The entry says that the rock fill, which nobody had proposed, closed the river twice. This piece goes through the preparation and the first closing as Cory set them out for the American Society of Civil Engineers in 1913, with the quantities from his tables; the companion “Fifteen days” takes the second closing and what the two of them taught. His paper is the only full account by someone in charge.
Getting there
When the summer flood of 1906 receded, the break opposite the wooden gate had widened from 600 feet “to more than ½ mile, and necessitated work on a far larger scale than had ever been suggested.” Cory’s first decision was transport. The opinions against rock notwithstanding, “the ability to get rock in large quantities and rapidly seemed to the writer to be so essential, and it was so obvious that much better transportation facilities were required,” that a branch railway was built from the Southern Pacific’s main line at a point seven miles west of Yuma, named Hanlon’s Junction, to the break. The surveyed line of the Inter-California Railroad, a Mexican subsidiary of the Southern Pacific whose construction the flood had stopped, ran a few hundred yards from the wooden gate, and its alignment was shifted to cross the Alamo where the diversion dam would best go. The stretch from the junction to the company’s land lay in the Yuma Indian Reservation, where a right of way would take time to get; “As it was feared that special permission might not be quickly obtainable, nothing whatever was said, but the line was simply built,” and permission was asked for afterwards, “and in due course was obtained.” Work began on 1 July; the first train of materials reached the Lower Heading on 15 August.
Rock came first from a quarry opened on the granite point at Pilot Knob where the concrete gate stood, on the land bought from Hall Hanlon at the very beginning; before the first closing was done it had a face 600 feet long and 40 feet high. A clay pit was opened between the quarry and the boundary; its clay was hard, needed blasting, “but it melts down in water,” and mixed with the cementing gravel of the railway’s Mammoth gravel pit, forty-one miles west on the main line, it made an impervious fill. Beyond these were four quarries at a distance: Declez, 195 miles west, which supplied the San Pedro breakwater with granite in twelve-ton blocks; a field of lava rock near Ogilby; an abandoned pit at Tacna, 52 miles east; and Patagonia, 370 miles east on the branch toward Nogales, which turned out ten-ton limestone. Brush for fascines and mattresses had to come from the south side of the break, with a wagon haul of about a mile, through growths six to eighteen feet high that only Indian labour could cut.
The railway supplied the rest on cost plus ten per cent, with freight at half a cent a ton-mile until the Interstate Commerce Commission’s rules forbade such arrangements at the new year. Two steam shovels went to the quarry and one to the clay pit; work trains were requisitioned until there were ten; a roundhouse foreman and an assistant master car repairer set up shops at Andrade with three carloads of repair parts. From the Lucin cut-off across Great Salt Lake came the steel side-dump cars of forty-five cubic yards, “locally known as ‘battleships,’” weighing about twenty tons and carrying 100,000 pounds with a permitted overload of ten per cent, which were loaded to 125,000 on the short haul from the quarry; eighty at first, about 300 in the end. Everything that came on to the branch line reported to the Southern Pacific’s assistant engineer, Eulogio Carrillo, who was lent to the company as a superintendent, and Carrillo reported to Cory, “and the entire force obeyed instructions as quickly and fully as though there were absolutely no connection between them and the Southern Pacific.”
The camp
Labour was the hardest thing to get. San Francisco’s rebuilding, Los Angeles’s growth, the Harriman lines’ construction on the west coast of Mexico, the Laguna weir and the Roosevelt dam were all hiring. The immigration laws kept Mexicans out of the United States, but the work was in Mexico, so the railway’s labour agent at Tucson was asked for five hundred men from central Mexico, to be shipped in bond from El Paso to Yuma and back across the line. “This plan was an utter failure, for two reasons. The Mexican officials did their best to prevent Mr. Heney’s agents from getting men started, and the 75 men who arrived were unable to stand the climate.” The Indian Service’s outing agent, C. E. Dagenette, then arranged for the tribes, “with the result that, by the time work was in full swing, practically all the men, women, and children of six Indian tribes were on the work: the Pimas, Papagoes, Maricopas, and Yumas, from Arizona; and the Cocopahs and Diegueños, from Mexico. These six tribes fraternized and got along together without any difficulties whatever, and constituted a separate camp of about 2 000 people.” About four hundred workmen could be relied on from it, at twenty cents an hour, with one man’s pay in every nine going to a woman who cooked. They bought their own supplies, and to avoid duty built their camp on the Arizona bank and crossed the dry channel below the break to work. They had to be paid weekly, and very few would work on Sunday or overtime, “regardless of how critical the stage of work may be when the whistle blows.” White labour drifted in from everywhere, drawn by the notoriety, and drifted out again through the saloons of Yuma, which was then “wide open.” A pile-driver foreman got fifty cents an hour, ordinary labour twenty-seven and a half to thirty, for eight to ten hours, less $22.50 a month for board.
Because so many kinds of men were on one job under Mexican law, the company asked the Mexican authorities to put the region under martial law, and a force of rurales under a military commandant policed the camps; liquor was barred, “and there was absolutely no disorder at any time.” Bunk houses were built with brush ramada roofs and screened against mosquitoes. The board was good and plentiful on purpose, “especially with lots of fresh vegetables,” at twenty-five cents a meal on the American side and forty in Mexico. Customs were a daily obstacle: the nearest Mexican custom house was at Mexicali, so manifests went there and back before goods could cross at Algodones, and emergency stores went over under the rule that let each person carry twenty Mexican dollars’ worth a day without a manifest, for which an inspector fined the company $3,000 a year later; the fine was remitted when the facts were explained. Once, when rock was needed at sundown and the officials at the line would not let trains pass after dark, their protests were disregarded, “for while the officials under the circumstances could not act otherwise, it would have been folly not to have disregarded their orders, considering the urgency of the matter.” After the work President Díaz recommended, and the Mexican Congress voted, a refund of about three quarters of all duties paid, more than $40,000.
The mattress
Work at the break began on 6 August, when the river had fallen to 24,500 second-feet and had left sand bars on both sides of the channel beside the wooden gate. Teams threw up embankments across the bars on the line of the diversion dam, brush jetties and the dredge Beta narrowed the stream to 600 feet, and the weaving began. The barge Silas J. Lewis, 116 feet long, had its decks cleared and skids rigged; steel cables an inch thick, eight feet apart, were anchored to dead men in the north bank and paid out from spools under the skids; on them, between vertical pins, the men built fascines of willow eighteen inches thick and a hundred feet long, bound with baling wire, and sewed each to the last and to the cables with galvanised cable and clamps. “When a length of mattress equal to the width of the barge was completed, the barge was slowly pulled from under it, and it caught the silt and at once settled heavily to the bottom. No kind of weighting whatsoever was required.” Two shifts wove and sank two mattresses, one on the other, 1,300 feet in all, in twenty working days, sixty-five feet of mattress a day. Several engineers with long experience of the river urged a canvas back sewn under it against washing; Cory thought that unnecessary. One layer was sunk in a curve twenty feet out of line through careless handling of the anchor lines, “This, however, was not important.”
A four-pile railway trestle with ten-foot bents went across on the centre line of the mattress, driven from both ends, and was ready for trains on 14 September. Brush fascines were laid between its bents, longways with the stream, and sunk with rock from Andrade; the battleships were loaded by steam shovel, hauled down and dumped from the trestle, “In this way a difference of 6 ft. in water elevation above and below this diversion dam was attained with no difficulty whatsoever.” Meanwhile the dredges cut the by-pass to the wooden gate, the steamer Searchlight sat in it for three days with its wheel going to hurry the erosion, and dynamite was used in the adobe. By 10 October only about 1,450 of the river’s 14,300 second-feet were not going through the gate.
The first dam
The gate failed on the eleventh. The entry tells it; what matters here is what the failure showed. “The lowering of the water above the diversion dam left it dry, except for a surprisingly small quantity of leakage, and enabled examination of the rock fill which had been produced by an ever-increasing proportion of rock with respect to brush. This condition of affairs seemed to indicate that the reasons urged why a rock fill dam of considerable height could not be built in a running stream were not altogether strong, and suggested the possibility of very quickly controlling the situation with a series of rock fill dams, each of which should sustain a head of not more than 4 ft.” As insurance, a rush order went by wire for more sewing cable, in case a diversion dam had to be built opposite the concrete gate instead. The trestles across the by-pass were repaired, a third was thrown across the stream fifty feet above the lower one, where the whole river now ran through an opening 130 feet wide, and the quarry, working night and day, produced about 5,000 cubic yards a day. The three dams across the by-pass went up so fast that by 29 October only ten per cent of the water was going through it; most of the rest, 8,600 second-feet, was going over the diversion dam on its mattress. There a second trestle, thirty feet upstream and parallel, had been rushed across, and from the two a rock fill dam turned the whole river, 9,270 second-feet, down the old channel on 4 November.
Leakage was stopped by dumping battleship loads of gravel and clay and puddling them with five fire nozzles throwing 225 gallons a minute each, which also flattened the slopes to about two and a half to one on the river side and two to one on the land side. The tracks over the trestles were raised as the fill grew, so fast that no attempt was made to recover the stringers or caps, and the top was carried to elevation 124, eight feet above the closing level. The daily record of carloads from each trestle was lost, “and the totals obviously signify nothing,” so Cory gave the quantities for the whole period from 1 August to 4 November instead: 2,200 cords of brush and forty miles of steel cable; 3,800 feet of trestle, 15,200 feet of eight-by-seventeen-inch Oregon pine stringers and 1,100 piles; 1,690 cars, 50,000 cubic yards, of rock, ninety per cent of it used between 11 October and 4 November; 841 cars, 32,000 yards, of gravel; 808 cars, 33,000 yards, of clay; 200,000 cubic yards of earth by teams and 200,000 by dredges; 200 to 500 head of mules and horses from July to 20 November; 200 men in June, rising to 1,000 on the day of closing. The river’s discharge had been 99,200 second-feet on 27 June, 24,400 when the channel work began on 6 August and 9,275 on the day it closed. The head on the dam at closing was 15.8 feet, and a week later 16.75. In its final form the Hind dam, named for Thomas Hind, superintendent at the Lower Heading from June 1906, has 400 feet of curve at the north end and 2,275 feet of tangent, with a rock core on brush mattress under its highest 600 feet; about eighty per cent of it was complete when the second break came on 7 December.
The levees were another matter. The plan called for a line of embankment from the concrete gate down to the dam and five or six miles south of it, eight feet wide on top, with its slopes two and a half to one on the river side and two to one on the land side, and its grade about six feet above the old high-water mark, and, deliberately, two and a half feet higher than the Reclamation Service’s levee opposite, “because, should the latter break, the damage would be far less than if the levees on the west side were to fail, with re-diversion of the river to the Salton Sea.” The ground was cleared and grubbed, but no muck ditch was dug. The desirability of one was fully realised and it was part of the design; experience in the valley was that cracked adobe leaked badly when water first came against it and then soaked tight, and the Reclamation Service’s levees opposite had been built without ditching too. “On the other hand, the money supplied by the Southern Pacific Company was for closing the break, and only for that purpose, until the re-diversion of the river was assured.” The borrow pits were taken from the land side, against usual practice, to leave the brush between levee and river to break the current. By 6 December about a mile and a half of levee had been completed above and below the dam, five miles more were under construction, and the ground was cleared for two miles beyond. The river came the next night. After it, every levee that stood was reconstructed with a muck ditch down its middle, and every new one built with one from the start, and in the flood of 1907, which put the water against them for their whole length, “in no case was any weakness apparent.”