[Dedicated to my friend Tony Sullivan (who urged me to reconsider my Sub-Roman Arthur)]
Last June I wrote the following piece as a sort of farewell to my dream of discovering a Dark Age Arthur:
After that blog, I pretty much went "all in" with my Artorius theory, one based primarily on my proposed reading of ARM.GENTES for the lacuna of the L. Artorius Castus memorial inscription. I've recently produced an entire book that argues for Castus = Arthur (LET NOT ANYONE ESCAPE FROM SHEER DESTRUCTION,
https://a.co/d/9EDrUev). But there remain some things that bother me about that argument, clever and comprehensive though it appears.
Chief among those things are the ANNALES CAMBRIAE dates given for Arthur (c. 516 and c. 537) and the inclusion of Badon and Camlan in the Arthurian list of battles. And I will get to my thinking on these matters in a bit.
But first I wish to concentrate on two odd characteristics that are present in the earliest tradition concerning the various Arthurs. I talking here, of course, about the prevalence of stones and horses in the sources. I have discussed both motifs before. But I have come to feel that they may be more important than I previously made them out to be.
As far as stones go...
Arthur of Dyfed has as a father one Petrus, a Roman name meaning "rock, stone." Arthur son of Bicoir kills the Irish prince Mongan with a stone. This is in one source referred to as a dragon stone, by which is probably meant obsidian. However, the stone in question may originally have come from a folklore reference to a large monolith at a possible Bicoir site (see
https://mistshadows.blogspot.com/2024/11/arthur-son-of-bicoir-of-kintyre.html). In the 1700s, Stanwix (for Etterby) was known as Arthur's fort (Arthuriburgum), and the huge Roman fort there housed the thousand-strong cavalry force of the Petriana. Some scholars hold that Petrianis of the NOTITIA DIGNITATUM is an actual nickname for the place, not a ghost-name. If so, it is interesting that the Petriana were named for a certain Titus Pomponius
Petra. Petra, once again, is stone or rock. The Arthur name in Dalriada is often thought to have come from the adjacent British kingdom of Strathclyde, with its capital at Dumbarton or Alclud, the Rock of Clyde, called
Petra Cloithe in Adamnan.
Alclud is referred to merely as Aloo, "the Rock", in Irish writings on/by St. Patrick.
And here is where the horses tie in (pun strictly intended). Arthur is first mentioned in the HISTORIA BRITTONUM in the context of Hengist's death. Hengist is a Saxon name meaning "stallion", and I have suggested that Hengist and Horsa (transparently "Horse") are Germanized names formed from the earlier British rulers of Kent like Epillus. The Irish Dessi-founded kingdom of Dyfed is replete with horse-lore (Gwri Gwallt-Eurin, Rhiannon, Manawydan). There is a horse-name (Eochaid Allmuir) at the head of the Dessi pedigree. The Dalriada of Kintyre was founded on the peninsula belonging to the British Epidii or Horse-people. The Dalriadans themselves had a horse-name (Eochaid Muinremuir) at the head of their royal genealogy.
The name Epidii includes the P-Celtic root epos, meaning "horse" (cf. Welsh ebol, "a foal").[3] The Q-Celtic equivalent would be *ekwos, which became Old Gaelic ech. It is suggested that they were named after a horse god, whose name could be reconstructed as *Epidios.
On the name Eochaid, Professor Jurgen Uhlich wrote to me the following:
"Echoed/Eochaid is in fact a guttural, not an io-stem (gen. Echdach; cf. attached), and the Primitive Irish equivalent written in Ogam) would be *EQADECAS, cf. the formally related Luguid, ge. Luigdech < (Ogam) LUGUDECCAS. The real later outcome of *eku̯o-dii̯os is Echdae, lit. ‘horse-like’. And yes, that vowel in the second syllable was o, not the Latinate i seen in Epidii."
Kent, called Cendlond in the ANGLO-SAXON CHRONICLE, may well have been confused with Kintyre/Cindtire, 'Land's End/Head', where Arthur son of Bicoir is situated. The ruling center of Arthur son of Petr was in Pembroke, earlier Penbrog or 'Land's End/Head.'
Now, sure, all of that may simply be coincidence. Or even just more spurious tradition. But, alas, along with the Arthurian battles it is really all we have when it comes to being able to localize our hero somewhere. The Arthurian birth story and the names of his parents are complete fiction (see
If we go with the earliest stratum of Arthurian lore, one that refers to him not as a king, but only as a soldier, and combine that with the pronounced equine element, we do end up with a figure who looks a lot like the eques/knight L. Artorius Castus. But there are two problems with that conclusion. First, there were a lot of members of the knight class in Britain during Rome's presence there. And it was common for non-knights who happened to be cavalrymen to put eques on inscriptions. We have a great many such examples. To claim that Castus was famous as an eques and so his name Artorius (which would have been linked to the British and Irish words for "bear", in any event) became popular among horse-loving Celts centuries after his floruit seems unlikely. Not impossible if Castus had achieved a quasi-mythical status in Britain, I suppose.
But when we toss the stone into the pond (!), the ripples that are created seem to have nothing whatsoever to do with Castus. Instead, we are forced to look at a Dark Age Arthur at Petrianis, himself a descendent of the Ala Petriana. Naturally, the name preserved over the centuries there may have come ultimately from Castus. Or it may simply be the name Arthur, derived from another man of that name. As Dr. Roger Tomlin has pointed out to me, Artorius "was a not uncommon name" among the Romans. We know that there were Artorii in the Dalmatian province over which Castus was procurator and some very good scholars think Castus may have not only have died in Dalmatia, but may well have been born there. We had Dalmatian units serving in Britain next to Castus's York and at the Carvoran fort on Hadrian's Wall. The British name of Carvoran was Magnis, "Stone, Rock." The Birdoswald Roman fort just to the west of Carvoran, with its Dark Age royal center, is in the Irthing Valley of the Camboglanna/Camlan fort. Dr. Andrew Breeze thinks the etymology of the river-name Irthing is "Little Bear", and I have tentatively placed the *Artenses (see the Welsh eponym Arthwys) or "Bear-people" in that valley.
However, it is important to note that neither Carvoran nor Birdoswald housed cavalry. On the other hand, the Stanwix garrison was the largest horse unit in all of Britain.
Petrianis also has some other points in favor of its being the Arthurian center during the Sub-Roman period. It was quite close to the Aballava/Avalana/"Avalon" Roman fort on the Wall, with its Dea Latis or "Lake Goddess." It was also near Concavata, a fort named for the raised bog in its vicinity (
https://mistshadows.blogspot.com/2024/12/a-new-theory-on-concavata-name-for.html). The dish known as a concavata was comparable to the gradalis/grail in medieval sources and I have proposed the fort at Drumburgh as the prototype for Arthur's Grail Castle. Lastly, Camlan at Castlesteads on the Wall is not far to the east of Stanwix.
In other words, Stanwix/Petrianis is the perfect locus in the context of other important Arthurian sites in the region.
This is the way it may have all worked:
1) A famous horseman Arthur of the 6th century, a descendent of the largest cavalry group in all of Britain, the Ala Petriana, resides at Petrianis. He fights a series of battles ranging from Buxton (Badon) at the southern edge of the old Brigantian territory all the way up to either a Caledonian Wood in Lowland Scotland or in Highland Caledonia itself. Battles from the Wall south are explicable in what Dr. Ken Dark called a war-leader who was attempting to maintain the prerogatives of the Roman Dux
Britanniarum. Battles to the North - which, frankly, accord much better with L. Artorius Castus - could without too much of a strain be construed as being due to either Arthur's conquest of that area or his being in alliance with the Gododdin (Roman Votadini). Remember, the HB does tell us Arthur was acting as a general in the wars of other British kings.
I'm reminded of the great Welsh poem THE GODODDIN, which tells of the disastrous British battle against the English at Catterick in Yorkshire. If such a force could come that far south from Edinburgh, then we might see Arthur doing something similar to help defend "Eidyn on the border" (see the PA GUR poem). We know Urien of Rheged, who came after Arthur, fought at Bremenium/Brewyn/High Rochester (an Arthur site), but also at Catterick.
The battles in the Manau Gododdin region, where we find the Roman period Maeatae (Miathi in the Irish sources),
could belong to Castus or even to Arthur son of Aedan, who is said to die variously while fighting against the Miathi or in Circinn further north (
https://mistshadows.blogspot.com/2025/05/the-discovery-of-ancient-pictish.html). It has always been deemed possible that the battles of several Arthurs have been combined to form the Arthurian list in the HISTORIA BRITTONUM, although the Badon and Camlan battles can only belong to the Dark Age British leader - especially if we hold fast to the dates in the AC.
2) The Dalriadans come into the Epidii territory, and the Dessi come into the Dyfed of Epona [1]. Within a fairly short period of time, and after intermarriage with the indigenous Britons, they decide they are thoroughly British, or desperately want to be. The Dyfed genealogy shows this, as the Irish version has the Irish names intact, while the Welsh version has stricken the Irish names and replaced them with Roman ones. One of the best ways for these horse-loving Irish-descended dynaties in Britain to assert their "Britishness" was to give royal sons a name of a famous British horse warrior, i.e. Arthur. The Petriana link inspired Petrus of Dyfed to name a son after this horse warrior of the Petriana. The Strathclyde Britons of Petra Cloithe/Alclud, through a niece of the king who was married to Aedan, passed the Arthur name to the Scots who were merging with the Epidii. Or it could have been the British Epidii themselves who passed along the Arthur name, with or without the "rock" connection (although we should not lose sight of the monolith at Dun Beachaire in Kintyre).
To me, the above schema seems a decent construct. Is it history? Well, who knows? It's the best that I personally can do after a quarter of a century of Arthurian research.
[1]
On Epona, I have this from John Koch's CELTIC CULTURE: A HISTORICAL ENCYCLOPEDIA. I have highlighted a relevant sentence of the selection.
Epona’s name is Celtic, specifically Gallo-
Brittonic (P-Celtic), and means ‘horse goddess’.
She is the most abundantly attested Celtic deity of
the Roman Empire. Evidence for her cult is strongest
in central and eastern Gaul, as well as the military
zones of the Rhine and Danube frontiers, and northern
Roman Britain. Within the military, the cult recurs
among, but was not limited to, cavalry units and units
recruited from Gaul. Epona is mentioned by the Roman
author Juvenal (see Greek and Roman accounts §9),
but we know of the cult mainly from inscriptions
and accompanying images, many on Romano-Celtic
altars with a focus cut into the top for the pouring of
libations or presenting other offerings. In the inscriptions,
almost all of which are in Latin, with far fewer
in Greek, she is sometimes called dea ‘goddess’ or regina
‘queen’ and is often grouped with other deities, for
example, the following on an altar at Pförring, Bavaria:
CAMPES(TRIBUS) ET
EPONAE ALA I
SING(ULARIUM) P(IA) F(IDELIS)
C(IVIUM) R(OMANORUM CVI P(RAE)EST
AEL(IUS) BASSIANUS
PRAEF(ECTUS) V(OTUM) S(OLVIT) L(IBENS) M(ERITO)
To the gods of the parade ground and to Epona,
the devoted and loyal first ala [auxiliary cavalry unit]
of singulares, Roman citizens led by the prefect Aelius
Bassianus, in fulfilment of a vow. (CIL III nos. 5910
and 11909)
From Auchendavy on the Antonine Wall in Scotland
(Alba), a Roman altar reads:
MARTI
MINERVAE
CAMPESTRIBVS HERC(V)L(I)
EPONAE
VICTORIAE
M(ARCVS) COCCEI(VS)
FIRMVS
C(ENTVRIO) LEG(IONIS) II AVG(VSTAE)
To Mars, Minerva, the Goddesses of the Parade
Ground, Hercules, Epona, and Victory, Marcus
Cocceius Firmus, centurion of the Second Legion
Augusta [set up this altar]. (RIB no. 2177)
A small bronze plaque for a donkey cart found at the
Gallo-Roman centre Alesia carries the punched
inscription:
DEA(E) EPON(A)E. SATIGENUS SOLEMNI(S)
FIL(IUS).V(OTUM).S(OLVIT).L(IBENS)
To the goddess Epona, Satigenus son of Solemnis
willingly fulfilled his vow.
The name Satigenus is Celtic.
The Epona cult was richly visual. Relief sculptures
often show her riding a horse side-saddle, the goddess
astride the horse being more common in the territory
of the Treveri in north-east Gaul. She sometimes
appears with a foal, particularly in the territory of the
Aedui. Images showing the goddess enthroned were
popular in the Rhine–Danube military frontier zone
and are probably to be understood as imperial iconography.
The figure of the horse cut into the hill at
Uffington may reflect a related cult in pre-Roman
Britain.
In Celtic studies, Epona is often mentioned—
and attempts have been made to recover her myth—in
connection with supernatural female characters in early
Irish and Welsh literature who have strong thematic
and narrative associations with horses, such as Macha
and Rhiannon (cf. also sovereignty myth), as well
as the Welsh folk custom of the Mari Lwyd. On the
likelihood that traditions of Epona have survived in
Continental chivalric romances, see Romance lyric
(also Arthurian literature [6] §2; courtly
love).
The root of the name Epona also occurs in Old Irish
ech ‘horse’ and the Gaulish month name EQVOS found
on the Coligny calendar, both from Indo-
European *ek´wos ‘horse’. On this type of divine nameformation,
cf. Damona; Matronae; Nemetona;
Sirona.
Primary Sources
Inscriptions. CIL 3, nos. 5910, 11909; RIB no. 2177.
