Geology of Kent and the Boulonnais

The work was originally undertaken in order to assist geological understanding at both ends of the Channel Tunnel.

The work involved not only researching the surface and subsurface geology of Kent, but also that of the Boulonnais. It relied heavily on the experience of both Chris Wood, the main author and Roy Shephard-Thorn.

Rocks of the Pre-Carboniferous basement of the area, to those of Quaternary age are discussed in stratigraphic order.

This paper has been uploaded onto the www to allow anyone interested in the Geology of Kent and the Boulonnais to have free access to it. This page is currently in draft form and is still under construction. In particular a number of figures are in preparation.

Pre-Carboniferous Palaeozoic basement

Late Palaeozoic

Carboniferous

Carboniferous Limestone

Inferior Oolite Group

Great Oolite Group

Kellaways Formation

Oxford Clay Formation

Corallian Group

Kimmeridge Clay Formation

Portland Group

Cretaceous

Purbeck Limestone Group

Wealden Supergroup

Lower Greensand Group

Gault Formation

Chalk Group

Lower Chalk Formation

Middle Chalk Formation

Holywell Chalk Member

New Pit Chalk Member

Upper Chalk Formation

Palaeogene

Palaeogene of NE Kent

Thanet Sand Formation

Upnor Formation

Harwich Formation

London Clay Formation

Palaeogene of the Boulonnais

Palaeogene of the Hampshire-Dieppe Basin

Neogene

Miocene-Pliocene

Inversion of the Weald

The Palaeozoic rocks of the Boulonnais belong to two separate tectonic units, the western extensions of the Dinant Synclinorium to the south and the Namur Synclinorium to the north, respectively. These two synclinoria are separated by a tight anticlinal structure (Condroz Zone), which formerly represented a barrier to northward transgression of the Devonian sea onto the Brabant Massif until mid-Devonian times. In the Dinant Synclinorium, Upper Silurian rocks are overlain conformably by Lower Devonian rocks. In contrast, the middle Devonian rocks of the Namur Synclinorium rest on folded Lower Palaeozoic rocks that were tectonised during the Caledonian Orogeny and overlie the Precambrian core of the Brabant Massif. In the course of the Variscan Orogeny the sediments of the two synclinoria were strongly folded and brought into juxtaposition along a major east-west low-angle thrust inclined to the south (Grande Faille du Midi). To the south of this thrust, the structural style is one of complex folds without any thrusts. Immediately adjacent to (and below) the Grande Faille du Midi, there is a zone of strongly folded and faulted rocks traversed by two major, closely spaced, low-angle thrusts, the Hydrequent Thrust and the Haut Banc Thrust. The thrust slice between the Grand Faille du Midi and the Hydrequent Thrust (Massif d'Hydrequent) comprises an overturned, tightly folded syncline of Devonian and Carboniferous rocks. The basal Hydrequent thrust brings Fammenian rocks into juxtaposition with folded Carboniferous rocks of the underlying thrust slice (Massif de Haut-Banc), which are in turn brought into juxtaposition with Coal Measures of the autochthonous Massif de Ferques by the Haut-Banc Thrust. The autochthonous succession (Middle Devonian to Westphalian Coal Measures) rests unconformably on folded Upper Silurian shales and is truncated to the north by the subvertical Landrethun Fault (Pruvost, 1992b) (**Fig.). The conventional view is that the Dinant Synclinorium has been thrust over the Namur Synclinorium and the underlying Brabant massif along the Condroz line, which is now expressed by the Grande Faille du Midi and its associated thrusts. The Grande Faille du Midi is a major seismic reflector and is commonly regarded as the Variscan (deformation) Front, i.e. the northern limit of transpressive movement of the mobile belt from the south against the foreland (cf. Lefort and Max, 1992, Fig. 2). This conventional interpretation has been challenged by Bouroz (1989), who considered that the fact that the Namur Synclinorium exhibited a greater degree of tectonism than the Dinant Synclinorium could only be interpreted by inferring a southward subduction of the Namur Synclinorium beneath the Dinant Synclinorium.

The extensive drilling programme in recent years has provided a detailed picture of the pre-Permian Palaeozoic floor, revealing an area of relatively simple anticlines and synclines to the north of the zone of low-angle thrusts traversing the Ferques inlier (Wallace, 1983, Fig. 11.3). Detailed palaeogeographical maps and analyses of the sedimentary history of northern France were produced by Colbeaux et al (1977). Figure ** (modified from insets to BGS 1:250 000 series maps) depicts the pre-Permian Palaeozoic floor for the entire area under review.

A recurring theme in this section of the Web Site will be the structural control exercised on the Mesozoic and Palaeogene rocks by the largely concealed Palaeozoic rocks of the London-Brabant Massif. Apart from the inlier of Palaeozoic rocks centred on Ferques in the Boulonnais, our knowledge of these rocks and some of the Mesozoic rocks that overlie them has come from boreholes and shafts sunk in connection with mining for coal and, more recently, from seismic surveys and boreholes in connection with the search for hydrocarbons offshore and the Channel Tunnel itself.

A contour map of depths in kilometres below sea level (Cameron et al, 1992, Fig. 11) to the top of the Lower Palaeozoic in southeast England reveals a small, relatively shallow, NW-SE basinal structure outlining the area occupied by the Carboniferous rocks of the Kent Coalfield. Here the base of the Carboniferous is as much as 2 km below the surface. The basin is bounded to the west and northwest by a topographic 'saddle' between 0.5 and 1 km down. Both these structures are aligned approximately parallel to a southeasterly extension of the conjectural edge of the Midlands microcraton (Chadwick, 1989, Fig. 9). To the west of the saddle, the top of the Lower Palaeozoic presents a gently inclined, irregular plane sloping down to depths of over 3.5 km. In this area, deep boreholes have proved Middle and Upper Devonian marine sediments overlying Lower Palaeozoic rocks. Within and immediately adjacent to the Kent Coalfield Basin, continental Middle Devonian rocks in Old Red Sandstone facies are present beneath the Carboniferous cover. On the crest of the saddle, Devonian rocks are apparently not preserved, and the top of the Palaeozoic floor is formed of Silurian or Ordovician rocks. To the north of the basin, undated Old Red Sandstone facies sediments as well as fossiliferous marine Devonian shales have been proved in deep boreholes. In east Kent, there is no borehole or seismic evidence for the existence of pre-Ordovician rocks. However, shallow marine and basinal Cambrian to Lower Ordovician sediments are more than 300 m thick on the southern flank of the London-Brabant Massif in Belgium (Walter, 1980) and may be similarly thick at the base of the Lower Palaeozoic sequence beneath the Dover Strait (Cameron, 1992). From the limited borehole evidence, it is inferred that east Kent and the Boulonnais occupied an area of shelf to intermediate depth seas situated to the east of a landmass of low relief during the Ordovician and Silurian. During the early Devonian (near the Pridoli-Lochkovian boundary), east Kent is assumed to have lain within a marine area on the basis of fossiliferous limestones, that are possibly of this age, in the Little Missenden borehole. No Lochkovian sediments are known from the Ferques inlier in the Boulonnais, but variegated red and green schists of this age have been proved in deep boreholes adjacent to the coast. Later in the Devonian, the position of east Kent oscillated between the littoral zone and shelf sea, with the North Downs-Thanet area representing the edge of the landmass to the north (Cope et al, 1992).

In the Bobbing borehole [TQ 874 652], Middle Jurassic (Great Oolite) rests directly on marine mudstones with a macrofauna of late Ordovician aspect and assigned a Caradocian age by acritarchs (Aster et al, 1969). Probable Wenlock neritic sediments containing an atrypoid and Eoplectodonta were proved in the old Cliffe borehole [TQ 719 785] (Cocks et al, 1971) and deeper water shales were found to the southeast, in the Chilham and Brabourne boreholes. In the Chilham borehole [TR 088 545], Upper Llandovery (Telychian) shales dipping at 80 degrees and yielding pyritised graptolites indicative of the Monograptus crispus Biozone were proved beneath Lias. In the Brabourne borehole [TR 077 423], red marls of possible earliest Jurassic age overlie an undated 9.7 m conglomerate that is of Triassic Dolomitic Conglomerate aspect, but probably not of that age. This conglomerate rests on steeply dipping shales and mudstones dated by acritarchs and chitinozoa to the late Llandovery or early Wenlock (Molyneux, 1991).

Evidence for the former existence of Upper Silurian sediments in east Kent is provided by a pebble of crinoidal limestone with a shelly fauna of probable Ludlow age in a conglomerate with a red sandstone matrix from a borehole near Chislet; the conglomerate is presumed to be of post-Ludlow age, possibly latest Silurian (Downtonian, i.e. Pridoli) or earliest Devonian (Dittonian, i.e. Lochkovian) in continental Old Red Sandstone facies (Holmes, 1981). In the Boulonnais, folded early Ludlow (Gorstian) graptolitic shales with Monograptus colanus have been proved beneath unconformable Middle Devonian (Givetian) sediments in the Ferques inlier, but at Caffiers these shales are directly overlain by the Gault (Pruvost, 1992b; Pruvost and Pringle, 1924). The Devonian sediments here belong to the Namur Synclinorium. To the south of the Grand Faille du Midi, i.e. on the northern flank of the Dinant Synclinorium, the Wirwignes borehole proved even younger Silurian rocks, of presumed Ludfordian age, with Dayia navicula. Farther south, the Samer Borehole was reported (Wallace, 1983) to prove early Devonian (Gedinnian, i.e. Lochkovian) marine strata overlying Upper Ludlow, but with no mention of intervening beds of Pridoli age.

Red and green mottled sandstones, conglomerates and siltstones of Old Red Sandstone facies were proved beneath Carboniferous Limestone in the Harmansole borehole [TR 142 529], just to the west of the coalfield and also in three boreholes near Chislet at the northern limit of the coalfield (Holmes, 1981). On palynological and plant macrofossil evidence, the rocks at Harmansole are of Givetian to possibly even earliest Frasnian age (Mortimer and Chaloner, 1972); they also yielded estherians and crossopterygian fish scales of either mid-or late Devonian age. An Old Red Sandstone facies conglomerate in one of the Chislet boreholes yielded the pebble of Ludlow limestone noted earlier. To the west of our area, fossiliferous shelf sea limestones with corals were cored in the Esso Tatsfield and Bletchingley boreholes and are actually exposed on the other side of the Channel in the Ferques inlier (Calcaire de Blacourt). Despite the proximity of the two areas, the Devonian succession in the Boulonnais at this time was predominantly marine, rather than continental as seen in the Harmansole borehole. In the late Devonian, with the extension of the sea to the north, east Kent is inferred (Cope ci at., 1992), to have lain partly in the littoral zone and partly in an area of shelf sea which extended eastwards to the Boulonnais, where shelf limestones with coral-brachiopod faunas (Calcaire de Ferques) were deposited. This interpretation is supported by the record of Frasnian marine shales with brachiopods including Cyrtospirifer verneuili from a borehole near Chislet (Holmes, 1981).

Although there is no direct evidence from boreholes, geophysical modelling (Shephard-Thorn et al., 1972; Shephard-Thorn, 1988) suggests that considerable thicknesses of Devonian rocks are probably present beneath the Carboniferous of the Kent Coalfield and its extension beneath the Dover Strait.

The Devonian succession in the Ferques inlier in the Boulonnais (circa 700 m) rests with marked discordance and a basal conglomerate on folded Silurian strata and belongs to the Namur Basin. The sequence here begins in the Middle Devonian (Givetian) with clastics containing plant debris, which are followed by shallow-water platform carbonates; earlier Devonian (Lochkovian) sediments belonging to the Dinant Basin are known from deep boreholes to the south of the Grande Faille du Midi, e.g. at Samer. The Ferques inlier succession is divided into six formations and is of international importance for it's abundant, diverse and well-preserved coral and brachiopod faunas. Brice (1985) gave a useful overview with a comprehensive bibliography and Wallace (1969, 1970) described the sedimentology and palaeoecology.

In earliest Carboniferous times, lagoonal conditions occurred in east Kent. With rising sea level, alternations of shales and thin crinoidal marine limestones were deposited. During the later Tournaisian and Visean, a persistent landmass extending from Wales eastwards into Belgium, the St. George's Land to London-Brabant High, dominated the palaeogeographical and depositional setting for the first time. To the south of this landmass, marginal shallow water platform carbonates with rich coral-brachiopod faunas were developed. For most of this period, the southern limit of the landmass oscillated backwards and forwards across east Kent, but during the late Visean (Brigantian) sea level rise it may have been situated temporarily far to the north in the area of East Anglia (Cope et al, 1992).

There is no evidence for Namurian sediments in east Kent: It is assumed that any marine deposits from the shelf seas fringing the landmass have subsequently been lost by erosion and that the greater part of east Kent formed part of the landmass itself at this time. Namurian rocks are also unlikely to occur beneath the Dover Strait.

In Westphalian times, east Kent and the Boulonnais occupied a narrow east-west paralic basin between the London-Brabant High to the north and the Variscan mountain range to the south. During the earlier part of the Westphalian, coal-swamp conditions prevailed in the area, with periodic short-lived marine incursions being registered by thin marine bands. The sediments deposited are predominantly argillaceous, with subordinate sandstones and numerous coal seams. Following uplift of the London-Brabant High, the former coal-swamp environment became increasingly dominated by the sedimentation of clastics resulting from the increased erosion of the uplands.

Within the Kent Coalfield basin, about 300 m of Tournaisian and Viséan Carboniferous Limestone are probably present beneath Westphalian Coal Measures, Namurian rocks being absent. Offshore seismic sections show that the limestones continue beneath the Dover Strait and southeast of Dover may be as much as 900 m thick (Cameron et al, 1992).

The Carboniferous Limestone of east Kent boreholes can he subdivided into a thin basal 'Lower Limestone Shale' division, comprising dark shales with thin limestones including crinoidal limestones; and an overlying 'Main Limestone', mostly massive crystalline limestones including some oolites. There are no boreholes penetrating the full thickness of the Carboniferous Limestone. The maximum known thickness is 135 m in the Trapham Borehole near the western margin of the coalfield, but it is probable that at least 300 m may be present beneath the central part of the coalfield. In 14 out of 31 boreholes, fossiliferous horizons in the higher part of the Limestone can be assigned on coral-brachiopod faunas (notably the occurrence of Linoprotonia corrugatohemisphaerica) to the Holkerian Stage of the Viséan (Mitchell, 1981), there being no evidence anywhere for the overlying Asbian and Brigantian stages. Rocks belonging to both these stages are assumed to have been deposited, but subsequently removed in the period of uplift and erosion during Namurian times. Tournaisian 'Lower Limestone Shale' marine sediments, rocks with Courceyan brachiopod-bivalve faunas have been proved some distance beneath the Holkerian in the Trapham Borehole (Mitchell, 1981) with no palaeontological evidence here or elsewhere in Kent for the intervening Chadian and Arundian stages, which may be thin or not represented (George et al, 1976). Even earlier Courceyan shales and limestones with bivalves indicative of a lagoonal environment are present in the 8 m of Dinantian 'Lower Limestone Shales' proved beneath Lias in the Harmansole Borehole.

In the Ferques inlier in the Boulonnais, the transition from the Devonian to the Carboniferous occurs in the tectonised Calcschistes de la Vallée Heureuse. The Carboniferous Limestone is represented by up to 200 m of Tournaisian-early Visean dolomites, rich in large crinoids at the base (Formation du Hure), overlain by at least 270 m of Viséan limestones divided into six formations (Colbeaux et al, 1985). Some of these limestones take a high polish and have been exploited for centuries in huge quarries as ornamental stone, incorrectly described as marble (Marbre), of which the most famous is perhaps the 'Marbre Napoleon'.

In the Boulonnais, the equivalent of the Carboniferous Limestone is followed by the Formation des Plaines, a succession comprising sandstones with brachiopods overlain by predominantly clastic sediments with thin coals that range in age (Colbeaux et al, 1985) from early Namurian to early Westphalian (Langsettian).

Following the prediction by Godwin Austen (1856) that coal would be found beneath the Mesozoic rocks of the Weald, Coal Measures rocks were finally proved in 1890 at -340 m OD in a borehole near Shakespeare Cliff (southwest of Dover, on the site of the abortive first Channel Tunnel workings) and at the Dover or Shakespeare Colliery (and the recent Channel Tunnel working site). Later exploration in connection with the development of the Kent Coalfield established that Westphalian Coal Measures non-sequentially overlying Carboniferous Limestone were preserved in an elongated WNW-ESE synclinal basin (Fig.**). This basin is clearly reflected in the Bouguer anomaly trends (Shephard-Thorn et al, 1972, Fig. 1).

The eroded upper surface of the Coal Measures slopes in a southwesterly direction from -244 m OD near Deal to depths greater than -396 m OD at Folkestone (Fig. **). The onshore limits of the basin are ill-defined and are possibly in part fault-controlled, notably on the western margin, where there is a rapid rise in the Carboniferous Limestone surface from about -915 m OD in the Bishopsbourne Borehole within the coalfield to about -320 m OD in the Harmansole Borehole, which is situated just outside the coalfield and where the contact between the Carboniferous Limestone and the Devonian is considered to be faulted. There is some evidence that the northern boundary of the coalfield is also fault-bounded (Shephard-Thorn, 1988).

Offshore, the Coal Measures are believed to extend at depth across the Strait of Dover. Borehole evidence suggests that the Kent Coalfield Basin may comprise two subsidiary troughs separated by an anticlinal structure to the southwest of the Snowdown and Tilmanstone collieries. In the northern trough, the St. Margaret's Bay Borehole penetrated 860 m of Coal Measures (Bisson et al., 1967), terminating an estimated 20 m above their base at about -1160 m OD. An even greater thickness (up to 968 m) of Coal Measures is inferred to be present in the southern trough.

The Coal Measures everywhere rest non-sequentially on eroded Carboniferous Limestone, Namurian rocks being absent. The contours on the upper surface of the Carboniferous Limestone broadly parallel those of one of the main coal seams (the Kent No. 6 seam). The surface of unconformity as seen in cores is sharp and there is little sign of deep weathering of the underlying limestones despite the major hiatus. However, in some cores the top of the limestone is deeply fissured with infills of arenaceous and pyritous Langsettian (Westphalian A) sediments extending several metres below the surface.

The Coal Measures of the Kent Coalfield are divided into two parts (for an outline coverage see Shephard-Thorn, 1988 and references therein). The Lower Westphalian Shale Division (213 m) at the base comprises mudstones with subordinate sandstones and includes 8 main coal seams and at least four marine horizons. It is overlain by the Upper Westphalian Sandstone Division (670 m), which comprises a lower coal-bearing succession with sandstones, thick mudstones and six main coal seams, of which the lowest (Kent No. 6) was formerly of considerable economic importance; and an upper succession with few workable coals in which thick sandstones form over 70 per cent of the lithologies. Faunal and floral evidence shows that the Shale Division belongs predominantly to the Langsettian and Duckmantian stages, although the earliest Langsettian is not represented; at the top there is a thin unit of Bolsovian measures. The Vanderbeckei and Aegiranum marine bands at the base of the Duckmantian and Bolsovian respectively contain some the richest marine faunas to be found at these levels anywhere in Britain. The Sandstone Division is entirely of Westphalian C and D age and there is no evidence for the existence of sediments of Stephanian or younger age.

Although a maximum of less than 900 m of Westphalian sediments has been proved in the Kent Coalfield, the geochemical maturity of the coals indicates that a considerable thickness of younger strata, probably of Carboniferous age, was deposited but removed prior to the Jurassic transgression (Hamblin et al., 1992). Seismic profiles suggest the existence of up to 1600 m of Westphalian sediments in the offshore extension of the coalfield beneath the Dover Strait, implying the possibility of greater thicknesses of coal-bearing strata than in the onshore part (Cameron et al., 1992).

In the Boulonnais, the coal basin is fragmented into several components (Becq-Giraudon et al., 1981) by normal faults trending N30 degrees and, in marked contrast to the Kent Coalfield, it is affected by thrusting associated with the Variscan deformation front. To the west, a small coalfield is preserved overlying Carboniferous Limestone in a shallow syncline at Strouanne on the coast northeast of Cap Gris-Nez (Wallace 1983, Fig. 11.3) and the presence of Coal Measures in a syncline beneath Calais was suggested by Godwin Austen (1856). The juxtaposition of Carboniferous Limestone on Carboniferous Limestone beneath the Haut-Banc Thrust locally preserves thin wedges of Coal Measures in the western part of the Ferques inlier and, farther to the east, thin coals of early Westphalian age were mined until 1936 in the Hardinghen Coalfield (Colbeaux et al., 1985; Leplat and Colbeaux, 1985).

At the close of the Carboniferous, the Variscan orogenic phase of folding, faulting and uplift resulted in the temporary retreat of the sea from the entire area of Britain, much of which became a desert landscape. Throughout the Permian, southern Britain remained land under continuing desert conditions and was unaffected by the later Permian (Zechstein) marine incursions to the north. No Permian or Triassic deposits are known from east Kent or the Boulonnais.

The Mesozoic sediments of east Kent and their continuation eastwards to the Boulonnais were deposited at the eastern end of an elongate, structurally controlled east-west Weald basin within the larger Wessex-Channel Basin. The Weald Basin was delimited from the remainder of the Wessex-Channel Basin by a WNW-ESE positive area, the Hampshire-Dieppe High (Hamblin et al., 1992, Fig. l4a), and was margined to the north by an intermittently emergent positive area of Palaeozoic rocks called the London Platform, which separated the Weald Basin from the North Sea Basin. The edge of the platform is marked by a zone of closely spaced normal faults of east-west (Variscan) trend throwing down progressively to the south (Chadwick, 1985, 1986). Mesozoic depositional history as known from the Kent Coalfield borings and the outcrops in the Boulonnais records repeated transgressions onto and regressions from this ancient massif (Lamplugh et al., 1923) as illustrated by Figs ***. These phases of coastal onlap and offlap were determined partly by tectonic uplift and subsidence, resulting from the periodic reactivation of deep-seated Variscan and preVariscan basement structures and partly by episodes of eustatic sea level change. The relationship in the Weald between faulting and folding in the Mesozoic cover and faults in the basement was discussed by Shephard-Thorn et al., (1972), Lake (1975) and Mortimore and Pomerol (1991), all of whom drew particular attention to the fact that the main structural lineations represent continuations of the main shear zones in northern France.

Within the depocentre of the subsiding Weald basin, situated well to the west of the area under review (for isopachytes of individual rock units see Sellwood et al, 1986), up to 3000 m of Jurassic and early Cretaceous sediments accumulated. The Wealden succeeds the Portland-Purbeck succession conformably in the Weald Basin but, towards the basin margin, it progressively onlaps the upturned underlying Jurassic strata to rest on the Palaeozoic floor in east Kent. The Jurassic-Lower Cretaceous basin fill was eroded and truncated in an intra-Aptian uplift phase (for cross-sections see Ruffell, 1992, Fig. 5), following which renewed transgression resulted in the deposition of the higher part of the Lower Greensand, followed by the Gault and Chalk. Minor basin inversion phases in the late Cretaceous, notably in Coniacian to early Campanian times (the so-called sub-Hercynian tectonic phases), controlled Chalk deposition to a not insignificant extent (Mortimore and Pomerol, 1991). A more important phase of inversion is seen in the folding, uplift and erosion that followed the end-Cretaceous period of regression and eustatic sea level fall. The highest beds of the Chalk were stripped off and the Chalk is everywhere overlain with considerable unconformity by early Tertiary sediments. However, the major inversion of the Weald Basin took place later in the Tertiary during the main (Miocene) Alpine orogenic phase.

Map showing the northern limits of sedimentation during the Jurassic and early Cretaceous

Mesozoic depositional history in the Boulonnais largely mirrors that in east Kent, but it was also strongly controlled by a mosaic of tectonic blocks delimited by N110 and N30-50 trending faults initiated at the close of or after the Variscan Orogeny. The longitudinal faults (N110) are dextral tear faults; movement of these faults in the Cretaceous resulted in anticlinal and synclinal folding with the same axial trend (Robaszynski and Amédro, 1986). Upper Cretaceous strata both onshore and offshore rest with slight angular discordance on the underlying strata in consequence of this folding phase. Some of these faults remained intermittently active until Quaternary times; differential synsedimentary movement of tectonic blocks had a profound effect on the sedimentation of the Chalk and correlative deposits (Mortimore and Pomerol, 1987). The most important of these tear faults is the arcuate Zone de Cisaillement (shear zone) Nord-Artois, which is situated approximately parallel to, and a little to the north of, the Grande Faille du Midi (Colbeaux et al, 1977, Fig. 1). This major shear zone forms a suture between the Ardennes and Brabant blocks and was probably initiated as early as the Cambrian over an inhomogeneity in the crust. It is marked by significant seismic activity and is expressed on both aeromagnetic and Bouguer anomaly maps. French geologists consider that this shear zone extends westwards beneath the Channel to Dungeness. Bergerat and Vandycke (1994) have documented palaeostress fields in the Cretaceous of east Kent and the Boulonnais based on an analysis of fracture patterns and fault-slip data. They have identified six successive tectonic events, ranging in age from Cretaceous to Recent. Strike-slip movement of early Cenomanian age may possibly be related to dextral movement of the Nord-Artois Shear Zone induced by an early sub-Hercynian tectonic phase.

During the Triassic, continued uplift of the Variscan mountain chains led to the re-establishment of the London-Brabant high, with its southern limit now permanently constrained by the final Variscan deformation front, which stretched from Belgium, via the Boulonnais and east Kent to the northern part of the Weald. Crustal extension in latest Triassic times followed by thermal relaxation led to the initiation of the Wessex-Channel structural/depositional basin and the beginning of a marine transgression from the west (Sellwood et al., 1986, Fig. 4). In the early Jurassic, the advancing Lias sea progressively inundated the western part of the Wessex-Channel Basin, but fully marine conditions were not established in the eastern part of the basin until later.

In the Brabourne Borehole [TR 077 423], (Pliensbachian) Lower Lias overlies undated red, purple and green marls above a 9.7 m thick conglomerate comprising pebbles of limestone, quartzite, calcareous sandstone and chert in a marl matrix. Although these much-discussed red-beds are of Triassic aspect (Lamplugh and Kitchin, 1911), the most recent view is that they are likely to be of earliest Jurassic age and comparable with the sub-Lias red-beds in the Ashdown No. 2 Borehole (Bristow and Bailey, 1972) to the west (Dr I. F. Penn: pers. comm., 1994). Similar undated red mudstones resting on Carboniferous Limestone were found beneath a conglomeratic calcareous sandstone of Rhaeto-Liassic age in the Warlingham Borehole (Worssam and Ivimey-Cook, 1971).

In the Boulonnais, relicts of Rhaeto-Lias continental clays and sands with plants are preserved in karst pockets in the Carboniferous limestones in the Ferques inlier (Corsin, 1951). Red marls, sandstones and conglomerates below Lias in the Framzelle Borehole (Pruvost, 1922a) near Cap Gris-Nez have been compared with the sub-Lias strata at Brabourne and are variously cited as Devonian, Triassic or Rhaeto-Lias. In the Boulogne Borehole, thin Pliensbachian marine sandstones rested discordantly on continental sediments, including sandstones and clays with lignite and plant debris, of inferred terminal Triassic or 'Infraliassic' age (Bonte, 1974).

In the depocentre of the Wealden Basin, the Ashdown No. 2 Borehole proved 380 m of Lias, with relatively nearshore facies at the base, indicating that open-shelf conditions had not yet been fully established (Bristow and Bazley, 1972; Hamblin et al, 1992). At Brabourne, the sandy base of the Lias is of Pliensbachian age. Farther to the east, the Lower Lias rests on Carboniferous Limestone at Harmansole, on Coal Measures within the area of the Kent Coalfield and, outside the coalfield, on the Lower Palaeozoic floor (Silurian) in the Chilham Borehole. The Lower Lias thins rapidly northeastwards onto the London Platform, and is overlapped by the Middle and Upper Lias, which themselves thin in the same direction: there is no evidence for overstep of Middle or even Upper Lias on to the Palaeozoic. In the northeast of the area, the Lias is absent due to erosion and is overstepped by younger Jurassic strata (Fig. **). The thickest development in the area under review is the 42.6 m recorded in the Brabourne Borehole, subdivided by Lamplugh and Kitchin (1911) into Lower Lias (24 m), Middle Lias (13.7 m) and Upper Lias (4.6 m). The Dover Colliery shaft proved 11.57 m of Lias, comprising Lower Lias (4.87 m). Middle Lias (4.87 m) and Upper Lias (1.83 m).

The Lower Lias is typically developed as grey and locally black fossiliferous clays with ammonites indicating the Prodactylioceras davoei Zone; lower beds at Brabourne have been tentatively attributed to the basal Pliensbachian Uptonia jamesoni Zone. The 1.2 m of Lower Lias in the Chilton Borehole are predominantly limestones with a basal bed containing compound corals (Montlivaltia). In several boreholes, e.g. those at Folkestone, Elham and Adisham, the base of the Lias contains small, well-rounded pebbles of Coal Measures sandstones and appears to be infilling minor irregularities in the underlying surface. In all the boreholes to the east of Brabourne, the Hettangian and Sinemurian stages of the Lower Lias are absent and the base of the Lias records the major early Pliensbachian transgression.

The Middle Lias (Upper Pliensbachian) in the east Kent boreholes can be broadly divided into a thin, relatively arenaceous or shaly unit overlain by a much thicker unit with persistent limestones. As the succession thins eastwards, limestones become predominant. In the Chilton shaft, the base of the Middle Lias is marked by a bed of pebbles of quartz grit, perhaps indicating short-lived uplift and erosion of the Palaeozoic landmass. Although Lamplugh et al (1923) inferred (in the absence of ammonites) that the two units corresponded to the Amaltheus margaritatus and Pleuroceras spinatum zones of the Upper Pliensbachian respectively, rhynchonellid brachiopod evidence (Ager, 1954) indicates that only the margaritatus Zone is present and that the top of the Middle Lias is missing due to erosion.

The Upper Lias (Toarcian) is represented in the east Kent boreholes by a succession comprising sandy limestones of the Dactylioceras tenuicostatum Zone overlain by clays and shales of the Harpoceras falciferum and Hildoceras bifrons zones. However, in the Brabourne Borehole, blue shales with ammonites indicative of the bifrons and Grammoceras thouarsense zones rest non-sequentially on green sandy Middle Lias limestones. At Dover Colliery, there is ammonite evidence for the tenuicostatum Zone and the lower part of the falciferum Zone; the overlying 9 m were not cored and cannot consequently be dated.

In the Boulonnais, the fully cored Boulogne Borehole proved, beneath the basal Bajocian conglomerate, a Lias succession comprising 13.10 m of Aalenian and Toarcian clays, marls and many limestones with Dactylioceras tenuicostatum at the base, resting on sandstones (0.8 m) with probable Pliensbachian bivalves (Bonte, 1974; Bonte et al, 1985). In other boreholes near the coast, the basal part of the Lias includes beds of ferruginous ooliths: in the Framzelle Borehole (Pruvost, 1922a), the lower of two such beds included rolled crinoids and rhynchonellid brachiopods. These horizons have been correlated with the Lotharingian (Upper Sinemurian) transgression in the Ardennes. Thin relicts of Lias with phosphates and conglomerates are also preserved in karst pockets in the surface of Carboniferous limestones in the Ferques inlier.

The top of the Upper Lias is missing in the east Kent boreholes, as are the Lower and Middle subdivisions of the Inferior Oolite. This is perhaps the result of the mid-Bajocian regression and erosive phase. The succeeding major late Bajocian transgression is recorded in the Dover Colliery shaft by 8 m of shelly sands with a basal conglomerate of small phosphatic nodules and pebbles of vein-quartz, which rest on eroded Upper Lias. No ammonites were found, but a diverse bivalve fauna including Trigonia suggests that the beds belong to the Upper Inferior Oolite, equating with the Stenoceras garantiana Zone, Upper Trigonia Grit and possibly extending up into the topmost Bajocian Parkinsonia parkinsoni Zone (Cope et al, 1980). Oolitic limestones resting on Upper Lias in the Brabourne Borehole have also been tentatively assigned to the Inferior Oolite on lithological grounds, but without any palaeontological support. To the northeast of Dover, the Inferior Oolite is overstepped by the Great Oolite.

Following the procedure adopted by the BGS for the English Channel (Hamblin et al, 1992), the Great Oolite Group in the east Kent boreholes is here broadly subdivided into the Fullers Earth, Great Oolite, Forest Marble and Cornbrash formations. However, the Kent succession has considerable palaeontological and lithological similarity to that developed to the west of Oxford, notably in the occurrence of rocks that may equate with the sandy facies of the Chipping Norton Formation, the Hampen Marly Formation and the White Limestone Formation (Lamplugh et al, 1923; Arkell, 1933; Cope et al, 1980).

The Great Oolite in our area exhibits what Arkell (1933) called 'the most important transgression of Jurassic times against the southern border of the London landmass.' The group overlaps the sandy Upper Inferior Oolite seen at Dover and (questionably) at Brabourne, overstepping progressively northeastwards on to Upper Lias, Middle Lias, Coal Measures and, finally, on to the pre-Carboniferous Palaeozoic floor (Ordovician) at Bobbing (Fig. **). There is also overlap of lower by higher beds of the Great Oolite Group. In some of the colliery shafts, the base was marked by a pebble bed composed of rocks eroded from the underlying strata.

In the Brabourne Borehole, the (inferred) Inferior Oolite is overlain by calcareous shales, with intercalated thin brecciated limestones towards the base. The only fossils were poorly preserved bivalves of no biostratigraphical value. These beds have been tentatively assigned broadly to the Fullers Earth (Lamplugh and Kitchin, 1911; Arkell, 1933). However, to the north and east of Brabourne in the direction of the shoreline, the base of the Great Oolite Group is marked by sands and sandy limestones (e.g. Dover Colliery shaft), or muddy limestones with thin beds of sandy limestone: these arenaceous beds were correlated by Arkell (1933) with the sandy facies of the Chipping Norton Limestone (Formation) of the Oxfordshire succession, i.e. (Cope et al, 1980) the basal Bathonian Zigzagiceras zigzag Zone.

The Great Oolite Formation comprises massive bedded, generally fine-grained, cream or grey oolitic limestones with sporadic intercalations of clay or marl. In the Dover Colliery shaft, 10.97 m were proved overlying 10.05 m of the basal sands and sandy limestones.

At Tilmanstone, shelly clays with abundant Praeexogyra hebridica beneath the main limestones were correlated by Arkell (1933) with the Hampen Marly Beds (Formation) of Oxfordshire.

The greatest thickness of the Forest Marble Formation in the area is the 5.48 m of bluish grey and pale green, poorly fossiliferous calcareous claystones with lignite overlain by yellowish oolitic limestone proved in the Dover Colliery shaft. In the thinner successions to the northeast, e.g. Tilmanstone (2 m), the Forest Marble is represented by greenish-black shelly and lignitiferous clays beneath limestones. Arkell (1933) suggested that, since there was no palaeontological evidence to show that the upper part of the Forest Marble was present in the Kent boreholes, there must be a significant non-sequence between the Forest Marble and the overlying Cornbrash. However, as the faunas at these levels are largely facies controlled, recognition of such a non-sequence on this basis should not be taken as conclusive (Dr. B. M. Cox: pers. comm.).

In the Boulonnais, Rhaetian continental sediments preserved in karst pockets in the Palaeozoic floor in the Ferques inlier are overlain locally by yellowish argillaceous sands, the Sables d'Hydrequent. These sands have been attributed by French authors to the Bajocian, which might suggest a correlation with the (Upper) Bajocian shelly grits in the Dover Colliery shaft. However, as noted by Ager and Wallace (1966), there is no palaeontological evidence for this interpretation. The only fossils known from this unit are poorly preserved moulds of bivalves near the top, which (by inference from Pringle, in Pruvost and Pringle, 1924) permit correlation with the (Lower Bathonian) sandy facies near the base of the Great Oolite Group.

These sands are overlain by marly limestones and marls, the Marnes d'Hydrequent, which contain oyster lumachelles composed of Praeexogyra hebridica and a diverse fauna including echinoids (Acrosalenia, Clypeus, Nucleolites) and brachiopods including Epithyris oxonica and Kallirhynchia concinna. These marls compare with the oyster-rich shelly clays at Tilmanstone. The marls pass up into limestones, the Calcaires de Rinxent, which equate with the lower part of the Great Oolite Limestone Formation. Any one of these three units may overstep the underlying Mesozoic strata to rest locally directly on the Palaeozoic floor. The Calcaires de Rinxent comprise finely bioclastic and slightly sandy limestones, passing up into cross-bedded and oolitic bioclastic limestones. The unit contains numerous hardgrounds and terminates in a hardground. The rich fauna is dominated by rhynchonellids (notably Kallirhynchia concinna), terebratulids and echinoids, with solitary and colonial corals occurring locally at the top. A pteridophyte flora (Lomatopteris, Otozamites and Pterophyllum) has also been documented. The record of Procerites (Gracilisphinctes) cf. laeviplex and the nautiloid Procymatoceras subcontractum places these limestones unequivocally in the Lower Bathonian (Bonte et al, 1985).

The overlying Oolithe de Marquise consist of bioclastic and oolitic limestones, commonly cross-bedded and terminating in a hardground. They have yielded no cephalopods, but the occurrence of common Burmirhynchia hopkinsi in the central portion indicates the Middle Bathonian and permits correlation with the White Limestone Formation in the Bath district.

The Forest Marble Formation is represented in the Boulonnais by the Marnes des Calhaudes (2.5 m), which comprise finely bioclastic limestones, bluish-grey to black marls and argillaceous limestones. The fauna is characterised by the rhynchonellids Burmirhynchia elegantula and Kallirhynchia yaxleyensis, associated with echinoids such as Acrosalenia lamarcki.

The Abbotsbury Cornbrash Formation in the east Kent boreholes comprises a variable impure limestone with a rich brachiopod, bivalve and echinoid fauna. The thickest and biostratigraphically most useful section is provided by the Tilmanstone Colliery Shaft in which the Cornbrash is a little over 7 m thick. Brachiopod records from Tilmanstone can be used to infer that both the Lower Cornbrash (uppermost Bathonian, Clydoniceras discus Zone) and the Upper Cornbrash (lowermost Callovian, Macrocephalites herveyi Zone) are represented (Callomon, 1955), although the admixture of species suggests possible reworking (Arkell, 1933). Macrocephalites cf. herveyi was recorded from the uppermost 2 in of the Cornbrash and some distance above Microthyridina lagenalis, the index of the youngest Cornbrash brachiopod Zone (Callomon, 1955). Macrocephalites was also recorded from the top of the Cornbrash in the Bobbing Borehole (Lamplugh et al, 1923) and from an unspecified level in the Cornbrash of the Guilford Colliery Shaft (Callomon, 1955). In the Dover Colliery shaft, 3.96 m of 'bluish grey streaky sandy clay, masses of indurated calcareous rock with soft brown nodules, and markings like borings' were tentatively assigned to the Cornbrash by Lamplugh and Kitchin (1911). The list of bivalves from these beds is more suggestive of Lower than Upper Cornbrash (Dr. B. M. Cox: pers. comm.), but a record by Callomon (1955) of a Macrocephalites from Dover in the Sedgwick Museum suggests that both subdivisions are likely to be present.

In the Boulonnais, the Cornbrash is represented by 1 m of irregularly bedded oolitic ragstone (Calcaires des Pichottes): both the uppermost Bathonian discus Zone and basal Callovian herveyi Zone have been recognised by the characteristic ammonites and brachiopods.

The Cornbrash in the east Kent boreholes is overlain by up to 2 m of clay with a distinctive bivalve fauna including Meleagrinella braamburiensis. This argillaceous unit has been traditionally referred to as Kellaways Clay but, following Page (1989), it should strictly be called the Cayton Clay (Dr B. M. Cox, pers. comm.).

The main mass of the Kellaways Formation comprises the Kellaways Sand Member, the Kellaways Rock of previous literature on the east Kent boreholes. Isopachytes for the formation (Sellwood et al, 1986, Fig. 11) appear to reflect the underlying Kent Coalfield basin. Cope et al. (1980) noted that this unit 'is very variable in thickness in Kent, ranging from c.6.5 m at Brabourne and Harmansole to c.13.2 m at Fredville. In addition, there is a marked lithological variation, the impure manly sandstones with ferruginous beds of the southern and eastern area (i.e. Dover) being replaced northwards by ferruginous marlstones which locally have the composition of very glauconitic 'millet-seed' ironstones.' The iron is mainly in the form of coffee-coloured polished limonite ooliths. The ironstones of the Kellaways Formation resemble those of the Corallian Group but the iron content is lower and there is a greater admixture of impurities.

The base of the Kellaways Sand is marked by the entry of abundant oysters such as Gryphaea (Bilobissa) dilobotes; these occur throughout the unit and become progressively larger upwards. Ammonites are very rare. Arkell (1933) noted that the record of a possible Proplanulites from the basal bed at Oxney (Lamplugh et al., 1923) suggested the existence of the Proplanulites koenigi Subzone [now Zone]. The only well authenticated record is a Kosmoceras medea from near the top at Guilford (Callomon, 1955), indicative of the K. medea Subzone of the Kosmoceras jason Zone; the record of the zonal index of the underlying Sigaloceras calloviense Zone from Tilmanstone (Lamplugh et al, 1923) was attributed questionably by Callomon also to K. medea. The Kellaways Formation thus extends biostratigraphically higher in Kent than in other areas and correlates with the lower part of the Lower Oxford Clay [Peterborough Member of Cox et al, 1992] of the Midlands (cf. Cope et al, 1980, Fig. 8).

In the Boulonnais, the Kellaways Formation is represented only by a thin unit (up to 5 m, but typically much less) of ferruginous oolitic marls, the Marnes ferrugineuses de Belle, with a rich bivalve and ammonite fauna including Sigaloceras calloviense.

This formation has been proved in several of the Kent boreholes. The total thickness varies from ca. 53 m at Brabourne to ca. 32 m at Dover. Lamplugh et al. (1923) subdivided the succession into three broad units. At the base, the 'Ornatus Beds' comprise brown and greenish brown marly clays containing bivalves and species of Kosmoceras. These beds are broadly equivalent to the Lower Oxford Clay or Peterborough Member of the Midlands. The zonal index of the Kosmoceras jason Zone has been recorded from the base of the Oxford Clay at Guildford and Tilmanstone (Callomon, 1955). Records of Erymnoceras sp. from Chilham and Fredville indicate the overlying Erymnoceras coronatum Zone. Lamplugh et al's record (1923) of Kosmoceras duncani suggests that at least part of the Peltoceras athleta Zone might also be represented.

The succeeding 'Renggeri Beds' of Lamplugh et al. (1923) are smooth bluish-grey clays [equivalent to the Stewartby and basal Weymouth Member of Cox et al, 1992] with rich faunas of Quenstedtoceras, including Q. lamberti and oppeliids including Creniceras and Hecticoceras. Cope et al. (1980) suggested that these beds might belong in part to the athleta Zone and the basal Oxfordian Q. mariae Zone, as well as the intervening Q. lamberti Zone. The recorded ammonites are most typical of the Q. lamberti Zone and the overlying Cardioceras scarburgense Subzone of the mariae Zone [B. M. Cox: pers. comm.]. The younger praecordatum Subzone of the latter zone is represented by Lamplugh et al 's 'Mariae Beds', smooth pale bluish-grey clays [Weymouth Member of Cox et al (1992)] with Quenstedtoceras mariae and perisphinctid ammonites. Elsewhere in southern England, the top of the Oxford Clay lies within the cordatum Zone. This is probably also the case in east Kent, although there is no recorded evidence of the latter zone within the 'Mariae Beds'.

In the Boulonnais, the Oxford Clay is represented by the Argiles de Montaubert (10-25 m), with rare ammonites and common serpulids; and the Argiles du Coquillot with basal argillaceous limestones (2 m) containing Quenstedtoceras lamberti (the equivalent of the Lamberti Limestone of British successions) overlain by up to 19 m of clays with small early Oxfordian ammonites in pyritic preservation. The former extensive exposures of these units in brickpits are no longer available.

Lamplugh et al (1923) recognised three subdivisions in east Kent, namely Lower Corallian, Corallian Limestones and Upper Corallian. The Lower and Upper divisions are lithologically similar, comprising clays and marls, marlstones and rubbly oolites, with minor beds of impure limestone. The lower part of the Lower Corallian includes up to 8 m of clays with limonitic oolitic grains and a concentration of crinoids ossicles (Millericrinus) near the top. The middle division comprises pale muddy coral-bearing limestones overlain by a development of Coral Rag comparable with that seen in the Corallian Group at outcrop. At Brabourne, corals occurred throughout the whole thickness of the limestones (40.84 m), this being the most extensive development of coral rag in England (Arkell, 1933). Near Dover, the admixture of polished limonitic grains in the sediment near the top of the Upper Corallian is sufficiently concentrated to form a significant thickness of 'millet seed' iron ore. The maximum thicknesses were found at Dover Colliery (4.8 m) and Farthingloe Borehole (4.5 m); while at Abbot's Cliff borehole, two beds of ore about 1 m and 2.4 m thick respectively were proved. The ore contains about 33 per cent iron and 15 per cent silica, with the reserves being estimated at 100 million tons (Lamplugh et al, 1923). This high Cr development of iron ore is approximately correlative with the Westbury Ironstone of Wiltshire.

A generalised succession for the Corallian Group of the area was given by Cope et al (1980), although not all the beds indicated in the lower part of the succession can be recognised at Dover. The ironshot marls yield abundant cardioceratids, probably largely indicative of the cordatum Zone. Above these, a thin unit (up to 4 m) of highly fossiliferous grey mudstones with black-coated fossils has been recognised in many of the boreholes. It contains a rich ammonite fauna of 'Aspidoceras', Peltoceras and perisphinctids, as well as Cardioceras and a diverse fauna of bivalves and brachiopods. The recorded ammonites are indicative of the C. cordatum Zone and/or the next youngest C. densiplicatum Zone; Cope et al (1980) assigned it to the latter. No ammonites are known from the succeeding 35 m of limestones, but up to 7 m of black clays with Amoeboceras overlie them in a few boreholes (e.g. Folkestone) and may represent the basal Ampthill Clay of eastern England. The succeeding sandy limestone, present in all the boreholes, has been equated with the Clavellata Beds of Dorset, but there is no palaeontological evidence to support this interpretation. This limestone is overlain by tough, silty clays without ammonites, presumed to correlate with the Sandsfoot Clay. The succeeding ironstones, discussed above, yield Ringsteadia, and are followed by pisolitic marls and clays, which approximately correspond to the Ringstead Clay and the complex condensed succession that includes the Ringstead Coral Bed.

From a maximum thickness in east Kent of about 104 m at Brabourne, the Corallian thins to about 88 in at Dover, with further reductions being recorded in the boreholes to the northeast. At Brabourne and Dover, the full succession is preserved beneath the Kimmeridge Clay, whereas in the localities to the northeast, pre-Cretaceous erosion has removed the Upper Corallian and higher part of the Corallian limestones.

In the Boulonnais, the greater part of the Corallian Group is represented by three thick units of clays, terminating upwards in coralliferous limestones, the lower two and the upper unit being interpreted (Bonte et al., 1985) as regressive and transgressive sequences respectively. The lowest sequence comprises the Argiles de Coquilot, the Marnes et Calcaires a Millericrinus and the Calcaires d'Houllefort, the latter having yielded a rich ammonite fauna. The middle sequence commences with the poorly fossiliferous Argiles de Selles, without ammonites, and ends in the Calcaires du Mont des Boucards, which have likewise yielded many ammonites. The top sequence comprises pyritic and sideritic black clays (Argiles du Mont des Boucards), rich in the oyster Deltoideum delta, with poorly bedded, lenticular coralliferous limestones (Calcaires de Brucquedal) near and at the top of the succession. These beds are followed by calcite-cemented, dark brown and red sandstones (Gres de Brunembert) overlain by a complex succession comprising oolitic limestones and marls (Oolithe d'Hesdin l'Abbe) surmounted by alternations of thin, bored limestones and thin marls (Caillasses d'Hesdigneul), which pass laterally into glauconitic sandstones (Grès de Questrecques). Bonte et al (1985) provided a graphic section of all these beds, which today are poorly exposed.

Despite their relative proximity, it is surprising that it is not possible to present detailed correlations of the successions in the east Kent boreholes with those of the Boulonnais. Without a modern reassessment of the fossils collected from Kent, it would be unsafe to attempt to update the Corallian part of the table given by Pruvost and Pringle (1924). The beds with Millericrinus are clearly equivalent on both sides of the Channel, as are the terminal pisolitic marls of Kent and the Oolithe d'Hesdin l'Abbe. Arkell (1935-1948; 1956) reviewed the ammonite evidence and concluded that, of the three coral limestones that might correlate with the Coral Rag of Kent, the Calcaires d'Houllefort equated with the Elsworth Rock of East Anglia (and were therefore older), while the fauna of the Calcaires du Mont des Boucards suggested a horizon younger than the top of the Clavellata Beds of Dorset, implying that both it and the succeeding Calcaires de Brucquedal represented coral rag developments within the equivalent of the Deltoideum-rich Sandsfoot Clay. There appears to be no evidence for the existence of an equivalent of the Clavellata Beds in the Boulonnais, suggesting that they may be missing at the basal Upper Oxfordian non-sequence. The Upper Corallian iron ore can be inferred to correlate broadly with both the Sandsfoot Grit of Dorset and the Grès de Brunembert in the Boulonnais. On ammonite evidence, the Oxfordian-Kimmeridgian boundary probably falls within the Caillasses d'Hesdigneul.

The CEGB onshore boreholes at Dungeness proved up to 25 m of Kimmeridge Clay comprising dark grey mudstones with thin calcareous beds and bands of phosphatic nodules (Lake and Shephard-Thorn, 1987). Approximately 6 km NE of Dungeness Point, offshore boreholes proved that Kimmeridge Clay formed the eroded core of a faulted anticline (Crosby and Fletcher, 1988) with a WNW-ESE trend.

Lamplugh and Kitchin (1911) and Lamplugh et al (1923) divided the Kimmeridge Clay in the east Kent boreholes into Lower Kimmeridge Clay, characterised by Nanogyra virgula and Upper Kimmeridge Clay, which was without E. virgula, but yielded Modiolus autissiodorensis; the Lower Kimmeridge Clay was more arenaceous and calcareous and of a more shallow water facies than the Upper Kimmeridge Clay (Smart et al., 1966).

The greatest preserved thickness of Kimmeridge Clay in the district was proved at Brabourne; here the succession was 79.85 m, subdivided into 19.5 m of Upper Kimmeridge Clay and 60.35 m of Lower Kimmeridge Clay. At both Brabourne and Ottinge, the Portland Group rests non-sequentially on the Kimmeridge Clay. Ammonite records quoted by Lamplugh et al (1923) suggest that the Upper Kimmeridge Clay may extend as high as the Pectinatites pectinatus Zone or even Pavlovia pallasioides Zone, as in theWarlingham Borehole (Callomon and Cope, 1971). Elsewhere, the Kimmeridge Clay, where present, is overlain by Cretaceous strata, which downcut progressively from west to east (Fig. **). By the Abbot's Cliff Borehole, the thickness of the formation has reduced to 46.63 m, and only 13.41 m of Kimmeridge Clay are present at Dover Colliery shaft, where the succession comprises glauconitic and ironshot impure clays, calcareous beds and limestones. Only the Lower Kimmeridge Clay is represented: the ammonite evidence reported by Lamplugh et al (1923) suggests that the youngest beds, beneath the Cretaceous unconformity, belong to the Aulacostephanus mutabilis Zone.

In the Boulonnais (Wignall, 1991), the argillaceous Kimmeridge Clay Formation of the east Kent boreholes is represented by a more marginal facies of silty mudstones and limestones alternating with thick sandstones and sands; towards the top, the facies becomes even more marginal and includes two horizons of phosphatised pebbles and reworked fossils (P1 and P2, or La Rochette Nodule Beds). The well-preserved trace fossils in the sandstones attest the shallow water environment (Ager and Wallace, 1970), while the argillaceous beds represent transgressive phases, with horizons yielding predominantly subserpenticone ammonites indicating maximum flooding surfaces (Herbin et al, 1995). Full details of the sequence stratigraphy of this succession were given by Roust et al (1993). Furthermore, Herbin et al (1995) have revised the ammonite biostratigraphy, based on accurately horizoned material. The basal beds are not exposed on the coast hut are seen inland: the Oolithe d'Hesdin label at the top of the correlative of the Corallian Group is overlain by some 4.87 m of hard marly limestones (Caillasses d'Hesdigneul) with the Kimmeridgian zonal index Rasenia cymodoce (Bonte et al, 1985).

The magnificent coast sections between Boulogne and Wissant (details in Ager and Wallace, 1970; Wignall, 1990) begin with the Argiles du Moulin Wibert of which the basal metre falls in the mutabilis Zone, while the remainder belongs in the succeeding emulous Zone. Much of the succession includes beds with significant Total Organic Carbon (TOC); the highest TOC values are found in a thick unit (up to 30 m) of laminated pyritic black shales (Argiles de Carillon), which correspond in part to the relatively organic-rich sediments of the Pectinatites elegans Zone in Dorset (Herbin and Geyssant, 1993). A higher organic-rich horizon is found in the Argiles de la Crèche (inferred P. wheatleyensis Zone), which is overlain by two thin limestones enclosed by the phosphatic nodule beds. The interval between these nodule beds may correspond to the highest of the organic-rich beds in the British sequence (huddlestoni-pectinatus zones), for which there is no ammonite evidence.

The equivalent of the higher part of the Kimmeridge Clay Formation is represented by a unit of sandy glauconitic clays alternating with sandy glauconitic limestones, the Argiles de Wimereux. These clays yield ammonites of the pallasioides Zone and terminate in a third horizon of phosphatised pebbles (P3 or Tour de Croi Nodule Bed). The nodule bed contains pebbles of Jurassic rocks (Pruvost and Pringle, 1924) and phosphatic Pectinatites and Pavlovia reworked from the pallasioides Zone, as well as indigenous iridescent ammonites comparable with those from the pallasioides Zone, Hartwell Silt and Swindon Clay of the English Midlands (Townson and Wimbledon, 1979). Although the Tour de Croi nodule bed was previously equated with the upper Lydite Bed of the latter area (i.e. the basal conglomerate of the Portland Group), the new ammonite evidence was interpreted by Towson and Wimbledon to show that the nodule bed represented an intra-Kimmeridgian (pallasioides Zone) event, perhaps partly equivalent to the Lower Lydite Bed. There is no evidence in the Boulonnais for the succeeding topmost Kimmeridgian zones (Pavlovia rotunda and Virgatopavlovia fittoni), which may be here extremely attenuated or absent (Townson and Wimbledon, 1979; Herbin et al, 1995).

About 13 m of Portland Beds were proved in the CEGB boreholes at Dungeness, the succession comprising mudstones and siltstones with cement stones, overlain by bioturbated shelly fine-grained sandstones (Lake and Shephard-Thorn, 1987). Offshore, the Portland Beds flank the anticline NE of Dungeness Point (Crosby and Fletcher, 1988). Evidence for the Portland Group in east Kent comes from the Brabourne, Hothfield and Ottinge Boreholes (Smart et al, 1966), where the successions are much thinner. The Portland Beds in the Brabourne Borehole are 9.44 m thick and comprise, from above, sandy limestones (including a bed of crystalline nodular ferruginous limestone, and a bed of oolitic limestone) with calcareous sandstone at the base; resting on sandy mudstone and bituminous calcareous sandstone with a basal bed of hard bituminous conglomeratic rock resting with a sharp junction on the Kimmeridge Clay. The much thinner succession in the Ottinge Borehole (5.25 m) is predominantly arenaceous and can he inferred to be largely glauconitic from the original description of beds with green grains.

In the absence of ammonites, the Portland Group succession at Brabourne must be interpreted on a lithological basis by means of the ammonite-dated successions in the Henfield and Warlingham boreholes outside the immediate area. In the Henfield Borehole, Glaucolithites occurred in the Portland Group immediately below the junction with the overlying Purbeck Limestone Group. At Warlingham, the ammonite records extend up to and including the kerberus Zone, indicating that the Portland Group there equates not only with the Portland Sand (as at Henfield), but also with the basal part of the Portland Stone of Dorset (Worssam and Ivimey-Cook, 1971). This may also be the situation in the Boulonnais. By analogy with Henfield and Warlingham, it is likely that the Portland Group in east Kent correlates with the Portland Sand and basal part of the Portland Stone of the type area, with a significant non-sequence at the contact with the overlying Purbeck Limestone Group (Lulworth Formation).

In the Boulonnais, the Portlandian (sensu anglico) is 20 m thick (Townson and Wimbledon, 1979). The Tour de Croi nodule bed is overlain by sandy glauconitic clays (Assises de Croi), which are followed by sandy nodular limestones. The higher part of the succession is developed as sands and calcareous sandstones (Gres des Oies), i.e. in Portland Sands facies. Towards the top, an intra-formational conglomerate of pebbles of Palaeozoic rocks, angular fragments of Portland limestones and rolled specimens of Laevitrigonia gibbosa (Poudingue de la Rochette) rests with strongly erosive contact on a bored surface of the underlying beds (Pruvost and Pringle, 1924).

The ammonite biostratigraphy has been reviewed by Townson and Wimbledon (1979), who concluded that the Kimmeridgian-Portlandian boundary probably lay within the lower part of the Assises de Croi. Epivirgatites spp. and Progalbanites albani from the middle Assises de Croi indicate the presence of the basal Portlandian albani Zone and the occurrence of fragmentary Glaucolithites body chambers in the upper Assises de Croi allow the recognition of the G. glaucolithus Zone. Ammonite records including Crendonites gorei suggest that the Galbanites okusensis Zone is likely to be present and there is good evidence, notably common Titanites giganteus, for the existence of the Galbanites kerberus Zone in the Lower and Middle Grès des Oies. No ammonites are known from the Upper Grès des Oies, but it is suggested that these may also belong in the kerberus Zone. The Boulonnais succession is essentially comparable with that found in the Warlingham Borehole and as developed at Swindon and Aylesbury.

In this account, the Cretaceous System is taken to begin with the Purbeck Limestone Group, which is here divided, following BGS practice and Clements (1992) into a lower Lulworth Formation and an upper Durlston Formation, with the base of the latter being drawn at the base of the distinctive Cinder Beds Member. Current research, reviewed by Allen and Wimbledon (1991) and by Feist et al (1995), places the base of the Cretaceous, as defined by the base of the Tethyan Berriasian Stage, at a low level within the Lulworth Formation.

About 55 in of Purbeck Limestone Group strata, including a thin group of evaporites and a possible representative of the Cinder Bed, were proved in the CEGB boreholes at Dungeness (Lake and Shephard-Thorn, 1987) and Purbeck rocks are also present on the flanks of the faulted anticline NE of Dungeness Point (Crosby and Fletcher, 1988). Strata belonging to the Purbeck Limestone Group were recognised by Lamplugh et al (1923) in the boreholes at Hothfield and Brabourne, where the succession was 20.72 m thick in each case. Purbeck strata are not present to the east of these localities. The junction with the Portland Group was not clearly seen at Brabourne. Details of the Brabourne succession and some fossil records were given by Lamplugh and Kitchin (1911). The occurrence of a 1 m thick breccia of fragments of mudstone and lignite in a matrix of greenish loam is noteworthy, as is the record of Protocardia purbeckensis indicating short-lived brackish or even quasi-marine conditions (Morter, 1984). About 6 m above the base of the succession at Brabourne, a layer of phosphatic nodules might possibly represent the Cinder Bed of expanded successions.

In the Boulonnais, the marine Portlandian is overlain by a thin (0.2-1.3 m), laterally variable succession (Calcaires des Oies) locally comprising tuffaceous concretionary algal limestones largely formed by the blue-green alga 'Spongiostromata' and closely comparable with the beds at the base of the Lulworth Formation in Dorset (Ager and Wallace, 1966). These limestones pass laterally into limestones composed of the bivalve Eocallista socialis, which is common in Dorset at the junction between the Portland and Purbeck groups (Townson and Wimbledon, 1979).

The Wealden Supergroup consists of two broad subdivisions, the Hastings Group below and the Weald Clay Group above. The former comprises an alternating succession of predominantly arenaceous formations (Ashdown Beds, Lower Tunbridge Wells Sands, Upper Tunbridge Wells Sands) and mudstones with lenticular sandstones (Wadhurst Clay, Grinstead Clay). The major sandstone units are typically overlain by pebble and/or bone beds marking transgressive events. The depositional environment of the Hastings Group has been reviewed in a series of papers by Allen (1959, 1975, 1989). The marked lithofacies change to mudstone sedimentation at the base of the Wadhurst Clay probably records the early Valanginian marine transgression at the end of the Continental 'Wealden' Formation period in Germany. This dating of the Wadhurst Clay as Valanginian is supported by charophyte evidence (Feist et al, 1995).

The change to mudstone deposition at the base of the Weald Clay Group is usually related to the early Hauterivian transgression in the marine realm. The Weald Clay Group comprises silty mudstones and muddy siltstones, typically with thin clay-ironstone beds near the base. In the middle part of the succession, there are groups of beds made up of low diversity or even monospecific gastropod associations, the so-called Small-'Paludina' beds (with Viviparus infracretacicus) and Large-'Paludina' Beds (with V. fluidram), in ascending order (Morter, 1978). A short distance above the Small-'Paludina' Beds, one or more horizons yielding quasi-marine faunas with the gastropod Paraglauconia [formerly cited as Cassiope] and bivalves including Cuneocorbula, Modiolus, Nemocardium and oysters (Praeexogyra) have been identified in the western Weald near Horsham, as well as in the Warlingham Borehole (Morter, 1978); charophytes from just above this level in the Warlingham Borehole indicate an early Barremian age (Feist et al, 1995). Above the (freshwater) Large-'Paludina' Beds, the depositional environment becomes increasingly saline, and the same quasi-marine molluscan assemblage reappears close to the top of the Weald Clay.

In the Hastings area, where the Hastings Group totals up to 385 m in thickness, the lower part of the Ashdown Beds is predominantly argillaceous (the Fairlight Clays of the old literature) and it is only the top 30 m that are dominantly sandy (Lake and Shephard-Thorn, 1987). The 'Fairlight Clays' have yielded a rich and well preserved flora, which includes pteridophytes, cycads and conifers. 20km to the east, the CEGB boreholes at Dungeness proved up to 165 m of Ashdown Beds, Wadhurst Clay and Tunbridge Wells Sands (Lake and Shephard-Thorn, 1987): although the boreholes started below the top of the Hastings Group, the total overall thickness can be inferred to be significantly less than at Hastings. The group crops out on the seabed to the north and south of the boundary faults of the faulted anticline NE of Dungeness Point (Crosby and Fletcher, 1988). The Hastings Group thins to the northeast and becomes progressively more arenaceous, so that the various subdivisions of the thicker successions, such as the Wadhurst Clay, can no longer be recognised. Near Dover, the sediments are particularly coarse and pebbly and the Hastings Group isopachytes record the position of a large valley trending in a southwest direction from the Palaeozoic uplands (Allen, 1967).

The succeeding Weald Clay Group is probably more or less complete in the western part of east Kent: Two 'Paludina' limestones 6 m apart were recorded 81 m from the top of the 122 m of Weald Clay in the Hothfield borehole near Ashford and an horizon with Praeexogyra is present 10 m below the top of the Weald Clay at Hythe. At Dover Colliery, the top of the Weald Clay at the contact with the Atherfield Clay is deeply burrowed. Here the occurrence of freshwater faunas with Viviparus and Unio in the top 1.5 m suggests that the highest part of the succession (with quasi-marine faunas) was removed by erosion at the time of the early Aptian marine transgression. In the St. Margaret's Bay Borehole, where the top of the Weald Clay is similarly burrowed, there is ostracod evidence for the higher part of the Weald Clay and a Large-'Paludina' limestone was found at the base above a downhole lithological change to arenaceous deposits of Hastings Group facies (Bisson et al, 1967; Worssam, 1978). Isopachytes for the Weald Clay (Worssam, 1978, Fig. 6) also show the valley identified in the Hastings Group near Dover and an additional narrow valley extending southwest from the basin margin in Thanet. In view of the lack of faunal evidence for the lower part of the Weald Clay in the east Kent boreholes, Worssam (1978) suggested that some of the supposed Hastings Group sands may actually be of Weald Clay age. This would fit with the picture, derived from seismic imaging, of the Weald Clay Group exhibiting a transgressive relationship to the underlying Hastings Group (Ruffell, 1992).

The Wealden thins from Ashford, where 217 m were recorded in a borehole (Smart et al, 1966), to 94.18 m at Brabourne and then to Dover Colliery where only 25.91 m remain. There is further thinning to the north, with the Weald Clay onlapping the feather-edge of the Hastings Group facies sands to rest directly on Coal Measures in a triangular area to the southwest of Ebbsfleet.

In the Boulonnais, the greatest known thickness of Wealden sediments is the 66.5 m of clays and sands recorded in a borehole at Wissant (Olry, 1904). Erosional relicts of grey lignitic clays of Wealden facies are found in pockets in Portland and Purbeck limestones near Boulogne and Wimereux. Sections of Wealden facies were formerly seen in quarries in the south of the Boulonnais near Nesles, but today the only available sections are in the eastern part of the area. Only the highest part of the succession is exposed, comprising up to 8 m of mottled reddish clays, overlain by black clays and white sands (Robaszynski and Amédro, 1986). Although Allen (1967) considered that the 'Wealden' sediments in the Boulonnais were comparable with the Hastings Group of southern England, palynomorph evidence (Herngreen, 1971) indicates a Late Barremian-Aptian age, pointing instead to a correlation with the higher part of the Weald Clay. There is thus an approximate parallel with the situation in east Kent, where sediments of Hastings Group facies may well be of Weald Clay age. The Wealden is overlain with erosive contact by either Lower Aptian or Lower Albian glauconitic sands, in each case with a basal concentration of phosphatic nodules.

The base of the Aptian in the Weald Basin is inferred to lie within the Weald Clay, by extrapolation from the Isle of Wight; where the early Aptian Vectis Magnetozone is situated within the correlative Vectis Formation (Kerth and Hailwood, 1988). There is a significant non-sequence in east Kent at the base of the Lower Greensand.

In Kent, the Lower Greensand Group at outcrop comprises, in ascending order, the Atherfield Clay Formation, the Hythe Beds Formation, the Sandgate Beds Formation and the Folkestone Beds Formation. The last three formations have their stratotypes in the vicinity of Folkestone. The Hythe Beds are overstepped from west to east by the transgressive Sandgate Beds and are already missing by Dover Colliery (Casey, 1961; Hesselbo et al, 1990; Ruffell and Wach, 1991). There are facies changes to the north-northeast as the Lower Greensand plunges undercover. The Atherfield Clay Formation in east Kent consists of bluish-grey, locally brown-mottled, sandy clay and pale grey, slightly glauconitic clay; typically with chocolate-brown clay in the lower part. It rests with sharp contact on the eroded, burrowed and bored top of the Weald Clay. Outcrop records and data from cored boreholes in the Kent Coalfield were reviewed by Casey (1961). There is a significant non-sequence at the basal contact: the Perna Bed and the overlying Chale Clay Member of the Isle of Wight succession are missing and the ammonite evidence (Casey, 1961) indicates that here the Atherfield Clay begins in the upper part of the Deshayesites forbesi Zone (D. callidiscus Subzone), the equivalent of the Crackers and Upper Lobster Beds. At Dover Colliery, the Atherfield Clay is 13.10 m thick and includes a bed of clay rich in Pinna robinaldina. There is a basal thin bed of grit with chert pebbles, bone fragments and Hybodus teeth. Simpson (1985) reinterpreted the ammonite evidence (Deshayesites spp., including the zonal index, as well as a single Roloboceras) to show that the succession could also include the equivalent of the Lower Lobster Bed (D. kiliani Subzone), albeit in the apparent absence of the subzonal index. However, Ruffell and Watch (1991) considered that the onlapping base of the Atherfield Clay in East Kent corresponded to the terminal forbesi Zone (callidiscus Subzone) Upper Lobster Beds transgression. The Atherfield Clay is overlain non-sequentially by the transgressive Sandgate Beds; the top of the clay exhibits greensand filled pholadid crypts, with the bivalve shells (Girardotia) still preserved (Lamplugh and Pringle, 1911, p1.1). In the St. Margaret's Bay Borehole (Bisson et al, 1967), the Atherfield Clay (5.13 m) yielded a rich fauna of molluscs, predominantly bivalves, as well as Deshayesites forbesi and Toxaster fittoni. Nearer to the margin of the basin, the Atherfield Clay at Harmansole is reduced to 0.6 m of unfossiliferous sandy clay with phosphatic pebbles.

The Hythe Beds Formation of the type area in east Kent comprises closely-spaced alternations of blue-grey limestones (locally known as rag) and poorly cemented, glauconitic, argillaceous sands (hassock). Fossils are preserved solid in the rag, but are typically crushed in the hassock. The Hythe Beds of the type area belong to the Lower Aptian Deshayesites deshayesi and Tropaeum bowerbanki zones. In the Maidstone area, the Hythe Beds succession includes strata of Late Aptian Cheloniceras (Epicheloniceras) martinioides Zone age, the Boughton Beds Member (Casey, 1961; Ruffell, 1993a,b; Worssam, 1993). The lower part of this unit in places comprises lagoonal deposits with plants and dinosaur remains, reflecting the martinioides Zone regression; towards the top, there are locally beds rich in sponges, with horizons of chert replacing the ragstones. Ammonite evidence (Casey, 1961) permits the recognition of the debile and gracile subzones, with the topmost subzone (buxtorfi) being possibly represented in the chert beds. Between Maidstone and Folkestone, erosion connected with the martinioides regression has resulted in the transgressive base of the Sandgate Beds resting on successively lower levels in the Hythe Beds. This major break in the Lower Greensand succession, recognised by Casey (1961) as the 'martinioides retrenchment' followed by the 'nutfieldiensis transgression', was identified (Hesselbo et al., 1990; Ruffell and Wach, l991) with the 109.5 Ma sequence boundary of the EXXON chart (Haq et al, 1988).

The original sections in Hythe are no longer available. The Otterpool Manor quarry [TR 112366], where 10.7 m of strata comprising the entire Lower Aptian Hythe Beds succession with the exception of the basal beds, were exposed beneath Sandgate Beds, is the standard reference section for the immediate area (Casey, 1961; Ruffell, 1993a), but is now backfilled. Most of this section was in rag and hassock facies; near the top of the Hythe Beds, a shell-rich calcareous sandstone with phosphatic nodules (Bed 30 of Casey, 1961) marked a transgression near the base of the meyendorffi Subzone of the Tropaeum bowerbanki Zone, above which the highest 1.5 m were developed in hassock facies with belemnites (Neohibolites ewaldi). At Mill Point, southwest of Folkestone, erosion has cut down to the transitoria Subzone of the bowerbanki Zone (Casey, 1961) and bored cobbles of rag limestone have been reworked into the basal Sandgate Beds. Ruffell (l993a) has recorded a 15 m Hythe Beds section at this locality by piecing together for the first time discontinuous exposures of the underlying beds.

The Sandgate Formation consists of glauconitic silty clays and silts, generally with a basal phosphatic nodule bed in glauconitic loam, which locally (e.g. Sellindge) yields the subzonal index of the highest (Cheloniceras (Epicheloniceras) buxtorfi) subzone of the martinioides Zone. Strata coeval with the highest Hythe Beds of the Maidstone area are thus represented northwest of Folkestone by a condensed sequence at the base of the Sandgate beds. At Mill Point, the base of the Sandgate Beds is marked by ferruginous boxstones impressed into the eroded top of the Hythe Beds. These boxstones include three components: debris from the erosion of Hythe Beds sediments; buff phosphatic nodules including C. (E.) buxtorfi; and an indigenous fauna of large brachiopods of nutfieldiensis Zone age. At Dover Colliery, the hiatus beneath the Sandgate Beds extends down to the top of the Atherfield Clay Formation (forbesi Zone). Farther to the north in the Ebbsfleet Borehole, the Sandgate Beds rest on Weald Clay and finally onlap the Palaeozoic floor (Kirkaldy, 1937). The lower part of the Sandgate Beds proper belongs to the Parahoplites nutfieldiensis Zone.

Nolaniceras spp. indicative of the (basal) N. nolani Subzone of the Hypacanthoplites jacobi Zone occur near the top of the formation west of Folkestone, together with a fauna dominated by Lamellaerhynchia caseyi (Casey, 1961).

The greater part of the Folkestone Beds Formation comprises typically white and/or yellow cross-bedded sands representing sandwave sedimentation under tidal conditions; the directions of foresets indicate movement to the southeast (Narayan, 1971; Bridges, 1982). Erosion surfaces marked by stratigraphical hiatuses identified by means of ammonites (Casey 1961) divide the Folkestone Beds as traditionally understood into three separate packages of sediment, which have been informally named the lower, middle and upper Folkestone Beds (Hesselbo et al, 1990) and studied in a sequence stratigraphical context. The upper Folkestone Beds of this classification, comprising the sediments of the topmost Lower Albian Douvilleiceras mammillatum Superzone overlain by the basal bed of the Gault, was treated by Owen (1992) as a separate unit of highly condensed beds, the Gault-Lower Greensand Junction Beds. Owen considered that the separation of this unit from the underlying and overlying beds was justified because horizons within it form major glide planes and are consequently of significance in civil engineering projects.

Spatial and stratigraphical relationships between the three units of the Folkestone Beds are highly complex: at no single locality is the entire succession preserved and the contact with the Sandgate Beds was only ever exposed in the coastal exposures on either side of Folkestone. Traced southeastwards from Brabourne to Folkestone (Smart et al, 1966, Fig. 1), successively higher units of the Folkestone Beds sands are preserved beneath the condensed mammillatum Superzone.

In the key, but now obliterated Sandling Junction sandpit section, 14 m of (lower) Folkestone Beds comprising uppermost Jacobi Zone Hypacanthoplites anglicus Subzone (terminal Upper Aptian) sands were formerly exposed. Soft rotted ironstone concretions in clay-rich sands near the base of the succession in both the pit and in the nearby railway cutting yielded Hypacanthoplites simmsi. A bed of partly phosphatised hollow nodules in the pit contained rare ammonites of the anglicus Subzone (H. simmsi and H. anglicus), together with a rich molluscan fauna, including a gastropod with the intestines exceptionally preserved in phosphate (Casey, 1960). The same ammonites were found in a bed of ferruginous sandstone (Red Bed) towards the top of the anglicus sands. There is no evidence in the exposed part of the Folkestone Beds here of the underlying Hypacanthoplites rubricosus Subzone, although this subzone is represented in the complex basal nodule bed east of Folkestone (see below). The recent discovery in a temporary motorway section near Lenham of indigenous H. simmsi in nacreous, rather than phosphatic preservation, in a bed of silty clay inferred to be at the top of the Sandgate Beds (Ruffell and Owen, 1995) is of significance in this context.

These lower Folkestone Beds sands were overlain with erosive contact (erosion surface LG3 of Hesselbo et al, 1990) by Lower Albian H. milletioides Subzone sands with sandstones containing the subzonal index. The LG3 hiatus includes the Aptian-Albian boundary: the phosphatic nodules (anglicus nodules) marking this hiatus include ammonites of the anglicus Subzone, but there is no evidence for the basal Albian farnhamensis Subzone.

In the two Folkestone sections, the lower Folkestone Beds anglicus sands seen to the west are no longer found and the milletioides Subzone sediments are overlain by similar beds of regularis Subzone age, the two subzonal packages constituting the middle Folkestone Beds of Hesselbo et al (1990). At Mill Point, west of Folkestone, the anglicus nodules rest directly on the Sandgate Beds and the milletioides Subzone is only 5.5 m thick; to the east at East Cliff, the bored surface of the Sandgate beds was formerly seen (Casey, 1961) to be overlain by a complex nodule bed in which the black anglicus nodules rested on spherical ferruginous nodules with a Hypacanthoplites rubricosus Subzone fauna. Here the milletioides Subzone sediments are reduced to 60 cm of sandy clays and the greater part of the (middle) Folkestone Beds belong to the Leymeriella regularis Subzone. Ruffell and Wach (1991) considered that the 107.5 Ma sequence boundary of the EXXON chart (Haq et al, 1988) was probably expressed by the end of sand deposition in regularis Subzone times. The Gault-Lower Greensand Junction Beds sensu Owen (1992) contain three important fossiliferous horizons. The base is marked by the kitchini nodule Bed, which yields phosphatised ammonites of the Sonneratia kitchini Subzone. The Main mammillatum Bed, some distance higher, includes fossils in two states of preservation incorporated in a locally cemented grit: phosphatic pebble fossils derived from the Cleoniceras floridum and Otohoplites raulinianus subzones; and partly phosphatised, indigenous fossils of Protohoplites (Hemisonneratia) puzosianus Subzone age. The Sulphur Bed at the top of the traditional Folkestone Beds and so named from the mineral jarosite (hydrated potassium iron sulphate: a decomposition product of pyrite) comprises greyish, pyritic phosphatic nodules set in a dark grey loam with pyrite (Ruffell, 1990). Owen (1992) has recorded Otohoplites crassus from the Sulphur Band at an inland section near Folkestone and has consequently assigned this bed to the Otohoplites ebulliences Subzone, but noted that there was also some admixture of derived puzosianus Subzone material. In the nodules immediately overlying the Sulphur Bed at Copt Point there are rare specimens of Hoplites (Isohoplites) steinmanni (?= eodentatus), indicating the eponymous subzone of the Hoplites dentatus Zone and the base of the Middle Albian as interpreted by Casey (in Smart et al, 1966) and by Destombes (1977).

In the Boulonnais, there is no record of earliest Aptian marine sediments equivalent to the Atherfield Clay Formation and the lower part of the Hythe Beds Formation. The succession begins with a glauconitic sandstone with Cheloniceras (Formation de Cat-Cornu), which is found in beach exposures near Wissant. Elsewhere in the Boulonnais, a period of intra-Aptian erosion has removed this sandstone, and it is represented at Menty in the south of the area only by a lag of sandy phosphatic nodules at the base of the Formation de Verlincthun, resting on the burrowed top of Wealden white sands. The nodules have yielded brachiopods and bivalves, as well as a rich ammonite fauna which indicates a concentration of the Deshayesites grandis Subzone of the D. deshayesi Zone and the Tropaeum bowerbanki Zone. The Formation de Cat-Cornu thus correlates with the higher part of the Hythe Beds Formation of the type area. The overlying Formation de Verlincthun comprises glauconitic sands with one or more horizons of large oysters (Ostrea leymerii, Aetostreon latissimum and Rastellum macropterum), over-lain by cross-bedded white sands. In the absence of ammonites, these beds have been correlated on general similarity of facies with the Sandgate Beds Formation, although there is no sign of the martinioides Zone nodule bed at the base.

The white sands at the top of the Formation de Verlincthun terminate in an irregular omission surface and are overlain by the Formation de Wissant, about I m of glauconitic sandy clays with black phosphatic nodules and sandy phosphatic concretions. The latter yield a rich ammonite assemblage comprising species of Hypacanthoplites including H. anglicus and H. rubricosus, indicating the eponymous subzones of the H. Jacobi Zone. This bed is the equivalent of the complex rubricosus/anglicus nodule bed at the base of the Folkestone Beds Formation in the Folkestone East Cliff section. The latter formation (including the Gault-Lower Greensand Junctions Beds of Owen) is represented at

Wissant by the Formation des Gardes, comprising only I m of dark green coarse glauconitic sands and sandstones intercalated between two beds of phosphatic nodules designated P1 and P2 by French stratigraphers. P1 yields reworked Hypacanthoplites together with indigenous Birostrina salomoni and Burrirhynchia leightonensis as well as ammonites in pebble preservation that are indicative of the Leymeriella regularis Subzone and also the Cleoniceras floridum Subzone of the mammillatum Superzone. In the context of the Folkestone Beds succession, the lithological equivalent of the main mammillatum bed rests here directly on the expanded correlative of the rubricosus/anglicus nodule bed, the regularis Subzone sands and sandstones of Folkestone being unrepresented by sediment. However, in the southeast Boulonnais near Samer, sediments of this age, comprising glauconitic sands with Leymeriella were reported beneath P1 by Robaszynski and Amedro (1986).

The higher part of the Formation des Gardes in the vicinity of Wissant is laterally variable and there are significant differences between the section normally visible (see Robaszynski and Amedro, 1986) and that occasionally exposed in the area of sand dunes 0.8 km northeast of Wissant (Owen, 1992). In the latter section, a terminal bed of large, flattened greyish brown nodules with an early Otohoplites bulliensis Subzone ammonite fauna is separated by only 5cm from a bed of glauconitic clay containing pyrite-encrusted phosphatic nodules with a pusozianus Subzone fauna (Owen, 1992, Fig. 16). In the normally exposed section, these two beds merge to produce the phosphate bed P2 with the subzonally mixed ammonite fauna listed by Robaszynski and Amédro (1986). The lithological and ammonite subzonal correlation between the two Wissant successions and the succession at Folkestone is highly complex (Owen, 1992, Fig. IS) and will not be discussed further here.