significance is to be attached to the personal coloring of this passage will be considered at length later. But even granting that this personal quality is the reflection of actual experience, the burden of proof upon the supporters of this hypothesis is a heavy one. The only other historical patron besides Eormanric to whom the singer refers with anything like a personal touch, who could have been known to a man who had attained his majority by the beginning of the third quarter of the fourth century, is Gunther, the historical Gundicarius, who fell in battle in 437. Ealhhild cannot be the daughter of Audoin and the sister of Alboin, as she is almost universally regarded. Eadgils, if he has any historical position, must be moved back into the fourth or early fifth century. The citations of the various peoples of course prove nothing. Nor will the list of monarchs and chiefs in the "innweorud Eormanrices" (ll. 112 ff.) help the case for putting the kernel of the piece in the fifth century. Dr. Garnett seemed to think otherwise: "He speaks .. distinctly of his comradeship with the Goths when they were contending against the bands of Ætla (Attila)." The probability that this is a purely fictitious list of rulers seems as great as that there is no sober record of fact in the list of peoples in 11. 82 ff. Müllenhoff has compared the mechanical use of sōhte ic with the equally mechanical use of ic was in the passages preceding which fall under the suspicion of being spurious. The "innweorud Eormanrīces" is a jumble of names, a few of which belong to history, but of widely different periods, as Theodric (l. 115), not the Frankish monarch, but the Gothic king (died, 526), the Hunnish Attila (died, 543), and the Lombard Egelmund, who reigned in the early days when the Lombard people were still in their seats in the north of Europe. Others belong to saga, like Becca; the Bikki who betrayed Randver and Swanhild in the Eormanric story; Sifeca and Heathoric, who are the traditional Sifke and Heidrek, and the equally imaginary pair Hlithe and Incgentheow, whom Grundtvig explained as Hlödh and Angantyr. The mythical Harlung brothers appear as the Herelingas, Emerca and Fridla. Others are utterly unknown, or at best darkly conjecturedWulfhere and Wyrmhere, Rædhere and Rondhere, whose names

have a suspiciously "decorative" look, Rumstan and Secca and Becca and Withergield and Aliso and Hungar.' Is it safe to read any serious personal experience into all this? Let it be granted that it preserves an early form of the Gothic saga-the chieftain Wudga, the later German Wittich in the Dietrich of Bern story, here appears in his proper setting if we allow him an historical counterpart in the old Gothic hero Widigoia or Widigauja. As to his friend and companion Heime or Hama, there are no conjectures to help out a decision as to his identity, save that he is the constant companion of Wudga, and so probably of like nationality. It is clear that after fiction and probable interpolation have been cut out of this passage there is little to base historic truth upon. The description of the contests between the Goths and the Huns shows discrepancies. The very mention of a series of battles instead of one great contest may indicate epic error, and the strife of the followers of Eormanric with the Huns of Attila about the Vistula is puzzling.' Dr. Garnett suggested: "It . . . . seems not unlikely that Widsith's lays on the conflicts between the Goths and the Huns really related to those which took place under Hermanric's immediate successors, but that the passage has been altered by a later poet, for whom Attila was the representative of the obliterated Hunnish nation, now passing into the domain of legend. Is it not more probable that this change took place in the oral tradition upon which such an account as this must rest, and that the passage in its present form was composed by a man who really had a wrong conception of the facts? But the possibilities of theorizing on the basis of the introduction of new names in the place of old ones are so varied that it is hardly profitable to carry this train of thought further. It seems evident, however, that if matters are as confused as this, no sound conclusions as to the life of the singer can be drawn from the Eormanric catalogue. If the Eadwine (1. 117) is Audoin,

993

1 Cf., for discussions of these names, Kögel, Gesch. der deutschen Litt., Vol. I, pp. 146 ff.; Binz, Beiträge, Vol. XX, p. 207; Bugge, Beiträge, Vol. XII, pp. 69 ff.; Jiriczek, loc. cit.

2 Cf. Heinzel, Hervararsaga p. 517.

3 Jagic, Arch. für slavische Philol., Vol. XI (2), pp. 305 ff., makes a similar suggestion, which Garnett may have had in mind, as his reference to modern Slavonic scholars suggests. Jagic remarks: "Attila, der legendhafte Eponym des Hunnenvolkes, möchte einen älteren Namen leicht verdrängt haben; die umgekehrte Änderung ist kaum wahrscheinlich."

the father of Alboin, it would seem that its present date, at least, must be pretty late.

The amount of actual testimony to the composition of any part of the poem by a man who had actually "seen Eormanric," then, is very small. It cannot be regarded as otherwise than highly dangerous, on the basis of such slender evidence, and the preservation of so small a part of the original poem, to refer its nucleus to so remote a period as the late fourth or early fifth century, a time when, as Professor Saintsbury has pointed out in this connection, no modern European language has left any traces of its existence.

The hypothesis that the poem reflects actual historic events of the fifth century, and that Widsith, if he were a real person, lived in that age, has been far more generally credited. The great learning and authority of Müllenhoff made his remarks on the historical events and the ethnology of the piece of very great weight. His arguments were accepted, in general, by Walker,' after a careful summary of the evidence up to 1885. They are especially worth notice as having formed the basis for the more elaborate studies of Möller and ten Brink.

Müllenhoff cut out as spurious 11. 75-87, and remarked that 1. 88 would make a good continuation for 1. 74: "V. 88 schliesst sich auch vortrefflich an v. 74 an. V. 88 bezieht sich offenbar auf v. 70, und die ganze folgende Schilderung von des Sängers Verhältnis zu seinem Herrn und zur Ealhhild, der Tochter Eadvine's tritt erst ins rechte Licht, wenn unmittelbar das Lob Elfvine's, des Sohnes Eadvine's, voraufgeht, und umgekehrt auch dieses, wenn jene unmittelbar folgt." It is not difficult to agree with Müllenhoff that the passage has all the earmarks of spuriousness, but that there was originally no gap between 11. 74 and 88 seems an unwarrantable assumption. The two hardly make a faultless connection. What does ealle prage mean? One of the commonest errors of the Liedertheoretiker was supposing that because an interpolated passage had been removed, and the beginning and end of the gap made good sense, no loss had taken place. It would be easy enough, as has been often pointed out in this con

nection, to cut out long passages in modern poems, so that no one unfamiliar with their original condition would guess that anything had been taken away. It seems quite likely here that the passion for mere information, for making this an "instructive" poem, may have led to the sacrifice of matter that would have explained the vague indication in the prologue that Widsith had accompanied Ealhhild to the home of Eormanric. But this is a mere conjecture. The main point is to examine Müllenhoff's interpretation of the part which has been preserved.

His argument is closely connected with his investigation into the identification and position of the different peoples. It really arises from the discussion of the location of the Myrgings.' They are treated in ll. 41-44 as the same folk as the Suevi or Swæfs; one of the exploits of Offa, the Anglian king, is that

āne sweorde

merce gemærde wið Myrgingum

bi Fifeldore: heoldon forð sippan

Engle and Swæfe, swa hit Offa geslōg.

They are mentioned separately after the Swæfs in 1. 22, but this does not necessarily mean that they cannot have been a division of the same people. As for the Swæfs themselves, Müllenhoff notes that their position according to the poem is "noch ganz in der stellung wie die Suebi in den ersten jhh. an der Elbe und Oder." But he thinks that the Myrgings were not a folk of this region. "Dass die Myrginge hier kein theil der Svæfen, etwa alte Holsteiner sind, beweist ihre verbindung mit der Langobarden an der Donau und in Pannonien." The Lombards in the time of Alboin were occupying lands in modern Austria, south of the Danube, and west of its southern course from Buda-Pesth downward, having crossed about 547 from the region lying east of the river. They were thus in an advantageous position to make their descent upon Italy in 568. Müllenhoff is convinced that the Myrgings were not up in Holstein, or thereabouts, as all the indications in the poem lead one to infer, but that they extended into much more southerly territory, not at a great distance from the Lombards in Pannonia. This view is all the more surprising, as the 1 Haupt's Zeitschrift, Vol. XI, pp. 278 ff.

location of "Fifeldor" in the passage just quoted is believed to be the river Eider, in the lower part of the Danish peninsula.' Peoples in these early days were subject to migration, however, and a very convincing reason might lead to placing the Myrgings farther south than would otherwise be believed.

2

Müllenhoff argues that they were neighbors of the Lombards because Ealhhild, the daughter of the Lombard Eadwine (Audoin) married the Myrging prince Eadgils. But it has been shown that there is no statement in the poem to that effect, and, furthermore, that there are grave objections to that interpretation. Müllenhoff thinks this marriage could not have taken place if the Myrgings had been restricted to Holstein: "Es kann aber der Langobardenkönig in Pannonien keine interesse gehabt haben seine tochter nach Holstein zu verheiraten. Der Myrgingenname muss eine viel grössere ausdehnung gehabt haben." Various reasons, he says, tend to confirm the conclusion that they may have been neighbors. The Saxons and Swabians are known to have followed Alboin into Italy, and they came from a district, "das von der Elbe durchströmte und östlich anliegende land von der Donau bis zur Ostsee," where the people were known as Maurungani, as the map of the Geographer of Ravenna indicates. Moreover, according to Mullenhoff, the Lombard saga of Paul the Deacon puts Mauringaland "eben dahin." Finally, the name seems to be preserved in relatively modern place-names; cf. the minnesinger Heinrich von Morungen, the curtis Moranga in pago Morangano in the Vita Meinwerci, etc. Hence, he thinks, one cannot doubt the linguistic identity of "Maurungi, Mauringi, Myrgingas," and the chain is complete.

1 Mallenhoff, Deutsche Mythologie, ed. Meyer, Berlin, 1875, Vol. I, p. 198; BosworthToller, A. 8. Lexicon, etc.

2 In showing Eadwine, the father of Ealhhild and Elfwine, to have been the Lombard king Audoin, he points out the fact that the name occurs in 1. 117 in connection with other Lombard heroes-Egelmund, Hlithe (if Ettmüller's conjecture be correct). Cf. n. 1, p. 344. The mention of the Wid-Myrgings in the following line seemed to him additional testimony to the close connection which he wished to establish between the peoples. But it is evident from the preceding lists that no sound conclusions can be drawn from grouping, otherwise one would have to see relations between the Greeks and the Finns in 1. 20. Müllenhoff himself says: "Eine strenge ordnung, wie im guten mhd. epos bei dergleichen aufzählungen, weiss ich freilich nicht nachzuweisen "(p. 276). In the days of Egelmund, the Lombards were near what we may believe to have been the seats of the Myrgings, a consideration which may per. haps have a little weight in the matter. The map in Hodgkin's Lombard Invasion, p. 80, will be found useful.