Excavation of a Roman industrial building east of Harville Road, Wye, Kent
In 2019 the cropmark (Fig. 1) of a possible Roman building and leat (an open watercourse conducting water to a mill) east of Harville Road was investigated, with permission of the landowner, by students from the Kent Archaeological Field School (KAFS) directed by Dr Paul Wilkinson. The site was located to the west and south of the village of Wye, about four miles north east of Ashford. The area of investigation was located to the east of Harville Road and west of the Great Stour River. The OS location to the centre of the site is NGR 604783 146517. Surveyed OD heights show that the OD height of the leat bed to the north is 30m aOD, whilst 100m to the south the OD height is 31m aOD giving a fall of one metre in 100m.
[fg]jpg|Fig 1: Red arrows indicate the Roman building and the Roman water course or leat. The direction of water flow to the sea is denoted by the external arrow. Google Earth 2003.|Image[/fg]
[fg]jpg|Fig 2: Location of the Roman building on the KCC HER at NGR 604783 146517 (red dot & arrow)|Image[/fg]
Geophysics resistivity surveys at Wye
Three resistivity surveys in March 2019, September and October 2020 at the site by members of KAFS, uncovered evidence of significant Roman buildings including the possible site of a Roman mill (Fig. 3 & Fig. 4)
The image (Fig. 3) shows an overlay of the three surveys in grey scale, dark is higher resistivity, light is lower resistivity. Darker areas of higher resistivity are often the result of stone, brick and other dense building materials in walls and footings. The dark areas in the South East corner were excavated in 2019 by KAFS and proved to be a Roman era building thought to be the mill. All of the surveys were aligned on the NE/SW axis of the mill to make it easier to interpret the data.
The rectangular pattern at the eastern edge of the overlay was probably an outbuilding or substantial enclosure of dimensions 5x10m. There may be other faint, much larger rectangular features around this possible enclosure. To the west were a dense range of features which were likely the remains of a significant building, although much disturbed by deep ploughing. These may in part be the remains of a building partly excavated in 1972, containing a hypocaust floor (Fig. 5).
[fg]jpg|Fig 3: The W. Martin survey shows in some clarity the Roman Industrial Building (red arrow) the subject of this report and additional buildings to the NW (red line).|Image[/fg]
[fg]jpg|Fig 4: The W. Martin geophysical survey shows the Roman industrial building investigated by KAFS and the area of an earlier fieldwalking exercise|Image[/fg]
[fg]jpg|Fig 5: Jim Bradshaw c.1972.|Image[/fg]
This building may extend over an area of 20m x 20m. Continuing to the north and west was more evidence of linear features on the same alignment as the mill. These could be a further range of buildings. A previous magnetometry survey of the field showed some features at the NW edge but there are only faint signs of this in the resistivity survey.
Finally there was an intriguing NE/ SW linear feature at the edge of the overlay which seemed to point to further demolished structures to the South on the same alignment as the mill, and might be connected to the central features.
All in all, this suggests a very exciting site and promises to yield more interesting results if further surveying can be undertaken in the area to the South. Resistivity is highly revealing of buildings and buried structures, but can be laborious on large sites.
The KAFS members, Rebecca Parr, Stewart Brown, Zoe Schofield, and Beatrice Nicholas, who participated in the three surveys in the cold, wind and rain will no doubt be rewarded in a future life.
The Roman building was part excavated after the geophysical survey by 25 students who attended the KAFS field school training week in April 2019.
The survival of the Roman building was good, with stone walls and some areas of sandy gravel floors, all covered by a layer of collapsed Kent ragstone walls (Fig. 7)
[fg]jpg|Fig 6: Area strip, map, sample (looking north)|Image[/fg]
[fg]jpg|Fig 7: View of Roman foundations and cross wall (Looking East)|Image[/fg]
[fg]jpg|Fig 8: Roman foundations and later Anglo- Saxon cill beam slot- red arrow (Looking NW)|Image[/fg]
[fg]jpg|Fig 9: Extent of the 2019 excavation (Looking North)|Image[/fg]
On top of this collapsed strata and cutting through it were cill beam slots of a timber building on a different alignment, identified by the pottery as belonging to the 5th-6th century AD (Fig. 8).
The revealed Roman building is about 32m long and about 10.5m wide, built of mortared Kentish ragstone with the collapsed walls indicating a height of about 2.5m for the outer walls (Fig. 9).
Archaeological finds and features were present in Area 1 and included robbed out Kentish ragstone walling in Test Pit 1. Surface finds of pottery sherds retrieved included fine ‘Belgic’ grog-tempered ware dating to c.25BC-AD200. Test Pit 2 included demolition deposits over Roman floors still in situ and twelve sherds of pottery, including Roman pottery from c.130-420 AD. Test pit 3 also produced Roman pottery.
The south wall of the Roman building was also exposed in this area and was built of flint nodules set in an off-white mortar with small well-rounded gravel inclusions (Fig. 10).
The south wall measured about 16m in width and continued to the south-east beyond the exposed building.
The central area of Area 1 revealed the remains of an oven with pottery sherds of Roman and Late Roman jars dating from AD 150-300. In addition five fragments of worked lava stone were recovered, one with a raised kerb which suggests a Roman origin.
The diameter of one fragment measures about 650mm indicating it was a millstone. All fragments are dressed with a pecked grinding surface.
Running north-south but ata different angle to the Roman building, a 23m x 0.40m cill beam slot inside the main Roman building has been dated by pottery sherds to the Early Saxon period.
A Roman external room to the east overlaid the route of the waterway or leat and a section cut through it outside the building revealed a Roman cut waterway about 3.50m wide and 2m deep with Roman pottery dating from c.43-250/300 and sherds of Early Saxon pottery retrieved from the bottom silt. The area was not fully excavated but the initial investigation did reveal some large quern stone fragments.
The excavation revealed the remains of a Roman industrial building with a later Anglo-Saxon timber structure built at a slightly different angle, suggesting the earlier Roman building had been demolished but the water resource was utilised by later settlers in the area.
[fg]jpg|Fig 10: Foundations and section through floor.|Image[/fg]
[fg]jpg|Fig 11: Aerial of KAFS site showing the exposed Roman Industrial Building. April 2019 (Looking E)|Image[/fg]
[fg]jpg|Fig 12: The 2003 aerial from Google Earth shows the site of Roman Industrial Building and the water leat running north and south with water flow strongest as it comes off the River Stour.|Image[/fg]
[fg]jpg|Fig 13: Area 1.|Image[/fg]
The archaeological evaluation on land at Harville Farm, Wye in Kent has succeeded in clarifying the form and function of a stone built Roman industrial building which seems to have been rebuilt as an Anglo-Saxon building in the late 5th century, presumably to utilise the surviving Roman watercourse or leat. The Roman building has been named by the historian Paul Burnham the ‘Wilkinson Building’ in ‘Discovering Roman Wye’ published by the Wye Historical Society in 2023.
For more information, you can access the full report on the Kent Archaeological Field School website here: www.kafs.co.uk/reports.aspx
Dr Paul Wilkinson is the director of Swale and Thames Archaeology, the founder of The Kent Archaeological Field School and a published author on topics including Pompeii, Beowulf and Archaeological practice.