But whatever may be the real origin of the popular tradition, the incontestable fact remains, that once upon a time the peak-shaped fell, now called Burg-work (Borgarvirki), towering to the height of some 800 feet above the level of the sea between the two steads of Mickle-Burg (Storaborg) and Little-Burg (Litla-Borg) in Willowdale, was transformed by the labour of man into a military fortress. We ourselves had an opportunity of visiting the work in our trip to Iceland in 1871, and to inspect the by no means inconsiderable fortifications thrown, in the shape of walls made of large flat slabs, across all clefts in the natural basaltic rock which offered access to the top, standing over four feet thick, and in some places as many as ten feet high. An interesting and minute description of the work is given by Dr. B. M. Olsen, a native of the neighbourhood, in "Arbok hins islenzka fornleifafelags 1880 og 1881," pp. 99-113, accompanied by a critical dissertation on the Burg-Work tradition, and he, a first-rate antiquary and scholar, comes to the conclusion that, since in the whole history of that country-side there is no event with which the really great works of fortification on the peak can be connected, unless it be Bardi's war with the Burgfirthers, we are not authorized at present to reject the existing tradition as utterly unhistorical.

The chronology of our saga has given great trouble hitherto. Its central date is, of course, the year of the Heath-slayings, which by some is placed at 1013, others at 1014 or 1018, and by the saga itself at 1021. Vigfusson declares in favour of 1014, relying on the statements of "Grettir's saga", "that the Heath- slayings befell in the autumn that Grettir spent in Iceland after his first journey abroad, but that year was 1014" ("Timatal", 460, cf. 473-474). He attaches particular weight to the evidence of the old Resenius' annals, which also place the Heath-fight in 1014.

At the time when Vigfusson wrote his "Timatal", he, in common with contemporary scholars, believed that the annalistic writings of Iceland were as old as the historical, and the dates of the former were independent of the latter. This opinion, which originated with the Northland annalist, Bjorn Jonsson of Skarthsa, in the seventeenth century, is radically refuted by Gustav Storm in his excellent edition of "Islandske Annaler indtil 1578", where a whole array of evidence is brought together to show, that annalistic writing in Iceland could not have begun till a few years before 1300. For the saga period, therefore, the evidence of the annals has no real weight, since their dates depend on the evidence of the sagas themselves, according as the annalists were able to reason them out in each particular case. In this instance, thus, the evidence of Resenius' annals falls through as worthless, since evidently it depends on Grettir's saga. But what does that saga's evidence amount to?

In chapter xxviii we are told that Grettir came on a visit to his kinsman and former superior playmate, Audun of Audunstead in Willowdale, and let loose his horse to graze in the home-mead "where the grass was highest" (lothnast, highest and thickest). This visit then happened in June, before the mowing of the home-mead began; mowing of home-fields having at all times in Iceland begun, in ordinary years, at the end of June or in the first week of July. Grettir, wanting to square old scores with Audun, falls to wrestling with him, in the midst of which scuffle Bardi arrives and separates the wrestlers. Grettir now offers Bardi to join his expedition, "for I have heard that thou art bent on going south to Burgfirth this summer." Bardi accepted the offer gladly and (chap. xxxi) rode home to Asbiornsness, and then to his foster-father, "who gladly received him, and asked what he had earned in the way of helpful following," etc.

This statement of Grettla's we can pronounce at once as false. It is invented on the basis of the Heath-slayings' story; but as we know it now, at least, there is no mention made in it of any meeting between Bardi and Grettir at any time, much less of Thorarin's disapproval of Bardi's engagement of Grettir, which in "Grettir's saga" is circumstantially related, and Thorarin's harangue kept exactly in his wary, half-pious vein and anxious care not to spoil his fosterling's chances by the admission into his band of any whose fetch was one of lucklessness. It would be incomprehensible how such an incident could ever have dropped out of the Heath-fight's story having once got into it. But there are more serious objections to be noted. Grettir could not possibly have heard rumours in June or July of that which was not resolved upon till "seven weeks were left of summer," i.e., the latter end of August, and then in strict secrecy, no one knowing the least about it till the Sunday, when six weeks were left of summer, that Bardi broke the secret in the folk-mote at Thingere. That Bardi, therefore, as the Grettla clearly gives to understand, should have been abroad recruiting his force in June or July, is out of question, of course. Why, the whole plot of the Heath-slayings' story turns really on one hinge, namely, the observance of absolute secrecy as to Thorarin's intentions, until they could be carried out in a shorter time than it would take the rumour of them to cross the mountains. This statement of Grettla, therefore, which hitherto has served as a key-stone of the chronology of our saga, is in itself of no worth, being a mere fabrication. If it should happen to relate to the right year, it would be by accident only.

Now the landmarks of time that our story itself supplies are the following: the year that Bardi was outlawed at the Althing he went abroad, but was shipwrecked on the northern coast of Iceland, and spent the winter with Gudmund of Maddervales (Mothruvellir) in Eyiafirth; the next winter he was in Norway; the next to that in Denmark, and in the following summer he set sail for Iceland, arrived on the north coast, and -- "By this time Gudmund was dead." Now the year of Gudmund's death was 1025; so, counting back these years of Bardi's outlawry, we see that he was in Denmark, 1024-1025, in Norway, 1023-1024, at Maddervales, 1022-1023; consequently the Thing at which he was outlawed was that of 1022, and the Heath-fight accordingly befell in 1021. Against this evidence of the saga itself Grettla's fictitious statement goes for nothing, of course. Vigfusson is by no means indifferent to these chronological facts, though he does not, on account of the great importance he attaches to Grettla's evidence, see his way to accept them. And it cannot be denied that a variety of difficult points is raised by accepting the evidence of our story. But to disallow it, considering that we have to deal with the oldest Icelandic saga, preserved in the oldest of all the saga vellums from Iceland, is obviously contrary to all rules of sound criticism. However, the whole question requires fresh overhauling, which it would be idle to attempt within the limited space of a preface to a translation of the saga.