Blackhill Wood

Report by Bruce Glendinning & Andrew Dunwell, Centre for Field Archaeology, University of Edinburgh

The tower site at Blackhill Wood (NN 8452 1075) lies c.900m north of Ardoch Roman fort, and is intersected by a length of the perimeter defences of a Roman temporary camp of possibly Severan date. A plan and profile of the site were produced by Roy in the 18th century, at which time it survived as a substantial rectilinear earthwork. Exploratory excavations were conducted in 1974 by St Joseph, who defined the tower site as a double-ditched annular enclosure c.14m across internally and c.23m across externally, with an east-facing entrance. Three of the four sockets supporting the corner posts of the timber tower were identified within the enclosure.

Since the time of St Joseph’s excavations, the condition of the site has deteriorated markedly as a result of its infestation by rabbits. By 1997 only vestigial surface traces of the tower site could be discerned. As Blackhill Wood is a scheduled ancient monument, and in the absence of realistic alternatives for the management of the site, Historic Scotland commissioned CFA to excavate the site. The fieldwork took place in August 1997.

The tower site was found to comprise a sub-rounded enclosure bounded by two ditches and three banks or ramparts placed

within, between and outside the ditches. The inner rampart was constructed of turf, and is paralleled by the turfworks at other Gask frontier sites; the outer banks were of gravel upcast from the ditches. The overall width of the enclosure was 25.2m – 26.3m, and the interior area inside the inner ditch was 12.3m N-S by 11.2m. The various enclosing works appear to relate to a single phase of construction; there was no trace of ditch re-cutting. The causewayed entrance to the enclosure lay to the ESE, but was largely obscured by tree roots.

The ditches differed in size, with the outer being the slighter (2.5 -3.4m wide by 0.3-0.5m deep), while the inner measured 2- 2.5m wide by 0.5 – 0.8m deep. Upon excavation, it was confirmed that the outer ditch did not continue around the southern side of the site where the ground slopes away steeply. An oven, with a stone paved base, was located cut into this slope, and may be associated with either the tower or the subsequent temporary camp. Within the enclosure the sockets of three of the four corner posts of the tower were identified (the same as located by St Joseph). The position of the fourth socket is believed to lie beneath an oak tree, and is thus probably destroyed. The NE post (PH2) showed evidence of two superimposed post pipes. The tower defined by the posts measured c.3.7m NE-SW by 3.5m. Within the area bounded by the post holes, i.e. the interior of the tower, a platform was preserved comprising of paving laid on top of a levelled surface of gravel and burnt clay. Three stratigraphic phases were identified within this area, although these cannot be related to the structural phases identified within PH2.

A small amount of excavation was carried out on the defences of the temporary camp, which was demonstrated to be stratigraphically later than the tower. A buried turf line was located within the inner ditch of the tower and sealed beneath the rampart of the camp. The camp ditch had a standard V-shaped profile. The nature of its fills suggests a natural infilling rather than deliberate backfilling. Finds as with all Gask sites were rare. They included an unstratified sherd of Roman coarse ware; a square headed nail from the interior floor of the tower; and a stone incised with linear grooves from the sharpening of iron blades from the inner enclosure ditch. There were no finds from the temporary camp. Regrettably, soil profiles across the site had been heavily disturbed by rabbit burrowing and tree root action, and no opportunities for palaeoenvironmental sampling occurred.

The archaeological results of the excavation which are of particular importance include:

1. The form and stratification of the enclosing works and internal area of the tower site proved more complex than had been suggested on the basis of St Joseph’s exploratory excavations. The enclosure was proved to be sub-rounded, rather than annular in form. It is of similar size and shape to other recent Gask tower sites examined by David Woolliscroft, such as Shielhill South.

2. The presence of a turfwork enclosed by the ditches indicates that such features were not restricted only to single-ditched Gask tower sites further north. The width of the Blackhill Wood example, c.1m, is significantly less than for other identified examples (e.g. Parkneuk 2.74m; Gask House 2.74m; Moss Side 4.3m). This may reflect the limited area available within the enclosure due to the provision of a second ditch at Blackhill Wood.

3. The presence of two post sockets within the NE corner post of the tower suggests structural complexity, and the likelihood of rebuilding of the structure at some stage. Evidence for rebuilding has been similarly identified at Shielhill South and Greenloaning. The evidence for tower complexity at Blackhill Wood contrasts with that from the enclosing works, which reflect a single phase of construction.

Further details are contained in a Data Structure Report deposited in the National Monuments Record of Scotland, or can be obtained from the authors. Post-excavation analyses and the production of a final report are ongoing, and scheduled for completion in 1998. The project was wholly sponsored by Historic Scotland.

A long term research project to study the Romans north of the Antonine Wall