As has been pointed out in the chapter on the Flatey Book (p. 137), the expedition of Thorfin Karlsefni must have followed those of Leif and the rest of Eric’s family at a considerable interval of time. Though this has not been generally realized, it is not a mere matter of opinion, but rests upon cogent and conclusive evidence when once the known points of chronology are closely examined. Apart from this, it is evident on consideration that it would involve a very curious coincidence if Karlsefni arrived in Greenland exactly at the time when the efforts of Eric’s sons at exploration were exhausted. It is therefore far more unlikely in the case of Karlsefni than in that of Thorvald, assuming the latter to have conducted an independent expedition, that the landfalls were the same as those made by Leif. If we accept Hauk’s version of the story, Leif’s voyage took place in a.d. 1000, and in any case it cannot have been many years later, while 1020 is as early as we can reasonably place Karlsefni’s expedition. For this reason, apart from any others, it is right to assign to this voyage a separate chapter and independent consideration.
Karlsefni’s starting-point, we are told, was not from the neighbourhood of Eric’s home at Brattahlid, but from the Western Settlement (Godthaab), and the ‘Bear Islands’. The latter name was apparently applied to Disko, far to the northward, but it is difficult to suppose that Thorfin sailed so far in the opposite direction to his objective. It is more probable that the name refers to some islands in the immediate neighbourhood of Godthaab. One has only to remember the frequent occurrence of such local names as Bjørnuren, Bjørnlien, in Norway, to realize that nomenclature of this character is often repeated, indeed one need not go outside this saga for an instance of such a repetition (in the neighbourhood of Markland).
Possibly the Western Settlement was visited for recruiting purposes. The visitors from Iceland, as we are told, only accounted for 80 men out of the 160 eventually taking part in the expedition; the original Icelandic crews, after a winter in Greenland, would probably need to be brought up to strength, and the better part of 100 volunteers must have been difficult to collect in so small a colony.[100] Mr. Babcock, p. 97, seems to think that the shortest way to Labrador via the north was already known in Greenland, and he also, curiously enough, considers it the safest route. On the question of danger there is room for difference of opinion, but it may be pointed out that progress from north to south or vice versa is frequently impeded by ice till a late date in the summer. The very slow Moravian mission ship, sailing from London, often reaches the stations on the Labrador coast before the Newfoundland steamer service, since, sailing from east to west, she travels across instead of along the ice-barrier. Karlsefni’s ultimate and principal objective being to the south, he would hardly have deliberately undertaken so dangerous, unexplored, and roundabout a course, even if he had known of the possibility, which seems extremely doubtful. As a basis for calculation we may therefore safely put the point of departure in the neighbourhood of Godthaab.
From this point we are told that the expedition sailed for two days with a north wind, i.e. in a southerly direction. It should be pointed out that the map occurring opposite page 106 of Mr. Babcock’s treatise is very misleading as to the courses which it suggests. It contains no meridians, and is tilted westward at an angle of nearly 40 degrees, with the result that the Western Settlement of Greenland is brought almost exactly north of the neighbourhood of Nain on the Labrador coast, which is the point selected by the author for Karlsefni’s landfall in Helluland. As a matter of fact there are not far short of 10 degrees of longitude between the two places, and the course between them is very far to the west of south. Mr. Babcock appears to have chosen this point on the coast of Labrador in order to retain the statement made as to the voyage having occupied but two days. The distance being about 450 miles, the author is compelled to assume a speed of nearly ten miles an hour, in support of which he cites statistics as to the speed of yachts and other modern sailing vessels. Now, as we have seen in Chapter V, this seems far beyond the capacity of ancient Icelandic ships, and, since on this point we have definite evidence, it is impossible that the time can have been correctly stated, even if we suppose the very nearest point on the Labrador coast to have been the land first sighted. It is moreover difficult to suppose that Karlsefni made the nearest point; he had no clue to its position, and his ultimate objective, for which he had a guide in the directions of his predecessor, Leif, lay far to the south.
Nor is a long coasting voyage along the shores of Helluland in any way suggested by the text; in fact it is inconsistent with it. In the summer, or still more in the spring, Karlsefni would almost certainly have been greatly impeded by ice off the Labrador coast, but no mention is made of any such feature. We must therefore either abandon the figure, two days, altogether, which—having regard to its repetition later on—is possibly the right course, or we must substitute some plausible alternative. Reeves suggests ‘sjau’ (seven) for ‘tvau’ (two), but in the manuscripts numbers seem to be usually given in figures. A possible amendment would be five (u), as, if the light stroke connecting the verticals in writing this figure had become erased by time, íí and u would be almost identical in Icelandic manuscript. This would be equivalent to 750 miles at average speeds, and would bring land more nearly to the south of the starting-point well within range.[101] It is, however, safer on the whole to decide that we have no reliable guide to the distance.
The question of the situation of Karlsefni’s landfall in Helluland has been already discussed (Chapter VI, p. 230), and we can only adhere to the conclusion there arrived at, viz. that there is a slight balance of probability in favour of Labrador as against Newfoundland, but that both countries would almost certainly have been assumed to be one and the same. Anyone who doubts this probability has only to look at the maps reproduced on p. 364 of vol. 2 of Dr. Nansen’s In Northern Mists, where the same confusion is shown to have been made in the case of Corte Real.
The question of Markland has also been treated at an earlier stage, and the improbability of the south-easterly course on which the identification of this country with Newfoundland mainly depends has been pointed out. Whatever theories we adopt as to the situation of the various lands, it is clear that the courses given in the Saga of Eric the Red and Hauk’s Book must at some point be abandoned. For example, Storm identifies the coast of Nova Scotia with that followed by Karlsefni after arrival at Keelness. The lie of this coast is a great deal nearer west than south, which is the direction given, and the same applies to the coast of New England after passing Cape Cod, which seems to be the alternative. A uniform southerly course is excluded. Again, two days of open sea from Newfoundland to Cape Breton, or from Cape Sable to Cape Cod, especially the former, would indicate a westerly rather than a southerly course for the expedition. If, on the other hand, we assume the explorers to have coasted Newfoundland to Cape Ray, the course to Nova Scotia is corrected at the expense of the distance. The upshot of all this is that, as already indicated, a course given in this version of the saga is a most unsatisfactory piece of evidence on which to found an important conclusion. Moreover, Eric’s Saga is silent as to this deflexion to the south-east, which consequently rests upon Hauk’s unsupported authority. This editor may merely have thought that, as the island next mentioned lay to the south-east, such a course was necessarily implied.
On this island off the shore of Markland to the south-east we are told that the explorers killed a bear, conferring in consequence the name Bjarney (Bear Island) on the place in question. It has been generally assumed that this must necessarily mean a polar bear. But Karlsefni was acquainted with Norway, where the European bear still exists and must then have been common, so that one would think that a bear which was not white would equally be called a bear. I would further suggest that this would be the case even if no bears other than the polar species had previously been known to members of the expedition. But secondly, supposing a polar bear to be meant, there does not seem any violent improbability in the idea that one should be found, in the eleventh century, so far south as Nova Scotia. At a far later date, Arctic fauna had a much more southerly habitat than at present. Walrus were regularly hunted in the Gulf of St. Lawrence, as is shown by a number of passages in Hakluyt. As to the polar bear itself, its restriction to its present northerly habitat appears to be even more recent. In Labrador, as far south and inland as Eagle River Falls, Sandwich Bay, Captain Cartwright records in his diary under date July 22, 1778: ‘Numbers were in sight. I counted thirty-two white bears, and three black ones; but there were certainly many more.’ In earlier days Jacques Cartier found a polar bear between Newfoundland and the Funk Islands, while both Cabot and Corte Real found the same animal on what was probably Newfoundland, and cannot certainly have been far north of it. It may further be pointed out that white bearskins are mentioned more than once in the Algonquin Legends of New England and Nova Scotia, collected by C. G. Leland. As to bears on islands, whether white or black, Cartier found them on Brion Island, so there is no improbability in this feature. If a polar bear is meant, Sable Island seems a possible location for Bjarney, but in any case there are many islands off the Nova Scotian coast which would fulfil the conditions.
Until, however, the expedition reaches Keelness, we are on very uncertain ground, and it would be imprudent to insist upon any definite conclusion. We may in fact, at this stage, so far as our information hitherto has taken us, be either at the north-eastern extremity of Nova Scotia or in the vicinity of Cape Cod, according as our identification of Helluland and Markland agrees with Storm or otherwise. We may however fairly say that the choice lies between these two localities. Any other theory breaks down at the first touch of criticism.
But when the description of Keelness given in the saga is compared with what we know of the Nova Scotian coastline, one meets at once with a very formidable objection to Storm’s theory. For here began Furdustrands, the Wonderful Beaches, so called from their great length, and thus described:—‘It was a desolate place, and there were long beaches and sands there.... They gave the beaches a name, calling them Furdustrands, because the sail past them was long.’ It appears too, as already hinted, that this feature was sufficiently marked to give rise to the application of the name to a large district, extending at least to Straumsfjord (cf. second song of Thorhall the Hunter, see also p. 227).
Now the coast of Nova Scotia cannot, to an unprejudiced eye, be said to comprise any continuous beach of a really remarkable length. On the contrary, it is both indented and rocky. Mr. Babcock clearly sees this difficulty; his remarks on the subject have been already referred to (p. 234). He requires a continuous stretch of at least 100 miles for Furdustrands, and this estimate compares favourably with those put forward by Storm and most of his adherents. Now, to meet the objection which is here raised, Mr. Babcock postulates a rise in the Nova Scotian coastline since the eleventh century sufficient to account for what is otherwise a fatal discrepancy in its present appearance. He frankly admits that there is no direct evidence of such a phenomenon, and indeed that ‘locally there is some scientific opinion that this probably has not occurred’. But this is not the most that can be said. In the first place, the early explorers who followed on the rediscovery of the country found the coast exactly as it is to-day. The upheaval postulated must therefore have taken place, if at all, within an even shorter period than that allowed by Mr. Babcock. Thus Champlain writes: ‘All the coast which we passed along from Cape Sable to this place (Canso) is moderately high and rocky, in most places bordered by numerous islands and breakers.’ Of Cape Breton Denys says (Green Island to Louisburg), ‘All the coast is nothing but rocks.’ Thenceforward ‘nothing but rocks’ is a phrase constantly repeated, but one looks in vain for any mention of a beach. Later on, ‘leaving there (St. Ann’s harbour) and going to Niganiche (Ingonish) one passes eight leagues of coast having shores of rock extremely high and steep as a wall.... Niganiche is not a bit better.’ Similarly right on to Cape North. We have not much room left for these long and wonderful beaches, which so struck the Norsemen immediately on their arrival at Keelness, and which were so impressively long to sail past. It is true, as we have seen, that there are white sands near the south-western end of the peninsula, but the numerous indentations break up the coastline, and besides, the description requires a cape facing a ship approaching from more northerly latitudes.
In the second place, had there been such a change as that suggested by Mr. Babcock, at so recent a date, there must necessarily have been positive geological evidence of it. When a beach rises from the sea, particularly if it be of such great extent as is required in the present case, traces of the former sea-level remain, in the form of raised beaches, water-worn rocks, or remains of marine fauna. In Nova Scotia such things are found indeed, but dating from a period far antecedent to that with which we are at present concerned. The formation appears to be contemporaneous with the existence of some form of mammoth, whose remains have been found, and in many places the course of these beach-deposits is cut through by river valleys which have been formed since. (See Dawson, Acadian Geology.) Now if these vestiges, dating from a period antecedent to the existence of human remains, are still to be traced, it is clearly impossible that no evidence should survive of what is alleged to have happened at a date which is, geologically speaking, yesterday. Mr. Babcock’s theory must accordingly be abandoned, in spite of his careful, ingenious, and elaborate argument, and, this being so, we are still faced with an insuperable difficulty in the way of associating Furdustrands with Nova Scotia.
Now let us turn to the alternative suggested, and consider Cape Cod to be Keelness. Karlsefni has now indeed been brought to a coast meriting the name bestowed, ‘a desolate place, with long beaches and sands.’ Not only does the Cape Cod or Barnstable peninsula, as Horsford saw, comply with the description, but beyond this point the name Furdustrands might appropriately be applied to nearly the whole Atlantic coastline of the United States. Passing the shores and sand-hills of Cape Cod and Monomoy, from Chatham at the heel of the promontory to Nobska Point at the entrance to Buzzard’s Bay, the coast, as the United States Pilot describes it (p. 341), ‘is low and sandy.’ If the course lay to the south of Nantucket and Martha’s Vineyard the same description would apply. Here there is a slight break formed by the indentations of Buzzard’s and Narragansett Bays, but the former is masked from the sea, until passed, by the Elizabeth Islands, and the latter by the islands at its mouth; the prospect throughout was an unattractive one, and these bays, from the sea, might easily pass unnoticed, while from Point Judith, west of Narragansett Bay, to the entrance of Long Island Sound, by Watch Hill Point, there is still, in the words of the Coast Pilot, ‘a low beach with lagoons inside and higher wooded land at the back.’ Until arrival at Straumsfjord, no attempt was made to land. As Mr. Babcock notices, the episode of Hake and Hekja is obviously an interpolation from another source, as no vines had been found up to the time of Thorhall’s versified comments on the subject. We have consequently to look for but one indentation, that presented by Straumsfjord. The answer to Mr. Dieserud’s objection to Cape Cod that the wild grape flourishes there close to the sea, and must therefore have been found, is that no landing was made there.[102] Karlsefni, unattracted by the prospect, sailed on, and consequently the discovery was deferred. ‘They went their ways, till interrupted by a fjord’, so one might almost interpret the language of the saga. In any case, the likeliest fjord to attract attention on a coasting voyage would be one lying right in the track of the ship. And such a fjord, if my conjecture is right, was Straumsfjord.
Dead in the course of a ship following the coast westward from Cape Cod lies Long Island Sound. Though not, strictly speaking, a fjord, it has, until the East River channel, leading to New York, is explored, exactly the appearance of one. It is very narrow at each end, and its greatest breadth, fifteen to sixteen miles, is only maintained for about twenty miles in its central part. Until the sound was explored by Adriaan Block in 1614, it was probably not known that Long Island was separated from the mainland.
At the mouth of the sound is an important island, Fisher’s Island, with an extreme length of six miles, between which and the less important Gull Islands runs a strong tidal stream, appropriately known as the Race. This is sufficiently formidable to necessitate the warning of the Coast Pilot,—‘Sailing vessels in the vicinity of the Race, or navigating along the southern side of the Sound near Gull and Plum Islands, should give them a wide berth when the ebb stream is running, or they may be drawn into one of the passages before aware of their danger.’ ‘There is always a strong tide-rip in the Race except for a period of about thirty minutes slack between the turn of the streams.’
Long Island is of interest to naturalists as a meeting-place for equatorial and arctic species of birds, and was a centre of the whaling industry as late as the first part of the nineteenth century, and Douglas, as already mentioned, in his Summary of the Planting of the British North American Settlements (1760) mentions specially that small whales affect the flats of Long Island. Altogether this sound appears to fulfil in every respect the requirements of Straumsfjord. The mainland immediately to the north of Fisher’s Island is hilly, though the mention of mountains at Straumsfjord may have another significance, which will be dealt with later on.
Now if we assume that the dispute between Karlsefni and the unruly Thorhall took place on Fisher’s Island or the mainland near it, the arguments of the two men would run somewhat as follows: Thorhall asserts that Leif’s landfall in Wineland must lie to the north of Keelness (Cape Cod), because Leif could not possibly have arrived on the coast which the later expedition had just explored, after leaving Markland, without previously sighting land. Karlsefni, on the other hand, regarding Keelness as the northernmost extremity of the country, has observed that from that narrow promontory the land has widened indefinitely as its southern coast was explored, and his view ‘that the region which lay more to the south was the larger’ may be paraphrased thus: the northern extremity of the country was obviously so narrow that Leif’s landfall could hardly have passed unobserved, whereas, here, to the south, the country is of enormous extent, so that, while we know everything there is to the north, to the south we may find anything. This appears to me a more reasonable explanation of this rather obscure passage than Dr. Nansen’s, viz. that it ‘was evidently due to the assumption that it (Wineland) was connected with Africa’.[103] Of such an assumption no real trace can be found, except in a later Icelandic geography, ‘thence it is not far to Wineland the Good, which some think is connected with Africa.’ To a geographer, anxious to place his countries within the limits of the known world, such a theory would be eminently natural. Confused by classical notions of the all-encircling Mare Oceanum, and hampered by the limitations imposed by early religious orthodoxy, primitive science would tend to deny the possibility of land connected with the known world on the farther side of the Atlantic; and to Africa, as the most westerly part of the world to the south of Iceland, the newly discovered lands would naturally be attributed; but it is hardly likely that Karlsefni would be hampered by geographical theories—at any rate there is no real trace of it in the saga.
Coming now to the furthest limits of Karlsefni’s expedition, at Hóp, it is obvious that we are provided in this case with a description which affords us more promising data than those with which we have hitherto been forced to be content. If we combine the information given in Eric’s Saga with that provided by the Flatey Book account of Leif’s camp, which clearly refers to the same place, the description becomes even more distinctive.
We need a land-locked bay, largely barred by shoals, guarded on one side of the entrance by a cape facing north, and on the other by an island, or something which might pass for one on a hasty visit. Into this bay a river must flow, which expands into a lake-like widening near its mouth, and then narrows, so as to divide the lake from the bay. This river must flow in from the north, as the Skrælings who visited the camp are said to have come from the south. A minor point, which is not so reliable as the remainder, is the mention of salmon in the river, which is included in the Flatey Book description.
Now it is manifestly not every river-estuary or land-locked bay which will conform to such a description in all, or even in nearly all, particulars. If therefore we find, in a suitable part of the American coast, a place which fulfils every one of these requirements, we may make our identification with something approaching certainty.
Now if the entrance of Long Island Sound be accepted as the site of the Straumsfjord base, the furthest limit of the exploration, at Hóp, can be made to fit the requirements of the story in a really remarkable way. I am convinced that it is a mistake to look for all the places mentioned in Karlsefni’s voyage within the restricted limits which seem to have contented other students of the subject. It seems to me illogical, when we hear of voyages of two or three days covering very considerable distances, to suppose when the saga says, ‘they sailed a long time,’ that we can be content to look for all the places mentioned in the course of a year’s exploration within a few hours’ sail of one another. It took a long time to sail past Furdustrands, and it was a long way from Straumsfjord to Hóp. The latter place is therefore to be sought about as far on from Straumsfjord as Straumsfjord was from Keelness. One has, moreover, to bear in mind, in searching for likely landfalls, that it is by no means every inlet which is likely to attract the notice of sailors on a coasting voyage. Openings which lie directly in their course, of which the situation selected for Straumsfjord is an example, are really far more likely to be explored. Now, about as far to the west of the entrance to Long Island Sound as Cape Cod lies to the east of it, the direction of the coast-line undergoes an abrupt change. And exactly in the angle formed by this change of direction is a bay, fulfilling all the requirements of Hóp. It is a land-locked estuary, largely barred by shoals, with a river running into it from the north, which widens into a lake among hills a short distance from the mouth. The approach involves a westerly course between a cape running north and an island. This is the bay or estuary of the Hudson River, constituting the modern approach to New York.
This was described by its first recognized discoverer, Verezzano, in 1524, in the following words: ‘We found a very pleasant situation among some steep hills, through which a very large river, deep at its mouth, forced its way into the sea.... We passed up this river, about half a league, when we found it formed a most beautiful lake, three leagues in circuit.’
Juet, in his account of Hudson’s visit to the same place, describes the estuary itself as a lake, and adds, ‘the mouth of that land hath many shoalds, and the sea breaketh on them as it is cast out of the mouth of it.... To the northward off us we saw high hills.... This is a very good land to fall with, and a pleasant land to see’.
De Laet, in his account of Hudson’s discovery, states, ‘he (Hudson) found there also vines and grapes, ... from all of which there is sufficient reason to conclude that it is a pleasant and fruitful country.’ Even the salmon, reported in the Flatey account of Leif’s voyage, in which, as has been pointed out, the description is largely borrowed from Karlsefni’s Hóp, appear formerly to have existed here. At any rate, Hudson is stated to have found them in this river, both by Juet and De Laet.
It is claimed that the analysis of Karlsefni’s voyage which has been attempted above presents no real difficulty, and is open to far fewer objections than any alternative theory. It is inconsistent with no fact alleged in the saga with the exception of the southerly course, and this, as has been shown, has to be abandoned on any hypothesis. It is the only theory which really gets over the Furdustrands difficulty; it provides a Straumsfjord and a Hóp which are both inherently probable landfalls, and which correspond in every particular with the details given. It does not seem to me that nearly as much can be said for the accepted theory of Nova Scotia, or for any other alternative. One further point must now be referred to. At the end of the section of Eric’s Saga and Hauk’s Book dealing with the last voyage and death of Thorvald Ericson comes a sentence which is quite differently rendered in the two versions. According to the Saga of Eric the Red, it runs, ‘They intended to explore all the mountains, those which were at Hóp, and those which they found.’ Hauk, however, gives it as follows: ‘They considered that the mountains which were at Hóp and those which they now found were all one, and so were close opposite to one another, and that the distance from Straumsfjord was the same in both directions.’ The word translated ‘intended’ in the first case, and ‘considered’ in the second, is the same, and the first part of the sentence is therefore nearly identical in the original, except for the omission of the words ‘at kanna’ (to explore) in Hauk’s rendering.
From this passage, as given by Hauk, it has been understood by Storm and some other authorities that after rounding Keelness the explorers came upon mountains which they imagined, rightly or wrongly, to belong to the same range as others which they had met with at Hóp.
Now the first point which occurs to one in this connexion is that the passage in question had, at an earlier date than that of any extant manuscript of the text, already become so corrupt as to be unintelligible. We can hardly regard the later half of the sentence as a gloss by Hauk: it is not characteristic of his work to make so considerable an addition to the matter copied. Still less can we suppose that the compilers of Eric’s Saga, who never retained any prejudice in favour of making sense of a passage, introduced the words ‘to explore’. It looks, in fact, as if at a very early date two inconsistent attempts had been made to interpret a phrase the meaning of which was already dubious. It is therefore a very dangerous passage on which to found any important conclusion.
Secondly, as has been already suggested, the passage about Thorvald bears all the marks of an interpolation. It comes between two sentences referring to the return to Straumsfjord which look as if the saga-writer were taking up the thread of his principal theme after a digression. It follows immediately after what is obviously information from a fresh source—the passage beginning ‘Some men say’. It introduces Thorvald suddenly for the first time, if we accept the purer version of Eric’s Saga (cf. p. 126). It is embellished with a speech plagiarized from elsewhere, a form of treatment without parallel in the saga. Towards the end of the suggested interpolation the words ‘they went back’ are twice repeated in Eric’s Saga. In these circumstances it seems fairly safe to regard this passage as having formed no part of the original story.
But if this be so, the sentence now under consideration, which mentions Hóp and Straumsfjord, cannot belong to the interpolated matter, but must be part of the original saga, and in this case it cannot refer to the topography of Thorvald’s voyage, but to the relation between Straumsfjord and Hóp.
In the third place, it seems unlikely that unscientific explorers would recognize two ends of a range of mountains as belonging to one another if separated by a long sea-voyage; the phrase ‘þat staediz mjök svá á’ (were therefore close opposite one another) seems to refer to a closer connexion, such as that of two sides of the same hill, which would be much more readily recognized.
The conclusions to be drawn are therefore:
1. The passage is too corrupt to allow of any important argument being based on it.
2. It is at least doubtful whether it refers to Thorvald’s voyage at all.
What follows is therefore put forward rather as an interesting suggestion than as a vital part of the main argument. But assuming that the sentence under consideration refers to the relation between Straumsfjord and Hóp, we know that mountains or hills were features of the landscape of both these places, and such features are not elsewhere specifically mentioned. If I am right in supposing Straumsfjord to be Long Island Sound and Hóp the estuary or lower waters of the Hudson, it would be quite correct to say that hills visible from the one place would also be visible from the other. If, as seems probable, the camp or base at Straumsfjord lay near the island at its mouth, it would also be true to say that any such mountain would be about the same distance from that camp, whether approached via Long Island Sound or by a route to the south of Long Island. As the explorers did nothing else, till the first winter at Straumsfjord, except investigate their surroundings, it is more than likely that they cruised sufficiently far up the sound to be able to see hills also visible from the Hudson valley. If this interpretation could be relied on it would therefore afford a strong confirmation of the topography suggested in this chapter, and I feel that this may be the correct explanation of the passage. It is safer, however, to treat the sentence as irremediably corrupt, and to conclude that the information it appears to contain may be a mere gloss, or may express a mistaken notion of the explorers. It is one of the many points as to which certainty is impossible, but it equally cannot afford a valid argument against theories which would otherwise be acceptable.