I have now finished the examination of all those incidents in the Grail Quest romances which are obviously derived from some other sources than Christian legend, and which are, indeed, referred by pronounced adherents of the Christian-origin hypothesis to Celtic tradition. I have also claimed a Celtic origin for features hitherto referred to Christian legend. This examination will, I trust, convince many that nearly all the incidents connected with the Quest of the Grail are Celtic in their origin, and that thus alone can we account for the way in which they appear in the romances. The latter are, as we have seen, in the highest degree inconsistent in their account of the mystic vessel and its fortunes; the most cursory examination shows the legend to be composed of two parts, which have no real connection with each other; the older of these parts, the Quest, can easily be freed from the traces of Christian symbolism; this older part is itself no homogeneous or consistent tale, but a complex of incidents diverse in origin and character. These incidents are: the rearing of the hero in ignorance of the world and of men; his visit to the court of the King, his uncle; his slaying of his father’s murderer, the trial made of him by means of the broken sword; his service with the Fisher King; his quest in search of the sword and of the vessel by means of which he is to avenge the death or wounding of his kinsman; his accomplishment of this task by the aid of a kinsman who is under spells from which he will not be loosed until the quest be ended; the adventure of the stag-hunt, in which the bespelled kinsman tests the hero’s skill and courage; the hero’s visit to the Castle of Talismans; the prohibition under which he labours; his failure to accomplish certain acts; the effects of his failure; his visit to the Magic Castle, the lord of which is under the enchantment of death-in-life; his visit to the Castle of Maidens; his visit to the Castle Perillous; and his deliverance of the captive damsels by means of the trials which he successfully undergoes. To one and all of these incidents Celtic parallels have been adduced; these have in each case been drawn from stories which present a general similarity of outline with the Grail romances, or share with them similar guiding conceptions, whilst at the same time they are so far disconnected with them that no hypothesis of borrowing can account for the features they have in common. The inconsistencies of the romances have been explained by the fusion into one of two originally distinct groups of stories, and this explanation is confirmed by the fact that traces of this fusion may readily be found in the parallel Celtic tales. These latter, when studied by scholars who never thought of comparing them with the Grail romances, have been found to contain mythical elements which other scholars had detected independently in the romances. Those features of the romances which have perplexed previous students, the Fisher King and the omitted question, have been explained from the same group of Celtic traditions, and in accordance with the same scheme of mythical interpretation which have been used to throw light upon the remainder of the cycle. Finally, the one Celtic version of the Grail Quest, the Mabinogi, which presents no admixture of Christian symbolism, has been shown, when cleared of certain easily distinguishable interpolations, to be genuine in character, and to present the oldest form of one of the stories which enters into the romances.

I have tried not to force these parallels, nor to go one step beyond what the facts warrant. I have also tried to bear in mind that a parallel is of no real value unless it throws light upon the puzzling features in the development of the romances. I thus rest my case, not so much upon the accumulative effect of the similarities which I have pointed out between the romances and Celtic tradition, as upon the fact that this reference of the romances to certain definite cycles of Celtic myth and legend makes us understand, what otherwise we cannot do, how they came by their present shape. It now remains to be seen if this reference, can in any way explain the Christian element in the legend, which I have hitherto left almost entirely out of account. Birch-Hirschfeld’s hypothesis is condemned, in my opinion, by its failure to account for the Celtic element; although I do not think an explanation of a late and intruding feature is as incumbent upon me as that of the original Celtic basis of the legend is upon him, I yet feel that an hypothesis which has nothing to say on such a vital point can hardly be considered satisfactory. It is the Christian transformation of the old Celtic myths and folk-tales which gave them their wide vogue in the Middle Ages, which endowed the theme with such fascination for the preachers and philosophers who used it as a vehicle for their teaching, and which has endeared it to all lovers of mystic symbolism. The question how and why the Celtic tales which I have tried, not unsuccessfully I trust, to disentangle from the romances were ever brought into contact with Christ and His disciples, and how the old mystic vessel of healing, increase, and knowledge became at last the sacramental cup, must, therefore, be faced. The hypotheses set forth in the preceding page might be accepted in their entirety, and the merit of this transformation still be claimed, as Birch-Hirschfeld claims it, for the North French poets, to whom we owe the present versions of the romances. On first reading Birch-Hirschfeld’s book, I thought this claim one of the flaws in his argument, and, as will be seen by reference to Chapter IV., other investigators, who accept the Christian origin of the larger part of the legend, hold that it has been shaped in these islands, or in accordance with Celtic traditions now lost. I think we can go a step farther. A number of myths and tales have been used to illustrate the romances. In them may be found the personages through whom probably took place the first contact between Celtic mythic tradition and Christian legend.We must revert for one moment to the results obtained in Chapter III. by an examination of the way in which the Grail and its fortunes are mentioned in the romances. We there distinguished two forms of the distinctively Christian portion of the legend, the Early History. In both Joseph is the first possessor and user of the holy vessel, but in one its farther fortunes are likewise bound up with him or with his seed. He, or his son, it is who leads the Grail host to Britain, who converts the island, and by whom the precious vessel is handed down through a chosen line of kings in anticipation of the promised Knight’s coming. In the other form, on the contrary, Joseph has nothing to do with Britain, which is converted by Brons and his son, Alain; Brons is the guardian of the holy vessel, and, in one version, the fisher of the mystic fish, whilst in another his son takes this part. There is repeated insistence upon the connection between the Grail host and Avalon. Finally Brons is the possessor of “secret words,” and may not die until he has revealed them to his grandson.

This account is, we saw, later in form than the Joseph one. As we have it, it was written after the greater portion of the Conte du Graal, after that redaction of the Early History made use of by the author of the Queste and of the first draft of the Grand St. Graal. Its influence only makes itself felt in the later stages of development of the legend. But none the less it clearly represents an older and purer form of the Early History than that of the Queste and of Chrestien’s continuators. It has not been doctored into harmony with the full-blown Arthurian legend as the Joseph Early History has. It is still chiefly, if not wholly, a legend, the main purport of which is to recount the conversion of Britain.

Such a legend is surely more likely to have been shaped by Welsh or Breton monks than by North French trouvÈres. And when we notice the Celtic names of the personages, and their connection with the Celtic paradise, Avalon, there can remain little, if any, doubt respecting the first home of the story. We may thus look upon Brons, owner of a mystic vessel, fisher of a mystic fish, as the hero of an early conversion legend. But the name Brons has at once suggested to most students of the cycle that of Bran. The latter is, as we saw in the last Chapter, the representative of an old Celtic god of the otherworld. He is the owner of the cauldron of renovation. He is also the hero in Welsh tradition of a conversion legend, and is commonly known as Bran the Blessed. Unfortunately the only explanation we have of this epithet occurs in a late triad, to which it is not safe to assign an earlier date than the fourteenth century. He is described therein as son of Llyr Llediath, “as one of the three blissful Rulers of the Island of Britain, who first brought the faith of Christ to the nation of the Cymry from Rome, where he was seven years a hostage for his son Caradawc.”[138] But if late in form this triad may well embody an old tradition. It gives the significant descent of Bran from Llyr, and thereby equates him with Mannanan Mac Lir, with whom he presents otherwise so many points of contact. It is quite true that the Bran legend, as is pointed out to me by Professor Rhys, is mentioned neither in the earliest genealogies nor in Geoffrey. But it should be noted that the Grand St. Graal does bring one member of the Brons group, Petrus, into contact with King Luces, the Lucius to whom Geoffrey ascribes the conversion. Again, the epithet “blessed” is applied to Bran in the Mabinogi of Branwen, daughter of Llyr. I have placed this tale as a whole as far back as the eleventh-tenth centuries, and my arguments have met with no opposition, and have won the approval of such authorities as Professor Windisch and Monsieur Gaidoz. But the Mabinogi, as we have it, was written down in the fourteenth century; the last transcriber abridged it, and at times did not apparently understand what he was transcribing. By his time the full-blown Bran legend of the triad was in existence, and it may be contended that the epithet was due to him and did not figure in his model. On the other hand, Stephens (Lit. of the Cymry, p. 425) quotes a triad of Kynddelw, a poet of the twelfth century, referring to the three blessed families of the Isle of Britain, one of which is declared by a later tradition to be that of Bran.[139] Again, the triads of Arthur and his Warriors, printed by Mr. Skene, Four Ancient Books, Vol. II., p. 457, from MS. Hengwrt, 566, of the beginning of the fourteenth century, and probably at least fifty years older, mentions the “blessed head of Bran.”[140] On the whole, in spite of the silence of older sources, I look upon the epithet and the legend which it presupposes as old, and I see in a confusion between Bran, Lord of the Cauldron, and Bran the Blessed, the first step of the transformation of the Peredur sage into the Quest of the Holy Grail. In the first capacity Bran corresponds to the Lord of the Castle of Talismans. From the way in which the fish is dwelt upon in his legend, it may, indeed, be conjectured that he stood to Peredur in some such relation as Finn-eges to Fionn. As hero of a conversion legend he came into contact with Joseph. We do not know how or at what date the legend of the conversion of Britain by Joseph originated. It is found enjoying wide popularity in the latter half of the twelfth century, the very time in which the romances were assuming their present shape. WÜlcker (Das Evangelium Nicodemi in der abendlÄndischen Literatur, Paderborn, 1872) shows that the legend is not met with before William of Malmesbury; and Zarncke, as already stated (supra, p. 107), has argued that the passage in William is a late interpolation due to the popularity of the romances.[141] But to accept Zarncke’s contention merely shifts back the difficulty. If William did not first note and give currency to the tradition, the unknown predecessor of Robert de Borron and of the authors of the Queste and Grand St. Graal did so; and the question still remains how did he come by the tradition, and what led him to associate it with Glastonbury. Birch-Hirschfeld, it is true, makes short work of this difficulty. The fact that there is no earlier legend in which Joseph figures as the Apostle of Britain is to him proof that Borron evolved the conception of the Grail out of the canonical and apocryphal writings in which Joseph appears, and then devised the passage to Britain in order to incorporate the Arthurian romances with the legend he had invented. It is needless to repeat that this theory, unacceptable on a priori grounds, is still more so when tested by facts.

But Joseph under other aspects than that of Apostle of Britain is worthy of notice. The main source whence the legend writers drew their knowledge of him was the Evangelium Nicodemi, the history of which has been investigated by WÜlcker. The earliest allusion in western literature to this apocryphal gospel is that of Gregory of Tours (WÜlcker, p. 23), but no other trace of its influence is to be met with in France until we come to the Grail romances, and to mystery-plays which relate Christ’s Harrowing of Hell. In Provence, Italy, and Germany the thirteenth and twelfth centuries are the earliest to which this gospel can be traced. In England, on the contrary, it was known as far back as the latter quarter of the eighth century; Cynewulf based upon it a poem on the Harrowing of Hell, and alludes to it in the Crist; the ninth century poem, “Christ and Satan,” likewise shows knowledge of it, and there is a West-Saxon translation dating from the early eleventh century.

Whence this knowledge and popularity of the gospel in England several centuries before it entered prominently into the literature of any other European people? WÜlcker can only point by way of answer to the early spread of Christianity in these Islands, and to the possibility of this gospel having reached England before it did France or Germany. He also insists upon the early development of Anglo-Saxon literature.

Whether the fact that the apocryphal writings which told of Joseph were known here when they were unknown on the Continent be held to warrant or no the existence of a specifically British Joseph legend, they at all events prove that he was a familiar and favourite legendary figure on British soil. It would be rash to go any farther, and to argue from the inadequacy of the reasons by which WÜlcker seeks to account for the early knowledge of the Evangelium Nicodemi in England, that Joseph enjoyed particular favour among the British Christians, and that it was from them the tidings of him spread among their Saxon conquerors.

The legendary popularity of Joseph in these islands, though not in any special capacity of Apostle of Britain, is thus attested. Let us admit for argument’s sake that the conversion legend did first take shape in the twelfth century, is it not more likely to have done so here, where the apocryphal writings about him were widely spread, than in France, where they were practically unknown? And why if Borron, or any other French poet, wanted to connect the Holy Vessel legend which he had imagined with Arthur, should he go out of his way to invent the personages of Brons and Alain? The story as found in the Queste would surely have been a far more natural one for him. And why the insistence upon Avalon? We have plain proof that Borron did not understand the word, as he explains it by a ridiculous pun (supra, p. 78).[142]

These difficulties are met in a large measure if we look upon Bran (Brons) as the starting point of the Christian transformation of the legend. In any case we may say that a conversion legend, whether associated with Joseph or anyone else, would almost inevitably have gravitated towards Glastonbury, but there are special reasons why this should be the case with a Bran legend. Avalon is certainly the Welsh equivalent of the Irish TÍr na n-Og, the land of youth, the land beyond the waves, the Celtic paradise. When or how this Cymric myth was localised at Glastonbury we know not.[143] We only know that Glastonbury was one of the first places in the island to be devoted to Christian worship. Is it too rash a conjecture that the Christian church may have taken the place of some Celtic temple or holy spot specially dedicated to the cult of the dead, and of that Lord of the Shades from which the Celts feigned their descent? The position of Glastonbury, not far from that western sea beyond which lie the happy isles of the dead, would favour such an hypothesis. Although direct proof is wanting, I believe that the localisation is old and genuine: Bran, ruler of the otherworld, of Avalon, would thus come into natural contact with Glastonbury; and if, as I assume, Joseph took his place in the conversion legend the association would extend to him. The after development of the legend would then be almost a matter of course. Bran, the ruler in Avalon, would pass on his magic gear (cauldron, spear, and sword, as in the case of the Tuatha de Dannan) to Bran the Blessed, who would in his turn transfer them to Joseph. And once the latter had entered into the legend, he would not fail to recall that last scene of the Lord’s life with which he was so closely associated, not by any pseudo-gospel but by the canonical writings themselves, and thus the gear of the old Celtic gods became transformed into such objects as were most prominent in the story of the Passion and of the scene that immediately preceded it. The spear became that one wherewith Christ’s side was pierced. As for the vessel, the sacramental nature is the last stage of its Christian development; its original object was merely to explain the sustenance of Joseph in prison, and to provide a miraculous refreshment for the Grail host, as is shown by the Early History portion of the Conte du Graal and by the Queste. In a dim and confused way the circumstances of the Resurrection helped to effect the change of the pagan resuscitation-cauldron into a symbol of the risen Lord. And some now lost feature of the original legend—some insistence upon the contents of the vessel, some assimilation of them to blood—may have suggested the use to which the vessel was first put.

This hypothesis assumes many things. It assumes a Bran conversion legend, of which the only evidence of anything like the same date as the romances is a single epithet; it assumes that the hero of this legend was originally an old Celtic divinity; it assumes a Joseph conversion legend, for which there is really no other evidence than that of the romances; it assumes the amalgamation of the two legends, and that Joseph took over in a large measure the rÔle and characteristics of Brons. And when it is recollected that the primary assumption, the identification of the two Brans, rests in a large measure upon the appearance of the fish in the Brons legend, that this fish is nowhere in Celtic tradition associated with Bran, that it is associated on the other hand with a being, Fionn, whom we have compared with Peredur, but that it is absent from the Peredur-saga, the hypothesis must be admitted to be of a tentative nature. I fully appreciate the force of the objections that can be urged against it; at the same time it has the merit of accounting for many puzzling features in the legend. When in the same story two personages can be distinguished whose rÔle is more or less of the same nature, when the one personage is subordinated in one version and has disappeared altogether from the other, it is quite legitimate to conclude that two originally independent accounts have become blended, and that one has absorbed the other. The hypothesis is on safe ground so far. It thus explains the presence of Brons in the legend, as well as his absence from some versions of it; it has something to say in explanation of the connection with Glastonbury; it explains in what way the Celtic traditions were started on their path of transformation; and it provides for that transformation taking the very course it did. There is nothing to be urged against it on a priori grounds; once admit the premisses, and the rest follows easily and naturally. Its conjectural character (the main objection to it) is shared in an even higher degree by the other hypotheses, which have essayed to account for the growth and origin of the legend, and they have the disadvantage of being inherently impossible.

In the light of the foregoing investigations and hypotheses we may now amplify the sketch history of the whole cycle given in Chapter III. The Peredur-saga probably came into existence in much its later form at an early date in the Middle Ages. A number of older mythical tales centered in a, perhaps, historical personage. The circumstances of his life and adventures may have given them not only cohesion, but may also have coloured and distorted them; nevertheless they remained, in the main, mythical tales of the same kind as those found all over the world. One of these tales was undoubtedly a Cymric variant of the Celtic form of the Expulsion and Return formula; another dealt with the hero’s journey to the Land of Shades; traces of many others are to be found in the Mabinogi. Another Celtic worthy, Gwalchmai, was early associated with Peredur, and the two stood in some such relation to each other as the twin brethren of a widely spread folk-tale group. Curiously enough, whilst comparatively few incidents in the Peredur-saga were worked up into the version which served as immediate model to the North French romances, that version contained many adventures of Gwalchmai’s which have not been preserved in Welsh. We can trace three main crystallizations of the original saga-mass; one represented by the Proto-Mabinogi contained the feud quest, and, probably, some only of the other adventures found in the present Mabinogi; the second, based more on the lines of the Expulsion and Return formula, is represented by the Thornton MS. romance; in the third the feud quest was mixed up with the hero’s visit to the Bespelled Castle, and those portions of the Gwalchmai-saga which told of his visit to Castle Perillous as well as to the Bespelled Castle. Whilst the Proto-Mabinogi was probably in prose, the Proto-Conte du Graal was probably in verse, a collection of short lais like those of Marie de France. Meanwhile, one of the chief personages of the older mythic world which appear in the Peredur-saga, Bran, the Lord of the Land of Shades, of the Bespelled Castle, of the cauldron of healing, increase and wisdom, and of the knowledge-giving salmon, had become the Apostle of Britain, his pagan attributes thus suffering a Christian change, which was perfected when Joseph took the place of Brons, bringing with him his gospel associations and the apocryphal legends that had clustered round his name. Thus a portion of the saga was Christianised, whilst the other portion lost its old, fixed popular character, owing to the fusion of originally distinct elements, and the consequent unsettling both of the outlines and of the details of the story. Incidents and features which in the earlier folk-tale stage were sharply defined and intelligible became vague and mysterious. In this state, and bearing upon it the peculiarly weird and fantastic impress of Celtic mythic tradition, the story, or story-mass rather, lay ready to the hand of courtly poet or of clerical mystic. At first Christian symbolism was introduced in a slight and meagre way—the Brons-Joseph legend supplied the Christian meaning of the talismans, and that was all. But the Joseph legend was soon vigorously developed by the author of the work which underlies the Queste and the Grand St. Graal. He may either not have known or have deliberately discarded Brons, the old Celtic hero of the conversion, as he certainly deliberately thrust down from his place of pre-eminence Perceval, the Celtic hero of the Quest, substituting for him a new hero, Galahad, and for the adventures of the Conte du Graal, based as they were upon no guiding conceptions, fresh adventures intended to glorify physical chastity. With all his mystic fervour he failed to see the full capacities of the theme, his presentment of the Grail itself being in especial either over-material or over-spiritual. But his work exercised a profound influence, as is seen in the case of Chrestien’s continuators. Robert de Borron, on the other hand, if to him the merit must be assigned, if he was not simply transcribing an older, forgotten version, was a more original thinker, if a less gifted writer. Although he was not able to entirely harmonise the conflicting accounts of which he made use, he yet succeeded in keeping close to the old lines of the legend whilst giving a consistent symbolical meaning to all its details. His work came too late, however, to exercise the influence it should have done upon the development of the legend; the writers who knew it were mere heapers together of adventures, and the very man who composed a sequel to it abandoned Robert’s main conception.

The history of the Legend of the Holy Grail is, thus, the history of the gradual transformation of old Celtic folk-tales into a poem charged with Christian symbolism and mysticism. This transformation, at first the inevitable outcome of its pre-Christian development, was hastened later by the perception that it was a fitting vehicle for certain moral and spiritual ideas. These have been touched upon incidentally in the course of these studies, but they and their manifestation in modern as well as in mediÆval literature deserve fuller notice.