The 2002 season at Barcombe, undertaken by the UCL Field Archaeology Unit and the Mid Sussex Field Archaeological Team, has recently come to an end, and the site of the main villa building has now been backfilled. As well as bringing to your attention some new discoveries that we have made this year, this interim report will also update our current thinking on various aspects of the site (cf. Sussex Past and Present Number 95, 6-7). Although absolute [?final] dating for most of the features is not yet available, it is possible to relate many features to one another based on their stratigraphic relationships and by the pottery and other finds that have so far been examined. In 2002 we extended the trench to include a large area to the south?cast of the main villa building, whilst also extending the trench to the south and east in order to investigate features partially exposed in 2001. We will attempt to deal with the site chronologically.

The earliest features we have at the site appear to date from the Bronze Age, and comprise a circular ring ditch, some 20 metres in diameter, possibly originally surrounding a barrow, together with two shallow linear features, running east to west across the site, which may be field boundaries. The barrow ditch, which is over a metre deep, has produced a few sherds of pottery and some pieces of flintwork.

The next phase of activity at the site is a newly discovered roundhouse (ie number 3), which provides the first evidence of settlement. It is located in (? on) a terrace cut into the slope immediately in front of the later villa, and partly lying over the by then filled-in Bronze Age ring ditch. The terrace had been filled in and covered over with almost half a metre's depth of domestic refuse, comprising broken pottery, animal bone and seafood shells, together with later building debris, discarded and broken metal and bone tools, and at least two hob-nailed boots. This material appears to have accumulated from the 1st century through to the final abandonment of the main villa building, and it was only towards the end of the excavation this year that we managed to remove all of this overburden and establish that there is a roundhouse below it. The roundhouse, which is some nine metres in diameter, comprises an outer wall made of wattle and daub, of which the stakeholes survive for much of the circumference. In places larger posts supplement the stakes, with a further internal group of larger posts that presumably provided support for the roof. A concentration of stakeholes on the south-west side of the roundhouse probably indicates the presence of a doorway. There is no evidence of a central hearth, but later activity may have removed this. Inside the roundhouse terrace, five shallow linear gullies were found running parallel to one another from north-west to south-east. It is not clear what these were, but it is possible that they may be internal divisions. The dating for roundhouse 3 has not been fully established, but how these features functioned here is little, if my, diagnostic Roman material associated with it, so it is possible that it is Late Iron Age. Part of a second terrace immediately to the south of roundhouse 3, appears to be another roundhouse of similar type and size, but the majority (three-quarters) of this structure lies outside the investigated area.

The next phase of occupation is the roundhouse (number 1) which was found last year underlying the main villa building. In 2001 it was thought that the burnt clay area within the roundhouse was part of the structure, but its archaeomagnetic date of 140-210AD (at 95% confidence) would have made it an extremely late roundhouse for southern Britain (see Current Archaeology 179 (2002), 487). After further work this year, it is now clear that the burnt clay area is stratigraphically later, as a number of post- and stake-holes that belong to the roundhouse were found scaled below the burnt clay. The enclosure ditch found last year, and thought to be associated with this roundhouse, has now been traced to the eastern edge of the site where it continues outside the trench. A ditch running north to south and turning to the west in the south-east extension was discovered this year, and could be a continuation of this enclosure ditch. On its west and north sides, the ditch, has a line of large postholes spaced approximately two metres apart along its outside edge. Other features may be associated with this phase, including a hearth or oven found in a pit just to the north of the enclosure ditch a north-south aligned ditch, and the large pit in front of the villa, found last year. An extensive area of flint metalling, which respects both the enclosure ditch and roundhouse (1), may also belong to this phase. A possible smaller roundhouse (2) was also found in 2001, just to the south of, and possibly overlapping, the larger one. It is still not clew whether this is a roundhouse as its gully and associated post- and stake-holes only survive as a half circle, and we do not currently know how its dating fits into the site phasing.

The function of the line of post-holes found in 2001 on the outside of the enclosure is also still uncertain, but it is possible that after the ditch had been filled in it was replaced by a fence on the same alignment. This next phase of occupation is represented by the simple rectangular flint building found last year. This building, which measures 10 x 9 metres, has narrow footings of flint bonded in clay, possibly originally supporting a timber-framed structure. It only survives as a flint footing on its west and north sides, with a possible robbed out eastern wall found this year. There is no evidence for a south wall, and only one possible internal dividing wall, also mostly robbed out. It is possible that some of these walls were rebuilt during the next phase and incorporated into the later building. The fence line referred to above appears to have subsequently been purposely dismantled, as the resulting post-pipes we all filled with a similar mortar-flecked fill with the occasional pieces of Roman tile. A shallow gully running north from the building also has a similar fill, suggesting it was filled in at the same time. It is likely that this happened when the rectangular building was demolished to make way for a larger building in the next phase.

There are various features that appear to be contemporary with the first masonry building phase, or at least pre?date the construction of the later winged corridor villa, some of which are industrial, and perhaps connected with the construction of the final building villa. These include a possible lime kiln and associated pit, the latter sealed beneath the burnt clay area dated to 140-200AD, and various quarry pits that were perhaps dug to provide clay for daub walls, and then filled in with rubbish and soil. These pits all seem to have similar dates of circa late 2nd/early 3rd century.

The new masonry building seems to have been constructed sometime in the mid 3rd century. Work this yew has concentrated on the cast end of the building and on investigating wall junctions in an attempt to determine whether it was built all in one go, or in several phases. Unfortunately as most of the walls have been robbed out to the bottom of the foundations, we only have occasional glimpses of this detail. It seems likely that the first phase comprised the rectangular series of rooms on the north side. The large wing rooms (Rooms 1 and 4) were then added, together with the corridor. At the east end there is no rear corner room to match Room 5 at the west end instead there is a smaller room (10) with a small square structure tacked onto its north side. This latter structure may be an external staircase, perhaps needed to access the ground floor that may have been some 1.5 metres above current ground level on this steeply sloping site. At a later stage the west wing room (1) was reduced in size by the addition of a dividing wall at its north end, which also seems to have extended the corridor right to the west boundary of the building. It is possible that other dividing walls were added later, such as the wall between Rooms 6 and 7 which appears to butt onto the north wall of the corridor.

Where wall footings have survived, they have been constructed of irregular flint nodules (from the South Downs) bonded in a chalk mortar. The flints have been laid into a footing trench, with web flint being carefully placed interlocking with the surrounding flints to give a very solid structure; in many places it is possible to identify in which direction the builder had worked during the construction of the wall. Below the flints there is sometimes a basal layer of chalk blocks that have been set into the natural clay. Between the bottom footing layer of chalk or flint and the mortared flints there is often a thin layer of re-deposited natural clay which seems to have been purposely deposited here to create a flat surface above the basal layer onto which the bottom layer of mortared flints were laid. Where underlying features such as filled in ditches or pits occur, the builders went to great lengths to ensure that they did not cause subsidence of the villa walls. The wall footing trenches were dug deeper, in some cases right into the bottom of the pit or ditch, and were packed with flints to ensure stability. In the south-cast corner of Room 9, an earlier pit was partly emptied of its original rubbish fill, and the empty pit was then packed with flints that interlock with the wall footings to reinforce this load bearing corner.

There is very little evidence for the structure of the villa above the wall footings, but finds in a few of the rooms and the corridor suggest that these areas may have had a flooring of plain red tesserae. In Rooms 1 and 8 small white, grey and red tesserae suggest the former presence of mosaics. Fragments of the painted wall plaster recovered from the backfilled robber trenches also indicate the presence of painted walls and perhaps ceilings in at least some of the rooms. It also suggests that at the time the villa walls were robbed they may have stood to a height that meant that plaster was still adhering to them. Apart from the finding of fragments of box-flue tile, there has been no evidence of a hypocaust heating system in this building However, the box-flue tile fragments that have been found (including some re-used as tesserae) have included some in constructional phases of the building, suggesting that nearby another, demolished building, had a hypocaust system.

The south-east extension was opened this year to investigate what we had assumed were agricultural outbuildings and/or accommodation for servants. However, within one room we discovered fragments of an in-situ tessellated floor. The floor was made from cubes (c. 25mm square) cot from tiles, and has survived in six small fragments where the remains of a wall footing have protected it from the ploughing which destroyed the remainder of the floor. Sections through the robbed-out wall lines of this building recovered more tesserae, painted wall plaster, and roof tiles, all of which suggest that this is fairly high status building, and not one of the agricultural or servants' quarters buildings we were expecting. Next year we hope to investigate this area further. A number of Roman ditches and pits were also found in the south-east extension.

After the abandonment of the villa, which is currently thought to have taken place c. AD300, there is no evidence of activity until the later Saxon period. In the corner between Room 11 and the corridor, an alignment of three postholes contain demolition material and cut through the Roman midden. Although nothing Saxon was recovered from these postholes, they were aligned with a large bell-shaped cess pit which contained a very humic fill with numerous animal bones and Saxon pottery. This group of features could be our first evidence of Saxon "squatter" occupation, possibly a shelter or building constructed against the remains of the villa walls.

The final phases of activity are associated with the robbing out of the flints from the walls of the villa to provide building material for the adjacent parish church and/or other local buildings. We knew from last year's excavations that this stone robbing started in the 11/12th centuries as evidence by pottery finds from the robber trenches. Finds of clay pipes suggest that such activities continued into the post-medieval period. In the south-east extension a shallow, curving gully was discovered running from the north-east to the south-west cutting through the top of the earlier Roman ditches. This had been partly filled in with flint nodules, but the fill also produced large quantities of 11th/12th century pottery. Within the area enclosed by this gully was an oval, bellshaped pit with a narrow slot cut at each end and containing a humic fill with early medieval pottery. This is interpreted as a cesspit, with the slots supporting a seat or plank. Other medieval pottery and finds came from this area, but as yet there is no evidence for structures. The gully is not deep enough to have been a boundary ditch, but could have been a simple water channel to protect the uphill side of the cess pit from water run-off from the higher part of the site. It is possible that the gully may have enclosed a medieval encampment for workers involved in the robbing out of the walls. This may have been a major job, involving a team of labourers for many months especially if needed to provide stone, to build the parish church, and the workers may have needed temporary accommodation located alongside the villa, perhaps using simple shelters or tents which would have left traces in the archaeological record.

Next year it is planned to resume excavations in May. Again it is proposed to provide both volunteer and training opportunities. Further details can be obtained here, or from Helen Dixey (Tel: 01273 845497).

Chris Butler (Mid Sussex Field Archaeological Team) and David Rudling (University College London Field Archaeology Unit).