"The Columbus Landfall at Samana Cay" by Joseph Judge Senior Associate Editor, "National Geographic" Magazine Finding the landfall of Columbus at Samana Cay in 1492 is not difficult for it is the landfall the log describes, as Captain Augustus V. Fox demonstrated a century ago, and as the work of the National Geographic team confirmed. As Fox concluded: "The track which I have laid down (from Samana Cay) was chosen because it appears to be the only one that can be made to fit the courses, distances, and descriptions in the log book." That was true in 1881. It is true today, because it is true in the real world. We did not begin with the premise that the problem had been solved, however, only with the premise that it could be solved. Despite expected errors of transcription in the log, and despite the vagaries of navigation in 1492, the general outline of the track through the islands is clear, and sufficient numbers of geographic features are given to enable us to connect the track to modern geography. We believe that we have established the landfall within any parameter of reasonable doubt. The track from Samana Cay is invariably and almost completely as described by the log. We isolated 30 distinct points of reference--estimates of distance, trending and length of coastlines, location of villages and anchorages, bearings of the fleet--and found that 28 of them are verified by modern geography and archaeology. An East-West Voyage The voyage from San Salvador to Cuba is, in the main, east- west, with legs that begin about 20 miles somewhere to the south of San Salvador: westward along Santa Maria de la Concepcion's 30 mile coast, westward across a sea passage for about 27 miles to Fernandina and westward from Isabela across the open sea for about 70 miles to the Sand Islands. A return leg of southeast-east from Fernandina to Isabela of about 30 miles confirms the 27-mile west passage. San Salvador thus lied about 100 miles east of the Sand Islands and about 55 miles east of Fernandina. It lies about 20 miles north of Santa Maria and more than 30 miles northeast of Isabela. The solution to the landfall, obviously, is to find a place to attach this general track to the modern Bahamas. The Sand Islands are certainly today's Ragged Island Range. If the western terminus of the general track is anchored to the Ragged Islands, it is immediately apparent that the landfall was at Samana Cay. It cannot be otherwise without assuming gross error throughout the log. Many, struggling with false landfalls, have made just that assumption and have devised tracks that depend upon clerical errors and unique translations. Invariably, those tracks begin with a selected landfall island, which is going about the problem in the wrong, inductive way. We believe that the landfall island should be the last island identified after a number of deductive steps. Thus, we began at either end. Luis and Ethel Marden calculated the transatlantic track from Gomera, and I looked for a known point from which to begin a backward track, which, to me as to Morison, is the logical path to the landfall. By agreement, neither part of the team revealed its progress to the other. The Backward Track It is agreed that the Ragged Islands are the Sand Islands so the backward track begins there. The fleet raised the islands on the afternoon of October 25 from about 14 miles east, but we have no way of knowing what latitude they were on. We do have excellent estimates of distance made good from Isabela, from which the ships had departed at midnight of October 24. The log gives us a day's run on October 25, from dawn until 3 P.M., of between 45 and 47 miles. On the previous night, between dusk and dawn, they had been in a tropical downpour and Columbus estimated they had made only about 6 miles. On that dusk, however, he had estimated his position as about 20 miles southeast of Cape Verde, the southwestern cape of Fernandina. Since we know the heading away from Isabela, west-southwest, this estimate of position gives us the first opportunity to attach the track to modern geography since the bearings terminate in specific points of geography--a high, green southwestern cape to the northwest, and a small island, perhaps two, lying at the north end of a larger island, Isabela, distinguished by a high and visible promontory. In terms of modern maps, we are looking at the cape still called Verde on southwest Long Island, and at French Wells, where two small islands lie off the north end of Fortune Island, or Long Cay. Since the voyaging around Isabela begins and ends at the same place, the Cape of the Small Island, it is of no consequence to the backward track. We look next to the reciprocal legs of the voyage from Fernandina to Isabela, and this takes us west, then northwest to the southwestern end of Long Island. Reversing the "around the island" direction given in the log brings us next to the east coast of southern Long Island, where three features--a village anchorage, a harbor with an island in its mouth, and an east-west coast, must be present--as they are, at Adam's Hole or Burrough's Cove, at Little Harbor, and at Strachen Cay westward. Next, eastward across a 27 mile sea passage to the western end of a coast that runs eastward for another 30 miles. It is the north coast of Crooked and Acklins islands, a single coast unified by a reef. There is no other coast in the central Bahamas that fits this description. Twenty miles north from the Northeast Point of Acklins lies the landfall--Samana Cay. By the testimony of the log, the backward voyage leads only to Samana Cay. The Transatlantic Track The Mardens had made an initial finding that astonished them as it would any seafaring person. The transatlantic track drawn for Morison by Admiral John McElroy had not been adjusted for leeway nor, more importantly, for current. It was, in effect, a railroad track from Gomera to south of Watling Island. The Mardens plotted the daily course from the log's headings and estimates of distance made good, adjusted for current by reference to the Atlantic Pilot Charts, adjusted for leeway based on the estimate of ship's performance by a British naval architect, and corrected for magnetic compass variation based on the figures of von Bemmelen. Dr. Marc Auslander of the Thomas Watson Research Institute, also a sailor, checked the computations. The Marden track ended 10 mile northeast of Samana Cay. It is a less reliable guide than the backward track because of the uncertainty over magnetic variation in 1492, but it is the first corrected track and it indicates that Samana Cay is the most probable target. Samana Cay, The Island Other than Fox, no historian or archaeologist who ever wrote on the landfall had ever been to Samana Cay--not one. Known for many years as Atwood's Cay or "Samana of the French" at a time when Crooked Island was shown as Samana, the island is nine miles long on an east-west axis and about a mile wide at its widest part. It is shaped, as Oviedo said of the landfall island, exactly like a bean pod. Two cays, the nearest one larger, extend to the east and are enclosed with the rest of the island by a reef. The north coast is very foul ground but in the middle of the south coast there is a large elliptical lagoon, called Sanbo Bay, enclosed by the reef and two small cays and the coast of a small peninsula. Our expeditions to the island convinced us that it is a remarkable match for the features Columbus cites--namely, it is very green, the historic use of the island being as a source for cascarella tree bark; it is very flat, averring about 2 feet across most of the center; it is encircled by the reef with openings on the south side that permit access to good anchorages; it has many waters, some 80 lakes, ponds, and water holes; and it has good candidates for each of the three distinctive pieces of geography the log notes--a "laguna en medio," a port area, and a piece of land made like an island that is not one. The laguna is probably the most famous of these features. The presence of large lagoons in the middle of watling had been advanced for years as one of the strong arguments for that island-- and one must ask, by the way, why, if Columbus landed on Watling Island after 33 days at sea, he did not fill his water barrels. On Samana, I have identified the long lagoon that lies behind the dune line above the beach at Sanbo Bay as the lagoon he would have seen from his anchorage on the southwest coast, just west of the Bay. Dr. Robert Fuson, however, believes that the obvious is true and that the large lagoon is the large lagoon. The port area described in the log lies between the reef and the island; on Samana it is the 4 mile long area between the reef and the coast lying east of Propeller Cay, one of the two small islands that help enclose the lagoon. It is the same dimension as Graham's Harbor, the candidate on Watling. The piece of land made like an island lies, in my view, immediately to the east of the lagoon and makes up its eastern shore. It is in fact a peninsula that cannot be seen as one because a creek separates its northern end from the beach. Inside the tongue of land is a narrow lagoon, now lined with mangrove forest. Dr. Fuson believes, on the other hand, that the island that is not one is in fact one today--the nearest of the two eastern cays, separated now from the island by a shallow swale, and the exact geological duplicate of Cut Cay, the candidate on Watling. We had to show evidence of an aboriginal population at the time of Columbus and we found several sites on the dune line behind the beach above the lagoon identified by large quantities of Palmetto Ware, a ceramic made by Arawak-speaking peoples prior to 1520 and probably before that. We found trade ware from the earliest contact period, a part of a small ceremonial figure, and a large piece of Spanish Olive Jar of unknown date. There is nothing that can be shown on Watling that is not present on Samana Cay. We then took up the voyage to Cuba from Samana Cay. After a century of debate, we are somewhat embarrassed by its simplicity. Santa Maria de la Concepcion: 15 by 30 miles Columbus left Samana Cay on the afternoon of October 14 and began sailing southwestward, or southward--it matters little since a number of apparent islands began to rise on the horizon ahead. These were the hills of Acklins and Crooked islands, so many that he did not know at first which way to go. He selected what seemed to be the largest and arrived off the Northeast Point of Acklins Island too late in the day to find an anchorage in good light. While standing off and on that night, the ships were carried to the east by a tide and in the morning they could see the 15 mile eastern coast of Acklins down to the end of visibility at Creek Point, and a long coast running to the west which they followed for six hours, or about 30 miles, anchoring at sunset in the lee of the western cape at Portland Harbor. The log correctly estimates the distance between Samana Cay and Northeast Point at about 20 miles, and the distance down to Creek Point at about 15 miles, and the coast west to Portland Harbor at about 30 miles. Fernandina: 25 Miles West, Village, Harbor, East-West Coast The following day, late in the morning, the ships departed and sailed westward across the Crooked Island Passage, correctly estimated at about 25 miles, and stood off the coast outside Lower Burrough's Harbor. In the morning they went in to anchor at Adam's Hole or Burrough's Cove and visited with the Indians who had a village there, on the northern hillside. About noon they sailed north to Little Harbor and anchored off it while the ships' boats went into the harbor looking for water--Columbus having thought the divided mouth of the harbor was the mouth of a river. After two hours wait, they resumed sail north to a point opposite Strachen Cay where the coast turned west and ran past the later site of Clarence Town before resuming its northwestward trend. The log accurately gives these relationships and distances. When the wind veered to the northwest, the ships turned around and ran through the night, in heavy rains after midnight, setting courses to keep off the coast they had seen that day--SE, SSE, sometimes due E. As in similar conditions they would encounter on the 24th, they made only a few miles. At dawn it was still cloudy and threatening rain so Columbus set his course for the lee of the southeast cape of Long Island where he anchored until the weather cleared. After it cleared later in the day, he sailed "around the island" as far as Cape Verde and probably beyond and anchored without going ashore. Isabela: Cape of the Small Island, Cape of the Lagoon At dawn on the 19th Columbus fanned the ships around a SE heading and sailed for about three hours until landfall was made on Fortune Hill to the east. Turning to that coast, another three hours brought them to French Wells, where Columbus noted Rat and Goat Cays lying off the north point of Fortune Island. Thus far, the log has depicted accurately the distances sailed and the geography encountered from San Salvador to Santa Maria to Fernandina to Isabela; now the log estimate for the coast of Isabela is 12 leagues, or about 36 miles, west. We will return to this point. They followed the coast of Fortune to the southwest, pausing opposite Fortune Hill to find an anchorage but the waters off the shore were too shallow. Finally they closed the coast at the concave, white-sand shore of Walter's Bay. As some point, they moved the anchorage around the south end of Fortune to one where the large lagoon running up the eastern coast was visible. They proceeded, at dawn of the 20th, to try a crossing of the Bight of Acklins on an east and north heading to find the elusive village with the gold but eventually ran out of water. Columbus could see the line of small cays running southwestward and realized that was the long way around and he would be better off returning to French Wells and seeking a way east from there. The caravels made it to anchor that night but Santa Maria stayed at sea and the ships were reconvened at French Wells before noon of the following day, where they stayed because of a dead calm until midnight of the 24th. On a west-southwest course, they made the run past the position off Cape Verde to the Ragged Islands. All of this is given accurately in the log with the exception of the 12 league west coast of Isabela which Columbus estimated from his position at the Cape of the Small Island. Yet a 36 mile coast running west is an impossibility for these reasons: the ships would have picked up such a coast long before three hours on a SE course away from Long Island; there would be no northern end to such a coast; and the headings into the Cape of the Small Island, NNE, and away from it, WSW, would take the ships over land. A coast with a north point that is approached from the west, returned to from the southwest on a north-northeast heading, and departed from on a west-southwest heading, can only trend NE-SW. Other than this impossible coast, Columbus has given us a picture of Fortune Island. In fact, the log gives us a picture of a voyage from Samana Cay to the Ragged Islands so generally accurate that one could sail it today with only the log as a guide. Inexplicable Passages What and where, then, are the inevitable difficulties? The inexplicable passages in the log are these: Columbus notes that he sees a larger island to the west as he is sailing the coast of Santa Maria, and again notes, while anchored at the western cape of that island, that he was anxious to sail to this other larger island he had seen. Long Island cannot be seen while sailing west along the Acklins-Crooked coast nor can it be seen while sailing west along Rum Cay's coast. There are other extraordinary claims of visibility in the log, notably that of Babeque on November 19 when Columbus "sees" it from 65 miles to its west. Fox believed he was looking for the high cumulus clouds that build up over Bahamian islands. We know that the caravels wee forward of Santa Maria, had the duty of finding a good anchorage for the night, and could sweep out large areas of visibility--but we really do not know what he meant when he reports seeing an island that cannot be seen at the distance given. There is a famous missing word in the entry for the 17th, after the wind has changed and the ships have turned around and the log notes that something was too light to anchor by. Morison, Jane/Vigneras, Lyon, Thomas, and all other but Dunn and Kelley give the word "wind" in the missing position. Since this would definitely put an end to the track from Watling Island, which requires an heroic sail of 70 miles to begin in bad weather at this point, Dunn and Kelley supply the word "light." But the remark is made after the night's courses have been given; it would be superfluous to note that there was no light in the middle of a dark and stormy night. The Watling Track's Many Versions We naturally compared the completed Samana Cay track to the one from Watling Island which Morison had declared put an end to the problem and we could find nowhere along it any of the principal matches between modern geography and the log that are so evident on the Samana track. We also took note of the corrected versions of the original track--Morison correcting Morison, Obregon correcting Morison, Obregon correcting Obregon, Keegan correcting Obregon, and Keegan correcting Keegan. None of them provide a close correspondence between log descriptions and modern geography. Only a single island, Rum Cay, rises on the horizon; the apparent archipelago described by Columbus is missing. Rum Cay's actual measurements in modern nautical miles are 3 for the east coast and 7 for the south. To make them fit Columbus's given 15 by 30 miles requires two assumed clerical errors, one for each coast, and an assumption on the length of the Spanish league, in terms of miles, that would result in Columbus looking at a 3 by 7 mile island and measuring it as 5 by 10 of those miles. But all of the assumptions ever made cannot cover over the fact that it took the ships six hours to sail west from cape to cape on Santa Maria. In good weather they made 5 knots. They would have sailed the coast of Rum Cay in a little over an hour, not six. Once at the western end of Rum Cay. there is no place to anchor. Sandy Point, the given anchorage has no safe holding ground. The distance to Long Island from Rum Cay is far shorter than the log gives. There is no village with an anchorage west of Rum Cay on the windward and steep-to coast of northern Long Island; mariners are advised to keep off at least a mile and a half. If the harbor Morison called Newton Cay is taken as the divided harbor, the nearest cape is too close. It is, in fact, the sharp northern point of Long Island. there is no east-west coast at northern Long Island. It is simply astonishing to us that seaman-historians like Morison or his advocates could look at the foul northern point of Long Island, which ships are advised to give a three mile berth, and call it an east-west coast. It is not. Like the archipelago and the anchorage on Santa Maria, it is missing. Nor do we credit the 70 mile sail, unreported in the log, from that point all the way down to Cape Verde at the southern end of this very long island, beginning in weather similar to that in which the fleet would make only 6 miles, and in the face of Columbus's expressed intention to seek an anchorage, and the entry which follows--"After it cleared, I sailed ..." From Cape Verde, the headings given by the log--about 3 hours southeast and about three east--run to French Wells, not Bird Rock. Finally, we see no reason to move the position off Cape Verde that Columbus gives for the evening of the 24th, as both Morison and Obregon do, the first closer to the cape and the latter far to the west, with the bearing changed from NW to NE. We find the Watling track a complete failure in its accordance with the date given in the log, and we do not believe that because a prominent American historian said it has to be so that it is so. Obviously, to the objective mind, it is not so. We conclude that the good Captain Fox had it right. This opinion is shared by people who know the ground--the historian Robert Fuson, who has written more on the landfall than any other; George Lewis, who owns the western end of Rum Cay and pointed out several years ago there is no safe holding ground at Sandy Point; Dr. Keegan's former colleague Steven Mitchell, who concludes the Watling track is simply too flawed; Frank Snyder, commodore of the New York Yacht Club, who has sailed Samana waters; and many others. We regard the action by the Bahamian government at the meeting of Ibero-American Columbus anniversary committees in San Juan, during which Dr. Davidson Hepburn, the director of that government's Columbus commission requested Dr. Obregon to introduce a resolution naming Watling Island the "official" landfall noting more than pandering to tourism. The resolution was passed without discussion, comment, or debate--a bleak moment in landfall scholarship. the official newsletter of the Bahamian Columbus Commission hailed this action has "a noteworthy coup." Based on this action, the America 500 Group, which is organizing a sailing fleet for 1992, declares in its newsletter that the landfall "is a closed argument." This is depressingly familiar is it not? But it will have consequences. In the future a younger generation of scholars without axes to grind will have no difficulty fixing the landfall at Samana Cay, and as they look back on 1992 and the arrival of the reconstructed ships off Watling Island, it will be regarded with wry amusement as one of history's more embarrassing moments. Reprint permission granted by author.