Novels I Classical Literature I (Pseudo)-Scientific I Children's Books I Comics I Music & Movies I Various stuff
What's New I Sitemap I Arthurian articles I History of Britain, 407-597 I View guestbook I Sign guestbook I Poll I About me I Links I Search


Vortigern Studies > Faces of Arthur > Arthurian Articles > August Hunt (4)

Guest Author:
August HuntVisit August Hunt's website: The Quest for Arthur's Grave

August Hunt, (1960), published his first short stories in his high school newspaper, THE WILDCAT WIRES. These were followed by stories and poems in THE PHOENIX literary magazine of Clark Community College, where he received a writing scholarship. Transferring to THE EVERGREEN STATE COLLEGE in Olympia, WA, he continued to publish pieces in local publications and was awarded the Edith K. Draham literary prize. A few years after graduating in 1985 with a degree in Celtic and Germanic Studies, he published "The Road of the Sun: Travels of the Zodiac Twins in Near Eastern and European Myth". Magazine contributions include a cover article on the ancient Sinaguan culture of the American Southwest for Arizona Highways. His first novel, "Doomstone", and the anthology "From Within the Mist" are being offered by Double Dragon (ebook and paperback). August, a member of the International Arthurian Society, North American Branch, has most recently had his book "Shadows in the Mist: The Life and Death of King Arthur" accepted for publication by Hayloft Publishing. Now being written are "The Cloak of Caswallon", the first in a series of Arthurian novels that will go under the general heading of "The Thirteen Treasures of Britain", and a work of Celtic Reconstructionism called "The Secrets of Avalon: A Dialogue with Merlin". 

Faces of Arthur is a part of Vortigern Studies


Faces of Arthur Index

Vortigen Studies Index


Arthurian Articles
click here

The Spirit of the Wood

August Hunt

WHO WAS MERLIN – or, rather, what was Merlin?

This question has intrigued and vexed countless students of the Arthurian tradition for centuries.  Was he someone who panicked and ran away from the Battle of Arfderydd?  Who lost his sanity in the battle and lived like a wild beast in the woods?  Had he really been a great bard of the chieftain Gwenddolau?  If he were a madman, by what mechanism did his insane utterances become recognized as prophecies?  Why was he also called Llallogan or Llallawg?  Why was he dealt a triple sacrificial death akin to that meted out to the god Lugh (Welsh Lleu)? 

These questions are important in and of themselves, of course. But for our purposes they take on a more profound significance.  For by answering them to the best of our abilities in an objective way, can we say definitively that originally Merlin had belonged to a class of druidic priests?  Or that he had performed some vital function for such a priesthood?

In Geoffrey of Monmouth's History of the Kings of Britain, Merlin (= the Welsh Myrddin) is associated with Amesbury’s Stonehenge on Salisbury Plain and with Mount Killaraus (= Killare next to the Hill of Uisneach, the center of Ireland), while in Geoffrey's Life of Merlin the great sage is placed atop a mountain in the Scottish Caledonian Wood.

Fragments of the Life of St. Kentigern tell of a madman/prophet named Lailoken, who is explicitly identified with Merlin, and who is found on a "rock" at Mellodonor (modern Molindinar Burn) within sight of Glasgow and at Drumelzier (modern Dunmeller) in Scottish Borders. Lailoken is said to have been buried near Drumelzier.

Before Geoffrey introduced Merlin into the Arthurian saga by substituting him for Ambrosius of Dinas Emrys (a hill-fort in Gwynedd, Wales) and Wallop, Hampshire, the madman/prophet had divided his time between Carwinelow (the fort of his lord Gwenddolau, near Longtown in Liddesdale, known now as the Moat of Liddel), nearby Arthuret (scene of the Battle of Arfderydd, in which his lord was slain and he went mad), the Lowland Caledonian Wood with its mountain and the court of King Rhydderch Hen/Hael. Rhydderch belongs at Dumbarton in Strathclyde, although Geoffrey makes him a Cumbrian king.

In Geoffrey the Caledonian mountain remains unnamed. This is unfortunate, in that by finding this mountain we might learn a great deal more about Merlin’s identity.  And, incidentally, we would have a much firmer fix on the location of Arthur’s seventh battle, which occurred in the Caledonian Wood.

Merlin’s Caledonian Wood mountain is mentioned in one other source:  the 13th century French verse romance by Guillaume Le Clerc entitled Fergus of Galloway. The Fergus romance is distinguished by the author’s knowledge of Scottish geography. To quote from Cedric E. Pickford in Arthurian Literature in the Middle Ages:

“His [Guillaume's] Scottish geography is remarkably accurate... In the whole range of Arthurian romance there is no instance of a more detailed, more realistic geographical setting.”

The modern translator of Fergus, the late D.D.R. Owen, has made similar remarks on this romance. The notes and synopses in his translation also remind the reader that various elements of the Fergus mountain episode were adapted from Chretien's Yvain and Perceval and the Continuations of the latter.

But it remains true that only Fergus actually names Merlin's mountain and purports to give us directions on how to get there. The hero Fergus starts his journey to the mountainn not as Nikolai Tolstoy (in his The Quest for Merlin) claims at the Moat of Liddel, where Merlin fought and fled in madness, but at Liddel Castle at Castleton in Liddesdale.  Tolstoy uses 1) Guillaume's directions and the placement of King Rhydderch at Dumbarton 2) Merlin's affinity with the stag in Geoffrey's Life of Merlin 3) the incorrect identification of Merlin's Galabes spring/s (these are at Wallop in Hampshire, not in the Scottish Lowlands) and 4) the great height of the Black Mountain to select Hart Fell at the head of Annandale as Merlin's mountain.

There are marked problems with each of these "guidelines" used by Tolstoy. Firstly, the directions given are incredibly vague and hence can be used to chart a course from the Moat of Liddel to just about anywhere:

“[Fergus] comes riding along the edge of a mighty forest... Fergus comes onto a very wide plain between two hills. On he rode past hillocks and valleys until he saw a mountain appear that reached up to the clouds and supported the entire sky...”

Secondly, Fergus's mountain is given two names, neither of which match that of Hart Fell: Noquetran (variants Nouquetran, Noquetrant) and Black Mountain. The latter is obviously a poetic designation only, the primary name being Noquetran. 

And thirdly, there is no edifice of any kind atop or on the flanks of Hart Fell which could have been referred to as ‘Merlin’s Chapel’.  As described, this edifice must be an ancient chambered cairn.  Such monuments are often associated with Arthurian characters. 

The hill-name Noquetran looks to me like a Norman French attempt at a Gaelic hill-name, with the first component being cnoc, English knock, “hill”.  As the French render English bank as banque and check as cheque, Cnoc/Knock became Noque-.

The secret to correctly interpreting the –tran component lies in a closer examination of Professor Owen’s notes on the Fergus romance.  For lines 773-93 he writes:

This adventure [of the Noquetran] is largely developed from elements in C.II [the Second Continuation of Chretien’s Perceval].  There Perceval fights and defeats a Black Knight in mysterious circumstances.  Earlier, he had found a fine horn hanging by a sash from a castle door.  On it he gave three great blasts, whereupon he was challenged by a knight, the horn’s owner, whose shield was emblazoned with a white lion.  Perceval vanquished this Chevalier du Cor and sent him to surrender to Arthur.  At his castle he learned of a high mountain, the Mont Dolorous, on whose summit was a marvellous pillar… fashioned long ago by Merlin.

For lines 4460 ff, Owen writes:
Mont Dolorous, which also appears in C.II (see note to II. 773-93 above), is here associated with Melrose and is probably to be identified with the nearby Eildon Hills…

In the Fergus romance, the Noquetran episode comes first.  The horn hangs from a white lion (cf. the lion on the knight’s shield in the Perceval Continuation) in the Noquetran chapel, where Merlin had spent many a year.  In front of the chapel is a bronze giant, apparently a statue, whose arms are broken off by Fergus, causing the giant’s great bronze hammer to fall to the ground.  Later in the romance, Fergus goes to the Dolorous Mountain or the Eildons and encounters there a club-wielding giant in the Castle of the Dark Rock (reminiscent of the Black Mountain name applied to the Noquetran).

As it happens, the Eildons are noteworthy for having three major ancient monuments atop two of their three hills.  On the Eildon North Hill is the largest hill fort in Scotland, the probable oppidum of the Selgovae tribe. Here also is a Roman signal station.  On Eildon Mid Hill is a large Bronze Age cairn. The CANMORE database (of the Royal Commission on the Ancient and Historical Monuments of Scotland) has this on the cairn:

This cairn is situated on the SW flank of Eildon Mid Hill about 100ft below the summit, at a height of some 1285ft OD. It has been much robbed and now appears as a low, irregular mo und of stones, about 50ft in diameter, from which a few boulders protrude to indicate the possible former presence of a cist.

Merlin's Chapel, the Bronze Age Cairn Atop Eildon Mid Hill (Photo Courtesy John Young)

More remarkable was the presence below the cairn of the following (also from CANMORE):

A group of seven bronze socketed axes, found on the lower western slopes of Eildon Mid Hill in 1982, is now in the Royal Museum of Scotland (RMS). More can be found on the axes in the Proceedings of the Society of Antiquaries of Scotland, Volume 115 (1985), pp.151-158.  An abstract of this article follows:

A group of bronze socketed axes from Eildon Mid Hill, near Melrose, Roxburghshire
Brendan O'Connor and Trevor Cowief.

In 1982, a group of seven socketed axes was found on the lower western slopes of Eildon Mid Hill, Ettrick and Lauderdale District, Borders Region. Although recovered from redeposited soil, the axes probably represent a hoard of the Ewart Park phase of the late Bronze Age. The find reinforces what appears to be a significant local concentration of contemporary metalwork around the Eildon Hills.

The Eildon Hills and the relative position of the axes.


On 9 August 1982, several bronze socketed axes were discovered by Messrs W and A Wilson (uncle and nephew) on the margin of the rifle range on the lower western slope of Eildon Mid Hill, the central and highest of the three peaks which form one of the most conspicuous landmarks of the Scottish Border country (illus 1 -2). They were intending to use Mr William Wilson's metal detector slightly further uphill in order to search for shell-cases and cartridges in the area of the targets. While they were walking up the roughly trodden grassy path on the northern margin of the range, with their detector switched on but not consciously in use, their conversation was interrupted by a signal from the machine. The tone indicated a non-ferrous metal, and closer scanning with the detector suggested that there were at least four soil anomalies within a small area. The Wilsons decided to investigate the source of these signals: clearance with a trowel of a small patch of the tenacious turf revealed, to their surprise, not spent ammunition but a socketed axe of bronze (catalogue no 7); a second discrete signal was found to emanate from a further axe (7) also lying on its own just under the turf. The source of a third signal was revealed to be a cluster of three axes lying close together (3-5). Realizing the growing archaeological significance of their find, and not wishing to disturb the site further, the Wilsons responsibly left undisturbed in the ground the sources of what by then appeared to be two further signals of a similar nature. They replaced the disturbed turf as best they could to mask the site, and immediately reported their discovery to the Ancient Monuments Division of the Scottish Development Department, one of whose Inspectors, Dr N Fojut, in turn notified the National Museum. On 17 August, one of the writers (TGC) visited the finders to inspect the axes already recovered and to view the site of their discovery, and the following day, he and Mr lan Scott of the NMAS, with the assistance of Aidan Wilson, investigated the immediate area of the find (NGR NT 542325).

A small trench, 2m by 2m, was set out centred on the find-spot of the cluster of three axes already removed (cf illus 1, d). Following removal of the coarse turf, the points from which the five axes had been retrieved became clear: these showed up as irregular depressions in what appeared to be the natural subsoil of reddish clayey loam with plentiful stones (mostly the local felsite). Scanning of the trench with the detector relocated the positions of the two signals not investigated by the Wilsons the previous day. One of these emanated from a slightly darker patch of humic soil: on removal, this proved simply to be a deeper pocket of topsoil, occupying a slightly damper, clayier depression in what seemed once again to be the natural subsoil. A further socketed axe (6) lay at an angle on the side of the depression 10-15 cm below the present ground surf ace: perhaps on account of the damper matrix, the axe was in a noticeably more corroded condition than the others when found.

Surprisingly, the source of the final signal appeared to emanate from natural subsoil with no obvious trace of any feature on its surface. Removal of a small area of this supposedly natural 'subsoil' revealed the source of the remaining signal to be a further axe (2) and threw some light on the context of the group of axes as a whole. It became clear that this last axe was not, as first thought, lying in undisturbed ground, but rather was lying in compacted redeposited soil apparently occupying the side of a natural gully or channel in the hillside. In the time available, it was not possible to excavate the presumed channel nor determine its width or depth, but the circumstances which led to the incorporation of redeposited soil in natural features seem clear enough for the construction of the rifle range must have involved considerable smoothing out of the contours and irregularities of the hill-slopes. Churning up of the ground, the infill of erosional features such as gullies caused by water run-off, and the compaction of the area by machines could account for the formation of the deposit on and in which all seven axes lay. Indeed, gullies of the type envisaged can be seen elsewhere on the slopes of the Eildons (see illus 2). The axes are likely to have been moved bodily in a load of earth and soil, dumped and then slightly dispersed (axes 2 and 6 were separated by a distance of 1-5 m). Following the removal of the final two axes, the excavated soil and turf were replaced and the site restored to its original appearance as far as possible. Finally the surrounding area was scanned with the metal detector, but no further non-ferrous anomalies were noted…


In view of their discovery in redeposited soil we cannot be absolutely certain how the axes were originally deposited. However, their number, their proximity and their similar condition all suggest that they came from a hoard, probably close to their eventual find-spot. Whether the seven axes recovered in August 1982 comprised the whole hoard remains uncertain. On the other hand, it is possible, though less likely, that more than one separate deposit was originally involved…

Eildon Mid Hill - the group of socketed axes (scale 1:3).

I would see in ‘Noquetran’ or Noquetrant a Gaelic cnoc or Anglicized ‘knock’ plus one of the following:

G. dreann – grief, pain (cf. Irish drean, sorrow, pain, melancholy)

G. treana, treannadh – lamentation, wailing

In other words, Noquetran is merely a Gaelic rendering of the Old French Mont Dolorous!

The bronze hammer Fergus causes to be dropped near Merlin’s Chapel on the Noquetran is a folk memory of a bronze socketed axe being deposited on the slope below the Eildon Mid Hill cairn or, more probably, of such an axe being found on the site prior to Guillaume Le Clerc’s writing of the Fergus romance.  Merlin’s Noquetran chapel is the Eildon Mid Hill Bronze Age cairn.

Melrose Mountain, Black Mountain and Castle of the Dark Rock are all designations for the Eildons.  The hill-name Eildon is found in 1130 (_Place-Names of Scotland, 3rd Edition, James Brown Johnston) as Eldunum and in 1150 as Eldune.  While various etymologies have been proposed, the most commonly favored one is G. aill, “a rock, cliff”, plus OE dun, “a hill”.  The “Fergus” romance’s “Castle of the Dark Rock” (Li Chastiaus de la Roce Bise) may stand for the hill-fort on Eildon North Hill, with Eildon being perceived as composed of aill, rock, plus not dun, “hill”, but instead - “OE dun, a colour partaking of brown and black; ME dunne, donne, dark-coloured: Ir. Dunn, a dun colour: Wel. Dwn, dun, swarthy, dusky: Gael. Donn, brown-coloured”
[From Toller’s and Bosworth’s Anglo-Saxon Dictionary]

So why were the Eldons identified with the Dolorous Mountain/Noquetran?  The answer may lie in part with Nikolai Tolstoy’s astute observation that the lion Fergus thinks should be roaming over the mountain-top, but which he finds inside the ‘chapel’ is an error or substitution for the god Lugos (Welsh Lleu, Irish Lugh).  In Welsh, Lleu’s name could sometimes be spelled Llew, and the latter is the normal spelling for the Welsh word “lion”.  Merlin’s associations with Lleu will be briefly discussed below.  For now, suffice it to say that the Dolorous Mountain undoubtedly got its name because Lugos or Lugh was at some point wrongly linked to Latin lugeo, “to mourn, to lament, bewail”.  Such mistakes in language could easily have occurred when going from Celtic to Old French. 

The Dolorous Mountain is thus, properly, “Lugos Mountain”.  And the Lugos/Lugh/Lleu mountain is particular is Eildon Mid Hill, the highest of the Eildons, with its Bronze Age cairn.   Such an identification of the Dolorous Mountain has implications for the Dolorous Garde of Lancelot, especially given that Lancelot himself is a late literary manifestation of the god Lugh, something first discussed long ago by the noted Arthurian scholar Roger Sherman Loomis. 

We know of five Lugh forts in Britain, four known and one unlocated.  Of the former there is Dinas Dinlle in Gwynedd, Loudoun in East Ayreshire, Luguvalium or Carlisle in Cumbri and Lleuddiniawn or “Lothian”, land of the Fort of Lugh.  Luguvalium has been interpreted as containing a personal name *Lugovalos, ‘Lugos-strong’, but I believe this name is instead a descriptive of the fort itself as being ‘Strong as Lugh’.  Dr. Graham Isaac, a Celtic language specialist at The National University of Ireland, Galway, agrees with me that this could well be the case. 

Then there is the Lugudunum or “Hill-fort of Lugh” of the Ravenna Cosmography.  This place is (see Rivet and Smith’s “The Place-Names of Roman Britain”) situated somewhere roughly between Chester-le-Street and South Shields.  The only good candidate would seem to be Penshaw Hill, which the Brigantes Nation Website calls “the only triple rampart Iron Age hill-fort known to exist in the north of England.”  Penshaw Hill is associated with the famous Lambton Worm, a monster not unlike the two worms or dragons of Lleu’s hill-fort of Dinas Emrys in Gwynedd, Wales. 

The Eildons are noted for the stories of “Canobie” or Canonbie Dick and Thomas the Rhymer of Ercildoune.  Canonbie is near to both the Carwinley of Myrddin’s/Merlin’s lord Gwenddolau and Arthuret Knowes, the scene of the Battle of Arfderydd in which Myrddin was driven mad.  The 13th century Thomas is credited with meeting an elf-woman under the Eildon Tree (whose location is now marked by a stone) and being taken under the Eildons to the land of Faery.  He is also credited with a prophecy concerning Merlin’s grave at Drumelzier:

When Tweed and Powsail meet at Merlin’s grave,
Scotland and England that day ae king shall have.

[According to P.C. Bartrum in “A Welsh Classical Dictionary”, under his entry for Llallogan, this prophecy was first published by Alexander Pennycuick in 1715.]

The story of Canonbie Dick presents Thomas as a wizard from past days, and I will quote it in full:

A long time ago in the Borders Region there lived a Horse Cowper (trader) called Canobie Dick. He was both admired and feared for his bold courage and rash temper. One evening he was riding over Bowden Moor on the West side of the Eildon Hills. It was very late and the moon was already high in the night sky.

He had been to market but trade that day had been poor and he had with him a brace (pair) of horses, which he had not been able to sell. Suddenly, he saw ahead of him on the moonlit road, a stranger. The stranger was dressed in a fashion that had not been seen for many centuries. The stranger politely asked the price of the horses.

Now Canobie Dick liked to bargain, and was not worried by the strange man's looks. Why, he would have sold his horses to the devil himself, and cheated him as well, given half a chance. They agreed a price which the stranger promptly paid.

The only puzzle was that the gold coins he used to pay were as ancient as his dress. They were in the shape of unicorns and bonnet pieces (Scottish coins from 1400s and 1500s). However, Canobie Dick shrugged his shoulders. Gold was gold. He smiled to himself, thinking that he would get a better bargain for the coins than the stranger had got for the horses.

When the stranger asked if he could meet him again at the same place, Canobie Dick was happy to agree. But the stranger had one condition: that he should always come by night and always alone.

After several more meetings, Canobie Dick became curious to learn more about his secret buyer. He suggested that 'dry bargains' were unlucky bargains and that they should seal the business with a drink at the buyers home.

"You may see my dwelling if you wish," said the stranger; "but if you lose courage at what you see there, you will regret it all your life."

Canobie Dick was scornful of the warning, after all he was well known for his courage and the stranger seemed harmless enough. The stranger led the way along a narrow footpath, which led into the hills between the Southern and central peaks to a place called the Lucken Hare. Canobie Dick followed but was amazed to see an enormous entrance into the hillside. He knew the area well but had never seen before such an opening or heard any mention of it.

They dismounted and tethered their horses. His guide stopped and fixed his gaze on Canobie Dick. "You may still return," he said. Not wanting to be seen as a coward, Canobie Dick shook his head, squared his shoulders and followed the man along the passage into a great hall cut out of the rock.

As they walked, they passed many rows of stables. In every stall there was a coal black horse, and by every horse lay a knight in jet black armour, with a drawn sword in each hand. They were as still as stone, as if they had been carved from marble.

In the great hall were many burning torches. But their fiery light only made the hall more gloomy. There was a strange stillness in the air, like a hot day before a storm. At last they arrived at the far end of the Hall. On an antique oak table lay a sword, still sheathed, and a horn. The stranger revealed that he was Thomas of Ercildoun (Thomas the Rhymer) the famous prophet who had disappeared many centuries ago.

Turning to Canobie Dick he said, "It is foretold that:

He that sounds the horn and draws that sword, shall, if his heart fails him not, be king over all broad Britain. But all depends on courage, and whether the sword or horn is taken first. So speaks the tongue that cannot lie."

The stillness of the air felt heavy. Canobie Dick wanted to take the sword but he was struck by a supernatural terror, such as he had never felt before. What, he thought, would happen if he drew the sword; would such a daring act annoy the powers of the mountain?

Instead he took the horn and with trembling hands put it to his lips. He let out a feeble blast that echoed around the hall. It produced a terrible answer. Thunder rolled and with a cry and a clash of armour the knights arose from their slumber and the horses snorted and tossed their manes.

A dreadful army rose before him. Terrified, Canobie Dick snatched the sword and tried to free it from its scabbard (sheath). At this a voice boomed:

"Woe to the Coward, that ever he was born,
Who did not draw the sword before he blew the horn"

Then he heard the fury of a great whirlwind as he was lifted from his feet and blasted from the cavern. He tumbled down steep banks of stones until he hit the ground. Canobie Dick was found the next morning by local shepherds. He had just enough trembling breath to tell his fearful tale, before he died.

A similar story is told of Alderley Edge in Cheshire, only in that version the wizard is Merlin and the sleeping knights are King Arthur and his men.   My guess is that in the case of the Canonbie Dick story, Thomas the Rhymer has taken the place of Merlin.  This is not a new supposition, but combined with my identification of Myrddin’s Noquetran with Eildon Mid Hill as the Dolorous Mountain, the argument is significantly strengthened.  For “Fergus” was written around 1200 AD, while Thomas is thought to have lived c. 1220-1298.  At some point Thomas was substituted for Merlin at his chapel/cairn on Eildon Mid Hill.

If I am right and the Eildons are Merlin’s Mountain at the center of the great Celyddon Wood, then we can allow for the Celyddon as being thought of as the ancient woodland which covered much of the area surrounding the Eildons.  When we combine this with the fact that Merlin was obviously wandering in the wood in the vicinity of Drumelzier when he was captured by Meldred, it is fairly obvious that the Celyddon, which in this context means merely a great forest of the Scottish Lowlands, extended at least to the Lammermuir Hills in the north and perhaps as far as the Cheviots in the east.

As the Roman Dere Street crosses the Teviot near Jedburgh and the Tweed at Newstead, and four of Arthur’s other battles were fought on or near Dere Street, it is likely that Arthur’s Celyddon Wood battle was fought in part of this ancient woodland on or near Dere Street somewhere between Teviotdale and Lauderdale. 

Indeed, there were four great ancient forests surrounding the Eildon Hills: the Jedforest, whose Capon Tree oak is one of the oldest such trees in all of Britain; Teviotdale itself, which was covered by huge oaks and ash trees in the 12th century; the Ettrick Forest of Selkirkshire; the Lauder Forest, an immense forested track encompassing Lauderdale that still existed up until the 17th century.  Apples, or rather crab-apples, the very species of tree Merlin takes refuge under in the early Welsh poetry, were also present in this region. The St. Boswell’s Apple is thought to be 150 years old and is the largest of its kind in Scotland.  Thomas the Rhymer, taken to Fairyland at the Eildons, is given an apple by the Queen of Fairy.

We may be able to pinpoint the location of the Coed Celyddon battle for precisely.  Although Geoffrey of Monmouth has been justly criticized for producing stories of early British kings rather than histories, we cannot discount the possibility that at least occasionally his account of Arthur’s reign may preserve accurate historical traditions.  When he has Arthur face the Saxons in the Caledonian Wood, he tells us that
Arthur… ordered the trees round that part of the [Caledonian] wood to be cut down and their trunks to be placed in a circle, so that every way out was barred to the enemy.

This circular palisade Arthur constructs in the Caledonian Wood made me think of the semi-circular Catrail dyke.  To quote the description of the Catrail from the Royal Commission on the Ancient and Historical Monuments of Scotland (courtesy Mark Douglas, Principal Officer for Heritage and Design, Planning and Economic Development, Scottish Borders Council):

This is the linear earthwork, known as the Catrail which runs from Robert’s Linn to Hosecoteshiel and embraces the entire head of the Teviot basin from the Slitrig to beyond the Borthwick Water.  It is not continuous, in the sense that is incorporates streams or woodland, and its engineering, in short sectors clumsily joined and imperfectly aligned, stamps it as a product of somewhat inexperienced communal effort.  The date of its construction is doubtful, though it is certain that it fits only Anglian territorial dispositions.  Oman notes the opinion that it might mark a political boundary after Aethelfrith’s victory at Degsastan in 603, and this is indubitably the earliest possible date for such a work.

Map with the course of the Catrail river.

As the date of the construction of the Catrail is “doubtful”, could we not propose that it was built slightly earlier, say in the 6th century, the time of Arthur?  The Catrail runs through what was the great Ettrick Forest, an extensive woodland which according to Nicola Hunt, Projects Officer of the Borders Forest Trust (private communication), ‘covered from Dumfriesshire to the Galashiels area’.  Thus this forest was very near to Merlin’s Eildon Mid Hill. 

The derivation of the name Catrail has been disputed.  The most commonly disseminated etymology is Cymric cad or cat, ‘battle’, plus Cymric rheil or rhail, wrongly defined as ‘fence’, probably for a presumed wooden palisade that may have once existed atop the dyke’s bank.  Unfortunately, as rheil or rhail does not mean ‘fence’.  It is ‘rail, track’ and is not attested in Welsh until the 18th century.  According to Professor Richard Coates (Onomastics and Director of the Bristol Centre for Linguistics at UWE, and Hon. Director of the Survey of English Place-Names) rheil/rhailis doubtless a borrowing from English and so Catrail cannot be derived from Cat + rheil/rhail.

A cad/cat + rhigol (the ‘rhill’ mentioned below), the latter being Welsh for ‘rut, groove, (long narrow) channel; trench, furrow, ditch, gutter’, cannot have become Catrail – even through a process of Anglicization.  Rhigolis pronounced something like “wriggle”.  Again, Professor Coates has assured me that a hypothetical Cad + rhigolcannot have become Catrail.  This is so despite the attractiveness of “Battle-ditch” as a meaning for the earthwork’s name.   Dr. Andrew Breeze, the Celtic onomastic expert in Pamplona, has assured me that rhigol “is unknown before the sixteenth century and is a loan from English”.

Helen Darling, Part-Time Local Studies Librarian, Library Headquarters, St. Mary’s Mill, Selkirk, was kind enough to send me the following on this feature-name:

“KENNEDY, William Norman: Remarks on the ancient barrier called "The Catrail", with plans  In  Proceedings of the Society of Antiquaries of Scotland 1857-1860  pp117 - 121

Considerable diversity of opinion exists as to the derivation of the term, it being variously stated by different authors.  Chalmers calls it "the dividing fence", or "the partition of defence"; Jeffrey, "a war fence or partition - Cat signifying conflict or battle, and Rhail a fence"; others from Cater a camp, and Rhail a fence, a dividing fence among the camps; others, again, from Cud a ditch, and Rhail a fence, the ditch fence or boundary; while another class call it "the Pictswork ditch", attributing the formation of it, and all other ancient artificial remains in the district through which it passes, to the Picts, - a race regarding  whom very mythical traditions continue to float about and receive credence; but almost all writers concur in attributing its formation to the Britons, subsequent to the withdrawal of the Romans from this country.

 CRAIG-BROWN, T.  The History of Selkirkshire or Chronicles of Ettrick Forest.  Vol. I The Shire - The Parishes of Ettrick, Kirkhope, Yarrow, Roberton, Ashkirk, Innerleithen, Peebles, Stow, and GalashielsEdinburgh: David Douglas, 1886, page 47

Concerning the origin of the word "Catrail" there have been as many different suggestions as about its purpose:

CHALMERS. - Cad, a striving to keep; and Rhail, a division.  Together - the dividing fence.

JEFFREY. - Cat, a struggle; and Rhail.  Together - a war-fence.

MACKENZIE. - (Quoted by Gordon).  An old Highlandword signifying wall or ditch of separation.

VARIOUS. - Cater, a camp; and Rhail Cud, a ditch; and Rhail.  Cad-rhill - war-trench.

MISS RUSSELL. - Cader (Welsh), a defence; and Rail.

An amateur philologist has ventured to suggest the Saxon words camp-trail as at all events a possible origin, but the probabilities are in favour of a British or Cymricderivation; and we incline to cater-rhail, a camp fence.  Cater, Celtic for camp, as in the celebrated Cater-thun, is undoubtedly derived from the Latin, castra.  Until the Romans came, there was no need for such a word in the native vocabulary, and it must be kept in mind that the diggers of the Catrail were Romanised Britons.

That there was along the Catrail a terrible struggle for supremacy between the Cymriand the Saxons may be inferred from the frequent occurrence of names beginning with the word "Cat", or "Cad", Cymric for battle or conflict.  Nor far from its northern end is Cat-pair, and nearer, Cat-ha(?).  Caddonwe have seen derived from Cad afon, battle river.  Cat Craig is a hill on Tinnis farm in Yarrow, traversed by the Catrail.  A little further up is Catslack, and there are pools near by known as Cat Holes.  Finally, near the end of the trench is Catlee.  It is necessary, however, to be on one's guard against etymological inferences, and these names are submitted with due reserve.”

The noted Scottish place-name expert Watson said that the Catrail name ‘may be compared with Powtrail [now Potrail], the name of a head-stream of Clyde’ whose ‘meaning is obscure’.  As Henry Gough-Cooper of the Scottish Place-Name Society confirmed, the Pow- of Powtrail is almost certainly Cymric pwll, ‘pool’, but also ‘stream’. 

After noticing on the map that the Powtrail has many pronounced bends or turns, I would follow Professor Richard Coates’ (Onomasticsand Director of the Bristol Centre for Linguistics at UWE, and Hon. Director of the Survey of English Place-Names) suggestion of Welsh traill, ‘turn, a turning’, as the stream’s second component, supplying us with a meaning of ‘stream that turns’ or the ‘turning stream’. 

There is precedence for connecting traill with a stream or river-name: Chepstow’s earlier Welsh name was Ystraigl or Ys-traigl, ‘a turn’, this being a reference to the bend of the River Wye at the site of the town. 


However, such a derivation for Powtrail does not help us with the Catrail name.


Henry Gough-Cooper of The Scottish Place-Name Society has informed me that Alan James prefers a different etymology for the Powtrail.  In his paper ‘A Cumbric Diaspora?’ (in ‘A Commodity of Good Names’, ed. Padel and Parsons, Shaun Tyas 2008), James mentions the Catrail in a footnote on Powtrail, which he speculates may be Cumbric “*polter-eil  ‘stream of a wattle fence of woven hedge’: for *eil see ‘The Poems of Taliesin’, ed. I. Williams (Dublin 1968), pp. 85-6.”


According to Gough-Cooper, *Polter is stream-naming word found in northern Cumberland, parts of the Solway basin and south-west Cheviots, e.g. see EPNS Cumberland I, 8, 24, 62; II, 373; plus Polterkened in the Lanercost Cartulary. Alan James suggests perhaps *pol + extension *duvr, the latter being the word for ‘water’.”  One might compare for this last element the river-names Kielder and Calder, both from Welsh caled, ‘hard’, and dwfr (OBrit dubro-), ‘water, stream’. 


Williams, on the pages cited by James, says that:

Eil was used of any sort of construction involving plaiting, wattling, and presumably would denote a defensive construction... Cf. perhaps Irish aile, ‘a fence’…


The problem with the –ail of Catrail being the word for ‘fence’ is what to make of the first component, Catr-. 


I would propose that the first component was derived from Gaelic cathair, 'fort'.  Even though, as Dr. Richard Coates has made clear to me, the medial th in cathair is silent (cf. Welsh caer, 'fort') and should not have yielded Catr-, in Highland Scotland we do find the Brown and White Caterthun (cathair + dun) forts.  If cathair could become cater in these names, certainly it could have become the Catr- of Catrail.


But if the Catrail is the ‘Fort-fence’, what fort are we talking about?

John Barber, Elaine Lawes-Martay and Jeremy Milln (in “The Linear Earthworks of Southern Scotland; Survey and Classification”, Transactions of the Dumfriesshire and Galloway Natural History and Antiquarian Society, Series III, Volume LXXIII, 1999), summarize their study of the Catrail thusly, bringing the dyke into connection with the fort next to Merlin’s Eildon Mid Hill, the center of the Celyddon/Caledonian Wood:

The natural segmentation of the Scottish landscape into separate territories may also account for the relative absence of sites used as local-administration boundaries. These administrative units are usually fairly clearly defined by the nature of the terrain and separated from each other by distinct, natural features, such as ranges of hills of mountains. This factor may also account for the paucity of lengthy earthworks. Of those examined in this study only the Catrail seems likely to have functioned as a large scale territorial boundary. In association with other natural features, it may have functioned in respect of the hillfort of North Eildon Hill in the same role as the earthwork called ‘The Dorsey’ seems to have functioned in respect of the royal site of Eamhain Macha (Lynn 1981). The discovery that the line of the Dorsey is continued across an apparent gap, once bogland, by a wooden palisade perhaps points to new avenues of approach to similar areas on the Catrail.”

In the same Transaction of the DGNHAS the Catrail is described in detail:

The Catrail by Jeremy Milln

Prior to this report, as many as fifty writers have mentioned the Catrail, the most recent being the RCAHMS which, in the autumn of 1945, carried out the only previous systematic survey (RCAHMS, 1956). All those portions given credence by RCAHMS (and many other earthworks which have on occasion been hypothesized as forming part of the Catrail), were re-surveyed by the CEU in 1984. The major part of the earthwork examined in 1945 had survived, but increasing land-use pressure includingforestry, drainage and cultivation, has been responsible for a rapid deterioration in its overall condition. The erroneous but popular concept of an earthwork extending for some 86 kmin a broad arc from Peel Fell on the border with Northumberland to Torwoodlee, north of Galashiels, was first expressed in the Itinerarium Septentrionale (Gordon 1726, 102-4). Chalmers (1886, 240-3) regarded it as having formed a frontier between Romanized Britons in the west and the Saxons in the east, running from the Forth to the Tyne, and including the Northumbrian Black Dyke. Murray (1864, 37-8) seems to have been the first to point out that the Catrail incorporates two separate and distinct earthworks. Craw’s plan of 1924 (Craw 1924, 42) shows one segment running from Roberts Linn to Hoscoteor possibly Clearburn, and a second, the Pictswork, extending from LinglieHill to Torwoodlee near Galashiels.Some succeeding writers were prepared to accept the curtailed Catrail, but Smail (1879) and particularly Lynn (1898) adhered in detail to the older theory. Lynn’s account is the more meticulous, but he was unable to distinguish between the true linear earthwork and the intermittent remains of old 80 GAZETTEER OF GROUP 1 SITES roads and field-boundaries; most notable is his misinterpretation of branches of the Minchmoor Roadin Selkirkshire, which have since been clarified (Inglis 1924, 205-6). The RCAHMS account of the course of the Catrail (1956, 479-83), as well as for the Pictswork (1957, Nos 126-7), is questionable only in matters of minor detail. The course of the Catrail is shown in Fig. 3. The following account examines it sector by sector beginning at its alleged origin at Peel Fell, on the border (Gordon 1726, 102). Earlier reports suggest thatthe Catrail ran for about 9 kmfrom Peel Fell to Roberts Linn via the Wormscleuch, Liddel, Dawston and Cliffhopevalleys. The only definite linear earthwork now visible on this line is an indeterminate fragment at CaddrounburnCulvert (152). As the RCAHMS survey points out (1956, 480), the topography of the steep-sided Roberts Linn forms a convenient natural break to its course here.
With the exception of a 350 m gap to the west of the Langside Burn (281) the earthwork from Roberts Linn is continuous as far as the head of Barny Sike(282) and it is only beyond this point that uncertainty occurs. Smail (1879, 113), following Gordon and Chalmers, asserts that it crossed WhitehillLower to the north of the enclosure; there is indeed a linear earthwork at this height (201), but no evidence of a connection past Pyot’sNest (NT 4817 0530). Furthermore, this earthwork is markedly off line with the bank on the south side of the ditch. This suggests that it is one of a number of linear outworks delimiting the ground around Pyots’ Nest fort on its landward side where it is overlooked by Whitehill. A short length of linear outwork passing close to the south-west side of the enclosure at the Dod (NT 4727 0602) is aligned on 201 and could perhaps be considered part of the Catrail (Ian Smith, pers comm), but it is isolated and nowhere recognized in the literature. Elsewhere, the line of the Catrail incorporates streams, suggesting that the course from the head of Barny Sikemight be along Barny Sikeitself and the DodBurn. The next section would thus seem to be 283 and, though this earthwork is discontinuous, for the most part its character is similar to that recognized elsewhere. RCAHMS (1956, No 481) suggests its sudden extinction above Priesthaughmight be explained by an area of valley side scrub carrying the work to the top of 284, 900 mto the north. It is debatable whether the short stretch of bank from Peel Brae to the Dod-Priesthaughroad continued west across the road to the earthwork on the Allan Water; though Lynndoes record a cropmarkin the cultivated field (Lynn 1898, 79). The connection with the Allan Water, which almost certainly maintained the line to Doecleuch, might alternatively be made by a 300 mlength of earthwork, consisting of a ploughed-down bank, running along the valley floor north of the farm at Priesthaugh (288). The next surviving fragment, 285, lies on the north-west slopes of DoecleuchHill, and continues the RCAHMS Inventory entry (1956, No 481), which cites a point (NT 460 061) on the south-east of that hill. The area is disturbed by rig and furrow and it was not discernable in the present survey. It is possible however,that the work ran from the steep lip of Doecleuchat NT 4609 0634 to join the truncated end of 285 at the DodBurn-TeindsideRoadand part of the line of this is preserved by an old footpath. 285 may have continued as far as Chapel Cleuch, just south-west of Old Northhouse (it is marked on a plan of 1857 inthe library of the Society of Antiquaries of Scotland), but the area is now marshy and no earthwork can be distinguished. The course of the Catrail, as far as the Muselee Burn, continued by the Northhouse and TeindsideBurns and 286, has never been disputed. Lynn (1898, 73) claimed that a short length was to be seen on the west side of the MuseleeBurn opposite the north-west end of 286, but he seems to have been misled by a track. By implication, it ran north/north-west across the rough pasture to pass close by the west side of the enclosed earthwork at Broadlee. As the latter certainly belongs to the enclosure, there can be no doubt that the work was continued to the Borthwickvalley by the MuseleeBurn. THE LINEAR EARTHWORKS OF SOUTHERN SCOTLAND 81 Many writers (Smail 1879, 113; Lynn 1898, 78; Craw 1924, 42) assume that 134 continued the earthwork from the head of a dry gully above the Hoscotepolicies. The RCAHMS rightly recognizes this as a work of a completely different trend and follows Kennedy (1860, 121) in accepting 287 as bridging the hiatus between the south end of this earthwork and the HoscoteBurn, with a course meeting the burn at NT 3845 1115. The case for accepting 287, though, is not proven, as this earthwork shows evidence of recent modification, perhaps as a field boundary, and the characteristics of the Catrail, if they existed, are obscured. Lynn’s idea of a ‘branch line’ continuing north-west past Hoscoteshielto include 12 should therefore be dismissed. Murray’s theory (1864, 37) should also be mentioned here, namely that the work terminated not on the Dean Burn but on the Clear Burn. Two earthworks, Quarry Rig (14) and Clear Burn (10), carry on the line from near the upper end of the HoscoteBurn to a point just above the head of Buck CleuchLinn. The remains of both earthworks are badly damaged by drainage and afforestation, but their alignment and character are similar to that of the Catrail proper (see below). The sectors of the ‘core monument’, to which the name Catrail has been restrictedare as follows - 281 Catrail (NT 5385 0262 to NT 5005 0365; 259-322 m OD; 4140 m) Roberts Linn - LangsideBurn The largest continuous portion of the true Catrail. Runs west from the mouth of Roberts Linn across the north ridge of Leap Hill to the west/north-west as far as the Langside Burn, a distance of c 4 km. Crosses 11 small burns flowing north into the Slitrig Water and the spurs and ridges between, irrespective of contour. Clearly visible, but in poor condition.On steep slopes its character has been degraded by natural surface drainage and on the more level ridge tops it has been obscured by the accumulation of peat. Consists of a low, rounded bank with a ditch, generally the more noticeable feature, on the uphill side.Today it rarely exceeds an overall width of 6 mand height of 0.5 m. Peat accumulation may account for some of the considerable diminution since it was first recorded 250 years ago. RCAHMS notes a counterscarp bank in one place which has vanished since 1945, but generally the work is not badly affected by human activity. Entire portion now covered by plantations within which it occupies a generous corridor, although a number of narrow surface drains cut the work, but vulnerable to damage in the course of thinning and felling and by heavy machinery using the break in which it lies. 282 Catrail (NT 4975 0405 to NT 4815 0502; 30-451 mOD; 1970 m) Foot of the Pike - Barny SikeThis earthwork, visually the most impressive surviving portion of the Catrail proper, extends from the upper limit of once cultivated land to the west of the LangsideBurn. Runs steeply across the ridge of the Pike, across the PenchriseBurn and a further, lower ridge of high ground to terminate at the head of the Barny Sike, a feeder stream of the DodBurn. Largely as described by RCAHMS with a definite ditch traceable over most of its course, flanked on the north/north-east by a main bank and on the south/south-west by a slighter counterscarp bank. It is obscured by peat moss on the more level ground at the summit of the Pike, in the valley of the PenchriseBurn, and on the moor to the west of the modern track from Peelbraehopeto Stobs, but the general character of the work is clear, with overall dimensions of 5.5-6.5 mwide and 0.7-1.2 m high. Between the top of the Pike and the modern track the earthwork is followed closely, and at points crossed by, the present boundary between the lands of PenchriseFarm and those of the Forestry Commission. The latter’s land was ploughed and planted in 1981 and it is evident that despite this portion being scheduled, it has been badly disfigured, particularilyin the area near the PenchriseBurn. 82 GAZETTEER OF GROUP 1 SITES 283 Catrail (NT 4825 0575 to NT 4770 0590; 305-366 m OD; 480 m) Head of dry hollow west of the DodBurn - Shoulder of Peel Brae The position of this earthwork is somewhat dislocated from the Catrail, but similarities in its course and character allowed its acceptance by the RCAHMS. Running from the head of a dry linear gully to the west of the Dod Burn, it ascends the ridge of Gray Coat along a slightly sinuous course, is crossed by the property boundary along the spine of that ridge and finishes abruptly on the shoulder overlooking the Allan valley. Consists of a ditch with a bank to the north.RCAHMS suggests that its segmented appearance is due to its construction in short sections, however, the occurrence of localized patches of moss and the considerable disturbance wrought by an old track, the ‘Thieves’ Road’, and banks associated with 185 render this observation most uncertain. 284 Catrail (NT 4675 0535 to NT 4630 0440; 259 m OD; 90 m) Foot of Peel Brae - Modern Dod/Priesthaughroad Consists of a broad low bank some 4 macross. No ditch. Runs from near the foot of the steep Peel Brae ridge where it has an abrupt, round terminus, to the Dod/Priesthaugh road, by which it is clearly cut. The RCAHMS has followed the received opinion that it is part of the Catrail, although the differences in form suggest that its position and alignment may well be fortuitous. Abutted near its upper end by a length of eroded head-dyke running south. 285 Catrail (NT 4561 0672 to NT 4535 0672; 256-305 m OD; 340 m) DoecleuchHill/Gray Hill - Old NorthhouseSwamp Visible on the west side of the Teindside-Priesthaughroad between DoecleuchHill and Grey Hill. Runs west, down to the edge of the boggy ground south-east of Old Northhouse. Consists of a ditch with a bank on the north and a slight counterscarp bank to the south. Best-preserved at east end near the road.There is a reduction in the vertical dimension of both themainbank and ditch to the west end, with the lowest 20 mmuch modified by run-off and artificial drainage. Both sides currently being eroded by ploughing.286 Catrail (NT 4133 0960 to NT 4047 1038; 271-297 m OD; 1140 m) TeindsideBurn - MuseleeBurn Visible in the moss at the head of the TeindsideBurn as a large ditch, probably enlarged by seasonal erosion. Runs north-west to the parish boundary by which time the ditch is 0.35 mdeep: the main bank to the north-east, 2.5 mwide and 0.25 mhigh, and a counterscarp bank to the south-west, 1.6 mwide and 0.1 mhigh. Easily traced to its termination on the MuseleeBurn, with dimensions similar to those above, though the counterscarp bank is discontinuous. Towards the north end where the work, unusually, is overlooked by high ground to the north-east, the main bank changes to the south-west side. Broken in four places by droving tracks. Has suffered erosion from livestock.Obscured by an area of peat moss just north-west of the parish boundary. 287 Catrail (NT 3890 1160 to NT 3735 1135; 259-283 m OD; 520 m) North-east corner cultivated lands of Girnwood - Tributary of the Dean Burn Commences at the track which passes to the north of the cultivated lands of Girnwoodand borders a new conifer plantation. Runs north/north-west across two small tributary streams of the Dean Burn to end abruptly at the bank of a third.RCAHMS regards only the southern two-thirds of this length to THE LINEAR EARTHWORKS OF SOUTHERN SCOTLAND 83 be part of the Catrail (NT 3890 1160 - NT 3796 1216) dismissing the remainder as being ‘probably an agricultural division’. However, the line is unbroken and on the same alignment, suggesting a common usage if not origin. The ditch is 2.5 mwide, very shallow, and with a bank on the east/north-east side and in the former section only is an exiguous bank. The main bank at some 0.8 mis surprisingly high, which suggests that if this earthwork belongs to the Catrail it has been raised and the ditch widened, giving credence to the RCAHMS suggestion that it was used as an agricultural division; the current boundary, a stone wall, runs alongside. Much of the earthwork now runs within a young conifer plantation and the drainage ditches associated with this development cut it in a number of places. 288 Possibly Catrail: Priesthaugh (NT 4646 0480 to NT 4660 0510; 219-223 mOD; 300 m) Runs from the edge of a small copse of deciduous trees across the valley floor north of PriesthaughFarm to a thicket. Consists of a low, rounded bank, much reduced by cultivation, and a very slight ditch on its east side.The bank is nowhere greater than 0.3 mhigh, its cropmarkbeing some 3 mwide. Runs along a gap in the line of the Catrail, but it may simply represent an old drainage dyke. 289 Possibly Catrail: DoecleuchHill (NT 4562 0640 to NT 4561 0672; 305-312 mOD; 310 m) Runs from the south end of Gray Hill along the ridge of DoecleuchHill just east of its crest, through a mature conifer plantation towards the middle of a large field of improved pasture. South end severely damaged by cultivation. North end overrides the filled ditch of the Catrail (285) and abuts its bank close to the modern road between Teindside and Dod.
Where both are still visible, the bank (2 mby 0.25 m) lies on the east side of the ditch (1.5 by 0.3 m). Badly damaged by ploughing and plantation trenches, the only well-preserved section being the 15 m at the north end used as a field boundary at present.Probably an early agricultural dyke and not part of the Catrail.


The best derivation for the name Myrddin has been proposed by Dr. G. R. Isaac of The National University of Ireland, Galway:

"The attested name MYRDDIN reflects an earlier, not directly attested
*MYR-DDYN, with the second element DYN 'man, person', and the first element
MYR- which is found in the name of the Old Irish goddess-type figure
MORRIGAIN (who also prophesies), and in English night-MARE, and also in
several Slavic words (see Indogermanisches etymologisches Woerterbuch p.
736). The basic meaning was 'supernatural being, elf, goblin, phantom' or the like.  So *MYRDDYN was originally something like 'elf-man'. Note his patronymicon MYRDDYN FAB MORFRYN 'elf-man son of elf-hill'.

"The form Mórrígain with long vowel in the first syllable, so 'Great Queen',
occurs from the Middle Irish period on, due to folk etymology, as we call
it (the alteration of a form due to the mistaken belief in a false
etymology). T. F. O'Rahilly thought this was the best form, and enshrined
it in his EARLY IRISH HISTORY AND MYTHOLOGY, followed by some later
commentators. However, O'Rahilly is no reliable source for linguistic
analysis, and in this case, he is certainly wrong. As Thurneysen had
already demonstrated in his DIE IRISCHE HELDEN- UND KÖNIGSAGE, the form with long first vowel is late, and the earlier, correcter, form is
Morrígain with a short vowel in the first syllable, which cannot be
connected with the adjective mór 'great' which has a long vowel (and of
which the earlier form is már in any case)."

This Myr-ddyn > Myrddin is a much more satisfactory explanation than the previously offered theory that Myrddin was derived from the city name of Carmarthen, ancient Caerfyrddin, Roman Moridunum. This last notion derives from the fanciful HISTORY OF THE KINGS OF BRITAIN by Geoffrey of Monmouth. In this source, Myrddin is found as a boy at Carmarthen. The whole story is Geoffrey’s alteration of Nennius’s tale of the boy Ambrosius being found at Campus Elleti in Glamorgan.


The ancient Classical writer Procopius (in his HISTORY OF THE WARS, VIII, XX. 42-48) said:

Now in this island of Britain the men of ancient times built a long wall, cutting off a large part of it; and the climate and the soil and everything else is not alike on the two sides of it. For to the south of the wall there is a salubrious air, changing with the seasons, being moderately warm in summer and cool in winter… But on the north side everything is the reverse of this, so that it is actually impossible for a man to survive there even a half-hour, but countless snakes and serpents and every other kind of wild creature occupy this area as their own. And, strangest of all, the inhabitants say that if a man crosses this wall and goes to the other side, he dies straightway… They say, then, that the souls of men who die are always conveyed to this place.”

From the Welsh poem "The Dialogue of Myrddin and Taliesin" (BLACK BOOK OF CARMARTHEN), we learn that at Myrddin’s Battle of Arfderydd:

Seven score chieftains became gwyllon ;
In the Wood of Celyddon they died.

Gwyllon or "Wild Ones" is a word deriving from gwyllt, "wild". The Welsh epithet for Myrddin is, of couse, Gwyllt. Myrddin Gwyllt is Myrddin "the Wild".

But as Tolstoy pointed out, there is something odd about these two lines.  The gwyllon or “Wild Ones” are equated with the warriors who died in the battle!  The word “died” in the poem’s second line is Middle Welsh daruuanan.  Modern Welsh has darfyddaf or darfod, which according to the authoritative “Geiriadur Prifysgol Cymru” (A Dictionary of the Welsh Language) has the following meanings:

To come to an end, end, conclude, finish, complete, terminate, cease; expire, die, languish, weaken, fail, fade, decline, perish

There is thus no ambiguity in the poetic passage we are considering.  The warriors who became “Wild Ones” did not simply “go away” or “disappear” – they died.  In this context, then, to become gwyllon means to become a roving spirit that has left its battle-slain body behind.  To exist as a “Wild One” is to exist in spirit-form after the death of the body.

The Christian medieval mind either could not accept this notion of wandering spirits or, just as likely, misunderstood it.  The gwyllon were transformed into living madmen who leapt or flitted about the forest much as did their Irish counterpart, Suibhne Geilt. 

In another Myrddin poem, "Greetings" (BLACK BOOK OF CARMARTHEN), we are told by Myrddin himself:

The hwimleian speaks to me strange tidings,
And I prophesy a summer of strife.

Hwimleian or "Grey Wanderer" is yet another word for a spirit or spectre.

Myrddin "the Wild" was thus the soul of a dead man, wondering the woods of Celyddon. We must presume that it was believed that certain individuals with necromantic skills could communicate with or perhaps even channel such spirits and thus extract from them information only available to the dead, e.g. prophecies.

Difficult as it is for us to believe in this modern age, ancestral spirits were once worshipped or placated with offerings.  The 'Dis Manibus' or "For the Divine Spirits of the Dead" was a common heading for Roman period memorial stones. Morddyn the "Elf-Man" may have been such a spirit of the dead.

In the case of Merlin, then, we have a warrior-bard who is slain in battle at Arfderydd.  His spirit wandered the Lowland Scottish forest, but had its proper home atop the Lugh Mountain of Eildon Mid Hill.  There, in “Merlin’s Chapel”, the Bronze Age cairn whose interior symbolized the Otherworld/Underworld and which was also a portal to the Otherworld, the spirits of the dead who had worshipped Lugh in life congregated.  Although thanks to Geoffrey of Monmouth and others Myrddin/Merlin came to be conflated with Lugh or perhaps seen as a human avatar of the god, the most judicious interpretation of the early sources does not warrant such an identification of ghostly warrior-bard with divine entity. 


We need now address the problem of Myrddin's identification with the madman Lailoken, W. Llallogan or Llallawc. In his edition of Geoffrey of Monmouth's Life of Merlin, Basil Clarke proposes that Myrddin was originally a certain Llallawg or Llallogan who went mad at Arfderydd. Lailoken/Llallogan, suggests Clarke, may derive from a placename similar to the Gaulish village Laliacensis, which was named after the family of Lollius Urbicus, a second century Governor of Britain.

However, the word llallawc/llallogan (linked to W. llall, "other", a reduplicated form of ail/eil) is used in Welsh poetry, where it denotes something like "lord" or "friend" or "dear friend". W.F. Skene, in his The Four Ancient Books of Wales, translates the word as "twin brother". Ifor Williams translates it as "friend" in The Poems of Llywarch Hen. Egerton Phillimore and A.O.H. Jarmon make a case for it being a true personal name.

The crux of the issue is found in the Welsh poem, "The Prophecy of Myrddin and Gwenddydd, His Sister". There Gwenddydd addresses her brother Myrddin as both Llallogan (once) and Llallawc (several times). The question then is: are we to see Llallogan/Llallawc as merely a term of endearment applied to Myrddin or as a separate person with whom Myrddin was identified?

The MS. Cotton Titus A. XIX contains the tales of "Kentigern and Lailoken" and "Meldred and Lailoken". Jocelyn's 12th century Life of St. Kentigern also mentions Lailoken in its last chapter. In both, Lailoken is identified with Merlin.

In the story of Meldred and Lailoken, Merlin suffers a triple death at the hands of Meldred's shepherds.  The triple death is a motif found in Celtic literature and was a special kind of death meted out to kings, heroes and gods alike.  It is, in essence, a form a human sacrifice, made sacred by the use of three simultaneous methods of killing.  In their book The Life and Death of a Druid Prince, Anne Ross and Don Robins hypothesize that each method of killing was sacred to a specific god.  The three Celtic divinities in question have been tentatively identified with Taranis the thunder god, Esus the lord and master and Teutates, the overall god of the people.  Doubtless there were regional counterparts to these deities. 

Nikolai Tolstoy used the example of Lailoken's triple death to more closely link Merlin with the god Lleu/Lugh, as Lleu in Welsh tradition also suffers a triple death. 

The true importance of Merlin's triple death on the Tweed near the Dunmeller of Meldred has been overlooked, however.  For it clearly demonstrates that by the time of the composition of the Life of St. Kentigern, any memory of the poetic device which employed a state of madness to represent a post-death spectral existence had been lost.  No one understood any longer that Merlin or, rather, Myrddin, had died at the Battle of Arfderydd.  And because his madness was taken literally, a story was concocted which sought to demonstrate his divine nature by making him the victim of a three-fold human sacrifice.

That such a form of human sacrifice was actually practiced has been proven by examination of the Lindow Man, who was struck in the head by an axe, choked with a garrote and drowned in a pool (see again The Life and Death of a Druid Prince).  By undergoing such a sacred killing, the Lindow Man was acting the role of a god like Lleu.  And by acting such a role, he became the god – a sun god who was seasonally killed and reborn. 

The corollary of the three-fold death was, of course, resurrection.  The god Lleu, who in death is represented as a Jupiter-eagle in an oak tree, was brought back to life by Gwyddion.  The Lindow Man as Lleu or a similar sun god, by dying, helped insure the seasonal rebirth of the sun god.  For as the sun died each year, so was it reborn.  There cannot be life without death.

Merlin under his nickname Llalogan thus becomes the god in death.  But in life, he was a warrior serving his lord Gwenddolau, who also perished at Arfderydd.

The model for Myrddin’s/Lailoken’s death at Dunmeller may have been borrowed from another “elf” figure.  In Welsh tradition we find a character called Gyrthmwl Wledig (variants Gwerthmwl, Gwrthmwl, etc.).  In Triad 1, he is mentioned as Chief Elder of Penrhyn Rhionydd in the North. Arthur is Chief Prince of the same place, while Cynderyn Garthwys [St. Kentigern] is the Chief Bishop. We have seen above that Lailoken was brought into connection with Kentigern.

In Triad 44, immediately after mention of the Arfderydd battle is made, Gyrthmwl's sons are said to have ridden onto Allt or Rhiw Faelawr or Faelwr in Ceredigion. This is the site of the fort known as Dinas Maelawr (modern Pendinas), the home of the legendary Maelor Gawr or Maelor "the Giant". Dinas Maelawr is also called Castell Maylor and was built upon a high hill or ridge beside the river Ystwyth. Gyrthmwl’s sons ride up the hill to avenge their father, whom Maelor is said to have killed.

We have seen that the Life of St. Kentigern places Merlin at Dunmeller, modern Drumelzier, on the Tweed. It is here that he is captured by the chieftain of Dunmeller, Meldred (cf. Maldred, the name of the early 11th century Scottish Lord of Carlisle and Allerdale in Cumbria), and it is Meldred’s shepherds who kill him.

The chieftain Gyrthmwl in Triad 63 is referred to as "Ellyll Gyrthmwl Wledig", the word ellyll meaning "spirit, phantom, ghost, goblin, elf, fairy". In her note to Triad 63, Rachel Bromwich suggests that the word ellyll in this context may be related to that of the Gwyllt epithet used for Myrddin and the Irish Geilt epithet used for the madman Suibhne.

To quote the relevant Triad in full:

Tri Tharv Ellyll Ynys Brydein:
Ellyll Gvidawl
Ac Ellyll Llyr Marini
Ac Ellyll Gyrthmvl Wledic

Three Bull-Spectres of the Island of Britain:
The Spectre of Gwidawl
The Spectre of Llyr Marini
And the Spectre of Gyrthmwl Wledig.

Ellyll is cognate with Irish Ailill, a personal name with the same meaning.  Rather remarkably, the best etymology for ellyll and Ailill is yet another reduplicated form of ail/eil, “other”.  To quote from Professor Daniel Melia of the University of California, Berkeley (personal communication):

“I’m citing CELTICA3, 1956, which reads as follows in its entirety:
  ‘The contracted form of Ailill gen. Ailella in all the manuscripts of the genealogies which I have read (Rawl. B 502; Laud 610 ; LL ; BB; Lec. ; H 2.7) is always Aill-, Aill-a.  These contractions are quite abnormal.
    Ailill is without a doubt cognate [1] with Welsh ellyll "ghost, elf, etc." and this suggests that the older form of the name was Aillill which became Ailill with the same kind of dissimilation we find in cenand < cenn-fhind and menand < menn-fhind. [2]
    Aill-, Aill-a would be perfectly normal contracted forms of Aillill, Aillilla and probably go back to a time when these were the current forms.
[1] Unless the absence of syncope points to its being borrowed; but in that case we should expect it to be indeclinable.
[2] Mennand Wb. 9 c 34 may be a pre-dissimilation form. ‘

The Geiriadur Prifysgol Cymru agrees with this meaning (perhaps influenced by O'Brien's article, however.)  They see a connection, as do you, with a reduplicated form of proto-celtic *allo- "other" as in the Gaulish tribal name "Allobroges" < *allo "other" + bro- "border" -> "country".  The Wurzburg glosses on the Pauline Epistles (Wb.) date from ~600-~750, so the form "Aillill" would presumably, by O'Brien's argument, have still been current, at least amongst literate intellectuals, in that period.”. 

Dr. Ranko Matasovic of “The Etymological Lexicon of Proto-Celtic” (again via private correspondence) says

“Welsh ellyll is indeed cognate with Mir. Ailill, but these names cannot be related to English elf, which is frfom Germanic *albiyo-.  It is certainly possible that these words contain the stem *al- (actually *h2el-, in a more modern notation), the plain pronominal stem that meant ‘other, different’ (Lat. alius, Gr. allos, etc.).  I would add that Alladhan, the name given to Llallogan in the Irish Suibhne Geilt story, would appear to be from Irish allaid, ‘wild’.  The most likely etymology for allaid is the same al- root, as an ‘Other’ is someone who lived beyond the civilized world and was hence barbarous or ‘wild’, a stranger or foreigner or an enemy, i.e. someone deemed dangerous because he did not belong to one’s native land. The evolution of meaning would be similar to the development from Latin silvaticus, ‘belonging to woods’ to French sauvage.” 

Dr. Graham Isaac of The National University of Ireland, Galway, and Professor of Celtic Thomas Charles-Edwards of Jesus College, Oxford, both agree with this derivation for ellyll/Ailill.  

In my article Camelot and Other Arthurian Centres (Vortigern Studies Website), I showed how Camelot derives via the Old French from an earlier Romano-British Campus Alletio.  Alletio is the name of a god whose name contains the same ail/eil root.  Anne Ross has suggested Alletio may, therefore, be the ‘God of the Otherworld’.  If so, it is rather remarkable that this divine name would appear to be similar is not identical in meaning with both Llallogan/Llallawg and ellyll/Ailill (and with Irish Alladhan; see below). 

The first version of the Triad 44 in Peniarth MS. 27 adds that a giant Maelwr was slain in his fort by Gyrthmwl's sons in revenge for the murder of their father. Lailoken, as already mentioned, is killed by the shepherds of Dunmeller. Gyrthmwl, given that Penrhyn Rhionydd has been very plausibly identified with the Rhinns of Galloway (see Bromwich's note to Triad 1), looks to be the 8th century Bishop of Whithorn, Frithuwald.

The discrepancy in date between Frithuwald and Kentigern (d. 603) is probably due to Nennius's statement that the Bernician chieftain "Friodo(l)guald", i.e. Frithuwald, was ruling in the last quarter of the 6th century. This Friodo(l)guald is mentioned along with Deoric son of Ida, Hussa, Urien Rheged, Rhydderch Hen, Gwallawg and Morgan. Interestingly, in one of the Wessex royal pedigrees a Frithuwald is made the father of Woden.

One might hypothesize that the Bernician chieftain was captured by British enemies based at Dunmeller.  As often happened with war captives, he was selected because of his high political status to be given us as a triple sacrifice.  Probably the sacrifice was intended to bring good fortune in battle with the Saxons.  By dying a triple death, Frithuwald became an incarnation of the god Lugh, who appears to have undergone a seasonal death and rebirth. 

Now Frithuwald in OE almost certainly means either "Peace-ruler" or "Protection-ruler". But Gyrthmwl Wledic (wledic is cognate with the Germanic wald) may well have been interpreted by the Welsh as "Ffridd"-ruler, i.e. "Forest-ruler". Irish has frith, "a wild, mountainous place, a forest, a deer forest", Welsh has ffridd, "wood, wooded land, a mountain pasture, a sheep walk" and ME has frid, "deer park", from AS fyrhth(e), "wood, woodland".

Triad 63, which refers to Frithuwald/Gyrthmwl as a ‘tharv ellyll’ or "bull-spectre", also has a variant, wherein the phrase ‘charw ellyll’ or "stag-spectre" is substituted.

If "stag-spectre" is the more proper form, then we would have yet another reason why Gyrthmwl’s death story was borrowed for that of Myrddin, as Geoffrey of Monmouth tells us that Merlin took the form of a stag. 

To sum up: it seems fairly certain that the story of Gyrthmwl the Ellyll’s death at Dinas Maalwr on the Ystwyth was borrowed in order to provide Myrddin with a death-story at Dunmeller on the Tweed.  We should remain content with the notion that Myrddin is simply Myrddin, the ‘Elf-man’.  After all, we have plenty of evidence for Aelf- or “elf” names among the Anglo-Saxons.  Aelfric is a good example.  The proper name Ailill (see above), the early Irish cognate of Welsh ellyll, “elf”, was quite common.  There would appear to be no reason, therefore, to propose that Myrddin only became the Elf-man after his death at Arfderydd. 


According to Dr. Graham Isaac (personal correspondence), the name of Myrddin’s sister (called Ganieda in Geoffrey of Monmouth) “… is perfectly transparent as it stands, meaning ‘Bright like the Day’, with gwen, the feminine form of gwyn ‘white, bright’, and dydd ‘day’.”  I have found, however, an Old Irish feminine name Findad/Findadh/Findath, mother of the 6th century St. Fintan of Clonenagh.  Findadh, which means “fair, blonde lady”, looks to me to be cognate with Welsh Gwenddydd. 

Geoffrey has her build Merlin’s “house” or astronomical observatory with its 70 doors and 70 windows, where he can watch Phoebus (Apollo the sun god) and Venus, it has been customary to see in her the goddess Venus herself.  One thinks of Virgil’s description of the goddess in his AENEID as ‘dea candida’, the ‘Bright or Shining Goddess”.  This megalithic house is the prototype for the “stone”, i.e. capstone of a cromlech or dolmen, under which the Lady of the Lake entraps Merlin in later romance.

Are there grounds, then, for seeing Gwenddydd as a goddess?  As she is presented to us in Geoffrey of Monmouth’s Life of Merlin, she would appear to be mortal and to function as a priestess.  But as I have written elsewhere, the Lady of the Lake was none other than the Irish goddess Nemhain (“Frenzy”).  So if the Lady of the Lake’s entrapment of Merlin derives from Gwenddydd’s building of a megalithic observatory for her brother, might not we tentatively identify her with Nemhain?  Could not Gwenddydd be an epithet for the goddess, rather than just a name?

Gwenddydd’s building of an astronomical observatory for her brother has often been compared with Merlin’s own building of Stonehenge.  And, indeed, authorities have interpreted Gwenddydd’s building of Merlin’s house as a reference to Stonehenge. The problem, of course, is that Stonehenge does not have 140 gaps between its circle of stones – or even 70, if we take the doors and windows to be the same openings duplicated.  Furthermore, the context of the “Life of Merlin”, which describes the building of the observatory by Gwenddydd, leaves little doubt that this particular stone circle is somewhere in the Scottish Lowland Caledonian Forest – not on Salisbury Plain in England.

Given that it was Geoffrey who first identified Merlin with the Ambrosius (“Divine or Immortal One”) who was said to be Lord of Gwynedd and who can be shown to be not a historical personage, but the god Lleu, and given that Lleu and Mabon the youthful sun god were identified in Welsh tradition (they are both placed in death at Nantlle, the ‘Bright Stream’ or “Stream of Lleu”, in Gwynedd), the only real possibility for Gwenddydd’s stone circle is that of the Lochmaben Stane (originally Clach-mabon, the ‘Stone of Mabon’) on the north side of the Solway.  While now only a single standing stone remains, a credible mid-19th century witness claims there was a large stone circle on the site whose stones were cleared to make room for cultivation.

Currently, there is no stone circle in the Scottish Lowlands with 70 stones.  This does not mean there never was such a circle, as some have been denuded of many of their stones and others are known to have been completely removed.  The closest such circle to the Lowlands would be Cumbria’s Long Meg and Her Daughters at Little Salkeld, sometimes called the Maughanby Circle.  According to Ekwall, Maughonby preserves the Old Welsh personal name Merchiawn.  There was a famous Dark Age Merchiawn, father of the Cynfarch or ‘March’ who left his name at the Mote of Mark hill-fort in southern Dumfriesshire.  It was this Cynfarch who was the brother of Uther, the father of Arthur (see my Shadows in the Mist: The Life and Death of King Arthur).  Whether the Maughanby Merchiawn was Arthur’s grandfather is not, alas, something we can establish.

One wonders, however, if the 70 doors and 70 windows are to be taken seriously as gaps between the upright stones of a circle.  After all, Geoffrey claims Venus is to be watched from the house Gwenddydd builds, and we know this planet’s greatest elongations take place approximately 70 days before and after inferior conjunction, i.e. Venus passes from greatest elongation as an evening star to greatest elongation as a morning star in about 140 days. 


Lastly, it might be revealing to investigate the name of Myrddin’s lord, Gwenddolau.  As it turns out, Gwenddolau is a Cymracized form of another name which is present in several early historical sources.  Gwenddolau, who perished c. 573, represents the Pictish king Cindaeladh/Cennalath of the Irish Annals (Tigernach and Ulster, respectively), who died. c. 580. In the Pictish Annals, Cindaeladh/Cennalath is called Galam cennaleph, cennaleph being a byname.

Galam Cennaleph ruled alone 552-3

G. Cennaleph ruled with Brude son of Maelcon 553-4

Brude son of Maelcon ruled 554-84

Peredur (“Praetor”, a title for the Roman governor at York) and his brother Gwrgi were present at the Battle of Arfderydd in 573, but according to the Welsh Annals they did not die until 580 – exactly the year that saw the death of Cindaeladh/Gwenddolau. 


There are some other Merlin sites whose locations are uncertain. Four of these are the Fountain of Barenton, the Forest of Broceliande, a tomb of Merlin and Merlin's springs of Galabes.

The Fountain of Barenton is none other than the mineral springs of Berrington (Berinton) near Tenbury Wells in Herefordshire. 

Broceliande is usually put in Brittany where, however, the name has not survived. Chretien says absolutely nothing about his knight leaving Britain. In this Chretien was right. Broceliande, also called Brecheliante, has been improperly identified with Breckland in Norfolk, the land of the brakes. I say improperly because the name Breckland was coined in 1925 by the archaeologist W. G. Clarke (information courtesy Edwin Rose, Norfolk County Council archaeologist).

Merlin's Forest of Broceliande (in Old French Briosque + "land") should instead be connected with Dumfries.  While Dumfries was once thought to be the "Fort of the Frisians", authorities beginning with Chalmers (see Watson) correctly identified the <fries> component of the place-name with Gaelic preas, Angl. pres(s), gen. phris, Angl. -fries, gen. pl. preas, (b)p(h)reasach, "bush, copse, thicket".  Niviane is Nefain mother of Urien Rheged and, quite possibly, a borrowing of the Irish goddess Nemhain. 

A previously unlocated grave of Merlin is said by the Prose Lancelot to be in the Perilous Forest of Darnantes atop a mountain. Darnantes or Dar-nantes is the River Dore, which flows through the Golden Valley in the Black Mountains. Dore is either from French D'ore, "golden", or W. dwr, "water", while -nantes is from W. nant, "stream, brook". The Perilous Forest of the Dore River must be in this area, which is still forested to this day. Only a couple miles west of the Dore is Mynydd Merddin, "Myrddin's Mountain", one of the traditional Welsh sites of Merlin's tomb. Watling Street ran down the Golden Valley to Caerwent (information courtesy Chris Chandler of the National Monuments Record, citing James G. Wood's 1993 article "Primary Roman Roads Into Herefordshire and Monmouthshire and the Crossings of the Severn"). However, as Mynydd Merddin is an outlier of the Black Mountains, this could well be a relocation of Merlin’s mountain in the North (see above).

As for Merlin's spring or springs of Galabes, Geoffrey of Monmouth places this site in the region of the Gewisse. In a note to his The Quest for Merlin (pp. 270-271), Tolstoy suggests that Geoffrey may have substituted the Gewisse for Nennius's Guunessi. This would mean, of course, that Galabes would be found in Guunessi. Tolstoy would appear to be mistaken here. Merlin's Galabes is pretty plainly Nennius's Guoloph, i.e. Wallop, the site of a battle between Aurelius Ambrosius and Vitalinus. Now there is a Wallop stream in the Shropshire of Vortigern, but there is another in Hampshire, the Wallop Brook, site of the villages known as the Wallops. Hampshire is within the territory of the Gewisse.

The Cair Guorthirgin of Guunessi has been identified with a site at Nant Gwrtheyrn near the northwest coast of Llyn between Yr Eifl and Nefyn (see the Gwrtheyrn entry in P.C. Bartram's A Classical Welsh Dictionary). Guunessi (Gwnnws, Gwynnys) is now a farm two miles south of Nant Gwrtheyrn.

The Spirit of the Wood is Copyright © 2008 and 2009, August Hunt. All rights reserved. Used with permission.

Comments to: August Hunt


VortigernStudies is copyright © Robert Vermaat 1999-2009. All rights reserved