What mainland and what islands would Columbus have been most likely to have believed were shown awaiting discovery by anyone daring enough to cross the Ocean Sea? Everything suggests that, far from a ‘New World’, what the Admiral actually expected to find at the end of his first crossing of the Atlantic was the remote and fabulous eastern extremity of the Old World – quite specifically Japan and China as they had been described in Marco Polo’s Travels and other sources.
This was not a zany, way-out idea on Columbus’s part but was the consensus view of geographers, mariners and merchants of his day. All accepted that the earth was a sphere and that it should be possible to sail round it in both directions. None knew of the existence of the Americas. All accepted, at least theoretically, that this meant Japan and China in the extreme east might be fetched more quickly, safely and easily by sailing west across the Atlantic from Europe, than by means of the arduous overland route that Marco Polo had taken to the Court of the Great Khan in the thirteenth century …
Such ideas were in wide circulation and had been expressed in clear visual form on maps and globes prepared before Columbus ever crossed the Atlantic. The classic example is the Behaim globe, completed at the beginning of 1492 -which Columbus is known to have seen in the months before his first voyage.19 Redrawn here in plan form (overleaf) this globe by the geographer Martin Behaim (Martin of Bohemia) shows the British Isles, Spain, North Africa and the Canary islands separated from Cipango (i.e., Japan), China, ‘Greater India’ and the Indonesian archipelago by an Ocean Sea about one-third wider than the Atlantic.20 In between there is no sign of the New World – of course, because Columbus would not discover it until later in 1492 – but Behaim has installed for good measure some ‘mythical’ islands, including the island of Saint Brendan and also Antilia. It is noteworthy that he represents the island which he labels Antilia as rather small and insignificant – nothing like the large and roughly rectangular landmass shown under the name of Antilia on the 1424 chart. But, weirdly, a landmass with something of the traditional rectangular, north-south shape of Antilia does appear much further west on the Behaim globe – lying in the Ocean Sea off the Chinese mainland. Behaim has labelled it Cipango (Japan) and surrounded it with numerous smaller islands.
Coast outlines from Martin Behaim’s 1492 globe. Based on Fiske (1902).
Other maps of the period that depict Cipango in the same Antilia-like manner include the Yale-Martellus world map of 1489 and the Contarini-Rosselli world map of 1506.21
What all have in common is an Ocean Sea far wider that any sailor of the fifteenth century would have dared to cross, Columbus included. All the more reason to suppose that the chart upon which he relied to make the crossing did indeed show the width of the Atlantic accurately – still a formidable enough distance to travel, but possible … possible.
Speculation 3: to Asia with a map of the Americas?
I want to reinforce the point here that Columbus, in possession of our hypothetically accurate but outdated ‘Tyrian sea-fish’ chart of certain parts of the eastern seaboard and islands of the Americas, could have used it successfully to guide his little fleet of caravels to the New World while yet remaining absolutely convinced that the coast and islands he had reached were parts of eastern Asia. Conditioned by Polo and Ptolemy, Columbus’s conception of the eastern extremity of Asia is likely to have been close or identical to that shown on the Behaim globe. Sailing west, he would, in other words, have been expecting first to find Antilia, then the island of Cipango (set amongst numerous other islands as Polo had indicated). And after Cipango he would have expected to arrive at the great curving peninsula of the Chinese province of Mangi, described by Marco Polo, with its fabulous capital of Zaitun.
Comparison of east coast of America with east coast of Asia.
Meanwhile – please remember this is speculation – what Columbus and Pinzon actually had to guide them was an antediluvian chart not of the coast and islands of Japan and China but of the Americas and the Caribbean between Florida and Venezuela. The chart showed pre-deluge Bimini (connected to Andros and the exposed Great Bahama Bank) as a large island with a shape and orientation roughly similar to that of Antilia on the 1424 Venetian portolan, but showing an even closer resemblance to Cipango on the Behaim globe.
We know from the inundation maps that antediluvian Bimini was also surrounded by other islands – as Columbus expected Cipango to be. Imposing his preconceptions on the chart, it is therefore quite possible that he mistook for Cipango what was in fact the cartographic ghost of antediluvian Bimini 6000 or more years ago, and that he mistook the Central American mainland that lay beyond for the Mangi peninsula.
It is usually argued that Columbus’s dreamlike and almost hallucinatory misunderstandings of the geography of the region he discovered arose out of his deep belief that he was sailing to Asia on the one hand and his actual experiences in the New World on the other. But I suggest – again, speculation only – that the real source of the dissonance between expectation and experience was that Columbus’s chart showed antediluvian features that had been long submerged by 1492 and that therefore could not be found no matter how frantically he searched for them. Despite these ‘maladjustments’, however, the islands and mainland of the part of the New World he had arrived in matched up well enough with his expectations of the islands and mainlands of Asia (see page 539) for him to convince himself that he was indeed in Asia.
Entries from the Journal of the first voyage make this extremely clear. First landfall was made at San Salvador on 12 October 1492,22 a point very close to the group of large antediluvian islands that existed around Bimini down to 6000 years ago. If the chart that Columbus used to get his fleet across the Atlantic had shown these ghost islands – the largest of which he believed to be Cipango – then he would have been disappointed and disoriented when he failed to find any large islands at all in the area. He might well have concluded that the chart on which he had placed so much reliance was after all inaccurate, or he might have concluded that he had failed to follow his course properly.
The Journal suggests that Columbus believed his fleet could have been carried too far to the north-east by currents on the transatlantic crossing.23 It is therefore of interest that on leaving San Salvador he chose to sail a compensatory route south and west, through the characteristic tiny cays and sandbars that dot the seascape today, trying to pick up intelligence en route about the whereabouts of the large island of Cipango:
Sunday 21 October 1492. I shall presently set sail for another very large island which I believe to be Cipango according to the indications I receive from the Indians on board. They call the island Colba [Cuba]. [From there] I am determined to proceed on to the mainland, and visit the city of Guisay [Qinsai] and deliver the letters of Your Highnesses [Ferdinand and Isabella of Spain] to the Great Khan, demand an answer and return with it.24
Tuesday 23 October. It is now my intention to depart for the island of Cuba, which I believe to be Cipango from the indications these people give of its size and wealth, and will not delay any further here …25
Wednesday 24 October. This must be the island of Cipango, of which we have heard so many wonderful things. According to the globes and maps of the world I have seen, it must be somewhere in this neighbourhood.26
Columbus did not complete the exploration of Cuba on his first voyage, and on his second voyage he changed his mind about its identification with Cipango and decided that it was part of the mainland of south-eastern China instead. This was because islanders told him that ‘Cuba had no end to the westward’, and referred him for further particulars to ‘the people of Mangon, a province towards the west’.27 As Charles Duff explains:
The name Mangon inflamed the imagination of Columbus, who immediately identified it with the Mangi of Marco Polo, the southern province of China, ‘the most magnificent and the richest province that was known in the eastern world,’ according to Polo.28
[Columbus] was now – as it happened ?
?? within two or three days sail of the western end of Cuba, the discovery of which would have disillusioned him concerning its connexion with the mainland of Asia. As it was, he turned back firmly convinced that Cuba was the eastern extremity of the Asiatic continent. And in that belief every person on board expressed his concurrence by a solemn signed deposition. Columbus never afterward abandoned his conviction – he remained unshaken to the end of his life. The dream or fantasy was to him a reality …29
Despite the constant stream of new discoveries and rapidly improving maps that followed the voyages of Columbus, the dream remained a reality for many others as well. Thus, an inscription placed next to the coast of Asia on the Contarini-Rosselli world map of 1506 informs us that ‘Columbus sailed westward to the province of Ciamba, the region of China opposite Cipango.’30
Last but not least, improbable though it may seem, it is known that Columbus finally decided that the island of Hispaniola was the Cipango of his dreams.31
As noted earlier, Gregory Mcintosh has presented a compelling case that a copy of an original map drawn by Christopher Columbus in which Cuba is represented as part of the Central American mainland is incorporated into the world-famous Piri Reis map of 1513. It is therefore intriguing to note – on exactly that section of Piri’s map that was derived from Columbus – that a large ‘ghost’ island with approximately the same shape, dimensions and north-south orientation as antediluvian Bimini is prominently depicted. What seems to seal the identification with antediluvian Bimini – as the reader may confirm at a glance from the zoomed window overleaf – is that this ghost island is clearly marked with a row of huge stone slabs laid out in a manner that strongly resembles the layout and appearance of the slabs in the now-submerged Bimini Road. Mcintosh does not comment on this peculiar megalithic image on the 1513 map; however, he does believe that he can explain the presence there of the non-existent island itself without any recourse to ghosts from before the flood.
It’s all terribly simple, he argues. This large north-south oriented island cannot be found today because it is just the result of a dishonest – or at any rate self-deluding – representation by Columbus of the island of Hispaniola to make it look more like the island of Cipango that he had convinced himself it was.32
Now a glance at any modern atlas will show that Hispaniola (today divided between Haiti and the Dominican Republic) has an east-west rather than north-south orientation and that an island of roughly the right size, in roughly the right place and with roughly the right east-west orientation to be Hispaniola does actually appear on the Piri Reis map. However, Mcintosh ignores that option and reminds us (correctly) that Columbus shared the general conception shown on the Behaim globe, etc., of Cipango as an island with a north – south orientation.33 So wedded was the great explorer to that idea, alleges Mcintosh, that in maps made on his second voyage (one of which Piri Reis copied) he simply flipped Hispaniola 90 degrees so that it now lay north-south -with the end-result that: ‘The shape and orientation of Hispaniola on the Piri Reis map is strikingly similar to that of Cipango shown on maps of the fifteenth and sixteenth centuries.’34
Piri Reis’s ghost island in the Caribbean.
‘It is difficult to accept the fact,’ adds Mcintosh:
Exposed Great Bahama Bank, 6900 years ago.
that Columbus, perhaps the greatest navigator of his time, would contort the orientations of Cuba and Hispaniola as much as 90 degrees out of place … [Yet] for Columbus in 1495–96, when the map was made that Piri Reis was later to use, to have turned Hispaniola 90 degrees to its correct position was to admit it was not Cipango and that his ‘Enterprise of the Indies’ was a failure.35
This is actually a moot point and Mcintosh passes over it too quickly. Columbus may well have shared the general preconception, apparently based on nothing more than vague reports of Marco Polo, that Cipango was a north-south-oriented island, but he was a practical man as well as a dreamer, and a good navigator by all accounts. I do not think he would have persisted in the notion that Cipango was oriented north-south if the island that he believed to be Cipango turned out, in practice, to be oriented east-west. Either he would have decided that he had not yet found Cipango, or that the old travellers’ tales about Cipango being oriented north-south had been wrong and that his own scientifically measured east-west alignment for the island should be substituted on future maps.
Hispaniola was discovered by Columbus on his first voyage, when he named it not Cipango but ‘La Isla Espanola’ – ‘The Spanish Island’. Mcintosh’s identification of Hispaniola with the strange north-south island on Piri’s map is therefore strengthened by the fact that Piri labels it, literally, ‘The Island Named The Spanish Island’. Additional support comes from a place-name on the island -Paksin vidad. ‘This name,’ says Mcintosh, ‘is undoubtedly Navidad, the name of the first settlement founded in the New World on the north coast of Hispaniola.’36
The peculiarly ambiguous identification of Hispaniola with Cipango that Mcintosh believes Columbus was anxious to make filtered through to others and survived long after it was known that Hispaniola was definitely not Cipango. Thus, in a legend on the Ruysch map of 1507 we read that ‘what the Spanish have named Hispaniola is also Cipango’.37 Likewise, Ononteus Finnaeus, in his world map of 1534, labelled Hispaniola as Cipango.38
The picture is then further complicated by other sources in which we find Hispaniola being identified not with Japan but with Antilia – for example in letters from the explorer Amerigo Vespucci published in 1506.39 And this, in turn, may well relate to what was apparently a widely held opinion after Columbus’s first voyage, particularly amongst the Portuguese, that ‘the islands he had discovered were the islands of the legendary Antilia and not the coast of Asia’.40 Indeed, this is the reason why the Caribbean islands on modern maps are still called the Antilles today.41
Amidst such cartographic confusion over place-names and attributions I think there is room to respect the quality of the theory that Mcintosh has put forward while remembering that it is only a theory and that there are other possible explanations of the island thought to represent Hispaniola/Cipango on the Piri Reis map. It is possible, for example, that the place-names on the island which so strengthen its identification with Hispaniola (‘The Island Called The Spanish Island’ and ‘Paksin vidad’) were not present on the Columbian original but were put there speculatively by Piri Reis himself.
In view of correlations with Ice Age topography identified on other maps of the period, and of the special importance given to the anachronistic map showing the end of the Western Sea that the Admiral was said to have possessed before discovering the Americas, I remain open to the possibility that all along what Columbus thought of as representations of Cipango and its surrounding islands on his mysterious chart could have been ghosts of the antediluvian islands of the Great Bahama Bank.
Taken to the limit, this line of reasoning might even suggest that the model for early cartographic representations of Cipango (conceived of as an island that could be reached by sailing west from Europe) was not provided by vague travellers’ reports sent back across the breadth of Asia as has hitherto been supposed, but was in fact derived from the representation on ‘Tyrian sea-fish’ maps (one of which had fallen into the hands of Columbus) of the large antediluvian island of Bimini.
But if the ghost of antediluvian Bimini did provide the model for early representations of Cipango, then, logically, it could not also have doubled up as the model for the legendary island of Antilia (which often appears on the same maps as Cipango).
Is there a model for Antilia?
That thing between Columbus and Pinzon again …
We’ve seen evidence, both from the inscriptions of Piri Reis and from the Journal of the first voyage, that Columbus possessed a chart of the Atlantic -and that it was considered so important as a guide to the crossing that it was passed back and forth between Columbus in his flagship the Santa Maria and his second-in-command Martin Alonso
Pinzon, captain of the Pinta. The presence and exact character of this chart seem enigmatic when we remember the intensive demand for it from the two captains (who shared it but apparently did not possess individual copies), its obvious practical utility to them throughout the voyage, and the fact that it got them to the New World. Such a useful chart of the Atlantic cannot be explained against the background of the cartographic knowledge of the time. On the contrary the part that it played in the success of the Columbian voyages must be weighed up against the background of the abysmal ignorance of even the greatest mapmakers in Europe of the true circumstances of an Atlantic crossing and of the real appearance of the coast and islands on the western side. To have followed the speculative vision of Behaim in his famous globe, or of others like him,42 would have been disastrous, even though their work represents the cream of fifteenth-century mapmaking and was known to Columbus. Indeed, as one commentator has observed, if his chart had been based on the Behaim scenario, ‘Columbus could not even have known of the whereabouts of the New World, much less discover it.’43