Monument record 866/1/3 - Iron Age hillfort at Eddisbury

Eddisbury is a bivallate (double rampart) hillfort situated on an outcrop of the Cheshire Sandstone Ridge overlooking the Weaver Valley to the east. Like many of the hillforts along the Ridge, it was constructed to make best use of the contemporary landscape by reinforcing the scarp on each side with a double rampart and intervening ditch. The defences survive quite prominently along the northern and north-western sides; however, they have been affected by plough damage and quarrying. The inner rampart is 15m wide and up to 6m high and is revetted with dry stone walling visible along the north face. The intervening ditch occupies some 10m width between the two ramparts and survives to a depth of 0.5m. The outer rampart is 16m wide at its base and rises to 5.5m high with no external ditch. The area enclosed by the defences is 5.4ha in extent.

Eddisbury is essentially comprised of three areas; the main body of the hillfort including an entrance on the eastern line of the rampart, the north-western end adjacent to Old Pale Farm including a second entrance, and Merrick’s Hill to the south-east (or the Chamber in the Forest) where medieval reoccupation included a royal hunting lodge. Eddisbury was first included on the Schedule in 1934. The hillfort is well preserved around much of its defences despite some plough damage and quarrying activity. There is a strong potential for surviving below ground remains within the hillfort. (1)

The earliest archaeological investigations by Varley in 1936 to 1938 comprised the excavation of four areas and a number of sections along the defences. Based on his findings, Varley identified seven broad phases of activity at Eddisbury that had been largely accepted until more recent excavations between 2010 and 2011. In general terms, Varley’s phasing consisted of an early palisade enclosure, possibly Bronze Age in date, covering Merrick’s Hill and the main area of the hillfort. This was subsequently replaced by a univallate (single rampart) hillfort, possibly including an extension to the west. Eddisbury was subsequently rebuilt as a bivallate hillfort with more complex defences and the addition of a second guard room at the eastern entrance. Varley also maintained that the hillfort was deliberately demolished in the Roman period. Reoccupation of the hillfort as an unenclosed settlement was then identified in the 4th-6th centuries before two phases of mid to late Saxon settlement culminating in the construction of a burh in AD 914 by Aethelflaeda. At the start of the medieval period, however, the burh had been laid waste and only Merrick’s Hill at the east end was re-occupied as a royal hunting lodge called the Chamber in the Forest. (2, 3)

There have been several surveys of Eddisbury beginning with the Ordnance Survey in the 19th century and later by Forde-Johnston in 1960 (4), Ferguson in 1977 (5), Longley in 1979 (6) and the RCHME in 1987 (7, 8). In addition, geophysical survey has been carried out on a number of occasions in 1990 (9), 2004 (10), 2006 (11) and in 2010 (13). These surveys have added to our understanding of the surviving features of the hillfort through the identification of substantial quarrying along the southern line of the defences as well as the potential for further below ground features of archaeological interest.

In 2009 a new project to study the hillforts of Cheshire included a new excavation of Eddisbury. The work was carried out in two phases between 2010 and 2011 and included 15 trenches across Merrick’s Hill and the main enclosure by volunteers under the guidance of the Habitats and Hillforts Project leader and Earthworks Archaeology (17) as well as a further five trenches at Merrick’s Hill by the University of Liverpool (18). The results provided more detailed phasing of activity at Eddisbury based on stratigraphic analysis as well as scientific dating including radiocarbon, archaeomagnetic and Optical Stimulated Luminescence (OSL) dating techniques.

The evidence suggested three phases of activity prior to the construction of the hillfort. Phase 1 comprised a possible pit and hearth dated to c.1870 BC (CHER 866/1/1). This was followed in around 770 BC by a palisade enclosure (CHER 866/1/2). Other pre-hillfort evidence included traces of at least one stone circular feature tentatively interpreted as either a hut or cairn.

Phase 4 was identified as the earliest phase of hillfort activity (CHER 866/1/3) with the construction of a primary inner rampart and segmental inner ditch; it is likely that the hillfort from the outset enclosed the entire extent of Eddisbury including Merrick’s Hill and the north-western end where the original entrance was located. This was followed by the addition of the primary phase outer rampart and outer ditch and may have also included the creation of the eastern entrance (Phase 5). A secondary phase of re-cutting of the inner rampart and ditch occurred soon (Phase 6) after followed by alterations to the eastern entrance. Evidence from both Varley’s excavations and the 2011 work suggested the southern guardroom at the eastern entrance was burnt down in Phase 7 with radiocarbon dating placing this at around 375 BC. The outer rampart and ditch were later re-cut in around AD 10-130 (Phase 8) followed by the ultimate abandonment of the hillfort in the transition between the Iron Age and the Roman period (Phase 9).

Phase 10 dates to around the mid 3rd to mid 4th century AD and appears to be Romano-British occupation of the interior (CHER 868/1) before the site was left again until the late 9th to 10th century (CHER 868/2) when Eddisbury was fortified as a burh site (Phase 11). The medieval period saw the eastern end of the hillfort reoccupied as the royal hunting lodge the ‘Chamber in the Forest’ (Phase 12) (CHER 837/1/2).

The main period of occupation dates to the mid Iron Age and the construction of a single rampart hillfort, although no scientific dating for the primary inner rampart has been identified, the stratigraphy suggests the initial phase of hillfort construction dates to the 5th century BC followed by a later reconstruction between 400-200 BC. Varley’s excavations in the 1930s recorded evidence at the north-west end (Area 1), comprising traces of the inner rampart constructed with a base deposit of boulder clay with a revetted dry stone wall against the outer face and a core of sand, the whole measured 5.8m in width to the east of the entrance and 4.45m to the west. The berm comprised a gentle rock slope some 1.8m wide between the rampart and the inner ditch. The ditch appeared asymmetric in form with a vertical inner face and a sloping outer face. It measured 3m in depth, 7.15m wide at the mouth and 1.95m wide at the base. (2)

The northwest entrance appeared to have been a fairly substantial feature with a levelled in-turned entrance with evidence for a central stop gate post and one on either side. A cobbled road surface approximately 2.3m wide was recorded between the entranceway and bordered by a fence line of upright oak timbers with horizontal oak beams separated by puddled clay walling. Within the area of the hillfort, a series of hearths were recorded set into a clay floor stratigraphically contemporary with the inner rampart (2). No clear dating evidence has been provided, however, and it was not included in the recent re-investigations of the site. A re-interpretation of the evidence based on discoveries elsewhere at Eddisbury suggest that the inner rampart in fact represents two phases of construction as seen elsewhere (recorded by Garner as Phase 4 and Phase 6).

The excavations along the north and eastern lines of the defences provided more extensive evidence from both the 1930s excavations (Areas 2 & 3) as well as the more recent work in 2010 (trenches 1, 3, 4, 6, 7 and 14). The primary inner rampart in trench 1 was constructed from re-deposited red clay with sandstone fragments [108] and measured 9.5m in width and 1.3m in height. The outer face of the inner rampart was revetted with a dry stone walling that has survived to a height of 0.8m [110]. Some distance to the east, the inner rampart [705] was again identified as 5.88m wide. Survival here was less extensive and the primary bank survived only to a height of approximately 0.7m. Although there was no definite remains of the dry stone walling it is likely that it continued along the outer face. In trenches 4 and 6 the primary inner rampart [405] was only located to the south of the eastern entrance where it survived to a height of 0.65m and a width of at least 5m (17). Although it had previously been suggested that the original hillfort comprised only the main area at Eddisbury with the rampart following the line of the present field boundary between it and the north-west end, this has since proven to be unlikely based on detailed topographic survey by English Heritage and the results of the 2010 excavations (EH, 17). The cutting of the secondary inner rampart has been radiocarbon dated from deposits to around 400-200 cal BC. The recutting at this point involved the creation of a bank some 2m wider than previous and likely meant a proportionate increase in height. Evidence of the eventual collapse of the inner rampart has only rarely been recorded with evidence of a probable collapse of the rampart in trench 3 consisting of a spread of sandstone rubble [715] (17). Traces of a possible collapse layer were also encountered during excavations by Liverpool University (18).

The primary inner ditch, lying between the inner and outer ramparts, was recorded in Area 2 in the 1930s where it was only briefly mentioned (2). The ditch [14012] was again encountered in trench 14 in 2010 where it was recorded as being approximately 4m wide at the mouth and 1.6m deep with a flat base some 1.95m wide. More extensive remains were encountered in trenches 3 & 7 where it was revealed the ditch was constructed in a segmental fashion comprising substantial pits [320 & 322] separated by a narrow ridge of bedrock between them. At this point, the ditch was approximately 5.5m wide and 1.8m deep (17).

At some point after the initial construction of the hillfort, an outer bank and ditch were added to create a bivallate hillfort. First recorded by Varley in Area 2, the outer bank was set approximately 11m from the inner ditch and was described as consisting of a lower core of sandstone rubble with a low revetting wall on the outer face with an upper core of smaller sandstone rubble with an outer ditch measuring 3.96m deep with a width of 7.62m at the mouth and 0.9m at the base. At the time Varley suggested the upper core was suggestive of later reconstruction in the Saxon period (2). This evidence has since been re-evaluated in light of recent discoveries and it is now suggested that the lower core represented the primary outer rampart (Phase 5) creating a bank that has survived to a width of 3.7m and a height of 1m. The bank was subsequently heavily truncated (Phase 8) and rebuilt to create a much larger feature some 8m in width and 1.8m high and it was at this point that the stone revetment was inserted. The bank was likely also identified in trench 3 where a feature consisted of a series of dumped deposits measuring 7.7m in width and 0.7m high was recorded, however, it was uncertain at this time how this bank fits into the current sequence of events (17).

The primary outer ditch was only recorded in trench 3 of the recent excavations where it measured 4.5m in width at the mouth and 2.5m deep. Cut into the bedrock, it was possible to record tooling marks made during its original construction. The primary phase of silting in the base of the first ditch produced a radiocarbon date of 400-200 cal BC. The outer ditch was subsequently recut to a depth of 1.75m, radiocarbon dating from the base fill of the ditch suggest this occurred in around AD cal 10-130. (17)

The eastern entrance was a significant construction consisting of a hollow way cut into the solid bedrock leading to a passageway through the rampart. The passage was defined by a series of post holes cut to a depth of approximately 0.9m (2), given the substantial nature of the postholes it has been suggested that they could have supported a sizeable timber superstructure (17). The passage measured 2.4m wide and 16m in length with a level floor cut into a series of flat ledges. At the eastern end (outside the defences), sections of dry stone walling [408, 409] were recorded between the final two postholes that appeared to represent an embellishment of the timber structure. The walls curved round the outer face of the inner rampart to form the stone revetment. Inside the defences evidence of two guardrooms were identified. The southern guardroom may have been the earlier of the two and comprised a sunken chamber cut into the bedrock to a depth of 0.6m. It was accessed through a passageway between two postholes making up the southern line of the entrance passage. The floor of the chamber consisted of a clean grey sand overlain by a charcoal rich deposit. A third deposit of sand discoloured by heat suggest the guardroom was destroyed by fire, radiocarbon dates from this layer indicate this occurred between 375-160 cal BC. The north guardroom was at a higher level than the south and required a step up from the entrance passage to enter it. The interior of the chamber contained a black occupation layer containing a few fragments of pottery and a single flint flake (2). It was noted in 2010 that the walls survived to a height of 0.8m (17).

At Merrick’s Hill, excavations in the 1930s (Area 4) revealed traces of a badly degraded rampart with stone revetments on the outer and inner face of the rampart as well as traces of the inner ditch along the west and eastern lines of the defences (2). Later excavations in 2010 (Liverpool Area 1) revealed only a fraction of the primary inner rampart [487], surviving to a width of 0.84m and a height of 0.60m. The excavations at Merrick’s Hill also posited a third, southern entrance, however the evidence for this being an Iron Age feature is uncertain (18).

Occupation within the hillfort has so far only been located on only a few occasions with evidence of a possible internal cobbled track and traces of a hearth from trench 4 close to the eastern entrance, on this occasion, two radiocarbon dates were taken suggesting activity between 410-260 cal BC. In trench 14, towards the centre of the hillfort, a series of postholes representing a possible structure were also recorded (17).

For air photos,see NJH 3500/8, 9, 15-24; 3501/10, 11; SRW Delamere 2 & 3.

Publication of the investigations undertaken by the Habitats and Hillforts Landscape Partnership Scheme 2009-2012 (30)

<1> English Heritage, Various, Schedule Entry (Scheduled Ancient Monuments Amendment) (Scheduling Record). SCH2950.

<2> Varley W J, 1950, Excavations at the Castle Ditch, Eddisbury, 1935-1938, 102/1-68 Varley W J 1950 (Article in Journal). SCH7299.

<3> Varley W J, 1937, Eddisbury and the Castle Ditch (Newsletter). SCH7296.

<4> Ordnance Survey, 1870-1982, Ordnance Survey Archaeological Record Card, SJ56NE1 1979 (Index). SCH2487.

<5> Forde-Johnston J, 1962, The Iron Age Hillforts of Lancashire and Cheshire, p.9-46 (Article in Journal). SCH6913.

<6> Ferguson, A, 1977, Field survey of Eddisbury hillfort (Unpublished Report). SCH7297.

<7> Royal Commission on the Historical Monuments of England (RCHME), 1987, Castle Ditch hillfort, Eddisbury, Cheshire, reconsidered: The excavations of 1935-38 in the light of recent field survey (Monograph). SCH7298.

<8> Royal Commission on the Historical Monuments of England (RCHME), 1987, Castle Ditch (Eddisbury) hillfort, Archaeological Survey Report, R3560 (Client Report). SCH7305.

<9> Davies, M.G., 1990, The Castle Ditch, Eddisbury. A report of recent fieldwork and a reconsideration of occupation phasing (Unpublished Report). SCH7300.

<10> South Trafford Archaeology Group, 2004, Geophysical survey undertaken at Eddisbury hillfort by South Trafford Archaeology Group (Unpublished Report). SCH7301.

<11> Gifford and Partners, 2006, Old Pale, Delamere, Cheshire. An Archaeological Watching Brief, R2675 (Client Report). SCH5003.

<12> Gifford and Partners, 2006, Archaeological Recording At Eddisbury Hillfort, Delamere, Cheshire, R2607 (Client Report). SCH4655.

<13> Archaeophysica, 2010, Eddisbury Hillfort, Cheshire. Geophysical Survey Report (Client Report). SCH6436.

<14> National Museums & Galleries on Merseyside, 2005, An Archaeological Desk-Based Assessment of the Sandstone Ridge ECOnet Partnership. (Client Report). SCH4718.

<15> Oxford Archaeology North, 2007, Habitats and Hillforts of the Cheshire Sandstone Ridge, Cheshire: Archaeological Condition Survey, R3894 (Client Report). SCH5260.

<16> Oxford Archaeology North, 2008, Habitats and Hillforts of the Cheshire sandstone Ridge: Archaeological Desk-Based Assessment, R3893 (Client Report). SCH5259.

<17> Cheshire West and Chester Historic Environment Service, 2012, Report on an archaeological evaluation at Eddisbury Hillfort (DRAFT), R3556 (Unpublished Report). SCH7285.

<18> Mason, R, 2011, Excavations at Merrick's Hill, Eddisbury Hillfort 2011 (Unpublished Report). SCH7327.

<19> Varley W J, 1936, Recent investigations into the origins of Cheshire hill-forts (Article in Journal). SCH7314.

<20> Roeder, C, 1951, Prehistoric Glimpses of Eddisbury Hundred (Cheshire) (Article in Journal). SCH7313.

<21> Longley D, 1979, Prehistoric Sites in Cheshire, /43 (Report). SCH2719.

<22> Royal Archaeological Institute, 1845-Present, Archaeological Journal, 3/89-90 Cotton M A 1955 (Journal/Periodical). SCH338.

<23> Society of Antiquities London, 1921-2014, The Antiquaries Journal, 19/93+244 Varley W J (Journal/Periodical). SCH284.

<24> Lancashire & Cheshire Antiquarian Society, 1883 -, Transactions of the Lancashire & Cheshire Antiquarian Society, 51/51-59 Varley W J 1936 (Journal/Periodical). SCH3293.

<25> Forde-Johnston J, 1976, Hillforts of the Iron Age in England & Wales, /59.103 + 143 (Book). SCH1347.

<26> Challis A.J. & Harding D.W., 1975, Later Prehistory from the Trent to the Tyne: Catalogue and illustrations., 20/ii/44-45 Challis A J & Harding D W 1975 (Monograph). SCH7282.

<27> Shone W, 1911, Prehistoric Man in Cheshire, /57-58 (Book). SCH2710.

<28> Longley D, 1977-1978, Longley Archive, /no.458 & 460 (Paper Archive). SCH2005.

<29> Leah M, 2006, Site visit to Eddisbury Hillfort, January 2006 (Written Communication). SCH7302.

<30> Garner, Dan and others, 2016, Hillforts of the Cheshire Ridge (Book). SCH8403.

<31> Cheshire Historic Environment Record, 1990-2001, Jill Collens and Rob Philpott's Aerial Photographs, LM 90.2013, 10/08/1990 (Aerial Photograph). SCH5403.

<32> Higham N J, 1984-9, Professor Nick Higham's Aerial Photographs, 1987/E/8-9, 31/01/1987 (Aerial Photograph). SCH7659.