the same sort as that which has been roughly described; it can scarcely be doubted that the age which is exhibited in the Poem of the Cid, the Song of Roland, and the lays of the Crusaders (la Chanson d'Antioche, for instance), was similar in all essentials to the age we find in Homer and the Nibelungenlied. Servia, too, has its ballad-cycles of Christian and Mahometan warfare, which suppose an age obviously heroic. But it hardly falls in with our scheme; Servia, at this time, might have been expected to have gone well past its Heroic Age. Either, then, it was somehow unusually prolonged, or else the clash of the Ottoman war revived it. The case of Servia is interesting in another way. The songs about the battle of Kossovo describe Servian defeat—defeat so overwhelming that poetry cannot possibly translate it, and does not attempt it, into anything that looks like victory. Even the splendid courage of its hero Milos, who counters an imputation of treachery by riding in full daylight into the Ottoman camp and murdering the Sultan, even this courage is rather near to desperation. The Marko cycle—Marko whose betrayal of his country seems wiped out by his immense prowess—has in a less degree this utter defeat of Servia as its background. But Servian history before all this has many glories, which, one would think, would serve the turn of heroic song better than appalling defeat and, indeed, enslavement. Why is the latter celebrated and not the former? The reason can only be this: heroic poetry depends on an heroic age, and an age is heroic because of what it is, not because of what it does. Servia's defeat by the armies of Amurath came at a time when its people was too strongly possessed by the heroic spirit to avoid uttering itself in poetry. And from this it appears, too, that when the heroic age sings, it primarily sings of itself, even when that means singing of its own humiliation.—One other exceptional kind of heroic age must just be mentioned, in this professedly inadequate summary. It is the kind which occurs quite locally and on a petty scale, with causes obscurer than ever. The Border Ballads, for instance, and the Robin Hood Ballads, clearly suppose a state of society which is nothing but a very circumscribed and not very important heroic age. Here the households of gentry take the place of courts, and the poetry in vogue there is perhaps instantly taken up by the taverns; or perhaps this is a case in which the heroes are so little removed from common folk that celebration of individual prowess begins among the latter, not, as seems usually to have happened, among the social equals of the heroes. But doubtless there are infinite grades in the structure of the Heroic Age.

The note of the Heroic Age, then, is vehement private individuality freely and greatly asserting itself. The assertion is not always what we should call noble; but it is always forceful and unmistakable. There would be, no doubt, some social and religious scheme to contain the individual's self-assertion; but the latter, not the former, is the thing that counts. It is not an age that lasts for very long as a rule; and before there comes the state in which strong social organization and strong private individuality are compatible—mutually helpful instead of destroying one another, as they do, in opposite ways, in savagery and in the Heroic Age—before the state called civilization can arrive, there has commonly been a long passage of dark obscurity, which throws up into exaggerated brightness the radiance of the Heroic Age. The balance of private good and general welfare is at the bottom of civilized morals; but the morals of the Heroic Age are founded on individuality, and on nothing else. In Homer, for instance, it can be seen pretty clearly that a "good" man is simply a man of imposing, active individuality[2]; a "bad" man is an inefficient, undistinguished man—probably, too, like Thersites, ugly. It is, in fact, an absolutely aristocratic age—an age in which he who rules is thereby proven the "best." And from its nature it must be an age very heartily engaged in something; usually fighting whoever is near enough to be fought with, though in Beowulf it seems to be doing something more profitable to the civilization which is to follow it—taming the fierceness of surrounding circumstance and man's primitive kind. But in any case it has a good deal of leisure; and the best way to prevent this from dragging heavily is (after feasting) to glory in the things it has done; or perhaps in the things it would like to have done. Hence heroic poetry. But exactly what heroic poetry was in its origin, probably we shall never know. It would scarcely be history, and it would scarcely be very ornate poetry. The first thing required would be to translate the prowess of champions into good and moving narrative; and this would be metrified, because so it becomes both more exciting and more easily remembered. Each succeeding bard would improve, according to his own notions, the material he received from his teachers; the prowess of the great heroes would become more and more astonishing, more and more calculated to keep awake the feasted nobles who listened to the song. In an age when writing, if it exists at all, is a rare and secret art, the mists of antiquity descend after a very few generations. There is little chance of the songs of the bards being checked by recorded actuality; for if anyone could write at all, it would be the bards themselves, who would use the mystery or purposes of their own trade. In quite a short time, oral tradition, in keeping of the bards, whose business is to purvey wonders, makes the champions perform easily, deeds which "the men of the present time" can only gape at; and every bard takes over the stock of tradition, not from original sources, but from the mingled fantasy and memory of the bard who came just before him. So that when this tradition survives at all, it survives in a form very different from what it was in the beginning. But apparently we can mark out several stages in the fortunes of the tradition. It is first of all court poetry, or perhaps baronial poetry; and it may survive as that. From this stage it may pass into possession of the common people, or at least into the possession of bards whose clients are peasants and not nobles; from being court poetry it becomes the poetry of cottages and taverns. It may survive as this. Finally, it may be taken up again by the courts, and become poetry of much greater sophistication and nicety than it was in either of the preceding stages. But each stage leaves its sign on the tradition.

All this gives us what is conveniently called "epic material"; the material out of which epic poetry might be made. But it does not give us epic poetry. The world knows of a vast stock of epic material scattered up and down the nations; sometimes its artistic value is as extraordinary as its archaeological interest, but not always. Instances are our own Border Ballads and Robin Hood Ballads; the Servian cycles of the Battle of Kossovo and the prowess of Marko; the modern Greek songs of the revolt against Turkey (the conditions of which seem to have been similar to those which surrounded the growth of our riding ballads); the fragments of Finnish legend which were pieced together into the Kalevala; the Ossianic poetry; and perhaps some of the minor sagas should be put in here. Then there are the glorious Welsh stories of Arthur, Tristram, and the rest, and the not less glorious Irish stories of Deirdre and Cuchulain; both of these noble masses of legend seem to have only just missed the final shaping which turns epic material into epic poetry. For epic material, it must be repeated, is not the same thing as epic poetry. Epic material is fragmentary, scattered, loosely related, sometimes contradictory, each piece of comparatively small size, with no intention beyond hearty narrative. It is a heap