The Mammoth Book of King Arthur

by Mike Ashley

  Such speculation, however, takes us no nearer to identifying the location precisely. There are certainly many who do not believe that Badon does equate to Bath and presume that Badon Hill is a specific location. Kenneth Jackson demonstrated that if Badon was the site of a hill fort, and therefore known as Din Badon to the British, it would convert into the Saxon Baddanbyrig, evolving into the English Badbury. Badbury in Wiltshire, Badbury Rings in Dorset and Badby in Northamptonshire are all recorded as Baddanbyrig in tenth century records. There is also Badbury Hill in Oxfordshire, near Faringdon.

  All of these sites except Badby have Celtic hillforts associated with them. That at Badbury, in Wiltshire, is now called Liddington Castle, a name I shall use to avoid confusion. Although an interesting case could be made for each of the sites, Liddington seems the most suitable by virtue of its location. It stands just above the Ridgeway, the ancient British trackway that runs through the Chilterns to the Marlborough Downs, towering above the neighbouring land. Over 277m (900 feet) at its highest point, it had a strong vantage point over the neighbouring territory, and was within sight of other major hill forts, including Barbury Castle. Liddington stands as the frontal defence against a northern or eastern attack, with its back line of defence at the Wansdyke. It allows for greater flexibility than Solsbury Hill at Bath, which would not only have to have conceded a significant Saxon advance, but is also a highly restricted site.

  Like Liddington, Badbury Rings in Dorset is an open site, and would have been a primary focus for any Saxon advance from the coast around Poole Bay. However, it is not a focal point for a major breakthrough. Badon was decisive because it repulsed the Saxons in their advance into the west. The Saxon target would have been the rich territories of Cirencester and Gloucester, and their advance would have been either from the south, where Aelle had established his base in Sussex, or from the east, around Lindsey. Aelle’s base after 491 was at Pevensey, but there were no major Roman roads in that area. So if Aelle were to strike toward Cirencester he would have had to move along the coast to Chichester, then follow the Roman road to Silchester and from there to Cirencester. That road goes right past Liddington. This is far more likely than working all the way along the coast as far as Badbury Rings, and then striking north for Bath along the Ackling Dyke, with the intention of taking the Fosse Way up to Cirencester.

  A Saxon advance from Lincoln towards Cirencester would have taken a direct route along the Fosse Way. However, the ASC tells us that though the Saxons had been making steady territorial gains in the south, and archaeological evidence reaches the same conclusion for East Anglia, Lindsey and further north along the coast, central Britain was untamed territory. The Angles would have needed to make far more gains towards the Midlands before risking an assault on the golden lands of the Cotswolds. The alternative would have been to march from East Anglia down Icknield Street to Verulamium (St. Albans), and then either follow Ackeman Street to Bicester and then to Cirencester, or follow the Ridgeway to Swindon and up to Cirencester. The latter route would, again, take them right past Liddington. There are no other logical routes that would take them past the alternative sites of Badbury Rings or Badbury Hill, and certainly not to Badby in Northamptonshire.

  It would of course have been possible for the Saxons to sail round the coast and up the Severn Estuary, a daring tactic considering the treacherous currents around Land’s End, though one of which they were capable. Then they would either strike directly at Gloucester or divert along the Avon towards Bristol and Bath. But why go to such lengths when they could have marched to Bath from the south anyway?

  Not everyone accepts that Liddington is Badon, or that Badon need necessarily be in the south. Alternative suggestions run from Dumbarton in Strathclyde and Bowden Hill in the Lothians, to the Wrekin in Shropshire or Caer Faddon near Welshpool. Dumbarton is a difficult one to accept if we believe that the battle was between the British and the Saxons in the 490s. The Gaels, the Irish Scotii of Dál Riata, called it Din Brithon (the “Fortress of the Britons”), certainly not Din Badon. In any case Gildas, who allegedly came from this area, would call it by its British name, Alclud.

  Bowden Hill, near Linlithgow, relies on little other than the similarity of the name and the fact that, like Bath, it is on a River Avon, albeit six hundred miles away. In 1710, the antiquarian Sir Robert Sibbald identified it in his Account of Linlithgowshire as having been the site of a major battle, and thereafter fancy took over. There is another Bowden in the Scottish Borders, a village on the southern slopes of the Eildon Hills. Since a possible site for the previous battle at Agned is also in the Eildon Hills, and since this held the major British hill fort in the area, it begs closer inspection, although it would seem strange for such a notable battle to be named after Bowden and not Eildon.

  There is also a Bowden in County Durham, between the towns of Willington and Crook. It has all but vanished today, and the location is only worth noting because it is close to Vinovium/ Binchester.

  Caer Faddon is the locale for Badon given in the Celtic tale The Dream of Rhonabwy (see Chapter 8) but its tradition as the site for a key battle may have derived from later conflicts.

  The Wrekin is championed by Frank Reno, who also draws upon The Dream of Rhonabwy, but interprets the directions differently. The Wrekin was a major hill fort outside present-day Telford, near the old Roman town of Viriconium. This was the fourth largest town in Roman Britain and continued to be occupied, in various stages of disintegration and repair, well into the seventh century. What is significant about Viriconium is that it had a major set of baths which almost certainly survived into the seventh century. Indeed, part of the outer wall, known now as the Old Work, is still standing after 1800 years. Though it seems scarcely creditable that Gildas would refer to Viriconium as Badon, it is possible that in his delight for word-play he would nickname the Wrekin as the Hill of the Baths. He may also have been alluding to the Breidden Hills, one of the probable sites for Caer Faddon, which can be seen from the Wrekin.

  Another suggestion is Mynydd Baidan in mid Glamorgan, south of Maesteg. Alan Wilson and Baram Blackett suggest that the name baidan is derived from the Celtic for “to dare”, which is beiddio in modern Welsh. North of Mynydd Baidan is Maescadlawr, which they translate as the “area of the battle field”. They believe this may be the site of the second battle of Badon in the year 667, but it’s one worth considering for the original battle.

  For completeness I should mention Laurence Gardner’s suggested site at Dun Baetan, near Carrickfergus in Ulster. In Bloodline of the Holy Grail, Gardner refers to the conflict between the Scotii kings of Dál Riata and their Irish overlords. In their battle for independence, the Scots defeated the Irish at Dun Baetan in 516, but were defeated there in 575. It was this second battle, according to Gardner, at which the young Artúir mac Aedan was present. Despite the internal logic and consistency of Gardner’s argument, there is an inherent problem in accepting that the British would celebrate a victory in Ireland by Irish settlers in Britain, and Gildas specifically states that the victory at Badon was against the Saxons, not the Irish.

  Before plotting the locations, let us turn to Arthur’s final battle.

  2. Camlann

  Camlann is not included in Nennius’s list. This may be because the original compiler did not want to sing of a defeat but of Arthur’s victories, ending at the triumph of Badon, or because the original list was compiled during Arthur’s reign, and therefore before Camlann. Its absence from the list is not necessarily critical, although it will inevitably raise doubts about whether it was fought by the same Arthur who fought the others. Curiously, Camlann is not mentioned anywhere else by Nennius, nor is it referred to by Gildas. Its first appearance is in the Welsh Annals under the year 93 (539AD), the year “in which Arthur and Medraut fell.” It also appears in several of the Welsh Triads, where the clear implication is that it arose out of a quarrel between Gwenhwyfar, Arthur’s queen, and Gwenhwyfach, Gwenhwyfar’s sister and the wife of Mor
dred, and, in the way of such things, a quarrel led to a battle. Geoffrey of Monmouth typically took it out of all proportion and has Mordred abduct and seduce Gwenhwyfar and seize the kingdom while Arthur is away. The Triads regard it as one of the “Three Futile Battles”, emphasising that it mushroomed out of nothing. This has the feel of authenticity, a memory of Britain’s greatest hero brought low by a pointless quarrel.

  Geoffrey of Monmouth places the battle at Camelford in Cornwall, based on no more than the name – the river Camel was known as Cambla – and possibly the proximity to Tintagel. The bridge over the Camel here is known as Slaughterbridge, though this probably refers to a battle between the British and Saxons during the reign of Egbert of Wessex in 823. In fact, the name may not refer to a battle at all as it could be derived from the old Saxon word slaggy for muddy, as in Slaggyford in Northumberland. In 1602 the antiquarian Richard Carew, one time High Sheriff of Cornwall, developed the Arthurian connection in his Survey of Cornwall by identifying a stone near the Camel as being the spot where Arthur died. This stone, however, bears the inscription Latini ic jack filius Mogari, recording the burial of Latinus, son of Mogarus, and was probably brought to the site years before to form part of a footbridge across the river. Although many Arthurian legends have developed in this area, it is difficult to find any basis for them.

  The word Camlann means either “crooked bank” (cam glan) or “crooked enclosure” (cam llan), a phrase which must describe thousands of locations across Britain. The River Cam in Somerset is a likely contender. It is a tributary of the Yeo and flows from the hills near Yarlington to join the Yeo just outside Yeovilton, near Ilchester. En route it passes by the impressive hill fort of Cadbury Castle, long believed to be the original Camelot. Excavations by Leslie Alcock in the late 1960s showed that Cadbury Castle was significantly refortified from 470 onwards, for at least two generations. It was both a defensive fort and an inhabited village right through the Arthurian period. If it were occupied by Arthur then Camlann may have been fought right on his doorstep. The Cam twists through a vigorous series of bends about a kilometre away at Sparkford.

  In 1935 O.G.S. Crawford proposed that the name was originally Camboglanna, a Roman fort along Hadrian’s Wall at what is now Birdoswald. It is certainly true that here the river Irthing twists its way around the site in a very crooked glen but, as Geoffrey Ashe has highlighted, the name Camboglanna, in evolving towards Camlann, would for centuries have been known as Camglann, and that ought to be how it is recorded in the Welsh Annals and any other near-contemporary sources. The fact that every source records it as Camlann suggests a much older name. Nevertheless, Camboglanna has another connection of interest. Some thirty kilometres to the west, at what is now Burgh-by-Sands, was the fort of Aballava, which became Avalana by the sixth century. The legend has Arthur taken to Avalon to heal his wounds after Camlann. Intriguingly, there are dedications at several of the forts along Hadrian’s Wall, including Aballava, to Latis, the goddess of lakes and water. August Hunt has suggested she may be the basis for the Lady of the Lake legend.

  Another northern site frequently suggested is Camelon, near Falkirk in the Lothians, just north of the Antonine Wall. It has also been suggested as the original for Camelot, most recently in David Carroll’s Arturius, A Quest for Camelot. Laurence Gardner has Artúir mac Aedan fight at both Camelon and Camboglanna, in his battles against the Picts. Camelon was a significant Roman fort which had been strengthened in the 140s at the time the Antonine Wall was built. The Roman name of the fort is no longer known, although the village that grew up around it gained the British name of Caermawr (“Great Fort”) so is unlikely to be confused with Camelot. It is difficult to know when it was abandoned, because the site was substantially robbed and subsequently built over. The presence of a nearby Romano-British temple, now called Arthur’s O’en, suggests a stable period of occupation. Nevertheless, the fort was almost certainly abandoned by the mid third century and steadily fell into ruin.

  There are several locations in Wales still called Camlan today. Two of these are near Dolgellau, near the village of Mallwyd, a name which may mean “battle ground”. This is the area advocated by Blake and Lloyd in Pendragon and by Alan Wilson and Baram Blackett in Artorius Rex Discovered. It is close to other locations connected with Vortigern and Ambrosius, and to several of the suggested sites from the battle list, especially the rivers Glen and Dubglas. It is also close to one of the suggested sites for Llongborth (discussed below), showing that it might fit a pattern of struggles within the Welsh princedoms. There is also a stream called Afon Gamlan just north of Dolgellau, emphasising how common the name is in the area.

  Another site is also associated with the battle. This is Cwm Llan, a valley on the southern flanks of Snowdon, close to the fort of Dinas Emrys. Peter Bartrum draws attention to the legend about this battle as recorded in Y Brython. It tells how Arthur and his men were heading from Dinas Emrys towards the pass over Snowdon at Cwm Tregalan, and met the enemy in Cwm Llan (“the Valley of the Lake”). Arthur was able to push the enemy back but at the top of the pass they were ambushed in a hail of arrows. Arthur was killed and buried where he fell at a cairn called Carnedd Arthur, and the pass is still called Bwlchysaethau, “the Pass of the Arrows”. A steep climb down from the pass takes you to Llyn Llydaw, supposed to be the home of the Lady of the Lake. Nearby is supposed to be Ogof Llanciau Eryri (“The Cave of the Young Men”) in which, rather like the Christian legend of the Seven Sleepers of Ephesus, the seven who survived Camlann are supposed to be sleeping, awaiting their call to fight again for Arthur.

  The Welsh sites are tempting because of the continuity of the name, but we should not forget that the name would have been just as common throughout Britain before the Saxon settlement. Nevertheless, the Welsh sites suggest a link with Arthur of Dyfed, who probably had conflicts with Gwynedd in this region. These sites do not, however, fit comfortably with any for Badon, raising again the question of whether the battles were fought by two different Arthurs.

  3. The Saxon sites

  Before we map out all of the above locations, we need to remind ourselves of the known Saxon battles during the Arthurian period. If Arthur’s twelve battles were all against the Saxons or Angles then, although the ASC was not given to recording defeats, there may yet be some hints.

  We have determined that the Arthurian period ran from about 480–520, and the ASC lists these battles during and around those years.

  477.

  Aelle fought Welsh at Cymenes ora, [who] fled into the wood Andredes leag.

  485.

  Aelle fought Welsh near the margin of Mearcrædes burnam.

  491.

  Aelle besieged Andredes cester.

  495.

  Cerdic fought Welsh at Cerdices ora, which is on or near the coast.

  501.

  Port landed at Portsmouth and killed a noble young Briton.

  508.

  Cerdic killed the British king Natanleod, after whom the land as far as Cerdices ford was named Netley.

  514.

  Stuf and Wihtgar fought the Britons at Cerdices ora.

  519.

  Cerdic fought the Britons at Cerdices ford.

  527.

  Cerdic fought the Britons at Cerdices leag.

  530.

  Cerdic took the Isle of Wight at Wihtgaræsbyrg.

  Only a few of these provide much help, especially since we know that some of the names may arise retrospectively, such as Port and Portsmouth. However, even though the individual’s names may be suspect, the locations are probably more accurate, if they can be traced. Thankfully a few are easier to identify than others.

  Cymenes ora (ora meaning shore) appears in a charter of Selsey Abbey as Cumeneshore relating to a grant of land by Caedwalla, king of Essex, to the abbey in 673. The surviving copy is not contemporary, leading some authorities to term it a forgery, but regardless of whether or not the abbey owned the land, the description must still be accurate. It states that
Cumeneshore was a stretch of coast between Pagham and Selsey Bill, now known as The Owers, much of which has long since eroded away. The Andredes leag is taken as the vast forest of Anderida, the Weald, which at that time densely covered much of Sussex and west Kent. Its western extremity was just north of Selsey, around Midhurst and Petersfield, so the ASC entry does hold together. In 491 Aelle besieged Andredes cester, the Roman fort at Anderitum [Pevensey]. Pevensey, some 80km east along the coast from Selsey, is the same spot that William the Conqueror chose to land nearly six hundred years later.

  These first three entries seem to make clear the spread of Aelle’s territory. The archaeology, however, does not wholly support this. The Saxons made little inroad into what became Sussex, and certainly not in the area around Selsey. One would expect a successful landing there to result in Aelle capturing the Roman town of Noviomagus [Chichester], but not only is there no mention of this, there have been no archaeological discoveries of any fifth-century Saxon sites there. The main area of Saxon settlement in the fifth century was between the rivers Ouse and Cuckmere, and by the sixth century it had expanded westward into the area between the Ouse and the Adur. There must be some significance in the battle of 485 “near the margin of Mearcrædes burnam”. This name has been interpreted as “the river of the frontier agreed by treaty”, suggesting that Aelle had an agreement with the British that the Saxons could settle on one side of a river only. We do not know which river, but it must be either the Cuckmere or, more probably, the Ouse. A major hill fort, Mount Caburn, rises above the Ouse where it is joined by the Glynde Reach. Caburn is one of the few Celtic names surviving in Sussex, originally Caer Bryn: “strong fort”. It seems likely that the original Saxon settlers, despite their initial victory at Cymen’s shore, must have settled between the Ouse and the Cuckmere, where they were guarded on the west by the fort at Caburn and on the east by the fort at Anderida. In 485 the Saxons sought to break out across one of the rivers, probably the Ouse at Caburn. Unusually for the ASC, this is not recorded as a victory. It simply says, “they fought the Welsh”. No doubt they were contained and pushed back across the river and later, in 491, broke out across the Cuckmere and slaughtered the British at Pevensey.