^{page 220 note 1} Certain problems considered in the following pages were discussed very briefly in a paper read at the meeting of the Modern Language Association at Princeton University, in December, 1908.

^{page 220 note 2} Cf. Kittredge, Harvard Studies and Notes, vol. viii, pp. 227 ff.

^{page 222 note 1} Antiq. Tidsk., 1852–3, p. 130, cf. Bugge, below.

^{page 222 note 2} Consult Müllenhoff, Beowulf, Berl., 1889, pp. 55 ff.; ten Brink, Beowulf, Strassburg, 1888, pp. 185 ff.; Symons, Germ. Heldensage, Strassburg, 1889, p. 44, and Paul's Grundriss, vol. iii, p. 649; Symons' views have been taken from the later work, “Züge aus dem anglischen mythus von Béaw-Biar. … wurden auf den dänischen sagenhelden (BöÐvarr-)Bjarki, durch Ähnlichkeit der Namen veranlasst, übertragen”; Boer, Die Beowulfsage, Arkiv f. nord. Filol., vol. xix, pp. 45 ff., cf. esp. pp. 47 ff.; Kluge, Eng. Studien, vol. xxii, p. 144; Bugge, Paul-Braune Beiträge, vol. xii, pp. 55 ff., cf. note in Grundtvig, Danmarks gamie Folkeviser, iii, p. 801; Sarrazin, Anglia, vol. ix, pp. 195 ff., Eng. Studien, vol. xvi, pp. 71 ff., vol. xxiii, pp. 242 ff., and vol. xxxv, pp. 19 ff., also his Beowulf-Studien, Berl., 1888, pp. 13 ff. References to Paul's Grundriss in the present paper are always to the second edition.

^{page 223 note 1} Danmarks Heltedigtning, Köb., 1903, vol. i, p. 135.

^{page 223 note 2} Hrólfs saga kraka og Bjarkarímur, udgivne for samfund til udgivelse af gammel nordisk litteratur, Köb., 1904. Cf. p. xxii. In his Oldnorske og Oldislandske Litteraturs Historie (1898) he called it “en rigtignok svag afglans af det fra Bjovulf bekendte Grendelsagn,” vol. ii, p. 832.

^{page 223 note 3} Heusler's reviews of Olrik are to be found in Anz. für deutsches Altertum, vol. xxx, pp. 26–36; Zts. für deutsches Altertum, vol. xxxvi (NF), pp. 57–87. It is perhaps worth while to give Heusler's comment in full: “Die frage nach dem Zusammenhang von Béowulf-Biár-Biarki behandelt O. s. 134 ff., 244. 248 behutsam und einleuchtend. Nur wenn, nach ausweis der Biarkarímur, der bär an die stelle des geflügelten Ungeheuers tritt, und wenn man das bluttrinken Hialtis als die spitze der erzählung gelten lässt, bleibt ein zusammengesetztes motiv übrig: ‘ein held kommt von Schweden (Gautland) an den Dänenhof und tötet ein ungetüm, das durch sein nächtliches erscheinen die hofmannen in schrecken hält ‘—ein motiv, dessen ähnlichkeit mit dem von Béowulf doch wol über den zufall hinausgeht. Und dann wird man es nicht ganz abweisen, dass der name Biarki (=Bericho) den etymologisch unverwanten, aber ähnlich klingenden namen Biár (= Béaw) angezogen habe, und dass dadurch der Rolfskämpe Biarki inhaber jenes fabulosen abenteuere wurde.”—Anz., loc. cit., p. 32.

^{page 224 note 1} Arkiv, vol. xxi, p. 276.

^{page 224 note 2} Zts. für Volkskunde, vol. xiv, p. 250.

^{page 224 note 3} Paul's Grundriss, vol. ii, p. 842.

^{page 224 note 4} Vol. i (1907), p. 29.

^{page 224 note 5} Paul's Grundriss, vol. ii, p. 993.

^{page 225 note 1} The Hrólfssaga has been edited by Rafn, Fornaldarsögur, Cop., 1829, vol. i; by V. A'smundarson, F. A. S., Reykjavik, 1891, vol. i; by Finnur Jónsson, Cop., 1904. Danish translation by Rafn, Nordiske Kœmpe-Historier, Cop., 1821, vol. i. There is an excellent German translation by P. Herrmann, Die Geschichte von Hrolf Kraki, Torgau, 1905. This contains much useful supplementary material; parallel passages from related sources, etc. For further bibliography consult Herrmann, p. 4. The above rendering is based on Jónsson's text; but I have followed Herrmann's example in not keeping the present tenses, which interchange with the preterits in a way disturbing to narrative in modern English.

^{page 227 note 1} Jónsson, Hrólfssaga, pp. 68 ff.; Herrmann, pp. 73 ff.

^{page 228 note 1} Cf. Jónsson's Introduction, esp. pp. xxvi ff. See also his Oldnorske og Oldislandske Litterature Historie, vol. ii, pp. 829 ff., and Mogk, Paul's Grundriss, vol. ii (2nd ed.), pp. 841 ff.

^{page 228 note 2} Heltedigtning, p. 69.

^{page 229 note 1} Jónsson, Hrólfs saga kraka og Bjarkarímur, esp. pp. xxviii ff.

^{page 230 note 1} Jónsson, pp. 139–140, Herrmann, p. 73.

^{page 230 note 2} Jónsson, pp. 141–142, Herrmann, p. 75. These two passages are paraphrased by Olrik, Heltedigtning, pp. 116–117. Indeed, it is best not to attempt to render the elaborate rhymes and repetitions of the original too literally, cf. Herrmann's note, p. 2. The sense is occasionally obscure, and the ms. defective.

^{page 230 note 3} Hrólfssaga, etc., p. xxii.

^{page 231 note 1} Saxo, ed. Holder, Strassburg, 1886, p. 56.

^{page 231 note 2} Heltedigtning, pp. 116 ff., pp. 134 ff.

^{page 232 note 1} Cf. Jónsson, Hrólfssaga, etc., p. xxvii.

^{page 232 note 2} Cf. Boer, Archiv, vol. xix, p. 52, “Ganz willkürlich ist schliesslich die annahme, der zug, dass Bjarki Höttr-Hjalti das blut des bären trinken lasst, sei in der dänischen sage die pointe der erzählung. … Mit gleichem rechte kann man solchen behauptungen gegenüber vollständig entgegengesetzte axiomata aufstellen.”

^{page 233 note 1} Zeitschrift für deutsches Altertum, vol. xxxvi, (NF), p. 62. Italics are mine.

^{page 233 note 2} Beovulf, Berlin, 1889, p. 55.

^{page 234 note 1} Studien zur Sagengeschichte Englands, i, pp. 249 ff. Cöthen, 1906.

^{page 234 note 2} Printed in Gaimar, Lestorie des Engles, ed. Hardy and Martin, London, 1888, vol. i, pp. 339 ff. This is a more accurate text than those of Bright and Michel, (cf. Introd., p. xlvii). For general criticism, cf. Introd., p. lii f. The passage here reproduced will be found on pp. 343–4.

^{page 235 note 1} It is interesting to note, in passing, that the women and girls made songs in his honor, “mulieres ac puellæ de eo in choris canebant,”—an incident for the attention of students of the development of popular poetry.

^{page 236 note 1} Cf., on this general subject, Deutschbein, loc. cit., Olrik, Arkiv, vol. xix, (1903) pp. 199 ff; and Heltedigtning, pp. 215 ff; for the Sivard-saga, Langebek, Scriptores rerum danicarum medii aevi, Hafniae 1774. vol. iii, pp. 288 ff; for the account in Saxo, Holder, p. 345, cf. Herrmann, Gesch. von Hrolf Kraki, p. 52. The bear-father episode may be well seen in Cosquin, Contes Pop. de la Lorraine, vol. i, (Jean de l'Ours, etc.). The reference in Olrik's Archiv article to the bear's ears in the son, (shown in Hartland, Legend of Perseus, to be very wide-spread), should read Legend of Perseus, iii, 24.

^{page 238 note 1} It is perhaps unnecessary to give references to the location of the hall Heorot, and its identity with the residence of Hrolf. Cf. Olrik, Heltedigtning, p. 16, where the fate of the hall foreshadowed in Beowulf ll. 81–85 is explained by the events at Hrolf's death; and O.'s general discussion, pp. 188 ff.

^{page 238 note 2} Müllenhoff, Beowulf, p. 46, “Der ruhm Hrothgars (Hroars) ist in der nordischen sage auf seinen neffen Hrothulf (Hrolf Kraki) übergegangen.”

^{page 240 note 1} For a criticism of ten Brink, cf. Boer, Arkiv, vol. xix, pp. 50 ff. Boer's views seems too much affected by his theory of a dragon-myth to be impartial; cf. p. 58, “auch in Saxo's quelle war das ungetüm schon aller Wahrscheinlichkeit nach ein fliegender drache; ein dem Grendel ähnlicher unhold wäre bei ihm unmöglich zu einem bären geworden,” etc. Skeat, it will be remembered, advanced the theory that there was so much in the figure of Grendel to suggest a bear that this might explain his origin. (Journal of Philology, vol. xv, pp. 120 ff.)

^{page 240 note 2} On this general subject, see Brandl's review of Olrik's destructive criticism of the parallel to the dragon fight in Beowulf afforded by an adventure of Frotho I, as related in Saxo's second book. This parallel, elaborated by Sievers, (Ber. der Gesells. der Wiss. zu Leipzig, vol. xlvii, pp. 175 ff.) has been generally accepted. Brandi defends it, in part, … “das Vorhandensein von Verschiedenheiten hebt die Beweiskraft der Übereinstimmungen nicht auf, gibt nur dem Nachahmer etwas von Originalität.” (Paul's Grundriss, vol. ii, p. 997). Note changes in the visualization of Grendel and his mother in the stories of Grettir and Orm. Or consider the variations in the shape and attributes of the monster in the Chapaluc or Cath Paluc legends (E. Freymond, Artus' Kampf mit dem Katzenungetüm, Halle, 1899, esp. pp. 45 ff.). There is much about the methods of the author of Tristram de Nanteuil, who worked over the old Chapalu motive, with such changes and elaborations as he saw fit, to remind one of the processes in the reshaping of the Hrólfssaga. (Freymond, pp. 26 ff.).

^{page 241 note 1} Paul-Braune, Beiträge, vol. xii, p. 56.

^{page 241 note 2} Boer, Zeitschrift fur deutsche Philologie, vol. xxx, p. 65, ascribes the Beowulf passage in the Grettissaga to a bookish person of the late thirteenth century, denying any proof here that the story was displaying an “unbewusste neigung” to attach itself to popular heroes. He admits the probability of its having been familiar in the north, however. He can see no force in the parallel in the Orm story. His arguments do not seem to have proved convincing,—both Olrik and Brandi speak of the latter as an analogue.

^{page 242 note 1} Englische Studien, vol. xxii, p. 145.

^{page 242 note 2} Neue Beowulf-Studien, ibid., vol. xxxv, pp. 19 ff. The article is dated Oct., 1904.

^{page 242 note 3} It may be observed that Hrunting. is expressly stated to be a good sword, hit œt hilde ne-swāc, 1461, which does not suggest the character of Hött. It is no reproach to a sword if it cannot survive an attack on a supernatural creature when wielded by a mighty hand,—Beowulf's sword Naegling breaks in the dragon fight, (2680 ff). The bone-throwing contest is not much like the flyting with Hunferth. Sarrazin further equates Hjalti with Wiglaf, and says that the speeches of Wiglaf recall those of Hialto in Saxo Grammaticus, to which attention had been called by Bugge. Wiglaf's sword, too, has, according to Sarrazin, “gewissermassen” the function of the old demonic sword, which we have seen is, on his theory, to be equated with Hjalti. All these correspondences I confess myself unable to follow. The resemblance between these speeches has been admirably criticised by ten Brink as due to the formal character of Germanic poetry in a given situation, “Ähnlichkeit der Situation ruft Ähnlichkeit der Ausführung von selbst hervor. … Es muss in der germanischen Poesie eine Art Typus fur die Fassung derartiger Reden gegeben haben, der trotz aller Variationen immer durchschimmerte.” (Beow., pp. 191–2). It will be shown presently that there does not appear to be any connection between the dragon-fight at the end of Beowulf and the fight at the end of the saga, in which a porcupine-troll plays a minor role.

^{page 243 note 1} Jónsson seems to have misunderstood this point, cf. p. xxii, “Hertil skal föjes, at i rimerne (v, 5–14), er der endnu tale om en ‘graabjorn,‘ der kommer og dræber Rolvs faar og kvæg i foldene; denne dræbes af Hjalte med et sværd, hvorved hann ogsaa fik sit tilnavn: Hjalte, og bliver nu hirdmand.” This cannot be the case; the name which the king gives him is not Hjalti, but hinn hjartaprúÐi, a poetical variation of his appelation hinn hugprúÐi in other Icelandic monuments, cf. Snorra Edda, ed. Jónsson, Cop. 1900, p. 108. If the king were the first to bestow this name on him, his mother could not say “Átta eg son er Hjalti hét,” etc. before the bear-killing.

^{page 244 note 1} Cf. Herrmann, note 12, p. 131.

^{page 244 note 2} On the general subject of names cf. Olrik, pp. 137 ff.

^{page 244 note 3} Beowulf-Studien, 1888, p. 47.

^{page 244 note 4} Englische Studien, vol. xvi, p. 82; ibid., xxiii, pp. 245 ff.

^{page 245 note 1} Hrólfssaga, Cap. xxxiii, Jónsson, p. 102.

^{page 246 note 1} For a general discussion of this matter, cf. Brandi, Paul's Grundriss, vol. ii, pp. 992 f., Olrik, p. 137 note, and p. 244 note; and Heusler, Anzeiger, vol. xxx, pp. 26 ff. The quantities of the vowels have been marked in this passage, in order to make the linguistic discussion perfectly clear; elsewhere the marks of length have been purposely omitted, save in some titles in Scandinavian.

^{page 246 note 2} Grundriss, vol. iii, p. 649.

^{page 246 note 3} Arkiv, vol. xix, pp. 19 ff.

^{page 246 note 4} Cf. Beowulf-Studien, 1888, p. 47, and Holthausen's review, Literaturblatt, 1890, No. 1, p. 15.

^{page 247 note 1} Beowulf, Berl., 1889, pp. 8 ff

^{page 247 note 2} Sitzungsberichte, loc. cit., p. 181. Cf. the statements of Koegel, Zeit. für deutsches Alt., xxxvii, pp. 268 ff.; Binz, Paul and Braune, Beiträge, vol. xx, pp. 153 ff.; Symons, Grundriss, iii, pp. 648 ff.

^{page 248 note 1} Grundriss, ii, p. 999.

^{page 248 note 2} Heltedigtning, p. 246.

^{page 248 note 3} Archiv, xix, pp. 28 ff.

^{page 248 note 4} Beowulf, Heidelberg. 1906, p. vii.

^{page 248 note 5} Cf. Englische Studien, vol. xvi, pp. 73 ff.

^{page 248 note 6} Vol. i, p. 31.

^{page 249 note 1} Beowulf, Cambridge, Mass., 1904, p. ix.

^{page 249 note 2} Beowulf: Scyld Scefing—Beowulf—Healfdene.

A.S. Chronicle: Sceaf. … Sceldwa—Beaw.

Aethelweard: Scef—Scyld—Beo.

Wm. of Malmesbury: Sceaf—Sceldius—Beowius.

^{page 249 note 3} Olrik makes Sceaf originally Sceafa, king of the Lombards in Widsith. This is denied by Chadwick and Heusler. See below, p. 259.

^{page 249 note 4} Cf. Olrik, pp. 239 ff., esp. p. 246.

^{page 250 note 1} Paul-Braune, Beiträge, vol. xx, pp. 155 ff.

^{page 251 note 1} Brandl points out that place-names are only of significance “wenn sie erst in einer zur sage stimmenden Relation auftreten” (Archiv, p. 152), and Symons, Grundriss, iii, p. 650) recognizes that the testimony of the place-names, apart from the present passage, is “weniger entscheidend.” For full reference to Brandl's article, cf. note, p. 263.

^{page 251 note 2} Gray-Birch, Cartularium Saxonicum, London, 1887, vol. ii, p. 364. I have modernized the A.S. character for w.

^{page 252 note 1} to bares anstigon.

^{page 252 note 2} Gray-Birch, iii, p. 84.

^{page 253 note 1} “There is an interesting approximation of the expressions beowan hammes and grendles mere in Cart. Sax., No. 677. The conjunction has been used as an argument to prove the local distribution of the Beowulf legend, and to found an historical generalisation.

I am induced by a recent reappearance of this argument to point out that grendles is not a proper name. The Charter has fugel mere, wvdu mere, grendles mere. The word grendel stands alone in C. S. 1103, and gryndeles sylle occurs in C. S. 996. In the former it is the ‘grindle,’ i. e., drain—see note ad loc. and Halliwell. In the latter the sense is ‘the grindle dirtpond’ (see Grein s. vv., sol, sylian) i. e., the dirty pond into which the drain runs (from gryndeles sylle to russemere). Hence in C. S. we have a series fugel mere ‘the bird-pool,’ wudu mere ‘the wood-pool,’ grendles mere ‘the cess-pool.’“ (Academy, May, 1894, p. 396.)

^{page 253 note 2} Binz, loc. cit., p. 157, note 3.

^{page 253 note 3} Cf. Heinzel, Anzeiger, vol. xvi, p. 267; ten Brink, Beowulf, p. 217, Anm. 2; Binz, p. 155. Binz objects to Heinzel's criticism of Beas broc on the ground that the strong inflectional form indicates a divine or mythic being. He refers to Kögel, Zs für deutsches Alt., xxxvii, p. 272, who says “mannesnamen nach göttlichen wesen pflegen in schwacher form, aus kompositis verkürzt, aufzutreten.”

^{page 255 note 1} Grundriss, loc. cit., p. 993.

^{page 255 note 2} Chadwick, p. 146. Mr. Chadwick does not propose this as a solution, but merely as a possibility.

^{page 255 note 3} Brandi (p. 993) admits, “Für die Verständlichkeit der Erzählung war sie kein Vorteil; Sagen zeigen daher in der Kegel das entgegengesetzte Bestreben, namensverwandte Gestalten zu vereinigen.”

^{page 257 note 1} Heltedigtning, p. 247. “Kampen med Grendel i Danernes kongehal har formodenlig faaet sin skikkelse ud fra forestillingen om hans danske byrd.” I am not sure that I fully understand Olrik's argument at this point.

^{page 257 note 2} Anzeiger, p. 32.

^{page 260 note 1} On this general subject, see Olrik, Heltedigtning, pp. 223 ff.; Chadwick, Origin of the Eng. Nation, pp. 269 ff.; Binz, Paul-Braune, Beiträge, vol. xx, pp. 147 ff.; Möller, Altengl. Volksepos, pp. 43 ff.; Müllenhoff, Beowulf, p. 9.

^{page 260 note 2} For a discussion of this, see G. Schütte, Oldsagn om Godtjod, pp. 13–33, Cop., 1907. See esp. his summary, p. 31 f. “Kun hos enkelte Forskere finder vi fuld Udprægning af de hinanden modsatte Standpunkter: yderst paa Mytesiden staar Scherer og Kögel, yderst paa den ‘flade Euhemerismes’ Side staar Wilhelm Müller.”

^{page 260 note 3} Mogk, Paul's Grundriss, vol. iii, p. 244.

^{page 260 note 4} Archiv, loc. cit.

^{page 263 note 1} “Müllenhoff hat Grendel für die Nordsee, Mogk für einen Walfisch, Laistner für einen Nebel erklärt; Breca gilt bei Müllenhoff für den Sturm, bei Möller für den Golfstrom, bei Sarrazin für die untergehende Sonne, bei Heinzel nur für einen berühmten Schwimmer. Daraus ersieht man, wie wenig es möglich ist, den alten mythischen Kern noch herauszuschälen.” Sitzungen der Berliner Gesellschaft für das Studium der neueren Sprachen, 12–26 Feb., 1901. Archiv, vol. 108, p. 153.

^{page 266 note 1} Custom and Myth, p. 62.

^{page 267 note 1} The Mabinogion, translated by Lady Guest, with notes by Alfred Nutt, London, 1902, p. 332.

^{page 269 note 1} Sitzungsber. der Gesell. der Wiss. zu Leipzig, vol. xlvii, p. 175.

^{page 270 note 1} Ll. 535 ff.

^{page 270 note 2} It is hardly necessary to explain that the word “myth,” as used in the present paper, does not mean merely an invented story, something having no existence in fact, but “a traditional story in which the operations of natural forces and occurrences in human history are represented as the actions of individual living beings, especially of men, or of imaginary extra-human beings acting like men.” See the full definition in the Century Dictionary under myth. It will be observed that a mere folk-tale, even one to the hero of which divine attributes have been given, does not belong under this definition, unless it can be shown that the adventures narrated were conceived at some time as explaining abstract ideas or natural phenomena.

^{page 272 note 1} P. 10.

page 273 Note:—Since the above was sent to the printer, Professor Gummere has expressed his opinion in regard to mythological elements in Beowulf in no uncertain way. “Undoubtedly one is here on the border-land of myth. But in the actual poem the border is not crossed. Whatever the remote connection of Beowulf the hero with Beowa the god, whatever this god may have in him of the old Ingævonic deity whom men worshipped by North Sea and Baltic as god of fertility and peace and trade, whatever echo of myths about a destroying monster of invading ocean tides and storms may linger in the story of Grendel and his horrible mother, nothing of the sort comes out of the shadow of conjecture into the light of fact. To the poet of the epic its hero is a man, and the monsters are such as folk then believed to haunt sea and lake and moor. Hrothgar's people who say they have seen the uncanny pair speak just as real rustics would speak about ghosts and strange monsters which they had actually encountered. In both cases one is dealing with folk-lore and not with mythology. When these crude superstitions are developed by priest and poet along polytheistic lines, and in large relations of time and space, myth is the result. But the actual epic of Beowulf knows nothing of this process; and there is no need to regard Grendel or his mother as backed by the artillery of doom, to regard Beowulf as the embodiment of heaven's extreme power and goodwill.” (The Oldest English Epic, N. Y., 1909 pp. 5 f.) A statement more completely in accord with the point of view in the present article could scarcely be desired.—W. W. L.