Two fifteenth-century Maps of "Zipangu": Notes on the Early Cartography of Japan

[206]

The first news of Japan was brought to Europe by Marco Polo at the end of the thirteenth century. While islands off the cast coast of Asia are shown on mappemondes of the fourteenth century,it had long been believed that Martin Behaim of Nürnberg, author of a famous globe in 1492, was the first mapmaker to show Japan in its proper location, and to provide it with a set of legends referring to its position and to its treasures.

But it was Fra Mauro, the last of the great cartographers of the Middle Ages, who first named Japan. On his world map of 1459, now in St. Mark's Library in Venice, a single island off the coast of Asia is shown, named "Ixola de cimpagu."(2) This is the sole legend applied to it.

In 1940, Roberto Almagià called attention to the existence of a detailed map of "Zipangu" in a manuscript "Insularium." (XXIX.25) by Henricus Martellus in the Biblioteca Mediceo-Laurenziana in Florence (3). Moreover, "Zipangu" appears in a form closely resembling that of the Florence atlas map on the large world map (measuring 108 by 190 centimeters), signed by Martellus, which was presented in 1962 to the Yale Library.(4) These two maps, dated approximately 1488-90, represent the first attempt to portray Japan in any detail, in the West. Further, it is likely that a map or maps, similar to these, may have served as a model for Martin Behaim's famed globe of 1492, and may also have been known to Columbus prior to his first voyage.

On the Martellus world map, Japan-Cipangu is shown as a large [207] island, with its longer axis oriented north-south, lying between 261 and 270 degrees east, 8 and 31 degrees north. (See first plate following page 224 - reduced about 2 1/2 times.) It is the largest island off the coast of Asia and is identified by name; the legends it bears are now, unfortunately, almost illegible. It has a distinct shape, however, and in its northern third is bisected by the Tropic of Cancer.

Of Henricus Martellus himself hardly anything is known. Like several of his German contemporaries, printers, etchers, designers, he sought and found fortune in Italy; he worked in Florence, and was associated with the mapmaking workshop of Francesco Rosselli there. His known works include two codices of Ptolemy's geography, and five codices of the "Insularium Illustratum" or "Book of Islands" (including the one already mentioned as being in Florence), as well as the world map in the Yale Library.

"Insularia" were popular early collections of texts illustrated with maps, and Martellus' version is a copy, with additions, of the earliest and best known of these, the "Liber insularum arcipelagi" of Cristoforo Buondelmonti. These collections included large and detailed maps of important islands of the Aegean and eastern Mediterranean; Martellus' was the first to include a world map as well. The five codices of Martellus' "Insularium" are in the British Museum, the University Library of Leiden, the Condé Museum at Chantilly, France, the Biblioteca Mediceo-Laurenziana of Florence (already mentioned), and the James Ford Bell Collection in the University of Minnesota Library. For many years the first three of these were well known, and the world map contained in the British Museum version was given considerable prominence by students of the cartography of the pre-Columbian period. In 1940, in the article already cited, Almagià included a detailed study of the "Insularium", pointing out first of all, that the version at Florence was the oldest of the five, having served as a working copy for the cartographer, and second, that it contained a detailed map of "Cipango". (See second plate following page 224 - reduced about one-half.)

This map, numbered XIII in the Florence codex, shows "Cinpangu Insula", in considerable detail, with rivers, mountains, several cities, and legends describing the island. In the northern portion, a legend states that the ships coming hence from the province of "Quinsay" [208] make the voyage in the winter, and return to their home ports in the summer. This statement corresponds exactly with Marco Polo's description of trade in the "Sea of Chin", the sea off the eastern coast of Asia:

the vessels engaged in the trade from the ports of Zaitun and Kinsai do not reap large profits, being obliged to consume a whole year in their voyage, sailing in the winter and returning in the summer. (The Travels, Everyman's Library ed., New York, 1918, pp. 329-30.)

A clear and unmistakable reference to the monsoons and to the dependence of Chinese mariners on that seasonal wind system, contained in Polo's narrative, is applied here specifically to Japan.

On the east coast of the island, above the legend "Cinpangu ciuitas regalis", a forest of nutmeg trees is shown and identified. Below the large-lettered words "CINPANGV INSVLA", another legend describes the island as being one hundred and fifty thousand paces from the province of Mangi, and very large. Here again it is easy to identify the source of the statement, from chapter II, book III in Polo's narrative: "Zipangu... is situated at the distance of about fifteen hundred miles from the mainland of Mangi, [i.e., the coast of China]" (The Travels, pp. 323-324). Finally a legend in the southern portion of the island refers to a forest of ebony trees.

Martellus' map of Zipangu, ascribed to the end of the 1480s, may thus be described as a map of the island showing far more detail than its only predecessor, Fra Mauro's mappemonde. But is it possible to date it earlier than Behaim's globe of 1492? Can we establish a connection between this map of the "Insularium", the Martellus world map at Yale, and Behaim?

The Yale world map was originally drawn by hand on large sheets of paper, and mounted on cloth. In the course of time the glue, used for mounting, has fused paper and cloth together and legends in many parts of the map are illegible as a result. Infra-red photographs of the map indicated that the island "Zipangu" had several legends on it, but these could not be deciphered on the photograph. On 23 February 1965, the map was examined under several types of light by the Curators of Maps and of Medieval Manuscripts of the Yale Library, together with the author. Under intensive incandescent light, several features of interest were revealed.

[209]

The ribbon-shaped decoration, located on the island below the Tropic of the Cancer, contains the name of the island: the letters "ZIP..." are legible. Below it, in the first line of a long legend, the words "h.c insula dist..." could be read. Both the wording and the script are identical with those of the Florence "Insularium" map. Further, although the other legends on the map defied attempts to read them, it was ascertained that the rivers and mountains shown on the island were drawn in positions corresponding very closely with the Florence map. Three towns could be identified, all on the west coast of the island and again identical with those placed on the Florence map. The shape of the island on the Yale and Florence maps, the wording and writing of at least one legend, and the visible details of topography, drainage, and cities furnish convincing evidence that these maps stem from the same cartographer and that they were both drawn within a short period of time.

In attempting to establish a connection between the Florence "Insularium" map and Behaim's globe, we are able to refer to contemporary evidence concerning the making of the globe, as well as to the image of "Zipangu" on these two works. On 26 August 1494, Jörg Holzschuher, a city official of Nürnberg, drew up a statement of the expenses incurred by the city in commissioning Martin Behaim's globe. It includes the following: "Item paid Mr. Merten Beham for a printed mappa mundi, embracing the whole world, which was used for the globe and is to be hung in the town office, 1 fl. 31 b."(5) In the context of the rest of the account these words can well be interpreted to mean that Behaim used a large world map as a model for his globe. R.A. Skelton, in an unpublished typescript, draws attention to the fact that, according to this account, the city paid for painting and mounting the world map that served as a model. The Martellus Yale Map, although not printed, is drawn and colored by hand and mounted on cloth. Furthermore, Behaim's globe is graduated in longitudes and latitudes and the Martellus Yale map is the only other non-Ptolemaic world map of the fifteenth century to show such graduations. The use of the Martellus world map or a printed version of it would fit in with the use of an Italian source by Behaim, and the [210] fact that close correspondences can be detected between his globe and the "lost map" of Paolo Pozzo Toscanelli.

In the case of Japan, a comparison of the island's outline on the Martellus "Insularium" map at Florence, the Martellus world map at Yale, and the Behaim globe shows striking similarities (See illustration). When we turn to the legends on Behaim's globe, more numerous and detailed than those on the Martellus "Insularium" map, we find that Martellus shows forests of nutmeg and ebony trees, while Behaim refers to forests of nutmeg and of pepper. There is no reference on Behaim's globe to the distance from the China coast to Zipangu, but there is no need for it, since the globe is graduated in degrees, and the value of the degrees in leagues is given. Instead, Behaim states that the island has a circumference of 1,200 miles, and refers to its treasures of spices and precious stones, and to its character as an independent kingdom.(6)

If we accept the priority of Martellus' map over Behaim's globe, there still remains the problem of identifying Martellus' sources for his map of Japan-Zipangu. Marco Polo's narrative, the first report on Japan in the West, was, as we have seen, the source of the legends, but what was the source for the island's outline?

Islands, whether real or imaginary, have fascinated cartographers for many centuries. As G.R. Crone has pointed out, the "terre des puceles" in the French version of the famous so-called letter of Prester John is the forerunner of many an "island of women" in literature, among them the "isle of women" off the coast of Japan.(7) And R. A. Skelton writes that

a strong family resemblance can be recognized in the conventional outlines given to islands reported or rumored but not seen or surveyed. We find it in the Catalan mapmaker's Greenland; in the Antillia of many is 15th century charts; in the mysterious ixola otenticha of Andrea Bianco's Atlantic chart of 1448; and in the Japan of the world maps by Fra Mauro (1459) and Henricus Martellus (about 1490).(8)

Japan on Fra Mauro's world map is one of many islands, distinguished only by its being labeled "cimpagu". Henricus Martellus [212] gives it a far more definite shape, and a north-south orientation, that reappear on his world map, and on Behaim's globe. Is this outline based on an unknown source, or is it a stylized attempt to locate and identify an island well known from the narrative of Marco Polo, the greatest traveler of medieval Europe?

In the Hall of the Coats of Arms ("Sala dello Scudo") of the palace of the Doges in Venice (where the incumbent Doge's arms were displayed during his tenure of office), Gastaldi, in 1540, painted on the walls a series of maps. Two centuries later, in 1761-62, the Venetian cosmographer Grisellini repaired and refurbished these maps. Among them there is a map of the Far East, showing the itinerary of Marco Polo and of other famous Venetian travelers. A story, widely circulated and long believed, maintained that this wall map was based on a map brought back by Marco Polo himself. The principal source of the story was Ramusio, who printed it in his collection of voyages. Almagià, in his introduction to the facsimile edition of Fra Mauro's map, thoroughly discredits the story, pointing out that maps available to travelers in Polo's time were most probably of the mythical, medieval mappamundi variety, and Polo, though unquestionably an excellent observer, was not likely to have possessed cartographic talents as well.(9)

Since there does not seem to be any proof of the Venice wall map's having been copied from an original in Marco Polo's possession, it is far more likely that Ramusio circulated his story to enhance the authenticity of his own version of Polo's travels. As for the outline of the Martellus "Insularium" map of Japan, there is enough of the style of fifteenth-century cartography in the modeling of its coastline to allow us to classify it as another imaginary isle, except for the accuracy of the legends placed upon it.

Important as they are for the history of the mapping of Japan, Martellus' image of Zipangu and his placing it on his Yale world map of 1489 are even more significant because of the role they played in the planning of Christopher Columbus' first voyage of 1492-93. Paolo di Pozzo Toscanelli, in his letter to the confessor of the King of Portugal, placed Zipangu 85 degrees west of the prime meridian, then believed to pass through the Canary Islands. It was Toscanelli's [213] reasoning as well as his own estimate of the distance between Zipangu and the Canaries that helped convince Columbus of the feasibility of his enterprise. The Portuguese historian Joao de Barros, in his Decades of Asia, says that since Columbus

...read a good deal in Marco Polo who spoke of Oriental matters, of the kingdom of Cathay, and of the mighty isle of Cypango, he reached the conception that over this Western Ocean Sea one could sail to this Isle Cypango and other unknown lands.(10)

When Columbus made his landfall on 12 October 1492, his Journal showed the distance covered to have been 1,123 leagues, equivalent to a little over 89 degrees of longitude west of the Canaries.(11) The Martellus map shows Zipangu 90 degrees west of the Canaries.

The Journal of Columbus provides further, and conclusive proof on 13 October 1492, Columbus, after describing the people and products of the island of his first landfall, writes: "But, in order not to lose time, I wish to go and see if I can make the island of Cipangu."(12) On 24 October, the Journal continues:

This night at midnight I weighed anchor... for the island of Cuba... This I am doing, for I believe that, if it be as all the Indians of these islands and those whom I carry with me in the ships give me to understand by signs, for I do not know their language, it is the island of Cipangu, of which marvellous things are recounted, and on the spheres which I have seen and in the drawings of mappemondes, it is in this region.(13)

Zipangu was the most prominent island in the sea off the east coast of Asia, as shown on the Martellus map of 1489. But there were many others, conforming to Marco Polo's statement that "... this sea contains 7440 Islands, mostly inhabited." (The Travels, p. 329.) Thus wrote Columbus, on 14 November 1492, in his Journal:

There he saw so many islands that he could not count them... And he says that he believes that these islands are those without number, which in the mappemondes are placed at the end of the east.(14)

In 1515, the son of Martin Alonso Pinzon, who was one of Columbus' captains on his first voyage, affirmed in court that [214] in or before 1492 his father had seen in Rome a map which showed that there, 95° to the westward (from Spain) by an easy passage, he would find between north and south a land of Sypango.(15)

The discovery of Martellus' "Insularium" map of Japan at Florence, and of his great world map, now at Yale, enables us to say that it is conceivable that Columbus or one of his companions was familiar with one or both of these maps, and that they strengthened his resolve to search for golden Zipangu.

Some twenty years after Martellus' maps showing "Zipangu", the name Japan first appeared in a Western treatise on Southeast Asia, in Tome Pires' "Suma Oriental". (16) But it was not until the middle of the sixteenth century that the Venetian mapmaker Zaltieri showed a single island labeled "Giapan" in its proper position off the Asian coastline, and "Zipangu" disappeared forever from the map.

Notes