This also, like the story of Achilles, is fit for a stage in which the characters are left free to declare themselves in their own way, unhampered by any burden of history, any purpose or moral apart from the events that are played out in the dramatic clashing of one will against another. It is not vanity in an historian to look for the historical origin of the tale of Troy or of the vengeance of Gudrun; but no result in either case can greatly affect the intrinsic relations of the various elements within the poems. The relations of Achilles to his surroundings in the _Iliad_, of Attila and Ermanaric to theirs, are freely conceived by the several poets, and are intelligible at once, without reference to anything outside the poems. To require of the poetry of an heroic age that it shall recognise the historical meaning and importance of the events in which it originates, and the persons whose names it uses, is entirely to mistake the nature of it. Its nature is to find or make some drama played by kings and heroes, and to let the historical framework take care of itself. The connexion of epic poetry with history is real, and it is a fitting subject for historical inquiry, but it lies behind the scene. The epic poem is cut loose and set free from history, and goes on a way of its own. Epic magnificence and the dignity of heroic poetry may thus be only indirectly derived from such greatness or magnificence as is known to true prosaic history. The heroes, even if they can be identified as historical, may retain in epic nothing of their historical character, except such qualities as fit them for great actions. Their conduct in epic poetry may be very far unlike their actual demeanour in true history; their greatest works may be thrust into a corner of the epic, or barely alluded to, or left out altogether. Their greatness in epic may be quite a different kind of greatness from that of their true history and where there are many poems belonging to the same cycle there may be the greatest discrepancy among the views taken of the same hero by different authors, and all the views may be alike remote from the prosaic or scientific view. There is no constant or self-consistent opinion about the character of Charles the Emperor in old French poetry: there is one view in the _Chanson de Roland_, another in the _Pèlerinage_, another in the _Coronemenz Looïs_: none of the opinions is anything like an elaborate or detailed historical judgment. Attila, though he loses his political importance and most of his historical acquisitions in the Teutonic heroic poems in which he appears, may retain in some of them his ruthlessness and strength; at other times he may be a wise and peaceful king. All that is constant, or common, in the different poetical reports of him, is that he was great. What touches the mind of the poet out of the depths of the past is nothing but the tradition, undefined, of something lordly. This