from the attendants at other dams is to the effect that needles are only put in place or removed from navigable passes when the fall is reduced to 5 or 5 feet.

MOVABLE DAMS ON UPPER SEINE.

In the same volume of the Annals, MM. Chanoine and Lagrené give an elaborate description of the twelve dams constructed on the Seine between Montereau and Paris, with full details of the methods used in constructing the foundations, and of all items of cost. This will be given at sufficient length for the purposes of the present report from the later report of M. Cambuzat, but reference is made to it for the benefit of those who may desire to go more fully into this subject. The following items are, however, quoted here, as they are not referred to by M. Cambuzat.

Careful observations on the river, before and after the construction of the dams, showed that they produced no sensible effect on the heights of floods. When they are submerged a very slight ripple on the surface is all that can indicate their presence. The works themselves were not injured, especially if the lock-gates were opened before the arrival of the flood. The only effect was a slight deposit of sediment in the locks.

The time allowed for passing a fleet of eight boats through a lock is twenty minutes, half of which is used in filling or emptying the lock, and the other half is needed for getting the boats in and out, and for closing and opening the gates. The use of a submerged chain for towing has increased this time to twenty-eight minutes. A single tow-boat can be passed through in fifteen minutes. The larger the lock the better, when navigation is habitually carried on by fleets, as there is a very great gain in being able to pass a whole fleet through in one lockage. The average time necessary for raising such wickets as are used on the Upper Seine is two minutes and ten seconds for each wicket.

In dropping a number of wickets there is sometimes considerable difficulty in moving the last ones, on account of the pressure on the tripping rod of those first dropped. This is remedied by passing an iron band under the caps of the axle of the wicket, so that when the wicket is down it is raised a little higher above the tripping-rod, and there is less danger of the weight of water pressing it against the rod. As it is important that the lock-tender should have some index to show him, at all times, the number of the wicket on which the trippingrod is acting at the moment, several have been invented, three of which are described. They are geared so as to move either with the trippingrod or with the axle of the pinion that works it.

Should a wicket remain standing after the others have fallen, the locktender, after the current through the pass has somewhat lessened, approaches it from below in a skiff, and drops it by a side-blow on the

prop.

If it becomes necessary to swing the weir-wickets at a time when the overflow is nearly sufficient to swing them of itself, it is unsafe to use the service-boat, unless it is fastened to a line of piles above the weir, as any pressure against the wickets, even though the boat has long fenders, might swing them.

A full discussion is given of a modification in the manner of connecting the weir wickets with their horses, so that after these wickets swing, the level of the pool may be kept constant as the discharge increases. This change, however, seems not to have been adopted in practice, and therefore need not be considered here.

JOINVILLE DAM ON THE MARNE.

In the same volume of the Annals is an article by M. Malézieux, en gineer of ponts et chaussées, on the Joinville dam, from which the following is taken. Joinville is about 113 miles above the fortifications of Paris, measured by the river, and about 4 measured by the St. Maur Canal.

A dam had been in existence on the Marne at Joinville for forty years. It was constructed in order to raise the water to a minimum depth of 5 feet in the St. Maur Canal, and at the same time to create a fall, which caused the construction of important manufactories along the banks of this canal. The necessities of navigation having been greatly changed, the first thought was to raise the old dam 2 feet, which would be enough to give 5 feet of draught in the river between the upper end of the St. Maur Canal and the lower end of the Chelles Canal. But the city of Paris became, in 1864, the owner of the manufactories of St. Maur; and from that time the motive power was used to pump up the Marne water, in order to supply the Vincennes lakes and the elevated parts of the capital. The city requested that the Joinville dam should be raised 34 feet, and offered to share in the cost. This was about the highest elevation that could be permitted without serious inconvenience, if not on account of the submersion of the banks above, at least on account of the heights of the cargoes of the charcoal-boats which pass through the tunnel of the St. Maur Canal. The proposals of the city were accepted by the state.

It was found inadvisable to rebuild the old dam, and an entirely new one was built about 1,400 feet below the old dam, so as to leave a basin of a little more than 12 acres below the mouth of the St. Maur Canal. This canal is a cut-off by which boats pass through two-thirds of a mile of canal instead of eight miles of river. As navigation goes almost entirely through the canal, there is little need of a navigable pass, and accordingly it is only 39 feet wide. The weir is 207 feet in length. The wickets are 3 feet 7 inches high, and their tops are 8 feet 2 inches above the bottom of the St. Maur Canal. The lift of the dam in low-water is 7 feet.

Experience on the upper dams of the Marne having shown that the Chanoine wickets were liable to many accidents, and the Marne not being a river subject to sudden heavy floods, and there being, on account of the special peculiarities of the situation, but seldom occasion for opening the pass of this dam, it was thought best to close the latter by a needle-dam, and to use the Desfontaines wickets on the weir.

The Desfontaines system has been in operation at the Damery dam since 1857, and at the Courcelles dam since 1861, but is still not well known, and M. Malézieux gives a description of it, from which we only extract such matter as has not been mentioned already.

The weir is contained between two rows of sheet-piling, 26 feet apart. The up-stream capping-pieces are even with the permanent sill, at 3 feet 8 inches below the level of the pool; the down-stream caps are at the level of low-water. Before the piles were driven, the area to be covered with beton was dredged. The least thickness of beton below the level of the low-water below the dam was taken at 4 feet, but this depth was occasionally exceeded.

When

The first thing is to construct the main body of masonry, which contains a cavity 7 feet wide, with a cross-section composed of a quadrant adjoining a rectangle, and extending from one end of the weir to the other. The two edges of the opening have cut-stone borders. this is finished it is divided into sections, or drums, each 5 feet long, separated by large transverse diaphragms of cast iron. These plates enter the masonry to a depth of 3 inches, and each one has two openings cut through it; one is on the up-stream side near the top of the quadrant, and is wider than it is high, making a horizontal opening

the other is on the down-stream side, is higher than it is wide, and forms a vertical opening. The latter is made of such a size that a man can without much difficulty pass through it from one end of the weir to the o her.

Each drum now received a large plate-iron wicket, which has near the middle a horizontal hinge. This hinge is a cast-iron tube, which contains a wrought-iron spindle from one end to the other. Its extremities rest on the two diaphragms, at the level of the upper border but nearer the lower. The wicket itself is formed of three arms, each 7 feet 11 inches long, and of two pieces of sheet-iron two-tenths of an inch in thickness.

The arms pass through the hinge, and inside of it are shaped as collars, in which the axis turns; they completely fill the annular space between the latter and the inside surface of the hinge.

The counter-wicket, starting from the hinge, is bent backward 1 foot 4 inches, and then becomes parallel to the wicket. The object of this is to prevent the counter-wicket from extending above the horizontal opening when the wicket is down. When the wicket is up the down-stream side of the counter-wicket coincides with the up-stream side of the vertical opening. It is prevented from passing this position by a wooden sill, against which its lower edge rests, by two moldings on the diaphragms that correspond in shape to the sides of the counter-wickets, and by a molding on the hinge which comes in contact with its upper edge. There is a play not exceeding one-seventh of an inch between the counterwicket and the walls of its drum. To secure such accuracy of fitting the iron-work is planed, and the stone-work of the cylinder is covered with Portland cement, which is accurately shaped by a cutting-tool turning on the axis of the wicket. When the wicket is up the counterwicket is in contact on all sides with the projections, and as its surfaces of contact are faced with India rubber, there is no leakage. When the counter-wicket is in any other position there is a slight waste of water. Each drum is divided by the counter-wicket into two compartments, and the top of each is covered by a sheet of plate-iron, which is fastened to the sole, to the diaphragms, and to the hinge; this combination supports the hinge itself.

The abutment, which is 9 feet 10 inches thick and 25 feet long, is pierced through the middle by a longitudinal culvert whose sections vary as follows: At each extremity are two parts, each 4 feet 3 inches in length and 3 feet 3 inches wide by 4 feet 3 inches high, into which a man can easily enter; the up-stream culvert has its sole 1 foot 3 inches above the low-water line below the dam, while the down-stream one has its sole exactly at this level. In the middle, with a length of 9 feet 10 inches, and separated from the end parts by two chimneys or wells, which establish a communication with the top of the abutment, is the essential part of the arrangement, that in which the distribution of water is effected and on which depends the maneuvering of the weir. In this central part the culvert is divided into two rectangular conduits, one above the other, and only separated by a five-eighths plate of cast iron; these conduits are each 2 feet 7 inches wide and 2 feet high. An ordinary cast-iron valve placed at the upper end shuts one of these openings while unmasking the other; there is a similar valve at the lower end. A balance-beam, to which are attached the rods of both valves, communicates to the down-stream valve the movement caused in the upstream one, or, more accurately, the opposite movement.

In the body of masonry which separates these conduits from the outer wall of the abutment, and consequently from the first drum of the weir, two other conduits are constructed, but no longer superposed. One of them, which is placed on the up-stream side and at the higher elevation, prolongs the horizontal opening of the diaphragms until a connection is made with the upper conduit in the abutment; the other prolongs the vertical opening to a connection with the lower conduit. (See the plates.)

The method of raising and lowering the wickets has already been explained.

It is essential to the successful use of drum-wickets that there should be a fall at the dam, and this is obtained by raising the permanent part of the weir. The obstacles to raising are the current and a deficiency of fall, both of which are least in low-water and greatest in high. The friction of the arms on the axis must also be overcome. It is not likely that the axes of the wickets will get untrue, as they are short and are held by three pairs of collars. The resistance of the water in the upper compartments when the wickets are lowered prevents them from coming down with a shock.

At Joinville, when the river below the dam is at low-water mark, (zero of the water-gauges,) a fall of 4 to 6 inches is sufficient to raise the wickets. If the pass is completely shut, which forces the water to flow over the permanent part of the weir, it takes about two minutes to raise the wickets and one minute to lower them. On the 5th of March, 1868, the water-surface standing at 7 feet above the dam and at 4 feet 7 inches below it, twelve minutes were required to raise the wickets, and five to lower them. On another occasion, when the water above stood at 6 feet 7 inches, and at 5 feet below, a half hour was required to raise the wickets. This is about the limit at which the Joinville wickets will work. To extend this limit would be easy, but there would be no practical advantage in so doing, and it would be a useless increase of cost, for when the water-surface stands at a height of 4 feet 7 inches below the dam, the wickets have no service to render to navigation, and the necessities that are ordinarily served by the Saint Maur pumps are more than satisfied.

It was naturally to be expected that there might be trouble at Joinville on account of the sediment and gravel that would be deposited in the drums. Experience, however, has shown that the strong currents that can be sent through the upper and lower compartments by suitably working the valves in the pier and in the abutment are sufficient to remove such deposits entirely. Care is taken to have gratings at the heads of the culverts, in order to keep out leaves, sticks, &c., and there are special valves that can be used to cut off all communication between the culverts and the upper pool during the winter season when the water is muddiest.

One great merit of Desfontaines' wickets is that they are more nearly water-tight than any others. The interval of four-tenths of an inch between the wickets can easily be reduced to two-tenths or less.

If a boat should be driven by the current against one of the wickets, and should injure it, it can at once be replaced from the spare wickets which are always in store. Such an accident could hardly happen except when the river was tolerably high, and as the upper compartments are always in communication with the upper pool when the wickets are up, in most cases the only result will be that the counter-wicket will drive back the water that presses against it, and the wicket will gradually yield to the shock and lie down.

The drums can be laid dry for repairs whenever the water below the dam does not stand higher than 3 feet on the gauge, (which is the case during half the year,) by building little dams in the piers and in the abutment, for which recesses are left. This, however, necessitates that 3 feet 7 inches be drawn off from the pool. At Joinville this plan was considered objectionable, and instead of it a line of triangular frames, 3 feet 7 inches high and 9 feet 10 inches apart, was placed on the weir above the wickets. These frames lie down like the trestles of a Poirée dam. They are used as the supports of a temporary dam of timbers 81⁄2 inches by 4 in section, which have claws that connect them with the uprights of the frames. They are easily placed from a skiff, and the dam is made water-tight by a covering of tarred canvas, held down by weights. After the top of the weir is exposed, a little traveling-crane, running on rails and movable from one end of the weir to the other, is

used for handling the wickets. In less than a day the covers can be taken off, the wickets can be raised and examined, and all can be replaced. No pump is used unless it is thought advisable to empty the drums.

Special precautions were taken to prevent wash below the weir of the Joinville dam, as the use of a notched bar to permit the half lowering of wickets was, for simplifiation, rejected. The greatest difficulty was encountered below the pass. The following is a description of this danger, and of the means taken to meet it:

The lower end of the sole of the pass is exposed to still more dangerous corrosions, and it may chance that the whole Marne will pass through this narrow opening of only 39 feet. In fact a scour, to the depth of 13 feet below low-water, was very soon formed at a distance of 100 feet below the pass. In order to keep the ultimate scour at a suitable distance from the dam, bearing in mind the difficulty of making the necessary repairs to a covering 2 feet 4 inches below low-water, the rock-work for a distance of 50 feet was covered, not by a masonry pavement, but by a strong flooring of oak timbers, bolted carefully to the caps that connected the six rows of piles.

The trestles of the pass are 9 feet 10 inches high and the one nearest to the pier is 6 feet 7 inches distant from it, so that a space has to be made for it in the pier, and this recess is separated from the culvert by a special casting. The service-bridge is 6 feet 7 inches below the top of the pier, and the latter is reached by a short ladder and four steps in the pier itself. A supplementary bridge is provided, whose floor is 4 feet higher than that of the main bridge, but it is seldom used, as the latter answers for most purposes.

The iron-work of the wickets cost 49,844 francs, or $9,470.

The counter-wickets of the Marne weirs are generally of the same length as the wickets, but at the Courcelles Dam the former are 43 inches longer, and these wickets are in consequence perceptibly more readily maneuvered than any others. The longer the counter-wickets the greater the height of the river at which the raising of the wickets becomes impracticable.

To satisfy the local wants of the Marne navigation, M. Desfontaines adopted the following rules:

1. To make the height of the permanent part of the weir equal to half the difference of level in low-water. 2. To limit the sinking of the drums below the low-water line, at the foot of the dam, to from 12 to 16 inches. But these rules were neither general nor absolute.

There are some advantages in a high weir, provided it is not too costly on account of its length; such as greater security than a movable curtain, greater ease of establishing, examining, and repairing its movable parts, and less exposure of the latter to being covered by saud or sediment.

Besides this, the height secures the initial fall, which is necessary for the maneuvering of the wickets.

M. Malézieux believes that there is no practical difficulty in the way of making deeper drums and higher wickets on the Desfontaines system.

IMPROVEMENTS ON POIRÉE NEEDLE-DAMS.

In Vol. XX, Annales des ponts et chaussées, 1870, M. Saint Yves takes up the subject of the improvements practicable on Poirée needle-dams. The substance of his remarks is as follows:

One of the troubles with a needle-dam, especially when high, is its lack of tightness. After a careful investigation, which is given in full in the original, he recommended that the cross-section of the needles should be a semi-regular hexagon, obtained by taking an equilateral tri