weir with end contractions and of the same crest length, and (iii) that the flow over rectangular weirs with end contractions varies as L1-02 and H1-47, being given by Q = 3.10 L1·02 H147, where Q is the discharge in cubic feet per second, L the length in feet, and H the head in feet. This formula and others given later are believed to give the flow within ± 1 per cent. Measurement of Water.-An extensive series of experiments on fully-contracted rectangular and Cippoletti1 (i.e., end slopes of 4 to 1) weirs of lengths 6 inches to 36 inches was carried out by Messrs. Steward and Longwell at the experimental station at the Boise Project of the United States Reclamation Service during 1911.2 The measurements were absolute, and after making slight corrections (not exceeding 1 per cent.) to compensate for the proximity of the bottom of the measuring channel to the sill of the weir, Table I (Appendix) was drawn up, using the coefficients given by these investigators as a basis, and this Table of coefficients has been used throughout the present experiments. 66 General Arrangement (Figs. 1).-A 49-foot gauging-weir on the River Alwen was raised temporarily, and the water diverted to flow through a channel 3 feet wide cut down to rock, and with a gradient of 1 in 125 to a controlling penstock, thence to the upper or meter" pool with a gradient of 1 in 25; after passing over the rectangular weir in this pool, the water flowed over the weir under investigation in the "experimental" pool, and was led back to the river by a channel blasted out in the rock to a 1 in 15 gradient. The backing up of the river formed a pool some 45 feet wide and about 100 feet long. This comparatively large area compensated in a very satisfactory manner any small variations in the stream-flow. After a few trials it was found possible to fix on the penstock a scale which enabled any desired rate of flow to be readily obtained. Pools (Figs. 2 and 3).—The pools were practically alike in size and shape. In plan the width at 12 feet upstream was 10 feet, and at 10 feet upstream the width had become equal to that measured at the weir, namely, 12 feet. Below the sill of the weir there was 3 feet of water, and this obtained to a line parallel to the weirs and 8 feet upstream, beyond which the bottom rose gradually to meet the discharge channel from the upper or the inlet channel, as the case might be. "Weir with Free Overfall and Constant Coefficient of Contraction," Minutes of Proceedings Inst. C.E., vol. lxxxv, p. 454. 2 W. G. Steward and J. S. Longwell, "Experiments on Weir Discharge," Trans. Am. Soc. C.E., vol. lxxvi, p. 1045 et seq. introduced in places where leakage might possibly have occurred. This was protected from scour by light boarding or pitching. Occasionally small slips of the sides of the intake channel took place, and at intervals the upper pool was cleared of any material which may have been carried in, so that its full depth was maintained. Considerable protection from wind was afforded by banks of spoil deposited round the pools. 66 Weirs. The sharp edges of the weirs were formed of 3-inch by -inch steel plates having bevelled edges which projected inch beyond the bevelled edge of the 3-inch planks to which the plates were screwed (Figs. 3). The metal sill of the rectangular or meter" weir 3 feet 6 inches long was attached to the middle of a 9-inch by 3-inch plank 12 feet long, the top edge of which was planed and set truly horizontal. Two pieces of 3-inch plank held together by stout battens carried the end plates of the weirs, and the bottom edge of the lower piece was planed so that the combination might slide on the 12-foot plank, previously referred to, thus allowing a weir of any desired length (not exceeding 3 feet) to be readily obtained. The joint between the fixed and sliding pieces was made by lightly caulking with red lead and oakum, and leakage behind the sliding pieces was prevented by a packing of clay. The arrangements for the triangular and trapezoidal or "experimental" weirs were of a similar nature. Care was taken during construction to ensure the plane of the weirs being vertical, and in setting either of the weirs to a definite opening every precaution was observed that the length should be exact to the nearest 4 inch, and that the inclination of the ends was exactly as desired. Measurement of Head.-In each pool at 9 feet back from the weir and 4 feet off the axial line of the pool, the lower portion of a pine ranging rod was driven into a hole bored in the rock, and a domeheaded screw was screwed into it, so that the top of the dome was exactly level with the sill of the weir. From time to time the top of the dome was checked against the sill with a surveyor's level, and occasionally the horizontality of the sill was also checked, though this latter operation was really unnecessary, as the 3-inch planking forming the end of the pool was carried down to and tied into the rock. The rod was surrounded by two 9-inch stoneware pipes set vertically one above the other; the lower one rested on bricks placed at the bottom of the pool, and was filled with broken stone for 12 inches up; the upper pipe was kept inch from contact with the lower one by packing pieces placed at the bottom of the socket, and the joint was made up with clay except for 1 inch. These stilling boxes answered quite well. There was no appreciable |