both of which are attributed to the twelfth century by Heusler and Ranisch¹². Unlike the poem on the battle between the Goths and the Huns, neither of these forms a story complete in itself. They presuppose the existence of a saga in some form or other, presumably oral, dealing at least with the fight at Samsø; and the existence of such a saga in the twelfth century is confirmed by the account of the same event given by Saxo¹³.

A third stage in the literary development of the heroic legends is represented by the written saga itself, which has evidently been formed by the welding together, with more or less skill as the case may be, of several distinct stories, and of more than one literary form. A particularly striking instance of this is to be found in the Hervarar Saga with its stories of the Heroic and Viking Ages, the poems dealing with the fight on Samsø, the primitive Riddles of Gestumblindi and the early poem of the battle between the Goths and Huns¹⁴. Something of the same kind has also taken place in the composition of the Thættir of Nornagest and of Sörli respectively, though into the former has entered a considerable element of folk-tale which is introduced with a certain naïveté and no little skill alongside the old heroic legends. As has been already mentioned, these three sagas, like others of the same type, appear to have been written down in the late thirteenth or the early years of the fourteenth century. On the other hand most if not the whole of the Saga of Hromund Greipsson appears to have been composed early in the twelfth century, but we do not know when it was first written down.

A fourth stage is represented by the Icelandic Rímur which are for the most part rhyming metrical versions of the sagas and which date from the fourteenth and fifteenth centuries. As an illustration of this stage I have translated a few stanzas from the Gríplur, a Ríma based on an early form of the story of Hromund Greipsson¹⁵. The Rímur are, so far as we can judge, somewhat wearisome paraphrases of the prose stories, and while the metre and diction are elaborate in the extreme, the treatment of the story is often mechanical and puerile. Comparatively few of the Rímur have as yet been published and the Gríplur is the only one known to me which is primarily concerned with any of the sagas contained in this volume.

The ballads, both Faroese and Danish¹⁶, belong to a fifth stage in the life of heroic legend in the North; but their origin and history is by no means so clear as that of the Rímur, and it is at present impossible to assign even approximate dates to more than a few of them with any degree of certainty. I have touched on this question at somewhat greater length below¹⁷; and I would only add here that some Danish and Swedish ballads, e.g. Ung Sveidal¹⁸, Thord af Haffsgaard¹⁹ and perhaps Her Aage²⁰, appear to be derived more or less directly from poems of the Viking Age, such as Fjölsvinsmál, Thrymskvitha and Helgakvitha Hundingsbana I—without any intermediate prose stage.

A careful study of the Faroese ballads as a whole might enable one to determine something more of the relation of ballads to 'Literature'²¹ and of the various ballad forms to one another, such as that of the short and simple Ballad of Hjalmar and Angantyr to the longer and more complicated Ballad of Arngrims Sons. Simplification and confusion are among the chief characteristics of popular poetry²²; but it is to be noted that in the case of the Hervarar Saga confusion set in long before the days of the ballad—as early as the saga itself, where there must surely be at least one case of repetition of character²³. In reality, considering through how many stages the ballad material has passed, one is amazed at the vitality of the stories and the amount of original groundwork preserved. A careful comparison of the Völsunga Saga and the Faroese cycle of ballads generally classed together as Sjúrðar Kvæði—which, be it observed, were never written down at all till the nineteenth century—brings out to a degree literally amazing the conservatism of the ballads on the old heroic themes.

Readers who desire to make further acquaintance with the 'Stories of Ancient Times' as a whole will find a further account of the subject in Professor Craigie's Icelandic Sagas (p. 92 ff.). More detailed accounts will be found in Finnur Jónsson's Oldnorske og Oldislandske Litteraturs Historie²⁴, Vol. ii, pp. 789-847, and in Mogk's Geschichte der Altnordischen Literatur in Paul's Grundriss der Germanischen Philologie, Ed. ii, 1904, Vol. II, pp. 830-857, while a discussion of the heroic stories will be found in Professor Chadwick's Heroic Age, chs. i-viii. For a full bibliography of the texts, translations, and general literature dealing with the Fornaldarsögur collectively, see the annual Islandica, Vol. V, pp. 1-9, compiled by Halldór Hermannsson and issued by the Cornell University Library, 1912.

Footnote 1: Cf. Saxo Grammaticus, Dan. Hist., Book