We have now to consider the question whether myth is a necessary element in the formation of heroic poetry. It has been noticed that historical persons figure in many stories of the Heroic Age, while others do not contain a single character whose historical existence can be authenticated. These latter stories are believed to be wholly mythical in origin, though they may not show any supernatural features in their final form. But even in stories of the former type it is held that some of the characters are almost always of mythical origin, and that their association with historical characters is a secondary development—due to confusion or to poetic imagination.

In the last chapter we put forward the view that Beowulf, the hero of the poem, has been confused with a mythical character of the same name, and that the adventure with the dragon originally belonged to the latter. It can scarcely be doubted that Scyld Scefing, the father of this earlier Beowulf, was also a mythical character. The only element in his story common to English and Scandinavian tradition is that he is regarded as the ancestor of the Scyldungas or SkiÖldungar, the Danish royal family, and all analogies suggest that he came into existence as their eponymus. The brief account of him given in the poem might, except in two particulars, be applied to almost any successful king of the Heroic Age. One exception relates to the story of the funeral ship, on which the dead king's body is sent out to sea. In spite of Prof. Olrik's doubts[195] I cannot but think that this is a reminiscence of ancient custom. The other is the reference to his arrival as an infant, likewise by sea—a story told more fully of Sceaf in certain English chronicles. The only question here is between myth and folk-tale. The story may fairly be classed under the latter head, though I think its origin is ultimately to be sought in a ritual myth. Scandinavian authorities, apart from Saxo[196], record nothing distinctive of SkiÖldr, except that he was a son of Othin and the husband of the goddess Gefion[197]—which again points to myth.

In Scyld-SkiÖldr we have the case of an eponymous ancestor appearing in the introduction to a poem which deals largely with the fortunes of his descendants. But there is no evidence that his own deeds ever formed the subject of an independent heroic poem. It would be somewhat hasty therefore to use this case as an argument for the origin of characters who are brought before us in the main action of heroic poems.

Next we may take the story of Weland, as to the mythical origin of which nearly all scholars seem to be agreed. It has indeed a historical or semi-historical connection in the fact that Weland is represented as the father of Widia (Wudga), i.e. the early Gothic hero Vidigoia mentioned by Jordanes (cf. p. 27); but this is held to be a secondary element in the story. In its original form the story is believed to have dealt only with the incidents related in VÖlundarkviÐa, viz. (i) the adventure with the swan-maidens, (ii) Weland's imprisonment by Nithhad and his revenge[198]. Behind the story itself however there lies a widespread belief in the existence of a supernatural smith of this name. Several places in Germany (Westphalia and Holstein) are reputed to be the scene of his operations, while in this country he is connected with the cromlech called Wayland Smith, near Ashdown in Berkshire. In its ultimate origin this belief is traced to the myth of a fire-demon. Certainly it is to be noted that the name Weland is of a very exceptional type—apparently participial in form. One can hardly help suspecting that it once had a definite meaning, though this cannot now be determined with certainty[199].

Now there can be little doubt that the adventure with the swan-maidens is derived from a folk-tale. In this part of the story there is no indication of a fire-demon, or even of a smith, while analogies for the incident are fairly common both in Teutonic lands and much farther afield. We may confine our attention therefore to the second and better known part—that which deals with Weland's imprisonment and revenge.

It is manifest that this story departs very decidedly from the ordinary standard of heroic poetry—firstly in the fact that the hero is here clearly represented as a smith, and secondly in the cruelty, treachery and vindictiveness ascribed to the chief characters. These are features which would be in place either in myth or folk-tale. But we may note further that there are analogies for part of this story, just as much as for the incident of the swan-maidens. As an example we may take Saxo's account of the robbing of Mimingus by Hotherus (p. 70 f.). Mimingus is a satyrus, i.e. clearly either an elf or dwarf, who dwells in a cave in an almost inaccessible forest. Hotherus surprises and binds him and then takes away his sword and a magical ring. A connection between the two stories is shown even in the name, for Weland's most famous sword is called Mimming[200]. For a more remote parallel we may compare the story of Loki and Andvari. Indeed the spoiling of a dwarf is quite a common motif in Northern tales, while at the same time such beings are constantly credited with extraordinary skill in metallurgy. The distinctive feature in the story of Weland, apart from the revenge, is that sympathy is on the side of the smith.

It is the end of the story—where Weland (VÖlundr) rises into the air and flies away—that is supposed to point most clearly to a fire-myth. But this feature cannot be traced except in the Norse version[201]. Moreover here we have also the adventure with the flying swan-maidens, in whose case there is no suspicion of such a myth. Setting aside this incident the story is perfectly explicable as a folk-tale founded on actual experience. There can be no doubt that in the Heroic Age—and indeed in much earlier times—princes were especially anxious to obtain slaves, whether foreigners or not, who were skilled in metallurgy. And it is by no means incredible that such slaves were sometimes lamed in order to prevent any attempt at escape—although, quite apart from this explanation, smith's work may be regarded as a vocation natural to the lame man, just as minstrelsy to the blind. Further, it is likely enough that servile smiths, when cruelly treated, would take any opportunity that presented itself of avenging themselves on their masters. For the murder of Nithhad's sons we have a somewhat striking historical parallel in Eugippus' Life of St Severinus (cap. 8). Feletheus, king of the Rugii, who were settled on the Danube in the time of Odoacer, had a young son who one day was entrapped by some goldsmiths in the queen's service. They threatened to take his life; but the saint intervened and rescued the boy on condition of the smiths obtaining their freedom[202].

What seems to me to be really the most remarkable feature in the story is that a person in this position should come to be made the subject of heroic or semi-heroic poetry; for it is plain enough from many sources, especially Saxo's History, that smiths were generally regarded with deep aversion. In Deor's Elegy Weland is said to be a more distinguished man than Nithhad; in VÖlundarkviÐa he is called a chief of the elves, while the introduction makes him the son of a king of the Finns. Yet, except in the late ThiÐreks Saga, his father's name is never given, and none of our authorities credit him with possessing a following of his own.

Now is there any real necessity for the assumption that Weland's relationship to Widia is a secondary development? It is found in two of the three national versions of the story[203], and hence dates back in all probability at least to the sixth century. It is not found in the VÖlundarkviÐa; but then Widia is altogether forgotten or unknown in Northern tradition. Moreover there is a distinct reference to offspring of Weland and Beaduhild at the end of the poem. Once grant that the relationship is old and the reason for the heroic treatment of the story becomes obvious. It is merely the reflection of the son's fame upon the father. As Widia is never said to be the son of anyone else the probability is that he was supposed to be illegitimate, and that a story was soon current as to his being the offspring of a union between a princess and a bondsmith. In such a case there would be a natural tendency to the accumulation of material from folk-tales about his parentage.

If this view is correct the story must of course come originally from the Goths or some neighbouring people. I cannot see that the Westphalian traditions are any more conclusive than the Berkshire cromlech as to its original home. If Weland was a character of folk-tale and his name had at one time a definite meaning, these local traditions may have been quite independent of the heroic story. The real difficulty seems to me to lie in determining the amount of material from folk-tale contained in the latter. We need not entertain any doubt as to the adventure with the swan-maidens. But what about Nithhad and Beaduhild? The latter name is not obviously framed to suit the character or circumstances of the unfortunate princess. Nithhad might be explained more easily in this way; yet a Gothic prince of that name is recorded by Jordanes (cap. 22). At all events there is no conclusive evidence in either case that these characters did not originally belong to the story of Widia.

Another story which is believed to be of wholly mythical origin is that of HeÐinn and HÖgni. In KÛdrÛn it appears as the introduction to a much longer story, from which point of view it bears a superficial resemblance to the story of Scyld and possibly to that of Weland. But since this feature is peculiar to the German poem and the second story seems to be entirely unknown from other sources, we can hardly do otherwise than treat the story of HeÐinn and HÖgni as an independent narrative.

This must have been one of the most popular stories of the Heroic Age, since it can be traced in England, as well as in Germany and the North. None of the characters however can be traced in any historical work[204], and the time to which it refers is quite uncertain. The Northern version of the story contains a supernatural element in the endless battle which forms its conclusion. It is generally held that this is the oldest element in the story and that Hild, whose name means 'war,' was really a valkyrie. The whole story then is to be regarded as a myth of 'unceasing strife between conflicting powers[205].' But we may naturally ask whether it is truly scientific, when dealing with a story known from three separate national traditions, to regard as the original element a feature found in only one of the three. It may be urged of course that the reconciliation, which in KÛdrÛn[206] takes the place of the tragic ending found in the Northern version, rendered it necessary to drop the mythical element; and again that we have extremely little information about the English form of the story. The latter remark is certainly true; but the little that we do know practically precludes the possibility of a mythical interpretation. Strictly speaking the passage in Deor's Elegy (cf. p. 56) is not a reference to the story at all, but a matter of fact statement by the poet that he had been in the service of the Heodeningas. Hence, in view of the fact that this passage—together with Widsith, v. 21—is probably by far the earliest reference to the story which we possess, I cannot regard the mythical interpretation as anything more than an extremely doubtful hypothesis[207].

Thus far we have been dealing with stories which are supposed to be of entirely mythical origin. Now we have to consider certain cases in which elements undoubtedly historical are believed to be blended with myth. As examples of this type we may take the stories of Waldhere and SigurÐr. In both cases the historical elements are practically the same.

In the former case it should be mentioned that the mythical theory is by no means universally accepted. Those scholars however who do adopt this interpretation base their view upon a supposed connection between the story of Waldhere and that of HeÐinn and HÖgni. The chief points of resemblance between the two are as follows: (i) The heroine is called in the one case Hiltgund (Hildegyth), in the other Hildr (Hild). (ii) Both stories deal with abduction (so-called) and then with fighting. (iii) The man (Waldhere, HeÐinn) who carries off the girl has in both cases to fight with a man called Hagen or HÖgni. (iv) In both cases the combatants have previously been friends—though strictly this feature applies only to the Northern version of the HeÐinn story. Now the first consideration carries no weight at all; for half the feminine names which occur in Anglo-Saxon heroic poetry contain the element -hild- (e.g. Beaduhild, Hildeburg), while in the Continental and Scandinavian authorities also they are extremely common. Again, the last consideration obviously has little validity, except when taken in conjunction with the other two. But these (the second and third) points of resemblance are, it seems to me, altogether misleading. To begin with it is hardly correct to apply the term 'abduction' to an escape of hostages, such as the story of Waldhere relates; at all events the conditions have nothing in common. Then the fight which follows is not, as in the case of HeÐinn and HÖgni, with an aggrieved father, or indeed with a pursuing force of any kind; it is an unprovoked attack made by a third party in the hope of plunder. Lastly, the part played by the person called Hagen (HÖgni) is quite different in the two cases. In one he is the injured father who is wholly responsible for the fight; in the other he is a vassal of Gunther (Guthhere), who is only drawn into it, with great reluctance, through the obligation of avenging his nephew (Patufrit), who has already been slain.

It is true that a different version of the story appears in ThiÐreks Saga af Bern[208]. Here HÖgni (Hagen) is represented as pursuing the fugitives on behalf of Attila, while Guthhere does not appear. Now it has been widely assumed that this version is an independent and more original form of the story than that contained in Ekkehard's work, in spite of the fact that it does not make its appearance till nearly three centuries after the latter—and probably nearly six centuries after the composition of the Anglo-Saxon poem. But the lapse of time in itself provides a perfectly adequate explanation of such divergencies, especially if we bear in mind the unfavourable conditions under which the heroic stories were preserved in Germany during the early Middle Ages. As the stories gradually became forgotten two tendencies are constantly observable: (i) to connect stories or incidents which originally were quite distinct, (ii) within the individual story to lose sight of all except the outstanding characters and incidents. Hence it is only in accordance with what we might expect that two different sets of opponents of the hero should be confused. For a parallel we may compare the late North German ballad on Eormenric's death (cf. p. 9), which describes how Theodric with eleven companions broke into the king's castle and slew him. It is generally agreed that this ballad is due to confusion of some kind with the story of HamÐir and SÖrli[209], who were likewise enemies of Eormenric, though not connected in any way with Theodric.

The saga itself really contains evidence which points to an earlier form of the story agreeing with that given by Ekkehard. For the hero is called Valtari af Vaskasteini[210], and there can be no doubt that this expression is to be explained by the rocky defile mentioned by Ekkehard (v. 490 ff.) in the Vosges (saltus Vosagus)—on the confines of Guthhere's dominions. Further it is to be noted that the story is introduced as an episode in the relations of Theodric and Eormenric, and that the hero is represented as a nephew of the last named. All scholars are agreed that this is due to late combination—and no doubt rightly. Yet Waldhere is associated with Theodric and Eormenric also in a number of German poems which are quite independent of the saga, and consequently it is by no means improbable that these combinations both preceded and helped to bring about the disappearance of Guthhere from the story.

In its earlier form the affinities of the story with that of HeÐinn and HÖgni are, as we have seen, scarcely worth consideration. Yet apart from this supposed connection there is no case for believing it to have a mythical foundation—except the assumption that myth is a necessary ingredient in every heroic story. Whether it is to be regarded as history or fiction is of course quite a different question and one which we shall have to consider later.

Of all the stories of the Heroic Age probably none has been more frequently referred to a mythical origin than that which deals with SigurÐr (Siegfried). It is held by the great majority of scholars that the Nibelungenlied and the corresponding Edda poems—or rather the earlier poems or legends on which both were based—came into existence through the amalgamation of an essentially mythical story with historical traditions of Attila and the fall of the Burgundian kingdom[211]. The original elements of the former are believed to have been as follows: A young prince is brought up by a cunning smith in a forest, away from his father's home. On reaching manhood he gains an immense treasure by killing a dragon; also he releases a maiden by overcoming difficulties and dangers, by fire or water, which were insurmountable to any other person. These two adventures are connected by many scholars. Later, the hero falls into the hands of foes, who slay him and take for themselves his wife and treasure.

It is held that this story was originally a myth of light and darkness—applying however to the course of the year as well as to that of the day. SigurÐr himself is a 'light-hero' and Brynhildr a 'sun-maiden' whom he releases at the dawn, while the treasure represents the blossoms of summer which the light-hero likewise wins by destroying the dragon of winter. Then, in the evening or autumn, he has to yield to the powers of darkness or winter. The original name of these powers was Niflungar or Nibelunge, a name connected with Old Norse nifl, 'mist,' Niflheimr, 'Hades.' Their chief representative is HÖgni or Hagen, who, like SigurÐr and Brynhildr, belongs to the mythical elements of the story. Many scholars also hold that the powers which destroyed the hero and appropriated his wife and treasure were originally identical with those from which he had won them at the beginning; and this view seems to be more or less involved by the interpretation given above, since day and night, winter and summer are constantly alternating with one another.

Now it will be obvious at once that the story as thus reconstructed differs greatly from both the forms in which it has come down to us. Indeed the only original feature preserved in both versions is the slaying of the dragon by SigurÐr. But it is only in the Norse version that the hero gains the treasure thereby; in the Nibelungenlied this is obtained by a different encounter, with two princes named Schilbung and Nibelung, while in the Seyfridslied it really belonged to certain dwarfs, the sons of Nybling[212]. Again, only the Norse version records that SigurÐr released Brynhildr—from a perpetual sleep with which she had been punished; the incident is not connected with the dragon adventure. The Seyfridslied does relate that the hero rescued a maiden from a dragon; but here the maiden is Kriemhild, whom the dragon has carried off from her home. Further, it is only in the German version that Hagen kills the hero. In the Norse version the actual perpetrator of the deed is Gutthormr, but the instigator is Brynhildr herself. Lastly, neither version of the story makes Brynhildr or the treasure return to their former owners, although—in the German version only—the first owners of the treasure bear the same name as those into whose possession it comes after the hero's death.

It appears then that the original form of this story has been greatly obscured in both versions. The explanation given is that, through confusion with a historical tradition, the Burgundian kings, Gunnarr (Gunther) and his brothers, have taken the place (as well as the name) of the Niflungar. Hence, in order to form a just estimate of the theory it is necessary to examine the various mythical characters separately. These are—in addition to the dragon—Brynhildr, SigurÐr and the Niflungar, including HÖgni.

The evidence for believing that Brynhildr was originally a mythical character lies chiefly[213] in the identification of her with the valkyrie SigrdrÍfa (cf. p. 13), who is mentioned only in the Norse version. The identity of the two characters is clearly recognised in the HelreiÐ Brynhildar and also in the prose authorities. On the other hand it is not recognised in GrÍpisspÁ, which is supposed to be a late work, while the other poems leave it uncertain. The evidence therefore on the whole is not very strong.

In the case of SigurÐr the evidence, apart from the valkyrie incident, rests upon his being the slayer of the dragon. But it is agreed that this part of the story must be connected with the similar adventure attributed by Beowulf to Sigemund (Sigmundr); so that the question at issue is whether the exploit was first related of the father or the son. We have already seen (p. 123) that the argument in favour of the latter based on the agreement of the Norse and German authorities is in reality misleading. Hence the balance of probability is in favour of believing that the incident has been transferred to SigurÐr from Sigemund.

In the case of the Niflungar the evidence depends upon the interpretation of the name. The use of the name is certainly somewhat curious. In the Seyfridslied (Part II) the dwarfs are called sons of Nybling. In the first half of the Nibelungenlied Nibelung is the name of one of the brothers slain by Siegfried in his youth, while the people who become subject to him, together with the treasure, on the death of the brothers, are collectively called Nibelunge. In the latter half of the poem however the same name is applied to the Burgundians. In the Norse version Niflungar always means Gunnarr and his people (i.e. the Burgundians), except perhaps in the expressions arfi Niflunga, hodd Niflunga (in AtlakviÐa), by which the prose authorities at all events understood the treasure which SigurÐr had taken from FÁfnir. The explanation given for this double use of the name is as follows. Originally it belonged to the mythical enemies of SigurÐr, i.e. HÖgni and his people—whether these were identical with the former owners of the treasure or not. Later, when HÖgni became associated with the historical Burgundian kings (Gunnarr etc.), the use of the term was extended so as to embrace them also. But it is to be observed that the interpretation of the name Niflungar as 'children of mist' or 'darkness' is not free from difficulty. In the Edda it is twice written Hniflung-[214], and on both occasions the H- alliterates, whereas alliteration with n- is never found. This fact suggests that the original form of the name was Hniflungar and that the form without H- is due to later influence—presumably on the part of scribes—from German sources, where of course the H- (before -n-) would regularly be lost at a much earlier date. If so the name cannot originally have had any connection with O. Norse nifl, etc.

In all these three cases then the evidence for the mythological interpretation of the story seems to be at best inconclusive. But we have yet to consider the case of HÖgni; and here it must be remarked that the demonic character of HÖgni is quite essential to the mythological theory. In the Norse version Gutthormr is a mere instrument and the person really responsible for the murder is Brynhildr herself—a feature obviously incompatible with the interpretation which we are discussing. HÖgni's mythical origin is as necessary for this interpretation as that of SigurÐr or Brynhildr. In order to maintain the theory the mythical character must be vindicated in all three cases alike.

Now it has been remarked that Hagen (Haguno, Hagano etc.) is not uncommon as a personal name even in quite early times. This is a curious fact if the name had such associations[215]. But there is a much more serious difficulty. It is altogether contrary to reason or probability to separate Hagen the vassal of Gunther in the Nibelungenlied from Hagen the vassal of the same Gunther in the story of Waldhere. His character in the two cases is certainly quite different. He is brave in both; but in the former he is both faithless and cruel, whereas in the latter he is an honourable man who is reluctantly drawn by circumstances into a course of action of which he heartily disapproves. But this is precisely the character borne by HÖgni in the Norse version of the story of SigurÐr—a fact which is the more remarkable since this type of character is extremely rare in heroic poetry. The agreement between the story of Waldhere and the Norse version seem to me to render it overwhelmingly probable that the character which they ascribe to HÖgni was that which he originally bore.

I am not arguing now to prove that HÖgni was a historical person. He may be of fictitious origin or even mythical (though the latter seems to me extremely improbable). But if so clearly he must have been taken either from the story of SigurÐr into that of Waldhere, or from the story of Waldhere into that of SigurÐr. Which of the two he belonged to originally is a question of minor importance. The essential point is that an earlier German form of the story of SigurÐr must be the link between the Norse version and the story of Waldhere; for there is no evidence that the story of Waldhere itself was ever known in the North. The conclusion to which we are naturally brought is that in this earlier German form of the story HÖgni bore the same character which is attributed to him in the Norse version. This character however is of course totally incompatible with a demonic origin; and here, it seems to me, the interpretation which we are discussing hopelessly breaks down.

I cannot help thinking that the investigation of the whole story has been greatly prejudiced by the application of wrong methods. There can be no doubt that a story of some kind—in which the adventures of SigurÐr were already combined with those of the historical Burgundian princes—was in existence long before the date of the earliest extant records, and that from this story, whether it was embodied in a single poem or consisted only of a mass of lays or legends, both the Norse and German versions are ultimately derived. It seems to me that, before trying to ascertain the origin of the various elements contained in the story, the object should be to determine the main features of this common foundation. The way to achieve this end is surely not by arbitrarily selecting one feature from the Norse version and another from the German, but by bringing together all the various features which the two have in common. To carry out such a process systematically would be quite beyond the scope of this book, but a brief outline of the scheme may not be out of place here.

First then we will take the part of the story relating to the hero's early adventures, which is preserved mainly in a different set of authorities from the rest. The chief German authority is the late Seyfridslied which, as we have seen, is really made up of two different ballads, inconsistent with one another. The Nibelungenlied contains only allusions to this part of the story, the action proper beginning shortly before the hero's arrival at Worms. The Norse version is given in the trilogy, ReginsmÁl, FÁfnismÁl and SigrdrÍfumÁl, as well as in VÖlsunga Saga, which is derived from these poems. The account given in ThiÐreks Saga is mainly a combination of the German and Norse versions, though it has one or two features peculiar to itself.

In this part of the story the common elements are very few in number. (1) SigurÐr kills a dragon; (2) SigurÐr gains a great treasure. In the Norse version the two adventures are combined, but in the Nibelungenlied the treasure belonged to Nibelung and his brother who had quarrelled and who are both killed by the hero. It may be noticed that Reginn and FÁfnir are also brothers who have quarrelled over a treasure, and they too are both killed by SigurÐr; but FÁfnir has become a dragon—or perhaps a reptile. The Seyfridslied, Part I mentions only the killing of a dragon (serpent), while Part II unites the acquisition of the treasure with the killing of a dragon—a fiery dragon—but states that the hero erroneously thought that the treasure belonged to the dragon. Really it belonged to the three sons of the dwarf-king Nybling, who are friendly to the hero and not killed by him. As a further common element we may mention (3) that SigurÐr is brought up by a smith. This story is found in the Seyfridslied, Part I, and in ThiÐreks Saga—practically also in the Edda, since Reginn is represented as a smith. Again, (4) both in the ballad and in Norse prose authorities SigurÐr breaks the smith's anvil, though the circumstances are quite different. It is doubtful whether we should connect the eating of FÁfnir's heart, which enabled the hero to understand the birds, with the German story that he became invulnerable by bathing in the dragon's blood. Further, we have seen that the awakening of the valkyrie in SigrdrÍfumÁl has practically nothing in common with the rescue of the maiden (Kriemhild) from the dragon related in the ballad (Part II). Lastly, it is to be observed that though the hero's father has the same name (Sigemund) in all authorities, there is great discrepancy as to his childhood. In the Edda he is posthumous but knows his parentage, in the Nibelungenlied he is brought up at his father's court, in the ballad, Part I, he leaves his home and goes to the smithy, in Part II and ThiÐreks Saga he does not know his parentage—in the latter indeed he is a foundling and suckled by a hind. It will be seen that this part of the story is permeated throughout by the supernatural and marvellous.

From the time of the hero's arrival at the Burgundian court we may take the Nibelungenlied for the German version, while the Norse one is best represented by the poems from the fragmentary SigurÐarkviÐa I to AtlamÁl. For the earlier portion we have also to use VÖlsunga Saga and the prose Edda in place of certain poems which are lost (cf. p. 13). In this part of the story the elements common to the two versions are far more numerous and striking. (1) SigurÐr comes to the Rhineland (Worms in the German version) and marries a sister (GuÐrÚn, Kriemhild) of King Gunnarr (Gunther). (2) SigurÐr in supernatural disguise wins Brynhildr for Gunnarr. (3) SigurÐr again in supernatural disguise sleeps with Brynhildr and takes from her a ring[216]. (4) Brynhildr quarrels with SigurÐr's wife, and the latter shows her the ring[216]. (5) Brynhildr bitterly resents the treatment she has received and devises the hero's death. (6) SigurÐr is killed by treachery; but the versions differ in regard to the perpetrator of the deed. (7) The hero's widow is for a long time irreconcilable, but eventually is married to Atli (Etzel). (8) Gunnarr with HÖgni and many others are invited to Atli's home. (9) The gold is sunk in the Rhine. (10) Gunnarr and HÖgni are captured alive and the rest killed in Atli's land. (11) A demand for the gold is made and refused. (12) Gunnarr and HÖgni are killed.

It will be seen that the supernatural is here confined to (2) and (3); indeed these are almost the only incidents in which it occurs in this part of the story. There is a difference between the two versions in regard to the character of the supernatural disguise. In the Norse version, where the two incidents are combined, SigurÐr and Gunnarr have exchanged forms; in the German Gunther is present in both cases, though Siegfried, who has rendered himself invisible, is the real actor. Several other important differences between the two versions have already been noted (p. 13 f.). In addition to these each version has of course many characters and incidents peculiar to itself.

Of the discrepancies enumerated on p. 13 f. the fifth is by far the most important, since Kriemhild's revenge for Siegfried forms the central motif of the second half of the Nibelungenlied. In the Norse version no such central motif is to be found. In the prose piece DrÁp Niflunga Atli's conduct is attributed to revenge for the death of Brynhildr, while in VÖlsunga Saga it is ascribed to his lust for SigurÐr's gold. But in the poems themselves no real explanation is given, and the connection between this part of the story and that relating to SigurÐr is scarcely more than a personal one—viz. that GuÐrÚn, Gunnarr and HÖgni figure in both. This however is a phenomenon for which parallels are to be found in other heroic stories, e.g. those of Beowulf and Weland. It is now held—and doubtless rightly—by the majority of scholars that the unity of interest imparted to the Nibelungenlied by the motif of Kriemhild's revenge is a later improvement on the somewhat disconnected story given in the Edda. For our present purpose however the question is immaterial, since it is not contended that this part of the story is of mythical origin.

In spite of the discrepancies noted above it cannot be denied that the two versions contain a remarkable number of identical features in this part of the story—a fact which renders all the more striking the very slight amount of agreement in the part dealing with the hero's early adventures. Unless all analogies are misleading the conclusion to which we are driven is that the original story began more or less where the Nibelungenlied begins, and that the hero's youthful adventures are later accretions, such as we see gathering round the childhood or ancestry of other heroes, e.g. Biarki (cf. p. 120). We may add also the cases of StarkaÐr, Hagen in KÛdrÛn and perhaps Witege (cf. p. 135). They appear to be derived, in part at least, from folk-tales. One of these—affecting probably only the Norse version—may be identified with the Scandinavian story of Svipdagr and MenglÖÐ (cf. p. 12), a variety perhaps of that of the Sleeping Beauty. Another is that of the forest dwarf who forges or preserves a magical sword. We may note that in ThiÐreks Saga the smith is called Mimir, a name which recalls Saxo's Mimingus (cf. p. 133)[217]. The story of the treasure-guarding dragon may also be included in this category, though strictly perhaps it belongs rather to popular belief than to folk-tale. From the fact that some of these elements are common to both versions we may probably infer that the process of accretion had begun before the story reached the North. Yet there do seem to be some indications of a reflex influence[218]—from the North or some region exposed to Northern influence—upon the development of the story in Germany.

In addition to folk-tales we must take into consideration also a tendency which is often associated with them—the desire to account for an obscure name. This seems to be the most reasonable explanation of the names Nibelung and Nibelunge in the first part of the German epic and Nybling in the ballad—all denoting the original owners of the treasure. We have seen that in the Norse version, as well as in the latter half of the Nibelungenlied, Niflungar means the Burgundians. May we not suppose that it was really a dynastic name[219] like Scyldungas, Uffingas, Merewioingas? In that case of course hodd Niflunga (hort der Nibelunge) ought to mean the family treasure of the Burgundian kings. But is it quite certain that AtlakviÐa does not use it in this sense? That it is identified with FÁfnir's treasure in later authorities may be due to subsequent German influence. As for the fact that the name Nibelunge is used for the Burgundians only in the second half of the German epic, may not this spring from some stylistic peculiarities of the 'common foundation'? It is not necessary to suppose that the latter was all the work of one author or even of one generation.

In dealing with questions such as these we cannot hope to get beyond a reasonable hypothesis, since the paucity of common features between the two versions admits of few definite conclusions. But from the time when the hero arrives at the Burgundian court the case is quite different. In spite of certain discrepancies there is no difficulty in determining the main outline of the story. Even in the most important point of all—the true cause of the hero's death—the two versions are really in full agreement. Gutthormr does the deed in one version, Hagen in the other; but in both alike it arises out of the bitter resentment cherished by Brynhildr, owing to the deception which has been practised upon her. We have seen that this motif is incompatible with the current mythological interpretation of the story. But more than this, it is plainly not a motif derived from mythology at all, but from real life.

It must not be overlooked that the Brynhildr and HÖgni of the Norse version are in the nature of character-studies. Both appeal to our sympathies, though we do not approve of the actions which they commit or allow. Here we are in a region of thought as alien as possible to that of the folk-tale. But it is also alien to that period of thought, which was most open to the influence of folk-tales, the period which we have called Stage III in the history of German poetry. In such a period the person who destroyed the hero must necessarily be a villain as black as Hell. Between the instigator of the deed and the perpetrator, who by this time was Hagen—whether this was so originally or not is immaterial—the choice was made, not unnaturally in the circumstances, in favour of the latter, while the former was allowed, awkwardly enough, to drop out of the story. Thus the peculiarities of the German version may be explained quite naturally as modifications of an earlier form similar to the other—modifications necessitated by the conditions under which heroic poetry was preserved in Germany. The effect produced is somewhat similar to that which would be obtained by converting a modern problem play into a popular melodrama.

The conclusion then to which we are brought is that the supposed traces of myth, so far as they have any foundation at all, are due to late accretions to the story, while the central motif in both versions alike is by no means of a mythical character, but essentially human. Consequently the story of SigurÐr stands quite on a line with the other stories of the Heroic Age. Most of them contain elements which may be interpreted as mythical; but these elements are always most prominent in the latest forms of the story. It must not escape notice that those scholars who most strongly uphold the mythical interpretation base their arguments chiefly on such works as the Seyfridslied and ThiÐreks Saga af Bern. The explanation is that myth is a growth which requires time to develop. Even Beowulf is no real exception to the general rule, for in the latter part the hero is probably confused with a namesake whose story may have been of considerable antiquity, while the only character in the poem who is quite clearly of mythical origin is the first ancestor of the Danish royal family[220].