Interpreting traces of present-day activity along Offa’s Dyke isn’t always straightforward.
In previous posts, I’ve reflected on a range of practices by which walkers leave their mark on Offa’s Dyke: sticking erect bird feathers into cracks on stiles and posts along the National Trail.
At Bronygarth in October, I noticed a further dimension that was new to me. In addition to a bird feather stuck into the top of a post beside a stile between a field and a lane on the National Trail, there was, suspended by a lace, a single well-worn walking boot (right foot I think).
Across the lane, in a nearby field, the second had been discarded.
What’s going on here?
Is the feather associated with the boot or are they unrelated?
Many questions remain unanswered…
Were the boots discarded as uncomfortable by a frustrated walker? If so, what did they use to walk onwards to the next Blacks or Millets?
Was one of them then tied to the post by the same person at the same time, or by different people at different times?
Could this be a memorial to a dead walker?
Does it celebrate walking?
Or else was this the makeshift use of an abandoned boot by the feather depositor?
Could the farmer be pointing out the refuse left by walkers by suspending one in a prominent position?
Leaving worn-out boots behind as a sort of salute to the trail is common on other trails, so I suspect that’s what happened here. They would have been carried to the spot purposely, the walker already wearing their replacements. Usually both boots are tied to a post together, so my best guess is that they weren’t secured very well (especially since you imply the first one was hanging by a single lace), and the second one was stolen by a fox. I’ve heard several stories of foxes stealing footwear, playing with it and then discarding it in the middle of a field or garden when they get bored; possibly the smell of sweaty feet interests them!
If we follow this interpretation, what I witnessed was the result of multiple agents: human intentional and accident, and then fox….
Maybe we’ll never know the precise sequence that led to this eerie deposit. However, there are a range of other inscribing and depositing practices along Offa’s Dyke. Someone seems to have taken the trouble, for instance, to balance a mushroom on a fence-post beside the dyke.
Meanwhile, on beech trees along the dyke, there is the occasional graffito.
Whether long-distance walkers or local people, walking footpaths along the dyke rarely yields litter, but occasionally one encounters more evidence than the trail and the stiles…
Complementing my recent blog-post about the CPAT dig at Chirk Castle, I wanted to explore other nearby sections of Offa’s Dyke a fresh. I’d previously discussed the well-preserved ditch and bank of Offa’s Dyke at Bronygarth as part of a broader discussion of the design and landscape context of the monument between Craignant and Chirk: see this link.
Whereas at Chirk Park, the ditch has long been filled in, the slope down from Chirk Castle to the Ceiriog does preserve a more substantial ditch, although heavily denuded and difficult to photograph.
In contrast, looking south from Chirk, one of the best-preserved sections of Offa’s Dyke anywhere is on the south-side of the Ceiriog valley as it rises from Bronygarth towards the Nant Eris.
At various points on this route, the ditch is clearer, and as one gets higher up, at certain points one can witness the width and depth of the V-shaped ditch.
Indeed, the opening image of this blog, taken looking S along the line of the dyke, reveals clearly the scale of the ditch, albeit after more than a millennium of erosion. Add to this another 1-1.5m of depth and you get a sense of how the CPAT dig’s dyke may have once appeared soon after creation.
The Offa’s Dyke Collaboratory visited the dig in its early stages two weeks ago, as part of the Offa’s Dyke Conference. Following this link for details of the event and visit. Returning to the site at the end of the dig, I was privileged to see the final state of the CPAT trench. I defer to CPAT themselves to disseminate the results in due course. However, with the permission of Ian Grant, I share some preliminary thoughts on what are truly monumental results. On a personal note, it is the first time I’ve had the opportunity to see first-hand what relatively very few archaeologists have seen: a completely excavated section of Offa’s Dyke’s ditch!
The results astounded me and photographs struggle to convey the monumental scale of the ditch of Offa’s Dyke that CPAT revealed. Ian and his team of experienced archaeologists and volunteers did not simply follow the ditch-cut’s sloping angle and presume they’d reached the bottom at just over 2m. Instead, Ian’s archaeological experience paid off and he followed the ditch as its cut a distinct sharp-angled slot down another entire metre!
Just let that sink in: Offa’s Dyke has just been shown to have had a whopping 5-6m wide and 3m-deep ditch near Chirk Castle! This ditch has completely disappeared over the centuries, and particularly during post-medieval landscaping, just as the bank itself is heavily denuded at this point and might have originally been over 2m high (based on comparisons with other excavations (Ray and Bapty 2016: 186). Offa’s Dyke’s bank and ditch – notwithstanding any pallisade or other features that defined the original construction, has been revealed in this section at least to have been a truly monumental barrier to be reckoned with. Completely lost at this point until the CPAT dig, the rediscovery and scale of the dyke are phenomenal.
Charcoal samples from the ditch might assist with further dating of the dyke and its immediate post-construction filling. Moreover, further work on Offa’s Dyke at Chirk Castle is urgently needed in future years to ascertain the character of the heavily denuded bank material – surviving in only a fraction of a metre of material where the bank was, and in a mixture of material, including some large stones that might have formed part of a revetment, slumped into the ditch. Further work is also needed to ascertain the presence and character of quarry ditches to the east of the bank, and any counter-scarp bank to the west of the ditch. For now, I simply want to celebrate these excellent results, mirroring what CPAT achieved earlier this summer in relocating Wat’s Dyke just south of Erddig Hall.
A bit of context
Keith Ray and Ian Bapty’s recent book reviews published evidence from surveys excavations for the scale and character of Offa’s Dyke ditch. They question the statement that Offa’s Dyke is usually 2m deep (by Hill and Worthington 2003) and cite Mainstone (Shropshire) and Redwood Lodge, Buttington (Powys) as excavations showing a 3m deep ditch combined with a varying width of between 6m and 9m. Elsewhere at Orseddwen, south of Selattyn Hill, it was suggested that the ditch was only 3m wide and 1m deep. In the accompanying notes, Ray and Bapty query whether the ditch was fully bottomed at these locations. It is important to note that this isn’t simply an issue of archaeologists not identifying the depth of the cut, but identifying the location of the ditch itself. Given bank material might slump into the ditch differentially on the west side of the ditch, encourages the possibility that an unbottomed excavation would misidentify a later slump line as the ditch profile, but also misplace the ditch west of its actual line. To my eye, such a situation of horizontal shift in the bottom of the slump lines occurs in the top of CPAT’s section at Chirk Castle. A further distraction are later ditches that look like a ditch for the dyke, but are instead later recuts for land-drainage purposes as at Bronwylfa Road, Esclusham, Wrexham (Ray and Bapty 2016: 181). Such factors can distract archaeologists, especially when dealing with a ditch on such a truly gigantic scale.
Ray and Bapty then turn to the evidence from Ffrydd Road, Knighton which indicates an 8.5-9m wide and 3m-deep ditch at one point, but only 4m wide and yet still a consistent 3m deep ditch at another. Significantly, a slot at the base was identified, and one wonders if this was fully bottomed out given the depth of the slot identified at Chirk. Moreover, the Chirk evidence confirms Ray and Bapty’s argument that, the ‘most common as-dug ditch-profile appears to be a V-shape’ (Ray and Bapty 2016: 182), but raises concerns about whether a slot has been missed by some previous excavations.
Notwithstanding the validity of Ray and Bapty’s concluding point that a ‘suite of build practices was brought into effect that complemented the suit of placement practices’ (Ray and Bapty 2016: 184), the Chirk evidence might well suggest that Offa’s Dyke, outside of situations where the dyke sat at the top of steep slopes where a shallow ‘notch’ might suffice for a ditch, might have been of varying width but a relatively uniform 3m in depth. Further work is clearly needed to ensure ditch-cuts are followed to conclusive bottoms in order to reveal the full monumental breadth and depth of Offa’s Dyke as suggested by the Chirk evidence.
Last year, Dr Keith Ray and I explored the line of Wat’s Dyke around Holywell and Basingwerk, Flintshire. Recently, I went back to explore Wat’s Dyke at its significant northernmost section in more detail. I came to some further (and I think rather significant) observations regarding how the Dyke is behaving as it navigates the Greenfield Valley.
What Keith and I failed to do was explore in detail the surviving line of the dyke in the Greenfield Valley as it ascends from Basingwerk south up a south-east side of the valley towards the upland around Holywell called The Strand. Here, in Strand Coed (Strand Wood), the Dyke isn’t identified with surety.
Part of the problem with our previous visit was the fact that Fox’s description is so difficult to unravel without a detailed knowledge on the ground. In his 1934 report, he states:
Wat’s Dyke was until recently visible as a partially levelled bank with deep W. ditch, 150 years in length, on high ground E. of Holywell, known as the Strand. The construction of a row of cottages has caused the ditch to be filled in for use as a roadway, but traces of the bank still survive. There is now no visible evidence of the Dyke on its traditional line northward from the Strand along Strand Walk, but the alignment of this Walk on to the point where the tiny valley of the Strand becomes a ravine is very suggestive of its former existence.
So Fox argues that the path known as Strand Walk might follow the line of the dyke from the relatively flat upland area east of the Greenfield Valley where it is heading for the Bagillt Brook – the Strand – northwards as it descends down into the Greenfield Valley.
Well, I agree with Fox that there is no demonstrable trace of the dyke in terms of bank or ditch. However, the line of the Walk he describes is unquestionably a notch below the top of the slope that is a very good candidate for having previously been the line of the ditch. The bank, if it had ever existed, would have been the slope itself, and at the top of the slope where the modern field boundary runs.
So I would suggest that Fox is overcautious in his determination here, and the Walk is in fact following the ditch of the dyke as it navigates the east side of the ravine at a consistent c. 2.5-3m vertically downslope of the top of the ridge. We can look at this another way: there is no other way the Dyke could have navigated a path between the surviving section of to the north (see below) and the historically attested line of the Dyke on the Strand to the south. If we don’t believe this is the route: there is no other option for how it would have once existed elsewhere.
The implications of this are, however, not made clear by Fox. The Dyke neither hugs the top edge of the Greenfield Valley south of this point: it heads across flat hilltop towards a high point by/at the Holywell Windmill before descending towards the Bagillt stream. Likewise, it doesn’t continue north along the edge of the valley which would have required it to jump around another ravine to attain higher ground once more. Instead, it uses the ravine here to descend north/ascend south between high ground and the valley bottom. In so doing, it marks a course direct for the spur that became occupied by the Cistercian abbey of Basingwerk in an unquestionable fashion.
The next piece of Fox’s text is significant, since it relates to a section of the dyke that is arguable surviving: the most northerly section of the dyke extant:
This ravine opens out above Meadow Mills on the flank of the deep trough valley carved out by the Holywell stream: and here a small piece of bank is seen, 21ft. in breadth, with an over-deepened W. ditch, which is possibly part of the Dyke. It seems to die out in the railway cutting. Beyond this point northward nothing recognisable as Wat’s Dyke is to be seen.
Here, the Dyke is discernible, and while maps make it seem as if it is continuing its route following the contours, the closer examination shows it is doing anything but. As mentioned above, the surviving section is cutting across the contours, heading down into the valley north/ascending up to the top of the ridge of the Strand Coed ravine to the south. Therefore, if this is indeed the line of the Dyke, it is doing something very significant at this point: dropping in an unprecedented fashion down into the valley, not to cross it, but to follow it down towards the estuary of the Dee at Basingwerk.
Despite being a shorter monument than Offa’s Dyke, because its course navigates lower elevations, there are more situations where it navigates modern conurbations. This is a challenge for the preservation of Wat’s Dyke, but offers a host of opportunities to engage large numbers of people with this significant early medieval monument. Sadly, for most people living in suburban and urban areas, Wat’s Dyke is an invisible giant.
So on New Year’s Eve I decided to head out to explore/re-explore a few locations where the important and enigmatic Wat’s Dyke intersects with the largest town in North Wales: Wrexham and its immediate suburbs. I then tweeted. supported by images of the dyke at Ty Gwyn Lane, Wat’s Dyke Primary School, and beside the Premier Inn opposite Wrexham railway station:
This received a flurry of interest in basic and important questions about Wat’s Dyke: what is it? where can we see it? It dawned on me that I couldn’t easily answer their questions. I did, however:
give my own summary of the monument: likely to be early 9th century and built as part of a complex Mercian frontier zone, a successor to Offa’s Dyke, it runs 38 miles from Basingwerk on the Dee estuary to Maesbury on the Morda;
But what about a guide to where Wat’s Dyke runs? Surely I could direct them to something already published and coherent? Sadly no!
Detailed Ordnance Survey maps will help the walker find sections of the monument, sure. However, they aren’t fully reliable and appropriately annotated for many sections. The Wat’s Dyke Way Heritage Trail is useful, but has no information about the monument itself along the course. Coflein will allow you to find details on individual points along its course, but not about how it looks and how publicly accessible the locations are. More/different details can be found on Archwilio – the online Historic Environment Record hosted by the Welsh Trusts: it gives you a line of points where Wat’s Dyke can be located, and details for each point. However, the ‘dots’ don’t always match up with the stretches of the monument that remain extant. Entries sometimes contain no information, and some sections of the dyke have no description. Again, access issues aren’t easy to match up with public rights of way, although a combination of Coflein/Archwilio with an OS map will get you to most locations.
Still, none of these are designed as public-facing and user-friendly for someone wanting to explore this monument in their neighbourhood. Moreover, the only published and easy-to-appreciate maps are those on the macro-scale: charting the dyke’s overall course in relation to the broader landscape. For instance, Fox’s 1955 Archaeologia Cambrensis map shows Wat’s Dyke from Dee to Morda, with the sections where it survives as an earthwork marked clearly in relation to rivers and in relation to steep river valleys. He also marks areas where he postulates it crossed cultivated land around Wrexham, Ruabon, Gobowen and Oswestry.
In 1997, Margaret Worthington produced a different version, focusing on the topography more. However, it is not detailed enough to let anyone navigate the precise route of Wat’s Dyke in relation to the existing landscape.
These are not designed for people to visit the monument and the scale prevents this in any case.
So how do we expect laypersons to visit and appreciate Britain’s second-longest early medieval linear earthwork if we cannot even put up a coherent heritage trail and heritage boards, or even web resources to guide people with a simple map??
This is especially important for Wrexham, where there is commemorated in buildings and a road – Wat’s Dyke Primary School, Wat’s Dyke Way – and where the monument runs through the western side of the town from Bryn Alyn hillfort and Pandy down to the Clywedog and Erddig Hall. In this area, while heavily damaged and lost in many places due to post-medieval/industrial activities and modern housing, it affords tens of thousands of people with the ready opportunity to visit a section viewable from public accessible land on their doorstep. No matter how humble, damaged and overgrown it might be, surely there must be easy digital resources and apps to help facilitate engagement with it? Even Erddig NT have nothing coherent and up-to-date to say about it, and the heritage board at Erddig castle bears out-of-date information despite being new.
Something far more slick could be developed, but let me sketch a rough attempt to answer the question ‘Where can I visit Wat’s Dyke in Wrexham’, focusing on 8 locations in the suburbs of town, and of course focusing on publicly accessible locations.
Let’s take each location in turn from North to South, identifying access, what you can see, and what it tells us, as well as links to previous blog-posts about each.
I’ve previous posted about this section here. The Archwilio entry is short and basic:
The dyke is badly damaged by quarrying and agricultural activitity. Present for c150m to north of A483 road. Further to north line marked by hedgebank. Remains consist of bank max 1.3m high, hollow to west. Southern end of this section of Dyke cut through by the building of the Wrexham by-pass.
Via a footpath running south from Plas Action Road just after the Bluebell Estate and before it heads west over the railway line, one can traverse a good stretch of Wat’s Dyke as far as the A483.
The ditch is largely filled in, and the bank denuded significantly, as the Dyke forms the boundary of post-medieval fields. However, don’t give up (as I did before) until you hit the bank of the A483 where the dyke has been destroyed for some distance by the road-building until one can see it again on the south-side of Ty Gwyn Lane (see point 2, below). This is because, to the south, close to the A483, the bank survives increasingly better, until it is almost 2m high. Furthermore, at the southernmost part of this section, a gully to its west is large and wide, implying the presence of a silted up ditch.
What’s the dyke doing here? It’s crossing flat ground with good views westwards, between the Alyn and Clywedog.
2. The School Dyke
Ty Gwyn Lane and Wat’s Dyke Primary School – Wat’s Dyke – Wrexham De221 – Primary Reference Number (PRN) : 106673
There is no Archwilio description of this section in Garden Village. On the south side of the A483, from Ty Gwyn Lane up to Wat’s Dyke Primary School, there is a good section. You can park on Ty Gwyn Lane, or else on Wat’s Dyke Way and then walk into the grassed area kept clear of houses and explore the dyke.
The bank again varies in height, and is higher to the north, beside Ty Gwyn Lane.The ditch does actually survive as a wide, low depression, particularly at the top of the hill at the southern end of this section, beside Wat’s Dyke Primary School.
South of the primary school, the Dyke can be seen for a short stretch but then disappears beneath the back property boundaries of houses.
The dyke is here navigating a modest but significant high-point between the Alyn and Clywedog. Its precise route optimises visibility west and maintains visual communication eastwards.
3. The Railway Dyke
Crispin Lane – Wat’s Dyke – Wrexham De191 – Primary Reference Number (PRN) : 106675
I confess I have cycled past this section of Wat’s Dyke for 10 years, but because it’s heavily overgrown, and it isn’t publicly accessible, I wasn’t precisely sure where it lay. So on my cycle I hadn’t realise Wat’s Dyke was lurking on the other side of the fence in Network Rail-owned land beside the railway line! Walk down Crispin Lane and you can catch glimpses of it in winter through the fence! The Archwilio entry reads:
Two sections of Wat’s Dyke bounded to the west by Crispin Lane, which probably overlies the original western ditch. Large, impressive and well-preserved stretches of upstanding dyke. Northern section (SJ33015128 to SJ33005122) some 80m long, 10m wide and 1.5m in height. The eastern face and top of the monument are obscured by dumped material, probably associated with the construction of the adjoining railway sidings. Southern section some 75m long and 8m wide, height increasing from 1m at southern end (SJ32975111) to 2m at northern terminal (SJ32995117). For approximately 60m of its length, the eastern face has been artificially scarped, possibly as a result of the construction of the adjoining railway sidings (Cadw, 2003).
In this area, the Dyke is running over flat/modestly sloping down before it descends to the Clywedog.
4. The Premier Dyke
Premier Inn, south of Wrexham Railway Station – Wat’s Dyke – Wrexham De191 – Primary Reference Number (PRN) : 106675
Park in the Premier Inn car park, or else walk in from the Mold Road, and between the new hotel (2012) and the 19th-century railway line is a reconstructed section of Wat’s Dyke, displayed but without signage, and relocated following archaeological excavation (Grant pers. comm.).
5. The Cambria Dyke – Wat’s Dyke – Wrexham De165n – Primary Reference Number (PRN) : 106678
Named by me the Cambria Dyke because of its proximity to Coleg Cambria. Park at Morrisons or beside Bellvue Park, and walk along Ruthin Road. From the east side of the rise where the road crosses the railway bridge, one can look down over a property boundary between allotments and houses (to the east of it) and a public footpath heading south to Bersham Road, with Coleg Cambria’s car park on the western side. This is Wat’s Dyke! The footpath follows the former line of the ditch. At many places the bank has been damaged and disturbed by 19th-century walling and 20th-century fencing, but just south of Ruthin Road the bank is almost 2m high.
In this section, and Wrexham cemetery, the dyke is approaching the Clywedog valley, so that it hits a prominent point before negotiating its norther side.
Park in or beside Wrexham cemetery and visit Wat’s Dyke within its grounds! The ditch has gone but the bank survives as a low earthwork running south from Bersham Road to the top of the hill before the cutting created by the Ruabon Road breaks it off.
The dyke lacks a separate PRN within the cemetery grounds, where it is discernible, if covered with 19th-century graves. Indeed, this is the only place where its edges are marked by heritage lines in the two driveways in the cemetery that bisect its line. Notably, the Welsh is on the ‘Mercian’ (east) side’, the English on the ‘Welsh’ (west) side of the monument!
Early medieval-inspired Victorian gravestone in the form of a ‘Celtic’ (i.e. Hiberno-Norse wheel-headed) cross
I was struck by the large number of early medieval-inspired graves of late 19th century date overlaying the early medieval monument! The funerary landscape incorporating such an important ancient monument is a rare thing.
This section of the Dyke is only separated a short distance from the Cemetery Dyke (6) by the cutting for the Ruabon Road. Indeed, one can see the Cemetery Dyke in profile from the location where the Court Wood Dyke begins.
To visit, walk in from Wrexham along the Ruabon Road. Alternatively, use the NT car park at the Woodland Classroom off Hafod Road, and take the path over the stream and up into Court Wood, taking the left-hand fork in the path.
Again, it is important to note the cluelessness of the National Trust signs, which mention the Wat’s Dyke Way, but don’t give visitors an explanation of where and what Wat’s Dyke was.
Variously positioned under or within back property boundaries, a substantial and significant section, recently cleared of vegetation by NT volunteers, can be viewed in two prominent stretches.
Once upon the ridge, one can discern the traces of the monumental bank and revealing ditch
At the south-east end of the ridge, Wat’s Dyke is lost as it descends towards the valley and towards Erddig Castle. Presumably it once crossed the valley floor and I think there are subtle traces of it doing so.
8. The Big Wood Dyke
Wat’s Dyke – Erddig Big Wood, excavation 1985 – Primary Reference Number (PRN) : 17766
Park at Erddig National Trust car park or walk into the estate from Wrexham or from the NT car park at the Woodland Classroom off Hafod Road, one can see a significant section of Wat’s Dyke within the woods between Erddig Hall and Erddig Castle. See my discussion here. and here, and also here.
An evening presentation and update on the latest revelations from the archaeological investigations in the Lancaut Peninsular & Spital Meend fort: Wednesday 28th November 7pm (doors open 6.30pm) Drill Hall, Chepstow, NP16 5HJ
Find out about the final results from the small partnership project that has conducted walkovers, geophysical surveys and targeted excavations around the medieval church yard at the Lancaut DMV and in the promontory fort at Spital Meend, Tidenham, Gloucestershire. The questions the project hoped to shed light on included whether Lancaut was a Celtic (British) / Early Christian monastic complex, how and when Spital Meend fort was occupied and used, and how these related to Offa’s Dyke. The partnership is between the Forest of Dean Building Preservation Trust, who own Lancaut Church, Historic England, Gloucestershire Archaeology, Herefordshire Archaeology, Offa’s Dyke Collaboratory, Southern Marches Archaeological Practice Ltd., Wye Valley AONB Partnership and local landowners, with funding from Heritage Lottery Fund and Gloucestershire Environmental Trust.
The free evening event is on Wednesday 28th November at the Chepstow Drill Hall, 7pm. Doors open 6.30pm, light refreshments available.
Thanks to the staff and logistics of CPAT (notably Dr Penelope Foreman) with additional financial support from the University of Chester, the fourth Offa’s Dyke Collaboratory event successfully took place at the Marches School, Oswestry, Saturday 15th September. This follows on from 3 previous events:
This fourth event aimed to bring together the convenors and members, focusing on identifying the rationale and achievements of the Offa’s Dyke Collaboratory, together with reports on local groups up and down the Anglo-Welsh borderlands. The event was combined with a field trip in the afternoon to see CPAT’s dig on the Chirk Castle estate.
The event was attended by c. 50 individuals, including some celebratory figures of Welsh and Borders archaeology – including Margaret Hill and Sian Rees.
The morning comprised of a series of talks by convenors and members. Together these presentations show how, 17 months since inception, the Offa’s Dyke Collaboratory is making real progress in building momentum for new research on the linear earthworks of the Anglo-Welsh borderlands. As well as fostering new work on dykes’ design and construction, new work is afoot to explore their landscape contexts and biographies of use and reuse.
The convenors’ talks were by
Dave McGlade of the Offa’s Dyke Association
Dr Paul Belford – Clwyd-Powys Archaeological Trust
Andrew Blake – Wye Valley AONB
The second half of the conference comprised of talks by local groups introduced by Dr Keith Ray:
Alan Brown – Caer Alyn Archaeology
Niall Heaton – Trefonen Rural Protection Group
Mel Roxby-Mackey and Ian Mackey – CoSMM (Community Stewardship for Mercian Monuments)
Ray Bailey – Offa’s Dyke Collabatory North
Dick Finch – Offa’s Dyke Collabatory Gloucestershire
After lunch and a chance to chat with the audience and fellow-speakers, we re-convened in the National Trust car park at Chirk Castle for a brief sermon from the mound by Dr Paul Belford, before visiting the CPAT dig.
Before we dispersed, there was a brief opportunity to get a photo of 5/6 of the Collaboratory convenors (L-R): Dr Keith Ray, me, Dave McGlade, Dr Paul Belford, Andrew Blake.
A Research Network for Offa's Dyke, Wat's Dyke and Early Medieval Western Britain