Guest Author:
Neil McDougall

Vortigern Studies Index

.Wansdyke Project 21
is part of
Vortigern Studies

Abstract

This paper explores the possibility that the Roman road did indeed form a part of the Wansdyke scheme, and in a second part looks at one of the key areas where this theory can be proved or disproved - the east bank of the Avon, east of Bath and north of Braford-on-Avon. It is at this point where archaeological remnants of a connection should be looked for between West Wansdyke (which ends at the Horsecombe vale) and the Roman road, which rises from Bathford and crosses West Wiltshire towards the Avon south of Laycock and then meets East Wansdyke at Morgan's Hill.

Features in the landscape

North and east of the Avon between Bath and Bradford on Avon, there are a sequence of enigmatic earthworks and other archaeological features. What are they? Quite a few archaeologists have visited the features and I have had no strong suggestions. There are at least sixteen features which at this stage can only be described as a collection of linear and apparently unfinished earthworks, cross-dykes, enigmatic humps, small stockades and the probable remains of Roman features.

A - Features around Monkton Farleigh
When examining a diverse range of archaeological features in landscape, I'd like to follow the approach suggested by Barbara Yorke, who suggested that attempts should be made to place the earthworks and features found on both sides of the Avon valley into a historical context and, if possible, into a historical sequence[1]. It is sensible to start by describing a feature close to Monkton Farleigh on a prominent hilltop overlooking Bath'.

1)  Monkton Farleigh (ST800657)
These are twin linear earthworks which contain stone alignments, they may be elements of a track, a linear boundary or a cross dyke. The alignments were brought to the attention of English Heritage in 1998[2] and their age and function remain undetermined. These alignments run at 90 degrees to the Bath/London Roman road, which supports English Heritage's initial view that the alignments are 'two sides of a trackway'. The obvious question that then arises and, remains unanswered, is why have a track way without a hilltop settlement. These linear alignments run at 90 degrees to the Roman Road. Sian William's argument[3] that they are 'two sides of a track way' is rather facile, as the obvious question (which then remains unanswered) is why have a track way without a hilltop settlement.

The feature from below, with pre-war defenses on top of the hill. Click the images to enlarge.

It would be very sensible to trench across the space between the stone alignments. As demonstrated in the photographs, these are massive deep-set stones which appear to form the outer sections of a wall (i.e., the remaining stones provide the edges of 80cm wide filled wall). The alignments can be linked to other features (as explained below).

The feature from above, with pre-war defenses on top of the hill. Click the images to enlarge.

I discovered an object that may be related to this linear feature. It is a corroded copper alloy clasp with an embossed, garlanded vase decoration. It was found about 6m away from the alignments in a ploughed field. Unfortunately, the object is difficult to date (it may be classical, ie late Roman, or neo-classical, ie early Georgian, in origin) and its' function remains unclear[4].

2) Avon Valley - Sheephouse farm (ST 794640)
An enclosure lies on the east bank of the Avon opposite what I assume was once a fortified position of the manor of Claverton (ST 785640). Both positions overlook the Claverton ford. If the earth bank described in 1) is directly aligned to the track way, this feature could be linked to it.

3) Bathford Hill – Rowbarrow Wood (ST 791657)
A mound of uncertain appearance, but most likely man-made (see picture below) appears on the hillside south of Bathford. If the earth bank described in 1) is directly aligned to the track way, this feature could be linked to it. The mound at Rowbarrow Wood is at 90 degrees to the alignment of the track way, and the enclosure is at the terminus of the track way if the track way follows the natural contour of a dry valley/track which cuts below the Bradford on Avon road at Dry Arch (ST 798644).

Ken Dark[5] described this mound as a possible 'motte', certainly it is an unusual man-made feature which is in an interesting strategic position. The feature is shown by Albany Major[6], and to the east of the mound are 'scoped-out earthbanks' which were once believed to part of Wansdyke, before OGS Crawford discredited that interpretation. According to the current National Monuments Record description[7], "neither of the earthworks (the scoped out-out earthbanks) is of archaeological significance". Another archaeologist suggested that the mound could be the 'row barrow', which gave the wood its name, i.e. a multiple grave!

4) By Brook valley – Bathford (ST 799670)
An enigmatic earth bank may also be related to the linear earthwork described in 1), although we could best discuss it as another unexplained feature (see the two images below). It is possible that it is an embankment associated with the trams that were used to link the underground stone quarries with the railway - however, these are shown on the 1901 OS maps, and were positioned on the County constituency boundary where it crosses the road to Kingsdown (ST 803668).

5) By Brook valley - Bathford (ST795666)
The Roman road rising from Bathford to Farleigh Down (see the two images below). This feature as seen in the pictures might have been created to stabilize the road. One of the archaeologists who were working on the Avon Archaeological Trust's study of the West Wansdyke was convinced that this was a Wansdyke-like feature.

6) By Brook valley – Kingsdown (ST 814674)
A ditch and bank seem to block the northeast approach to Farleigh Down. Running from the escarpment from NE to SW for about three hundred metres, this feature may be even longer yet ploughed down (above). This may well be a medieval field boundary although the modern field boundaries make no use of it.

B - Features
The second part of this list

7) Avon valley – Limpley Stoke (ST 785627)
This is another earth bank of approximately 3m high, that runs for about 60m between the river just north of Dundas Aquaduct and the woodland to the east. Effectively complicating movement up and down the riverbank, the bank is similar in form to 3) above, with no discernable ditch. It is possible that these, and maybe other features, are English Civil War defences. The Royalist forces moved on Bath to take the city, moving via Bradford and Monkton Farleigh (which illustrates the strategic significance of these two locations) in 1643 AD.

8) Avon valley – Limpley Stoke (ST 780627 to ST 790624)
This is a deep track which undoubtedly was used to remove stone quarried close to the ridge. The track is probably older than the quarry (the quarry at ST791623 makes the track unusable to the east), and a westerly extension would take it across Claverton Down to Widcombe (where it would join an ancient track called ‘Aethelburg’s Way’ and then meet the Avon opposite the fortified core of the Roman city).

Interestingly, it is followed by the County boundary. Note also the ‘lost’ Bath to Poole Roman road at ST770595 converging on the valley track leading down to Widcombe.

The track eastwards is a footpath traced from ST 797622 to ST799621, and again from ST812616 to ST818613. The track as this stage is adjacent to the recently discovered villa (see below) at ST819614, and interestingly a large number of field boundaries on this plain follow the alignment of the track.

The probable route of the track can be traced through Bradford on Avon, following the contour to ST837609 where a similarly aligned track takes it past the earthwork at ST846605 (see 16, below) to the Avon.  On the same alignment is Cold Harbour at ST899587 (which is commonly believed to denote a Roman/post-Roman structure associated with a road). Interestingly, this alignment would precisely link together Bath and Andover (a Roman garrison town).

9) Avon valley – Limpley Stoke (ST 781603)
This is the stockade that probably gave Limply Stoke the second element of its name (see the three photos below). Limply Stoke, alongside South Stoke ST748613, North Stoke ST702691 and Englishcombe ST717628, were the earliest Saxon settlements on the fringes of what may well have been a largely abandoned Bath.

The photo below shows the commanding view from this location down the Avon valley to Bradford on Avon. This location would also provide a position for signalling movement down the valley. As far as I know, this feature has not been identified or excavated.

10) Avon valley – Winsley (ST 793611)
Another small stockade, again in a significant strategic position in relation to the contours of the valley-side and the view up and down the valley. Again, as far as I know, this feature has not been identified or excavated.

11) Avon valley – Winsley (ST 805604)
There is another stockade at the junction of two tracks, as seen by the three images below:

12) Avon valley – Bradford-on-Avon (ST 823607)
This may be an appropriate stage to outline what a believe is an ancient track linking together the former hillfort at Budbury and Little Solsbury Hill (ST768680).

My interest in the origins of Wirtgernesburg (which is probably an early-English rendering of Vortigern's burg) and Badon (which is probably an early-British name for Bath circa 550 AD, and the strong source for the identification of Bath as the probable location of the siege of Mount Badon) started as I attempted to identify the age and usage of a track close to the Budbury hillfort. It was very clear that the medieval development of Bradford on Avon centred upon this track, which again links to the earthwork in Great Bradford Wood (ST845605). The track followed the ridge of the north valley side, between the modern A3108 and the road from Bradford on Avon to Turley.

The route of the track westwards can be traced as it emerges from Belcombe Court as a bank which crosses the 80m contour at ST813606, from here it probably follows the modern road through Turleigh, past 11, above, through to Murhill (again on the modern road), through Conkwell Wood and along the modern road to ST797636 where it follows the footpath to the important crossing point of Dry Arch (ST797644).

From this point ST750658, the track was marked with a sequence of upright stones[8] (one of which remains).  A similar marking stone was present at the Bradford on Avon end of the track (at ST821607) in the late 19th century. This indicated, perhaps, that the track was both ancient and significant.

From the point at ST795658, the path divides, one path veers to the right at a right angle.  This was probably a path used by the Clunic monks to walk between the monastery at Monkton Farleigh and the Abbey in Bath. Another path falls down the valley side through Bathford, and connects with the Roman road at ST875671.  A simple approach can then be made to Little Solsbury Hill (ST768680), following the track from ST 774675 to the hilltop through the hillfort gateway.

As Major and Burrow explain[9], this track marks the boundary between Wiltshire and Somerset, which gives the track a further strategic significance.  It is also possible that the track forms a link between the Jurassic Way (thought to be an ancient track from Bath to Leicester used prior to the construction of the Fosse), and portions of the Ridgeway which reach down to Warminster. The eastern route of this link probably follows the line of the River Biss from its junction with the Avon, to the east of Bradford on Avon, and then following the line of the A350 which takes the high ground to Westbury/Warminster.

Another piece of evidence of the significance of the highland overlooking the Avon and By Brook valleys, is an early English ‘military’ road identified by Grundy[10]. He traces this 'military road' from very early Charters, as running from Kilmington Common (ST770360) via Mapperton Hill through Maiden Bradley, Horningsham and to Cley Hill (ST840450), then to Chapmanslade, then Beckington, then Rode, and then along the route of the modern B3109 to Bradford on Avon. After following the route of the modern A363 towards Bath, the ‘military road’ veers off, at Farleigh Wick through Monkton Farleigh, following the northern road through the village as far as Kingsdown (ST811670). Here, the ‘military road’ follows an odd loop along the contour of the ridge northwards to ST814680 where it turns sharply and runs in a south-easterly direction through Blue Vein at ST673830, to Fiveways at ST671839, and then along the route of the modern B3109 to Corsham, and finally through Biddestone ST86735 and ultimately to Malmesbury.

Interestingly, the road diverts away from the direct route between Bradford on Avon and Corsham (the modern B3109), but makes a very unusual sortie around Kingsdown which overlooks the confluence of the Avon and By Brook valleys and the city of Bath.

13) Avon valley – Bradford-on-Avon (ST 821614)
The Roman villa (or possibly villas, as they may well be two significant buildings) is positioned in the grounds of St Laurence School. The building has been surveyed from crop marks which are very distinct and from geophysical surveys. The villa/s is/are substantial and probably were occupied by high status individuals. Apparently, there were twin towers, which may have been used to provide a view across the Avon valley. Significant fragments of high quality stone work (some with engraved Roman numerals) have been discovered nearby.

14) Avon valley – Bradford-on-Avon (ST 823609)
The Iron Age promontory fort at Budbury was extensive, making use of the steeply sloping valley side north of the Avon. Geoffrey Wainwright, who until recently was the Senior Archaeologist at English Heritage, excavated the fort in the late sixties, and found firm evidence of ‘Dark Age’ usage[11].

I’d imagine this to be the site of Wirtgernesburg, i.e. the location of the battle described by William of Malmesbury (who dated it to 652AD). It would have been an important defensive position at any time when opposing factions held Wiltshire, the Cotswolds and Somerset. If there was civil war between Vortigern and Ambrosius Aurelianus and the former (as Morris suggests) held the Cotswolds while the latter held Wiltshire (as tradition asserts), then Bradford on Avon would be the south eastern stronghold of the Cotswolds at the entrance to the Avon.

As explained above, the Saxon and Medieval development of Bradford on Avon can be traced along the route of the track described in 12) above, and it would be possible that Wirtgernesburg would survive as a separate and distinct name which distinguished the hillfort from the developing town! There may well have been strong connections between this site and 13) above.

15) Avon valley – Bradford-on-Avon (ST 842611)
This is a stockade; again this has not been recorded or excavated.

16) Avon valley – Bradford-on-Avon (ST 845605)
Finally, there is the earthwork in Great Bradford Wood. This is probably connected with the track described in 12) above, and a possible Roman road river crossing (see 8, above). Robert Vermaat has identified this witha possible location for Wirtgernesburgh[12].

There are similar defensive positions on either side of the river approaches to Malmesbury, and these may have been constructed in the 10th or 11th centuries, and there is a similar position to the west of Bradford on Avon at ST818606, which probably dates from this period.

Conclusion

Alcock observed thirty years ago[13], that, 'it is of course typical of our period (367 - 634AD) that one of its major archaeological monuments (Wansdyke) should defy attempts to date it, to set it in its political context, or to understand its tactical function'. 

Robert Vermaat and Neil McDougall have independently pursued geographical approaches to address slightly different questions. Robert sought to identify connections between known sections of Wansdyke, whilst Neil was examining the geographical and historic connections between 'Dark Age' Bradford on Avon and Bath. These independent inquiries have reinforced the claim, that the segment of land to the south of the Roman road between Bathford and to the south of Laycock, and bounded by the curve of the Avon, with Bradford on Avon the south-eastern-most point, was one of the most strategically valuable pieces of land in South West Britain.

This claim is partially dependent upon the case that one of the hilltops surrounding Bath is the most likely location of the siege of Mount Badon, circa 500AD, (and the Monkton Farleigh hilltop and the hillfort at Budbury are possible locations)[14].

However, leaving aside the search for convincing evidence for the most important of all Arthurian battles, the Bath, Bradford on Avon and the Avon valley retained primary strategic and historical importance, from at least as early as 577 to 1013 AD.

The strategic value of Bath was demonstrated by the coronation of Edgar, in the city in 975 AD. This unified Mercia and Wessex under Edgar's brief, but successful kingship. Bath's symbolic value, on the Avon boundary between the previously separate and hostile kingdoms, was again reinforced by the surrender of the West Saxon Ealdormen to Swein Forkbeard, the Danish leader in 1013 AD. This triggered his final ascent to the English throne, after previously accomplishing the surrender of Oxford, Winchester and London. 

Centuries earlier, the defeat of the British by the West Saxons at Dyrham in 577AD, a few miles north of Bath, broke the post-Badon dominance of the Romano-British in the South West. West Saxon settlement followed almost a century later, after they defeated what was probably an alliance of Mercian and British forces, at Bradford on Avon in 652 AD (the Wirtgernesburg battle noted by William of Malmesbury[15]). A second battle of Badon occurred in 665 AD (recorded in the Welsh Annals[16]), this was convincingly described by the Birkitts[17] as a last ditch attempt by the British to regain territory to the south-east of the Severn and, they argue that it must have been in the vicinity of Bath and to the west of Bradford on Avon (which in the Birkitt's view adds further support to a Bath location for the siege of Mount Badon, circa 500AD. 

Despite Wansdyke reaching the southern fringe of the Bath, there is so little evidence of the significance of the city during 'Arthur's Britain' (using Alcock's description for the period of our primary interest) that the Bath/Badon place name association has been denied strongly by Manco[18]. This impasse is mirrored in Wansdyke studies as well. Thirty years have passed since Alcock wrote his tribute to the resilience of Wansdyke to yielding its enigmatic status. Despite advances in archaeological techniques, are we any closer to understanding the date, functions and geopolitical significance of the structures?

The observation of the Bath archaeologist, Peter Davenport[19], that we might be more successful in discovering clues and evidence of Bath's 'Dark Age' history by searching the city's rural hinterland, rather than expecting the fragments revealed in occasional rescue excavations in the city to answer deep and fundamental questions, is probably also true for Wansdyke. Understanding how the gap in Wansdyke worked defensively, might provide insights which help unravel the enigma.   

Amongst the ancient tracks, the last vestiges of Roman occupation, the hillforts and the stockades that served to defend a stronghold that retained Vortigern's name for six centuries after his death, and a city venerated by the British for its hot springs and  celebrated as a 'marvel of Britain', we should be able to prepare convincing archaeological hypotheses and the opportunity to test them to establish whether the features that we identify, and describe, are contemporary with Wansdyke and serve a common strategic purpose.  Here also, we might uncover more of the archaeology of  'Dark Age' Bradford on Avon and Bath'.

Notes:

^[1] personal correspondence: 'looking at the map you sent I am struck by how many earthworks there are in this part of the Avon valley especially on the east bank. There are parallels with the Test Valley in Hampshire. One could postulate a series of territories centred on the Avon each of which had its own earthwork.  There seems to be the potential for an extremely interesting archaeological project that would aim to establish the relative dates and functions of all the earthworks in the Avon valley between Bradford on Avon and Bath, starting in the late Bronze Age when territories seem to have been first laid out, a procedure which might be accompanied by the building of earthwork enclosures. The fifth/sixth century would be one period one would expect to find activity, and the late Saxon period is another one of potential interest when Anglo-Saxon lords were beginning to build ring worked manorial sites for themselves. The program would probably require an extensive field walking campaign and perhaps exploratory excavation through test- pits.'
^[2] "West Country amateur historian discovers prehistoric earthwork on a hill", at: http://www.thenoiseroom.com/archnews/nd/archive/17898.htm (retrieved october 2001)
^[3] personal correspondence.
^[4] Received wisdom is that it is neo-classical. I'm not altogether convinced. The British Museum accepted that it could have been made in the late Roman period but they thought this unlikely.
^[5] personal correspondence.
^[6] Major, Albany F. and Burrow, Edward J. 1926: ibid., p. 77.
^[7] National Monuments Record, Unique identifier: 203326, NMR nr. ST76 NE31:
Earthworks, including as supposed course of the Wansdyke, now discredited, are noted between the river Avon and Bathford Hill - see AO/62/301/2. The supposed Wansdyke is described by Crawford as a 'scoop' and bank in the hill, possibly an old quarry and he states that the 'decayed walls' in Rowbarrow wood do not look ancient. The earthwork between ST 78396566 band ST 78756560n is an old field boundary. In Rowbarrow Wood, between ST 79156509 and ST 79336511 are two sinuous banks parallel and about 20 meters apart. (only the southern bank is published on the 25". They appear to form the edge of a low narrow ridge jutting eastwards from Bathford Hill behind. Stone digging along the middle or spine has produced a broad scoop or hollow leaving the bank like features - some 2.0m high - to each side. Neither of the earthworks is of archaeological significance.
^[8] Major, Albany F. and Burrow, Edward J. 1926: ibid., p. 77.
^[9] ibid.
^[10] Grundy, G.B. 1918; 'The Ancient Highways and Tracks of Wiltshire, Berkshire and Hampshire', in: Archaeological Journal vol. LXXV.
^[11] Wainwright, Geoffrey, 1970: An Iron Age Promontary Fort at Budbury, Bradford on Avon’, in: WANHS vol. 65.
^[12] Vermaat, Robert: Wirtgernesburh, Bradford-on-Avon, in: Vortigern Studies, at: http://www.vortigernstudies.org.uk/artcit/wirtgern.htm (retrieved october 2001)
^[13] Leslie Alcock, 'Arthur's Britain', p. 350,  Allen Lane The Penguin Press, 1971.
^[14] These arguments are addressed in detail in an Arthurnet (http://lists.mun.ca/archives/arthurnet.html) dialogue between Neil McDougall, Robert Vermaat and others, in late December 2000 and early 2001.
(http://lists.mun.ca/cgi-bin/wa?S2=arthurnet&D=0&H=0&O=T&T=1&q=&s=
Badon%2FWansdyke&f=&a=&b=) .
^[15] William of Malmesbury, 'The Kings before the Norman Conquest' Translated by Joseph Stephenson (1856),  p 19, Llanerch Publishers, 1989.
^[16] Historia Brittonum - Nennius: British History and the Welsh Annals, Latin and trans. John Morris, History from the Sources 8, (Chichester 1980).
^[17] Tim and Annette Burkitt, 'The Frontier Zone and the Siege of Mount Badon: A review of the evidence for their location', pp. 81 - 93, 'Somerset Archaeology and Natural History, vol 134, 1990.
^[18] Jean Manco, 'Saxon Bath: the legacy of Rome and Saxon rebirth', Bath History (1998), also at http://www.buildinghistory.org/bath/saxon/.
^[19] Director of the Bath Archaeological Trust.

Features in an old Landscape is ©2001 Neil McDougall

Comments to: Neil McDougall

VortigernStudies and Wansdyke Project 21 are copyright © Robert Vermaat 1999-2007.
All rights reserved