.F3 b/ It*-. ..^mhhJI,, ,,,J],',*;;3j;j; ••►•'••«•«-». ■• '•'•>• •m*'.„..,mT:; _.,;, .x.;r.-'/ LiL*..".'n:i'.'i"f*"""''*"i"'"'^ ,i«J7.7i..AJ{.i ..'.n^ ! "»» <• I ■ -.. - ►.. ,,, _L :! 'v*f"'-* .»-.< i«...»4 ...^ " •• *■ --•• . ■■^^^VJ^tRJ^V .. 'T-:'^f'>, Z, .-., ^■•-' -- '-■,x,':V ••i«-»4.„. .,>::: ..i!.-;,-:.;,: «„;:■.- • ...t^.l •^•<"t'«'. V.-. ......... ,.. A ;. '• ■ " '■'.• jrrr^' •»'' ^.. , -T", w' ;*•.*•' • • ■ -1 -. , . . . ,,,,^^.'''.1'"'"' ^^•2'^^^''^'4.,-l ::: ^^**^-^»s?u. .:.:;' • 4.^-,. .- V. /; ,:•..•;.,>-•-. •..;,;;;,.•;;;;• •-.. -.'^ -*.".-,;. I.';., '!.;,*:•..' i^:®i^5??p?« ;.SS h7'UO^ PRESENTED BY / l^llOy Ube mniverstti? ot dbtcaao FOUNDED BY JOHN D. ROCKEFELLER SIR PERCEVAL OF GALLES A STUDY OF THE SOURCES OF THE LEGEND A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF ARTS AND LITERATURE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (department of ENGLISH) BY REGINALD HARVEY GRIFFITH THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS Hgentg THE BAKER & TAYLOR COMPANY NEW YORK CAMBRIDGE UNIVERSITY PRESS LONDON AND EDINBURGH XTbe ^niversitp of dbicaao FOUNDED BY JOHN D. ROCKEFELLER SIR PERCEVAL OF GALLES A STUDY OF THE SOURCES OF THE LEGEND A DISSERTATION SUBMITTED TO THE FACULTY OF THE GRADUATE SCHOOL OF ARTS AND LITERATURE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY (department of ENGLISH) BY REGINALD HARVEY GRIFFITH THE UNIVERSITY OF CHICAGO PRESS CHICAGO, ILLINOIS Copyright 1911 By The Univbesity of Chicaoo All Eights Reserved Published March 1911 Gin OCT 13 19U Composed and Printed By The University of Chicago Press Chicago, Illinois, U.S.A. 0/ PREFACE In making this investigation, many obstacles besides the scanti- ness of time allowed by classroom duties have had to be overcome. The University of Texas library is not a large one, and, in the field immediately concerned, is weak. Access to needed books has been had only in summer vacations and in libraries a thousand miles and more away from Austin. The difficulty continues a very present one. In seeing the book through the press, I have not been able to verify references by a comparison with original authorities, but have had to rely upon my manuscript notes. To hope that no errors have crept in is unreasonable; but I trust the reader will find them few, and will believe that I have made a painstaking endeavor to avoid them. In seeking the origins of the Perceval tale, I have circum- scribed the interpretation of "origins.'^ It is the immediate ancestry, not the ultimate source, that has here been sought. I have made no inquiry into Old Irish literature in the expectation of pointing out its parallels to the Perceval tale, if such there be; nor any into folklore domains in the hope of tracing the tale or its elements to an origin in custom, myth, or rehgion. Finally, the Grail problem lies outside the limits of this investigation, since no allusion to the Grail occurs in the English poem which is taken as the point of departure. In several ways this study is incomplete, as perhaps any study of its kind must be. The number of tales discussed is large, for I have mentioned every tale I have found that appears to throw light on the origin of the tale of Perceval as it is told in Sir Perceval of Galles; but the collection makes no pretense to finality. There are doubtless many variants now unknown to me. If the reader will indicate any such, I shall feel much beholden to him. In especial, the tale which is studied in chapter III (the tale in which a despised youth avenges an insult to his king and relieves his relatives from the attacks of an army that, slain every day, is restored to life every night by a hag with a reviving cordial) is VI PREFACE intrinsically most interesting, and would surely repay investiga- tion. J. F. Campbell says his MSS contained variants. Still others are doubtless procurable. Any tales, too, that appear akin to the story of the secondary heroine, the lady whom Perceval kissed and so brought into reproach, will be welcome additions. The courtesies I have received from many people are remem- bered most kindly and with a lively sense of obligation. To Professor John M. Manly, of the University of Chicago, I owe a debt of gratitude for inducting me into the mystery and fascina- tion of mediaeval romance. The late Alfred Nutt, whose recent death seems a personal loss to me, was very kind when I ventured to seek him in his house of business some years ago. Miss Jessie L. Weston was cordially friendly when I had opportunity to dis- cuss Sir Perceval with her one summer. To the books and articles of the many students who have preceded me my indebtedness is writ large on every page. The authorities of Lincoln Cathedral, of the British Museum, of the Library of Congress, and of the university libraries at Yale, Harvard, Chicago, Wisconsin, and Texas I desire to thank heartily for their many favors. And to my colleagues and very good friends. Professor Callaway, Professor Campbell, and Dr. Law, my very best thanks are due for criticism and many another deed of kindness; all of them have ''read proof" for me; how can friendship do more? — unless it be to "read proof" twice, as Professor Campbell and Professor Callaway have both done. R. H. Griffith Austin, Texas, U.S.A. March 7, 19 11 CONTENTS PAGE Introduction . i Statement of the problem. — Materials from which evidence is to be adduced: a condensed bibliography. — ^Resume of opinions of some of the most important students. — ^Lines of investigation in the present study. Chapter I: The Hero's Forest Rearing . . .14 The four incidents to be considered. — Comparison of SP and C by siunmaries. — ^Evidence of other versions, by summaries. The father's marriage tournament, in SP and W. The father's death, in SP, PC, W, Pd, Card, Fool; some comments. The widow's flight, in SP, PC, W, Pd, Card, Fool. Boyish exploits, in SP, W, Pd, Card, Fool, Ty; coirmient; evidence of C. — ^Table. — Argument: C as the source; any other version as the source. — Conclusions. Chapter II: The Hero's Awkward Attempts to Follow Instructions 29 The four incidents to be considered; materials not sufficient for an argument; summary of SP. — ^Two divisions of two incidents each.— Table: SP, W, Card, Ty, Fool, C, Pd, PC— First, or Religious Instruction-Forest Knights portion; comment. — Second, or Advice-Tent adventure portion: Advice, summaries of SP, W, Card, Ty, C, Pd; Tent incident, summaries of SP, W, C, Pd; comment. — ^Tentative conclusion. Chapter III: The Red Knight- Witch-Uncle Story . 40 The five incidents to be considered. — Summaries of SP and C. — Their differences. — Incident of the arrival of the hero at court; summaries of Pd, Ty, Card, Fool; comment. — ^The Red Knight- Witch-Uncle story: summaries of SP, Pd{a), Pd{h), G, Red Sh, Red Sh Variants, Conall; table of incidents; table of particulars; summaries of another set of tales, arranged in four groups. — Comment and argument on the Red Knight- Witch-Uncle story; the Insult, its types; the Meeting with Relatives, the number of meetings, the Uncle, the Three Young Men, the two Women; the Witch; the Death of the Insulter. — Recapitulation. — The incor- poration of the story into the frame-tale. vii Vlll CONTENTS PAGE Chapter IV: The Relief of the Besieged Lady . .78 The six incidents to be considered. — Summaries of SP and C. — Their four notable differences. — "The Saracen Influence"; sum- maries of SP, W, Conall, Saudan Og, Pd(b) ; difficulties in the way of the argument. — Recapitulation; comment and argument on the "Saracen Influence." — ^A new problem stated. Chapter V: The Rescue of the Lady Falsely Accused 94 The nine incidents to be considered; two groups, the first of seven, the second of two incidents. — Summary of SP. — ^An incident- outline of SP, C, W, Pd. — The Tent Lady's history: differences between SP and C — ^The Snow Scene in C. — The Tent Lady- Giant story: summaries of Yv and LF {SP not repeated); table of incidents; table of particulars. — ^Threads that bind this story together: Tent Lord's suspicion; the Lady's ring; the Giant combat and Gawain's relative. — The evidence of W concerning this story. — SP and C compared again. — The hero's mother. — His wife. — ^The end of the tale. Conclusion . . 116 JR,esume of the foregoing chapters. — ^The test of S3mthesis. — A-Stage: summary of frame- tale; tales showing it. — ^B -Stage: summary of Red Knight-Witch-Uncle story; tales showing it; the process of absorption; summary of the resultant account. — C-Stage: simmiary of Tent Lady-Giant story; process of absorp- tion; smnmary of the resultant account. — D-Stage: cleavage; one branch subjected to Saracen Influence; development within this stage. — ^E-Stage: the Grafl story incorporated. — F-Stage: the Swan Ejiight story incorporated. — ^Diagram. — Geographical home of the sources. — ^A final word on the evidence adduced. — SP independent of C. — SP probably not the translation of any French poem. INTRODUCTION The problem to which the following pages address themselves concerns the origin of the mediaeval English poem Sir Perceval of Galles, whether or not it is the offspring of a romance composed in French by Crestien de Troyes and now commonly known as Perceval le Gallois, ou le Conte du Graal. The materials from which to draw evidence for an argument are a group of tales' gathered from widely separated places — from England, Wales, Scotland, and Ireland, from France, Germany, and Italy. They may be Hsted as follows: 1. SP. — Sir Perceval of Galles {ca. 1370) is a Middle-English metrical romance, preserved, with some imperfections and sHght irregularities, as 143 sixteen-Hne stanzas (2,288 lines), in a single MS, the Thornton MS of Lincoln Cathedral. It was printed by J. O. Halliwell in The Thornton Romances (pp. 1-87), for the Camden Society in 1844, and reprinted at the Kelmscott Press in 1895. Its dialect is Northwest Midland, its date about the middle of the fourteenth century (some of its phrases are quoted in Chaucer's "Sir Thopas"); its author is unknown, but its rhyme-scheme and plot-structure indicate that the coniposer was not without practice.^ 2. C. — Crestien's tale of Perceval, Le Conte du Graal {ca. 1175), is an uncompleted poem of about 9,300 lines in Old French. It usually appears as part of a mass of verse that grew up around it. This composite mass developed because of the desire of other poets to finish what Crestien left unfinished. No single book contains all of it. There are sixteen MSS, the longest of which stretched its meter to the length (impos- sible, let us hope, outside of an antique song) of more than 63,000 lines. A prose redaction was printed in Paris in 1530, and Potvin edited the larger part of the "poem" as Perceval le Gallois, ou le Conte du Graal, in six volumes, Mons, 1866-71. Besides Crestien, three other contributors are known by name, Wauchier (Gaucher, Gautier), Manessier, and Gerbert; but the limits ' I have unifonnly used the word "tale" to mean the whole account any one author gives of his hero; "story" to mean a group of incidents more closely bound to each other than to other incidents in the tale in which they stand — a circle within a circle, so to speak. * For working bibliography see A. H. Billings, "A Guide to the Middle English Metrical Romances," Yale Studies in English (igoi), 125 fif. 2 SIR PERCEVAL OF GALLES of the portions they contributed are uncertain. At least two writers prefixed introductions to Crestien's lines. One of these introductions and the portions by Crestien and Gerbert are the parts of the "poem" I have used most. 3. PC. — The second of the two introductions just mentioned is about 800 lines long; once thought by some scholars to be by Crestien, it is now considered the work of an anonymous contrib- utor, and is referred to as one of the "pseudo- Crestien" portions. It is preserved in two MSS, Mons and British Museum Add. 36,614; its substance appears in part in the prose redaction of 1530; and it is printed in full from MS Mons by Potvin. The first introduc- tion (Potvin, 1-484) may be referred to as Elucidation. 4. G. — Gerbert's '' Continuation" is preserved in two MSS, but it has not been printed. I have had to rely upon two resumes, one given by Potvin (Vol. VI) and the other by Miss Weston in The Library (magazine), January, 1904. Gerbert's 10,000 lines appear in the MSS between the parts by Wauchier and Manessier. If the reader will imagine Potvin's edition revised so as to place Gerbert's lines before Manessier's, he may gather from the appended table an idea of the various parts of the Conte. Author Lines Nature of Contents Assigned Date Anonvmous 1-484 485-1,282 1,283-10,601 10,602-34,934 34,93 S-ca. 45 ,000 ca. 45,000-fa. 63,000 "Elucidation"; ^ Grail's mystery and winners Death of Perceval's father; flight of his mother Perceval's deeds; Ga- wain's adventures Adventures of Perceval and others Ditto Ditto 1220-30 1220-30 I175 I 190-1200 1216-25 I2I0-''20 Anonymous Crestien de Troyes .... Wauchier, and Inter- polators Gerbert Manessier, and Inter- polators For discussions of these tales see the books mentioned on pp. 7 ff., infra, and the authorities to which they in turn refer. 5. W. — Parzival ( ?i 200-1 2 16), a Middle-High- German poem by Wolfram von Eschenbach, has been preserved in many MSS and edited by several scholars. I have used editions by K. Bartsch {Deutsche Classiker des Mittelalters, Leipzig, 1875-77), P. Piper {Deutsche National-Litteratur, Stuttgart, 1890-92), and K. Lach- INTRODUCTION 3 mann (4th ed., Berlin, 1879); translations by Hertz and by Botticher into modern German; and the translation by Miss Weston into English. The poem is arranged in sixteen books, averaging about 1,500 lines each. Books I-VI and XIV are the ones I have used most. My references are to Bartsch's edition. 6. Pd. — Peredur (?i 2 50-13 50), a Welsh prose tale in which the hero is Perceval under another name, is preserved in the Welsh Red Book of Hergest, dating from i3oo(?) to i35o(?). It was translated into English by Lady Charlotte Guest {The Mabinogion, 1838-49), and into French by J. Loth (in D. de Jubainville's Cours de Litt. Celtique, Vol. IV, Paris, 1889). Reprints of Lady Guest's text issued by D. Nutt, 1902, 1904, by Dent, 1906, and by other publishers have made Peredur the most easily accessible of all the versions of the Perceval tale. My references are by pages to Nutt's reprint (when no name is given) and to Loth's translation. I have not had opportunity to see The White Book Mabinogion: Welsh Tales and Romances Reproduced from the Peniarth MSS, edited by J. Gwen- ogvryn Evans, PwUheli, 1909. In his review of this volume {Folk Lore, June, 1910, pp. 237-46), Nutt comments upon Evans' Introduction. 7. Ty. — Tyolet (?i25o), a French lai preserved in a single MS, was printed by G. Paris in Romania^ VIII (1879). The 704 lines of the poem fall into two parts : {a) the early life of Tyolet and his coming to court (1-320); {h) the adventure of the White Stag, whereby Tyolet wins a wife (321-704). I have used all of the first part, and the concluding lines of the second. 8. Card. — Carduino (?i375), an Italian poem, was published from a unique MS by Rajna in 1873 {Poemetti Cavallereschi, Bologna). A portion of the poem is wanting in the middle. There remain two cantos, one of thirty-five eight-line stanzas, the other of seventy- two. My references are to stanzas. Card is the most primitive of its group of four tales; the others are Libeaus Desconus (LD), Bel Inconnu (BI), and Wigalois (Wig). For an excellent study of the group cf. W. H. Schofield, "Studies on the Libeaus Desconus," Harvard Studies and Notes, IV, 1895. 9. Yv. — Yvain (?ii65), by Crestien de Troyes, ed. by W. Foerster, Halle, 1887, and later years. References are to the edition of 1906. 4 SIR PERCEVAL OF GALLES 10. LF. — The Lady of the Fountain (?i25o), the Welsh version of the Iwain tale, is accessible in Lady Guest's Mabinogion (Nutt's reprint, pp. 167 ff.) and in Loth's translation (ref. as for Pd, supra), pp. I ff. On 9-10, see a valuable essay by A. C. L. Brown, "Iwain: A Study in the Origins of Arthurian Romance" {Harvard Sttidies and Notes, VIII, 1-147), 1903; and Foerster's comment on Brown's book, Yvain (ed. 1906), p. xlix. Besides the materials already mentioned there are some folk- tales still current that furnish evidence. These tales are told of different heroes, and no one of them relates more than a portion of the adventures attributed to Perceval. Often, indeed, it requires a comparative study to show that the adventures are akin. The citation of these tales, however, makes it possible for us to study the evolution of the Perceval tale. They are presented in three groups. THE SCOTCH GROUP 11. Fool. — Amadan Mor, or the Lay of the Great Fool.^ 12. Red Sh. — The Knight of the Red Shield. 13. Conall. — Conall Gulban.^ 14. Een. — How the Een Was Set Up.^ 1 1-14 are from J. F. Campbell, Popular Tales of the West Highlands (four vols., London, 1890-93): Fool, III, 160-93; Red Sh, II, 451-93; Conall, III, 199-297; Een, III, 348-60. 15. Manus. — A Tale of Young Manus. Maclnnes and Nutt, "Folk and Hero Tales of Argyllshire," Waifs and Strays of Celtic Tradition, II (1890), 338-75. ' There are other versions of the Lay, which may be spoken of as variants: var. a is O'Daly's, in Transactions of the Ossianic Soc, VI, 161-207; var. h, "Amadan Mor and the Gruagach of the Castle of Gold," in Curtin's Hero Tales of Ireland, 140-62; and var, c, "The Amadhan Mor," in Kennedy's Bardic Stories of Ireland (1871), 151-55. ' Conall var. a, "The Adventures of Conall Gulban," is in Kennedy's Bardic Stories of Ireland, 156-60; its variations do not help us. Dr. D. Hyde says (Beside the Fire [London, 1890], p. xxxii)' "On comparing [Campbell's Conall] with an Irish MS, by Father Manus O'Donnell, made in 1708, and another made about the beginning of this century, by Michael O'Longan, of Carricknavar, I was surprised to find incident following incident with wonderful regiilarity in both versions." * There are several versions of Fionn's youthful deeds, which only in part parallel those of Perceval. A second version is "The Boyish Exploits of Finn MacCmnhail" in Transactions of the Ossianic Soc. (Dublin, 1859). A third is "The Birth of Fin MacCvunhail"; see chap, iii, 64, infra. INTRODUCTION 5 1 6. Big Men. — Fin MacCoul in the Kingdom of the Big Men. 17. Ransom. — Fionn^s Ransom. 16-17 are in J. G. Campbell, "The Fians," Waifs and Strays of Celtic Tra- dition, IV, 175-91, 242-57. THE IRISH GROUP 18. Lonesome. — The King of Erin and the Queen of Lonesome Island. 19. Kil A. — Kil Arthur. 20. Fear Duhh. — Fin MacCumhail and the Fenians of Erin in the Castle of Fear Duhh. 18-20 are in J. Curtin, Myths and Folk-Lore of Ireland (Boston, 1890), 93-113,175-85,221-31. 21. Coldfeet. — Coldfeet and the Queen of Lonesome Island. 22. Lawn D. — Lawn Dyarrig, Son of the King of Erin, and the Green Knight of Terrible Valley . 23. Faolan. — Fin MacCool, Faolan, and the Mountain of Happi- ness. 21-23 are in J. Curtin, Hero-Tales of Ireland (Boston, 1894), 242-61, 262-82,484-513. 24. Mananaun. — King Mananaun. 25. Red Belt. — The Champion of the Red Belt. 24-25 are in W. Larminie, West Irish Folk Tales and Romances (London, 1893), 64-84, 85-105. 26. D^yerree. — The Well of D^yerree-in-Dowan. D. Hyde, Beside the Fire (London, 1890), 129-41. 27. Dough. — Amadan of the Dough. 28. Hookedy. — Hookedy-Crookedy . 27-28 are in S. MacManus, Donegal Fairy Tales (New York, 1900), 29-57, 95-133- 29. Golden Mines. — Queen of the Golden Mines. S. MacManus, In Chimney Corners (New York, 1899), 37-53. 6 SIR PERCEVAL OE GALLES THE WELSH GROUP 30. Kg of Eng. — King of England and His Three Sons. J. Jacobs, More English Fairy Tales (1894), 132-45. Although in a book of English tales, this tale, so a note implies, came from a gypsy in Wales. BRETON TALES The two tales of Morvan Lez Breiz, and Peronnik V Idiot I have not used: primarily, because they offer no help; Morvan offers only the battle against a black giant, the ^'More du Roi," which bears but the faintest likeness to a part of SP, and gives no help at all, and Peronnik is like SP in only two places — the beginning and the end — and only vaguely similar there; and secondarily because de la Villemarque's Morvan has been discredited, and Souvestre's Peronnik has been suspected of being not altogether a folk-tale, not altogether free, i.e., from "cooking." TEUTONIC TALES Nutt, "Mabinogion Studies," Folk Lore Record, V (1882), 1-32, compares Red Sh with the Faroese Hognilied, with parts of the Volsunga and the Thiarek Sagas, and with the Hilde legend (mentioned in Bartsch's Kudrun, pp. v-viii). I have not been able to get at the books for a proper study of the Hild story, and cannot tell whether it is akin to the "Red Knight- Witch-Uncle story" of chapter III, infra, or not. Two versions given in Magntisson and Morris' Three Northern Love Stories (London, 1901) give no evidence of kinship; the same statement holds for Saxo Gram- maticus; for the brief outlines in Bartsch's and in Symons' editions of Gudrun and in Schofield's translation (pp. 193-94) of S. Bugge's The Home of the Eddie Poems (London, 1899); and for the discus- sions in Paul's Grundriss (2d ed.), Ill, 711, in F. Panzer's Hilde- Gudrun (1901), and in F. E. Sandbach's Niehelungenlied and Gudrun in England and America (1903). For the tales that I have listed the source has been carefully stated by all the collectors except MacManus. INTRODUCTION 7 Other tales are referred to, but bibliographical information con- cerning them is given in notes. On the propriety of using these tales, see the note on p. 41, infra. In the seventy years since the matter began to be much discussed, almost every shade of opinion possible has been expressed concern- ing the relation of the English Sir Perceval to the French Conte du Graal. The English poem makes no mention of the Grail, yet, paradoxically, every scholar who has studied the origin of the Grail legend has been forced to consider the Sir Perceval. Digests of the body of the literature that has thus grown up are to be found in several places; e.g., in A. Nutt's Studies in the Legend of the Holy Grail (London, 1888); in E. Wechssler's Die Sage vom heiligen Gral in ihrer Entwicklung bis auf Richard Wagners Parsifal (Halle, 1898); and in Miss J. L. Weston's Legend of Sir Perceval (London, 1906-9). On the more restricted subject, the relation of the Eng- lish poem to the French, a good working resume is given by Miss A. H. Billings, in her ''Guide to the Middle EngHsh Metrical Romances," Yale Studies, IX (1901). It does not seem advisable to recapitulate here all the opinions scholars have expressed, but the leading ones, arranged in groups, may be stated. FIRST group: GERVINUS and GASTON PARIS 1 87 1. Gervinus, G. G. Geschichte der deutschen Dichtung. 3 vols. Leip- zig. Fiinfte vollig umgearbeitete Auflage. — i, 576-77: "Wir haben oben die bretagnischen Volkslieder von Morvan erwahnt, die von seiner einsamen Wald- erziehung erzahlen, und wie er seine Mutter, nach Ritterthaten diirstend, ver- lasst, die dann der Gram um ihn todtet. Ob diese einfache Sage zuerst an dem Namen Morvan oder an welchem anderen gehaftet habe, ist gleichgiiltig; gewiss ist sie der Kern und Rahmen der Sage, deren Held im 12. Jahrh. in walschen und romanischen Erzahlungen die Namen Peredur und Parzival fiihrt. In einer sehr volksthiimlichen Gestalt, die an jenem einfachen Kern am treuesten festhalt, ist die Sage in einem spaten, strophisch abgetheilten, burlesken Gedichte eines englischen Bankelsangers des 14. Jahrhs. erhalten, das einem alteren bretagnischen Lai nacherzahlt sein mag." 1883. Paris, Gaston. "Perceval et la legende du Saint-Graal," Bulletin de la Societe Historique et Cercle Saint-Simon, II (November, 1882, Paris): "Le conte de Perceval appartient a la tradition galloise, recueillie de la bouche 8 SIR PERCEVAL OF GALLES des conteurs et musiciens gallois par les jongleurs et trouveurs normands ou frangais apres la conquete de I'Angleterre. La forme la plus authentique de ce conte nous est sans doute representee par un poeme anglais du XIIP siecle, Sir Perzivell, dans lequel le graal ne joue encore aucun role Le Sir Perzivell s'appuie certainement sur un poeme anglo-normand perdu, et nous offre un specimen des romans biographiques qui forment la plus ancienne couche des romans frangais du cycle breton" (pp. 98-99). 1888. [Paris, Gaston.] Histoire liiieraire de la France, Ouvrage commence par des Religieux Benedictins, etc. (Paris), Vol. XXX: "L'editeur, M. Halli- well, le (SP) regardait tout simplement comme un abrege tres sommaire du Perceval de Chretien et des continuations de ce poeme. Une telle opinion n'est pas soutenable Le 'Sir Percevelle'remonte done a une autre source, et sans doute a un poeme anglo-normand (p. 259) La vraie place de 'Sir Percevelle' dans revolution du cycle toujours amplifie de Perceval a, au contraire, ete parfaitement discernee par un savant qui est un poete M. Wilhelm Hertz .... a montre que le poeme anglais nous represente, sous une forme assez voisine de I'original, quoique alteree, un des elements primordiaux qui sont entres dans la composition du conte gallois et du roman frans:ais. II faut ajouter, comme nous I'avons dit, que ce poeme repose tres probablement sur un poeme anglo-normand, derriere lequel on peut avec vraisemblance chercher un conte purement celtique" (p. 261). SECOND group: steinbach, nutt, and kolbing 1885. Steinbach, Paul. Ueber den einfluss des Crestien de Troies auf die altenglische literatur. Diss., Leipzig: "Dass diese annahme des beriihinten literar-historikers (Gervinus) betreffs der vorlage unseres gedichtes entschieden eine irrige zu nennen ist, wird die folgende untersuchung zeigen (p. 28) Bis vers 820 folgt der engl. dichter genau dem gauge der erzahlung des franzo- sichen gedichtes (hier bis vers 2400), mit nur wenigen und nicht bedeutenden abanderungen, .... dagegen manchen punkt weglassend und stark kiirzend. Von vers 821 findenwir .... Cr. mehr oder weniger freibenutzt (35) In einen, urspriinglich bretonischen iiberlieferungen entstammenden rahmen hat er [the English poet] in freier kiirzender bearbeitung, unter benutzung einiger vielleicht bei den in England wohnenden Bretonen vorgefundenen volkstiimlichen ziige, teils alteren, teils neueren ursprungs, und unter hinzu- fiigung einiger an die schilderung von kampfen in den Chansons de geste erinnernder partien, das Crestiensche werk 'Li contes del graal' bis ca. v. 6000 eingeschoben, indem er sich dabei im ersten telle seines gedichtes (bis v. 821) mehr, im letzteren weniger an dasselbe anlehnt und zugleich mit bemerkens- werter konsequenz jede beriihrung mit der gralsage vermeidet" (41). 1881. Nutt, Alfred. "The Aryan Expulsion-and-Return Formula in the Folk and Hero-Tales of the Celts," Folk-Lore Record, IV, 1-44: "Schulz's opinion that the English romance is a translation or a close imitation of a twelfth- INTRODUCTION 9 century Breton poem is probably correct. The romance represents at any rate an independent and, in many respects, older treatment of the subject than the Mabinogi" (ii). 1888. Nutt, Alfred. Studies in the Legend of the Holy Grail, etc., London: Nutt restates Steinbach's view, and adds, "The use of Chrestien by the author of Sir Perceval seems, however, uncontestable: and, such being the case, Stein- bach's views meet the difi&culties of the case fairly well" (150). 1891. Nutt, Alfred. "Les derniers travaUx allemands sur la legende du Saint Graal," Revue Celtique; same art., Folk Lore, II, Appendix: "Mais M. Golther a-t-il parfaitement raison? II n'expose nulle part sa these d'une fagon claire, mais je ne crois pas aller au dela de sa pensee en la formulant ainsi: Chrestien a le premier traite le sujet de la quete du Graal et de la lance qui saigne; tout ce qui a ete ecrit depuis releve de son roman inacheve et a ete ecrit dans le but de le completer; a la verite il avoue avoir puise a une source anteri- eure, mais cette source est entierement perdue et n'a eu aucune influence sur les autres ecrivains du cycle [Folk Lore, p. xxv] Je n'ai pu que me rencontrer avec des erudits distingues, en y reconnaissant des traits archaiques [ in SP], L'auteur, on le sait, laisse absoliunent de cote tout ce qui, chez Chrestien, se rapporte au Graal. La faute en est toujours, d'apres M. Golther, aux allures enigma tiques du poete frangais; dans le doute, le traducteur anglais s'est abstenu. Voila une reserve dont on trouverait difiicilement un second exemple chez les ecrivains du moyen age. Mais lui aussi a connu non seule- ment Mennecier, auquel, d'apres indication formelle de M. Golther, il a emprunte la fin de son roman, mais aussi Gerbert, auquel, ex hypothesi, il a d