As they have for millennia, winds and currents in the South Atlantic circle anti-clockwise in a huge oval loop from the Cape of Good Hope in the south to the ‘bulge’ of Africa in the north. At the Cape, the mariner meets the Benguela current that carries him due north up the west coast of Africa. After some three thousand miles, the current starts to hook first to the north-west, then westwards to South America. Off the coast of South America the current continues its anti-clockwise movement, running southwards off Brazil and Patagonia down the east coast as far as Cape Horn before sweeping to the east, back to South Africa. If a sailing ship, carrying sufficient supplies and robust enough to withstand the ‘Roaring Forties’ – powerful winds that circle the globe for hundreds of miles north and south of the latitude that gives them their name – were to hoist its sails off South Africa and sail before the wind and current, then several months later, having crossed thousands of miles of ocean in this great anti-clockwise loop, it would return more or less to where it started. An illustration of this is provided by the epic voyage of a very brave and distinguished submarine captain, now Vice-Admiral Sir Ian McIntosh KBE CB DSO DSC, once captain of the submarine squadron in which I served. He wrote to me:
In March 1941 I was a Sub Lieutenant in a merchant ship taking passage to Alexandria. She was sunk by gun-fire by an armed commerce raider some 500 to 600 miles west of Freetown at about 08°N 30°W. The 28-foot standard wooden lifeboat, ‘authorised’ capacity 56, finally had 82 souls on board.
Even when I had repaired the shrapnel holes in the hull and the boat was reasonably dry I could not get her to sail closer than 5 or 6 points to the wind, a brisk NE Trade. This would never have allowed us to reach Africa, and to run before the wind to Brazil some 1,600 miles distant by the route chosen seemed preferable.
The circulatory winds and currents in the South Atlantic Ocean.
The plan was to steer due west until reaching 33° West then alter course to SW. This made full use of the NE Trades and gave us only a few days of shifting winds (and some most welcome rain) in the Doldrums before picking up light SE Trade winds. We made the [South American] coast on the 22nd day, ran NW along the coast looking for a suitable landing, which we found on the afternoon of the 23rd day.
I had estimated a maximum of 28 days for rationing purposes knowing that the equatorial currents were helping us but I had no idea at our latitude and time of the year whether they were a quarter knot or more than 1 knot, so I disregarded it in my noon DR [Dead Reckoning] positions.16
It is entirely feasible that the treasure fleets did reach the Cape of Good Hope where they would have been swept by the wind and current around the Cape and up the west coast of Africa to the ‘Garbin’ described by Fra Mauro. What I now urgently needed was independent evidence that this had happened. I pondered this question for months. Then I had a stroke of luck. John E. Wills Jr, Professor of History at the University of Southern California, and Dr Joseph McDermott, Professor of Chinese at Cambridge University, England, suggested to me that although the charts and records of the treasure fleets in China had been destroyed, there might be copies in Japan, for Japanese scholars were particularly interested in the early Ming era.
Subsequent research revealed that Ryukoku University in Kyoto held a copy of a Chinese/Korean chart known colloquially as the Kangnido. The Korean ambassador had presented Zhu Di with this extraordinary world map in 1403 after his inauguration as emperor. The original map, however, has been lost, and the Ryukoku version of the Kangnido was extensively modified after 1420. It is nearly square and strikingly large, measuring 1.7 by 1.6 metres. Painted on silk, it remains in excellent condition, its colours little faded by the passing centuries. It is ‘nicely organized and well worth admiration. One can indeed know the world without going out of the door.’17
The Kangnido gave a grandiose panoramic view of the world as seen in the early fifteenth century, and was compiled from many different sources. Names for Europe were in Persian Arabic, central Asia came from the Mongols, China and south-east Asia from old Chinese maps. Europe was covered in names as far north as Germany (named Alumangia). Spain was depicted, as were the straits of Gibraltar leading into the Mediterranean and the North African coast with the Atlas mountains. Europe, Africa, Asia, Korea and China were in their correct positions relative to one another, though Korea, perhaps for reasons connected with national pride and its traditional rivalry with Japan, was shown vastly larger than it should be and Japan much smaller. Nonetheless, it was an extraordinary piece of mapmaking.
For the moment, the part of the Kangnido that interested me most was Africa. So accurately does the Kangnido depict the coasts of East, South and West Africa that there cannot be a shred of doubt that it was charted by someone who had sailed round the Cape. Europeans did not reach South Africa for another sixty years; Arab navigators on the west coast never sailed south of Agadir in modern Morocco, eight thousand kilometres away, and the Mongols never reached Africa at all. The accuracy of the Kangnido told me that Mauro/da Conti’s description made absolute sense. A Chinese navigator could indeed have reached ‘Garbin’ and then drawn the Kangnido. Still I had no precise location for Garbin save that, from the shape of the coastline shown on the Kangnido, it appeared to be near the Bay of Biafra, off western Nigeria. It was a problem I would have to address later. For now, I felt justified in assuming that the ‘junk’ referred to by Fra Mauro and drawn on his planisphere was from the treasure fleets, for Chinese merchant ships did not sail beyond Kilwa in East Africa. The Kangnido was much less accurate when it came to the ‘bulge’ of Africa north of the Bay of Biafra, so I next turned my attention to that part of the voyage. If they had managed to survey the coast of southern Africa with such accuracy, why was the bulge of West Africa not shown on the Kangnido chart?
(i) The Kangnido map showing Africa.
(ii) The Kangnido map corrected for longitude.
(iii) Modern Africa.
By the time the Chinese fleets reached the Bay of Biafra, they had sailed some three thousand miles north from the Cape. I assumed that they rounded the Cape on their outward journey some time in August. At their average speed of 4.8 knots it would have taken about twenty days to sail from the Cape to ‘Garbin’. They would have reached it in late August or early September 1421, the end of summer and towards the end of the rainy season. As I well know from my own time at sea in the South Atlantic, there is an extraordinary natural phenomenon in this part of Africa. Starting in the Bay of Biafra, the south equatorial current runs first to the north past São Tomé e Príncipe Island (where the bulge starts) then hooks westwards to flow due west along the south coast of the bulge, past Nigeria, Ghana and the Ivory and Gold Coasts until it peters out a thousand miles out into the Atlantic around 21°W. This massive body of cold water flows westwards with considerable speed the whole year round; a minor change occurs in summer when it extends further north to reach 5°N, a similar latitude to Monrovia in modern Liberia.
This current would have had two important implications for the Chinese: they would have been carried due west for some 1,800 miles, but they would not have known that this had happened. At this stage of their voyage the Chinese could only measure longitude by estimating their speed through the water, and if the great body of water was itself moving, either against them or with them, there was no way that they could determine their position with any accuracy, any more than a man walking up an escalator can judge the distance he has travelled by the number of paces he has taken. With mounting excitement, I realized that the charts drawn after they had entered the south equatorial current had to be adjusted to take account of this discrepancy, and the land they showed moved by up to 1,800 miles further to the west. I went back to my copy of the Kangnido and adjusted the land north of the Bay of Biafra to allow for this longitudinal error. The result was startling: the familiar outline of Africa became immediately recognizable. It appeared that the Chinese had been carried by the wind and current to the ‘bulge’ of Africa forty yea
rs before the first Europeans set eyes on it.
The south equatorial current gave them a ‘free ride’ westwards until the current petered out a thousand miles into the Atlantic. By then, they were in the south-east trade belt and being blown towards the coast of Senegal. In the wet season, running from April to October, the Sénégal current off this coast of West Africa reverses its normal direction and runs northwards along the coast at a rate of 0.6 to 1 knot. Yet again the junks would have had a free ride, this time to the north for around five hundred miles until the current itself petered out off Dakar, the modern capital of Senegal. By then they were in the belt of the north-east trade winds, blowing them south-west to the Cape Verde Islands. These lonely islands, then unknown to Europeans, were to play a vital part in unravelling the mystery of the Chinese voyages.
I checked and rechecked my calculations. By late September the junks that left the Cape of Good Hope in August would have found themselves approaching the Cape Verde Islands from the north-east. The design of the ships and the prevailing winds and currents would have prevented these flat-bottomed, broad-beamed monsters from sailing south at any point. It was now clear that Fra Mauro’s account was entirely possible and that the Cape Verde Islands could have been the ‘Isole Verde’ reached by the ship or junk from India, forty days after leaving the Cape of Good Hope; they even had the same name. At 4.8 knots, the speed the treasure fleets averaged over all six great voyages, this would have taken forty days. Vasco da Gama took thirty-three days to make the same passage in the closing years of the century.
The journey to the Cape Verde Islands.
To have been called the ‘Isole Verde’, the islands Niccolò da Conti described to Fra Mauro must have been strikingly green. I knew the Cape Verde Islands well, having sailed through them in HMS Newfoundland. They are divided into two groups, and the windward (balaventos) are significantly wetter than the leeward (sotaventos). Of the windward islands, the biggest, highest, wettest and greenest is Santo Antão. It is an island of savage grandeur, awesome and eye-catching from the sea, particularly to a mariner seeking fresh water. The Chinese admirals would have been approaching from the north-east on the trade winds, and from that direction they would have sighted Santo Antão first. On the north coast of Santo Antão, clearly visible from miles out as you approach from the north-east, there is a dramatic volcano. Streams pour down its sides and rush through lush valleys to the sea around what is now the small settlement of Janela. That strip of coast would have been an obvious and immediate place for the Chinese fleets to anchor and obtain water. If the Chinese had indeed landed there, I was confident that a legacy of their visit should exist.
The Cape Verde Islands were uninhabited when the first European, Cà da Mosto (1432–88), a Venetian explorer in the service of Henry the Navigator, arrived in 1456, so I could not expect to find goods that the Chinese had traded for food, such as the blue and white plates that were their currency on the south-east coast of Africa. On the Cape Verde Islands the Chinese could have obtained any amount of food and water for nothing. The seas teemed with swordfish, sole, shark, octopus, crayfish and tiny sweet mussels, the island was lush with fruit, and flocks of tame birds could be picked up by hand, for they had never learned to be wary of humans, as Cà da Mosto’s crew found to their joy thirty-five years later. Nonetheless, there should have been other mementoes. A carved stone similar to the one erected by Zheng He on the estuary of the Yangtze stands at Galle, near Dondra Head in southern Sri Lanka. Inscribed in Chinese, Tamil and Persian, it extols the virtues of Hinduism (the local religion), Buddhism (Emperor Zhu Di’s faith) and Islam (the religion of most Indian rulers in the early fifteenth century). There are other similar stones near Cochin and Calicut. I wondered if a carved stone might have been erected here.
The Chinese were always careful to respect local sensibilities; the language school in Nanjing, the Ssu-i-Quan, was, after all, set up by Zheng He specifically to train interpreters, and the fleets on this sixth voyage carried interpreters fluent in seventeen different Indian and African languages. It was highly probable that they had also left a stone on one of the Cape Verde Islands, carved with inscriptions in a language they thought people from the surrounding areas would understand. Such stones were always sited in prominent places where they would readily be discovered by others – what would be the point of erecting a monument to your achievements and then hiding it where it would never be found? If such a stone existed, the first Europeans should have found it when they reached Santo Antão thirty-five years later.
I referred to the journals describing Antonio da Noli, Cà da Mosto and Diego Alfonso’s first voyages to the islands, and discovered that they had indeed found a large, free-standing stone near the coast at Janela. The stone still stands there today, in a dramatic setting framed by encircling mountains, beside the Ribeira de Penedo. Until a century ago, a clear, rushing stream tumbled down the side of the volcano, but now the stream has dried up and the stone is surrounded by agave plants. The stone, called locally Pedra do Letreiro (Stone of Letters), is of red sandstone, some three metres high and covered with inscriptions from top to bottom. The later carvings are in medieval Portuguese, commemorating the death of a mariner, Antonio of Fez, but underneath them I could see more calligraphy, unfortunately obscured by moss and lichen. The stone was so badly weatherworn and defaced by recent graffiti that it was very difficult to decipher the underlying calligraphy. A series of experts had tried – first a Frenchman, M. Chevalier, in 1934, then several learned Portuguese and Cape Verde historians over the past twenty years. They could tell me what the calligraphy was not – it was not Arabic, Judaic, Berber, Tifnaq, Aramaic, Phoenician, Latin, or any other European language – but they could not tell me what it was.
After receiving the necessary approval from the Cape Verde authorities, some of the lichen was removed. This revealed two pieces of calligraphy. I hoped that, helped by computer enhancement, I would at least be able to determine the language, but the calligraphy was quite extraordinary, unlike anything I had ever seen in my travels anywhere in the world. It appeared to have two characteristics: whorls like interlocking ram’s horns, and a number of concentric circles.
My first thought was that it could be medieval Chinese, either the Zhu Qi Shan script or ‘Flowinghand’. I sent photographs to experts at the Forest of Steles in Xian, China. Once the Temple of Confucius, it is now a museum and library holding a huge collection of steles, or engraved stone tablets, a timeless memorial to the Chinese written language. It was neither script. Could it be Tamil, similar to the writing on the stone the Chinese erected in southern Sri Lanka? It does resemble Tamil, but not closely enough. Nor is it Swahili, the lingua franca of the east coast of Africa. I then wondered if it could be another Indian language, perhaps one of the thirteen shown on today’s high-denomination Indian banknotes. Could the Bank of India help? I faxed them a photo of a small section of one of the pieces of calligraphy.
‘It looks like Malayalam,’ they replied.
It was a language I had never even heard of. I faxed again.
‘Where was this language spoken?’
‘It was the language of Kerala.’
‘Was it in use in the fifteenth century?’
‘Yes, it had been in common use since the ninth century.’
Once I’d put down the phone, I punched the air in my excitement. In 1421, Kerala’s capital was Calicut, the great port of India from which the Chinese had sailed. Once again, Fra Mauro and Niccolò da Conti seem to have been correct: a ship or junk from India appeared to have reached the Cape Verde Islands before the Portuguese arrived.
I next trawled through the learned experts’ research18 to see whether they had come across another, similar stone while they were attempting to decipher the writing at Janela. They had, but not in the Cape Verde Islands. The other stone was sited at the Matadi Falls in the Congo. My first impression was that this was a wild goose chase. Why should a ship voyaging from India have visited a wat
erfall in Africa? But closer examination revealed the Matadi Falls to be at the upper navigable limit of the Congo River, where a mariner may anchor in beautiful surroundings and obtain clean, fresh water. A succession of ships had done just that down the centuries, from the Portuguese in 1485 to the Chinese today. The river pours over a series of cataracts before reaching the falls. The carved stone stands sentinel above a dark pool near the foot of the falls, where in days gone by fishermen would sit motionless while prostitutes patrolled the banks awaiting the arrival of the crews of foreign ships sailing upstream to water and gather provisions.
I had to retrace my route back to the coast of Africa to investigate this discovery. Like its Janela counterpart, the Matadi Falls stone has calligraphy beneath medieval Portuguese. The Portuguese writing once again commemorated a deceased comrade, here the navigator Álvares. There is less underlying material than at Janela, but experts confirmed that it was the same calligraphy, that once again it looked like Malayalam. Its identity appeared to have been solved, although the concentric circles remained a mystery. It was likely that the Chinese had come here on their journey up the African coast. Not only is the Matadi Falls an ideal place for watering, it is ‘in the middle of the west coast of Africa’, and fits the description of ‘Garbin’ given by Fra Mauro. It is a busy port today.
Once again, Fra Mauro and da Conti appeared to have been vindicated: a ship sailing from India ‘around the year 1420’ seems to have reached Garbin. This does not, of course, guarantee that the ship was Chinese rather than Indian, but Indian ships were sailing with the Chinese fleets. It appeared the Chinese had reached there. The simple and obvious explanation was that the calligraphy carved on the ‘Garbin’ and Janela stones was inscribed by stonemasons travelling with the Chinese fleet, just as they had carved inscriptions in foreign languages at Dondra Head, Cochin and Calicut.
1421: The Year China Discovered the World Page 10
