In the summer of 1902, the U.S. Senate voted on the route by which an artificial river would join two oceans. Behind that vote lay bankruptcies, volcanoes, mosquitoes, and a handful of men who knew how to count earth in cubic yards. The fork was narrower than people tend to think.
By the beginning of the 20th century, the idea of an interoceanic canal through Central America was no longer a dream but an accounting problem. More than two routes were considered—across Tehuantepec in Mexico, through Darién on the border between Panama and Colombia—but by 1902 the real competition had narrowed to a pair: Panama and Nicaragua. The Panamanian isthmus was shorter—some 50 miles from ocean to ocean—and it already had a railroad, built back in the 1850s. Its chief advantage, though, was this: over twenty years, the French had removed millions of cubic yards of rock from Panamanian soil. That work had cost thousands of lives and ruined tens of thousands of shareholders, but it had been done. The Nicaraguan route looked different. It began at San Juan del Norte on the Caribbean coast, ran some 30 miles up the San Juan River, emerged into Lake Nicaragua—a freshwater sea 107 feet above sea level—and ended in a short artificial cut to the Pacific near Rivas. Its total overland length was around 100 miles, twice that of the Panamanian option, but vessels would have traveled a substantial part of the way on water that already existed.
The Walker Commission, which studied both routes twice, initially leaned toward Nicaragua. Politically, that option seemed cleaner: no need to buy the legacy of a French scandal, no need to negotiate with the New Panama Canal Company, which sat on the wreckage of de Lesseps’s empire and wanted $109 million for its assets—a figure Congress considered obscene. When the French cut the price to $40 million, the arithmetic changed. The Walker Commission revised its recommendations and in January 1902 came out in favor of Panama. But the Senate was not yet convinced.
The story of how Philippe Bunau-Varilla mailed Nicaraguan postage stamps bearing smoking Momotombo to senators is often told as the anecdote that decided the canal’s fate. That is an oversimplification, but not an invention. Bunau-Varilla, a French engineer who had worked personally on the Panamanian construction project and had ties to the New Company, did indeed use the stamps as a visual aid: there, you see, the Nicaraguans themselves admit their country is a volcanic zone. A few weeks before the Senate vote, on May 8, 1902, Mount Pelée erupted in Martinique, destroying the city of Saint-Pierre and killing around 30,000 people. It was not a Nicaraguan volcano, or even a Central American one, but it set nerves on edge. Bunau-Varilla was quick to remind them: the tectonic fault line where plates collide runs straight along Lake Nicaragua and Managua. Earthquakes in the region were documented—1844 was one of the earliest known episodes, while destructive repeats in 1931 and 1972 later confirmed that the fears had not been empty.
But to reduce the decision to stamps is to ignore everything else. The Panama lobby worked on several fronts: lowering the price of the French assets, engineering arguments about the route’s length, calculations of the cost of dredging the San Juan River. The Nicaragua lobby—and it existed too, backed by the influential Alabama senator John Tyler Morgan—pressed the political cleanliness of the deal and the route’s proximity to the ports of the American South. On June 19, 1902, the Senate passed the Spooner Act, authorizing construction through Panama but—and this is crucial—providing for Nicaragua as a fallback route if the Panama deal fell through. The act was not a bet on one horse, but an insurance policy with two tabs.
From there, events moved down a corridor the Spooner Act had cracked open but not predetermined. The United States signed the Hay–Herrán Treaty with Colombia, granting Washington rights to build in Panama, then a Colombian province. In January 1903, the Colombian Senate refused to ratify it. The reasons were clear: Colombia found the terms humiliating and wanted more money. Formally, that was the moment when the Nicaraguan fallback plan should have kicked in. But instead, in November 1903, Panama declared independence from Colombia—with support that can hardly be called covert: the American cruiser Nashville stood offshore, preventing Colombian troops from landing.
Two weeks after the revolution, Bunau-Varilla—now serving as envoy for the newborn republic—signed a treaty with Secretary of State Hay, transferring control of the Canal Zone to the United States. On February 23, 1904, Panama received $10 million and the promise of an annual annuity. On May 4, 1904, the United States paid $40 million for the French property—rails, excavations, equipment, hospitals, and, essentially, the right not to begin from scratch. The Nicaraguan option was not rejected for engineering reasons—it was bypassed politically. The question I cannot shake: what if the Colombian Senate had said yes? Or, more interesting still—what if the Roosevelt administration had not dared support the separatists and had turned to the fallback plan?
The 1907 design, drafted after construction in Panama had already begun, gives us a detailed picture of what a Nicaraguan canal might have been. The route called for roughly three locks on the Caribbean side, lifting ships to lake level, transit across Lake Nicaragua, then another roughly three locks on the Pacific descent near Rivas. The maximum height of terrain that had to be overcome reached 200 feet—almost twice the height of the watershed on the Panamanian route. An alternative route through the Punta Gorda River made it possible to bypass part of the San Juan River, but required additional earthworks.
The San Juan River is not a navigable artery but a temperamental tropical stream. Its channel grew shallow in the dry season and flooded in the wet; controlling its flow demanded dams, embankments, and constant dredging. Lake Nicaragua itself was an asset—a gigantic natural reservoir—but also a problem: its water level fluctuated, and its southwestern shores approached the volcanic chain. Put crudely, the builders would have been handed a canal whose blueprint nature was constantly trying to rewrite.
There was another factor people spoke little of aloud in 1902: the San Juan River formed the border between Nicaragua and Costa Rica. A shifting channel—and tropical rivers shift their channels regularly—turned an engineering task into a diplomatic one. The territorial dispute between the two countries had dragged on for decades and remains unresolved to this day. Building a canal along a disputed border is like running a water pipe through the wall between neighbors who do not speak to each other: technically possible, legally a minefield.
Ferdinand de Lesseps, the conqueror of Suez, arrived in Panama convinced that Central America could be cut through as Sinai had been—with a sea-level canal, no locks required. The Isthmus of Suez was flat, dry, and made of sand. Panama was wet, mountainous, and composed of clay that slid. The difference between the two isthmuses is the difference between digging a trench in a sandbox and tunneling through a soggy cake. The Universal Interoceanic Canal Company began work in 1881, and by 1889 it had gone bankrupt. By various estimates, between 20,000 and 22,000 workers died—mostly of yellow fever and malaria. The financial scandal that followed the bankruptcy shook the Third Republic: bribed deputies, ruined small investors, court cases.
When the Americans began construction in 1904, they inherited the Culebra Cut—later the Gaillard Cut—the railroad, and an experience that could be studied in reports. But they did not immediately absorb the central lesson. In 1905–1906, a fierce dispute raged within the American engineering leadership: should they build a sea-level canal, as de Lesseps had dreamed, or use locks? President Roosevelt convened an international board of consultants, and most of its European members came out in favor of a sea-level canal. American engineers, who knew the terrain better, insisted on a lock system. The final decision in favor of locks was approved only in 1906—two years after work began.
The lock design turned the Chagres River from an enemy into an ally. Instead of fighting its floods, engineers dammed it and created Gatun Lake—at the time, the largest artificial lake in the world. Ships rose through three flights of locks, traveled 24 miles across the lake, passed through the Gaillard Cut, and descended to the Pacific. The volume of earthworks was radically reduced compared with what a sea-level canal would have required. This was not surrender before nature, but a translation of the problem into another language: instead of “remove the mountain”—“raise the water.”
William Gorgas arrived in the Canal Zone in 1904 with experience fighting yellow fever in Havana. His first two years were a battle not so much with mosquitoes as with his own superiors: chief engineer John Wallace and his successor John Stevens initially considered sanitation spending excessive. Gorgas insisted on accepting the mosquito theory—the idea that yellow fever and malaria were carried by mosquitoes of the Aedes and Anopheles genera, respectively. When he was given the resources, results came quickly: draining puddles and swamps, treating standing water with oil, installing window screens, building water mains and sewers, creating a hospital network. By 1906, yellow fever in the Canal Zone had been virtually eliminated. Malaria declined more slowly, but steadily.
Transferring that experience to the Nicaraguan route would have been harder than it seems. Panama’s Canal Zone was a compact strip that could be controlled like a military camp. The Nicaraguan route stretched for a hundred miles, much of it running along a river and lake shores—that is, along ideal mosquito breeding grounds. Draining puddles around a construction site ten miles across is one thing. Draining the shores of a lake with an area of 8,264 square kilometers is a different order of task. This arithmetic of scale fascinates me: the same methods, the same science, but geography turns the solution into a problem.
There was a historical precedent, too. In 1780, a British expedition commanded by the young Horatio Nelson traveled up the San Juan River to seize a Spanish fort. Of its 1,800 men, the expedition lost 1,500—almost all to tropical disease. Nelson survived, but was evacuated half-dead. One hundred and twenty years later, medicine knew things about mosquitoes that Nelson did not, but the hydrology of the San Juan River had not changed: it was still a warm, slow river, marshy in its lower reaches, ideal for Anopheles.
Suppose the Colombian Senate said no—which it did—but the Roosevelt administration did not support the Panamanian separatists and turned to the Spooner Act’s fallback plan. What then?
The first question is timing. The Panama Canal took ten years to build (from 1904 to 1914) with the French head start already in place. The Nicaraguan route began with a blank slate: no railroad, no cuts, no port infrastructure. The canal was twice as long. Dredging the San Juan River was work that had to be repeated after every rainy season. Under the most optimistic scenario, construction would have taken 15–20 years, meaning the canal would not have opened before the 1920s. That means: during the First World War, the United States has no interoceanic canal. The Pacific Fleet cannot quickly move into the Atlantic. Military logistics are rearranged.
The second question is cost. The Panama Canal cost roughly $375 million (in dollars of the time), including the purchase of French assets. The Nicaraguan project, according to Walker Commission estimates, would have cost as much or more—while the $40 million saved on French property would have been swallowed by the scale of the new earthworks and the need to build all infrastructure from scratch.
The third is geopolitics. The Canal Zone in Panama became an American enclave that existed until 1999. A Nicaraguan Canal Zone would have created a similar enclave, but in a country bordering Costa Rica, which disputed ownership of the San Juan River. The United States would have found itself the arbiter of someone else’s territorial dispute—and, given the history of American interventions in Nicaragua in the 20th century (the Marines were in the country, with interruptions, from 1912 to 1933), one can assume the presence would have been even heavier.
The fourth is Panama’s fate. Without the canal, Panama probably remains a province of Colombia. No revolution in 1903, no Canal Zone, no annuity, no Panama City as an international financial center. The railroad across the isthmus continues to operate, but the isthmus remains a periphery—not a hub of global trade. Nicaragua, meanwhile, receives what Panama received: money, dependency, military presence, and a society split between those who live off the canal and those whom it deprived of land.
Alexander von Humboldt, who explored Central America in 1799–1800, considered a Nicaraguan canal feasible. Half a century later, Cornelius Vanderbilt built a transit route through Nicaragua for gold seekers rushing to California, and he too thought about a canal—but ran up against a shortage of machinery, volcanoes, earthquakes, and yellow fever. Another half-century on, American engineers had calculated the route down to the last lock. And still they chose Panama.
The Panama Canal was not the triumph of inevitability, but the outcome of a chain in which every link could have come apart. If the French had not gone bankrupt, there would have been no assets to sell. If the New Company had not lowered its price from $109 million to $40 million, the Walker Commission would not have changed its recommendation. If Mount Pelée had not erupted six weeks before the vote, the volcanic argument would have sounded quieter. If Colombia had ratified the treaty, there would have been no need for a revolution. If Roosevelt had not sent the cruiser, the revolution might have failed.
The canal, opened on August 15, 1914—two weeks before the Battle of the Marne began—became one of the greatest engineering works in history. But it might have run through different land, across another lake, past other volcanoes. The Nicaraguan project never vanished completely: it resurfaced in the 1910s, the 1930s, and the 2010s—most recently as a Chinese concession in 2013, which never came to fruition. The San Juan River still flows toward the Caribbean Sea, Lake Nicaragua still lies 107 feet above sea level, and the tectonic plates beneath it still have not come to an agreement.