[103] Cf. Parzival, Book III. l. 937 et seq. I unfortunately omitted to note the reference in the prose Lancelot. The passage is on p. 127, vol. iii. of M. Paulin Paris's abridged edition.

[104] Cf. Parzival, Hertz, n. 66, p. 495.

[105] Cf. Lais inédits, M. Gaston Paris, Romania, vol. viii.

[106] Lancelot's eagerness to receive knighthood should be compared with that of Parzival. Thus Lancelot says to Yvain, 'Dictes a monseigneur le roy qu'il me face chevalier comme il a promis—car ie le veuil estres sans attendre plus,'—and again, 'ie ne seray plus escuyer.' prose Lancelot, ed. 1533, vol. i. Cf. this with Parzival, Book III. ll. 1001-2, 'nune sûmet mich nicht mêre phleg mîn nâch riters êre,' and 1158-9, 'i'ne wil niht langer sîn ein kneht, ich sol schildes ambet hân.' The correspondence is striking.

[107] 'En verité ce varlet n'est mye bien sage, ou il a este mal enseigné.' Yvain suggests that a woman has forbidden him to tell his name (which might be compared with Parzival, Book III. l. 1464). By his speech he must be de Gaulle. Ed. 1533, vol. i. (The 1533 edition has in each volume a summary of chapter contents, thus reference is easy.)

[108] MS. 751, fol. 144 vo., quoted by M. Paulin Paris in vol. iv. of Romans de la Table Ronde, p. 87.

[109] This Dame de Nohan is probably the same as the Dame de Noauz mentioned in the Charrette, l. 5389.

[110] Cf. Romania, vol. xxvi. p. 290.

[111] Legend of Sir Gawain, p. 65.

[112] M. Marillier in a review of the Voyage of Bran and Legend of Sir Gawain, contained in Revue des Religions (July-August 1899), is inclined to connect the adventure of the Fier Baiser ascribed to the son with the adventure of the Marriage of Sir Gawain ascribed to the father. Both are disenchantment stories, and both appear to belong to the class of disenchantment by personal contact. The point is an interesting and a suggestive one.

[113] The character of the fairy and the nature of Lancelot's upbringing demand a special study, for which, so far, the materials are not available. The Lady of the Lake touches on the one hand the Queen of the Other-World, on the other, Morgain la Fee. I understand that a study on the characters of Lady of the Lake, Vivienne, and Morgain, is being prepared under the direction of Dr. Schofield. For the details of Lancelot's childhood, we must wait till a critical edition of the prose Lancelot shows us whether we have any variants or traces of early redactions, to bridge the gulf between the poem of Ulrich van Zatzikhoven and the final prose romance.

[114] Cf. Introduction to M. Paulin Paris's Romans de la Table Ronde, p. 81 et seq., also M. de Villemarqué's Merlin, p. 121.

[115] Dr. Wechssler's interesting study on 'die verschiedenen Redaktionen des Graal-Lancelot Cyklus' will be referred to later on. It is an excellent statement of certain aspects of the problem, but further research shows some of his conclusions to be very doubtful. His judgment with regard to the Queste variants is certainly at fault.

[116] l. 8050 et seq.

[117] Cf. Rhys, Studies in the Arthurian Legend, chap. iii. The author remarks that to this day in some parts of Wales it is held an insult, as implying a reflection on her moral character, to call a girl Guinevere.

[118] 'Arthur gave in charge all that he had to Mordred and the queen. That was evil done that they were born, for the land they destroyed with sorrows enow. And at the end themselves the Worse (devil) began to destroy that they there forfeited (lost) their lives and their souls, and ever since are loathed in every land, that never a man will offer prayer for their souls.'

[119] This line is lacking in the oldest MS., but can be supplied from the later recension: 'Man knew not, in sooth, whether she were dead (and how she hence departed), whether she herself were sunk in the water.'

[120] The Merlin of course deals with a period anterior to this liaison, but as we possess it, it has been, as we saw above, redacted under the influence of a tradition of which the amours of Lancelot and Guinevere formed an integral part.

[121] Cf. Legend of Sir Gawain, p. 76 et seq.

[122] On this point, cf. my Legend of Sir Gawain, Mr. Maynadier's Wife of Bath's Tale (both in Grimm Library), and M. Marillier's article in Revue des religions (July-August, 1899), already referred to.

[123] I have purposely omitted Tristan, as, though a Celtic hero, he is only indirectly connected with Irish tradition.

[124] I am glad to find that M. Gaston Paris evidently holds this view, as in a note to his discussion of the tradition that Roland was Charlemagne's son as well as his nephew, in the Histoire Poétique de Charlemagne, he refers to Gawain as holding the same position.

[125] The above remarks of course refer to Gawain as connected with Arthur; originally he was probably independent. As our knowledge stands at present, the parallels between Gawain and early Irish tradition appear to belong mainly to the Ultonian cycle; while in the case of Arthur the parallels are rather to the Ossianic.

[126] In some versions eighty.

[127] As far as English opinion goes, the popularity of Tennyson's version of the Arthurian tales has operated disastrously in confusing the question. Not long ago a writer contributed to a review an article on the subject, in which he contended for the essential identity of the Tristan and Lancelot stories, naming among other parallels the fact that in both cases the hero is sent to fetch home his lord's bride—an addition due to Tennyson; Lancelot in the genuine story being unborn at the date of the marriage. As regards the Idylls, it can only be said that whereas Malory's juxtaposition of half a dozen different compilations made confusion of a subject already more than sufficiently complex, Tennyson's edifying rearrangement of Malory made that confusion 'worse confounded.' Malory is highly valuable for the Arthurian legend in his proper place, when critically compared with other versions; and has a separate and independent position as an English classic. The Idylls of the King may perhaps also be considered an English classic, but is entirely outside the range of critical Arthurian scholarship, and should never be quoted as evidence for the smallest tittle of Arthurian romance.

[128] I am not quite certain on this point. Certainly the Perceval story is earlier than we commonly suppose, and I think we may find that it had reached the ecclesiastical ascetic stage at quite an early point in the evolution of the Lancelot story.

[129] Cf. Wechssler, Über die verschiedenen Redaktionen des Graal-Lancelot-Cyklus, p. 17.

[130] Merlin, Sommer's ed., chap. xxvi. p. 343; Perceval, l. 9546 et seq.; Parzival, xii. ll. 1306-7, xiii. l. 542 et seq.; also my Legend of Sir Gawain, p. 75 et seq.

[131] I have purposely excluded the Melwas-Meleagant story from this comparison. I am not clear that it was, in its origin, a tale of conjugal infidelity; it rather appears to me to be a Pluto-Proserpine abduction tale. The abductor may at one period have been Guinevere's lover; but, as we now have it, the queen is the innocent victim of violence. Further, it is evident that the abductor had ceased to be the lover before the introduction of Lancelot into the story (cf. Lanzelet). Therefore, if originally an infidelity story, we are met by the same perplexing gap in the tradition as we find in the Mordred version.

[132] Cf. references under heading 'Gawain.' They are scattered throughout the book.

[133] Cf. Grand S. Graal, ed. Hucher, pp. 271 and 289-93.

[134] Dr. Wechssler's caution is quite right, nevertheless I think we may eventually find that Borron was really the author of some sort of a cycle.

[135] Dr. Wechssler contends for this, as the correct title, rather than Grand S. Graal.

[136] Cf. supra, p. 17.

[137] Cf. supra, p. 14.

[138] Cf. supra, p. 9.

[139] Die Sage vom Heiligen Gral, in ihrer Entwicklung bis auf Richard Wagner's Parsifal: Halle, 1898.

[140] Obviously added by M. Paulin Paris.

[141] On this point I need only refer to M. Gaston Paris, Introduction to the Huth Merlin, p. viii.

[142] I do not discuss here how far this romance represents the original Borron-Perceval poem. As it stands, it is certainly not Borron's work. The question is, are we to consider it the work of a later writer, or does it represent an early Perceval romance, worked over for cyclic purposes?

[143] Some years ago, when preparing my translation of the Parzival, I found in the Gesta Comites Andegavorum a summary of the closing events of Arthur's life closely agreeing with that of the Didot Perceval. The connection between Perceval and Angevin tradition has not, in my opinion, received sufficient attention.

[144] We have seen reason to believe that the original Perceval story did early affect the Lancelot, and this argument, which we used at first of the independent, becomes strengthened when we examine the cyclic form.

[145] If this be true, it would throw an interesting light on the conjunction of the Queste and Perceval li Gallois in the well-known Welsh MS. translated by the Rev. R. Williams. The compiler of the MS. may have had versions of the two Lancelot cycles before him and have taken the Queste from each, perhaps doubtful which was the right version.

[146] Hucher, vol. i. p. 421.

[147] Quoted by Professor Heinzel: 'Über die französischen Gralromane,' p. 177. The parallel passage is on p. 279, vol. ii. of Dr. Evans' translation, The High History of the Holy Grail; but it is not included in the Welsh translation.

[148] Professor Heinzel's study did not come into my hands till the MS. of this chapter had been sent to the press. The support afforded to my theory by the above expression of opinion was most welcome to me. A point which deserves notice in connection with this romance is the appearance in it of the above-named Briant des Illes, and the story of the death of Lohot, King Arthur's son. So far as I know, no other prose romance knows either of these characters, but Chrétien refers to both in his Erec, ll. 6730 and 1732. I think it is possible that the name given by Wolfram von Eschenbach to Arthur's son, Ilinot, may rest upon a misreading of Lohot; the story connected with the latter is certainly curiously archaic in detail.

[149] I cannot at all agree with Dr. Wechssler's view that the Galahad Queste has been largely worked over; on the contrary it has been the least tampered with of all the Arthurian romances. I shall show this presently by comparison of texts.

[150] The worst fault of Dr. Wechssler's Grail study is that he predicates the distinctive traits of Perceval as being of Galahad—to whom they never in any sense belonged. Galahad is not Perceval's understudy, much less is he his original: he is an absolutely and entirely independent creation. The only quality they have in common is that of virginity, which is not of them, but of the monkish redactors of the legend. It is certainly no part of the primitive Perceval tale.

[151] The passage which represents Gawain as admitting himself to be the slayer of eighteen out of the twenty-two knights who have lost their lives in the Queste, Baudemagus, his dearest friend according to the Merlin Suite, among them, should, I think, be printed at the end of the Queste, not at the beginning of the Mort Artur, where it is now generally found. It is entirely in accordance with the tone of the first named romance, and out of keeping with the latter. Moreover, both the Dutch Lancelot and the 1533 version print it in the former position. The compiler of the Tristan has generally been supposed to be the first to introduce the vilification of Gawain's character; in the light of Dr. Wechssler's suggestion it would be interesting to examine whether this presentment is to be found in the Tristan before its contamination with the later Lancelot-Map cycle. I think there were peculiarities in the original Gawain story, which, misunderstood by later compilers, helped to cast a false light on his character, but it is open to question whether it was the Tristan compiler or the author of the Galahad Queste who was the original propagator of calumny.

[152] The Queste writer dwells upon instances of heroes betrayed through their love of women—Samson, Solomon, etc. If he had known the earlier Lancelot-Borron story, with the instance of Merlin's betrayal by the lady who brought up Lancelot, he would surely have made use of so very à propos an illustration.

[153] I suspect this sword of being the sword of the original Perceval story, for which an edifying legend has been invented. It probably belongs to a very early stage of the tradition. I hope some day to make it the subject of special study.

[154] Cf. the Perceval of Chrétien, and more especially the Parzival of Wolfram, with the hero of the Didot Perceval or Perceval li Gallois. I consider the two first represent the independent, the two latter the cyclic form.

[155] It may be noted here that in Wolfram's version of the Perceval story—a version which, as we have seen, has certainly influenced the Lancelot legend—the Grail-bearer, Repanse-de-Schoie eventually becomes the mother of Prester John. The circumstance that the details of the begetting of Galahad are found in the Lancelot, and not in the Queste, suggests the consideration that the author of this latter romance may have worked over the section of the Lancelot in question, so as to bring it into superficial accord with his story. Or he may have worked in conjunction with one of the later redactors.

[156] Chrétien does not appear to know anything about him: in the Charrette, for instance, had he known Bohort as represented in later legend, he would certainly have made him, and not Gawain, undertake the conflict with Meleagant, for which Lancelot threatens to be too late. The role of 'helpful friend,' played by Gawain in the earlier versions of the legend, is passed over to Bohort in the later.

[157] On this point cf. what I have said before as to the development of the Chansons de Geste; p. 92 note.

[158] 1. Edited by Dr. Jonckbloet, 2 vols., 1850, will be referred to as D. L.

2. Edition in 3 vols., a complete copy is contained in the Douce collection in the Bodleian Library, referred to as 1533.

3. Morte Arthur, edited by Dr. Sommer, vol. iii., Sources of Malory, the sections entitled The Lancelot Proper, The Quest of the Holy Grail, and La Morte au Arthur; all three are referred to as S.

4. Queste del Saint Graal, ed. Furnivall—Q.

5. Morte Arthur, Sommer (vol. i. text)—M.

6. The Welsh Queste (ed. Rev. R. Williams, 1876), which I have also consulted, being, in its available form, the translation of a translation, scarcely affords reliable ground for comparison; it is, moreover, a very free rendering of the text. Nevertheless, as it is well to make use of all available versions, I have, in the cases where the original text appears to be fairly represented, added this reading under the heading W.

[159] Cf. Jonckbloet, Roman van Lancelot, vol. i. p. lvii.

[160] To speak quite correctly it really begins rather before the Agravain proper. I have noted this further on. M. Paulin Paris remarks (Romans de la Table Ronde, vol. v. p. 296), with regard to the Agravain, that we find it 'le plus souvent copié isolément, ou bien complétement séparé des autres parties.' One of the useful hints of this scholar which might have earlier been taken into consideration.

[161] In this connection it is amusing to find Dr. Wechssler (Sage vom Heiligen Gral, pp. 166-167) remarking complacently that the achievement of the adventures announced by the Grail Messenger 'wird nirgends erzählt.' The Dutch Lancelot has been edited and available for fifty years. I must own that the result of my examination of this, and of the version of 1533, equally available, has been to seriously shake my belief in the soundness and reliability of foreign criticisms of the Arthurian cycle. It is quite clear that the material at our disposal, limited as it is, has not yet been properly examined.

[162] The romances not being named in the D. L., I have adopted for convenience' sake the names given to them by M. Gaston Paris.

[163] Abstracts of these episodic romances are given by M. Gaston Paris, in vol. xxx. of Hist. Litt. de la France.

[164] Dr. Sommer says, and correctly, that the 'pomier' must be the older version.

[165] This account of Lancelot being found asleep and carried off by three queens should be compared with that of Renouart found sleeping and carried off to Avalon by three 'fays.'

I assume throughout that Dr. Sommer's summary correctly represents his text, but I admit that I have my doubts on this point; certainly in the Queste section he gives some most mistaken readings; indeed, apart from the evidence of D. L. and 1533 the whole Lancelot-Queste section needs revision. It is unfortunate that some foreign scholars have been so ready to accept Dr. Sommer's statements without taking the trouble to verify them.

[166] I do not think this is a proper name, but the equivalent of Grave = Count.

[167] No other version mentions, as does M., that the ladies won their living by 'al maner of sylke werkes,' but the whole story looks so like a copy of Yvain's adventure at the Château de Pesme Aventure that I think it may have been in his source.

[168] Of course M. Paulin Paris's book, being greatly condensed and modernised, cannot be used for textual criticism; but the compiler was a scholar of very wide learning, and there are numerous notes and hints, which we, of a later generation, make a great mistake in disregarding.

[169] This lady, never mentioned by M., plays an important rôle in the prose Lancelot.

[170] Here I take the opportunity of saying that I entirely dissent from Dr. Sommer's assertion that Gareth is the equivalent of the French Guerresches rather than Gaheret. It is this latter (in the D. L. Gariëtte) which M. renders by Gareth. I have paid a good deal of attention to this question, and have come to the conclusion that, although in the descriptive summary of King Lot's sons, found in the Lancelot, Guerresches (Gurrehes) is said to be the youngest, save Mordred, and Gawain's favourite, yet the adventures ascribed to Gaheret (variants, Gaheriet, Gariëtte, Garhiës) throughout mark him as the original of Gareth; a point which etymology alone would, I think, decide in his favour! This much is certain, wherever M. and the French versions can be compared we find Gaheret and not Guerresches. When Dr. Sommer takes it upon himself, as he does in the quotations from the French contained in the Mort Artur section, to arbitrarily change the Gaheret of all the foreign versions into Guerresches, because the latter agrees with his preconceived ideas, he is setting what I must consider as a most undesirable precedent; we cannot take these liberties with the texts and hope to arrive at a satisfactory and scientific conclusion. As pointed out in my review of Dr. Wechssler's Grail Study, once allow such a substitution, and what is to prevent us from a series of editions emendated to suit the personal views of each editor? I think myself that Gaheret and Guerresches may originally have been one, but that confusion arose from Mordred being sometimes considered as Lot's, sometimes as Arthur's, son, and that a tradition of four sons of King Lot having been established early in the evolution of the romantic story, the personality of the third was doubled to make up the correct number. This is only a suggestion, but there is certainly a confusion as to identity in the French versions, though there is no confusion as to the original of M.

[171] It seems likely that this was in M.'s source, as we read that the old man has a spear in his hand, 'and that spere was called the spere of vengeaunce.' But the old man never speaks of it to Bors.

[172] As regards the mention of Galahad and Lancelot in 1533, I find I have no special note. They are certainly not in D. L. and the two versions are in such habitual accord that I think I must have noted it had they differed here. Still, I think it only fair to point out my omission.

[173] On p. 200 of the Studies there is a mistake. Dr. Sommer speaks of the fight between Bors and Perceval and their healing by the Grail. It should, of course, be Hector, not Bors. We may note here that in this instance the Grail is stated to be the dish out of which Our Lord ate the Paschal lamb in the house of Simon the Leper; there is no mention of its containing the Blood of Christ, or of its being borne by a maiden as in M.

[174] There is no mention of Balyn's sword: this is clearly an interpolation of M.

[175] This passage throws into strong relief the absolute unreality of the Galahad Queste. The hero knows all about the Grail, its keeper, where it is to be found, his own relation to it. He has grown up under its shadow as it were. Nor need he fulfil any test to gain it: in all the records of his adventures there is no temptation such as that undergone by Perceval or Bohort; he is as fit to become keeper of the Grail (for this and not Grail-King he practically becomes) when he leaves Arthur's court as when he finally, after a series of aimless adventures, arrives at Corbenic. Contrast this with the earlier versions: the hero knows nothing of the Grail; not till after he has beheld the Talisman and failed to accomplish the necessary test does he even hear the name; when he would make amends for his negligence he can no longer find the castle, and not till he has proved himself worthy through long-continued trial is the opportunity once lost again offered to him. Neither do the inhabitants of the Grail Castle know their deliverer; they hope that it may be he, since they believe none other might find the way, but they do not know him, whereas Galahad is well known to the dwellers in Corbenic.

[176] Dr. Sommer's description of the swearing of the questers, on p. 210 of the Studies, is utterly wrong. In every version Arthur calls on Gawain to swear first, when Baudemagus interposes, saying that he who is to achieve the quest should be the first to swear. Consequently Galahad swears first, and is followed by Lancelot, Gawain, Perceval, Bohort, Lionel and Hélie le Blank. Baudemagus is in no instance the first to swear.

[177] Dr. Sommer's summary is again misleading, and entirely misrepresents the general character of the incident.

[178] Studies, p. 212.

[179] Cf. Dr. Sommer's remark on p. 212. I cannot recall a single instance in which the equivalents to M. give any other reading.

[180] On p. 212, Dr. Sommer states that Q. does not, at this point of the story, say what becomes of Perceval. This is wrong; Q. distinctly says he leaves Lancelot to return to the recluse.

[181] In his summary of the conversation on p. 213, Dr. Sommer again misrepresents his text—all agree in saying that Perceval asks his aunt about his mother and 'parens,' not that the aunt asks Perceval!

[182] The adventure of Perceval on the rock agrees closely with that of Mordrain in the Grand S. Graal. There also are two ships—in one a man who encourages, in the other a woman who tempts, him. In both cases the woman accuses the man of being an enchanter; in both her ship is covered with black silk, and she departs in a tempest. Cf. Hucher, Le S. Graal, vol. ii. pp. 354, et seq.

[183] S. Graal, ii. p. 444.

[184] As I said before, this may be due to the influence of Morien, but we must not overlook the fact that this poem certainly has some curious points of contact with the Parzival of Wolfram von Eschenbach, which also knows of the hero (or more accurately here, his son) regaining his kingdom, which he also does in Perceval li Gallois.

[185] The scribe of the original MS. may have had to condense on account of space here, which is contrary to the usual practice of 1533; but in a printed edition it is not easy to decide the real value and significance of such omissions.

[186] 1533 ten, representing the number as thirteen, Galahad taking the place of Our Lord. This is a point on which we might expect to find different readings, according as the compiler held, or did not hold, Judas to have been present at the Institution—a question on which a difference of opinion has always existed.

[187] This is the passage to which I referred in connection with the Yvain sources, p. 76. This son of King Claudas is, no doubt, the same who played such a valiant part in the war between Lancelot and his father, related at great length in the Lancelot.

[188] This arrival of the nine knights at the Grail Castle, and their share in the Grail revelation, is a striking proof of the unreality of the Galahad Queste quâ quest, on which I have remarked elsewhere. Who are these knights? What claim have they to be admitted to a feast so holy that even King Pelles and his son are excluded? Practically they are as much achievers of the Quest as Galahad himself. The fact is the writer is so taken up with the religious symbolism of the relic that in exaggerating and insisting on symbolic details he loses sight of the real point of his story. I very much doubt whether any one but the Grail Winner himself ought to reach, or was ever contemplated as reaching, the Grail Castle, much less be witness of the full explanation of the relic. To this it may be objected that Gawain reaches it; but Gawain was certainly at one time looked upon as the Grail Winner, and I believe it is only in this character that he ever found the castle. The accessibility of Corbenic is a very weak point of the Galahad Queste.

[189] I cannot agree with M. Gaston Paris's suggestion that this passage, which he takes as part of the Mort Artur, refers to an earlier Queste redaction. A Queste giving a full account of the fate of so many of the knights engaged would be of portentous length, and there is absolutely no sign of this Galahad Queste having existed in another form. I regard it as a summing up, by the author, of the general results of the expedition, a postscriptum which enabled him to have a final fling at his bête-noire Gawain. The addition of Baudemagus's name may have been his work, or that of a copyist, and designed to give point to his accusation. Whether the tradition that he should be killed by Gawain arose from this passage, or was incorporated in the Merlin from another source we cannot say. The Baudemagus tradition demands study. In the Merlin he is represented as but six years older than Gawain, whose dearest friend he is, but in the Charrette he appears as quite an old man, whose son, Meleagant, is the contemporary of Gawain and Lancelot; while in the prose Lancelot and Queste he appears as the devoted friend of the family of King Ban, sharing the adventures of these young knights on an equal footing. The whole presentment is hopelessly confused. The frequent reference to the Arthurian records, as kept in the 'almeryes' at Salisbury, appears to me to be a parallel case to the allusions in the Charlemagne Romances to the records at S. Denys. I suspect there is as much, or as little, truth in the one ascription as in the other.

[190] Cf. Studies, p. 214. Dr. Sommer uses as an argument for this the difference of spelling in the name of Corbenic, but this proves nothing. D. L. has at least four ways of spelling this word, and sometimes a variant occurs in the space of a few lines. The general character of the name is always preserved, and in MSS. that have been frequently copied, to say nothing of printed, the substitution of one letter for another is too frequent to call for remark.

[191] Dr. Wechssler in his Lancelot study announces solemnly, 'So viel aber steht für uns fest, dass Malorys Quelle für sein sechstes Buch nicht die Branche eines Cyklus, sondern ein selbständiges Originalwerk gewesen ist' (Gral-Lancelot, p. 35). But we now see it was beyond any doubt part of a cyclic work.