Among the most fascinating of the surviving records of the work is a series of Army Signal Corps films made down in the Cut. Watching these rare old motion pictures (now in a collection at the National Archives), seeing the trains cut back and forth across the screen, seeing the dynamite go off and tiny human figures rush about through clouds of dust and smoke, one senses too how extremely dangerous it all was. At one point, when a shovel suddenly swings, Goethals can be seen to jump nimbly out of the way.
Bishop and those others who described the spectacle from the cliffs above had very little to say about such hazards. But year after year hundreds of men were being killed or hideously injured. They were caught beneath the wheels of trains or struck by flying rock, crushed to death, blown to bits by dynamite. “Man die, get blow up, get kill or get drown,” recalled one black worker; “during the time someone asked where is Brown? He died last night and bury. Where is Jerry? He dead a little before dinner and buried. So on and so on all the time.”
Construction of the canal would consume more than 61,000,000 pounds of dynamite, a greater amount of explosive energy than had been expended in all the nation’s wars until that time. A single dynamite ship arriving at Colón carried as much as 1,000,000 pounds– 20,000 fifty-pound boxes of dynamite in one shipload–all of which had to be unloaded by hand, put aboard special trains, and moved to large concrete magazines built at various points back from the congested areas.
At least half the labor force was employed in some phase of dynamite work. Those relatively few visitors permitted to walk about down in the Cut saw long lines of black men march by with boxes of dynamite on their heads, gangs of men on the rock drills, more men doing nothing but loading sticks of dynamite into the holes that had been drilled. The aggregate depth of the dynamite holes drilled in an average month in Culebra Cut (another of those statistics that defy the imagination) was 345,223 feet, or more than sixty-five miles.* In the same average month more than 400,000 pounds of dynamite were exploded, which meant that all together more than 800,000 dynamite sticks with their brown paper wrappings, each eight inches long and weighing half a pound, had been placed in those sixty-five miles of drill holes, and again all by hand.
Difficulty was had at first in determining how much dynamite to use in a single shot, depending on the depth of the holes, the spacing of the holes, and the character of the rock, which could be anything from basalt to the softest shale. The foremen responsible for the loading and tamping learned by trial and error. Different grades of powder were tried, different kinds of fuses and methods of firing.
Premature explosions occured all too often as the pace of work increased. “We are having too many accidents with blasts,” Goethals noted in June 1907. “One killed 9 men on Thursday at Pedro Miguel. The foreman blown all to pieces.” Several fatal accidents were caused when shovels struck the cap of an unexploded charge. Another time a twelve-ton charge went off prematurely when hit by a bolt of lightning, killing seven men. Looking back years later, one West Indian remembered, “The flesh of men flew in the air like birds many days.”
The worst single disaster occurred on December 12, 1908, at Bas Obispo. More than fifty holes had been drilled in the solid rock on the west bank of the Cut and these had been loaded with some twenty-two tons of dynamite. The charges had been tamped, the fuses set, but none of the holes had been wired since the blast was not scheduled until the end of the day. As the foreman and one helper were tamping the final charge, the whole blast went off, by what cause no one was ever able to determine. Twenty-three men were killed, forty injured.
As time went on the men became extremely proficient and accidents became comparatively rare considering the volume of explosives being used and the numbers of laborers involved. Still, more men would be killed, and very often, as at Bas Obispo, there would be too little left of them to determine who they were.
The shovels in the Cut set records “never anticipated,” as Goethals noted, and in the eyes of most beholders they became something more than mere machines. They had personality and gender–usually feminine, yet they were also likened to Theodore Roosevelt–and accounts of their prodigious feats of strength, as well as their agility, acquired a kind of mythical quality. The Canal Record’s full-page reports on their performance were read as avidly as baseball scores.
The peak was in March 1909, when sixty-eight shovels, the largest number ever used at one time in the Cut, removed more than 2,000,000 cubic yards, ten times the volume achieved by the French in their best month. The record for a single shovel was set in March 1910, when a ninety-five-ton Bucyrus (No. 123), working twenty-six days, excavated 70,000 cubic yards. More astonishing is the realization that the vast rift in the earth at Culebra was dug entirely by what, comparatively speaking, was a mere handful of machines. The volume removed from the Cut was 96,000,000 cubic yards. So even allowing for replacements, the average shovel dug well over 1,000,000 cubic yards, despite the worst kind of punishment year in, year out. No machines had ever been subjected to such a test and their record was a tribute to the men who designed and built them.
The shovels were deployed along the entire nine miles of the Cut, but in one section just to the north of Gold Hill they were stacked one above another at seven different levels, while seven parallel tracks carrying the dirt trains were kept constantly busy. “There were any amount of . . . trains, which were going in every direction,” noted a young English tourist in her diary; “they must be very well arranged.” In fact about 160 trains a day were running in and out of the Cut, and the degree of planning needed to handle such traffic can be further appreciated when it is taken into account that most of the track had to be shifted–removed, replaced, relocated–time and again. There were 76 miles of construction track within the nine-mile canyon, while in the Central Division as a whole there were 209 miles, not counting the Panama Railroad. In any one year well over a thousand miles of track had to be shifted about within that area just to keep the work moving in the Cut. And to complicate the problem further still, the bottom of the Cut, the main work level, kept steadily contracting in width the deeper the Cut became.
No one part of the operation–not the drilling, the blasting, the shoveling, the dirt hauling–could ever be permitted to interfere or disrupt another. So consequently every move was the result of very careful study. All shovels, every mile of track, every one of the hundreds of rock drills in use, were located daily on a map at division headquarters at Empire. Careful estimates were made as to the progress of each individual steam shovel, when it would have to be repositioned, when tracks would have to be shifted, what effect such moves would have on the disposition of drilling and blasting crews. So neatly was everything coordinated, so smooth were communications, that at the close of each day locomotive crews, as an example, had only to check the assignment boards at the roundhouses to see exactly what they were to do the day following. Traffic in and out of the Cut was directed from towers at either entrance by yardmasters who kept in telephone contact with the various dumps and with a half-dozen small towers strung out along the line of excavation. The yardmasters, who took their orders from the chief dispatcher at Empire, directed the passage of each loaded train to a particular dumping ground and ordered the right of way for the train when it hit the main line of the Panama Railroad. When the empty trains returned, it was the yardmaster again who distributed them to the shovels.
The dumping grounds–the other end of the system–were located anywhere from one to twenty-three miles from the Cut. Sixty-odd locations were used in the course of excavation, and though much of the spoil was simply gotten rid of–that is, put to no useful purpose–a very considerable part of it served to build earth dams, to build embankments on the new line of the railroad, and to create the huge new Naos Island breakwater at the Pacific end. To keep the flooding Chagres from backing up into the Cut as the great trench deepened, an earth dike was thrown across the north end, at Gamboa, seventy-eight feet above sea level.
All the dumps were carefully engineered, with tracks on several terraces. At each dump was another yardmaster who reported the arrivals and departures of trains and his “readiness for spoil,” who ordered the distribution of loaded trains to the several dumping tracks, and who, in addition, directed the movements of the Lidgerwood unloaders as well as two additional pieces of equipment that had since come into use: the dirt spreader and the track shifter.
Both devices were of vital importance to the efficiency of the entire system, since the least delay at the dumping end at once decreased progress in the Cut. The dirt spreader was a railroad car with big steel blades mounted on either side, these operated by compressed air. Once a train had been unloaded, its spoil dumped beside the tracks, the spreader came through, pushed by a locomotive, and did the job of several hundred men working with shovels. The track shifter, an even cruder-looking piece of equipment, was the creation of William Bierd, former head of the Panama Railroad, who had built the first one in the shops at Gorgona shortly before Goethals’ arrival. It was a huge crane-like contraption that could hoist a whole section of track–rails, ties, and all–and swing it in either direction. And since the tracks at the dumps had to be shifted constantly, to keep pace with the loads being delivered, it was an extremely valuable adjunct. Bierd’s own creation could shift track about three feet, but subsequent models, built after he resigned, could reach as much as nine feet. With one such rig, fewer than a dozen men could move a mile of track in a day, a task that would have taken not less than six hundred men working by hand.
The largest of the dumps were at Tabernilla (fourteen miles beyond the north end of the Cut), Gatun Dam (the most distant location), Miraflores, and La Boca, the largest, which had been renamed Balboa. Some of the dumps covered as much as a thousand acres, and in the rainy season they became great seas of mud, with tracks slipping and sinking five or six feet. At Tabernilla, more than 16,000,000 cubic yards of spoil were simply dropped in the jungle. At Balboa, 22,000,000 cubic yards were deposited, with the result that 676 acres were reclaimed from the Pacific as a site for a new town.
By far the most troublesome of the dumps was the Naos breakwater, where, as at Gatun Dam, spoil from the Cut was dumped from a huge trestle, this one being extended slowly across the mud flats of the bay. At first everything went as hoped. But then the soft bottom sediments began to give way beneath the heavier material being poured on top. Overnight whole sections of trestle and track would vanish into mud and everything would have to stop until they were replaced. In some areas the vertical settlement exceeded a hundred feet, while the slippage sideways was three times worse. In time not a single foot of the long trestle remained where it had been to start with. By 1910 well over 1,000,000 cubic yards of spoil had been dumped into the breakwater and still it was a mile short of Naos Island. To reach the island, ultimately, would require 250,000 cubic yards of earth and rock from Culebra, which was ten times what had been originally estimated.
“Culebra Cut was Hell’s Gorge,” one steam-shovel man would write, recalling the heat and dust and noise. Nor were the rains any less of a problem than in times past. In 1908 and again in 1909, the years of the heaviest work, well over ten feet of rain fell. To check the torrential runoff, to reduce the chance of landslides, Goethals did what the French had done: he had diversion channels dug parallel to the Cut. But he greatly expanded on their plan. The channel on the east side of the Cut, known as the Obispo diversion, ran for a distance of five and a half miles and had a minimum width of fifty feet. To build this ditch, and another similar to it on the opposite side, the so-called Camacho diversion, required another 1,000,000 cubic yards of excavation. And very possibly they were a mistake, as Goethals himself later conceded, since they were dug too close to the Cut and water seeping from them below ground may have been the cause of several of the more disastrous slides.
All technical problems at Panama were small problems compared to the slides in the Cut. The building of the great dam at Gatun, for so long the most worrisome part of the plan, turned out to be one of the least difficult tasks of all. A tremendous man-made embankment simply grew year by year at Gatun, extending a mile and a half across the river valley, a ridge of earth that was to be fifteen times as wide at its base as it was high. At the eastern end were the beginnings of the Gatun Locks; in the center were the beginnings of what was to be the dam’s giant concrete spillway. Two big outer walls of “dry” spoil were built first as a base for the embankment. These toes, as they were called, were nearly half a mile apart–the river, meantime, having been turned into an old diversion channel built by the French–and into the space between them was pumped hydraulic, or “wet,” fill, a solution of blue clay, which when dry would create a core almost as impervious as concrete. There was no lack of controversy over the project as time went on (much of it stirred up by Philippe Bunau-Varilla, who was convinced that Goethals did not know what he was doing), and once, on November 20, 1908, a section about two hundred feet long slipped sidewise and sank nearly twenty feet at the point where the dam crossed the old French canal. In the face of a storm of criticism and alarm in the newspapers, Goethals insisted that the situation was not serious and as it turned out he was perfectly correct. The damage was repaired; the work went on.
The slides, however, were a wholly different matter. The first occurred early in the fall of 1907, or just as Goethals was beginning to feel he had things under control.
The Cucaracha slide, located on the east bank of the Cut just south of Gold Hill, was the slide that had given the French such grief. On the night of October 4, 1907, after days of unusually heavy rain, Cucaracha “started afresh.” Without warning, an avalanche of mud and rock plunged into the bottom of the Cut, destroying two steam shovels, obliterating all track in its path. And for days afterward that same part of the slope, about fifty acres in area, kept moving down and down, slipping anywhere from ten to fifteen feet a day. “It was, in fact, a tropical glacier–of mud instead of ice,” Major Gaillard noted in an article for Scientific American, “and stakes aligned on its moving surface and checked every 24 hours by triangulation, showed a movement in every respect similar to stakes on moving glaciers in Alaska upon which the writer has made observations in 1896.” After ten days, when the slipping stopped, 500,000 cubic yards of mud had been dumped into the canal.
In 1910 Cucaracha let go twice again, burying shovels, track, locomotives, flatcars, and compressed-air lines. The entire south end of the Cut was bottled up for months. Within a year Gaillard reported that the worst of the slides were over, but in fact they were still to come. From 1911 on, as the Cut grew very much deeper, the slides occurred season after season and grew increasingly worse. “No one could say when the sun went down at night what the condition of the Cut would be when the sun arose the next morning,” Bishop wrote. “The work of months and years might be blotted out by an avalanche of earth or the toppling over of a small mountain of rock.” There were slides at Las Cascadas, La Pita, Empire, Lirio, East Culebra–twenty-two slides all together. Cucaracha was almost never still. It took three months to dig out the rock and mud dumped into the Cut by slides in 1911. In 1912 more than a third of the year, four and a half months, was spent removing slides. On one day more than a hundred trains would roll out of the Cut; the next day there would be none, because a monstrous slide had occurred.
Steam shovels were buried so deep in mud that only the tips of their cranes were left protruding. Hundreds of miles of track disappeared or were twisted into crazy roller-coaster patterns. In one bizarre instance a shovel and track were picked up by a landslide and were deposited unharmed halfway across the floor of the Cut.
On some of the terraced slopes the ground crept ever so slowly, barely inches a day, which was never enough to do any serious damage, but for two years gangs of men had to be kept constantly at hand, day after day, moving the track back to where it belonged.
At another place a slow but relentless slide kept perfect pace with the steam shovel
working at its base. The shovel never had to move; as much as it dug, the slide replenished.
For the engineers the problem was not merely the size of the slides. They were also confronted with a type nobody had anticipated. Those slides that had beset the French, like the comparative few experienced by Wallace and Stevens, were normal, or gravity, slides–Cucaracha being the largest and most destructive example. As explained earlier, they nearly always occurred in the rainy season, when a top layer of soft, porous material slid from the sloping plane of underlying rock, “like snow off a roof,” as one American said. But the new variety, and much the worst, were what geologists classified as structural break and deformation slides. They were due not to sliding mud, but to unstable rock formations, the height of the slopes, and, in part, to the effects of heavy blasting. As the Cut deepened, the underlying rock formations of the slopes lost their lateral support and were unable to withstand the enormous weight from above. It was as if the flying buttresses had been removed from the wall of a Gothic cathedral: the exposed wall of the Cut simply buckled outward under its own load and fell. Rains and saturation actually had little to do with such slides. In fact, some of the most horrendous happened during the dry season.
David McCullough Library E-book Box Set Page 391