By Gunnar Thompson, Ph.D., July 1, 2000

Abstract (or go straight to the Article Text)

Discovery of an early 15th century world map is shedding new light on ancient tales of New World exploration.  The map in question is an authenticated Venetian map by Albertin DeVirga circa 1414.  This map includes a Hyperborean continent called “Norveca” with features recognizable as the East Coast of North America.  The accuracy of this coastline and the Nordic nomenclature suggest that DeVirga’s Hyperborean continent is a copy of the long-lost map from Inventio Fortunatae

Numerous 16th century geographers including Gerhard Cramer, Abraham Ortelius, and Richard Hakluyt gave credence to the testimony of a Dutch journalist who recorded the exploits of Franciscan friars in the New World.  According to the journalist, an English friar had traveled throughout “Dusky Norway” using an astrolabe to record the locations of places that he visited on the way.  The biographers of Columbus, Bartholomew Las Casas and Ferdinand Colon, were explicit in stating that the friar’s travelogue, Inventio Forutnatae, told of lands in the far west. 

Following the trail of the Inventio Fortunatae through history and into the s of map libraries has led to the identification of 13 maps showing the location of Florida before Columbus.  Temporal analysis of the coastlines on these maps shows that the macro-peninsula creeps downward from Arctic regions to Tropic seas.  By 1492, Portuguese maps show the macro-peninsula situated directly west of Europe on the Tropic of Cancer—precisely where Columbus> expected to find land.

Placement of the macro-peninsula within a few hundred miles of the actual location of Florida confirms that ancient mariners conducted scientific surveys of the East Coast of North America.  Early knowledge of the phenomenon of magnetic declination and use of astrolabes instead of compasses for mapping explains how ancient surveyors were able to make a scientific map.  The effort to make such a map can be traced through historical sources to English scientist Roger Bacon and Oxford University.


Article By Gunnar Thompson, Ph.D., July 1, 2000

 Medieval superstitions and biblical visions added colorful embellishments to the mariners’ tales of exploration.  When cartographers back home set about translating field reports into geographical coordinates, the resulting maps were often whimsical and misleading.  Nevertheless, ancient maps contain clues that reveal increasing knowledge of New World territories long before the traditional advent of European exploration in the 15th century.  

Many of the clues to ancient maritime exploration can be found along the northern borders of European maps in a region that is often called “Hyperborea.”

Hyporborea—Land Beyond The North Wind

According to one Roman myth, there was a temperate land at the North Pole where a race of happy people, the Hyperboreans, dwelled amidst orchards and meadows.  This land was variously called “Hyperborea,” the “Polar Regions” or the “Magnetic Isles.”  Presumably, this latter designation was derived from a belief that a huge magnetic mountain was situated at the top of Earth.  The magnetic force of this mountain (Figure 1) supposedly explained why lodestones pointed towards the North.[1] 

Apparently there was some disagreement among Roman philosophers regarding the nature of the Polar Regions.  By the time Claudius Ptolemy was writing his geography and making maps in the 2nd century, there was a popular belief that Earth had climatic zones.  Members of the Ptolemaic School ridiculed myths about Hyperborea because they were convinced that the Polar Regions were quite frigid and consequently uninhabited.  This controversy endured through the 14th century.  In 1266, the English scientist Roger Bacon insisted that the Hyperborean tale was accurate in spite of the climatic zones.  He felt compelled to invent the concept of a unique situation at the North Pole that enabled mountains to trap solar heat in order to give the Hyperboreans a warm place to live.  He made this effort, he said, in order to explain why Pliny the Elder and numerous mariners had identified temperate lands at the North Pole:[2] 

How far habitation extends north, Pliny shows through actual experience and by various authors.  For habitation continues up to that locality where the poles are located; and where the day lasts six months and the night for the same length of time.  Martin, moreover, in his description of the world agrees with this statement whence they maintain that in those regions dwells a very happy race.

Modern historians have criticized Bacon for his apparent ignorance of the Polar climate.  Actually, this is an example of a situation where the English scientist felt compelled to bend scientific explanation to the wisdom of experience which he always insisted must take precedence over theoretical models.  Thus, he credits “actual experience” for his unpopular stand that seems contrary to the theory of climatic zones.  In other words, he was aware of mariners who had used the compass to reach the Magnetic Polar Regions—located in the temperate seas near Hudson Bay.  Like his contemporaries, Bacon was unaware that there were two North Poles—a geographic pole situated at the frigid top of Earth and a magnetic pole situated south of the Arctic Circle.  Irish and Roman accounts told of voyages to temperate lands such as “Thule” in the northwest.  We also find confirmation in Geoffrey of Monmoth’s History of Arthur that mariners from Wales had reached the “Magnetic Isles” in the 6th century.[3]  So, Roger Bacon had plenty of “experiential” support for his belief that there was such a place as Hyperborea.  Even in his own time, English friars, fishermen, and traders were sailing to places in the far northwest that were variously called the “Icelandic Isles,” “Wineland,” “New Land,” “Dusky Norway,” “Great Ireland,” and “Greenland.”  All of these were titles for settlements, trading posts, or hillforts situated in North America.[4] 

Most Medieval maps included a region along the northern periphery of Europe that was called “Hyperborea.”  This placement followed Roman tradition that is seen in 15th century copies of maps by Pomponius Mela (1st century) and Macrobius (5th century)—Figure 2.  The Roman maps include a great gulf along the coast of Hyperborea that may well be an early representation for the Gulf of Mexico.  On the 1st century map, the gulf is shown near the North Pole.  It is called “The Caspian Sea.”  One region of this Caspian Sea that is mentioned in the geographical text by Pomponius Mela is called the Sinus Hyrcanius or “Gulf of Hurricanes.”  Aristotle describes hurricanes as being most common in autumn in his Meteorology Book II, Chapter 6.[5]  Since hurricanes are generally considered to be storms of the Western Atlantic or Caribbean, it makes sense to regard the Gulf of Hurricanes on the 1st century map as showing that region.[6]  This interpretation is confirmed by the Macrobius map which portrays the Caspian Sea as being directly across from Europe or in a region that the Romans regarded as “India Superior”—hence the “Indies” of Columbus.  The configuration of this western Caspian Gulf on the Macrobius map closely follows the actual shape of the Gulf of Mexico — with a peculiar macro-peninsula in the approximate location of  

Theories About Ancient Maps of the New World

The idea that ancient maps might actually show New World isles gained credence with the publication of a treatise on maps by the Argentine scholar Dick Ibarra-Grasso.[7] He noted that an eastern mainland on a map by the Roman geographer Claudius Ptolemy circa 140 AD was in the approximate location of Peru. Ptolemy called this mainland “Cattigara.” His map shows the mainland east of Asia across the Sinus Magnus or “ Great Gulf.” Since Magellan identified this “Great Gulf” as the Pacific Ocean, Ibarra-Grasso reasoned that Cattigara (being the eastern shore of the Pacific) had to represent Peru as it was known to the Romans.[8]

Several other scholars believe they have found evidence of New World lands on ancient maps. Mexican historian Gustavo Vargas Martinez has identified a map by Henricus Martellus Germanus (1489) which might include the coast of Brazil.[9] Hjalmar Holand has suggested that the Polar Region of Martin Behaim’s globe (1490-92) portrays isles in the vicinity of the Gulf of St. Lawrence.[10] In 1965, the British map historian, R.A. Skelton, and his associates at Yale University introduced the “Vinland Map” which they thought included the region of North America between Greenland and Cape Cod.[11]

These examples have limited value in a scientific study because they represent only isolated specimens. Evidence which Ibarra-Grasso, Martinez, Skelton, and Holland have given to support their theories tends to be anecdotal, literary, and conjectural. Each contender represents a different area of the globe: Peru, Brazil, the East Coast of North America, and the Polar Regions. Holand’s theory must take into consideration that Behaim’s globe has suffered from decay and questionable reconstruction.[12] The Yale map remains under a cloud of controversy in spite of reports from UC-Davis that the composition of its ink and paper were similar to those used in the Gutenberg Bible.[13] Lacking any systematic basis for comparison, these documents hardly challenge the popular belief that: “Columbus put the New World on the map.”

Following the lead of Ibarra-Grasso, researchers at the New World Discovery Institute began a review of cartographic materials seeking evidence to support or refute the theory of New World isolation in ancient times. The appropriate test of the theory seemed to be the date at which scientific cartography of the New World actually began. If Columbus was the first person to use scientific equipment and techniques to measure the location of Caribbean isles, then he was effectively the first person to begin the process of putting the New World on scientific maps. On the other hand, if someone else used scientific equipment in the New World at an earlier date, we would expect to find cartographic evidence.

Consideration of such a possibility was not unreasonable. Indeed, Romans had produced fairly accurate surveys of the Mediterranean region by the 2nd century—hence maps in the Ptolemaic tradition. These maps employ longitudes and latitudes for the accurate arrangement of land areas. The Chinese had produced highly accurate maps of Asia in the 12th century. One example of this craft, an inscribed stone tablet, is on exhibit at the British Museum. By the 12th century, Arabian scholars had devised sophisticated astrolabes, and they had improved upon Ptolemy’s map of the Oecumene by determining the accurate length of the Mediterranean Sea.

There was another reason for believing that there might be evidence of early scientific mapping of the New World. An English scientist, Roger Bacon, had proposed the creation of a scientific map of the world in 1266.[14] In accordance with Ptolemy’s guidelines, the map was to have a base meridian along with longitudes and latitudes. Bacon realized that contemporary knowledge of the world was inadequate, so he envisioned field expeditions to survey unknown regions (principally in the Western Hemisphere).

Although Bacon died before his proposal could be implemented, the concept of a scientific map inspired the deans at his alma mater—Oxford University. In the early 14th century, William Rede and Simon Burton trained cadres of friars in surveying and celestial observation with the intention of making Bacon’s dream a reality.[15] They established an astronomical observatory for making the essential baseline measurements of celestial phenomena, and they built a factory for the manufacture of astrolabes.

Oxford friars with astrolabes traveled about the Northern Regions from Norway to Greenland by the mid-14th century. It is hardly surprising that the Medici Atlas of 1351 shows the Arctic isle of Greenland (Aloland) for the first time in its correct position northwest of England. This positioning was dependent upon use of astrolabes. These instruments enabled the geographical placement of land areas by measuring their relationship to celestial objects. It was a more laborious process than using compass bearings to construct maps. However, maps based on compass coordinates led many mariners and cartographers to believe that Greenland was due north or even east of Norway. This misplacement was a consequence of the geophysical phenomenon of magnetic declination. Due to the location of the magnetic pole near Hudson’s Bay, isles in that direction including Greenland have an erroneous compass bearing of “north” when in fact their geographical location relative to Norway is almost due west.

According to testimonials by Gerhard Mercator, an Oxford Franciscan traveled in the so-called “Polar Regions” for the purpose of making a map during the 14th century.[16] Mercator’s information on the subject was derived from the journal of a Dutch traveler, Jacob Cnoyen. According to Cnoyen, the friar’s travelogue, Inventio Fortunatae, was completed circa 1360 and presented to King Edward III.

The Inventio Fortunatae (or Discovery of The Fortunate Isles) is not unknown to historians. Columbus mentioned the document in a letter to Bristol merchant John Day in 1496. The mariner’s son Ferdinand wrote that his father had seen the travelogue prior to his first voyage west. In addition to Mercator, cartographers John Ruysch (1508) and Cornelius Judaeus (1593) also attributed the Polar Regions on their maps to the Inventio. English historians made a fruitless effort to find the friar’s travelogue in the late 16th century. Since that time, it has been presumed lost.

Most historians have regarded Cnoyen’s account of voyages to the “Polar Regions” as a blatant fantasy. This is due to the extreme cold near the North Pole and the mythical design of the Arctic on maps that Mercator attributed to Roman philosophers and the Franciscan survey of the Far North. Mercator’s maps show a magnetic mountain at the North Pole surrounded by four gigantic isles. The magnetic mountain was a fantasy of Roman philosophers who sought an explanation for why magnetic compass needles pointed towards the north.[17] Several historians regard Cnoyen’s account of voyages to temperate isles beyond Iceland as a plagiarized version of the Arctic derived from Icelandic folklore.

Researchers at the New World Discovery Institute realized that the friar’s tale of Arctic exploration included elements of fact that deserved closer scrutiny. According to Mercator, the English Franciscan traveled to a place called “the magnetic regions.” Roman philosophers had speculated that the “magnetic regions” were situated at the North Pole—site of the mythical magnetic mountain. However we found testimony in the Nancy Manuscript (1424) that the pole of the Western Hemisphere was situated at 66°N.[18]  This second “pole” would have been a magnetic pole in the thinking of most Medieval clerics who had not yet conceived of Earth having a pole of rotation. But what was a magnetic pole doing so far south? 

We found our answer by tabulating the record of Earth’s wandering magnetic pole. When James Ross and William Peary identified the location of the magnetic pole in 1818, it was located near Lancaster Sound at 71°N-96°W. Subsequent field measurements recorded the location of the pole in 1903, 1945, 1959, 1963, and 1980. During this time, it moved northwest to 77°N-102°W.[19] Thus successive measurements indicate a generally northwestward movement of 6°N and 6°W during a time period of less than two centuries. By extrapolating the movement of the magnetic pole back in time, it is not unreasonable to suspect that the pole might have been even farther south in the 14th century. Although the report of a magnetic pole at 66°N in the Nancy Manuscript is probably a crude approximation of its actual latitude, it represents the earliest historic reference to the location of the magnetic pole. A competent surveyor who traveled west from Iceland in that era could have noted the discrepancy between the location of Polaris (in winter months) and the direction for north indicated by a magnetic compass.

Another intriguing idea emerged from our review of historical documents. We reasoned that the terms “magnetic regions” and “Hyperborea” might correspond to Roman ideas of temperate lands near Hudson’s Bay. Until the 16th century, most cosmographers regarded the magnetic pole as being congruent with the celestial pole, although a few forward thinkers had already calculated magnetic declination by the 15th century.[20] So it seemed possible to us that a failure to distinguish between the two kinds of poles (geographic versus magnetic) might have resulted in some of the discrepancies we encountered on ancient maps.

Since Medieval mariners depended upon magnetic compasses, it seemed like a reasonable possibility that geomagnetism and magnetic declination were factors that had to be considered in the study of ancient cartography. If ancient mariners mistakenly believed that isles found in temperate climates near the magnetic pole were located north of Europe, instead of in their true positions towards the west, then this geophysical phenomenon might explain the misplacement of New World isles in the Polar Regions on Medieval maps. Thus, Mercator’s account of 14th-century surveying expeditions in the Polar Regions suggested to us that this was the area on maps where we should look for evidence of a scientific survey of the New World.

DeVirga’s Map of Hyperborea

After nearly two years of examining old maps in s and portfolios in museum collections in Eastern Europe, Scandinavia, and the United States, our efforts were rewarded in November of 1994. We found a copy of what appears to be the Franciscan geographical survey hidden along the northern margins of a map by the Venetian cartographer Albertin DeVirga.[21] This map (Figure 3) is variously dated between 1411 and 1415—although 1414 seems to be the year that scholars use most often. Albert Figdor found the map in a Croatian antique store in 1911. Viennese Professor Franz Von Wieser authenticated the map in 1912 after comparing it to another DeVirga map in the Bibliotheque National in Paris.[22] The map was photographed in 1912 and again in 1932. Although the map has subsequently disappeared, the photographs contain sufficient details to confirm that scientific surveys of the New World took place prior to the 15th century.

The photograph, which we found in an atlas of antique maps, proved to be the key to our search for evidence of scientific maps of the New World. DeVirga’s map has two unusual continents. In addition to the customary Old World continents, there is a continent called “Norveca” attached to the northwest coast of Norway; and a second continent called “Ca-paru or Great India” situated southeast of Asia. Von Weiser assumed that the extra continents were marvelous impressions of Greenland and Japan.[23] But it seems more likely in the context of ancient cartography that we are seeing early versions of North and South America

Norveca was a Medieval name for a Norwegian province (a.k.a., Norbega). By the 16th century, this region of North America was known as Nor-um-bega which some writers regard as a clear derivative of the Nordic word Nor-bega. This region corresponds to the Roman legend of a Hyperborean paradise. The Oxford Survey of this region (Figure 4) follows the 14th century Oxford/Medici cartographic practice of showing Aloland (a.k.a., Greenland) west of Norway. This version of the Northern Regions, first seen in Medici maps ca. 1350, follows the known expeditions of English Franciscans to map the north. Although Venetians, Majorcan Jews, and the Arabs had the technical capability of making such maps, only the English Franciscans are known by historical documentation to have surveyed the region.

One historical document pertaining to the naming of New World territory as a Nordic province (Norveca) is the declaration of sovereignty by King Haakon IV of Norway-Sweden in 1261.[24] This declaration placed all the isles from the Baltic to the North Pole under Nordic jurisdiction. Among the Nordic “Polar Isles” were Iceland, Greenland, and Landanu (that is, “the New Land”). We can identify Landanu as an early title for the region of Newfoundland or Nova Scotia, because Icelandic sagas specify that Landanu was situated towards the southwest. Thus, historical and cartographic data combine to reveal that the Nordic concept of “Polar Regions” included temperate isles near the magnetic pole of Hudson’s Bay.

The most distinguishing feature of the Hyperborean continent on DeVirga’s map is a macro-peninsula that appears to terminate near the North Pole above a huge gulf called “the Caspian Sea.” A map by the 1st century Roman geographer Pomponius Mela has the “Caspian Sea” directly north of Europe near the North Pole—at about the same place where it appears in the DeVirga map.

Florida on Ancient Maps

After examining the DeVirga map, we suspected that the macro-peninsula of Hyperborea/Norveca might be an early version of Florida. We would naturally expect to find a macro-peninsula above a huge gulf on scientific maps of America, because that is the actual configuration of the New World shoreline. The apparent misplacement of this macro-peninsula near the North Pole on DeVirga’s map can be understood in the context of the infancy of cartography. In this case, the cartographer attempted to show land areas from two hemispheres within the confines of a single circle.  He also had to deal with the inherent difficulties encountered when portraying Arctic regions where the effects of converging meridians and magnetic declination are most pronounced.

Having identified one ancient map with a macro-peninsula in the shape of Florida, our next goal was to find more examples. We realized that a single map was subject to varying interpretations. On the other hand, a sufficiently large collection of maps would enable scientific analysis that would transcend the inadequacies of individual specimens. We proposed to examine a sample of maps on the basis of several parameters including provenance, date of the drawing, name of the land form, configuration of the shore, and geographical position as indicated with respect to longitude and latitude. If these parameters were consistent with real world characteristics, or if they showed significant relationships with each other, then we would have a basis for ruling out the random influence of Medieval fantasy.

Subsequent to our identification of Florida on the DeVirga map, we have found 13 maps dating before Columbus that feature a macro-peninsula across the Atlantic from Europe. Leading scholars have authenticated all of these maps; many of them can be found in atlases of the Late Medieval and Early Renaissance. Data from these maps are summarized in Tables I and II.

Table 1: Macro-Peninsula on Ancient Maps

author/map          date   source              lat. N   miles

Macrobius          440    Roman                45°      ?

Sanudo              1320    Italian                 45°      ?

DeVirga            1414    Ven./Fran.       polar?     ?

Yale Vinland    1440    Swiss/Fran.        38°      ?

Florentine         1447    Florence             40°      ?

Genoese            1457    Genoa                 40°      ?

Florida              2000    Mod.Atlas          25°   3,500

Table II: Portuguese Macro-Peninsula/Antillia

Bianco (a)         1436    Ven./Port.           35°   1000

Bianco (b)        1436    Ven./Port.          30°    1000

Mauro               1459    Ven./Port.           25°      ?

Toscanelli         1474    Flor./Port.          20°   4,500

Martellus          1489    Ger./Ven.            23°   4,000

Martellus          1490    Ger./Ven.            23°   4,000

Behaim             1492    Ger./Port.           23°   4,500

Florida              2000    Mod. Atlas         25°   3,500

The latitudes of the tip of the macro-peninsula can be estimated by comparing their locations to reference points in Europe. Estimating the distance of mainland west across the Atlantic is a more speculative matter as few of the maps give any indication of longitude or the equatorial distance between points of reference.

Cartographers adjusted the latitude of the macro-peninsula over time bringing it closer to the actual location of the southern tip of Florida (at about 25°N). Macrobius (440) and Sanudo (1320) placed the macro-peninsula at about 45°N. Henricus Martellus (1489 & 1490) and Martin Behaim (1492) brought the macro-peninsula down to the Tropic of Cancer (23°27’N) or just a bit south of the actual latitude of Florida. This is the kind of increasing accuracy we might expect if explorers and cartographers refined their measurements over time. We noticed that in Table II, maps attributed to the Portuguese or the cartographers working in their service have a particularly high degree of accuracy. The Bianco maps are only 5° too far north, although the longitude seems to be off by 2,500 miles. This error could be attributed to the infancy of cartography or simply the perspective of the map. By 1474, the Portuguese estimate of longitude is within a thousand miles, and by 1489, the latitude is right on the mark for Florida

This degree of accuracy with respect to both longitude and latitude can not be attributed to coincidence.  Indeed, historical accounts confirm the role of ancient New World explorers in the pre-Columbus effort to make a scientific map of the Western Isles. We have already considered the 14th century expeditions of Oxford Franciscans. Prince Henry The Navigator of Portugal and King John II sent numerous explorers west seeking the location of mainland referred to as “Antillia” or the “Isle of Seven Cities” in the 15th century.[25] Among these Lusitanian navigators were Goncalo Cabral, Joao Fernandes, Vincent Dias, Diego de Tieve, Dom Fernao, Joao Vogado, Jao Vaz Corte-Real and sons, Ruy Concalves, Fernao Telles, Fernao Dulmo, and Martin Behaim—to name a few. Columbus, himself, believed that the Portuguese had succeeded in reaching Antillia and the mainland beyond—although not by the route that he proposed.

Of the 13 maps in our sample, 5 are known to have come from Portuguese sources, and another 4 are from cartographers with Portuguese connections. These include the English Franciscans who served King Edward III. Their knowledge was apparently passed on to Prince Henry The Navigator who was the great-grandson of Edward III. Andrea Bianco and Fra Mauro were Venetians serving the Portuguese. Paolo Toscanelli was a Florentine whose correspondence with King John II and Columbus is sometimes credited with launching the “Enterprise of The Indies.” Martin Behaim, a German expatriate working as a cartographer in Lisbon, was associated with the same guild as Columbus and his brother Bartholomew. Henricus Martellus (a German monk working in Florence) made maps showing the latest Portuguese surveys of Africa

The maps in our sample manifest increasing accuracy with respect to longitude or the distance of the macro-peninsula west of Europe (Figure 5). A temporal analysis indicates that the location of the macro-peninsula gets closer to the actual location of Florida. Although most of the earlier maps lack sufficient reference points for estimating the distance separating the so-called “Asian mainland” from the west coast of Europe, it is possible in the later examples to estimate the distance. This can be done by multiplying the number of degrees of longitude across the ocean times the given distance of a degree (as determined by contemporary estimates) or by comparing the space on the map to the known distance between two points on the European coast.

The earliest map allowing such an estimate is Andrea Bianco’s world map of 1436. On this map, the macro-peninsula is attached to a large island called “Antillia.” This name signifies “land opposite tile” (the Roman name for Iceland). The distance west to Antillia is about twice the north-south length of the Iberian coast—or about 800 miles. Since Antillia is also placed west of the Azores by about an equal distance between the Azores and Lisbon, we have another possible estimate of the distance at almost 2,000 miles. The disparity in the two estimates can be attributed to the difficulty in determining longitude without a chronometer and the lack of uniformity in portraying land areas on Early Renaissance maps.

By 1492, three cartographers (Toscanelli, Martellus, and Behaim) indicated that the macro-peninsula (called “India Tercer” or “Zaiton”) was situated between 90° and 110° west of Europe. Since they believed that Earth had a diameter of about 18,000 miles (the current estimate by Alfraganus), we can calculate the distance between Europe and the overseas coast on the Tropic of Cancer at about 4,000 miles.[26] This estimate is about 500 miles over the actual distance between Florida and Europe. Columbus came up with a similar estimate of 3,000 miles from the Canary Islands to Japan. Considering that contemporary scholars believed that the coast of China was considerably farther west, modern historians have been perplexed by the incredible accuracy of the Columbus estimate. 


The increasing accuracy of Late Medieval maps with respect to the configuration of the coastline across the Atlantic from Europe seems to have resulted from the cumulative knowledge of numerous explorations. There is a continuous record of Portuguese voyages to Antillia or The Isle of Seven Cities at the same time that cartographers in their service increased the accuracy of the location of the macro-peninsula (Zaiton/Antillia/Florida) across from Europe.

The importance of this macro-peninsula for New World exploration in the 15th century is that Marco Polo identified Zaiton as the location of thriving ports in Cathay (China). According to Polo, Toscanelli, Martellus, and Behaim, the Island of Cipangu (Japan) was situated between 500 and 1,500 miles east of Zaiton. Japan was the initial destination of Columbus in 1492 from which he expected to visit China and the Indies. When he found land about where it was located on his maps (which were based on Portuguese sources), he was convinced that he had reached his destination—Japan. Subsequently, he deduced from a comparison of actual coastlines to his map that Cuba was the macro-peninsula (Zaiton); and he assumed that Hispaniola (Haiti) was Japan.[27] Historian Peter Martyr indicated that this Japan of Columbus was previously known to the Portuguese as “Antillia.”

Columbus seems to have been misled by bogus Portuguese maps. Considering the sophistication of Portuguese geographical knowledge and the king’s desire to protect commercial interests in Africa and India, it is conceivable that they issued maps showing Zaiton 4,000 miles west of Europe in a deliberate effort to mislead Spanish competitors.[28]

The collection of maps in our sample attests to a gradual modification of a place called “Antillia” or “Isle of Seven Cities” during the course of numerous Portuguese voyages into the western Atlantic. We can trace the cartographic evolution of Antillia into Florida by comparing the Bianco map of 1436 with the so-called “Cantino Map” of 1502—Figure 6.  Antillia has the same name, same shape, and about the same location on both maps.[29] Geographical features on the coast of the macro-peninsula of the Cantino Map are identical to those used on Martin Waldseemuller’s map of 1507—at which time the macro-peninsula is joined to mainland that he calls “America”—Figure 7.  Most subsequent maps of the New World naming “Florida” derive from Waldseemuller’s map or from the Cantino Map—which can ultimately be traced back to Bianco’s map of 1436.

Following voyages by Amerigo Vespucci past Cuba in 1497, the Portuguese calculated the distance between Europe and the macro-peninsula at 65°. This is where it appears on Waldseemuller’s map in 1507. The actual distance in degrees of longitude between Europe and Florida is about 71°. This increased accuracy was made possible because Vespucci had invented a new method for calculating longitude.[30]

The presence of mainland on maps having the configuration of Florida before it was “officially” discovered by Ponce DeLeon has not escaped the notice of mainstream historians.  Nebenzahl noted that the Florida-like coast on the Cantino Map (1502) was “the greatest unsolved cartographic puzzle of the period.”[31] A similar coastline is found on the Caveri map of 1504. Strangely, historians credit DeLeon with “discovering” Florida in spite of the fact that it was already on the map.[32] Incredibly, DeLeon’s two expeditions produced no known map of the region nor is there any apparent modification visible on subsequent maps of Florida.

If the 15th century copy of the Macrobius map is an accurate reproduction of the original Roman document, then we seem to have cartographic evidence of Roman voyages to the New World. Aristotle and Pliny The Elder mentioned voyages to isles in the west—indicating the possibility of some familiarity with mainland in that direction.[33] Thus, it should not come as any great surprise that the configuration of mainland west of Europe is similar to that of the Americas on ancient Roman maps. We also find the macro-peninsula on a map by Marino Sanudo in 1320—suggesting that the Roman tradition continued to influence geographical concepts during the High Middle Ages.

While the case for Roman voyages remains speculative, our examination of cartographic evidence does not support the theory of New World isolation prior to Columbus. Data on the configurations of western coastlines with respect to the longitude and latitude of the macro-peninsula do not reflect the kind of random features we would anticipate if they were merely the result of “Medieval fantasies.” On the contrary, they manifest a surprising degree of uniformity and scientific accuracy. Maps of this region were refined over time in the context of exploratory voyages so that prior to Columbus, the Portuguese had an accurate idea of the location of Florida. This knowledge led to an assessment on the part of Columbus that his proposed venture to sail west was feasible. After Columbus used an astrolabe to calculate the latitude of his landfall in 1492, he was convinced that he had found land where he expected it to be.


[1] According to Jacob Cnoyen’s Travels in the Northern Regions, in Taylor, E. (1956), “A Letter Dated 1577 from Mercator to John Dee,” in Imago Mundi, Vol. 13, “the philosophers” were the source for the belief in a magnetic mountain at the North Pole.

[2] Bacon, R. Opus Majus (1266). Translation in Burke, R. (1928) The Opus Majus of Roger Bacon (University Press, Philadelphia), p. 327.

[3] See The Hakluyt Society 1893, The Voyages of Captain Luke Foxe (No. 88), p. 13-15, and Jacob Cnoyen’s Travels in the Northern Regions, in Taylor, E. (1956).

[4] See Thompson, G. (1996) The Friar’s Map (Argonauts, Seattle). 

[5] Alder, M. The World of Aristotle, Encyclopedia Britannica, Chicago, 1952, p. 470.

[6] Most dictionaries and encyclopedias indicate that the word “hurricane” is derived from the Native American word “hurakan” encountered often in Mayan tradition. It is generally regarded as a post-Columbus import to Europe, however it is seen on numerous Medieval maps near the western Caspian Sea and was known to Aristotle.

[7] Ibarra-Grasso, D. (1970) La representacion de America en Mapas romanos de tiempos de Cristo (Eddiciones Ibarra-Grasso, Buenos Aires). Harley, J. and Woodward, D., Eds. (1987) The History of Cartography (Chicago, U. Press).

[8] Boorstin, D. (1983) The Discoverers, (Random House, New York), p. 264, identifies Magellan as calling the Pacific “Ptolemy’s Great Gulf.”

[9] Martinez, G. (1996) America en un mapa de 1489 (Taller Abierto, Mexico).

[10] Holand, H. (1956) Explorations in America before Columbus (New York), pp. 303-5.

[11] Skelton, R., Marston, T., and Painter, G. (1965) The Vinland Map and The Tarter Relation (Yale, New Haven).

[12] Babcock, W. (1922) Legendary Isles of the Atlantic (Geographical Society, New York), p. 47, notes that the current reconstruction of Behaim’s globe may not be as accurate as some of the copies made before the repairs were done.

[13] See Thompson, G. (1996) The Friar’s Map (Argonauts, Seattle), pp. 65-71.

[14] Bacon, R. in Burke, R. (1928), p. 311.

[15] Poole, A. (1958) Medieval England (Clarendon, Oxford), p. 592.

[16] Mercator’s documentation of the Franciscan survey is contained in his map of 1565. It is also found in Taylor, E. (1956) Imago Mundi, Netherlands, 13, pp. 56-68.

[17]Taylor (1956). Mecator attributes the idea of a magnetic mountain to “the Philosophers”—suggesting that Romans used the compass.

[18] Nansen, F. (1911) In Northern Mists (AMS, New York), Vol. I, p. 262. Thompson (1996) pp. 145-149, 282-285.

[19] Thompson (1996) pp. 285-87. In 1989, the magnetic pole was near 75°N-101°W.

[20] Collinder, A. (1954) Marine Navigation, p. 46, mentions that a Flemish gnomon dated 1451 is marked for declination. Waters, D. (1992) Mariner’s Mirror, UK 78 (4), pp. 398-99, indicates knowledge of declination as seen on Portuguese maps by 1375.

[21] It is not uncommon for historians to exclude the northern portion of Martin Behaim’s globe out of a misconception that the area is of no significance to New World discovery.

[22] This map and its authentication have been the subject of several recent articles. Durst, A. (1996) Cartographica Helvetica, Switzerland, 13, pp. 10-21. Thompson, G. (1996) WAML Info. Bul. USA 27 (2), pp. 65-77. Thompson, G. (1999) WAML Information Bulletin USA 30 (4).

[23] Ca-paru is the earliest known portrayal of Peru. The coastline is very similar to the coast of Peru on a modern map. Thompson (1996), pp. 153-174.

[24] Tornoe, J. (1965) Columbus in the Arctic (Broggers, Oslo).Thompson (1996), 46-53.

[25] Thompson (1996), pp. 175-244.

[26] Nebenzahl (1990), p.13, calculates the distance between Europe and Japan on maps by Martellus and Behaim at 3,500 miles or 90°.  Harley, J. (1990) Maps and the Columbus Encounter (Maier, Milwaukee) indicates the distance at 4,250 miles. 

[27] Thompson (1999).

[28] Thompson, G. (1999).

[29] Thompson (1999).

[30] Pohl, F. (1991) Enc. Americana, 28, p. 56, says Vespucci used this method to estimate the circumference of Earth to within 50 miles of the correct distance.

[31] Nebenzahl (1990), p.34.

[32] Nebenzahl (1990), p. 34, suggests that Vespucci might have been responsible for the Florida map—but it predates his voyage of 1497. Other mariners who might have had a role in the cartography include John Cabot and Nicholas of Lynn—an Oxford monk that Richard Hakluyt identified as the author of Inventio Fortunatae in Thompson (1996), p. 75-153).

[33] Galvano, A. (1563) in Babcock (1922) mentions that Roman mariners found mainland across the Atlantic at 50°N.