reference in this connection; and it appears the more unlikely that any map of the American discoveries was drawn before “the age of written history.” The narratives or logs of the expeditions, handed down with remarkable particularity in the stories of the saga tellers, conveyed the geographical circumstances of these voyages with sufficient precision to satisfy their audience and to make a map even more supererogatory, and indeed inconsistent with this mode of transmission, than a written transcript. We have also to remember that the Norse voyages to America, unlike those to Greenland and along its coasts, were isolated episodes, covering only a brief span of time, and that (so far as we know) they were not followed up by any regular navigation.

But a map, like other graphic records, cannot be substantially communicated by word of mouth. Even an observer, if he draws from visual memory, and without the aid of sketches made on the spot, a map of the coasts which he has traversed, must generalize and formalize his outlines; and this process necessarily plays a much greater part in the construction of a map by someone who does not carry in his mind’s eye an image of these coasts from direct visual experience. If (as we think probable) no map illustrating the American landfalls of the Norsemen was drawn before the twelfth century at the earliest—that is, long after the death of the adventurers who made them—its chain of transmission must have been verbal, not graphic. In these circumstances we could expect from such a map no more than a general indication of the coasts in question and of their most conspicuous features, with the data of direction from known lands, of location, of trend, orientation, and form stated with no greater precision than a verbal relation, whether spoken or written, could offer. This hypothesis will be tested by scrutiny of the outline in the Vinland Map.

The two principal sources for the story of the Vinland voyages, namely the Saga of Eirik the Red and the Tale of the Greenlanders, appear to represent (respectively) Icelandic and Greenlandic versions of the Norse discovery of America. Their divergences in recording the sequence of events are profound and various. With the detailed arguments respecting the degree of authority to be ascribed to the two accounts we need not concern ourselves here. As a working assumption, we shall (with William Hovgaard and Gathorne-Hardy) take it that both narratives contain authentic elements, and we may extract the geographical information which they furnish on the discoveries.

In A.D. 986 Bjarni Herjolfsson, sailing from Iceland for Greenland, is driven to the south of his course by “the persistent northerly and easterly winds which often prevail in the North Atlantic during the early summer months.” He sights land, wooded and not mountainous, and knowing himself to be south of his latitude, turns northward, leaving it to port; after two days’ sailing he comes to another land, flat and wooded; sailing

out to sea for three days before a southwesterly wind, he comes to a third land, mountainous and with ice upon it, which he takes to be an island; and the same wind, strengthened in force, brings him in four days’ sail to Cape Herjolfsnes in the south of Greenland.

Fourteen years later Bjarni goes to Norway, where he tells his story. On his return to Greenland in A.D. 1001, he sells a ship to Leif Eiriksson for a voyage to the lands discovered in the west. Leif, sailing the other way along Bjarni’s homeward track, comes to the country which Bjarni found last, a flat, rocky coast with glaciers behind, which Leif names Helluland; and then to Bjarni’s second land, low and wooded with wide expanses of white sand, which he calls Markland. Sailing out to sea for two days, he discovers another mainland, with an island north of it, and, entering “the sound which lay between the island and the cape which ran north from the mainland,” he runs west past the cape; after running aground, the Norsemen take their ship up a river and into a lake, where they erect houses. Here “day and night were more equally divided than in Greenland or Iceland,” as they note from observation of azimuths of the midwinter sun in the forenoon and afternoon. If these bearings refer precisely to sunrise and sunset, they indicate a northern limit of roughly 50° latitude for the position. The country has no frost in winter and bears vines, so that it is given the name Vinland. After wintering there, Leif returns “with a fair wind” to Greenland in the next spring.

In the following year (1004) Thorvald Eiriksson, Leif’s brother, takes a ship to Vinland and winters in Leif’s old camp. In the next summer the country to the west is explored by boat and found to be wooded, with white sands and many islands. The summer after, Thorvald sails east “and along the more northerly part of the country,” to a cape which he names Kjalarnes (Keelness) and eastward to fjords and a headland, where Thorvald is killed in an affray with natives and buried, so that the headland is called Krossanes (Crossness). After the winter his crew returns to Greenland.

A year later (1008) Thorstein Eiriksson sets out for Vinland in his brother Thorvald’s ship, but is carried by weather far out into the Atlantic and finally reaches the Western Settlement in Greenland, where he dies.

About the year 1019 the Icelander Thorfinn Karlsefni comes to Greenland, where in the next year he marries Thorstein’s widow and sets out on an expedition with two ships to make a settlement in Vinland. They sail north to the Western Settlement and the Bjarneyar (Bear Isles), and after two days at sea with a north wind they come to land with many flat stones and with arctic foxes; they thought this was Leif’s HeIluland. Sailing thence for two days, still

with a north wind, and changing course “from south to southeast,” they reach a wooded land, which they call Markland, with an island offshore. Coasting southward for two days, they come to Keelness, naming the long sandy beaches Furdustrands; and while they lie at anchor in one of the bays, grapes and self-sown wheat are found. Sailing on, they come to a fjord with an island and strong currents, which they name Straumsfjord; here they winter. In the spring Thorhall, a member of the expedition, takes one ship north in search of Leif’s Vinland. Karlsefni with his companions in the other ship coasts southward “a long time” until they come to a river flowing through a lake across a gravel bar to the sea; on this estuary, which they call Hop, they find vines and self-sown wheat. After wintering there, they return to Straumsfjord, where they pass the third winter before returning with a south wind to Markland and Greenland.

Next year (ca. 1024) Freydis, Leif’s sister, goes with two ships to his camp in Vinland.

Such, in barest outline, are the geographical data which the narratives of the Vinland voyages yield and which have been the subject of widely differing interpretations and identifications.

As to the discovery of three distinct land masses or sections of continental coast, Hovgaard prudently observes that “different names may have been given to the same land, and the same name may have been applied to different lands by different explorers. Thus … the Markland of one expedition may have been the Vineland of another, and the Helluland of one expedition may not have been the Helluland of another.”

The conditions of climate and vegetation ascribed to the three lands, while by no means consistent, indicate certain limits of latitude. Thus the wild vines found by Leif in Vinland and by Karlsefni at Hop would today locate these regions south of Nova Scotia, as would the reference to winters free of frost; but evidence of this kind must be interpreted in the light of climatic changes, and allow for the milder climate which (as meteorologists agree) prevailed in the eleventh century.

One must also consider the description of these lands and of their prominent features. Helluland is stony; Markland low and wooded, with beaches; Furdustrands has long sandy beaches; Keelness is a northpointing cape, with a sound or bay on the west; Straumfjord has strong currents; Hop is a landlocked tidal estuary. It is the difficulty of reconciling these physical attributes with the other circumstances of the Norse discoveries that has occasioned the strongest conflict of opinion regarding the identification of the lands named in the narratives. The possibility that, in the process of transmission, details may have been transferred from one voyage to another, or even introduced into the saga narratives from some extraneous source, cannot be disregarded; and this gives a certain air of unreality to much of the debate. (In Mark Twain’s words, “the researches of many commentators have

already thrown much darkness on this subject, and it is probable that, if they continue, we shall soon know nothing at all about it.”)

It must be admitted that the Vinland Map reproduces very little of the geographical detail noted in the narratives of the Norse voyages. We find only the outline of a single land mass, divided into three by two deep inlets and the prominent capes, pointing northward or northwestward, on the south of their entrances. The narratives furnish no authority for the combination of the three lands discovered into a single great island, nor for the tracing of a continuous east coast, still less of a west coast. Are we to suppose these to be cartographic constructions evolved by the author of the map to justify his theoretical concept of Vinland?

We may observe on the map that Iceland is laid down a little north, and Greenland a little south, of the true positions. If these necessary corrections are made, it is not difficult to reconcile the relative positions of Atlantic lands in the map with the navigational data of the Norse voyages. We can see, in the first place, how Bjarni, sailing from Iceland and driven south of his course for Greenland by northeasterly winds, inevitably made a landfall on the American coast perhaps as far south as Nova Scotia, perhaps on Newfoundland or Labrador; his second and third landfalls lie northward of his first, and thence a southwesterly wind brings him, doubtless on a course more east than north, to Cape Herjolfsnes. Leif’s course to Helluland and Markland follows that of Bjarni in reverse; and, after crossing a sound and passing a north-pointing cape to the east of it, he comes to the land which he named Vinland. This topographical description is plainly reflected in the delineation of the more southerly inlet and cape in the Vinland Map. This cape perhaps also corresponds, in the cartographer’s mind, to the Kjalarnes of Thorvald’s and Karlsefni’s voyages. Karlsefni adopts a different outward course; and if the “Bjarneyar” of his narrative correspond to the similarly named islands of the old chorography of Greenland already cited, he may have gone as far north as Disko Island (69°-70° N) before turning his ship’s head westward. Hovgaard’s suggestion that the open coastal channel had to be followed “at least as far as the Western Settlement” before Karlsefni could find a passage through the East Pack neglects the probability that at this date Davis Strait was free of ice. With a northerly wind, and thus (we may suppose) on a southwest or westerly course, Karlsefni crosses Davis Strait and picks up land in two days. If we discount the map’s longitudinal error, these courses agree reasonably enough with the delineation of Greenland and the most northerly section (Helluland) of Vinlanda Insula .

Thus, if we were able to take the delineation in the map “at its face value,” by supposing its scale to be equivalent to that of

the European coasts, correlation of the American lands depicted in it with the geographical conditions prescribed by the written sources would justify the identification of the two large inlets as Hudson Strait or Ungava Bay and as the Gulf of St. Lawrence, and of the three land sections respectively as Baffin Land, Labrador (perhaps with Newfoundland), and the Maritime Provinces of Canada.

The delineation of Vinlanda Insula in the map does not, in our view, derive from a cartographic model drawn from experience. The general appearance and conventional style of the outlines, the elements evidently born of theory or conjecture, and plainly contrasting with a few features which are clearly and boldly depicted as they are described in the narratives; the want of particularity elsewhere and the omission of numerous other geographical details mentioned in the texts—all these aspects of the design persuade us that it owes its ultimate, if not immediate, origin to a graphic reconstruction of the geography of the voyages, compiled from saga accounts or from hearsay, and generalized in transmission.

That the author of the Vinland Map made use of a map prototype compiled in this way is of course very probable. The prototype, whose maker must have enjoyed access to the texts of the saga narratives, was- with little less probability—drawn in Iceland.

There are, as Mr. Skelton points out at length, many differing identifications offered by scholars as to just what actual places Helluland, Markland, and Vinland may have been.

We need not refer here to the supposed Norse remains discovered in North America—the Dighton Rock, the Kensington Stone, the Newport Tower—nor to the hypotheses founded on them regarding the extent of the Norsemen’s travels. It is within possibility that archaeological investigation may uncover betterauthenticated relics of Norse exploration or settlement and so help to define the regions of discovery more confidently and precisely than is at present permissible.

In 1960, in fact, the Norwegian explorer Helge Ingstad identified as Norse a habitation site on the north coast of Newfoundland, in about 51° 36′ N and 55° 32′ W, some five miles west-southwest of Cape Bauld. The site, near the hamlet of L’Anse-aux-Meadows, is on an old beach terrace, one hundred yards from the sea in Épaves Bay, on the east side of Sacred Bay, and the excavations carried out by Dr. Ingstad in 1961, 1962, and 1963 have uncovered the foundations of seven turf buildings, with indications of smith’s workings. The largest structure measures 60 x 45 feet (thus much bigger than any Eskimo building), and one of its five rooms is a large hall with an open stone hearth. Carbon 14 tests of charcoal samples are reported to yield dates from the seventh to the eleventh centuries, with a cluster about A.D. 900, and there are signs that the site was subsequently visited and stripped by Eskimos. It is understood that, at the time of my writing, Dr. Ingstad is preparing his report for

publication. Canadian archaeologists seem to accept that the settlement cannot be Eskimo but “to reserve judgment on whether it is indeed Vinland,” as claimed by Dr. Ingstad.

Such caution is certainly justified. On the one hand, it appears possible that Norse sites have hitherto gone unrecognized or unrecorded, whether along the American mainland to the south and west, or on the coasts of Labrador, Ungava Bay, and Baffin Land to the north. On the other hand, the saga accounts mention several wintering-places of the Norsemen, and it is hardly to be expected that, even if the Viking origin of the settlement at L’Anse-aux-Meadows can be established, it will yield any evidence to determine which expedition established itself there, and when. The location of the site, on the west side of Cape Bauld and at the entrance of Belle Isle Strait, together with the configuration of the adjacent coasts, does indeed call to mind the north-pointing cape, lying as a wedge between the ocean on the east and a sound on the west, where Leif built his winter quarters in Vinland. It recalls, no less, the Kjalarnes of Thorvald and of Karlsefni, and may suggest that the settlement at L’Anse-aux-Meadows was the work of the Karlsefni expedition (thus supporting Hovgaard’s conjectural geography for this voyage).