THE NORTHERN KINGDOMS

To give some idea of the political landscape of Arthur’s Britain, it might be helpful to examine some of the “Men of the North” and the kingdoms they controlled.

The most northern of these kingdoms was, of course, the ancient territory of the Votadini or Gododdin, which in the Roman period is believed to have stretched from the Wear or the Tyne through Northumberland and the Lothians to the Forth.

The term ‘Lothian’ appears to have been of Dark Age origin, which as we have seen stands for an original Lleudiniawn, ‘Place of the Fort of [the god] Lugus’. There is an eponymous king recorded in the Life of St. Kentigern called Leudonus, i.e. Lleuddun, and his kingdom in Welsh was known as Lleuddunion. He was supposed to have ruled from Traprain Law, which was earlier called Dunpelder, the ‘Fort of the Spear (shaft)’.

In the late 6th century, the king of the Votadini was, apparently, one Mynyddog Mwynfawr. He is said to have ruled from Din Eidyn or Edinburgh and was the son of a certain Ysgyran, and probably succeeded Clydno Eidyn. The Gododdin poem implies that the Britons who fought the English at Cattraeth assembled at Mynyddog’s court at Edinburgh. Clydno Eidyn, in turn, was the son of Cynfelyn son of Dyfnwal Hen. Myynyddog is also given the epithet ‘Eidyn’ meaning, undoubtedly, ‘of Eithne’. Once again, Eidyn is likely the British form of Eithne, mother of the god Lugh in Irish tradition.

Pabo Post Prydain, the ‘Pillar of Britain’, is the son of Ceneu son of Coel Hen, both famous chieftains of the North. Pabo is spelled Pappo in the genealogies appended to the HB. Coel Hen’s name is believed to be preserved in Kyle in Ayreshire.

A son of Pabo is Dunod Fwr, who is probably the chieftain who fought against the Rheged princes in Erechwydd, which itself is usually placed somewhere in Cumbria. We may relate this Dunod to Dent in NorthWest Yorkshire, his lands here being termed the ‘regio Dunotinga’, kingdom of the descendents of Dunod. From John Morris’s The Age of Arthur:

“DENT: regio Dunotinga is one of four districts of north-western Yorkshire overrun by the English in or before the 670s, Eddius 17 [Life of Wilfrid]. The passage is overlooked in EPNS WRY 6, 252, where the early spellings Denet(h) are rightly related to a British Dinned or the like, and Ekwall’s derivation from a non-existent British equivalent of the Old Irish dind, hill, is properly dismissed. EPNS does not observe that Dent was, and still is, the name of a considerable region, and tha thte village is still locally known as Dent Town, in contrast with the surrounding district of Dent…. Regio Dunotinga plainly takes its name from a person named Dunawt, Latin Donatus, as does the district of Dunoding in Merioneth, named from another Dunawt, son of Cunedda.”

The regio Dunotinga was associated with the Ribble and other places in the north of the West Riding. As the Dent River is a tributary of the upper Lune in Lonsdale, and Upper Lonsdale seems to have been within the canton of the ancient Carvetii tribe, it is likely that Dunot was himself descended from the ‘People of the Stag’. The Carvetii (see Cerwyd/Cerwydd below) ruled over what we now think of as Cumbria and adjacent areas.

Bran son of Ymellyrn is associated with both Dunawt of Dent and Cynwyd of Kent (see below for the Cynwydion).  The patronymic here is transparently from Old Norse a, river, plus melr, sandbank, identifying his region with Ambleside in Cumbria just to the west of the River Kent. 

Another son of Pabo’s is Cerwyd or Cerwydd, who is otherwise completely unknown. This name is transparently an eponym for the Carvetii tribe. We have just seen that Dunod’s Dent seems to have been a part of the territory once covered by this ancient tribal kingdom.

The form Cerwydd as a direct eponym for the Carvetii is not possible; we would need Cerwyd for an exact linguistic correspondence. However, as Cerwydd means ‘stag-like one’, we can say with a fair degree of certainty that he does represent the People of the Stag.

As for Pabo, father of Dunod, we may situate him at Papcastle (Pabecastr in 1260), the Derventio Roman fort in Cumbria. Pap- is thought to be from ON papa, papi, for ‘hermit’, but this seems an unlikely name for a ‘ceaster’. Instead we should look to early W. pab, ‘pope’, i.e. papa, pl. pabeu, and Llanbabo church of St. Pabo in Anglesey. Pabo's Chester would seem to do quite nicely. We could then locate Pabo within the Carvetii kingdom of his sons Cerwyd/Cerwydd and Dunod.

I would add that Pabo’s epithet ‘Post’ or ‘Pillar’ is possibly a reference to the Solway, which is believed to be from OScand. sul, ‘pillar or post’, and vath, ‘ford’. It has been proposed, quite reasonably I think, that the pillar or post of the Solway is the Lochmaben Stone at Gretna Green. A ‘papa’ or ‘father’ of the post/pillar named for the Divine Son Mabon makes for an interesting combination of place-name elements!

However, it is true that the Papcastle fort is not on the Solway.  The name of the Roman period fort here – Derventio – was named for the river Derwent, the ‘oak-river’ or ‘river in an oakwood’.  As the oak was a very sacred tree to the early Celts, it is possible the ‘post’ or ‘pillar’ that gave its name to Pabo was an oaken one and thus an indirect reference to the place-name.

Sawyl Benisel ("Low-head"), yet another son of Pabo, is dated c. 480 CE. On the Ribble, not far south of ‘regio Dunotinga’, is a town called Samlesbury. The place-name expert Eilert Ekwall has Samlesbury as ‘Etymology obscure’, but then proposes OE sceamol, ‘bench’, as its first element, possibly in the topographical sense of ‘ledge’. A.D. Mills follows Ekwall by saying that this place-name is probably derived from scamol plus burh (dative byrig). However, sceamol/scamol is not found in other place-names where a ‘ledge’ is being designated. Instead, the word scelf/scielf/scylfe, ‘shelf of level or gently sloping ground, ledge’ is used.

I would suggest as a better etymology for Samlesbury: ‘Sawyl’s fort’. There are, for example, Sawyl place-names in Wales (Llansawel, Pistyll Sawyl, now Ffynnon Sawyl). Sawyl is the Welsh form of the name Samuel.

Dr. Andrew Breeze of Pamplona, a noted expert on British place-names, agrees with this proposed etymology:

“I feel sure you are right. The form surely contains the Cumbric equivalent of Welsh _Sawyl_<_Samuel_. Your explanation of this toponym in north Lancashire is thus new evidence for Celtic survival in Anglo-Saxon times.”

Now that we have placed Pabo and his descendents on the map, we need to investigate what has been explained as an intrusion on their pedigree.

An Arthwys and his father Mar are both inserted into the Pabo genealogy. Instead of Pabo son of Ceneu son of Coel Hen, we have Pabo son of Arthwys son of Mar son of Ceneu, etc. This same Arthwys is made the grandfather of a Cynwyd of the tribal group known as the Cynwydion (of the Kent river in Cumbria - Kent being from Kennet, which in Welsh is Cynwyd), of Gwenddolau of Carwinley (Caer Gwenddolau just a little north of Carlisle) and father of Eliffer (Eleutherius) of York. Eliffer in another pedigree is the son of Gwrgwst Ledlum (Fergus Mor of Dalriada) son of Ceneu son of Coel Hen.

Mar is made the father of Lleenog, father of Gwallog of the kingdom of Elmet (a small kingdom centreed about Leeds, probably from Welsh elfydd, ‘world, land’), but in another pedigree it is Maeswig Gloff, i.e. Maeswig ‘the Lame’, who is father of Lleenog.

Mar looks to be an attempted eponym for the Mor/Mer-ingas of Westmorland, although as the name is also written 'Mor', this could be yet another reference to Fergus Mor.  Fergus otherwise occurs in the early genealogies of the Men of the North as Gwrwst Ledlwm, the father of Meirchiaun Gul of Cumbria and Eliffer Gosgorddfawr of York.  

Maeswig Gloff (Masguic Clop in the Harleian genealogies) was, presumably, a ruler of the vast Plain or Vale of York. His name appears to be from *Magos-vicos, ‘Fighter of the Plain’. However, I should not neglect to point out that the Roman fort at Burrow Walls, Workington, Cumbria, was named Magis, formed from British *magos, ‘plain’. Papcastle of Pabo is on the Derwent only a few miles east of Magis, itself at the mouth of the same river.

The name Arthwys has frequently been brought into connection with that of Arthur/Artorius. This name is from Arth-, ‘Bear’, + –(g)wys(*weyd-so- 'knowledgeable'), which in the early period was comparable to Irish fios, ‘knowledge’. Hence he was the ‘Knowledgeable Bear’. Dr. Andrew Breeze has made a case for the river Irthing containing the word Arth, ‘bear’. From his article “Celts, Bears and the River Irthing” (Archaeologia Aeliana, 5^th series, volume XXXII):

Irthing, which has early forms Irthin, Erthina, and Erthing, would also make sense as ‘little bear’, with a Cumbric diminutive suffix corresponding to Middle and Modern Welsh –yn (Old Welsh –inn), as in defynyn ‘droplet’ from dafn ‘drop’ or mebyn ‘young boy’ from mab ‘boy’.  As the th of Arth is pronounced like that of English bath, but that of Irthing like that of brother, the process of voicing here would take place after borrowing by English, not before.”

The claim has been made that Arthwys should be Athrwys, as this spelling is found in later sources. The argument would seem to have some support as the name Athrwys is found in Wales. If it was Athrwys, the first element would be W athro 'teacher' (< PIE *pH2tro:w- ‘uncle’). However, as Professor Ranko Matasovic has pointed out to me via private correspondence, while we have plenty of examples of Arth- or bear names, other than the presumed Athrwys, we have absolutely no other extant names containing athro.

[NOTE: Arthwys can be interpreted as a territorial designation, rather than strictly as a personal name. Welsh has a -wys suffix, which derives from Latin –enses.  A discussion of this suffix can be found in John T.  Koch's Celtic Culture, among other sources.  Regedwis, for example, is 'people of Rheged' - or maybe better, 'inhabitants of Rheged'. The entry for -wys (1) in the University of Wales Dictionary confirms it as a Latin borrowing and as a nominal plural ending, giving the examples of Gwennwys, Lloegrwys and Monwys. Could –wys, then, be a suffix used for the people who live on a certain river?  Like on an Arth or Bear River?

When I put this question to Dr. Delyth Prys of the place-name experts at The University of Wales, Bangor, he replied: “I've no independent evidence for this, but river names are sometimes used as the name for a more general area and by extension it could be the people of the Arth (area)." Now, if the Irthing is not from ir-t, but from erth/arth +inga (belonging to, not descendents of), it would be the 'tun belonging to Arth' or belonging to the bear.  But if the river itself were originally the Arth/Erth, then the tun itself would belong to the river.

Alan James of BLITON states that river-names can sometimes be also the names of adjacent regions, or - probably more correctly - some river-names may have originally have been regional names (or vice versa). This may have been the case with Llwyfenydd/ Lyvennet of Urien. The kind of river-names that seem to double as district names tend to be ones that refer to local terrain, etc., but that may just because such topographical names are more obviously linked to the area. Again, rivers were sometimes boundaries, but they're as likely to flow through a territory perceived as one as to divide such a territory into two. A hypothetical Arth/’Bear’ region could have included both the Irt and the Irth of Irthington, not necessarily been bounded by them.]

According to the early Welsh genealogies, Gwenddolau ('white dales'), who belonged at Carwinley in Cumbria, was the son of Ceidio.  Ceidio as a name is a hypocoristic form of a longer two-part name that begins with *cad-, 'battle.'

Recently, I thought to look for a relic of Ceidio in place-names.  As he was a son of the Arthwys who stands for the *Artenses or People of the Bear of the Irthing Valley, my attention was caught at first by Powcady between the King Water and the Cambeck not far from the Camboglanna Roman fort at Castlesteads.  Early forms for Powcady were late: Pocadie, Pokeadam.  But Alan James proposed that this contained a typical pol- element 'pool in a stream, stream' plus cad-, 'battle', plus perhaps a -ou plural suffix.  I wondered if it could instead contain the name Ceidio/Keidyaw/Ceidiaw.

Powcady is at a footbridge over Peglands Beck, which was earlier known as Polterkened. See

https://spns.org.uk/wp-content/uploads/2019/03/Alan_James_Brittonic_Language_in_the_Old_North_BLITON_Volume_II_Dictionary_2019_Edition.pdf

As Polterkened (or at least Kened, as polter may have been added later) was this stream's ancient name, a *pol- of a different name on the same watercourse would designate a pool in this location.  I asked Alan James whether this could be 'Ceidio's Pool.'  He responded:

"Poll Ceidio isn't impossible, though it should be lenited *Geidio (but lenition is a bit iffy in Cumbric pns). So, no, not impossible."

I would very tentatively propose, therefore, that the name Ceidio son of Arthwys/Artenses is preserved at Powcady.

Over the years, I've explored different possibilities for the location of the famous Armterid/Arfderydd battle, at which Merlin (Myrddin) went made and fled into the Caledonian Wood.  But only recently have I been able to settle on one particular site.

The place was called ‘Weapon-fierce’ (courtesy Andrew Breeze; https://digitalcommons.brockport.edu/jlo/vol2/iss1/1/), i.e. Arm-terid or Arf-derydd.  According to Breeze, this was the name for the stream at Carwinley which marked the northern boundary of Arthuret parish.

As for fixing the site of the battle, we have one possible clue.  In “Lailoken and Kentigern" Carwinley or Caer Gwenddolau is called Carwannock and the battle is said to take place between the latter and the Lidel on a plain. I had proposed that -wannock was either derived from Cumbric gwaun, “high and wet level ground, moorland, heath; low-lying marshy ground, meadow” or might be a hypocoristic form of Gwenddolau. Gwen-ddolau itself looks to be a place-name, as it means, literally, “White dales” (dol being “meadow, dale, field, pasture, valley”). Brythonic place-name expert Alan James confirmed both possibilities for me:

“The meaning of derivatives of *wāgnā in the Brittonic languages is primarily ‘level, marshy ground’, whether upland or lowland; developments include gwaun ‘a meadow’ in Welsh, goon ‘downland, unenclosed pasture’ in Cornish. Br -āco-/ā-> -ǭg is an adjectival and nominal suffix, indicating ‘being of the kind of’, ‘association with’, ‘abounding in’, the stem-word. It occurs very widely in river-names, hill-names and other topographic names. It 's not diminutive, though in hypocoristic personal names like Gwennock it might be affectionate.”

Thus Carwannock and Carwinley are the same place.  This is confirmed, in fact, by the primary sources. In the St. Kentigern VITA fragment (Titus A. XIX ff. 74-75b) the language is in campo qui est inter Lidel et Carwannock (see https://www.persee.fr/doc/roma_0035-8029_1893_num_22_88_5789 ). But according to the 15th century edition of John of Fordun the battle took place in campo inter Lidel et Carwanolow situato (Ifor Williams quoted in PNCmb I p51 n1).

But what of modern Arthuret, which is considerably to the south of Carwinley?

Derydd as similar to L. torridus, dried up, also Irish tioradh, drying, tíraid, dries.

Guess what is a tributary of the Hall Burn IN ARTHURET PARISH?

THE DRY BECK (see http://www.lakesguides.co.uk/html/lgaz/lk00252.htm).

Dry Beck in Arthuret Parish

So this stream is Terydd/Derydd - probably the original name for the entire Hall Burn, which flows past Arthuret proper.

The question is then what is Arm-/Arf-.  Although almost all sources had Ard- and not Arf-, I'm aware of the lectio difficilior requirement here.  So what is Arf-/Arm-?

There is Gaelic airm, 'place', and that has been proposed, but it's really not very convincing, given its total absence otherwise in Britain.  From Alan James' BLITON:

*arμ (f?)

Early Celtic *armā- > Br *armā-; O-MIr, G airm.

‘Place, location, whereabouts’.

Proposed by I. Williams, see PNCmb pp. 51-2, in [bellum] Armterid AC573 (in London, BL MS

Harley 3859). There is no other evidence for the word in P-Celtic, nor does the Goidelic form

seem to occur to as a place-name generic. If a Brittonic cognate had existed and survived, it 

would have fallen together as it did in Goidelic with adopted Lat arma ‘arms’ (Welsh arf). See

Arthuret Cmb, below.

a2) The river-name Armet Water MLo (Stow), PNMLo p. 75, SPN² p. 241, and the territorial

name Armethe Stg (Muiravonside), PNFEStg p. 38, could formally be + -ed if adopted early

enough by Northumbrian Old English speakers to retain –m- (LHEB §§98-100, pp. 486-93);

however, such a formation would be be unlikely to involve *arμ. An early hydronymic element

is possible, see ERN p. 149 (discussion of R. Erme Dev), and *ar in river-names.

b2) Arthuret Cmb PNCmb pp. 51-2 ? + -*tērïδ. Arthuret church stands on a prominent bluff

overlooking the Border Esk about 2 miles south of Longtown. Williams’s identification of the

battle-site with Arthuret is plausible, given the strategic location, though it should not be regarded

as certain. On the burgeoning of stories surrounding this battle in mediaeval Welsh literature, see

Rowlands (1990) pp. 109-14. See also discussion of Carwinley under cajr.

I have long maintained that the Arm-/Arf- spelling is a poetic development and does not represent a real place-name.  After extensive discussion with Dr. Simon Rodway of The University of Wales, I've gotten agreement on this point.  In addition, I had proposed the Arderydd "variant" might derive from either Ar-derydd, 'in front of the Derydd' or Ardd-derydd, 'the height of the Derydd.'  This last seemed the most reasonable to me, as the rotwyd or rhodwydd Arderys [sic] was a circular, earthen dyke fortification guarding a ford.  This is discussed by Bromwich in her Triads and by Sir Ifor Williams in a note to his Taliesin edition. At Arthuret this fortification was atop the hill adjacent to the ford over the River Esk. 

- Rachel Bromwich

The actual origin of rhodwydd is debated.  Ifor Williams thought it from rhawd + gwydd.  But I think the GPC now has it right, with rhod from the word for 'wheel', and gwydd being the same as in gwydd4, 'tumulus', cf. gwyddfa, 'height, eminence, promontory.'

Here is what Dr. Rodway had to say on the subject:

"I think the best explanation is *Ardd-derydd < *Ardo-torridus - both variants can derive from this, and there would be good motivation for alteration in order to avoid a car crash of dentals following syncope of the composition vowel.  (1) Arfderydd: dd and f sometimes interchange, e.g. afanc ~ addanc.  (2) Arderydd fricatives can be lost after r in post-syncope consonant clusters.  Analogy could have played a part in both forms - as it was famous as the site of a battle, arf 'weapon' might have seemed appropriate, and for Arderydd we have plenty of place-names containing ar 'in front of, opposite', e.g. Arfon, Arberth etc."

Thus the location of the Arderydd battle was the 'dry' stream at Arthuret.  

For the best discussion of the actual fortification at Arthuret, consult W.F. Skene's NOTICE OF THE SITE OP THE BATTLE OF ARDDERYD OR ARBERYTH

(https://archaeologydataservice.ac.uk/archiveDS/archiveDownload?t=arch-352-1/dissemination/pdf/vol_006/6_091_098.pdf):

"About a mile south from Longtown is the church and rectory of Arthuret, situated on a raised platform on the west side of the river Esk,which flows past them at a lower level; and south of the church and parsonage there rise from this platform two small hills covered with wood, called the Arthuret knowes. The top of the highest, which overhangs the river, is fortified by a small earthen rampart, enclosing a space nearly square, and measuring about 16 yards square." 

Do note, however, that Skene was wrong to look towards the Moat of Liddel as the actual site of Arderydd.  Liddel Strength, as it is otherwise known, is over a kilometer north of the northern boundary of Arthuret parish and is, needless to say, nowhere near the Dry Beck.

Incidentally, Dreon son of Nudd, another famous hero at the Arthuret battle, is likely a son of the Nudd mentioned on an early 6^th century tombstone at Yarrow Kirk. 

Not far west of the Carwinley of Gwenddolau on the coast of Galloway is the fort of Caerlaverock. The name of this fort is referred to in Welsh tradition as the ‘Lark’s Nest’ and it is said to have been the cause of the Battle of Arfderydd (Arthuret). But ‘lark’ is itself either a mistake or pun for the personal name Llywarch, in this case Llywarch Hen son of Elidir Lydanwyn. Llywarch was first cousin to Urien Rheged. Caerlaverock is, therefore, Caer Llywarch.

There is another interesting reference to a place in Cumbria that I might mention.  In the ‘Cambridge’ group of Historia Brittonum MSS., an interpolation tells us that Vortigern is said to have built “Guasmoric near Carlisle, a city which in English is called Palme castre.”  Palme castre has long been erroneously identified with the Old Carlisle Roman fort one mile south of Wigton in the parish of Westward.  There is a double-error in the Historia Brittonum, for Guasmoric itself is not the same place as the Palme castre fort.

Guasmoric must be Gwas Meurig, the “Abode of Meurig or Mauricius.”  This is clearly an attempt at rendering the Gabrosentum Roman fort in Cumbria at Moresby.  According to both Ekwall and Mills, Moresby (Moriceby, Moresceby) is Maurice’s By, Maurice being a Norman name and -by being Old Scandinavian for “farmstead, village, settlement”.  Whether we can propose an original Welsh Meurig underlying Maurice is questionable.  In all likelihood, the interpolation is late and Guasmoric represents Maurice’s By.  If originally a Meurig place-name, this may commemorate the 6th century Meurig son of Idno son of Meirchion, who married a daughter of Gwallog of Elmet. Cynfarch son of Meirchion may have left his name at the Mote of Mark in Dumfries. 

As for Palme castre, this is a place now called Plumpton (Plumton, ‘tun where plum trees grow’; see Ekwall) in Cumbria.  Directly between Plumpton and Plumpton Foot is the Voreda Roman fort.  Rivet and Smith (The Place-Names of Roman Britain) list the fort as being “at Old Penrith, Plumpton Wall, Cumberland, beside the river Petteril”.  Voreda means ‘horse’ in British.

As archaeology has shown us, there were two main centres for the Carvetii kingdom. One was the ancient tribal centre near Brougham, the Roman Brocavum, with its triple sacred henges at Eamont. One of these henges is actually called King Arthur’s Round Table and another the Little Round Table. There is evidence in the form of a concentration of inscriptions at Brougham that the primary Carvetii deity worshipped at these henges was a horned god (doubtless a stag, given that Carvetii means ‘People of the Stag’) named Belatucadros.

But there was also an important region called variously Erechwydd or Yr Echwydd, mentioned in connection with Urien, his sons, Gwallog son of Lleenog of Elmet and with Dunod Fwr. No wholly satisfactory identification of Erechwydd has yet been made, but it would seem to be somewhere in or close to Cumbria.

What we do know about Erechwydd is that the Er- prefix is not the definite article yr, even though the name is sometimes wrongly written ‘yr echwyd’ in the poetry, but a form of Ar-, as found in other place-names, e.g. Arfon. Ar- as a prefix originally meant ‘in front of’. But it came to have the senses of ‘upon, on, over, at, in, across from’.

The National Dictionary of Wales defines echwydd as ‘fresh (of water, as opp. to salt); fresh water’. However, although this meaning has been extrapolated from the contexts in which the word is used, no good etymology had yet been proposed.

I asked Graham Isaac if the word could come from ech, ‘out of, from’, plus a form of the Indo-European root *ued, ‘wet’. His response was:

“The etymology echwydd < *exs-wed-yo-, or *exs-ud-yo- (either would probably do it) seems plausible enough.”

The literal meaning would then be the ‘out-water’, but the sense of the word would be simply ‘flowing, fresh water’. Again, the Welsh texts which use this word leave no doubt that we are talking about fresh water emerging from springs or lakes.

So where was Erechwydd/Yr Echewydd, the ‘Place by the flowing, fresh water’? Our clue lies not only in the name of the region, but in the battles fought there between Dunod Fwr of the Dent region and Gwallog of Elmet against Urien’s sons. These engagements are recounted in the Llywarch Hen poetry. Given that Urien Rheged seems to have had his origin in Galloway (where we find Dun Ragit, the ‘Hill-fort of Rheged’), and both Dunod and Gwallog had kingdoms in southeastern Cumbria and just southeast of Cumbria, respectively, the most logical place to seek Erechwydd would be the twin valleys of the Eden and Petteril.

A Roman road led from the south up through the valley of the river Lune right past Dunod’s Dentdale. This road continued north to the Eden Valley. Another Roman road led west from Leeds and joined with the Lonsdale road. Gwallog could have taken this route to the Eden or he could have gone north up Dere Street and then cut over through the Pennines at Stainmore.

The Eden and Petteril Valleys were the heartland of the ancient Carvetii kingdom. Not only did the twin valleys provide the obvious natural route from Carlisle towards Lancaster and York, the area has been shown to have supported a widespread and occasionally dense pattern of rural settlement in the Roman period.

It is even possible that Erechwydd as a regional designation can be more precisely localized within the Eden and Petteril Valleys. The headwaters of the Petteril lie just west-northwest of Eamont. We have already discussed the importance of Eamont with its sacred henges. The river Eamont (a back-formation from the name Eamont itself, from AS ea-gemot, ‘river-meet’, i.e. confluence) and Lowther join at Eamont Bridge and continue for a short distance eastward to the Eden. There was also, of course, a nexus of Roman roads at Eamont.

In my opinion, the Anglo-Saxon place-name ea-gemot/Eamont may overlie an original British Echwydd. Ekwall thought Eamont refers to the confluence of the Eamont and the stream from Dacre, although given the location of the Brougham/Brocavum Roman fort at the juncture of the Eamont and Lowther, it makes much more sense to see this ea-gemot as the confluence of the latter two rivers. If I am right, then Arechwydd was the Eamont area, specifically the land at and around the Brougham fort and the three Carvetii henges. The ‘out-water’ would be a reference specifically to the Eamont, which is formed by the outflow from the Ullswater, the second largest lake in Cumbria.

Just a few miles south-southest of Eamont is the Lyvennet Beck, a tributary of the Eden. This has been identified with the Llwyfenyd over which Urien is said to have been ‘ruler’ (Welsh teithiawc).

In the Strathclyde genealogy proper, we find a Garbaniaun son of [Ceneu son of] Coel Hen. This Garbaniaun has a son named Dumngual Moilmut or Dyfnwal Moelmul.  Both names are, rather transparently, forms of the Dalriadan prince Gabran (Garbaniaun shows a metathesis of Gabran, plus a territorial suffix, as in Gwrtheyrniaun, a region named for Gwrtheyrn/Vortigern; cf. with Garban for Gabran in the Irish Book of Lecan) and his son Domnall.  The Bran son of Dumngual/Domnall of the British pedigree is probably the attested Bran son of Aedan son of Gabran.

I should note that scholars have preferred to see in Garbaniaun the Roman Germanianus.  However, Germanianus is a rare Latin name, and why it should have appeared among the Starthclyde Britons at this time is very hard to explain.  There was a 4th century Prefect of Gaul bearing this name, but no one else of any note, so far as our records tell us.

While we need not take these apparent intrusions of Irish Dalriadan royal names into the British Strathclyde genealogy at face value, they probably do indicate the existence of marriage ties between the Strathclyde Britons and their neighbors, the Dalriadans.  Such marriage ties are hinted at in the records which pertain to the history of Scottish Dalriada (see John Bannerman’s Studies in the History of Dalriada, Edinburgh and London, 1974).

According to the Historia Brittonum, the British name for Bamburgh was either Din (“Fort”) Guayrdi or Din Guoaroy.  The name has remained a problem for philologists and no satisfactory etymology has been proposed.

I would suggest the Welsh word gwyar, ‘blood’, plus an ethnonymic suffix. In this case, Gwyar is a proper name, possibly the mother of the famous Arthurian hero Gwalchmai. Alan James has informed me that the medial syllable would have been syncopated, so we could expect a form such as *Gwyardi.  This fits Guoaroy better than, say, Welsh gwaered, ‘declivity, downward slope.” In the case of Guoaroy, the 'o' could be a miscopying of 'ꝺ', 'insular d'.

Din Gwyardi, the ‘Fort of the People of Gwyar.’

William of Malmesbury said that Gwalchmai had been buried at Ros (Rhos) in Wales.  This may be a relocation for Ross Low at Bamburgh. 

The Welsh Triads place Gwalchmai’s grave on the Parret in Somerset, but this is doubtless because Gualganus, a form of his name, was wrongly linked to the Anglo-Saxon Chronicle’s Cenwalh, who fought the British at that river.