Behaim, Martellus and Columbus

(Arthur Davies held the Reardon-Smith Chair of Geography at the University of Exeter from 1948 to 1971.) The Nuremberg Globe of Martin Behaim, made in 1492, was almost identical with the world map of about 1490 by Henricus Martellus, of Florence. Efforts to establish a common prototype in German cartography failed. A large world map, signed by Martellus, came to light about 1960 and was given to Yale University. It proves to be an assembly of tracings on paper sheets of a world map, developed by the Columbus brothers by 1490. It is suggested that an earlier version of this Columbus map was seen by Behaim in 1485 in Lisbon. The Yale map is, in fact, in the hand of Bartholomew Colombo: Martellus merely assembled and mounted the sheets and drew a frame around them.

In 1492, Martin Behaim constructed his famous Nuremberg globe when residing in that city, his home as a boy. It was designed to demonstrate the ease with which one could sail westward to Japan and China, to the Cipango and Cathay of Marco Polo. It showed Cipango as only 80° to the west of the Canaries and Cathay some 35° more. Behaim had hopes of leading a voyage westward to Asia and sought the backing of the Emperor Maximilian. The Emperor shrugged it off by passing it on to King John of Portugal in a letter written by Dr. Monetarius, dated 24 July 1493:

"Maximilian, invincible King of the Romans, who, through his mother, is himself a Portuguese, intended to invite Your Majesty through my simple letter to search for the eastern coast of the very rich Cathay... At your pleasure you can secure for this voyage a companion sent by our King Maximilian, namely Don Martin Behaim, and many other expert mariners, who would start from the Azores islands and boldly cross the sea."

This was virtually identical with the proposals of Columbus to Portugal in 1485. Unknown to Behaim, Columbus had discovered the Bahamas and the Antilles and had returned to Spain by March 1493. It sets a problem: two men with identical plans to reach Asia, at the same time in history and with similar cosmographical concepts. Columbus, in the copy of the Toscanelli letter (made on the flyleaf of the Historia Rerum of Pope Pius II) gave Cathay as 130° west of Lisbon and Cipango as ten spaces, or 50°, west of the legendary island of Antillia. On fifteenth century maps, Antillia was placed at 30° to 35° west of Lisbon so that Columbus made Cipango extend from 80° to 85° west of Lisbon. Behaim went to Lisbon in 1484, and soon was in high favour with King John who placed him on his mathematical junta and knighted him in February, 1485. It was this junta which reported on the proposals of Columbus to sail west to Cipango and Cathay and rejected them during that summer. Behaim had ample opportunity to study the proposals of Columbus and the map which accompanied them. This was the origin of Behaim's later plan to sail to Asia and the source of his concepts of the distances involved. Both were carbon copies of the ideas and map of Columbus. No criticism of Behaim is implied. As far as he knew in 1492, Columbus had not gained support for his enterprise in Portugal or Spain and there was no logical or moral reason why he should not seek German support.

A more baffling problem is the astonishing resemblance between the 1492 Nuremberg globe of Behaim and the 1489 map of Henricus Martellus, now in the British Museum (Plate VII). The representation of East Asia is identical in them. No one at that time had any knowledge of the true position and outlines of East Asia. One was obviously copied from the other for in each the coasts were hypothetical (i.e. invented), unless they had both copied from a common prototype which first revealed such hypothetical coasts. There are, however, important differences between these two cartographic works. Kimble (1938) showed that, as far as 13° south, the nomenclature of West African coasts is 80 per cent identical in the two maps but bears no relation to the nomenclature of any other map of the period. South of that limit, however, the Martellus map gives the outlines and nomenclature of the voyage of Bartholomew Diaz in 1487-88, while the Behaim globe has invented nomenclature corresponding to nothing in Portuguese cartography. The correspondence between the Behaim globe and the 1489 Martellus map consequently ended in 1485, when Diego Cão had returned from his first voyage after reaching 13° south. This was the time when Behaim, as a member of the junta, studied the map and proposals of Columbus.

The 1489 Martellus map extends from the Canaries to the east coast of China. No meridians, parallels or scales of longitude are given, but estimates based on measurements of the map indicate about 230° from Lisbon to the coast of China, or 240° from the Canaries. These agree with the distances on the globe of Behaim. The coast of Cathay is approximately 130° west of Lisbon on the Martellus map, on the globe of Behaim and in the Columbus-Toscanelli concept. The minor differences in location between these three must be seen against the enormous exaggeration of the extent of Eurasia which they exhibit in comparison with all previous estimates. The Catalan Atlas of 1375 made it 116° east from the Canaries to the coast of China; the Genoese map of 1457 made it 136°; and the Fra Mauro map of 1459, about 125° (Fig. 1). It is actually 141° from the Canaries to Shanghai. Who had most to gain from such a reckless exaggeration of the extent of Eurasia and who was the first to do so? Columbus, surely. His entire hopes of gaining support from King John in 1485 for such an enterprise as sailing westward to Cathay rested on his argument that it lay only 130° to 140° to the west. The Behaim globe and the Martellus map seem designed to plead the same cause. But Martellus had no such ambition or motive. It is difficult to avoid the conclusion that he copied a map which was originally designed to support the ideas of Columbus. Yet that map could not have been completed until early in 1489 for it had complete details of the discoveries of Bartholomew Diaz in his voyage of 1487-88, when he doubled the Cape of Good Hope and reached the Indian Ocean. He returned from this voyage in December 1488 and, within a year, full details, including rich nomenclature, had appeared on the map of Martellus made in Italy; this despite the utmost secrecy on the part of King John of Portugal. Even today, the best source for information on the voyage of Diaz is the Martellus map of 1489. The policy of secrecy of King John was shattered in one great leakage by someone in a unique position to know all the details.

Other features of the Martellus map indicate an origin in Lisbon. The representation of East Asia, from the Arctic to 35° south of the equator is new; it extends far to the east of the Ptolemaic limit of 180°; its nomenclature is completely new and based on Marco Polo. A manuscript copy of Polo and his travels was given by the Doge of Venice to Prince Pedro in 1427 and was thereafter in the King's Treasury in Lisbon. Copious marginal notes, in the handwriting of Bartholomew and his older brother Christopher, are found in the Admiral's copy of the printed book of Marco Polo, published at Louvain between 1485 and 1490.

The Arctic coasts on the Martellus map and those of north and north-west Europe resemble those of the Fra Mauro planisphere of 1457-9 rather than those of the Ulm Ptolemy of 1482. From Normandy to Sierra Leone the coasts seem to be based on Portuguese discoveries. In the Martellus map the influence of the Fra Mauro map is pronounced. The coasts charted by Diaz have been fitted into the circular outline of the world of the Fra Mauro hemisphere, presenting such a marked trend to the south-east that the Cape of Good Hope seems to be due south of the Persian Gulf, whereas it is due south of the Adriatic. The Indian ocean is open to the south, as in the Fra Mauro planisphere. Due south of the Malay peninsula, at 28° south, there appears an enormous peninsula which widens and turns north to join China, again with the largely circular concept of the Fra Mauro map. It does not exist in fact, and seems to be a relic of the Donus Nicolaus map of the world in the Ulm Ptolemy, cut away by the circular outline of Fra Mauro. The Fra Mauro map was made at Murano, near Venice. It was commissioned for King Alfonso V of Portugal and was in Lisbon from 1459 on. Someone with access to it and to the reports of Bartholomew Diaz drew the prototype of the Martellus map. Two peculiar features in the region of South Africa suggest that Bartholomew Columbus was the someone!

1. By 1486, the mathematical junta (using Master Joseph Vizinho's simplified tables of solar declination based on those of Abraham Zacuto of Salamanca made in 1473) had solved the problem of establishing latitude by measuring the height of the midday sun (Albuquerke, 1971). The actual latitude of the Cape of Good Hope is 34° 22' S based on measurements on land by Diaz, who landed three times on the south coast. Measurements of altitudinal height of the sun by astrolabe or quadrant were accurate on land but could be 2° or more out on the rolling deck of a ship. Yet the Martellus map shows South Africa extending across the frame of the map to 45° south. The only other claim that the Cape was at 45° south is in the hand of Bartholomew Columbus.

In the volume of Imago Mundi, found amongst the possessions of Christopher Columbus after his death, there are numerous notes or postils, written in the margins or below the printed matter. No. 23 is in the handwriting of Bartholomew, and was identified as his by Bishop Bartoloméo de Las Casas, who knew him well. I have made a long study of the writing of the Columbus brothers and state, without a shadow of doubt, that it is in the hand of Bartholomew. It reads:

"Note that in the year '88 in the month of December arrived in Lisbon Bartholomew Diaz, Captain of three caravels which the Most Serene King of Portugal had sent to try out the land in Guinea. He reported to the same Most Serene King that he had sailed beyond Yan 600 leagues, namely 450 to the south and 250 to the north, up to a promontory which he called Cabo de Boa Esperanca [Cape of Good Hope] which we believe to be in Abyssinia. He says that in this place he found by the astrolabe that he was 45 degrees below the equator and that this place is 3100 leagues distant from Lisbon [c. 19850 km]. He has described this voyage and plotted it league by league on a marine chart in order to place it under the eyes of the Most Serene King himself. I was present in all of this."

Bartholomew was present when Diaz reported to King John. It indicates that he was high in the confidence of the King as an expert cartographer, otherwise he would not have been present at such a secret occasion. It shows that he had the task of adding new discoveries to a Portuguese world map. He alone maintained that Africa reached to 45° south, as on the Martellus map.

2. No one in Lisbon knew of this 45° assertion. It was done probably in Seville after he joined his brother, entering the postil in Imago Mundi and altering his prototype map at the same time. Two unusual features of the Martellus map reveal this late alteration:

(a) Africa has been extended to 45° south only by showing it as breaking through the frame of the map. It seems that the prototype originally showed the Cape at 35° south, well clear of the frame at about 41° south, as one would expect of a competent cartographer.

(b) The second peculiarity is a legend off the east coast of Africa which reads: ultima navigatio Portug. A.D. 1489. This dates the legend as 1489, probably in January of that year, just before Bartholomew went to Seville. This legend has baffled scholars. On the face of it, seeing it on the Martellus map, it asserts that Diaz had proceeded north along the east coast of South Africa to beyond Natal. His furthest point, in fact, was the Rio de Infante (Great Fish River) on the south coast. This is at 34° south. The legend is also at 33° to 34° south. It appears to be north of Natal only because Africa is shown as extending to 45° south. It is conclusive evidence that the prototype originally terminated at 35° south, with the legend correctly placed near it. When Bartholomew altered the prototype map to 45° south, he was unable to remove the legend.

What purpose was served by extending Africa to 45° south? It was not to influence King John, who knew that the Cape was at 34 1/2° south. It was to influence the Catholic Sovereigns who were in the dark owing to the intense secrecy by Portugal regarding discovery. This suggests that the alteration was made in Seville. It suited Columbus admirably and it is likely that Bartholomew made the change at his direction. Columbus hoped Spain would support a voyage westward to Cipango, 85° away, and to Cathay 130° to the west. At that latitude one degree was thought to be 50 miles (80km) - according to the Toscanelli letter - so that Japan was only 4250 miles (7200km) west. To reach India around Africa would involve sailing north-to-south for 39° + 45° = 84°, each of 67 miles (108 km); then north to India, another 45° + 15°; together with 83° of eastering (Lisbon to Mangalore). All told such a voyage to India would total 227° or 15000 miles (24000km). The extra ten degrees in shifting the Cape to 45° south meant more than twenty degrees extra distance in a voyage to India. Moreover, and perhaps this was the decisive factor, it would take Portuguese ships to nearly 5O° south to round Africa, into what Diaz had already found to be the roughest seas encountered anywhere in the world. Bartholomew, in 1512, gave evidence in the Pleitos (the great lawsuit of the Columbus family versus the Crown of Spain) and declared that he had gone about with his brother in Spain helping to gain support for his enterprise. His prototype map would have been a powerful argument.

Another peculiar feature of the Martellus map is the enormous peninsula commencing due south of the Malay peninsula at 28° south, thereafter widening to reach China. It is a relic of the Ptolemaic world maps, and it needs a name to identify it in argument. It bears a rough resemblance to the hind leg and huge paw of a tiger which is facing west. At the risk of sarcastic comment I propose to call it the Tigerleg peninsula! It did not exist in fact but, on the prototype map of Bartholomew and on the Martellus map, it served an important purpose. It seems to render impossible voyages of Arab ships or Chinese junks between Ceylon and China. Although the Columbus brothers knew that Marco Polo had returned from China by this sea-route, they inserted this great obstruction of Tigerleg by 1485. It would show King John that even if the Portuguese reached India they could not reach the Spice Islands (which were on the equator cast of Tigerleg) without having to make long voyages into the southern stormy seas. For King John and for the Catholic Sovereigns it showed that Spain could easily reach Cathay and then the Spice Islands, secure front all interference from Arabs or Portuguese in the Indian Ocean. Columbus may have deceived himself regarding the existence of Tigerleg but, since it suited his plans so admirably, one suspects that he asserted its existence just as he later made the Cape of Good Hope to be at 45° south.

Tigerleg must have formed part of the map and the proposals which Columbus made to King John of Portugal in 1485, for it was embodied in Portuguese cartography thereafter for half a century. It was in the prototype drawn by Bartholomew, copied by Martellus in 1489. The globe of Behaim has it to 30° south. The Laon globe, made in France about 1493 (probably by Bartholomew Columbus when he served Anne, Regent of France) has Tigerleg to 40° south. It is shown to more than 25° south in the Cantino map of 1502, the Canerio map of 1504, the Waldseemuller maps of 1507, 1513 and many others. It was one of the boldest and cleverest concepts of the Admiral.

The "Yale" Martellus

In 1963, Alexander O. Vietor (1963), Map Curator of Yale University, reported a gift by an anonymous donor "in the form of a magnificent painted world map signed by Henricus Martellus, approximately six feet by four feet" (180x120cm) (Plate VIII). Signora Carla Marzoli of Milan, in a private communication, stated that this large map "had left Italy into the possession of a family centuries ago and had been lodged in a Swiss bank for safety, for a long time." In 1959, through trade channels, she learned that this map was for sale. She examined it and saw its connection with the Martellus map in the British Museum and with the conceptions of Columbus. At her request, Roberto Almagia and Raleigh Skelton examined it and pronounced it authentic. She bought it and sold it to the donor, who gave it to Yale. Scholars all over the world owe her and the donor a great debt of gratitude. The fact that the map had been put away in a Swiss bank for perhaps 60 years explains why it went unobserved.

Vietor described it as extraordinary for size and precision of execution. He went on to state:

"It is painted in what seems to be tempera over a base of paper in sheets of different sizes, the whole backed up with a large framed canvas, much in the manner of a painting... It has graduations of latitude and of longitude in the margins, the first instance of longitudes being shown on a map... On this map Cipango is placed 90 degrees from the Canaries."

I am greatly in the debt of Alexander O. Vietor for further information in correspondence and for providing me with infra-red photographs of the map for close study. The cast-west extent from the Canaries to the coast of China is about 235°, which agrees with the Toscanelli-Columbus concept and with the Martellus map now in the British Museum. The two maps are nearly identical but there are two major differences. The "Yale" map has Cipango in the ocean at 30° east of Cathay. It extends from 7° north to 31° north of the equator. Its shape, outline and position are identical with the representation of Cipango on the Behaim globe. The Martellus map in the British Museum is less than 20 inches (50cm) from west to east, and is on a scale one-quarter that of the Yale map. Its size was governed by the two folios of the codex (or collection) on which it is drawn. To have included Cipango would have required a scale reduced to one-sixth, too small to permit of legible nomenclature. The second major difference is that, while the coasts from Normandy to Sierra Leone, on the Yale map, are based on the world map of Donus Nicolaus of 1482, the corresponding coasts on the British Museum Martellus uses a Portuguese map for them. Nevertheless, the Cape of Good Hope is at 45° south on the Yale map, Tigerleg extends to 33° south and Africa breaks through the frame of the map in both. One map has been copied from the other, at one-quarter the scale. The Yale map was the original, or prototype, and was the joint product, as will be seen, of the Columbus brothers. As such, it is of enormous value, academic and financial.

The sheets of paper on which the Yale map is drawn are of different sizes, which excluded the possibility that they were printed map sheets, for they would then have had to be the same size to fit within the map portfolio. In a private communication of June 1972, Vietor stated that X-ray examination had revealed no evidence of printing on the paper sheets and that everything was hand-drawn, lettered and coloured. He added: "unfortunately the physical state of the map makes reading the legends almost impossible." The few names which are legible, mainly in the Indian Ocean, are in formal capitals, beautifully executed, as are two lengthy legends in the lower corners of the map. Fortunately, Martellus copied most of the nomenclature and legends on to the smaller map, now in the British Museum, so the "Yale" Martellus can recover them without difficulty.

Little is known of Martellus. In 1489 he produced a codex with maps and text, which he called Insularium Illustration. It is a series of handsomely drawn and painted islands, with accompanying descriptive text in Latin, all done on parchment folios bound in leather-covered boards. It is listed in the British Museum as Additional Manuscript 15760. The authorities kindly allowed me to study it in 1972 and my particular thanks are owed to Miss Janet Backhouse, who devoted much time and care to establishing that it was purchased in 1846 from a London dealer named Thorpe, who paid 70 guineas for it at one of the sales of the collection of the Duke of Sussex. That is all that was known for certain of its history. It survives in almost perfect condition for it has been so little used in 500 years that its colours, features and lettering, preserved from light, dust and the chemical impurities in air, seem as fresh as when Martellus made it in 1489. Folio I commences with an inscription in Latin giving the title Illustrated Islands by Henrici Martelli Germani and stating that it will present "all the islands of our Mediterranean Seas". Martellus was commissioned to do this codex because he was versed in the new and beautiful humanist script of Italy, used throughout in the text, and for his talents as a painter and decorator of maps. Mountains are in brown and gold, rivers in light blue, while forests have tree symbols in green. In every case, the island is set in a dark blue sea which extends to the rectangular margin of the picture. Each island-picture is surrounded by a frame, drawn and painted in ivory to give the impression of a carved wood frame, suggesting a picture hanging on a wall. Such picture-islands occupy nearly all the folios, making an Atlas of Islands, and are then followed by a superb picture-map of Italy, taken from the 1482 world map of Donus Nicolaus. The islands in this codex were copies from the Ulm Ptolemy, from the Isolario of Bartolomeo de li Sonetti, printed in Venice in 1485, and from other sources.

Then comes the most original and remarkable world map of the century. It was not foretold on Folio I and is clearly an unexpected addition to the codex of picture-islands. It uses the homeopathic (heart-shaped) projection, as far as can be judged, for it lacks meridians and parallels and scales. Next come three regional maps, striking because of the enormous nomenclature on the coasts. The first is of western Europe and Morocco, cut off in the east through the centre of France; the second starts from this line and extends cast as far as Naples and Tunis. These two maps, it is evident, were copied from one original. The third map is of the Black Sea region. Martellus had been required by his patron to include in the codex a world map and regional maps which had just come into his possession in 1489. His great talent for such an exacting task was his skill as a draughtsman-cartographer experienced in altering the scale of maps of islands that he was copying. He did this to fit the folios of the codex.

Almagia (1940) stated that he had identified no less than three manuscript maps of the world signed by Martellus, all virtually identical with that in the codex in the British Museum. He dated them to 1490. They all omitted Cipango but he found a separate map of Cipango in a codex which had the same outlines as those in the globe of Behaim. The three manuscript maps were larger in scale and showed more detail than the codex world map. The sequence now becomes clear. The sheets of the "Yale" Martellus, together with certain regional maps, were acquired by the patron. Martellus assembled the paper sheets, stuck them on a canvas backing, made his characteristic rectangular picture frame for it and coloured the seas in dark blue. Then he signed it. He then made a copy on a quarter of the scale, covering two parchment folios of the codex, to be included in it. He copied the regional maps and included them. Lastly he made three manuscript maps from the codex world map, larger in scale.

When Columbus left Lisbon in 1485 for Spain, Bartholomew, with his highly trained skills as a cartographer in the Genoese style, stayed on in the map workshop of King John II. He was engaged in building up a large map of the world based on Donus Nicolaus and on Portuguese charts. It was, like all important maps at that time, drawn on sheets of parchment which could be joined together almost invisibly, and mounted on linen. This large map, 180 cm by 120 cm, formed a standard Portuguese world map, continually added to by new discoveries, including those of Cão and Diaz. By the beginning of 1489, Columbus faced poverty and failure in Spain; his pension had been ended in 1488 and he no longer had free board and lodging from Medina-Celi or the Marquess de Moya. Bartholomew prepared to join him in Spain and help his project. They needed money and, in particular, the vital and continued support of the Bank of St. George in Genoa. They got both. Money could be obtained from the sale of maps kept secret in Portugal. Before leaving Lisbon, Bartholomew copied maps of convenient size. The large standard world map he had to copy in some secrecy and, because of its size, he needed 11 sheets of paper, cheaper, and thinner and quieter than parchment. These sheets of the "Yale" Martellus were tracings in the hand of Bartholomew. Early in 1489 he left Lisbon. He went first to Seville to help his brother and there altered the Yale map by substituting another sheet of paper which showed Africa to 45° south. Then he visited Italy to sell maps and to gain the support of the Bank of St. George. Antonio Gallo, Chancellor of the Bank of St. George, was also official chronicler of Genoa. After Columbus returned from discovering Asia, as was claimed by him at the time, his name rang through Europe. Yet Gallo took the occasion to record his knowledge (in his chronicles) of the Columbus brothers. What is astonishing is that he gave all the credit for conceiving the enterprise to Bartholomew, who first thought it out and entrusted it to his older brother, who was more used to the sea. This account of Gallo was copied, almost verbatim, by Serenega in 1499 and by Giustiniani in 1516. Bartholomew left Genoa as a youth in 1479 and officially made only one further visit to Italy, in 1506. Gallo could have acquired knowledge of the role of Bartholomew only from his own lips, in 1489. Thereafter the Bank supported Columbus from time to time with credits.

Again, in 1489, maps appeared in a Venetian Atlas by Soligo, drawn in Venice, now listed in the British Museum as Egerton MS 73. They show in complete detail the coasts explored by Portugal down to 1484.

In 1951, Marcel Destombes (1952) described two maps which he had studied in the Biblioteca d'Estense, now in Modena. One was a chart of the Atlantic and bordering lands, listed as GGA 5a. This map originally extended further north, west and south but had been cut so that it now extends from Normandy to Sierra Leone and eastwards to Naples and Tunis. Destombes concluded from rhumb lines that the map was designed to extend west as far as the legendary islands of Antillia and Satanaxia. He assigned this map without hesitation to Bartholomew Columbus, on the basis of its excellent lettering and its Genoese style of cartography. The same collection bad a second map, A5a, which is a magnificent Portuguese map in colour, dated as c. 1471. It is the earliest known example of Portuguese cartography and suggests that Bartholomew left Lisbon by sea with maps drawn by himself and by others. It may throw light on the strange absence of Portuguese maps of 1460 onward. It is known that in 1490 Columbus sold maps to earn a living in Seville.

GGA 5a provided the coasts from Normandy to Sierra Leone in the codex world map, replacing those of Donus Nicolaus, perhaps on the advice of Bartholomew. It also provided the first and second regional maps which extended east to Naples and Tunis. One need look no further for the patron of Martellus. It was Hercule d'Este, Duke of Ferrara, who had made his court into a centre of learning, of arts and cosmography. His own leaning was to cosmography.

Behaim copied from the map submitted to King John in 1485 by the Columbus brothers. On his globe, South Africa, between 23° and 28° south, extends a horn of land for hundreds of miles due eastward into the Indian Ocean. It is a relic of the Fra Mauro map, which had Africa extending eastward to this extent, but in a great curve from the equator. It indicates that, in 1485, Columbus used Fra Mauro for Africa south of the equator in the map he submitted to the Portuguese monarch.

The Martellus maps of 1489 show that Columbus and his brother Bartholomew relied on authority in their 1485 map, on Donus Nicolaus and Fra Mauro. But they altered authority in three ways even at that stage. First, by extending Asia eastwards to 240° from the Canaries. Second, by inventing the great obstruction of Tigerleg. Third, by placing Cipango as only 90° west of Lisbon. In 1489, they extended Africa to 45° south. Such alterations were all to the advantage of Columbus, unscrupulous no doubt, but they earn our admiration for the sheer audacity of the great discoverer.

References