St. Michael's Church, built within the grounds
of the Roman fort at Burgh-By-Sands
It is not my purpose in this chapter to deal with what I consider to be the misidentification of Glastonbury with Avalon. Others have presented a detailed case against the fraudulent claim of Glastonbury as the final resting place of King Arthur, and I added some of my own arguments to my previous book, THE MYSTERIES OF AVALON.
Here I wish to restrict my attention to the only known place in Britain to actually have born the name Avalon prior to the time of Arthur as well as to this place’s proximity to both Arthur’s Camlann at Castlesteads and his possible ruling center towards the west end of Hadrian’s Wall.
Obviously, the possible location of his grave at Avalon is of great interest to anyone seeking to demonstrate the reality of a historical Arthur. Geoffrey of Monmouth’s ‘Insula Avallonis’ or ‘Isle of Avalon’ is held by most Arthurian scholars to be a purely mythological designation - no matter where one chooses to localize it.
From a philological standpoint, the –on terminal of Avallon or Avalon demands an original terminal fronted by a broad vowel. Thus there is a problem trying to equate the word with Welsh afallen, ‘apple tree’, or Cornish avallen. This problem can be overcome in two ways: 1) by evoking an attested Continental place-name, e.g. Aballone, modern Avallon, in France or by 2) allowing for the possibility that the plural form of Welsh afal, afalau, cf. Cornish avalow and Breton avalou, at some point underwent a fairly common miscopying of u/w as n.
As it happens, the only known site in all of Roman Britain to bear an ‘Avalon’ name is the Aballava fort at Burgh-By-Sands, 5. miles west of Stanwix on Hadrian’s Wall. This fort is under 14 miles west of Castlesteads. The name Aballava is found listed in the various early sources in the following forms:
Aballava – Rudge Cup and Amiens patera
Aballavensium – RIB inscription No. 883
Avalana, Avalava – Ravenna Cosmography
Aballaba – Notitia Dignitatum
It is the one spelling in the Ravenna Cosmography that stands out here. The v of Aballava/Avalava has been rendered as an n, yielding the spelling Avalana. This is exactly the type of spelling we would need to end up with Geoffrey of Monmouth’s Latinized Avallonis.
The Celtic derivational suffix –ava of Aballava, British *-aua, is now found in the –au of Welsh, giving as a meaning for Insula Avallonis ‘Island of the Apple-trees’.
An Arthur who fell at Camlann/Camboglanna at Castlesteads could easily have been carted along the Roman road or brought down the river system in this region to Burgh-By-Sands.
Camboglanna is on the Irthing, a tributary of the Eden River. The Eden empties into the Solway Firth very near Aballava/Avalana.
Two dedications to a goddess Latis were made at the Birdoswald Roman fort, 7 miles east of Castlesteads, and at Aballava. The first (RIB 1897) is addressed to DIA LATI and the second to DEAE LATI. Latis comes from a British root similar to Proto-Celtic *lati-, ‘liquid, fluid’, and Proto-Indo- European *lat-, ‘wet’. Some authorities have seen in her a goddess of beer (cf. Old Irish laith, ‘ale, liquor’), but here she is manifestly a goddess of open bodies of fresh water, i.e. she is a literal ‘Lady of the Lake’. Burgh-By-Sands was, in fact, surrounded by vast marshlands. Although these lands have long since been drained, the area is still called ‘Burgh Marsh’. We can be fairly certain, then, that the Avalon fort was on an island of sorts, the true ‘Insula’ of Geoffrey of
Monmouth’s apple-tree Otherworld.
Topography dictated the position of the Aballava fort. There was an important crossing of the Solway at Burgh and the existence of this crossing may have influenced the siting of the Roman fort here. The fort sits atop a low hill on the highest ground at the east end of the village. The church sits within the south-east corner of the fort and is partly built of Roman stones. The modern road lies on the line of the Wall. Burgh is one of the least explored and understood of all the forts on the Wall. Although earlier visitors presumed a fort here, no remains were visible.
Excavations north of the church in 1922, when a new burial ground was formed, resulted in the location of the east wall, 6-7 ft thick, with an earth backing, and the east gate of the fort, with a road leading out. Within the fort, stone buildings running north-south were interpreted as barracks-blocks. The Roman levels and buildings were all badly preserved.
The sketch plan of the site prepared on the basis of these discoveries suggests a fort measuring 520 ft north-south by 410 ft east-west, giving an area of nearly 5 acres. Excavations on several occasions between 1978 and 2002 south and east of the fort has led to the discovery of buildings, presumably of the civil settlement. The bath-house, south of the fort, was destroyed in making the canal, itself replaced by the railway line, now also abandoned. Further south, the tombstone of a Dacian tribesman may indicate the location of the cemetery. Recent excavations have failed to clarify the location, size and date of the Wall fort at Burgh. We do know the stone fort lay astride the Wall, but the Wall ditch was infilled and re-cut before it was constructed. It is possible that the fort to the south of the Wall at
Moorhouse was retained for some time before being succeeded by a replacement astride the Wall.
As stated above, the actual Roman period cemetery at Burgh-By-Sands/Aballava is said have been to the south of the fort. When I enquired about the tombstone of the Dacian tribesman found in this cemetery, Tim Padley at the Tullie House Museum in Carlisle informed me of the discovery of two other fragments. All three are listed in the Roman Inscriptions of Britain as follows:
IVL PII... TINVS CIVES DACVS
2047 (tombstone) D M S
2048 (tombstone) VII
Alas, according to Mr. Padley, the placement of the cemetery to the ‘south of the fort’ puts it, in his words, ‘near the vallum, possibly destroyed by the canal and railway.’
The tombstone fragments were in the care of Tullie House when they disappeared.
While it is impossible to know whether Arthur was buried in the Roman period cemetery of the Aballava fort, this cemetery must remain a primary candidate for the location of his grave.
In a section of my book THE MYSTERIES OF AVALON, I included the following note detailing one of the supposed sites for Arthur’s grave. As it happens, this tradition matches the one that places Camlan on the Afon Gamlan in NW Wales.
A Note on Northwestern Wales as the Site of Arthur’s Grave
There are a few Camlans/Gamlans in northwestern Wales or Gwynedd. The presence of these sites has prompted various Arthurian scholars to propose that Arthur fought his last and fatal battle in this region. The modern champions of this notion are Steve Blake and Scott Lloyd, whose book PENDRAGON: THE DEFINITIVE ACCOUNT OF THE ORIGINS OF KING ARTHUR, was released in 2003 by Lyons press.
We cannot ignore these Camlans or Gamlans (the most noteworthy being the Afon Gamlan, a river) when searching for a historical Arthur. Unlike the placement of Camlan (or Camlann) in Cornwall, something done by Geoffrey of Monmouth in his HISTORY OF THE KINGS OF BRITAIN, Gwynedd can claim to possess real candidates for Arthur’s final battle site. The only other known site that qualifies linguistically is much further north – Camboglanna on Hadrian’s Wall, which I have discussed above in Chaper 3.
Blake and Lloyd place their trust in a very late medieval source, the VERA HISTORIA MORTE DE ARTHURI, a work dated in extant MSS. to c. 1300, although perhaps to originals dating between 1199 and 1203. According to Blake and Lloyd, the VERA HISTORIA probably was written in Gwynedd. I will not contest this point, as it may well be correct.
The importance of the VERA HISTORIA lies in its placement of Arthur’s interment – and thus of Avalon – in Gwynedd. Although Blake and Lloyd are familiar with the Gwynedd tradition which places Arthur’s grave at Carnedd Arthur near Cwm-y-llan or Cym Llan (an error for Cwm Llem, the Valley of the river Llem), they choose to ignore this bit of folklore and instead settle on Tre Beddau near Llanfair, well to the east on the Conwy River, as the actual burial place of the king. They deduce this from the fact that the VERA HISTORIA states that the grave is near a church of St. Mary (in Welsh, Llan-fair), and that archaeologists have recently uncovered a Dark Age or 6th century cemetery at Tre Beddau.
[Note: Cwm Llan is a very clumsy attempt at rendering Camlan, and is obviously spurious tradition.]
Unfortunately, the authors of PENDRAGON also choose to ignore the description of the burial place of Arthur as preserved in the VERA HISTORIA. In their own words, the burial of Arthur after Camlan is told as follows:
“… the VERA HISTORIA describes the funeral of Arthur as taking place at a chapel dedicated to the Virgin, the entrance to which was so narrow that the mourners had to enter by first forcing their shoulder into the gap and then dragging the rest of their body through the opening. While the funeral took place inside the chapel, a large storm blew up and a mist descended, so thick that is was impossible to see the body of Arthur – which had been left outside, as it would not fit into the chapel. Following the storm the mourners came out to find that the body had gone and the tomb prepared for Arthur was sealed shut, ‘such that it rather seemed to be one single stone’.”
Now, this passage quite obviously DOES NOT portray a 6th century Christian cemetery. Rather, it is a fitting description of a ‘chapel’ comparable to the “Green Chapel’ of SIR GAWAIN AND THE GREEN KNIGHT. In other words, the said ‘chapel’ is a Neolithic chambered tomb, whose passage is so tight as to barely allow the entrance of the mourners.
Furthermore, we are talking about TWO conjoined passage tombs – one that is the chapel of the Virgin, and the other which mysteriously receives the body of King Arthur. In all of Gwynedd, there is only one such ancient monument: that of the double chamber tomb of Dyffryn Ardudwy not far west of the Afon Gamlan.
One of the two chambers of Dyffryn Ardudwy is actually known as Coetan Arthur or Arthur’s Quoit. The “Virgin” is here a Christian embellishment on what would have been a pagan goddess associated with the Otherworld site.
The grave of Arthur discussed in the VERA HISTORIA is thus a product of folklore only. It can thus be dismissed as an actual grave of Arthur.
Granted, we cannot so easily dismiss the Camlans/Gamlans in northwestern Wales. Since writing this, Dr. Jessica Hughes of CADW has sent me information via snail-mail that adds important details to the description of the Dyffryn Ardudwy chambered tombs. To quote Dr. Hughes:
“The Chambered tomb at Dyffryn Ardudwy has been known as Coetan Arthur in the past, indeed antiquarian reports of the site refer to Dyffryn as ‘Coetan Arthur’. However, the name appears to refer to the whole of the monument as opposed to a particular chamber. Interestingly (and maybe somewhat confusingly), one mile to the east of Dyffryn lies another chambered tomb known as ‘Cors-y-Gedal’. This was also known in the past as ‘Coetan Arthur’… Regarding whether there is a church of St. Mary in proximity to Dyffryn Ardudwy, I have found a church 4 miles north of Dyffryn in the village of Llanfair. “
The enclosed Detail Report on this Church of St. Mary states that Llanfair was dedicated to Mary “by at least the 12c when Gerald of Wales and Archbishop Bladwin stayed there in 1188…”
Here is the COFLEIN listing for the second chambered cairn:
“A rather tapering rectilinear cairn, c.31m NESW by 14.5m, showing at its eastern end a number of orthostats, partly supporting a tipped capstone, c.3.6m by 3.0m & 0.45m thick: a spindlewhorl, thought to be IA, is said to have come from under the capstone.”
Both of these chambered tombs are directly west of the Afon Gamlan.