Surprisingly little attention has been paid by editors and commentators to the geography and chronology of Gulliver's Travels. Sir Henry Craik, in his Selections from Swift, found the geography worth a fairly extended passage,¹ Mr. G. R. Dennis, in his edition of Gulliver,² commented on some of the cruces, and Mr. Harold Williams devoted some space in the introduction of his edition to a discussion of the maps.³ Mr. Williams also provided his readers with the most satisfactory commentary we possess on the difficulties and inconsistencies of the time-scheme.⁴ The conclusions reached by these scholars, and, indeed, by almost all students of Swift who have occupied themselves with the problem, were that Swift worked out for his book a detailed framework in time and space, but that it is (at least as it has come down to us) so imperfect that it is impossible to reconcile it with itself or to be sure, in many instances, of the author's intentions. Recently, however, Professor John R. Moore has suggested that the geography and chronology of the Travels are so nonsensical as to indicate intentional confusion: in other words, they are part of a satiric burlesque of travel literature.⁵

Undoubtedly the framework of the Travels presents difficulties. Unless some happy chance brings to light more material like the Ford letters we shall have to rely on conjectures with regard to a number of points. Yet it is possible to solve a good deal of the puzzle, and to come much closer to certainty than is usually believed.

Two geographical authorities are spoken of in the Travels—Nicolas Sanson and Herman Moll. The only mention of the former is found in the second voyage, where the "little Book" carried by Glumdalclitch is described (2.2.8) as "not much larger than a Sanson's Atlas" There is nothing to indicate that Swift ever looked inside the covers of this book, which he cites only because it is the largest he can think of. There was, indeed, little reason for using Sanson as an authority. His atlases had all been published in the seventeenth century, and the first quarter of the eighteenth had seen great advances in geographical knowledge, especially in the neighborhood of those regions in which Swift was to place his imaginary countries. Swift's obvious course was to avoid, as far as possible, any gross contradictions of accepted cartography, and he could not have effected this more surely and easily than by following the maps of Moll. This eminent English engraver of Dutch extraction had industriously gathered geographical data from the turn of the century, and by Englishmen, at least, his maps were widely accepted as the standard. As a matter of fact, however, the early eighteenth-century maps were much of a muchness, at least so far as the general physical structure of the world was concerned. Even in the regions of the South Seas (a term which sometimes included the North as well as the South Pacific) the cartographers tended to agree, only venturing to show occasional divergences in minor details, and often attempting to hedge by the use of faintly drawn coast lines which warned the reader not to put too much faith in what was, at best, unverified information.

Swift's remarks about Moll are entirely different in tone from his incidental reference to Sanson. Gulliver ventures (4.11.3) to disagree respectfully with Moll in the matter of the southeastern portion of New Holland, which is, he insists, at least three degrees west of the position it occupies in Moll's maps. This is enough to show that Swift had consulted Moll with attention: it also shows that he felt it safe, in the current state of geographical knowledge, to dispute minor points even with an expert. Whether he relied on any other sources it is impossible to say. No atlas appears in the catalogue of his library as it existed at the time of his death, but he did own numerous books of travel, most of which contained charts of some kind.⁶ Among them was Dampier's New Voyage around the World (1697) which, it is almost certain, Swift read intensively. This book contains some of Moll's earliest maps, including one of the world, and another of the Netherlands East Indies and the northern part of New Holland or Australia. Some of the place names Swift uses in the Travels differ in form from those used by Moll: but these Swift may have taken from the mass of travel literature which he read during the period in which the book was composed.⁷ In any event, a fairly extensive search among the maps and atlases of the period has brought to light no map which agrees more closely than Moll's with Gulliver's geography, and it will be convenient to use the former's A New & Correct Map of the Whole World (1719) as a basis for discussion.⁸

Swift's primary geographical problem was not a difficult one. He had to find, in the unexplored portions of the globe, locations for seven imaginary countries, only two of which were of considerable size. It was desirable that these locations should not be so close to either pole as to be obviously unfit for human habitation, and Swift apparently felt also that they should not be too close to the equator: the climates do not seem to differ much from that of England. It was also necessary to take care that these countries did not lie too close to each other, or to trade routes generally known to Europeans, though, on the other hand, they could not lie too far from these routes, lest it should be difficult to account for Gulliver's arrivals and departures. Accordingly we find that the three smallest countries, Lilliput, Blefuscu, and Houyhnhnmland, are placed not far from Australia, although on different sides, and that Brobdingnag, Balnibarbi, and the islands described in the third voyage are situated in the North Pacific, by far the largest area which remained unexplored by Europeans in 1720.

An examination of Moll's map shows the astuteness of Swift's choices.⁹ Most traders with the Orient in the eighteenth century preferred to hug the coast line of Asia, sailing to China through the multitudinous islands which lie between Siam and Australia. Consequently the Malay Archipelago was well known to mariners, save for New Guinea, the eastern and southern coasts of which were in dispute, some cartographers boldly uniting the island with Australia, while others remained noncommittal. The northern and western coasts of Australia had been determined with considerable accuracy, but the eastern and southeastern shores were not even guessed at, although there was a general agreement that it was possible to sail between Australia and New Zealand, the southwestern corner of which appears in Moll's maps. Exploration in the North Pacific had been checked by the disinclination of the Japanese to deal with Europeans: even the Dutch, who had a monopoly of occidental trade with Japan, had for nearly a century been limited to the port of Nagasaki. There was, therefore, almost no knowledge of the geography of northern Japan and of the regions which lay beyond. It was not known how far Iesso (Yezo) extended, whether it was an island or a peninsula, or whether any lands lay to the north or the east of it, although many maps vaguely indicated a territory to the east called "Company's Land"—claimed by the Dutch East India Company on the ground of explorations early in the seventeenth century. This land was separated from Yezo by the "Straits of the Vries," which led to a hypothetical northwest passage to Europe, hope for the discovery of which had not yet been abandoned. The eastward extent of "Company's Land" was frankly a matter of conjecture, some maps indicating only the western tip, others showing (though in faint lines) a coast stretching almost due east to within a short distance of the North American coast, close to the "Straits of Annian." Western North America was perhaps the most mysterious region of all. The cartographers were not even sure whether California was an island or a part of the main continent. Many maps, including Moll's, showed the Gulf of California extending northward until it rejoined the Pacific at the entrance to Puget Sound, north of Cape Blanco. San Francisco Bay was undiscovered, and to the north and west of Puget Sound lay a perfect and absolute blank.

With so much of the globe at his disposal it would seem that Swift should have had no trouble in arranging his hypothetical countries without coming into too violent conflict with accepted geography. And yet a series of misfortunes has led to a very general misunderstanding of his intentions. The chief misfortune was the fact that Swift, because of the secret method employed to publish the Travels, was unable to read proof and correct errors of detail. It is impossible to determine to what extent the errors were the fault of the author, the transcriber, and the printer, but that there were errors is undeniable. The second misfortune was that the maker of the maps for the original edition was careless or stupid or both. He was guilty of mistakes for which there can be no excuse—the misspelling of names, the miscalculation of comparative distances, the misplacing of localities with respect to each other when there was no discrepancy in the text from which he worked. But despite all this, and despite the fact that no one believes that Swift approved the maps, certain assumptions which the original engraver made have been accepted by modern scholars without question, with the result that Swift has been charged with errors of which he was not guilty. The engraver's practice seems to have been to read the text until he came to the first description of the location of the land Gulliver was visiting, then to assume that this description was accurate, and that any conflicting supplementary data must be reconciled with it, or ignored, if reconciliation proved to be impossible. It seems never to have occurred to him to assemble all of the geographical data in a given voyage, to try to harmonize the whole and, if conflicts appeared, to seek a solution of the difficulty without giving preference to any statement because it came early in the tale rather than late. And yet such a procedure would have prevented a series of blunders which begins with the first voyage.

The wreck which cast Gulliver upon the shores of Lilliput occurred "Northwest of Van Diemen 's Land" and "in the Latitude of 30 Degrees 2 Minutes South" (1.1.5). Two areas bear the name "Van Dieman's Land" (or "Dimen's Land") in eighteenth-century maps—north-western Australia and Tasmania—but as the former lies between ten and twenty degrees south latitude it cannot have been intended here by Swift unless the passage in the Travels is hopelessly corrupt. But there are difficulties even if one assumes that the reference is to Tasmania. A glance at a modern map will reveal what was apparent to the eyes of the engraver of the original maps for the Travels: that a point northwest of Tasmania and in latitude 30° 2' S. lies not in the ocean but well inland in Australia. A timid soul might have solved this discrepancy by moving Lilliput a little to the south of Australia, in latitude 32° or 33° S. The engraver was made of bolder stuff. He moved Tasmania some forty-five degrees to the west, and placed Lilliput in the Indian Ocean, due south of the western end of Sumatra. Australia was ruthlessly erased from the map. It is not surprising to find that this was only the beginning of the engraver's liberties. Both Lilliput and Blefuscu are drawn on a scale far too large, as compared with Sumatra, and Mildendo appears as "Mendendo."

Yet after all this manipulation of the facts there are more difficulties to come. When Gulliver left Blefuscu he sailed first north and then northwest (1.8.9). "My Intention," he says, "was to reach, if possible, one of those Islands, which I had reason to believe lay to the North-East of Van Diemen's Land." On the third day he was picked up about twenty-four leagues from Blefuscu, in latitude 30° S., by a ship sailing on a southeasterly course in its return from Japan "by the North and South-Seas" Either an eighteenth- or a twentieth-century map will show the ridiculousness of this account if we assume that Lilliput lies either south of Sumatra, as Motte's engraver thought, or in the Great Australian Bight, as some later commentators have suggested. It represents Gulliver as attempting to sail entirely around Australia in search of an island on which to land, and it describes the ship's captain as laying a southeasterly course from the neighborhood of Sumatra in order to reach England. If we adopt the theory which seems to be most favored today—that Van Diemen's Land is northwestern Australia, and that the position of Lilliput is about 15° S. and 120° E., similar difficulties arise: they can be solved only by supposing that the island Gulliver was searching for lay northwest of Australia, and that the ship which rescued him was on a southwesterly course—in other words, by changing the text in four places in order to achieve a result which is uncertain and unsatisfactory.

In all the speculation about this crux it does not seem to have occurred to anyone to question the accuracy of the first datum of all. Yet if one supposes that Lilliput lies northeast of Tasmania, instead of northwest it, all difficulties vanish. The position thus indicated lies in the South Pacific about midway between Brisbane and the North Cape of New Zealand. Gulliver's project of sailing northwest from Blefuscu in the hope of reaching an island northeast of Tasmania is seen to be a thoroughly practicable plan, and the ship captain's southeasterly course is recognized as entirely normal, since traveling from the Orient to England by sailing around Australia was an accepted alternative to following the shorter but more difficult route through the Netherlands East Indies. This solution disposes of all difficulties so simply and completely that there can be little doubt of its correctness.

The geography of Brobdingnag is a much less complicated matter. Even Motte's engraver was right as to its approximate location, though he was characteristically careless in his scale, making the country far too small. Brobdingnag is a peninsula joined to the northern portion of the North American continent (2.4.1). Gulliver's observation (2.4.2) that it is "terminated to the North-east by a Ridge of Mountains" suggests that the main axis of the peninsula runs from southwest to northeast. The dimensions of the country (in which Swift is a little more generous than the map will allow) are "six thousand Miles in Length, and from three to five in Breadth." The most helpful data about the location are given in the last chapter of the voyage. The eagle which captured Gulliver carried him fifty leagues from the coast (presumably the southern coast) of Brobdingnag, and dropped him somewhere near 44° N. 143° E. According to Moll's map this position is just off the coast of Yezo, in the "Straits of the Vries." This would fix one part of the southern coast of Brobdingnag in the vicinity of 45° N. 143° E., though the southernmost tip may lie some distance further to the southeast. A coast line running west-northwest and east-southeast may be extended as far as 20° N. 170° W. without interfering with either Moll's geographical data or the lands described in the third voyage, as their positions are worked out hereafter. From this point the eastern coast line may tentatively be carried north-northeast to 40° N. 140° W., and thence in a still more northerly direction to 70° N. 90° W., from which point the high mountain range which, according to Gulliver, separates Brobdingnag from the rest of North America, may be supposed to run northwest. The coast line, however, must be pure conjecture, since we are given only enough information to locate the points of Gulliver's arrival and departure, which lie very close to each other. The storm at the beginning of the second voyage caught the crew of the Adventure just east of the Moluccas, about three degrees north of the equator, and drove them a long distance east-northeast, after which a strong west-southwest wind carried them about five hundred leagues to the east. Such a course could easily have brought them to a position off the southern coast of Brobdingnag if it lay in the region suggested above. No doubt the point at which the landing was made was east of Flanflasnic, which seems to have been close to the southwest tip of the country. Such a location, of course, disposes of the hypothetical "Company's Land," but Swift would have felt no compunction over depriving the Dutch of territory.

It has been remarked that the dimensions of Brobdingnag are too large to be fitted into the map of the world as known to the eighteenth century.¹⁰ The exaggeration was not, after all, very serious. The southern coast, as described above, would have given the country a width of three thousand miles, which might have been increased, farther up the peninsula, to well over four thousand. If the North American shore line above Cape Blanco bent backward to the east, so that the dividing mountain range was on the far side of the pole, a long axis of a good deal more than five thousand miles might be secured. Swift may, of course, have been careless in his calculations: he may even have been deceived as to the area at his disposal because he used a flat map instead of a globe. Most probably, however, he was not so much concerned with precise geography here as he was with providing the Brobdingnagians with a country suitable to their size: note, in the same passage, the mountains thirty miles high. In the first voyage Swift had ignored his scale deliberately for the sake of a picturesque incident—the troop of cavalry exercising on his handkerchief—and this slighter deviation for a more important purpose would probably not have worried him. If this is the case, however, it is pertinent to inquire why he chose just these dimensions as adequate for a country of giants. If one divides them by twelve, in accordance with the scale, they become five hundred miles in length and from two hundred and fifty to four hundred sixteen and twothirds miles in breadth. The long axis of the British Isles, from Cape Wrath to the Isle of Wight, is about five hundred and sixty miles. The greatest width, taken at right angles to this axis, is from Bray Head, in Ireland, to Scarborough, about four hundred and fifty miles; elsewhere the width is commonly between two and three hundred miles. The figures are at least suggestive.

There is, however, one difficulty about the geography of Brobdingnag which becomes manifest only at the beginning of the third voyage. Here we learn (3.1.8) that Gulliver was boarded by pirates in the neighborhood of "the Latitude of 46 N. and of Longitude 183" (that is, 177° W.). This point, if previous calculations are correct, lies within the boundaries of Brobdingnag. But there are several good reasons for suspecting that the bearings thus given are inaccurate. In the first place, even if this position did lie within the ocean, it would have been an odd spot for eighteenth-century pirates to lurk about in search of prey. In the second place, it is hard to see how Gulliver's sloop could have reached this place by the course he describes. Finally, the bearings are inconsistent with the movements of Gulliver during the remainder of the third voyage. We have once more to do with an initial error which has been accepted without question, and which has consequently darkened counsel.

Motte's engraver outdid himself. He began by placing Balnibarbi (which he makes an island rather than a part of a continent) in latitude 43° N., a few hundred miles due east of Yezo. He was then confronted with the information (3.7.1) that Luggnagg lies northwest of Balnibarbi, and that this island, in turn, lies south-east of Japan. Lest there should be any doubt about this last point, Gulliver gives the exact position of Luggnagg—29° N. 140° E. The engraver compromised: he drew Luggnagg southwest of Balnibarbi, so that it lay southeast of the northernmost part of Japan, and a good ten degrees east of the position given by Gulliver. His confusion of mind led him to commit the additional errors of attaching the island of Glubbdubdrib to Luggnagg, and placing the port of Maldonada in the latter island as well.

The process of working backward from data given late in the voyage produces, however, a clear and consistent picture. To begin with, the position of Luggnagg is described so circumstantially, and agrees so well with Gulliver's description of his voyage from thence to Japan, that we may feel ourselves on fairly safe ground. Balnibarbi is to the southeast of Luggnagg, and at a considerable distance from it, since Gulliver's voyage between the two countries occupied a month. Still, the length of the voyage was due in part to unfavorable winds, and perhaps we should look for the port of Maldonada in the neighborhood of 24° N. 149° E. This port is situated on the southern or western coast of Balnibarbi, since the island of Glubbdubdrib lies southwest of it. Lagado is about a hundred and fifty miles from Maldonada, and apparently south of it. Now the island from which Gulliver was rescued by the Laputans was ninety leagues (about three hundred and fifty miles) southwest by west of Lagado (3.2.8), and consequently about four hundred and fifty miles southwest by south of Maldonada, if the previous suppositions have been correct. By this reasoning Gulliver was marooned near 20° N. 145° E.—a long way from 46° N. 177° W. Which of these positions agrees more closely with Gulliver's account of his movements before he encountered the pirates?

Three days out of Tongking on a trading voyage, Gulliver's sloop was caught in a storm which drove it for five days, first northeast and then east, after which the weather became fair, but with a strong gale from the west for ten days. It is clear that the only northing the ship made occurred during the first part of the five days' storm. It is equally clear that no storm could drive a sailing vessel twenty-five degrees northward (more than sixteen hundred miles, even if the ship had no eastward motion) in five days. The commonly accepted position for Balnibarbi, east of Yezo, is an impossible one.

Once more, then, it is necessary to follow Gulliver's story with a map before us, in order to discover what Swift really intended. The Gulf of Tongking is a partially enclosed body of water from which Swift desired Gulliver to be carried into the Pacific. The storm which bore the sloop first northeast and then east (3.1.4) was designed to bring it first opposite Hainan Strait and then eastward through that passage into the open ocean. A ship on such a course would be following the twentieth parallel closely: ten days' progress due east with a strong following gale, partly of storm force, might easily bring it to 144° E. (a run of about thirteen hundred miles), although the precise longitude of the scene of the capture of the sloop may have been a few degrees east or west of this position. That the latitude was not far from 20° N. seems almost certain. How the erroneous "Latitude of 46. N. and of Longitude 183" found its way into the text it is impossible to explain: it may be an overlooked detail surviving from an early draft; it may be a mere error of transcription or a printer's mistake.

The final geographical problem, although it is the simplest of all, has nevertheless led to differences of opinion. In this case the customary geographical details are wanting at the beginning of the voyage, and calculations must be based on a brief passage near the end (4.11.3). From this we learn that Gulliver surmised that Houyhnhnmland lay west of New Holland (Australia), and about 45° S. He therefore decided to sail due east, in the hope of coming either to the southwest coast of Australia, or to some island to the west of that continent. To his surprise, two days' sailing brought him to the southeast point of Australia. Dennis as sumed that this was a mistake for southwest,¹¹ and in this opinion he has been followed by other commentators. The last part of Gulliver's account, however, confirms the reading of the original text. To begin with, the southwestern point of Australia is approximately 34° S., not 45° S. More important is the fact that if this had been the region intended, the geography would require further revision. After rounding the cape which he had first reached, Gulliver saw the ship which eventually rescued him approaching from the north-north-east (4.11.6)—a patent impossibility by either eighteenth- or twentieth-century maps if Gulliver had been near the southwestern tip of Australia. What Swift really meant was that Gulliver had reached the southern point of Tasmania, which, on Moll's map, lies about 44° S., about half a degree further south than it does in actual fact. Swift, in this passage, commits himself on two points which had not been finally determined by eighteenth-century cartographers: that Tasmania was joined to Australia, and that its southern extremity was "at least three Degrees" west of the location shown on Moll's map. Houyhnhnmland, therefore, lies a short distance due west of the southern tip of Tasmania, at about 44° S. 142° E. Motte's engraver was less at fault than usual in placing it at about 40° S. 125° E.

The foregoing geographical scheme for Gulliver's Travels requires only two alternations in the text of the first edition. The time-scheme of the book is more detailed, and cannot be straightened out quite so easily, though the serious difficulties are confined to the third voyage. The second and fourth voyages are extremely simple. There is a slight slip at the beginning of the second. Gulliver returned from Blefuscu on April 13, 1702, and, as he remarked in the last paragraph of the first book, he sailed away again after two months. The actual date of his departure, as we learn from the first paragraph of the voyage to Brobdingnag, was June 20, but in the same paragraph his stay in England is given as ten months. The discrepancy was corrected by Motte in the fourth octavo, though Ford failed to note it in his emendations. The same paragraph contains another minor chronological error. The twenty-day gale which set in on April 19 is said to have ceased before May 2. There is a considerable probability that the former date should read "the 9th of April." Printers habitually read off a fairly long phrase or sentence from a manuscript and set it from memory: if the compositor of the Travels worked in this manner it is easy to understand how the similarity of "9th" and "19th" could have brought about the mistake. Another possible example of this kind of slip occurs near the end of the third voyage.

The remainder of the second voyage presents no difficulties. Gulliver landed in Brobdingnag on June 17, 1703, and was taken to the farmhouse on the same day. He began his tour of the country "upon the 17th of August, about two Months after [his] Arrival," (2.2.6) and reached Lorbrulgrud, after ten weeks, on October 26 (2.2.7-8). Here, after a "few Weeks" (2.3.1), he was bought by the Queen. There are no further statements about the times of Gulliver's adventures until the second paragraph of the last chapter of the voyage, where he observes, "I had now been two Years in this Country; and about the beginning of the third, Glumdalclitch and I attended the King and Queen in a Progress to the South Coast of the Kingdom." This progress must have taken place during the summer of 1705, and the arrival at Flanflasnic, at the end of the journey (2.8.3), presumably occurred late in the summer. This agrees exactly with the only other pertinent bit of information which Swift gives us. " … I never went out of the Ship," Gulliver asserts, "till we came into the Downs, which was on the 3d. Day of June, 1706, about nine Months after my Escape."

The chronology of the last voyage is equally simple. Once again there is a minor discrepancy between the dates and the elapsed time of Gulliver's sojourn in England between voyages. He returned from Laputa on April 10, 1710, and sailed from Portsmouth on the second of August following (3.11.7: 4.1.1): this is a little short of four months, instead of five, as the first edition has it. Gulliver was marooned in Houyhnhnmland on May 9, 1711 (4.1.3): after this there are no clues to the calendar until the latter part of the book. The quadrennial assembly which decided upon the exile of Gulliver met at the vernal equinox, that is, about September 21, since the country lay south of the equator (4.8.16), and about three months before Gulliver's departure (4.9.1). This last is a loose statement: apparently the three months' interval occurred between the assembly and the notice to depart given to Gulliver by his master, which on this supposition should be dated December 21 or thereabouts. Gulliver was allowed two months to build a boat, but completed its construction in a little more than six weeks (4.10.9, 12). The date given for his departure, February 15, 1715 (4.11.1) is consequently in strict accord with his other statements.

Two passages elsewhere in the text have led to some misunderstanding. Gulliver begins the fourth paragraph of the eighth chapter: "Having lived three Years in this Country, the Reader I suppose will expect that I should, like other Travellers, give him some Account of the Manners and Customs of its Inhabitants…. "In the seventh paragraph of the twelfth chapter, recounting his capture by the Portuguese sailors, Gulliver says, "I told them, I was born in England, from whence I came about five Years ago…. " There is no discrepancy between these two statements. Gulliver was speaking in round numbers. His voyage had lasted more than nine months before he was marooned. He dwelt with the Houyhnhnms precisely three years, nine months, and six days; when he was picked up by the Portuguese, on February 20, 1715, he had been absent from England exactly four years, six months, and eighteen days—a little nearer five years than four. The editor of the edition of 1735, not remembering these facts, emended a passage a little later on (4.11.10), in which Gulliver spoke of his three years' residence in Houyhnhnmland, to read "five years' residence." This mistake persists in many modern editions.

The time-scheme of the first voyage falls into two parts, each detailed and coherent in itself, and not inconsistent with the other. Gulliver sailed from Bristol on May 4, 1699 (1.1.4), and on the return from a prosperous voyage to the Pacific Ocean his ship foundered off the coast of Lilliput on November 5. The year is not stated, but subsequent events show that it was 1700. On the morning after the wreck Gulliver found himself in captivity, and on the next day he arrived at Mildendo. About three weeks later (November 28, approximately) he had learned the language of Lilliput sufficiently well to permit him to converse with the Emperor (1.2.6), and to be partially intelligible to the officers who searched his pockets. Their inventory is dated "the fourth Day of the eighty ninth Moon" of the Emperor's reign (1.2.9). The articles by the terms of which Gulliver gained his freedom are dated "the twelfth Day of the Ninety-first moon" (1.3.18). "Moon," presumably, means a lunar month of twenty-eight days, in which case the date of Gulliver's release came at the end of January or the beginning of February, 1701. "About a Fortnight" later Reldresal made a visit to Gulliver to acquaint him with the political state of the kingdom (1.4.4). Here, in the middle of February, the first series of dates comes to an end: there is no clear indication of the length of time spent by Gulliver in maturing his plan for the attack on Blefuscu and carrying it into effect. A second series of dates begins, however, with Gulliver's capture of the Blefuscudian fleet. The plot against Gulliver broke out "less than two Months" after he had refused to help enslave Blefuscu, which he seems to have done immediately after his great exploit (1.5.4, 5). The ambassadors from Blefuscu arrived in Lilliput "about three Weeks" after their naval disaster (1.5.6), but this period seems to have been overlapped by the two-month interval just referred to, and if this is true it may be ignored in the chronology. A "considerable Person at Court" warned Gulliver of the plot just before it was to have been put into execution, and in consequence Gulliver fled to Blefuscu three days later. The exact date of this flight is fixed by the statement (1.6.19) that Gulliver resided in Lilliput for nine months and thirteen days: he departed, therefore, on August 18, 1701. Reckoning backward about two months (the duration of the plot) gives the middle of June as the time of the falling-out of Gulliver and the Emperor. By this calculation the whole affair of the fleet occupied about four months: but this period may have been shorter if it took Gulliver some time to persuade the Emperor to order the inventory, in which case the dates immediately preceding the naval victory should be set somewhat later.

The remainder of the calendar is straightforward. Gulliver discovered a derelict boat on August 21, three days after his arrival in Blefuscu (1.8.1): he spent ten days in making paddles, and brought the boat into the royal port for repairs (1.8.2). He finished these in "about a Month" (1.8.7), and set sail on September 24, 1701 (1.8.9). Two days later he was picked up by an English vessel, which eventually landed him in the Downs on April 13, 1702.

The chronology of the third voyage is by far the most complex and unsatisfactory. The Hope-well sailed from England on August 6, 1706, and is said to have reached Fort St. George (Madras) on April 11, 1707: here the ship stayed for three weeks, and then spent an unspecified time in sailing to Tongking (3.1.3). Shortly after the arrival in that port Gulliver set out in a sloop, and eighteen days later he was captured by pirates (3.1.4), who set him adrift. Five days later still he was found by the Laputans and taken up into their flying island (3.1.8-10). He learned the language in "about a Months time" (3.2.18), and spent two months there all told (3.4.2). As the date of his departure is given specifically as February 16 (3.4.6), it is possible by reckoning backward two months and twenty-three days to set November 24 as the approximate time of departure from Tongking. Working still farther backward, we find that the voyage from Madras to Tongking occupied the period from the beginning of May to the middle of November: this is perhaps a little long, but not at all implausible. Going forward from February 16 again, Gulliver spent "a few days" with Lord Munodi (3.4.16), and one or two more in the Grand Academy of Lagado. Perhaps a week more was spent in the overland journey of a hundred and forty miles to the port of Maldonada (3.7.1). Gulliver then spent ten days on the island of Glubbdubdrib (3.7.6), and another fortnight in Maldonada waiting for his ship, which carried him to Luggnagg in a month. This all adds up to something over two months, which agrees with the date given for the arrival in Luggnagg—April 21 (3.9.1).

At this point difficulties begin. The year should be 1708, but in the first edition it is 1711, which is clearly wrong. The edition of 1735 made the apparently obvious correction to 1708, ignoring the fact that Gulliver, after what is clearly a short stay in Luggnagg (he had not even time to learn the language), departed for Japan in 1709. One easy way out of the maze is to substitute 1708 for 1707 as the year of the arrival in Madras and 1709 for 1711 as that of the landing in Luggnagg. But this gives an abnormally long voyage from England to Madras,¹² and adding a year to the time spent between Madras and Tongking is equally unsatisfactory. For the moment it will be well to adopt 1709 as the year of arrival in Luggnagg and take a fresh start.

The sojourn in Luggnagg, brief as it is, produces its own crux. Gulliver landed on April 21, and left on May 6, 1709 (3.11.4). In the meantime, however, he observes (3.9.7), "I stayed three Months in this Country…. " This statement, which is in a passage not primarily concerned with the chronology, and which is some distance from other statements of that kind, may be due to an oversight in revision: there is no possible way of harmonizing it with the text.

Three weeks after Gulliver's departure from Luggnagg he landed in Japan at the mouth of Tokio Bay (3.11.4). He was immediately sent to Yedo (Tokio), about forty miles away, and his audience with the Emperor probably took place on the following day, May 28. He was then convoyed south by a body of troops on the march to Nangasac (Nagasaki), in which city he arrived on June 9, 1709, "after a very long and troublesome Journey." The distance, by Moll's map, is over five hundred miles: in actual fact it is a good deal greater. The average daily distance traveled would therefore have been more than forty miles—a practical impossibility if much of the trip was made overland. But as Yedo and Nagasaki are on different islands, some of the traveling, and perhaps most of it, must have been by sea.

From Nagasaki Gulliver sailed to Amsterdam in the Amboyna, and thence to England in a small vessel. According to the first edition, the landing in Amsterdam took place on April 16, 1710, and the arrival in England on April 10 (3.11.6, 7). Some editors have assumed a simple transposition of these dates by the printer: others have preferred to change the second date to April 20. It seems more likely that this is another instance of a compositor's error like that conjectured at the beginning of the second voyage: if so, the true date of the debarkation at Amsterdam was April 6, and the similarity between "sixth" and "sixteenth" led to the confusion in the mind of the typesetter.

The final difficulty with the chronology of this voyage is found in the last paragraph, which stated, in the first edition, that Gulliver's absence from England had been for "Five Years and Six Months compleat." The elapsed time from August 5, 1706, to April 10, 1710, is a little over three years and eight months. The mistake suggests one possible reason which may underlie all the chronological discrepancies of the voyage. It is worth remarking that the other serious errors occur in the ninth and tenth chapters, which are concerned with Luggnagg. If in the original draft the voyage had been intended to occupy over five years, and especially if a visit to another country had intervened between the stay in Balnibarbi and the visit to Luggnagg, then the date "1711" (3.9.1) and the other and lesser discrepancies in these chapters are easily explained as oversights in revision. The explanation is tempting, but it must be admitted that there is no external evidence to support it.

Only one statement remains to be discussed. "Thus, Gentle Reader," says Gulliver at the beginning of the twelfth chapter of the last voyage, "I have given thee a faithful History of my Travels for Sixteen Years, and above Seven Months." Here, at least, Swift was on his guard: the voyage to Lilliput began on May 4, 1699; the final return to England took place on December 5, 1715. This accuracy is more characteristic of Swift's dealing with time in the Travels than is the carelessness which is evident in a few places at the end of the third voyage. One possible explanation of the major discrepancies has already been proposed, but of course any or all of these, as well as the trifling mistakes, may have been due to copyists or typesetters. Swift's complaints in the Letter to Sympson have a ring of sincerity: "I find likewise, that your Printer hath been so careless as to confound the Times, and mistake the Dates of my several Voyages and Returns; neither assigning the true Year, or the true Month, or Day of the Month: and I hear the original Manuscript is all destroyed, since the Publication of my Book." (Par. 4.) The natural question, of course, is: If this was the case, why did not Ford include the correct dates in his list of emendations, or in his corrected copy of the first edition? To this there can be no positive answer: the most probable is that the errors had already crept into the manuscript, which Ford followed mechanically in his collation, not concerning himself even with the most glaring chronological inconsistencies. There is at least some evidence for this in the fact that he passed over the date "1711," in 3.9.1, and the obvious contradiction of dates at the end of the third voyage which brought Gulliver home six days before his intermediate sojourn in Amsterdam.

Notes

¹ Craik, Sir Henry, Selections from Swift, London, 1893, pp. 441-4.

²Gulliver's Travels (ed. Dennis), in the notes, passim.

³Gulliver's Travels (ed. Williams), pp. lxxix, lxxx.

⁴Gulliver's Travels (ed. Williams), pp. 459-90 passim.

⁵ "The Geography of Gulliver's Travels" J.E.G.P., 40.214 (1941).

⁶ See Williams, Harold, Dean Swift's Library, Cambridge, 1932.

⁷Correspondence, 3. 134, 137; The Letters of Swift to Ford, pp. 36-7.

⁸ This map, which is included in various atlases, was apparently the most recent Moll map of the world at the time when Swift began the composition of the Travels, though an examination of other Moll maps between 1709 and 1745 indicates that the cartographer made no changes in his data during this period which would have affected Gulliver's geography.

Frederick Bracher has recently published (Huntington Library Quarterly, 8.1.59 [1944]) an article entitled "The Maps in Gulliver's Travels," in which he shows that the artist who drew the maps in the first edition used Moll's map of 1719. He agrees with the present author that neither Swift nor any of his friends was in any way responsible for the 1726 maps with their numerous errors, and gives reasons for believing that they were drawn by John Sturt and engraved by Robert Sheppard.

⁹ The map of the Gulliverian hemisphere in this volume is based on Moll's map of the world.

¹⁰ Moore, John R., "The Geography of Gulliver's Travels," J.E.G.P., 40.217-20 (1941).

¹¹Gulliver's Travels (ed. Dennis), p. 295, note.

¹² The late Professor Walter Graham pointed out to me that eight months (as indicated in the first edition) was the normal time for a voyage from England to India, according to the eighteenth-century records of the East India Company.