^{page 337 note 1} Philip Ogden, A Comparative Study of the Poem Guillaume d’ Angleterre with a Dialectic Treatment of the Manuscripts, 1900. Dr. Ogden has unfortunately printed only a few excerpts from his dissertation, so that it is impossible with justice to criticize his study in detail. Though I have treated rather more stories than he does, I wish to express my obligations to his work for the suggestion of some material that I might otherwise have overlooked. It is only fair to say, however, that for most of the versions treated I am not indebted even to the indefinite hints found in Dr. Ogden's fragment, as my study had made considerable progress before I knew of his work. It is to be hoped that he will soon publish the results of his completed study.

^{page 339 note 1} See chap. v.

^{page 340 note 1} For general treatment and history, see Klebs, Apollonius aus Tyrus, 1899.

^{page 340 note 2} P. 297.

^{page 340 note 3} Ogden includes them and even (p. 28) connects them closely with the Spanish Cifar, because one of the twins in that romance is lost by straying away as is the Epidamnian Menaechmus. But see chap. vi.

^{page 341 note 1} See chap. vii.

^{page 341 note 2} See Kittredge, Arthur and Gorlagon, 1903, p. 241, note, for a partial list of studies.

^{page 342 note 1} P. 23.

^{page 342 note 2} Beiträge zur Crescentiasage, 1866.

^{page 342 note 3} The Constance Saga, 1902 (Palaestra xxiii).

^{page 344 note 1} Ogden refers to a story in the Sanskrit epic, Ramayana as the oldest version of the motive. It is in reality rather an analogue of The Calumniated Wife than of the Man Tried by Fate. The Ramayama is now generally regarded as later than the Mahabharata but perhaps existed in its primitive form as early as the fifth century b. c. (Monier-Williams, Indian Epic Poetry, p. 3; Macdonell, Sanskrit Literature, p. 309). The story in question is found in the seventh book, an addition to the original work that dates from the second century b. c. or later. In the translation by Manmatha Nath Dutt, 1891-94, the tale is found, pp. 1717 ff. It is summarized by Monier-Williams, p. 88. Rama banishes his wife Sita because of aspersions that have been cast on her chastity. He sends her to the hermitage of Valmiki, the poet, where she gives birth to twin sons. These boys are reared in the solitude and learn the Ramayana. When they grow up they wander to their father's court, where they recite the poem and by means of questions are recognized. Rama sends for Sita and on her arrival prays the goddess Earth to bear witness to her purity if she is innocent, whereupon the earth opens and receives her. The allusions to the poem in the story mark the seventh book, which is the last, as a work much later than the body of the epic. It is, however, earlier than any of the tales that belong to the cycle of the Man Tried by Fate. This does not make it advisable to include in our study a narrative essentially foreign to the cycle.

^{page 344 note 2} Hindoo Tales or, The Adventures of Ten Princes freely translated from the Sanscrit of the Dasakumaraeharitam, by P. W. Jacob, 1873, pp. 19-22, 51-56; Une tétrode on drame, hymne, roman et poème, trans. H. Fauche, 1862, ii, pp. 18-21, 41-44.

^{page 344 note 3} Macdonell, p. 332. Jacob, p. vi, gives its date as the eleventh century.

^{page 345 note 1} R. Spence Hardy, Manual of Budhism, 1860, pp. 116 ff.

^{page 346 note 1} It may be considerably earlier. See the following tale.

^{page 346 note 2} Tibetan Tales translated from the … Kah-Gyur by F. A. von Schiefner, translated from the German by W. R. S. Ralston, 1882, pp. 257 ff.

^{page 346 note 3} P. xxvi.

^{page 346 note 4} As the twenty-sixth tale. Translation of R. F. Burton, Supplemental Nights, vol. i, pp. 319 ff.; of Habicht, von der Hagen, and Schall (Breslau ed. Tausend und eine Nacht), vol. xiv, pp. 133 ff.

^{page 347 note 1} See, e. g., Burton, Supplemental Nights, I, p. 191, note.

^{page 347 note 2} Translation of Habicht, etc., x, pp. 115 ff.

^{page 348 note 1} Translation of Sir Wm. Ouseley, ed. W. A. Clouston, 18S3, pp. 46 ff.

^{page 348 note 2} P. 19.

^{page 348 note 3} See Clouston, p. xxxv.

^{page 349 note 1} Pp. xxxvi fl.

^{page 349 note 2} P. li.

^{page 349 note 3} Translation of Habicht, etc., x, pp. 166-179. Seventh of the Viziers.

^{page 349 note 4} Found also, it will be remembered, in Turki and Malay.

^{page 349 note 5} Clouston, pp. 87 ff.

^{page 350 note 1} I cite from the translation by Weil, Lewald's ed., 1834-41, iv, pp. 543. It is given more briefly in the Habicht translation, xi, pp. 214 ff.

^{page 350 note 2} In Jellinck's Beth. Hamidrasch, i, 72. I take the story from Joseph Perles, Zur rabbinischen Sprach- und Sagekunde, 1873, p. 58, with some additions from the account of it by Israel Lévi, “The Pious Israelite or Histoire d'un homme qui ne voulait pas jurer,” Revue des études juives, xi, p. 229. Perles gives the climax of the story somewhat indistinctly. Lévi's article also appears in his Trots contes juives, 1887, p. 25.

^{page 351 note 1} Perles says that the youngest son was carried away with his father, while the other sons were taken by sailors. This seems to be a fault in translation simply.

^{page 351 note 2} Levi, p. 228; d'Ancona, Poemetti popolari italiani, 1889, p. 421.

^{page 351 note 3} August von Haxthausen, Transcaucasia, English translation, 1854, p. 374.

^{page 351 note 4} See chap. iv.

^{page 352 note 1} History of the Forty Vezirs, written in Turkish by Sheykh Zada, done into English by E. J. W. Gibb, 1886, 13th Vezir's Story, pp. 151-161. The collection was first translated into a European language by P. de la Croix as L'Histoire de la Sultane de Perse et des Visirs, 1722, but this is incomplete. Our tale appears as the 9th Vizier's Story. From the French came the Turkish Tales, 1809, referred to by Dunlop, Hist, of Fiction, ed. 1842, ii, p. 352. Another incomplete translation was made by Belletete, Contes Turcs en langue turque, extraits du Roman intitule Les Quarante Vizirs, 1812, in which our story does not appear. In the later German translation by Behrnauer, Die vierzig Veziere oder weisen Meister, 1851, which is as complete as Gibb's, our tale appears as the 14th Vizier's Story.

^{page 352 note 2} See Gibb, pp. viii, ix.

^{page 353 note 1} In the version of de la Croix the story is varied in manner which would make it of the highest interest could we be assured that the text was tolerably accurate. The lady whom the prince meets in the pit bears two sons to him but is later placed by pirates on an island, where she becomes queen. To her and to her two children the prince is later reunited. I have not seen de la Croix's translation and take this information, as well as the earlier reference, from Gibb, p. 416 f.

^{page 354 note 1} The most detailed account of the versions yet attempted is that of Knust in Dos obras didácticas y dos leyendas, Madrid, 1878, pp. 107 ff. I have catalogued these, together with several discovered later in The North-English Homily Collection, 1902, pp. 49 ff. Wagner, The Sources of El Cavallero Cifar, 1903 (reprinted from Revue Hispanique, x), p. 13, note 1, mentions a Latin version which “follows almost literally the version of the Legenda Aurea,” in ms. 93, Bibl. Nac. of Madrid; and p. 14, note 4, an Old French version in ms. 9446, of the same library, previously alluded to by P. Meyer, Bull, de la societé des anciens textes français, iv, 57.

^{page 354 note 2} Sept., tom. vi, die 20, pp. 123 ff.

^{page 354 note 3} The Latin versions differ a good deal on this point. Jacobus a Voragine in Legenda Aurea, cap. clxi, ed. Graesse, p. 712, says, for example, that according to some accounts Christ spoke through the figure on the cross.

^{page 354 note 4} The form of the names is different in the several Latin versions.

^{page 355 note 1} In the Speculum Historiale by Vincent of Beauvais, x, cap. 58-61, and in the Leg. Aurea, ed. Graesse, p. 714, he is definitely offered the choice of trial in this life or sorrow in the next. I quote from the former: “Die ergo si modo temptationem vis accipere aut in fine vite.” Eustace chooses the former and prays for the virtue of patience.

^{page 356 note 1} Into the question as to whether the “Crestien” named in the beginning of the romance was the poet of Champagne I cannot enter. It is sufficient to say that the arguments of Eudolf Muller in his dissertation, Untersuchung über den Verfasser der altfranzösischcn Dichtung Wilhelm von England, 1891, and of Foerster in Christian von Troyes, Sämtliche erhaltene Werke, iv, pp. clxiv ff., have not yet been satisfactorily answered. Professor Foerster concludes (p. clxvii) that “der Wilhelm erst nach dem Cligés entstanden muss.”

^{page 356 note 2} Chroniques Anglo-Normandes, 1836-40, iii, pp. 39-172.

^{page 357 note 1} iv, pp. 253-360.

^{page 357 note 2} Pp. 173-211. Both poems were reprinted from Michel by Giles, Scriptores Rerum Gestarum, Willelmi Conquestoris, 1845, pp. 179-297.

^{page 357 note 3} Ed. by Knust, Dos obras didácticas y dos leyendas, pp. 171-247.

^{page 357 note 4} Printed by Knust, pp. 299-403.

^{page 357 note 5} So Foerster believes, as against Knust who thought that it came from a Latin original. Ogden (p. 27) says that it was taken from a hypothetical Anglo-Norman reworking of the original Guillaume, “although the resemblance in episode between the Estoria and Chronica would point to some connection between the two.” As this judgment rests upon some assumptions as to related stories that I shall discuss later (see chap, v), I need only state here that I treat the various forms of Guillaume as a group by themselves.

^{page 359 note 1} Wilhelm von Wenden ein Gedicht Ulrichs von Eschenbach, ed. W. Toischer, Prag, 1876, in Martin's Bibliothek der mittelhochdeutschen Litteratur in Boehmen, Bd. i.

^{page 359 note 2} P. xxxiii.

^{page 360 note 1} P. xxiii.

^{page 361 note 1} Die gute Frau, Gedicht des dreizehnten Jahrhunderts, ed. E. Sommer, Zts. f. deutsches Alterthum, ii, pp. 385-481. An analysis of it was made by F. Wolf, Ueber die neuesten Leistungen der Franzosen, 1833, p. 75.

^{page 361 note 2} Sommer, p. 385, Vogt in Paul's Grundriss, ii¹, 294.

^{page 362 note 1} Ed. H. Michelant, Historia del Cavallero Cifar, 1872 (Bibl. des litt. Vereins in Stuttgart, cxii).

^{page 363 note 1} Michclant refers to it as probably of the first part of the fourteenth century. Knust, Dos obras didácticas, p. 93, believes that it was composed early in that century. Wagner, Sources of El Cavallero Cifar, pp. 9 ff., comes to the same conclusion from a careful consideration of the prologue.

^{page 365 note 1} Ed. Halliwell, The Thornton Romances, 1844, pp. 88-120; G. Schleich, Sir Ysumbras. Eine engl. Romanze des 14 Jahrhunderts im Anschluss an die Vorarbeiten J. Zupitzas, 1901 (Palaestra, xv). Utterson, Select Pieces of Early Popular Poetry, 1825, i, pp. 77-112, reprinted Copland's version.

^{page 365 note 2} See Schleich, pp. 65 ff., where detailed descriptions are given.

^{page 365 note 3} So much it seems Safe to say. See Schleich, p. 98, Wilda, Über die örtliche Verbreitung der zwölfzeiligen Schweifrehmstrophe, 1887, p. 39, places it in “eine südliche Gegend Nord-Englands,” but his attribution is not justified.

^{page 365 note 4} Schleich, p. 88, note, says: “Für die Bestimmung der Zeit, in der unsere Romanze entstanden ist, weiss ich keine andere Handhabe als das Alter der ältesten Hs. -C-, die in die zweite Hälfte des vierzehnten Jahrhunderts gesetzt wird.” Since he shows that ms. C (Gonville and Caius College 175, Cambridge) stands in the fourth degree from the original, it is safe to assume that the work was composed at least as early as 1350.

^{page 367 note 1} Ed. J. J. Eschenburg, Denkmäler altdeutscher Dichtkunst, 1799, pp. 347-362. F. H. von der Hagen, Minnesinger, Pt. 4, p. 640, note, says that a better version than that printed by Eschenburg from a fifteenth century print is found in the Munich Meistersinger ms.

^{page 367 note 2} Eschenburg, p. 345, says that it belongs “gegen die Mitte und in die letzte Hälfte des funfzehnten Jahrhunderts.”

^{page 368 note 1} Early English Versions of the Gesta Romanorum, ed. S. J. H. Herrtage, E. E. T. S., xxxiii, ext. ser., 1879, pp. 87 ff.

^{page 368 note 2} Herrtage, p. xix.

^{page 368 note 3} Ed. Oesterley, pp. 444 ff.

^{page 369 note 1} Danmarks gamle Folkeviser, ii, pp. 605-607, from Eentzel's ms. 48. The ballad contains thirty-one couplets, with refrain.

^{page 369 note 2} I give a rough paraphrase of what Köhler, Zts. f. rom. Phil., iii, 276, calls a “leider zum Theil lückenhaften und nicht recht verständlichen Lied.”

^{page 370 note 1} Two versions are given by Simrock, Die geschichtlichen deutschen Sagen, no. 20, p. 46, that by Guido Görres (1805-1852), and no. 21, p. 47, another which is called simply a “Volkslied.” Gruntvig, Folkeviser, ii, 605, note, remarks the connection with Eustace.

^{page 371 note 1} Vol. i, pp. 179 ff.

^{page 371 note 2} Seemingly the word indicates the Romans, with Mani as a degenerate form.

^{page 371 note 3} Sung by Jeanne le Rolland at Pluzunet, in 1867. I cite from Luzel's translation.

^{page 371 note 4} Sung by Marie-Anne Lenoan, of the commune of Duault.

^{page 372 note 1} I use the following abbreviations for the narratives:

^{page 378 note 1} This recalls the horn and ring of the Horn saga, with the many parallel stories cited by Child, English and Scot. Pop. Ballads, i, 17, note. Add Will Stewart and John, Percy Folio Ms., iii, 226.

^{page 379 note 1} See chap. vii.

^{page 380 note 1} See chap. vii.

^{page 380 note 2} See chap. v.

^{page 380 note 3} See p. 344, note.

^{page 380 note 4} P. 23.

^{page 381 note 1} P. 346.

^{page 382 note 1} Supplemental Nights, i, p. 191, note.

^{page 382 note 2} See E. Kuhn, Barlaam und Joasaph, 1893, and the illuminating essay on the same subject by Gaston Paris, Poèmes et légendes du moyen-age, pp. 181 ff.

^{page 382 note 3} Burton, as above, and Clouston, Pop. Tales and Fictions, ii, p. 64.

^{page 382 note 4} See p. 354.

^{page 382 note 5} In De Imaginibus, Orat. iii, ed. Opera omnia, 1712, i, p. 372.

^{page 383 note 1} Poemetti popolari italiani, 1889, p. 421.

^{page 383 note 2} P. 26.

^{page 383 note 3} See p. 387, note.

^{page 384 note 1} See chap. vi.

^{page 384 note 2} The addition in Armenian that a genius appeared to the hero is simply a commonplace of Oriental folk-lore.

^{page 385 note 1} It seems better to delay the discussion of the theme to chap. vii.

^{page 386 note 1} See p. 382.

^{page 386 note 2} See p. 384, note.

^{page 386 note 3} See p. 351.

^{page 387 note 1} See p. 351.

^{page 387 note 2} In connection with Eustace, it must be remembered that Ælfric's translation of the Latin version, which ten Brink (Gesch. der engl. Litt., i, 127) dates “um das Jahr 996,” follows the version of the Acta Sanctorum with accuracy and fullness. See p. 383, above.

^{page 387 note 3} Zur rabbinischen Sprach- und Sagekunde, p. 64.

^{page 387 note 4} Revue des études juives, xi, 232. He also remarks that Eustace is at least an analogue of the tales.

^{page 388 note 1} A similar trait is found in Ysumbras and Gesta, but with very different detail. No connection is possible. Indeed, it is probable that the incidents in the two European versions are independent of one another. See also chap. v.

^{page 388 note 2} See pp. 348, 349.

^{page 388 note 3} There is a similar weakening of statement in Armenian.

^{page 388 note 4} The immediate source of the Forty Viziers is, of course, the Forty Morns and Eves.

^{page 390 note 1} Forty Viziers and Jewel-Merchant have independent changes at this point.

^{page 390 note 2} The Arabian follows Shah Bakht here, perhaps a case of “contamination.”

^{page 392 note 1} See p. 388, note.

^{page 393 note 1} I summarize from the French translation of J. C. Mardrus, vol. v, pp. 1-149. The tale is found in Habicht's translation, vol. v, pp. 3 ff. It has been mentioned in a similar connection as the present by the German translator; by d'Ancona in Scelta, p. xi, and Poemetti, p. 393 (for full titles see below); by Tettau, Über einige bis jetzt unbekannte Erfurter Drucke, 1870, pp. 309, 310; and by Köhler, Zts. f. rom. Phil., iii, p. 277.

^{page 394 note 1} I shall cite this tale as Kamaralzaman.

^{page 394 note 2} Summary given by Jonathan Scott, Bahar-Danush or Garden of Knowledge, 1799, vol. in, p. 277. I cite the tale as Bharam.

^{page 394 note 3} As noted by d’ Ancona, Scelta, p. xix, a similar story is found in the Bibliothèque des romans, aôut, 1777, pp. 51 2. in Les aventures du prince Abdulselam et de la princesse Chelnissa. It purports to be of Turkish origin and differs considerably from the other versions, though perhaps nearer Kamaralzaman than any other.

^{page 395 note 1} Folk-Tales of Kashmir, J. H. Knowles, 1888, pp. 306 ff.

^{page 395 note 2} This trait recalls Shah Bakht and the Armenian, where the hero is chosen king by an elephant and a white eagle respectively. It seems to be something of a commonplace of Oriental story and cannot be regarded as furnishing evidence of any schematic connection between the two motives. The influence of Shah Bakht in some form or other may, however, account for the appearance of the trait here.

^{page 395 note 3} I cite this tale as Kashmir.

^{page 395 note 4} From W. Radloff, Proben der Volkslitteratur der türkischen Stämme Süd-Sibiriens, 1872, iv, p. 482. It is mentioned by R. Köhler, Zts. f. ram. Phil., iii, 277. I cite the tale as Tartar.

^{page 396 note 1} Ed. by Michelant and Meyer, Soc. des anciens textes, 1894. Analyzed by Köhler, Germania, xvii, 63, and by Littré, Hist. litt. de la France, xxii, 807. The former compares it with Der Busant, as does Meyer in his Introduction. For textual comment see Mussafia, Zur Kritik und Interpretation romanischer Texte, Wiener Sitzungsberichte, Phil.-hist. Classe, 135, xiv, 1-72; 136, vii, 35-48.

^{page 396 note 2} “Il ne faut pas se le dissimuler, le poème fut peu répandu.” Meyer, Introd., p. xl.

^{page 396 note 3} Meyer, p. xxxv.

^{page 397 note 1} I cite the romance as Escoufle.

^{page 397 note 2} Ed. von der Hagen, Gesammtabenteuer, 1850, Pt. i, pp. 331-366. Summary given by Tettau, p. 301. Compared with Eseoufle by Kohler, Germania, xvii, 62; with Kamaralzaman and Pierre by Tettau, p. 310.

^{page 397 note 3} Tettau, p. 303.

^{page 398 note 1} I cite the tale as Busant.

^{page 398 note 2} Ed. A. d'Ancona, (a) in the series Scelta di curiosità litterarie inedite o rare, no. lxxxiii, 1867, (b) in Poemetti popolari italiani, 1889, pp. 393 ff.

^{page 398 note 3} D’ Ancona, Poemetti, p. 406. D'Ancona compares it with Kamaralzaman and Pierre.

^{page 399 note 1} I cite the romance as Ottinello.

^{page 399 note 2} I have used the edition published by Gamier.

^{page 399 note 3} G. Paris, Rom., xviii, 511.

^{page 399 note 4} See d'Ancona, Scelta, p. xxv; Meyer, L'Escoufle, p. xxx.

^{page 400 note 1} I cite the tale as Pierre. Paris, Rom., xviii, 511, refers to the romance Paris et Vienne, written by Pierre de la Sippade in 1432, “qui offre plus d'un rapport avec Pierre de Provence.” I have not been able to use this work. In the 1903 reissue of Political, Religious, and Love Poems, E.E.T.S., orig. ser. 15, Dr. F. J. Furnivall prints (pp. 293-300) a few fragments of a Peare of Provence and the fair Maguelone from a fifteenth century ms. belonging to Lord Clifden of Lanhydrock. We thus have bits of an English translation made in the same century as the original. A complete copy would be of greatest value, accordingly.

^{page 400 note 2} Published by G. Pitrè in his Fiabe novelle e racconti popolari siciliani, 1875, i, pp. 123-132. I owe the reference to Kohler, Zls. f. rom. Phil., iii, 277.

^{page 400 note 3} I cite the story as Mandruni. V. Imbriani in his notes in Pitrè, iv, pp. 374, 5, gives as variants of the Sicilian tale: (A) A story (no. xxii) in Porrelane of M. Sabadino Degli Arienti Bolognese, concerning the adventures of the son of a king of Portugal with the daughter of a king of England, ed. 1531. (B) Adventure of Sifanto in the seventeenth canto of Mondo Nuovo by Tomaso Stigliani da Matera, 1628. (C) Novella 56 in Duccento Novelle of Celio Malespini, 1609, concerning adventures of Orio and Pulicastra. As being late and literary these versions can safely be neglected. I have simply verified and adapted the references.

^{page 402 note 1} Work cited, pp. 309, 310.

^{page 402 note 2} He believed, of course, that Pierre was a work of the twelfth century.

^{page 402 note 3} Poemetti, pp. 410 ff.

^{page 402 note 4} Scelta, p. xlvi.

^{page 402 note 5} Poemetti, p. 410.

^{page 403 note 1} Introduction to L'Escoufle, pp. xxviii-xxx.

^{page 403 note 2} The view held by D'Ancona, Poemetti, p. 421.

^{page 403 note 3} M. Meyer notes, p. xxxii, that in the Reman de la Violette the ring of Euriaut is carried off by a tame swallow and is recovered by her lover. See ed. of F. Michel, 1834, p. 187.

^{page 403 note 4} Unfortunately I cannot vouch for the accuracy of the summary which I have used of this version.

^{page 404 note 1} It must be remembered that in Pierre the treasure motive is imbedded in a larger plot, though easily separable.

^{page 406 note 1} As far as they go, some versions like Hubert being fragmentary.

^{page 407 note 1} Armenian, see p. 386.

^{page 408 note 1} For example, by Gruntvig, Knust, Kohler, d'Ancona, Foerster, Sommer, and Tettau, as already cited, and by Steinbach, Uber den Einfluss des Crestien de Troies auf die altenglische Literatur, 1885, pp. 41 ff. It is necessary to emphasize the fact, because no one save Ogden has hitherto made a comparative examination of the entire group and because he denies the influence of Eustace altogether. He does not treat Sakarias, Breton, and Hubert.

^{page 408 note 2} Wilhelm is included in the group because of its close relationship to Guillaume, which is discussed below. The Breton, is included because of its marked similarity to Ysumbras, as is shown further on.

^{page 409 note 1} Work cited, pp. 309-318.

^{page 409 note 2} Poemetti, pp. 413-417.

^{page 409 note 3} P. 417.

^{page 410 note 1} Wilhelm, a close relative of Guillaume, does not carry the matter further than to state the fact of the payment made for the sons, which is proof that d'Ancona was right in regarding B as non-essential to the Man Tried, by Fate.

^{page 411 note 1} See above.

^{page 411 note 2} I believe that Gesta shows contamination from the Placidus of the Latin Gesta Romanorum, see below.

^{page 412 note 1} Tettau's opinion, see p. 409, above.

^{page 412 note 2} iv, p. clxx.

^{page 413 note 1} Breton also makes him a king, but the change is not unnatural in a folk-song and seems to have been made independently.

^{page 414 note 1} This change may be due to the influence of the treasure story, where the hero and heroine elope, or possibly to some obscure connection with the motive of The Calumniated Wife. For the analogous relations of the latter see the next chapter.

^{page 414 note 2} In Ysumbras there are three sons, and the ordor of loss is two sons, wife, last son.

^{page 414 note 3} Tettau, pp. 316, 317, traces this episode to the Arabian Nights and a Sanskrit tale.

^{page 415 note 1} Wilmotte in Le Moyen Age, ii, 189, remarks that the life of the youths in the forest “rappelle invinciblement un Episode de Tristan,” the youth of Tristan, I suppose. But such scenes are not uncommon in the romances.

^{page 416 note 1} In Gesta a nightingale interests him by its song, and the interpretation is given by an old man. The essential fact is, however, as stated above.

^{page 416 note 2} See p. 418.

^{page 416 note 3} It must be noted that the details are differently related in the two.

^{page 417 note 1} See p. 415, ante.

^{page 417 note 2} Breton places neither hero or heroine in a high station before the discovery.

^{page 418 note 1} Wagner, pp. 20 ff., comes to this conclusion and also studies the provenance of various traits in which Cifar differs from the legend. I shall therefore limit my discussion for the most part to the events which are paralleled in Eustace and its derivatives, referring the reader to Dr. Wagner's admirable monograph for interpolated material. Dr. Ogden, p. 28, connects Cifar with the Greek comedy Didymi, “which was taken into the Latin in the Menaechmi of Plautus.” He is led to this somewhat surprising conclusion “in view of the incident of the twins, and the manner in which one is lost by straying away in the city streets.”

^{page 418 note 2} See p. 416. Wagner notes the similarity in Ysumbras, but not in Wilhelm.

^{page 419 note 1} Wagner notes (p. 22) that the elder son's loss while his father sleeps beside a spring recalls Octavian. For this see the following chapter.

^{page 419 note 2} P. 23 f.

^{page 419 note 3} What appears to be a borrowing from the Crescentia saga in the circumstances of the discovery is discussed by Wagner, pp. 25 ff.

^{page 419 note 4} See p. 378.

^{page 420 note 1} P. 29.

^{page 420 note 2} iii, p. 907.

^{page 422 note 1} iv, clxxv. Ogden comes to much the same conclusion, p. 27.

^{page 423 note 1} See p. 337.

^{page 424 note 1} Varied in Ysumbras by the choice of poverty in youth or age.

^{page 424 note 2} See p. 385.

^{page 425 note 1} Papstfabeln des Mittelalters, 1863, p. 26, note. He also says that it is found “in einer deutschen Cölner-Chronik.”

^{page 425 note 2} Creizenach, Geschichte des neueren Dramas, i, p. 236, note, says that he has been unable to find any such statement in the edition of the Mirabilia Romae by Parthey, 1869. I cannot find it in that edition nor in the edition of Graesse, Beitrage zur Literatur und Sage, 1850, pp. 1-26.

^{page 425 note 3} Creizenach, i, p. 236.

^{page 425 note 4} Ed. Keller, Fastnachtspiele, pp. 900-955, no. 111. I cite it as Joan.

^{page 426 note 1} I think that it is unnecessary to assume with Creizenach that Schernberg introduced the episode himself. Dollinger must have had some basis for his statement. We may conclude that the legend of Joan, which did not rise till the thirteenth century, took over the episode in the fifteenth.

^{page 426 note 2} A. M. Tendlau, Fellmeiers Abende, Mahrchen und Geschichten aus grauer Vorzeit, 1856, no. 13, pp. 105-109. The same story appears in G. Levi, Parabeln, legenden, und gedanken aus Thalmud und Midrasch, 1863, pp. 153, 154, where it is ascribed to “Jalkut Ruth pag. 165 a.” In this form an Arabian proffers the wealth and for a period of six years. The outcome is the same. It would be better to substitute this Jewish variant for that given by Tendlau, but it was found too late.

^{page 427 note 1} See Benfey, Pantchatantra, i, pp. 495-499, Andrew Lang, Perrault's Popular Tales, 1887, pp. xlii-li, and J. Bédier, Les fabliaux, 1893, pp. 177-193. Benfey, p. 497, recognizes the alternative form apparently.

^{page 427 note 2} L. Gonzenbach, Sicilianische Märchen, 1870, no. 21, p. 130.

^{page 428 note 1} Ogden, p. 23. See pp. 342, 380, above.

^{page 428 note 2} See Kittredge, Arthur and Gorlaqon, in Harvard Studies and Notes, viii, 1903, pp. 240 ff.

^{page 428 note 3} P. 241.

^{page 428 note 4} Portuguese Folk-Tales, collected by Pedroso, translated by Henriqueta Monteiro, 1882, no. xxix, pp. 116 ff.

^{page 429 note 1} Asbjörsen and Moe, Norske Folkeeventyr, 1852, no. 8, pp. 42-47.

^{page 430 note 1} Gonzenbach, Sicilianische Märchen, no. 20, pp. 124 ff.

^{page 430 note 2} Kinder und Hausmärchen, 1856, iii, p. 7.

^{page 431 note 1} See W. A. Clouston, Popular Tales and Fictions, 1887, i, pp. 198-205, and Lang, Perrault's Popular Tales, pp. lx-lxiv, for brief studies of the theme.

^{page 431 note 2} Grimm, Kinder und Hausmärchen, i, p. 9.

^{page 432 note 1} Hartland in his study of The Forbidden Chamber in the Folk-Lore Journal, iii, p. 212, gives a variant of Grimm's tale from Leskien and Brugman, Litauische Volkslieder und Märchen, 1882, no. 44, p. 498. It follows the other closely except that the heroine enters the chamber and betrays herself by smearing her lips with blood from Christ's wounds—a transference presumably from the latter part of the tale as found elsewhere, e. g. in the Sicilian.

^{page 432 note 2} A. and A. Schott, Walachische Mährchen, 1845, pp. 90-96.

^{page 433 note 1} Fr. Schonwerth, Aus der Oberpfalz. Sitten und Sagen, 1857-59, iii, p. 317.

^{page 433 note 2} Haupt und Schmaler, Volkslieder der Wenden, 1841, (Anhang: Wendische Märehen, no. 16), ii, p. 179.

^{page 433 note 3} Two variants of the theme are given by Hartland, p. 213, from Waldau, Böhmisches Märchenbuch, 1860, p. 600, and Comparetti, Novelline popolari italiane, 1875, i, no. 38, p. 156. They add only the variation that the heroine is finally praised and rewarded for her obstinacy or constancy in denial, since by means of it the benefactor is released from a spell. This seems like a perversion of the story.

^{page 434 note 1} It is probable that the paper promised by Professor Kittredge, Arth. and Gorlaaon, p. 241, note 4, will discuss these matters.

^{page 434 note 2} See p. 427.

^{page 436 note 1} See p. 355, note.

^{page 436 note 2} Chap, iii, table vi.

^{page 436 note 3} See Darmesteter, De Floovante, 1877, Bangert, Beitrag zur Geschichte der Floaventsage, 1879.

^{page 436 note 4} Work cited.

^{page 436 note 5} Die Octavian-sage, 1884.

^{page 436 note 6} Vollmöller, Oetavian, 1883 (Foerster's Altfranz. Bibliothek, iii).

^{page 436 note 7} Sarrazin, Oetavian, Zwei mittelenglische Bearbeitungen der Sage, 1885 (Kölbing's Altengl. Bibliothek, iii).

^{page 437 note 1} ms. fr. 1452, Bibliothèque nationale, as yet unpublished. See Vollmöller, p. xvi.

^{page 437 note 2} Michel et Monmerqué, Theatre français au moyen-age, 1842, pp. 551-608.

^{page 437 note 3} See Rajna, Ricerhe intorno ai Reali di Francia, 1872, who gives its date as the second half of the fourteenth century (p. 323).

^{page 437 note 4} In Octavian proper, both French and English, the elder son is seized by an ape, rescued by a knight who is immediately set upon by robbers, and sold by the latter to a merchant from Paris; the younger son is seized first by a lioness and secondly by a griffin.

^{page 437 note 5} In Octavian proper the second son is found by the empress and carried by her to Jerusalem.

^{page 438 note 1} See chap, iii, table xi.

^{page 438 note 2} The transference of the boys from hand to hand recalls the sixth century Sanskrit tale (see chaps, ii, iii, iv). It is just possible that there may be some underground connection between the two. Since I have found no real evidence of this, however, I prefer the explanation given above, which is favored alike by chronology and geography.

^{page 438 note 3} The opinion of Streve, p. 3.

^{page 438 note 4} Sarrazin, Octavian, pp. xliv, xlv, compares Octavian with Ysumbras and points out a few literal resemblances. He believes that each “auf eine erweiterte fassung der Eustachiuslegende zurückgeht,” but says that “Der Isumbras scheint ein verdünnter, verwässerter aufguss des Octavian zu sein.” I believe that I have sufficiently discussed the matter above.

^{page 438 note 5} Danish: Namnlös och Valentin. En Medeltids Roman, ed. G. E. Klemming, 1846; Low German: Valentin und Namelos die niederdeutsche Dichtung, ed. W. Seelmann, 1884; and fragments of a Dutch version, ed. G. Kalff, in Middelnederlandsche epische Fragmenten, 1885-86.

^{page 439 note 1} See Seelmann, p. liii.

^{page 439 note 2} See Bangert, pp. 16, 17.

^{page 439 note 3} Ed. Halliwell, The Thorton Romances, 1844 (Camden Society), pp. 121-176.

^{page 439 note 4} Ed. E. Adam, E. E. T. S., extra ser. li.

^{page 439 note 5} Zielke, Untersuchungen zu Sir Eglamour of Artois, p. 48, says the beginning of the fourteenth century.

^{page 439 note 6} Adam, p. v.

^{page 440 note 1} Pp. xxx ff.

^{page 440 note 2} Schleich's edition of Sir Eglamour will probably supply this. See his Ymmbras, p. vi.

^{page 440 note 3} P. 60.

^{page 440 note 4} Pp. xxi f.

^{page 440 note 5} P. 16.

^{page 441 note 1} It is altogether probable that the source of the two had the double form as in Torrent, which was simplified in Eglamour.

^{page 441 note 2} Uggeri has the griffin, but again probably from the Italian form of the Octavian saga. See below.

^{page 441 note 3} Holland, Chrestien de Troies, 1854, believed that Octavian, Eglamour, and Torrent all derived from Eustace. His opinion has been held rather uncritically by succeeding scholars.

^{page 441 note 4} I cite from the summary by P. Rajna, Uggeri il Danese, Part iii, Romania, iv, pp. 401, 402.

^{page 442 note 1} Über die Sage von Ogier dem Dänen, 1891, p. 124.

^{page 442 note 2} Rajna, Romania, ii, p. 154.

^{page 442 note 3} Romania, iv, p. 423. Rajna also enumerates other points of likeness between Uggeri and Fioravante.

^{page 443 note 1} No edition of the romance has appeared, as far as I know. R. Ruths, Die französischen Fassungen des Roman de la belle Helaine, 1897, p. 57, gives a good summary of it, on which I base my notes.

^{page 443 note 2} Suchier, Les oeuvres poétiques de Beaumanoir, 1884, p. xxvii, and Ruths, p. 1, note, give the date as the thirteenth century; Paris, Litt. franç. au moyen age, 2nd ed., p. 254, places it in the early fourteenth.

^{page 443 note 3} Über die Sage von Offa und ryo, Paul and Braune's Beiträge, iv, pp. 500-521, and the introduction to Oeuvres de Beaumanoir.

^{page 443 note 4} The Constance Saga, 1902 (Palaestra, xxiii).

^{page 443 note 5} Conveniently reprinted in the Originals and Analogues, published by the Chaucer Society, 2nd ser., vii, pp. 73-84.

^{page 444 note 1} See the plan, Constance Saga, p. 13.

^{page 444 note 2} The opinion of Suchier, Oeuvres de Beaumanoir, p. xxix.

^{page 444 note 5} Novella, x, 1.

^{page 444 note 4} Oeuvres de Beaumanoir, pp. xxxi, xxxii.

^{page 444 note 5} Boeve de Haumtone, ed. Stimming, Bibl. Normanniea, 1899.

^{page 444 note 6} Ed. Kölbing, E. E. T. S., ext. ser., xlvi, xlviii, and lxv (from several mss. of which the Auchinleck is the earliest and has the date given). I disregard the versions of the saga other than the French and English, which contain the episode, since they are derivative.

^{page 445 note 1} The episode includes vv. 3543-3962 of the English version.

^{page 445 note 2} See pp. 413-415 for a discussion of this group.