else but mere transcripts from MH., but I cannot reproduce them all, and must limit myself to the most important ones:

MH.

3714. "nay," sayd the quene, "that wylle I not." 3853. hyt ys bot hevynesse of yower blode;

3879. A-gaynste hym openyd the gatys of hevyn; 3884. Syr lancelot eylythe no thynge but gode.

M.

855*, 13. Nay sayd the quene that shal I neuer do 858*, 21. It is but heuynesse of your blood.

859*, 6. & the yates of heuen opened ayenst hym 859*, 8. syr Launcelot ayleth no thynge but good.

The last part of the final chapter of book xxi. contains I think incidents of three different kinds; those invented by M., as Ector's praise of Lancelot and the enumeration of the knights who return to their own land; those which M. has in common with the Thornton MS. "La Morte Arthure," such as the succession to the throne of England by "Constantyn that was syr Cadores sone of Cornwayl"; lastly those M. must have borrowed from some French source we no longer possess, such as the statement, that Bors, Ector, Blamour, and Bleoberis undertake a pilgrimage to the Holy Land.

MALORY'S COMPILATION REPRESENTS A LOST "SUITE DE LANCELOT."

F, at the end of the critical examination of all that concerns Lancelot in Malory's rifacimento, we cast a retrospective glance at the results arrived at, and bear clearly in mind how Sir Thomas Malory dealt with the English metrical romances, or with the "Quest of the Holy Grail," or with the "Merlin," we can have little or no doubt as to one point: The differences, the altered sequence of incidents, and the additions revealed by a comparison of Malory with the Prose-Lancelot, cannot be attributed to the Englishman, but must have been present in his sources. Malory must thus have derived his information from one or several sources; these were nearly related to, or perhaps based upon, the VulgateLancelot, and already contained those features noticeable in "Le Morte Darthur" which are at variance with the Vulgate-Lancelot.

What hinders us from assuming only one source for Malory's account of Lancelot? Contradictions in various portions of the "Lancelot" section would demonstrate the use of several sources, but

none such have been revealed by our close examination of Malory's text. We may thus confidently advance the hypothesis that Malory knew but one source for his "Lancelot" section, and we may explain the existence of this source by analogy from the "Suite de Merlin and from the so-called enlarged "Tristan." Both these romances, the "Merlin" and the "Tristan," exist in a Vulgate and a modified version; is it not more than likely that such was the case with the most popular of all the Arthurian romances, that of "Lancelot of the Lake"? Just as the modified versions of the "Merlin" and the "Tristan were intended to replace the infinitely long Vulgateversions of these two romances, so it was with our hypothetical "Lancelot." Let us call this version, which we assume was used by Malory, the "Suite de Lancelot."

This "Suite de Lancelot " apparently met with no more success than the "Suite de Merlin," though both have the decided advantage over the Vulgate-versions of greater precision and of dealing more exclusively with their heroes.. Strange to say, both the "Suite de Merlin" and the "Suite de Lancelot" had well-nigh the same fate, both being rescued from oblivion by Sir Thomas Malory, whose compilation was the only authority for the "Suite de Merlin" until the discovery of the Huth MS. some fifteen years ago, and still continues to occupy this position for the "Suite de Lancelot." We may hope, however, that one of these days, either in France or in England, the confirmation of the hypothesis now urged will turn up.

Assuming, for argument's sake, the existence of such a "Suite de Lancelot," it remains for us to reconstruct its form from the materials furnished us by Sir Thomas Malory's rifacimento.

We must first ask ourselves, what portion of the Vulgate-Lancelot this "Suite" was intended to replace. If we glance on the table which faces page 177, and sets forth the relationship of the various MSS. of the "Lancelot " at the British Museum to the printed edition of Paris, 1513, and to "Le Morte Darthur," and if we follow the solid and dotted perpendiculars which traverse the large rectangle forming the frame of the table, we find that all the portions which Malory has in common with the "Lancelot " belong-save a fragment of part i.-to the second, third, and fourth parts. The "Suite de Lancelot" was thus evidently intended to replace a small portion of the first and the three last parts, or, to put the matter more concisely, the contents of vols. ii. and iii. of the edition printed at Paris in 1513.

we

To get an idea of the structure of the "Suite de Lancelot " have but to examine the various books of "Le Morte Darthur" dealing with Lancelot. These are :—

VOL. III.

8

I. Book vi. We have seen (supra, pp. 178-190) that chapters i._xi. are derived from the Vulgate-Lancelot, whereas chapters xii.-xviii. contain matter not to be identified therein.

attribute to the writer of the "Suite."

2. Book xi.

3. Book xii. chapters i.-x.

These we must, therefore,

4. Books xiii.-xvii., or "The Quest of the Holy Grail."

For these five books Malory must have had recourse to the Vulgateversion of the "Queste," &c., which is evident from his contradicting himself several times in the twelfth and thirteenth books, as I have shown, page 214. Whether the "Queste" was originally included in the "Suite de Lancelot," whether it was intercalated therein from the Vulgate, or whether, finally, it occurred there in a condensed form, is of course impossible to say; but I think the latter most likely, and that the writer of the "Suite" arranged the Vulgate-Queste in the same way as he did the remainder of the Vulgate-Lanceloti.e., he excluded all that did not concern Lancelot or his son Galahad, perhaps even the Boors section. Malory may have substituted the VulgateQueste for this curtailed account.

5. Books xviii.-xxi.

It must be left uncertain whether the "Suite de Lancelot " began with the events told at the beginning of Malory's sixth book, or whether other incidents preceded these events; it is, however, very probable, though by no means certain, that the "Suite" began with Arthur's expedition against the Roman emperor Lucius; at least the following facts are suggestive in this connection: Malory remarks, at the opening of book vi., that Lancelot is the first knight whom the "frensshe book" mentions after Arthur's return from Rome. The Roman expedition is found in the Vulgate-Lancelot, briefly told, but resembling Malory's account in its main features-e.g., the imperial embassy to Arthur, the Roman invasion of Burgundy, Gawayne's valour, Arthur's sending the emperor's body to the Senate at Rome, as the tribute claimed from him, &c.--but it occurs in the fourth part, after the fruitless siege of Gannes and Gawayne's being seriously wounded by Lancelot, where it is evidently entirely out of place. Further, by supposing that the "Suite" began with the Roman expedition, we have at once a source for the first part of the English metrical romance "La Morte Arthure" (Thornton MS.)' which Malory copies so servilely in his fifth book, and which is devoted to Arthur's Roman war.

1 Branscheid has succeeded in pointing out (Anglia, viii. pp. 179–236) various points which this romance has in common with the old English Chronicles, but I do not think it at all likely that the poet took the trouble to combine his information from so many sources, but rather that he had only one.

The fact that in the two last books, where Malory, besides his French source, had a doublet in the English metrical romance "Le Mort Arthur” (Harl. MS. 2252) he freely used, as we have seen, its very words and phrases, also points in this direction.

From the beginning of book vi.-leaving aside book vii., which is most probably a prose-rendering of a lost French romance of "Syr Gareth"; books viii.-x., which are devoted to the life and adventures of "Syr Trystram;" and books xiii.-xvii., which faithfully reproduce the Vulgate "Queste del Saint Graal"-to the end of book xxi., what Malory relates of Lancelot may reasonably be looked upon as parts of one complete whole, a "Suite de Lancelot."

The existence of such a "Suite de Lancelot " at once explains the numerous difficulties which Malory's text affords to the critic. The last portion (11, 1318–3969) of the English metrical romance "Le Mort Arthur " (Harl. MS. 2252) represents the conclusion of the "Suite de Lancelot," and thus we have, besides the Prose-Lancelot, which the writer of this romance used for his first part (ll. 1–1181), a second source, which, as we have seen, contradicts the Prose-Lancelot in various points, and which the poet used for his second part (11. 1318– 3969).

As far as Malory's account of Lancelot is concerned, we attribute all modifications, omissions, and additions to the writer of the "Suite de Lancelot," and we can further explain the following points:

Malory, possessing only a "Suite de Lancelot," lacked, was even perhaps altogether ignorant of, the first part of the Vulgate-Lancelot ; hence the extraordinary fact that he says nothing of Lancelot's birth and early life, nor even of his arrival at Arthur's court, but abruptly introduces him with a few vague phrases.

Finding Guenever's abduction by Meleagaunt of interest, the writer of the "Suite" embodied this episode, which occurs in the first part of the Vulgate-Lancelot at a later stage, and using partly the account found in the Vulgate-Lancelot, partly some other French, or perhaps even Welsh, source, gave it the form it assumes in Malory's book xix., and then, to avoid the immediate sequence of two such similar episodes as that of Guenever, Mador, and Lancelot, and that of Guenever, Meleagaunt, and Lancelot, altered the sequence of incidents in book xviii. Thus is confirmed' the theory of M. Gaston Paris concerning Malory's

1 M. Gaston Paris arrived at the result that Malory must have derived his information for his book xix. from two different sources-viz., the first part from some lost French or Welsh poem, the second from the Prose-Lancelot. In the above-named chapter, I have shown the correctness of M. G. Paris' theory, but I have added: The contamination of the two sources may be attributed either to Malory or to the writer of his source, and the latter seems to me the more probable. Considering now the nineteenth book

book xix., which I have reproduced and completed in the chapter headed "Le Roman de la Charrette and the Prose-Lancelot " (supra, pp. 232-249).

The writer of the "Suite" further added the three last chapters of book xix., in which the superiority of Lancelot over all the other knights of the Round Table is shown by his alone being able to heal the wounds of "Syr Vrre." He here shows great ingenuity in presenting his hero Lancelot at the climax of his fame and glory, previous to his disgrace and distress.

All the additions in the eighteenth, twentieth, and twenty-first books, such as the greater detail in all that concerns Lancelot and the enumeration of the various parts of his kingdom which he distributes among his faithful companions (book xx. chapter xviii.), were added by the writer of the "Suite" with a view to justifying the title of his romance more than in the case of the Vulgate-Lancelot, in which very much of the narrative has nothing whatever to do with Lancelot.

66

When I expressed the hope in the foregoing pages that some day a French MS. might be discovered confirming my theory respecting the lost "Suite de Lancelot," I little thought that I was myself destined to discover a part at least of such a MS. While studying the romance of "Tristan" for the third section of my Studies on Syr Thomas Malory's Le Morte Darthur," I was led to pay special attention to the British Museum MS. Add. 5474, which contains the so-called enlarged "Tristan," wrongly attributed to Hélie de Boron. Though Malory has not used this version, but only the Vulgate of the "Tristan," for his account of the life and adventures of Syr Trystram, I undertook the drudgery of reading through the entire MS., in the hope of finding some clue to the three last chapters of Malory's book xii., which are not to be traced in the Vulgate "Tristan." And richly were my endeavours rewarded! Besides other interesting points,' I found that this MS. (Add. 5474) contains, from fol. 144 recto, col. i. line 2 from the bottom, to fol. 162 recto, col. ii. line 19

in its relation to the other books, and Malory's capacities as a writer, I decidedly transfer to the writer of the "Suite de Lancelot" the part provisionally attributed by M. G. Paris to Malory, and assume that Malory found his information for book xix. ready in his French source-i.e., the "Suite de Lancelot."

Considering, further, the fact that M. G. Paris evidently proved that the poem of Chrétien, "Le Roman de la Charrette," is the direct source of the corresponding section in the Prose-Lancelot, that Robert de Boron's poem is the source of the Prose-Merlin, and that the same is the case as regards the "Roman de Tristan," I pronounce decidedly in favour of the theory according to which the poetical- preceded the prose

versions.

1 Discussed later on in the section on the "Tristan," p. 284, &c.