26 Fighting Nature

As recently as the year 1900, the Imperial Valley had not a single civilized inhabitant, and not one of its hot, arid acres had ever been cultivated. It now has a population of more than forty thousand, with churches, banks, ice factories, electric-light plants and fine school buildings, in half a dozen prosperous towns, and its 400,000 acres of cultivated land have produced, in the last six or eight years, crops to the value of at least $50,000,000. The history of this fertile oasis in the Colorado Desert will forever be connected with the name of E. H. Harriman. He did not create the Imperial Valley, nor did he develop it; but he saved it from ruinous devastation at a time when the agency that had created it threatened capriciously to destroy it, and when there was no other power in the world that could give it protection.

—George Kennan, The Salton Sea (1917)

Amid his trials of 1906, Harriman found to his dismay that the list of foes arrayed against him included nature itself, which unleashed its wrath in the form of two major catastrophes. Although the last thing Harriman needed was more fights, these offered a refreshing difference. They were clean and elemental, pitting force against force with the stakes clear and the outcome uncluttered by human conniving. Or so Harriman thought when he took up the challenges.

Information was vital to the running of a far-flung empire. To get it reliably and privately, Harriman had in 1901 leased a private wire from Western Union between New York and Chicago, then built an extension from Chicago to Omaha. By using the Union Pacific’s wires west of Omaha, Harriman had a communication link that enabled him to talk to people in the West by telegraph as a later generation would by telephone. He would fire messages at his secretary, who tapped them out on the key and then read him the response when it came.¹

Sometimes nature thwarted this system. In November 1903, just as Harriman was preparing to leave for the ceremonies opening the Lucin cutoff, a monster storm knocked out telephone, telegraph, and rail service, isolating Harriman at Arden for more than a week. A secretary named W. V. Hill managed to get a brief message to him by a circuitous route and was told in reply to get to Arden the best way he could. Dutifully Hill started out, trudging through the rain on foot along the railroad track. As he neared Arden, a small figure in rubber boots came striding down the track toward him. It was Harriman, who had come to meet his man and make sure he was all right after so soggy a journey. This was Hill’s first glimpse at the human side of Harriman. In the office, Harriman had a way of keeping his staff constantly on edge awaiting his call that rang out like a rifle shot.²

When they were finally free to leave Arden, Harriman asked Hill if he had ever been out west. The young man shook his head. “I’m going in a few days,” Harriman’said, “and I want you to come along.” In this abrupt way Hill, who was not yet a stenographer, found himself in charge of Harrimans deluge of correspondence, interviews, and appointments on a trip lasting several weeks. This trial by fire prepared Hill for many such trips, including one that arose unexpectedly on the morning of April 18, 1906, when a series of violent tremors shook the Pacific coast and engulfed San Francisco in a wall of fire.

Hill got word of the disaster that same morning and rushed to Harrimans residence in New York with the news. Harriman went at once to his office and ordered no expense spared in restoring communications with San Francisco. A line was patched through to Oakland Pier, which had escaped damage. For much of the day Harriman digested reports from his men on the scene and sent out a steady stream of orders. Southern Pacific trains were to carry survivors out of and relief supplies into the city at no charge and cooperate in every way with civic and military officials.³

No one could match Harriman in his ability to direct a complex operation far from the scene, but that was not good enough for him. He told Hill to get ready to leave the next morning for San Francisco. Civic leaders had rebounded well from the shock, and the army was there in force. Hundreds of people were already toiling furiously to contain the damage and organize the relief effort, yet Harriman felt obliged to join them. The Southern Pacific controlled most of the city’s transportation arteries and was vital to the relief work. Harriman wanted to make sure the work got done the right way—his way—and he wanted to meet the challenge directly.

For three days his special train rumbled across the continent, scooping up hundreds of telegrams at every stop. Harriman exhorted his officials to keep supply trains moving at all costs and to make all company facilities available to local authorities. Early on the morning of April 22 his train rolled into Oakland and parked on the outer edge of the yard, where his car became command headquarters. The devastation had spared Oakland and most of the waterfront area. Harriman’s order to general manager E. E. Calvin had been a laconic “Do all you can,” and he was gratified to learn how much had already been done. Calvin had sent agents to buy carloads of food in nearby towns and run them into San Francisco for army troops to distribute. Company ferries carried hundreds of refugees to Oakland at no charge and returned with tons of supplies. Company riverboats also brought in supplies.⁴

Since the Ferry Building had escaped damage, Calvin threw open its gates for use as a relief center. The railroad opened information bureaus throughout the city and evacuated thousands of people at no charge. With connecting roads also providing free tickets to whatever destination the riders chose, more than two thousand cars filled with refugees left San Francisco in the two days after the quake. The staff of the company hospital worked to exhaustion tending victims of the quake and twice had to flee when flames overtook first the hospital and then the car barns used as a temporary site. They finally settled their patients on an athletic field near the yards in San Mateo.⁵

Shortly after his arrival, Harriman met with a committee of state, local, and business leaders formed to deal with the crisis. He listened to their plans for rehabilitation, then fired off a stream of suggestions to get the work started. After all of them were adopted, the Southern Pacific men secured an office building for themselves and set to work rebuilding it. For nearly two weeks Harriman’stayed at the pier, orchestrating the flow of supplies, firing telegrams in all directions, issuing studiously upbeat statements to the press, stamping the crisis with his unique blend of energy, efficiency, and optimism.⁶

No other business figure rushed to the scene; for once, Harriman had center stage to himself. Every report from him stressed two themes: the marvelous spirit of everyone on the scene and the determination to rebuild. “The rich and the poor have to be cared for alike,” he noted in one telegram home, “and it is wonderful how courageous and hopeful they all are. It is the kind of spirit upon which can be depended the successful return of upbuilding and prosperity.” The earthquake was history; all eyes must turn to restoring the glory that was San Francisco. His positive attitude boosted everyone’s spirits, and it was not confined to words. The barrage of telegrams from his car included appeals, public and private, for funds to help this work as well as the victims. Harriman put $200,000 from his own roads into this kitty.⁷

Even to his critics Harriman emerged in this crisis as a symbol of conquering adversity through sheer will. In one stroke he had done more to clean the tarnished image of the Southern Pacific in California than had been managed in three decades. His response, declared one editor, put Harriman “before the people of the Pacific coast in an entirely new and entirely favorable light.” Another hailed the “Harriman monopoly” for proving itself “in the hour of need the strongest and most faithful friend that the city of San Francisco could have had.”⁸

Harriman was surely aware of the public relations coup he had scored, but it was strictly a fringe benefit. He had not planned the trip west or calculated its effect; he had simply reacted to a crisis and gone instinctively to the scene. The workload had been staggering—thousands of telegrams to handle as well as meetings, inspection tours, reports, interviews, and decisions large and small. One day he was standing with Calvin on the deck of a ferryboat when a stranger approached Calvin with a hard-luck story about his wife having died from the shock. The man had no money, knew no one in the city, and desperately needed transportation back east for himself and his wife’s remains. Without hesitation Calvin wrote out an order for free travel.

“How do you know the man’s story was true?” Harriman demanded after the man had left. Calvin admitted that he didn’t, but he thought it was better to take a chance than run the risk of denying help to someone in need. Harriman nodded. “It is well that you reached that conclusion,” he said, “because if you had not done so, I would have taken the case out of your hands and given him the money myself.”⁹

When the pace finally slowed, Harriman turned wearily to Hill and quipped, “We have been a little busy.” The secretary replied hesitantly that there were still hundreds of telegrams that Harriman had not yet seen. Harriman lugged the bundles to his bed, where he fell asleep reading them. On the trip home he asked first for a slow run so he could rest, then changed his mind and ordered full speed ahead. When his special train paused at Green River, Wyoming, he sent for the engineer and fireman. They trudged back to his private car slathered in coal dust, expecting to be reprimanded for something. Instead Harriman’shook their hands, thanked them for running through the rugged canyons without disturbing his sleep, and gave them a week off with full pay.¹⁰

Harriman was in an expansive mood, tired but pleased with himself. The trip home turned into another record run, landing him back in New York in just under three days. The papers loved it, and a larger than usual gaggle of reporters was waiting for him at Grand Central Station. After greeting his family, Harriman joked with the reporters, then planted himself against a wall to field their questions. He wanted to talk about San Francisco, not the record run, and he deluged them with facts, figures, and opinions on the city. Later he even put together a short article on the subject.¹¹

Ironically, Harrimans response to the disaster was about the only thing that earned him public praise during 1906, and its benefits were quickly subsumed by the parade of negative publicity that summer and fall. Even his response to the quake had unfortunate repercussions. In his tributes to those on the scene, Harriman had lavished praise on Mayor Eugene E. Schmitz, who was widely considered corrupt and unscrupulous. A few months later Schmitz and an ally were indicted on charges of bribery and extortion, which tainted Harrimans own falling star. The usually voluble Theodore Roosevelt said not a word of praise for Harriman’s work in the earthquake crisis, but he was quick to note his role in another catastrophe.¹²

In his year of discontent, when so much else was going wrong, Harriman could not have predicted that an earthquake would compound his miseries. Neither could he have guessed that a runaway river in the West would complicate his difficulties with Roosevelt. Like most of the controversies that plagued Harriman, this one began somewhere else with someone else and somehow managed to find him.

Today the Imperial Valley of southern California is a garden of breathtaking beauty and fertility from which flows a large portion of the nation’s fruits and vegetables. In 1900, however, it was a barren alluvial plain without a single inhabitant—most of it lying below sea level, its parched surface swept by howling sandstorms and scorched by blazing sunlight that sent temperatures well past a hundred degrees. To the northwest lay a huge basin known as the Salton Sink, once the home of an ancient sea the size of Great Salt Lake.¹³

A hundred miles long, thirty-five miles wide, and a thousand feet deep at its lowest point, the sink had been dry for centuries. A salt company working its bed provided the region’s only industry. The Southern Pacific’s tracks passed alongside the sink, taking full advantage of the land’s low profile. Both the sink and the valley owed their origins to the volatile Colorado River, which carved a jittery route between California and Arizona. Long ago the river had flowed into a Gulf of California that extended a hundred miles farther north, but its fast-moving current deposited so much silt that a dam formed, cutting the gulf in half and creating the Salton Sea. Over the centuries the river emptied into the lower gulf until the sediment built up high enough to divert it back to the Salton; then, after following this route for years, it reversed course again and returned to the gulf.

The plain created by this process was a natural hothouse with ideal soil and sunlight for growing almost anything if a source of water could be found. As early as 1853 a geologist named William P. Blake, who was the first to examine the sink closely, declared boldly that the plain if irrigated would “yield crops of almost any kind.” The obvious source was the Colorado River, which was dangerous to tamper with because of its fast currents and unpredictability at flooding times. Nevertheless, in 1900 the California Development Company (CDC) started work on an irrigation channel and created a Mexican subsidiary to extend the work below the border.¹⁴

In May 1901 the company completed its first cut into the river just opposite Yuma. To skirt a series of sand hills between the river and the valley, a canal was dug in Mexico toward one of the dried riverbeds that once carried water to the sink. On this riverbed water flowed northward into the valley. The promoters sold water rights and organized a land company to lure settlers. They succeeded beyond their wildest dreams; settlers arrived so rapidly that by 1905 the Imperial Valley boasted ten thousand inhabitants. A network of irrigation ditches spread across the plain, 120,000 acres of land went under cultivation, towns sprang up, and the Southern Pacific laid a branch line into the valley.

Natures battleground: This map shows the Imperial Valley, the Colorado River, and the Salton Sink that was soon to become a sea. The island in the river can be seen right center, just above the Yuma Project. (Smithsonian Institution Report, 1907)

Prosperity meant a growing demand for water. Giddy with the ease of their success, the promoters were happy to oblige. Their engineering costs had been remarkably low, and they saw no reason to increase them. In their euphoria they ignored the fragility of the resource on which all life in the Imperial Valley depended. By one estimate, a single day’s water supply for the valley carried enough silt to form a levee twenty feet high, twenty feet wide, and a mile long. While the demand for water soared, the supply declined as silt clogged the canal and ditches. Yet the company did nothing to dredge or sluice or create settling basins for the canal.

By 1904 the main canal had become so clogged that it could no longer meet the valleys needs. When some crops failed for lack of water, disgruntled farmers filed damage claims. Once the summer flood level dropped, the promoters faced a crisis. The dredging tools at their disposal could not possibly clear the clogged four-mile section of the main canal in time to provide an adequate water supply for the coming season. Their only hope was to open a new heading from the river to a point on the canal beyond the silted section.

Opening a new cut involved the risk of winter flooding, but records kept at Yuma showed that the river had flooded only three times in the past twenty- seven winters. Late in September a cut was made about four miles south of the original heading at a point where the river divided around an island. From this spot the engineers could reach the main canal by dredging only a quarter-mile through easy material. The new opening looked to be the perfect shortcut; it turned out to be the gateway to disaster.

The company did not install a head gate. It did not yet have formal permission from the Mexican government (the cut was on Mexican soil), and it wanted to keep the cut open until summer to maximize scouring action. But the river had other ideas. By an unhappy coincidence it chose that season to launch one of its periodic reversals of pattern. Two February floodings served notice of this intent but caused no damage. An engineer estimated that the heading could have been closed then for $5,000. In March the company protected the cut with a dam of piles, brush, and sandbags only to have a third flood sweep it away. “This was sufficient notice to us,” wrote the company engineer with masterful understatement, “that we were up against a very unusual season.”¹⁵

Hurriedly the company threw up another dam; still another flood carried it away. Normally the Colorado began its rise in the spring, peaked in July, and subsided in August. This year the waters began rising in the winter and kept rising. The series of floods sent water surging over the banks of the canal toward its ancient repository, the Salton Sink. If this flow continued, the old, long-dry riverbeds would spring to life, the sink would fill, and the Colorado might shift from its present to its ancient bed.

By the spring of 1905, the promoters saw bankruptcy or worse staring them in the face. Lacking the funds and equipment to contain the river, they approached the largest corporation in the valley, the Southern Pacific. A skeptical Julius Kruttschnitt heard their plea for a loan and sent them packing. Desperate for money, they went directly to Harriman, who had no personal interest in the valley and learned of its situation for the first time in this interview. Nor was he yet aware how serious a danger the river posed. The promoters assured him that the heading could be closed for $20,000 at most; the rest of the $200,000 they sought would go to pay off old debts and expand the irrigation system.¹⁶

Harriman could not help but be impressed at the bold idea of transforming a desert into a garden. He was familiar with irrigation projects elsewhere along his lines and quick to grasp the potential business Imperial Valley might generate for the Southern Pacific. But he also liked to control anything he went into, and he disliked going into ventures with men he did not know. In his usual fashion, he sent a trusted official to investigate the matter.

Something in the official’s report struck a responsive chord in Harriman. He decided to grant the loan even though Kruttschnitt and others advised strongly against it, but on terms that gave him working control. The Southern Pacific held 51 percent of CDC’S stock as collateral and got the right to name the president and three directors. Shortly after signing the papers on June 20, 1905, Harriman asked Epes Randolph to take the presidency of CDC. A gifted civil engineer, Randolph had already served Harriman well as the head of Southern Pacific’s Arizona and Mexican lines.

Events soon made Harriman wish that he had not been so adamant about controlling the company. At the time, however, the project was for him a small-change operation. No one yet grasped how vulnerable the valley was to the whims of the river. It was, observed an engineer later, “a Holland—500,000 acres of an American Holland.” Late in June an engineer from the federal Reclamation Service examined the cut and described the situation as “not serious, but sufficiently alarming to require some attention.” A few weeks later Randolph went down to inspect the heading for the first time and was appalled by what he found.¹⁷

Already the river had widened the new cut from forty to one hundred feet and deepened it from eight to twenty feet, and its current was scouring furiously. The flow of water through the cut had reached an estimated eight thousand cubic feet per second and was increasing by quantum leaps. Randolph realized that the river was overwhelming the gap and shifting its course. If that process continued, the Imperial Valley would be wiped out. The fall of the Colorado River from the break to the gulf was only a hundred feet compared to four hundred feet from the break to the bottom of the valley. Since the distance was about the same either way, the river, if it broke through, would follow the steeper decline with a velocity that would make it almost impossible to contain.¹⁸

And it was gathering force fast. Although the river was falling, the amount of water pouring through the cut was rising. Already the overflow accumulating in the Salton Sink had flooded out the saltworks there; before long it would threaten the Southern Pacific tracks. Randolph hurried to tell Harriman that the situation was far worse than he had been told. Not only the Southern Pacific tracks but the valley itself was in jeopardy. The entire $200,000 would not touch the job; no one could predict how much was needed, but the tab might run as high as $750,000. The work, Randolph added gravely, would be “a very difficult undertaking.”¹⁹

Harriman chewed on this information, then asked, “Are you certain you can put that river back into the old channel?”

“It can be done,” Randolph assured him.

“Go ahead and do it.”

Randolph and his engineers tried first to divert the river around the eastern side of the island by erecting a barrier on the side nearest the cut. When the opening continued to widen, he got Harrimans consent to try a more elaborate barrier dam across the west fork. Randolph started work in September and had not quite completed the dam by November. There seemed no reason to hurry; the river was at low ebb and flood season was months away. Yet he barely had time to admire the new barrier before disaster struck.²⁰

The only thing more unpredictable than the Colorado River was the Gila River, which drained into the Colorado at Yuma and was notorious for sudden, violent floods. On the last two days of November the Gila sent a raging wall of water clogged with driftwood roaring down on Yuma. Rising at the rate of a foot an hour, this torrent overflowed the island and carried much of it away, ripped the new barrier dam to shreds, and poured into the heading with a fury that widened the breach to six hundred feet. Through it the river flowed with mounting force, spreading over the plain in search of old riverbeds on its way to Salton Sink.

Randolph’s worst fears had come true. The canal and ancient bed had become the river, carrying its waters to the sink, which was fast becoming a sea again. If the heading could not be closed before next springs flood season, the Colorado might keep its new course. The Imperial Valley would then become a freshwater lake, and the Southern Pacific would find sixty miles of its track under water. Two threats faced the residents of the valley: the surging waters might submerge their land, or they might destroy the system of irrigation ditches on which all life in the valley depended. They might drown or die of thirst.

The river could not be shut out entirely; it had to be contained and controlled. After a testy debate among the engineers over what to do, Randolph decided to pursue two plans at once. A concrete head gate set in rock would be installed at a point where the river could be turned into a newly dredged channel as a permanent source of water for the valley. In addition, a wooden head gate would be put just above the breach to divert the river while a permanent dam was constructed. Through the winter Randolph pushed both projects only to find himself in a losing race. The concrete gate was not finished until late June 1906; the wooden gate was not ready until mid-April despite round-the-clock efforts.

Throughout 1906 the Colorado remained at near record levels, discharging that year more than twice its usual annual flow. The concrete gate could not be made ready in time, and the wooden one was overwhelmed by the river. When Harriman rushed to San Francisco in April to help fight the earthquake, Randolph went to meet him with more bad news. All efforts to force the Colorado back into its bed had failed, and the river was pouring unchecked into the valley. A giant dredge under construction in San Francisco and urgently needed in the valley had been delayed by the earthquake. The funds loaned to CDC had been used up. Randolph shrugged helplessly and asked what Harriman wanted him to do.

Standing amid the debris of one disaster, pondering the momentum gathering around another one, Harriman did not hesitate. The earthquake had already disrupted business on the Southern Pacific and tied up its lines. There was no way of knowing what effect the runaway river would have or how much it would ultimately cost. Nor was there a compelling reason why he should get deeper into it other than his love of a challenge and his instinct for doing what should be done when adversity struck. He had shown that instinct by hurrying to San Francisco, and he revealed it again by giving Randolph another $250,000 to spend on harnessing the river.

While Harriman battled the earthquake, the feisty Randolph returned to the valley with renewed determination. When the summer flood crested late in June, the river poured through a cut now half a mile wide at a rate of seventy-five thousand cubic feet per second. Its restless waters spread across a ten-mile area on their way to the Salton Sea, which had grown to an area of four hundred square miles and was rising seven inches every day. Five times that summer the Southern Pacific moved its track to higher ground. The saltworks lay beneath sixty feet of water, two towns had been damaged, and thousands of acres of crops were flooded or badly eroded. Another thirty thousand acres lay parched because their irrigation ditches had been swept away.

In the water’s relentless path the soft, powdery delta soil crumbled into new beds and rapids, which soon turned into cascades and then falls, some of them eighty feet wide and a thousand feet high. One engineer estimated that in nine months the surging river dumped into the Salton Sea an amount of silt four times the entire yardage of the Panama Canal. “The water simply tore things in the valley to pieces,” he exclaimed. “It was perhaps the most extensive and remarkable geological action of recent geological times.”²¹

Runaway river: One of the channels cut by the Colorado River as its runaway waters headed toward the Salton Sink. Note the lone figure standing above the cut on the right. (Smithsonian Institution Report, 1907)

Randolph found himself staring at a crisis that had no precedents. Other rivers had overflowed their banks and then receded; no one had ever tried to force a runaway river back into its bed. Forty or fifty top engineers visited the site and could agree only on what would not work. Randolph decided to go with his gut feeling, which was to erect a rock dam. An ordinary barrier dam could not withstand the enormous pressure of the water pouring through the gap, so Randolph tried a deceptively simple notion: keep piling rocks into the breach until more remained than were washed away and the flow gradually diminished. Once the gap was closed, a permanent system of dams and levees could be built.

Nearly all the engineers dismissed Randolphs plan as impractical, yet Harriman’supported it at once. This was an extraordinary commitment given the high stakes and the doubts of so many distinguished engineers, but Randolph was his man and he had confidence in him. “I do not believe,” Randolph marveled later, “that any man whom I have ever known, except Mr. Harriman, would have undertaken it.”²²

Feeling the full weight of the responsibility thrust on him, Randolph spared no effort in mounting his campaign. The Southern Pacific ran a spur line to the site and furnished all the work trains needed. Randolph borrowed three hundred of the giant side-dump cars or “battleships” used on the Lucin project. The CDC deployed its fleet of three light steamers and several barges. To get the vast amount of rock needed, Randolph tapped every quarry within four hundred miles and opened a new one. Clay came from a deposit just north of the Mexican border and gravel from the Southern Pacific’s huge pit forty miles west of the opening. Massive quantities of timber and steel cable poured in from Los Angeles, along with pile drivers, steam shovels, and other heavy equipment. Randolph recruited laborers wherever he could find them—Indians from six different tribes in the region, Mexicans, and drifters.

Lost towns: A view from temporary dikes of the submerged towns of Calexico and Mexicala in the Imperial Valley. (Smithsonian Institution Report, 1907)

On August 6 they went to work in earnest. The plan was to dam the breach enough to shunt the water to the bypass where the wooden gate had been installed and then close the heading permanently. It took two shifts of men working around the clock twenty days to fashion a brush mattress as a base for the dam. Once this was put in place, the crews threw a railway trestle across the gap. On September 14 the first battleships rumbled across the trestle to dump loads of rock on the mattress. Gradually the flow shifted to the bypass, but with it came masses of driftwood banging against the gate. On October 11 the gate broke off and floated away. Behind it came the river, turning the bypass into the main channel and leaving the new dam standing dry.

Harriman responded to the bad news with a shrug and sent Randolph back to work. The dam had acquitted itself well, convincing Randolph that he was on the right track. The trestle was extended across the bypass, and three thousand loads of rock were dumped into the bypass until both openings were dammed and connected by huge levees to form a barrier half a mile long. Gobs of clay and gravel were pumped into the openings to seal the dams. On November 4 the river slid grudgingly back into its old bed.

Everyone breathed a sigh of relief and turned to the task of dismantling the works they had built. Then, on December 7, the fickle Gila River sent another of its killer floods boiling into the Colorado. The new dams held firm, but one of the earthen levees south of them developed a leak. Originally built by CDC, the levees had been reinforced but not rebuilt. In short order the river gnawed a hole in the levee; three days later it was pouring through a thousand-foot-wide gap on its way to the Salton Sea. A disheartened Randolph, watching all his work come undone, realized that it was not enough to dam the river. Full protection required a twenty-mile system of levees along the bank as well.

The bad news brought by Randolph this time put Harriman in a dilemma. He had already spent $1.5 million trying to contain the river, and Randolph estimated it would take at least that much more. The twenty miles of levees would be expensive because they required excavations as deep as ten feet to anchor them on material that could not be eroded away again. The total traffic from the valley did not come close to the cost, let alone the profits. Harriman could move the roads tracks again for only $60,000, but another, more sensitive issue nagged at him.²³

The responsibility for this mess clearly belonged to the CDC, which had opened the original heading, but this fact might easily get lost in a controversy over who was to blame. The Southern Pacific had put itself in the position of being morally and legally responsible for what the CDC had done, bearing out yet again the truism that no good deed goes unpunished. The full implications of this dawned on Harriman when Randolph suggested that he had done all any private citizen could be expected to do in fighting the river. The time had come to enlist government aid in the fight rather than drain the railroad’s treasury. In that dreary December of 1906 the last thing Harriman wanted to do was ask Theodore Roosevelt for help. But Randolph was right: the more Southern Pacific assumed the burden on its own, the more liable it became for whatever went wrong.²⁴

Reluctantly Harriman on December 13 sent Roosevelt a lengthy telegram outlining the situation. “In view of the above,” he concluded, “it does not seem fair that we should be called to do more than join in to help the settlers.” Two days later he got exactly the reply he expected from Roosevelt. “I assume you are planning to continue work immediately on closing break in Colorado River,” the president said blandly. “I should be fully informed as to how far you intend to proceed in the matter.”²⁵

Harriman glowered at the reply, his eyes red with fatigue and his body clenched in pain from hemorrhoids that required surgery again. To the growing list of grievances between himself and Roosevelt had been added a runaway river, a catastrophe that could easily be blamed on him by one so practiced at the art as the president. Harriman had no choice now but to play the controversy through, and to his dejected spirit it was difficult to know which posed the gravest threat: the river or the president.