Above the gates shutters are placed which can temporarily close the slui shutters have their axis of rotation fixed on the floor, and they fall up strea they are raised they are held nearly vertical by two retaining-chains faste floor; when they are laid down they are held in a horizontal position by spri which are fastened on their upper surfaces, and catch of themselves on which is fixed transversely on the floor. A slight rotation of this bar is disengage the latches, and thus free the shutters.

These shutters will at once be recognized as the counter-shutters of system, and their assistance is necessary in order to start the American ga proceed to explain.

In order to shut the sluice, the attendant commences by closing the valves of the culverts in the piers; he then opens the up-stream valves a the latches of the shutters; the latter are immediately brought up b they close the sluice and stop the flow of water through it; the low and the upper pool rises. This difference of level causes a pressure u which finally raises them until they meet the wooden heurters placed in the piers.

The water, still rising in the upper pool, soon flows over the shut space between them and the gates; the former being thus relieved that kept them up yield, partly to the effect of their own weight, pull of the retaining-chains, and fall down on the floor, where their la under the stop-bar.

The maneuver of opening is still simpler. It is limited to clos valves in the piers and opening the down-stream ones. The wate loses the pressure of the upper pool, and a portion of it flows into t the gates lie down in their chamber.

Experience shows that the gates begin to rise when the fall fre lower end of the sluice is at least two feet. The time used in closin therefore, upon the time needed by the river to generate this fa opening only lasts three minutes, and the gates immediately sink The American system, which has just been described, gives remarks:

In the first place, the gates are not balanced around their axes fontaines' wicket, nor are they received by a water-tight cavity, sion, they retain considerable weight, which produces a decided

The calculations which the author gives show th of the upper gate is 20,840 pounds, and that of t gate is 49,756 pounds. The width of the upper gat of the lower gate, measured from its axis of rotatio friction-rollers, at its upper end, is 20 feet. Wher the center of the friction-roller of the lower gate is the axis of rotation of the upper gate. When t pressure that must be applied to the under surfac order to start them is 709 pounds per running 1 square foot.

Thus the mere weight of the apparatus requires a differen in order to put the gates in motion.

In this system there are other resistances to raising. P gate turn around a wooden axle 31 feet long, which is he interior diameter. This large diameter is the cause of over, between the collars each axle turns in a hollow we find here a double difficulty; in fact, if we leave m the hollow quoin, imitating the practice with lock-gat age, and the loss of pressure of the sustaining water m other hand, we reduce the play to fractions of an ine sediment will cause friction that cannot always be ove Each axle, although perfectly true when the gates a a little while, and by its bending may produce abnor lars.

The friction of the axles is not the only thing that seven rollers at the upper end of the counter-gate. The lever-arm of the lifting force, exercised by t very feeble at the beginning of the movement.

Finally, the leakage under each of the four edge gate, must be considerable, and there must be water that put the apparatus in motion.

h needles having sections ons have shown that, owing difficulties in placing them te, it has been decided to The square. He suggests for ternate with square needles. altogether satisfactory; and per sides are slightly rounded. r, with the view of overlapping tions of T-needles and of those ce, but this will not be objectiont the sill. He also suggests the for which he gives the necessary ve to be handled by mechanical in the pass of the Melun Dam, in 1 of 7 feet 2 inches, and there was Sections of these forms are given

e supporting-bars are necessary in since their adoption on the Lower des has almost entirely ceased. ht be advantageous to have trestles

further up stream than that of the counter-wicket, and at a slightly higher level, so that when the wicket is down it covers the props, and its end rests on the frictionrollers.

Each compartment of the circular conduit communicates with the upper conduit by a hole in the dividing-wall, which always remains open.

Each compartment of the circular conduit is, at will, put in communication with the lower conduit by a hole in the dividing-wall, and this hole is opened or shut by means of a valve, according to the needs of the service.

We can now explain the working of the apparatus. When the counter-wicket turns around its axis of rotation under the influence of the hydraulic pressure, the props take the wicket in reverse, and make it describe an angle of 70°. In its last position, and throughout its movement, the wicket rests upon the rollers of the props.

If the new weir had no other arrangements than those just described, its maneuvering would not differ from that of the Desfontaines weir. The combination by which the independence of movement of the wickets can be assured is yet to be described.

Let us suppose the dam down, and let us imagine that each one of the holes connecting with the lower conduit is closed by a water-tight valve; whatever may be the pressure brought to bear upon the up-stream faces of the counter-wickets, none of them will budge, for there is a play of one-seventh of an inch around the perimeter of each, and consequently there is an equality of pressure on the two faces of each counter-wicket. As soon as one valve is opened the counter-wicket of the compartment, thus relieved from pressure on its down-stream side, will commence moving, and will set up the wicket which it controls. The valve once closed, the equilibrium of the pressure will be immediately re-established on the two faces of the counter-wicket, and the wicket, pressing on the rollers with the pressure due to the lift of the dam, will carry back the counter-wicket to its first position as it falls down itself. In order to permit this movement of lowering, it is necessary, even when the wicket is up, that there should be a water communication between the up and down stream faces of the counter-wicket. Thus, contrary to what happens in the Desfontaines system, the counter-wicket, when the wicket is up, does not make a water-tight connection with its seat, but is supported on a certain number of points, so as to permit leakage around its perimeter.

Maneuvering will then be reduced to that of the valves, of which I have just spoken, and can easily be controlled by means of a rod with projections, which is placed in the lower conduit. By suitably arranging the projections of this rod, all the wickets can be lowered in succession, in whatever order may have been determined in ad

vance.

The raising will be done in inverse order.

The axis of rotation of each valve is provided with a projecting finger, (Fig. 66,) whose direction is parallel to that of the valve. This finger, when pushed by the projection, remains horizontal while the projection rests upon it.

When the projection is drawn back by a movement of the rod, the valve, which is now free, is swung down by the counter-weight, which is fastened to it for this object. It has been foreseen that a case might occur in which this counter-weight would be insufficient to overcome the adhesion of the India-rubber facing and to cause rotation.

A counter-finger, likewise fastened to the axis of the valve, and at right angles to the finger, receives the push of the tripping-rod through a stud, which forms a kind of counter-projection. When the movement has once begun, the counter-weight will always suffice to swing open the valve, and the counter-finger will not be in the way of the advance of the bar while the movement of lowering the wickets is in progress. This brief description already shows that, by placing the upper and lower conduits outside of the druins, it is much easier to increase their cross-section; this is an important consideration in order to avoid loss of pressure.

It can also be seen that the relative positions of the wicket and the counter-wicket lead, as a first consequence, to the possibility of a reduction of the length of the latter. In fact, for a dam constructed under conditions similar to that at Joinville, a lift of 6 feet 7 inches compels the adoption of a counter-wicket descending to 8 feet 2 inches below the permanent crest. In the system described in this chapter, the counter-wicket, describing an angle nearly double that described by the wicket, can furnish the same quantity of effective work with a shorter length, and the depth of the excavation in the masonry is thus diminished about 2 feet.

The weir is contained between two piers or abutments, which are penetrated by the two rectangular conduits; they end in aqueducts, which pass through the entire length of each bounding wall, and are perpendicular to the direction of the dam. These aqueducts have three valves: The first, commencing up stream, is between the opening for taking water and the end of the raising-tube; it is seldom closed, except during floods of muddy water, and during the cleansing operations, which will be described next. The second valve is between the two conduits, and is generally closed. Lastly, the

third valve is placed below the emptying-conduit, and is always open, except in rare

cases.

As a rule, the up-stream conduit is left in constant communication with the upper pool at both ends, and the lower conduit in communication with the lower pool. The only work to be done to regulate the level of the pool consists, as I have said, in shifting the tripping-rod.

If the upper rectangular conduit should receive sedimentary deposits, it is only necessary to close the upper valve of one of the piers and to open the middle valve of the same pier. A very rapid current will be formed in the conduit, and will clean it out. A similar maneuver will assure the cleanliness of the lower conduit.

However, deposits, if they should occur, need not be dreaded, since they will be localized in the two conduits at a distance from the drums, and they could have no injurious effect upon the march of the counter-wickets, except by slightly diminishing the cross-section of supply or discharge.

The establishment, above the wickets, of a line of trestles, carrying a foot-bridge, is demanded by general reasons which apply to all wicket-dams; in fact, it is important to give the lock-tender an easy circulation from one shore to the other. In addition, the line of trestles permits the construction of a temporary dam for repairing, either with the view of momentarily replacing the wicket-dam by one of needles, or of moderating the current by means of a screen of needles placed in front of lowered wickets, or of regulating the inclination of the wickets, as I have explained in speaking of the Joinville Dam. In conclusion, although a foot-bridge above is not indispensable for the Desfontaines or the Cuvinot Dam, I think that there ought to be no hesitation about establishing one.

The author then calculates the amount of the forces that act on the different parts of the apparatus, selecting the wicket nearest the pier, as in this system this should be the last wicket raised. Even in this case there would be no difficulty in raising the wicket. There may be some trouble with the friction-rollers, but the proper test of the system is to try it, which, apparently, has not yet been done.

He suggests that the masonry in the weir might be greatly reduced by omitting the upper and lower conduits, and opening direct communication between the drums and the upper and lower pools. The objection to this is that the drums might be filled with sediment and vegetable matter brought down by the current.

POIRÉE NEEDLES.

M. De Lagrené states that experiments with needles having sections. of the form of regular or semi-regular hexagons have shown that, owing to the bending of the needles and to practical difficulties in placing them in the positions which theory would indicate, it has been decided to give up these sections and to return to the square. He suggests for trial needles shaped like a T, which are to alternate with square needles. This solution, however, does not appear altogether satisfactory; and he also proposes for trial needles whose upper sides are slightly rounded aud are provided with strips of India-rubber, with the view of overlapping adjacent square needles. The upper portions of T-needles and of those with India-rubber flaps ought to be square, but this will not be objectionable, as the greatest leakage is nearest the sill. He also suggests the use of hollow needles made of planks, for which he gives the necessary calculations. Such needles would have to be handled by mechanical means. Two such needles were tried in the pass of the Melun Dam, in September, 1872, when there was a fall of 7 feet 2 inches, and there was no difficulty in putting them in place. Sections of these forms are given on Plate 10, Fig. 62.

The author thinks that intermediate supporting-bars are necessary in high needle-dams, and states that since their adoption on the Lower Seine, in 1868, the breaking of needles has almost entirely ceased.

M. Cadot suggested that it might be advantageous to have trestles

with two stories of needles. He proposed to fasten pins to the upper legs of the trestles, a little below the tops of the lower needles, and to place on them the sill for the second story of needles. This sill should be of T-iron, hooked on to these pins, and should be put in place after completion of the lower story of needles. This plan would make two foot-bridges necessary at different heights, and it might happen that the removal of the upper story of needles would not sufficiently lower the water to expose the lower foot-bridge. That, however, might be remedied by limiting the two-story arrangement to a part of the dam, and making the rest of long needles. If the removal of the latter did not suffice to uncover the lower bridge, the needles of the lower story would have to be removed by a boat. This plan, although not yet in use, seems to promise good results.

SHUTTERS WITH PONTONS.-KRANTZ SYSTEM.

The plan of dam proposed by Chief Engineer Krantz in 1868, and now in course of construction on the Lower Seine, at Port Villez, suggests new combinations, which I will explain from the description of the inventor which accompanied his plan.

M. Krantz first decided on the principle that every movable dam should satisfy the following conditions:

1. It should be maneuvered by the aid of the natural forces of the water-course, properly brought in play, and without exposing the attendants to any risk.

2. The whole apparatus should at all times be subject to human control.

3. It should, spontaneously, correct the slight changes of level in the pools, and should rarely require the intervention of man.

4. It should only be composed of strong parts, capable of resisting violent shocks.

5. It should only require for its establishment constructions similar to those which are habitually built on our rivers.

6. It should be sufficiently tight.

7. It should be applicable to lifts greater than those of the present dams.

Then comes the description, which I copy verbatim, allowing the inventor to speak for himself:

"The dam which we are about to describe is planned to sustain a lift of 9 feet 10 inches.

"When lowered, it should not project above the horizontal plane drawn 2 feet 7 inches below low-water.

"We could easily have chosen a different relief, as the system is well adapted to it. But, on the one hand, the depth of 2 feet 7 inches is that of the navigable passes constructed on the Lower Seine, particularly at Suresnes; and, on the other hand, a lift of 9 feet 10 inches is that of our highest needle-dams. By choosing these dimensions we render the comparisons which we intend making more simple and more conclusive. "The essential parts of the dam are: (Plate 12, Figures 67, 68, 69, 70.)

"1. The lockette, (éclusette, or little lock,) by which the water which serves to work the apparatus is distributed under the proper pressure.

2. The dam proper, which includes the pouton with the upper wicket, and its valves, and the water-conduit.

"The lockettes are placed at the ends of the different sections of the dam, each of which may have as great a length as 300 feet.

"Their number, and the spaces between them, depend upon local circumstances, with which we have no concern at present.

"As to the dam, whatever be its importance, it is divided into elements 9 feet 10 inches long, which act simultaneously under the action of the same forces, while at the same time preserving a mutual independence.

"The lockettes are of metal, and rest upon a masonry base. They are hollow, and communicate with the upper and lower pools.

"Two sets of valves, with vertical axes, placed near the ends, permit at will the interruption or the establishment of communication between the central part of each lockette and the pools.

"Whence it follows that the lockette is, in fact, what its name indicates-a kind of lock, of small size, in the chamber of which one can, at will, by a suitable movement of the valves, maintain either the level of the upper pool, or that of the lower pool, or an entirely different intermediate level.

"The side wall of the lockette which adjoins the dam is pierced at the height of the conduit by a rectangular opening, through which flows the water destined to raise he apparatus, or by which it escapes when the dam is to be lowered.