In Sir Cyril Fox’s full survey of Offa’s Dyke, the section at Craig Forda to the south of Oswestry Old Racecourse is part of a two-page discussion. While Fox’s description of the Dyke is limited here, no sections were taken south of the Racecourse, it clearly left its mark on Fox’s thinking about the overall function and importance of Britain’s longest linear earthwork.

Why? Despite being swathed in woodland, the topography here is dramatic. Indeed, together with the Llanymynech to Llynclys section, at Craig Forda Offa’s Dyke most closely resembles the Wye Valley stretches of any part of Offa’s Dyke north of the Wye.

Here, the Dyke maintains a long stretch at the top of the hill scarp looking out over the Morda. Before woodland enclosed around the abandoned earthwork, one can imagine how, newly built, and with woodland and vegetation cleared from its line both east and west, it would have been a striking and imposing sight, dominating the immediate surroundings to the west. Along this stretch, Offa’s Dyke maintains a constant level just ‘below the 1000-ft contour’ (Fox 1955: 61-62) overlooking the vale of the Morda. Fox says specifically:

‘The choice of so dominating an alignment is notable, influencing conclusions as to its purpose and significance’.

Fox 1955: 61.

His profiles of Racecourse Wood are actually to the north of the section I’m considering here: on land without public access. Hence, they aren’t directly useful in appreciating the sections now on the Offa’s Dyke Path. However, they reflect the overall pattern to the south. Both sections show the dyke having an eastern quarry ditch and one has a berm that is now a track. This mirrors what we see to the south in broad terms (see also Ray and Bapty 2006: 190-192).

Fox regarded these sections as never having had a ditch due to the steepness of the scarp. He describes is as a ‘doubly scarped’ hill-side, above to form the bank, below to form the ditch. He argues that the bank is formed entirely from material from the east side (Fox 1955: 61). Yet, as it descends to the Morda he argues it is a ‘broad bank’ marked by both an upper and lower ditch composed of quarried limestone (Fox 1955: 62). Fox argues that the presence or absence of the ditches need not reflect a change of design or designer, but adaptions prompted by the ‘mountainous alignment’ (Fox 1955: 62). I wouldn’t necessarily disagree, but I think the full extent and character of a ditch must be proven with excavation and its absence shouldn’t be presumed relying on surface evidence alone.

In these photographs, you get a sense of the scale of the earthwork even without a surviving ditch.

There are multiple points that are not appreciated in Fox’s review, including the scale and dominance of the Dyke as it overlooks the Morda. Here, I want to address how, situated at the top of the break of slope, Offa’s Dyke has a striking rocky composition here. Offa’s Dyke navigates varied geologies, and here at Craig Forda is looks very different from the stretches one the Dyke one sees elsewhere amidst fields in the Vale of Montgomery, Clun Forest or indeed to north around Ruabon and Brymbo.

I would go further, however. Notwithstanding the Devil’s Pulpit overlooking Tintern and associated sections north and south thereof overlooking the Wye Valley where the Dyke is cut into bedrock and is basically comprised of that natural bedrock, this ridge-top stance is most assuredly one of the most ‘rocky’ parts of Offa’s Dyke. Indeed, unrecorded by Fox, the Dyke actually incorporates a huge rocky spur where it kinks before continuing its descent to cross the Morda and on to Trefonen.

Large boulders at the base of the bank of Offa’s Dyke, Craig Forda
Large boulders embedded in Offa’s Dyke, Craig Forda
Large boulders incorporated into Offa’s Dyke, Craig Forda
On top of the rock outcrop incorporated into Offa’s Dyke
The rock outcrop at Craig Forda incorporated into Offa’s Dyke

Together with Llanymynech where Offa’s Dyke ascends to ‘Craig Offa’, this section firmly makes the point that Offa’s Dyke’s bank, while often comprised of earth, turves, pebbles and other loose small stones, was a truly stone-built monument in places. It certainly was made to look like a stone-built monument here. Not only are there excavated sections elsewhere where the monument incorporates basal deposits of large stones, there are instances of possible dry stone walling/revetment as well as the natural bedrock cleared back so its layers resemble dry stone (as discussed in part by Ray and Bapty 2016: 185-188).

Further still, the scale of the large stones and outcrops incorporated at Craig Forda really did surprise me and seems to have neither been fully surveyed or evaluated in publications to date. And while we cannot rule out later quarrying and stone moving to account for some of these stones, I feel it is sufficient evidence to suggest that calling Offa’s Dyke a ‘linear earthwork’ isn’t entirely accurate based on available surface examination and excavation data. Where stone was locally available, the monument was verging on the megalithic. Was it even created to look geologic: to look part of an ancient world, a wonder from the age of Offa I? If so, these geologic dimensions might have been integral to the ideology of Offa’s Dyke: created to look as old as the hills, a monument fit for legends.

Fox, C. 1955. Offa’s Dyke: A Field Survey of the Western Frontier-Works of Mercia in the Seventh and Eighth Centuries AD. London: The British Academy.

Ray, K. and Bapty, I. 2016. Offa’s Dyke: Landscape and Hegemony in Eighth-Century Britain. Oxford: Windgather.