Underworld: The Mysterious Origins of Civilization

by Graham Hancock

  The pilot is said to have used unfamiliar nautical instruments, presumably in conjuction with the map.

  We’ve already seen that neither on da Gama’s 1498/9 voyage, nor on Cabral’s of 1500/01 – and indeed not until after 1510 – did the Portuguese have the opportunity to chart the north-west coast of India between Goa and the Indus delta. The evidence of this is in the record of the voyages and also, obliquely, in the Cantino map of 1502, which draws on the latest knowledge that the Portuguese had acquired along the way. Ironically, the very absence of an accurate portrayal of the Kathiawar peninsula in the Cantino map, an absence that still persisted in 1510 when the Reinal map of the Indian Ocean was drawn, provides further convincing evidence that the Portuguese did not chart north-west India until after 1510 – because if they had they would have done a much better job of it (at least as good a job as they did on the coasts of Brazil also discovered on the 1500/01 voyage). They would certainly not have overlooked such a prominent feature as the Kathiawar Peninsula of Gujerat with its two great gulfs of Kutch and Cambay (the latter offering particularly rich trade potential). If we accept in addition that a Gujerati pilot of some repute seems to have been known to the Portuguese, it becomes all the more incredible to imagine that the most precise navigators and mapmakers of the fourteenth and fifteenth centuries could have charted the coast of their pilot’s home region and failed to make an accurate representation of it.

  In short, everything suggests that the Portuguese were not there, and did not chart those coasts until after 1510, and that the representation of north-west India which appears in the Cantino and Reinal maps must therefore have been borrowed by them from a pre-existing local map.

  What better candidate for such a map than the very one that Guzarate showed da Gama and that da Gama so admired on his first crossing to Calicut in 1498?

  Quick detour to Oceania

  One of the several intriguing possibilities suggested by the Guzerate story is that a tradition of accurate map-making with its roots lost in prehistory -perhaps the same tradition that also nourished Marinus of Tyre in the Mediterranean and that eventually expressed itself in the medieval portolans – survived amongst both Arab and Indian navigators in the Indian Ocean right up until the time of the European voyages of discovery.

  The quality of the maps derived from the Indian Ocean tradition was recognized in the fifteenth and sixteenth centuries by the great Portuguese mariners like da Gama (and others as we shall see in later chapters). But there is evidence that these maps and the navigational system that lay behind them had also influenced other cultures in much earlier epochs. I note in passing that in his detailed study of the astonishing achievements of Micronesian and Polynesian navigators in their discovery of the Pacific between approximately 2000 BC and 1000 AD Dr David Lewis draws attention to ‘some remarkable similarities between what has been recorded of ancient Indian Ocean systems of non-instrumental navigation, unquestionably the older, and their Pacific counterparts’.100

  Lewis points out that ‘the magnetic compass … was preceded in the Indian Ocean by a star compass … a compass-card marked in star points’.101 Strangely, the archaic Indian Ocean star compass proves to be very similar to star-compasses of the far Pacific:

  No fewer than eighteen of the thirty-two star points appear to be identical in the Indian Ocean and the Pacific systems …102 [There is] every reason to believe that what we term ‘Polynesian-Micronesian’ navigation was merely part of a system once practised through all the Asian seas, and which very probably did not even originate in Oceania at all.103

  Ice Age India?

  We will encounter other traces of the same lost system when we reach China and Japan in later chapters. But our concern for the moment remains with its impact on European maps of India produced in the early days of the age of discovery. We’ve seen that the Cantino and the Reinal maps (1502 and 1510 respectively) were drawn before the Portuguese exploration of India’s coastlines was complete and that a likely explanation for this that is that both were copied from a pre-existing local source map – perhaps the very map that Guzarate showed da Gama.

  Having a shared common source, or deriving from different but closely similar sources, provides a simple explanation for why the Cantino and Reinal maps are so much alike in almost all respects and also, crucially, why both contain similar mistakes. As I was already aware from Sharif Sakr’s first report (see chapter 14) these mistakes include the absence of the Kathiawar peninsula with its characteristic Gulfs of Kutch and Cambay; a distinct bulge in the north-west corner of India; enlargement of many small island groups, and a south-westerly orientation (with what Sharif describes as ‘distinct lips’) of the southern tip of India. In his e-mail of 23 February 2001 he then makes the crucial observation that:

  While these deviations are all errors relative to a modern map of India, they in fact match up extremely well with Glenn Milne’s map of India 21,300 years ago at LGM. This inundation map shows a large indent at the mouth of the Indus, a bulge obscuring completely the Kathiawar peninsula, enlarged Lakshadweep and Maldives islands, and, most surprisingly, a SW-pointing ‘mouth’ shape at India’s southern tip that is virtually identical to that shown by Reinal.104

  It seems to me that these correlations, and the others that Sharif reported on 10 August 2001, are obvious, striking and speak for themselves. The only questions that need to be asked about them are: (1) do they result from the workings of coincidence? Or (2) are they there because the source maps for Cantino and Reinal were originally drawn at the end of the Ice Age – perhaps not as far back as the LGM but certainly before the final inundation of the Gulfs of Kutch and Cambay which created the Kathiawar peninsula around 7700 years ago?105

  We already know, and nobody would dispute, that the maps of Claudius Ptolemy have now survived in human culture for almost 2000 years and that they incorporate far older streams of ideas, some certainly going back as far as the sixth century BC and some probably much further.

  In the light of Masudi’s testimony confirming the late survival amongst the Arabs of the maps of Marinus, it by no means seems far-fetched to suggest, with Nordenskiold, that the Marinus ‘branch’ of cartography was never ‘lost’ at all but simply transformed itself into the portolan tradition. Otherwise we have the paradox of ‘the most perfect cartographic work of the Middle Ages’106 appearing suddenly, from nowhere, with no prior evolution. And since we already accept that Ptolemy incorporated ideas much earlier than his own in the making of his maps, why shouldn’t we accept that Marinus did so too?

  The Reinal and Cantino maps are portolans that extend far beyond the normal portolano area of the Mediterranean and the Black Sea. And while they do incorporate a few Ptolemaic ideas about the shape of the world, both maps are more distinguished by their stark differences from – and superiority to – Ptolemy. How much of this is due to Marinus? And how old might the oldest information be that could have been included in the Marinus maps? Could some of it have been as old as the last Ice Age when India did actually look the way it is portrayed by Cantino and Reinal?

  If there is any possibility that the latter scenario is correct, then it would become interesting to work out what precise period during the 10,000-year post-glacial meltdown between 17,000 and 7000 years ago is portrayed by the Indian coastlines on the Cantino and Reinal maps.

  Final report on Reinal

  Sharif Sakr to Graham Hancock

  15 August 2001

  It seems that every time I go back to comparing the Reinal map of 1510 and the Milne map for 11,500 BC, I find that the correlation is even better than I previously thought. My latest revision highlights the great affinity between the latitudinal positions of the ‘erroneous’ features on Reinal’s non-Ptolemaic Indian coastline and the correlating features on Milne’s inundation map.

  Milne’s map, in harmony with bathymetric maps of India’s outer shelf, clearly shows a large gulf at the latitude of today’s Indus river delta. I call this feature the �
�Indus Gulf, simply because before the postglacial period the Indus river may have emptied here. In my first e-mail I correlated the ‘Indus Gulf with the only gulf shown on Hapgood’s tracing of Reinal’s map in roughly the right place. This correlation is not perfect: the portolan gulf is the wrong shape and it lies too far north (because Reinal’s Tropic of Cancer is too far north, continuing a Ptolemaic error). Moreover, this northern gulf on the Reinal might be better matched with Sonmiani Bay (and the mouth of the Porali river), which lies to the north of the Indus and which was well known to Arab geographers of the time because of the important seaport of Daibul. This northern part of the map is so inaccurate that it is difficult to be sure of anything.

  But the Bodleian photograph reveals another large gulf on the Reinal, not shown properly in Hapgood’s tracing, which exactly matches the Indus Gulf on Milne’s map in terms of shape, size and latitude. This gulf lies south of Reinal’s erroneous Tropic of Cancer, and at exactly the right latitude relative to, for example, the eastern tip of Oman on the opposite side of the Indian Ocean. It lies well outside the area covered by the old Ptolemaic model and is therefore very likely to have been present on the mysterious non-Ptolemaic source that Reinal used.

  When we correlate the gulf shown in Milne’s map with this gulf on the Reinal map, the latitudinal positions of Reinal’s other ‘errors’, relative to each other and to this northern landmark, make far more sense. Overleaf is my final matching of ‘errors’ on the Reinal map to features on the Milne map – just follow the numbers:

  India’s coastlines in Reinal map of AD 1510.

  Today this is the mouth of the Indus river, which is a delta. But on both Reinal’s and Milne’s maps, it is marked by a wide gulf.

  A large bulge that in both Reinal’s and Milne’s maps replaces the Kathiawar peninsula that exists today.

  An island (or island-group) which is depicted on both maps but which does not exist today.

  A gulf which on both maps is much smaller than the Gulf of Cambay that exists today.

  A large island (or island-group) which is depicted on both maps but which does not exist today.

  An island at the same latitude as the northernmost Lakshadweep island (approximately 12 degrees north) is shown on both Reinal’s and Milne’s maps. No island exists there today.

  India’s coastlines in 11,500 BC.

  The Lakshadweep islands, which exist today but which are enlarged in both Reinal’s and Milne’s maps.

  The tip of the sub-continent. Both maps show the tip of the sub-continent somewhat like a bay, wide but not deep, facing south-west towards the northern Maldives -very different from the south-east-facing tip that exists today.

  A tiny island which is depicted on both Reinal’s and Milne’s maps next to the southern tip of the sub-continent. No island exists there today.

  The Maldive islands, which exist today but which are enlarged in both Reinal’s and Milne’s maps.

  How likely is it that such extensive and detailed correlations could have come about by chance?

  22 / The Secret Memories of Maps

  Polo’s explanation of the size accorded Ceylon on the chart was that the chart’s geography originated at an earlier time before much of the island had been submerged.

  Thomas Suarez

  There is a saying that in ancient times the noble isle of Sumatra was joined to the main, until mountainous seas eroded its base and cut it off.

  Camoes, The Lusiads, 1572

  Imagine setting off on a journey along the hippy trail to Afghanistan and the East in 1971 and not getting home again until 1995.

  Though more of a merchant adventurer than a hippy, that’s what Marco Polo did in the dangerous days of Kubilai Khan. He left Venice in AD 1271, travelled to the East via the Black Sea, Persia, Afghanistan and the Pamirs, spent seventeen years in China and seven on the road and at sea, and returned to Venice in 1295. Later he composed a book, Il milione (‘The Million’), known in English as the Travels of Marco Polo, which was to become a geographical classic.1

  Polo’s account of his outbound journey – almost entirely overland – and of his long residence in China, contains little of relevance to the mysteries we are exploring in Underworld. His return journey, however, begun around 1292, is of much greater interest to us here. It includes the first-ever notice by a European of the existence of Japan – which Polo called Cipango (or ‘Zipangu’) from the Chinese Jih-Pen2 – and it describes the epic sea voyage that he undertook on his way home, beginning at the eastern Chinese port of Ch’uan-chou (modern Quanzhou, opposite the island of Taiwan), sailing south around Vietnam and Cambodia, across the Gulf of Thailand, around the Malay peninsula, through the narrow Strait of Malacca that separates the peninsula from Sumatra, thence across the Bay of Bengal to Sri Lanka, around Cape Comorin, north along the west coast of India to the Gulf of Cambay, and finally across the Arabian Sea to Hormuz at the entrance to the Persian Gulf.3 Thus it was that Marco Polo made familiar to Europeans the names and descriptions of many places that would not be heard of again until the Portuguese exploration of India and the Indies in the sixteenth century, more than 200 years later.

  Though Polo himself states frankly that he has never visited Japan – and thus that what he has to say about it is second-hand and perhaps inaccurate – the notion of the mysterious island kingdom of Cipango that he planted in European consciousness at the end of the thirteenth century was later one of several powerful influences that spurred Christopher Columbus forward in his crossings of the Atlantic at the end of the fifteenth century. This was so because Columbus -underestimating the circumference of the earth and knowing nothing of the existence of the Americas or of the Pacific Ocean – believed that he could reach Cipango, and thence the Chinese mainland beyond, by sailing directly westwards across the Atlantic from Europe. Columbus is also likely to have calculated that Cipango would be reached after only a relatively short journey towards the west – for he had read Marco Polo, who describes Cipango, erroneously, as lying ‘far out to sea’ fully 1500 miles to the east of the Chinese mainland4 (the true distance is nowhere much more than 500 miles). Polo goes on to inform us that:

  Marco Polo’s return voyage from Quanzhou to Hormuz.

  Cipango … is of considerable size; its inhabitants have fair complexions, are well made, and are civilized in their manners. Their religion is the worship of idols. They are independent of every foreign power, and governed only by their own kings. They have gold in the greatest abundance, its sources being inexhaustible … The entire roof [of the sovereign’s palace] is covered with a plating of gold, in the same manner as we cover houses … with lead.5

  ‘Gold in the greatest abundance,’ echoes Columbus in a marginal note beside this passage in his own copy of Marco Polo’s Travels — now preserved at the Biblioteca Colombina in Seville.6 We will return to Columbus, and his obsessions.

  A ‘map’ of antediluvian Sri Lanka?

  After traversing the Bay of Bengal, commenting en route on ‘the island of Andaman’ (described as ‘a very big island’ inhabited by ‘a cruel race’ of cannibals with heads, teeth and eyes like those of dogs)7, Marco Polo’s homeward voyage brought him to ‘the island of Zeilan – Ceylon – modern Sri Lanka.8 In his account of Sri Lanka, which further illustrates his already established tendency to exaggerate distances (in this case approximately tenfold) the Venetian traveller nevertheless makes certain observations about the ancient geological history of the region that come remarkably close to the truth:

  The island of Zeilan presents itself. This, for its actual size, is better circumstanced than any other island in the world. It is in circuit 2400 miles, but in ancient times it was still larger, its circumference then measuring full 3600 miles, as the Mappa-Mundi says. But the northern gales, which blow with prodigious violence, have in a manner corroded the mountains, so that they have in some parts fallen into the sea, and the island, for that cause, no longer retains its original size.9 (Emphasis added.)

&nbs
p; This is the translation of William Marsden (1754–1836) from the Italian of Giambattista Ramusio’s printed edition, dated 1553.10 The more recent (1958) translation of Ronald Latham provides clarification of some elements of the same passage:

  The traveller reaches Ceylon, which is undoubtedly the finest island of its size in all the world. Let me explain how. It has a circumference of some 2400 miles. And I assure you that it used to be bigger than this. For it was once as much as 3600 miles, as appears in mariners’ charts of this sea. But the north wind blew so strongly in these parts that it has submerged a great part of this island under the sea. That is why it is no longer as big as it used to be.11 (Emphasis added.)

  In yet another translation we read again that Ceylon in Polo’s day has a circumference of: ‘2400 miles … in old times it was greater still, for it then had a circuit of about 3600 miles, as you find in the charts of the mariners of those seas’.12 (Emphasis added.)