| Summary : The Neolithic Windmill Hill monument encloses an area of 8.45 hectares, making it one of the largest in England; it was also one of the first to be recognised as a causewayed `camp' or enclosure. The finds have been important to establishing the extent of transport of artefacts and materials during the early Neolithic, the nature of early farming, and the development of pottery styles. The enclosure has three circuits, defined by the inner, middle and outer ditches. The smallest, inner ditch has no trace of a bank, and the north-west part probably had an entrance. The middle ditch is circular, may have traces of an inner bank, and a possible entrance in the form of a wide causeway, slightly offset from the entrance in the inner ditch. The largest, outer ditch has a bank which in parts survives to 0.7 metres high and 5 metres wide. The site was excavated by H G O Kendall in 1922-23, Alexander Keiller in 1925-29, I F Smith in 1957-58, and A Whittle in 1990. The finds included Neolithic flint artefacts such as arrowheads, axe-heads, a sickle blade and scrapers. Numerous animal bones and skulls as well as 'ceremonial chalk cups' were also found. A type of Neolithic pottery found on sites across Wessex was first identified here and has hence taken the name of the site: Windmill Hill type pottery. A radiocarbon dating programme now places construction between 3700-3601cal BC, and the continuation of its main use until about 3500 or 3300 cal BC, with the construction of the monument as a whole taking place over the span of just one or two generations. The date and the location in the Avebury area associate it with the origins of a unique monument complex, of which the earliest elements may have been some of the nearby long barrows. Later additions included Silbury Hill, the Avebury complex, the Longstones enclosure, and the West Kennet Farm enclosures. In the Early Bronze Age, round barrows were built on the hill itself and on the slope to the south. |
More information : (Centred SU 086714) Neolithic Camp (NR) (1)
A Neolithic causewayed camp on Windmill Hill, with three concentric banks and ditches (see plan) (2-3). There was Neolithic activity on the hill prior to the construction of the inner circuit, which cut some of a cluster of pits, and of the outer circuit, the bank of which overlay artefact scatters, pits and postholes. Outside the enclosure to the north there were four pits of Beaker or Early Bronze Age date. Others may have extended beyond it. Further pits and possible postholes were encountered outside the circuits to the east, during the excavation of a square enclosure which may have been a bedding trench for posts. The very few finds from these features would be compatible with a Neolithic date and, since at least one was cut by the enclosure, they may all have pre-dated it. Here again, pits may have extended beyond the excavated area. The date of the square enclosure itself remains uncertain. Bank structure was preserved only in the outer bank in the east of the circuit. Here, Smith's excavations showed that the first stage of bank construction had been the two low parallel mounds of topsoil and weathered chalk which may have been intended as setting out lines for the bulk of the bank. The fills of all three ditches were asymmetrical, arguing the former presence of internal banks even where none survived. All three were rich in artefacts and food remains. The absence of any surviving bank inside the inner and middle ditches was attributed to the practice of backfilling material from the bank over deposits placed in the ditch, which at the same time accounted for their integrity and good preservation, with the at least partial survival of the outer bank corresponding to a lower level of cultural material in the fills of the outer ditch. The plotting of individual sherds in 1957-8 showed that, in all but one of the sections then excavated, only Neolithic Bowl pottery (Windmill Hill Ware) was present in the primary and secondary fills, with Peterborough Ware, Grooved Ware (Rinyo-Clacton Ware), Beaker, Early Bronze Age and Roman pottery all occurring together in the tertiary fills. All the struck flint from the hill had been brought there, since the chalk in the vicinity does not contain flint. This also applies to the extensive flint scatter on the slopes of the hill to the south of the enclosure. (2)
Excavated by Keiller between 1925-1939 and by Smith 1957-8. Finds from the top of the primary silt included hearth areas with bones of cattle, sheep, goats, pigs, dogs and occasional human bones. Artifacts from the same level included stone implements, leaf arrowheads, scrapers, and other flint implements, and a profusion of sherds of baggy-shaped, round bottomed pottery for which Windmill Hill (3) became the typesite. Above this silt-surface were sherds of Peterborough and other Rinyo-Clacton wares and beakers of types A and B, as well as many leaf arrowheads and petit tranchets, polished flint axes, and fragments of axes of various foreign stones. Chalk phalli and chalk cups or lamps were found. Radio carbon dating has shown that the enclosure was constructed circa 2500 BC and that the site was occupied from circa 3000 BC. The majority of finds went to Avebury Museum. (2-3)
Centred SU 08677144. The neolithic enclosure of eleven hectares encircles the crest of Windmill Hill. Generally only the excavated parts show the causewayed nature of the eccentric circles and little survives of the associated banks. The outer ring is visible as are 0.6m high scarp around the south side and on the east and north-east discontinuous scarp/banks and ditches (averaging 1.2m high and 0.3m deep) are visible, although now partly mutilated by later quarrying. On the north and north-west the perimeter descends the steep Hillside as a series of shallow scoops but its subsequent ascent through woodland on the western side cannot be located. Of the middle ring only the eastern side, excavated by Keiller, can now be traced on the ground and this comprises a 0.5 metre deep ditch with intermittent causeways. The inner excavated ring is identifiable throughout as an 0.3m deep causewayed ditch.
Resurveyed at 1:2500. See Archive AO 2273. (4)
Excavations at Windmill Hill in 1988 under the direction of Alasdair Whittle formed part of a wider programme of research in the Avebury region (eg see also SU 16 NW 69, SU 07 SE 19). Excavation was small in scale and focused on the ditches. There was an abundance of animal bone deposited in spreads and dumps, and samples potentially rich in environmental data were taken for analysis. Pottery from inner and middle ditches is plain Windmill Hill pottery throughout, while in the outer ditch this occurs only in primary contexts. Later contexts in this ditch contain Ebbsfleet then Mortlake-Fengate, Beaker and Grooved ware. Thus it may be that the outer ditch is later than the inner pair. A grave containing an adult inhumation was found beneath the outer bank, as were pits and post-holes. Interim notes on the 1988 excavations have been published (5-7) as has a note on radiocarbon dates ranging from 4870 plus/minus 70bc to 4550 plus/minus 80bc obtained from bone samples recovered in 1988 (8). (5-8)
Additional references - discussion of causewayed and other Neolithic enclosures including Windmill Hill. (9-18)
Note on flint report in Smith's publication of Keiller's and her own excavations. (19)
Additional bibliography. (20)
A notable discovery of Keiller's was a square enclosure located outside the eastern side of the outer ditch and between two round barrows at c.SU18877139 (SU 07 SE 16 barrows A and B). Kinnes (21) mentions it in a discussion of Neolithic mortuary houses but notes that the uncertainty over its date leads to difficulties in explaining function or status. (21)
The causewayed enclosure was surveyed by staff of RCHME Salisbury in 1990 at the request of Alasdair Whittle of University College, Cardiff. The survey also forms part of an RCHME project to record the monuments of the Avebury World Heritage Area.
The enclosure consists of three slightly eccentric circuits of interrupted ditch occasionally accompanied by internal banking. The largest of the circuits encloses c9.5ha. Extensive traces of multi-period field system (see SU 07 SE 67) surround the neolithic enclosure and at least three later burial mounds (see SU 07 SE 16, 21) overlie the line of the enclosure. Refer to the archive report for a full discussion of the surveyed earthworks. (22)
The history of archaeological investigation at Windmill Hill, including both published and unpublished documentary sources, was reviewed as part of the RCHME project on Industry and Enclosure in the Neolithic. See archive report for details and full bibliography. (23)
Additional reference. (24)
The neolithic enclosure was mapped from aerial photographs. Windmill Hill has been photographed from the air many times begining in 1925 (25), and includes some taken of Keiller's excavations of the inner circle in 1929. (26-30)
Additional references. (31-34)
Excavation of the enclosure in 1988 included sections across all three ditches, and was accompanied by earthwork and geophysical survey, and a detailed reassessment of the records and finds from Keiller¿s excavations. Results included the following.
Earthwork survey showed that the slight trace of a bank, previously unobserved, may survive within the middle ditch and that an apparent double eastern entrance, formed by two breaks in the outer bank some 20 metres apart, both corresponding to causeways in the ditch, is not original, since the north gap is crossed by a vestigial bank and the bank terminals on either side of the south one are sharply truncated.
A magnetometer survey defined the degraded south-west part of the middle ditch for the first time. The definition of the south and west parts of the outer ditch was enhanced sufficiently to indicate that the segments here were, like those excavated to the north, longer than those of the other circuits. Few internal features were recognised, among them an additional round barrow between the middle and outer ditches, possible pits, and a few weakly defined linear anomalies. The poor definition of the southern parts of the middle and outer circuits was shown to be due to cultivation in the historic period.
More of the artefact scatter and features beneath the Outer Bank were explored, including the grave of an adult male in which the corpse had lain exposed for some time.
The outer of Smith's marking out banks was clearly visible, the inner less so. Pottery from the extenuated tail of the bank was multiperiod, but dominated by sherds of Ebbsfleet Ware.
Also uncovered were three postholes which aligned with a row already exposed, and others running at right-angles to them. The structure(s) of which they formed a part, may have pre-dated pits sealed by the bank; or there is the possibility of timbering connected with the outer bank.
Soil micromorphology indicated a prolonged human presence, including possible cultivation, before the construction of the outer bank. Molluscan analysis confirmed disturbance to the soil prior to the construction of the outer bank and emphasised that the surrounding area before and during the construction and early use of the enclosure was predominantly wooded, more so than that of local long barrows. In the outer ditch, woodland species were predominant throughout the primary and lower secondary fills. Open conditions were reflected only from the upper secondary fills onwards.
The radiocarbon dates were used to suggest at least three, perhaps four, phases: first, pre-enclosure activity sometime in the first half of the fourth millennium cal BC; second, the construction of the enclosure in the middle of the fourth millennium cal BC, together with the digging of a burial into the old land surface; third, some activity represented under the tail of the outer bank and at the junction of the primary and secondary fills of the middle ditch, dating to the latter half of the fourth millennium cal BC; and fourth, continued deposition in secondary ditch contexts at the end of the fourth millennium or the beginning of the third millennium cal BC.
Pre-enclosure activity on Windmill Hill was thought to begin in Phase B (circa 4000-3500 cal BC) and to continue into Phase C (circa 3700-3100 cal BC), during which the enclosure was built. It was suggested that the primary use of the enclosure lasted at least 200 years, and the possibility of a succession of circuits was suggested. Molluscan evidence suggested the possibility that the outer circuit might have been added later than the middle and inner ditches. (35)
Scheduled. Windmill Hill Neolithic causewayed enclosure is situated in close proximity to an early prehistoric field system and group of eight Bronze Age round barrows on a prominent but low hill north west of Avebury. The enclosure has a small circular area surrounded by three roughly concentric rings of banks and ditches. The ditches are interrupted by causeways, created as a result of the gang construction method used to form the enclosure. The outer limits of the enclosure form an oval ring aligned roughly south-west to north-east. The dimensions are 400 metres by 300 metres at the widest points. The enclosure was in use from about 3000 BC to around 2500 BC. The finds uncovered during excavation included Neolithic flint artefacts such as arrowheads, axe-heads, a sickle blade and scrapers. Numerous animal bones and skulls as well as 'ceremonial chalk cups' were also found. A type of Neolithic pottery, found on sites across Wessex was first identified here and has hence taken the name of the site. (36)
See source 48 for an online version of this designation.
The classic Neolithic 'causewayed enclosure', with three concentric but intermittent ditches. Large quantities of animal bones found here indicate feasting, animal trading or rituals, or perhaps all three. (37)
In 1992-3 the lithic scatter on the south slope of the hill was investigated by geophysical survey, fieldwalking, test-pitting and excavation. The exercise focussed on North Field,. Both fieldwalking and test-pitting returned highest densities in the north of the field, close to the enclosure, and continuation of one row of test-pits downslope showed that falling densities were genuine, and not simply an effect of colluviation. Bronze Age material tended to concentrate around the known round barrows. Investigation of magnetic anomalies led to the discovery of two Neolithic pits. A cluster of intercutting early Neolithic pits 100 metres south-east of the enclosure contained artefacts and animal bone, and more abundant charred cereals and wild plant foods than those in the enclosure ditch fills. The pits in this group yielded a quern fragment and two rubbers. Two more widely spaced late Neolithic pits 200 metres south of the enclosure yielded lower levels of cereal and higher levels of hazelnut shell along with far more abundant charcoal than in the earlier pits. (38)
Additional references. (39-45)
Windmill Hill was included in a research programme into the dating of the early Neolithic causewayed enclosures of southern Britain and of Ireland, using chronological estimates produced by Bayesian statistical analysis of radiocarbon dates. The results suggested that the first circuit at Windmill Hill was probably excavated in 3680-3650 cal BC. The inner ditch was probably constructed in 3670-3645 cal BC. The outer circuit was probably constructed in 3670-3635 cal BC. The middle circuit was probably constructed in 3640-3615 cal BC. It is probable that the inner ditch was dug first, and that the middle ditch was dug last. It is uncertain whether the outer ditch is earlier than the inner. All three ditches appear to have been dug in the 37th century cal BC, probably within the lifespan of a single individual, or over two generations.
The modelling confirmed that all three circuits remained open at the same time, for a long period. Deposition in the main use of the enclosure took place over a period of around 350 years. The model estimates that this main phase of deposition ended in all three ditches in the middle decades of the 34th century cal BC. This ending seems to represent a change in the use of the enclosure, rather than a complete cessation of activity. Later activity is represented in the ditches of the causewayed enclosure. It is possible that there was something of a hiatus during the last third of the fourth millennium cal BC. Activity continued into the third millennium cal BC and beyond, as seen in the abundant finds of later Neolithic and Early Bronze Age pottery, and associated animal bone. The estimates for the dates when the circuits were constructed are remarkably similar. It is cautiously suggested that activity probably ended some time during the 35th or 34th century cal BC. The dating programme radically alters the understanding of Windmill Hill. It appears that construction took place in the 37th century cal BC and continued in its main use until the 35th or 34th centuries cal BC. Windmill Hill might have been the first major act of monumentality in the area.
The main model suggests a span of a lifetime or of two generations (taking a generation as 25 years) for the construction of the monument as a whole. It began with the construction of an enclosure with a ditch whose segments were quite short and pit-like. Perhaps within the same generation was the construction of the larger outer circuit with longer, deeper and wider ditch stretches, earthwork banks, and potentially even timbering. Then came the middle circuit, a slighter construction, with pit-like segments accompanying longer stretches of ditch, and indications of a slight bank. This further served to elaborate the division of space on the hill. The first deposits in the inner ditch were varied, both the bones and other material may have been processed or possibly even curated. They may represent the remains of meals or feasts associated with initial construction. They should still have been visible in some way as the outer ditch was constructed. The use of the monument as a whole took place over a long period of time, and the end result may not have been what was originally envisaged. (46)
Detail added to the aerial survey transcription of windmill hill from a 2006 lidar survey. This included hollow way features to the east of the monument, which are probably more recent in date. (SU 0887 7138). (47)
The designated record of the causewayed enclosure can be accessed online from the National Heritage List for England. (48) |