Wat’s Dyke

The view west from outside Erddig Hall over the valley of the Black Brook

Running for 38 miles, seemingly continuously from the Dee estuary at Basingwerk to Maesbury Marsh south of Oswestry, Wat’s Dyke is the second longest earthwork known from early medieval Britain, only over-shadowed in scale by its neighbour: Offa’s Dyke.

As Cyril Fox put it:

‘Wat’s Dyke, throughout its course from the Dee to the Middle Severn Valley, marks the boundary between the lowland of the English Midlands, and the hill country of northern Wales’ (Fox 1934, 211)

and its course:

‘is designed to include as much country as lowlanders could conveniently occupy or control’ (ibid.).

Wat’s Dyke in Erddig Woods, looking north

This west-facing earthwork, was thought by Fox, and subsequent by many other scholars, as a less grandiose and shorter predecessor to Offa’s Dyke (Fox 1955; Ray and Bapty 2016). Building on Fox’s first important and systematic survey, the last half century has seen the work of many small-scale interventions by commercial archaeologists revealing its morphology at particular points but yielding no dating opportunities. Likewise, the long-term survey and excavations by David Hill and Margaret Worthington (involving over 60 interventions on its course) have served to both confirm and extend Fox’s inferences regarding its continuous and coherent construction and the lack of demonstrable gateways along its course (Worthington 1997; Hill and Worthington 2003; Worthington and Grill 2015). This work as corrected misunderstandings by Fox, although much of it remains partially published or disparately published. For instance, many steep scarps where Fox thought the monument was absent have been reinterpreted by Hill and Worthington’s fieldwork, confirming the dyke’s presence. The consensus has remained, however, that Wat’s Dyke was earlier than Offa’s Dyke: a monument of the 7th or 8th-century kings of Mercia, perhaps built by the powerful Mercian ruler who preceded Offa: Aethelbald (r. 716-757).

Wat’s Dyke, Erddig Wood, looking from the bank over the ditch to the scarp slope down to Black Brook, looking NW.

A decade ago, the opportunity for a large-scale opening of the earthwork ahead of development at Gobowen, allowed a more extensive excavation of Wat’s Dyke. Crucially, this facilitated a dating programme of the monument by Laurence Hayes and Tim Malim at this location (published in Hayes and Malim 2008). Their work suggested that Wat’s Dyke might well be later, not earlier, than Offa’s Dyke by between a decade to perhaps half a century. This might be crudely summarised as suggesting Wat’s Dyke was a 9th-century monument with a ditch to rival in size that known from many stretches of Offa’s Dyke, even if the bank itself seems less monumental than achieved for parts of Offa’s.

The implications of this work have really yet to fully percolate academic debates, let alone heritage interpretation. As a successor to, rather than a predecessor to, the late 8th-century Offa’s Dyke, it was perhaps the work of King Coenwulf, whose reign was dominated by accounts of campaigns against his Welsh foes. The king died at the northern end of Wat’s Dyke: Basingwerk. It might equally have been the work of his successors.

Whoever built it, it clearly aimed to stop up the clear gap in the northern extent of Offa’s Dyke and this might plausibly be regarded as a response to the increasing military power and influence of Gwynedd along the Flintshire coast and areas formerly dominated by the kingdom of Powys in the early 9th century. Consequently, it was Wat’s Dyke, not Offa’s Dyke, that defined the frontier to the Norman Conquest, as indicated by Margaret Worthington’s work on hidated and unhidated manors at Domesday, whose distribution corresponds to the line of Wat’s Dyke (hidated on the ‘English’ side, unhidated on the ‘Welsh’).

This is by no means certain, of course, and conversely, Ray and Bapty (2016), in their new book on Offa’s Dyke, venture an alternative suggestion. They argue that Wat’s Dyke might have been integral part of an Offan frontier, and thus an extension of, and perhaps used contemporaneously with, Offa’s Dyke.

So Wat’s Dyke mapped out and perpetuated a geo-political faultline in the Early Middle Ages between the powerful kingdom of Mercia, itself fluid in its fortunes and scale, and a series of British rivals to the west. In this frontier zone, Mercia had absorbed local politics during the 7th century and may have cultivated a distinctive linguistically and culturally variegated population living and working on either side of linear earthworks built as both defensive and aggressive statements and territorial devices. Offa’s and Wat’s Dyke might be seen as successively offering visually control land to the west and manage/restrict the movement of people and resources along and across their lines. They were also ideological statements of Mercian hegemony, as Ray and Bapty articulate in their new book.

What is clear, however, is that Wat’s Dyke, despite running through historic landscapes and NE Wales’s largest conurbation – Wrexham – remains poorly appreciated and understood, including by heritage professionals. A primary school is named after it, a few roads, but it is generally ambiguous in the consciousness of historians, archaeologists and local people.

Looking west from Erddig towards the coal tip of Bersham colliery and Ruabon Mountain beyond


Wat’s Dyke at Erddig

I recently visited the National Trust site of Erddig to explore the house, gardens and woodland, as well as to revisit the site of the earth-and-timber Norman castle there overlooking the Clywedog. Wat’s Dyke runs through it along the top of the eastern slope descending from Erddig Park to the Black Brook, and running into, and obscured by, the later castle on the promontory. My former student, Rachel Swallow, has discussed this castle and suggested its relationship with Wat’s Dyke is far from coincidental (Swallow 2016).

Erddig Hall – the line of the dyke is probably hidden beneath the gravel leading from the woods beyond straight towards the position of the photographer

I would suggest that very few visitors to this property appreciate the true importance of the landscape here for understanding the monument, and the long-term history of the Mercian frontier and subsequent Anglo-Norman March of Wales.

Wat’s Dyke survives in key stretches near Erddig Hall. Fox regarded the flanking spur of the motte as ‘almost certainly not, in its present form, Wat’s Dyke’, and it fades to a berm beside the motte. But then, to the south from the castle:

… then as the W. slope eases, a bank appears. This bank ends in the artificially levelled flat by E of Erddig Park. The sequence of earthwork forms, dependent on terrain, is normal; and wholly, or in part, the structure must be recognised as Wat’s Dyke. The Dyke probably existed, but has been destroyed, in the neighbourhood of the house [Erddig Hall], but traces of both bank and ditch are again present at one point a little farther on, on the edge of the steep wooded slope above the valley floor, by the rookery. A little farther on again, the steep and regular contour of this river scarp is broken by two small re-entrants. Here, the large and characteristic structure of the Dyke reappears, the intervening spur being cut off by a large bank and ditch (scrarp 22 ft, overall 48 ft), and the furst spur isolated in a similar manner.

Fox surveyed in particular by the Rookery and included a profile of the monument at this point. However, there are 7 successive points to make here about this stretch in the wood between the castle and Erddig Hall that seem to contrast with the way the monument has been described in the publications by Fox (1935) and Worthington (1997):

  1. The dyke survives remarkably well in the woods of Erddig Park as a large bank and ditch as Fox describes and is comparable in general morphology to other well-preserved stretches of the dyke on breaks of slope:
  2. However, Fox fails to describe adequately how its use of natural topography creates a counterscarp bank out of the natural ground surface between the natural steep slope down to the Black Brook and the ditch. Worthington argues this takes place elsewhere on steep slopes, yet at Erddig Park this actually occurs at the top of the slope on relatively flat ground, rather than adapting the scarp. My understanding is that, with only a few possible exception inadequately explained by Fox and subsequent writers, Wat’s Dyke is not supposed to possess a counterscarp bank;
  3. In this form, the bank would have to have originally been of monumental proportions or topped by a high palisade and wall-walk and/or watch towers to allow those on the bank to survey anyone approaching up the steep slope from the Black Brook. This makes no sense in military terms, although the bank, set so far back, would still have allowed longer distance views dominating lands westwards;
  4. By the same logic, rather than the ‘performative’ nature of Offa’s Dyke (as discussed recently by Ray and Bapty) which makes it seem imposing as one approaches from the west, the ditch and bank were effectively concealed from the view of those approaching from the valley;
  5. It raises the question: are we heard looking at a monument comprised of two banks – the bank itself and its equally monumental counterscarp created by the cutting of the ditch?
  6. If so, then the counterscarp would need to be accessible to allow defenders/patrols to visually control the valley below.
  7. This contrasts with the other stretches of Wat’s Dyke I’ve seen where it is presumed to utilize natural topography or it commands clear views to its west.

I’m happy to concede if I’ve misunderstood the monument or identified a section modified in later centuries. Still, the obvious solution is that it is best regarded here as twin-banked monument, or that its counterscarp on the tip of the slope was accessible for surveillance into the valley and served as the primary defensive feature, not the ditch and bank to its rear.

dsc09304Heritage Interpretation

Of course, following my recent critical discussion of the display of the dyke at Chirk, you might guess where this is going. It is clear that the monument is not fully understood and the one new NT sign board (beside the entrance to the castle), while creating a striking cross-section visualisation of the motte and bailey of the Anglo-Norman castle, renders Wat’s Dyke incomprehensible because:

  1. The line of Wat’s Dyke isn’t represented on a map and so it won’t be clear where visitors should go to view it;
  2. Some attempt at a cross-section or artist’s reconstruction would at least help articulate to visitors what Wat’s Dyke might have looked like (not easy to do given our limited evidence of any wooden or other superstructures upon the bank, or other features associated with it);
  3. The sign contains weird stylistic error of capitalising ‘Century’ and not having a hyphen for the adjectival use of what should be: ‘eighth-century’ or ‘8th-century’, not ‘8th Century’;
  4. Crucially, Wat’s Dyke is inexplicably attributed to the 8th century, which shows it is working on information at least a decade old and without factoring in the evidence from Hayes and Malim’s 2008 publication.

In summary, more consideration is needed to understand Wat’s Dyke in this location and explain it to visitors.



Fox, C. 1934. Wat’s Dyke: a field survey, Archaeologica Cambrensis LXXXIX Part II, 205-78

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

Hayes, L. and Malim, T. 2008, The date and nature of Wat’s Dyke: a reassessment in the light of recent investigations at Gobowen, Shropshire, in S Crawford and H Hamerow (eds), Anglo-Saxon Stud Archaeol Hist 15, 147–79.

Hill, D. and Worthington, M. 2003, Offa’s Dyke: History and Guide, Stroud: Tempus.

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

Swallow, R. 2016. Cheshire castles in the Irish Sea cultural zone, Archaeological Journal 173: 288-341

Worthington, M. 1997. Wat’s Dyke: an archaeological and historical enigma, Bulletin of the John Rylands University Library of Manchester 79(3), 177–96.

Worthington Hill, M. and Grigg, E. 2015. Boundaries and walls, in M C Hyer and G R Owen-Crocker (eds), The Material Culture of the Built Environment in the Anglo-Saxon World, Liverpool: Liverpool University Press, 162–80.