Tuesday, January 31, 2017


Llandewi Aberarth, church with hill fort above it

Llandewi Aberarth

In previous blog posts, I proposed that the name Arthur or Artorius was a Latin decknamen on an Irish name meaning "Bear-king", and that the king in question was Ceredig son of Cunedda of Ceredigion, known in the Anglo-Saxon Chronicle as Cerdic of the Gewessei.  Furthermore, I had theorized that the bear personal names in the Ceredigion genealogy could best be explained by the fact that in mid-Ceredigion there was an Arth or 'Bear' River, right next to the Afon Aeron.  Aeron or *Agrona, 'Slaughter goddess', suggested to me that the "Bear" might also have been a divine river.  If so, the bear names of Ceredig and his descendants were due to the fact that the ruling center of the kingdom was somewhere on the Afon Arth.

Subsequent research to find this ruling center of Ceredig/Arthur revealed only two possible sites. One is Dineirth, and the other Llandewi Aberarth.  What I needed to determine is which of the two places showed potential for sub-Roman or early medieval habitation.

According to Frances Foster, Archive & Library Assistant, Archive and Library Team, Royal Commission on the Ancient and Historical Monuments of Wales:

"I have checked the information we hold on the above site [Dineirth] and have not found any mention of Roman occupation or activity.  The description on the Ordnance Survey National Archaeological Record card states that the site was originally a Clare Castle (Richard de Clare) founded prior to 1136.

In R.E. Kay’s description he suggests that ‘the first Norman adventurers may have found the earthworks of a small promontory fort at their chosen site which they heightened and strengthened’.  He also says that ‘the Castle of Dineirth (or Dinarth) was probably founded by Richard de la Mare and follower of Richard de Clare in 1110.

Again in Cadw’s scheduling description there is no mention of sub-Roman occupation of this site." 

Llandewi Aberarth may be different. On the possible ruling center of the kingdom of Ceredigion, Toby Driver, Senior Investigator (Aerial Survey) for the RCAHMW and author of HILLFORTS OF CARDIGAN BAY, says

"I would find it very difficult to decide on a place for a ‘ruling centre’, although a coastal position may be better connected than one inland [emphasis mine]... The Arth has a very restricted catchment, and I am not aware of any other major archaeological sites along its reaches. That said the presence of a major castle at Dinerth, the Viking hog-back stone at Llanddewi-Aberarth and the coastal hillfort, along with the ancient fish traps on the foreshore at Aberarth, all add up to a high concentration of interesting sites. Certainly it is an interesting locale along the Cardigan Bay coastline."

Here is the description of the hill fort at Llandewi Aberarth from COFLEIN (http://www.coflein.gov.uk/en/site/419561/details/hillfort-st-davids-church-llanddewi-aberarth):

NPRN 419561

Map Reference SN46SE

Grid Reference SN47616337

Unitary (Local) Authority Ceredigion

Old County Cardiganshire

Community Dyffryn Arth


Broad Class Defence

Period Iron Age
Site Description The church of St Davids sits below and to the east of the triangular summit of a prominent coastal hill which has been suggested as a possible Iron Age hillfort by the Dyfed Archaeological Trust, citing both its general shape and the presence of Bronze Age finds just below the summit.

Royal Commission aerial photographs of the hilltop taken during the drought summer of 2013 (on 12th July) appear to clarify parched banks and ditches of a large triangular or pear-shaped hilltop enclosure, generally following the lines of present field boundaries and walls but of more massive construction, suggesting a prehistoric origin. This main hilltop enclosure measures approximately 278m x 110m, enclosing 2.7 hectares. On the east side of the ridge is a discrete, smaller rectangular earthwork enclosure within the larger one, following the line of the ridge but with a rounded southern end, into which the cemetery of St Davids Church encroaches. It is possible that this is also a prehistoric or early medieval enclosure [emphasis mine], perhaps representing a first phase, or a contraction of a larger site. Recorded during RCAHMW aerial reconnaissance (frames AP_2013_2951-53) and also during winter conditions on 26th February 2014, highlighting the earthworks (frames AP_2014_0311-12, 319, 324-5).

If I'm right about the Arth River being the location of the Dark Age ruling center belonging to Ceredig and his descendants, then the hill fort at Llandewi Aberarth must be considered the primary contender.  Of course, only excavation of the site, and in particular of the smaller enclosure on the east side of the ridge, would be able to demonstrate whether sub-Roman or early medieval habitation occurred there.

I'm still trying to procure at least a preliminary report on the hill fort at Llandewi Aberarth, but so far have had no success obtaining such a document.  It may not exist. However, if such a report is forthcoming, I will immediately make it available here.  

Saturday, January 28, 2017


Here is my final arrangement of the battles of Cerdic/Ceredig/Arthur.  For a more detailed discussion of them, please see some of my more recent blog posts.  

A few notes on the map:

1) Ceredicesora or "Cerdic's shore" has been thought to be the Ower near Calshot.  This is a very good possibility for a landing place.  However, the Ower further north by Southampton must be considered a leading contender, as it is quite close to some of the other battles.

2) Cerdicesleag or "Cerdic's wood" I would tentatively identify with Lee (leag)-on-the-Solent.  I pick this location because of the mention of Stuf (= Stub/b) both before and after the Cerdicesleag battle. Lee-on-the-Solent is just a little bit west of Stubbington, the settlement of the descendents of Stuf/Stubb.  It is also just across the Solent from the Isle of Wight, which was given to both Wihtgar and Stuf.  

3) Bitton on the Somerset Avon is thought to be the Roman period Trajectus (= Tribruit).  While the place may have been called this, it was named for the actual crossing of the Severn from here to Caerwent in Wales.  

4) Wallop, here in green letters, was supposedly the location of a battle of Ambrosius.  The great Danebury Hill camp is at Nether Wallop and I would make the case for this being the actual battle site.

Here are the correspondences of the Historia Brittonum battles and the modern place-names:

Glein = Ower 1 or Ower 2

Dubglas in Linnuis = River Blackwater

Bassas = Charford

Castle Guinnion = Wihtgarasburh

City of the Legion = Caerleon

Tribruit = Bitton (Severn Crossing from Caerwent)

Breguoin/Agned (Bremenium/Egnatius) = Liddington Castle/Badbury

Badon = Badbury/Liddington Castle


Some discussion has been made of Chearsley and Notley in Buckinghamshire as indicators that Cerdic of Wessex's battles originally belonged there, rather than in Hampshire.  This is quite simply wrong.

Charford, for example, has known early forms such as Cerdeford.  Notley is not remotely possible for Natanleod/leag.  Here are the two Buckinghamshire place-names as discussed briefly by Dr. Richard Coates of The English Place-Name Society:

"The bulk of early spellings are of the type Cherdes-, and nothing definitely points to Ceolred (no <l>) or Cerdic (no <c>).  Notley (Bucks) is just a straight up and down hnutu + leah, as all the medieval spellings show. And Natan- won’t give Nut- spellings."

Wednesday, January 25, 2017


Cadbury Castle, Somerset

As I mentioned in my previous blog piece http://mistshadows.blogspot.com/2017/01/arthurs-badon-at-last-evidence-for.html, Badda of the Badbury/Baddanbyrig hill forts is an attested English name that is also found spelled Bada. It is the spelling with a single /d/ that should interest us the most.

I was thinking about the several Badburys and recalled that a similar pattern of hill fort names exists further to the west in England, in territory that remained much longer in control of the British.  I'm referring, of course, to the Cadbury forts.  There are four of these:

Cadbury Camp in North Somerset
Cadbury Castle in Devon
Cadbury Castle in Somerset
Cadbury Hill in North Somerset

To these we may compare -

Badbury Rings in Dorset
Badbury in Wiltshire
Badbury Hill in Berkshire
Badbury in Northants.
Baumber in Lincolnshire (Badeburg in the Domesday Book; Bada's or Badda's burg, according to                                                 Ekwall)

It is fashionable in older books (even Mills!) to find the first component of this place-name as a hypothetical OE personal name *Cada.  But more recent scholars (like noted English place-name expert Dr. Richard Coates) agree with me that the known Brythonic name (found in an Arthurian context!) Cadwy/Cato/Cado/Cattw, etc., more likely explains Cad-.  Although Cadwy is included in some royal genealogies and is presented to us as a human hero, his name might well be a hypocoristic form of a divine name like the Gaulish god Caturix, 'Battle-king'.  Dr. Coates response to my query about whether Cadwy could stand for Caturix or similar was merely "Yes."

The Welsh word cad as listed in the GPC:


[Crn. cas, Gwydd. cath, Gal. Caturīges; fel elf. mewn e. fel catberth, caterwen, catffwl ac enw nant fel Cadnant golyga ‘cryf, nerthol, mawr’]

eb. ll. cadau, cadoedd.

a  Brwydr, ymladdfa, rhyfel, ymryson, ymdrech, helynt:

battle, conflict, war, strife, struggle, trouble. 

In other words, the Cadbury hill forts are elevated defensive earthworks that were deemed sacred to a sort of divine personification of battle.

I would suggest that the same is true of the Badbury forts - and, indeed, that this and only this can explain why here are several such forts all bearing the same name.  

In the famous, though exceedingly brief, account of Badon in Gildas, we are told a great slaughter of the enemy occurred at the hill.  Gildas's word here is the Latin strages. [For those who would like to study for themselves the relevant sections on Badon in Gildas, in Latin text form and in good modern English translations, please see http://people.clas.ufl.edu/jshoaf/2014/07/07/gildas/.  As I've mentioned before, Bede pretty well repeats what Gildas has to say on the battle, using stragis for strages.] Strages or 'slaughter', as it happens, is one of the meanings of OE beado, a word that occurs in numerous uncompounded and compounded personal names in a fairly wide variety of spellings.

From the Anglo-Saxon dictionary of Bosworth and Toller:

BEADO, beadu; g. d. beadowe, beadwe, beaduwe; f. Battle, war, slaughter, cruelty; pugna, strages :-- Gúþ-Geáta leód, beadwe heard the War-Goths' prince, brave in battle, Beo. Th. 3082; B. 1539. Wit ðære beadwo begen ne onþungan we both prospered not in the war, Exon. 129b; Th. 497, 2; Rä. 85, 23. Beorn beaduwe heard a man brave in battle, Andr. Kmbl. 1963; An. 984. Ðú þeóde bealdest to beadowe thou encouragest the people to slaughter, Andr. Kmbl. 2373; An. 1188. [O. H. Ger. badu-, pato-: O. Nrs. böð, f. a battle: Sansk. badh to kill.]

So on the model of the Cadbury forts, I propose that Bada or Badda is merely a spelling variant of a beado-derived name, and that Gildas's comment on the strages that took place there is not only a description of the result of military action at the location, but an actual definition of Bada/Badda as 'Slaughter.'  When I passed this idea along to Dr. Coates, he commented "Well, to my mind this is ingenious."

If I am right, this is yet more evidence in support of Badon = Badbury, in this case the one now called Liddington Castle in Wiltshire.  

Tuesday, January 24, 2017


Barbury Castle, Wiltshire

Here is the 'Badon Problem' in a nutshell, and it is one I've been wrestling with for decades: the place-name itself is universally held by Celtic linguists and place-name specialists to be a British form of the Old English Bathum or 'Baths.'  The philological and phonological arguments are very solid.  

But there are major problems with this identification.  Firstly, Gildas - who claims the battle happened when he was born, and who should, therefore, have some authoritative knowledge regarding the event - claims the Saxons were soundly defeated there.  This seems quite impossible for Bath in Somerset, as the English has not penetrated anywhere near this far into the West Country at this point in history.  

Slightly more likely is the other 'bathum', namely Aquae Arnemetia at Buxton in Derbyshire. Buxton, however, does not fit in at all with the other battles as I've laid them out in the South of England.

I've recently identified Arthur's City of the Legion and the following battle of Tribruit with Caerleon-on-Usk and the Trajectus or Severn crossing to the Avon.  Bath is, of course, on the Avon, and so would fit in well with this grouping as Badon.  But, the intervening battle of Mount Breguoin/Agned must give us pause.  

I've shown how the meaning of the root of the place-name Breguoin (= Brewyn/Bremenium) matches that of the root found in the Liddington place-name that was given to the Badbury hill fort in Wiltshire.  Agned is for Egnatius, the name of a Roman governor who became associated with Bremenium.  This is too much of a coincidence, and I'm fairly certain Breguoin is here a substitute for the English place-name Liddington.  

'Farm/settlement on the stream called Hlyde (= the noisy one)'. The water-course is on record as the Lid Brook.
Elements and their meanings
hlȳde (Old English) A noisy stream (literally a loud one.
tūn (Old English) An enclosure; a farmstead; a village; an estate.

Watts; Cambridge Dictionary of English Place-names 372
E. Ekwall; Dictionary of English Place-Names 297
A.D. Mills; Dictionary of English Place-Names 298
"Gover, Mawer and Stenton, The Place-Names of Wiltshire" 282

I have noticed that there is a Lid Brook a few miles ENE of Bath as well.  While I tried making a case for Bury Wood Camp being named for the Lid Brook, the truth is that the hill fort is between two branches of the Doncombe Brook, with the Lid being a fair distance to the south.  Nor could I make a case for the Lid Brook once being the original name of the By Brook river, for this last was originally called the Weaver. Here are the early forms and the etymology for the By Brook (as kindly supplied to me by Ally McConnell, Archivist, Wiltshire and Swindon Archives, Wiltshire and Swindon History Centre, who was drawing from THE PLACE-NAMES OF WILTSHIRE):

Essentially the same information is to be found in Eilert Ekwall’s book English River Names (1928 edition):

What this means is that the 'Badon' identified with Breguoin has to be Liddington Castle.

I've also demonstrated that the so-called Second Battle of Badon of the HISTORIA BRITTONUM would appear to be a reference to the ANGLO-SAXON CHRONICLE'S account of Cenwalh's being driven to Ashdown by Wulfhere.  Ashdown is only a few miles from the Liddington Badbury and we might assume that the ravaging went right past the hill fort.  

As I see it, there are two possible solutions to the 'Badon Problem:'

1) Very early on the superficially similar names Bathum/Badon and Baddan(-byrig) were confused in the Welsh tradition. Gildas or his source may not have been responsible for this, but subsequent copyists certainly could have been.  Thus what should have more closely resembled Baddan- ended up, merely by mistake, as Badon.  Eventually, the incorrect form became accepted as canonical, while the correct one was forgotten.


2) Modern linguists are, at least in this one case, overly strict in disallowing the possibility that Badon actually represents Baddan(-byrig) and not Bathum. In other words, that some otherwise unexplainable development took place which transmuted Baddan- to Bathum.

My own feeling is that it is far more likely the ancient Welsh are the ones to blame for the place-name switch.  They were far less astute in the science of language evolution than today's scholars.  In this respect, then, the linguists and place-names experts are both right and wrong: they are right in that Badon cannot be the linguistic equivalent of Baddan-, but they are wrong if they insist that precisely because this is so Badon cannot be the Liddington Badbury.  

For what it's worth, Liddington Castle has my vote as Arthur's Badon. 

Friday, January 20, 2017

Arthur's Badon at Last: The Evidence for the Badbury at Liddington Castle

Liddington Castle Ditch

I've struggled for some time with the hill-name Agned, site of one of Arthur's battles according to the HISTORIA BRITTONUM.  While I was seeking the hero's military activities in the North of Britain, I was quite satisfied that Agned stood for the personal name Egnatius, a Roman governor who did rebuilding of the fort of Bremenium.  This argument received enthusiastic support from language experts.  Bremenium is Arthur's Breguoin, found as Brewyn in a poem devoted to Urien's campaigns.

But when I came to realize that Arthur was Cerdic of Wessex or, more properly, Ceredig son of Cunedda, and had to shift the theater of war to southern England and Wales, Agned once again became a problem.  I wondered if it might be for Agnes (Agnetis), another excellent solution for the hill-name.  I then attempted to foist Agnes onto Little Solsbury Hill at Bath, thinking this Christian saint may have been a substitute for the virgin pagan goddess Sulis Minerva.  But this argument was forced in the extreme, as there is no evidence for the worship of St. Agnes at Bath.

While we can opt for Agned being a corruption of agued, a word meaning "distress, dire straits, anxiety' or the like, and propose that it is not really a true place-name at all, but merely a descriptor, this solution takes some special pleading.  Such is always the case when you propose an alternate reading, thus assuming, quite dangerously, that the word in question should be something other than what it is.  We often resort to this kind of linguistic trickery when we find ourselves ignorant of the original form.

What I'm realizing now is that I can keep Egnatius as the name underlying Agned. I would suggest that what happened with the three hill names in the battle list is as follows:

Liddington is a name based on a stream that ran near the hill fort of Baddanbyrig/Badbury.  The stream-name is from OE hlyde, 'loud', and conveys the sense of the 'roaring' water.  As it happens, the root of Bremenium/Brewyn/Breguoin is Welsh brefu, used of a stream by the Roman fort of that name.  The meaning of brefu is the same as that of English hlyde, something I've mentioned before in previous essays.  So what seems to have happened here is that whoever compiled the battle list knew the meaning of English Liddington, a name applied to the Badbury hill-fort, and so substituted Breguoin, a Welsh name with the same meaning.  After this substitution took place, someone else who did not know Breguoin was for Badbury at Liddington added Egnatius/Agned to this list as a second name for Breguoin. I've also pointed out that the Roman fort of Bremia on the Afon Brefi in Ceredig's Ceredigion is from the exact same root as Bremenium.    

I can now state my opinion that Badon, despite the philologists insistence that this MUST be for Bathum, 'baths', is actually Badbury at Liddington.  Furthermore, the Barbury hill-fort close to Liddington, the "Fort of the Bear", was named for Arthur, whose name was by the Welsh connected with their own word for bear, arth.  Finally, Wanborough hard by Liddington Castle was site of the Roman town of Durocornovium.  Cornovium contains the same place-name element as Cernyw or Cornwall and Cornovii (the tribe inhabiting what later became Powys).  Arthur in Welsh tradition is always associated with Cornwall. 

A great deal of ink has been spilt on the most probable etymology for the Baddanbyrig name. This comes from an attested, NOT a hypothetical name, Badda.  Badda itself is demonstrably Germanic. Dr. Richard Coates recently shared with me why the name can't come from the Brythonic:

"Badda, if borrowed, and if we take the double <dd> seriously, is difficult to link to a Brittonic etymon.

British */t/ > Britt. */d/ would show up as OE /d/, not as a geminate (double).
British */tt/ > Britt. */θ/ would show up as OE */θ/, written with thorn, but never <dd>.
British */d/ > Britt. */ð/ would show up as OE /d/ or /ð/, depending on the period, for which the spelling <dd> is most unlikely.
British */dd/ seems to have yielded simple Britt. */d/ (Jackson LHEB 428, on credu), and would show up as OE /d/, not as a geminate (double).

So I conclude that Badda cannot be of Celtic  origin, particularly because Late British did not have geminate /dd/."

I would note in passing that as with several similar Old English names (e.g. Eadda/Eada, Hadda/Hada; see M. Redin's STUDIES IN UNCOMPOUNDED PERSONAL NAMES), Badda is found spelled with only one /d/, i.e. as Bada.

Badda is actually found on a coin as the name of a moneyer of King Edward the Elder (899-924); see https://www.acsearch.info/search.html?similar=1146410.


[NOTE: The following is taken from an old post of mine at http://www.facesofarthur.org.uk/articles/guestdan2.htm.]

There is one possible clue to identifying Badon. It lies in a comparison of the Welsh Annals entry for the Second Battle of Badon and the narrative of the Anglo-Saxon Chronicle.  The actual year entry for this Second Battle of Badon reads as follows:

665 The first celebration of Easter among the Saxons.  The second battle of Badon. Morgan dies.

The "first celebration of Easter among the Saxons" is a reference to the Synod of Whitby of c. 664.  While not directly mentioned in the Anglo-Saxon Chronicle, nor the Anglo-Saxon version of Bede, there is an indirect reference to this event:

664 … Colman with his companions went to his native land…

This is, of course, a reference to Colman's resigning of his see and leaving Lindisfarne with his monks for Iona.  He did so because the Roman date for Easter had been accepted at the synod over the Celtic date.  

While there is nothing in the ASC year entry 664 that helps with identifying Badon, if we go to the year entry 661, which is the entry found immediate prior to 664, an interesting passage occurs:

661 In this year, at Easter, Cenwalh fought at Posentesburh, and Wulfhere, son of Penda, ravaged as far as [or "in", or "from"] Ashdown…

Ashdown is here the place of that name in Berkshire. It is only a half dozen miles to the east of Badbury and Liddington Castle.  A vague reference to ravaging in the neighborhood of Ashdown may well have been taken by someone who knew Badon was in the vicinity of Ashdown as a second battle at Badon.

I've recently identified Posentesbyrig as "Pascent's Burg".  Leading English place-name authority Dr. Richard Coates had this to say when I asked him if this etymology worked:

"I see no absolute barrier to Posent – Pascent. Welsh <sc> is the cluster [sk], which would be rendered in OE as  “esh” since OE had no cluster [sk] before a front vowel, even in the earliest times. “Esh” would normally also be spelt <sc>, but that’s a coincidence. It’s possible for “esh” to appear very occasionally as <s>, even before the conquest, as in Ryssebroc for Rushbrooke (Suffolk) in the mid-10th century."

Unfortunately, we have no way of knowing where Pascent's Burg was located.  Pascent son of Vortigern ruled over Buellt and Gwrtheyrnion.  But the Vortigern family was also said to have originated at Gloucester.  William of Malmesbury claims that Bradford on Avon was once called Vortigern's Burg, but this is surely not right, as Cenwalh of Wessex is said to fight at both Bradford on Avon and Posentesbyrig in the Anglo-Saxon Chronicle.  While Pascent's Burg and Vortigern's Burg may refer to the same place, if these aren't nicknames for Gloucester we cannot know Posentesbyrig's location.  

The battle listed before Posentesbyrig in the ASC is Peonnum, fought against the Welsh by Cenwalh. This is thought to be Penselwood in Somerset, found in the Domesday Book as Penne or Penna.

A battle at Gloucester, nicknamed Pascent's fort, would make sense between the Mercian king Wulfhere and the Wessex king Cenwalh.

Tuesday, January 17, 2017

St. Aaron, the goddess Aeruen and the City of the Legion

Roman fort of Chester on the River Dee

The SS Julius and Aaron were said by Gildas to have been martyred in the City of the Legion[s], urbs legionis.  The question has always been "Which City of the Legion?"  For there were several such in Britain, although the Welsh of the time appear to have known about only two of them.  

From Keith J. Fitzpatrick-Matthews article on the Cities of Britain in the Historia Brittonum (http://digitalcommons.brockport.edu/cgi/viewcontent.cgi?article=1022&context=jlo:

Scholars have made their own cases for the City of the Legion being other than Chester or Caerleon. Andrew Breeze, for example, opts for Leicester (http://digitalcommons.brockport.edu/cgi/viewcontent.cgi?article=1024&context=jlohttp://onomastics.ru/sites/default/files/doi/10.15826/vopr_onom.2016.13.1.002.pdf), while P.J.C. Field prefers York (http://www.heroicage.org/issues/1/hagcl.htm).  Andy Seaman ("Julius and Aaron 'Martyrs of Caerleon': in search of Wales' first Christians," ARCHAEOLOGIA CAMBRENSIS 164, 2015, 201-219) maintains a more traditionalist approach, accepting the relatively late evidence for the presence of these saints at Caerleon-On-Usk.

If we do accept the notion that Gildas's City of the Legion had to be either Chester or Caerleon-on-Usk, is there a way that we can decide between the two?  I believe there is, although the argument in favor of one of these forts relies on accepting a hypothetical substitution of a fictional Christian saint for an authentic pagan goddess.

The following summary on the authoritative etymology for the name of the Chester Roman fort is drawn from http://marikavel.org/angleterre/chester/accueil.htm:

* Eilert Ekwall : 

Chester Chs.  ( Deoua c 150 Ptol, Deva 4 IA, ciuitas Legionum, Legacaestir, Brit Carlegion c 730 Bede, Legaceaster c 890 OEBede, 894 ASC, Ceaster 1094 ASC (E), Cestre DE]. Deva, the earliest name, is identical with the river-name DEE. Chester is on the Dee. Chester must also have been called Lat Castra legionum, which is the source of OE Legacaestir, and also, with substitution of Welsh caer 'fort, city' for castra, OW Cair Légion (or urbs Legionis) c 800 HB, Welsh Caerlleon. Later Chester supplanted the longer name Legacaestir.


* A.L.F RIVET & Colin SMITH, The Place-Names of Roman Britain, pages 336 / 337 : 

- Ptolémée, II,3,II : Dnoua; Legiwn K Nikeforos (DEVA, LEGIO XX VICTRIX); variante : Deouana ( = DEVANA). 

- Inscription CIL XIII.6221 (Worms) : ... DEVAS.

Itinéraire d'Antonin :

- 4692 (Iter II) : DEVA LEG. XX VICI

- 4825 (Iter XI) : DEVAM (forme accusative)

- 4828 (Iter XI) : DEVA.

Ravenna 10644 : DEVA VICTRIS.

These forms hardly call for comment, except DEVAS 'from Chester' in two of the inscriptions; Holder I. 1274 suggested that this is for Deva(ti)s. In Ravenna's form, presumably read from a map, it seems that the legionary adjective has transferred itself to the whole settlement (i.e. it is not simply a case of omission of 'Leg. xx'; compare its entry for Isca Augusta). Victris shows Vulgar Latin -s for -x ; no Greek transcription is involved, as R&C thought.

DERIVATION. The name is British *Deua 'the goddess (par excellence ?)', more properly the name of the river ( > Dee) transferred as often to the place. Older Celtic *deiuo-s 'god', Indo-European *deiuos are postulated; cf. Latin divus, and cognâtes in many languages. Jackson mentions derivatives : Old Welsh duiu, Welsh dwy as in Dubrduiu, Dyfrdwy 'water of Dwy' (= Dee), and discusses the development of the name in Anglo-Saxon (LHEB 375, 629). The name was very widespread : in Britain Deva 2 and Deva 3, Devionissum, Devona; abroad, numerous Deva rivers and settlements in Ireland, Gaul and Spain, also in compounds such as Deobriga; and in many personal names, especially of Gaul. Many names show divo- rather than devo-; there may in some cases have been assimilation to Latin divo-, but Ellis Evans GPN191-93 explains that apart from this, 'Gaulish è (from Indo-European ei) probably had a very close pronunciation, which could account for the -i-'; Whatmough DAG 456 thinks that 'ei giving è > i is possibly dialectal' within Gaulish. Many names show Deo-, Dio- after loss of -v-. The belief of the Celtic peoples in the divinity of water, or more strictly of the présence of a divinity in the water, is widely attested by these names. Other British Dee rivers, not recorded in ancient sources, have the same origin (see map).

IDENTIFICATION. The Roman legionary fortress at Chester (SJ 4066). Victrix reflects the cognomen of the XX Légion, acquired in the suppression of the Boudican Revolt.

Note. Bede in II, 2, gives evidence of a possibly alternative ancient name which continued in use for some centuries : ...ad civitatem Legionum, quae a gente Anglorum Legacaestir, a Brettonibus autem rectius Carlegion appellatur. If this had endured, it would clearly have produced a *Caerleon like that of Monmouthshire. Watson CPNS 383-84 does indeed mention that Chester was Caer Lleon in Old Welsh. O. G. S. Crawford in Antiquity, IX (1935), 287, argues that Arthur's ninth battle as recounted by Nennius, in urbe Legionis, took place at Chester, because Caerleon-on-Usk was always so described in Welsh literature, while York - the third and last legionary centre - was never called 'city of the legion'. The proposal is certainly attractive. (Bede's mention of Caerleon by a similar name is taken from Gildas, a different case.)

Why is all this important?  Because, as it happens, we know the actual name of the deva/goddess of the River Dee.  Her name was AERFEN, found also as Aeruen as early as the 12th century.

I have found a mention of Aerfen in Celtic Folklore Welsh & Manx Vol II by J. Rhys page 441:

“The Dee has in Welsh poetry still another name, Aerfen, which seems to mean a martial goddess or the spirit of the battlefield, which is corroborated and explained by Giraldus, who represents the river as the accredited arbiter of the fortunes of the wars in its country between the Welsh and the English.”

In Celtic Britain also by J. Rhys on page 68 he refers to the Dee or Deva of North Wales as having another name in Welsh literature; Aerven or the genius of war; it was supposed to indicate the frequent wars between the Welsh and the English by eating away its bank on the Welsh or on the English side, as the case might be."

We might assume that if the goddess Aerfen were indeed ancient, she served the same purpose during the warfare that ensured between the Welsh and the Romans.

Welsh place-name expert Professor Hywel Wyn Owen (in his 'Dictionary of the Place-Names of Wales') says that Llyn Tegid/Bala Lake was called Llyn Aerfen in the 15th century.  The Afon Dyfrdwy flows into and out of Llyn Tegid.

Here is the GPC listing for the name Aerfen:


[aer1+elf. anh. (?cf. tynghedfen)] 


?Clodfawr mewn brwydr:

?renowned in battle. 

12g. GCBM i. 192, Baranres aeruleit, aeruen—y waedlafyn / A’e waedlan disgywen.

15g. GLGC 391, Efô a orfydd a fo aerfen, / efô a esyd cyfraith Foesen [i Domas ap Rhys].

Digwydd Aerfen fel e. duwies ac e. ar Afon Dyfrdwy, cf. D, Aerfen, Dyfrdwy, Dea fluvius. Aerfen bengrech felen fawr, a gw. B vii. 123-4.

The 'renowned' meaning here must relate to Welsh ban,   "high (of status, standing, &c.), lofty, exalted, dignified, excellent, renowned, conspicuous; tall, high."  But this strikes me as a poor derivation.

We might think Aerfen is not a goddess name at all, but merely a locative.  -fen could be for Welsh man, men, myn, "particular place or spot, location, position, part", making for a 'Slaughter-place', a sort of nickname for the Dee.  

However, in my opinion the best etymology for Aerfen might well be (Dr. Simon Rodway) Proto-Celtic *Agro-bena, "Slaughter-woman."  To this name we may liken the Welsh goddess Cerridwen. As stated by P.C. Bartram in his A CLASSICAL WELSH DICTIONARY:

"Ifor Williams thought that the original and authentic form of the name [Cerridwen] was Cyrridfen, meaning ‘crooked woman’, [Kyrridven in the Black Book (BBC 9.6 and 15.2)], rather than Ceridwen, ‘fair and loved’ (Chwedl Taliesin, pp.3-4; TYP p.308)."

 Marged Haycock agrees that the last element of the divine name Cerridwen is ben - 'woman'.

What I would propose is that a form of the Dee goddess's name Aerfen/Aeruen led to the substitution by Christians at Chester of the Biblical name Aaron.  Please understand I am not here in any way suggesting a linguistic relationship between the two names!  Far from it.  Instead, the superficial similarity was enough to warrant the replacement of the goddess with the saint.

As for Aaron's companion Julius, well, who can say?  In the Roman period Julius was an incredibly common name.  However, if I might speculate further one cannot help but recall that the Chester Roman fort was founded by the then-governor Sextus JULIUS Frontinus (http://www.british-history.ac.uk/vch/ches/vol5/pt1/pp9-15).  Frontinus was himself a priest, a distinguished member of the College of Augurs.  Could it be that in Julius and Aaron of Chester we have a folk memory of Frontinus, builder of the original fort on the Dee, and of Aerfen, the goddess of the river?

River Dee (Aerfen) Northeast of Glyndyfrdwy

Friday, January 6, 2017


King Bladud with one of his pigs, Bath

Bladud is first introduced to us in the fanciful history of Geoffrey of Monmouth, where we learn he is credited with founding Bath.  Subsequent details were added to this early British king's legend, although most of the accretions appear to be quite late.  Some of these additions are also fraught with detectable error.  For example, the notion that Bladud founded a university at Stamford is baed solely on the fact that Stamford was originally OE stan, 'stone, plus ford, and there is a Stanford on Avon in Leicestershire next to a Swinford (Swine-ford).  This particular Swinford was identified by some folklorist with the Swineford on the Somerset Avon, which in typical aetiological fashion was explained as being the place where Bladud crossed the river with his pigs.

However, it was the tale of the founding of Stamford's school with philosophers brought from Athens that led me to investigate Bladud's real identity.  

Athens was famous for its Lyceum, a temple to Apollo that served as a great center of philosophical learning.  Scholars have before commented on the story of Bladud's flying the temple of Apollo in London, where he perished, as a borrowing from the Classical story of Daedalus and Icarus. Daedalus in one version of the Greek myth is given wings by Athena (the Greek form of the Roman Minerva, who was syncretized with Sulis at Bath) and he hangs them up as a dedication in the temple of Apollo. 

The Lyceum was named for Apollo Lyceus.  Now, while this sun god's divine epithet has been subject to varying interpretations both in ancient times and modern, it was and still is often associated with the Greek word lykos for 'wolf.'  This is curious, as Bladud is from the Welsh blaidd, 'wolf', plus udd, 'lord', for 'Wolf-lord.' 

Only a couple of miles north of Bath at Nettleton Shrub was a Roman period temple to Apollo Cunomaglus (Cunomaglos).  While it has been customary to translate cuno- as 'hound', Xavier Delamarre (http://ralphhaussler.weebly.com/wolf-mythology-italy-greek-celtic-norse.html) has shown that cuno- in these early contexts could mean 'wolf' as well. In fact, Delamarre renders Cunomaglus as "Seigneur-Loup" or Wolf Lord - an exact match to the name Bladud.

In my previous blog post on Cunedda at Bath (http://mistshadows.blogspot.com/2017/01/cunedda-at-badonbathumbath.html), I discussed Cunomaglus in another context:

"In this year [577] the English under Ceawlin defeated the British at Deorham, modern Dyrham in South Gloucestershire, where there is a hillfort now called Hinton Hill.  As a result of this victory, Ceawlin took the British cities of Gloucester, Cirencester and Bath.  Bath is, of course, Arthur's Badon, a battle the Welsh Annals places around 516 A.D.

One of the British kings who fell in the Battle of Deorham is named Conmail.  As Dyrham is very near Nettleton Shrub - in fact, only several kilometers distant - and both sites are only a few miles from Bath, I take this Conmail to either be the god Apollo Cunomaglos, whose shrine was at Nettleton Shrub or a local chieftain who had assumed the name of the god."

What I would propose, then, is this:  Bladud is a later substitute for the pagan god Cunomaglos the Wolf-Lord.  As Apollo the sun god, he was credited with the founding of Sulis's shrine at Bath.  This is appropriate, as the most accepted etymology for the divine name Sulis is an ancient Brittonic root meaning 'sun', cf. Latin Sol, genitive Sulis.  While other theories abound, the similarity of the names would have been sufficient to bring the goddess Sulis under the aegis of Apollo. 

Like Bladud, Apollo was a shepherd - twice, in fact.  

According to at least one early source, he transformed himself into a great boar:

"Erymanthos, son of Apollon, was punished because he had seen Aphrodite after her union with Adonis and Apollon, irritated, changed himself into a wild boar and killed Adonis by striking through his defenses."

- Ptolemy Hephaestion, New History Book 1 (summary from Photius, Myriobiblon 190) (trans. Pearse) (Greek mythographer C1st to C2nd A.D.)

Apollo is not otherwise brought into close connection with swine, so far as I've been able to determine.  As Apollo Maponus or Mabon he is associated very closely in early Welsh tradition with the monstrous boar Twrch Trwyth,  This god is the master of three hounds during the hunt of the boar:

"And Mabon son of Mellt went with the two dogs of Glythfyr Ledewig in his hand, and Drudwyn the whelp of Graid son of Eri."

- Culhwch and Olwen

Of course, we have no idea whether swine were a part of the cult of Apollo Cunomaglus.


Belief, Legend and Perceptions of the Sacred in Contemporary Bath
Marion Bowman
Vol. 109 (1998), pp. 25-31
Published by: Taylor & Francis, Ltd. on behalf of Folklore Enterprises, Ltd.

Bladud of Bath: the Archaeology of a Legend

Bladud: The Flying King of Bath
A. T. Fear
Vol. 103, No. 2 (1992), pp. 222-224
Published by: Taylor & Francis, Ltd. on behalf of Folklore Enterprises, Ltd.


Dyrham Camp hillfort, Hinton Hill, South Gloucestershire

In my book THE ARTHUR OF HISTORY, I demonstrated that Ceawlin of the Gewessei was one of the names the great Cunedda went by. Also presented in that work was my observation that the genealogy of Cunedda and his sons had undergone a curious reversal in the Anglo-Saxon Chronicle. As I've recently identified Cunedda as Uther Pendragon, and his son Ceredig as Arthur, the ASC's year entry for 577 is of remarkable interest.  

In this year the English under Ceawlin defeated the British at Deorham, modern Dyrham in South Gloucestershire, where there is a hillfort now called Hinton Hill.  As a result of this victory, Ceawlin took the British cities of Gloucester, Cirencester and Bath.  Bath is, of course, Arthur's Badon, a battle the Welsh Annals places around 516 A.D.

One of the British kings who fell in the Battle of Deorham is named Conmail.  As Dyrham is very near Nettleton Shrub - in fact, only several kilometers distant - and both sites are only a few miles from Bath, I take this Conmail to either be the god Apollo Cunomaglos, whose shrine was at Nettleton Shrub or a local chieftain who had assumed the name of the god.  Of Conmail I will have more to say in my upcoming piece "Bladud and the Founding of Bath."  As Dyrham is on an escarpment of the Cotswolds, which were named for the Celtic goddess Cuda, is it probable that Cuthwine, the son of Ceawlin according to the ASC, is "Cuda's friend" (wine meaning 'friend' in Old English). Cf. Cutha son of, variously, Ceawlin or Cynric (Cynric = Cunorix son of Maquicoline/Ceawlin), and the name Cuthwulf.  Cutha is generally taken as a hypocoristic name, but I would make the case for all of these names originating with Cuda.   [NOTE: in 556 Ceawlin is said to have fought at Barbury hillfort in Wiltshire.  This is either the Fort of the Bear or, according to Ekwall, the Fort of a man called Bear.  I've made the case before that this 'bear' is a reference to Arthur, whose name could be connected with the Welsh word arth. 'bear.']

There is no way to reconcile the dates of Arthur's presence at Bath (Little Solsbury Hill) with that given for Ceawlin at Dyrham.  But the presence of the Gewessei here points to my latest identification of Ceredig's/Arthur's battles in the region as most likely being correct.  If Arthur were fighting along a line from the sites in Hampshire to those at Bath, Bitton (Trajectus/Tribruit) and Caerleon (urbs legionis), and he was based in NW Wales, the enemy was the British kingdom of Dumnonia and the English were considered allies in this conflict.  

Monday, January 2, 2017


Little Solsbury Hill

The Antonine Itinerary XIV (see Rivet and Smith) describes a route from Isca (Caerleon on Usk) to Venta Silurum (Caerwent) and thence across the Severn trajectus to Abone (possibly Sea Mills on the River Avon).  The route then continues to Aquae Sulis or Bath/Badum/Badon.

The hill of Agned is placed chronologically between the Severn Trajectus and Bath/Mount Badon.  Discounting for a moment Breguoin (Bregomion, Bregion, etc.), might this hill have been in the vicinity of Bath?  Or, might it have been another name for a hillfort at Bath?

I believe I finally have the solution to this vexing problem.  As I’ve mentioned before, the English name Badum (in Welsh, Badon) was chosen because the Roman-British name of the place was Aquae Sulis, the ‘Waters of [the pagan goddess] Sulis.’  Now, Sulis was identified at the site with Minerva, the Roman version of Athena Parthenos (Latin Parthenus), ‘the Virgin or Maiden.’  This immediately reminds us of Geoffrey of Monmouth’s ‘Castle of Maidens’ title for Mount Agned. 

It gets better.  A hillfort over Bath is to this day called Little Solsbury Hill.  Sols- here is believed by most scholars to be for the goddess Sulis.  So where does Agned fit in?  My earliest research, during which I consulted the best of the Celtic linguists and place-name specialists, showed that the most regular etymology for Agned was a Welsh form of Latin Agnetis, the genitive form for St. Agnes.  I thought nothing of this at the time, for I was woefully ignorant about that particular saint’s VITA.  But here is the relevant abbreviated version of her Life:

“Saint Agnes was twelve years old when she was led to the altar of Minerva at Rome and commanded to obey the persecuting laws of Diocletian by offering incense. In the midst of the idolatrous rites she raised her hands to Christ, her Spouse, and made the sign of the life-giving cross.”

Some good resources on Agnes and her relationship with Minerva:


So while there is no evidence for the worship of St. Agnes at Bath, I feel that what we have in the Agned name is simply this: the hillfort of Sulis was the site of Arthur’s famous battle.  As neither Sulis nor the Roman goddess of virgins with whom she was identified could be mentioned by monkish writers like Gildas, the place was referred to by the name of the saint who was the patron of virgins – Agnes.  Thus Mount Agned or Mont Agnetis is Little Solsbury Hill at Bath.

I’ve earlier discussed Mount Breguoin as a Welsh attempt to translate a hill-name near the Lid Brook just NE of Little Solsbury Hill.  In all likelihood, then, all three hill names in the Historia Brittonum are, in typical Celtic fashion, meant to designate the same hillfort, i.e. the Fort of Sulis.