No matter how far the chroniclers of old departed from fact in the details of their narratives, they grouped the incidents around a central figure, a magnificent ancient hero; and, more than that, they specified the actual locality in which that hero had won his renown. But just as they magnified the hero out of all proportion, so they extended the area of his realm beyond all possibility: hence the difficulties that meet us in the search for truth.

Of the Celts, Ralph Waldo Emerson has perhaps left us, in brief, the best record. He sums up the greatness and the importance of the race by saying that of their beginning there is no memory, and that their end is likely to be still more remote in the future; that they had endurance and productiveness and culture and a sublime creed; that they had a hidden and precarious genius; and that they “made the best popular literature of the Middle Ages in the songs of Merlin and the tender and delicious anthology of Arthur.” This race was not likely to take a narrow view of its possessions, or to assign a small territory to its greatest monarch. Its claim may be preposterous, but that comes of the consciousness of superior strength and of daring imagination. Britain was not large enough for the Celts; they required not a country but a continent. And when their songs were sung, their stories told, and their great Arthur’s name celebrated throughout the west, they boldly affirmed that the west was his, and that he had subdued and ruled the whole civilised world. Arthur’s England became in their eyes the perfect realm, the ideal place; and the survival of this idea may be discovered in the works of the poets, old and new.

was Llywarch HÊn’s allusion to the slaughter of bards, evincing his belief in their sacred character. Song was to the Cymry at once education, a vent for national feeling, and a memorial of great events. The bard ranked beside the artisan as one of the pillars of social life. He had only one theme, his country’s hope, misfortune, and destiny; and, as M. Thierry has aptly said, the nation, poetical in its turn, extended the bounds of fiction by ascribing fantastic meaning to the words. “The wishes of the bards were received as promises, their expectations as prophecies; even their silence was made expressive. If they sang not of Arthur’s death, it was a proof that Arthur yet lived; if the harper undesignedly sounded some melancholy air, the minds of his hearers spontaneously linked with the vague melody the name of a spot, rendered mournfully famous by the loss of a battle with the foreign conqueror. This life of hopes and recollections gave charms, in the eyes of the latter Cambrians, to their country of rocks and morasses.” How much we really owe, then, to historic fact and how much to bardic song the accounting of Camelot and Avalon, Tintagel and Almesbury, as the famous and redoubtable spots of Arthurian accomplishments and occupation, would be difficult to decide. Literary genius from the first centres in the minstrel, who is both composer and singer, who stimulates to action and records events, who is himself “doer” and “seer.”

But for this rich and sustained Celtic influence our literature would be poor indeed, would be less romantic, less poetic, and lacking in the vitality of human passions, human hopes and aspirations, human suffering and despair. For the dominant note in Celtic literature—and this particularly applies to the Arthurian legend which, despite its boasts, is a story of failure—is an indefinable melancholy, an exquisite regret; the poetry may be, as Matthew Arnold said, drenched in the dew of natural magic, and the romances may be threaded with radiant lights, but there always remains the underlying sombreness of texture or the overhanging cloud-darkening of the scene. Joyous music concludes in a minor key or is broken by a sudden note of pathos. The Celtic bards sang of war, but though the heroes always went forth bravely to battle it has been recorded that they “always fell.” Victories are less frequently celebrated than defeats are mourned. The glory of the Celt was vast and transcendent, but from minstrel-times it was a fading glory. Work as the history-weavers might with the golden shuttles of romance their tears mingled with the gleaming strands, and the tissue as it left the loom was a medley of broken lights and shadows. Nevertheless, the pictures they have left us of chivalrous times remain unsurpassed for the grandeur of their conception: they remain the model and despair of all ages.

The description of Arthurian England, the “Logris” of the chroniclers, comports with the suggestions of romance, but ill accords with the facts.[5] Even if we grant the Round Table and the Quest of the Grail, the fact remains that the times were barbarous and that the Britons of the sixth century had only reached the outer borders of civilisation. The exploits of the knights themselves are indicative of a prevailing state of lawlessness verging perilously upon absolute savagery. Appalling rites were practised in the castle strongholds, and the life neither of man nor woman was deemed precious. The romancers themselves do not disguise that the purpose and the methods of the knights were little superior to the purpose and methods of those whom they warred against; and the common practice of the knights to “reward themselves” in their own ways for victories achieved disposes at once of the contention that their motives were unselfish, or that their chivalry was pure and disinterested. The England of King Arthur was therefore by no means like to be the ideal land of peace, beauty, and content which poets have imagined. Neither can we concede the whole claim to Arthur’s undisputed possession of the entire kingdom. The freedom with which the chroniclers spoke of the king’s unmolested journey north, south, east, and west, only proves that they made an unwarrantable use of names. Among the places loosely mentioned or referred to at random in the romance, or perchance confused in the writers’ minds with places within a small area, we must count all those beyond the Severn and Trent, unless we adopt the alternative theory and accept the north as Arthur’s realm. To these we add all the large proportion of places, more or less fantastically named, which seem to have had no existence out of the chroniclers’ brain. Where shall we look for Carbonek, for the land of Petersaint, for Joyous Isle, for Waste Lands, for Lonazep, for Goothe, for Case, for the Castles of Grail, La Beale Regard, Pluero, Jagent, and Magouns? to say nothing of a host of others. And are we to be deluded by the familiarity with which Jerusalem, Tuscany, Egypt, Turkey, and Hungary are spoken of, into believing that these distant places were really visited by Arthur and his knights? Even if we were to concede all the localities mentioned in Malory’s work we should be confronted by a new difficulty in the Mabinogion, where quite a fresh series of towns and countries is mentioned in addition to many of the old ones. But while in the Mabinogion the west of Europe is almost exclusively dealt with, the English, French, and German historians would be content with nothing less than the best part of the hemisphere. No petty view, however, must be taken of the Arthur-land of romance. If Caerleon was his capital, we must believe that he was not unknown north of the Humber, and that he had a castle in old Carlisle. Calydon and Brittany, Ireland and Wales, acknowledged his power and felt his sway. The Roman himself met Arthur face to face; knights carried his fame to Constantinople—so the early historians asseverate, and so they doubtless sincerely believed.

But the more cautious student will confine his attention to a group of but half-a-dozen places in South Wales, Devonshire, and Cornwall, and will doubt the truth of tradition even when it mingles with the nomenclature of the romance. Of Lyonnesse whelmed beneath the waves we have no knowledge; it is a lost and perhaps half fabulous region. Cameliard, whose boundaries are fairly well known, is strewed with doubtful relics, and preserves a multitude of strange stories. These are all that remain to us when we have traversed King Arthur’s land. Lyonnesse is reported to have been a region of extreme fertility, uniting the Scilly Isles with Western Cornwall. The hardy Silures were the inhabitants of this tract, and were remarkable for their industry and piety. No fewer than one hundred and forty churches testified to the latter quality, and the rocks called Seven Stones mark the site of their largest city. Tradition is untrustworthy as to any great cataclysm, but the Saxon chronicle declared that Lyonnesse was destroyed by a “high tide” on November 11, 1099. The assumption is that where the sea now sweeps with tremendous force, between Land’s End and the Scillies, once lay a fair region, another Atlantis, which formed no unimportant part of King Arthur’s realm. The etymology of the name Scilly is more or less doubtful. The word has been identified with Silura, or Siluria, the land of the Silures—that is, South Wales. Malory’s Surluse, or Surluce, reminiscent of the French Sorlingues, if it be not Scilly must remain unidentified. The first mention of it is in the history of La Cote Male Taile, where it is said that Sir Lancelot and the damsel Maledisant (afterwards known as Bienpensant) “rode forth a great while until they came to the border of the county of Surluse, and there they found a fair village with a strong bridge like a fortress.” A later reference shows that it was in and about Cornwall that the knights were at this time staying and seeking adventures with the king; and the “riding forth a great while to the border of the country of Surluse” would fit in with the idea that Cornwall and Scilly were not then divided by the sea, but formed part of the kingdom of Lyonnesse. Sir Tristram, who is essentially a Lyonnesse knight, was sought in the country of Surluse when he had vanished during the period of King Mark’s treachery; and there seems no doubt that, though an accessible part of the kingdom, it was a considerable distance away, and perhaps somewhat out of the beaten track. Sir Galahalt, “the haut prince,” was its ruler, and he was resorted to by the knights; but we are distinctly told that “the which country was within the lands of King Arthur,” and for that reason Sir Galahalt could not even arrange a joust without obtaining his sovereign’s consent. Again, Sir Galahalt was known as Sir Galahalt “of the Long Isles,” which admits of a fair deduction, and seems not without its significance in this argument.

The “guarded Mount,” dedicated to St. Michael, overlooks the long Atlantic waves, the waste of waters, and “towards Namancos and Bayona’s hold,” and this Ultima Thule is thronged with traditions of Arthur and his lost territory. Grim, cavernous Pengwaed, or Land’s End, with its granite rocks; the Lizard, and Penzance, the last town in England, are all stored with these old memories; and the waves flooding the bays tell of that younger time over which hangs perpetual shadow. This is the Lyonnesse of Tennyson’s imagining, the

where long hillocks dip down to the sea-line, where the coast spreads out into shifting treacherous sand, and where amid the dreary plains the Silures fought their battles for life and freedom.[6] At Vellan, Arthur slaughtered so many Danes that the mill next day was worked with blood. Land’s End still shows its “Field of Slaughter,” and by the coast Arthur and Mordred met during the last conflict. Lyonnesse may have included Armorica also, still rich with its incomparable traditions and its unsurpassed folk-songs. For once the people of Brittany, Cornwall, and Wales, speaking practically the same tongue, lavished all their poetic wealth upon the Arthurian cycle of legendary history, claimed the knights in common, and each still claims to possess the more famous shrines. Merlin’s forest thus becomes a part of Lyonnesse; Joyous Gard (as we shall presently see) can still be found in Brittany, instead of Northumberland; and Avalon, instead of being a pilgrim’s resort in Somerset, is an island off the Breton coast, seen dimly from the wild moorland country, strewn with dolmens, and reaching down to a shore of silvery sands. Between the orange-coloured rocks “the sea rushes up in deep blue and brilliant green waves of indescribable transparency. On a bright summer day the whole scene is one of unspeakable radiance. Delightful little walks wind round the western headland, where more groups of rock appear, as weird and fantastic as the first.”[7] And across the stretch of azure sea lies the dim islet which Breton legend affirms is King Arthur’s resting-place. When we consider the French sources of the history compiled by Geoffrey, Wace, and Map, the reasonableness of believing that Avalon was at first located in Brittany becomes at once apparent, and the wonder is that in this and many other cases the transference of the scenes to England should have been so complete or that English equivalents should have been so readily accepted.

The more obscure names of places would doubtless be identified if the search were more assiduous in Brittany than in Britain, and if the original Breton nomenclature were used as a basis. Tristram, Iseult, and Lancelot at least are French, and the prevailing tone of the romances in which they figure is French; we must look to Brittany for some part of the scenery.[8] At various times it has been stated that Sir Lancelot’s Joyous Gard was none other than Alnwick, or else Bamborough Castle, in Northumberland, a structure which dates from the year 554, and may have been the site of an earlier stronghold.[9] But why Sir Lancelot, a Breton Knight of Arthur’s Court, whose exploits are confined to Lyonnesse, the southern portion of King Arthur’s territory, should have had his castle located in the north cannot be determined, unless we so far revise our opinions as to credit (as some have done) the existence of a Scotch knight of that name. Instead of looking to Northumberland for Sir Lancelot’s stronghold, and endeavouring to identify Bamborough as his residence, why not turn straightway to France, his native land, and accept such facts as are there to be found? The chronicle of Malory itself says that Joyous Gard was “over sea.” Beyond the forest of Landerneau may still be seen the traditional site of a Chateau de la Joyeuse-Garde, with an ancient gateway and a Gothic vault of the twelfth century remaining. Here at least we find the name; the Breton regards the spot as that which Lancelot, the Breton knight, claimed as his own; and the scene is in that Armorica from which the original traditions sprang, or, at least, where they took earliest root.[10] In addition to Joyous Gard, Brittany boasts of its Tristan Island in the Bay of Douarnenez, named after the “Tristan des LÉonais” who was the rival of King Mark. King Mark, too (“Marc’h,” in the original, signifying horse, and so named because of his pointed ears), has his own locality, for according to Breton legend he was not ruler of Cornwall but of Plomarc’h, which place lies a little to the east of Douarnenez and contains the ruins of his “palace.” But Renan justly inquired, if Armorica saw the birth of the Arthurian cycle, how was it that we failed to find there any traces of the nativity?

Cameliard is a tract in some respects not so hard to define or locate as Lyonnesse. The town of Brecknock, three miles from which is Arthur’s Hill, seems to have marked one of its borders, and its capital was a now undiscoverable city, Carohaise. Ritson believes that Arthur’s kingdom could not have been considerable, and he is disposed to grant him the lordship only over Devon and Cornwall, with perhaps some territory in South Wales, the land called Gore or Gower. Be that as it may, his name, by a series of links, extends from Cornwall to Northumberland, from the Scillies to London, and from London to Carlisle. The British tribe, the Silures, to which Arthur belonged, occupied the region now divided into the counties of Hereford, Monmouth and Glamorgan. Brecknock and Radnor may have been added, and it is certain that Arthur had supreme dominion over Cornwall and part of Somerset and Devon. Any “kings” of these places, such as Erbin, father of Geraint, must have been tributary to him. Tacitus has left us an account of the valour, the determination, and the warrior qualities of the Silures, who had Iberian blood in their veins. It was after the Roman and Saxon invasions that they removed their seat of Government from London to Siluria, Arthur having his court at Caerleon. The Britons were a Christian race, for that religion had been introduced among the Latinised Brythonic tribes before the end of the second century. This race prevailed over the Goidels and Ivernians in the territory, and on the recall of the Roman legions one of the Brythons succeeded the Dux Britanniarum and thus became the head of the Cymry (or Cambroges, “fellow-countrymen”). Saxon Cerdic and his son Cymric for twenty years found it impossible to break through the forest districts west of the Avon, which formed the outwork of the British forces; and we may almost take it for granted that at one time the whole of the west country was in Arthur’s power, a line from Liddlesdale in the north to the southern extremity of Lyonnesse, taking in Cumberland, Wales (and perhaps Staffordshire and Shropshire), Devon and Cornwall, roughly marking the boundary. But his reported excursions north of the Trent and to the east counties would also lead to the inference that for some time the tribe overran the major part of the country. Hence we can account for the large number of scattered memorials of the monarch found in all parts of the land, though superstition may have attached his name to many places where he was absolutely unknown. Arthur’s Seats, or Quoits, abound. They are to be found both in North and South Wales, and the name seems to have been given to any rock or commanding situation which in the popular fancy was fit to bear it. In Anglesey, in the wooded grounds of Llwyliarth a seat of the Lloyd family, a rocking stone, the famous Maen Chwf, is called Arthur’s Quoit. Cefn Bryn ridge in Glamorganshire, an imposing elevation, is crowned with a cromlech, together with numerous cairns and tumuli. The cromlech, known as Arthur’s Stone, is a mass of millstone grit fourteen feet long and seven feet two inches deep, and rests upon a number of upright supporters each five feet high. In the Welsh Triads this cromlech, which is near the turnpike road from Reynoldstone to Swansea, is alluded to as “the big stone of Sketty,” and it ranks as one of the wonders of Wales. Another such stone is to be found in Moccas parish, Herefordshire, the cromlech in this case being eighteen feet long, nine feet broad, and twelve feet thick, and supported originally by eleven upright pillars. The colossal king was to have colossal monuments. Brecknockshire has several imposing memorials of Arthur. Five miles south of Brecon rise the twin peaks of the mountain range, and they are designated Arthur’s Chair. A massive British cromlech adjoining the park of Mocras Court is called Arthur’s Table. On the edge of Gossmoor there is a large stone upon which are impressed marks resembling four horse-shoes. Tradition asserts that these marks were made by the horse King Arthur rode when he resided at Castle Denis and hunted on the moors. Between Mold and Denbigh is Moel Arthur, an ancient British fort, defended by two ditches of great depth. At Rhuthyn (Ruthin) in the vicinity King Arthur is said to have beheaded his enemy Huail (Howel), to whom Gildas refers. The record might be extended indefinitely, though no valid argument can be based upon any of the facts. The indiscriminate use of Arthur’s name often shows an extravagance of imagination and a reckless disregard of what is appropriate. Between Mold and Ruthin, for instance, is Maen Arthur, a stone which popular fancy has adjudged to bear the exact impression of the hoof of the king’s steed. There is something like substantial reason for believing that the British hero was connected with Monmouth, Cardiff, and even with Dover, and either the Arthur of the Silures or another British chief seems to have reached Carlisle—that is, if the chronicles did not confuse Cardoile with Carduel. The Cumbrian Arthur figures in two ancient ballads, “The Marriage of Gawaine,” and “The Boy and the Mantle,” while Scott’s poem of Arthur and his Court at Carlisle is, of course, too well known to need more than a reference. In the time of Baeda Carlisle was known as Lugubalia, which name by corruption became Luel. The British prefix Caer, a stone fort, made the name Caer-Luel, and as such it was long known. It gradually degenerated into Carliol, and finally became Carlisle. That the ancient city should have become confused with Caerleon is natural and explicable. Yet Arthur’s connection with a portion of the north is strongly insisted on. Where Wigan now stands he fought a famous battle. Pendragon Castle in Westmoreland claims him as its founder; and passing by easy stages we find ourselves confronted with a Northumbrian Arthur. From this point the transition to Scotland itself is extremely easy, the lowland part of that country being claimed as the veritable Cameliard.

According to no mean authority, we must leave England entirely and search in the North alone for the sites, not only of King Arthur’s battles, but for all the places connected with his exploits and his residence. Badon is then found in Linlithgowshire at Bowden Hill, and the great battle of Arderydd is located at Arthuret in Liddlesdale. The Scotch Merlin and the Scotch Lancelot are the king’s companions, and a Scotch Gildas is the historian. The resting place of Avalon is then found in the caverns of the Eildon Hills, and the voice to rouse him from his charmed sleep will echo through them and “peal proud Arthur’s march from fairyland.” As a curious fact it may be mentioned that nearly all the heroes of the “Four Ancient Books of Wales” are traced to Scotland, and admittedly in the Arthurian legend the British king was connected with as northern a place as the Orkneys by the marriage of his sister to the king of those islands. Of King Arthur, the Scotch ballad rudely tells that when he ruled that land he “ruled it like a swine.” The story of the king was the diversion of James V., who may have known that Drummelziar on the Tweed could boast of a Holy Thorn like Glastonbury, that there was an Arthur’s Oven on the Carron near Falkirk, and that Guinevere’s sepulchre was at Meigle in Strathmore. Edinburgh, or Agnet, is positively represented as the site where the Castle of Maidens stood, and the lion-shaped Arthur’s Hill is supposed to confirm the tradition that here the king abode and made his name.[11] His tomb is pointed out in Perthshire, and all the machinery of the romances is claimed as of Scotch origin and invention. The names of localities are traced, and by transporting Arthur boldly to the Lowlands we account more easily for his rapid incursions into Northumberland and of the country north of the Trent, if we cannot for his equally rapid journeys to Dover and Almesbury and Winchester.

Are not the interchangeability of names and the duplication of persons and places susceptible of a very simple explanation? Caerleon, or Carduel, was confused with Carlisle, each in itself a fitting and likely place for Arthurian exploits; the historians were grievously misled as to Winchester and the part it occupied in the romances; and we know now that various contradictions simply arose from the confusion in the minds of the chroniclers, who never seemed to have been quite certain whether Caledonia and Calydon were not one and the same, whether Camelot was inland or by the sea, whether Joyous Gard was a few days’ or a few months’ journey from Cornwall, whether Camelot was in England or in Wales, whether Arthur’s “owne castell” at Tintagel could be reached by “riding all night” from London, or whether Lyonnesse was Cornwall or Brittany. A hundred topographical complexities meet us wherever we look, and the sole conclusion of the matter is that Geoffrey and his successors inextricably mixed Scotch, Welsh, and Armoric details both in regard to the stories and the localities. The historians made no effort to be consistent in their allusions, to reconcile contradictory statements, or to account for abrupt changes of scene from the South-West to the North. While they endeavoured to concentrate Arthur’s kingdom in South Wales and Cornwall they made occasional sweeps to Berwick and Edinburgh, and annihilated the distance between Dover and Carlisle. To add to the confusion there were names, especially in the Lowlands of Scotland and in the West of England, of the same derivation, and, as Mr. Glennie has demonstrated, it is as easy to discover a Caledonian Caerleon, Avalon, or Camelot as it is to discover any of them in the district once called Cameliard. The unravelling of the skein, which became more and more entangled as new hands developed the romances, is now almost an impossibility. Arthur’s own name was changed, and it has been affirmed that he is still confused with Arthurius of Gwent, and with others of like name who were distinct persons. The conclusion of the whole matter must be that names in the romances are a source of error and confusion; that different significances were attached to them by the chroniclers themselves, and that if the truth be ever established totally new meanings may be expected.

Let me here give one instance of possible confusion of names, and broach a somewhat bold theory. The name Camelford, the scene of the last battle, is by some said to be derived from the Anglo-Saxon gafol, meaning “tribute,” the spot so called marking the ford where of old time tribute was paid. The name Guildford is also declared to have a similar signification, and, in fact, to be but a variation of Camelford. If this be so, a curious point arises. Guildford is mentioned towards the close of the Arthurian history. Sir Lancelot and the king having parted company, it is recorded that Arthur “departed towards Winchester with his fellowship. And so by the way the king lodged in a towne called Astolat, which is now in English called Gilford.” Upon this Mr. Aldis Wright observes: “Guildford in Surrey is no doubt the place alluded to; but I am not aware that the name of Astolat or Astolot (Caxton) is given to it in any authentic history.” It may be argued that King Arthur would be more likely to pass through Guildford, Surrey, than through Camelford, Cornwall. But his starting point is not certain, and it must be specially noted that the Winchester to which he was making his way was not Winchester in Hampshire but “Camelot, that is, Winchester” (Book XVIII., c. 9). The unauthorised and even absurd interpolation that Camelot was Winchester at once changes the whole argument. Disregarding this misleading explanation we find that Arthur was on his way to Camelot from one of his Courts, and if Camelot was in Somersetshire it is most likely that Camelford would be one of the intermediate stages. But the importance of the whole contention is this: Astolat, as frequently mentioned in connection with the “faire maide” Elaine and Sir Lancelot’s worthiest love episode, is undiscoverable. The name is unknown outside romance; and though we are assured that it is “now in English called Gilford,” no authority can be found for the assertion. Besides, Guildford in Surrey was rather beyond the borders of the British Kingdom, even granting occasional excursions to Middlesex and Kent. But if Guildford were synonymous with Camelford, as the derivation permits us to believe, then Astolat was none other than Camelford, and at once there are light and order where formerly prevailed obscurity and confusion. Another point worth mention is that, although tradition marks Camelford as the actual scene of important events in the Arthurian history, and although from its situation, its proximity to Tintagel, and its steep hill suitable to be crowned by a baron’s castle such as Sir Bernard of Astolat possessed, we may safely surmise that it was well known to the ever-journeying knights, yet the actual name of Camelford is never mentioned in the chronicles. As it was of Anglo-Saxon origin, this omission would easily be accounted for in the earliest records, while if Astolat was the traditional name it is at once clear how it could equally be applied to Camelford and to Guildford. We must of course remember that where the chroniclers themselves sought to elucidate they too often confused; the finger-posts they set up have started many upon weary and fruitless journeys, and the guidance offered with such confidence turns out most commonly to be the most random of guesses. If, however, we may place the slightest credence in the “Astolat, which is now in English called Gilford,” as much can be said for “Gilford” being “Gafolford” or Camelford, as for its being “Gyldeford” or Guildford. The stretch of low-lying level fields on either side of the Camel, the sharp-peaked hills in the distance, the dark meres among the hills, and the angry sea lashing against the rocks visible a mile or two away, all accord with the typical scenery of King Arthur’s realm, and make us not unwilling to believe that famous Astolat was here to be found.

When all is told, when all the searching is ended, it is found that some half-dozen places only stand out pre-eminent from the host of localities in the West in each of which only a single seed seems to have germinated; and these half-dozen places, like the last citadels of the hero, resist every effort and assault of the invader to dislodge the traditions of Arthur. I have not attempted to write a history of these places, but only to say something of their aspect to-day and of the chief events and ancient traditions linked with their names. Now and again I mention facts of later date for the purpose of showing that these famous spots have continued to be the centres of activity and connected with great characters; but in the main I confine myself to the legends of Arthur and to the episodes of chivalry. To have attempted more would have entailed not only a far more comprehensive work, but the treatment of the subject in a more scientific spirit than is here displayed. The object has been to deal rather with the romantic side than with the technical, for which the deep scholarship of a Rhys or a MÜller alone can be the qualification. It is necessary to premise also that of the most conspicuous Arthurian localities nothing but the bare tradition can be recorded. That tradition lives and is cherished, but its origin is undiscoverable. The sap lingers in the branches, but the roots are detached and lost. The legend is spread everywhere, but there are no verities. The visitor to the Arthurian scenes finds nothing but eponymous names and superstitions—indeed, the evidence present leads him to other conclusions than those he seeks. He looks for a British encampment, and he finds a post-Roman; he looks for a relic of Arthur, and he finds one of Antoninus. What is persistently ascribed to the British hero, or associated with his times, is either intangible or is irreconcilable with existing facts. Castles he is said to have inhabited were built centuries after his death, and there can only remain the free speculation that they mark the site of a former structure of which no trace remains and of which no record was made. Spots which are called King Arthur’s grave, or his seat, or his hunting-ground, or his camp, neither he nor his band, it often happens, could ever have been near. We look for persons, and we find a crowd of phantoms; we eagerly watch for demonstrations, and we find myth and fable; we hope to see the clear page of history, and we find a page that is undecipherable or blotted with shadows. Records are effaced, song and story delude, the track to truth is almost closed. Everything crumbles into dust at the touch, like Guinevere’s golden hair, and nothing is now left but the pure romance. And some of us may be content and almost glad to have it so.