Last Train to Paradise (13 page)

Dredges scooped out the muck and marl and sandy sea bottom in areas too shallow for barges to maneuver, and dumped their spoils to create man-made islands still visible from the bridge today. Next, the pilings—anywhere from seventeen to twenty-one of them, depending on the size of the pier—were driven into the exposed limestone cap formed by the remains of vast, ancient coral reefs.

Crews found the top layer of the limestone to be a relatively stable crust that varied from three to six feet in depth. Below that, however, the deposits were softer and pocked here and there by caves and layers of sand. In some places, pile drivers were unable to penetrate the cap rock more than a foot. In others, the operators would have to give up searching for a solid footing and hope that the crust they’d driven would hold. Some of the pilings might be twenty feet from top to bottom, others might be fifty.

Once the pilings were in place, helmeted divers were called in to aid in setting the outer forms, or cofferdams, then pouring the bottom cement seals, which were left to harden for four days. Once the seals had set, the water and debris were pumped out of the cement-bottomed cofferdams, and wooden molds were lowered inside and secured to the base. While the molds were different for the support piers than for the spandrel arch bases, the principle involved was the same. Workmen descended inside these inner molds to assemble the reinforcement rods, the ends of which would project above the tops of the forms to tie into the next section.

Then derricks mounted on nearby cement-mixing barges would transfer concrete to be poured into the forms, bucket by agonizing bucket. Once this step was completed, the rest of the work would take place above the waterline.

In the case of the spandrel arches forming the south end of the Seven Mile Bridge, the arch-ring molds could now be set in place, and the rest of the more complicated process would proceed. For the steel girder supports, what remained was a simpler matter. A last prefabricated form was set atop the pier that poked out of the water, and after more reinforcing rods were assembled inside, the form, sometimes rising as high as twenty feet above the water, was filled with concrete. After the concrete was allowed to harden for a few weeks, the forms were removed to be used again, and the steel support spans could be laid from pier to pier.

The support spans used in the Moser Channel were eighty feet long and nine feet high, so that the final roadbed lay nearly thirty feet above the ocean. Each of the huge girders weighed nineteen tons and had to be ferried to the work site on barges pulled by tugs, then lifted into place by massive Derrick Boom No. 9, a seventy-foot monster mounted on yet other barges.

When a pair of the spans had been secured to iron footings atop the piers, they were joined together by a series of cross braces and angle braces riveted into place. Once the spans were rigidly braced, crossties could be bolted atop them, and the rails fixed into place.

An account of an inspection tour over the sixty- to eighty-foot elevated spans by one of the workers gives a sense of the breathtaking nature of the work: “After dinner walked the girders (fourteen to eighteen inches wide). Ties were laid for a bit, but farther out there is nothing to walk on but the girders about twenty-six feet above the waters,” he wrote, before adding this understated afterthought: “It makes for some bad walking when the wind is blowing.”

If no problems intervened, crews could normally complete as many as four support piers in a week, and ironworkers could join as many spans in the same time frame. During one shining week in 1909, with support piers ready and everything falling perfectly into place, crews set a record by joining twelve spans, a rate that surely must have pleased “the Chief,” as Flagler had become known.

In an entry in his own diary dated June 29, 1909, Flagler noted that the last of the concrete support piers for the Knight’s Key Viaduct had been completed, and by August, he wrote to an associate that good progress had been made on the Bahia Honda Bridge as well. Rails would soon reach his ultimate destination, Flagler predicted:

“Unless we are hindered by unfavorable weather or delays in the delivery of steel for bridges, I think we shall have the track ready to run trains to Key West on or before the first of February [1910].”

Meanwhile, Flagler was hardly content to sit idly while work progressed. He kept a close eye on every decision, firing off dozens of memoranda and letters to Krome and his supervisors, often wondering if some task might be undertaken in a more efficient or economical manner.

In August of 1909, for instance, Flagler wrote to Krome, referring to the tides that were passing beneath those bridges they were struggling to build: “The thought often occurs to me that there is an immense amount of power going to waste between the piers on the viaduct and the three bridges, and I am often wondering whether it would not be feasible to utilize this power by making electricity with floodgates. Can’t you rig something up?”

Krome’s reply was a masterful rendition of dodging and double-talk, managing to reassure his employer that the manufacture of electrical power by harnessing the tides was a matter well worth looking into but one that would have to wait a bit while other fish were fried. Whether Flagler was fully palliated or not is unclear, but Krome was permitted to turn his attentions back to the task of reaching Key West in time to meet his boss’s brave predictions.

Given the way things had been proceeding, most observers felt that for once Flagler was not exaggerating when he said the line would reach Key West by the first of the year. But, once again, nature had its own plans.

On October 10, with work on the Knight’s Key Viaduct very nearly wrapped up, and Flagler ready to think about having timetables to Key West printed, the U.S. Weather Bureau reported that a hurricane had struck western Cuba and was now approaching the Florida Keys. The Weather Bureau had established an observation station on Sand Key, about eight miles southwest of Key West, where steadily dropping barometric readings suggested that the storm was strengthening far more rapidly than had been anticipated.

By 8:30
A.M.
on the morning of October 11, winds had reached seventy-five miles per hour on Sand Key, and weather observers were ordered to take shelter in a nearby lighthouse. Shortly thereafter, the anemometer cups were sheared away by winds estimated at one hundred miles per hour. Trees and fences were flattened, and the tides had engulfed the entire island. At ten-thirty, a massive surge carried the weather station away into the seas, where it was shattered and sunk.

W. R. Hawkins, resident civil engineer at Marathon from 1909 to 1912, kept meticulous notes on his experiences with the FEC. On the evening of October 10, he read the bulletin issued by the Weather Service in Washington, and, after spreading the word, copied the message down in his careful hand:

“NE storm warning. Tropical disturbance west of Havana, moving NW by N with apparently increasing intensity. Increasing NE wind tonight along gulf coast of central and southern Florida.”

There was no mention of the Keys in the bulletin, but Hawkins understood the implications. There was only one way to get from Havana to the mainland coast, and he and the others were in the middle of that path.

Early the next morning, the worst had happened. “Raining and blowing hard at 5
A.M.,
” Hawkins wrote. “Tides came very high. Nearly everyone on Boot Key [where a maintenance facility was located] went to Marathon [on nearby Key Vaca]. Had to hold on to the rails a good part of the way walking up the track to freight yard.”

And it is little wonder the men had to make their way along like crabs, bent over and clinging to the tracks. Hawkins had no access to a wind gauge, but the Weather Bureau estimated that winds had reached 125 miles per hour by the time they hit Marathon, where Hawkins and some three thousand others were quartered. On Knight’s Key, gusts were powerful enough to lift five of the nineteen-ton girder spans that had not yet been fully secured to their support piers, tossing them into the sea.

Earle Hartridge, one of bridge-builder C. S. Coe’s assistant engineers, described his ordeal in a letter: “We experienced one of the worst hurricanes at our Pigeon Key Camp, the center passing over us—we thought we would never survive it but we did by making a kind of crater in the top of the cement stored in one of the warehouses, where we rode it out, although the warehouse itself was blown away.”

Not everyone was so lucky. J. H. Brown was aboard Houseboat Beta, and, according to Hawkins, refused to go ashore when he and his companions were ordered to high ground at Marathon. “The houseboat was blown off,” Hawkins writes, “and poor old Brown was killed or drowned.”

Thirteen more men, the crew of the tugboat
Sybill,
died when their craft was driven down, but the shelters held fast for the men who had heeded the orders to come on land. There were no confirmed reports of other deaths.

Workers who ventured out after the worst had passed expressed amazement at the storm’s caprice. One man walked the completed portion of the bridge to where the massive, tons-heavy girders had been ripped away into the sea, and found himself staring in amazement at a keg of nails that still perched undisturbed at the edge of one of the now-empty piers.

While Krome’s precautions undoubtedly had saved the lives of many, the damage to the railway itself was massive. Some accounts were telegraphed from Miami to the rest of the nation, describing the extension as a total loss and claiming that forty men had actually died.

Flagler was quick to dispute these claims, estimating that he had suffered no more than $200,000 in damages. Other surveys suggested that the costs would rise into the millions: more than forty miles of roadbed and track had been washed away in the Upper Keys, with boulders weighing as much as ten tons dislodged by the surge of storm waters.

Still, Flagler was not deterred. In a letter to James Ingraham, he admitted that the company’s losses had been formidable, “but to the general public we must all keep a stiff upper lip and admit nothing.” To other associates, he urged, “My recommendation is to hoist Key West’s flag high, keep it waving and let it bear the inscription, ‘Nil Desperandum.’ ”

In an aside to Ingraham, however, Flagler gave some indication that his respect for the power of the storms was growing: “I say to you in confidence that I don’t want to invest a dollar in building improvements on any of the Keys.”

It seems that Flagler’s narrowing of focus and his flag-waving paid off. By November 8, less than a month after the storm had supposedly wiped out the line, the company had somehow managed to restore passenger service to Knight’s Key, with construction and supply trains following in the wake.

17

Learning Curve

As it turned out, Krome and Coe made one heartening discovery in the storm’s aftermath: all of the piers and viaduct supports that had so recently been set for the Seven Mile Bridge, as well as those of the great Long Key span, had emerged from the battering of winds and tides unscathed.

It was considered a major victory, for while the engineering team had designed those support structures to withstand forces four times as great as any storm had ever produced in the Keys, experiences of the last five years had taught everyone involved with the project that there was a vast difference between figures on paper and what was actually experienced in the lovely but treacherous Keys. It was with a great sense of relief, then, that work resumed on the pilings and arch supports across the broad channel.

What had happened in the Upper Keys, along the stretch of the railway constructed according to more conventional standards, was another matter, however. Local residents and federal consulting engineers alike had from the first maintained that adding fill instead of bridges wherever possible had created the potential for disaster, and the effects of the hurricane of 1909 proved the critics right.

A glance at the map illustrates the issue. Between the southern tip of the Florida peninsula and the island of Cuba, a distance of more than 150 miles, lie the waters of the Florida Straits, which constitute a broad channel connecting the Atlantic Ocean with the Gulf of Mexico. Another 150 miles or so southwest of Cuba lies the Yucatán Peninsula of Mexico.

Geologists theorize that the entire distance separating the United States from Mexico was once an unbroken stretch of land, and the Florida Keys that dot the waters between Cuba and the mainland United States are its vestigial remnants, all that remain after eons of erosion by wind and storm and tide. And in the existence of these vestiges, attractive as they may be to tourists, developers, and railroad builders, lies the problem, especially when a monster storm comes boiling up from the Caribbean, pushing a wall of water ten or twenty or thirty feet high ahead of it.

Until the railroad builders arrived, those storm surges, formidable as they might be, would course along the natural passageways that eon after eon of preceding surges had gradually cut and deepened through the reefs. Even tidal waves tend to follow the path of least resistance, and much of that force would simply be shunted into tremendous rivers of water surging through the natural cuts northwestward toward the Gulf. Once the storm had passed on and the waters were released from the driving force of the winds, the currents reversed direction and poured back through the channels separating the Keys toward the Atlantic. If something were to be suddenly dropped in the path of these surges, obviously, there would be hell to pay.

The problem was serious enough in the Middle to Lower Keys, where the passage to the Gulf was relatively unimpeded, the whole process of tide reversal spread over vast, open seas. But in the Upper Keys, the storm-driven waters would barely pass through the channels before they piled up against the mainland. With nowhere to go, a tidal wave would roar back across the Upper Keys much more quickly and with much greater force than to southward. The picture is a simplified one, but it lays out the nature of the wave action quite clearly.

When the railroad builders chose to close off natural channels, then, they were building what were essentially unreinforced dikes, stretches of earth and gravel perhaps twenty feet wide and four to six feet high, capped by wooden ties and railroad tracks. Such structures would have to stand against the forces that had carved through the Keys since time immemorial.

The result in 1909, of course, was no contest. Forty miles of track were swept away, boulders the size of automobiles carried off like pebbles. Even some of the barges that had been purposely sunk to avoid their being dashed to pieces on the shoals had been slung across the sea bottom and destroyed, or buried beneath tons of tide-shifted sands.

In this way, then, the Keys themselves taught Flagler and his men some valuable lessons. Where six miles of bridges were originally planned, eighteen miles would now be built to allow for natural tidal flow.

In addition, the design of the railroad bed itself would be changed. Originally the practice was to dredge or blast for marine marl in the shallows just offshore, then use that material as the substratum for a thick layer of crushed rock and gravel brought in from elsewhere and piled up until the proper grade height was achieved. In the aftermath of the storm, Krome and his men made two key observations. The first was that the worst effects of the storm had been created not by the advancing waves that slammed ashore on the Keys, but by the tides rushing back out to sea. They also noted that while the imported rock and gravel had been swept away by the tides, along with the ties and rails, quite often that substratum of limestone marl remained intact.

Flagler himself pondered these findings and, on October 16, wrote to FEC operations manager J. P. Beckwith with an unprecedented suggestion. They would reverse the normal roadbed building process from now on, Flagler said, first laying down a base of rock and then covering it with the slimy mixture brought up from the nearby sea bottom. Once in contact with the air, the marl hardened, forming a kind of natural concrete seal that was not unlike the very constitution of the neighboring landmass upon which the roadbed was built, and one far more likely to resist the action of storm-tossed waves than the riprap normally used atop grade.

Flagler’s theory proved correct and saved the railroad untold thousands over the ensuing years, but it was only the beginning of an ongoing process of construction research required by the singular project. Engineers were also concerned about the effects of the salt spray upon the steel deck spans and reinforcing girders. An FEC paint laboratory established in Marathon would determine that no paint could be devised that would provide protection against the local conditions for more than two years. As a result, painting crews would be required to work year-round on a job that would never cease, so long as the railroad stood.

The issue of wave and salt action on the concrete support piers was also investigated, but there the news was more optimistic. The results gained by observing concrete samples submerged in laboratory tanks, and by the detailed record-keeping on piers already put in place, suggested that while the concrete might erode slightly after being set (up to one-sixteenth of an inch), the process would eventually stop, impeded by the formation of algae and other marine-aided deposits that effectively sealed the concrete.

By the beginning of 1910, then, work had resumed full force, all of it guided by a directive Flagler issued to Krome in the aftermath of the hurricane: “From this time on let us bear in mind that what we need is permanent construction, even though it costs a little more money and takes considerably more time. I would rather be two years completing the line to Key West and have it permanent, than to have a repetition of the disaster of the hurricane of the 11th.”

Permanent construction of the sort Flagler was referring to would not come cheaply, however. Early in 1910, Flagler wrote to John Carrere, designer of the Ponce de Leon Hotel in St. Augustine, that repairing the damage caused by the hurricane had actually cost him $1 million, and reiterated that it had taught him a valuable lesson about upgrading the quality of the work. He estimated that it would require at least another $9 million to push the track to Key West, a figure that did not include the costs of a terminal and docking facilities.

“As it is,” he told Carrere, “I have got to finish the work out of my income and I cannot expect to live long enough to do that.”

As a result, Flagler confided, he had taken an unprecedented step, a secret “known to but one other person.” After talking the matter over with Krome, Flagler had gone into debt for the first time since the project had begun, authorizing J. P. Morgan to issue $10 million in bonds to underwrite the completion of the route. It was a momentous step for Flagler, a clear admission that the project had become more formidable than even he had reckoned.

“I should feel less anxiety,” he concluded, “were it not for the Governmental raid upon the Standard Oil Company [from which Flagler derived his most dependable flow of income]. Perhaps this will clear up in a few months and enable me to take a little more hopeful view of the future.”

While marine crews labored to complete the various sections of the Seven Mile Bridge, land-based crews pushed track southward along the Lower Keys—Little Duck, Ohio, Missouri—and Key West–based crews worked northward over Stock Island, Boca Chica, Big Coppit, Sugarloaf, Cudjoe, Summerland, Ramrod, Little Torch, and Big Pine Keys, some forty miles of work altogether.

Weekly work reports filed by Coe give some sense of the nature of the effort:

. . . for the week ending January 29, 1910.

Pile Driver #1 drove dolphins [preliminary pilings] in Moser Channel up to Thursday when she moved over to the rock pile and replaced the dolphins torn out by the blow of the 21st. [Note that it was not only hurricane weather that plagued the project.] On Friday she moved on line of Pigeon Key Trestle #2 and on Saturday drove 13 piles in the trestle.

Pile driver #11 worked in Moser Channel all week replacing the anchor and fenders piles of the piers up to Saturday when she commenced driving on Pier #54.

Pump Barge #12 cleaned off the old seals ready to allow the cofferdams to be replaced.

Catamaran #3 set five cofferdams for the week. The last cofferdam set being #45. The cofferdams have fitted over the old seals very nicely so far. [The references are to repair of pier bases not yet completed when the hurricane hit.]

. . . Derrick Barge #9 completed setting the steel on Knight’s Key Bridge. Settling the last span on Wednesday. She set the first span in Moser Channel on Friday and set two more on Saturday, making 4 for the week.

50 recruits were received from Key West on Thursday which brings our force up to the limit.

The last is a telling comment, for it suggests that regardless of all calamity, and in spite of the rigors of the work, Flagler was by now having relatively little difficulty in finding men willing to join his effort. Another progress report issued at about the same time breaks down the workforce headquartered on Pigeon Key in greater detail:

The 1910 U.S. Census provided a further breakdown of the workforce stationed at Pigeon Key: 61 of the men came from 28 different states, with New York providing the most, at 12. The other 150 men tallied came from a welter of countries, including 77 from Spain, 33 from Grand Cayman, and 13 from Ireland. Only 5 black workers were listed.

The breakdown offers some insight into the FEC’s evolving hiring practices. The Spanish workers had always been “stayers” in Flagler’s eyes, and as word got back to their countrymen about improved living and working conditions on the project, more and more of them came to sign on. Native Caymaners, likewise accustomed to the climate and the insects, had also come to constitute a significant part of the workforce.

In addition, the camp on Pigeon Key offered accommodations that were far superior to those of the early outposts built on the Upper Keys. Most men were quartered in sturdy bunkhouses and had access to an infirmary on site. Each of the dormitories housed up to sixty-four men, and each had a reading and recreation room with electric lights powered by a generator. There was a domestic staff that washed and changed linens once a week.

A writer visiting the camps for the
Railroad Gazette
noted that while mosquitoes were “large and fierce,” all the bunkhouses and porches were screened. “The men get up at 5:00 in the morning, take a bath . . . have breakfast at 5:30, work from 6:00 to 11:00, have dinner, go back to work at 12:00 and knock off at 5:00, with supper from 5:30 to 6. Sundays are rest days.” Many workers reported that conditions at the camp were superior to those at their own homes.

Also, supervisors had learned to look the other way on some matters of policy. The derelict freighter
Senator
was now anchored permanently in Boot Key Harbor, just offshore from the busy work camps on Pigeon and Long Keys, and was doing a thriving business as a bar and house of prostitution. And whether or not Flagler cared much for the idea, by now the FEC was making no efforts to send the
Senator
packing.

Wherever they were from, and whatever their duties and their off-duty proclivities, the workforce by this point seemed to have rallied about the drive to complete the project so that the Chief could “ride his own iron” to Key West while he was still able. Though the Seven Mile Bridge remained to be finished, and a last great channel was yet to be crossed, the race to finish was on.