EMAIN ABLACH AND ARBERTH, WALES

When it comes to localizing the Irish Emain Ablach, the tradition that identifies this particular Otherworld with the Isle of Man may be correct.  Manannan, due solely to his name, would naturally have been associated with the island.  I have shown in my Arthurian battle researches that the Welsh Manawydan was in or on the border of Manau Gododdin when he fought at the Tribruit (North Queensferry trajectus on the Firth of Forth). Manau Gododdin has left some local place-names - Slamanann, 'the mountain of Manau', Clackmanann, 'the stone of Manau' and, possibly, Dalmeny.

However, we have no independent place-name evidence that the Isle of Man was ever called Emain Ablach.

Manawydan, the Welsh counterpart of Irish Manannan, is placed at the court of Arberth (modern Narberth) in Dyfed with the horse goddess Rhiannon (Epona Regina). Arberth is the place 'at/near/beside the perth.'  Perth has an interesting etymology.  

Narberth Castle, Dyfed, Wales

Here is the listing from the GPC:

perth

[yr e. lleoedd Crn. (Pen)berth, (Pen)perth, e. tref Perth (Gael. Peart) yn yr Alban, Gwydd. C. ceirt (< *queirt) ‘pren (afalau)’: ?< Clt. *ku̯erku̯-t- (?< IE. *perku̯u- ‘derwen’, cf. Llad. quercus ‘derwen’, Sans. parkatī- ‘pren ffigys sanctaidd’, H. S. furh (> S. Diw. fir)) neu ?cf. pren, prysg]

eb. (bach. perthen, ll. perthenni) ll. perthi, perthydd, perthau.

Gwrych, clawdd, sietin; llwyn, prysglwyn, dryslwyn, draenllwyn; tir prysglwyn, cefn gwlad; jyngl; hefyd yn dros. ac yn ffig.:

hedge; (thorn-)bush, brake, thicket, copse, coppice; (the) bush, countryside; jungle; also transf. and fig. 

It will be noticed that the cognate to perth in Irish is ceirt.  And here is the eDIL listing for ceirt:

1 ceirt

Cite this: eDIL s.v. 1 ceirt or dil.ie/8502

n apple-tree ; the letter q in Ogham (see Vendryes, RC xliv 313 ff. ): c.¤ .i. abhall, O'Cl. tri foilcheasta inn oghaim .i. c.¤ ┐ gedal ┐ straiph. In baile i mbi c ria n-u is queirt is scribhtha and, ut est cuileand ┐rl., Auraic. 440 . queirt dano, is o chrand rohainmnighead .i. abhull ut dicitur. clithar boaiscille .i. elit gelt cuert .i. aball, 1184 . ? Cf. bille .i. c.¤ , Corm. Y 186 .

So, in other words, from the Irish perspective, a Manawydan, i.e. Manannan, at Arberth would be the sea god at the apple thicket.'

We should recall, of course, that there was a 7th century Arthur of Dyfed. He belonged to a dynasty that had been founded by the Irish Deisi. 

Should Arberth, then, be considered an Avalon?

A PAGAN GRAIL NEAR ARBERTH?

The cauldron of Diwrnach is taken to Dyfed, where it is left at the house of Llwydeu son of Cil Coed. Llwydeu or Llwyd has been linked to Ludchurch, Welsh Eglwys Llwyd, hard by the stream of Cil Coed in Pembroke. The Otherworld castle of Llwyd of Cil Coed is probably the ancient fort that stands atop the hill overlooking Ludchurch.


The notion that Llwyd may be a Welsh version of the Irish hero Liath son of Celtchair, whose name is preserved in the famous fairy hill in County Longford called Bri Liath, is certainly significant. Bri Leith was for a time the home of the goddess Etain Echraide, that is, Etain ‘Horse-rider’. Midir (*Medio-rix, ‘King of the Middle’, i.e. of Midhe), the god who owned Bri Liath, possessed a magical cauldron, which was stolen from him by Cu Roi. The fortified hill at Ludchurch may well have been thought of as the Welsh counterpart of Bri Liath in Ireland and, hence, became the respository of the horse goddess’s patera. This is especially true since the Dark Age ruling dynasty of Dyfed, as mentioned already above, was of Irish Dessi origin.

A NEW ETYMOLOGY FOR ISAMNION, AND AN OLD ETYMOLOGY DEFENDED FOR EMAIN

Navan Fort

ISAMNION

The place-name Emain (and variants; see http://www.placenamesni.org/resultdetails.php?entry=10437) has given toponymists fits for a very long time.  The standard folk etymologies "Twins" or "Neck-Pin" found embedded in Irish myth or medieval glossaries are not considered to be satisfactory  The debate continues as to whether Emain should be associated with Ptolemy's Isamnion promontory.  

Currently, the only proposed derivation that is deemed somewhat acceptable is the one proposed first by T. F. O'Rahilly and recently favored by Gregory Toner. They hold that Isamnion, minus the standard Greek suffix, comes from *isa-monis, 'holy mountain/mound.'  Professor Jurgen Uhlich has demonstrated to me how *isa-monis could have yielded Emain:

"O’Rahilly’s suggestion that ‘Emain might go back to *Isamonis or the like; the syncopated genitive, Emna, would be analogical (cf. Lemain, gen. Lemna)’ (EIHM 13) is at least phonologically sound – and there would not have been anything ‘shortenend’ at the beginning, but *isa- regularly > e- by syncope. O’Rahilly also adds a reference to Pokorny in ZCP 21 (1940), 127, where Isamnion is compared to have yielded first Emnae, which would then have been transformed to emain under the influence of the word for ‘twins’ – which is where your semantic considerations might come in in support."

There is nothing wrong with this proposed etymology, other than the fact that *isa-, so far as I can tell, is not found elsewhere in Celtic.

Irish has is, 'below, under' (see eDIL).  Welsh has is also, with the same meaning.  The IE root is believed to be *ped-su.  This element has been proposed for Isamnion, but in combination with a different construction for the remainder of the place-name.  Warner (in https://www.academia.edu/7150776/Ptolemys_Isamnion_promontory_rehabilitation_and_identification) discusses the proposed etymology of Patrizia de Bernardo Stempel:

"...if [the word means] ‘the point standing below’, it would be the yo- adjective derived from the compound formed by the adverb yielding Irish ís ‘below’ and the Goidelic out-come of *sta-mno- , alternatively, it may represent the gen. plur. of the similarly formed name of the inhabitants,‘(promontory) of the people standing below.’"

This seems an awkward and ambiguous meaning for a place-name.  What does the akron stand below?  de Bernardo Stempel suggests (2007 paper) that 'standing below' is to be interpreted with reference to Robogdion promontory (the next 'above' or to the north) from the standpoint of the mariner.  I find this explanation to be strained and not very plausible.

It is unfortunate that an extant Irish word has not been brought into connection with Isamnion.

Take off the Greek ending, and cf. Isamn- to the words/definitions pasted below.  Note especially essamin (as smaller vowels are more easily dropped) and es(s)amnae. The same word yielded a personal name meaning 'the fearless one.'  In the case of the Isamnion Akron, we would be talking about a tribal name, giving us something like the 'the Promontory of the Fearless Ones.'

It would leave Emain unanswered as a place-name, of course. We will consider the latter below.

fear Proto-Celtic *obno-, Gaulish -obnus, -omnus ‘fear’, [Primitive Irish probably omno, according to Professor James Mallory), Early Irish omun ‘fear’, Welsh ofn ‘fear, terror, fright, anxiety; awe’, Cornish own ‘fear’, Breton oun (Middle Breton), aon ‘fear’

The relevant eDIL listings:

es(s)amain
Cite this: eDIL s.v. es(s)amain or dil.ie/20379
Forms: esamnu

adj (omun) fearless, bold, daring: am essamin do thabirt testassa airibsi I am bold to give evidence Wb. 16a26 . (with gen.) am essamin se precepte fearless in preaching 23b7 . cia fiu esámain (gl. quam temerarium) Ml. 27d8 . nibbát ecal ocal opond esamain be neither timid nor violent, precipitate and rash IT i 213.19 . is esamain (the question) is a bold one (?) Anecd. ii 12.3 . (Adv.) tanic co hesamain ┐ confergach i n-agaid Patraic, Trip. 456.1 . co dána ┐ co hesamain, PH 5137 . Compar.: it essamnu de they are the more fearless Wb. 23b12 . esamnu (gl. audentior) Ml. 29a12 . esamnu...du écnduch dǽ bolder to blaspheme God ib. 11 . Cf. esamain .i. lasamain, ut est oc esamain airbert i cuairt, O'Dav. 772 (cf. 799 ). esomain (essomuin v.l.) .i. obann no lasamain, ut est is esomain naillech .i. is lasamain fri tabairt luighi `he is a sudden swearer, i.e. he is on fire to take an oath' 781 . pn As sobriquet: mac Elcaib Essamain `son of E. the fearless' MacCongl. 75.8 ; 119.24 (Esomain). Cf. es(s)omon.

es(s)omon
Cite this: eDIL s.v. es(s)omon or dil.ie/20552
Forms: essemón

n I A cessation of fear, hence a respite; a truce: ronbith essoman lais co comtastar fer do let us have a truce with him TBC-I¹ 1235 (essomon, Eg. essemón LU). essomon tri laithi ┐ teora n-aidchi, RC xxv 346.14 . dorónsat essomon trī laa ┐ n-oidchi etarru, ZCP viii 312.17 . ailthus E. imm esomon asks them for a truce 313.5 . gaid essomon bliadna do a respite, delay Anecd. iii 59.3 . digensat essomon fri soegul nO., 60.13 .

II A welcome: fearus Eochaidh esemain (.i. fáilti) fris, ZCP v 523 x . fuidis Guaire essomuin friu, ACL iii 3.9 . easomain .i. fáilte, O'Cl. Cf. es(s)amain, es(s)amnae.

es(s)amnae
Cite this: eDIL s.v. es(s)amnae or dil.ie/20380
Last Revised: 2019
Forms: esamni, esamni

n iā, f. (es(s)amain)

(a) boldness, confidence: ní nuae dunni ind esamnae so dit ditin-siu (gl. de protectione tua fiducia) Ml. 110d12 . a s. int [leg. inn] esamni ind firiéin dut chreitim-siu (gl. fiduciam) 'the confidence of the righteous man to believe thee', 142b2 . in n-esamni nad lugimem (gl. confidentiam non minimam) 14b10 . in esamni (gl. impudentiam) 25b9 . tre essamni cumachti through confidence of power Wb. 17a7 . d s. ní rabæ di esamni Duaid David had not the boldness Ml. 33c17 . is danatus ┐ is essamna mor do Throianaib tuidecht co Grecu, LL 399b36 = TTr.² 245 . dál fria hessamna a tryst with boldness (?) MacCongl. 95. 7 . Cf. fornes fir iar mbag esamna the declaration (i.e. test, proof) of a man (comes) after (his) vaunt of confidence (?) O'Dav. 799 = forneis iar mbagh esomna, 781 . oc esamna airbert .i. ar in airbert lasamain, 799 = oc esamain airbert, 772 . As npr. m.: Rogen mac Essamna Emna maic Blaith, ZCP viii 292.13 .

(b) a welcome: ferais esomni fris `she made him welcome' RC xi 446.68 = feraiss eussoman (.i. failti) friss ZCP iii 251.8 . ferais Medb esomne fris 'Medb welcomed him', BDC 104.18 . ferais Cormac esomne frie 'Cormac welcomed her', BDC 110.96. Cf. es(s)omon.

ess-, es-, e-, é-
Cite this: eDIL s.v. ess-, es-, e-, é- or dil.ie/50223
Last Revised: 2019
prefix etymologically identical with the preverb as- and the preposition 7 a 'out of' ( GOI § 834 ). In earlier formations ess- became e- followed by gemination before stops and é- before r, l, n. In many cases the older form of the prefix is difficult to distinguish from é-, an allomorph of privative an-, in the written language (e.g. éccra, 1 etal). Already in the Old Irish Glosses the form es(s)- is generalised to preconsonantal position (e.g. escarae, esba(e), esrantaid). ess- forms privative and negative compounds with nouns and adjectives. Nouns prefixed by privative ess- can become adjectives as in énirt (nert), es(s)amain (omun). o- and ā-stem nouns can become i-stem adjectives as in the aforementioned examples, or remain unchanged like ecal (gal). The adjectival function may be eclipsed by substantivised use (e.g. esert, eisrecht, eistech). With adjectives the prefix has negating function and often seems equivalent to an- + adjective (e.g. esidan, ainidan; eisinnraic, aininnraic). For examples see the individual entries.​

omun , ómun
Cite this: eDIL s.v. omun , ómun or dil.ie/33853
Last Revised: 2019
Forms: ómun, ómun, hómon, hómun, ómun, homun, homun, omun, (h)omno, omna, ómnu, úamun, omun, uamain, oman

n u and o, m. O.Ir. n s. ómun , Ml. 33c7 , 59a18 , (with art.) int ómun , 42d9 ; hómon, Wb. 6a7 ; hómun, Wb. 16a21 ; d s. ómun , Thes. ii 21.35 ( BCr. 33b14 ); but n s. homun, Ml. 79b2 , 96a10 ; a s. homun, 55c10 ; omun , Wb. 29d15 ; g s. (h)omno, Ml. 128b7 , 9 . omna, Laws text. a p. ómnu, LU 599 (o-st.). In Mid.Ir. always ŏmun (as attested by rhyme), pos- sibly the orig. form; the Gaulish and Brythonic cognates show ŏ, cf. W. ofn (see Pedersen Vgl. Gr. i 49 ). From ómun comes the Mid.Ir. variant form úamun, which is often used indifferently with omun and finally supersedes it. In Mid.Ir. treated as o-st. (occas. as fem., d s. uamain, TBC-LL¹ 1345 , LL 34a17 ). úamhan omhan (o, m.), IGT Dec. § 17.19 . úamhain omhain (f.) § 150.6 .

I As subst. fear, the state of being afraid , absol. or with gen. of cause of fear; transl. Lat. timor, metus in Glosses. oman .i. ecla, O'Dav. 1319. int ómun fil la suidib, gl. cum metuant, Ml. 42d9 . nos gébad ule crith ┐ uamun, FA 14 ( LU 2070 ) = oman, LB. ecla nó uamun, CRR 20. ros-lín oman, PH 1121. uaman, 1244 . fer . . . cen omon, TBC-LL¹ 31. foce[i]rd a n-oman cach tuaith casts . . . into dread, Fianaig. 12.15 . With obj. gen.: robú mór a homun lium-sa (gl. metuebam), Ml. 96a10 . ómun Dáe the fear of God, 33c7 , cf. 55c10 , Wb. 16a21 . omon Fíadat, ZCP v 499 § 4. ar ecla ┐ ar úamun Conculaind, TBC-LL¹ 1608 . cromfait a cinnu ar m'ecla ┐ ar m'uamain, 1345 . cid do chana dúib ar n-uaman-ne do beth foraib? what causes you to be afraid of us? PH 3399. tre omun fochide, Wb. 29d15 . ar ómun pían n-iffirnd, FA 33 ( LU 2282 ) = ar uamun, LB. omun báis, PH 3830. (with proleptic poss. pron.) ata a uaman oruinde co n[d]ingna sé fasach do crichaib S. we are afraid he will devastate, ZCP vi 295.8 . tig an bás . . . an tan is lugha bhíos a uamhan orra, TSh. 546. eagla Dé . . . tar omhan chréatúra, 6819 . Rarely with subj. gen.: in toirsi dorigne Ísu, noco n-ar a omun fén doróine `through fear for himself', PH 3026 . dorochar-sa fri talum for mét mo omain, 3730 . g s. (a) ermitiu omno Dáe, Ml. 128d9 . cor omna `a contract of fear' (one entered into under intimidation), Laws iv 208.32 Comm. fonascc omna, v 376.3 , cf. 186.7 . (b) boegal i n-ecmais omain daring in the absence of [cause for] fear, PH 3010 ; cf. 3730 . pl. in sense terrors, causes of fear: frisbered omnu huain, ACC 16 ( RC xx 164 ) = ómnu, LU 599 , cf. O'Dav. 1319. all fri homna, Anecd. i 68 § 155. sciath fri homna i n-ernaidhe, AFM i 524.16 (poem).

As pred., is o.¤ LA fears, is apprehensive : is omun lium-sa do chlod-sa I fear thou art vanquished, PH 1793. is omun leam-sa . . . saethar do chaithimh fria gnimradh idhul, Marco P. 1 . is omain leis a mac do facbail, Laws ii 164.22 Comm. Ml. 96a10 (quoted above). As pred., is o.¤ DO is a cause of fear to: ní húamun duit misi 'you need not fear me', FDG² 10.304 . With cop. and dat. pron.: isam ómun ar a aisnéis, LU 5349 = omun , TBC-LL¹ 873 YBL. isim omun ar in mbein `I have fear of the woman', RC xvii 132.22 , cf. MacCarthy 398.4 .

II As adj. generally as pred.

(a) afraid, apprehensive : dia mbad úamain ṁ'ḟer if my husband were a coward, TBC-LL¹ 36 = imeaclach St., omhnach Add. nibtis omain neich they feared no one, ZCP iii 38.25 .

(b) fearful, terrible : bás ┐ cach ní homain, Corm. p. 13 s.v. ceal (g s. ?). ge fuair bas tre re[i]m n-uaban `in his dread career', Celt. Rev. iv 214.6 (possibly g pl.).

Compds. Huamanchenn (npr. m.), ZCP viii 326.20 . mac Huamanchind, ib. 23 . omungnath (< gnáth `wont, habit') a state of fear? o. dorognad, Ält. Ir. Dicht. i 41 § 32 ( Rawl. 115b27 ), cf. Wortk. § 7. Marstrander RC xxxvi 339 takes the word as npr. m., comparing Smirgnath ( § 38 in same poem), Faelgnath, Domungnath, BB 158a27 .

es(s)amnae
Cite this: eDIL s.v. es(s)amnae or dil.ie/20380
Last Revised: 2019
Forms: esamni, esamni

n iā, f. (es(s)amain)

(a) boldness, confidence: ní nuae dunni ind esamnae so dit ditin-siu (gl. de protectione tua fiducia) Ml. 110d12 . a s. int [leg. inn] esamni ind firiéin dut chreitim-siu (gl. fiduciam) 'the confidence of the righteous man to believe thee', 142b2 . in n-esamni nad lugimem (gl. confidentiam non minimam) 14b10 . in esamni (gl. impudentiam) 25b9 . tre essamni cumachti through confidence of power Wb. 17a7 . d s. ní rabæ di esamni Duaid David had not the boldness Ml. 33c17 . is danatus ┐ is essamna mor do Throianaib tuidecht co Grecu, LL 399b36 = TTr.² 245 . dál fria hessamna a tryst with boldness (?) MacCongl. 95. 7 . Cf. fornes fir iar mbag esamna the declaration (i.e. test, proof) of a man (comes) after (his) vaunt of confidence (?) O'Dav. 799 = forneis iar mbagh esomna, 781 . oc esamna airbert .i. ar in airbert lasamain, 799 = oc esamain airbert, 772 . As npr. m.: Rogen mac Essamna Emna maic Blaith, ZCP viii 292.13 .

Isamnion was once linked to Rinn Seimhne, the modern Islandmagee.  But scholars are now (based on their reading of Ptolemy's questionable coordinates) opting for the Cooley Peninsula (or Cooley Point) much further south.  I still think Seimhne makes for an excellent relic of Isamn-.  We would need to opt for Ptolemy's Logia being Carnlough and his Winderis being Glenarm.  Both Isamnion and Seimhne are on the northeastern coast of Ireland. But if Seimhne is not Isamnion, the resemblance between the two headland names is still remarkable. In fact, it is downright uncanny. I feel that ignoring the apparent congruence would be to our peril.

Islandmagee, Co. Antrim, Ireland 

EMAIN

Epona Between Two Foals, Marble Votive Relief

Epona with Male and Female Foals, Bronze Statuette

Epona and Two Horses

Despite all other theories, there is no good reason to refuse to accept the meaning of 'Twins' for Emain Macha. Macha gives birth to two foals in her capacity as horse goddess.  Their names are Fir and Fial.  Fir is Irish fir, 'true', with a form firi.  This is cognate with Welsh Gwri, the son of the horse-goddess Rhiannon.  Rhiannon or the 'Divine Queen' is a reflection of the Regina title given to Epona.  Epona is frequently depicted with either two or four horses.

The problem comes in when we ask what could such horse twins possibly have to do with Emain Ablach, the Otherworld island of apple trees?  I mean, other than the fact that horses love apples!

Well, in the Welsh myth, Rhiannon becomes the wife of Manawydan son of Llyr, the counterpart of the Irish Manannan son of Lir.  The sea god is himself closely associated with horses (as was his Classical counterpart, Neptune/Poseidon. Manannan's favorite horse was Enbarr, whose name (according to the eDIL) means 'foam, froth.'  In this case, the foam/froth of a wave. A fitting name for the horse of the sea god. One of the gessi of Cuchulainn was that he was not to see the stud-team of Mac Lir (in AIDED CON CHULAINN). This stud-team would be two horses used to draw the god's chariot over the sea. 

Poseidon was given the epithet Hippios, 'of the horses'.  Pausanias describes one of his temples thusly:

Pausanias, Description of Greece 8. 10. 2 :

"By the foot of the mountain [Alesion in Arkadia] is the sanctuary of Poseidon Hippios, not more than six stades distant from Mantineia. About this sanctuary I, like everyone else who has mentioned it, can write only what I have heard. The modern sanctuary was built by the Emperor Hadrian, who set overseers over the workmen, so that nobody might look into the old sanctuary, and none of the ruins be removed. He ordered them to build around the new temple. Originally, they say, this sanctuary was built for Poseidon by [the mythical brothers] Agamedes and Trophonios, who worked oak logs and fitted them together. They set up no barrier at the entrance to prevent men going inside; but they stretched across it a thread of wool. Perhaps they thought that even this would strike fear into the religious people of that time, and perhaps there was also some power in the thread. It is notorious that even [the mythical hero] Aipytos, the son of Hippothoos, entered the sanctuary neither by jumping over the thread nor by slipping under it, but by cutting it through. For this sin he was blinded by a wave that dashed on to his eyes, and forthwith his life left him. There is an old legend that a wave of sea-water rises up in the sanctuary . . . at Mantineia the sea rises after a very long distance, and quite plainly through the divine will."

For more on Poseidon's relationship to horses, see https://www.theoi.com/Cult/PoseidonCult.html.

We thus seem to have ample reason to connect Manannan with horses in general, and with two horses in the case of his chariot (see https://books.google.com/books?id=Uz1MDAAAQBAJ&pg=PT142&lpg=PT142&dq=chariots+in+the+tain+were+for+two+horses&source=bl&ots=kZ-z7MUmMY&sig=ACfU3U0mL9QPzDsTxml--77zUV_g8nvcHw&hl=en&ppis=_e&sa=X&ved=2ahUKEwjFtMbn2b3nAhXMIDQIHTq6AqoQ6AEwCXoECAUQAQ#v=onepage&q&f=false).  The hero Cuchulainn had two supernatural horses for his chariot, one being named 'The Grey of Macha' (see https://en.wikipedia.org/wiki/Liath_Macha_and_Dub_Sainglend). The 'Twins of Ablach', then, were presumably Otherworld apple trees that belonged to Manannan's divine horses. However, as with the Welsh borrowed form Afallach, Ablach could be used as a personal name in the Irish:

1 ablach

Cite this: eDIL s.v. 1 ablach or dil.ie/69

adj o, ā (aball) having apple-trees: lis a.¤ crandcaem, Acall. 2738 v.l. Usually in Emain A.¤ , a legendary island, equated by Chotzen, Études Celt. iv 255 ff. , with Avalon; see also under Emain. i mbrugh Eamhna abhlaighe (:thairrngire), DDána 120.25 . As n. pr. mad cuablaich (mocuablaigh, v.l.) a.¤ arann /mad comuirgil / muirgel manand, IT iii 34.5 (Meyer, Meyer Prim. 16.12 ). ?romarbsat i cath abla a.¤ lugach is ler . . ., LL 204b38 . A.¤ ingen Domnaill, LL 17270 . As n loc. Ablach Owley , Hog. Onom.

EMAIN ABLACH AND ARBERTH

When it comes to localizing the Irish Emain Ablach, the tradition that identifies this particular Otherworld with the Isle of Man may be correct.  Manannan, due solely to his name, would naturally have been associated with the island.  I have shown in my Arthurian battle researches that the Welsh Manawydan was in or on the border of Manau Gododdin when he fought at the Tribruit (North Queensferry trajectus on the Firth of Forth). Manau Gododdin has left some local place-names - Slamanann, 'the mountain of Manau', Clackmanann, 'the stone of Manau' and, possibly, Dalmeny.

However, we have no independent place-name evidence that the Isle of Man was ever called Emain Ablach.

Manawydan, the Welsh counterpart of Irish Manannan, is placed at the court of Arberth (modern Narberth) in Dyfed with the horse goddess Rhiannon (Epona Regina). Arberth is the place 'at/near/beside the perth.'  Perth has an interesting etymology.  

Narberth Castle, Dyfed, Wales

Here is the listing from the GPC:

perth

[yr e. lleoedd Crn. (Pen)berth, (Pen)perth, e. tref Perth (Gael. Peart) yn yr Alban, Gwydd. C. ceirt (< *queirt) ‘pren (afalau)’: ?< Clt. *ku̯erku̯-t- (?< IE. *perku̯u- ‘derwen’, cf. Llad. quercus ‘derwen’, Sans. parkatī- ‘pren ffigys sanctaidd’, H. S. furh (> S. Diw. fir)) neu ?cf. pren, prysg]

eb. (bach. perthen, ll. perthenni) ll. perthi, perthydd, perthau.

Gwrych, clawdd, sietin; llwyn, prysglwyn, dryslwyn, draenllwyn; tir prysglwyn, cefn gwlad; jyngl; hefyd yn dros. ac yn ffig.:

hedge; (thorn-)bush, brake, thicket, copse, coppice; (the) bush, countryside; jungle; also transf. and fig. 

It will be noticed that the cognate to perth in Irish is ceirt.  And here is the eDIL listing for ceirt:

1 ceirt
Cite this: eDIL s.v. 1 ceirt or dil.ie/8502
n apple-tree ; the letter q in Ogham (see Vendryes, RC xliv 313 ff. ): c.¤ .i. abhall, O'Cl. tri foilcheasta inn oghaim .i. c.¤ ┐ gedal ┐ straiph. In baile i mbi c ria n-u is queirt is scribhtha and, ut est cuileand ┐rl., Auraic. 440 . queirt dano, is o chrand rohainmnighead .i. abhull ut dicitur. clithar boaiscille .i. elit gelt cuert .i. aball, 1184 . ? Cf. bille .i. c.¤ , Corm. Y 186 .

So, in other words, from the Irish perspective, a Manawydan, i.e. Manannan, at Arberth would be the sea god 'next to the apple thicket.'

We should recall, of course, that there was a 7th century Arthur of Dyfed. He belonged to a dynasty that had been founded by the Irish Deisi. Should Arberth, then, be considered an Avalon?

COULD EMAIN ABLACH = EMAIN MACHA?

We should note in the above GPC definition "Llad. quercus ‘derwen'."  This is 'Latin, quercus 'oak.'

I have before me Damian McManus' A GUIDE TO OGAM.  His entry on Q Cert, the ogam letter, reads as follows:

"This letter name is undoubtedly related to Welsh perth 'bush' and cognate with Latin quercus < *kwerkw-, ultimately from the root *perkw."

At Emain Macha a huge ritual structure was made of wood, with oak posts and a large central oak trunk.  It is a reasonable speculation to relate this structure (destroyed and infilled soon after its construction) to the Red Branch house of Irish legend.  In any case, the two Emain localities are unique in that both seem to have something to do with sacred trees.  

Emain Macha - oak/quercus

Emain Ablach - apples/ceirt

The question we should ask at this point, especially given what we've just learned about Manawydan and Rhiannon's Arberth as a possible Avalon, is whether E. Ablach and E. Macha should be identified with each other. This would have happened through a learned error, I suppose, in which a monkish scribe took the quercus of Emain Macha to be Irish ceirt, 'apple', leaving the way open for the later substitution of the word ablach.  

Another strange way of looking at the word Emain is to use it in its single application.  It can be 'twin', as in one of a pair.  So Emain Macha could be read 'the twin of Macha', while Emain Ablach would be 'the twin of Ablach.' Macha's twin is Ablach, and Ablach's twin is Macha.

All of this is, admittedly, a bit far-fetched.  But strange things happen during the creation and evolution of mythology.   

MUIREDACH MUINDERG, KING OF ULSTER - AND MATERNAL GRANDFATHER OF ARTHUR?

If I'm right, and Arthur's father Sawyl Benisel of Ribchester had as his wife a daughter of the Irish Dal Fiatach king Muiredach Muinderg, I thought it would be interesting to take a look at where the latter had his chief court.  We cannot, of course, know whether this Irish princess was Arthur's mother. But it is a distinct possibility.  

Fortunately, the Irish place-name and national monument folks have done their homework.  The case has been made for two rulings centers for Muiredach, both in Co. Down: Drumnabreeze and Dromorebrague.  I have posted below links and excerpts from the relevant Websites.

***

http://www.placenamesni.org/resultdetails.php?entry=11215

Drumnabreeze, County Down

Show the Map

Origin

perhaps Ir. Dromainn Brís ‘ridge of (the) high ground’

Background

7km E of Lurgan

par: Magheralin bar: Iveagh Lower, Upper Half

Drumnabreeze was part of the O’Lavery district of Moira in 1609 (CPR Jas I 395a), and although it was let to Stephen Haven of Dromore by 1631 (Inq. Ult. (Down) §16 Car. I), Laughlin Roe O’Lawry was still resident in 1663 (Sub. Roll Down 273).  It is on rising ground east of the Lagan valley, with a hill in the centre 297 feet high, from which the view ‘ranges considerably over the flat ground adjoining the Lagan’ (OSNB).  The first element seems to be one of the words for ‘ridge’, droim or its derivatives dromainn or droimne. John O’Donovan suggested that the final element was brí ‘upland’ (gen. plural breg later brígh).  As noted by the editor Margaret Dobbs, a name like Droim na mBrígh invites comparison with Druimne Breg ‘ridge of the high ground’, the location of Uachtar, the earliest chief site of Dál Fiatach in east Ulster (Descendants Ir xiii 324, 336, 338).  However, there is no archaeological evidence for an important historical settlement in this area, and Dromorebrague in the parish of Aghaderg is a more likely location.

If the element brí is accepted, it is possible that the final [z] of the modern pronunciation is due to the addition of English plural -s, which is usually pronounced as [z] after a vowel.  There are forms without -s in the Census of 1659 and the Down Subsidy Roll (1663).  A plural -s has been added to this element in English spelling in other place-names, for example Mac

Muiris na mBrígh,‘MacMorris of Brees’ Co. Mayo (Misc. Ir. Annals 1406).  Dean Mooney mentioned Brees Castle from Caisleán na mBrí, Brees in Clanmorris, Mayo, referred to as (is) na Brighibh hi ccloinn Muiris in 1595 AD (AFM vii 1988).  Another possibility is that [z] represents the genitive ending of bríoghas, an unattested collective form of brí (cf. Ó Máille 1989-90). This would be bríos in modern spelling.

***

NOTE:  There once existed a fort at Drumnabreeze.  I found this information here:

https://www.jstor.org/stable/20566485?seq=1

The argument in favor of Dromorebrague is unsupportable, in my opinion, as one must opt for a place-name with an intrusive element (mor, 'big, great'). Drumnabreeze is, on the other hand, quite perfect for our needs.

Still, for the sake of full disclosure, the reader will find below the relevant entries on Dromorebrague and its hillfort.

***

http://www.placenamesni.org/resultdetails.php?entry=11800

Dromorebrague, County Down

Show the Map

Origin

Ir. Droim Mór Breá ‘big ridge of the high ground’

Background

Dromorebrague must be the townland of Shankill district called in the early 17th century Ballydromore, although it is now known locally just as Brague. There is a hill-fort at 489 feet near the north-east boundary of the parish (ASCD 148).  The townland is high ground and Droim Mór ‘big ridge’ would need no explanation as the basic name of the place. The two place-names in this parish including the element brague have been understood as containing bréige, the genitive of bréag ‘a lie’, and thus having the sense of an adjective ‘false’ (Glancy 1956, 78).  The term buachaill bréige ‘false lad’ was often used for a standing stone which might look like a person when seen from far off (Joyce ii 435; Ó Ceallaigh 1952-3(b) 36-7).  The -brague element has also been studied by Dean Mooney.  He thought that Dromorebrague derived from Droim Mór Bréagtha ‘big ridge for playing games’, where -brague represented the genitive of bréagadh, ‘deceiving’ and thence ‘beguiling’, used in this case of a hill where people once assembled to play games (Mooney 1956(a), 26-7)  (Mooney MS 11).  However, this would strain the meaning of bréagadh, and bréagtha would be unlikely to be anglicized brague. It is possible that -brague could represent the genitive singular or plural breagh (standard breá) of brí ‘height, plateau’, with final -gh  delenited to [g] in the modern form.  Dromorebrague might then be part of the district Droim Breá or Dromanna Breá ‘ridge (or ridges) of the height (or heights)’, which was the earliest settlement of Dál Fiatach in east Ulster. 

https://apps.communities-ni.gov.uk/nismr-public/Details.aspx?MonID=7762

On top of a hill with excellent views. The site is impressive, although some 52m of the N extent have been removed, where the bank & ditch would have been situated. The site, 71m E-W diam., stands proud of the field surface & is surrounded by an arc of waterlogged ditch running round the outside of the bank, which varies in height. At WSW, the bank is 1.25m above the interior, 5.4m wide & 2.35m above the ditch, which is 1.4m wide & 1m deep. The interior gradually slopes to N. See SM7 for further details. An archaeological evaluation was carried out on a proposed development site NE of the hillfort. Two test trenches were excavated across the development area. No artefacts or features of archaeological significance were uncovered [SOC 20/09/04].

SMR Number DOW 034:032 view on map

 

Edited Type: HILLFORT

Specific Type Specific Period

Townland: DROMOREBRAGUE

 

Council: ARMAGH CITY BANBRIDGE AND CRAIGAVON

County: DOW

Grid Ref: J1344041630

Protection: Scheduled

Parish: AGHADERG

Barony: IVEAGH UPR;UPR HALF

Town:

General Type: HILLFORT

Condition: SUBSTANTIAL REMAINS (Vast majority definable)

General Periods:

Submit

IRON AGE

PREHISTORIC

https://apps.communities-ni.gov.uk/nismr-public/docs/DOW/DOW_034/DOW_034_032/Public/SM7-DOW-034-032.pdf

https://apps.communities-ni.gov.uk/nismr-public/docs/DOW/DOW_034/DOW_034_032/Public/SM7-DOW-034-032A.pdf

https://apps.communities-ni.gov.uk/nismr-public/docs/DOW/DOW_034/DOW_034_032/Public/SM7-DOW-034-032B.pdf

MIGHT ARTHUR REALLY HAVE BEEN BURIED AT BURGH BY SANDS/'AVALON'?

St. Michael's Church, Burgh By Sands

***

ABALLAVA (entry from THE PLACE-NAMES OF ROMAN BRITAIN by Rivet and Smith):

Rudge Cup and Amiens patera: ABALLAVA

Inscription, RIB 883, IN C(U)NEUM FRISIONUM ABALLAVENSIUM

Ravenna, 10729 : AVALANA; variant AVALAVA.

Notitia Dignitatum, XL47 :  Prafectus numeri Maurorum Aurelianorum, ABALLABA. 

DERIVATION. The name is from Celtic *aballa-; there is Gaulish avallo = 'poma' in the Vienna Glossary. The modem languages show Welsh afall 'apple-tree', Welsh afal 'apple' and Breton aval, Irish ubhall, abhall. Cognate Germanic words include English apple and others which go back to a common North European base *abl-. Latin nux avellana 'hazel-nut' may involve a borrowing from Celtic, though there was possibly a native Italic representative of the group (see below). Romano-British Aballava can hardly refer to a fruit or to a single distinguished tree, although it might (as with other tree-references in place-names) allude to a sacred tree. A more mundane explanation is that the name is a collective 'orchard ' ; Irish abhall is used in this sense in some parts. The name is paralleled in Gaul : Aballo TP = Aballone AI  l604 > Avallon (Yonne, France), and there are other places called Avalon in France. In Italic lands there was Abella  (Campania), described as malifera 'rich in apples' in the Aeneid VII.740.

The name has the derivational suffix -ava (British *-auà), as in Galava, Manavia, and abroad Genava, etc. ; see Holder I.305. It is now represented by Welsh -au, Breton -aou -ou.

IDENTIFICATION.  The Roman fort at Burgh-by-Sands, Cumberland (NY 3259).

Note. Presumably the Avalon of Arthurian legend has a like origin. The name appears first in Geoffrey of Monmouth, and the equation with Glastonbury was made with the discovery of the pretended tomb of Arthur at the abbey in 1191 : 

Hic iacet sepultus inditus rex Arturius in insula Avalonia. 

This was adopted in Welsh as Ynys Afallon 'Isle of Apples', perhaps a Celtic version of the Greek Hesperides".

***

CAMBOGLANNA (entry from THE PLACE-NAMES OF ROMAN BRITAIN by Rivet and Smith)

Rudge Cup : CAMBOGLANS

Amiens patera : CAMBOG[LANI]S

Ravenna 10711 (= R&C 131) : GABAGLANDA 

Ravenna 10736 (=R&C 167)  CAMBROIANNA

ND : In Seeck's edition the entry at XI,44 reads : Tribunus cohortis primae Aeliae Dacorum, AMBOGLANNA

M. W. C. Hassall in Aspects of the ND (Oxford, 1976), 113, supposes a lacuna in an early MS of ND, and restores it as follows : 

XL43a : Tribunus cohortis primae Aeliae Dacorum, [BANNA

XL44 : Tribunus cohortis secundae Tungrorum], CAMBOGLANNA

See further BANNA. The presence of cohors II Tungrorum at Castlesteads is attested on RIB 1981-83 and 1999.

The sources need some elucidation. On the restored form of the Amiens patera, see p. 230. The Rudge Cup and the patera tell us that a locative plural in -is- is being recorded (compare Mais, locative plural of Maia, on the same vessels); in the other texts Camboglanna is therefore a nominative or accusative neuter plural, not a feminine singular. Ravenna's first form has initial G for C, a common scribal error, and has at some stage lost -m-, which was often abbreviated, like -n-, in medieval MSS. Ravenna's second form has not previously been equated with Camboglanna. R&C indeed specifically deny the possibility; they emend the Cambroianna of the text to *Cambolanna, by no means unreasonably, and fmd an etymology for its second element in British *landa, *lanna (see VINDOLANDA), placing this *Cambolanna in S.W. Scotland. However, it should be noted that R&C's emendation leaves the original one letter short, whereas equation with Camboglanna does not; also, duplications in Ravenna are much more common than R&C allowed (see Chapter V). The wide separation of the two names in Ravenna's list is no bar to uniting them; compare the duplication of the name of another Wall-fort, Maio-Maia, again widely separated in the text. The Cosmographer was working here from two maps, whose differing scripts led him to duplicate in differing ways. See also Crawford's earlier study in Antiquity, IX (!935) > where he had other arguments for maintaining the separateness of the two places and for locating *Cambolanna well to the north of the Wall. Finally, ND's omission of the initial C- may be merely a scribal accident or an assimilation to Latin ambo- 'both'.

DERIVATION. For *cambo-, see the previous entry. British *glanno- is now Welsh glann 'bank, shore', found also in British Glanum (if for *Glannum) and Glannoventa. The preésent name is thus 'curved bank' or 'bank at the bend'.

IDENTIFICATION. The Roman fort at Castlesteads, Cumberland (NY 5163), beside the Cam Beck.

Note. Survival of the name for a time may be indicated if the Camelon of Harleian MS 3859, where Arthur fought two of his battles (one placed at A.D. 537), is really for Camlann or Cambglan. See discussion in Antiquity, IX (1935), 289-90, and Modern Philology, XLIII (1945), 56; also LHEB 437. Hassall notes, in support of thé identification with Castlesteads, that the Cambeck, the river at the site, may preserve the first element of the old name.

***

Proximity of Camlan and Avalon on Hadrian's Wall

Over the years I written a great deal on Arthur's Avalon.  Is it a pleasant fiction, a nebulous Otherworld paradise?  Or might it have been a real place?

Well, there definitely was an Avalon, and it lies not too many miles to the west of the Camlan where Arthur died.  We are talking, of course, about the Roman fort of Aballava (or Avalana) at Burgh By Sands.  There was a great marsh here (Burgh Marsh), and a Roman period goddess Dea Latis (almost certainly 'Goddess of the Lake').  

Apples in Celtic tradition are emblems of the Otherworld, and a place named Aballava or "Apple Orchard' may have designated an ancient British sacred center.  We need not posit that Arthur was conveyed to an island of goddesses, for it was customary to build early Christian centers atop pagan shrines.  In my mind, therefore, it is not at all difficult to imagine Arthur being conveyed from Camboglanna/Castlesteads to Burgh By Sands.

The only problem for me in this scenario is what his burial at Aballava would mean for the geopolitical state of affairs then prevailing. I've only recently written this piece on what Badon and Camlan might mean for an Arthur who had been born at Ribchester:

https://mistshadows.blogspot.com/2020/01/a-ribchester-arthur-and-battles-of.html

In my book THE ARTHUR OF HISTORY, I vacillated between the Irthing Valley or Stanwix (the Roman period Uxellodunum, which housed the largest cavalry unit in all of Britain) as Arthur's power center.  If he had ruled from Banna/Birdoswald or Camboglanna/Castlesteads, it would be hard to explain an interment at Burgh By Sands.  In addition, if Moderatus/Modred/Medraut belonged to an emerging Rheged centered in Annandale, accepting Aballava as the final resting place of Arthur becomes increasingly difficult. 

However, regardless of where Arthur's chief court lay, we must remember that although he perished in the conflict at Camlan, so, too, did his opponent.  There may have been a sort of pyrrhic victory. And that may mean that Arthur's territory remained intact and relatively safe from any immediate renewed attack.  If Aballava were within the region his successor controlled, he might still have been buried there.

A NOTE ON ABALLAC SON OF BELI AND ANNA

At the head of some of the Welsh genealogies we find Aballac (or Afallach) son of Beli and Anna.  It is well known that Aballac represents a Welsh attempt at Irish Ablach, a word meaning "abounding in apples" or "having apple trees", and found in the Irish Otherworld island place-name Emhain Ablach.

Given that this is so, I have no problem subscribing to the view that Beli here is a substitute for Irish bile, 'a sacred tree.'

But if this is so, who is Anna?  And why are these three personages set at the head of genealogies for the ancestry of Coel Hen of the North and of Cunedda, who was wrongly said to come from Manau Gododdin at the head of the Firth of Forth?  Manannan mac Lir was associated with Emhain Ablach,  sometimes misidentified with the Isle of Man, and the Welsh Manawydan is placed at North Queensferry/Tribruit in the PA GUR poem.  Queensferry was either in or on the eastern border of Manau Gododdin, where the two extant place-names Slamannan and Clackmannan delineate part of that region's extent.  

Well, I believe the clue to Anna's identity lies in the Welsh tradition which makes Afallach the father of Modron, wife of Urien of Rheged.  As I've demonstrated before, the nucleus of Rheged was Annandale, and it was in Annadale that we find Lochmaben.  Not far to the southeast is the Clochmaben Stone.  Both are named for the god Maponus/Mabon, son of Modron.  

As I read it, Anna is merely a Christian substitute for a goddess akin to the Irish Anu.  As Rivet and Smith make clear in their THE PLACE-NAMES OF ROMAN BRITAIN, the River Annan, found in Classical period sources as Anava, has as its root the same word we find in Anu.  There is a St. Ann's hamlet on the Kinnel Water tributary of the Annan, and a St. Ann's Well at Beaumont in the parish of Burgh By Sands ( = Aballava/Avalana/'Avalon').  The River Annan is across the Solway from the Bowness-on-Solway Roman fort, which is just two forts to the west of Burgh By Sands along Hadrian's Wall.

One of the 'waths' or fords over the Solway led from Burgh By Sands to the Clochmaben Stone:

Courtesy https://solwayshorewalker.wordpress.com/2015/08/22/waths-ford-and-border/

Could it be, therefore, that Aballac/Afallach was actually localized at Burgh By Sands?  Was this the earlier Welsh tradition, before it was supplanted by the false identification of Glastonbury with Avalon?

PRECEDENCE IS ALL: EIL FOUND SPELLED ELI or ELIWLAD = EILWLAD

When exploring whether Eliwlad could have come from an original Eilwlad, it was necessary to do more than merely consult top Welsh specialists.  While they could give their weighty opinions on the possibility of such a metathesis as eli from eil having occurred, I still needed to show that this actually happened in early MSS.

I have just succeeded in doing so.

The line pasted at the top comes from A Glossary of Mediaeval Welsh Law Based Upon the Black Book of Chirk by Timothy Lewis, Manchester University Press, 1913.

https://books.google.com/books?id=GbTnAAAAIAAJ&pg=PA146&lpg=PA146&dq=%22Eil%22%2B%22eli%22%2B%22welsh%22&source=bl&ots=763uQuJyNn&sig=ACfU3U2VTj_oF6TzuSGbivzqozp8KBc3qg&hl=en&ppis=_e&sa=X&ved=2ahUKEwiRibvSm5XnAhVNoZ4KHUAlDTQQ6AEwCnoECAQQAQ#v=onepage&q=%22Eil%22%2B%22eli%22%2B%22welsh%22&f=false

In that source, eli is plainly given as an alternate spelling for eil.

The Black Book of Chirk is a very early MS.  From https://www.maryjones.us/jce/chirk.html:

"The Black Book of Chirk Y Llyvyr Du or Weun

ca. 1350

Sometimes called "The Chirk Codex" the Black Book is the earliest collection of the Welsh laws codified by Hwyel Dda, the tenth century king of Gwynedd, though it was composed in the thirteenth century, presumably copying from an earlier, lost manuscript. There is a contemporaneous Latin version of the Laws, Peniarth MS 28, which is best known for being one of the few fully-illustrated manuscripts of Wales.

It currently resides at the National Library of Wales."

One caveat from Dr. Simon Rodway:

"I doubt this is an example of miscopying, but rather of scribal carelessness.  This MS, Peniarth 29 (The Black Book of Chirk), is notorious for the sloppiness and eccentricity of its scribes.  See the detailed discussion by Paul Russell, ‘Scribal (In)competence in Thirteenth-Century North Wales: The Orthography of the Black Book of Chirk (Peniarth MS 29)’, National Library of Wales Journal, 29 (1995‑96), 129‑76."

But, still, we have eli for eil, and this example shows that the right misspelling of the latter could occur.

I have asked the head of MSS. at the National Library of Wales, Dr. Maredudd ap Huw, if he can think of any other examples of eli for eil.  Should he be able to produce such results, I will post those here.

I am now putting my research to rest with the following declaration: I'm confident I have finally, after some two and a half decades of searching, "discovered" the historical Arthur of the 6th century.

EXAMPLE TWO FOUND!

In "The Red Book of Hergest" version of CULHWCH AND OLWEN, we find the personal name Eli rendered as Eil.  This instance occurs in a listing of Arthur's champions.  What should read "Eli and Trachmyr" is instead written "Eil and Trachmyr."

The two images below were taken from -

The Text of the Mabinogion: And Other Welsh Tales from the Red Book of Hergest, John Gwenogvryn Evans, Sir John Rhys, 1887, p. 140 (MS. 842)


https://books.google.com/books?ppis=_e&id=XlEGAQAAIAAJ&dq=%22Red+book+of+hergest%22%2B%22eil%22%2B%22eli%22&q=Eil+%28cf.+Eli%29#v=snippet&q=Eil%20(cf.%20Eli)&f=false

and

https://digital.bodleian.ox.ac.uk/inquire/Discover/Search/#/?p=c+1,t+ac%20eil%2C%20a,rsrs+0,rsps+10,fa+,so+ox%3Asort%5Easc,scids+,pid+9bf187bf-f862-4453-bc4f-851f6d3948af,vi+49792518-1334-4f3e-b894-2dcf72eb6c36

For comparison, here is MS. 818 (110, 7 in Evans and Rhys), where Eli is written properly:

CADGUALLAUN LIU GRANDSON OF SAWYL BENISEL = CAEDUALLA REX BRETTONUM?

Aerial View of Site of the Ribchester Roman Fort

In 2004, Alex Woolfe published an article in NORTHERN HISTORY that may have considerable bearing on my theory that the legendary Arthur came from a royal family based at Ribchester:

https://www.academia.edu/313147/Caedualla_rex_Brettonum_and_the_passing_of_the_old_north

In this piece, Dr. Woolfe very plausibly suggests that a great chieftain of the 7th century named Cadwallon Lyw (the chieftain, lord, leader, ruler, governor; see GPC), usually assigned to the Gwynedd dynasty, should instead be attached to the family of Sawyl Benisel: 

"It remains now to identify an alternative and better placed candidate for Bede's Caedualla rex Brettonum. What is sought is a northcountryman, a neighbour to both Deira and Bernicia, who was appropriately placed to be over-king of both and who had afloruit in the early to mid seventh century. Happily a candidate readily presents himself. Harley pedigree 19 reads:

Catguallaun liu [1] map Guitcun map Samuil penissel map Pappa post Priten map Ceneu map Gyl hen"

I've read his entire paper - more than once now.  And it seems to me that if he is right, this would reinforce my own idea that Arthur was a son of Sawyl.  How?  Well, only because it would be proof that Sawyl's family (and, presumably, kingdom) was extremely powerful in the North during the Dark Ages.  And, that by being so, it would have been fertile ground for the production of a significant hero prior to Sawyl's grandson. 

The case could even be made that Arthur's earlier victories against the Saxons in the North, including that of Badon, had set the stage for the later military successes of Cadwallon.  

[1]

Dr. Woolfe argues, for instance, that the Ynys Glannauc (Enislannach in Gerald of Wales) battle of Cadwallon was a relocated event. I agree.  In my opinion the proper location is the Llan Lleennawc, "Lleenog's enclosure", probably here a fort or fortified settlement and not, as later, a church.  This place is found mentioned in the Book of Taliesin and incorporates the name of the father of Gwallog of Elmet.  

The Kingdom of Elmet lay to the south of Birdoswald on Hadrian's Wall.  Birdoswald/Banna was ruled over by Sawyl, and Ekwall etymologizes the place-name as containing W. buarth, "fold", plus the OE personal name Oswald.  I would suggest that the Roman period Banna was also Cadwallon's fort, and that it was renamed after Oswald killed the former at Heavenfield.  

XI.

GWALLAWC

[BT 29]

"Ef differth aduwyn llan lleennawc"

On the identification of Heavenfield, where Cadwallon met his death, see https://mistshadows.blogspot.com/2020/05/where-exactly-was-hefenfeldheavenfield.html.

CEIDIO SON OF ARTHWYS AND POWCADY, CUMBRIA?

King Water Valley (photo courtesy Andrew Smith)

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.