70. Stones cut with the same technique as Masada-no Iwafune at a depth of 20 metres, Yonaguni, Japan.
71. Jomon magatama curved stone ornament, Japan. Compare with 72, right.
72. Maltese curved stone ornament in National Museum of Malta. The only one of its kind ever found in Malta, it is identical to the magatama of the Jomon of Japan.
73. Small stone circles, Komakino Iseki, northern Japan. Compare with 74, below.
74. Small stone circles at a depth of 30 metres, Kerama, southern Japan.
75. Iseki Point, the main underwater monument of Yonaguni.
76. The author diving beside the ‘Face’, Yonaguni.
77. Underwater megalith, Yonaguni.
78. View down on Centre Circle, Kerama, Japan, at a depth of 30 metres.
79. Centre Circle and Small Centre Circle (foreground), Kerama.
80. Centre Circle, Kerama.
81. The author diving beside the parallel megaliths, Yonaguni, Japan.
When we compare the 1424 chart with the inundation map sequence (pages 632–4), no obvious correlation emerges down to as late as 14,600 years ago, when Kyushu, Honshu and Shikoku were still so firmly bonded together by lowered sea-levels that even the Bungo Strait did not exist.
Within just another thousand years, however, around 13,500 years ago, the inundation maps show that a squarish inlet topped by a narrow, north-east-trending, fjord-like channel very similar to the portrayal of the Bungo Strait on the 1424 chart had opened up.
The correlation remains much the same on the inundation map for 12,400 years ago, although it is possible to detect a slight opening to the north-west, not shown in the 1424 chart, in what was to become the Suo Gulf.
By 10,600 years ago, however, the correlation is much less precise, with both the Suo and Iyo Gulfs opening up into fat cloverleafs north-west and north-east of the Bungo Strait.
Finally, by 8900 years ago, the submergence of the shorelines around the Inland Sea approaches today’s levels, Shikoku, Kyushu and Honshu begin to emerge as separate islands, and the geography of the 1424 chart becomes and remains an anachronism.
Bearing in mind the limitations of inundation science – these maps are models, based on the latest data but do not claim 100 per cent accuracy – there is no doubt that the best correlation between the 1424 chart and the actual appearance of this part of Japan comes not in 1424 but in a specific and clearly demarcated 1100-year time-window between 13,500 years ago and 12,400 years ago.
Coincidence? Or the leavings and memories of ancient world maps preserved amongst mariners since the end of the Ice Age in fragments, and copies of fragments, and fragments of copies?
Satanaze from the 1424 Pizzagano chart.
What about Taiwan?
As I was considering the implications of this interesting problem it occurred to me – since Satanaze appears together with Antilia on the 1424 chart – that the two islands probably appeared together on the source map too. In that case the treatment of Antilia/Taiwan on the 1424 chart could serve as a useful control to speculation about Satanaze/Japan. If, for example, it should turn out that Taiwan’s 1424 portrayal was best matched by the modern appearance of the island and bore no resemblance to the inundation maps then it would make it more likely that any Satanaze/Japan correlations were just coincidences. On the other hand, if Antilia and ancient Taiwan matched up well to one another, and especially if they were to do so in the same time-window as Satanaze/Japan, then I thought this would make it much more likely that the similarities had been derived from a common source map that had contained accurate depictions of Japan and Taiwan as they had looked at the end of the Ice Age.
At the beginning of the meltdown around 16,400 years ago lowered sea-levels meant that Taiwan was not an island but was, instead, fully integrated with the east coast of China. The inundation maps show its distinctive narrow south-eastern tip, which has changed its appearance very little over time, protruding as a peninsula from a vast antediluvian landmass extending eastwards for hundreds of kilometres from the present Chinese coast. These long-lost coastal plains, fertile with the silt of the ancient Yangtse and Yellow rivers, were wide and extensive enough to incorporate the entire Korean peninsula much further north, completely filling the basin of the Yellow Sea, and the Bo Hai and Korea Bays (page 635).
The situation of Taiwan had not dramatically changed two millennia later, as represented in the inundation map for 14,600 years ago. We can see there (page 635) that it has made some progress towards its eventual destiny as an island but that it is still very much fixed to the mainland and as such offers no correlation with the 1424 chart of Antilia/Taiwan. In fact the inundation maps show that Taiwan did not become an island, and thus did not even become eligible for comparison with Antilia, until 13,500 years ago (above).
It is notable, therefore, when we compare its appearance at that time to the outline of Antilia, that we immediately find a tantalizing resemblance, but certainly not an exact one. For, although the inundation map shows Taiwan as an island of roughly the right shape, it also shows a distinctive peninsula protruding from the mid-latitudes of its west coast that is not to be seen anywhere on Antilia. Instead, the 1424 chart gives us a second smaller island, named Ymana,5 roughly where the peninsula on the inundation map ends.
The next inundation map in the sequence, which shows Taiwan as it looked 12,400 years ago, is where things get interesting. Very strikingly the peninsula has vanished and what remains is an island of the right size and in the right location to match Ymana (page 638).
Again, is it a coincidence?
Antilia from the 1424 Pizzagano chart.
Taiwan 12,400 years ago.
Here the logic that led me to look for Antilia/Taiwan correlations at the end of the Ice Age (as a control on the apparent correlations I had noticed between Satanaze and Japan in the same period) works in reverse to reduce the likelihood of coincidence still further. Of course it still could be a coincidence. The fact is, however, that the representations of Antilia and Satanaze on the 1424 chart not only appear to have captured characteristics of Taiwan and Japan as both looked during the meltdown of the Ice Age but, far more impressively, as both looked during exactly the same ‘window’ between 13,500 and 12,400 years ago.
Objections
There are two important objections to this line of reasoning, which must be registered and responded to immediately.
First, despite its steep coastlines, Japan did undergo significant changes to its appearance at the end of the Ice Age when one antediluvian island – Satanaze on the 1424 chart – was filleted into segments by the rising seas to form modern Kyushu, Shikoku and Honshu. Taiwan’s even steeper coastlines, by comparison, have changed much less since it first became an island around 13,500 years ago. Thus to the extent that Robert Fuson is right at all to identify Antilia as a map of Taiwan, then it could, theoretically, be a map of Taiwan in almost any epoch after 13,500 years ago. As such isn’t it too vague and general an indicator to be useful for any particular purpose or to draw any specific conclusions from?
My response to this objection is that Antilia has more information to reveal than at first meets the eye.
Superimposition of 1424 Antilia on modern Taiwan.
A good starting point is Fuson’s own superimposition of Antilia on to a modern map of Taiwan. As the reader will observe, other than the overall dimensions and the roughly rectangular shape being about right, the correlations between the coastlines of the two islands are not in fact particularly good (and would make no case in themselves were it not for the many other convincing comparisons between Antilia and Taiwan that Fuson is able to present).6
When compared with the modern map Antilia does best in the south-east -where both it and Taiwan come to a distinct sharply pointed, south-east-facing cape. But in the south-west, north-west and north-east the island depicted on the 1424 chart extends many kilometres beyond the coastal margins of Taiwan as it looks today.
Is it anothe
r coincidence that two out of these three supposed ‘mistakes’ in Antilia’s portrayal of the main island of Taiwan would make perfect sense if the source maps showed Taiwan as it looked around the end of the Ice Age? There is no match at any date for the triangle of land that Antilia adds on to Taiwan in the south-west. But the extra lands that Antilia also claims in the north-west and the north-east of Taiwan do correlate closely with extra lands – then still above water in precisely these areas – that show up on the inundation map for 12,400 years ago. Since 12,400 years ago is also the date that provides the best fit for the island of Ymana on the 1424 chart, coincidence seems to me an explanation that is increasingly difficult to defend …
Superimposition of 1424 Antilia on Taiwan as it looked 12,400 years ago.
However, it is precisely here that a second objection must be registered and responded to. One of Robert Fuson’s proofs that Antilia is Taiwan, cited in chapter 24, is that: ‘Taiwan also has something else that Antilia must have. And that is a small island to the west. On the 1424 Pizzagano chart it was called Ymana. Today it is the Peng-Hu group, or Pescadores (Islands of the Fishermen).’7 It is the Pescadores, exaggerated into a single larger landmass by cartographers’ errors, that Fuson speculates served as the model for Ymana.
My response is that the location of the Pescadores in relation to the main island of Taiwan – marked today by little more than dots on the map – is not identical to the location of Ymana, but considerably further south. By contrast, as we’ve seen, the inundation map for 12,400 years ago provides a single antediluvian island of the right size and in the right location to be Ymana. The same map shows us that the Pescadores were, at that time, still part of the Chinese mainland and lay at the tip of a peninsula some 200 kilometres south of my antediluvian candidate for Ymana. They did not finally become islands -initially just one island – until around 10,600 years ago and thereafter were gradually broken up into the many much smaller remnants that still survive today.
It remains entirely possible that Fuson is right and that it was the Pescadores that served as the model for Ymana – though I note in passing that they would have more closely resembled Ymana when they were consolidated by lowered sea-level into one island around 10,600 years ago than at any much later date.
I was therefore overtaken by an irresistible feeling of curiosity when news came to me from my Japanese friends that extensive underwater ruins had been discovered in the Pescadores. Lying off the south shore of a tiny island called Hu-Ching – it means ‘Tiger Well’ – the ruins were said to consist of two gigantic walls crossing each other at right angles extending from a minimum depth of just 4 metres to a maximum depth of more than 36 metres. It was too tempting a prospect to pass up and Seamen’s Club were willing to fund one more trip. Santha and I packed our dive gear and flew out to Taiwan at the end of August 2001.
But I’m getting ahead of my story. Before we fast-forward to Taiwan we need to rewind to the end of chapter 27 and the mini-expedition to Japan’s Ryukyu archipelago that I made in March 2001 with the German geologist Wolf Wichmann. The reader will recall that Wolf and I left Yonaguni, the westernmost of the Ryukyus, after failing to reach agreement on the provenance of the underwater structures there. Our next destination was Naha, capital city of the much larger island of Okinawa, where we would be a one-hour journey by boat from what are perhaps the most extraordinary and enigmatic underwater structures in all of Japan – the great stone circles of Kerama.
29 / Confronting Kerama
I agree that this is very amazing and very strange, even to me, how these structural buildings could be formed. Patterns like these, I haven’t seen formed by nature.
Dr Wolf Wichmann, geologist, Kerama, Japan, March 2001
Although I usually refer, in shorthand, just to ‘Kerama’, the correct term is ‘the Keramas’ – for this is in fact a group of small islands, including Aka, Zamami, Kuba and Tokashiki, lying in the Pacific Ocean about 40 kilometres due west of Naha, the capital of Okinawa.
The islands are poignantly beautiful, with verdant hills, rugged, rocky coasts and sand-fringed beaches, and they are separated from one another by expanses of crystal-clear water ranging in intensity from the palest turquoise to the deepest midnight blue. The whole area is a marine nature preserve renowned for the great numbers and varieties of whales and dolphins that congregate there.
And at the end of the Ice Age? The story that Glenn Milne’s inundation maps tell is that down to about 14,600 years ago Kerama remained attached to the southern end of Okinawa by a thick, curving tongue of land. Okinawa was itself at that time a much larger and wider island than it is today with many kilometres of low-lying, gently sloping plains extending both east and west of its present coastline. Indeed, it is on these now inundated plains off its south-western coast that Okinawa’s own underwater monuments – the ‘step-pyramids’ and ‘terraces’ off-shore of Chatan, described in chapter 1 – are located. And at that time there was continuous land between Chatan and Kerama …
Looking further through the inundation sequence we find that by 13,500 years ago the Kerama-to-Okinawa land-bridge had been severed and 20 kilometres of water lay between the two. But it is also clear that Kerama at that time had not yet broken up into smaller units. Further detail is difficult to resolve, but the maps indicate that this single, larger Kerama may have survived, with minimal diminution, until as late as 10,000, perhaps even 9000, years ago – though since parts of it were steeply sloping, and parts flat, not all of it would have been submerged at the same moment even then. It would have been around this time, 9000–10,000 years ago, that Kerama’s stone circles would have been inundated.
The circles lie under almost 30 metres of water, 10 kilometres south-east of Aka island, at the intersection of latitude 26 degrees 07 minutes north and longitude 127 degrees 17 minutes east. A few jagged rocks just break the surface near by, with waves constantly crashing over them, but otherwise the site is completely exposed in open water.
The constraints
Kerama, March 2001
The March 2001 dives with Wolf Wichmann were funded and filmed by Channel 4 on a rushed, money-saving schedule – two working days for Yonaguni, and one for Kerama. In practice this meant that if the weather turned sour – which it frequently does in the Ryukyus – we would not be able to dive at Kerama at all. And even if the weather god was with us, the sea god might not be: the currents at Kerama are often so severe that you have to fight the water continuously if you want to stay in one place.
When humans fight water, water wins. I’ve seen divers lose their masks and have their regulators pulled from their mouths by the Kerama currents. I’ve seen desperate, breathless struggles to stay on top of the site, or to help others stay there, and not get swept away into the wild blue yonder. I’ve seen fit young adults crawl back on to the boat exhausted, literally trembling with fatigue. So what I’ve learned, after several unpleasant experiences of that sort, is that it’s just not worth diving there when the current flows. It’s better to anchor the boat tight with a couple of lines fore and aft, put a buoy in the water, watch how it bobs, and wait for a lull.
If there’s a lull.
Briefing
Kerama, March 2001
We set out from Okinawa soon after 9 a.m. on what turned out to be a reasonably fine morning with waves of less than a metre. Once again we were working with the great local diver Isamu Tsukahara and his very professional team, and using his fast, spacious cabin cruiser as our dive boat. Mitsutoshi Taniguchi, the original discoverer of the circles, had come up to join us from his home on Miyako island further to the south. And Kiyoshi Nagaki had also volunteered to dive with us that day.
We began to sight the Keramas after about an hour of steady running to the west, and as we drew closer Wolf explained to me their basic structure evident from areas of bare rock along the coasts and from scars left by earthfalls that had uncovered the underlying strata in the hills. Rather like Malta in the far-off Mediterra
nean, it seemed that these islands had been formed out of huge deposits of coralline limestone (i.e., corals turned to rock) that had been laid down under ancient seas as much as 50 million or 100 million years ago and then subsequently exposed and inundated again, exposed and inundated again, with more coral growth taking place in the epochs of inundation but later itself being fossilized and exposed. In some places sedimentary layers of softer limestones, comparable to Malta’s globigerina layers, lay on top of a coralline core. In others coralline outcrops formed the surface layer itself, glaring white in the morning sunlight.
Plan drawing of Kerama’s stone circles and associated structures. Based on Kimura.
By 10.30 a.m. we were manoeuvring into position over the dive site. Isamu Tsukahara – who always takes the hardest work on himself – went down to set the anchors and the buoy. This must have required an almost superhuman effort on his part, since the current was flowing strongly enough to create visible turbulence on the surface, but he calmly and capably succeeded and was soon back on board none the worse for wear. Then we all sat around and waited, listening to the creaking anchor ropes as the current tried to rip the boat free and send it spinning back to Okinawa. The buoy, rather distressingly, had been sucked completely underwater by the force of the flow, and no diving was going to get done before it popped back up again.