Before passing on to examine the voyages themselves, with a view to identifying so far as possible the territory explored, it is advisable to clear the way by the discussion of two questions, the first of which provides by its solution an approximate standard for the measurement of certain distances recorded, while the second provides a rough northerly limit to the possible situation of Wineland. The two questions are not in any way connected, except as being preliminaries to any trustworthy inquiry: as such they may conveniently be dealt with in one chapter, which may be skipped by the unscientifically inclined.

‘Dægr sigling’.

In the early days with which the present volume is concerned, the only method of measuring distances at sea was necessarily by time. No astronomical observations capable of giving results even approximately exact can then have been understood, and it is a curious fact in the history of navigation that even the simplest form of log for calculating the rate of progress was not introduced until comparatively modern times. The most natural method of measuring nautical distances in these circumstances would be by means of units corresponding to the usual divisions of time. We should therefore expect to find one unit representing an hour’s sail, another representing a voyage of twelve hours, and for use over long tracts of open sea possibly a unit based on the average progress during a period of twenty-four hours.

Now the standards of nautical measurement found actually to have been used by the Icelanders are primarily two—the ‘vika’, and the ‘tylft’ or dozen, which, as its name implies, represented twelve of the first-named units. It will be found useful for the present inquiry to establish first of all, with as much certainty as possible, the distances represented by the ‘vika’ and the ‘tylft’.

In a fifteenth-century manuscript incorporated in the collection of scientific treatises known as Rímbegla the following passage is to be found (p. 482): ‘Between Bergen and Nidaros (Trondhjem) there are about four degrees, so one degree comes to about a nautical “tylft”.’ Pausing here, we may observe that the voyage from Bergen to Trondhjem was evidently recognized to be four nautical ‘tylfts’. The passage continues: ‘now a degree on land and a “tylft” at sea are equal, and there are two “tylfts” in a day’s (dægur) sailing.’ To the expression used for a day’s sailing attention will have to be directed later on, but for the present it may be allowed to stand in the non-controversial form into which it is translated above. Taken as an accurate statement of the case, this quotation from Rímbegla has given us the following table:

Now if a day’s sail be taken here as equivalent to twenty-four hours, we have precisely the divisions of distance which, as I said at the outset, we ought to expect where the measurement is effected by time. A ‘vika’ represents an hour’s run, a ‘tylft’ twelve hours, and a day’s sail twenty-four hours. Whether, the geographical distances which they are alleged to represent have been correctly stated is another matter, into which we may now look a little more closely.

It is evident that the assumed correspondence between a ‘tylft’ and a degree, which, having regard to the state of navigation in the saga period, must in any case have been accidental, rests upon the hypothesis that the length of a voyage from Bergen to Trondhjem is four degrees or 240 nautical miles. The difference of latitude between the two places is in fact little more than three degrees, and even the rhumb-line connecting Bergen and Trondhjem is not 240 nautical miles in length; this error, however, need not necessarily have any effect on the author’s calculation. But on working out the shortest distance covered by a ship sailing from the one place to the other, it is apparent that in calling this distance four degrees a serious under-statement is made which vitiates the conclusion arrived at. This distance as sailed at the present day is said to be 318 nautical miles, and calculation or inspection of a chart will show that it is impossible to bring it much below 300, so that if this represents four ‘tylfts’, calculated by time, as it probably did, our table must be revised as follows:

This estimate is corroborated to some extent by the scale of Icelandic sea-miles (vikur) given in Troil’s Letters on Iceland (1780), where they are represented as nine to a degree or equal to 6⅔ miles each. Exact correspondence is of course not to be expected in standards of measurement arrived at by so rough a method as the time occupied on an average voyage.

Another ‘tylft’ capable of measurement is that given in the Greenland sailing directions attributed to Ivar Bardson. Here the distance so described is that between Reykjanes (lat. 63° 24′ N., long. 22° 40′ W.) and Snæfellsnes (lat. 64° 55′ 30″ N., long. 23° 59′ 40″ W.). Calculation gives the length of a rhumb-line between the two points as 73.54 miles, according to which a ‘vika’ would be about 6.12 miles, which once more justifies the assumption that something over six rather than five miles must be the correct measurement of this unit.

If the line of reasoning has been correct so far, it follows that the average rate of speed on an Icelandic voyage under favourable conditions would be something over six knots. The next thing to ascertain is the highest speed possible under exceptionally favourable circumstances. Fortunately this point is also capable of determination. It is unnecessary, and probably misleading, to enter, as Mr. Babcock does, into calculations based on the speed of modern ships. In the saga of Olaf the Saint (see this saga in Heimskringla, §125), one Thorar Nefjolfson accomplished what was evidently regarded as a remarkable feat by sailing from Norway (Moeri) to Eyrarbakki in Iceland in the space of four days and four nights. There is in this case no ambiguity about the meaning of eight ‘dægra’, the period recorded, for Thorar himself refers to the fact that four nights previously he was with the King in Norway. The starting-point may safely be taken as Stad, which lay in the Söndmöre district, since we know from other sources that it was the usual place of departure for Iceland, as indeed its geographical position at the extremity of the westerly trend of the coast-line from Trondhjem would render inherently probable. The geographical position of Stad is 62° 11′ N., 5° 8′ E. The distance to Eyrarbakki (63° 51′ 45″ N., 21° 7′ W.) round the most southern point of Iceland (63° 23′ 45″ N., 19° 5′ 5″ W.) comes in round figures to about 730 nautical miles. This would represent a rate of about 7·6 knots, and though this is probably too little, as the course can hardly have been so direct and we know neither the precise place of departure nor the exact times of start and finish, we shall be safe in assuming that anything appreciably over eight knots was beyond the extreme powers of an Icelandic vessel.

According to our calculations, then, the average distance covered in twelve hours with a fair breeze would be about seventy-five miles, and having obtained these important data we may now proceed to consider more particularly the unit of distance uniformly employed in the story of Wineland, namely the ‘dægr sigling’ or day’s sail.

In its strictly scientific signification there can be no doubt that a ‘dægr’ is a period of twelve hours. The Rímbegla (not the treatise already cited, but another incorporated in the same collection) is explicit upon the point. ‘In a day there are two “dægra”, in a “dægr” twelve hours’ (p. 6). In nautical phraseology, in which the word most commonly occurs, it cannot be denied that it is sometimes used with the same meaning. The passage already quoted, recording the voyage of Thorar Nefjolfson, is a case in point. On the other hand, the statement of the Rímbegla treatise already cited, that there are ‘two tylfts in a “dægur” sailing’, must clearly be interpreted as meaning twenty-four hours, since even 120 nautical miles could not be covered in twelve hours at what we have found to be the extreme speed of an Icelandic sailing ship, and we should always hesitate to assume the identity of local or technical usage with accurate scientific terminology.

One might, for example, be led seriously astray by taking the length of a mile from a geographical textbook and applying it under all circumstances to any distance called by the same name.

The author of the last-mentioned passage in the Rímbegla seems indeed to use ‘dægr’ and ‘dag’ interchangeably, for he goes on to say that ninety degrees of the earth’s circumference would take forty-five ‘dag siglingar’, and the complete circumnavigation of the globe would occupy 180 ‘dag siglingar’. If this passage stood alone it would doubtless be possible to explain the first ‘dægr’ as a mere verbal slip; it is accordingly necessary to examine the matter from a different standpoint, and to investigate the distances said to be covered by a given number of ‘dægra sigling’.

A convenient passage for this purpose occurs in Landnáma I, 1. The writer is evidently endeavouring to fix the position of Iceland by reference to well-known points on all sides of it. With this object he makes the following statement:

‘Wise men say that from Norway from Stad it is seven “dægra” sail west to Horn on the east of Iceland; but from Snaefellsnes where the distance is shortest, there are four “dægra” of sea to Hvarf in Greenland[92].... From Reykjanes in the south of Iceland there are five “dægra” of sea to Jolduhlaup in Ireland, to the south, while from Langanes in the north of Iceland there are four “dægra” of sea to Svalbarda in the Polar Sea (Hafsbotn) and it is a “dægr” sail to the uninhabited parts of Greenland from Kolbein’s island (Mevenklint) north.’

Let us examine these statements seriatim.

Horn in the east of Iceland may either mean the modern Cape Horn, the most easterly point of the country, or more probably East Horn a little further to the south-west. My reason for preferring the latter place is that it appears to have been the most easterly Horn known as such to the authors of Landnáma. It is referred to shortly afterwards in describing the discovery of the land by Gardar, and was evidently not the most easterly point of the country, for Gardar is said to have arrived to the east of it. The position of the most easterly part of Cape Horn is 65° 5′ N., 13° 27′ 45″ W.; that of East Horn is 64° 20′ N., 14° 25′ W. The distances from Stad to these two places respectively are 524·67 and 543·46 miles. In this case, therefore, it is clear that seven periods of twelve hours are meant, and the distance covered in each ‘dægr’ corresponds closely with the average ‘tylft’ at which we have already arrived, being from 74·9 to 77½ miles according to the objective chosen. It is clear from this that we are here dealing with averages, and not, as Storm suggests, with records, for the rate is but 6·4 knots at the outside, which apart from what we know of Thorar Nefjolfson’s voyage is obviously nothing extraordinary, while the journey between these two points must in all probability have been traversed more frequently than any of the others here referred to.

Hvarf (turning-point) in Greenland was either Cape Farewell or one of the promontories such as Sermesok lying immediately to the north-west of it, and for our present purpose it will be fair enough to calculate the distance to Cape Farewell.

This works out at about 631 miles in a direct line, and it is at once evident that four periods of twelve hours are quite insufficient to cover the voyage. On the other hand, four days of twenty-four hours suit remarkably well, the rate being about 6½ knots.

Let me now deal with one or two possible objections. First it may be urged that the version of the passage which specifies Hvarf as the objective may be wrong, and that the coast of Greenland immediately west of Snæfellsnes is the point of measurement. The words ‘west to Greenland’, which take the place of any mention of Hvarf in the alternative reading, may seem to bear out this view, but a glance at the chart will show that all the courses laid down must be interpreted with considerable freedom, and that Hvarf answers as closely to west of Iceland as, say, Ireland to south of Reykjanes. The real answers to the objection, however, are first that no one can ever have completed an uninterrupted voyage to a point in Greenland due west of Snæfellsnes, having regard to the ice barrier which at this point intervenes between the coast and the open sea; and next that the distance to Greenland due west of Snæfellsnes, about sixteen degrees of longitude, is at least 400 miles, and is therefore an equally impossible distance to cover in forty-eight hours sailing. Finally, it is surely more probable that a point regularly passed on the voyage between Iceland and Greenland should be chosen for measurement than an undefined locality in an unexplored region hundreds of miles out of the track of practical navigation. The next objection will possibly be that I have measured the distance on the rhumb-line, whereas it appears from the old sailing directions that this was by no means the usual course adopted. The course laid down in the directions attributed to Ivar Bardson appears to lie west for a day and a night and then in a south-westerly direction parallel to the belt of ice. Now first of all it must be remembered that a rhumb-line course is not actually the shortest, and if a day and a night due west be laid down on the chart and the remainder of the distance be calculated from say longitude 29·45 W., the resulting distance will not be very materially increased, but will come to somewhere about 645 miles, which can still be covered in four days, at a rate of about 6·7 knots. In point of fact probably all the courses with which I am dealing would in practice be longer than I have estimated them, and the average rate which I have deduced from them should be slightly increased, while the same does not apply to the rate of eight knots which I have taken as the maximum, since in this case a liberal allowance for deviation has already been made. If it be said that my maximum and average rates are in such circumstances brought rather close together, I reply that in fact a gale does not bring with it a very great advantage in speed over a fair sailing breeze, as the effect of the sea raised is to neutralize much of the gain which might otherwise be anticipated. If the distance actually travelled between Snæfellsnes and Hvarf be increased even to 700 miles, the rate is not much over seven knots, or well within the limits assigned. For these reasons the distance given in Landnáma between the two points seems to me to be a correct statement, but ‘dægra sigling’ must here be interpreted as days of twenty-four hours.

Similarly in the case of the voyage from Reykjanes to Jolduhlaup in Ireland. This cannot by any means be brought within the space of five ‘dægra’ of twelve hours each. Approximately the nearest points in Ireland may be taken as about 688 miles distant. Malin Head in the north and Erris Head in the west of Ireland are almost equidistant from Reykjanes, the former being some 685, the latter 690 miles from the starting-point. There is no real reason to suppose that any point in Ireland so near to Iceland is the true position of Jolduhlaup. It is evident on the other hand that a very few more miles will make the distance recorded perfectly consistent with five days of twenty-four hours. If we bring our ship into Sligo Bay the distance will be 718·6 miles, or ten ‘tylfts’ of 71·8. This would be perfectly consistent with the standards of distance already considered, but of course Jolduhlaup may easily have lain even farther away than this from Reykjanes. The name is generally taken to mean ‘wave-run’, and is sometimes spelt Olduhlaup. Joyce[93] attributes a Scandinavian origin to the name of Olderfleet close to Larne Harbour, and as ‘hlaup’ and ‘fljót’ are both common terminations meaning ‘stream’, this word in an Icelandic form (Oldufljót) would be practically identical with Olduhlaup. The author above quoted says that the first part of Olderfleet is a Scandinavian corruption of Ollorbha, the Celtic name of Larne water, but whether the true derivation be from this word or ‘oldu’ a wave is a question which applies equally to Olderfleet and Jolduhlaup and affords no ground of distinction between them. As far as names are concerned the two may well be identical. Larne would be the first important harbour after entering the North Channel between Scotland and Ireland, and may well have been chosen therefore as a well-known point for the measurements in Landnáma. From Reykjanes to Rathlin Island off the entrance of the North Channel is about 713 miles, thence to Larne would be about thirty-seven more, making a distance, if there be anything in this conjecture, of some 750 miles to be covered in the five ‘dægra’—ten ‘tylfts’ of seventy-five miles, which corresponds exactly with our amended table.

It has been objected that some of the MSS. do not read ‘five dægra’; this is true, but the alternative (three dægra) does not help those who contend for a twelve hours ‘dægr’, while even if we adopt the arbitrary emendation of the version printed at Skalholt in 1688 and read ‘eight dægra’, the rate of travel, even to the nearest point, would be too rapid to be normal. We have therefore once more a statement remarkably consistent with our data if we interpret a ‘dægr’ as twenty-four hours, and wholly impossible if a ‘dægr’ must universally be considered as only twelve.

In estimating the distance from Langanes to Svalbarda we are confronted with the difficulty that we do not know where the latter place can have been. I am content, however, to admit that in this case a dægr of twelve hours seems to be indicated. Four times twenty-four hours would penetrate too far into the Arctic regions to be at all probable, while Jan Mayen seems best to fulfil the conditions of a spot to the north of Langanes, situated in the Polar Sea.

From the point of Langanes to the southern extremity of Jan Mayen is about 296 miles, or ‘tylfts’ of 74 miles, the route in summer would at this point normally be clear of ice, and altogether it seems probable that Jan Mayen rather than Spitzbergen, as sometimes suggested (840 miles away), is the place described as Svalbarda.

The last distance recorded is from Kolbein’s Island (Mevenklint) to the uninhabited coast of Greenland lying to the north. The position of Mevenklint is in lat. 67° 10′ N., long. 18° 30′ W., and the nearest point on the Greenland coast would be about lat. 69° 40′ N., long. 22° 48′ W. The distance would therefore be 177·45 nautical miles, and so it is evident that it could not be covered by a voyage of twelve hours. In twenty-four hours, however, under exceptionally favourable conditions, the whole distance could be traversed, and in any case in that period of time a ship would be likely to have got as close to the land as the ice would permit. It is not likely that this particular voyage, which is not included in all the texts of Landnáma, was sufficiently often accomplished to enable a fair average to be taken; the allusion is more probably to a special case within the knowledge of the authors, which would in all likelihood have taken place on an exceptionally favourable opportunity.

Now the conclusions to which we are forced by the consideration of all these distances recorded in Landnáma are as follows:

1. Only two out of the five voyages are at all compatible with a ‘dægr sigling’ of twelve hours.

2. These two appear to be very accurately recorded, which raises a presumption in favour of the correctness of the other data. In the voyage from Stad to C. Horn we have exactly seven ‘tylfts’ of 74·9 miles to cover in seven dægra, in that from Langanes to Svalbarda (if Jan Mayen is meant) four ‘tylfts’ of seventy-four miles each in four dægra.

3. Either the remaining three are hopelessly inaccurate, or a ‘dægr sigling’ in these cases means twenty-four hours.

4. If they are inaccurate, it is a most remarkable coincidence that they can all be made accurate by adopting the basis of twenty-four hours.

Thus, taking the average of seventy-five miles in twelve hours at which we had previously arrived:

The distance from Stad to C. Horn would take 6·9 or practically seven days of twelve hours (given as seven dægra).

If the alternative Horn be taken the voyage would occupy 7·1 days.

From Snæfellsnes to Hvarf would be 4·1 days of twenty-four hours (given as four dægra).

In sailing from Reykjanes to any part of Ireland one could not arrive before the fifth day of twenty-four hours was well advanced, and it would be easy to find a point which would occupy exactly the time prescribed. From Langanes to Jan Mayen the distance is correct within eight miles, which may easily be accounted for by slight differences in the points of arrival or departure.

From Mevenklint to Greenland would occupy 1·16 days of twenty-four hours.

Thus the discrepancies are so slight that even if the rate had to be limited to this average, the statements would be as correct as so vague a unit as a day’s journey would permit, and of course the variation in speed must have been greatly in excess of anything required absolutely to justify these estimates in the smallest detail.

That in the case of three out of five statements such a correspondence should be fortuitous seems to me to be out of the question.

It will doubtless be objected that I am not justified in interpreting the same word in the same passage by two different periods of time. The compilers of the Landnámabók, however, expressly disclaim personal responsibility for the statistics recorded. They are based on the reports of ‘vitrir menn’, men that is with the requisite special knowledge, and once it is admitted that the meaning of the expression ‘dægr’ may have varied from place to place, there is nothing extraordinary in a discrepancy of this nature being exemplified in a passage based on information gathered from different informants in the east and west of Iceland.

It is comparatively easy to see how such a discrepancy in nautical use may have arisen. Evidently ‘dægr sigling’ was the usual nautical expression for a day’s sail. This is shown not only by the fact that it is nearly always in a nautical context that the word ‘dægr’ makes its appearance, but also by the opening sentence of the Landnámabók’s preface, which renders Bede’s words ‘sex dierum navigatione’ by ‘sex dægra sigling’ as the obvious equivalent. Now of course until the exodus brought about in Scandinavia by the policy of Harold Haarfagre, the vast majority of the voyages undertaken by Norsemen were along the coast of Norway and the adjacent countries, and were carried on almost entirely by day, the ships putting into a convenient haven almost every night. The coast of Norway, before the days of lighthouses, cannot have been a pleasant place to navigate in the dark, and in fact we almost always find it recorded, as an exceptional occurrence, when any motive induced the seamen of this period to sail day and night without stopping. A day’s journey in a ship would therefore in the normal course be equivalent to the distance covered in a ‘dægr’ of twelve hours, and thus the application of this word to a nautical day’s journey doubtless began. Then, when colonial expansion and viking enterprises made continuous open-sea voyages more common, two courses would be open to those who wished to record the distance travelled. They might take the nautical expression ‘dægr’ as referring to the twelve hours actually occupied in sailing under old conditions, or they might take it as extending to the period during which the ships of less venturous seamen had usually lain at anchor. A man who had taken—say—four ‘dægra’ to sail between two points, stopping at night, would actually have travelled but forty-eight hours, but the time occupied from point to point would have been four days of twenty-four hours. According to the aspect of the question which struck a sailor accustomed to this method of reckoning he would be likely to call a continuous voyage of four days either four or eight ‘dægra’. Thus a variety in local usage might quite naturally spring up which would account for the discrepancy which has given rise to the difficulties with which I have been endeavouring to deal.

Of course it is but seldom in passages where this expression is used that we have any data at all to enable us to say which meaning should be attached to the word. In the sagas of Wineland the word ‘dægr’ occurs perhaps with unusual frequency, and to my mind every passage where it is there employed might be prayed in aid of the argument that a ‘dægr sigling’ was frequently twenty-four hours. But to use these passages at this stage would be to argue in a circle, and we must be content to rest the assumption that the word was so used on the data of which use has been made in the foregoing argument, reserving to ourselves the right in subsequent investigation of the voyages to accept what is there stated with regard to distances sailed, even though on the hypothesis that a ‘dægr’ can only mean twelve hours the statements made are clearly incredible.

The ‘eyktarstad’ problem.

In the account of Leif’s sojourn in Wineland, contained in the Flatey Book, will be found a passage which has given rise to more acute controversy than any other in the story. It runs as follows:

‘Sol hafðe þar eyktarstad ok dagmálastad um skamdegi’—the sun had there eykt place and breakfast place on the shortest day, or, as rendered in our translation, p. 42, ‘on the shortest day the sun was up over the (Icelandic) marks for both nones and breakfast time’.

Now one may note in passing that, whatever the significance of the words, they are evidently not the sort of thing which a romanticizing saga-writer would introduce from his own imagination. This is admitted by the most adverse critics of the authority which reproduces them.

In view of the attitude taken up by some modern writers, it is important to point out their entire independence of anything to be extracted from the rival version. They go far to disprove, if disproof be necessary, the theory that the Flatey Book account is borrowed from the Saga of Eric the Red.

But at this point in the inquiry we are less concerned with this than with the precise significance of the expression used, and though the question has finally been solved, and nothing new can be added, it is necessary, for the sake of readers unfamiliar with the subject, to devote some space to the matter.

The Icelanders, possessing no clocks or scientifically constructed dials, were in the habit of estimating the time of day by the position of the sun above the horizon. With this object they marked eight points upon the horizon, utilizing hills and natural objects where such were conveniently situated, and erecting cairns in places which were otherwise undistinguished. This method of time-keeping, crude as it was, persisted down to very recent times, if indeed it is not still in use in some parts of the country. Henderson, who visited Iceland in 1814–15, describes the method in some detail (Iceland, vol. i, p. 186), and gives the names and time-equivalents of the various points as follows:—

In an earlier work,[94] the same divisions of time are mentioned, but with some difference in the equivalents, thus:—‘Otta is with them three o’clock in the morning; Midur morgon or Herdis rismal, five o’clock; Dagmal, half past eight; Haadege, eleven; Nonn, three in the afternoon; Midur afton, six in the morning (sic: obviously should be ‘afternoon’); nattmal, eight, and midnatt twelve o’clock at night.’ A little thought will make apparent the reasons for these discrepancies in time, for not only is the method exceedingly rough, but of course the horizontal bearing or azimuth of the sun at a particular time is not the same throughout the year, and it also varies with the latitude. For example, taking the latitude of Iceland as 65°, and the obliquity of the ecliptic in a.d. 1000 as 23° 34′, which is substantially accurate, and calculating the sun’s bearing at three o’clock p.m. throughout the year, we get:—

while on shifting the latitude to 51° 30′ (about that of London) we get a bearing of 68° 17′ for 3 p.m. at midsummer.

It appears, however, from the fact that one of the eight points was midnight, and another ‘hádegi’ (high day or noon), that the scheme would aim at dividing the equinoctial day into three-hour intervals. Dagmál would then be about 9 a.m. and Nón 3 p.m. The latter word originally meant the ecclesiastical ‘nones’ (3 p.m.) and in old Icelandic ‘eykt’ is used as synonymous with ‘nones’.

In the Icelandic Ecclesiastical Code, or Kristinret, instructions are given for the correct location of the mark for ‘eykt’. ‘It is eykt’, says the law, ‘when the south-west airt is divided into three, and the sun has passed two divisions and has one to go.’ This gives us a bearing of S. 52° 30′ W. for ‘eykt’ or nones, which would be, in Iceland of the eleventh century, pretty correct for 3 p.m. between the equinoxes and the summer solstice, during nearly the whole time, that is, when the sun would be visible at this hour in these northerly latitudes. (See accompanying diagram.)

Now the root error of all the earlier commentators who attempted the elucidation of the passage under consideration consisted in treating ‘eykt’ not as a solar bearing, but as a definite clock time. Three o’clock clearly would not do, for sunset at 3 p.m. on the shortest day in winter indicates a latitude too far north to correspond in any way with the climate indicated. Torfaeus, the earliest writer on the subject, accordingly interpreted ‘the south-west airt’ as the whole quarter between south and west, and dividing the time between noon and 6 p.m. (equinoctial west) into thirds he arrived at 4 p.m. as the time of sunset, which with 8 a.m. for Dagmál gave an eight hours day, or a latitude of approximately 49° N. Of course, for the reasons already given, the bearing corresponding to such a division of the horizon (S. 60° W.), assuming the latter to be justifiable, would not unalterably represent 4 p.m. even in Iceland, and the clock time for which the bearing stood in Iceland would be indicated by a wholly different position of the sun in another latitude.

Next came what may be called the school of Rafn, who claimed to have located the Wineland of the sagas with certainty in the neighbourhood of Rhode Island. For them an interpretation which resulted in a latitude of 49° was unsatisfactory. They accordingly prayed in aid a passage from Snorri’s Edda, in which the winter is said to begin at the point where the sun sets in ‘Eyktarstad’. It was known that winter, according to the Icelandic calendar, began in the week preceding the 18th of October, and observation in the latitude of Snorri’s home showed that the sun set there on the 17th of October at 4.30 p.m. As the passage is drawing a distinction between autumn and winter it could hardly refer to the Icelandic winter beginning about the 18th of October, for as Vigfusson has pointed out with regard to this division of the calendar, which persists in modern Iceland, it is a division of the year into summer and winter only, and leaves spring and autumn out of account.[95] But it led Rafn and his followers to assert, in the teeth of all the other evidence, that ‘eykt’ was not a point but a period of time, and that ‘eyktarstad’ was a point which could be interpreted as 4.30 p.m. apparently in any latitude! This, with ‘dagmálastad’ at 7.30 a.m. gave a day of nine hours, from which Rafn claimed to deduce the latitude—to a second of arc—as 41° 24′ 10″, an observation which, accepting Rafn’s theory as to the locality visited, would be beyond the accuracy of a modern sextant. Unfortunately for this surprising result, the method of calculation was hardly so correct, for, apart from the fallacy of treating the local time as transferable, no correction was made for the effects of refraction, &c., and the declination assumed was not that of the eleventh but of the nineteenth century.

It remained for Dr. Gustav Storm to point out the correct way of utilizing the data supplied. Assuming the instructions in the Kristinret to apply to an observation recorded of an earlier day, and assuming the passage to mean that the sun set at the precise moment of ‘eykt’, the amplitude, or distance from the west at setting, of the sun on the shortest day in Wineland was 37½. We may assume, as the observer would have been looking across the land, that the lower edge of the sun was at least 19′ above the actual horizon, and this being so no allowance for refraction or dip of the horizon need be made before working out the formula:—sec : lat : = sin : amp : cosec : decl. Professor Turner, of the Oxford University Observatory, has kindly supplied me with the corrected declination for the year a.d. 1000, viz: 23° 34′ 8″. We need not trouble about the seconds, as we know neither the precise moment of the solstice nor even the year with certainty; omitting these the problem works out as follows:—

The latitude therefore would be about 48° 57′ N. This, however, correctly understood, gives only the northern limit beyond which the observation could not have been made.

It might be argued that the refinements enjoined in the Kristinret were not likely to have been in operation in these primitive times. There seem to have been eight day-marks, two of which represented midnight and noon respectively, and it would seem more natural therefore for men who attached no particular importance to the hour of 3 p.m. such as was subsequently associated with the time of nones, to divide their horizon into equal parts, which would serve, at any rate at the equinoxes, accurately for 6 a.m. and p.m. and mid-day, while dagmál and eykt would occupy the points midway between the others, and stand, less accurately, for 9 a.m. and 3 p.m. The answer to this criticism probably is that it was found necessary to divide the day into equal watches: anyhow, such an interpretation cannot be correct, for an amplitude of 45° would give a latitude of 55° 34′ up to which this bearing would be visible, and this would be too near the latitude of Greenland to be remarkable, while nothing is clearer than that the writer of the passage was endeavouring to record a marked and surprising difference from the length of the winter day to which Greenlanders were accustomed.

This in fact, rather than a precise determination of latitude, would seem to be the object of the statement, taken as a whole. It is as if one were to say, ‘I could breakfast, or shave, by daylight all the year round.’ It by no means follows from the passage that dagmálastad and eyktarstad are meant to be understood as sunrise and sunset; in fact, it would involve an extraordinary coincidence if they were. There were only eight points in general use by which the time of day could be measured or expressed, and to say therefore that the sun was up at a particular time does not indicate that at that precise moment it was on the horizon. Indeed if Rafn had been content with probabilities instead of trying to make the passage support an exact determination of latitude, he would have made out a fairly strong case, so far as eyktarstad was concerned, for the locality which he identified with the explorers’ camp. The chances are that, over a background of wooded and hilly ground, actual sunrise and sunset were invisible, and that the sun was well up at the time of passing over the points recorded. I have calculated roughly the altitude of the sun in eyktarstad at the time in question, and I make out that even so far south as 40° it would not be as much as 5½°. Even assuming that the time of sunset was meant, it would not require any very great unevenness of the horizon to produce the effect of sunset at this point in the latitude supported by Rafn, and it is almost certain that the locality indicated was much nearer to this latitude than to the northern limit of the observation.

Taken with their context, the words seem to be an illustration of the greater equality of day and night referred to in the opening words of the sentence. Their real value lies in the fact that they embody a remark of a circumstantial and business-like character, which goes far to support the historical authenticity of the narrative. It is not the sort of thing that a romancer would invent, it is the sort of thing that a traveller would notice. Secondarily, though in all probability the words indicate a much more southerly latitude, they make it impossible that the site of the observation was north of (roughly) 49°. To consider them as a deliberate attempt to fix latitude is to lose sight of all probabilities. Let any who still adhere to this interpretation go and fix marks for themselves, and endeavour therefrom to ascertain the latitude. The south point could of course be fixed accurately, by the place of the shortest shadow or various other well-known devices. The time equivalents given by Henderson and Troil do not, however, suggest that it was so fixed as a rule. But without instruments to measure the angles for the other marks correctly to—say—2° would be very difficult indeed, while the marks themselves would probably subtend an appreciable angle. An error of one degree will be more than reproduced in the latitude. Any change in the exact position of the observer would be likely to cause an inaccuracy of at least this extent; so that if the locality visited is to be identified at all it certainly will not be by the use of this passage, on which so many commentators have expended so much fruitless ingenuity.