It had been the “monstrous experiment” of the plan for the Gatún Dam that had attracted most criticism in the United States. In fact, there was nothing unprecedented about the dam apart from its scale. More justified were worries about its siting—over two deep geological gorges filled with dubious alluvial material. The plan was to block the Chagres valley at Gatún with an enormous, essentially earth structure one and a half miles long. The height was adjusted in the course of the building, but ended up being 115 feet above sea level, 30 feet above the planned level of the lake at 85 feet above sea level. At its top it was to be 100 feet wide, then 500 feet at the lake's surface and almost half a mile wide at the bottom. Between the two gorges across the dam site was a hill of solid rock, which would serve to anchor the structure and to provide the site for the spillway, through which the waters of the Chagres would be funneled and thence flow to the sea at San Lorenzo. Consisting of a convex wall 800 feet long and nearly 100 feet high, the spillway was to be the new jailer of the formidable Chagres River. For the planned locks to work properly, it was important that the level of the new river should not deviate from its 85-foot level by more than 2 feet. Within the wall of the spillway were a series of openings with massive steel gates. When all the gates were lifted, the spillway could cope with a discharge of 182,000 feet per second, judged to be the worst that the Chagres could throw at it. Next to the spillway a hydroelectric station was planned, to generate enough electrical current from the falling water to supply all the canal's energy needs. The triple-tiered double locks would be built at the dam's eastern end.
The dam itself was well suited to the height and length requirements of the chosen site, but more important it was to be constructed largely of the materials excavated from elsewhere along the canal line. A dam in concrete would have been equally feasible, but a great deal more expensive. Earth dams are not complex structures and are sometimes also referred to as gravity dams, because that is exactly how they perform; the resultant lake is held back purely by the huge mass of the dam and thus the friction at the base between the dam structure and the existing ground on which it is founded. The principle of earth dams is based on the fact that most clays are impervious. To construct such a dam, one builds in layers. First, outer walls of hard rock are laid down on both the upstream and downstream faces. This is to hold the earth and clay filler in the construction phase, but also to protect against scour in the permanent case (from the lake on the upstream face and storm water on the downstream). The next layer inside both faces is then normally constructed of soil and smaller rocks/stones, and then the central core is composed exclusively of appropriate clay. The clay can either be placed “dry” and compacted using rollers, or liquefied and pumped in. As the slurry drains and dries it will harden and become impervious.
At first all went well. To prevent organic matter forming a potentially porous layer under the dam, the six-hundred-acre site had to be meticulously cleared of its rich cover of tropical jungle. This task was completed by the end of 1907, and the dam site was barren. It reminded one visitor of the scene of a battle in some unimaginable war to come: “an ugly denuded waste of land … stubble was everywhere, and standing out like pockmarks were hundreds of black ash heaps where the greenery had been burned. Across this soggy wasteland … a dredge in the Chagres sucked mud from one place and vomited it into another; and dynamite crews sent up enormous geysers of rock and water. Men in gangs of forty to a hundred swarmed about the valley, all in the blue shirts and khaki trousers of the Zone Commissary, while the air was filled with the babel of more than twenty languages.” As well as the dam site, it was necessary to clear all the large trees from the future route of the shipping lane through Lake Gatún. Barbadian Edgar Simmons described this process: up to fifteen holes were cut in the trunks of the trees, which were then stuffed with explosives: “Three sticks of dynamite, with a cap and coil, about 18 inches long, and covered with mud. So all are set for evening,” he wrote. “After the 5:15 passenger train pass for Panama, we start lighting. Some of us has up to 65 or 72 holes to light… Nine of us start out, each one with two sticks of fire in our hand, running and lighting, at the same time trying to clear ourselves before the first set begin bursting on us. Then it's like Hell… it was something to watch seeing the pieces of trees flying in the air.”
In 1907 the Chagres flowed through four routes to the sea: its old riverbed, the French canal, and the two diversion channels the French had built either side of their waterway. All four flowed through the site of the new dam. The Western Diversion was the lowest and this was widened and deepened, while the other three channels were dammed without any complications. The plan was to start work on the eastern end of the dam and the spillway, and then close the West Diversion when it was time for the river to be diverted through the spillway.
Shovels started cutting a channel through Spillway Hill while suction dredges thoroughly cleaned out the old river and canal beds. Then two parallel lines of trestles were erected at 30 feet above sea level along the upriver and downriver faces of the dam some 4,000 feet apart. On these were laid rail tracks to carry cars of rocks to create the two stone “toes” that would hold the innards of the dam in place. Locomotives from Culebra or the Mindi Hills started arriving regularly and slowly the rock walls began to crawl across the wide valley. Meanwhile a suction dredge began pumping in between the walls a mixture of clay and water sucked up from the bed of the old French canal behind the dam site. Toward the end of 1908 a solid wall of rock, 60 feet high, had been laid out along the extremities of the dam.
Back in the States, the doubters were still vociferous, and now Bunau-Varilla had added his comments: water would seep through, emptying the lake, he said; the dam should have been located at Bohío (as per the French lock plan), where there was better bedrock; Gatún Dam would develop fissures and would be damaged by even a light earthquake. Then, at end of November came the news that everybody had been dreading. “Collapse of the Gatún Dam” read one headline in the United States. “Chagres River plunges through Gap in Isthmian Wall.” It turned out that an American journalist in Panama had noticed that a large section of the southern “toe,” some 200 feet, had slipped by 20 feet where it crossed the line of the old French canal. Goethals denied any serious problem and was backed up by John Stevens, who wrote an article denouncing “the outbreak of yellow fever journalism in regard to the Gatún Dam.”
But with their predictions of doom seemingly confirmed, the press kept the story running. The New York Times, a longtime enemy of Roosevelt and his Panama route, declared that the canal had to be begun anew at Nicaragua. Taft was quickly dispatched to Panama with some experts in tow to find out the truth about the dam. He confirmed Goethals's line that the slip was not serious, and rounded on critics of the great project, accusing them of creating a “fire in the rear… calculated to break down the nervous system of those persons on the Isthmus working day and night, tooth and toenail to build the greatest enterprise of two centuries.”
The damage to the “toe” was repaired, and early in 1909 the first concrete was poured at the spillway site. Throughout the year, the massive dam structure began to take shape, rising to block the valley. By the end of 1910 four dredges and more than ten locomotives were at work, adding between them over a million cubic yards a month to the mass of the dam. As the new chasm was being created in the Cut, much of the spoil removed ended up as the new mountain at Gatún. The peak year of building was 1911, when there were two thousand men at work and over a hundred trainloads of rock and earth were being dumped every day. There were further alarms: in October 1911, 1,000 feet of the dam's eastern end settled a few feet, and in August the following year about 800 feet of the crest of the dam dropped 20 feet. On other occasions, as with the walls of the Cut, the dam writhed and twisted as it sought a new equilibrium, wrecking track and threatening the lives of the locomotive drivers. But these were largely movements within the fill, rather than a slip of the entire structure, and as the clay center of the dam dried out and hardened, so the period of greatest danger passed.
With the Panama Railroad relocated on higher ground, it was time to start filling the new lake. At the end of April, preparations were complete for the closing off of the West Diversion. The engineers knew that the final taming of the Chagres River would not be easy. For one thing, delays meant that the river level was 7 feet above its lowest dry-season level, and more rains were on the way. Furthermore, the West Diversion ran directly over a deep gorge, with mud to great depths. It was also about 20 feet lower than the bottom of the spillway, through which the river was now to be directed. This meant that the river would need to rise by this height before it would begin to find the new outlet.
Huge trestles were driven into the mud of the channel, and hundreds of flatcars, loaded with rock from Culebra, were readied nearby. The plan was simply to unload the rock into the current faster than it could be washed away.
Starting at either side, rocks were dumped from the trestles around the clock for four days. Initially, all was fine, but as the channel contracted, the flow picked up power and pace. When it had shrunk to 80 feet wide and 6 feet deep, as fast as the spoil was dumped it was washed away. Even rocks weighing a ton were lifted up and carried off by the powerful Chagres current. A great mass of old twisted French rails was brought up to the site and thrown off the trestle into the open section. As predicted, the rails snarled on the trestle piles, forming a web that started to catch and hold the bigger rocks. But this also transferred huge pressure onto the trestle itself, which groaned and trembled, then broke, with parts starting to move downstream. The water rushed through again, and parts of the surviving dam slid and collapsed.
But the battle was not lost. Men clambered out onto the trestles to repair them, more cars were brought up frantically to dump rock, and a dredge started pumping clay upstream of the dam breach to reduce the force on the trestles. Finally the dam held, and as the water at last found the outlet of the heavily regulated spillway, the destructive force of the Chagres, the bane of the French effort, was gone forever.
With the West Diversion closed, the water was allowed to back up and the lake began to rise at 2 inches a day. Eventually the water would cover 164 square miles of jungle, as well as several villages and much of the digging of the French. Some villagers refused to move out and had to be forcibly evicted. For many of them the drowning of the Chagres valley was like a biblical flood. A way of life based on the river, which long predated Columbus's first voyage, was coming to an end. “It is not hard to realize why the bush native does not love the American,” wrote Harry Franck, who took a trip out on the growing lake in a police launch in mid-1912. “Suppose a throng of unsympathetic foreigners suddenly appeared resolved to turn all the world you knew into a lake, just because that absurd outside world wanted to float steamers you never knew the use of, from somewhere you never heard of, to somewhere you did not know.”
he first task in constructing the locks had been to excavate to bedrock the basins in which they were to be built. For Gatún, this meant digging a hole a mile and a quarter long, 200 yards wide, and 50 feet deep—the removal of nearly 7 million cubic yards of rock and silt. Most difficult was the lowest of the three basins, whose bottom was 66 feet below sea level, and whose sides constantly slipped. As the engineer in charge wrote, “No one expected on returning to work in the morning to find things as they were left the evening before.” Even though there were temporary dams at both ends of the lock basins, there was a constant danger of seepage and floods. American Harry Cole worked on the locks on the Pacific side—one pair at Pedro Miguel, two at Miraflores. In the lowest basin on the Mi-raflores locks—”an extremely dangerous place to work”—he experienced the same problems. “Sometimes, in the rainy season,” he wrote, “even the small rivers became large rivers and often overflowed and inundated much of the construction work, putting even our drainage pumps out of action. Canal banks would break loose, cover up lock wall foundations, fill up culverts, submerge railroad service tracks and cause weeks of delay to clean up and put the work in order again. Those were our heart-breaking times.”
As excavation continued on the lower basins, in late 1909 actual construction started on the completed upper basin of the Gatún Locks. The lock component of the 1906 plan had attracted almost as much criticism as the dam. Certainly, the locks were a gamble. Nothing of the sort even remotely as large had been built before, and the choice of concrete as the primary building material, mainly to save time, was even more daring. Today, concrete technology is an industry in its own right, as it can be a more difficult material to work with than it first seems. In the early 1900s little was known about the properties of concrete and the basic materials (sand, gravel, cement, and water) were simply thrown together in a fairly fixed proportion, mixed, and poured. That the locks at the Panama Canal are still working nearly a hundred years later without any major problem in the concrete is testament to the standard of design and construction employed.
The structural design of the locks was unremarkable. They were to be constructed of “mass concrete,” that is, there was to be little use of steel reinforcement bars. This type of structure, like the Gatún Dam, relies mainly on the enormous mass of material present to withstand the loads imposed upon it. In fact, the sidewalls of the locks are structurally similar to miniature dams. The massive width of the bottom of the wall, some 50 feet, creates an equally large frictional resistance to the force of the water pushing against it when the lock is full. The wall becomes thinner and thinner as the height increases and the applied hydraulic force reduces.
The key to building a major concrete structure was the same as that for the excavation works at the Cut: creating an organized and balanced production line. In this, Goethals was equally as successful at the locks as Stevens had been in the Culebra Cut. To build any concrete structure, the wet concrete must be poured into giant “molds,” referred to as shutters. Traditionally, these are made of wood. In massive structures such as the Panama locks, using wood becomes extremely wasteful, time-consuming, and thus expensive, as the molds can only be used two or three times before being scrapped. Goethals therefore opted to use shutters made of steel, which has a very high initial cost, but a relatively low total lifetime cost. Such a shutter weighed many hundreds of tons and constituted a major feat of precise manufacture in itself. These main shutters were then mounted on rails, running parallel to the walls, so that they could be easily moved.
An enormous amount of concrete was required. For the Gatún Locks alone, 2 million cubic yards would be poured. The cement came from the States, about five million bags or barrels of it. Obviously it made sense to source the stone and sand from nearer to the site. The Commission sent a party to try to secure sand from San Blas, which although ninety miles from Colón, had the best quality for cement. But this area was inhabited by Cuna Indians, who had long memories: “They did not look with favour on visits from the white men, whom they suspected were searching not for sand but for gold,” reported one of the engineers. The Americans were led before a seventy-year-old local chief, “seated on a block of timber, and he motioned his visitors to a seat on the sand at his feet. An air of great solemnity surrounded the proceedings.” The visitors outlined their plans to connect the two oceans by a canal, and how this would bring a better price for the Indians’ coconuts and ivory. “The Chief listened,” the account continues, “but when the story was finished he said that God had given the Indians their country, the land and the water, and that which God had given to the Indians they would neither sell or give to the white man.” That was his final word, and permission to anchor for the night was only granted on the condition that the Americans left the next morning and did not return. So the rock for the gravel, including old stones from the fortress of San Felipe de Todo Fierro, ended up coming from Portobelo, where a crushing plant was built, and the beach at Nombre de Dios was stripped of its sand, all to feed the voracious appetite of the Gatún Locks. On the Pacific side suitable rock and sand were found much closer to the lock sites, but to supply Gatún required barges negotiating the dangers of a sea journey of between twenty and forty miles. Any stoppages from bad weather would upset the brilliant mechanized system Goethals had now established to deliver the mixed concrete to where it was needed. Once the barges from up the coast arrived at Colón, they made their way up the newly redredged old French canal to Gatún, where unloading facilities and a mixing plant had been established. The concrete was assembled by the use of an ingenious system of small electric cars carrying buckets passing under hoppers containing sand, crushed stone, or cement. These would end up above one of the eight giant mixers, underneath which other similar cars waited for the finished concrete.
Chamber cranes at Miraflores
Spanning the site were four giant, 800-foot-long cableways, supported on either side of the great basin by towers 85 feet high. As each pair of buckets was filled with 6 tons of liquid concrete, the car moved off and was then picked up by a hook on the cableway and taken off to the site. At the same time, an empty skip returning from the site was set down on the railcar.
The system worked brilliantly. In one year a million cubic yards of concrete were delivered to be molded and set. The walls were built in 36-foot sections all the way to the top, which took about a week. Then the scaffolding, shutters, and cableway towers, all rail-mounted, were simply shifted along to the next section.
On the Pacific side concrete was batched on-site and lifted to the required position using a tower crane. Here even more concrete was used for the single lock at Pedro Miguel and the double tier at Mi-raflores. A small dam was also required at Miraflores in conjunction with the locks, forming the tiny Miraflores Lake, but this structure presented no serious difficulties.
The lock basins are the construction marvel of the canal. They are massive, a hundred thousand tons of concrete each, 1,000 feet long and more than 100 feet wide. Seen full of water they are impressive enough, but those who stood on the empty floor and gazed upward at the sheer, featureless walls, more than 80 feet high, taller than a six-story building, were awestruck. “These locks are more than just tons of concrete,” commented one such visitor, “they are the gate to that pathway of which Columbus dreamed and for which Hudson died. They are the answer of courage and faith to doubt and unbelief. In them are the blood and sinew of a great and hopeful nation, the fulfillment of ancient ideals and the promise of a larger growth to come … I left with the feeling that follows a service in a great cathedral.”
he mechanical marvel of the canal was the machinery of the locks. For one thing, the entire system was powered by electricity (supplied from the hydroplant next to the spillway), a great innovation at a time when steam- and horse-powered systems were the norm, and the electrification of factories was in its infancy. The power that did the hard work of lifting or lowering ships was, of course, gravity, the filling of a lock with water from the lake or a higher lock or the release of water to farther down the system. Water was drained or admitted through tunnels 18 feet in diameter, built lengthwise within the center and wide walls of the locks. Running perpendicular to these were smaller tunnels or culverts under the floors of the locks, fourteen per chamber. Each had five evenly spaced openings in to the floor, so that when water was admitted the turbulence would be minimized. The main culverts had large steel gates as valves. The principle was ancient: to fill a lock, the valve at the lower end would be closed and the upper one opened, and vice versa to drain it. Each lock had a lift of an unprecedented 28 feet, and the system aimed to raise or lower a ship this distance in just fifteen minutes.
Worries voiced in the U.S. press prompted the adoption of a series of stringent safety measures to avert the worst-case scenario of a ship smashing through the upper lock, causing the lake to pour through the breach. No ship would pass through the locks under its own steam. Instead the movement of the ships would be controlled by a group of powerful electrical cars, mounted on rails at the lock's edge, each in control through a windlass, a steel cable, attached to the ship. Each lock had a double gate, and a chain ran in front of the mouth of the top lock to catch and slow any vessel approaching in a dangerous way. Failing that, an emergency dam could be lowered across the channel within minutes.
All of the mechanisms for the lock operation were controlled from one position, where a working replica of the locks had been made, with the controls for each machine adjacent to its model. In addition, a system of safeguards meant that it was impossible, for instance, to open a culvert valve if the relevant lock gates were not firmly shut.
The manufacture of the lock gates was the only substantial part of the canal work that was subcontracted. A Pittsburgh firm, McClintic-Marshall, started assembly and installation in May 1911 at Gatún, August at Pedro Miguel, and at Miraflores in September 1912. The lock gates were designed to close into a flattened V. When open they slotted into a recess in the lock wall. Each individual gate was 65 feet wide and 7 feet thick and varied in height from 47 to 82 feet depending on its position. The highest were the bottom gates at Mi-raflores, which had to cope with the Pacific tides. The gates were opened and closed by giant wheels within the lock gates. These needed to be hugely powerful machines, but the gates weighed less than they looked. They were hollow, consisting of steel plates riveted on to a steel frame, and watertight so relatively buoyant. At their base were rollers, which ran on enormous steel plates embedded into the floor of the lock.
Few ships at that time needed anything like the space the giant locks provided, so intermediate gates were installed so that vessels less than 600 feet long could pass through more quickly and with less expenditure of water. So in all there were forty-six pairs constructed and installed. The great steel gates were one of the most spectacular sights on the Isthmus, particularly while the lock chambers were still empty and their huge size could be appreciated.
However, the work on the gates also figures, along with the explosion at Bas Obispo, as the worst memory of the construction period for the West Indian workers. The money was good—McClintic-Marshall paid US$0.25 an hour, more than double the basic labor wage—but the work was among the most dangerous anywhere. As a West Indian work song from the time had it:
Yuh gets more money for that job than working in the cut
But it all depends muh honey on if yuh don't get hut
For if you ever get a drop yuh surely have to die
For dem gates lawd gad gal is seventy five feet high.
In mid-1912, when work on the gates was well advanced, Harry Franck went to the Gatún Locks looking for a man accused of theft. “I found myself racing across the narrow plank bridges above the yawning gulf of the locks, with far below tiny men and toy trains, now in and out among the cathedral-like flying buttresses, under the giant arches past staring signs of “DANGER” on every hand… I descended to the very floor of the locks, far below the earth, and tramped the long half-miles of the three flights between soaring concrete walls … On them resounded the roar of the compressed air riveters and all the way up the sheer faces, growing smaller and smaller as they neared the sky, were McClintic-Marshall men driving into place red-hot rivets, thrown at them viciously by negroes at the forges.”
This was the worst job of all. The riveters, using heavy pneumatic hammers, worked on scaffolding that normally consisted of chains running down the face of the gate to which hooked planks were attached one above the other. There were four men per plank, and any sudden movement could unbalance and unhook the platform; if one came down from above, it usually took a few lower ones with it.
Eustace Tabois remembered a particularly gruesome accident at Gatún. “We were working there one Sunday,” he said. “I was inside the gate, right at one of the portholes, and I just happened to look out. I saw a shadow come down like that. And when I look out I saw this man down below. The scaffold break away with him and he went right down and the plank down there with him a spike went right though his head. Kill him dead, kill him dead. Man used to die every day.” “Scuffles would break away,” said James Ashby, “hurting many and sometimes killing men instantly. Lord how piercing!” As another West Indian commented, “The family of those men working on those locks were always fearful as to who may be next to fall.”
It was the same on the Pacific side. Jamaican Nehemiah Douglas was working on the giant gates at Miraflores when the cable holding his plank broke, “kill[ing] some men, on the spot. The amount of blood that flowed gave the appearance of a little gully,” he later wrote, “and when I saw what appeared to be an island of blood, I got nervous, I think, because how I got down, I do not know; but I got down and ran like never run before, straight home in Paraiso.” Many stayed away after witnessing incidents like this, even if they had pay to collect, but Douglas returned and was soon after hit by a crane and received a fractured skull.
Nevertheless, with the money that could be earned, men still “poured like sand” to go and work on the locks. However many were injured or killed, “there were always others to take their places without hesitation.” This became more urgent as the project progressed. It was obvious that the well-paid work was not going to be around for that much longer. The canal project was now nearing its triumphant conclusion.