Monuments which are now considered to be henge-enclosures have been known about for many years, not least because of their substantial earthworks. However, it was not until the 1960s and a campaign of excavation and research under the direction of G J Wainwright that the characteristics of the class became clear. Key excavations were those at Durrington Walls, Wiltshire (Wainwright and Longworth 1971), Marden, Wiltshire (Wainwright 1971) and Mount Pleasant, Dorset (Wainwright 1979). The most recent synthesis of data relating to henge enclosures and related sites is by Harding and Lee (1987).
The most distinctive components of all henge-enclosure, as they survive today, are those comprising the boundary earthworks, namely the bank and internal ditch.
The internal ditch is bed-rock cut, typically between 7m and 16m wide, steep sided with a flat bottom, and between 2m and 5.7m deep. The ditch terminals are variously expanded slightly, in-turned, out-turned, or square butted, and are sometimes deeper than the rest of the ditch circuit. When excavated the primary silts of the ditch fill are usually found to be fairly undisturbed, although hearths and stakeholes were found on stabilization horizons sealing the lower fills at Durrington Walls and Mount Pleasant respectively.
The banks were constructed of material quarried from the ditches. Soil and turf form the core with an envelope of redeposited bedrock above. No certain traces of revetments have been found. At Marden the bank survives to a maximum height of 2.75m and a greatest width of 40m; where excavated it proved to be 13.5m wide and only 1.0m high. Variations in the size and scale of the bank have been noted at other sites too. Estimates of the original size of the bank at Durrington Walls suggests that it was 30m wide and 3m high when new.
At all recorded henge-enclosures the bank and internal ditch are separated from each other by a berm for at least part if not the whole circuit. The width of the berm varies considerably even within a single monument, at Durrington Walls it ranges from 6m to over 20m across.
The enclosure earthworks are punctuated by entranceways. Durrington Walls and Marden each have two entrances while Mount Pleasant appears to have had four. In all cases the entranceways are fairly wide; 22.8m and 30m in the case of Durrington Walls, 10.5m and 27m at Marden. No certain traces of gates or temporary arrangements to close the entranceways have been found at investigated sites.
In plan, the earthworks of henge-enclosures define slightly irregular shapes of sub- circular and oval outline, the boundaries having been mostly laid out as a series of nearly straight alignments. Most are full enclosures, but at Marden the south-western quarter lacks a boundary earthwork and seems to have been delimited by the river Avon. Internally, henge-enclosures range from about 185m across at Knowlton South to 387m at Durrington Walls. Externally the extremes, represented by the same two sites are 240m and 479m respectively. Marden is slightly unusual in being more oval in plan, approximately 429m x 280m internally and 392m x 537m externally.
Investigations within henge-enclosures have been limited in extent as even quite large excavations such as those at Durrington Walls only involved a small percentage of the total interior. What is, however, fairly clear is that henge-enclosures contain a relatively high concentration of archaeological features representing a range of components. This is also borne out by geophysical surveys at Durrington Walls.
The fairly high density of activity in henge-enclosures contrasts with what is known of standard henge monuments. A further distinction is the fact that henge- enclosures do not have the distinctive flat, almost peneplained, interior so common among standard henges.
One difficulty with understanding the components found within henge-enclosures is knowing whether they are all of the same date and whether they necessarily belong to the period when the henge-enclosure was in use. Some internal features may well pre- date the construction of the boundary earthworks but this can rarelt be proven.
Round houses have been found within two henge-enclosures, Durrington Walls and Marden. At Marden the structure was represented by a ring of substantial postholes some 10.5m in diameter together with several internal postholes. At Durrington Walls the structure was slightly larger, 14.4m in diameter, probably of more than one phase, and comprising a central setting of four large posts surrounded by a ring of smaller posts.
A Timber circle, similar in form to those known elsewhere as single monuments, forms a freestanding component within Durrington Walls and a similar structure lies within a small henge at Mount Pleasant. The Durrington Walls timber circle (southern circle) has at least three main phases, and it may be noted that the entrance to the circle is aligned with the south-eastern entrance to the henge.
At Mount Pleasant a small type I henge some 43m in diameter (internal) lies within the henge-enclosure in the south-western quarter of the site. The entrance to this henge is to the north, and within it there is a timber circle. In all respects this henge is identical with those that occur elsewhere as single monuments. A circular earthwork towards the southern end of the Marden henge-enclosure may also be the remains of a henge.
Within Durrington Walls, situated very near to the timber circle already referred to, was an elongated oval hollow some 12m by 6.7m interpreted as a midden because of the large quantities of dark soil, pottery, animal bones, stone tools and ash that were found within it. Stakeholes around the two ends may have supported uprights that sserved to contain the debris or to screen it from view.
Fences comprising lines of postholes and stakeholes were found at Durrington Walls. One formed a sort of facade to the timber circle while others flanked a pathway leading to the entrance into the northern circle. All the postholes and stakeholes forming these recorded fences were very shallow and it is probable that the visible lengths were part of a wider pattern of boundaries dividing-up the site.
Pits seem to be fairly rare within henge-enclosures, but one small example was located at Marden. Being only 0.59m in diameter and 0.1m deep it probably represents only the truncated remains of what was originally a more substantial feature. Truncatation of the original ground surface may be the reason that pits seem so rare.
Burials have been found at Marden and Mount Pleasant, in both cases being inhumations placed in the lower ditch fills above the primary silts. Occasional finds of disarticulated human bone are recorded at all of the excavated sites.
The small number of recorded henge-enclosures precludes any detailed typology of the class, although a two-fold division may provisionally be proposed as follows:
A. Henge-enclosures with complete circuits of boundary earthworks (eg. Durrington Walls, Wiltshire).
B. Henge-enclosures with incomplete circuits of boundary earthworks (eg. Marden, Wiltshire).
Similified ground-plans illustrating a selection of henge-enclosures are given on Figure 1.
Henge-enclosures were variously reused in prehistoric times. Bowl barrows were built at Marden and Mount Pleasant, perhaps reflecting the earlier importance of the site. At Mount Pleasant the timber circle within the small henge was replaced by a stone setting (?cove) in early Bronze Age times, and at about the same time a massive palisaded enclosures was constructed inside the ditch of the henge-enclosure, probably utilizing the part-filled ditch in a defensive way. At Durrington Walls traces of Iron Age settlement were found within and around the enclosure.
The purpose and function of henge-enclosures are not precisely known. Interpretations include ritual enclosures and large settlements. Given that considerable quantities of domestic debris and what appear to be round houses are found alongside other components such as henges and timber circles which are more regularly attributed ritual functions it may be appropriate to see henge-enclosures as residential foci that include provisions for the ceremonial needs of the community as well as their domestic needs. Studies of patterning in the deposition of pottery, flintwork and animal bone at Durrington Walls (Ricahrds and Thomas 1984) suggests that different parts of the site were indeed used in different ways; to take animal bones as an example it was found that pig bones were most common at the timber circle whereas cattle bones predominated at the round house (northern circle). The great size of henge-enclosures, the amount of labour needed to build them, and the large quantities of timber and materials required has led to the suggestion that these sites were central places of some sort, perhaps the residences of holy men or the great chief of the territory, or both.