40

the dam. The river, as shown by fig. 1, for 200 or 300 miles, flows through a hilly country, from above Colorado City, on the Texas and Pacific Railroad, and in its course absorbs the waters of the Concho, the San Saba, the Llano, and. the Pecan Bayou. All of the country drained by these tributaries is hilly, with the exception of a few miles along the head of the Colorado and the Concho.

When the break occurred the silt in the immediate vicinity of the dam would have flowed out had there been no water. Just above the part of the dam that gave way was a plateau whose surface was on an average about 18 feet below the crest of the dam. At the time of the break the lake level was 11 feet above the crest of the dam, making the depth of water on the plateau 29 feet. The torrent poured over this plateau with immense velocity, as shown in Pls. X and XI. The silt on the plateau was cut away with such swiftness that in three hours it was swept almost clean. Only a slight amount remained, as can be seen from fig. 8; the men are standing on the original soil of the plateau, the silt appearing behind them.

In the main channel the upper surface of the silt was 38 feet below the crest of the dam, giving, at the time of the break, a depth of water in the main channel of 49 feet. After the water level dropped below the plateau the current was confined to a narrow gorge. The flood continued in the main channel for several days. A week after the failure the silt along the shores of the former lake was cut into fantastic shapes by the currents of the river and those of many mountain gorges. The silt in contact with the dam undoubtedly increased the pressure against it, but that portion of the dam across the main channel where the silt was 28 feet deep and where the pressure was greatest was on a good hard rock foundation and successfully resisted it. There was practically no waterlogged drift in this silt; the soundings indicated mud bottom. The silt deposit kept the river at Austin muddy for months after the failure. (It is still muddy-November 1, 1900.)

Pl. VIII, B, is a view of the lake from Mount Bonnel, looking up the river, two weeks after the break, with the freshet still on. The extreme right-hand part of the curve is the mouth of Bull Creek. For purposes of comparison, the silting up of reservoirs is best reduced to heights and depths on a square-mile base. To derive an expression for the amount of silt deposited in a given time, let x equal depth, in feet, of silt deposited in a year by each foot of water in the reservoir, and let h equal the original depth of the reservoir. The depths of water are at the end of one, two, three, etc. n years:

• First year, h - h x= h (1- x);
• Second year, h (1 - x) -h(1-x)x=h(1-x)²
• Third year, h(1-x)²2-h(1-x)2x=h(1-x)³
• Fourth year, h(1-x)³-h(1-x)³x=h(1-x)4.

Hence, if we let d=depth of water in it years, we have d=h (1-x)n. In 1897 h=81, d=50, n=4; .. x=0.1135.

In 1900 we have d=0.52h, n=6/23, (1-x)/203, Q=0.52; ." .x=0.09343.