Bigbury Camp and its associated earthworks: recent archaeological research

Bigbury Camp and its associated earthworks: recent Archaeological research

christopher sparey-green

Over the last decade research in the area of Bigbury Camp, Harbledown and Rough Common, west of Canterbury, has identified a wide range of sites in the vicinity of the known hillfort, proving this to be only one element of a wider complex of earthworks in and around the woodland of the South Blean and extending north of the A2 (Fig. 1).1 Fieldwork undertaken by the writer in 2008-2010 for the Kent Wildlife Trust at Bigbury and in the South Blean reserves was complemented by a LiDAR survey which confirmed the linear earthworks extending from the hillfort into the South Blean, west of Chartham Hatch. A community project on behalf of the Wildlife Trust followed this with a more detailed study of woodland history (Bannister 2013). These surveys also confirmed the existence of earthworks north of the A2, extending from Manson Wood east to the major enclosure in Homestall Wood, the focus of fieldwork in 2015-18.²

The Blean is an extensive area of woodland interspersed with pasture and hop fields occupying high ground west of the Stour valley and the city of Canterbury. The major known monument is the Iron Age hillfort of Bigbury Camp on the western side of the valley, other earthworks beyond to the west and north-west setting the hillfort in a landscape with a long and complex history predating the woodland cover. This paper summarises recent research within areas of the parishes of Harbledown and Rough Common, Thanington Without, Chartham and Dunkirk. Aerial survey and recent fieldwork suggests the complex extends even further north-east along the Stour valley to St Thomas Hill and Sturry, and south to earthworks along the south-facing scarp of the Chartham Downs, in the parishes of Chartham and Petham, east of the valley.

The earlier elements of a palimpsest landscape are outlined here, concentrating on earthworks of later prehistoric or early Roman character mostly west of the Stour. The sites will be described in a clock-wise direction starting at Bigbury before considering the woodland west of Chartham Hatch and then moving to the areas either side of the A2 and the woodland above Upper Harbledown. Lastly, outlying features in the area of Rough Common and St Thomas Hill will be briefly considered, as well as sites on Chartham Downs.³

The Topography and Landscape History of the Stour Valley south-west of Canterbury

The Blean occupies an irregular plateau projecting north-east from the North Downs, west of the Stour. The area under consideration extends over almost 17 sq. km within the southern section of the massif, centred on the bowl of land now occupied by the hop fields and orchards around Denstead Farm (NGR 6101 1579) (Fig. 1). This is here termed the ‘Denstead Bowl’, and is bounded on the south by the high ground extending south-west from Bigbury to Chartham Hatch and west to Rhode Common near Selling. Earthworks within the woods here overlook a scarp slope to the Stour, further linear features on the Chartham Downs on the opposite bank perhaps also enhancing that south-facing barrier. West of the river the high ground rises to 120m aod at Dunkirk with a steep drop beyond to the distant Medway valley and the Graveney Marshes.⁴ From Dunkirk a ridge descends east, bounded by a wide valley on the north dropping towards the village of Blean, the ridge occupied by earthworks extending from Church Wood to Willows and Homestall Woods. Beyond Rough Common, St Thomas Hill overlooks St Dunstan’s and the original crossing of the Stour at Canterbury. Springs on the perimeter of the Denstead Bowl have created steep-sided streams which feed the river Cranbourne as it leads east to the Stour below Harbledown. The earthworks in Willows Wood and Homestall Wood are partly defined by these gullies.

The topography reflects a varied geology with, on the west, the underlying Chalk overlain by sands and gravels of the Thanet and Woolwich beds which extend into the Denstead Bowl. To the west and north these are overlain by the London Clay, in turn covered by areas of River Terrace gravels. East of the Stour the Chartham area is Chalk with superficial deposits of sand, gravel and clay.

The prehistory of the Blean woodlands remains to be researched but the earliest documentary references date to the middle Anglo-Saxon period (Holmes and Wheaten 2002).⁵ The origin of the name may derive either from a root referring to the existence of a boundary or descriptive of rough ground (Wallenberg 1934). Settlements such as Chartham Hatch and Upper Harbledown were established as assarts carved out from the wood in the former case, or by settlement along the Roman road line in the latter. The medieval Pilgrims’ Way, in origin probably a prehistoric ridgeway, approaches Bigbury from the south-west, before converging on this route descending to the river crossing at St Dunstan’s. Extensive old-established woodland also covers the area south of Chartham on the east bank.

History of research in the area

The identification of the Canterbury area with events recorded in Caesar’s commentaries is a topic beyond consideration here but the local tradition that Chilham Castle was a Caesarian camp and that Julliberrie’s Grave on the opposite bank of the Stour was the burial place of Q. Laberius Durus, the only named casualty of the second expedition, was recorded by Camden in 1586.⁶

The earliest reference to ancient earthworks in the vicinity of Canterbury was made a century later by Dr Robert Plot of Oxford who, while visiting Canterbury in 1693, reported visiting a double entrenchment 3 miles from the city, presumed by him to be the stronghold attacked by Caesar in 54 bc (Rawlinson 1714).⁷ This is perhaps an early identification of Bigbury but during the following one and a half centuries the site was lost and is not marked on maps of the eighteenth century, both the first edition one-inch Ordnance Survey maps and the Tithe Apportionment map referring only to Bigberry Wood.⁸

The existence of a major archaeological site was revealed by the groups of iron work and other finds recovered during gravel digging on the hilltop in the period 1861 to 1902 and at later dates. The entrenchments were first recognised as those of a hillfort by Hussey in 1874 and later by Rice Holmes as a camp conforming to Caesar’s definition of a woodland stronghold and the likely site of the assault described.⁹ In the same period, Vine promoted the complex earthworks on Barham Down above Bridge on the Lesser Stour as the site of the first major encounter with British forces and the location of both British strongholds and a major Roman base, but without reference to Bigbury.¹⁰

Following the nineteenth-century finds, the site was the subject of two major campaigns of excavation, the first in 1933-4, the second in 1978-80, the latter providing the first radiocarbon dating (Jessup and Cook 1936; Thompson 1983; Clark and Thompson 1989). Important work on the defences near the putative site of the east gate was also undertaken in 1962-3 while sections of the south-eastern defences enabled further radiocarbon dates to be obtained (Jenkins 1963; Blockley and Blockley 1989).

During 2008-11 the writer conducted walk-over surveys of specific compartments within the South Blean reserves of the Kent Wildlife Trust and observed fence construction within the hillfort defences (Sparey-Green 2010a and 2012).¹¹ This was followed up by a more extensive landscape history project, studying the later history of the woodland, its ownership and management (Bannister 2013). Further survey of the hillfort and excavation within its environs identified early prehistoric occupation on the eastern hillside below the camp.¹² The Homestall Wood earthworks have since been the subject of survey and limited excavation during 2014-7, the results of this fieldwork summarised here.¹³

Bigbury Camp hillfort: defences and finds

Only major features of the hillfort will be highlighted here with a re-assessment of the main defences and the context of the metalwork and other finds recovered from the site in the late nineteenth and early twentieth centuries. The earthworks comprise three main elements, the early cross-ridge dyke, the main contour-following en-closure on the spur overlying this earlier boundary and the annex descending to lower ground on the north, close to a stream originating in Hunstead Wood (Fig. 2). The cross-ridge dyke has been confirmed by geophysical survey on the north of the spur; it remains to be traced south of the Bigbury Road and the Pilgrims’ Way. It can be dated to the second or third centuries bc on pottery from the ditch fill (Thompson 1983, 246). The inner hillfort defences encompass 10.7ha of the ridge and spur and follow an almost figure of eight outline. These defences have been considerably damaged, presumably by quarrying and agricultural activity, obscuring much of the circuit on the south and south-east where the deep hollow-ways of the Pilgrims’ Way occupy the site of what had probably been the east gate. Survival is best on the north side where the inner defences were sectioned at five points in the central re-entrant and on the north-east spur, firstly by Jessup and Cook and later by Thompson. Here the defences look north towards Homestall Wood (Plate I). The damaged south-western side, south of the putative west gate, was also investigated at four points by Jessup and Cook and later by the Blockleys. The defence varied between bivallate on either side of the putative eastern and western gates and univallate along the steep slope to the north. The northern annex of 3.3ha is univallate but with a substantial counterscarp on part of the circuit. The south and south-eastern margins are now largely destroyed but may have been univallate, perhaps pierced by another entrance; recently identified outer works are described below.

Where the hilltop defences are best preserved on the north side of the circuit the bank of sand and gravel appeared to be 3.5m wide and 0.8m high but was fronted only by a steep scarp (Thompson 1983, 242-5). To the north-east, in Jessup and Cook’s cuttings 1, 2 and 8, a spread of gravel from the rampart sealed a shallow scoop filled with occupation and burning debris, the natural scarp at its front cut in a marked step and fronted by a rounded V-cut ditch 3.6m wide and 1.5m deep filled with gravel and sand (Jessup and Cook 1936, 156-8). Cutting 12 also produced burnt debris of a similar character at the rear of the bank (op. cit. 160-162). Further north-east in cutting 14 the spread gravel bank sealed a mass of burnt debris 0.7m deep (op. cit., 162-165).¹⁴ Pit 35 of the 2010 fencing operation, close by, encountered a similar destruction deposit apparently sealed by the rampart, finds here including a copper alloy and iron harness ring (Plate II), pottery and charcoal which provided a radio-carbon date in the late fourth or early third century bc.¹⁵

Of the four sections dug on the damaged south-west defences, Jessup’s cutting 6 and the Blockleys’ trench XII revealed an inner ditch 3.5m wide and 1.4m deep of asymmetric profile with steeper outer face, as in a ‘Punic’ ditch of Roman type (Jessup and Cook 1936, 158; Blockley and Blockley 1989, 244). In the latter case a possible palisade trench was identified on the inner edge, recalling the marked step cut in the front of the north-east defences. The bank survived in Blockleys’ trench I as a 4m wide band of clean gravel, the front overlapped by burnt deposits in turn sealed by loam and pebbles (op. cit., 241). This site produced pottery similar to that from the earlier excavations and datable to the fourth to first centuries bc (Thompson 1989, 248).

The entrances have not been explored but a western access onto the ridge and the ancient route south-west is now destroyed or inaccessible beneath the Bigbury Road. A corresponding access to the east probably lay in an area now heavily intersected by hollow-ways on the line of the Pilgrims’ Way. The section cut by Frank Jenkins in the grounds of Woodside, immediately to the north, identified a ‘glacis’-style bank, the face of which yielded clay sling bullets, the ditch containing settings for timber posts, suggesting some defensive structure perhaps associated with that entrance.¹⁶ On the north side a re-entrant in the main defence is likely to be the site of a gateway giving access to the interior from the annex (Fig. 3). Below this, one of the two gaps in the lower northern earthwork served as an exit to the lower ground and the River Cranbourne, the eastern breach favoured in recent survey work, although the western is preferable as occupying a shallow re-entrant in the defences. The levelling of the long southern side may well have erased an entrance giving onto the postulated annex and an outer defence towards the Stour, now obscured by the Parish Road.

The interior is known from the casual finds in gravel digging and the limited excavations within the northern and south-western defences, the following summarising the settlement evidence before re-appraising the metalwork finds. The 1978-80 excavations revealed a water hole and traces of one or possibly two circular huts in the lee of the northern defences, these features producing a ploughshare and copper alloy harness link associated with burnt debris (Thompson 1983, 246-51). Investigation of the annex produced traces of a burnt timber structure and an ironworking site with an anvil remaining in situ (op. cit., 251) suggesting this may have served as a work area, animal corral with water source and trading area with controlled access to the main hilltop settlement.

The finds-groups from the interior

The exact context of the finds groups from quarries within the southern defences is uncertain but at least three groups of material were recovered between 1861 and 1895.¹⁷ Of the 1861 finds one group was buried in a cut at least 7ft (2.1m) deep by at least 12ft (3.7m) in extent and filled with ‘deep gravel’, the floor of the pit covered in approximately 1-1.5ins (40mm) of black soil, interpreted as ‘a layer of turf which had become decomposed’.¹⁸ Further finds in 1866 ‘at the same locality’ comprised a mass of iron, one copper alloy buckle, and fragments of a ‘Roman cinerary urn’, ‘the whole ... deposited about 3½ft (1.2m) below the surface and upon a layer of burnt wood or earth’.¹⁹ The Maidstone Museum collection holds other objects including two shackles, three ‘vase-headed’ iron and bronze lynch-pins and one ring-headed pin besides other delicate bronze and iron fittings which recall items recovered from burials of this period.²⁰ The finds from 1895 were recorded as from an area approximately 25m to the north of the original find spot, this collection comprising mostly agricultural iron work, often pairs of implements and, notably, shackles and a complete slave chain.²¹ At a later date, the find of a single shackle with barb-spring padlock is notable as of Roman type and, if part of the main finds group, might question the origin of this equipment (Jessup 1938, 175-6; Manning 1972, 230). In the Canterbury Museum are also a spear head and socketed adze, while pieces of iron plate with copper alloy plate coating on the convex face, are perhaps fragments of a helmet. A small bronze nail cleaner, similar to others found in IA burials, should also be noted. A whetstone of fine-grained iron-stone in the form of an animal head is likely to be a touch stone for the testing of gold.²²

Outer works of the hillfort

On the east side of Bigbury, the line of the Parish Road/ Tonford Lane may obscure outworks which extended both north-east to Faulkner’s Lane and south-west into Bigbury Wood where they survive as a double or triple bank and ditch approximately following the 35m contour (labelled (I) on Figs 1 and 2). On the next bluff to the south-west, below Howfield Wood Farm, the LiDAR shows a curving triple earthwork (II) following the 50m contour, this interrupted by a possible entrance 50m wide defined by banks. A single bank linked to the northern terminal descends to the Victoria to Dover railway line, reappearing below it as a prominent terrace (III) following the 25m contour for 500m south-westward to Howfield Lane. The northern section of this triple work and the descending linear are followed by the parish boundary between Chartham and Harbledown and Rough Common, this boundary continuing to the floodplain. Here it follows the southern side of an old oxbow in the Stour, within the curve of which air photographs indicate growth marks of features now lost in quarrying.²³ The terrace is traceable further south to Langdane Wood, before possibly turning to follow the steep north-eastern side of the combe to The Rough and New Town Street (IV).²⁴ Close to the river an isolated section of double bank (V) on the 15m contour survives in pasture between the A28 and the Ashford to Canterbury West railway. These earthworks are un-dated but the sections coinciding with the parish boundary suggest an early origin for at least the triple earthwork and an origin as outworks of the hillfort.

On the high ground beyond the hillfort, a track heading west to No Man’s Orchard adjoins a linear earthwork (VI) which heads to a prominent bluff before descending to the stream north of Chartham Hatch. A section across this for drainage purposes at TR 1128 5749 revealed a dump construction over traces of a de-humified old ground surface, suggestive of a work of some age. The Pilgrims’ Way follows the ridge, earthworks here erased by farming activity and the extensive quarrying. Three early Roman pottery jars and the separate find of a glass jug are likely to be grave-goods and evidence for later settlement here.²⁵

Earthworks in the South Blean

Earthworks in the woodlands beyond the hillfort and its associated outer works can be traced over 7 sq. km of the high ground west of Chartham Hatch, through Hunstead, Denstead and Joan Beech Woods (Fig. 1).²⁶ The primary features are substantial banks and ditches which precede other works recognisable as wood banks or hollow-ways linked to the exploitation of the woodland. The first and most extensive earthworks are a double bank and ditch (VII) following the sinuous ridge for approximately 1,500m between the 100m points in Fright Wood and Nickle Wood and the 105m spur above Bower Wood.²⁷ This substantial bivallate earthwork, 15m wide overall, with ditches to the south, is interrupted by a possible entrance at the highest point of Fright Wood. At one point it is crossed by an embanked trackway, this clearly stratigraphically later and similar to the medieval Radfall further north in the Blean. This track approaches from Chartham Hatch and continues through Fright Wood before descending a combe towards the North Downs Way and the Pilgrims’ Way. The southern spur, above Nickle Farm, is surmounted by a slight rectilinear enclosure.

To the north-west, the 100m spur within Denstead Wood is bounded on the south-east by an earthwork (VIII) which extends for 500m between two water courses. Below Primrose Hill this terminates at the source of a stream feeding north-east to the Cranbourne while to the south-west it descends to the stream which feeds the peat bog in Hunstead Wood before draining north-east past Bigbury.²⁸ Near the mid-point this earthwork is a bank 3m high from the much-silted ditch on the south-east, the whole work 18m wide and adjacent to an early iron working site (Plate III).²⁹ Beyond the Hunstead stream this earthwork appears as a slighter boundary adjoining the present track which heads south-west before it crosses the previously mentioned linear (VII) and skirts the 105m spur above Bower Wood. At this point VII ascends the spur and follows the ridge west for at least 500m, petering out near the boundary with Joan Beech Wood. The lower boundary (IX) continues westward on the uphill side of the existing track close to the 80 or 85m contour, the north-south boundary between Denstead and Joan Beech Wood, crossing its line. Beyond this woodland boundary slight earthworks adjoin the track while another follows the 100 and 110m contours.³⁰ The latter (X) may then turn sharply south to skirt the prominent triple-lobed spur south of Joan Beech Wood, dropping down to the 80m contour and continuing on a tortuous route for another 1,000m before rising to the 105m contour at a point above the Selling Railway tunnel.³¹ Here the earthwork bank and ditch (XI) is interrupted by an extensive post-medieval quarry, beyond which a further 600m can be traced close to the 100m contour towards Rhode Common.³² The open pasture and orchards on the hillside below Joan Beech and Bower Wood may preserve traces of terraces from ancient fields and there is also record of a Roman burial from this area. The earthwork enclosure in Perry Wood 2.5 km to the south-west, occupies an outlying block of high ground with commanding views south and west and may be an outpost of this boundary system.

Earthworks west of the ‘Denstead Bowl’ and north of the A2

To the north, numerous features of post-medieval date near the source of the Cranbourne are omitted from this survey. One earthwork (XII), however, starting close to the headwaters in South Bishops Den is notable for its construction in linear sections for 1.5km, extending through Fishpond Wood to the junction of Denstead Lane and the A2 and thence north-east into Church Wood.³³ Of a markedly different character, this is formed of straight sections with angled junctions, the sections approximately 200 or 400m in length, the south-western end at approximately 70m aod dropping to the 60m contour to the north-east. Set on the gently sloping western side of the ‘Denstead Bowl’, where best preserved in woodland adjacent to Denstead Lane, it consisted of a bank 8m wide with a similar width of ditch on the downhill, south-east side. An irregular track or hollow-way crosses it aligned on the junction of the lane and the A2.

To the east, past identification of a barrow near the Roman road line draws attention to two prominent mounds (XIII) visible on LiDAR south of the Cranbourne and immediately north of Denstead Farm.³⁴ Within the southern margin of Church Wood a linear earthwork (XIV) continues the alignment of that south of the road, the angled north-eastern return possibly abutting the uphill side of a rectilinear enclosure, now partly lost.³⁵ Upslope, in Manson Wood and the eastern edge of Church Wood, a complex bank and ditch system (XV) extending for a distance of 600m on the 95m contour is cut by the New Road track. To the west, the double bank and ditch appears to be reinforced with a third, short returns at either end overlying it. A gap to the east may be an original entrance. East of the New Road there is again a triple profile, the slight bank on the south 4.5m wide fronted by traces of a ditch, the middle bank 5m wide fronted by a 2m wide ditch. The third and northern bank is 4m wide fronted by a 2m wide ditch, separated by a 5m wide berm on which quantities of iron slag indicate an earlier smelting site. East and south of the sharply-defined triple terminal a series of hollows or water-filled pits may well be the source of iron-stone nodules in the River Terrace gravels.³⁶ To the east, a section of earthwork (XVI) on the southern edge of Church Wood may be a remnant of a more extensive system, partly erased by activity along the woodland margin north of the A2.

Earthworks in Homestall and Willows Wood

The two major earthworks, XVII and XVIII, within these woods are separated by a major linear boundary XIX (Figs 4 and ⁵).³⁷ Earthwork 1 in Homestall Wood (XVIII) is the largest element in the complex, the 35ha polygonal enclosure occupying a spur at between 80 and 60m aod and bounded to north and east by a stream linking to the Cranbourne.³⁸ Defined by a single bank and external ditch, the bank is 8m wide, the ditch of similar width. The circuit of 2,300m appears regular in profile, laid out in ten sections each varying between approximately 100 to 450m, with no surface indications of internal quarry pits. Seven interruptions to the bank suggest original entrances, in five cases without sign of causeways in the external ditch, suggesting access by wooden bridges or the later removal of any solid causeway. Excavation at Site A on the north-west side revealed a bank of alternate layers of clay and gravel, without obvious trace of revetment and sterile of finds or debris. Traces of two ovens or fire pits to the rear of the bank produced only charred plant remains, unfortunately not suitable for dating purposes. A very limited investigation of the ditch revealed its upper fill was derived from the slighted rampart, this overlying an apparent metalled surface on which lay a sherd of native grog-tempered pottery of first-century bc date. The outer edge was ill-defined, the lower fill un-excavated but the section suggested a double profile.

On the south-west side, at Site E, the defence was masked by the woodland boundary, the silted ditch visible in the pasture beyond, only its inner lip falling within the excavation. A remnant of the bank within the wood sealed an old ground surface and bank make-up containing first-century bc native pottery. The rampart front had been truncated by a compacted surface, the overlying soil producing both native wares and imported amphorae, Terra Rubra and Gaulish White Wares of the late first century bc or early first century ad. A notable find here was an iron, socketed projectile, possibly a ballista bolt.

At the highest point within the interior, on Site B, a rectangular enclosure, the 22ha Earthwork 2, had sides 160 by 140m, with well-defined corners except on the damaged north-east side. Investigation close to an apparent interruption of the eastern side showed it to be defined by a ditch with external bank. Beneath the latter was a metalled surface sterile of finds, sealed by soil in which wheel-ruts could be traced. The bank material above contained a range of pottery of the late first century bc or early first century ad (Plate IV). The interior is largely inaccessible but a depression outside the south-east corner is the source of a stream draining north-east to the encircling stream. Over the whole interior of Earthwork 1 tree-throw holes have produced several groups of pottery of late Iron Age date, including sherds of amphorae of uncertain origin.³⁹ The lack of terra sigillita/samian imports from excavated contexts and surface collections is notable. A geophysical survey of a strip within the western earthwork revealed a complex pattern of features, many on a diagonal alignment, but without further survey these are difficult to interpret.

On the south-east side of Earthwork 1 an external funnel-like entrance, Earthwork 3, linked to an outer earthwork following the 50m contour. Survey of the inner section showed this to be off-set 10m to the north of a gap in the inner rampart, its northern terminal truncated to accommodate it, suggesting Earthwork 3 was an addition (Fig. 6). The outer works were 50m wide overall with external banks 15m wide and 1.5m high with slight internal ditches. These works extend downhill for 80m and increase in scale, a central bank creating a steeply-graded double access which continues into open ground as a much-denuded, splayed outer entrance. Outer works extend on either side, that to the north-east dropping to the encircling stream, the other continuing round to the south-west as a substantial bank for 300m, crossing a re-entrant to gain higher ground to the west and continuing as a boundary to Homestall Wood, before linking with Earthwork 1 on its south side. A hollow-way, much altered by quarrying, cuts this and descends from the wood to the line of Watling Street in Upper Harbledown. Slight earthworks on the open hillside above the road line may be field terraces associated with the village. Later activity within Earthwork 1 is marked by a wood bank running east-west through its centre and continuing westward, also by quarries for gravel and a settlement plot within the eastern interior. The latter adjoined a track from the hollow-way to Rough Common, the existence of a homestead indicated by the wood’s name.

On the next spur to the west, Earthwork 5 (XVII) defined an irregular oval area of approximately 10ha on a spur above the 75m contour. It is best-preserved on the north-west where a double bank and external ditch can be traced, the remainder of the incomplete circuit marked by a single defence. This can be identified on the west, and the north and east sides where it follows a stream dropping towards the Cranbourne. The south, down-slope side may be marked by two lines of single bank, 100m apart. Other than calcined flint in the northern stream bed no finds have been recovered from this site. Between this enclosure and Earthwork 1, a single bank and ditch, Earthwork 4 (XIX), extends from the eastern side of the stream for at least 600m northward before being lost on the high ground. Where best preserved this is a prominent bank 4m wide with a ditch 5m wide on the west. The northern margins of the plateau extending from Dunkirk to Homestall have not been closely surveyed, the only nearby site being the extensive Iron Age settlement at Seasalter, 6km to the north on the edge of Graveney marsh.⁴⁰ North-east of Homestall, excavations on St Thomas’s Hill at St Edmund’s School and Turing College on the University of Kent campus, 1.5km distant, have revealed extensive prehistoric and late Iron Age occupation, hut circles and substantial boundary ditch on sites overlooking the Stour Valley to the south-east. Nearer Homestall, traces of levelled earthworks may be identifiable in the Rough Common area.⁴¹

South of Bigbury, but on the opposite bank of the Stour, the Chartham Downs forms a significant barrier to traffic and may have been enhanced by defensive works.⁴² These overlook the type-site of the Swarling Iron Age cemetery which can now be seen to adjoin an extensive enclosure to the south.⁴³

discussion

In the following, the hillfort defences will first be considered followed by an interpretation of the finds groups from the interior and an overview of the wider earthwork complex, including those within Homestall and Willows Woods, without describing the elements of the complex in a strictly chronological sequence. Detailed survey and the dating of many elements, especially the linear features, is still required and this must remain a preliminary overview of the historical setting.

Bigbury Camp has received most attention and has long been recognized as a native stronghold set astride the ancient route represented by the Pilgrims’ Way. Of its inner defences, the north-west rampart showed major destruction deposits were incorporated beneath the clean gravel of the extant bank, the adjacent ditch sections cut back to a steeper inner profile and free of burnt debris. On the south-west, burnt debris was interleaved between a truncated gravel bank topped by further clean material, the nearby ditch section devoid of burnt debris. In both cases the debris of burnt structures included first century bc pottery, consistent with the date for the interior occupation and all but one of the C-14 determinations from the defences.⁴⁴ As suggested by the Blockleys, the composition of the inner bank questions the circumstances of its construction, since on both the north-west and the south-west it incorporated fresh destruction debris and gravel from a re-cut ditch. Of the five ditch sections, three suggested a stepped or asymmetric V-cut profile of Roman ‘Punic’ type. On the north-eastern side the ‘glacis’ bank, fronted by obstacles set in the ditch, was of a markedly different pattern and may have remained from the original east gate.⁴⁵

The destruction of interior settlement structures and the re-modelling of the defences are seen here as almost contemporary events related to an attack and rapid re-occupation of the site in the mid first century bc. Such a re-profiling of the ditches and incorporation of destruction debris in the rampart would allow the interpretation of these works as part of a larger temporary over-night camp, created by Roman forces in haste.⁴⁶

The collections of finds from the hillfort are derived from two locations, those from the 1860s and 1870s distinct from the 1895 collection.⁴⁷ The 1861 account allows restoration of the original funerary interpretation, this collection the goods accompanying a high-status burial of late Iron Age type. The large pit containing a wide range of gear including horse harness, lynch-pins and iron tyres supports the existence of a chariot within it, the soil layers the decayed organic grave-goods or the rotted timber revetment.⁴⁸ The fire dogs, pot hangers and cauldron equipment would be consistent with a furnished grave, the clay bricks, if from a fire place, suitable accompaniments to cooking equipment. The possible presence of two slave shackles is problematic in view of the later finds.⁴⁹ The date of this potential burial can only be gauged from the style of the metalwork, unless the decorated vessel was an unfamiliar import, such as a Gallic butt-beaker. That would then accord with an interment of the post-Caesarian period.⁵⁰

Few context details are available for the 1895 collection of tools and agricultural equipment but these are here seen as distinct and items lost during the destruction of the hillfort rather than interred as grave goods at a later date.⁵¹ The slave chains and animal hobbles within this group may be equipment from native trading activity but, as noted above, the reported inclusion of two shackles amongst the 1860s finds prevents a definitive interpretation. The identification of a Roman-style shackle amongst later reported finds also questions the date and origin of the equipment. The touchstone and the copper alloy nail cleaner are intrinsically significant, the former used for the testing of coin, the latter an item of toilet equipment associated with burials. The possible fragments of a helmet are now of increased interest in view of the discovery of the Bridge helmet.⁵²

Beyond Bigbury Camp, the finds of complete pottery and glass vessels from the Chartham area are further evidence for burial, perhaps associated with later Roman settlement in the area. The potential barrow sites in the ‘Denstead Bowl’, admittedly only prominent and symmetrical features visible on LiDAR, and another barrow reported near Watling Street, have received no detailed appraisal but would be comparable to the barrow group and other evidence for early high status burials on the site of the city itself (see below, endnote 59).

The system of earthworks identified in the surrounding landscape extends over 17 sq. km north-west of the Stour but also extend east of the valley; outlying crop or parch mark sites and native settlements extend north-east towards the University of Kent campus. Within this complex three groups of defensive features and strong points can be identified, exploiting open ground and utilizing areas of existing woodland, this perhaps the very landscape and defensive system described by Caesar (B. Gall. V, 9, 4-5). The first group comprises Bigbury Camp and its out-works, Monuments I-VI (Fig. 1). The second would comprise the linear defences on the South Blean ridge, Monuments VII-XI. The third the more regular linears and the defensive enclosures on the north, XII-XIX. Site XIII denotes the potential burial mounds in the Denstead Bowl. The first is a series of contour-following enclosure boundaries on the west bank below the hillfort, perhaps annexes close to water but also controlling the use of river crossings. The second consisted of a system of sinuous and sometimes multiple banks and ditches along the scarp of the South Blean, these varying in scale and of more than one phase. This perhaps long-lived system could have controlled access from the south along a predecessor of the Pilgrims’ Way and from the Stour headwaters. On the east bank, the Chartham Downs, reinforced by linear defences along its scarp face, could have complemented this as a ‘stop-line’ on the east bank. While not closing a circuit around Bigbury, these could have been part of a ‘territorial oppidum’ incorporating the hillfort within a defensive system facing south and south-east. This is the native element of the defensive system, its focus, Bigbury Camp, a site destroyed and crudely re-defended as outlined above.

The third group would comprise the regular linear boundaries west and north of the ‘Denstead Bowl’ and the defensive earthworks in Willows and Homestall Woods, these seen here as of a different character. The linear earthworks extending north-east onto the high ground, although much denuded, comprised a south-facing bank and ditch in regular straight alignments and extending for almost 2.5km to Church Wood where a small rectilinear enclosure marked its termination. That Watling Street crosses it at an acute angle suggests the road is later and unrelated, the system perhaps impeding access from the south before its existence.⁵³ The triple earthwork on the scarp above would have duplicated one section of the linear and occupied the site of an earlier bloomery site perhaps within an area of woodland area cleared for fuel.⁵⁴ A further, now much truncated boundary may have continued east along the present woodland margin, such a system defending the western flank of Earthworks 1 and 5.⁵⁵ As already noted, the major 35ha Earthwork 1, one of the largest in Kent, has been recently listed as a presumed successor to Bigbury, an oppidum constructed in the aftermath of the Caesarian campaign. While Bigbury and its outworks can be characterised as a native hillfort or oppidum, Homestall is different and strategically placed to overlook this native stronghold to its south, its regularly-constructed defences and polygonal plan pierced by entrances without solid causeway entrances, its primary defences provisionally dated to the period of Bigbury’s destruction. These works follow a Roman pattern of construction but, as already noted, are too substantial to be the hastily constructed temporary camp on Day 2 of the campaign, here suggested as being improvised by a re-occupation of Bigbury itself.⁵⁶ Homestall 1, however, would correspond with the camp reported as under construction at the start of the renewed inland campaign, following the repair of the fleet and the construction of the expanded coastal base.⁵⁷ Earthwork 5 may either be an outer guard-post during the earlier temporary works at Bigbury or ancillary to the construction of Homestall 1. The intervening linear 4 seems tactically superfluous unless a pre-existing cross-ridge dyke retained as an outer western defence to the latter, on its completion. The re-use and adaptation of Homestall can be more certainly placed in the period between the Caesarian and Claudian campaigns, the central Earthwork 2 datable to the late first century bc or early in the next, this distinctive enclosure comparable to Late Iron Age sacred enclosures recently discovered in Kent and elsewhere in South-East Britain.⁵⁸ The outer entrance, Earthwork 3 is then suggested as a remodelling of the south-east entrance providing an elaborate and formal entrance from the valley to this focal site. As to the proposed high-status burial in Bigbury Camp this would necessarily fall in the aftermath of the failed campaign, preceding the foundation of Canterbury and its own cluster of early funerary monuments.⁵⁹

As noted above, since at least the late sixteenth century attempts have been made to identify sites in the Canterbury area as associated with Caesar’s inland campaign. Dr Plot was possibly the first to recognise Bigbury, but it was not until 1967 that Shepherd Frere recognised both the significance of Bigbury for the Caesarian account and anticipated, plausibly, a Caesarian camp nearby in the Harbledown area. Hugh Thompson was later convinced of Bigbury’s ‘special place in British History’, this echoed by the Blockleys who were tempted ‘to see the burnt deposits in front of the rampart as evidence of a conflagration after Caesar’s attack on Bigberry in 54 B.C.’.⁶⁰ This paper builds on this earlier work using the results of as-yet limited fieldwork over the last decade, aided by new techniques, particularly remote survey with the aid of LiDAR, but also dependent on the writer’s close reading of the ancient literary sources. The sequence of events outlined here is a hypothesis, setting the archaeological sequence as far as that can be reconstructed within the historical record of Caesar’s campaign and, hopefully, advancing the study of Canterbury’s origins.⁶¹

acknowledgements

The study of Bigbury Camp and the South Blean grew out of the author’s involvement in a survey of woodland compartments in the Bigbury to South Blean reserve of the Kent Wildlife Trust, this commissioned of the Canterbury Archaeological Trust (CAT). The author is indebted to these bodies for their support and to the latter body for assistance in preparing this publication.

The author is grateful to Simon Pratt of CAT for processing enhanced print-outs of the LiDAR survey. The staff of Canterbury City Museum kindly provided access to the records and finds from past work at Bigbury Camp. The writer must acknowledge the kind permission of the landowner John Wilson Haffenden and the assistance of the woodland manager, Rick Vallis of Silva Woodland Management, in the conduct of fieldwork here, funded by the late Hugh Toller and other anonymous donors. Neil Morris initially identified the earthworks in Homestall and Willows Woods and has been a mine of information about the sites and their environmental context. The fieldwork was supported by Professor Paul Bennett, the excavation carried out by individual staff members of CAT, supported by Keith Parfitt and members of the Dover Archaeology Group and other volunteers. He is grateful to Dr Andrew Bates for carrying out the geophysical survey in 2017 and to the University of Kent for the loan of equipment. Nick Watts has assiduously checked the woodland for finds from tree-boles. The writer also acknowledges the assistance of Roger Goodburn and Sally Stow for their hospitality and in arranging a meeting with Professor Shepherd Frere to discuss Homestall in 2014.

The author is very grateful to the Roman Research Trust and private donors for funding the preparation of illustrations by Canterbury Archaeological Trust.

bibliography

Allen, T. and Willson, J., 2001, ‘Sunset Caravan Park and Church Lane East, Whitstable’, Canterbury’s Archaeology 1998-1999, 10-11.

Andrews, P., Booth, P., Fitzpatrick, A.P. and Welsh, K., 2015, Digging at the Gateway, Archaeological landscapes of south Thanet, The Archaeology of East Kent Access (Phase II), Volume 1: The Sites; Volume 2: The Finds, Environmental and Dating Reports, Oxford Wessex Archaeology, Monograph No. 8.

Bannister, Nicola R., 2013, South Blean Woods Landscape History Project, Kent Wildlife Trust, Full report 2010-2013, May 2013, available via:- http://www.kentwildlifetrust.org.uk.

Barclay, A.J. and Stevens, C.J., 2015, ‘Radiocarbon Dating’, in Andrews et al. 5, Vol. 2, 569-581.

Bates, A, 2017, ‘Making the invisible visible; New Survey and investigation of the Iron Age Hillforts of Bigbury and Oldbury in Kent’, unpubl. ph.d. thesis, https://kar.kent.ac.uk/66317/ (accessed 18.7.2019).

Bayard, D., 2018, L’occupation des oppida gaulois par l’armée romaine à la fin de la Repub-lique: L’example du “Camp César” de la Chausée-Tirancourt, in Reddé 2018, 157-178.

Black, E.W., 2015, A Survey of Selected Late Iron Age and Roman Sacred Sites in Eastern England, with particular reference to Essex, The Essex Society for Archaeology and History, 6, 106-163.

Blockley, K. and Blockley, P., 1989, Excavations at Bigberry, near Canterbury, 1981, Archaeologia Cantiana, cvii, 239-251.

Booth, P., 2009, ‘Roman Britain in 2008, Southern Counties, Kent’, Britannia, 40, 276.

Booth, P., 2012, ‘Roman Britain in 2011, Southern Counties, Kent’, Britannia, ⁴³, 350-1.

Booth, P., 2014, ‘Roman Britain in 2013, Southern Counties, Kent’, Britannia, ⁴⁵, 393 and fig 26.

Booth, P., 2016, ‘Roman Britain in 2015, Southern Counties, Kent’, Britannia, ⁴⁷, 358-9.

Bowman, S.G.E., Ambers, J.C. and Leese, M.N., 1990, ‘Re-evaluation of British Museum radiocarbon dates issued between 1980 and 1984’, Radiocarbon 32(1), 59-79.

Boyd Dawkins, W., 1902, ‘On Bigbury Camp and the Pilgrims’ Way’, Archaeological Journal, 59, 211-218.

Brent, J., jnr., 1861, ‘Roman cemeteries in Canterbury, with some conjectures concerning its earliest inhabitants’, Archaeologia Cantiana, iv, 27-42.

Bronk Ramsey C., 2014, ‘Bayesian Analysis of Radiocarbon Dates’, Radiocarbon 51(1) 337-360. Last updated 5th September 2014. https://c14.arch.ox.ac.uk/oxcal/OxCal.html.

Bushe-Fox, J.P., 1925, Excavation of the Late-Celtic Urn-Field at Swarling Kent, Report of the Research Committee of the Society of Antiquaries of London, V, London.

Canterbury Archaeological Trust, 2015, Canterbury’s Archaeology 2013-2014, annual review of the Canterbury Archaeological Trust.

Clark, A.J. and Thompson, F.H., 1989, ‘Revised radiocarbon dates for three hillforts in Kent and Surrey’, Antiquaries Journal, LXIX, pt II, 303-307.

Copley, G.J. (anntd and ed.), 1977, Camden’s Britannia, Kent, edition of 1789, ed. R. Gough, Hutchinson, London.

Cunliffe, B.W., 1972, ‘The Late Iron Age metalwork from Bulbury, Dorset’, Antiquaries Journal, LII, 293-308.

Farley, J., Parfitt, K. and Richardson A., 2014, ‘A Late Iron Age Helmet Burial from Bridge, near Canterbury, Kent’, PPS, 80, 379-388.

Fitzpatrick, A.P., 2019, ‘Caesar’s landing sites in Britain and Gaul in 55 and 54 BC: Critical places, natural places’, in Fitzpatrick and Haselgrove, 135-158.

Fitzpatrick, A.P. and Haselgrove, C., 2019, Julius Caesar’s Battle for Gaul, New Archaeological Perspectives, Oxbow, Oxford and Philadelphia.

Flinders Petrie, W.M., 1880, ‘Notes on Kentish Earthworks’, Archaeologia Cantiana, xiii, 8-17.

Frere, S.S., 1999, Britannia, Folio Society, London.

Goodburn, R., 2015, ‘A Personal Memoir’, in ‘Sheppard Sunderland Frere, Historian and Archaeologist’, Britannia, 46, 1-5.

Gould, N., 1862, ‘Proceedings of the Association’, Journal of the British Archaeological Association, 18, 272-3.

Hoffman B., 2013, The Roman invasion of Britain, Archaeology versus History, Pen and Sword, Barnsley, S. Yorks.

Hogg, A.H.A., 1975, A Guide to the Hill-Forts of Britain, London.

Holmes, W. and Wheaten, A. (eds), 2002, The Blean: The Woodlands of a Cathedral City, The Blean Research Group.

Hussey, R.C., 1874, ‘The British Settlement in Bigbury Wood, Harbledown’, Archaeologia Cantiana, ix, 13-15.

Jenkins, F.J., 1963, ‘Interim Report by Mr. Frank Jenkins, F.S.A. on Excavations at Bigberry Camp, Harbledown , near Canterbury, 1962-3, Reports, 1963, Appendix 1, Archaeologia Cantiana, lxxvii, xlvii-xlviii.

Jessup, R.F., 1932, ‘Bigberry Camp, Harbledown, Kent’, Archaeological Journal 89, 87-115.

Jessup, R.F., 1938, Objects from Bigberry Canp, Harbledown, Kent, Antiquaries Journal, XVIII, No. 2, 174-176.

Jessup, R.F., 1939, ‘Further excavations at Julliberrie’s grave, Chilham’, Antiquaries Journal, XIX, no. 3, 260-281.

Jessup, R.F. and Cook, N.C., 1936, ‘Excavations at Bigberry Camp, Harbledown’, Archaeo-logia Cantiana, xlviii, 151-168.

Lane, R., 2012, ‘Interim report on archaeological works undertaken on land adjacent to Giles Lane, St Edmund’s School, St Thomas Hill, Canterbury, Kent’, unpubl. client report, CAT, 2012/39.

Lane, R., 2014, ‘An Archaeological excavation at Turing College, University of Kent, Canterbury, Kent: interim report’, unpubl. client report, CAT, 2013/176.

Legendre, R.-M. and Piechaud, S., 1985, ‘’Une sepelture á outils du debut de La Tène á la Chaussee-sur Marne (Marne)’, Prehistoire et Protohistoire en Champagne-Ardenne, 9, 57-66. 

Lock, G. and Ralston, I., 2017, Atlas of Hillforts of Britain and Ireland, available at https://hillforts.arch.ox.ac.uk.

Mackinder, T., 2005, ‘Furfield Quarry Finds’, KAS Newsletter, 66, Autumn 2005, 14-15.

Manning, W.H., 1972, ‘Iron work Hoards in Iron Age and Roman Britain’, Britannia, 3, 224-250.

Margary, I.D., 1973, Roman Roads in Britain (3rd edn), London.

Morris, N., n.d., ‘Hunstead Wood. Its history explored through physical features of the wood’, unpubl. mss.

Morris, N., 2010, ‘Homestall Wood, Stock Wood and Willows Wood at Harbledown: Earthworks recorded’, unpubl. mss.

Niblett, R. et al., 1999, The Excavation of a Ceremonial Site at Folly Lane, Verulamium, Britannia Monograph 14, London.

Parfitt, K., 1998, ‘A late Iron Age burial from Chilham Castle, near Canterbury, Kent’, PPS, 64, 343-351.

Payne, 1893, Collectanea Cantiana.

Philp, B.J., 1960, A Romano-British Villa site at Swarling, Archaeologia Cantiana, lxxiv, 74, 186-190.

Pilbrow, J., 1871, ‘Discoveries during excavations at Canterbury’, Archaeologia, XLIII, 151-164, Society of Antiquaries of London.

Poole, C., 2015, ‘Fired Clay and Briquetage’, in Andrews et al., 289-323 at 304.

Rady, J., 2010, Thanet Earth, Monkton, Canterbury’s Archaeology 2008-2009, 33rd annual report of the CAT, 1-16.

Rawlinson, R. (ed.), 1714, ‘A copy of a letter from Robert Plot, Ll.D., Keeper of the Ashmolean Museum in the University of Oxford: design’d to be sent to the Royal Society in London’, Miscellanies on several curious subjects: Now first Published from their RESPECTIVE ORIGINALS, 44, London.

Reddé, M., 2018, L’armée romaine en Gaule a l’époque républicaine. Nouveaux témoignages archéologiques, Glux-en-Glenne, Bibracte, 28.

Reddé, M., 2019, ‘Recent archaeological research on Roman military engineering works of the Gallic War’, in Fitzpatrick and Haselgrove, 91-112.

Rice Holmes, T., 1907, Ancient Britain and the Invasions of Julius Caesar, Oxford.

Roberts, G., 1999, Woodlands of Kent, Ashford.

Smith, C.R., 1868, Collectanea Antiqua, Etchings and Notices of Ancient Remains, VI, 262.

Sparey-Green, C., 2010a, ‘Preliminary survey of earthworks in the vicinity of Bigbury Camp and the Blean Woods’, Canterbury’s Archaeology 2008-2009, 33rd annual report of the CAT, 32-35.

Sparey-Green, C., 2010b, ‘Homestall Wood Earthworks, Harbledown, Kent’, KAS Newsletter, 86, Winter 2010, 14-15.

Sparey-Green, C., 2012, ‘Bigbury Camp, Harbledown, Canterbury’, Canterbury’s Arch-aeology 2010-2011, 35th annual report of the CAT, 27-28.

Sparey-Green, C., 2016, ‘Archaeological Evaluation of Earthworks in Homestall Wood, Harbledown, Kent, Interim report on 2015 season and proposal for 2016’, unpubl. client report.

Stead, I., 1967, A La Tène III Burial at Welwyn Garden City, Archaeologia, CI, 1-62.

Thompson, F.H., 1983, ‘Excavations at Bigberry, near Canterbury, 1978-80’, Antiquaries Journal, LXIII, 237-277.

Thompson, F.H., 1993, ‘Iron Age and Roman Slave-Shackles’, Archaeological Journal, 150, 57-168.

Thompson, I., 1989, ‘The Iron Age Pottery’, in Blockley and Blockley, 246-250.

Vine, F.T., 1887, Caesar in Kent, An Account of the Landing of Julius Caesar and his Battles with the Ancient Britons; with some Account of early British Trade and Enterprise, London.

Watson, S., 2017, ‘Roman Britain in 2016, Southern Counties (East), Kent’, Britannia, 48, 424-5.

Wallenberg, J.K., 1934, The Place-Names of Kent, Uppsala; Lundequistska Bokhandein.

Wheaten, A., 2016, Short Wood, otherwise known as Church Wood, KAS Newsletter 104, 40-45.

Wheeler, R.E.M., 1943, Maiden Castle Dorset, Reports of the Research Committee of the Society of Antiquaries, London, XII.

Woolley, C.L., 1925, ‘The Iron-smelting Works’, in Bushe-Fox, 49-53.

endnotes

¹ Work previous to 2008 is referenced elsewhere in this paper. More recent study of Bigbury and Homestall Wood is covered by Morris n.d. and 2010; Sparey-Green 2010a, 2010b and 2012; the LiDAR survey covering both sites was conducted by Peter Crow of the Forest Research Unit for Landscape Modelling and the Blean Partnership http://www.the blean.co.uk/blean-lidar-project/lidar-map/accessed 1.7.2020. Only some images appear accessible. In Fig. 1 sites other than Bigbury Camp are indicated as Monuments I-XIX, this designation specific to this paper and including Earthworks 1-5 in Homestall and Willows Woods as XVII-XIX. While the majority of sites are shown the Chartham Downs to the south-east and Rough Common or St Edmund’s Hill area to the north-east are beyond its limits.

² This research, initiated in a private capacity, has been supported by Canterbury Archaeological Trust (CAT) and the University of Kent.

³ Sections of the South Blean west of Chartham Hatch and Bigbury Camp are Kent Wildlife Trust reserves. The A2 follows Roman Watling Street, Road 1b from Canterbury to Rochester (Canterbury UAD MKE 4004; NMR TQ 96 SE 44; Margary 1973, 43). The earthworks in Homestall, Willows and Church Woods lie within woodland managed by Silva Woodland Management Ltd but also extend west into Manson Wood, part of the RSPB reserves.

⁴ From this high point the Thames Valley, Isle of Sheppey and even the hills of Essex near Billericay, 55km to the north-west, are visible.

⁵ A charter of ad 785/791 records the grant of Blean woodland by Offa, King of Mercia to Ealdbeorht and his sister Selethryth, https://esawyer.lib.cam.ac.uk/ viewed 4.11.2019; charters 123, 125 and 1614.

⁶ Camden’s Britannia, anntd and edt. Copley, 1977, 39 and 42. The derivation of the place name from Jul-ham is seen by Copley as a fictitious ‘lead-in’ to the association with Caesar. Excavation of the Neolithic long barrow showed that burials of late Iron Age or Early Roman date had been inserted in the east end (Jessup 1939, 264-8). More recent finds from the area south of the Castle include the mirror burial dated to the mid first century bc (Parfitt 1998).

⁷ The author is grateful to Colin Flight for providing this reference.

⁸ The term -bury denoting a defended earthwork is derived from OE burh but often appears in the form borough, berry or perry, the latter seen in the name Perry Wood in Selling to the south-west.

⁹ First edition OS 25 inch map; Hussey 1874, subsequently listed in Flinders Petrie 1880, 9; Rice Holmes 1907, 256-7, 337 and 685. For the description of the woodland strongholds see Caes. B. Gall. V, 9, 4-6; 21, 3.

¹⁰ Vine 1887, 165-176 and 178-198. Vine refers to ‘the strongly defended oppidum at Durovernum’ (op. cit. 216-7) but makes no specific mention of Bigbury Camp. His favoured site on the Barham Downs outside Bridge is now seen as a palimpsest of prehistoric to post-medieval earthworks.

¹¹ In 2008 Neil Morris and Tony Nuthall surveyed the known earthworks and other outworks. For the work by CAT in 2008-9, see endnote 1 above. The cross-ridge dyke was confirmed in a geophysical survey of the western interior of the hillfort undertaken by Lloyd Bosworth of the University of Kent. Survey of the northern annex was undertaken in 2016 by Adrian Oliver as part of Leicester University’s ‘Footsteps of Caesar project’.

¹² Andrew Bates 2017, chapters 2, 5-7. His fieldwork showed lengthy prehistoric occupation of the hillside below the camp and identified the extent of recent activity which has truncated the eastern defences.

¹³ Other than references in endnote 1, investigations to date are summarized in Booth 2014 and 2016 and Watson 2017. Bigbury Camp is site EN 3771 in the Oxford University Hillforts Atlas (Lock and Ralston 2017). Surviving sections of the defences and the interior of Bigbury Camp are listed as a Scheduled Ancient Monument Kent 51, the whole site entered as HER Number TR 15 NW 33.

¹⁴ The initial Cuttings 1 and 2 in 1933 show an ‘occupation layer’ sealed by gravel and truncated by the stepped, inner edge of the ditch. In 1934 the adjacent cutting 8 identified a more pronounced burnt layer uphill of the ditch, this described as ‘more than the remains of an ordinary domestic fire’ and sealing no structural traces other than a hearth. In cutting 12 the mix of pottery, charcoal and burnt debris including sherds of a pear-shaped jar scattered throughout the 0.3m deep deposit (Jessup and Cook 1936, 157; 160, fig. V, 11). Significantly, in situ structural remains other than the hearth were absent; these deposits were simply spread destruction debris.

¹⁵ Sparey-Green 2012, 12. The contractor’s pit adjoined the site of Jessup and Cook’s trench 14. The harness ring would have parallels in the late Iron Age while the pottery from here would be of similar date and consistent with the 1933-4 finds. The single C-14 date was, however, considerably earlier and may represent residual material or old timber.

¹⁶ A fuller report of the 1963 excavations by Frank Jenkins will appear in a future volume of Archaeologia Cantiana.

¹⁷ The records of the finds groups and the accounts of the circumstances of their discovery are difficult to reconcile, the collections in Canterbury, Maidstone and Manchester Museums neither closed groups nor complete, the former Museum containing additional material from the 1930s and 1950s. Jessup (1932, 90, 95-111) summarised finds to 1932, Thompson producing the first detailed catalogue of the three major groups (1983, 265-275, figs 13-20 and plates 32-39). The initial finds were located at Hussey’s point E (1874, 15) within the quarry behind the southern defences (approximately TR 1168 5747, south of Bigberry Farm). For the 1895 findspot, see endnote 21. The 1902 finds were recovered in 1895.

¹⁸ The first report (Brent 1861, 33) saw the finds as from burials, the list including a share, coulter, cattle goad, ‘the iron tire of plough or chariot wheels’, horse bit and iron links or traces (? pot hangers). He also reports another Roman grave containing iron fire-dogs. Gould (1862, 273) then provides context details of the main find and describes the six triangular clay bricks laid close together ‘in a circle’, elements of a tripod and pot-hangers, the ‘bail-handle’ of a cauldron, a ‘large knife’ (? one of the extant sickles), a ‘powerful snaffle-bit’ and an urn which ‘points to a sepulchral interment’ ... the ‘paste bespeaking a Celtic origin, though the ornamentation is peculiar’. Significantly, Brent again refers to ‘wheel tires’ before describing new finds from 1866 (Smith 1868).

¹⁹ This group comprised sickles, iron rings, part of an iron rod, a ferrule, and a small engraved bronze buckle. The 1861 and 1866 finds are now shared between Canterbury City Museum and Maidstone Museum.

²⁰ Possibly fittings from harness or game boards and similar to objects from Iron Age burials such as at Welwyn Garden City (Stead 1967, 27-38). The origin of these is unknown but the shackles are similar to those in Manchester (see endnote 21). It is possible that these, and perhaps the firedog, were amongst the thirteen iron objects from ‘Harbledown’ donated by R. Howard White to Maidstone Museum in 1870 and part of the 1861 find (Jessup 1932, 98). Further iron work and pottery finds were donated in 1887 but details are lacking.

²¹ Boyd Dawkins 1902, fig. 1, at point A, approximately TR 1169 5749, and not close to Bigbury Cottages as on later OS maps. This collection, found in or before 1895, and held in Manchester City Museum, comprised two spear heads and a short sword or dagger, iron and wood-working tools, two billhooks and two sickles, rusted together, a coulter and two plough shares, horse harness including two snaffle bits and bronze plated ring, five pot hooks and chains, two slave shackles and a slave chain (op. cit. 213-216). Amongst the ironwork Thompson recognised two animal hobbles (Thompson 1993, 136-8, ill. 108-110). Reference is also made to ‘brown pottery’ and ‘flat bottom of a vessel in greyware’ recovered from the spoil in 1896.

²² The spearhead (Cat. No. 1153) and socketed adze (271b) are unpublished additions to the collection. The form of the copper-alloy coated iron fragments suggest they were fragments of a helmet rather than a vessel (Canterbury City Museum 1950/120). Nail cleaners have been found in late Iron Age burials at Welwyn Garden City and the ‘War Cemetery’ at Maiden Castle, Dorset (Stead 1967, 27, fig. 15; Wheeler 1943, fig. 92, 7). A toilet-set is associated with a burial from Deal (Bushe-Fox 1925, 18-19, Pl IV, Fig 3). The author is grateful to the late Andrew Savage for the suggestion that the whetstone of fine-grained iron-stone is similar to the material of a medieval touchstone from Canterbury (Jessup 1938, 175-6; City Mus. 1952, 120.1/8234). Pottery sherds of Late Iron Age type include one base-ring (registered as 1952.120. 2) and others marked 6927. The collections and records in Canterbury, Maidstone and Manchester deserve further research.

²³ HER TR 15 NW 328. An early prehistoric site close on the south (TR 15 NW 641) produced important environmental material.

²⁴ Bates (op. cit., 198-199, fig. 23) draws parallels with field lynchets but out-turns either side of the gap in the triple earthwork and the dog-leg in the parish boundary suggest an entrance out-work. A profile and enhanced LiDAR print, kindly provided by Simon Pratt of CAT, shows the earthwork below the railway is 30m wide with a much denuded but substantial uphill bank.

²⁵ Jessup 1932, 99-101, 114-5; Payne 1893, 130. Few details of the origin of this material are available but burials are recorded from Nickhill Farm and Hatch Green, south-west of Chartham Hatch (HER TR 05 NE1-2).

²⁶ The earthworks were initially identified during survey of specific compartments within the Kent Wildlife Trust reserve and later confirmed by LiDAR survey (Sparey-Green 2010a; Booth 2009 and Booth 2012). Reference to the catalogue accompanying Bannister 2013 is included here but omitting most features associated with woodland and re-interpreting some entries. The complex is also included in Bates (2017, Chapter 7).

²⁷ The first linears would correspond with, from east to west, Bannister Features 8 and possibly 100, 4 and 10, 5 and 6. In Nickle Wood the earthworks are interrupted by a gravel pit of some age.

²⁸ The main bank and ditch, Bannister earthwork 18 and 92. Iron working debris has been recorded from the low hill immediately south-east of the main earthwork (MKE 3986).

²⁹ TR05 NE 7, the ironworking debris of Iron Age to early medieval type includes cinder and burnt clay from a bloomery site.

³⁰ Bannister 14 but also traceable on the uphill side of the existing track. Close to the 100m contour above, 176 and 177 may represent sections of a parallel contour-following linear. The LiDAR also shows a substantial earthwork on the east side of the Bower Wood spur, this extending for over 400m to the 100m contour. The spur is flanked by ancient stream lines on either side.

³¹ Bannister 30-33 and 37. The quarry is earthwork 40 and is occupied by a veteran oak.

³² Bannister 24-26 and 29 which may continue along the northern side of the grounds of Woodlands Cottage. Traces of two rectilinear enclosures may be recognized on the high ground to the north. The linear earthwork here may be masked by re-use as footpaths or hollow-ways and quarrying visible on the LiDAR to either side of it.

³³ Bannister 97 and 147. The central section has been levelled by agriculture but can be traced in grassland and on some air photographs. A sinuous and irregular feature extending east-west across the stream line and heading up slope westward, Bannister 72, 88-90, 93, is visible on Google Earth continuing towards Winterbourne Wood and may be a hollow-way linking with Boughton under Blean.

³⁴ The line of Watling Street, Margary route 1b, is likely to have had some antecedent westward from the high ground. The mounds or hillocks lie above the 65m contour north of Denstead Farm at TQ 1002 5786 and TQ 1008 5780. The LiDAR suggests other mounds may exist to the west, on the south side of the Cranbourne. See endnote 53.

³⁵ Bannister 149 and 150.

³⁶ Bannister 143; Booth 2014, 393. The ironworking site and triple earthwork in Manson and the west end of Church Woods is recorded in the Canterbury UAD MKE 3987, NMR TR 05 NE 8.

³⁷ Monument XVII is Earthwork 5, XVIII comprises the linked Earthworks 1-3 and XIX the intervening linear 4, to conform with the order here. For details of this complex see Figs 4 and 5.

³⁸ Holmes and Wheaten (eds, 2002, 126) had originally suggested a geological origin but the LiDAR survey confirmed Neil Morris’s (2010) identification of the polygonal enclosure. Documentary sources do not indicate the earthworks as those of a medieval deer park or as boundaries within the Blean woods (Wheaten 2016). Initial results were summarised as Sparey-Green 2010a and 2010b; Booth 2014 and 2016; Watson 2017). The site is not scheduled but has double entries in the HER as TR 15 NW 1599 and 2399, referencing the LiDAR survey in 2010 and an unpublished document by Neil Aldridge. Earthwork 1 has been classed as a hillfort and listed as site EN 3823 in the Oxford Hillforts Atlas (Lock and Ralston 2017).

³⁹ Nick Watts and the writer have assiduously checked ground disturbances, resulting in both the recovery of pottery finds and the observation of structural details from the earthworks.

⁴⁰ The Seasalter site is now classed as a potential oppidum in the Hillforts Atlas (Lock and Ralston 2017, EN 3997). The marshes may have been partly open water, finds of briquetage suggesting it was a centre of salt winning (Allen and Willson 2001).

⁴¹ Lane 2012 and 2014. Amongst the most significant finds at Turing College were fragments of Dressel 1 amphorae of the mid first century bc. Google Earth images from 2003 and 2011 suggest the presence of boundaries and enclosures on playing fields of Canterbury College and possible field terraces on the valley-side to the south.

⁴² Google Earth images confirm the steep south-facing scarp is occupied by two sinuous linear features, impracticable as trackways and presumably earthwork barriers which extend from Shalmsford Street east to Perry Hill (HER TR 15 SW 45; 47 and 49). Several rectilinear enclosures lie on the high ground including an Iron Age/early Roman enclosure below Iffin Wood (Philp 1960).

⁴³ Crop marks of rectilinear enclosures centred at TR 128 524, visible on Google Earth 4/21/2007. For Swarling see Bushe-Fox 1925. The traces of ironworking are exceptional for the apparent mine in the iron-bearing gravels; the burnt floor described by Woolley may have been used for ore-roasting.

⁴⁴ Bates 2017, 47-53 quotes the original Thompson determinations (BM 1530 and BM 1768), now revised by Barclay and Stevens 2015, 579. These samples, from the interior settlement, are set at 2080 ±45 bp and 2060±50, 175 to 85 cal bc and 160-60 cal bc, respectively. The archaeomagnetic date for the same context was revised to 300-90 cal bc (Clark and Thompson 1989). Sample (HAR-5030) from the south-west defences, at 1930± 70 bp provides a date of 100 cal bc-cal ad 250 (Blockley and Blockley 1989, 250; revised by Bronk Ramsey 2014). A single sample of wood charcoal from below the northern rampart (UBA-18135) is considerably earlier, with a date of 2245 ±26 bp, 390 -348 cal bc, this single sample recovered during contractors’ work (Sparey-Green 2012, 33). All at 95.4% confidence.

⁴⁵ The clay sling bullets from the Jenkins’ 1962 excavation are paralleled at other British hillforts, their use as incendiary weapons by the Gauls described by Caesar in his account of the attack on Quintus Cicero’s camp in 54 bc (B. Gall. V, 43, 1). As with other weapons of the period their use does not need to be limited to one side, the scatter of projectiles representing the attacker’s ‘expended munitions’.

⁴⁶ Caesar makes specific reference to the urgent need to create an overnight camp (B. Gall. V, 9, 7). The hilltop alone would not have accommodated four legions and the cavalry, but its occupation would have denied the site to the enemy, the annexes giving additional space and access to water. The scale of Earthwork 1 suggests this could not have been constructed in the time scale but it is possible that 5 might have been an adjunct to these works. This is discussed further below.

⁴⁷ The two groups of finds from the 1860s, now shared between Canterbury and Maidstone, and the 1895 collection in Manchester, are one of the most important collections of late IA artefacts in the country, complementing the evidence for ironworking in the area, activity itself worthy of detailed research.

⁴⁸ The pit would be paralleled by other Late Iron Age burials, such as the chamber at Folly Lane, St Albans (Stead 1967, 44, fig 28; Niblett 1999, 30-45). A funerary context was most recently suggested by Hogg 1975, 133, but dismissed by Thompson (1983, 252) who saw all the equipment as lost in the destruction of the hillfort (op. cit., 256). With the interpretation offered here both may have been right. The soil descriptions given by the gravel-diggers are likely accurate since they would seek clean aggregate and thus be discountenanced by archaeological deposits, organic deposits or calcined bones.

⁴⁹ Such bricks have often been seen as loom weights, as at Thanet Earth, Monkton (Rady 2010, 7), but, at Ebbsfleet, they could serve as ‘heat sinks’, oven floors or supports for salt-boiling vessels (Poole, 2015, 304). The Bigbury bricks were described as clustered in a similar manner to those at Ebbsfleet. A slave chain and fire dog from near a barrow at Lord’s Bridge in Cambridgeshire could show that such equipment was placed in native burials, but any symbolism is obscure (Thompson 1993, 60, ill. 1).

⁵⁰ Finds from the interior of Bulbury Camp, Dorset, can be cited as possible gravegoods and a metal-working hoard (Cunliffe 1972). The copper alloy fittings, fragments of a mirror, glass beads, a sword, firedogs, and copper alloy bowls were grave goods, the tools, including a lump hammer, axe and a ship’s anchor and chain, as a blacksmith’s hoard. A late Iron Age burial at Chaussée-sur-Marne, near Rheims in north-east France, has also produced a set of iron woodworking tools as grave goods (Legendre et Piechaud 1985).

⁵¹ Thompson encountered lost equipment within destruction deposits elsewhere on the site (1983, 250-1).

⁵² Thompson’s 1993 study of slave-shackles shows that such equipment was in common use by Romans and Britons. In Britain, shackles were also recorded in the case of Caesar’s envoy Commius who, arrested by the Britons on landing in 55 bc, was brought to him in chains (B. Gall. IV, 27, 3). Caesar records the taking of British obsides during both British campaigns (B. Gall. IV, 27, 5; 38, 4; V, 20, 3; 22, 4) but these hostages are unlikely to have been so restrained. Later in the century, slaves, presumably of native origin, are recorded as a British export (Strabo IV, 5, 1-2). The Bridge helmet, dated to the mid first century bc was entirely of copper-alloy (Farley et al. 2014).

⁵³ See notes 34 and 35. A post-Roman barrier, such as the Faesten Dic in open country in west Kent, would here be set across Watling Street at right angles linking the high ground.

⁵⁴ Notes 29, 36 and 43 above. The earthwork post-dates the metal working since the bank incorporates bloomery debris. The juxtaposition of earthwork and bloomery site would be arch-aeologically significant for C-14 dating of the sequence here. Caesar’s observation of the limited iron resources in the coastal region may even be a reference to an industry dependent on the iron-stone within the gravels, as here (B. Gall. V, 12, 4).

⁵⁵ Such a linear defence (bracchium) could form a corollary to Republican camps of polygonal or oval plan and sometimes linking with outposts (Reddé 2019, 93-5, figs 6.1-6.3). Homestall Earthwork 1 would be compatible with such defences, the linear perhaps a ‘stop line’ set up against chariot action (B. Gall. V, 15).

⁵⁶ The use of native defences by Roman forces is exemplified in Northern Gaul, most notably at Chausée-Tirancourt where geophysical survey and excavation has identified the tent-lines from a republican army encampment within the oppidum (Bayard 2018).

⁵⁷ From its first edition in 1967 Shepherd Frere’s Britannia suggested that Caesar’s first inland camp was ‘still to be located. It must occupy some 150 acres (60.7 ha), perhaps at Harbledown’ (Frere 1999, 23). See also now Goodburn 2015, 4-5. Frere was right to identify the location but, as postulated here, Homestall 1 is too substantial to be the temporary camp but could be the works referred to as under construction during the two-day battle on day 13 and 14 of the campaign (B. Gall. V, 15,3; 16,1; 17,1). The western linear boundaries, the proposed bracchium, would have acted as a defence on the western flank.

⁵⁸ Black 2015. Similar enclosures have recently been found in Kent at Cheeseman’s Green, Ashford and at Furfield Quarry (Mackinder 2005).

⁵⁹ Canterbury’s early origin was most recently exposed by discoveries at the Marlowe Arcade where more of a late Iron Age triple-ditched enclosure was recorded (Canterbury Archaeological Trust 2015, 10-11 and 18). Attention should also be drawn to the exceptional copper alloy krater of Mediterranean origin from Palace Street, inverted over cremated bone and other grave goods, evidence for an early high-status burial subsumed within the later Roman city (Pilbrow 1871, 11-12, Site 63).

⁶⁰ From Thompson 1983, 237 and 254-9; Blockley and Blockley 1989, 241.

⁶¹ As recently as 2013 Birgitta Hoffman (p. 21), stated that no physical trace existed for Caesar’s presence in Britain. The potential form of that evidence, however, is shown by the continental sites (Reddé 2018) and the British situation must now be reviewed in the context of sites such as Ebbsfleet, on the east coast of Kent, proposed as the naval camp for Caesar’s campaign, in competition with the long-standing claims for the traditional site at Deal and Walmer (Fitzpatrick 2019). We are now on the cusp of identifying the traces of the British campaign of 55 and 54 bc.

Fig. 1 General plan of the earthworks in the area of Bigbury Camp, the South Blean and the woodland north of the A2 (Watling Street); the Stour valley to the bottom right.

Fig. 2 Bigbury Camp, the earthworks as surviving showing the recorded metal findspots, the main areas of quarrying and the sites of excavations between 1936-1981.

Plate I View north from Bigbury Camp; the Cranbourne valley and Upper Harbledown in middle ground and Homestall Wood on the skyline (photo by C. Sparey-Green).

Plate II Copper alloy ring with iron core, a (?) baldric or harness fitting from beneath the Bigbury northern inner defence (photo by the late A. Savage © CAT).

Fig. 3 Bigbury Camp, the northern entrance to the Inner Defences, showing the existing earthworks, recent sandpits and the excavation trenches 1978-1981

Plate III View south-west to main earthwork at Denstead Wood, South Blean; the figure is standing in the silted ditch with bank to the right (photo by C. Sparey-Green).

Fig. 4 The LiDAR survey of the Willows and Homestall Woods earthworks showing Earthworks 1-5. LIDAR data Co. Forestry Research based on Unit for Landscape Modelling and the Blean Partnership Data.

Fig. 5 Plan of Earthworks 1-5 in Homestall and Willows Woods, based on LiDAR survey, ground survey and excavations at Sites A, B and E. LIDAR data Co. Forestry Research based on Unit for Landscape Modelling and the Blean Partnership Data.

Plate IV Homestall Wood Earthwork 2 east side, looking north-west, showing (in foreground) the truncated external bank and underlying wheel ruts; in the middle distance the internal ditch is shown by the dip in the fence-line (photo by C. Sparey-Green).

Fig. 6 Homestall Wood, Earthwork 3, outer south-east entrance to Earthwork 1.