GGAT 93 Church Hill Enclosure and Roman site Survey and Excavation
The Church Hill excavation and survey was a project carried out jointly with the Department of Classics, Ancient History and Egyptology of Swansea University and the Royal Institution of South Wales (Friends of Swansea Museum). The fieldwork for this project took place over three weeks at the end of June and the first half of July. Thirteen students and two lecturers took part from the university, and there was some additional assistance from RISW.
Following clearance of herbaceous vegetation that had grown since the Forestry Commission cleared the site in the spring, two trenches were laid out. It had been intended to cut one trench 2x30m across the enclosure bank into the interior, and one 2x20m in the area of Roman finds to the south of the enclosure, but the location of standing trees and stumps made this impossible. Trench 1, outside the enclosure was 2x19m, Trench 2 across the bank was 2x10m, and following discussion with Cadw a third trench, Trench 3, also 2x10m, was cut in the interior.
Trench 1 ran east-west, with its eastern end intersecting the modern trackway. A wall running north-south was discovered, 2m from the eastern end of the trench and partly on the line of the trench. It was 0.60m wide and survived one course high above the foundation, built of limestone blocks with a well-made face on the east side, but with the west side in the same smaller rubble blocks as the core. It had not apparently suffered any ill effects from vehicles passing along the track. A deposit of rubble to the west of this wall overlay a cobbled surface, which did not apparently extend all the way to the west end of the trench, though time constraints prevented all but a small portion from being uncovered. More extensive excavation was carried out to the east of the wall. Here the corresponding rubble overlay a loose deposit containing combustion products, probably from some industrial process, which in turn overlay a layer of clay burnt red and yellow. Small patches of unburned yellow clay were noted adjacent to the east section.
Church Hill, Gower: Excavation (Horizontal Scale 1m) Trench 1 with Roman wall in foreground
Trench 2 was cut north-south through the bank on the southern side of the enclosure at the point that the trackway ran across it. Again, the vehicular traffic was shown to have done relatively little damage to the archaeology. The bank was about 4m wide and up to 0.7m high. It was constructed from large rubble blocks laid as a facing with an internal rubble core, and a line of regularly placed rubble blocks down the centre. Approximately 2.5m to the south there was a probable foundation, also of large uncut fieldstones about 0.8-1.0m in width. The intervening area contained a loose packing of rubble and earth fill together with larger blocks from the wall overlying a clay layer. At the southern, outer foot of these bank deposits was a rubbish deposit including very large quantities of oyster shell, but the relationship between the two could not be clarified in the time available for the excavation. The inclusion of pottery in the bank shows that it must have been constructed in or after the Roman period, and therefore the site does not appear to be one Roman-period occupation continued in and around a later prehistoric enclosure.
Trench 3 running northwest-southeast in the interior contained very little in the way of structural information. A patchy cobbled surface was noted, but this did not extend all the way to the northwestern end, which contained only a patchy deposit of rubble in a similar soft clay matrix as noted at the west end of Trench 1.
The topographic survey showed the enclosure in relief, together with the topography inside and outside. It also indicated the extent to which the present track and its predecessors had penetrated below the present ground surface.
Church Hill, Gower: Finds Recording
All datable finds that were recovered from all trenches were Roman. The initial assessment of the pottery indicates that the date range was mainly 2nd-3rd century, with only a few later forms noted; it included samian as well as coarsewares, although the quantity of finewares was not large. The relative proportions of the vessels was however unusual. This and the low numbers of fineware vessels suggested to the pottery expert that the site was not a villa; however, the finds associated with the building’s structure show that it is likely to have been of high status.
Brick and tile were frequent finds, and there were some fragments of opus signinum, and tesserae; a fragment of wallplaster also came from earlier surface finds. Preservation of animal bone was good, and the assemblage includes deer and wild boar, another likely indicator of high status. A few fragments of copper alloy and iron were discovered, which were not considered worthy of report, but no coins.