The Arthurian Battles (when confined to the North)
Of course, we can default to L. Artorius Castus, the 3rd century Roman prefect of the Sixth legion at York, who may have led legionary vexillations against the Maeatae and Caledonii confederations during the Emperor Septimius Severus' invasion of northern Britain. But if we want to keep Badon and Camlan in the mix, we must have our sub-Roman/Dark Age chieftain.
And, as I never get tired of reminding my readers, the list could be a composite one, made up of the exploits of several Arthurs. I remain very unhappy with the far Northern battles when it comes to seeing them as belonging to a Dark Age chieftain based on the Wall or in Northern England. They look to be either battles belonging to Aedan son of Dalriada or L. Artorius Castus. Only Castus really works for those battles plus the southern ones, minus Badon. If Castus fought at Camboglanna (which is quite possible, as Septimius Severus and Caracalla are attested there), he obviously didn't die there. The idea that an Arthur based in northern England or on the Wall might have been fighting for or in alliance with the Gododdin against Picts is not impossible, but also not particularly attractive.
I'm quite content letting go of Uther Pendragon, a.k.a. St. Illtud, for the reasons I've discussed in the past (more recently here: https://mistshadows.blogspot.com/2026/08/the-terrible-ghost-of-illtud-soldier.html). I think it is highly likely that Geoffrey of Monmouth or a predecessor had merely adopted the Uther name for the sake of poetic assonance or because a splendid story could be made out of Uther's epithet Pendragon. It is important to note, however, that if this were done it could only be because the father of the famous Arthur was not known. Had it been known (and, again, we don't find Uther attached to Arthur until we reach the Galfridian material), we would either have a record of it in the Welsh poems, stories and hagiography or someone post-Galfridian would have objected to the identification and sought to set it right.
If for the sake of argument we take out Uther, and we take out L. Artorius Castus, we are left with a group of men generally referred to as the genuine historical Arthurs. I'm here posting an image of p. 48 in CONCEPTS OF ARTHUR by Thomas Green, wherein these men are described:
The consensus view is that the first of the historical Arthurs is Artur son of Aedan of Dalriada. I have written about him before, most recently in the following piece, where I bring attention to the alternate father provided for him in the Irish sources, viz Conaing, the Irish form of the English word for "king":
The name Arthur, being British, had to come into the Dalriadan royal family via the Britons, and most likely through Strathclyde. It is known that Aedan married the niece of a British king (Rhydderch of Alclud, Aedan's contemporary?) and that he had a daughter named Maithgemma ("Bear" in the Irish). We all know that Artur may well have been associated with the Irish art, 'bear", cognate with the British arth.
Now, as always, our real problem here is dating. The HB and the AC nicely sandwiches Arthur between 488 and 547. The AC provides dates for Badon and Camlan at c. 516 and c. 537. This is way too early for any of the historical Arthurs, and way too late for the early 3rd century L. Artorius Castus. An intentional backdating of Arthur in order to place him at Gildas' Badon is certainly possible (although purists hate the idea, quite naturally!). But if we want to allow for the Northern battles as they appear on the map above (with Badon = Buxton of Aquae Arnemetia and the Bathamgate and Camlan as Camboglanna/Castlesteads), and we must accept the provided chronology, we are forced to seek another solution to the problem.
It would be wonderful if we could accept the dubious testimony of Nicholson and Burns who, in the late 1700s/early 1800s recorded a local tradition that Etterby hard by the Stanwix/Uxellodunum/Petriana Roman fort on the west end of Hadrian's Wall was known as Arthur's burg. But the tradition is quite late, and we have no independent confirmation of such a tradition. Improbable that Etter- was wrongly interpreted as a corruption of Arthur. But, the first element of the Etterby place-name (from OFr. Etard) features in Arthurian romance (see ETTARD in Christopher W. Bruce's THE ARTHURIAN NAME DICTIONARY). Thus we must view this potential power center with some suspicion (although, see more on Petriana below).
I've tried to make circumstantial cases for Birdoswald and Ribchester as well, but both collapsed under a sea of empty suppositions, often based on unreliable Welsh tradition (such as genealogies) and Roman period history and archaeology.
All of these efforts, although they can be made superficially attractive, even compelling on occasion, cannot prove the presence of a Dark Age Arthur in any of the proposed locations.
The real question is: why, prior to Geoffrey of Monmouth, do we not possess a father or a kingdom for Arthur? Why, from the 9th century HISTORIA BRITTONUM on, is he lacking a patronymic?
Why was this Arthur left as a ghost for so long? I've been racking my brain over this for some time. None of the explanations offered by other Arthurian researchers stand up to scrutiny. I can resort to the Dalriadan Conaing being mistaken for just the noun "king", rather than being seen as a personal name. But this idea does nothing to resolve our chronological inconsistency.
Well, I keep coming back to this article:
In that piece, I examined two prominent motifs found in the historical and traditional material relating to Dalriada and Dyfed. The first concerned a prevalent equine association, while the second displayed a propensity for rocks.
Now, we have to be very careful here, for the Ala Petriana would have been withdrawn from Uxellodunum in the early 5th century. It is not unreasonable, however, to assume that some relic or attempted preservation (duplication?) of a martial horse culture might have remained at the fort.
The Gallic cavalry's predilection for the worship of Epona is historical fact, and the Petriana was originally formed in Gaul. Most scholars of Celtic religion think that the Dyfed Rhiannon is a manifestation of Epona. If so, then the name of a great sub-Roman war-leader in the North who traced his descent to a member of the Ala Petriana (named for a Petra) might well have appealed to a chieftain of Dyfed named Petrus. This is especially true given the Eochaid and Artchoirp at the head of the Deissi genealogy.
Similarly, a Scottish group whose tribal name (Dal Riada) and ancestor Eochaid bore names related to riding and horses, who had settled in the Kintyre of the Epidii or Horse-tribe, might also find the name Arthur attractive. Especially if the name came from British Petra Cloithe/Alclud/"Aloo".
Yes, it is true there were other Roman cavalry units stationed in Britain over the centuries. And if we wish to see such a horse connection for Arthur (other than the obvious fact that Castus was of the equestrian class), we could look to those. But if the goddess Epona is significant in this regard, we must confine ourselves, it seems to me, to Gallic groups - like the Ala Petriana.
We would not be talking about the Sarmatian horse at Ribchester, for example.
Sheppard Frere nicely sums up the strategic importance of the Petriana fort in his Britannia: A History of Roman Britain:
“The western sector of the Wall was the most dangerous… both on account of the nature of the ground and because of the hostile population beyond it. It is not surprising to find, then, that at Stanwix near Carlisle was stationed the Ala Petriana… Such regiments are always found on the post of danger; and the prefect of this Ala was the senior officer in the whole of the wall garrison. Here, then, lay Command headquarters, and it has been shown that a signaling system existed along the road from Carlisle to
York, which would enable the prefect at Stanwix to communicate with the legionary legate at York in a matter of minutes.”
Dr. Roger Tomlin commented on this passage (personal communication):
"Stanwix was certainly the base of the ala Petriana, the only milliary ala in Britain, so its prefect would have been senior to the other prefects and tribunes on the Wall. Its geographical location also is significant, as Frere says."
Without knowledge of a 5th-6th century Northern Arthur's father's name, and lacking a point of origin for our theoretical hero, my placement of the HB and AC battles, along with the shared characteristics of the Dyfed and Dalriadan Irish-descended dynasties impels me to point very tentatively to Petriana as the sub-Roman Arthurian power center.
P.S. I won't lie; I rather like the proximity of Petriana to Camboglanna, as well as to Roman Aballava/Avalana (modern Burgh-By-Sands) with its Dea Latis/"Lake Goddess" and Concavata (modern Drumburgh), a possible prototypical Grail Castle.
