WATER SUPPLY. 355 WATER SUPPLY. Springs may be developed and sometimes havo their How inereased liy digging a basin or well aiound tliem, or by driving a tunnel or gallery into a hillside, liy either of these means the How of a number of springs may often be united. The walling up and covering in of springs is fre- ((Uently desirable to exclude foreign matter of all sorts and to keep the water cool. Underground waters other tluin springs are developed by sinking wells, building inliltra- tion galleries, driving tunnels, and, occasionally by the construction of submerged dams. The latter have been employed to intercept per- colatirg waters flowing through valle.vs in the West, chiefly for use in irrigation. Well-sink- ing, dam-building and tunnel-driving are de- scribed under their projier heads. Infiltration galleries are generally formed by digging a trench and walling up its sides with timber, brick, or stone, laid with open joints. The top is geu,?rally tightly covered and the bottom left open. works and irrigation are described under those beads, and the various classes of pumps are treated under PUMi's ANu Pumping MAcili.NEny. Estimates of the proliable yields of drainage areas should be made only after due considera- tion of all the factors reviewed above. This is particularly true of underground supplies. It is a rare thing to find a single group of wells that yield, year after year, more than ."j.OOO,- 01)0 to Kl.OOO.OOO gallons a day, hut .sonjetimes a number of groups may be made tributary to one water-supply system 'liy proper pipe connections and pumping plants. The yield of surface supplies, as has been stated, is far better known than that of under- ground waters. The accompanying table, taken from Turneaure's Public ll'o/cr fiuppHca, gives the drainage area and yields of 1.5 streams in the United States, differing widely in size and location. The records, as will be seen, are mostly for quite long periods and include the average yearly, minimum yearly, and average half-yearly Sudbury Cochituate Mystic Connecticut Croton Upper Hudson Genesee Passaic Perkiomen Tohiclion Nesliomiuy Potomac Savannali Des Plaines Dpper Mississippi 75.2 18.87 26.0 1023.4 338.0 450.0 106.0 82.2 15.2 102.2 139.3 110.43 729.4 63.0 326.5 Average yearly Rain Indies 1875-97 1863-96 187S-961 1871-85 1870-94 1888-96 1894-96 1877-93 1884-97 1884-97 1884^-97 1886-91 1884-91 f 1889
1889-95
1885-99 45.77 47.08 43.79 44.69 48.38 39.70 39.82 47.08 47.98 50.17 47.88 45.47 45.41 30.66 26.57 Flow inches 22 . 22 20^33 19.96 25.25 24.57 23.36 12.95 25.44 23.62 28.43 23.24 24.03 22.19 6.61 4.90 Per cent. 48.6 43.2 45.6 56.5 50.3 59.0 32.5 64.0 49.2 60.7 48.5 62.7 48.9 21.6 18.4 Tear of minimum flow Rain inches 37.23 31.20 31.22 40.02 38.. 52 33.49 31.00 35.64 38.67 38.34 36.30 37.03 43.10 32.38 22.86 Flow inches 11.19 9.76 9.32 18.25 14.54 17.40 B.67 15,23 15.66 18.75 16.19 14.. 50 10.26 3.19 1.62 Per cent. 34.1 31.3 29.3 45.6 37.8 63.2 21.5 43.7 40.4 49.0 44.3 39.2 37.7 9.9 7.1 Average for December-May Rain inches 22.98 22.97 22.11 20.13 23.39 18.20 19.58 22.47 23.27 24.28 23.04 22.13 21.51 13.01 Flow inches 17.52 14.37 15.12 17.96 17.81 16.23 10.20 18.22 16.52 20.42 17.05 16.27 13.03 Averasre for June-Xovember Per Rain riow| Per cent, inches indies' cent. 76.0 64.7 68.4 89.1 76.1 S9.0 52.2 81.1 70.8 83.3 74.0 73.7 60.6 22.61 24.10 21.06 24.. 56 24.99 21.05 20.24 24.39 24.71 26.12 24.44 23.34 23.90 16.30 4.711 5.46 4.34 7.38 6,76 7.13 2.75 7.19 6.93 8.00 6.10 7.77 9.16 1.14 20.8 22.7 22.4 29.7 27.0 33.8 13.6 29.5 28.3 30.6 25.0 33.0 38.4 7.0 The water of running streams is sometimes diverted to the canal, head-race, or intake pipe by means of a dam, or in the case of a large stream by a wing dam, extending only part way across. Intake pipes are often laid on the bed of large streams or lakes luitil a point of sufficient depth or remoteness from the shore is reached, where it terminates in a crib or other arrangement for the jirotection of the exposed end and the strainer frequently placed upon it. Beneath rivers ;ind for a number of the cities on the Great Lakes, the water sup- ply is sometimes drawn through tunnels. These terminate at the .^hore end in the pumping sta- rainfall, run-off, and percentage of run-off to rainfall. A fall of 1 inch of rain on a square mile of area is equal to 2,323,200 cubic feet, or about 17.375,000 gallons. A total flow of 1 cubic foot per second is equal to 046,300 gallons a day. A more detailed study of the records would shov that after making allowance for evapora- tion from water surfaces, there would be a negative yield in some months; or, in other words, the river would run dry were it not for artificial or natural storage. It is to meet just such contingencies that storage reservoirs are provided. There are two ways of considering storage: (1) The amount required to supple- .iveraffe .rearly rainfall in inche.«. three driest years in succession Square miles required to yield 1.000.000 gallons, in severe drouplit Maximum storatre room ad- visable (1-1,000,000 g-allonsl CITY Per square mile tiibutary drainage Per 1,000.000 gallons actually delivered 3$ 44 40 35 30 29 17 2..50 1.70 3,00 5.00 7.00 6.25 9.00 260 205 280 2.50 20O 225 150 650 New York San Francisco No. 1 2 • ■• 3 349 840 1.250 1.400 1,406 1.350 ■• 4 Oakland tion and at the river or lake end in a vertical shaft, protected above by a timber and stone crib and tower. Pumping plants are often important parts of water-supply systems. Their relations to water- ment the natural flow so as to give the required daily supply; (2) the economic limit of the storaga development. The deficiency is then to be made good by establishing the storage capa- city indicated, provided the expense involved is
