"Columbus Was A Latitude Sailor!" by Arne B. Molander in "Encounters: A Quincentenary Review" (Summer 1990, pp. 12-15) If Samuel Eliot Morison had analyzed the voyages of Cabot and Cartier prior to his monumental work, Admiral of the Ocean Sea, he would never have bought the popular rationale that Columbus navigated by dead reckoning (DR). For when Morison later wrote "The European History of Discover: The Northern Voyages," he openly admitted that Cabot's 1497 "principle of navigation, the time-honored latitude sailing, is revealed by his latitude observations." Not only did Morison recognize the "time-honored" latitude sailing (LS) methods of Cabot and Cartier, but he also unabashed- ly credited them with extremely accurate trans-Atlantic results- only a four-mile error for Cabot and a three-mile error for Cartier! Given these LS performances, it is hard to understand why Morison had earlier assumed that Columbus missed his 1492 latitude by ninety miles, using dangerously inaccurate DR. The extreme accuracy difference between LS and DR derives entirely from the ways compass errors are monitored because these navigation methods are otherwise identical on an east-west course. Either system would simply have accumulated the distanc- es made good on each westerly heading. The DR pilot, however, might have detected compass variations by noting azimuth-angle deviations from Polaris, a measurement most easily made in the full darkness of midnight. In contrast, the LS pilot would have detected the integrated effect of compass variation by noting elevation-angle deviations of Polaris and the circumpolar stars, a measurement made during the brief twilight interval when the star and horizon are both visible. (Note that each of Columbus's three recorded compass checks occurred during either evening or morning twilights, as required by LS, rather than at the midnight preference for DR.) The best achievable accuracy for the suggested DR azimuth checks would probably be about 2 degrees. This estimate consid- ers that the pilot must first point to Polaris in its 3 degree circuit of the north celestial pole, then offset his point to account properly for the star's position in its circuit, next lower his arm 28 degrees in a true vertical to the horizon, and finally read this projection against a compass marked only at 11.25 degree intervals! While this somewhat daunting task aboard a rolling ship may have been feasible for a 150-mile coastal run in the Mediterranean (5 miles error), it could easily have generated a disastrous 100-mile latitude error on a 3000-mile ocean crossing. Another limitation of this correction scheme is that it merely reestablishes true north to 2 degrees accuracy. It does not give the pilot any indication of the course correction necessary to bring him back to reference latitude, nor does it measure the effects of current and leeway. On the other hand, the LS pilot easily could have maintained his latitude within 60 miles crossing the Atlantic by using the 1 degree accuracy of the quadrant to measure the height of Polaris above the horizon. But he almost certainly would have fine-tuned this reading with the altitudes of the brighter circumpolar stars as they brushed the northern horizon. The Arabs had compiled a comprehensive list of second magnitude circumpolar stars for each latitude and season of the year. They were so expert at LS that Ahmed Ibn Majid, a contemporary of Columbus, scolded some of his fellow pilots for ignoring small tabulated fractions because "they have said that an error of up to quarter of an isba (22 1/2 nautical miles) is not an error. But this is a distance, espe- cially when making landfalls on routes which are almost due east and west." At least four conditions are needed for precise LS naviga- tion. First, the ship's course must lie along a constant lati- tude for the circumpolar star to reach the same minimum elevation angle each night as it sweeps under the north celestial pole. Second, the circumpolar star must be of at least second magnitude brightness to be visible with 1 1/2 degrees of the horizon. Third, the circumpolar star for Columbus should lie between 24 and 27 degrees from the north celestial pole. Further from the pole, it would disappear below his horizon; closer, he would lose the precision afforded by brushing the horizon. Fourth, the ship's crossing must occur during that season when the circumpo- lar star brushes the horizon during twilight hours. All four of the conditions were met for Columbus by the constellation Ursa Major, whose corner star, Dubhe, would have skimmed the northern horizon at an elevation of 1 1/2 degrees in 1492. Using this marvelous natural stellar navigation system, is it conceivable that Columbus could have failed to notice elevation angle changes on the horizon equivalent to a quarter of the moon's diameter? Such an accessible level of precision would translate into only an 8-mile latitude error, thereby certifying Northeast Providence Channel as his gateway to the New World! Given that Columbus was able to navigate accurately by means of LS, we must now establish his need for this precision. I think he had at least three strong motivations for achieving such accuracy. First, there was the safety of a rescue fleet retrac- ing the course if an unknown reef had claimed the Pinta and Santa Maria in mid-Atlantic. The Nina could then have led rescuers along the identical track with LS navigation. On the other hand, the vagaries of compass, currents, and leeway would have forced a DR search team to zigzag across the trackless ocean. Second were the political issues raised by the 1479 treaty of Alcacovas which divided Iberian exploration grants along the 27th parallel. Columbus understandably declined to follow the birds into Portuguese territory despite Martin Alonso Pinzon's importuning on October 6. I think Columbus reluctantly clung to the 28th parallel while Morison already had him drifting aimless- ly across the demarcation line. Third was Columbus's proposal in the diario prologue to make a "new chart for navigation. . .picturing everything by latitude from the equinoctial line and by longitude from the west etc." How would Columbus have created such a chart if he did not even know his latitude with respect to the Canary Islands? What do the independent experts have to say about Columbus's method of navigation? As far as I know, they all believe he used LS rather than DR. In The Oxford History of Technology, Volume III, Eva Taylor eloquently explained why DR would not do for trans-Atlantic navigation when she wrote that "[t]he methods of navigation developed first for the enclosed Mediterranean Sea. . .were inadequate for the new ocean navigation of the early fifteenth century sponsored by the Portuguese Prince Henry the Navigator. . . . In effect, the new method of navigation pro- posed was what is termed 'running down the latitude'." In his "Conquest of the North Atlantic," G. J. Marcus summarized his chapter on navigation with "It cannot be too strongly emphasized that latitude sailing was the underlying principle of all ocean navigation down to the invention of the chronometer. . . . It was practiced by Arab pilots. . .Christopher Columbus on his voyage of 1492-93. . .Vasco da Gama and predecessors. . .John Davis. . .the Earl of Cumberland. . . . Many other cases might be quoted from this and the following century." The only contem- porary definition of his navigation method comes to us from Antonie Galvano's "The Discoveries." In this 1555 text, the Governor of the Moluccas (who had presumably arrived at his post by LS navigation across the Indian Ocean) wrote of Columbus". . .and some affirmed that they were the first that sailed by latitudes." These published positions are confirmed in private corre- spondence from three navigation experts, none of whom is biased in favor of a northern landfall. Tim Severin wrote me (Arne Molander) on 17 May 1983: "Yes, I agree with you entirely that the probability is that Columbus used latitude sailing on his first voyage of discovery. Indeed, I find it difficult to imagine what other method he would have preferred. It is the only method that takes account of ocean current, leeway, and any other variables which are inadequately covered by dead reckoning navigation." Thor Heyerdahl answered my query on 8 June 1981 with: ". . .It seems very likely to me that Columbus made use of latitude sailing when he pushed into the unknown ocean as this seems by far the most logical thing to do when you have rather primitive means of navigation." Finally, in an unguarded moment, Samuel Eliot Morison wrote on 17 November 1972: "Of course Columbus was attempting latitude sailing, but his journal shows very clearly that he changed course toward the end, to follow the birds." (Morison seems to have forgotten that the navigation argument focuses on the period prior to October 7 and has little to do with his later WSW course to follow the birds.) In summary, latitude sailing was a much more accurate navigation technique than dead reckoning. It had been introduced to the Iberian peninsula by Prince Henry in the mid-fifteenth century; stellar conditions were perfect for its use along the 28th parallel in October of 1492, and all independent experts believe Columbus would have used it. Accurate navigation would have led Columbus to a landfall at Egg Island in Northeast Providence Channel. From there a route to Cuba by way of New Providence, Andros, and Long Island matches the diario features with unassailable precision! Permission granted by publisher.