about 32 cubic feet a second (1,250 miner’s inches), which could be diverted from Texas Creek at a point 480 feet vertical above the Bloomfield flume.  An aqueduct about 4,000 feet long, partly of ditch and partly of flume, was needed to bring the water from the catchment dam on the creek to the brow of the gorge.  The vertical head for the pipe could therefore be from a maximum of 460 feet down to any lesser head; with a head of 460 feet, the pipe would be 4,790 feet long; and with a head of 220 feet, the length would be 4,290 feet.  Assuming a maximum tensile strain upon the iron of 16,500 pounds per square inch, with the formula for the greatest head of about

d = (.359 l/h)^{1/5}, [or, v = 68 (dh/l)^{1/2}, and Q = 32],

and a lower value of the coefficient in the last equation for the lesser heads, it was found, by calculation, that the least cost could be obtained with a head from 300 to 350 feet.  The head fixed upon was 303.6 feet, with a length of 4,438.7 feet.  A profile of the pipe, with nearly the same horizontal and vertical scales (horizontal scale, showing slope lengths), is given in Fig. 14; details are given in Figs. 15 and 16.  The pipe was of double riveted sheet iron, made in lengths of about 20 feet, and of the following thicknesses: 

1,349 linear feet,    0.083 inch thick.
220    "            0.095     "
240    "            0.109     "
250    "            0.120     "
320    "            0.134     "
610    "            0.148     "
1,450    "            0.165     "

Some of the iron was of the very poorest quality; the pipe was made by contract in San Francisco, without the supervision of an inspector, as the contractors were a firm of good reputation; the bad quality of the iron was not detected until too late to have it corrected.  Since then, the writer has always had such pipes—­the mines of which he has been the manager using large quantities—­made directly on the ground where they are to be used; the pipe makers, in the latter case, always reject such sheets as are too much below in thickness the standard gauge, and those which show in passing through the rolls the bad quality of iron; tests of each joint by hydrostatic pressure would add too much to the cost.

[Illustration:  FIG. 16.]

The maximum tensile strain upon each of the seven thicknesses of iron used was intended to be 16,500 pounds per square inch.  Some of the sheets were below the standard gauge, so that, in reality, the tensile strain is sometimes as high as 18,000 pounds.  The mean diameter of the pipe was 1.416 feet.  The entrance into the pen-stock was tapered, so that the coefficient of contraction was about 0.92.  For pressures not exceeding say 380 feet, the joints were put together stove-pipe fashion.  For greater pressures, the joints were made by an inner sleeve riveted on one end of the joint, with an outer lap-welded band, as shown by Fig. 15; lead was run into the space between the outer band and the pipe, and then tightly