The field was still open. Aside from Nicaragua and Panama, there had been surveying expeditions in Mexico, where, as elsewhere, concessions were granted, then resold, but no work was ever started. The San Blas route, where the Isthmus spans only thirty miles, had been visited by British and American explorers, who had expressed optimism about a “very remarkable depression” in the area's 1,000- to 1,500-foot mountainous spine, not that they had actually reached it. The region that had attracted the most excitement, however, was Darién, the section of the Isthmus spanning the present-day Panama-Colombia border, still inhospitable jungle today.
In 1850, an Irish doctor, Edward Cullen, had announced in England that he had found a short and convenient canal route from Caledonia Bay to the Gulf of San Miguel, with a break in the Continental Divide at only 150 feet above sea level. He also claimed that the previous year he had painlessly walked the route from coast to coast in a few hours. It seemed a sensational discovery. But subsequent investigations by an internationally manned surveying team had found no pass across the Continental Divide at this point at less than 1,300 feet above sea level. During the course of the exploration, several teams got lost and were either ambushed by hostile Cuna Indians or died of starvation or disease.
“It is proved beyond all doubt,” the leader of the surveyors had reported, “that Dr. Cullen never was in the interior, and that his statements are a plausible net-work of fabrications.” The Irish doctor wisely disappeared, though he emerged later as an army medic in the Crimea. It seems that his entire story was a gigantic lie, a dream built on wishful thinking, which had cost the lives of more than a dozen men.
But many were undeterred and optimistic stories still circulated. South of Caledonia Bay, a broad river, the Atrato, penetrates deep inland, offering a canal excavation of as little as twenty-eight miles. For this reason several French and British expeditions had been organized and sent to the area. Soon after Cullen's sensational claim, Wall Street millionaire Frederick M. Kelley heard a rumor of a “Lost Canal of Raspadura” linking deep inland, tributaries of the Atrato and the Tuyra rivers flowing to opposite oceans. In 1852 Kelley funded an expedition to investigate, led by John C. Trautwine, one of the experienced Panama Railroad engineers. His report was emphatic—there was no “Lost Canal”—but Kelley, described in his later years as “mystical and imaginative,” was clearly a man in the grip of an idea. As well as corresponding with and meeting the by now ancient Hum-boldt, he toured the capitals of Europe looking for support for a canal, and over the next twenty years spent an estimated $150,000 of his own money on further optimistic surveys in Darién, few of which even managed to venture far inland.
Rear Admiral Charles H. Davis, in his 1866 report for the U.S. Senate, had also favored Darién, in particular the more westerly route between Paterson's Caledonia Bay to the Gulf of San Miguel. On the map it looked simple: the narrow barrier between the oceans, the probable route taken by Balboa, is only forty miles, and large rivers promised to reduce further the necessary length of a canal there.
So the destination was Caledonia Bay when the first Grant expedition to leave the United States for the Isthmus headed out from Brooklyn Navy Yard on January 22, 1870, commanded by thirty-three-year-old Lieutenant Thomas O. Selfridge. “The Department has entrusted to you a duty connected with the greatest enterprise of the present age,” read his orders from Admiral Ammen. The weather was clear and bright.
For two months his hundred-man team, heavily armed to deter Indian attacks, searched the high mountain range near the Atlantic coast, but found no sign of Cullen's pass, and their thoroughness, something of a new departure for Darién surveys, left no room for doubt. From there Selfridge moved on to San Blas. There, like previous expeditions, they were turned away by Cuna Indians before completing an overland crossing of the Isthmus, but they saw enough to bury the hope of a “remarkable depression” and quickly to discard a sea-level or lock canal. Even if the long string of locks needed to lift ships to over a thousand feet could be built, every lock canal needs a supply of water at its summit elevation. Each time a lock is used, thousands of gallons of water pass “downstream.” At the top of the mountains of San Blas there were only trickling brooks. The only option was a tunnel, as others had concluded, and this time the estimate was at ten miles long. Nevertheless, because it was still the shortest route, Selfridge refused to rule it out as he led his exhausted team back to New York.
He was back in Darién at the end of the year, still chasing shadows. On December 29, 1870, he sailed to the deep, beautiful Gulf of Urabá into which flows the Atrato River. This was where Balboa had founded the now abandoned town of Santa María more than three hundred years before and where Kelley's explorers had hoped to find the “Lost Canal.”
Selfridge himself was now succumbing to the wishful thinking, or “excessive optimism,” that is a recurring part of the canal story. He was much taken by the Atrato basin and allowed himself to imagine that “it [would] one day be covered with sails from every clime.” Self-ridge had been joined by fellow naval officer Edward P. Lull, who carried out a hydrographic survey of the river's gulf, while Selfridge himself headed upstream. He eventually came up with a plan that involved more than twenty locks and a tunnel five miles long. Tunnels were particularly problematic, even without geological considerations. They had to be high enough for the tallest sail-bearing masts, but there was also the problem presented by the accumulation of steam from engine-powered vessels. Nevertheless, Selfridge was keen, and further expeditions were sent to take another look.
Meanwhile, separate expeditions had set off to investigate the other possibilities suggested by Humboldt, Davis, or both. On November 11, 1870, a party of surveyors landed at Tehuantepec, the narrowest point in Mexico, where the Isthmus is 130 miles wide, with a low range of mountains. Although it was the possible canal site nearest to the United States, the surveyors found little else to recommend it. The lowest pass was over 750 feet above sea level, and the only possible canal would be 144 miles long and need an impossible 140 locks.
There were two Grant expeditions to Nicaragua. The first got off to a calamitous start, when the commander of the expedition was drowned while attempting to land. But a second expedition headed by Selfridge's previous companion Edward Lull and Cuban-born Ancieto Menocal (who had worked on the Vanderbilt survey back in 1849) conducted a thorough survey, running a line of levels from coast to coast, measuring river flows and depths, and preparing tables, maps, and charts. They confirmed the earlier survey's findings that there existed a pass only 153 feet above sea level in the narrow neck of land between Lake Nicaragua and the Pacific. This was more than 100 feet lower than anything found at Panama. Lull and Menocal suggested a canal with ten locks on either side, estimated to cost nearly $65 million.
It was not until the beginning of January 1875, after a further expedition had explored other routes around the Atrato River, that the Panama survey team left New York for Colón. From the surveys conducted for the railway, the Panama route had been the best known at the outset, so it was left to last. The expedition was led by Lull and Menocal, and for two and a half months some one hundred men explored and measured the land between Colón and Panama and up into the Chagres watershed. Nevertheless, both Lull and Menocal had pretty much decided on their preference for the Nicaragua route, and they were unable to imagine a realistic design for a canal at Panama. The line of their proposed waterway largely followed that of the present-day canal. It was to be a lock canal, with a summit level of 124 feet, reached by twelve locks at either side. The problem with Panama, they correctly judged, was the volatility of the Chagres River. They suggested that what was needed was a huge, 2,000-foot-long viaduct sufficiently high to allow the floodwaters to pass under it. But the river was required as well, to supply water to the summit level, so the engineers decided that ten miles of aqueducts would have to be built. In all, it was just too impractical and expensive.
When the Inter-Oceanic Canal Commission reported in February 1876, there was a firm majority in favor of a route through Nicaragua. Tehuantepec was dismissed out of hand. Darién was too remote and wet; and the various Atrato schemes were too ambitious. All myths about “lost canals” and low passes in those areas had been exposed. Panama was too expensive, and what's more, “the deep cut would probably be subject to land-slides, from which the Panama Railroad has suffered seriously, and the canal would be exposed to serious injury from flood.”
From then on, for the next twenty-five years, U.S. canal policy was firmly focused on Nicaragua, which became lodged in the minds of the public as well, who were beginning to demand an American canal under American control rather than the neutral and open-to-all arrangement favored by the Grant Commission. In 1877, treaty negotiations were started with the Nicaraguan government. In contrast, the United States did not even have a diplomatic representative in Bogotá. Slowly, draft treaties were drawn up, while private companies jostled for the concession, and Ancieto Menocal prepared another, even more thorough, Nicaragua-bound surveying party.
Then, in May 1879, the United States public read in their papers news from Paris that upset all previous plans. Le Congrès International d'Etudes du Canal Interocéanique had approved the digging of a canal at Panama and events were moving fast. It seemed that there would be a canal, but it was going to be built by the French.