Stonehenge Riverside Project: 2007 Excavation II
Durrington Walls houses
A total of ten Late Neolithic houses have now been excavated at Durrington Walls since 2005. Seven of these are from the east entrance of the henge enclosure and all appear to pre-date the construction of the henge ditch and bank. This summer we completed the excavation of four of these at the East Entrance, three to the north of the avenue running from the River Avon to the Southern Circle (the large timber circle first investigated in 1967), and one sitting on the south bank of this avenue. North of the avenue there was a group of five houses terraced into the valley side. The uppermost of these was a large house and its smaller ancillary building, both separated from the other three houses by a fence. Rubbish, particularly burnt and worked flint, appears to have been thrown against the fence from both sides, allowing us to identify midden waste to specific households.
In 2007 we completed excavation of the most solidly built house (House 851) and of two of the houses east of its fence line. House 851´s west side was remarkable in having the base of its chalk cobb wall still surviving. It was spread up to a metre across but was probably 0.7m wide originally, and formed the base of a wall set outside the stakeholes which supported a superstructure of wattle and daub. Cobb is a traditional building material made from crushed, conglomerated chalk and this is the earliest evidence for its use. This is also the first time that the wall of a Neolithic house has been found anywhere in England (examples are known from Orkney off the north coast of Scotland). House 851 was almost square (interior dimensions 4.8m N-S x 5.2m E-W) and had its doorway at the west end of the south wall.
The central hearth was oval, set into a square plaster floor surrounded by beam slots which formerly supported wooden furniture in the form of beds along the east and west sides, a storage unit or dresser along the north side and smaller storage units in the southeast and southwest corners. The only other house to have been built with a cobb wall was the small ancillary building. All the others possessed simple wattle and daub or wattle walls.
The two houses east of the fence (House 800 and House 1360 to its north) were also square (about 5m x 5m) but were less well constructed. House 800 had two different phases, first with its walls aligned on the cardinal points and its entrance to the south and then with its corners facing towards the cardinal points and its entrance to the southeast. This second phase cut through the southern edge of House 1360 to its north, which appears to have been contemporary with the first phase of House 800. House 1360 had its corners facing towards the cardinal points and its entrance at the south end of the northwest side. Slots for two box beds were found along the southwest and northeast sides, and the hearth was tended by someone sitting on the northeast side.
Of the two houses on opposite banks of the avenue, the northern one (772) was excavated in 2005-2006. The house on the south bank (902) was located in 2006 but fully excavated in 2007. It was of very different form to those in the northern group and most similar to House 772. House 902 was built in two phases, both of which had a central entrance on the east side. Its walls were only set into the ground as lines of stakeholes in its western half. Some of these were angled to form interior buttresses to support the west wall. The walls bowed out, indicating that the house was originally D-shaped with the flat side to the east. The absence of stakeholes for the walls at the front of the house is matched by the plan of House 772, and suggests that these two houses were open on their east sides, looking down the avenue towards the river.
The extent of the Durrington Walls settlement
House 902 and the large midden to its southwest indicate that settlement activity was particularly dense beneath the southeast bank of the henge. The basal deposit of bank fill dumped on top of these layers (deriving from the topsoil dug from the henge ditch) was also packed with occupation debris and wall plaster, indicating that houses had also stood where the henge ditch was later dug.
Three evaluation trenches dug around the eastern edge of the henge bank showed that the occupation surface of this settlement, detected in Trenches 1, 5 and 6, continued around to the north, preserved beneath the henge bank. However, long-term ploughing since the Iron Age has removed all trace of this occupation east of the henge bank except for the deeper cut features such as pits.
Similar occupation layers, rich in pottery, bones, burnt deposits and other finds, were detected beneath the henge bank on the north side in 1951, on the south side in 1968 and on the west side in 1917.
The Late Neolithic settlement at Durrington Walls can now be envisaged as a large circular village of many hundreds of houses, set around an open area or arena which was occupied by the Southern Circle, itself surrounded by an arc of special buildings which included a timber circle (the Northern Circle, excavated in 1967) and two houses which were set within timber palisades and ditched enclosures which appear to have been kept clean (excavated in 2006). This settlement, with a circumference of almost a mile, would have been the largest village known in northwest Europe. If the density of houses from the 2004-2007 excavations (one house per 120sq m) is representative of their packing around the settlement then we might expect over 300 houses to survive beneath the henge banks.
Preliminary results from environmental analyses suggest that this was a seasonal settlement. The absence of carbonised grain or quern stones and the lack of bones from neonatal pigs and cattle, together with the evidence for culling of pigs in the midwinter period, suggest that people journeyed here with their pre-prepared foodstuffs and animals only at certain times of the year.
The south and west entrances to Durrington Walls
The south entrance of the henge enclosure appears to have been blocked, thereby severing the route to Woodhenge. We were unable to recover any dating evidence for this event and this part of the henge bank had been heavily damaged by Iron Age and later ploughing so that only the `ghost´ of the bank survived as an unweathered pediment of natural chalk. Amongst a variety of pits and postholes within our trench into the bank, there was a setting of small postholes that may have formed the walls of a small square house about 4m x 4m, with its walls facing the cardinal points. There was no trace of a surviving floor or hearth but a small pit within its interior contained sherds of Beaker pottery.
A 56m-long trench was dug 20m west of the henge enclosure´s west entrance. There was no surviving buried soil, indicating that later ploughing had destroyed such layers beyond the edge of the henge bank despite the build-up of colluvium on this side of the monument. Amongst the tree holes and debris from the WW1 camp to the west, the only prehistoric features were an Iron Age ditch and two Neolithic pits. There was no indication of a monumental entranceway or approach to the henge from this west side. The low density of Neolithic features indicates that the village probably did not extend this far from the henge bank.
The palaeochannel within the River Avon at Durrington Walls
Two small trenches were dug by machine into the palaeochannel. One of these was 30m east of the end of the Durrington Walls avenue. At the base of its 1.5m deep sequence, pieces of burnt hazel sat upon the channel bottom. These were covered by a sequence of peat-rich alluvial layers which contained no artefacts other than two struck flints. The second trench was about 200m south of the first and was slightly wetter. It was 1.6m deep and its lower deposits were composed of peat, turning higher up to peat mixed with alluvium. This second sequence probably results from accumulation within a Neolithic reed swamp of slow-moving water into which sediments were deposited from the Later Bronze Age or Iron Age onwards when the river´s environs were heavily cultivated.