LOOKING UP AT THE FAT, GRAY SKY, it was hard to tell when, or even if, dusk had arrived. Looking down, it was easy. One minute the water beneath our paddles was the color of tea; the next, it was black. We pushed on around a few more bends, then beached the canoes on a riverbank at a clearing in the forest.

Only Morgan Gnoundou—Mor-GAHN, he pronounced it, à la française—had any energy left. He cleared space for the tents with his machete, got a fire going, put a pot on to boil, and then bounded down to the river to dredge for crevettes—little shrimp with which to bait his fishing line. Sprawled on our Therm-a-Rests, the rest of us watched in awe.

It was early August, dry season in Gabon. That morning, eight of us, in four canoes, had set out from an abandoned logging camp on the upper Djidji River not far from the Congo border. The plan was to paddle downstream 100 or so miles to a take-out just above a spectacular cataract called Djidji Falls. In the process, we'd traverse the entire roadless expanse of 1,158-square-mile Ivindo National Park. Like all 13 of Gabon's national parks, it was created ex nihilo just over four years ago, largely at the urging of the Wildlife Conservation Society, the international nonprofit headquartered at the Bronx Zoo. No one, to our knowledge, had ever paddled the length of the Djidji before, but our real mission was to evaluate the river's touristic potential—something the WCS program director for Gabon, English-born biologist Lee White, was banking on.

"It's not a whitewater river; it's a wildlife river," White had told me a few days earlier, at his office in Gabon's capital, Libreville. "And I think it could be the premier wildlife river in equatorial Africa."

Sitting by the fire that first evening, I had my doubts. We'd seen animals, all right: monitor lizards, unidentifiable monkeys, and a large red forest antelope called a sitatunga, head up and stock still, as if posing in a diorama at the American Museum of Natural History. And we'd seen fantastic birds: sapphire-blue kingfishers, dinosaurish hornbills, flocks of gray parrots, and a huge, honking, iridescent thing called a hadada ibis.

But, truth be told, what we'd mostly seen were trees—fallen trees. Every 50 yards, it seemed, another giant mossy trunk lay across the river, blocking our path. Sometimes there was nothing to do but lift and shove our vessels straight over the top, Fitzcarraldo style. With two of the canoes, slender plastic things improbably mail-ordered from L.L. Bean, that was at least a semifeasible proposition. But the other two were big aluminum johnboats with square transoms—wider, more heavily laden, and about as portable as cast-iron bathtubs. Sometimes it took five or six of us, balancing precariously on branches and slippery trunks, to heave them over. We took to calling them "les vaches de mer"—the sea cows.

Then there were the snakes. The day before, while on a short pre-trip reconnaissance cruise, Gnoundou and Christian Mbina, a Gabonese environmentalist, had been hacking their way through a brambly snag a few hundred yards from the put-in when a fat black-and-tan snake dropped into the canoe—whereupon Mbina leaped out. Malcolm Starkey, a WCS project director for two of Gabon's parks, and our designated naturalist, was too busy laughing to identify the beast. "But," he said cheerily, "I'm 90 percent certain it wasn't venomous."

Less than reassured, we'd approached our first log crossings with a certain delicacy. But as the day wore on, it got harder and harder to worry. Ninety percent was good enough—the main thing was to keep the boats moving, a point that was driven home the first night as we sat sipping Scotch by the fire. "So, how far did we get?" our nominal leader, Bryan Curran, the director of projects for WCS Gabon, asked our navigator, Mark Boyer, a map specialist from the WCS office in New York. Boyer punched a few buttons on his GPS and raised his eyebrows dramatically. "It looks like we've covered a grand total of 4.8 kilometers," he said. "Of course, that's as the crow flies; we probably did twice that over the ground."

A stony silence ensued. We'd planned on spending a week on the river—which meant making 20 to 25 kilometers (12 to 15 miles) a day. Finally, Starkey spoke. "Hmmm," he said, trying to sound chipper. "Where are we on the map?" Before the trip, Boyer had laboriously created and laminated a set of custom maps by combining some old topos from French colonial days with more recent data gathered by a NASA shuttle mission. He pulled the first sheet out of his daypack, squinted at his GPS again, and laughed. "We're not yet on it," he said.