6. Straumsfjord, with its island and strong currents, is too circumstantially described to be an invention. 7. The topographical characteristics of Hóp, apart from the meaning of the name, which seems to be a land-locked tidal estuary, are confirmed by the evidence of both independent versions. We must therefore accept its main features—extensive shoals, and a river running through a lake into the sea. These then are our points of departure. To these we may safely add, as a general rule, points as to which the independent versions agree. The savages, though equally well authenticated, and valuable as evidence of the general truthfulness of the story, are not included, since, whatever the opinion we ourselves have formed, it may still be considered arguable by some that they were Eskimo. In any case they do not help us to fix any situation more closely than our other data. If they were Indians they might occur anywhere within the area of our inquiry, if Eskimo they cannot carry Wineland with them north of the 49th parallel, or away from the vines from which it derived its name. Their existence, if established, would only prove a more southerly migration of the Eskimo than has been hitherto generally accepted. _The Labrador Theory._ In spite of all this some writers have strenuously maintained that the full scope of all the voyages recorded should be confined to the Labrador coast. These are not generally to be found among those who have specialized on the subject. They are more usually those who, like Weise (_Discoveries of America to 1525_), deal with the matter incidentally, as part of a wider historical study. Their view, for the most part, seems to be connected with a sceptical attitude towards the sagas as a whole. It is, indeed, independent of the story except in so far as this supplies some corroboration of the bare fact that the Norsemen discovered America. Its advocates mainly argue on independent grounds that bold sailors like the Norsemen, having got so far as Greenland, must occasionally have been driven to Labrador. Nothing that is recorded of Wineland can really be brought into line with such theory, except possibly the skrælings, who are made the most of for that purpose with very inconclusive results. The ‘eyktarstad’ observation (see previous chapter, p. 211), a most circumstantial point in the story, rules out the whole of Labrador. The climate, too, is altogether inappropriate, and, of course, the vines and corn become an absurdity. Apart from these things one may ask where, on the Labrador coast, we are to find three distinct land-forms, with wholly different characteristics, and separated from one another by days of open sea.