¹ Facsimile and Text of the Book of Taliesin, ed. J. G. Evans (Llanbedrog, 1910), p. 54, 1. 16–p. 56, l. 13.

² S. Turner, History of the Anglo-Saxons, ed. 5 (London, 1828), iii, 617.

³ Ibid., p. 634.

⁴ Ibid., p. 636.

⁵ Nennius, Historia Brittonum, ed. F. Lot (Paris, 1934), pp. 130, 201. On Taliesin cf. Cymmrodor, xxviii (1918).

⁶ Zts. f. Celt. Phil, ii (1899), 127 f.

⁷ Poems of Taliesin, ed. J. G. Evans (Llanbedrog, 1915), pp. 126–129. E. Davies, Mythology of the Druids (London, 1809), 513–526.

⁸ J. Rhys, Hlbbert Lectures (1887), pp. 263. f.; Studies in the Arthurian Legend (Oxford, 1891), 300 f., 307 f.; Celtic Folklore, Welsh and Manx (Oxford, 1901), ii, 679.

⁹ Pedeir Keinc y Mabinogi, ed. I. Williams (Cardiff, 1930), p. liii. Rev. Celt., xxxiii, 460 f.

¹⁰ R. S. Loomis, Celtic Myth and Arthurian Romance (N.Y., 1927), 91 f., 201–204, 211–213, 330. Speculum, viii (1933), 428 f.

¹¹ Perlesvaus, ed. W. A. Nitze and others, ii (Chicago, 1937), 154 f.

¹² H. Newstead, Bran the Blessed in Arthurian Romance (New York, 1939), pp. 21, 25 f., 68, 91, 148, 150.

¹³ T. Stephens, Literature of the Kymry (Llandovery, 1849), 192–204.

¹⁴ In W. F. Skene, Four Ancient Books of Wales (Edinburgh, 1868), i, 264–266.

¹⁵ Malory, Morte d'Arthur, ed. J. Rhys, Everyman Library, i, xxii–xxiv.

¹⁶ The Welsh text may be found in Skene's Four Ancient Boohs, ii, 181 f., or in works cited in notes 1 and 15.

¹⁷ pendernc gwlat ri: Lloyd-Jones suggests, “ruler of the kingly domain,” or “sovereign of a country's king.” Williams translates, “the Supreme, the King-Ruler.”

¹⁸ Williams and Jackson both note that py must be scribal error for ry.

¹⁹ kyweir: cf. discussion infra. Williams translates “ready.”

²⁰ kywirwas ae ketwi. Lloyd-Jones suggests as alternative, “it was a faithful youth that it guarded.”

²¹ preideu. Williams prefers “herds of cattle,” refers to its use in this sense in his Pedeir Keinc, p. 231, and suggests that Gweir had been turned into a cowherd.

²² parahawt yn bardwedi. Williams translates, “our bardic prayer will last.” Williams and Lloyd-Jones agree that wedi represents gweddi, “prayer.”

²³ cerd ochlywir. Lloyd-Jones writes: “Though this may stand for cerdd o chlywir, it is much more likely that ochlywir is connected with gochlywet… The rhyme is -it, and one should perhaps read cerdd gochlywit, ‘song was heard.‘ ”

²⁴ pedyr yckwelyt. Both Lloyd-Jones and Williams refer to a passage in the Red Book of Hergest, pedryfan dwfyn pedrychwelit. Williams translates this as “four-cornered is the world, four-sided.” Lloyd-Jones sees in pedyrychwelyt a verbal form, 3rd sing. pres. ind. or imperative or impf, or aor, impersonal, and suggests translating, “it was borne.”

²⁵ yg. All authorities think it probable that yg is the common contraction for vyg, “my.'

²⁶ kynneir. Lloyd-Jones suggests “praise.”

²⁷ o. Williams prefers translating “of.”

²⁸ gwrym. Williams and Jackson believe that the word is gwrm or gwrwm, a noun meaning “dark blue”; Williams suggests a reference to enamel work. Gruffydd sees no objection to gwrym, meaning “ridge.” Lloyd-Jones holds both meanings possible.

²⁹ ny rytyghit. Lloyd-Jones suggests the alternative translation “it would not be forsworn.” Jackson and Gruffydd favor the translation given. Williams suggests that the -t-is an infixed pronoun, the verb is ynghit (from wnc, “near”), and the ry- expresses possibility. The meaning would then be, “he cannot approach it.”

³⁰ lluch lleawc. Lloyd-Jones and Williams take these to be two adjectives, “shining, flashing,” and “destructive, death-dealing,” and ignore the connections pointed out in the commentary below. Gruffydd says: “We seem to have here 1) lleu lleawc (vide Kulhwch ac Olwen) 2) llew (flaw) lleminawc.”

³¹ idaw rydyrchit. Lloyd-Jones and Williams agree that “was lifted to it” is impossible. The former suggests, “appears to him,” the latter, “was sought from it.” Neither seems very plausible.

³² yn llaw leminawc yd edewit. Williams compares llaw lleminawc with Irish lugleimnech, “making short jumps,” and translates the whole passage, “at a quick pace he departs from it.”

³³ trafferth lethrit. Gruffydd prefers “mighty in prowess,” Williams, “famed was the disaster,” Lloyd-Jones suggests, besides the translation given, “glorious in difficulty.”

³⁴ Vedwit. Lloyd-Jones writes, “‘Carousal’ is as good a suggestion as any other that might be offered, although meddwdawd and meddweint are the usual forms.” Williams refers to a little known saint, Meddwid. Baring-Gould and Fisher, Lives of the British Saints, iii, 458. Cf. discussion below.

³⁵ kerd glywanawr. Williams writes: “Read clywanor to rhyme. Cf. Canu Aneirin, ll. 298–299, ‘song is heard.‘” Lloyd-Jones offers three possibilities, “song is the ruler of honour,” “song in praise of a leader,” “a song that will be heard.”

³⁶ ynys pybyrdor. Gruffydd suggests that this is in apposition to the subject of wyf, and means “the mighty lord of the isle.” Lloyd-Jones suggests that it is in apposition to kaer pedryfan, and means “stout defense of the isle.” Cf. discussion below.

³⁷ echwyd amuchyd. Lloyd-Jones and Jackson admit the translation given. Lloyd-Jones prefers, “by noon and by night there will be mixed bright wine….” Williams denies that eckwyd riming with muchyd can be ech$wUydd, “mid-day”; thinks that dwfyr, “water,” has been omitted after pybyrdor; suggests the translation, “fresh (or drinking) water is mixed with jet,” a reference to their hardships on the voyage. “Then comes the contrast: but before their hosts they were wont to drink sparkling or clear wine.” To this interpretation one might object that probably not the voyagers but the dwellers in the kaer drank the wine supplied by the spring. Cf. discussion below.

³⁸ rigor. Williams, Gruffydd, and Jackson challenge the rendering “king-dwarf.” Lloyd-Jones finds it difficult “since this would imply the rhyme -orr instead of -or (the usual rhyme in the stanza).” Williams asserts that rigor represents the mutation of Grigor, “Gregory,” but offers no explanation of the name. Emending rigor to rigorn seems possible since the poet admits another imperfect rime (gosgord) in this stanza. Either “king-dwarf” or “royal horn” would fit admirably, since we know from other sources that Annwn contained a pigmy king and the magic horn of King Bran. Cf. discussion below.

³⁹ Ny obrynafi lawyr llen llywyadur. Williams and Lloyd-Jones interpret lawyr as a compound of llaw and gwyr, “mean men, common folk.” Williams substitutes llyadur for llywyadur, and translates the line, “I set no value on book-reading folk.” It marks “the contempt of the bard for the llaw-wyr, … who read their mss., but know nothing of Arthur's feats in Faeryland.” Jackson questions translating obrynafi as “I set value on.”

⁴⁰ golud. Cf. discussion below. Williams writes: “After caer soft mutation occurs so that the original must be taken to be Coludd. I know of no instance of coludd with any other meaning than ‘entrails, bowels.’ … Cf. the use of perfedd for ‘middle’ and ‘bowels.‘ Caer Goludd, the Middle Fort, or the Fortress in the Middle of the Earth, the Mediterranean Fort!”

⁴¹ Cymmrodor, xxviii, 238.

⁴² W. Stokes, E. Windisch, Irische Texte, iii (1891), 211–216, T. P. Cross, C. H. Slover, Ancient Irish Tales (New York, 1936), pp. 504 f.

⁴³ S. H. O'Grady, Silva Gadelica (London, 1892), ii, 199 f. Stokes, Windisch, op. cit., iv (1900), 107 f.

⁴⁴ Stokes, Windisch, op. cit., i (1880), 214, l. 20; 215, ll. 6, 11; 217, l. 6; 218, l. 26; 219, l. 17; especially 227, and translation of same in A. H. Leahy, Heroic Romances of Ireland (London, 1905), I, 184.

⁴⁵ Cross, Slover, pp. 488–90. Zts. f. Cell. Phil., xvii (1927), 195–205.

⁴⁶ R. S. Loomis, Celtic Myth, pp. 49 f. PMLA, xlvii (1932), 320–323.

⁴⁷ Cockayne, Leechdoms, Wortcunning, and Starcraft in Early England (Rolls Series), iii, 233.

⁴⁸ J. Hertel, Die Himmelstore im Veda und im Awesta (Leipzig, 1924), pp. 41, 44.

⁴⁹ Revue des traditions populaires, x (1895), 571. Cf. ibid., xvi, 119 f.; F. M. Luzel, Contes populaires de Basse-Bretagne (1887), i, 6, 17, 33, 58; Rev. Celt., ii (1873–75), 289 ff.

⁵⁰ Ulster Journal of Archaeology, 1860, p. 111; Royal Soc. of Antiquaries of Ireland, Journal, xxix, 5; xl, 288; xli, 46.

⁵¹ For example in O'Grady, Silva Gadelica, ii, 290 f., and P. W. Joyce, Old Celtic Romances (Dublin, 1920), p. 252 ff.

⁵² Joyce, op. cit., pp. 143 f., 157 f., 164 ff. A. O'Kelleher, G. Schoepperle, Life of Colum Cille (Chicago, 1918), p. 389 ff. See specially A. C. L. Brown in Mod. Phil., xiv (1916), 67 ff.

⁵³ Pedeir Keinc, ed. I. Williams, pp. 99–101. Folklore, xviii (1907), 146–151.

⁵⁴ Pedeir Keinc, p. 55 f. J. Loth, Mabinogion (Paris, 1913), i, 159 f.

⁵⁵ Chrétien de Troyes, Conte del Graal, ed. A. Hilka (Halle, 1932), p. 136.

⁵⁶ Pedeir Keine, p. 4. Loth, Mabinogion, i, 87.

⁵⁷ Two ms. texts survive and are printed in Y Great (London, 1805), pp. 339–341 and in Baring-Gould and Fisher, Lives of the British Saints (London, 1913), iv, 377. The translation is made from the former. Cf. Charlotte Guest, Mabinogion (London, 1849), ii, 325. For a somewhat similar story cf. Minor Poems of the Vernon MS., ed. F. J. Furnivall, pt. ii, EETS, vol. 117 (1901), 484 ff.; A. B. Van Os, Religious Visions (1932), 173.

⁵⁸ Cf. however, the faery castle in Thomas of Erceldoune, ed. J. A. H. Murray (London, 1875), p. 12 f., and the Grail castle in Didot Perceval (J. L. Weston, Legend of Sir Perceval, ii, 57).

⁵⁹ On Gwynn cf. Guest, Mabinogion (1849), ii, 323 ff.; Black Book of Carmarthen, ed. J. G. Evans (Pwllheli, 1906), p. x–xv.

⁶⁰ Proceedings of the Roy. Irish Acad., xxxiv (1917–19), c, 145 ff.

⁶¹ Ibid., p. 147.

⁶² Very clearly the elements of the equation appear in Fiacc's Hymn (c. 800), which asserts that the Irish used to worship the peoples of the síd (tuatha síde), and in the conclusion to Serglige Conculaind (A version, eleventh century), where we read that the power of the demons was great before the advent of the Faith and they were called the folk of the síd (aes síde). Stokes, Windisch, Irische Texte, i (1880), p. 14, l. 41; p. 227. A. H. Leahy, Heroic Romances of Ireland (London, 1905), i, 184.

⁶³ Loth, Mabinogion, i, 314 n., 315. J. G. Evans, While Book Mabinogion, col. 484.

⁶⁴ Facsimile and Text of Bk. of Taliesin, ed. Evans, p. 20, ll. 8 f.

⁶⁵ Pedeir Keinc, ed. Williams, p. 100.

⁶⁶ Two recent efforts to interpret the Celtic Other-World as a land of the dead are those of Prof. A. C. L. Brown in Speculum xv (1940), 3, and Dr. A. H. Krappe in Rev. celt., xlviii (1931), 94. Neither seems to offer satisfactory proof from texts of the period and the peoples in question. The latter admits, moreover, on pp. 108 f. that there are no dead, “jamais des morts,” in his land of the dead, and that is a fatal admission.

⁶⁷ E. S. Hartland, Science of Folklore, pp. 40–44.

⁶⁸ Cf. the famous story of King Edwin's council in 627 (Bede, Ecclesiastical History, Bk. ii, ch. 13) and the Anglo-Saxon elegies of The Ruin and The Wanderer.

⁶⁹ K. Meyer in Sitzungsberichte der Preussischen Akademie der Wissenschaften, 1919, p. 537 ff. Cf. also on Donn, Proceedings of Roy. Irish Acad., xxxiv, c. 163–165.

⁷⁰ Anthropos, xxvi (1931), 438.

⁷¹ Folklore, xviii (1907), 121 ff., 445 ff. Hastings, Mythology of All Races, iii (Boston, 1918), p. 123. Hastings, Encyclopaedia of Religion and Ethics (1909), ii, 689–696. J. A. MacCulloch, Religion of the Ancient Celts (Edinburgh, 1911), pp. 370–376. L. Gougaud, Les Chrétientés celtiques (Paris, 1911), p. 25. Romanic Rev., iv (1913), 178 ff. T. P. Cross, W.A. Nitze, Lancelot and Guenevere (Chicago, 1930), p. 71. K. Meyer, loc. cit., p. 545.

⁷² A. Le Braz, La Légende de la mort, ed. 3 (Paris, 1912), i, p. xxiii–xxv.

⁷³ S. H. O'Grady, Silva Gadelica, ii, 202 f. Stokes, Windisch, Irische Texte, iv (1900), 111.

⁷⁴ O'Grady, op. cit., ii, 111. Stokes, Windisch, op. cit., iii, 11–13. Cf. p. 13, “dochuatar ar cúla chúm Tuaithe dé Danann tréna cleamhnus.”

⁷⁵ The idea is exquisitely expressed in the Odyssey, iv, ll. 561 ff. “‘But it is not thy destiny, O Menelaus, child of Zeus, to die in horse-pasturing Argos. The immortal gods will send thee to the Elysian plain and the verge of the world, where fair-haired Rhadamanthus dwells, where life is easiest for man. No snow falls there, nor any violent storm, nor rain at any time; but Ocean ever sends forth the clear, shrill blast of the West wind to refresh mankind; because thou hast Helen to wife and they count thee to be son-in-law to Zeus’ ”

⁷⁶ Cross, Slover, Ancient Irish Tales, pp. 488–490, 248–253. O'Grady, Silva Gadelica ii, 290 f.

⁷⁷ Perlesvaus, ed. Nitze and others, I, ll. 38, 1081, 1646, etc.

⁷⁸ P. Piper, Höfisches Epik (Stuttgart), ii, 181.

⁷⁹ Heinrich von dem Türlin, Krone, ed. Scholl (Stuttgart, 1852), 11. 29532–34.

⁸⁰ J. A. MacCulloch, Medieval Faith and Fable (London, 1932), pp. 98–100.

⁸¹ Mod. Phil., xxxiii (1936), 232–235.

⁸² M.L.N., li (1936), 28–30. Le Braz, op. cit., 3rd ed., I, p. xxxii. In a number of Breton folktales in which the heroine marries a dead man, it is clear that originally she married the personified sun. This proves that little reliance can be placed on modern folktales as evidence of a medieval death-cult. Cf. Rev. celt., ii, 302.

⁸³ H. Newstead, Bran the Blessed in Arthurian Romance, p. 137, n. 7.

⁸⁴ Ed. W. Foerster (Halle, 1890), ll. 5436 f., 5463.

⁸⁵ H. O. Sommer, Vulgate Version of the Arthurian Romances, iii (Washington, 1910), 143, l. 35 f.

⁸⁶ Ibid., i, 289, l. 24 f. For other examples cf. PULA, xlviii (1933), 1007 ff.

⁸⁷ J G. Evans, White Book Mabinogion, cols. 473, 486. Loth, Mabinogion, I, 291, 318 f.

⁸⁸ W. F. Skene, Four Ancient Books of Wales, i, 274.

⁸⁹ Ibid., p. 266.

⁹⁰ Rhys, Hibbert Lectures,² p. 248.

⁹¹ Facsimile and Text of Book of Taliesin, ed. Evans, p. 34, l. 8.

⁹² T. Wiliems, Dictionarium Duplex (London, 1632). Cf. Lloyd-Jones, Geirfa Barddoniaeth Gynnar Crymraeg, Rhan 3, p. 271.

⁹³ Rhys, Celtic Folklore, ii, 679.

⁹⁴ E. Anwyl, Poetry of the Gogynfeirdd (Denbigh, 1909), pp. 96, 121. C. Guest, Mabinogion, Everyman Library, p. 350.

⁹⁵ Gwalchmai and Gwalhavet, names invented or substituted by the Welsh for the French Arthurian names of Galvain and Galahot (?), seem to have been provided arbitrarily with a patronymic Gwyar. Cf. Evans, While Book Mabinogion, col. 469. Rhys's surmises in regard to these names (Studies in Arthurian Legend, p. 166–169), seem baseless, since Galahad is demonstrably a Biblical substitution (A. Pauphilet, Études sur la Queste del S. Graal (Paris, 1921], p. 136) and Galvain has no original connection with Gwalchmai. Cf. PMLA, xliii (1928), 385–390; liv (1939), 656 ff.

⁹⁶ Loth, Mabinogion, ² ii, 267.

⁹⁷ Ibid., i, 40.

⁹⁸ Rhys, Hibbert Lectures, ² p. 27 f.

⁹⁹ Evans, White Book Mabinogion, col. 492 f.

¹⁰⁰ Evans, White Book, col. 493.

¹⁰¹ Rhys, Celtic Folklore, ii, 678. Cymmrodor, xxviii, 238. Enc. Brit., ed. 11, v, 642. Rhys and others, as Prof. Jackson points out to me, were mistaken in thinking the noun síd by itself meant a “fairy.”

¹⁰² Facsimile and Text of Book of Taliesin, ed. Evans, p. 34, l. 8. For poor translation of whole poem cf. Skene, Four Ancient Books, i, 274.

¹⁰³ Cymmrodor, xxviii, 197.

¹⁰⁴ For rendering of these lines cf. ibid., p. 236; Rhys, Arthurian Legend, p. 301; Rhys, Celtic Folklore, ii, 678.

¹⁰⁵ Cormac's Glossary, ed. J. O'Donovan (1868), p. 114. R. S. Loomis, Celtic Myth and Arthurian Romance, p. 187.

¹⁰⁶ Cf. supra notes 42, 43.

¹⁰⁷ Cross, Slover, Ancient Irish Tales, p. 505.

¹⁰⁸ Zts. f. Celt. Phil., xviii (1929), 193. On Goibniu's Feast cf. Atlantis (London), iii (1862), 389 n.; Stokes, Windisch, Irische Texte, iv, pp. 177, 189, 327.

¹⁰⁹ Pedeir Keinc, ed. I. Williams, p. 46 f. Loth, Mabinogion, ² i, 148 f.

¹¹⁰ Rhys, Arthurian Legend, pp. 269, 394. Loth, op. cit., i, 145, n. 2. Pedeir Keinc, p. 214 G. Owen, Description of Penbrokeshire, ed. H. Owen (London, 1892), p. 112.

¹¹¹ Rhys, Celtic Folklore, i, 171 f. A variant of this island tradition is that of the submerged country. Ibid., i. 382—ii, 419.

¹¹² Quoted ibid., I, 161. Cf. also Wirt Sykes, British Goblins (London, 1880), p. 9 f.; Rhys, Arthurian Legend, p. 269–271.

¹¹³ Pedeir Keinc, pp. 46. f. Loth, op. cit., ii, 149.

¹¹⁴ Pedeir Keinc, p. 47. Loth, op. cit., i, 150. Though Williams questions (Pedeir Keinc, pp. 220–222) the translation of both ysbydawt and urdaul, with all deference to his authority I cannot share his skepticism. Cf. Loth. i, 390.

¹¹⁵ Cf. n. 102 and n. 103.

¹¹⁶ This startling adjective probably refers to the tradition found in Peredur, Wolfram's Parzival, and other romances, where the Grail castle is the scene of lamentation. Rev. celt., xlvii (1930), 40.

¹¹⁷ Pedeir Keinc, p. 215 f. Cymmrodor, xxviii, 197.

¹¹⁸ Rev. celt., xlvii, 39 ff.

¹¹⁹ Pedeir Keinc, p. lii f. Malory, Morte d'Arthur, Everyman Library, p. xxii, n. 2.

¹²⁰ W. J. Gruffydd, Math Val Mathonwy (Cardiff, 1928). Transactions of Honourable Soc. of Cymmrodorion, 1912–13, p. 4 ff.

¹²¹ J. Rhys, Hibbert Lectures, p. 263 f.; Studies in the Arthurian Legend, p. 10; Malory, Morte d'Arthur, Everyman, p. xxv.

¹²² Evans, White Book Mabinogion, col. 462.

¹²³ Ibid., col. 498 f. Loth, Mabinogion, ² i, 334 f.

¹²⁴ Ed, J. J. Parry (Urbana, 1925), ll. 908–940. On date cf E. Farai, Légende arthurienne, prem. partie (Paris, 1929), ii, 37

¹²⁵ On this name cf. Romanic Rev., 1938, p. 176 f.

¹²⁶ Pedeir Keinc, ed. I. Williams, p. 100. It should be clearly understood that Margan is not the original Welsh name of the dwywes o Annwfyn, but like Gwalchmai, Lawnselot, Bort (Loth, Mabinogion, II, 285), is an importation from France. So unfamiliar was Morgan in Wales that the redactor of Geraint, finding this name in his French original, plus the mysterious Breton word Tud, took her to be a male physician. Cf. Rev. celt., xiii (1892), 496 f.; Romania, xxviii (1899), 322 ff. For other hypotheses about Morgan Tud cf. L. A. Paton, Studies in the Fairy Mythology of Arthurian Romance (Boston, 1903), p. 259 ff.; Loth, Contributions à l'étude des romans de la Table Ronde (Paris, 1912), p. 51. The original of Morgan is the former goddess Modron. In addition to the evidence I have already adduced in my Celtic Myth, 192 f. and Romanic Rev., 1938, p. 176 f. note that according to a Welsh tale (Aberystwyth Studies, iv, 105–109; J. G. Evans, Hist. MSS. Comm., Report on MSS. in Welsh Language, i [1898], 911) the mother of Owain and Morfudd by Urien, who was Modron (Loth, Mabinogion, ² ii, 284), announced herself as the daughter of the King of Annwn. All the evidence fits together beautifully, as to the identity of Modron with Morgan, her original divinity, her connection with the island paradise of Avallach, and its equation with Annwn.

¹²⁷ On this passage, cf. T. D. Kendrick, The Druids (London, 1927), pp. 139 f.; L. A. Paton, Studies in the Fairy Mythology, pp. 43 f.; Rev. celt., x (1889), 352.

¹²⁸ Cf. supra n. 81.

¹²⁹ MP, xxxviii (1941), 302 ff.

¹³⁰ F. Lot in Romania, xlv (1918–19), 1 ff.

¹³¹ H. M. and N. K. Chadwick, Growth of Literature, i (Cambridge, 1932), 103 f., 263, 461.

¹³² Ibid.

¹³³ Cross, Slover, Ancient Irish Tales, pp. 199, 328, 353. Banquet of Dun na n-Gedh, ed. J. O'Donovan, p. 51.

¹³⁴ Cross, Slover, op. cit., p. 28. Cf. V. Hull in Zeits. f. Celt. Phil., xviii, 86. On Celtic caldrons cf. A. C. L. Brown, in Kittredge Anniv. Papers, p. 235 ff.

¹³⁵ Stokes, Windisch, Irische Texte, iii, 205 f. Banquet of Dun na n-Gedh, p. 51. On elaborate etiquette of meat-portions cf. Transactions of Roy. Irish Ac., xviii, Antiquities, pp. 200–211.

¹³⁶ Stokes, Windisch, iii, 210.

¹³⁷ MS. Peniarth 51, Nat. Lib. Wales, Aberystwyth, p. 170. “pair dyrnwch.”

¹³⁸ MS. Peniarth 77, p. 214. “Pair Tyrnoc (written above: dyrnawg) cawr, o rhoid ynddo gic i wr llwrf, digon byth nis berwai, ac o rhoid i wr dewr digon a verwai ar vrys, ag yno y caid gwahan rhwng y dewr ar llwrf.”

¹³⁹ Cross, Slover, op. cit., pp. 506 f. Stokes, Windisch, iii, 211 ff. Romanische Forschungen, xlv (1931), 72.

¹⁴⁰ If the Head of Annwn be identified here with Bran vab Llyr, as the analogies with the mabinogi of Branwen would indicate, then the caldron of the Head of Annwn must possess the functions of Bran's caldron, which restored to life dead warriors who were cast into it. But the Pedeir Keinc, where we find this property assigned to Bran's caldron, are notoriously muddled, and we may legitimately question whether this is not a case in point; for surely it is a somewhat repellent notion to cast corpses into a cooking vessel, and nowhere else, if I am not mistaken, though the references to caldrons in medieval Irish and Welsh literature are numerous, is one of them used for this purpose. And it is not difficult, I believe, to see how this mistake arose. That the feasts of Bran, like Manannán's Feast of Age, preserved those who partook from sickness, old age, and decay, seems fairly patent, and it seems natural to infer that since caldrons provided the boiled meat at such banquets, Bran's caldron was credited with the virtue of preserving youth and of curing wounds. Now perhaps it was the famous classical legend of Medea's caldron into which she plunged the aged Aeson and restored his youth, or, more probably, the Irish tradition of the resuscitation of slain warriors by immersion in a well or bath (cf. C. O'Rahilly, Ireland and Wales [London, 1924], p. 108; A. H. Krappe, Balor with the Evil Eye [New York, 1927], pp. 132 ff.), which has led the author of Branwen to attach to the reviving caldron of Bran this unseemly method of healing. Thus, perhaps, the property of preserving and restoring the health of those who partook of its contents, so suitable to the vessel of the Head of Annwn, where there was neither age nor decay, has been transmogrified into the grotesque faculty of restoring life to the corpses cast into it.

¹⁴¹ Lists of the Thirteen Treasures have been printed in E. Jones, Bardic Museum (London, 1802), pp. 47 ff.; Y Brython, 1860, p. 372; C. Guest, Mabinogion, Everyman Library, pp. 328 f.; Y Greal (London, 1805), p. 188.

¹⁴² Rhys (Arthurian Legend, p. 313) points out that the Irish cognate mias is used in Matth, xiv, 8, for the charger on which John the Baptist's head was placed. Cúroí Mac Dáire possessed a caldron, vat (dabach), drinking horn, and platter (mias). Zts. f. Celt. Phil., iii, 39. Thurneysen, Irische Helden- und Königsage (Halle, 1921), p. 445.

¹⁴³ H. Newstead, Bran the Blessed in Arthurian Romance, pp. 20, 68, 86 ff'.

¹⁴⁴ Evans, White Book Mabinogion, col. 237.

¹⁴⁵ But not from Chrétien's Ivain. Cf. A. C. L. Brown, in Romanic Rev., iii, 143 ff.; Loomis in Mod. Lang. Notes, xliii (1928), 215; R. Zenker, Ivainstudien, Beiheft, Zts. f. Rom. Phil., lxx (1921).

¹⁴⁶ MS. Peniarth 77, Natl. Library of Wales, pp. 215, 213.

¹⁴⁷ T. Jones in MLR, xxxv (1940), 403 f. supports this translation and shows that two conflicting concepts of Bran's nature existed. On calet cf. Cymmrodor, xxviii, 192.

¹⁴⁸ Zts. f. Celt. Phil., i (1896), 288. H. Newstead, op. cit., pp. 23, 95 ff.

¹⁴⁹ Newstead, op. cit., pp. 163–167.

¹⁵⁰ MP, xxii (1924), 113.

¹⁵¹ Geoffrey of Monmouth, Vita Merlini, ed. Parry, 15–19. Skene, Four Ancient Books, I, 66, 174. Nennius, Historia Brittonum, ed. Lot, pp. 75 f.

¹⁵² Canu Aneirin ed. I. Williams (Cardiff, 1938), pp. xxxviii, xl, 17, l. 416.

¹⁵³ Loth, Mabinogion, ² ii, 311, n. 1.

¹⁵⁴ Jones, Bardic Museum, p. 47. Rhys, Arthurian Legend, p. 354; Hibbert Lectures, p. 155.

¹⁵⁵ Cf. supra, n. 123.

¹⁵⁶ Evans, White Book Mabinogion, cols. 461, 464, 466.

¹⁵⁷ Skene, Four Ancient Books, ii, 51. Malory, Morte d'Arthur, Everyman Lib., i, xix. J. G. Evans, Facsimile of Black Book of Carmarthen (Oxford, 1888), f. 47 v.

¹⁵⁸ Ibid.

¹⁵⁹ Rev. celt., xli, 489. C. O'Rahilly, Ireland and Wales, pp. 114 f. Sgilti Ysgawndroed, “the Nimble-footed,” is of course the swift Fenian hero, Caeilte.

¹⁶⁰ W. J. Gruffydd, Math Vab Mathonwy.

¹⁶¹ Rev. celt., xii (1891), 127. It occurs in MS. Harley 5280, Brit. Mus., f. 69a, a ms. of first half of sixteenth century.

¹⁶² E. Hull, Cuchullin Saga, p. lvi f.

¹⁶³ R. Flower, Cat. of Irish MSS. in Brit. Mus., ii (London, 1926), 319. Prof. Vernam Hull would date it still earlier, Zts. f. Celt. Phil., xviii, 80, 89.

¹⁶⁴ Cf. supra n. 160 and C. O'Rahilly, Ireland and Wales, pp. 117–119.

¹⁶⁵ Speculum, viii (1933), 419 f. For influence on Chrétien's Charrette, cf. M. F. Speyer in Romanic Rev., xxviii (1937), 195 ff.

¹⁶⁶ Lancelin is recorded in Brittany as early as 1069. Cf. H. Morice, Mémoires pour servir de preuves à l'histoire ecclésiastique et civile de Bretagne (Paris, 1742–46), i, 432. Zimmer, Lot, and Bruce all recognize the influence of Lancelin. Cf. Zts. f. Franz. Spr. u. Lit., xiii (1891), 43 ff.; Romania, xxv, 12; J. D. Bruce, Evolution of Arthurian Romance (Baltimore, 1923), i, 193.

¹⁶⁷ Merlin, ed. G. Paris, J. Ulrich (Paris, 1886), i, 223, n. 1, 225 f., 230. On the original identity of Balaain and Beaumains cf. PMLA, liv (1939), 656 ff.; Loomis, Celtic Myth, 250–252, 348.

¹⁶⁸ Loomis, Celtic Myth, pp. 348 f. For confusion of Gales and Gaule, cf. Sommer, Vulgate Version, i, 293, n. 10. In a ms. of Wace's Brut (Bib. Nat. fr. 375, f. 219r) “Franche” and “Cambrie” are both equated with Gales.

¹⁶⁹ Cf. also the figure of King Luces in the Estoire del S. Graal. MLR, xxiv (1929), 425–427.

¹⁷⁰ Though Hiberus seems to be the best authenticated form, Prof. Hammer kindly informs me that two mss., Lincoln Cathedral, No. 98, f. 91v, and Camb. Univ. Lib., Mm. 5. 29, f. 93r, give Hibernius.

¹⁷¹ Loth, in Rev. celt., xvi (1897), 84. E. T. Griffiths, Chantari di Lancelotto (Oxford, 1924), p. 186. PMLA, xliii (1928), 386 ff.

¹⁷² PMLA, xliii, 388; xlviii (1933), 1018–21.

¹⁷³ Loomis, Celtic Myth, pp. 7, 11–15, 22, 228. PMLA, xlviii, 1005, 1015–21, 1032.

¹⁷⁴ Loomis, op. cit., pp. 59 f. A. Buchanan in PMLA, xlvii (1932), 316 f. PMLA, xlviii, 1024–26.

¹⁷⁵ Loomis, op. cit., pp. 93 f. Romania, liv (1928), 518; lxiii (1937), 388.

¹⁷⁶ Rhys, Arthurian Legend, p. 301.

¹⁷⁷ J. Strachan, Introduction to Early Welsh (Manchester, 1909), p. 12.

¹⁷⁸ K. Meyer, A. Nutt, Voyage of Bran (London, 1895), i, 169.

¹⁷⁹ Ériu, xii (1938), 181.

¹⁸⁰ A. C. L. Brown, Iwain, Studies and Notes in Philology and Lit., viii (Boston, 1903), 79 f. Rhys, Arthurian Legend, p. 302.

¹⁸¹ Cf. n. 178.

¹⁸² Noted by Brown, Iwain, p. 80, n. 1. Cf. also Loomis, Celtic Myth, 110–114.

¹⁸³ Ed. M. R. James (Oxford, 1914), pp. 13 f. Trans. F. Tupper, M. B. Ogle (London, 1924), pp. 15–18. Cf. MP, xxxviii (1941), 301.

¹⁸⁴ Giraldus Cambrensis, Opera, vi (London, 1868), ed. J. F. Dimock, 75 f. Trans. Itinerary through Wales, Everyman Lib., pp. 68 f.

¹⁸⁵ Rhys, Celtic Folklore, i, 158–160, 166–169.

¹⁸⁶ Evans, White Booh Mabinogion, col. 433–436, 445–451. On Gwiffert cf. Zenker in Zts. f. Franz. Spr. u. Lit., xlviii (1926), 30–37.

¹⁸⁷ There is no comprehensive treatment of the dwarfs in Arthurian romance. Besides Zenker's article just cited and the tales summarized below, cf. Loomis, Celtic Myth, p. 22, n. 36, p. 160; Durmart le Gallois, ed. E. Stengel (Tübingen, 1873), pp. 125 f.; Webster in Harvard Studies and Notes, xvi (1934), 208 f.; C. Potvin, Perceval le Gallois, v (Mons, 1870), 41–78; H. O. Sommer, Vulgate Version, iii, 280 ff.

¹⁸⁸ Evans, White Book Mabinogion, col. 482. Loth, Mabinogion, ² i, 309; ii, 252, n. 4.

¹⁸⁹ Chrétien de Troyes, Erec, ed. W. Foerster (1890), l. 2005, reading of mss. H and C.

¹⁹⁰ Rev. celt., iii (1876–78), 85. On Beh cf. J. Rhys, J. Brynmor-Jones, Welsh People (1909), pp. 39–13, 133; W. J. Gruffydd, Math Vab Mathonwy, pp. 173–177; H. Newstead, Bran the Blessed in Arthurian Romance, pp. 164–167; Geoffrey of Monmouth, Historia Regum Britannìae, ed. A. Schulz (Halle, 1854), p. 238.

¹⁹¹ Arthurian Legend, pp. 335–337. On relationships of Beli, Avallach, Modron, and Bran cf. Cymmrodor, ix (1888), p. 170, n. 5, p. 174; viii (1887), 85; Romanic Rev., xxix (1938), 176 f.; Pedeir Keinc, ed. Williams, pp. 218, 226; Loth, Mabinogion, ² ii, 284; Loomis, Celtic Myth, pp. 189–193. There were other more or less historical Belis: Beli ab Mynogan, Beli ab Benlli Gawr, Beli ab Rhun, Beli ab Elfin.

¹⁹² Aberystwyth Studies, iv (1922), 105–109. J. G. Evans, Hist. Mss. Comm., Report on Mss. in Welsh Language, i (1898), 911.

¹⁹³ Transactions of Third International Congress of Religions (Oxford, 1908), ii, 236 f. J. Rhys, J. Brynmor-Jones, Welsh People (1909), p. 43 n. Rhys, Celtic Folklore, ii, 432. T. Jones, in MLR, xxxv (1940), 404, questions this interpretation of biw Beli and Beli wirawt.

¹⁹⁴ Chrétien, Erec, ll. 1993–2011. Cf. MP, xxxviii (1941), 292 ff. The Irish too knew this concept of a high-minded king of dwarfs. Cf. Cross, Slover, Ancient Irish Tales, pp. 471–475.

¹⁹⁵ Perceval le Gallois, ed. Potvin, iv (Mons, 1868), 33–53. On this story cf. Rev. celt., xlvii (1930), 46 ff.; Bennett in Speculum, xvi (1941), 43 ff.

¹⁹⁶ Sommer, Vulgate Version, v, 251–254, 263–268.

¹⁹⁷ Ibid., pp. 393–396.

¹⁹⁸ Ibid., pp. 339–403.

¹⁹⁹ Ibid., iv, 343–347. The story has been contaminated by the Lit Périlleux motif. Cf. PMLA, xlviii (1933), 1013–18. On Pelles cf. Romanic Rev., xxviii (1938), 354 f. and Bruce's learned but mistaken article in MP, xvi (1918), 113 ff.

²⁰⁰ Perlesvaus, ed. Nitze, i, 24.

²⁰¹ Sommer, Vulgate Version, v, 403. On traditional origin of this episode cf. MLR, xxiv (1929), 425–427.

²⁰² J. L. Weston, Legend of Perceval (London, 1906–09), ii, 40.

²⁰³ Perlesvaus, ed. Nitze, i, 112.

²⁰⁴ Ibid., I, 61, 164–166, 264 f.

²⁰⁵ Ibid., i, 54–58.

²⁰⁶ Ibid., i, 390.

²⁰⁷ Ibid., i, 402–406. The author has made two episodes out of one as was his practice. Cf. ibid., i i, 162; Newstead, op. cit., p. 105.

²⁰⁸ Ed. F. Lot, Bibl. de l'Ecole des Hautes Etudes, vol. 273 (Paris, 1934), p. 156.

²⁰⁹ Cf. Perlesvaus, ed. Nitze, ii, 152 f.

²¹⁰ On various medieval traditions of buildings of glass cf. references furnished by Patch in PMLA, xxxiii (1918), 610 n.; by L. A. Paton, Studies in Fairy Mythology, p. 40 n. 2; by J. D. Bruce, Evolution of Arthurian Romance, i, 200, n. 14.

²¹¹ Perlesvaus, ed. Nitze, ii, 58. Cymmrodorion Record Series, ii (London, 1901), p. 410. On date cf. Tatlock in Speculum, xiii (1938), 139–152.

²¹² Ed. W. Foerster (1890), ll. 1946 ff.

²¹³ Ibid., p. 264, l. 36.

²¹⁴ Perlesvaus, ed. Nitze, ii, 275. Cross, Slover, Ancient Irish Tales, pp. 594 f. Historia Meriadoci, ed. J. D. Bruce (Baltimore, 1913), pp. xxxiii, 31. A. C. L. Brown in Romanic Rev., iii (1912), 159.

²¹⁵ Rhys, Arthurian Legend, p. 301.

²¹⁶ On date cf. Perlesmus, ed. Nitze, ii, 73–89; Romanic Rev., xxxvii (1937), 351 f. For previous discussions of these resemblances cf. R. S. Loomis, Celtic Myth, pp. 201–204; Speculum, viii (1933), 428 f.; Perlesvaus, ed. Nitze, ii, 151–155.

²¹⁷ Perlesvaus, i, ll. 9543–9603.

²¹⁸ Zts. f. Celt. Phil., ii (1899), 130.

²¹⁹ Evans, White Book Mabinogion, col. 70. The inferior Red Book text adds that the chains were attached to the bowl.

²²⁰ M. O'Donnell, Life of Columcille, ed. A. O'Kelleher, G. Schoepperle (Urbana, Ill., 1918), p. 399. The cowl was still preserved in Kilmacrenan in Donegal in his day. Ibid., p. xl, n. 3; p. 403.

²²¹ MP., xiv (1916), 385 ff.

²²² Perlesvaus, ed. Nitze, i, ll. 9605–34.

²²³ Ibid., l. 9763.

²²⁴ Ibid., ll. 9809 f.

²²⁵ Pedeir Keinc, ed. I. Williams, pp. 27, 51.

²²⁶ Ibid., p. 51.

²²⁷ Ibid., p. 9.

²²⁸ J. L. Weston, Legend of Perceval, ii, pp. 20 ff.

²²⁹ Loomis, Celtic Myth, pp. 183–186, 194, 262. Speculum, viii, 423 f.

²³⁰ Peredur, Wolfram's Parzival, Perlesvaus, Manessier.

²³¹ Pedeir Keinc, p. 50.

²³² I say “cognate” because manifestly the Welsh texts are somewhat remote from the oral traditions from which they and the Breton conteurs ultimately drew.

²³³ Revue celt., xxxiii (1912), 452 ff.

²³⁴ On inconsistency of characterization in the Matter of Britain cf. Newstead, Bran the Blessed in Arthurian Romance, p. 149 and n. 10; T. Jones in MLR., xxxv, 403 f.; A. B. Hopkins, Influence of Wace on the Arthurian Romances of Crestien de Troies (1913), pp. 80–102.

²³⁵ PMLA, xliii (1928) 386 ff.; liv (1939), 656 ff. On Gwrvan as a separate name cf. T. Stephens, Lit. of the Cymry (Llandovery, 1849), p. 132.

²³⁶ R. S. and L. H. Loomis, Arthurian Legends in Medieval Art (New York, 1938), pp. 32 f. PMLA, xliii, 388 f.

²³⁷ R. S. Loomis, Celtic Myth, pp. 330–338.

²³⁸ Here is the explanation of much that Miss Weston (Legend of Perceval, ii, 301 ff.) found so provocative of speculation.

²³⁹ Loomis, Celtic Myth, pp. 206 ff. Sone de Nansai, ed. M. Goldschmidt (Stuttgart, 1899), ll. 4339 ff., 17017 ff.

²⁴⁰ Bruce, Evolution of Arthurian Romance, i, 350, n. 16.

²⁴¹ Revue celt., xlvii (1930), 56. Romantiche Forschungen, xlv (1931), 70. Speculum, viii, 430.

²⁴² Newstead, op. cit., 86–95.

²⁴³ Historia Meriadoci, ed. J. D. Bruce, p. xxxiv.

²⁴⁴ Ibid., pp. 43 f.

²⁴⁵ On heathenism in bardic poetry cf. H. M. and N. K. Chadwick, Growth of Literature, i, 469–471; Morris-Jones in Cymmrodor, xxviii, 238–248. On heathenism in custom and saga cf. Loomis, Celtic Myth, pp. 39–51, 184–187.