One cubic foot of water equals 7.48 gallons. On August 9, 1906, 300 million cubic feet of water passed through a ruptured irrigation intake, into the Imperial Valley, every hour. The Colorado River changed its course a year before. From a point just inside the Mexican border, it ran through a half-mile-wide crevasse, flooding towns and farms and creating the Salton Sea. Five attempts to close the intake — known as “Rockwood’s Gate” — had failed.
Harry Thomas Cory. He ordered trains, battleships, and flatcars awaiting his signal at every major quarry in the Western United States.
“The situation was so new,” writes Margaret Romer, “that engineers had nothing upon which to base their opinions. They all disagreed. About the only point they agreed on was that something had to be done at once.”
How do you lay siege to a river? That charge fell to Harry Thomas Cory. He’s the star of two novels — Harold Bell Wright’s Winning of Barbara Worth and The River by Ednah Aiken — and Ronald Coleman plays him in the movie Barbara Worth. Yet his credentials don’t suggest heroics. A former professor of civil engineering, Cory served as advisor to E.H. Harriman, railroad magnate, and as assistant to Epes Randolph of the beleaguered California Development Company. Cory was modest, bookish, precise (his Imperial Valley and Salton Sink ranks among the most thorough books ever written). Nothing in his past indicated he could lead like a general and save Imperial Valley from inundation. Though Cory was an elegantly dressed, soft-spoken Midwesterner, a valley local said Cory “savvys the country like a burro does the cook shack.”
In the summer of 1906, Cory made two decisions. First, he rejected previous methods of dam building: pilings, sandbags, silt levees. The river bullied them away. Plus, the soil was so soft, when a dredge ran aground on a sand bar it sank.
To prevent rock from sinking or tumbling downstream, Cory ordered vast amounts dropped on the crevasse day and night. “The ability to get rock in large quantities and rapidly seemed so essential,” he wrote, “and it was obvious that better transportation facilities were required.”
In previous attempts to close the intake, crews floated materials on barges. Cory’s second decision: he ordered a branch of the Southern Pacific railroad to run from the mainline, seven miles west of Yuma, to the break. Trains would haul materials to within a few hundred yards at first, then on a trestle across the river. Construction began July 1. On August 15 the first train reached the site.
“The shape of the channel at this point,” writes David O.Woodbury, “was like a great foot, with its toes to the west. Just beyond the bend of the ankle, sand bars had constricted the passage to an instep 600 feet wide. Beyond, the river flared out again onto the delta. It was at this narrowest point that Cory planned to build his dam.”
Cory inherited an arsenal of onsite equipment: two clam-shell dredges, two barges, pile drivers, the 91-foot steamer Searchlight, and the 75-foot St. Valliers. Using funds from the Southern Pacific, he added worktrains, plows, Fresno scrapers, and steel side-dump railroad cars. Also known as “battleships,” the dump-cars weighed 20 tons and carried 100,000 pounds. They “were frequently loaded to 125,000 lb., as the trip between the Andrade quarry and the break did not exceed 4 miles.” In August, Cory conscripted 80 battleships. By January of 1907, he had almost 300.
Hired hands posed another problem. The San Francisco earthquake — April, 18, 1906 — drew thousands of usually reliable “hobo laborers” north. The Imperial Valley climate that, writes Aiken,“dulls vanity and wilts collars,”also took a toll.
With government aid, Cory hired 400 Indians from the “Pimas, Papagoes, Maricopas, and Yumas from Arizona, and Cocopahs and Diegueños from Mexico.” They and their families lived in a tent camp on the Arizona side of the old channel. A worker earned 20 cents an hour. Every group of nine received an extra day’s pay for a squaw to cook meals. Their first task: construct a 13,000- foot, double-strength, woven brush mattress as a foundation for the rock fill. For 20 days, they worked 11-hour shifts, around the clock, producing 80 feet of matting per shift.
Brush-cutting, chopping into gnarly mesquite and arrow weed, makes even severe conceptions of hell look pattycake. The temperature is 110 to 125, with no wisp of wind. Sunlight blinds. Mosquitoes swarm. The mind plays tricks. On more than one occasion, a cutter stared down at the thick stalks assembled at the bottom of a cliff, imagined they were water, and dove to his death.
Cory also hired “floating labor,” people drawn to projects receiving national attention. Few stayed for long. “A work train ran into Yuma every night for provisions and supplies, returning early in the morning, and it always carried a considerable number of cheerful capitalists out and sadder and wiser men in.”
The site was a Babel of shouted orders, whistles, and curses, though workers heeded one cry most of all: “The RIVER!” This meant a new surge of reddish water — three, five, even six feet above the surface — rushed their way. The worst wasn’t the water but the snarls of debris (called “drift”) it carried. Most threatened areas: the pile driver hammering 90-foot posts into the crevasse; and a barge, moored to a pile, where crewmen wove the brush mattress around thick steel cables.
Because no one could see what was coming, the cry became most fearsome at night. So Cory suspended large acetylene lamps across the channel and created a city’s worth of illumination in the midst of nowhere.
Contrary to many engineers, who claimed it would sink in the quicksand-like river bottom, the brush mattress held fast. Then workers manned pile drivers and built a ten-foot-wide railroad trestle across the break. The first battleship dumped its load. A single-file armada followed, hauling a mountain’s worth of rock, gravel, and clay.
Two weeks later, the massive wall became visible on the surface. But workers had no time for elation. Drift, from pebbles to tree trunks, gathered near the headgate and began to scour and sidecut the banks. Barges dropped rock fill, but not fast enough. The river rolled it away. On October 5,Cory ordered a second pile/trestle constructed a few yards from the gate. On October 11,at 11:00 a.m., the first train edged onto freshly laid tracks. But, writes Woodbury, “the hastily built trestle could not stand the strain. It buckled in the middle and half the train fell in. Heavy timbers snapped off and charged down upon the gate.”
At 2:30 that afternoon, battered by drift and sheer hydraulic force, the giant headgate rose from its foundation, made a splintering screech, and cracked, the top two-thirds tumbling downriver, toward the original trestle.
A dump train stood at the south end, its battleships still loaded with rock. The engineer, seeing the wooden headgate rush toward him like a wounded whale, gunned the throttle. The train lurched across creaking tracks, reaching the north side just before pilings gave out, a large section of the dam caved in, and the river carved a new channel to the desert.
Looking back on his decisions, Cory wrote in 1915: “As in many irrigation and other projects in the west, the garment had to be cut according to the cloth. The sum total of events resulted in carrying out the project along lines which were far from ideal.” Cory’s next move, as he and Epes Randolph inspected the damage, was a leap of faith. “Let’s quit fooling with gates,” he’s quoted as shouting, “what this feller needs is rock, and more rock, and more rock” — and not even a brush mattress beneath them.
To force the Colorado back to its original channel, Cory repaired the original trestle dam and ordered a second barrier-dam trestle built beside it, from both shores at once. Trains loaded with stone came from as far as 400 miles. They had right-of-way even over passenger locomotives. At the site, they waited for their turn on side-spurs. By late October, trains on both trestles dumped into the gap between them, around the clock. On October 29, 90 percent of the water flowed down the old Colorado channel to the Gulf of Mexico. The end was near. After battleships continued their barrage for the next six days, a majority of the workers packed up, elated, and went home.
Then came a cry: “The RIVER!”
On December 5, a flood flashed down the Gila and fused with the Colorado in a reddish-brown maelstrom. The river rose from 9000-second feet to 45,000 and slammed into Cory’s trestles and nearby levees.
Three miles south of the site, a crew was grading the original channel. The steamer Searchlight went to save them. But the west-side levee crumbled in three places. The Colorado rediverted itself —and beached the Searchlight. The men walked back up the channel, to their camp, on dry land.
Remi Nadeau: “This time the breakthrough proved to most observers that the Colorado had, in the course of centuries, reached the stage of leaving the gulf once more and swinging north into the dead sea its delta had created. The inevitable process had merely been hastened by Rockwood’s original Mexican cut. For the first time the engineers realized the full magnitude of the geological forces they had been fighting.”
Cory had another problem: his crews, around 1100 men and their equipment, had gone. Also, efforts thus far cost $1.5 million, and the work would require at least that much more. And the townspeople of Calexico and other settlements were storming mad. They’d been promised a steady supply of water, not the biblical Flood.
In what became the “battle of the telegrams,” E.H. Harriman, president of Southern Pacific Railroad and funder of earlier efforts, and President Theodore Roosevelt haggled about who should finance the operation. Roosevelt called Harriman “an undesirable citizen” because he wouldn’t continue support. More telegrams. When Roosevelt eventually promised congressional action, Harriman shouted at his staff, “Turn the river at all costs!”
During the days it took the two men to agree, Cory ordered trains, battleships, and flatcars awaiting his signal at every major quarry in the Western United States. Each train would have only a “five-minute headway” before the next would begin. When he got the word, on December 20, 1906, Cory wired, “Go!”
They had two months until flood season, no time for planning. Woodberry: “Corners were cut everywhere; this was no construction project but a rescue in which 1500 men were to work night and day.” To fill a crevasse 1100 feet wide and 40 feet deep, the trains came in hordes. They used battleships, but Cory preferred flatcars, since they could haul much larger rocks. To unload them, workers split boulders with “pop-shots” of dynamite.
Four times they built trestles across the crevasse. But each time, when near completion, a freshet pushing tons of drift demolished their efforts. On January 27 at 5:00 p.m., the first trestle reached the southern shore. That night, 145 cars dumped rock into the crevasse. Workers began a second trestle. Many said the roar of the river, train whistles, shouts, fountains of spray when boulders broke the surface were so loud, one barely heard the dynamite.
On February 2, a flash flood hit the first trestle, taking out three pilings. Drift ruptured the second. The river leapt into the gap. But most of the second trestle held — a minor victory. Encouraged, for the next 192 hours, an army of exhausted workers battled through warm desert days and cold winter nights. They rebuilt the trestles, closed the crevasse, and rerouted the river. On the morning of February 11, 1907, they won.
Nadeau: “The mighty Colorado was bent on revisiting the Salton Sink and filling it to its brim — a process it had repeated at intervals through past ages. Undoubtedly the greatest geological change in the world’s recorded history had been frustrated here by the hand of man.”
- Aiken, Ednah, The River (Grosset & Dunlap, 1914)
- Cory, H.T., The Imperial Valley and The Salton Sink (Newbegin, 1915)
- deBuys, William, Salt Dreams: Land & Water in Low-Down California (University of New Mexico Press, 1999)
- Nadeau, Remi, The Water Seekers (Crest Publishers, fourth edition, 1997)
- Romer, Margaret, “A History of Calexico,” Southern California Quarterly 12 (1922), 26–66.
- Woodbury, David O., The Colorado Conquest (Dodd, Mead & Company), 1941