Prehistoric Coastal Navigation: The Illa-Tiki and La Manteña expeditions

Section 1

1: Inspiration by Thor Heyerdahl

2: Connecting the pre-Incan and Aztec worlds

Section 2

1: The Illa-Tiki expediton

Section 3

1: La Manteña: New expeditions, same teredo navalis

2: From Ecuador to Columbia

3: From Columbia to the gyre



Balsa raft explorer John Haslett (photo courtesy of


































































The launch of Illa-Tiki, 1995 (photo courtesy of )






















































Sketch upon which experimental archaeologists such as Thor Heyerdahl and John Haslett based their constructions of balsa rafts (after a sketch made in Ecuador by F.E. Paris in the 19th century).


















































































































La Manteña (photo courtesy of .



Section 1: Heyerdahl, guara, and the perils of prehistoric coastal navigation


1: Explorer John Haslett, inspired by Thor Heyerdahl and the Kon-Tiki expedition


            As a sailor and an avocational historian with an interest in the archaeology of Central and South America, explorer John Haslett had long been fascinated by the Kon-Tiki expedition.  In the 1980s, Haslett became determined to follow Thor Heyerdahl across the Pacific, by building a balsa raft and sailing it from Mexico to Hawai’i.

By the spring of 1994, Haslett had formulated his plans to such an extent that he sought the advice of Thor Heyerdahl himself.  The Norwegian responded by stating his belief that the Hawai’ian islands were well within the range of a balsa raft.  But the success of such a voyage depended heavily on the effective use of the guara, which Heyerdahl himself had demonstrated in experiments conducted five years after Kon-Tiki.  To solve the problem of crossing the doldrums, Heyerdahl suggested that the raft could be paddled, depending of course on the size of the raft and the number of the crew.  Heyerdahl reminded Haslett that for two hundred years Spanish caravels crossed the Pacific from Mexico directly to the Philippines, but always had to return on the Japan Current that carried them north of the Hawai’ian islands. 

In addition to his contact with Heyerdahl, Haslett conducted research on a group of cultures occupying pre-Columbian Ecuador and known collectively as the Manteños.  It was in this Manteño area of coastal Ecuador where—along with the maritime culture areas of Taiwan—that archaeologists had placed the development of the guara centerboard.



2: Connecting coastal South America with coastal Central America by raft


            Haslett was intrigued by archaeological evidence linking the Manteño areas with those of the Aztecs of Mexico.  Similarities between culture areas of Western Mexico and Ecuador included ceramics, clothing and, especially, techniques of metal production.  Explorer Gene Savoy believed that the two areas had been joined in prehistory by totora reed boats, and sought unsuccessfully to demonstrate such a link in the voyage of the Feathered Serpent in 1969.  Haslett considered it much more likely that the giant balsa raft freighter, with its massive cargo of trade goods, made a more suitable vehicle for such contact.  To demonstrate that a balsa raft from Ecuador could reach Mexico, however, one could not simply drift with prevailing currents.  The raft would have to be sailed like any sailboat. 

As Savoy had discovered, the northwestern coast of South America is a difficult place for primitive craft.  Winds and currents play havoc with coastwise navigation.  The waters are filled with rocks and reefs; surface and shore by pirates and bandits.  If the raft lost control near the shore, a wreck was certain.  If one took a course too far offshore, the raft would get caught in the Humboldt Current and be carried 4,000 miles into the Central Pacific.  Then came the real challenge: to cross the Equator and the Gulf of Panama without getting stuck in the endless circulations of the Equatorial Counter Current, the same ocean current in which Vital Alsar’s balsa raft La Pacifica was stranded for over one hundred days in 1966-67.

If he reached the coast of Mexico, Haslett would complete a voyage that no other primitive replica raft had ever navigated.  Even so, with the fiftieth anniversary of the Kon-Tiki expedition only a few years away, Haslett did not feel that this would be accomplishment enough.  He felt that if he reached Mexico he could then launch a personal adventure: to reach Hawai’i by using the prevailing winds and currents that would then be at his back.  If successful, the whole dual experiment would offer both an experimentally demonstrated mechanism for contact between Central and South America, and offer a speculative route to the Hawai’ian islands for prehistoric mariners from the Americas.  While the latter voyage was considered largely outside the scope of prehistoric possibility, the former was of primary interest to scholars examining maritime interactions between traditional cultures of Central and South America.



Section 2: The Illa-Tiki expedion


Haslett set out for Ecuador in January of 1995.  There he harvested nine huge balsa logs, and had them shipped by flat bed truck to the coastal village of Salango.  The twenty-ton raft, an almost precise copy of Heyerdahl’s Kon-Tiki, was launched on March 23rd, 1995.  With Heyerdahl’s permission, Haslett christened his balsa raft Illa Tiki, the fire god.  Thousands of local villagers turned out for the launch, and dozens of them insisted on climbing on board the raft as it took to the sea.  Haslett feared disaster as the first wave rolled into the raft, but it merely lifted the balsa logs and drenched those on board.

            Thirty-eight days later, from the top of the thirty-five foot mast, Haslett sighted Panama.  In his first month at sea he had learned volumes about the coastal sailing abilities of a large balsa freighter.  With the hope that he could modify the raft and improve it’s sailing characteristics, Haslett decided to anchor in Panama for a time. 

It was while anchored in Panama that Haslett discovered that the nine balsa logs that made up the Illa Tiki were hopelessly infested with the wood boring teredo navalis.  The raft had been in salt water for more than eighty days.  The buoyancy of the raft had been so compromised that only modern pesticides would stop the attack.  Haslett knew he had to reject such a course, for it would mean the end of his archaeological experiment and the beginning of a mere adventure, as well as a severe environmental hazard for the harbor in which the raft was anchored.

Crushed and dispirited, Haslett had no choice but to beach the Illa Tiki permanently on the shores of the Bay of Ciruelos in Panama.  He realized that he had fallen into a familiar trap of mistaking adventure for science.  If he was to make a real contribution to understanding prehistoric navigation, he would have to start all over.  “On the Illa Tiki expedition we backed into the science,” said Haslett, “because we started out merely to copy the adventure of Kon-Tiki.


“The voyage to Hawai’i had been my primary goal.  Now I realized that if we were serious about testing a prehistoric balsa raft we were going to have to sail it from point A to point B and test a scientifically-grounded hypothesis.  What we learned from Illa Tiki was that everything we knew about balsa rafts was wrong” (Haslett, personal communication, August 27, 2000).


Section 3: The La Manteña expeditions


1: New expedition; same teredo navalis


            By the fall of 1996, John Haslett had regrouped from the destruction of the Illa Tiki balsa raft off the coast of Panama a year earlier.  With the fiftieth anniversary of the Kon-Tiki expedition only months away, Haslett wanted to both commemorate Heyerdahl’s voyage and continue his own research in prehistoric Manteño coastal navigation by organizing his second balsa raft expedition from Ecuador.  With good luck, his second raft would be ready to sail from Salango, Ecuador in August, 1997, a half century to the day after Kon-Tiki landed at Raroia. 

            To honor the ancient mariners whose technology he was recreating, Haslett decided to name his raft La Manteña-Huancavillca after both the Manteño culture and its immediate neighbor to the south.  A local Salango shipwright, Maestro Enrique Guillen, once again oversaw the construction of the raft.  And as with the Illa Tiki, nine balsa logs were lashed together with one and a quarter inch hemp rope.  The longest log, which the locals dubbed “The Pope,” was nearly sixty-one feet long.  On top of these logs Haslett laid eleven crossbeams of heavy cocobolo wood.  Forty, six-inch thick bamboo logs lashed to the cocobolo formed the deck, atop which was a bamboo cabin topped with a roof of thatch palm fronds. 

Haslett used another wood, guayacan, for the ten-foot long, sixteen inch wide guara centerboards, and two thirty foot cocobolo logs to form a forward and aft masts.  The fore and aft rigging was a radical departure from the square sail mounted on a bipod mast on the Illa Tiki.  It was another indication that the new balsa raft would be a sailing, not a drifting, vessel.  This meant new sails as well, and the expedition’s archaeologist, Cameron McPherson Smith, spent a week with a Guayaquil tent-maker supervising the sewing of two triangular lateen sails.  Gone was the familiar square sail carried by Kon-Tiki and so many other balsa rafts.  With this new design, Haslett believed he had produced the truest copy of a prehistoric Manteño balsa freighter ever built in modern times.

            To try and prevent La Manteña-Huancavillca’s destruction by teredo navalis, Haslett spent much of his time after Illa Tiki consulting with specialists on the marine mollusk.  On the first expedition, the balsa logs had been cut and then shipped by truck to Salango, unlike the spring of 1947 when Heyerdahl had floated his balsa logs down river from the interior.  William Willis and Vital Alsar had also floated their logs down the freshwater rivers of Ecuador, and Haslett speculated that immersion in fresh water might have offered a kind of protection to the logs once they reached the salt water home of the teredo.  Perhaps the logs of the first expedition, having missed any freshwater immersion, had not been ‘innoculated’ against the invertebrate. 

On this second expedition, Haslett planned to sail La Manteña-Huancavillca up into coastal estuaries and river mouths whenever possible.  In theory, such periodic natural shifts into freshwater environments would destroy any saltwater parasites clinging to the raft.  The river excursions would also test the inshore navigability of the raft, while putting the crew in contact with upriver tribes that might have been part of the prehistoric trade network of the Manteño.  Since Haslett also believed that another method for defeating the teredo could have been the periodic removal from the water and drying of the raft itself, the crew could also conduct surveys for potential prehistoric balsa raft landing sites. 

            Like Tim Severin, who experimented with natural lacquer as a way to guard his bamboo raft Hsu Fu against wood boring mollusks, Haslett also experimented with possible natural coatings to try to fend off the teredo.  There was tar, found in naturally occurring tar pits and seeps along the Pacific coast and a plausible prehistoric anti-fouling paint.  There was also the juice of the barbasco fruit, which was suggested to Haslett by local knowledge.  This juice, whose active ingredient is called rotenone, was used to stun and catch fish, and one suggested method was to hand bags of basbasco underneath the raft, creating clouds of rotenone to frighten away any lurking teredo. 

Experiments with submerging different blocks of balsa wood in seawater, however, showed that the juice had no effect on preventing a teredo infestation.  To compromise, and insure that this second balsa was not destroyed, Haslett coated the seven middle logs of the raft with antifoliant paint, and the two outboard logs with a mixture of 20% rotenone and 80% water.  If was difficult if not impossible to know which if any of these natural methods were known or used in prehistory.  The experiments provided at least some data on their effectiveness or lack thereof.

Haslett also planned two potential alterations from his itinerary on the Illa Tiki expedition.  If La Manteña-Huancavillca reached Mexico, Haslett considered returning to Ecuador, thereby demonstrating a plausible prehistoric trading round trip.  Or Haslett could construct two rafts: one to reach Mexico and continue on to Hawaii, and a second raft to sail from Panama to Ecuador to explore the return voyage from Central to South America.  In either case, the voyage to Hawaii was now relegated to a secondary adventure, without any theoretical baggage on board. 

Delays caused by funding and other problems—such as recruiting two separate raft crews—forced Haslett to scale back his second expedition to a single raft.  Completed in Salango in September, La Manteña-Huancavillca was sixty feet long and weighed more than twenty tons.  On September 28, 1997, the raft was pushed into the sea by a horde of locals assisted by a bulldozer.  For two weeks, La Manteña-Huancavillca lay at anchor in forty feet of water while Haslett and his crew provisioned the raft for the expedition.  It was two weeks Haslett was soon to regret.



2:  From Ecuador to Columbia


On October 12th, during the local annual Balsa Festival, as the crew prepared the raft to be towed to sea, more than seventy residents of Salango poured onto the raft and most of the day was involved in setting them ashore.  A fishing boat towed the raft two miles to sea and set it adrift.  The crew raised the new triangular sails and headed north.  Under overcast skies, Haslett tried to stay about thirty miles offshore as he navigated the raft along the coast.

Just twelve days after leaving Salango, Haslett sent two crewmembers overboard to check on the condition of the balsa logs.  He was stunned to discover that teredo had already infested La Manteña-Huancavillca.  The outboard rotenone-treated logs were severely infected, and the antifoliant paint on the main logs had flaked away in several places, allowing colonies of teredo to take hold of the raft.  A balsa raft seemingly possessed acres of surface area, and short of immersing the logs in a vat of DDT, there were bound to be unprotected spots where teredo could enter.  Once a single teredo bored into the wood, it multiplied and consumed the raft from the inside out.

The infestations forced Haslett to drive the raft ashore.  Adverse currents kept the raft away from Panama, so La Manteña-Huancavillca was steered toward Colombia, to a place called La Playa de la Muerte, the appropriately named beach of the dead.  Haslett was stranded, mystified as to why the relentless teredo seemed to single out his raft after a coastal voyage of 700 miles.  Thor Heyerdahl, Willis Willis, Vital Alsar, none seemed to have had to deal with the destruction of their balsa rafts by teredos.  Kon-Tiki had survived 4,300 miles of open-ocean and is now on permanent display in Oslo, Norway.  Willis’ Seven Little Sisters cross 6,800 miles of ocean and was brought ashore in Samoa and put on display.  La Balsa crossed the entire Pacific, more than 8,000 miles, and is now on display in Alsar’s hometown in Spain. 

As Haslett sat on the lonely Columbia beach, question after question filled his mind.  Was it the very coastal nature of the voyage of La Manteña-Huancavillca that led to its infestation?  Or the prolonged ride at anchor in Salango after the launch, a two-week period that could have allowed teredo navalis to get a grip on the raft?  How did the earlier balsa raft expeditions manage to avoid this menace?  What of the prehistoric mariners he was trying to study?  Was it possible that they never had to deal with the teredo?  Could it be that Europeans unknowingly introduced the teredo to the Pacific?  Burrowed into the wooden hulls of Spanish caravels, did this ‘alien species’ spread from Acapulco to the Philippines, from Jakarta to Hawaii, attached to the hulls of Dutch, British, or French exploration, trading, or military  fleets?  Was it something more systematic and cultural, like regular switching of balsa logs, or charring the logs to keep the voracious teredo at bay? 

Crewmembers began to leave the expedition and stray back to their homes, leaving Haslett with no good alternatives.  Refusing to give in, Haslett replaced the balsa logs, and soaked these logs with two coats of primer, three layers of natural brea, or tar, a coat of white cement, and two more coats of tar.  He replaced four departed crewmembers with two Columbians eager to join the adventure.  The new, smaller crew now spoke primarily in Spanish.  By early January 1998, the refitted La Manteña-Huancavillca II was ready to head to sea again, aiming for the west coast of Mexico.  Ahead lay a treacherous stretch of Columbian coastline frequented by pirates known to board boats, kill their crews, and make off with the cargo. 



3:  From Columbia to the gyre


Soon after the raft headed north from the beach in Columbia at about a half a knot, it was caught in a pattern of circular ocean currents, known as a ‘gyre,’ off the coast of Costa Rica.  La Manteña-Huancavillca II found itself unable to break out.  Irregular winds failed to fill the sails, and two near-hurricanes battered both the raft and its crew.  Like Vital Alsar in 1966-67, weeks passed with little progress.  Then, as the raft wound yet again around the gyre, Haslett discovered that his teredo nemesis had reappeared through seven layers of tar.  As the raft slowly began to sink, the crew retrieved a balsa log, of all things, that they noticed floating by, and jammed it under the raft to increase its buoyancy. 

After fifty days of aimless drifting, the main deck was awash.  A brief east wind gave Haslett a temporary hope of maneuvering out of the gyre.  The wind also brought a heavy storm, washing equipment overboard and canting the mainmast.  When the storm subsided, Haslett cut away the bamboo cabin, and threw overboard anything that could be spared and was biodegradable.  With the raft lightened by nearly a ton, the crew had a notion to make a run for Hawaii.  But the gyre refused to give the raft up. 

With the raft steadily sinking after another two weeks in the gyre, with a total absence of wind to push the raft, Haslett had reached the end of his tether.  He had smoked cigars all his life, and one afternoon on the raft he could not understand why his face was surrounded by cigar smoke, something that had never happened before.  It suddenly struck him that the air was so perfectly still that no even cigar smoke dissipated.  He called a halt to the experiment.  The crew was rescued by a cutter of the Costa Rican Coast Guard, but not before the captain of the cutter ordered Haslett to chop the drifting hazard apart and let the balsa logs float away separately. 

Despite all his preparations and experiments, Haslett had been defeated a third time by a small marine invertebrate.  It could be argued that prehistoric mariners knew how to defeat these pests.  Perhaps they took a different route—further offshore to Cocos Island, for example, the island around which the dreaded gyre circulates—or sailed in a different season.  Or perhaps the ocean currents and wind patterns were simply different a thousand years ago.  But even more than the frustrating teredo infestations, the voyage of La Manteña-Huancavillca II had shown exactly how dangerous, lengthy, and potentially endless the trade route from Ecuador to Mexico could be.