Site of Castlesteads Roman Fort
After his many astonishing victories, Arthur is said to have perished with Medraut at a place called Camlann.
Camlann has long been linked with Camboglanna, the ‘Crooked Bank’, a Roman fort towards the western end of Hadrian’s Wall. The only other candidates for Camlann are in NW Wales (the Afon Gamlan and two other Camlanns near Dolgellau), and these do not have anything to do with the Northern Arthur. For what looks to be a relocation of Arthur to a Camlan in NW Wales, see Appendix III.
Crawford pointed out that the best etymology for Camlann would actually be B. *Cambolanda, ‘Crooked Enclosure’, an utterly unknown name, but Jackson had no problem with the derivation from Camboglanna.
Those who point to Camelon on the Antonine Wall are ignorant of the fact that this place was originally called Carmuirs. It was renamed Camelodunum in 1526 by the antiquarian Hector Boece. He did this because the Camelon fort has been identified with the Colania of Ptolemy and the Ravenna Cosmography. Colania was confused with Colonia or Colchester, itself called Camulodunum.
The Castlesteads fort sits on a high bluff overlooking the Cambeck valley and the break on the mosses to the north-west which carries the modern road from Brampton to Longtown. The site was drastically leveled in 1791, when the gardens of Castlesteads House were laid out and today nothing is visible of the fort aprt from the southern edge of the fort platform, while the view described above is obscured by trees. The Cam Beck has so far eroded the north-west front of the fort that the side gates now lie only 50 ft from the edge of the gorge. From east to west the fort measures 394 ft and it is thought to have been originally about 400 ft square, covering some 3.75 acres, though it is not impossible that the fort faced south rather than north and was therefore somewhat larger.
Excavations in 1934 revealed the east, west and south walls of the fort, the east and west doubleportal gates and south-west angle tower. The gate towers were built one course deeper than the fort wall, whose foundations were the normal
6 ft wide. All walls had been heavily robbed, but roof-tiles occurred in a number of the towers at ground-floor level, suggesting the possibility of oven-bases, as at Birdoswald, rather than collapsed roofs. Space allowed only for the identification of one ditch, 16 ft wide. No contact has been made with any internal building, but an external bath-house was located and partly dug in 1741.
Castlesteads is unique along the whole Wall for sitting between the Wall and Vallum but not being attached to the former; presumably either its pre-existence or the lie of the land dictated its location.
A carved stone dated roughly 466-599 CE was found at Castlesteads. Because in the past the inscription has been read wrongly, i.e. upside down as ‘BEDALTOEDBOS’, this has been considered a corrupt attempt at the divine name BELATUCADROS, altars to whom were found here in a Roman context.
However, I have parsed the inscription as actually reading ‘SUB DEO LAUDIB[US]’, which according to Professor David Howlett of Oxford can be translated as ‘with the accompaniment of praises of God’. Thus this stone clearly denotes a Christian presence at Castlesteads during the time of Arthur.
In fact, this area may have been a Christian center during the generation preceding that of Arthur (see the note on the home of St. Patrick in Chapter 5 below).
The Name Medraut/Modred (= Mordred)
On February 26, 1996, I received a letter from Professor Oliver Padel of Cambridge. This was in response to a query I had sent him some time earlier in which I proposed that the name Medrawt – born by the personage who died with Arthur at Camlann – may represent the Roman name Moderatus. What Padel had to say on this possibility is important enough for Arthurian studies to be reprinted in full below:
“Not much has been done on the name of Medrawt or Mordred… In an article on various words in Welsh with the root med, Medr-, Ifor Williams suggested that the name might be connected with the Welsh verb medru ‘to be able, to hit’; but he did not develop the idea, only mentioned it in passing.
Middle Welsh Medrawt cannot formally be identical with Old Cornish Modred, Old Breton Modrot (both of which are recorded, indicating an original Old Co.Br. *Modrod), since the Welsh e in the first syllable should not be equivalent to a Co.Br. o there.
What people do not seem to have asked is what this discrepancy means: we can hardly say that Welsh Medrawt is a different name, since it clearly belongs to the same character as
Geoffrey’s [Geoffrey of Monmouth] Modredus < Co.Br. Modrod.
Which is ‘right’? I would suggest that the Co.Br. form is the ancient one, and that the Welsh form has been altered, perhaps indeed by association with the verb medru.
That was already my conclusion, but I did not have a derivation for Modrod. However, Modrod would be the exact derivative of Latin Moderatus, as you suggest. Your suggestion is most attractive, and neither I nor (so far as I know) anyone else has previously thought of it.
Like you, I should be relucatant to say that Modrod couldn’t have a Celtic derivation; but it fits so well with Moderatus that I personally don’t feel the need to look further.”
If Medrawt or, rather, Modrod, is Moderatus, this may be significant for a Medraut at Cambloglanna on Hadrian’s Wall, for we know of a Trajanic period prefect named C. Rufius Moderatus, who left inscriptions at Greatchesters on the
Wall and Brough-under-Stainmore in Cumbria (CIL iii. 5202, RIB 1737, 166-9, 2411, 147-51). The name of this prefect could have become popular in the region and might even have still been in use among Northern British noble families in the 6th century CE.
Camlann and the Grave of Osfran’s Son
The purpose of this essay is to prove, once and for all, where Arthur’s Camlann battle site was located. Or, more accurately, where Welsh tradition happen to place it!
It is fairly well known that the Welsh record seven survivors of Camlann. Yet, to my knowledge, no one has sought to plot these personages out on a map. To do so may help us pinpoint a geographical region in which Camlann was believed to be situated.
One of the seven – Geneid Hir – it a difficult and otherwise unknown name. P.C. Bartram (in “A Welsh Classical Dictionary: People in History and Legend up to about A.D. 1000) suggests the name may be corrupt and offers an unlikely identification with a personage named Eueyd or Euehyd Hir (often rendered Hefeydd). However, I would see in Geneid ‘Cannaid’, “white, bright, shining, pure, clean, radiant,” an epithet substituted for the original title Ceimiad, ‘Pilgrim’, of St. Elian. Elian had churches on Mon/Anglesey and in Rhos, Gwynedd.
Sandde Bryd Angel looks to be a pun for the Afon Angell, Aberangell, etc., places immediately to the south of the Camlan on the Afon Dyfi in Merionethshire.
Morfran son of Tegid is from Llyn Tegid, now Bala Lake in Gwynedd.
St. Cynfelyn is of Llancynfelyn in Ceredigion just below the Afon Dyfi.
St. Cedwyn of Llangedwyn in Powys, while somewhat further removed than the rest, is still in NW Wales.
St. Pedrog of Llanbedrog is on the Lleyn Peninsula in Gwynedd, just opposite the three Camlans in Merionethshire.
St. Derfel Gadarn is at Llandderfel near Bala Lake in Gwynedd.
Needless to say, if we “triangulate” with all these names/places, we find at the center the three
So which one is the right one?
Only one way to know for sure: we must find the Camlann that is claimed as the gravesite of Osfran’s son. This reference comes from the ‘Stanzas of the Graves:’
Bet mab Ossvran yg Camlan,
Gvydi llauer kywlavan…
The grave of Osfran’s son is at Camlan,
After many a slaughter…
[“The Black Books of Carmarthen ‘Stanzas of the
Graves’, Thomas Jones, Sir John Rhys Memorial
Lecture, 1967, Critical Text and Translation.]
While –fran of Osfran looks like Bran or ‘Raven’, the Os- does not look at all right for a Welsh name. I suspected Ys- and after a first search failed, I defaulted to bryn or ‘hill’ as the original of –bran. Thus I was looking for an Ysbryn.
And I actually found him – or, rather, it! [See “An
Inventory of the Ancient Monuments in Wales and Monmouthshire: VI – County of Merioneth”, p. 98, RCAHMW, 1921.]
On the Mawddach River in Merionethshire there is a Foel Ispri. It used to be Moel Ysbryn and was the legendary residence of Ysbryn Gawr or Ysbryn the Giant. If we go north on the Mawddach we run into its tributary the Afon Gamlan, i.e. the Water of the Crooked Bank.
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.
A Note on Northwestern Wales as the Site of
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
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
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.
Conclusion Regarding Arthur’s Welsh Camlan
The next question, of course, is what Arthur was it who died at the Afon Gamlan – assuming the
Welsh tradition is historical in nature? It can’t be a Northern Arthur. It could be Arthur of Dyfed, but if so, the 537 A.D. date given for Camlann in the Welsh Annals is a gross error.
There is an Arthur son of Bicoir ‘the Briton’ in the Irish sources, but this may well be Arthur of Dyfed, whose father’s name occurs in a number of variants, including Petuir, which could have become Bicoir. Complicating all of this is the presence of a Dark Age Beccurus stone a couple dozen kilometers NW of the Afon Gamlan at Gesail Gyfarch, Penmorfa. It is tempting to see in this Beccurus the Bicoir (genitive form of the name?) father of Arthur.
The problem is that both Arthur of Dyfed and Arthur son of Bicoir, whether the same man or different, have floruits well after the Badon and Camlann dates claimed for Arthur in the Welsh Annals. The most logical explanation for this is not that the Welsh Afon Gamlan is the wrong location for the Camlann battle, and that the Welsh have relocated the site here from Camboglanna on Hadrian’s Wall.
Numerous such relocations have been found in the early Welsh sources. To cite merely one example, Rhydderch of Strathclyde is said in the “Stanzas of the Graves” to lie at Abererch on the Lleyn Peninsula in Gwynedd.