The companion “Battleships” goes through the preparation and the first closing of November 1906. This piece takes the second, after the river had broken through the levee a quarter of a mile south of the new dam on the night of 7 December, and ends with what Cory, Newell of the Reclamation Service, and the engineers who discussed Cory’s paper made of it all.

The second closing

The second break was Clarke’s. Thomas Hind went to the levees, and Clarence Clarke, resident engineer of the Southern Pacific’s Tucson division, came on 20 December as superintendent of the closing. From that date the Mexican company was paid directly by “Epes Randolph, Agent S. P. Co.,” and the development company had nothing to do with the work. There was no mattress this time: “experience during the first closing left little doubt that there would be any particular difficulty in making the second closing without any brush mattress foundation. Even had such been deemed desirable and worth the delay in time, very little brush was available nearby because of the large quantity used for the first closing. It was felt that the work was standardized and consisted in throwing ordinary railroad trestles across the break and from them making rock fill dams in series.”

The current struck where the south end of the Hind dam had been and was deflected sharply, and few who saw the erosion there “believed it possible to hold this erosion from going entirely through the structure,” but immense quantities of riprap held it, with the water in front cutting to thirty-eight feet. Two trestles were started, with five pile-drivers, one at each end of each and a floating machine in midstream; they had to curve upstream to meet the levee on the south, so the ninety-foot piles were driven in a strong quartering current, two complete driving outfits were kept in reserve against the drivers overturning, and two boats waited below to pick up any man thrown overboard. On 28 December, with one trestle nearly done, a flood with unusual drift tore out about a third of it. “Three times this occurred, resulting in the loss of a large quantity of bridge material.” The telegram from Los Angeles on 14 January about the exhaustion of piles in Southern California is in the entry. On 20 January the first trestle was finished for the fourth time, with the stringers in place and the track two thirds laid; seventeen bents of the second remained; 176 battleships of rock and 100 flat cars of big rock stood loaded on the sidings; a mile of levee had been rebuilt at each end of the dam; 1,100 men and 1,000 head of stock were at Andrade and on the river.

“On January 27th, the lower trestle was completed, and dumping rock from it began at 5 P. M. By daylight 145 ‘battleships,’ containing 6 600 cu. yd. of rock had been unloaded as another flood from the Gila began arriving, but it caused little trouble. On February 10th at 11 P. M. the second break was closed, and all the water was again going down the old channel.” The river’s discharge had been 12,500 second-feet when the work began on 20 December, 13,800 when the dumping began, and about 20,000 on the day it closed. The materials of the second closing were 4,000 feet of trestle, of which 1,800 feet was carried away before either trestle was complete or any rock dumped, “When rock dumping began, no more trestle was lost”; 16,000 feet of stringers; 1,200 piles; 46,000 cubic yards of earth by teams, making 960 feet of earth dam to the levee; and 2,357 carloads, 66,000 cubic yards, of rock before the water was shut off, with 221 cars of gravel and 203 of clay. Cory’s Table 14 gives the rate: between 27 January and 2 February, 606 battleships and 601 flat cars of rock; between the third and the ninth, 439 and 91; on the tenth, to midnight, 92 and 285. After the water was shut off, 956 carloads of clay and 873 of gravel finished the Clarke dam, which was made impervious by dumping alone, without puddling, and whose stringers were all taken out and whose top and sides were finished with the gravel spreader from the levee work. The clay pit closed on 3 March and the Hind-Clarke dam was finished, but for its surfacing, on 16 March. In the whole of it, from the second break to the finished dam, 1,517 battleships, 1,240 flat cars, 956 cars of clay and 1,052 of gravel were used, and the railway’s equipment on the river between December and July came to 82,485 car-days.

Newell watched. “The stones used were as large as could be handled or pushed from the flat cars by a gang of men or by as many men as could get around a stone. In some cases the pieces were so large that it was necessary to break them by what are called ‘pop-shots’ of dynamite laid upon the stone while it rested on the cars.” He heard in them “artillery in action,” and “the whistle signals, the orders to the men, and the bustle of an army working day and night to keep ahead of the rapid cutting of the stream.” He also recorded a thing Cory does not: that trains of flat cars came from quarries on the Santa Fe and the San Pedro road as well as the Southern Pacific, “and in fact, it is stated that even high-grade cut stone en route for building was requisitioned and diverted to fill the need.” Cory, before the House committee a year later, said that for three weeks two divisions of the railway, about twelve hundred miles of main line, “were practically tied up because of our demands for equipment and facilities,” that about three thousand flat cars of rock had been brought in from the far quarries, and that “It was simply a case of putting rock into that break faster than the river could take it away.”

What the river taught

Cory called what the two closings showed “entirely unique and in some respects unexpected.” First, that the brush mattress “is not only unnecessary, but adds to the cost, both in time and money, provided rock is thrown in at a reasonably rapid rate.” The current was expected to carry small rock away indefinitely and the big rock to sink into the silt “to indeterminate but very great depths”; in fact “a relatively small quantity of ‘battleship’ rock sufficed to blanket the bottom of the stream with a mattress of rock which fulfilled essentially the same function as a mattress of brush.” Rock of all sizes was best, as a steam shovel loads it; stones of one to eight tons, which only a derrick can put on a flat car, were useful but not necessary for raising a low place in a hurry, and were unloaded by crowds of men with pinch-bars, the cars chained to the stringers against upsetting. “One man was killed during each closing by going overboard with a large rock, and these were the only serious accidents. No equipment was lost during either closing.”

Second, that the water going over the dam while it was built did not breach it. “Once, in the building of each of these structures, a local settlement occurred, resulting of course, in large quantities of water going over the relatively short lowered section.” The engineers had held that this would breach the barrier hopelessly; “In point of fact, continued dumping of even small rock soon stops such so-called breaks, although that results in the waste of much material.” The real nuisance was drift caught by the trestle, which theory said the rock would break up and carry down and practice said threatened the trestle; a sharp kink in one of Cory’s photographs is where drift pushed the whole structure out of line, and two or three bents of the upper trestle were lost to it. Two trestles were used in both closings, thirty feet apart in the first and fifty in the second, so that the rock mattress would extend far enough upstream; after the first closing it looked a needless precaution for heads of fourteen feet or less, and in the second, “Everything worked so well that it seemed safe to do a little experimenting, and, practically speaking, the second break was closed from the lower trestle alone.” The rock fill under the trestle was 1,125 feet long and the river was let pour over 680 feet of it; the most that went over the crest was 27,000 second-feet; settlements were handled with about a hundred second-feet going over each foot of dam. “The writer would say that experience along the Colorado River would not justify using the rock fill barrier dam method for quantities in excess of half that flow, on an average, although possibly much greater rates of discharge over such a structure might be safe.”

He set the Laguna weir beside it. In December 1906 the Reclamation Service’s contractors closed the central gap of the government’s weir above Yuma by the same method, two trestles 400 feet apart, rock fills from each, the whole river forced through the sluiceways at the ends, about 600 men in two ten-hour shifts, no lives and no equipment lost, “and the time required after beginning to dump rock was about 2 weeks!” The cost of the rock in place there, with everything charged, was $1.04 a cubic yard. Of the men who had said the thing could not be done, Cory said only that the method “had not been advocated, or indeed seriously considered, by a single man.”

The dispute that followed in the Society’s discussion was about what the method meant. Sellew, the Reclamation Service’s engineer on the Yuma project, who had been “an innocent bystander while some of the work described was in progress,” honoured the fighters and entered his protest “when the closing of this crevasse, which, from the nature of things, had to be accomplished by rough and ready methods, is heralded as a solution of all engineering problems pertaining to the Lower Colorado.” The real problem, he said, was bank caving, and the real question was storage at the headwaters. Clarke, “being one of the men referred to by Mr. Sellew as ‘red-blooded fighters,’” wanted it on record that the crevasse “was closed by railroad men, who used methods with which they were familiar; they tried no experiments and devised no new engineering practice. They knew they were not whipped at any stage of the game, and they kept at the job because they had done the same thing on a smaller scale many times and were confident of ultimate success.” The problem of the lower Colorado, he said, had three parts, the direction of the river’s flow, the exclusion of its floods by levees, and the safe diversion of water into canals, and none of them called “for the invention of new methods of river control”; a railway track on top of every levee, with gravel and rock on cars behind it, was the answer to the bank, as the saving of 1,400 feet of levee north of the Clarke dam in January 1907 had shown, when a pile jetty ninety feet long built a bar in a Gila freshet in eight days. The two men agreed, as Cory noted in closing, that the track on the levee was indispensable, and disagreed about almost everything else, and the Society printed both.