What was apparently in the minds of those who made the conjecture referred to was a passage in Landnáma (i. 22) which tells how one Ari Marsson was driven by storms to ‘Hvítramannaland, which some call Ireland the Great; it lies westward in the ocean near Wineland the Good: it is called six days’ (dægra) sail west from Ireland: Ari did not succeed in getting away from thence, and he was baptized there. This story was first told by Rafn the Limerick-farer, who had been long at Limerick in Ireland. Thorkel Gellison (uncle of Ari the Learned) stated that Icelanders say, who had heard it from (Earl) Thorfin in the Orkneys, that Ari had been recognized in Hvítramannaland, and did not succeed in getting away from it, but was held in great honour there.’ In the Eyrbyggja Saga a similar story is told, though the name of the strange country is omitted, of one Björn Asbrandsson, who was cast in the same way upon a land to the south-west of Ireland, where he was subsequently recognized by an Icelander named Gudleif Gudlaugson. This story does not appear to me sufficiently relevant to the subject in hand to warrant more detailed notice, though the curious will find ample mention of it in other works on the Wineland question. Apart from the irrelevance of these stories, those familiar with the laws of evidence will doubtless agree that lands where a hero is said to have made his final disappearance, reported as they must necessarily be on hearsay testimony, are on a very different footing from countries whose explorers returned to describe them in person. The only value—either one way or the other—of this passage from Landnáma lies in the mention of Wineland the Good as a place known and acknowledged to exist at a period long antecedent to the date of any extant manuscript of these voyages. The proximity of Hvítramannaland to Wineland is presumably a conjecture by the authors of Landnáma, who would naturally tend to connect with one another any unknown lands reported in a westerly direction. It seems to me highly improbable that Wineland found any mention in the original story told by Rafn from Limerick. At any rate, no one can be justified in basing an argument on the assumption that it did, as does Dr. Nansen,[91] when he says, in support of his argument that the Celtic imagination has played a large part in corrupting the traditions of Wineland, ‘Ravn must have heard of both Hvítramannaland and Wineland in Ireland, since otherwise he could not have known that one lay near the other.’