No abstract

Author proposes to give a short description of the construction, working, and efficiency of the tidal flour-mills in use in Andalusia, more especially in the province of Huelva, which are, he believes, little known in England, and which are interesting not only from their extreme simplicity and primitiveness, but also as being perhaps the earliest form of wheel of the turbine class, having come down in their present mode of construction almost unaltered from Moorish times, if not from a still earlier date.These mills are thickly studded all over the '' Marismas," or salt-marshes, which border the lower portions of the Andalusian rivers, within the tidal range. They are built across the " esteros," or creeks, by which these marshes are intersected, at the most suitable points, where the firmest foundation is obtainable, and a sufficient amount of storage available, the natural capacity of the " estero " being increased by means of rough excavations, so as to retain at spring-tides from 500,000 to 1,000,000 cubic feet of tidal water. The total head rarely exceeds 7 feet at springs, and the mill is usually worked until it falls below 3 feet. A spring-tide

Proceedings.] THE CITY OF BIRMINGHAM FROM MID-WALES. '73were used in place of bends, on all horizontal and vertical curves Mr. Bruce. of the 48-inch siphons, and were found to work satisfactorily. Mr. Bruce had once had to carry a 32-inch main, conveying water from Tchar Lake to Bombay, over the Kurla Creek, which was used by salt-barges.H e adopted a hydraulic arch, somewhat similar to that described in the Paper, of flanged steel pipes, springing from heavy castings bolted down to brick wells filled with concrete, and with an air-valve at the summit. Referring to the filteringmaterials, it would have been more economical to adopt the plan of covering the floor with bricks on edge, covered with others laid flat, and spreading upon the latter 3 inches of each of the two finer grades of gravel, which, while saving 18 inches in the depth of the beds, would have been equally efficient.The rate of filtration adopted seemed very slow considering the purity of the water.

Proceedings.]DISCUSSION ON IXPOUXDIXG RESERT-OIRS. CORRESPONDENCE ON IMPOUNDING RESERVOIRS. [Minutes ofDr. -4. VOR analysis of the problem had convinced Delocre that, in consequence liraunmiihl. of the varying relation of the pressures, continuous profiles, such, for instance, as logarithmic curves, were not possible; he therefore divided the cross-section into three portions. Professor Kreuter remarked truly that, in order for the line of pressure to change from its fixed position in portion No. l to its position in portion No. 2 of Delocre's section, an intermediate part should be introduced. This part was of trapezoidal form. The equation for the breadth in the third portion was given as an algebraic expression, that in the fourth (base) portion being determined by a logarithmic curve. The principal difficulty consisted, after the breadth in the entire cross-section of the dam were found, in SO joining together the four portions as to ensure the fulfilment of the necessary conditions of equilibrium. This was an easy matter in the first and second portions, but in the last two the equation contained an integral that could only be evaluated step by step. The effect of the vertical component of the waterpressure on the profiles, hitherto neglected, was next considered by a combined graphical and analytical method.Engineers might, perhaps, consider this theoretical treatment of the question too cumbrous for ordinary use. Two replies might be given to this : first, the work involved in the computation was less laborious than it might at first sight appear to be; and, secondly, the time spent in arriving at a solution of the question was of little importance in relation to the certainty introduced into the design of dams by rigorous mathematical treatment. nrr. A. Fairlie Mr. A. FAIRLIE BRUCE observed that the strength of the Tansa Rruce.dam might have been increased had it been constructed with an inner skin of 5 to 1 Portland cement concrete, say 5 feet thick at the bottom, diminishing to 1 foot at the top. That would not have added much to the cost, and would have prevented water from finding its way into the heart of the structure, which, while it might cause little or no leakage, must, by its pressure, impair the stability of the dam. It would have been interesting to know why rubble masonry was preferred to rubble concrete for the work. The crushing strength of the mortar compared unfavourably with that of Portland cement mortar. With cement concrete two weeks old, in which the mortar was composed of 1 part of cement to 3 parts of sand, he had, in arched ribs, obtained nearly as good results as those quoted in the Paper at six to eleven months ; and at twelve weeks he had obtained 2,447 lbs, per square inch, which exceeded the maximum Tansa test. Judging from the quantity ofwater delivered, and the small loss by leakage, the workmanDownloaded by [ UNIV OF TEXAS AT AUSTIN] on [11/09/16].

No abstract