Underworld: The Mysterious Origins of Civilization

by Graham Hancock

  In the midst of the Eastern Sea there are three magic mountain islands, Pheng Lai, Fang-Chang and Ying-Chou, inhabited by immortals. We beg to be authorized to put to sea … to go and look for the abodes of the immortals hidden in the Eastern Ocean.21

  The target of this voyage, which did receive the emperor’s blessing, is stated to be far off ‘in the midst of the Eastern Sea’, but again there is no consensus as to its location. Hsu Fu went to look for it with a well-stocked fleet, said to have been carrying large numbers of young men and women and ‘ample supplies of the seeds of the five grains’22 – which suggests settlement plans. The Shih Chi records that he ‘never came back to China’.23 But confusingly, the same chronicle also reports other voyages – equally fruitless in terms of any definite discovery – which sought the same islands much closer to the Chinese coast:

  From the time of the Kings of Chhi [c.378 BC] … people were sent out into the ocean to search for the islands of Pheng Lai, Fang-Chang and Ying-Chou. These three holy mountain isles were reported to be in the midst of Po-Hai [the Gulf of Bo Hai], not so distant from human habitations … Many immortals live there, and the drug which will prevent death (pu ssu chih yao) is found there, but the difficulty [is] that … before you have reached them … these three holy mountain isles sink down below the water – or else a wind suddenly drives the ship away from them. So no one can really reach them …24

  Convergence

  All taken together it seems fair to say that the Chinese myths contain very much the same sort of strange brew as do their Japanese counterparts – of an entrance to the Underworld, of enchanted islands and of kingdoms beneath the sea. But where Japanese traditions specify the location of the Kingdom of the Sea-King as being somewhere in the Lu-Chu islands, Chinese references to Pheng Lai, Fang-Chang and Ying-Chou, ‘the islands of the Sea Mage’,25 are contradictory as to location – varying from Hsu Fu’s unspecified destination in the midst of the Pacific Ocean to somewhere extremely close to home like the Bo Hai Gulf (which lies between the city of Tianjin and Korea Bay at the northern end of the Yellow Sea).

  Perhaps the contradiction is less than it seems, however, for Hsu Fu is venerated as a Kami in Japan. There he is the Kami Jofuku whose tomb-shrine exists to this day at Shingu in Wakayama Prefecture of southern Honshu,26 which, like Cape Ashizuri in nearby Shikoku, overlooks the course of the Black Current. If there is truth to this strange tradition of Hsu Fu’s settlement at Shingu then it suggests that the islands of the Sea Mage ‘hidden in the Eastern Ocean’ to which he had directed his expedition must all along have been somewhere in the vicinity of southern Japan.

  Although this cannot be confirmed, I suspect that the convergence of myths from both China and Japan does hint at something very real – perhaps even a shared memory of a lost land with ‘palaces and towers’, once believed to be enchanted and inhabited by ‘immortals’, that now lies beneath the sea.

  Lu-Chu and Bo Hai

  And where are we to look for this lost land, should we wish to rediscover it? Across the two traditions the only clear pointers given to its whereabouts are that it is to be found somewhere in the Lu-Chu islands – effectively anywhere along the arc from Taiwan to Kyushu – or near the northern terminus of the Yellow Sea in the Bo Hai Gulf. These locations are not proximate but are at opposite ends of the same region. Both are highly plausible as potential locations for ‘palaces beneath the sea’.

  At the end of the Ice Age we know that the Ryukyu islands were larger than they are today. There was therefore ample room along their antediluvian shores for any number of ‘palaces’ to be built – and later submerged as sea-levels rose. Moreover, as I’ve endeavoured to show in the preceding pages, a number of extraordinary underwater structures that seem increasingly likely to have been made by humans at the end of the Ice Age have already been found around the Ryukyus.

  Likewise, if we look at the Bo Hai Gulf on Glenn Milne’s inundation maps, we discover that it too has an interesting story to tell. Down to 14,600 years ago it was dry and far from the sea (opposite, above). By 13,500 years ago, however, we observe that the Yellow Sea has penetrated deeply inland towards the modern coast of China and has carved out the Korean peninsula for the first time – but the Bo Hai Gulf is still dry (opposite, below).

  Then we come to the map for 12,400 years ago (page 660). In between Shikoku and Honshu we see, still well preserved, the correlation with the Bungo Strait much as it is portrayed on the 1424 chart. And at the northern end of the Yellow Sea we see that the Bo Hai Gulf has at last succumbed to partial inundation. Within it, rather strikingly, an island has materialized. Though it is beyond the resolution limits of Milne’s computer model, it is perfectly possible that the single island shown could then or at some stage afterwards have been divided up into three smaller islands exactly as the Shih Chi seems to remember. Either way its presence in the palaeo-Bo Hai Gulf is intriguing and obliges us to wonder exactly what it was that inspired the Chinese in the third and fourth centuries BC to make so many real voyages into the Eastern Ocean in search of islands that did not exist.

  Could it have been a legacy of ancient maps copied from copies of copies of copies of even more ancient maps, the originals of which had been drawn before the rising sea-levels at the end of the Ice Age gave the world its post-glacial face? If not, what other explanation is there? After all, the Bo Hai Gulf has been in its present form for at least the last 9000 years – so what possible reason could the chronicler of the Shih Chi around 90 BC have had to imagine that there could ever have been any dry land in it, let alone a group of islands? Why should the Chinese have gone to such lengths to seek out those islands – over a period of two centuries – when it was plainly such a fruitless enterprise? And must we resort to ‘coincidence’ again to explain what appears to be a piece of truly anachronistic geographical knowledge in the possession of the mariners who pursued the search – i.e., that an island or islands, which could never be found because it had sunk beneath the waves, did once exist in the Gulf of Bo Hai?

  Think about it. What are the odds against the Chinese seafarers of 2300 years ago getting their palaeo-geography so right purely by chance? Or we can come at the question from another direction? How likely is it that China’s historical quest for the three ‘holy mountain isles’ in the Gulf of Bo Hai was inspired by meaningless myths – as orthodox historians must conclude – when we now know that an island or islands did exist in the Gulf of Bo Hai 12,400 years ago and did subsequently ‘sink down below the water’ as the Shih Chi maintains?

  The palaeo-island was gone within 1800 years, as we see on the next inundation map in the sequence (for 10,600 years ago, above). The maps also show that the entire span of the island’s existence did not exceed 3000 years, since it had not yet taken shape 13,500 years ago

  And though the Shih Chi does not give us a written description of the palaeo-geography of the Yellow Sea for 13,500 years ago (as it seems to for 12,400 years ago), it is a remarkable anomaly of history that something looking very much like a graphic representation of the Yellow Sea and its coastline 13,500 years ago has survived. Now kept in the ‘Forest of Steles’ in Xian, it is a good Chinese map (Needham describes it as ‘magnificent’),27 carved in stone in AD 1137, called the Hua I Thu (‘Map of China and the Barbarian Countries’).28 In a refrain familiar from our investigation of the sometimes strangely anachronistic portolans of the West it is known to have been based on older sources.29 Nobody can be absolutely sure how much older. But if ever there was a land in which we might expect to encounter an ancient map-making tradition, then that land is surely China.

  Hua I Thu Chinese map of AD 1137.

  At around the time that the maps attributed to Marinus of Tyre were being circulated in the Mediterranean, a great Chinese geographer, Chang Heng (AD 78–139) was producing maps of unbelievably high quality in China. Like Marinus, he is often credited by historians with having introduced a grid system for maps – it being said of him that he had ‘cast a net
work of coordinates about heaven and earth, and reckoned on the basis of it’. The title of one of his lost books was ‘Discourse on Net Calculations’ and there was also a ‘Bird’s-Eye Map’.30

  It is clear, however, that Chang Heng, who is considered one of the ‘fathers’ of scientific cartography in China, must himself have been the ‘son’ of a much earlier and older tradition – for one does not reach his level of sophistication without a vast store of prior knowledge and experience to build upon. That such a store or archive did exist, and that it did contain extremely ancient material, is confirmed in the dynastic chronicles which also give prominence to the works of another great Chinese geographer, Phei Hsiu (AD 224–271):

  Phei Hsiu made a critical study of ancient texts, rejected what was dubious [outdated by climate change?], and classified, whenever he could, the ancient names which had disappeared [because inundated?]; finally composing a geographical map in 18 sheets. He presented it to the emperor, who kept it in secret archives.31

  The chronicles also cite the full text of Phei Hsiu’s preface to his Atlas, in which he laments the loss of geographical knowledge from earlier times (emphases added):

  The origin of maps and geographical treatises goes far back into former ages. Under the three dynasties [Hsia, Shang and Chou,32 c. 2000–1000 BC]33 there were special officials for this [Kuo Shih). Then when the Han people sacked Hsien-yang, Hsiao Ho collected all the maps and documents of the Chhin. Now it is no longer possible to find the old maps in secret archives, and even those which Hsiao Ho found are missing; we have only maps, both general and local, from the later Han time. None of these employs a graduated scale (fen lu) and none of them is arranged in a rectangular grid.34

  The implication is not only archives of maps going back thousands of years but also that the rectangular grid was known very early in Chinese history, then fell into disuse under the Han in the first millennium BC, and was then later reintroduced by Chang Heng, the contemporary of Marinus of Tyre, when he cast his ‘network of coordinates about heaven and earth’.

  So we have a confirmed cartographic science in China from around 2000 years ago (Chang Heng, Phei Hsiu), and references to an ancestral tradition more than 2000 years older than that – which presumably was itself not new in 2000 BC when ‘special officials’ already existed dedicated to the archiving and probably copying of ancient maps.

  It is against this long background, therefore, which disappears into prehistory and has no known beginning in China, that we should evaluate the Hua I Thu – the Chinese map of AD 1137 said to have been based on older sources – which I have claimed shows the Yellow Sea and the Korean peninsula not as they looked in AD 1137 but as they looked 13,500 years ago. Although other interesting issues are raised by the Hua I Thu, I will confine my remarks here to the north-eastern segment of the map, around the Yellow Sea. In the diagrams overleaf the reader may compare the Yellow Sea and the Korean peninsula as they appear on a modern map with the same areas on the inundation map for 13,500 years ago and on the Hua I Thu. It will be observed that an excellent level of correspondence does in fact exist between the latter two and that the Hua I Thu’s portrayal, though a bad one of the Yellow Sea as it looked in 1137 – and as it still looks today – is rather a good one if it represents the Yellow Sea 13,500 years ago. Particularly noticeable is the absence on both the Hua I Thu and the inundation map of the Shantung peninsula, a prominent feature of the northern end of the Yellow Sea, which the rising waters began to carve out some time after 13,000 years ago and which took on its modern form about 10,000 years ago.

  (Left) The Yellow Sea as shown on the Hua I Thu map of AD 1137.

  (Below left) Modern map of the Yellow Sea.

  (Below right) The Yellow Sea as it looked 13,500 years ago.

  Nor can it be claimed that the Chinese of 1137 were simply ignorant of the Shantung peninsula. On the contrary we can prove that they knew it very well – because another map, also carved on stone in 1137 and also preserved in the Forest of Steles at Xian, shows it very clearly and with great accuracy much as it looks today. Called the Yu Chi Thu (‘Map of the Tracks of Yu’), it too is a copy of an earlier original but, as Joseph Needham observes, ‘has a more modern look’ than the Hua I Thu and seems to belong ‘to a different tradition’.35

  Could this be because the Hua I Thu’s portrayal of the Yellow Sea was derived directly from a very ancient source map – perhaps stored with many others in the Imperial archives – while the Yu Chi Thu incorporates the results of Chinese maritime expeditions that we know had explored the region thoroughly at least as early as the third century BC?

  Cutting Korea down to size

  Deferring for the moment our parallel interest in the lost islands and sunken kingdoms of the Ryukyu archipelago, what do we have so far concerning Korea and the northern end of the Yellow Sea?

  In summary we have geographical traditions, recorded in the Shih Chi, which place the lost ‘islands of the Sea Mage’ in the Bo Hai Gulf. There was an island in the right place 12,400 years ago. We also have a Chinese map copied from earlier sources on to stone in AD 1137 that anomalously fails to show the Shantung peninsula and that greatly narrows the Yellow Sea between China and Korea. However, there was no Shantung peninsula, and the Yellow Sea was narrowed in exactly this way 13,500 years ago.

  Reduced to bare essentials, therefore, what the Shih Chi and the Hua I Thu both proclaim is that Korea was larger in the past than now. This is completely true. Yet as the inundation maps show, the Korean coastline has remained unchanged for the last 9000 years – having done all its shrinking in the 5000 years prior to that. It follows that if these are memories of a formerly much larger Korea then they must be at least 9000 years old.

  Japan too preserves such memories – if they are memories. In the Fudoki we read of an exploit of Sosano-wo-no-Mikoto, the great Kami called Brave-Swift-Impetuous-Male, whom we encountered in chapter 26. Seeing that parts of the Korean peninsula are much larger than they should be, he removes them, draws them away (‘slowly, slowly, like a river boat’) and sews them on to Japan.36 I have nothing to say about the latter part of the myth, but I do think the bit about the subtraction of land from Korea is interesting: ‘Perceiving that it had a portion in excess, he took up a spade, wide and flat like the breast of a maiden, and thrust it into the land, parting it asunder as one cuts the gills of a huge fish, and severing it.’37

  Sosano repeats this process with several other parts of Korea until he is satisfied38 and, presumably, the peninsula has taken on its present shape.

  The myth is called ‘the drawing of the lands’. What it conjures up in my mind are not images of a spade shaped like a maiden’s breast, attractive though the concept may be, but inundation maps of this area between roughly 14,600 and 10,600 years ago. These do show lands being ‘sliced away’ piecemeal as the basin of the Yellow Sea filled up to allow the Korean peninsula to emerge. Therefore, although it may have come about by chance and have no significance whatsoever, what confronts us in this text is another time-capsule of accurate geographical information on the region as it looked during the meltdown of the Ice Age.

  The Fall

  There are, I think, too many such time-capsules of ancient geography scattered across too many sources from too many lands – myths and folklore, maps and traditions – for every example to be explained away as coincidence. I am convinced that something must lie behind all this and that the odds are rising in favour of a significant forgotten episode in the story of civilization localized in time at the end of the Ice Age. The hypothesis I have followed, which receives virtually unlimited support from world deluge myths, is that the discontinuity – some might call it the Fall – was a direct product of episodes of post-glacial flooding and linked cataclysms. So it follows that the evidence for what we have lost – which might explain how and by whom the world came to be mapped more than 12,000 years ago – should be found on the bottom of the sea.

  The entire ‘arc’ from Taiwan in th
e south, north-eastwards through all the islands of the Ryukyu archipelago, brushing the tip of Kyushu, leaping across the Korea Strait and thence into the Yellow Sea, Korea Bay and Bo Hai Bay, encloses an area with enormous potential for underwater discoveries.

  For me it is an underworld – an ancient domain of forgotten ancestors. Like the others we have entered in this book – in the Mediterranean, in the Atlantic, in the Indian Ocean – I believe it will have to be explored thoroughly one day if we really want to know the truth about our prehistory.

  But by March 2001 I was also beginning to feel that I personally had done all I could to initiate the necessary exploration – and after four years of diving in the Ryukyus I had every reason to expect that the trip with Wolf Wichmann would be the last I would need to do. As I reported in chapter 29, however, Komakino Iseki changed my mind. The resemblance of its circles of river stones to the circles of river stones I’d glimpsed on the sea-bed at Kerama had to be followed up.

  Then came the news that an underwater site had been discovered in Taiwan’s Peng-Hu archipelago – the Pescadores Islands – and my Japanese colleagues and I began to plan a short expedition there for late August 2001. Since Taiwan is so close to Okinawa, it obviously also made sense to redive Kerama on the same trip.

  Getting from base-3 to point ‘D’

  August 2001, Taiwan

  We’re nearing the end of a long story and this is not the place to introduce new characters, plots or locations. But I will mention some of the qualities of Taiwan which place it firmly amongst the usual suspects.