed by late-Silurian tectonic tightening.

General structural variations

The coastal section shows a progressive change in fold orientation as it is followed from Siccar Point to Pettico Wick. Axial planes of the western folds dip SE at 75° or more, but between Midden Craig [NT 838 702] and Lumsdaine Shore [NT 877 700] the inclination is steeply to NW. South-east inclinations, generally as low as 65°, prevail between Shilments [NT 883 698] and Pikie's Cove [NT 890 694], east of which the axial planes are nearly vertical, dipping in all but two cases steeply to SE. An analogous variation in axial plunge is also observed. The folds west of Siccar Point plunge SW at 20 to 55°. From Redheugh Shore [NT 830 703] to Little Rooks [NT 854 710] the direction is NE at 25 to 58°. Folds eastward from Fast Castle revert to a SW plunge but the limbs are more nearly parallel and the angle of plunge, exceptionally up to 30°, is generally less than 15°. East of Little Pits [NT 896 693] the hinges are approximately horizontal. Scattered observations inland do not suggest any systematic areal pattern of plunge variation.

The general direction of strike between the coast and the Eye valley is N060°, but in the Whiteadder north-west of Preston the direction is between N045 and N025°. In the eastern corner of the outcrop, in Buskin Burn [NT 895 662] west of Coldingham, a northerly strike prevails over a cross-strike width of some 500 m. A narrower zone of northerly strike occurs at Houndwood [NT 842 641], and on Dalks Law [NT 848 649] to north-east is an equally narrow zone of easterly strike.

Folds

While individual folds are displayed at many inland exposures, it is on the coast that every aspect of structure is most comprehensively discernible. Faulting is universal though actual displacements are rarely recognisable. In general folds are broad-limbed with narrow hinge-zones, and it is possible to identify major anticlines and synclines. In a 5 km cross-strike section there are 30 major reversals of dip. Unfolded limbs are between 15 and 550 m wide and contain up to 520 m of strata. As neither axes nor limbs are always mutually parallel it will be apparent that the widths and stratal thicknesses of the limbs vary considerably along strike.

The folds themselves are generally concentric with well rounded hinges, the massive greywackes in particular preserving their thickness across them. Typical examples are the Hawk's Heugh [NT 861 709], Moorburn Beach [NT 888 695] and Heathery Carr [NT 8903 6936] synclines and Brander anticline [NT 872 705]. Structures in more thinly bedded strata may be much more complex. They may, as in the tight anticline of Pikie's Cove [NT 8900 6940], be considerably drawn out in the axial plane, with minor subsidiary faulting and considerable bedding-plane slip (Figure 5). Alternatively additional subsidiary folds develop in the thinner-bedded strata. In the monoclinal anticline [NT 8835 6970] south-west of Mawcarr Stells a rounded arch in greywackes is succeeded upwards by an M-shaped fold in flaggy and shaly beds. Similarly at Midden Craig [NT 8382 7080], shales and flaggy greywackes in the core of a broad anticline form an asymmetric N1-shaped fold, with significant drawing out of the shales in two of the axial planes. A widespread phenomenon, which is clearly characteristic of the general structure, is the change in the fold-profile within the axial plane. A fold may change shape within a metre or two, either by the development of flexures in planar beds or by the accentuation of existing flexures, to such an extent that, lacking continuity of exposure, the structures would not be recognised to be the same. The anticline of Fast Castle is a multiple structure incorporating a central syncline [NT 8600 7104] which, within 100 m NE along the axis, is transformed from a simple slightly distorted fold into a sharp W-shaped structure, with the constituent synclines about 2 m apart. Similarly, the syncline [NT 8686 7076] south of Souter appears on the cliff-face as a simple structure, but on the foreshore the hinge zone is markedly disharmonic.

The spectacular anticline of Brander Cove [NT 8719 7055] carries a double fold on its north-west flank. In seaward exposures it is isoclinal, the anticline plunging at 30° to SW, the axial plane very steep to NNW, and the syncline difficult to trace and probably faulted. To the south-south-west the subsidiary folds are more open and more widely separated, but have a similar orientation. The main fold on the other hand has a very low axial plunge to seaward and a near-vertical axial plane. As the subsidiary folds diverge south-westward from each other so do they as a pair diverge from the main fold. A similar pattern may be seen [NT 8418 7023] in a syncline which can be followed south-west to Tod's Rock [NT 8406 7017]. The NW limb of this broad-crested fold is itself folded close to the main axis by an anticline and syncline, nearly isoclinal, with only about 1 m of strata between them; but 50 m to SW, at lower structural levels, no such complications are recognised. The folds pass out to sea within a few metres to NE and where seen again 700 m to NE near Black Bull [NT 8483 7063], the detailed structure is not comparable. Because folds change shape within short distances even large folds cannot be traced through isolated exposures. It is possible to follow the major coastal folds inland for about 2 to 2.5 km where exposure is reasonably good, but farther extrapolation is speculative.

Unusual structures

There are a few abnormal structures which appear to be unique. An unusually complex structure forms part of a major syncline at Little Pits [NT 8956 6927] (Figure 6). Four closely spaced isoclinal folds, within inverted strata dipping steeply to SE, are cut off upwards by a low-angle fault, beneath which the crest of the more north-westerly anticline is markedly rotated in the north-west direction of hade.

Strata on the beach [NT 8975 6915] below Uilystrand Brae between two well defined round-crested anticlines are overall approximately horizontal. In detail, however, these thinly bedded strata are crumpled into many small impersistent folds with low axial plunges to SW and a variety of cross-sectional shapes. One mid-shore cross-strike section 4 m long consists of two flat-hinged, steep-limbed, box-folds. A few metres to landward the central steep limb of the two folds is replaced by a small faulted monocline. Some 5 m to seaward from the box-folds many more folds develop and the structural details are difficult to correlate between the sections. The folds are box-shaped, with angular or rounded hinges.

Another abnormal fold occurs in association with the anticline of Thrummie Carr [NT 901 691]. On the shore it appears as a well rounded fold with limbs dipping steadily at 45° to NW and SE. Above to the south-west on Biter's Heugh [NT 8997 6900] it has a geniculate form, the SE limb nearly horizontal, the NW one dipping steeply to NW, an apparent rotation reminiscent of that described from Little Pits. Shiells and Dearman (1966, p.233, fig. 2) described this structure as a parasitic fold developed on the NW limb of the main fold, showing down-dip movement contrary to the orientation of a typical drag-fold.

The strata in Silverwells Dean [NT 883 662] which have yielded Ashgillian fossils are structurally complex. Two antiforms with crests 2 to 3 m apart, are overturned to NW with a steep overturned median limb and considerable minor fracturing of thin greywackes and contortion of the shales. For 300 m to NW of this locality, and possibly farther to the south-west, the closely folded strata, apparently Silurian, are overturned in the opposite sense and both boundaries of the Ashgillian rocks are probably faulted, the southern one being a major fault. Confirmatory brecciation, veining and slickensiding are seen.

Faults

Faults are numerous throughout the outcrop. Many of those seen close to fold-axes are syntectonic but most faults are not clearly related to the folding. About two-thirds of the faults are high-angle with dips steeper than 50°, and of these some 55 per cent show predominantly horizontal displacement, with strike-distribution as shown in (Figure 7). A majority, about 65 per cent, are sinistral faults striking about N 015°, while the dextral faults strike around N 150° and N 015°. The orientation of these lateral faults may be interpreted as resulting from near-horizontal compression from NNW to SSE, resulting in an overall sinistral or counterclockwise displacement of the rocks.

Faults which have a predominantly vertical displacement are evenly distributed between high-and low-angle, but reverse faults (over 80 per cent of the total) are relatively more abundant in the low-angle category (Figure 7). Most low-angle reverse faults hade to NNW or SSE, but most high-angle faults hade to SSE. Those which hade to SSE are generally more steeply inclined than those hading to NNW. Many faults are very closely associated with folds, in some cases replacing limbs. In other cases fold axes are displaced by small reverse faults. Most low-angle reverse faults hade in the same direction as the axial planes of nearby folds, and most of the high-angle faults hade in the opposite direction. The two sets of faults can be interpreted as being complementary, and related to the horizontally compressive folding forces in an analogous fashion to the formation of complementary sets of lateral faults, but with vertical relief of stress, rotated slightly in the same sense as the asymmetry of the folds. Other commonly recorded features of reverse faults are their occurrence in groups of several parallel or subparallel fractures and the curved form of the fracture-surfaces.

There is no consistent age relationship between normal and reverse faults. Most normal faults strike ENE and about two-thirds hade to SSE at an average of just over 40°, like that for all the normal faults.

Fault displacements up to 20 m may be seen in exposures, but many faults certainly have much greater throws. The only apparently major fractures observed occur on the coast west of Fast Castle between Green Stane [NT 8485 7072] and The Little Rooks [NT 8526 7093], where two parallel zones of sheared and brecciated greywacke and shale run close to the strike of the north-western limb of the Tod's Rock–Little Rooks syncline. The sheared zones, about 15 m apart, vary in width up to at least 5 m and dip at 60 to 75° to NNW, whereas the strata dip generally at 80° or more to NNW or SSE. In the narrow strip to seaward of the northerly shear the strata dip steeply to SSW, and there may be a similar rotation in poorly exposed rocks to south. The shears are considered to be high-angle reverse faults with NW hade. They run out to sea at each end of the exposures, but do not reappear on their linear extrapolation around Tod's Loup [NT 836 702].

A syncline exposed on the foreshore [NT 830 703] near Hirst Rocks (Figure 8), displays bedding-plane fractures and a narrow crush-zone cutting across the fold at a high angle to the axial plane.

Southern outcrop

The southern outcrop extends from Eyemouth to the northern outskirts of Berwick. Its boundary is seen in only a few places, but mapping indicates that it is largely faulted. The rocks are magnificently exposed on the coast between Eyemouth and Burnmouth. Inland exposures are abundant in the Eye Water but elsewhere they are mainly small and isolated. Minor intrusions are numerous, especially in the eastern part.

Lithology

The strata comprise a thick succession of well-bedded turbidites, closely folded and steeply dipping. By comparison with the Lammermuir outcrop individual greywackes are thinner, few exceeding 50 cm. Shaly beds are usually less than 75 cm thick, and generally comprise 17 to 39 per cent of the strata in a section. A high proportion of the strata consists of alternating beds of greywacke and shaly siltstone or mudstone varying between 25 and 300 mm in individual thickness. These observations suggest that, compared with the Lammermuir outcrop, the Eyemouth strata are more distal turbidites.

Petrography

A typical greywacke, from Ramfauds [NT 9511 6454], consists largely of subangular and subrounded grains of quartz, up to 0.35 mm across, set in an abundant calcareous matrix. Some grains of feldspar and rather rounded lithic fragments are also present, of the same size as the quartz grains, as well as many small grains of carbonate. Small flakes of muscovite or sericite and grains of magnetite and hematite are moderately abundant. Features which distinguish the greywackes of the southern outcrop are the preponderance of quartz grains over feldspathic or lithic fragments, the common occurrence of a calcareous matrix, and the general fineness of grain-size. Muscovite and iron-ores are generally present but igneous rock fragments and accessory minerals are rarely recorded, though observation is hampered by the fine grain-size and abundant carbonate.

Rocks described in the field as siltstones or shales appear in thin section as very fine-grained greywackes or coarse siltstones, the larger grains being subangular quartz along with tiny flakes of muscovite and/or sericite. The finest-grained rocks are grey mudstones with a flinty conchoidal fracture, dominated by sericitic or chloritic material which in some cases shows parallel orientation in very small-scale cross-bedding. The original material and texture are often obscured by very fine-grained carbonate, which is developed in abundance. Thin persistent red mudstones south of Elgy Loch and Agate Point are similar to the grey beds, but distinguished by abundant fine-grained hematite, which in some specimens shows apparently primary differences of concentration in very thin bands. Cross-lamination on a microscopic scale is locally developed. The persistence and sharpness of definition of the red beds suggest a primary origin.

In addition to these primary red beds, secondary reddening is widespread. Some is probably penecontemporaneous as is shown by its irregular patchy development. Other reddening appears to be due to modern weathering as is indicated by its prevalence in river bed exposures and by the shore section at Ramfauds [NT 9501 6449] where reddening is more pronounced close to sea-level. Intensive reddening at the top of Blaikie Heugh [NT 9529 6266] appears to be due to the intrusion of a porphyrite dyke.

Thin veins of calcite are not uncommon throughout the outcrop. In coastal exposures between Elgy Loch [NT 953 644] and Hawk's Ness [NT 955 629] they include baryte in places and lie mainly in bedding-planes, but near Hawk's Ness they are concentrated near a high-angle fault trending SSE. A feature of this fault, and of many others along the coast, is the development of coarse angular breccia (Plate 6). These breccias, 1 to 2 m wide, composed of fragments of greywacke some 5 cm across, occur in fracture-zones and close to fold-axes particularly on the coast between Whalt Point [NT 956 639] and Fancove Head [NT 957 625]. In several cases the displacement across the fractures appears to be no more than a few metres. The geographical restriction of this phenomenon may be more apparent than real, but the brecciation is so distinctive that it is unlikely to have escaped observation and comment if well exposed elsewhere. The structure in this area follows a regular pattern of parallel paired folds and is in this way significantly different from that elsewhere along the coast.

Sedimentary structures

Sedimentary structures of the same types as those seen in the Lammermuir outcrop are common here in the south. The relative preponderance of sole-structures over upper-surface structures is maintained, but in the southern outcrop the ratio of bedding planes with flute-casts to those with groove-casts is more than twice that in the main outcrop. This difference may be related to the higher proportion of finer-grained rocks in the southern outcrop, and may be compared with a similar variation noted by Schenck (1970) from Lower Palaeozoic flysch-like rocks in Nova Scotia. Though data are sparse, flute and groove casts trending ENE, between N 050° and N 080°, predominate and show that the currents were mostly from ENE.

Structure

Tectonic structures in the southern outcrop are more complex than in the Lammermuir outcrop (Geikie, 1863; Mackenzie, 1956; Dearman and others, 1962). On the coast they are simplest in the zone, 800 m wide across strike, between Whalt Point and Fancove Head mentioned above, within which are more than 45 fold pairs, the longer limbs dipping at around 70° to WNW and the shorter limbs more steeply to SSE, locally inverted. Individual folds plunge to SW more steeply than 45° and most of them have narrow, sharply curved, hinge-zones and relatively long subplanar limbs. Most folds are concentric, but shearing and axial thickening are locally developed in the thinner-bedded strata, with multiple zig-zag folds in some hinge zones (Plate 7). Axial faulting is common but displacements are small. The strata generally young to NW.

South of Fancove Head the structural pattern is less clear, partly because of the abundance of minor intrusions, but appears to be similar.

On Whalt Point and to the north-west structures are similar to those to south but the SW plunges of the folds are generally steeper, between 75° and vertical. More complex structures prevail beyond an inlet [NT 9544 6400] 180 m north-west of Whalt Point, west of a N – S reversed or sinistral fault which runs out to sea [NT 9546 6416] between the high and low water at Agate Point. On a headland [NT 9543 6403] just west of the fault a pair of downward-facing tight folds plunge at 83° to NE. The acme of structural complexity is reached within a zone 160 m wide, bounded by the Agate Point fault and a parallel fault which runs close to high water west of John's Roads. Many of the details of this zone have been discussed and interpreted by Dearman and others (1962). The resurvey confirms their essential finding, that the axes of the closely spaced folds have been folded along their length, causing variations in plunge which are locally abrupt. At Agate Point most of the folds are downward-facing and plunge steeply to NE but in places the plunge can be seen to steepen and to tilt over to SW. These variations are consistent with the effects of overturning from north-east to southwest.

The interpretation by Dearman and others (1962) of the folding and faulting on the promontory [NT 9535 6415] at the south end of John's Roads is also confirmed by the resurvey. At this locality at least six folds are seen within a cross-strike distance of 10 m, plunging at about 25° to NE, and downward-facing on the evidence of sedimentation features (Figure 9). Their SW continuation is interrupted by a cross-strike fault beyond which fold-axes are prominently displayed, showing continuous variation in plunge from SW, through vertical, to NE. On both sides of the promontory these folds face SW, but plunge is variable in a series of sideways closing, open structures. At first sight it might appear that the downward-facing NE-plunging folds, north of the cross fault, have been overturned along their axes from SW to NE. Dearman and others (1962) comment on the similarity between the John's Roads and Agate Point structures but point out that the sense of overturning is opposed. lf, however, the folds at both localities were generally upward-facing before the compressive overturning then the sense of overturning in both cases is most simply inferred to be from NE to SW.

The fault which marks the western limit of the John's Roads–Agate Point zone of complexity closely follows the axial trace of a rounded steep-limbed anticline which is well displayed on the foreshore [NT 9529 6413] SW of John's Roads and for 50 m to N. The plunge is very steep to N at each end of the exposure, vertical near the middle, and about 45° to S, downward-facing, within the northern half. Close to this line, a short distance to the north, comparable folds are exposed on the shore [NT 9529 6422]. At least five folds within a cross-strike distance of 10 m constitute an open anticline plunging generally at very high angles to NNW. The NE-facing limbs, vertical or overturned, are steeper than those which dip NW at between 45 and 90°.

Within a few metres to the west of this anticlinal line the beds are folded on NE axes and traversed by an associated axial plane cleavage. As the beds are followed across the N-plunging anticline the angle between bedding and cleavage is maintained, clear evidence that the anticline is a significantly later structure than the NE folds.

These latter folds, plunging WSW at upwards of 45° and in some cases passing into axial-plane faults, are abundant to NW. Most of them are tight structures with limbs dipping at around 65° to NW and more steeply to SSE, but more open rounded folds, with lower angles of plunge and less steeply inclined SE limbs, become common towards Elgy Loch [NT 9525 6440]. This inlet follows the axis of an anticline of this type, with dips of 65° to N 330 and 57° to N 170° and an axial plunge of 25° to N260°. Flanking it 10 m away on either side are synclines similar in form and orientation. A few metres beyond the north-west syncline, but diverging to north-east to form the south-east side of Elgy Rocks [NT 9526 6444], runs the broad axis of another anticline of this group. Transverse current ripple-marks are common in this area, and on the crest of this fold at Elgy Rocks their alignment with the axial trace direction has resulted in their modification into parallel mullions 10 cm wide.

A succession of folds is well displayed on Elgy Rocks, a headland isolated at high tide by a channel along a fault dipping steeply to NE. Towards the seaward end 18 folds were recorded in a 50 m SE traverse. Their aces plunge at between 20 and 50° to WSW, with both open and sharp hinges, the general dips of the limbs being 60° to NW and 80° to SSE. Variations in fold-development are common, both along strike and vertically. In the north-west part of the traverse several well-defined folds die out within 10 m along strike, and a series of eleven folds in a 15 m section is represented by an open syncline in the cliff about 75 m to SW. A spectacular exception to the rounded folds which prevail in this area is displayed near the landward end of Elgy Rocks (Figure 10) where the folds appear to be displaced by SE overthrusting, with many minor faults parallel to the thrusts, the general dip of the strata and the axial planes of the folds. Developments of this type appear to be highly localised.

Immediately north-west of Elgy Rocks, 14 well-defined folds are recorded in 120 m across the strike. The plunge is at 20 to 50° to WSW and the limbs facing SSE are generally a little steeper than those which oppose them. The most north-westerly anticlines of this group are double-crested. For 60 m to the NW of this the strata dip at 60° to NW [NT 9511 6453] and are cut by low-angle faults, both normal and reversed, hading between 50 and 75° to SSE with displacements of 1 to 6 m. Small accommodation faults of this type are common throughout the outcrop but measurement of displacements is in most cases not possible.

On the west side of the inlet known as Ramfauds [NT 950 645] south of Nestends, the beds are very closely folded on NE axes. At least 28 folds occur in 120 m across the strike, in most cases in pairs plunging to WSW, the anticline to NW of the syncline. One or two folds plunge to NE at up to 60°. Good examples of square-shouldered box-folds are seen on the cliff at Nestends [NT 9500 6460]. The most northerly anticline is well rounded at lower levels, and the next one to the south, cut by an undulating low-angle fault, has a very broad crest and could be regarded as a S-facing monocline.

Along the N-facing coast between Nestends and Eyemouth Harbour structures are abnormally irregular. Features include faults and folds trending N to NNE, folds with smoothly curved axial plane traces, small-scale downward-facing folds plunging variously to NE and SW, and, near the western end, minor folds plunging 30° to N 310° which are impressed on a thin NNE sill as well as on the containing strata. However, a basic pattern of NW dips and paired folds, as seen to the south-east, prevails over most of this area. The profusion of small-scale folds and faults continues south on the foreshore east of the harbour.

Between Ayton Mains [NT 935 622] and Eyemouth the Eye Water runs nearly on strike, and most folds plunge at low angles to NE. South of Ayton Mains the strike is generally between E and ENE. Near the new bridge [NT 935 615] inverted strata younging to SE seem to form one limb of an overturned upward-facing anticline which plunges to ENE. Localised N-plunging folds are seen 80 m [NT 9254 6080] and 310 m [NT 9231 6071] above Ayton Bridge. Elsewhere away from the coast exposures are too small and scattered to permit any analysis of the pattern of folding.

Faults

Half the faults dip at more than 60° and only 5 per cent less than 30°. Lateral and vertical displacements appear to be about equally common, as do sinistral and dextral, and normal and reversed faults. By comparison with the Lammermuir outcrop the preferred orientations of sinistral and dextral faults are very poorly defined, but observations suggest dominant trends of N 160° for sinistral and N 140° for dextral faults. The inferred direction of maximum compression, N 150°, is slightly oblique, in a clockwise sense, to that inferred from the orientation of bedding. An analogous relationship is observed in the main outcrop. Vertical displacement was recorded in 29 faults, 13 normal and mostly low-angle, and 16 reversed equally both high- and low-angle.

Axial-plane fractures and small cross-cutting accommodation faults appear to have a common origin with the main folding. Laterally persistent N –S faults, for example at Hurker's Haven [NT 9538 6259], Agate Point, John's Roads, Ramfauds, and north-east of the Eyemouth Lifeboat House [NT 9468 6456], suggest a relationship with the minor intrusions, and with the post-Dinantian Berwick Monocline.

Coldingham Beds

The Coldingham Beds (or Formation, properly speaking) were defined by Shiells and Dearman (1963) as the greywackes, shales and yellow-weathering calcareous siltstones which crop out on the foreshore from Coldingham Bay [NT 917 666] to the mouth of Fleurs Dean [NT 925 655]. Apart from the excellent shore exposures, the beds are also seen in stream exposures which however add little information. On both NW and SE sides, the outcrop is fault-bounded. Total thickness can only be roughly estimated as about 200 m.

The Coldingham Beds have yielded no fossils and are structurally more complex than any others in the district. Shiells and Dearman's (1963) structural analysis, which was confirmed in the resurvey, recognised an additional folding phase not developed in the other Silurian outcrops. This suggested that the beds belonged to a deeper tectonic level and that they were pre-Llandovery, Shiells and Dearman (1966) indicating that they might even be Dalradian. On the geological map they have been classed as ? Ordovician. Recently Casey (1983; and Casey and Oliver, 1984) has shown that soft-sediment deformation affected the Coldingham Beds and that they have a lower level of metamorphism than the strata of the Lammermuir outcrop. These findings also apply to the adjacent Linkim Beds which have recently yielded probable Wenlock microfossils (Molyneux, 1981 and in press); and Casey (1983) proposed that both formations were deposited on a Southern Uplands accretionary prism as trench-slope basin deposits which suffered pre-lithification deformation during gravity sliding down the trench wall. The new evidence for a later rather than earlier age is convincing, but the geological map went to press before it could be incorporated.

Lithology

The Coldingham Beds consist of greywackes, greywacke-siltstones and shales, with many beds of grey, yellow-weathering calcareous siltstone which are a distinguishing feature. The strata are thinly bedded, individual beds being generally 10 to 25 cm thick and rarely up to 1 m. Carbonate is abundant in the matrix of the greywackes, in which the grains are quartz with a very low proportion of feldspar or lithic fragments. Small amounts of muscovite are generally present, and in some specimens the grains show a clear parallel orientation which may be of tectonic origin. Fine-grained sericitic material is abundant and secondary hematite is locally common both in the matrix and, with carbonate and silica, in veins.

Intense veining and shearing combine to obscure sedimentary structures and graded bedding, but the few observations obtainable agree with the younging directions implied by Shiells and Dearman (1963).

Structure

The excellent coastal exposures show abrupt changes of dip and strike with frequent interruptions of bedding and folding, giving a first impression of chaos. Shiells and Dearman (1963) studied individual areas in detail and proposed a tectonic history which includes an early phase of abundant tight or isoclinal folds, formed as open structures under NW–SE compression, and so common as to be effectively a diagnostic feature of the Coldingham Beds. Axial plunge directions and inclinations vary markedly within short distances, the early folds being refolded along their lengths under NE– SW compression. Deformation before lithification was demonstrated by Casey (1983) by ductile pinch and swell structures and boudinage without microscopic strain, intense ductile folding without cleavage, faults with sheared sand and mud showing no microscopic strain or breakage, and mudstone-filled fractures in early carbonate veins.

Folds

Examples of deformation of these beds are given in the following paragraphs.

Cross-folded isoclinal structures are well displayed on the south side of Milldown Point [NT 9190 6635]. Within 2 or 3 m along their lengths, the fold axes plunge variously to W and E at high and low angles, and the flexures in the crests and troughs, very abrupt in some cases, are clearly visible (Shiells and Dearman, 1963, fig. 4). Near HWM [NT 9217 6612] west-south-west of Yellow Craig isoclinally folded beds dipping E exhibit horizontal axes on the south side of a small stack and pronounced S plunges on the north side, some 10 m distant. Axes of refolding can seldom be followed from one isocline to the next, adjacent fold-axes being generally askew to one another, as can be seen west and south-west of Yellow Craig.

The early folds are further deformed by open concentric folds, which are displayed, for example, on the south side of Milldown Point. An example south of Yellow Craig [NT 9290 6525] illustrated by Shiells and Dearman (1963, fig. 6) comprises flat-lying isoclines in which the axial surfaces dip at low angles to E and W, folded about subhorizontal axes which trend nearly N – S, approximately parallel to the axes of the isoclines themselves. Generally though, axes of warping and of isoclinal folding are mutually oblique. Rounded open folds, not directly associated with isoclinal structures, are widely developed. For example, on the headland [NT 9215 6621] west of Yellow Craig a spectacular rounded anticline is disharmonic. It has very steep limbs and plunges at 35° to N 070°, but the beds above and below are only slightly flexed.

Faults

Numerous faults cut the Coldingham Beds, the great majority closely associated with the intense folding, as shears of unknown magnitude nearly parallel to bedding or axial planes. At many localities beds are pinched and squeezed by abundant fold-related shears.

Other faults not clearly linked with the folding generally strike between NNE and ENE. The north-easterly dykes of Linkim Shore and elsewhere may have been emplaced along faults of this group.

Linkim Beds

Shiells and Dearman (1963) defined the Linkim Beds (Formation) as the strata cropping out between the mouth of Fleurs Dean and Callercove Point, and consisting of alternating greywackes and shales with pervasive reddish iron-staining. The strata are much less intensely folded and fractured, and the lateral continuity of beds is much better preserved than in the Coldingham Beds. The total thickness is about 200 m.

Lithology

The Linkim Beds form a succession of evenly bedded massive greywackes, generally 10 to 30 cm thick (rarely exceeding 60 cm), and rather thinner closely-laminated siltstones. Thicknesses of individual strata appear to be well maintained, although in only a few localities may they be followed for more than about 10 m. Sedimentary structures are common, and include flute- and load-casts, graded bedding, and planar, sigmoidal, and ripple-drift cross-lamination.

The reddish brown colour, due to hematite in the matrix, is universal, with local gradations towards grey or purple. A coarse-grained purplish greywacke in thin section proved to consist largely of subangular grains of quartz and grains and fragments of feldspar and fine-grained siliceous and argillaceous rock with flakes of muscovite.

Metamorphism

Casey and Oliver (1984) demonstrated that the metamorphic grade of both the Coldingham and Linkim beds is transitional between the zeolite and prehnite-pumpellyite facies, and is lower than that for the Southern Uplands generally. This supports the idea of a high structural position and relatively late age.

Palaeontology

Shales near Linkim Kip [NT 928 655] have recently yielded fragments of Monograptus sp. and acritarchs identified by S.G. Molyneux (1981) as including ?Dictyotidium stenodictyum Eisenack 1965, Lophosphaeridium sp., Micrhystridium spp., Multiplicisphaeridium cf. picorricum (Cramer, 1964) Lister, 1970, Tylotopalla caelameniculis Loeblich, 1970, T. aff. robustispinosa (Downie, 1959) Eisenack, Cramer and Diez, 1973 and T. wenlockia Doming, 1981. Molyneux interprets this flora, with caution because of its poor preservation, as of probable Wenlock age.

Structure

Sedimentary structures show that most of the Linkim Beds on the shore section and in the Ale Water are inverted. Only near Callercove Point are significant areas uninverted. Dip directions are very variable, and small areas of complex folding not uncommon, but the overall pattern is of a broad antiform with general dips of the order of 30°. As with the Coldingham Beds, soft sediment deformation is pervasive (Casey, 1983) though the Linkim Beds are more coherent. They show isoclinal folds sheared along axial planes with ductile folding; brittle and ductile extensional fractures are seen together; and intrastratal disruption in which sandstone layers have been broken while unlithified (Casey, 1983).

Folds

The following paragraphs give examples of folds typical of the Linkim Beds.

Between 100 and 300 in west of Callercove Point, the strata are closely folded within a zone of eastward dips of 45 to 60°. Four sharp folds are seen [NT 9302 6519] within a cross-strike width of 25 m, the strata dipping ESE at 60 and steeper. Flute-casts show that the most south-easterly fold is a downward-facing syncline, a tight overturned antiform plunging NNE at 18 to 28°. The other folds also appear to face downwards, the most north-westerly plunging at 50 to SSW. About 50 m to north four downward-facing folds are exposed on the shore [NT 9301 6524] within a cross-strike distance of 3 m, the strata and the axial planes dipping steeply E and the hinge-lines plunging to N. Repeated folding of this type is seen on the foreshore to the east the best example at [NT 9313 6523] (Figure 11).

On the western face of Linkim Kip [NT 9282 6545] inverted strata dipping at 30° to NNW are crossed by a fold pair with axial planes dipping at 45 to 50° to NNW. About 100 m W of Linkim Kip [NT 9271 6546], in strata dipping to N and NW at 2(7 to 30° and probably inverted, a pair of folds is developed, Z-shaped in plan, the hinges about 6 m apart. The median limb is vertical and strikes NE, and the more northern fold is an antiform plunging NE at 20°. Between 60 and 75 m E of Linkim Kip [NT 9290 6543] an S-fold in thinly bedded greywackes dipping at 30° to N has a median limb dipping at 35 to N 085°, and axial planes at 35° to NE. (Figure 12). The axial region of the lower fold is somewhat fractured and a thin porphyrite intrusion in the median limb is slightly discordant to the bedding.

The asymmetry of all the minor folds implies that they are developed within the lower inverted limb of a major recumbent anticline, tilted to NE and broadly warped about a NE axis. Downward-facing folds in the E-dipping strata seen farther east show no clear sense of vergence and may lie close to the axial region of the anticline. The most easterly strata are uninverted and they may therefore lie on the upper or NE limb of the major anticline.

Faults in the Linkim Beds fall into two categories broadly similar to those affecting the Coldingham Beds. Syntectonic fractures are developed close to the axial planes of local folds or to the bedding-plane direction in unfolded strata; and a later set of faults cuts across the folds trending NE and hading SE.

Inland exposures

Away from the shore the only significant exposures of the Linkim Beds are in the Ale Water, in a narrow gorge near Alemill [NT 913 635] where thin-bedded purple greywackes, lithologically the same as those on the coast, are cut by a few thick porphyry intrusions. In North Wood the dip is generally 40 to 50° to NNW and locally to SSE, with some evidence of inversion. Near Alemill for about 150 m the beds are flat-lying, undulating and inverted but in the most south-easterly exposures, the dip is steeply to E or SE. Two or three downward-facing folds are seen or inferred between 120 and 50 m above Alemill Bridge and a further downward-facing syncline some 50 m below it. Sedimentary structures clearly show the inversion. The structural pattern is the same as on the shore.

Formation boundaries

An intricate rectilineal boundary between the Coldingham and Linkim formations has been mapped on the foreshore below Bennison's Brae [NT 926 655], which with the difference in tectonics and lithology, confirms that the boundary is a fault. Shiells and Dearman (1963) considered that it dipped S at a low angle, and Casey (1983) believes it to be a thrust, with the Coldingham Formation structurally underlying the Linkim Formation.

Chapter 3 Lower Devonian

Lower Devonian strata are largely volcanic: the andesite and basalt lavas shown on the map include tuffs, and the Lower Old Red Sandstone is composed mainly of locally derived volcanic debris. The presence of these coarse volcaniclastic sediments to east and south of the lava outcrop suggests that they are laterally equivalent to and derived from flows which die out to the south-east and south (see horizontal section), though the possible alternative of a major syncline, with the sandstones generally younger than the lavas, cannot be dismissed.

The main outcrop of Lower Devonian strata extends inland from the coast between St Abb's Head and Eyemouth, and volcanic rocks also crop out at Mordington [NT 95 57]. Because of the extensive drift cover, boundaries can only be fixed approximately. Those that follow fairly straight lines are thought to be faults while more sinuous lines are taken as stratigraphical, as shown on the map. Lower Devonian rocks rest unconformably on a highly irregular topography of Silurian strata.

Abundant and widespread intrusions, having mineralogical and textural affinities with the lavas, do not cut younger rocks, and are therefore assigned to the same igneous cycle which is part of the widespread Caledonian calcalkaline province. Three radiometric ages are available: 408 ± 5 Ma (Rb-Sr isochron) for the Stoneshiel Hill (Cockburn Law) [NT 770 590] diorite (Brown and others, 1985), 400 ± 9 Ma (K-Ar biotite) for the Bell Hill [NT 9171 6807] lamprophyre which cuts Lower Old Red Sandstone (Rock and Rundle, 1986), and 398 ± 13 Ma (Rb-Sr isochron) for the Milldown Point [NT 9193 6643] keratophyre which cuts the Coldingham Beds (Institute of Geological Sciences, 1971, p.57; determined by C.C. Rundle). These ages are close to the Silurian– Devonian boundary, and imply that the Lower Old Red lavas and sandstones in this district may be late Silurian rather than Lower Devonian.

Bell Hill pre-volcanic sediments

Pre-volcanic sediments crop out only at Bell Hill [NT 916 680] where at least 120 m of conglomerates and sandstones (devoid of lava clasts) rest on Silurian greywackes and are overlain by lavas at White Heugh (Figure 13).

Lower Devonian strata are in unfaulted unconformable contact with the underlying Silurian on the sea-cliffs [NT 9176 6803] south-east of Bell Hill, where greywackes are overlain by fine and coarse breccias, pebbly sandstones and flaggy siltstones, dipping at 45° to N. Fine-grained greywacke-conglomerate is seen in an old quarry [NT 9171 6807], adjacent to the lamprophyre dyke. At the northern end of Bell Hill are crags of conglomerate consisting of angular and subrounded pebbles of greywacke up to 30 cm across. Immediately northeast of the St Abb's Head fault coarse greywacke-breccia forms a stack [NT 9180 6797] on the shore; and just to north and north-east along the base of White Heugh [NT 9185 6801] sandstones and siltstones resembling those south-west of the fault are overlain by lava.

Lavas

Petrography

The lavas include olivine-phyric basalts and basaltic andesites and feldspar-phyric andesites some of which include types with a variety of ferromagnesian phenocrysts. In general the rocks are so altered that the distinction between basalt and andesite is likely to be unreliable. Those with a higher proportion of olivine (over 5 per cent) are considered as basalts, while the more felsic rocks trend to dacite. Because quartz is common in amygdales, as veins, locally as pseudomorphs and as a secondary mineral in the matrix, the silica content will be higher than in fresh rocks, and dacite is in fact confined to one outcrop near Eyemouth.

Tuffs interbedded with the lavas contain lava fragments which are generally too altered for petrographic assignment, but where recognisable, the fragments are of the same types as adjacent lavas.

Exceptionally, feldspars are fresh but albitisation and replacement by sericite, kaolin and carbonate are general. Olivine is represented only by alteration products such as chlorite, bowlingite (saponite), carbonate, hematite or quartz. Orthopyroxene is always pseudomorphed by bastite or bowlingite, and clinopyroxene by carbonate, chlorite, etc. Hornblende is altered to chlorite, carbonate, smectite and hematite. Fresh biotite forms small phenocrysts and occurs locally as spongy crystals in the matrix.

The rocks may be grouped according to their phenocrysts into 'nine types.

  1. Olivine-phyric and microphyric basalts and basaltic andesite. The content of olivine pseudomorphs is variable. Exceptionally up to 14 per cent (S54165) is recorded but those with less than 5 per cent are best regarded as olivine-bearing basaltic andesites. Olivine-phyric rocks predominate in the St Abb's Head area, occur as lower flows in the Auchencrow–Coldingham area, and are common at Mordington. The pseudomorphs after olivine are set in a groundmass of commonly flow-aligned tiny laths or microlites of plagioclase with much hematite dust and granules and, in some cases also with carbonate pseudomorphs probably after clinopyroxene grains. The plagioclase is generally albitised or replaced by kaolin (S52751). In rare fresher samples (S49460), (S51843), (S52755), (S54160), (S54163), (S58451), (S63716) the feldspar is zoned from labradorite (An55) to andesine, commonly with a little intersertal alkali feldspar. Amygdales are common. A few rocks approach dolerite in grain-size. In some specimens the original presence of clinopyroxene and/or orthopyroxene is suspected in the matrix. Pseudomorphs after orthopyroxene occur as small grains (S50509), (S50514), (S51817), (S51818) commonly forming a corona to the olivine. In two exceptional specimens (S50514), (S54167) some residual clinopyroxene is preserved. In others (S49460), (S56517), (S51816), (S51817) biotite occurs as late spongy plates in the groundmass or associated with felsic 'segregation' patches or veinlets containing alkali feldspar and quartz.
  2. Olivine-orthopyroxene-phyric rocks There are transitions (S50504), (S52375), (S54167) between the subgroup of olivine-phyric rocks with orthopyroxene in the groundmass and those with orthopyroxene and olivine microphenocrysts (S49457), (S50518), (S51843), (S52756). Spongy biotite may also occur.
  3. Orthopyroxene-phyric rocks The previous group passes into those with microphenocryst pseudornorphs after orthopyroxene alone (S52752), (S58448), (S58449) or accompanied by very little olivine (S58440). Rocks of this group occur sparsely in the Mordington area and in the Auchencrow-Coldingham area where they may represent a flow above the basal olivine-phyric lavas.
  4. Olivine-(?)clinopyroxene-phyric rocks In a specimen (S50542) from Ale Water the olivine pseudomorphs are accompanied by smaller carbonate pseudornorphs which may represent clinopyroxene rather than orthopyroxene.
  5. Plagioclase-olivine-phyric rocks Only a few have been noted, all from the Ale Water exposures. Three (S50547), (S50549), (S51860) are extremely decomposed but one sample (S51875) is fresher and contains numerous phenocrysts of fresh plagioclase zoned from labradorite (An55 at core) to andesine, and sparse large pseudo-morphs after olivine. Microphenocrysts of somewhat resorbed biotite occur and some carbonate may be after clinopyroxene.
  6. Plagioclase-pyroxene-phyric rocks A number of andesites occur containing phenocrysts (up to about 3 mm long) and microphenocrysts (generally less than 1 mm long) of plagioclase with microphenocryst pseudomorphs of orthopyroxene with or without clinopyroxene (S50555) - (S50556), (S50565), (S50575), (S51890). This type forms most of the outcrops on the coast at Eyemouth and the more south-easterly lavas in Ale Water, as well as the lowest flow exposed at Mordington.
  7. Plagioclase-hornblende-phyric and plagioclase-hornblende-pyroxenephyric types The first of these types (S50557), (S50559), (S63719), (S63720) is characterised by microphenocrysts up to 1.4 min long, of argillised plagioclase, with pseudomorphs in chlorite, quartz and carbonate after cathedral hornblende, as phenocrysts and as smaller crystals in a fine-grained rather leucocratic matrix. In the Auchencrow-Coldingham area one of the higher flows is of this type (S63719), (S63720) and a stratigraphically higher tuff (S51842) contains fragments of this type. A lava (S50557), (S50559) in the Ale Water and a tuff (S54159) at Mordington are also of this type. The plagioclase-hornblende-pyroxene-phyric types are represented by single flows in the Ale Water (S50560), (S51882), (S51899), (S51900) and Mordington (S54186) (S54187), (S54191), (S55969) sequences. The lavas contain phenocrysts of hornblende and plagioclase, and smaller pseudomorphs after orthopyroxene (S54186), (S54187). Phenocrysts of biotite occur in two rocks (S54191), (S55969).
  8. Biotite-hornblende-quartz-feldspar-phyric dacite or rhyodacite This type (S50576), (S51891) occurs only at the base of the cliff [NT 9432 6478] south of Eyemouth Fort. The rock contains phenocrysts and microphenocrysts of fresh and altered biotite, completely argillised feldspar, quartz, and smaller pseudomorphs after hornblende. These are set with smaller crystals of the same minerals in a cryptocrystalline matrix much altered to clay minerals and quartz. The precise identification of the rock as dacite is uncertain because of the complete alteration of all feldspar.
  9. Biotite-feldspar-phyric rock An extremely altered volcanic rock (S54175) of a type unique in the district forms a high flow at the northern end of the Mordington sequence. Microphenocryst pseudomorphs, in chlorite and hematite, apparently after biotite, with sparse phenocrysts of carbonated feldspar, occur in a matrix containing flow-aligned small pseudomorphs, probably after biotite, and microlites of albitic plagioclase with much disseminated carbonate and chlorite.

St Abb's Head

Lavas at St Abb's Head are mostly olivine-phyric or microphyric basalts and andesites, some of which contain pseudomorphs after orthopyroxene. Olivine contents range up to 12 per cent (S51795). The lowest flows, totalling at least 100 m in thickness, are red, finely-vesicular, generally flow-brecciated, thinly-bedded and largely composed of tabular lava fragments. These flows crop out between Horsecastle Bay Fault and White Heugh (Figure 13), where they directly overlie the Bell Hill sediments and form a sequence younging generally to NW. A younger sequence of lavas and tuffs, some 500 m thick and younging to SE crops out between Pettico Wick and Horsecastle Bay. These lavas are purplish, fine-grained olivine- and olive-pyroxene-phyric with flow-aligned plagioclase microphenocrysts in the matrix. Upper and lower parts of flows are vesicular. Flows are up to 50 m thick and interbedded tuffs up to 60 m thick.

White Heugh to Horsecastle Bay

White Heugh is the apt name given to the 40-in high SE face of a narrow promontory [NT 9190 6803] composed of lava, overlying the Bell Hill sediments. Lava of similar type (S50514) forms the cliffs of Hardencarrs Heugh [NT 919 681], beyond the narrow inlet north of White Heugh which may mark an ENE fault or the outcrop of a relatively soft tuff. The St Abb's Head Fault is exposed at the cliff top [NT 9176 6803] at the head of the inlet. The face of Hardencarrs Heugh appears to be the top of a lava-flow dipping at 45° to N 120°, the overlying rock, seen on the north-east continuation of the cliff, being a highly vesicular tuff or auto-brecciated lava. In this area the lavas are traversed by closely spaced steep joints which trend NE, SE and locally N. Northwards to Horsecastle Bay [NT 918 685] the rocks are red, finely vesicular and commonly brecciated. Locally they are stratified, and largely composed of tabular fragments, tens of centimetres in size, giving the general appearance of a flow-brecciated, thinly bedded, blocky lava (Plate 8). Massive, unbroken, fine-grained lava (S50517)-(S50518) has a more limited occurrence, notably on the north side [NT 9200 6839] of Burnmouth Harbour, (the little bay north of Wuddy Rocks), and capping the cliffs of Horse Castle where flows up to 5 m thick are recorded. In general the dip is 15° to ESE but tuffs below the Burnmouth Harbour lava dip at 40° to NE and the lava itself has an undulating base dipping locally to NW. Close-jointing, calcite-veining, and faulting are common, most frequently in planes dipping very steeply to W and SE, and locally at between 30 and 60° to NW. Horsecastle Bay is eroded on a NE fault downthrowing NW over 500 m, which limits the outcrops of the red, vesicular tuffs and tabular blocky lavas, which are hardly represented in the younger sequence north-west of the fault.

Pettico Wick to Horsecastle Bay

The oldest lavas of this north-west sequence crop out in the northern part of St Abb's Head [NT 910 693] from where there is an upward succession of flows to SE (Figure 13). Most are fine-grained andesite or basalt generally purple but locally with a blue, grey or red tinge, with minute red or brown phenocrysts which are ferruginous pseudomorphs after olivine, pyroxene or amphibole. A glassy groundmass with abundant tiny prisms of feldspar, shows a flow pattern and a characteristic felted or ‘andesitic’ texture. Many flows, especially their upper and lower parts, are vesicular, some highly so, the amygdales most commonly siliceous or chloride and in some cases with carbonate. Central parts of the lava flows are finer-grained, less vesicular and less weathered than marginal parts. At several localities ramifying irregular bodies of bright red rock occur within the duller purple lava. Under the microscope the red rock is seen to be brecciated but otherwise similar to the adjacent lava, and its ramifying form suggests that it may represent the brecciation of lava in site by intrusion of magmatic gases. The differences between the central and marginal parts of the flows led Geikie (1887) to describe the lavas south of Cauldron Cove [NT 916 689] as intrusive, but all are now interpreted as extrusive.

Within 100 m of the St Abb's Head Fault at Pettico Wick [NT 9077 6918] coarse bedded agglomerate 5 m thick is seen at two localities near high water mark, separated by one NW fault and limited to NE by another. In cliff exposures separate lava flows are indicated by brick-red bands, 1 to 2 m thick, which form the tops of the flows. Outcrops of more easily weathered inter-flow tuffs form hollows in a dip and scarp topography. The lowest mappable tuff can be followed from Mire Loch [NT 9120 6886] to the cliff-top [NT 9127 6928] 170 m west of the lighthouse, from where the dip slope of the top of the underlying lava descends south-east to the foot of the cliff [NT 9141 6928], where lithic tuff less than 1 m thick lies between two lava-flows. Small exposures of tuffaceous sandstone occur here and there by the lighthouse road, overlying vesicular lava at the sharp bend [NT 9110 6902] and underlying lava crags to the east. The top of this overlying flow is defined by a bright red band visible on the cliff north of the lighthouse, the tuff above being traceable south from the cliff-top [NT 9134 6924] along a drift-filled depression. An old quarry [NT 9129 6910] shows 1 m of flaggy, finely conglomeratic tuff, dipping at 34° to N 130° and its full thickness here is estimated at 5 to 6 m. The lava beneath, on which the lighthouse stands, is 40 to 50 m thick. Small faults cut the cliffs of St Abb's Head and are responsible for the alignment of gullies and stacks such as Cleaver Rock.

From the lighthouse to Cauldron Cove the cliffs show a succession of SE dip-slopes eroded on tuffs between lava-flows. Though details are somewhat obscured by NE-SW faulting, three lava-flows with intervening tuffs can be traced south to the boundary fault. The lowest flow is that on which the lighthouse stands. Overlying tuff is exposed to a thickness of 2 m low on the cliff [NT 9155 6913] west of Clafferts Rock, near the foot of an extensive dip-slope, and to 5 m by the path [NT 9133 6887] 390 m south-south-west of the lighthouse. Along the shore of Mire Loch [NT 9130 6874], is a 30 m section of coarse bedded tuffs whose total thickness is about 60 m. These beds dip at 37° to N135°.

The succeeding lava, about 15 m thick, is exposed beneath tuff 2 m thick on the seacliff [NT 9153 6909] north-west of Harelaw Cove, dipping at 60° to N 125°. Cliffs on the south side of Harelaw Cove, terminating on the east at Black Gable [NT 9167 6905], are formed by the next lava in upward succession (S51816), (S51817), (S51818), 35 m thick, finely porphyritic and purple. Ramifying brick-red breccia attributed to gas-flow in this lava is exposed south-west of Black Gable [NT 9164 6900].

At the head of Cauldron Cove [NT 9159 6889] tuffs and lava are thrust 20 m NW over the bedded tuffs at the foot of the dip-slope of the Black Gable flow. The thrust-plane lies close to the stratification but steepens downwards. On the south side of Cauldron Cove, lava in the lower cliff is overlain by 27 m of tuff, the lower part of which is thick-bedded (beds up to 25 cm thick) and the rest thin-bedded (about 5 cm). The depression running south-west from Cauldron Cove is eroded along the outcrops of these tuffs and of several NE-trending faults. At the south-east end of Mire Loch, 8 m of well bedded pebbly tuffs are exposed in the upper part of an old quarry [NT 9147 6857], and are overlain by 5 m of lava.

Above the Cauldron Cove tuffs, the 60 m high cliff of Batty's Heugh [NT 9165 6885] consists of lavas at least 50 m thick, in places somewhat brecciated and with irregular red areas which may indicate separate flows. Lava over 50 m thick is exposed on the SW face of Kirk Hill, [NT 9160 6865], where there are ovoid amygdales of chalcedony up to 5 cm long, in some cases in the form of agate. Cliff faces east of Kirk Hill are defined by joints and faults.

Pebbly tuffs crop out on the western shore [NT 9183 6857] of Horsecastle Bay, where they are 27 m thick, in beds between a few centimetres and nearly 2 m thick, dipping at 13 to 27° to NE and ENE. In their coarsest lithologies, some 8 m below the top, they contain many large well-rounded fragments of lava. These tuffs are correlated with those of Cauldron Cove and those south of Waimie Carr, but faulting makes precise correlation doubtful.

The lava beneath the Horsecastle Bay tuffs is exposed in the SW corner of the bay [NT 9180 6851] where there appears to be an upward transition from compact lava, through scoriaceous and fragmented lava, to tuff becoming progressively more heterogeneous in its pebble content. The top of the tuffs and the basal 15 m of the overlying lava form the west and south faces of East Hurker [NT 9185 6865]. The topmost metre or so of the tuff is locally unusually reddened, and the base of the lava, which is amygdaloidal and includes many agates, is irregular and cuts down NE into the tuff.

Auchencrow to Coldingham

This outcrop is not well exposed, the best sections being in streams west of Auchencrow, in the Eye valley west of Reston, and around Coldingham. While most exposures are of lava, interbedded tuffs and tuffaceous sediments are also seen. The succession youngs to SE, but evidence is too sparse to map individual flows.

The lowest exposed flows are olivine-phyric or micro-phyric rocks similar to those at St Abb's Head (S49458), (S52751), sometimes with orthopyroxene (S49457), (S49459), (S49460), (S50539), (S52375), (S52755), (S52756). Biotite occurs locally as small spongy plates in the matrix (S49460), (S52375), (S52756). Higher flows include pyroxene-phyric and hornblende-plagioclase-phyric (S63715) types.

The most westerly exposures are in Fosterland Burn 240 m NW of the road [NT 8280 6068], where purple fine-grained olivine-basalt or andesite is seen (S52751), 30 m downstream of greywacke. About 60 m SE of the lava fine-grained tuff dips at 20° to SSE, and 140 m below the road-bridge is a further small exposure of tuff. The base of the Upper Old Red Sandstone hereabouts was mapped on very slight evidence, and the Crossgate water bore [NT 8233 5986] drilled in 1979 suggests that the base lies about 400 m farther south than shown.

North-west of Greenburn, finely porphyritic lava, has been quarried [NT 8363 6125] beside the burn 570 m above the bridge. Rock in the stream nearby appears to be a welded tuff, and fragmental tuff is seen in the next exposures, 90 and 130 m downstream. Amygdaloidal lava some 20 m thick succeeds these tuffs downstream and is overlain by 20 m of bedded pebbly tuffs dipping at 50° to N 170° [NT 8379 6112]. Fine-grained orthopyroxene-phyric andesite (S52752) is exposed downstream, in crags on the left bank and in an old quarry [NT 8385 6100]. Finally massive sandy tuff is seen 150 m above the bridge. In the next stream to the north-east, exposures of fine-grained lava occur down to 450 m N of Coldlands [NT 8438 6149], the rock typically carrying small altered ferromagnesian phenocrysts, some being pseudomorphs after olivine. No tuffs were observed.

Evidence for the N – S fault near here is in the Eye Water 70 m W of Howburn Bridge [NT 8532 6245] where veined and brecciated lava (S63719), (S63720) is exposed to within 8 m of veined and contorted greywacke. Also, greywackes in the stream [NT 8533 6276] just south of the main road crop out close to lavas and tuffs upstream near Houndwood House where they are cut by steep N – S joints.

Between Howburn and Coveyheugh [NT 869 624] fine-grained purple lava seen in the Eye Water is in many exposures characterised by prominent small phenocrysts of feldspar, in addition to the more commonly occurring altered ferromagnesian minerals; flow-banding is also locally developed. Similar rock is exposed in the railway cutting [NT 8679 6244] at Coveyheugh, where it is locally tuffaceous or brecciated.

At Lemington [NT 861 631] fine-grained locally feldspar-phyric olivine-bearing purple lavas (S58440) are exposed in a 200 m gorge, and similar rocks (S50539), (S51843) are seen in the stream 200 m N of the farm. Lava with small feldspar phenocrysts is exposed in a ditch [NT 8693 6345] south-east of Hillend. About 60 m uphill to the north purple micaceous flaggy sandstone dips anomalously at 35° to NNW. Similar sandstones, not thin-sectioned, are seen [NT 8870 6306] south-west of Blackhill, and near the farm several crags [NT 8878 6345] are formed of a hornblende-andesite breccia (S51842). Fine-grained basalt with olivine phenocrysts (S49460), is exposed on Gallows Law [NT 8764 6432] and in a ditch 320 m to the south-west as well as in an old quarry 250 m to NE. One kilometre to the north-east on Greenlaw Knowe [NT 885 649], fine-grained lava with olivine and orthopyroxene phenocrysts, (S49457), (S49458), (S49459), locally highly amygdaloidal, has been quarried from several prominent crags.

Coldingham Law [NT 907 658] is a prominent crag-and-tail feature, elongated ESE, with many exposures of fine-grained amygdaloidal andesite, with small phenocrysts of feldspar and ferromagnesian minerals. Flow-banding is seen in places. In Milldown Burn soft, red, finely porphyritic (hornblende and/or pyroxene) andesite is exposed for 260 m upstream from the faulted junction with the Coldingham Beds [NT 9105 6615].

Ale Water and Eyemouth

In this area lavas with interbedded tuffs more often than not appear to young to NE, though faulting, folding and limited exposure preclude establishing a general sequence. The lavas show a preponderance of plagioclase-pyroxene-phyric basalts, some with olivine or hornblende pseudomorph phenocrysts. On the coast south of Eyemouth Fort [NT 9432 6478] the lavas are altered dacite (S50576), (S51891), unique in this district. There are good exposures only in the Ale Water and on the sea cliffs.

Below Alemill Bridge [NT 916 634], immediately downstream from Linkim Beds greywackes and faulted against them, is a 25-m thick conglomerate packed with small rounded pebbles of greywacke and including a thin bed of fine-grained sandstone which lies vertically on a strike of N 055°. The conglomerate is thought to be a basal bed to the lava sequence, and the displacement on the fault to be small and merely a response under stress to the lithological contrast at an original unconformity. Similar relationships are seen on the coast at Callercove Point [NT 9316 6524] (see below). Downstream the conglomerate is succeeded by some 200 m of olivine-orthopyroxene and olivine-plagioclase basalts (S50547), (S51860), (S58442), which include three interbedded sandstones and finely pebbly tuffs. The thickest flow is 90 m thick, the thickest tuff about 30 m. The tuffs include subangular fragments of finely porphyritic pale red-brown lava, up to about 1 cm across, in a veined matrix of fine-grained purple lava which is indistinguishable from the immediately underlying olivine-plagioclase-hornblende basalt (S50549). In many places this forms an autobrecciated fine-grained agglomerate. In this section and in others downstream there are many changes of dip and strike, with faulting, so that while the rocks generally young downstream, the succession is obscure.

For the next 500 m downstream lavas predominate, brecciated in places. They include tuffs at a point [NT 9215 6303] 710m below Alemill Bridge, which appear to be vertical on a strike of N 095°. Most of the lavas upstream of the tuff are feldspar-phyric, commonly with olivine, and in one case with phenocrysts of partly resorbed biotite (S51875), while those downstream carry hornblende (S50557), (S50558), (S50559). Pyroxenes or their pseudomorphs occur in most of the rocks of both groups. North of Little Dean at a sharp left-hand bend [NT 9215 6275] the Ale Water enters a short gorge in plagioclase-clinopyroxene-phyric lava (S50561), (S51874), which includes tuffaceous bands. Lavas downstream (S51873) are of the same type but also include hornblende-bearing types (S50560) and become conspicuously amygdaloidal. At the western end of Linthill Haugh [NT 9235 6270] amygdaloidal olivine-basalt or andesite (S51870) is coarsely brecciated in places.

Lava and tuffs are exposed in a 300 m gorge south of Linthill and past the mouth of Little Dean Burn [NT 9279 6264]. Downstream from the confluence bedded pebbly tuffs are succeeded by plagioclase-pyroxene-basalts (S50555) with olivine. The most easterly exposure of the lava sequence is at the sharp left-hand bend [NT 9301 6264] 735 m W of Old Linthill. An usually fresh specimen (S50554) of feldspar-phyric trachybasalt was obtained from this locality, and similar but more altered rocks occur nearby to the west. These distinctive rocks could be intrusive porphyrites.

There are small quarries in fine-grained purple lava in the grounds of Highlaws [NT 9352 6367] and north of Linthill (S51882) [NT 9259 6356]. Small exposures of lava occur east of Hallydown and on the by-pass 300 m NE of Highlaws [NT 9385 6100].

Lavas exposed east and west of Hairy Ness are mostly plagioclase-pyroxene-andesites (S50565) - (S50566), (S50572), (S50575). Many are autobrecciated. The more northerly exposures, around Hairy Ness, are locally flow-banded, and contain phenocrysts of hornblende as well as of feldspar and pyroxene (S51899), (S51900).

Rocks exposed on the coast [NT 9432 6478] immediately south of the Upper Old Red Sandstone outlier at Eyemouth Fort are pale greyish purple dacite with phenocrysts of feldspar, hornblende, and biotite, and a well developed flow structure (S50576), (S51891). These are the most acid of all the Lower Devonian lavas within the district. Most of the exposed dacite is autobrecciated, consisting of an accumulation of clearly defined fragments, usually many centimetres in size, showing an irregular flow pattern, with no significant matrix of different material. Spherical knots of malachite are locally common, up to about 5 cm in diameter, surrounded by a bleached aureole about 5 cm wide. To the south [NT 9429 6464] the lavas are brecciated pyroxene-andesite (S51890) while to the north there are further exposures of purple pyroxene-andesite, (S51887) well jointed and calcite-veined. Similar rock (S50575), partly brecciated, forms a narrow faulted inlier on the cliff [NT 9435 6495] 60 in to NW, across the headland and is exposed on the foreshore towards Hairy Ness.

Hairy Ness is formed largely of fine-grained hornblende-andesite (S51899) but there is a roughly rectangular enclave of brecciated lava, some 200 m2, on the foreshore [NT 9405 6503] 100 m SW of the headland. A short distance to the north-west [NT 9403 6507], lava (S51900) shows prominent flow banding which dips at 40° to NNW.

South-west of the narrow rectangular inlet known as Weasel Loch [NT 9395 6503] coarsely brecciated lavas are interbedded with unbroken lavas in beds over 3 m thick. Dominant fractures are steep and NE-trending; others trend ENE. Fine-grained pyroclastic rocks are only seen [NT 9398 6504] 30 m E of Weasel Loch, where a 15-cm lens of shaly tuff lies within lava.

Mordington

Faulted boundaries against Silurian to the north and east are inferred from indirect evidence. There is a large disused quarry in lava [NT 938 563] near Blinkbonny.

The general dip is to SE, and petrographic variations indicate a succession of four lava-flows with two additional flows intervening in the NE corner between the third and fourth. Coarse-grained tuffs separate the second and third flows forming an outcrop, up to 300 m wide, striking N 035° and passing midway between Deans Hill and Mordington Mains. Pyroclastic rocks were not observed to the north-west and are not common to the south-east.

The lowest flow sampled is a plagioclase-orthopyroxene-phyric andesite (S54162), which is succeeded to the south-east by flows of olivine-microphyric basaltic andesite with olivine contents 5 to 8 per cent (S54157), (S54160), (S55968). South-eastwards again is a tuff (S54159) containing fragments of plagioclase-hornblende-biotitephyric andesite, and then a flow of andesite with phenocrysts of plagioclase, microphenocryst pseudomorphs after hornblende and smaller pseudomorphs after orthopyroxene (S54186), (S54187), (S54191), (S55969). Microphenocrysts of fresh biotite also occur in the last slice. In the north-east of the area the next higher flow is an orthopyroxene-microphyric andesite (S58448), (S58449), succeeded by an unusual fine-grained leucocratic chloritised and carbonated lava (S54175) with flow-aligned pseudomorphs after biotite and flow-aligned microlites of albite. This unique rock is exposed in a crag [NT 9588 5786] 1100 m SSW of Lamberton Shields. Farther south, the higher flows are olivine-microphyric basalts containing tiny pseudo-morphs after orthopyroxene (S54163), (S54165) and, in one slice (S54167), larger bowlingite (saponite) pseudomorphs after orthopyroxene.

The pyroclastic rocks include finely pebbly tuffs with sparse matrix, tuffaceous sandstones with scattered small pebbles, and agglomerates resembling the tuffs but more coarsely pebbly. The pebbles are nearly all of hornblende-andesite lavas with a general increase in homogeneity with increase in size.

The inlier in North Crowlea Plantation [NT 9252 5692] shows olivine-basalt (S63716) containing small pseudomorphs after both ortho- and clinopyroxene.

Lower Old Red Sandstone

Reston and Ayton

The southern part of the main Lower Devonian outcrop, in the Eye valley from Reston to Ayton, is mapped as Lower Old Red Sandstone, and is composed mainly of sandstones derived from the erosion of basalts and andesites like those seen at outcrop in adjacent ground to the north-west and north. The characteristic lithology is a reddish brown coarse-grained sandstone with abundant small (up to 5 mm) sub-rounded fragments of fine-grained lava of a variety of yellow, grey and red colours. Tuffs, siltstones and mud-stones also occur and locally there are coarser conglomerates or agglomerates. Cornstone is seen in one exposure. Away from the Eye and Ale Waters, exposures are sparse. A fragment of Pteggotus sp. from Auchencrow Burn [NT 86 59] confirms the Lower Devonian age.

In the Ale Water for about 1 km E of Ale Water Bridge [NT 892 645], there are good exposures of coarse-grained tuffs, with some beds of agglomerate, tuffaceous sandstone and shaly tuff. In the cliff [NT 8971 6420] 580 m below the bridge the interbedded coarse-grained and medium-grained tuffs include veins and concretions of calcite. Several tuffs contain carbonate either as small fragments, segregations, or cement. The strata are disposed in a broad anticline, plunging south, but dip and strike are very variable. About 150 m of strata are exposed in the eastern limb, rather more than in the western. Some 1060 m east of Alewater Bridge amygdaloidal lavas and sandy tuffs are separated from greywackes of the Coldingham Beds by a gap of 6 m marking the line of the boundary fault.

In the Eye Water 125 m east of Coveyheugh [NT 8700 6251], coarse-grained tuff carries rounded igneous pebbles up to 35 cm in diameter. Similar conglomerate on the opposite bank and in the nearby railway cutting is interbedded with fine-grained tuff. Agglomerate is exposed beside a footbridge [NT 8803 6221] 200 m NE of Reston church, and just below the road-bridge [NT 8862 6229] micaceous flaggy sandstone and siltstone, calcareous in part and with calcareous concretions, dips at 30° to N 070°. Close by, the Reston No. 1 Water Bore [NT 8860 6239], proved sandstones to a depth of 50 m. The topmost 28 m are mainly coarse-grained tuffaceous and pebbly rock in beds around 2 m thick, but lower strata are mostly fine-grained and grey, with many carbonised plant fragments, quite unlike any rock exposed nearby.

South-east of Swinwood Mill tuffaceous rocks intermittently exposed beside the Eye Water, [NT 8917 6187] to [NT 8923 6198], are unusually calcareous and include nodular cornstone and cornstone-conglomerate. Thin section shows the cornstones to be calcite-mudstones fissured or broken into small rounded or mamillate globules separated by calcareous siltstone, both components being somewhat feldspathic. A 5-cm band of cornstone from the north-east end of the exposures includes abundant carbonised plant fragments. Most of the rocks here are well bedded, pebbly, coarse and fine-grained sandstones and mudstones, all to some extent tuffaceous and calcareous. The cornstones appear to be of pedogenic origin.

There are further exposures of tuffaceous sediments in the Eye Water for nearly 2 km below Reston Bridge. Pebbly beds are more common east of the dyke [NT 8966 6217], tuffaceous sandstones being predominant to the west. Some 1600 m below the bridge [NT 9017 6196] the rock contains igneous pebbles up to 10 cm across. These exposures show an anticline plunging south, perhaps continuous with that in the Ale Water. There is a complementary syncline to the west between the cornstones and Reston Bridge.

Downstream, coarse agglomeratic tuffs some 430 m thick are exposed south-west of Aytonlaw Farm for 650 m downstream from [NT 9065 6109]. They have been quarried hereabouts. Red tuffaceous mudstone is seen on the left bank [NT 9094 6095]. In Horn Burn [NT 9142 6021], 400 m W of its confluence with the Eye, coarse pebbly tuffs dip at 30° to N 130°.

North-east of the Eye Water there are few scattered exposures. Tuffaceous sandstone in the stream [NT 9002 6247] 500 m N of East Reston Mill dips at 35° to N 080°. Fine pebbly tuff is seen in ditches north-east of Aytonlaw [NT 9183 6138] and north of the main road 300 m to NW. Some 10 m of coarse and fine bedded tuff is exposed in an old quarry [NT 9175 6227] 200 m NW of Aytonwood House. In the gorge of Little Dean Burn, 780 m to E [NT 9251 6237] coarse-grained tuffs dip at about 40° to N 220°. Both west and east of the tuff are extensive exposures of agglomerate, with rounded igneous pebbles up to 30 cm in diameter, which is considered to lie close above the top of the main Lower Devonian lava sequence.

Tuffaceous coarse-grained and pebbly sandstone is seen in places on the bypass north of Ayton. Coarse-grained tuff and agglomerate are exposed for about 230 m in the railway cutting [NT 913 600] northwest of Prenderguest, dipping E at 30°. At the foot of the western slope of Millerton Hill [NT 9059 5955] an exposure of agglomerate includes lava pebbles up to 15 cm across, and for 280 m to NNE pebbly and sandy tuffs are exposed at many points, dipping eastwards at 25 to 30°. South-west of Millerton Hill the boundary between the pebbly lithologies of the Lower and Upper Old Red Sandstone is determined on soil evidence, but with difficulty because of the common occurrence in the younger formation of reworked material from the older.

Among the buildings at Restonhill [NT 896 607] pebbly tuffs are seen in places dipping ESE at 20°. Restonhill Water Bore, [NT 8950 6041] 250 m S of the farm, proved 32 m of sandy and conglomeratic tuffs, with lava pebbles up to 5 cm across. The Silurian inlier at Stoneshiel [NT 871 603] is mapped on the evidence of rock fragments, and there may be other small outcrops 500 m to S and 600 m to SE of the house.

Thinly bedded tuffaceous sandstones and mudstones are exposed in Auchencrow Burn downstream from the railway bridge [NT 8618 6022]. Immediately south of the Chirnside road thin beds of mudstone show sand-filled desiccation-cracks in the lower part of a well-bedded 10-m section. Higher sandstones and siltstones a few metres downstream [NT 8642 5980] contain indeterminable plant stems. The strata in all these exposures dip at between 35 and 55° to between N 110 and N 130°. It was from them that a fragment of Pteggotus sp. was collected during the primary survey (Geikie, 1863, p.27).

Vents

St Abbs agglomerate

St Abbs stands on a volcanic vent recognised as such by the chaotic distribution of igneous and pyroclastic rocks on the well-exposed shore and cliffs, and on the extensive occurrence of very coarse agglomerates unlike any rocks of the bedded sequence of St Abb's Head. No sequence can be deduced within the vent. Some rocks appear to be of direct magmatic origin and so are considered to be intrusive; others are pyroclastic, mainly agglomerates, but in places include finer-grained pebbly tuffs in which bedding is confidently recognised. The different rock types are irregularly interjacent.

The andesites are all highly altered, but most have the textural character of lavas and contain small plagioclase phenocrysts, which differentiates them from the lavas of St Abb's Head. It cannot be determined whether andesites in the vent are to be regarded as lavas contemporaneous with the agglomerates and subsequently collapsed into the vent, or as intrusives emplaced within them at a later stage of igneous activity, as Geikie (1887) thought.

The north-west limit of the vent is marked on the shore northwest of Ship Rock [NT 9169 6771] by a 100 m wide zone of brecciated and veined Silurian siltstone with some lava. This marginal zone is considered to be the fractured wall of the vent, and its preservation suggests that little faulting has occurred here since igneous activity ceased. The veins, which in places follow the general north-east grain, consist mainly of silica, carbonate and hematite. Hematite veins are common also in the adjacent greywackes.

At and south-east of Ship Rock are exposures of porphyrite, brecciated fine-grained andesites and olivine-basalts, tuffs and agglomerates. Very coarse agglomerate is exposed at Black Craighead [NT 9184 6753] with lava blocks up to 1.8 by 0.6 m. Coarse agglomerate occurs eastwards to beyond the harbour entrance and south into the west harbour, but on the south side of this harbour and the east side of the outer harbour the rock is a fine-grained hornblende-pyroxene-andesite with small feldspar phenocrysts.

Between the harbour and The Kip [NT 9187 6683] coarse agglomerate with subangular fine-grained andesite or porphyrite clasts has intricate boundaries against fine-grained intrusives which predominate farther south. Finer-grained tuffs are seen in places. The prominent stack of The Kip is composed of intrusive rock overlying agglomerate. South of The Kip only small areas of tuff or agglomerate are seen within extensive foreshore exposures of fine-grained igneous rock, mainly pyroxene-andesite with phenocrysts of feldspar and hornblende or olivine in some cases.

Inland exposures of the vent-rocks are few and small, the best being round St Abbs Church [NT 9173 6737], in small crags of finely porphyritic lava-like rock, and in the stream [NT 9142 6761] 400 m to NW where heavily veined greywacke abuts against tuffaceous sandstone and brecciated lava, resembling rock on the shore north-west of Ship Rock. The vent boundary appears to be steeply inclined to NNW.

Callercove Point (Eyemouth volcanic vent)

Vent rocks are exposed only on the cliffs and foreshore south-eastwards from Callercove Point. Like the St Abbs vent they comprise an assortment of igneous and fragmental rocks of intermediate to basic composition within which, to an even greater degree than at St Abbs, no systematic pattern of distribution has been discerned.

The igneous rocks are all very altered but such mineral and textural evidence as survives indicates that they were originally andesites (S51878) or porphyrites, or possibly basalts in a few cases. Traces of the felted andesite texture are common as are feldspar phenocrysts or their pseudomorphs and in these respects the rocks resemble the lavas which abut against them to south-east, at an ill-defined boundary. The vent andesites may be intrusive or extrusive or completely fragmented xenoliths formed at a different level.

Agglomerates which occur widely in the vent are composed of discrete blocks of lava varying in coarseness upwards from finely pebbly, and are generally more readily eroded than the andesites which stand out above them on the shore.

The Silurian –Devonian boundary is best exposed at the northwest corner of Callercove Point, on a N-facing cliff 8 m high just above HWM [NT 9316 6524], where greywackes of the Linkim Beds dipping steeply E abut against vertical Devonian greywacke-conglomerate at a steep surface which is concave to W (Figure 14). Against the lower, W-sloping part of the boundary-surface the greywackes are dragged down, as if they had moved upward along the surface over the conglomerate below. The boundary, complicated by an andesite intrusion of the type seen in the vent, is intermittently exposed over a distance of some 200 m but nowhere else is it so clearly displayed. In a gully in the cliff [NT 9307 6515] 190 m SW of Callercove Point there appears to be an intermittent transition from greywacke to conglomerate, implying that the conglomerate rests on a reconstituted layer of variably fragmented greywacke. On the foreshore the boundary is displaced dextrally by two ENE faults. The junction between the two formations in these exposures is therefore interpreted as essentially an unconformity, rotated into a near-vertical position and in that process acting as a preferred surface for the development of a fault. A small outlier [NT 9314 6523] of conglomerate and andesite lies within greywackes 15m NW of the main boundary.

At a minor headland [NT 9325 6523] on the south side of Caller Cove is an outcrop 20 m (N – S) by 12 m in extent of regularly-bedded flaggy 'tuff, composed of fragmented greywacke, siltstonc and mudstone, possibly a reconstituted xenolith of Silurian country rock fragmented initially by volcanic gas, a 'non-igneous tuff'. Similar rocks form several small outcrops farther south-east.

Intrusive rocks

Minor intrusions, petrographically associated with the lavas, cut the Silurian rocks and are numerous near the coast and around Stoneshiel Hill in the west. A few dykes also cut the Lower Devonian.

Granitic rocks

Grey 'granite' and grey and black 'diorite' are exposed in places on Stoneshiel Hill [NT 775 580] and on the south-east slopes of Cockburn Law (Sheet 33E) north-west of the Whiteadder. The intrusion is dated at 408 ± 5 Ma (Brown and others, 1985). Silurian rocks adjacent to the intrusion are hornfelsed; and the adjacent Upper Old Red Sandstone includes pebbles of both hornfels and granite.

Walker (1925, 1928) and Midgley (1946) described the succession of consolidation, from marginal hybrid hypersthene-diorite, through granophyric microgranite and granite to a central adamellite. Exposures are poor, so that the boundaries between the various types are somewhat speculative. Being more resistant the marginal rocks form a topographic contrast at their contact with the country rock. The straight south-western margin of the intrusion may be faulted. The zone of hornfelsed greywacke is much wider north-east of the intrusion, extending at least 1300 m to the vicinity of Elba. In this direction the roof of the intrusion probably declines at a low angle. Since the resurvey, a quarry (now disused), [NT 7751 5935] 400 m SW of Cockburn East, was opened in quartz-diorite (S75513), (S75514), (S75515), (S75516), (S75517), (S75518). A group of small exposures, around [NT 7716 5904] 800 m SW of Cockburn East and on the west side of the river, show several rock types from adamellite (S48210) and granodiorite (S48909), (S48229) to quartz-diorite (S48226).

A small granitic intrusion at Lamberton [NT 968 590] (Walker, 1925) is exposed in the stream and is bounded to the east by post-Dinantian faulting. It is a hornblende-biotite-granodiorite containing pseudomorphs, in chlorite and carbonate, after hornblende, locally chloritised biotite, plates of somewhat sericitised andesine, subordinate plates of rather turbid orthoclase and anhedral quartz. The granite is cut by a number of thin lamprophyre dykes which trend approximately N 010°.

The Stoneshiel and Lamberton granites may belong to the Tweeddale batholith (Lagios and Hipkin, 1979).

North-west of Callercove Point, the Linkim Beds are cut by dyke-like bodies of coarse-grained acid igneous rocks, best classed as quartz-diorites, which may well be separated parts of one intrusion. One dyke (S51830), over 20 m wide, runs ENE across the foreshore from a point 20 m ENE of Linkim Kip [NT 9283 6548]. Intrusions of porphyrite (S51831), thermally altered by the diorite, occur immediately to the south-west and to the south-east but separated by screens of greywacke. Another diorite dyke, 7.5 m wide (S63694, (S63696), 150 m W of the first, can be followed across the foreshore north-west of Linkim Kip and into the cliff. The diorites are highly altered and coarse-grained, containing large plates of kaolinised sodic plagioclase and pseudomorphs after hornblende and mica in an aggregate of sericitic material and hematite, subordinate anhedral quartz and a little orthoclase. The thermally altered porphyrite contains much new biotite as scattered small flakes and as aggregates replacing the ferromagnesian phenocrysts.

Porphyries

Rocks described under this heading with the symbol F on the map include quartz-porphyries, feldspar-porphyries, porphyritic microgranodiorites and acid porphyrites. Porphyries are distinguished from porphyrites in that the feldspars in the former ar dominantly alkali feldspars and there is a higher proportion of micas and amphiboles, while in the porphyrites plagioclase predominates. As the rocks have suffered secondary alteration and the types grade into each other, allocation of specimens is not always certain.

Many porphyries with restricted exposure are not shown on the geological map.

Quartz-porphyries occur only in Silurian rocks within 5 km of Eyemouth. There are only two localities where relationships with the country rocks are seen: in the Ale Water [NT 911 637] above Alemill, and between Fleurs [NT 917 653] and Linkim Shore [NT 924 657]. In the Ale Water (S51859), (S58441) the south-east margin of the intrusion, which is acid porphyrite in part, dips SE at 30° beneath greywacke, whereas the north-west margin is defined by a zone of broken rock trending NNE. The Linkim Shore quartz-porphyry also includes more albitic rock-types which may be classified as quartz-microdiorite and keratophyre. Parts of the shore-intrusion are breccia-like; close planar jointing is common, with rounded folds locally, giving the rock a superficial resemblance to thinly bedded sediment.

The quartz-porphyry north of Peelwalls House is well exposed beside the Eye Water [NT 9227 6033]; its junction with the Silurian appears to be irregular. Another exposure (S52763) occurs in Horn Burn [NT 9157 6030] 700 m to W, cutting Lower Old Red Sandstone. Quartz-porphyry was also recorded (S54168) in the north-east corner of the Mordington outcrop of Lower Devonian lavas in a small old quarry [NT 9592 5797], and at one locality [NT 9501 5786] in a water-main trench (S58447).

The only feldspar-porphyry shown on the published map is the intrusion [NT 913 581], known only from soil-debris (S63677), 800 m SW of Whiterig. Pink coarse-grained feldspar-porphyry (S58453) with close joints dipping steeply to N 250° crops out immediately east of the head of Eyemouth Harbour [NT 9459 6111]. Poor exposures along the steep bank to the south-west are of similar rock, but a specimen collected 350 m to SW [NT 9440 6383] is a fine-grained lamprophyre or porphyrite, though apparently part of the same intrusion. A porphyry intrusion lies just within the Linkim Beds on the shore [NT 9256 6554] east of Fleurs.

The quartz-porphyries and feldspar-porphyries contain phenocrysts of oligoclase, commonly sericitised, and plates of mica (or pseudomorphs) in addition, in the first variety, to phenocrysts of quartz. In some specimens (S58453), (S63677) the micas are biotite, commonly (S54168), (S58447) replaced by chlorite, but in others the phenocrysts are plates of white mica (S58441). Small pseudo-morphs, in chlorite or sericitic material after hornblende, occur in a few slices (S52382), (S58447), (S62701), (S63677). The matrix of the porphyries is generally fine-grained containing alkali feldspar and finely granular quartz.

The main rock (S51826) of the Linkim Shore dyke contains numerous phenocrysts of turbid albitic plagioclase, and less commonly of quartz and white mica, in a matrix containing plates of quartz and albitic feldspar. Sericite is common. Some specimens from inland outcrops (S51834), (S51835) are similar but have larger phenocrysts, and in one (S51835) the larger feldspars are completely argillised. The matrix of these two specimens is a finely granular mosaic of quartz, alkali feldspar, and much sericite. Other specimens from Linkim Shore differ somewhat. One (S51827) contains scattered phenocrysts of albitic plagioclase in a matrix of moderately flow-oriented slender laths of that mineral, with interstitial quartz, small plates of sericite, grains of leucoxene, and hematite and carbonate in veins and patches. Another (S50535) is similar but coarser, the feldspar laths forming a plexus and phenocrysts being less common.

Intrusions of porphyritic microgranodiorite are individually significant only within the aureole of the Stoneshiel Hill granite, although they form minor phases in other intrusions. In the Whiteadder [NT 7703 5843] 630 m NW of Cockburn Mill a 6-m dyke (S48217) strikes NE. A larger irregular intrusion (S49420) is seen by the river [NT 7857 6015] 200 m S of Elba, and opposite Elba the greywackes are cut by a 1.2 m wide dyke with a NNE trend.

The microgranodiorites from this area contain feldspar phenocrysts, some of which are partly sericitised oscillatory-zoned andesine, as well as biotite locally (S49420) or completely (S49414), (S48217) altered to chlorite. Altered hornblende also is present in some rocks (S49420). Much epidote, possibly an aureole effect, occurs in two (S49420), (S48217). The matrix generally has a granular appearance and is composed of turbid sodic plagioclase, clearer potash-feldspar and quartz. In one slice (S48217) crystals of apatite are patchily coloured and pleochroic from pale brown to purplish grey.

Acid porphyrites are here grouped with the porphyries because they contain phenocrysts of both sodic plagioclase and mica. In the Ale Water [NT 907 639] south of Whitecross acid porphyrite cuts the greywackes; its margins appear to be steep. There are some acid porphyrite dykes on the shore to the north-east (not shown on the map), and also (recognised as soil-debris) (S63712) in the lava outcrop south-east of Hallydown [NT 92 64]. Two sill-like intrusions which occur in the bay west of Fancove Head [NT 954 627] are 5.2 and 6.7 m thick. The western one shows tourmalinisation (S63710). Similar intrusions are exposed near Breeches Rock [NT 957 621] up to a thickness of 15 m. Although they run parallel to the general strike of the greywackes it is clear that the intrusions are discordant dykes. The cross-cutting relationship is well displayed on the southern face of Burnmouth Hill [NT 956 611] and in the railway cuttings between Burnmouth and Hilton Bay, where several dykes, including an acid porphyrite 14 m wide, cut highly folded greywackes. Dykes up to 30 m wide are seen in Catch-a-penny quarry [NT 959 602].

Many dykes cut the Silurian rocks of Lamberton Moor. Some are over 40 m wide. Tourmalinisation is a recurrent characteristic of the acid porphyrites between Fancove Head and Lamberton Moor. They are difficult to distinguish from porphyrites, and several examples may be cited where one intrusion includes rocks of both classes.

Acid porphyrites cut Devonian lavas near the stream [NT 9552 5783] 1400 m WNW of Lamberton church, and in small old quarries (S54171), (S54173) up to 250 m distant to N and E. Acid porphyrite with phenocrysts of feldspar, biotite, and hornblende (S54166) was seen adjacent to amygdaloidal lava in a trench [NT 9518 5711] 1695 m WSW of Lamberton church.

Undifferentiated felsic dykes are mapped south of Elba. The more northerly [NT 785 601] is 4.5 m wide. At the other locality [NT 786 597] three parallel NNE-trending dykes occur close together on the left bank of the river. They are, from west to east, weathered feldspar-porphyry 7.3 m, hornblende-porphyry (spessartite, (S49431) 2.4 m, and a feldspathic dyke at least 9 m wide.

The acid porphyrites are rather altered rocks characterised by phenocrysts of sodic plagioclase, generally rather sericitised and probably albitised. In a few exceptional specimens (S54173), (S63707) phenocrysts of andesine exhibit oscillatory zoning. In addition, biotite (S54171), (S63710) or chlorite pseudomorphs after biotite occur (S54166), (S54171), in one instance (S63706) with epidote. Phenocrysts of white mica composed of multiple flakes, occur in some specimens (S50534), (S64470) and pseudomorphs after hornblende have been noted. The matrix is variable and generally altered; in some specimens (S63712) it is silicified. Small laths of sodic plagioclase and poikilitic plates of quartz are common and delicate intergrowths of micropegmatite occur in interstitial patches in one sample (S63707). Tourmaline is widespread in porphyrites in the Lamberton Moor area, occurring (S63706), (S63708), (S63709), (S63710) as sparse isolated crystals generally enclosed in feldspar phenocrysts or in the chlorite pseudomorphs after ferromagnesian minerals. Radiating clusters are abundant in one slice (S63704). It is commonly pleochroic from blue to pale pink with some zoning to brown. Epidote aggregates occur.

Keratophyres

Though mineralogically similar to the quartz-porphyries and acid porphyrites, keratophyres have been separated primarily on their distinctive field appearance, being very pale cream in colour, even-grained, close-jointed, hard, and apparently highly siliceous, though in fact they have a rather variable mineralogy.

At Milldown Point [NT 919 664] keratophyre forms a NE dyke between 6 and 17 m wide, with irregular margins which to some extent follow the highly folded strata of the enclosing Coldingham Beds; but it is not clear whether intrusion pre- or post-dates folding. The quartz-porphyry of Linkim Shore [NT 924 657] includes keratophyre-like rock. In the Ale Water [NT 903 642] north-east of Whitfield, exposures of keratophyre extend for 250 m. Towards its eastern edge the keratophyre loses its distinctive megascopic character and has a more felsic appearance. Keratophyres are recognised only in the Coldingham Beds. Samples from Milldown Point give a Rb-Sr date of 398 ± 13 Ma, approximately end Silurian, which may be the date of alteration, rather than of intrusion.

Specimens from Milldown Point (S52489), (S68070) (S68071), (S68072), (S68073), (S68074), (S68075), (S68076) and Milldown Burn (S52487), (S52488), (S68064), (S68065), (S68066), (S68067), (S68068), (S68069) are very similar. They contain small anhedral to euhedral phenocrysts of quartz and of largely carbonated albite. The matrix is very fine-grained, leucocratic, and sericitised, and encloses isolated small laths of translucent albite and grains of quartz, resulting in a characteristic sago-like texture in thin section. In one specimen (S68064) the feldspar laths are moderately flow-aligned. Features of other specimens include rare microphenocrysts of mica pseudomorphed by sericitic aggregate, and feldspar laths pseudomorphed by kaolinite.

Specimens (S51853), (S51854), (S51855), (S65995) (S65996), (S65997), (S68077), (S68078), (S68079), (S68080), (S68081), (S68082), (S68083), (S68084), (S68085) from Ale Water show variation in texture and alteration, and differ in several respects from those of Milldown. Almost all contain phenocrysts and microphenocrysts of albitic plagioclase, locally with carbonate and kaolinite, and commonly intensely sericitised (S65996), (S65997), (S68079), (S68080), (S68081), (S68082), (S68083), (S68084). Some microphenocrysts are of carbonate and micaceous aggregate, pseudomorphing euhedral hornblende and biotite. In the matrix of some specimens tiny laths and tablets of albitic feldspar are set in poikilitic plates of quartz. The matrix of others is granular and contains anhedral quartz (sometimes only small amounts), feldspar and sericitic mica. A specimen (S51855) from a 30-cm apophysis consists of flow-aligned small laths of albitic plagioclase with a little interstitial quartz, very few phenocrysts and carbonate pseudomorphs after hornblende.

The keratophyres are all somewhat altered. At Milldown the feldspar phenocrysts are commonly carbonated, whereas sericitisation is more general at Linkim Shore and the Ale Water. Carbonate and sericite are common in the matrix everywhere. Small pyrite crystals are characteristic, and form elongate needles in a carbonate vein at Milldown (S68070). Microprobe analysis shows them to be essentially pyrite with arsenic (up to 6 per cent), and copper and nickel up to 0.5 per cent (M. T. Styles, pers. comm). A patch of baryte was noted at Milldown Point (S52489), and in the Ale Water (S68077) pyrite is associated with a quartz-baryte vein. Kaolin occurs locally in some veinlets. Alteration may have been penecontemporaneous with emplacement or due to later mineralisation, though this has not been noted in the country rocks.

Porphyrites and andesites

Most of the porphyrites are NNE- or NE-trending dykes, mainly cutting Silurian rocks. Of those which cut Devonian lavas, many have an andesitic texture and so could be extrusive.

Porphyrite dykes (S49410), (S49413), (S49423), (S49424), up to 4.5 m wide occur at three localities in the Whiteadder Water within 400 m above and below Elba [NT 786 600]. On the shore north-east of Lumsdaine [NT 879 700], close to the vents, there are a number of dykes with an andesitic or plagiophyric texture, between 0.2 and 3.0 m wide. The broadest has a variable composition and may be part of the major vent intrusions; the thinner ones are concordant with the greywackes. Another andesitic dyke, 2 m wide (S47963), west of Mawcarr Stells, runs parallel and close to a monocline [NT 8847 6974]. Intrusive andesite (S47993), (S47995) 650 m NE of Pilmuir [NT 907 679] is bounded to the south-east by coarse tuff with fragments of andesite and greywacke. In a crag 30 m to SW greywackes are cut by thin andesite intrusions (S47991), (S47992), (S61761). These exposures are taken to mark the site of a Lower Devonian volcanic vent.

An intricate network of porphyrite dykes up to 10 m wide (S63686) cuts the Linkim Beds [NT 930 652] west of Callercove Point. The complexity of ramification increases south-east towards the Eyemouth Vent, with which these dykes are associated.

On the headland north of the harbour-mouth at Eyemouth [NT 9467 6464] a sill-like porphyrite (S61764), 1.2 m thick, is repeatedly displaced by NW-plunging anticlines and sinistral faults. It can be followed north-north-east for 120 m to a major fault. An andesitic intrusion of uncertain orientation is exposed over 20 m of the left bank of the Eye [NT 9418 6359] at the site of a bridge opposite Netherbyres. The bridge abutments are on brecciated greywacke, screens of which occur within the intrusion. The andesite dyke beside the Eye [NT 9389 6244] opposite Millbank is 4.5 m wide in the railway cutting, and is a basaltic variety accompanied by much parallel brecciation, slickensiding and quartz veining with some baryte.

Between Eyemouth and Mordington the numerous porphyrite dykes are not easily separable petrographically from the acid porphyrites already described. At the cliff-top [NT 9528 6269] west of Fan-cove Head a concordant dyke of altered porphyrite (S52381) 0.9 to 1.5 m wide forms the face of an old quarry in greywackes, which are unusually reddened for 12 m to the east. The porphyrite dyke south of Breeches Rock [NT 958 617] (S64469) is a concordant intrusion up to 45 m wide, which includes thin lenticular screens of greywacke, and several parallel off-shoots (S62702) diverging from its eastern margin. It is clearly a tongued intrusion which has penetrated in a selective lenticular manner among the steep strata. The same phenomenon is seen on a coarser scale on the moor [NT 961 593] 800 m W of Hilton Bay, where two porphyrite dykes 40 and 30 m wide converge northwards.

Intrusive andesitic porphyrite is a major constituent of the volcanic vent at St Abbs, as large irregular bodies south of the harbour (S50528), (S50531), (S50577), (S56544), (S56545), (S56546), (S56548), and as smaller dyke-like bodies (S50527), (S50529), (S58434) to the north. Small outcrops occur in the Eyemouth vent (S50553), (S51867) where the composition is mainly andesitic with the characteristic texture preserved.

Porphyrites have been noted within lavas in the Ale Water and Little Dean, 735 to 1000 m W of Old Linthill and also north of Mordington [NT 9552 5783], but relations with the country rocks are not clear. In the old railway cutting east-north-east of Millbank are two intrusions of olivine-basalt. The southern one [NT 9397 6248] is 7.5 m wide; and the northern one is over 40 m wide, adjacent [NT 9404 6261] to the south side of the quartz-dolerite dyke shown on the map.

The porphyrites, which are all altered, are characterised by phenocrysts of sodic plagioclase, probably albitised and locally sericitised (S62702), (S63686). In a few specimens (S49410), (S49423)–24) residual fresh plagioclase is preserved within the albitised phenocrysts. Other phenocrysts or microphenocrysts include pseudomorphs after hornblende (S50567), (S52381), (S64469) or pyroxene (S54174), (S61764), (S62702), (S63703), chlorite pseudomorphs after biotite (S49423), (S49424) or fresh biotite (S49410). Some of the more severely altered rocks may be intrusive andesites in which the matrix, replaced by quartz, clay or carbonate, has lost its original texture. In one slice (S49413) within the aureole of the Stoneshiel Hill granite, pseudomorphs after ferromagnesian minerals are replaced by epidote which also partly replaces the plagioclase microphenocrysts.

Intrusions of andesite or basalt can be matched petrographically with similar lavas. They include olivine-phyric types, some (S47993) with pseudomorphs after orthopyroxene in the matrix, olivine-orthopyroxene-phyric or microphyric types (S47991), (S47992), (S47995), (S63713), (S63714) and orthopyroxene-microphyric andesites (S47963), (S47996), (S51867), (S61761). Feldspar-phyric varieties are common, the feldspar phenocrysts accompanied by microphenocrysts of pyroxene (S56548), (S58436) or hornblende (S56546) or both (S50528), (S56544), (S56545), (S56546). Many of the rocks (S50577) are extremely altered. The feldspar is commonly albitised and replaced by carbonate and clay minerals, and olivine, hornblende and pyroxenes are represented by pseudomorphs in a variety of materials. Silification of the matrix to form poikilitic areas of quartz is also observed (S50527), (S50529), (S56548), (S58436). Many of the rocks are argillised or sericitised (S50553). Apatite is a common accessory mineral as small generally colourless, but occasionally yellow or orange prisms.

Lamprophyres

Narrow lamprophyre dykes are recorded in all parts of the Silurian outcrop, and locally in the Lower Devonian. A spessartite (S48204) (i.e. characterised by hornblende and feldspar), cuts altered greywackes in the Whiteadder [NT 7739 6048] 400 m SE of Retreat House and others (S49417), (S49422) are seen downstream near Elba. Most of the lamprophyres are recorded on the coast, and are characterised by well-formed crystals of biotite and augite or olivine.

On the shore [NT 8442 7030] 400 m SW of Black Bull a decomposed xenolithic dyke 0.66 m wide (S47989), (S47990) dips at 60° to N 265°. About 400 m SE of Mawcarr Stells [NT 8883 6946] is a 0.9-m dyke (S47999), (S48000), (S48001), (S48002) and 1 km to E [NT 8989 6912] is another (S47962), about 2 m wide.

On Bell Hill a lamprophyre dyke (S47957), dated at 400 ± 9 Ma using the K-Ar biotite method (Rock and Rundle, 1986), can be followed for 70 m SSW from an old quarry [NT 9171 6807], and on the same alignment on the shore [NT 9189 6847] east of Horsecastle Bay, a similar dyke (S51805), (S51806), (S51807) 3 m wide cuts agglomerate and has two thin off-shoots to SE. At the cliff top [NT 9177 6805] just east of the Bell Hill exposure a sill-like body of lamprophyre 4 m thick, interpreted as a south-east apophysis of the dyke, intrudes Lower Old Red Sandstone.

East of Eyemouth a number of lamprophyre dykes are seen. Two similar dykes crop out north of Scout Point [NT 9545 6420], 0.5 (S63698) and 1.0 m (S63700) wide and up to 3.5 m apart, converging and possibly crossing over each other. The dyke (S61736) at Horse Head [NT 955 631] is 0.6 to 1.2 m wide.

Apart from those in the south-west part of the district, the lamprophyres are varieties of minette (biotite-lamprophyre) but with differences in the ferromagnesian minerals present. Some (S47957), (S51805), (S51806), (S51807) have pseudomorphs after olivine as the only other ferromagnesian consistuent. One group (S48000), (S48001), (S61736) has pseudo-morphs after olivine and also in carbonate after pyroxene, as microphenocrysts. In the chilled edges (S47999), (S48002) of such intrusions the olivine pseudomorphs form abundant phenocrysts, but those after augite and biotite are less common.

In a specimen from a dyke which cuts the Lamberton Beach granodiorite the feldspar of the groundmass commonly forms radiating sheaves. In one slice (S47962) large ocelli occur which are more leucocratic than the host rock and contain alkali feldspar, some biotite and intersertal patches of chlorite. The ocelli are mantled tangentially by curved flakes of biotite. Two rather altered rocks (S47989), (S47990) contain thin flakes of chloritised biotite and numerous pseudomorphs, in hematite-speckled chlorite, after hornblende. Larger pseudomorphs after olivine and possibly augite also occur. In some augite-biotite-lamprophyres (S63698), (S63700) the feldspar occurs as radiating sheaves.

Within the aureole of the Stoneshiel Hill granite dykes of chloritised, carbonated fine-grained hornblende-lamprophyre (S49417), (S49422) and others of thermally altered relatively fresh spessartite (S48204), (S49431) and allied porphyrite (S48202) crop out in the Whiteadder Water.

Basalts and dolerites

The vents rocks which crop out on and near the coast around Hollow Craig [NT 878 699] north-east of Lumsdaine form a complex of intrusive masses of olivine-basalts or dolerites and greywacke-breccia, the two rock types being distributed according to no clear pattern. A number of ENE-trending dykes of similar lithology (S47969), (S47983) emanate from the larger masses. The intrusive rocks (S47972), (S48016), (S48017), (S48018), (S48019), (S48021) are dark green, with grey and black variations, speckled with small amygdales, and of medium to very fine grain. Xenoliths of greywacke occur, in places so abundant that the rock grades into a breccia of angular and rounded clasts of greywacke and shale up to 0.15 m across. Locally the intrusive rocks and the breccia matrix weather red and yellow. The rocks of the complex differ from the other Lower Devonian intrusions in their textures and mode of occurrence and therefore are attributed to a different phase of igneous activity, a more deep-seated emplacement of Lower Devonian magma or possibly an early Carboniferous phase. The tholeiite dyke which traverses the complex is of late Carboniferous age.

Numerous irregular intrusions associated with the vent complex are of decomposed olivine-basalt or dolerite. A few specimens (S47969), (S48017) of relatively fine-grained rock contain, in addition to the large pseudomorphs after olivine, small microphenocrysts of bastite probably after euhedral orthopyroxene. Though differing somewhat in texture, in particular lacking the flow orientation of the feldspars, these rocks resemble olivine-orthopyroxene-phyric lavas. Some specimens (S48006), (S48016), (S48018), (S48019), (S48021) are much coarser, ranging in grain-size up to coarse basalt or fine dolerite, and contain phenocrysts, commonly large, of olivine completely replaced by carbonate, quartz, and/or chlorite. Pseudo-morphs in chlorite or basalt occur in some specimens but in others (S48007), (S48008) pyroxene is pseudomorphed by carbonate. Scraps of biotite and intersertal quartz occur in many specimens. Though the rocks are generally highly chloritised and carbonated, and feldspar where not carbonated is albitised and locally argillised (S48021), a few less altered rocks (S47972), (S47983), (S48007), (S48008) occur in which fresh labradorite (An66 at core) is preserved. It seems probable that these rocks represent hypabyssal, more coarsely crystalline equivalents to the olivine-phyric lavas.

Chapter 4 Upper Old Red Sandstone

Rocks of Upper Old Red Sandstone facies crop out both north and south of Lammermuir, resting unconformably on Silurian or Lower Devonian, and succeeded gradationally and conformably by grey, fossiliferous late Tournaisian sediments. The Upper Old Red Sandstone formed as a fluvial piedmont to the older strata (Mykura, 1983; Paterson and others, 1976; Smith, 1967).

The term Devono-Carboniferous is used on the map rather than Upper Devonian, because the Upper Old Red Sandstone spans the system boundary, the position of which is not known. The lowest fossiliferous Carboniferous strata contain CM Zone miospores (late Tournaisian) so at least the upper part of the Upper Old Red Sandstone is also Carboniferous, the base of which lies within the formation.

A sparse fauna of fish, mostly Bothriolepis spp. (Plate 9) but including Grossilepis sp., has been described by Miles (1968) and is generally taken to indicate a Famennian age, with the lowest strata possibly of Frasnian age (House and others, 1977).

As the Upper Old Red Sandstone rests on an uneven unconformable base, its thickness is variable. In the northern outcrop it is 250 to 500 m thick; in the south it is between 200 and 1000 m thick, but in places such as north of Preston and east of Foulden Deans it is much thinner, of the order of 75 m. South-east of the Border, near Cumberland Bower, it is entirely overlapped by the Cementstone Group.

Northern outcrop

Upper Old Red Sandstone is best exposed in a 4-km coastal section centred on Siccar Point [NT 812 710]. The coast is approximately parallel to the strike, with about 200 m of strata exposed between Greenheugh Point [NT 799 710] and the cliff [NT 825 705] south of Meikle Poo Craig, and a further 100 m southwards to the basal unconformity [NT 828 703] near Hirst Rocks. On the northern side of Pease Bay about 170 m of strata appear to follow above the beds at Greenheugh Point. This leads to a total thickness of 450 to 500 m, but the surface of unconformity is so irregular that west of Siccar Point there is no more than about 350 m of strata.

The base of the Upper Old Red Sandstone is exposed in spectacular detail at Siccar Point (Figure 15) [NT 812 710], and can also be seen on the shore south of Hirst Rocks, in the minute outlier on the cliffs above, in Pease Burn [NT 7931 7030] S of Pease Bay, and at two points in Tower Burn (Heriot Water) 800 m farther west. A basal breccia or conglomerate is developed at Siccar Point, Hirst Rocks, and one of the Tower Burn localities, but elsewhere the lowest rocks are sandstones.

Siccar Point

Siccar Point (Plate 1) has outstanding historical associations, stemming from the visit paid by James Hutton in 1788, accompanied by Sir James Hall and Professor John Playfair, which did much to confirm Hutton's developing ideas in the uniformitarian philosophy of geology. It has become a place of international geological pilgrimage, and the impressive display of sedimentary and tectonic processes seldom fails to excite the admiration of even the most jaded visitor. The headland is formed of Silurian greywackes dipping to NNW at about 90°, overlain by Old Red Sandstone bedded greywacke-conglomerate. On the north side of the point this forms an area of wave-cut slope about 100 m long and up to 20 m wide, which dips seawards with the bedding at about 15°. The surface on which the conglomerate rests is in detail very uneven, some of the beds of greywacke standing up more prominently than others. The overlapping of the beds of conglomerate against the greywacke surface is well displayed. The concept of structural succession, the Silurian rocks having clearly evolved through a longer and more complex history than the Old Red Sandstone, and the similarities between ancient and modern sedimentary processes, are expressed in a most self-evident way, and it is easy to share with Hutton and his companions the illusion of being transported backwards in time through the several stages of development ((Figure 15) and Playfair, 1805, pp. 71– 72).

The conglomerate varies considerably in thickness. In places it is at least 6 m thick, but higher up, on the south-eastern side of Siccar Cove [NT 8127 7094], it is absent, the lowest beds being of sandstone. Its absence at other localities, and this local variation, are probably due to its having been deposited preferentially in hollows on the original surface. Only greywacke pebbles are seen, subangular or subrounded, somewhat flattened, and showing imbrication indicating S or SE flowing currents. The sandy matrix is sparse.

The small Silurian inlier in Tower Burn [NT 785 702] is partly fault-bounded but in places the unconformity is clearly visible. The Devonian rocks are crumbly red mudstone and siltstone with ribs of sandstone, locally yellowish green and dipping about 30° to N. Upstream [NT 7851 7005], where the main Silurian outcrop begins, a basal conglomerate at least 4 m thick lies on a very uneven surface and includes partings of sandstone up to 15 cm thick and veins of barite up to 4 cm wide. The northward dip of the Old Red Sandstone persists, at 25 to 30°, as does the steeper SE inclination of the Silurian. In Pease Burn [NT 7931 7030] 210 m NW of Woodend, red sandstone dipping at 35° to N rests on vertical NE-striking greywackes.

West of Siccar Point the cliffs of Silurian are capped by Old Red Sandstone, with basal conglomerate developed locally to about 3 m in depressions. In the bay immediately south of Siccar Point the rocks are traversed by a group of parallel faults with downthrow to SSE, which cause the conglomerate to be exposed again at sea level (Figure 15). Near high water mark in the corner of the bay [NT 8121 7083] is a minute inlier of Silurian, only 20 m by 6 m, above which the conglomerate is about 5 m thick, although the succeeding 6 m, or so, of red sandstones also include many beds with pebbles. Further downfaulting leads to the exposure of higher sandstones immediately south-east of the inlier, but the lowest beds seen on the shore, about 100 m to SE, contain greywacke pebbles.

Redheugh Shore (Hirst Rocks)

The base of the Old Red Sandstone is well exposed on the shore south of Hirst Rocks where the west limit of the Silurian is at a N–S fault downthrowing W, 20 m NE of an old boathouse [NT 8267 7026]. A parallel down-W fault, 10 m farther E, displaces the unconformity sinistrally by about 24 m; and several other faults nearby have similar effect. Eastwards for at least 100 m good exposures show the unconformity and basal conglomerate up to 4 m thick, succeeded by sandstones with pebbly layers. Towards the east end of the exposures younger sandstones successively overlap older at the base. Sandstones exposed 20 m NW of the boathouse are full of pebbles and cobbles of greywacke, and the Silurian must lie only a few metres below the surface.

The small outlier which caps the cliff of Lansey Bank [NT 832 702] shows fine- to medium-grained sandstones with one or two coarser-grained beds up to about 30 cm thick.

Above the basal conglomerate south of Hirst Rocks, sandstones are continuously exposed north and west for about 200 m, the top of this section occurring at low water mark at Meikle Poo Craig [NT 822 708]. The succession is interrupted by faults at several points and correlation of beds across them is not always certain. It is also distinctly possible that some faults were active during deposition; and some beds show considerable lateral variation. The whole section comprises some 190 m of strata: 30 m mainly sandstone, succeeded by 45 m mainly mudstone, then 25 m of sandstones, another 30 m largely mudstone and finally 60 m mainly sandy mudstones with sandstones prominent at top and bottom.

As noted above, sandstones full of greywacke fragments exposed just north-west of the old boathouse are thought to be close to the unconformity. The lowest 30 m of strata exposed here are thinly bedded red sandstones, with beds of mudstone up to 30 cm thick and thin coarse-grained beds with fragments of greywacke, the latter predominating in the lowest 3 m. A further 45 m of beds underlie the lowest prominent sandstone seen on the east-facing Red Heugh [NT 825 703], a 60-m cliff which demonstrates some aspects of the faulting and lateral variation which affect the succession. These 45 m of beds are mainly red mudstones, with thin sandstones near the base and sandstones up to 1 m thick near the top. The mudstones include many scattered sand-grains. The middle of Red Heugh cliff is occupied by about 25 m of very thinly bedded bright red-brown sandstones, with well separated more massive beds, none of which is much more than 1 m thick. In the basal 9 m are two massive red sandstones about 4.5 and 2.5 m thick. The upper thicker one is pale cream-coloured at top and bottom, and has a thin lens of mudstone near the top. Down-dip to N, this bed thins. Above the two thick sandstones is an alternating succession of blocky and shaly mudstones and thin or shaly sandstones, including a 12-m mudstone with a few thin green sandstones. The highest rocks on the cliff lie in the lowest part of a 30-m group of red mudstones which includes a few sandstones. Thin sandstones and mudstones alternate in the overlying 5 m and are succeeded by a 1.2-m red sandstone, with a channelled base. A further 3 m higher is a 3-m bed of red and grey sandstone, divided by two 20-cm partings of red mudstone, forming a prominent reef along the foreshore. Younger rocks to seaward are mainly sandy mudstones, but sandstone beds become more numerous. The thicker beds, up to 1.8 m, are generally cream-coloured or very pale yellowish grey, one of them including calcareous bullions up to about 1 m in diameter. The highest beds seen are two massive fine-grained pale purple sandstones, each between 3.5 and 4 m thick and separated by 1 or 2 m of unexposed rock. They are underlain by about 17 m of red sandy mudstones with many thin hard beds of pale grey sandstone.

Throughout the section the sandstones are composed of small subrounded grains of quartz. Although a few beds are cream or pale grey in colour, they are predominantly a bright reddish brown; some are highly micaceous. Current-bedding is common, either upwardly truncated planar foreset beds, or multiple sets of ripple-drift lamination. The planar beds indicate deposition by eastward-flowing currents, and the ripple-drift evidence suggests currents towards east and north-west. The mudstones are predominantly shaly and sandy with thin layers of tiny pebbles in places, and are universally red except for small pale green reduction spots which are locally common in the lower beds.

Greenheugh Point

The Old Red Sandstone resting on the Silurian at Kirk Rigging [NT 804 710], 650 m E of Greenheugh Point, is tentatively correlated with the uppermost 70 m or so of the Redheugh section. The evidence is partly structural and partly lithological, neither of which are very satisfactory criteria. About 88 m of strata are continuously exposed on the foreshore east of Greenheugh Point, the lowest beds apparently within 10 m of the base, which is not exposed, though there are a few small exposures of conglomerate. The lowest 4 m of the continuous section consists of sandstones, current-bedded and with well developed ripple-drift bedding. They are succeeded by 63 m of sandy mudstones, with thin partings of shaly sandstone. The mudstones are in thick beds: seven of them between 4 and 11 m thick. The overall proportion of mudstone to sandstone is nearly 4 to 1, compared with less than 3 to 1 at Meikle Poo Craig. The thickest sandstone among the mudstones is a 1.8-m red and cream bed with strong planar cross-bedding. In general the sandstones are red, with pink and cream colours developed in many beds, and yellow weathering in one case. Green reduction spots are locally common and many bedding-planes show this same green colour. Planar and ripple-drift current-bedding are common, with foreset beds dipping variously to east and to west. The mudstones are generally red, with many small scattered sand grains. Thin beds of sandstone, some of which are shaly, are common at some levels. Most of the mudstones in the upper half of the section are partly cream-coloured, in bands and irregular patches. Succeeding the 63 m of mostly mudstone strata, are 21 m of thick-bedded current-bedded sandstone with a few mudstone partings, seen near low water mark.

Bothriolepis hicklingi (Plate 9) has been collected from loose blocks on the foreshore between Meikle Poo Craig and Greenheugh Point (Miles, 1968).

Westwards towards Greenheugh Point the same beds are exposed on the shore, and the succession continues upwards towards the north through a further 90 m of strata. Similar alternations of sandstones and sandy mudstones are seen, but in the upper strata, current-bedded sandstones predominate and form thicker units, nine of which are between 2 and 6 m thick. Honeycomb-weathering is a feature in some beds. Between Pease Sands and Greenheugh Point the same transition is seen, from mainly mudstone in the lower beds to a predominantly sandy succession above.

Pease Bay

So far as may be judged, there is little difference in stratigraphic level between the highest rocks seen east of Pease Sands and the lowest seen on the west side, some 500 m distant. There, at Red Rock [NT 791 711], red medium- to coarse-grained cross-bedded sandstones with subsidiary silty mudstones are exposed to a thickness of at least 10 m. The mudstones include yellowish green ribs and coarsely sandy layers. In the corner of the bay some 150 m farther N silty mudstone is seen on the shore and massive cross-bedded sandstone above it in the cliff. About 240 m N of Red Rock [NT 7910 7134] the strata are cut by E– W strike-faults downthrowing N. North of the faults there are abundant deep red and brown iron-stained veins and nodules, the latter often lying in the bedding-planes. Northwards about 170 m of red, brown and grey cross-bedded locally pebbly sandstones are exposed. In the higher beds are many irregular ribs and lenses of iron-stained nodular cornstone which form prominent ridges on the wave-cut rocky foreshore. The highest strata include a prominent bed of red sandstone with veins and lenses of cream-coloured chert, about 1.5 m below the horizon adopted for the base of the Calciferous Sandstone Measures, in the embayment called Eastern Hole [NT 7900 7157].

Inland exposures

Sandstones, siltstones, and mudstones are exposed over a 0.5 km length of Hazeldean Burn [NT 782 708], south-west of Linhead, and in the railway cutting to SE. The sandstones are cross-bedded, medium- to coarse-grained, highly micaceous in places, and exhibit a variety of colours, predominantly red, but locally purple, lilac, brown, yellow, or grey. The paler colours are more common in the upper beds. The strata dip generally N by W at up to 15°.

Exposures in Tower Burn south-west of the railway [NT 786 703], in the lowest beds, are described above. In Pease Burn also [NT 793 703] 25 m of red sandstone are exposed above the unconformity. In these burns the beds dip at 20 to 35° to N. Medium-grained cross-bedded red sandstones, with beds of mudstone and shale up to 3.5 m thick, are exposed in Cockburnspath Burn [NT 790 709], and dip at up to 35° to NNE.

Red sandstones and shaly and sandy mudstones are exposed at a few points in Old Cambus Burn downstream from the basal unconformity. About 640 m E of the West Mains [NT 8108 7026] a gorge is cut in red blocky mudstones dipping at 20° to NE. In the lower part of Piperdean Burn, which joins from the south, sandstones and sandy mudstones are seen in places. About 140 m S of the confluence [NT 8138 7017] ripple-drift sandstone 4.5 m thick with beds of sandy mudstone overlies a 0.6 m coarse sandstone containing mud-stone pellets and fragments of Grossilepis brandi; G. sp., Bothriolepis sp. and Holoptychius sp. (Miles' (1968) Hazeldean Burn locality 28). The combined streams drop the last 30 m to the shore through a short steep sandstone gorge [NT 821 706]. In Littlecleugh Burn [NT 8227 7011] at Redheugh Farm red sandstones and mudstones dip 45° to N.

Structure

Beds in this northern outcrop dip at 20 to 30° predominantly to N. Excellent shore exposures show considerable faulting and local variations of dip. A close relationship between these is well displayed between Siccar Point and Meikle Poo Craig where structures can be followed for 700 m along strike (Figure 16). A group of small faults at a very acute angle to the strike runs near high water mark in a zone up to 40 m wide on the N side of a sharp parallel anticline, which can be followed for over 200 m, with a complementary syncline about 10 m to the S. At the E end of its exposure [NT 8179 7073] the axial trace of the anticline swings abruptly from an azimuth of N 105° to N 170°, and is then obscured by beach deposits. Dips on the N limb swing from a 'normal' 20° to N 010° through 20° to N 070° and finally 35° to N 115°. The more compressed S limb, with a general dip of 20° to N 185°, steepens in the bend of the trace to 50° to N 213° and then 65° to N 275°. The eastward dips of the northern limb decrease very rapidly away from the fold. A further 800 m to E [NT 8255 7050] a minor fault similarly curves through 90°, but over a greater distance, striking SW, S and then SE as it is followed southwards for 100 m.

Such variations in the orientation of structures in beds close to an underlying unconformity may result from adjustment within the sediments to basal irregularities.

Complex folding and faulting is also seen on the foreshore 350 m SE of Greenheugh Point (Figure 17). The folds are confined to a small area [NT 8020 7085] between two NNE faults, and consist of a sharp-crested anticline and complementary flat-troughed syncline a few metres to S trending N 105°. Beds adjacent to the western NNE fault are tilted up against it, forming a faulted asymmetric anticline plunging NE.

Many faults strike NE or NNE and it is suggested that they represent the reactivation of Caledonian movements. The beds are also cut by overthrust faults which bade WNW and by other normal and reverse faults.

As previously noted, Red Heugh cliff [NT 825 703] displays many small normal faults, hading both N and S, the S-hading faults being less steep. The N-hading fault cutting the middle of the cliff strikes N 115° and throws about 20 m, whereas the curved S-hading, SE-striking fracture in the N half of the cliff throws about 17 m. The faults here have the appearance of a mere accommodation of lithological differences and depositional irregularities, and total displacement across the cliff seems to be only two or three metres down S.

Southern Outcrop

The Upper Old Red Sandstone is poorly exposed in the long southern outcrop, and the base and top can be located in only a few places, but the general distribution is readily identified by fragments in the soil and the red sandy drift.

Duns Wood

An isolated mass of red sandstone, with apparently no significant lateral extension, was exposed in 1972 above the intrusive dolerite in Borthwick Quarry [NT 769 544]. Its outcrop straddles the south-west corner of the district. As seen at the top of the 18 m high west wall of the quarry, the sandstone filled a pocket some 6 m deep and 20 m from N to S, the bedding following the curved base, which sloped at about 30° from the northern edge towards the middle. The S edge, approximately vertical, had the appearance of a post-intrusion fault on an ESE strike, against which the bedding of the sandstone was completely disrupted. Vertical boreholes a few metres west of the deepest part of the pocket were said to have proved at least 13 m of sandstone, but there are indications of dolerite near the surface about 100 m W of the face. The sandstone is interpreted as an outlying remnant of the original roof of the sill. A larger remnant in Castle Wood, 1.5 km to NE is seen in an old quarry [NT 781 556] where dolerite is overlain by sandstone, not obviously altered by the intrusion, at an approximately horizontal junction.

Oxendean

Rock is extensively exposed in Oxendean Burn, from the edge of the district downstream to Mill Dam [NT 781 558], below which exposures are few. Bothriolepis hayi was collected from exposures [NT 7706 5605] to [NT 7709 5605] 400 m WSW of Oxendean Tower (Miles, 1968). Alternations of red sandstone and mudstone make up the succession; many of the sandstones have a muddy matrix and many of the mudstones include thin sandy ribs. The sandstones are generally thin-bedded and fine-grained, with a few more massive beds and coarser textures. The dip is 4 to 17° between E and SSE. In Cumledge Burn [NT 791 563] the highest Old Red Sandstone strata crop out against a SE fault.

Whiteadder Water

The base of the Old Red is exposed at several points in Baramill Plantation [NT 775 570] and on the old river-cliff to the north-west, where a number of small Silurian inliers occur. The disposition of these indicates that the river-cliff coincides approximately with the unconformity. In a stream [NT 7733 5714] in the plantation, coarse-grained pebbly sandstone is exposed adjacent to a greywacke inlier, and a short distance to the north-east conglomerate is seen close to the main greywacke outcrop. Most exposures hereabouts are of red fine-grained sandstones and mudstones which are best seen on the steep right bank of the Whiteadder. In general the beds dip at 10 to 20° to ESE.

Exposures are abundant in the Whiteadder near Cockburn Mill, from just below the ford [NT 775 578] upstream to the margin of the Stoneshiel Hill granite. Above the mill, outcrop distribution is complex because of the mutual proximity of intrusive granite, greywacke and sandstones, and of the uneven unconformity beneath the sandstones. Several small inliers of greywacke protrude through them close to the margin of the granite. In this area the sandstones include beds with pebbles of hornfelsed greywacke up to 12 cm across, as well as of granite. South of the mill the sandstones are mainly fine-grained, and mudstones are thicker and more abundant than in the lower beds upstream. The colour is generally red with some white and grey beds. The strata dip SE at up to 20°, but their attitude is very varied in detail, especially near the Silurian inliers and faults. Several localities between Cockburn Mill and Baramill Plantation have yielded fragmentary Bothriolepis spp. (Miles, 1968).

Sandstone is exposed on the east side of Preston Haugh where the highest beds, within a few metres of the volcanic rocks, are in many cases white rather than red. There are isolated exposures at Primrosehill [NT 7835 5776], and by the road [NT 7980 5916] 600 m N of Baird's Covert.

Slighhouses–Billie Mains

Upper Old Red Sandstone is exposed all along Lintlaw Burn, from west of Slighhouses [NT 814 593] to south of Billie Mains, with variable dips between E and S. Medium and fine-grained red sandstones predominate with mudstone and muddy siltstone at a few localities. Current-bedded pebbly sandstone is seen [NT 8354 5846] 1 km downstream from the Slighhouses road-bridge, and a further 250 m downstream the sandstone contains pebbles up to 25 mm of vein-quartz, sandstone, and other lithologies. Some of the higher beds farther downstream are calcareous and, in one exposure, nodular. Coarse-grained calcareous sandstone, laterally very variable, forms the most easterly exposure [NT 8508 5808], where 1.6 m of rubbly, conglomeratic, laminated rock overlies 0.75 m of massive rock, conglomeratic at the base. Current-bedding direction here is S but not far away it is ENE. Strata in this 2 km length of Lintlaw Burn dip consistently between SSE and S, mainly at less than 20°. A noteworthy feature is the occurrence of pebbly beds well above the base.

Crossgate water bore (1979) [NT 8233 5986], drilled since the map was compiled, proved that the Lower Devonian outcrop extends farther south than shown. Slighhouses water bore (1978) [NT 8262 5922] was sunk in fine-grained red sandstone to a depth of 117 m, so the base of the Upper Old Red Sandstone is now known to crop out between the two bores; and an isolated exposure of tuffaceous rock faulted against red sandstone in Fosterland Burn [NT 8357 5984], 1 km to the E, is now considered to mark the southern limit of Lower Devonian strata, rather than belonging to a small fault-bounded inlier. The convolute outcrop of the base of the Upper Old Red Sandstone, as shown on the map between Mayfield and Auchencrow Mains, was adopted from the original survey, but a better estimate of the position of this boundary would be a line from Bunkle Church [NT 811 595], passing north of Slighhouses, through the Fosterland Burn locality, and following a broad are concave to the south passing south of Auchencrow Mains, to merge with the published line east of Billie Castle. The first surveyors recorded 'red sandstone and marl' in the railway cutting [NT 861 600] at Auchencrow Mains, but this could be either Upper or Lower Old Red Sandstone.

Pebbly sandstone, coarse- to medium-grained, is exposed in the stream [NT 8470 5971] 400 m W of Billie Castle, one of the few localities at which the Upper Old Red sandstones are clearly calcareous. Fine-grained laminated and blocky sandstones are seen just west of the castle and within the fortification works. About 450 m downstream, high on the steep right bank [NT 8550 5946], is an exposure of cream-coloured quartz-sandstone, but the few other exposures east of the castle are somewhat atypical in that the sandstones are purple rather than red and in one case, coarsely feldspathic and possibly tuffaceous. However, on the left bank [NT 8573 5942], about 100 m ENE of the last exposure, there is much red sandstone debris. Further support for the inclusion of all these rocks in the Upper Old Red Sandstone is given by the gradual downstream change of dip from S to SW in sympathy with the postulated arc of the formation boundary.

Between the Billie Castle and Draden burns red sandstone debris is locally abundant and there are small exposures in Fosterland and Draden burns. South of Billie Mains purple and brownish red sandstone 4.5 m thick is exposed in the stream bank [NT 8514 5850]. The upper half of this sandstone is micaceous and laminated, with green bands, the lower half massive, but laminated in detail, with current-bedding indicative of derivation from the west.

In Billie Burn [NT 8546 3749], 450 m NE of East Blanerne, massive fine- to medium-grained cream-coloured sandstone 0.3 m thick, locally with knobbly weathering like a cornstone, dips gently to WSW. Marginally higher in the succession, 120 m downstream, highly calcareous yellow and grey sandstones overlie red sandy mudstone. The topmost 75 mm is a yellow sandy cornstone which overlies 0.3 m of fine-grained saccharoidal sandstone. Apparently underlying this, 10 m farther downstream, is a 0.6 m purple flaggy sandstone, which includes a flat-lying calcareous nodule measuring 250 by 115 by 115 mm. On the right bank [NT 8533 5722] 170 m downstream, is a 2.5-m red mudstone, with small quartz grains, and green reduction veins, spots, and irregular layers. Within the mudstone and at the base are two beds of calcareous medium-grained sandstone, pale cream and red, each 0.3 m thick. The highest exposure assigned to the Old Red, 30 m upstream from the mapped top, is a 0.3-in coarse-grained calcareous cream and pink sandstone, dipping at a low angle to WSW. The development of calcareous sandstones and cornstones is a feature of these topmost beds of the formation, of which some 20 m crop out in Billie Burn.

Chirnside

West of Oldcastles unusually coarse-grained red sandstone was dug in the railway cutting and has been quarried nearby [NT 8569 5822]. About 1100 m SSW of the farm stiff red and green clay in the floor of a narrow valley [NT 8604 5720] is interpreted as deeply weathered mudstone. Pits dug to about 1.4 m just north of Chirnside [NT 8698 5670] revealed medium- and coarse-grained red sandstone completely broken into blocks embedded in red sandy clay, and similar rock was seen [NT 8773 5810] 350 m N of Harelaw. Excavations [NT 8672 5654] 600 m NW of Chirnside school were in red sandstone, with only a little glacial drift.

Harelaw Quarry [NT 883 572] is now largely filled in and overgrown, but the original survey noted 'soft red sandstone and sandy marl' dipping at 5 to 16° to SSE, and that Ayton Castle had been built of this stone. This quarry has also yielded Bothriolepis wilsoni (Miles, 1968; = 'B. ornata' of Geikie, 1863, p.38). A hollow 450 m to NE [NT 8863 5741] may be an old quarry; red sandstone is exposed beneath drift. Good exposures near the base of the formation occur in a stream S and SE of Causewaybank. Red sandstone dipping at 20° to N 170° is seen 250 m S of the farm [NT 8778 5874]. About 160 m to NE, high on the left bank, sandstone and conglomerate are near the surface, and a further 120 to 160 m to NE conglomerate is exposed on the right bank on both sides of the road. Conglomerate is also exposed 1200 m to E near the bridge [NT 8909 5865] at Blackburn, and between 220 and 290 m downstream from there to NE. Pebbly soils, with red sandstone debris in places, are conspicuous in the dissected terrain immediately north of the crossroads [NT 9057 5797] at Edingtonhill Plantation (South), but to the north-east tuffaceous debris is common and the uncertainty of the position of the boundary between Lower and unconformably overlying Upper Old Red Sandstone is increased by the probable inclusion in the younger formation of pebbles of the older.

Red sandstone has been quarried at many points south of the high part of Chirnside. Medium-grained reddish brown sandstone is exposed [NT 8695 5631], in an old quarry 250 m NW of the school. About 700 m E of Maines House red clay with cornstone nodules is exposed in a stream [NT 8907 5666], and 40 m to S a 0.6-m pale grey medium-grained calcareous sandstone is seen, dipping at 20° to N 155°. Grey mudstone a further 10 m to south is the highest rock seen in this vicinity, and is taken to lie within the Cementstone Group.

Foulden

Conglomerate north-west of New Farm [NT 92 57] is mapped on the basis of pebbly soil. Red sandstone is seen by the lane [NT 9228 5658] 600 m N of Burnbank, and in the wood 250 m SW of New Farm where it has been worked in two small old quarries [NT 9266 5639]. It is also seen along the stream [NT 932 564] 400 m ESE of New Farm. There was a sandstone quarry in the wood [NT 9314 5585] 100 m NE of Foulden church. Conglomerate and red sandstone, variously shaly and conglomeratic, are exposed in and beside the stream between 500 and 250 m NW of the bridge at Foulden Deans. In old workings on the south bank [NT 9355 5597] exposures of gravel are interpreted as weathered conglomerate, with rounded and flattened pebbles, up to 10 cm long, mostly greywacke with some igneous rock. Red and brown sandstone, with some partings of grey shale, is exposed at intervals downstream from the bridge. Grey micaceous siltstone 320 m below the bridge is near the top of the formation, the base of the Cementstone Group being drawn at a cementstone bed a further 30 m downstream [NT 9394 5524]. In these exposures the dip is generally S, increasing southwards from 17 to 30°. Conglomerate north of the lava inlier is exposed in Wheatland Burn [NT 943 556], where it contains boulders of lava up to 60 cm across and of greywacke up to 30 cm. A NE fault crosses the stream 80 m S of the bridge, the sandstone nearby showing an abnormal NW dip of 40°. South of the fault the sandstones are mainly fine- or medium-grained and of various shades of red, brown, and purple, with some green patches. Many beds are notably calcareous and some have the rubbly nodular appearance of cornstones. Thin beds of siltstone occur in places and a few of the sandstones include beds with siltstone fragments. The dip is generally SSW increasing to 20° at the top of the formation, which is again drawn at a thin cementstone [NT 9423 5510], 270 m SSW of the bridge. Eastwards the Upper Old Red Sandstone is believed to be overlapped by the Cementstone Group, which rests directly on the Silurian beyond Cumberland Bower, 450 m E of the Border.

Eastern outliers

Eyemouth

The conspicuous red cliff on the north-west side of Eyemouth bay is made up largely of an outlier of Upper Old Red Sandstone conglomerate, resting unconformably on Lower Devonian lavas and agglomerates. The headland which it caps is the site of Eyemouth Fort, and the south-west limit of outcrop approximately follows the line of fortifications across the isthmus. Both conglomerate and basal unconformity are well exposed on the cliffs.

At a point [NT 9439 6486] 120 m S of the headland the unconformity is near high water mark, the cliff above being composed of conglomerate, with a few beds up to 1.5 m thick of pebbly current-bedded sandstone and blocky siltstone with scattered sand grains. These beds are weakly 'calcareous; calcite coats joints, and there are deposits of tufa. The largest clasts in the conglomerate are of pinkish brown acid porphyrite, generally rounded and up to 60 cm across. Other clasts are sandstones, tuffaceous sandstones and greywackes, both pale greenish grey fine-grained and purple medium- to coarse-grained. Along this cliff the surface of unconformity declines gently to NNE. A very small inlier of lava is seen towards the headland.

From 20 to 110 m WSW of the headland another inlier forms the cliff-face to within 3 m of the top, and is bounded against the conglomerate on the wave-cut platform to the north-west by a steep reverse fault along the high water mark, and is terminated to the north-east by a reverse fault hading 45° to ESE. Along this cliff the overlying conglomerate is cut by low-angle reverse faults trending ENE. A little bay [NT 942 649] 200 m SW of the headland is bounded to the east by a steep fault downthrowing 9 m to WNW. The base of the conglomerate can be followed in the cliffs on each side of the bay, and conglomerate also crops out in the central area of the bay. On the west headland [NT 9419 6493] conglomerate 12 m thick includes a 2.5 m bed of current-bedded sandstone. Angular fragments of purple igneous rocks are common near the base, but most of the pebbles are of pale greenish grey fine-grained greywacke. Greenish grey and purple sandstone forms the matrix of the conglomerate, and thin lenses of current-bedded sandstone occur throughout. The base is generally flat-lying but there are marked local irregularities.

Greywacke conglomerates in the gateway to Highlaws [NT 9365 6359] and 500 m SE of Hallydown [NT 9270 6430] may mark two small outliers of Upper Old Red Sandstone, not shown on the map.

Chester Hill

The conglomerate forming the outlier capping Chester Hill and Ayton Hill is well exposed on the north-east face of the former [NT 953 602], and is also seen [NT 9448 6012] on the northwest side of the latter. Clasts are rounded, up to 30 cm across, mainly greywacke and acid porphyrite, in a sparse matrix of reddish brown sandstone. Except in some finer-grained beds, the conglomerate is ill-sorted.

There are no exposures in the small outlier of conglomerate [NT 949 585] 1 km ENE of Greenfield, which is mapped on the occurrence of fragments.

Burnmouth

At Burnmouth the uppermost 50 m of the Upper Old Red Sandstone is exposed on the shore west of the harbour, with inverted dips of 55° to 80° towards WNW and WSW. To the west it is faulted against Silurian greywackes. Red sandstones make up most of the succcession, with intercalations of mudstone and cornstone.

The sandstones are in beds 0.45 to 9 m thick, with considerable lateral variation, and vary in texture from argillaceous and fine-grained to coarse-grained with quartz and mudstone pebbles up to 5 cm in diameter. The mudstones generally contain scattered sand grains. Current-bedding is common in the sandstones, in planar, trough, and ripple forms, and indicates deposition by currents flowing southwards. Within the succession five fining-upward depositional cycles have been recognised, each beginning with coarse-grained cross-bedded sandstone, in places with an erosive base, and ending with fine-grained sandstone or mudstone and cornstone. T. E. Smith (1967, 1968) has described these rocks in detail and interprets them as channel and overbank deposits laid down by streams probably in a semi-arid environment. The upward transition from red beds to the grey of the Cementstone Group he ascribed (1968, p.353) to a change of depositional conditions from flood-plain to lagoonal, from oxidising to reducing. He asserted that in the Tweed Basin this change takes place only once (in contrast to the conditions at Pease Bay, where several alternations of red and grey beds are seen) and attributed it to a rapid marine transgression over a wide flood-plain. The Cementstone Group is also much thinner bedded, 0.01 to 0.3 m compared with 0.6 to 4.5 m in the Old Red Sandstone, though there is no change in mineralogy or petrography to suggest a different source of sediment.

Lamberton

Upper Old Red Sandstone on the cliffs at Lamberton, overlooking Hilton Bay, is faulted to the east against Carboniferous strata. To the south it rests unconformably on granite and to the west appears to be unconformable on the Silurian though the junction is not exposed. The best exposures are in the railway cutting [NT 968 590] 350 m NE of Lamberton Shields, where the rock is a coarse conglomerate composed of pebbles of greywacke and dyke rocks, poorly stratified, in a red sandy matrix full of minute lithic fragments. The weathered top of the rock has, as Geikie noted (1863, p.40), a confusing resemblance to glacial gravel, but the pebble content is clearly pre-Carboniferous. Pebbles of the adjacent granite occur in the conglomerate.

Chapter 5 Carboniferous

The Upper Old Red Sandstone of Devono-Carboniferous age is conformably overlain by grey fossiliferous undoubtedly Carboniferous strata of late Tournaisian age. In the northwest Cockburnspath area these grey strata belong to the Calciferous Sandstone Measures (Figure 18). In the south and east Whiteadder and Lamberton areas they belong to the Cementstone and succeeding groups.

In the Cockburnspath area the base of the Calciferous Sandstone Measures is defined somewhat arbitrarily on the coast section at the incoming of grey strata (see below). In the Whiteadder area the base of the Cementstone Group is similarly defined, except in the extreme west where the base of the (post-Old Red) Carboniferous is taken at the base of the Kelso Lavas which underlie the Cementstone Group. At Burnmouth the base is taken at a change in sandstone petrography.

The two southern outcrops, Lamberton and Whiteadder, lie on the northern fringe of the Northumberland Trough and are physically linked through the Berwick area, though the coastal rocks cover a much wider stratigraphic range than those inland. Strata of the northern and southern outcrops differ considerably, being deposited in separate basins on opposite flanks of the positive Lower Palaeozoic area of the Lammermuirs, though both show features derived from their marginal position in the basins.

Classification and zonation

Classification and zonation are shown in (Figure 18). In the north-western Cockburnspath outcrop, the lowest strata yield CM Zone miospores of Ivorian age (Clayton, 1971; Neves and others, 1973), implying that the highest part of the conformably underlying Upper Old Red Sandstone is

Carboniferous. The Cove marine bands (Wilson, 1952), near the top of the Cockburnspath sequence, are correlated with the Macgregor Marine Bands (Wilson, 1974) and yield an Asbian (Lower Dibunophyllum zone) fauna, and also contain NM Zone miospores (Neves and others, 1973).

The only stratigraphically useful fossils from the Cement-stone Group in the Whiteadder valley are CM Zone miospores from strata at Foulden (Clayton, 1985) between 100 and 200 m above the base. In the west the Kelso Lavas occur at the upward change from red to grey sediments and mark a significant palaeogeographical change (Leeder, 1971, 1974). The base of the lavas is conventionally taken as the base of the Carboniferous, though in fact they are well up in the Tournaisian.

In the coastal outcrops around Lamberton, the Lamberton Limestone and Marshall Meadows Marine Band are correlated with the Cove marine bands on faunal and floral evidence (Wilson, 1974; Neves and others, 1973), and with the Dun Limestone of northern Northumberland on general lithological grounds (Fowler, 1926). This horizon, which defines the base of the Lower Limestone Group, correlates with the Ladies Wood Limestone about the middle of the Lower Liddesdale Group in central Northumberland (Frost and Holliday, 1980).

Below the Lamberton (Dun) Limestone, the Scremerston Coal, Fell Sandstone and Cementstone groups are defined lithologically (Fowler, 1926). The Scremerston Coal Group is much thinner and poorer in coals than it is only a few kilometres to the south. Both TC and Pu zones miospores are recorded from the uppermost Cementstone Group strata in the Marshall Meadows Borehole (Neves and others, 1973), which also yielded NM and TC miospores from the Scremerston strata.

It should be noted (Figure 18) that the Lower Limestone Group of the Scottish Midland Valley is wholly younger than the Lower Limestone Group of Northumberland. The Cementstone Groups of each region, while broadly correlative, are not exactly so; and the precise relationship between the Fell Sandstone and Middle Border groups is uncertain.

General stratigraphy

The Cockburnspath succession is the easternmost representative of the Cementstone and Lower Oil-Shale groups of the Scottish Midland Valley (Cameron and Stephenson, 1985). Fluvial sandstones form most of the sequence, with intercalations of marine and fresh-water mudstones, coals with seatearths, and cementstones.

In the south-east at Lamberton the strata were deposited in a different sedimentary basin, but fortuitously fill nearly the same time span. They include representatives of the lowest four groups of the Northumbrian succession, all somewhat thinner than farther south. In the Whiteadder valley only the lower part of the Cementstone Group is preserved, but this is thicker (some 600 m) than the whole group at Burnmouth (450 m).

The Cementstone Group comprises thin-bedded mud-stones, silty mudstones, siltstones and sandstones with subordinate cementstones, deposited in a flat, low-lying semi-arid coastal plain subject to periodical inundation (Smith, 1967; Scott, 1971; Belt and others, 1967; Leeder, 1974; Heward, 1981; Anderton, 1985). The sandy strata are distal fluvial, and the cementstones which are thin, muddy or sandy ferroan dolomites are generally regarded as lacustrine. Other noteworthy features are a restricted, non-marine fauna, occasional rooty horizons, absence of coals, fragmentary plants (Scott and others, 1984), and thin beds and nodules of gypsum.

Very well-sorted fluvial sandstones make up the Fell Sandstone Group, which is virtually unfossiliferous (Smith, 1967). Fluvial sandstones are subordinate in the Scremerston Coal Group which is largely mudstone, silty mudstone and thin sandstones with marine beds, coals and seatearths, showing cyclic deposition in a fluvial and deltaic flood-plain subject to occasional marine incursions. The Lamberton Limestone at the base of the succeeding Lower Limestone Group is a thoroughly marine limestone with corals and brachiopods. Higher strata are mainly sandstone but comprise only part of the lowest depositional cycle.

Palaeontology

Fossils found during the first survey were determined by Salter (in Geikie, 1863, pp.57–58). Fish from Foulden were described by White (1927) and Moy-Thomas (1938), and the fauna and flora of this unique locality have been described in a series of papers by Almond, Andrews, Briggs and Clarkson, Clayton, Forey and Young, Gardiner, Pollard, Scott and Meyer-Berthaud, Waterston, and Wood and Rolfe (all 1985). Floras were reviewed by Scott and others (1984) who list the 26 studies made by Long between 1960 and 1979, mostly on plants from this district. Miospores from Cockburnspath and Marshall Meadows were noted by Clayton (1971) and Neves and others (1973). H.H. Wilson described the Cove marine bands fauna, and all the marine faunas were reviewed by R.B. Wilson (1974).

Cockburnspath area

Lower Carboniferous rocks in the Cockburnspath area are 320 m thick (Figure 19) and are conformable on the Upper Old Red Sandstone, from which they are distinguished by the development of thin beds of cementstone, a general upward colour change from red to grey, and a change of fauna from sparse fish to relatively abundant molluscs and ostracods. Plant remains also become locally abundant. Except for 65 m of strata cut out by the Cove Fault, all the measures are exposed on the coast.

As stated in Chapter 4 the base of the Calciferous Sandstone is taken at 1.5 m above a distinctive red cherty calcareous sandstone which crops out on the shore [NT 7910 7155] 450 m N of Red Rock. The basal 18 m are thinly bedded shales and siltstones with beds of cementstone and sandstone. The cementstones are individually up to 230 mm thick, and several close together form a group 1 m thick. They are fine-grained and grey or brownish, locally reddened. The sandstones are fine-grained, generally pale brown or purple, and cross-bedded or ripple-bedded. Some of the lowest beds are weakly calcareous and show prominent honeycomb weathering. A 1.14-m sandstone below the Eastern Hole Conglomerate contains nodules of sandy cementstone in its uppermost layers. The Eastern Hole Conglomerate, 13 m above the base, is a 0.75 m-thick breccia of cement-stone fragments set in a pale red sandy matrix, and containing plant fragments and fish scales. Overlying strata include a 230 mm-thick bed of cementstone with carbonaceous plant fragments and the marine bivalve Sanguinolites sp., several more cementstones and a 6 m sandstone which is fine-grained, honeycomb weathered, and includes siliceous beds and many carbonaceous micaceous partings. Carbonaceous plant debris is abundant at many horizons in both sandstones and shales.

The Horse Road Sandstone, 47 m thick, crops out at Horse Road Rock [NT 7910 7165]. The central 30 m is a strongly cross-bedded sandstone, greyish white with a green tinge, with weakly calcareous brownish ribs up to 0.6 m thick, plant fragments and lenses with fragments of cementstone and shale. Large round calcareous and ferruginous doggers occur in the uppermost 2 m of this central section. The basal beds of the Horse Road Sandstone are thin-bedded and cross-bedded, and fine- to medium-grained with some coarse lenses. The topmost beds are cross-bedded and ripple-bedded, and include dark grey silty wisps. Blue-grey carbonaceous micaceous mudstone, 2 m thick, follows, then brown and grey sandstones, 10 in thick, and siltstones 10 m thick, with some cementstone lenses, poorly exposed in a bay [NT 7878 7157] 300 m E of Cove Harbour.

The dip increases to 45° and the major, southern, branch of the Cove Fault is thought to pass through these siltstones, nearly along strike. On the northern downthrow side of the fault is a breccia, with fragments of cementstone up to 75 mm across, varying in thickness between 0.3 and 1.5 m, deeply red at the top. The overlying Kip Carle Sandstone, forming the minor headland [NT 7875 7160] from which it is named, is 21 m thick with a 1.2-m shaly parting near the top, and is medium-grained with coarse-grained patches, pale brown and yellowish brown, but purplish red towards the basal beds which contain many bivalves. For about 5 m above the sandstone the mudstones and siltstones, with ironstone nodules and ostracods, are crushed. A layer of leathery clay, up to 0.38 m thick, is developed in the lower part, and angular sandstone fragments occur near the top and the dip increases to 70°. This marks the outcrop of the northern branch of the Cove Fault.

In higher beds the carbon content increases markedly, in the form of plant stems, roots, and comminuted debris, and of thin seams of sulphurous coal. Eight seams of coal, up to 300 mm thick, occur in mudstones between 7 and 18 m above the crushed rocks. About 2 m higher the top of this group is marked by a bed of carbonaceous mudstone underlain by 0.75 m of grey seatearth. These beds may be the local equivalent of the Scremerston Coal Group of Northumberland, which is 300 m thick at the eponymous locality, 100 to 150 m at Lamberton [NT 970 585], here at Cove less than 25 m, and 3 km to WNW in the Birnieknowes Borehole some 80 m thick (Institute of Geological Sciences, 1969, pp.32 and 114). These differences demonstrate the attenuation towards the Lower Palaeozoic land-mass.

Heathery Heugh Sandstone, which succeeds the coal group and forms the headland on the south side of Cove Harbour, is 56 m thick with a median parting of 3.4 m of soft mudstone. The rock is red and brown, varying to purple and yellow in places, medium-grained with many coarser-grained bands, especially in the top 6 m of the lower sandstone. Ironstone nodules, red or purple, occur throughout and are locally abundant in the basal 1.2 m. The lowest 8 m has bands of honeycomb weathering. The upper sandstone, 26 m thick, is markedly cross-bedded, and similar to the lower. Shaly and flaggy sandstones and soft carbonaceous shales succeed the Heathery Heugh Sandstone, in units between 1 and 3.5 m thick. Plant fragments are common, and silty shale 13 to 15 m above the sandstone contains bivalves. Dark grey shale 0.7 m thick with thin partings of coal lies 27 m above the sandstone. Just over 2 m of brownish siltstone and blue-grey shale separate the coal from the weakly calcareous sandstone,' 4.5 m thick, which constitutes the Cove Lower Marine Band, containing brachiopods, marine bivalves, crinoid debris and plant fragments. It is well exposed beside the south wall of Cove Harbour [NT 7851 7166], 45 m from its landward end, whence it may be followed east-north-east across the foreshore. Near the north end of the south wall [NT 7853 7171], about 15 m higher in the succession, the Cove Upper Marine Band is exposed. Intervening beds are mainly fine- to medium-grained sandstone, with siltstone and mudstone at top and bottom, and a 75 mm coal seam in the middle, underlain by about 1.5m of carbonaceous seatearth. The sandstone below the seatearth is red or brown, locally deeply red, calcareous in part, and with small-scale cross-bedding, while that above the coal is yellowish brown, (white in the lowest 60 cm), carbonaceous, and rooty. The upper marine band is a 0.45 m thick rusty-weathered greyish brown siltstone, with abundant brachiopods and bivalves in addition to other marine fauna. A list of the fauna from the Cove marine bands is given on p.54, and two species are shown in (Plate 9).

About 9 m of alternating thin beds of sandstone, siltstone, and shale, with two thicker beds of soft sandstone, separate the upper marine band from the 25-m thick Cove Harbour Sandstone above, which forms the headland on the north side of the harbour. A tunnel was driven through the sandstone to give access to the harbour from the north-west. The sandstone is medium-grained red or yellow-brown, strongly cross-bedded near the top. Just above the middle is a lenticular conglomerate, about 5 m thick, of fragments of sandstone, siltstone and mudstone in a matrix of greyish white sandstone. Further conglomerates occur higher up.

Thinly bedded mudstones and sandstones crop out on the foreshore between Cove Harbour and Reed Point, forming the 20 m of strata between the Cove Harbour and Bilsdean sandstones. The Cove Oil-Shale is a faintly brownish, dark grey shale, 30 cm thick, with small crystals of pyrite and pyritised plant stems, the top half weathering bright red. It is underlain by 15 cm of grey seatearth and friable blocky silty mudstone, grey at the top but mainly reddish purple or greenish yellow; and overlain by blue-grey mudstone with ironstone ribs and plant fragments, which thins eastwards across the shore from 1.3 m to zero, an overlying sandstone resting on the oil-shale near low water mark. Higher strata include mudstone, 3 m thick, mottled greenish grey and purple with yellowish brown patches, and with an irregular calcareous band near the base which forms a distinct feature on the foreshore. The Bilsdean Sandstone, which forms Reed Point and the cliffs to the west, is red and brown, medium-grained, with some coarse-grained beds and calcareous patches. A few neptunian sandstone dykes occupy fissures in the underlying rock.

The lithologies and fauna of the Calciferous Sandstone Measures indicate that they were deposited in a deltaic or fluvial environment, with a few marine incursions. The cementstones are interpreted as being deposited in quiet confined lagoonal conditions where carbonate concentration could increase to saturation point. Intermittent exposure to the air resulted in desiccation and sun-cracking in some beds. Further evidence of uplift is afforded by cementstone-breccias and the conglomerates in the Cove Harbour Sandstone, the products of short episodes of erosion of lithified sediments. The broken cementstone appears to have been redeposited more or less where it had previously lain; the conglomerate was transported by stronger fluvial currents, of the type which formed major sandstones. Coal-seams, oil-shales, and thinly bedded shales and sandstones were deposited in deltaic or flood-plain swamps, brackish embayments, and shallow streams, in an environment subject to periodic change as the rate of sedimentation overtook or lagged behind that of intermittent subsidence. Dr R. B. Wilson (1974) has demonstrated that over south-east Scotland early Carboniferous marine incursions were not individually widespread, although the Cove Lower and Upper bands are local representatives of the widespread Macgregor Marine Bands, to which the Lamberton Limestone also belongs. The succession of marine faunas over a much wider region suggests that the open sea lay to the east (Wilson, 1974, p.48).

Whiteadder Valley

Cementstone Group outcrops in the Whiteadder valley form the northern edge of those underlying the Merse of Berwickshire. Most of the exposures are in the Whiteadder. West and south of Preston, a lava flow up to 8 m thick on tuff up to 6 m thick, are the most northerly representatives of the Kelso Lavas, and are taken to define the base of the Group, but elsewhere the base is defined on a change of colour from red to grey and the incoming of cementstones and of plant and bivalve fossils. The Birrenswark lavas, penecontemporaneous with the Kelso lavas, are dated at 355 Ma (Francis, 1983 quoting de Souza, 1979), slightly younger than the 360 Ma generally accepted age for the Devonian–Carboniferous boundary.

Kelso Lavas

Basalt lavas overlie tuff and red and green mudstone with dolomitic nodules in a temporary section at the southern edge of the district [NT 7917 5437] 800 m SW of Rulesmains. Northwards the next exposure is in Cumledge Burn [NT 7913 5634] 100 m W of its confluence with the Whiteadder, where the beds are disturbed by a fault, and no detailed succession could be established, though amygdaloidal basalt is underlain by tuff. Further faulted exposures occur in the Whiteadder [NT 7845 5675] 250 m W of Preston Bridge, where red sandstones, tuffaceous in part, lie south-west of the fault, and grey sandstones with bands of tuff north-east of it. Fine-grained feldsparphyric basalt is exposed at the bridge and underlying tuff is seen immediately to the south-east. Basalt is exposed at several points on Whitelaw Braes, on the left bank of the Whiteadder some 500 m SW of Primrosehill. In the river at Anger My Heart [NT 7835 5700], 650 m W of Cumledge Mill, the lava is invaded by a sill-like mass of columnar Dunsapie-type basalt (Tomkeieff, 1945). Several exposures of tuff lie immediately to the west, locally with blocks of barite-veined basalt. Quartz-veining is common in the basalt all along these exposures, which indicate that tuffs and tuffaceous sandstones occur above and below a thin flow of veined, amygdaloidal, fine-grained basalt. The lower tuffs are generally red and interbedded with mudstones and cornstones.

There are no exposures of volcanic rocks east of Primrosehill, but abundant blocks of fine-grained basalt near Bishop's Well Plantation [NT 810 581] and on a hillock [NT 838 573] north-east of Hammerhall, suggest that the eastward concealed extension of the lavas may locally extend north to crop.

Lava at Preston Bridge was classed by Tomkeieff (1945, p.55) as a highly altered basalt of Dunsapie-Dalmeny type, whereas that on the bank of the Whiteadder south of Primrosehill has a lower part of Dunsapie type with some Markle characteristics, and an upper part of Jedburgh type of mugearitic affinities. Sliced rocks in the Survey collection show considerable variation, most being highly decomposed, commonly apparently albitised. They include rocks (S48230), (S49436) that perhaps because of alteration have a mugearitic aspect, and also coarse ophitic olivine-basalt (S32670), (S49435). Some specimens (S32674), (S49437) contain phenocrysts of altered feldspar. Specimens (S32669), (S48231) from the columnar sill at Anger My Heart are of Dunsapie basalt characterised not only by phenocrysts of olivine, augite and plagioclase, but also by xenocrysts of bronzite, with a reaction rim of augite which encloses blebs of olivine pseudomorphed by green alteration products. The distinctive reaction rim is similar to that around pyroxene xenocrysts in the Hog Fell plug north-east of Langholm (Elliot in Lumsden and others, 1967, p.202). In another rather altered specimen from this locality no fresh bronzite is preserved but there are pseudomorphs after bronzite surrounded by a reaction rim. Farther downstream very altered basalt contains carbonated feldspar and pseudomorphs after olivine as phenocrysts, and also pseudomorphs after bronzite with the distinctive reaction rim (S32673), (S48212).

An unusual rock (S32672) from Anger My Heart may represent tuff or xenolithic basaltic slag. It contains fragments of generally fine-grained altered basalt and also grains of quartz, many of which are mantled by needles of inverted tridymite.

Cementstone Group

Numerous sections are available, particularly in the banks of the Whiteadder, but they cannot be stratigraphically related to one another, because of lack of distinctive horizons, absence of progressive upward lithological change, and variable structure. (Figure 20) illustrates the lithological characters of the group. Only a selection of exposures are detailed below.

In the Whiteadder valley, dips are mainly southwards at less than 20°, but exposures are sufficient to show that there is a succession of broad folds plunging at low angles to S or SSW. Between Broom House [NT 803 567] and Edrington Mains [NT 943 546] five synclines with intervening anticlines are recognised, generally asymmetric, the SW-facing limbs being marginally steeper and narrower. Axial regions are not well defined, and are 0.5 to 2 km apart. The east–west compression forming the folds may have accentuated original irregularities on the depositional basin floor.

Preston to Edrom

Strata overlying the Kelso Lavas are exposed on the right bank of the Whiteadder at Crumble Edge [NT 7930 5638], 600 m SSE of Cumledge Mill, where there are 60 m of alternating yellowish sandstone and grey mudstone, the latter containing thin beds of cement-stone generally in distinct groups ((Figure 20), column 1). The sandstones and mudstones commonly contain plant material.

Black and dark grey mudstones and siltstones, both with thin cementstones, and pale brown micaceous carbonaceous sandstone are exposed at intervals over 200 m in a stream [NT 794 580] 1100 m ENE of Primrosehill. These rocks are near the base of the Cementstone Group. Higher beds of similar lithology exposed 600 m downstream [NT 7960 5743], just below the road, have abnormally high dips of 50 and 67° to ESE.

Between Preston and East Blanerne there are no significant exposures north of the Whiteadder. Along the river, and in the old river cliffs which border the alluvium, there are frequent exposures of a repetitive succession of yellowish sandstones, grey mudstones, and thin cementstones. Plant fragments are common and there are several beds of calcareous mudstone with bivalves. Dips are generally southwards at 10° and 20°, swinging to WSW near Edrom.

Preston Quarry [NT 8042 5704], 400 m W of Marden, exposes the thickest sandstone seen in all this area, upwards of 8 m of false-bedded buff rock with carbonaceous and micaceous layers, including beds and lenses of conglomerate containing angular fragments of plant-bearing shale in a sandstone matrix. Higher strata seen in the river west of Broom House include many cement-stones up to 0.3 m thick. Calcareous shales with bivalves crop out in the stream [NT 8066 5701] 150 m W of Marden. On the right bank of the Whiteadder [NT 807 568] south-west of Marden discontinuous cementstones exhibit diachronism in relation to the interjacent shales, being developed in them at levels progressively higher towards the west.

A section at Kilnick Plantation [NT 8143 5677] ((Figure 20), col. 6) shows groups of cementstones (including one bed some 60 cm thick) and shales with bivalves (Plate 9) together with two sandstones each about 1 m thick. On the left-bank cliff [NT 8150 5660] 1150 m of Broom House, brown and grey, locally calcareous, mudstones with cementstones and cross-bedded argillaceous sandstones are exposed to a total of 15 m.

Massive buff sandstone, with large-scale current bedding, was formerly exploited in an extensive overgrown quarry [NT 836 563] immediately cast of Todheugh. The cliffs here are made of this sandstone.

Buff coarse-grained sandstone, cut by a basic dyke trending ESE, is seen at several points on steep slopes [NT 805 547] 600 to 700 m WN of Manderston House, and sandstone was worked on both sides of the valley immediately west of Buxley, where white blocky siltstone and dark grey mudstone are also exposed. The eastern quarry [NT 8084 5484], and another, 500 m to SSW [NT 806 544] where rock is not now exposed, are both abnormally extensive.

Chirnsidebridge to Allanton

South-west of East Blanerne the river cliff 500 m long is composed mainly of shaly mudstones and thin cementstones dipping at 5° to 15° to SE. Near the south end of the cliff [NT 8480 5671] is an 11 m wide fault zone, with unknown throw, hading at 55° towards SSW.

Basal Cementstone Group beds are exposed in Billie Burn, west of the railway [NT 852 568]. They are mainly grey shales with thin cementstones dipping generally to SW at less than 20°. Shales with cementstones are exposed in the roadside 250 m SSW of the station; in an old quarry [NT 8516 5650] south of the road 2 m of sandy shale includes two cementstones, 0.23 and 0.15 m thick, and a lenticular sandstone 0.3 m thick; and yellow sandstone was quarried [NT 8530 5665] 120 m S of the former station. In the Whiteadder up to 150 m W of the railway bridge a succession of thin cementstones dips at a low angle to SW or S.

Cliffs up to 12 m high extend 75 m N from the (old) Chirnside Bridge [NT 8518 5625] at Rockhouse Bank. About 65 m N of the bridge a reverse fault of unknown throw hales steeply NNE. Some 12 to 15 m of strata are seen, on each side, in beds up to 1.5 m thick but generally of the order of 0.5 m ((Figure 20), col. 2). Lithologies are sandstones, shales, and cementstones, with transitional calcareous sandstones and sandy cementstones. Sandy strata form 70 per cent of the succession and calcareous strata form 60 per cent. Of the latter, fine-grained cementstones make up one-fifth in beds less than 30 cm thick, whereas sandy cementstones may attain 1.2 m. Current-bedding and wavy bedding are common, and there are rapid variations in thicknesses. One calcareous sandstone, about 1 m thick, has a markedly channelled base.

On the prominent bend in the Whiteadder south of Chirnsidebridge, 25 m of strata are exposed in a 450 m long cliff. Lithologies are similar to those at Chirnside Bridge, but the strata are more thinly bedded and there is a higher proportion of mudstone. Transverse ripple-marks striking N 115° are seen at the north end [NT 8497 5582], and elsewhere foreset beds dip variously to SE and NW. Thin sandstone beds are locally markedly lenticular as are a few of the cementstones, which show all gradations between massive beds and layers of isolated nodules. The strata are predominantly grey and brownish, but at one level there are several beds of reddish brown mudstone and sandstone, up to 0.3 m thick, with greenish grey beds above them.

At the sharp bend in the river [NT 855 557] north-north-west of Stuartslaw sandy cementstones up to 1 m thick form the hulk of a 7.5-m section. Thin cementstones and shales are intermittently exposed for nearly 1 km downstream. Some 30 m of strata are displayed in the 300 m long cliff-section [NT 866 555] south-east of Ninewells. The lower 12 m are predominantly arenaceous, and the succeeding 7 m which is poorly exposed is also mostly sandstone. Over half the strata are calcareous but there are very few fine-grained cement-stones. Several sandstones are 1.5 to 2.5 m thick, with trough cross-bedding a feature in some. The topmost 15 m of the cliff-section ((Figure 20), col. 4) are much more argillaceous than the beds below, shaly mudstone with sandy wisps and layers amounting to nearly half of the total. About 27 per cent of the rocks are calcareous, most of them sandy. One bed, of variable thickness between 25 and 330 mm, is a coarse-grained grit with many pellets of mudstone and other rocks. This bed and some others have sand-filled desiccation cracks. Nodular ironstone forms more or less continuous beds between 75 and 100 mm thick, at four horizons in the shales. The highest beds in the section are locally brown or purple and one shaly sandstone is noticeably green, but most are dull grey, pale buff or cream-coloured. Transverse current ripple-marks were noted, striking generally NNE.

About 50 m downstream from the top of this section, on the same left bank [NT 8659 5530], strongly cross-bedded medium-grained sandstone forms a cliff about 8 m high. Planar, trough and ripple cross-lamination are well developed, indicating deposition from NNW or NW. Beds are up to 0.75 m thick, with interbedded shaly sandstones up to 0.25 m. A farther 50 m downstream, on the opposite bank [NT 8645 5526], massive sandstones are exposed, with cross-bedding indicating deposition from both NE and NW.

Mudstones exposed [NT 8624 5512] east-south-east of Stuartslaw include thin fine-grained cementstones and a layer of irregularly shaped, flattened nodules of fine-grained cementstone about 0.2 m across. Predominantly shaly beds crop out in the Blackadder Water between Allanbank and Allanton, but south-south-west of Allanbank most exposures are of sandstone.

A 13-m section ((Figure 20), col. 7) on the right bank of the Whiteadder [NT 8674 5447], 250 m below Allanton Bridge, where the dip is 8° to N 170°, is at about the same horizon as strata in the lower reaches of the Blackadder, and higher in the succession than those to the north-east of Allanbank. Mudstones and siltstones make up 80 per cent of the strata here, interbedded with fine-grained cementstones, mainly about 25 mm thick but exceptionally attaining 450 mm. Ostracods occur below a median sandstone.

Whitehall to Hutton Castle

At the sharp left-hand bend [NT 8779 5473], 1 km E of Allanton, is a 13-m section lithologically similar to the previous section which is nearly on strike, though cementstones are fewer and thinner and sandstone is more prominent. Near the middle of the section, forming a waterfall, is a 1.6-m bed of pale brown, fine-grained sandstone, finely laminated, ripple-bedded in the lower half and with plants and roots in the upper. Plant debris and carbonacous, micaceous layers are recurrent features of this section. Several mudstones and siltstones are patchily calcareous, and cementstone nodules are locally developed, in some cases in recognisable layers. Siltstone-filled desiccation cracks are up to 30 mm deep in one mudstone. At the base of the section, above a 20-mm bed of hard, brownish grey cementstone, mudstone 1.1 m thick contains plant debris in its upper blocky part and shell fragments in the lower shaly beds.

Lower beds are exposed on the north-bank cliff [NT 877 550] at the right-hand bend about 300 in downstream, and for a further 800 m downstream, east of Whitehall. Mudstones predominate here, with cementstones up to 0.6 m thick, and sandstones up to 1.3 m. Bivalves occur in sandy mudstone about 300 m E of the ford.

March Burn, the left-bank tributary which joins the Whiteadder [NT 8880 5520] 700 m SE of Nether Mains, cuts into solid rock in its lower reaches. Sandstone 20 m thick has been quarried near the sharp left-hand bend [NT 8862 5550] 500 m ESE of Nether Mains. The rock in the burn is brown and cross-bedded, but very hard, bright red and finely conglomeratic at the base. Strata above and below are grey shales with thin cementstones and sandstones. On the left bank [NT 8886 5582] 400 m above the bend, grey mudstone 6 m thick includes eight beds of cementstone up to 200 mm thick, of which several are distinctively yellow-weathering. On the steep banks of the valley hereabouts, there are six small old quarries, probably for sandstone which is well exposed in the largest [NT 8895 5585].

Dark grey shaly mudstone exposed in a small stream [NT 8909 5659] south of Lazybeds Plantation is thought to lie very close to the base of the Cementstone Group.

Rocks exposed near the mouth of March Burn and on the left bank of the Whiteadder for about 200 m downstream are mainly shaly mudstones, with some yellow-weathering cementstones. Sandstones include carbonaceous, micaceous layers and one thin bed contains many large carbonised plant stems.

About 350 m NNE of Hutton Castle Barns a small stream exposes about 19 m of strata on the steep right bank of the Whiteadder [NT 8918 5469]. Shaly mudstone and cementstone constitute 84 percent of this section; in 7.5 m near the middle there are 13 cementstones, up to 280 mm thick and amounting to 1.93 m in total. In a disused quarry [NT 8925 5495] at Edington Mill, 10 m of massive pale brown sandstone is seen. Planar and ripple cross-lamination and convolute lamination are picked out by dark micaceous layers. In another old quarry about 150 m to E similar sandstone is exposed, and a third quarry a further 50 m to E exposes 12 m of dark reddish purple massive sandstone, in current-bedded posts between 0.6 and 3.7 m thick, separated by thin partings of grey micaceous mudstone.

On the south bank [NT 8935 5475], across the river from the old quarries, a 32-m section is exposed. The topmost 12 m are not readily accessible, but comprise about 4.5 m of massive current-bedded sandstone, underlain by 7.5 m of mudstones with many cementstones. The 20 m of strata examined in detail ((Figure 20), col. 8) are mainly mudstones and siltstones, in almost equal proportions, with cementstones making up about 12 per cent of the total thickness, and sandstones about 20 per cent. The sandstones are up to 1.07 m thick, mainly brown or purple, commonly with planar and ripple cross-lamination. An unusual feature of this exposure is the occurrence throughout of gypsum in pink or white veins, nodules, and vugs. Generally the veins are less than 15 mm wide, but one layered vein is 40 mm thick; and vugs up to 75 mm in diameter are seen in one of the lowest beds.

About 600 m to SSW [NT 8913 5411] the Hutton Castle Barns water bore (1927) penetrated the basal 142 m of the Cementstone Group and a further 41 m of Upper Old Red Sandstone. Nearly 80 percent of the Cementstone Group strata are described as 'fakes', shaly sandy mudstones, the remainder being mudstones and sandstones in equal abundance, and thin cementstones and ironstones amounting in total to 6.5 per cent of the whole. About 85 per cent of the beds are grey or greenish grey, the remainder being reddish or brownish, most of which occur between 75 and 115 m depth, where red and grey beds alternate in bands which are mainly less than 1 m thick. More than half of the sandstones were recorded in the basal 37 m, and only here is any sandstone thicker than 0.75 m. Most of the cementstones are in the uppermost 45 m of the borehole, though the thickest is 0.9 m thick at 58.5 m. They are pale grey, locally greenish, and argillaceous, but one or two bands are sandy and might be termed calcareous sandstones. Thin beds of ironstone are common below 60 m, both fine-grained and sandy, the latter forming the thickest beds, up to 0.45 m. Cementstone and ironstone occur at a few horizons as scattered nodules. The two lithologies grade into each other. Gypsum occurs throughout, principally as joint-fillings and veins, but 'ribs' up to 25 mm thick are common in the topmost 50 m, and at 106 m a layer 75 mm thick was recorded. Small concretions, locally kidney-shaped, were recorded at a few horizons, most often in the topmost 20 m, and at one point flat masses of pink gypsum occurred at the intersections of many white veins. Gypsum crystals are scattered generally in the shales. The record of this borehole suggests that gypsum is much more widely developed in the Cementstone Group than the single surface occurrence would imply. Plant debris, bivalves, ostracods, and fish-scales were recorded from several horizons in the topmost 60 m, the ostracods in four bands between depths of 12.6 and 16.0 m.

The base of the Cementstone Group is much nearer the surface in the water bore than would be expected by calculation from outcrops. Some of the discrepancy may be explained by diachronism in the inception of the grey Carboniferous facies, but concealed tectonic complexity, which may be no more than broad undulation of strata, must be the principal cause.

Local tectonic complexity is apparent in exposures at the right-hand bend of the Whiteadder [NT 8985 5480] 900 m NE of Hutton Castle Barns where dips to WSW, E and S are seen in close proximity. About 13 m of strata are exposed on the cliff: mudstones with a few sandstones, and several cementstones up to 0.20 m thick in the predominantly red and green upper beds. On the cliffs 500 m to SSE [NT 9015 5445], 36 m of strata are exposed dipping at 15° to N 135°. The topmost 21 m, not readily accessible, include a lens of massive sandstone which dies out completely from a maximum thickness of 3.7 m. Lower down are four well-separated sandstones up to 2.7 m thick, showing planar and/or ripple cross-lamination. Cementstones up to 28 cm thick occur throughout the section, in which the beds are red, purple, brown and green, but not grey.

Broadmeadows to Foulden

At the next bend in the river [NT 9020 5483], 1400 m WNW of Broadmeadows House, 11 m of generally silty strata, red, brown and green in colour include a multiple cementstone 0.51 m thick and a white flaggy sandstone which thickens east about 1 m to at least 3 m by cutting downwards across the underlying strata. A further 500 m to E, purple and pale pink sandstones are exposed on the left bank [NT 9068 5490] in two beds, 4.3 and 2.0 m thick, between which are a 0.3-m cementstone and red mudstones and siltstones. On the same bank a further 400 m downstream [NT 911 548] two sections in grey mudstones include many cementstones, and pale grey, fine-grained, ripple-laminated sandstones. Several riverside exposures north of Broadmeadows House are in similar strata, in which several beds are characterised by the occurrence of orange-coloured vugs.

On the left bank of the Whiteadder [NT 9228 5470] S of Foulden Newton and 130 m below Hutton Bridge, 22.5 m of strata are continuously exposed ((Figure 20), col. 5). Mudstone, siltstone, and cementstone make up nearly 90 per cent of the succession, sandstone forming only six well separated beds between 0.20 and 0.86 m thick. Ripple-lamination is common in the sandstones, and the thickest bed includes many carbonaceous micaceous layers in its upper part. Near the bottom of the section there is a 6-m sequence of siltstone, with some sandstone and mudstone, in which cementstones are absent, but in general they are developed at intervals of no more than 2 or 3 m. The thickest cementstone is made up of two equal leaves each 0.23 m. Some of the beds are clearly nodular, and scattered nodules and calcareous patches are developed in several mudstones and siltstones. Two of the cementstones are impersistent within the exposure, varying in thickness from zero to 0.08 and 0:25 m. Most of the rocks are grey or buff, but reds, purples, browns, and greens are also seen in significant proportions.

Exposures of mudstones, sandstones, and cementstones occur in Foulden Burn. About 120 m SW of Foulden Newton [NT 9214 5513] shales with thin sandstones and cementstones between 100 and 200 m above the base of the Cementstone Group, have yielded a rich fauna and flora (Almond; Anderton; Andrews; Briggs and Clarkson; Clarkson; Clayton; Forey and Young; Gardiner; Pollard; Scott and Meyer-Berthaud; Waterston; and Wood and Rolfe; all 1985). The exposure is a Site of Special Scientific Interest. North of Foulden Newton exposures are mostly of sandstone.

Clarabad to Edrington

About 350 m NNW of Clarabad [NT 9255 5455] is a 16-m section in mudstones, siltstones, sandstones and cementstones. The thickest sandstone, at 1.75 m, is white, fine-grained, massive, well cemented, ripple-laminated in part and contains carbonised plant debris. The cementstones are up to 0.38 m thick, one showing a variation between 0.05 and 0.25 m, and cementstone also occurs as nodules in some of the siltstones and mudstones. Red coloration is seen at a few horizons but the great majority of the rocks are grey.

About 350 m E of Clarabad beds are exposed to 24 m on the right bank. The topmost 18 m are inaccessible shales and cementstones, with some sandstone, but the lower beds consist of 4.0 m of grey sandstone, fine- to medium-grained, massive but with planar and ripple cross-lamination, with ripple-laminated siltstone above and below for about 1 m. Similar sandstone, exposed to 4.6 m, has been quarried on the opposite bank [NT 9294 5447] 200 m to NNW, and 3 m of similar sandstone forms the top of a 7-m section on the river bank upwards of 100 m E of the old quarry. Sandstone up to 2.7 m thick is exposed at intervals high on this left bank to 220 m E of the old quarry. The lower part of the 7-m section consists largely of calcareous siltstones, with nodular cementstones near the top, and calcareous sandstone, 0.36 m, and cementstone, 0.30 m, near the base.

Just east of Clarabad Mill, opposite the mouth of the Lambsmill Burn, sandstone is exposed to 7.6 m in an extensive old quarry [NT 9340 5416], and a sandstone cliff 4.5 m high forms the right bank of the river for a further 250 m to the east. On the opposite bank, just below the confluence, 3.4 m of rock are exposed, mainly purple mudstones, with four thin sandstones and a cementstone.

Rock is exposed at many points in Lambsmill Burn and its tributaries. On the right bank [NT 9336 5472] 900 m W of Edrington Mains, grey mudstone and siltstone, 8 m thick, locally red or greenish, overlie 2.7 m of sandstone, grey in the upper half with plant roots, brown and mainly massive below. Thin cementstones are common in the mudstone, closely grouped at two horizons to produce layered beds 0.6 m thick. About 800 m WNW of Edrington Mains sandstone was formerly worked in a small quarry [NT 9350 5494], and 150 m to S, on the opposite bank of Lambsmill Burn, an exposure of 5.3 m of beds consists almost entirely of grey siltstone, with eight fine-grained and sandy cementstones up to 0.15 m thick. On the south bank at a sharp right-hand bend [NT 9363 5473] 600 m W of Edrington Mains, is a 13.0-m section with a dip of 20° to N 260° ((Figure 20), col. 2). There are three beds of sandstone in the upper part, 0.84, 0.53 and 1.30 m thick, and several cementstones in the lower part. Ripple-lamination is a feature of the two thinner sandstones. The thickest and lowest is made up of massive posts and thin silty partings. Bivalves occur in a mudstone about 3 m from the top of the section.

Rocks exposed at frequent intervals in the western branch of the stream, which flows south past Foulden Deans, form an alternating sequence of grey and buff shaly and blocky mudstones, micaceous cross-bedded and flaggy sandstones, and thin cementstones, locally ochreous. The highest bed exposed, on the left bank [NT 9395 5497] 420 m NW of Edrington Mains, is pale brown sandstone at least 5.5 m thick, fine- to medium-grained, massive in part and ripple-laminated in part with silty micaceous layers. The base of the Cementstone Group is taken at the base of a 0.2-m compact, fine-grained cementstone, 350 m below Foulden Bridge. In the eastern branch of the stream exposures are less common. Rocks seen up to 50 m downstream from the Old Red Sandstone are mainly sandstones, locally red and coarse-grained, with beds of brownish grey silty mudstone and doubtful traces of bivalves. They dip at 40° to SW. In this stream the basal bed of the Cementstone Group is a fine-grained cementstone, 0.30 m thick, grey with red staining, which is underlain by reddish brown sandstones.

Around Edrington Mains the drift is thin and there are some minor exposures. Mudstones and siltstones, with a thin cement-stone, are exposed to 3 m high on the left bank of the Whiteadder [NT 9442 5404] 570 m SSE of the farm. Many worm-trails are seen in a dark grey mudstone. Just to the south, nearer the river, 12.5 m of strata are exposed in a wooded cliff 100 m long, the beds dipping at 15° to S. The upper part of the section, includes a 1.5 m massive grey to pink sandstone of fine to medium grain. Overlying mudstones contain many thin cementstones. Most of the beds are grey or brownish grey, but red and purple at a few horizons.

Two extensive sections in south dipping beds at slightly different stratigraphic levels are seen beside the Whiteadder within 400 m S of the district limit. Both include sandstones at least 7.3 m thick, in one case strongly cross-bedded, in the other more massive with layers of mudstone pellets; and they also include closely grouped cementstones at three horizons, with aggregate thicknesses of between 0.84 and 1.27 m. At many levels, the mudstones and siltstones are red or purple with green patches or bands.

Lamberton and Burnmouth

The narrow coastal outcrop exposes a sequence similar to that in Northumberland, and the lithostratigraphic classification established there can be applied to the Berwickshire strata (Figure 21), which are grouped as follows:

Lower Limestone Group: 50 m (lower beds only) Thick sandstones and thin mudstones with Lamberton Limestone at base
Scremerston Coal Group: 180 m Sandstones and mudstones with thin coals and cementstones. A few marine horizons
Fell Sandstone Group: 180 m Cross-bedded quartz-sandstones and thin shales
Cementstone Group: 480 m Thick sandstones with interbedded rhythmic units of sandstones, shale, and cementstone

Only the upper half of the Scremerston Group crops out onshore, but submarine investigation has shown that the full thickness is about 180 m (Eden and others, 1969, p.243).

Cementstone Group

The whole group is completely exposed on the foreshore at Burnmouth, and the highest beds are exposed near high water mark for a further 1 km S to Catcairn Bushes. To the west the strata are faulted against Silurian greywackes but at Burnmouth Harbour [NT 959 610] the basal beds follow conformably on the Upper Old Red Sandstone. Over much of the exposure the strata have an inverted WSW dip, locally as low as 55°, but eastwards they become vertical and then dip at 80° to ENE where they are succeeded by the Fell Sandstone Group. The succession consists of cross-bedded, micaceous sandstones separated by alternating beds of mud-stone, sandstone, and cementstone. Typical sections are shown in (Figure 22).

Argillaceous rocks in the basal 130 m are dominantly reddish or purplish, although lacking the bright reds of the underlying Old Red Sandstone. Cementstones are generally less than 0.3 m thick, in many cases sandy, and in some cases internally banded. These last become more common and thicker upwards. One horizon of bivalves and one of fish fragments were recorded, and plant fragments are common. Three thick sandstones are between 8.5 and 14.6 m thick, all showing cross-bedding and ripple lamination; the lower two are grey, the uppermost purple.

In the succeeding 150 m of strata the mudstones are dominantly grey and shaly; the cementstones are mainly banded, and generally thicker and more numerous than below; some are algal. Horizons with serpulids and bivalves are quite common, as are those with rootlets and other plant fragments. It is in this group of strata that the cementstones reach their maximum development and the intervals between major sandstones are greatest. There are two of these, the lower is 15 m thick and purple, the upper one is 9 m thick, grey coloured and has carbonaceous layers. Both show cross-bedding and ripple lamination and include lenses of crystalline calcareous sandstone.

In the topmost 200 m of the Cementstone Group cement-stones are few, thin and sandy and there are many coloured beds: reds, purples, browns, and greens. Only one horizon of bivalves and two of fish fragments were observed and, while rootlet beds are more common than below, other plant fragments are relatively rare. One highly carbonaceous coal-like mudstone was noted, about 5 mm thick. The major sandstones become thicker (up to 28.5 m) and nearer together towards the top of the succession. They are generally red or purple, cross-bedded and patchily calcareous.

In general, most of the mudstones and some of the sandstones, as well as the cementstones, are sufficiently calcareous to effervesce with cold dilute hydrochloric acid. Large concretions of calcareous sandstone up to a metre in diameter are common in some beds. The major sandstones vary considerably in thickness due to wash-outs and channel-filling.

Smith (1968, p.352), interpreted the upward change from Old Red Sandstone to Cementstone Group as a change from fluvial flood-plain deposition to lagoonal deposition, though petrographically and mineralogically the sandstones of the two formations are similar and palaeocurrent directions in both are southward.

Sandstones in the Cementstone Group and the conditions of their deposition are discussed by Smith (1967), most of the material studied having come from the Burnmouth outcrop. The 13 thick sandstones which make up about 30 per cent of the Group are commonly highly micaceous, somewhat argillaceous, and, in the grey beds, carbonaceous. They are medium- to fine-grained, well sorted, sub-arkoses and quartz-arenites. Within beds upward fining is common, with an upward succession of sedimentary structures, from trough cross-lamination at the base, in many cases with locally derived conglomerates, to planar cross-lamination, and ripple lamination at the top. This succession may recur several times in one sandstone body but there are also reversed, upward-coarsening sequences. Internal erosional junctions and abrupt upward decreases of grain-size are not uncommon. Massive and uniformly laminated sandstones are common, and most sandstones have erosional bases in places. Convolute lamination occurs locally. The thick sandstones generally persist along the exposed strike of 700 m, but vary in thickness with basal scouring, and may coalesce with sandstone below, with an intraformational conglomerate at the .junction. Coalescence of sandstone bodies, both vertically and laterally, is the essential process in the formation of the thick sandstones and explains the complex vertical sequences of sedimentary structures and grain-size variation.

About 50 per cent of the rocks of the Group are sandy micaceous mudstones, the remaining 20 per cent being thin cementstones and argillaceous sandstones in approximately equal proportions. They form symmetrical cycles of the form cementstone-mudstone-sandstone-mudstone-cementstone, of which there are over 100 between 0.6 and 3 in thick. Cementstones are typically between 7 and 15 cm thick. Ripple lamination is common in the sandstones.

There are several types of cementstones. Where the term is unqualified, 'cementstone' is an argillaceous dolomitic rock with mainly silt-grade carbonate and at least 10 per cent of clay minerals. `Sandy cementstone' is a dolomitic rock with an abundance of elastic quartz and some mica, as well as the clay mineral content. `Banded cementstone' consists of alternating bands of 'cement-stone' and fine-grained 'sandy cementstone'. There are no hard and fast boundaries between these types, and `cementstone' grades into carbonate mudstone, and 'sandy cementstone' into carbonate-bearing sandstone, or what has been called 'calcareous sandstone' in the Whiteadder outcrop. The texture of the rocks is similar to that interpreted by Pettijohn (1957, p.419) as indicative of the dolomitisation of elastic limestones. Internal structures, such as flaser bedding, ripple lamination, scour and fill, mudcracks and burrows, indicate a shallow-water environment with gentle currents, as in a lagoon, with intermittent incursion of sea-water as a source of carbonate. The presence of dolomitic clasts in intraformational conglomerates implies that the alteration of the original limestone took place early, probably as a result of increasing magnesium concentration in residual brines in contact with calcareous muds and sands. Examples of recent dolomite formation by this process are given by Deffeyes and others (1965), Illing and others (1965), and Shinn and others (1965).

Fell Sandstone Group

Strata of this group crop out on the coast from Ross Point to Hilton Bay. They are steeply inclined throughout, dipping at about 75° to ENE in the north, but southwards gradually changing to inverted dips of 55° to WNW and 40° to WSW. At the southern extremity of the outcrop the beds are transected by the main boundary fault which limits the Carboniferous rocks westwards, and to the east the succcession is truncated by a NNE branch-fault, almost parallel to the strike. Submarine exploration showed that the top of the Fell Sandstone Group is probably only a few metres above the highest strata exposed near Heathery Carrs [NT 9690 5960] (Eden and others, 1969). The exposed beds, 160 m thick, consist almost entirely of sandstone, white, yellow, or pink in colour, cross-bedded and locally convoluted. Some bands are coarsely gritty with quartz pebbles up to 25 mm across. Veins and nodules of ironstone occur locally, and there are discontinuous beds of red sandy mudstone, up to 5 m thick, two beds being seen at one point to converge along the strike as the intervening sandstone thins out.

The sandstones are predominantly very well sorted quartzarenites (Smith, 1967, p.293), somewhat arkosic or lithic near the base of the Group. In contrast to those of the Cementstone Group they are only sparsely micaceous, with almost no biotite, and with very little argillaceous matrix. Planar cross-lamination, convolution, and overturned foreset beds are common sedimentary structures but lack any clear systematic distribution. The direction of palaeocurrent flow is SSW, a few degrees more westward than in the underlying formation. The base of the Fell Sandstone Group is erosive, cutting down by as much as 6 m into the beds beneath; one channel was observed with a depth of 3 m and a width of 24 m. The basal 10 m or so of sandstone include scattered pellets of locally derived mudstone.

Scremerston Coal Group

The Scremerston Group was formerly economically important in north Northumberland where many coal-seams were worked, the last pit closing during the 1960's. The group is 300 m thick in the mining field, immediately south of Tweedmouth, but in this district, 13 km to NNW, offshore of Ross Point (Eden and others, 1969, p.243) the thickness is 180 m and in Marshall Meadows Borehole, just south of the Border, the group is only 96 m thick. Only the topmost 90 m of strata crop out at Lamberton.

At Hilton Bay the highest beds with steep inverted W dip, form an outcrop up to 50 m wide, faulted on the west and succeeded eastward by the Lamberton Limestone and higher strata of the Lower Limestone Group. The main outcrop on the coast to the south is in more gently dipping beds on the E limb of the Berwick Monocline. The Lamberton Limestone lies near the top of the cliff, with Scremerston Group strata in the lower cliff and foreshore. Southwards exposures continue for 2 km beyond the Border.

The cliff-sections are mainly inaccessible, but it can be seen that sandstone channelling is responsible for considerable local variation in the succession, and variations on a smaller scale are also evident.

Few beds other than the coal below the Lamberton Limestone, the Marshall Meadows Marine Band, and the thicker sandstones can be followed for more than a few hundred metres. Mudstones form rather more than half the total section. The Marshall Meadows Marine Band is 15 to 22 m below the Lamberton Limestone and is best exposed at Marshall Meadows Point [NT 982 571], where it is a 2 m grey, bioturbated, sandy mudstone with brachiopods and bivalves. At Lamberton Beach [NT 970 588] the marine strata are in two beds separated by 1.8 m of unexposed strata. The upper bed,1.5 m thick, is of yellow sandstone with brachiopods and burrow trails, whereas the lower consists of 0.9m of grey mudstone with bivalves, its base 16.4 m below the Lamberton Limestone. In the steep beds at Hilton Bay [NT 969 592] crinoid columnals in a cementstone nodule 13.5 m below the limestone are believed to mark the same marine horizon. Nodules and bands of cementstone are developed here in purple and grey shale and sandstone between 11.7 and 18.4-m below the limestone. In the Marshall Meadow Borehole the marine band is a 0.83 m limestone 20.68 m below the Lamberton Limestone. One of the most striking of the thick channel sandstones is that seen on the cliffs above Lamberton Beach. In 200 m it thins southwards from 14 to 4.5 m, and within a further 200 m the overlying sandstone cuts down to join it. The 14-m sandstone appears to have developed from three successive stages of down-cutting and channel filling.

At least 12 coal seams are present, but all are less than 25 cm thick. Most are in a group close to the base of the Lamberton Beach exposures, the others lying beneath the Lamberton Limestone and the Marshall Meadows Marine Band. Seatearths occur at other horizons. Horne (1837, p.189) records that three seams were worked on the shore at Lamberton, about 75, 33, and 14 cm thick, and that a 1.8 m seam was seen at extreme low water. Geikie (1863, p.45) quotes an account by 'the last tenant who tried to work these coals', who drove an adit in a thin coal exposed in the sea-cliff. The adit extended WSW from the cliff for about 270 m to a shaft which was sunk to 37 m below adit level to the worked coal. This consisted of 51 cm of sulphurous coal on 60 cm of seatearth on 41 cm of coal.

The adit is still visible, about 70 m N of the Border [NT 9788 5755], and the shaft is one of two marked on the first Ordnance Survey six-inch map (1860). There is no record of mining at Lamberton more recent than this unprofitable venture. Lamberton Farm No. 2 bore (1909). sited just above high water [NT 9739 5806] 450 m NW of Meg's Dub, proved a coal-waste 1.16 m thick at a depth of 3.40 m. One of the objectives of the Marshall Meadows Borehole was to prove the full sucession of the Scremerston Coal Group and establish the presence or absence of workable seams north of the Tweed. Only thin seams were encountered, (p.54 ).

Strata between the thick sandstones are mainly shales, mudstones, and thin sandstones. Plant debris is common, stems and leaflets as well as rootlets, the latter forming many seatearths.

Sandstones are quartz-arenites (Smith, 1967, p.296) and vary from fine-grained, muddy, micaceous and carbonaceous in the thinner beds, to medium-grained and siliceous in the thick channel sandstones which have an upward fining sequence (in many cases interrupted and in part recurrent) of trough cross-lamination, massive or uniform lamination, convolution, planar cross-lamination, and ripple lamination. Transverse current-ripples are common, up to 25 mm deep and 150 mm in wavelength, and rib and furrow in sets up to 25 mm thick and 75 mm wide. Depositional currents were towards the SSW.

Lower Limestone Group

Only the lowest 50 m of the Lower Limestone Group is present in this district, at Hilton Bay and Lamberton. The western part of both outcrops lie in the steep west limb of the Berwick Monocline but most of the outcrop is in the relatively flat-lying, undulating eastern limb. The basal member, the Lamberton Limestone, is not exposed along the western margin of the southern outcrop where it is probably faulted out.

The Lamberton Limestone, the Dun Limestone of Northumberland, is 1.2 m thick in the Berwickshire exposures, but at Marshall Meadows thickens to 1.6 m. It has a few shaly partings at Hilton Bay. Beneath a locally developed purple or yellowish brown weathering skin the colour is grey. The fauna includes colonial corals, notably Lithostrotion, gigantoproductoids, Girvanella, and crinoid columnals.

Overlying beds are grey shales, with marine bivalves and brachiopods in the lower beds and plants in the upper. The shales are 1 to 5.5 m thick, the variation in thickness due to variable down-cutting by a pink cross-bedded channel sandstone which caps the cliff from Lamberton Beach, where it is at least 7.5 m thick, to Marshall Meadows Point, where it is at least 24 m thick. At Spittal, 6 km SE of Marshall Meadows, the persistent Woodend Limestone occurs 40 m above the Dun, but it has not been recognised in this district; it may be washed out by the thick sandstone. At Hilton Bay the plant-bearing shales are succeeded by about 15 m of yellow, brown, and red sandstone, convoluted at certain levels, which is seen in the cliff to be folded into a large monoclinal syncline facing eastward, the Berwick Monocline. The beds on the eastern limb here dip at 0 to 35° to E.

There are few inland exposures. Close to the boundary fault, in a little valley [NT 9682 5849] 500 m SE of Lamberton Shields, 3 m of coarse-grained pink sandstone forms a prominent group of crags.

Planar cross-lamination dips at 33° to N 135° and the general dip of the strata is 35 to 45° to E. Some 600 m SE similar sandstone is exposed to 3.7 m in a well-defined meltwater channel just west of the railway [NT 9722 5807], and 500 m farther, at The Cove [NT 9746 5763], 6 m of white even-grained sandstone is seen in another channel. At the last two localities water-worn caves have been excavated in the rock. In this southern area, west and south-west of Meg's Dub, many ice-moulded mounds and ridges are composed of sandstone, but many others are essentially of sand and gravel.

The Lamberton Shields Water Bore (1948) [NT 9679 5848] was sited extremely close to the boundary fault, 30 m W of the first of the sandstone exposures. It proved 49 m of sandstone and sandy shales, underlain in downward succession by 0.05 m of coal, 3.12 m of crinoidal shelly limestone with shale partings, and 4.37 m of shelly mudstone, all dipping at about 70° and forming an inverted succession through the Lamberton Limestone and its underlying coal-seam. The true thickness of the limestone is 1.07 m. Beds below, to the total depth of 76.20 m, dip at a low angle, and the two groups of beds are clearly on opposite limbs of the Berwick Monocline. An outcrop of the Lamberton Limestone may be expected between the borehole site and the sandstone exposure nearby to east.

In general lithology and sedimentary structures Lower Limestone Group strata are similar to those of the Scremerston Coal Group, with the addition of thin marine limestones. Palaeocurrent directions are generally to the south or south-west.

Structure

The major structure is the Berwick Monocline (Shiells, 1963), which extends from Burnmouth to Berwick and plunges below the present level of erosion just south of the Tweed. The fold profile is exposed on the cliff south of Hilton Bay, where a thick sandstone steepens westwards, within 50 m, from a gentle eastward dip near low water mark to vertical and steeply inverted. North of Hilton Bay Carboniferous rocks close to the boundary fault against the Silurian dip steeply westwards, towards it. They are inverted here, but eastwards they pass through the vertical and dip steeply to the east.

South of Hilton Bay the beds on the coast all have low dips, and therefore lie on the gently dipping eastern limb of the monocline. The axial trace of the fold is thought to run parallel to the boundary fault, some 50 m east of it, but exposures are few. The steep western limb can be followed to beyond the Tweed but the boundary fault dies out north of the river, and gently dipping Carboniferous rocks extend westwards to join the Whiteadder outcrops of this district.

The fold thus corresponds with Shiell's description of an east-facing monoclinal syncline, tighter in the north, the steep limb broken by a high-angle reverse fault dying out southwards. South of Berwick the syncline has a complementary monoclinal anticline on its western side.

At Hilton Bay a large fault splays off the boundary fault northwards with a trend of N 010°, and limits the narrow outcrop of the Scremerston Group. Submarine investigation to the north (Eden and others, 1969, fig. 1) demonstrated the recurrence of the group and suggests that the fault is a dextral wrench with a horizontal displacement of about 200 m.

About 250 m to the S another fault splays off the boundary fault south-eastwards with a NE downthrow of 100 m. It also displaces the monoclinal axis sinistrally.

The Berwick Monocline may be related to E– W compression against a Lower Palaeozoic block. The structure is most tightly compressed in the north, away from the relief afforded by the depression of the pre-Upper Devonian rocks between the Southern Uplands and the Cheviot Hills. Thus it is that the W limb is steep and faulted in the north but to the south the profile is more open and the limb unbroken. East–west compression could be responsible for both sinistral displacement along the SE fault noted above and the gentle folding in the Whiteadder valley.

Marshall Meadows Borehole

The Geological Survey drilled the Marshall Meadows Borehole [NT 9797 5686] in 1970 to prove the succession in the Scremerston Coal Group, much of which is not exposed and which south of the Tweed includes workable coal seams. Their possible northward extension is of both scientific and economic interest. A graphic section of the borehole is given in (Figure 21).

Strata above the Lamberton Limestone are mainly coarse-grained, brown and purple sandstones, with interbedded mudstone and siltstone with plant fragments. Marine fossils occur in dark grey mudstone above the Lamberton Limestone which is dark grey, 1.58 m thick, with a varied neritic fauna including Lithostrotion junceum (Fleming). Below the limestone is a 5-cm carbonaceous shale, on a 23-cm coal seam on 4.6 m of seatearth. Most of the underlying 15 m of rock is purple, white and grey, medium-grained sandstone. At the base of this is a 1-m bed of dark grey ferruginous mudstone with bivalves and plant and fish debris, followed by 56 cm of calcareous and pyritous sandstones and mudstones, on an 81-cm crinoidal and Shelly limestone with some mudstone partings, which constitutes the Marshall Meadows Marine Band at 74.7 m. Three coal-seams, between 9 and 15 cm thick, occur within the underlying 14 m of strata, which are mainly sandstones and seatearths, with plant fragments, carbonaceous debris, occasional coal partings up to 5 mm thick and sphaerosiderite. Three miospore species known to occur near the base of the zone of Raistrickia nigra–Triquitrites marginatus, the NM Zone, were collected between 97.8 and 99.8 m depth (Neves and others, 1973, p.48). Underlying strata are sandstone, silty micaceous mudstone with plants and carbonaceous layers and three thin seatearths. Thin coals, seatearths and marine mudstones occur between 141.76 and 146.30 m. The detailed section is:

Thickness cm
Mudstone with Lingula fragments 3
COAL, pyritous and soft 10
Seatearth, soft and slickensided 69
Mudstone with Lingula fragments, ferruginous 3
Cementstone, ferruginous: plant stems and bivalves 69
Mudstone, carbonaceous, with ribs of pyritous coal; plant stems, megaspores, fish-scales 43
Sandstone with rootlets; bioturbated: carbonaceous layers and plant stems 107
Siltstone with beds of slumped sandstone 56
COAL, pyritous, and carbonaceous mudstone 8
Siltstone; some plant debris 5
COAL, pyritous, crushed, with carbonaceous mudstone 33
Sandstone with abundant rootlets 51

No further coal seams were intersected in the borehole although plant debris is common and there are a few root beds. The base of the Scremerston Coal Group is taken at 146.3 m. Miospores at 143.6 m indicate a horizon in the upper part of the zone of Perotrilites tessellatus–Schulzospora campyloptera, the TC Zone (Neves and others, 1973, p.48).

Sandstone is the predominant lithology in the next 68 to 213.8 m forming the Fell Sandstone Group. Below this to the bottom of the borehole at 321.62 m, 10 per cent of the sequence consists of thin cementstones and calcareous sandstones in beds and groups of beds up to 1 m thick, while sandstones, including shaly and calcareous beds, make over half the sequence. Miospores from below 266 m represent the zone of Lycospora pusilla, the Pu Zone (Neves and others, 1973, p.43).

Technical difficulties led to abandonment of the borehole at a shallower depth than originally intended; but because the Cementstone Group was reached it followed that the main objective, proving the Scremerston Group succession, had been achieved, though this succession was more attenuated than had been expected.

List of fossils

The Calciferous Sandstone Measures contain non-marine fossils at many horizons: plant remains, the worms Serpula sp. and Spirorbis sp., the bivalves Modiolus cf. laths (Portlock) and Naiadites sp., ostracods, estheriids and other arthropods and fish remains. At a few horizons, poorly preserved, scarce bivalves identified as Lithophaga cf. lingualis (Phillips), Myalina sp., Sanguinolites sp. and Schizodus sp. indicate marine conditions of brief duration.

Fully marine conditions are represented by the Cove Upper and Lower Marine Bands, the Lamberton Limestone and the Marshall Meadows Marine Band (Wilson, 1974, p.42). The combined fauna from these includes the corals Lithostrotion junceum (Fleming) and L. pauciradiale (McCoy), the bryozoa Fenestella sp. and Penniretepora sp., the brachiopods Buxtonia sp., Gigantoproductus aff. maximus (McCoy), Leptagonia cf. analoga (Phillips), Productus redesdalensis Muir-Wood, Punctospirifer cf. scabricosta North, Rugosochonetes cf. hardrensis (Phillips) and Trigonoglossa sp., the bivalves Aviculopecten spp., Edmondia arcuata (Phillips), Leiopteria hendersoni (Etheridge jun.), Sanguinolites striatus Hind, Schizodus axiniformis (Phillips), Streblopteria? redesdalensis Hind and the goniatite Beyrichoceratoides redesdalensis (Hind). Several of these species are confined to the Macgregor Marine Bands in eastern Scotland and the fauna has been traced as far north as Fife (Wilson, 1974).

Early Carboniferous intrusions

The dolerite intrusions of Duns Law [NT 785 546] and Little Duns Law [NT 783 554] are circular in plan, with no exposures though blocks of dolerite and agglomerate are strewn about. They are interpreted as volcanic vents, associated with the nearby extrusive rocks.

The larger outcrop of dolerite, which includes Duns Wood and extends into the adjacent districts, is regarded as an early Carboniferous sill. Fine-grained dolerite, exposed to 4.5 m in an old quarry [NT 781 557] 500 m N of St Mary's Cottage, is overlain by Old Red Sandstone, the junction being nearly horizontal. Elsewhere sparse evidence indicates that the base of the dolerite rises gently south-west. Sandstone, overlying dolerite at least 18 m thick, is exposed at the north edge of Borthwick Quarry, in the extreme south-west corner of the district [NT 769 544], where the top of the dolerite locally slopes south at 30° but is displaced by an ESE fault with N downthrow. Columnar jointing is developed with the columns inclined at 30° to N.

The quarry produces roadstone and low-shrinkage aggregate. The rock is coarse-grained analcime-bearing olivine-basalt or dolerite (S33908), (S33909), (S34689), (S48233) containing numerous crystals of fresh olivine, laths of zoned calcic plagioclase and purplish brown augite as small, commonly aggregated, prisms and rarely as phenocrysts. Alkali feldspar occurs intersertally, and locally intersertal patches of analcime are found. The iron ore occurs as plates, and, characteristically, as stout rods. In the quarry near St Mary's Cottage the rock is finer grained and has purplish trachyophitic augite. The iron ore occurs as small plates, not as rods, and pseudomorphs after olivine are common.

At the top of Red Heugh cliff [NT 8250 7035], a poorly exposed basalt dyke is at least 20 m wide, steeply inclined and trending at N 055°; but it is not seen among the well-exposed sandstones on the shore below, only about 100 m to NE. About 250 m to SE, at the cliff-top [NT 8270 7013] on the west side of a deep gully, loose blocks indicate the presence of another dyke similar in type and strike. A specimen (S63717) from the eastern dyke is an olivine-basalt of Markle type (MacGregor, 1928) with abundant pheocrysts of plagioclase and less commonly of pseudomorphs after olivine. A few small flakes of biotite occur. A specimen (S54183) from a fallen block differs slightly in having scattered large augites and is a transitional type between Markle and Dunsapie, or might be regarded as grouped with the Feldspathic Dunsapie type of Clark (1956, p.46).

Two small intrusions at Mordington are assigned to the Lower Carboniferous from their petrographic affinities. About 170 m WNW of the summit of Witches' Knowe [NT 9549 5723] is a fresh, coarsely ophitic olivine-basalt (S64382), and about 280 m SW of the summit fine-grained olivine-dolerite is seen in another small exposure [NT 9542 5703]. Neither the extent nor the orientation of these intrusions is known, so they are not shown on the map.

Late Carboniferous intrusions

Dykes of quartz-dolerite or tholeiite following east-northeasterly courses are assigned to the late Carboniferous episode of igneous activity at about 295 Ma (Fitch and others, 1970) in which many dykes and sills of this type were intruded in Scotland and northern England. Within the district most of the intrusions are dykes, and will be described from north to south. As the tholeiitic variety, i.e. with a glassy mesostasis, may occur in a single dyke with normal quartz-dolerites, the two types are dealt with together.

A sill-like intrusion of tholeiite (S47966), (S47968), (S48010) –2), up to 4.5 m thick, cuts greywackes and vent rocks on the shore and cliffs [NT 88 70] north-east of Lumsdaine. It is a highly amygdaloidal rock and is markedly reddened in its more south-easterly 'upper' layers. On the beach [NT 891 693] 100 in SE of Heathery Carr a dyke of vesicular tholeiite (S47959), (S47960) about 2.7 m wide strikes from N 080° to N 055° as it is followed eastwards. In the cliff it is inclined steeply to NW.

A dyke of quartz-dolerite (S51837), with tholeiitic margins, follows a slightly sinuous ENE course beneath the south wall of St Abbs harbour [NT 920 672]. At the landward end it is 14 m wide but eastwards it widens locally to 30 m. .Joints parallel to the margins and veins at right angles are prominent features. The SE margin, irregular in detail, is fine-grained, locally banded, and clear-cut. The NW margin is less well exposed but a porphyrite xenolith 0.3 m across is seen within this part of the dyke.

A series of exposures between Reston and the coast south of Eyemouth may mark the course of a single quartz-dolerite dyke. On the left bank of the Eye Water [NT 8967 6217] 250 m NW of East Reston Mill the Lower Devonian is cut by a dyke of purple micaceous quartz-dolerite (S51850) 18 to 21 m wide, its NW margin inclined SE at 70°. It is traversed by steep joints dipping to N 160°, approximately parallel to the local course of the dyke, which is partly exposed again (S63721) on the opposite bank [NT 8939 6205] 300 m to WSW. The fact that it is not seen in the river for a further 300 m upstream suggests that its continuation is to SW, rather than to WSW, or is offset. An isolated outcrop of purple amygdaloidal tholeiite in the stream [NT 9180 6246] 150 m W of Little Dean may be the same dyke. On the same line a further 1.5 km to the east quartz-dolerite (S63674) is well exposed in the Ale Water [NT 9346 6264] 300 m W of Old Linthill, forming a dyke 20 m wide steeply inclined to N 170°. As at St Abbs length-parallel joints are common, as are steep quartz veins at right angles. Westward the dyke appears to be slightly displaced by the Devonian boundary fault. Eastward, on the wooded ridge of Ale Kip Rock, immediately west of the Eyemouth–Ayton road, purple dolerite is seen to a width of about 9 m, and east of the road it forms the high rocky bluff of Kip Rock [NT 9396 6264] (S51892) on the left bank of the Eye Water. At both localities the dyke maintains its steep southerly dip. On the right bank of the Eye the dyke is about 30 m wide and it is seen again in the disused railway cutting. Dolerite is exposed at the cliff top [NT 9553 6336] 300 m S of Scout Point (S63711) and intermittently inland for 300 m. At the cliff-top a sill-like apophysis 0.5 m wide leaves the south margin in a south-west direction, the margin itself here running irregularly to SE. The dyke here is at least 35 m wide and inclined very steeply S. Its central part is eroded to form a deep high cave between two minor headlands composed of marginal dolerite and adjacent indurated greywacke.

A 3-m dyke of tholeiite exposed [NT 9573 6191] 150 m S of Breeches Rock, and 400 m to S [NT 9587 6154] a similar, rather sinuous, E–W dyke (S49003), between 6 and 9 m wide, appears to be cut off westwards by the boundary fault. Another tholeiite dyke (S49005), 6 m wide, crosses the foreshore ENE from the harbour wall [NT 9591 6094] at Burnmouth, and is displaced at intervals, by about its own width to right or left. About 450 m to WNW of the harbour wall, in the gully [NT 9550 6107] south of Burnmouth Hill, a dyke of fine-grained amygdaloidal quartz-dolerite (S49002) strikes ESE and may be a continuation of the harbour dyke. A fine-grained quartz dolerite (S52764) is exposed in the railway cutting [NT 9401 6106] north of Flemington where it is only 0.6 m wide, but can be traced for 50 m, steeply inclined to N 325°. Adjacent greywackes are traversed by many veins of quartz and calcite.

About 200 m N of Hilton Bay an E–W dyke of amygdaloidal quartz-dolcrite is 12 m wide on the shore and in the railway cutting [NT 9658 5955], but it is not seen again in the moderately well exposed ground farther west.

At the western edge of the district an E–W dyke of quartzdolerite cuts the granite of Stoneshiel Hill, and is exposed in the Whiteadder Water [NT 7720 5882] where it is 18 m wide.

An isolated exposure of micaceous quartz-dolerite is seen among sandstones [NT 8055 5471] 600 m NW of Manderston House.An abundance of similiar debris in the soil some 200 m NW of the exposure indicates that the intrusion is a dyke about 15 m wide with a trend of about N 115°.

Rocks of the quartz-dolerite suite include some very coarse-grained (S51837), (S51850), (S51892), (S63674), some finer grained but crystalline and better described as quartz-basalt (S49002), (S52764), and also tholeiitic (used in a petrographical sense) dolerites containing an intersertal mesostasis of glass which may be devitrified. Specimens from chilled thinner dykes or margins of thicker dykes (S48012), (S49005), (S52254) contain pseudomorphs after olivine. The rocks are generally very altered but the plagioclase is commonly fresh. In the fresher rocks some fresh pale clinopyroxene is preserved (S49002), (S51837), (S51892), (S52764), (S63674), (S63721), which in (S63711) shows rod-like central pseudomorphs in verdite which may represent altered orthopyroxene. Generally however the pyroxene, which occurs as subhedral (locally subophitic) to euhedral plates or prisms, is replaced by carbonate or chlorite. Scraps of pale brown hornblende occur in one specimen (S51850). In the coarser rocks quartz and a quartzo-feldspathic mesostasis occur intersertally with iron ore as plates, locally skeletal. In the finer grained `tholeiitic' or glassy rocks the iron ore commonly occurs as needles forming a reticulate meshwork (S47960), (S47966), (S49003). Chlorite or bastite pseudomorphs after pyroxene occur as microphenocrysts in some fine-grained rocks (S47968), (S48010), (S48011), (S48012). Devitrified glass may be confined to intersertal patches but may partly fill amygdales (S47959) in some tholeiites.

Chapter 6 Quaternary

Most of the Quaternary phenomena seen in the district date from the last major glaciation, the Devensian, and more recent times. At their maximum, Devensian ice-sheets blanketed the entire district, moving in a generally southeastward direction; but major topographic features influenced the movement, thickness and duration of the ice and its deposits. Extension of the ice-sheets into the North Sea directly affected the development of the drainage. Interpretation of Quaternary geology depends largely on studies outwith the district (Sissons, 1983).

For description the district is divided into five topographic areas. They are the north-western lowlands, the broad upland zone of the Lammermuir Hills (extending to the coast between Fast Castle and St Abbs), the lowland area of the upper Whiteadder and lower Eye, the upland area between Chirnside and Burnmouth, and the lowland area of the lower Whiteadder.

North-western lowland

The low ground around Cockburnspath is largely occupied by sands and gravels overlying boulder clay (or till), and forms the eastern extremity of the coastal plain which extends from beyond Dunbar.

The undulating topography probably owes more to subsequent erosion than to deposition, though the high proportion of clay in the matrix of the gravel and the common occurrence of lenses of sand and gravel in the boulder clay suggest strongly that the deposits all originated in close association with glacier ice. The best exposure of sand and gravel is in the disused sand-pit [NT 774 704] south-west of Chapelhill, which shows 1 m of pebbly sandy clay on 1 m of gravel on 6 m of red-brown sand, strongly cross-bedded and with coarse lenses. The moundy terrain, the many meltwater channels, and the abnormally high sand-content all suggest that this deposit is a glaciofluvial outwash, unlike the more directly ice-derived gravels on the lower ground to north and east. At the cliff-top [NT 7952 7059] south-east of the ford at Pease Bay, the basal 4.5 m of gravel is underlain by brown boulder clay, and smaller exposures on the cliffs to the north-west show a similar succession.

At the Pease Bay locality the boulder clay has a bed of calcite-cemented gravel 0.6 m thick about 10 m below the top. This gravel, which appears to form one bed, is exposed at many points though its position relative to the gravels and boulder clay is variable. At Heathery Heugh [NT 7856 7157] it overlies gravel, whereas 450 m SE, near Red Rock, it forms a 0.5-m band between red-brown sandy pebbly clay above and chocolate-brown boulder clay beneath. On the left bank of Pease Burn [NT 7915 7009] west of Woodend it overlies sandy boulder clay, and on the cliff [NT 7975 7085] to the north-east it forms the basal 1.5 to 1.8 m of the gravels, with chocolate-brown sandy boulder clay immediately below. It is seen again in Greenheugh Burn, about 300 m to E, where it is at least 2.5 m thick, and also nearly 2 km farther east, at the cliff-top [NT 8190 7063] 700 m ESE of Siccar Point, where it lies at the base of a 3-m exposure of bedded gravel. It is thought to be formed by the precipitation of calcium carbonate at springs emerging at the base of the gravels. On Lansey Bank [NT 8327 7006] about 1 km E of Redheugh the base of the drift deposits, underlying boulder clay and gravel, is a calcite-cemented layer of coarse angular scree-like debris.

On the cliffs north of Redheugh, gravel is seen at several points at the base of the boulder clay, the best exposure, to 4.5 m thickness, being just south of the mouth of Hazeldean Burn [NT 8211 7050]. Poor stratification and sorting suggests that this gravel was deposited directly from the ice, and that clay and fine sand have been washed out by circulating groundwater.

The wedge-shaped area of gravel which extends southwards from Tower Farm to Penmanshiel Cottage has a generally plane surface with a low southward gradient, and is a terrace deposited from meltwater flowing south-east through to the Eye valley. The gravel is coarse and its pebbles mainly greywacke.

Boulder clay in the Cockburnspath area is generally red-brown or chocolate-brown, with some grey and yellow patches, sandy and pebbly but locally of a stiff consistency. It includes lenses and beds of gravel and fine sand, with pebbles of greywacke, sandstone, basalt, trachyte and quartzite. The clay appears to be thickest in the Heriot Water valley east of the railway, where the stream flows in a gorge whose unstable banks are about 15 in high.

The recurrent interbedding of gravel, sand and boulder clay is evidence of penecontemporaneous reworking by meltwater of the initial deposits, causing a degree of size-sorting of the constituent materials. This phenomenon is noted elsewhere in the district, for example, in the Eye valley and in the coastal area south of Burn-mouth.

Meltwater channels in the higher parts of the Cockburnspath area, of which the most spectacular are those south-west of Tower Farm, were formed by streams flowing south-east along the lower slopes of the Lammermuir Hills. The alluvium-filled valley east of the village is the prolongation of a well-defined channel which runs from the adjacent district. North of Old Cambus West Mains, another series of meltwater channels runs parallel to the coast. The lowest cuts a sinuous narrow rocky gorge in Silurian greywackes and, beyond Old Cambus Quarry [NT 808 705], a broader grassy valley through the softer Old Red Sandstone. The channels originated at a time when the flow of meltwater along the foot of the Lammermuirs was held to the south by ice in the sea, and were progressively abandoned as the ice wasted, until ultimately short northerly seaward streams were established.

Behind the sands of Pease Bay a shingle beach, largely grass-covered, extends up to the 25-ft (7.6 m) contour, one of the few examples in the district of the post-glacial shoreline which is widely developed farther west. The cliffs east of Greenheugh Point are generally subject to landslip, ideal conditions for which are afforded by the seaward dip of the strata.

Lammermuir Hills

This is the upland area north of Preston, Auchencrow [NT 85 61] and the lower Eye valley. The ground rises abruptly along the northern edge to altitudes of 200 to 250 m above OD, which are maintained in rounded summits across the Silurian outcrop to the highest points in the district, 269 and 262 m, which lie on the Drakemire Strips [NT 80 60] to Horseley Hill [NT 83 62] ridge, close to the southern edge of Lammermuir. The upland is deeply dissected. Aligned valleys in the ground north of the Eye Water, and the parallel elongation of rock exposures, are due to erosion by the ice-sheet. Southwest of the Eye valley erosion was not so severe. Differences in the distribution of solid rock exposure reflects the dynamics of the ice-sheet. On steep slopes facing north or west the ice was strongly erosive, whereas in hollows and on lee slopes erosion was reduced and deposition increased.

Crag and tail

Erosive power of the ice is well demonstrated near St Abbs where crag-and-tail features form narrow parallel ridges running southeastwards from knobs of greywacke, some of which show glacial striae. Over-deepening on the stoss side of the crags forms hollows, that occupied by Coldingham Loch [NT 89 68] being the outstanding example. The narrow drift-filled valleys between the crag-and-tail ridges are glacially eroded and later served as meltwater channels. In this coastal area glacial erosion was intensified by the obstructive effect of the high ground, combined with the eastward confluence with North Sea ice. Post-glacial erosion was intensified by meltwater from the north-west, flowing between ice-sheets on the higher ground to south-west and in the sea.

West of Eyemouth crag-and-tail topography and continuation of south-eastward alignment of drainage persist across more subdued terrain.

Crag-and-tail is also developed between Old Cambus [NT 80 69] and Dalk's Law [NT 84 65], but features are more rounded, rock exposures relatively rare, and meltwater erosion much less. Both here and in the upland area to west and south many of the areas mapped as solid rock occupy smooth hill-tops and slopes with little relief. Erratics confirm the former presence of glacial drift, but the brown soil and abundance of greywacke debris indicate that solid rock lies at shallow depth. There are undoubtedly pockets of boulder clay within these areas, and over much of the high ground there is a considerable depth of rock-waste overlying the truly solid rock. In such upland uncultivated land, recognition of superficial deposits depends on subjective interpretation of minimal evidence.

Boulder clay

Over most of the Lammermuir area the boulder clay is red or reddish brown and sandy with many small stones, most of which are little worn and locally derived. Soil colour varies between brown and red. Where it is abnormally red, other evidence shows that solid rock is near the surface. In places the boulder clay is abnormally sandy and pale brown or orange-brown, as for example between Grantshouse and Coveyheugh, where up to 4.5 m of orange-brown gravelly clay is interpreted as a basal facies of the boulder clay.

Some of the widespread occurrences of pale brown generally finely sandy clay with only a few small erratics, may be weathered and re-worked boulder clay with a degree of size-sorting of the constituents. Some may be re-worked pre-Devensian boulder clay which was preserved in pre-glacial valleys like that of the Eye Water, but deposits of earlier glaciation are not confidently recognised. In the dean south-east of Lemington [NT 861 631], several exposures show about 6 m of bedded gravel with beds of sand, between boulder clays, the upper one at least 7.5 m thick red or purple and sandy, and the lower one red and at least 3 m thick. In places the top metre of gravel is cemented. Similar clay and gravel relationships beside the Eye near Reston (below) suggest the possibility that the bedded gravels and lower boulder clay may be pre-Devensian.

In places boulder clay is reworked, and sorting and redeposition form more homogeneous deposits of clay, sand or gravel. Gravel which at several points lies between solid rock and boulder clay may be a coarser lag deposit from the latter. In the Lammermuir area such gravels are seen on the cliff-tops [NT 9167 6789] 550 m SW of Wuddy Rocks and [NT 9394 6495] 480 in NW of Eyemouth Coastguard Station. The best exposure of redeposited clays is in the disused brick and tile works [NT 922 632] about 500 m W of Linthill, where stiff purple clay with few stones, at least 1.5 in thick, appears to lie within red and brown sandy stony boulder clay and may have been deposited in a lake impounded against ice in the Eye valley. About 250 m north of the tile works a borehole [NT 9218 6349] proved 1 in of similiar stiff clay beneath 2.7 in of clayey, coarse gravel, which forms a flat expanse here and appears to be outwash from a meltwater channel from the north-west. Abnormally tough red and purple boulder clay occurs at two points [NT 9314 6264], [NT 9340 6261] by the Ale Water south-east of the tile works.

Boulder clay is exposed to a maximum thickness of 9 m in the streams which drain Coldingham Moor: in Buskin Burn [NT 9005 6590], in the Ale Water near Alewater House [NT 893 646] and Press Castle [NT 871 654] and [NT 8589 6508], and in Mid Grange Burn [NT 8634 6582]. In the valley north-east of Penmanshiel Cottage [NT 800 681] brown stony clay is at least 9 m thick on the south-east side, but the steep north-west bank is in solid rock covered by a thin veneer of boulder clay. The greatest recorded thickness of superficial deposits in the district is 20.1 m in a water-bore [NT 8785 6297] at Heugh Head, where 3.3 m of gravel overlies 16.8 m of boulder clay. On the high ground north-west of Auchencrow boulder clay is only locally thick, as in the depression south of Warlawbank [NT 832 619], drained by a stream which 480 m N of the road [NT 8367 6120] has its left bank in solid lava and the right in 7.5 m of red boulder clay.

Coastal area

Gravel deposits on the steep seaward slopes [NT 83 69] north of Haud Yauds occur as ridges on the high ground and as terraced delta-like spreads on the lower ground. The ridges, 3 to 4.5 m high, steep-sided, flat-topped and sinuous, extend for about 400 m immediately NW from two shallow cols [NT 8376 6952], [NT 8388 6958]. Small old quarries show rounded pebbles and cobbles up to 0.3 m across in poor exposures of unstratified poorly sorted gravel. The ridges are best interpreted as eskers, formed by englacial meltwaters flowing south-east, up and over the cols.

As the general level of ice fell northwards, meltwaters were able to cross the watershed at progressively lower levels. The spread of gravel south of Hirst Rocks [NT 83 70] reflects this lowering of drainage level. On the lower slopes, principally between the 122 and 152-m contours, broad terrace-like surfaces are developed at several levels, in some cases resembling small deltas deposited from the south. The terracing is clearly visible, in suitable light conditions, from the vicinity of Siccar Point. Near the cliff-tops mounds of coarse gravel stand up locally above the general level, and appear to be ice-contact deposits.

Poorly stratified fluvial gravel with a very sparse sandy matrix is seen, up to 3 m thick, at the cliff-top [NT 8603 7097] south-west of Fast Castle, resting locally on boulder clay, but whether or not boulder clay also overlies it is not clear. At the cliff-top [NT 8680 7084] southwest of Souter sand and fine gravel, finely laminated, is exposed to 3 m, underlain by unsorted clay-gravel 6 m thick, with cobbles up to 0.6 m across, but locally almost stoneless. Near Souter evidence of four erosion levels takes the form of small pockets of wet boulder clay, swampy head, or gravelly benches.

Dowlaw Burn is a post-glacial stream cutting across the glacial grain of the landscape and beheading streams which at one time flowed south-east. The character of its valley changes as it crosses the alternating ridges and hollows, sections with steep rock walls alternating with more open stretches across drift-filled depressions. The first rocky section lies north-west of Coldingham Common, and is followed by a broad shallow section, beyond which the stream plunges over a waterfall [NT 8530 6922] into a sinuous gorge 25 to 90 m wide and 27 m deep. A final open section is followed by the spectacular gorge of Dowlaw Dean, winch is headed by a second waterfall and is a wider straighter gorge than the middle one, up to 100 m wide and 50 m deep, with walls much less steep and extensively scree-covered. The stream is graded to the top of a waterfall near its mouth, a level determined by ice formerly occupying the shore and the sea.

Coldingham

Moundy deposits of sand, with a minor proportion of gravel, which cover much of the ground for about 2 km inland from Coldingham Bay, are meltwater deposits from the north-west. An exposure shows 2.5 m of cross-bedded sand overlying gravel at the roadside [NT 9022 6637] 440 in NNW of the Priory. North-east of the village [NT 907 663] two roadside sections show, in the higher one 1.5 m of medium and fine sand with some beds of clay and fine gravel, and in the lower the same thickness of medium and coarse bedded gravel. The fields to the east have a flat-topped rounded topography divided by deep, smooth-sided, gently sinuous channels. Cross-bedded sand with small gravel is seen to 4.5 m behind a building [NT 9153 6656]. At the southern end of the bay a conical mound, some 20 m high, is made up of a layered succession including sand, clayey to coarse in texture, and small gravel. All round the Yellow Craig headland thin boulder clay underlies fine sand, the boundary marked by a line of springs, below which swampy hillwash largely conceals the clay.

Two low mounds of fine sand [NT 899 654] south-east of Temple Hall are associated with high erosion features along Abbey Burn. Similar high-level deposits occur south of Bee Edge [NT 9000 6445] and alongside Hill Burn [NT 8945 6745] north-east of Coldingham, where a flat-topped ridge about 4.5 m high is composed of sand capped by gravel.

Grantshouse–Aytonwood

Glaciofluvial gravels, associated locally with meltwater channels, occur at intervals in the Eye valley between Grantshouse and Houndwood. North of the road [NT 8054 6560] 6 m of bedded gravel is seen, with a few large boulders and some thin beds of sand. Close-packed bedded gravel forms a cliff behind the houses to the east. Gravel forming a conspicuous cuspate ridge with a steep southern face is exposed in a railway cutting and was worked [NT 8255 6460] to the east where it overlies boulder clay. The shape of the ridge suggest that it could be a kame-moraine deposited at an ice-front, but it is not essentially different from other high terraced gravels hereabouts. Coarse rounded gravel with a sparse sandy matrix rests on rock in the cutting 250 m to SE [NT 826 643], and 500 m farther to SE gravel and sand are underlain by sandy boulder clay. Sand with gravelly layers is exposed in a stream bank [NT 8326 6370] north of Horseley. In Green Wood gravel terraces occur about 9 and 12 m above the flood-plain. Exposures show ill-sorted and poorly bedded gravel with well laminated, cross-bedded sand.

Low esker-like ridges [NT 901 625], 1.5 km S of Whitfield, are formed of reddish-brown subrounded medium sand and small gravel. Although the ridges are located immediately south of the mouth of a meltwater channel their shape suggests that they are ice-contact deposits, formed within or against ice standing to the cast.

Several mounds and ridges of gravel and sand near Little Dean [NT 921 625] have been worked at several points, the best exposure [NT 9248 6242] being of brown subangular coarse sand and small gravel. The more westerly are kame-like, the more easterly resemble glaciofluvial terraces. The extensive gravels immediately to the south-east are described in the next section.

Meltwater channels

Meltwater channels are abundant in the Lammermuir area. They vary considerably in degree of incision, and their recognition may be hindered by physical insignificance or integration into the present-day drainage system. Many followed lines of glacial erosion. For clarity many are omitted from the published map, but nearly all the southeastward strips of boulder clay would at some time have served as meltwater channels. The direction of flow shown on the map is that of the present-day downward slope, but confinement under hydrostatic head could cause the meltwater to flow uphill. Meltwater flowed generally eastward.

Features of particular interest include a series of parallel channels south-west of Westerside [NT 88 68] with successive courses joined by short lateral links, abandoned high-level courses of the Eye Water west of Houndwood [NT 84 63], the several series of marginal channels, many one-sided, around the western and northern slopes of Horseley Hill [NT 83 62], the prominent channels which cross the watershed [NT 797 594] north of Preston and those north-west of Blackhouse [NT 826 602].

Peat, landslip and raised beach

Peaty soil is common on the uncultivated uplands, but the areas of peat shown on the map generally occupy hollows on or near watersheds or saucer-shaped re-entrants at the heads of streams. Peat cuttings are known only on Drone Moss [NT 844 669] although the track which leads south-east from Long Moss [NT 857 683] is known as the Peat Road. Drone Moss is a flat-topped heathery dome of swampy peat surrounded by a narrow strip of swampy grey clay, and Penmanshiel Moss [NT 826 683] and the moss [NT 843 657] at the head of Atton Burn form similar domes above marginal swamps. The largest area of peat and associated clay is Long Moss [NT 855 685], which occupies a shallow elongate depression with a NW gradient of not more than 1 m 370 (0.27 per cent). The area is very swampy, especially around the little loch to the north-west, and in places the peat seems to be a floating mass no more than 70 or 80 cm thick.

Peat also forms along valleys where streams are no longer active, for example north-west of Grantshouse peat and swampy clay form the flat watershed across which the ancestral Eye Water formerly flowed south-eastwards. The peat lies between two alluvial fans which spread across from the south-west. The NW fan [NT 798 663] defines the present watershed and is a classic example of a corram, that is a watershed fan whose parent stream feeds to one valley or the other according to its varying course across the fan (Kendall and Bailey, 1908). Deposits hereabouts have been considerably disturbed in recent years by realignment of road and railway.

Except in the vicinity of Black Bull [NT 846 703], where large masses of greywacke have slipped down the cliff, landslip is insignificant. Extensive screes occur along the cliffs and locally inland, notably in Dowlaw Dean and west of Horsecastle Bay [NT 916 686].

Evidence of former high sea level is seen on Linkim Shore [NT 923 655], where the storm-beach shingle grades up into a grass-covered shingle flat which extends above the 25-ft (7.6 m) contour. The sands at Coldingham Bay are backed at their northern end by a flat beach some 20 m wide, the former shoreline back-feature coinciding approximately with the 25-ft (7.6-m) contour. Shingle beaches at Horsecastle Bay [NT 918 685] and Burnmouth Harbour [NT 919 683] merge inland with a flat grassy tract to which alluvium is graded from the north-west. The tract is composed of coarse well-rounded shingle, and lies between the 25-ft and 50-ft contours (7.6 to 15.2 m).

Upper Whiteadder And Lower Eye valleys

This area extends from the south-west corner of the district to the sea at Eyemouth, forming low-lying. ground largely drift covered. The upper Whiteadder is separated from the lower by a short gorge below Chirnsidebridge [NT 855 555], and was formerly linked to the Eye Water near Ayton via the misfit valley of Billiemire Burn.

The boulder clay is predominantly red or reddish brown and sandy, with many small erratics, giving rise to reddish or brown stony soils. Over the Old Red Sandstone the abundance of locally derived fragments of red sandstone in the lower part of the boulder clay obscures the distinction between it and sandstone fragmented in situ. Near Stoneshiel [NT 871 603] the boulder clay soil is locally full of angular pebbles of greywacke, taken to indicate an unexposed inlier of Silurian rock. On the lower slopes south-east of Ferney Castle [NT 88 60] the boulder clay soil is unusually gravelly.

Upper Whiteadder

Boulder clay up to 12m thick is seen by the Whiteadder and the Eye. At Slighhouses water bore [NT 8262 5922], drift 16.5 m thick was reported. The gravels which cover a large area [NT 84 57] east of Hammerhall appear to be both overlain and underlain by similar boulder clay.

Late-glacial gravels on the right bank of the Whiteadder Water extend from Oxendean [NT 77 56] nearly to Chirnsidebridge. A unique feature of these gravels is the esker of Oxendean, a discontinuous train of sharp ridges of gravel which can be followed from the edge of the district [NT 7691 5567] 750 in SW of Oxendean Tower for 1350 m to the north-east. Oxendean Burn cuts a gorge through the esker. Gravel ridges west of Hen Poo may also be eskers. West of the deep meltwater channel south of Broomhouse Mains [NT 808 558], the gravel terrain comprises many low mounds separated by a network of meltwater channels with clay floors. To the east, the topography is gently undulating and dissected by more widely spaced channels and small streams.

The general level of the gravel surface around Edrom and East Blanerne is 15 m above the flood-plain of the Whiteadder. On the left bank of the Whiteadder [NT 8483 5677] 430 m W of Chirnside station, 4.5 m of bedded gravel overlies solid rock. Southwards of East Blanerne there are three distinct terraces, the highest about 11 m above the gravelly flood-plain, but immediately north-east of the farm a terrace-like flat lies at a higher level, 17.5 m above the flood-plain of the Billiemire Burn. On the eastern side of this burn gravels are seen up to about the 76-m contour, as around Edrom, with one patch above 91 m. On this relatively steep west-facing slope they are considerably dissected, and the gravel mounds are interspersed with steep faces in solid rock and swampy clay-filled hollows. Traces of a former high-level course of the Whiteadder south-eastwards from Craigswalls [NT 84 55] to Allanton [NT 86 54] are represented by several small patches of gravel and fluvial channels north-west of Allanbank.

Billiemire Burn

The broad deep valley of Billiemire Burn and its copious deposits of gravel are evidence of its importance as a north-eastward channel for the Whiteadder glacier and its meltwater. Discontinuous ridges and mounds [NT 84 59] west-north-west of Billie Mains form gravel eskers deposited by englacial streams flowing ESE. Much red sandstone debris is mixed with rounded erratics. A pit [NT 8534 5878] exposes:

Thickness m
Gravel, poorly bedded, with flat-lying pebbles up to 0.15 m, and a few larger ones 1.2
Sand, brown and coarse above, pink and silty below; undulating and dipping gently east 0.15
Gravel as above, locally with very little matrix; some irregular blocks, up to 0.3 m in the lower half 2.4

Many of the pebbles are subangular, showing no preferred orientation and poor size-sorting. About 1 km to NE, thin dissected bodies of gravel are underlain by boulder clay. A terrace 7 m above the Billiemire Burn floodplain has a gentle eastward gradient.

On the south side of the valley gravels are banked up almost to the 107-m contour at their western end. The highest deposits are moundy but terracing is developed between 6 and 15 m above the flood plain. The general orientation of erosional and depositional features is ENE. The gravel pit at Causewaybank [NT 880 593] is in gravel, with interbedded sand making up about one-third of the deposit. The coarser gravels include cobbles up to 0.23 1n across, but boulders 0.6 m across have been extracted, though not seen in place. Bedding generally shows ripple-lamination and planar cross-bedding which dip to E or SE. In places beds are strongly lenticular and in addition lithological boundaries may be near vertical with contorted beds and small steep faults. The general environment of deposition is fluvial, with shifting channels and contemporaneous erosion and deposition. On the eastern face of the pit a channel about 30 in wide, filled with sand and fine gravel, cuts down about 2.5 m into the sand and coarse gravel below. The deposits have all the characteristics of kames which accumulated adjacent to ice filling the valley to the west.

Eye valley

The gravels of the Eye valley, from Coveyheugh [NT 869 624] to the sea, are terraced deposits. Terraces occur at two levels opposite Reston, 8 and 11 m above the floodplain, where a bore [NT 8860 6239] proved coarse bedded gravel 6 m thick. Downstream the higher terrace increases its relative elevation. Small exposures of bedded gravel are seen but the hacks of both terraces are followed by broad clay-filled depressions up to 1.5 m deep. North-east of East Reston [NT 903 613] both terraces are dissected by meltwater channels. North of Aytonlaw [NT 915 619] the upper terrace marks a former course of the Eye Water through a narrow rocky channel. North-east of Ayton the terracing is less clear and the higher deposits are dissected by meltwater channels. Deposits seen to 8 m [NT 9297 6167] 220 m N of Ayton Castle range from fine clayey sand to coarse gravel. Cross-bedding, asymmetric ripples, upward fining, wedge-bedding, angular discordance, irregular bedding junctions, disruption of beds and stepped boundaries are common features, with depositing currents flowing east and south.

Moundy gravels which extend well up the slopes around Huldies Park [NT 926 618] may be genetically linked to the adjacent deposits of Little Dean and derived from the north-west. The broad area of generally flat ground around Ayton Mains [NT 935 622] is a glaciofluvial terrace more than 30 m above the floor of the gorge. North of Old Linthill a sloping moundy sand and gravel terrace is 15 m above the flood-plain, banked against a mound, expanse of sand, extending about 1 km to west and formed of outwash down the Ale valley impounded against ice in the Eye valley. The only good sections are on the left bank of the Ale Water. A borehole [NT 9348 6298] proved 2.5 m of very fine sand underlain by 1.2 m of pebbly coarse sand. In the Ale Water [NT 9305 6280] 400 m SE of Linthill 10 m of red boulder clay with a 2.5 m bed of sand and gravel near the base is capped by 2 m of gravel.

North of Biglawburn [NT 938 634] the left-bank gravels are generally terraced. Sand and gravel 2 m thick, overlying boulder clay, was seen at the coastguard station [NT 942 645], but the terrace-like flat to north-west appears to be cut in gravelly clay.

On the cliffs to north [NT 9425 6475] boulder clay overlies gravel filling a channel, which may therefore be pre-Devensian. One section is as follows:

Thickness m
Boulder clay up to 3
Gravel, coarse and angular at base, faint stratification dips to N 4
Sand, medium, subrounded, closely and regularly stratified, isolated pebbles. Dip about 5° WNW 3
Gravel, coarse, especially at base, doubtful stratification dipping SE. Some boulders to 0.45 m, moderately rounded 3
Solid rock (Lower Devonian lava) 2

On the right bank of the Eye Water, gravelly mounds near Loanside [NT 870 615] may be kame deposits from ice standing in the low ground south-west of Reston. Gravel and boulder clay appear to be interbedded in the railway cutting [NT 8690 6130] 400 m W of Newmains, where 0.75 m of medium gravel lies within red boulder clay, and appears to be followed about 1 m higher by more gravel. From Reston eastwards terrace deposits are similar to those on the north bank. About 400 m WNW of Reston station by the Eye Water [NT 8739 6217] gravel 4.5 m thick overlies 12 m of red boulder clay. Between 350 and 600 m downstream boulder clay overlies gravel and sand. West of Swinwood Mill [NT 887 620] 7.5 m of gravel is exposed in the river bank. South-eastwards of East Reston [NT 903 607] the broad undulating flat of the main terrace is traversed by a network of ice-wedge polygons, a product of frost-contraction of the soil and infilling of the cracks with material of different hydrological properties (Greig, 1981). At the bridge [NT 8830 6170] 600 m E of Reston station fluvial deposits are apparently banked against boulder clay lying to the south. The section showed:

Thickness m
Gravel, with cobbles at base up to 0.10 m across 0.9
Sandy stony clay, wedging out to NW 0.75
Gravel, fine in the top 0.3 m (base not seen) 3.7

In the cutting [NT 9383 6138] 600 m W of Fairnieside, 1.5 m of fine sand is overlain by 3.7 m of boulder clay, the sand apparently an englacial fluvial deposit related to drainage towards Burnmouth. At Eyemouth terraces of gravel up to 10 m above the flood-plain, have clayey hollows near their outer margins. Small areas of sand and gravel north and south of Gunsgreenhill may be relics of a former high-level course of the Eye Water. Realignment of the road south of Eyemouth exposed sections up to 2.5 m high which included bedded sands and gravels, sandy clay, and two sandy boulder clays: a stiff clay with much angular greywacke debris and a sandier clay with smaller and more rounded erratics. The inter-relationships of the deposits were considerably disturbed, probably by frost-action, and no general pattern was apparent.

Originally the terraced cliff-top gravels at Eyemouth were thought to be raised beach deposits, albeit at an unexpectedly high level. The apparent absence of shell debris from them and their extensive persistence landward along the Eye valley suggest that they are essentially fluvial deposits, impounded to their 23-m level by coastal ice.

Meltwater channels and peat

Several meltwater channels cut across the ridge near Stoneshiel and Ferney Castle [NT 87 60]. Some can be followed as benches along the contours on the northward slope, crossing the ridge at successively lower levels eastwards, to be transformed into well-defined sinuous channels descending south-eastward towards Billiemire Burn. Many are between 3 and 4.5 m deep, an example [NT 8752 6037] 300 m E of Stoneshiel being 60 m wide and nearly 4 m deep. Another well formed channel is that north of Primrosehill [NT 784 578] which marks a former south-eastward course of the Whiteadder.

Large areas of peat lie in the Billiemire valley. That south of Billie Mains [NT 854 583] may mark a former watershed between the Eye and the Whiteadder. The 3-km stretch in the valley north of Causewaybank [NT 880 593] includes the present imperceptible watershed. A shallow borehole towards the west end of the peat [NT 8695 5944] proved it to be 2.7 m thick, underlain by at least 1.5 m of fine sand, sparsely pebbly.

Chirnside–Burnmouth upland

In this area extending from Chirnsidebridge [NT 85 56] to the sea, drift deposits are thin and consist largely of boulder clay with small patches of peat and alluvium.

The boulder clay reflects the nature of the underlying rock, and is a reddish brown sandy clay with many small stones. Over the Old Red Sandstone it is more sandy. In the coastal strip south of Burnmouth it is generally very sandy and unpredictably interbedded with sand and gravel. Inland sections are poor, the best being in the streams [NT 8909 5860] just south of Blackburn.

Sand and gravel

Significant deposits of sand and gravel are confined to a narrow coastal strip and appear to be the product of meltwater drainage between high ground to the west and sea-ice to the east. Knolls and ridges within and west of a col [NT 9515 6225], just above the 76-m contour, 600 m SW of Fancove Head appear to be derived from eastward drainage through the col, but gravel knolls [NT 945 622] above Redhall may be ice-contact kame deposits. In a pit [NT 955 613] on Burnmouth Hill the sand is rather clayey with rounded and angular pebbles. Sections at the cliff-top at Ross Point [NT 965 605] show 9 m of gravelly clay and, nearby, upwards of 0.6 m of sand with small modern shells. These sections are typical of others along the cliff-tops and in the railway cuttings between Burnmouth and Berwick, in respect of the close inter-relationship of till-like and fluvial materials and of the presence at these unexpectedly high levels, up to 60 m above OD, of sea-shells. These are mostly modern mussels and other bivalves and limpets, which have probably been transported by man in cultivating the land, or possibly by seabirds. They have not been found convincingly buried in undisturbed natural deposits.

Southwards from Catcairn Bushes [NT 967 596] many ridges of sand and gravel, oriented SSE, have been worked in places, notably west of the road [NT 9650 5930] above Hilton Bay. The group of ridges at Lamberton are similar in form to adjacent ridges of solid rock, both having been shaped by meltwater flowing south-east off Lamberton Moor. Exposures by the main road demonstrate the close association of sands and clays, ascribed to the partial fluvial resorting of glacial till. Lamberton Farm No. 1 Borehole [NT 9714 5764], proved gravel with some sand to a depth of 17.02 m. Within a meltwater channel flanked by solid rock 160 m to N excavations exposed 3 m of coarse sand with boulders and some thin seams of clay.

Meltwater channels

The map demonstrates the marked topographic difference between the ground north of Clappers [NT 945 575] which is deeply dissected with sharp drift boundaries, and the rest of the Chirnside–Burnmouth area where in general the topography is smoothly rounded and boundaries between drift and solid rock are not sharply defined, and are therefore generalised on the map. The pattern of meltwater channels shows that around Whiterig [NT 920 585] the grain of the country is mainly ENE, while from Bastleridge [NT 937 593] to Lamberton it is SE. In the deeply dissected area north of Clappers, however, the drift-filled depressions fall generally to SW. This topography is believed to have been initiated by north-eastward surge of ice from the lower Tweed basin. Progress of this ice was halted by the south-eastward ice-stream across Lamberton Moor. An extensive area of peat and swampy alluvium [NT 955 583] south of the Moor resulted from the ice-obstruction to the south, the meltwaters eventually overflowing east and south-east at a col [NT 9604 5780] 850 m NW of Lamberton church. Near the southern limit of this swampy area [NT 956 577] three ridges and cones of bouldery gravel are interpreted as terminal moraines of the southern ice. The lobe may have been wider, to the south-east, than the area of dissected topography, in which case the restriction of this topography must be related to the greater heterogeneity of the solid rocks there and to more effective post-glacial erosion to the southeast, where the smoother terrain is traversed by many broad meltwater channels.

Other examples of ice-impeded drainage, due to ice in the Billiemire–Horn Burn valley, are the areas of peat and swampy clay south-west [NT 890 582] and east [NT 900 590] of Blackburn. Both may have originated at about the same time, the latter being at a slightly lower altitude. A generally flat swampy bench, with shallow valley-parallel drainage channels, forms the south side of the Billiemire valley for about 2 km E of the Causewaybank gravels. Ice-obstruction may also be invoked to explain the formation of peat and clay swamps on Lamberton Moor [NT 959 589] and [NT 954 597].

The persistent meltwater channels north of Chirnside, at Whiterig [NT 920 585], and west of Lamberton Moor follow glacially scoured depressions. The col [NT 945 591] followed by the road south of Ayton Hill is parallelled on both sides by linear SE channels and benches at several levels, up to the summit of the hill itself. In the general area of Edingtonhill [NT 897 573] and elsewhere channels can be traced from the Billiemire valley south-east across the high ground and down towards the Whiteadder. Such 'up and over' channels will have been cut by water under hydrostatic pressure beneath the ice.

Landslip

Landslip is a feature of the Carboniferous sea-cliffs from Burnmouth southwards. Recent slips at Burnmouth have damaged property, and one at Catcairn Bushes necessitated realignment of the railway. The only significant body of scree is on the north-eastern face [NT 953 603], of Chester Hill, below the cliff of Old Red Sandstone conglomerate.

Lower Whiteadder Valley

This area, at the southern edge of the district, comprises the Whiteadder valley below Chirnsidebridge, and is a tract of low relief, covered largely by boulder clay. The valley of the Whiteadder Water in this reach appears to be a direct continuation of that of the Blackadder Water, whose course for some 18 km above their confluence at Allanton [NT 864 546] is steadily east-north-eastward. Four alluvial terraces are recognised near Allanton, up to 15 m above the flood-plain. Gravel is the major constituent except at the highest level, where clay predominates. Downstream the terracing is less clearly developed. East of Hutton Castle [NT 89 55] the river flows in a rock-walled gorge 30 m deep.

The boulder clay is predominantly reddish and sandy, but over the Cementstone Group outcrop it is browner and heavier. Close to solid it is locally green or greenish grey. In places it is notably gravelly, for example on a ridge [NT 859 551] just south of Stuartslaw, and at the top of the left-bank cliff [NT 8656 5523] due east of that farm where 3.7 m of gravel, unsorted and unstratified, is exposed. Nearby to the north-east 7.6 m of gravelly drift is seen. Exposures of thick boulder clay are 4.5 m at the cliff-top [NT 8513 5552] south of Newstead, and 6 m in an old quarry [NT 8895 5586] west-north-west of Edington Mains.

Ridges of gravel [NT 870 554] north of Broadhaugh suggest deposition from the glaciofluvial Whiteadder. Moundy gravels [NT 883 546] west of Hutton Castle Barns are probably kame deposits which accumulated close to the front of the Blackadder glacier. In Hutton Castle Barns (1927) water bore [NT 8913 5407], sunk through alluvium adjacent to the gravel, the superficial deposits consisted of 1.2 m of sandy clay underlain by 5.4 m of sand on a basal gravel 1.9 m thick.

Many meltwater channels dissect the slopes between Chirnside and the Whiteadder. Some are clearly courses of the ancestral river which it followed before cutting the gorge, whereas others drained the interfluvial ridge between the Billiemire and the lower Whiteadder. The dissected terrain resembles small-scale drumlin topography, especially in the area between Chirnside, Edingtonhill [NT 898 572], and Edington Mains [NT 897 556], but the relative importance of glacial moulding and glaciofluvial erosion is not altogether clear.

On the south bank of the Whiteadder, eastward channels predominate, many of them containing grey and dark brown alluvial clay, which in the past has been used for brickmaking. The only brickworks within the district [NT 937 540], 1 km E of Clarabad, lies in a broad hollow which has been cut through by the Whiteadder. This area is on the northern fringe of the lowland district of Berwickshire known as the Merse, the name being a form of the word 'meres' and derived from the profusion of small lochs which in earlier times were a dominant feature of the landscape.

The walls of the lower Whiteadder valley, steep cliffs of thinly bedded rocks of alternating lithologics capped by sandy boulder clay, are subject to landslips, which are numerous but mostly too limited in area to be shown on the published map.

Evolution of the river system

The river system has evolved over many geological periods, in response to changes in climatic and topographic circumstances, but to avoid unwarranted speculation, only late Quaternary events and the main rivers are considered.

At least four major anomalies are recognised in the present drainage pattern (Figure 23). These are the broad flat valley-watersheds in the Eye valley north-west of Grantshouse [NT 799 662] and in the Billiemire valley south-west of Ayton [NT 880 594], the deep drift-filled valley north-west of Drakemire [NT 795 616], and the eastward loop of the Whiteadder from Ellemford [NT 730 600] via Abbey St Bathans [NT 755 625] (both on Sheet 33E) and Elba [NT 786 605] to Preston Haugh [NT 774 576]. As the inset on the published map shows, it is believed that ice-streams were responsible for significant erosion of the three anomalous watersheds. Prior to glaciation the Grantshouse gap formed part of the course of an ancestral Eye Water which originated near the present source of Pease Burn, west of Ewieside Hill [NT 763 680], and at a former watershed [NT 795 683] east of that hill. Similarly the Billiemire gap was part of a pre-glacial system which fed the Eye from the high ground to the west and possibly also from the Whiteadder west of Edrom [NT 830 560]. From Ellemford to Preston it appears that the Whiteadder might have taken a direct ESE course through the Windshiel gap [NT 744 587], but rockhead elevations make this suggestion untenable, though the gap may have been deepened by glacial action. The Whiteadder between Abbey St Bathans and Elba is a continuation of the Monynut Water valley, but its further progress south-eastwards is obstructed by a high ridge of Silurian rock. In glacial times, and perhaps earlier, the main outflow from the Elba area may have been north-east towards Grantshouse, through the Drakemire gap.

At some stage in the glaciation a major ice-stream from the valley of the Whiteadder moved eastwards through the Billiemire gap, rather than south-east along the present course towards Allanton. Evidence in support of this hypothesis is afforded by the extensive spread of meltwater gravels around East Blanerne [NT 850 570] and Causewaybank [NT 878 590] and by the nature and degree of erosion across the Billiemire–Horn Burn watershed. The gorge between Chirnsidebridge and Allanton, where the Whiteadder and Blackadder join, suggests that there may have been a low divide here in pre-glacial times between these two rivers, but the evidence is by no means convincing. Above their confluence, for many kilometres, the two courses have almost identical profiles and there is no significant break of slope along the valley of the Whiteadder (Figure 24).

Glaciation of the Eye valley, as well as deepening and reshaping the watershed area near Grantshouse, also deepened the Ayton–Burnmouth valley, but there is no indication of significant fluvial activity along this line, either before or after glaciation.

During deglaciation the presence of remanent ice and of glacial deposits in some of the major valleys obstructed the dispersal of meltwater along them. Thus the Billiemire valley could not transmit all the Whiteadder water, which found its way south-eastward by a succession of channels from the vicinity of Billie Mains [NT 851 588] towards Foulden and in time, as the valley-ice wasted, finding and deepening the present course from Chirnsidebridge to Allanton. The widespread gravels in the Edrom area suggest that for a time water from the Blackadder valley from the south-west was fed towards Blanerne [NT 83 56] and the Billiemire gap. Obstruction of the Eye valley west of Houndwood [NT 842 639] led to the choking of the Drakemire valley and to the waters of the Monynut escaping south-west towards the White-adder. The profiles of the rivers suggest that a considerable deepening of the Abbey St Bathans loop of the Whiteadder has occurred since pre-glacial times and that its establishment was, therefore, a prolonged process.

During the early stages of deglaciation meltwater flowed into the Eye valley round the northern side of Ewieside Hill [NT 780 690], but as the ice to the north wasted downwards the flow was diverted north-east towards Pease Bay. The short steep course of this stream enabled it to cut back rapidly towards the Eye valley, resulting in the capture of the latter's head-stream from Edmond's Dean [NT 790 670] and the ultimate establishment of the flat watershed between there and Grantshouse.

The Eye gorge below Ayton is largely post-glacial, formed in response to falling base-levels as sea-ice wasted and, later, as sea level fell. The presence of boulder clay within it, mainly in the lower reaches, is proof that a valley existed prior to the last glaciation, and this evidence is supplemented by the pre-boulder clay fluvial gravels below Eyemouth Fort. A pre-glacial buried channel of the Eye is thought to be present, largely beneath the left bank, between Ayton and the sea, crossing the Ale valley between 700 and 1000 m west of its mouth. There is evidence of a buried channel within Eyemouth harbour at a depth of 12 m below OD, possibly associated with a rock-platform at 18 m BOD, beneath the sea at Burnmouth, ascribed to a late-glacial episode of low sea level, the Loch Lomond Re-advance.

Chapter 7 Mineral and groundwater resources

Metalliferous minerals

Unsuccessful trials for copper were made around 1760 and 1825 at Hoardweel on the east bank of the Whiteadder Water opposite Elba [NT 7867 6038] where three levels were driven north-north-eastward. The ores were chalcopyrite and tetrahedrite. (Milne, 1837; Stevenson, 1849; Wilson 1921).

Industrial minerals

Dolerite is worked in Borthwick Quarry [NT 770 545] in the south-west corner of the district by Wimpey Asphalt Ltd for roadstone and aggregate. Sand and gravel is dug in Causewaybank Pit [NT 880 593]. These appear to be the only working quarries.

Throughout the district there are numerous small quarries where greywacke, sandstone, lava, limestone or sand and gravel were dug in the past for local purposes. None attained sufficient reputation to be developed for a wide trade. A small quarry [NT 7751 5935] 400 m SW of Cockburn East was recently opened for a short time in diorite in the Cockburn Law intrusion.

Groundwater

Groundwater from springs and shallow wells satisfied the needs of many farms and isolated communities, but in recent years most of these have been given mains supplies. There are no major aquifers, though groundwater is obtainable from most formations, the Upper Old Red Sandstone being the best. Recharge to bedrock aquifers is restricted by widespread boulder clay. The public water supply is derived from upland gathering grounds, small river intakes, some springs and one borehole. Two springs at Ayton issuing from drift over Lower Old Red Sandstone and Silurian have a combined capacity of 0.3 million litres per day (Ml/d) and the Slighhouses bore [NT 8262 5922] in Upper Old Red Sandstone provides a further capacity of 0.4 Ml/d.

Groundwater in Silurian strata is limited to storage and transport within open cracks and joints. Supplies are small although water quality is good. In the Lower Devonian tuffs and lavas groundwater storage and transport is similarly in open cracks and joints. The Lower Old Red Sandstone has an overall permeability of the order 10-2 m/d. Sustainable borehole yields of only 0.5 l/s and 1.8 l/s were obtained from 50 m deep test boreholes at Reston and Ayton [NT 8860 6239] and [NT 9172 6153] respectively. The Upper Old Red Sandstone offers slightly more favourable conditions, although fissure flow is still important and storage within the pores of the rock limited. A 120 m deep test borehole at Slighhouses [NT 8262 5922] indicated a permeability of 10-1 m/d and storativity of 10-4, with a sustainable yield of 6 l/s. The groundwater is moderately mineralised with a bicarbonate concentration of about 240 mg/l.

Carboniferous strata are poor aquifers. Sustainable yields of only 0.5 l/s are recorded from various locations in boreholes of 30 to 75 m depth. Impermeable mudstone horizons restrict vertical groundwater movement and overall permeability is low (perhaps 10-2 m/d or less). Water quality is poor, with a high concentration of bicarbonate and other ions, making it generally unsuitable for consumption.

Groundwater is present in appreciable quantity in some of the superficial deposits such as alluvium and glacial sand and gravel. Water quality is favourable except where deeper water from bedrock is in hydraulic contact with the shallow aquifers.

References

ALMOND, J. E. 1985. A vermiform problematicum from the Dinantian of Foulden, Berwickshire , Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 41–47.

ANDERTON, R. 1985. Sedimentology of the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 7–12.

ANDREWS, S. M. 1978. A possible occurrence of Remigolepis in the topmost Old Red Sandstone of Berwickshire. Scott. J. Geol., Vol. 14, 311–315.

ANDREWS, S. M. 1985. Rhizodont crossopterygian fish from the Dinantian of Foulden, Berwickshire , Scotland, with a re-evaluation of this group. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 67–95.

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. Geol. J., Vol. 21, 101–125.

BELT, E. S., FRESHNEY, E. C. and READ, W. A. 1967. Sedimentology of Carboniferous cementstone facies, British Isles and eastern Canada. J. Geol., Vol. 75, 711–721.

BOUMA, A. H. 1962. Sedimentology of some flysch deposits. (Amsterdam: Elsevier.) 168 pp.

BRIGGS, D. E. G. and CLARKSON, E. N. K. 1985. Malacostracan crustacea from the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 35–40.

BROWN. G. C., FRANCIS, E. H., KENNAN, P. and STII.I.MAN, C. J. 1985. Caledonian igneous rocks of Britain and Ireland. Mem. Geol. Soc. London, No. 9, 1–15.

CAMERON, I. B. and STEPHENSON, D. 1985. British regional geology: The Midland Valley of Scotland. (London: HMSO for British Geological Survey.)

CARTER, A. V. F. 1966. Records of wells in the areas of Scottish one-inch geological sheets Peebles (24), Kelso (25), Berwick-upon-Tweed (26), Haddington (33), Eyemouth (34), and North Berwick (41, East Lothian portion). Water Supply Pap., Geol. Sure. GB.

CASEY, D. M. 1983. Geological studies in the Central Belt of the eastern Southern Uplands of Scotland. Unpublished DPhil thesis, University of Oxford.

CASEY, D. M. and OLIVER, G. J. H. 1984. Coldingham Beds and Linkim Beds. In OLIVER, G. J. H. and others, 1984, 251–253.

CLARK, R. H. 1956. A petrological study of the Arthur's Seat Volcano. Trans. R. Soc. Edinburgh, Vol. 63, 37–70.

CLARKSON, E. N. K. 1985. Palaeoecology of the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 97–100.

CLAYTON, G. 1971. A Lower Carboniferous miospore assemblage from the Calciferous Sandstone Measures of the Cockburnspath region of eastern Scotland. Pollen & Spores, Vol. 12, 577–600.

CLAYTON, G. 1985. Plant miospores from the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 21–24.

CLOUGH, C. T., BARROW, G., CRAMPTON, C. B., MAUFE, H. B., BAILEY, E. B. and ANDERSON, E. M. 1910. The geology of East Lothian. Mem. Geol. Sure. GB, (Scotland, sheet 33).

COCKS, L. R. NI., HOLLAND, C. H., RICHARDS, R. B. and STRACHAN, I. 1971. A correlation of Silurian rocks in the British Isles. Spec. Rep. Geol. Soc. London, No 1. (J. Geol. Soc. London, Vol 127, 103–136).

COOPER, D. C., PARKER, M. E. and ALLEN, P. M. 1982. Investigations of small intrusions in southern Scotland. Min. Reconn. Prog. Rep. Inst. Geol. Sci., No. 58.

CRAIG, G. Y. (Editor). 1983. Geology of Scotland. (Edinburgh: Scottish Academic Press.) 472 pp.

CRAIG, G. Y. and WALTON, E. K. 1962. Sedimentary structures and palaeocurrent directions from the Silurian rocks of Kirkcudbrightshire. Trans. Edinburgh Geol. Soc., Vol. 19, 110–119.

CURRIE, E. D. 1954. Scottish Carboniferous goniatites. Trans. R. Soc. Edinburgh, Vol. 62, 527–602.

DAVIES„A., McADAm, A. D. and CAMERON, I. B. 1986. Geology of the Dunbar district. Mem. Geol. Sure. G. B. (Scotland, sheet 33E).

DEARMAN, W. R., SWELLS, K. A. G. and LARwoon, G. P. 1962. Refolded folds in the Silurian rocks of Eyemouth, Berwickshire. Proc. Yorkshire Geol. Soc., Vol. 33, 273–285.

DEFFEYES, K. S., LUCIA, F. J. and WEYL, P. K. 1965. Dolomitisation of recent and Plio-Pleistocene sediments by marine evaporate waters on Bonaire, Netherlands Antilles. Spec. Pub. Soc. Econ. Palaeont. Miner., No. 13, 71–88.

DE SOUZA, H. A. F. 1979. The geochronology of Scottish Carboniferous volcanism. Unpublished Ph.D. thesis, University of Edinburgh.

DZULYNSKI, S. and WALTON, E. K. 1965. Sedimentary features of flysch and greywackes. In Developments in sedimentology, Vol. 7. (Amsterdam: Elsevier.) 274 pp.

EDEN, R. A., CARTER, A. V. F. and MCKEOWN, M. C. 1969. Submarine examination of Lower Carboniferous strata on inshore regions of the continental shelf of south-east Scotland. Marine Geol., Vol. 7, 235–251.

EDEN, R. A., SMALL, A. V. F. and MCQUILLIN, R. 1970. Preliminary report on marine geological and geophysical work off the east coast of Scotland, 1966–1968. Rep. Inst. Geol. Sci., No. 70/1.

FITCH, F. J., MILLER, J. A. and WILLIAMS, S. C. 1970. Isotopic ages of British Carboniferous rocks. C. R. 6me. Congr. Int. Stratigr. Geol. Carbonif. (Sheffield, 1967), Vol. 2, 771–789.

FOREY, P. L. and YOUNG, V. T. 1985. Acanthodian and coelacanth fish from the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 53–59.

FOWLER, A. 1926. The geology of Berwick-on-Tweed, Norham and Scremerston. Mem. Geol. Surv. G.B. (England, sheets 1 and 2).

FRANCIS, E. H. 1983. Carboniferous and Carboniferous-Permian igneous rocks. Pp.253–324 in Geology of Scotland. CRAIG, G. Y. (Editor). (Edinburgh: Scottish Academic Press.)

FROST, D. V. and HOLLIDAY, D. W. 1980. Geology of the country around Bellingham. Mem. Geol. Surv. G.B. (England, sheet 13).

GARDINER, B. G. 1985. Actinopterygian fish from the Dinantian of Foulden, Berwickshire , Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 61–66.

GEIKIE, A. 1863. The geology of eastern Berwickshire. Mem. Geol. Surv. G. B.

GEIKIE, A. 1887. Geology and petrology of St. Abbs Head. Proc. R. Soc. Edinburgh, Vol. 14, 177.

GEORGE, T. N., JOHNSON, G. A. L., MITCHELL, M., PRENTICE, J. E., RAMSBOTTOM , W. H. C., SEVASTOPULO, G. D. and WILSON, R. B. 1976. A correlation of Dinantian rocks of the British Isles. Spec. Rep. Geol. Soc. London, No. 7.

GREIG, D. C. 1971. British regional geology: The south of Scotland. (London: HMSO for British Geological Survey.)

GREIG, D. C. 1981. Ice-wedge cast network in eastern Berwickshire. Scott. J. Geol., Vol. 17, 119–121.

HEWARD, A. P. 1981. Lower Carboniferous fluvial facies in Northumberland and Borders. In Field guide to the modern and ancient fluvial systems in Britain and Spain. ELLIOTT, T. (Editor). (Keele.)

HOME, D. M. 1837. The geology of Berwickshire. Trans. Highland Soc., Vol. 5 n.s., 247.

HOUSE, M. R., RICHARDSON, J. B., CHALONER, W. G., ALIEN, J. R. L., HOLLAND, C. H. and WESTOLL T. S. 1977. A correlation of Devonian rocks of the British Isles. Spec. Rep. Geol. Soc. London, No. 7

ILLING, L. V., WELLS, A. J. and TAYLOR, V. C. M. 1965. Penecontemporary dolomite in the Persian Gulf. Spec. Rep. Soc. Econ. Palaeont. Miner., No. 13, 89–111.

INSTITUTE OF GEOLOGICAL SCIENCES. 1969. Annual Report for 1968. (London: HMSO.)

INSTITUTE OF GEOLOGICAL SCIENCES. 1981. Annual Report for 1979. (London: HMSO.)

IRVING, J. 1930. Four 'felstone' intrusions in central Berwickshire. Geol. Mag., Vol. 67, 529.

KELLING, G. 1962. The petrology and sedimentation of Upper Ordovician rocks in the Rhinns of Galloway, south-west Scotland. Trans. R. Soc. Edinburgh, Vol. 65, 107–137.

KELLING, G. 1964. The turbidite concept in Britain. Pp.75–92 in Turbidites. BOUMA, A. H. and BROUWER, A. (Editors). (Amersterdam: Elsevier.) 264pp.

KENDALL, P. F. and BAILEY, E. B. 1908. The glaciation of East Lothian south of the Garleton Hills. Trans. R. Soc. Edinburgh, Vol. 46, 1–31.

LAGIOS, E. and HIPKIN, R. G. 1979. The Tweeddale Granite - a newly discovered batholith in the Southern Uplands. Nature, London, Vol. 280, 672–675.

LEEDER, M. R. 1971. Initiation of the Northumberland basin. Geol. Mag., Vol. 108, 511–516.

LEEDER, M. R. 1974. Lower Border Group (Tournaisian) fluvio-deltaic sedimentation and palaeogeography of the Northumberland basin. Proc. Yorkshire Geol. Soc., Vol. 40, 129–180.

LEEDER, M. R. 1974. The origin of the Northumberland basin. Scott. J. Geol., Vol. 10, 283–296.

LEGGETT, J. K., MCKERROW, W. S. and CASEY, D. M. 1981. The anatomy of a Lower Palaeozoic accretionary fore-arc: the Southern Uplands of Scotland. Pp.495–520 in Trench fore arc geology. LEGGETT, J. K. (Editor). Spec. Pub. Geol. Soc. London, No. 10.

LEGGETT, J. K., MCKERROW, W. S. and EALES, M. H. 1979. The Southern Uplands of Scotland: a Lower Palaeozoic accretionary prism. J. Geol. Soc. London, Vol. 136, 755–770.

LUMSDEN, G. I., TULLOCH, W., HOWELLS, M. F. and DAMES, A. 1967. The Geology of the neighbourhood of Langholm. Mem. Geol. Sure. G.B., (Scotland, sheet 11.)

MACGREGOR. A. G. 1928. The classification of Scottish Carboniferous olivine-basalts and mugearites. Trans. Geol. Soc. Glasgow, Vol. 18, 324–360.

MACKENZIE, D. H. 1956. A structural profile south of Eyemouth, Berwickshire. Trans. Edinburgh Geol. Soc., Vol. 16, 218- 253.

McKERROW, W. S., LEGGETT, J. K. and EALES, M. H. 1977. Imbricate thrust model of the Southern Uplands of Scotland. Nature, London, Vol. 267, 237–239.

MIDGLEY, H. G. 1946. The geology and petrology of the Cockburn Law intrusion, Berwickshire. Geol. Mag., Vol. 83, 49–66.

MILES, R. S. 1968. The Old Red Sandstone antiarchs of Scotland: Family Bothriolepididea. Palaeontogr. Soc. [NT Monogr.]

MITCHELL, G. F. 1948. Late-glacial deposits in Berwickshire. New Phytol., Vol. 47, 262.

MOLYNEUX, S. G. 1981. Silurian acritarchs from the Linkim Beds. Rep. Br. Geol. Sure., No. PDL/81/138. (Unpublished.)

MOLYNEUX, S. G. In press. Acritarchs from the Linkim Beds (Silurian) of the Southern Uplands.

MOY-THOMAS, J. A. 1938. Carboniferous Palaeoniscids from Northumberland and Berwickshire. Geol. Mag., Vol. 75, 308–318.

MYKURA, W. 1983. Old Red Sandstone. Pp.205–251 in Geology of Scotland. CRAIG, G. Y. (Editor). (Edinburgh: Scottish Academic Press.)

NEVES, R., GUEINN, K. J., CLAYTON, G., IOANNIDES, N. S., NEVILLE, R. S. W. and KRUSZEWSKA, K. 1973. Palynological correlations within the Lower Carboniferous of Scotland and Northern England. Trans. R. Soc. Edinburgh, Vol. 69, 23–70.

OLIVER, G. J. H., SMELLIE, J. L., THOMAS, L. J., CASEY, D. M., KEMP, A. S. W., EVANS, L. J., BALDWIN, J. R. and HEPWORTH, B. C. 1984. Early Palaeozoic metamorphic history of the Midland Valley, Southern Uplands-Longford-Down massif and the Lake District, British Isles. Trans. R. Soc. Edinburgh, Vol. 75, 245–258.

PATERSON, I. B., BROWNE, M. A. E. and ARMSTRONG, M. 1976. Upper Old Red Sandstone palaeogeography. Scott. J. Geol., Vol. 12, 89–91.

PEACH, B. N. and HORNE, J. 1899. The Silurian rocks of Britain, Vol. 1, Scotland. Mem. Geol. Sure. G.B.

PLAYFAIR, J. 1805. Biographical account of the late Dr James Hutton. Trans. R. Soc. Edinburgh, Vol. 5, 39–99.

POLLAND, J. E. 1985. Coprolites and ostracods from the Dinantian of Foulden, Berwickshire , Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 49–51.

RAGG, J. M. and FUTTY, D. W. 1967. The soils of the country around Haddington and Eyemouth. Mem. Soil. Sure. G.B., Scotland, sheets 33 and 34.

RAMSAY, J. G. 1961. The effects of folding upon the orientation of sedimentation structures. J. Geol., Vol. 69, 84–100.

ROCK, N. M. S. and RUNDLE, C. C. 1986. Lower Devonian age for the Great Basal Conglomerate, Scottish Borders. Scott. J. Geol., Vol. 22, 285–288.

RUST, B. R. 1965. The sedimentology and diagensis of Silurian turbidites in south-east Wigtownshire, Scotland. Scott. J. Geol., Vol. 1, 231–246.

SCHENK, P. E. 1970. Regional variation of the flysch-like Meguma Group (Lower Palaeozoic) of Nova Scotia, compared to recent sedimentation off the Scotian Shelf. Spec. Pap. Geol. Assoc. Canada, No. 7, 127–153.

SCOTT, A. C., GALTIER, J. and CLAYTON, G. 1984. Distribution of anatomically preserved floras in the Lower Carboniferous in western Europe. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 75, 311 -340.

SCOTT, A. C. and MEYER-BERTHAUD, B. 1985. Plants from the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 13–20.

SCOTT, W. B. 1971. The sedimentology of the Cementstone Group in the Tweed Basin: Burnmouth and the Merse of Berwickshire. Unpublished PhD thesis, Sunderland Polytechnic.

SHIELLS, K. A. G. 1964. The geological structure of north-east Northumberland. Trans. R. Soc. Edinburgh, Vol. 65, 447–481.

SHIELLS, K. A. G. and DEARMAN, W. R. 1963. Tectonics of the Coldingham Bay area of Berwickshire, in the Southern Uplands of Scotland. Proc. Yorkshire Geol. Soc., Vol. 34, 209–234.

SHIELLS, K. A. G. 1966. On the possible occurrence of Dalradian rocks in the Southern Uplands of Scotland. Scott. J. Geol., Vol. 2, 231–242.

SHINN, E. A., GINSBURG, R. N. and LLOYD, R. M. 1965. Recent supratidal dolomite from Andros island, Bahamas. Spec. Pub. Soc. Econ. Palaeont. Miner., No. 13, 112–123.

SISSONS, J. B. 1983. Quaternary. Pp.399–424 in Geology of Scotland. CRAIG, G. Y. (Editor). (Edinburgh: Scottish Academic Press.)

SMITH, T. E. 1967. A preliminary study of sandstone sedimentation in the Lower Carboniferous of the Tweed Basin. Scott. J. Geol., Vol. 3, 282–305.

SMITH, T. E. 1968. The Upper Old Red Sandstone-Carboniferous junction at Burnmouth, Berwickshire. Scott. J. Geol., Vol. 4, 349 - 354.

STEVENSON, W. 1849. The geology of Cockburn Law and the adjoining district in Berwickshire. Trans. R. Soc. Edinburgh, Vol. 16, 33–46.

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. J. Geol. Soc. London, Vol. 144, 753–764.

STRACHAN, I. 1982. The distribution of graptolites in south-east Scotland, with notes on the fauna of Old Cambus Quarry. Scott. J. Geol., Vol. 18, 157–165.

THIRLWALL, M. F. 1979. The petrochemistry of the British Old Red Sandstone province. Unpublished PhD. thesis, University of Edinburgh.

THIRLWALL, M. F. 1981. Implications for Caledonian plate tectonic models of chemical data from volcanic rocks of the British Old Red Sandstone. J. Geol. Soc. London, Vol. 138, 123–138.

THIRLWALL, M. F. 1982. Systematic variation in chemistry and Nd-Sr isotopes across a Caledonian talc-alkaline volcanic arc: implications for source materials. Earth Planet. Sci. Lett., Vol. 58, 27–50.

TOMKEIEFF, S. I. 1945. Petrology of the Carboniferous igneous rocks of the Tweed Basin. Trans. Edinburgh Geol. Soc., Vol. 14, 53 75.

TOMKEIEFF, S. I. 1952. The Carboniferous igneous rocks of the Kelso district. Proc. Univ. Durham Philos. Soc., Vol. 11, 95.

WALKER, F. 1925. Four granitic intrusions in south-eastern Scotland. Trans. Edinburgh Geol. Soc., Vol. 11, 357–365.

WALKER, F. 1928. The plutonic intrusions of the Southern Uplands east of the Nith Valley. Geol. Mag., Vol. 65, 15–11.

WALTON, E. K. 1983. Lower Palaeozoic-stratigraphy, and Lower Palaeozoic-structure and palaeogeography. Pp.105- 166 in Geology of Scotland. CRAIG, G. Y. (Editor). (Edinburgh: Scottish Academic Press.)

WATERSTON, C. D. 1962. A new Upper Old Red Sandstone fish locality in Scotland. Nature, London, Vol. 196, 203.

WATERSTON, C. D. 1985. Chelicerata from the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci., Vol. 76, 25–33.

WEIR, J.A. 1974. The sedimentology and diagenesis of the Silurian rocks on the coast west of Gatehouse, Kirkcudbrightshire. Scott. J. Geol., Vol. 10, 165–186.

WHITE, E. I. 1927. The fish-fauna of the Cementstones of Foulden, Berwickshire. Trans. R. Soc. Edinburgh, Vol. 55, 255–286.

WILLIAMS, A., STRACHAN, I., BASSET, D. A., DEAN, W. T., INGHAM, J. K., WRIGHT, A. D., and WHITTINGTON, H. B. 1972. A correlation of Ordovician rocks in the British Isles. Spec. Rep. Geol. Soc. London, No. 3.

WILSON, G. V. 1921. The lead, zinc, copper and nickel ores of Scotland. Spec. Rep. Miner. Resour. G.B., Vol. 17.

WILSON, H.H. 1952. The Cove Marine Bands in East Lothian and their relation to the Ironstone Shale and Limestone of Redesdale, Northumberland. Geol. Mag., Vol. 89, 305–319.

WILSON, R.B. 1974. A Study of the Dinantian marine faunas of south-east Scotland. Bull. Geol. Surv. G.B., No. 46, 35–65.

WOOD, S.P. and ROLFE, W. D. I. 1985. Introduction to the palaeontology of the Dinantian of Foulden, Berwickshire, Scotland. Trans. R. Soc. Edinburgh, Earth Sci, Vol. 76, 1–6.

Appendix 1 List of six-inch maps

The geological six-inch maps covering and wholly or partly in the 1:50 000 Eyemouth 34 Sheet are listed below with the names of the surveyors and the dates of survey.

NT 75 SE Duns Cameron 1962
NT 75 NE Preston* Cameron 1962–3
NT 76 SE Abbey St Bathans Cameron 1962–3
NT 76 NE Ewieside Hill* Davies 1966–8
NT 77 SE Cockburnspath* Davies 1963
NT 85 SW Manderston House Cameron and Greig 1963–4
NT 85 NW Edrom* Cameron 1963
NT 85 SE Allanton McAdam 1968
NT 85 NE Chirnside Greig 1968–70
NT 86 SW Houndwood* Greig 1965–6
NT 86 NW Grantshouse Greig 1963
NT 86 SE Reston* Greig 1964–6
NT 86 NE Lumsdaine* Greig 1962–3
NT 87 SW Siccar Point" Greig 1962
NT 87 SE Fast Castle2 Greig 1962–3
NT 95 SW Paxton McAdam 1968
NT 95 NW Foulden* Greig and Davies 1966–70
NT 95 NE Lamberton* Davies, Greig, Monro and Smith 1963- 70
NT 96 SW Eyemouth* Greig 1964- 70
NT 96 NW St Abbs* Greig 1962–70
NT 96 SE Burnmouth* Greig and Smith 1963–70
1 Printed with 86 NW
2 Printed with 86 NE
* Available for purchase as dye-line copies

Appendix 2 Boreholes

The Geological Survey registered number is given after the name, and can be used to access more information. Except for Marshall Meadows all the boreholes are for water.

Appendix 3 Geological Survey photographs

All are in the D and MNS series. Those with numbers lower then D 1470 are black and white; those with numbers D 1471 and over or MNS numbers are available at a fixed tariff as colour prints and 35 mm slides. Some colour photographs are retakes of earlier black and white photographs. The MNS photographs are all oblique aerial views of the coast.

Views

Silurian

Lower Devonian

Carboniferous

Quaternary

Figures and plates

Figures

(Figure 1) Sketch map of the district

(Figure 2) Generalised coast section from Redheugh to Pettico Wick. This section is 6 km long (across strike)

(Figure 3) Dislocated masses in thin-bedded greywackes

(Figure 4) Current alignments in Lammermuir Silurian

(Figure 5) Anticline at Pikies Cove [NT 8900 6940]

(Figure 6) Folds at Little Pits [NT 8956 6927]

(Figure 7) Fault orientation in Lammermuir Silurian

(Figure 8) Faulted syncline at Hirst Rocks [NT 830 703]

(Figure 9) Folds at John's Roads [NT 9535 6415]

(Figure 10) Folds at Elgy Rocks [NT 9521 6441]

(Figure 11) Folds in Linkim Beds near Callercove Point [NT 9313 6523]

(Figure 12) Folds in Linkim Beds near Linkim Kip [NT 9290 6543]

(Figure 13) St Abb's Head geological map

(Figure 14) Callercove Point map and section [NT 9316 6524]

(Figure 15) Siccar Point: Hutton's unconformity [NT 812 709]

(Figure 16) Structures on shore east of Siccar Point [NT 8175 7075]

(Figure 17) Structures on shore east of Greenheugh Point [NT 8020 7087]

(Figure 18) Carboniferous classification

(Figure 19) Carboniferous sequence at Cockburnspath

(Figure 20) Cementstone Group sections in Whiteadder valley

(Figure 21) Carboniferous sequence at Lamberton

(Figure 22) Cementstone Group sections at Burnmouth [NT 960 610]

(Figure 23) Pre- and post-glacial drainage

(Figure 24) River profiles

(Succession) Geological succession in the Eyemouth district

Plates

(Plate 1) Siccar Point: Upper Old Red Sandstone unconformable on Silurian greywackes (D 1471) [NT 8126 7094]

(Plate 2) Coast west of Pettico Wick: folded Silurian (D 1230) [NT 9085 6915]

(Plate 3) Lowries Knowes: large ripple marks (D 1366) [NT 8530 6956]

(Plate 4) Pikie's Rock: flute and groove casts (D 1374) [NT 8902 6945] A Flute marks B Groove marks

(Plate 5) Brander Heugh shore: groove casts (D 2861) [NT 8732 7029]

(Plate 6) Scout Point: brecciated greywacke in anticline (D 1453) [NT 9550 6379]

(Plate 7) Scout Point: folded greywackes with subsidiary folds in the hinge zone (D 1452) [NT 9547 6374]

(Plate 8) Horsecastle Bay: brecciated lava (D 1408) [NT 9180 6848]

(Plate 9) Devono-Carboniferous and Carboniferous fossils 1 Fouldenia ottadinia X 2, Cementstone Group, Crooked Burn, Foulden. 2 Holoptychius nobilissimus x 1, Devono-Carboniferous, shore between Greenheugh Point and Pease Burn. 3 Pterinopectinella sp. X 2, Cove Lower Marine Band, shore between Reed Point and Eastern Hole, Cove. 4 Modiolus cf. latus x 2, Cementstone Group, Whiteadder Water, Kilnick Plantation, 880 m north of Edrom Mains. 5 Gigantoproductus ? x 1, Cove Upper Marine Band, shore between Reed Point and Eastern Hole, Cove. 6 Bothreolepis hicklingi X 1, Devono-Carboniferous, shore west of Old Cambus Burnmouth.