Bridgland, D.R. 1994. Quaternary of the Thames. Geological Conservation Review Series No. 7. JNCC, Peterborough, ISBN 0 412 48830 2.
The original source material for these web pages has been made available by the JNCC under the Open Government Licence 3.0. Full details in the JNCC Open Data Policy
D.R. Bridgland
The first of the three divisions of this chapter is devoted to sites associated with the pre-diversion Thames deposits of the Kesgrave Group. These deposits have only been accepted as the direct products of Thames drainage since the definition of the Kesgrave Sands and Gravels (Rose et al., 1976; Rose and Allen, 1977). However, the typical sequence of Pleistocene sediments in which they occur, which is widespread in East Anglia, has been the subject of research extending back over a century. The main components of this sequence are the Kesgrave Sands and Gravels, mapped until very recent years as 'Glacial' by the Geological Survey, and the (Anglian Stage) Lowestoft Till, which overlies the early Thames gravels over much of their distribution. Localized occurrences of outwash gravel and aeolian deposits (coversand and loess) have been observed between the Thames gravels and the till. Other stratigraphically important parts of the sequence are superimposed warm- and cold-climate fossil soils, which occur at the top of the Kesgrave Group deposits throughout the area.
The occurrence of gravels beneath the East Anglian till sheet was first noted by Wood (1867, 1870; Wood and Harmer, 1868), who devised a tripartite stratigraphical scheme, based on the till-sand-till sequence of north-eastern Norfolk (now classified as Cromer Till, overlain by Corton Sands, overlain in turn by Lowestoft Till). In this scheme, the three divisions were respectively classified as 'Lower Glacial', 'Middle Glacial' and 'Upper Glacial'. Wood's classification was subsequently used by the Geological Survey when mapping Essex (Whitaker, 1877, 1889; Dalton, 1880). The upper (Lowestoft) till was widely recognized in this area and gravels underlying it were assumed to equate with those classified as 'Middle Glacial' in Norfolk. This view was strengthened by the recognition, at a few sites (such as Maldon — see below), of a till below gravel of similar type (Whitaker, 1889).
Prestwich (1881) observed that the majority of gravels underlying the till in eastern Suffolk were composed of large quantities of rounded flint and quartz, with significant amounts of subangular flint and Lower Greensand material, suggesting a southern derivation, in contrast to the northern provenance associated with the 'Glacial' drifts (including the 'Middle Glacial Gravel' of north-east Norfolk). He noted that these gravels, which he termed Westleton Beds', ranged as far afield as Essex and the Thames valley. He later traced possible equivalents of these beds, which he regarded as pre-'Glacial' marine deposits, over much of southern England (Prestwich, 1890b, 1890c). The distinction between the 'Glacial Beds' and Westleton Beds' of the Braintree area was outlined in some detail by French (1891). Despite these early attempts at clarification, the deposits later to be recognized as early Thames gravels (belonging to the Kesgrave Group) were variously classified by early workers as both 'Glacial Gravel' and Westleton Beds'. Furthermore, the early Geological Survey mapping recognized only 'Glacial Gravel', a category that included both the early Thames deposits and true glaciofluvial deposits. Whitaker (1889, p. 299) emphasized that the term 'Glacial' (with a capital G), as applied by the Geological Survey, was 'a proper name for a geologic period' (this can be correlated broadly with the post-Cromerian Pleistocene of modern usage) and not to be confused with the adjective meaning associated with or produced by ice. Unfortunately, such confusion became commonplace, the deposits mapped as 'Glacial Gravel' being widely interpreted as glaciofluvial outwash until the 1970s (see below).
Salter (1896) disputed the evidence for the marine origin of the deposits in Essex classified as Westleton Beds' by Prestwich. He pointed out that the rounded character of many of the flints in these gravels was inherited from the Palaeogene pebble beds that were their immediate source. Salter (1905) later included these beds in a series of gravels, traceable along the Chiltern dip slope from the Goring Gap towards the North Sea (Chapter 3), which he regarded as the product of an early northeastward flowing drainage system that existed prior to the formation of the modern Lower Thames valley. Important evidence for the former existence of this drainage system, according to Salter, was the concentration of southern rocks in high-level deposits between the Lower Thames valley and the Chilterns. He realized that such gravels, containing abundant clasts of southern origin, were the products of right-bank tributaries of the old north-eastward flowing river.
Gregory (1894, 1922) regarded the bulk of the gravels of central Essex as 'pre-glacial' and divided them into 'High Level Quartzite Gravels' (or 'Danbury Gravels') and 'Low Level Quartzite Gravels' (or 'Brain Valley Gravels'), the latter coinciding largely with Prestwich's 'Westleton Beds'. Gregory believed that all these gravels had been deposited by streams flowing from the west, on a regional slope resulting from late Oligocene to early Miocene uplift. However, Solomon (1935) considered the two levels of 'Quartzite Gravels' recognized by Gregory to belong to a single series disturbed by tectonic activity, incorporating the Westleton Beds' (sensu Prestwich) and Wood's 'Middle Glacial Gravel'. Solomon classified these deposits as his Westleton Series', which he regarded as marine, although with a glacial intercalation in Norfolk. He ascribed lateral differences in composition to local variations in provenance within a single depositional basin. Solomon (in Clayton, 1957) claimed a distinction, based on heavy-mineral analysis, between two types of gravels in Essex: one, previously classified as 'Middle Glacial Gravel' or 'Westleton Gravel', pre-dated the Essex till sheet, whereas the other he regarded as a true outwash deposit, intimately associated with the till.
Warren (1955, 1957) reported that gravels deposited by the Thames in pre-Chalky/Jurassic Till (Lowestoft Till) times had been traced by Baden-Powell and himself from Oxford, across central Essex to the Clacton area. This interpretation, which confirmed the views of Salter (1905), was the first to indicate that Thames deposits could be recognized within the Pleistocene record of southern East Anglia. Based on clast lithology, but without systematic analyses, it anticipated the results of later work by Rose et al. (1976; Rose and Allen, 1977; see below).
Whilst progress was being made in the interpretation of the early gravel aggradations of southern East Anglia, research was being carried out in parallel on the glacial deposits that commonly overlie them. The 'monoglacial' model favoured by the 19th century workers was replaced by one that recognized four separate glaciations, as first envisaged in the Alpine region by Penck and Bruckner (1901–1909). However, deposits of only the last three of these were identified in East Anglia, with the last confined to the extreme north-western corner of Norfolk (West, 1963). The two previous glaciations were considered to be represented by superimposed tills in East Anglia, differentiated on the basis of composition by Baden-Powell (1948) into a lower 'Lowestoft Till' and an upper 'Gipping Till'. Confirmation of this distinction was provided by studies of till fabrics by West and Donner (1956), who found significant differences in this respect between the Lowestoft and Gipping Tills and suggested respective correlations with the continental Elsterian and Saalian glaciations.
Clayton (1957) proposed a complex subdivision of the glacial deposits of the Chelmsford area, in which he recognized an older, weathered and dissected till (Hanningfield Till), largely covering plateaux, and a younger 'sandwich' of deposits filling valleys, comprising lower and upper tills separated by gravel. To this later tripartite sequence he applied the names Maldon Till, Chelmsford Gravels and Springfield Till, in ascending stratigraphical sequence. The Maldon Till was thought to be highly localized, whereas the Springfield Till was considerably more widespread. The Chelmsford Gravels, generally equivalent to the 'Middle Glacial Gravel' of Wood and Harmer (1868), were interpreted as outwash deposits. Clayton (1957) suggested a correlation between the Hanningfield Till and the continental Elsterian (= Anglian) glaciation and between the later tripartite sequence (Maldon Till, Chelmsford Gravels and Springfield Till) and the Saalian Stage. However, in an appendix to a later paper he proposed correlations with glacial drifts in Norfolk and Suffolk that implied an Anglian age for the Hanningfield and Maldon Tills (seen as equivalent to the Cromer and Lowestoft Tills, respectively) and a Saalian (Gipping Till) age for the Springfield Till (Clayton, 1960). In this later paper he suggested a Hoxnian age for the Chelmsford Gravels, which would seem to preclude their interpretation as glaciofluvial sediments, although no explanation of the change of view was provided. Clayton's earlier interpretation of the gravels as glaciofluvial was largely reiterated by Baker (1971), who suggested that the area may have been occupied by a pre-glacial Thames and that the deposits of this river may have been reworked and incorporated in the outwash.
The distinction of the products of two separate glaciations in southern East Anglia was seriously questioned by Baker (1971), who found that the Hanningfield and Springfield tills could not be differentiated either by lithological, stratigraphical or morphological evidence in north-west Essex. Baker also suggested that the Maldon Till resulted from a minor advance of the main 'Chancy Boulder Clay' glaciation of East Anglia (see below, Maldon), implying that all the glacial deposits of Essex are the product of a single glaciation. Subsequent mapping and lithological analyses of tills throughout East Anglia and the East Midlands led to confirmation that only a single glacial episode is represented amongst the Chalk-rich glacial deposits of these areas, equivalent to the Lowestoft Till of Norfolk (Bristow and Cox, 1973; Perrin et al., 1973; Perrin et al., 1979). Separate recognition of the partly Scandinavian Cromer Tills/ North Sea Drift glaciation was maintained, but the Gipping Till in its type area (the Gipping valley, near Ipswich) was found to be nothing more than a weathered profile in the upper part of the Lowestoft Till. Evidence for a glaciation equivalent to the continental Saalian was, however, accepted in the West Midlands and, with a good deal of caution, in the Breckland (Turner, 1973); a new name Wolstonian' was given to this post-Hoxnian and pre-Ipswichian glaciation (Shotton, 1973b; Chapter 1). Straw (1979, 1983) has continued to argue for an earlier 'Wolstonian' glaciation in northern East Anglia, principally on geomorphological grounds, and Wymer (1985b) has described sites in Suffolk and Norfolk where glacial deposits might overlie Hoxnian sediments. However, the view that only a single (Anglian Stage) glaciation, albeit with multiple ice advances, can be recognized in Essex has been consolidated in recent reviews (Baker and Jones, 1980; Whiteman, 1987; Allen et al., 1991).
Even in the Wolstonian type area around Coventry there is growing uncertainty over the distinction between the deposits attributed to this glaciation and the Anglian deposits to the east (Perrin et al., 1979; Sumbler, 1983; Rose, 1987). There are clear stratigraphical reasons for considering the whole of the 'Chalky till' to have been deposited during a single glaciation, during the Anglian Stage. Doubts have, however, been expressed elsewhere in this volume about the tenability of any model that places all these glacial deposits within a single glacial episode. These doubts stem from stratigraphical correlations of Chalk-rich tills on the Cotswolds and in the Vale of St Albans with the Thames terrace sequence, which suggest that the glaciations recognized in these two areas may have been separated by a temperate episode (see Chapter 1; Chapter 2, Long Hanborough and Wolvercote; Chapter 3, Part 2). It may be feasible, in the light of the more complex chronology indicated by the record from deep-sea cores (Chapter 1), for two separate glacial episodes to be represented within the 'Chalky Till' of eastern England, both of them post-Cromerian (sensu West Runton) and pre-Hoxnian (sensu Hoxne).
Re-evaluation of the gravels underlying the till sheet of southern East Anglia has also taken place over the past quarter of a century. Hey (1967), following considerable reinvestigation of the deposits, concluded that the Westleton Beds' of the type area in north-east Suffolk were truly marine. However, he found elsewhere that gravels described by Prestwich (1881, 1890b, 1890c) under that name, and attributed by Solomon (1935) to his Westleton Series', differed from the type Westleton Beds and could frequently be shown to be younger. These gravels, generally mapped by the Geological Survey as part of the glacial sequence, were defined as the Kesgrave Sands and Gravels by Rose et al. (1976) and Rose and Allen (1977), who interpreted them as periglacial fluvial deposits of probable Thames origin. These authors demonstrated a distinction, on the basis of clast-composition, between the early Thames gravels and less extensive glacial outwash deposits, which they named Barham Sands and Gravels. They confirmed, therefore, the distinction made by Solomon (in Clayton, 1957) using heavy-mineral analysis.
Rose et al. (1976) proposed an important new stratigraphical scheme for southern East Anglia. From a study of sites throughout the area between Ongar (in the south-west), Thetford (in the north-west) and the coast, they found that the Lowestoft Till overlies widespread fluvial deposits of the pre-diversion Thames, their Kesgrave Sands and Gravels. The upper part of these deposits was frequently found to be rubified and clay-enriched, features indicative of pedogenic activity in a warm climate (Rose and Allen, 1977; Kemp, 1985a; Rose et al., 1985b). This horizon occurs immediately beneath the Lowestoft Till or its associated outwash (Barham Sands and Gravels), indicating that soil development occurred prior to the Anglian glaciation. It was therefore concluded that the uppermost parts of the Kesgrave Group gravels incorporate the illuvial horizon of an interglacial soil, named the Valley Farm Soil (Rose et al., 1976; Rose and Allen, 1977; Kemp, 1985a). As the Kesgrave Sands and Gravels overlie the Chillesford Beds at Chillesford Church Pit, which at that time were ascribed to the Pastonian Stage (Turner, 1973), the early Thames gravels were placed by Rose et al. (1976) and Rose and Allen (1977) in the subsequent cold stage, the Beestonian, and the palaeosol, its age constrained by the overlying Anglian glacial sediments, in the Cromerian Stage.
It is clear that intensely cold conditions prevailed following the formation of the Valley Farm Soil, even before the deposition of the Lowestoft Till; at many localities, including the GCR site at Newney Green, a periglacial soil is superimposed on the earlier warm-climate one. This periglacial soil, termed the Barham Soil, was recognized from both large- and small-scale structures related to frost activity; the larger structures include involutions and ice-wedge casts, whereas the smaller ones include disrupted clay skins and fractured gravel clasts (Rose et al., 1976, 1985a; Rose and Allen, 1977). As well as being developed in the upper parts of the Kesgrave Group gravels (incorporating the earlier Valley Farm Soil), the Barham Soil has been shown to have developed in Cromerian interglacial sediments in north Norfolk. In such cases, where overlain by Anglian Stage tills, this soil has been established as a 'soil stratigraphic unit' of early Anglian age (Rose et al., 1985a). Without these overlying deposits, in areas where Kesgrave Group gravels form the present land surface, it is possible to identify relict features of both the Barham and Valley Farm Soils (Rose et al., 1976, 1985a, 1985b; Rose and Allen, 1977; Kemp, 1985a). However, later temperate‑ climate soils have been developed in the top of the Lowestoft Till (Rose et al., 1978; Sturdy et al., 1978) and cryoturbation has probably occurred during several cold episodes since the Anglian glaciation, so the recognition of pedogenic features characteristic of the Valley Farm and Barham Soils is of equivocal stratigraphical value beyond the glaciated area. Loess and coversand have also been recorded between the Kesgrave Sands and Gravels and the Lowestoft Till (Rose et al., 1976, 1985a; Rose and Allen, 1977). These sediments, ascribed to the Anglian Stage, are frequently incorporated in the large-scale cryoturbation structures of the Barham Soil.
The recognition of a widespread Cromerian soil developed on the early Thames deposits of central and northern Essex led Rose et al. (1976) to suggest that the river had migrated to a more southerly route by that stage. However, subsequent work in the Vale of St Albans (Gibbard, 1974, 1977, 1978a; Green and McGregor, 1978a, 1978b; Chapter 3) and in eastern Essex (Bridgland, 1980, 1983a, 1983b, 1988a; Part 2 of this chapter) has indicated that the Thames was not diverted from its early course through Hertfordshire and central Essex until the Anglian glacial maximum. A possible explanation for this apparent contradiction is that the palaeosol seen beneath the Lowestoft Till in many gravel pits in southern East Anglia was formed on the terraces of the pre-glacial Thames valley, whereas the valley floor and channel occupied by the river immediately prior to its diversion, in the centre of the Mid-Essex Depression of Wooldridge and Henderson (1955), is deeply buried by Lowestoft Till and not exposed. In the unglaciated area east of Colchester, the lowest formation within the Low-level Kesgrave Subgroup has been ascribed to the Anglian Stage (Bridgland, 1980, 1983a, 1988a; Bridgland et al., 1988, 1990). The Valley Farm Soil is not developed on this formation (see below, St Osyth and Holland-on-Sea).
Study of the clast composition of the Kesgrave Sands and Gravels reveals important minor components of value as provenance indicators. The gravels generally contain 0.5–2% sponge-spicular chert derived from the Lower Greensand of Kent and Surrey (Bridgland, 1980, 1986b; Green et al., 1982). Carboniferous chert, for which there are a number of potential sources to the north and west, all of them outside the London Basin (Bridgland, 1986b), typically accounts for 1–2% of the total gravel content (Bridgland et al., 1990). Also important is a restricted suite of volcanic rocks, some of which have been tentatively traced to sources in North Wales (Hey and Brenchley, 1977; Whiteman, 1983). The abundant quartz and quartzite (up to 35%) must also have been derived from outside the London Basin, the Midlands and Welsh borderlands representing the most likely source areas (Bridgland, 1986b). As was recognized by Rose et al. (1976), these various non-flint components are, when found in gravels in the London Basin, indicative of a Thames origin. A similar assemblage of component rocks has been recognized in the early terrace gravels of the Middle Thames (Green and McGregor, 1978a; McGregor and Green, 1978; Gibbard, 1985) and in the Northern Drift of Oxfordshire, also believed to represent early Thames aggradations (Hey, 1986; Chapter 2). Whether these rock-types were introduced into the Thames system by glaciation(s), by fluvial transport in a more extensive catchment or by a combination of the two remains a subject of controversy (Bowen et al., 1986a; Bridgland, 1986b, 1988c; Hey, 1986; Whiteman, 1990; see Chapter 1).
Rose et al. (1976) and Rose and Allen (1977) recognized that their Kesgrave Sands and Gravels might represent a series of terrace aggradations, but the first formal subdivision was proposed by Hey (1980), who recognized the downstream continuation of his Westland Green Gravels of the Middle Thames (Hey, 1965; Chapter 3) within the older, higher-level part of the group
Recent work by Whiteman (1990) has shown that the Westland Green Gravels as recognized in Suffolk do not, in fact, correlate with the unit of the same name originally defined by Hey (1965) in Hertfordshire and the Middle Thames. By determining (largely from borehole records) the distribution of the various formations within the Kesgrave Group in Essex and Suffolk, Whiteman has demonstrated that the formation in Suffolk that has been termed Westland Green Gravels is, in fact, the downstream continuation of the Gerrards Cross Gravel of the Middle Thames (see Chapters 1 and 3). It is therefore necessary to make considerable changes to the schemes for correlating the pre-diversion gravels of the Thames valley with the formations of the Kesgrave Group that have been published in recent years (Hey, 1980; Green et al., 1982; Gibbard, 1983; Green and McGregor, 1983; Bowen et al., 1986a; Bridgland, 1988a).
Whiteman's work has wider repercussions than are outlined above. He was able to suggest correlations, based on altitude and variations in clast composition, between central Essex and the higher parts of the terrace sequences upstream in both the Middle and Upper Thames regions (see Chapters 3 and 2 respectively). He concluded that the earliest true Thames deposits, from the Stoke Row Gravel up to and including the Gerrards Cross Gravel, are the product of a very large river, with a catchment extending far beyond the present Upper Thames (as already envisaged by Hey (1986)). In East Anglia these formations have all been included within the High-level Kesgrave Subgroup.
The formations of the Low-level Kesgrave Subgroup, thought until recently to represent the downstream continuation of Middle Thames terrace formations between the Satwell and Winter Hill Gravels (inclusive), have been shown by Whiteman to fall entirely between the deposition of the Gerrards Cross Gravel (last of the High-level Kesgrave Subgroup formations) and the diversion of the river. Therefore an interval that in the Middle Thames is represented only by the Winter Hill and Rassler Gravels (see Chapters 1 and 3) is more fully represented by the Low-level Kesgrave Subgroup in Essex, which comprises four separate formations. Despite this revision, the correlation of the lowest of these with the Winter Hill Formation of the Middle Thames (Bridgland, 1980, 1983a, 1988a), based on stratigraphical evidence resulting from the diversion of the river, is upheld (see below, St Osyth and Holland-on-Sea).
Bridgland (1988a) has suggested that the deposits in northern Essex that are now classified as Low-level Kesgrave Sands and Gravels occupy a separate valley from that in which the formations of the High-level Kesgrave Subgroup were deposited. The latter valley lies further to the north and west and is separated from the former, in south-east Suffolk, by an interfluve of residual Red Crag (Allen, 1983). Further upstream the Low-level Kesgrave Subgroup formations are preserved in central Essex as the lowest tiers in a terrace 'staircase' of Kesgrave gravels that is largely obscured by later (Anglian) glacial deposits. Formations within the High-level Kesgrave Subgroup form the higher part of this 'staircase', which declines from north-west to south-east, towards the modern Chelmer and Blackwater valleys. To the south and east of these valleys, possible outliers of High-level Kesgrave Sands and Gravels have been identified on Danbury Hill and on the Tiptree Ridge (Hey, 1980; Bristow, 1985;
Important information about the age and stratigraphy of the Low-level Kesgrave Sands and Gravels has been gained in recent years from the discovery, at a number of sites in Essex, of inter glacial sediments interbedded with these early Thames gravels (Bridgland, 1988a; Bridgland et al., 1988, 1990; see below, Ardleigh, Little Oakley and Wivenhoe). It is hoped that the detailed interpretation of these sites will provide a framework for the relative dating of the lower Middle Pleistocene Low-level Kesgrave Subgroup, which can be extended to assist with the dating of the sequences in other parts of the Thames Basin and permit correlation with the sequences further north in East Anglia and on the Continent.
In addition to the recognition of individual (terrace) formations within the Low-level Kesgrave Subgroup, differences in gravel composition enable the distinction of western and eastern subdivisions of the lowest three of these formations on the Tendring Plateau (Bridgland, 1988a; Bridgland et al, 1990;
| West (Thames) | East (Thames-Medway) |
| Waldringfield Gravel | |
| Ardleigh Gravel | Oakley Gravel |
| Wivenhoe Gravel | Cooks Green Gravel |
| Lower St Osyth Gravel | Lower Holland Gravel |
The Lower St Osyth Gravel and its Thames-Medway equivalent, the Lower Holland Gravel, are so called because they are overlain by later deposits, the Upper St Osyth and Upper Holland Gravels, that are not typical of the Kesgrave Group. These deposits can be closely correlated with sediments and events further upstream in the old Thames valley. They indicate the cessation of normal Thames drainage when the river was blocked by the Lowestoft Till ice sheet. The Upper St Osyth and Upper Holland Gravels both contain distal outwash material from the ice sheet, but as the Medway valley was unglaciated, the latter river made a significant and uninterrupted contribution to the Upper Holland Gravel (Bridgland, 1983a, 1988a;